The spatial distribution of sedimentary compounds and their environmental implications in surface sediments of Lake Khar Nuur (Mongolian Altai)

Lake sediments are valuable natural archives to reconstruct paleoclimate and paleoenvironmental changes which consist of inorganic and organic sediment compounds of allochthonous origin from the catchment and of autochthonous production in the lake. However, for robust paleo‐reconstructions it is important to develop a better understanding about sedimentation processes, the origin of inorganic and organic sediment compounds and their distribution within the lake. In this context, modern process studies provide important insights, although environmental and anthropological changes can affect the spatial distribution of sediment compounds through time. Therefore, in this study the spatial distribution of grain size and geochemical proxies in 52 surface sediment samples from Lake Khar Nuur, a small high‐altitude lake in the Mongolian Altai with a small and anthropogenically used hydrological catchment, is investigated. The results show a distinct sediment focussing in the two deep basins of the lake, which therefore act as accumulation zones. In those accumulation zones, total organic carbon (TOC), total nitrogen (N) and their isotopic composition (δ13CTOC, δ15N) as well as n‐alkanes indicate that organic sediment compounds are a mixture of both allochthonous and autochthonous origin. While the recent catchment vegetation consists of grasses/herbs and the shrub Betula nana (L.) with distinct differences in their n‐alkane homologue patterns, those differences are not reflected in the sediment surface samples which rather indicates that grass‐derived n‐alkanes become preferentially incorporated in the lake. Extensive anthropogenic activity such as grazing and housing in the southern part of the catchment causes soil erosion which is well reflected by high TOC, N and sulphur (S) contents and 15N depleted δ15N values at the central southern shore, i.e. increased allochthonous sediment input by anthropogenically‐induced soil erosion. Overall, the surface sediments of Lake Khar Nuur origin from allochthonous and autochthonous sources and are focussed in the accumulation zones of the lake, while their distribution is both environmentally and anthropogenically driven.


