Increasing Anthropogenic Mercury Pollution over the Last 200 Years Revealed by Lagoonal Sediments from Hainan Island, South China

The toxic heavy metal mercury (Hg) has significantly enhanced the global Hg cycle influenced by human activities over the last century. In this study, we presented a high-resolution Hg deposition history between ∼ 1780 and 2015 AD in a sediment core from Xincun Lagoon, located in the southeastern Hainan Island, South China, and analyzed it in conjunction with geochemical elements, grain-size distribution, organic matter, and HYSPLIT backward trajectory simulation. The objective was to investigate the influencing factors affecting historical Hg deposition in relatively remote regions and assess the extent of the effects of natural background and human activities. The results showed that the Hg in the sediment was deposited primarily through atmospheric deposition, which was closely related to regional and even global human activities. Anthropogenic Hg contamination increased gradually from the 1830s to 1850s, possibly due to Hg emissions from Opium Wars I and II occurring in southeastern China. High broad peaks of anthropogenic Hg were observed during the 1910s to 1950s and in the 1980s, likely associated with the two world wars and modern Chinese wars. In addition, a further sharp increase in anthropogenic Hg from the mid-1970s to the present occurred, likely originating from the intense industrial activities in China triggered by the reform and opening-up policy of China in 1978 and some countries in Southeast Asia.


Introduction
Mercury (Hg) and its derivatives, especially methylmercury (MeHg), pose a potential threat to the environment and human health after biomagnification in aquatic food chains because of its toxicity, persistence, and bioaccumulation (Beckers and Rinklebe, 2017).Hg is emitted into the environment through natural processes (such as volcanic activity, weathering of rocks, geothermal activity, and forest fire) and anthropogenic activities (such as coal combustion, industrial manufacture, ore mining, and cement production) (Nriagu, 1989).The apparent onset of rising Hg concentrations and fluxes can be traced back to roughly the 1850s in preindustrial times, as evidenced by the reconstruction of Hg records from lake sediments, peat bogs, and ice cores in remote areas (Fitzgerald et al., 2005;Biester et al., 2007;Yang et al., 2010;Beal et al., 2015).However, widespread concerns regarding Hg toxicity and bioaccumulation have caused a massive reduction in Hg emissions in North America and Europe since the 1970s (Lindberg et al., 2007;Beal et al., 2015).
Anthropogenic activities have led to regional-and global-scale contamination of Hg due to the residence time of its element form, with Hg cycling between the atmosphere and terrestrial and aquatic systems (Fitzgerald et al., 2005;Pacyna et al., 2006;Lindberg et al., 2007;Liu et al., 2021).Recent compilations of large amounts of lake sediments derived from North America (Fitzgerald et al., 2005;Engstrom et al., 2014;Lepak et al., 2020) and Europe (Lindberg et al., 2007;Corella et al., 2018), and global emission inventory have demonstrated a consistent trend in atmospheric Hg depositional history, with high peaks between 1940 and 1990 AD attributed to the onset of World War II and industrial uses (Hylander and Meili, 2003;Cooke et al., 2020), and a subsequent decline.However, Asia has undergone rapid economic growth over the past few decades, making it a major contributor to anthropogenic Hg emissions, especially in China and Southeast Asia (Streets et al., 2005;Pacyna et al., 2006;Zhu et al., 2020).Therefore, asynchronous with the declining trend in North America and Europe since the 1970s, an upward trend in mercury emissions has been observed in China (Xu et al., 2011;Kang et al., 2016;Zhan et al., 2020).In fact, some records have already shown that there were significant rises in mercury emissions in China from the beginning of the industrial revolution due to enhanced human activities (Kang et al., 2016;Sun et al., 2016a;Zhan et al., 2020).
To distinguish and document temporal changes in natural and anthropogenic Hg discharges, quantifying the long-term history of atmospheric Hg deposition deduced from natural archives is essential (Hare et al., 2010).Lake sediments are excellent natural archives that provide an assessment of environmental changes in catchment transport and atmospheric deposition of Hg in connection with watershed ecosystems.Previous investigations of lake sediments have been widely used to reconstruct the temporal and spatial changes in Hg pollution histories in China (Zeng et al., 2017;Pan et al., 2020;Li et al., 2021;Wang et al., 2021), whereas most have focused on the relationship between Hg deposition and climate change, as well as human activities over a millennium time span.Although there have been various high-resolution archives including lake sediments, peat bogs, and coral skeletons from China (Yang et al., 2010;Bao et al., 2016;Sun et al., 2016a;Liang et al., 2022), high-resolution records from the remote region in South China are still scarce (Xu et al., 2011;Liu et al., 2012), especially in Hainan Island (Sun et al., 2016a;Ji et al., 2020).Sun et al. (2016a) analyzed Hg records in coral skeletons from the eastern Hainan Island and found a sustained increase in Hg concentrations after 1980 AD, which was attributable to regional industrial and urban activities.However, a sedimentary record derived from marine sediments in eastern Hainan revealed a decrease in Hg fluxes during the same period, reflecting a signal of global Hg production (Ji et al., 2020).There may be some discrepancies in historical Hg patterns in South China, and it is uncertain whether these differences actually exist or are ascribed to various time scales or natural archives.Thus, further investigation of new mercury accumulation records is necessary for a more thorough assessment of mercury deposition in South China.Noteworthy, it is critical to understand the depositional history of Hg in different regions since preindustrial times, and to explore potential sources of mercury deposition in conjunction with model simulations (Liang et al., 2022), for the purpose of developing national and even global environmental protection policies.
In this study, a high-resolution record of Hg observed from Xincun Lagoon in southeastern Hainan Island of China is presented.We aim to quantify historical variations in Hg accumulation and evaluate the impact of natural background and anthropogenic activities on Hg contamination in Hainan Island.This study will contribute to a better understanding of the release and accumulation of Hg in natural environment associated with human activities, from the perspective of sedimentary records from coastal lagoons, and in combination with model simulations.

