Interactive comment on “ Mercury dynamics and mass balance in a subtropical forest , southwestern China ”

This manuscript presents measurements of litterfall, throughfall, output stream water, forest floor and soil mercury concentrations, as well as air-soil exchange and other fluxes in a one year study in China. Those measurements were used to estimate mass balance of the forest under investigation. The topic is relevant to ACP, however, some results could be erratic. Furthermore, the scientific contribution is hindered by a lack of in-depth discussion considering results of similar studies conducted in the same region had been published. Moreover, the presentation has much room for improvement. My comments are listed below. Major concerns:


Introduction
Mercury (Hg) can cause damage to the environment and human health due to its extreme toxicity.It is well established that gaseous Hg can travel a long distance in the atmosphere so that aquatic systems in remote regions can be impacted by Hg pollution through deposition from the atmosphere (Lindberg et al., 2002a, b;Feng et al., 2009a, b).As a consequence, atmospheric deposition is the principal form of total Hg (THg) input to aquatic systems in remote pristine regions.Although Hg emissions must be reduced to mitigate current Hg contamination in surface water and fish, the magnitude of that reduction is a critical policy debate.
Forest ecosystem is generally regarded as an active pool of Hg.Hg transformation processes in the forest is considered a vital part of global Hg cycling (Ericksen et al., 2003;Sigler et al., 2009).Most of the Hg accumulated in canopy foliage comes from atmospheric sources, rather than root uptake (Ericksen et al., 2003;Stamenkovic and Gustin, 2009).The forest canopy is a major receptor of Hg in forested landscapes (St. Louis et al., 2001).The deposited Hg to the forest may produce a certain ecological risk on the biogeochemical cycle of Hg in the forest watersheds.Hg accumulated in the forest soil may be considered a source of both total and methyl Hg (MeHg) to aquatic ecosystems through runoff and/or stream flow.Moreover, Hg in the forest soil and decomposed litterfall can transfer into MeHg, resulting in increased MeHg levels in downstream wet areas.Thus the release of Hg compounds from the forest field can be considered an initial step of Hg mobilization in forested catchments, and seems to be of high importance for its mobility.
Biogeochemical mass balance studies quantifying Hg pools and fluxes in the forest ecosystem are essential for assessing current rates of Hg inputs to, retention within, and release from terrestrial ecosystems.Major research initiatives have improved our understanding of current Hg pools and fluxes (Grigal et al., 2000;Schwesig and Matzner, 2001); however, knowledge of the internal cycling dynamics controlling retention within and release from these ecosystems which are located in the elevated Hg emitting regions is still limited (Demers et al., 2007).China's rapid economic development is predicted to increase the emission of atmospheric Hg (Fu et al., 2008(Fu et al., , 2010a)).The coal burning leads to Hg pol-lution in industrial and urban areas, as well as remote areas due to the long-range atmospheric transport of Hg (Feng and Qiu, 2008;Fu et al., 2009).In this study, we conducted a fullscale investigation on the distribution of Hg in the throughfall, litterfall and precipitation for a year.At the same time, we calculated the output and input of Hg during the study period.Thus, the objectives of this study were to (1) evaluate the deposition and output fluxes of Hg in the forest field and the accumulation of THg in a subtropical forest soil pool of southwest China, (2) discuss Hg import and export characteristics via deposition and runoff and/or stream flow in the study field, and (3) explore the main factors affecting Hg deposition, retention within and output fluxes in the subtropical forest ecosystem.

