Rapid thermokarst evolution during the mid-Holocene in Central Yakutia,Russia

Abstract

The reconstruction of Holocene thermokarst landform evolution is important to understand the potential impact of current global climate change on permafrost regions. A multi-proxy approach was applied to analyse the sedimentological and biogeochemical characteristics as well as pollen and lacustrine microfossils of a core profile drilled in a small pingo within a large Central Yakutian thermokarst basin (alas). Age–depth modelling with macrofossil ¹⁴C ages reveals high thermokarst deposit sedimentation rates and a complete thermokarst sequence spanning about 900 years during the mid-Holocene between ~6750 and 5870 cal. yr BP. In total, three stages of thermokarst landscape evolution have been identified. Thermokarst processes were initiated at ⩽6750 to 6500 cal. yr BP. Terrestrial conditions changed quickly to lacustrine conditions, and a thermokarst lake rapidly emerged and grew to an estimated size of 120–600 m diameter and 7.5–15 m depth during only ~150 years between ~6500 and 6350 cal. yr BP. The decline of thermokarst processes and lake decrease may have been affected by local hydrological conditions between ~6350 and 5870 cal. yr BP but ceased completely after 5870 cal. yr BP, likely due to climatic changes. Clear evidence for long-lasting and stable lacustrine conditions was not obtained. The study emphasises that short-term warming led to very active permafrost degradation and rapid but locally variable modification of alas and thermokarst evolution.

Keywords

climate change Eastern Siberia Holocene thermal maximum palaeolimnology permafrost thermokarst deposits

Introduction

Climate-controlled thaw of ice-rich permafrost (i.e. thermokarst) causes substantial environmental changes, for instance, the increased release of greenhouse gases such as methane and carbon dioxide (e.g. Walter Anthony et al., 2016) or hydrologic alterations in tundra and taiga lowlands (Bring et al., 2016; Liljedahl et al., 2016). It may also pose a risk for infrastructure stability and local livelihoods (Istomin and Habeck, 2016; Streletskiy et al., 2015). Therefore, thermokarst studies are of relevance to understand future changes and related challenges in arctic and subarctic regions. While remote-sensing studies provide the most common tool to understand ongoing changes in thermokarst landscapes (e.g. Grosse et al., 2013), sedimentological and palaeoenvironmental analyses are necessary to look into past thermokarst dynamics and related hydrological and biogeochemical responses (e.g. Bouchard et al., 2016; Gaglioti et al., 2014). The mid-Holocene climate optimum (roughly between 6700 and 5000 cal. yr BP) in Central Yakutia (CY; Andreev et al., 2002; Nazarova et al., 2013) is of special interest because climate conditions and related environmental changes are at least partly comparable with those expected for the 21st century (Fradkina et al., 2005; Intergovernmental Panel on Climate Change (IPCC), 2013; Monserud et al., 1998).

Large-scale climate forcing triggers the general pattern of periglacial landscape evolution and thermokarst processes in permafrost regions (Grosse et al., 2013). However, it may also depend on temporally and spatially limited non-climatic factors (e.g. vegetation cover destruction, local erosion, forest fire and land-use change). Rising air and ground temperatures disturb the thermal equilibrium at the upper permafrost boundary and subsequently lead to permafrost thaw, ground-ice melt and surface subsidence. Typical permafrost degradation processes, thermokarst and thermo-erosion result in certain landforms such as large lake-filled basins and valley structures, respectively (e.g. Soloviev, 1973). Our classical understanding of thermokarst was developed by Soviet scientists using the example of CY alas landscapes (e.g. Katasonov, 1979; Soloviev, 1973) and expertly summarised by Czudek and Demek (1970). Most authors suggest that thermokarst evolution in eastern Siberia started around the Pleistocene-to-Holocene transition (e.g. Grosse et al., 2007; Katamura et al., 2006).

The formation of thermokarst lakes and basins generally starts with ground-ice (ice-wedge) melt and surface subsidence, creating a hummocky surface (i.e. baydzherakhs and thermokarst mounds) of remaining sedimentary ice-wedge polygon centres. Water ponds in troughs above melting ice wedges and polygon-centred ponds subsequently coalesce, forming shallow lakes. Once the water depths exceed lake-ice thickness, liquid water exists year-round, promoting the formation of unfrozen deposits (i.e. taberal deposits and taliks) below the lake. The surface subsidence below developing thermokarst lakes results in larger lake-filled thermokarst basins with steep slopes. Alases form when a lake drains or evaporates, leaving a deep depression with a residual lake. Further basin deepening might be prevented if ice-poor ground forms an insulating layer at a later stage or if underlying sediments have low ground-ice contents (Morgenstern et al., 2011; Ulrich et al., 2010). Thermo-erosion along the slopes and mass wasting cause lateral growth of thermokarst basins (Séjourné et al., 2015). Pingos (i.e. ice-cored mounds and bulgunnyakhy) usually evolve on the exposed alas floor by permafrost and ground-ice aggradation when the talik refreezes, allowing frozen ground to rise and an ice core to form.

Traces of periglacial landforms preserved in permafrost deposits indicate the interplay between past climate and landscape settings. Besides climate control, on-site periglacial morphology, hydrology and vegetation alter permafrost regimes and must be taken into account when interpreting late Quaternary permafrost chronologies. Thermokarst deposits preserved in late Quaternary permafrost have been mostly studied in palaeoenvironmental research in arctic Siberia, Alaska and Canada and aligned to different warm stages, but the subarctic (boreal) regions of CY in eastern Siberia have been less considered in this context so far. Most studies capture late glacial to Holocene thermokarst lake sediments that are widespread in northern Siberia (Biskaborn et al., 2013a, 2013b; Schleusner et al., 2015), Alaska (Farquharson et al., 2016; Lenz et al., 2016) and Canada (Fritz et al., 2012; Lenz et al., 2013).

However, there is a knowledge gap about the timing of Holocene lake expansion and lake decrease or drainage dynamics in permafrost regions in general and in CY in particular. The dynamics of thermokarst lakes and corresponding permafrost degradation and aggradation as well as peat accumulation and development and refreezing of taliks result in specific cryolithostratigraphy, sedimentology and physico-chemical properties of alternating lacustrine and terrestrial deposits (Bouchard et al., 2016; Katasonov, 1979; Morgenstern et al., 2013). These sedimentary archives have previously been less considered in CY (Katasonov, 1979); multi-proxy approaches with regard to differing thermokarst dynamics in different geomorphological conditions are rare.

Here, we present a multi-proxy record from a small pingo within a large CY alas basin. Our study highlights rapid thermokarst lake and alas evolution over time and aims to reconstruct related local and regional palaeoenvironmental conditions including climatic, permafrost, geomorphologic as well as floral and lacustrine faunal dynamics. The mid-Holocene periglacial landscape dynamics are, moreover, discussed and compared with modern thermokarst development and driving processes.

Study site

CY is bounded by the large Lena and Aldan rivers (Figure 1). The active layer (i.e. the seasonal thaw layer) reaches depths of 0.5–2.0 m, depending on vegetation cover, topography, soil type and subsurface water content. Large parts of CY are covered by late Pleistocene Yedoma Ice Complex (IC) deposits (up to 50 m in thickness) containing ~70–80% ice by volume, mainly consisting of huge syngenetic ice wedges (Soloviev, 1959, 1973). Alases and thermokarst lakes are widespread features in this region, originating from late Pleistocene to Holocene permafrost degradation. The region has a continental climate with low annual precipitation (223 ± 54 mm), mean annual air temperature of −9.8 ± 1.8°C, and high annual temperature amplitude of >60°C (1910–2014 average; National Oceanic and Atmospheric Administration (NOAA), National Climatic Data Center, https://www.ncdc.noaa.gov). Summer (May–September) precipitation is more than twice as high as than winter (October–April) precipitation; annual evaporation substantially exceeds annual precipitation, leading to distinctly arid conditions (e.g. Ulrich et al., 2017). Deciduous coniferous larch forests typically dominate CY vegetation. The major forest-forming species are Larix cajanderi, L. gmelinii, L. sibirica and L. czekanowskii with inclusions of Pinus silvestris communities (Troeva et al., 2010). Alases form grassland islands within the taiga forest and are dominated by halophytic steppe-like and bog plant communities, depending on hydrologic and usually saline edaphic conditions (Desyatkin, 2008). Traditionally, grassland areas within alases are extensively used for haying and pasture.

Figure 1.

Regional setting of the KB7 drilling location. (a) The Central Yakutian region within the zone of continuous permafrost (topography compiled using data from the ESA DUE Permafrost project (https://geo.tuwien.ac.at/permafrost/)). (b) The Central Yakutian thermokarst and alas landscape (Landsat 8 close-up, July 2013). (c) The Khara Bulgunnyakh (KB) alas as part of a larger alas system. Alas basins are clearly distinguishable from the taiga forests in the false-colour satellite image by the pinkish colour (Pleiades close-up, September 2012). (d) Photograph of the profile and drilling location KB7 on a small pingo within the KB alas.

