Influence of the biological carbon pump effect on the sources and deposition of organic matter in Fuxian Lake,a deep oligotrophic lake in southwest China

Haibo He·Zaihua Liu·Chongying Chen·Yu Wei·Qian Bao·Hailong Sun·Yundi Hu·Hao Yan

Abstract Biological carbon pumping(BCP)is a key process in which dissolved inorganic carbon in terrestrial aquatic ecosystems is utilized by aquatic autotrophs for photosynthesis and transformed into autochthonous organic matter(AOC).However,the mechanisms underlying BCP and the amount of generated AOC deposited effectively,are still poorly understood.Therefore,we conducted a systematic study combining modern hydrochemical monitoring and a sediment trap experiment in Fuxian Lake(Yunnan,SW China),the second-deepest plateau,oligotrophic freshwater lake in China.Temperature,pH,EC(electrical conductivity), DO (dissolved O2), [HCO3-],[Ca2+],SIc,partial CO2(pCO2)pressure,and carbon isotopic compositions of HCO3-(δ13CDIC)in water from Fuxian Lake all displayed distinct seasonal and vertical variations. This was especially apparent in an inverse correlation between pCO2 and DO,indicating that variations of hydrochemistry in the lake water were mainly controlled by the metabolism of the aquatic phototrophs.Furthermore,the lowest C/N ratios and highest δ13Corg were recorded in the trap sediments.Analyses of the C/N ratio demonstrated that the proportions of AOC ranged from 30%to 100%of all OC,indicating that AOC was an important contributor to sedimentary organic matter(OC).It was calculated that the AOC flux in Fuxian Lake was 20.43 t C km-2 in 2017.Therefore,AOC produced by carbonate weathering and aquatic photosynthesis could potentially be a significant carbon sink and may have an important contribution to solving the lack of carbon sinks in the global carbon cycle.

Keywords Carbonate weathering·Hydrochemical variation·Biological carbon pump effect·Sediment trap·Autochthonous organic carbon·Carbon sink

1 Introduction

Anthropogenic emissions of CO2from burning fossil fuels,industry,and land use have increased atmospheric CO2concentration to 409 ppm in 2018 since the beginning of the industrial period in the late 18th century(i.e.,over the Anthropocene).Many mass balance calculations for CO2,however,have revealed an imbalance in the atmospheric CO2budget(Broecker et al.1979；Ciais et al.2013；Melnikov and O'Neill 2006；Schindler 1999；Sundquist 1993).Identifying missing carbon sinks is,therefore,a top priority in the science of global climate change.

Cole et al.(1994)found that globally,lakes influence the terrestrial carbon budget significantly through the efflux of CO2to the atmosphere,indicating that they are carbon sources rather than sinks.However,when sediment accumulation is taken into account,research has shown that a large quantity of carbon from terrestrial sources is buried in aquatic sediments(Cole et al.2007；Xiao 2017).Other studies(Liu et al.2010,2011,2018)have suggested that the weathering of carbonate minerals,often coupled with aquatic photosynthesis(the biological carbon pump effect or ‘‘BCP''), may be a carbon sink and an important contributor to the missing global carbon.Rapid carbonate weathering(Liu and Dreybrod 1997)provides abundant dissolved inorganic carbon(DIC),and aquatic photosynthesis takes up the latter as its carbon source when the BCP effect occurs in aquatic ecosystems.This means that the autochthonous organic carbon(OC)in organic sediments,especially from karst basins, is a significant sink for anthropogenic CO2.Therefore,the carbon sink resulting from combining carbonate weathering with aquatic photosynthesis(CCW)might be very important for controlling climate change at any time scale(Liu et al.2011,2018).