Study area and sample collection
Xincun Lagoon is a nearly closed port in Lingshui County (Fig. 1), southeast Hainan Island, China, located between 18°24′N−18°26′N and 109°58′E−110°02′E, with an area of approximately 22 km 2 , an average depth of 5.7 m, and a maximum water depth of 11.2 m.There is no obvious runoff inflow in Xincun Lagoon, but only two streams with small watershed areas and runoff converge.The lagoon is connected to the South China Sea by a narrow mouth entrance of approximately 150 m, which is principally surrounded by granite.The width of the lagoon entrance has not changed significantly in recent decades (Fig. 1), according to a comparison of re-mote sensing images taken in 1988 AD and 2010 AD.The climate is mainly controlled by the East Asian monsoon system, dominated by the northeastern monsoon in winter and the southwestern monsoon in summer, with an annual average temperature of 24.7°C and precipitation of 1653.5 mm.
A sediment core of 64 cm in length was drilled at a water depth of ~ 4 m in May 2017 using a gravity piston sampler, marked as XC-03 (110°00′58″E, 18°25′16″N).After being transported to the laboratory, the upper 64 cm of the XC-03 core used in the present study was sectioned into 1 cm intervals in a PVC tube, yielding 64 subsamples, which were freeze-dried and preserved until analysis. 210Pb dating was performed using a high-purity germanium gamma spectrometer (GMX40P4, ORTEX, America) at Nanjing Institute of Geography & Limnology, Chinese Academy of Sciences.Odds samples numbered from 1−39 from the XC-03 core were selected for dating.The chronology was constructed using the constant rate of supply (CRS) model, which assumes that the 210 Pb flux to the sediment is constant over time (Appleby et al., 1986).In addition, an AMS (Accelerator Mass Spectrometry) 14 C date was measured at the Laboratory of Beta Analytic (U.S.) at a depth of 20 cm.

Mercury and other element analysis
Samples in the core were collected at 1 cm intervals and freeze-dried until Hg analysis.The Hg concentration (Hg C ) in bulk sediments was measured using a DMA 80 Hg analyzer (Milestone® DMA 80, Italy, detection limit 1.5 ng absolute).To minimize errors, each sample was analyzed at least three times, and the average analytical error was less than 5%.Major elements were analyzed using a PANalytical polarization energy dispersion Xray fluorescence spectrometer (Epsilon5, Almelo, Netherlands) with an accuracy of ~ 2%.Approximately 2-3 g of freeze-dried sample was pulverized using an agate mortar and pestle before being passed through a 200mesh sieve.Then, appropriate sieved samples and boric acid powder were pressed together under a pressure of 30 t for 30 seconds.Finally, all the samples were analyzed using the Epsilon5 spectrometer (Almelo, Netherlands).

Total organic carbon, dry bulk density, and grain size analysis
Total organic carbon (TOC) was determined using a multi N/C 3100 Analyzer (multi N/C®3100, Germany).
For TOC analysis, the freeze-dried sample was acid-digested in dilute hydrochloric acid (HCl) for 24 h in a beaker to remove carbonate until complete reaction.The sediment samples were then rinsed with deionized water three to four times and dried at 70°C.Approximately 200 mg of sample was weighed and passed through a 100-mesh sieve.Finally, all the samples were put into the machine and burned at 550°C.The dry bulk density (DD) was measured as the weight of the dry mass per unit volume (Janssens, 1983).A dry mass of exactly 1 cm 3 was determined by oven drying at 50°C until a constant weight was achieved.
Grain size distribution was measured by a Malvern Mastersizer 2000 Analyzer (England) with a measuring range of 0.02-2000 μm and an inaccuracy of 0.5%.The sample was pretreated to remove organic matter, calcium carbonate, and other impurities before the experimental determination.Organic matter was removed by adding approximately 10 mL of a 30% hydrogen peroxide (H 2 O 2 ) solution, and then 10 mL of 10% hydrochloric acid (HCl) was prepared to remove calcium carbonate.The sample solution was rinsed off with 2000 mL of deionized water and stored for 24 hours to remove acidic ions.Finally, before grain size analysis, the sample residue was treated with 10 mL of 0.05 mol/L sodium metaphosphate ((NaPO 3 ) 6 ) on an ultrasonic vibrator for 10 min to completely disperse the fine particles.Each sample was measured three times, and the average value obtained was used for analysis.

