East Europe extends from north to south about 3000 km. We can observe consecutive meridional change of natural zones named on the basis of their vegetation type: tundra, forest-tundra, zone of coniferous forests (taiga), zone of mixed and deciduous forests, forest-steppe, steppe, desert and mountain zones (Fig. 1).

Distributions of modern amphibians and reptiles cover at present almost the whole of Eastern Europe, and their limits of distribution are determined by different factors: presence of sufficiently long seasons with positive temperatures, the degree of water warming during amphibian larval development, and so on (Bannikov and Denisova, 1956). Notably, the distribution boundaries closely coincide with natural zone transitions (Fig. 2).

The northern boundaries of the distribution of terrestrial ectothermic vertebrates almost never transverse the limits of forest-tundra and are commonly located far to the south (only the areas of Salamandrella keyserlingii, Rana arvalis and Zootoca vivipara rarely extend into low bush and sedge tundra). Northern distribution boundaries of Bombina bombina, Pelobates fuscus, Bufotes viridis, Pelophylax lessonae, Pelophylax ridibundus and Coronella austriaca roughly coincide with the transition from coniferous taiga to mixed forest. The southern distribution limits of S. keyserlingii, Triturus cristatus, Lissotriton vulgaris, Bufo bufo, P. lessonae, Rana temporaria, Anguis fragilis, Z. vivipara, Vipera berus and northern distribution limits of Emys orbicularis, Eremias arguta and Vipera renardi generally coincide with the location of the forest-steppe zone (Bannikov et al., 1977 and Nikolaeva, 1986).

Every species is associated with certain types of biotopes (Ananjeva et al., 1998, Bannikov et al., 1977, Garanin, 1983 and Kuzmin, 2012). Both population density and the number of individuals inhabiting a given territory depend on the distribution of biotopes (Ratnikov, 1996a). Since adjacent natural zones can include identical biotopes, every amphibian or reptilian species may inhabit two or more zones; on the other hand, each zone is characterized by a specific association of species (Table 1). Therefore, the environmental conditions and possibly the natural zone in the vicinity of a locality can be reconstructed on the base of a fossil assemblage.

The following ecological types can be marked out among modern terrestrial ectothermic vertebrates of Eastern Europe (Ratnikov, 1996a): •

species of closed biotopes typical for the forest zone are: the Siberian Salamander S. keyserlingii, the Smooth Newt L. vulgaris, the Crested Newt T. cristatus, the Common Toad B. bufo, the Grass Frog R. temporaria, the Pool Frog P. lessonae, the Slow Worm A. fragilis, the Viviparous Lizard Z. vivipara and the Adder Viper V. berus. The Moor Frog R. arvalis is more characteristic for the forest zone, but also inhabits tall grass wet meadows and, therefore, has a very broad area, extending from steppes to forest-tundra. The distribution of the Smooth Snake C. austriaca also extends from forests to steppes, although this species prefers closed biotopes as well;

Of the above-mentioned natural zones, the forest-steppe zone is the richest in species number because it is inhabited by representatives of both closed and open biotopes. Since typical tundra species of amphibians and reptiles are absent in Eastern Europe, the forest-tundra is inhabited only by woodland taxa.

It is obvious that the covering glaciers, periodically coming from Scandinavia, have had an influence on the distribution areas of animals. First of all, ectothermic animals evidently disappeared from the territories covered by glaciers. But changes also occurred in the periglacial territories.

Not so long ago, all natural zones in the Quaternary period were considered to have been displaced under the influence of decreasing temperature across the glacier edge, and the animals within the zones migrated as the zones were displaced. The spatial location of ecosystems was believed to constantly fluctuate from the north to the south and back. However, experts on small mammals noted the occurrence during glacials of so-called “mixed faunas”, consisting of fossil assemblages of taxa (genera and species) that in the present day do not occur in the same biotopes; this concerns in particular steppe and tundra species (Agadzhanyan, 1972, Markova, 1994, Markova, 2004, Markova and Puzachenko, 2008 and Rekovets, 1995). Formation of such assemblages is possible if the forest zone is eliminated as a barrier between tundra and steppe. Therefore, the hypothesis of zones migration does not explain the available data. We need a hypothesis that explains both the formation of the “mixed faunas” and the restoration of a forest zone after a glaciation.