Site description
We conducted this research at Mt. Simian National Natural Reserve (106 • 22 -106 • 25 E, 28 • 35 -28 • 39 N), which is situated about 200 km away from Chongqing city (Fig. 1).Chongqing is the largest industrial city in southwest China, where combustion of coal accounted for more than 75 % of the regional energy supplies in recent years.The study area has a subtropical monsoon climate, with abundant rainfall every year.The mean annual temperature is 13.7 • C, with the highest and lowest records in August (average: 31.5 • C) and January (averages: −5.5 • C) respectively.The mean annual precipitation in the study area is 1522.3mm with a daily maximum up to 160.5 mm (Lv et al., 2014).There are four seasons in Chongqing: spring (March to May), summer (June to August), autumn (September to November), and winter (December to February), with a well-defined wetwarm season from June to October.The study area is typical of the region with hills of 1394 m and watersheds of about 100.1 km −2 .The evergreen broad-leaf forest selected in our research is believed to be one of the most representative vegetation types preserved in the study reserve due to the following reasons.First, it is one of the most complete forests located in the area between Chongqing and Guizhou.Second, it is almost all subtropical forest and until recent decades is one of Asia's least populated and most inaccessible areas.Third, it is the only largest and intact forest in the same latitude of the earth.Therefore, the evergreen broad-leaf forest was selected as the representative forest of subtropical vegetation in this research.

Sampling method of throughfall and precipitation
The throughfall samples were obtained from the evergreen broad-leaf forest where Hg dynamics have been investigated for 1 year, from March 2012 to February 2013.The precipita- tion samples were collected by automatic precipitation samplers (APS-3A, Changsha Xianglan Scientific Instruments Co., Hunan, China), which were placed on the forest field of the sampling site.The throughfall was collected and measured using the same rain gauges (APS-3A; Ma et al., 2015).Four rain gauges were randomly placed in each of the three 20 × 20 m 2 permanent observation plots in each plantation, resulting in 12 throughfall sampling points for each plantation.The containers were pre-washed with dilute (5 %) HCl and thoroughly rinsed with deionized water after each sampling.Moreover, the throughfall and precipitation samples were collected after each precipitation event from each site during the sampling period.
The volume-weighted mean concentration (VWM) is obtained with the formula introduced in Acid Deposition Monitoring Network in East Asia, 2012: where X t means the ion concentration in each precipitation event (ng L −1 ), V t means the volume of each rainfall (mm).Hg flux was determined by multiplying Hg concentrations by the volume of precipitation collected.Wet deposition fluxes of THg and MeHg were calculated according to the following equation: where F w is the annual THg or MeHg wet deposition flux (mg m −2 yr −1 ), C i is the VWM (ng L −1 ) of each rain sample, and P i (mm) is the precipitation or throughfall amount.

Sampling method of the stream
The stream and/or runoff was collected at the edge of the forest catchment.For the water yield of the stream and/or runoff, it was monitored in the outlets of the forest catchment by the local hydrological departments.Stream water samples for analysis were collected in seven sampling sites every 2 weeks from March 2012 to February 2013.The 250 mL Teflon bottles were used to collect the stream water samples.All Teflon bottles were cleaned with detergent, thoroughly rinsed with tap water, boiled in a 30 % HNO 3 solution (v/v) for 1.5 h, rinsed and filled with Milli-Q water.Trace-metal grade HCl was immediately treated to the samples to acidify them.It should be noted that the water samples were not filtered and thus represented the stream load of THg.The subtropical forest field in the study area is 100.1 km −2 .Annual water discharge of the study site is 1.86 × 10 8 m 3 (hydrological departments of Jiangjin district).The annual precipitation (Table 1) of the sampling site is slightly lower than the annual discharge.Therefore, it can be assumed that the calculation has a certain representativeness.Volume-weighted concentrations were computed by the stream/runoff collected during the study period.The fluxes were achieved by multiplying the average Hg concentration by the total amount of runoff during the study period.

Sampling method of litterfall and soil pool
The litter samples were collected by self-made litter collectors (0.5 m × 0.5 m), which were made from treated lumber with a screen bottom.During the study period, the collectors were placed at four different sites within the study field.
The litter collected was saved in brown paper bags and transported to the laboratory under 4