Our study site (61.832813°N, 130.642854°E) is south-west of the Khara Bulgunnyakh (KB) alas, which is part of a larger alas system (Figure 1). The study site KB alas is about 1500 m in diameter and about 10–12 m deeper than surrounding Yedoma uplands. In total, three large pingos within the KB alas reach heights of 8–12 m and diameters of 50–100 m. They are surrounded by shallow lakes. The alas system was already investigated by a Russian research group in the 1970s (Katasonov, 1979). They suggest that its evolution started around 12,000–10,000 years ago as they found stalks of reed grass dated to 9120 ± 200 cal. yr BP in the north-west of the alas system (Ébé alas; Figure 1) below 3.0 m depth. The drilled KB7 pingo is about 2 m high and about 30 m in diameter. The pingo top is about 128 m a.s.l.

Material and methods

Permafrost coring and sampling

The drilling was undertaken in August 2013. First, a pit was dug down to the permafrost table at 110 cm depth below surface (dbs). In total, three sediment samples were taken from the active-layer pit according to the described soil horizons (see text Supplementary Material, available online). The frozen core was then drilled from the permafrost table to 382 cm dbs. The core segments (each ~20–30 cm long) were cleaned, described, photographed and continuously sampled at <10–20 cm intervals according to cryolithological properties following French and Shur (2010). In total, 19 sub-samples were packed in plastic bags and transported back for laboratory analyses.

Sedimentology and biogeochemistry

Sedimentological and biogeochemical parameters were analysed for all sediment samples after freeze-drying. Grain-size analyses were carried out using organic-free (treated with 35% H₂O₂) sub-samples using a laser particle analyser (Beckmann Coulter LS 200). The mass-specific magnetic susceptibility (MS) was analysed using a Bartington MS2B dual-frequency sensor. The values are expressed in SI units (10⁻⁶ m³/kg). Total carbon (C_total) and total nitrogen (N_total) were measured with a Vario EL cube elemental analyser and are given as weight percent (wt%). The total inorganic carbon (C_inorg) content was taken from carbonate measurements conducted using the Scheibler method on an Eijkelkamp calcimeter apparatus by continuously adding 4N HCl to a sub-sample until the reaction ceased. The total C_inorg content was then calculated based on the volume of CO₂ produced during the Scheibler reaction. The total organic carbon (C_org) content was determined for each sediment sample by subtracting C_inorg from C_total. The C/N ratio is expressed as C_org/N_total. Stable C_org isotope ratios (δ¹³C) were measured on carbonate-free sub-samples with a Delta V Advantage isotope mass spectrometer (ThermoFisher Scientific^™). Values are expressed in delta per mil notation (δ, ‰) relative to the Vienna Pee Dee Belemnite (VPDB) Standard.

X-ray fluorescence analysis

In order to obtain quantitative information about the elemental composition of the studied sediments, all samples were analysed by x-ray fluorescence (XRF) spectrometry. Analyses were carried out using the energy-dispersive polarisation-XRF (EDP-XRF) SPECTRO XEPOS (SPECTRO Analytical Instruments Ltd) analyser. Samples were homogenised with a vibration mill, mixed with a wax binder (CEREOX Licowax at a ratio of 4 to 1) and pressed into 32-mm pellets. Measurements were conducted in a helium gas atmosphere; the contents of all elements from sodium to uranium were simultaneously determined and adjusted to sample weight.

In total, six elements were chosen for interpretation of thermokarst (lake) processes. Zircon (Zr), rubidium (Rb) and strontium (Sr) are interpreted as signals for detrital input and grain-size fluxes. Since Sr often substitutes for Ca in feldspar and Rb for potassium (K) in illite clay minerals, the Sr/Rb ratio was used as an additional proxy for grain-size changes (Biskaborn et al., 2013b; Kalugin et al., 2007). Variations in sulphur (S) are utilised to estimate organic inputs (Bouchard et al., 2011) but may also be interpreted to reflect changes in salinity (gypsum formation) and the presence of iron sulphides (Biskaborn et al., 2012; Siegert, 1979, 1987). Finally, the ratio of iron (Fe) to manganese (Mn) changes with the degree of oxidation during the accumulation of sediments in lakes and is thus used for estimating palaeo-redox conditions (e.g. Biskaborn et al., 2013b; Davison, 1993). Elements and elemental ratios that were additionally used for interpretation and discussion are given in Table S1 of Supplementary Material (available online).

Radiocarbon dating, age–depth modelling

Terrestrial plant remains and wood fragments from five samples and two bulk sediment samples from selected depths of the KB7 core and active-layer pit were radiocarbon dated at the Curt-Engelhorn-Centre Archaeometry (Mannheim, Germany) using the accelerator mass spectrometer (AMS) technique (Table 1). The conventional ¹⁴C ages were calibrated using the IntCal13 data set (Reimer et al., 2013) and SwissCal 1.0 (L. Wacker, ETH Zürich). The ¹⁴C ages are normalised to δ¹³C = −25‰ (Stuiver and Polach, 1977). Calibrated radiocarbon ages are given as cal. yr BP. The age–depth relation was constructed with Bacon 2.2 routines (Blaauw and Christen, 2011) using R software (R Core Team, 2012) based on the five calibrated radiocarbon dates obtained from wood and plant remains. The median ages were modelled for each centimetre along the core. The radiocarbon dates from the bulk sediments were not used for age–depth modelling.

Table 1.

Radiocarbon dating results for the KB7 core profile and from a pingo in a nearby alas. Calibrated ages refer to the 2σ range. Ages from bulk material were excluded from the age–depth model.

Laboratory no.	Sample name	Dated material	¹⁴C age (yr BP)	δ¹³C (‰)	Calibrated ages (cal. yr BP)	Mean age (cal. yr BP)
21306	KB7 60	Poales leaf remains and remnants of wood and roots	4984 ± 24	−26.3	5854–5650	5752 ± 102
21307	KB7 154-137	Wooden stems and bark	5587 ± 25	−34.2	6409–6307	6358 ± 51
23120	KB7 243-224	Wood and bark remains	5735 ± 27	−15.5	6633–6451	6542 ± 91
21308	KB7 264-258	Bulk material	12060 ± 40	−30.0	14030–13790	13910 ± 120
23121	KB7 285-265	Wooden remains	5767 ± 27	−18.5	6643–6493	6568 ± 75
23122	KB7 320-306	Wood and poales leaf remains	5790 ± 27	−15.5	6660–6504	6582 ± 78
21309	KB7 382-365	Bulk material	22210 ± 70	−30.7	27480–26190	26835 ± 645
Poz-51636	Pingo top	Molluscs	4460 ± 35	N/A	4966–5289	5128 ± 162

Analyses of lacustrine microfossils

Lacustrine ostracod and cladoceran fossils, that is, their resting eggs (ephippia), were studied to characterise past hydrochemical and ecological lake settings. For specimen counting and taxa identification, core samples with known weight (normalised to 40 g) were wet-sieved through 250-µm mesh screen, air-dried and examined under a stereo-microscope (Zeiss Stemi 2000-C). Ostracod species determination is based on Russian and European identification key textbooks (Bronshtein, 1947; Meisch, 2000) and rare regional ostracod literature (e.g. Pietrzeniuk, 1977; Wetterich et al., 2008). Since Yakutian cladocerans are poorly studied, we used available taxonomic literature (Benzie, 2005; Vandekerkhove et al., 2004) and recent publications (Frolova et al., 2013, 2014; Klimovsky et al., 2015).

Pollen analysis

A total of 18 samples, each consisting of 1–9 g of dry sediment, were treated for pollen analysis using the standard procedure including treatment with HCl and KOH, sieving (250 µm), treatment with HF and mounting in glycerine (Faegri and Iversen, 1989). One Lycopodium spore tablet was added to each sample in order to calculate total pollen and spore concentrations (Stockmarr, 1971). Pollen and spore residues mounted in glycerine were analysed under a Zeiss AxioImager D2 light microscope at 400× magnification. Pollen and spores were identified using a reference pollen collection and pollen atlases (Beug, 2004). Non-pollen palynomorphs (NPPs) were identified using descriptions and illustrations by Van Geel (2001). The microscopic analysis revealed moderately high pollen concentration and generally good pollen grain preservation in the upper 240 cm of the recorded core, allowing easy counting of ⩽300 terrestrial pollen grains per sample. In the lower part, the concentrations of pollen decreased, and only 100–200 grains per sample were counted. Percentages of all taxa were calculated based on setting the total of all pollen and spore taxa equal to 100%. Results of pollen analysis are displayed in the pollen diagram produced with the Tilia/TiliaGraph software (Grimm, 2004). Definition of pollen zones (PZs) was supported by CONISS software.