Importantly, temperature decrease is not the only influence of a glacier on an environment. Condensation of water in ice leads to the aridization of the climate and is accompanied by the reduction of forest and by the predominance of grassy vegetation (Velichko, 1973). Thus, the distribution of tundra during continental glaciations moved towards the south, whereas the steppe zone moved towards the north. At the same time, the forested area was divided into ever decreasing areas that were gradually reduced to rather small massifs. As a result, unique periglacial landscapes were formed. They were populated by original theriocomplexes that included tundra and steppe representatives. Analogues for them do not exist today (Baryshnikov and Markova, 2002, Markova and Puzachenko, 2008 and Rekovets and Nadachowski, 2007). That means that glacial ecosystems differed in their structure both from interglacial ecosystems from the past and from all ecosystems existing nowadays, which are also interglacial. In the background of these periglacial landscapes, refuges of forest vegetation persisted in river valleys.

Natural zones restored gradually after the retreat of glaciers. Forests did not come from the south, but appeared as a result of expansion of tree taxa in the refuges as humidity increased in the watersheds (Ratnikov, 2013).

Against the background of vegetation change, the distribution areas of “ectothermic” animals constantly varied. The distribution areas of two ecological types, the inhabitants of closed and opened biotopes, also changed (Ratnikov, 2009). The splitting of forests into a mosaic, with the subsequent reduction into refuges, led to the corresponding distribution fragmentation and area reduction of species in closed biotopes. Most likely, amphibians and reptiles did not survive in all refuges of forest vegetation. Then, after the retreat of glaciers and subsequent expansion of forests, the distribution areas of the populations inhabiting these refuges widened and eventually united.

During the fragmentation of forest areas at the beginning of glacial periods, the steppe species had opportunities to advance toward the north along the “corridors” of steppe between the “islands” of forests, and then to widen their distribution as these “islands” reduced in sizes. This expansion was bounded by the limits of glaciers distribution. During a deep subsequent penetration of continental ice toward the south, especially at the time of Don or Moscovian (= Dnieper) glaciations (MIS 16, 659–621 kyr BP and MIS 6, 186–127 kyr BP, respectively, according to Iossifova et al., 2009), areas of steppe species became smaller still. Then, during glacial retreat, the restoration of forest area reduced the open spaces and consequently, the range of their associated species.

In contrast with mammals, East European species of amphibians and reptiles do not include typical tundra forms, and “mixed” herpetofaunal complexes have not been observed. Only the degree of herpetofaunal diversity distinguishes assemblages from the different climatic periods (Ratnikov and Mebert, 2011).

The use of the herpetological information for biostratigraphical purpose has been detailed in Table 2. First of all, the presence of “ectothermic” animal remains strongly suggests the absence of ice cover in a given place. Such situations existed either between glaciations, or outside of glacier distribution.

As forest-linked herpetofauna within the limits of periglacial hyperzone were concentrated in a few refuges, fossil associations of amphibians and reptiles at glacial intervals show poor species diversity, consisting mainly of steppe elements. Change from forest associations to forest-steppe or steppe assemblages within the limits of modern forest zone is attributed to climate aridization as a result of the coming glaciation. Therefore, during paleogeographic reconstructions, it is important to determine what is a natural zone, as opposed to an assortment of possible biotopes. For this purpose, it is necessary to consider quantitative ratios of the remains of various ecological types (Ratnikov, 1996a and Ratnikov, 2002).

Ratios in a fossil assemblage can be different from the original ratios and require appropriate corrections. First of all, we must consider burial conditions, which have an influence on the composition of the fossil herpetofauna. The taphonomic type of locality regulates the area from which fossils are derived. Alluvial channel deposits contain remains of amphibians and reptiles from a significant part of a river basin and, therefore, they mostly reflect the composition of herpetofauna that inhabited this territory during the locality formation. Inundated, dead channel, limnetic, marsh, proluvial, deluvial, karst and cave deposits contain the herpetofauna remains from the close vicinity of the pond or locality (Ratnikov, 1996a and Ratnikov, 2002). Another correction is the degree of preservation. Agadzhanyan (1979) pointed out that the destruction rate of organic remains is higher in forest conditions than in steppes, due to the slower burial rate, the abundance of scavengers, and the presence of humic acids in soils. Therefore, even a few forest animal fossils can indicate a wide distribution of forest and bush biotopes. An equal ratio of forest and steppe animal fossils may be interpreted as a significant predominance of closed biotopes over open ones. Glacial intervals for modern steppe zones can theoretically be recognized by the emergence of desert species.