Hg volatilization from the forest field
A dynamic flux chamber (DFC) in series with Tekran 1110 synchronized dual-port sampling unit and Tekran automated Hg analyzer (2537X) were used to measure the emission rates of Hg 0 (Fu et al., 2010b).The special DFC method was described in detail in Ma et al. (2013).Hg emission fluxes were calculated by the equation below (Ma et al., 2013): where F is the flux (ng m 2 h −1 ); C out and C in are Hg 0 levels at the outlet and inlet of the Hg analyzer (ng m 3 ); Q is the flushing flow rate through the chamber (m 3 h −1 ); and A is the surface area of the soil exposed in the chamber (m 2 ).Hg emission fluxes across the air-soil interface were monitored seasonally during eight intensive field campaigns from spring 2012 to winter 2013 (Spring: 4-16 March, 8-15 May 2012;Summer: 5-12 July, 15-22 August 2012;Autumn: 15-22 September, 20-27 October 2012;Winter: 24-31 December 2012, 6-13 February 2013).Hg emission flux were monitored at three sampling sites in the evergreen broadleaved forest of Mt.Simian.Hg emission fluxes were measured continuously for 7 days for each sampling.Quality assurance was conducted by manually injected Hg to the ambient air and soil vapor of the Tekran analyzer before and after data collection.At the beginning and end of each measurement date, the Hg fluxes over a clean TeflonTM sheet in the field was measured and regarded as the chamber blanks.The chamber blanks in our research ranged from 0.48 to 0.64 ng m −2 h −1 , with an average of 0.54 ± 0.07 ng m −2 h −1 (n = 12).No blank value was needed to be subtracted from the flux due to no significant difference found.

Sample analysis and quality control
For THg and MeHg in water samples, the special method was described in Ma et al. (2015).Detailed introduction of the measurement of THg and MeHg in soil and litter samples can also be found in Ma et al. (2015).The detection limits of THg and MeHg in this research were 0.02 and 0.01 ng L −1 respectively.The dissolved total mercury (DHg), dissolved methylmercury (DMeHg), particulate Hg (PHg) and particulate MeHg (PMeHg) were analyzed following the EPA Method 1631 (US EPA, 1999).The method blank was lower than detection limits in all cases.And the equipment blanks for THg and MeHg were 0.04 and 0.02 ng L −1 , respectively.The detection limits of Tekran 2537X was 0.1 ng L −1 for GEM.The average relative standard deviation for the duplicate analyses of THg and MeHg was 5.2 and 5.4 %, respectively.Matrix spikes recoveries for THg and MeHg were both within an acceptable range, 89 to 115 % for THg, and 91 to 117 % for MeHg.Precision was determined by relative standard deviations.For duplicate samples, the precision was 5 % for THg in water samples, 9 % for MeHg in water samples, 8 % for THg in soil samples, and 4.1 % for THg in leaf tissues.