Results

Geochronology

The two bulk sediment samples show late Pleistocene ages (Table 1; Figure 2) and are interpreted to be overestimated as inherited ages of organic matter from the Yedoma IC deposits, which subsequently subsided or were erosionally re-deposited due to thermokarst processes and lake growth (Biskaborn et al., 2013b; Gaglioti et al., 2014). The five terrestrial plant macrofossil samples show a sequence consistent with the mid-Holocene age of the KB7 record. The continuous deposition is suggested by the general trend in the age–depth relationship (Figure 3, Table 1). Since the C_org content is approximately equal to C_total (see also Supplementary Material, available online), carbonate-derived reservoir effects are suggested to be insignificant. Transforming the sample depths into depositional ages, the part of the KB7 core profile between 365 and 60 cm dbs covers the time interval between ~6750 and ~5870 cal. yr BP. Sedimentation rates decrease from 0.6 cm/yr before ~6500 cal. yr BP to ~0.4 cm/yr between ~6500 and 6300 cal. yr BP and to ~0.2 cm/yr between ~6300 and 5900 cal. yr BP (Figure 3).

Figure 2.

KB7 profile and core including cryolithostratigraphy, grain-size distributions of all samples, calibrated results of the radiocarbon dating and photographs of the dated samples. For details on cryolithology, see Supplementary Material (available online).

Figure 3.

Age–depth model for the KB7 core profile based on Bacon 2.2 modelling routines (Blaauw and Christen, 2011) and calculated sedimentation rates from five AMS ¹⁴C dates calibrated with IntCal13 (Reimer et al., 2013). The red dotted line shows modelled median ages along the core and the grey stippled lines indicate the 95% confidence intervals of the modelled age–depth relationship. The transparent blue violin plots show the five calibrated AMS ¹⁴C dates. The upper left graph shows the iteration history. The middle and right graphs show prior (green lines) and posterior (grey histograms) density functions for accumulation rate and memory.

Sedimentological and biogeochemical characteristics

During field description, the KB7 core profile was characterised by differences in cryostructure, colour and sediment composition. For the cryolithological descriptions, see Supplementary Material (available online). According to its sedimentological characteristics and biogeochemical properties, the KB7 core profile was divided into four major sedimentary units (SUs). Peaks of grain-size distribution are located mainly in the coarse-silt fraction but all samples show bi- to multimodal distributions (Figure 2) and poor to very poor sorting. Additional minor peaks in the fine-sand fraction are mainly found in samples from the upper and lower sediment sequence; minor additional peaks in the coarse-clay to fine-silt fraction are predominantly shown in samples from the middle part of the core.

SU1 (<365–258 cm dbs, ~6750–6550 cal. yr BP)

SU1 is characterised by a general homogeneity in the grain-size distribution (mean: 27.3 ± 2.5 µm; Figure 4). The MS increases upwards and reaches a distinct peak at 285–258 cm dbs. This peak coincides with a peak in the S content, which increases upwards to a maximum value of ~4600 mg/kg (SU1 mean: 2130 ± 1490 mg/kg). Both peaks mark the very dark grey to black horizon between 265 and 258 cm dbs (Figure 2). In total, two possible explanations for the S peak are a strong but short-term input due to a fire event and/or the formation of iron sulphide minerals (greigite and mackinawite) by iron hydroxide reduction in warm shallow waters by the end of SU1 (Siegert, 1979, 1987). The C_org values in SU1 are very low with a mean of 0.9 ± 0.1%. The mean of the δ¹³C values in SU1 is −25.9 ± 0.3‰ while C/N ratios vary little, with values around 9. The comparatively high values of CaCO₃ in SU1, however, are quite remarkable. However, they decrease upwards (SU1 mean value 4.8 ± 1.5%). The Zr values vary between 330 and 443 mg/kg (SU1 mean: 383 ± 30 mg/kg) and the Sr/Rb ratios vary very little between 3.4 and 4.0 (SU1 mean: 3.7 ± 0.2); both suggest high detrital input and domination of the coarse-silt fraction. The Fe/Mn ratios of SU1 are low and vary little around 30, suggesting lower redox conditions and higher oxygen availability during sediment accumulation.

Figure 4.

Summary of sedimentological and biogeochemical parameters plotted against the modelled ages and depth below surface. Mean grain size is illustrated as white diamonds overlying grain-size distribution. The profile zonation according to sedimentary units is given on the right-hand side and highlighted by the grey and white areas.

SU2 (258–164 cm dbs, ~6550–6350 cal. yr BP)

The grain-size distribution in SU2 shows high variations and the lowest overall mean values of the whole sediment sequence (21.7 ± 7.1 µm). The mean grain-size values generally decrease upwards from 31.2 µm at 258–243 cm dbs to 11.9 µm at 215–204 cm dbs (Figures 2 and 4). The MS decreases suddenly at the transition from SU1 to SU2 and decreases upwards with a minimum at 215–204 cm dbs. A strong decrease at the transition to SU2 is also seen in the S content, but afterwards, it strongly increases again from ~1300 to ~4700 mg/kg at 203–186 cm dbs (SU2 mean: 2762 ± 1106 mg/kg). The strong increase in the C_org content to 6.4% (SU2 mean: 3.4 ± 2.2%) and in the C/N ratio to 12.3, as well as the sudden increase in the Fe/Mn ratio at 203–186 cm dbs, indicates a rapid change in organic matter decomposition and redox conditions, likely due to progressively anaerobic conditions in a growing thermokarst lake at the drilling site. Upwardly decreasing and lighter δ¹³C values (SU2 mean: −27.0 ± 0.2‰) suggest an initial change in organic matter source. CaCO₃ values in SU2 are lower (mean: 1.9 ± 1.0%) than in SU1. They stay more or less constant with values between 2.2 and 2.8% until 215–204 cm dbs but fall immediately afterwards to nearly zero. After reaching the highest values at 243–224 cm dbs, the Zr content (SU2 mean: 329 ± 78 mg/kg) and the Sr/Rb ratio (SU2 mean: 3.1 ± 0.5) decrease to minimum peaks at 203–186 cm dbs, may be related to changes in mean grain size.

SU3 (164–70 cm dbs, ~6350–5870 cal. yr BP)

SU3 is characterised by small variations in almost all parameters until the upper part of the active layer. The overall grain-size-distribution mean in SU3 (26.2 ± 8.5 µm) is only a little lower than in SU1 but higher than in SU2 due to sandy outliers. The MS in SU3 slightly decreases upwards, showing very low values. The S content is very high (SU3 mean: 3643 ± 1286 mg/kg) until the active-layer bottom at 110–90 dbs but declines strongly above. Changes in S content parallel changes in C_org values, suggesting that the S in SU3 is mainly organically bound. C_org values are the highest in the whole sequence (SU3 mean: 6.8 ± 1.8%) with a maximum of ~11% at the active-layer bottom. The δ¹³C values decrease at the transition from SU2 to SU3 and show their lowest values, down to −28.8‰ (mean: −28.2 ± 0.7‰). C/N ratios are roughly constant until the surface (SU3 mean: 10.7 ± 0.5). Increasing Fe/Mn ratios (48–66) indicate mainly reducing conditions; constant very low CaCO₃ values suggest acidic conditions. Low Zr values (mean: 285 ± 50 mg/kg) and smaller Sr/Rb ratios (mean: 2.8 ± 0.4) correspond to low detrital input and low sedimentation rates in SU3, respectively (Figure 4).

SU4 (70–30 cm dbs, <5870 cal. yr BP)

The upper part of the active layer is characterised by a strong increase in the mean grain-size distribution and the δ¹³C values and a decrease in Zr content and Sr/Rb ratio, suggesting a different sediment source. This observation agrees with the described lithological difference above and below a ~3- to 5-cm-thick layer of platy reed and other hydrophytes remains that is gently inclined from 60 to 90 cm dbs (see Figure 2 and Supplementary Material, available online). Lower C_org and lower S values in the upper active layer suggest comparably higher rates of organic matter decomposition.

Lacustrine microfossils

A total of 12 ostracod taxa was found at nine sample depths between 320 and 186 cm dbs covering the period between ~6650 and 6400 cal. yr BP. Ephippia fossils of six taxa were found between 320 (~6650 cal. yr BP) and 80 cm dbs (~5970 cal. yr BP) in 15 samples (Figure 5; Frolova et al., in press). Differences in ostracod and ephippia composition allow us to distinguish three lacustrine units (LUs) that are interpreted as different stages of rapid lake development.

Figure 5.

Count data of lacustrine ostracod and ephippia fossils in the KB7 core profile. Data are plotted against the modelled ages and depth below surface. Different lacustrine units (LUs, bordered by the dashed lines) are defined based on the presence and composition of the lacustrine microfossils (see also Frolova et al., in press).

LU1 (320–243 cm dbs, ~6650–6500 cal. yr BP)

The first lake stage is characterised by low abundances of ostracods and ephippia pointing to the unstable conditions of an evolving lake. The presence of D. pulex group indicates shallow nutrient-rich water with high algae density as well as neutral to alkaline pH (Flössner, 2000), and an electrical conductivity range from 0.15 to 1.24 mS/cm (Benzie, 2005). Only three species of the subgenus Ctenodaphnia are known from northern Eurasia (Popova and Kotov, 2013), that is, C. atkinsoni (Baird, 1959), C. similis (Claus, 1876) and C. magna (Straus, 1820). The latter was previously described from the study region (Klimovsky et al., 2015), and we attribute our findings of Ctenodaphnia sp. ephippia to species C. magna (Streble and Krauter, 2002), which prefers higher ionic content and alkaline conditions (Benzie, 2005) of shallow, warm and (hyper-)eutrophic waters (Bellmann, 1989; Jones, 1984; Kotov and Taylor, 2011).

