Holocene lake carbon sequestration,hydrological status and vegetation change,China

LingMei Xu,Yu Li,WangTing Ye,XinZhong Zhang,YiChan Li,YuXin Zhang

Key Laboratory of Western China's Environmental Systems (Ministry of Education), College of Earth and Environmental Sciences, Center for Hydrologic Cycle and Water Resources in Arid Region, Lanzhou University, Lanzhou, Gansu 730000,China

ABSTRACTLakes have received considerable attention as long-term sinks for organic carbon (C) at regional and global scales. Previous studies have focused on assessment and quantification of carbon sinks, and few have worked on the relationship between millennial-scale lake C sequestration, hydrological status and vegetation, which has important scientific significance in improving our understanding of lake C stocks and storage mechanisms. Here, we present a comprehensive study of pollen records,organic geochemical proxies,lake-level records,sediment accumulation rate(SAR)and organic C accumulation rate (CAR)in China since the Holocene.We also include numerical climate classification and lake-level simulations,to investigate variations of lake C sequestration,hydrological status and vegetation during the Holocene.Results indicate that the evolution of lake C accumulation showed an out-of-phase relationship with hydrological status and vegetation in China.Lake C accumulation exhibited an overall trend of increasing from the early to late Holocene in response to gradually increasing terrestrial organic matter input. However, China as a whole experienced the densest vegetation cover in the middle Holocene, corresponding to the mid-Holocene optimum of a milder and wetter climate. Optimal hydrological conditions were asynchronous in China; for example, early Holocene in Asian monsoon dominated areas, and middle Holocene in westerlies controlled regions.Our synthesis indicated that climate change was the main factor controlling the long-term variability in lake C accumulation,hydrologic conditions,as well as vegetation,and human influences were usually superimposed on the natural trends.

Keywords:lake sediment;millennial-scale;organic carbon accumulation;lake hydrological status;vegetation;climate change

1 Introduction

Distributed across various geological, geographical and climatic regions, lakes are generally sensitive to changes in the surrounding environment and the impacts that such changes have on their lacustrine deposit are of considerable importance (Wang and Dou,1998; Kortelainen et al., 2004; Elliott et al., 2006;Shen, 2013; Wang et al., 2015). In this regard, the C sequestration capacity of lakes easily varies according to the alterations in hydrological status and surrounding vegetation (Martini et al., 2006; Chen et al., 2008;Zhang et al., 2013;Wang et al., 2015). However, most studies on lakes in China were confined to estimating the total amount of C burial using lake records, while the long-term C accumulation of lakes and their responses to climate change have received little attention (Duan et al., 2008; Gui et al., 2013; Wang et al.,2015). For a more thorough understanding about the role of lakes in regional C cycling, the relationship between organic C sequestration, hydrological status and vegetation conditions of different lake types in various regions needs to be investigated (Kling et al., 1991;Herzschuh, 2006; Cole et al., 2007; Li and Morrill,2010;Zhang et al.,2013;Wang et al.,2015).

The formation of lakes in China generally has no zonality, distributed through all the climatic zones including tropic, temperate and frigid zones with altitudes ranging from 5,000 m above sea level (a.s.l.)to -155 m a.s.l. (Wang and Dou, 1998; Shen, 2013;Wang et al., 2015). In spite of covering only about 0.9% of China's entire territorial area, lakes show a significant carbon sequestration potential with total C accumulation rate of 1.98 Tg C/a (Duan et al., 2008),indicating China's lakes cannot be ignored in consideration of the regional C cycle. Recently, studies have focused on the long-term C accumulation of lakes.Wang et al. (2015) synthesized the records of C sequestration from 58 lakes with direct organic C measurements throughout China, and suggested that the Holocene lake C accumulation showed an overall increasing trend since 12 ka. Another synthesis derived from 42 palaeolake records in China indicated a high accumulation rate for lakes during the middle Holocene relative to the late Holocene (Zhang et al.,2013).Millennial-scale variations in lake hydrological status and vegetation have been conducted by many studies using numerous proxy-data from lakes (An et al.,2000; Yu et al., 2000; Herzschuh, 2006; Parmesan,2006;Chen et al.,2008;Qian et al.,2009;Long et al.,2010; Liu et al., 2013; Zhang et al., 2016). The most comprehensive lake-level study of Shen (2013) indicated that lakes on the Tibetan Plateau experienced a persistent and stable high lake level in the early Holocene, whereas overall high lake-levels with maximum moisture conditions were observed during the middle Holocene in the mid-latitude arid Asian region and the monsoon area. Li and Morrill (2010) reported that most lakes of the monsoonal Asia began to expand during the post-glacial, and reached the highest stands in the early Holocene. As for vegetation of China, Yu et al. (2000) provided two state-of-the-art palaeovegetation maps using modern and fossil pollen data, in which the forest biomes in eastern China systematically shifted northwards and extended westwards during the middle Holocene while steppe and even desert vegetation extended to eastern China at the last glacial maximum. At present, numerous studies have been conducted for lakes of China,but we still lack a fundamental and comprehensive understanding of the relationship between lake C accumulation, hydrological status and vegetation of China. In this paper, we synthesized records from 51 lakes in various K?ppen climate zones and different geographical regions of China; investigated the Holocene SAR and CAR variations of lakes at 1,000-year bin; assessed the factors controlling the temporal and spatial evolution of C accumulation;reconstructed millennial-scale lake hydrology and vegetation changes in different regions; as well as analyzed the relationship between lake C accumulation, hydrological status and vegetation. Our results provide insights into the mechanism of the carbon balance, and demonstrate a new perspective on global and regional carbon cycle research.

2 Materials and methods

2.1 Data set and analyses

Holocene lake records in this paper were considered multi-proxy data from different profiles for an individual lake (Figure 1, Table 1). We have selected records based on three criteria: (1) the record should have reliable chronologies and successive sedimentary sequences but not have depositional hiatuses during the Holocene; (2) indicators derived from the records must include sediment organic carbon (OC) and (3) lake-level records should be indicated by the effective proxies.Therefore, 51 lakes derived from various K?ppen climate zones and different regions of China were selected. All radiocarbon ages (14C dates) of individual lake were first corrected,and any possible old-carbon effects were removed according to original publications (Chen et al.,2008).Then,the corrected14C ages were calibrated to calendar years (cal. a B.P.) using the program of Calib 6.1.0 (Stuiver and Reimer, 1993). The calibrated ages were used to compile the variations of sediment and C accumulation rate, organic geochemical proxies,and total pollen concentration throughout the text (Figures S1-S4). Lake-level records, site information and representative pollen indicators of lakes in China are presented in Tables 1,S2 and 3,respectively.

2.2 Estimate of organic C accumulation rate

The organic carbon accumulation rate (CAR,g C/(m2·a)) at 1,000-year bins of deposition in each lake was gathered in this paper using the following equation(Müller et al.,2005)(Table S1):

Lake sediment accumulation rate (SAR, mm/a)was typically established by14C chronology. Organic carbon content (OC, %) was directly derived from 51 lake sediment records, and sediment density (ρ) and porosity (φ) could be calculated using Equation (2)(Alin and Johnson, 2007) and Equation (3) (Danielson and Sutherland,1986;Avnimelech et al.,2001).

The dry bulk density (DBD, g/cm3) for lake sediments without a measured value was then obtained using empirical relationships(Dean and Gorham,1998;Avnimelech et al.,2001;Kastowski et al.,2011;Wang et al.,2015).

Equation (4) was given by Dean and Gorham(1998), and Equation (5) was derived from Avnimelech et al.(2001).

Figure 1 Overview map showing lake sites selected in this study(numbers in symbols refer to lakes described in Table 1)and the geographical zones of China:the Qinghai-Tibet Plateau(TP),the monsoonal regions(MR)and northwest arid and semi-arid China(NAC)

2.3 Environmental variables and lake characteristics

This paper explored the relationship between lake average CAR of past 12 ka and a series of environmental variables or lake characteristics. Annual temperature, summer temperature, annual precipitation and summer precipitation were calculated based on the datasets of 0.5°×0.5° monthly gridded precipitation and temperature during 1961-2013 released by the National Meteorological Information Center (http://cdc.cma.gov.cn).Altitude, annual evaporation, mean depth,max depth,surface areas and catchment area of lakes were partly from Wang et al. (2015) and partly from available publications(Table S2).

2.4 Numerical climate classification

Dating from 1900, the K?ppen-Geiger climate classification system continues to be the most widespread climate system(Lohmann et al.,1993;Peel et al.,2007;Baker et al., 2010; Rubel and Kottek, 2010; Naipal et al.,2015).Based on the 0.5°×0.5°monthly gridded precipitation and temperature datasets of 53 years during 1961-2013, and a 0.5°×0.5° gridded elevation data derived from a quality-controlled global Digital Elevation Model (DEM), this paper produced a K?ppen's climate classification map for China,using the method given by Zhu and Li(2015).The map divided China's climate into 5 main climate groups, 12 types and 28 subtypes.Figure 5 shows that China's main climate regions are Arid (B),Temperate (C), Cold (D), Polar (E), and main climate types include steppe climate (Semi-arid) (Bs),desert climate (Bw), mild temperate with dry winter(Cw), mild temperate, fully humid (Cf), snow with dry winter(Dw),and tundra climate(ET).