Lacustrine OC can reflect a complex mixture of different sources,including autochthonous and allochthonous OC.When compared to autochthonous OC,allochthonous OC represents terrigenous sources such as vascular plant tissue and detritus from marshes and upland sources(O'Reilly et al.2014).Some studies show that lakes can bury considerable amounts of OC in sediments that include inputs from both aquatic primary production and terrestrially fixed organic matter(Galy et al.2011；Waterson and Canuel 2008；Huang et al.2017).However,the relative contributions of autochthonous and allochthonous sources,which regulates OC burial in sediment,and the mechanisms underlying these,remain to be determined.Traditional geochemical methods(C/N,δ13Corg)used to study organic matter can be ideal tools with which to determine the sources of organic matter in sediments(Meyers and Ishiwatari 1993；Ramaswamy et al.2008；Tue et al.2011).δ13Corghas been used as a proxy indicator of productivity(e.g.,Hollander and McKenzie 1991；Brenner et al.1999),whereas C/N ratio values indicate the relative contribution of allochthonous versus autochthonous organic matter in lake sediment (Meyers and Ishiwatari 1993； Meyers 1997,2003).

In view of this,a systematic study of hydrochemical monitoring and trap sediments was conducted in Fuxian Lake,the second-deepest plateau oligotrophic freshwater lake in southwest China.Here we address the following questions:(1)Are hydrogeochemical changes affected by the process of BCP?(2)Is the sedimentary OC in Fuxian Lake mainly derived from autochthonous or allochthonous sources?(3)How much is the sedimentary flux of organic carbon generated by the BCP effect in Fuxian Lake?(4)Do anthropogenic activities and climatic change affect sedimentary fluxes?

2 Field setting

Fuxian Lake(24°17′—24°37′N,102°49′—102°57′E)is an oligotrophic lake located in central Yunnan Province,Southwest China(Fig.1).The catchment area is 675 km2,while the lake has a surface area of 212 km2and a maximum depth of 158 m.It has an average altitude of 1723 m above sea level(asl).The lake basin is wide and deep in the north and narrow and comparatively shallow in the south,with one significant inflowing stream(Liangwang River)to the north,numerous small stream-and spring-fed inflows(e.g.,Luchong Spring),and a single outlet (the Haikou River). It has an estimated water retention time of 167 years(Wang and Dou 1998).Present climatic conditions in this region are typically dominated by the Indian Summer Monsoon and experience seasonality.The region is characterized by an average temperature of 15.6°C,with 951 mm of annual average precipitation,and 83%of the total annual precipitation occurs between May and October(http://www.yncj.gov.cn/).

3 Methods

3.1 Field monitoring

Physical and chemical variables were monitored in the water column and samples were collected quarterly from the northern,central,and southern regions of the lake throughout 2017.Temperature(T),pH,electrical conductivity (EC), and dissolved oxygen (DO) profiles were obtained at 10 m intervals from the surface(0.5 m)to the bottom(110 m)of the water column using a multiparameter water quality probe(WTW；Wissenschaftlich-Technische-Werkstaetten Technology Multiline 350i). The meters were calibrated prior to deployment using standards for pH(4,7,and 10),EC(1412 μs cm-1)and DO(0%and 100%).The measuring accuracy of the pH,T,DO,and EC was 0.01, 0.01°C, 0.01 mg L-1, and 0.01 μs cm-1,respectively.The major spring(the Luchong)and river(the Liangwang)located along the western and northern shore,respectively,(Fig.1)were also sampled.

3.2 Water sampling and analysis

Two sets of water samples were filtered in the field using 0.45 μm Millipore nitrocellulose filters and 20 ml of these samples were stored for measuring major cation and anion contents.Cation samples were acidified with concentrated ultra-pure HNO3to pH ＜2 to prevent complexation and precipitation. Both anionic and cationic samples were refridgerated at 4°C.

Anions(Cl-,SO42-,and NO3-)were measured by ICS-90 ion chromatograph,and major cations(K+,Na+,Ca2+,and M g2+)were determined with an inductively coupled plasma optical emission spectrometer(ICP-OES).Reagent and procedural blanks were determined in parallel to the sample treatment.The experimental detection resolution of these ions was 0.01 mg L-1.In addition,alkalinity(chieflyas HCO3-)were in situ titrated by an Aquamerck alkalinity test kit supplied by Merck(Germany),with an estimated accuracy of 0.05 mmol L-1.In addition,ion concentrations were used to calculate the CO2partial pressure(pCO2)and calcite saturation index(SIc)from a geochemical model with pH,temperature,and seven major ions concentrations using the PHREEQC3 model(Parkhurst and Appelo 1999).

