Geology and Mineral Resources of Estonia: MAIN PAGE



The territory of Estonia has been inhabited at least ten thousand years. The first tribes to settle in the area were hunters and fishers who could already use local raw materials, such as crystalline erratic boulders, gravel, sand and clay. At about 6000 BP inhabitants learned to make earthenware from clay and around 5000–4000 BP to apply carbonate rocks to building of town lets and fortified settlements. Since 1230, lime has been widely used as a binder. Red bricks, made of local clays and used as building material for strongholds and churches, provide Estonia’s historical buildings and architectural monuments with a specific geological splendour.

The first geological studies were carried out in Estonia more than 150 years ago. The long tradition of geological research in the area is due to the large and representative bedrock exposures providing excellent conditions for the study of Lower Palaeozoic rocks, and making Estonia a key region for solving several principal stratigraphic problems. The Palaeozoic rocks in Estonia enclose extraordinarily rich communities of well-preserved fossils, and a great number of new species and higher taxa have been established here. Ancient coastal formations of the Baltic Sea and relief forms, left behind by the last glaciation, are represented here more completely than in other regions. Within Estonia are found excellent examples of meteorite craters and the largest erratic boulders in northern Europe. All this makes Estonia’s geology unique in several aspects.

For centuries Estonia has served as an economic, scientific and cultural bridge between the East and the West. Already in the Middle Ages it was an arena of serious ideological conflicts. At the end of the Livonian War (1558–1583), southern Estonia fell under Polish rule. In the interest of restoring Catolicism, Jesuits opened a gymnasium in Tartu in 1583. After Estonia was taken over by Sweden, the Swedes founded a Protestant gymnasium in Tartu in 1630 to counterbalance the Jesuit school. In 1632, the Protestant gymnasium was changed into a university, which is one of the oldest and most prominent higher educational establishments in northern Europe. In the 17th century, Tartu University became a principal centre of education, science and humanistic ideas in the region. After re-opening in 1802, it developed into an outstanding centre of geological education and science in the former Russian Empire. The corresponding topics were also advanced in the Tartu (later Estonian) Society of Naturalists, founded in 1853.

In 1920, Estonian became the medium of instruction at Tartu University. In 1937, the Geological Committee of Estonia was founded. After the occupation and incorporation of Estonia into the Soviet Union, the most prominent geologists and a lot of promising young scientists left homeland and a new generation of geologists was trained. In 1947, the Institute of Geology of the Estonian Academy of Sciences was established and ten years later the Geological Survey of Estonia was founded. Both these institutions developed into important centres for geological, geophysical and environmental research in the northwestern portion of the Soviet Union and neighbouring countries.

The essential results of the research carried out during more than two centuries were summarised in multi-volume issues showing the directions and level of geological studies in Estonia (Geological Studies of the USSR, 50, Estonian SSR, Tallinn, 1968, 1972, 1973, 1974, 1977, 1984, 1987; History of Geological Sciences in Estonia, 1986). These, like most monographs in the field of geology issued in Estonia during the last decades, are in the Russian language and practically unknown to our western colleagues. Due to intensive drilling programmes and medium- and large-scale geological mapping, a lot of new geological information has been obtained. As there are currently no published general surveys on the geology and mineral resources of Estonia, the present monograph attempts to fill this gap. Its main purpose is not only to impart scientific information about Estonia’s natural environment, but to serve also for industrial and agricultural purposes encouraging the sound use of mineral resources in the present-day Estonia.

Mining of mineral resources has inflicted incurable wounds on Estonia’s nature. Another task of the present issue is to assist in drawing up main outlines of the strategy addressing improvement of the environment.

Most distinguished specialists of the Republic have participated in the compilation of this monograph. Its publishing has been made possible by the financial help of the Estonian Science Foundation (grant No. 1661), which is gratefully acknowledged. Thanks are due to the authors and all persons who have contributed to finalizing of this book. Special thanks go to Mrs. Helle Kukk for the revision of the English text, to Mr. Jüri Nemliher for the layout of this book and to Mr. Paul Pärkma for the drawings.


Anto Raukas and Aada Teedumäe



A. Raukas


The Republic of Estonia, the northernmost of the three Baltic States, is situated in the North-East of Europe, on the east coast of the Baltic Sea. The name Estonia is probably derived from Aists, the name the ancient Germans used to denote the Baltic tribes, living to the northeast of the Vistula River. In a written record the Aists (Aesti, Aestorium gentes) were first mentioned by the Roman historian Tacitus in the first century AD. The first written reference to the land of Estonians dates from 1154. On the order of Roger II, the king of Sicily, the Arab geographer and traveller Abu Abdallah Muhammad al-Idrisi designed a map of places in the world known in those times including Qalewany (Tallinn) in Astlanda (Estonia).

The territory of Estonia in nowadays boundaries extends from 57º30’34" to 59º49’12"N and from 21º45’49" to 28º12’44"E (Fig. 1). The northernmost point of Estonia is on the Island of Vaindloo (the Cape of Purekkari on the mainland), the easternmost point in the Town of Narva, the southernmost point is the Naha farmstead at Mõniste, and the westernmost point is on the Island of Nootamaa (the Cape of Ramsi on the mainland). The extreme length of the Estonian territory is 350 km from west to east, and 240 km from north to south. The length of the Estonian coastline is 3,780 km; of this 1,242 km are on the mainland and 2,540 km are divided among the islands.

Estonia has an area of 45,215.4 sq km of which 9.2% is taken up by islands and 4.6% is under inland bodies of water. Climatically, Estonia belongs to the mixed-forest subregion of the Atlantic continental region of the temperate zone, which is characterized by warm summers and moderately mild winters.

Geologically, Estonia is situated in the northwestern part of the East-European Platform. Structurally, it lies for the most part within the boundaries of the southern slope of the Fennoscandian Shield with only its extreme southwestern and southern parts forming the wings of the Baltic Syneclise and the Valmiera-Lokno Uplift, respectively.

As part of the vast East-European Plain, Estonia is a generally flat country (Photo 1), where uplands and plateau-like areas alternate with lowlands, depressions and large valley-like forms. The average height above sea level is approximately 50 m, relative heights of landforms do not as a rule exceed 20 m, being only seldom 50 and more metres. About 40 per cent of Estonia’s territory is at an absolute height of 50 to 100 m, and only one tenth has an elevation over 100 m above sea level (Fig. 1). The highest point in Estonia, the Suur Munamägi Hill (nearly 318 m), is located in the Haanja Heights.

Estonia displays a large variety of landscapes (Fig. 2). The northern part of the country consists of an extensive limestone plateau (Fig. 2), the northern edge of which forms a steep escarpment (Photo 2), known as the North-Estonian Klint (relative height up to 56 m). The narrow Fore-Klint Coastal Plain is situated in front of the Klint. The highest areas in the northern part of Estonia are the Pandivere Upland (166 m a.s.l.) and the Jõhvi Upland (81 m a.s.l.). To the south of the Pandivere Upland lies the gently sloping Vooremaa watershed (the Saadjärv Drumlin Field, with elevations up to 144 m a.s.l.).

Relatively high areas of North Estonia border on the Kõrvemaa and Alutaguse lowlands. To the south-west of the Pandivere Upland lies the Central-Estonian Plain which, gently sloping, passes over into the Võrtsjärv Depression. The Alutaguse Lowland turns into the Peipsi Depression.

In western Estonia the absolute height seldom exceeds 20 m and large areas are entirely flat. This is the region of the lowlands of West Estonia and West-Estonian (Moonsund) Archipelago. Some small elevations are Kõpu (63 m a.s.l.), Middle-Saaremaa (54 m), Sõrve (36.6 m) and Tõstamaa–Varbla (44 m); the scarps of the islands of Saaremaa and Muhu and those in the western part of the mainland form the West-Estonian Klint (up to 21 m a.s.l.).

In South Estonia the topography is more varied and differences in the altitude are greater than elsewhere in Estonia. The area has four topographic highs (Fig. 2): Sakala (up to 146 m a.s.l.), Otepää (217 m), Karula (137 m) and Haanja (318 m). They are separated from one another by the Valga and Hargla depressions and Võru Valley. The South-Estonian medium-height terrain (Ugandi Plateau, 40–100 m a.s.l.) is occupied by the South-West Estonian Plain.

The largest relief forms — plateaus, uplands, lowlands, depressions, the North-Estonian and West-Estonian escarpments (Aaloe & Miidel 1967) were formed in Pre-Quaternary times as a result of the long-term continental erosion (Tavast & Raukas 1982). Monoclinal bedding of bedrock strata and their different resistance to erosion resulted in the questa-like ancient topography (Orviku 1955). During all ice ages glacial erosion prevailed in North and West Estonia. These areas are characterized by a thin Quaternary cover and wide distribution of alvars against the background of Estonia’s generally flat topography. The erosional relief forms here are represented by both small (glacial scratches, etc.) and large (rock drumlins, hollows and troughs of glacial ploughing) ones.

In the transitional zone between the prevailing glacial erosion and accumulation areas in central Estonia, the most characteristic relief forms are of the erosional-accumulative type. Among them are drumlins (Photo 3) and drumlin-like ridges, including megaflutings, which may reach 13 km in length, and 80 m in height (Saadjärv Drumlin Field). The accumulation area in southern Estonia features gently sloping and undulating till plains, and morainic hills, with the latter being especially common on accumulative insular heights (Otepää, Haanja). In places, dump and push moraines stretch some tens of kilometres in length with relative heights up to 50 m (West-Saaremaa Elevation).

Glaciofluvial accumulative relief forms are widely distributed in Estonia, with classic eskers and kame fields formed, as a rule, in passive or dead ice (Karukäpp & Raukas 1976). Radial eskers are most common on the Pandivere Upland (Photo 4) and marginal eskers on the West-Estonian Lowland (Raukas et al. 1971). Fluvio- and limnoglacial kames either form separate fields or are scattered in hilly topography. As for the genesis, the glaciofluvial gravel and sandy plains are for the most part glaciofluvial deltas or outwash deltas. Less frequent are kame and glaciofluvial terraces and outwash cones.

Genetically and morphologically, the valleys of glacial meltwater discharge are diverse. These relief forms are most typical of southern Estonia where in places they form orthogonal valley systems. They include both radial and marginal valleys, some of which are formed under the ice (e.g. rills of discharge), while others came into being by epigenetical superimposing on subglacial topography under the conditions of jointed passive or dead ice, but also due to glacial breaks and intense joining of ice-dammed lakes. Glacial meltwaters often flowed along ancient valleys which had developed before the last glaciation (Tavast & Raukas 1982).

Characteristic of glacial terrain are also funnel- and saucer-shaped closed depressions — kettle holes, formation of which is associated with the melting of buried dead ice blocks (glaciokarst). Undoubtedly, in many cases, the process came to an end in the Late-glacial or at the beginning of the Holocene. However, it seems that some of the kettle-holes formed considerably later, with the process having started in the Boreal and coming to an end only in the Atlantic climatic period (Raukas & Rõuk 1995). Quite often kettle-holes are filled with peat, the thickness of which may reach 17 m.

In all stages of deglaciation considerable areas in front of glacier margins were occupied by glaciolacustrine basins of different ages (Raukas 1992a). These bodies of water have left behind deposits (mainly varved clays) and coastal relief forms (abrasional scarps, beach ridges, etc.) which are traceable at different levels, such as those on the slopes of the Otepää and Haanja heights.

The extensive glaciolacustrine plains, which were of great landscape-forming significance, occur only in the lower parts of the territory, particularly on the Alutaguse, Võrtsjärv, Peipsi and Kõrvemaa lowlands (Raukas et al. 1971). Due to the wide distribution of clayey deposits on the surfaces, modest absolute height and unfavourable drainage conditions these plains have undergone paludification, and many of them have turned into bog plains.

The effects of a variety of coastal processes can be found along both modern (Orviku 1974) and ancient coasts of the Baltic Sea (Raukas et al. 1965) and large lakes (Raukas & Tavast 1989). These features include wave-cut notches, scarps, abrasional platforms and plains, accumulative terraces, spits, barrier beaches, arrow-shaped spits, tombolos, bars, beach ridges, etc.). The development of the largest ancient coastal formations is connected to the transgressive phases of the Baltic Sea development (Raukas 1966).

More prominent aeolian relief forms (ridge-like longitudinal dunes, parabolic dunes, etc.) are spread along the ancient transgressive coastlines of the Baltic Sea. The height of dunes seldom exceeds 20 m. Formation of higher dunes was hindered by the small supply of sand, humid climate, continuing uplift of the Earth’s crust and by several other circumstances (Raukas 1968). Beside contemporary and ancient coastal dunes there are also inland aeolian formations.

Karst topography and underground features (Photo 5), resulting from the solution of rocks and leaching processes, are common on the outcrops of carbonate rocks in northern, western and central Estonia, and of limited distribution in the southeasternmost part of the Republic (Heinsalu 1977). However, because of the relatively small thickness of soluble rocks, low absolute heights of the terrain, short duration of the post-glacial evolution of the territory and for several other reasons, karst relief forms have rather modest dimensions in Estonia. Nevertheless, several features have been identified here, including karren, karst holes, open jointings, karst relicts, funnel-sinks of absorption, angular subsidence sink-holes, underground and disappearing rivers, small caves, and other karst phenomena. In some places one may find peculiar temporary karst lakes.

Beside karst caves there are suffosion caves (Heinsalu 1987). For the most part, they are found along the slopes of the ancient South-Estonian valleys and scarps of the North-Estonian Klint. Calcium carbonate precipitates out of spring water forming dome-shaped travertine terraces on valley slopes.

Gravitational relief forms composed of debris are for the most part distributed at the foot of the North-Estonian and West-Estonian klints. In the hilly topography of southeastern Estonia, deluvial processes are ongoing due to human impact.

Biogenic (organogenic) relief forms were formed by the plant growth (phytogenic forms) and animal activities (zoogenic forms). The largest phytogenic relief forms having great landscape-forming significance are the bog or telmatogenic plains, with their hummocks, mochezhinas and other relatively small features (Masing 1977).




A. Rõõmusoks, V. Puura, A. Raukas & E. Mark-Kurik


The essential natural resources of Estonia have been used for centuries. The first geological studies were carried out in the 17th century by members of staff at Tartu University. Publications of that time include G. Mancelius’ paper on earthquakes (1619) and M. J. Herbinius’ survey of waterfalls (1678). After reopening of Tartu University in 1802, the foundation was laid for the systematic scientific research into Estonia’s geology.

Orviku and Viiding (1986) differentiated several stages, substages and periods in the history of geological research in Estonia. Within the first stage, which covered the time span from the beginning of human settlement in Estonia up to the reopening of Tartu University, they distinguished three substages: (1) gathering of elementary empiric knowledge about local geological monuments; (2) limited use of local mineral resources; (3) rapidly expanding use of local building materials after Estonia had been occupied by Germans, Danes and Swedes.

During the second stage, the most noteworthy events were the establishment of the Cabinet of Mineralogy at Tartu University in 1820, the foundation of the Tartu (later Estonian) Society of Naturalists in 1853, the University reform in 1892, the October Revolution in 1917, the birth of the independent Republic of Estonia in 1918, the incorporation of Estonia into the Soviet Union in 1940 and the restoration of the independent Republic of Estonia in 1991.

This chapter will deal only with the main outlines of the historical studies.


Organisation of geological education, research and exploration

After reopening in 1802, the old (1632) Tartu (Dorpat, Derpt, Jurjev) University developed into an outstanding centre of geological education in the former Russian Empire. Moritz von Engelhardt (1779–1842), who was born in Estonia but received special education in Germany (Leipzig, Freiberg), became the first professor of mineralogy at Tartu University in 1820 (Photo 6). Students came to Tartu beside Estonian and Livonian districts also from Poland, Lithuania and other countries. Lecturing in the Estonian language started in 1920. Hendrik Bekker (1891–1925), the first Estonian professor of geology, defended his Ph.D. degree at London University in 1921 (Photo 7).

During 1920–1940, only a few geologists were trained at Tartu University. A special governmental body - the Geological Committee (Survey) of Estonia was formed in 1937, but soon after Estonia was occupied by the Soviet Union in 1940, it ceased to exist. During the German occupation (1941–1944) it operated as the Department of Geology of the Institute of Industrial Research and afterwards, with the similar staff, as the Department of Mineral Resources of the Central Research Institute of Industry of the Estonian S.S.R. In 1947, the latter gave its staff to the new institution — the Institute of Geology of the Estonian Academy of Sciences (Jaanusson 1994). Up to the beginning of the fifties this institute carried out both geological survey and basic research. The small scientific staff of the geological department of Tartu University concentrated its efforts in some special fields of research.

Only a few Estonian geologists survived through World War II. Since the end of the war, some 350 geologists have graduated from Tartu University. Several tens of Estonian-born geologists were trained at the universities in Moscow, Leningrad (St. Petersburg), Irkutsk, Tomsk (all in Russia) and Vilnius (Lithuania).

In 1957, the Geological Survey of Estonia was founded and the Institute of Geology became mainly engaged in basic studies. Since then, the geological, hydrogeological, geophysical and other studies, initiated by branch establishments of central institutions of the former Soviet Union after World War II, were gradually turned over to local organisations. In the course of complex studies aimed at geological mapping, mineral prospecting and exploration, tens of thousands boreholes penetrating deep into the sedimentary cover and more than 500 boreholes (Fig. 3) passing right through the sedimentary cover to reach the crystalline basement were made. Large-scale prospecting and exploration studies on oil shale and phosphorite were performed.

The stratigraphic studies aimed at compiling regional legends for geological mapping have always been of high priority in Estonia. Based on the results of large-scale field and laboratory studies and drilling programmes, researchers from the Geological Survey of Estonia began compilation of basic maps separately for the Palaeozoic bedrock and the Quaternary sediments, on a scale of 1:200 000. These were completed in 1975. The mapping on a scale of 1:50 000 is under way. A large number of complementary maps dealing with the geomorphology, hydrogeology, engineering geology, distribution of mineral resources, geological structure and several other aspects, and provided with the relevant comments, have also been compiled.

During the late 1970s and the 1980s, as a result of joint efforts of scientists of Estonia, Latvia, Lithuania and the Kaliningrad District of the Russian Federation, a set of 15 general geological maps of the East Baltic (1:500 000) was compiled and published (Grigelis & Puura 1980).


Evolution of scientific ideas

At the dawn of biostratigraphical studies in Estonia, E. Eichwald (1795–1876), first and foremost in several papers from 1840, introduced the terms Cambrian, Silurian and Devonian (sensu Murchison 1839) in the area (Photo 8). In 1856–1858, Fr. Schmidt (1832–1908) proposed a reasonably detailed, palaeontologically and lithologically motivated subdivision of the northern Baltic Lower and Upper Silurian (now Cambrian, Ordovician and Silurian) rocks on stages level for the outcrop area, and an excellent geological sketch map.

In the last century it was already understood that the extraordinary survival of sediments, defined today as Vendian and Palaeozoic, in a nearly original low lithified level with perfectly preserved rich assemblages of fossils of calcitic, phosphatic and chitinous skeleton could be due to a very low tectonic compression and a small depth of burial which, according to recent estimation, never reached deeper than some 800 m during the whole geological history. In the middle of the last century, also the general feature of the regional geological structure — a homocline with a very gentle southward inclination of the weakly disturbed Palaeozoic strata overlying the crystalline (“granitic”) basement, was described by Fr. Schmidt (Photo 9), C. Grewingk (1819–1887) and several others.

Since the pioneer work of palaeontologists of several generations, tens of thousands specimens of fossils have been collected during systematically arranged field studies. The collections stored at the Institute of Geology and at the Geological Museum of Tartu University include about 1000 type-specimens of new species of fossils. A great number of new genera and families have been established. The most numerous groups of fossils, mostly Ordovician and Silurian, which have been studied in particular detail, are marine invertebrata: articulate and inarticulate brachiopods (studied by A. Rõõmusoks, M. Rubel, L. Hints, I. Puura during the last decades), molluscs of different classes, arthropoda like trilobites (Reet Männil), ostracods (L. Sarv, T. Meidla), merostomata and different echinodermata (Ralf Männil), tabulate (E. Klaamann) and rugose (D. Kaljo) corals and stromatoporoids (H. Nestor), bryozoans (Ralf Männil), graptolites (D. Kaljo), conodonts (V. Viira) etc. Besides, chitinozoans of problematic origin (Ralf Männil, J. Nõlvak, V. Nestor) as well as fossil elements and fragments of endo- and exoskeleton of early representatives of vertebrata from the Late Silurian and Devonian (E. Mark-Kurik, T. Märss) have also been studied.

Based on the data on the distribution of fossils in the Vendian – Middle Palaeozoic sedimentary record, a stratigraphic scheme for the Baltic States and the East-European Platform, as a whole (Ralf Männil, D. Kaljo, K. Mens, E. Mark-Kurik, H. Nestor, L. Hints a.o.) and a detailed biostratigraphic subdivision of the Ordovician and Silurian rocks in the North Baltic area were elaborated. The ecostratigraphic studies, taking into account peculiarities in the distribution of different fossils in lateral changes of lithologies, provided a basis for the correlation of the successions of different facial belts of the Early Palaeozoic pericratonic Baltic Basin (Reet Männnil, D. Kaljo a.o.). An impact of global and regional geologic/tectonic and palaeoenvironmental processes, like the continental drift of the Baltic Continental Plate from the south polar position to the northern tropics during the Late Vendian – Devonian, the continental glaciation in Gondvana in the Late Ordovician, causing ocean-level changes, and pericratonic tectonic activities along the continental palaeomargins with specific influence on sedimentation and fossil communities, has been fixed in the geological sedimentary record of the Baltic Palaeozoic marine basin.

Estonia was among the first regions where the theory of continental glaciation in the middle of the last century was applied (E. Eichwald, Fr. Schmidt) and later the structure and formation of different landforms and the evolution of the Baltic Sea were described in detail (Karl Orviku, H. Kessel, A. Raukas, E. Rähni a.o.).


Structural studies

Up till the 1950s, the area of Estonia was structurally classified as a simple, almost horizontal homocline. Detailed investigations of the sedimentary cover and the studies carried out within large-scale drilling and geophysical survey programmes revealed a typical fault-and-block tectonic pattern of both the sedimentary cover and the underlying crystalline basement (Vaher et al. 1962, Puura 1979). Classifications of the fault-related linear structures were suggested during the detailed structural studies of the oil shale (Puura 1986) and phosphorite basins (Puura 1987). The recent local crustal movements and weak earthquakes occurring against the background of the postglacial crustal uplift (Orviku 1960c) are, to a certain extent, related to faulting lines (Sildvee & Vaher 1995). The tectonic map with the accompanying explanatory text for the whole East Baltic territory was published by Suveizdis et al. (1979).


Research into the crystalline basement and the Vendian-Cambrian strata

The existence of the crystalline (“granite”) basement under the sedimentary cover was recognized in Estonia in the middle of the 19th century (Schmidt 1881), but systematic geological research into the basement was initiated not until the 1960s. The studies showed that in northern and north-eastern Estonia the basement consisted mainly of migmatized metamorphic rocks of the amphibolite facies and in southern and southwestern Estonia of granulite facies (Puura et al. 1976). The recent Sm-Nd dating of folded rock assemblages indicated Palaeoproterozoic age of the orogenic continental crust in southern Estonia (Puura & Huhma 1993) and in the whole Byelorussian – Baltic granulite province (Gorbatschev & Bogdanova 1993). Thus, the basement of Estonia was considered as a continuation to the Svecofennian Domain of the Fennoscandian Shield.

The first evidence about the Vendian rocks was obtained in 1842–1845 from a 90-m-deep borehole in Tallinn (Helmersen 1851). The first borehole, which reached the crystalline basement at a depth of 162.8 m, was drilled in 1898–1899 at Aseri. According to Sokolov (1953), the thickness of the Vendian complex was 92.5 m there (Rüger 1923). As a result of recent studies (Mens & Pirrus 1971, 1974, 1980, 1986), a great lateral and vertical facial variability has been established and a detail lithostratigraphical subdivision presented.

As early as the first half of the 19th century, the Cambrian sequence was divided into a lower, Blue Clay Unit and an upper, Sandstone Unit (Engelhardt 1820, Strangways 1822, Eichwald 1825). Following Murchison’s concept of the Silurian system, Eichwald (1840) included these two units into the Silurian, and Schmidt (1858), more specifically, into the Lower Silurian. Within the Cambrian sandstone upon the blue clay, Linnarsson (1873) suggested to distinguish units equivalent to the Swedish Euphyton Sandstone and Fucoid Sandstone units. These two terms were widely used in Estonia until denomination on the basis of geographical names (Öpik 1933).

A great step forward was A. Mickwitz’s discovery of malacofauna in the Eophyton Sandstone (Schmidt 1888) and the finds of an olenellid trilobite which proved Lower Cambrian age of these beds, and also the finds of the brachiopod Mickwitzia which indicated the same age with the correlatable beds in Sweden. Öpik (1925, 1926, 1929) contributed further important biostratigraphical data on the Lower Cambrian and revised the terminology (Öpik 1933). He (Öpik 1956) was convinced that in places the terrigenous beds dated as the basal Ordovician could be of Late Cambrian age. An increased access to numerous cores of borings all over Estonia enabled the subsurface Cambrian to be studied in more detail. Since the 1960s, many papers dealing with various aspects of the Lower Cambrian sequence, including a detailed lithostrati-graphic classification (Mens & Pirrus 1977), have been published. In southeastern Estonia, some parts of the terrigenous sequence were supposed to belong to the Upper Cambrian (Volkova et al. 1981). Intense biostratigraphical studies aimed at determining a convenient basis for the Ordovician System indicate that a comprehensive lower portion of the sandstone, previously regarded as basal Ordovician, is actually Late Cambrian in age.


History of Ordovician research

The study of the Ordovician strata in Estonia was commenced with the descriptions of the North-Estonian Klint (Severgin 1808, 1809). Engelhardt (1820), Strangways (1821, 1822) and Eichwald (1825) pointed out the similiarity of the North Estonian and Scandinavian sequences. Strangways (1822) produced the first geological-lithological map which included also Estonia.

In the history of Cambrian-Silurian stratigraphy of Estonia, the second half of the 19th century is known as the Schmidt’s Epoch (Männil 1986). Friedrich Schmidt (1832–1908), a descendant of Baltic Germans, published a comprehensive survey of the Cambro-Silurian outcrop area in northern Estonia (Schmidt 1858) which included the stratigraphic classification of the Ordovician sequence in the East Baltic area and a bedrock map. The units of Schmidt (Table 1) are neither strictly lithostratigraphical nor based on guide fossils, but reflect the stages of development of the regional benthic fauna without any preference for any particular group of fossils. In modern terms, they could be compared to topostratigraphical units. This approach, together with the use of geographical names for the units, was followed by subsequent generations of geologists in the research of the Estonian bedrock.

Based on a detailed study of the post-Tremadocian Lower Ordovician in Ingria, Lamansky (1905) introduced a stratigraphic classification. It was primarily based on ranges of trilobite species which he could follow into northern Estonia. An important result of his study was the recognition of numerous breaks in the correlatable North-Estonian sequence.

In the late 19th and early 20th century, a number of important taxonomic monographs were published by both native and foreign specialists (Rosen 1867, Schmidt 1874, Pahlen 1877, Dybowski 1877b, Holm 1886, Mickwitz 1896, Stolley 1896b, Koken 1897, 1925, Hoyningen-Huene 1899, Jaekel 1899, Bonnema 1909), and the data on the Ordovician of Estonia was included in several other publications as well. Of particular importance was the series of monographs on trilobites by Schmidt (1881, 1885, 1894, 1898, 1901, 1904, 1906, 1907). With Bassler’s (1911) monograph on bryozoans the former Lower Silurian Series was, as far as the Estonian sequence was concerned, definitely elevated to the rank of a system, termed Ordovician.

In 1914, the renowned American geologists P.E. Raymond and W.H. Twenhofel examined Cambro-Silurian exposures in Ingria, Estonia and Scandinavia with the purpose of attempting a correlation with eastern North America. Raymond (1916) proposed five new terms (Table 1) and divided the Ordovician System in the Baltoscandian Basin into three series: Lower, Middle and Upper Ordovician. According to current correlations, the boundaries of these series correspond very closely to those between the Ordovician series of North America. Bekker (1921) introduced the term Stage (Table 1) for the basic chronostratigraphic unit (= beds in Schmidt 1881) and revised the spelling of geographic names to accord modern maps (Bekker 1922, 1925). Armin Öpik (1898–1983), the successor of Bekker on the chair of geology at Tartu University (Photo 10), focused, as far as the Ordovician stratigraphy is concerned, on studying the Tremadocian sequence (Öpik 1929). However, his main contribution was in the field of palaeontology, documented by a number of monographs on brachiopods (Öpik 1930b, 1933, 1934, etc.), trilobites (Öpik 1937b), ostracodes (Öpik 1937a) and some other groups. In a manuscript in 1934, he proposed the terms Viru Series for the Middle Ordovician and Harju Series for the Upper Ordovician of the region, which were first published by Luha (1940a). Karl Orviku (1903–1981) made a noteworthy contribution to knowledge of the Ordovician in Estonia mainly in the field of detailed lithostratigraphy (Photo 11). His excellent monograph on the lower Middle Ordovician of northern Estonia (Orviku 1940), in which he distinguished the Lasnamägi and Uhaku stages (Table 1), exerted great influence on the succeeding generations of Estonian geologists. His analysis of discontinuity surfaces (hardgrounds), in particular, both with regard to their morphology and importance as markers of stratigraphic breaks, received wide attention. In subsequent papers Orviku (Orviku 1960a, b) gave a detailed lithostratigraphic classification of the Lower Ordovician Volkhov and Kunda stages of northern Estonia.

The scientific activity of the succeeding generation (Jaanusson, Männil, Rõõmusoks, Kaljo, Aaloe and several others) started in an organisation known as the Section of Geology of Gustavus Adolphus Natural Science Circle. As a result of field studies and later publications, the Schmidt’s hitherto poorly known Lyckholm beds (Table 1) were divided into three separate stages (Jaanusson 1944b, 1956) and the stratigraphy of the Viru Series (Middle Ordovician) of northern Estonia was revised (Jaanusson 1945). In the upper Middle Ordovician, the occurrence of K-bentonites was observed (Jaanusson & Martna 1948), and their importance for correlation and stratigraphic classification recognized (Jürgenson 1958a, Bergström et al. 1995).

The progress in knowledge of the Ordovician sequence in Estonia after the war was largely due to the availability of numerous cores of drill borings. Ralf Männil (1924–1990) solved the somewhat confused stratigraphic terminology of the Schmidt’s unit F1a by introducing the term Nabala Stage (Männil 1958b). He also showed (Männil 1958c, d, 1960) that information from borings indicated Schmidt’s Vasalemma beds to represent a lithostratigraphic unit which also includes the beds of Keila Age. For the chronostratigraphic unit above the Keila Stage Männil proposed the term Oandu Stage, borrowed from the poorly exposed lithostratigraphic unit Oandu beds in northeastern Estonia (Öpik 1934) which was unknown to Schmidt. Kaljo, Rõõmusoks and Männil (Kaljo et al. 1958) emphasized the need of regional terms for Ordovician series and proposed the Oeland Series for the Lower Ordovician in the accustomed Baltoscandian usage. The post-Tremadocian Lower Ordovician was renamed Ontika Subseries.

As a result of a thorough examination of the subsurface Ordovician in Estonia, Latvia and Lithuania, Ralf Männil (Männil 1966, preliminary report 1964) summarized the development of the Baltoscandian Basin in the Ordovician Period. He showed that in an extensive west-central area of the East Baltic region (Livonian Tongue after Jaanusson 1976), including southern Estonia, the post-Tremadocian Ordovician rocks are basically of the same type, both lithologically and faunistically, as on the Swedish mainland north of Scania, and that they belong to a single Swedish-Latvian Facies Zone (Central Baltoscandian Confacies, Jaanusson 1976). The Estonian Facies Zone (North Estonian Confacies Belt, Jaanusson 1976) was proved to be conspicuously different in many respects. Põlma (1967, 1972, 1982) presented a comparative analysis of the carbonate lithology of these main regions and distinguished a lithologically transitional belt, characterised by transitional lithologies and interlocking pattern of various lithofacies from the south and north. The relation of the boundaries of the transitional belt to those of confacies belts is still disputed (Jaanusson 1995, Meidla 1996). Of great palaeogeographical interest was the discovery that the Varangu beds (Männil 1958a), which occur in a limited area of northern Estonia, correlate with the upper Tremadocian Ceratopyge Shale in Scandinavia (Viira et al. 1970, Kaljo & Kivimägi 1970). The stratigraphy of the Middle Ordovician (Viru) of North Estonia (stratotype area) was summarised in a comprehensive monograph by Rõõmusoks (Rõõmusoks 1970), and that of the Ordovician of Estonia in general, in a chapter of a separate book by the same author (Rõõmusoks 1983).

The rich knowledge of the Ordovician fauna of Estonia has increased the precision of correlations not only within Estonia but also with the Ordovician sequences elsewhere (Nõlvak & Grahn 1993, Männil & Meidla 1994, Jaanusson 1995).


History of Silurian research

M. von Engelhardt (Engelhardt & Ulprecht 1830) was the first to draw attention to the differences between the rocks now classified as Ordovician and Silurian. He noted that on mainland Estonia limestones containing orthoceratite cephalopods and trilobites were succeeded southwards by younger rocks with corals and pentamerid brachiopods.

In the monograph by Murchison, Verneuil and Keyserling (1845) the Silurian sequence of Estonia is briefly but adequately summarised. Murchison distinguished within the Silurian of Estonia (in the current sense) the following three units in ascending order (1) Pentamerus Limestone, (2) Limestone with corals, and (3) Limestone with Terebratulas.

Schrenk (1854) gave the first comprehensive lithological survey of the Silurian localities in Estonia. Schmidt (1858) continued the study in a much greater detail and with emphasis on fossil fauna. His classification of the Silurian sequence of Estonia into three groups is roughly comparable to Murchison’s tripartite subdivision and the distinguished groups correspond to the Llandoverian, Wenlockian and Ludlovian + Downtonian Series in the current sense (Table 2). The lower, Schmidt’s Group with Smooth Pentamerids (= Llandoverian), was subdivided into smaller units and the base of the Upper Silurian defined, in the outcrop area of northern Estonia, at the level of the present Ordovician-Silurian boundary in the area. Schmidt interpreted the Borealis Banks as local mass accumulation of shells (Muschelbänke) and regarded the unit to be somewhat artificial in a chronostratigraphic sense. Subsequently, Schmidt ( 1879, 1881) introduced capital Latin letters combined with Arabic numbers (G1, G2, G3, etc.) as symbols for this unit. He (Schmidt 1892) further refined his stratigraphic classification by distinguishing within the Upper Oesel Group Eurypterus beds and Ilionia (Didyma) beds which correspond to the Rootsiküla and Paadla stages of the current classification, respectively.

The Silurian fauna of Estonia received early attention, especially by Eichwald, in various papers. A contribution of great international importance was the monograph on Silurian (=Ordovician + Silurian) fishes by Pander (1856). The Silurian (s.str.) material was mostly derived from the Schmidt’s Upper Oesel beds of Saaremaa. The description of Pander’s new group Conodonta was partly based on specimens from Saaremaa.

The exquisitively preserved specimens of merostomes, fishes and some other uncommon groups from exposures at Viita in Rootsiküla on Saaremaa Island were of wide international interest. The exposures were located by Schrenk in 1852 and material from those beds formed the subject of many papers. The monographs by Rohon (1892, 1893) and Holm (1898) should be mentioned in the first place.

The Schmidt’s stratigraphic terminology was somewhat inconsistent which made Twenhofel (1916) to propose more adequate terms for some units (Table 2). Bekker (1922) revised the terminology to accord modern maps and started field work with an aim of studying the stratigraphy of the Silurian sequence of Saaremaa in more detail (Table 2). On account of his untimely death, only an outline of his stratigraphical results became published (Bekker 1925). The study was continued by Artur Luha (1892–1953) who assisted Bekker in the field. Regrettably, only a condensed version (Luha 1930) of his voluminous manuscript was published.

In 1929, Luha discovered a locality with a rich agnathan fauna at Himmiste-Kuigu on Saaremaa. Later studies (Aaloe 1963b) showed that the Himmiste beds were younger than previously believed (about middle of the Paadla Stage).

Luha (1933) improved the chronostratigraphic classification by restricting the term Jaani Stage to beds now known to correspond to the Lower Wenlockian (Table 2). Later Luha (1946) introduced the term Jaagarahu Stage for the unit comparable to the Upper Wenlockian. Teichert’s (1928) study on the Lower Llandoverian in the western part of mainland Estonia and on the Island of Hiiumaa is also worthy of mentioning.

Studies of the Silurian resumed in the mid-1950s. Numerous borings made it possible to extend examination of lithofacial relationship also south of the outcrop area, and facilitated establishing of a detailed lithostratigraphic classification (Aaloe 1958a, 1960, 1961, Aaloe & Kaljo 1962, Einasto 1962, etc.). Kaljo and Sarv (1966) specified equivalents to the Downtonian Series in the Estonian sequence. Examination of graptolites (Kaljo 1967, Kaljo & Vingisaar 1969) contributed to a more precise correlation of the sequence with other areas. As the Kaugatuma Stage proved to be a composite unit, Klaamann (1970a) distinguished its lower, Ludlovian part as a separate Kuressaare Stage. The discovery of Llandoverian K-bentonite beds in borings on Saaremaa and in the southwestern part of mainland Estonia by Jürgenson (1958a, 1964) contributed to correlations. Studies of lithology of the Silurian rocks received increased attention (Jürgenson 1966, 1974, etc.).

Biostratigraphic correlations were greatly facilitated by monographic descriptions of several groups of microfossils such as ostracodes (Sarv 1968), agnathan scales (Märss 1996), chitinozoans (V. Nestor 1994) and conodonts (Männik 1992b).

Silurian macrofauna was described in numerous papers, some of those monographic in character. Various groups were covered, including stromatoporoids (Nestor 1964, 1966), tabulate corals (Sokolov 1951b, 1952b, Klaamann 1961, 1962, 1964, 1966, 1970b), bryozoans (Astrova 1970), articulate brachiopods (Rubel 1970a, b), and trilobites (Männil 1982, 1992).

The impressive array of information from the Silurian of Estonia is systematised and discussed in several books, covering general features (Kaljo 1970c), facies and fauna (Kaljo 1977), communities and biozones (Kaljo & Klaamann 1982), ecology (Kaljo & Klaamann 1982) and lithology (Jürgenson 1988).


History of Devonian research

Engelhardt and Ulprecht (1830) provided the earliest information of the red sandstones and mentioned also finds of vertebrate teeth and bone fragments. Kutorga (1835, 1837) described the exposures at Tartu (Fig. 4). Eichwald (1840a, c) and Buch (1840) recognised very early that the terrigenous sequence was comparable to the Devonian Old Red rocks in Great Britain. Helmersen was the first to show the approximate outcrop area on a map. A similar map was provided by Murchison et al. (1845). The available information was summarised by Grewingk (1861, further specifications in 1878 and 1879, Photo 12).

Quenstedt (1838) suggested that the teeth and bones recovered from the sandstone belong to ancient fish-like forms. The fairly substantial material of fish remains was subject to monographic studies by Asmuss (1856) and Pander (1856, 1860).

Based on the field studies of the Upper Devonian carbonate rocks of the southeastern part of the Republic, Bekker (1924a) attempted a chronostratigraphic classification of the sequence of which the term Dubniki Stage is still in use. In the current stratigraphic practice the other terms are based on the sections in the neighbouring areas of Latvia and Russia where carbonate rocks of this age have a wider distribution.

The lithostratigraphy of the basal Old Red lithofacies in the outcrop area was subject to a detailed study by Orviku (1930c, 1932, 1935b). Obruchev (1933) distinguished two stratigraphic units, now regarded as the Pärnu and Narva stages. Subsequently, Orviku (1948) published an additional comprehensive paper on the stratigraphy of the Narva River Stage (=Narva Stage).

An important contribution to the knowledge of the Old Red sequence was the biostratigraphic classification, based on fishes from both Estonia and Latvia. The classification was developed by Gross (1933, 1940a, b, 1942, 1951) and later continued by E. Kurik (née Mark). Mark (1958) distinguished the Eifelian Aruküla Stage. The historically known Aruküla caves near Tartu (Photo 13) where, since 1831 during more than 20 years, H. Asmuss had excavated bones of large placoderms and other fishes, were selected for the type locality.

During several years between World War I and II, V. Paul (1934, 1939) excavated fossil fishes at Tamme near Lake Võrtsjärv and at Haaslava. The excavations, organized by E. Mark-Kurik, started in 1949 and lasted with a few intervals up to 1993 (more than 10 excavations were made at Karksi alone). Taphonomical study was provided during these excavations.

The overlying, lower Givetian unit was distinguished by Mark (1958) as the Burtnieki Stage with the stratotype in Latvia. The stratigraphy of the Devonian in Estonia was summarised by Mark and Paasikivi (1960), including new information obtained from examination of cores from numerous deep borings. Special attention was further paid on studying the mineral composition of terrigenous rocks, largely on the basis of core material. The studies of the kind were initiated by H. Viiding (1929–1988) and supplemented particularly by A. Kleesment (née Tamme). Numerous papers were published on the subject (Viiding 1962, 1964, 1965, 1976b; Tamme 1962, 1964; Mark & Tamme 1964). Based on lithological and mineralogical criteria, Kleesment (1966) and Kleesment et al. (1975) recognised in the borings of southeastern Estonia the presence of the Lower Devonian Tilžė and basal Middle Devonian Rēzekne stages, previously distinguished in Latvia and Lithuania. Mark-Kurik (1991a) has considered the Rēzekne Stage as a lower Devonian unit.

The Devonian stratigraphy in Estonia was subsequently summarised in the papers published by various authors in the book “Devonian and Carboniferous of the Baltic” (Sorokin 1981). Recently, Kleesment (1994, 1995) published important contributions to detailed lithostratigraphy of the Middle Devonian sequence in Estonia. Mark-Kurik correlated Estonian Devonian units with those in other areas, e.g. in Scotland and Latvia (Mark-Kurik 1981, 1991a, b, Kurik et al. 1989). Valiukevičius (1994) gave an acanthodian zonation for the Baltic Basin (including Estonia). The fossil fishes of the Devonian sequence continued to receive attention. Beside the above-mentioned papers by Gross, the contributions by Heintz (1930, 1934), Mark (1953a, b, 1956, 1963), Mark-Kurik (1968, 1973, 1993a, b, Mark-Kurik et al. 1991), Obruchev and Mark-Kurik (1965, 1968) deserve special mention. Mark-Kurik has also studied special problems of palaeoichtyology: palaeopatology (Mark-Kurik 1966), functional morphology (1984) and trophic relations (1995). Rich collections in Estonia have provided an excellent basis for studying of several fossil fish groups: antiarchs by Karatajūtė-Talimaa (1960, 1963), arthrodires by Obrucheva (1962, 1966), acanthodians by Valiukevičius (1985) and crossopterygians by Paul (1940), Vorobyeva (1977). Thomson (1940) pioneered in studying the Devonian plant remains (Photo 14). Later Vaitiekūnienė (Kleesment et al. 1975) and Kalamees (1988) studied spores and macroremains (including phytoleimma), respectively. A number of papers contain descriptions of the Devonian invertebrates from Estonia: ostracodes (Öpik 1935a, Polenova 1966), lingulates (Batrukova 1960, 1964, 1969, Gravitis 1981) and conchostracans (Mironova 1969).


History of Quaternary research

The first stratigraphical scheme of the Quaternary deposits was compiled by Schrenk (1854). Based on the then prevailing drift theory he, like Schmidt (1854) and Grewingk (1861), divided all Quaternary sediments into diluvial and alluvial sediments with several lithological varieties.

Estonia was among the first regions where the theory of continental glaciation was applied. Eichwald (1853) was the first in the Baltic provinces to consider the possibility that at least northern Estonia had once been covered by an ancient active glacier. Already in 1865, Schmidt (1865) clearly spoke about glacial sediments, and a bit later he (Schmidt 1869) described glacial and postglacial formations and differentiated four stages in the development of the territory, including the time of the invasion of big glaciers, the time of the melting of the glaciers, and the time of the final melting of the ice with a wide distribution of fresh-water lakes. Schmidt (1858) was the first who found the shells of Ancylus fluviatilis in beach deposits on Saaremaa Island and distinguished a fresh-water stage in the Holocene history of the Baltic Sea.

Helmersen (1869) explained the distribution of erratic boulders and formation of boulder clays and ice scratches with the joint action of the continental ice sheet, floating icebergs and erosional processes on land. Schmidt (1871) proved that in the glacial epoch a unitary glacier moved from Scandinavia over the Baltic Sea depression and Estonia. In 1879, Grewingk already spoke about several glaciations (Grewingk 1879), basing on the study of the different till beds in Tartu.

However, it was not until half a century later that Grewingk’s statement was confirmed by palaeontological data. In 1939, Orviku performed the first detailed studies on interglacial organogenous deposits at Rõngu (Orviku 1939) which, according to pollen zones (Thomson 1939a, 1941), were correlated with the typical Riss-Würmian (Eemian) interglacial deposits in Western Europe. By now, tills of five glaciations or big stadials have been identified in Estonia (Raukas 1978) and both Eemian and Holsteinian interglacial sediments described in detail (Liivrand 1991). Official stratigraphical schemes of the Quaternary (Raukas & Kajak 1995), Late-glacial (Pirrus & Raukas 1996) and Holocene (Raukas et al. 1995b) have been accepted and published.

Monographic studies of ice-marginal formations (Raukas et al. 1971), bedrock topography (Tavast & Raukas 1982), lithology of Quaternary deposits (Raukas 1978), modern (Orviku Kaarel 1974) and ancient (Kessel & Raukas 1967) coastal formations have been published. The overviews about the glacial history (Raukas 1995a) and the history of the development of geomorphology in Estonia (Raukas & Karukäpp 1993), imparting more information about the study history of different landforms and types of sediments, appeared recently in print.


The authors of the chapter would like to thank Prof. Valdar Jaanusson for his numerous valuable comments.


V. Puura, V. Klein, H. Koppelmaa & M. Niin


The early Precambrian crystalline rocks are covered by the Upper Vendian and Palaeozoic sedimentary rocks. The basic data for studies of the Precambrian has been obtained by means of boreholes and geophysical survey. Of about 500 boreholes passing right through the sedimentary cover, the deepest ones penetrate into the basement to a depth of up to 450 m.

The Precambrian basement in Estonia consists of two megaunits: the orogenic Svecofennian complex of metamorphic and plutonic rocks and the anorogenic complex of plutonic rapakivi granites and related rocks. Earlier views about the age of basement rocks (Puura et al. 1976, 1983) have considerably changed during the last decade due to the isotopic dating of the South-Estonian granulitic crust (Puura & Huhma 1993) and rapakivi granites of Fennoscandia (Rämö et al. 1996). The new standpoints, based on the results of these studies, underlie the recent joint publications on the Precambrian of the Gulf of Finland and surrounding area (Koistinen 1994, 1996; Laitakari et al. 1996). Based on recent results, a stratigraphic chart of Precambrian rocks of Estonia was compiled (Table 3).



Svecofennian orogenic metamorphic rocks

According to the degree of metamorphism and the composition of the metamorphic sequences reflected in the geophysical patterns, Estonia’s basement is divided into structural regions (Figs. 5, 6) which differ from each other in the volume of sedimentary and felsic to mafic volcanic rocks.

Comparison of the basement-forming rocks in Estonia, Finland and Sweden has shown that the metamorphosed volcanic and sedimentary rocks in Estonia’s basement have many features in common with the rocks in the Svecofennian orogenic complexes. In the first instance, it was established that the rocks of the Tallinn and Alutaguse structural zones and the Svecofennian complex are similar in lithology and have the same stage of metamorphism (Puura et al. 1983, Klein 1986). The orogenic supracrustal rocks of southern and western Estonia differ from the bulk of Svecofennian rocks of Scandinavia by their prevailingly mafic to intermediate composition and high‑grade metamorphism. This is in good correlation with high anomalies of gravity and magnetic fields (Puura et al. 1976, 1983, Koppelmaa et al. 1978). The Palaeoproterozoic age of the granulite complex of Estonia was dated by Sm‑Nb isotopic studies (Puura & Huhma 1993). Petrological signatures of mafic rocks in southern and western Estonia are concordant with those in the northern and northeastern parts of the territory.

In the Alutaguse Zone (Fig. 5) gneisses containing biotite, cordierite, garnet and sillimanite intercalate with biotite gneisses and form a complex of the same name (Puura et al. 1976). Within the Alutaguse Zone, in the area of Uljaste, Haljala and Assamalla, the basement comprises sulphidic black schists, quartzites, amphibole and pyroxene gneisses, marbles and pyroxene skarns.The rocks in the Alutaguse Zone derive from clastic successions with minor sequences of volcanic, sandy and carbonate rocks in the above-mentioned areas. The local uplifts of the basement in the Uljaste and Assamalla area consist primarily of quartzites.

The Al-rich, sillimanite‑garnet‑cordierite gneisses are, for the most part, medium‑grained, banded and migmatized by plagioclase‑microcline granite or pegmatite. The mineral composition of gneisses varies. Light minerals are represented by quartz, plagioclase (An 25–55) and microcline; dark minerals by biotite, cordierite, garnet and sillimanite. Muscovite is rare, while andalusite is occasional and rare. Chemically, the sillimanite‑ garnet‑ cordierite gneisses are similar to pelites and originate, in all likelihood, from psammitic to pelitic sediments. Microgneisses, rich in quartz, form interlayers with the highest sand content in these sediments.

The Tallinn Zone is characterised by the stratified Jägala Complex of intercalating sillimanite‑cordierite and biotite gneisses, intermediate to mafic metavolcanics, and leucocrate gneisses. In the WNW‑ESE‑trending zones, the primarily psammitic to pelitic metasediments intercalate with metavolcanites. The acidic metavolcanites alternate with more abundant intermediate to mafic metavolcanites.

The intermediate metavolcanic rocks are represented by metamorphosed to biotite‑hornblende and biotite gneisses which are fine‑ to medium‑grained and migmatized by microcline‑ plagioclase granites. The main minerals of biotite gneisses are plagioclase (An 30–45), quartz, biotite and, in places, microcline. The basic minerals of biotite‑hornblende gneisses, which are slightly more mafic in composition, are plagioclase (An 35–50), quartz, hornblende and biotite. These gneisses are andesitic (SiO2 55–63%, Na2O + K2 = 4.5–6.5%).

Typical felsic quartz‑feldspar gneisses are fine‑grained, rather massive or schistose granoblastic rocks. In single boreholes, rocks with relicts of blasto‑porphyritic texture (phenocrysts of quartz and plagioclase), indicative of their volcanic origin, have been found. Quartz (25–40%), plagioclase (An 20–40) and potassium feldspar form 85‑95% of quartz‑feldspar or granite gneisses. In the chemical composition (SiO2 65–76%, Na2O + K2O = 5.5–8%) the quartz‑feldspar and granite gneisses are similar to acidic volcanites (dacites, rhyolites) and arcosic sandstones.

Petrographically, the Al-rich gneisses of the Jägala Complex are similar to those in the Alutaguse Zone.

The basement in the West-Estonian Zone consists predominantly of the same assemblage of rocks as in the Tallinn Zone, although the stratified structure is not so well reflected in geophysical anomalies. The rocks, characteristic of this zone, are rather uniformly medium‑ and fine‑grained biotite‑hornblende gneisses and amphibolites, which have been migmatized by microcline granites. The amphibolites mostly occur as layered bodies and are intercalated with gneisses. In the amphibolites, plagioclase (An 35–55) and hornblende are the main minerals, however, they may also contain biotite, clinopyroxene and quartz. The main minerals of the biotite‑hornblende gneisses are plagioclase (An 30–50), quartz, hornblende and biotite, rarely potassium feldspar. There are also gneisses, the mafic parts of which consist entirely of biotite. North of Haapasalu, on the Noarootsi Peninsula, the gneisses also contain hypersthene. According to the chemical composition, the amphibolites (SiO2 45–53%, Na2O + K 2O = 3–4.5%) are referred to basalt, and the gneisses (SiO2 55–63, Na2O + K2O = 5–6.5%) to andesite.

In the Tapa Zone, a rock association, analogous to that of the West-Estonian Zone (amphibolites, biotite‑hornblende gneisses, in places pyroxene gneisses) occurs. The Jõhvi Zone (Fig. 5) is composed of the rocks of the Vaivara Complex. Magnetite quartzites occur in a limited area together with Al-rich and pyroxene gneisses. The latter contain interlayers of quartz‑feldspar and biotite‑amphibole gneisses. The fine‑ and medium‑grained pyroxene gneisses with a variable mineral composition display charnockitic and granitic migmatization. Orthopyroxene and biotite are always present. The content of clinopyroxene and hornblende varies from 0 to 25%. Of light minerals, plagioclase (An 40–55%) predominates, while quartz and potassium feldspar are often absent. Within the Jõhvi magnetic anomaly area, rocks of almost ultramafic composition comprising orthopyroxene, clinopyroxene, hornblende, biotite and plagioclase are occasionally encountered (5–10%). Biotite‑hypersthene gneisses containing plagioclase and quartz are also widespread. Chemically, the gneisses correspond to andesite.

The magnetite quartzites, fine-grained banded rocks in the Jõhvi area, contain besides quartz and magnetite, also garnet, orthopyroxene, clinopyroxene, hornblende, cummingtonite and biotite in different quantities. The average content (by microsections) of quartz is 30–40%, with the proportion of magnetite reaching 25–30%. The magnetite quartzites are cut by veins of pegmatoid microcline granite.

In the South-Estonian Zone, the metamorphic complex consists of hypersthene, clinopyroxene and amphibole gneisses, originating from mafic to intermediate volcanites, and possibly from greywackes. It also contains Al‑rich and minor members of felsic gneisses.

Different fine‑ to medium‑grained pyroxene gneisses, which have undergone charnockitic and granitic migmatization, are characteristic of southern Estonia. The primary structures of these gneisses have been obscured or obliterated. The characteristic mineral assemblage of the hornblende‑pyroxene gneisses is orthopyroxene + clinopyroxene + hornblende + biotite + plagioclase +/‑ potassium feldspar +/‑ quartz. The plagioclase is mostly antiperthitic, mainly andesine‑labradorite, rarely bytownite. The potassium feldspar is orthoclase‑microperthite. Quartz is rare. The hornblende‑pyroxene gneisses have been found mostly in boreholes in the surroundings of Pärnu and Viljandi where they occur as interlayers in acidic gneisses. The chemical composition of the amphibole‑pyroxene gneisses corresponds to basalt or basaltic andesite (SiO2 47–54%, Na2O+K20 = 3–5%), but the content of iron, magnesium and calcium differs noticeably. The increased content of magnetite (3–4%) is a specific feature of these gneisses. The essential minerals of the biotite‑hypersthene gneisses are plagioclase (mainly An 35–45%, in some cases An 70–80%), hypersthene, biotite and quite often quartz and potassium feldspar. Gneisses of this type occur typically in the vicinity of Tartu, Otepää and Laeva. Among the biotite‑hypersthene gneisses, both melanocratic and leucocratic varieties occur (SiO2 48–60%). Compared to the hornblende‑pyroxene gneisses, the biotite‑hypersthene gneisses are generally poorer in calcium, but richer in potassium and magnesium, which is evidently due to the weathering of the source rock and mixing with pelitic matter. In the rather rare quartz‑feldspar gneisses of southern Estonia, garnet or hypersthene and hornblende occur as accessory minerals.

The gneisses, formed at granulite facies in the South-Estonian and Jõhvi zones, contain hypersthene and accessory spinel, garnet and cordierite porphyroblasts (also in granitic veins) and sillimanite, the latter occurring as inclusions in cordierite. The biotite gneisses occur together with sillimanite‑cordierite gneisses. They are medium‑ to fine‑grained, often foliated migmatitic rocks, the main minerals being quartz, andesine, biotite and potassium feldspar. Garnet, cordierite, sillimanite and muscovite occur in small quantities. The content of dark minerals averages 20–25%. Compared with other gneisses, biotite gneisses have the highest content of quartz.


Svecofennian orogenic plutonic rocks

Traditionally, the granitoid rocks of southern Finland have been classified into four groups, based on their relationship to orogenic movements (Koistinen 1994, 1996). These groups are synorogenic (synkinematic), late‑orogenic (late‑kinematic), post-orogenic and anorogenic rocks (rapakivi granites). Practically, this classification is often expanded on the whole variety of plutonic rocks.

Compared to Finland, the Estonian basement is rather poor in orogenic plutonic rocks. Synorogenic granitoid complex and associated mafics are rare in Estonia. The late‑orogenic potassium granites which form the extensive W‑E‑striking belt in southern Finland, occur as small bodies and migmatite veins in the basement of northern Estonia.

In Finland, the granite migmatites fall into two distinct age groups which are related to early (1.9–1.87 Ga) and late orogenic (1.84–1.83 Ga, Koistinen 1996) granitoids. Like in southern Finland, where the both age groups of migmatites occur in the same metamorphic complex around the Potassium Granite Belt (Koistinen 1996), age classification of migmatites in Estonia’s basement is extremely complicated and, therefore, they are treated as one orogenic group. Occasionally, the classification of other plutonic rocks into the early and late orogenic groups is possible. Small bodies of synorogenic gabbronorite and gabbro, or metagabbro, cut by granite veins, occur in the Tapa Zone. The mafic rocks contain abundant hornblende and biotite of later origin. Similar rocks in northeastern Estonia form the considerably large Pada Pluton, which contains also diorite.

Drilling in southern and western Estonia has revealed some mafic, probably synorogenic rocks. There are small gabbro-norite plutons including Võru, Laeva, Pärnu, Vanaküla and several others, some of those with structural orientation. Gabbro-norite is a massive rock of coarse or medium grain size, which contains plagioclase (An 50–70), ortho- and clinopyroxene, hornblende and biotite of secondary origin. In northern Estonia, small granodiorite and quartz-diorite bodies have been found in single boreholes (Aruküla, Letipea). The quartz-diorite is orientated, medium‑grained and cut by veins of microcline granite. The small Utria body in the northeastern coastal area consists of massive medium‑grained gabbrodiorite.

The granite rocks of the Estonian crystalline basement are mainly migmatite granites: plagioclase‑microcline granites in northern and western Estonia, and charnockites and plagioclase‑orthoclase granites in the granulite facies area of southern Estonia. The charnockites consist of potassium feldspar, plagioclase (andesine) and, to a lesser extent, of biotite, hypersthene and hornblende, the latter three forming 5‑10 % of the rock. Quartz, microcline, oligoclase‑andesine and biotite are the main minerals in the migmatite granites of northern and western Estonia.


Svecofennian post‑orogenic plutonic rocks

In southern Finland, the 1.82–1.78 Ga tonalitic to monzonitic and granitic post‑orogenic intrusions are neither voluminous nor numerous (Koistinen 1996). However, they mark the final stage of the Svecofennian orogeny when the temperature of the crust was still high. Recently, a group of post‑orogenic granites was identified in the Estonian basement as well (M. Niin, unpublished report).

The Taadikvere body in Central Estonia consists of granodioritic – quartz-monzonitic rocks which are of preferred orientation and contain plagioclase and potassium feldspar phenocrysts. The medium-grained groundmass of the rock comprises quartz, plagioclase (An 32–36), potassium feldspar, biotite and hornblende.

The Virtsu body in western Estonia, consists of porphyritic rocks of quartz monzonite composition that are strongly crushed within the west-east oriented central Estonian cataclastic zone. The medium-grained groundmass of the rock consists of plagioclase (An 31–42), potassium feldspar, quartz and biotite with some admixture of hornblende, and numerous potassium feldspar and plagioclase phenocrysts, up to 2–3 cm in diameter.


Palaeoproterozoic metamorphism

The mineral paragenesis of metamorphic rocks and, partly, of early orogenic plutonic rocks is due to regional metamorphism which in the Svecofennian orogen in Finland occurred in several stages during 1.885–1.81 Ga as dated by isotopic studies (Koistinen 1996).

Metamorphic zoning (Fig. 7) is typical of the Svecofennian orogen. In the province as a whole, andalusite‑muscovite mica schists prograde into potassium feldspar‑sillimanite gneisses and migmatitic garnet‑cordierite gneisses. The neosomes in the migmatites differ markedly in type. In central Finland the granitoid area is surrounded by tonalitic and trondhjemitic migmatites, while potassium granitic neosomes occur in migmatites of southern Finland (Koistinen 1996). Potassium granitic migmatites extend into the North-Estonian zones of amphibolite metamorphism. In the southern part of Estonia, charnockite and enderbite migmatites characteristic for granulite facies area have been described.

The Precambrian basement of Estonia consists of rocks which have been subjected to high-grade metamorphism. In northern Estonia (Figs. 8, 9), amphibolite facies gneisses are most abundant, while granulite facies mineral assemblages occur locally as in the Jõhvi and Tapa zones. Assemblages marking transition from amphibolite to granulite facies occur in the vicinity of Uljaste and Assamalla (Klein 1986). The amphibolite facies gneisses in Estonia serve as an extension to those spread in southern Finland. The dominant metamorphic grade here is a high- temperature amphibolite facies with a local PT zoning from sillimanite‑potassium feldspar subfacies to granulite facies (Uljaste, Haljala). Geothermobarometry, mostly of the biotite + garnet + /‑ sillimanite assemblage and of cordierite, estimates prograde metamorphism at 600–700º C and 3‑5 kbar.

In the Jõhvi Zone, there are characteristic granulite mineral assamblages in cordierite‑garnet gneisses (hypersthene) and in mafic gneisses (two pyroxenes and spinel). In the Tapa Zone, the traces of granulite metamorphism have probably been partly removed by high-temperature retrograde metamorphism.

In southern Estonia, granulite facies gneisses form a large domain (Fig. 7), which extends from the Middle-Estonian fault zone (Saaremaa‑Peipsi Zone) to northern Latvia and further south through the beltiform Belarussian‑Baltic granulite domain. The conditions of metamorphism have been mainly studied from two drillcores — Kõnnu 300 and Varbla 502 (Koppelmaa et al. 1978, Hölttä & Klein 1991). The mineralogy of granulites varies. The widespread garnet and cordierite, formed by breakdown of biotite and sillimanite, indicate prograde metamorphism. Hypersthene coexists with garnet and cordierite although, so far, the sillimanite‑hypersthene assemblage has not been found. The PT‑conditions have been calculated using several geother-mometers and geobarometers (Hölttä & Klein 1991). These give temperature estimates for the prograde stage of metamorphism of 700–800ºC and pressure estimates of 5–6 kbar or more.

The age relations between the described zones are still ambiguous. The southern Estonian granulite area has been correlated with those in Finland. In the Haukivesi‑Kiuruvesi Complex metamorphism has been dated at 1.88 Ga (Korsman et al. 1984). High-temperature metamorphism in southern Finland, 1.83–1.81 Ga ago (Korsman et al. 1984), may be correlated with similar metamorphism in northern Estonia.

It is emphasized that the southern Estonian granulites (Fig. 10) formed under higher pressure than is characteristic of Svecofennian metamorphism. This suggests that the South-Estonian region represents a deeper crustal section (Koistinen 1996).



Palaeoproterozoic to Mesoproterozoic – rapakivi and related rocks of the Fennoscandian province


A large time span for the Fennoscandian rapakivi and related rocks’ plutonism at 1.65–1.54 Ga was established by isotopic studies (Rämö et al. 1996). Recently, it was stated that the province consists of four subprovinces separated areally and differing in age (Puura & Flodén 1996). The large plutons have a central position in the subprovinces, while the stocks and dike swarms occur in peripheries of the subprovinces. Volcanic rocks have survived as remnants in the vicinity of the main plutons. In the basement of Estonia, plutonic and related rocks of two, Vyborg and Riga‑Åland subprovinces occur (Table 4).


Palaeoproterozoic rocks of the Vyborg Subprovince, the 1.62–1.67 Ga age group

The Vyborg Pluton in southeastern Finland and adjacent offshore area has a central position in the oldest rapakivi subprovince. Its southern satellite spreads in the bottom of the Gulf of Finland, near the northeast coast of Estonia (Koistinen 1994).

Characteristic structures of this subprovince in Estonia are stocks of porphyritic potassium granites, chemically only little differing from the proper rapakivi (Kuuspalu 1975, Puura & Flodén 1996). Presuming that also the westernmost and smallest known but undated Taebla Stock (Fig. 5) belongs to the Vyborg Subprovince, then practically the whole mainland Estonia was influenced by rapakivi magmatism at about 1.65–1.64 Ga.

The Naissaar Pluton (55km × 25 km), the northern part of which extends under the Gulf of Finland, is composed of porphyritic granites cut by aplites. According to the chemical and mineral composition, this pluton is divided into two phases (Soesoo & Niin 1992). The more melanocratic granites of the first phase form the periphery of the pluton. The central part of the pluton is composed of leucocratic granites of the second phase, which have some similarities with the second and third phases of the Märjamaa Stock. The structure of the rocks is massive, in places (second phase) trachytoidic.

There are two generations of quartz (25–35%): crystals within microcline phenocrysts and anhedral crystals in the groundmass. In places, tabular microcline (35–45%) contains inclusions of plagioclase. Euhedral to subhedral zoned plagioclase (20–25%) is of oligoclase composition. Biotite forms small flakes containing euhedral crystals of zircon and apatite as inclusions. Muscovite and fluorite of post-magmatic origin replace plagioclase in the second phase of the intrusion. Hornblende occurs sporadically in granites of the first phase. The other accessory minerals are apatite, titanite, zircon and epidote. The opaque minerals are represented by magnetite and ilmenite.

The Neeme Pluton (25km × 20 km), the northern part of which is under the Gulf of Finland, is composed of coarse- and medium-grained pinkish-grey porphyritic granites cut by aplites (Soesoo & Niin 1992). By chemical and mineral composition, the rocks form two groups, possibly two phases. Two small bodies in the central and northeastern parts of the intrusion are more melanocratic, their chemical composition being partially similar to granodiorite. Partially assimilated xenoliths of surrounding gneisses with a diameter of about 20–30 cm have been found in some drill cores. The structure of rocks is massive, in places trachytoidic.

Quartz (20–25%) is generally euhedral, smaller grains are anhedral. Microcline (35–45%) occurs in groundmass in the form of rare phenocryst, up to 3–5 cm in diameter. Plagioclase (15–25%) is represented by euhedral crystals of oligoclase-andesine composition. Biotite (2–10%) forms unhedral flakes that contain small crystals of apatite, titanite, fluorite and zircon as inclusions. Hornblende occurs sporadically. Muscovite, apatite, fluorite, titanite, zircon, epidote and opaques are accessory minerals.

The Ereda Pluton (5 km × 15 km) is composed of homogeneous pinkish‑grey coarse‑grained porphyritic granites (Soesoo & Niin 1992). As there are only two drill cores available, it is difficult to correlate the Ereda rocks with those of other stocks. The mineral composition and structure of the Ereda granites and the Märjamaa and Neeme leucocratic type of granites have some similar features.

Two generations of quartz (30–35%) have been distinguished. Microcline (35–45%) is present as tabular crystals; large phenocrysts with a dimater of 3–4 cm are zoned. Plagioclase (20–30%) forms various euhedral, tabular and prismatic crystals of andesine composition. Biotite (5–10%) is altered. The accessory minerals are fluorite, apatite, zircon, epidote and rutile. Magnetite and hematite opaques occur.

The Märjamaa Pluton (40 km × 25 km) is composed of coarse-grained pink-grey porphyritic granitoids, sometimes cut by aplites (Soesoo & Niin 1992). According to the geophysical and drilling data, the pluton has features of ring structure and is accompanied by a smaller satellite. The contacts between the granites and surrounding Palaeoproterozoic gneisses are sharp. The round central part of the Märjamaa composite stock is represented by the most melanocratic and basic type of porphyritic granodiorites (first phase). In places it contains gneiss xenoliths, up to 20 cm in diameter. The second, intrusive phase, is represented by biotite and hornblende-bearing granites. The granites of the third phase (possibly the small individual Kloostri Massif in the northwestern part of the Märjamaa Intrusive) are more leucocratic in composition and, in places, have a trachytoidic structure.

Two generations of quartz (20–30 %) have been distinguished. The first generation consists partly of anhedral crystals between potassium feldspar individuals and partly of euhedral inclusions within microcline. The second generation occurs as anhedral crystals in the groundmass. Microcline (20–40%) is present as phenocrysts (diameter about 2–3 cm) and in the groundmass. The phenocrysts are often perthitic and contain inclusions of quartz, biotite and rare, titanite. Plagioclase (20–40%), which forms euhedral tabular or prismatic crystals, is represented by oligo-clase‑andesine. Anhedral crystals of biotite (2–10%) are often clustered together as swarms of small or large flakes. Hornblende (mainly in the first and, partly, in the second intrusive phase) and muscovite (in the third phase) occur sporadically. The content of the main opaque minerals, magnetite and ilmenite may reach 3–5%. Apatite, fluorite, zircon, titanite and epidote are accessory minerals.

The Taebla body, 6–7 km in diameter, is the smallest one distinguished by drilling of two wells. It is composed of homogeneous leucocratic porphyritic granites (Soesoo & Niin 1992). In terms of the mineral and chemical composition, the Taebla granites are similar to the rocks of the third phase of the Märjamaa Pluton and to the second phase of the Naissaar and Neeme plutons.

Geophysical data indicate that the Abja body (8 km × 5 km), which occurs in the structural zone of South-Estonian granulites, is ellipse-shaped. In the only drill core available, in the depth interval 550-635 m, medium-grained greenish-grey gabbrodiorites (SiO2 49–52%) with massive texture are cut by fine- and medium-grained pinkish-red potassium granites. The isotopic age (U-Pb, zircon of the gabbrodiorites is 1635 ± 7 Ma (Kirs & Petersell 1994). The age of the intersecting granites is 1622 ± 7 Ma. The gabbrodiorites comprise relatively euhedral plagioclase (An 33–39, 40–50%), hornblende (10–20%) and biotite (10–20%) with minor quartz and potassium feldspar. The content of accessory and opaque minerals is quite high; the most important being apatite (2–5%), titanite (1%) and titanomagnetite (2–6%).

Geophysical data suggests that the Sigula body of mafic rocks is a NE-trending dike (1.5km × 4 km). The one drill core available to date (depth interval 223.2-316.6 m) reveals that the pluton consists of inequigranular dark-grey massive diabase (SiO2 47–49%) with ophitic structure (Puura et al. 1983). The relatively large prismatic plagioclase crystals provide the rock with a slightly porphyritic outlook. The amount of plagioclase (An 55–63) is remarkable reaching 50–60%. Other minerals are hornblende (8–10%) and clinopyroxene (8–10%), biotite, orthopyroxene, quartz and potassium feldspar (all < 5%). The content of apatite (2–5%) and, especially, titanomagnetite (7–10%) is noticeably high.


Mesoproterozoic rocks of the Åland‑Riga Subprovince, the 1.54–1.59 Ga age group

The largest, Riga complex rapakivi‑anorthosite pluton (Bogatikov & Birkis 1973, Kuuspalu 1975) spread in the basement of western Latvia, southwestern Estonian Archipelago in the Gulf of Riga and Central Baltic proper, and the Åland Pluton belong to the Åland-Riga Subprovince. Defined and supposed rapakivi bodies in the northern Baltic seabed occur near the West-Estonian Archipelago (Puura et al. 1992, Koistinen 1994). It has been mentioned that the most intense rapakivi plutonism area coincides with the junction of the Baltic proper with the gulfs of Bothnia, Finland and Riga (Puura & Flodén 1996).

The rapakivi plutons and stocks are characterised by changeable magnetic and stable density properties and, as a whole, they have a considerably massive internal structure. Thus, they are easy to identify and contour by geophysical mapping and drilling.

The Riga Pluton is an essential representative of bimodal rapakivi-anorthosite complexes. The mafic part of the pluton locates in its southern part, in southwestern Latvia. In the central part, on the Kurzeme Peninsula, a variety of both typical vyborgite- and pyterlite-like and even-grained granites occurs. In the central and southern parts of the Riga Pluton, also quartz mangerites, mangeritic granosyenites and quartz‑syenites occur among mangeritic granitoids (Bogatikov & Birkis 1973). All these rocks have been formed in deep levels of the crust. However, in cores of two wells penetrating into the Riga Pluton in southwestern Estonia, on the islands of Ruhnu (3.4 m of crystalline rocks) and Saaremaa (at Kuressaare Town, 28.4 m), the rocks are represented by subvolcanic granite porphyries (granophyres) with micropegmatite matrix (Kuuspalu 1975, Puura et al. 1983).

Exterior to the northern part of the Riga Pluton (Fig. 5), the Undva well penetrates into a suite of rapakivi‑related volcanic rocks (Puura et al. 1983). The lower part of the sequence consists of plagioclase porphyrites, and the upper part of quartz porphyries.

Plagioclase porphyrites have some similarity with the rapakivi-related, more felsic porphyrites on Hogland, and with Dala porphyrites from the Transscandinavian Igneous Belt in central Sweden. The rocks are dark grey or black, in places, with pink-shaded massive and dense rocks. Their groundmass is fine- or very fine-grained, with microophitic texture, and consists of plagioclase (An 50–65, 65–75%), clino- and orthopyroxene (15–25%), hornblende (<5%), biotite, titanomagnetite, hematite and apatite. Prismatic and tabular phenocrysts of plagioclase (An 45–70), with an average size of 4mm × 5 mm (occasionally 30-40 mm), are rare forming about 3–10%. In terms of the chemical composition, plagioclase porphyrites are similar to andesites.

The quartz porphyries are brownish-red or pink massive rocks. Their fine-grained groundmass consists of quartz (30–40%), feldspar (40–50%), opaque minerals (hematite and magnetite up to 15%), chlorite, apatite and glass, and they have granophyric, radially fibrous and spherolitic texture. Small rounded phenocrysts of dark grey quartz are about 3–4 mm in diameter and make up 3–10% of the rock. Prismatic phenocrysts of plagioclase (An 1–7) and microcline-perthite are a bit larger (diameter about 5–8 mm, rarely 10–20 mm); their quantity varies between 20 and 30%.

The quartz porphyries of the Undva Member (Table 3) differ from those on Suursaari (Hogland) Island in colour and texture of groundmass. They comprise less and smaller phenocrysts, but the content of opaque minerals and apatite is higher which makes them more similar to some Dala porphyries in the Transscandiavian Igneous Belt in central Sweden.




K. Mens & E. Pirrus


The Vendian (Vendian Complex) as an independent stratigraphic unit, probably in the category of system, was distinguished in the early 1950s by B. Sokolov (1952a, 1953). Its stratotype area is in the western part of the East-European Platform.

The Vendian of the stratotype area includes three subdivisions in the rank of regional series, which in ascending order are Vilchan, Volyn and Valdai (Keller & Rozanov 1979b). The Valdai regional series consists of the Redkino (below) and Kotlin (above) stages, of which only the latter occurs in Estonia and forms the lowermost part of the sedimentary cover overlying the Proterozoic crystalline basement.

The current stratigraphic scheme of the Estonian Vendian was accepted in 1976 at the Baltic Stratigraphic Conference in Vilnius (Table 5).


Kotlin Stage

The unit in the rank of stage was defined as the upper part of the Valdai Series corresponding to “Laminarites” Clay on the East-European Platform (Resheniya… 1965). The name Kotlin was proposed by Sokolov (Männil 1958a) after Kotlin Island in the eastern part of the Gulf of Finland. Mens and Pirrus (1974, 1980, Gnilovskaya et al. 1979, Keller & Rozanov 1979b) determined the present stratigraphic extent of the stage and worked out its classification for the East Baltic area.

The Kotlin Stage is widespread in mainland Estonia, lacking only in its southwestern part and in some local structures, including Assamalla and Uljaste (Fig. 11). Stratigraphically, the most representative and thickest sections are situated in northeastern Estonia. In a westerly direction, the sections thin out rather rapidly and change in lithology.

The lower boundary of the stage coincides with the base of the sedimentary cover in Estonia, and is easy to determine. Some complications occur if the core yield is low or the core is distorted. The upper boundary is clear in eastern and central Estonia where the overlying rocks contain glauconite and mineralized skeletal fossils. In the northwestern part of mainland Estonia and on Hiiumaa Island the boundary is less distinct due to the lithological similarity with the overlying Lower Cambrian rocks. Identification of the Kotlin Stage is most complicated in the sections west of Keila (Fig. 12) where the lower part of the sedimentary cover comprises light-coloured loose quartzose sandstones with occasional lenses or interbeds of compacted multicoloured argillaceous rocks. As the latter rock type is lacking in the overlying Cambrian beds, this part of the sequence is conditionally regarded as the Kotlin Stage.

The Kotlin Stage is represented by siliciclastic rocks which accumulated under cool and humid climatic conditions (Pirrus 1992). This extensive, high-order cycle of deposition covered three shorter cycles divided as successive Gdov, Kotlin and Voronka formations. Multicoloured sandy-silty sediments consisting of low maturity and poorly sorted detrital material accumulated at the beginning of the Kotlin depositional cycle (Gdov Formation). Upwards in the section, the coarse-grained red-coloured deposits change into grey-coloured clayey sediments which accumulated during the stable phase of the Kotlin depositional cycle (Kotlin Formation). The cycle ends with the reappearance of multicoloured sediments of high maturity (Voronka Formation).

In recent years, acritarchs as the most abundant and widespread fossil group in the deposits of the Kotlin Stage have underlain the subdivision and correlation of the Vendian rocks (Volkova et al. 1983). Acritarchs are represented by a taxonomically simple assemblage consisting mainly of representatives of the genus Leiosphaeridia (Paškevičiene 1980). Microfossils are accompanied by vendotaenids, of which Vendotaenia antiqua Gn. is most common, while Aataenia reticularis Gn. is rare (Gnilovskaya et al. 1979). Besides microfossils and vendotaenids, fragments of shapeless organic matter occur on the bedding surface. All the above-listed palaeontological finds occur in the rocks of the Kotlin Formation which have promoted their accumulation and preservation. In some sections in the northeasternmost part of Estonia (Meriküla, Sinimäe), the grey argillaceous rocks of the upper member of the Gdov Formation comprise acritarchs and organic matter of irregular form.

The Gdov Formation. Asatkin (1937) derived the name from the Gdov beds used as a division to denote the sandy strata between the “Laminarites” Clay and the crystalline basement in the northwestern part of the East-European Platform. The Gdov Formation is considered as the lower part of the Kotlin Stage accumulated during the initial phase of the Late Valdaian transgression over the northwesternmost part of the East-European Platform, including the present-day Estonia, Latvia, and the western part of the Leningrad Region.

In Estonia, the Gdov Formation rests immediately upon the crystalline basement and spreads in subsurface lying in the northern, eastern and central parts of the Republic. Its thickness ranges from 0.2 to 58.3 m (Fig. 11-203, 11-102). The Venevere (Fig. 11-86) drill core in the interval of 287–322.8 m has been selected as a hypostratotype for Estonia (Fig. 12). The formation prevalently consists of multicoloured sandstones of various grain-size. The uppermost and, locally, also the lowermost part comprises a considerable quantity of reddish and purplish argillaceous rocks. The sandstones are represented by arkose and feldspatic varieties comprising besides quartz up to 50% of feldspars. Micas, both muscovite and green altered biotite, are occasionally found. The mineral composition of the clay fraction is rather stable throughout the formation, being characterised by illite-kaolinite suite (Pirrus 1970). On the basis of lithological features, the Gdov Formation is subdivided into three members which in ascending order are Oru, Moldova and Uusküla (Table 5).

The Oru Member occurs locally in the base of the Gdov Formation. It accumulated in depressions of the crystalline basement on account of its weathering crust. The greatest thickness of the Oru Member (6.7 m) has been recorded in the Jaama borehole (Fig. 11-104). The member consists of red unsorted clayey-sandy-gravely deposits (mixtite). Unlike the overlying members, the deposits of the Oru Member differ both structurally and mineralogically. Quartz is the prevailing mineral in sand and gravel (up to 90%); its grains are angular or subangular with nonsorted size distribution. Feldspar is not common (less than 10%). The clayey matrix consists mostly of kaolinite. All this suggests that these deposits were formed in deluvial fans as a result of intense weathering of acid crystalline rocks.

The Moldova Member overlies either the Oru Member or the crystalline basement. Its thickness is usually 30–40 m, maximum 50 m. The member consists of yellowish or pinkish arkose and/or feldspatic sandstones of various grain size and a few interlayers of multicoloured, frequently reddish, argillaceous rocks. Hence, the clastic material transported into the sedimentary basin originates from a close-lying area with a low degree of weathering of rocks.

The Uusküla Member is the most fine-grained and multicoloured part of the Gdov Formation. It consists of silty claystones intercalated with silt- and sandstones. The number and thickness of claystone layers increases eastwards. In the same direction the rocks gradually loose the red colour until in the easternmost sections they are predominantly grey. The proportion of sandstones is low and they are mainly represented by arkose and feldspatic types. Micas occur in remarkable quantities, but rock fragments are uncommon.

The composition and structure of the rocks suggest a rather low hydrodynamic energy of the sedimentation basin.

The Kotlin Formation. The name was introduced by Sokolov to designate the “Laminarites” Clay (Männil 1958a). The formation is spread in eastern Estonia, in a more typical form in its northeastern part attaining a thickness of 52.6 m in the Narva borehole (Fig. 11-33). In the west direction the thickness decreases quickly pinching out on the Tapa - Ellavere line (Figs. 12, 13). The formation is known only from core sections, and the interval of 109–150 m of the Meriküla core (Fig. 11-32) has been defined as the hypostratotype. The lower boundary is drawn at the level where multicoloured deposits turn grey. At the base of the formation, gravel and coarse-grained sand occur locally.

The dominant components of the formation are thinly laminated grey claystones with intercalating light-coloured very fine-grained sandstones or siltstones, or both. The lamination is complicated by the occurrence of dark-brown films of organic matter.

The rocks of the formation are low in sand and silt, the content of which in the upper- and lowermost parts only locally exceeds 50%. Quartz and feldspars (particularly K-feldspar) are the main detrital minerals of sand and gravel grain-size. The content of micas, including biotite and muscovite, is also notable. Their ratios depend on the type of rock. The rocks are characterised by a small content of both opaque and transparent allogenic minerals. Heavy minerals are dominated by siderite and pyrite of authigenic origin. Illite is a dominant clay mineral, the content of kaolinite ranges from 15 to 40%, the latter value being fixed in the lowermost part. Chlorite is common, in the middle part of the formation its average content is 15–20% (Pirrus 1970).

 On the basis of the lithological composition, the Kotlin Formation is divided into three members (Table 5).

The Jaama Member is made up of alternating grey-coloured massive siltstones and thinly laminated claystones. A few siderite nodules and organic matter films occur. This is the first stage in the large-scale clay accumulation in the eastern part of Estonia.

The Meriküla Member is the most typical unit of the Kotlin Formation. It is represented by the “Laminarites” Clay consisting of rhytmically alternating 0.5–0.8 mm thick pairs of dark-grey fine-dispersed clay layers and light-grey laminae higher in silt. The bedding plane is covered by dark-brown shapeless organic films. Vendotaenides, small flakes of mica and siderite nodules are common.

The fairly uniform mineral composition shows that the source areas must have been located relatively far from the depositional basin.

The Laagna Member has the most restricted distribution area, compared to other members of the formation. In the northeasternmost part of Estonia it is up to 6 m thick. The member consists of grey clayey and silty-clayey argillaceous rocks with many up-to-20-cm-thick intercalations of siltstone. The typical “Laminariates” Clay layers are absent. Scarce organic matter films and small siderite nodules occur suggesting the terminal phase of clay accumulation under weak hydrodynamic conditions.

The Voronka Formation was established by Mens and Pirrus (1971). Earlier, this part of the sequence was treated as two lower units of the post-Laminarites Sandstone or as the lower and middle parts of the Lomonossov Formation (Mardla et al. 1968). The type section of the formation is an outcrop on the lower reaches of the Voronka River, Russia (Mens & Pirrus 1971). Beyond the stratotype area, the formation is of subsurface occurrence being known in eastern and northern Estonia and in eastern Latvia. The Meriküla (Fig. 11-32) drill core in the interval of 90–109 m serves as a hypostratotype for the Voronka Formation (Mens & Pirrus 1980). The formation occurs between the overlying Lontova Formation and the weathering crust of the underlying Kotlin Formation (Mens & Pirrus 1969, 1970). In Estonia, the thickness of the formation ranges from 10 to 40 m. The Voronka Formation consists of variable siliciclastic rocks and represents a single upwards coarsening cycle from argillaceous rocks to well-sorted sandstones. The lower boundary of the formation is drawn on the basis of the change in colour. Based on lithological evidence, the formation is divided into the Sirgala and Kannuka members (Table 5).

The Sirgala Member consists of alternating multicoloured clays and siltstones with interlayers and lenses of light-coloured sandstones, the share of which increases upward the section. Most of detrital grains are subrounded quartz with a small quantity of feldspars (up to 10%) and micas (mainly muscovite). In the clay fraction, kaolinite slightly prevails over illite. Chlorite is uncommon.

The mineral composition suggests that these deposits derived from the weathered zone of sedimentary rocks.

The Kannuka Member consists entirely of light weakly cemented fine- to medium-grained quartzose sandstones with a few thin interlayers of multicoloured clayey siltstones, which are similar to the underlying deposits of the Sirgala Member. The clay fraction is dominated by kaolinite. Increase in the maturity in minerals upward the section is indicative of the redeposition of older sediments.




K. Mens & E. Pirrus


Cambrian rocks are widespread in Estonia. They are missing on the crest of the Valmiera-Lokno swell and on some peninsulas on the southern coast of the Gulf of Finland. Exposed Cambrian rocks are encountered in outcrops along the Baltic Klint, but mostly they are overlain by younger rocks and the basic data for studies has been obtained by means of boreholes.

The main pioneering work towards the subdivision of the Estonian Cambrian (then the Lower Silurian) was done by Eichwald (1854a) and Schmidt (1888) who worked out the lithostratigraphical subdivision and described the first fossils found in these rocks. A zonal division and modern nomenclature were introduced by Öpik (1933, 1956).

Up to the middle of this century, the Cambrian stratigraphy was based on outcrop sections and embraced the lowermost part of the Lower Cambrian succession of the studied area and the problematic Acrotreta Zone of the Upper Cambrian (Öpik 1956).

During the last fifty years, numerous deep borings were made which revealed the full thickness of the Cambrian rocks. Elaboration of Cambrian stratigraphy was greatly promoted by identification of plant microfossils (now known as acritarchs) by Naumova, Timofejev and Volkova on the East- European Platform. Palynological studies provided valuable data for establishing distinct acritarch assemblages, their ranges in the sequence and relationship with trilobite zones.

Elaboration of the present-day stratigraphical subdivision of the Estonian Cambrian (Table 6) was favourably influenced by international cooperation with researchers from neighbouring countries, and supported by IGCP projects No. 29 and 86. The results obtained were summarized in the stratigraphic scheme accepted in 1976 at the Baltic Stratigraphic Conference in Vilnius (Resheniya… 1978) and improved in 1983 by the Stratigraphic Conference on the Cambrian of the East-European Platform (Resheniye… 1986).

As the global standard is still under preparation, the boundaries between the Lower/Middle and Middle/Upper Cambrian are not strictly formal. There are no generally agreed names for the stages and their boundaries are unclear.

The Cambrian stratigraphic scale in the East-European Platform from the base of the Sabellidites cambriensis Zone to the top of the Acerocare Zone is based mainly upon the succession of trilobites, except the lowermost part where trilobites are lacking (Mens et al. 1990). Only part of the Cambrian is present in Estonia (Table 6). The lower boundary of the system is distinct in the studied area, and coincides with regional changes in the sedimentary conditions which led to the accumulation of normal marine sediments (Pirrus 1993). This level is marked by the appearance of primitive skeleton-forming organisms and changes in the composition of ichno- and phytofossils.

The upper boundary of the system is not obvious although, biostratigraphically, the Cambrian‑Ordovician transition in Estonia is relatively well studied (Mens et al. 1993). This is due to the circumstance that the IUGS has not yet passed the final decision on the Ordovician lower boundary.

Conventionally, the Cambrian is subdivided into three subsystems: Lower, Middle and Upper. The Lower/Middle Cambrian boundary is at the base of beds with Paradoxides, more exactly Eccaparadoxides insularis, and the Middle/Upper Cambrian boundary is at the base of the Agnostus pisiformis Zone for the East-European Platform (Table 6).

Rocks of all three subsystems are encountered in Estonia, but the degree of completeness varies. Compared to other subsystems, the Lower Cambrian rocks are most widespread and thickest. Their two-folded structure results from remolding of the basin prior to the Liivi transgression. Of the six regional stages established on the ground of the succession of acritarch assemblages in the Lower Cambrian on the platform, four are present in Estonia (Table 6).

The Middle Cambrian sequence in Estonia is entirely devoid of fossils and the regional stages have been established on the basis of lithological criteria.

The Upper Cambrian is documented on the basis of palaeontological evidence, derived from both the shelly fauna and acritarchs.

No regional stages (except the Kybartai and Deimena in the lowermost Middle Cambrian) have yet been differentiated in the rest of the Middle and through the Upper Cambrian in Estonia. The relevant rocks have been treated only in general lines by lithostratigraphic units.

Based on the stratigraphical completeness of the sections and facies conditions, Estonia’s territory is subdivided into the northern, western and southeastern regions (Brangulis et al. 1974, 1975, Table 6), with their characterisitic formations and members.


Lower Cambrian

Lontova Stage

The oldest Cambrian rocks in Estonia were formed during the Baltic evolutionary stage in the pre-trilobite Early Cambrian (Mens 1981c). The onset of sedimentation in the Early Cambrian in Estonia corresponds to the Platysolenites antiquissimus Zone defined as the Lontova Stage. The Rovno Stage, composed of the lowermost rocks of the Baltic evolutionary stage, is lacking in Estonia (Table 6).

The name Lontova was introduced by Öpik (1933) in the rank of beds to designate the “Blue Clay” proper. It corresponds to the upper part of the blue clays by Schmidt (1888) and Mickwitz (1911), the Lontova beds by Öpik without the uppermost layers with Volborthella tenuis (Öpik 1933, 1956) and to the Platysolenites antiquissimus Zone in the current use (Mens et al. 1990).

The rocks of the Lontova Stage crop out at the foot of the Klint and extend as a narrow belt from Tallinn to Narva. The main localities are the quarries at Kopli, Tammneeme, Kolgaküla, Kunda and Aseri (Fig. 14).

The stratotype of the stage is the Kunda quarry (Öpik 1933), subsequently complemented by the Lontova drill core in the interval of 14.0 to 88.3 m (Mens & Pirrus 1977). The stage in the stratotype section is incomplete since the deposits of the regressive phase of the Baltic sedimentary cycle are lacking. The stratigraphical completeness and thickness of the stage varies with regions: the rocks are at their thickest (ca 90 m) in northeastern Estonia (Fig. 14) and thin in a southerly direction due to post-sedimentary denudation. In Estonia, the Lontova Stage overlies, with a break in sedimentation, the Kotlin Stage. Its lower boundary, known only from core sections, coincides with the appearance of typical marine sediments in the succession (Mens & Pirrus 1977, Pirrus 1993).

The Lontova Stage is represented by siliciclastic rocks with a clear lateral variation of the ratio of rock types. Argillaceous rocks are prevailing in eastern and central Estonia, while sandstones dominate west of the Vihterpalu ‑ Häädemeeste line (Fig. 14).

A relatively diverse assemblage of skeletal fossils containing Sabellidites cambriensis Yan., S. sp., Platysolenites antiquissimus Eichw., P. lontova Öpik, P. spiralis Posti, Yanichevskyites petropolitanus (Yan.), Aldanella kunda (Öpik) together with pyritized casts of hyolithids and hyolitelmintes, hornlike chitinous (?) sklerits, fragments of brachiopods and agglutinated tubes of Onuphionella has been identified from the Lontova Stage (Mens & Pirrus 1977, Mens & Posti 1984). Some of the above-mentioned species like P. antiquissimus, Y. petropolitanus and casts of hyolithids occur throughout the stage, whereas the vertical range of the rest is more limited. On the basis of the earliest appearances of the index taxa, the Lontova Stage is subdivided into four parts, which in ascending order are the Sabellitides cambriensis beds, Platysolenites lontova beds, Aldanella kunda beds and P. spiralis beds (Mens & Posti 1984).

The distribution of some species shows a distinct facies control. Thus, the hornlike chitinous (?) sklerites have been found in the argillaceous rocks of the eastern part of the territory only. The tubes of Platysolenites and Yanichevskyites are rather rare in the well-sorted sandstones in the western part of the studied area.

Acritarchs from the Lontova Stage have been described by several investigators (Naumova 1960, Volkova 1968, 1973; Jankauskas & Posti 1973, a.o.), who all agree that the stage has an acritarch assemblage of its own, which contains besides leiosphaerides and tasmanites also marginats forms. The frequency and diversity of acritarchs in the assemblage clearly depend on palaeoenvironmental conditions and facies changes in the basin of sedimentation (Mens & Paškevičiene 1981).

Trace fossils from the Lontova Stage are diverse and comprise numerous ichnospecies, among them Phycodes pedum Seilacher (Palij et al. 1983).

In conformity with the ratio of rock types in the succession, two formations lateratelly replacing each other, have been distinguished in the Lontova Stage (Kala et al. 1981b).

The Lontova Formation was identified in the rank of formation by Männil in 1958 (Männil 1958a). Its type section in the Kunda quarry has been selected for the stratotype of the Lontova Stage (see above). The formation occurs in northern, eastern and central Estonia (Fig. 14), and is westwards laterally replaced by the Voosi Formation. The Lontova Formation is represented by greenish-grey and variegated argillaceous rocks with interbeds of coarse- to fine-grained sandstone in the lowermost and fine-grained sandstones in the uppermost part. The formation is subdivided in ascending order into the Sämi, Mahu, Kestla and Tammneeme members (Kala et al. 1970, Mens & Pirrus 1977).

The Sämi Member consists of alternating sandstones and argillaceous rocks containing glauconite and, occasionally, also flattened phosphatized pebbles.

The Mahu Member is made up of greenish-grey sandy or silty claystones with thin interlayers of sandstone.

The Kestla Member is characterized by homogenous multicoloured claystones with greenish‑grey, reddish‑brown and purplish interbeds, strips and spots. The admixture of sandy material is limited.

The Tammneeme Member consists of fine‑grained sandstones and greenish‑grey claystones and occurs in a limited area (Fig.15).

The Voosi Formation was defined on the basis of lithological evidence. Sandstones account for more than 50% of its composition (Kala et al. 1981b). Its stratotype is the interval of 237.5 to 300 m in the Haapsalu‑3 drill core (Fig. 14-124)

The formation is distributed in the northwestern part of Estonia; in a easterly direction it is gradually replaced by the Lontova Formation. Its lower part extends farther to the east (Fig.16). The thickness of the formation ranges from 72 to 14.6 m.

The formation consists mostly of quartzose sandstones which is the dominant type of rock on the islands of the West-Estonian Archipelago. Claystones are of minor importance and associate mostly with the upper part of the formation in mainland Estonia.

The formation is subdivided in ascending order into the Taebla, Kasari and Paralepa members.

The Taebla Member consists of light fine-grained sandstones with a few thin interbeds of silty claystones. Glauconite is not common.

The Kasari Member is represented by sandstones of various grain-size and with limited claystone content. The sandstones are rather rich in glauconite and in some places flat phosphatized pebbles occur. The sandstones of the Kasari Member are quite similar to those in the Sämi Member of the Lontova Formation.

The Paralepa Member consists of interbedded greenish‑grey (with a few purplish spots) argillaceous rocks, mainly silty claystones, and of fine‑grained sandstones.


Dominopol’ Stage

According to the currently accepted correlation (Resheniye… 1986), the lowermost part of the trilobite-bearing Cambrian on the East-European Platform, deposited during the Liivi evolutionary stage in the East Baltic area (Mens 1981c), is defined as the Dominopol’ Stage.

Previously, this unit in the same stratigraphical extent was referred to as the Lükati Stage (Aren et al. 1975, Keller & Rozanov 1979a), as the Talsy (= Lükati) Stage (Keller & Rozanov 1979b) or as the Talsy Stage (Birkis et al. 1970, Brangulis et al. 1981). It should be pointed out that the name Lükati in the rank of stage is also used in a more restricted stratigraphical extent (Mardla et al. 1968, Mens 1986).

The Dominopol’ Stage was distinguished by Kirjanov (Kirjanov 1969) with the stratotype section in the interval of 617.2 to 747 m of the Berezhki 2944 drill core situated in Volyn, the western Ukraine.

In Estonia, the Dominopol’ Stage is represented by three succeeding formations: Sõru, Lükati and Tiskre (Table 6) well recognizable on the basis of lithological and palaeontological evidence. The Dominopol’ Stage occurs on the islands of the West- Estonian Archipelago (except Ruhnu) and in the western, northern and central parts of mainland Estonia (Fig. 17). Only the upper part of the stage (Lükati and Tiskre formations) is exposed along the North-Estonian Klint. The outcrop extends from the Pakri Peninsula in the west up to the Narva River in the east. The main localities are Türisalu, Rannamõisa, Kakumägi, Lükati (Photo 15), Saviranna, Kunda, and Utria (Fig 17). The maximum thickness of the stage (76.6 m) has been fixed in the Kalana borehole where all three formations occur.

The stage consists of siliciclastic rocks, mainly sandstones. The lower boundary of the stage is lithologically distinct, accompanied by a change in the faunal composition and marked, in some places, by lenses of conglomerate.

The palaeontological finds in the lower part of the stage (Sõru Formation) are scarce. These are mostly trace fossils, rare shells of agglutinated foraminifers and a few acritarchs. The latter are represented by leiosphaerides and rare Globosphaeridium cf. cerinum (Volk.), Asteridium pallidum (Volk.), Loposphaeridium tentativum Volk. and Tasmanites bobrowskae Waz. (Mens 1986). This part of the stage corresponds to the Rusophycus parallelum Zone.

The middle part of the stage (Lükati Formation) is palaeontologically well characterized. The most typical species include Volborthella tenuis Schm., V. conica Schindewolf, Schmidtiellus mickwitzi (Schm.), Mickwitzia monilifera (Linnars.), torellellids, hyolithids, agglutinated foraminifers (“Lycatiella”), and in the west Platysolenites antiquissimus Eichw. has been identified in the basal beds. The acritarch assemblage in this part of the stage is abundant and diverse, with baltisphaerids dominating (Mens & Pirrus 1977). The middle and upper parts of the stage correspond to the Schmidtiellus mickwitzi Zone.

Of the upper part of the stage (Tiskre Formation), only its lower part (Kakumägi Member) is palaeontologically well characterized. It contains, particularly in the conglomerate lenses, Mickwitzia monilifera (Linnars.), M. formosa Wiman, M. concentrica Gorjansky, Paterina rara Gorjansky, Scenella discinoides Schm., S. tuberculata Schm., Bradoria? estonica Melnikova, Konicekion kundaensis Melnikova and fragments of trilobites. In the uppermost part (Rannamõisa Member), only occasional indeterminable fragments of brachiopods and trace fossils occur. Acritarchs have been identified in the Kakumägi Member, and only once they were found in the drill core of the Rannamõisa Member (Muraste-2 borehole). Its assemblage is much the same as in the Lükati Formation, except the appearance of Tasmanites piritaensis Posti et Jank.

Based on the differences in the palaeontological composition and distribution area as well as clear contacts between the three formations of the Dominopol‘ Stage, these parts of the sequence are considered independent stages (Mens 1986).

The Sõru Formation, resting transgressively either on the Lontova Formation or on the crystalline basement, occurs in the northwestern part of mainland Estonia and on the islands of the West-Estonian Archipelago, except Ruhnu (Fig.18).

Rocks of this formation are known only by core sections. The thickness of the formation ranges from 6.2 m at Vihterpalu to 58.2 m at Eikla (Fig.17-40, 178). The Tahkuna drill core (Fig. 17-34) in northern Hiiumaa in the interval of 100.5 to 147 m has been selected as the type section of the Sõru Formation (Resheniya… 1978, Kala et al. 1984a).

The lower part of the formation consists mostly of massive fine-grained quartzose-feldspatic sandstone with thin clay interbeds and films. The upper part is represented by a complex of interbedded argillaceous rocks and sandstones. In both parts the rocks are light-grey with greenish-grey shade, but red and purplish-red patches also occur .

The Lükati Formation, the most widespread division of the Dominopol’ Stage (Figs. 18, 19), lies transgressively on the Sõru Formation in the west and on the Lontova Formation in the east. It is often separated from the underlying units by conglomerate lenses containing pebbles of phosphatized sandstones (Mens & Pirrus 1975). Outside the distribution area of the Tiskre Formation, the upper part of the Lükati Formation is often weathered. The Lükati Formation, formed during the stable phase of the Liivi evolutionary stage, corresponds to the whole stratigraphical extent of the Dominopol’ Stage along the southern margin of its distribution area in Estonia.

In the type area in the vicinity of Tallinn, the formation reaches 20 m in thickness. The type section of the formation is an outcrop on the left bank of the Pirita River (Photo 15) in the eastern outskirts of Tallinn (Öpik 1933), complemented today by a drill core from the lower part of the formation (Mens & Pirrus 1977).

The lithologically monotonous formation consists of interbedded greenish-grey argillaceous rocks and very fine-grained sandstones. The latter form mostly 0.1—0.3-m-thick layers, some of which are hard, in places tightly cemented by poikilotopic carbonate.The upper surfaces of the sandstone layers are covered with ripple marks, the lower surfaces with casts of various trace fossils and mud cracks. Pyrite and glauconite are very common. In the lower part, glauconite often forms 1 - 3-cm-thick laminae.

The Tiskre Formation in the current use is interpreted as a unit containing the Tiskre beds (= Diplocraterion Sandstone) by Öpik (1933) or the Tiskre Formation by Männil (Männil 1958a) together with the Kakumägi beds (Scenella Zone) by Öpik (1933) or the Kakumägi Member of the Pirita Formation by Männil (Männil 1958a).

The Tiskre Formation is distributed in northern and western Estonia (Figs. 18, 19) where it overlies the Lükati Formation. The lower boundary is lithologically abrupt and marked by a change from argillaceous rocks to sandstones. Between Muraste and Aseri, lenses of Mickwitzia conglomerate occur at that level. Its type section is an exposure at the southern end of the Rannamõisa Cliff, 14 km west of Tallinn (Mens & Pirrus 1977). The formation is at its thickest (20 m) in the stratotype area.

The Tiskre Formation consists of light-coloured massive or thick-bedded sandstones with thin interbeds of greenish-grey argillaceous rocks (Photo 16). Within the outcrop belt the Tiskre Formation is divided into the Kakumägi (lower) and Rannamõisa (upper) members.

The Kakumägi Member is represented by poorly sorted sandstones with an admixture of clayey material. Conglomerate lenses within the sandy sequence are locally present. In the basal part, sandstones are often well cemented with dolomitic cement. Bedding is mostly lenticular, casts of mud cracks, slump-rolls and ripple marks are common.

The Rannamõisa Member consists of horizontally bedded winnowed fine-grained or very fine-grained sandstones with thin interlayers of argillaceous rocks. Glauconite is present. Ripple marks and convolute bedding occur throughout the member (Pirrus 1978).


Ljuboml’ Stage

The succeeding Aisčiai evolutionary stage terminates the Early Cambrian sedimentation in Estonia, and embraces deposits of the Ljuboml’ and Vērgale stages (Table 6). The former was earlier treated as a part of the Vērgale Stage (Mens et al. 1990) or as the Lower Vērgale Substage (Resheniye… 1986). Kirjanov (Kirjanov 1969) distinguished it in the rank of an independent stage and also as a formation with the stratotype section in the interval of 543.5 to 617.2 m of the Berezhki‑2944 drill core in Volyn, the western Ukraine.

The Ljuboml’ Stage is represented in western Estonia by the Soela Formation and in central and eastern Estonia by the Vaki Formation (Resheniya… 1978, Mens 1979, Kala et al. 1984a), and is known only from core sections. Since the boundary between the Ljuboml’ Stage and the overlying Vērgale Stage is difficult to determine, a map showing the distribution of the late Lower Cambrian rocks in Estonia was compiled jointly for these units (Fig. 20).

The Ljuboml’ Stage lies with a stratigraphic unconformity on the rocks of the Dominopol’ or Lontova Stage or on the crystalline basement (Ruhnu Island, borehole 257). Accordingly, the lower boundary is expressed variously. It is well recognizable in the areas where the underlying strata consist of crystalline or argillaceous rocks. In the latter case, the topmost part of the Lontova or Lükati Formation is often weathered. Identification of the lower boundary is most complicated in the sections where the Tiskre Formation of the Dominopol’ Stage and the Soela or Vaki Formation of the Ljuboml’ Stage occur simultaneously, because they are lithologically very similar and extremely poor in fossils. In that case, the lower boundary of the stage is tentatively drawn at the level of the essential change in the composition of detrital minerals (Mens 1979). Due to the general lack of fossils, the boundary between lithostratigraphical units is conditionally taken for the upper limit of the stage which is placed at the base of the Irben Formation. Since the boundaries of the stage have not been firmly fixed, its thickness is difficult to determine, but it is mostly less than 40 m.

Fossils are extraordinarily rare. Occasionally, casts of Volborthella, undeterminable fragments of inarticulate brachiopods, valves of agglutinated foraminifers and some ichnofossils, mostly of the genus Skolithos are encountered. Acritarchs have been found from the Soela Formation in the Varbla (Fig 20-188) drill core (440.1 m) and from the Vaki Formation in the Oostriku‑700 (Fig. 20-155) drill core (256.2 m). The Soela Formation contains an abundant and diverse acritarch assemblage prevailed by leiosphaerids. The occurrence of Goniosphaeridium varium (Volk.) and Skiagia ciliosa (Volk.) among acantomorphids suggests late Early Cambrian age (Keller & Rozanov 1979a). The acritarch assemblage from the Vaki Formation is pauperated, beside leiosphaerida it contains a few Goniosphaeridium volkovae Hagenfeldt and Comasphaeridium latviense (Volk.), and this unit can have a wider biostratigraphic bracketing than the Soela Formation.

The Ljuboml’ Stage is regarded as corresponding to the Holmia inusitata Zone.

The Soela Formation as an independent stratigraphic unit was distinguished recently (Mens 1979, Kala et al. 1984a). Earlier, this part of the sequence was treated either as the upper part of the Tiskre Formation (Kala 1972, Keller & Rozanov 1979b) or the lower part of the Kurzeme (now Irben) Formation (Mens & Pirrus 1972).

The type section is in the interval of 230.7 to 263.7 m of the Emmaste drill core (Fig 20-117), Hiiumaa Island. The formation occurs on the islands of the West - Estonian Archipelago and in the western part of mainland Estonia (Fig. 21). It is known only from core sections, and its lower boundary coincides with the lower boundary of the Ljuboml’ Stage (see above). The upper boundary is lithologically clear, only east of the Koluvere ‑ Rumba ‑ Pärnu line it is somewhat debatable. The formation lies transgressively on the Lower Cambrian rocks or the crystalline basement (Fig. 20-257). The thickness varies from 7.9 to 41.8 m, increasing from north to south.

The formation consists of weakly cemented light‑coloured fine‑grained feldspatic (subarkose) sandstones, containing up to 10% of coarse sand and gravel grains. In the lowermost part of the section, mainly beyond the area of distribution of the Tiskre Formation, interbeds of greenish‑grey siltstones and argillaceous rocks occur. Pebbles of greenish‑grey clayballs and cross‑bedding marked by mica flakes and glauconite grains are characteristic of the lower part of the formation.


Vērgale Stage

The Vērgale Stage, embracing the topmost Lower Cambrian in Estonia, was established in the present stratigraphical extent by Birkis et al. (1970). Previously, it had been used in a wider extent (Keller & Rozanov 1979b, Resheniye… 1986). The Vērgale‑46 drill core in the interval of 1293 to 1318 m has been selected as the stratotype section of the stage (Birkis et al. 1970, Brangulis 1989). Only the lowermost part of the stage occurs in Estonia. The Vērgale Stage is of subsurface distribution on the islands of Hiiumaa and Saaremaa and in the western part of mainland Estonia (Fig. 20). Its thickness decreases northwards being about 40 m at Seliste and less than 1 m at Tahkuna (Fig. 20-225, 34).

The lower boundary of the stage is tentatively drawn at the level of the base of the Irben Formation, near the level of the appearance of the Vērgale acritarch assemblage. It is less expressed beyond the distribution area of the Irben Formation where the whole post‑Liivi sandy Lower Cambrian succession has been distinguished as a joint Ljuboml’‑Vērgale unit or as the Aisčiai Group.

The faunal record of the Vērgale Stage includes Volborthella, agglutinated foraminifers, fragments of trilobites and inarticulate brachiopods, also ichnites are very common. The so-called Vērgale acritarch assemblage comprises Estiastra minima Volk., Skiagia ciliosa (Volk.), S. insigne (Fridr.), S. compressa (Volk.), S. orbiculare (Volk.), Comasphaeridium strigosum (Jank.), Asteridium spinosum (Volk.), A.lanatum (Volk.), A. tornatum (Volk.), Tasmanites volkovae Kirjanov, T. bobrowskae Waz., T. tenellus Volk., Dictyotidium priscum Kirjanov & Volk., Leiovalia tenera Kirjanov, Pterospermella solida Volk., Lophosphaeriduim truncatum Volk., Alliumella baltica Vanderflit, etc. with several variations in the species composition from section to section. On the ground of the palaeontological evidence, the stage corresponds to the Holmia kjerulfi Zone.