2.5 Lake-level simulations

We applied a set of models, the community climate system model (CCSM 3.0), a lake energy-balance and a lake water-balance model, to verify long-term changes in lake-level for the early (8.5 ka), middle (6.0 ka)and late (pre-industrial (PI)) Holocene. Model details are clearly described by Li and Morrill(2010).

Table 1 Information on lake records for organic C accumulation rate,lake level and vegetation changes of China 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Bangong Co Songxi Co Naleng Lake Chen Co Nam Co Zigetang Co Ngoin Co Paru Co Pumoyum Co Koucha Lake Ximen Co Qinghai Lake Donggi Cona Lake Gahai Lake Chaka Salt Lake Genggahai Lake Erlongwan Maar Lake Xingkai Lake TOC,Pollen TOC TOC,C/N,δ13C,Pollen TOC,C/N,Pollen TOC,C/N,Pollen TOC,δ13C,Pollen TOC,C/N,δ13C TOC TOC TOC,δ13C TOC TOC,C/N,δ13C,Pollen TOC,C/N TOC,C/N,δ13C,Pollen TOC TOC,C/N,δ13C TOC,C/N,δ13C,Pollen TOC,C/N,δ13C,Pollen TP TP TP TP TP TP TP TP TP TP TP TP TP TP TP TP MRMR-+*+-+--+**+++-- ++//*+-/-////+-- /- ++//*//+-//*/+/+/-*-+/*++-/////+/-//**Arid Arid Polar Polar Polar Polar Polar Polar Polar Polar Polar Cold Polar Arid Arid Arid Cold Cold Desert Steppe Tundra Tundra Tundra Tundra Tundra Tundra Tundra Tundra Tundra Dry Winter Tundra Steppe Steppe Steppe Dry Winter Dry Winter Cold Cold/////////Cold Summer/Cold Cold Cold Warm Summer Warm Summer High Low High High High Low Low High High Low Low Low Low High Low Low Low High Low Low Low Low Low High Low High High Low Low High Low Low Low Low Low High Low Low Low Low Low High Low High Low Low High High Low Low Low Low High High High Low High High High High High High High High High High High High Low High Low Low High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High Low High High High High Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low Low High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High ID Site Proxies Region Relationship between CAR and proxies TOC C/N δ13C TPC K?ppen climate zones 1st 2nd 3rd CAR 8.5-6 ka 8.5 ka-PI 6 ka-PI Lake level 8.5-6 ka 8.5 ka-PI 6 ka-PI Vegetation 8.5-6 ka 8.5 ka-PI 6 ka-PI to be continued

19 20 21 22 23 24 25 26 27 28 29 30 3132 33 Sihailongwan Maar Lake Moon Lake Erhai Lake Dian Ch i Xima Ch i Xingyun Lake Lugu Lake Hugangyan Maar Lake Dajiuhu Lake Gucheng Lake Jiangling Dahu Lake Chaohu Lake Taihu Lake Beihuqiao TOC,C/N,δ13C TOC,C/N,δ13C,Pollen TOC,δ13C,Pollen TOC,C/N,δ13C,Pollen TOC TOC,C/N,δ13C,Pollen TOC,Pollen TOC,C/N,δ13C,Pollen TOC,C/N,δ13C,Pollen TOC TOC,C/N,δ13C,Pollen TOC,C/N,δ13C,Pollen TOC,Pollen TOC,Pollen TOC,C/N,δ13C MR MR MR MR MR MR MR MR MR MR MR MR MRMR MR-++++-+++-+++++-*/+/+/+*/+- //---- -/+/+ +/- - / / +/-- +/++++/*+*- /Cold Cold Temperate Temperate Temperate Temperate Temperate Temperate Temperate Temperate Temperate Temperate Temperate Temperate Temperate Dry Winter Dry Winter Dry Summer Dry Summer Dry Summer Dry Winter Dry Winter Dry Winter Dry Winter Dry Winter Without dry season Without dry season Without dry season Without dry season Without dry season Warm Summer Cold Summer Warm Summer Warm Summer Warm Summer Warm Summer Warm Summer Warm Summer Hot Summer Hot Summer Hot Summer Hot Summer Hot Summer Hot Summer Hot Summer High Low High Low High Low Low High Low Low Low High Low High High High Low High Low Low High Low High Low/Low High///High Low High Low Low High Low High Low/Low Low///Low High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High High Low Low Low Low Low Low Low High Low Low Low High Low Low Low High High High High High High High High Low High High High Low High High High High High High High High High High High High High Low High High High ID Site Proxies Region Relationship between CAR and proxies TOC C/N δ13C TPC K?ppen climate zones 1st 2nd 3rd CAR 8.5-6 ka 8.5 ka-PI 6 ka-PI Lake level 8.5-6 ka 8.5 ka-PI 6 ka-PI Vegetation 8.5-6 ka 8.5 ka-PI 6 ka-PI to be continued Table 1 Information on lake records for organic C accumulation rate,lake level and vegetation changes of China

343536 37 3839 40 41 42 43 44 45 46 47 48 49 50 51 Longgan Lake Sanqing Ch i Bosten Lake Aibi Lake Yitang Lake Balikun Lake Wulungu Lake Manas Lake Zhuye Lake Juyanze Lake Huahai Lake Sanjiaocheng Daihai Lake Dali Lake Gonghai Lake Hulun Lake Tengger Nuur Yanhaizi Lake TOC TOC TOC,C/N,Pollen TOC TOC,C/N,δ13C TOC,Pollen TOC,C/N,δ13C,Pollen TOC,Pollen TOC,C/N,δ13C,Pollen TOC,Pollen TOC,C/N TOC,δ13C,Pollen TOC,C/N,Pollen TOC,C/N,δ13C TOC TOC,δ13C,Pollen TOC,Pollen TOC,C/N MRMR NAC NAC NAC NAC NAC NAC NAC NAC NAC NAC NAC NAC NAC NAC NAC NAC++*+-+-++++- - *-+-+//*/+/-/+/++- +///+////-/-/-////*/-////*//+++-*/++//-*/Temperate Cold Arid Arid Arid Arid Arid Arid Arid Arid Arid Arid Cold Cold Cold Arid Arid Arid Without dry season Dry Winter Desert Desert Desert Steppe Steppe Steppe Desert Desert Desert Desert Dry Winter Dry Winter Dry Winter Steppe Desert Desert Hot Summer Warm Summer Cold Cold Cold Cold Cold Cold Cold Cold Cold Cold Warm Summer Warm Summer Warm Summer Cold Cold Cold High/Low/High High Low Low Low Low High Low Low Low Low Low High High Low/Low/High Low Low Low Low Low High Low Low Low Low Low High High Low Low Low High Low Low Low Low Low Low High Low Low Low Low Low Low Low High/Low Low Low Low Low Low High High Low Low High High High Low High High High/Low Low High Low Low Low High High High High High High High High High High High High High High High High High High High Low High High High High High High High High Low/Low Low Low Low Low High Low High Low High Low Low Low Low Low Low High/Low High High Low Low High High High High High High High High High High High High High High High High High High High High High High High High High High High High High ID Site Proxies Region Relationship between CAR and proxies TOC C/N δ13C TPC K?ppen climate zones 1st 2nd 3rd CAR 8.5-6 ka 8.5 ka-PI 6 ka-PI Lake level 8.5-6 ka 8.5 ka-PI 6 ka-PI Vegetation 8.5-6 ka 8.5 ka-PI 6 ka-PI to be continued Tab le 1 Information on la ke records for organic C accumulation r ate,lake level and vegetatio n changes of China

3 Results

3.1 Organic geochemical proxies, pollen records and lake C accumulation rate

This paper describes the relationships between organic geochemical proxies, total pollen concentrations (TPC) and lake C accumulation rate according to their changing trends during the Holocene (Figures S1-S4).The"+"shows a similar changing trend of paleoclimatic indicators and CAR, the "-" represents the opposite changing trend, and "*" suggests completely different trends between them (Table 1).Therefore, we divided China into the Qinghai-Tibet Plateau (TP), monsoonal regions (MR) and northwest arid and semi-arid China (NAC), based on physiography, taking into account the relationships between the indicators and CAR (Figure 1, Table 1).Lakes in TP show totally different trends of TOC,C/N,δ13C,TPC and CAR.In addition,positive correlations between TOC, C/N, TPC and CAR were found in most lakes of MR and NAC, but the relationship of δ13C and CAR varied among the regions,for example; positive in southern MR, negative in northern MR and NAC.