Fig.1 Schematic map of the Fuxian Lake basin with sampling points indicated by filled circles(Luchong Spring,Liangwang River and Fuxian Lake).FX-1,FX-2 and FX-3 are sampling sites representing the northern,central and southern sectors of the Lake,respectively.The basin boundary(black dashed line)and lake depth contours(dotted lines)are from Liu et al.(2008)

The remaining water samples(60 ml)were filtered by pressure filtration through 0.45 μm cellulose acetate filters and stored in previously acid-washed HDPE bottles with air-tight caps and no air bubbles or headspace for subsequent measurement of δ13CDIC.One drop of saturated HgCl2solution was added to each sample to prevent microbial activity, and all samples were stored in a refrigerator at 4°C until analyzed by MAT-252 mass spectrometry.

3.3 Sediment sampling and analyses

Settling particulate material was collected quarterly in three sediment traps deployed throughout 2017.The traps were simple open-topped cylindrical polyethylene tubes with an aspect ratio of 7:1(105 cm in length,with an inner diameter of 15 cm).Material was obtained from depths of 40,80,and 110 m in the center of the lake(depth of water about 120 m at site FX-2,Fig.1).After recovery of the sediment traps,the material collected was transferred into plastic containers in the field and stored at 4°C prior to sub-sampling,then freeze-dried.The homogeneous dry material of the sediment traps was weighed,and mass accumulation rates k(g m-2day-1)of dry material were calculated using:

where m(g)is the dry mass collected,a(m2)is the open trap area,and t(day)is the time interval of exposure.

3.4 Measurement of organic carbon,inorganic carbon,and TOC/TN

Total carbon(TC)was measured using an elemental analyzer(Elementar-vario MACRO cube),which was calibrated using sediment standard materials (B2150 and AR2026)to an analytical precision better than 0.2%.Total organic carbon(TOC),total nitrogen(TN)content,and TOC/TN(atomic)ratios were determined for the sediments following acid treatment with 1.5 mol/L HCl for 24 h to remove inorganic carbon.The carbonate-free samples were triple-rinsed with distilled water to remove acid residues,dried at 60°C for 48 h,and measured using the MACRO analyzer.The total inorganic carbon(TIC)content was then calculated as the difference between TC and TOC.

The rates of accumulation of organic carbon(OCAR)and inorganic carbon(ICAR)(g C m-2day-1)were estimated using the following formula:

where k(g m-2day-1)is mass accumulation rates of dry material and C is the content of organic carbon(or inorganic carbon)(mg g-1).

3.5 Measurement of stable carbon isotope ratios

The stable carbon isotope ratios(δ13C)of organic carbon and inorganic carbon in the samples were determined using a MAT-252/253 mass spectrometer.Isotopic composition is reported relative to the international Vienna Peedee Belemnite (VPDB) standard. Analytical precision was typically better than±0.03%based on replicate measurements of an internal laboratory standard(Sun et al.2011).All isotopic compositions of samples mentioned in this paper are expressed as‘‘δ''values,representing deviations in per mil(%)from the standard:

where R is the13C/12C ratio of the sample or standard.

3.6 Calculation of OCAR from autochthonous and allochthonous sources

The values of the C/N ratios indicate the relative contribution of allochthonous versus autochthonous organic matter to lake sediments(Meyers and Ishiwatari 1993；Meyers 1997,2003).Autochthonous and allochthonous sources of OC in trap sediments can be calculated from the following:

where Cautoand Calloare the proportion of autochthonous and allochthonous OC,respectively,C/Ni is the C/N ratio of a given sample,C/Nauto is the C/N ratio of the autochthonous endmember(7.02),and C/Nallo is the C/N ratio of the allochthonous endmember(12.71).

The organic carbon accumulation rates from autochthonous and allochthonous OC can then be calculated by:

where OCARauto(g C m-2day-1)and OCARallo(g C m-2day-1)are the organic carbon accumulation rates from autochthonous and allochthonous OC,respectively.