The Irben Formation (Kala et al. 1984a) was earlier termed the Kurzeme Formation (Mens & Pirrus 1972). The stratotype of the formation is the type section of the previous Kurzeme Formation represented by the interval of 1329 to 1387 m of the Pavilosta drill core in Latvia (Lieldiena & Fridrichsone 1968).

The Irben Formation is distributed on the islands of the West-Estonian Archipelago and in the western part of mainland Estonia (Figs. 19, 20), and is known only from core sections. It rests conformably on the Soela Formation having the maximum thickness (42.4 m) in the Seliste drill core (Fig. 20-225). The lower boundary of the formation is marked by the appearance of argillaceous rocks. The upper boundary of the formation is erosional throughout its distribution area in Estonia.

The formation consists of interbedded clay‑ and siltstones with interlayers of fine‑grained sandstones, the number and thickness of which increase eastwards. Brown ferruginous oolith interbeds of goethite oolites are characteristic of the formation; in Estonia, these have been found only in the westernmost sections (Figs 19, 21). Owing to the unique lithology and wide distribution of argillaceous rocks, it is an important level for regional lithostratigraphical correlation (Pirrus 1986). Argillaceous rocks are greenish‑grey, but in the upper part they are locally dark-grey with a purplish or brownish shade of colour. Besides the wavy horizontal bedding, there are abundant ichnites in clay intervals responsible for a bioturbated structure of the “kråksten” type.

Silt- and sand-size fractions contain quartz (up to 85%), feldspars (up to 20%) and micas (usually 2‑10%, rarely up to 25%). Like in the Soela Formation, the content of heavy minerals is low. Clay minerals are dominated by illite.

The Vaki Formation occurs in central and eastern Estonia (Table 6), being probably a shallow‑water equivalent of the Soela and Irben formations. It is known only from core sections (Fig. 21). The formation rests with a stratigraphic unconformity on rocks of the Dominopol’ or the Lontova Stage. In the latter case, or if the Tiskre Formation is absent, the topmost layers of the underlying units are weathered (Mens et al. 1984).

The Vaki‑67 (Fig. 21 - 148) drill core in the interval of 284.4 to 322.0 m has been selected as a type section for the formation (Resheniya… 1978). The maximum thickness of the formation (more than 35 m) has been registered in its type area (Fig. 20).

The Vaki Formation consists of weakly cemented glauconite‑bearing light-coloured fine‑grained or very fine‑grained sandstones, or of both, with thin interlayers of greenish‑grey and bleached‑purplish argillaceous rocks. On some levels the latter are highly micaceous and often contain ichnofossils of Skolithos affinites filled with fine sandy material. Besides ichnites, the formation contains scarce fragments of inarticulate brachiopods and in the Oostriku‑700 (Fig. 20-155) drill core at the depth of 256.2 m also a pauperized acritarch assemblage (further above‑ the Vērgale Stage).


Middle Cambrian

Compared to the Lower Cambrian, the Middle Cambrian is of more limited distribution and known only in the subsurface occurrence (Pirrus 1991).

Due to the lack of fossils and a very low core yield, the Middle Cambrian deposits are stratigraphically and sedimentologically poorly studied. The Middle Cambrian age of rocks was proposed according to their geological setting between the palaeontologically characterized Lower and Upper Cambrian rocks, and justification has been derived from the correlations with adjacent areas where the corresponding rocks contain fossils. Under such conditions, it is not always possible to identify the Middle Cambrian boundaries with a certainty. The lower limit of the Middle Cambrian is taken at the base of a thin band formed of coarse‑grained non-glauconitic sandstones with gravel and quartz pebbles overlapping the Lower Cambrian strata or the crystalline basement. Notable is an essential change in the mineral composition of the rocks, marked by the disappearance of glauconite, increase in the degree of maturity, prevalence of allothigenous minerals in the group of heavy minerals, etc. In southearn Estonia, the lower boundary is drawn on top of the whethering crust of the Lontova Stage. The identification of the Middle Cambrian is most complicated in the sections situated on the northern margin of the Middle Cambrian distribution area where it is underlain by lithologically similar Lower Cambrian light‑coloured quartzose sandstones (Mens & Pirrus 1992).

Throughout the distribution area, the Middle Cambrian consists of siliciclastic rocks dominated by mature light‑coloured well‑sorted non-glauconitic quartzose sandstones.

The Middle Cambrian succession in Estonia is subdivided into the Ruhnu and Paala formations (Table 6). The former is a local equivalent of the Deimena Formation in western Latvia (Resheniye… 1986, Mens et al. 1984, Pirrus 1991) and, according to fossil evidence, corresponds to the Ptychagnostus praecurrens Zone (Mens et al. 1987, 1990, Hagenfeldt 1989b).

The Paala Formation is distributed mainly in the southeast of Estonia and is tentatively interpreted as an equivalent of the Sablinka Formation of the Leningrad and Pskov regions (Mens et al. 1990). On the basis of palaeontological data, the Sablinka Formation is referred to the Paradoxides paradoxissimus and P. forchhammeri zones Judging from the mineral composition, the Middle Cambrian succession in some sections, including Abja and Otepää (Fig. 22-233, 249), is not clear.

The Ruhnu Formation is distributed in the southwestern part of Estonia, while its range in the middle of southern Estonia is plausible. It lies transgressively on the Lower Cambrian and is overlain by the Upper Cambrian or Lower Ordovician rocks.

The Ruhnu drill core in the interval of 706.8 to 748 m (Fig. 22-257) has been selected as the type section for the Ruhnu Formation (Kala et al. 1984a). The maximum thickness of the formation (41.2 m) has been established in this borehole and it decreases towards the north.

The Ruhnu Formation is represented by light well‑sorted, fine‑ to very fine‑grained quartzose sandstones with only a few thin interbeds of dark‑grey, in the lower part sometimes variegated, argillaceous rocks. The basal part is often marked by a layer of very coarse‑grained sandstone containing over 10% of quartz with gravel or pebble grain-size, or with both. The top of the formation is usually cemented by carbonates and pyrite. Often an admixture of feldspars and micas (muscovite) occurs. Glauconite is lacking. Heavy mineral assemblage is characterized by the prevalence of transparent minerals dominated by zircon and tourmaline. Among opaque minerals, detrital leucoxene and ilmenite occur in almost equal quantities or the former prevails slightly. Clay minerals are dominated by illite, the content of kaolinite often reaches 30‑35%. Chlorite is rare reaching occasionally 10% in the lower part of the formation.

The Paala Formation occurs in central and southeastern Estonia transgressively on the crystalline basement or on the Lower Cambrian rocks the type section is the Viljandi drill core (Fig. 22-203) in the interval of 409.8 to 434m (Mens et al. 1984). The thickness of the formation varies greatly. Due to the low core yield, the determination of the boundaries is complicated and the thickness of the formation is unclear.

The formation consists of light quartzose non‑glauconitic middle‑ to fine‑grained sandstones with pellets of white kaolinitic clay. The grain-size of the rocks in the lower and upper parts of the formation is coarser than in its middle part where very fine- to fine‑grained varieties dominate. Feldspar and muscovite are not common. The heavy mineral assemblage is mainly composed of allothigenic minerals dominated by ilmenite. In the group of heavy transparent minerals, zircon is always the index mineral.

The character of the basal bed depends on the type of underlying rocks. On the crystalline basement, it consists of conglomeratic sandstone, locally with phosphatized pebbles (Laanemetsa, Fig. 22-269). Resting upon the argillaceous Lower Cambrian rocks, which are often weathered, the basal bed occurs like a variegated kaolinitic clay comprising coarse grains of quartz redeposited from the weathering crust.


Upper Cambrian

During the last two decades, the stratigraphic extent of the Estonian Upper Cambrian and its biostratigraphical subdivision has been established more precisely due to the progress in research on acritarchs, conodonts and lingulates (Volkova et al. 1981, Kaljo et al. 1986, Popov et al. 1989, Mens et al. 1993 etc.).

The Upper Cambrian rocks have been palaeontologically documented in two isolated areas - in northern and south-eastern Estonia (Fig. 23). They crop out along the Baltic-Ladoga Klint and in the river valleys crossing it. The main localities in Estonia are Ülgase, Valkla, Turjekelder and Suurjõgi.

In Estonia, the Upper Cambrian strata are distributed sporadically and dominated by sandstones, less than 20 m in thickness. Argillaceous rocks are of limited distribution forming grey to greenish-grey interlayers within light-coloured sandstones in the lower part of the succession and brownish-grey varieties in the upper part.

The Upper Cambrian succession is condensed and interrupted by several minor and major hiatuses. The Upper Cambrian biostratigraphy and interregional correlation are usually based on trilobites. In the Estonian Upper Cambrian sections, where trilobites have not been found, it is based on acritarchs, conodonts and lingulates (Popov et al. 1989, Mens et al. 1993). Altogether five acritarch-, four conodont-, and at least three lingulate-based biostratigraphic units are distinguished (see Mens et al. 1993).

The oldest Estonian Upper Cambrian acritarch assemblage recorded from the Petseri Formation (SE Estonia) is similar to the acritarch assemblage BK1 b by Volkova (1990, Volkova & Kirjanov 1995), except the occurrence of Leiofusa stoumonensis Vang, which has been found in the base of the acritarch-bearing part of the Petseri Formation. Based on the occurrence of the genera Stelliferidium, Cymatiogalea, Leiofusa and Veryhachium, which appeared in Olenus time (Potter 1974, Downie 1984), and the absence of the Impluviculus species, the rocks comprising this acritarch assemblage are considered as a stratigraphic equivalent of the lower or middle, or of the both parts of the Olenus & Agnostus Zone.

The next acritarch assemblage in the Ülgase Formation (Table 6) is similar to that described above, but differs in the presence of Veryhachium dumontii and representatives of the genus Impluviculus and in the lack or restricted distribution of typical Middle Cambrian species. As the appearance of the genus Impliviculus has been correlated with the uppermost part of the Olenus & Agnostus Zone (Downie 1984) and with the lowermost part of the P. spinulosa Zone (Martin & Dean 1988), we regard the Ülgase Formation tentatively as a stratigraphic equivalent of the uppermost Olenus & Agnostus and the lowermost Parabolina zones (Table 6).

The third Upper Cambrian acritarch assemblage in the Estonian succession is distinguished by the appearance of Trunculumarinum revinium (Vang.) Loeblich et Tappan, Dasydiacrodium caudatum Vang. corresponds to the microflora A4, i.e. T. revinium - D. caudatum assemblage sensu Martin and Dean (1988), to the assemblage BK-3 sensu Volkova (1990), to the uppermost P. spinulosa Zone and to the Leptoplastus Zone, as a whole (Table 6).

The next acritarch assemblage, showing a high taxonomic diversity and variation in different sections, contains a significant amount of diacroids and sometimes also endemic forms. On the East-European Platform, this assemblage was first described from the upper part of the Ladoga Formation distributed in the Leningrad Region (Volkova & Golub 1985) and is referred to as BK-4. In Estonia, the assemblage occurs together with conodonts of the Proconodontus Subzone in the Tsitre Formation, but these acritarchs have also been found in the Cordylodus andresi Zone in the Kallavere Formation. A relatively similar assemblage together with the trilobites of the Peltura scarabaeoides Zone has been determined from the Degerhamn section of southern Öland (Di Milia et al. 1989).

The youngest Upper Cambrian acritarch assemblage, which also contains an abundance of diacroids (Volkova 1989, 1990), can be distinguished by the appearance of Acanthodiacrodium angustum (Downie) Combaz and Dicrodiacrodium ramusculosum (Combaz) Volkova. It occurs together with conodonts of the C. proavus Zone, and may be also characteristic of the C. intermedium Zone (Volkova & Mens 1988).

Conodonts have been studied from a number of outcrops and drill cores. The number of specimens is small, representing mostly the genera Phakelodus, Furnishina, Prooneotodus and Westergaardodina. Eoconodonts (Proconodontus, Eoconodontus and Cordylodus) have been found only in the topmost Upper Cambrian. According to the conodont zonation worked out by Sergeyeva and Viira for the Baltic-Ladoga Klint area, the Upper Cambrian sequence is subdivided as follows (from below upwards): the Westergaardodina Zone with W. bicuspidata, W. moessebergensis and Proconodontus subzones, and the Cordylodus andresi and C. proavus zones (Kaljo et al. 1986).

Lingulate brachiopods are represented in the Upper Cambrian of Estonia by lingulids and acrotretids. Following the brachiopod zonation suggested by Popov and Khazanovitch (Popov et al. 1989) and adopted by Puura and Holmer (1993), four brachiopod zones can be distinguished. From below upwards these are the Ungula inornata Zone, the Ungula convexa Zone, the Ungula ingrica Zone and the Obolus apollinis Zone. The first three zones belong to the Upper Cambrian, while the latter one can belong partly to the Ordovician, depending on the position of the lower boundary of the Ordovician System.

The Estonian Upper Cambrian includes three succeeding lithological units in the rank of formation (Petseri, Ülgase, Tsitre; Table 6), and the lower part of the overlying Kallavere Formation.

The Petseri Formation introduced by Kajak (1967) in the rank of beds is known only from core sections in southeastern Estonia, from where it spreads east- and southwards. In the Petseri borehole in Russia (Fig. 23), serving as the type section, the formation is 10.7 m thick. In complete sections the Petseri Formation can be subdivided into three parts. The lower and upper parts are represented by light-coloured weakly cemented quartzose sandstones, whereas the middle part is predominantly composed of grey argillaceous rocks (Volkova et al. 1981). Sandy parts of the formation contain some glauconite and debris of inarticulate brachiopods. In the argillaceous part, shells and fragments of lingulates of the genera Ungula and Oepikites and the Dasydiacrodium setuensis - Leiofusa stoumonensis assemblage of acritarchs occur (Volkova et al. 1981, Volkova 1990, Paalits 1992a). On the ground of this acritarch assemblage, the Petseri Formation, or at least its middle part, is considered as a time equivalent of the lowermost part of the Olenus & Agnostus Zone (Table 6).

The Ülgase Formation was referred by Öpik (1929) to the Acrotreta Sandstone and assigned to the local Acrotreta-Lingulella Zone. Subsequently, this part of the succession was defined as the Ülgase Member in the limits of the Pakerort Stage (Müürisepp 1958). In the rank of formation it was first considered by Khazanovich and Missarzhevsky (1982). The formation with a thickness of about 10 m is better fixed in the vicinity of Tallinn and within some 50 km east and south of it. The Ülgase Formation consists of light-coloured very fine- to fine-grained sandstones with interbeds and lenses of greenish-grey clay in the lower part and brownish-grey thin films in the upper part. Its upper boundary is transitional and in the earlier papers the lower part of the overlying Tsitre Formation was regarded as belonging also within this formation (Mens 1984, Popov et al. 1989). The formation contains numerous lingulates of the genera Ungula (including U. inornata), Oepikites, Angulotreta and Cerotreta. The occasional conodonts belong to the genera Phakelodus, Furnishina and Prooneotodus (Kaljo et al. 1986, Mens et al. 1993). Torellella sulcata Missarzhevsky abound. In the argillaceous interlayers the acritarchs, forming the Impluviculus multiangularis - Veryhachium dumontii assemblage are numerous (Volkova 1982, 1990, Volkova & Mens 1988). On the basis of fossil evidence, the Ülgase Formation is assigned to the uppermost part of the Olenus & Agnostus Zone and to the lower part of the Parabolina spinulosa Zone (Table 6).

The Tsitre Formation was introduced by Popov and Khazanovich (1985) with the stratotype in the Turjekelder section. Earlier, this part of the succession belonged to the Kallavere Formation (Kaljo et al. 1986, Resheniye… 1986). Currently, the Tsitre Formation (Table 6) includes also the underlying beds containing kerogen-bearing argillaceous interlayers and differing in fossil record from the Ülgase Formation (Mens et al. 1993).

The Tsitre Formation expands as a narrow belt from Tallinn to Kohtla-Järve (Fig. 23). Its thickness in the outcrop sections is a bit more than 3 m. In the drill core sections its thickness is unclear due to the low core yield, but probably it is less than 10 m.

The formation is typically represented by light-grey weakly cemented fine-grained quartzose sandstones, with a few thin interbeds of variegated, dominantly brownish-grey clayey rocks. These interlayers are often accompanied by bedding planes covered with convex-up lingulate shells. This has been considered in drawing the boundary between the Ülgase and Tsitre formations.

The co-occurrence of Trunculumarinum revinium and Dasydiacrodium caudatum in the lower part of the Tsitre Formation suggests that these deposits belong to the upper part of the Parabolina spinulosa Zone (Martin & Dean 1988, Paalits 1992b). The upper part of the Tsitre Formation contains a rather rich fossil record and its relationship with the trilobite zones is shown in Table 6.

The uppermost part of the Cambrian is considered to be represented by the lower part of the Kallavere Formation. The latter is discussed with the rest of this formation within the Ordovician.





J. Nõlvak


In the Ordovician, epicontinental seas with extensive distribution of carbonate sediments had a greater extent than in any other period. The marine flora and fauna changed markedly in the course of the Ordovician. A number of major taxonomic groups (bryozoans, brachiopods, echinoderms, trilobites, ostracodes, chitinozoans and others) appeared or became common. In this respect, the Ordovician is one of the most interesting periods in the history of marine faunas, and Estonia is among the areas in the world where this fauna is well preserved and studied.

The Ordovician was characterised by an extreme biogeographical differentiation of both planktic and benthic faunas, but with different degree. This makes the worldwide correlation of the Ordovician rocks difficult and has resulted in numerous regional stratigraphic schemes. A series of detailed stratigraphical charts compiled for the East Baltic (see Resheniya… 1978, 1981, Männil & Meidla 1994, and literature cited in these papers), gave a relatively stable detailed local classification for Ordovician rocks and afterwards obtained the status of a regional standard for most of the East-European Platform (Männil 1990).

The large-scale biogeographical and facies differentiation within the Ordovician Palaeobasin of Baltoscandia is well expressed in the concept of confacies belts (Jaanusson 1976, Fig. 24). The territory of Estonia is divided between the North Estonia and Central Baltoscandian confacies.

An emended version of the correlation chart, presented for the Estonian succession ranging from 70 to 180 m in thickness (Table 7), is based mainly on the above-cited schemes. For practical reasons the present version is simplified and many smaller subdivisions have been omitted. Some subdivisions, defined earlier as formations due to overestimations or difficulties in their specifications, are treated as members. Often the lithounits (Table 7) have diachronous (wavy line) or topical (discontinuous line) boundaries or the unit serves as “topostratigraphic” unit (sensu Jaanusson 1976, p. 310) with their boundaries coinciding with the stage boundaries.


Oeland Series

Pakerort Stage

H. Heinsalu & V. Viira


The lowermost Ordovician Pakerort Stage (Table 7) distinguished by Raymond (1916) consists of two different lithotypes - the Obolus Sandstone (the Kallavere Formation, Männil & Rõõmusoks 1984) and Dictyonema Shale (Türisalu Formation, Müürisepp 1958, 1960b). During the last decades, both lithotypes have been studied in particular detail in terms of their industrial use and potential environmental impact. In northern Estonia, the so-called Obolus-Conglomerate (brachiopod coquina) occurs usually at the base of the Obolus Sandstone. The coquina was used as a good lithological marker in fixing the lower boundary of the Pakerort Stage and the Cambrian/Ordovician boundary in Estonia.

The Cambrian/Ordovician boundary (Photo 17) became an object of special international studies in the 1970s, and since then several different stratigraphical levels have been proposed as the lower boundary of the Ordovician system. According to most of stratigraphers, the definition of the Cambrian/Ordovician boundary should be based on conodonts and the horizon chosen should be close to, but below the lowest planktic (nematophorous) graptolites. Three biostratigraphic horizons of conodonts were considered as possible guides for marking the boundary level. These are the base of Cordylodus proavus, of C. intermedius and of C. lindstromi zones. Currently, the attention is focused on the first appearance of the conodont Iapetognathus n. sp. in the lower part of the C. lindstromi Zone just above the first appearance of the planktic graptolites R. praeparabola and R. parabola in the Dayangcha section (China). In Estonia, these conodont zones have been identified in the Kallavere Formation considered preliminarily (Männil & Rõõmusoks 1984) as the oldest part of the Ordovician sequence. Somewhat later the lower boundary of the system and of the Pakerort Stage was tentatively drawn at the level of the first appearance of Cordylodus (base of C. andresi Zone, Fig. 25) in the lower part of the Kallavere Formation (Kaljo et al. 1986, Resheniya… 1987, Männil 1990). But if a higher stratigraphical level (e.g. the base of the C. lindstromi Zone) will be accepted internationally for the boundary between the Cambrian and Ordovician systems, most of the Kallavere Formation must be excluded from the Ordovician (Norford 1991, Miller & Taylor 1995, Fig. 25).

As there are no distinct lithological changes on the boundaries of the conodont zones in most of the sequences, the Obolus Sandstone of the Kallavere Formation will be treated below as an entity (Fig. 26). Within that formation the main attention focuses on the distribution of conodonts, zones of which could serve as guides for the boundary between the systems. The following succession of conodont zones (in ascending order) has been established in the Obolus Sandstone and Dictyonema Shale (Kaljo et al. 1986): Cordylodus andersi, C. proavus, C. intermedius, C. lindstromi and C. angulatus. The Cordylodus andresi Zone has been established only in a few sequences in northern Estonia (Kidaste core in Hiiumaa, outcrops at Turjekelder, Vihula and Toolse). The C. proavus Zone occurs in all sections (except Turjekelder) where conodonts have been studied, and in some sections it is rather thick (Fig. 25). The zone is absent in most of the brachiopod coquina (Obolus-Conglomerate). The morphological variability of the zonal species of the C. proavus Zone suggests that the zone is discontinuous: sometimes the lower, sometimes the upper part is missing. Compared to other conodont zones in the Estonian sequences, the C. proavus Zone has the most distinct lower boundary which coincides or occurs close to the lower boundary of the Kallavere Formation.

The C. intermedius Zone is of limited distribution in Estonia. The index species has been established only in three sections (Mäekalda in Tallinn, Ülgase, Toolse). The C. lindstromi Zone, vice versa, is widespread. It is missing only on the Pakri Cape and possibly also in the Vihula section (Fig. 25). The specimens of C. lindstromi Druce et Jones found in Estonia and Australia (Nicoll 1991), are morphologically very similar (low base, small cavity with one or more pointed secondary tips).

The C. angulatus Zone occurs in all northern Estonian sections, except the Pakri Cape. Its lower boundary is well defined by the appearance of numerous specimens of C. angulatus Pander apparatus. In the western part of Estonia, it coincides with the lower boundary of the Suurjõgi Member, which is a good lithological marker (Fig. 26).

The co-occurrence of conodonts and graptolites differs considerably from area to area (Kaljo & Viira 1989). The Rhabdinopora flabelliformis group makes its first appearance at different levels in relation to the conodont zones. In the vicinity of Tallinn, it occurs in the top of the C. proavus Zone or in the C. intermedius Zone, east of Tallinn in the C. lindstromi Zone, and to the west of it in the C. angulatus Zone. This complicates the use of graptolites in the correlation of sections. Nevertheless, the graptolites serve as the most important group of fossils in establishing the upper boundary of the Pakerort Stage which falls into the organic-rich argillites of the Türisalu Formation.

Besides the above-mentioned conodonts and graptolites, lingulate brachiopods and acritarchs can be used for the subdivision of the Cambrian-Ordovician boundary beds (Kaljo et al. 1986, Puura & Holmer 1993, Mens et al. 1993, Paalits 1995).


Kallavere Formation

The quartzose sandstone with interbeds of dark argillite of the Kallavere Formation are distributed almost all over Estonia. It is missing only in a belt running from southwestern to eastern Estonia (Fig. 27). The formation is at its thickest (more than 17 m) in central Estonia. It consists of the Maardu, Rannu, Katela, Orasoja and Suurjõgi members, replacing each other in space or in time (Heinsalu 1981, 1987, Fig. 26).

Lithologically, the Kallavere Formation is dominated by quartzose sandstone, commonly weakly cemented, with the grain-size of 0.05-0.25 mm. It contains phosphatic brachiopod valves and their fragments, forming a distinct brachiopod coquina layer at the base of the Kallavere Formation (in the Maardu and Rannu members). The thickness of the coquina is only a few cm, except the Maardu (up to 1 m) and Rakvere (up to 4-6 m, sometimes even more) areas. The Kallavere Formation comprises lingulate brachiopods of the Ungula ingrica and Obolus apollinis zones (Heinsalu et al. 1987, Mens et al. 1993). In the conglomerate bed the most common species is Ungula ingrica, accompanied by species of the genera Schmidtites, Keyserlingia and Oepikites.

Dark argillite interbeds, 0.1 mm to 15 cm in thickness, occur generally above the brachiopod coquina. In northwestern Estonia, a 10–30-cm-thick dark argillite (Dictyonema Shale) bed with very thin interbeds of sandstone lies immediately on the lower boundary of the formation. In northwestern Estonia, west of the Kunda - Rakvere line, the uppermost part of the formation is represented by the so-called skeletal detritus layer of the Suurjõgi Member which consists of cross-bedded quartzose sandstone comprising brachiopod fragments with a size of 1-3 mm.The thickness of the Suurjõgi Member is about 1 m, except the Toolse - Vihula area where it exceeds 5 m (Fig. 26).


Türisalu Formation

Up to the 1970s, the Türisalu Formation (Müürisepp 1958, 1960a, b) was considered as the upper part of the Pakerort Stage. The studies of the distribution of graptolites (Kaljo & Kivimägi 1970, 1976) allowed to divide the formation between the Pakerort and Varangu stages. The older part is characterized by the occurrence of graptolites of the Rhabdinopora flabelliformis Zone and the younger part by the graptolites of the Kiaerograptus Zone. West of the Tallinn - Rapla line (Fig. 28) only the older part of the formation is represented. It comprises the Rhabdinopora f. flabelliformis (in the lower part) and R. flabelliformis multithecata graptolite subzones (Kaljo & Kivimägi 1970, 1976), and the upper part of the C. angulatus conodont Zone. The most complete stratigraphical sequences of the Türisalu Formation occur between Tallinn and Tapa where the lower part of the formation belongs to the Pakerort and the upper part to the Varangu Stage. All the sections of the Türisalu Formation east of this area are of Varangu Age.

The Pakerort Stage is represented mainly by dark-brown horizontal laminated graptolite argillite. The lamination is caused by the different content of organic matter (intercalation of darker and lighter laminae) or by different grain-size (Heinsalu 1990a, Kivimägi & Loog 1972). In some cases, wavy or cross-bedded structures or thin (a few cm) interbeds of light, often pyritized, quartzose siltstone occur in the lower part of the formation.

There is no lithological markers for identification of the boundary between the Pakerort and Varangu stages in the limits of the Türisalu Formation and the maps of the distribution of Tremadoc rocks have been compiled by formations (Figs. 27, 28, 29). The thickness of the Türisalu Formation is up to 7 m (Fig. 28).


Varangu Stage

H. Heinsalu & V. Viira


The later Tremadoc rocks, which belong to the Varangu Stage (Männil 1990, = Ceratopyge Stage, Männil 1966, Viira et al. 1970) and have a thickness of 4-5 m extend, as a relatively narrow (20-50 km) belt in northern Estonia (Fig. 29). In the argillites, the lower boundary of the stage can be established by the appearance of graptolites of the Kiaerograptus Zone and conodonts of the Paltodus deltifer pristinus Subzone. The appearance of adelograptids marks the lower boundary of the Varangu Stage in the lithologically quite uniform Türisalu Formation. The upper part of the formation differs from the lower part, which belongs to the Pakerort Stage, by the occurrence of interbeds of very fine-grained quartzose sands from some mm up to 4-5 cm in thickness. Frequently, these interbeds abound in pyrite concretions.The Toolse area, where the Toolse Member was defined, has been studied in particular detail (Kivimägi & Loog 1972, Heinsalu 1980).


Varangu Formation

The Varangu Formation, the youngest part of the Tremadoc, is widely distributed in northwestern Estonia (Fig. 29). It is at its thickest (ca 3 m) between Haljala and Kunda in northeastern Estonia where the Varangu Formation can be subdivided into three lithologically different parts. The lower and upper parts are predominantly clayey, consisting mostly of compact claystone which comprises glauconite and pyrite, scattered or concentrated in small lenses. The middle part is rich in glauconite and very fine-grained quartz, sometimes prevailing over pelitic material. The sand is hardly pyritized. A similar three-part sequence of the Varangu Formation occurs also on the Pakri Cape in northwestern Estonia, but its thickness there is only 0.3-0.4 m.

In most of western Estonia, the Varangu Formation is characterized by the greenish-grey compact silty clay or sandy deposits with glauconite grains. In some sequences the clays of the Varangu Formation are dark in colour which makes them similar to the Dictyonema shale.


Hunneberg Stage

T. Meidla


The Hunneberg Stage was introduced by Tjernvik (1956) as the Hunneberg Group in Sweden, based mainly on trilobite faunas. During several decades, the stage has been recorded as the lower substage of the Latorp Stage in Estonia (after Jaanusson 1960a, Männil 1966, Männil & Meidla 1994 etc.). Following Jaanusson (1982), Mägi (1984) and Hints et al. (1994) considered this unit in the rank of stage.

During the last years, Sweden has served as the key area for biostratigraphical research of the Ontika Subseries comprising the stages from Hunneberg to Kunda. Detailed studies of the earliest post-Tremadoc sequences by Lindström (1954), Tjernvik (1956), Jaanusson (1963) and several other researchers have been supplemented by recent studies of sequences and distribution of graptolites (Lindholm 1991 a.o.) and conodonts (Löfgren 1993a, b, 1994, 1996). In the East Baltic region, the stratigraphy of the Ontika Subseries has been studied by Lamansky (1905), Öpik (1930b) and Männil (1963a, b, 1966). In Estonia modern biostratigraphy of this interval bases mainly on conodonts studied by Viira (1966, Mägi & Viira 1976, Mägi et al. 1989).

In northern Estonia, the Hunneberg Stage is represented by poorly lithified glauconitiferous terrigenous sediments: glauconitic siltstones of the Klooga Member (thickness up to 2.9 m) and glauconite silt and sand of the Joa Member (up to 1.2 m) which together form the lower, main part of the Leetse Formation (Figs. 30, 31). The content of glauconite is increasing upwards. The glauconitic siltstones of the Klooga Member are dominated by quartz with a supplement of glauconite (Mägi 1970), while the silt- and sandstones of the Joa Member consist mainly of glauconite (50–70%) and quartz (about 10–20%; Mägi 1970, 1984, 1990). The lower boundary of the Leetse Formation and the Hunneberg Stage represents a well defined lithological marker level with the glauconitic sandstones and siltstones overlying conformably, sometimes with a discontinuity surface, the dark-brown argillites of the Türisalu Formation or light-grey clays of the Varangu Formation.

The maximum thickness of the stage reaches 4 m in northwestern Estonia, but usually it is less than 2 m. As the Hunneberg and Billingen stages have not been differentiated in most sections, the thickness map (Fig. 30) shows only total thicknesses for both stages. In western Estonia, mainly on the islands of the West-Estonian Archipelago, the absence of the Hunneberg Stage has been documented from several sections.

In Sweden, the Hunneberg Stage corresponds to the Megistaspis armata and M. planilimbata trilobite zones. The base of the stage is close to that of the Paroistodus proteus conodont zone (Fig. 32, Löfgren 1993a). In the East Baltic, the lower M. armata Zone has been established only with confidence in Latvia (Männil 1966, Ulst et al. 1982). In northern Estonia, the lowermost part of the Leetse Formation has generally been assigned to the Paroistodus proteus Zone (Männik & Viira 1990). In the Mäekalda section, the thin Klooga Member at the base of the Leetse Formation is referred to the Paltodus deltifer Zone (Fig. 31, Mägi 1984).

Among macrofossils, a distinctive assemblage of lingulate brachiopods (Goryansky 1969) has been recorded. It is characterized by Thysanotos siluricus (Eichwald) and Leptembolon lingulaeformis (Mickwitz) constituting the Leptembolon-Thysanotos assemblage, widely distributed in eastern and central Europe (Popov & Holmer 1994).

The fossil evidence from most of central and southern Estonia, is too fragmentary yet for the limitation of the Hunneberg and the overlying Billingen stages. The Hunneberg age of glauconitic sandstones (up to 0.5 m) has been established in the Karula core (Männil 1966), but in most cases the detailed stratigraphy of the undivided Hunneberg–Billingen strata is unclear.

In Latvia, and in some sections in southern Estonia, close to the Estonian - Latvian border, the mudstones of the Zebre Formation, reaching a thickness of 46 m in Latvia, have been considered as equivalents of the Varangu, Hunneberg and Billingen stages (Ulst et al. 1982, see also Fig. 30). The middle part of this formation (Zirni Member) is of Hunneberg age, as it yields a zonal trilobite Megistaspis planilimbata (Angelin) (Fig. 32) and several graptolites, including Tetragraptus phyllograptoides Strandmark, together with a zonal conodont Paroistodus proteus (Lindström) (ibid.). The occurrence of the peripheral parts of the Zebre Formation can be assumed in southern Estonia.


Billingen Stage

T. Meidla


The Billingen Stage (Tjernvik 1956, Jaanusson 1982), understood here in the sense of the upper Billingen Substage of the Latorp Stage (Jaanusson 1960a, Männil 1966, 1990, Männil & Meidla 1994), consists of two distinctive parts in northern Estonia. The lower one is represented by the glauconitic calcareous sandstones and limestones of the upper part of the Leetse Formation (Mäeküla Member, equal to BIβ by Lamansky (1905), whereas the glauconitic limestones of the lowermost part of the Toila Formation (Päite Member, equal to BIγ by Jaanusson (1951), form the upper half of the Billingen Stage (Table 7, Fig. 31). In some publications, including the detailed lithological study of the Volkhov and Kunda stages by Orviku (1960a), the Päite Member is interpreted as the lowermost unit of the Volkhov Stage. This interpretation is also in use in the Leningrad Region of Russia.

The original concept of the Billingen Stage (Tjernvik 1956) was based on the evidence from the trilobite faunas, but its lower boundary can best be recognized by the distribution of conodonts. In Sweden, it is situated fairly close to the boundary of the Prioniodus elegans Zone and this level is recognizable also in several sections of northern Estonia where it nearly coincides with the lower boundary of the Mäeküla Member in the upper part of the Leetse Formation (Fig. 31).

The Mäeküla Member consists of glauconitic sandstones which are replaced upwards by calcareous sandstones and glauconitic limestones. The lower boundary of this member is lithologically fairly distinct in the klint area, and is marked by the change from poorly cemented silts and sands to well cemented sandstones. For the purposes of correlation, the occurrence of the conodonts Prioniodus elegans Pander and Oepikodus evae (Lindström) is most important (Figs. 31, 32). The thickness of the Mäeküla Member varies from 0 to 0.5 m. Sandy material with the grain-size over 0.1 mm forms up to 80% of the rock, whereas glauconite grains make up some 80% of this fraction. The quartz content varies from 10 to 40% (Mägi 1984, 1990). The highest content of glauconite has been recorded in central northern Estonia. The macrofauna of the Mäeküla Member has not been described monographically but, according to the available evidence, its main, upper part is comparable with the Megistaspides dalecarlicus Zone of Sweden (Fig. 32).

The Päite Member is represented by limestones or dolomites which dominate in the easternmost sequences, with a low content of mainly fine-grained glauconite. The greatest thickness of the member is 1.13 m, and it decreases in the northwest direction. In the Leningrad Region, the presence of several distinctive lithological marker horizons within the equivalents of the Päite Member (roughly equal to the informal Red Dikari Member) has been demonstrated by Dronov et al. 1996, part of those can be supposedly distinguished in northeastern Estonia. On the islands of Väike-Pakri (Photo 18) and Osmussaar, the member is sandy and may contain a layer of calcareous, glauconitic sandstone at the base (Orviku 1960a). The Päite Member is roughly equivalent to the Megistaspis (Paramegistaspis) estonica Zone of Sweden (Fig. 32).

In central Estonia, the presence of the Billingen Stage needs further approval. Glauconitic sandstones occuring locally in a restricted thickness in this area, like in the Äiamaa and Võhma cores (Rõõmusoks 1960, 1983), may belong to this stage but their precise age is not yet clear. Further to the south (at Tartu), the reddish-brown, occasionally glauconitic dolomites may be tentatively assigned to the Billingen Stage. The distribution of the Zebre Formation in Latvia (Ulst et al. 1982) suggests that it extends as far as the southernmost Estonia. In Latvia, the red or mottled clays of the topmost Zebre Formation (Zante Member) contain Megistaspis (Paramegistaspis) estonica (Tjernvik) and a zonal conodont Oepikodus evae (Lindström) (Gailite & Ulst 1975, Ulst et al. 1982), characteristic of the Billingen Stage in several sections of northern Estonia (Mägi 1990, Einasto et al. 1996, see Fig. 31).

Fossil evidence from southern Estonia is too fragmentary to enable the limitation of the Hunneberg and Billingen stages over the study area. Figure 30 shows only their sum thickness. The absence of the Billingen Stage in several sections of the West-Estonian Archipelago (Saaremaa, Hiiumaa) should be mentioned.

The Mäeküla Member contains the oldest Ordovician representatives of articulate brachiopods of the genera Plectella, Panderina, Prantlina and Angusticardinia (Rubel 1961) bryozoans, ostracodes (H. Aru, pers. comm) and trilobites. Frequent occurrence of trilobites (“Megistaspis”) has been recorded in some levels (Orviku 1960a). In the Leningrad Region, the Mäeküla Member yielded the material for original definition of Conodonta by Pander (1830). The yet poorly studied fossil record of the Päite Member contains conodonts, brachiopods, trilobites and ostracodes.


Volkhov Stage

T. Meidla


The Volkhov Stage, corresponding roughly to the “Glauconit kalk” by Schmidt (1879, 1881), forms a lithologically distinctive unit in the sections of the North-Estonian Klint and nearby river valleys (Photo 19). The term “Volkhov” was introduced by Raymond (1916) as the “Walchow Formation” in a broader meaning (Table 1) corresponding to the lower part of the Ontika glauconitic limestones in northern Estonia. Lamansky (1905) was the first to introduce three substages in the Leningrad Region. In Estonia these are in ascending order the Saka, Vääna and Langevoja substages (Männil & Meidla 1994), conceptually largely based on trilobite zonation. The two first units are not accepted in Sweden.

In northern Estonia, the Volkhov Stage is represented by the main part of the Toila Formation, locally succeeded by the lower part of the Sillaoru Formation (Pada Member, Fig. 33) which has tentatively been assigned to the Volkhov Stage (Orviku 1960a, a.o.). In central Estonia, the Toila Formation is assumed to compose the whole Volkhov sequence. Southward the formation grades into the Kriukai Formation (Table 7, Fig. 34).

The Toila Formation is a complicated stratum made up of various, partly dolomitized glauconitic limestones resting on sandstones of the Leetse Formation. The lower boundary of the Volkhov Stage and the Saka Member is marked by a smooth discontinuity surface with “amphore-like borings” (Orviku 1960b) in the uppermost bed of the Päite Member (Orviku 1960a), known as Püstakkiht (Orviku 1961). In northern Estonia, the main part of the Toila Formation, corresponding to the Volkhov Stage, is subdivided into five members (Orviku 1960a) which are partly lateral equivalents. Only the lower, Saka Member consisting of dolomitized glauconitic limestone (up to 1.2 m) forms the base of the Volkhov Stage all over northern Estonia (Fig. 33). It is overlain by two laterally equivalent units: the Telinõmme Member in the west (interbedded greenish-grey limestones and marls, up to 2 m) and the Künnapõhja Member in the east (mottled dolomitic limestone, up to 1 m) (Fig. 33). East of Tallinn, the upper part of the formation is represented by the Kalvi Member (grey argillaceous glauconitic limestones) with a thickness of up to 1.7 m. West of Tallinn, it is composed of the Lahepera Member (glauconitic limestones, partly sandy or conglomeratic, up to 0.5 m) which is assumed to represent the youngest part of the formation, as it locally overlies the Kalvi Member.

According to the trilobite evidence, the lower, Saka Member comprising trilobites Megistaspis “elongata” (Schmidt) and Megistaspis “lata” (Törnquist), and a zonal conodont Baltoniodus navis (Lindström) (as referred to by Männil 1966 and Männil & Meidla 1994), represents the lower, Saka Substage in Estonian succession. The Telinõmme and Künnapõhja members comprising Paroistodus originalis (Sergeeva) may correspond to the middle, Vääna Substage, while the Kalvi and Lahepera members probably correspond to the upper, Langevoja Substage (Fig. 32). Ostracodes (Fig. 35) and brachiopodes (Rubel 1961) are common in the Toila Formation. Gastropods, cephalopods and cystoids have also been recorded.

The up-to-3.5-m-thick Toila Formation is poorly developed in northwestern Estonia (Fig. 34). In northeastern Estonia where the formation is at its thickest, the rocks have undergone extensive dolomitization.

The Sillaoru Formation, the “untere Linsenschicht” by Schmidt (1897), consists of two distinct members of oolitic limestones with noticeably different ages. The lower, Pada Member (up to 0.5 m of limestone with small ferriferous ooids and occasional glauconite grains) comprises Metaptychopyge truncata (Nieszkowski) (Resheniya… 1978), Ptychopyge angustifrons (Angelin) (Mägi 1990) and is apparently of Late Volkhov (Early Kunda?) Age. Its age relationship with the Lahepera Member of the Toila Formation remains still open due to the different distribution areas of these members (Fig. 31).

In central Estonia, the sequence of the Toila Formation comprises the same members with similar thicknesses as in northeastern Estonia (Saka, Künnapõhja and Kalvi). In southern Estonia, the formation grades into the Kriukai Formation (Table 7, Fig. 34) which consists mainly of reddish-brown marls, with limestone and mudstone intercalations. The thickness of these rocks does not exceed 20 m in Estonia, but in Latvia it is much greater, reaching 32.5 m. The age of the formation in southern Estonia has not been established biostratigraphically. However, in Latvia it displays a rich fauna of trilobites, ostracodes, conodonts, more rarely articulate brachiopods and lingulates (Megistaspis “limbata” (Boeck), Ptychopyge angustifrons Angelin, Tallinnellina primaria (Öpik), Rigidella mitis (Öpik) (Gailite & Ulst 1975, Ulst et al. 1982, Männil & Meidla 1994), suggesting the Volkhov Age (Sarv 1959, Meidla & Sarv 1990, Männil & Meidla 1994).


Kunda Stage

T. Meidla


The Kunda Stage (Kunda Formation by Raymond 1916) is represented by oolitic, glauconitic (Lamansky 1905) and sandy limestones corresponding to the emended Vaginatum Limestone by Schmidt (1897). A three-part subdivision of the strata, based on the trilobite zonation, was introduced already by Lamansky (1905) in the Leningrad Region. He also assumed the absence of the lower unit - the Asaphus expansus Zone in northern Estonia which was afterwards confirmed by several authors (Raymond 1916, Orviku 1960a, Männil 1966, a.o.). Orviku (1958b) proposed to name the Lamansky’s subdivisions in ascending order the Hunderum, Valaste and Aluoja substages, and presented a detailed lithostratigraphical description of the corresponding interval in northern Estonia (Orviku 1958b, 1960a, b; Fig. 33).

In northern Estonia, the Kunda Stage comprises the Valaste (corresponding to the Asaphus “raniceps” Zone) and the Aluoja (zones of Megistaspis obtusicauda and Megistaspis gigas) substages (Fig. 32, Table 7). In most of northern Estonia, the lower boundary of the Valaste Substage is drawn at the base of the oolitic limestone of the Sillaoru Formation, or locally within the unit, being marked by a discontinuity surface on the boundary between the Pada and Voka members, and the disappearance of the glauconite grains, characteristic of the underlying strata (Fig. 33).

In northwestern Estonia, the Kunda Stage is represented by the Pakri Formation, eastwards replaced by the upper part of the Sillaoru Formation and the Loobu Formation (Fig. 33). In northeastern Estonia, the Napa Formation forms the topmost part of the Kunda Stage and grades into the Rokiškis Formation in central Estonia. In southern Estonia, the entire Kundan sequence, including the Hunderum Substage, is represented by the Šakyna and Baldone formations (Fig. 36).

The Pakri Formation (Öpik 1927), up to 4.5 m of yellowish-grey sandy limestones and calcareous sandstones, sometimes with conglomerate beds, occurs in northwestern Estonia in the area west of Tallinn. In the westernmost part of its distribution area, the main, lower part of the Pakri Formation consists of up-to-4m-thick nodular kerogenous calcareous sandstone of the Suurupi Member, overlain by the thin (0.5 m) sandy limestone of the Osmussaar Member. In the surroundings of Tallinn, the formation is represented by limestones with quartz and glauconite grains, locally with a basal conglomerate (Kallaste and Jägala members, up to 0.8 m). On the Island of Osmussaar and, to a lesser extent, in the neighbouring mainland areas, a system of sedimentary dikes cuts through the Pakri Formation and the underlying Volkhov-Billingen strata (Puura & Tuuling 1988). The time of the formation of the dikes is dated as middle-late Kunda.

In most of northern Estonia (except the distribution area of the Pakri Formation) and in central Estonia, the basal part of the Kunda Stage consists of the oolitic limestone of the Sillaoru Formation (Resheniya… 1978, Männil & Rõõmusoks 1984). The main, Valaste time part of the formation (Voka Member, up to 0.6 m) consists of clayey limestones with abundant ferriferous ooids, developed around skeletal particles or glauconite grains (Mägi 1984). Among the skeletal particles, fragments of trilobites and ostracodes dominate (50-70%, Mägi 1984). The Voka Member generally serves as a good marker level in the North-Estonian sequence, although in restricted areas of northern and northeastern Estonia it overlies the thin oolitic Pada Member which differs from the main part of the formation by the presence of glauconite grains and has been included in the Volkhov Stage by Orviku (1960a) and subsequent authors. In northern and northeastern Estonia, the Loobu Formation constitutes the main part of the Kunda Stage. Detailed study of the formation (Orviku 1958b, 1960a) has revealed its two-part subdivision; both the lower and the upper parts consist of two laterally equivalent units. In central northern Estonia, east of Tallinn, the formation is represented by clayey limestone of the Nõmmeveski Member (up to 2 m) and glauconitic limestone of the Ubari Member (up to 2 m, Fig. 33). In northeastern Estonia, the lower part of the formation consists of glauconitic limestone of the Utria Member (up to 3 m), overlain by clayey limestones of the Valgejõgi Member (up to 4.7 m, Männil 1987). In the outcrop area, large nautiloids Cyclendoceras vaginatum (Schlotheim), Estonioceras ariense (Schmidt), Para-cyclendoceras cancellatum Eichwald etc., (Rõõmusoks 1960) are characteristic of most of the Loobu Formation. In northeastern Estonia, the rocks have undergone extensive dolomitization resulting in a mottled red colour and cavernous structure. The Loobu Formation reaches its maximum thickness (7 m) in the central part of northern Estonia (Fig. 36), in central Estonia it is less than 3 m thick (Resheniya… 1987). In that area the formation consists of grey, partly clayey glauconitic limestones, overlying the oolitic limestones and marls of the Sillaoru Formation (0.5 m).

The Napa Formation, an oolitic marl and limestone body (up to 4 m), is supposed to replace the upper part of the Loobu Formation in northeastern and central Estonia (Fig. 33).

The relation of the formations and members forming the Kunda Stage in northern Estonia is well demonstrated by Orviku (1960a). The correlation is largely based on the trilobite evidence. Asaphus “raniceps” Dalman has been identified in the Suurupi Member of the Pakri Formation and in the lower part of the Loobu Formation (Nõmmeveski Member). The Osmussaar Member comprises Pseudoasaphus globifrons (Eichwald), which is known from the upper part of the Loobu Formation (Ubari and Valgejõgi members). The Napa Formation is characterized by Megistaspis gigas Angelin (Resheniya… 1978, Mägi 1990). In terms of conodont zonation, the Valaste and Aluoja substages roughly correspond to the Eoplacognathus variabilis Zone (Fig. 32). The shelly fauna is represented by brachiopods, ostracodes, gastropods and cephalopods (Öpik 1927, Sarv 1959, Rubel 1961, Mägi 1990).

In central Estonia, the Napa Formation grades into the Rokiškis Formation (Fig. 36), which is represented by red mottled oolithic limestone (up to 15 m). The fauna of this unit is poorly known in Estonia. Based on Panderodus cf. sulcatus (Fåraeus) and Pinnatulites procera (Kummerow) recorded by Männil (Resheniya… 1987), the Kunda-Aseri age has been suggested. In southern Estonia, the sequence of the Loobu and Rokiškis formations grades into the Šakyna and Baldone formations, represented by grey glauconitic limestone and clayey red limestone, respectively. Palaeontologically, these units are poorly characterized in Estonia and the age relationship to the northern and central Estonian sequences is obscure. The fauna of the Šakyna Formation in Latvia contains trilobites, more rarely brachiopods and graptolites, the Baldone Formation is more fossiliferous (Gailite & Ulst 1975, Ulst et al. 1982). In this area the succession of the above-named formations comprises the entire Kunda Stage and stratigraphically the section of southern Estonia is the completest in this interval.

The thickness of the Kunda Stage demonstrates an obvious decreasing trend towards northwestern Estonia. In most of northern and central Estonia, it does not exceed 10 m, but in southeastern Estonia may locally reach 20 m (Fig. 36).


Viru Series

Aseri Stage

L. Hints


The term Aseri Stage was used first by Bekker (1922, 1923) for the Schmidt´s (1897) Upper Oolitic Limestone (Obere Linsenschicht). In nowadays understanding the Aseri Stage bases in a great deal on the studies carried out by Orviku (Jaansoon-Orviku 1927, Orviku 1929, 1930a, 1940), Rõõmusoks (1960, 1970) and Männil (1966).

In northern and central Estonia, the Aseri Stage is 0.1 - 5 m thick (Fig. 37) and consists of bioclastic limestones with unevenly distributed ooids, predominantly brown ferriferous (goethitic) ooids (Orviku 1940, 1960b). In places, the ooids are frequent in the lower and upper parts of the stage, but in the dolomitic limestones of northeastern Estonia they occur only in the upper part (Fig. 38). White phosphatic ooids are distributed mainly in the westernmost sequences. These, early Middle Ordovician oolitic limestones have been treated as the Kandle Formation (sensu stricto; Männil & Rõõmusoks 1984). Afterwards, Männil (1990, Männil & Meidla 1994) proposed the name Aseri Formation for the oolitic limestones of Aseri Age. Here preference is given to the term Kandle Formation, because in many cases the upper boundary of the Aseri Stage is difficult to determine; it may coincide with the upper boundary of the oolitic limestones or fall into the upper part of it.

The Kandle Formation is subdivided into the Malla (Männil & Rõõmusoks 1984) and Ojaküla (Orviku in Aaloe et al. 1958) members. The lower, Malla Member (Asaphus and Echinosphaerites limestones by Jaansoon-Orviku 1927) is lithologically the most variable part of the Kandle Formation and differs from the predominantly oolitic limestones of the Ojaküla Member (Cephalopod Limestone) in the occurrence of glauconite, e.g. in the surroundings of Jägala, or in the absence of ooids in some places or parts of the sequence. The thickness of the Malla Member decreases from 2.5 m in the eastern to 0.30 m in the central part of the klint area. West of Jägala, the Malla Member is missing and the Aseri Stage is represented by the 10–20-cm-thick sandy oolitic (mainly with phosphatic ooids) limestones of the upper, Ojaküla Member.

The Kandle Formation extends to central Estonia (Fig. 38) with the dominantly grey-coloured limestones of northern sections turning southwards brownish-grey or yellowish-grey. In southern Estonia, the stage is represented by the up-to-9m-thick reddish-brown limestones of the Segerstad Formation (Männil 1966, Männil & Meidla 1994). In the transitional area between the Kandle and Segerstad formations, reddish-brown and mottled limestones with occasional goethitic ooids (Männil 1990, Männil & Meidla 1994) have been distinguished. They belong presumably to the upper part of the Rokiškis Formation (Laškov et al. 1984). In practice, identification of the latter unit in sections seems in a great deal subjective, and its distribution area is difficult to determine.

The lower boundary of the Aseri Stage in recent use was defined by Orviku (Jaansoon-Orviku 1927, Orviku 1929). In contrast to Bekker (1922), he excluded from that stage the lowermost part of the oolitic limestones which comprises several early Ordovician (Oelandian) taxa (Ahtiella baltica Öpik, Antigonambonites sp., Megistaspis sp., Rõõmusoks 1970 p. 30). The boundary is marked by essential changes in the faunal composition, especially in trilobites and cephalopods (Rõõmusoks 1970, table 3, see also Jaanusson 1960a, 1963). Notable is the disappearance of the trilobite genus Megistaspis and appearance of Asaphus (Neoasaphus), represented at least by six species in northern Estonia (Rõõmusoks 1970; table 3). Asaphus platyurus, a characteristic species in the Segerstad Limestone in Sweden and Latvia (Jaanusson 1960a, Männil 1963b, 1966) occurs also in southern Estonia (Karula core, Männil 1966, fig. 12). Of new faunal elements, Echinosphaerites as a quite easily notable fossil is also worth of mentioning (Jaansoon-Orviku 1927, p. 15, 16; Orviku 1929, p. 9-11).

The data published on the distribution of ostracodes in the Aseri Stage in Estonia is scanty (Sarv 1959, Männil 1966). The ostracode Pinnatulites procera Zone of the Kunda Stage is replaced by the Piretella tridactyla Zone in the Aseri Stage (Meidla & Sarv 1990). In several core sections (Männil 1966, figs. 12-14), Euprimites effusus Jaanusson appears close to the lower boundary of the Aseri Stage.

The chitinozoans are known only in the grey-coloured rocks of the Kandle Formation which comprises the Cyatochitina regnelly and C. striata zones (Table 7). The boundary between these zones coincides with the boundary between the Malla and Ojaküla members.

The Aseri Stage corresponds to the lower part of the Didymograptus murchisoni graptolite zone and roughly to the Eoplacognathus suecicus conodont zone (Männil 1990, Männik & Viira 1990, Einasto et al. 1996).


Lasnamägi Stage

L. Hints


In northern Estonia, the fairly uniform Early Viru sequence of comparatively thick-bedded, hard bioclastic limestones abounding in discontinuity surfaces (Saadre 1992, 1993), was first distinguished as a separate unit - the Building Limestone (Baukalkstein), by Orviku (Jaansoon-Orviku 1927). Subsequently, this unit, determined mainly by the lithological criteria, was termed (Orviku 1940) the Lasnamägi Stage after the sections in the Lasnamägi quarry in the northeastern part of Tallinn. The Lasnamägi Stage is well-exposed also in some other sections, including Suhkrumägi (Photo 20) and Mäekalda (see Einasto et al. 1996, fig. A16) in the vicinity of the type section. In general lines, Orviku’s interpretation of the Lasnamägi Stage kept valid until the 1970s (Jaanusson 1945, Rõõmusoks 1960, 1970, Männil 1963a). In 1966, Männil (1966) stated that the Building Limestone comprises two distinct successive faunal associations of which the upper one with several characteristic trilobites, such as Xenasaphus devexus devexus (Eichwald), Asaphus (Neoasaphus) lepidus Törnquist, and graptolites including Gymnograptus linnarssoni (Moberg), is closely related to the fauna of the overlying argillaceous limestones of the Uhaku Stage. The beds containing the “upper” fauna were included (Männil 1966, 1976, Resheniya… 1978) to the Uhaku Stage, while the term Lasnamägi Stage was restricted to the lower half of the Väo Formation (Photo 21) in recent use (Männil & Rõõmusoks 1984), corresponding roughly to the Kallaste Substage by Rõõmusoks (1970), and to the lower part of the former Building Limestone.

The 4–10-m-thick Väo Formation (Fig. 39) is subdivided into three units; in ascending order these are the Rebala (the relatively argillaceous part, thickness up to 3 m), Pae (dolomites, up to 1.5 m) and Kostivere ( hard limestones, up to 6 m) members. Besides, in the stratotype area where the formation has a detailed bed-by-bed stratification, each layer has a name of its own given by quarry-workers (Mägi 1990, Einasto et al. 1996). Männil (1976) included to the Lasnamägi Stage the Rebala and Pae members and the lower part of the Kostivere Member, up to the discontinuity surface, above which Gymnograptus linnarssoni appears.

In northern Estonia, the lower, Lasnamägi part of the Väo Formation is up to 4.5 m and in the stratotype area at Lasnamägi up to 4 m thick (Männil 1976, Männil & Saadre 1987, Mägi 1990). Still in many sections the exact level of the upper boundary of the Lasnamägi Stage is not established and in Fig. 39 the total thickness of the Väo Formation is given. In southern Estonia, the Lasnamägi Stage is represented by red (lower part) to grey (upper part) bedded, mostly micritic limestones of the Stirna Formation (Ulst & Gailite 1976), equivalent to the Seby and Folkeslunda limestones of Öland Island and mainland Sweden (Männil & Meidla 1994). The Stirna Formation is up to 15 m thick (Fig. 39) which corresponds to the maximum thickness of the formation in Estonia and northwestern Latvia (Ulst et al. 1982, fig. 45). In the transitional area between the Väo and Stirnas formations in central Estonia, the oolitic lithofacies is developed (Põlma 1982, fig. 7).

In northern Estonia, the lower boundary of the Lasnamägi Stage falls into the upper part of the oolitic limestones, predominantly with goethitic ooids, of the Kandle Formation. It is marked by the discontinuity surface above which there appear brachiopods (Equirostra, Noetlingia), trilobites (Illaenus schroeteri (Schlotheim), Illaenus schmidti Nieszkowski, cephalopods (Lituites sp.) and others (Jaanusson 1945, Rõõmusoks 1970). The level of the appearance of phosphatic ooids in the top of oolitic limestones is used as the lower boundary of the Lasnamägi Stage, if the boundary discontinuity surface is absent or the palaeontological data are insufficient.

The lists of fossils published earlier for the Lasnamägi Stage (Rõõmusoks 1970, table 4) can be used with consideration that only the data from the Kallaste Substage by Rõõmusoks (1970) characterize the Lasnamägi Stage in recent meaning. In northern Estonia, the macrofauna is quantitatively dominated by sedentary forms, particularly articulate brachiopods and bryozoans (Jaanusson 1984). Cephalopods occur mostly in lower quantities, except northeastern Estonia where the lowermost beds of the stage abound in orthocones and where lituitids are also fairly common. The same groups of fossils seem to be relatively abundant in core sections as well (Hints & Põlma 1981).

Important information on the range of North Atlantic conodont zones (see Bergström, 1971 for the reference), graptolites and chitinozoans of the Lasnamägi Stage in northern Estonia is provided by Männil (Männil 1976, fig. 2; 1986, fig. 2.1.1). According to him, the stage is comparable to the Eoplacognathus foliaceus Subzone and the main lower part of E. reclinatus Subzone. Although the chitinozoans are not very dignostic for the distinction of the Lasnamägi Stage, the Cyathochitina sebyensis Zone is a good marker for the Aseri-Lasnamägi boundary beds (Table 7).


Uhaku Stage

L. Hints


The Uhaku Stage comprises, in the revised and amended form (Jaanusson 1960a, Männil 1966, 1976, 1990), the Caryocystites Zone (Jaansoon-Orviku 1927, = Uhaku Stage by Orviku 1940) and the upper part of the Building Limestone (Väo Formation). These two parts of the Uhaku Stage are considered also as substages (Männil 1976, 1990).

The thickness of the Uhaku Stage varies from 5-10 m in western to about 20-25 m in eastern Estonia (Fig. 39). In northern Estonia, the hard bioclastic limestones of the Väo Formation, forming the lower part of the Uhaku Stage, are of a rather stable thickness (4-5 m). The upper part of the Uhaku Stage, made up of relatively thin-bedded argillaceous limestones of the Kõrgekallas Formation, is subdivided (Table 7) into the Koljala, Pärtlioru and Erra members (Männil & Rõõmusoks 1984). The thickness of the formation decreases from about 18 m in northeastern to 1-2 m in northwestern Estonia (Figs. 39, 40). The lower boundary of the Kõrgekallas Formation and the Koljala Member, formed of argillaceous limestones with marly intercalations, supposedly coincides with the lower boundary of the Conochitina tuberculata Zone (Männil & Bauert 1986, p. 17; Table 7). In the Pärtlioru and Erra members, the argillaceous intercalations in the bioclastic limestones are partly kerogeneous. In the Oil Shale Basin in northeastern Estonia (Puura 1986), thin kukersite beds (up to 2 cm) occur, or they form together with limestones and marls distinct intervals (up to 1.6 m) between relatively pure limestones (Männil & Bauert 1986). These are the oldest kukersite beds in the Middle Ordovician sequence in northern Estonia. In central Estonia, the kukersite beds appear in the Kukruse Stage (cf. Männil 1966, 1986).

In southern Estonia and also in Latvia, the Uhaku Stage is represented mainly by micritic limestones with intercalation of bioclastic limestones and marls of the Taurupe Formation (= Furudal Formation in Männil 1966) with a thickness of 6 - 19 m. Only in a few sequences in western Latvia, the Taurupe Formation is over 20 m thick (Ulst et al. 1982, fig. 46). The limestones of the transitional belt between northern and southern Estonia are characterised by an interfingering pattern which resembles that of northern Öland in Sweden (Männil 1966). Both goethitic and phosphatic ooids occur in some places in the basal part of the Uhaku Stage testifying to continuous shift of the oolitic lithofacies in time (Põlma 1982, fig. 7, Pärnu and Ikla cores, Fig. 40).

In northern Estonia, the lower boundary of the Uhaku Stage falls into the lithologically rather uniform Väo Formation. In practice, a prominent discontinuity surface is used as a boundary marker above which several new taxa appear, some of them widespread and frequent throughout the Baltic Basin. Mass occurrence of the trilobite Xenasaphus d. devexus (Eichwald) is recorded from the Island of Osmussaar in the west as far as Ingria (L. Popov and R. Einasto, pers. comm., Alichova 1960, 1969) in the east. Of the graptolites of the Hustedograptus teretiusculus Zone, Gymnograptus linnarssoni (Moberg) is identified from the Oslo Region up to the Moscow Syneclise (Männil 1976). According to Männil (1986), the lower part of the Uhaku Stage corresponds to the Eoplacognathus robustus and E. lindstroemi subzones of the Pygodus serra Zone (Table 7). The upper part of the stage corresponds to the Pygodus anserinus Zone. The latter zonal species appears close to the base of the Kõrgekallas Formation. Nevertheless, on the basis of the distribution of conodonts, the boundary between the Lasnamägi and Uhaku stages is unclear, at least on the subzones level (Table 7). The chitinozoan Conochitina clavaherculi Subzone comprises the most part of the Väo Formation, including the strata with the first finds of G. linnarssoni (Männil 1986, fig. 2.1.1).

The Uhaku Stage comprises a varied sedentary benthic fauna, particularly articulate brachiopods, bryozoans and cystoids. Since there is no generally acknowledged interpretation of the Uhaku Stage, the lists of fossils presented by different researchers comprise taxa from different stratigraphical intervals (Rõõmusoks 1960, 1970; Männil 1963a, 1966).

Macrofossils are poorly known in the subsurface area of the Uhaku Stage (Fig. 40). The Taurupe Formation, which is distributed in southern Estonia, includes many elements, such as Nileus and Upplandiops (=Estoniops sp. n. in Männil 1966, fig. 12) among trilobites, and both Alwynella? and Christiania among articulated brachiopods, which are widely distributed in the Furudal limestones in Sweden (Jaanusson 1960a, 1963, Jaanusson & Ramsköld 1993).


Kukruse Stage

L. Hints


The Kukruse Stage (Kuckerssche Schicht by Schmidt 1879, 1881) as a stratigraphical unit comprises the commercially exploited oil shale (kukersite) seams (Chapter X) and the richest and most diverse faunal assemblage in the Ordovician of Estonia represented by more than 330 species and subspecies (Rõõmusoks 1970, table 10).

The stratigraphy of the Kukruse Stage has been dealt with in several papers (Rõõmusoks 1957, Männil 1984, Männil & Bauert 1984, 1986, Bauert 1993, Saadre & Suuroja 1993b). The bed-by-bed stratification of the kukersite complex with special sets of indices for the individual kukersite seams form the base for the correlation of the sequences within the kukersite basin (Männil 1984, fig. 2; Bauert & Puura 1990).

The thickness of the Kukruse Stage (Fig. 41) ranges from about 3 m in western to more than 20 m in eastern Estonia (Saadre & Suuroja 1993a). The stage consists of three lithologically distinct formations. The argillaceous bioclastic limestones with intercalations of kukersite (oil shale) and marls of the Viivikonna Formation (Männil & Rõõmusoks 1984) are distributed northeast of the line Osmussaar Island (southwestern Estonia) - Mehikoorma (south coast of Lake Peipsi) (Fig. 41). Based on the frequency of kukersite seams or the content of the kerogenous component, the Viivikonna Formation is subdivided into the Kiviõli, Peetri and Maidla members (Fig. 42). The Kiviõli (lower) and Peetri (upper) members differ from the Maidla (middle) Member by the occurrence of 10—14-cm-thick kukersite seams, while the middle part of the formation consists of kerogenous and variously argillaceous limestones (Männil et al. 1986, Bauert 1993, figs 3, 4). Due to the facies shift of the kukersite beds (Männil et al. 1986), the boundaries of the Viivikonna Formation are diachronous. As a result, the upper part of the Viivikonna Formation (Peetri Member) is missing in northeastern Estonia, but it is exposed in the vicinity of Tallinn (Fig. 42, Nõlvak & Hints 1996) and is well-known by core sections south of the outcrop area (Männil 1984, Männil & Bauert 1984, Männil & Saadre 1987).

Westwards, the Viivikonna Formation grades into the bioclastic limestones of the Pihla Formation with a thickness of about 3 - 6 m (Saadre & Suuroja 1993b) and southwards into the limestones with dark pyritized skeletal detritus and nodular intercalations of argillaceous marls of the Dreimani Formation (Fig. 41, Springis 1974). The thickness of the latter varies from 7 to 14 m and only in southeastern Estonia it is about 20 m, which is nearly the same as in eastern Latvia (Ulst et al. 1982, fig. 47).

For the lower boundary of the Kukruse Stage, Bekker (1923, 1924b) proposed the base of the lowermost commercially important kukersite seam “A” at the base of the Viivikonna Formation. The renovation of faunal association begins with the appearance of new bryozoans in seam “A”. Somewhat higher, in seam “C” several new species, including the brachiopods Bilobia musca (Öpik), Sowerbyella (S.) liliifera (Öpik), Estonomena estonensis (Bekker), and the trilobites Asaphus (Neoasaphus) nieszkowskii Schmidt, Estoniops exilis (Eichwald), Paraceraurus aculeatus (Eichwald) appear (Rõõmusoks 1970, table 9). In western and southern Estonia, the base of the Pihla or Dreimani Formation is used as the lower boundary of the Kukruse Stage. This level is marked by the appearance of indicator ostracodes Baltonotella kuckersiana (Bonnema), Conchoprimitia leperditioides Thorslund, Euprimites locknensis Thorslund and others, several of which are common with the lower part of the Dalby Limestone in Sweden (Männil 1966, Jaanusson 1976). At the same time, several early Viru taxa, such as Chasmops odini odini Eichwald, Sowerbyella (Viruella) uhakuana (Rõõmusoks), Platystrophia biforata (Schlotheim), Dianulites fastigiatus (Eichwald) and others, disappear close to the lower boundary of the Kukruse Stage (Rõõmusoks 1970, p.156, 157). The graptolite Orthograptus uplandicus whose range zone corresponds to the Kukruse Stage (Männil 1984) and the chitinozoa Cyathochitina savalaensis appear roughly on the lower boundary of the Kukruse Stage. In all likelihood, also the boundary between the North Atlantic conodont anserinus and tvaerensis zones (Männil & Bauert 1986) falls into the lower part of the Kukruse Stage.

The diverse assemblage of Kukruse macrofossils is represented first of all by bryozoans (more than 60 species), brachiopods (about 90 species) and trilobites (about 50 species: Rõõmusoks 1970, table 10) which form about two thirds of the species identified. The most abundant and diverse association occurs in the Kiviõli Member in the lower part of the stage. Still some species, such as Hesperorthis inostrantzefi inostrantzefi (Wysogorski), Echinosphaerites aurantium suprum Hecker, are notable due to their mass occurrence in the upper part of the stage (Rõõmusoks 1970, p. 169). The character of the distribution of some brachiopods and trilobites, such as Estlandia marginata magna Öpik, Otarion planifrons (Eichwald), Pharostoma nieszkowskii (Schmidt) and others, shows a facies shift from the lower part of the Kukruse Stage (Kiviõli Member) in northeastern to the upper part (Peetri Member) in northwestern Estonia.

In the core sections, macrofossils are of secondary importance due to their scarcity, especially in western Estonia (Fig. 40). Still, the occurrence of some species should be noticed. In some northernmost core sections, the brachiopod Kullervo panderi (Öpik) marks the lowermost part of the Kukruse Stage (Rõõmusoks 1970). In the outcrops, this species appears presumably in the kukersite seam “G”, which lies 1-4 m above the lower boundary of the stage. In the southern periphery of the Viivikonna Formation and in the Dreimani Formation, Asaphus (Neoasaphus) ludibundus Törnquist and Bilobia musca (Öpik) appear in the Kukruse Stage and in some areas Echinosphaerites becomes frequent.


Haljala Stage

L. Hints


Jaanusson (1995) proposed the term Haljala Stage for the unit which comprises the Idavere and Jõhvi chronostratigraphical subdivisions, previously regarded as separate stages. These two subdivisions, now classified as the Idavere and Jõhvi substages (Table 7), comprise most of K‑bentonite beds which lie below the thickest bed (“d” by Jürgenson 1958a) established in eastern Baltic. The substages are difficult to differentiate in southern Estonia, in areas where K-bentonite beds are uncertain or absent. Also the faunal distinction between the substages is rather inconsiderable (Põlma et al. 1988, figs. 9-11). In Estonia, the thickness of the Haljala Stage varies mostly from 10 to 20 m (Fig. 43).

The Idavere Substage comprises the regularly bedded hard bioclastic limestones of the lower, Tatruse Formation and argillaceous limestones with intercalations of marls and some thin K-bentonites of the upper, Vasavere Formation (emended by Männil & Meidla 1994). This substage has the most reduced sequence in northern Estonia and in some places in the surroundings of Tallinn it is entirely missing (Jaanusson 1945). The Tatruse Formation (Põlma et al. 1988) corresponds roughly to Schmidt’s (1881) original concept of “Itfersche Schicht” and the fauna recorded by him and his contemporaries from the “Itfer” belongs only to this formation. The Vasavere Formation contains usually two, but in the west sometimes up to 18 K-bentonite beds, which belong to the Grefsen complex of bentonites (Bergström et al. 1995). In the areas where only beds “a” and “b” (Jürgenson 1958a) are recognizable, the upper bed is regarded as the top of both the Idavere Substage (Männil 1963a) and the Vasavere Formation. In areas farther south where bentonite beds of the Grefsen complex disappear or in the west where they are numerous, a distinction between the Idavere and Jõhvi substages is difficult.

The Jõhvi Substage, which is at its thickest (more than 10 m, Fig. 44) in northwestern Estonia, comprises argillaceous bedded to nodular limestones with argillaceous intercalations in the middle part (Männil & Rõõmusoks 1984, Põlma et al. 1988). These limestones form the lower part of the Kahula Formation (Männil & Meidla 1994). A fairly persistent K‑bentonite bed (bed “c” by Jürgenson 1958a, “Sinsen K-bentonites” by Bergström et al. 1995) occurs close to the boundary between the middle and upper parts of the Jõhvi Substage.

In southern Estonia, the Haljala Stage, 8 - 18 m in thickness, is represented by argillaceous limestones with thin K-bentonite beds and in places with phosphatic ooids of the Adze Formation (Ulst et al. 1970).

In the outcrop area, the lower boundary of the Haljala Stage and the Tatruse Formation is formed by a conspicuous discontinuity surface – a hardground which in places is penetrated by cavities, some 5 cm or even more in diameter at the surface and extending sometimes some 40 cm downwards (Põlma et al. 1988). The limestones above the basal discontinuity (Kisuvere Member) comprise up to 16% of quartz sand. The most detailed stratification of the lowermost beds of the Haljala Stage is based on chitinozoans. The oldest part of the stage, the Armoricochitina granulifera and Angochitina curvata zones (Männil 1986, fig. 5.1.1, Nõlvak & Grahn 1993) occurs in the Laeva area in eastern central Estonia (Fig. 45) where the stage has the maximum thickness (ca. 25 m, Fig. 44, core No. 285). These two zones do not occur in the northenmost sequences, where the Lagenochitina dalbyensis Zone forms the basal part of the substage.

The gap on the boundary between the Kukruse and Haljala stages is rather well expressed by differences between faunas, especially of brachiopods and trilobites in northern Estonia. Both these groups are represented in the Idavere Substage with about 40 species, a few of which occur also in the underlying Kukruse Stage (Rõõmusoks 1970, table 12). The occurrence of several Kukruse bryozoans in the upper part of the Idavere Substage (in the Vasavere Formation) is seemingly of facies origin. The lower boundary of the Haljala Stage is also marked by rather sharp changes in the composition of ostracodes (Põlma et al. 1988, figs. 7, 9-11), though some typical Idavere - Jõhvi species, e.g. Braderupia asymmetrica (Neckaja) appear in the top of the Kukruse Stage. The frequent occurrence of Leiosphaeridia above it, is a rather good marker for the lower boundary of the Haljala Stage. The base of the Haljala Stage is close to both the graptolite Diplograptus multidens Zone and the conodont Baltoniodus gerdae Subzone (Männil 1990, Jaanusson 1995).

The changes in the faunal composition on the transition between the Idavere and Jõhvi substages are continuous which is clearly revealed in core sections, especially by ostracodes (Põlma et al. 1988). Several new macrofossil taxa, such as Toxochasmops maximus (Schmidt), Clinambon anomalus (Schlotheim), presumably appear somewhat higher (1.5 - 2 m, Männil 1963a, b) of the boundary K-bentonite bed between the Idavere and Jõhvi substages.

In the core sections located far away from the outcrop area, the alga Mastopora concava (Eichwald), spicules of Pyritonema subulare (Roemer) and also some brachiopods (Bilobia) occur (Fig. 40), but there is no characteristic species among macrofossils for determination of the lower boundary of the Haljala Stage.


Keila Stage

L. Hints & T. Meidla


In most of northern Estonia, the Keila Stage (Kegelsche Schicht, Schmidt 1881) comprises the argillaceous bioclastic limestones, with intercalations or occasionally thicker (up to 4 m) intervals of relatively pure limestones of the Kahula Formation (Table 7). Only in a restricted area in northwestern Estonia, the upper part of this formation is replaced by the Vasalemma Formation where the greatest thickness of the Keila Stage (more than 30 m) has been recorded (Fig. 46).

Initially, due to the unclear relationship between the fossilifereous argillaceous limestones of the Keila Stage and the Schmidt‘s “Wassalem’sche Schicht”, the term Keila-Vasalemma Stage was introduced by Bekker (1922, see also Öpik 1930b). Later, Jaanusson (1945) and Männil (1958c, 1963b, 1966) subdivided the Keila Stage into several members and defined the lower boundary of the stage on the level of the thickest K-bentonite (bed “d” by Jürgenson 1958a, see also Jaanusson & Martna 1948, Vingisaar 1972). The composition of the Kahula Formation and the distribution of members overlying the boundary K-bentonite is shown in Figure 47.

The lowermost part of the Keila Stage (Kurtna Member) is represented by argillaceous limestones. The Kurtna Member is overlain by relatively pure limestones, in places with argillaceous intercalations of the Pääsküla Member. This unit, although differently understood by stratigraphers (Nõlvak 1996), can be identified in the core sections of northwestern Estonia as a complex of biomicritic limestones, up to ca. 7 m in thickness (Põlma et al. 1988, Fig. 47). It may be replaced by intercalating argillaceous bioclastic and biomicritic limestones with a thickness of up to 20 m (Ainsaar 1991), seemingly corresponding to a longer time interval than the Pääsküla Member in the sense of Põlma et al. (1988).

The younger part of the Keila Stage comprises the Saue and Lehtmetsa members, the fossiliferous argillaceous limestones and detrital marls with thin layers of argillaceous limestones, respectively. Contemporaneously, the formation of carbonate buildups (interpreted as reefs, Raymond 1916, or bioherms, Männil 1960, or mud mounds, Põlma & Hints 1984) has been developed in northwestern Estonia. They belong to the Vasalemma Formation, nearly constituting the upper half of the Keila Stage in the surroundings of Keila - Vasalemma, whereas a distinct eastward shift of the corresponding facies is recorded during late Keila time (Fig. 48). The Vasalemma Formation, in thickness up to 15 m, consists of several principal lithotypes (Männil 1960, Põlma 1967, Hints 1996). The most characteristic type of rock is the bedded grainstone (cystoid limestone), which is intercalated with clayey limestones in the lower part of the formation. Cystoid limestone consists mainly of skeletal sand particles aggregated with pure calcite cement (content of terrigenous material less than 3%). The grainstones contain irregular buildups, measuring up to 10 m vertically and up to 300 m horizontally and consisting of pure limestones with a low content of skeletal sand (less than 10%) and terrigenous material (up to 6%), occasionally with inclusions of fossiliferous marls. The buildups mostly lack the reef-like framework and are considered as carbonate mounds. Still, in some “mounds” the edrioasteroid Cyathocystis rhizophora Schmidt is frequent and may form frame-like structures. The lower and middle parts of the Vasalemma Formation contain a number of species common with the Kahula Formation - Estlandia pyron silicificata Öpik, Clinambon anomalus (Schlotheim), Horderleyella? kegelensis (Alichova), Sowerbyella (S.) cf. forumi Rõõmusoks a.o., which indicate the Keila Age of corresponding rocks.

In southern Estonia, the Keila Stage comprises the upper part of the bioclastic limestones of the Adze Formation and clayey limestones and marls, which contain some species common with the Blidene Formation in Latvia (Ulst et al. 1982). The lowermost part of the Mossen Formation may also be of Keila Age (Table 7, Meidla 1996). Still, the correlation of the Lukštai and Blidene formations with a unit of siltstone and silty limestone identified in southern Estonia (Ainsaar 1995) needs to be adjusted. Due to this uncertainty, the identification of the Keila Stage is complicated in the transition between the distribution areas of the Kahula and Adze formations.

The total thickness of the Kahula Formation may exceed 30 m, and in northwestern Estonia its main part corresponds to the Keila Stage. In general, the thickness of the Keila Stage part of the formation (mostly 10-15 m) decreases in the southeast direction. In the same direction, the formation becomes lithologically more uniform and argillaceous. In southern Estonia, the thickness of the equivalents of the Keila Stage presumably does not exceed 10(?) m.

A rich and diverse fauna of bryozoans, brachiopods, trilobites, echinoderms and other sedentary and vagile groups (see Rõõmusoks 1970) is distributed in the Kahula Formation. In the upper part of the formation, corresponding to the Keila Stage, several macrofossil taxa are common with the Haljala Stage, but a specific component in this particular association comprises last representatives of several brachiopod genera (Clinambon, Cyrtonotella), trilobites (Asaphus (Neoasaphus) nieszkowskii Schmidt and Toxochasmops maximus (Schmidt)), crinoids (Ristanacrinus marinus Öpik and different baltocrinids) or species characteristic of the Keila Stage only Keilamena occidens (Männil), Longvillia asmusi (Verneuil), Horderleyella? kegelensis (Alikhova). The data on macrofauna come mostly from northern Estonia. In southern Estonia, the Keila Stage is characterised by a brachiopod - trilobite association, which comprises several taxa (Skenidioides, “Sampo”, Eoplectodonta), appearing on a higher stratigraphical level in northern Estonia or being related to the Scandinavian faunas.

The Keila Stage presumably corresponds to the uppermost part of the Diplograptus multidens and the lowermost part of the Dicranograptus clingani graptolite zones (Männil 1990). The lower boundary of the stage, the level of the K-bentonite bed “d” corresponds to the lower boundary of the chitinozoa Angochitina multiplex Subzone (Table 7) and is close to the Northern Atlantic conodont superbus Zone (Männik & Viira 1990).


Oandu Stage

L. Hints & T. Meidla


In northern Estonia, the Oandu Stage comprises rocks of two different lithofacies forming the Vasalemma and Hirmuse formations. The Vasalemma Formation, distributed in northwestern Estonia, consists of bedded fine- to coarse-grained bioclastic limestones with irregular bodies of aphanitic massive limestones (carbonate buildups). These rocks were identified first as the Hemicosmites Limestone (Eichwald 1854a) or Wasalemm’sche Schicht (Schmidt 1881). Vasalemma, as the name of a chronostratigraphic unit was applied also to the rocks of another lithofacies – the argillaceous limestones and marls, named Oandu beds by Öpik (1933) and the Hirmuse Formation by Männil and Rõõmusoks (1984), which are exposed on the banks of the Oandu River in northeastern Estonia (Rõõmusoks 1953, Aaloe et al. 1958). Later studies (Männil 1958c, 1960) showed that the lower and middle parts of the Vasalemma Formation are of Keila Age (Fig. 48) and the name Oandu was proposed for chro-nostratigraphic unit of post-Keila Age. The Oandu Age of the uppermost Vasalemma Formation is presumed by the appearance of the corals Lyopora tulaensis (Sokolov), Eofletcheria orvikui Sokolov and the brachiopods Rhynchotrema? parva Oraspõld, Rostricellula nobilis (Oraspõld), Dactylogonia luhai (Sokolskaya) (Männil 1960, Rõõmusoks 1970), or it is supposed by the disappearance of Leiosphaeridia and brachiopods of the Keila Stage (Rummu core, Põlma et al. 1988). In northern Estonia, the Oandu Stage is restricted in thickness (1-4 m, Fig. 49); only in the limits of the Vasalemma Formation it is up to 6 m thick.

In the stratotype area in northeastern Estonia, the lower boundary of the Oandu Stage and the Hirmuse Formation, is known only by the core sections where it is marked by a sharp discontinuity surface with up-to-35-cm-deep pockets, on the upper boundary of the Kahula Formation (= Kahula Group, Männil & Meidla 1994). Below this level, a great number of Middle Ordovician species and even genera common with several older stages, including the brachiopods Cyrtonotella, Estlandia, trilobites Asaphus (Neoasaphus) nieszkowskii Schmidt, Pseudobasilicus, ostracodes Tetrada (Tetrada) harpa (Krause), Polyceratella spinosa Sarv (Fig. 50) disappear. Notable is the disappearance or sharp decrease in the frequency of the acritarch Leiosphaeridia which is abundant in the Keila Stage (Fig. 51). This fossil seemingly can be used for the preliminary establishing of the above-mentioned boundary in core sections, especially when the uppermost part of the Kahula Formation is more argillaceous and possibly belongs to the Lehtmetsa Member of late Keila Age (Fig. 47, Põlma et al. 1988, fig. 32). A new complex of fossils with the ostracodes Bolbina rakverensis (Sarv), Klimphores minimus (Sarv), Disulcina perita perita (Sarv), brachiopods Howellites wesenbergensis (Alichova), Equirostrata wesenbergensis (Teichert) appears near the lower boundary of the Oandu Stage (Fig. 50, see also Põlma et al. 1988). In some cases, these species are found even below the boundary discontinuity surface, seemingly they occur in the deep pockets filled with deposits of Oandu Age. Due to the essential changes in the faunal composition (Männil et al. 1966, Hints et al. 1989), several authors have suggested to use the lower boundary of the Oandu Stage as the regional subseries or series boundary (Jaanusson 1945, Rõõmusoks 1956).

The Hirmuse Formation thins out within a rather short distance in the southern direction, and in many places in central Estonia the Oandu Stage is represented only by the Tõrremägi Member of the Rägavere Formation with a thickness less than one metre (Fig. 49). This area separates the rich and diverse fauna of bryozoans, brachiopods, echinoderms and trilobites of the Hirmuse Formation in northern Estonia (Põlma et al. 1988) from the relatively rich brachiopod and trilobite fauna in the marls and argillaceous limestone in southern Estonia, corresponding presumably to the Lukštai Formation. Beside some species common for northern and southern Estonia (Howellites wesenbergensis Alichova, Rhactorthis kaagverensis Hints a.o.), several brachiopods (Reushella magna Hints, Laticrura sp. Skenidioides sp., Leptellina? sp.) have been identified only in the latter region and are also known in the Lukštai Formation in Lithuania or in the Moldå Limestone in Sweden (Jaanusson 1982, fig. 7).

Identification of the Oandu Stage is most complicated in southeastern Estonia where the black shales and the overlying marls of the Mossen Formation are distributed (Karula core in Fig. 51). The shales, encountered in various sections have been included to the Keila (Meidla 1996), Oandu (Hints 1975) or Rakvere (Männil 1966) Stage. The overlying marls of the Priekule Member are correlated with the uppermost Oandu and/or the Rakvere Stage. In some sections, the marls below the black shale (about 4 m in the Karula core, Fig. 51) comprise brachiopods known in the other sections (Otepää, Laeva) mainly in the beds presumably of Oandu Age. In favour of this age testifies also the disappearance of Leiosphaeridia at a depth of about 3 m below the shales. At the same time, the ostracode record allows to suppose Keila Age of the lowermost part of the Mossen Formation (Meidla 1996). The contradiction in interpreting the age by macrofossils and ostracodes may be caused by insufficient data available, or it may indicate the patchy distribution of deposits in the Keila - Oandu boundary interval in southeastern Estonia.

Still, in most of southern Estonia, the Oandu Stage can be identified most realiably on the basis of ostracodes. The lower boundary of the Oandu Stage is marked by the appearance of Sigmoopsis granulata Sarv, Bolbina rakverensis Sarv, Pelecybolbina illativis Neckaja and Klimphores minimus (Sarv), and a general rapid faunal change which occurred throughout the Estonian part of the palaeobasin (Meidla 1996).


Rakvere Stage

L. Hints & T. Meidla


The “Wesenbergsche Schicht” by Schmidt (1881) corresponds roughly to the Rakvere Stage in nowadays understanding (Männil 1958b, 1963a, Kõrvel 1962, Põlma et al. 1988). In northern Estonia, the Rakvere Stage forms the lowermost, relatively thick part of Late Viru and Harju pure micritic (aphanitic) limestones which intercalate with more or less argillaceous varieties. The cycles of different lithotypes generally constitute distinct lithostratigraphical units (Põlma 1982, Hints et al. 1989), whereas the clayey parts of the cycles are characterized by the appearance of abundant new taxa .

The Rakvere Stage consists of the Piilse and Tudu members (Kõrvel 1962) which form the main part of the Rägavere Formation. The stage is at its thickest (28 m) in western Estonia (Fig. 52) and it thins notably in the southeastern direction. The lower, Piilse Member with a thickness of up to 27 m (Rõõmusoks 1983) consists of pure, in places dolomitized micritic limestones with a low content of terrigenous material (3 - 9%) and skeletal sand (less than 5%, Kõrvel 1962, Põlma et al. 1988). The member is characterised by a distinct pyritic pattern, following abundant burrows in the former sediments. The upper, Tudu Member is up to 10 m thick and differs from the Piilse Member in the higher content of skeletal sand (commonly 15 %, Põlma et al. 1988), in the occurrence of thin, up to 3 cm thick kukersite layers and rare and weakly developed pyritic patterns.

Southwards the limestones of the Rägavere Formation become more argillaceous and in southern Estonia they are supposedly replaced by the carbonate marls of the Priekule Member in the upper part of the Mossen Formation (Männil & Meidla 1994, Meidla 1996). On the basis of chitinozoan distribution (Nõlvak & Grahn 1993) it is supposed that in some places (Ruhnu and Ohesaare cores) the Rakvere Stage is missing.

The data on the distribution and composition of macrofossils in the Rakvere Stage, particularly in the Tudu Member is scanty due to relatively few outcrops. The Rakvere Stage with the relatively sparse macrofauna of bryozoans, brachiopods and trilobites is characterized by frequent and diverse association of calcareous algae (Cyclocrinites, Rhabdoporella etc., Kõrts et al. 1990). From the Rakvere Stage up to the end of Ordovician, calcareous algae and their fragments dominate in the composition of skeletal particles (Põlma 1972, 1982) where they may account even for 97.6%.

Unlike macrofossils, the ostracode record of the Rakvere Stage is rich, comprising more than 80 species (Meidla 1996). Several distinct associations have been recorded in the lower part of the stage corresponding approximately to the Piilse Member of the Rägavere Formation (Meidla 1996, fig. 47). Valuable is the ostracode record of the upper part of the stage which contains only sparse macrofauna. This interval, nearly equal to the Tudu Member, corresponds to the Daleiella admiranda Subzone (subzone of Daleiella sp. n. in Meidla & Sarv 1990, Table 10), a range zone prominent in the sections of northern and central Estonia. The appearance of several long-ranging taxa, such as Steusloffina cuneata (Steusloff), Medianella blidenensis (Gailite), Pullvillites laevis (Abushik & Sarv ), etc., has been recorded within this interval (Fig. 50).

The lower boundary of the Rakvere Stage is lithologically more or less distinct in northern Estonia where it coincides with the pyritized discontinuity surface on the top of the Tõrremägi Member in the lower part of the Rägavere Formation. The appearance of several new brachiopods, including Microtrypa estonica Rõõmusoks, Platystrophia lutkevichi satura Oraspõld, P. quadriplicata Alichova, Sowerbyella (Sowerbyella) raegaverensis Rõõmusoks, Vellamo wesenbergensis (Pahlen) and trilobites Chasmops wesenbergensis (Schmidt), Encrinuroides seebachi (Schmidt), Pharostoma pediloba (Roemer) and others, above this boundary shows the renovation of faunal associations. However, macrofossils are very scarce, particularly in core sections, and cannot be used for the purposes of detailed biostratigraphy. The same applies to ostracodes, because most of the species characteristic of the Late Ordovician ostracode fauna appear below this boundary, in the Tõrremägi Member of the Oandu Stage which represents a facies similar to the Rakvere Stage. Also the zonal chitinozoa Fungochitina fungiformis, characteristic of the Rakvere and Nabala stages, appears in the Tõrremägi Member. The suggestion to include the Tõrremägi Member to the Rakvere Stage (Meidla 1996) follows partly the earlier wider interpretation of that stage, according to which the Oandu beds by Öpik were included to the Rakvere Stage (Jaanusson 1945, Alichova 1960).

The Rakvere Stage corresponds roughly to the lower part of the graptolite Pleurogratus linearis Zone and the chitinozoan Cyathochitina angusta Subzone of the Fungochitina fungiformis Zone (Nõlvak & Grahn 1993, Table 7).


Harju Series

Nabala Stage

L. Hints & T. Meidla


The Nabala Stage was distinguished by Männil (1958b) as the lower part of the Schmidt’s (1858) “Lyckholm’sche Schicht”. He subdivided the stage into the lower, Paekna and the upper, Saunja substages (see also Öpik & Laasi 1937, Jaanusson 1994). Nowadays they are used as lithostratigraphical units (formations) which represent the Nabala Stage with a total thickness of 10 to 35 m in northern and partly in central Estonia (Fig. 53). The Paekna Formation is up to 16 m thick and comprises predominantly argillaceous bioclastic limestones intercalating with micritic limestones. The thickness of micritic interlayers is usually about 0.1 m, but occasionally it may reach 1-3 m. The up-to-28-m-thick micritic limestones of the Saunja Formation are lithologically uniform and occur all over Estonia. Their thickness decreases towards the south until it is only 0.3 m (Meidla 1996). South of the Muhu - Mustvee line, the lower Paekna Formation is replaced by the Mõntu Formation. This 3–7-m-thick complex consists of argillaceous bioclastic limestones with rare thin (5–30 cm) layers of micritic limestones containing glauconite (Oraspõld 1995).

On the transition from the Rakvere to the Nabala Stage the shelly fauna undergoes notable renovation. Of about 150 species and subspecies occurring in the Nabala Stage, only one third is common with the Rakvere Stage (Männil et al. 1966). Several new species of brachiopods, including Bekkeromena semipartita (Roemer), Ilmarinia sinuata (Pahlen), Laticrura rostrata Hints, Sulevorthis lyckholmiensis (Wysogorski), Pseudolingula quadrata (Eichwald), appear in the Nabala Stage. Some of these species are missing in the Saunja Formation, but appear again in the overlying Vormsi Stage. The micritic limestones of the Saunja Formation contain a notably abundant and diverse fauna of molluscs (about 30 gastropod and more than 10 cephalopod species).

The lower boundary of the stage is exposed only in the Paekna quarry (Nõlvak & Meidla 1990). It is marked by a series of uneven discontinuity surfaces, above which there appears a new association of chitinozoans, including the zonal Armoricochitina reticulifera (Grahn). The latter can be used as the most reliable fossil for the identification of the lower boundary of the Nabala Stage in the core sections in central and southern Estonia.

The composition of ostracodes changes remarkably on the Rakvere - Nabala transition. Several new taxa, including Disulcina perita explicata Sarv, Tetrada neckajae Meidla, Oepikella luminosa Sarv a.o., appear in the lower part of the Nabala Stage, but mainly somewhat higher of the lower boundary of the Paekna Formation (Fig. 50). For this reason the lower boundary of the stage is marked better in the ostracode record by the disappearance of the species Disulcina perita perita (Sarv) and Daleiella admiranda Meidla (a zonal species) in the uppermost part of the Rakvere Stage (Meidla 1996). Brachiopods are found mostly in the lower Paekna Formation and among them new faunal elements Pseudolingula quadrata (Eichwald) and Sulevorthis lyckholmiensis (Wysogorski) appear close to the lower boundary of the Nabala Stage (Fig. 54).

The Nabala Stage corresponds to the middle part of the Pleurogratus linearis graptolite Zone (Table 7) and the upper part of the North Atlantic superbus conodont Zone. In the ostracode record, the summary differences between the Paekna and Saunja formations are not significant, but the very uneven distribution of ostracodes in the Saunja Formation should be mentioned (Meidla 1996).


Vormsi Stage

L. Hints & T. Meidla


The Vormsi Stage (Jaanusson 1944b, = middle part of the Lyckholm’sche Schicht, Schmidt 1858) consists of a facies succession of bioclastics limestones (Kõrgessaare Formation, up to 21 m) in northern Estonia, argillaceous limestones with glauconite (Tudulinna Formation, up to 17.1 m) in central Estonia and black shales (Fjäcka Formation, up to 4.5 m) in southern Estonia. The thickness of the stage decreases from 10 - 20 m in northern Estonia to 0.3 m in southern Estonia (Ohesaare core, Fig. 55). In the transitional area between the Kõrgessaare and Tudulinna formations, interfingering of these units can be followed (Oraspõld 1982a, figs. 3, 4).

The association of the diverse shelly fauna of corals, bryozoans, brachiopods, molluscs and trilobites includes some 200 species (Rõõmusoks 1967) in northern Estonia in the Kõrgesssaare Formation. Southwards this fauna is replaced by a specific and less diverse association.

The Tudulinna Formation is characterized by an association of brachiopods, comprising species of the genera Dicoelosia, Christiania, Skenidioides, Leptellina?, and a facies dependent ostracode association prevailed by Uhakiella curta Sidaraviciene, Medianella blidenensis (Gailite) and Rectella nais Neckaja (Meidla 1996). The Fjäcka Formation comprises a brachiopod association typical of shally facies, consisting mainly of inarticulated small brachipods Paterula, Hisingerella a.o., and of a few articulate brachiopods, such as “Sericoidea” and Onniella (Fig. 56). In general, the association is similar to that in the Mossen Formation.

The lower boundary of the Vormsi Stage coincides with a lithologically sharp boundary in most of Estonia. Above that boundary the frequency of macrofossils and ostracodes increases notably. In the Kõrgessaare Formation, several new species appear, including the corals Proheliolites dubius (Schmidt), Kenophyllum siluricum (Dybowski), Streptelasma (Streptelasma) distinctum Wilson, the brachiopods Eoplectodonta schmidti (Lindström), Equirostra gigas Schmidt, Sampo hiiuensis Öpik, Glyptorthis plana Oraspõld, Triplesia insularis (Eichwald), the first Dicoelosia (Wright 1968) and the trilobites Encrinurus moe Männil a.o. Ostracodes are dominated by the species common with the older strata (Fig. 57). The earliest conodonts of the ordovicicus Zone occur also in the Vormsi Stage (Männik1992b).The zonal species Amorphognathus ordovicicus has been recorded from the basal part of the Kõrgessaare Formation and also from the Tudulinna Formation.

In spite of distinct lower and upper boundaries of the Vormsi Stage, the detailed correlation of the formations belonging to this stage is not yet very clear. The distribution of zonal chitinozoans allows to suppose that the oldest part of the Vormsi Stage is missing in central and southern Estonia (Nõlvak & Grahn 1994). In these areas the Vormsi Stage corresponds to the Tanuchitina bergstroemi Zone which forms the upper part of the stage in northern Estonia, overlying the Fungochitina fungiformis Zone (Table 7) of the lower part of the stage in this area.

The distribution of ostracodes (Meidla 1996) does not support the correlation schemes based on chitinozoans.

The topmost part of the Vormsi Stage correponds to the chitinozoa Acanthochitina barbata Subzone (Table 7). The level of the disappearance of the index species marks well the traditional upper boundary of the Vormsi Stage.


Pirgu Stage

L. Hints & T. Meidla


The Pirgu Stage (Jaanusson 1944b) is a lithologically variable (Table 7) and thick (up to 66 m, Fig. 58) stratigraphical unit (=upper part of the “Lyckholm’sche Schicht”, Schmidt 1858). The notable changes in the thickness of the stage, sometimes within a short distance, are due to various reasons, e.g. the development of mud mounds, denudation during late Pirgu and/or Porkuni time, intensive tectonical movements and changes in sea-level.

In northern Estonia, within the stage two succesive rock units of grey-coloured limestones are distinguished - the lower, Moe and the upper, Adila Formation (Rõõmusoks 1960) which correspond approximately to the former Nyby (Jaanusson 1944b, Männil 1966) and Piirsalu (Jaanusson 1945) substages.

The Moe Formation, up to 40 m in thickness, consists of micritic and bioclastic nodular or bedded limestones with argillaceous intercalations. The lower part of the formation contains abundant calcareous alga Palaeoporella? (=Dasyporella in Männil 1966) and in some places (Hoitberg on Vormsi Island, Võhma core in central Estonia) typical carbonate mounds are developed, quite similar to the Boda mounds in the Siljan district of Sweden.

The Adila Formation comprises predominantly bioclastic limestones with a thickness of 10-15 m. Numerous discontinuity surfaces and cyclically alternating pure and argillaceous limestones are characteristic to the upper part of the formation. In this topmost part of the formation, the pentamerid brachiopod Holorhynchus has been recorded in the Island of Hiiumaa (Hints 1993). The boundary between the Moe and Adila formations coincides with the boundary between the rugata and bergstroemi chitinozoan zones (Table 7).

Within a rather large area in central Estonia, the Pirgu Stage is characterized by the interfingering of different rock units (Oraspõld 1975a, Table 7), whose correlation with the northernmost and southernmost sequences is complicated. In this transitional area, the lowermost part of the stage consists of argillaceous bioclastic limestones with glauconite (Tootsi Member), overlying the upper part of the Vormsi Stage, corresponding to the chitinozoan barbata Zone. The Tootsi Member contains a distinct association of shelly fauna (Fig. 59). This unit is succeeded upwards by the grey-coloured, sometimes red mottled marls and highly argillaceous limestones of the main part of the Halliku Formation whose relationship with the Moe Formation is not yet very clear (Männil & Meidla 1994). The diverse and abundant ostracode fauna of the Halliku Formation comprises taxa (Fig. 60) common with the upper part of the Moe Formation (Meidla 1996). Most uncertain is the age of the Kabala Member on the transition between the Pirgu and Porkuni stages in central and westernmost Estonia. According to the ostracode record, the formation contains two different associations. The older association comprises the Pirgu species Brevibolbina pontificans Schallreuter, Bullaeferum tapaensis (Sarv), whereas the species Apatochilina falacata Sarv and Gryphiswaldensia plicata Schallreuter from the supposedly younger part are common with the fossiliferous part of the Ärina Formation of the Porkuni Stage.

In southern Estonia, in the limits of the central confacies belt, the Pirgu Stage is represented by red-coloured or mottled argillaceous limestones and mudstones of the Jonstorp and Jelgava formations, including the Kuili Member (Table 7).

According to the traditional understanding, the lower boundary of the Pirgu Stage coincides with the lower boundary of the Moe Formation in northern Estonia. This level is underlain by the chitinozoan Acanthochitina barbata Zone. The zone has also been established under the red-coloured limestones of the Jonstorp Formation (Ruhnu core). On this basis, the lower boundary of the latter formation has been equalized with the lower boundary of the Pirgu Stage in southern Estonia. In terms of graptolite zonation, the stage boundary corresponds to the base of the D. complanatus Zone (Männil 1990). The presence of Climacograptus supernus Elles et Wood in the upper part of the Pirgu Stage suggests its correlation with the Dicellograptus anceps Zone (Männil 1976, 1990).

The lower boundary of the Pirgu Stage is poorly reverberated in the distribution of most shelly fossil groups. As a rule, the new faunal elements appear 1 - 2 m above (or even higher) of the distinct lithological changes (Fig. 59). In northern Estonia, the biostratigraphical boundary is best expressed by the appearance of the concentrations of the alga Palaeoporella?. In central and southern Estonia, the faunal renovation is best revealed in the ostracode record (Meidla 1996, see also Fig. 60).

The Pirgu Stage comprises three different assemblages of macrofauna, related to different facies zones of the palaeobasin. In northern Estonia, in the grey-coloured Moe and Adila formations a rich assemblage of large articulated brachiopods Plaesiomys solaris (Buch), Equirostra gigas Schmidt, Triplesia insularis (Eichwald), Luhaia vardi Rõõmusoks, corals Sarcinula, Catenipora, Palaeofavosites, and stromatoporoids, together with different molluscs, is distributed (Fig. 59). In the ostracode composition nonpalaeocopes are dominating: the associations of Steusloffina cuneata-Medianella blidenensis and S. cuneata-Olbianella fabacea (Meidla 1996, Fig. 60) occur. In central Estonia, in the Halliku Formation the most common representatives of the shelly fauna seem to be brachiopods and rugose corals (Fig. 59). In the red-coloured deposits of southern Estonia, only a few macrofossils, mainly brachiopods and trilobites, have been recorded, while trilobite and echinoderm fragments dominate in the skeletal sand (Männil et al. 1968).


Porkuni Stage

L. Hints & T. Meidla


The Porkuni Stage (Borkholm’sche Schicht by Schmidt 1881) represents the topmost Ordovician stage (Raymond 1916, Bekker 1922). Up to the 1960s, the stage was included to the Silurian System (Öpik 1930b, 1934, Aaloe et al. 1958, Alichova 1960, Rõõmusoks 1960). In Estonia, the Porkuni Stage is represented by variable deposits of shallow-water facies (Männil 1966; Oraspõld 1975b, 1982b; Rõõmusoks 1983), with a thickness of about 10 m in northern and up to 18 m in southern Estonia (Fig. 61). In northern Estonia, the stage is supposedly represented by its older part only, because afterwards, during late Porkuni time, this area turned into dry land as a result of the glacioeustatic sea-level lowering (Oraspõld 1975b).

In northern Estonia, the Porkuni Stage is represented by the Ärina Formation comprising a succession of dolomites (Röa Member), stromatoporoid-tabulate reefs (with surrounding facies) and oolitic or sandy limestones (Kamariku Member) in the top. The assignment of the Röa Member (0.5-5.5 m of dolomites) has been problematic over the years. Some researchers have assigned it to the Pirgu Stage (Rõõmusoks 1991), others to the Porkuni Stage (Rosenstein 1943, Jaanusson 1956, Resheniya… 1987). The unit, usually poor in fossils, yields some species common with the Pirgu Stage (Rõõmusoks 1989). In many sections the lower boundary of the member is lithologically sharp, except the areas where the topmost part of the Adila Formation is dolomitized. The upper boundary is transitional. Here, the assignment of the Röa Member to the Porkuni Stage is conventional.

Small reef bodies, recorded in the middle part of the Ärina Formation (2-3 m high, up to 20 m wide, traditionally treated as the Tõrevere Member), yield the tabulate corals of the Palaeofavosites rugosus community (Klaamann 1986) and the stromatoporoids Clathrodictyon mammillatum (Schmidt), Ecclimadictyon porkuni (Riabinin) a.o. The reefs are surrounded by skeletal limestones (Vohilaid Member, up to 3.7 m) and kerogenous limestones (Siuge Member, up to 2.6 m; see Oraspõld 1975b, Rõõmusoks 1983), apparently representing the pre-reef and inter-reef facies, respectively. In the western part of mainland Estonia, the Vohilaid Member, which is often represented by pure skeletal sand in sparry (?) calcite matrix, contains thin (up to 20 cm) layers of oolitic limestone, whereas the ooids make up 10-45% of the rock volume (Oraspõld 1975b).

In core sections, the common succession of the three lithotypes of the “reef complex” begins with skeletal limestones which are overlain by kerogenous and reef limestones (Fig. 62). The rocks contain a rich and diverse macrofauna of corals, brachiopods, gastropods etc., (more than 150 species and subspecies; Männil 1962, Rõõmusoks 1970). The associations characteristic of the particular lithotypes have many species in common: rugose corals Konodophyllum rhizobolon (Dybowski), Streptelasma (Streptelasma) giganteum (Kaljo), brachiopods Streptis undifera (Schmidt), Schmidtomena acuteplicata (Schmidt), trilobite Platylichas mastocephalus (Öpik) and others. Among microfossils, ostracodes are abundant (Meidla 1996), whilst conodonts are extremely rare (Männik 1992b).

South of the distribution area of the Ärina Formation, distinction of the Porkuni Stage is complicated. In some sequences (Ohesaare core) the Porkuni Stage is obviously missing. In many sections in central Estonia (Are and Kahala etc., Fig. 62), the topmost part of the Ordovician sequence is represented by 1- 2m-thick dolomites, which may correspond to some part of the Ärina Formation (?Röa Member). The distribution area of these dolomites coincides roughly with the area of the pre-Silurian (early Silurian?) channeling where the erosion reached the pre-Porkuni rocks (Perens 1995).

In southern Estonia, the Porkuni Stage is represented by the peripheral parts of the Kuldiga and Saldus formations (Ulst & Gailite 1982). The Kuldiga Formation of bioclastic limestones and marls, overlain by the silty and sandy limestones of the Saldus Formation, comprises the cosmopolitan Hirnantia fauna (Rong & Harper 1988). Hirnantia sagittifera (M’Coy), Dalmanella testudinaria (Dalman), Plectothyrella crassicosta (Dalman), typical elements of that fauna, have been identified in the core sections of Ruhnu, Ikla and Taagepera. These species appear in the lower part of the Kuldiga Formation roughly on the level where the ostracodes common with the Ärina Formation and a zonal chitinozoa Spinachitina taugourdeaoui (Eisenack) disappear. The new ostracodes appearing in the Kuldiga Formation seem to have an extraordinarily wide geographical distribution and probably form a part of the Hirnantia fauna sensu lato (Meidla 1996).

The youngest Ordovician deposits corresponding to the Glyptograptus persculptus graptolite Zone are identified only on the western coast of the East Baltic (Ulst 1992). There is no certain evidence on the occurrence of shallow-water deposits of the persculptus Zone in Estonia, although they may be present as the unfossiliferous topmost Ordovician or even in strata assigned to the lowermost Silurian (Kaljo & Hints 1996).

Concluding the data on the terminal Ordovician in Estonia, it should be mentioned that the Porkuni Stage, in the presents limits, comprises rocks of different age. The oldest part of the stage is present in the stratotype area in northern Estonia, while the most complete sequences presumably occur in southern Estonia. The appearance level of the Hirnantia fauna, which may be correlated with the lower boundary of the Hirnantia Stage in Scandinavia, lies seemingly in the lower part of the Porkuni Stage in the East Baltic.




H. Nestor


The first stratigraphical classification of the Silurian rocks in Estonia was worked out by Schmidt (1858, 1881, 1892). Bekker (1922, 1925) and Luha (1930, 1933, 1946) established the present nomenclature of the Silurian regional chronostratigraphical units - regional stages. Lithostratigraphical divisions have been adequately defined in the monograph “The Silurian of Estonia” (Kaljo 1970c). Further amendments to the stratigraphical nomenclature and correlation with the sequences of the adjacent areas have been published in the unified regional stratigraphical charts of the Baltic Republics (Resheniya… 1978) and of the East-European Platform (Resheniya… 1987). The latest version of the Silurian stratigraphical chart, approved by the Stratigraphical Commission of Estonia, was published by H. Nestor (1995a) and is followed in the present publication (Table 8 ).