3.2 Organic C, sediment and C accumulation

The variations of lake sediment OC content, sediment and C accumulation rate varied significantly by region and through time during the Holocene (Figure 2). Lakes in TP, MR and NAC show completely different changing trends in OC and SAR,while CAR of lakes indicate a growing trend in most regions, except for MR during the late Holocene(Figure 2).The overall OC content of lakes in China suggest a mid-Holocene peak, but sediment and C accumulation rate exhibited a similar changing trend of increasing from the early to late Holocene(Figure 2).

The sediment OC content of Chinese lakes ranged from 0.27 (Songxi Co) to 45.90% (Dahu Lake), with a mean value of 5.17%; lake sediment accumulation rate ranged from 0.07 (Xingkai Lake) to 3.44 mm/a(Gucheng Lake), with a mean value of 0.64 mm/a;lake C accumulation rate ranged from 0.40 (Xingkai Lake) to 63.92 g C/(m2·a) (Gucheng Lake), with a mean value of 8.07 g C/(m2·a). The observed peak of the sediment OC content was 8.60% in MR, three and two times higher than the average content of TP and NAC(Table S1).The average sediment and C accumulation rate were 0.56 mm/a, 6.95 g C/(m2·a) in TP;0.61 mm/a,9.45 g C/(m2·a)in MR;and 0.75 mm/a,7.60 g C/(m2·a) in NAC, which did not differ significantly among the regions (Table 2).The lake C burial in China was estimated as 8.80 Pg, with the highest value of 3.80 Pg contributed by lakes in the TP,intermediate value of 3.00 Pg by those in the MR, and the lowest of 2.00 Pg by those in NAC(Table 2).

Figure 2 Temporal variation patterns of regional organic C content(a),sediment accumulation rate(b)and C accumulation rate(c)of lakes in China

3.3 Lake sediment CAR and environmental variables or lake characteristics

Correlation analysis was applied to evaluate the factors controlling lake carbon sequestration in China(Figure 3, Table S2). Results suggest that the average CAR of overall lakes in China was not significantly related to environmental variables or lake characteristics,but there were close correlations between them in various regions of TP, MR and NAC (Figure 3). The Holocene mean CAR of lakes in TP was negatively correlated with altitude (R2=0.18, α=0.001) (Figure 3f), water mean depth (R2=0.31, α=0.001) (Figure 3g)and water maximum depth (R2=0.16, α=0.001) (Figure 3h). Lake CAR in MR show positive correlations with summer temperature (R2=0.12, α=0.001) (Figure 3d) and evaporation (R2=0.68, α=0.001) (Figure 3e),and negative correlation with lake area (R2=0.25,α=0.001) (Figure 3i). Lake CAR in NAC had positive correlations with summer precipitation (R2=0.12,α=0.001) (Figure 3c), water mean depth (R2=0.19,α=0.001) (Figure 3g) and water maximum depth(R2=0.24,α=0.001)(Figure 3h).

Table 2 Lake areas(Duan et al.,2008),organic C content,sediment and C accumulation rate,as well as estimated regional C burial for various geographical zones of China

Figure 3 Relationships between Holocene C accumulation rate and environmental variables or lake characteristics.(a)Annual mean precipitation;(b)annual mean temperature;(c)mean summer precipitation;(d)mean summer temperature;(e)evaporation;(f)altitude;(g)mean water depth;(h)maximum water depth;(i)log(Lake area,km2);and(j)catchment area/lake area.Colors show different regions in China:TP(green circles),MR(pink squares),NAC(blue triangles)

3.4 Lake simulations and lake-level records

Holocene lake-level changes in China were simulated by CCSM 3.0, lake energy and water balance model, and results show that simulations were consistent with most lake level records (Figure 4).Namely, lake level in MR and TP exhibited a coherent changing trend of decreasing from the early to late Holocene, while overall high lake-levels in NAC were observed in the middle Holocene (Figure 4,Table 1).

3.5 Lake C accumulation, hydrological status,vegetation

Millennial-scale variations in lake C accumulation, hydrological status and vegetation in China between the early (12-8 ka), middle (8-4 ka) and late(4-0 ka) Holocene were calculated or reconstructed in this study (Figure 5, Table 1). Lake C accumulation of most lakes in China suggest a growing trend of increasing from early to late Holocene (Figure 5a), whereas a decreasing trend for the evolution of lake hydrological status through the Holocene was found in the regions of TP, MR, eastern NAC, and lakes in western NAC show optimal hydrological conditions in the middle Holocene(Figure 5b).In addition, overall lakes in China experienced favorable vegetation conditions in the middle Holocene, and underwent their worst vegetation conditions at the late Holocene (Figure 5c). Consequently, lake C accumulation, hydrological status and vegetation of China indicated relatively strong temporal and spatial variability in different geographical areas, while they were not synchronous during the Holocene in each K?ppen-Geiger climate zone.

Figure 4 Simulated lake-level changes for 8.5 ka minus 6.0 ka(a),8.5 ka minus PI(b),and 6.0 ka minus PI(c).Points show locations of lake sites,with color indicating the direction of lake level change implied by the paleo record

Figure 5 The temporal and spatial evolution of Holocene lake C accumulation(a),hydrological status(b)and vegetation(c).Colors show different evolution patterns of lakes in China:PI>6 ka>8.5 ka(green points);PI>8.5 ka>6 ka(orange points);8.5 ka>6 ka>PI(red points);8.5 ka>PI>6 ka(blue points);6 ka>8.5 ka>PI(yellow points);6 ka>PI>8.5 ka(purple points);no data available(pink points).The K?ppen's climate classification map for China is calculated based on the 0.5°×0.5°monthly gridded precipitation and temperature datasets of 53 years during 1961-2013,and a 0.5°×0.5°gridded elevation data derived from a quality-controlled global Digital Elevation Model(DEM).This map only contains the mainland of China,and the south China sea,Taiwan region,Diaoyu Island,etc.are not include in the map

Table 3 Information on fossil pollen records in representative pollen indicators from China 1 2 3 4 5 6 7 8 9 10 11 1213 1415 16 17 18 19 20 21 22 Bangong Co Songxi Co Naleng Lake Chen Co Nam Co Zigetang Co Ngoin Co Paru Co Pumoyum Co Koucha Lake Ximen Co Qinghai Lake Gahai Lake Chaka Salt Lake Donggi Cona Lake Genggahai Lake Erlongwan Maar Lake Xingkai Lake Sihailongwan Maar Lake Moon Lake Erhai Lake Dian Chi High//High High High//High High/High High///High High/Low Low/High//Low High High//High High/High High///High High/Low Low/High//Low High High//Low Low/Low High///Low Low/High Low/High/High/High High//High High/Low High///High Low/High//High/Low/Low Low//High High/Low High///High High/High//Low/Low/Low Low//Low Low/Low Low///Low High/Low////Low High Low High///Low/Low High///High Low/Low Low///Low High Low Low///Low/Low High///Low Low/Low High///Low High Low Low///High/High High///Low Low/Low High///Low Low/Low//High High/High High///High Low/Low High///High Low/Low//High High/High High///High High/Low High///High High/High//High High/High High///Low High/High High/Low/High Low Low Similar//High Low/High Low///High High/High//High/High High Low Low//Low Low/High High///High High/Low//High/Low High Low Low//Low High/Low High///Low High/Low/////////////High////Low Low/Low Low////////////High////Low Low/Low Low////////////High////High Low/Low Low/Low//Low////High High/High High///High High/Low Low/Low//Low////High High/Low High///High High/Low High/High//Low////Low Low/Low Low///Low High/High High///Low/Low///Low//Low///////Low Low Low//High/Low///Low//High///////Low High Low//High/High///High//High///////High High High////////High///////High//Low Low High////////High///////High//Low High Low////////Low///////Low//High High Low ID Site ArtemisiaⅠⅡ ⅢChenopodiaceaeⅠⅡⅢPinusⅠⅡⅢBetulaⅠⅡⅢCyperaceaeⅠⅡⅢQuercusⅠⅡⅢA/CⅠⅡ ⅢTreesⅠⅡ ⅢHerbsⅠⅡⅢto be continued Notes:Ⅰ,8.5-6 ka;Ⅱ,8.5 ka-PI;Ⅲ,6 ka-PI

