INTRODUCTION

A significant volume of Upper Neopleistocene deposits in the Kolyma Lowland is occupied by cryopedolites—silty strata processed during their accumulation by synlithogenic soil formation, permeated by thick ice veins, and called the ice complex or yedoma (Kaplina et al., 1978; Tomirdiaro, 1980; Kaplina, 2009). The specific accumulation of deposits determined the absence of differentiation of the material of strata and general morphological homogeneity (Gubin, 1984). Gubin (2002) showed that cryopedolites of Karginsky (Molotkovsky) age, correlated with the formations of marine isotope stage (MIS) 3 through radiocarbon dates and cryopedolites of Sartansky age, corresponding to MIS 2, are similar in morphological structure: uniform coloration, the presence of highly dispersed plant detritus and thin grass roots, layering, and cryotextures. Analysis of the particle size distribution of these deposits also showed their similarity—the absolute predominance of dust with prevailing particle sizes within 0.05—0.01 mm. These cryopedolites are characterized by a neutral or slightly alkaline pH (7–8.5) and similar Corg values (0.8–1.8%) both along and across the bed, which, along with uniquely high content of mobile forms of phosphorus (up to 60 mg/100 g), indicates significant participation of living organisms in their formation (Zanina, 2006). The similarity of the above characteristics indicates a certain similarity of the accumulation conditions of these strata. They differ in a slightly higher amount of organic material and increased content of coarse humus in the Sartansky formations, as well as in more contrasting differences in the Corg content between the layers of these deposits (Gubin, 1994, 1998). Processing and humification of the fossil organic material and the presence of peated material with significant participation of mosses observed in the Karginsky deposits indicate better heat supply in the summer period of their formation and higher humidity as compared to the Sartansky time (Gubin, 2002). The Karginsky strata contain fossil burrows of rodents and gophers (Gubin et al., 2001, 2003a; Zanina, 2005; Lopatina and Zanina, 2006; Zanina et al., 2011), as well as four beds of buried soil of different age (Gubin and Zanina, 2013, 2014; Lopatina and Zanina, 2020a, 2020b). The lower bed of buried soil (BC) is attributed to the Early Karginsky pedocomplex (40 ka or older), and the three overlying beds are attributed to the Late Karginsky one (BC I, 37–35 ka; BC II, 33–31 ka; and BC III, about 28 ka). No buried soils were found in the Sartansky deposits of the studied region, which may be due to the increasing severity of the climate and the increasing inflow of mineral material.

A significant amount of fossil plant material (remnants of trunks, branches and stems, roots, as well as leaves, fruits, seeds, spores, and pollen of plants) was identified in the Upper Pleistocene deposits of the Kolyma Lowland; seeds of higher plants, in particular, of catchflies (Yashina et al., 2002), were determined as capable of germination. The relatively rapid transition of fossil plant remains into the buried state and their subsequent storage in the layer of annual summer thawing at relatively low above-zero temperatures, ensured by high cold reserves in permafrost sediments, resulted in good preservation of the studied plant material. Gubin et al. (2003b) concluded that the transformation of plant material occurred only in the upper part of the seasonally thawed layer, in the zone with diurnal and seasonal temperature variations and, accordingly, with repeated phase transitions of water and active action of soil solutions and microbiota. Organic material resided in this zone only one or two months a year during the first few dozen years. Further, with increasing burial depth and before the transition to permafrost, summer temperatures did not exceed 0 to +3°С, which contributed to a decrease in microflora activity and good preservation of organic material. The analysis of organic (spores and pollen) and mineral (phytoliths, diatom algae, and sponge spicules) microfossils from yedoma deposits of the Kolyma Lowland revealed a certain number of damaged forms, suggesting that some of them were destroyed during repeated freeze–thaw cycles that accompanied the transition of deposits into permafrost.

The purpose of this work is to perform a microscopic study of the biogenic components of the sample: organic (pollen, spores, wood and herbaceous detritus) and biogenic silica (phytoliths), to identify them as bioindicators of the differences between cryopedolites of Karginsky and Sartansky ages in the Kolyma Lowland, to detail reconstructions of vegetation and landscape of the time intervals considered, to analyze the conservation of microphytofossils, and to determine its potential use in paleoecological studies and in the Kolyma Lowland.

MATERIALS AND METHODS

The material for this study was a series of samples taken from cryopedolites of Karginsky and Sartansky age and burrow fossils in three key Upper Paleostocene outcrops in the Kolyma Lowland: Duvanny Yar, Stanchikovsky Yar, and Zeleny Mys (Sher, 1971; Kaplina et al., 1978, 1980; Gubin, 1984; Murton et al., 2015; Lopatina and Zanina, 2020a, 2020b). The sections under consideration are located between 68° and 69° N in the belt of modern pre-tundra sparse forests of the northern taiga subzone (Fig. 1).

