Freshwater flows that occurred in the epoch of Pleistocene glaciations during the drainage of ice-dammed lakes in the intermont basins of Russian Altai are among the world’s largest megafloods. Despite the large array of 14С-, 10Be-, OSL-, and IR-SL-dates of glacial, lacustrine, and flood deposits, the chronology of Pleistocene events in Russian Altai remains debatable. There exist the following major alternative views about the scales and time of formation of glaciers, ice-dammed lakes, and outburst floods: (1) the largest megafloods in the Kurai and Chuya basins occurred at the end of the Late Pleistocene during marine isotope stage (MIS) 2, which was determined by the maximum glaciation area in the Pleistocene over this period [1, 2]; (2) the major floods occurred at the beginning of the Late Pleistocene during MIS 5 or even MIS 6 of the Middle Pleistocene in the course of the glacial maximum, while during MIS 2 ice-dammed lakes were not formed in the Chuya and Kurai basins due to the valley-cirque character of the glaciers [3], or in the outlet of the Kurai basin during MIS 2 there was only a small lake, which was dammed by a large glacier and did not flood [4, 5]; (3) the formation of the lakes and their catastrophic outburst floods occurred during different epochs of the Late Pleistocene; the latter relatively large ice-dammed lake in the Kurai basin existed during MIS 2 and drained catastrophically ~ 19‒16 ka [6, 7]. The different ideas about the age and scales of glaciations and glaciogenic floods determined the differences in the reconstructions of the age and genesis of the Inya and Saldzhar formations that compose the high and medium terraces in the valleys of the Chuya and Katun rivers: both formations are related to the floods in MIS 2 [1, 2] or are separated by their age and the deposition of the even younger Saldzhar formation is confined by the upper limit of 90 ka (MIS 5) [4]. Thus, it is still urgent to solve many controversial issues of stratigraphy and paleogeography during the Pleistocene in Altai.

In this work, the chronological reconstructions of glaciations, ice-dammed lakes, and glaciogenic floods in Russian Altai were performed based on the study of 13 sections from the Chuya basin to the mouth of the Sema River (Fig. 1). The complex of applied methods included the following: geomorphological, lithological, grain-size, X-ray phase, morphoscopic analysis of quartz grains, micromorphological, microfaunistic, palynological, and a complex group bioanalysis, as well as the methods of absolute geochronology. OSL- and IR-SL-dating was performed at Silesian University of Technology, Poland; 14С LSC-analysis was done at Sobolev Institute of Geology and Mineralogy, Siberian Branch, Russian Academy of Sciences, and at the Institute of Geography, Russian Academy of Sciences, 14С AMS-analysis, at the Institute of Geography, Russian Academy of Sciences, and the Institute of Archaeology and Ethnography, Siberian Branch, Russian Academy of Sciences (sampling, producing graphite targets for the IGANAMS and GV dates, respectively) and at the University of Georgia and Novosibirsk State University (measuring for the IGANAMS and GV dates, respectively). The 14С-dates are reported calibrated (2σ) in the CALIB Rev 8.2 program (http://calib.org).

Fig. 1.
figure 1

Position of the studied sections in the basins of the Chuya and Katun rivers in Russian Altai.

Full size image

Our study made it possible for the first time to determine the age of the highest (2100‒1945 m a.s.l.) lacustrine terraces in the Chuya basin, to date the base of the glacial formation of the Chagan section, the key section for determining the number and ages of glaciations in Russian Altai, by modern methods, to substantiate the existence of the ice-dammed lake during the entire MIS 2 in the Kurai basin and for the first time to identify two associations of ostracods in its deposits, and to determine the age and different genesis of the deposits in the middle terraces in the upstream part of the Katun river, as well as in the mouths of its inflows Inya and Sema.