23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 Xima Chi Xingyun Lake Lugu Lake Dajiuhu Lake Huguangyan Maar Lake Jiangling Dahu Lake Chaohu Lake Taihu Lake Gucheng Lake Beihuqiao Longgan Lake Sanqing Chi Bosten Lake Aibi Lake Yitang Lake Balikun Lake Wulungu Lake Manas Lake Zhuye Lake Juyanze Lake Huahai Lake/Low High Low High Low/High Low Low///High Low/Low Low/Low Low//Low High Low Low Low/Low High High///High High/Low Low/Low Low//Low Low Low Low High/Low High High///High High/Low High/High Low////High High High/Low Low Similar///High Low/High Low/Low High////High High Low/Low High Similar///High High/High Low/Low High////High Low Low/Low High High///Low High/Low Low/High Low//Low Low Low/High High High Low Low////Low/Low//High Low/Ⅱ/Low High Low/Low Low Low Low High////High/Low//High Low/Ⅲ/High High High/Low Low Low Low High////High/High//Low High//Low High Low/Low/Low Low///////Low Low//High/Ⅱ/Low High Low/Low/Low Low///////High Low//High/Ⅲ/High Low High/Low/Low Low///////High High//Low//Low Low/High Low Low Low Low////High//Low///High/Ⅱ/High Low/Low Low Low High Low////High//Low///Low/Ⅲ/High Low/Low High High High Low////High//Low///Low//Low Low Low Low High High Low High/////Low////Low Low//Low Low Low Low High High High High/////Similar////High Low//High High High High High High High High/////High////High High///Low Low High Low/High High Low///Low Low/Low High Low High Low///High Low Low Low/High High High///Low High/Low High High High Low///High Low Low Low/High High High///High High/High High High High High////Low High Low High High/////////Low/High/////Low High High High High/////////Low/High/////High High High Low High/////////High/High///Low/Low Low High High High/////Low//Low//Low//Ⅱ/Low/High Low High Low High/////High//High//Low//Ⅲ/Low/High Low High Low High/////High//High//High//ID Site ArtemisiaⅠⅡ ⅢChenopodiaceaeⅠⅡⅢPinusⅠBetulaⅠCyperaceaeⅠQuercusⅠⅡⅢA/CⅠⅡ ⅢTreesⅠⅡ ⅢHerbsⅠto be continued Table 3 Information on fossil pollen records in representative pollen indicators from China Notes:Ⅰ,8.5-6 ka;Ⅱ,8.5 ka-PI;Ⅲ,6 ka-PI

45 46 47 48 49 50 51 Sanjiaocheng Daihai Lake Dali Lake Gonghai Lake Hulun Lake Tengger Nuur Yanhaizi Lake Low High//High High/Low Low//High High/High Low//Low High/Low Low//Low High/Low Low//Low Low/Low Low//Low Low/High Low//Low Low/Low High//Low Low/Low High//Low High//High//High High//High//High High//High//High High/////Low Low/////Low High/////Low High//Low///High//High///High//High///High/Low High//High Low/High High//High High/High High//High High//Low//Low///High//Low///High//High///High//High///Low//High///Low//Low//ID Site ArtemisiaⅠⅡ ⅢChenopodiaceaeⅠⅡⅢPinusⅠⅡⅢBetulaⅠⅡⅢCyperaceaeⅠⅡⅢQuercusⅠⅡⅢA/CⅠⅡ ⅢTreesⅠⅡ ⅢHerbsⅠⅡⅢto be continued Table 3 Information on fossil pollen records in representative pollen indicators from China Notes:Ⅰ,8.5-6 ka;Ⅱ,8.5 ka-PI;Ⅲ,6 ka-PI

4 Discussion

4.1 Lake C accumulation

This study found that most lakes in China show an overall increasing trend in C accumulation during the Holocene, with relatively high accumulation rate in the middle or late Holocene. The high accumulation rate during the middle Holocene was closely related to an optimal environment characterized by a warm and humid climate (Zhang et al.,2013;Wang et al., 2015). However, the high accumulation rate during the late Holocene could be explained by terrestrial organic matter input due to land-use change (Kortelainen et al., 2006; Liu and Fan, 2010; You and Liu,2012;Wang et al.,2015).

The C sequestration capacity of lakes is very sensitive to alterations of the surrounding environment(Martini et al., 2006; Chen et al., 2008; Zhang et al.,2013; Wang et al., 2015). Organic geochemical proxies and pollen records, with abundant information of regional climate and environment during the deposition period, play an important role in Holocene climate change studies (Davis, 2000; Balascio et al.,2013; Li et al., 2013). Therefore, this paper divided China into TP, MR and NAC, based on physiography,taking into account the relationships between organic geochemical proxies, total pollen concentration and lake C accumulation rate. Lakes of TP show totally different trends of the proxies and CAR due to the low biomass compared with other regions of China.Positive correlations between TOC, C/N, TPC and CAR were found in most lakes of MR and NAC, that is, high TOC, C/N, TPC usually corresponded to warm-humid climatic conditions. However, the relationship of δ13C and CAR varied among the regions,for example; positive in southern MR, negative in northern MR and NAC, which could be attributed to climate effects of plant δ13C (O'Leary, 1988; Li et al.,1999). Pyrophilous C4plants with light δ13C values were widespread in the warm-humid East Plain of southern MR, while cold or dry conditions prevailing across northern MR and NAC caused the expansion of pyrophilous C3plants with heavy δ13C (Yin and Li,1997;Li,1999;Ehleringer et al.,1997;Ehleringer and Bj?rkman,1977;Zhou et al.,2013).

Regression analysis suggest that the average CAR of past 12 ka for all lakes in China was not significantly related to lake characteristics (Figure 3), which illustrated that lake C accumulation is a result from comprehensive function of multiple factors, involving geological,physical,chemical,biological and anthropic processes (Shen, 2013). However, on regional scales, several factors could partly explain the millennial-scale variability in lake C accumulation. In NAC,lake CAR is positively correlated to summer precipitation; indeed, vegetation growth and watershed hydrology in arid areas were strongly influenced by moisture conditions and the rainfall of the regions was mainly concentrated in the summer months (Wang et al., 2005; Romero et al., 2012, 2013). Additionally,positive relationship between lake CAR and water depth was observed in NAC, because the shallower lake water was favorable for the oxygenic mineralization of lake sediments (Sobek et al., 2013; Wang et al.,2015).In MR,high summer temperature generally means higher primary productivity, associated with abundant lake organic C produced in situ or transported from terrestrial ecosystems (Zhang et al., 2013;Wang et al., 2015), and thus positive relationship between lake CAR and temperature was found in MR.However, the low C burial in large lakes of MR could be attributed to the high oxidation of terrestrially fixed C to CO2(Kortelainen, 2004). Negative correlation between lake CAR and altitude in TP is easy to understand such that high altitude usually correspond to sparse vegetation with relatively low temperature and moisture. Except for climate factors, lake C accumulation was also impacted by human activity.Extensive agricultural practices and deforestation accelerated soil erosion and thus considerably increased sediment accumulation of lakes (Whitmore et al., 1994;Saito et al.,2001;Ran and Lu,2014).

4.2 Lake hydrological status

The Holocene lake hydrologic changes for different geographical regions were reconstructed based on lake level records derived from 51 lakes of China(Figure 4, Table 1). The regions of MR, TP and eastern NAC show gradually weakened hydrological conditions through the Holocene, namely a wetter early Holocene, a moderately wet middle Holocene, and a dry late Holocene. However, lakes of western NAC experienced favorable hydrological status in the middle Holocene, suboptimal hydrological conditions at the early Holocene and poor hydrological status during the late Holocene.

Lakes in MR,TP and eastern NAC were dominated by the Asian monsoon and those in western NAC were influenced by the westerly winds (Li and Morrill, 2010; Shen, 2013). Palaeoclimate studies revealed that strong summer Asian monsoons and weakened mid-latitude westerlies prevailed during the early Holocene (Herzschuh, 2006; Chen et al., 2008; Li and Morrill, 2010; Jin et al., 2012; Shen, 2013). As a result, considerable precipitation with less lake surface evaporation occurred in Asian monsoon dominated areas, while low-level water vapor transport from the North Atlantic was observed in the westerlies controlled regions (Parmesan, 2006; Chen et al., 2008;Jin et al., 2012; Shen, 2013). Another explanation of the early Holocene regional difference in lake hydrological status could be attributed to air movement induced by summer insolation (Ye and Gao, 1979;Rodwell and Hoskins, 1996; Herzschuh, 2006; Chen et al., 2008). Intense heating of low latitudes region caused intensified subsidence of air masses to TP and inevitably dry climate in NAC (Ye and Gao, 1979;Broccoli and Manabe, 1992; Rodwell and Hoskins,1996; Herzschuh, 2006; Chen et al., 2008; Shen,2013). Moreover, under deglacial boundary conditions at high latitude, cool air and sea-surface might have reduced water vapor transport to western NAC during the early Holocene (Chen et al., 2008; Li and Xu, 2016). In the middle Holocene, the Asian monsoon began to retreat while the westerlies were intensified and expanded (Fleitmann et al., 2003; Yuan et al., 2004; Shao et al., 2006; Chen et al., 2008; Li and Morrill, 2010). Furthermore, reduction in glacier cover would increase the surface air temperature, thereby inducing more vapor transported to NAC (Numaguti,1999; Chen et al., 2008). Consequently, lake hydrological conditions for areas dominated by the Asian monsoon show gradually weakened hydrological conditions through the Holocene, and optimal hydrological conditions occurred during the middle Holocene in the westerlies controlled regions.It should be pointed out that TP was typically characterized by discharges of glacial-melt water; and thus the weakening lake hydrologic status after the middle Holocene in TP was also connected with ice-sheet reduction (Liu et al.,2013;Shen,2013;Li et al.,2014).