4 Results

4.1 Spatial-temporal physicochemical variations in river,spring,and Fuxian Lake water

Temporal and spatial variations in water chemistry and carbon isotopes in Luchong Spring,Liangwang River,and surface water from Fuxian Lake are summarized in Tables 1 and 2.All parameters in the spring and river water showed little seasonal variation,with the coefficients of variation(CV)generally being ＜1%(Table 1,Fig.2).In addition,the spring and river water had a lower pH,DO,SIc,and a higher[Ca2+]and pCO2than the surface water from Fuxian Lake.

The vertical profiles of lake water temperature displayed thermal stratification and the thermocline was located between 20 and 50 m from April to October 2017.In contrast,it was well mixed between 0 and 50 m in January 2017(Fig.3).The hydrochemical facies of the water column under different seasons are shown in a Piper diagram in Fig.4.The major cations were Ca2+and Mg2+,and themajor anion was.The hydrochemical type was of simple HCO3-—Ca—Mg,indicating typical karst water.

Table 1 Statistics on the seasonal variations of physicochemical parameters in the spring and river

The vertical variations in pH,DO,and SIcin the water column were similar to that of T during thermal stratification in April,July,and October,indicating that the vertical water exchange was effectively limited by thermal stratification.These values were much higher in the epilimnion than in deep water.In contrast,the EC,[Ca2+],,and pCO2in the epilimnion were lower than those in deep water during both summer stratification and winter mixing(Fig.3).

The δ13CDICvariations in spring,river,and lake water in the 4 months are presented in Fig.5 and Tables 1 and 2.δ13CDICvalues of spring and river waters exhibited no discernible seasonal variation,with the CV values of 0.1 for both spring and river waters.However,the δ13CDICvariations of surface waters varied seasonally,being most negative in winter.Figure 5 shows the depth profiles of δ13CDICin Fuxian Lake obtained between January and October 2017.The lake was homogeneous with respect to δ13CDICin the epilimnion between 0 m and 50 m during the winter mixing period.δ13CDICvalues decreased in the hypolimnetic waters.

4.2 Contents,accumulation rates and carbon isotope ratios of organic and inorganic carbon in trap sediments

Throughout the year (January 2017—January 2018) of sediment trapping in Fuxian Lake,the collected material revealed distinct changes(Table 3).Total accumulation rates(total AR)derived from sediment traps were substantially higher during summer and fall(1.00—2.95 g C m-2day-1)than during spring and winter(0.44—0.70 g C m-2day-1).

There were higher values(92.21—107.78 mg g-1)of TOC in the winter and spring, while higher TIC(63.54—80.83 mg g-1)occurred in the summer and fall.There were positive correlations between inorganic carbon accumulation rates(ICAR)and organic carbon accumulation rates(OCAR),with both reaching their highest values in summer and fall(ICAR:0.07—0.24 g C m-2day-1；OCAR:0.08—0.13 g C m-2day-1)(Table 3).The C/N ratio is a frequently-used index feature of sediments,which was essentially lower than 10 in the current study.There were positive correlations between δ13Corg(-20.88 to-25.37%)and δ13Ccarb(-0.35—0.007%),with higher values in summer and fall(Table 3).

The calculated average contributions of autochthonous OC in these trap sediments ranged from 30 to 100%.The OCARautoranged from 0.02 to 0.13 g C m-2day-1,with an average of 0.04 g C m-2day-1(Table 4).

5 Discussion

5.1 Mechanisms for the temporal-spatial variations of water chemistry and carbon isotope in Fuxian Lake:role of biological carbon pump effect

Fuxian Lake has a large open water surface area and volume and a small catchment area,with a long water retention time of 167 years(Wang and Dou 1998).It receives water from more than 20 small rivers,indicating that the body of the lake has a strong buffer effect on inputs from allochthonous sources.In contrast,Liangwang River,the largest river entering the lake,showed seasonal fluctuations in the physicochemical parameters which are consistent with values in the surface water of Fuxian Lake during the sampling year(Fig.2).However,these data cannot explain variations in the depth profiles from Fuxian Lake.All parameters in Luchong Spring showed an opposite trend to that of the surface water of Fuxian Lake(Fig.2).Therefore,the physical and chemical parameters in the depth profiles of Fuxian Lake were not controlled by river and spring input.