To reconstruct the period of existence of the largest paleo-reservoir with the level up to 2100 m a.s.l., which was likely to be common for the Kurai and Chuya basins, the section of the lacustrine terrace deposits on the left bank of the Yustyd River in the eastern part of the Chuya basin served as the key section (section 1 in Fig. 1). A series of sloping strandlines up to the mark of 2100 m was formed by the lake in slope deposits of the rocky slope; however, at its bottom (2055 m a.s.l.) the terrace has a more complex structure. In the coastal cliff 7‒8 m high, colluvial and alluvial deposits overlain by a layer of boulder perluvium (moraine outwashed by the lake) are exposed (bottom-up). The OSL-dates of 85.1 ± 5.8 ka GdTL-3903 and 81.8 ± 5.9 ka GdTL-3904 for the sand lenses in alluvium show that the ice-dammed lake that left the maximum high strandlines on the slopes of the Chuya basin existed later than 90‒80 ka, i.e., in the Late, rather than Middle Pleistocene. At the same time, the fact of existence SE Altai glaciation during MIS 6 of the Middle Pleistocene was confirmed by three coincident OSL-dates: GdTL-3902 169.1 ± 8.9 ka, GdTL-3900 168.3 ± 9.1 ka, and GdTL-3901 166.0 ± 9.2 ka for stratified sands at the base of the thick (up to 45 m) lens of glacio-lacustrine deposits at the base of the moraine deposits in the Chagan reference section in the southwestern part of the Chuya basin (section 3, Fig. 1). These dates made it possible to refine the magnetostratigraphy of the glacial deposits in Altai: the episode of the geomagnetic field anomaly, the Chagan event, dated conventionally at about 266 ± 30 ka in the 1970s based on a single TL date [8], was established to have begun much later—at about 170 ka.

The upper chronological boundary for the formation of the high lacustrine terraces in the Chuya basin is the oldest of 14С-dates of multilayer cutans (films) on the fragments of rocks in the proluvium terraced by the lakes at the base of the Kurai Ridge at a height of 1945 m a.s.l. (section 4, Fig. 1), 26351 ± 254 cal BP IGANAMS 7059. It was obtained by carbon of carbonates in the sparitic layer, which was formed first on the fragments during soil formation after drainage or the decrease in the lake level below 1945 m. The ancient lithogenic carbon that is present in the sample could have led to obtaining an older date, which can be interpreted only as a reference age marker when the soil began to form. Thus, the uppermost (2100‒1945 m) lacustrine terraces were formed in the Chuya and probably Kurai basins between 90‒80 and ~26 ka. Correspondingly, in the case of rapid draining of the lake (lakes) with such levels, catafluvial (diluvial) deposits in the Chuya and Katun valleys did not begin to accumulate until 90‒80 ka (which does not exclude the accumulation during draining of older but smaller lakes and the participation of alluvial processes in the accumulation of the Inya formation in general). These data are consistent with the OSL- and 14С-dates that indicate the existence of the latter, smaller, catastrophically drained ice-dammed lake in the Chuya basin between 37 and 11 ka [9]. We updated its upper age limit: ~14 ka the level of the lake, if it existed by that time, was below 1890 m; two coinciding 14С AMS-dates IGANAMS 7059 14 174 ± 130 cal BP and GV04026 14 453 ± 369 cal BP were obtained for the thighbones of a Marmota baibacina skeleton lying in situ in the deposits at the base of the Sailyugem Ridge (section 2, Fig. 1).