The Silurian sequence in Estonia consists of ten regional stages grouped directly into the series of the global chronostratigraphical standard. In most cases the boundaries of the regional stages and series have been considered more or less congruent, based on the graptolite or conodont datings (Kaljo 1962, Viira 1982). An exception is the Wenlock/Ludlow boundary which is only conventionally fitted with the junction of the Rootsiküla and Paadla stages. The lower limit of the Silurian System coincides with the boundary between the Porkuni and Juuru stages. It is proved by the presence of the Hirnantian trilobites and brachiopods in the Kuldiga and Saldus formations of the Porkuni Stage and records of Stricklandia lens prima Williams from the lowermost beds of the Varbola Formation of the Juuru Stage (Kaljo et al. 1988b).

Based on the sharply expressed lateral, facies changes of the Silurian rocks, the Mid-Estonian and South-Estonian confacies belts have been distinguished (Kaljo 1977). The Mid-Estonian Confacies Belt is dominated by various lime- and dolostones, rich in shelly fauna. The belt covers the islands of the West-Estonian Archipelago and the western and central parts of mainland Estonia (Table 8). In the latter area, the Silurian sequence is less complete; its upper part has undergone severe dolomitization. The South-Estonian Confacies Belt consists mostly of marl- and mudstones with a more unilateral deeper-water shelly fauna, graptolites and planktonic microfossils (chitinozoans). Within the confacies belts separate sets of lithostratigraphical units have been established.

Many parts of the Silurian sequence have a clearly expressed cyclical nature, especially in the more shallow-water Mid-Estonian Confacies Belt. In such cases a cyclostratigraphical unit, the so-called beds consisting of alternating types of rocks with a certain trend of succession, has been distinguished and treated as a subdivision of formation. In some cases formations can be subdivided into members.


Llandovery Series

Juuru Stage

The Juuru strata were established by Schmidt (1858) as the Bed (“Jördensche Schicht”), later transferred to the rank of Stage (“Stufe”) (Schmidt 1892). Nestor and Kala (1968) determined the present stratigraphical extent of the stage and worked out its classification. With the Juuru Stage they united the Tamsalu Formation, earlier treated as an independent stage, and the lowermost beds of the Raikküla Stage (now the Karinu Member). The former Hilliste Member of the Juuru Stage was recently expanded and raised into formation rank partly corresponding to the Raikküla Stage (Männik 1992b, Nestor 1995a).

The Juuru drill core in the interval of 0.4-16.2 m has been selected as a neostratotype for the Juuru Stage (Nestor 1993). The Juuru Stage spreads on the islands of Hiiumaa and Saaremaa and in the western, central and southern parts of mainland Estonia. The outcrop extends as a west-eastwards widening belt (4 to 25 km) from midsouthern Hiiumaa as far as the eastern slope of the Pandivere Upland. The main localities are ancient coastal cliffs at Kallasto and Pullapää, quarries at Hilleste, Kirimäe, Karinu, Tamsalu and Rakke (Kamariku) and a well in the ancient Varbola stronghold (Fig. 63). The full thickness of the stage varies from 20.1 m in the Asuküla borehole to 63.7 m in the Viljandi borehole (Fig. 63).

The stage is dominated by biomicritic limestones (packstones, wackestones) rhythmically intercalating with thin layers of marl- and mudstones (argillites, clays) and containing interlayers of sparitic limestones (grain- or rudstones). The proportion of marlstones increases southwards and the number of sparitic interlayers towards the north-west and upwards in the sequence.

The lower boundary of the stage coincides with the base of a thin band of micro- to cryptocrystalline limestone of the Koigi Member or, if the latter is absent, with the base of the marl- or mudstones of the Varbola and Õhne formations overlying various sparitic limestones of the Porkuni Stage, including bioclastic and oolitic grainstones, lithoclastic rudstones of shallow-water origin. Above the boundary, the brachiopod Stricklandia lens, the chitinozoans Ancyrochitina laevaensis and Spinachitina fragilis or the conodont Ozarkodina ex gr. oldhamensis appear.

The Juuru Stage contains a rather rich benthic shelly fauna, whereas planktonic fossils are rare. The most characteristic species are ( abbreviations in brackets: vr - Varbola Formation, tm - Tamsalu Formation, õh - Õhne Formation, pt. - part) Clathrodictyon boreale Riabinin (vr, tm), Paleofavosites paulus Sokolov (vr, tm, õh), Stricklandia lens prima Williams (vr, lower pt.), S. lens lens Williams (vr, upper pt.), Zygospiraella duboisi (Verneuil) (vr), Borealis borealis (Eichwald) (tm), Acernaspis estonica Männil (õh), Calymene ansensis Männil (vr, tm), Aitilia senecta Sarv (vr), Steusloffia eris Neckaja (vr, tm, õh), Ozarcodina ex gr. oldhamensis (Rexroad) (vr, tm, õh), Distomodus kentuckyensis (Branson et Mehl) (vr, tm, õh), Ancyrochitina laevaensis Nestor (õh, basal pt.), Spinachitina fragilis Nestor õh, basal pt.), Conochitina postrobusta Nestor (õh), Dimorphograptus confertus Nicholson) (õh, top), Pribylograptus incommodus (Toernquist) (õh, top). Records of S. lens prima, A. laevaensis and S. fragilis from the basal part of the stage suggest that the base of the Juuru Stage lies on the level of the Parakidograptus acuminatus Zone (Cocks 1971, Nestor V. 1994). Graptolites D. confertus and P. incommodus from the top of the stage confirm that the upper boundary of the stage roughly coincides with the boundary between the Orthograptus vesiculosus and Coronograptus cyphus zones (Kaljo & Vingisaar 1969).

In the Mid-Estonian Confacies Belt, the Juuru Stage is divided into the Varbola (below) and Tamsalu (above) formations. In the South-Estonian Confacies Belt, the Õhne Formation corresponds to both of them (Figs. 64, 65).

The Varbola Formation is represented by nodular biomicritic limestones (skeletal to coquinoid pack- and wackestones) with thin intercalations of marlstone. The formation contains tempestitic interlayers of skeletal grainstones, often with intraclasts, the number of which increases upwards in the sequence and northwestwards in the space. Brachiopods of the Stricklandia Community are characteristic to the formation. The thickness of the formation varies from 8.8 m in the Pusku borehole to 24.6 m in the Käru borehole. The 0.1—3.5-m-thick Koigi Member of micritic (aphanitic) limestones is developed at the base of the Varbola Formation.

The Tamsalu Formation consists of various, prevailingly sparitic limestones (skeletal and pelletal grainstones, coquinoid or lithoclastic rud- and floatstones). The thickness of the formation varies from 8.8 m in the Pusku 2 borehole to 18.5 m in the Rumba borehole. The formation is subdivided into the Tammiku (below) and Karinu (above) members.

The Tammiku Member is typically represented by a bank of coquinoid limestone consisting of shells and debris of the brachiopod Borealis borealis. The thickness of the bank reaches 13.5 m on the Pandivere Upland. In the same area, the Karinu Member consists of skeletal and pelletal grainstones and bio- or lithoclastic rudstones. South- and westwards the latter are replaced by fine-grained grain- and packstones with numerous hardgrounds.

The Hilliste Formation consists of a highly variable assemblage of rock types in which the most characteristic are crinoidal limestones (grainstones) with coral-stromatoporoid bioherms. The formation also contains fine-grained pelletal and skeletal grain- and packstones and micritic limestones. The formation corresponds to the upper part of the Tamsalu Formation (Karinu Member) and to the lower part of the Raikküla Stage (Nestor 1995a). It occurs on Hiiumaa Island and in the vicinity of Haapsalu - Rohuküla and Rapla - Käru in mainland Estonia.

The Õhne Formation is represented by marlstones, mudstones and micritic limestones. It corresponds to the whole stratigraphical extent of the Juuru Stage in southern Estonia. The rather poor fauna corresponds to the brachiopod Clorinda Community. The maximum thickness (63.7 m) has been fixed in Viljandi 91 borehole. The thin, up-to-2.7-m-thick Puikule Member of marlstones and the overlying, up-to-8-m-thick Ruja Member of micritic limestones occur in the basal part of the Õhne Formation along the southern and eastern margins of the area of distribution of the formation.


Raikküla Stage

The Raikküla beds were originally defined (Schmidt 1858) as the “Intermediate zone” (Zwischenzone) between the strata with Pentamerus borealis and P. oblongus. In 1881, Schmidt introduced the geographical name - Raikküllsche Schicht. Kaljo and Vingisaar (1969) presented the currently used subdivision of the stage for southern Estonia. Perens (1992) and H. Nestor (1995a) modernized the classification for the outcrop area. The Mõhküla beds, earlier attributed to the Adavere Stage, were replaced into the Raikküla Stage as they are separated from the rest of the Adavere Stage by a structural disconformity (Nestor 1995a). However, since the stratigraphical level of the Mõhküla beds was changed only recently, it is not yet reflected in the limit between the outcrops of the Raikküla and Adavere stages on the printed geological maps.

The Raikküla-Paka scarp and Raikküla drill core in the interval of 0.5 to 35.0 m have been defined as the composite stratotype of the stage (Nestor 1993). The Raikküla Stage is distributed on the islands of the West-Estonian Archipelago and in the western, central and southern parts of mainland Estonia. The outcrop extends as a latitudinal, eastward widening belt (6 to 45 km) from southern Hiiumaa as far as the southeastern slope of the Pandivere Upland near Palamuse. The main localities are active or abandoned quarries at Pusku, Orgita, Keava, Mündi, Kalana and Rôstla, and ancient coastal scarps (inland cliffs) at Pakamägi and Raikküla-Paka. The thickness of the stage varies from 16.3 m in the Murika borehole to 176.3 m in the Ikla borehole (Fig. 66) and decreases abruptly in the northwest direction due to the end-Raikküla denudation of the upper layers of rocks.

The Raikküla Stage consists of a variety of carbonate rocks. The most characteristic are micritic (micro- and cryptocrystalline) limestones cyclically interbedding with marl- or mudstones in the south and with different bioclastic limestones (wacke-, pack- and grainstones) in the north. In the northernmost sections of central Estonia, the shallowing-up sedimentary cycles may end with argillaceous primary dolomites. In the southernmost sections, the marl- and mudstones contain graptolites on certain levels. In central Estonia, in the Paide - Pärnu belt of faults and eastwards, the Raikküla rocks are strongly dolomitized.

The lower boundary of the stage coincides with the base of a band of marl- or mudstones overlain by thick deposit of monotonous micritic limestones of the Järva-Jaani beds in the north and Slitere Member in the south. In the area of distribution of the most shallow-water sequences of the Hilliste and Raikküla formations the boundary is less definite. Above the boundary, sparse graptolites of the Pristiograptus cyphus Biozone and chitinozoans of the Conochitina electa Biozone (C. electa, C. maennili, etc.) appear.

Fossils are of uneven distribution in the rocks of the Raikküla Stage. The widespread micritic limestones and different dolostones (from pure dolomite to dolomitic marl) contain occasional macrofossils. The most characteristic species are (abbreviations in brackets: rk - Raikküla Formation, nr - Nurmekund Formation, sr - Saarde Formation, - upper part, - middle part, - lower part): Clathrodictyon clivosum Nestor (rk, u. pt.), Parastriatopora celebrata Klaamann (rk, u. pt.), Borealis pumilus (Eichwald) (nr), Borealis borealis osloensis Mjork (nr), Meifodia ovalis Williams (sr), Hermannina hisingeri (Schmidt) (rk, nr), Bythrocyproidea sarvi Neckaja (nr), Icriognathus cornutus Männik (rk, l. pt.), Kockelella manitoulinensis (Pollack, Rexroad et Mehl), (rk, nr), Conochitina electa Nestor (rk, nr, sr, l. pt.), C. iklaensis Nestor (nr, sr), Spinachitina maennili Nestor (sr), Coronograptus cyphus (Lapworth) (sr, l. pt.), C. gregarius (Lapworth) (nr, sr, m. pt.), Demirastrites triangulatus (Harkness) (sr, m. pt.), D. convolutus (Hisinger) (sr, u. pt.). The presence of zonal graptolites shows that the Raikküla Stage spans from the C. cyphus Biozone to the D. convolutus Biozone.

The Raikküla Stage consists of the Raikküla, Nurmekund and Saarde formations, laterally replacing each other from north to south (Figs. 64, 65). The upper part of the Hilliste Formation is of Raikküla Age (Table 8).

The Raikküla Formation is distributed in central and western Estonia, in the Lääne, Rapla and Järva counties. It is represented by two shallowing-up sedimentation cycles star-ting with biomicritic or micritic limestones, succeeded by ske-letal grainstones, pelletal or coral-stromatoporoid limestones, and ending with argillaceous lagoonal dolostones. These cycles are treated as the lower and upper subformations (Nestor 1995a). The thickness of the formation varies from 30 m in the Kiideva borehole to 56 m in the Käru borehole. The upper layers of the formation have undergone considerable denudation and in the westernmost sections of mainland Estonia the upper subformation thins totally out.

The Nurmekund Formation south and east of the Raikküla Formation consists of five sedimentary cycles which begin with a relatively thin layer of marlstone or argillaceous limestone. The main, middle part of the cycle is represented by wavy-bedded micritic limestone, the upper part by bioclastic limestones containing numerous discontinuity surfaces. In central Estonia, the formation is strongly dolomitized, particularly its upper half. The first, third and fifth cycles from below are thicker and more complete, the second and fourth being thinner and less typical. In ascending order, the cycles are termed the Järva-Jaani, Vändra, Jõgeva, Imavere and Mõhküla beds (Table 8). Westwards the upper beds gradually thin out and on Saaremaa Island only the Järva-Jaani and, partly, the Vändra beds are present. The thickness of the formation ranges from 16 m in the Murika borehole to 73+ m in the Võhma borehole.

The Saarde Formation is distributed in southwestern Estonia. It consists of cyclically alternating deposits of horizontally-bedded micritic lime- and marlstones or mudstones and is subdivided into six members. The lowermost, rather thin mudstone member, comprised mostly of argillites, has no name and was earlier included in the Õhne Formation of the Juuru Stage. In ascending order, the Slītere, Kolka, Ikla, Lemme and Staicele members follow. The shaly mudstone interlayers in the Ikla Member abound in graptolites of the Demirastrites triangulatus Zone. In other members graptolites are less frequent. In its full thickness (176.3 m) the Saarde Formation occurs only in the Ikla borehole.


Adavere Stage

The Adavere Stage as a stratigraphical unit was established by Schmidt (1858) as the uppermost unit (zone 6) of the group of smooth pentamerids (“Gruppe der glatten Pentameren”). Afterwards it was termed the Esthonus-Schicht (Schmidt 1881), Addifer Formation (Twenhofel 1916), Adavere Stage (Bekker 1922). Kaljo (1962) fitted the upper boundary of the stage with the Llandovery and Wenlock boundary and included in it the marlstones of the present Velise Formation. Recently, Perens (1992) and Nestor (1995a) excluded the Mõhküla beds and replaced them into the Raikküla Stage.

The Päri quarry in western Estonia has been selected as the neostratotype of the stage (Nestor 1993, 1995a) and the Kirikuküla core at the depth of 50.3 m may be treated as the boundary stratotype of the stage. The Adavere Stage is distributed in the southernmost part of Hiiumaa Island, on Saaremaa and Muhu islands and in the southwestern part of mainland Estonia as far as the Viljandi fault. The outcrop extends as a 10—15-km-wide belt from the southernmost Hiiumaa Island and the Soela Strait over Matsalu Bay up to the vicinity of Türi - Vändra being denudated eastwards the Paide-Pärnu belt of disturbances. The main localities are the Saastna coast, Päri quarry, river banks at Päärdu, Jädivere, Velise, Valgu, Vändra and ditches at Lätiküla and Valgu (Fig. 67). The thickness of the stage increases westwards — from 10.7 m in the Ristiküla borehole to 56.3 m at Nässumaa.

The Adavere Stage is represented by thin-bedded to nodular biomicritic limestones (wackestones to packstones) with Pentamerus oblongus (below) and marl- to mudstones (above). The former unit is treated as the Rumba Formation and the latter as the Velise Formation. The clay content increases westwards. The lower boundary of the stage coincides with the strongly pyritized erosion surface at the base of the nodular biomicritic limestones of the Rumba Formation, transgressively overlying different strata of the Raikküla Stage. The Adavere Stage contains rather rich shelly fauna of Pentamerus (below) and Clorinda (above) communities. Microfossils (chitinozoans, ostracodes, conodonts) are more frequent in the mud- and marlstones of the Velise Formation, almost devoid of corals and stromatoporoids. The most characteristic species are as follows (abbreviations in brackets: rm - Rumba Formation, vl - Velise Formation): Clathrodictyon variolare (Rosen) (rm), Mesofavosites obliquus Sokolov (rm), Angopora hisingeri (Jones) (vl), Palaeocyclus porpita (Linnaeus) (vl), Prodarwinia speciosa (Dybowski) (rm), Pentamerus oblongus (Sowerby) (rm), Stricklandia laevis (Sowerby) (rm), Dicoelosia baltica Musteikis et Puura (vl), Encrinurus (Nucleurus) rumbaensis Rosenstein (rm), Calymene frontosa Lindström (vl), Beirichia valguensis Sarv (rm), Longiscella caudalis (Jones) (vl), Conochitina emmastensis Nestor (rm), Eisenackitina dolioliformis Umnova (rm, vl), Angochitina longicollis Eisenack (vl), Pterospathodus celloni Walliser (vl), P. amorphognathoides Walliser (vl), Spirograptus turriculatus (Barrande) (vl), Monograptus discus Törnquist (vl), Monoclimacis griestoniensis (Nicol) (vl).

The presence of the index species of graptolites (S. turriculatus, M. griestoniensis) and conodonts (P. celloni, P. amorphognathoides) in the upper half of the Adavere Stage demonstrates that most probably the stage corresponds to the Monograptus sedgwickii to Monoclimacis crenulata biozones.

In Estonia, the Adavere Stage consists of the Rumba (below) and Velise (above) formations. The Rumba Formation spreads on the islands of the West-Estonian Archipelago and in the southwestern part of mainland Estonia. It is represented by horizontally-bedded to nodular biomicritic limestones (wackestones, packstones) with clayey partings and scattered shells or tempestitic accumulations of the brachiopod Pentamerus oblongus. The formation consists of twelve low-grade sedimentary cycles beginning with argillaceous rocks (marlstones, argillaceous limestones) and ending with a layer of pure, hard limestone (Einasto et al. 1972). Westwards the clay content of the rocks increases and on Saaremaa Island marlstones are prevailing in the sequence of the Rumba Formation. A characteristic yellowish-green tuffaceous (metabentonite) interlayer (8 to 18 cm) occurs at the level of the base of the upper third of the sequence.

The thickness of the Rumba Formation is mostly 15 to 19 m and it decreases at the western and eastern margins of the distribution area. Local hiatuses occur in the Ohesaare and Are sections.

The Velise Formation overlies the Rumba Formation and consists of different marlstones and mudstones up to plastic clays. The mostly greenish- to bluish-grey rocks are south- and eastwards replaced by red-coloured (purple) varieties. In the southwesternmost sections (Ohesaare, Ruhnu) graptolites are present in the dark-grey interlayers of argillite. Thin (0.5 to 5.0 cm) metabentonite interlayers are characteristic to the formation. The thickness of the formation is greatest in northwestern Saaremaa, reaching 37 - 38 m in the Viki and Eikla sections. In the southeast direction, it decreases until thinning out in the Ristiküla section, eastern Pärnumaa.


Wenlock Series

Jaani Stage

The Jaani Stage was defined by Luha (1933) as a marlstone unit corresponding to the lower part of the “Untere Oeselsche Gruppe (Stufe)” by Schmidt (1858, 1892). Kaljo (1962) separated the lower part of the marlstones (now the Velise Formation), corresponding to the uppermost Llandovery, and joined it with the Adavere Stage. V. Nestor (1984) determined the position of the upper boundary in the subsurface area. Aaloe (1960, 1961) subdivided the Jaani Stage into the Mustjala, Ninase and Paramaja members. Later, Aaloe & Kaljo (1962) distinguished the Tõlla Member for the South-Estonian subsurface area.

A historical stratotype of the Jaani Stage is the sea shore with the Paramaja Cliff in the vicinity of the Jaani Church (Resheniya… 1987, Nestor 1993). The Ohesaare drill core at the depth of 345.8 m may be treated as the boundary stratotype of the stage. The Jaani Stage spreads on Saaremaa and Muhu islands and in the southwestern part of mainland Estonia (Pärnumaa and southern Läänemaa). The outcrop runs along the northern coast of Saaremaa and Muhu islands and the southern coast of Matsalu Bay towards the Vändra Borough (Fig. 68). The main localities are the cliffs at Undva, Suuriku, Ninase, Panga, Liiva, Paramaja, Kautliku, Püssina and Uisu, the quarries at Koguva and Anelema (lower part), and the river bank at Jädivere. The thickness of the stage increases westwards and varies from 24.2 m in the Lihula borehole to 70 m in the Kaugatuma borehole.

The stage consists mainly of various marl- and mudstones. Limestones (skeletal wacke-, pack-, grain- and boundstones) are of minor importance and occur only in the upper half of the stage in the northwestern part of Saaremaa Island (Ninase Member). The lower boundary of the stage has been made congruous with the Llandovery/Wenlock boundary (Kaljo 1962), established by the appearance of the graptolite Cyrtograptus murchisoni in the Ohesaare drill core above the depth level 345.8 m and in other sections by chitinozoans of the Margachitina margaritana Zone (Nestor V. 1994). Lithologically, it usually coincides with a certain increase in the carbonate content of rocks.

The Jaani Stage contains rather rich shelly and planktonic faunas with characteristic deeper-water elements (graptolites, chitinozoans, trilobites). The most typical species are as follows (abbreviations: M - Mustjala Member, P - Paramaja Member, N - Ninase Member, T - Tõlla Member, - upper part, - lower part): Stromatopora impexa Nestor (M, u. pt.), Halysites senior Klaamann (M), Thecia podolica (P), Neocystiphyllum keyserlingi (Dybowski) (P), Leptaena rhomboidalis (Wahlenberg) (M,P), Eocoelia angelini (Lindström) (N), Pseudobollia krekenawaiensis Neckaja (T,P), Craspedobolbina (C.) mucronulata Martinsson (N,P), Beirichia (B.) suurikuensis Sarv (N,P), Calymene orthomarginata Schrank (T,P), Encrinurus punctatus (Wahlenberg) (P), Conochitina cf. mamilla Laufeld (T,N,P), Calpichitina acollaris (Eisenack) (P), Pterospathodus amorphognathoides Walliser (T,M,, Kockelella ranuliformis (Walliser) (M,N,P), Cyrtograptus murchisoni Carruthers (T), Monograptus riccartonensis Lapworth (T), M. flexilis Elles (P). The presence of the index species of graptolites proves that the Jaani Stage spans from the C. murchisoni Biozone to the M. flexilis Biozone.

In Estonia, the Jaani Stage is mainly represented by the Jaani Formation. Only in the southernmost sections, the lower part of the stage has been treated as the Tõlla Member of the Riga Formation (Figs. 69, 70).

The Jaani Formation consists of marlstones and, to a lesser extent, of bioclastic and biohermal limestones. The lower part is formed by the Mustjala Member comprising argillaceous marlstones (Figs. 69, 70), which are often dolomitized, particularly in eastern sections. In the middle of the sequence of the Jaani Stage, the carbonate content increases abruptly and, respectively, the upper half of the Jaani Formation is represented by calcareous marlstones or argillaceous limestones of the Paramaja Member in the eastern part of Saaremaa Island, on Muhu Island and in mainland Estonia. In the northwest direction the Paramaja Member is laterally replaced by bioclastic limestones (wackestones to grainstones) of the Ninase Member containing also bioherms. In many sections, a tongue of the Paramaja marlstones overlaps the Ninase limestones.

In Estonia (Tõlla, Ikla, Ruhnu, Ohesaare drill sections) the Riga Formation is represented only by its lower part - the Tõlla Member, which is characterized by graptolite-bearing grey mudstones. Northwards it is replaced by the greenish-grey marlstones of the Mustjala Member and upwards with the marlstones of the Paramaja Member, both belonging to the Jaani Formation (Figs. 69, 70).


Jaagarahu Stage

The present unit was established by Luha (1933) provisionally as the Muhu-Kurevere Stage, later as the Jaagarahu Stage (Luha 1946). It corresponds roughly to the upper half of the “Untere Oeselsche Gruppe (Schicht)” by Schmidt (1858, 1881). The subdivision of the stage has been recurrently changed (Bekker 1925, Luha 1930, Aaloe 1970, Aaloe et al. 1958, 1976, etc.). Recently, some additional units, including the Jamaja and Riksu formations, were introduced (Resheniya… 1987, Nestor 1995a) and the Muhu dolomites by Luha (1930) were re-established as a formation (Nestor 1995a). V. Nestor (1984) determined the scope of the stage in subsurface area.

The historical stratotype of the stage is an abandoned quarry at Jaagarahu supplemented with the Jaagarahu drill core in the interval of 0.3 to 21.4 m (Aaloe 1970). The Jaagarahu Stage spreads on Saaremaa and Muhu islands and in the southwestern part of mainland Estonia (Pärnumaa and southern Läänemaa). The outcrop extends as a 10—30-m-wide belt from Vilsandi and Vaika islands through northern Saaremaa and Muhu as far as Eidapere and Tori at Tallinn - Pärnu railway (Fig. 71). The main localities are the quarries at Jaagarahu, Tagavere, Koguva and Anelema, and recent and ancient coastal cliffs at Vilsandi, Abula, Panga (Photo 22), Pulli (Oiu), Üügu, Püssina, Kesselaid, Salevere and Kirbla (Photo 23). The thickness of the stage is variable and increases southwestwards from 32.3 m in the Viki core to 145.0 m in the Ohesaare core.

In the western part of Saaremaa, the Jaagarahu Stage is dominated by comparatively pure limestones, while dolomites are prevailing in the eastern part of Saaremaa, on Muhu Island and in mainland Estonia. Reefs (bioherms and mounds) are widespread in the Jaagarahu Stage, especially in its lower part (Vilsandi beds and Kesselaid Member). In the South-Estonian Confacies Belt, the lower part of the Jaagarahu Stage is represented by the marlstones of the Jamaja Formation, and the upper part by the nodular biomicritic limestones of the Sõrve Formation. Temporal analogues of the latter formation are absent in northern sequences due to the long stratigraphical hiatus (Nestor & Nestor 1991). In the Mid-Estonian Confacies Belt, the lower boundary of the stage has been drawn by an abrupt increase in the carbonate content of the rocks coinciding with the base of the Jaagarahu and Muhu formations. In the more argillaceous sequences of the South-Estonian Confacies Belt and transition area, the lower boundary is determined by the appearance of chitinozoans of the Linochitina cingulata Biozone (Nestor V. 1994) at the base of the Jamaja and Riksu formations.

The Jaagarahu Stage contains a wide spectrum of fossils from lagoon-related eurypterids and thelodonts to deep-water communities of chitinozoans, ostracodes and trilobites. Severe dolomitization has destroyed skeletal remains over a vast area in the eastern part of the stage (e.g. in the Muhu Formation). The most characteristic fossils are as follows (abbreviations: jg - Jaagarahu Formation, jm - Jamaja Formation, srv - Sõrve Formation, rks - Riksu Formation, mh - Muhu Formation, V - Vilsandi beds, M - Maasi beds, - upper part): Vikingia tenuis (Nestor) (jgV), Ecclimadictyon astrolaxum Nestor (jgM), Favosites mirandus Sokolov (jgV), Thecia confluens (Eichwald) (jgM), Coenites juniperinus Eichwald (jgV,M), Acervularia ananas (L.) (jgV), Kodonophyllum truncatum (L.) (jgM), Dolerorthis rustica (Sowerby) (jm), Howellella cuneata Rubel (jg,mh,srv), Encrinurus balticus Männil (jm), Warburgella estonica Männil (jgM), Craspedobolbina insulicola Martinsson (jm), Leptobolbina quadricuspidata Martinsson (srv), Conochitina lagena Eisenack (jm, rks), C. pachycephala Eisenack (jm, srv), C. cribrosa Nestor (srv), Ozarkodina sagitta rhenana Walliser (jgV), Kockelella walliseri Helfrich (jm), K. amsdeni Barrick et Klapper (srv), Monograptus flemingii Salter (jm, srv), Gothograptus nassa (Holm) (srv,, Logania taiti (Stetson) (srv). Taking into account the few findings of zonal species of graptolites in the Ohesaare core, it seems that the Jaagarahu Stage probably spans from the M. flexilis Biozone (partly) to the Gothograptus nassa Biozone (Table 8). The main, lower part of the Jaagarahu Stage consists of the Jaagarahu, Muhu, Riksu and Jamaja formations laterally replacing one another (Figs. 69, 70). The upper part of the stage is represented by the Sõrve Formation which is distributed only in the South-Estonian Confacies Belt; in the northern sequences a stratigraphical cap corresponds to it.

The Jaagarahu Formation occurs in northwestern Saaremaa and consists of very variable, prevailingly sparitic limestones of shallow-water origin. Coral-stromatoporoid limestones, including reefs or bioherms, and fine-grained skeletal and pelletal grainstones are the most widespread rocks. In some places they are dolomitized. The formation contains some bands of lagoonal argillaceous dolostones, the so-called eurypterus and pattern dolomites which divide the formation into three subunits: the Vilsandi, Maasi and Tagavere beds. The Vilsandi beds comprise an abundance of large bioherms. The Maasi beds contain biomicritic interlayers of deeper-water genesis. The Tagavere beds are capped by the thickest (5-8 m) deposit of lagoonal dolomites treated sometimes as the Selgase Member. The thickness of the Jaagarahu Formation varies from 32 to 46 m (Fig. 71 ).

The Muhu Formation is distributed in northeastern Saaremaa, on Muhu Island and on mainland Estonia north of the Pärnu latitude. It consists mostly of rather monotonous flaggy dolomites containing numerous large massive reef-mounds (Fig. 70, Photo 23) in its lower part (Kesselaid Member). Almost everywhere the thickness (20 to 40 m) of the formation is uncomplete due to the post-Silurian denudation.

The Riksu Formation bounds the Jaagarahu and Muhu formations from the south, and spreads along the southern coasts of Saaremaa, the Tõstamaa Peninsula and around Pärnu Bay. It is mostly represented by nodular biomicritic to micritic limestones containing layers of argillaceous limestones and marlstones causing a cyclical nature of the sequence. In the eastern part of the distribution area, the rocks are usually dolomitized. In places (Pärnu, Kihnu, Seliste), the Riksu Formation is underlain by a tongue of the Jamaja Formation and, in most places, it is overlain by the tongue of the Jaagarahu or Muhu formations. The thickness of the Riksu Formation varies from 34.5 m in the Nässumaa borehole to 50.8 m in the Kaugatuma borehole.

The Jamaja Formation forms the lower part of the Jaagarahu Stage in the South-Estonian Confacies Belt. The formation is represented by different marl- and mudstones. The thickness of the formation reaches 95.2 m in the Ohesaare section.

The Sõrve Formation overlies the Jamaja Formation in the southernmost sections of Estonia (Ohesaare, Ruhnu, Ikla). It is represented by biomicritic to micritic nodular limestones (pack- and wackestones) similar to the Riksu Formation but lying stratigraphically higher in the sequence and corresponding to the hiatus in northern sequences. The thickness of the formation reaches 49.8 m in the Ohesaare boring.


Rootsiküla Stage

The Rootsiküla strata were established by Bekker (1925) as the Rootsiküla-Kaarma Substage of the Saaremaa Stage (= “Obere Oeselsche Gruppe” by Schmidt 1858, 1881). Later Luha (1933) raised the unit into the stage rank and introduced the name Kaarma. Einasto (1970) motivated the use of the name Rootsiküla. He also defined the boundaries of the stage and subdivided it into beds. Viita quarry, the historical stratotype, has been destroyed. The Kipi drill core in the interval of 25.6 to 53.6 m has been treated as the hypostratotype of the stage (Einasto 1970, Nestor 1993). The Rootsiküla Stage spreads in middle and southern Saaremaa, in the western part of the Tõstamaa Peninsula and on Kihnu and Ruhnu islands. The outcrop forms a 4–10-km-wide belt running through the central part of Saaremaa from Atla to the Kübassaare Peninsula. On mainland, it reaches the Seliste Village on the Tõstamaa Peninsula (Fig. 72). The main localities are the coastal cliffs at Elda, Soeginina, Anikaitse and Kübassaare, the Vesiku Rivulet and an abandoned quarry at Pamma. In Estonia, the full thickness of the stage varies from 20 to 40 m and increases rapidly southwards, towards the Kurzeme Peninsula (Fig.72).

The Rootsiküla Stage consists of various skeletal, pelletal, lithoclastic, coquinoid and micritic limestones cyclically interbedding with argillaceous sedimentary dolomites (the so-called Eurypterus and pattern dolomites). Limestones form the lower and dolomites the upper part of the shallowing-up sedimentary cycles. Microbial-algal structures (oncolites, stromatolites) are frequent. The limestones are often dolomitized and in the eastern part of the distribution area the whole sequence consists completely of dolomites. The lower boundary of the stage has been determined at the base of a stratum of skeletal pack- or grainstones forming the lowermost part of the Viita beds which disconformably overlie the first thick stratum of lagoonal dolomites (Selgase Member of the Jaagarahu Formation) in the Middle-Estonian Confacies Belt and unnamed members of skeletal grainstones at the top of the Sõrve and Riksu formations in the South-Estonian Confacies Belt. In the Ohesaare boring these basal-Rootsiküla nodular packstones contain the Beirichia subornata ostracode fauna characteristic to the Mulde Marls on Gotland and correlatable with the Gothograptus nassa graptolite zone.

The Rootsiküla Stage contains sparsely distributed and specific fossil biota. Eurypterids, thelodonts, leperditian ostracodes, specific gastropods, bivalves, oncolites, stromatolites are common indicating the shallow, near-shore environments. The most typical species are (abbreviations: rt - Rootsiküla Formation, Vt - Viita beds, K - Kuusnõmme beds, Vs - Vesiku beds, S - Soeginina beds): Araneosustroma stelliparratum (Nestor) (rtK), Parastriatopora commutabilis Klaamann (rtK,S), Howellella cuniculi Rubel (rtVt), Straparollus (S.) helicites (Sowerby) (rtVs), Murchisonia (Hormotoma) compressa Lindström (rt), Hermannina phaseola (Hisinger) (rt), Bingeria vesikuensis Sarv (rtVt), Beirichia subornata Martinsson (rtVt), Balteurypterus remipes tetragonophtalmus (Fischer) (rtVt,Vs), Ctenognathodus murchisoni (Pander) (rt), Ozarkodina bohemica bohemica (Walliser) (rtVt), Logania martinssoni Gross (rt), Tremataspis schmidti (Rohon) (rtVt,K,Vs). The presence of Ozarkodina bohemica bohemica enables to date the Rootsiküla Stage as top Wenlock - basal Ludlow.

In Estonia, the Rootsiküla Stage is represented by the Rootsiküla and Sakla formations laterally replacing each other (Fig. 70). The Rootsiküla Formation is distributed on Saaremaa Island, except its easternmost part. The formation is represented by cyclically alternating limestones (often secondarily dolomitized) and argillaceous sedimentary dolostones. Limestones are prevailingly skeletal and pelletal grainstones containing in places oolites, oncolites and intraclasts. Biomicritic and micritic limestones also occur on some levels. Argillaceous dolomites and dolomitic marlstones form the upper part of the sedimentary cycles. They are laminated Eurypterus dolomites or massive bioturbated pattern dolomites. Four cycles have been distinguished in the Rootsiküla Formation (Einasto 1970) defined as beds. The lowermost, Viita beds begin with comparatively normal-marine biomicritic limestones and end with typical Eurypterus and pattern dolomites. The Kuusnõmme beds form a thin uncomplete cycle with coral-stromatoporoid or oncolitic limestone in the lower and pattern dolomite in the upper part. The Vesiku beds begin with sparitic limestones, pelletal, oolitic, coquinoid or lithoclastic grain- or floatstones and their dolomitized counterparts and end with the thickest band of Eurypterus and pattern dolomites. The Soeginina beds form an untypical cycle with totally dolomitized porous grainstones in the lower, thick stromatolite band in middle and pattern dolomites in the upper part. The Sakla Formation is developed in southeastern Saaremaa, on Kihnu Island, and on the Tõstamaa Peninsula in mainland Estonia. It is represented by comparatively monotonous, thick-bedded, bioturbated dolomites with numerous pyrite patterns and undefinite cyclicity.


Ludlow Series

Paadla Stage

The present unit was treated by Schmidt (1892) as the Ilionia Schichten (Beds) of the “Obere Oeselsche Gruppe”. Bekker (1925) introduced the geographical name Paadla and considered these beds as a substage of the Saaremaa Stage. Aaloe (1963b) included equivalents of the present-day Himmiste beds earlier correlated with the Kaarma Stage. Aaloe et al. (1976) redefined the upper boundary of the stage excluding the Tahula beds and gave the current stratification of the stage (Table 8).

The historical stratotype of the stage - Paadla quarry, has been destroyed. The Kuressaare-GI (Kingissepa) drill core in the interval of 19.8 to 43.4 m has been chosen as the neostratotype of the stage (Nestor 1993). The rocks of the Paadla Stage occur in middle and southern Saaremaa, on Kihnu and Ruhnu islands and in the western part of the Tõstamaa Peninsula in mainland Estonia. The outcrop forms a 12–20-km-wide belt passing through midsouthern Saaremaa from the vicinity of Karala to Kõiguste and extending eastwards as far as the Tõstamaa Settlement on mainland (Fig. 73B). The main localities are cliffs at Roopa and Katri, quarries at Lümanda, Himmiste-Kuigu, Kogula, Kaarma and Uduvere and the walls of the Kaali meteorite crater. The full thickness of the stage varies from 2.8 m in the Kihnu borehole to 28.4 m in the Kaugatuma borehole, increasing westwards (Table 9).

The Paadla Stage consists of various bioclastic, pelletal and argillaceous limestones containing coral-stromatoporoid bioherms and biostromes in the west and different primary and secondary dolomites in the east. The base of the stage coincides with the top of the Soeginina pattern dolomites of the Rootsiküla Stage, overlain by argillaceous limestones and dolomites with Didymothyris didyma and Ilionia prisca.

In Estonia, the Paadla Stage contains a rather specific shallow-water fauna of corals, stromatoporoids, agnathans, brachiopods and molluscs in the north-west, more diverse shelly fauna in the south-west, and almost barren dolomites in the east. The most typical species are as follows (abbreviations: pd - Paadla Formation, tr - Torgu Formation, kh - Kihnu Formation, S - Sauvere beds, H - Himmiste beds, U - Uduvere beds, - middle part, - upper part): Conochitina latifrons Eisenack (tr), Angochitina elongata Eisenack (tr), Parallelostroma typicum (Rosen) (pd), Lophiostroma schmidtii (Nicholson) (pd), Thecia swindereniana (Goldfuss) (pd), Laceripora cribrosa Eichwald (pdU), Phaulactis cyathophylloides Ryder (pd), Didymothyris didyma (Dalman) (pd, tr), Howellella elegans Muir-Wood (pd, tr), Ilionia prisca Hisinger (pd, tr), Cardiola interrupta Sowerby (pd, tr), Megalomphala taenia (Lindström) (pdU), Hemsiella hemsiensis Martinsson (pdU, tr, Neobeirichia nutans (Kiesow) (tr), Hammariella pulchrivelata Martinsson (pdU, tr), Amygdalella paadlaensis Sarv (pd, tr), Balizoma obtusus (Angelin) (pdU), Ozarkodina crispa (Walliser) (pdU, tr, O. cf. snajdri (Walliser) (tr, Tremataspis mammillata Pander (pdH, kh), Phlebolepis elegans Pander (pd, tr, kh), Andreolepis hedei Gross (pdU, tr The presence of chitinozoan species Conochitina latifrons and C. lauensis enables to correlate the strata of the Paadla Stage in Estonia with scanicus to tauragensis (leintwardinensis) graptolite zones of middle Ludlow in Latvia which suggests that the basal Ludlow beds are probably absent (Nestor & Nestor 1991).

In Estonia, the Paadla Stage is represented by the Paadla, Torgu and Kihnu formations (Aaloe et al. 1976) laterally replacing one another (Fig. 73). The Paadla Formation occurs in the southern part of Saaremaa, except the Sõrve Peninsula. It is dominated by argillaceous biomicritic to sparitic limestones and dolomites with bands of marlstones, coral-stromatoporoid biostromes, pelletal and coquinoid (Didymothyris) limestones. The formation is subdivided into the Sauvere, Himmiste and Uduvere beds (Klaamann 1970a). The Sauvere beds are represented by nodular argillaceous bioturbated biomicritic limestones (pack- and wackestones), containing small bioherms in the west and being gradually replaced by argillaceous dolomites towards the east. The Himmiste beds are mainly represented by micro- to cryptolaminated argillaceous dolomites with the remains of eurypterids and agnathans. They also contain bands of pelletal-skeletal grainstones at the base and (less often) at the top. The well-known Kaarma building dolomite is tentatively attributed to these beds now (Einasto in Kaljo & Nestor 1990, p. 173). The Uduvere beds are represented by variable rocks of shallow-water genesis: skeletal-, pelletal- lithoclastic-, oncolitic grainstones, packstones and rudstones, interbedded with bands of marlstones, coral-stromatoporoid biostromes, etc. Rocks are partly or totally dolomitized, particularly east of the Kuressaare Town.

The Torgu Formation spreads on the Sõrve Peninsula and Ruhnu Island. It mainly consists of nodular argillaceous biomicritic limestones with rather rich shelly fauna, but corals and stromatoporoids are rare.

The Kihnu Formation is distributed on the Tõstamaa Peninsula and Kihnu Island. It is represented by monotonous platy dolomites (below) and argillaceous dolomites (above) of reduced thicknesses containing agnathans of Paadla and Kuressaare ages, respectively (Einasto et al. 1977), and consequently spanning from the Paadla to Kuressaare Stage.


Kuressaare Stage

The Kuressaare Stage was separated from the Kaugatuma Stage by Klaamann (Klaamann 1970a). Later on, the Tahula beds were added to the stage from among the Paadla Stage (Aaloe et al. 1976). The stratotype section is the Kuressaare-GI (Kingissepa) drill core in the interval of 1.5 to 19.8 m (Aaloe et al. 1976). The Kuressaare Stage spreads in the southernmost Saaremaa, on Ruhnu and Kihnu islands and in the southwestern part of the Tõstamaa Peninsula. The outcrop forms a 2-to-10km-wide belt along the southern coast of Saaremaa Island (Fig. 73). The rocks of the stage crop out in temporary excavations and ditches in the Town of Kuressaare and its surroundings. The full thickness of the stage varies from 5.4 m in the Ruhnu to 27.4 m in the Ohesaare borehole (Tab. 9).

The Kuressaare Stage consists of different marlstones (below) and nodular argillaceous biomicritic limestones (above), both containing interlayers of skeletal, lithoclastic and coquinoid grain-, float- and rudstones. The base of the stage coincides with a sharp increase in the clay component and appearance of the elements of a new microfossil assemblage: Pterochitina perivelata, Ozarkodina remscheidensis aff. scanica, Calcibeirichia altonodosa, Thelodus sculptilis.

The Kuressaare Stage contains a rich assemblage of shelly fossils, especially ostracodes.The brachiopod Atrypoidea prunum is extremely numerous and forms coquina banks in the upper, Kudjape beds of the stage. The most typical species are as follows (abbreviations: T - Tahula beds, K - Kudjape beds): Pterochitina perivelata (Eisenack) (T, K), Conochitina granosa Laufeld (T, K), “Parallelopora” ornata Mori (K), “Paleofavosites” moribundus Sokolov (K), Entelophyllum articulatum (Wahlenberg) (K), Tryplasma loveni (M.Edw. et Haime) (K), Atrypoidea prunum (Dalman) (T, K), Calcaribeirichia altonodosa Sarv (T, K), Plicibeirichia numerosa Sarv (K), Retisaculus sulcatus Gailite (K), Limbinariella malornata Sarv (K), Calymene flabellata Männil (K), Pulcherproetus kuressaarensis (Männil) (K), Ozarkodina remscheidensis aff. scanica (Jeppsson) (T, K), O. snajdri parasnajdri Viira et Aldridge (T, K), Thelodus sculptilis Gross (T, K). The Kuressaare Stage has been indirectly correlated with the upper part of the Ludlow Series.

In Estonia, the Kuressaare Stage is represented by the Kuressaare Formation which is subdivided into Tahula beds (below) and Kudjape beds (above) (Aaloe et al. 1976). The Tahula beds mainly consist of argillaceous or dolomitic marlstones with bands of various bio- and lithoclastic limestones. The content of the calcareous component increases northeastwards.

The Kudjape beds are represented by nodular argillaceous biomicritic limestones containing coquinoid interlayers with Atrypoidea prunum and numerous colonial rugose corals.


Přidoli Series

Kaugatuma Stage

Twenhofel (1916) introduced the name Kaugatoma in the sense of the upper subdivision (Zone) of his Oesel Formation (=“Upper Oeselsche Gruppe” by Schmidt 1858, 1881). The present-day limits and stratification of the stage were proposed by Klaamann (1970a) who separated the Kuressaare Stage as an independent unit. The historical stratotype of the stage is the Kaugatuma Cliff supplemented by Kaugatuma-GI drill core in the interval of 0.6 to 37.2 m (Resheniya… 1987, Nestor 1993). The rocks of the Kaugatuma Stage are distributed on the southern peninsulas of Saaremaa Island, and also on Ruhnu and Abruka islands. They crop out in the northern part of the Sõrve Peninsula and on the Roomassaare, Muratsi, Vätta and Leina peninsulas (Fig. 73). The main localities are the cliffs at Kaugatuma and Lõu and the abandoned quarries at Muratsi, Väike-Rootsi and Äigu. The full thickness of the stage varies from 41.6 m in the Ruhnu borehole to 85.7 m in the Sõrve-514 borehole (Table 9).

The Kaugatuma Stage is represented by interbedded marlstones and bioclastic to coquinoid limestones displaying certain cyclicity. In the lower part of the cycle, marlstones are dominating; in the upper part the limestone interlayers become more frequent and the cycle ends with a thick (2-4 m) deposit of crinoidal limestones. In the upper cycles and southwards, the role of limestone layers decreases. The lower boundary of the stage coincides with a notable increase in the clay component. Higher in the sequence, there appear species of ostracodes Amygdalella nasuta, Sleia equestris, Frostiella groenvalliana, Neobeirichia buchiana; chitinozoans Ancyrochitina fragilis; conodonts Ozarkodina remscheidensis eosteinhornensis etc.

The Kaugatuma Stage contains a rich shelly fauna, particularly ostracodes. The guide fossils of the stage are as follows (abbreviations: Ä - Äigu beds, L - Lõo beds): Ancyrochitina fragilis Eisenack (Ä, L), Fungochitina pistilliformis (Eisenack) (L), Densastroma astroites (Rosen) (Ä), Actinostromella vaiverensis Nestor (Ä), Parallelostroma tuberculatum (Yavorsky) (Ä), Favosites pseudoforbesi muratsiensis Sokolov (Ä), Syringopora blanda Klaamann (Ä, L), Cystiphyllum cylindricum Lonsdale (Ä), Atrypoidea prunum (Dalman) (Ä), Stegerchynchus pseudobidentatus (Rybnikova) (Ä, L), Acaste dayiana Richter et Richter (Ä), Pulcherproetus nieszkowskii (Männil) (Ä), Amygdalella nasuta Martinsson (Ä, L), Sleia equestris Martinsson (Ä), Frostiella groenvalliana Martinsson (Ä), Nodibeirichia tuberculata (Klöden) (L), Crotalocrinites rugosus (Miller) (Ä, L), Ozarkodina remscheidensis eosteinhornensis (Walliser) (Ä), O. remscheidensis remscheidensis (Walliser) (L), O. confluens nasutus (Viira) (L), Thelodus admirabilis Märss (Ä), Nostolepis gracilis Gross (Ä, L). The presence of Ozarkodina remscheidensis eosteinhornensis in the lower part of the Kaugatuma Stage shows that its base roughly corresponds to the Ludlow/Pridoli boundary.

In Estonia, the Kaugatuma Stage is represened by the Kaugatuma Formation which is subdivided into the Äigu (below) and Lõo beds (above) (Fig. 73). The Äigu beds consist of two regressive sedimentary cycles, sometimes regarded as the Lower and Upper Äigu beds (Nestor 1995a). In the lower part of these cycles, marlstone layers are prevalent; in their upper part limestones dominate. Among the latter, coarse-grained crinoidal limestones are the most typical rocks, but interlayers of coquinoid or bio-lithoclastic limestones are also quite common, among these bands with Atrypoidea prunum. Both cycles are capped by a thick deposit of crinoidal limestones. The Äigu beds roughly correspond to the ostracode Frostiella groenvalliana Biozone.

The Lõo beds also consist of two sedimentary cycles of the same type but marlstones prevail throughout the whole sequence. Bioclastic to coquinoid limestones occur as thin intercalations. A thicker band of crinoidal limestones occurs at the top of the lower cycle considered sometimes as the Lower Lõo beds. The Upper Lõo beds lack crinoidal limestones at the top. The Lõo beds roughly correspond to the Nodibeirichia tuberculata Biozone.


Ohesaare Stage

The Ohesaare strata were originally established by Bekker (1925) as a substage of the Saaremaa Stage and were raised into the stage rank by Luha (1933). Klaamann (1970a) defined the lower boundary of the stage. Aaloe et al. (1976) distinguished the Kaavi Member. The Ohesaare Cliff is the historical stratotype of the stage. Ohesaare-2 drill core at the depth of 4.10 m has been selected as the boundary stratotype of the stage.

The Ohesaare Stage crops out in the southern part of the Sõrve Peninsula and spreads also on Ruhnu Island under the Devonian cover. The only exposures of the stage are the Ohesaare and Loode cliffs (Fig. 73). In Estonia, the upper limit of the stage is erosional and the stage does not reach its full thickness anywhere. The thickest section (33.7+ m) has been recorded in Kaavi-568 boring.

In Estonia, the Ohesaare Stage is represented by the Ohesaare Formation which mostly consists of argillaceous-dolomitic marlstones or calcareous mudstones with thin intercalations of partly to totally dolomitized bio- to lithoclastic limestones. At the base of the stage there is a rather thick (4-5 m) deposit of various thin-bedded bioclastic to micritic limestones with thin intercalations of marlstone. In the upper part of the sequence the argillaceous-dolomitic marlstones are of red colour and contain silt and sand admixture. This part of the sequence is regarded as the Kaavi Member. The lower boundary of the stage coincides with the junction between the marlstones of the Lõo beds and the platy bioclastic limestones in the basal part of the Ohesaare Stage. Above this level there appear some new elements among ostracodes (Juviella piltenensis, Nodibeirichia protuberans), chitinozoans (Urochitina sp. sp., Eisenackitina lagenicula) and vertebrates (Poracanthodes punctatus, Goniporus alatus), etc.

The Ohesaare Stage contains a rather rich shelly fauna and a diverse association of agnathans and fish remains. Most of the palaeontological records come from the Ohesaare locality and characterize the lowermost part of the stage. From the Kaavi Member (K) only vertebrate fossils have been identified up to now. The species characteristic of the whole stage include Eisenackitina lagenicula (Eisenack), Urochitina cf. simplex Eisenack, U. verrucosa Eisenack, Favosites forbesi ohesaarensis Klaamann, F. vectorius Klaamann, Fistulipora tenuilamellata (Bassler), Eridotrypa parvulipora Ulrich et Bassler, Shaleria dzwinogrodensis (Kozlowski), Collarothyris collaris (Rubel), Grammysia obliqua (McCoy), Tentaculites scalaris (Schlotheim), Lonchidium inaequale Eichwald, Calymene conspicua Schmidt, Eophacops serotinus Männil, Juviella piltenensis Gailite, Nodibeirichia protuberans (Boll), Klodenia leptosoma Martinsson, Orcofabella testata (Gailite), Ozarkodina confluens nasutus (Viira), Poracanthodes punctatus Brozen, Tylodus deltoides Rohon, Goniporus alatus (Gross) (K), Nostolepis alta Märss (K).

The presence of the conodont species Ozarkodina remscheidensis remscheidensis allows to correlate the Ohesaare Stage with the upper Pridoli.




A. Kleesment & E. Mark-Kurik


The first data about the Devonian of Estonia and fossils date from the first half of the 19th century (Engelhardt & Ulprecht 1830, Kutorga 1835, 1837). A great contribution was made by Asmuss (1856), professor of Tartu University, with his valuable collection of fish fossils from the Aruküla caves (Photo 13). Grewingk (1861, 1879) was the first to describe and correlate the Devonian strata with neighbouring areas. Eichwald (1854a) presented the first reasonably complete description of the cross bedding of the Devonian outcropping strata.

Systematic research into the Devonian was started in the first half of the 20th century. During the 1920s-1940s, the outcropping Devonian strata were described and correlated by Bekker (1924a), Orviku (1930c, 1932, 1935b, 1946, 1948), Gross (1930, 1931, 1933, 1934, 1940b, 1942), Obruchev (1931, 1933) and Bölau (1943, 1944). In general outline, the classification and correlation dating from that period are valid todate (Sorokin 1981).

In the second half of the century, the lists of the fish fossils related to Devonian strata have been essentially improved (Obruchev & Mark-Kurik 1965) and new stages differentiated (Mark 1958, 1964). Age correlation of the Devonian strata in Estonia has been revised and adjusted to the internationally acknowledged scale (Mark-Kurik 1991a, 1993c, Valiukevičius 1988, 1994).

In connection with a medium-scale geological mapping thousands of boreholes were made which enabled research into buried Devonian strata (Kajak et al. 1970, Kajak & Kajak 1983, 1986, Kõrvel et al. 1970, Väärsi et al. 1971, 1981, Vanamb et al. 1975, Kala et al. 1981a, Polivko et al. 1981, Arvisto et al. 1987). In geological mapping, the stratigraphical schemes worked out by the Commission of Stratigraphy of the Baltic and East-European Platform were taken into account (Resheniya… 1978, Rzhonsnitskaja & Kulikova 1990).

Mineralogical studies initiated by H. Viiding and carried on by A. Kleesment imparted much valuable information for subdividing and correlating the strata (Viiding 1962, 1964, 1965, 1976b, Kleesment 1976, 1977, 1984). The zonal scheme was worked out by J. Valiukevičius (1988, 1994) and Valiukevičius et al. (1985) on the basis of acanthodian scales.

The present chapter is based on the whole bulk of data available on the subject. The sources include the publications and original material of the authors of the chapter, descriptions of sections stored in the Estonian Geological Fund, the results of grain-size and mineralogical analyses, based mainly on fraction 0.1–0.05 mm. Use has also been made of H. Viiding’s unpublished results. The recent stratigraphical scheme of the Devonian in Estonia (Table 10) is based on complex studies and was accepted in 1995 by the Devonian Working Group of the Estonian Stratigraphic Commission.


Lower Devonian

A. Kleesment & E. Mark-Kurik


The Lower Devonian with a total thickness of up to 51.5 m is spread in restricted areas in southern Estonia. It is represented by three stratigraphical units of different age, separated from each other by stratigraphical gaps (Table 10, Figs. 74, 79, 80).


Tilžė Stage

Liepinš was the first to acknowledge the Tilžė beds as an independent stratigraphic unit - the Lower Stoniškiai Formation (Liepinš 1955). Paškevičius (1959) determined the present stratigraphic extent of the stage and gave it the present name. Faunistically, it was determined by Karatajute-Talimaa (1962). The stratotype of the Tilžė Stage is in the interval of 1104.5–1212 m of the drill core Stoniškiai-1 in southwestern Lithuania.

The Tilžė Stage is spread in southeastern Estonia and covered with rocks of the Rēzekne Stage. It has been determined only in the Laanemetsa and Värska drill cores, but is assumed to be present also in the Väimela drill core. The sediments are absent in the Mõniste High. The total thickness of the stage varies from 2.1 to 17.7 m (Fig. 74).

The Tilžė Stage lies with a stratigraphical unconformity on the rocks of different stages of Ordovician age (Figs. 75, 76, 77). It has yielded several thelodonts of stratigraphical value: Turinia pagei (Powrie), Turinia sp., Nikolivia gutta Kar.-Tal. and N. elongata Kar.-Tal. The other fossil fishes (psammosteid heterostracans, acanthodians) are identified only on the group level (Sorokin 1981).

In the Baltic region, the Tilžė Stage is represented by the Tilžė Formation which in Estonia is composed of grey and purplish-grey horizontally bedded silt- and sandstones with interlayers of grey clay and yellowish-grey dolomite. Silt- and sandstones are predominantly strongly cemented with dolomitic or gypsum (Värska) matrix. Siltstone is often mottled, conglomeritic sandstone occurs in the basal part.

The rocks of the Tilžė Formation are quartzose or feldspatic arenites with the content of quartz 60–85%. Micaceous arenites (content of micas up to 60%) occur in some places.

The heavy fraction is dominated by the group of transparent allothigenic minerals (50–70% ). Garnet with a considerable supplement of zircon is prevailing. Tourmaline and apatite are also important (Fig. 78).


Ķemeri Stage

The stage was established by Liepinš (1960, 1964). The Ķemeri drill core in the interval of 461 to 547.65 m has been selected as a neostratotype for the Ķemeri Stage (Sorokin 1981). The probable Ķemeri rocks in Estonia have not revealed any fossils which makes the correlation of sections rather difficult.

In Estonia the presumable Ķemeri Stage occurs in a limited area in the southwestern part of the Republic and is identified only in Ikla, Ipiku, Tõlla and Abja drill cores where its thickness varies from 5.9 to 8.4 m (Fig. 79). It lies with a stratigraphical discordance on the Silurian rocks and is covered with deposits of the Rēzekne Stage (Figs. 76, 77).

In the Baltic Region, the Ķemeri Stage is represented by the Ķemeri Formation (Table 10). In Estonia the formation consists of light-grey and pinkish poorly sorted horisontally thin-bedded sandstone and dolomite cemented with conglomeratic sandstone in the basal part. It includes interlayers of grey clay and dolomitic marl, seldom bluish-grey stabby siltstone.

Mineralogically, the rocks of the Ķemeri Formation are predominantly quartzose arenites with the quartz content reaching 80-98%. In the heavy fraction, ilmenite and transparent allothigenic minerals dominate, accounting for 17–57% and 37–53%, respectively. Among the latter group zircon is clearly prevailing, but garnet and tourmaline are also important (Fig. 78).


Rēzekne Stage

The Rēzekne Stage was established by Lyarskaya (1974) with the stratotype in Akniste-5 drill core (interval 361.8 - 487.8 m) in southeastern Latvia. Earlier it was treated as the Kemeri (Liepinš 1952) or Viesite (Liepinš 1964) Formation. After destruction of Akniste-5 drill core, the interval of 427 - 446 m in Ludza-15 drill core in eastern Latvia was selected for the neostratotype of the Rēzekne Stage (Lyarskaya 1978). In Estonia, the corresponding strata were earlier treated as the Pärnu Stage (Mark & Paasikivi 1960) and the Viesite Formation (Kleesment 1966). As the Rēzekne Stage they were first mentioned in 1975 (Kleesment et al. 1975). The age of the stage was palaeontologically determined by Mark-Kurik (1991a).

The Rēzekne Stage is spread in southern Estonia and covered with the rocks of the Pärnu Stage (Figs. 75, 76, 77). The best examined section is the Mehikoorma drill core (interval 220.3 - 246.2 m, Kleesment et al. 1975). The total thickness of the stage varies from 0.7 m in the Asuküla and Kaagvere boreholes up to 51.5 m in the Kavastu borehole. The stage is at its thickest in eastern Estonia (Fig. 80).

The Rēzekne Stage is characterized by greenish-, purplish- and light-grey sandstone. In southeastern Estonia, the upper part of the section is represented by dolomitic marls. The stage lies with a stratigraphical unconformity on the different stages of Ordovician or Silurian age, in a few cases on the rocks of the Tilžė Stage (Värska, Laanemetsa, Väimela), in the Mõniste High it overlies the basement complex. The lower part of the section consists of sandstones with dolomitic matrix, often conglomeratic (Figs. 75, 76, 77).

The fossils known from the Rēzekne Stage are largely confined to the Rēzekne Formation. An equivalent of the latter unit, the Lemsi Formation contains a few unidentified fish remains (Sorokin 1981). Fossil fishes coming from the Rēzekne Formation include Schizosteus sp., Psammosteidae gen. indet., Cephalaspidida gen. indet., Kartalaspis belarussica Mark-Kurik in litt., Antiarchi gen. indet., Laliacanthus singularis Kar.-Tal., Diplacanthus kleesmentae Valiuk., Acanthodes ? sp. B Valiuk., Acanthodes? sp. C Valiuk., Ptychodictyon ancestralis Valiuk., Cheiracanthus gibbosus Valiuk., Markacanthus parallelus Valiuk., Ectopacanthus flabellatus Valiuk., Rhadinacanthus primaris Valiuk., Nostolepis sp., Pruemolepis wellsi Vieth-Schreiner, Crossopterygii gen. indet. The presence of otoliths is noteworthy.

Invertebrates comprise “Lingula” sp., Ostracoda, Glyptasmussia? sp., Gastropoda. Microfossils include a simple conodont and miospores: Retusotriletes simplex Naumova, Leiotriletes microrugosus (lbr.) Naumova, L. simplex Naumova [rz1], Stenozonotriletes conformis Naumova [rz1], Acanthotriletes perpusillus Naumova [rz1], A. parvispinosus Naumova [rz1], Archaeozonotriletes memorabilis V. Umnova [rz1], Emphanisporites rotatus McGregor [rz1], Dibolisporites cf. eifeliensis (Lanninger) McGregor [rz1], Diatomozonotriletes devonicus Naumova [rz1], Retusotriletes cf. priscus V. Umnova [rz2], Leiotriletes cf. insuetus V. Umnova [rz2], Granulatisporites cf. rudigranulatus Staplin [rz2], cf. Calamospora pannucea Richardson [rz2]. The above list shows that the miospore content differs in the lower [rz1] and upper [rz2] parts of the Rēzekne Formation. The lists of fossils are by Kleesment et al. 1975 and Valiukevičius 1994 (miospores identified on generic level are not indicated).

In Estonia, the Rēzekne Stage consists of two formations, laterally replacing each other. In eastern Estonia, the Rēzekne Formation expands as far as Lake Võrtsjärv, west of it the Lemsi Formation occurs (Sorokin 1981, Figs. 77, 80).

The interval of 220.3–246.2 m of the Mehikoorma drill core has been selected as the parastratotype of the Rēzekne Formation. The thickness of the formation varies commonly from 10 to 30 m (Fig. 80). The lower part of the section is represented by grey, purplish- and pinkish-grey loose sandstone, with interlayers of brownish-black silty platy clay in its basal part. On contact with the underlying Ordovician or Silurian carbonate rocks (3–5 m) the sandstones are strongly cemented with dolomitic matrix. In southeastern Estonia, the upper part of the Rēzekne Formation is represented by grey silty dolomitic marl up to 10 m in thickness, in other regions by an up-to-1–2-m-thick layer of grey siltstone or silty sandstone (Figs. 75, 76, 77). The sandstone of the Rēzekne Formation is fine- and very fine-grained.

Mineralogically, the sandstone of the Rēzekne Formation is predominantly feldspatic arenite with the quartz content of 70–85%. The sand component of the dolomitic marl contains 60–75% of quartz. The heavy fraction is dominated by allothigenic transparent minerals. In sandstone its share is commonly 45–60% and in dolomitic marl it forms 30–50% of the fraction. This group is dominated by garnet (50–70%), accompanied by zircon (15–25%, Fig. 78). The content of garnet is relatively high in dolomitic marls.

The stratotype of the Lemsi Formation is the interval of 69.8–85.8 m of the Kihnu drill core (Sorokin 1981). The thickness of the formation is commonly 11–20 m (Fig. 77). It mostly consists of light grey, yellowish and brownish, most rarely of purplish loose sandstone, which in the basal contact part with Silurian carbonate rocks is strongly cemented with dolomitic matrix. The upper 0.6–2.9 m of the section consist of greenish and purplish-grey siltstones or very fine-grained sandstone, often strongly cemented with dolomitic matrix. The sandstone of the Lemsi Formation is predominantly fine- and medimum grained.

Mineralogically, the sandstones of the Lemsi and Rēzekne formations are similar. Only in the sandstone of the Lemsi Formation the content of zircon is higher, particularly in the Kanaküla – Tõlla – Ipiku area where it dominates over the garnet.


Middle Devonian

A. Kleesment & E. Mark-Kurik


The Middle Devonian is the completest part of the Devonian section in Estonia with both the Eifelian (Pärnu, Narva, Aruküla) and Givetian (Burtnieki, Gauja, Amata) standard stages being present (Mark-Kurik 1993c). The total thickness of the Middle Devonian rocks reaches 400 m. The wide outcrop area contains excellent exposures (Figs. 81, 82, 84, 86, 87, 88).


Pärnu Stage

The Pärnu strata were established as an independent stratigraphical unit by Orviku (1930c, 1932). The name “Pärnu” (“Pernu”) was first used by Obruchev (1933). Palaeontologically, it was distinguished as the Schizosteus heterolepis Zone by Gross (1942) and as a stage by Mark-Kurik (Mark 1958). The stratotype is the bank of the Pärnu River near the settlement of Tori. The exposures occur on the banks of the Pärnu and Navesti rivers in central and southwestern Estonia, including Oore dairy — the boundary outcrop with the Narva Stage (Fig. 81).

The Pärnu Stage is spread in southern Estonia. The outcrop forms two narrow wedgeform areas in the northwestern and northeastern parts of the distribution area. The total thickness of the stage ranges commonly from 15 to 47 m (Figs. 75, 76, 81). In the Võrtsjärv Depression, only the topmost part of the section, up to 8 m in thickness, is represented.

The Pärnu Stage is characterized by light-yellow fine-grained cross-bedded sandstone. In most of the distribution area it lies conformably on the Rēzekne Stage. The topmost layer of the Rēzekne Stage in the southeastern part of the area is represented by dolomitic marl which is overlain by sandstone of the Pärnu Stage. In the western part, the boundary between these stages is difficult to establish because of their similar composition. In the northern part of the of distribution area, the Pärnu Stage lies with a stratigraphical disconformity on the Silurian and Ordovician carbonate rocks.

The majority of the fossils of the Pärnu Stage are confined to the Tori Member. The Tamme Member has revealed gyrogonites (?) of charophyte algae and, probably, unidentified lamellibranchs and rare fish remains (Orviku 1930c). In the Tamme Member Valiukevičius (pers. comm.) has identified scales of acanthodians Cheiracanthus gibbosus Valiuk., Rhadinacanthus primaris Valiuk., Cheiracanthus brevicostatus Gross and Acanthodes? sp. D. Fossil fishes occurring in the Tori Member are: Schizosteus heterolepis (Preob.), Psammolepis toriensis (Mark-Kurik), Tartuosteus sp., Actinolepis tuberculata Ag., Homostius sp., Byssacanthus dilatatus (Eichw.), Archaeacanthus quadrisulcatus Kade, Diplacanthus kleesmentae Valiuk., Acanthodes sp. B? Valiuk., Porolepis sp., Glyptolepis sp., Osteolepididae, Dipnoi?. Invertebrates (lingulates) are extremely rare. Common is fossil flora including macroremains of Hostinella sp., and Psilophytites sp., and miospores: Periplecotriletes tortus Egorova, Emphanisporites rotatus McGregor, Retusotriletes raisae Tchib., R. devonicus Naumova, R. concinnus Kedo, R. incomptus Kedo, R. planituberculatus Kedo, Dibolisporites antiquus (Kedo) Arkh., Hymenozonotriletes marginodentatus Kedo, H. altus Kedo, H. ludzus Kedo, H. longus Arkh., Calyptosporites velatus (Eisenack) Richardson, C. tener (Tchib.) Obukh. var. concinnus Tchib., Camarozonotriletes apertus Kedo, Sinuosisporites sinuosus (V. Umnova) Arkh., Punctatisporites tortuosus (Tchib.) Arkh. (data from Sorokin 1981, Valiukevičius et al. 1986, modified according to Abukhovskaya (pers. comm.)).

In Estonia and adjacent areas, the Pärnu Stage is represented by the Pärnu Formation. In Estonia, the formation (Table 10) is divided into the Tori (below) and Tamme (above) members.

The Tori Member is dominantly represented by yellow, light-grey or purplish-grey loose cross-bedded sandstone. Strongly cemented sandstone with dolomitic matrix forms only a basal layer with a thickness of 0.03 to 2 m on the Silurian or Ordovician carbonate rocks. Commonly, it contains pebbles of the underlying sediments. The sandstone is dominantly fine-grained, in the bottommost part medium-grained sandstone is developed. The thickness of the Tori Member varies greatly and irregularly and is highest in the Tsiistre drill core. In some places it is absent (Taagepera drill core, Fig. 77).

The Tamme Member is represented by interbedding loose and dolomitic-cemented greenish-, pinkish- and purplish-grey sandstone containing thin interlayers of siltstone and clay. The complex is horizontally-bedded. Commonly, yellowish-grey sandy dolomite (dolostone) with a thickness of 0.5 to 1 m occurs in the topmost part of the section. Strongly cemented sandstones with dolomitic matrix contain irregular vugs with a diameter of 1 to 15 cm, fulfilled with loose sandstone. The sandstone of the Tamme Member is fine and very-fine grained. The thickness of the member varies irregularly from 2 m (Valga-324 borehole) to 30 m, which is the full thickness of the Pärnu Formation (Taagepera borehole, Fig. 77).

Mineralogically, the sandstone of the Pärnu Formation is quartzose and feldspatic arenite with the quartz content of 75–85%. The heavy fraction is dominated by transparent allothigenic minerals (about 50%), among which garnet with a considerable supplement of zircon is prevailing (Fig. 78). The sandstones of the Tori and Tamme members are quite similar; only in the Tamme Member the content of garnet is higher and that of zircon is lower.


Narva Stage

As an independent stratigraphic unit the Narva Stage was distinguished and termed by Obruchev (1933). The left bank of the Narva River near Gorodenka, and the banks of the Gorodenka Brook and the Poruni River near the place where these watercourses flow into the Narva River in northeastern Estonia, make up the stratotype area of the stage. The outcrops are concentrated in northeastern Estonia, with the Narva and Sirgala quarry sections (Fig. 82) being most noteworthy.

The Narva Stage is spread in southern and eastern Estonia. The outcrop area extends as a 10–30-km-wide belt from Ruhnu to Halliku. Besides, there is a separate area in northeastern Estonia and a few isolated spots near the outcrop line. The total thickness ranges from 30 to 109 m, increasing from north to south (Fig. 82).

The lower boundary of the stage coincides with the base of a carbonate breccia or dolomitic marl, 0.2 to 10 m in thickness, which overlie sandy dolomite or sandstone of the Pärnu Stage (Figs. 75,76).

On the basis of palaeontological (acanthodians, inarticulates, brachiopodes, spores), lithological and mineralogical characteristics, the Narva Stage has been divided into three substages (Table 10) traceable from the stratotype area up to eastern Belarus (Valiukevičius et al. 1986, Kleesment et al. 1987).

The Narva oil shale quarry section, 10 km southeast of the Sirgala Settlement, has been selected for the stratotype of the lower, Vadja Substage. Its parastratotype is the outcrop on the left bank of the Narva River, 300 m downstream from the mouth of the Gorodenka Brook. The thickness of the substage in the stratotype area is about 16 m, in Estonia it varies from 10 to 31.9 m (Figs. 75, 76, 83).

The Luutsniku-451 drill core in the interval of 317–377.7 m has been selected as a stratotype for the Leivu (middle) Substage. In the stratotype area it is exposed at the Poruni and Narva rivers. The thickness of the substage in northeastern Estonia is about 5 m and it increases considerably in a southern direction reaching 60.7 m in the Luutsniku drill core (Figs. 75, 76, 83).

For the stratotype of the upper, Kernavė Substage the interval of the Ledai borehole in central Lithuania was proposed.

The Narva Stage is rich in fossil fishes, particularly its upper part - the Kernavė Substage [k] where the majority of the macroremains come from. In the Vadja Substage [v] and especially in the Leivu Substage [l] the fishes are more scarce. The fish fauna of the stage consists of Schizosteus striatus (Gross) [k], Pycnolepis splendens (Eichw.) [l?,k], Cephalaspidida, Actinolepis tuberculata Ag. [k], Holonema sp. A Mark-Kurik [v], Holonema sp. B Mark-Kurik [k], Homostius latus Asm. [k], Coccosteus cuspidatus Miller ex Ag. [k], Protitanichthys? sp.n. Mark-Kurik [v], Byssacanthus dilatatus (Eichw.), Asterolepis estonica Gross [k], Cheiracanthoides estonicus Valiuk. [v], Acanthodes? sp.C [v], Cheiracanthus crassus Valiuk. [v], Rhadinacanthus balticus Gross, Acanthodes? sp.B, Acanthodes? sp.D, Cheiracanthus brevicostatus Gross, C. longicostatus Gross, Ptychodictyon distinctum Valiuk. [l,k], P. rimosum Gross [l, k], Cheiracanthus sp. [l,k], Diplacanthus sp. [l,k], Acanthodes ? sp.A. [l,k], Cheiracanthus intricatus Valiuk. [k], Nostolepis kernavensis Valiuk. [k], Cheiracanthoides proprius Valiuk. [k], Markacanthus costulatus Valiuk. [k], Minioracanthus laevis Valiuk. [k], Ptychodictyon sulcatum Gross [k], Diplacanthus carinatus Gross [k], Acanthodii gen.n. Valiuk. [k], Archaeacanthus quadrisulcatus Kade, Haplacanthus marginalis Ag., Homocanthus gracilis (Eichw.), Thursius fischeri (Eichw.) [k], Osteolepididae, Glyptolepis quadrata Eichw. [k], Glyptolepis sp., Onychodus sp.[v,k], Dipterus serratus (Eichw.)[k], Dipnoi [v, l], Cheirolepis sp.sp., Orvikuina sp. [v, l], O. vardiaensis Gross [k].

Invertebrates are represented by lingulate brachiopods (Bicarinatina sakalana Rõõmusoks et Gravitis a.o.) which are especially numerous in the upper portion of the Leivu Substage. Ostracods (Lepertitia tartuensis Öpik var. geographica Hecker, Ostracoda inc. gen.) are less common and so are unidentified lamellibranchs in the Kernavė Substage. Conchostracans Pseudestheria pogrebovi Lutk., Trigonestheria triangularis Mir., Glyptoasmussia quadrata Mir., G. aff. willweratica Nov., Concherisma eifelense Raym., Ulugkemia sinuata Lutk., U. mesodevonica Mir., Asmussia membranacea Pacht, Praeleaia quadricarinata Lutk. are characteristic for the Vadja Substage and the lowermost part of the Leivu Substage. The lists of animal fossils are given after Sorokin (1981), Valiukevičius et al. (1986) and Valiukevičius (1994).

Microspore assemblage from the Narva Stage is restricted to the Vadja Substage. It includes Retusotriletes raisae Tchib., R. devonicus Naumova, R. concinnus Kedo, R. incomptus Kedo, R. planituberculatus Kedo, R. cf. brandtii Streel, R. fragosus Arkh., R. microsetosus Kedo, R. lanceolatus Kedo, R. luxispinus Kedo, R. engurensis Kedo, R. clivosiformis Kedo, Hymenozonotrietes cf. marginodentatus Kedo, H. altus Kedo, H. cf. echiniformis Kedo, H. ludzus Kedo, Camarozonotriletes apertus Kedo, Grandispora protea (Naumova) Moreau-Benoit, Phylotecotriletes triangulatus Tiw. et Schaarschmidt. The Vadja Substage has also revealed acritarchs, not yet identified (data from Valiukevičius et al. 1986, modified). Gyrognites (?) of charophyte algae (Sycidium) have been found from all members of the Narva Stage. The Kernavė Substage comprises poorly preserved plant macroremains.

In Estonia and adjacent regions, the Narva Stage is represented by the Narva Formation with a highly variable lithology. Its lower part consists mostly of dolomitic marl with interlayers of dolomite and dolomitic clay. Siltstone, very fine-grained and silty sandstone intercalating with interlayers of dolomitic marl and clay, forms the upper part of the Narva sequence. The sequence of the formation is divided into three members corresponding in volume to substages (Table 10, Fig. 83).

In the basal part of the lower, Vadja Member the breccia of dolomitic marl with unsorted irregular pebbles of dolomite, dolomitic marl and siltstone are common. In general this member is characterized by a thin-bedded complex of dolomitic marl, dark-grey to black silty clay and pale yellowish-grey dolomite which often includes crystalline dolomite, chalcedony, pyrite or sphalerite filled vugs. The detrital component of the rocks in the lower part of the unit is mineralogically relatively mature with the quartz content reaching 50–80%. In the upper part, the rocks are often micaceous containing quartz 20–50, feldspars 15–30 and micas 20–60%. The heavy mineral suite is prevailed by Fe hydroxides or pyrite, in some interlayers baryte is dominating. Nonopaque heavy minerals are dominated by garnet, followed by zircon (Fig. 78).

The middle, Leivu Member is prevailed by dolomitic marl. The section varies both lithologically and in thickness. Within the section four beds have been distinguished (Valiukevičius et al. 1986, Kleesment 1995). Two lower beds are thinning towards the north-east (Fig. 83). The lowermost bed of grey dolomitic marl contains a remarkable amount of silty-sandy particles with a diameter up to 1–2 mm. The next bed is a grey thin-bedded complex formed of intercalating dolomitic marl, dolomite and dolomitic clay. The third bed from the bottom comprises interlayers of grey silt- and sandstone. The topmost bed consists of reddish-brown, purplish-grey and grey mottled massive argillaceous dolomitic marl which serves as a good correlative level. Mineralogically, the detrital part of the rocks in the lowest bed belongs to the feldspatic arenites, in other beds – to the arcosic and micaceous arenites. The heavy fraction is dominated by Fe hydroxides, more rarely by pyrite or micas. The beds differ from one another by the nonopaque heavy mineral suite. In the two lower beds garnet is clearly dominating, while in the overlying beds it is accompanied by apatite, zircon and tourmaline. Significant is the presence of sphen (titanite) and corund in the two lower beds (Fig. 78).

The upper, Kernavė Member consists of brownish-red and grey loose and dolomite-cemented silty sandstone with intercalations of siltstone, dolomitic marl and clay. The complex is horisontal-, lenticular-, more rarely cross-bedded. Mineralogically, the rocks of the Kernavė Member belong to the arcosic arenites with the quartz content of 50–80%. In heavy fraction micas (30–60%), more rarely Fe hydroxides (up to 92%) are prevailing. The heavy nonopaque mineral suite is mostly dominated by apatite, followed by tourmaline, zircon and garnet (Fig. 78).


Aruküla Stage

The Aruküla strata were distinguished as an independent stratigraphical unit by Gross (1940b, 1942) and transfered to the rank of regional stage by Mark-Kurik (Mark 1958). The stratotype is the river bank near the Tartu Jaani Cemetery (Photo 24); not far from it are the Aruküla caves (Photo 13). The other important localities are Kallaste, Viljandi, Tamme and Õisu (Fig. 84). The rich material collected there during more than a hundred years contains the majority of the fauna characteristic of this unit.

The Aruküla Stage is spread in southern and southeastern Estonia. The outcrop area forms a 17–55-km-wide belt extending from Ruhnu Island and Ikla Settlement in the west to Pala and Mehikoorma settlements in the east (Fig. 84). The total thickness of the stage in Estonia ranges from 66.0 to 97.2 m.

The Aruküla Stage consists of reddish-brown cross-bedded sandstone, interbedded with siltstone, clay, and dolomitic marlstone. It lies everywhere above the Narva Stage. The lower boundary of the Aruküla Stage coincides, in general, with the lower surface of the first significant uncemented reddish-brown sandstone layer above the dolomitic siltstone or marl of the Narva Stage (Figs. 75, 76). The topmost part of the Narva Stage often shows a greenish-grey siltstone layer; the overlying Aruküla sandstone is mostly inequigranular. In the Võrtsjärv Depression, this boundary is often difficult to establish (Mark & Tamme 1964).

The Aruküla Stage is rich in fossil fishes known from its all three subdivisions (Table 10). Characteristic are psammosteid heterostracans and several placoderms, both arthrodires and antiarchs. Fishes from the Viljandi [vl] and Kureküla [kr] beds are better known than those from the Tarvastu beds [tr]. The Aruküla fish fauna includes: Tartuosteus giganteus Gross, Pycnosteus palaeformis Preobr. [vl], Ganosteus artus Mark-Kurik, Psammolepis proia Mark-Kurik [vl, kr], cephalaspidids [vl], Actinolepis tuberculata Ag. [vl,kr], Holonema obrutshevi Mark [vl], Homostius latus Asm., Heterostius ingens Asm., Coccosteus grossi O.Obr. [vl,kr], Millerosteus? sp [tr], Byssacanthus dilatatus (Eichw.) [vl,kr], Asterolepis estonica Gross [vl,kr], Archaeacanthus quadrisulcatus Kade, Haplacanthus marginalis Ag., Homacanthus gracilis (Eichw.), Rhadinacanthus balticus Gross [vl], R. multisulcatus Valiuk., Diplacanthus sp. [vl], D. carinatus Gross, D. gravis Valiuk., Markacanthus costulatus Valiuk. [vl], Minioracanthus laevis Valiuk. [vl,kr], M. alius Valiuk., Acanthodes? sp. A, Acanthodes? sp. B, Acanthodes? sp. D, Cheiracanthus brevicostatus Gross, C. longicostatus Gross, Ptychdictyon rimosum Gross, P. sulcatum Gross, Gyroptychius pauli Vorob., Glyptolepis sp. sp., Dipterus sp. sp., Conchodus sp. [kr], Orvikuina sp.n., Cheirolepis sp., Tartuosteus? luhai Mark-Kurik [kr, tr], Pycnosteus pauli Mark [kr,tr], Nodocosta pauli Gross [kr], Thursius estonicus Vorob. [kr], Hybosteus sp. [kr], Tartuosteus maximus? Mark-Kurik [tr], Nostolepis sp.n. Valiuk. [tr], Ptychodictyon distinctum Valiuk. [tr], Porolepis? sp. [tr].

Invertebrates of Aruküla Age are mostly known from the Viljandi beds. These are: ostracods Leperditia tartuensis Öpik, Aparchitellina taehtverensis (Öpik), Drepanellina orvikui (Öpik), Pontocypris rulescens (Öpik), lingulates Bicarinatina bicarinata (Kut.), B. ugalana Rõõmusoks et Gravitis and unidentified lamellibranchs. Trace fossils are fairly numerous in the Kureküla beds. Lingulate brachiopods occur also in the Tarvastu beds.

Plant remains consist of gyrogonites? of charophyte algae (Viljandi beds) and some poorly preserved fragmental branches. The list of fossils is from Sorokin (1981) and Valiukevičius (1994, modified).

In Estonia, the Aruküla Stage is represented by the Aruküla Formation. On the basis of lithological and mineralogical evidence, three cyclic units are distinguished in the Aruküla Formation. These cycles occur in all sections of Estonia and adjacent areas and are defined as the Viljandi (lower), Kureküla (middle) and Tarvastu (upper) beds of the Aruküla Formation (Table 10). Each unit begins with relatively coarse and poorly sorted sandstones of a mature mineral composition, but ends with a clayey-silty complex (Figs. 75, 76, 85, Kleesment, 1994).

The lower, Viljandi beds are dominated by very fine sandstones, often platy or slaty-bedded. The Kureküla beds are characterized by irregularly cemented interbeds of variegated siltstones, pockets of white sandstone, lenses of conglomeratic sandstone and interlayers with large clay pebbles. The section of the Tarvastu beds contains typically conglomeratic interbeds and surfaces and crusts of Fe hydroxide.

Mineralogically, the rocks of the Aruküla Formation are predominantly quartzose and feldspatic arenite with the quartz content of 60–90%. Micaceous arenites (content of micas 20–50%) occur as thin interbeds. The heavy fraction is dominated by ilmenite (30–60%) and transparent allothigenic minerals (15–40%). Among the latter, garnet and zircon are most significant. Tourmaline and rutile are also important. On the lower boundary of the formation the content of zircon and apatite increases significantly, and staurolite appears (Fig. 78).


Burtnieki Stage 

As an independent stratigraphical unit the Burtnieki strata was distinguished by Gross (1940b, 1942). Into the rank of regional stage it was raised by Mark-Kurik (Mark 1958). The stratotype is the bank of the Salaca River, 12 km northwest of Lake Burtnieki in northern Latvia. In Estonia, main exposures are situated at Helme and on the banks of the Ahja (Photo 25) and Võhandu rivers. The outcrops of Karksi, Härma, Koorküla and Essi are known as localities of fossil fishes (Fig. 86).

The Burtnieki Stage is spread in southeastern Estonia. The outcrop area forms a 25–50-km-wide belt stretching from Ipiku and Valga in the west to Mehikoorma and Karisilla in the east. The total thickness ranges from 60.6 to 94.5 m (Fig. 86).

The Burtnieki Stage is mainly represented by light (white, yellowish, pinkish and greyish-brown) fine-grained medium- to weakly-cemented cross-bedded sandstones with interlayers of siltstone and clay. The stage lies everywhere above the Tarvastu beds of the Aruküla Stage (Figs. 75, 76, 85). The topmost layer of the Tarvastu beds is, as a rule, represented by reddish or variegated (purplish-grey to reddish-brown) siltstones, which are overlain by white, yellowish-, brownish- or purplish-grey poorly sorted loose sandstones of the Burtnieki Stage.

The Burtnieki Stage is divided into the Salaca (below) and the Abava (above) substages (Table 10). Mark-Kurik (1993a, b) has treated the latter as an independant stage.

Fossils coming from different parts of the Burtnieki Stage, the Härma [hm] and Koorküla [kr] beds and the Abava [ab] Substage belong mainly to fishes: Tartuosteus maximus Mark-Kurik [hm], cephalaspidids [hm], Pycnosteus tuberculatus (Rohon) [hm,kr], Ganosteus stellatus Rohon, Psammosteus bergi (Obr.) [hm], Actinolepis magna Mark-Kurik [hm,ab], Tropinema haermae (Mark) [hm], Homostius latus Asm. [hm,kr], Heterostius ingens Asm. [hm,kr], Coccosteus markae O.Obr. [hm], Asterolepis sp.1 Kar.-Tal.[hm], Homacanthus gracilis (Eichw.) [hm], Nodocosta sp. [kr], Acanthodes? sp. A, Acanthodes? sp. B, Acanthodes? sp. D, Acanthodes sp. [ab], Cheiracanthus brevicostatus Gross [hm,ab], C. longicostatus Gross [hm], Cheiracanthus sp. [ab], Ptychodictyon rimosum Gross [hm], P. sulcatum Gross [hm], ? Ptychodictyon sp. [hm], Diplacanthus carinatus Gross [hm], D. gravis Valiuk. [hm], Acanthodii gen.n. Valiuk. [hm], Markacanthus alius Valiuk. [hm], Rhadinacanthus multisulcatus Valiuk. [hm], Nostolepis sp.n. Valiuk. [hm], Gyroptychius elgae Vorob. [hm], Glyptolepis? karksiensis (Vorob.) [hm], holoptychiids [hm], Psammolepis sp.sp., Byssacanthus sp.sp. [kr,ab], Hamodus lutkevitshi Obr. [kr,ab], Panderichthys? sp. [kr,ab], Psammolepis abavica Mark-Kurik [ab], Psammosteus sp.sp. [ab], Watsonosteus sp.n.? [ab], Livosteus? sp. [ab], Plourdosteus? panderi O. Obr. [ab], Asterolepis essica Lyarsk. [ab], Microbrachius cf. dicki Traq. [ab], Chondrichthyes? [ab], Laccognathus sp. [ab], Osteolepididae [kr, ab], Onychodus? sp. [ab], Dipnoi [ab], Moythomasia? sp. [ab], Cheirolepis sp. [ab] (Sorokin 1981, Valiukevičius 1994, modified).

Of other fossils, silicified wood has been found in the Härma beds and rare lingulates and various remains of the pteridophyte Pseudosporochnus estonicus Kalamees in the upper clayey part of the Abava Substage (Kalamees 1988).

In Estonia and adjacent areas, the Burtnieki Stage is represented by the Burtnieki Formation. On the basis of the lithological and mineralogical data, three cyclic units are distinguished in the Burtnieki Formation. These cycles are observable in all sections of Estonia and defined as the Härma (lower), Koorküla (middle) and Abava (upper) beds (Table 10). Each unit begins with relatively coarse-grained light, variegated (yellowish, pinkish, greyish and brownish) sandstones of a mature mineral composition and ends with clayey silt layers (Figs. 75, 76, 85, Kleesment 1995).

Lithologically and mineralogically (Fig.78), these three beds are rather similar. The sandstones are prevailed by fine-grained fraction which usually forms 50–70% of the rock. The rate of medium-grained and very fine-grained sand fractions is variable, forming 10–30 and 6–20% of the rock, respectively. The share of other fractions rarely exceeds 5%. The predominating thickness of the cross-bedded sandstone series is 20–30 cm. They are dipping to the south, southwest, and southeast. In the Härma beds, the southwest inclination is prevailing, while in the Koorküla and Abava beds the inclination directions are more variable. Siltstones are mostly medium-cemented, variegated, clays are strongly silty, grey, and reddish-brown.

Mineralogically, the rocks of the Burtnieki Formation are predominantly quartzose and feldspatic arenites with the quartz content of 70–90%. Micaceous arenites (content of micas up to 50%) occur as rare thin interbeds. The heavy fraction is dominated by ilmenite (45–65%). The share of allothigenic transparent minerals in the heavy fraction is in general 15–30%. This group is dominated by zircon (40–70%). Of other accessory minerals, tourmaline (7–20%) and staurolite (3–15%) are more important, noteworthy is the appearance of kyanite (Fig. 78). The share of tourmaline is greatest in the Abava beds where it makes up 10-30% of the group of transparent allothigenic minerals.


Gauja Stage

The Gauja Stage was formally established by Liepinš (1951), although the corresponding stratigraphical unit as a Stage already existed in the scheme of Kraus (1934) and had been distinguished by Gross (1942) as “Oredesch-Stufe”. In different periods it has been treated as a separate stage or as the lower part of the Šventoji Stage (Sorokin 1981). The stratotype of the Gauja Stage is the bank of the Gauja River between Cēsis and Sigulda in northern Latvia. In Estonia more important localities are the banks of the Piusa, Pärlijõgi and Mustjõe rivers, Tuhkvitsa Brook and sand quarries near the railway station at Piusa (Fig. 87).

The Gauja Stage is spread in a restricted area in southeastern Estonia. The outcrop area forms a 14–30-km-wide belt which extends from Valga and Luutsniku in the west to Karisilla and Petseri in the east. The total thickness of the stage in Estonia ranges from 78 to 79.8 m (Fig. 87).

The Gauja Stage consists mostly of weakly- to medium-cemented white and light- to yellowish-grey cross-bedded sandstones. It lies everywhere on the topmost clayey-silty complex of the Abava beds of the Burtnieki Stage (Figs. 75, 76, 85). On the contact level the sandstone is often rich in carbonate cement.

The lower and upper parts of the Gauja Stage differ in fossils. The lower, Sietiņi Member has yielded fossil fishes: Psammolepis venyukovi Obr., P. paradoxa Ag., P. heteraster Gross, P.alata Mark-Kurik, Plourdosteus livonicus (Eastm.), Asterolepis ornata Eichw. sensu Ag., Bothriolepis? sp., Glyptolepis baltica Gross, Laccognathus panderi Gross and Megadonichthys kurikae Vorob. in litt. In the Sietiņi Member also some large fragments or stems, or both, of silicified and ferriferous wood have been found (Sorokin 1981, modified).

In the Lode Member, only plant macroremains (Hostinella sp., Archaeopteris sp., A. fissilis Schalh.) and miospores are known. The miospore assemblage includes: Retusotriletes rugulatus Riegel, Samarisporites triangulatus Allen, S. eximius (Allen) Loboziak et Streel, Geminospora micromanifesta (Naumova) Arkh., G. lemurata Balme, emend. Playford, Ancyrospora sp. cf. A. incisa (Naumova) M. Rask. et Obukh., Dictyotriletes sp. cf. Reticulatisporites perlotus (Naumova) Obukh., Perotriletes sp. cf. Rugospora? impolita (Naumova) Tchib. (Blieck et al. 1996).

In Estonia and adjacent areas, the Gauja Stage is represented by the Gauja Formation. In the latter, two cyclic complexes can be distiguished, corresponding to the Sietiņ (lower) and Lode (upper) (Table 10) members, established by Kuršs (1992). The Sietiņi Member consists mostly of sandstones, with variegated siltstone in the topmost part. The lower part of the Lode Member is represented by light, mainly white sandstones, its upper part is dominated by siltstones and clays (Figs. 75, 85).

The sandstones of the Gauja Formation are fine-grained. The share of fine-grained particles is usually 55–65%, the content of very fine-grained particles is 22.5%, on an average. The cross-bedded series are 5–40, mostly 15-30 cm thick, predominately inclined to the southwest, south and southeast. Characteristic are brown iron-rich surfaces, pebbles of purplish-brown and grey clay, quartz and Fe hydroxide.

The siltstones, which form ca 20% of the section on an average, are usually clayey, represented by variegated, grey and brownish varieties. Clays (average 15%) are strongly silty, grey, and purplish-grey, often dolomitic.

Mineralogically, the rocks of the Gauja Formation are predominantly quartzose arenites with the quartz content of 80–94%. The heavy fraction is dominated by ilmenite, transparent allothigenic minerals make up 20–30%. In the latter group, the leading mineral is zircon, although in the Lode Member tourmaline often dominates (Fig. 78). The Sietiņi and Lode members differ notably in the composition of clay minerals. In the Sietiņi Member, the average share of hydromicas is 75% and kaolinite 25%. In the Lode Member, these values are 45 and 55%, respectively. The Lode Member is the most kaolinite-rich level in the Devonian sequence of Estonia.


Amata Stage

The Amata Stage was formally established by Liepinš (1951), although the corresponding stratigraphical unit as Stage existed in the scheme of Kraus (1934) and had also been distinguished as the “Podsnetogor-Stufe” by Gross (1942). The stratotype of the stage is situated at the lower course of the Amata River in Latvia. In Estonia, the main outcrops are the banks of the Piusa (Loosi) and Peetri rivers (Karisöödi), and in the vicinity of Vastseliina (Fig. 88).

The Amata Stage is spread in a restricted area in southeastern Estonia. The outcrop area forms a 5–10-km-wide belt from Mõniste and Ape in the west to Petseri and Dekshino in the east. The total thickness in boreholes ranges from 12–21 m (Fig. 88), but in the outcrops on the banks of the Piusa River it reaches 30 m.

In Estonia, the Amata Stage is represented by sandy-silty sediments alternating with frequent clay interbeds. The stage lies everywhere on the grey clay of the Gauja Stage and starts with a layer of breccia-like sandstone (Figs. 75, 85). According to Kuršs (1992), in the lower Amata layers the cross-bedded series are inclined to the north which is not typical of this part of the Devonian.

The Amata Stage contains Psammolepis undulata (Ag.), Psammosteus praecursor Obr., P. maeandrinus Ag., Asterolepis radiata Rohon, Bothriolepis prima? Gross, B. cellulosa? Pand., Panderichthys rhombolepis Gross and Laccognathus panderi Gross (Sorokin 1981).

In Estonia and adjacent areas, the Amata Stage is represented by the Amata Formation. According to borehole data, the predominating rock type in the Amata Formation is the greenish- and purplish-grey siltstone which forms on average of 45% of the section. In outcrops, however, the sandstones are prevailing. The sandstones of the Amata Formation are light to yellowish-grey, more rarely reddish-brown, fine-grained, medium- to strongly-cemented, with indistinct cross-bedding, the inclination of which varies in wide limits. The sandstones often contain pebbles and lens-shaped interlayers of clay, more rarely quartz pebbles. Clay interlayers are usually purplish-grey and -brown and form 30% of the section as an average.

Mineralogically, the sandy-silty rocks of the Amata Formation are predominantly quartzose arenites with the quartz content being 80-90%. The heavy mineral suite is dominated by ilmenite, the share of transparent allothigenic minerals is relatively great, varying from 26 to 40%. Among this group zircon is predominating. It is followed by tourmaline, staurolite and rutile in almost equal amounts ( Fig. 78, Kleesment 1995). The assamblage of clay minerals is dominated by hydromicas with the average content of 95%.


Upper Devonian

K. Kajak


In southeastern Estonia, the Upper Devonian is represented by carbonate rocks, the thickness of which reaches 47 m in the Parmu borehole. The outcrop belt of rocks of the Upper Devonian Pļaviņas, Dubniki and Daugava stages has a complicated configuration. In the area where the clayey-sandy sediments of the Middle Devonian Gauja and Amata stages crop out, single outlier-islands of carbonate rocks are encountered, the largest being Saarlase and Loosi (Fig. 89).


Pļaviņas Stage

The Pļaviņas Stage and the Pļaviņas Formation have been named after the exposures in the vicinity of the Town of Pļaviņas in Latvia (Liepinš 1951). Currently, these exposures are under the waters of the Pļaviņas reservoir and, therefore, the outcrops near Izborsk (Irboska) have been selected as the neostratotype for the Pļaviņas Stage.

In Estonia and adjacent areas, the Pļaviņas Stage has a thickness of 27–32 m. In the vicinity of Izborsk it is 37 m thick (Fig. 90). The lower boundary of the stage is lithologically clear — the clayey sandy deposits of the Amata Stage are overlain by carbonate rocks of the Pļaviņas Stage. Based on the palaeontological and lithological characteristics, the Pļaviņas Stage has been subdivided into the Snetnaya Gora, Pskov and Chudovo substages.

The Snetnaya Gora Substage has been named after a type section near the Snetnaya Gora Monastery in the vicinity of the Town of Pskov in Russia. In Estonia, the rocks of the substage crop out at the Peetri River upstream of Karisöödi, at the Pärli River near the Saarlase Mill, in the Rõuge Ööbikuorg, in the environs of Loosi and Vastseliina (Fig. 89). According to the borehole data, the thickness of the substage is ranges from 5.5 to 12 m, and increases from west to east (Fig. 90).

The substage is represented by rhythmically alternating yellowish- and greenish-grey micro- to cryptocrystalline argillaceous silty dolomite (MgO 16%, CaO 24%) and dolomitic marl (insoluble residue 30%, MgO 13%, CaO 19%), less frequently by clay. Dolomites and dolomitic marls contain silty interlayers. The complex is micro- and thin-laminated. Imprints of cubical salt crystals are found in dolomite. In northern regions, thin sand interlayers occur. At the Peetri River, the lower part of the section is composed of clay, and the upper part of dolomite.

The fossils, occasionally found in the section, are represented by the brachiopods Camarotoechia aldoga Nal., conchostracans Asmussia vulgaris Lutk. and the fishes Psammosteus meandrinus Ag., Ctenurella pskovensis (Obr.) and Bothriolepis cellulosa Pand., Grossilepis tuberculata (Gross), Moythomasia perforata (Gross).

The Pskov Substage has been named after the type section on the bank of the Velikaya River near Pskov in Russia. The exposures occur in the same area where those of the Snetnaya Gora Substage are situated. According to the borehole data, the Pskov Substage is 7–13 m thick, on the base of the exposures in the vicinity of the Izborsk Castle (Russia) it is about 17 m thick (Fig. 90).

The Pskov Substage is represented by grey, in the lower part by pale purplish limestone. The rate of dolomitization grows to the west. In the Karisöödi area, the lower part of the substage consists of dolomite (MgO 20%, insoluble residue 6%) with 3–10-cm-thick clay interlayers. In the east (Tsiistre, Hino, Vungi), the substage is mainly represented by thin-layered, often cavernous dolomite (MgO 20%), partly silty-argillaceous (insoluble residue 12–20%), in the upper part of the section it is calcareous in places (CaO 30-35%). On the east margin of the Haanja Heights (Tiirhanna, Parmu), the lower beds are represented by dolomite, the upper ones by limestone.

The Pskov Substage is rich in fossils. The brachiopods Anatrypa micans (Buch), Atrypa velikaya Nal., Ladogia meyendorfii (Vern.), Ripidiorhynchus pskovensis (Nal.) dominate. Calcareous algae have also been found.

The Chudovo Substage was differentiated on the basis of the exposures in the vicinity of the Town of Chudovo, Russia. In Estonia, the Quaternary cover is thick and the rocks of the Chudovo Substage are not exposed. The substage crops out near the Pskov - Riga highway. Based on the key fossils Ripidiorhynchus tschudovi (Nal.) and Anatrypa heckeri Nal., the age of the substage has been established in the Izborsk outcrops in Russia. The thickness of the substage in the boreholes reaches 13 m. In places (Laura, Vungi, Parmu), the lower boundary of the substage is marked by a pyritized discontinuity surface.

In the easternmost part of its distribution area (Vungi, Parmu), the Chudovo Substage is represented by micro- and cryptocrystalline limestones (CaO 44–49%, insoluble residue 5–10%). Dolomitization of rocks increases westwards and the substage consists of micro- and fine-crystalline dolomitic limestones (CaO 32–34%, MgO 16–17%, insoluble residue 3–7%) to coarse-crystalline cavernous dolomites (CaO 28–29%, MgO 20%, insoluble residue 3–7%). Dolomitic facies is spread west of Misso.

The Pskov and Chudovo substages are lithologically very similar and sometimes it is expedient to treat them together as the Izborsk Member (Table 10).


Dubniki Stage

The Dubniki Stage and the Dubniki Formation have been named after the former gypsum quarry which is situated east of Izborsk Town in Russia (Bekker 1924a). In the walls of the quarry up to 12.5 m thick bed of greenish-grey clay with gypsum and dolomite interlayers was exposed. Fossils are represented by the brachiopods Comiotoechia bifera (Phill.), Ripidiorhynchus strugi (Nal.) and the ostracod Acratia benevaensis Zasp.

The Formation covers a limited area in the southeastern-most part of Estonia, but the exposures do not occur there because of the thick Quaternary cover (Fig. 89). The thickness of the stage reaches 10 m (Fig. 90) in the boreholes . The section consists of bluish-grey marl (CaO 33%, MgO 3% and insoluble residue 27%) and argillaceous dolomitic marl (CaO 7%, MgO 6%, insoluble residue 58%) with clay and dolomite interlayers.


Daugava Stage

The Daugava Stage and the Daugava Formation have derived the name from the exposures on the banks of the Daugava River, Latvia (Liepinš 1951). In Estonia, the uppermost 8.5 m of the Devonian section in the Parmu borehole belong to the Daugava Stage (Fig. 90). In the borehole, the stage is represented by argillaceous micro- and cryptocrystalline limestones (CaO 49%, MgO 2%, insoluble residue 27%).




A. Raukas & K. Kajak

Structure of the Quaternary cover

Estonia belongs to the zone of glacial erosion or moderate accumulation and, therefore, the Quaternary cover is rather thin. In northern Estonia, on the outcrops of the Ordovician and Silurian carbonate rocks it is usually less than 5 metres. Occasionally, on the so-called alvars, it is even lacking (Photo 26). The Quaternary cover is at its thickest (Fig. 91) in the Haanja and Otepää heights (often more than 100 m) and in the buried valleys of southern Estonia (at Keskküla 207 m). Thick Quaternary deposits are also encountered on lee-sides of heights and elevations (e.g. the Saadjärve Drumlin Field in the “shadow” of the Pandivere Upland) and in front of escarpments oriented against the movement of glaciers which favoured the accumulation of a considerable amount of sediments. In the fore-klint area, the deposits are more than 100 m and reach 143 m in the ancient Harku Valley (in the vicinity of Tallinn). In the Gulf of Finland, the greatest thicknesses have been established in megadrumlins to the north of Tallinn, for example, on the Island of Prangli, where the Pleistocene deposits are up to 123 m thick. Here the Quaternary cover comprises several till layers (Kajak 1961, 1965a, 1995, Raukas 1978).

About 95% of the Pleistocene cover is formed of glacial and aqueoglacial deposits. Glacial sediments, 70% by volume and 47.7% by surface area, dominate (Raukas 1978). Of wide distribution are also glaciolacustrine (6.8% by area) and glaciofluvial (3.1%) deposits (Fig. 92). Five till beds, often of great thickness, are to be noticed more or less distinctly. Only in a few cases they are separated from one another by deposits containing spores and pollen of interglacial or interstadial origin which considerably aggrevates the correlation and dating of the glacial strata.

In terms of glacial stratigraphy and lithogenesis, the bedrock valleys and interlobate “insular heights” are the main objects of interest. Deep ancient river valleys, which were further eroded by glaciers and their meltwaters, vary in morphology and sediment facies infill. They may be filled with glaciofluvial deposits of the last glaciation (Pada Valley), glaciolacustrine deposits of the last glaciation (Selja Valley) or one till bed with under- or overlying glacio-aquatic deposits (Kunda Valley). There are also many valleys with a complicated structure comprising several till beds and accompanying glacio-aquatic deposits. Valleys of the first three types are characteristic of northern Estonia, whereas valleys with a complicated structure prevail in the southern part of the Republic (Tavast & Raukas 1982). The deposits of radial and marginal valleys also have certain areal differences. The deposits of radial valleys have to a large extent been reworked by glaciers and contain less older deposits. With regard to their age, the Upper Pleistocene and Holocene deposits predominate in the buried valleys of northern Estonia and Middle Pleistocene deposits in the valleys of southern Estonia (Raukas & Tavast 1987). The deposits of ancient valleys are the most suitable objects for stratigraphical studies because they contain less erratics than uplands. In the latter, the blocks of bedrock and older Quaternary sediments have been displaced not only horizontally, but occasionally also a considerable up-thrusting or folding, or both have taken place. Older beds are thus found standing more or less vertically in a position tens and even hundreds of metres above their normal stratigraphic position (Raukas & Gaigalas 1993).

The so-called “accumulative insular heights” (the name is derived from the “island-like” position in the topography) form three belts with a N-S orientation in the East-European Lowland (Aboltinš et al. 1989). Estonian insular heights- Haanja and Otepää - belong to the Latgale Zone. These heights are characterized by a hummocky topography and by a considerable thickness of Quaternary deposits (up to several hundred metres). They have a rather great altitude, distinct slopes, plateau-like forms of glaciolacustrine origin in water-divide areas, and usually a bedrock core. During their development they have undergone the following four morphogenetical stages: (1) subglacial, (2) englacial, (3) marginal accumulation, and (4) a dead ice stage (Aboltinš et al.  1989). The heights have formed between rather large ice lobes as a result of frequent redeposition of older deposits, accounting for the mosaic pattern of sediments. Representative outcrops on heights have served as the main areas of stratigraphic investigations for over a century. As a result of redeposition of interglacial and interstadial sediments, the number of supposed interglacials could be erroneously increased. This, in turn, may lead to an older age being assigned to the tills separating them and to misleading palaeogeographical conclusions (cf. e.g. discussion in Liivrand 1991). A precise correlation of Pleistocene deposits assumes great knowledge of glaciosedimentation processes and elaboration of special dating methods.


Classification and composition of deposits

The classification of the Estonian Quaternary deposits is based on the study of genetical types of deposits resulting from the development of a certain dynamic form of accumulation, and playing qualitatively different role in the structure and history of the formation of the sedimentary cover (Raukas 1978). Genetical types can be merged into paragenetical series, groups and subgroups, and, at need, into smaller taxonomical units (facies and subfacies).

Among the Estonian Pleistocene deposits six paragenetical series occur: eluvial, organogenous, colluvial and deluvial, aqueous, glacial and subaerial.

In the paragenetical series of eluvial deposits, areal and linear (along the crevasses and zones of tectonic faults) crusts of weathering, soil horizons on the boundaries of stadial or phasial beds of different age, and the deposits and relief forms of permafrost (cryogenous eluvium) occur in some places. Cryogenous phenomena (occurrences of cryoturbation, ice wedges, bedding disturbances and structural grounds) are most widespread in the Younger Dryas sediments (Photo 27).

Deposits of organogenous paragenetical series occur as interglacial peat and gyttja in few places (Rõngu, Karuküla, etc.). Organogenous deposits did not accumulate in late-glacial times and the thin interlayers of peat are most probably redeposited.

Colluvial and deluvial deposits are relatively frequent in front of Ordovician and Silurian escarpments and in the hilly topography of southeastern Estonia where in late-glacial times solifluctional processes resulting from the melting of buried ice played a significant role.

Deposits of aqueous paragenetical series are represented by several genetical types of different lithological composition. Among those, proluvial and subsurface aquatic deposits are rare. Little is also known about the interglacial and interstadial alluvial deposits represented by intermorainic silty-sandy sediments (Valguta, Peedu) and sandy gravelly deposits in the ancient buried valleys. Late-glacial alluvial deposits of the last glaciation occur in the terraces of a great many river valleys in southern Estonia where their thickness usually ranges between 3 and 10 m.