4.3 Vegetation

Based on the synthetic analysis regarding organic geochemical proxies, total pollen concentrations and eight representative pollen indicators of 51 lakes in China, we reconstructed the Holocene vegetation history in various regions (Figures S1-S4, Table 3).Lake sediments in MR indicated less terrestrial plants and aquatic plankton were produced in the early Holocene, and the region was dominated by pine forest with patches of herbs and shrubs. However, notable increase of TOC, C/N and total pollen concentration suggest that paleoecology improved markedly in the middle Holocene such that deciduous broadleaved forest appeared to extent over the northern region of MR and southern region was dominated by evergreen broadleaved forest. During the late Holocene, along with the decrease in temperature and precipitation,the broadleaved forest degenerated and then pine forest expanded again. In addition, indicators from lakes Bosten, Aibi, Yitang, Balikun, Wulungu, and Manas in western NAC suggest an arid early Holocene with widely distributed dry steppe.The increase in Artemisia and A/C ratios between 4-8 ka indicate that the environment became wetter and the regional vegetation was converted to desert-steppe. During the late Holocene, A/C ratios implied a decreasing trend of effective moisture but drainage basins were still covered by desert-steppe. Nevertheless, eastern NAC shows that arid steppe dominated the region in company with patches of pine forests during the early Holocene; intensive mixed pine and broadleaved forests widely developed in the middle Holocene; and arid stepped reexpanded in the late Holocene. The region of TP, with an average elevation of more than 4,000 m a.s.l.,experienced two different patterns of vegetation and landscape evolution through the Holocene but coherently show an optimal vegetation cover in the middle Holocene. In high-altitude regions of northwest TP, millennial-scale reconstructions of vegetation suggest a dominance of alpine meadows in the early and late Holocene, and the coniferous mixed forest tended to dominate the middle Holocene. On the other hand, all types of pollen in southwest TP indicate the presence of desert steppe during the early Holocene, alpine steppes with desert elements in the middle Holocene, and the re-expansion of desert steppe in the late Holocene.

Although watershed vegetation in China show a regionally-coherent pattern where vegetation flourished during the middle Holocene, the limiting factor of the regional vegetation was asynchronous; for example, precipitation and temperature in TP, MR and eastern NAC, and precipitation in eastern NAC.Therefore, in spite of considerable precipitation in TP,MR and eastern NAC during the early Holocene, the vegetation coverage was sparse owing to the cold environment (Fleitmann et al., 2003; Yuan et al., 2004;Shao et al., 2006; Chen et al., 2008; Li and Morrill,2010).In the middle Holocene,the intensity of precipitation was still strong and the temperature reached the warmest period of the Holocene; thus, relatively large-scale plant cover prevailed in those regions(Shen, 2013; Li and Xu, 2016). However, the evolution of lakes in western NAC were generally dominated by mid-latitude westerly winds which brought considerable precipitation in the middle Holocene, thereby flourishing vegetation in this region was observed at the middle Holocene (Tao et al., 2010; Xue and Zhang,2011;Zhang and Li,2015).

4.4 Lake C accumulation, hydrological status and vegetation

Since the migration and deposition of organic matter from upstream to downstream are interlinked by hydrologic cycle, lake C accumulation can be considered as a synthesis of hydrology and vegetation conditions for a drainage basin (Jasper and Gagosian,1990;Turcq et al., 2002; Zhang et al., 2013; Crann et al.,2015; Wang et al., 2015). Indeed, the strong basinwide hydrodynamic conditions usually correspond to abundant exogenous fluxes of organic C,and the luxuriant vegetation in the watershed means considerable sources of terrestrial organic matter. Accordingly, although various climate-forcing mechanisms acted on different geographical regions, lake C accumulation of most individual lakes in this study set was relatively high during the middle Holocene. However, the accumulation peak of lake C in the late Holocene was observed to be closely related to human activity that was generally characterized by pastoralism in TP and agricultural practices in MR and NAC (Makohonienko et al.,2004;Zhao et al.,2007;Han,2014;Ran and Lu, 2014). Human interference accelerated soil erosion, and thus resulted in considerable C burial in lakes and rivers (Whitmore et al., 1994, Saito et al.,2001;Ran and Lu,2014).

5 Conclusions

Millennial-scale variations in lake C accumulation, hydrological status and vegetation in China between the early (12-8 ka), middle (8-4 ka) and late(4-0 ka) Holocene were calculated or reconstructed based on direct organic C values, lake-level records,pollen records and organic geochemical proxies. Our synthesis indicates that variations of lake C accumulation show an out-of-phase relationship with hydrological status and vegetation during the Holocene. Lake C accumulation exhibited an overall trend of increasing from the early to late Holocene in response to gradually increasing terrestrial organic matter input. However, China as a whole experienced the densest vegetation cover in the middle Holocene, corresponding to the mid-Holocene optimum of a milder and wetter climate. Furthermore, a decreasing trend for the evolution of lake hydrological status through the Holocene was found in Asian monsoon dominated areas, while lakes in westerlies controlled regions show optimal hydrological conditions in the middle Holocene.

Table S1 Organic C content,sediment and C accumulation rates of individual lake cores at 1000-year bin during the Holocene 1 2 3 4 5 6 7 8 9 10 Bangong Co Songxi Co Naleng Lake Chen Co Nam Co Zigetang Co Ngoin Co Paru Co Pumoyum Co Koucha Lake OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))2.62 1.23 20.12///5.05 0.70 12.45 0.97 1.42 13.08 1.31 0.47 5.21 1.48 0.61 7.22 1.75 0.73 9.46 33.14 0.27 6.64 1.34 0.24 2.61 15.42 0.10 2.33 2.10 1.24 18.15///5.22 1.17 20.77 0.98 1.36 12.57 1.00 0.51 4.79 1.31 0.63 6.97 2.30 0.61 8.95 28.25 0.28 6.90 1.33 0.25 2.73 11.95 0.90 19.79 1.16 1.25 12.86///5.94 0.72 12.59 0.96 1.42 12.99 1.10 0.28 2.77 1.20 0.62 6.46 3.07 0.75 12.28 29.03 0.30 7.33 1.20 0.27 2.72 11.93 0.51 11.22 0.78 1.25 10.13 0.33 0.28 1.23 6.16 0.66 11.46 0.98 1.33 12.30 1.24 0.19 2.03 1.45 0.63 7.32 2.86 0.59 9.47 25.83 0.32 7.64 1.12 0.28 2.75 11.94 0.37 8.17 0.79 1.26 10.27 0.34 0.15 0.66 5.83 0.65 11.36 0.96 0.46 4.22 1.22 0.16 1.66 1.48 0.62 7.33 1.92 0.55 7.46 25.76 0.33 8.02 1.03 0.29 2.74 12.43 0.29 6.39 1.00 1.27 11.98 0.29 0.11 0.46 6.60 0.64 10.85 1.21 0.38 3.98 1.47 0.16 1.88 1.64 0.61 7.62 3.59 0.51 8.75 24.69 0.35 8.35 0.89 0.31 2.64 11.22 0.24 5.19 0.86 1.28 10.98 0.33 0.10 0.43 5.64 0.64 11.32 0.93 0.38 3.38 1.34 0.12 1.35 1.90 0.63 8.42 3.23 0.47 7.89 22.28 0.37 8.64 0.81 0.32 2.59 9.02 0.20 4.28 0.74 1.29 9.99 0.31 0.08 0.35 4.49 0.65 11.59 0.75 0.35 2.72 1.33 0.12 1.35 1.98 0.62 8.45 4.04 0.43 7.56 22.50 0.38 9.01 0.78 0.33 2.64 5.62 0.17 3.47 0.76 1.30 10.26 0.25 0.08 0.27 5.20 0.67 11.83 0.48 3.17 18.32 1.49 0.12 1.42 2.41 0.63 9.45 1.17 0.39 4.05 22.72 0.40 9.41 0.69 0.35 2.57 6.56 0.15 3.14 1.94 1.30 19.74 0.14 0.07 0.16 5.20 0.67 11.97 0.68 0.55 4.05 2.03 0.11 1.47 2.67 0.61 9.53 0.59 0.36 2.40 23.69 0.41 9.81 0.97 0.36 3.29 11.28 0.14 2.98///0.11 0.06 0.12 3.82 0.68 11.83 1.02 0.49 4.67 1.56 0.11 1.32 1.73 0.31 3.91///21.18 0.43 10.07 1.28 0.38 3.99 16.49 0.12 2.81///0.14 0.06 0.14 2.75 0.69 10.89 0.98 0.52 4.85 1.06 0.25 2.46/////////1.55 0.39 4.60 15.27 0.11 2.55 Fan et al.,1996 Li et al.,1994 Kramer et al.,2010ab Zhu et al.,2009 Zhu et al.,2008 Wu et al.,2007 Wu et al.,2006 Bird et al.,2014 Lü et al.,2011 Herzschuh et al.,2009 ID Site 1 ka 2 ka 3 ka 4 ka 5 ka 6 ka 7 ka 8 ka 9 ka 10 ka 11 ka 12 ka References to be continued