| INTRODUCTION
Lake sediments are valuable natural archives providing long records of past limnological, hydrological, and anthropological change. However, robust reconstructions require an understanding of how sediments and the recent climate/environmental signals are incorporated into the sediment archive and how the sediment archive is distributed within the lake basin. Studies investigating the composition and spatial distribution of lake surface sediments have a long history in the field of limnology. Over the last decades, the spatial distribution of sedimentological, inorganic, organic and biogeochemical proxies (Anderson et al., 2008;Dearing, 1997;Ju et al., 2010;Kastner et al., 2010;Shuman, 2003;Thomas et al., 1972;Vogel et al., 2010;Wang et al., 2015;Yu et al., 2015;Yu et al., 2018), as well as heavy metals (Guo et al., 2018;Moore, 1980;Onyari & Wandiga, 1989), pollen (DeBusk, 1997;Li, 2018Li, , 2019Zhao et al., 2006), diatoms and ostracods (Anderson, 1990;Martens & Tudorancea, 1991;Shah et al., 2017;Yu et al., 2019;Zalat & Vildary, 2005) was analysed in lake surface sediment samples by several studies. While most of these studies show that sediments and sedimentary compounds accumulate in the deep basins within lakes (i.e. sediment focussing), they mainly focussed on large hydrological settings with large lake surfaces.
Therefore the spatial resolution of the analyses and the related process understanding are limited due to the low sample size compared to the lakes surface. Thus, there is a lack of studies investigating lake surface sediments in high spatial resolution to gain a better understanding about modern sedimentation processes, the origin of different sedimentary compounds and their distribution within the lake.
Sedimentary processes in hydrological systems can best be understood by the grain size and elemental [e.g. aluminium (Al), iron (Fe)] distribution which is directly related to transport mechanisms and transport energies and indicate allochthonous input into the lake (Håkanson & Jansson, 1983;Mclaren & Bowles, 1985;Wen et al., 2008;Wündsch et al., 2018). Both proxies can be used to trace sediment focussing in deeper parts of the lake below the mud deposition boundary (Anderson et al., 2008;Lehman, 1975;Wang et al., 2015).
The origin of organic sedimentary compounds in lake sediments (allochthonous vs. autochthonous) can be disentangled by total organic carbon (TOC) and total nitrogen (N) contents, as well as their stable carbon and nitrogen isotopes (Meyers, 1994;Meyers, 2003). Since TOC can represent a heterogeneous organic matter pool in lake sediments that comprises organic carbon of different origins and preservation states, n-alkanes can be complementary for source detection.
They are a more homogenous part of the organic matter pool, and depending on their chain-length distribution, they are of autochthonous (aquatic) or allochthonous (terrestrial) origin. While n-alkanes produced by aquatic (submerged) macrophytes have a dominance of the shorter chains nC 21 and nC 23 (Ficken et al., 2000;Sachse et al., 2004), n-alkanes produced in the leaf waxes of higher terrestrial plants have a dominance on the longer chains from nC 27 to nC 35 (Eglinton & Eglinton, 2008;Shepherd & Wynne Griffiths, 2006). Besides being source-specific, the homologue patterns of leaf wax n-alkanes can be used to differentiate between vegetation types. Although homologue patterns can have limited differentiation power on a global scale (Bush & McInerney, 2013), they can differentiate well between grasses and herbs (i.e. nC 31 and nC 33 ) and woody shrubs and deciduous trees (i.e. nC 27 and nC 29 ) in many regions of the world including semi-arid Mongolia (Bliedtner et al., 2018;Liu et al., 2018;Schäfer et al., 2016;Struck et al., 2020). Up to now, studies analysing the spatial distribution of n-alkanes and their origin in surface sediment samples are very limited , and studies that combine both sedimentological and inorganic as well as organic and biogeochemical proxies in small lakes with small catchments and a clearly defined catchment-geomorphology are still scarce. However, such lakes and their respective catchments have the great advantage of a high sediment connectivity, that is rapid sediment transport from the catchment into the lake, while sediment re-working in the lake is limited (Oldfield, 1977).
Therefore, the aim of this study is to investigate the spatial distribution of sedimentological and geochemical proxies of surface sediments in Lake Khar Nuur, a small high-altitude lake with a small drainage basin in the Mongolian Altai. We aim to develop a better understanding about recent sedimentation processes, the origin of different sedimentary compounds and how those compounds are distributed within the lake. Although environmental changes can affect the spatial distribution of sediment compounds through time, we further aim to improve proxy-understanding and paleoenvironmental interpretations using modern analogues in lake sediments. More specifically, we address the following objectives: • We will investigate the origin and focussing of allochthonous and autochthonous inorganic and organic sediment compounds in the lake by grain size distribution, elemental composition (Al, Fe), TOC and N and their stable isotopes (δ 13 C TOC , δ 15 N) as well as nalkanes.
• We will investigate distinct differences in the vegetation types of the catchment by leaf wax n-alkane homologue patterns and if those vegetation types can also be detected in the lakes surface sediments.
• We will identify traces of anthropogenic activity and soil erosion in the catchment in the lakes surface sediments by elemental sulphur (S) as well as TOC and N contents and their stable isotopes.

| STUDY SITE
Lake Khar Nuur is a small para-glacial lake located at 2,486 m above sea level (a.s.l.) in the Mongolian Altai (48 37 0 N, 88 56 0 E) and has a small drainage basin (44.8 km 2 ) that covers altitudes from 2,486 to 3,156 m a.s.l. (Figure 1a (Walther et al., 2017). The bedrock of most parts of the catchment is composed of friable black clay shales, whereas only the morainal deposits in the western parts of the catchment consist of granite. Steep slopes (>16 ) occur in the northern, western and south-eastern parts of the catchment, but flatter areas are noticeable in the central southern part (Figures 1c, 2a).
On the steep slopes of the catchment, very shallow Leptosols occur, whereas Mollisols with humus rich topsoils ($50 cm) and underlying permafrost developed in the flatter areas of the central southern part of the catchment. At those flatter areas, several gullies exist (Figures 1b,2c) and favour increased runoff and sediment transport into the lake during snowmelt and precipitation events. The other parts of the catchment only produce little episodic runoff and no perennial inflow into the lake exists (Figure 1b, c). Today, most parts of the catchment are anthropogenically used during the summer month, although the central southern part is preferential used by grazing and nomadic housing ( Figure 1d). Additionally, a road was constructed around the southern part of the lake (Figure 2b).