Fig. 1.
figure 1

Location of the studied transects in the Kolyma Lowland. (1) Duvanny Yar, (2) Stanchikovsky Yar, (3) Zeleny Mys.

Material of cryopedolites used for the analysis of Karginsky deposits from the Duvanny Yar section was collected from profiles I PP (R-1310 and R-1332) and II PP (R-1333) of the Late Karginsky pedocomplex. Material from a fossil rodent burrow (R-1311) and cryopedolite above and below its level was studied. Burrows 1321 and 1075 were described from the same layer. Samples of cryopedolites from the Stanchikovsky Yar section were collected from profile I of the Late Karginsky pedocomplex (R-08-03), as well as from gopher burrow 1010 and cryopedolite located at the same level. Material from two burrows 1208 and 923 were studied from the Zeleny Mys section. The dates obtained for the buried soils and burrow material are shown in Fig. 2. The study of material from burrow fossils, in our opinion, is methodologically justified, as it was previously shown that the bottoms of both modern and fossil burrows were on the border of the seasonally thawed layer and rapidly transitioned to the perennially frozen state (Zanina, 2005). This was proved by numerous finds of rodent remains, including their soft tissues in fossil burrows (Gubin et al., 2003a; Faerman et al., 2017). Our earlier publications provide a partial palynological characterization of the deposits of these sections (Lopatina and Zanina, 2006, 2020a, 2020b). The Sartansky deposits were sampled mainly on the VI outlier of the Duvanny Yar section. A series of samples P-818 collected at intervals of 2 m and two isolated test samples P-6 and P-11A were taken. In order to identify possible differences in the chemical and paleobotanical characteristics of the material in the center and at the margins of the mineral block in the context of its complex genesis during the growth of ice veins, a series of Duv 54–59 test samples were collected. Sample P-819 was taken in the Stanchikovsky Yar section (Fig. 2).

Fig. 2.
figure 2

Sampling sites within the Duvanny Yar, Stanchikovsky Yar, and Zeleny Mys sections.

In the present work, we counted all microbiomorphs in the preparation, determined the taxonomic composition of the fossil flora, and quantified damaged microfossils. Owing to the mass occurrence and satisfactory preservation of palynomorphs, spore and pollen analysis is recognized as one of the leading paleoecological methods used for the study of Upper Paleopleistocene deposits of the Kolyma Lowland (Kaplina et al., 1978, 1980; Githerman, 1985; Vasil’chuk, 2003; Lopatina and Zanina, 2006; Zanina et al., 2011; Murton et al., 2015). Since the taxonomic composition of the fossil palynoflora in these deposits is rather uniform, the stratigraphic breakdown of the strata in question and vegetation reconstruction are usually made using quantitative characteristics. Such an approach is not always objective, considering the high activity of aeolian processes during the sedimentation in this region. The spectra may contain significant amounts of imported pollen, which obscures the role of some taxa in the vegetation composition, e.g., grass pollen, an insignificant amount of which does not reflect its participation in the local vegetation (Lopatina and Zanina, 2016).

Phytolithic analysis does not reveal all the plants growing on the territory, but only those that form stable forms with a characteristic morphology (conifers, grasses, sedges, a number of dicotyledonous grasses, mosses). Identification of taxa to family or genus is not always possible; this method, however, allows complexes characterizing a particular phytocenosis to be singled out (Goljeva, 2001). Phytoliths do not contain imported specimens, since plant debris is less mobile compared to spores and pollen.

The palynological and phytolithic analyses have their own fields of application, their limitations require certain corrections when interpreting the results obtained. The palynological analysis reflects a character of the regional vegetation, while the phytolithic analysis is focused on the local vegetation of habitats and is used to confirm the presence of certain taxa determined by the spore and pollen method.

The samples were processed using the separation procedure by Grichuk without treating the macerate with acetolysis mixture; macerates for palynological analysis were additionally treated with concentrated hydrofluoric acid to purify them from mineral particles (Pyl’tsevoi…, 1950; Paleopalinologiya, 1966). Palynomorphs were counted up to 200 grains, and the amount of damaged palynomorphs was recorded. The content of the so-called underdeveloped pollen, found in Quaternary deposits of Northeast Asia by Vasil’chuk (2005, 2007; Vasil’chuk and Vasil’chuk, 2018), was also estimated. The underdeveloped pollen is pollen with a smooth, thin exine, sometimes with a fine-grating, usually with three vaguely expressed furrows, without any other clear morphological features, presumably belonging to a motley grass. Since its systematic position is unclear (we may only speculate that morphologically similar underdeveloped pollen grains might have been produced by representatives of the families Ranunculaceae, Lamiaceae, and Rosaceae), its quantity was not included in the calculated sum, and its content per 200 specimens of spores and pollen was counted. The quantitative content of phytoliths was determined by counting them in five vertical rows of 24 × 24 mm glass. The percentage of each form of phytoliths with clear forms was calculated; indeterminate and corroded phytoliths were counted separately, and their ratio to the total sum of uncorroded forms was calculated. Images of microphytofossils were taken using a Tescan Vega 3 LSU electron microscope at the Collaborative Use Center of the Institute of Physicochemical and Biological Problems of Soil Science, Russian Academy of Sciences (Pushchino).