Interglacial and interstadial lacustrine deposits are represented by gyttja (Karuküla, Rõngu) and silty-sandy sediments (Otepää, Sudiste). Late-glacial terrigenous lacustrine sediments of the last glaciation are to be found under many contemporary bogs. The boundary between the Pleistocene and Holocene lacustrine deposits is rather clear and easy to notice due to an abrupt increase of organic matter in the Holocene deposits or carbonates in the form of lacustrine lime.

Marine interglacial and interstadial deposits are distributed in the fore-klint area (Prangli Island) and on the islands (Kihnu) of the Gulf of Liivi (Riga). The deposits of the Baltic Ice Lake, occurring as bottom and coastal formations, are conventionally also regarded as marine deposits.

As mentioned above, deposits of glacial paragenetical series form a great part of the Pleistocene cover in the Republic (Fig. 2). All the deposits of the glacial paragenetic series can be divided into two paragenetic groups: glacial drift deposited by glaciers on ground (subaerial tills) and underneath ice shelves (subaqueous tills). Among the subaerial varieties the lodgement, superglacial (ablation) and frontal (margin) tills and among the subaqueous varieties iceberg and submarine ablation tills can be distinguished as genetic types (Raukas 1978). On grounds of detailed micropalaeontological, geochemical, geomorphological, structural and other studies, it has been proved that the tills in Estonia are mainly of continental subaerial genesis (Raukas 1973).

Marine microorganisms (diatoms, foraminifers, ostracodes, etc.) are very rare in Estonian tills. Microfauna and -flora has been discovered in greater amount only in a narrow fore-klint area which in the past, and most likely during the interglacial stages as well, was occupied by the sea or big glaciolacustrine basins. The borehole at Vääna-Jõesuu provides an excellent illustration of the above. However, also there the content of foraminifers and diatoms in till is remarkably lower than in intermorainic silty clays (Raukas & Liivrand 1971). The occurrence of foraminifers and diatoms predominantly in the intermorainic layer evidences of the fact that the microfauna and -flora, at least partially, had redeposited here from the bottom of the Gulf of Finland where marine conditions were of repeated existence in the past.

Geochemical and most of the lithological evidence support the theory of subaerial genesis of Estonian glacial deposits. For instance, this is indicated by the great similarity between the lithological composition of tills and underlying rocks, poorly sorted sediments, lack of new authigenous formations of marine origin, poor rounding of clasts and the increase in roundness towards the south and south-east, i.e. in the direction of the supposable movement of continental ice, but also the distribution of indicator (index) boulders, a relatively high content of clayey particles and a high degree of compaction in tills, the orientation of clasts in the direction of advance of ice masses, parallel to glacial striae and almost horizontal in position, the occurrence of glaciodislocations (Photo 28) etc. (Raukas 1973).

Lodgement tills are of the widest distribution. The formation of subaerial tills is thought to have taken place both beneath an advancing glacier (lodgement till) and as a result of bottom melting of a passive glacier during the degradation of glacial covers (basal melt-out till). In Estonia, deformation tills are quite frequent. They are formed by subsole drag underneath the moving glacier (Raukas 1978).

Due to the flatness of the topography, the content of supraglacial material in tills is quite low in Estonia. By contrast, the importance of englacial and basal debris is much greater and quite different. In most cases, the material of basal debris predominates, whereas in places, lodgement tills composed of the material of englacial debris (the so-called erratic tills) are also found. Besides, local tills consisting entirely of local sedimentary material are spread (Photo 29). The vast majority of lodgement tills belong to the intermediate group between local and erratic varieties containing local and allochthonous (far-transported) material in different ratios. According to Gaigalas (1969), it would be expedient to call them transitional lodgement tills. In transitional lodgement tills the local sedimentary material is prevailing.

Ablation tills resulting from surficial melting and gravity flowing of superglacial and englacial debris are represented by flow tills and melt-out tills. Usually they are difficult to distinguish from lodgement tills. Exceptionally, lodgement tills are immediately overlain by ablation tills.

Frontal (margin) tills are present in end moraines which fall into push and dump moraines. Among dump moraines stationary and recessional forms are distinguished. There occur also end moraines of complicated structure which bear traces of ice pressure and are overlain by glaciofluvial deposits, or vice versa, as well as interior peripheral moraines of push character developed between dead and active ice (Raukas et al. 1971). Frontal tills consist of various squeeze lodgement, deformation and flow-till facies with injections of aqueoglacial deposits and bedrock erratics.

The formation of lithological and mineral composition of tills depends on a number of factors, such as the composition and topography of the underlying bedrock, the dynamics of the movement of glacier, the location of material in the body of the glacier, the character of accumulation, the nature and intensity of weathering of material, etc.

Numerous investigations have enabled to elucidate that on its way the glacier accumulated in its body great quantities of local bedrock material (Raukas 1969, a.o.). At that, maximum content of rock particles (about 60-80% ) from a certain stratigraphic unit is usually traceable near the distal boundary of the outcrop of the unit. Already at a distance of 6-8 km from the bedrock boundary, the amount of rock particles from the corresponding unit does not exceed 20-30% of the total (Raukas 1978). The content of erratic material in a typical lodgement till does not usually exceed 5-10%, but in englacially formed till it amounts to 100%. The transport distance of clasts is greatly dependent upon the resistance of rocks. Claystones and weakly cemented sandstones have travelled no more than 15 km, fine-grained limestones 120 km, dolomites 300 km and resistant varieties of crystalline rocks 800-900 km. During transport the till becomes enriched with more resistant clasts. For example, as crystalline rocks are more durable than carbonaceous ones, the South-Estonian tills on sandstones are enriched with crystalline clasts. At the same time, the fragments of carbonaceous rocks become enriched with dolomites as the tougher ones (Raukas 1978).

The local bedrock exerts also a remarkable influence on the mineral composition of tills (Raukas 1974, a.o.). For example, territorial differences are easy to trace on the outcrops of Cambrian clays and siltstones, Ordovician and Silurian carbonaceous rocks and Devonian sandstones. However, even smaller dependences can be traced from territorial aspect and with respect to the composition of different minerals. Thus, the content of carbonaceous minerals decreases abruptly and that of quartz increases southwards from the outcrops of carbonate rocks. The content of amphiboles, pyroxenes and other minerals of heavy subfraction, typical of the outcrops of Precambrian rocks of Finland, gradually decreases in the southern and southeastern direction. Correspondingly, the quantity of weathering-resistant minerals typical of the underlying Palaeozoic rocks, such as garnet, zircon, tourmaline, rutile, etc., increases. Great variations in the proportions of these minerals occur depending upon local conditions. In favour of the above speaks the content of garnet and zircon in the tills of southern Estonia where the underlying Devonian rocks display distinct regularities with respect to those minerals (Raukas 1974).

During the various glaciations, the movements of glaciers have differed (Tavast & Raukas 1982). This enables correlation of till beds on the basis of lithological (Table 11) and mineralogical (Table 12) data.

Of lithological methods, most promising for solving the problems of the Pleistocene stratigraphy and palaeogeography seems the study of crystalline indicator (index) boulders, the content of which in deposits was only slightly influenced by the differences of the local bedrock, and has remained almost stable over vast areas (Raukas 1963b).

The paragenetical subgroup of glaciofluvial deposits is divided into englacial and periglacial genetical types with frequent transitions between them. The deposits of radial eskers and fluviokames are conventionally regarded as englacial glaciofluvial deposits. The deposits of glaciofluvial deltas (Photo 30), sandurs and marginal eskers are identified as periglacial. Glaciofluvial deposits are mostly characterized by a highly variable grain-size composition and structure, and the great variation in lithological and mineralogical composition, everywhere closely connected with the composition of the adjacent till and the bedrock (Raukas 1978).

The maximum distance of transport of pebbles in glaciofluvial streams extends to 16-20 km, being naturally controlled by a great many additional factors, e.g. the hardness of rocks, the width of the outcrop of regional stages, the bedrock topography, etc. (Raukas et al. 1971). In the course of the formation of glaciofluvial deposits, the content of resistant rocks and minerals increases on account of less stable fractions that are crushed or destroyed during their transportation by water streams. Usually the content of crystalline rocks in gravel and pebble fractions of glaciofluvial deposits is 10-15% higher than in tills, whereas the content of carbonate rocks in them is accordingly lower. The content of metamorphic (predominantly gneisses) and coarse-grained magmatic rocks (predominantly rapakivi and pegmatites) is 5-10% lower than in the initial tills (Raukas 1978).

Glaciolacustrine deposits are also divided into englacial and periglacial genetical types. Englacial deposits form plateau-like limnoglacial kames (Raukas et al. 1971) and are included in superposed kames (Kajak 1963). They are of the widest distribution in northeastern Estonia, in the vicinity of Iisaku-Illuka, while periglacial deposits are most common in the Otepää and Haanja heights.

Periglacial glaciolacustrine deposits, predominantly silts and varved clays (Photo 31), are more frequent in western Estonia, in the fore-klint area and on the Narva Lowland (Fig. 2), but also in the numerous river valleys of southern and northern Estonia (Pirrus 1968). The thickness of varved clays reaches 26-27 m. The similar mineral composition of tills and varved clays, and the high content of weathering-resistant minerals in clays points to the insignificant role of chemical changes in the transformation of the initial morainic material into glaciolacustrine clay (Pirrus & Raukas 1963).

Deposits of subaerial paragenetical series are located along ancient shorelines of the Baltic Ice Lake and the larger local ice lakes where they form coastal dunes, up to 12 m high, or in the form of a thin mantle covering the sandy beach ridges. More seldom, they are represented by hillocky plains. Fine-grained sand prevails in the deposits.


Stratigraphical subdivision

Several local and regional stratigraphical schemes have been compiled for Estonia (in 1956, 1957, 1961, 1963, 1970, 1976). These were mainly correlative parts of the schemes of the European portion of the former Soviet Union or the Baltic States and Belarus (Orviku 1956, 1960d, e, Raukas 1978). In the scheme compiled by Kajak et al. (1976) local geographical names were for the first time used to denote stratigraphic units. Over a period of more than 15 years, the scheme served as a basis for large-scale geological mapping and applied works in the Republic.

On May 6, 1993, a new official stratigraphical chart of Quaternary deposits of Estonia (Table 13) was accepted by the Estonian Stratigraphic Commission (Raukas & Kajak 1995). The scheme was approved as a correlative part of the stratigraphical chart of the Baltic States at the II Stratigraphic Conference in Vilnius (May 9-14, 1993).

In the Quaternary stratigraphy, the age of tills is of special interest as it enables correlation of lithologically similar formations over vast areas (Raukas 1978). The age of tills is generally determined by bedding conditions, by their position with respect to interglacial or interstadial deposits. Unfortunately, the latter are rather uncommon. Besides, the advancing glaciers of the succeeding glaciations crushed most of unconsolidated intermorainic organic deposits which today are often found as erratics embedded in younger sediments. The deposits of the Prangli (Eemian, Mikulinan) interglacial, both continental (Rõngu) and marine (Prangli), serve as key sediments in stratigraphic subdivision and correlation of the Pleistocene cover. The Karuküla (Holsteinian, Likhvian) deposits are more complete in the Karuküla section in southwestern Estonia. The spore and pollen spectrum of all other intermorainic sections is not clear, as these sediments often contain reworked pollen. The most important type sites are shown in Figure 93.

In the Estonian stratigraphical chart (Table 13), lithostratigraphic terms have been used as basic units. As a fundamental unit, formation is used in a meaning of glacial and interglacial episodes in the event stratigraphy. Formations are the three-dimensional sedimentary bodies which have been formed by a specific geological process in the time span of one clear geological event. Big stadial episodes in a meaning of event stratigraphy are comparable with subformations. Using close in the meaning but not synonymous chrono- stratigraphical (e.g. the Prangli Stage), climatostratigraphical (Prangli Interglacial), lithostratigraphical (Prangli Formation) and event stratigraphical (Prangli Interglacial Episode) terms has been avoid. Although interglacial sediments are differentiated on the basis of spore and pollen and other fossil evidence, and pollen assemblage zones underlie their description, for the unity of the scheme, even here lithostratigraphical terms (Prangli and Karuküla formations) were preferred.

Lower Pleistocene deposits are absent in Estonia, and the oldest Middle Pleistocene deposits identified so far in the official chart belong to the Upper Sangaste Subformation. Taking into account some similarity of pollen spectrum of sandy clayey sediments in the Otepää buried valley (Harimägi borehole No. 323 at a depth of 143.3-169.2 m) with the Turgeliai and Belovežje subformations in Lithuania and Belarus (Kajak & Liivrand 1967), Kajak (1995) includes those beds to the Middle Sangaste Subformation and underlying tills and glaciofluvial deposits (Otepää borehole No. 2 at the depth 123.7-173.7 m) to the Lower Sangaste Subformation. Liivrand (1991) includes all the mentioned sediments to the Järva Formation.


Middle Pleistocene

Sangaste Formation

The Sangaste Formation is correlated with the Dainava Stage in the southern Baltic, the Oka Stage in the European part of Russia and with the Elsterian Stage in Western Europe.

The lowest diamicton unit in Estonia termed as the U p p e r S a n g a s t e S u b f o r m a t i o n (named after the Sangaste Parish north-east of the Town of Valga) till, is very compact, brownish, sometimes greenish in colour with indications of shearing. It rests directly upon the bedrock and is found only in the bottommost part of ancient valleys. The thickness of the till bed is small - 15 m at Puiestee, 10.7 m at Sudiste and 5.4 m at Mägiste. Borehole 177 (Puiestee) at a depth of 169.0–207.0 m was chosen for the stratotype section (Raukas et al. 1993). The clast composition is different: in southwestern Estonia crystalline rocks are clearly prevailing (up to 95%), but in southeastern Estonia their amount is only 25–60%.

The high content of Vyborg rapakivi and Suursaari quartz porphyries in southeastern Estonia and the absence or a very low content of rapakivi from southwestern Finland suggest that the deposition of till was due to the southward flowing ice. The poorly sorted diamicton is richer in clay particles than the uppermost till units. The latter are high in kaolinite (up to 30–35%) derived from the weathered bedrock. Due to the influence of Devonian sand- and siltstones, the sand and silt fractions of till are richer in quartz and contain less feldspars and carbonates than other till units (Raukas 1978).

According to bedding conditions (Kajak 1995), to the Upper Sangaste Subformation belong grey and brownish till beds and glaciofluvial deposits below organogenous bog and lacustrine deposits in the Karuküla and Kõrveküla sections, up to 23 m in thickness.


Karuküla Formation

The Karuküla Formation (interglacial) is palynologically correlated with the Butenai Stage (interglacial) in the southern Baltic, the Likhvinan Stage in the European part of Russia and the Holsteinian of Western Europe.

The Karuküla type site is situated in southwestern Estonia, in the Pärnu County, about seven kilometres south of the Town of Kilingi-Nõmme (Fig. 93). It displays continental deposits and was first described by Orviku (1944). The name of the stratotype proposed by Kajak et al. (1976) is inaccurate because the section is actually situated in the Keskküla Village. Due to the rather long history of investigations and wide recognition of the site, changing of the stratotype’s name was considered unpurposeful.

The Middle Pleistocene (Likhvinan, Holsteinian) age of the section was first suggested by Danilans (1966) and Voznyachuk (1966) and later established by Liivrand (1984, Velichkevich & Liivrand 1976, 1984).

The information currently available on the Karuküla site has been derived through the study of about 70 boreholes and excavations. The interglacial deposits are probably of allochtonous bedding (Levkov & Liivrand 1988). There seem to be three large and two small erratics and two lumps of Holsteinian deposits within one stratigraphic level measuring 105 m horizontally and 3.25 m vertically (Liivrand 1991). The Karuküla section and its palaeobotanical characteristics have been described in detail in several publications (Liivrand 1972, 1984, 1990, 1991). Another well investigated site of the Karuküla Formation is at Kõrveküla near Tartu (Liivrand & Saarse 1983).


Ugandi Formation

The Ugandi Formation, called after an ancient South-Estonian and North-Latvian area, where those deposits are most widespread, is correlated with the Žeiminiai Formation in Lithuania, Kurzeme in Latvia, Middle Russian in Russia and Saale in Western Europe. In some places Middle Ugandi interstadial beds have been described. Borehole 6 on Prangli Island (depth 75.5-123.0 m) and borehole 268 at Valguta (13.1-35.0 m) have been established as the unit and boundary stratotypes for northern and southern Estonia, respectively (Raukas et al. 1993).

The till of the L o w e r  U g a n d i  S u b f o r m a t i o n, which is correlated with the Dniepr till in Russia and the Žemaitia till in Lithuania, is reddish-brown both in northern (Prangli, Naissaar, Suurpea) and southern Estonia (Mägiste, Lanksaare, Sudiste) and up to 50 m thick (Mägiste). The till is compact and occurs in uplands mainly in sheltered position or rests in ancient valleys. The clast lithology (high content of Vyborg rapakivi in eastern Estonia) indicates that the Lower Ugandi till was deposited by the southward flowing ice (Raukas 1978). Clasts in northern Estonia are almost completely fragments of crystalline rocks, whereas in southern Estonia their composition reflects both the local provenance (up to 10% of local Devonian sand- and siltstones) and the influence of the outcropping carbonaceous rocks on the way of the moving ice (50-60% carbonaceous clasts). Of all Estonian tills, it has the highest content of sandy fraction. In clay fraction illite (50-70%) prevails, but also the content of kaolinite is rather high (20-45%).

M i d d l e  U g a n d i sands, silts, loams and sandy loams often contain rebedded pollen and their stratigraphic position is not clear (Liivrand 1991).

U p p e r  U g a n d i till is massive to slightly stratified, reddish-brown in the fore-klint area (Prangli, Juminda) and grey in northern (Sõrve, Saadjärv) and southern (Rõngu, Suur-Munamägi) Estonia. The till unit is up to 70 m thick (Suur-Munamägi) and often cemented with carbonates. According to its composition (absence of Vyborg rapakivi and quartz-porphyries from the Island of Suursaari), the till entrained by southeastward flowing ice. Practically all the clasts in the fore-klint area originate in the crystalline basement, but in other areas carbonate rocks prevail (65-80%). In southern Estonia, this till has the highest content of silt particles and the lowest content of Devonian material. Illite (65-80%) prevails and the content of kaolinite (10-20%) is low in the clay fraction.

The aqueoglacial deposits of the Lower (Puiestee 60 m) and Upper (Vääna-Jõesuu 60 m) Ugandi subformations are rather thick and variable in composition (Raukas 1978).


Upper Pleistocene

Prangli (Rõngu) Formation

The Upper Pleistocene in Estonia begins with the well-known Eemian interglacial deposits in Western Europe and Mikulinan in Eastern Europe. In the Regional Scheme of the Baltic area this interglacial is called the Merkine Interglacial after the town in southeastern Lithuania. The Eemian (Mikulinan) deposits, both continental (Rõngu) and marine (Prangli), correlated on the basis of the pollen assemblage zones, are in good stratigraphical agreement (Liivrand 1991).

The continental Eemian deposits at Rõngu were investigated in particular detail about half a century ago (Orviku 1939, Thomson 1939a, 1941). Later, complementary investigations were carried out in several other sections (Küti, Kitse, Peedu) by Liivrand (1977) and Kajak (1995).

In the sixties, marine Eemian deposits were found on Prangli Island in the Gulf of Finland (Kajak 1961) and subject to palynological (Liivrand & Valt 1966, Liivrand 1974, 1987, 1990, 1991) and diatom (Cheremisinova 1961) studies. Later, marine deposits of the Prangli Formation were found in several places (Põhja-Lehtju, Väike Tütarsaar, Kihnu a.o.).

A stratotype section at a depth of 67.6-75.5 m in borehole 6 on Prangli Island and a parastratotype for the continental deposits in borehole 264 (2.3-7.8 m) and excavation II (2.0-5.8 m) on the lands of the Vaeva Farm, 2 km west of Rõngu, were established for the Prangli Formation (Raukas et al. 1993). The name for the formation was proposed by Kajak et al. (1976).


Järva Formation

The name of the formation was proposed by Kajak et al. in 1976 after the Järva County in central Estonia where a typical grey till of the last glaciation is widespread in the drumlins and lowland near the Town of Paide. The Järva Formation is correlated with the Nemunas Formation in Lithuania, the Baltia in Latvia, the Valdaian Stage in Russia and the Weichselian in Western Europe. The Vääna-Jõesuu (13-70 m) and Kitse boreholes (0-31.1 m) were chosen for stratotype sections in northern and southern Estonia, respectively (Raukas et al. 1993).

The Kelnase beds were named after the village on Prangli Island. In the Prangli section, they are represented by clayey silts with the pollen spectra characterized by an increasing quantity of Betula nana (40-80%) and herbs (tundra species). Gramineae and Cyperaceae are common. Selaginella selaginoides, Lycopodium alpinum and Artemisia arctica are present. A cryophilous and hydrophilous vegetation refers to the approaching glacial advance (Kajak et al. 1976, Liivrand 1991).

The Valgjärve beds, named after the lake in southern Estonia, are represented by grey till in northern and purplish-grey till and related aqueoglacial deposits in southern Estonia. The purplish-grey till was proposed for a specific stratigraphical unit by Orviku (1939) and described lithologically by Orviku (1958a) and Raukas (1963a, 1978). In the Kitse borehole No. 19 near Lake Valgjärv at a depth of 4.2-31.1 m, the till of the Valgjärve bed covers the organogenous deposits of the Prangli (Rõngu) Formation (Kajak 1995).

The Savala beds named after the village in northeastern Estonia belong to the M i d d l e J ä r v a S u b f o r m a t i o n (Kajak et al. 1976). The type section (borehole 7854, depth 25.8-30.2 m) is situated in the Savala ancient valley about 120 km east of Tallinn. It is mainly filled with grey-coloured varved clays. The pollen and spore composition of the intermorainic layer suggests dry periglacial conditions (Liivrand 1985, 1991). The Savala interstadial warming was not accompanied by any substantial development of forests.

The Võrtsjärve beds, named after Lake Võrtsjärv, are represented mainly by tills of different colour of the last glaciation and aqueoglacial deposits above and beneath the till. In several places some till layers with thin intermorainic interstadial or interphasial sediments occur (Orviku 1939, Raukas 1963a). Tills of the last glaciation on the Cambrian blue clays, sand- and siltstones in the fore-klint area are bluish-grey, mostly clayey and contain mainly clasts from Finland and the bottom of the Gulf of Finland. On the crystalline basement, the till is brown or reddish-brown. Stony tills on the Ordovician and Silurian bedrock are enriched with the local carbonaceous material (Photo 29). The constituent clasts are mainly angular. Tills on the Devonian sand- and siltstones are reddish-brown. The rather well-rounded local carbonaceous and erratic crystalline material occurs in various ratios in the cobble and pebble fractions and are under the influence of the Devonian bedrock, comparatively rich in sand and silt fractions (Orviku 1958a, Raukas 1978). In the stratotype area - the basin of Lake Võrtsjärv, both grey carbonaceous (Valma) and reddish-brown (Tamme) tills are widespread.


Stratigraphy of Late-glacial deposits

The Upper Järva Late-glacial deposits are divided into Arctic (Oldest Dryas, Bølling, Older Dryas) and Subarctic (Allerød, Younger Dryas) chronozones (Table 14). According to the decision of the INQUA Congress in Paris in 1969, the Holocene/Pleistocene boundary is accepted as 10,000 14C years.

Traditionally, the Late-glacial interval in Estonia starts from the accumulation of Rauna interstadial deposits in central Latvia (Kajak et al. 1976). In the Raunis section, interstadial sands with alternating layers of silt and clay, which contain peat and plant remains, lie between two layers of till to the southeast from the Town of Cēsis, on the right bank of the Raunis River. Organic remains from the Raunis section have been dated in several laboratories (13,390±500: Mo-196; 13,250±160: TA-177; 13,320±250: RI-39) and the results obtained are in good agreement (Punning et al. 1968).

In mainland Estonia and on the islands of the Gulf of Finland, Eemian (Prangli) deposits or pre-Weichselian tills are in some places (e.g. Prangli Island) overlain by four till beds, the exact age of which is uncertain. The upper till beds are regarded as stadial ones of Late Weichselian age representing secondary oscillations of the ice sheet. It is also possible that a two- or three-layered till beds may consist of contemporaneous basal and ablation tills from a single glacial event (Raukas 1963a). Locally, Haanja/Otepää and Pandivere/Palivere tills are separated by terrigenous layers containing subfossil molluscs (Kameri, in Orviku 1939) and pollen assemblages of a cool character. Tills of the Haanja, Otepää, Pandivere and Palivere stadials have specific colour and lithological composition (Raukas 1963a, 1978) and can be regarded as lithostratigraphical units of the lowest taxonomical rank (beds).

In some places intermorainic layers have been dated by the 14C method, but the results are contradictory. In the Kurenurme section, southeastern Estonia, remains of Salix wood were taken from sandy loam overlying Haanja till (Ilves et al. 1974). Quite reliable radiocarbon datings of the wood (12,650±520: TA-57) and organic detritus (12,420±100: Tln-35) indicate that these deposits accumulated at the beginning of the Bølling Interstadial. Unfortunately, the process of the deposition of the organic material is not clear (Karukäpp 1986) and this hampers the usage of the section in the till stratigraphy. In the Kaagvere section southeast of Otepää, the dates obtained on the reddish-brown till (15,150±575: TA-50, >30,000: TA-36) suggest redeposition of older interglacial material. The palynological characteristics of interstadial layers between the stadial till beds are not clear either. Probably, these layers contain a lot of material redeposited from older interglacials (Liivrand 1991).

Palynological studies of pre-Allerød deposits above the till beds in Estonia give evidence of severe climatic conditions throughout the Arctic period. They do not permit the layers related to the Bølling amelioration to be distinguished. Such deposits may be present in southern but hardly in northern Estonia. In the section of Haljala (Männil & Pirrus 1963), a pollen assemblage suggesting a brief interval of warming, possibly Bølling, has been reported from a sandy interlayer at a depth of 10.5-11.2 m. However, as its redeposition in the sandy sediment is not excluded, the kind of pollen data interpretation must be taken with great caution (Pirrus & Raukas 1969), and the more that no corresponding warm interval is known from any other site in Estonia.

Deposits of Older Dryas age occur both in northern and in southern Estonia. The lower boundary of the Older Dryas is undefined in Estonia (Kajak et al. 1976), but probably it is the boundary between the Otepää and Kurenurme beds.

In the light of the pollen evidence, the retreat of the ice margin from the Haanja position started during the transitional from Oldest Dryas to Bølling time and the deglaciation of the Estonian territory was completed during the second half of the Allerød (Pirrus & Raukas 1969).

According to Reet Pirrus (Pirrus & Raukas 1996), some more or less clear trends in the vegetation history could be given (Table 15).


Older Dryas Chronozone (Dr2)


Artemisia - Chenopodiaceae palynozone

The Older Dryas Chronozone is represented by glaciolacustrial varved clays or rhythmically laminated silts and sands and overlain by lacustrine silts and clays. In the southern part of Estonia, minerogenic lacustrine sediments may contain minor amounts of plant remains. The thickness of deposits ranges from 1.3 to 11.3 m. In the Older Dryas about 12,000 years ago the Baltic Ice Lake formed and corresponding deposits started to accumulate.

This zone is characterized by high herb pollen percentages (Artemisia, Chenopodiaceae, Helianthemum, Cyperaceae, Gramineae, and several other species of primary vegetation) along with Betula nana L.


Allerød Chronozone (Al)


The Allerød Chronozone is represented by lacustrine clays and silts (0.15-1.85 m in thickness) with blackish-grey interlayers and the Baltic Ice Lake sediments. Scattered plant remains, mostly leaves and stalks of Bryales moss are common in lake sediments.

The Allerød Chronozone is subdivided into two pollen zones (Pirrus & Raukas 1996): a) Pinus-Betula Zone (ALa), b) Pinus Zone (ALb).

The lower boundary of AL Chronozone is fixed with a rather distinct increase of AP pollen and decrease of herbs (Artemisia, Chenopodiaceae) and Betula nana L.

Characteristic of AL Chronozone is the prevalence of tree pollen. Betula shows a rapid increase and towards the uppermost part of the zone Pinus increases distinctly and has its Late-glacial culmination. At the same time, herb pollen is at its minimum. Betula nana L. is constantly present in low percentages. The variety of bog and meadow species of terrestrial herbs and water plants has increased. Xerophytes, halophytes, heliophytes and tundra plants are continuously present, but in low values. Fine preservation and abundance of pollen as well as the regularity of pollen curves indicate their bedding in situ.


Younger Dryas Chronozone (Dr3)


Artemisia - Betula nana Zone

The younger Dryas Chronozone is represented with the Baltic Ice Lake and Yoldia Sea sediments and by lacustrine silts and clays, often with fine-grained sand interlayers. Bryales remains are scattered or occur as thin layers, occasionally abounding in hydrotroilite. The thickness of lacustrine deposits ranges from 0.2 to 4.0 m.

The zone boundary AL/DR3 is placed at the strong and rapid increase of the content of herb pollen (particularly Artemisia) and Betula nana L. This zone is characterized by remarkably high frequency of herb pollen ranging from 40-60%. Maximum values of Betula nana L. pollen in different profiles range from 20 to 25%. The Late-glacial culmination of Picea is either in the lowermost (Võru, Visusti, Haljala) or uppermost (Remmeski, Loobu) part of the pollen zone.

The boundary DR3/PB is placed at a rapid increase of tree pollen, prevailingly Betula (about 80%, in SE Estonia up to 90%) and Pinus (about 20%).


Holocene deposits and their stratigraphical subdivision

The Holocene continental deposits, occasionally rather thick, occur practically everywhere above the Pleistocene deposits. Unfortunately, the offshore and nearshore marine deposits are characterized by numerous unconformities and rapid facies changes and in many sequences gaps cover longer time spans than the preserved strata. The main stages in the Baltic Sea history are known from the very beginning of the century, but they have never been properly defined as stratigraphical units (Hyvärinen & Raukas 1992). Therefore, the stratigraphical scheme of the Holocene deposits (Table 16) is mainly based on the continental deposits. The existence of the four major phases in the postglacial history of the Baltic – the Preboreal Yoldia “Sea”, the Ancylus Lake, the Litorina Sea and the Limnea Sea – is recognised.

In Estonia, there are 9836 peat bogs and about 1150 lakes greater than 1 ha in area (Mäemets 1976, Orru 1992). The peat is at its thickest (16.8 m) in the small Vällamäe kettle hole. The peat deposits are usually 8-10 m thick. The greatest thickness of organic lacustrine deposits is 18 m (Väimela-Alajärv), lake marl 6-7 m (Tapa, Kulina), travertine 5-6 m (Loosi, Rõuge), alluvial deposits 15 m (Väike-Emajõgi) and aeolian deposits 15-20 m (Sininõmme, Kõpu, Rannametsa). In the first half of the century, the palynological approach (Thomson 1925) was applied to the stratigraphical studies in Estonia, at the end of the fifties, physical dating methods were taken into use (Ilves et al. 1974). P. Thomson investigated lake and mire deposits in about 20 localities and modified his first (Thomson 1925) Estonian Holocene stratigraphical scheme in several high standard publications (Thomson 1926, 1929, 1930, 1933). Some 30 years later K. and L. Orviku published the next Holocene stratigraphical scheme (Orviku 1956, Orviku L. 1960).

The first official stratigraphical chart of the Estonian Holocene deposits was compiled under the leadership of Prof. K. Orviku and accepted in May 1976 (Kajak et al. 1976). The second official stratigraphical chart presented in this book (Table 16), was compiled by R. Pirrus, A. Raukas and S. Veski (Raukas et al. 1995b). The part of the scheme dealing with continental deposits is based on the studies by H. Kessel, R. Pirrus, A. Sarv, L. Saarse, K. Kimmel, T. Koff, L.-K. Königsson, S. Veski and A. Poska. The investigations of J. Lutt, H. Kessel and A. Raukas underlie the subdivision of marine deposits. The scheme was accepted at the session of the Estonian Stratigraphical Commission on May 6, 1993 and a week later it was approved at the Stratigraphical Conference of the Baltic States in Vilnius. The regional chart for the Baltic States was approved at the same time. They both followed the Scandinavian scheme (Mangerud et al. 1974). The charts have parallel subdivisions for the continental and marine deposits.

According to the international rules, stratigraphical charts should be based on the unit and boundary stratotypes. Unfortunately, up to now there are no officially accepted stratotype sections for Holocene deposits in Estonia or in the other Baltic States. This work is in progress.

Each site with its own local pollen assemblage biozones is effectively its own stratotype, but no stratotype can exist for the regional pollen assemblage biozones, which are artificial synthesis (Turner 1989). The same type of artificial synthesis is the proposed local stratigraphical chart (Table 16), based on the multiple sections throughout Estonia, all having their own characteristics. As the pollen zones are time transgressive, the boundaries between palynological chronozones have not been drawn and this makes the chart useful and applicable in Estonia as a whole.






Water-bearing formation

R. Perens & L. Vallner


Basic data

In terms of groundwater formation and circulation, the groundwater system in Estonia can be divided into three principal hydrostratigraphical units.

1. T h e   Q u a t e r n a r y   d e p o s i t s. The sandy and clayey Quaternary deposits and peat form porous aquifers with mainly unconfined groundwater which are directly influenced by meteorological conditions. The whole infiltration water percolates into the Quaternary cover and the greater part of groundwater discharge flows through it. The upper portion of the Quaternary cover or sporadically all Quaternary deposits belong to the aeration zone where a lot of water circulates by the agency of capillary force or evaporates, in addition to the filtration flow.

2. T h e   b e d r o c k. The terrigeneous and carbonate Palaeozoic and Proterozoic rocks form porous, fissured and karstified, mostly confined aquifers, which are isolated from each other with aquitards of different isolation capacity. In the karst cavities near the ground, shallow groundwater flows very fast and its chemical composition is close to that of the surface water. However, the deeper strata contain pre-Quaternary groundwater, which is high in total dissolved solids (TDS) and moves very slowly under natural conditions.

3.  T h e   c r y s t a l l i n e   b a s e m e n t. Predominantly pre-Quaternary groundwater in the fissures of igneous and metamorphic rocks contains a high rate of TDS and under natural conditions it is sporadically almost stagnant. The lower portion of the crystalline basement serves as an aqifuge for the whole overlying water-bearing formation in Estonia (to the exclusion of the hypothesis about water originating and arising from the depths of the Earth’s crust).

Aquifer and aquitard are the units of detailed hydrogeological stratification of the water-bearing formation. Aquifer is a relatively homogeneous water-bearing layer or rock with similar water conductivity and storage capacity yielding water in a useable quantity to a well.

According to the hydraulic conductivity value K*, the degree of the permeability of water-bearing strata is the following:

                                102 ≤ K < 1 very low

                                 1 ≤ K < 3 low

                                 3 ≤ K < 10 medium

                                10 ≤ K < 30 high

                                30 ≤ K < 70 very high

                                K > 70 extremely high

Permeability in a lateral direction can be up to 100 times higher than in a transversal direction.

Aquitards are the strata, the transversal conductivity Kt of which is generally less than 10-2 m/d. The following degrees of impermeability can be distinguished:

                                10-2 > K > 10-4 weak

                                10-4 ≥ Kt > 10-6 medium

                                10-6 ≥ Kt > 10-8 strong

                                Kt < 10-8 very strong

Not a single aquitard with the above-mentioned filtration characteristics has an absolute isolation ability. According to this gradation, even strong aquitards are permeable to up- or downward groundwater flows, the total amount of which in large areas can extend up to 104 m3/d.

Aquifers which lie one over another are not necessarily isolated with aquitards. The rocky complex consisting of aquifers and aquitards with similar hydraulic characteristics but with different rock types is termed aquifer system.

In terms of the real water supply, the aquifers and aquifer systems can be subdivided into sufficiently water yielding aquifers and aquifer systems (with specific capacity of wells correspondingly q > 0.1 l/(s×m) ≈ 10 m3/(d×m), K > 1 m/d) and weakly water yielding aquifers and aquifer systems   (q < 0.1 l/(s×m), K < 1 m/d). Aquifers and aquifer systems can be sufficiently or weakly water yielding either sporadically (locally) or in the whole distribution area. According to the aforenamed criteria, at least 23 aquifers and 4 regional aquitards can be distinguished in the water-bearing formation of Estonia (Table 17, Figs. 94, 95).


Water in the Quaternary cover

The technogeneous deposits (tQIV) in settlements mostly consist of stuff and building waste. Besides, there are 495 dumps of different size in Estonia. In the mining region extensive areas are under spoil heaps and oil shale plateaus. The water in technogeneous deposits is usually highly polluted. The water leaking through ash hills and dumps is dangerous to the environment.

The boggy deposits (bQIV) are mostly represented by peat. Under natural conditions, the water generally lies at a depth of 0.1...0.5 m, being deeper only in the fields of milled peat. The conductivity of peat is 0.3...1 m/d. The inflow of water into experimental pits amounts to 1...10 m3/d per 0.5...1 m of drawdown. Bogs recharge from precipitation, while the replenishment to fens is also from lateral flows of unconfined groundwater and vertical flows of confined groundwater from deeper strata. The water of boggy deposits has a nasty taste and smell and is practically not used for the water supply.

The aeolian deposits (vQIV) are mainly represented by fine-grained and well- sorted sands of dunes on the northern coast of Lake Peipsi and on ancient and present beaches of the Baltic Sea. Due to its chain-like morphology, the upper and greater portion of a dune is usually dry; water occurs in the lower part at a depth of 10...15 m from the surface. The yield of wells does not usually exceed 1 to 5 m3/d. At the sea coast the occasional intrusion of brackish sea water can take place.

The lacustrine deposits (lQIV) occur in limited areas in intermittent strata of loamy sand, loam and sapropel mostly in the Alutaguse and Võrtsjärve lowlands. The strata are poorly permeable and not suitable for water supply.

The alluvial deposits (aQIV) are represented by gravel, sand, loamy sand and loam of river valleys with a total thickness of up to 15 m. Due to the limited distribution, they do not have any practical importance.

The marine deposits (mQIV) are up to six, occasionally even more metres thick and consist mostly of sand and coastal gravel which are found in Lower Estonia. In places, the water level can lie close to the ground. Water can sometimes be weakly confined due to the occurrence of clayey interlayers in sands. The discharge of 2...4-m-deep wells ranges from 10 to 60 m3/d and this water is used in many households.

The glaciolacustrine deposits (lgQIII), with a total thickness of 5...10 m, cover a large area and are represented by fine-grained sand, loamy sand and varved clay. Sands and light loamy sands are weakly or sufficiently water yielding with their conductivity varying from 0.1 to 5 m/d. Many wells with the discharge ranging from 0.5 to 20 m3/d have been sunk into these deposits.

Varved clay (lgQIII) with a thickness of up to 22 m and transversal conductivity less than 10-4 m/d, forms medium and strong local aquitards all over Estonia. The largest aquitard (30 km2) occurs in the catchment of the Kasari River in western Estonia. Varved clay effectively protects deeper aquifers from pollution.

The glaciofluvial deposits (fQIII-II) form frontal aprons, eskers and deltas and occur in some buried valleys. They consist mostly of sand and gravel, the conductivity of which is generally 5...10 m/d, in some places even up to 100 m/d. Due to this, the wells tapping the glaciofluvial deposits are generally high yielding. Glaciofluvial sediments in buried valleys are usually confined by clayey glaciolacustrine deposits and till. Public water intakes with a pumping rate of up to 10,000 m3/d tap glaciofluvial aquifers in the buried valleys at Vasavere near Jõhvi, at Raadi-Maarjamõisa in Tartu, and at Männiku-Pelguranna in Tallinn.

The glaciogeneous sediments (gQIII) cover almost 2/3 of Estonian territory. Weakly or sufficiently water yielding are the loamy-sandy varieties of till and sporadically spread lenses of sand and gravel in till with a thickness of a couple of metres. The conductivity of loamy-sandy till ranges from 0.01 to 1.0 m/d. The majority of up-to-10-m-deep dug wells all over Estonia get water from till. The discharge of these wells is predominantly 0.2...2 m3/d. Usually the water table is at a depth of 1.5...3 m from the surface, quite often it is at a depth of 8...12 m, in the Otepää and Haanja heights occasionally even 20 m below the ground. In late summer, shallow (2 - 5 m) wells in loamy-sandy till often run dry. The loamy-sandy till with the conductivity of 10‑3...10-4 m/d is considered a weak or medium aquitard.


Water in the bedrock

The Upper Devonian aquifer system (D3) consists of karstified and fissured dolomites and dolomitized limestones of the Dubniki and Pļaviņas stages. The total thickness of this aquifer system is 17...25 m and it covers some 500 km2 in southeastern Estonia (Fig. 94). The siltstone of the Snetnaja Gora Member with interlayers of clay in the lower portion of the Pļaviņas Stage forms an aquitard with medium isolation ability. The aquifer system is overlain by the Quaternary cover with a thickness of 40...80 m. Groundwater is mostly confined and its potentiometric surface lies at a depth of 3...8 m from the ground. Big sink-holes through which melt- and rain-water percolates fast into the karstified bedrock strata occur at Rõuge, Meremäe, Meeksi and some other places. Conductivity of karstified carbonate rocks varies between 1...50 m/d. According to this, the specific capacity of wells ranges from 0.1 to 6.0 l/(s×m), predominantly it is 1 l/(s×m). Due to its limited occurrence, the Upper Devonian aquifer system is used for the public water supply in a few places only.

The Middle Devonian aquifer system (D2) is extending in southern Estonia (Fig. 94) between the Gulf of Liivi (Riga) and Lake Peipsi. It consists of sand- and siltstones with interlayers and lenses of clay of the Amata, Gauja, Burtnieki and Aruküla stages. Clayey material prevails in the Amata Stage, forming with the Snetnaja Gora Member a weak or medium aquitard between the Upper and Middle Devonian. One third of the volume of this aquifer system includes clayey rocks which serve as weak or medium aquitards and, for this reason, probably form several confined aquifers of local distribution (Verte 1965). The occurrence of the latter has not been sufficiently proved yet.

The northern boundary of the distribution of the Middle Devonian aquifer system lies approximately on the Häädemeeste - Mustvee line. To the south from this line, the thickness of the aquifer system increases up to 250 m in the Haanja Heights. The aquifer system outcrops only occasionally in deeper river valleys, elsewhere it is covered with Quaternary deposits, ranging 5...80 m in thickness. In uplands the potentiometric surface lies at a depth of 10...15 m from the surface, while in lower areas a lot of flowing wells are encountered (Tõrva, Valga, Antsla, Võru, etc.). The absolute height of the potentiometric surface ranges from 80 to 130 m in the Otepää and Haanja heights, in the Sakala Upland it is between 50...80 m.

The lateral conductivity of aquifer system is rather equable: predominantly 1...3 m/d. Transmissivity reaches 200...500 m2/d in the Sakala Upland, Otepää and Haanja heights, elsewhere it is usually less than 100 m2/d. The storage coefficient amounts to 5×10-5...10-3. The discharge of wells changes between 200...700 m3/d per 3...7 m of drawdown. The specific capacity of wells is predominantly 0.4...1 l/(s×m). The Middle Devonian aquifer system is used for the public water supply mainly in the areas south of the Häädemeeste - Põlva line, but also in the towns of Tartu, Viljandi, Elva and Kallaste.

The Narva regional aquitard (D2nr) consists of layers of siltstone, dolomite, marl and clay of the Narva Stage with a total thickness of up to 90 m. In southern Estonia, these layers form the uppermost effective bedrock aquitard, the transversal conductivity of which is 10-5...10-4 m/d, in places 10-6 m/d or even less. The clayey layers of the Narva Stage serve as a regional aquitard for the whole Baltic Artesian Basin (Juodkazis 1989). The rocks in the upper portion of the stage supply water for the area between Tartu and Mustvee and for Ruhnu Island. The specific capacity of wells is 0.06...0.2 l/(s×m). The Narva aquitard separates the Middle Devonian aquifer system from the underlying water-bearing strata.

The Middle-Lower-Devonian aquifer system (D2-1). The Narva aquitard is underlain by the water-bearing layers of the Pärnu Stage (Middle Devonian) and Rēzekne and Tilžė stages (Lower Devonian) which consist of fine-grained weakly cemented sand- and siltstones with interlayers of clayey and dolomitized sandstone. Together with the underlying Silurian strata the layers are used for the public water supply in Pärnu, Viljandi and Tartu. The association of water-bearing strata has been named the Middle-Devonian-Silurian aquifer system and the united account has been kept of its water extraction and water resources (Savitski et al. 1996). However, in view of the collector characteristics of the rocks, it would be more correct to treat the complexes of terrigeneous and carbonate rocks separately.

In southern Estonia, the Middle and Lower Devonian aquifer system with a thickness of up to 100 m lies at a depth of more than 200 m below sea level (Fig. 95). The water is predominantly confined. In lowlands, where the potentiometric surface extends above the ground, flowing wells occur. In the uplands, the potentiometric surface is at a depth of 10...20 m below the ground.

Due to its good collector characteristics, the water yielding capacity of sandstone is relatively high. The discharge of wells is predominantly between 260...700 m3/d by drawdown of 4...10 m. The specific capacity of wells ranges from 0.25 to 1.0 l/(s×m). Conductivity of sandstones is mostly 2...6 m/d, rarely 8...10 m/d. Transmissivity of the aquifer system is 50...500 m2/d, the storage coefficient ranges from 0.001 to 0.15.

The Silurian-Ordovician aquifer system (S-O) is an important source of water supply in the regions north of the Pärnu - Põlva line and on the islands of the West-Estonian Archipelago (Fig. 96). It consists of diverse limestones and dolomites with clayey interlayers. The upper portion of the rocky complex with a thickness of 30 m is extremely cavernous, with numerous cracks and fissures (Heinsalu 1977). Karst cavities form some half-a-metre-high canals trending in the direction of bedrock fissures. Caverns are especially abundant in dolomites and dolomitized limestones. Close to the ground, bigger karst cavities, a couple of metres high and some twenty or thirty metres long, occur in some places. Karst phenomena and fissures are most abundant in the carbonate rocks forming the upper part of the bedrock (Photo 32) - the weathering zone, usually 1...3 m, rarely 5...10 m in thickness. The deeper the lying depth of carbonate rock, the less fissures and cavities; such rocks generally turn into an aquitard. Southward from the Ikla - Elva - Mehikoorma line the Silurian-Ordovician rocks practically yield no water.

Besides traditional aquifer pumping tests, the impeller method (flowmeter logging) has been widely used in studying the filtration characteristics of Estonia’s bedrock (Perens 1984, Perens & Paltanavičius 1989, Perens et al. 1994). The results indicate (Fig. 97) that the Silurian-Ordovician carbonate rocks have fragmentary water conducting zones with parallel lamination and an abundance of fissures. In these 1...2-m-thick zones groundwater flows in a lateral direction (wells included). Water conductivity zones are separated from each other by 5...10-m-thick layers in which groundwater flows predominantly in vertical fissures. Only about 13% of the whole length of the rock complex is covered by lateral water zones (Perens 1984). As an average, there are about 5 water conducting zones per 100 m of vertical cross-section. In a lateral direction the water conducting zones are fragmentary and their stratigraphical level may be more or less the same only within a couple of kilometres. The water zones of quite different stratigraphical levels can be found in wells lying only a few hundred metres from each other.

According to the logging data gathered in more than 300 wells, about half of water in these wells is provided by the upper portion of cross-section with a thickness of 15 m and average transmissivity of 400 m2/d. Downwards the total transmissivity of carbonate rocks is evenly increasing and at depths of 50 and 75 m it is 630 and 700 m2/d, respectively. As new water conducting zones are very rare in deeper layers, the depth of 75 m can be considered the lowest border of sufficiently water yielding layers of the whole Silurian-Ordovician aquifer system. In western Estonia and on islands, the thickness of the main water yielding portion of the Silurian-Ordovician aquifer system is only 30...40 m. In most of Estonia, the total transmissivity of the Silurian-Ordovician aquifer system usually varies between 100...500 m2/d, depending essentially on the distribution of zones of tectonic disturbances. In those areas transmissivity increases and often exceeds 1000 m2/d.

The lateral conductivity in the carbonate bedrock is variable: 10...50m/d in the topmost 20 m, 5...8 m/d at a depth of 20...50 m, and only 1...2 m/d at a depth of 50...100 m. The lateral conductivity of deeper strata does not usually exceed 1 m/d, although occasionally strata with considerably higher conductivity can be found even as deep as 200 m from the surface (Fig. 97). The data of 235 pumping tests in the oil shale region of northeastern Estonia have shown that the lateral conductivity of Ordovician carbonate rocks near the ground is predominantly between 5...300 m/d, while at a depth of 80...100 m it is only 0.1 m/d (Riet 1974, 1976).

According to the water budget calculations, the transversal conductivity of the layers between the lateral water conducting zones is 10-5...10-2 m/d (Vallner 1980, Jõgar 1983). These interlayers serve as weak or medium aquitards confining the local aquifers of different range, the distribution of which is not yet completely clear. In northeastern Estonia (Verte 1965, Gazizov 1971, Jõgar 1977, Savitski et al. 1996, a.o.), including the region of oil-shale mines with plentiful experimental data, the following aquifers and aquitards can be distinguished (from top downwards): Nabala-Rakvere aquifer (O3-2nb-O2rk), Oandu aquitard (O2on), Keila-Jõhvi aquifer (O2kl-O2jh), Jõhvi-Idavere aquitard (O2jh-O2id), Idavere-Kukruse aquifer (O2id-O2kk), Uhaku aquitard (O2uh) and Lasnamäe-Kunda aquifer (O2ls-O1kn). The occurrence of aquitards and aquifers in the Silurian-Ordovician aquifer system has also been proved by hydrogeological modelling. The satisfactory calibration of the more extensive filtration models is impossible without this distinction (Vallner 1996b).

The average lateral conductivity of the Silurian-Ordovician aquifer system is 8.1 m/d. Higher values of conductivity have been recorded in the limestones of the Raikküla Stage (24.2 m/d) and in the Adila Formation (17.0 m/d). The average conductivity of the Idavere and Uhaku stages is only 0.3 and 0.7 m/d, respectively (Perens 1989).

Water in the fissure systems and karst cavities of the carbonate bedrock flows relatively fast. In the outcrop area of aquifer system it recharges from Quaternary deposits and, for this reason, can easily become polluted in the areas with a thin Quaternary cover (Photo 26). The upper portion of aquifer system enfolds an aeration zone which in North-Estonian uplands is up to 20 m and in alvars only a couple of metres thick. Elsewhere, the Silurian-Ordovician aquifer system is more or less confined by the Quaternary cover and the uppermost aquitards of the carbonate bedrock. On the foot of uplands and in river valleys the potentiometric surface can often be 0.5...2 m above the ground which is the reason for many springs and flowing wells (Heinsalu & Vallner 1995). The North-Estonian watershed serves as the main area of head generating. In the Pandivere Upland, the groundwater level is at a height of up to 110...120 m above sea level.

The specific capacity of wells tapping the upper portion of the aquifer system ranges from 1 to 3 l/(s×m), being 3...5 l/(s×m) on average. The specific capacity of wells deriving water from deeper strata does not usually exceed 1 l/(s×m). The average yield of wells by drawdown of 5...10 m is usually 400...900 m3/d. The average specific yield of unconfined aquifer system is 0.02...0.06, depending on the degree of fissuration and karstification. The storage coefficient changes between 10-6...10-3.

The Silurian-Ordovician regional aquitard (S-O) consists of limestones, marls, siltstones, clays and argillites of the Toila, Leetse, Varangu and Türisalu formations (Lower Ordovician), extending at a length of 30 km southward from the North-Estonian Klint. Farther in the south, the aquitard includes all Silurian and Ordovician rocks. Its thickness increases from a couple of metres in the vicinity of the klint up to 200...350 m near the southern border of Estonia. Conductivity is very variable: the lateral conductivity changes between 0.001...1 m/d, extending sometimes up to 2...5 m/d (Fig. 97), the transversal conductivity ranges from 10-7 to 10-5 m/d (Vallner 1980, 1996a).

The Ordovician-Cambrian aquifer system (O-C) underlies the Silurian-Ordovician regional aquitard extending under most of Estonia, except the North-Estonian coastal region, Mõniste-Lokno uplift area and the islands of the West-Estonian Archipelago (Fig. 98). The aquifer system includes the Kallavere-Tiskre aquifer (O1kl- C1ts) in mainland Estonia which consists of fine-grained sandstone and siltstone of the Lower Ordovician Kallavere Formation and the Lower Cambrian Tiskre Formation. The thickness of the aquifer system is 20...60 m, it increases from north to south. The depth from the ground increases from 10...20 m at the North-Estonian Klint up to 500 m on Estonia’s southern border. The main recharge area is the Pandivere Upland where water from Ordovician strata leaks downward through the Silurian-Ordovician regional aquitard and disperses in radial directions as confined filtration flows. There the absolute altitude of potentiometric surface is up to 70 m under natural conditions. In Lower Estonia, Võrtsjärve, Alutaguse, Valga and Varnja-Värska lowlands and in the Väike-Emajõgi Valley, the potentiometric surface of the Ordovician-Cambrian aquifer system extends above the ground. The aquifer system is an important source of public water supply in northern Estonia and it is also intensively used in the towns of Pärnu, Viljandi and Tartu. This has caused several depressions of potentiometric surface (Fig. 98).

The lateral conductivity of the Ordovician-Cambrian aquifer system mostly ranges from 1 to 3 m/d. Transmissivity tends to decrease southwards. However, due to the thickness of water-bearing strata it is 80...130 m2/d in central and southeastern and only 25...50 m2/d in northern Estonia. The yields of wells are predominantly 430...600 m3/d per 10...15 m of drawdown. The specific capacity of wells changes between 0.2...0.4 l/(s×m). The storage coefficient is 2.5×10-5...6×10-3; the specific yield of the aquifer drained is 0.12...0.14.

On the islands of the West-Estonian Archipelago, the Ruhnu (C1rh) and Soela-Tiskre (C1sl-ts) aquifers, isolated by the Irben (C1ir) aquitard, belong to the Ordovician-Cambrian aquifer system. They consist of the Lower Cambrian sand- and siltstone (Ruhnu and Soela formations).

The Lükati-Lontova regional aquitard (C1lk-C1ln), spread in most of mainland Estonia, is represented by siltstones and clays of the Lower Cambrian Lükati and Lontova formations. The total thickness of layers decrease from 90...100 m in northern Estonia to pinching out on the Mõisaküla - Vastseliina line in southern Estonia and in western Estonia (Fig. 99). This aquitard has a strong isolation capacity; the transversal conductivity is predominantly between 10‑7...10‑5 m/d (Vallner 1980, 1996a, unpublished report). In the West-Estonian Archipelago, the Ordovician-Cambrian aquifer system is isolated from the underlying Cambrian-Vendian aquifer system with clays and clayey siltstones of the Lower Cambrian Lükati and Sõru formations.

The Cambrian-Vendian aquifer system (C-V). Cambrian-Vendian terrigeneous rocks occur all over Estonia, except the Mõniste-Lokno uplift area. The water yielding portion consists of sand- and siltstones with interlayers of clay. The difference between the cross-sections of western and eastern Estonia is obvious (Figs. 94, 95, 99). East of the Rakvere - Põltsamaa - Otepää line, the up-to-53m-thick clays of Kotlin Formation (V2kt) divide the aquifer system into two aquifers. The upper, Voronka aquifer, consists of quartzose sand- and siltstone with a thickness of up to 45 m in northeastern Estonia. The lower, Gdov aquifer is formed of up-to-68m-thick complex of mixed-grained sand- and siltstone.

In northern Estonia, the aquifer system is confined by 60...90m-thick clays of the Lontova Formation. Westwards from the Tallinn - Pärnu-Jaagupi line, the Lontova Formation is gradually replaced by interbedding clay and sandstone of the Voosi Formation, which attain a thickness of 90 m in southwestern Estonia. On the West-Estonian islands, the Vendian deposits have also been pinched out and the water-bearing terrigeneous rocks consist only of Cambrian sand- and siltstones with interlayers of clay.

The Cambrian-Vendian aquifer system is the most important source of public water supply in northern Estonia. Intensive water extraction has led to the formation of two extensive depressions of potentiometric level (Fig. 99).

The Voronka aquifer (V2vr) in eastern Estonia consists mainly of quartzose sand- and siltstones of the Voronka Formation, up to 40 m in thickness.The conductivity of rocks ranges from 0.6 to 12.5 m/d, being 2...6 m/d on average. Transmissivity decreases from 100...150 m2/d in northern Estonia to 50 m2/d and even less in the southern direction. In northern Estonia, the specific capacity of wells ranges from 0.2 to 5.8 (average 2) l/(s×m). In central and southern Estonia, it is 0.1...0.3 l/(s×m). Under natural conditions, the potentiometric level in the coast of the Gulf of Finland was 1.5...5.5 m above sea level.

The Gdov aquifer (V2gd) consisting of mixed-grained sand- and siltstone with the thickness predominantly between 40...65 m lies directly on the Precambrian basement. The clay of the Kotlin Formation serves as an upper confining unit. In northern Estonia, the conductivity of water-bearing rocks is 0.5...9.2, average 5...6 m/d. Transmissivity in northeastern Estonia is 300...350 m2/d; it decreases in a southerly and westerly direction to 100 m2/d or less. Specific capacities of wells differ, the average value being 1.5...2.5 l/(s×m). Since most of wells tap both the Gdov and Voronka aquifers, their specific capacity is considerably higher. The potentiometric surface was 3...5 m above sea level under natural conditions in the coastal area of northern Estonia .

Westward from the line where the Kotlin clays are pinching out (Fig. 99), the Cambrian and Vendian water-bearing rocks form the steady Lontova-Gdov aquifer. On the West-Estonian islands, the Vendian sediments are absent and all the rocks deeper than the Lükati-Sõru aquitard (C1lk-sr) form the Voosi aquifer (C1vs). The productivity of wells on the islands of the West-Estonian Archipelago is 3...4 times lower than in northern Estonia.

The crystalline basement (PR1) comprises groundwater in its upper weathered and fissured portion only. The specific capacity of wells does not exceed 0.1...0.2 l/(s×m). Natural potentiometric surface refracts the heads of the overlying Cambrian-Vendian aquifer system. Flowing wells occur in some places. Currently, the water stored in the crystalline basement is not used for water supply.

In extensive regional hydrogeological reviews the Estonian water-bearing formation has been treated as the northern slope of the Baltic Artesian Basin extending from the Gulf of Finland up to Minsk and Warsaw (Juodkazis 1989). This viewpoint, although based on structural and geological aspects and a certain unity of formation of palaeohydrogeological conditions of deep groundwater, is still a theoretical construction. From the applicational point of view, the Estonian water-bearing formation should be considered as an independent artesian basin because the exchange of underground water with the neighbouring areas is less than 0.1% of the total annual groundwater recharge (Vallner 1980).


Groundwater flow

L. Vallner


Main components of the budget and flow systems

In Estonia, groundwater is recharged from rain- and melt-water percolating through unsaturated soils. Mean annual precipitation is in the range of 500...750 mm or 1370...2060 m3/(d×km2), and is lower on the coast and some 10...20% higher on uplands.

The net infiltration (I) for Estonia (total groundwater recharge minus evaporation from the zone of saturation or ca-pillary fringe) has been calculated preliminary from the budget equation comprising the main components of groundwater flow (Vallner 1976, 1980):


I = R + Q + M - W ± V ± S


where R is the groundwater discharge (base flow) m2/d to streams; Q is the pumpage from layers; M is the direct seepage of groundwater to the sea; W is the flux from streams into aquifers (induced recharge); V is the subsurface exchange of groundwater between Estonia and surrounding areas and S is the storage change.

The long-term groundwater discharge to streams (R) has been estimated on the basis of observations carried out during several decades at more than 100 hydrological gauging stations all over Estonia. Apart from the gauging stations, many irregular measurements of the low flow have been made approximately in 1000 stream cross-sections. The gained sporadic low flow data were modified to average base flow value by statistical methods using regular observations of gauging stations (Vallner 1980). The pumpage data (Q) were obtained from state institutions checking the groundwater use. The subsurface fluxes to the sea (M) and groundwater exchange with adjacent areas (V) were calculated by Darcy’s formula.

The total groundwater discharge to the channel network of Estonia (R) is approximately 7,700,000 m3/d, but its intensity varies with regions (Fig.100). Average pumpage from wells and mines (Q) reaches 1,000,000 m3/d causing the inverse fluxes (W) in an amount of ca. 500,000 m3/d from surface bodies of water into aquifers. Direct groundwater seepage to the sea (M) averages 800,000 m3/d. In the west and north, Estonia is bounded by the sea. In the east, its border runs along the Narva River and central portion of Lake Peipsi, in the south the border generally coincides with the divides of major streams. As a result, the exchange of groundwater with adjacent territories (V) is quite insignificant - not more than 10,000 m3/d. For a long-term period the storage change S = 0. The total net infiltration (I) calculated by the budget equation averages 9,000,000 m3/d.

Groundwater discharge to the channel network (the base flow) and the instrumentally checked pumpage make up about 90% of the sum of the right-side members in the above water budget equation. Therefore, the value of net infiltration estimated by budget equation is probably more authentic than that based on indirect data, such as the air temperature and atmospheric humidity, evapotranspiration, etc. (Lerner et al. 1990). After completing the general budget of groundwater flow main components, the distribution of heads was simulated for the whole Estonian water-bearing formation (Vallner & Tobias 1984). Thereby the net infiltration was estimated by the trial and error method pursuing an optimum coincidence between the modelled and measured data.

The total amount of groundwater in the cracks and pores of Estonia’s water-bearing strata is nearly 2000 km3. Both the distribution of the groundwater head and the direction of subsurface flows depend on spacial relations of the topography, surface bodies of water, and impermeable crystalline bedrock. Besides, human impact on the groundwater flow has been continuously increasing during the last five decades. Based on the total effect of the topography and geological structures, three main groundwater flow systems can be recognized in Estonia (Vallner 1980, Tóth 1995)

The local flow system enfolds chiefly the unconfined or locally confined shallow groundwater moving from its recharge area toward the nearest ditches, creeks, rivers, and discharging directly to lakes or to the sea in coastal areas (Fig. 101). The length of the upper branches of the local flow system usually does not exceed a few kilometres, but lower branches which are not drained by small surface waterbodies and discharge to the middle courses of rivers can reach 10...15 km in length. The vertical thickness of the local flow system mostly ranges from 10 to 30 m.

The intermediate flow system takes its rise from Upper Estonia (Fig. 2) where the height of the terrain above sea level is more than 40 m. In the Harju Plateau, Otepää and Haanja heights, Pandivere, Jõhvi, Sakala and Karula uplands and in the Vooremaa area, the local maxima of the groundwater potential energy occur. In the Haanja and Otepää heights the groundwater table is 120...280 m above sea level, while in uplands and plateaus its height mostly ranges from 50 to 100 m. The groundwater head declines down from above in uplands; the difference of heads in some aquitard can be up to 50...60 m, and the corresponding head gradient up to 2.5...3. Consequently, a downward groundwater flow recharges the underlying aquifers from the overlying ones in highlands.

Groundwater moves from Upper Estonia towards Lower Estonia where the surface altitudes are 0...40 m above sea level. Branches of the intermediate flow system discharge to surface waterbodies between absolute heights of 30...40 m in the Võrtsjärv, Alutaguse, Valga and Varnja-Värska lowlands, and in the Väike Emajõgi Valley. The portions of the intermediate flow system formed in the Sakala Upland and in the western part of the Harju Plateau are drained by the lower course of the Kasari River between altitudes 0...30 m, while the lower branches of this flow system discharge to the offshore Baltic Sea. The intermediate flow system is completely confined. Its branches, approaching the discharge areas, bend up and recharge the local flow system from the underside. It is proved by decreasing of the groundwater head from below upward observed in lake depressions and lower course valleys of most significant rivers where a lot of deep flowing wells occur. A portion of the intermediate flow system formed in the Harju and Viru plateaus, Pandivere and Jõhvi uplands discharges through springs or seeps out from layers on the North-Estonian Klint. The lateral length of the intermediate flow system is up to 100 km.

The regional flow system takes its rise in those parts of the Haanja and Otepää heights where the groundwater table is 180 to 280 m above sea level. There the head declines down from above attesting to the existence of a downward groundwater flow. This flow reaching the absolutely waterproof portion of the crystalline basement changes its direction and bends towards the discharge areas which are situated in the depressions of the Baltic Sea and the Gulf of Finland. The regional flow system underlies the intermediate flow system enfolding the portion of the lithosphere where the confined groundwater moves directly into the sea under natural (predevelopment) conditions. In western Estonia, the regional flow system includes the strata of the Ordovician-Cambrian and Cambrian-Vendian aquifer systems lying higher than 250 m below sea level. The Cambrian-Vendian aquifer system belongs completely to the regional flow system west- and northward from lakes Võrtsjärv and Peipsi. The length of lower branches of the regional flow system can reach 300 km between the Haanja Heights and the central part of the Baltic Sea.

The diversity of rock permeability complicates and modifies the head distribution in the subsurface hydrosphere. Therefore, the groundwater movement is mainly parallel to bedding in aquifers, but transversal in aquitards. The latter are often disconnected at tectonic disturbances where relatively intense vertical groundwater flow may occur. The configuration of the potentiometric surface and pumpage data show that all aquitards are more or less permeable in Estonia. Evidence is derived from declining of the head of the Ordovician-Cambrian aquifer system in the opposite directions - from the Otepää Heights and the Pandivere Upland towards the Emajõgi Valley. There the groundwater flows rise up and discharge to the river regardless of the Narva and Silurian-Ordovician confining aquitards with a total thickness of up to 300 m. The transversal permeability of regional aquitards is an eminent, but not completely clarified evidence until now (Tissot & Welte 1978, Brace 1980, Tóth 1995). Apart from porosity, it may well be caused by microfissurization of rocks.

While the potential direction of the groundwater movement is marked by the flow systems described above, the actual quantity and velocity of subsurface flows depend on permeability of layers. Various co-influences of head distribution and permeability are expressed by the vertical zoning of the groundwater flow. In general, the velocity of the groundwater movement decreases with the increase of the flow depth.


Vertical zoning

The  z o n e   o f  a c t i v e  w a t e r  e x c h a n g e  enfolds the upper portion of the water-bearing formation (Table 18, Fig. 101). All aquifer systems overlying the Lükati-Lontova aquitard north of the Pärnu - Tartu line belong to this zone. Southward the zone is limited from underneath by the Ordovician aquitards or the altitude of 250 m below sea level. In the zone of active water exchange, an upper subzone of fast groundwater flow may be distinguished which includes the Quaternary cover and Devonian aquifers less than 50 m below sea level and the Silurian-Ordovician aquifer system, as a whole. Farther down, a subzone of moderate groundwater flow occurs comprising the portion of the Middle Devonian aquifer at a depth of 50-250 m below sea level and the Ordovician-Cambrian aquifer system down to the lower boundary of the zone of active water exchange.

The  s u b z o n e  o f  f a s t  f l o w  is immediately influenced by climate. In this subzone the infiltration is accumulating and evaporation from the aeration zone takes place. From the subzone of fast flow all underlying strata are recharged and also the whole groundwater flow discharging to stream channels permeates through this subzone. As mentioned above, in Estonia the total net infiltration reaches 9,000,000 m3/d, which makes an average of 200 m3/(d×km2) or 73 mm per year, but its actual value significantly varies with areas (Fig. 100). Infiltration is most intensive (350 to 770 m3/(d×km2) in the uplands of northern Estonia, where the carbonate bedrock abounds in karst phenomena. In an area of about 1000 km2 in the central part of the Pandivere Upland, where the channel network is entirely lacking, most of rain- and meltwater, which has not been removed by evapotranspiration, percolates into karst interstices. Owing to the high permeability of the karsted carbonate bedrock, a significant portion of this newly formed groundwater quickly discharges to adjacent streams. That’s why the groundwater table is relatively deep in the upland, being often 7...20 m below the ground surface. With such a depth, the discharge by evaporation from the aeration zone is negligible and the net infiltration reaches its maximum in Estonia - up to 900 m3/(d×km2).

In the uplands of southern Estonia, the net infiltration is twice as high as its average value. Highlands form only 16% of Estonia’s total area but, nevertheless, about 40% of the total groundwater recharge takes place just on the uplands and their slopes. On the plateaus of northern and southern Estonia, the net infiltration is 170...260 m3/(d×km2) but in lowlands it is less than 170 m3/(d×km2).

Mires impede the recharge of deeper groundwater by infiltration. Fens are recharged from rising groundwater flows and bogs are usually located above an effective local aquitard which restricts the downward filtration. Similar conditions occur in the domains of glaciolacustrine varved clays covering an area of some thousand square kilometres.

The intensity of groundwater discharge to the channel network varies in a wide range, but its mean value is 170 m3/(d×km2). On the Harju and Viru plateaus, the groundwater discharge into streams averages 130 m3/(d×km2). The groundwater discharge to the channel network is at its highest, reaching 400...900 m3/(d×km2) in some topographic catchments around the Pandivere Upland where a lot of springs occur. In this area, the groundwater discharge averages 300 m3/(d×km2).

S p r i n g s are essential sources of recharge for many streams in northern Estonia. Commonly, they are associated with tectonic faults in the carbonate bedrock and occur in groups. The central part of the Pandivere Upland where the permanent channel network is lacking, is surrounded by a belt of gravity springs at an height of 80 to 100 m above sea level. Unconfined groundwater discharges through these, often intermittent springs (Heinsalu & Vallner 1995). A significant portion of confined groundwater formed in divides is discharged through ascension springs which are situated farther away from the upland in places where tectonic faults cross the river valleys. There groundwater moves upward along tectonic disjunctions of aquitards. About half of rising groundwater discharges directly to the streams dispersively or through subaquatic springs.

The total discharge of some spring groups draining the karsted bedrock is up to 50,000 m3/d during wet periods, but mostly it ranges from 500 to 10,000 m3/d. Besides the Pandivere Upland, there are a lot of springs in the upper course of rivers flowing out from the Harju Plateau and also in the upper course of the Navesti River. All in all, some 1500 springs occur in North-Estonian uplands and plateaus, but predominantly their discharge is less than 50 m3/d. The abundance of springs points to both the high permeability and vulnerability of the carbonate bedrock.

In the uplands and plateaus of southern Estonia, the subzone of fast flow includes mostly till, but also Devonian sand- and siltstones. The intensity of the groundwater discharge into the channel network ranges from 130 to 350 m3/(d×km2). Of an abundance of springs occurring in this area, the vast majority have a discharge of up to 10 m3/d, and only a few springs 1000 to 1500 m3/d. In Lower Estonia (Fig. 2), the intensity of groundwater runoff usually ranges from 50 to 100 m3/(d×km2), i.e. less than the average. This is due to the low and weakly dissected topography and low permeability of upper aquifers. In the Fore-Klint Lowland, the groundwater runoff reaches 200 m3/(d×km2) which is higher than the average value. The klint abounds in springs, but their discharge usually does not outnumber 10...20 m3/d during the dry period. The total discharge of springs averages 500,000 m3/d in Estonia forming about 7% of the total groundwater discharge to the channel network.

The total groundwater discharge from the subzone of fast flow directly to the sea is 700,000 m3/d. In Estonia, the total lengths of the channel network and coastline are 31,000 and 3800 km, respectively (Hang & Loopman 1995). Thus, the intensity of groundwater discharge per a unit length of the channel network is about 250 m3/(d×km) and per a coastline unit - 180 m3/(d×km). These parameters are in good accordance and prove the reliability of the performed water budget calculations.

In northeastern Estonia, the subzone of fast flow enfolds the oil shale mines where the groundwater head has withdrawn up to 60 m. As a result of dewatering of mines, drawdown occurs in an area of about 700 km2 and the induced recharge of Ordovician aquifers from the channel network reaches 400,000 m3/d.

Of the total amount of groundwater formed in Upper Estonia, 430,000 m3/d flows laterally into Lower Estonia; of that ca. 170,000 m3/d into the Fore-Klint Lowland. The groundwater discharge directly to Lake Peipsi is 150,000 m3/d.

Actual velocity of groundwater movement is variable in the subzone of fast flow. During wet periods, in the outcrop of the carbonate bedrock a lot of intermittent gravity springs, recharged from flows of shallow groundwater, come into being. Experiments carried out by means of dyestuffs have shown that the velocity of such groundwater movement is up to 5000 m/d. In deeper layers the actual velocity of groundwater v is much smaller and it is calculated by the formula v = IK/n, where I is hydraulic gradient, K is conductivity, and n is porosity.

In the above lateral water-conducting zones of the carbonate bedrock, the actual groundwater velocity v predominantly ranges from 1 to 10 m/d under natural conditions. In transversal fissures connecting these zones it is mostly 0.001...1 m/d. Consequently, in the karstified carbonate bedrock of high permeability it may take a month for the polluted groundwater to reach from the ground surface to a depth of 30 m, where usually the cased portion of a well ends. During another month, it may cover some 300 m in a lateral direction. After a year, the pollution may be found at a distance of about one kilometre from the pollution source. Approximately such velocities of groundwater movement have been observed in cases of the oil pollution on the outcrop of carbonate bedrock. In extremely permeable aquifers, groundwater may flow from the surface to a depth of 30...60 m even in a few days. However, in the Silurian-Ordovician carbonate bedrock the actual lateral velocity of groundwater movement ranges from 0.1 to 3 m/d. It is lower in Lower Estonia and in deeper layers.

In South-Estonian highlands, the hydraulic gradient of downward groundwater flows is between 0.01...0.4, but the gradient of lateral flows ranges from 0.0001 to 0.01. The lateral velocity of groundwater is 0.02...0.2 m/d in sandstones; the transversal velocity ranges from 0.001 to 0.005 m/d. In loamy till, the velocity of groundwater movement usually does not exceed 0.001 m/d, but in glaciolacustrial or glacial sandy loam it is up to 0.1 m/d. The velocity of groundwater movement is 0.0005...0.001 m/d in peat, 0.001...0.15 m/d in sand and 10...15 m/d in gravel.

A  s i n g l e  w a t e r  e x c h a n g e  may take place in some short and highly permeable branches of local flow system during a couple of months. Sometimes such branches with a length of up to 0.5 km occur between the local divide and the nearest stream. If the permeability of flow system is moderate or low, then under the same conditions a couple of years are required for a single water exchange. In the case of very low permeability of local flow system, when conductivity is 0.01...0.1 m/d (loam, peat, fine sand), it may take 50...150 years.

The time needed for a single water exchange has been calculated assuming that the existing groundwater will be replaced gradually along the branches of flow system which are isolated from one another. Actually, a more or less intensive water exchange takes place between all adjacent branches of groundwater flow. The most intensive water exchange occurs across the groundwater table directly affected by infiltration and evaporation. If an unconfined aquifer consists of Quaternary deposits or sandstone, the average thickness of which is 10 m and porosity 0.2, then at a common infiltration rate of 200 m3/(d×km2), the single water exchange takes about 25 years. Under the same conditions, in the carbonate bedrock with the porosity of 0.02 it takes only 3...4 years.

The  s u b z o n e  o f  m o d e r a t e   f l o w   is recharged from the overlying subzone of fast flow. The amount of this downward flux is 400,000 m3/d, i.e. only about 4% of total net infiltration. Of that, 160,000 m3/d leaks through the Silurian-Ordovician regional aquitard with average intensity of 9.5 m3/(d×km2) in northern Upper Estonia. The intensity of downward fluxes penetrating the tectonic faults and recharging the underlying Ordovician-Cambrian aquifer system reaches 25...50 m3/(d×km2) or even more in the Pandivere Upland. The portions of the Middle-Lower Devonian aquifer system and the underlying Silurian aquifers which both belonging to the subzone of moderate flow are recharged through the Narva regional aquitard in South-Estonian highlands. In that case, the total amount of downward fluxes is 240,000 m3/d and average intensity 20 m3/(d×km2). The lateral groundwater flux from Upper to Lower Estonia along the subzone of moderate flow is 220,000 m3/d. Of that, 200,000 m3/d returns into the overlying subzone of fast flow by uprising filtration. The flow to shelf deposits and from there into the sea averages 100,000 m3/d, almost an equal amount goes to the underlying subzone of slow flow. Pumpage from the subzone of moderate flow was about 74,000 m3/d in 1995.

The length of lateral groundwater flow branches belonging to the intermediate flow system ranged from 50 to 250 km under natural conditions, the hydraulic gradient was mostly 0.0003...0.0004. The actual velocity of groundwater movement was probably 0.005 m/d in Devonian and Cambrian sandstones and 0.05 m/d in Silurian carbonate rocks. At such velocities, a complete water exchange could have taken place only in the Silurian layers during the last 10,000 years, i.e. since the time the ice sheet retreated from Estonia’s territory (Raukas & Rõuk 1995). During the same period, in sand- and siltstones the groundwater could have moved forward only some thirty or forty kilometres.

Owing to the intensive pumpage, a piezometric depression has formed in the subzone of moderate flow. Local cones occur in Haapsalu, Paldiski, Vasalemma, Tallinn, Kohtla-Järve, Pärnu, and Tartu where the drawdown ranges from 20 to 94 m. At the time being, the groundwater moves in many different directions towards the intakes in the subzone of moderate flow. The hydraulic gradient is 0.005...0.01 or even higher (Fig. 98). The actual velocity of the groundwater movement predominantly ranges from 0.1 to 2 m/d.

The  z o n e  o f  p a s s i v e  w a t e r  e x c h a n g e  under the subzone of moderate flow enfolds the Silurian-Ordovician regional aquitard and all underlying strata south of the Tartu latitude (Fig. 101). Farther in the north, it comprises the Lükati-Lontova regional aquitard, the Cambrian-Vendian aquifer system, the water-bearing portion of the crystalline basement, and in the West-Estonian Archipelago partially also the Ordovician-Cambrian aquifer system. In the zone of passive water exchange, the water moves at a considerably lower actual velocity than in the zone of active water exchange. Therefore, the above-mentioned relatively thick regional aquitards belong to the zone of passive water exchange. Under natural conditions, the zone of passive water exchange recharges from overlying strata in South-Estonian highlands only, in the Pandivere Upland such recharge is negligible.

The  s u b z o n e  o f  s l o w  f l o w  embraces the upper portion of the zone of passive water exchange to a depth of 350 m below sea level. Under natural conditions, the downward filtration from the overlying subzone of moderate flow into the subzone of slow flow averaged 30,000 m3/d which was only 0.3% of the total net infiltration. The uppermost branches of the intermediate flow system with a length of up to 50 km, which belonged to the subzone of slow flow rised up in the region of Lake Peipsi recharging the subzone of moderate flow from below. The lower branches headed along the regional flow system towards discharge areas in the depressions of the Gulf of Finland or central Baltic Sea. The lateral hydraulic gradient of deep groundwater flows ranged from 0.0001 to 0.0003.

The calculated velocities of deep groundwater movement are between 0.0005...0.005 m/d under the above-described conditions, which means that during the last 10,000 years the deep groundwater could have move forward only by some twenty or thirty kilometres and a complete water exchange along flow branches was impossible. This viewpoint is proved by the isotopic composition of Cambrian-Vendian fresh water which shows values of d18O from -1.8 to 2.2% (Vaikmäe & Vallner 1989). Such water must have originated from the thawing of the ice sheet at the end of the Pleistocene and its preservation in the subzone of slow flow is only due to an extremely low velocity of groundwater movement.

At present, pumpage from the subzone of slow flow is about 110,000 m3/d. Pumping wells are mostly situated in the coastal area of northern Estonia within 20 km from the sea. Pumpage is most intensive in Tallinn and Kohtla-Järve where the local centres of piezometric depression have formed and the maximum drawdowns reach 25 and 50 m, respectively (Figs. 99, 102). At the present time, the water moves to the centres of piezometric depressions in the subzone of slow flow. North of groundwater intakes the direction of flows is from the sea to the mainland, i.e. contrary to that in predevelopment conditions. Therefore, an encroachment of brackish sea water into coastal aquifers is taking place in the near-shore area of northern Estonia.

The  s u b z o n e  o f  v e r y  s l o w  f l o w  enfolds the lower portion of the zone of passive water exchange at a depth greater than 300 m below sea level. This subzone includes the lower strata of the Ordovician-Cambrian and Cambrian-Vendian aquifer systems (Fig. 101) south of Elva latitude which comprise water with TDS ranging from 1 to 22 g/l. Due to the lack of experimental data, the velocity of groundwater movement in this subzone is not yet clear. In the lowermost portion of the Estonian water-bearing formation which lies at a depth of 500...700 m below sea level in the vicinity of Ruhnu Island (Fig. 94), the water may be stagnant (Mazor 1995, Mazor et al. 1995). In any case, the water has not become fresh in the subzone of very slow flow during the postglacial period though the TDS of water might have decreased.


Water budget of aquifer systems

In Estonia, the total net infiltration enters first the Quaternary cover (Fig. 103). The downward flow from the Quaternary deposits into the underlying bedrock averages 5,300,000 m3/d, while the direct discharge into the channel network is 3,000,000 m3/d. Discharge through springs is 43,000 m3/d and 320,000 m3/d flows directly to the sea. The flux of bedrock water rising upward and recharging the Quaternary cover from below is 4,100,000 m3/d.

The amount of water flowing from the Quaternary cover into the Upper and Middle Devonian aquifer systems reaches 960,000 m3/d and the recharge from the underlying Middle Devonian aquifer system is 510,000 m3/d. About ¾ of inflow discharges in the form of lateral flows to the channel network across the streambeds. The remaining amount discharges through springs, is extracted by pumping or leaks into deeper layers.

The Middle-Lower Devonian aquifer system is recharged from the Quaternary deposits on its outcrop and farther in the south from the overlying Middle Devonian aquifer system. The total inflow is 350,000 m3/d of which some 80% rises upwards into the overlying Middle Devonian aquifer system or discharges into streams through the Quaternary deposits. Approximately 15% is pumped out or drained directly by the sea, while 5% goes into the underlying Silurian-Ordovician aquifer system.

Of the total downward flow formed in the Quaternary cover, the Silurian-Ordovician aquifer system receives 75% or 4,000,000 m3/d. The induced recharge of the carbonate bedrock due to the dewatering of mines is 500,000 m3/d. The upward recharge from the underlying Ordovician-Cambrian aquifer system reaches 70,000 m3/d in Lower Estonia. Discharges to the channel network through Quaternary deposits and springs are 3,000,000 and 450,000 m3/d, respectively. Pumpage from the Silurian-Ordovician aquifer system, including mine water, averages 760,000 m3/d and discharge into the sea through the shelf is 300,000 m3/d.

The Ordovician-Cambrian aquifer system is recharged mainly from downward flow reaching 150,000 m3/d which comes from the overlying Silurian-Ordovician aquifer system. Of the above amount of inflowing water, one third recharges the underlying Cambrian-Vendian aquifer system, one third rises up into the overlying Silurian-Ordovician aquifer system in Lower Estonia and one third is pumped out or flows into the sea.

Pumpage from the Cambrian-Vendian aquifer system was approximately 110,000 m3/d in 1995. About half of it leaks through the Lükati-Lontova aquitard from the Ordovician-Cambrian aquifer system, and another half has formed on account of lateral flows coming from the side of the Gulf of Finland and central Estonia.


Fluctuation of the groundwater table under natural conditions

Natural fluctuation of the groundwater table has been investigated in many places all over Estonia. The statistical analyses of observation data have shown (Vallner 1982) that the groundwater table in the Quaternary deposits covering the outcrop of the Devonian rocks is 0.1 m higher in the beginning of the year than the annual average level of the water table. Later on, the groundwater table is lowering more or less evenly until the beginning of March when it is below the mean level by 0.05 m. When the air temperature rises above 0oC, the melt-water will percolate into the soil and the spring phase of intensive infiltration starts, lasting until snow has melted in the last decade of April. The amplitude of the spring rise of the groundwater table is 0.4 m whereby the maximum point exceeds the annual mean level by 0.35 m. In late spring and in summer, the amount of groundwater mostly decreases due to the intensive evapotranspiration of the soil moisture. As a result, the groundwater table will lower during about 140 days until the first or second decade of September when the decrease of evapotranspiration caused by lowering of the air temperature will be balanced with infiltration. The amplitude of groundwater level lowering is 0.6...0.7 m in the warm period. The minimum point of the groundwater table is below the annual mean level by 0.3 m in September. The summer lowering of the groundwater table can be retarded or even changed to rising because of occasional rain periods. Intensive infiltration recurs in autumn when a systematic rainfall starts and evapotranspiration is low. Then the groundwater table rises until the soil will freeze in December. During a cold winter the amount of groundwater predominantly decreases due to restricted infiltration.

The fluctuation of the groundwater table in western Estonia is very similar to that in southern Estonia as described above. Only the fluctuation amplitudes in southern Estonia are by 0.1 m less in winter and spring. It may be explained with the higher air temperature in winter, owing to which a portion of melt-water percolates into the soil or discharges directly to stream channels before the main thawing period starts in spring.

The annual amplitude of the groundwater table fluctuation is significantly greater in the karstified and fissured carbonate bedrock of northern Estonia. The spring rising reaches 0.7 m and the summer lowering is about 0.9 m. In the areas of abundant karst phenomena, the annual amplitude of fluctuation can range from 4 to 6 m, exceeding occasionally even 10 m. The spring maximum point and the summer minimum point arrive by one decade earlier than in southern Estonia. Such relatively great fluctuation amplitudes are caused by the karst cavities in the carbonate bedrock which accumulate a significant amount of water in spring, but this water quickly discharges to streams in summer.

Seasonal fluctuations of the water table are remarkably small in peat. The spring rising amplitude does not exceed 0.1 m and the summer lowering averages 0.2 m. The winter lowering lasts until the third decade of March and the spring maximum point arrives in the middle of May.

In confined bedrock aquifers the seasonal fluctuation of the head is commonly similar to the fluctuation of the groundwater table, but the amplitude decreases with depth. Owing to the intensive pumpage, the character of natural seasonal fluctuation can be more or less perverted.


Composition and properties of groundwater under natural conditions

V. Karise


Zone of active water exchange

Infiltration water, comprised in the active water exchange zone of the Estonian groundwater system, obtains the chemical composition typical of groundwater mostly in the aeration zone (Table 19, Fig. 95). The upper 30...50 metres of the active water exchange zone are characterised by oxidized state, while in the lower part a transition from oxidizing to reducing conditions takes place (Põllumajanduslik ... 1994). The passive water exchange zone is entirely under reduced state.

In the active water exchange zone, calcium and magnesium carbonates are practically the only dissolved compounds. Therefore, regardless of the lithological compositon and the redox state of the groundwater the HCO3-Ca-Mg (frequently also HCO3-Mg-Ca) type of groundwater is formed, with the content of dissolved mineral salts under natural conditions being 0.1...0.6 g/l, most commonly 0.3-0.4 g/l. The content of free CO2 in the upper part of the active water exchange zone is prevailingly 20...30 mg/l (the boundary content limits are 0.5...50 mg/l, occasionally even 100 mg/l). With the pH values 7.2...7.6, the concentration of balanced HCO3- in the water is 200...400 mg/l. When carbonates dissolve, then together with HCO3- also Ca2+ and Mg2+ reach water in proportional amounts (most frequently 40...95 and 11...30 mg/l, respectively). Besides, the water is enriched with Na+, Cl- and SO42- (2...20, average 10 mg/l) originating from precipitation or soil.

In the active water exchange zone, neither the groundwater composition nor the amount of dissolved mineral salts is controlled by the lithological composition of rocks. Evidence is derived from the uniform chemical composition of water in springs and the amount of mineral salts dissolved in water. As an exception serves the water stored in the Quaternary sands, in which the proportion of quartz reaches 90%, and also the water stored in Devonian sandstones in southern Estonia in the areas where the Devonian is covered by sand, not by till. In that case the total content of dissolved salts (TDS) may reach 0.1...0.2 mg/l. This kind of water is unsaturated with carbonates, and has maintained the potential ability of dissolving carbonates. In practice, the cases are known, when groundwater in sandy areas has corroded concrete well curbs. This kind of water is used for removing scale from steam boilers.

Groundwater in the active water exchange zone is generally weakly alkaline (pH = 6.8...7.6). The unsaturated water in sandy regions is slightly acid (pH = 5.5...6.5). The water in bogs is acid (pH = 3.0... 5.0).

In the upper part of the active water exchange zone, the groundwater always contains free oxygen (O2). Its content is much the same as in the surfce water, which in Estonia is 8...12 mg/l, as an average (Simm 1975). Due to the presence of free oxygen, the redox potential (Eh) of the water is always positive, because even with a small amount of free oxygen available, Eh cannot be less than +0.17...+0.18V (Posokhov 1975). The higher the concentration of free oxygen in water, the higher the redox potential (maximum +0.7V). Generally, the water under oxidized state does not contain iron, because Fe-oxides and -hydroxides are insoluble in water, which means that Fe comprised in water-bearing rocks does not reach groundwater. However, in several cases Fe2+ (0.7...5.0 mg/l) has been determined in the water of bored wells tapping the Devonian sandstones or Silurian and Ordovician carbonate rocks, and in several cases also in the water of springs flowing out from Devonian sandstones which is indicative of reduced state at that depth. Aqueous environment, where pH = 7.0 and Eh < +0.3 V, is reducing in respect of iron. The latter stays dissolved under such conditions, although there is a small amount of free oxygen available in the water (Carrels & Christ 1965, Shvartsev 1982).

In natural, uncontaminated groundwater which is in oxidized state the content of NO3- is commonly 5...6 mg/l. In the spring water flowing out from Devonian sandstones the concentration of NO3- is only 1...3 mg/l, in the water of excessively damp areas and peatlands it is less than 1 mg/l. This is due to reducing conditions under which, as a result of denitrification, part of the initially dissolved NO3- has been reduced to free oxygen (N2) which volatalizes (Põllumajanduslik... 1994).

In northern Estonia, in the areas with a thick (up to 100 m) Quaternary cover where sediments contain buried organic matter, emissions of burning gas from bored wells have sometimes been recorded (Keri, Prangli and Mohni islands, Viinistu, Püssi). The gas comprises methane, hydrogen, nitrogen, hydrogen sulphide and, to a lesser extent, also helium, argon, oxygen and carbon dioxide and other compounds (Voytov et al. 1982). Frequently, there are gas emissions from the bored wells tapping the bedrock. In this gas the main component is free nitrogen (N2) which accounts for 80...90% of gas volume. Sometimes, hydrogen sulphide is emitted together with the gas. In places, particularly in northern and southwestern Estonia, the water of borings contains helium in a rather high concentration (up to 0.6 ml/l) which originates in the crystalline basement and reaches groundwater through tectonic disturbances (Tibar 1987).


Zone of passive water exchange

A great part of the Estonian water-bearing formation is situated in the passive water exchange zone with reduced state. Due to combined effect of several factors, such as the very slow groundwater flow, connate water occasionally present in rocks, soluble mineral salts, etc., the chemical composition and TDS in groundwater in the passive water exchange zone differ with regions. The HCO3-Mg-Ca water with the TDS ranging from 0.5 to 0.6 g/l is of limited distribution and occurs in the Middle-Lower-Devonian aquifer system along Pärnu - Tartu - Viljandi line. The HCO3-Cl-Na-Mg-Ca water with the TDS in between 0.3...1.5 g/l is more widespread. It occurs in central and southern Estonia in the Middle-Lower Devonian, Silurian-Ordovician and Ordovician-Cambrian aquifer systems. The Cl-HCO3-Na-Ca and Cl-HCO3-Ca-Na water with the content of dissolved mineral salts 0.4...1.0 g/l is stored in the Cambrian-Vendian aquifer system in northern Estonia.

In several places around Tallinn, the content of δ18O in this water is -18...-22‰, but in the area of Loksa, Võsu, Kunda, Toila and Salutaguse it is -11.9...-16‰ (Vaikmäe & Vallner 1989, Savitskaja & Viigand 1994). In all likelihood, in the former case we have a typical glaciogenic palaeowater, in the latter case the glaciogenic water seems to have mixed with infiltration water formed under the conditions of moderate climate.

The Cl-Na, Cl-Na-Ca and Cl-Ca-Na water with the TDS from 2 to 22 g/l is widespread in the passive water exchange zone. This type of water has been established in the Ordovician strata on Hiiumaa Island (Kärdla), Ordovician-Cambrian aquifer system in southern, Cambrian-Vendian aquifer system in northeastern, southwestern and southeastern Estonia, on the islands of Saaremaa and Ruhnu and in the Lower Proterozoic strata in northern and southwestern Estonia.

The SO4-Cl-Ca-Na water with the TDS up to 4.6 g/l, which very rarely occurs under such conditions as prevailing in Estonia, is found at a depth of 260 m at Värska in the Middle-Lower Devonian aquifer system (Fig. 94). Formation of sulphate-rich groundwater is due to the occurrence of gypsum in those layers south of Värska. Since the depth interval 250...800 m is charactericed by reduced state, the water there does not contain O2 or NO3-; SO42- is absent or present in a very low amount (1...5 mg/l). The content of Fe2+ is occasionally very high (up to 10 mg/l) and pH of deep water is usually 7.8...8.5. One reason is the low concentration of free CO2 (often 0...2 mg/l, occasionally up to 6 mg/l). The content of HCO3- is also low, commonly 90...120 mg/l, seldom more.

Total hardness of Estonian groundwater differs in a wide range: it is soft in non-carbonaceous sediments, hard or very hard in carbonate-rich strata. Total hardness of bog water is 1...25 mg/l as CaCO3, the hardness of water stored in sands under the influence of bog water is 20...85 mg/l, in extensive sand areas 65...195 mg/l and in sedimentary rocks 145...360 mg/l. Total hardness of mineral water with the TDS reaching 22 g/l, may be up to 4250 mg/l.



The concentration of microelements in the water stored in oxidized state in the active water exchange zone is generally very low. The low content of some physiologically important microelements, such as fluor and iodine in drinking water, may cause health troubles. The incidents of endemic struma and caries, in some regions more numerous than in others, are associated with the low content of iodine and fluor in drinking water, respectively (Kuik 1961). The groundwater, which occurs under reducing conditions in the passive water exchange zone, is richer in microelements. In some places the content of microelements is even in excess of the optimum value established for drinking water. For instance, in the Silurian-Ordovician aquifer system the content of fluorides (F-) is 5.5...7.2 mg/l in western and southwestern Estonia, 3.2 mg/l in Tartu and 2.4 mg/l at Abja, being well in excess of the standard established for the drinking water in Estonia and exceeding the level permitted by the World Health Organisation, which is 1.5 mg/l (Kuik 1963, Viigand & Vatalin 1992, Guidelines... 1993). This explains the incidences of fluorosis in these areas (Kuik 1961). In Pärnu, the content of fluorides in the Silurian-Ordovician aquifer system is almost optimal - 0.8...1.0 mg/l (Boldõreva et al. 1993), being elsewhere well below it.

In the Ordovician-Cambrian aquifer system, the content of microelements is low in the areas where the TDS is less than 1.0 g/l. South of the Pärnu - Viljandi - Tartu line where the Ordovician-Cambrian aquifer system stores Cl-Na-(Ca) water with the content of dissolved mineral salts up to 14 g/l (Ikla), the content of microelements is higher. Thus the contents of cadmium (Cd2+), lead (Pb2+) and lithium (Li+), in excess of the norms established by the drinking water standard in Estonia, have been registered at Värska (Fig. 94). Since the water derived at Värska is used as a mineral water, and half of the amount produced is dilluted, then it does not pose any threat to human health.

In the groundwater stored in the Cambrian-Vendian aquifer system east of Tallinn, the concentration of iodides (J-) is heightened, reaching 120...280 μg/l (Kuik 1961). This water, if used for drinking, will cover the need for iodine in the population. Generally, the content of all microelements increases in northern Estonia towards the east in this aquifer system. In northeastern Estonia, in some bored wells the content of cadmium (Cd2+), lead (Pb2+) and lithium (Li+) is slightly in excess of the norms. In this aquifer system the concentration of microelements is at its highest at Värska in southeastern Estonia where in Cl-Na-Ca water the TDS ranges from 6.0 to 19.0 mg/l. In this water the content of cadmium (Cd2+), lithium (Li+), manganese (Mn6+) and lead (Pb2+) is in excess of permitted boundary limits. This type of Värska water is not used directly for drinking, but due to its medicinal effect it is used in baths and as curative drinking water. The origin of the microelements in the water of Ordovician-Vendian and Cambrian-Vendian aquifer systems is not yet unambiguously clear. They may partly reach the water from the steel casing of wells, continuously corroded by salt water stored in reduced state, however, part of microelements is evidently of natural origin (Savitskaja & Viigand 1994).

In several regions of Estonia, the heightened concentrations of bromides (Br-) have been detected (Fig. 94, Table 20):

- in the Ordovician-Cambrian aquifer system: 31 mg/l at a depth of 645...658 m at Ikla and 50...54 mg/l at a depth of 707...784 m on Ruhnu Island;

- in the Cambrian-Vendian aquifer system: 13 mg/l at a depth of 540...555 m at Kuressaare, 16...17 mg/l at a depth of 520...535 m and 51...56 mg/l at a depth of 540...600 m at Värska;

- in the water of the crystalline basement 51...61 mg/l at Hirvli and Pudisoo.


Mineral water

The groundwater in which the TDS is 2 g/l or more is rated as mineral water in Estonia. A. Verte was the first to predict the existence of different types of mineral water in Estonia (Photo 33). The first mineral water deposit was discovered at Pärnu in 1959 by the researchers of the Geological Survey of Estonia. For the purposes of structural geological investigations a well, deeper than 500 m, was sunk in Pärnu. It tapped the Lower Proterozoic crystalline bedrock and yielded Cl-Na water with the TDS about 22 g/l (Vingisaar 1978). Mineral water has been found in 16 different sites all over Estonia. At Värska, it occurs in four aquifers, at Kuressaare and Arumetsa in two aquifers (Table 20, Fig. 94).

In Estonia the bottling and marketing of mineral water was started in 1968. Currently, the Cl-Na-Ca water of the Ordovician-Cambrian aquifer system obtained at Värska with the content of TDS ranging from 2.0 to 2.2 g/l is bottled. The mineral water derived from Värska and Kuressaare is used for curative purposes both for drinking and in baths. In earlier years, the mineral water derived at Arumetsa, Häädemeeste, Ikla, Kuressaare and Kärdla was also bottled and sold. During the bottling, the water is often enriched with carbon dioxide (5...7 g/l). The reserves of Estonian mineral waters are estimated at about 6000 m3/d. Total reserves of salty and salt water are very large in Estonia and may amount to hundreds of cubic kilometres.



The temperature fluctuations caused by meteorological factors occur in the upper part of Estonia’s water-bearing formation with an average thickness of some 18 m (Jürima 1984). The maximum thickness of this zone, marking the depth of the so-called neutral layer, reaches 30 metres in the Pandivere Upland. In western Estonia and on islands, the thickness of the zone subjected to annual temperature fluctuations is 10...15 m and in the coastal plain of northern Estonia it is 5...11 m. The temperature of water in springs and up-to-30-m-deep wells in the Pandivere Upland ranges from +4oC to +6oC, elsewhere it is +6...+7oC. The water is coldest in March- April when a great quantity of melt-water percolates into the ground, and warmest in September - October. The temperature in the neutral layer is stable, being prevailingly +7oC, in some places also +6oC. Downwards the temperature rises steadily (Fig. 104). At a depth of 50 m in the area of Silurian and Ordovician carbonate rocks, the temperature is +6.2...+6.3oC, in southwestern Estonia under Devonian sandstones +9.4...+9.5oC and in Lontova clays on Estonia’s north coast +7...+8oC. At a depth of 100 m the average temperature is +7.6oC. In the uplands with intensive groundwater recharge the temperature is lower than the average (+6.5...+7.2oC), whereas in northeastern Estonia and on the Pärnu Lowland it is higher, being +9.2...+9.5oC and +9.8...+10.2oC, respectively. At a depth of 200 m in the carbonate bedrock of central Estonia, the temperature is only +8.0oC which shows that the subsurface groundwater percolates quickly to a greater depth. At the same depth in northeastern Estonia, the temperature is +14.0...+14.5oC and in southwestern Estonia +10.3...+10.6oC. At this depth, the highest temperature (+15...+16oC) has been measured in the crystalline basement of northeastern Estonia (Jürima 1984).

The mean value of geothermal gradient in Estonia is 1.2oC/100 m, and it increases with depth being 1.0oC/100 m in Silurian-Ordovician carbonaceous rocks and Devonian sandstones, 2.0...3.5oC in deeper bedded Cambrian sandstones, and 5.0...6.0oC (average 4.0oC) per 100 m in the underlying Cambrian clays and silts (Lükati-Lontova regional aquitard) (Jürima 1984).


Man-made changes of groundwater quality

L. Savitskaja

Pollution load on groundwater

As a result of extensive economical activity and high vulnerability of the uppermost aquifer, the shallow groundwater is in places heavily polluted and therefore unfit for drinking. The point-pollution sources are different constructions and pipelines in poor condition, such as boilerhouses, fuel storages, storages of chemicals and manure, settling basins, sewerage, leaching beds, gas stations, landfills, burial places of domestic animals, etc. Extensive fuel leakages have occurred on military airfields (Fig. 105) and in railway junctions. The majority of gas stations that were state-owned during the Soviet period contaminate environment with oil products. The asphalt concrete plants using primitive equipment and technology (Tiitso, Riisipere, etc.) are also sources of extensive pollution of groundwater.

In the Ida-Viru County, essential point-pollution sources are the spoil heaps of oil shale mines and ash plateaus of thermal power plants (Fig. 105). The groundwater leaching from oil shale ash is polluted with phenols and compounds of heavy metals, its pH-value is 12 and even more. Extensive surface water and groundwater pollution has been caused by cracking processes accompanying fires in oil shale mines (Vallner & Sepp 1993, Vallner 1994).

Non-point pollution is caused by the misuse of mineral fertilizers on arable lands and the underutilization of slurry from pig farms, but also by the treatment of fields with toxic chemicals. In the Ida - Viru County powerful thermal power plants annually eject into the atmosphere 120,000 tonnes of fly-ash and 80,000 tonnes of aerosol fractions containing harmful elements (S, F, Cl, V, Cr, Ni, Br, Sb, Cd, Pb, As) and radioactive isotopes (Õispuu & Rootamm 1994). The ash and gaseous pollutants transported by wind over large areas percolate together with rainwater into the groundwater. Higher concentrations of these pollutants in the atmosphere promote formation of sulphuric and nitric acids. The impact of acid rains is mitigated by fly-ash from thermal power plants which may spread to a distance of 100...150 km (Frey et al. 1987).

In 1994, about 14,000,000 tonnes of solid waste was generated in Estonia; of that amount 46 % by the oil-shale-based energy production, 40 % by oil shale mining, 8 % by the chemical industry and only 6 % in other spheres (Keskkond... 1995). Of 1,962,000,000 m3 of waste water produced, 70 % was the cooling water heated through use in an industrial process which raised the temperature of both surface and groundwater. The amount of waste water needing purification was 380,000,000 m3. Approximately 1% of waste water (1,800,000 m3) was discharged directly into soil and groundwater. Compared to 1989-90, the total industrial capacity has decreased and the technology has improved. As a result, the amount of solid waste and waste water has decreased by about 20 and 40 %, respectively.

In 1994, a total of 41,000 tonnes of mineral fertilizers (as N) and 1,100,000 tonnes of manure were used in Estonia, which is about 6.5 times less than during 1988-89. The average pollution load of solid waste was of 320 t/km2, and that of untreated waste water 8,700 m3/km2. The amount of mineral fertilizers (as N) was 0,04 t/ha and that of manure 1.1 t/ha. The pollution load was highest in the Ida - Viru County due to oil shale enrichment waste and ash of thermal power plants. In 1994, the average pollution load of solid waste in this region was 3,840 t/km2 (1.6 times less than in 1990), while the load of waste water in need of purification was 69,000 m3/km2.

In northeastern Estonia, the Ordovician carbonate rocks covered with a thin Quaternary mantle are intensively polluted with shale oil and phenols. In the undermined area (about 200 km2) the lowering of groundwater level has resulted in oxidation of pyrite in the aeration zone, due to which the content of sulphates in groundwater has increased up to 650 mg/l (under natural conditions it is less than 20 mg/l).

Groundwater is polluted with oil products at Tapa (Fig. 105) and in its surroundings. Since 1966, several big accidents of fuel tanks and constant leakage of fuel pipelines took place on the former military objects of this region. Particularly intensive pollution with oil products has been recorded on the former military airfield, where the stopcocks, valves and pipelines of fuel tanks were leaking continuously. Near the railway oil receiving centre, an oil lake had formed on the surface. Due to carelessness or leakage from boiler-houses and fuel tanks, groundwater is polluted with oil products also in Tallinn, Tartu, Rakvere, Kohila, Rapla, Tamsalu, Aruküla and several other places. In these regions, the water supply is more or less disturbed.

In the area of the former galvanic departments and in the surroundings of landfills, groundwater contains heavy metals. Due to the diffusion of filtrates from Tuula landfill near Keila, the concentrations of hazardous compounds in the neighbouring wells are the following: Pb2+ – up to 0.13 mg/l, Mn2+ – up to 1.4 mg/l, Cd2+ – up to 0.0035 mg/l (Tennokesse et al. 1992); the maximum permissible concentrations of these elements in drinking water are 0.01, 0.1 and 0.003 mg/l, respectively (Eesti standard...1995).

In the regions of agricultural activity, groundwater is contaminated mainly by nitrogen compounds (Fig. 106), but the concentration of chlorides and sulphates has also increased to 40...60 mg/l (under natural conditions 20 mg/l), as an average. In the surroundings of Viiratsi piggery near Viljandi, the concentrations of ions are the following: NH4+ – up to 80 mg/l, Fe2+3 – up to 128 mg/l, Cl- – up to 262 mg/l and SO42- – up to 330 mg/l (Boldõreva et al. 1992). The high content of ammonium and iron, as well as the lack of nitrates indicate reducing conditions. In Estonia, the maximum permissible content of NO32- in drinking water is 45 mg/l (Eesti standard...1995). In 1990, the groundwater did not meet the requirements established for drinking water in 40...70 % of the total number of shallow wells (depth to 15 m) in southern, 20...40 % in northern, 30...60 % in central Estonia and in 10 % of wells on the islands of the West-Estonian Archipelago (Põllumajanduslik... 1994). In the wells with a depth of 30...100 m, the content of nitrogen compounds was remarkably lower, especially in southern Estonia (Tennokesse et al. 1992).


Groundwater protection

Due to a thin aeration zone (mainly 1.5...3 m) the unconfined groundwater is generally weakly protected against surface pollution. The problems are acute in northern and central Estonia where the Quaternary cover is less than 2 m thick or practically lacking (alvars) and the fissured and highly ca-vernous limestones crop out on the surface.

On the basis of numerous experimental data and calculated infiltration velocity of waste water, certain criteria for the assessment of the degree of natural protection of shallow groundwater from agricultural and municipal pollution have been worked out (Savitskaja 1987).

Groundwater is considered unprotected on alvars (the thickness of the Quaternary cover is less than 0.5 m) and in the areas where aquifers are covered with a up-to-2-m-thick layer of loamy sand (conductivity value K=0.1...0.5 m/d), or up-to-20-m-thick layer of sand or gravel (K=1..5 m/d). Groundwater is weakly protected when the thickness of the confi-ning layer of loamy sand ranges from 2...10 m, or that of the clay layer (K=0.0001...0.005 m/d) is less than 2 m. Groundwater is regarded as moderately protected when aquifers are overlain by a 10...20-m-thick layer of loamy sand, or with a 2...5-m-thick layer of clay, and protected when the thickness of the confining layer of loamy sand exceeds 20 m, or that of clay layer is over 5 m. In case of intensive water consumption, these criteria must be corrected, because under such conditions the transport velocity of pollutants in soil may increase noticeably.

The map showing the degree of shallow groundwater protection demonstrates that groundwater is weakly protected or unprotected in ca. 40 % of Estonia’s area. A comparison of the maps of groundwater protection degree and nitrate concentrations (Fig. 106) has shown that the intensity of distribution of nitrates depends closely on groundwater protection degree.

Since 1991, the impact of agriculture on groundwater state has noticeably decreased due to manifold reduction of the use of fertilizers and decrease in the number of domestic animals. Long-time observations of 19 springs on the slope of the Pandivere Upland show that the content of nitrate in spring water has decreased which also indicates to general lowering of pollution load in this region (Savitski et al. 1996).

In 1994, a decree of the Minister of the Environment addressing reduction of pollution load, enacted restrictions to the use of fertilizers (Table 21) and the number of domestic animals by regions.

In Estonia’s environmental strategy, priority has been given to the problems of groundwater protection, the most important being elimination of sources of groundwater pollution and regulation of groundwater use. The Water Code of Estonia and the Law of Sustainable Development enact general regulations for economically effective and environmentally sound use of water resources.


Groundwater extraction and safe yield

L. Vallner & L. Savitskaja

Groundwater use

Until 1944, the groundwater was mostly used for domestic needs only. According to studies of J. Kark, there were only a few hundred bored wells, deeper than 50 m in Estonia which were used by public waterworks in central parts of towns and predominantly for the purposes of food industry. Pumping of water from a few production wells did not induce any cone of depression worth of mentioning. There was only one public water intake tapping the glaciofluvial water-bearing gravel and sand of the Meltsiveski buried valley which was put into operation in Tartu with the pumping rate of 12,000 m3/d (Orviku 1946). The total number of wells was approximately 250,000. Predominantly these were domestic wells with a depth of up to 8 metres.