LU2 (243–164 cm dbs, ~6500–6350 cal. yr BP)

The second lake stage represents optimum conditions for the studied lacustrine faunae as seen in sharply increased ostracod and cladoceran abundances and highest species diversity. Overlapping environmental ranges of Candona candida (O.F. Müller, 1776), Candona muelleri jakutica (Pietrzeniuk, 1977), Cyclocypris ovum (Jurine, 1820) and Physocypria kraepelini (G.W. Müller, 1903; synonym for Physocypria fadeewi (Dubowsky, 1927) as described in Pietrzeniuk, 1977) indicate pH values from 7 to 8.7 and an electrical conductivity range from 0.1 to 0.8 mS/cm (Wetterich et al., 2008). The presence of the phytophilic species Cypridopsis vidua (O.F. Müller, 1776) points to well-oxygenated permanent waters with rich (littoral) vegetation (Meisch, 2000). As summarised for LU1, the presence of Ctenodaphnia sp. and, in particular, of C. magna fossils suggests warm, (hyper-)eutrophic and alkaline conditions, while the increase in Daphnia species counts supports rising lake level and an enlarged open-water (limnetic) zone during LU2. Ceriodaphnia sp. and Simocephalus sp. are littoral species of which the latter is related to present vegetation (Kotov et al., 2010).

LU3 (164–80 cm dbs, ~6350–5970 cal. yr BP)

The third lake stage lacks ostracods and exhibits decreasing count numbers and the lowest diversity of cladoceran remains per sample (Figure 5). The taxa Simocephalus sp., Ctenodaphnia sp. and C. magna are completely absent in LU3. The absence of D. pulex above 134 cm dbs (~ 6250 cal. yr BP) points to further deterioration of lacustrine conditions, perhaps lake cooling, acidification or desiccation (or drainage). The ongoing presence of ephippia of the D. longispina group, however, points to at least semi-aquatic conditions of small water ponds, which disappeared after about 5970 cal. yr BP.

Pollen data

In total, 43 pollen and spore taxa could be finally identified in 17 out of 18 sediment samples. The pollen diagram is subdivided into three PZs based on changing pollen taxa composition and abundances (Figure 6). The differences in pollen taxa composition and percentages throughout the KB7 record PZs are interpreted as successive series of local vegetation changes. Larix pollen grains and stomata could not be recorded even if Larix is the dominant forest-forming species today, which was at least also regionally present during the mid-Holocene as reported by, for example, Fradkina et al. (2005).

Figure 6.

Percentage diagram of pollen and non-pollen palynomorph (NPP) data from the KB7 core profile. The pollen zones (PZs) shown on the right-hand side are based on CONISS cluster analysis.

PZI (365–243 cm dbs, ~6750–6500 cal. yr BP)

PZI is notable for low pollen and spore concentrations. The dominant tree pollen are Picea (⩽33%) and Betula (⩽35%); more or less significant are the percentages of Pinus s/g Haploxylon. Among herbs, Artemisia, Cyperaceae and Asteraceae pollen prevail. The PZI is characterised by relatively high percentages of Lycopodium spores and very high abundances of Glomus spores (>300 in one sample).

PZII (243–123 cm dbs, ~6500–6200 cal. yr BP)

PZII is characterised by a sharp increase in Betula pollen (⩽60%) and decrease in Picea (maximum 8%). Among herbs, Poaceae (⩽20%), Artemisia (⩽12%), Cyperaceae (⩽10%), Asteraceae (⩽10%), Chenopodiaceae (⩽5%) and Thalictrum (⩽10%) dominate. Lycopodium spore percentages decrease sharply as do Glomus chlamidiospores. However, PZII could be divided at 164 cm (~6350 cal. yr BP) into two subzones due to the distinct increase in overall pollen and spore concentrations, distinct changes in herb composition and occurrence of aquatic plants.

PZIII (123–60 cm dbs, 6200–5870 cal. yr BP)

In the PZIII, Pinus s/g Diploxylon pollen increases by ⩽60%. Picea pollen also increases (⩽15%) while Betula slightly decreases (⩽25%). Among herbs, Cyperaceae (⩽12%) and Asteraceae (⩽5%) dominate. Botryococcus remains become more abundant in this zone.

Discussion

Local landscape and thermokarst evolution at the KB7 core location

Thermokarst initiation (⩽6750–6500 cal. yr BP)

For the lowest part of the KB7 sediment sequence, we hypothesise initial thermokarst created by melting Yedoma ice wedges, ponding water and surface subsidence (Figure 7a). Increasing air temperatures and precipitation above today’s average values at the onset of the mid-Holocene climate optimum (~6700 cal. yr BP; Fradkina et al., 2005; Nazarova et al., 2013) probably additionally promote the initiation of thermokarst processes at the KB7 site. The low tree pollen counts and high amounts of Glomus and Lycopodium spores in PZI (Figure 6) indicate, however, open forest conditions at the study site and frequent soil and vegetation cover disturbance, possibly due to active thaw processes and related surface subsidence (Andreev et al., 2011). More convenient conditions for lacustrine organisms did not prevail before ~6650 cal. yr BP (Figure 5). Low numbers of ostracods and cladocerans in the lowest core parts further point to slowly changing subaerial thermokarst conditions at the KB7 site until about 6500 cal. yr BP. The very low C_org content, the low C/N ratios and the high δ¹³C values also reflect inhibited bioproductivity under terrestrial conditions during that time. The low Fe/Mn ratio in SU1 (Figure 4) suggests subaerial conditions during sediment deposition, likely indicative of well-drained soil conditions that slowly changed to shallow and well-mixed waters (Biskaborn et al., 2013a). A period of very low water levels between ~6750 and 6650 cal. yr BP is supported by high values of the detrital sediment and grain-size indicators at the very bottom of the KB7 sediment core (see also Table S1 in Supplementary Material, available online). The high carbonate content despite the absence of ostracods in this part suggests high evaporation under subaerial conditions and its authigenic origin from Yedoma IC deposits, which are usually rich in carbonate (Katasonov, 1979; Schirrmeister et al., 2011a). We suggest that the lowest part of the KB7 core represents thermokarst initiation that is characterised by in situ thaw and subsequent Yedoma deposit subsidence. An indication of its Yedoma origin could also be seen in the late Pleistocene age of the bulk sediment sample from the bottom of the KB7 core (26835 ± 645 cal. yr BP). The low level of organics in SU1 also suggests further organic matter decomposition of the Yedoma deposits. This could have been the results of further thaw processes within a talik below the developing thermokarst lake as described from taberal deposits found below Yedoma thermokarst lakes elsewhere (Farquharson et al., 2016; Schirrmeister et al., 2011a; Wetterich et al., 2012). At the end of this initial thermokarst stage, a small, shallow thermokarst lake was formed that probably froze to the bottom in winter (Figure 7a). The lacustrine faunae described above for LU1 indicate warm eutrophic or hypereutrophic waters with higher ionic contents and alkaline conditions, which likely have arisen in shallow water during the short Central Yakutian summers. Typically, iron sulphides (greigite and mackinawite) are formed under such conditions (Siegert, 1979, 1987) as the result of sulphate-reducing bacteria and subsequent reactions of hydrogen sulphide with metal ions (Biskaborn et al., 2012). This would explain the increased S content within the sediment sequence until ~6550 cal. yr BP. However, as the Fe/Mn ratio points to rather oxidising conditions, another event must have additionally led to the peak in the S concentration and the MS values at the end of SU1 (Figure 4). We interpreted the dark grey to black horizon between 265 and 258 cm dbs which includes many microscopic charcoal fragments (Figure 2; see also Supplementary Material, available online) as a marker of a large fire event that could have occurred at the study site around 6550 cal. yr BP. Forest fires are often considered as one of the local triggers for thermokarst processes (Brouchkov et al., 2004; Czudek and Demek, 1970; Edwards et al., 2016; Katamura et al., 2009). Subsequent lake-shore erosion could have washed charcoal-rich sediments into the lake (Katamura et al., 2009). As bulk sediment ages typically overestimated the age of a stratigraphic layer, the bulk sediment age of the black horizon between 265 and 258 cm (13910 ± 120 cal. yr BP) is interpreted as a mixed signal from older reworked carbon sources of the surrounding Yedoma deposits and younger (i.e. contemporary) organic material from the burned catchment vegetation (Biskaborn et al., 2013b; Gaglioti et al., 2014). Synchronously, minor peaks in the C/N ratio, δ¹³C values and Sr/Rb ratio indicate short but strong changes in sediment source. Fire events occur frequently in CY, and the charcoal-rich material could originate from an older fire event. Furthermore, according to Brouchkov et al. (2004), fires and related vegetation disturbance do not always cause thermokarst processes. However, such an event in connection with increasing regional temperatures and precipitation would have contributed to the significant acceleration of thermokarst lake growth as seen afterwards within the KB7 core data (Brouchkov et al., 2004). The dominance of Betula pollen in PZII (Figure 6) is seen as succession after a fire event at the study site (Brown et al., 2015).