RESULTS AND DISCUSSION

The palynological analysis of the studied deposits showed that the samples from the Karginsky cryopedolites are dominated by pollen of grasses and shrubs, mainly Poaceae and Cyperaceae. Subordinate taxa of grasses and shrubs of various ecological confinement (Ericaceae, Caryophyllaceae, Asteraceae, Ranunculaceae, Fabaceae, Valeriana, etc.) are encountered as isolated specimens. The content of pollen of trees and bushes (Larix, Pinus s/g Haploxylon, Betula sect. Nanae, Duschekia) is usually 10–12%. An increased pollen content of this group (30% and higher) was encountered in the spectra 1310 АС 1 and 1333АС 1 from the Duvanny Yar section, 08-03 АС 2 from the Stanchikovsky Yar section, and 1208 and 923 from samples of the Zeleny Mys. Spores occupy a subordinate position in the spectra. The spectra from burrows are generally similar to those of cryopedolites, with a high content of pollen of cereals (more than half of all palynomorphs; test samples 1075, 1010) and Caryophyllaceae (test samples 1075, 1311), which is caused by the use of these plants by mammals (Fig. 3, Plate I). Underdeveloped pollen in the spectra of samples from the deposits corresponding to MIS 3 is rare or absent; only four samples contain 20–30 specimens per 200 palynomorphs. The amounts of redeposited pollen of Pinus, Picea, Tsuga, Podocarpus, Juglans, Myrica, and Ulmus and spores of Cyathea were found to be 1‒4 specimens per 200 palynomorphs.

Fig. 3.
figure 3

Content of spores and pollen in Karginsky deposits of the Duvanny Yar, Stanchikovsky Yar, and Zeleny Mys sections. See Fig. 2 for legend.

figure a

Plate I . Spores, pollen, fossil diatom algae, and sponge spicules from the Karginsky and Sartansky deposits of the Kolyma Lowland. (1, 2) Lycopodium; (3–5) Pinus s/g Haploxylon; (6) Cyperaceae; (7, 8) Poaceae; (9) underdeveloped pollen; (10) Ranunculaceae; (11) Sphagnum; (12, 15, 16, 17) diatom algae shells; (13, 14, 18) sponge spicules.

The pollen content of trees and shrubs in the spectra from Sartansky cryopedolites of the sections under consideration usually does not exceed 10%. About half of the total amount of all palynomorphs is pollen of Poaceae and Cyperaceae (approximately 10%). The taxonomic diversity and amount of pollen of grass is decreased compared to the Karginsky cryopedolites. The spore content is about 20% with the predominance of Selaginella rupestris (L.) Spring. (Fig. 4, Plate I). Underdeveloped pollen was observed in all spectra, with an average content of 40–50 grains per 200 palynomorphs. Redeposited palynomorphs (Pinus, Picea, Tsuga, Ulmus, Carya, and Osmunda) are rare (1–7 specimens per 200 palynomorphs).

Fig. 4.
figure 4

Percentage of spores and pollen in the Sartansky deposits of the Duvanny Yar and Stanchikovsky Yar sections. See Fig. 2 for legend.

The results of palynological analysis of deposits from the sections under consideration are given in publications Giterman (1985) and J. Murton et al. (2015). On the basis of his study of spores and pollen from the Karginsky deposits of the Kolyma Lowland, including the Duvanny Yar and Stanchikovsky Yar sections, Giterman concluded that there was an alternation of warm and cold periods. The spectra of warm intervals are characterized by a significant content (about 40%) of pollen from trees and shrubs (larch, dwarf pine, tree and shrub birch, and alder); Poaceae, Cyperaceae, and motley grass dominate in the grass and shrub pollen group, which constitutes approximately half of the total amount. The beginning of cold phases is dominated by pollen grains of various herbaceous plants; the spectra of the end of these phases are dominated by pollen of Poaceae, Artemisia, motley grass; Selaginella rupestris prevails among spores. Spectra from Sartansky deposits are dominated by Poaceae, Artemisia, and Selaginella rupestris.