Since 1945, unbalanced industrialization, urbanization and militarization of Estonia was carried out by Soviet authorities. The need for a centralized water supply increased rapidly and the number of deep bored wells augmented from a few hundred in 1950 to 3300 in 1964 (Arkhangel´sky 1966). Since 1964, 200...300 new wells were bored every year, but at the same time there was a need to liquidate a lot of amortized wells. In 1991, the number of wells deeper than 60 m reached 8400 in Estonia.

The total groundwater extraction grew from about 30,000 m3/d in 1950 to 470,000 m3/d in 1991 (Boldõreva et al. 1993). During the first two decades, the rate of water extraction increased steadily, thereafter began to decelerate (Fig. 108). During 1964-94, the contribution of different aquifer systems in the total groundwater supply was following: Quaternary, Middle Devonian and Middle Devonian-Silurian aquifer system - 6...16% each, Silurian-Ordovician aquifer system - 22...25%, Ordovician-Cambrian aquifer system - 9...12% and Cambrian-Vendian aquifer system - 34...39% (Savitski et al. 1995). Consequently, the proportion of the aquifer systems exploitation did not change significantly during 30 years.

Until 1991, approximately 60% of total groundwater supply was consumed in towns, of that nearly half in Tallinn and in the Kohtla-Järve region; 15% in Tartu, 5% in Pärnu and the remaining 30% in other towns. In rural areas, the deep bored wells were predominantly used to supply the settlements and big livestock farms.

In 1995, about 290,000 m3 of drinking water was extracted from bored wells, which is 60% less than in 1990 (Kivit et al. 1991 & Savitski et al. 1996). Approximately 80% of this water was spent for supplying of towns and industrial settlements, the remaining 20% was used in agricultural regions (Fig. 109). The decrease of groundwater extraction has been caused by the decline of industrial and agricultural production and by the more sustainable use of groundwater. As an average, 110,000 m3/d was extracted from the Cambrian-Vendian aquifer system, 59,000 m3/d from the Silurian-Ordovician aquifer system, and 17,000...47,000 m3/d from each of the remaining aquifer systems (Table 22).

Currently, groundwater forms 70% of consumed drinking water in Estonia. Only in Tallinn and Narva the consumption of purified surface water exceeds that of groundwater, while elsewhere in Estonia groundwater is the only source of public water supply. The major amount of groundwater is extracted from about 20,000 bored wells from which only a half are deeper than 20 m. The depth of bored wells, used in public water supply, remains predominantly in the interval of 80...150 m. However, in central Estonia some wells have a depth of up to 450 m.


Groundwater safe yield

In Estonia, the groundwater safe yield is considered to be an amount of groundwater which could be withdrawn during a calculation period without producing an unpermitted deterioration in the quality of water pumped. Since 1963, the groundwater safe yield has been determined on the basis of detailed hydrogeological investigations to get a clear conception about the perspectives of public water supply. To calculate the safe yield Qs the modification of Theis’ equation was mostly used (Bochever 1968):

where s is a given suitable and really feasible drawdown in a given point A of wellfield; n is the total number of wells under consideration; Qi is the given constant discharge of a well i; T is the transmissibility of the aquifer, and Wi (u) is the so-called well function with the argument u = ri2S/4Tt where ri is the radial distance of the well i from the point A; S is the storage coefficient, and t is the time since beginning of pumping. The safe yield was calculated as a sum Qs = Q1 + Q2 + ... + Qn by the trial and error method.

In calculating the safe yield, a series of simplifying assumptions, such as infinite areal extent of an aquifer, its uniform thickness, homogeneity, isotropy, etc., were posed. In spite of this, in most cases the calculated prognoses fitted quite well to observations, made later in operation of water intake. The necessary hydrogeological parametres for calculating the safe yield were taken from the materials of hydrogeological mapping of Estonia in scale 1:200 000 and partially 1:50 000 but also from the results of previous hydrogeological exploration, including pumping tests and groundwater quality analyses.

Valuable information was gained from the observation network of regular hydrogeological monitoring, which was established in 1946 on some water intakes in northern Estonia and in the region of oil shale mines. Later on the observation network of the state geological survey was steadily widened and developed: in 1990, it consisted of 846 observation wells (Kivit et al. 1991). The aquifer systems were monitored all over Estonia. Besides, 60...120 observations wells were bored on more essential water intakes and in mining districts (Tallinn, Kohtla-Järve, Vasavere) Groundwater level was measured at observation points once every three days and the chemical analysis of water was made once every three months to determine the content of 8...12 main components. The sanitary state of groundwater was steadily controlled by sanitary survey and the amount of consumed groundwater was registered by the owner of the well.

Until 1995, the safe yield of 131 well fields was estimated (Savitski et al. 1996) at about 550,000 m3/d which is nearly three times as high as the real extracted amount (Table 22, Fig. 109). About 40% of the safe yield associates with the Cambrian-Vendian aquifer system, while the Quaternary and Ordovician-Cambrian aquifer systems provide only 8 and 5%, respectively. The safe yield of the Quaternary and Silurian-Ordovician aquifer systems is rapidly recurrenting, but these systems are vulnerable to pollution and require extensive zones of sanitary protection. The safe yield of the Middle Devonian-Silurian, Ordovician-Cambrian and Cambrian-Vendian aquifer systems will be restored slowly, or it will never be restored (glaciogeneous water in the Vendian strata). To ensure sustainable use of the safe yield, the slowly recurrenting groundwater must be used only for drinking water supply.

The regional safe yield for main aquifer systems and bigger water intakes was simulated by means of an analogue computer in 1976. The total, so-called presumptive safe yield, determined by this method with relatively smaller accuracy, is 1,530,000 m3/d of which 42% falls to the Silurian-Ordovician aquifer system and 15% to the Cambrian-Vendian aquifer system (Table 22).

In 1994, about 70% of the total pumpage was got by intakes having an estimated safe yield and enfolding 925 production wells. The remaining portion of water was derived by 7,000 wells scattered over Estonia. Despite some deviations, the used safe yield generally meets the drinking water requirements (Eesti standard... 1995). The increased contents of iron and manganese have been found in the water of Devonian strata in southern Estonia, and the increased content of fluorine in the water stored in the Silurian strata in western Estonia. The water of some wells tapping the Cambrian-Vendian aquifer system contains iron, manganese and chlorine in excess of the sanitary norms (Kohtla-Järve, Jõhvi, Ahtme, etc.). To reduce the concentration of hazardous components, the groundwater must be treated before use.


Consequences of the intensive groundwater use

The most serious consequences of intensive groundwater use include the formation of regional depressions of potentiometric level which has caused cardinal changes in the direction and velocity of filtration flows in the lower portion of the water-bearing formation (Figs. 102, 103). As a result of heavy pumpage, the groundwater inflow into the strata increases. If in natural conditions the inflow into the Cambrian-Vendian aquifer system was less than 30,000 m3/d, then in 1990 it was 163,000 m3/d or even more. Drawdown contours show that half of the increased inflow is coming from the seaside. Consequently, the other half is supplied by downward, upward and southward lateral flows. The downward flow provides the Cambrian-Vendian aquifer system mostly with fresh water, but the rising flow may be connected with upconing of brackish water from the crystalline basement. Lateral flows conduce the transport of connate brackish water from the deeper portion of an aquifer or sea-water intrusion to groundwater intakes.

The sea-water intrusion into water intakes can take place first in Tallinn where a lot of production wells tapping the Cambrian-Vendian aquifer system are situated close to the sea. However, in spite of several threatening prognoses the intrusion of sea-water into water intakes in Tallinn has not been observed yet. The increase of TDS in the water of some wells from 0.5 up to 1.3 g/l has been observed in the northern part of Tallinn, but the data of isotope analyses confirm that it is caused by the inflow of brackish water from the lower part of aquifer system or from the crystalline basement (Savitski et al. 1995). The potential intrusion of sea-water into coastal bored wells, tapping the Silurian-Ordovician aquifer system, is a serious problem also in Pärnu, Haapsalu, Kuressaare and some other places.

The potentiometric surface of the Ordovician-Cambrian aquifer system has dropped at least 7 metres in the central part of the Pandivere Upland due to regional impact of water consumption. Deep local depressions have also formed in Tartu, Rakvere and several other places (Figs. 102, 110). Because of the significant head withdrawal, pollutants can easily intrude into the Ordovician-Cambrian aquifer system from surface in Tallinn, Tapa and Kohtla-Järve.

To intensify the water supply of Kohtla-Järve urban area, a groundwater intake with the planned capacity of 25,000 m3/d using the glaciofluvial aquifer of the Vasavere buried valley was put into operation in 1971. The intake was within the Kurtna Landscape Reserve featured by wooded hillocks and ridges with a lot of picturesque lakes, many of them with unique hydrobiological biocenoses. Unfortunately, already at the pumping rate of 10,000 m3/d the water table of several lakes lowered below the acceptable minimum level and the unique biocenoses were irreversibly damaged (Mäemets 1987).

In Tartu, many historical buildings have been founded on timber piles which were initially submerged by groundwater. After 1960, the pumping from the Toome-Meltsiveski buried valley and bedrock aquifers was significantly increased which caused a drop of the groundwater level by 2.5...3.5 m in the old part of Tartu. The upper portion of timber piles remaining in the aeration zone were intensively attacked by the wood borer Cossonus parallelepipedus and began to decay quickly (Oll 1967). As a result, the buildings started to crack. Decaying of architectural memorials could have been avoided by correct groundwater pumping; now tens of millions of US dollars are needed for their reparation.

An uneven land subsidence occurs in the area of buried valleys permeating the territory of Tallinn (Arbeiter et al. 1982, Vallner & Lutsar 1966). The total subsidence amounted to 0.6...0.8 m and its maximum annual rate reached 36 mm in the vicinity of the trading port in 1964. Deformation of the ground caused the bench-marks to shift, spoiled the designed inclination of sewage collectors, created cracks in the walls of buildings. The subsidence was caused by the compaction of marine and glaciolacustrine clays confining the Cambrian-Vendian aquifer system in buried valleys. Due to intensive pumping from this aquifer system, the potentiometric surface was lowered by 20...25 m and the compacting intergranular pressure in the overlying highly compressible Quaternary aquitard increased. However, in 1972, the potentiometric surface of the Cambrian-Vendian aquifer system dropped below the bottom of the confining Quaternary aquitard and since then land subsidence has gradually expired. Calculations have shown that the maximum subsidence does not exceed one metre in the area of the old trading port of Tallinn (Vallner 1989). An analogous subsidence, only of smaller extent, occurred in Pärnu, where the compaction of Quaternary clayey sediments has been caused by the essential lowering of potentiometric surface in the underlying Silurian-Ordovician aquifer system due to intensive pumping.


Dewatering of mines and drainage of arable land

As a result of dewatering of oil-shale mines, the groundwater level of Quaternary and Ordovician aquifers has lowered by 15...65 metres and several local cones of depression, influencing each other, have formed over an area of 600 km2 between Purtse and Narva rivers. In 1984-94, depending on precipitation, 600,000 to 900,000 m3 of water was pumped out from the mines every day. The yearly amount of the water pumped out from the mines was 1.4...1.9 times as high as the total annual groundwater supply during the last decade. However, the proportion of groundwater in total mine water does not exceed 20...50%, the remainder is formed of surface water intruding directly into goafs through cracks, ventilation holes, unsealed wells, etc., or precipitation which accumulates in open-pit quarries. During a wet period, 3 to 10 times as much water is pumped out from mines as during droughts (Gazizov 1971, Norvatov 1987, Savitski 1980). More than 60% of the whole amount of annual intrusion intrudes into the mines, located in the northern and central part of the oil shale region. It takes place during 4...5 months of water-abundant periods in spring and autumn. Drawdowns of the groundwater head increase southward in accordance with the dipping depth of the commercial oil shale layer. Depressions of the head are extending to a distance of 4...8 km outward from the mines.

The shallow wells, situated in the area of mine influence, often dry up. Dug into Quaternary cover or hewed into carbonate bedrock, they usually do not reach deeper than half a metre below the lowest groundwater level under natural conditions (Vallner 1996b). A groundwater drawdown of only 0.5 m caused by mine dewatering unfit these wells for water supply in dry period. In the areas, suffering from the greater drawdown, the dug wells are out of use during a longer period or they dry up completely. Besides of intensive pumping from the Vasavere intake, the mine dewatering is another reason why the groundwater table has lowered in the Kurtna Landscape Reserve (Kurtna... 1996). Owing to the depressurization of Ordovician carbonate aquifers in the area of the Jõhvi Upland, the recharge of the underlying Ordovician-Cambrian aquifer system has decreased causing the lowering of potentiometric surface by up to 8 metres (Vallner 1996b).

Since 1950, about 11, 000 km2 of arable land and woodland have been drained by ditching and covered trenches. Altogether about 75% of arable agricultural land and about 30% of woodland has been drained. In the drained area, the groundwater level has dropped a metre, but the confined aquifers of the carbonate bedrock were also tapped in places which increased the local groundwater runoff. Therefore, an apprehension of the groundwater overdraft arose. However, the more detailed calculations showed that due to the lowering of the groundwater level, evaporation from its surface decreased inducing an increase of the total infiltration in the drained areas by 15 mm (Vallner & Metsur 1988). Thus, as a result of the agricultural land drainage and mine dewatering, the groundwater recharge has not diminished, but, on the contrary, augmented.




Deep structure

V. Puura & R. Vaher


In the light of geophysical studies of the deep crustal structure conducted in the Fennoscandian (Baltic) Shield during the last decades and the much sparser data available on the adjacent Russian Platform area, the position of Estonia in the centre of the thickest crustal domain within the northwestern part of the East-European Craton becomes evident (Fig. 111). Main features of the regional crustal structure were formed during the Svecofennian orogeny (1.8 to 1.9 Ga). In the present morphostructure, the Baltic Sea Drainage Basin conforms well with the Svecofennian Crustal Domain (Puura & Flodén 1997). The Baltic Sea Depression (sensu stricto) occupies the central position in both the drainage basin and the crustal domain. In terms of the crust’s age, Estonia locates within a unitary super- and morphostructure. The apparent differentiation in the crustal thickness (Moho depth) in Estonia and adjacent areas (Fig. 111) can be explained by both Svecofennian orogenic processes and subse­quent crustal changes induced by tectonic and magmatic events.

Compared with Fennoscandia, less information is available on the deep structure of Estonia where the first data were obtained only a few decades ago. The data of the magnetotelluric survey (Fig. 112), carried out in 1970‑72 (Andra et al. 1974), gave 75 km for the mean depth of the highly conducting region in the mantle. At three stations, it was also possible to estimate the depth (140 to 180 km) of another highly conducting region.

Based on the studies, carried out on the FENNOLORA seismic refraction profile in 1979 and on the Sovetsk - Kohtla-Järve profile (Fig. 111) in 1983-86, Luosto (1991) demonstrated an almost east-west oriented Moho depression extending from the eastern coast of Sweden over the Baltic Proper to the Baltic States. On these profiles above the assumed trough of the depression, the crust is 55 to 64 km thick. Later studies showed that on the Baltic Sea profile between the islands of Gotland and Saaremaa the crust is thinner with its thickness ranging from 41 to 44 km (Ostrovsky et al. 1994). Hence, the Moho depression is located mostly on mainland (Fig. 111).

According to the calculations by Sadov and Penzina (Ankudinov et al. 1994), based on the Sovetsk - Kohtla-Järve profile, the crust is up to 64 km thick in the middle of the depression (Fig. 111) southeast of Riga, Latvia, and 46 to 51 m thick in the Estonian part of the profile.

In 1976, a seismic survey was carried out close to the epicentre of the Osmussaar earthquake of October 25, 1976. As a consequence, Bulin (1978) suggested the preliminary depth of about 42 to 47 km for the Moho depression in northwestern Estonia (Fig. 113, left). The reinterpretation (Bulin et al. 1980) gave smaller depth values of 36 to 44 km for this area (Fig. 113, right) suggesting a local Moho uplift (Fig. 111). Considering the depth of the Moho surface in southern Finland (Luosto 1991), it seems that in Estonia the thickness of the crust increases slowly from 44 to 51 km towards the south.

In the Moho depth map (Fig. 111), the thickest crustal area in southeastern Estonia and its gradient zone are referred to the Svecofennian orogenic features, while the rised Moho surface in the West-Estonian Archipelago and in northern Estonia are contiguous to crustal thinning areas developed during the rapakivi-age continental rifting and basaltic underplating (Puura & Flodén 1996).

Like in the whole Russian Platform, the crust in Estonia is divided into two structural stages: (1) the strongly disturbed and metamorphosed Precambrian basement, and (2) an unconformably overlying thin (less than 800 m) cover of little deformed and gently tilted sedimentary strata. The crust is broken into blocks by deep-seated faults (Fig. 111).


Basement features

V. Puura


As shown above, Estonia belongs to the Svecofennian Crustal Domain, the structural patterns of which have been studied in particular detail in the Fennoscandian Shield. In Estonia, the data for crystalline basement studies has been basically obtained by geophysical (gravimetry, magnetome­try, electrometry etc.) survey and deep drilling (Puura et al. 1983, Koistinen 1994, 1996). As a whole, the main struc­tural elements of the buried basement of Estonia are quite well established: 1) areas of folded metamorphic rocks, 2) plutonic rocks, 3) regional fault zones. The folded structure of the orogenic rocks can be characterised only in general lines, basing on observations of geophysical anomaly field patterns (Fig. 6) and drill core samples. Localities of anorogenic plutonic rocks with areas large enough for geophysical determination have been well mapped by deep drilling (Fig. 114). Fault zones of different age and different depth of formation (brittle or ductile deformation) have been studied by detailed geophysical profiling and veryfied by drilling.

Traditionally, the folded basement of Estonia has been divided into several structural zones (Fig. 114) differing in the rock composition and level of regional metamorphism (Puura et al. 1983). In the northern ‑ northeastern part, the basement consists of metamorphic belts of amphibolite facies (Tallinn and Alutaguse structural zones) with local high-grade blocks (Jõhvi, Tapa, etc.). In the south and south-west, a large area of prevailing mafic to intermediate granulites with enderbites-charnockites has been distinguished, while mafic and intermediate rocks of metabolite facies occur in the west and northwest. The boundaries of the zones and blocks are usually marked by faults occurring as specific gradient zones in geophysical anomaly fields. Late-, post- and anorogenic plutonics accompanied by nonlinear anomalies show their cutting position in the linear anomaly fields of metamorphic rocks (Fig. 114).

According to the recent reinterpretation of the tectonic structure of the Fennoscandian Shield (BABEL... 1990, 1993, Korja 1993) and adjacent areas, considering the new data on the deep crustal processes in the region (Koistinen 1996, Puura & Flodén 1996), the Svecofennian structure of the basement is composed of a collage of metasedimentary basins squeezed between crustal blocks of the island arc origin (Korja 1993). The roots of the Svecofennian mountains have preserved along crustal bulges exceeding 50 km in thickness. In this interpretation, the Tallinn Structural Zone of the North-Estonian basement (Puura et al. 1983) belongs to a prevailingly volcanic block of presumably volcanic arc origin. Its continuation might be observed in southern Finland (Koistinen 1994). The Alutaguse Zone in northeastern Estonia is probably a fragment of a large sedimentary basin presently exceeding the St. Petersburg and Novgorod areas in Russia.

Already in the early stages of investigation, it was stated that the structure of the basement in southern and southwestern Estonia differs from that in northern and northwestern Estonia (Fotiadi 1958, Puura et al. 1983). The South-Estonian granulite area belongs to the Belarussian‑Baltic assemblage of beltiform granulite and amphibolite facies tectonic sheets (Grigelis & Puura 1980, Puura et al. 1984). Within the Belarussian-Baltic granulite subdomain, the origin and age of rock protoliths are probably the same as within the Svecofennian of Fennoscandia (Puura & Huhma 1993, Gorbatschev & Bogdanova 1993). The regional high-grade granulite metamorphism coupled with the regional high rock density and magnetization in the area called the Baltic Geophysical High (Fotiadi 1958, Puura & Huhma 1993) conceal the geological and potential field patterns characteristic for the proper Svecofennian in that area. Thus, the high-grade, prevailingly metavolcanic rocks in southern Estonia probably belong to the late Svecofennian stacked tectonic sheets.

Within the Svecofennian Domain, a number of orogenic fault systems can be distinguished. The undulating shear zones, shown on the map of the Precambrian basement (Koistinen 1994) in southern Finland, follow the direction of the Tampere subduction zone (Korja 1993) and are considered as being of early orogenic age. The curved but somewhat more straight fault zones, which coincide with the boundaries of the granulite belts of the Belarussian-Baltic Province, are probably of late orogenic origin (Grigelis & Puura 1980, Koistinen 1994). The general fault pattern of the Svecofennian Domain, including the Estonian basement, conforms with the deep crustal features, such as changes of Moho depth (Fig. 111) or the position of conductivity zones, or both (Korja 1993).

Like in the shield area, the structural zones of the basement, well observable in gravity and magnetic anomaly patterns, differ in assemblages of metasedimentary or metavolcanic rocks. The most clearly observable structural patterns of the orogenic crystalline rocks in drill cores are their prevailingly steep tilting of schistosity or, randomly, original bedding, and very variable stage of mig­matization (or char-nockitization). They occur everywhere in metasedimentary and metavolcanic rocks, but are of less intensity in early orogenic plutonic rocks. Their texture and composition are dealt with in more detail in Chapter III. The Paldiski-Pskov Zone occurs as a major marginal block assemblage of the Belarussian-Baltic granulite province (Gorbatschev & Bogdanova 1993).

The rapakivi and related magmatic structures formed 1.65‑1.54 Ga, are about 200‑300 Ma younger than the Svecofennian orogenic rocks. The rapakivi‑anorthosite magmatism was related to and coupled with basalt underplating of the thick Svecofennian continental crust resulting in thinning of the crust in the area of the Vyborg Pluton (Elo & Korja 1993) and Åland Pluton (BABEL... 1993), but also in the northern part of the Baltic proper (Ostrovsky et al. 1994) and in northwestern Estonia (Fig. 114). Near the Åland Pluton, crustal signatures typical for continental rifts have been reported (BABEL... 1993). Probably, the crustal structure in the northern and western parts of Estonia (Fig. 111) carries traces of the above rifting, basaltic underplating and Subjotnian Rapakivi magmatism.

The Fennoscandian Rapakivi Province consists of four subprovinces (Koistinen 1994, Puura & Floden 1996). In all subprovinces of Subjotnian rapakivi, major polyphase bimodal (granite-anorthosite) volcano-plutonic complexes take a dominant position evident in deep seismic sounding and gravity field signatures (Elo & Korja 1993). The rapakivi-age diabase dike swarms and minor granite stocks occur around the main plutons designating the areas of different subprovinces. In Estonia, the fault-controlled Sigula offite gabbro body probably belongs to the Subjotnian diabase dike series.

The oldest, 1.62‑1.67 Ga, Vyborg Subprovince has an eastern central position in the Province. The southern satellite body of the Vyborg Pluton in the central eastern part of the Gulf of Finland, and 7 rapakivi stocks in Estonia belong to the Vyborg Subprovince (Fig. 114, Table 4)). Individual plutonic bodies in Estonia measure some 300-1000 km2 in diametre and have distinct specific geophysical signatures.

The 1.54‑1.58 Ga Riga-Åland Subprovince in the central southwestern part of the province is represented in Estonia (Gulf of Riga, Ruhnu and western Saaremaa, central eastern part of the Baltic Proper) by the norheastern and northern wing of the Riga Pluton, largest in the province. Near the northern border of the Riga Pluton, the rapakivi‑related local Undva volcanic sheet has been penetrated by a drill-hole.

During and after the rapakivi magmatism, faulting of the crust was of great importace. Positioning of plutons and minor stocks, as well as dike swarms indicates the prevailing northwest-southeast and northeast-southwest systems of faults. Major Svecofennian deep faults became reactivated as it was on the Åland - Märjamaa direction. However, the formation of new fault lines was a most frequent feature of rapakivi-time deformation. The Central-Estonian (Saaremaa - Mustvee) east-west striking fault zone probably belongs to this group (Puura 1979). Tectonic block and fault movements were expressed on the earths surface as structural displacements of hundreds and, possibly, thousands of metres.

In areas adjacent to Estonia, in the Late Precambrian time span of 1.5-0.6 Ga at least five tectonic events, some of them coupled with mafic dike magmatism have been documented. There is no information on their occurrence in Estonia, as yet.

As a whole, a pattern of different‑age fault systems, intersecting each other, and partly reactivated fault systems is a characteristic feature of the basement structure (Pobul & Sildvee 1975, Puura 1979). However, the majority of fault zones observed in geophysical fields are still difficult to date.

Much more stable geodynamic environments were established in the Svecofennian Domain after the Subjotnian rapakivi igneous activity. However, the results of the first proper planation occurred only in the beginning of the Jotnian sedimentation ca 1.4 Ga, which became the first evidence of a stable craton. However, the final peneplanation of the region was succeeded just before the Late Vendian inundation and sedimenta­tion.

The late pre-Cambrian break in sedimentation coinciding with the development of the pre‑Vendian peneplain, resulted in the formation of the major angular disconformity in the geological sequence of Estonia which usually serves as a basis in structural studies of the sedimentary cover. Under the sedimentary cover, the upper part of the basement is weathered. In drill cores, clayey minerals as evidences of weathering have been determined in crystalline rocks at a depth 1-100 m below the basement surface.


Cover structure

V. Puura & R. Vaher


The sedimentary bedrock of Estonia is divided into three tectonic stages, named mostly by geologists of the Baltic States (Suveizdis 1979): Baikalian (Vendian and Early Cambrian, Lontova Stage incl.), Caledonian (post-Lontova Cambrian, Ordovician, Silurian and Early Cambrian, Tilžė Stage incl.) and Hercynian (post-Tilžė Devonian). The tectonic stages are separated with regional unconformities. The solid bedrock is covered with a blanket of unconsolidated Quaternary deposits. The tectonic style of the sedimentary bedrock is to a certain degree laterally variable.

The most prominent structure of cratonic type is the Estonian Homocline (Puura 1974) which extends from the Gulf of Finland to northern Latvia (Fig.115). The sedimentary bedrock strata have a very gentle (6 to 18') regional southward dip here (Puura & Mardla 1972). Local variations in dip are well traceable in the meridional and, particularly, in the latitudinal section (Fig. 115). In southwestern Estonia, there is no distinct boundary between the Estonian Homocline and the Baltic Syneclise. Provisionally, it may be the contour line of -550 m on top of the basement.

In the southeast, the Estonian Homocline borders on the Võru Saddle (Vaher 1972) which is anticlinal in E‑W section and synclinal in the perpendicular section. In the middle part of the saddle the basement lies at the level of -500 m, descending eastwards towards the Moscow Syneclise, and westwards towards the Baltic Syneclise. It rises northwards towards the Estonian Homocline, and southwards towards the Valmiera‑Lokno Uplift. The buried Võru Saddle is clearly visible in the Baikalian and Caledonian rocks, but the unconformably overlying Hercynian strata have a fairly regular amount of dip in the general southern direction. This overlying structure is called the Estonian‑Latvian Homocline (Suveizdis et al. 1977, Brangulis et al. 1984).

The Valmiera‑Lokno Uplift in the Estonian‑Latvian border zone is 20 to 30 km wide and 200 km long. It consists of four minor uplifts (Fig. 116): Valmiera, Smiltene (both in Latvia), Mõniste (in Estonia), and Haanja‑Lokno (partly in Estonia, partly in Russia). On the crest of the highest, Mõniste Anticline, the basement lies at the level of -230 m, descending northward to -500 m, and southward to -1,000 m. There are deep-seated faults in the northern and southern flanks of the composite uplift (Fig. 115). The distribution of Vendian and Lower Palaeozoic rocks in southeastern Estonia shows that this structure was formed principally in the Late Silurian and has a long history (Vaher et al. 1980a). In various parts of the major uplift, the basement is covered by sedimentary rocks of different age: Vendian on the Haanja‑Lokno Uplift, Cambrian on the Valmiera Uplift, and even Devonian on the crest of the Mõniste Uplift. As the composite uplift began and de­veloped in the Late Silurian and Early Devonian, from the crest at Mõniste over 200 m of Lower Paleozoic rocks were eroded. Figure 116 shows the lateral variety in the extent of the erosion at that time. A further uplift of the structure in the Middle Devonian was weaker. On the crest of the Haanja‑Lokno Uplift (Fig. 117), the contact between the Middle and Upper Devonian lies at a level of 130 m, descending northward to 90 m and southward to -50 m. The anticline in the Upper Devonian strata is a result of post‑Devonian uplift. The axis of the recent uplift, as established by repeated levelling data, is shifted to the north, and found straight above the northern slope of the old Mõniste Anticline (Fig. 118).

The Estonian Homocline, the largest structure in Estonia, is complicated by many minor structural features of various kind described in detail in northeastern Estonia (Vaher et al. 1962, Puura 1986, 1987). The oldest of them are placanticlines ‑ brachyanticlines or domes of sedimentary strata formed over pre‑Vendian monadnocks of the basement. Most of the linear structures of various trend are probably of Late Silurian - Early Devonian age, but some of them possibly formed already in Vendian and Cambrian‑Ordovician transition times or, otherwise, after Devonian sedimentation. The diapir-type near-surface structures in northeasternmost Estonia are probably of Pleistocene age. Meteorite impact structures may have any age. Small isometric depressions in the oil shale basin are undated yet.

Dome-like plain-type folds (placanticlines, Puura & Kala 1978), some 1 to 6 km in diameter and 30 to 130 m in height on top of the basement (Table 23), have been found so far only in northern Estonia (Fig. 115). The net result of the folding is a relative local elevation without a corresponding depression. Thus, there are only brachyanticlines and domes rising above the regional dip. The folds become more pronounced with depth, the lowermost strata wedge out on the flanks of the folds, and there is a thinning of the strata above the crests of the folds (Table 24).

The best known, Uljaste Placanticline was discovered during the exploration for oil shale in 1930 (Reinwald 1935), and studied by boreholes finished in the basement in 1960‑62 (Vaher et al. 1962, 1964) and 1975‑77 (Puura & Kala 1978). On the flanks of the structure (Fig. 119), not only all Vendian Strata but also the Lower Ordovician Pakerort and Varangu stages and the Mäeküla Member of the Billingen Stage wedge out (Vaher et al. 1964). Three layers of conglomerate, containing quartzite pebbles transported from the crest of the basement monadnock, were found in the Vendian (Fig. 119 - 47) on the southern flank of the fold. In the claystone of the Dominopol’ Stage the amount of glauconite and particles of silt and sand size increases toward the crest. Above the crest, the Lontova Stage is 12 to 18 m thinner, and the Lower Ordovician rocks show specific facial changes. In places, basal beds of the Billingen Stage contain rounded fragments of the underlying graptolite argil­lites (Fig. 119 - 49) and pebbles of Obolus sandstones (Fig. 119 - 45, 46, 49), both of the Pakerort Stage. Iron oolithes were found in an exceptional place: in the lowermost strata of the Saka Member of the Volkhov Stage.

The above structures remind monadnocks described in the seabed of the western central Baltic (Flodén 1984) and plain‑type folds in the Mid‑Continent Area (Clark 1932) of the North American Craton. Afanasyev and Volkolakov (1981) classified the Uljaste Placanticline as a supratenuous fold generated by unequal compression of sediments over the buried basement hill. However, it must be admitted that the principle has certainly been reconsidered because the above facial changes in the Lower Ordovician cannot be explained by this mechanism. According to Vaher et al. (1980a), an uplift took place during deposition in the Vendian, the crest area being a sourceland from time to time. Mens et al. (1981) showed that the repeated events of uplift took place during regional interruptions of Vendian and Cambrian sedimentation accompanied by significant erosion. Movements were renewed in the Early Ordovician and after the Ordovician, possibly in the Early Devonian. Thus, the total folding of the covering rocks is due to both differential compaction and repeated uplift.

Narrow, some 1-to-4-km-wide linear zones of disturbances intersect both the sedimentary cover and the crystalline basement. These zones divide the homocline into a number of blocks with different sizes. In Estonia, serving as a transition zone between the Fennoscandian Shield and the Baltic Syneclise, the amplitude of block displacement does not exceed some 50 m (Puura 1986). In the central part of the Baltic Syneclise (Latvia) a few similar blocks show displacements of the basement surface up to 600 m (Suveizdis 1979).

The zones of disturbances have been studied in detail in northeastern Estonia (Vaher et al. 1962, Puura 1986, 1987) which is the best-known portion of Estonia because of extensive exploration for oil shale and phosphorite (more than 10,000 boreholes per 2900 km2). Figure 120 shows a typical network of boreholes. As a rule, the zones of disturbances occur as a flexure above a basement fault combined with anticline in the upthrown side and with syncline in the downthrown side (Vaher et al. 1978). Total vertical amplitude (At) in the sedimentary bedrock consists of components, known as anticlinal (Aa), monoclinal (Am), and synclinal (As), as shown in Figure 121. There are fracture zone(s) on the more steeply dipping limb and, in places, fault(s) in the fracture zone. Main parametres of the zones of disturbances are given in Table 25.

The folds in the zones of disturbances are in general very gentle and not high, usually 5 to 10 m and never more than 50 m. Anticlinal belts are ascribed mostly to fault movements in the basement because they are very asymmetric in shape. The Viivikonna Anticline (nose in Fig. 120) is an excellent example of that kind of structure. The beds in its southeastern limb have a very gentle average dip of 20' (max. 55') towards southeast, in the northwestern limb up to 7° towards southwest (Fig. 121A). In the 125‑m-wide fracture zone the intensity of jointing increases toward the centre, and some oil shale beds are substituted by karst clay. In the 40-to-50‑m-wide part of the structure, known as the shatter zone, the carbonate rocks are so intensively cracked and mixed with karst clay that it is difficult to tell whether there are faults in the zone or not. The uplifts are often called monoclines, because the opposite flank is nothing but a flexure (Fig. 121B).

Alterations of deformed rocks within the zones of disturbances show a long and multistep development of the zones, and fluid migration in the upper crust (Vaher et al. 1962, Pichugin et al. 1976), including metasomatic dolomitization of limestones, sulphide lead and zinc mineralisation, karst, etc. (Fig. 122).

The Aseri Zone was studied in 1975‑77 at Viru‑Nigula (Fig. 123) by a profile of six boreholes finished in the basement (Vaher 1983, Puura 1987). If the thinning of some Upper Vendian (Uusküla and Voronka) and Lower Cambrian (Kestla+Mahu) members above the crest of the anticline (Table 26) is due to vertical movements during deposition, the uplift started in the Late Vendian, ceased for a time, renewed at the end of the Vendian, and ceased again for a time. A further uplift took place in the Early Cambrian (Lontova Age). It was followed by a long‑term stable period which, as found out in southern sections of the structure, runs at least to the Wenlock (Early Silurian). The age of the folds in the Ordovician and Silurian strata is probably Early Devonian. Along some of the zones of disturbances one can recognize contrasting movements in succeeding periods. Thus, it appears that the Ahtme Monocline in the Ordovician strata is due to a rise of the northwestern block, while a minor fault in the Ahtme Zone is indicative of a later rise of the southeastern block (Fig. 121B).

Probably, specific disturbances (without roots in the basement) which occur in the sedimentary cover of the northeasternmost part of Estonia are of glaciotectonic nature. According to Jaansoon-Orviku (1926), the markedly tilted (20 to 70° SSE or SSW) Lower Ordovician strata in exposures on the slopes of three hills (Pargimägi, Põrguhauamägi and Tornimägi) at Vaivara belong to erratics. Miidel et al. (1969) have described folded rocks (Fig. 124a) on the western slope of the Tornimägi Hill (Photo 34) and erratics (Fig. 124b) in the section of the Pargimägi Hill, the hills themselves being end moraines. In a northerly and norheasterly direction, between Sillamäe and Narva, some narrow linear structures (Fig. 125) are due to clay diapirs (Vaher & Mardla 1969). One anticline was studied in an artificially produced trench where the dip of the Lower Ordovician carbonate strata little by little increases toward the centre of the structure until reaching 40°. Carbonate rocks were eroded from the central part of the anticline, and the exposed Lower Cambrian (Dominopol’ Stage) siltstone beds are virtually vertical in the middle of the structure. A similar structure was demonstrated by Orviku (1930b) at Narva - Kalmistu where the dip of the Lower Ordovician carbonate strata and the Lower Cambrian siltstone beds is 20 and 550, respectively.

In the drill core of borehole No. 314 (Fig. 126) at Sinimäe, in the centre of another anticline, the Vendian strata are practically undisturbed above the basement of normal altitude. The claystones of the overlying Lontova Stage are disturbed by numerous slickensides ever increasing upward in number. Clay- and siltstones of the Dominopol’ Stage are severely disturbed, and nearly three times as thick as normal. Evidently, the claystone was locally squeezed upward along the lines of minimum resistance.

In places, the general direction of the clay diapirs coincides with the usual trend of the above linear zones of tectonic disturbances in northeastern Estonia. Thus, the diapirs might inherit the position of pre-existing disturbances.

Small depressions of unknown origin, 0.1 to 1km in diameter, have been found in oil‑shale mines. The largest (1.3 to 2 km), Sämi (Kabala) depression is well-known (Vaher et al. 1962) owing to exploration for oil shale in 1950‑52. This is an oval structure with a slightly wavy eastern limb (Fig. 127), in the centre of which the Ordovician strata lie 50 m below the level of surroundings. The amplitude on top of the basement is not known.

On the Island of Hiiumaa, a large impact crater is buried under Upper Ordovician carbonate rocks (see Ch. XI, 2). It is nearly 4 km in diameter and 540 m deep (Puura & Suuroja 1992). The diameter of the largest, Neugrund crater is 7 km.


Neotectonics and recent crustal movements

A. Miidel & R. Vaher


Under neotectonic movements the authors mean and deal with the Neogene‑Quaternary crustal movements. Since deposits from the Upper Devonian through Neogene are absent in Estonia, the reconstructions of the neotectonic movements are to a great extent based on the analysis of the bedrock topography. The latter has been regarded as consisting of polygenetic planation surfaces of different age. It has been suggested that either Miocene‑Pliocene (Mozhayev 1973, Šliaupa et al.1982), Miocene (Isachenkov 1982, 1983) or Mesozoic and Palaeogene (Meshcheryakov 1965) planation surfaces are found in the bedrock topography in Estonia. According to Mozhayev (1973), the total Neogene‑Quaternary uplift reaches 200 m.

The formation of the bedrock topography started at the end of the Devonian, and is polygenetic in character. Besides long‑term crustal movements and various continental denudational processes, its development was controlled by glacial erosion in the Pleistocene. Isachenkov (1982) claimed that in the northern part of the Baltic region, glacial erosion reached the amount of 80 m, which often exceeds the relative height of the bedrock surface features treated as planation surfaces by other authors. Taking into consideration remnants of erosional topography of Middle Devonian age, established in western Russia (Sammet 1961, Tuuling 1988b, 1990) and in Estonia (Puura et al. 1987, Kiipli 1989), the bedrock topography represents, in places, exhumed surface of very old age.

The sloping of the Earth’s surface toward the present Baltic Sea is a Cenozoic feature. It is important to know when the general slope was formed. According to Karabanov et al. (1993), in the Late Oligocene - Quaternary, the East-European Craton south of the Gulf of Finland and east of the Tornquist’s line experienced a regional inclination (inversion) from the Ukrainian‑Voronezh Uplift to the Baltic Sea. According to Puura (1980), denudational processes were activated earlier, in the Palaeogene by an asymmetric uplift of Fennoscandia, and a concurrent uplift of the southwestern part of the East-European Craton. The Baltic Sea depression as a region of minimal uplift accumu­lated waters from surrounding areas.

Highly disputable is the role of local differentiated movements. Rähni (1973b) was of the opinion that a number of the bedrock surface features corresponded to blocks, mobile in the Quaternary period. During the course of the past 12,000 to 13,000 years some of those blocks rose at least 30 m, others sank 10 to 20 m. On the neotectonic map of the Soviet Baltic Republics (Šliaupa et al. 1982) five neotectonic uplifts (Haanja, Keila, Nuia, Pandivere and Võru) and two neo­tectonic depressions (Pärnu and Võrtsjärv‑Mustvee) are shown, all in a good accordance with bedrock surface features.

The question now is: how many structural features of the sedimentary bedrock are actually in agreement with palaeolandforms? A special study, the morphotectonic analysis in northeastern (Vaher & Tavast 1979) and southeastern (Vaher et al. 1980a) Estonia has proved that direct morphostructures are infrequent in those areas (Raukas et al. 1988). Of the 18 zones of disturbances, studied in detail in northeastern Estonia, only the Viivikonna Anticlinal (Fig. 120) is expressed in the bedrock topography. On the other hand, of the six dome‑like placanticlinales, known in northern Estonia, four are expressed in the bedrock topogra­phy (Fig. 119), one is not expressed at all, and for one there is too little information available.

The structural-topographical relations in the sedimentary bedrock are rather complicated in the Haanja Heights (Fig. 117). The southern part of the heights intersects with the steeply-dipping flank of the anticline whose amplitude accounts for more than 120 m on the bottom of the Upper Devonian Pļaviņas Stage. This flank is not expressed in the bedrock surface. The amplitude of the northerly dome on the above‑mentioned bottom is some 50 m (Paasikivi 1966). Thus, the possible amplitude of relatively young movements does not exceed 50 m. It forms about one half of the relative height of the Haanja Heights in view of its bedrock surface.

The fairly large Pandivere Upland is intersected by six zones of disturbances trending mainly northwest to southeast, and with the steeper-dipping flanks of the anticlines mostly to the northwest. They divide the area into several blocks with a throw relative to one another of 10 to 20 m (Fig. 115). A very gentle southward regional dip is characteristic of the area on the whole, and no traces of major block (or fold), corresponding to the Pandivere Upland, have been found (Vaher & Tavast 1979, Karukäpp & Tavast 1985).

Disagreement between a medium‑size bedrock surface feature and structural features of the sedimentary bedrock is best revealed in the Ahtme Elevation in northeastern Estonia (Fig. 120). Against a background of homoclinal dip of the Middle Ordovician Kukruse Stage, the Ahtme Zone of diturbances with the total vertical amplitude of up to 18 m, is distinctly observable. It intersects the elevation diagonally and is not expressed in the bedrock surface.

The above provides a basis for the conclusion that the formation of the bedrock topography has been but little effected by postglacial and recent blockwise differential movements. Morphotectonic analysis enables one to prove the occurrence of those movements in a few cases only.

The area of Fennoscandian continental glaci­ation was repeatedly subjected to crustal downsinks and uplifts caused by ice sheet’s loading and unloading. The recent and present uplift of Fennoscandia, including the northwestern part of Estonia, is an evidence of glacioisostatic relaxation of the crust.

The Late-glacial and Holocene crustal movements have been studied in more detail. Based on numerous distance diagrams compiled, starting from Ramsay (1929) and ending with Kessel and Raukas (1979) (Fig. 128), the following conclusions have been reached:

(1) in the Late-glacial and Holocene, this area experienced crustal uplift and tilting;

(2) the amount and rate of the uplift decreased from the northwest to the southeast;

(3) the rate of the uplift has been decreasing from the Late-glacial to the present time.

The total uplift in Estonia without the eustatic factor is at least 85 m; however, extrapolating the glacial lake level G1 along the 355° azimuth up to the Kõpu Peninsula (Hiiumaa Island), the total uplift reaches 115 m (actually, the highest shoreline has there an height of 60.8 m a.s.l.). The amount of the total uplift in southeastern Estonia is not known so far. It has been estimated at about 25 m (Orviku 1960c).

Most of the uplift was evidently realized at the end of the Late-glacial. The preliminary calculations, now a little out of date, suggest that in northwestern Estonia the uplift pro­ceeded at a rate of 30 mm per year, and reached the amount of 65 m in the time interval from the end of the Middle Dryas up to the Pre-Boreal (Kessel & Miidel 1973).

In the Holocene, the rate of uplift decreased considerably. At that time, the average rate of uplift on the Kõpu Peninsula, in Tallinn and Pärnu was 5.0, 4.0 and 1.0 mm per year, respectively.

Through the postglacial, the rate of uplift probably decreased unevenly. As is seen from the gradient curve (Fig. 129), the rate of uplift decreased during the periods lasting from 9,500 to 9,000 and 6,800 to 5,300 years ago. Thereafter, the rate stabilized to decrease again some 4000 years ago (Kessel & Miidel 1973).

The trend surface analyses have revealed that during Ancylus time northwestern and western Estonia rose fast, but in Litorina time these areas experienced considerable sinking (Miidel 1995). Irregularities in uplift are also expressed in the change of the direction of tilting from 155° (local ice-dammed lakes) to 130‑135° (recent crustal movements) (Pärna 1962, Kessel & Miidel 1973, Kessel & Raukas 1979, Miidel 1995).

The gradient of the Baltic Ice Lake shorelines changes across the line Pärnu ‑ Navesti River ‑ Narva (Pärna 1962), forming a hinge line in the crustal tilting. To the northwest from the hinge line, or belt, the rate of uplift was higher than in the southwest. The belt coincides in places with tectonic dislocations in the basement and Palaeozoic rocks. It is assumed (Orviku 1960c, 1969), that in the Late-glacial the dislocations became active. Regionally, this belt partly belong to the Svecofennian deep‑seated fault zone Vyborg ‑ Pärnu ‑ Liepaja (Puura 1979, Raukas & Hyvärinen 1992). Another hinge zone in the area of the Gulf of Finland was demonstrated by Donner (1966, 1969, 1970). The hinge zone, expressed as a change in shoreline gradients between Finland and Estonia, must be controlled by tectonics (Donner 1970).

By means of geodetic, mareographic and cartographic methods it has been established that crustal movements have continued up to the present (Zhelnin 1960, 1964, 1966, Vallner & Zhelnin 1975). According to the most recent map (Vallner et al. 1988), the major part of Estonia is rising at a rate up to 3 mm per year northwest of the line Lake Võrtsjärv ‑ Tartu), whereas south­eastern Estonia is sinking at a rate of up to 0.8 mm per year (Fig. 130). In general outline, the isolines of the recent crustal movements coincide with the late‑ and postglacial ones, but reveal a great deal of irregularities. This may be due to block movements along rejuvenated, NE‑SW‑oriented fault zones (Sildvee & Miidel 1978, 1980; Vallner et al. 1988). The most prominent gradient zone along the line Pärnu ‑ Tapa ‑ Kunda coincides with the zone of deep‑seated fault in the bedrock and partly with the above-mentioned hinge line. River activity in the Pärnu River basin has been noticeably influenced by crustal movements in the zone (Miidel 1966b, 1991, Sildvee et al. 1973). Recently, it was demonstrated (Miidel 1994) that the Pärnu ‑ Kunda gradient zone may lie at northeastern border of the central part of a present regional uplift anomaly, established by Svensson (1989, 1991).

Vallner (1978, 1981; Vallner et al. 1988) divided Estonia into five annual velocity planes of vertical movements (southern, central, southeastern, southwestern, and western Saaremaa), moving in parallel or under a small angle relative to one another.

During the past 20 years, the problems of seismicity have become topical. Prior to the Osmussaar earthquake on October 25, 1976, Estonia had been considered aseismic. The unexpectedly strong Osmussaar earthquake aroused interest in seismic phenomena. As a result of numerous studies (Klaamann 1977, Bulin 1978, Avotinya et al. 1988, Sildvee 1988, 1991, Nikonov & Sildvee 1992, Sildvee & Vaher 1995), the former concepts about the seismicity of Estonia have been entirely changed.

According to Sildvee and Vaher (1995), twenty four macroseismic and numerous small (M<3.5) instrumental events have occurred in Estonia since 1602 till 1991 (Fig. 115), their intensity being 3 to 7, magnitude 1.5 to 4 and depth of the focus 5 to 14 km (for local shocks about 5 km). The most powerful was the Osmussaar earth­quake (magnitude 4.7, intensity 6 to 7), with a epicentre 5 to 7 km northeast of the island. Its impact was felt over an area of 191,000 sq km (Kondorskaya et al. 1988). According to macroseismic evaluations, the depth of its focus was about 13 km, according to instrumental studies - 10 km. After Kondorskaya et al. (1988) the epicentre is connected with the fault running from the Central Baltic via the northwestern part of Hiiumaa Island to the coast of Finland near Hamina. However, it is not excluded that the epicentre lies at the fault line Lahti‑Porkkala (Sildvee & Miidel 1980, Raukas & Hyvärinen 1992).

The distribution of the earthquakes in Estonia is not random. Their occurrence must be, to a certain extent, correlated with geology and tectonics. The majority of macroseismic epicentres are located in the northwestern part of Estonia, i. e. in the region where the rate of the uplift is highest. A considerable number of macroseismic events occur in the Paldiski ‑ Pskov gravity and aeromagnetic anomaly zone which is related to a deep‑seated fault (Sildvee & Vaher 1995). The epicentres of small instrumen­tal events, concentrated in the northern part of Lake Võrtsjärv and its surroundings, may be located in a minor cross fault. It is of interest that they are situated in a gradient zone of recent crustal movements.




Formation of the Earth’s crust

V. Puura & V. Klein

Formation of main features of the continental crust in the region

The development of the Earth’s crust and its main para-meters in the northern Baltic and adjacent areas (Puura & Flodén 1996) was controlled by two processes: 1) creation of the crust during the Palaeoproterozoic Svecofennian orogeny, and 2) changes in the juvenile continental crust caused by bimodal rapakivi‑anorthosite magmatism (Palaeo- to Mesoproterozoic).

Recent basic achievements in geophysical and geological studies of the Earth’s crust have changed the general views on the formation of the earliest, Precambrian continental crust.

First, the applicability of the plate tectonic fundamentals to the Proterozoic crustal growth was veryfied in our crustal domain, on the example of the Palaeoproterozoic Svecofennian crust in the northern part of the Gulf of Bothnia (BABEL... 1990). It became evident that also the Precambrian structures in the Baltic basement should be interpreted in terms of plate tectonics.

Second, according to the general time trends in crustal thickness, the Proterozoic crustal domains have the thickest crust (40-55 km) compared with the domains formed earlier (27-40 km in the Archean) or later in the Phanerozoic (Durrheim & Mooney 1994). Consequently, the extraordinarily thick crust of the Palaeoproterozoic Svecofennian Domain (50-65 km) fits well to the global regularities.

Third, in the areas subjected to crustal shortening due to compression during the orogenic stage, lateral displacements of deep crustal slabs into the upper crust are common. In the overview of the Precambrian crust in the Fennoscandian Shield and adjacent Russian Platform (Gorbatschev & Bogdanova 1993), the model of crustal stacking as the mood of formation of the beltiform Belarussian-Baltic, granulite terrain (including southern and western Estonia) was introduced.

After the formation of the juvenile continental crust, substantial changes may have taken place in its composition and thickness. One of the most significant processes of the crustal deformation accompanied by crustal thinning is rifting and the related magmatism in tensional stress fields. Formation of the rapakivi and related rocks of Fennoscandia has long been interpreted as a result of magmatism in tensional tectonic environments. The bimodal composition of magmas was due to the melting of both the lower crust and uppermost mantle (Rämö 1991). The bulk of felsic magmas intruded into the uppermost crust and extruded on its surface. The mafic magmas only in places reached the Earth’s surface (anorthosite plutons and diabase dikes and sheets), while the bulk of their volume remained in the low levels of the crust. Crustal thinning due to basaltic underplating was stated around the Vyborg Rapakivi Pluton (Elo & Korja 1993) and in the area of the Åland Pluton (BABEL... 1993). The deep crustal signatures of the rifting type were identified by seismic sounding near the Åland Rapakivi Pluton in the southernmost part of the Gulf of Bothnia. The whole area of crustal thickening down to 50‑40 km in the central part of the Svecofennian orogenic domain was interpreted as a result of the bimodal anorthosite‑rapakivi magmatism in tensional environments (Puura & Flodén 1996) in the time span of 1645-1540 Ga (Rämö et al. 1996).

Thus, the thick crust in southeastern Estonia is considered as a survived primary Svecofennian crust, while the considerably thinned crust of northern and western Estonia is due to rapakivi‑time reworking, including basaltic underplating.


Formation of the continental crust during the Svecofennian orogeny

A principal question is what kind of conditions existed in the region before the Svecofennian orogeny. By means of Sm-Nb isotopic studies of typical Svecofennian rocks of southwestern and central Finland (Huhma 1986) and granulites of central and southern Estonia (Puura & Huhma 1993) it was established that there were no Archean crustal remnants in the internal areas of the Svecofennian orogen. Thus, the Svecofennian crust was formed as a juvenile feature in an area of oceanic crust. It can be supposed that oceanic environments existed in an area larger than the present Svecofennian crustal domain, because the latter was formed during the subsequent process of substantial crustal shortening.

However, little is known about the crustal processes before 1.9 Ga. In the border zone of the Karelian Archean continental domain, ancient rift structures older than 2 Ga have been documented. It was concluded that they originate from the times of opening of the pre‑Svecofennian ocean. The ophiolite series of the Outokumpu area in the Karelian‑Svecofennian transition zone has been considered to be a fragment of the pre‑Svecofennian oceanic crust of an age ca. 2.02 Ga which was obducted on the edge of the Karelian Craton during the Svecofennian orogeny (Koistinen 1981).

Plutonic, volcanic and sedimentary orogenic suites, isotopically dated up to now in the internal parts of the Svecofennian Domain, show ages of magmatism and metamorphism in the time span of 1.9-1.8 Ga (Huhma et al. 1991). It means that no direct evidences on pre‑Svecofennian rocks have been found there so far. However, at least two circumstances might point to the existence of the sialic crustal compounds before the final Svecofennian orogeny. The SHRIMP dating of detrital zircons from clastic metasedimentary rocks of the Svecofennian interior (Huhma et al. 1991, Claesson et al. 1993) has shown that at least in Finland and Sweden there existed source rocks with an age of ca. 2.1‑1.9 Ga. Basing on the geochemical and isotopic studies of the South-Estonian granulites, it was concluded that original volcanic rocks were created in the crust having sialic signatures (Puura & Huhma 1993). As a whole, the data obtained suggest probable existence of a considerably thick sedimentary, or volcanic layer, or both, on top of the mafic ocean floor before the island arc magmatism started. Somehow it reminds recent geological environments transitional from oceans to continents.

Basing on deep seismic (BABEL... 1990) and magnetotelluric soundings (Korja et al. 1993) and on the interpretation of petrological data of igneous rocks (Hietanen 1975), the fossil volcanic island arc structures have been determined in the northern part of the Svecofennian Domain. The Vihanti - Pyhäsalmi and Tampere zones in Finland belong to the kind. Considering the petrological data, the early orogenic metavolcanic and -plutonic rocks in the Baltic area carry the same signatures.

The fossil island arc systems with deep (up to 80 km) fossil roots of subduction zones have been identified in the northern part of Svecofennia (BABEL... 1990). Taking into consideration the isotopic signatures, the early orogenic igneous rock bodies in the internal parts of Svecofennia have been interpreted as of Paleoproterozoic juvenile origin (Huhma 1986).

It has been stated that in the Fennoscandian Shield area, the thick Svecofennian continental crust was created during a short time span between about 1.9 and 1.77 Ga (Huhma 1986, Gorbatschev & Bogdanova 1993, Koistinen 1996). The early stage of the Svecofennian island arc plutonic and volcanic activity occurred during 1.90‑1.88 Ga: intrusive Svecofennian rocks in Finland have zircon ages 1890+/‑10 Ma which include early‑, syn‑, late‑ and post‑orogenic rocks of central Finland. Younger Svecofennian rocks lie within the potassium granite‑migmatite area in southern Finland and probably in Estonia. Potassium migmatite‑granites dated at 1840-30 Ma evidence of a large secondary west‑east‑striking zonal magmatism within the young (1850 Ma) orogen. In the same zone, post‑tectonic granitoids at 1820-1770 Ma occurred (Koistinen 1996). In Estonia, the possible age analogues are Virtsu and Taadikvere plutons (Fig. 5).

Judging by the isotopic data concerning the formation of the Svecofennian crust in Finland, the Estonian basement probaly belongs to an area where both early orogenic at 1.890+/‑10 Ma island arc and late orogenic within‑crustal zonal granite magmatism occurred. The accretion of Svecofennia to the present SW border of the Late Archaen Karelian crustal domain and the late orogenic tectonic, metamorphic and magmatic transformations came generally to an end at ca. 1.8 Ga.

The crust of Svecofennia, the thickest one compared with older and younger crustal domains of the East-European Craton (Platform), was created during several stages. After the pre‑orogenic oceanic igneous activity and sedimentation, the overwhelming magmatic and sedimentary activity of the island arc (subduction) stage created rock assemblages together with the interarc sedimentary basins available for later observations. Intraorogenic deformation and metamorphism together with within‑crustal partial melting (granitoids) changed the primary rock assemblages. Several phases of deformation and metamorphism have been documented and dated in the time span of 1.9‑1.8 Ga.

The crustal shortening and corresponding thickening started with the formation of subduction zones and island arcs above them. The above early orogenic igneous rocks and the first stages of deformation, metamorphism and partial melting (migmatisation) were related to extraordinarily intense processes of subduction contemporarily occurring in several island arc systems within Svecofennia (Gorbatschev & Bogdanova 1993). Erosion of island arcs became a source for sedimentary depressions between the arcs, the largest of which in Estonia is the Alutaguse Zone.

The next stage of crustal shortening occurred as compression and fragmentation of fossil island arcs and sedimentary basins, with the second stage of general folding and metamorphism. In the late orogenic period, other phases of faulting and infracrustal granitoid magmatism occurred, the latter being represented by the potassium granite intrusions and migmatites. The step‑wise cratonisation of Svecofennia finished with the intrusion of the so‑called post‑orogenic granitoids at 1.82-1.77 Ga.

Already before the rapakivi magmatism at about 1.65 Ga, the young crust was deeply eroded up to the levels of mainly amphibolite, in places to granulite grade. Starting already from the early orogenic island arc stage at about 1.9 Ga, single mountain chains in marine surroundings became subjects of intense erosion. In this early stage, depositories of sediments located in the between‑arc space. Afterwards the whole orogenic domain uplifted at about 1.87-1.85 Ga and turned into a continent with mountain chains. Intense erosional processes planated the original complex mountain relief. Erosional debris was removed outside the whole domain and their depositories are not known yet.


Crustal changes during the Meso‑ to Neoproterozoic

The above principal reconstruction of the Svecofennian crust during rifting and rapakivi and related magmatism at 1.65-1.54 Ga was coupled with one more tectonic blocking of the crust and differentiated movements of the blocks reaching kilometres (Koistinen 1996). The volcanic complexes with volcanic topography above the rapakivi plutons diversified the mountain environments. The concurrent and subsequent erosion planated the Earth surface composed of both rapakivi and pre‑rapakivi rocks. Before the first epicontinental Jotnian sedimentation at about 1.4 Ga, the first perfect peneplain was worked out in the Svecofennian Domain, as it can be concluded from the survived fragments of Jotnian clastic sedimentary basins in the Bothnian Sea ‑ the Satakunta and Lake Ladoga areas (Koistinen 1996).

However, the post‑Jotnian (about 1.2 Ga) and younger (at about 1.1-0.9 and 0.6 Ga) events of mafic dike magmatism evidence about the repeated blocking of the crust. The gentle depressions, diversified by faults with horsts and grabens identified in the Jotnian sedimentary cover, carry information on the reaction of the crust on both tensional and compressional stresses in the region. The latter data have been gathered from the neighbouring areas of Fennoscandia and the Baltic Sea (Koistinen 1996, Puura & Flodén 1996).

The pre-Late Vendian/Cambrian peneplain has been worked out upon the complex target composed of Svecofennian orogenic, rapakivi‑time anorogenic, and Jotnian and later epicratonic sedimentary and within-cratonic plutonic rocks.


Crustal changes during the Phanerozoic

The pre‑Late Vendian/Cambrian peneplain marking the pre‑Phanerozoic Earth’s surface has fully survived in Estonia and surrounding areas under the sedimentary cover. As shown above, the total thickness of the survived Palaeozoic sediments is less than 800 m. According to some estimations, the total rate of erosion of the Phanerozoic sedimentary cover reached 500‑800 m. During the Late Caledonian compressional phase, local vertical displacements in zones of disturbances did not generally exceed 50 m in Estonia. As an exception serves the Riga-Pskov fault zone on the Estonian‑Latvian border where it reached 600 m. All the other tectonic phases have caused weaker deformations.

As a summary, the crustal conditions during the Phanerozoic can be classified as most stable cratonic ones. Since the opening and spreading of the North-Atlantic Ocean, the Fennoscandian‑Baltic region as the northwestern part of the East-European Craton was situated in a most quiet tectonic regime in terms of plate interactions which promoted isostatic levelling of the crust. During the Cenozoic, the Svecofennian Domain became a morphostructure with the slightly elevated transitional zones and subsided (locally even submerged) interior. The primary Svecofennian crust is thick and light. The secondary crust, reworked by rapakivi-anorthosite magmatism, in the Svecofennian interior is thinned and denser. The buoyancy differencies between these crustal sections are considered as having been responsible for the formation of the Baltic Sea drainage basin with the Baltic Sea in the centre (Puura & Flodén 1997).

The late Cenozoic glaciations caused repeated glacioisostatic downwarping and uplift of the Earth’s surface in and around the Fennoscandian centre of glaciation. At present, the upper crust is in a stage of postglacial uplift. The glacioisostatic crustal movements were superposed to the Cenozoic Baltic depressional morphostructure and, thus, diversified the morphology of the latter. They have been more localised and short‑term compared with its larger and long‑term features.


Vendian-Tremadoc clastogenic sedimentation basins

K. Mens & E. Pirrus


After the formation of the Palaeo-Mesoproterozoic crust, a long-lasting continental period followed. A deep denudation shear on the bedrock marks this important discontinuity in the pre-Vendian geological history of the region. The topmost 0.5 to 150 metres of the basement have been changed due to weathering.

In general outline, the bedrock succession in Estonia is divided into three parts (in ascending order): the mainly terrigenous part of Vendian-Tremadoc age, the Ordovician-Silurian carbonate formation and the Devonian, predominantly terrigenous rocks.

The Vendian-Tremadoc rocks occur all over the East- European Platform, particularly in its western and central parts. In Estonia, they form a complex with a thickness of up to 220 m and are non-existent only in the area of the Lokno-Mõniste-Valmiera Uplift (Fig. 131).

Despite the stratigraphic incompleteness and several minor or major hiata, the general stability and regional uniformity of the succession is remarkable. To provide a better understanding of Estonia’s geological development during the Vendian-Tremadoc, the palaeogeographical situation in much larger area than Estonia will be discussed (Figs. 131-137, Rozanov & Lydka 1987).

The sedimentation of the lower, terrigenous part of the bedrock in the area under consideration was discontinuous due to the cyclically repeated transgressions and regressions related to the tectonic evolution of the platform in general. In the northwestern part of the area, seven evolutionary stages of deposition may be distinguished. The sedimentary complexes formed during these stages differ in the distribution, structure, thickness, facies composition, lithological characteristics, number of hiata and several other features (Fig. 132). In ascending order they correspond to the following stratigraphic intervals (Tables 5, 6): Valdai Subseries; Baltic, Liivi, Aisčiai and Deimena series, late Middle Cambrian - early Late Cambrian and late Late Cambrian - Tremadoc (Fig. 132). The former four complexes are characterised by easily defined cyclic structure and regular alternation of lithotypes reflecting the transgressive-regressive nature of sedimentation. Three upper complexes occur in Estonia as remains of the complexes, most completely developed in the west and east. These complexes are mostly represented by sandy deposits formed under shallow-water conditions and complicating the interpretation of the development.


Valdai evolutionary stage

Widespread typical platform facies in the northwestern part of the East- European Platform are known from the second half of the Vendian. The subsidence of the Earth crust at the beginning of the Vendian embraces only the narrow linear depressions, the so-called aulacogenes, seated deep in the pre-platform basement far from Estonia.

The early Vendian volcanic, sedimentary-volcanic and sedimentary rocks of continental origin deposited during the second half of the Drevljan Stage and are confined to the extensive Orsha Depression which partly reached the area under consideration (Fig. 133A).

The Valdai complex, representing the Upper Vendian in the East-European Platform, is of wider distribution and similar in character throughout the platform. Its deposits overlie transgressively older accumulations of the platform bedrock cover or occur immediately on the Palaeo-Mezoproterozoic basement.

The distribution of the Valdai complex testifies to the emplacement of a new structural framework on the East-Europen Platform (Craton) which existed throughout the terminal Neo-proterozoic and initial Palaeozoic. Its principal characteritic feature is an extensive eastern transgression resulting in the formation of the large Moscow Basin, which influenced upon sedimentation in the northeastern part of the platform, including Estonia.

The Valdai section can be divided into two great sedimentary cycles: Redkino (below) and Kotlin (above).

Redkino Age. The Redkino transgression was discontinuous with intermittent regressions. Evidence is derived from the rhythmic development of the strata. Each new transgression started with a basal coarse-clastic sandy unit and graded into a clay unit corresponding to a submerged maximum. Deposits of the regressive phase have not usually preserved.

The Redkino rocks, not known in Estonia, are distributed immediately beyond its boundaries - in the Leningrad Region and eastern Latvia (Fig. 133B). According to the specific features, such as the occurrence of soft-bodied Metazoans, glauconite, early diagenetic pyrite and phosphates, they accumulated under typical marine conditions.

The end of the Redkino accumulation is marked by an uplift in an extensive area. It changed sedimentation conditions and broke the link with the ocean.

Kotlin Age. The rocks of the Kotlin Stage are the only representatives of the Valdai evolutionary stage in Estonia. The uplift of the northwestern part of the platform at the end of the Redkino Age led to a slight erosion and subsequent subaerial weathering of the topmost beds of the Redkino Stage. This short-term continental period was followed by an extensive Kotlin transgression. The development of the Moscow Basin proceeded from the structural framework of the preceeding Redkino Age, being the major subsidence zone in the centre of the platform. Changes in the framework can be seen in the northwest where the sedimentation spread extensively westwards and most of Estonia was submerged (Fig. 133 C).

In Estonia, the rocks of the Kotlin Stage rest directly on the weathered crystalline basement. At the beginning of the Kotlin transgression, the mellow surface of the weathered basement was partially destroyed and redeposited in a basin as follows from the composition of the basal Oru Member. Upwards the share of the clastic material derived from the weathered crust decreases. Nevertheless, the enchanced content of feldspars, micas and kaolinite and their poor roundness in rest of the Gdov Formation (Table 6) indicate low-grade reworking of the clastic material inherited from the surrounding areas of igneous or metamorphic, or both, rocks. The grain-size decreased constantly upwards until the formation of a thick stratum of the “Laminarite clays” (Kotlin Formation). The clay component of this argillaceous rock is low in kaolinite and high in illite-chlorite (Pirrus 1970, 1983), which may be partially due to the redeposition of Redkino deposits into the Kotlin basin (Pirrus 1980, 1987).

The assemblage of authigenic minerals, total lack of the body fossils, weakly developed varved structure of clays and low content of boron (Bitjukova & Pirrus 1979) suggest accumulation of the argillaceous deposits of the Kotlin Age in a basin similar to ice-dammed lake. The above-mentioned deposits wedge out abruptly without transitional facial belts along the meridian in central Estonia (Fig. 133C). A similar kind of accumulation is observed in the recent arctic sedimentation regime which is in accordance with palaeospatic reconstructions of Late Proterozoic time (Scotese et al. 1979).

The terminal phase of the Kotlin clay accumulation in Estonia is marked by some shallowing of the sedimentary basin and the appearance of interbeds of sand- and siltstones (Laagna Member).

The succeeding uplift of Estonia’s territory is marked by the weathered uppermost part of the Kotlin Formation (Mens & Pirrus 1969, 1970). The short-term (small thickness of the weathering crust, absence of kaolinite-zone) subaerial development was followed by the formation of the shallow-water Voronka basin, the variegated silty-clayey deposits of which (Sirgala Member) are upwards gradually replaced by well-sorted quartzose sandstones (Kannuka Member). More likely, this part of the Vendian sequence represents a small separate sedimentary cyclithe of regressive origin.

The present limits of the Kotlin Stage are erosional, except the northwesternmost part (Fig. 133C) and cannot, therefore, serve as real boundaries of the Late Vendian sedimentary basin.


Baltic evolutionary stage

Like in the preceeding Vendian, the Baltic sedimentary stage was influenced by the sinking of the Moscow Depression. After the uplift of the entire Vendian basin area at the end of the Vendian, accompanied with the denudation or weathering, or both, processes, the marine basin again revolved into the northwestern part of the platform. The conditions of sedimentation changed considerably from the beginning of the Cambrian. The configuration of the transgression remained unchanged, but the rock composition is indicative of a distinct rearrangement in the hydrochemical regime of the basin. Despite the continuing accumulation of sandy-clay deposits, a typical marine assemblage of authigenic minerals (glauconite, phosphates, early diagenetic pyrite, etc.) appears. This shows that by the time of the Early Cambrian accumulation, the sedimentary basin had a link with the ocean and was subject to long-term normal marine regime.

The reconstruction of the palaeobasin of the Baltic evolutionary stage is complicated enough, due to intensive denudation processes prior to the accumulation of younger, Cambrian deposits.

Rovno Age. The Rovno deposits are absent in Estonia, but distributed immediately east of its boundaries. The sequence of the Rovno Stage starts, as a rule, with grit- or coarse-grained, or both, sandstones followed by an accumulation of argillaceous rocks. In some sections, a distinct bipartite structure – two smaller scale cyclithes (rhythms) can be recognized which differ in palaeontological characteristics (Paškevičiene 1986). If the upper cyclithe is absent, the topmost beds of the lower cyclithe are usually weathered.

The lack of coarse-grained rocks within the present-day distribution area provides a basis for the supposition that during the Rovno Age the basin might have been larger and the area of Estonia could have been submerged, at least partially. This is in good accordance with the present western and northern limits of the Rovno Stage which are characterised by argillaceous rocks accumulated in the inner part of the basin (Fig. 134A).

The character of the Rovno/Lontova boundary beds in the Leningrad Region and eastern Latvia shows continuous deposition which was not the case in the other sections. In this area, the deposits of the Baltic sedimentary stage form a unique cyclithe, the basal part of which is formed by Rovno deposits.

Lontova Age. The Lontova sedimentation in the second half of the Baltic evolutionary stage took place during a marine transgression which advanced from the east and was more extensive than the previous one. The Lontova rocks in the northwest are of much wider areal extent than the underlying Rovno rocks.

Most of the present-day Estonia was submerged in the Lontova Age (Fig. 131, 134B). A relatively small amount of coarse clastic material and the small proportion of sandstone in the basal beds of the stage point to a slow transgression and smooth topography of both the basin bottom and surrounding source area.

The Lontova sequence is dominated by clays. Huge accumulation of argillaceous deposits is typical of the Lontova Age over the whole East-European Platform. The prevailing illite with an admixture of chlorite in the clay component and a relatively mature character of clastic grains provide a basis for the supposition that the Vendian siliciclastic rocks of surrounding areas were the main source of sandy-silty material of the Lontova Stage.

The Lontova sedimentation is characterised by relatively smooth changes in facies environment. Evidence is derived from the gradual decrease in sand and increase in clay content from west to east (Fig. 134B). The changes in sedimentation were caused mainly by changes in the depth of the basin (Kala et al. 1981b). In Estonia, two distinct lithofacies belts may be distinguished: prevailingly sandy facies in the west (Voosi Formation) and clayey facies in the east (Lontova Formation). The former is interpreted here as marginal shallow-water deposits formed not far from the shoreline, the latter as deposits which accumulated under quiet hydrodynamic conditions farther from the coast (Mens & Pirrus 1986). On the basis of clay content, some detailed zones can be distinguished within the limits of the clay facies belt (Fig. 134B).

The recent areal extent of the Lontova Stage, except the northwestern part of Estonia, is erosional as is suggested by the lithofacies distribution. Probably, the primary northern boundary of the basin was situated farther in the north. The occurrence of lithologically and palaeontologically similar rocks in the vicinity of Torneträsk (Bergström & Gee 1985, Vidal & Moczydłowska 1996) suggests that this part of the platform was submerged by the basin having depositional conditions similar to those in the Lontova basin in Estonia. Probably, a passage may have existed across the recent Fennoscandian Shield connecting the sedimentary basin in the north with the Lontova Sea in the south (Keller & Rozanov 1980, Puura et al. 1987).

The Lontova rocks in southeastern Estonia are partly or completeley reduced (Kala et al. 1981b). This is due to the change in the structural framework on the pre-trilobite/trilobite Cambrian boundary and tectonic activity in the Lokno-Valmiera uplift area (Mens 1981b). The well-preserved crust of weathering in the uppermost part of the Lontova Stage in southeastern Estonia shows that erosional processes occurred mainly before the Liivi evolutionary stage.

The end of the Baltic evolutionary stage was marked by the regression of the sea and extensive changes in the structural framework.

The Moscow Depression, which so far functioned as the main subsiding area, underwent an uplift and stayed for a long time (up to the Middle Cambrian) above the sea level. In the western part of the platform, a new subsiding area – the Baltic Basin formed. It was opened to the west and had a direct connection with the ocean.


Liivi evolutionary stage

Dominopol’ Age. The beginning of the trilobite-bearing Cambrian was associated with a considerable reconstruction of the structural framework of the East-European Platform. The onset of the Liivi transgression was related only to the westernmost margin of the platform where a number of longitudinal gulf-like depressions were formed. The Baltic Basin, as one of those depressions, was situated at the southern slope of the Fennoscandian Shield (Fig. 134C).

The rocks of the Liivi evolutionary stage belong to the Dominopol’ Stage and are represented by alternating arenaceous and argillaceous deposits, accumulated in a shallow-water basin, and providing favourable conditions for various organisms. The flooding events with intervening water low-stands up to short-term uplifts can be established on the grounds of the composition of the Dominopol’ sequence. Its lower part (Sõru Formation) is represented mainly by arenaceous rocks with several interbeds of coarse clastic sediments, while claystones are less frequent. The succeeding part of the sequence (Lükati Formation) consists of argillaceous rocks and is terminated by well-sorted fine-grained sandstones (Tiskre Formation). All the above-mentioned lithounits are separated by hiata on boundary level, which in some places are marked with conglomerate lenses. The occurrence of phosphorite nodules and phosphatized pebbles on the lower boundaries of lithounits and the presence of strong bioturbation on several levels in the middle part of the succession indicate slow sedimentation with periods of non-deposition even regardless of the high content of clay material in the Lükati Formation.

The Liivi evolutionary stage in Estonia terminates with a gradual uplifting from the south as evidenced by weathering traces in the topmost Lükati Formation.


Aisčiai evolutionary stage

Ljuboml’ - Vērgale Age. Arenaceous-argillaceous strata corresponding to the Ljuboml’ and Vērgale stages have deposited during the first half of the Aisčiai evolutionary stage. The Rausve and Kybartai stages are missing in Estonia probably due to the postsedimentation denudation prior to the Middle-Late Cambrian. The formations of the Aisčiai evolutionary stage overlie rocks of different age down to the crystalline basement.

The Aisčiai transgression, like the preceeding Liivi transgression, advanced from west to east but its extent was more spacious (Fig. 135A).

The composition and palaeontological data of these rocks refer to a generally shallow basin. The presence of interformational breaks, generally associated with ferruginous oolith interbeds, an abundance of glauconite and phosphates on the bedding planes (Pirrus 1986) and intense bioturbation of argillaceous beds indicate slow sedimentation with periods of non-deposition during the whole interval.

The mineral composition of the clastic component suggests that the surrounding unflooded areas were mainly covered by sedimentary rocks. The occurrence of immature minerals is indicative of the uplifts of the crystalline basement and the existence of small islands within the basin.

The well-preserved crust of weathering on the underlying argillaceous rocks and the low content of coarse clastic material in the basal beds suggest a relatively slow marine transgression and a peneplaned topography of the bottom of the sedimentary basin and the source area.

The subsidence of the Baltic Basin in the course of the Aisčiai evolutionary stage was not simultaneous; it began in the north-west, spread southwards and in a stable phase (Vērgale Age) of the transgression, the whole Baltic Basin was submerged (Vidal & Moczydłowska 1996). According to the facial distribution of rocks, the eastern coastal limit of this basin was probably not far from the recent distribution area of the Ljuboml’ and Vērgale rocks. The northern boundary of the basin must have been north of the present limit of the rocks embracing most of mainland Sweden where synchronous rocks occur as small patches.


Deimena and late Middle Cambrian – early Late Cambrian evolutionary stages

The Middle Cambrian sketch-map is a generalization of the geological development of Estonia (Fig. 135B), although it presents data only on two evolutionary stages - Deimena (older) and the late Middle Cambrian - early Late Cambrian (younger), which is represented only by the rocks of the Paradoxides paradoxissimus Zone in Estonia. For the early Late Cambrian part of this unit see Figure 135C.

The Middle Cambrian is represented by incomplete successions over the whole northeastern part of the East-European Platform. The most incomplete erosional sections are in the East Baltic where the Middle Cambrian rocks are in places missing and the highly complicated evolution of the area is difficult to follow.

At the Lower/Middle Cambrian boundary, the sedimentation was relatively continuous (Mens 1981a, Hagenfeldt 1989a, b, Eklund 1990). At the beginning of the Middle Cambrian the deposition continued in the limits of the Aisčiai sedimentary stage. During the Eccaparadoxites insularis time (Table 6) the regressive deposits of the Kybartai Stage accumulated, but they have not preserved in Estonia.

After the short post-Aisčiai continental period, Estonia, at least its southern part, was again submerged. The Paala and Ruhnu formations, which in Estonia consist of sandy rocks, are Middle Cambrian and separated from the Early Cambrian rocks by hiata. In a limited area in southeastern Estonia, the Middle Cambrian rocks rest directly on the crystalline basement (Fig. 131).

The Ruhnu Formation spreading in southwestern Estonia is considered as belonging to the Ptychagnostus praecurrens Zone (Mens et al. 1990). Two sedimentary areas with varying lithofacies can be distinguished in the western part of the area under consideration. The easternmost area is characterised by the accumulation of well-sorted fine- to medium-grained quartzose sandstones yielding a few thin interbeds of brownish claystone. Evidently, they formed under high hydrodynamic conditions not far from the shoreline in a shallow-water basin expanding from the west. The mineral composition, sorting and roundness of grains suggest that the sandstones originate from the sedimentary rocks east of the basin.

Westwards, on the islands of Gotland and Öland, the above-mentioned sandstones are replaced by grey or greenish, or both, shales with intercalations of silt- and sandstones designated to the Eccaparadoxides oelandicus Zone, like the Faludden Sandstone on Gotland (Hagenfeldt 1994). Towards the west, the role of silt- and sandstones decreases and interlayers of dark grey shales appear.

The Eccaparadoxides oelandicus deposition is followed by a general regression. Evidence is derived from abrupt changes in faunas and commonly also in the lithology (Bergström & Gee 1985). The subsequent flooding was of the largest areal extent. Thus, in the middle of the Middle Cambrian (Paradoxides paradoxissimus time) sedimentation was renewed in the Moscow Depression. Estonia with the adjoining areas in the west and east served as a seaway between the two intensively subsiding areas: the Moscow Depression in the east and the “Alum Shale Basin” in the west. The latter was opened towards the Iapetus Ocean bordering the East-European Platform from the west and supplying the depositional basin with nutrient-rich water.

Rocks of the Paradoxides paradoxissimus time in Estonia are represented by the Paala sandstones, disseminated in its southeastern part. Probably, these deposits accumulated in the offshore conditions, and in that case the advance of the basin was from the east. This viewpoint is supported by the decrease of the share of argillaceous rocks and palaeontological evidence of coeval deposits in a easterly direction.


Late Cambrian – Tremadoc evolutionary stage

The uppermost part of the Vendian - Tremadoc clastogenic succession from the middle of the Upper Cambrian to the base of the Arenig forms a unitary epoch in the history of the evolution of the Baltoscandian sedimentary basin. The basin was opened to the west - towards the Iapetus Ocean. Eastwards it extended as far as the Moscow Syneclise.

The completeness of stratigraphical sequences of this interval varies greatly. The sediments of the Late Cambrian - Tremadoc usually rest on the eroded surface of Lower or Middle Cambrian rocks. The subsequent erosion removed large parts of the succession, and in Estonia only minor remains of the Upper Cambrian - Tremadoc rocks are preserved. The sequence is dominated by clastic rocks represented by varigrained sand- and siltstones and black shales (alum shales - kerogenous argillites). They contain rare interlayers of the brachiopod coquina (“Obolus conglomerata”) and greenish or greyish clay.

On the basis of the spatial distribution, thickness and number of hiata, three major depositional settings (facies belts) may be recognized in the northwestern part of the East-European Platform: a black shale facies in the west, a common siliciclastic marine sandy-silty-clayey facies in the east and a transitional facies between them including Estonia’s area (Mens et al. 1993).

The western facies belt was characterised by an accumulation of fine muds rich in organic matter (Alum Shale Formation) (see Thorslund 1960, Martinsson 1974, Andresson et al. 1985, Bergström & Gee 1985, Dworatzek 1987).

As is generally acknowledged, the Alum Shale Formation was deposited in anoxic shallow-water conditions, influenced by upwellings which brought nutrient-rich organically productive water from the bordering Iapetus Ocean in the west. The condensed black shale succession during a long-lasting period is indicative of the great stability in the sedimentary environment.

The Upper Cambrian of the eastern facies belt is dominated by sandstones, but both the share of argillaceous rocks and stratigraphic continuity increase rather quickly eastwards (Mens et al. 1990). The presence of glauconite and phosphates and faunal evidence suggest deposition in the marine basin with changeable hydrodynamic conditions.

The Upper Cambrian - Tremadoc succession of transitional facies belt is very condensed and interrupted by several hiata of different duration. The preserved deposits belong to those of the deepest part of the sedimentary basin. The deposits of more elevated areas were presumably reworked and redeposited during the subsequent period. A levelling of the sea floor took place over a large area in the East Baltic. The Upper Cambrian - Tremadoc together with the late Middle Cambrian (from the P. paradoxissimus Zone) form the two latest evolutionary stages of the Vendian - Tremadoc clastogenic sedimentation (Figs.135C, 136). The Late Cambrian part of the former includes the rocks from the A. pisiformis till the middle P. spinulosa zones, represented by the Petseri and Ülgase formations (Table 6), indicating comparatively stable depositional environments and flattened sea floor (Fig. 135C).

The terminal evolutionary stage is characterised by the increasing instability of the depositional environment. The lithological composition and palaeontological evidence of the Tsitre and Kallavere formations demonstrate the deposition under shallow-water conditions and uneven bottom topography complicated by a lot of mobile sand bars. Within the terminal evolutionary stage four phases are distinguished (Fig. 136). They are determined by the rapid changes in the sedimentation environment, characteristic to the end of the clastogenic sedimentation. The deposits of the latest phase, coinciding with Varangu time, were of wider distribution than the underlying sandy deposits (Fig. 136D). The distribution pattern of the pre-Arenig bedrock (Fig. 137) was shaped by the post-Tremadoc erosion.

The lithofacies development of the Tremadoc - Arenig boundary indicates a transition from clastogenic to carbonate sedimentation, accompanied by intensive glauconite formation.


Ordovician and Silurian carbonate sedimentation basin

H. Nestor & R. Einasto


During the Ordovician and Silurian from the Arenig to the end of the Pridoli and even at the very beginning of the Devonian (Lochkovian), Estonia was part of the northern flank of a shallow cratonic sea in which carbonate and fine-clastic sediments accumulated (Fig. 138). In the earlier stages of development this sea extended from Norway to the Volga area, and from the Finnish to the Belarussian-Mazurian Pre-cambrian massif. During the final stages of development, the basin was restricted to the Baltic Syneclise in the East Baltic area and North Poland. The nuclear part of the basin in the Baltic area is treated here as the Palaeobaltic Basin s. s.

Männil (1966) published a monograph on the development of the Baltic (Baltoscandian) Basin (s.l.) in the Ordovician. Kaljo and others composed a set of Silurian lithofacies maps for Estonia (Kaljo et al. 1970) and for the East Baltic area (Kaljo & Jürgenson 1977). Nestor and Einasto (1977) created a facies-sedimentary model for the Silurian Palaeobaltic Basin, further developed by Einasto (1986). A brief summary on the basin development and facies models was published by Nestor (1990a).


Tectonical and facies settings

Two main structural elements of the Baltic Ordovician Basin (s.s.) were defined by Männil (1966):      1) the marginal or Estonian-Lithuanian Confacies Belt, 2) the central or Swedish-Latvian Confacies Belt (= Livonian Tongue, Fig. 138). A transitional zone between the above-mentioned belts has also been distinguished. The first belt roughly corresponds to the southern slope of the Fennoscandian Shield and to the northwestern slope of the Belarussian-Mazurian Anteclise (Massif), the second belt - to the Baltic Syneclise. The marginal confacies belt was dominated by shallow-water carbonate sediments with a lot of discontinuity surfaces, while the relatively deeper-water central belt comprised predominantly clayey sediments. The present-day Estonia is situated mostly within the marginal confacies belt and transitional zone. Typical facies of the central belt reach only the southernmost part of Estonia. During the basin development, the limit of the main confacies belts gradually shifted southwestwards.

In the Early and Middle Ordovician, the western part of the East-European Platform as far as the Moscow Syneclise was slowly subsiding and covered by a shallow, epicontinental sea with a comparatively weak bathymetric differentiation and extremely slow sedimentation rate. At the end of the Ordovician, since the Late Caradoc and especially during the Silurian, the upheaval of the northwestern margin of the craton got dominance in connection with the closing of the Iapetus Ocean. At the same time, on the southwestern margin of the craton, belonging to the sphere of influence of the Palaeo-Tethys, the subsidence of the basin floor intensified. As a result of different tectonic movements, a comparatively deep, “starved” of sediments, intracratonic basin depression was formed within the limits of the central (axial) confacies belt in western Latvia, western Lithuania, the Kaliningrad District and northern Poland where hemipelagic argillaceous deposits accumulated. At the same time, the sea gradually retreated from the northwestern and central parts of the East-European Platform and the basin evolved from an epicontinental to a gulf-like pericontinental sea. During the Silurian, the influx of the fine-clastic material progressed from the Scandinavian Caledonides and it gradually infilled the starved depression.

During the Ordovician and Silurian, drastical climatic changes took place. The Baltica Craton drifted from the southern high latitudes to the tropical realm (see Scotese & McKerrow 1990, Torsvik et al. 1992, etc.). It induced the growth of the sedimentation rate of carbonates and development of such types of deposits which are characteristic of the arid and tropical climate (e.g. sedimentary dolostones, pelletal and oolitic deposits, coral-stromatoporoid reefs, etc.). These characteristics were totally lacking during the Early and Middle Ordovician when the Baltic Basin was situated definitely in the temperate climate zone (Jaanusson 1973a). The first signs of warm climate (appearance of tabulate corals, stromatoporoids and reefs) became evident in the Middle Caradoc (Oandu Age), but it was not until the very end of the Ordovician (Porkuni Age) that they got prevalence.

In a sedimentary basin, the facies distribution may be generalized by means of facies models presenting a lateral succession of facies along the reconstructed bathymetric profile. According to the facies model worked out by Nestor and Einasto (1977, etc.), five main facies belts can be differentiated in the Baltic Silurian Basin: tidal flat/lagoonal, shoal, open shelf, transitional (basin slope) and a basin depression (Table 27). The first three facies belts formed a carbonate shelf or carbonate platform and the latter two - a deeper pericratonic basin with fine-clastic deposits.

Later on, Einasto (1986) elaborated two modifications of the basic Silurian facies model: one for the periods of weak and another for the periods of intense supply of terrigenous material. In the first case, the pure lime muds were widespread all over the open shelf and transitional belt; in the second case, argillaceous muds diluted carbonate sedimentation on the open shelf and periodically even in the shoal belt. It also became evident that the Silurian models are not applicable to the Early and Middle Ordovician which climatically and tectonically differed considerably from the Late Ordovician and Silurian. During the Early and Middle Ordovician, sediments accumulated under moderate climatic conditions. The main source of the carbonate component was skeletal material and its production was extremely slow. Therefore, the amount of loose skeletal material on the sea floor was limited. On the other hand, the effect of waves on the bottom was weak, because they subsided gradually in the shallow epeiric sea and lost their energy before reaching the shore. Due to these two circumstances, presumably there did not form notable accumulations of winnowed skeletal sands (grainstones) characteristic of the shoal belt of the basic Silurian facies model. Typical lagoonal carbonate sediments were also lacking, because evaporation was very weak. The position of the lagoonal and shoal facies belts was probably occupied by the belt of nondeposition, represented by a hardground (discontinuity surface). As mentioned above, in the Early and Middle Ordovician, a deeper, sediment-starved axial depression was not developed in the Baltic Basin, and the basin floor formed an evenly and weakly tilted ramp. Accordingly, only three facies belts: a nondepositional belt, an upper and a lower ramp were distinguished in the Early and Middle Ordovician facies models of the Baltic Basin (Nestor 1990a, Kõrts et al. 1991). In the development of the Baltic post-Tremadoc Ordovician and Silurian basin, five stages were differentiated (Fig. 139):

1. The transgression stage (Arenig - Llanvirn). In the marginal part of the basin the deposition was very slow and with many breaks. Within the limits of the upper ramp, micritic skeletal calcarenites accumulated, sometimes containing silt, scattered pebbles and abundant glauconite grains, goethite and francolite ooids and impregnated hardgrounds. On the lower ramp, coinciding with the central confacies belt, mainly red-coloured calcareous-argillaceous deposits (argillaceous limestones and marls) were formed. The deposits of the lower ramp were 2 to 10 times as thick as those in the coeval upper ramp.

2. The unification stage (Llandeilo - Early Caradoc). Along the whole extent of the bathymetric profile (ramp) grey calcareous - argillaceous sediments (argillaceous limestones and marls) accumulated although the general trend towards an increasing clay component and decreasing content of bioclasts in the offshore direction remained. Deposits of the upper ramp contained an admixture of light-brown kukersite kerogen and also pure kukersite interlayers. Interbeds of volcanic ash (metabentonite) were also characteristic of that stage of the basin evolution.

3. The differentiation stage (Late Caradoc - Middle Llandovery). A deeper axial depression of the basin was formed and facies zonation, typical of the Silurian developed. The supply of the basin with clastic material was periodically extremely low, and comparatively pure calcareous muds were deposited during those periods on the open shelf and in the transitional belt, while condensed dark graptolitic shales formed in the basin depression. Also an agitated-water shoal belt with pelletal and skeletal sands and a lagoonal belt with dolomicritic deposits (dolomitic marls, argillaceous dolomites) developed. During this stage, the periods of low terrigenous mud influx cyclically alternated with the phases of more intensive supply with the fine clastic material. In the latter case, deposition followed the basic Silurian facies model characteristic of the stage of stabilization of the basin development.

4. The stabilization stage (Late Llandovery - Early Ludlow). Moderate influx of the fine-clastic material, which partly deposited in the lagoonal and open shelf belts but mostly in the transition belt, resulting in side-filling of the “starved” depression with deposit wedges (lenses) and gradual progradation of the carbonate shelf margin was characteristic of this stage. Facies zonation was clear and the basic Silurian facies model (Table 27, Fig. 139d) reflects the situation during that stage of evolution. However, the facies belts were not equally developed during the different phases of the basin development. The deeper-water facies were widespread during the transgressive phases (the end of the Llandovery and the beginning of the Wenlock), while shallow, marginal-marine facies were best developed during the regressive phases (the end of the Wenlock and the beginning of the Ludlow) of the basin development.

5. The infilling stage (Late Ludlow - Pridoli). Intense influx of terrigenous material filled the basin depression and also diluted carbonate sedimentation on the open shelf where bioclastic limestones were replaced by bioclastic marls. Even in the shoal belt skeletal sands interlayered with marls.

Generally, certain facies models were typical only of certain stages of basin evolution, however, in some cases different types of sedimentation could also alternate. For example, during the Late Caradoc to the Middle Landovery, the periods of low and moderate supply with terrigenous clastic material alternated cyclically and sedimentation followed the models charcateristic of the differentiation and stabilization stages, respectively.

In the following, the evolution history of the basin will be presented with an emphasis on the situation in Estonia. A distinct cyclicity in the basin evolution was summarized by Einasto (1995) and is figured in the present work (Fig. 140). Nine high-rank macrocycles of eustatic origin have been established in the evolution of the Baltic Ordovician and Silurian Basin. They are separated from each other by subregional sedimentation breaks of different duration, increasing in onshore direction. Lower-rank, meso- and microcycles have also been distinguished which, besides the fluctuation of the sea level, were also induced by climatic cyclicity and pulsatory supply with terrigenous material. A meridional cross section of the Ordovician and Silurian rocks across Hiiumaa and Saaremaa islands and the Kuramaa Peninsula (Fig. 141) shows spatial and temporal relations of the main facies commented in the text below.


Transgression stage of development

This, the first stage in the evolution of the Baltic carbonate sedimentation basin coincided roughly with the Arenig and Llanvirn (s.s.), i.e. with the time interval from the Hunneberg to Lasnamägi ages. However, the upper limit of the stage was quite transitional and felt in the middle of a high-rank eustatic macrocycle (see Fig. 140). It is defined by the disappearance of goethide and francolite ooids in the marginal confacies belt and marine red-beds in the central confacies belt, by the appearance of kukersite kerogen and by an abrupt increase in the content of terrigenous clay material, characteristic of the next - unification stage of development.

At the beginning of the transgression, the general structural setting of the post-Tremadoc Ordovician deposition basin developed. Two main facies (sedimentation) belts: an upper (inner) and a lower (outer) ramp were formed within the Baltic Basin sensu stricto (Fig. 139a). The former coincides with the marginal or Estonian-Lithuanian Confacies Belt by Männil (1966), the latter corresponds to the central, Swedish-Latvian Confacies Belt (Männil 1966) or the Livonian Tongue by Jaanusson (1973a).

The marginal, upper ramp deposits were represented by comparatively pure bioclastic carbonate muds with an admixture of glauconite in the lower and goethite or francolite ooids (pseudo-ooids) in the upper part of the sequence. A lot of discontinuity surfaces (hardgrounds), corresponding to shorter or longer sedimentation breaks, but also scattered lithoclasts and skeletal grainstone interlayers are characteristic of the deposits of the most peripheral part of the basin. In the deeper-water, central confacies belt (Livonian Tongue) much thicker deposits of purple calcareous-argillaceous muds with rare discontinuities accumulated. These were treated as the deposits of lower (outer) ramp. Outside the Baltic area, on the western margin of the East-European Platform (Scanian Confacies Belt), the purple calcareous argillaceous muds were replaced by the deepest-water black kerogenous graptolitic muds (see Männil 1966, Figs. 50-55).

The stage of transgression may be subdivided into two substages of development. The first substage roughly corresponds to the Arenig (Hunneberg to Volkhov ages) and represents a full eustatic macrocycle (Fig. 140). The second, Llanvirn (s.s.) substage (Kunda to Lasnamägi ages) embraced only half of the next – Llanvirn - earliest Caradoc macrocycle.

The Arenig macrocycle (substage) began with a well-known world-wide transgression. In the marginal confacies belt it started with the deposition of a very thin band of glauconitic sands and silts of the Leetse Formation (Hunneberg and Billingen regional stages). In the central confacies belt (Livonian Tongue) comparatively thick deposit of purple to greenish-grey terrigenous muds (Zebre Formation) accumulated. In the course of progressive transgression, the bulk of the carbonate component gradually increased and the area of all facies belts extended remarkably.

During the Volkhov Age, glauconitic, bioclastic calcareous muds (Toila Formation) accumulated in the marginal part of the basin. There were numerous interruptions in sedimentation marked by impregnated hardgrounds. At the same time, in the central belt of the basin the purple calcareous-terrigenous muds (Kriukai Formation) expanded their area. A wide transitional zone, including all of southern Estonia, with variegated bioclastic calcareous-argillaceous muds was developed between the marginal and central confacies belts (Fig. 142).

The Arenig macrocycle ended with an extensive eustatic regression as a result of which the sea receded from the whole upper ramp area and a karstified erosional surface with deep corroded pockets and furrows developed on top of the Volkhov Stage. At the time of the regression maximum, in the central confacies belt the purple calcareous argillaceous muds were replaced by grey calcareous muds (Šakina Formation).

The second, Llanvirn substage of basin development roughly corresponded to the Llanvirn sensu stricto (Kunda, Aseri and Lasnamägi ages). In the central belt of the basin, the deposition of purple and variegated argillaceous-calcareous muds (Baldone, Segerstad, Stirnas formations) continued. The deposits being represented by argillaceous and biomicritic limestones became only more calcareous than the earlier ones. In the marginal part of the basin, deposition of bioclastic calcareous muds was ongoing, but instead of glauconite, goethite and francolite ooids were distributed on certain levels of the Loobu, Kandle and Väo formations. The Early Ordovician transgression reached its maximum spatial extent during the Kunda Age (Saadre 1992). However, a slow deepening of the basin, especially in its peripheral part, continued until the end of the stage of transgression and during the next, unification stage as well (Fig. 140).


Unification stage of development

This stage in the basin development covered the Llandeilo and Early Caradoc interval from the Uhaku to Keila ages. It was a period of relative tectonic and eustatic stillstand on the East-European Platform which terminated the general transgressive phase in the development of the Ordovician basin. The most characteristic features of the stage were remarkable intensification of the influx of fine terrigenous material, prevailingly from the northeast, and considerable equilization of environmental conditions in the marginal and central parts of the Baltic Basin (s.s.). All over the Baltic area mainly bioclastic argillaceous-calcareous muds accumulated. Besides lateral variation in the ratio of argillaceous and calcareous components, also a distinct lower rank cyclicity, expressed in alternation of more and less argillaceous deposits, was apparent. The latter reflects the pulsatory supply of the basin with fine terrigenous material rather than the sea-level fluctuation.

Characteristic of the earlier part of the unification stage (Uhaku and Kukruse ages) was accumulation of light-brown organic matter - kukersine which formed kukersite interlayers and dispersed admixture in calcareous sediments. During the second half of the unification stage (Haljala and Keila ages), numerous volcanic ash (metabentonite) intercalations formed giving evidence of growing volcanic activity in the adjacent Iapetus Ocean, perhaps connected with its transition from the opening phase to the closing state. It is remarkable that the same phenomenon, i.e. the presence of numerous metabentonite interlayers, is also characteristic of the highly argillaceous Late Llandovery and Early Wenlock sediments which formed during the Silurian transgression maximum. The above-mentioned two parts may be treated as substages of the unification stage.

The Llandeilo - earliest Caradoc substage of development (Uhaku and Kukruse ages) was a specific period of accumulation of kerogenous, kukersine-bearing deposits in northeastern Estonia and in its vicinity (Fig. 143). The depocenter of these deposits was situated near the present-day outcrop belt. It is supposed (Männil et al. 1986, Bauert 1989, Kõrts et al. 1991) that most of the kukersine was allochtonous and derived from algal-microbial mats, covering the flat sea-shore (tidal flat), from where it was transported into a broad near-shore sinking of the sedimentation basin. In rest of the marginal confacies belt and transition area, monotonous bioclastic argillaceous-calcareous muds (Kõrgekallas and Dreimani formations) were deposited (Fig.143). In the sediments of a wide transition zone (Dreimani Formation), the skeletal particles are heavily pyritized and scattered goethite ooids occur. In the axial part of the Livonian Tongue, the bioclastic muds became more argillaceous, but other characteristics remained almost the same. In northern Estonia, the Llandeilo - earliest Caradoc evolution substage terminated with a short erosional sedimentation break at the end of the Kukruse Age (Fig. 140).

The Early - Middle Caradoc substage of development (Haljala and Keila ages) was a complete eustatic macrocycle with a short transgressive phase at the beginning, longer stillstand period in the middle and drastical shallowing and regression at the end. It began in northern Estonia with the deposition of a thin basal layer of calcareous silt (Kisuvere Member of the Kahula Formation). The silt probably derived from the Kärdla impact crater (Põlma 1982) and was transported eastwards by longshore drift. The macrocycle in consideration is characterised by the occurrence of the most argillaceous sediments in the whole Ordovician sequence and by the presence of numerous metabentonite interlayers. Only at the beginning (Tatruse Formation) and at the end of the macrocycle (Pääsküla and Saue members of the Kahula Formation), purer bioclastic calcareous sediments occurred in northern Estonia, in the marginal confacies belt. During the middle part of the macrocycle, bioclastic calcareous-argillaceous muds accumulated in the marginal confacies belt, while in the central belt highly argillaceous bioclastic muds (Adze Formation) deposited.

The macrocycle and the whole unification stage ended with remarkable shallowing and regression. In northwestern Estonia, in the Vasalemma area at late Keila and early Oandu times, a reef complex with bryozoan-microbial carbonate mounds and intermound pelmatozoan grainstones was formed in the shallow-water high-energy environment. Elsewhere in northern and central Estonia, a probable hiatus existed between the Keila and Oandu stages. At the same time, highly argillaceous muds of the Blidene Formation (marls and mudstones) formed in the axial part of the Livonian Tongue (Fig. 144). In the wide transitional zone, they were replaced by slightly more calcareous muds which in their most peripheral part in southwestern Estonia, easternmost Latvia and Lithuania contain fine siliciclastic material (see Ainsaar 1995). The latter deposited at the time of regression maximum when the sea supposedly withdrew from the whole upper ramp area, including northern Estonia.


Differentiation stage of development

This stage of evolution corresponds to the Late Ordovician and earliest Silurian covering the time interval from the Oandu Age up to the end of the Raikküla Age. The beginning of the evolution stage roughly coincided with the general tectonic inversion in the western part of the East-European Platform, i.e. transition from the transgressive to regressive phase in basin development. On the other hand, it also reflected transition from humid, moderate climatic conditions to arid subtropical-tropical climatic conditions. Evidence is derived from the appearance of the first corals and stromatoporoids, beginning of the formation of organic build-ups, pelletal, aragonitic sediments, extensive accumulation of pure lime muds, etc.

The most characteristic features of this evolution stage included: 1) development of considerable lateral lithological differentiation of sediments, 2) rapid increase in the sedimentation rate and thickness of deposits, 3) formation of condensed deposits of dark kerogenous muds (shales) in the central confacies belt of the basin, cyclically interbedded with thicker deposits of greenish-grey to purple muds (mudstones and marls), 4) cyclical alternation of deposits of pure lime muds (micritic limestones) with bodies of argillaceous lime muds (marls, argillaceous limestones) in the marginal part of the basin and transitional zone. The former three peculiarities were probably connected with transformation of the western margin of the Baltica Continent from passive to active state due to the beginning of the gradual closure of the Iapetus Ocean. It caused different tectonic movements in the Baltic Syneclise and its surroundings; as a result, deeper intracratonic depression began to form within its limits. A possible reason for cyclical alternation of the deposition of pure and argillaceous lime muds may be an interchange of the arid and humid climate periods characteristic of the big glacial epochs in the Earth history, including the Late Ordovician - Early Silurian glacial epoch. The arid periods caused a relatively low influx of terrigenous material and, consequently, deposition of purer lime muds, while during the humid periods more argillaceous muds were deposited.

In the evolution of the Baltic Basin, the cyclically alternating periods of low and high influx of terrigenous material are well recognizable. During the low influx (“limy”) phases rather pure, light lime muds (micritic or calcilutitic limestones) deposited on the open shelf and in the transitional facies belt (see Fig. 139c). During the high influx (“clayey”) phases, on the open shelf bioclastic argillaceous-calcareous muds (nodular argillaceous biomicritic limestones) were deposited, while in the transitional belt prevailingly terrigenous muds (marlstones, mudstones) accumulated (Fig. 139d). Nine of such cycles of alternating low- and high-influx phases have been recognized in the development of the Baltic Basin during Late Ordovician - earliest Silurian times.

The low-influx (“limy”) phases were: 1) Rakvere (Rägavere Formation), 2) late Nabala (Saunja Formation), 3) early Pirgu (Moe and Svėdasai formations), 4) mid-Pirgu (Oostriku, Baltinava, Parovėja formations), ?5) latest Pirgu (Taučionys Formation), 6) early Juuru (Koigi, Ruja, Sturi members), 7) early Raikküla (Järva-Jaani beds, Slītere Member), 8) mid-Raikküla (Jõgeva beds, l. pt., Ikla Member), 9) late Raikküla (Mõhküla beds, Staicele Member).

The high-influx (“clayey”) phases were correspondingly: 1) Oandu (Hirmuse, Lukštai formations), 2) early Nabala (Saunja, Mõntu formations), 3) Vormsi (Kõrgessaare, Tudulinna, Meilūnai formations), 4) early-mid Pirgu (Adila, Halliku, Ukmerge formations), 5) mid-late Pirgu (Ludza, Kuiļi formations), 6) Porkuni (Ärina, Kuldiga, Saldus formations), 7) middle and late Juuru (Varbola, Tamsalu formations, Rozeni Member), 8) early-mid Raikküla (Vändra beds, Kolka Member), 9) mid-late Raikküla (Jõgeva beds, u. pt., Lemme Member).

In most cases, the distribution of the pure lime muds was restricted to the open shelf and transitional facies belt (see Fig. 138c). Basinwards the lime muds were replaced by dark brown kerogenous graptolitic muds (Mossen Formation of the Rakvere Stage, Dobele Formation of the Raikküla Stage) or by purple calcareous-argillaceous muds (Jonstorp Formation of the Pirgu Stage, Remte Formation of the Raikküla Stage). However, in some cases light lime muds covered also extensive areas in the central part of the basin (Saunja Formation of the Nabala Stage, Parovėja Formation of the Pirgu Stage, Stūri Member of the Juuru Stage); more deeper-water deposits are not represented in the Baltic area. Lateral transition of the open shelf lime muds into the deposits of the near-shore agitated-water shoal belt has been established only in the Raikküla Age (Figs. 144 and 66). In this case, they were gradually replaced by fine-grained skeletal-pelletal deposits, in places containing numerous corals and stromatoporoids.

The deposition of the pure lime muds probably proceeded under somewhat specific hydrochemical conditions. The deposits of the micritic limestones (calcilutites) contain extremely sparse organic remains. Most frequently, the Late Ordovician micritic limestones comprise skeleton particles of specific dasycladacean algae Cyclocrinites, Vermiporella, etc. The micritic limestones of the Raikküla Stage often contain detritus of problematic dendroid graptolites. The presence of such specific biotic elements suggests that a certain generative role of the biochemical factor in the formation of pure lime muds is not excluded. The micritic limestones of the Pirgu Stage, especially those of the Moe Formation, are richest in skeleton particles of Vermiporella and other dasycladaceans. The latter and the adjacent Tootsi Formation also contain specific carbonate mounds (Hoitberg, Niiby, Ruunavere, Kaugatuma, Paatsalu, Võhma) which are analogous to the well-known Boda mounds in Central Sweden (see Nestor 1995b).

The differentiation stage of the basin development consisted of two similar bathymetric macrocycles, one corresponding to the Late Caradoc - Ashgill from the Oandu to the Porkuni stages, and another to the Early-Middle Llandovery from the Juuru to the Raikküla stages. These macrocycles began with a slow, gradual deepening of the basin and finished with a rapid shallowing, regression and, in places, with intensive erosion of the older deposits.

The Late Caradoc-Ashgill macrocycle began with a short transgressive episode in the Oandu Age when for the first time in the post-Tremadoc history of the basin development typical anoxic depression facies - dark-brown kerogenous graptolitic muds (Mossen Formation) were formed in the Baltic Syneclise. At the slope of the Belarussian-Mazurian Anteclise the Oandu deposits transgressively overlap the older strata and in northern Estonia they also disconformably overlie the Keila deposits.

After a rapid initial deepening during the Oandu Age, there followed a relative stabilization and levelling of sedimentation conditions culminating during late Nabala time when monotonous pure lime muds (Saunja Formation) expanded over the whole East Baltic area. A new deepening impulse and bathymetric differentiation followed in the Vormsi Age when deposition of dark kerogenous muds with graptolites (Fjäcka Formation) was restored in the central depression of the Baltic Syneclise (Fig. 145).

The Pirgu Age was a variable period in the basin evolution with two distinct episodes of accumulation of pure lime muds. A general regressive trend of basin development is recognizable in the Pirgu Age. During the first, early Pirgu phase of deposition of lime muds (Moe Formation) purple terrigenous muds (Jonstorp Formation) were deposited in the central confacies belt, while during the second, mid-Pirgu phase of lime mud deposition (Oostriku, Baltinava, Parovėja formations), monotonous pure lime muds spread all over the Baltic Basin. The Pirgu Age ended with an obvious sedimentation break of different duration in different places. Extensive local sedimentation gaps with considerable erosion of lower-lying strata have been established in the areas of the Lower Nemunas and Irbe structural elevations (see Männil 1966, Kaljo et al. 1988b). In these and some other places, the Porkuni rocks rest disconformably on the Pirgu strata showing that the Porkuni Age began with a certain transgression event. During the first half of the Porkuni Age a diverse complex of shallow, agitated-water carbonate sediments (skeletal sand and silt, coral-stromatoporoid bioherms, bioclastic calcareous muds of the Ärina Formation) formed in the marginal confacies belt. In the central belt bioclastic calcareous-argillaceous muds (Kuldiga Formation) deposited at the same time. An abrupt shallowing took place in the middle of the Porkuni Age and deposition of shallow-water calcarenitic sediments (bioclastic, oolitic, lithoclastic sand, silt and gravel) with remarkable admixture of siliciclastic silt and sand shifted into the central confacies belt, forming high-energy shoal deposits of the Salduse Formation (Fig. 146). In the peripheral part of the Baltic Basin, including northern Estonia, subaerial conditions existed at that time. In the mid-Estonian transitional zone some 15-to-30-m-deep erosion channels (Tootsi, Jõgeva, Ruskavere) were recently discovered (Perens 1995, Ainsaar 1995). The end-Ordovician drastical shallowing event (or events) in the Baltic Basin was definitely connected with the global glacio-eustatic drop of the ocean level (Kaljo et al. 1991), perhaps combined with certain tectonical upheaval, especially at the end of the Pirgu Age.