| Bathymetry
A bathymetric map of the lake floor was created using a Lowrance HPS 5 Fishfinder working with a frequency of 100 kHz. A $240,000 depth-measurements were conducted and the respective drive-lines are shown in Figure 3(a). The map was created using the software Sonar Viewer 2.1.2 and Esri ArcGIS 10.5 applying the 'spline with barriers' interpolation tool.
F I G U R E 1 (a) Location of the study site in western Mongolia. (b) Satellite image of the study site with the hydrological catchment of Lake Khar Nuur, the lake bathymetry with the two basins A and B, the surface sediment sample locations with sample identification (ID) and the viewpoints of the images in Figure 2. The grey shaded area shows the morainal deposits.

| Sampling and laboratory analyses
For this study, the sediment surface (0-1 cm) of the lake floor was sampled at 52 locations (for sample locations see Figure 1b) using a Van Veen Grab sampler (Lie & Pamatmat, 1965). All samples are from a water depth > 2 m (Figure 1a). Except for grain size analyses, samples were freeze-dried (−50 C; > 72 h), ground and sieved to < 40 μm.

| Grain size analyses
For grain size analyses, sample aliquots were treated with hydrogen peroxide (H 2 O 2 , 15%, 30%) to remove organic matter and subsequently treated with hydrochloric acid (HCl, 15%) to remove carbonates. As a dispersant, 5 ml sodium-pyrophosphate (Na 4 P 2 O 7 Á10 H 2 O) were added to each sample and shaken for > 2 h. Before the measurement, the fraction > 2 mm was removed from the samples by sieving due to technical reason. It has to be noted that the > 2 mm fraction was only present in the samples 38,40,45,46,48,50 and 51 (grey sample points in Figure 3b), which, however, was not further quantified. The < 2 mm grain-size distribution was determined with a laser diffraction particle size analyzer (LS 13320, Beckman Coulter, Brea, California, USA). Samples were measured with the aqueous liquid module in several 60 s cycles until a reproducible signal was obtained.

| Carbon and nitrogen analyses and bulk isotopic analyses
Carbon and nitrogen contents and their stable isotopes were analysed with an Elementar Analyser (vario EL cube) coupled to an isotope ratio mass spectrometer (Isoprime precision). $30 mg of untreated sediment per sample was weighted into tin boats (Elementar, 6 × 6 × 12) and measured for total carbon (TC), total nitrogen (N) and the nitrogen isotopic composition (δ 15 N). Prior to the analyses of TOC and the carbon isotopic composition (δ 13 C), carbonates were removed from the samples with 1 N HCl at 60 C for 8 h. Samples were subsequently washed with ultrapure water to pH neutrality and $30 mg were weighted into tin boats. The analytical precision of the δ 15 N and δ 13 C analyses was checked against certified standards (L-Prolin, EDTA and USGS65), and gave an analytical error < 0.25‰ and < 0.05‰, respectively. The δ 15 N and δ 13 C are given in their delta notation against Air and the Vienna Pee Dee Belemnite (VPDB).
Total inorganic carbon (TIC) was calculated by subtracting TOC from TC. The molar ratio of TOC and N was calculated based on their respective molecular weights (Equation (1)). Relative errors were calculated based on triplicate measurements (TC: < 0.1%, N: < 0.9%, TOC: < 0.8%).