Murton et al. (2015) indicated that the spore and pollen spectra from the Duvanny Yar section are characterized by different combinations of pollen from trees and other plants and are difficult to distinguish from each other. Thus, in Zone D (5–26 m above the river water level, 48–42 ka), spectra with a predominance of Poaceae and mixed grasses alternate with the spectra with considerable amount (up to 60%) of pollen of trees and shrubs (Pinus s/g Haploxylon, Larix, Betula, Duschekia). Spectra from Zone C (26–33 m above the river level, 33–30 ka) characterize deposits that accumulated before the maximum of the last glaciation. Poaceae, mixed grasses, and Selaginella rupestris dominate; pollen of trees and shrubs is less than 10%; and Larix is rare. The spectra of zone B (33–37 m above the river level, 25–17 ka), corresponding to the maximum of the last glaciation, contain 20–60% of pollen of trees and bushes, 20–60% of grass (Poaceae and motley grass), and 10–40% of Selaginella rupestris. The spectra of this zone differ from those of zones D and C by the lower diversity of taxa and the absence of larch.

A palynological analysis of the Karginsky cryopedolites, burrows from these deposits, and the Sartansky cryopedolites from sections of the Kolyma Lowland, taking into account data obtained in earlier studies by other authors, revealed that the spore-pollen complexes from these deposits differ little from each other. It is rather difficult to establish differences between them because of approximately identical taxonomic composition and variations in the content of palynomorphs. Thus, the percentage of tree and shrub pollen in the Karginsky cryopedolites is 10–15%, with individual spectra containing less than 10% or more than 30%. Tree and shrub pollen in spectra from Sartansky deposits is usually less than 10%, but there are also spectra containing up to 16%. The low content of pollen of this group in the spectra from the Sartansky deposits is pointed out by Giterman (1985), while Murton et al. (2015) pointed out that the spectra from the strata dated to 25–17 ka are characterized by a high content of pollen of this group, which was, in their opinion, imported. Thus, the amount of tree and shrub pollen is not a criterion for subdividing these strata. The taxonomic composition of pollen of this group in the deposits under consideration is similar. However, it should be emphasized that the spectra from the Sartansky deposits do not contain larch pollen, single grains of which were detected in most spectra from the Karginsky deposits. The pollen of grass and shrubs in the spectra from the deposits corresponding to MIS 3 and MIS 2 is dominated by Poaceae. The codominants of this group in the spectra from the Karginsky cryopedolites are Cyperaceae, the pollen content of which is often similar to that of Poaceae, and in some samples even higher, which is, in our opinion, the difference between the spore and pollen spectra from these deposits. The noticeable amount of sedge pollen in the spectra from the Karginsky deposits was determined by Giterman (1985). In addition, most of the spectra from these deposits are characterized by the taxonomic diversity of grass pollen. The Poaceae family is monodominant in the samples from the Sartansky cryopedolites, constituting more than half of the total sum of palynomorphs, while the pollen content of Cyperaceae is usually less than 10%. The amount of grass pollen is somewhat reduced here. The composition of spores in the deposits under consideration is similar, but their amount in the Sartansky deposits is higher owing to Selaginella rupestris (up to 20%). Thus, according to the palynological analysis, the cryopedolites of Karginsky and Sartansky ages have the following differences: the spectra from the Karginsky deposits commonly contain isolated grains of Larix pollen; they usually have a more diverse composition owing to grasses; the codominant of Poaceae in this group are Cyperaceae; the Sartansky spectra contain more spores owing to Selaginella rupestris.

In respect to the analysis of underdeveloped pollen in the studied spectra, the following should be noted. Usually, these are small (on average 15 µm) pollen grains with three indistinct grooves, without any clear morphological signs of exine structure, usually smooth and thin. Presumably, this is immature grass pollen that did not have time to evolve because of the disturbance in vegetation. The analysis of undamaged fossil pollen with atypical morphological structure are given, in particular, in the works of Ananova (1966), Dzyuba (2007), Vasil’chuk (2005, 2007); Vasil’chuk and Vasil’chuk (2018), and Evstigneeva (2017). According to these researchers, the appearance of pollen of this kind resulted from abrupt, extreme changes in natural conditions, such as fires, increased ultraviolet radiation, intense volcanic activity, lower temperatures, lack of humidity, etc. Ananova (1966) noted the presence of this pollen in deposits associated with glaciation; the pollen composition of the spectra containing this type of pollen is usually poor. She cited the study of modern pollen from immature inflorescences: it was usually crumpled, had weak or, on the contrary, dense coloration, unclear morphology, and exine stratification. According to (Vasil’chuk, 2005) and (Vasil’chuk and Vasil’chuk, 2018), palynospectra from Upper Neopleistocene deposits of Northeast Asia may contain underdeveloped pollen in the amount of 60–80%. It is rare in modern surface samples, because it is rapidly destroyed by microfauna. The maximum content of underdeveloped pollen is recorded in surface spectra from the hypoarctic tundra, where plants produce pollen in appreciable amounts, but pronounced temperature variations during the meiosis phase have a negative effect on its formation. Such disturbances in the region under consideration may be associated with early and rapid formation of snow cover, late snow melting or the snow falling on flowering plants, sharp temperature drops, floods occurring during the growing season, etc. For the underdeveloped pollen to survive, it should almost immediately get into the conditions of low temperatures and freeze (Vasil’chuk, 2005, 2007). The authors of this article detected isolated specimens of underdeveloped pollen only in a few spectra of surface samples from the Kolyma Lowland.