11 12 13 14 15 16 17 18 19 20 Ximen Co Qinghai Lake Donggi Cona Lake Gahai Lake Chaka Salt Lake Genggahai Lake Erlongwan Maar Lake Xingkai Lake Sihailongwan Maar Lake Moon Lake OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))9.72 0.02 0.40 1.86 0.44 5.89 0.98 0.34 3.19 2.06 2.61 36.45 0.08 3.54 5.37 1.96 1.19 16.18 5.29 0.49 9.75 0.42 0.06 0.32 9.81 0.06 1.21 12.40 0.52 11.44 9.53 0.24 5.24 0.90 0.44 3.84 1.13 0.19 1.97 1.77 1.52 19.73 0.07 1.28 1.64 3.37 0.76 12.70 15.57 0.49 11.20 0.31 0.06 0.27 11.41 0.09 2.07 15.86 0.51 11.63 9.66 0.56 12.15 0.41 0.43 2.23 1.39 0.19 2.21 1.89 1.41 18.91 0.12 0.74 1.50 4.14 0.65 11.41 13.27 0.50 11.17 0.37 0.06 0.30 14.11 0.14 3.04 17.63 0.50 11.54 9.28 0.57 12.28 0.69 0.42 3.12 1.29 0.27 2.91 1.34 1.17 13.08 0.13 0.52 1.15 4.22 0.59 10.38 14.25 0.51 11.43 0.44 0.06 0.35 13.95 0.17 3.92 19.23 0.49 11.42 9.69 0.50 10.86 1.62 0.41 5.10 1.56 0.17 2.09 1.56 0.97 11.79 0.25 0.40 1.44 3.29 0.55 9.16 14.88 0.52 11.70 0.45 0.07 0.36 13.96 0.22 4.86 17.12 0.48 11.04 11.18 0.61 13.29 2.24 0.40 5.85 1.71 0.16 2.10 1.59 1.13 13.89 0.65 0.33 2.34 3.80 0.52 8.95 7.48 0.53 10.98 0.55 0.07 0.43 14.24 0.26 5.78 14.11 0.47 10.57 12.04 0.25 5.59 3.69 0.40 6.81 1.22 0.24 2.51 1.63 0.99 12.30 0.66 0.28 1.99 2.65 0.49 7.67 3.28 0.53 10.14 0.55 0.07 0.44 13.52 0.30 6.64 10.32 0.46 10.00 11.50 0.15 3.30 4.94 0.39 6.91 1.56 0.21 2.59 1.51 1.13 13.49 0.97 0.24 2.18 3.51 0.47 8.03 8.08 0.54 11.40 0.55 0.07 0.45 11.60 0.34 7.38 10.08 0.45 9.76 7.91 0.20 4.28 8.03 0.38 5.63 1.06 0.25 2.40 1.73 3.00 38.47 1.28 0.21 2.25 3.10 0.46 7.49 3.14 0.55 10.41 0.61 0.07 0.49 11.01 0.38 8.22 7.21 0.44 9.18 8.77 0.15 3.20 5.85 0.37 6.51 0.72 0.20 1.53 1.66 0.79 9.90 1.07 0.18 1.80 1.40 0.44 5.05 4.04 0.56 10.86 0.60 0.07 0.49 8.92 0.42 8.87 5.68 0.43 8.73 6.18 0.17 3.46 7.57 0.36 5.66 0.68 0.18 1.35 1.15 0.75 7.66 0.88 0.17 1.45 4.45 0.43 7.59 6.72 0.57 12.37 0.51 0.07 0.45 6.67 0.45 9.35 6.99 0.42 8.73 6.58 0.20 4.14 5.98 0.36 6.20 0.50 0.19 1.16 0.70 0.66 4.92 0.55 0.16 0.99 3.01 0.42 6.81 1.27 0.57 6.20 0.55 0.08 0.49 6.46 0.50 10.21 8.47 0.41 8.70 Zhang and Mischke,2009 Shen et al.,2005a Opitz et al.,2012 Guo,2012 Liu et al.,2008a Song et al.,2012 You and Liu,2012 Wu and Shen,2010ab Liu et al.,2005b Liu et al.,2010 ID Site 1 ka 2 ka 3 ka 4 ka 5 ka 6 ka 7 ka 8 ka 9 ka 10 ka 11 ka 12 ka References to be continued Table S1 Organic C content,sediment and C accumulation rates of individual lake cores at 1000-year bin during the Holocene

21 22 23 24 25 26 27 28 29 30 31 Erhai Lake Dian Chi Xima Chi Xingyun Lake Lugu Lake Huguangyan Maar Lake Dajiuhu Lake Gucheng Lake Jiangling Dahu Lake Chao Lake OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))0.32 0.38 1.64 2.12 0.49 6.92 6.87 0.22 3.68 1.07 0.05 0.59 10.96 0.40 8.76 2.53 0.64 11.90 29.56 0.40 9.72///0.76 2.77 21.79 24.36 0.26 6.14///0.37 0.40 1.91 2.30 0.47 6.90 6.36 0.12 2.08 0.46 0.77 2.46 9.60 0.39 8.39 4.30 0.64 12.65 44.82 0.24 6.21///1.00 1.66 15.65 11.68 0.05 1.08///1.75 0.31 4.00 2.09 0.45 6.32 5.28 0.12 2.10 1.99 0.56 2.95 8.97 0.38 8.07 6.74 0.65 13.33 48.28 0.20 5.24///3.76 0.95 16.43 12.55 0.11 2.33///1.57 0.44 5.35 1.65 0.43 5.38 3.47 0.12 1.95 6.45 0.47 9.71 11.04 0.37 7.99 7.31 0.65 13.48 43.48 0.18 4.66///7.18 0.67 10.78 20.04 0.05 1.07///1.79 0.46 6.01 3.25 0.41 6.83 2.86 0.11 1.79 8.00 0.42 8.84 11.66 0.35 7.77 10.69 0.65 14.09 28.99 0.17 4.15///1.86 0.51 6.76 58.04 0.10 2.65///1.73 0.48 6.15 2.35 0.39 5.77 3.85 0.11 1.90 10.09 0.37 8.03 10.03 0.34 7.37 7.19 0.65 13.50 24.94 0.16 3.84///1.03 0.42 4.07 76.53 0.08 2.17///2.17 0.50 7.15 2.37 0.37 5.51 3.12 0.01 0.16 10.60 0.34 7.43 8.27 0.33 6.97 4.92 0.65 12.95 27.26 0.15 3.67 5.55 7.69 130.57 0.72 0.35 2.70 52.10 0.05 1.42 0.83 1.50 12.50 2.78 0.52 8.23 1.55 0.35 4.24 2.13 0.10 1.49 11.77 0.32 7.02 8.06 0.32 6.71 12.10 0.65 14.35 36.97 0.15 3.64 6.09 2.90 59.16 0.66 0.31 2.26 46.36 0.14 3.52 0.37 1.50 7.23 2.24 0.55 7.97 1.58 0.33 4.03 2.06 0.10 1.43 9.44 0.30 6.41 9.01 0.31 6.52 12.49 0.65 14.43 34.78 0.14 3.48 6.40 2.04 41.91 0.56 0.29 1.87 71.93 0.18 4.94 0.36 1.50 7.11 2.18 0.58 8.30 1.58 0.31 3.77 2.49 0.10 1.52 9.08 0.28 6.01 8.44 0.29 6.24 12.32 0.65 14.43 18.09 0.14 3.13 8.03 1.66 34.99///75.21 0.15 4.07 0.46 1.50 8.41 1.54 0.60 7.21///2.16 0.10 1.41 5.85 0.27 4.73 8.09 0.28 5.95 10.73 0.65 14.23 7.28 0.13 2.73 5.63 1.44 29.09///58.62 0.08 2.22 0.52 1.50 9.21//////2.57 0.10 1.50 3.77 0.26 4.52 7.23 0.27 5.58 8.48 0.65 13.88 6.69 0.13 2.64 5.40 1.28 25.87///43.48 0.11 2.92 0.54 1.50 9.43 Shen et al.,2005b Wu et al.,1998 Yang et al.,2004 Zhang et al.,2014 Zheng et al.,2014;Zheng,2014 Mingram et al.,2004 Ma et al.,2008 Wang et al.,1999 Xie,2004 Xue et al.,2007 Hu et al.,2015 ID Site 1 ka 2 ka 3 ka 4 ka 5 ka 6 ka 7 ka 8 ka 9 ka 10 ka 11 ka 12 ka References Table S1 Organic C conten t,sedimen t and C accumulation rates of individ ual lake cores at 1 000-year bin durin g the Holocene o be continued t