Figure 7.

Schematic diagram showing landscape evolution at the study site as discussed in this article. Different landscape stages interpreted from the sedimentology, palaeo-bioindicators, palynology and cryolithology of the KB7 core profile are distinctly mirrored by thermokarst stages (a–d) that presently exist in Central Yakutian thermokarst landscapes.

Thermokarst intensification and rapid lake growth (~6500–6350 cal. yr BP)

The time after 6500 cal. yr BP is characterised by distinct changes in all considered proxies. Most remarkable is the sharp increase in ostracod and ephippia abundance (LU2 in Figure 5), which suggests that rapid lake growth led to optimum conditions for the studied lacustrine faunae within a few decades. According to Fradkina et al. (2005), one phase of maximal climate warming during the mid-Holocene was reached during ~6400–6000 cal. yr BP with annual temperatures ⩽1.5°C higher and precipitation greater than today. During this phase, the vegetation cover destruction around the KB7 lake by the proposed fire event could have further accelerated permafrost degradation and intensified lateral thermo-erosion (Brown et al., 2015; Jorgenson and Osterkamp, 2005). At the same time, increasing water depth by ongoing lake-bottom subsidence and talik growth is very likely after water depth exceeded maximum lake-ice thickness as known from modern thermokarst lakes and numerical simulation of talik development (Grosse et al., 2013; Kessler et al., 2012; Ling et al., 2012; West and Plug, 2008). The talik formation induced a progressive warming and degradation of adjacent permafrost and further lake-basin subsidence and expansion (West and Plug, 2008). Due to the distinct changes in SU2 and LU2 (Figures 4 and 5), we suggest the existence of a reasonably large and deep thermokarst lake from ~6450 cal. yr BP at the latest (Figure 7b). As is shown by the shift to smaller mean grain-size distribution and the strong decrease in the detrital sediment indicators at this time (Figure 4), our coring location became more distal to the sediment source, indicating an enlarged open-water zone with lake expansion (Lenz et al., 2013). According to Lyons and Finlay (2008), rapidly thawing permafrost surrounding the lake could have increased nutrient and weathering product fluxes to the lake; this is usually marked by increased phosphorus and nitrogen values between ~6450 and 6350 cal. yr BP (see Table S1 in Supplementary Material, available online). Increasing bioproductivity under lacustrine conditions after ~6450 cal. yr BP is inferred by rising values of C_org, C/N, δ¹³C, S and the Fe/Mn ratio in SU2. Additionally, oligo- to eutrophic, neutral to slightly alkaline and warm-water conditions are suggested from the lacustrine fauna composition between 6500 and 6350 cal. yr BP. The lake must have then very quickly reached its maximum expansion. Potential reasons that inhibit further deepening are the presence of low ground-ice contents of sediments underneath or accumulating ground-ice-depleted deposits, which insulate and prevent further talik development (Farquharson et al., 2016; Morgenstern et al., 2013; Ulrich et al., 2010).

Decline of thermokarst processes, lake cessation (~6350–5870 cal. yr BP)

Even if the climate conditions continued to be optimal for thermokarst processes (Fradkina et al., 2005; Monserud et al., 1998; Nazarova et al., 2013), we suggest that thermokarst lake growth at the KB7 site slowed and ceased after ~6350 cal. yr BP. Declining thermokarst processes at the KB7 site probably caused by sandy deposits with low ice contents underlying the ice-rich Yedoma deposits at ~6–8 m dbs (Katasonov, 1979; Morgenstern et al., 2013). Stronger evaporation under high summer temperatures and years of less precipitation would have led to gradual lake-level lowering (Nazarova et al., 2013; Riordan et al., 2006). This could also have caused lateral lake expansion to decrease. The gradual lake-level lowering could have been preceded by a lake drainage event. However, thermokarst lake drainage events usually cause a clear disruption in the lacustrine data setting and a change to terrestrial depositional conditions (e.g. Fritz et al., 2016; Lenz et al., 2016; Morgenstern et al., 2013), which we cannot see in our data. Furthermore, as already concluded by Pestryakova et al. (2012), lakes in CY seem to be less vulnerable to lake drainage. But small peaks in the C/N ratio, δ¹³C and S values and the small increase in detrital input parameters at about 175 cm dbs (~6380 cal. yr BP) could result from a sudden but only partial drainage and subsequent lake-level drop. This would have caused a short-term change to more terrestrial depositional conditions and an increasing incorporation of vascular land plants as organic carbon source (Meyers, 2003). Subsequently, a shallow lake remained surrounded by bluffs a few metres high marking the former lake deepening into the Yedoma IC deposits and related surface subsidence (Figure 7c). Some sediment input from slumping lake-shore bluffs during wet years resulted in short-term increases of the grain-size mean (Figure 4). The appearance of aquatic plants in the PZIIb pollen data, Daphnia resting eggs and the low δ¹³C values prove the continued presence of a water body. The lower C/N ratios together with the high C_org values and the high Fe/Mn ratios indicate a nutrient oversupply, high bioproductivity and good organic matter preservation under reducing conditions due to very shallow, warm waters. Anoxic semi-aquatic conditions are suggested by the constantly high S values in SU3 that accompany black streaks in the sediments (Figure 2; see also Supplementary Material, available online). Iron sulphide precipitation is typical for shallow thermokarst lakes and could also result from prolonged ice cover that freezes down to the lake bottom (Biskaborn et al., 2012). Considering the gradual decrease and total disappearance of aquatic faunal remains above the 80 cm sample depth, the lake probably disappeared almost completely after ~5870 cal. yr BP. We suggest the thin layer of platy hydrophytic plant remains at about 60 cm dbs as termination of the lacustrine facies (see description above). Cryoturbation structures just below 60 cm dbs (see Supplementary Material, available online) could indicate a former soil surface and/or very shallow water that froze totally during winter and strong active-layer freeze-thaw cycles (e.g. Edwards et al., 2016). Finally, a small alas basin remained, surrounded by and connected to neighbouring thermokarst lakes by remnant Yedoma deposit bridges (see Figure 7c). Within the small alas, subaquatic to subaerial conditions probably still exist.

Permafrost aggradation and pingo growth (after 5870 cal. yr BP)

For SU4 origin, we hypothesise massive sediment input via the breakthrough of neighbouring lakes and the associated collapse of Yedoma remnants bordering the small alas. These processes could have resulted in higher sand content but the rather mixed signal of other parameters in the uppermost sediment sample (30–70 cm dbs; Figures 2 and 4). The shallow depth of the disappearing lake, which likely froze to the bottom during wintertime, probably caused permafrost aggradation and talik refreezing. After lake disappearance and exposure of the ground to air temperatures, the sediments froze quickly from above as commonly observed in drained thermokarst lake basins (e.g. Lenz et al., 2016), which probably led to organic carbon preservation and the still high C_org contents in SU3. Downward moisture migration during freezing finally resulted in the lenticular-layered cryostructure in the lower parts of the whole sediment sequence (French and Shur, 2010). The rapid superposition of the lacustrine sediments could, moreover, have led to organic matter preservation because of a syngenetically rising permafrost table (Schirrmeister et al., 2011b). The tree-pollen-dominated assemblage at the sample depth, 30–70 cm dbs, could be interpreted as a signal for sediment input from forest-covered Yedoma uplands, because alas basins are usually dominated by grasslands due to salinisation of the alas active layer after lake disappearance (Desyatkin, 2008). However, the high abundance of Pinus subg. Diploxylon pollen in PZIII (Figure 6) can also be explained as a final stage of succession when Betula was replaced by Pinus on the surrounding uplands. This event could have coincided with natural expansion of Scots pine forest between 6400 and 5300 cal. yr BP as a result of warm, moist CY climate conditions as discussed by Katamura et al. (2006).

The coalescence of numerous thermokarst lake basins, progressive thawing of the remaining Yedoma deposits between them and further flattening of the ground-ice-depleted lake-basin centres finally resulted in the development and growth of the large KB alas (Figure 1). With the further regress of the alas slopes, the today’s KB7 drilling site is removed even further from the sediment source. This and the later pingo formation with increasing exposure of the sediments to erosion could explain the hiatus in the uppermost part of the KB7 core profile (Wetterich et al., 2012; Figure 7d). The bleached soil colour and persistently high sodium values in the upper core profile (see text and Table S1 in Supplementary Material, available online) are interpreted as signs of increasing active-layer salinisation likely resulting from high summer evaporation (Larry Lopez et al., 2007). Alterations in soil chemistry and organic matter decomposition due to active-layer depth changes can, thus, not be excluded. Active-layer depths deeper than observed are very likely for the mid-Holocene thermal maximum time of deposition and might be indicated by cryostructure changes (Fritz et al., 2016).