Hg concentrations and deposition fluxes in throughfall and litterfall
THg concentrations in the throughfall ranged from 3.2 to 62.5 ng L −1 for the individual samples, and the average level of throughfall was 24.1 ± 7.9 ng L −1 .Canopy density did have an effect on THg and MeHg concentrations (the forest cover is more than 90 % and the canopy density is 0.9).THg concentrations measured in the throughfall of the subtropical evergreen broadleaved forest (24.1 ± 7.9 ng L −1 ) were significantly higher than those measured in the open field (10.9 ± 3.1 ng L −1 ).Similar to THg concentrations, MeHg concentrations in the throughfall were nearly 2.5 times higher than that in precipitation (p = 0.004, n = 49).THg concentrations in the throughfall were consistently at their highest in the cold season (Fig. 1), which was probably due to the lower rainfall but elevated atmospheric Hg in this season.In the subtropical region of China, the monsoondriven climate of northwest China does not bring much precipitation in the cold season.At the same time, atmospheric stability is high during the cold period, and pollutants like atmospheric Hg do not spread easily, which contributes to the higher scavenging ability of Hg in the atmosphere.The warm season (from April to September) is influenced by the southeast monsoon, and the rainfall increases greatly, which leads to lower concentration of Hg (Fig. 1).The study station has obscure seasons and clear rainy and dry seasons.The deposition fluxes of THg through throughfall also showed the seasonal variation characteristics, with higher fluxes appearing in the wet season (June to August).The deposition fluxes of THg through throughfall in summer at Mt. Simian accounted for more than 40 % of total annual Hg deposition.However, it is still in September and October that a higher throughfall flux is observed.This may be because the rainfall in the 2 months is mainly effected by Indian Monsoon, contributing to a higher rainfall (Fu et al., 2010a) compared with other months.The minimum values for THg deposition occurred in the cold season.
During the surveillance, THg in the throughfall was evaluated to be 32.2 µg m −2 yr −1 .The deposition fluxes of THg through throughfall in Mt.Simian were lower than those investigated in the southwestern cities of China, such as Guizhou and Chongqing (Precipitation: 8.4-62.1 µg m −2 yr −1 , throughfall: 15.6-292.1 µg m −2 yr −1 ; Guo et al., 2008;Feng et al., 2009a, b;Wang et al., 2007).However, the deposition fluxes of THg through throughfall in Mt.Simian were approximately 2-10 times higher than those reported in remote areas of North America and Europe (Precipitation: 3.1-10.0µg m −2 yr −1 , throughfall: 6.7-23.0µg m −2 yr −1 ; St. Louis et al., 2001;Keeler et al., 2005).Obviously, the THg fluxes at Mt. Simian were higher than other sites abroad.The reason perhaps was that Mt.Simian had considerably more dense forest canopies.As one of the National Natural Reserves of China, it has preserved the best subtropical evergreen broad-leaved forest of China.The forest cover in the reserves reaches over 90 %.The increased THg concentrations in throughfall mainly resulted from the dry deposition of Hg on the vegetation, followed by the washout of throughfall.Another possible reason for the elevated deposition fluxes, which may be the most important one, probably related to the increased atmospheric Hg concentrations in the past 30 years due to China's fast economic development.This area, especially Chongqing city, has a large demand for energy consumption, and about 70 % of which is from coal combustion.The annual mean gaseous elemental Hg (GEM) concentration in the middle of Chongqing city (9.6-31.9ng m 3 ; Wang et al., 2006), more than 200 km away from the study site, tripled compared with global background level (Lindberg et al., 2002a, b), which corresponded to the high annual deposition flux of Hg in the study area.It is also reported that the GEM concentration in the study area is as high as 3.8 ± 1.5 ng m −3 (Ma et al., 2015), even if it is situated in a natural subtropical forest reserve.The MeHg flux was 0.45 µg m −2 yr −1 , which was higher than those measured in other areas.While MeHg/THg in the throughfall samples was 1.3 %, which was a relatively high value compared with other studies (0.4-0.8 %; Lee et al., 2000;Demers et al., 2007;Choi et al., 2008;Fu et al., 2010a;Guo et al., 2008;Larssen et al., 2008).Here, the higher ratio of MeHg to THg in throughfall samples may suggest that the contribution of MeHg from throughfall cannot be ignored and should be taken care of in future studies.After all, accumulation of MeHg in the soil might have caused serious risks in the functioning of natural downstream ecosystems.
The deposition fluxes of THg through litterfall are shown in Table 1.The average concentrations of THg and MeHg in leaf litter were 106.7 ± 18.3 ng g −1 (SE = 2.6, N = 60) and 0.8 ± 0.4 ng g −1 (SE = 0.2, N = 60), respectively.The deposition flux of THg through litterfall was estimated to be 42.9 µg m −2 yr −1 in the measurement field, which was remarkably higher than the input flux through throughfall.It is also considerably higher than litterfall fluxes reported from other regions (St. Louis et al., 2001;Demers et al., 2007).GEM can be absorbed by stomata and detained in the leaf tissue (Ericksen et al., 2003;Fu et al., 2008Fu et al., , 2010a)).Therefore, we believed that the elevated litterfall input fluxes directly related to the increased GEM concentrations, even in remote areas.