Some lakes are supersaturated with CO2and consequently emit CO2to the atmosphere(Sobek et al.2005；Cole et al.1994).Cole et al.(1994)estimated that the mean partial pressure of CO2of lake water could be as high as three times that of the overlying atmosphere.However,our results showed that there existed both efflux and influx of CO2in Fuxian Lake during the study period(Table 2,Fig.3).Furthermore,pCO2and DO followed opposite trends Fuxian Lake at different water depths of the water column over the year (Fig.6), indicating that strong aquatic photosynthesis(with high DO concentration as an indicator)in lake waters might be an important carbon sink as this results in the absorption of CO2directly from the atmosphere,such as what occurs in the ocean(Passow andCarlson 2012).In the euphotic zone,the photosynthesis of aquatic plants consumes CO2in the water and produces O2.With increasing depth,light intensity decreases,and the photosynthetic metabolism of aquatic phototrophs weakens.

Table 2 Statistics on the seasonal variations of physicochemical parameters at different depths in Fuxian Lake

Table 2 continued

Fig.2 A comparison of a temperature,b pH,c EC,d DO,e[Ca2+],f[HCO3-],g SIc,h pCO2 and i δ13CDIC determined in 2017 in Luchong Spring,Liangwang River,and the surface water of Fuxian Lake(FX-1,FX-2,FX-3；water depth of 0.5 m).Dashed line represents CO2 concentration in equilibrium with the atmosphere

Stable carbon isotopic composition(δ13C)is usually used to constrain the carbon sources(Han et al.2010；Cui et al.2017).The main controls on the δ13CDICin most lake waters are:(i)river and groundwater contributions；(ii)CO2exchanges of the water-atmosphere interface；(iii)carbonate mineral precipitation or dissolution reactions；(iv)CO2contributed by degradation of organic matter；and(v)photosynthetic utilization of12C into organic carbon(McKenzie 1985；Falkowski and Raven 1997；Xu et al.2006；Lamb et al.2007；Li et al.2008；Chen et al.2018).In this study,the δ13CDICof spring and river waters in the Fuxian Lake basin had smaller coefficients of variation(spring,-8.5%；river,-9.9%；Table 1),which was in good agreement with the trend in the surface water of Fuxian Lake.However,this was unable to account for the relatively large variations of lake water (CV:-121,027.4%—3705.7%； Table 2). As already noted,higher pCO2(than atmospheric)and lower pH values,favoring chemically enhanced emission of CO2rather than influx,were observed in the surface waters during winter mixing in January.Theoretically,this would result in a higher δ13CDIC, but this was contrary to the results observed by our monitoring.Furthermore,some studies have confirmed that lacustrine water with more negative δ13CDICvalues reflected increased decomposition of organic matter(Wachniew and Ro´˙zan´ski 1997)and precipitation of carbonate minerals (Valero-Garce´s et al.1999).In the present study,however,the increasingly positive δ13C values obtained from January to July(Fig.5),when the decomposition of organic matter and the calcite deposition collected in the sediment traps both increased gradually,indicated that the relative importance of these two mechanisms could not be determined by direct observations.Indeed,the increasingly negative hypolimnetic values may partly be due to a weakening of photosynthesis and degradation of organic carbon, which produced the13C-depleted DIC(Parker et al.2010).

Variations in δ13CDICvalues are strongly associated with biological processes；for example,changes in the13C/12C ratio are a response to algal productivity(McKenzie 1985； Chen et al. 2017). Over time, the increased biomass would assimilate proportionally larger amounts of12C,leaving the DIC pool enriched in13C.This concurs with the more positive δ13CDICin the surface water in July and the heavier δ13CDICin the epilimnion than in the hypolimnion(Fig.5).This indicates how important the influence of aquatic photosynthesis is on the stable carbon isotopic compositions of DIC.