| Leaf wax analyses
Total lipids of 49 surface sediment samples, 13 topsoil samples n-Alkane concentrations ( P n-alkanes) are given in μg g −1 dry weight and were calculated as the sum of nC 21 to nC 35 . The odd- We used a normalized n-alkane ratio that we modified according to the differences between the two prominent vegetation forms in the Khar Nuur catchment and that we calculated to detect possible con- To further obtain information about n-alkane contributions from aquatic and terrestrial plants in our surface sediment samples, we quantify the relative contribution of n-alkanes derived from submerged and floating macrophytes relative to n-alkanes derived from leaf waxes of terrestrial plants (P aq 0 ratio; Equation 4; modified after Ficken et al., 2000). Aquatic plants (of submerged macrophyte origin) are dominated by short-chain n-alkanes (typically nC 21 and nC 23 ; e.g. Sachse et al., 2004). Grasses/herbs and shrubs show a nC 31 and nC 29 predominance in Mongolia (Struck et al., 2020).

| Data analyses
The hydrological Khar Nuur catchment and its geomorphological attributes were derived from a digital elevation model (SRTM) with a spatial resolution of $30m × 30 m. Based on the sedimentological and geochemical results of the surface sediment samples, spatial distribution maps were generated using the 'spline with barriers' tool in Esri ArcGIS 10.5. Pearson's r-values were calculated to detect correlations of the analysed proxies and a student's t-test was performed to determine significant differences (α = 0.05). All statistical analyses were performed using the statistical software OriginPro 2017.

| Bathymetry and grain size distribution
The bathymetric map of Lake Khar Nuur shows two basins within the lake. Following the morphology of the catchment, the first basin However, the molar TOC/N ratio ranges between 1.22 and 6.60 and reflects only a very small range.
The spatial distribution of Al, Fe and TOC/N is also reflected in significant correlations with water depth (r: Al = 0.87, Fe = 0.81 and TOC/N = 0.59; α < 0.05; Table 1). Although TOC, N and TIC are also significantly correlated with water depth (r between 0.56 and 0.75; α < 0.05; Table 1) and Al (r between 0.61 and 0.80; α < 0.05;
Significant correlations exist between the spatial distribution of the n-alkane concentration, the n-alkanes nC 21 , nC 31 and the P aq 0 (r between 0.67 and 0.99; α < 0.05; Table 1). While the OEP is significantly anti-correlated with water depth (r = −0.62; α < 0.05), there exists no significant correlation between the modified n-alkane ratio and water depth (r = −0.23; α > 0.05; Table 1). Moreover, there is a significant anti-correlation between the P aq 0 and the modified n-alkane ratio (r = −0.73; α < 0.05; Table 1).

| Allochthonous versus autochthonous sediment and inorganic and organic sediment compound origin in Khar Nuur sediments
Lake sediments and their inorganic and organic sediment compounds are a mixture of allochthonous material from the F I G U R E 4 Spatial distribution of (a Al, (b) Fe, (c) S, (d) TOC, (e) N, (f) TIC, (g) δ 13 C TOC , (h) δ 15 N and (i) molar TOC/N ratio of Lake Khar Nuur. The lower and upper scale limit is based on the lowest and highest measured values. Note that for the distribution map of S an upper limit of 2,000 ppm instead of the detected 4,900 ± 500 ppm was used to highlight differences of the spatial distribution. Also shown are the slope and the catchment of Lake Khar Nuur [Colour figure can be viewed at wileyonlinelibrary.com] catchment and autochthonous material produced within the lake (Meyers & Ishiwatari, 1993). Therefore, it is an important precondition to disentangle both signals to robustly interpret transport processes, inorganic and organic sediment compound origin/sources and the environmental conditions under which they accumulate.