Interglacial herpetofaunas are characterized by an increased diversity and the presence of thermophile southern forms. Interglacials differed from each other in their climatic parameters (Bolikhovskaya, 1995 and Ratnikov, 2013). Therefore, distribution areas of amphibians and reptiles in interglacial optimums may also have had different outlines. Such fluctuations may depend not only on changes in climate and borders of natural zones, but also on other factors, such as the assortment of forming biotopes, the presence of competitors, predators, food, etc. The cumulative pressure of these factors could lead to initially inexplicable distributions. As a result, distribution areas sometimes varied only slightly, but sometimes they expanded, or even disappeared within significant territories. Some species of amphibians show changes of distribution that can be used in regional stratigraphy for defining of large age intervals. For example, the Mongolian Toad (Strauchbufo raddei) living nowadays in Eastern Russia, Mongolia, Manchuria, Korea, North and central China (Kuzmin, 2012), lived in Eastern Europe from the Pliocene up to and including the Middle Pleistocene (Ratnikov, 1996b, Ratnikov, 2002 and Ratnikov, 2009) (Fig. 3). Numerous reliable findings of the Green Toad (B. viridis) appear here only from the Late Pleistocene onwards. Only two bones of this species have been found in more ancient sediments together with S. raddei: one comes from the Pliocene of the locality of Dolinskoe and another from the il’inka horizon of the locality of Staraya Kalitva (Ratnikov, 1996b, Ratnikov, 2002 and Ratnikov, 2009) (Fig. 4). Two bones, one of S. raddei, and another of B. viridis, have been found in Morozovka. This site has been estimated by various experts as either Late or Middle Pleistocene. Assuming that the bone of S. raddei was not redeposited, both species would have coexisted at the beginning of the Late Pleistocene. The finding of the Mongolian Toad therefore points to a sediment age of that is older than the Late Pleistocene, whereas the Green Toad remains can, most likely, be attributed to a younger age. The distribution area of the Siberian Salamander (Salamandrella sp.) at the end of the Early Neopleistocene covered most of the East European plain (Fig. 5).

As an example of an analysis of an amphibian-reptile association, I will now discuss the herpetofauna of the Mastyuzhinka site, located in the basin of the river Ikorets (a tributary of the river Don) near of the town of Liski in the Voronezh area. Characteristic of the association is that more than half the anuran bones belong to small (juvenile) individuals. The site was probably a shallow pool with quiet, warm water in which young amphibians developed. The animals, which died either here or on the nearby land, were buried in this reservoir.

According to modern classification (Frost, 2015), of the composition of the fossil amphibians and reptiles fauna from the Mastyuzhinka site is as follows: Salamandrella sp. – 114, Caudata indet. – 3; Bombina cf. variegata – 6, Bombina sp. – 3, P. fuscus – 1, P. cf. fuscus – 3, Pelobates sp. – 115, S. raddei – 32, Bufonidae indet. (not Bufo) – 26, Bufonidae indet. – 32, R. arvalis – 21, R. cf. arvalis – 6, Rana sp. – 8, Ranidae indet. – 71, Anura indet. – 63; L. agilis – 10, L. cf. agilis – 3.

At least six species present in this association are indicative of favorable (warm and damp) conditions at the time of burial. Three species (Bombina variegata, Salamandrella sp., S. raddei) do not live in the Voronezh area today. Remains of S. raddei exclude a Late Pleistocene age for the herpetofauna. Salamandrella sp. reduces the age interval further as this form is only found in after-Don deposits of the Early Neopleistocene in central Eastern Europe, suggesting the Muchkap or Ikoretsk interglacial.

The prevailing quantity of Salamandrella sp. remains (the species of closed biotopes) testifies to the existence of forest conditions in vicinities of the site. Moreover, species of open biotopes (S. raddei, L. agilis) are also present, suggesting that this forest is within the limits of the forest-steppe zone. The numerous remains of Spadefoot would seem to confirm this hypothesis. As the site is within the limits of the modern forest-steppe zone, we can conclude that conditions during deposition were not necessarily drier than today, so it was, most likely, an interglacial optimum.

It is obvious that use of such methodology requires knowledge of the history of the species inhabiting the study region. Potential candidates for this purpose are B. variegata, Bufo verrucosissimus, E. arguta, Elaphe dione and Zamenis longissimus, the remains of which are also discovered in localities distant from their modern distribution areas (Ratnikov, 2009).

The dynamics of past distribution areas of modern amphibian and reptilian species, based on fossil remains recovered in archaeological and/or palaeontological localities, allows the following stratigraphical inferences: •

the presence of the “ectothermic” animal remains suggests the absence of ice cover in a given place at the time of burial of the herpetofauna;

Nevertheless, it is necessary to consider each case carefully within the framework of the history of the locality.