To better understand the concept of the mechanisms controlling the processes underpinning hydrochemistry in terrestrial aquatic ecosystems,atmospheric/soil CO2uptake was modeled by coupling carbonate weathering with an aquatic photosynthesis mechanism (CCW) (Liu et al.2018),as follows:

where x and 1-x are stoichiometric coefficients.

According to Eq(6),carbonate weathering coupled with aquatic photosynthesis (the biological carbon pump effect—‘‘BCP'')forms organic matter.This suggests that variations of hydrochemistry in the lake waters weremainly due to aquatic photosynthesis,and thus predominantly controlled by the BCP effect.

Fig.3 Depth profiles of the physiochemical parameters monitored in winter(a),autumn(b),summer(c),and autumn(d)at sites FX-1,FX-2,and FX-3 in Fuxian Lake(1/2017—10/2017).The dashed line represents the CO2 concentration in equilibrium with the atmosphere

5.2 Sources of sedimentary organic matter in trap sediments of Fuxian Lake

The C/N ratios of organic matter can be used effectively to indicate the relative contribution of allochthonous versus autochthonous organic matter to lake sediment(Meyers and Ishiwatari 1993；Meyers 1997,2003).Generally,C/N ratios for phytoplankton,soil,and terrigenous plants are 6.6(Redfield 1963),10—13(Gon~i et al.2003),and 20(Meyers 2003),respectively.In this study,the average value of C/N of lake sediment OC was 8.81(Table 3)；this is consistent with a composite of the three,with the aquatic plant proportion being predominant.δ13Corgvalues may also be used to determine the relative contributions of autochthonous organic matter to lake sediment(Meyers and Ishiwatari 1993；Meyers 1997).Aquatic plants tend to have lighter δ13Corgvalues due to the isotopic fractionation of aquatic photosynthesis(Falkowski and Raven 1997).As shown in Fig.7,data from Fuxian Lake trap sediments were within the range for autochthonous organic matter and were different from the range for allochthonous organic matter. In addition, there was a positive relationship between δ13Corgand δ13Ccarb(R2=0.51,P ＜0.05,n=9；Fig.8)in this year-long study of trap sediments,suggesting both inorganic and organic carbon reflected variations in the isotopic composition of contemporaneous DIC(Meyer et al.2013).This confirmed that organic matter in the Fuxian Lake sediments was chiefly derived from aquatic plants.

Fig.4 Piper diagram of water samples from Fuxian Lake

Fig.5 A comparison of seasonal variations(1/2017—10/2017)in δ13CDIC at sites FX-1,FX-2 and FX-3 in Fuxian Lake

In the following,we quantitatively determine the proportions of the autochthonous and allochthonous organic matter in lake sediments by examining C/N ratios.

Terrigenous allochthonous C/N ratios were 12.71(n=6) using the average values of soil data in this watershed collected during this study and Ding et al.(2016),which were close to the experimental range of end element values(10—13)of soil organic carbon(Gon~i et al.2003).It is important to note that we regard soil as a typical allochthonous source because the direct input from terrestrial vegetation is relatively small.Liang et al.(2018)found that the C/N ratio of primary producers in ten lakes with different nutrient levels in Eastern Yunnan had low C/N ratios,with mean values of 7.68,slightly higher than the mean value of 6.69 from the present study of Fuxian Lake.Similarly,autochthonous the C/N ratio(7.02)was determined by the average value of these studies.Based on the Eq.(4),the proportions of autochthonous OC ranged from 30 to 100%(Table 4).The mean value of the proportions was 58%in trap sediments,which was comparable to that(61%)obtained in a recent study(Chen et al.2018),indicating that lake OC was largely affected by the BCP effect.