| Grain size and elements
A fundamental proxy to differentiate sediment transport energy and transport regimes in lake sediments is the grain size distribution. The most prominent sediment transport regimes can be identified by the Mode 1 which describes the most frequently occurring particle size in each sample (Blott & Pye, 2001). The grain size distribution in the surface sediments of Lake Khar Nuur exhibits high Mode 1 values in the nearshore samples compared to low values in the two basins ( Figure 3b). Consequently, the grain size distribution follows the direction of sediment transport into the lake, indicating a higher energy regime near the shore and a lower energy regime towards the central part of the lake where sediments are focussed (Håkanson & Jansson, 1983;Mclaren & Bowles, 1985;Wang et al., 2015). This sediment focussing is likely caused by higher transport energies near the several small periodically inflows which mainly occur along the depth contours of the catchment during precipitation events and snow melt ( Figure 1b). The clay and silt fractions are transported towards the lake centre of basins A and B where they can settle at calmer water conditions. Sediment focussing of the finer fractions in deeper waters of Lake Khar Nuur is further supported by the significant anticorrelation of water depth and Mode 1 (r = −0.45, α < 0.05; Table 1; Figure 3b). This typical sediment distribution was also found in previous studies (Anderson et al., 2008;Thomas et al., 1972;Wang et al., 2015). It has to be noted that the > 2 mm fraction was removed before grain size measurements due to technical reasons. While the > 2 mm fraction was not quantified but present in some samples (grey sample points in Figure 3b), it is likely that the > 2 mm fraction would lead to higher Mode 1 values in those samples. However, the > 2 mm fraction was exclusively found in the nearshore samples which supports our findings that high transport energy regimes occur near the periodical inflows at the shoreline.
The decrease in grain size from the shoreline to the basins is accompanied by an increase in the concentration of the allochthonous elements Al and Fe. Both elements are typically bound to the fine grain size fraction (i.e. clay and feldspars) (e.g. Wen et al., 2008;Wündsch et al., 2018) and therefore significantly anti-correlate with the Mode 1 in Lake Khar Nuur (r = −0.49 and −0.55, respectively; α < 0.05; Table 1; Figures 3b, 4a, b). Since Al and Fe are thought to be of allochthonous origin in lacustrine systems (Wen et al., 2008;Wündsch et al., 2018), their significant correlation with water depth (r = 0.87 and 0.81, respectively; α < 0.05; Table 1) indicates the high accumulation of allochthonous material within the deep zones of basins A and B. According to Anderson et al. (2008, and references cited therein), sediment accumulation mostly occurs below a mud deposition boundary depth which is further described by Wang et al. (2015) as the sediment accumulation zone that is least influenced by wind-driven turbulences and other processes. Thus, this accumulation zone provides the best suited sediments for paleoenvironmental studies, and for Lake Khar Nuur, a distinct sediment focussing occurs in the accumulation zones in the two basins A and B (Anderson et al., 2008;Lehman, 1975;Wang et al., 2015).