The Early-Middle Llandovery macrocycle began with a glacio-eustatic rise of the sea level and deposition of pure lime muds (Stūri, Rūja, Koigi members) on wide areas of central and eastern East Baltic. Only in the deepest-water residual depression (South Estonia - Kurzeme) the lime muds were replaced by calcareous-argillaceous muds (Õhne Formation). During the Juuru Age, a bathymetric differentiation and development of regular facies zonation, interrupted by the end-Ordovician hiatus, denudation and levelling of the sea floor topography, were gradually restored, but it was not until mid-Raikküla time (Ikla/Jõgeva beds) that a deep-water central depression with dark kerogenous graptolitic muds was finally restored and since then, until middle Ludlow time, it functioned as a sediment-starved depression where deposition rate could not keep pace with the subsidence of the sea floor. A full set of five main facies belts (Table 27) was completely established and a shelf-type sedimentation finally replaced the ramp-type sedimentation which prevailed during most of the Ordovician. The most characteristic feature of the shelf-type sedimentation was formation of thick deposit wedges in the transitional facies belt which led to the side-filling of the basin depression and gradual progradation of the carbonate shelf edge. A gradual side-filling effect is well visible in the cross-section of the Llandovery rocks in western Estonia (Figs. 66, 141).

The Early-Middle Llandovery macrocycle ended with extensive local sedimentation breaks and denudation of earlier deposited strata in the marginal parts of the basin. One area of denudation was situated in western Estonia where the north-westwards increasing erosional hiatus cut the older strata down to the Järva-Jaani beds. Still more remarkable erosional break was developed on the opposite flank of the Baltic Basin, in eastern Lithuania, where the Early and Middle Llandovery deposits were subject to denudation all over the carbonate shelf as far as the eastern limit of the Baltic Syneclise with depression facies of dark graptolitic shales of the Dobele Formation (Fig. 147). The local (subregional) character of the nondeposition and changeable extent of deposition breaks suggest the tectonic nature of the upheaval, most probably induced by the beginning of the collision of the Laurentia and Baltica continents.


Stabilization stage of development

The stabilization stage of basin evolution embraced the main part of the history of the Baltic Silurian Basin from the beginning of the Late Llandovery (beginning of the Adavere Age) up to the end of the Middle Ludlow (end of the Paadla Age). The most characteristic features of the stabilization stage of evolution were: 1) a moderate influx of the fine terrigenous material, 2) the presence of a comparatively deep, starved axial basin depression with continuous sedimentation of dark-grey argillaceous deposits (mudstones and shales) with graptolites, 3) a perfect shelf-type facies zonation beginning from marginal-marine, lagoonal dolomitic muds and ending with dark -grey graptolitic muds in the depression belt (see Table 27 and Fig. 139d). At the beginning of the stabilization stage, during the transgressive phase of the basin development, deeper-water facies had a wide distribution. During the regressive phase, in the second half of the stabilization stage, shallow near-shore facies were well developed.

In the marginal part of the basin, a general regressive trend of evolution was characteristic to the stabilization stage and it led to the transformation of the Baltic Basin into a gulf-like pericontinental sea. On the background of the general shallowing trend, smaller-scale sea level fluctuations were characteristic of the basin evolution. The stabilization stage may be subdivided into two eustatic macrocycles of unequal duration and completeness. The first, perfect deepening-shallowing macrocycle corresponds to the Late Llandovery - Middle Wenlock (Adavere to Jaagarahu ages) and the second, uncomplete one to the Late Wenlock and Early Ludlow (Rootsiküla and Paadla ages).

The Late Llandovery - Middle Wenlock macrocycle began in the Baltic Basin with a rather long transgressive (deepening) phase, lasting from the beginning of the Adavere Age up to the mid-Jaani time. The transgression proceeded in two steps (Nestor 1972). It began with an extensive deposition of bioclastic argillaceous-calcareous muds with coquinite accumulations of the brachiopod Pentamerus oblongus (Rumba Formation) in the the open shelf facies belt, transgressively overlying different strata of the Raikküla Age. In the central depression of the basin, condensed dark-brown kerogenous graptolitic muds of the Dobele Formation continued to deposit. During the next step of deepening, corresponding to late Adavere (Velise) time, argillaceous sediments of the depression (Jurmala Formation) and transitional facies belts (Velise and Švenčionys formations) covered the whole East Baltic area. The transgression was especially extensive in eastern Lithuania where marly deposits of the Švenčionys Formation disconformably overlie the Ordovician strata. The Late Llandovery - Early Wenlock transgression expanded even into the Moscow Syneclise (Resheniya… 1987). During the late Adavere time, purple terrigenous muds accumulated in the western part of the Baltic Basin (Kurzeme, Sõrve, Gotland areas).

Deposition of the deeper-water, highly argillaceous sediments continued also at the beginning of the Wenlock, Jaani Age (Tõlla beds of the Rīga Formation, Mustjala Member of the Jaani Formation, Sutkai beds of the Paprieniai Formation). The highly argillaceous sediments of the Late Llandovery and the earliest Wenlock contain numerous thin metabentonite interlayers. It leads to the supposition that the Late Llandovery – Early Wenlock sea-level high-stand was probably induced by the acceleration of the sea-floor spreading accompanied by intensification of volcanic activity in the Iapetus Ocean.

In the middle of the Jaani Age, an abrupt shallowing developed in the peripheral part of the Baltic Basin. Reefs and skeletal sand bar deposits of the shoal facies belt began to form in the Gotland area (Högklint Formation), northwestern Saaremaa (Ninase Member of the Jaani Formation) and eastern Lithuania (Jačionys Formation). However, in the basin depression the facies of dark-grey graptolitic muds even expanded its area at that time (Fig. 148). As a result, extremely rapid early-mid Wenlock regression along the perimeter of the basin finally transformed the Baltic Basin s.s. into a gulf-like pericratonic sea (the “Baltic Gulf”). The most drastic upheaval and regression took place in the Scandinavian part of the sea due to the progressing rise of the Caledonides.

During the Jaagarahu Age, the gradual shallowing continued at the margins of the basin, interrupted by short deepening episodes. This resulted in a cyclical alternation of high-energy shoal deposits (winnowed skeletal-pelletal sand and silt, coral-stromatoporoid banks and reefs) with different lagoonal-littoral dolomitic muds (Eurypterus- and bioturbated pattern-dolomites). The Eurypterus-dolomites probably formed in a brackish-water environment (Einasto 1968). Three of such shallowing-up mesocyclithes have been recognized in the Jaagarahu Formation and are treated as the Vilsandi, Maasi and Tagavere beds. Basinwards these intercalating shoal and lagoonal deposits of the Jaagarahu Formation were successively replaced by the bioclastic argillaceous-calcareous muds (Riksu Formation), greenish-grey calcareous-argillaceous muds (Jamaja Formation) and finally by dark grey terrigenous muds with graptolites (Rīga Formation). During late Jaagarahu time, a long sedimentation break and erosion of the earlier deposits took place all over the shelf plateau (Nestor & Nestor 1991). At the moment of the maximum shallowing, at the end of the Jaagarahu Age, a band of thinly interbedded marls and limestones (Ančia Member) was formed in the central depression of the basin, marking the end of the Late Llandovery - Middle Wenlock macrocycle in the basin development.

The Late Wenlock - Middle Ludlow macrocycle corresponds to the Rootsiküla and Paadla ages. It is represented by the cyclical alternation of winnowed skeletal-pelletal grainstones with sedimentary argillaceous dolostones (Eurypterus-, pattern- and microlaminated dolomites). Unlike the Jaagarahu cycles, the shoal and lagoonal facies of the Late Wenlock - Middle Ludlow mesocycles extended over a very wide area of southwestern Estonia, covering the whole levelled shelf plateau. Southwards these marginal-marine facies were rapidly replaced by the dark-grey graptolitic muds of the Siesartis and Dubysa formations which shows that at that time a rather steep gradient existed between the shelf plateau and basin depression. The open shelf and transitional facies belts were heavily reduced (Fig. 149). It means that a platform-type sedimentation with a very wide belt of marginal-marine facies was typical of the Late Wenlock - Middle-Ludlow macrocycle of basin development.

On the background of the cyclical sea-level fluctuations a faintly expressed regressive-transgressive trend is perceivable in the basin evolution. The break point, i.e. the regression maximum was probably reached by the end of the Vesiku time (mid-Rootsiküla) and further a very slow, gradual transgression followed. It is likely that several sedimentation gaps existed during the time-interval in consideration, but direct evidence is still lacking and only a probable hiatus has been revealed between the Rootsiküla and Paadla stages (Table 8).


Infilling stage of development

It was the final epoch in the evolution of the Ordovician-Silurian carbonate sedimentation basin at the western margin of the East-European Platform which corresponded to the time interval from the Late Ludlow (Kuressaare Age) up to the earliest Devonian (Tilžė Age). An abrupt increase in the supply of the basin with terrigenous material, coming from the Scandinavian Caledonides, was a characteristic feature of the infilling stage. The previous side-filling of the basin depression was changed by the total infilling and shallowing all over the central part of the Baltic Syneclise where the deposition of terrigenous graptolitic muds was now replaced by the accumulation of the greenish-grey calcareous-argillaceous muds with benthic biota. The formation of the graptolitic muds of the depression belt migrated to the platform margin in northeastern Poland where a thick clayey-silty complex of distal turbidites (Siedlce beds) was formed. Even in the shelf area, corresponding to the open shelf and shoal environments, the accumulation of the terrigenous material was so heavy that it diluted the carbonate sedimentation. As a result, the formation of bioclastic marls got dominance all over the shelf area. However, from time to time rather thick (3 to 5 m) deposits of crinoidal grainstones, coquinite banks with Atrypoidea prunum, thickets of rugose and tabulate corals were formed. Such thick deposits of crinoidal grainstones, containing cross-bedding, ripple marks and other signs of the agitated-water environment mark the end of several shallowing-up mesocycles. They are most typically developed in the Kaugatuma Stage where four of such cyclithes (Lower and Upper Äigu beds, Lower and Upper Lõo beds) have been distinguished. A clear facies zonation was characteristic of the moments of the formation of crinoidal grainstones in the shoal belt (Fig. 150), while during the rest of time with intensive influx of clayey component, the lateral differentiation of facies was much poorer.

The infilling stage began with a brief transgressive event during the Kuressaare Age. This Late Silurian transgression reached its maximum at the beginning of the Pridoli (early Kaugatuma time). The transitional facies belt with deposition of calcareous-argillaceous muds (Šilalė beds of the Minija Formation) was very wide and extended from southwestern Latvia to northwestern Poland (Fig. 148). In onshore direction it was gradually followed by open shelf biomicritic marls, coarse-grained crinoidal gravel and sand of the shoal belt and lagoonal dolomitic muds. The latter have preserved only in the Lithuanian part of the basin. During the late Kaugatuma and Ohesaare times, the general facies pattern remained the same but all facies belts migrated gradually southwestwards.

By the beginning of the Devonian (Tilžė Age), only a remnant lagoon-like body of water was preserved in northern Kurzeme and south-western Lithuania. It was characterized by a clastic-dominated near-shore belt and argillaceous-dolomitic muds in its offshore part.



Devonian sedimentation basin

A. Kleesment


Estonia belongs to the northwestern part of the Devonian Main Field of the East-European Platform which in the Devonian was situated in the equatorial region on the Laurussia Continent formed at the end of the Silurian. Its development was influenced by both tectonic movements and eustatic sea-level oscillations (Ziegler 1988). In the territory under consideration, epicontinental shallow sea sediments accumulated. In the Early Devonian and at the beginning of the Middle Devonian, this sea had a connection with the ocean in the northeast, later in the east and south-east. The sequence has a very complex cyclic structure indicative of pulsatory nature of the sedimentation process. The regression, which had started at the end of the Silurian, continued at the beginning of the Devonian and alternated with a short-term transgression in the middle of the Early Devonian during the Ķemeri Age. A significant long-term transgression started in the Rēzekne Age at the end of the Emsian and lasted until the end of the Leivu Subage in the second half of the Eifelian when a regression with a duration up to the end of the Givetian started. A new transgression started at the beginning of the Late Devonian when the northernmost marginal part of the sedimentary basin formed and extended to a restricted area in southeastern Estonia (Tihomirov 1967, Kuršs 1992).

At the beginning of the Early Devonian, in the Lochkovian Age, the filling of the relict basin formed during the Silurian regression continued. The deeper axial part of this basin opening to the south-west was located in the Baltic Syneclise. Its northern margin, however, reached the present-day Estonia (Narbutas 1984, Kuršs 1992) where sandy sediments deposited. The region of the Mõniste Uplift served as a continental denudation area (Fig. 151A). The influx of terrigenous material into this basin was mainly from the Scandinavian massif in the north-west. The highly variable mineral composition of the sediments becomes more uniform in a southerly direction.

The sediments of the Lochkovian Tilžė Age occur in a restricted area in Estonia. Palaeontologically, they have been determined from the Laanemetsa and Värska core sections in southeastern Estonia (Sorokin 1981). The finds of thelodont scales in the Ruhnu core, which have redeposited from the Tilžė Stage into the basal beds of the Rēzekne Stage (Mens et al. 1992), show that earlier these sediments were distributed also in southwestern Estonia where they were afterwards subject to denudation. This supports the opinion of Kuršs (1975) and Narbutas (1984) who maintain that in the Tilžė Age the sediments formed as two tongues in the northern wing of the relict basin.

In the second half of the Lochkovian Age, sedimentation was interrupted in the northwestern part of the East-European Platform. The geocratic period, separating the Caledonian and Hercynian tectonic stages, began. In the Baltic area it covered the second half of the Lochkovian and the beginning of the Pragian. This period is characterized by marked denudation in the course of which the sediments in the marginal area of the basin were partly denudated. A new marine flooding invaded Baltic from the southwest. In the second half of the Pragian in the Ķemeri Age, sandy sediments started to accumulate here, while in the deeper part of the basin - the Polish-Lithuanian depression, sandy-silty sediments deposited. The weathering crust which had formed during the hiatus was redeposited. Therefore, Ķemeri sediments contain kaolinite in notable amounts and the sandy fraction has a relatively mature mineral composition. Additional terrigenous material was transported into the basin from the Scandinavian massifs. Possibly, the influx took place along the site of the present Gulf of Riga (Kuršs 1975). In the northern part of the distribution area sandstones are much coarser. The northern margin of the Ķemeri Basin reached Estonia’s territory and nearshore sandy sediments accumulated here (Fig. 151B). Supposedly, the Ķemeri sediments occur in a number of boreholes in southwestern Estonia (Ipiku, Ikla, Tõlla, Abja), although palaeontologically they have not been dated anywhere in Estonia.

The Ķemeri transgression was relatively short. At the end of the Pragian and at the beginning of the Emsian, the whole East-European Platform was governed by continental conditions and the earlier sediments were denudated. A new westerly transgression took place at the end of the Emsian during the Rēzekne Age. The sea flooded a great part of the East-European Platform, the waves reaching as far as the Moscow Syneclise (Kuršs 1992). The northern part of the basin extended to Estonia’s territory leaving behind sandy sediments of nearshore shallow sea. In southeastern Estonia, mostly offshore sandy-silty sediments deposited. Towards the end of the Rēzekne Age the sea became deeper and in southeastern Estonia normal-sea carbonate sediments accumulated. The influx of the terrigenous component was mostly from the northwest, from the Scandinavian massif. The grain-size of sediments and the areal location of zircon-rich belts suggest two possible ways of influx (Fig. 151C). The mineral composition shows that granitic massifs were subject to denudation.

At the beginning of the Pärnu Age, the basin retreated. The selective concentration of garnet possibly indicates that in the first half of the Pärnu Age the coastline stayed for a long time on the Strenči - Valga - Väimela - Värska line (Fig. 152A). In the middle of the Pärnu Age, the marine basin started to extend to the north, north-east and east, over a great part of the East-European Platform. In Estonia, mainly nearshore sandy sediments accumulated, but in the west evidently subwater delta sediments deposited. The influx of terrigenous material from the Scandinavian granite massifs (Fig. 152A) continued. The salinity of the Pärnu basin was presumably somewhat higher which is evidenced by the occurrence of gypsum. Gypsum is particularly characteristic of the areas located father in the east and in the southern part of the Baltic. Owing to intensive freshwater influx, the salinity was normal in the northwestern part of the basin. At the end of the Pärnu Age, beside sandy sediments also carbonate sediments with abundant sandy admixture accumulated. The occurrence of syneresis cracks and pyrite-rich surfaces in carbonate sediments indicates periodical hiata in sedimentation at that time.

The Middle Devonian marine transgression in the East- European Platform reached its maximum in the Narva Age when a shallow basin with carbonate sedimentation formed the major part of the platform. The center of sedimentation was in the Baltic Syneclise where the thickness of sediments reached 150-180 m (Kuršs 1992). The northwestern part of the basin covered the current site of Estonia (Fig. 152B, C). In the Narva Age, three subages in the basin evolution are distinguished: Vadja, Leivu and Kernavė (see Fig. 83).

The basin widened markedly in the Vadja Age when almost the entire Estonian territory was included in the sedimentation area (Fig. 152B). In the northern part of the East- European Platform dolomitic muds with variable clay content were accumulated. In the northwestern part of the basin, the sequence starts with a peculiar landslide breccia or layers of breccia-like domerite. The genesis of breccia is also related to a break in sedimentation, but a more probable reason seems to be an underwater landslide caused by tectonic movements. The occurrence of syneresis cracks and pyritic surfaces suggests that water was shallow and the basin episodically turned dry. The presence of gypsum and caverns, formed due to the leaching out of halite, is indicative of brackish-water conditions at the time of deposition during the Vadja Age. Owing to intensive freshwater influx from the north, the salinity of water in Estonia’s area was somewhat lower. The influx of detrital material from the north-west continued but, additionally, some material was also derived from the Kola massif in the north-east (Fig. 152B, C). As the result of the northeasterly influx across the Ordovician outcrop, the sediments of the Vadja Subage have yielded redeposited Ordovician conodonts and crinoidal stem fragments from a height of up to 10 m above the Ordovician/Devonian contact. High corundum concentration in the region of the Navesti and Narva rivers shows that these watercourses served as the influx channels, but it also evidences of the proximity of the shoreline. The character of sediments (fine alteration of dark-grey silty clays with domerites and dolomites, the occurrence of psammitic material) suggests sedimentation in the tidal belt (Hettinger 1995).

The basin turned shallower and the sea retreated southward at the end of Vadja Subage. The earlier deposited layers were denudated and weathered. This level is marked by a sandstone layer with a mature mineral composition and high roundness of minerals. It is overlain by dolomites or domerites enriched with sandy material. By the beginning of Leivu Subage, the sea had retreated back to the limits of the Baltic Syneclise. Soon a new flooding followed and the basin widened northwards reaching the maximum extent at the end of the Leivu Subage. This basin was shallow, with increased salinity and carbonate sedimentation. Its basic source area was as earlier in the Scandinavian massifs. The main influx of terrigenous material was into the region of the present-day Lake Võrtsjärv depression characterized by a NW-SE facies belt enriched with sandy-silty material. The influx from the north-east continued as well (Fig.152B). Marked alteration of the mineralogical composition, first of all a decrease in garnet and titanite, and an increase in apatite beginning from the second member of the Leivu Substage, but also the changes in the typomorphic varieties of minerals indicate that new massifs were subjected to denudation (Fig. 78). Accumulation of red-coloured sediments at the end of the Leivu Subage refers to a humid type of weathering in the denudation area (Kuršs 1992).

The Kernavė Subage was characterized by a regression trend. As a result of more intensive influx of terrigenous material and fresh water, the basin turned into a waterbody of normal salinity with rich biota where mostly red sandy, in the Baltic Syneclise also normal marine grey carbonate sediments deposited. In the territory of the present-day Estonia, which was situated in the northwestern marginal area of this basin, sandy and variegated clayey-carbonate sediments of nearshore shallow sea accumulated. The offshore facies belt with the accumulation of silty-carbonate sediments embraces only a restricted portion of the southeasternmost part of Estonia. As previously, the influx of terrigenous material was via the present Võrtsjärv Depression and northeastern Estonia (Fig. 152C). Changes in the mineral composition indicate some alteration of the initial massifs (Fig.78).

The shallowing trend in the basin development continued in the Aruküla Age when terrigenous sedimentation came to dominate. The cyclic structure of the sequence suggests frequent sea-level fluctuations against the background of general regression trend. Three main cyclic complexes permit the distinction of three successive levels - the Viljandi, Kureküla and Tarvastu beds, corresponding to definite evolutionary stages (Kleesment 1994, Fig. 85). Supposedly, the sea retreated finally from Estonia during the initial phase of the accumulation of Kureküla and Tarvastu deposits, in which short-term hiata in sedimentation occur. The preserved sediments, however, are mainly of marine genesis. The character of sediments suggests the deepening of the basin to the southeast, while the pattern of garnet concentration refers to a fairly stable position of the shoreline on the Häädemeeste - Abja - Viljandi - Aakre - Slantsy line (Fig. 153A). The influx of terrigenous material was as formerly from the north-west and north-east. The appearance of staurolite among accessory minerals shows that besides granitic massifs also metamorphic rocks were subject to denudation (Fig. 78). A general increase in the maturity of the mineral composition and the growth of the degree of roundness of grains upwards (Kleesment 1994) confirm that in the Aruküla Age the rate of redeposition was high.

The slow retreat of the marine basin, interrupted by temporary transgressions, continued in the Burtnieki Age. Against the background of the general shallowing trend, three regressive-transgressive stages can be distinguished in the sediments of the Härma, Koorküla and Abava beds (Kleesment 1995, Fig. 85). The regressive stages were characterized by short-term breaks in sedimentation. In Estonia, mainly near-shore facies and delta sediments deposited. The shoreline, marked by garnet concentration, shifted to the south-east compared with its location during the Aruküla Age (Fig. 153 A, B). Underwater delta sediments have been identified in the Joosu quarry (Kuršs 1992). Source terrigenous material came as before from the Scandinavian massifs where the area of denudation had reached metamorphic rocks. Evidence is derived from the higher proportion of staurolite and kyanite among accessory minerals (Fig. 78). Redeposition processes came to play an ever increasing role as suggest the growth of the quartz, zircon and rutile content and an increase in the degree of grain roundness upward the section.

The pulsatory retreat of the marine basin and partial redeposition of older sediments continued in the Gauja Age and were accompanied by a constant influx of fresh detrital material from the Scandinavian metamorphic massifs. Concurrently with the continuing retreat of the sea and probable humidification of climate (Kuršs 1992), the processes of subaeral weathering were intensive in the second half of the Gauja Age. The sediments, formed at that time, are characterized by a high degree of maturity which is revealed in the high concentration of quartz, zircon, tourmaline and rutile, and large amounts of kaolinite among clay minerals. In Estonia, nearshore marine sandy sediments, in the south-east sandy-silty sediments, deposited in the Gauja Age when a great part of the territory was under delta (Kuršs 1992, Fig. 153C).

In the Amata Age, the palaeogeographical situation remained unchanged with respect to the source area and the general configuration of the marine basin. In Estonia, sandy and sandy-silty sediments of nearshore shallow sea accumulated (Fig. 153D). The scantiness of coarse terrigenous material, the lack of fresh minerals and the increasing roundness of grains suggest that the Scandinavian massifs were peneplaned and mostly older sediments were redeposited. Irregular inclination of cross-bedded series is indicative of complicated sedimentation conditions (Kleesment 1995).

At the end of the Amata Age, the link with the ocean in the south-west broke. A new, Frasne transgression started from the Moscow Syneclise in the east. The northwestern point of the sea, which covered the middle part of the East-European Platform and was characterized by Pļaviņas carbonate sedimentation, reached southeastern Estonia. In general, the salinity of the basin somewhat increased, but at the current site of Estonia it was normal due to freshwater influx (Sorokin 1978). Sammet (1971), basing on the presence of gypsum-rich deposits east of Estonia, has treated the sediments of this area as lagoonal. As earlier, the influx of terrigenous material was from the north.



Evolution of life during Vendian – Devonian


H. Nestor & E. Mark-Kurik


The sedimentary cover of Estonia was formed during the first half of the Phanerozoic and records, therefore, only more primitive forms of life: marine algae, cyanophytes, invertebrates, the earliest vertebrates and vascular plants. Fossils are very unevenly distributed throughout the sequence in which, besides the strata extremely rich in fossils rather extensive stratigraphical gaps and almost barren rocks occur.

The fossils are extremely rare in the Vendian and Cambrian clastic deposits, accumulated on southern high latitudes (Fig. 138). They contain neither elements of the famous Ediacara fauna nor diverse Cambrian assemblage of archaeocyathids which obviously settled only the equatorial climatic zone. Acritarchs and problematic thallus-like organical remains - Vendotaenites are the only fossils found from the Vendian strata of Estonia. Such poor and specific content might be explained by freshened-water conditions which probably existed during Late Vendian Kotlin time.

The very sparse Lower Cambrian fossil assemblage contains single representatives of probable annelids (Platysolenites), pogonophores (Sabellidites), gastropods (Aldanella), problematic phylum Agmata (Volborthella), trilobites (Schmidtiellus) (Photo 41:1), jellyfishes (Medusites), phosphatic-shelled brachiopods from the class Paterina (Paterina, Mickwitzia -Photo 39:1), monoplacophores (Scenella), ichnofossils (Skolithos, Diplocraterion) indicating a rather high diversity of phyla comparing with the Vendian biota. From the Upper Cambrian (Ülgase, Tsitre, lowermost Kallavere formations) representatives of the class Lingulata (Ungula, Angulotreta, Schmidtites, etc.) and the earliest conodonts (Westergaardodina, Furnishina, Cordylodus, etc.) have been identified.

In the earliest Ordovician (main part of the Pakerort Age), the first possible bryozoan (Marcusodictyon) and dendroid graptolites (Rhabdinopora) appeared, belonging to the earliest representatives of these groups. The first chitinozoans date from the Upper Tremadoc (Varangu Stage). In the Arenig, during the Billingen Age, articulate brachiopods appeared in the Baltic area in the Volkhov Age, cephalopods and cystoids were added. These were just brachiopods, trilobites and cephalopods that played a leading role in the latest Early Ordovician faunas in Estonia.

During the Middle Ordovician, a very rich assemblage of brachiopods, ostracodes, trilobites, bryozoans, cystoids occurred in the area under consideration. This assemblage contained a number of endemic taxa as many of clitambonaceans among brachiopods, asaphids and cheirurids among trilobites etc. It allows to speak about the special North-European or Baltoscandian faunal province (Jaanusson 1979). This was definitely a temperate-climate fauna as it totally lacked reef-building corals and stromatoporoids, rather wide-spread on some other cratons (North America, Siberia, Kazakhstan, North China, Australia) at that time. The first and very primitive solitary rugose corals (Primitophyllum, Lambelasma), as obviously less dependent on climate, appeared during the Haljala Age. The earliest known echinoid - Bothriocidaris (Photo 35) made its appearance during the same age which is much earlier than in any other region of the world. The first stromatoporoids (Stromatocerium) and tabulate corals (Lyopora, Eoflecheria) are known since the Middle Caradoc, the Oandu Age. In several groups (bryozoans, brachiopods, trilobites, conodonts) some North American faunal elements (e.g. Howellites, Zygospira, Rynchotrema, Bumastoides, Belodina) immigrated. This suggests the beginning of a gradual remission of provincial differences between the Baltic region and North America which progressed during the whole Late Ordovician. Only at the beginning of the Silurian, the fauna in the Baltic area became more or less cosmopolitan as during the end-Ordovician mass-extinction only the most tolerant, wide-ranged taxa survived while the more specialized, endemic elements died out (Nestor et al. 1991).

In the Silurian, a clear lateral differentiation of ecological assemblages took place. Within the shallow-water faunas the role of corals and stromatoporoids rose abruptly, while in deeper-water associations graptolites, chitinozoans, trilobites and ostracodes preponderated. In most of taxonomical groups lateral ecological communities have been distinquished (see Kaljo & Klaamann 1982, 1986).

During the Silurian, a remarkable evolutionary diversification took place among the coral and stromatoporoid faunas reaching its acme in the Late Llandovery - Wenlock time. The earliest known representatives of several higher taxonomical groups, such as stromatoporoid orders Actinostromatida, Syringostromatida and tabulates Theciida, Syringoporida, originated from the Baltic region.

Agnathans and fishes were also very rapidly evolving groups. The first vertebrates in the Baltic area (thelodont Loganellia) are known from the Middle Llandovery (Raikküla Stage), whilst in some other regions (North America, Australia) agnathans were present at least since the end of the Early Ordovician (see p. 245). Such delay in appearance might be explained by nektobenthic mode of life of the early agnathans which prevented their immigration into the Baltic Basin until the beginning of the closure of the Iapetus Ocean. During the Silurian, the representatives of gnathostomes gradually appeared in the Baltic area: first acanthodians (Gomphonchus) during the Late Llandovery (Adavere Age), bony fishes - osteichthyans (Andreolepis) in the Early Ludlow (Paadla Age). They both are the oldest representatives of the corresponding groups. Chondrichthyans (elasmobranchs) are known since the Early Pridoli (Kaugatuma Age).

Since the Middle Wenlock (Jaagarahu Age), eurypterids became rather common in near-shore, lagoonal facies. However, the appearance of these arthropods in the Baltic area was obviously connected with arising environmental conditions favourable for their settlement as well as their preservation in the fossil record. In some other regions, the eurypterids are known since the Cambrian already. Nearly at the same time, the earliest representatives of another group of merostomates - xiphosurans (Bunodes, Pseudoniscus) – appeared for the first time in the fossil record.

On the one hand, progressive increase in the role and diversity of corals, stromatoporoids, ostracodes and especially vertebrates was the most remarkable trend in the evolution of the Silurian faunas in the Estonian area. On the other hand, it was accompanied by a decrease in the diversity of brachiopods, trilobites and some other groups.

During the Devonian, mainly terrigenous or carbonate terrigenous sedimentation took place in the marginal marine environment in Estonia, while in the Frasnian carbonate deposition predominated again.

The fossils of the lowermost Devonian, coming from the boring cores, are rare. Moreover, a significant part of the Lower Devonian rocks is lacking because of a stratigraphical gap between the uppermost Lochkovian and the upper Emsian. Fish faunas belonging to the Euramerica Province are important throughout the Devonian. Fishes and plants (macroremains, miospores and gyrogonites? of charophyte algae) are the most common fossils in the Upper Emsian and Middle Devonian rocks. Among invertebrates, lingulate brachiopods (Bicarinatina) and conchostracans (Glyptoasmussia, Ulugkemia, etc.) are more numerous in comparison with ostracodes, gastropods and bivalves. In the Middle Devonian, from the Narva to the Burtnieki Age, fishes reached the greatest variety. All main fish groups (agnathans, heterostracans and osteostracans) were present. In the Baltic area, the characteristic members of the fish assemblages were psammosteid heterostracans (Tartuosteus, Pycnosteus, Ganosteus), placoderms (Homostius, Asterolepis), numerous acanthodians and crossopterygians (osteolepidids, Glyptolepis). Dipnoans (Dipterus) and actinopterygians (Orvikuina, Cheirolepis) were also frequent. In the late Middle Devonian Gauja Age the fish fauna became gradually poorer. Among the psammosteids Psammolepis predominated; among crossopterygians Laccognathus was frequent. Later on, in Amata and, particularly, in Pļaviņas time, the psammosteid genus Psammosteus and the antiarch Bothriolepis were most significant. Among the fishes, in the latter time, ptyctodont Ctenurella and, actinopterygian Moythomasia were common. The invertebrate fauna of the Pļaviņas Age includes brachiopods, stromatoporoids, tabulate corals, etc.

Vascular plants show changes from the Hyenia flora with Psilophytites and Hostinella of the Pärnu Age and Pseudo-sporochnus of the Burtnieki Age to the Archaeopteris flora of the Gauja Age. The latter flora is characteristic of the Late Devonian.

The scarcity of the Devonian fossils in comparison with those of the Early Palaeozoic depends on their specific preservation conditions rather than on unsuitable conditions of life.


Ordovician chitinozoans

J. Nõlvak


Eisenack, in his pioneer works on chitinozoans from the early 1930s, studied mainly Ordovician and Silurian glacial erratic boulders in the South-Baltic coastal area. The biostratigraphical potential of chitinozoans was proved by Ralf Männil (1971). Nõlvak and Grahn (1993) elaborated a detailed Ordovician biozonation scheme for the whole Baltoscandia, which is presented in Table 7.

The chitinozoans were exclusively marine organisms, their palaeogeographic distribution shows clearly a pelagic mode of life. They are treated as an extinct planktic group of unknown affinity ranging from the Ordovician to Devonian. Chitinozoans display a rather indistinct provincialism at the generic level. Data concerning the lateral associations of chitinozoans are uncertain yet.

The huge amount of subsurface sections, excellent outcrops and good preservation in the Ordovician and Silurian sequences of Estonia have made chitinozoans (Photo 36:1-12) useful for stratigraphical purposes (Männil 1971, Nõlvak 1972, etc.). They have been used for compilation of all recent stratigraphical charts (Resheniya… 1978, 1987). Compared to other groups, chitinozoans give most precise stratification and correlation on many levels, ecpecially in the Viru and Harju series (Table 7). Beginning from the first record of chitinozoans in the uppermost Tremadoc up to the topmost Ashgill, 15 zones and 8 subzones have been defined (Nõlvak & Grahn 1993).

The earliest record of chitinozoans comes from the topmost layer of the stratotype section of the Varangu Stage in northern Estonia where Lagenochitina esthonica Eisenack appears. The post-Tremadoc Ordovician has a relatively high taxonomic diversity of chitinozoans with more than 130 species being recorded. Only heavy secondary dolomitization in some parts of the sequence (mainly in the lowermost and uppermost Ordovician) has caused the absence of chitinozoans. They do not occur in high-energy grainstones, reef facies and marine redbeds either.

A relatively rich assemblage of chitinozoans has been found beginning from the lowermost beds of Arenig, from the glauconite-rich Leetse sandstones (Hunneberg Stage). It is diversest in the Llanvirn - Lower Caradoc strata (from the Kunda to Haljala stages) where the average number of species reaches 25 (Fig. 154). Throughout the Ordovician, the mean number per stratigraphical unit is 19 (Kaljo et al. 1996). Higher up in the sequence, the taxonomic diversity is below the average with some exceptions in the lower Nabala and Vormsi stages.

In general, the following main changes can be distinguished in the dynamics of the chitinozoan diversity. During early Volkhov to Kunda time, the most intensive origination of taxa took place. The Kukruse - Haljala interval was a time of rapid diversity fluctuations, characterized by many short-ranging species. The late Keila extinction and diversity minimum in the Oandu Age coincided with a considerable sedimentological change in the sequence (Hints et al. 1989). This relatively short episode might be called the “Oandu crisis” and is probably globally observable. At the beginning of the Nabala Age, the intensive origination started. Late Pirgu - early Porkuni time is characterized by mass extinction, only four taxa ranged into the lowermost Silurian, across the Ordovician - Silurian boundary.


Silurian chitinozoans

V. Nestor


The investigation of Silurian chitinozoans in Baltoscandia was initiated by Eisenack (1931-76), continued by Laufeld (1974) and Grahn (1995) in Sweden, by Ralf Männil (Männil 1970) and Viiu Nestor (1976, 1984, 1990, 1992, 1994) in Estonia.

The evolution of the Silurian chitinozoans (Photo 36:6-12) was rather slow on the higher taxonomic level (genera, families). Most of the genera entered into the Silurian from the Ordovician and continued in the Devonian, but the species assemblage and diversity experienced considerable changes throughout the Silurian (Nestor V. 1992). In the Silurian of Estonia, four main cycles of evolution of chitinozoans can be distinguished (Fig. 155 A) on the basis of the diversity and taxonomic innovation of chitinozoan assemblages. At the end of each cycle, extensive disappearance of taxa took place (Fig. 155 B); at the beginning of the next cycle abundant new elements appeared. All cycles are characterized by the presence of taxa with specific morphological features, typical of a definite stratigraphic interval. The Early and Middle Llandovery cycle was still characterized by the presence of some specific Ordovician genera (Cyathochitina, Spinachitina) and by the scarcity of mucronated vesicles. In the Late Llandovery - Early Wenlock cycle, chainlet and copulated forms (Densichitina, Margachitina) made their appearance. Mucronated forms (Conochitina) and those, in which the spines were arranged in rows on the vesicle surface (Gotlandochitina), gained abundant distribution. The Middle Wenlock is characterized by a high abundance and diversity of chitinozoans only in the sections of southwestern Estonia. At that time, species with a spongy and reticulated ornamentation occured. At the end of the Late Wenlock, 80% of taxa became extinct. At the beginning of the Ludlow - Pridoli cycle, in the Estonian sequence 90% of the chitinozoan assemblage got renewed. A big variety of cylindro-spherical, ovoidal and lenticular forms appeared gradually up to the end of the Kaugatuma Age, when the diversity of chitinozoan decreased considerably. The above-mentioned cycles were also closely related to the sea-level fluctuations in the Baltic Basin during the Silurian.

In the Silurian sequence of Estonia, 31 chitinozoan biozones have been distinguished (Nestor V. 1990, Fig. 15), five of which are called interzones as they contain scarce chitinozoans without specific forms (see Table 8). The lower limit of the biozone is usually defined by the first appearance of the zonal species or by the disappearance of a number of species occurring in the previous zone. Good relationship with the regional graptolite biozonation allows to use chitinozoan data for age determination in shelly sequences (Nestor V. 1994).

A wide spectrum of environmental parameters has affected the ecology and distribution of chitinozoans in different facies belts. The abundance and diversity of chitinozoans reach the maximum in the sections of southern and southwestern Estonia where marls and argillaceous limestones are predominating. In more carbonate sections of middle Estonia their variability and number decrease, and the occurrence becomes more sporadic.

The distribution of chitinozoans was also controlled by the temperature of water which can be generally associated with palaeolatitude. For a global biozonation of Silurian chitinozoans, the index species, irrespective of palaeolatitudes and palaeoplate configuration, were selected avoiding usage of endemic taxa. Almost all index species are represented in the Silurian succession of Estonia (Verniers et al. 1995, fig. 4).


Algae and vascular plants

A. Kõrts & E. Mark-Kurik


The Vendian-Silurian flora in Estonia is represented by problematic organic-walled thallus-like macrofossils — Vendotaenides (originally found and named by Eichwald as Laminarites-type algae), calcareous algae, Gloeocapsomorpha as the main precursor organism of oil shale organic matter, acritarchs as a group of palynomorphs of controversial affinity, but possibly comprising also algal cysts, and prasinophyte palynomorphs Tasmanites and Leiosphaeridia. Oncolites and stromatolites, biosedimentary structures characteristic of intertidal and subtidal marine environments, have also revealed microbial-algal fossils.

Taxonomical studies of Estonian calcareous algal palaeoflora date back to works by Eichwald (1840, 1854a), Dybowski (1877a) and Stolley (1893, 1896a, b, 1897, 1898). Table 28 summarizes the present knowledge of the distribution of calcareous algae in the Ordovician and Silurian and presents the species list. The distribution of acritarchs in the Vendian, Cambrian and Ordovician has been treated in several papers (Paškevičiene 1980, Volkova et al. 1983, Uutela & Tynni 1991, Mens et al. 1993).

The Lower Ordovician in Estonia lacks calcareous algal macrofossils, but the presence of algal debris in carbonate rocks has been mentioned by Põlma (1982). Cyclocrinitids (Cyclocrinites Eichwald, Coelosphaeridium Roemer, Mastopora Eichwald), unique Palaeozoic dasycladacean macroalgae evolved by the Middle Ordovician (the earliest-known are Coelosphaeridium excavatum Stolley in the Aseri Stage and Coelosphaeridium kohtlense Bekker in the Kukruse Stage), diversified and formed an abundant flora in the Baltoscandian shallow sea during the late Middle Ordovician Rakvere Age. As the biology of these organisms is not sufficiently understood, the dasyclad affinity of Cyclocrinineae Pia has been recently questioned by Nitecki & Spjeldnaes (1992), suggesting a separate taxonomical unit for these extinct algae.

The first dasyclads had appeared already in the Precambrian, but reached remarkable abundance in the Ordovician. During the Middle Ordovician, another group of dasycladacean algae - the vermiporellids (Vermiporella Stolley, Rhabdoporella Stolley) appeared and differentiated. These algae are abundant in the Upper Ordovician Vormsi and Pirgu stages, extending their range up to the Jaani Stage in the Silurian. Stratigraphical range of the genus Rhabdoporella Stolley in Estonia is based on data from the Rapla (Vormsi - Pirgu stages) and Seliste (Juuru - Jaani stages) cores where Rhabdoporella is abundant in certain layers. Recent unpublished finds from the Kuldiga Formation of the Porkuni Stage in the Taagepera and Ruhnu boreholes confirm that, in all likelihood, Rhabdoporella like Vermiporella, formed algal mats occupying large areas in the shallow epicontinental sea (Jux 1966) and had the widest geographical distribution in Baltoscandia and North America (Poncet & Roux 1990). The occurrence of numerous specimens of Rhabdoporella pachyderma Stolley in the Porkuni Stage is associated with the Hirnantia-fauna.

Solenoporacean algae (Rhodophyta) emerged in the Cambrian, but being rare at that time, they arose in the Middle Ordovician nearly simultaneously in Estonia (Solenopora filiformis in the Idavere Stage), Scotland and North America (Roux 1991) comprising two genera – Solenopora Dybowski and Parachaetetes Deninger. The latter genus is more typical of Ludlow patch reefs in Estonia.

In the Wenlock and Ludlow, the cyclocrinitid-vermiporellid flora was replaced by different algal communities consisting of solenoporaceans, codiacean Dimorphosiphon Hoeg and various microalgae, including Hedstroemia Rothpletz, Wetheredella Wood, Ortonella Garwood and Bevocastria Garwood (Radionova & Einasto 1986).

The Devonian plant remains (Photo 37:1-10), particularly those coming from the Tori locality on the Pärnu River are known since Eichwald’s studies (Eichwald 1854b, a.o.). Karpinsky (1906) included into his description of charophytes the gyrogonites (“trochiliscs”) collected from Tartu. Thomson (1940) described both macroremains and miospores from Tori and Küllatova. The early Middle Devonian miospores were figured and listed by Vaitiekuniene (Kleesment et al. 1975, Valiukevičius et al. 1986) and Kedo (Sorokin 1981). Küllatova miospores were recently identified by Loboziak (Blieck et al. 1996). Yurina (1988) revised Thomson’s material from Tori and Küllatova and preliminarily identified plants from the clay quarry at Joosu. The latter locality yielded a new fossil species showing sporangia (Pseudosporochnus estonicus), described by Kalamees (1988), who also redescribed plant macrofossils from Tori. In 1982, one more rich plant locality – Kose (Oore) downstream of Tori – was discovered.

Table 29 shows three well-known stratigraphical units with plant macroremains identified on generic and species level: (1) Tori Member, (2) upper clayey part of the Abava Substage and (3) the Lode Member. (1) The Tori Member includes Hostinella sp. and Psilophytites sp. belonging either to Pteridophyta or Progymnospermopsida (Kalamees 1988). Earlier these plants, known as Aulacophycus sulcatus Göpp., Asteroxylon and Aneurophyton, were considered as psilophytes. (2) Pseudosporochnus estonicus (Kalamees 1988) from Joosu is a pteridophyte. (3) According to Meyen (1987), the fossils Archaeopteris fissilis Schalh. and Archaeopteris sp., identified from the Lode Member by Yurina (1988), are progymnosperms. Hostinella is also reported from the same level (Thomson 1940).

If to list all Devonian flora occurrences, there is hardly any stratigraphical unit without plant remains. Rather common are gyrogonites(?) of charophyte algae (Sycidium, Trochiliscus), particularly in the Tamme Member of the Pärnu Stage, called earlier the “Trochilisken-Sandstein” (Orviku 1930). Charophyte algae occur also in all three members of the Narva Stage and in the Viljandi Member of the Aruküla Stage. Calcareous algae are known from the Pskov Substage of the Pļaviņas Stage (Mark & Paasikivi 1960). The Burtnieki and Gauja stages have yielded silicified and ferriferous wood. Rocks of the Narva, Aruküla and Burtnieki stages contain on several levels unidentified plant remains.

Besides macroremains, plant microfossils also occur on certain stratigraphical levels. Miospores have been established in the grey-coloured rocks of the Rēzekne Stage (Lower Devonian) and of the Tori and Vadja members (Middle Devonian). They occur also much higher up, in the Lode Member of the Gauja Stage. Acritarchs have been discovered from the Vadja Member of the Narva Stage.

Table 30 shows the stratigraphical range of the miospores and the probable position of the spore zones given by Avkhimovitch et al. (1993) for the Devonian of Eastern Europe. The Periplecotriletes tortus (PT) Zone is well established in the Tori Member by the presence of the index species and Calyptosporites velatus. The miospores Punctatisporites tortosus and Hymenozonotriletes ludzus (= Grandispora ludza), characterizing the earlier, Diaphanospora inassueta (DI) Zone and established in the Rēzekne Stage of the southern Baltic and Belarus (Avkhimovitch et al. 1993), occur in the Estonian sequence in the Pärnu Stage. The Tori Member (Pärnu Stage) and the Vadja Member (Narva Stage) have quite a number of common spore species. The Vadja assemblage seems to be older than that of the Rhabdosporites langii (RL) Zone, characteristic of the Kernavė Member in Lithuania. The spore assemblage in the Lode Member (Gauja Stage) can be considered as that of the Accyrospora incisa- Geminospora micromanifesta (IM) Subzone (Blieck et al. 1996). In Belarus, the miospores of this subzone occur in the lower portion of the Lan’ Stage which is an approximate equivalent of the Gauja Stage (Valiukevičius et al. 1995).



H. Nestor


In the Ordovician and Silurian strata of Estonia, 88 valied species of stromatoporoids have been described (Rosen 1867, Nicholson 1886-91, Ryabinin 1951, Nestor 1960, 1964, 1966). The full list of the species, belonging to 26 genera and 16 families, and representing all orders of stromatoporoids except fine-cylindrical amphiporids, was published recently (Nestor 1990c). Stromatoporoids (Photo 38:1-3) are continuously present in all regional stages of Estonia beginning from the Lower Ashgill (Vormsi Stage) and ending with the Lower Pridoli (Kaugatuma Stage). However, some earliest representatives of stromatoporoids (Stromatocerium canadense and S. sakuense) occur in the Middle Caradoc (Oandu Stage) already. Shorter gaps in the distribution, explained with unfavourable ecological conditions or local stratigraphical hiatuses, occur at the Llandovery/Wenlock and Wenlock/Ludlow boundaries.

Stromatoporoids appeared in the Estonian sequence later than in North America, North China or Australia where the earliest indubitable stromatoporoids have been recorded from the Llanvirn - Llandeilo strata already. It has been explained with the location of the Baltica Continent in the southern temperate climate zone until the Ashgill time when it migrated finally into the equatorial belt (Webby 1980).

In the Ordovician of Estonia, stromatoporoids are rare, except its topmost part - the Porkuni Stage. A few species of the most primitive, vesicular stromatoporoids (Order Labechiida) have been recorded from the Oandu (Stromatocerium), Pirgu (Cystostroma) and Porkuni (Pacystylostroma) stages. Plumatalinia, a problematic intermediate form between labechiidae and reticulate stromatoporoids (Actinostromatidae), occurs in the Pirgu Stage (Table 31). In the latest Ordovician, representatives of the sublaminate stromatoporoids (Order Clathrodictyida) became rather common: Clathrodictyon appeared during the Vormsi Age and Ecclimadictyon in the Porkuni Age.

In the Llandovery, clathrodictyids flourished. Clathrodictyon Nich. et Murie and Ecclimadictyon Nestor became dominating genera. They formed more than 80% of stromatoporoid specimens. Labechiids (Pachystylostroma, Forolinia, Rosenella and Labechia) were the second abundant group. During the Llandovery, the first representatives of several families appeared in the Estonian sequence. Thus, during Raikküla time, Intexodictyon - the earliest known representative of the Family Atelodictyidae (Table 31), and Plectostroma, the first certain representative of the Order Actinostromatida, made their appearance. During the Adavere Age, Petridiostroma was added among clathrodictyids as the most ancient representative of the Family Gerronostromatidae. At the same time, a very peculiar form “Stromatopora” elegans Rosen (=Pachystroma) appeared, showing the closest affinities to the Family Pseudolabechiidae. Thus, during the Llandovery, the first genuine laminate stromatoporoids (Atelodictyidae and Gerronostromatidae) and different branches of reticulate stromatoporoids (Actinostromatidae, Pseudolabechiidae) were gradually added to the prevailing fauna of the sublaminate and vesicular stromatoporoids.

During the Wenlock, the enrichment and diversification of the stromatoporoid fauna continued. In Jaani time, the first known microreticulate stromatoporoid Densastroma appeared. It belonged to the family Densastromatidae and played an important role later in the Silurian. At the same time, Stromatopora appeared in the Estonian sequence, being one of the earliest representatives of the Order Stromatoporida, i.e. stromatoporoids with irregularly amalgamated skeletal elements. Simplexodictyon validum Nestor, an early representative of the tripartite-laminated stromatoporoids (Order Stromatoporellida), has been recorded from the Maasi beds of the Jaagarahu Stage. Vikingia tenuis (Nestor) was the main frame builder in the Jaagarahu reefs (Vilsandi beds); it may be treated as a possible ancestor of the Order Syringostromatida with clinoreticulate microstructure of vertical skeletal elements (Nestor 1994). During the Wenlock the role of clathrodictyids decreased considerably; labechiids have not been recorded from Estonia.

During the Early Ludlow (Paadla Age), the diversity of the stromatoporoid fauna reached its maximum (13 genera from 12 families) in Estonia. In different taxonomical branches new elements were added, e.g. Lophiostroma among labechiids, Plexodictyon among clathrodictyids, Pseudolabechia in Actinostromatida, Syringostromella among stromatoporids, and Parallelostroma - the first representative of the Order Syringostromatida in the Estonian sequence. However, representatives of all the above-mentioned genera are known from somewhat earlier strata in other regions (Nestor 1994). In the Late Ludlow (Kuressaare Age) and Early Pridoli (Kaugatuma Age), the diversity of stromatoporoid assemblages decreased considerably due to an increase in the clay content of sediments and bad exposure of the corresponding strata. Stromatoporoids have not yet been recorded from the Silurian Ohesaare Stage.

The above leads to the conclusion that favourable climatic conditions for constant colonization of the Estonian area by stromatoporoids were established during the Ashgill. The beginning of the Silurian was characterized by comparatively unilateral fauna of sublaminate clathrodictyids (Clathrodictyon, Ecclimadictyon) with an admixture of labechiids. During the Llandovery and Wenlock, representatives of most of the orders and families were gradually added, and in the Early Ludlow (Paadla Age) the diversity of the stromatoporoid fauna reached its maximum, falling after that rapidly due to a progressive increase in the influx of clayish clastic material from the raising Caledonides.

Stromatoporoids were highly facies-dependent organisms with comparatively narrow ecological niche. The richest and most diverse stromatoporoid association occurred in the high-energy shoal facies belt, represented in fossil record by coral-stromatoporoid boundstones, skeletal and coquinite grain- and rudstones (Nestor 1990c). They were rather numerous also in the moderate- to low-energy open-shelf facies belt where biomicritic deposits (nodular skeletal packstones) were accumulated. Parallel successions of imperfectly deliminated lateral communities have been distinguished for shoal and open-shelf environments, consisting of 23 stromatoporoid communities (Nestor 1990c). Up to now, no definite biogeographic provinces have been established for the Late Ordovician and Silurian stromatoporoids (Nestor 1990b).


Tabulate corals

M.-A. Mõtus


Estonian tabulate corals (Photo 38:4-5) have been studied in particular detail by Sokolov (1951a, 1952b, 1955) and Klaamann (1961, 1962, 1964, 1966, 1983). The monographic studies of the above researchers formed the basis for the Estonian tabulate coral taxonomy.

The earliest fossil of possible tabulate corals was reported from the Lower Cambrian Moorowie Formation of South Australia (Fuller & Jenkins 1994). Genuine tabulates with the genus Cryptolichenaria appeared in the Early Ordovician of Texas, Pennsylvania and Siberia (Kaljo & Klaamann 1973).

In the Middle Ordovician, tabulates were of wider distribution. Lichenariids and tetradiids were common for Sibera and North America. The Late Ordovician was characterized by a greater variety of tabulates; favositids, heliolitids, tetradiids, sarcinulids were widespread in different parts of the world. In the Silurian and Devonian, a great diversification of tabulate corals took place. In the Silurian the most common were favositids, halysitids, heliolitids, theciids and syringoporids, reaching the maximum diversity in the Wenlock. Halysitids disappeared at the end of the Silurian. Some new families, including Micheliniidae, Cleistoporidae, Trachyporidae, appeared already at the end of the Silurian, but became widespread in the middle of the Devonian together with pachyporids, alveolitids and coenitids. Big differences occurred also on the generic level. At the end of the Devonian, nearly all subfamilies of favositids disappeared as did all theciids, syringolitids, alveolitids and also the order Heliolitida. In the Carboniferous, auloporids, micheliniids, cleistoporids and palaeacids were well diversified with syringoporids dominating. The Permian fauna did not differ much from the Carboniferous one, but was poorer and less widely distributed. Tabulate corals became extinct at the end of the Late Permian.

Several palaeobiogeographic faunal regions - Americo-Siberian, Central-Asian and European (including the Baltic area), have been distinguished (Kaljo & Klaamann 1973). Due to the favourable climatic and shallow-water conditions in the Estonian part of the Baltic Basin, tabulates played a significant role in the Late Ordovician and Silurian faunal assemblages.

In Estonia, the earliest tabulates are known from the Late Caradoc. Lyopora tulaensis Sok., Saffordophyllum grande Sok., Eoflecheria orvikui Sok., occur in the Vasalemma reef facies of the Oandu Stage (Table 32). At the beginning of the Late Ordovician, the conditions for tabulates were unfavourable, therefore the fossils are rare. Only one species, Catenipora obliqua (Fischer - Benzon) has been recorded from the Nabala Stage. Diversification of the Late Ordovician tabulate fauna began in the Vormsi Age with the appearance of the first Paleofavosites - P. schmidti Sok. and P. borealis Tshern. The most ancient heliolitids (Wormsipora, Esthonia, Protaraea) are also known from the Vormsi Stage. Sarcinuliids and halysitids (Catenipora) became common for the first time. During the Pirgu Age diversification continued. The tetradiids (Cryptolichenaria multiplex Klaam.) appeared for the first time in the Baltic area. The halysitid Eocatenipora was widely distributed and early heliolitids were well developed. The Late Ordovician favositids were blooming in the Porkuni Age. The first records of Porkunites, Mesofavosites and Priscosolenia are from the same age.

The beginning of the Silurian was a time of rapid diversification of Mesofavosites and Paleofavosites. In the middle of Llandovery, Multisolenia and Parastriapora were added to Favosites and favositids became the dominant group of tabulates in the Silurian. The Juuru Stage is characterized by a few genera of tabulates. Paleofavosites and Mesofavosites were the most common genera at that time. Catenipora was common at the beginning of the age.The oldest representatives of the genera Halysites (H. priscus Klaam.) and Favosites (F.antiquus Sok.) appeared. In the Raikküla Stage representatives of the genus Paleofavosites are less numerous than in the Juuru Stage whereas Favosites is more abundant. Such genera as Parastriatopora, Multisolenia, Syringopora, Vacuopora and Sinopora appeared at that time. In the Adavere Stage the assemblage of tabulates is more diverse, but less endemic than the Raikküla assemblage. The first alveolitids and coenitids occur, auloporids and halysitids (Catenipora) are widespread. The morphology of the tabulates in the Adavere Stage is quite different from those appearing in the Wenlock.

The Jaani Age was characterized by major changes in the tabulate fauna: Syringolites, Thecia and Mastopora appeared first at the end of this age; Mesofavosites and Catenipora disappeared. Paleofavosites, Mesofavosites, Favosites, Catenipora and Halysites were poorly represented in the Jaani fauna. During the Jaagarahu Age, the diversity of tabulates rose again. That can be explained by more favourable enviromental conditions in a widespread shoal facies. The new genera Cladopora and Romingerella appeared at that time. In the Jaagarahu fauna the typical Silurian tabulate genera were almost fully represented, but at the end of the age Multisolenia and Halysites disappeared from the Estonian sequence, which preceeded their disappearance in surrounding areas. The Rootsiküla Stage is characterized by rare Favosites and an abundance of Parastriatopora commutabilis, specific to the stage.

In the Ludlow, tabulates had a low generic diversity. Of those, Favosites was important. A few representatives of Thecia, Romingerella, Laceripora and Syringopora have been found from the Paadla Stage. Favosites was more diverse than in the Rootsiküla Stage. Most tabulates disappeared at the end of the Paadla Age. The Kuressare Stage has a very poor record of tabulates and only a few species of Favosites and Aulopora have been found. No changes at the generic level took place at the Ludlow - Pridoli transition. The last few species of Paleofavosites occur in the Kaugatuma Stage, whereas Favosites is quite common. Syringopora and Mesosolenia are rare. The tabulate fauna of the Ohesaare Stage does not differ significantly from that of the Kaugatuma Stage and only a few species have been recorded.

The Estonian tabulate faunas reveal quite clear differentiation into lateral communities (Klaamann 1986) as shown in Table 33. The formation of communities was influenced by the water depth, hydrodynamics and other factors. The depth of water determined the species composition of communities, while the other agents controlled the shape of coral colonies and the diversity and number of lateral communities. The more the enviromental conditions differentiated, the greater the number of lateral communities was.


Rugose corals

D. Kaljo


Knowledge of the Ordovician and Silurian rugose corals of Estonia (Photo 38:6-7) is mainly based on the studies by Eichwald (1854-60), Dybowski (1873/4), Weissermel (1894), Reiman (1956, 1958) and Kaljo (1956, 1958, 1961, 1970b, 1996). During the recent decades, Neuman (1969, 1986), Scrutton (1988) and Weyer (1973, 1982, 1983, 1993) have published several papers describing only a few new taxa but improving considerably the taxonomy of corals identified earlier. The number of the known species-level taxa, slightly exceeds one hundred, but the share of undescribed forms might be at least 20-30%.

Rugose corals made their first appearance in the Middle Ordovician of North America. In Estonia, they are represented by Primitophyllum primum Kaljo and Lambelasma dybowskii (Kaljo) occurring in the Haljala Stage and undoubtedly having the habitus of the most primitive tetracorals. In general, the Ordovician rugose coral assemblages were dominated by simple streptelasmatid corals provided only with tabulae between septa and often having a dilated septal apparatus. The first corals with well developed dissepimentarium appeared at the very end of the Late Ordovician and gained predominance later in the Silurian. The Ordovician Period, however, ended with a serious extinction of corals (first of all species- and genus-level taxa, particularly streptelasmatids), and the earliest Silurian (Rhuddanian) was a low-diversity period dominated by Ordovician carry-overs. Later, a stepwise increase in the diversity followed until the maximum was reached in the Wenlock (Kaljo 1996). Morphological differentiation was remarkable. New types of septa, stereozones, calices, many colonial forms, etc. appeared which formed a base for taxonomical diversity. The most characteristic were different cystiphyllids, kodonophyllids, entelophyllids, lykophyllids, arachnophyllids, etc. The Late Silurian shows a decline of rugose corals in general, but a few new elements appeared in the Pridoli, among them the so-called “Devonian” elements (Acanthophyllum, Lyrielasma, etc., Scrutton, 1988).

The above general evolutionary pattern is well observable in Estonia. Apart from the above-mentioned primitive rugosans, Kenophyllym and Streptelasma appeared in the Keila Age, and Borelasma and the first tryplasmatid Estonielasma, in the Oandu Age. The first Grewingkia was identified at the end of the Middle Ordovician. It means that rugose corals were scarce in the Middle Ordovician of Estonia, but their diversity was already comparatively high.

The Late Ordovician was mostly dominated by streptelasmatids (Kenophyllum, Streptelasma, Grewingkia, Helicelasma, Dalmanophyllum), but there occurred also rare lambelasmatids or calostylids s. l.: Coelostylis (Vormsistylis), Neotryplasma, Calostylis, Estonielasma. The end of the period (Porkuni Age) was marked by the incoming of the first paliphyllids (Paliphyllum, Strombodes) and staurids (Palaeophyllum).

The Silurian rugose corals in Estonia are the most diversified in the following stratigraphical units: (1) the upper Aeronian Rumba Formation (Dinophyllum, Entelophyllum, Prodarwinia, Phaulactis, etc.); (2) the Middle Wenlock Jaagarahu Formation (Acervularia, Spongophylloides, Microplasma, etc.); (3) the Ludlow - Pridoli (Entelophyllum and Tryplasma were most common, but in the Kaugatuma Stage also Cystiphyllum, Holmophyllum, Strombodes and the first representatives of Acanthophyllum appeared).

The distribution of these corals shows a distinct facies control. Reliable records of rugose corals from the Silurian lagoonal and shelf depression facies are lacking. These corals were scarce also in the Borealis and Pentamerus banks and stromatoporoid biostromes, but rich assemblages occurred in the reefs and their surroundings (e.g. Hilliste reefs of the Juuru and Raikküla stages, Sepise outcrop of the Jaagarahu Stage, etc.). A diverse assemblage of rugose corals occurred also in the shallow part of the open shelf. However, the share of solitary corals in it was higher than in reef environments, and the role of colonial corals decreased. Up to now, only a few species have been identified from the deeper (outer) shelf (Porpites porpita from the Velise Formation, etc.), but many new taxa have not been described yet.

By now, no suggestions for the biozonations of rugose corals have been made, but Kaljo (1961, 1970b, 1996) has listed the characteristic species for stratigraphic units.

Biogeographically, Estonian rugose corals belonged to the Baltoscandian (or North European) Province which had some connections with the North American - Siberian and also with the Middle Asian provinces. These connections, as well as the share of the widely distributed and endemic corals, were changing during the time under discussion. The importance of endemic corals was relatively high before the Wenlock.


Inarticulate brachiopods

I. Puura


Inarticulate brachiopods with calcium phosphate and calcitic shell, formerly referred to the class Inarticulata, have been recently assigned to three new subphyla, and several classes (Williams et al. 1996). The representatives of the subphylum Linguliformea, formerly known as phosphatic inarticulate brachiopods, are abundant in the Cambrian and Lower Ordovician of Estonia, common in the Middle and Upper Ordovician, rare in the Silurian and rather common in the Devonian. The representatives of the subphylum Craniiformea are known from the Ordovician and Silurian.

These brachiopods have been studied since the early 19th century. In their earlier works Eichwald, Pander, Mickwitz, Kutorga, Huene, Walcott, Bekker and several other authors described many new species and genera from the Cambrian and Ordovician of Estonia (see Puura 1990 for review). More recent taxonomical studies of the East Baltic and Scandinavian Cambrian and Ordovician faunas, discussing the systematic position of Estonian taxa, include Gorjansky (Gorjansky 1969), Biernat (1973), Holmer (1986, 1989), Popov and Nõlvak (1987), Popov and Khazanovich (1989), Holmer and Popov (1990, 1994), Puura and Holmer (1993), Popov and Holmer (1994) and Popov et al. (1994).  Rare Silurian and Devonian lingulate brachiopods have been described by Popov (Popov 1981) and Gravitis (Gravitis 1981), respectively.

The earliest brachiopods known from Estonia, Mickwitzia monilifera (Linnarsson) (Photo 39:1) and Paterina rara Gorjansky from the Lower Cambrian, are tentatively assigned to the class Paterinata.

The representatives of the class Lingulata are most abundant in the Upper Cambrian and the lower part of the Pakerort Stage where obolid coquinas form the deposits of shelly phosphorites. In this interval, the successive assemblages have been used for defining four lingulate zones: Ungula inornata (Mickwitz), U. convexa Pander, U. ingrica (Eichwald) and Obolus apollinis Eichwald (Popov & Khazanovich 1989). These zonal species, alongside with more than ten other species, apparently restricted to Baltoscandian Basin, can be used for correlations across the basin from Sweden to Lake Ladoga (Popov & Khazanovich 1989, Holmer & Popov 1990, Puura & Holmer 1993).

In the kerogenous Dictyonema Shale of the upper part of the Pakerort Stage, rare lingulate brachiopods Eurytreta sp. and Lingulella aff. L. tetragona Gorjansky occur.

The Hunneberg Stage is characterized by an assemblage of more than ten lingulate species dominated by Thysanotos siluricus (Eichwald) and Leptembolon lingulaeformis (Mickwitz). In the Billingen Stage, acrotretids Acrotreta subconica Kutorga and Myotreta crassa Gorjansky occur.

The characteristic assemblage of the Volkhov Stage includes Acrotreta tallinnensis Holmer and Popov, Rowellella rugosa Gorjansky and Myotreta estoniana (Biernat). Myotreta crassa Gorjansky, Biernatia rossica (Gorjansky) and Eosiphonotreta verrucosa (Eichwald) range from the Volkhov to the Kunda Stage, while Eoconulus cryptomyus (Gorjansky) ranges from the Volkhov to the Aseri Stage and Conotreta mica (Gorjansky) from the uppermost Kunda to the lowermost Uhaku Stage. Siphonotreta unguiculata (Eichwald) ranges from the Aseri to the Uhaku Stage.

In the Kukruse Stage lingulate brachiopods are represented by Biernatia holmi Holmer and Schizotreta elliptica (Kutorga). Carbonate-shelled craniate brachiopods (class Craniata) are represented by Philhedra baltica Koken, Orthisocrania planissima (Eichwald) and about ten more species assigned to the genera Philhedra, Orthisocrania, Pseudopholidops and Paracraniops.

In the Haljala Stage, the lingulate Alichovia ramispinosa Gorjansky and the craniates Philhedra metatypotheisa Huene, Orthisocrania curvicostae Huene occur. O. depressa (Eichwald) ranges from the Jõhvi Substage of the Haljala Stage to the Keila Stage and Philhedra kegelensis Huene from the Keila Stage to the lowermost Oandu Stage.

From the Nabala Stage, Gorjansky (Gorjansky 1969) has reported Pseudolingula quadrata (Eichwald) and Lingulops mirus Gorjansky. An assemblage from the Vormsi and Pirgu stages of the Viljandi core includes Acanthambonia portranensis Wright, Rowellella minuta Wright, Spondylotreta cf. parva Wright, Eoconulus semiregularis Biernat, Paterula sp. and Schizotreta sp. (Popov & Nõlvak 1987, Popov et al. 1994).

The class Craniata is represented in the upper half of the Ordovician by Pseudopholidops stolleyana in the Oandu Stage and by about ten species assigned to the genera Philhedrella? (Rakvere and Nabala stages), Petrocrania? (Vormsi Stage) and Pseudocrania (Porkuni Stage).

The only two lingulate species so far described from the Silurian of Estonia include Opsiconidion aldridgei (Cocks) from the Raikküla and Jaani stages and Eschatelasma rugosum Popov from the Jaani Stage (Popov 1981).

From the Devonian of Estonia, Gravitis (1981), has reported lingulate brachiopods Bicarinatina bicarinata (Kutorga), B. ugalana Gravitis from the Aruküla Stage and B. sakalana Gravitis from the Narva Stage. The latter two species have also been reported from some Latvian sections.

Most of the species discussed above are restricted to the East Baltic and Scandinavia, but some lingulate species have a wider distribution and are of interest for biogeographic and palaeogeographic studies. For instance, the Upper Cambrian species Angulotreta postapicalis is known, except for Estonia, also from Novaya Zemlya and North America (see Puura 1990). Thysanotos siluricus and Leptembolon lingulaeformis, typical of the Hunneberg Stage in Estonia, are known from about the equivalent stratigraphic level from the South Urals, Poland and Bohemia (Popov & Holmer 1994). The Upper Ordovician Acanthambonia portranensis and Rowellella minuta are known from the Upper Ordovician of Ireland (Wright 1963) and the Lower Silurian of Wales and England (Cocks 1979).


Ordovician articulate brachiopods

L. Hints & A. Rõõmusoks


Reseach into articulate brachiopods, which form one of the main groups in the Ordovician shelly fauna (Photo 39:2-5), was started in northern Estonia and adjacent areas in the 19th century (Pander 1830, Eichwald 1860) and continued by Öpik (1930b, 1934, a.o.), Alichova (Alichova 1951, 1953) Rõõmusoks (1959, 1981, 1989,  a.o.), Rubel (1961), Oraspõld (1956) and others. As a result, more than 300 species of Articulata have been described. They belong to about 130 genera, among which the representatives of the orders Orthida and Strophomenida dominated during the Early and Middle Ordovician (Table 34). The orders Pentamerida, Rhynchonellida and Spiriferida represent mainly the latest Middle and Upper Ordovician. The occurrence of numerous endemic elements in the Baltic Basin has enabled to distinguish a separate Baltic Province (Williams 1973) or a specific Baltoscandian fauna (Jaanusson 1973b). During the Ordovician, the changes in the composition of brachiopod fauna were caused in a great deal by a succesive decrease of the endemic elements and increase of the immigrants from the different faunal provinces (Rõõmusoks 1967, 1970). The dynamics and comparison of different contemporaneous Ordovician brachiopod faunas, including the Baltoscandian, has been analysed by Jaanusson (1973b, 1976, 1979, 1984). Reviews on the brachiopod fauna in Estonia and data on the stratigraphical distribution of species have been presented in several publications (Männil 1966, Männil et al. 1966, Hints et al. 1989, Hints 1990), in particular detail by Rõõmusoks (1967, 1970, 1983).

Different brachiopod associations are characteristic of northern and southern Estonia belonging to different confacies belts (Fig. 24). The richest and diversest brachiopod fauna has been described from the variably argillaceous limestones in northern Estonia. The micritic (aphanitic) limestones (Rägavere and Saunja formations) comprise relatively few brachiopods, which mainly occur in the argillaceous interbeds. An impoverished brachiopod fauna is characteristic of the red-coloured units (the whole Lower Ordovician, Jonstorp Formation) in southern Estonia.

The earliest articulate brachiopods (orthids Prantlina, Panderia, Ranorthis, dalmanellids Paurorthis, plectambonitids Plectella a. o.) appear in the glauconitic sandstones of the Mäeküla Member in the lower part of the Billingen Stage (see Table 34). In the carbonate rocks of the Volkhov and Kunda stages, this earliest short-living brachiopod fauna is replaced by a new fauna comprising the orthids Productorthis, Orthambonites. Orthis, plectambonitids Ahtiella, Ingria and taxa of Clitambonitidina (Antigonambonites, Gonambonites, Progonambonites), mostly endemics in the Baltic Basin.

At the beginning of the Middle Ordovician, the last Early Ordovician (Oelandian) brachiopods (Ladogiella, Lycophoria) became extinct and the Middle Ordovician (Viruan) brachiopod fauna, which comprised several new strophomenids (Christiania, Leptestia), started to form. Somewhat later, several new taxa characteristic of the lower half of the Viru Series (Sowerbyella, Leptelloidea, Hesperorthis, Clitambonites, a.o. were added (Table 34). Several of these brachiopods have been found in the core sections which allows to suppose their wide distribution within the North Estonian Confacies Belt (Fig. 32). From southern Estonia, the early and middle Viru brachiopod fauna is insufficiently known. Alwynella, Bimuria, Sampo? (?=Leptellina) serve as Scandinavian faunal elements.

A remarkable renovation of the brachiopod fauna took place at the Keila - Oandu transition coinciding with the period of essential environmental changes in the basin. Most of the brachiopod species and many genera, existing earlier in the basin, disappeared at the end of Keila time (see Rõõmusoks 1970). The new, post-Keila time brachiopod fauna comprised several taxa which had immigrated from North America and North Europa (Rynchotrema, Rostricellula, Camerella, Dactylogonia; Rõõmusoks 1967). Some of those (e.g. Reushella) have been found only in central Estonia. In southern Estonia, the changes in the composition of the brachiopod fauna are not clear. However, since Keila - Oandu time, the occurrence and diversity of articulate brachiopods decreased notably, especially in the black shales (Keila - Oandu and Vormsi stages) and red-coloured limestones (Pirgu Stage).

In northern Estonia, the late Middle and Upper Ordovician brachiopod faunas were dominated by relatively large-shell brachiopods (Plaesiomys, Platystrophia, Vellamo, Triplesia, Bekkeromena, “Leptaena” and others). After gradual appearance of Dicoelosia (Vormsi Age) and Eospirigerina (Pirgu Age), and presumably also of Holorhynchus (latest Pirgu Age), the Baltic fauna became close to the Hiberno-Salairian fauna (Jaanusson 1979). At the same time, several descendants of endemic brachiopods occur (Ilmarinia, Apatorthis, Equirostra). The number of brachiopod genera decreased towards the end of the Ordovician. Thus, in the Vormsi and Pirgu stages, the brachiopods are represented by more than 35 genera, about half of which range over into the Porkuni Stage where they are associated with a few new taxa. Streptis and Meristina can be treated as really new taxa, whereas some other genera (Reushella, Laticrura, Rhynchotrema) recurred after their first entrance in Estonia during the late Viru time. Articulated brachiopods are quite common fossils in the reef complex of the Porkuni Stage (main part of the Ärina Formation) in spite of their decreased diversity. This brachiopod fauna, named tentatively the Streptis association (Hints 1993), disappears at the end of the Ordovician. Up to now, there are only two Ordovician species Onniella trigona Rubel and Eospirigerina porkuniensis Rubel which have been identified also from the lowermost Silurian.

During the Porkuni Age, the so-called Hirnantia fauna, described from different Ordovician basins (Rong & Harper 1988), was distributed in southern Estonia. Several typical representatives of that fauna (Hirnantia, Dalmanella, Plectothyrella, Hindella) have been established in the Kuldiga Formation (Ruhnu, Ikla, Taagepera drill cores) which commonly represents the lower and middle parts of the Porkuni Stage in the central East Baltic (Ulst et al. 1982, Oraspõld 1975b). Up to now, the relationships between the Streptis and Hirnantia faunas are unclear. The preliminary data on the isotopic composition of the topmost Ordovician strata suggests that these faunas may be partly contemporaneous, but the Hirnantia fauna has existed longer.


Silurian articulate brachiopods

M. Rubel


The Silurian articulate brachiopods (Photo 39:6-7) were numerous and diverse in carbonate facies of the cratonic seas including the Estonian area (Table 35). Up to now, about 200 species from 98 genera have been described from the Estonian Silurian by different authors in a lot of papers, including monographs by Sokolskaya (Sokolskaya 1954), Rubel (1963, 1970a) and Modzalevskaya (1985). At least three well-known evolutionary lineages in the genera Stricklandia, Dicoelosia and Pentamerus have been studied and used for the dating of rocks (Rubel 1971, 1977, Musteikis & Puura 1983, Johnson et al. 1991).

The appearance of a large number of genera at the beginning of the Silurian was connected with the expansion of many cosmopolitan Silurian brachiopods all over the world, excluding spiriferids which were lacking in the Baltic area at that time. The second increase in the number of genera in the Jaani Age was probably connected with the new transgression of the sea, tied with the global sea-level rise at that time. After the Late Wenlock crisis, there followed a gradual decrease of the Late Silurian brachiopods in Estonia reflecting the stepwise retreat of the sea.

The general impoverishment trend can be proved also by the brachiopod communities. Thus, the Linoporella, Borealis-Pentamerus, Stricklandia-Zygospiraella and Meifodia-Clorinda communities are clearly recognizable in the Llandovery; the Stegerhynchus, Whitfieldella and Dicoelosia-Skenidioides communities, widespread in the Wenlock, turned into low-diversity Didymothyris-Salopina, Atrypoidea and Homoeospira-Delthyris communities in the latest Silurian time (Rubel 1970b, Kaljo & Rubel 1982).


Bivalves and rostroconchs

M. Isakar


The Ordovician and Silurian bivalves (Photo 40:5-6) and rostroconchs have been described in Estonia since the middle of the last century (Eichwald 1840b, 1842, 1860, Schmidt 1858, 1861, 1881). In this century, Bekker (1921) and Öpik (1930a) described some new bivalves from the Kukruse Stage. Teichert (1930) recorded an Ordovician rostroconch Ischyrinia from the Upper Ordovician. During the last years, some more Ordovician and Silurian bivalves of Estonia have been described (Isakar 1985, 1990, 1991, Isakar & Sinitsyna 1985, 1993, Sinitsyna & Isakar 1987, 1992, Kiselev  et al. 1990).

Bivalves have a rather limited stratigraphical value due to their rarity and poor preservation. They usually occur as casts. However, some species are more valuable, at least in stage-level correlation, e.g. Ahtioconcha auris, Ilionia prisca and Grammysia obliqua. Most of bivalve and rostroconch specimens have been collected from the North and Central Estonian confacies belts.

In Estonia, the first bivalves (some undetermined small nuculoids) and rostroconchs (Ischyrinia and Eopteria) appeared during the Kunda Age (Table 36). During the Aseri Age, small Similodonta and Cleionychia were added. The fact that the earliest pelecypod fauna was dominated by infaunal forms (nuculoids) suggests that the ancestral mode of life of the class was infaunal (Pojeta 1971). During the Early Ordovician, rostroconchs underwent their greatest radiation (Pojeta & Runnegar 1976). At least four rostroconch species (Ischyrinia norvegica, I. triangularis and two species of Eopteria) existed during the Kunda Age in Estonia.

Ahtioconcha Öpik from the Kukruse Stage is the earliest probable pteriacean known so far (Pojeta 1971). Species diversity increased considerably during the Middle Ordovician, reaching the peak (13 species) in the Kukruse Age (Fig. 156). Three genera – Tancrediopsis, Dystactella? and Ahtioconcha, occur only in the Kukruse Stage. In the Middle Ordovician, the common and numerous pelecypod genera (unfortunately usually badly preserved) were Modiolopsis, Cleionychia, Plethocardia?, Vanuxemia?, Ambonychiopsis and Aristerella. Ulrich (1894) described Aristerella as having the left valve smaller than the right. Most of the Estonian representatives of Aristerella have markedly right convex inequivalved shells but “Aristerella” from the Jõhvi Stage has an even plano-convex inequivalved shell and obviously belongs to a separate genus. In the Porkuni Stage there occurs a specific pelecypod association - Similodonta, Ctenodonta ?, some undetermined small nuculoids, Modiolopsis, Mytilarca, Ambonychia, Ambonychiopsis, Cleionychia ? and Pterianea. The small rostroconch Hippocardia with fine hood and long tubular rostrum occurs also in the Porkuni Stage.

Bivalves have not been found from the Landovery. This may be due to unsuitable conditions for preservation of the shells.The bivalves association of the Jaani Stage is dominated by small nuculoids - Nuculoidea, Deceptrix ?, Prae-ctenodonta and Orthonota, Praecardium, Grammysia (the latter comes from the Nässuma borehole, see Fig. 3 - 223). A massive, thick-shelled Megalomus formed banks in the shoal deposits of the Jaagarahu Age. Mytilarca and Modiolopsis occur also in the same stage. From the Rootsiküla Stage, Palaeopteria, Modilopsis, Modiodonta have been recorded. One type of pterioids inhabited the near-shore shallow-water environment with reduced salinity (see Einasto 1968), forming about a 0.8-m-thick deposit. A lot of bivalves: Ilionia, Kogulanychia, Pteronitella, Palaeopecten, Modiolopsis, Ptychopteria, Megalomoidea, Goniophora and Palaeopteria inhabited in the nearshore environments of the Paadla Age. The Kuressaare and Kaugatuma stages are characterized by Pteronitella, Ilionia and various Pterioidea. In the latter, a rostroconch Mulceodens and a small Nuculoidea have also been found. In the Ohesaare Stage, Grammysia, Ilionia, various Pterioidea, Palaeopecten, Actinopteria and Modiolopsis? are represented.

The above shows that in the Ordovician representatives of the superfamilies Cyrtodontacea, Modiomorphacea and Ambonychiacea dominated, while in the Silurian Pteriacea were most common.



M. Isakar


The first data on the Ordovician and Silurian gastropods (Photo 40:1-4) from Estonia were published in the last century by Eichwald (1840b, 1842, 1860), Schmidt (1858, 1861, 1881) and Koken (1896, 1897, 1898). The latter also published a monograph on the Ordovician gastropods (Koken 1925). Teichert (1928) recorded a new Silurian gastropod species Cyclonema hiiumaa from Hiiumaa Island and Öpik (1930a) described some new gastropods from the Kukruse Stage. During the last years, the Ordovician and Silurian gastropods have been described by Isakar (1990, 1991, 1995) and Kiselev et al. (1990). The earliest gastropod in the Estonian sequence – Aldanella kunda (Öpik) – has been known from the Lontova Formation already, i.e. from the pre-trilobite Early Cambrian rocks.

More than 200 species of gastropods from 63 genera have been identified from the Ordovician and Silurian of Estonia. Since all this material has been collected from northern and central Estonia, we have not been able to record the changes in the lateral distribution of gastropod associations.

The earliest Ordovician gastropod (Subulites huenei Koken) was presumably found from the Lower Ordovician glauconite limestone of the Volkhov Stage (Koken 1925). The diversity of the gastropod association in the next, Kunda Stage is much higher, comprising about 40 species which belong to 20 genera (Fig. 157). In the carbonate rocks of the Kunda Stage, there appeared bellerophontaceans – Salpingostoma, Sinuites, Tetranota, Bucania, Cyrtolites, Temnodiscus; pleurotomariaceans – Pararaphistoma, Clathrospira, Lophospira, Brachytomaria; euomphalins – Lesueurilla, Lytospira, Ecculiomphalus, Helicotoma; trochonemataceans – Proturritella, Spirotomaria, platyceratacean – Holopea, clisospiracean – Clisospira and subulitacean – Subulites (Table 37).

At the Early/Middle Ordovician transition between the Kunda and Aseri ages, the gastropod fauna underwent an essential change. The taxonomic diversity reduced remarkably; this is characteristic of the carbonate sequences throughout the whole Baltoscandian region (Jaanusson 1976). During the Lasnamägi, Uhaku and Kukruse ages, the new genera Cymbularia, Tropidodiscus and Eotomaria appeared, respectively. There were rather abrupt changes in species and generic levels at the Kukruse/Idavere and Keila/Oandu transitions. Many Idavere genera (Salpingostoma, Temnodiscus, Cymbularia, Sinuites, Ecculiomphalus, Lesueurilla, Clathrospira, Subulites and Holopea) have also been recorded from the erratic boulders in Germany (Neben & Krueger 1973). At the Keila /Oandu transition, a significant renovation of the Middle Ordovician fauna took place. Practically all gastropod species and numerous genera (Kokenospira, Tropidodiscus, Temnodiscus, Lesueurilla) disappeared. The diversity of the gastropods of the Oandu Age was relatively low and quite distinct from the assemblages of the preceding ages: platyceratid gastropods (Cyclonema and Platyceras) and rare pleurotomariacean Pseudocryptaenia appeared at that level. The Rakvere Stage consists mainly of calcilutites which contain a gastropod fauna quite different from that of the preceding stages. The genus Murchisonia and new species in Subulites, Pararaphistoma, Mimospira made their appearance at that time.

Most of the gastropod genera in the Nabala Stage (16) are long-ranging. Tetranota conspiquaa is a species occurring only in the Nabala Stage. During the Vormsi Age, the number of the gastropod genera reached 20 which is the maximum for the Late Ordovician. The argillaceous limestones of the Kõrgessaare Formation contain a rich and diverse assemblage of gastropods. The lower boundary of the stage is marked by a sharp change in lithofacies, but is comparatively weakly expressed in the distribution of gastropods. The species occurring only in the Vormsi Stage include Straparollus vortex and Cymbularia aequalis. Sinuites, Bucania, Cymbularia, Helicotoma and Straparollus reappear in the Vormsi Stage, and range into the Pirgu and Porkuni stages. In the Pirgu Age, the gastropod diversity decreased to 14 genera. The only Ordovician index gastropod - Maclurites neritoides - has been recorded from the Adila Formation (Resheniya… 1987). In the Porkuni Stage, a specific gastropod association is distributed. It consists of the euomphalaceans Helicotoma, Straparollus, pleurotomariaceans Lophospira, Mourlonia, Cataschisma, trochonematacean Trochonema, anomphalaceans Pycnomphalus, Anomphalus, pseudophoraceans Umbonellina, Trochomphalus, murchisoniaceans Ectomaria, Murchisonia. Three genera – Anomphalus, Trochomphalus and Umbonellina – are restricted to this stage.

At the Ordovician/Silurian boundary, the gastropod fauna changed considerably – Helicotoma, Cataschisma, Anomphalus, Clisospira, Trochomphalus, Umbonellina, Ectomaria disappeared finally. The lowermost Silurian strata are characterized by a comparatively poor gastropod association, dominated by new species of pleurotomariaceans together with the first loxonemataceans (Isakar 1990). In the Wenlock gastropods are more abundant, with 12 genera being recorded from the Jaani and 8 genera from the Jaagarahu Stage. Practically all these genera (except Murchisonia) disappear at the Jaagarahu/Rootsiküla transition. From the upper part of the Rootsiküla Stage, only Straparollus (former “Platyschisma helicites”) and Murchisonia have been recorded. In the Paadla Age, there were 9 genera of gastropods, but during the Kuressaare, Kaugatuma and Ohesaare ages, the gastropod taxa reduced in number – only 3-4 genera have been identified. The Silurian index gastropods – Straparollus (“Platyschisma helicites”) sp. and Murchisonia compressa, have been recorded from the Rootsiküla Stage (Vesiku beds) and Paadla Stage, respectively (Resheniya… 1987).

The diversity and abundance of gastropods increased essentially during the transgressive phases of basin development coinciding with the Kunda, Kukruse, Jõhvi, Vormsi, Jaani and Paadla ages (Fig. 157). It shows that gastropods tended to prefer open-shelf conditions.



L. Sarv & T. Meidla


The first data on the Lower Palaeozoic ostracodes from Estonia (Photo 40:7-10) were published in the middle of the last century by Eichwald (1854a, 1860), Schrenk (1854) and Schmidt (1858). The latter published a monograph on Silurian leperditiids (Schmidt 1873). In the first half of this century, the ostracodes from the certain stratigraphical levels (Volkhov, Uhaku, Kukruse, Aruküla stages) were described (Bonnema 1909, Öpik 1935a, b, 1937a). The regular study of Estonian ostracodes was started in the 1950s by Netskaya (1953, 1958, 1966, 1973), Stumbur (1956) and Sarv (1959, 1968, 1977, 1980). Currently, these studies are carried on by Meidla (1996).

The earliest ostracodes, known from Estonia, are bradoriids Bradoria? estonica Melnikova and Konicekion kundaensis Melnikova from the Lower Cambrian Tiskre Formation. The mass appearance of ostracodes took place in the Early Ordovician Volkhov Age with incoming of the oldest ctenonotellids, eurychilinids and leperditellids (Sarv 1972). The first tvaerenellids, bolbinids and tetradellids appeared at the end of the Early Ordovician (Kunda Age). The Middle Ordovician (Aseri to Rakvere ages) was an acme for the representatives of Beyrichicopa - especially ctenonotellids, also tvaerenellids and tetradellids. The species diversity increased considerably from the beginning of this period reaching the peak during the Kukruse Age. Noticeable renovation of the fauna in the Idavere Age was followed by the period of stabilization lasting until the Keila Age. At the beginning of the Oandu Age most of the species were replaced and the number of metacopids began to increase, reaching 1/3 by the end of the Ordovician. During this period tetradellids, oepikellids, tvaerenellids and bollids were dominating among beyrichicopids, while the importance of ctenonotellids decreased considerably. The highest species diversity for the Ordovician Period was reached at early Pirgu time (ca 120 taxa according to Meidla 1996).

By the beginning of the Silurian, a rich and diverse Ordovician palaeocope fauna disappeared and only podocopes continued their evolution. In the Early Llandovery the first, strictly Silurian craspedobolbinids appeared, the oldest beyrichiids came in during the Middle Llandovery. The Wenlock Epoch was marked by the appearance of early cavellinids and Silurian primitiopsids. The species diversity of the Silurian ostracodes reached the maximum in Ludlow time, partly also at the beginning of Pridoli, gradually decreasing by the end of the period.

The Devonian ostracodes of Estonia are poorly known. Some specimens of leperditiids (probably Leperditia geographica Hecker) were found from the Middle Devonian Narva Stage and 7 species of different families were described by Öpik (1935a) from the Aruküla Stage. These species are the most ancient Devonian ostracodes recorded from the East- European Platform.

Altogether, some 400 Ordovician and 300 Silurian ostracode taxa have been identified in Estonia. In the stratigraphical distribution of the Ordovician ostracodes the Lower Ordovician (Volkhov and Kunda stages), lower Mid-Ordovician (Aseri - Kukruse stages), middle Mid-Ordovician (Idavere - Keila stages) and the topmost Mid-Ordovician - Upper Ordovician (Oandu - Porkuni stages) complexes can be distinguished. The vertical range of the Silurian ostracodes has been divided into the Llandovery (Juuru and Raikküla stages), topmost Llandovery - Wenlock (Adavere - Rootsiküla stages) and Ludlow - Pridoli complexes. All stages have their own characteristic species or species complexes, underlying the subdivision of sections and stratigraphical correlations (Sarv 1959, 1968, Kaljo 1970c, Meidla 1996). On the basis of index species, ostracode zones have also been established (Table 7, see also Meidla & Sarv 1990).

During the Early Ordovician, the species assemblage of ostracodes was rather uniform throughout Estonia. In the Middle and Late Ordovician, the ostracode faunas were different in northern and southern Estonia. In the lateral distribution of the Silurian ostracodes some characteristic features have been mentioned (Nestor & Einasto 1977). Lagoonal sediments have yielded only representatives of large leperditiids characteristic of the dolomitic rocks of the Raikküla and Rootsiküla stages. Ostracodes of the shoal facies belt are less studied. Most of the Silurian ostracodes of Estonia have been found from the sediments of the open shelf facies belt. They differ noticeably from the ostracodes of the transitional facies belt, the differences being more distinct in the Ludlow and Pridoli.

The Ordovician and Silurian ostracodes have been successfully used for correlation purposes within the limits of the Baltic region. More long-distance correlations based on the common genera and species are possible, first of all, with Podolia, Norway, Great Britain and Canada (Abushik & Sarv 1983, Vannier et al. 1989). For a long time, ostracodes have been applied to dating of erratic boulders in northern Germany.


Ordovician trilobites

A. Rõõmusoks


The Ordovician trilobites of Estonia (Photo 41:2-4) were described in a series of profound monographs by Schmidt (from 1881 to 1907) and Holm (1886). Important additions were later provided by Öpik (1937b) and Männil (1957).

During the Tremadoc and the Hunneberg Age, in Estonia the environment was unfavourable for trilobites. As a result of an environmental change during the following Billingen Age, a varied trilobite fauna appeared in northern Estonia (H. Pärnaste and particularly V. Jaanusson, personal communications). It consisted mainly of genera known from contemporaneous or even earlier strata of the Central Baltoscandian Confacies Belt in Sweden. In northern Estonia, the trilobite fauna remained diverse throughout the rest of the Ordovician (altogether some 100 genera or subgenera, see Table 38). The “Chasmops” praecurrens, “C.” maximus and “C.” eichwaldi groups are virtually new genera with several new species not yet described.

Faunal differences between the North Estonian and Central Baltoscandian confacies belts are clearly reflected also in the distribution of trilobites. The central belt extends to southern Estonia (Männil 1966, Männil et al. 1968, Jaanusson 1976), but as chances of finding macrofossils in drill cores are limited enough, there is at present no information available from the Lower Ordovician of southern Estonia. In the Middle and Upper Ordovician, 19 genera have been established in both northern and southern Estonia, whereas 5 genera of the central belt (“S” in Table 38) have never been recorded from northern Estonia. The differences between the North Estonian and central belts are still more conspicuos when the faunas from individual stages are compared.

Biogeographically, the benthic fauna of the Estonian Ordovician can be regarded as belonging to a separate Baltoscandian Province (for a review see Jaanusson 1979). With regard to trilobites, this province is characterized by a great taxonomic diversity of large asaphid trilobites, especially Megistaspis, Ptychopyge and the related genera in the Ontika Series, and Neoasaphus and the related forms in the Viru Series (“Asaphid Fauna”, Whittington 1966; “Asaphid Province Fauna”, Whittington & Hughes 1972). Further characteristics include the taxonomic diversity of the Pterygometopinae in the upper Ontika and Viru rocks (Jaanusson & Ramsköld 1993), Chasmopinae in the Viru and Harju series and various lichids.

The temporary immigration of benthic genera, whose affinities are mainly with the North American midcontinent region, is in the trilobite fauna of northern Estonia reflected by the appearence of Bumastoides and, possibly, Achatella (Achatella).

There is an apparent general tendency towards an increasing difference in the trilobite faunas of the North Estonian and Central confacies belts from the mid-Viru upwards with a culmination in the Harju Series. The normal Upper Ordovician fauna of the central belt, mostly dominated by trinucleids, has a few genera in common with the North Estonian Belt. The uppermost Ordovician Hirnantia fauna with its characteristic trilobites Dalmanitina (Mucronaspis) and Brongniartella, is known also from southern Estonia.


Silurian trilobites

R. Männil


The Silurian trilobites of Estonia were described by Nieszkowski (1857, 1859), Holm (1886) and Schmidt (1881-1907) in the last century already. In this century particular families have been studied, such as encrinurids (Rosenstein 1941, Männil 1958e, Männil R. P. 1977a, Edgecombe & Ramsköld 1996), calymenids (Männil R. P. 1977b, 1983), phacopids, etc. As a result, about 35 trilobite genera (Table 39) and 100 species have been recognized up to now. Most common representatives of the Silurian trilobite fauna in Estonia are calymenids, encrinurids and proetids which form more than half of the described species. Encrinurids prevailed in the Llandovery, calymenids and proetids in the Wenlock, and particularly in the Late Silurian.

During the Late Ordovician, the typical Ordovician families gradually disappeared and only a few genera crossed the Ordovician/Silurian boundary. The basal Silurian is recognized by the appearance of the family Phacopidae and several new genera, including Calymene and the typical Llandovery genera Acernaspis, Opsypharus, Elsarella, etc. The lowermost Llandovery contains an impoverished fauna. A relatively diverse and rich assemblage is known from the Tamsalu Formation of the Juuru Stage containing Acernaspis, Calymene and encrinurine trilobites of the “Encrinurus” variolaris plexus (Strusz 1980, Edgecombe & Ramsköld 1996) in the shoal facies, and locally abundant Opsypharus and Stenopareia in bioherms. From the overlying shallow-water deposits of the Raikküla Stage, only sparse trilobites have been found. In southern Estonia, the deeper-water facies of the Juuru and Raikküla stages are characterized by rare trilobites, predominantly by different species of Acernaspis and encrinurines. The maximum rise in species diversity has been recorded from the uppermost Aeronian - Rumba Formation of the Adavere Stage, characterized by the appearance of a number of new species and some short-ranging genera (Radiurus, Distyrax, a.o., Männil 1992). In general, the fauna of the Rhuddanian and Aeronian ages was relatively homogenous in terms of the dominating genera. The following Telychian transgression caused a notable evolutionary change. In Estonia, it coincides with the boundary between the Rumba and Velise formations and is characterized by the disappearance of the genus Stenopareia and encrinurine genera of variolaris plexus which are replaced by Encrinurus and Wallacia. At this boundary an almost complete turnover of species assemblage took also place.

Across the Llandovery/Wenlock boundary, the trilobite fauna changed remarkably. The genus Acernaspis disappeared finally, while Proetus (s.str.) and Dalmanites came in, the latter being abundant in southern Estonia. Calymenids became prevalent occurring in all trilobite-bearing facies. The role of proetids increased. During the Jaagarahu Age, the genera Cyphoproetus, Pseudotupolichas and some new species appeared. Higher up in the sequence, the diversity decreases considerably due to long regression, and no trilobites have been found from the uppermost Jaagarahu and Rootsiküla stages.

The Upper Silurian trilobite faunas of Estonia are of shallow-water origin and unilateral, greatly dominated by Calymene and Pulcherproetus. No trilobites are known from the lower part of the Paadla Stage. The species, recorded from the Uduvere beds, belong to Struszia, Pulcherproetus, Calymene and rare lichids. Trilobites of the Kuressaare Stage are of very low diversity. Only Calymene and Pulcherproetus are known from the Kudjape beds.

In the Pridoli, the diversity and frequency of trilobites increased. The open shelf sediments of the Kaugatuma Stage yield numerous specimens of Pulcherproetus and three species of Calymene, accompanied by rare representatives of Eophacops and Acaste. The same genera occur also in the Ohesaare Stage, but the fauna differs on the species level.

In general, the Silurian trilobites had a wide environmental range, but because of their strong dependence on lithofacies, their generic and specific compositions changed considerably along the palaeoslope of the Baltic Silurian Basin (Männil R.P. 1986). Accordingly, several trilobite communities, related to the different facies belts, have been established (Männil 1982a, b). In the Llandovery, the shallow-water shoal facies belt was dominated by a styginid-illaenid fauna which was replaced by a calymenid-encrinurid fauna in the open shelf environment and an encrinurid-phacopid fauna in the transitional and depression facies belts. During the Wenlock, calymenids and lichids (in bioherms) were prevailing in the shoal belt, being replaced by the encrinurid-proetid-calymenid fauna in the open shelf environment and by a rich dalmanitid-calymenid fauna in the transitional and depression facies belts. The Late Silurian trilobite faunas were similar to each other by dominating genera because of their more or less uniform shallow-water origin all over the area in consideration.



L. Hints & G. Stukalina


In spite of their great importance in the shallow water faunal associations, echinoderms in Estonia have been studied insufficiently and unevenly. On some levels, the dissociated skeleton elements of different echinoderms, especially of pelmatozoans, form an essential part of the bioclastic material (skeletal sand) in the composition of carbonate rocks (5 - 35%, in some cases up to about 90%, see Põlma 1982). Frequent occurrence of some cystoids is characteristic of distinct beds of the Middle Ordovician, the “Echinosphaerites Limestone” in the lower part and the “Hemicosmites Limestone” in the upper part of the Viru Series (see Chapter IV).

Eichwald (1860) was the first to describe Estonian echinoderms. Afterwards they were studied by Schmidt, Jaekel, Hecker, Ralf Männil and others (references see Hecker 1964). According to Ralf Männil (unpublished data), in Estonia some 50 genera and 150 species of Ordovician echinoderms have been established by more or less complete skeletons. They were dominated by cystoids and crinoids (more than 100 species). To the last group belong also most of the taxa established by columnals.

The diversest and richest associations of echinoderms (cystoids, crinoids, eocrinoids, edroiasteroids, carpoids, paracrinoids, ophiocystoids, a. o.) occurred in northern Estonia. In southern Estonia, the environmental conditions were presumably quite unsuitable for the distribution of echinoderms during the most of the Ordovician.

In Estonia the first echinoderms (crinoids) appeared in the Billingen Stage (Table 40), in the glauconitic sandstones of the Mäeküla Member. Upwards, in the Volkhov and Kunda stages, several cystoids (Cheirocrinus, Echinoencrinites) and eocrinoids (Bockia, Rhipidocystis) appear. The upper part of the Kunda Stage is characterized by the first appearance of some crinoids known by columnals (Babanicrinus, Schizocrinus, Baltocrinus) which are widespread in the Middle Ordovician rocks (Table 40).

In the lower and middle parts of the Viru Series, the echinoderms are represented by the largest number of taxa in the Ordovician. At the base or in the lower part of the series, there appear new cystoids (Echinosphaerites, Scoliocystis), crinoids (Hoplocrinus) and some eocrinoids (Rhipidocystis, Bockia) become frequent. In offshore facies in central and southern Estonia, some taxa (e.g. Echinosphaerites) have a wider stratigraphical distribution than in onshore facies in northern Estonia. Rich and diverse association of echinoderms characterizes the Oandu Stage in northern Estonia (Table 40). At that, the carbonate mounds of the Vasalemma Formation comprise edrioasteroid Cyathocystis which seems to form a frame in some mounds. The inter mound bioclastic limestones consist mainly of pelmatozoan columnals and skeleton elements of cystoids. Among the latter ones, the most well-known is Hemicosmites.

In the upper part of the Viru Series (Rakvere Stage) and in the Harju Series, the distribution of echinoderms is uneven. They are rare in the micritic (aphanitic) limestones, but their columnals are abundant (Table 40) in the talus facies of reefs or in the argillaceous limestones.

The data on the distribution of echinoderms in the Silurian sequence of Estonia are very scanty. The pelmatozoan columnals occur (in places abundantly) in the argillaceous limestones, but there are few data on their taxonomic composition. In the Adavere Stage, Myelodactylus and Glyptocrinus can be mentioned as taxa distributed widely in the Llandovery rocks in the East-European Platform and some other regions. Inadunate crinoid Pisocrinus is characteristic to the Jaani and Jaagarahu stages, to the deep-water deposits lying between graptolite and shelly facies (Rozhnov et al. 1989). The Paadla and Kuressaare stages are characterized by the occurence of Crotalocrinites which presumably appears first in the uppermost Llandovery, in the Adavere Stage. In the lowermost Pridoli, in the Kaugatuma Stage, Crotalocrinus is a rock-forming fossil in the cyclically recurrent deposits of crinoidal limestones.

The data available from the uppermost Silurian suggests that Leptocrinites, Eucalyptocrinites and Anthinocrinus are of biostratigraphic importance for the Kaugatuma Stage and Hexacrinites and Cicerocrinus for the Ohesaare Stage.



D. Kaljo


In the Early Palaeozoic biota of Estonia, graptolites played a leading role only in a few time intervals and in limited areas. They occurred in the Tremadoc of northern Estonia and in the Aeronian and Early Wenlock of southwestern Estonia. Except for the Tremadoc Dictyonema Shale, these graptolite-bearing rocks (marlstones, argillites) give evidence of the distribution of deeper, outer shelf margin and basin facies in Estonia.

In carbonate rocks of the post-Tremadoc Ordovician and Silurian, graptolites occur sporadically in Estonia. Beside rare finds of graptolite remains on slab surfaces, many new occurrences have been established by processing rock samples to obtain organic-walled microfossils (Männil 1976, Kaljo & Männil 1990).

Taxonomically, these graptolite associations from argillites and carbonate rocks are completely different: nearly all Tremadoc graptolites belong to the order Dendroidea. From the Varangu Stage Didymograptus (?) primigenius has been identified (Kaljo & Kivimägi 1976); in most cases it is probably a dendroid Kiaerograptus (bithecas were not seen because of poor preservation). In the Tremadoc sequence of Estonia, the following succession of biozones has been established (in ascending order): the Rhabdinopora flabelliformis (with R. f. socialis and R. f. norvegica, etc.) and R. anglica - R. multithecata zones in the lower and the Adelograptus - Kiaerograptus Zone in the upper Tremadoc. The lowermost Tremadoc zone (R. parabola) has not been established in Estonia.

Higher in the Ordovician and Lower Silurian benthic dendroids (Dictyonema, Thallograptus, Inocaulis, Estonicaulis, etc., Obut 1953, Obut & Rytsk 1958) occur sporadically.

A specific assemblage of dendroid graptolites (Rhadinograptus jurgensonae, Leveillites, Mastigograptus, Crinocaulis, etc., Obut 1960) was described from the Raikküla Stage in central Estonia (Kursi, Laeva, Survaküla, Kanaküla, Nurme, etc. core sections). However, a graptolite origin of some of these fossils is doubtful.

In the post-Tremadoc Ordovician and Lower Silurian, the sporadic occurrences and, naturally, the graptolite-bearing facies were dominated by planktic representatives of the order Graptoloidea. The last order was a rapidly evolving group providing good key species for correlations and biozonation.

From the Ordovician carbonate rocks of Estonia, Öpik (1927) and Jaanusson (1960b) described several species of graptoloids. As a summary, Männil (Kaljo & Männil, 1990) listed the following most significant graptoloids: Didymograptus pakrianus from the Kunda Stage, D. acutus from the Aseri Stage, Climacograptus distichus from the Lasnamägi Stage, Gymnograptus linnarssonii from the Uhaku Stage, Amplexograptus bekkeri and Nemagraptus gracilis from the Kukruse Stage and A. cf. fallax (= A. baltoscandicus, Jaanusson 1995) from the Haljala and Keila stages, Pseudoclimacograptus cf. scharenbergi from the Haljala Stage, Climacograptus spiniferus from the Rakvere Stage and from the lower part of the Nabala Stage, Archiretiolites regimontanus and Rectograptus gracilis from the Saunja Formation (upper part of the Nabala Stage). The last species seems to occur until the lower part of the Pirgu Stage. From the uppermost Pirgu Stage (Kabala Formation), Climacograptus supernus has been recorded.

The same sporadic style of occurrences continued in the Silurian carbonate rocks: e.g. Pseudoclimacograptus hughesi in the Juuru Stage, Paraclimacograptus estonus and Coronograptus aff. gregarius in the Raikküla Stage, rare specimens of Climacograptus in the lowermost part of the Adavere Stage. However, a more complete picture on the graptolite diversity changes can be obtained from the subsurface sections south of Pärnu and in the Gulf of Riga. Based mostly on the data from the Ikla (Kaljo & Vingisaar 1969), Holdre (Ulst 1970), Ohesaare (Kaljo 1970a) and Ruhnu (unpubl.) cores, the following succession of graptolite zones has been established (see Table 8): Dimorphograptus confertus, Coronograptus cyphus, C. gregarius and Demirastrites triangulatus, a gap in the succession in the territory of Estonia, Spirograptus turriculatus, Monograptus crispus, Monoclimacis griestoniensis, Oktavites spiralis (partly), Cyrtograptus murchisoni, Monograptus riccartonensis. Higher in the Estonian sections, the graptolites become sporadic again; only Pristiograptus sardous, Monograptus flexilis, M. ex. gr. flemingi and Gothograptus nassa have been recorded from the upper part of the Wenlock sequence at Ohesaare (Kaljo 1970a).

All these graptolite data have proved highly valuable for stratigraphical correlations.



V. Viira & P. Männik


The history of study on conodonts dates back 140 years (Pander 1856). The systematic study of the Ordovician and Silurian conodonts of Estonia (Photo 42:1-14) started in 1966. In a number of papers and two doctoral theses (Viira 1974, Männik 1992b) the ranges of the Lower Palaeozoic conodonts were established and the faunas of all main stratigraphic units were described. Conodonts from the Upper Silurian (Viira 1982b, 1983), Upper Llandovery (Männik 1992a, Jeppsson & Männik 1993) and the Cambrian-Ordovician boundary beds (Kaljo et al. 1986, 1988a, Viira et al. 1987) have been studied in more detail. Multielement taxonomy in conodont studies was accepted in the seventies. New Silurian conodont apparatuses were described and illustrated by Viira (1982b, 1994) and Männik (1992a, b).

The Estonian Lower Palaeozoic permits to trace the evolution of the conodont faunas from the beginning of their appearance in the Late Cambrian through their highest diversity in the Early Ordovician up to the slow decline in the Silurian. All the seven known orders of Conodonta and representatives of 27 families (from 43) have been found in Estonia (Table 41).

The oldest conodonts are proconodontids belonging to the family Cordylodontidae. They are most numerous in the Cambrian-Ordovician boundary beds (Kaljo et al. 1986, 1988a). The Early Tremadoc zonation is based on the evolutionary lineage of Cordylodus. All zones of this age are defined in the Kallavere and Türisalu formations. The appearance of the first striated coniform protopanderodontids (Paltodus, Paroistodus, Drepanoistodus, Drepanodus, Scolopodus, etc.) marks the beginning of a very diverse conodont fauna in the Varangu and Hunneberg stages. The occurrence of the first prioniodontids with combshaped or platform P elements (Oistodus, Acodus, Oelandodus, Prioniodus, Oepikodus, Periodon, Baltoniodus, etc.) followed in the Billingen and Volkhov stages (Viira 1966, 1970, Mägi et al. 1989). Species and subspecies of Baltoniodus are the most frequent components of the fauna in the interval of the Kunda to Kukruse stages. The evolutionary lineage of platformed prioniodontid genus Eoplacognathus is used as a basis for the zones and subzones in the same stratigraphic interval. During the Aseri Age, the laterally furrowed coniform panderodontid genus Panderodus appeared passing the rest of Ordovician and Silurian. The genus Amorphognathus, the best known member of the family Balognathidae, is widespread, especially in the Kukruse Stage (A. tvaerensis), and has a long range upwards. The A. superbus Zone corresponds to the interval from the Keila Stage up to the Nabala Stage, and the A. ordovicicus Zone to the Vormsi, Pirgu and Porkuni stages. Some representatives of the prioniodontids are also known from the Upper Ordovician, e.g. the earliest Icriodella species, Phragmodus and Plectodina have been found in the Vasalemma Formation. A belodellid Hamarodus with rastrate elements is characteristic of the Nabala Stage, particularly of the Mõntu and Saunja formations. The occurrence of the youngest protopanderodontids – Strachanognathus – is limited to the Pirgu Stage. By the end of the Ordovician, the diversity of conodont faunas considerably decreased - only 10 families of 21 survived during the end-Ordovician extinction.