Mercury emission from soils under the canopy
The emission characteristics and air-surface exchange of GEM from the subtropical forest field have been investigated during eight intensive field campaigns from 2012 to 2013.At the forest field, GEM released from soils had the characteristic of obvious diurnal and seasonal variations.Day and night GEM fluxes were statistically different (t test, p < 0.001), with nighttime emissions considerably lower than that in daytime in all seasons (Fig. 2).Average fluxes of Hg in spring, summer, autumn and winter were 12.2 ± 5.1, 14.2 ± 4.7, 9.9 ± 2.5, and 3.1 ± 1.1 ng m −2 h −1 , respectively.It can be seen that the highest value occurred in summer, followed by spring and fall, while the lowest value was observed in winter.The average fluxes of Hg in spring (12.2 ± 5.1 ng m −2 h −1 ) were slightly lower than that in summer (14.2 ± 4.7 ng m −2 h −1 ), which was different from other studies (Larssen et al., 2008;Fu et al., 2010b).It appeared that warm temperature with low canopy density in spring at the mid subtropical forest was more likely to release GEM.Perhaps the primary reason lies in the fact that the forest canopies are lushly and well spaced in spring, and thus the forest can receive more sunlight.Therefore, the reduction rate of Hg 2+ by photochemical, thermal and biogenic processes probably increased.
This research indicated that Hg fluxes of forest field were far lower than those observed from contaminated areas such as heavily air-polluted areas in eastern Guizhou (33-3638 ng m −2 h −1 ) (Wang et al., 2007), some cities in southwest China (15.0-44.4ng m −2 h −1 ) (Qiu et al., 2006), dry landfills (46.5-22.8ng m −2 h −1 ) (Zhu et al., 2010) and wetlands (20-500 ng m −2 h −1 ) (Lindberg et al., 2002b).But the emission of GEM elevated in comparison with those reported from other places (−5.4-4.2 ng m 3 , Lindberg et al., 2002a,  b; 1.7-8.4ng m 3 , Travnikov, 2005).At Mt. Simian, the estimated net GEM fluxes were released from soils during the warm season and slightly volatilized during the cold season.Hg deposition was only observed in several nights of the cold season during the study period.Hg released from the snowair interface was extremely low comparing with the soil-air interface.It was supposed that the Hg 0 flux was zero from snow-covered surface (Huang et al., 2012).At most subtropical areas, especially mid-subtropical forests, however, there were short winter seasons with unstable snow cover, and the snow cover season only tended to occur in January.So we assumed that there still existed Hg 0 emission in December and February in winter.Therefore, the annual total net Hg emission flux was 18.6 mg m −2 yr −1 .

Hg concentrations and out-flux in stream water
Annual volume-weighted concentrations of THg and MeHg were measured at the outlet stream of the forest field of Mt.Simian.The mean concentrations of THg and MeHg in the outflow stream were 3.9 ± 2.0 and 0.2 ± 0.08 ng L −1 , respectively.THg and MeHg concentrations in stream water draining the upland in our research were slightly higher than those reported in literature (Fu et al., 2008(Fu et al., , 2010a;;Larssen et al., 2008).THg concentrations in runoff and/or stream water in rainy seasons (4.6 ± 2.0 ng L −1 ) were significantly higher than those in dry seasons (3.3 ± 1.8 ng L −1 ), which can probably be attributed to the soil erosion and runoff (Ma et al., 2015).It is known that if a remote forest field does not have other obvious Hg pollution sources, Hg concentrations in the runoff and/or stream water can represent risks from a solitary watershed.Numerous studies showed that the catchments of remote forest were regarded as filters between atmosphere and hydrosphere (Lee et al., 2000;Larssen et al., 2008;Ericksen et al., 2003).The fate of Hg stored in the forest soils can be divided into three parts.One part of them transfers through food webs, threatening the balance of forest ecosystems; the second part of them is released into the atmosphere again; the third part of them probably transfers with the runoff and/or stream, becoming one of the Hg sources of downstream aquatic ecosystem.Therefore, to a certain extent, the role of forested catchments as Hg filters can be characterized by Hg output (runoff/stream) from the forest field.This study showed that, even though Hg deposition fluxes in throughfall was high, THg concentration in stream and/or runoff was lower than that in contaminated sites under the same geological background.This indicated that subtropical forest field had the filtering effect of Hg in precipitation and throughfall.On the other hand, the lower concentration in stream and/or runoff indicated that the study area did not suffer from severe anthropogenic Hg pollution.Steam output of THg was calculated by multiplying the average THg concentration in stream water (3.9 ± 2.03 ng L −1 ) by the water discharge rate in the forest field of the study site (annual water discharge: 1.86 × 10 8 m 3 , from hydrological departments of Jiangjin district).The export flux of THg via runoff and/or stream was 0.73 kg yr −1 .The subtropical forest field in the study area is 100.1 km −2 .So the export flux of THg through stream water was 7.23 µg m −2 yr −1 , which tripled the values reported in the catchments of Sweden (1.6-1.8 µg m −2 yr −1 , Lee et al., 2000;2.4µg m −2 yr −1 , Larssen et al., 2008).Our results indicated that the output fluxes of MeHg via stream water were 0.08 µg m −2 yr −1 , which was similar to or slightly larger than other results (0.03-0.07 µg m −2 yr −1 , Lee et al., 2000;0.05µg m −2 yr −1 , Schwesig and Matzner, 2001).As we mentioned above that Mt.Simian was one of the most complete forest and until recent decades was one of Asia's least populated and most inaccessible areas.Aver-age Hg concentration in the soil detected in previous research was 0.28 mg kg −1 (Ma et al., 2013), which indicated that it was not an obvious geological Hg hotspot.Therefore, the elevated Hg fluxes in stream water were probably attributed to the great atmospheric Hg depositions.At the same time, our preliminary research results also illustrated that forest runoff and soil erosion could increase Hg output from subtropical forest catchments (Ma et al., 2013).But the total output fluxes of THg and MeHg were far lower than the input fluxes via wet deposition (32.2 µg m −2 yr −1 for THg and 0.5 µg m −2 yr −1 for MeHg).This study showed that the subtropical forest was able to exert purification effect of filtration, even under the condition of elevated deposition of Hg.