| Carbon and nitrogen contents and bulk isotopic composition
The accumulation zones in the two basins in Lake Khar Nuur are also evident in the characteristic spatial distribution of inorganic and organic sediment compounds. The highest TOC and N concentrations, as well as highest molar TOC/N ratios occur in both basins, resulting in significant correlations of TOC, N and TOC/N ratio with water depth (r = 0.75, 0.71 and 0.59, respectively; α < 0.05; Table 1; Figure 4d, e, i). Low molar TOC/N ratios in the two basins would classically indicate a predominantly autochthonous origin of organic material since values < 10 are thought to be representative for aquatic plants (Meyers, 1994). However, TOC, N and their molar ratio also significantly correlate with the allochthonous elements Al and Fe (Table 1) (Meyers, 1994;Pyankov et al., 2000). Changes in δ 13 C of C3 plants indicate variations in plant water use efficiency and thus drought stress (Rao et al., 2017;Struck et al., 2020), but those changes in water use efficiency would only explain a few ‰ differences and not the large differences at the northern and central southern shoreline.
Another explanation are aquatic plants that can have δ 13 C TOC values in the isotopic range of C3 plants (−30 to −25‰) when they use dissolved carbon dioxide (CO 2 ) from the water column which is in isotopic equilibrium with the atmosphere (Meyers, 1994). However, when the availability of dissolved atmospheric CO 2 (δ 13 C = −7%) is limited and aquatic plants begin to use dissolved bicarbonate ion (HCO 3 − ) (δ 13 C = 1%), their isotopic composition becomes enriched in 13 C, resulting in up to 10‰ more positive δ 13 C TOC values (Bradley, 2015;Yu et al., 2015). The HCO 3 − origin in the lake from formerly deposited back-dissolved TIC can be of autochthonous origin by evaporitic carbonate precipitation and/or biogenic incorporation (formation of carbonate shells by plants and animals) (Jones & Bowser, 1978;Ju et al., 2010;Kelts & Hsü, 1978;Yu et al., 2018). Additionally, TIC can derive from allochthonous sources such as limestones in the catchment and/or aeolian dust which is an important carbon source in semi-arid regions (Jones & Bowser, 1978;Grunert & Lehmkuhl, 2004;Kelts & Hsü, 1978). Although we cannot completely rule out the contribution of allochthonous carbonates in the Khar Nuur sediments, they are rather unlikely because the bedrock of the catchment is not made of calcareous bedrock which is supported by modern 14 C-ages of a dated water plant and a surface sediment (see Supporting Information, Table S1). Thus, since the so-called 'hardwater effect' from old calcareous bedrock is not evident in the Khar Nuur catchment, TIC should mainly be derived from carbonate precipitation and biogenic sources.
However, samples with 13 C enriched δ 13 C TOC values along the northern shore have likewise low TIC concentration. Therefore, we suggest that the strong enrichment is due to a dissolution of the formerly deposited inorganic carbon in the lake water and that the southward exposure of the northern shore enables enhanced aquatic productivity in the lake, also below the ice cover (eight to nine months) (Gibson et al., 1999;Loose et al., 2009). However, we have to mention that a more detailed picture of the TIC sources would be derived by further mineralogical analyses.

| Leaf wax n-alkanes
A more homogenic part of the organic matter pool in lake sediments are n-alkanes. Based on their chain-length distribution one can differentiate whether they have a allochthonous origin from the leaf waxes of terrestrial plants (nC 27 -nC 35 ) or a autochthonous origin from aquatic (submerged) macrophytes (nC 21 and nC 23 ) (Bliedtner et al., 2018;Eglinton & Eglinton, 2008;Ficken et al., 2000;Sachse et al., 2004;Struck et al., 2020). In Lake Khar Nuur, the spatial distribution Overall, the n-alkane preservation is good with OEP values above 5, and higher values at some sites close to the shore probably document input of poorly degraded allochthonous n-alkanes.
It is important to note that n-alkanes represent only a small fraction of the total organic matter pool in lake sediments and are possibly less influenced by degradation effects on shorter timescale compared to TOC, N and their stable isotopes (Eglinton & Eglinton, 2008;Meyers, 2003). Nevertheless, this does not hamper the overall picture that the heterogenic organic matter pool as well as the more homogenic n-alkanes comprise a mixture of autochthonous and allochthonous organic matter sources.