To the end of the mid-Holocene climate optimum in CY after ~5000 cal. yr BP (Fradkina et al., 2005; Monserud et al., 1998; Nazarova et al., 2013), the KB alas had probably almost reached today’s appearance. Considering very slow pingo growing rates of a few millimetres per year (Soloviev, 1973), the existence of ⩽12-m-high pingos within the KB alas proves that permafrost aggradation and talik refreezing must have started not long after that time. This conclusion is supported by our dating of mollusc remains found in 2012 atop a comparable-sized large pingo in a nearby alas, which reveal thermokarst lake existence at this location until at least 5128 ± 162 cal. yr BP (Poznań Radiocarbon Laboratory ID: Poz-51636, unpublished data; Table 1).

The late Holocene in CY is characterised by unstable, generally colder climate conditions than before (Andreev et al., 2002; Fradkina et al., 2005). Widespread thermokarst processes probably decreased significantly during the late Holocene (Nazarova et al., 2013). Furthermore, the low-lying areas of mature alases are considered to be geomorphologically stable with regard to thaw processes because Yedoma deposits below have completely thawed (Brouchkov et al., 2004; Soloviev, 1973). Occasional KB alas flooding during wet years cannot be excluded, and we know of current artificial annual alas flooding by the local population to counteract soil salinisation (personal communication, A. Fedorov). However, sedimentation in the margin areas of alases is rather negligible even under current conditions as indicated by often overgrown and geomorphological stable alas slopes (Czudek and Demek, 1970; Soloviev, 1959; Ulrich et al., 2017). The development of the small KB7 site pingo is considered to be an initial stage; further heaving and pingo growth can be expected. Katasonov’s (1979) drilling results show that the underlying KB alas area sands contain flowing groundwater even during recent decades, leading to the formation of injection ice and possibly further pingo growth.

Past and present thermokarst lake evolution in CY

It is generally assumed that the emergence of large alases requires development time of several hundred to a thousand years (Soloviev, 1973), but the formation of thermokarst lakes can occur over just decades or a few hundred years (Brouchkov et al., 2004; Fedorov et al., 2014) and has been observed to be very rapid especially during recent times (Ulrich et al., 2017). Bosikov (1998) suggests that thermokarst could reach a mature, stable stage as alas within 100–200 years in CY, and Brouchkov et al. (2004) assume that 200–300 years are required to completely thaw a Yedoma sequence below a growing lake. Rapid thermokarst dynamics and very active thermokarst processes are currently widespread in CY. These are usually associated with and initiated by natural and anthropogenic destruction of the forest cover in combination with current climate change (Brouchkov et al., 2004; Fedorov et al., 2014). Exceptionally high thermokarst lake subsidence and growth rates since the 1990s are due to increasing air and ground temperatures and precipitation (Fedorov et al., 2014; Ulrich et al., 2017).

The initiation and expansion of thermokarst processes during the early and mid-Holocene also occurred in concert with rising temperatures (Kaplina, 2009; Katamura et al., 2006; Romanovskii et al., 2000). Additionally, rising precipitation accelerated permafrost thaw and large thermokarst basin development during the Holocene climate optimum (Grosse et al., 2013; Monserud et al., 1998). However, some studies of thermokarst lake growth in arctic regions reveal that lakes are growing slowly over the whole Holocene time period but with different stages of thermokarst intensity (Biskaborn et al., 2013b; Edwards et al., 2016; Morgenstern et al., 2013; Pestryakova et al., 2012). Stages of high thermokarst intensity that are connected with lake expansion and permafrost degradation alternates with stages of low intensity in which thermokarst processes ceased and permafrost can aggrade in drained areas (Morgenstern et al., 2013). Furthermore, many studies in arctic permafrost regions show that thermokarst lakes can grow to very large expanses (>100–500 ha) before they drain, leaving a large drained thaw-lake basin (e.g. Grosse et al., 2013; Lenz et al., 2016; Morgenstern et al., 2013). With regard to the data obtained from the KB7 core and the KB alas morphology (Figure 1), we assume that thermokarst lake evolution at the study site was not an ongoing Holocene process but rather was limited to short-term favourable forcing climatic condition phases leading to very active thermokarst processes and rapid but locally variable landscape modification. Clear evidence for long-lasting, stable lacustrine conditions was not obtained from this study. The evolution of the large KB alases is rather the result of numerous comparable small lake basins coalescing. The formation of large alas systems in CY as described already by Soloviev (1959, 1973) is probably the result of similar processes.

The exceptionally high KB7 core sedimentation rates are interpreted to indicate rapid, highly dynamic lake evolution during the mid-Holocene climate optimum. Usually, Holocene sedimentation rates in thermokarst lakes are far lower in Eastern Siberia (⩽0.4 mm/year; e.g. Biskaborn et al., 2012, 2013b). For the KB7 core, it is assumed that the calculated sedimentation rates (Figure 3) do not represent continuous sedimentation over time. Rather, they indicate frequent but, with decreasing thermokarst activity, decreasing slumping processes delivering sediment material from collapsing lake-shore bluffs during lake expansion (Biskaborn et al., 2013b; Lenz et al., 2013). The extensive resedimentation of Yedoma upland material that increasingly propagated across the lake-basin floor with growing distance between the sediment source and the coring location finally resulted in the relative sedimentological similarity of the whole KB7 core (Lenz et al., 2013; Murton, 1996).

According to our own remote-sensing analyses, expansion rates of individual CY thermokarst lakes were measured at up to ~4–7 m/yr, mainly forced by wet, warm years during the last decades (Ulrich et al., 2017). For the same timescale, Fedorov et al. (2014) found accelerated thermokarst activity with increasing subsidence rates of up to ~18 cm/yr. Considering conservative estimates of average expansion and subsidence rates of CY thermokarst lakes under current warm climate conditions of 0.8–4.0 m/yr and 5–10 cm/yr, respectively (Brouchkov et al., 2004; Fedorov et al., 2014; Ulrich et al., 2017), the KB7 site lake could have reached 120–600 m diameter and 7.5–15 m depth during 150 years. This shows how fast thermokarst processes happened during warm climate conditions in the past and are happening today. According to the projected increase in precipitation and air temperature in arctic and subarctic regions (Bring et al., 2016; IPCC, 2013) and based on the knowledge of past warm climate conditions, we can expect that CY permafrost degradation and thermokarst processes will further increase and accelerate in the future.

Conclusion

The sedimentological and biogeochemical properties of the KB7 core profile in combination with lacustrine microfossils and palynological data enabled the reconstruction of thermokarst landscape dynamics and related environmental conditions at the KB7 core location during the mid-Holocene climate optimum. Dynamic landscape changes during only ~900 years between ~6750 and 5870 cal. yr BP were recorded by the 382-cm-long core profile covering thermokarst initiation, lake expansion and disappearance and talik refreezing. This was followed by pingo growth in the mature, large KB alas basin.

Our study of permafrost response to the interaction of climate warming and local conditions is indeed first of all a local record. However, thermokarst-shaped landscapes are widely distributed in circumarctic periglacial lowlands. Therefore, we believe that our record holds potential for comparison with other locations. Permafrost thaw during warm periods, especially during peak times such as the mid-Holocene is of importance beyond CY in particular with view on expected environmental changes for the 21st century.

Footnotes

Acknowledgements

We greatly appreciate the efforts of all Russian and German colleagues in organising and supporting the fieldwork and laboratory analyses. We thank, in particular, Peter V. Efremov and Avksenty Kondakov (Permafrost Institute Yakutsk) for drilling assistance as well as Birgit Schneider (Leipzig University) for XRF laboratory analyses. Juliane Wolter (AWI Potsdam) helped greatly to determine the dated plant material. LF is especially thankful to Dr Alexey A. Kotov (Severtsov Institute of Ecology and Evolution) for his helpful consultation on ephippia identification. This article benefited by English proofreading and valuable comments from Candace S. O’Connor (Scientific Editor, Fairbanks, Alaska). We are also very thankful for valuable comments and constructive reviews by three anonymous referees. All data for this article are properly cited and referenced in the reference list. Data may be available upon request to the corresponding author.

Funding

The study contributes to the Arctic Ecological Network (Arc-EcoNet) funded by the German Federal Ministry of Education and Research (BMBF grant no.: 01DJ14003 to SW). MU was supported by the German Research Foundation (DFG grant no.: UL426/1-1). The ephippium study of LF was supported by the Russian Scientific Foundation (RSF grant no.: 16-17-1018) and by the Russian Foundation for Basic Research (RFBR grant no.: 15-05-04442). The pollen study of NR was performed according to the Russian Government Programme of Competitive Growth of the Kazan Federal University and supported by the Altai State University (grant no.: 14.Z50.31.0010) and by the Russian Science Foundation (RSF grant no.: 14-50-00036).

References

Andreev

Klimanov

Sulerzhitsky

(2002) History of vegetation and climate of Central Yakutia during the late Glacial and Holocene. Botanical Journal 87(7): 86–98 (in Russian).

Andreev

Schirrmeister

Tarasov

et al . (2011) Vegetation and climate history in the Laptev Sea region (Arctic Siberia) during late Quaternary inferred from pollen records. Quarternary Science Reviews 30: 2182–2199.

Bellmann

(1989) Spinnen, Krebse, Tausendfüßer. Europäische Gliedfüßer (ohne Insekten). München: Mosaik Verlag (in German).