The Ordovician and Silurian boundary is marked with the first appearance of ozarkodinids and prioniodinids (exceptional is the genus Erraticodon occuring in the Middle Ordovician already). Conodonts are rare in the lowermost Silurian. Only a single prioniodontid (Distomodus), a pectiniform ozarkodinid (Ozarkodina), a prioniodinid (Oulodus) and some coniforms (Walliserodus, Dapsilodus, Decoriconus) characterize the Juuru Stage (Männik 1992a, b, Männik & Viira 1993). The diversity of the conodont faunas considerably increased during the Raikküla Age when the earliest representatives of the platform ozarkodinids (Kockelella) and pterospathodontids (Pranognathus) appeared. An extremely rich fauna, dominated by the pterospathodontids (Pterospathodus, Apsidognathus, Astropentagnathus), ozarkodinids (Aulacognathus, Ozarkodina) and panderodontids (Panderodus, etc.) characterizes the Velise Formation and the lowermost part of the Jaani Formation (Männik 1992a, b, 1994, Männik & Viira 1993, Jeppsson & Männik 1993). In the Jaagarahu Stage and upwards, the conodont fauna is represented mainly by ozarkodinids Ozarkodina, Kockelella, Ctenognathodus; the prioniodinid Oulodus. Ctenognathodus murchisoni is characteristic of the Rootsiküla Formation, particularly of the Vesiku beds on Saaremaa (Viira 1982b). Mainly three species – Ozarkodina confluens, O. excavata, and Oulodus siluricus – are abundant in the fauna of the Paadla Stage. Ozarkodina roopaensis is characteristic of the Sauvere and Himmiste beds on Saaremaa. Biostratigraphically important are Ozarkodina snajdri and O. crispa in the Paadla and Kuressaare stages. The fauna of the three succeeding units – the Kuressaare, Kaugatuma and Ohesaare stages – consists mainly of three species and their subspecies Ozarkodina confluens, O. remscheidensis and Oulodus elegans (Viira 1982a, 1983).

The Ordovician conodont faunas from Estonia belonged to the North Atlantic Province (cold-water realm). The conodont zone succession, generally accepted for this province, is applicable also in Estonia (Table 42).

The world-wide conodont zonation for the Silurian System, recently revised, is basic for the local Estonian zonation. As many Silurian conodont species have evident ecological control, the near-shore zones have been additionally introduced (Table 43).


Early vertebrates

T. Märss


Silurian vertebrates (agnathans and fishes) of Estonia (Photo 43:1-3) have been studied relatively well. The first description of an osteostracan Thyestes verrucosus was published by Eichwald (1854a). Later on several papers and monographs have dealt with the morphology and microstructure of the exoskeletons of agnathans and fishes, with their biostratigraphy, phylogeny and palaeoecology (Pander 1856, Rohon 1892, etc., Robertson 1938, etc., Hoppe 1931, Gross 1967, etc., Märss 1986, etc.).

Estonian Silurian vertebrates were a well advanced and diverse group of animals containing several representatives of early agnathans and gnathostomes. The first finds come from the Jõgeva beds of the Raikküla Stage (Middle Llandovery). However, in other regions vertebrate remains have been recorded from much earlier strata. A problematic agnathan Anatolepis has been described from the Upper Cambrian to the Lower Ordovician rocks from Spitsbergen, Greenland and North America. The oldest confirmed agnathans, closely related to the heterostracans, come from the Lower Ordovician of Australia (Ritchie & Gilbert-Tomlinson 1977). The first thelodonts, another group of agnathans, have been described from the uppermost Ordovician of the Timan-Pechora Region (Karatajute-Talimaa 1997). Gnathostomes also appeared in the (?) Late Ordovician, but they became widely distributed beginning from the Early Silurian: acanthodians in the Baltic area (Märss 1986) and China (Pan Jiang 1988), elasmobranchs in Mongolia (Karatajute-Talimaa et al. 1990, Karatajute-Talimaa & Novitskaya 1992), placoderms in the mid-Silurian of China (Pan Jiang 1988). The earliest osteichthyans - actinopterygians are known from the Late Silurian of Baltic (Märss 1986, Fredholm 1988) and the Central Urals.

Thus, the appearance of early vertebrates was a gradual process which lasted through the whole Ordovician. In the Early Silurian, all main agnathan groups (thelodonts, heterostracans, osteostracans, anaspids, galeaspids) already existed, and during the Silurian, representatives of all higher taxa of fishes (placoderms, acanthodians, chondrichthyans, osteichthyans) appeared. The Silurian was the period of several innovations and radiations in the evolution of early vertebrates (Kaljo & Märss 1991).

From the Silurian of the Baltic area, 55 vertebrate species belonging to 33 genera have been recorded (Table 44). During that period, three main evolutionary stages of vertebrates can be documented.

(1) The Raikküla - Jaagarahu evolutionary stage when single species of thelodonts, an osteostracan, an anaspid and acanthodians occur.

(2) The Rootsiküla - Paadla stage with a wide distribution of thelodonts, anaspids, osteostracans (in the Paadla Stage also heterostracans and actiopterygians).

(3) The Kuressaare to Ohesaare stage with a diverse vertebrate fauna of both agnathans and gnathostomes.

Using the distribution of vertebrate species in Estonia and Latvia, the regional biozonal succession has been compiled (Märss 1982, improved in 1990 and 1996, see also Table 8) which served as a base for the Global Vertebrate Zonal Standard (Märss et al. 1995).

The Silurian vertebrates, belonging to the nekton and nekto-benthos, inhabited all facies belts of the Palaeobaltic tropical sea: lagoonal, shoal, open shelf, transitional and depression (Märss 1986, 1989, 1990). The skeletal elements of the same vertebrate species occur both in the carbonate and terrigenous rocks that allows to treat the corresponding terrigenous sediments and fish remains in them as marine and not as continental (Märss 1989). During the Early Silurian and the beginning of the Late Silurian in the Palaeobaltic Basin the agnathans preponderated over gnathostomes, especially in the shallower parts of the sea. Beginning from the Ludlow in the deeper-water open shelf and transitional environments the gnathostomes prevailed, and during the Pridoli gnathostomes predominated in all facies belts.

Judging by palaeogeographic reconstructions, the distribution of the Silurian vertebrates in the Baltic region, northern North America and northern Greenland shows that they were mainly tied to the tropics, occurring between latitudes 40 degrees north and south. Agnathans and fishes with different species dwelled also in the temperate zone (e.g. finds from southern Siberia and Bohemia). Five faunal provinces can be distinguish on the base of vertebrates: the eastern and northern Europe and Severnaya Zemlya, Siberia, Tuva, China, and North America (at least partly).


Devonian fishes

E. Mark-Kurik


The study of the Devonian fishes in Estonia started in the fourth decade of the 19th century. During the current century the monographical studies were carried out by Gross (1933, 1940a, b, etc.) concerning all fossil fish groups, by Heintz (1928, 1930, 1934), Mark-Kurik (1973, Mark 1953a) and Obrucheva (1962, 1966) on arthrodires, by Karatajute-Talimaa (1960, 1963) on antiarchs, by Obruchev and Mark-Kurik (1965, 1968) on psammosteid heterostracans, by Vorobyeva (Vorobyeva 1977) on crossopterygians and by Valiukevičius (1985) on acanthodians

The fishes, known from Estonia, characterize mainly the flourishing period of the Devonian fish faunas coinciding with the Middle Devonian. Biogeographically, they belong to the eastern part of the Euramerica Province (Dineley & Loeffler 1993). All main groups, such as agnathans, placoderms, acanthodians, crossopterygians, dipnoans and actinopterygians, are present showing a great diversity. Dominating forms are psammosteid heterostracans which outside the NW of the East-European Platform are uncommon. Placoderms (arthrodires, antiarchs, Photo 43:6) are also numerous. Real giants (Photo 43:4-5) occur among the representatives of the above groups (e.g. psammosteids Tartuosteus, Pycnosteus and arthrodires Homostius and Heterostius), but also in other groups (Crossopterygii, Dipnoi). Acanthodians are frequent members of fish assemblages. Tesserated cephalaspidids and actinopterygians, known by their microremains, seem to be quite common. Rare chondrich-thyan remains have also been discovered. Decline of the Middle Devonian fish fauna can be observed roughly since the Gauja Age and further on when the assemblages became much less variable (Tables 45, 46).

During the last decade, biozonation based on different fish groups was worked out and included into the Devonian correlation chart of the East-European (Russian) Platform (Rzhonsnitskaja & Kulikova 1990). However, the zonal subdivision of the Devonian in Estonia and Latvia is largely based on publications by Gross (1942, 1950). Gross established the psammosteid, antiarch and some Upper Devonian acanthodian zones (Mark-Kurik 1993c). The acanthodian zonation has been elaborated by Valiukevičius (1994, 1995, Valiukevičius et al. 1995, etc.). A summary of the recent zonal schemes is given by Mark-Kurik (Blieck et al. 2000; Table 47). The rather detailed psammosteid and placoderm (antiarch and arthrodire) assemblage zones, especially characteristic of the Middle Devonian, have been established on the basis of macroremains collected from exposures. The acanthodian bio- or interval zones, based on microremains (scales), are more universal as they can be distinguished using samples from both outcrops and drill cores. But the stratigraphical ranges of the Middle Devonian acanthodian zones, particularly those of the late Middle Devonian, are comparatively long.

In some cases, the dependence of the content of the fish assemblages on the character of the bottom of the basin (sandy, muddy) is rather apparent. The representatives of psammosteids and cephalaspidids obviously preferred sandy bottom (Valiukevičius et al. 1986). In establishing the completeness of the faunal assemblages, the reconstruction of trophic relations was used (Mark-Kurik 1995). Incomplete assemblages may be caused by the taphonomic loss. In the relatively complete assemblages various secondary and tertiary consumers were represented beginning with the top consumer, the largest predator.

The stratigraphical ranges of fish species, mostly the Middle Devonian ones, are given in Tables 45 and 46. Among psammosteids and placoderms (Table 45) the species confined to one or two stratigraphical units of lower rank (members, beds) are rather numerous whereas the acanthodians (Table 46) show a limited number of species with short range.



Bedrock topography

E. Tavast


The Estonian bedrock is cuesta-like and more complicated in southern Estonia. The maximal amplitudes of absolute heights of the bedrock surface are close to those of the modern topography, being minimal in the ancient valleys (Harku Valley -145m) and maximal in the Haanja Heights (+166m).


History of research

In Estonia, the peculiarities of the bedrock topography are easier to study than in neighbouring areas, because the Quaternary cover is rather thin, especially in northern Estonia (Fig. 158).

The research into the Estonian bedrock topography goes back to well over a century. The first information on the bedrock topography was issued by Schmidt (1854, 1883b) who described the cuesta‑like topography of Estonia with abrupt northern and rather flat southern slopes. Later, main attention focused on single relief forms, particularly the escarpments. The Sub‑Cambrian Peneplain (Giere 1938, Fromm 1943), Cambrian and Silurian escarpments (Jaanusson 1944a, Martinsson 1958), Vendian (Witting 1910, Nilsson 1913, Büchting 1918), Ordovician (Tammekann 1928, 1940, Martinsson 1958, Künnapuu 1959, Miidel 1992) and Silurian (Jaanusson 1947, Aaloe 1958a, Aaloe & Miidel 1967) escarpments were described in particular detail.

The first scheme of the bedrock topography was compiled by Tammekann (1928, 1949). More detailed zonation, based on the lithological complexes, was carried out by Orviku (1955). He distinguished several large features, including the Depression of the Gulf of Finland, the North-Estonian Escarpment (Photos 2, 18), the North-Estonian Plateau with the Pandivere Upland, the Middle-Estonian Lowland, the Middle- and Upper-Devonian plateaus, the depressions of Võru-Piusa, and lakes Võrtsjärv and Peipsi. A large quantity of information about the bedrock topography has accumulated in the Geological Survey of Estonia which was summarised by Kajak (1966). He was the first to show the isolines of the bedrock topography after every 20 m. To the large features defined by Orviku, he added the West-Estonian Lowland and several medium forms, including the Ahtme Elevation and the Luuga-Narva and Ojamaa lowlands. On a more detailed map compiled by Miidel (Raukas et al. 1971), the isolines were drawn after every 10 m. In 1977, a map of the bedrock topography of Estonia was composed by Tavast on a scale of 1:200 000 (Tavast & Raukas 1978). It was based on the data from nearly 2500 boreholes and dugholes, but also on the materials of geophysical evidence and earlier maps. Some information has also been obtained on the bedrock topography of the coastal areas of the Baltic Sea (Amantov et al. 1988, Tavast & Amantov 1992). In the Gulf of Finland, the Sub-Vendian, Vendian-Cambrian and Ordovician escarpments were distinguished. Owing to the geological mapping of Estonia, there are rather top-quality maps available on the bedrock topography of many regions (Saaremaa a.o.) but, unfortunately, only the bedrock map in the scale of 1:2 500 000 (Kajak 1995) has been published until now.


Genesis of the bedrock topography

The bedrock topography of Estonia has developed under the effect of different geological processes. In the course of long-term continental period, the development of bedrock surface was controlled by erosional processes, and by the beginning of the Pleistocene, a complicated bedrock topography had formed. In the Pleistocene, the bedrock topography was significantly affected by glaciers; in the Holocene the waters of the Baltic Sea (Photo 44) and several terrestrial geological agents played an important role.

The bedrock topography has been affected by both passive and active geological factors, such as tectonic movements, heterogeneity of the lithological composition of the bedrock, erosional, gravitational and other geological processes, eluvial and karst phenomena, cosmic factors and human activities.

Some researchers point out the leading role of the tectonic movements (Doss 1913, Nikolayev 1962, Rähni 1973b), others maintain that the main relief forming factor was the long‑term denudation on the lithologically heterogenous bedrock (Schmidt 1883b, Hausen 1913a, Markov 1931b). Many researchers (Giere 1938, Fromm 1943, Tammekann 1949) stress the importance of the fluvial erosion in forming the cuesta-like topography of the bedrock.

A rather big group of scientists gives prevalence to glacial erosion (Grewingk 1879, Lewinski & Samsonowicz 1918, Isachenkov 1976). Giere (1932) maintains that the North‑Estonian Klint had retreated under the influence of glacial erosion from the line of Suur‑Pakri ‑ Lavansaari ‑ Seiskari to the present position.

According to the calculations by Isachenkov (1982) and Makkaveyev (1976), the Pleistocene glaciers had removed a layer of rock up to 50 - 60 m in thickness from the bedrock surface on the southern slope of the Fennoscandian Shield. In the ice lobe depressions, the thickness of such a layer could have reached even 100 metres.

Most of contemporary researchers are of the opinion that the bedrock surface has formed under the influence of dif­ferent geological processes (Tavast & Raukas 1982).

Two or three stages (Giere 1932, Tammekann 1949, Martinsson 1958) can be distinguished in the history of the formation of the bedrock topography. In the pre-Quaternary period, active tectonical movements and long-term denudational processes prevailed.

In the Pleistocene, the genesis of the bedrock topography was influenced by glacial erosion and erosion by interglacial rivers and the Eemian Sea. Wave erosion of different stages of the Baltic Sea and denudational continental processes changed the bedrock topography in post‑glacial time. During all geological epochs, the tectonic movements have played an important role in the formation of the bedrock topography, which is revealed through: 1) the regional sinking or decelerated denudation; 2) the differentiated movements of the blocks of the crystalline basement which formed new morphostruc-tures; 3) geological structures which have already lost their tectonical activity (faults, crevice zones). The latter forms promoted the selective denudation of the bedrock.

We have compared the block structure of the Estonian basement, bedrock and contemporary topography (Vaher & Tavast 1979, Tavast & Vaher 1982, Raukas et al. 1988) and found that direct morphostructures are infrequent in our area, and such local structures which could correspond in the plan and sign to large forms of bedrock topography are not observed in the Estonian territory.

For example, the most prominent topographic phenomenon of the area ‑ the Haanja Elevation, is located in the eastern part of the Mõniste‑Lokno plain‑type fold. On the crest of the anticline the basement lies at a depth of 230 m below sea level, descending northwards to 500 m, and southwards to 1000 m below sea level. According to the distribution of Vendian and Lower Palaeozoic rocks in southeastern Estonia, this structure was formed principally in the Late Silurian (Vaher & Tavast 1979).

The southern part of the elevation intersects with the high‑dipping slope of the anticline which amplitude accounts for more than 120 m on the floor of the Upper Devonian Pļaviņas beds. The height of the northerly dome on the above-mentioned floor amounts to 30 m. Thus, the amplitude of relatively young movements does not exceed 30 m. It forms only about one third of the relative height of the Haanja Heights in view of its bedrock surface, and less than 15 m of the relative height of topographic elevation. Consequently, these were exogenic processes, not tectonic movements, that dominated in its shaping (Raukas et al. 1988).

The question is what role, if any, linear dislocations played in the formation of the bedrock and recent topography. Extremely dense network of boreholes in the area of Estonian oil shale deposit permitted Vaher (Vaher & Tavast 1979) to draw the contour lines for that region at 1 m intervals. However, with one exception, even in this case ex­pression of the linear structures in the bedrock topography was not detected in the recent topography of northeastern Estonia. If these structures were activated by inherited postglacial movements, some attack on the de­nudated bedrock surface must have occurred though it may well be invisibly minute.

Summarizing the above, we should like to stress that the formation of the bedrock and contemporary topography has been only little affected by postglacial and recent blockwise differential movements.

Although most of linear structures and destruction zones are not reflected in the bedrock topography, they have influenced on the formation of some single forms or relief complexes. As an example serves the slope of the Pandivere Elevation in the destruction zone (Rõuk & Tavast 1982). Miidel (1966b, 1971) measured the direction of tectonic joints in the North‑Estonian river valleys crossing the Ordovician Escarpment. The results were compared with the direction of the river valleys in the places where measurements had been carried out. In all cases, tectonic joints coincided with the direction of the valleys which suggests that the linear tectonic disturbances played an important role in the formation of modern river valleys. In the same areas, the ancient valleys and karst phenomena are often connected with tec­tonic joints (Heinsalu 1977). Such connection is quite understandable, because the linear erosion and karst processes are naturally more intensive in the destruction zones of the bedrock.

In distinguishing the role of tectonic movements in the genesis of the bedrock topography, we have to mention two different aspects: the indirect and direct. The general rise of the Earth crust created the conditions for the erosion‑denudation processes and determined their intensity. Even the dead tectonic structures (joints, zones of disturbances) are highly conducive to the denudation and dissolution processes. The role of the differentiated neotectonic movements in the genesis of the single relief forms of the bedrock topography was rather modest, however, these movements had a great influence on the formation of the bedrock topography over vast areas.

In the development of all active endogenic geological processes the lithological heterogeneity has been and will be of very great importance, since it accounts for uneven glacial erosion, denudation, the specific arrangement of the river drainage, the distribution and morphology of karst forms, etc. (Tavast & Raukas 1978, Vaher & Tavast 1979). All meridio-nally elongated depressions and lowlands (Middle Estonian, Võrtsjärv, Peipsi, the Gulf of Riga and Vidzeme) have developed in the easily erodable Middle Devonian siltstones, clays and sandstones.

The lithological heterogeneity of the bedrock affected not only the genesis of large and medium relief forms, but it also controlled the formation of small ones (Photo 23). Already Schmidt (1854, 1883b) called attention to the long-time denudation in the formation of hard rock hillocks elongated from the north-west to south-east. The height of such hillocks is usually less than 10 m. The areas between the hillocks (Winkler 1920, Tammekann 1928) are composed of less durable limestones.

The intensity of erosional processes in pre‑Quaternary time, during interglacial and in late‑ and postglacial times depended on many factors, such as the amount of water, the geological structure and the shape of the catchment area, the inclination of the valleys, the stream velocity, and the climate.

The most impressive example of the pre‑Quaternary river erosion is the ancient buried valleys. Most of them probably formed in the Late Palaeogene when the Earth’s crust was much higher than at present due to the riftogenesis in the North‑Atlantic area (Puura 1980). A complicated net of valleys was formed in and around the contemporary Baltic Sea.

There are several schemes of Estonia’s buried ancient valleys compiled by many researchers (Tammekann 1928, 1949, Kajak 1970, Raukas et al. 1971, Tavast 1992).

Marine and lacustrine erosion affected the bedrock topography during interglacials and late‑ and postglacial times. Due to neotectonic movements, which were more intensive in the northwestern than in the southern part of Estonia, the shoreline was subject to rather rapid changes here and only during the transgressive phases of the Baltic Sea, when the rise in the water‑level more or less equalled the uplift of the Earth crust, the position of the shoreline somewhat stabilized and the erosion of the bedrock was considerable (Kessel & Raukas 1967).

By our opinion, the wave erosion was one of the essential factors responsible for the formation of alvars (Photo 26), in the development of which three main stages can be distinguished: pre-glacial, Pleistocene and post-glacial (Jürgenson & Tavast 1986). During the first stage, erosion and denudation were intensive. During the second stage, erosion and accumulation of gla­ciers were the most characteristic processes. In the post‑glacial stage, the Baltic Sea with its transgressions and regressions participated in the formation of alvars. After the final retreat of the sea from Estonia, various less important continental processes followed.

Human activities have affected upon the bedrock topography considerably during the last decades, especially in the regions where phosphorite, oil shale, clay and limestone are mined. Man has also influenced onto the bedrock topography by dredging rivers and laying out the foundation for constructions. As the bedrock topography is comparatively flat, the construction of high‑ and railways hasn’t brought about remarkable changes in the topography, however, this effect tends to increase. For instance, a 5-m-deep and several-kilometres‑long channel with a width of approximately 50 m, has been blasted into the Ordovician rocks at Lasnamäe in Tallinn.

The zonation of the bedrock topography of Estonia is presented on the basis of relative and absolute heights, taking into consideration the lithological composition of the bedrock. In the area under consideration the following large forms were proposed (Tavast & Raukas 1978, Tavast 1992, Figs.159, 160): I ‑ Depression of the Gulf of Finland, II ‑ Viru‑Harju Pla­teau, III ‑ Livonian Lowland, IV ‑ Middle‑Devonian Plateau; V‑ Cen­tral‑Estonian and Lake Võrtsjärv depressions, VI ‑ Lake Peipsi Depression, VII ‑ Valga and South‑Estonian lowlands, VIII ‑ Upper‑Devonian Plateau. Also the medium (Pandivere and Sakala elevations, Ordovician, Silurian (Photo 23) and Devonian escarpments), small (elongated eminences and hollows) and tiny forms (ice scratches) are distinguished.

To sum the above up, it should be pointed out that the bedrock topography has formed under the effect of different geological factors. Unlike some other geological factor, it has continuously affected the formation of the Quaternary deposits and landforms, being thus responsible for the inherited nature of the litho‑ and morphogenesis during the Quaternary as a whole (Tavast 1992).



Ice ages

A. Raukas & K. Kajak


Quaternary glaciations covered on the East-European Plain a vast territory, extending over polar and subpolar regions in the north and reaching as far as the Don and Dnieper rivers in the south. The deposits of the Lower Pleistocene are absent in Estonia and even the Middle Pleistocene sequence is rather uncomplete. Nevertheless, Estonia was one of the first regions where the theory of continental glaciations was elaborated, and the concepts of the glacial litho- and morphogenesis worked out here have contributed to solution of topical questions concerning the dynamics of ice sheets and formation of glacial landforms and deposits.

During all glaciations, Estonia was affected by the Baltic and Peribaltic ice streams which moved at different rates during different glaciations and stadials of glaciations. For example, the lithology of till beds and the orientation of clasts in tills suggest that the ice flow direction during the Late Ugandi (Warthe) and Valgjärve (Early-Middle Weichselian) glaciations was mainly from north-west to south-east, during the Early Ugandi (Saale) and Late Sangaste (Elster II) glaciations from north to south. There were naturally different local movements depending upon the bedrock topography.

Areas of accumulation and erosion remained relatively stable through time. Ancient valleys, interlobate massifs, lee-side areas of bedrock elevations, as well as escarpments oriented transverse to the ice movement, acted as areas of accumulation (Tavast & Raukas 1982). Intensive erosion took place on bedrock elevations and in ice lobe depressions, and the Quaternary cover is correspondingly very thin.

Alternation of till beds with interglacial and interstadial sediments has been controlled by the cyclically changing climate. At the beginning of all glaciations, the centre of glaciation was in the Scandinavian mountains owing to which circumstance a northwest-southeast direction of the ice movement was prevalent. As climate deteriorated further, the centre of the glaciation extended eastwards and a north-south movement of the ice became dominant. This direction is recorded by the distribution of indicator boulders in the tills. At the end of the glaciations when climate improved, the glaciation centre was transferred once again to the Scandinavian mountains, and the direction of the ice movement was again northwest-southeast (Raukas 1961).

The bedrock topography exerted an influence over the distribution of ice-marginal depositional zones which are mainly connected with the slopes of bedrock elevations and with the depressions in the bedrock. Various types of glacial landforms are also related to certain elements of the bedrock topography, which facilitates the determination of their ge-nesis.

Drumlins (Photo 3) most frequently form large crag and tail formations on the distal side of bedrock elevations, e.g. the Saadjärve Drumlin Field where before the drumlinization thick older deposits occurred (Kajak 1965b), and in depressions where the glacier desintegrated into lobes which moved at different speeds (Rõuk & Raukas 1989).

End moraines are usually distributed on the proximal slopes of elevations (Photo 45) blocking the movement of glaciers and thus causing the accumulation of till. Marginal eskers (narrow deltas) are mainly spread in bedrock depressions. Radial eskers occur under different topographic conditions, frequently in ancient valleys. Glaciofluvial deltas are characteristic of regions with the bedrock surface inclined towards the ancient glaciers (northern Estonia). Hilly glacial relief (Photo 46) is widely distributed in areas with a rapidly changing bedrock topography (Tavast & Raukas 1982). In southeastern Estonia, the formation of hilly landscapes was greatly affected by Middle Pleistocene sediments which in the inner part of the heights are much thicker than in the surrounding lowlands (Kajak 1965a, 1995).

The evidence obtained by analysing the internal structure of the basement and bedrock, the thickness of different facies within the sedimentary cover complexes, and the maps showing the structure and contours of different key beds, provide the basis for the conclusion (Raukas et al. 1988) that the tectonic framework has considerably and repeatedly changed with time. The block structure of the basement is weakly reflected in the bedrock and glacial topography. Linear elements, such as faults in the basement, are identified as the most active and long-lasting forms, but even they are rather poorly reflected in the bedrock and present topography. For instance, the Haanja Heights, which is the highest region in the Baltic Republics, is situated on the tectonically active Mõniste basement rise, but low rates of neotectonic uplift and the relatively thick Quaternary cover show that the tectonic movements have not been decisive in the formation of this highland area (Tavast & Raukas 1982).

Although the concept of the extinction of Pleistocene glaciers by way of stagnant ice fields separated from the margin of the active glacier was known already before the turn of the century, there nevertheless, still exist disagreements as to the extension and thickness of such dead ice fields (cf. e.g. discussion in Karukäpp & Raukas 1976).

The vast majority of investigators believe that the formation of stagnant ice fields was determined, first of all, by the change in the dynamics of the glaciers due to climatic conditions: ablation exceeded the inflow of the ice from the accumulation zone.

Since the hilly topography is mainly distributed in southern Estonia (Photo 45), it was assumed for a rather long time that the glacial relief of southern Estonia was formed chiefly due to the effort of stagnant ice, whereas that of northern Estonia was shaped by active ice (Kajak 1963). At that, it was supposed (Rähni 1963b) that the glacier lobes in northern Estonia continued to be active practically until their final melting. Nowadays these assumptions have been refuted and the significant role of both stagnant and active ice in both regions mentioned is quite clear (Karukäpp 1979).

In a very generalized form, Estonia’s territory may be divided into lobe depression regions with a rather flat topography reflecting the stadial and phasial halts of the glacier margin, and ice-shed regions of insular heights with a rather complicated topography which have been formed during the course of a long time interval under the influence of active, passive and dead ice (Aboltinš et al. 1989).

Most likely, the waters of the Holsteinian Sea did not reach Estonia. Continental deposits of Holsteinian (Likhvin) age have been established in the southwestern (Karuküla section) and southeastern (Kõrveküla section) Estonia (Liivrand 1984).

The Karuküla organic deposits, overlain by a thin (1 m) reddish-brown till of the last glaciation, rest on the grey till of Middle-Pleistocene age. According to Liivrand, the Holsteinian organogenic deposits at the Karuküla site are erratics. According to Kajak (1995), these deposits are glaciotectonically compressed and disturbed, but in autochtonous bedding.


Description of the section (from top downwards):


                Layers                                    Thickness in metres

Reddish-brown till, upper

                part is weathered                                                  1.00 - 2.00

Yellowish unsorted sand with inclusions

                of organic matter                                 0.10 - 1.25

Forest peat with pieces of wood                        0.15 - 0.70

Forest peat/Phragmites peat transition layer 0.05 - 0.15

Phragmites peat                                                  0.25 - 0.40

Gyttja                                                                    0.05 - 0.60

Yellowish-grey silty clay                                     0.10 - 0.35

Unsorted sand or grey till                                   0.50 +


Pollen zones in the Karuküla section after Reet Pirrus (Orviku & Pirrus 1965) and Liivrand (1984, 1991) are the following (Fig. 161):

K1 - lower part of gyttja-Betula (Betula nana) and Pinus.            Rise of Picea, Alnus and Ulmus curves;

K2 - upper part of gyttja-Picea and Alnus maxima. Quercus,     Ulmus, Trapa natans and spores of Osmunda are        present. Immigration of Tilia;

K3 - Phragmites peat - Tilia, Quercus and Ulmus maxima.

  Picea is frequent;

K4 - lower part of forest peat - Picea, Abies and Carpinus


K5 - upper part of forest peat - Betula and Pinus with Picea.

  Alnus and broad-leaved species are present.

The pollen zones of the Karuküla section (Fig. 161) correspond to the pollen zones of Holsteinian deposits in Latvia and Lithuania (Liivrand 1984, 1991). They all demonstrate a wide spread of conifers, an early appearance of Picea and Alnus (maxima occurred already before the climatic optimum) and a very limited occurence of Corylus.

Holsteinian deposits in the Kõrveküla section occur within an old buried valley near Tartu. They are represented by gyttja and loam, covered by reddish-brown till and underlain by glaciofluvial deposits. The palynological spectrum coincides with that of the Karuküla section (Liivrand & Saarse 1983).

As mentioned above, both marine and continental Eemian deposits are found in Estonia. Interglacial deposits were supposed to occur in the basin of the Gulf of Finland already long before they were discovered in the Rõngu section (Orviku 1939). Thomson (1934), in his work dealing with the finds of mammoths in northern Estonia, pointed out that all the remains of these animals (molar teeth at Paljassaar, Pirita, Ülemiste and Lüganuse; incisiors and shin-bones at Ihasalu) occurred well preserved in the fore-klint area. This is indicative of their short transportation, and refers indirectly to the occurrence of interglacial layers north of the klint. Basing on the drilling data by Mickwitz (1908) and Öpik (1929) on the islands of the Gulf of Finland where submorainic organogenous layers were reported, Zans (1936) correlated the layers with the Portlandia Sea (Skærumhede) sediments of the Eemian Interglacial in Denmark (Jessen & Milthers 1910).

Abundant places of gas emanation in the bottom deposits of the Gulf of Finland were also indicative of submorainic organogenous layers, e.g. between the Viimsi Peninsula and Malusi Island at Viinistu (Eplik 1935), on Keri Island and in several other places. On the Island of Keri, gas was used in household and in the lighthouse.

Drilling on the Island of Prangli (Fig. 91) brought some clarity to the stratigraphy of Pleistocene deposits in the Gulf of Finland (Kajak 1961). Interglacial deposits composed of greenish-grey silts and comprising subfossil molluscs and plant remains with vivianite were discovered at a depth of 66.0 - 77.7 m in seven boreholes. They are underlain by waterlain glacial deposits and two beds of brownish till with different thicknesses (6 and 15 m), and overlain by four grey tills (see the description of the section). Interglacial deposits occur in an area of about 2.5 km2 and are best preserved in borehole 6 (see the description below) which was chosen for a stratotype section (Raukas & Kajak 1995).

All the four Järva tills are separated by thin layers of waterlain glacial deposits, in places (between the upper two tills) by varved clays with plant remains which has enabled Kajak (1961) to speak about interstadial sediments.

In the area of the Gulf of Finland, the stratigraphic subdivision of the Pleistocene deposits is hampered by the limited distribution of interglacial and interstadial deposits and abundant erratics. Interglacial deposits in a secondary position stand out because of their differing lithological and palaeontological composition in the sections studied. The recognition of erratics derived from marine sediments is aided by a study of their hypsometric position. Marine Eemian interglacial deposits in the vicinity of St. Petersburg, Lake Ladoga and on the coast of Luga Bay are located above the coastline of that interglacial, which is a clear indication of their secondary position (Malahovsky & Sammet 1982). Only in the Prangli section they seem to be in situ.

Pollen diagrams, compiled and interpreted by Liivrand (1972, 1984, 1991, a.o., Liivrand & Valt 1966) show the whole cycle of development of vegetation between the Late Ugandi (Wartha) and Early (or Middle) Järva (Weichselian) glaciations (Fig. 162). Three complexes are most distinct:

(1) spore-and-pollen spectra of late-glacial deposits of the Late Ugandi glaciation characterizing varved clays and the lower part of marine sediments;

(2) spore-and-pollen spectra of marine Prangli (Eemian) interglacial deposits;

(3) spore-and-pollen spectra of early glacial deposits and the upper part of marine deposits.

Late-glacial deposits of the Late Ugandi glaciation are characterized by the following pollen zones (Fig. 162):

LUg1 (Ms1) - Betula nana L., abundant herb pollen (up to 70 %). The herbs are dominated by Artemisia (up to 60 %) and Chenopodiacea (up to 30 %). Eurotia ceratoides and Polycnemum are constantly present. The quantity of pine pollen increases at the end of the zone. The formation of varved clays and the lowermost part of grey loam is referred to that zone.

LUg2 (Ms2) - Pollen zone of tree-like birch and pine. Substantial decline in Betula nana L. and herbaceous plants. The zone indicates a short-term climatic warming.

LUg3 (Ms3) - Birch pollen zone. Birch pollen dominates the spectra (95%) consisting mainly of Betula nana L. (90%) which refers to approaching climatic cooling.

After the Late Ugandi glaciation, the climate was cold and very dry (cryoxerotic stage).

On the Island of Prangli, the marine interglacial deposits are characterized by the same pollen zones (Fig. 162) which Jessen and Milthers (1928) differentiated in the Eemian interglacial deposits in western Europe (zones c - i) and Grichuk (1961) in the Mikulinan interglacial deposits on the East-European Plain (zones M2 - M8).

c(M2) - Birch and pine pollen zone. Pollen of tree-like birch prevails. B. nana L. and B. humilis Schrank occur in small amounts.

d(M3) - Pine and birch pollen zone. Spruce disappears, hazel and afterwards also alder appear.

e(M4a) - Pine and birch pollen zone. Oak and elm also occur. The proportion of hazel and alder increases.

f(M4b) - Oak and elm pollen zone. The lower maximum of hazel and alder.

f(M5) - Willow pollen zone. The second half of hazel and alder maximum.

g(M6) - Hornbeam pollen zone. The proportion of hazel and alder decreases.

h(M7) - Spruce pollen zone. Abrupt decline in thermophilous species.

i(M8) - Pine pollen zone. Hazel, alder and hornbeam continue to occur in very small amounts.

The Eemian interglacial is characterized by two equally well-pronounced climatic stages - thermoxerotic and thermohygrotic. In the former stage, oak and elm were widely spread, whereas willow, hornbeam and spruce pollen maxima are typical of the latter stage.

According to Liivrand (1984, 1991), the further development of the vegetation is shown by the appearance of herbaceous plant associations and tundra species which are indicative of the effect of the starting climatic cooling of the Järva (Weichselian) glaciation (Fig. 162)

To this period is attributed the formation of the upper part of marine sediment complex and that of grey varved clays of the Kelnase Subformation (Table 13).

QIII jr1kl - Birch pollen zone. Strong representation of Betula nana L. Herbaceous plants are represented by sedges and loose-bunch grasses. In varved clays the pollen of Artemisia arctica (Cham.) Wallr. and the spores of Lycopodium alpinum L. and Selaginella selaginoides (L.) Link occur. The climate was cold and humid (cryohygrotic stage).

Diatom analyses, performed on the interglacial deposits from borehole 4 (Cheremisinova 1961, Znamenskaya & Cheremisinova 1962), discovered two stages in the development of the Eemian Sea which correspond to two diatom complexes. One of them characterizes the lower part of the deposits and the beginning of transgression. In the composition of the diatom complex there have been identified cold-favouring fresh-water relict forms, such as Cocconeis disculus (Shum.) Cl., Diploneis domblittensis (Grun.) Cl. a.o., and the fresh- and salt-water forms Pinnularia sp., Epithemia sp., Neidium sp., but also marine shallow-water forms, such as Hyalodiscus scotius (Ktz.) Grun., Actinocyclus Ehrenbergii Ralfs, Grammatophora sp. and others. According to Cheremisinova, the co-occurrence of offshore-marine species with cold-water relict forms marks the start of the sea-water’s intrusion into the glacial lake. The late-glacial pollen zones a+b correspond to this period.

The second complex of diatoms is typical of the period of interglacial transgression and is related to the middle part of marine loams at a depth of 60.5 - 67.5 m below sea level. The diatom complex indicates normal salinity of water, but there are shallow-water forms, such as Melosira sulcata (Ehr.) Ktz., Actinocyclus Ehrenbergii (Bail) Ralfs, Hyalodiscus scoticus (Ktz.) Grun, Grammathophora sp. a.o. Besides, there also occur typical thermophilous Eemian species: Synedra Gaillonii (Bory) Ehr., Navicula abrupta Greg., Coscinodiscus antiquus Grun., C. granulosus Grun., C. perforatus Ehr. According to the pollen evidence, this period corresponds to the interglacial unit characterized by pollen zones M2-M8.

Sea water intruded into the Gulf of Finland early, already in the post-glacial of the Late Ugandi Glaciation and marine conditions existed here throughout the whole Eemian interglacial. Figure 163 shows the two principally different constructions of the distribution of the Eemian Sea (Raukas 1991a).

According to Grichuk, the Baltic Sea was connected with the North Sea via the Skagerrak, Kattegat and Danish Sounds through the present lake system of Vänern and Mälar in Central Sweden and the area of the current Kiel Canal on the Jutland Peninsula (Gerasimov & Velichko 1982). Like Lavrova (1961), he assumes a connection between the Eemian and White Sea basins through the system of shallow sounds and the lakes of Onega and Ladoga. This means that the Eemian and Boreal transgressions must have been synchronous, as demonstrated by dating techniques including the ESR-method (Molodkov & Raukas 1988).

According to the second reconstruction, the contours of the Eemian Sea closely coincided with the Litorina Sea limit and therefore this basin of water could have been linked with the ocean only via the Skagerrak, Kattegat and Danish Sounds. Transgressive waters would have inundated the Lake Ladoga depression and small areas of the Vistula River valley (Blagovolin et al. 1982). To our mind, the first version seems more reliable.

In terms of pollen zones, the marine interglacial sediments on Prangli Island are in good agreement with the other well-known Eemian interglacial deposits in the St. Petersburg (Mga section, a.o.) and other regions. In several other parts of the Gulf of Finland, the seismoacoustic evidence indicates the three-fold division of the Pleistocene deposits (Kiipli et al. 1993), analogical to that discovered on Prangli Island. This is also reflected in the occurrences of tills of two glaciations with marine Eemian deposits in many supra-aquatic (islands of Eksi, Rammusaar, Malusi) and sub-aquatic (banks of Kuradimuna, Bezymyannaja, Moksei a.o.) drumlins, but also in the area of the Island of Mosčnyi. The layers under consideration are of almost horizontal bedding; the top of the intermorainic layers is located at a height of 60-70 m and the base at a depth of 75-90 m below sea level.

Very complicated and still unsolved is the post-Eemian geological history of the territory under consideration. Contradictory opinions have been expressed on the Early Weichselian stage. Finnish researchers maintain that in the Early Weichselian their territory was free from ice; in the Baltic countries the vast majority of investigators are of the opinion that at that time there was a thick ice sheet, almost as large as during the Late Weichselian.

The oxygen-isotope curves from deep-sea sediments show that the volumes of water, stored in glaciers, were almost equal in the first and second halves of the Late Pleistocene. This does not mean that the development of the glaciers was synchronous in different areas. On the basis of geomorphological indications it is rather difficult to reconstruct palaeoglaciological parametres, and further, the potentials of geochronological methods are too limited to allow unambiguous conclusions.

In the former Soviet Union, several investigators (I. Krasnov, E. Zarrina, A. Raukas, L. Serebryanny) maintained that the glaciation culminated on the East-European Platform in the Early Weichselian when the Brandenburg (Bologoe) marginal moraine was formed, others (M. Makarycheva, N. Chebotareva, A. Velichko a.o.), on the other hand, either deny an Early-Weichselian advance of the ice sheet or consider it to have been of a rather limited distribution. According to Nilsson (1973), the glaciers did not reach Scania in southern Sweden at that time. In Finland, a Lower-Weichselian till is represented by a dark-grey clayey variety (Rainio & Lahermo 1984), correlated with the Suintio Till (Bouchard et al. 1990). The so-called purplish-grey till in southern Estonia is referred to the Lower Weichselian and so is the lower grey till overlying the Eemian in the fore-klint area and in the North-Estonian Plateau, and also the related waterlain glacial deposits (Raukas 1978). Lower-Weichselian tills are also found in Latvia and Lithuania. In the new stratigraphical scheme of Estonia (Table 131) the purplish-grey till belongs to the Lower Weichselian Valgjärve Subformation (Raukas & Kajak 1995). Near Valgjärv, this bed overlies Eemian sediments (Kajak 1995).

Some authors maintain that instead of a single increase in the volume of the ice sheet and its melting in the study area there were numerous rapid fluctuations of the ice margin in the Early Weichselian. Based on pollen diagrams, these authors have differentiated in northwestern Russia three distinctly differing cycles in the development of the vegetation in the Early Weichselian (Malahovsky & Spiridonova 1981), reflecting independent warm oscillations. The first, Upper-Volga cycle was characterized by the prevalence of a boreal vegetation. In the succeeding, Tosna cycle alongside with the above-mentioned complex, nemoreal species were spread, whereas in the third, Berezai cycle a hypoarctic vegetation dominated. These interstadials are separated from each other by Kurgolovo, Kileshi, Bologoe and Jedrovo stages, whose pollen spectra show the prevalence of nonarboreal plants. At first, spores dominated and then the proportion of herb pollen increased. From a geological and geochronological point of view it may, however, be concluded that these interstadials are not yet sufficiently grounded, and that their pollen spectra are not clear enough.

The radiocarbon ages are beyond the limits of the method, and the spore and pollen data are difficult to use in the correlations.

Detailed studies have been performed on classical Early-Weichselian interstadials in the Netherlands and Denmark. In Denmark, the Brørup interstadial deposits, characterized by pollen spectra of birch and pine forests with spruce (Andersen 1961) were studied beyond the boundaries of the Weichselian glaciation, however, not far from its margin. The preceding Rodebæk interstadial deposits, which are embedded between solifluction layers in the same region, are characterized only by pollen of dwarf birch, juniper and herbs. Further south, in the Netherlands, spruce and pine forests with birch and hazel were present during the Brørup interstadial, whereas during the Amersfoort interstadial, correlated with the Danish Rodebæk, there were only birch and pine forests. Quite often the Rodebæk (Amersfoort) interstadial is connected with the Brørup. In western Europe, after the Brørup interstadial, the Odderade interstadial has been distinguished with its stratotype being located in Schleswig-Holstein, Germany. The Brørup and Odderade interstadials in northwestern Europe were warmer and of longer duration than other Weichselian interstadials with boreal forests. They are correlated with the isotope substages 5a and 5c in the deep-sea chronology.

It is worth mentioning that in the Netherlands in Western Europe, the Middle Weichselian interstadials (Moershoofd, Hengelo and Denekamp) were established long ago. Their age succession has been determined on the basis of radiocarbon datings of an order of 30,000 - 50,000 years. According to spore and pollen data they were characterized by a cold-favouring vegetation.

In the City of St. Petersburg, in a section on Grazhdanski Prospect, a still warmer Middle Valdaian interval has been established (Malahovsky et al. 1969). According to E. Spiridonova’s pollen analysis, this section displays three climatic warmings and three coolings. The warmings are characterised by the culmination of pine, spruce and tree-like birch. However, at the same time, a great quantity of herb (20%) and dwarf birch pollen shows the absence of closed forests. Interpretation is also complicated by the redeposited interglacial pollen of hazel (up to 8%), alder (up to 15 %), broad-leaved and other species of trees, which undoubtedly contribute to the rise in the arboreal tree pollen percentage. During climatic coolings the share of dwarf birch (50-60 %) and herbs (up to 50 %) increased remarkably. Evidently, tundra-like periglacial vegetation was widespread at that time. The section has yielded an radiocarbon date of 40,380 ± 800 (LU-22) years, corresponding to the end of the first warming. Many sections in the East-European Platform correlate with the key section on Grazhdanski Prospect in terms of radiocarbon and palynological data. These Middle Valdaian (Weichselian) sections are not complete; they are mostly located in the vicinity of the marginal zone of the Valdaian glaciation.

In such a complicated situation, it would be rather difficult to correlate these sections with those in western Finland (Vimpeli, Oulainen), which some authors refer to the Eemian interglacial, and others to a Weichselian interval. To our mind, a reliable conclusion has been drawn by Donner (1983, 1984) who considers them as representing an undisturbed Early-Middle Würmian interstadial. They are correlated with the Jutland interstadial in Sweden, the Peräpohjola interstadial in northern Finland, and the Brørup interstadial in Denmark. Beside spore and pollen data, the conclusion is also supported by the fact that the in situ section of Oulainen, with its continental organogenous deposits, is undoubtedly located below the maximum level of the Litorina Sea which means that it lies also below the Eemian Sea level. The Vimpeli section is located at a height of 117-120 m above sea level, i.e. 10 m higher than the Eemian marine section at Norinkylä. Therefore, it cannot be Eemian either, but should be attributed to the Oulainen interstadial.

The radiocarbon ages do not contradict this conclusion. The sample of wood from the Vimpeli section yielded the ages ≥43,000 (Hel-1404) and ≥50,300 (Su-925), peat ≥43,000 (Hel-1405) and the fraction of humus from the same sample ≥40,300 (Hel-1406). The sample of gyttja from Oulainen section gave the ages ≥48,000 and 63,000 (+5500, -3200) years (GrN - 7982). Eight TL sand samples above the interstadial deposits in the Oulainen section yielded an average age of 94,000 ± 4000 years and three samples from the lower part - 121,000 ± 5000 years (Donner 1984).

From a stratigraphical point of view, the location of the Vääna-Jõesuu section (Fig. 91) in the Vääna klint bay, 20 km west of Tallinn, has remained unclear. In that section, Weichselian deposits with a total thickness of 71.1 m were studied (Raukas & Liivrand 1971).


The description of the section:


    Layers                                                  Thickness in metres


Yellowish-grey sand, of varying grain size      0.00 - 11.50

Sandy gravel with pebbles                                 11.50 - 13.45

Glaciolacustrial silty clay with

  indistinct varved structure                                               13.45 - 13.80

Grey compact till, rich in crystalline rocks      13.80 - 53.10

Dark-grey silty clay, with organic matter               53.10 - 56.00

Grey compact till, rich in crystalline rocks      56.00 - 71.10

Below follows Cambrian clay.


Taking into consideration its location in the fore-klint area, only 35 km southwest of the Island of Prangli, and also the similar geological structure and petrographic composition of tills, the Vääna-Jõesuu section is well correlated with the section of Prangli Island. The layer with organic matter lies here higher than on Prangli Island which might indicate its younger age. This is confirmed by pollen evidence.

In the Vääna-Jõesuu section, the quantity of pollen of broad-leaved trees (oak, hornbeam, linden and elm) amounts to 80% in the sum of trees. The amount of hazel and alder reaches 300 %. At the first sight, this concentration is entirely comparable with the quantity of thermophilous tree pollen in interglacial deposits. However, a distinct successional development of vegetation is not observed here. Pollen of the different thermophilous species occurs throughout the intermorainic interval, but the absolute content varies, being less in the lower part of the interval. As an exception serves linden, the concentration of which is lower than that of the other broad-leaved species, and in the lower half of the interval, it is represented only by occasional grains. To solve this problem, E. Liivrand used variograms for tills and intermorainic layers. In the tills redeposited pollen of thermophilous species was discovered in different amounts, but with similar composition. In the tills the pollen is practically entirely redeposited, and in the present case, prevailingly from the underlying Eemian layers. It was demonstrated that the pollen composition of interglacial thermophilous species practically did not change throughout the 70-metre-thick Weichselian section. Interglacial deposits were subjected to the most intensive abrasion during the accumulation of intermorainic clays, in which the preservation of pollen is much worse than in tills. Besides, in the same sections one can observe the inversion of spectra which is indicative of redeposition. During the accumulation of the lower till layer, the Upper Eemian layers with abundant pollen of hornbeam were eroded. Afterwards the lower layers rich in oak pollen were also subject to erosion.

The occurrence of an extremely high amount of redeposited interglacial pollen in intermorainic clays at Vääna-Jõesuu hampered the study of in situ pollen of the periglacial vegetation which could have persisted here under the ice-free conditions. According to E. Liivrand, this is shown also by the maximum of herbaceous plant pollen (up to 18 %), accompanied by an increase in the pine and birch pollen, spores of Sphagnum and a small content of redeposited pollen of thermophilous species (only 5-12 %).

Diatoms (see Raukas & Liivrand 1971) are most likely also redeposited from the Eemian deposits, although thermophilous species, typical of this sea, have not been recorded.

G. Nedesheva found in the tills and intermorainic clays foraminifera, sometimes in great amounts. Taking into account their ecological needs, it was concluded that the intermorainic clay accumulated in a shallow sea basin where the water temperature did not exceed +3oC and the salinity was below normal (less than 30‰). But, since similar species have been found in great amounts also in the underlying till, at a depth of 58 m, the foraminifera may well have been redeposited (Raukas & Liivrand 1971).

The intermorainic interstadial and stadial clays of the Vääna-Jõesuu section may be referred to the Middle Weichselian, although direct evidence in favour of this conclusion have not yet been found. On the basis of their stratigraphical position and spore and pollen characteristics, it is difficult to connect them with the sections of Oulainen and Vimpeli in Finland, or with intermorainic sections in southern Estonia and in the Leningrad District.

The problems concerning the deglaciation of the last ice sheet and the occurrence of Gotiglacial interstadials and stadials will be discussed in the next chapter.




Deglaciation history

R. Karukäpp & A. Raukas

General data about the glacial dynamics

Estonia and the Gulf of Finland south of the Salpausselkä ice-marginal formations was freed from the continental ice in Gotiglacial time. The final phase of the Gotiglacial was characterized by a marked differentiation of the radial flow in the thin marginal parts of the ice sheet, resulting in the formation of tension zones, and areal and linear concentrations of drift. Selective erosion and accumulation under the conditions of convergent movement of the ice lobes towards the subglacial uplands led to the formation of the Haanja and Otepää glacial accumulative insular heights (Raukas & Karukäpp 1979).

A dynamic system of fractures developed during the final stage of the deglaciation; along these fractures glaciofluvial deposits were subsequently accumulated. Eskers are practically absent in the areas of the Daniglacial phase of deglaciation, and they are not typical of the early part of the Gotiglacial either. Eskers often occur within the zone of glacial erosion where the average thickness of glacial drift is small, for example, on the Pandivere Upland.

As the thickness of the ice diminished, the movement of the marginal lobes was highly controlled by the underlying bedrock topography (Tavast & Raukas 1982). The lobes reduced in size because the supply from the central parts of the ice sheet decreased. As a result, the length of the margin of the retreating ice sheet diminished towards the end of the Gotiglacial.


Ice-marginal formations

Marginal positions of the ice sheet in the present topography are marked by interrupted chains of end moraines and glaciofluvial formations (Fig. 164D). Most frequently they seem to represent events of temporary stagnations of the ice margin when the glacial regime was close to equilibrium. More often than in the southern part of the Baltic area, extensive proglacial bodies of water were present in front of the receding ice margin. This is reflected in the relatively common occurrence of glaciofluvial deltas (Photo 47) and marginal eskers, in the absence of typical sandurs, and in the wide distribution of glaciolacustrine sediments. The primary relief was often reshaped by wave action and shore processes of the proglacial lakes.

The concentrated unequal subglacial and marginal accumulation was typical of the early stages of the Gotiglacial in southern and southeastern Estonia (Raukas & Karukäpp 1979), where the Haanja and Otepää glacial accumulative insular heights were formed (Fig. 164A, B). On the basis of the composition of tills belonging to different glaciations or glacial stages, the orientation of elongated clasts in till or linear elements of topography, the direction of the ice flow, and the stages of development of insular heights have been established (Hang & Karukäpp 1979, Raukas & Karukäpp 1979).

During the subglacial stage, the several glaciodynamic structures, including ice-folds, fractures and shear plains overloaded by till and erratics from subglacial surface, were formed. The convergent direction of ice flows caused intensive accumulation around the subglacial elevations as initial centres. Both, the movement direction and rapid accumulation led to stagnation of great portions of thick glacier ice in the central areas of insular heights, which initiated the second, englacial stage of development of the insular heights. This stage marked the beginning of the glacial stagnation and retreat of the active ice in southern and southeastern Estonia.

High relief of specific large landforms of complicated structure in the centre of the heights (Photo 46) is the result of glaciotectonics of extremely great horizontal and vertical stresses in the beginning of the englacial stage of morphogenesis, causing horizontal displacement and considerable upthrusting of sediments. Towards the end of this stage, the area of stagnant ice gradually rose; active accumulation and relief formation took place on the widening transition belt between active and stagnant ice (Photo 48). The glacial deposits still contained great blocks of buried dead ice (Fig. 164B).

The above-mentioned stage was followed by peripheral marginal accumulation (Aboltinš 1972) and the final, dead ice stage when intensive glaciokarst, slope and erosional processes took place (Karukäpp 1985). Various crevasse fillings, in some places covered with flow till and other solifluctional deposits, were formed.

Linear ice-marginal formations are typical of the Pandivere and Palivere zones.

In the Pandivere zone, they are represented by push end moraines (Photo 45), glaciofluvial deltas (Photo 47) and marginal eskers. The height of the formations ranges from some metres up to 40-50 metres (Sinimäed at Vaivara, Photo 34) depending on the intensity of accumulation and ablation, and on the abundance of drift in the glacier. The North-Pärnumaa marginal esker chain has the greatest, more or less continuous length of 150 km (Fig.164C). It was eroded and partly levelled by younger bodies of water. Marginal features of the Pandivere Stadial also occur on the northwestern slope of the Pandivere Upland. East of the Pandivere Upland the ice margin followed the line Vaivara - Laagna - Narva.

The Palivere line (Fig. 164D) is clearly marked by marginal eskers in northwestern Estonia and by extensive glaciofluvial deltas near and east of Tallinn (Raukas 1992b). The Upland of Western Saaremaa and Sõrve end moraines are the westernmost continuation of the Palivere line. The eastern part of the line is submerged by the present Gulf of Finland.

A number of different views has been expressed on the course of deglaciation and correlation of the marginal formations in the area. There is no disagreement as to the ice margin positions in western Estonia where continuous chains of marginal formations of various types are present. On the other hand, the reconstructions suggested for the St. Petersburg district and for the areas covered by the waters of the present Gulf of Finland differ significantly (Zarrina & Krasnov 1965, Serebryanny & Raukas 1966). Analysis of the Gulf of Finland’s bottom topography, based on large-scale depth charts, reveals a series of subparallel marginal features (Karukäpp & Vassiljev 1992) which do not seem to fit in with the conventional schemes. Usually, these features show smaller dimensions than the radial forms of glacial topography, represented by megadrumlins and eskers. However, the available information does not yet allow any reliable reconstruction of the submarine continuation of the Palivere line.


Ice-dammed lakes

Research into the deglaciation history is in many respects based on the analysis of the raised beaches of proglacial lakes. The coastal formations of local ice lakes are usually small, up to 2 m high (Raukas 1992a), and unclear in outline owing to the plain topography, the occurrence of erosion-resistant rocks and deposits in the distribution areas of the lobes, and to the small dimensions, shallowness and brief existence of the lakes, on account of which the action of the waves was inconsiderable (Raukas 1986).

Coastal formations of the Baltic Ice Lake, which formed after the retreat of the continental ice to the north and northwest from the Pandivere Upland, are clearer in outline (Ramsay 1929, Kvasov & Raukas 1970). The littoral formations of the different phases of the Baltic Ice Lake are represented by glaciofluvial deltas.

The delta surfaces in northern Estonia were mostly formed at altitudes of 38–50 m (Männiku) or at about 20 m (Potsepa) and above 70 m (Kemba, Voose). Their vertical distribution shows a more or less clear concentration at certain levels of altitude (Pärna 1960). In most cases, the delta surfaces are even (Photo 47) and gently inclined towards the distal margin, or slightly convex. Proximal slopes of the highest deltas mark the contact with the glacier during the existence of local ice lakes (Kemba, Mustamäe ).

Deglaciation of the Pandivere Upland and the area north of the upland is related to two types of glaciofluvial deltas (Karukäpp & Tavast 1985) where the upper (70–80 m) level (Kemba, Valgejõe) has contact slopes with the glaciers and inclination of bedding to the distal southern slope of the delta. The lower (45­–48 m) level formed as the result of the streams flowing from the Pandivere Upland to the north and transporting glaciofluvial deposits back in the opposite direction — into the lowering ice lakes. The latter are of no significance for palaeoglaciological reconstructions, but indicate the local altitude of the water level before the Palivere Stadial.


Lithostratigraphical evidence

The various tills of different ages from the last glaciation indicate the stadial-oscillatory character of the deglaciation. The distribution of tills and index boulders in them implies that there were at least four stadials during the last glaciation in Estonia. Each of the stadials is characterized by a certain direction of continental ice flow, differing from that of the other stadials and, accordingly, by a certain combination of index boulders (Raukas 1963a). The Haanja, Otepää and Sakala stadials are characterized by the prevailing southeasterly direction of the ice movement, the Pandivere Stadial by the southerly or even southwesterly direction (west of the Pandivere Upland) and the Palivere Stadial again by the southeasterly direction (Fig. 164) with the corresponding composition of index boulders, mainly rapakivi from southwestern Finland, granites and rapakivi from the Åland Archipelago, Baltic red quartz-porphyries and olivine diabases from Satakunta, all of which are practically absent south of the ice-marginal formations of the Palivere Stadial (Raukas 1992b).

The presence of till-covered layers of interstadial character, the distribution of index boulders and a number of other lithostratigraphical observations provide evidence for a significant event of ice front oscillations. For instance, at Männiku in the southwestern part of Tallinn, glaciofluvial varved clays of the Pandivere Stadial are covered by glaciolacustrine varved clays with a thickness of up to 20 m (Raukas & Rähni 1966, Karukäpp & Miidel 1972), indicating a glacier readvance.


Biostratigraphical evidence

Interstadial or interphasial layers between the different till beds occur at many sites both in northern and southern Estonia. Palynological data from these layers are available, for example, on the Island of Prangli (Raukas & Rähni 1966, Raukas 1978). Unfortunately, they contain a lot of redeposited pollen that hampers the correlation of sections and the establishment of palaeogeographical conditions during the interstadial events.

Of greater consequence in dating of deglaciation of the area are the pollen data from till-covered Pleistocene deposits which suggest severe climatic conditions throughout the Late-glacial. Local vegetation just set in to develop and the concentration of redeposited pollen in sediments is high. That is why the Bølling sediments in Estonia do not reveal any clear palynological characteristics (Pirrus & Raukas 1996).

For dating of different steps of the deglaciation, it would be essential to know whether till-covered Bølling deposits occur in Estonia or not. Such deposits may be present in southern Estonia, but hardly in the north. In the section of Haljala (Männil & Pirrus 1963), a pollen assemblage suggesting a brief interval of warming, possibly Bølling, has been reported from a sandy interlayer at a depth of 10.5–11.2 m. However, in view of the circumstance that the pollen might have redeposited into the sandy sediment, the interpretation of the pollen data remains uncertain, and the more that there has not been established any other corresponding warm interval elsewhere in Estonia.

Deposits of Older Dryas age occur both in northern and southern Estonia. They are represented by varved clays, silts and sands with horizontal varve-like bedding. Compared with Allerød deposits, their pollen floras are poorly preserved, suggesting redeposition. Upwards the amount of redeposited pollen usually diminishes. The lower boundary of the Older Dryas is undefined in Estonia (Kajak et al. 1976).

Allerød deposits (about 10,800–11,800 yr BP) are present even in the northernmost parts of Estonia (Pudisoo, Haljala, etc.). This suggests that Estonia was almost totally ice free by the beginning of Allerød, except the northwestern part (Pirrus & Raukas 1969). Allerød deposits differ by the large amount of organic remains they contain and are therefore considered a stratigraphic reference horizon throughout the Baltic States (Kabailiene & Raukas 1987). In the Allerød, the whole Baltic area experienced considerable warming which highly contributed to the rapid ice retreat.

The Upper Dryas deposits (10,000–10,800 yr BP) in Estonia are characterized by a high content of herbs (40 - 60% of the total amount of spore and pollen) and dwarf birch. The Younger Dryas cooling gave rise to the development of tundra vegetation once again and undoubtedly promoted activization of the ice cover close to Estonia.

The accumulation of organic sediments in the lakes of southeastern Estonia started only at the end of the Younger Dryas about 10,300–10,200 yr BP, substituting the accumulation of sand, silt and clay which had been clearly prevailing during the Late-glacial. The oldest organic sediments have been dated at Saviku ca. 10 200±90: TA-328 yr BP (Sarv & Ilves 1975).

In the light of the pollen evidence, the retreat of the ice margin from the Haanja (Luga) position started during Bølling time and the deglaciation of the Estonian territory was completed during the second half of the Allerød (Pirrus & Raukas 1969). Hence, the deglaciation of the Gulf of Finland took place towards the end of the Allerød and in the beginning of the Younger Dryas.



In Estonia, deglaciation has been dated with conventional varve chronology, noncorrected radiocarbon chronology and TL methods. Up to now, all the above-mentioned methods are prone to big errors and uncertainities. None of them can, therefore, be assigned universal validity. The majority of 14C dates of intermorainic or submorainic sequences are younger than one would expect on the basis of the conventional methods. A good example is the submorainic sections at Petruse (12,670±200; 12,080±120) and Viitka (10,950±80) in the hilly area of southeastern Estonia which was freed from ice at least 13,000 years ago. This may result from redeposition, caused by glaciokarst processes (Raukas & Karukäpp 1994). At the same time, depending on the “hard water” effect, some organic layers above the uppermost till have given 14C dates of 13,000 - 14,000 yr BP (Pirrus & Raukas 1969). In dating of geological objects with the TL method, the most complicated problem is the establishment of the zero point for the time of the formation of the sediments studied. All the dates obtained are substantially older than would be expected on the basis of traditional deglaciation chronology.

Somewhat better results have been obtained using conventional varve chronology. To establish a local chronology, Rähni (1963a) studied varved clays in northeastern Estonia. The clay series investigated by Markov (1931a) from the Luga and Neva basins provided a good basis for further correlations. A connection between the varve diagrams from northeastern Estonia and from the Luga Basin in the Leningrad District was established by using the so-called drainage varves. These marker horizons served as the basis for the dating of the ice retreat from the Pandivere Upland (Raukas et al. 1971).

The attempts to create a local chronology for western Estonia on the basis of varved clays have not yet proved successful, although according to Pirrus (1968) and other authors, clays suitable for varve chronological studies are present there.

The attempts to determine the age of the Salpausselkä ridges by means of varved clays have been hampered by the difficulties of correlating the floating Finnish chronology with the Swedish time scale. A connection between the Swedish and Finnish chronologies was proposed by Strömberg (1990) on the basis of step-by-step correlations of varves from east-central Sweden via Åland to southwestern Finland. According to these correlations, and including the recent revisions of the Swedish time scale (Cato 1987), Sauramo’s zero year should be dated at 8693 BC (ca 10,650 yr BP).

Markov and Krasnov produced a large number of varve graphs from sites in the Leningrad District       (Markov & Krasnov 1930) and in southern Karelia (Markov 1931b). According to these data, there was a constant retreat of the ice margin in the northern Onega area with an average rate of 160 metres a year. Unfortunately, the northwesternmost proximal equicess by Markov in Karelia lies about 100 km distalwards of the southernmost equicess by Sauramo (1923, 1929). By calculations of the rate of retreat, Markov thus arrived at the conclusion that the ice retreat from Petroskoi (Petrozavodsk) took place in the year -3300 in Sauramo’s system. According to the conventional dating of Sauramo’s zero year, this would correspond to approximately 13,900 yr BP.

Extending Sauramo’s chronology to the Leningrad area, Markov (1931a) estimated that the retreat of the ice from St. Petersburg to Viipuri would thus have taken about 250 - 350 years.

Markov (1931a) was not able to correlate varved clays from the Luga Basin near Kingissepp (Jamburg) with those from the Neva Basin. The clay sequences in this area are characterized by the occurrence of two drainage varves, one at the level of 79–80 years and the other at the level of 111 years from the base of the varve sequence. They are explained as reflecting two events of ice-lake drainage from the Neva to the Luga Basin, the connection between the basins being regulated by the ice margin leaning against the klint edge near the Koporje Village.

According to Rähni (1963a), the Pandivere ice-marginal position is correlated in time with the drainage of the Neva Ice Lake at Koporje. The first drainage event can be dated at 12,080 and the second at 12,050 yr BP (Raukas et al. 1969). When the two large ice-marginal water bodies joined up, synchronous drainage varves were formed in the clay sequences of northeastern Estonia and the Luga Basin (Rähni 1963a). The temporary closing of the connection between the two bodies of water was caused by an oscillation of the ice margin 12,080–12,050 yr BP during the Pandivere event. In 1984, Karukäpp reinvestigated a section of the Luga Basin to check the varve correlation with northeastern Estonia (Rähni 1963a) by palaeomagnetic methods (Bahmutov et al. 1987). As the result, it was established that the drainage varves in the Luga Basin and northeastern Estonia differ in the composition and, therefore, the direct correlation on the basis of the drainage level is not possible (Bahmutov et al. 1987). The palaeomagnetic data from the studied clay sections did not provide an indisputable basis for a revision of the earlier established age of the Pandivere Stadial (Raukas et al. 1969).


Concluding remarks

The territory of Estonia was freed from the continental ice in Gotiglacial time. Against the background of a gradual climatic warming there probably occurred remarkable cooling periods, which caused halts or even new advances (Palivere Stadial) of the degrading ice cover, marked in nature by distinct ice-marginal formations. According to traditional approach, Estonia was finally cleared of the continental ice about 11,000 years ago, but before the glaciers temporarily reinvaded the West-Estonian Archipelago and northwestern Estonia. This Palivere Belt is well marked with end moraines, marginal eskers and glaciofluvial deltas and can be traced from the distribution of index boulders (Raukas 1992b).

It should be pointed out that the revised Swedish varve chronology (Cato 1985, Lundqvist 1986) calls for some readjustment of the dates obtained earlier. It may be suggested that the recession from the Estonian ice marginal zones could have started some half thousand years earlier than hitherto assumed (Donner & Raukas 1989, Karukäpp et al. 1992). However, the data available today do not allow any definite revision of the old deglaciation chronology (Raukas 1986).



Evolution of the Baltic Sea

A. Raukas


Estonia has been a maritime nation from time immemorial. The oldest Estonian towns belonged to the venerable Hanseatic League. A great number of Estonians earned their living out at sea and many of them never returned. In consideration of the vital role the sea has played in the life of the Estonians, the investigation and management of coastal areas is a prime priority in Estonia.

Already the ancient inhabitants of the coastal area noticed that in a course of time submarine boulders and small islets dangerous for seafarers crop out from the water, former islands join the land and fishing villages withdraw from the sea-shore. In the 13th century, a big portion of the present old Tallinn was under the sea, with water reaching as far as the town wall, now half a kilometre away from the sea-shore.

The first investigators erroneously associated these phenomena with a gradual fall of the water level in the sea basin (Kozakevich 1848). In the second half of the century, the tectonical uplift of the area was a generally accepted truth, however, up to now opinions differ as to the grounds of recent Earth’s crust movements.

In the history of the Baltic the following five stages have been distinguished in Estonia:

1) The Baltic Ice Lake in the Allerød and Younger Dryas (Fig. 165 A);

2) The practically fresh-water Yoldia “Sea” at the end of the Younger Dryas and at the beginning of the Pre-Boreal (Fig. 165 B);

3) The Ancylus Lake in the second half of the Pre-Boreal and Boreal (Fig. 165 C);

4) The Litorina Sea in the Atlantic and in the beginning of the Sub-Boreal (Fig.165D);

5) The Limnea Sea in the Sub-Boreal and Sub-Atlantic covering the last 4000 years.

Taking into account similar ecological conditions and general development, the transitional Echeneis and Mastogloia phases are referred to the Ancylus and Litorina stages, respectively. The last episode, called the Mya Sea, is characterized by conditions similar to those at present. Therefore the recognition of an independent Mya Stage in the history of the Baltic in Estonia is not justified.

The evolution of the Baltic Sea is characterised by alternation of transgressions and regressions. For a long time it was believed that transgressions were more or less synchronous all over the Baltic. However, recent studies have shown that due to the different tectonical situation, they culminated at different times in different parts of the Baltic (Kessel & Raukas 1984). Similarily, the velocities and tendencies of the shoreline displacements varied. If in the central and northern Baltic the ancient beach deposits and coastal formations are located above the contemporary sea level, then in the southern Baltic the synchronous formations are up to 40-50 metres below the water level (Fig. 166).

The diatom and mollusc evidence shows that the rise in the salinity and water temperature was not simultaneous either, which complicates stratigraphical correlations of near- and offshore sediments in different parts of the Baltic.

For instance, in Blekinge (southern Sweden) saline conditions established some 8500 years ago, after the waters of the Baltic were reunited with the ocean. Diatoms show that in western Estonia saline influence began about 7800-8000 years ago. A clearly brackish mollusc fauna enclosed in Litorina deposits, appeared there about 7000 years ago (Kessel 1965).

In general, the epicontinental character of the Baltic Sea is reflected in all processes responsible for the coastal morphology and near- and offshore sedimentation. Tides were insignificant. Much more important were seasonal variations and short-term changes in water level induced by strong winds. Longshore drift was predominant in Estonia’s northeastern and southwestern coastal waters. Elsewhere it was obstructed by the rugged shoreline and shallowness of water. In those areas material of local origin accumulated. On the basis of the composition and morphology of shingle and gravel grains, the mechanism of the formation of beach deposits can be easily elucidated and the initial rocks defined (Kessel & Raukas 1967). Contrasts in sediment calibre may be correlated with contrasts in wave energy related to the past and present near-shore gradients.

Owing to the moderate crustal uplift of the area, the shoreline remained stable through transgressions. The related coastal formations observable in a vast area are up to 6-7 m thick  (Photo 49) and morphologically clear (Raukas 1966). This means that the transgressive shorelines of the Baltic Ice Lake (BIII), Ancylus Lake (AI) and Litorina Sea (LI + LII) can serve as key horizons for palaeogeographical reconstructions.

The treatment of the Baltic stages varies with areas and authors and, therefore, they are not always comparable with each other. The units distinguished on the basis of littoral diatoms are variable because changes in salinity were expressed differently in different areas. Any precise stratigraphical subdivision on the basis of the faunal evidence is difficult (Photo 50) because this evidence mainly comes from shallow-water deposits which seldom form continuous sequences. In actual fact, the subdivisions in different countries are based on mixed criteria and the results are not inspiring. A combination of morpho-, litho-, bio- and chronostratigraphy has lead to confusion (Hyvärinen & Raukas 1992). The main stages in the Baltic Sea history are generally known from the beginning of the century, but they have never been properly defined as stratigraphic units.

Ignatius et al. (1981) subdivided the Baltic off-shore sediments into three main lithological units, repeatedly observed in cores: (1) glacial varved clays; (2) massive sulphide-rich transitional clays and (3) massive or finely laminated postglacial muds with a relatively high organic content. The transitional clay unit can be divided into two subunits: the lower part is characterized by a strong sulphide colouring, while the uppermost part is formed of grey clay with occasional dark streaks. In all likelihood, the deposition of post-glacial muds started at the beginning of the Litorina Stage when brackish-water conditions were established. In a broad sense, the transitional clay unit correlates with the Ancylus Stage, while the uppermost transition clay (sometimes referred to as “upper Ancylus”) correlates with the Mastogloia Stage, separating the Ancylus and Litorina stages (Hyvärinen et al. 1992b). The boundary between the transition clays and postglacial muds is well-defined and consistent with the stratigraphical horizon. According to Ignatius et al. (1981), it marks the boundary between the Ancylus and Litorina stages. The fact that some of the lithological units are diachronous across the Baltic Sea, while others are clearly synchronous, can cause misunderstanding when used in conjunction with climatic chronozones (Winterhalter 1992).

On the Estonian shelf, Lutt (1992) differentiated seven more or less distinct lithological units which could be included to four main sedimentation stages of late- and postglacial time (Kiipli et al. 1993): (1) basal part: formation of glacial and glaciofluvial deposits; (2) stage of Late-glacial vast ice-dammed lakes; (3) stage of post-glacial great lakes; (4) stage of marine sedimentation: formation of mostly stratified pelitic and silty sediments, containing brackish-water fauna and flora. The off- and near-shore deposits are characterized by numerous unconformities and rapid facies changes (Lutt 1992) and often gaps in the sections are prevailing. Therefore, in terms of palaeogeographical conclusions, the classical supraaquatic sections with buried organic sediments and lagoonal deposits often impart much more information.

The Baltic Sea was formed some 12,000 years ago after the readvance of the ice margin from the northern slope of the Pandivere Upland in northern Estonia (Fig. 167), as a result of which the isolated big ice-dammed basins west and east of the elevation joined up (Kvasov & Raukas 1970).

The Baltic Ice Lake, the first stage in the history of the Baltic, had coastal formations at five different levels of which the Palivere and BIII levels (corresponds to the formation of Salpausselkä I ridges in southern Finland) were of transgressive character (Fig. 166). It was assumed for a long time on the basis of varve chronology and radiocarbon dating (Raukas et al. 1969) that the Palivere Stadial (Fig. 168) occurred about 11,200 years ago. But the results obtained by dating the Salpausselkä ridges (Donner & Raukas 1989) by means of revised Swedish varve chronology (Cato 1985, Lundqvist 1986), in which 430 years is added to the age of the drainage varve in the Swedish time scale, suggest that the Palivere zone was formed about 400-500 years earlier than hitherto supposed (Raukas 1992b).

The retreat of the continental ice cover across the lowland at Mt. Billingen in Central Sweden established a connection between the Baltic Ice Lake and the ocean in the west, causing a rapid lake level drop (Fig. 166). In Estonia, near Pärnu, it was estimated at about 25-30 m (Talviste 1988). This event occurred according to the revised Swedish varve chronology 10,690 varve years BP (Cato 1987), and is considered as the beginning of the Yoldia Sea Stage (Fig. 165B). Svensson (1989) has proposed 10,300 BP as the time for the final drainage of the Baltic Ice Lake. In connection with the abrupt drop of the water level, there should be clear changes in the erosion-accumulation areas and in the grain-size of the bottom deposits but, unfortunately, these phenomena have not yet been precisely studied. Only in a few seismic profiles and boreholes, the boundary between late-glacial and post-glacial clays is distinct.

After the drainage of the Baltic Ice Lake, slightly saline conditions could have existed in the so-called Preboreal Yoldia Sea near the Swedish coast, but not in the open off-shore waters and in the eastern Baltic (Raukas 1991b). In terms of shore displacement, the Yoldia Sea in the eastern Baltic was regressive due to the rapid crustal uplift (Fig. 166). Unfortunately, the lowest point reached by the Yoldia regression cannot be precisely determined. The clearest coastal forms of the Yoldia Sea (YII) in northern Estonia are probably located several metres higher than the maximum level of the succeeding Ancylus Lake in the same area (Raukas 1994).

The hydrology and salinity of the Yoldia Sea have been subject to much discussion. A detailed account of the ecological conditions in the area of the Baltic-Atlantic straits in western and central Sweden was provided by Freden (1988). Recent diatom studies in the Kalmarsund area (Håkansson in Svensson 1989) suggest that there was only a brief (100-150 years) influx of salt water during the middle part of the Yoldia Stage, which did not reach the open Baltic or the Gulf of Finland (Hyvärinen et al. 1992a). According to Gudelis (1979), there were only fresh-water basins of various sizes near the Lithuanian coast in the Pre-Boreal.

Due to the limited connection with the ocean, and abundant meltwater supplied from the nearby ice sheet (Fig. 165B), the salinity of the Yoldia Sea was low, and the brackish-water malacofauna, including rare dwarf forms of Portlandia (Yoldia) arctica, were preserved in a rather small western part of the Baltic. The single specimens of P. arctica, found in tills and marine deposits in southern Baltic (Kliewe & Janke 1978) and in Latvia, have probably been redeposited from the Eemian sediments. These are known to occur in the coastal zone of Latvia in tills younger than the Eemian marine deposits (Dreimanis 1970).

The Yoldian offshore sediments in the area under consideration are typically barren of diatoms, or may contain a very sparse flora consisting mainly of Melosira islandica, which is the dominant species in the richer floras of the Ancylus Stage upwards (Hyvärinen et al. 1992a).

The presence of Yoldian “brackish-water” diatomaceous taxa in some coastal regions and littoral sediments of the Baltic States and Leningrad District together with fresh-water forms is probably due to redeposition processes. This means that the Yoldian offshore diatom assemblages are mainly fresh (Raukas 1994).

The crustal uplift, being more rapid than the eustatic sea-level rise, closed the connection between the Yoldia Sea and the ocean, initiating the Ancylus Lake Stage (Fig. 165C). Opinions differ as regards the time of this event, however, most frequently 9600 BP is offered. Although, the ice sheet had already strongly receded by that time, its influence on the rapid Ancylus Lake transgression was considerable. The maximum of the transgression was asynchronous in different regions, and in the area of the Gulf of Finland it culminated somewhat 9200-9000 BP (Haila & Raukas 1992).

The transitional stage of the Echeneis Sea between the Yoldia and Ancylus, recognized originally by Thomasson and later by Sauramo and several other authors, is obviously based on erroneous correlations and interpretations of diatom stratigraphy (e.g. Svensson 1989, Hyvärinen & Eronen 1979). As there seems to be no valid evidence for this stage, it should be dropped (Kessel & Raukas 1988, Kessel et al. 1988).

The Ancylus Stage derived its name from the fresh-water snail Ancylus fluviatilis whose remains were found in the beach deposits on the islands of Gotland and Öland and on the coast of Estonia. The diatom flora of the Ancylus period consists of oligohalobous species, the so-called Ancylus forms. Melosira islandica ssp. helvetica is the dominant species in planktic assemblages. In the off-shore lithostratigraphy, the Ancylus stage is represented mainly by massive, sulphide-stained clays. It is, however, difficult to draw a boundary between the Yoldia and Ancylus off-shore sediments.

According to the biostratigraphic data, the salinity of water in the Ancylus basin near Estonian coast could not have exceeded 3‰, because Ancylus fluviatilis is not able to inhabit the water with a salinity higher than that (Kessel & Raukas 1979). The Ancylus transgression was of the order of 10-15 m in the area of the Gulf of Finland, but in the southern parts of the Baltic its amplitude was larger.

Due to a strong erosion of the new outlet channel through the Danish Straits, a rapid regression started over the whole of Ancylus Lake. After a few hundred years, the lowering of the water level slowed down and around 8 500 BP the rising ocean level reached an equiniveau-position with the level of the Ancylus Lake. This event, the establishment of a sound connection between the two bodies of water, may be regarded as the end of the Ancylus Lake Stage.

The end of the Ancylus Lake and the beginning of the Litorina (Mastogloia) Sea is marked in the bottom sediments with a distinct boundary (Ignatius et al. 1981), but the reason, responsible for such a great change of sedimentary conditions was until recently unclear. Winterhalter (1992) has explained it with the inflow of saline water, which caused massive flocculation and deposition of suspended mineral particles. Due to the decrease of particulate material in suspension, sunlight could penetrate deeper down resulting in a rapid increase in organic productivity. The abundance of organic matter in conjunction with a marked decrease in fluvial transported mineral matter led, by his opinion, to a drastic change in the character of deposition.

To our mind, this is not the most likely explanation, because according to the diatom flora and mollusc fauna, the salinity changes both in the Ancylus Lake and Litorina Sea were not rapid at all. So, in the Arcona Depression and at Rügen Island, the saline conditions were established at about 8000 yr BP (Kliewe & Janke 1978), on the southwestern coast of Finland at about 7400-7300 yr BP (Eronen 1974), and at the head of the Gulf of Bothnia only at about 7000 yr BP (Eronen 1982).

The regression, following the maximum of the Ancylus transgression (Fig. 166), is often reported as having been rather rapid (Eronen & Haila 1982, Svensson 1989) and, according to Estonian material, had evidences of “catastrophic” outflow similar to the Baltic Ice Lake drainage. Already in the sixties, Kessel and Raukas (1967) identified a low Ancylus level (AVI), the deposits and relief forms of which have only partly preserved in the contemporary topography and are buried under the transgressive deposits of the Litorina Sea. It was established that the water level dropped about 20 m. According to Tavast (1995), in the Partsi gravel pit on Hiiumaa Island (Fig. 169) the Ancylus mollusc fauna, represented by Ancylus fluviatilis (5.5 %), Lymnaea baltica (60.7 %), Bithynia tentaculata (18.4 %), Valvata piscinalis (6.7 %), Pisidium amnicum (6.2 %) and Sphaerum nitidum (2.5%), is covered with a typical brackish-water Litorina fauna, including Cerastoderma glaucum (78.6%), Macoma baltica (13.6%), Hydrobia ulvae (5.1%) and Littorina littorea (1.1%), which provides a clear evidence of considerable paleogeographical changes in the Ancylus/Litorina transitional period. A Lymnaea baltica shell sample from Ancylus sediments has yielded an ESR calendar age of 8860 ± 70 yr BP corresponding to about 8000 conventional non-corrected radiocarbon yr BP. A Cerastoderma glaucum shell sample from the Litorina sediments has given an ESR age of 6310 ± 720 calendar years BP or about 5500 14C yr BP (Molodkov 1995).

If during the course of 300 years the water level in the Ancylus Lake rose 15-20 m above the ocean level (Eronen & Ristaniemi 1992), then the annual rate must have been 5-6 cm. During the same time span, the ocean level rose at a rate of about 1 cm per year (Pirazzoli 1991).

According to Svensson (1991), a rapid regression of 8-10 m on Gotland followed shortly after 9300 yr BP and the rate of regression during the next 600 years was only around 0.5 m per century. At the same time, on the Island of Hiiumaa, where the transgressive sediments of the Ancylus Lake at Kõpu are fixed at a height of 42-45 m (Kents 1939) and the regressive sediments at Partsi at a height of 10-15 m, the regression amplitude was much higher (including neotectonical uplift 30-35 m). In view of this, the regression rate between 9000 and 8000 yr BP should have been there at least 3.0-3.5 cm annually (Raukas et al. 1996).

The Litorina Sea, named after the gastropod genus Littorina, is the stage of maximum salinity in the history of the Baltic Sea. It is separated from the Ancylus Lake by the Mastogloia Sea, a transitional stage of low salinity, often regarded as a sub-stage of the Litorina Sea in a broad sense (Hyvärinen et al. 1988). The salinity of the sea water in this basin near the Estonian coast varied between 8 and 15‰ (Kessel & Raukas 1979).

Shore displacement curves of various authors show a varying number of fluctuations during Mastogloia-Litorina-Limnea time in the course of the last 8500 years. The nature of these fluctuations has been the subject of a great deal of discussion. If these fluctuations were not local, they should have been eustatic in origin and reflect the cyclicity in the global sea-level change. Taking into account a single marine Holocene transgression (Tapes/Litorina) with a broad culmination between about 8000 and 6000 BP on the western coast of Norway (Kaland 1984), a similar sea-level change in the Baltic seems to be most reliable (Hyvärinen 1979). All other short-term relative sea level fluctuations were probably local in origin and they were caused by climatic and tectonic factors (Raukas 1991b).

The transgression culminated at different times depending on the local rate of isostatic uplift (Fig. 166). Along the western sector of the Finnish coast, west of Helsinki, the eustatic rise never exceeded the isostatic uplift. In this area the interval 8000-6000 yr BP is characterized by a very slow regression, indicating that the eustatic and isostatic movements were nearly equal (Hyvärinen et al. 1992b). After about 6000 to 5000 yr BP, a more or less uniform regression seems to have prevailed also all over the Estonian coast. In Lithuania, the regression of the Litorina Sea in the Early Sub-Boreal about 4300 yr BP was interrupted by a new slight transgression which caused the sea level to rise a few metres (Gudelis 1979). In some papers, a weak Sub-Boreal transgression is reported also in other areas of the eastern Baltic. In Estonia this transgression is not fixed.

The Limnea Sea is the final stage in the history of the Baltic. It was first recognized in 1886 by G. Lindström on the basis of the disappearance of the genus Littorina and the introduction to the Baltic of the fresh-water mollusc Lymnaea ovata (Drap.) f. balthica Nilss. The lower boundary is gradual. It is proposed to define this boundary at 4000 years BP (Raukas et al. 1992), in agreement with the usage common in Sweden (e.g. Freden 1979). According to Kessel (Kessel 1958, 1961b, 1965), Lymnaea ovata (Radix peregra f. baltica) immigrated to the coastal waters of Estonia about 4000 yr BP and already at the beginning of the Limnea Stage, 2-37% of all mollusc shells in the North-Western Archipelago of Estonia belonged to this species; Lymnaea stagnalis appeared about 2500 and the typical fresh-water species Bithynia tentaculata about 1700 years ago (Kessel 1965). The index species of the Litorina Sea - Littorina littorea and L. Saxatilis, passed away from the Estonian coastal waters only about 1500 years ago when the salinity dropped below 7-8‰, critical for those species (Kessel 1958). Lagoon and shallow-water deposits of Limnea age are usually characterized by weakly brackish diatom assemblages similar to recent assemblages. The Limnea shorelines, recognized in Estonia, are all regressive. The salinity was about 10‰ in the western part and about 5‰ in the eastern part of the Gulf of Finland. Around 2500 years ago, the salinities were still 2 to 3‰ higher than today (Kessel 1965).

In spite of more than a hundred-year-long investigations of the Baltic Sea sediments and shorelines in Estonia, some conclusions are still rather hyphothetical and contradictory. So, for the establishment of real water-level fluctuations and shoreline displacements for selected time slices we need much better knowledge of the actual rates of tectonic and glacio-isostatic movements of the Earth’s crust which undoubtedly differed in every concrete area in time and space. As to the chronologically fixed events, truthful information is needed concerning the possible correlation between 14C, ESR, OSL, varvo-chronological and tree-ring data. Besides, agreements in using 14C dates (corrected and/or noncorrected against tree-ring data) should also be reached.

To reconstruct palaeoenvironment, much greater knowledge of the climatic conditions (circulation of currents, storm frequencies, wave activities, etc.) in the past is needed. On the contemporary seashores of Estonia, differences between low and high water level are about 1-1.5 metres, however, in the event of exceptionally fierce storms they may reach 2-3 metres. It means that contemporary beach ridges may be a couple of metres higher above the normal water level and some distance inland from the shore. Due to slope processes, wind and water erosion, it is sometimes difficult to establish the actual heights of beach ridges.

For the further understanding of the suspectibility of the Baltic Sea ecosystem to global changes and elaboration of strategies for the integrated management of the sea and coastal zone, the reconstruction of the basin development since deglaciation will be of the first-rate importance. The study of sea level displacements is an important prerequisite to successful modelling of palaeoenvironmental processes and to better understanding of former ecosystems. Owing to the moderate uplift of the Earth’s crust, which differed remarkably with the areas, the coastal relief forms in Estonia are diverser and more pronounced than in neighbouring areas. In view of the above, the Estonian coasts are of great scientific significance since, on the one hand, they allow a glimpse into the late- and postglacial geological history of the Baltic Sea and, on the other hand, make an excellent laboratory where geomorphological and environmental processes and the effects of the changes taking place on land and in near-shore areas can be studied in particular detail.

During the last decades, Estonian beaches have repeatedly suffered heavy storm damage (Photo 51). Therefore, the management of the coastal zone should be carefully planned taking into consideration the possible sea level rise to be expected in the future due to the “greenhouse effect”.



Holocene terrestrial processes

River activity

A. Miidel & T. Hang


The rather young drainage system in Estonia has formed and developed during the last 13,000 years. The formation of river valleys was closely related with the deglaciation of the territory, the evolution of the Baltic Sea and unequal glacioisostatic uplift. Besides, other factors affecting the development of river valleys included the bedrock topography, lithology of Palaeozoic rocks, geomorphology and lithology of Quaternary deposits. The location and orientation of rivers was also controlled by tectonic joints (Tammekann 1926, Teichert 1927a, Miidel 1966a, b, 1971, 1982). It is particularly obvious with the rivers cut in the bedrock.

On the ground of geomorphology and geology, Orviku (1960c) divided the Estonian river valleys into three groups, distinguishing between the rivers in the catchment area of: (1) the Gulf of Finland, (2) the Väinameri and the Gulf of Riga, and (3) lakes Peipsi and Võrtsjärv.

The rivers of the above-mentioned groups differ clearly in the shape of the longitudinal profile (Fig.170). The North-Estonian rivers have convex longitudinal profiles. In the lower courses, the stream gradient is great, often more than 2-3, occasionally even 5-8 m/km (Miidel 1963). In many cases, 30-50% of the total fall of rivers takes place within short sections (Photo 52) forming only 3-25% of the total river length. The longitudinal profiles of the West-Estonian rivers flowing into the Gulf of Riga and Väinameri are straight, with gradients remaining more or less constant at the whole length of the rivers (Fig.170). Concave longitudinal profiles are typical of the South-Estonian rivers flowing into the southern part of L. Peipsi and L. Võrtsjärv. For instance, in the longitudinal profile of the Piusa River, the lower course forms about 46% of the river’s total length, but only 13% of the total fall takes place within this section (Liblik 1966). Numerous studies (Orviku 1960c, Miidel 1963, 1966b, Hang & Miidel 1987) have shown that the longitudinal profiles of Estonian rivers reflect main elements of the bedrock topography, glacial landforms and deposits, furnishing thus an excellent example of a young drainage system with the unevennesses of the profiles not yet levelled.

The thickness of alluvial deposits in the North- and South-Estonian river valleys differs notably. In northern Estonia, the thickness of fluvial deposits increases from the upper course (1-3 m) towards the middle course (6-7 m), from there onward it decreases steadily. In the lower course, the deposits are only 1-2, occasionally 4-5 m thick (Miidel & Raukas 1965, Miidel 1966b). The valleys in southern Estonia have a rather thick alluvial cover in the lower courses - up to 10-15 m together with peat (Fig. 171; Kajak 1959, Miidel 1966b).

It is interesting to follow the changes in the relation of overbank and channel facies along the rivers. However, it is sometimes very difficult to distinguish between these facies, particularly in places where the river valleys have been cut in fine-grained glaciolacustrine deposits. In these cases, river deposits and facies are lithologically very similar to the source material. Nevertheless, the general tendencies have been determined. In the valleys of northern Estonia, the proportion of channel deposits increases from the middle courses downstream - from 40-50 to 60-70% of the total thickness. By means of detailed grain-size analysis it has sometimes been possible to distinguish between point bar and channel lag subfacies (Hang & Niedzialkowska 1994). Oxbow deposits are of limited distribution. In the lower courses, where incision dominated in the Holocene, palaeochannels are only partly filled with sediments. They are often empty and dry. In the middle reaches, oxbow deposits are of wider distribution. In the grain-size they are very similar to overbank deposits which makes their identification rather problematic.

In the South-Estonian valleys, the role of different facies is just reversed (Fig. 171). The share of channel facies diminishes downstream and in the lower courses it forms less than 20-30% of the whole fluvial sequence. The channel facies is at its thickest (4-5m or 50-60%) in the middle reaches of the rivers (Hang 1995). The overbank facies is characterised by a high content of fen peat. In the lower reaches, its thickness ranges from 5 to 10 m (Kajak 1959, Miidel 1966b, Miidel & Tavast 1981, Miidel et al. 1995).

The grain-size and mineral composition of alluvial deposits is immediately controlled by the stream gradient, the bedrock and Quaternary deposits. The main influx of sediments into the channel is by creep and sliding from the undercut valley fill of various genesis, and from older river terraces. Thus, the grain-size and mineralogy of river deposits are strongly influenced by local geology. It is reflected most distinctly in the character of channel deposits. In northern Estonia, channel deposits often contain boulders, cobbles and shingles of the Palaeozoic carbonate rocks. The channel deposits of specific character are formed in the valleys, cut in till. As the finer fractions are washed out, coarse material (gravel, boulders, cobbles) remains in the channel. The further downcutting is hampered by the accumulated layer of very coarse material. Rapids with boulders of crystalline rocks typically occur in those channels. In the middle courses of rivers, where mainly fine-grained glaciolacustrine sediments are eroded, the channel deposits contain an abundance of fine-grained material with clay and silt fractions (sometimes more than 60%). In the overbank deposits, fine sand and coarse silt prevail. The mineral composition of alluvial deposits also reveals a great influence of local geology. In both channel and overbank deposits among light minerals (fraction 0.1 - 0.25 mm) quartz and feldspars dominate. The content of carbonate minerals is usually less than 5%. In the lower courses, just downstream the North-Estonian Klint, glauconite is present in notable amount. Among the heavy minerals, amphiboles, garnet, magnetite and ilmenite are prevailing. This kind of mineral association is typical of the deposits occurring in the area of the last glaciation. It should also be mentioned that the overbank and channel facies do not significantly differ in mineral composition (Miidel & Raukas 1965).

In southern Estonia, where Devonian terrigenous rocks crop out, the alluvial deposits consist mainly of sands with a variable grain-size. For example, the overbank sediments of the Piusa River are dominated by fine (0.1-0.25 mm, 55%) and medium sand (0.25-0.5 mm), while the channel facies consists prevailingly of medium (55%) and coarse (0.5-1.0 mm, 25%) sand.

According to Aivo Lepland (unpublished data), guartz (over 90%) and feldspars are the prevailing light minerals (fraction 0.1-0.25 mm) in both channel and overbank facies. Among heavy minerals, amphiboles, garnet and magnetite dominate. Compared to northern Estonia, the content of zircon, tourmaline, staurolite and leucoxene in the South-Estonian rivers is clearly higher. This is explained with the influence of Devonian terrigenous rocks in which these minerals are common.

The morphology and development of the river valleys in Estonia has been discussed in a number of studies (Tammekann 1926, Künnapuu 1957, Kajak 1959, Orviku 1960c, Arold 1960, 1971, Linkrus 1963, Hang et al. 1964, Liblik 1966, Miidel 1966b, 1967, Miidel & Tavast 1981, Eberhards & Miidel 1984, Hang & Miidel 1987, Miidel & Raukas 1990, 1991, Hang 1995, Raukas & Miidel 1995, a.o.). The results of these studies have shown that the morphology and evolution of river valleys in northern and southern Estonia was substantially different. In the north, the valleys are better developed in the lower courses where deep, V-shaped valleys, short canyons and flat-floored valleys prevail. Their depth ranges from 15 to 35 m. The valleys are at their deepest a little downstream of the North-Estonian Klint. In the upper and middle courses, either V-shaped or only channel valleys have developed in the zones of marginal glacial formations, or flat-floored valleys occur in the plains of various genesis. Usually the depth of these valleys does not exceed 10 m. Morphological valley is missing in the peatlands, which are common in the upper and middle courses of the North-Estonian rivers. The development of these valleys has been significantly affected by the Baltic Klint - a steep Palaeozoic escarpment which rises up to 56 m above sea level. The Baltic Klint emerged from the sea step-by-step; earlier in the east after the drainage of the Baltic Ice Lake and later in the west, during the regression of the Ancylus Lake. The klint with its cap from resistant carbonate rocks turned into a permanent base-level for rivers. Together with the numerous related waterfalls (Photo 19) the klint prevented the erosion from penetrating further inland (Miidel 1967). This explains why under the conditions of the lowering sea-level and continuous land uplift numerous terraces were formed only in the fore-klint reaches of the North-Estonian valleys.

The number of the terraces in the lower courses (Photo 53) ranges from 3 (Pühajõgi Valley, Tammekann 1926) to 11 (Valgejõgi Valley, Linkrus 1963). They are represented by small segments, measuring 30 - 300 m in length and 10 -100 m in width. The surface of the terraces is more or less inclined downstream and towards the channel. They usually have a thin alluvial cover of coarse sand and gravel, the overbank facies is lacking. The altitude and the time of formation of the terraces have been correlated with the numerous shorelines of the Baltic Sea (Tammekann 1926, Künnapuu 1957, Arold 1960, Linkrus 1963, Hang et al. 1964, Miidel 1967, a.o.). It is worth of mentioning that not a single Baltic Sea stage or level has a corresponding terrace in the North-Estonian valleys. Consequently these, mainly erosional (rock-cut and fill-cut), terraces belong to the type of unpaired terraces. The spectrums of North-Estonian river terraces are opened towards the mouth and, having a fan-like shape, point to the continuous rejuvenation caused by the land uplift and lowering base-level (Fig.172; Miidel 1967, Hang & Miidel 1987).

The land uplift had a different effect on the development of rivers in northern and western Estonia. The isobases of the transgressive shoreline of the Baltic Ice Lake indicate a remarkable difference in the uplift (nearly 20 m) between the upper and lower courses of the rivers flowing to the northwest. For the rivers, flowing to the south-west or even west, it is less than 8 m. In the first case, the distribution of the uplift along rivers caused a relatively strong tilting of the surface to the south-east. As the monoclinally bedded bedrock had a slight southward inclination and the topography was flat, shallow valleys with wet boggy floodplains and without a single terrace on the slopes, developed in the middle reaches (Miidel 1963, 1966b). It is quite possible that the development of rivers was characterized by a modest incision or even by a slight aggradation and shifting of the channel zone back and forth, more or less at the same level (Raukas & Miidel 1995). With the rivers flowing into the Gulf of Riga and Väinameri, the uplift along channels was relatively moderate and more or less equal. The river activity in western Estonia depended mainly on eustatic sea-level changes and local geology. In both cases, the regression of the Baltic Ice Lake brought about considerable elongation of the rivers (Orviku 1969).

Two stages of development have been distinguished in the evolution of the North-Estonian river valleys (Hang & Miidel 1987). The first stage started with the formation of glaciofluvial deltas in the Baltic klint bays east of Tallinn and lasted until the rising of the klint out from the sea. This moment marks the outset of the second stage which is still in progress. In the timescale, the boundary of the stages varies with the regions. In the western part, as far as the Jägala River, it coincides with the regression of the Ancylus Lake and the following drop in the Baltic Sea level. In the eastern part, the second stage commenced earlier, after the retreat of the Baltic Ice Lake. During the first stage, fluvial activity was probably modest, while the second stage was a period of intense downcutting, leading to the formation of deep valleys in the lower courses of the rivers.

In southern Estonia, V-shaped valleys are prevailing in the upper and middle courses where they have cut into the complicated topography of the Haanja and Otepää uplands. Due to the changing glacial topography, V-shaped valley sections alternate with flat-floored ones. The depth of the valleys varies between 15-30 m and the width ranges from 200 to 500 m. Descending from the uplands to the surrounding plains, valleys reach their maximum depth (30-45 m). Towards the river mouth, the depth of the valleys gradually decreases. However, at the mouths, valley bottoms lie 6-15 m lower than the present water-level of lakes Peipsi and Võrtsjärv (Kajak 1959, Liblik 1966, Miidel 1966b, Miidel & Tavast 1981). As an average, these valleys are 0.7-1 km, occasionally up to 2 km wide. The wide floodplain is subject to paludification in the middle and lower courses where the thickness of the resultant fen peat reaches 5-10 m (Orviku 1960c, Liblik 1966, Miidel 1966b, Miidel et al. 1995).

The number of terraces left on the valley slopes by river activity (Hang et al. 1964, Liblik 1966, Hang 1995, Hang et al. 1996) ranges from six (Võhandu Valley) to 16 (Ahja Valley). The terrace levels form some groups, 2-4 levels in each (Fig. 173). In the Piusa Valley, the highest terraces are 500-600 m wide (Liblik 1966) and up to 300 m long. However, their relative height is only 2-2.5 or even less metres. The terraces were cut mostly into glaciolacustrine and glaciofluvial deposits, but in some cases also into till and bedrock. In the rock-cut terraces, the alluvial cover is thin (1-1.5 m) consisting of coarse sand and gravel with pebbles. It is sometimes difficult to distinguish alluvial deposits from glaciolacustrine or glaciofluvial ones. Nevertheless, the terraces are undoubtedly erosional.

The development of rivers in Estonia started from the south according how the territory was freed from the ice cover. The large Võru Valley came into being some 12,600 yr BP, when the glacier of the Otepää Stadial came to a halt along the line Kulje - Talabsi islands - Elizarovo, and the Pihkva Ice Lake I was formed (Raukas & Rähni 1969). The meltwaters flowed westwards towards the Gauja Basin. Many geologists have recognized the Võru Valley as a marginal meltwater spillway (Hausen 1913b, Hang et al. 1964, Raukas & Rähni 1969, Raukas et al. 1971, Kvasov 1975, Miidel & Raukas 1991, a.o.). However, after Lepland (1991) and Hang (1995), fine-grained glaciolacustrine deposits with horizontal bedding in the consistence of more or less horizontal terraces in the Võru Valley indicate glaciolacustrine origin of those terraces. Accordingly, it was supposed that there was no meltwater flow from the east to the west along the Võru Valley, and this depression was occupied by ice-dammed lakes in which the water-level gradually sank. The formation of river terraces began in the Piusa, Võhandu and Ahja valleys with the lowering of lake-level down to the altitude of 75-70 m. The further retreat of the glacier and lowering of the water-level in ice-dammed lakes led to the incision in the valleys. When the glacier margin stopped on the line Kaiu - Gdov, the outflow from the Peipsi Ice Lake l took place via the Väike-Emajõgi Valley to the south (Raukas et al. 1971). Further retreat of the glacier to the north led to the formation of the Emajõgi Valley and later the Viljandi Valley (Raukas et al. 1971), through which meltwaters flowed from the Peipsi ice-dammed lake to the west.

After Miidel and Raukas (1991), the terrace formation processes in southern Estonia completed in the Bølling. If the formation of the youngest (lowermost) terraces are correlated with the Männikvälja - Iisaku marginal formations (Fig. 174), belonging to the Pandivere Stadial, the formation of terraces should have ended about 12,500 yr BP (Hang 1995, Hang et al. 1996). The varve chronology (Raukas et al. 1971) shows that the terrace formation process developed extremely quickly - during the course of 200-300 years (Hang 1995).

The low position of the valley bottoms in the southern part of lakes Peipsi and Võrtsjärv indicate the continuous downcutting in the river valleys after the terrace formation due to the lowering base-level. As a result, in the south the lake depressions dried up and the rivers cut down even as deep as 13-16 m below the present lake-level (Kajak 1959, Orviku 1960c, 1969, Hang et al. 1964, Miidel 1966b, Miidel & Tavast 1981, a.o.). Probably it took place during the Younger Dryas or later (Orviku 1960c, Raukas & Rähni 1969, Hang et al. 1996). As a result of the uneven land uplift, the base-level rose. Alluvial deposits started to accumulate and re-deepened valleys were subject to paludification. In the Optjok River, it commenced during the Early Boreal (Miidel et al. 1995), in the mouth of the Võhandu River during the Boreal (Pirrus & Tassa 1981) and in the mouth of the Emajõgi River in the Late Atlantic (Thomson 1939b, Sarv & Ilves 1975). The process is still in progress. It has been supposed that the floodplains in the South-Estonian valleys were formed in the Holocene. Dating of oxbow sediments from the Piusa and Võhandu valleys indicates that the meandering of rivers and the formation of accumulative floodplains with organic deposits started in the Pre-Boreal or Early Boreal together with the rise of the base-level (Fig. 175).

As stressed above, the river valleys in northern and southern Estonia have developed in different ways (see Fig. 176).


Lake Peipsi

A. Miidel & A. Raukas


Lake Peipsi (3555 sq km) on the border with Russia, ranks fourth in size among the European lakes. Its average depth is 8 m and maximum depth 15.3 m. The lake consists of three main parts: the northern section - Peipsi proper is connected with the southernmost part L. Pihkva through the narrow, strait-like L. Lämmijärv.

Lake Peipsi lies in a vast ice-lobe depression which in the Late Weichselian was occupied by ice-dammed lakes. The contours and water-level of these bodies of water were controlled by the northward retreat of the glacier and opening of new outlets.

During the Otepää Stadial some 12,600 years ago (Raukas et al. 1971), when the ice margin retreated to the line Kulje - Lisje - Talabsi islands - Elizarovo, a big ice-dammed lake was formed in the southernmost part of the lake depression (Raukas & Rähni 1969, Fig. 177a). This was Pihkva Ice Lake I with the water-levels 95, 85 and 75 m a.s.l. From this lake, meltwaters flowed to the southwest along the Võru Valley (known also as the Piusa-Võru-Hargla Valley) into the proglacial lakes of the Gauja Basin (Fig. 177a). In the Võru Valley, group A terraces were formed at altitudes between 71.5 and 95 m (Liblik 1966). According to Kvasov (1975, 1979), this lake was only the western part of the large Privalday Ice Lake which came into being about 2000 years earlier and had now an outlet via the Võru Valley.

Lately it was stated (Lepland 1991, Hang 1995, Hang et al. 1996) that the terraces in the Võru Valley are glaciolacustrine, not glaciofluvial in origin because they consist of horizontally bedded fine-grained silts and sands and the terraces have no inclination towards the supposed flow direction. In all likelihood, the Võru Valley was a wide strait which connected ice-lakes in the east and west (Fig. 177a). Further studies are needed to estimate the origin of the terraces in the valley.

Afterwards, when the lake level dropped to 75 m, meltwater flow to the west ceased. It is possible that the outflow was restored along a new strait via the upper courses of the Ahja and Võhandu rivers. However, this connection existed only for a short period.

When the glacier retreated to the line Mehikoorma - Pnevo - Remda, where a dump moraine was formed, the water level sank to 62-60 m a.s.l. and Pihkva Ice Lake II came into being (Raukas & Rähni 1969, Raukas et al. 1971, see also Fig. 177b). The inflow was from the surrounding heights, occupied by dead ice. The formation of the Ahja and Võhandu valleys started (Fig. 174). There was probably no outflow. The next belt of ice marginal formations has been established on Piirissaar Island and Knyazya Gora (east coast of L. Peipsi, Fig. 178). The water level was at a height of 60 m a.s.l. (Raukas & Rähni 1969). According to Hang (1995), the terraces of group B in the Ahja and Piusa valleys were synchronous with the Piirissaar glaciofluvial delta and Knyazya Gora end moraine line. Thus, at Knyazya Gora the water level must have been somewhat higher than 60 m a.s.l. The outflow was probably via the Väike-Emajõgi Valley (Fig. 174). After Hausen (1913b), at that time there existed a large Pihkva Ice Lake with its water level 75 m a.s.l. From this lake meltwaters flowed to the west through the Võru Valley and to the southwest via the Valga-Valmiera sandur into the Gauja basin. According to Kvasov (1975, 1979), the waters of L. Novgorod, a remnant of the split L. Privalday, flowed into Pihkva Ice Lake through a short valley in the vicinity of the town of Porkhov. Later, the connection between Novgorod and Pihkva ice lakes was via a spillway in the middle course of the Luga River. However, Kvasov maintains that Kemba and Voose ice lakes on the northwestern slope of the Pandivere Upland were also synchronous with L. Novgorod, but actually they were formed later when the glacier had retreated northwards (Pärna 1960, Raukas & Rähni 1969, Raukas et al. 1971).