The Kurai basin did not remain dry during the entire MIS 2, as considered in [3]: for the deposits composing the lowest (1525 m) accumulative–abrasion lacustrine terrace at the base of the Kurai Ridge foreberg (section 8 in Fig. 1), the OSL-dates were obtained: 25.1 ± 1.7 ka GdTL-3487 and 19.0 ± 1.3 ka GdTL-3486 from the depths of 1.6 and 1.1 m. respectively [10]. The dates mark sudden changes in the reservoir depth and in the character of lacustrine sedimentation, which was confirmed by the ostracod fauna analysis (Fig. 2). The alternation of gray and yellow clays (2.5‒1.6 m, the deposits in the deeper lake with the changing level) includes monospecies accumulations of Leucocytherella sinensis Huang (earlier identified as Leucocythere sp. 1 and sp. 2 [11]) (Fig. 2). High dynamics of the reservoir ~25‒19 ka is indicated by the formation of lens-like, horizontal- and cross-stratified interbedding of sands and clays, with inclusions of scree and fragments of the weathering crust by the foreberg rocks (1.6‒1 m). The depth lowered significantly about 19 ka, which triggered the development of Leucocythere dorsotuberosa Huang, Leucocytherella sinensis, Ilyocypris tibeta Peng et al., and Potamocypris arcuata (Sars) (1.0‒0.55 m) (Fig. 2). The samples from the lower parts of both associations contain the highest number of ostracods, up to 60 valves in the lower and many hundreds of valves in the upper association. Single gastropods were encountered together with ostracods. Microphytoplankton was recorded in situ across the entire section: numerous cells of green algae Botryococcus, Pseudopediastrum boryanum, Ovoidites, and single Schizosporis. Pollen grains in situ are single. Pollen of the hardy-shrub Alnus sp., Betula sp., B. sect. Nanae, Salix, Pinus sylvestris, Pinus sibirica, Pinus sp., Picea sp., and Abies sp. was encountered; pollen of herbs and subshrubs belongs to Artemisia, Amaranthaceae, Ephedra, Polygonaceae, Ericaceae, Ranunculaceae, Rosaceae, Saxifragaceae, and Valerianaceae; moss spores (Sphagnum, Encalypta) and ferns (Botrychium) are present. The vegetation during the deposition of sediments was probably treeless, and open landscapes dominated. A sudden drop in the lake level about 19 ka led to intensive erosion of the Neogene deposits at the base of the Kurai Ridge and the fast redeposition of well-preserved pollen of the Miocene flora and numerous shells of Cenozoic ostracods. Coal formations served as the source of redeposited (primarily at the base of the section) fragments of fish skeletons, teeth, and scale; we note that these were the first finds of coal ichthyofauna in Southeastern Altai. About 19 ka, the upper level of the Kurai paleolake, which was dammed by the Maasheisk glaсier, was evidently not lower than 1730 m, and it spread to the Chuya valley between the Kurai and Chuya basins. This is indicated by the lacustrine terraces in the eastern part of the Kurai basin and the OSL-date of 18.8 ± 1.2 ka GdTL-3485 for lacustrine-delta deposits, a layer of coarse-grained sand with horizontal lens-shaped low-angle wavy bedding 30 cm thick and overlain by fine-grained horizontally-layered sand and massive clay loam at the level of 1730 m in the mouth of the Kuektanar River, which is 20 m above the Chuya water line (section 7, Fig. 1). The fact that the sand grains are unrounded in the dated basal layer is likely to show the short-term existence of the lake at this level. The Kuektanar glacier moraines underlying lacustrine deposits and moreover the glacier itself blocked up the Chuya valley before 19 ka. The local landslide–moraine–dammed lake was formed upstream the Kuektanar mouth later, ~16 ka [12] and as the OSL-dates showed 10.76 ± 0.68 ka GdTL-3482, 10.51 ± 0.67 ka GdTL-3483, and 10.15 ± 0.72 ka GdTL-3481 [10] (section 5, 6, Fig. 1), it existed until the onset of the Holocene. The weak roundness of quartz grains from the horizontally layered sandy deposits of this lake points to the close position of the source areas.

Fig. 2.
figure 2

Section 8 of the deposits of the Kurai ice-dammed paleolake and the results of its integrated study. The changes in the depth were reconstructed according to the data of the X-ray phase analysis. N is the number of ostracods in the sample. (1) Clays, (2) clay loams, (3) sand, (4) unrounded fragments of clay, (5) scree, (6) broken rock, (7) boulder, (8a) sampling places for ostracod, palynological, complex biological, and X-ray phase analyses, (8b) samples for OSL-dating. Ostracods: (1) Leucocytherella sinensis Huang, left valve: (1а) lateral view, (1b) inner view, 2.4‒2.5 m; (2) Ilyocypris tibeta Peng et al, right valve: (2а) lateral view, (2b) dorsal view, 0.85‒0.95 m; (3) Potamocypris arcuata (Sars), left valve: (3а) lateral view, (3b) dorsal view, 0.85–0.95 m; (4) Leucocythere dorsotuberosa Huang, right valve: (4а) lateral view, (4b) dorsal view, 1.0‒1.05 m.

Full size image

The recent data complement the reconstructions of the Kurai paleolake: earlier, the OSL-dates in the interval of 19‒16 ka were obtained for the catafluvial and lacustrine deposits in the lacustrine terraces with the marks of 1570 and 1650 m [7, 13]. Its final drainage occurred no later than 16 ka: for Chuya alluvium (rhythmical alternating of clays and sands exposed in sections 9 and 10 in the right-bank scarp 3‒12.7 m high) in the outlet of the Kurai basin, near Baratal area, we calculated the OSL-dates of 16.7 ± 1.1 ka GdTL-3478, 16.3 ± 1.0 ka GdTL-3479, 14.7 ± 1.0 ka GdTL-3488, 12.2 ± 0.7 ka GdTL-3478, and 12.0 ± 1.4 ka GdTL-3489. Thus, the recent data do not allow us to agree that ice-dammed lakes did not occur in the Kurai basin during MIS 2 [3] or that only local Baratal Lake dammed by the Chibit glacier existed continuously in the outlet of the Kurai basin without draining or level changes for at least 21‒14.4 ka [5].