The amounts of underdeveloped pollen detected in the spectra of samples from the Karginsky deposits of the Kolyma Lowland are usually insignificant, while Sartansky deposits mostly contain a substantial amount of such pollen (40–50 specimens per 200 palynomorphs). In our opinion, the presence of underdeveloped pollen in the studied spectra is very important since it indicates both well-developed grass cover, repeated adverse conditions during the growing season, and the low temperatures of the surface soil layer contributing to the rapid freezing of palynomorphs and their transition into the perennially frozen state. Destruction of underdeveloped pollen occurs first of all owing to the alteration of its biochemical properties, controlling the strength of its shell (Dzyuba, 2007).

The content of damaged palynomorphs in MIS 3 cryopedolites from the studied outcrops of the Kolyma Lowland averaged 8–9%; samples with both relatively high (14%) and low (3%) contents of damaged palynomorphs were recorded. The number of damaged spores and pollen in MIS 2 formations is lower and ranges from 1 to 5%. Spores and pollen with physical-type damage (fractures and cracks) predominate in the studied Upper Pleistocene strata; palynomorphs with chemical and biotic damage (thinning of the exine, caverns), mainly resulting from chemical effects of microbes, are rare and sporadic. This is probably due to the prevalence of low temperatures during the year and suppressed microbiological activity within the cryolithozone. The pollen of Pinus s/g Haploxylon with breaks and cracks prevails among damaged pollen from Karginsky formations; representatives of the Poaceae family are noted in subordinate amounts. Pollen of Larix, Betula, Cyperaceae, and Caryophyllaceae and spores of Selaginella rupestris with damage of this type were recorded rarely and sporadically. Isolated grains of broken pollen of Pinus s/g Haploxylon and Poaceae and underdeveloped pollen were found in samples from the Sartansky deposits; pollen of Betula, Caryophyllaceae, and Asteraceae and spores of Selaginella rupestris were recorded sporadically. Musina and Sahibgareev (1983) analyzed the preservation of palynological remains in deposits and found that large-sized palynomorphs and pollen of Gymnospermae are most susceptible to destruction of the physical type. It is possible that these two reasons are responsible for the noticeable content of damaged pine pollen in the deposits. However, it cannot be ruled out that some of it was imported and may have been damaged during transportation. The stable presence of isolated broken grains of cereal pollen can probably be explained by the fact that this family is dominant in the spectra, so in this case we have a large number of grains for statistical analysis. Almost all pollen grains of cereals and sedges are crumpled.

The relatively large amount of pollen with ruptures and cracks in the Karginsky strata was probably caused by the following. The period correlated with MIS 3 is characterized by seasonal temperature rises, during which there was thawing of ice veins or stagnation of surface water on the denser underlying strata. The pollen entering the sediment swelled under the conditions of increased hydromorphism, and then, with the onset of a cold period, it dried up and froze out; repeated cycles like that caused formation of ruptures and cracks on the spore and pollen shells. The temperature regime of MIS 2 was more stable, the pollen that fell into the sediment was exposed to low temperatures, froze, and, accordingly, was better preserved.

The phytolith content in the material of Karginsky cryopedolites varies from a few dozen to several hundred varieties. Most of them are elongated smooth forms formed in the columnar parenchyma of leaves and attributed to dicotyledonous grasses, although some of them could have also been formed by monocotyledonous plants. Assembling them into one group was justified by the fact that these forms are encountered in almost all plants of the humid zone; other forms are characteristic of monocotyledonous plants, in particular, cereals, while dicotyledonous plants are usually devoid of such other forms. It cannot be excluded that the predominance of elongated smooth phytoliths is due to their structure: a solid cylindrical shape and absence of perforations and inner cavity. Such a configuration allows them to be better preserved in deposits, being less exposed to destructive influences. These forms are poorly informative and are not used as a diagnostic, but in our case, they emphasize the significant role of herbaceous associations (meadow grasses and cereals) in the vegetation. The content of lanceolate and various trapezoidal forms (threelobate, lobate asymmetrical, polylobate symmetrical, globular), characteristic of sedges and grasses growing on moist soils, is significant. This is also confirmed by the constant presence of moss tissue (sphagnum and polytrichum). It should be noted, however, that specific conical sedge phytoliths are very rare in Karginsky deposits, probably having been destroyed by fossilization. The echinate phytoliths characteristic of xerophytic vegetation were found in a small amount, and the blocky polyhedral particles characteristic of the Pinus genus are isolated (Fig. 5, Plate II).