Table 3 Dry weight,TOC and TIC contents,C/N ratios,δ13Ccarb,δ13Corg,OCAR,ICAR and Total AR determined from sediment traps at site FX-2,Fuxian Lake,from different water depths in different seasons

Table 4 The contribution percentages of autochthonous OC and allochthonous OC to the bulk OC in the sediment traps at site FX-2,Fuxian Lake,from different water depths in different seasons

Fig.6 Relationship between pCO2 and DO in winter(a),autumn(b),summer(c)and autumn(d)at sites FX-1,FX-2 and FX-3 in Fuxian Lake

5.3 Autochthonous organic carbon sedimentation in the lake

5.3.1 Synchronous increase in OCARauto and ICAR caused by the BCP effect

Fuxian Lake was saturated with calcite throughout the year and precipitated calcite in summer and fall in response to increases in pH caused by increased rates of aquatic photosynthesis.This increased organic productivity,due to the strong BCP effect, may cause greater precipitation of CaCO3(Kelts and Hsu¨1978).According to Eq.(6),when the BCP effect occurs in aquatic ecosystems,there will be a synchronous increase in OCARautoand ICAR.In the current study, the significant positive correlation between OCARautoand ICAR in the trap sediments(R2=0.69,P ＜0.006, n=9； Fig.9) illustrate this. However, the observed increase in OCARautowith depth can be explained either by input from the euphotic zone,lateral input into the traps,or resuspension of OC,particularly in the very deep trap(Usbeck et al.2003).This phenomenon can also be shown in ICAR.Therefore,we chose the upper trap(40 m)to conservatively calculate the OCARautoas the actual autochthonous OC produced by the BCP effect,mainly to reduce the effect of resuspension.By that calculation,the AOC flux in the Fuxian Lake was 20.43 t C km-2in 2017.

Fig.7 Distinctive source combinations of C/N ratios and δ13Corg values of the trap sediments,soil,foliage,and phytoplankton in Fuxian Lake

Fig.8 Correlations between δ13Corg and δ13Ccarb in the trap sediments during the monitoring year,1/2017—1/2018

5.3.2 Possible impacts of anthropogenic activities and climatic changes

Fig.9 Correlations between OCARauto and ICAR in trap sediment during the monitoring period January 2017—January 2018

There is now a general agreement that the biogeochemical cycles of phosphorus and nitrogen have had considerable impacts on global lake productivity(Schindler et al.1999).The increased OCARautofrom July to October may have been caused by increased river-borne nutrients and phosphorus inputs to the euphotic zone from the more than 20 short rivers across the basin of Fuxian Lake.Indeed,the concentrations of TP in the surface water were significantly higher in the wet season,which was mainly caused by pollutants contained in runoff(Yao et al.2017).The temperature could be an alternative explanation for the high OCARautofrom July to October.The proportion of OC deposited(i.e.,the burial efficiency)will increase with high temperatures due to enhanced aquatic photosynthesis(Lewis 2011；Brothers et al.2008；Sobek et al.2014),though warmer water temperatures can lead to mineralization and less organic carbon burial,as previously discussed(Gudasz et al.2010).

6 Conclusions

By combining modern hydrochemical monitoring and a sediment trap experiment in Fuxian Lake(Yunnan,SW China),we conducted a systematic study to understand how much of the AOC was generated by the BCP effect can be deposited effectively and what the underlying mechanisms were.The main findings of this study were as follows:

1. Temperature, pH, EC (electric conductivity), DO(dissolved O2), [HCO3-], [Ca2+], SIc, and partial pressures of CO2(pCO2),as well as carbon isotopic compositions of HCO3-(δ13CDIC)in Fuxian Lake water, all displayed distinct seasonal and vertical variations.This was especially apparent in an anticorrelation between pCO2and DO,indicating that the variations of hydrochemistry in lake waters were mainly controlled by the photosynthetic metabolism of aquatic autotrophs.

2. The low C/N ratios and high δ13Corgwere recorded in the trap sediments.From observed C/N ratios,the proportions of AOC were estimated to be 30%—100%of all OC,indicating that AOC was an important contributor to sedimentary OC.

3. It was calculated that AOC flux in the Fuxian Lake was 20.43 t C km-2in 2017,and this was affected by synergistic effects,including nutrient availability and climatic factors.

4. AOC formed by coupled carbonate weathering and aquatic photosynthesis(the BCP effect)may be a significant potential carbon sink and may make an important contribution to resolving the problem of the missing carbon sink in the global carbon cycle.If this process holds for all global lakes,the challenge would be to enhance carbon sequestration via the BCP effect.

AcknowledgementsThis work was supported by the National Natural Science Foundation of China(Nos.41430753,U1612441).