The paleontological and geochronological studies of the Saldzhar formation in the Katun River valley and in the mouths of the tributaries showed that its deposition was long-term and related to frequent floods during the entire MIS 2. The formation was also found to include lacustrine deposits corresponding to the long-term inter-flood periods. For example, in the mouth of Sema, a left tributary of Katun, at the base of the section of catafluvial (from [1, 3]) deposits composing the 60-m-high terrace (section 13, Fig. 1), in the 35-cm thick layer of thinly-alternating clays and fine-grained sands, we identified numerous fauna of gastropods, charophyte remains, tubes of lugworms; the lake was shallow (up to 1.5‒2 m deep), heated, with weak hydrodynamics. For the gastropod shells from the roof of this layer, we obtained the AMS 14C-date of 22 338 ± 151 cal BP GV04025. For a large (100 × 2 m) lens of coarse-grained sand lying in the section above this layer, we obtained the OSL-date of 32.3 ± 3.2 ka GdTL-3899; however, we noted that it could be older due to incomplete light exposure of quartz grains in the aqueous medium [10], which was confirmed by 14C-dating of gastropod shells. Despite the inversion, both dates indicate that the terrace deposition started no earlier than at the boundary between MIS 3 and MIS 2. In the Inya valley, at the base of section 12 (Fig. 1), which was proposed earlier as a parastratotype of the Saldzhar formation dated to MIS 4 or even MIS 5 [3, 4], we have recently obtained two 14С-dates indicating that the grassy swamp deposits are overlain by the sediments of diluvial–lacustrine cyclites ~16‒17 ka [10]. Draining of the last ice-dammed lake in the Kurai basin at this time [7] and the passage of the flood across the main Katun River valley could result in a sudden change in the sedimentation character in the Inya valley. The fact that the deposition of the Saldzhar formation in the Katun valley in the area of the Chuya mouth was completed at the very end of the Late Pleistocene (during the late MIS 2) is also indicated by the OSL-dates we obtained for the sediments on the edge of the Katun left-bank terrace 70 m above the water line (section 11, Fig. 1). Cutting into the middle terrace did not begin here until 16‒12 ka, i.e., at the very end of MIS 2 and the onset of the Holocene [10]. Thus, the chronological and paleontological data for sections 10‒12 contradict the ideas in [3, 4] that the Saldzhar formation and the Chuya and Katun medium terraces it created had accumulated as early as 90 ka and had an exclusively catafluvial genesis.

In general, the results of multidisciplinary studies showed that the bordering mountains of the Chuya basin were affected by glaciation during MIS 6 in the Middle Pleistocene. However, the ice-dammed lake with the highest (up to 2100 m) terraces in the Chuya basin existed later than 90‒80 ka in the Late Pleistocene. This benchmark limits the influence of possible drainage of the lake on the accumulation of the Inya formation in the Chuya and Katun valleys. It was confirmed that the large ice-dammed lake with a water level of at least 1730 m existed during MIS 2 in the Kurai basin. It finally drained no later than 16 ka. Its rapid drainage could initiate the formation of diluvial–lacustrine cyclites ~17‒16 ka in the Inya mouth. In general, the Saldzhar formation was deposited in the Katun valley between the Chuya and Sema mouths over a long time, during the entire MIS 2, as a result of both catastrophic and gradual hydrological processes. It is necessary to revise the concept adopted by SibRISC in 2018 (Novosibirsk) as a working hypothesis, which assumes the absence of ice-dammed lakes in the intermont basins of Russian Altai in MIS 2, the deposition of the Saldzhar formation and the medium terraces it composed in the Chuya and Katun river valleys before 90 ka, and the Inya formation and the high terraces ~150 ka during MIS 6. Multidisciplinary study of the deposits in the basins of these rivers needs to be continued.