Dynamics and transport of Hg based on forest field
THg content in the forest field (forest floor and soil profiles) of Mt.Simian was shown in Table 2.The THg stocked in the forest soil was approximately 20192 µg m −2 (average soil depth is 98 cm), while that in the organic floor was 5148 µg m −2 (average litter depth is 19 cm).THg content in soil profile was three times more than the organic horizon in the subtropical forest field.The active pool (the upper 22 cm, O i ) of THg represented 41 % of the total storage of the study area.In the soil profile, THg content in the organic horizon (O i ) is obviously higher than those in the other horizons.At the same time, the organic matter is well decomposed under a warm and rainy subtropical climate, which has high affinity to Hg in soil.Due to the good adsorption and reduction of organic matter, the organically bound contents of Hg could be released into the environment again during the decomposition of organic matter.
The ultimate fate of Hg in the terrestrial ecosystem may depend upon its delivery and incorporation into the forest floor.And the average Hg fluxes were also estimated.Input of THg to the forest field of Mt.Simian included net throughfall and litterfall depositions (St. Louis et al., 2001;Fu et al., 2010b).Annual throughfall and litterfall deposition fluxes of THg in Mt.Simian were 32.2 and 42.9 mg m −2 yr −1 , respectively (Fig. 3).Litterfall deposition inputs were estimated to be 134 % of the throughfall deposition at the forest field.In the study forest field, the predominant pathway of Hg fluxes to the forest floor was via litterfall (57.1 %).A majority of atmospherically deposited THg was released through Hg 0 at a rate of 18.6 µg m −2 yr −1 .Steam outflow of THg from the wetland was 7.2 µg m −2 yr −1 .The ratio between output and input of THg was 0.34 at the subtropical forest field of Mt.Simian, which was significantly higher than others (0.02-0.04, Lee et al., 2000;0.16, Larssen et al., 2008;0.30, Fu et al., 2010b).The apparently higher ratio between the output / input fluxes of THg may represent an important ecological risk.
The THg stored in the forest field was 982 times larger than the annual THg output by stream and/or runoff outflow and soil volatilization, and 338 times larger than the input estimated from wet and dry depositions (Fig. 3).The estimates of the deposition flux of THg in this study were much higher than values reported from northeastern American (3.8-12.6 µg m −2 yr −1 ; Driscoll et al., 2007) and Norway (7 µg m 2 yr −1 ; Larssen et al., 2008), probably indicating a significant impact of heavy regional Hg emissions from industry and urban on local Hg deposition.The reason perhaps was that highly soluble Hg 2+ was easily stripped from the atmosphere and deposited locally.Higher wet deposition can illustrate the remarkable influence of local Hg emissions on Hg accumulation in the regional forest field.The THg flux through litterfall was 1.5 times larger than that through throughfall due to greater input of litter mass and higher Hg concentrations in the litter.Annual exports of THg in stream water of the study area (3.2-9.5 µg m −2 yr −1 ) were not accorded with those reported from northern forest catchments (1.0-3.4 µg m −2 yr −1 ; St. Louis et al., 2001;Grigal et al., 2000).An amount of the atmospherically deposited THg was lost via emissions at a rate of 18.6 µg m −2 yr −1 .Compared with stream outflow, the evasion of Hg from forest soil played a more essential role in THg outputs.
Compared to the ratios of output to input flux with other places (Larrsen et al., 2008;Fu et al., 2010b), the higher ratio may be greatly affected by the elevated deposition.Therefore, regional emission of Hg may have a stronger influence on forest ecosystems, in which the deposition of THg through litterfall and throughfall, being affected more by local and regional changes of Hg emissions and cycling, were the main paths for Hg entering into soil surface (Demers et al., 2007).However, in this study, the outflow of Hg via runoff output and the soil-air interface accounted for a small fraction of Hg budget in the study area.And the accumulation pattern of Hg in forest floor and soil profiles was seasonal.As we mentioned above that Mt.Simian was one of the most typical subtropical forest systems and the least accessed area, average Hg concentrations in all soil surfaces of this area were below 0.30 mg kg −1 (Fu et al., 2010b;Ma et al., 2013).Therefore, the accumulation of Hg in soil would be enhanced with time.At the same time, the ultimate fate of deposited Hg de-pends upon the biogeochemical processes that have not been well quantified within the ecosystem.Hg dynamics during litter decomposition, for instance, need to be considered as a whole so that we can better understand controls on long-term accumulation of Hg in the forest ecosystem and its delayed release to surface water.