Benzie

JAH

(2005) The Genus Daphnia (Including Daphniopsis) (Anomopoda: Daphniidae). Guides to the Identification of the Microinvertebrates of the Continental Waters of the World 21. Leiden: Kenobi Productions, Ghent & Backhuys Publishers.

Beug

(2004) Leitfaden der Pollenbestimmung für Mitteleuropa und angrenzende Gebiete. München: Verlag Dr. Friedrich Pfeil.

Biskaborn

Herzschuh

Bolshiyanov

et al . (2012) Environmental variability in northeastern Siberia during the last ~13,300 yr inferred from lake diatoms and sediment-geochemical parameters. Palaeogeography, Palaeoclimatology, Palaeoecology 329–330: 22–36.

Biskaborn

Herzschuh

Bolshiyanov

et al . (2013a) Late Holocene thermokarst variability inferred from diatoms in a lake sediment record from the Lena Delta, Siberian Arctic. Journal of Paleolimnology 49: 155–170.

Biskaborn

Herzschuh

Bolshiyanov

et al . (2013b) Thermokarst processes and depositional events in a tundra lake, northeastern Siberia. Permafrost and Periglacial Processes 24(3): 160–174.

Blaauw

Christen

(2011) Flexible paleoclimate age-depth models using an autoregressive gamma process. Bayesian Analysis 6: 457–474.

10.

Bosikov

(1998) Wetness variability and dynamics of thermokarst processes in Central Yakutia. In: Proceedings of the 7th International Permafrost Conference, 23–27 June 1998. Collection Nordicana, No 57. Yellowknife, NT: Université Laval, Centre d’études nordiques, pp. 71–74.

11.

Bouchard

Francus

Pienitz

et al . (2011) Sedimentology and geochemistry of thermokarst ponds in discontinuous permafrost, subarctic Quebec, Canada. Journal of Geophysical Research 116: G00M04.

12.

Bouchard

MacDonald

Turner

et al . (2016) Paleolimnology of thermokarst lakes: A window into permafrost landscape evolution. Arctic Science. Epub ahead of print 22 November. DOI: 10.1139/AS-2016-0022.

13.

Bring

Fedorova

Dibike

(2016) Arctic terrestrial hydrology: A synthesis of processes, regional effects, and research challenges. Journal of Geophysical Research Biogeoscience 121: 621–649.

14.

Bronshtein

(1947) Fresh-Water Ostracoda: Fauna of the USSR Crustaceans, vol. 2, number 1. Moscow: Soviet Academy of Science Publishers (in Russian).

15.

Brouchkov

Fukuda

Fedorov

et al . (2004) Thermokarst as a short-term permafrost disturbance, Central Yakutia. Permafrost and Periglacial Processes 15: 81–87.

16.

Brown

DRN

Jorgenson

Douglas

et al . (2015) Interactive effects of wildfire and climate on permafrost degradation in Alaskan lowland forests. Journal of Geophysical Research Biogeoscience 120: 1619–1637.

17.

Czudek

Demek

(1970) Thermokarst in Siberia and its influence on the development of lowland relief. Quaternary Research 1: 103–120.

18.

Davison

(1993) Iron and manganese in lakes. Earth-Science Reviews 34(2): 119–163.

19.

Desyatkin

(2008) Soil Formation in Thermokarst Depressions (Alas) of the Cryolithozone. Novosibirsk: Nauka (in Russian).

20.

Edwards

Grosse

Jones

et al . (2016) The evolution of a thermokarst-lake landscape: Late Quaternary permafrost degradation and stabilization in interior Alaska. Sedimentary Geology 340: 3–14.

21.

Faegri

Iversen

(1989) Textbook of Pollen Analysis. Chichester: John Wiley & Sons.

22.

Farquharson

Walter Anthony

Bigelow

et al . (2016) Facies analysis of yedoma thermokarst lakes on the northern Seward Peninsula, Alaska. Sedimentary Geology 340: 25–37.

23.

Fedorov

Gavriliev

Konstantinov

et al . (2014) Estimating the water balance of a thermokarst lake in the middle of the Lena River basin, eastern Siberia. Ecohydrology 7(2): 188–196.

24.

Flössner

(2000) Die Haplopoda und Cladocera (ohne Bosminidae) Mitteleuropas. Leiden: Backhuys Publishers (in German).

25.

Fradkina

Alekseev

Andreev

et al . (2005) East Siberia (Based on data obtained mainly in Central Yakutia). In: Velichko

Nechaev

(eds) Cenozoic Climatic and Environmental Changes in Russia. London: The Geological Society of America (Special Paper 382), pp. 105–120.

26.

French

Shur

(2010) The principles in cryostratigraphy. Earth-Science Reviews 101: 190–206.

27.

Fritz

Wetterich

Schirrmeister

et al . (2012) Eastern Beringia and beyond: Late Wisconsinan and Holocene landscape dynamics along the Yukon Coastal Plain, Canada. Palaeogeography, Palaeoclimatology, Palaeoecology 319–320: 28–45.

28.

Fritz

Wolter

Rudaya

et al . (2016) Holocene ice-wedge polygon development in northern Yukon permafrost peatlands (Canada). Quaternary Science Reviews 147: 279–297.

29.

Frolova

Ibragimova

Ulrich

et al . (in press) Reconstruction of thermokarst lake history during the mid-Holocene based on subfossil Cladoceran analysis (Siberia, Central Yakutia). Sibirskiy Ekologicheskiy Zhurnal 4 (in Russian).

30.

Frolova

Nazarova

Pestryakova

et al . (2013) Analysis of the effects of climate-dependent factors on the formation of zooplankton communities that inhabit Arctic lakes in the Anabar River Basin. Contemporary Problems of Ecology 6(1): 1–11.

31.

Frolova

Nazarova

Pestryakova

et al . (2014) Subfossil cladoceran remains from sediment in thermokarst lakes in northeastern Siberia, Russia. Journal of Paleolimnology 52: 107–119.

32.

Gaglioti

Mann

Jones

et al . (2014) Radiocarbon age-offsets in an arctic lake reveal the long-term response of permafrost carbon to climate change. Journal of Geophysical Research Biogeoscience 119(8): 1630–1651.

33.

Grimm

(2004) TGView 2.0.2 (Software). Springfield, IL: Illinois State Museum.

34.

Grosse

Jones

Arp

(2013) Thermokarst lakes, drainage, and drained basins. In: Shroder

(ed.) Treatise on Geomorphology. San Diego, CA: Academic Press, pp. 325–353.

35.

Grosse

Schirrmeister

Siegert

et al . (2007) Geological and geomorphological evolution of a sedimentary periglacial landscape in Northeast Siberia during the Late Quaternary. Geomorphology 86(1–2): 25–51.

36.

Intergovernmental Panel on Climate Change (IPCC) (2013) Climate Change 2013: The Physical Science Basis. Contribution Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (ed Stocker

Qin

Plattner

G-K

et al .). Cambridge: Cambridge University Press, pp. 3–29.

37.

Istomin

Habeck

(2016) Permafrost and indigenous land use in the northern Urals: Komi and Nenets reindeer husbandry. Polar Science 10(3): 278–287.

38.

Jones

(1984) Daphnia magna and D. atkinsoni (Crustacea: Cladocera); new records in Scotland. Journal of Natural History 185: 689–696.

39.

Jorgenson

Osterkamp

(2005) Response of boreal ecosystems to varying modes of permafrost degradation. Canadian Journal of Forest Research 35: 2100–2111.

40.

Kalugin

Daryin

Smolyaninova

et al . (2007) 800-yr-long records of annual air temperature and precipitation over southern Siberia inferred from Teletskoye Lake sediments. Quaternary Research 67: 400–410.

41.

Kaplina

(2009) Alas complex of North Yakutia. Kriosfera Zemli 13(4): 3–17 (in Russian).

42.

Katamura

Fukuda

Bosikov

et al . (2006) Thermokarst formation and vegetation dynamics inferred from a palynological study in Central Yakutia, Eastern Siberia, Russia. Arctic, Antarctic and Alpine Research 38: 561–570.

43.

Katamura

Fukuda

Bosikov

et al . (2009) Charcoal records from thermokarst deposits in central Yakutia, eastern Siberia: Implications for forest fire history and thermokarst development. Quaternary Research 71: 36–40.

44.

Katasonov

(ed.) (1979) Structure and Absolute Geochronology of Alas Deposits in Central Yakutia. Moscow: Nauka (in Russian).

45.

Kessler

Plug

Walter Anthony

(2012) Simulating the decadal- to millennial-scale dynamics of morphology and sequestered carbon mobilization of two thermokarst lakes in NW Alaska. Journal of Geophysical Research 117: G00M06.

46.

Klimovsky

Becker

Sinev

et al . (2015) Cladocera (Crustacea, Branchiopoda) of Central Yakutia. 4. Systematics-faunistic and zoogeographical analysis. Zoological Journal 94: 1367–1378.

47.

Kotov

Taylor

(2011) Mesozoic fossils (>145 Mya) suggest the antiquity of the subgenera of Daphnia and their coevolution with chaoborid predators. BMC Evolutionary Biology 11: 129.

48.