Fig. 5.
figure 5

Contents of phytoliths in the Karginsky and Sartansky deposits of the sections of the Duvanny Yar and Stanchikovsky Yar. See Fig. 2 for legend.

Plate II.
figure 6

Shapes of phytoliths from the Karginsky and Sartansky deposits of the Kolyma Lowland. (1, 6) Elongated smooth form; (2) elongated spiky form; (3, 5) lanceolate-shaped with a spike; (4) lanceolate-shaped with a wide base; (7, 13–15) trapezoidal lobate forms; (8–10) rounded forms; (11) elongated wavy form; (12) blocky polyhedral form; (16–20) corroded forms of phytoliths.

The content of phytoliths in Sartansky cryopedolites is insignificant (up to 20 specimens) with a low variety of morphological forms; they are mostly small (10–20 μm). The low content of phytoliths may be related to the weak accumulation of amorphous forms of silica in plant organs and/or to weak humification of plant debris and the dominance of detritus formation in the accumulation of strata. Most specimens are so deformed that it is impossible to diagnose their original shape. This probably indicates both scarcer and more suppressed vegetation and a more pronounced continental climate during the formation of these strata. As in the spectra from the Karginsky deposits, the predominance of elongated smooth forms and the stable presence of moss remains were observed here; the number of forms characteristic of plants of humid habitats, common to spectra corresponding to MIS 3, however, is significantly reduced. Thus, lanceolate-shaped forms are absent here, trapezoidal forms are represented exclusively by lobate asymmetric forms and are not found in all samples. Elongated wavy forms characteristic of cereals, absent in the material from the Karginsky deposits, were found here. Phytoliths of conifers were detected in the spectra of two samples (Fig. 5, Plate II).

Corroded and highly pitted phytoliths were recorded in the samples studied; their content in the samples from the Karginsky strata and from the Sartansky deposits were 2–9% and 7–11% of the total amount, respectively. The study revealed several types of damage to phytoliths: thinning of small elements (stems, spines, etc.), corrosion of the massive body (the appearance of caverns and through perforations), scale-like damage (formation of small chips, tightly adjacent to each other, creating the impression of scales), and chipping resulting from fractures (Plate II). It should be noted that damage was also detected on other siliceous microfossils, sponges, and diatom shells, often found in the studied material. Sponge spicules of rounded shape with a hollow central channel in the amount of 10–20 specimens were found in the samples from formations correlated with MIS 3 and MIS 2 (Plate I). Approximately half of the specimens were broken, and a number of specimens with a scale-like corrosion were found. Diatom shells were detected only in the samples from the Karginsky strata; they are an indicator of increased hydromorphism. The amount of damaged forms with thinned shells (especially in the area of areolae), cracks, and destruction of walls is up to 30%. It can be assumed that the destruction of siliceous microphytofossils in the permafrost zone was indicative of the environment-seasonally thawed layer with alternating cycles of freezing and thawing. No corroded phytoliths were detected in buried deposits, which may be related to their short residence in the seasonally thawed layer and to the weak destruction of plant material during fossilization. The buried Late Pleistocene soils contain 1–2% of corroded phytoliths, which may be due to both the milder, humid conditions of their fossilization and the large total amount of forms (Lopatina and Zanina, 2020).

The samples from the strata under consideration were analyzed for plant detritus sized 10–200 µm. Its usual component in the Karginsky deposits are specific forms with fringed pores, characteristic of larch wood. Significant amounts of moss detritus and epidermal remains of Cyperaceae, Poaceae, and Ericaceae, differing in stomatal complexes, were found. The composition of plant detritus from the Sartansky cryopedolites is poorer; it contains fossil cereals and mosses.

Khasanov (1999) and Gubin et al. (2001, 2003a) presented the results of the study of coarse plant remains from burrows of gophers in the sections under consideration. Remains of woody vegetation (Larix cajanderi Mayr. wood) were found only in burrows of the Zeleny Mys section; shrubs (fragments of twigs and leaves of willows) are common in the litter. Psychrophytes (plants growing on moist and cold soils) are most frequent among grasses in the complex of carpological remains, with the predominant species being Potentilla nivea L., Bistorta vivipara (L.) S.F. Gray, Poa arctica R. Br., Rumex arcticus Trautv, and different types of sedges, e.g., Carex bonanzensis Britt., C. vesicata Meinsh., and C. norvegica Retz. Seeds of steppe plants such as Poa attenuata Trin. and P. botryoides (Trin. ex Griseb.) Roshev, Silene stenophylla Ledeb, pioneers of dry habitats (Draba, Plantago canescens Adams), and pioneers of wet habitats (Ranunculus repens L.) are also present in a significant amount. These species are found in modern tundras of northern Yakutia and allow the vegetation cover in the vicinity of the burrow to be identified as tundra with participation of disturbed habitats occupied by pioneer and steppe vegetation. The only fossil burrow of a small rodent from Sartansky deposits is described in the Duvanny Yar section. It contained fragments of Dryas sp., lingonberry Vaccinum vitis-idaea L., and polar willow Salix polaris Wahlenb and isolated seeds of dry habitat pioneers Draba sp., Plantago canescens, Taraxacum lateritum Dahlst, and steppe plants, Allium strictum Schrad., Poa botryoides, and psychrophyte Potentilla nivea, as well as fossil mosses and lichens. The taxonomic composition of the above macrofossils indicates the existence of a low-productive steppe community in the vicinity of the burrow against the background of tundra vegetation (Gubin et al., 2003a).