Conclusions
In this study, the mass balance and transport of Hg in southwestern China were first measured at a subtropical forest, Chongqing, China.Results revealed that litterfall deposition inputs were the predominant pathway (42.9 mg m −2 yr −1 , account for 57.1 %) of Hg flux to the forest floor.Annual deposition fluxes of Hg through throughfall were 32.2 m −2 yr −1 , accounting for 42.9 % of the Hg inputs.Researchers should pay more attention to the higher ratio of MeHg to THg in the throughfall deposition when they model the biogeochemical cycling in a typical local forest watershed.For the output process, the exchange of Hg (18.6 µg m −2 yr −1 ) across the forest field-air interface was an essential part of the biogeochemical cycle of Hg.The runoff and/or steam outflow of THg from the wetland was 7.2 mg m −2 yr −1 , which was lower than that in contaminated sites under the same geological background.Therefore, we may conclude that (1) the study area does not suffer from severe anthropogenic Hg pollution; (2) the forested field has the filtering effect of Hg in precipitation and throughfall, even in the elevated atmospheric Hg area.
The forest field (forest floor and soil profiles) plays an important role in the cycling of THg and MeHg.In reality, it is just another problem created by the accumulation of Hg, which would be a potential risk affecting the output of Hg in the long term.Terrestrial ecosystems that have accumulated more Hg may ultimately emit them to the wetlands and surface water, finally affecting the entire aquatic ecosystems.Therefore, it is a signal that we should not ignore.In this case, however, any changes in the forest floor like deforestation or forestland degradation may strongly affect Hg budget of the region.

Figure 1 .
Figure 1.Volume-weighted mean concentrations of THg and MeHg and deposition fluxes of throughfall in the evergreen broad-leaf forest from March 2012 to February 2013.

Table 1 .
Mean values of THg and MeHg concentrations and deposition fluxes in throughfall and litterfall.

Table 2 .
The concentrations and contents of THg in forest floor and different soil layers.