Kotov

Sinev

Glagolev

et al . (2010) Cladocerans (Cladocera). In: Alekseev

Tsalolihin

(eds) The Determinant of European Russia Freshwater Zooplankton and Zoobenthos: Zooplankton, vol. 1. Moscow: KVM Partnership Scientific Publications, pp. 151–276.

49.

Larry Lopez

Brouchkov

Nakayama

et al . (2007) Epigenetic salt accumulation and water movement in the active layer of central Yakutia in eastern Siberia. Hydrological Processes 21(1): 103–109.

50.

Lenz

Fritz

Wetterich

et al . (2013) Postglacial landscape dynamics in the West-Canadian Arctic – Results from a lake record on a push moraine (Herschel Island, Yukon Territory). Palaeogeography, Palaeoclimatology, Palaeoecology 381–382: 15–25.

51.

Lenz

Grosse

Jones

et al . (2016) Mid-Wisconsin to Holocene permafrost and landscape dynamics based on a drained lake basin core from the northern Seward Peninsula, Northwest Alaska. Permafrost and Periglacial Processes 27(1): 56–75.

52.

Liljedahl

Boike

Daanen

et al . (2016) Pan-Arctic ice-wedge degradation in warming permafrost and its influence on tundra hydrology. Nature Geoscience 9: 312–318.

53.

Ling

Zhang

et al . (2012) Modelling open-talik formation and permafrost lateral thaw under thermokarst lake, Beiluhe Basin, Qunghai-Tibet Plateau. Permafrost and Periglacial Processes 23(4): 312–321.

54.

Lyons

Finlay

(2008) Biogeochemical processes in high-latitude lakes and rivers. In: Warwick

Laybourn-Parry

(eds) Polar Lakes and Rivers – Limnology of Arctic and Antarctic Aquatic Ecosystems. New York: Oxford University Press, pp. 137–156.

55.

Meisch

(2000) Freshwater Ostracoda of Western and Central Europe. Heidelberg; Berlin: Spektrum Akademischer Verlag.

56.

Meyers

(2003) Applications of organic geochemistry to paleolimnological reconstruction: A summary of examples from Laurentian Great Lakes. Organic Geochemistry 34: 261–289.

57.

Monserud

Tchebakova

Denissenko

(1998) Reconstruction of the mid-Holocene palaeoclimate of Siberia using a bioclimatic vegetation model. Palaeogeography, Palaeoclimatology, Palaeoecology 139: 15–36.

58.

Morgenstern

Grosse

Günther

et al . (2011) Spatial analyses of thermokarst lakes and basins in Yedoma landscapes of the Lena Delta. The Cryosphere 5: 849–867.

59.

Morgenstern

Ulrich

Günther

et al . (2013) Evolution of thermokarst in East Siberian ice-rich permafrost – A case study. Geomorphology 201: 363–379.

60.

Murton

(1996) Thermokarst-lake-basin sediments, Tuktoyaktuk Coastlands, western arctic Canada. Sedimentology 43: 737–760.

61.

Nazarova

Lüpfert

Subetto

et al . (2013) Holocene climate conditions in central Yakutia (Eastern Siberia) inferred from sediment composition and fossil chironomids of Lake Temje. Quaternary International 290–291: 264–274.

62.

Pestryakova

Herzschuh

Wetterich

et al . (2012) Present-day variability and Holocene dynamics of permafrost-affected lakes in central Yakutia (Eastern Siberia) inferred from diatom records. Quaternary Science Reviews 51: 56–70.

63.

Pietrzeniuk

(1977) Ostracoden aus Thermokarstseen und Altwässern in Zentral-Jakutien (Ostracods from thermokarst lakes and old branches of Central Yakutia). Mitteilungen aus dem Zoologischen Museums in Berlin 53: 331–362 (in German).

64.

Popova

Kotov

(2013) Latitudinal patterns in the diversity of two subgenera of the genus Daphnia O.F. Müller (Crustacea: Cladocera: Daphniidae). Zootaxa 3736: 159–174.

65.

R Core Team (2012) R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing.

66.

Reimer

Bard

Bayliss

et al . (2013) IntCal13 and Marine13 radiocarbon age calibration curves 0–50,000 years cal BP. Radiocarbon 55: 1869–1887.

67.

Riordan

Verbyla

McGuire

(2006) Shrinking ponds in subarctic Alaska based on 1950–2002 remotely sensed images. Journal of Geophysical Research 111: G04002.

68.

Romanovskii

Hubberten

H-W

Gavrilov

et al . (2000) Thermokarst and land-ocean interactions, Laptev Sea Region, Russia. Permafrost and Periglacial Processes 11: 137–152.

69.

Schirrmeister

Grosse

Wetterich

et al . (2011b) Fossil organic matter characteristics in permafrost deposits of the northeast Siberian Arctic. Journal of Geophysical Research 116: G00M02.

70.

Schirrmeister

Kunitsky

Grosse

et al . (2011a) Sedimentary characteristics and origin of the Late Pleistocene Ice Complex on north-east Siberian Arctic coastal lowlands and islands – A review. Quaternary International 241: 3–25.

71.

Schleusner

Biskaborn

Kienast

et al . (2015) Basin evolution and palaeoenvironmental variability of the thermokarst lake El’gene-Kyuele, Arctic Siberia. Boreas 44: 216–229.

72.

Séjourné

Costard

Fedorov

et al . (2015) Evolution of the banks of thermokarst lakes in Central Yakutia (Central Siberia) due to retrogressive thaw slump activity controlled by insolation. Geomorphology 241: 31–40.

73.

Siegert

(1979) Minerologic-petrographic characteristic of alas deposits. In: Katasonov

(ed.) Structure and Absolute Geochronology of Alas Deposits in Central Yakutia. Moscow: Nauka, pp. 44–61 (in Russian).

74.

Siegert

(1987) Greigite and mackinawite in quaternary deposits of central Yakutia. Mineralogicheskiy Zhurnal 9(5): 75–81 (in Russian).

75.

Soloviev

(1959) Cryolithic Zone of the Northern Part of Lena-Amga Interfluve (in Russian). Moscow: Izdatel’stvo Akademii SSSR, pp. 142.

76.

Soloviev

(1973) Guidebook: Alas thermokarst relief of Central Yakutia. In: Second International Conference on Permafrost, Yakutsk 13–28 July 1973. Yakutsk: USSR Academy of Sciences, Section of Earth’s Sciences, Siberian Division.

77.

Stockmarr

(1971) Tablets with spores used in absolute pollen analysis. Pollen et Spores 13: 615–621.

78.

Streble

Krauter

(2002) Das Leben im Wassertropfen; Mikroflora und Mikrofauna des Süsswassers. Kosmos Naturführer. Stuttgart: Kosmos-Verlag (in German).

79.

Streletskiy

Anisimov

Vasiliev

(2015) Permafrost degradation. In: Haeberli

Whiteman

Shroder

JF Jr

(eds) Snow and Ice-Related Hazards, Risks and Disasters. Amsterdam: Elsevier, pp. 303–344.

80.

Stuiver

Polach

(1977) Discussion: Reporting of 14C data. Radiocarbon 19(3): 355–363.

81.

Troeva

Isaev

Cherosov

et al . (2010) The Far North: Plant Diversity and Ecology of Yakutia, Plant and Vegetation 3. Dordrecht: Springer Science + Business Media B.V., 390p.

82.

Ulrich

Matthes

Schütze

et al . (2017) Differences in behavior and distribution of permafrost-related lakes in Central Yakutia and their response to climatic drivers. Water Resources Research 53: 1167–1188.

83.

Ulrich

Morgenstern

Günther

et al . (2010) Thermokarst in Siberian ice-rich permafrost: Comparison to asymmetric scalloped depressions on Mars. Journal of Geophysical Research 115: E10009.

84.

Van Geel

(2001) Non-pollen palynomorphs. In: Smol

Birks

HJB

Last

(eds) Terrestrial Algal and Siliceous Indicators; Tracking Environmental Changes Using Lake Sediments. Dordrecht: Kluwer Academic Press, pp. 99–119.

85.

Vandekerkhove

Declerck

Vanhove

et al . (2004) Use of ephippial morphology to assess richness of anomopods: Potentials and pitfalls. Journal of Limnology 63: 75–84.

86.

Walter Anthony

Daanen

Anthony

et al . (2016) Methane emissions proportional to permafrost carbon thawed in Arctic lakes since the 1950s. Nature Geoscience 9: 679–682.

87.

West

Plug

(2008) Time-dependent morphology of thaw lakes and taliks in deep and shallow ground ice. Journal of Geophysical Research 113: F01009.

88.

Wetterich

Grosse

Schirrmeister

et al . (2012) Late Quaternary environmental and landscape dynamics revealed by a pingo sequence on the northern Seward Peninsula, Alaska. Quaternary Science Reviews 39: 26–44.

89.

Wetterich

Herzschuh

Meyer

et al . (2008) Evaporation effects as reflected in freshwaters and ostracod calcite from modern environments in Central and Northeast Yakutia (East Siberia, Russia). Hydrobiologia 614: 171–195.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.27 MB