32 33 34 35 36 37 38 39 40 41 Taihu Lake Beihuqiao Longgan Lake Sanqing Chi Bosten Lake Aibi Lake Yitang Lake Balikun Lake Wulungu Lake Manas Lake OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR g C/(m2·a)OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))//////1.53 3.45 41.40 1.86 0.22 2.96 5.05 1.06 18.79 0.40 0.51 2.58///2.52 0.30 4.52 0.67 0.44 3.20 2.21 1.71 24.60//////1.38 0.58 6.60 1.27 0.10 1.13 4.47 1.03 18.24 0.50 0.49 2.94 0.53 1.00 6.19 2.54 0.29 4.43 0.65 0.38 2.72 1.32 0.65 7.17//////0.62 0.61 4.25 1.14 0.08 0.83 4.17 1.00 17.51 0.64 0.47 3.35 0.20 1.05 3.20 1.52 0.28 3.38 0.37 0.32 1.56 0.49 0.34 1.99//////0.27 0.64 2.42 0.98 0.07 0.64 5.11 0.96 17.01 0.80 0.46 3.74 0.29 1.10 4.44 1.21 0.27 2.89 0.61 0.26 1.77 0.13 0.27 0.57 0.46 0.29 1.64///0.44 0.67 3.65 0.72 0.06 0.46 4.38 0.94 16.58 0.97 0.44 4.08 0.23 1.17 3.97 0.88 0.27 2.31 0.75 0.20 1.57 0.14 0.22 0.52 0.47 0.30 1.74 3.61 1.94 33.28 0.91 0.70 6.22 0.90 0.06 0.49 3.42 0.91 15.29 1.20 0.43 4.49 0.21 1.24 3.88 1.09 0.26 2.56 0.91 0.14 1.23 0.09 0.20 0.34 0.49 0.32 1.87 3.38 1.20 20.21 1.05 0.73 7.09///3.92 0.87 15.08 0.87 0.41 3.58 0.19 1.33 3.87 2.00 0.25 3.44 0.68 0.08 0.57 0.18 0.18 0.51 0.46 0.33 1.88 3.56 2.43 41.53 0.12 0.76 1.58///3.60 0.84 14.38 0.79 0.40 3.25 0.16 1.45 3.79 2.06 0.24 3.38 1.10 0.02 0.20 0.25 0.17 0.60 0.43 0.35 1.88 3.62 2.67 45.70 0.24 0.83 2.88///3.05 0.82 13.33///0.22 1.59 5.22 2.66 0.23 3.65 0.57 0.04 0.25 0.27 0.15 0.60 0.46 0.37 2.08 3.61 2.89 49.45 0.35 0.90 4.15/////////1.33 1.80 19.99//////0.16 0.15 0.37 0.51 0.39 2.40 3.57 3.03 51.79 0.92 0.99 8.83/////////0.47 2.11 12.07//////0.33 0.14 0.61 0.39 0.43 2.16///0.94 1.09 9.89/////////0.26 2.71 10.02//////0.10 0.13 0.23 Shu et al.,2007 Wei et al.,2016 Yang et al.,2002 Song et al.,2016 Zhang et al.,2007 Wu,1995 Zhang and Li,2015 Xue and Zhang,2011 Jiang et al.,2007 Rhodes et al.,1996 ID Site 1 ka 2 ka 3 ka 4 ka 5 ka 6 ka 7 ka 8 ka 9 ka 10 ka 11 ka 12 ka References Table S1 Organic C content,sediment and C accumulation rates of individual lake cores at 1000-year bin during the Holocene to be continued

42 43 44 45 46 47 48 49 50 51 Zhuye Lake Juyanze Lake Huahai Lake Sanjiaocheng Daihai Lake Dali Lake Gonghai Lake Hulun Lake Tengger Nuur Yanhaizi Lake OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))OC(%)SAR(mm/a)CAR(g C/(m2·a))0.27 1.88 7.16///0.64 0.50 3.51///0.69 2.08 15.47 4.30 2.39 42.20 9.83 2.38 51.37 0.97 1.56 14.34 1.70 0.25 3.18 0.23 1.18 4.03 0.91 0.55 4.88 1.71 1.37 17.48 0.45 0.13 0.69 1.20 1.14 11.92 0.93 1.89 16.98 4.76 1.61 28.62 9.55 1.69 36.37 0.96 1.51 13.88 1.39 0.27 3.03 0.23 1.26 4.25 1.44 0.20 2.30 1.07 1.30 12.75 0.30 0.17 0.70 2.17 0.90 12.89 1.46 1.72 20.11 5.17 0.94 16.67 11.07 0.95 20.80 1.66 1.45 18.22 1.37 0.28 3.20 0.31 1.30 5.48 1.38 0.48 5.45 1.10 1.20 12.01 0.46 0.22 1.24 3.92 0.64 11.17 1.35 1.54 17.28 5.02 0.66 11.71 12.33 0.68 15.09 0.71 1.40 10.51 1.20 0.30 3.20 0.56 1.32 8.50 1.13 0.43 4.34 1.27 1.10 11.90 0.69 0.28 2.11 3.49 0.50 8.53 1.33 1.36 15.13 5.64 0.51 9.00 12.94 0.54 11.97 0.24 1.34 4.66 0.80 0.33 2.74 0.33 1.34 5.91 1.26 0.50 5.39 0.80 1.00 8.15 0.79 0.38 3.08 2.88 0.41 6.54 1.46 1.18 13.82 4.77 0.42 7.49 14.79 0.44 9.96 0.41 1.28 6.58 0.66 0.38 2.72 0.23 1.35 4.61 1.20 0.39 4.10 0.81 0.90 7.44 0.81 0.48 3.93 2.57 0.34 5.26 2.05 1.00 13.90 5.92 0.36 6.20 19.34 0.37 8.62 3.51 1.22 20.71 0.53 0.43 2.68 0.27 1.37 5.28 0.23 0.72 2.42 0.83 0.80 6.69 1.10 0.62 6.19 5.24 0.30 5.28 2.66 0.81 12.62 5.87 0.31 5.39 20.41 0.32 7.51 2.46 1.16 17.51 0.60 0.53 3.59 0.17 1.38 3.78 0.25 0.24 0.85 0.60 0.70 4.77 0.57 0.81 5.29 3.66 0.26 4.50 1.63 0.63 7.79 5.54 0.27 4.80 25.44 0.28 6.77 0.45 1.10 6.04 0.68 0.76 5.61 0.77 1.38 11.06 0.26 0.34 1.26 0.65 0.60 4.30 0.36 1.03 4.84 5.59 0.23 4.14 1.84 0.46 6.03 5.59 0.24 4.28 19.59 0.25 5.87 0.28 1.04 4.09 0.79 1.25 10.13 0.82 1.39 11.59 0.21 0.19 0.61 0.46 0.51 2.88 0.34 1.35 6.10 6.47 0.21 3.65 2.20 0.32 4.56 3.25 0.22 3.64 23.86 0.23 5.44 1.25 0.98 10.50///0.79 1.40 11.37 0.06 0.09 0.14///0.87 1.75 15.09 4.66 0.20 3.51///1.38 0.21 2.33 17.70 0.21 4.84 1.65 0.93 11.62///0.87 1.41 12.09 Li et al.,2013 Herzschuh et al.,2004 Wang et al.,2013 Zhang et al.,2000 Xiao et al.,2006 Fan et al.,2015 Chen et al.,2013 Hu et al.,2000 Ma et al.,2004 Chen et al.,2003 ID Site 1 ka 2 ka 3 ka 4 ka 5 ka 6 ka 7 ka 8 ka 9 ka 10 ka 11 ka 12 ka References Table S1 Organic C content,sediment and C accumulation rates of individual lake cores at 1000-year bin during the Holocene to be continued