| The vegetation distribution in Khar Nuur sediments using leaf wax n-alkanes
Long-chain n-alkanes have the potential to differentiate between vegetation types by their homologue patterns (Bliedtner et al., 2018;Marseille et al., 1999;Schwark et al., 2002). The fundamental assumption of this approach is that the n-alkane chain-lengths C 27 and C 29 are thought to be produced by deciduous trees/shrubs and the chainlengths C 31 and C 33 by grasses/herbs (Bliedtner et al., 2018;Marseille et al., 1999;Schwark et al., 2002). This holds also true for the Khar Here, the aforementioned allochthonous input of eroded topsoil material possibly leads to higher production rates of aquatic (algae) n-alkanes via nutrient loading, which can also produce to some amounts the chain-length nC 27 (Ficken et al., 2000) and possibly contribute to the n-alkane ratio. In this context, it is notable that the P aq 0 and the n-alkane ratio are significantly anti-correlated (r = −0.73; α < 0.05; Table 1)   Those findings are supported by the conspicuous distribution of the element S, which shows highest concentrations at the central southern shore (Figure 4c). Sulphur is a part of the proteinogenic amino acids which are involved in several enzyme-reactions during the degradation of plants and faunal remains in the digestive tract and partly defecated (Brosnan & Brosnan, 2006;Kertesz & Mirleau, 2004 Traces of anthropogenic activity and soil erosion in the Khar Nuur sediments are further supported by the significant correlation of TOC, N and S (r between 0.66 and 0.96; α < 0.05). Thus, anthropogenic activity in the catchment by grazing and human occupation most likely leads to the characteristic spatial distribution of those elements in the lake. However, trace-elements of anthropogenic activity are difficult to disentangle and more detailed analyses of source-specific biomarkers are necessary to confirm this assumption.

| CONCLUSIONS
Our grain size and biogeochemical investigations on surface sediment samples from Lake Khar Nuur, a small lake with a small hydrological catchment from the Mongolian Altai, gave the following results: • The grain size distribution shows higher transport energy levels at the shoreline compared to the two basins of Lake Khar Nuur. The accumulation of finer sediments in those basins is accompanied by high amounts of the inorganic sediment compounds Al and Fe, which are of allochthonous origin. Consequently, the two basins act as the main sediment accumulation zones where sediment focussing is observed.
• Disentangling the autochthonous and allochthonous origin of organic material in the surface sediments is not trivial. While TOC and N accumulate preferentially in the accumulation zones, molar TOC/N ratios indicate high autochthonous production throughout the lake, which is mostly underlined by 15 N enriched δ 15 N values.
The respective 13 C depleted δ 13 C TOC values could indicate contributions from both allochthonous C3 plants and autochthonous production or at the shorelines changes in available dissolved inorganic carbon (CO 2 ) for aquatic use. However, those proxies reflect the whole heterogeneous organic matter pool which most likely comprise a mixture of both allochthonous and autochthonous sources. In contrast, the more homogenous and source-specific nalkanes are constituents of the organic matter pool and indicate that n-alkanes from allochthonous and autochthonous sources are apparent in the surface sediments with higher proportion of allochthonous compounds compared to autochthonous ones. from Lake Khar Nuur, where nC 31 is the dominant n-alkane throughout the lake. This indicates that grasses become preferentially incorporated into the lake, also at those sites in the catchment where Betula nana (L.) is growing.
• Anthropogenic activity in the southern Khar Nuur catchment is clearly observable in the surface sediments from the central southern shoreline since TOC, N and δ 15 N indicate increased allochthonous input that is most likely due to increased soil erosion. This is accompanied by increased input of faecal remains as possibly indicated by increased contributions from the element S. However, those are only indirect indicators of anthropogenic activity and more direct analyses such as source-specific biomarkers are necessary to confirm this assumption.
Overall, sediments deposited in the accumulation zones in the two basins of Lake Khar Nuur reflect the recent environmental conditions of the lake and its hydrological catchment and are therefore ideal sediments for paleoenvironmental reconstructions. However, the proxies which possibly will be used in such paleoenvironmental studies have to be carefully evaluated by modern reference studies, although one has to be aware that environmental changes can affect the spatial distribution of sediment compounds through time. Future work on surface sediments should focus on dating the different organic matter compounds for more information about the time of the integrated signal, as well as on compound-specific isotopes of the more homogenous organic matter compounds.

DECLARATION OF COMPETING INTEREST
The authors declare no competing interests.

DATA AVAILABILITY STATEMENT
All data used in this study is available either from the corresponding author on reasonable request or from the public resources as referenced in the text.