The analysis of spores, pollen, phytoliths, and plant detritus, taking into account the results of the study of carpological material, indicates occurrence of a tundra landscape with the prevalence of herbaceous groups (cereals, cereal-sedge grasses, and cereal-grasses) during the MIS 3 interval. The predominance of cereals in the vegetation was confirmed by the analysis of phytoliths, plant detritus, and the contents of fossil burrows, where ample remains of these plants were discovered. The predominance of psychrophytes in the complexes of carpological remains and the presence of diatom algae shells and forms typical of moist cenoses among phytoliths indicate the wide propagation of tundra meadows on moist soils. They may have developed as a result of an increase in the thickness and moisture content of the seasonally thawed layer due to thawing of ice wedge upon some warming of the climate. The presence of grass pollen in the spectra studied is rare. Analysis of palynological remains from subrecent spectra of the Kolyma Lowland showed that the pollen of grasses and shrubs in them does not always reflect the diversity of local vegetation, and its amount may not reflect the adequate presence of some families (e.g., Onagraceae, Lamiaceae, Ranunculaceae Liliaceae, etc.). This may be due to the low pollen productivity of these plants under severe climatic conditions and their transition to vegetative reproduction (Lopatina and Zanina, 2016). We can assume, therefore, a significant role of diverse grasses and shrubs in the composition of vegetation of the Karginsky and Sartansky time. The ecological interpretation of grass pollen in the spectra is difficult, since the families of Caryophyllaceae, Asteraceae, Fabaceae, Ranunculaceae, Polygonaceae, etc., are represented by a large number of species of different ecological confinement. Their presence is, however, generally consistent with this reconstruction. The diverse composition of phytoliths reflects the rich vegetation. Sedges, whose pollen is subdominant in the spore and pollen spectra, grew in waterlogged areas, and finds of their fruits are common in burrows. At the same time, melting of ice wedge triggered intense slope and erosion processes; newly formed screes were populated by pioneer vegetation, Draba and plantain, replaced in the process of successional changes by steppe, cereal and motley grass associations. The presence of forms with spicules, characteristic of xerophytic, including steppe, flora, in the complex of phytoliths confirms this conclusion.

Tree and shrub pollen in the spectra under consideration is usually 10–12%, sometimes reaching 30% or even more. It can be assumed that the variations, although insignificant, in the ratio of vegetation groups and major and minor taxa in the spectra of samples from the Karginsky cryopedolites reflect climatic fluctuations during the Karginsky time and a periodic increase or, conversely, a decrease in the role of sparsely vegetated larch forests in the territory occupied mainly by grass-shrub tundra vegetation. Only Larix and Salix were identified in the complex of macrofossils. The absence of plants of this group cannot be explained by mineralization as a result of burial and soil formation, since the strata under consideration contain a significant amount of fossil herbaceous and moss vegetation that normally undergoes humification and mineralization more easily than wood. Paleosoil studies did not reveal any signs of forest soils in the Karginsky cryopedolite layers (Gubin and Zanina, 2013). Taking into account very poor preservation of larch (Larix cajanderi Mayr) pollen and its low migration ability, a single find of its wood in the burrow of Zeleny Mys and remains of its wood found in plant detritus may indicate the presence of small patches of larch forest in the reconstructed landscape. The wood of this genus found in the burrows of the Arctic ground squirrel, an inhabitant of open spaces, may indicate only sparse larch forest in its habitat. Given the high productivity and migration ability of pine, it is usually difficult to estimate its role in the vegetation cover. A significant part of it found in the subrecent spectra of the Kolyma Lowland was proved to be imported (Lopatina and Zanina, 2016). Karginsky and Sartansky deposits contain many broken forms among the specimens of Pinus s/g Haploxylon pollen, which, in particular, may be a consequence of transportation and suggests its extraneous origin. However, findings of conifer phytoliths, which usually have good preservation owing to the significant amount of resinous substances in the needles, preventing their rapid destruction, indicate the participation of dwarf pine in the vegetation, possibly as undergrowth in larch forests or small thickets in valleys.