The further withdrawal of the glacier northwards with a following new advance to the line Kaiu - Gdov about 12,250 yr BP (Fig. 177c) led to the formation of Peipsi Ice Lake I (Raukas et al. 1971). Different views have been expressed as to the height of the lake level. According to Raukas and Rähni (1969), in the north-west the lake level was at a height of 86 m a.s.l., at Kaiu 75 m and at Mehikoorma 40 m a.s.l. (Fig. 178). As is known, Peipsi Ice Lake corresponds to the South Peipsi Ice Lake by Hausen (1913b) where the water level was only 36-37 m a.s.l., but Hausen failed to establish any shoreline there. Hang (1995) thinks that the terraces of group C in the Ahja and Piusa valleys were formed when the glacier came to a halt at the line Kaiu - Gdov. Considering the altitude of river terraces in these valleys (54-50 and 51-48 m a.s.l.) and the uplift gradient, the water level at Kaiu and Gdov might have been about 51-56 m a.s.l. (Fig. 178).

During the Pandivere Stadial when the glacier readvanced again and came to a halt along the Männikvälja - Iisaku - Vaivara ice marginal formations (Fig. 177d), Peipsi Ice Lake (PeIII after Raukas & Rähni 1969) formed a single body of water with glacial lakes in the east (Luuga, Neeva a.o.). Kvasov (1975, 1979) termed it L. Ramsay which corresponded approximately to Great Peipsi Ice Lake by Hausen (1913b). According to Raukas and Rähni (1969), the water level in the lake was 80 m a.s.l. at Saare, 70 m at Iisaku and 43 m a.s.l. at Kavastu (Figs. 177d, 178). But after Hausen (1913b), it was 53 m a.s.l. at Iisaku. Hang (1995) associates the highest level of the terrace group D in the Ahja and Piusa valleys (both 41 m a.s.l.) with the shoreline at a height of 47-45.5 m (Fig. 174) which was determined between Kallaste and Kavastu by Liblik (1969). On this basis, the calculations have given 50 m a.s.l. for the height of the lake level at Iisaku. At the same time or a bit later, the Emajõgi Valley was formed. At Tartu, the water level was 42-43 m a.s.l. (Mieler 1926, 1927) and its lowering led to incision of meltwaters into Devonian rocks.

After the retreat of the ice from the northern slope of the Pandivere Upland in the vicinity of Männikvälja (Fig. 177a) and Uljaste, ice lakes west and east of the Pandivere Upland joined up (Fig. 167). The event is acknowledged as the beginning of the Baltic Ice Lake (Kvasov & Raukas 1970).

After the retreat of the glacier into the Gulf of Finland, the water level dropped and the Peipsi Depression was isolated from the glacial lake, situated in the Gulf. It is supposed that the southern part of the Peipsi Depression dried up in the Younger Dryas or at the beginning of the Holocene (Orviku 1960c, Raukas & Rähni 1969, Miidel & Tavast 1981, a.o.). However, opinions have also been expressed that it happened considerably earlier - about 12,000 years ago (Kvasov 1975, 1979, Hang et al. 1996). In both cases, it is supposed that the northern part of the depression was occupied by L. Small Peipsi, into which the Velikaya River and its tributaries (Emajõgi, Võhandu a.o.) discharged (Fig. 177e).

At the beginning of the Pre-Boreal Chronozone, a shallow lake existed in the southern part of L. Peipsi (Pirrus et al. 1985, Davydova & Kimmel 1991, Miidel et al. 1995, Hang et al. 1996) in which either silts (in the mouth of the Optjok River, Värska Bay, at the Meeksi Brook) or fine-grained sands (in L. Lämmijärv) were deposited (Fig. 178). It is not exactly known what was the altitude of the water level, but in the mouth of the Optjok River it must have been considerably lower than today (Miidel et al. 1995). Pollen evidence (Pirrus & Tassa 1981, Pirrus et al. 1985, Davydova & Kimmel 1991, Miidel et al. 1995) suggests that at that time the lake was, in general, shallow and its surroundings were paludified. However, there were also deeper areas in the lake, the greatest depth being at least 7.5 m in L. Lämmijärv (Fig. 178).

At the end of the Pre-Boreal (mouth of the Optjok River, Fig. 178) or at the beginning of the Boreal (Värska Bay, Fig. 178), the silty deposits became overlain by organic rich silt or fen peat containing silt. Probably, this abrupt change in sedimentation was caused by a lowering of the lake level. When the formation of fen peat commenced at the mouth of the Optjok River, the water level must have been at least 10 m lower than at present (Miidel et al. 1995, Hang et al. 1996). The southern part of the lake basin was occupied by a swamp. The lowermost water level in the Holocene between 10,000-9,000 yr BP, was marked by a break in sedimentation (Davydova & Kimmel 1991, Miidel et al. 1995, Hang et al. 1996).

The slow rise of the water level, following its low stand in the Pre-Boreal, started in the mouth of the Optjok River in the first half (Miidel et al. 1995) and in Värska Bay in the second half of the Boreal (Pirrus & Tassa 1981). In the deepest part of the lake, silt sedimentation started anew, but the lake remained shallow (Davydova & Kimmel 1991). This period corresponds to Small Pihkva Lake by Rähni (1973).

During the Atlantic Chronozone, the accumulation of reed peat with shell fragments continued at the mouth of the Optjok River (Miidel et al. 1995). In the beginning of the period, paludification started in the mouth of the Kunest River (Fig. 178), in the second half - in the mouth of the Rovya River (Miidel et al. 1975) and Emajõgi River (Thomson 1939b, Sarv & Ilves 1975). At the end of the chronozone, paludification on the Island of Gorodets (south-east of Piirisaar Island) commenced (Miidel et al. 1975). In the north, the lake submerged the area between Omedu and Rannapungerja, its waters extending as far as an ancient shoreline at Raadna (Fig. 178). The water level was 5-6 m higher there than at present. According to Rähni (1973), this lake was the Atlantic Peipsi. It is possible that the outflow from the lake via the Narva River was formed in the Atlantic.

In the second half of the Sub-Boreal, the water level rose rapidly (Miidel et al. 1995, Hang et al. 1996). The rate was so high that the development of a bog at Laane ceased and the bog was submerged (Pirrus et al. 1985). Swampy conditions spread around the mouth of the Samolva River (Miidel et al. 1975).

The water-level rise continued in the Sub-Atlantic. According to Rähni (1973), the rapid water-level rise (the Atlantic Great Peipsi) was followed by an abrupt fall of 1.5-2 m. In the Sub-Atlantic, peat deposits were buried under lake deposits in some places.

The water-level rise in the southern part of the depression is associated with the intense glacioisostatic uplift which was faster in the north (Hausen 1913b, Ramsay 1929). The water-level rise was controlled by both tectonic and climatic factors (Miidel et al. 1995, Hang et al. 1996).

The curve of water-level changes, compiled for the southern part of L. Peipsi (Hang et al. 1996), demonstrates a very fast water-level lowering at the end of the Late Weichselian and the succeeding continuous rise since the beginning of the Boreal and fastening in the Sub-Boreal Chronozone (Fig. 179).

The data concerning the Holocene water-level changes in the southern part of L. Peipsi and the Emajõgi (Sarv & Ilves 1975) and Väike-Emajõgi (Pirrus et al. 1993) rivers are in good agreement.

The lake continues to retreat southwards. According to Vallner with co-authors (Vallner et al. 1988), the northern part of the depression is rising at a rate of 0.2-0.4 mm, whereas the southern part is sinking at a rate of 0.8 mm per year. The tilting of the lake depression makes the water to flow from north to south. As a result, the banks of L. Pihkva are suffering from ever increasing erosion; wide stretches of lowlands around the lake have become paludified. In 1796, the area of Piirissaar Island in the southern part of L. Peipsi proper was 20.08 km² (Mieler 1926), to date it is only 7.39 km².

The water-level is slowly rising or more or less stable at the northern coast. Evidence is derived from coastal bluffs in dunes and buried peaty deposits. This is due to the relatively hard rocks, outcropping in the upper course of the Narva River, and long-shore drift, obstructing the outflow (Kajak 1964, Miidel 1966b, Raukas & Tavast 1996).

The study of the evolution of Lake Peipsi has a long history, but many topical problems have remained unsolved. There are several gaps in the deglaciation history because of the lack of clear ice-marginal formations in the Peipsi Lowland and in its vicinity. Late-glacial river terraces have not yet been properly dated and, frequently, the water-level fluctuations are the best traceable indirect markers (the absolute height of the flat tops of eskers and kames, scarps, developed on the banks of the ice-dammed lakes, boulder fields, etc.). In the light of the present knowledge, it may be supposed that there were no long-lasting halts during the retreat of the ice cover promoting the formation of clear terrace surfaces. It is not excluded that the lake depression was filled with passive and dead ice. In this case, the evolution of the lake in the Late-glacial may be interpreted in a principally different way.

There are much more data available on the Holocene history of the lake which was controlled by the neotectonic movements, well dated by means of peat accumulation. Like the majority of lakes in the Northern Hemisphere, Peipsi has a more open eastern and a more swampy and overgrown western bank. Due to the prevailing south-westerly and westerly winds, the active erosion-accumulative or erosional shores are spread in the eastern and northern (Photo 54) parts of the lake, while the swampy coasts overgrown with bushes, bulrush and reed are characteristic of its western (Photo 55) and southern parts (Fig. 178).


Lake Võrtsjärv

A. Raukas & E. Tavast


Võrtsjärv, the second largest lake in Estonia, has a surface area of 270.7 km². Its maximum length is 34.8 km and maximum width 14.8 km. The length of the weakly dissected shoreline measures 96 km, maximum depth is about 6 m, average depth 2.8 m, long-term water level stand 33.68 m, volume 756 million m3 of water, catchment area 3380 km2. The main tributaries number 18, the outflow is via the Emajõgi River (Mäemets & Raukas 1995).

The lake depression was formed in pre-Quaternary time, but during the course of thousands of years its shape has been radically altered by the glaciers. The orientation of the drumlins and clasts in the till of the last glaciation (Raukas & Tavast 1990) shows a more or less meridional direction of the ice movement between the Sakala and Ugandi plateaus.

The bedrock, mainly sand- and siltstones of the Middle Devonian Aruküla Stage, is exposed on the lake’s steep east bank at Tamme (up to 8.5 m high), Trepimägi and Petseri (Fig. 180). Devonian rocks are rich in quartz (75-90 %) and micas (1-10 %). Heavy minerals are dominated by ilmenite, zircon, garnet and tourmaline (Kleesment 1994). The same minerals are widespread in the lake deposits. The carbonates (up to 14.1 % of fine sand fraction), amphiboles and pyroxenes (up to 30 % of heavy subfraction) occurring in beach deposits have been washed out from glacial sediments (Tavast 1990), rich in those minerals (Raukas 1978).

The depression around the lake is covered with till, glaciofluvial sand and gravel, glaciolacustrine silt and clay, gyttja, lake marl and peat (Orviku L. 1958). In some places aeolian and alluvial sandy-silty sediments occur. The thickness of the deposits is mainly 5-10 m, seldom more (Raukas 1978).

Vast stretches of the low-lying shore are overgrown with reed and bulrush. In the west and east, reed forms an up-to-100-metre-wide belt. The small depth, high water temperature in summer and increased concentration of mineral nutrients promote overgrowing of the lake. The encroaching reed will hardly reduce the area of beaches available for recreational use (Tavast et al. 1983b). The best sandy beaches are in the north at Vaibla and in the east in Vehendi Bay. Figure 180 shows different types of contemporary beaches and prevailing grain-size of shore deposits.

The bottom sediments consist mostly of fine sand and silt, sapropel (up to 9 m) and lacustrine lime (up to 8 m thick). In the northern part of the lake, sapropel and lacustrine lime either form a thin layer or are entirely absent (Veber 1973). About two thirds of the topmost part of the lake’s sediments consist of sapropel (gyttja) and sandy sapropel (Fig. 181), with the total volume of about 200 million m3, together with the lake marl ca 360 million m3 (Veber 1973). Organic matter forms 87-92% of the sapropel. Silty clay, lacustrine lime and other types of sediments are less abundant. In places, especially in the northern part where the bottom sediments are absent, the lake depression exposes varved clay or till. In the southern part of the lake, the sediments are much thicker than in the northern part indicating a gradual rise of the water level in the southern portion of the basin (Pirrus & Raukas 1984). Figure 182 presents a characteristic cross-section of lake sediments.

The lake has a complicated history (Fig. 182). Glacial lakes of different shape and size were formed immediately in front of the retreating ice cover about 12,600 (Fig. 183A), 12,250 (Fig. 183B) and 12,050 (Fig. 183C) yr BP during the Otepää, Sakala and Palivere stadials, respectively (Raukas 1986). The outflow from those lakes was first to the south via the Väike-Emajõgi Valley to the basin of the Gauja River. Afterwards new outflows were formed to the west via the Viljandi Valley. In the Younger Dryas, at the end of the Late-glacial when meltwater of the glacier diminished, Lake Ancient Võrtsjärv (Orviku 1973) came into being in the Võrtsjärv Depression. Due to the neotectonic uplift, which was more intensive in the north-west, the outflow to the west gradually diminished and closed in the Early Holocene. In the depression, Lake Big Võrtsjärv (Orviku L. 1958) was formed (Fig. 183D). The water level in the northern part of the basin was 4-5 m higher than today. At the beginning of the Middle Holocene, about 7500 yr BP, an outflow to the east developed via the Emajõgi Valley (Fig. 183E), and gradually the lake acquired its present contours (Orviku 1973).

At present, due to the uneven neotectonic uplift of the lake’s depression, Lake Võrtsjärv is steadily retreating southwards inundating new areas (Jaani 1973). The water level in the lake is very unstable (in 1922 the maximum annual amplitude was 2.2 m, the maximum difference of the water table is up to 3 m, the rise during the spring flood up to 174 cm).

In the last century, the water level of the lake was about one metre higher than today. In the 1920s, the outlet of the Emajõgi River was thoroughly dredged, and the lake level dropped. Lake Võrtsjärv is rich in phytoplankton, 36 fish species inhabit the lake. Valuable commercial fish form about 70 per cent of the total catch (Mäemets & Raukas 1995). The regulation of water level would create more favourable conditions for the spawning of valuable fish and survival of their larvae. However, beforehand, careful study of potential environmental consequences is needed (Raukas & Tavast 1990). Without doubt, rising of the lake level will promote erosion processes both in Lake Võrtsjärv and on the banks of the Emajõgi River.


Small lakes

L. Saarse


There are about 1500 small lakes with a surface area less than 10 km2 located irregularly in different landscape regions in Estonia (Mäemets & Saarse 1995). The Haanja and Otepää heights, the Saadjärve Drumlin Field and the Kurtna Kame Field are dotted with lakes; in the West-Estonian Lowland their number is small. Because of the infilling and overgrowing, the number and area of lakes is constantly decreasing, but due to the land uplift in the west and north-west (up to 3 mm per year) isolation of new waterbodies from the Baltic Sea is in progress. More than half of small lakes are glacial in origin and scattered in Upper Estonia. Basins of small lakes are filled with Late-glacial sand, silt and clay, Holocene organic and calcareous deposits which store information on the postglacial stratigraphy, vegetational history and climate change. The longest lake records start from the Older Dryas, most commonly from the Younger Dryas. Organogeneous deposition began at the beginning of the Holocene with some delay in most kame field lakes.

In the following, we shall deal with the evolution of the main types of small lakes (Saarse 1990) through the palaeoecological regions (Saarse & Raukas 1984).


Lakes of the Middle and Upper Devonian plateaus

The Otepää and Haanja heights are characterized by a highly disjointed topography containing some four hundred lakes with a small surface area, disjointed bottom relief and varying trophic conditions (Mäemets 1977). Glacial and residual lakes are the basic types in this region. It is almost impossible to distinguish between the lakes originating from the irregularities of the glacial drift and lakes of glaciokarst origin. Glaciolacustrine beds are commonly absent, lacustrine clayey-silty sediments are thin or non-existent (Mäetilga, Vaskna, Tuuljärv, Kurgjärv, Väikjärv, Päidla, Mähe, Ahvenjärv, Räbijärv, Juusa; Fig. 184) (Ilves 1980, Mäemets 1983, Ilves & Mäemets 1987, Saarse 1994). Steep slopes and jointed catchment topography promote abundant input of minerogenous matter into the lakes. In the littoral belt of some lakes lacustrine lime and calcareous gyttja have deposited. In L. Kurgjärv peat is buried under gyttja (Fig. 185). Buried peat occurs also on the steep slopes of the lakes Pangodi, Mäetilga and Tuuljärv, obviously drifted down during the solifluction. Pollen and 14C records of bottom deposits indicate that organic sedimentation started at different times in the Pre-Boreal, but the lakes themselves, at least some of them, were evidently formed since the Haanja Stadial.

The basins of the residual lakes were first occupied by proglacial lakes. After the ice recession and disappearance of proglacial lakes, residual water bodies remained as separate lakes (Tamula, Pangodi, Kirikumäe, Pulli, Pühajärv). The mentioned evolutional changes are fixed in their bottom deposits: till or glaciofluvial sand and gravel covered with glaciolacustrine varved clays and laminated sands and silts, overlain by gyttja or calcareous gyttja. The thickness of organic lacustrine deposits varies from 2-3 m in Koobassaare, Saarjärv and Neitsijärv, up to 8 m in L. Pühajärv (Photo 56), 9 m in L. Vaskna and 11 m in L. Vagula (Ramst 1992). Biostratigraphical studies have been carried out from sequences of lakes Tamula (Pirrus 1969), Mäetilga, Kõverjärv (Mäemets 1983), Tuuljärv, Vaskna (Ilves & Mäemets 1987), Pulli, Kirikumäe, Punso (Saarse 1994), Pühajärv (unpublished), etc. Most of the sequences are supplemented by radiocarbon dates. From Pangodi the pollen diagram is absent and only radiocarbon dates are available (Ilves 1980).

Lakes in kettles left by melting ice blocks into pre-existing valleys and lakes formed in irregularities of hilly topography due to uneven deposition, are characteristic of the Sakala Upland and the Ugandi Plateau. Lake basins are elongated, they have uneven floors and a rather complicated sediment composition. In the Holocene part of the lacustrine deposits, calcareous and minerogenous gyttja dominates (Saarse et al. 1996a). Deep basins with steep underwater slopes and negligible assemblages of aquatic plants are poor in lacustrine deposits (Viljandi, Mäeküla, Parika, Pärsti; Ramst 1992), whereas in shallow (5-10 m) lakes organogenous deposits have accumulated (Õisu, Võistre, Kariste, Veisjärv). According to the pollen stratigraphy, minerogenous lacustrine sediments formed since the Allerød (Saarse et al. 1996a; Pirrus 1969), and were succeeded in the Pre-Boreal by organic and calcareous deposits. Varved clays are absent in the lakes which served as a drainage route for the glacial meltwater and were, thus, subject to erosion. The lakes Päidre (Pirrus 1969, Saarse et al. 1995b), Võistre (Saarse 1994), Sinialliku and Viljandi (Lõokene 1979) have been palynologically studied, and only from L. Päidre about 20 radiocarbon dates are available (Saarse et al. 1995b).


Lakes of the Central Estonian Watershed

This region includes the Pandivere Upland and the Saadjärve and Türi drumlin fields. The drumlin lakes in the Saadjärve Drumlin Field are also glacial in origin, resulting from a complicated combination of deposition and erosion. These lakes with uneven longitudinal and transitional profiles are elongated towards the direction of the ice movement and contain a complete stratigraphic succession of Late-glacial and Holocene deposits (Pirrus & Rõuk 1979, Saarse & Kärson 1982, Pirrus et al. 1987). Older Dryas sediments are represented by a 1-3-m-thick complex of yellowish-brown varved clays. Allerød units are dark-grey and greenish-grey silts and silty clays with dispersed organic matter. Younger Dryas laminated silts are coarser, frequently with fine-grained sand and moss interbeds. The boundary between the Late-glacial and Holocene deposits is sharp: the minerogenous deposits are replaced by organic or calcareous ones. In shallow and medium depth lakes, during the Holocene two carbonate-rich units separated by an interlayer of organic gyttja were formed (Fig. 186). The total thickness of these units reaches 6 m in L. Elistvere, 7.5 m in L. Pikkjärv (Saarse & Kärson 1982) and 8.5 m in L. Raigastvere (Pirrus et al. 1987, Saarse et al. 1996a). These lakes were formed during the Otepää and Pandivere stadials of the ice recession. First they were submerged by a local proglacial lake, at the bottom of which varved clays deposited. When the threshold of the proglacial lake was freed of ice, the water level dropped and independent development of the inter-drumlin lakes started. Pollen diagrams are available from lakes Raigastvere (Pirrus et al. 1987), Pikkjärv (Pirrus & Rõuk 1988), Soitsjärv (Pirrus & Rõuk 1979), Elistvere (Pirrus, unpublished), Prossa (Kihno, unpublished), Saadjärv, Kaiu (Zirna & Pirrus 1961). The sequences from lakes Raigastvere, Elistvere and Prossa have been dated by the radiocarbon method (Ilves 1980, Pirrus et al. 1987).

On the Pandivere Upland the Ordovician and Silurian limestones crop out or lie under a thin Quaternary mantle causing the calcite dissolution in cracks, favoured by the zones of tectonic dislocations, now occupied by karst lakes (Raukas 1993). The karst lakes in sink-holes are temporary bodies of water, devoid of lacustrine sediments. The glaciokarst lakes of esker ridges (Photo 57) and kame fields are most widespread near Nõmmküla (Joonuks 1967).

The Äntu group of lakes has been studied in particular detail (Saarse & Liiva 1995). L. Äntu Sinijärv is rare with its highly transparent water and thick lacustrine lime unit which accumulated throughout the Holocene. The thickness of lacustrine calcareous deposits in other lakes is moderate, commonly 2-4 m, in L. Äntu Valgejärv it reaches 5 m; gyttja in L. Neeruti Eesjärv is up to 6 m thick (Saarse 1994).

A special group of lakes are the alkaline dammed valley lakes which existed during the Early Holocene, some of them until the end of the Atlantic. Their sediments consist of Late-glacial clayey units and Holocene lacustrine lime with basal peat in several basins (Kärsa, Lehtse, Tapa, Vatku, Kadrina; Männil 1961), indicating the low lake level status in the Pre-Boreal and at the beginning of the Boreal. The average thickness of lacustrine deposits is 2-3 m, in L. Vatku – up to 7 m.

Because of the prevalence of highly calcareous deposits, few lacustrine sequences of this region are radiocarbon dated (Äntu Sinijärv, Valgjärv, Linaleo; Saarse & Liiva 1995).

Intermediate Estonia (Kõrvemaa) the transitional region between the West-Estonian Lowland and the Pandivere and Sakala uplands, abounds in glaciokarst kettle-hole lakes (Viitna Pikkjärv, Linajärv, Vohja Kõverjärv, etc.). Their formation is closely related to the formation of esker ridges and kame fields (Saarse 1992). Kettle lakes in pitted outwash are rich in organic gyttja (Nikerjärv, Matsimäe, Viitna Pikkjärv, Linajärv). The lake sediments in morainic topography are mostly characterized by calcareous gyttja (Kiruvere; Fig. 187). The lacustrine sedimentation started at least in the Younger Dryas with the formation of sandy-silty beds. In some lakes minerogenous deposition continued in the Pre-Boreal and even in the Boreal (Udriku Suurjärv, Vohnja Kõverjärv; Saarse 1994). In most of the lakes studied in this region, gyttja accumulated since the Pre-Boreal onwards. The thickness of lacustrine deposits varies between 2.5-3 m, reaching 5 m in L. Vohnja Kõverjärv, 8 m in L. Nikerjärv, and 8.7 m in L. Viitna Linajärv. Palynologically studied lakes are Viitna Pikkjärv and Linajärv (Pirrus, unpublished), Kõverjärv, Udriku Suurjärv and Kiruvere (Saarse 1994). Radiocarbon dated deposits come from lakes Viitna Pikkjärv and Linajärv.

The Peipsi and Võrtsjärv lowlands and the Alutaguse area are rich in mires, residual and telmatogenic lakes and kame field lakes. As elsewhere, the kame field lakes are rich in gyttja with a small amount of minerogenous and calcareous compounds. The basal glaciolacustrine and lacustrine clay and silt in kame field lakes are commonly absent. In L. Valgejärv (Illuka Kame Field), basal peat is buried under gyttja (Saarse et al. 1985). The depth and bedding condition of the lacustrine deposits are variable due to the differences in the bottom topography, water chemistry, trophic stage and biological productivity. In the studied lakes, the average thickness of gyttja is 2-4 m (Saarse et al. 1985), maximum 8 m in L. Räätsma. The sequences of lakes Haugjärv, Räätsma, Martsika, Konsu, Liiv-, Lina- and Ümarjärv have been biostratigraphically studied and except the two former, supplemented by radiocarbon dates (Saarse et al. 1985, Koff 1994). The deposition of gyttja started at the beginning of the Pre-Boreal in L. Konsu, and at the end Pre-Boreal in the other studied lakes.

The lakes situated east of L. Peipsi often have bottom peat (Kalli, Lahepera) (Paap et al. 1981, Orru 1995), obviously Pre-Boreal or Boreal in age, unfortunately not yet bio- and chronostratigraphically studied.

Residual coastal lakes in northwestern and western Estonia are few in number on the terraces of the Baltic Ice Lake (Imsi, Kaisma, Järlepa, Järveotsa, Karujärv, Mustjärv), but rather numerous in the modern coastal area. Lakes, left on the terraces of the Baltic Ice Lake, occupy depressions in bedrock hollows or in the gently rolling glacial relief, later modified by the Baltic Ice Lake. These lakes are shallow, but rather large in area. Their independent development began at the end of the Late-glacial, after the area had emerged from the Baltic Ice Lake. The basal glacial and glaciolacustrine deposits are commonly covered by clayey lacustrine units, lacustrine lime and gyttja. Pollen records are available from lakes Järlepa, Järveotsa, Karujärv and Mustjärv (Poska 1994, Saarse 1994).

The number of residual coastal lakes is high in northern and western parts of mainland Estonia and on the islands of the West-Estonian Archipelago. Their formation started after the regression of the Yoldia Sea (Kahala, Rummu, Ülemiste). Those lakes occupy former depressions, but some of them are dammed up by spits, bay-mouth bars and beach ridges, mostly formed during the Ancylus transgression (Ülemiste, Pitkasoo, etc.). The basins of coastal lakes are filled with organic or clayey gyttja. In some lakes (Ülemiste, Käsmu, Ermistu), the lithostratigraphy of the bottom beds is more complicated, comprising also calcareous or minerogenic deposits (Fig. 188). The mineral bottom of lakes Ülemiste, Ermistu, Valge- and Mustjärv is covered by Pre-Boreal peat or peaty gyttja, overlain by lacustrine bed.

Of about 80 lakes on Saaremaa Island, most are young polyhalobous bodies of water with thin or non-existent lacustrine deposits, which isolated from the sea recently due to the land uplift. The oldest lakes on Saaremaa are located on the Western Saaremaa Elevation. According to pollen and 14C records, the accumulation of organic deposits started very soon after the emergence of this area from the sea, on the transition from the Pre-Boreal to the Boreal (Pelisoo, Pitkasoo, Käesla; Saarse & Königsson 1992, Kessel & Raukas 1967). During the Ancylus regression, these basins became shallow and paludified.

The biostratigraphy of lacustrine sequences on the terraces of the Baltic Sea was studied by Kessel (1961a, Kessel & Raukas 1967). New high-resolution pollen diagrams are available from lakes Kahala (Poska, unpublished), Maardu (Veski 1992, 1996b), Ermistu, Tõhela, Kiilaspere, Mustjärv (Veski, unpublished), Pitkasoo, Surusoo and Vedruka (Saarse & Königsson 1992, Veski 1996a, Poska, unpublished) which, besides the Baltic Sea history, provide information on human activities (Poska 1994, Veski & Lang 1996).


Palaeogeographic conclusions

About 10% of Estonia’s small lakes have been geologically studied. This is far from being sufficient for drawing any fully acceptable conclusion on the development of these lakes. Lakes in southern Estonia with almost complete Late-glacial lithostratigraphic record since the Older Dryas, have not been subject to detailed studies. Lakes in central Estonia, particularly in the Vooremaa area, contain almost full Late-glacial sequences which have been studied well. The lakes situated in northern and western Estonia (Ülemiste, Maardu, Kahala, Ermistu, Mustjärv), have been studied in detail, but they have less complete Late-glacial sequences.

Late-glacial biostratigraphic records are rather representative in Raigastvere, Kirikumäe, Päidre and Kahala lake deposits. Complete Holocene records are available from most of the small lakes in Upper Estonia, but also from some lakes in Lower Estonia, e.g. Ermistu, Pitkasoo, Kahala. Several lakes, including Kirikumäe, Tuuljärv on the Haanja Heights, Päidre on the Sakala Upland, Raigastvere in the Saadjärve Drumlin Field, Äntu on the Pandivere Upland, Kahala and Maardu on the North-Estonian Plateau, Ermistu on the West-Estonian Lowland and Karujärv on Saaremaa Island, can serve as Holocene reference sites.

The sediment composition in small lakes is variable and differs with regions. Organic, rather homogenous sediments are characteristic to the kettle lakes in pitted outwash (Kirikumäe, Viitna, Kurtna, Aegviidu), calcareous sediments prevail in the lakes fed by groundwater (Valgjärv, Äntu Sinijärv, Väinjärv), and mixed sediments are typical of lakes with calcareous tilly catchment (Päidre, Võistre, Järveotsa, Ülemiste).

Lacustrine sequences store information on the evolution of lakes, on the vegetation and the Baltic Sea history (in western Estonia), but also on human activities and climatic changes.

Lake level fluctuations serve as an useful tool for the moisture balance control. Lakes respond to changes in the local hydrological balance by changing in depth and area (Saarse & Harrison 1992). The lake level reconstruction of the 13 Estonian lakes (Saarse et al. 1995a) shows that the most pronounced lake level lowering occurred about 9000-8000 and 4000-3000 yr BP. The lakes in Estonia were at the highest level about 9500, 7000, 3000 yr BP and at the present. This indicates drier conditions during the second half of the Pre-Boreal and Boreal and in the middle of the Sub-Boreal, and wetter conditions at the beginning of the Pre-Boreal, in the Early Atlantic, and at the end of it. The long-term trend in climatic change has been explained as a result of changes in the Earth´s orbital parameters (Berger 1978, Berger et al. 1995). In the Pre-Boreal and Boreal, the Northern Hemisphere insolation was about 8% greater than present in summer and 8% less in winter, creating very severe continental conditions and higher evapotransipation during the summer season (COHMAP Members 1988). Latest model simulations have shown that main forcing for the lake level fluctuations is precipitation and evaporation, as well as cloudeness (Harrison et al. 1993).

In western Estonia, the lacustrine sequences have been used to reconstruct the Baltic Sea history (Kessel 1961a) and shore displacement curves (Kessel & Raukas 1979). The Latest studies on coastal lakes (Maardu, Kahala, Ülemiste, Ermistu) suggest that the Ancylus transgression started about 9500-9300 yr BP (Saarse et al. 1990, Raukas & Hyvärinen 1992, Saarse et al. 1995b). Prior to the transgression, there was a low water stand during which peat was formed in several depressions.


Aeolian activity

A. Raukas


In Estonia, where both inland and coastal dunes are encountered, the aeolian redistribution of fine aqueoglacial and beach material was highly controlled by land uplift, palaeoclimatic parameters, including wind direction and activity, soil moisture, but also by the grain-size of initial sediments. Due to the limited supply of sand, concentration of heavy storms to autumn and winter periods and high precipitation rate, dunes are relatively low (mainly 5-15 m) and rendered stationary by vegetation. At present, there is practically no dune sand movement in Estonia.

On several occasions in times past, the aeolian sands posed a great threat to inhabitants and the environment. At the end of the 19th century, a mobile dune on the Island of Saaremaa endangered the Kärla church, pastor’s mansion and farmsteads in the vicinity. The advance of the dune was stopped by a pine stand (Tiismann 1924). The same author describes the movement of sand in the surroundings of the Ristna lighthouse where the strong southwesterly winds picked up a mass of sand and deposited it behind the doors. Every time the islanders had to work days to cart the sand off and clear the access to the houses.

The intensification of agriculture and large-scale land improvement in the 1950s-60s brought about huge fields with a steppe-like appearance and caused deflation on sandy and peat soils. For instance, during April 29 - May 11, 1974, some 16.2 tonnes of dry soil per hectare was carried away by wind action from the Apometsa fields of the former Ranna State Farm in the present-day Harju County. In the same year, during May 1-7, the cloud of dust, blown up from the fields of the Paluküla and Tubala villages on Hiiumaa Island, reached a height of a few tens of metres and deteriorated traffic conditions on the road. A layer of sand, up to 30 cm in thickness, deposited on the road; on the roadside its thickness reached 75 cm (Kees 1992).

Particles of silty soil and dry peat start moving already when the wind speed is as low as 3-4 m/sec. With the wind speed of about 15 m/sec, the amount of soil set into motion generates already a surface dust storm (Photo 58). In Estonia, the wind speed may reach 40 m/sec at a height of 10-20 m above the ground. If the wind blows with a speed of 35 m/sec during several minutes, it may cause great damage to agriculture.

There are some 200,000 hectares of land endangered by deflation in Estonia. On Hiiumaa Island, such fields make up two thirds of the arable lands. When the danger was understood, fields larger than 50-60 hectares on the lowlands and 20-30 hectares on the elevations were prohibited. Belts of trees were planted to shelter fields, while gentle peat soils were planned for a long-term use as grasslands. As a result of the liquidation of state large-scale agriculture in Estonia, the area of fields endangered by deflation has essentially decreased.

In Estonia, continental dunes occur in the Iisaku-Illuka area (Rähni 1959). These parabolic and transverse formations, 0.8-2.7 km long and up to 15-20 m high, indicate a westerly-northwesterly palaeowind direction (Zeeberg 1993). The west and northwest, windward slopes of the dunes are gently sloping (3-18°), while the opposite, leeward slopes are much steeper (18-24°). Small coversand hillocks occur on top and on the slopes of dunes.

Most probably, these coversands, dunes and drift sands originate from the Younger Dryas and the beginning of the Pre-Boreal when a significant regression of Lake Peipsi took place in the Alutaguse Lowland, and the so-called Small Peipsi was formed (Raukas & Rähni 1969). The sand and silt material in the area was formed locally from glaciofluvial and glaciolacustrine deposits reworked by the waves of ancient Lake Peipsi. Dunes were commenced here immediately after source deposits became available, and the process stopped when they became overgrown with vegetation.

The inland dune areas in northeastern Estonia are covered with pine forests and surrounded by bogs. The maximum ages of the dunes may be inferred from the sediments or soils from which they started to develop. These, mostly glaciolacustrine plains and kame fields, formed some 12,200 years BP (Raukas et al. 1971). However, the sand became available for redeposition not until it had drained and dried, possibly after the retreat of the glacier of the Palivere Stadial some 11,000 yr BP.

Unfortunately, this cannot be checked by palynological sampling of dune sands or dating of organic remains. In 1988, attempts were made to solve the problem by means of the TL-method (Raukas et al. 1988) however, all the dates obtained suggested much younger ages, between 4000 - 7100 yr BP.

Coastal dunes occur in Lower Estonia where sandy-silty sediments were available for aeolian processes. The largest dunes, up to 20-25 m in height, formed during the transgressive phases of the Baltic Ice Lake, Ancylus Lake and Litorina Sea (Eltermann & Raukas 1966). Rising sea levels brought about shoreline erosion, destruction of fore-dune and beach ridge vegetation, and initiation of transgressive dunes (Cooper 1958). The most prominent dunes associated with the Baltic Ice Lake transgressive shoreline are located in the Lahemaa National Park and on the Kõpu Peninsula on Hiiumaa Island. The dunes of the Ancylus Lake occur on the West-Saaremaa Elevation, on the Tõstamaa and Kõpu peninsulas and near Häädemeeste. The dunes related to the Litorina Sea transgressive shoreline are found at Sininõmme, Rannametsa and Tõstamaa. The dunes are most numerous on west-facing shores, where the prevailing winds are westerlies and south-westerlies (Fig. 189).

As a result of the uplift of the Earth’s crust, coastal dunes are nowadays situated at some distance from the contemporary shore and at different heights above sea level (Eltermann & Raukas 1966, Martin 1988). At the present seashore only low, a few metres high fore-dune ridges occur, e.g. at Kloogarand, Narva-Jõesuu and Valgerand. The biggest contemporary dunes with specific morphology, termed “basket-trap” dunes by Orviku (1933b), are located on the north coast of Lake Peipsi (Fig. 189). Some small dunes occur also around Lake Võrtsjärv (Tavast et al. 1983b).

Estonian ridge-like coastal dunes, the length of which ranges from 50 to 150m and the width from 20 to 50 m, are elongated in the direction of the prevailing winds (Fig. 190). They usually parallel each other in echelon-like series (Orviku 1933b). The length/width ratio is mostly 2:1 - 4:1 (Fig. 190). The windward slope (5-20°) is often scattered with small, secondary coversand hillocks. The leeward slope is steep (25-40°) and even. Occasionally, e.g. in the Litorina Sea dunes at Rannametsa (Photo 59), both the wind- and leeward slopes are almost equally abrupt (25-45° and 25-30°, respectively). This may be due to the esker buried under aeolian deposits (Raukas 1988) or the later marine erosion of the windward slope during the stages of high water level (Eltermann & Raukas 1966, Martin 1988). In the areas with variable winds, as for example on the Kõpu Peninsula, a clear orientation of dunes is absent or they are oriented in two or three different directions (Eltermann & Raukas 1966).

As a result of uneven migration of dunes, during which the central ridge blowouts moved downwind and the low-lying arms fixed with vegetation lag behind, parabolic dunes came into being. In such dunes the windward slope has an inclination of 10-20° and the leeward slope 20-30°. Both, the ridge of parabolic dunes and the fields made up of such dunes, contain multiple reaping-hook segments. Erosional forms in dune fields are only some metres deep.

During the invasion of the sea, the migration of the deposits up to the coastal slope resulted in their mechanical and mineralogical differentiation which, in its turn, brought about accumulation of well-sorted sand on the shore (Raukas 1966). In Estonian dunes, fine-grained (0.1-0.25 mm) sand prevails. In 25.4 per cent of analyses (Raukas 1968) it formed 80%, and in 9 per cent of analyses even 90 % of the dune material. Medium-grained (0.25-0.5 mm) sand predominates in 18 per cent of the analyses, whereas in about 10 per cent of cases the coastal dunes contain more or less equal amounts of both fine-grained and medium-grained sand, or fine-grained sand and coarse-grained (0.05-0.1 mm) silt. The size of sand grains in dunes depends mainly on the composition of the initial deposits. Compared to the initial deposits, the dune-sands are much better sorted and lower in both coarse and fine fractions (Raukas 1968).

The aeolian sands of Estonia are mostly bimineral and less frequently oligomineral, consisting mainly of quartz and feldspars (Raukas 1968). In northern Estonia and on the islands of the West-Estonian Archipelago, where the initial deposits are rich in carbonates, in some places the dune-sands contain, next to quartz and feldspars, also great amounts of carbonates (at Palivere 11.2 - 26.2 %, in the Püha Andreas Dune Field on the Kõpu Peninsula 2.2 - 13.1 %, at Lümanda 2.0 - 6.8 %). In comparison with the initial deposits, the aeolian sands are somewhat richer in heavy and more or less isometric and weathering-resistant minerals (quartz, zircon, etc.). The content of lamellate and tabular minerals (especially micas and chlorites), and the minerals liable to weathering and wear (carbonates, feldspars, etc.) is lower. The aeolian sands contain more SiO2, and less Al2O3 and K2O than the initial deposits. This fact agrees with the results of mineralogical analyses (Raukas 1968).

According to the interior structure, Estonian dunes may be divided into several types (Raukas 1968). The structure of dunes differs in longitudinal and cross-sections. In cross-sections the sets of lamina are usually more or less symmetrical in relation to both slopes. According to the classification of Botvinkina (1965), a wedge-like inclined stratification predominates. Concavo-convex wave-like sets of lamina are rarer and they mostly occur in longitudinal sections (Fig. 191).

Depending on their closeness to the ancient shorelines, the Estonian coastal dunes are traditionally classified as the Ancylus transgressive shoreline dunes, Litorina dunes, etc., however, they may have been repeatedly reblown and rather young in age. Forest cuttings and fires, military actions (Kroodi, Värska) and other kinds of human activities may trigger the movement of surficial sand. The soil profile, occasionally several metres thick (e.g. Lemmeoja), shows that at least some dunes became overgrown with vegetation immediately after the deposition and never moved again. Wet environmental conditions, sparse population and rapid spread of vegetation prevented extensive redistribution of loose sandy sediments by wind in Estonia.


Genesis and development of mires

M. Ilomets


Like in other lowland countries, the surface structure and climate in Estonia favoured the formation and expansion of mires. Estonia has a rather dense network of rivers and a high proportion of lakes. Per 100 km2, there are about 22.8 km of rivers, about 50 km of streams and channels and 3 lakes with a mean surface area of 1.1 km2. The climate varies from submaritime in the western coastal region to subcontinental in the easternmost region. Average annual precipitation is in the range of 500‑700 mm. Mean temperature in July is 16.5° to 17.5° C, in February -4.0° to -7.5° C.

Peatland is a term applied to any peat covered area ameliorated or untouched, mire is a general term for any peat covered area in virgin state. Mires are divided into two basic types: ombrotrophic (raised) bogs which are totally fed by atmospheric precipitation, and minerotrophic fens with additional feeding by ground and surface waters. Peatlands cover about 1,009,101 ha, i.e. 22.3 % of Estonia’s land surface (Orru 1992).

Estonia belongs to the hummock-ridge bog region. The boundary between the Baltic Coast Bog Province and the East Baltic Bog Province divides its territory into two parts (Botch & Masing 1983). The main phytogeographical boundary crossing Estonia in a SSW‑NNE direction (Laasimer 1965) corresponds well with the geomorphological differences that follow the maximum transgression limits of the Baltic Sea. The western part, the so-called Lower Estonia, was covered by the Baltic Ice Lake and Holocene stages of the Baltic Sea, while in the eastern part or Upper Estonia local periglacial lakes lingered briefly (Fig. 192).

Following the classification by Masing (1975, 1988a), minerotrophic mires are subdivided into soligeneous, topogeneous, limnogenous and transitional. The water in soligeneous or spring fens is commonly calcium-rich, sometimes with a very high Ca content. Schoenus nigricans and Juncus subnodulosus communities are found in such calcium‑rich spring fens in the Island of Saaremaa.

Extremely rich fens are predominantly distributed on the carbonate‑rich substrates of western Estonia. Myrica gale, Schoenus ferrugineus and Cladium mariscus fens occur in the western part of Saaremaa Island and in some places on the west coast of mainland Estonia. Rich, particularly tall (Carex acuta ‑ C. elata ass.) and low sedge fens (Carex nigra ‑ C. panicea ass.) are more common in the eastern part of the country.

Flood‑plain fens are related to the South-Estonian rivers where ground water plays an important role.

Poor fens (transitional or mixotrophic mires) are divided into transitional fens and wooded transitional bogs. Different categories of transitional fens are distinguished. Sedge‑moss fens (Carex lasiocarpa and C. Iasiocarpa ‑ C. rostrata ‑ C. Iimosa ass.) occur mostly on floodplains around lakes in western and central Estonia; elsewhere they are rather rare. Myrica‑Schoenus moss fens with Sphagnum patches on calcium‑rich substrate are spread in western Estonia. Wooded transitional bogs often form a belt around large ombro­trophic bogs, particularly in northern Estonia.

Ombrotrophic mires are divided into moors and bogs. Moor heaths on thin peat and underlying pure sand occur in depressions between dunes in the western coastal area of the mainland and on western Estonian islands (particularly on Hiiumaa). On the basis of the density of tree canopy, two types of bogs are commonly distinguished: wooded bogs and open bogs. The bog margins are usually covered with bog forests. A bog in its early stage of development may be entirely covered with pine forest. Unpatterned, open or wooded bogs dominate in western Estonia, while patterned bogs with strings, hummocks and pools are common in the eastern part of the country.

Regionally, on the basis of differences in the vegetation and mire‑complex types, Estonian bogs are divided between East-Estonian convex and West-Estonian plateau bogs. According to Allikvee & Masing (1988), there are 8 mire districts and several subdistricts in Estonia.


Mire initiation

The terrestrialization of water basins, mainly lakes, as a result of infilling (Photo 60) and the paludification of mineral soils are two alternative ways in mire initialization. In all likelihood, the mire formation is controlled by hydrothermal conditions. So, the terrestrialization should be dominating there where the water table is dropping as a result of decreasing humidity or increasing evapotranspiration. In the case of increasing precipitation and cooling climate the paludification processes ought to have some advantages. Naturally, there are several other factors, such as lithology, geology, neotectonics, giving also an impulse to the mire formation. However, on such a rather uniform and compact territory like Estonia, the factors other than climate and, to a certain extent, also the neotectonics, can be eliminated as having been stable throughout the Holocene. Here we have to consider that the larger the data set in use, the higher the objectiveness of the results.

With this in mind, for the purpose of mire formation studies, stratigraphical sections of the deepest points of the mires, larger than 300 ha, were selected from the peat resources inventarization data of the Peat Group of the Estonian Geological Survey. Of the 467 mires under consideration, 345 are ombrotrophic bogs (Table 48). The distribution of mires between Lower and Upper Estonia is uneven - 35 and 65%, respectively. In both cases, some 25% of mires are limnic in origin.

Peat samples from 35 Estonian mires were dated in the radiocarbon laboratories of the Institute of Zoology and Botany (Tartu) and Institute of Geology (Tallinn). Of the total of 500 datings, about 240, which underlie the present calculation of peat increment values (Ilomets 1995), were obtained on the stratigraphically well-studied cores. As the data are rather unevenly distributed between different peat types, e.g. 65% of samples are characterizing the ombrotrophic Sphagnum dominated peats, the average peat increment values may be fitted for bog peats rather than for minerotrophic fen peats. It is supposed that the calculated mean values may characterize the growth rate of the peat in all 467 mires studied. Following the arrangement and thickness of peat types in the stratigraphical sections, the likely age of every peat deposit was calculated. The calculations yielded an age of ca. 10,000 yr BP for the peat-mineral subsurface contact for several mires in Upper Estonia. The oldest peat sample, collected from the Vaskna fen, gave a radiocarbon age of 9930 ± 70 BP; TA-1600.

The first marked increase in paludification (Fig. 193) in Estonia took place about 8500 yr BP. During the next paludification interval around 7100 - 6100 yr BP, about 13% of mires were formed. This interval of intensive paludification can well be correlated with the transgression phase of the Litorina Sea (7000 - 6800 BP, Kessel & Punning 1984). In Upper Estonia, about 24% of topogeneous mires started to develop about 5100 - 4100 yr BP and ca. 13% between 3500 and 2700 yr BP. In Lower Estonia, there were several periods with not a very clearly defined intensification of paludification processes: 6800 - 6400, 4500 - 3700 and 2700 - 2000 yr BP.

The data indicate that the first increase in terrestrialization (Fig. 194) started some 2000 years later than the first paludification rise up (Fig. 193). Most of limnogeneous mires (67%) in Upper Estonia were formed within a rather limited interval during the second half of the Atlantic Chronozone between 6500 and 4500 yr BP. About one third of topogeneous mires came into being 5400 - 3800 yr BP. In Upper Estonia, this was the most powerful paludification interval during the Holocene. In this part of Estonia, intensive mire formation started evidently some 700 years earlier and completed ca. 500 years before it commenced in Lower Estonia. Although the glacioisostatic uplift in Estonia reached its zero level about 5000 yr BP (Punning 1985), the terrestrialization frequency increased in Lower Estonia at about 5500-5000 and 4500 - 4000 yr BP. It should be pointed out that in the Holocene, there was only one period - around 4500 yr BP, when the processes of paludification and terrestrialization increased simultaneously in Lower and Upper Estonia (Fig. 193). The plant species composition (mostly Bryales and Carex species) of initial communities, as reflected in the lowermost peat layers, indicates that the paludification of forests might have been of no importance, but in most cases the peat formation started in the lagoons during the regressive phases of the Litorina Sea development. Under the conditions of slow water-level decrease, the shallow fresh-water lagoons, which had isolated from the sea, were filled up with peat. However, in some cases, a thin (ca 2 - 5 cm) layer of gyttja accumulated there. Most probably, there was no paludification increase in Lower Estonia and all the mires, which formed during that interval, are limnic in origin.

More than 30% of mires in Upper Estonia started to develop during the time span from 2500 to 6500-4000 yr BP. Of those mires, about 2/3 are of limnic origin. In all likelihood, this is a result of the integrated effect of several factors, favouring mire formation. One of the reasons of continuous infilling of lakes may be related to the stagnation or even decrease of the water level as a result of deterioration of climatic conditions. Digerfeldt (1988) has shown that in southern Sweden the major decrease of water level in lakes occurred 6500-4600 yr BP.

Most of Upper Estonia was freed from the ice in the Pandivere Stadial about 12,000 yr BP (see Ch.VIII.8). It must have taken at least some 5000 years, before a lake became infilled with mud and the process of terrestrialization could start. The longevity of the interval also indicates that if the infilling had already reached a certain level, then even the water-level rise after the stop of glacioisostaic uplift some 5200 yr BP, was incapable of stopping the terrestrialization processes.

Mires reached the ombrotrophic stage of development at well distinguished intervals and synchronously in Upper and Lower Estonia (Fig. 195). The oldest bogs in Estonia date from about 8000 yr BP. However, there are three intervals of intensive mire formation which started 7000, 6000 and 5000 yr BP and lasted for about 500 years each. Quite a small part of our raised bogs was formed during those intervals. It is worth of mentioning that the two first intervals indicate the beginning of bog stage mostly in Upper Estonia, while in Lower Estonia the bog formation intensified only 5000 yr BP. The most intensive bog formation started ca 4000 yr BP and during the succeeding 2000 years (up to 2000 BP) more than 40% of the bogs under consideration came into being. This time span can be divided into two periods: 4000 - 3400 and 3000 - 2000 yr BP. During the first period, bogs developed at equal rates in the two parts of Estonia, while during the second period the process was more intensive in Lower Estonia. The last intensive bog formation interval when more than 30 mires, today larger than 300 ha, reached the raised bog stage showed marked differences between Lower and Upper Estonia. Anyway, the calculations indicate that in Upper Estonia some ten bogs were formed during rather restricted time spans 1500-1200 and 700-500 yr BP, but in Lower Estonia it happened only between 1200 and 800 yr BP.

Although mires may reach the raised bog stage at different times, the temporal dynamics of bog stage initialisation shows the importance of external factors giving impact to these processes. These factors ought to be different from those influencing the temporal dynamics of mire initialisation as the latter process cannot be well correlated with the temporal dynamics of mire origin and as mires are developing simultaneously all over Estonia.

As mentioned above, 40% of the bogs under consideration developed during the course of ca. 2000 years (4000 - 2000 BP). In all likelihood, this phenomenon results, at least partly, from autogenic succession. A mire may reach the ombrotrophic stage after a certain time span, most importantly determined by the combination of geomorphological, water quality (supply of mineral components) and peat accumulation rate peculiarities. The research data indicates that the time span needed is at least 1000 years, commonly 2000 - 2500 years. Therefore, by autogenic reasons the expansion of raised bogs and, consequently Sphagnum could not start earlier than 4000 yr BP, as most of mires developed between 6500 and 4000 yr BP.



In view of the different origin, the development of limnogeneous and topogeneous mires is analysed separately (Ilomets 1992).

The most probable sequences of limnogeneous mires are:

1. In the case of deeper lakes and thicker lake sediments:

Phragmites stand → Betula-Phragmites fen → Betula-Picea-Carex forested fen → Carex open fen → Carex-Sphagnum poor fen → hollow-ridge complex with Eriophorum vaginatum and Sphagnum species → pool-ridge complex with S. fuscum on hummocks

2. In the case of shallow water bodies with a thick layer of lake sediments the two most probable pathways are:

2.1. Phragmites stand → Betula-Phragmites fen → Betula-Picea fen forest → Betula-Picea-Pinus fen forest → Pinus bog forest → hollow-ridge complex with Calluna vulgaris and Sphagnum → pool-ridge complex with S. fuscum on hummocks.

2.2. Bryales quaqing mat → Carex-Bryales quaqing fen → Carex open fen → Carex-Sphagnum poor fen → Sphagnum poor fen → hollow-ridge complex with Eriophorum vaginatum and Sphagnum species → pool-ridge complex with S. fuscum on hummocks.

There is a high probability that in the course of the limnogeneous development all the communities reach the Carex dominated fen stage (Fig. 196) and thereafter the ombrotrophic stage over the mixotrophic Carex/Sphagnum community.

So, it may be concluded that the type of origin may play a rather important role in the developmental pattern of a mire. Most surprisingly, in very many cases this phenomenon may affect the developmental peculiarities in the ombrotrophic bog stage. Thus, for topogeneous mires Sphagnum magellanicum dominated lawn communities are not holding a very important position between the pine forest state and the S. fuscum dominated hummock-ridge state. In the case of limnogeneous mires the same S. magellanicum lawn communities keep a central position between Eriophorum dominated lawn and S. fuscum dominated hummock-ridge communities.

The formation of topogeneous mires as a result of paludification may have started with the formation of at least 15 different peat types. The most probable sequences of peat types are:

1. In the case of paludification of forested areas:

Betula-Pinus herb rich fen forest → Betula-Phragmites forested rich fen → Betula-Picea wooded rich fen → Picea-Betula-Pinus poor fen forest → Pinus bog forest → hummock-ridge complex with Sphagnum magellanicum → hollow (pool)-ridge complex with S. fuscum on hummocks.

2. In the case of open (treeless) depressions, temporarily overflooded:

Bryales mat or Sphagnum carpet → Carex open rich fen → Carex poor fen → Carex-Sphagnum poor fen → Sphagnum poor fen → hollow-ridge complex with Eriophorum vaginatum and Sphagnum species → pool-ridge complex with S. fuscum on hummocks (Fig. 197).

The two sequences are converging into the same point - a patterned Sphagnum dominated bog. The first case emphasizes the paludification of forests and indicates that forested mires can develop up to the ombrotrophic bog stage as mire forests. The second case illustrates the development of open areas and leads to minerotrophic Carex dominated open fen sites. Further development can be characterized by a gradual rise of the role of Sphagnum species.


Soil formation

L. Reintam

Pre-Pleistocene pedogenesis

Although the territory of Estonia became dry land already at the end of the Devonian, there is no direct evidence concerning the formation of soils in the Pleistocene, to say nothing of the more ancient Neogene, Palaeogene and entire Mesozoic era. The soils formed during these times have either been completely destroyed by later geological events or transformed beyond recognition by glacial erosion. Therefore, the reconstruction of soil history must be based on the contemporary analogies, proceeding from the theoretical assumption that the plant, faunal and microbal organic matter as well as its derivates are always and everywhere recognized as the formative power of pedogenesis, and soils are considered as the result of and a pre-requisite for the production processes (Reintam 1978).

In the Late Neogene, coniferous and deciduous forests were spread in Central and North Europe, the animal life was represented by the species of herbivores and carnivores, much the same as today (Raukas 1987). This means that the herbaceous ground vegetation in forests, and perhaps even grasslands in localities of thin forests and/or treeless plains had to vouch for the existence of pasture food chain as well as of energy fluxes and turnover of substances there. There is no information as to the thickness of weathering crust on Ordovician and Silurian calcareous deposits and Devonian sandstones with clayey interlayers. Presumably, it was extremely thin or entirely lacking. It is possible that biological weathering and pedogenesis on limestones and dolomites under the herbaceous forest or grassland vegetation were developed by the way of argillization and humus accumulation analogously to those of Holocene origin.

The contemporary calcareous tills in the North-Estonian Plateau and in central Estonia contain 8.8-11.4% and 11.5-13.1% of clay-size (<0.001 mm) particles, respectively. Fine earth in unchanged tills could not result from the mechanical crushing by the moving glacier. It was probably formed as a result of Pre-Pleistocene weathering and pedogenesis. The same conclusion is possible for the organic carbon, the small amount of which (0.1–0.2%) is always characteristic of unchanged tills and cannot represent the phenomenon of modern pedogenesis. The Pre-Pleistocene soils were destroyed by glaciers. However, the organic matter, clay or any secondary product of chemical and biological weathering, accumulated in these soils, did not disapear but, mixed with pure mineral matter, participated in the formation of tills.

Under the herbaceous sites, on the Ordovician and Silurian limestones as well as on their desintegrated weathering products, there were excellent preconditions for the intensive humus accumulation and rendzina formation. Most likely, the Pre-Pleistocenic Rendzic Leptosols* were characterized by the variable thickness, stony debricity of active section and different productive capacities. It is possible that the progress of argillization of dolomitized limestones and overlying terrigenous sediments resulted in the development of cambic properties and the formation of Pre-Pleistocenic Cambisols. The latter could have occurred besides Rendzic Leptosols, on limestones only there where the weathering rate was as high as in the Holocene and exceeded two grams clay per sq.m per year (Reintam 1975). An analogy to that may be also assumed on Upper-Devonian limestones, though on Ordovician and Silurian limestones the pedogenesis is some millennia older.

Quite a different scenario applies to the Middle-Devonian (Old Red) sandstones during the Pre-Pleistocene. Presumably, the bulk density of their upper strata was smaller and water permeability greater than in the topmost part of the contemporary sandstones underneath the thick cover of Quaternary deposits. Base unsaturation and slight coatings of sesquioxides on the grains of quartz are typical of Devonian sandstones. These chemical relationships, accompanied by the relatively high water permeability and low moisture capacity, had to result in poorer phytocoenoses than those on calcareous deposits. The eluviation of solum, podzolization and formation of Podzols were possibly analogous to some contemporary podzolic (spodic) formations on sands under the stands of Eucalyptus or Agathis, or both, in humid regions of Africa, South-Eastern Asia and Australia (Dudal 1990, Eswaran et al. 1995). Bearing in mind the contemporary analogies, podzolization and formation of humus-illuvial (carbic) podzolic sections in Pre-Pleistocene time could have been sporadic. Sandy Ferralic Cambisols and Arenosols could have occurred under more favourable mineralogical or phyto-coenotical conditions.

It is difficult to suppose that the lessivage and formation of Luvisols took place on the Devonian sands, but their possible occurrence on clays and different clayey products of the weathering of calcareous strata cannot be excluded. Undoubtedly, the phenomena of surface, perched and ground gleyization could have also been involved in the Pre-Pleistocene pedogenesis. In spite of eventual high level of ground water table in limestones and dolomites, their extremely poor capillarity hardly favoured the realization of ground gleyization there. The progress of the latter could have been plausible on Devonian sands which in places contain different ferrous relics up today.


Holocene pedogenesis

In the Late Pleistocene and Holocene, the pedogenesis is related to climatic changes, alteration of forests and the evolution of the Baltic Sea. Although direct data on the temporal and spatial changes in soils during that time have been received in the course of the last decades or centuries, the final product in the kind of ped sections is still assumed to reflect soil’s past and present, memory and moment everywhere (Sokolov & Targul´yan 1976). World experience and information on the regularities in any interaction within different plant-soil systems enable the reconstruction of trends and nature of soil formation and evolution in dependence on climatic, lithologic and phytocoenotic situation (Fig.198). Without the highly informative palaeobotanic evidence, reconstructions of this kind would be impossible.

According to pollen evidence, after the retreat of the last glacier a tundra-like vegetation appeared in Upper (Watershed) Estonia (Thomson 1929, Lippmaa 1935, Laasimer 1965). Proglacial lakes and permafrost tilly landscapes occurred in front of the retreating glacier. Without doubt, the presence of water-saturated fine earth in tills was an immediate prerequisite for the permafrost formation there. It is hardly possible that permafrost could have formed in limestones and in local till in which coarse fraction formed more than 75%. That is why simultaneously with the tundra situation at the time of ice lakes, arctic steppe (meadow) associations expanded on unfrozen tills, poor in coarse fractions, in central and southern Estonia (Grichuk & Grichuk 1950). Textural peculiarities of Late Pleistocene deposits and the contemporary situation in Russian and Canadian permafrost areas (Tedrow 1977, Morozova 1981) suggest that cryoxerophilic meadows on (Gelic?) Leptosols and even on Kastanozem-like formations could have occurred on limestone outcrops (alvars), North-Estonian stony till, drumlins and end moraines with an extremely coarse texture, at least in the core.

The permafrost weathering in combination with surface gleyization became, probably, prevalent under the tundra vegetation, especially in the upper section of the Quaternary stratum which had seasonally been subject to the alternating frozening and thawing processes. Besides the slight textural differentiation, an eluviation of fine-dispersed particles tends to be also characteristic of contemporary Gelic Gleysols in tundra. If to presume that soils on tills in the Late-Pleistocene tundra were much more waterlogged than their contemporary analogies then, in addition to the surface runoff, an intensive internal stream and migration of substances ought to have been developed in Gelic Gleysols there. This had to result in the progress of decalcification and decolmatation (Gerasimov 1960) of solum above the frozen till and translocation of products from Upper to Lower Estonia. The highly similar mineral composition of Upper Estonian tills and varved clays (Pirrus 1968) allows to conclude the formation of the latter from the silt and clay decolmatized from the section of Gelic Gleysols of the Pandivere and Sakala uplands. This, in its turn, gives rise to the presumption that pedogenesis in Late-glacial tundras was connected with the formation and accumulation of fulvic humus, characteristic of eluvial Gelic Gleysols, surface gleyization, weathering of primary minerals and migration of the latter’s ferrous products. As a result of these processes, a gley horizon, light in texture and improverished with bases and sesquioxides, was formed above the permafrost in Gelic Gleysols of the Late Pleistocene.

In all likelihood, such a gley horizon in the forest conditions ensuing the tundra situation, was transformed, in dependence of the calcareousness of tills and further development of moisture relationships, vegetation and pedogenetic details, into the eluvial horizon of contemporary Luvisols, Planosols, Podzoluvisols, or even some Podzolic formations. The bisequal or even multisequal texture of geological-pedogenetic origin, characteristic of a lot of contemporary luvisolic, planosolic and podzoluvisolic sections, has probably also formed as a result of permafrost weathering and lessivage in the Late Pleistocene.

Like in contemporary dry site types, in the conditions of cryoarid (subarctic) steppes, the underground herbaceous biomass had highly to exceed the above-ground biomass. On this bases, a progress of the humus-accumulative process can be expected. Without doubt, the arctic rendzina formation took place there. Thus, Rendzic Leptosols in the Pandivere Upland and on the drumlins, end moraines and Upper-Devonian limestones, have a remarkable absolute age beginning from the time of ice lakes. Like the contemporary arctic pedogenesis, the rendzina formation in the Late Pleistocene took probably place at an extremely slow rate. Therefore, rendzinas in Upper Estonia are thin and skeletal, and do not differ from their younger varieties in Lower Estonia. At that, the formation of Gelic Leptosols and their further evolution over Calcari-Gelic Cambisols up to Calcaric Cambisols cannot be excluded, except for the arctic Rendzic Leptosols. Dwarf Ferric and Ferri-Gelic Podzols with Al-Fe-humus ochric epipedon could have developed on sands of different genesis. Their further evolution into deep Ferri-Carbic and Carbic Podzols tends to be quite natural during the Holocene.

The clear differentiation of pedogenetic processes apparently began under the boreal pine-birch stands depending on the calcareousness, layeral texture, chemical, mineral and moisture relationship of parent deposits. Herbaceous ground vegetation, established by pollen analyses, did not apparently form on sands inhabited already in the Late Pleistocene by green mosses and shrubs. In other sites, the humus-accumulative process, induced by herbaceous plants, was common everywhere. Without doubt, beginning from under the boreal stands the pedogenetic transformation and translocation of mineral particles was differentiated in dependence of the structure, texture and composition of parent strata (Fig. 199) into the sections of clay-accumulative (A–Bm–C), clay-translocative (A–EL–Bt or Bmt–C), stagnic ferric-accumulative (A–(Baf)–ELg–Bt–C), and hydrolytic-eluvial (A–E–B–C and/or A–E–Bh–C) origin in both automorphic and hydromorphic conditions.

The lasting retreat of the glacier and the drainage of proglacial lakes decreased soil overmoistening in general. However, ground gleyization in micro- and mesohollows could have intensified due to the continuously high level of the water table in these hollows. The disappearance of permafrost had led to eluviation of the top of former Gelic Gleysols, and their gleyic formation changed into podzolic and luvic on noncalcareous and calcareous tills, respectively. In several other places, the underlying till, frozen prior to the regression of ice lakes, served after the thawing as a stratum for the stagnation of perched water favouring the progress of stagnic, surface-gleyic and luvic properties. Contemporary occurrence of stagnic, stagni-gleyic and surface-gleyic soils as well as Histosols of atmospheric nutrition tends to demonstrate an intensification of seasonal surface waterlogging against the background of a general decrease in overmoistening regime in Upper Estonia at that time.

Surface gleyization stopped and argillization in situ, described by Bystryakov (1988), initiated on well-drained tills, rich in carbonates and alumosilicates, on the Pandivere and Sakala uplands and on drumlins. Therefore, besides Rendzic Leptosols developed already in cryoarctic conditions, Calcaric Cambisols and Luvisols differentiated under the Pre-Boreal pine and birch stands are associated with calcareous yellowish-grey till in the Pandivere Upland and on drumlins. Calcaric and Stagnic Luvisols, Planosols and Podzoluvisols differentiated on two-layered glacial deposits and/or reddish-brown tills of medium and heavy texture. Some Podzoluvisols and Podzols tend to be associated with coarse permeable sandy and/or loamy sandy noncalcareous tills and glaciofluvial deposits. Figure 198 shows the formation and progress of Holocene pedogenesis through time. The gradual formation of soil sections and genetic-evolutional connections as functions of the moisture conditions as well as the chemism of parent deposits are shown in Figures 199 and 200, respectively.

During the Ancylus Lake, climate became more arid and warm (Raukas 1995c). With the increase in pine in the composition of Ancylian stands the aggressivity of humus substances rose and hydrolytical processes intensified in soil formation. It was probably impossible on calcareous materials, because the supplies of alkaline earths in biological turnover were scarcely less than those in modern cyclings (Kõlli & Reintam 1970). Podzolization intensified on sands and noncalcareous reddish-brown till of light texture. It is not excluded that also seasonal decrease in stagnic and surface gleyic phenomena developed prior to Ancylus time on bisequal deposits of pedogenetic-geological origin. This could result in the local progress of cheluviation there. In the conditions of arid summer characteristic of boreal forests, the hydrolysis of minerals and cheluviation of products was irregular, because at the same time accumulation of weathering and pedogenetic products formed within the warm season in situ favoured similar to contemporary situation (Firsova 1977).

It may be supposed that as a result of the Boreal regression of the Baltic Sea, the importance of ground gleyization decreased, and reoxidation of reduced ferrous, manganese, etc. compounds became evident in deep subsections. Remnant neoformations and concretions in the Bt- and BC-horizons of Stagnic Luvisols, Planosols and Podzoluvisols deeper than one metre as well as ferric-ferrous hydrocarbonates in yellowish-grey calcareous till serve as an indication of the existence of these relics in the conditions under which the contemporary seasonal hydromorphism is lacking.

The warming of climate and consecutive increase in humidity which had begun at the end of the Ancylus period reached the maximum during the Litorina Sea stage. The so-called Atlantic climatic conditions gave rise to the progressive development of broad-leaved herbaceous forests favouring intensive production phenomena, turnover of substances, humus-accumulative process and biological weathering. The development of rendzinas on limestone and stony till was going on simultaneously with the accumulation of secondary weathering products of alumosilicates. Naturally, amorphous sesquioxides prevailed during the first stages. Their step-by-step transformation into crystalline hydroxides was probably similar to the processes characteristic of the contemporary slightly continental subtropics and marine-continental temperate climate (Zonn 1982, 1986). Biogenic argillization, accompanied by the breakdown and leaching of carbonates, and accumulation of nonsiliceous sesquioxides intensified under the herbaceous broad-leaved stands, rich in hazel underwood. The presence of calcium and magnesium in the interlayer structure of clay minerals and in clays, as a whole, can be hardly qualified as a result of modern pedogenesis (Reintam 1971). This seems to be characteristic of pedogenesis on calcareous-alumosilicate tills from the very beginning.

So, it must be emphasized that the culmination of argillization, the progress of ferrolysis simultaneously with the accumulation of nonsiliceous sesquioxides, lessivage and the formation of stagnic properties (pseudopodzolization) became widespread under the broad-leaved forests of the Atlantic Chronozone and are still ongoing. The preconditions for the podzolization were (and are) absent on texturally medium and heavy, and chemically rich deposits. That is why Holocene podzolization is mainly restricted to light and chemically poor deposits under pine stands.

The synchronous progress of Cambisols and Luvisols on calcareous tills was probably induced by local textural peculiarities, formed prior to the Atlantic climate and broad-leaved forest vegetation. The influence of lessivage and ferrolysis on the differentiation of luvic sections cannot be excluded, whereas perhaps due to the multisequal development of ferrolysis described by Brinkman (1979) layeral distribution of amorphous iron took place against the background of its pedogenetic profile accumulation. This had been more considerable on glacial deposits, the bisequal texture of which was formed as a result of Pre-Atlantic pedogenetic-geological processes. Seasonal stagnation of perched water at the juncture of the subsections of different texture was accompanied by the lessivage, ferrolysis, deferritisation and autoaluminisation (Pedro et al. 1974), and resulted in the formation of Stagnic Luvisols and Planosols in the same way as in contemporary humid forests with a high productive capacity, intensive turnover of substances and pedogenetic activity. Yellowish-brown Baf-horizon below the humus one and above the whitish horizon with stagnic and surface gleyic properties was probably formed as early as in the time of the Litorina Sea. It represents the biogenic accumulation of weathering products above the seasonal perched water table in the conditions of variable hydrothermic relationships. Appearantly, part of amorphous iron accumulated has a subsoil (tilly) origin and has been translocated into the topsoil by the mediation of roots and falling litter. Iron hydroxides are quite stable against the reduction and leaching, because modern stagnic and gleyic properties develop under the seasonal perched water with depth, and biogenic accumulation prevents from topsoil degrading.

Surely, pine stands and Podzols did not disappear on different sands within warm and humid Litorina time. Nevertheless, the development of podzolization in depth could have been inhibited by the deficiency of basic constituents able to neutralize the acid products of ground litter humification. That is why the acidoic chelates precipitated immediately in the top of soil section, very close to the bottom of forest floor. The formation of humus-illuvial (carbic) horizons of different morphology took place. They are often present as very thin stripes in modern Podzol sections and barely have contemporary genesis during some last millennia. The presence of oak in Litorina pine stands is acceptable, whereas the mixed oak-pine falling litter was a significant temporary abation against podzolization, especially in deep sandy massifs on valley terraces. Relics of bygonal humus-accumulativity and/or -illuviation can be found there.

Taking also into consideration different stagnic and gleyic expressions of surface- and ground gleyization in lowlands and depressions of glacial and postglacial topography, it is possible to conclude that the contemporary diverse soil cover (Fig. 200) developed in the Atlantic period. At that, against the background of humus-accumulative phenomena, the ecological situation was more favourable for the progress of argillization, lessivage, ferrolysis and surface gleyization simultaneously with the properties of ferritization and/or ferrallitization above the seasonal perched water, but also for the ground-gleyization in the conditions of high ground water table in depressions and, perhaps, for the podzolization in coarse, permable and poor deposits. Plausibly, accumulative and eluvio-accumulative processes developed more and eluvial ones less intensively than prior to and after the Atlantic Chronozone. Evidence is derived from the comparison of any soil type in the transgression area of the Ancylus Lake and in Upper Estonia where the absolute age of the soil cover is some 4,000–5,000 years greater.

As a result of the Baltic Sea regressions, new territories were subjected to pedogenetic phenomena. Although Rendzic Leptosols, Rendzic and Calcaric Gleysols of Ancylus, Litorina and Limnea age are thinner than their older analogies, the similar substantional-regime characteristics as well as profile structure are typical of all the soils of those types. This testifies to the considerable rate of pedogenesis in its early stages and to its decelerating stabilization with time. The presence of crystalline constituents in calcareous till made the formation of Cambisols and their further evolution into Luvisols possible even in the territory of the Ancylus transgression where they are thinner than in the Pandivere Upland or in the drumlin area. It is difficult to suppose that gleyed and gleyic formations could have occurred prior to the modern automorphic Rendzic Leptosols and Cambisols. Nowhere, all of a sudden and total appearance of vegetation accompanied by synchronous soil formation was incredible, because water regression from morainic hillocks had been temporarily distanced from the process in concave forms of undulating morainic topography.

Alternation of different Gleysols in morainic hollows covered by Ancylus sediments, gleyic formations in transitional areas, and Rendzic Leptosols, Cambisols and/or Luvisols on micro- and mezohillocks is characteristic of current natural situation and tends to demonstrate the phytocoenotic and pedogenetic mosaicity already within the Ancylus and Litorina stages. At that, the progress of argillization in situ can be partly connected with thin Holocene sediments on tilly underground and with the weathering phenomena of the latter in the alternating hydrothermic conditions. As stony till, drying analogeously to the modern situation deeply up, did not favour the argillization on local watersheds, the total occurrence of littoral Fluvisols in the entire area of the Ancylus transgression tends to be scarcely supposable. Fluvisols of limnic nature could have occurred only on the concave forms of relief near the regressive and sometimes once again offensive lake.

In connection with the climatic re-cooling and penetration of spruce into the composition of stands about 4,000–5,000 years ago, spruce-decidous forests became prevalent on tills. Pine stands preserved on deep sands and in droughty limestone and stony till areas. Although the acidity of forest humus could have increased under the impact of spruce falling litter, intensification of hydrolytic-eluvial processes was hindered by calcareousness and richness in bases of parent materials and former soil strata as well as by the quite high ashness, characteristic of spruce litter up to now. Naturally as usual, decidous, especially hazel underwood and widespread herbaceous ground vegetation was of a great importance in the prevention of any destructural-eluvial pedogenetic phenomenon. On the basis of these regularities, it is possible to conclude that in the Sub-Boreal and Sub-Atlantic pedogenesis continued following the same trends and directions as before, whereas the mineral and chemical composition (potential) of parent tills and Holocene sediments, exerted as usual, an impact on the phytocoenotic processes and affected interactions between organic matter and mineral strata. Probably, due to these interdependences, the stabilization of plant-soil systems on different parent deposits was reached which, in its turn, enabled the vicinity existence of various soil combinations (Rendzic Leptosols, Cambisols and Luvisols in central Estonia on yellowish-grey calcareous till; Luvisols, Planosols, Podzoluvisols and even Podzols in southern Estonia on reddish-brown till) within a short distance.

As a result of climatic cooling, the mobility of humus substances could have grown and an increase in the migration of organic-mineral complexes as well as in the compaction of illuvial horizons became feasable there where these phenomena had not previously taken place. Of course, the stagnation of perched water and the progress of stagnic and surface-gleyic properties intensified against the background of former lessivage and/or podzolization. Therefore, Planosols and Stagnic Luvisols of bisequal texture have a great territorial importance, whereas many-featured morphology and regional variable properties are characteristic of them (Kokk et al. 1985).

During the last two or three millennia, the pedogenesis and soil properties have been considerably affected by human activities (Photo 61). The development of burnt-over tillage brought about the same phenomena found in Karelian and Vepsian materials (Reintam & Moora 1983): a decrease in acidity and humus stability, an increase in humus fulvicity, exchangeable bases and base saturation. Clear-cutting of forests, disappearance of forest floor and the opening of soil surface to the weather conditions resulted in the intensification of soil leaching and lessivage from the beginning of human agricultural impacts. Simultaneous use of harvesting residues, manure, composts, crop rotation, etc., gave a rise to the man-generated accumulation of complementary organic-mineral complexes into the humus horizon, accompanied by the gradual growth of the latter in depth. Although the amounts of organic matter participating in soil processes have decreased and the turnover of substances in the plant–soil system has become ever opened in arable soils, there is no reason to suppose principal changes in pedogenetic trends and directions as a result of human agricultural activity. In principle, all the Holocene pedogenetic processes and formations are characteristic of both forest and arable lands. Highly different Anthrosols are special and rare.

As a successive drying up of the thin-rooted and weather-opened topsoil is always accompanied by the formation of water-permeable cracks and cleveages in fields, the intensification of lessivage tends to be typical of arable soils from the beginning of permanent agriculture. These phenomena are prevented by the ground litter in the forest ecosystems. In its turn, lessivage leads to the formation of textural argillic horizon instead of cambic one, and stagnation of seasonal perched water in subsoil. That is why also a certain anthropogenic intensification of ferrolysis (pseudopodzolization), and progress of Stagnic Luvisols and Planosols have been found as a result of tillage and cultivation. Advantageous translocation of clay in arable soil conditions enables to explain the light texture of the humus horizon of some Rendzic Leptosols and Cambisols without the eluvial differentiation of their entire sections.

The differentiation of current gleyic soil sections and the character of pedogenetic processes inducing gleyization tend to be universal in different moisture relationships. Without doubt, not only ferrolysis and lessivage, but also any eluvial phenomenon, in dependence on parent material, were favoured by the gleyization within the entire Holocene. As these conditions were less conductive to the argillization and ferrallitization, cambic, ferralic and/or argic properties are weak in the sections of some Gleyic Cambisols, Gleyic Luvisols and Cambi-Calcaric Gleysols. A great number of Eutric Histosols have been evolutionated from Gleysols through their Histic subtypes.


Current soils

Automorphic Leptosols, Cambisols, Luvisols, Planosols, Podzoluvisols, Podzols and Arenosols with their gleyic subtypes and eroded kinds cover 42.0%, different Gleysols 32.5%, and Histosols (mires and bogs) 23.2% of Estonia’s territory (Table 49). Fluvisols (alluvial and saline litoral soils) form 2.1% and technogenic rendzic formations on reclaimed land 0.2%. By texture, sandy soils make up 26.7%, peaty soils - 23.7%; loamy sands, loams and clays occupy 17.0, 27.8 and 4.8%, respectively. Of the latter, 50.1, 38.2 and 11.8% are arable, correspondingly (Reintam 1995, Kokk 1995). Nearly 50% of the underground consists of Ordovician and Silurian calcareous sediments, 75% of soil parent material is calcareous.

Rendzic Leptosols (Rendzinas) are formed on limestones and dolomitized limestones, on strongly calcareous stony till and on coarse glaciofluvial materials. They represent an original soil formation having been developed on stony calcareous deposits within any historical stage of pedogenesis. As Rendzic Leptosols occurred already in the Pre-Pleistocene, their (although temporarily interrupted) dissemination is steady and typical of northern and central Estonia. Calcareousness higher than 10% already in the humus horizon, short skeletal non- or slightly differentiated profile, calciphilous vegetation and relative richness in humus have always been characteristic of Rendzic Leptosols.

Cambisols on calcareous yellowish-grey or reddish-brown tills are characterized by the argillization in situ of the entire solum or Bm-horizon under the humus one, relative accumulation of nonsiliceous hydrates of sesquioxides, favourable hydrothermal, redox and biological relationships. Their independent progress is related to the Holocene, because eventual Pre-Pleistocenic Cambisols were destroyed by glacier.

Luvisols on calcareous tills have much the same early genesis as Cambisols, but they are characterized by morphologically-distinguished lessivage (translocation of clay). A–EL–Bt (Bmt) section (solum) is also rich in nonsiliceous sesquioxides, biologically active and like that of Cambisols rich both in humus and available moisture. In spite of territorial closeness, the pedogenetic differentiation of Luvisols from Cambisols commenced under the boreal herbaceous forests is still ongoing.

Stagnic Luvisols (brown pseudopodzolic soils) are formed on bisequal glacial deposits of pedogenetic-geological origin under the seasonal stagnation of perched water at the juncture of layers of different texture. These soils are characterized by an accumulative Baf-horizon, rich in amorphous iron, between humus (A) and stagnic eluvial (Elg) horizons. The progress of Stagnic Luvisols started as early as the Boreal against the background of Gelic Gleysols after the regression of tundra and permafrost. It culminated under the broad-leaved forest vegetation of Litorina age.

Planosols (light pseudopodzolic soils) on tills of loamy and clayey texture are characterized by the presence of a white-coloured stagnic horizon, rich in amorphous iron and ferric concretions, immediately below the humus horizon. Significant textural argillization is characteristic of these soils, however, some surface gleyization can be also found. Genetically, they are similar to Stagnic Luvisols with which they have progressed synchronously during the Holocene.

Podzoluvisols on bisequal glacial, glaciofluvial, aeolian, etc. sediments above glacial ones have both podzic and stagnic properties in their multisequal section. Textural contrasts between layers make possible the synchronous development of a great number of pedogenetic processes. Prior to the distinguishing of Stagnic Luvisols in FAO-ISRIC system, the latter were interpreted as Podzoluvisols, because there are some similarities in their genesis and profile structure.

Podzols on noncalcareous reddish-brown loamy sandy till have a humus horizon in their top. This is usually lacking or rarely present on sands of different nature. Flowing regime, good filtration, low-intensive ash-deficient turnover of substances and agressive mobile humus are characteristic of Podzols. Humus-illuvial (Carbic, Carbi-Ferric) subsection is widespread in sandy Podzols. Podzolization and Podzols have always been connected with sands not only within the Holocene, but also in the Pre-Pleistocene. Therefore, Podzols (as Rendzinas) have occurred in any stage of climate and vegetation.

Arenosols on coarse glaciofluvial sands have a slightly ferritized cambic and weakly differentiated profile. Their formation tends to be connected with moisture deficiency and arid herbaceous forest vegetation since the ice-lake stage. Since they resemble both weak Cambic Podzols and Ferralic Cambisols, they are not yet distinguished as a distinct entity in soil maps.

Gleyic subtypes of any automorphic soil type are characterized by some raw humus in A–horizon, stagnic properties in topsoil and gleyic properties in subsoil. As an exception serve Stagnic Luvisols, Planosols, Podzoluvisols and loamy sandy Podzols, in which ground-gleyization is lacking, and surface gleyization in the kind of bluish marble and iron neoformations occurs in ELg–Btg (Bg, BCg) subsection. The importance of gleyic phenomena has been changing in time and space during the Holocene up to now.

Gley-Podzols (Podzolic Gleysols) are formed on noncalcareous sandy deposits. Carbic and/or Carbi-Ferric nature is mostly characteristic of them. Their Holocene history and genesis tend to begin from the time of ice lakes. Later they developed and differentiated under pine stands.

Gleysols have been the prevailing soil formation since the permafrost tundra stage and they include different hydromorphic types of ground and/or combined surface-perched-ground aquic nutrition. Rendzic Gleysols, Cambi-Calcaric and Luvi-Calcaric Gleysols develop on calcareous till, but most of Eutric and Dystric Gleysols on graded deposits of the Baltic Sea transgressions. On different water-laid sediments, the base-saturated Gleysols on flat territories or on microlowlands often alternate within a short distance with the acid Podzol-Gley formations on microheights. The properties of Gleysols highly depend on the water nutrition as well as on the chemism of water and parent material. Together with Rendzic Leptosols and sandy Podzols, Gleysols represent the third formation which is spread and developed within any stage of phytocoenotic and pedogenetic activities. At that, Histic Gleysols have always occupied the transitional territory from Gleysols to Histosols.

Histosols (mires and bogs) have a peat (Histic) horizon with a depth of more than 30 cm. They have developed from Gleysols within the Holocene, but also as a result of the eutrophication of waterbodies. Eutric Histosols (lowland mires) are characterized by a ground-water or flooding regime, Dystric Histosols (transitional mires and raised bogs) – by atmospheric nutrition.

Fluvisols are the current formation of seasonal inundation and accumulation of alluvial, lacustrine, etc. suspensions. They have been steadily formed within the Holocene and even Pre-Pleistocene, prior to any following geological event arose a rechange in pedogenetic phenomena.





Prehistoric times

A. Poska & L. Saarse


Interpretation of the standard pollen diagrams in terms of human impact during Mesolithic and Early Neolithic times when man’s economy was mainly based on hunting, fishing and gathering, is a complicated task. The development of woodlands during that time was controlled by climatic, edaphic and other ecological factors rather than by human influence (Poska & Königsson 1996). Nevertheless, the Stone Age people also utilized the environment around habitation sites, causing disturbances of local importance and introducing in this way more favourable conditions for the flourishing and spread of light-demanding and nitrophilous herb species (Behre 1988, Poska 1994, Veski 1992, 1996a, b).

The first traces of primitive farming in Estonia date from Neolithic time. Cattle rearing and crop cultivation were properly introduced during the Bronze Age (Lõugas 1992). With more advanced agriculture, man´s influence upon the environment greatly increased.

The systematic research on human impact in Estonia was undertaken by palaeoecologists and archaeologists in the 1990s when coastal areas of Saaremaa Island (Jõhvikasoo), northern (Maardu, Kahala, Tondi) and northeastern Estonia (Kunda, Narva) were studied in the frame of the PACT Project (Hackens et al. 1996). As the coastal area is supposed to be a cradle of Estonian farming, the above-mentioned regions were chosen to describe the development of the early agriculture. During the mentioned project, fossil field remains were discovered and dated at Maardu and Saha-Loo (Lang 1996).

The following is an attempt to give an overview of the data available on the history of the colonization and land-use based on the biostratigraphical evidence in Estonia. The discussed material covers almost the whole Estonian territory with different landscape regions and a time span from the beginning of the Mesolithic until historical times.

More than 30 pollen diagrams were examined to find indications of human impact on the environment; 23 diagrams were selected to treat this problem (Fig. 201). The chosen localities, except some earlier-mentioned sites, were studied mainly for stratigraphical purposes. Most of the used sequences were dated. The time scale of five undated sequences (Kunda, Vedruka, Tõhela, Surusoo and Pitkasoo) is based on the correlation with the sites in the closest surroundings. All radiocarbon dates are uncalibrated.

Biostratigraphical data has been summarized in terms of human impact using the method suggested by Behre (1981). Based mainly on the works of Behre (1988) and Hicks (1990), the following plant groups were separated:

1. Cultivated land (Cerealia, Secale, Triticum, Triticum spelta t., Hordeum, Avena, Cannabaceae, Cannabis t., Centaurea cyanus, Fagopyrum, Polygonum convolvulus)

2. Fresh meadows (Ranunculaceae, Ranunculus acris t., Plantago sp., Plantago lanceolata, P. m/m t., P. maritima, Potentilla, Cerastium, Achillea t., Saussurea t., Solidago t., Cirsium t., Caryophyllaceae, Trifolium, Linum catharticum, Rhinanthus t., Centaurea scabiosa, C. jacea, Scrophulariaceae, Helianthemum, Succisa t., Alcemilla t., Hypericum t., Gypsophila muralis, Vicia cracca, Valeriana)

3. Ruderal communities (Artemisia, Chenopodiaceae, Chenopodium album, Brassicaceae, Urtica, Plantago m/m t., Onagraceae, Chamaenerion, Polygonaceae, Polygonum aviculare, P. persicaria, Rumex ac/ac t., R. acetosella, Centaurea sp., C. nigra, Malvaceae)

4. Grazed forest (Pteridium, Melampyrum)

5. Dry pastures (Juniperus, Cerastium t., Campanula).

Conclusions considering forest disturbances are based on the abrupt changes in tree pollen curves, finds of light-demanding NAP taxa, increase in charcoal dust amount and a general rise in anthropogenic indicator graphs. A fluctuating character of these indications is considered to show different phases of settlements: establishment, flourishing and abandonment. Appearance of Cerealia pollen together with other evidence is supposed to characterize the cultivation practice.

The first traces of human habitation in Estonia come from the Mesolithic (9500-6000 yr BP, Kriiska 1996a). Settlements of this period were commonly situated along the water-edges, in ecothons where the plant communities are sensitive to disturbances and even a relatively weak stress may yield pronounced and long-lasting consequences. The nomadic character of the Mesolithic people, the successive utilization of the different biotopes (coastal and inland sites) after seasonal food availability may have enlarged the influenced area considerably. In Estonia, the oldest known settlement site, dated to 9600±120 (TA-245) and 9575±115 (TA-176; Kessel & Punning 1969) was situated near the present-day Pulli Village on the Pärnu River bank (southwestern Estonia). Detailed palynological studies in the surroundings of the settlement site have not yet been performed. The Kunda Settlement (northern Estonia), which appeared somewhat later and gave a name to the Mesolithic Kunda Culture of Estonia, has been better studied (Photo 62).

The finds of Kunda Culture are known from different parts of Estonia. Based on biostratigraphical material, the net remains found by Indreko (1932) from the former sea bottom at Siiversti in the Narva area are referred to the Boreal/Atlantic limit (Jaanits et al. 1982), being the oldest of this kind in northern Europe (Kriiska 1996c).

Particularly abundant Kunda Culture finds come from the Võrtsjärv Lowland. The Siimusaare, Jälevere, Tamme, Lepakose, Umbusi and Moksi dwelling-places have been examined by archaeologists (Jaanits et al. 1982, Jaanits 1992). The biostratigraphical material is scattered or absent in the direct vicinity of the above-mentioned settlement sites. Still, some evidence of landscape disturbance in the area during Mesolithic time may be found on the available pollen diagrams (Ilves & Sarv 1970).

In eastern and southeastern Estonia, the Neolithic dwelling-places at Akali and Kullamäe are well known at the mouth of the Emajõgi River (Jaanits 1959, Moora et al. 1988). The bio- and chronostratigraphical data refer to the existence of the habitation in the Akali area earlier than recorded by archaeologists.

During the last decades, new Late Mesolithic settlement sites have been excavated on the islands of Saaremaa (Võhma, Kõnnu; Lõugas 1982) and Hiiumaa (Kõpu; Lõugas 1988, Kriiska 1996b). The last biostratigraphic studies in the vicinity of the ancient settlement sites on Saaremaa: Pelisoo, Surusoo, Jõhvikasoo (near Võhma); Pitkasoo (Naakamäe) and Vedruka (Loona, Kurevere) have yielded supplementary information concerning human activities during Late Mesolithic and Neolithic times (Saarse & Königsson 1992, Poska 1994, Hansson et al. 1996, Veski 1996a). New palynological evidence from the region points at the presence of Mesolithic settlers not only in the Võhma, but also in the Käesla (Naakamäe?) and Kihelkonna (Loona?, Kurevere?) areas. Pollen evidence from Kõivasoo Mire (Hiiumaa) shows forest disturbance about 6000 yr BP which is consistent with the latest data from archaeological excavations in the Kõpu area (Kriiska 1996b).

Up to the end of the Stone Age, the economy of man was mainly based on hunting, fishing and gathering. In the Neolithic, indications of human impact traceable with biostratigraphic methods, were much the same as during Mesolithic time. Nevertheless, primitive crop farming techniques, such as slash-and-burn agriculture, were introduced to Estonia already during the Neolithic, and settlements started to spread all over the country (Jaanits 1992). The expansion of the farming activities led to the introduction of some new plant species (both cultivated and accompanying ones), and the need for cleared land increased.

During the Neolithic, the first more extended indications of the forest clearance appear on the pollen diagrams between 5000-4500 yr BP. The earliest finds of cereals in the Maardu area are dated to about 3800 yr BP (Veski 1992), in the Saadjärve Drumlin Field to 4495±35 yr BP (Kõrenduse; Pirrus & Rõuk 1988) and ca. 4100 yr BP (Velise; Veski, unpublished data), being older than in northern Estonia. Such distribution of the first cereal findings indicate that the tillage could have been introduced to Estonia from the south and southeast direction. The crop farming spread then all over the country, but gained importance in man’s economy only in areas with favourable natural conditions (soils, water, etc.).

In the Bronze Age, farming gained importance, overruled fishing and hunting, and became basis of man’s economy. The Late Bronze farming revolution (Veski & Lang 1996) led to the drastic changes in land use and the spread of cultivation from easily tilled alvar soils to the other land. Due to the evolution of agrarian techniques, the habitation moved from coast to inland. The main plant species cultivated during the Bronze Age were: Triticum dicoccum, T. monococcum, Hordeum vulgare, Linum usitatissium, Pisum sativum, Lens esculenta (Behre 1988). In contrast to the Stone Age forest clearances, after which a regeneration of forest normally occurred, the clearances in the Bronze Age often brought about permanent changes in ecosystems. The formation of the majority of Estonian alvars probably started during the Bronze Age.

The Iron Age, as a whole, is characterized by well established quickly spreading and developing farming. New agrarian techniques enabled man to extend the size and amount of the fields and pastures and explore less favourable parts of Estonia. The final revolution in land occupation started in the Late Iron Age about 1100 yr BP.

Regardless of relatively high density of sites investigated in terms of pollen analysis on Estonian territory during the last decades, the quality of the determinations and the availability of datings only in a few sequences corresponds to the level required for investigations of prehistoric human impact. Still, on the basis of the discussed data some outstanding regions and periods in the prehistoric land use are proposed:

1. The biostratigraphical records show the same Stone Age habitation centres in Estonia as proposed by archaeologists (Jaanits et al. 1982, Kriiska 1996a, b):

a) North-Estonian Coast,

b) Võrtsjärv Lowland,

c) Eastern part of the Peipsi Lowland (lower course of the

 Emajõgi River),

d) Islands of Saaremaa and Hiiumaa.

Palynological indications of the human influence upon the vegetation in southeastern Estonia do not coincide with the archaeological records. As the data available are rather episodic, further investigations are needed.

2. Based on the available biostratigraphical evidence, the following regional phases of the human impact may be distinguished:

a) At 9000 yr BP, the start of a period with indications of woodland disturbances in the coastal area of the Baltic Sea, on lake shores and in river valleys of northern and northeastern Estonia.

b) At 7000-6500 yr BP, the second period of forest disturbances occurred in northern Estonia and on the Võrtsjärv Lowland, at ca. 6500-6000 yr BP on the islands of Saaremaa and Hiiumaa, and in eastern Estonia.

c) At 5000-4500 yr BP, the relatively well defined period of increasing human impact is traceable almost all over Estonia. The first Cerealia pollen finds are recorded in eastern Estonia (Kõrenduse).

d) At 3800 yr BP, pollen spectra of several diagrams indicate increasing forest clearance, and in some cases, also a primitive crop farming (Cerealia and Plantago lanceolata pollen in L. Maardu; Veski 1996b).

5) At 3300-3200 yr BP, indications of extensive forest clearances and crop farming in northern Estonia, on the Võrtsjärv Lowland and on the Island of Saaremaa.

6) The first finds of the Secale cereale pollen are known already from 2500 yr BP (Ilves & Mäemets 1987), but obviously it was then growing as a weed in crop fields.

The real introduction of Secale cereale is likely to have happened at about 1400-1500 yr BP.


Historic times

L. Saarse


Population. At the beginning of the Middle Ages the population increased, favoured by climate and social relations. The inhabitants settled all over Estonia, even the Alutaguse and Kõrvemaa areas were sparsely populated. A historical document dating from 1228, names Hiiumaa a “desert” island, but it seems not to have been the truth (Tarvel 1992). Estimates place the population of Estonia in the 13th century at about 100,000-150,000 people (Tarvel 1992). The main settlement type was village, in southeastern Estonia scattered habitation prevailed. The Liber Census Daniae (Eisen 1920, Johansen 1933) contains data about Estonian villages, showing that the 13th century Estonia was a typical agrarian country with villages and fenced strongholds, into which manors started to integrate.

During the course of the next three (14th-16th) centuries, the population of Estonia doubled. According to Vasar, some 250,000-280,000 people lived in rural districts. Together with citizens, the population of Estonia could not have been more than 300,000 at the end of the 16th century (Tarvel 1992). However, since there is very little demographic data about the ancient inhabitants of Estonia, this number is very approximate. The main fields of economy were crop farming and cattle rearing, about one fourth of the yield was exported to Finland, Sweden, the Netherlands, Germany and even to Portugal.

Population dynamics and economy were fundamentally affected by wars, epidemics and famines, causing the death of thousands of people. After the Russian-Livonian War (1558-83), 50-70% of farms were deserted or abandoned. After the Polish-Swedish War (1600-29), about 87% of farms in the present-day Tartu County stayed desolated (Tarvel 1992). The fighting and outbreaks of epidemics and famine ravaged the country and reduced the Estonian population nearly by two-thirds (from 250,000-280,000 inhabitants before the Livonian War to about 70,000-100,000 in 1625). The period following those hard times witnessed a rapid economic development; deserted farms were taken into use and the yield export expanded once again. The number of population increased and before the Great Famine there were about 375,00-400,000 inhabitants in Estonia (Palli 1992).

The regional mapping in 1684 shows that the distribution of farms and rural settlements was much the same as in recent times (Troska 1992). After the Great Famine (1695-97), the Northern War (1700-21) and the plague that ravaged Estonia in 1710, the survivors numbered 150,000-170,000 in 1712. The consequences of the epidemic were most horrible in western Estonia.

In the 18th century the rural population increased. In the middle of the century, Estonia had 330,000 - 340,000 inhabitants, of those only 5% lived in towns (Palli 1992). In 1797, there were 17,212 farms in the Estonian Province (North Estonia). In the beginning of the 19th century, the number of farms in Estonia was at its highest - 18,879 (Troska 1992); thereafter it started to decrease. In the middle of the 19th century, the size of the population had doubled, reaching about 750,000 inhabitants, 14% of whom lived in towns. The population was concentrated to villages and, thus, Estonia still stayed a typical agricultural country. Since the 1950s, the urban population rapidly increased, forming 71.4% of the total population in 1990 and showing a drastic reorientation in the Estonian economy (National....1992). The present-day Estonia is a highly urbanised industrial country.

Landuse. Feudal relations, foreign conquest and division of Estonia into several parts between different powers were the main reasons, why agriculture flourishing at the end of the 12th century, made a slow progress in the Middle Ages. Unfortunately, the written sources do not give the full clarity of the medieval tools and tillage. For this reason, it is not known when exactly the manuring of fields started. It was first mentioned in historical documents in the 15th century, which does not mean that the manure had not been used earlier. The main tools were wooden plough, harrow, hoe, sickle and scythe. As to the cereals, most important was Secale cereale (rye) cultivation, followed by Hordeum (barley) and Avena (oat). The share of Triticum (wheat) was low, being more widely grown in western and southern Estonia. In Livonia (South Estonia), Linum (flax) was also cultivated and exported (Kivimäe 1992). From horticultures cabbage, carrot, onion, parsley and hop were grown. Besides crop farming, animal husbandry was also a very important branch of economy and a booty for the conquerors (Henriku...1993).

After the Polish-Swedish War (1600-29), the economic development in Estonia gained in momentum following the same trend as in the succeeding centuries. This was promoted by improved tillage and manuring the fields. Three-field-crop- rotation system which dominated up to the 17th-18th centuries was accomplished, slash-and-burn agriculture lost its importance in northern and western Estonia, but remained in use in southern Estonia (Livonia). In northern Estonia, the clearing of stony fields was widely practiced. The maps of the 17th century show stone edges and cairns.

The first data on draining of land by ditches in Estonia comes from the 14th century in the Maardu area, but these were only single attempts. It was not until the 17th century that large-scale amelioration was undertaken in the estate fields of northern Estonia and on the Island of Saaremaa.

In the late 19th century, a large number of tenants began to buy farms in perpetuity. This brought about a new rapid development of agriculture, particularly crop farming and animal husbandry. The production of timber made a rapid progress, the area under arable land was expanded and the forested area as a whole, decreased to ca. 20% by the beginning of the 20th century. The situation turned critical and called for afforestation. Forest management has a long tradition in Estonia. The first orders prohibiting felling of forest were put into force on the islands in 1297, the status of forest keepers was introduced in 1696, regulations for forest management in 1782 and forest plantations have been cultivated since the last century (National... 1992). Currently, about 48% of Estonia’s land area is taken up by forests, the main forest-forming species being pine, birch and spruce. This means that during the last 50 years the forested area has increased 2.2 times, mainly on the account of the overgrowing of abandoned fields and meadows with brushwood. Land drainage, air pollution, intensive felling, etc., have affected the spread of swampy black alder and boreo-nemoral hardwood-spruce forests which are currently rare or on the verge of extintiction. In recent years, the first signs of damage caused by acid rains and new species of forest pests have begun to manifest themselves (National... 1992).

In this century, the Estonian agriculture has passed three revolutionary changes: (1) 1918-19, implementation of the land reform which touched mostly manors and dispossessed big landowners of land, (2) large-scale collectivization in 1949, and (3) privatization and establishment of small private farms in the 1990s. The agrarian structure, which has a long historical background, being developed in the frame of private farmsteads and estates, was completely destroyed during the Soviet period. Small patches of arable land, tracking landscape peculiarities, were united into huge fields. One third of the former fields were left uncultivated and almost as large area of natural grassland was cultivated into fields. In the Haanja and Otepää heights and on the Sõrve Peninsula about 50% of previous fields were abandoned because, being too small and scattered, they were not profitable for cultivation with heavy machinery. The use of heavy machines resulted in the compaction and degradation of arable soils, affecting their porosity and moisture content. Large fields suffer from wind erosion, the fields on the hummocky terrain are subject to surface water erosion. But there were also positive aspects of the collective farming: large fields were cleared up of stones, fields were regularily limed and fertilized, and almost 66% of fields were subject to amelioration. However, today the out-of-date drainage system needs reconstruction to avoid paludification of drained fields.

During the Soviet power, the grasslands, meadows and natural or seminatural pastures which in 1939 formed 24.5% of Estonia’s area (National... 1992) were partly cultivated into grassland, partly afforested and partly became overgrown with brushwood.

Ecological problems. The development of the landscape and environment has been mainly controlled by two factors: natural (climate, geological-geomorphological-hydrological conditions, soil texture, etc.) and social (human activities). We have reached the state where special expenditures are needed for finding out major agents affecting soil fertility and production level, for controlling the exploitation of national resources along rational complex line and for ensuring that man´s living conditions will not deteriorate any more.

In the 1950s, Estonia became a region of very intensive agriculture and industry. This caused damage to soils, inland and sea water, air and to the environment, as a whole. The soils and water were subject to pollution by the wastes of urban and rural settlements, industry, big farms and military camps. From some rather big urban settlements, like Tartu, where sewage treatment plants are lacking, the unpurified waste water is discharged directly into the rivers. The main pollution load comes from northeastern Estonia which is the region of mines, electric power stations and chemical plants. Self-ignition in oil-shale waste hills, harmful chemicals and minerals, phenols and oil have contaminated soils and rivers and reached the sea. The water of the Gulf of Finland contains nutrients, oil products, phenols, heavy metals, toxic elements, the concentration of which is rather high near Tallinn, Sillamäe and Narva.The residual of petrol and rocket fuel from military airfields, camps and depositories has polluted the soil and entered the groundwater (at Tapa, Kärdla, Aruküla). Excessive use of pesticides and herbicides in agriculture and set up of big rural settlement without proper utilization of domestic wastes, has enlarged the pollution danger to the country.

The situation is especially critical in the areas with a high concentration of population, mines and industrial enterprises (Tallinn, Kunda, Kohtla-Järve, Sillamäe, Narva). The area spoilt by mining, buildings and industrial activities amounts to 30,648 ha (Ranniku 1993). Since the beginning of the use of oil shale deposit, 858 million tons of oil shale have been mined; the production culminated in 1980 (Paalme 1995). The total area under the oil-shale mines is about 11,000 ha. Of that some 9000 ha damaged by surface mining has been reclaimed (Paalme 1995). About 6000 ha of land is under milled peat fields. The mining of phosphorites was terminated in 1991, but 59 ha of the mined-out area still remain to be recultivated. Most of the exhausted areas are usually reforested, some turned into fields, some into settlements.

The re-establishment of the independent Estonian Republic in 1991 brought about several changes in the landuse and environmental policy. Big farms were liquidated and foundation of small farmsteads is under way. In 1883, the land structure was the following: arable land - 14%, meadows - 21%, heath (pastures) - 14%, mires - 21% (Kahk 1992). After one hundred years, arable land made up 25%, meadows 6.04%, forests 44.05%, water basins 2.00%, and others (including mires and brushwoods) 22.91% of Estonia’s total area (Järv 1994). In the historical retrospective, during the last 100 years the arable area has increased by 78%, and forested area 68%, while the area under meadows and pastures has drastically decreased - 5.8 times. It leads to the conclusion that the number of cattle in the 19th century could have been even bigger than nowadys and the manuring of fields very extensive. At the beginning of 1994, there were 10,179 farms with a total area of 252, 200 ha. The average size of a farm was 24.8 ha, of that 43.2% were fields, 33.8% woodlands and 16.1% grasslands and shrubbery (Sein 1994). Compared with the beginning of the 19th century, the total number of farms has decreased 1.8 times. The centres of arable farming have also shifted. At the beginning of Medieval times, the arable farming was flourishing in northern Estonia (Henriku... 1993). Now the percentage of arable land is highest in the Tartu County (34.7%) and lowest in the Ida-Viru County (10.9%). The area of forest is largest on the Island of Hiiumaa (58.4%) and smallest in the Tartu County (38.8%).

The recent restructuring of economy has brought about several positive trends in the environmental management. The decrease in waste water and pollution load with efficiency of the new sewage treatment plants have improved the water quality in rivers (Pachel & Loigu 1995). But still some rivers are polluted with nutrients (Keila, Vääna, Selja, etc.), some of them even with heavy metals, phenols and oil products (Purtse, Kohtla). Especially large pollution load comes from the ash hills of the joint-stock companies Kiviter, Silmet, and Eesti Põlevkivi. The Kroodi Brook is a real waste water drain, receving untreated or poorly treated domestic and industrial waste water from the Maardu Settlement and from the joint-stock company Eesti Fosforiit.

As to air pollution, the situation is worst in the Kunda, Kohtla-Järve and Narva areas. The total emisson of pollutants into the atmosphere was highest in 1980.  In 1994, it was only half as high, owing to the several international conventions Estonia has recently jointed and the better technology introduced. Still, in 1994 the volume of pollutants, emitted into the atmosphere, was 354,000 tonnes, including 161,500 tonnes of solid, 141,100 tonnes of SO2 and 14,600 tonnes of NOx (Saar 1995). Of the total air pollution load, about 50% comes from transport and about 50% from the other sources.

In conclusion, it should be pointed out that during half a century material-and-pollution-extensive large-scale industry and agriculture have caused serious environmental problems in northern Estonia. Exploitation of mineral recources and utilization of waste have deteriorated the groundwater quality, contaminated soils and air. Luckily, several problems which were topical during the last decades of Soviet power in agriculture, like those related to large farms and point pollution sources, huge fields and soil erosion, intensive use of agrochemicals, etc., are gradually loosing their actuality.

The above shows that during the course of the historical times human impact has considerably increased and its character changed. Only one century ago Estonia was a typical agrarian country with well developed rural settlements and farms orientated to cattle rearing and crop farming. Roughly 20% of the territory was cultivated. To date, arable land forms ca. 25% of Estonia’s area; nearly 75-80% of the territory has undergone land improvement. It means that anthropogenic stress on our environment has been tremendous. Hopefully, the implementation of the private-farm-system in agriculture will promote restoration of traditional nature management habits in Estonia.




The Mineral Resources Classification System, worked out by the Estonian Commission on Mineral Resources, is built up on internationally accepted principles and is in good accordance with the last (1979) Classification of the UN Committee on Natural Resources. Category T corresponds to category R1, category R to R2, category P to R3 in the UN classification. The term “active” expresses the economic (E) status of resources; while the term “passive” marks their subeconomic (S) status.

In this chapter, the term “output” means the amount of mined out resources, the mining losses are not included.


Kukersite oil shale

H. Bauert & V. Kattai


The Baltic Oil Shale Basin (ca 50,000 km2) is situated prevailingly in northeastern Estonia with a part of it extending eastward into Russia. Three well-explored oil shale deposits – Estonia, Leningrad and Tapa – occur within the oil shale basin. The boundary between the Estonia and Leningrad deposits is purely geopolitical and it runs along Estonia’s border with Russia. Both deposits are currently mineable, while the Tapa oil shale deposit in central Estonia is considered to be a prospective one.

At present, the Estonia deposit is the largest commercially exploited oil shale deposit in the world; its total resources exceed 7 × 109 tonnes of oil shale. The resources of the prospective Tapa deposit are in order of 2.6 × 109 tonnes. The Estonia deposit has been mined continuously since 1919, with the maximum annual output of 29.7 million tonnes in 1980. Currently, kukersite oil shale is mined in six underground mines and in three open-cast pits. The total annual output is about 13 million tonnes.

The main oil shale (kukersite) sequence is of Middle Ordovician (Llandeilo - Early Caradoc) age. It contains up to 50 laterally continuous kukersite seams which can be traced at a length of 250 km in the eastwest direction. Basinward, however, they lose organic matter rather abruptly and are no longer distinguishable from the host limestone beyond 40-50 km.


Geological setting of the kukersite oil shale

Minor kukersite-type organic matter (OM) accumulations have been recorded throughout the whole Middle Ordovician time in the Baltic Oil Shale Basin (Männil 1966, Rõõmusoks 1970, Põlma 1982). The major kukersite OM accumulation, however, took place during early Middle Ordovician time (Llandeilo - Early Caradoc, Männil 1990) in an area presently known as the Baltic Oil Shale Basin which is located on the southern slope of the Fennoscandian Shield. The kukersite OM accumulated in an area larger than 50,000 km2 extending from western Estonia to St. Petersburg in Russia. In its present outline, the main kukersite accumulation area resembles an elongated crescent (Fig. 202) and apparently represents the southern and central parts of the original deposit. The actual shape and the northward extent of the original depositional basin remains unknown because