Table S2 Site information of Chinese lakes in this study 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Bangong Co Songxi Co Naleng Lake Chen Co Nam Co Zigetang Co Ngoin Co Paru Co Pumoyum Co Koucha Lake Ximen Co Qinghai Lake Donggi Cona Lake Gahai Lake Chaka Salt Lake Genggahai Lake Erlongwan Maar Lake Xingkai Lake Sihailongwan Maar Lake Moon Lake 33°40′N 34°37′N 31°06′N 28°56′N 30°42′N 32°04′N 31°28′N 29°47′N 28°33′N 34°00′N 33°22′N 36°32′N 35°18′N 37°08′N 36°41′N 36°11′N 42°18′N 44°57′N 42°17′N 47°30′N 79°30′E 80°16′E 99°45′E 90°35′E 90°40′E 90°50′E 91°30′E 92°21′E 90°26′E 97°12′E 101°07′E 99°36′E 98°32′E 97°33′E 99°07′E 100°06′E 126°22′E 132°25′E 126°36′E 120°52′E 4,241 4,870 4,200 4,420 4,718 4,560 4,532 4,845 5,001 4,540 4,261 4,583 4,090 2,850 3,200 3,000 722 69 916 1,194 604 25 1.8 38 1,920 191.4 61.3 0.1 290 18 3.6 4,456 229 35 105 2 0.3 4,380 0.5 0.0269 28,714 1,605 470 148 10,610 3,430 1,081 2.97 1,233.9 8850 29,660 3,174/11,600/0.4 36,400 0.7/5 8//30 28.5/////21/8 30//4.5//41/36.7 31 33 39/1.2/6.9 63.3 32.8 98 15/1.8 36 10 506.5 2,465//1,250 2,350 925.1 1,000/1,770 4,400/1,502/2,000 2,074.1 1,716/587.2//109.5 126.9 852.3 358.0 430.2 400.3 486.8 550.7 390.3 508.8 812.0 333.0 354.4 252.6 284.3 325.1 787.0 549.6 849.0 485.5-5.6-6.8-2.2 1.8-0.8-2.7-2.0-2.2-2.8-6.2 0.0 1.0-2.8-0.3 3.5 3.8 4.5 4.4 2.6-2.5 29.0 34.2 175.1 82.5 97.6 91.8 106.0 112.1 87.4 101.3 153.4 71.9 73.4 55.3 61.9 67.1 160.9 100.4 170.3 106.6 5.23.95.58.77.87.08.35.64.24.04.1 11.37.4 10.8 12.0 13.9 20.0 20.2 17.9 14.8 Campo et al.,1996 Li et al.,1994 Kramer et al.,2010ab Feng et al.,2004 Zhu et al.,2008 Li et al.,2009b Wu et al.,2006 Bird et al.,2014 Nishimura et al.,2014 Aichner et al.,2010 Zhang and Mischke,2009 Matsumoto,2005 Opitz et al.,2012 Guo,2012 Liu et al.,2008a Song et al.,2012 Liu et al.,2008c Wu and Shen,2010ab Liu et al.,2005b Wu and Liu,2012 ID Site Latitude(°C)longitude(°C)Altitude(m)Lake surface area(km2)Catchment area(km2)Mean depth(m)Max depth(m)Evaporation(mm)Precipitation(mm)Temperature(°C)Summer precipitation(mm)Summer temperature(°C)References to be continued

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Erhai Lake Dian Chi Xima Chi Xingyun Lake Lugu Lake Huguangyan Maar Lake Dajiuhu Lake Gucheng Lake Jiangling Dahu Lake Chao Lake Taihu Lake Beihuqiao Longgan Lake Sanqing Chi Bosten Lake Aibi Lake Yitang Lake Balikun Lake Wulungu Lake 25°47′N 25°04′N 25°47′N 24°20′N 27°42′N 21°09′N 31°28′N 31°15′N 30°02′N 24°41′N 31°31′N 31°22′N 30°22′N 29°58′N 33°55′N 42°05′N 44°51′N 40°31′N 43°40′N 47°12′N 100°12′E 102°41′E 100°05′E 102°47′E 100°50′E 110°17′E 110°04′E 108°55′E 112°24′E 115°00′E 117°23′E 120°07′E 119°56′E 116°06′E 107°56′E 87°03′E 82°23′E 94°58′E 92°48′E 87°17′E 1,974 1,886 3,800 1,740 2,690 87.6 1,780 139 1,956 255 450 62 1,186 13 1,564 1,048 194 1,054 1,575 479 256.5 330/34.7 57.7 2.3 16 23.4/0.8 770 2,338.1/316.2/1,002.4 542 150 116 927 2,785 2,866/386 216 3.5 34 37.2//13,349//5,511/55,600 50,621/4,500 32,000 10.2 5/7 38.4//7//2.7 1.9/3.8/8.2 1.4/0.6 8 20.7 8/11 105.3 22/10//3.8 2.6/4.6/17 2.8/1.1 16 1,160 1,685/1,192///1,400-1,900//1,604////1,800-2,000 1,315 2,486 2,250 1,844 915.0 1,092.2 915.0 980.5 797.4 1,612.9 1,105.5 1,192.9 1,123.4 1,635.3 1,048.8 1,156.5 1,406.7 1,453.6 1,008.6 90.6 153.5 52.9 200.0 128.9 13.1 11.5 13.1 14.6 9.7 23.2 15.3 17.8 16.8 18.4 16.0 16.1 16.6 17.0 5.7 8.3 8.2 8.5 3.1 5.0 178.8 208.8 178.8 178.4 179.3 269.7 163.3 169.5 151.3 200.1 154.1 159.3 172.7 178.8 163.6 18.5 18.3 11.2 35.7 17.8 17.9 15.6 17.9 18.6 15.2 27.6 24.1 26.6 26.6 25.1 26.4 26.3 26.6 26.9 14.5 21.8 23.9 22.5 17.7 21.8 Shen et al.,2005b,2006 Wu et al.,1998 Yang et al.,2004 Yang et al.,2004 Zheng,2014 Liu et al.,2005a Zhu et al.,2006 Wang et al.,1999 Xie,2004 Zhong et al.,2010 Wang et al.,2008 Li,2013 Wei et al.,2016 Yang et al.,2002 Song et al.,2016 Zhang et al.,2007,2010 Wu,1995 Zhang and Li,2015 Tao et al.,2010 Liu et al.,2008b ID Site Latitude(°C)longitude(°C)Altitude(m)Lake surface area(km2)Catchment area(km2)Mean depth(m)Max depth(m)Evaporation(mm)Precipitation(mm)Temperature(°C)Summer precipitation(mm)Summer temperature(°C)References Table S2 Site information of Chinese lakes in this study to be continued

41 42 43 44 45 46 47 48 49 50 51 Manas Lake Zhuye Lake Juyanze Lake Huahai Lake Sanjiaocheng Daihai Lake Dali Lake Gonghai Lake Hulun Lake Tengger Nuur Yanhaizi Lake 44°56′N 39°03′N 41°53′N 40°25′N 38°10′N 40°33′N 43°18′N 38°54′N 49°00′N 42°27′N 40°08′N 86°20′E 103°40′E 101°51′E 98°48′E 102°57′E 112°39′E 116°37′E 112°14′E 117°25′E 110°42′E 108°27′E 367 1,309 892 1,200 1,469 1,220 1,226 1,860 545 1,092 1,180 750 628 24//134 238 0.36 2,339 28.618 11,000 41,600 3,500 14,400/2,289 783/37,214/2,000//2//7.4///0.7///4.1//16.1 11 109 1.20.5 2,000-2,600 2,600 3,000/1,162 1,632/1,400-1,900 2,360 2,604 170.9 111.9 46.6 65.6 146.3 395.9 372.1 449.3 278.5 207.2 223.5 7.18.49.18.78.63.91.14.3-0.75.26.8 19.6 22.7 9.7 12.6 27.0 82.8 80.5 91.9 65.2 43.8 47.4 24.1 22.6 25.4 22.5 21.4 17.9 16.9 17.6 18.2 20.8 21.1 Rhodes et al.,1996 Li et al.,2009a,2011 Kai and Wünnemann,2009 Wang et al.,2013 Zhang et al.,2000 Xiao et al.,2004,2006 Fan et al.,2015 Chen et al.,2013 Wen et al.,2010 Guo et al.,2012 Chen et al.,2003 ID Site Latitude(°C)longitude(°C)Altitude(m)Lake surface area(km2)Catchment area(km2)Mean depth(m)Max depth(m)Evaporation(mm)2,100 Precipitation(mm)Temperature(°C)Summer precipitation(mm)Summer temperature(°C)References Table S2 Site information of Chinese lakes in this study to be continued

Figure S1 Relationship between lake C accumulation rates(g C/(m2·a))and sediment organic carbon content(TOC,%).Gray bars suggest lake C accumulation rates and black curves indicate TOC

Figure S2 Relationship between lake C accumulation rates(g C/(m2·a))and carbon nitrogen ratio(C/N).Gray bars suggest lake C accumulation rates and black curves indicate C/N

Figure S3 Relationship between lake C accumulation rates(g C/(m2·a))and organic carbon isotope(δ13C,‰).Gray bars suggest lake C accumulation rates and black curves indicate δ13C

Figure S4 Relationship between lake C accumulation rates(g C/(m2·a))and total pollen concentration(×103 grains/g or grains/cm3).Gray bars suggest lake C accumulation rates and black curves indicate total pollen concentration(TPC)

Acknowledgments:

This work was supported by the National Natural Science Foundation of China (Grant Nos. 41822708 and 41571178), the Strategic Priority Research Program of Chinese Academy of Sciences (Grant No.XDA20100102), the Fundamental Research Funds for the Central Universities (Grant No. lzujbky-2018-k15),and the Second Tibetan Plateau Scientific Expedition(STEP) program (Grant No. XDA20060700). The authors give special thanks to Fengju Zhang for her help in the calculation of lake C accumulation rate.