Spores occupy a subordinate position in the spectra; however, large amount of moss shoots were found in in burrows, and the phytolith spectra steadily contain their characteristic forms as well as tissues, suggesting that the moss cover was quite well developed.

The microphytofossil analysis indicated that the tundra plant communities, predominantly cereals and grasses, occurring on the territory in question during the MIS 2 interval were more impoverished compared to those of MIS 3. A significant amount of underdeveloped pollen, presumably of motley grass, discovered in the Sartansky deposits suggests development of meadows and, possibly, unfavorable conditions of the growing season that prevented full formation of pollen grains. A significant reduction in the number of forms characteristic of wet cenoses and the presence of pollen and carpological remains of Asteraceae, Caryophyllaceae, Fabaceae, Potentilla, and Draba among the phytoliths indicate a wide distribution of dry, well-drained areas. The abundance of Selaginella spores in the spectra indicates the presence of dry rocky screes on the slopes.

CONCLUSIONS

The study of microphytofossils as bioindicators of plant communities from the Karginsky and Sartansky deposits of the Kolyma Depression made it possible to obtain additional information on the genesis of the strata studied, to more fully characterize the composition of the vegetation and landscape, and to detail the environment reconstruction.

The composition of palynomorphs in samples from the Karginsky and Sartansky deposits is essentially similar; grass and shrub pollen predominate with prevalence of Poaceae. Their differences, which should be taken into account in stratigraphic breakdown, are as follows: more diverse composition of grasses, participation of Cyperaceae as codominant, presence of single Larix pollen in the spectra from Karginsky deposits, and increased number of spores due to Selaginella rupestris and underdeveloped pollen in Sartansky formations.

The typical inhabitants of field and meadow grasses and cereals predominate in the complex of phytoliths from the Karginsky strata; a significant amount of sedges and cereals growing on moist soils and phytoliths of mosses is present, and a single content of forms characteristic of xerophytic flora and conifers is also noted. The content of phytoliths in Sartansky cryopedolites is lower compared to that in Karginsky strata, their sizes are smaller, and the diversity of morphotypes is lower, which indicates the existence of relatively sparse and depressed vegetation in that period. There is a stable presence of phytoliths of dicotyledonous grasses and moss remains; however the number and diversity of forms characteristic of moisture-loving plants, common to spectra from strata corresponding to MIS 3, are significantly reduced. Phytolithic analysis confirmed the presence of some plants, in particular, mosses, numerous shoots of which were found in burrows, although their participation in palynological spectra is low, as well as motley grass, determined in the spectra as single grains and underdeveloped pollen. It has been established that the plant groups include the Siberian dwarf pine, the pollen of which can be regarded as imported in the spectra in the absence of this genus in the complex of micro- and macrofossils.

The known morphology of modern pollen formed under unfavorable environmental conditions and processes associated with its formation made it possible to identify underdeveloped pollen, presumably belonging to the motley grasses in the spectra. Its ubiquity suggests well-developed grass cover, predominantly unfavorable conditions of the growing season, and low temperatures of the surface soil layer, contributing to its preservation. The presence of pollen of this kind in the fossil spectra can be used as an additional criterion in the reconstruction of the paleoclimate and landscapes.

The composition of spores, pollen and phytoliths from the Karginsky strata suggests that a variety of landscapes existed on the considered territory during the given time interval: communities of wet and dry tundras, larch sparse forests, and tundra bogs, as well as habitats occupied by pioneer steppe vegetation. The formation of cryopedolites during the whole Sartansky stage took place in more uniform, cold and dry conditions. The data on microbiomorphs from these deposits confirm the active participation, up to complete dominance, of herbaceous, mainly cereal associations in the complex with a moss cover during this period.

Analysis of the preservation of palynomorphs from MIS 3 and MIS 2 deposits of the Kolyma Lowland has shown that the chemical and biological damage caused by the action of microbes is scarce here. This may be explained by the predominance of low temperatures during the year and low microbiological activity. The content of palynomorphs with physical damage, caused, most likely, by alternating desiccation and wetting, usually does not exceed 10% in Karginsky strata and 5% in Sartansky strata. The period of accumulation of Karginsky deposits is characterized by a more pronounced manifestation of summer temperature increase, accompanied by melting of ice veins and stagnation of water in the depressions of the terrain. The increased humidity made palynomorphs deposited in the sediment swell, and with the onset of the cold period, they froze and dried out; repetition of such cycles resulted in the formation of lacerations and cracks on the shells of spores and pollen. The conditions of the Sartansky period were more stable; spores and pollen were better preserved at low temperatures. This is also indicated by the abundant finds of underdeveloped pollen in these deposits. The data obtained add to the ideas on the character of plant material transformation during cryogenesis and may prove useful for interpreting data on the microphytofossils from Pleistocene deposits of the cryolithozone.