Mercury accumulation rates
The sedimentary rates, sedimentological conditions of the lake, and Hg emissions in the lake watershed can affect Hg flux in sediments.Therefore, it has been suggested that Hg accumulation rates are more suitable than Hg concentrations to reflect the depositional history of Hg in sediments.The Hg accumulation rate (Hg AR , ng/(cm 2 •yr) was calculated using the following equation: where Hg C is the mercury concentration (ng/g), DD is the sedimentary dry bulk density (g/cm 3 ), and SR is the sedimentation rate (cm/yr) calculated from the agedepth model.

Enrichment factors and anthropogenic mercury
Atmospheric Hg signals can be masked by terrestrial Hg fluxes and changing erosion rates (Bookman et al., 2010).
In lakes, which accept a significant part of sediment Hg via erosion of catchment soils, Hg is often bound to organic matter and/or minerals and finally transported to the lake (Fitzgerald et al., 2005;Perry et al., 2005).Therefore, an accurate assessment of natural mercury from soils is essential.
The enrichment factor (EF), calculated by normalizing the distribution of elements in sediments to reference diagenetic elements, provides technical approaches to evaluating metal pollution history.As a typical lithogenic element, aluminum (Al) is regarded as one of the most reliable erosion indicators and implies terrestrial input resulting from strong erosion in the watershed (Fitzgerald et al., 2005;Wang et al., 2022).Therefore, Al was used as a lithogenic reference to normalize Hg C for tracing the natural Hg contribution derived from soil erosion (Tylmann, 2005;Boës et al., 2011).The equation is as follows: where Hg EF is the enrichment factor; Hg sediment /ref sediment is the Hg/Al ratio in sediment; and Hg background /ref background is the Hg/Al ratio in the natural background.
The accumulation rate of anthropogenic Hg (Hg ARA, ng / (cm 2 •yr) was evaluated by Hg EF and Hg AR , representing the anthropogenic input of Hg into sediments, as follows (Boës et al., 2011;Hermanns and Biester, 2013):

HYSPLIT backward trajectory simulation
Potential sources of mercury in Xincun Lagoon were explored using the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT) model (Stein et al., 2015; (1970( , 1975( , 1980( , 1985( , and 1990 AD) AD).Finally, the transport routes of air masses for the summer and winter seasons were further clustered into four sub trajectory groups that traveled toward Xincun Lagoon, revealing the potential sources of mercury deposited in the lagoon.Since the 210 Pb dating method is only applicable to the last 100 to 150 yr, the 36-64 cm chronological sequence of the XC-03 core was established based on the average sedimentation rate of the upper 1-35 cm layers with the interpolation method using Origin 2022b software, revealing an age range of 1905 to ~ 1780 AD (Fig. 2c).

Mercury deposition history and possible factors affecting mercury deposition in Xincun Lagoon
As illustrated in Fig. 3, sedimentary Hg levels of Xincun Lagoon were the lowest and had no obvious fluctuation before ~ 1830 AD and rose gradually from 1830 to 1970 AD with two high intervals (e.g., the 1900s-1920s, 1930s-1950s).After ~ 1970 AD, Hg levels exhibited a steep upward trend and reached maximum values in the 2010s.Overall, the Hg deposition history in Xincun Lagoon agreed well with other Hg records from around China (Yang et al., 2010;Xu et al., 2011;Kang et al., 2016).
In light of the results reported here from a large suite of lake-sediment records, it is of great significance to analyze the role of climate change, debris input, organic matter (OM) and clay absorption, and atmospheric deposition in Hg accumulation (Lucotte et al., 1995;Perry et al., 2005;Hao et al., 2013).Lake sediments can record direct Hg-containing effluents, natural Hg inputs from catchment erosion, river runoff containing accumulations of atmospheric Hg deposits in terrestrial environments, and direct atmospheric Hg deposition.Since there are no coal-fired power plants, smelters, or waste incineration plants around Xincun Lagoon, atmospheric Hg point source pollution can seem to be ignored.In addition, there is no significant runoff inflow, so the sources of Hg deposited in the lagoon mainly come from catchment erosion and atmospheric Hg, including anthropogenic and natural sources.Xincun Lagoon is only connected to the South China Sea by a narrow mouth entrance (Fig. 1), as this entrance has not been altered much recently or even in the longer time, the capacity of the water exchange between the lagoon and the sea could not have a considerable effect on the change in Hg levels in the lagoon.Hg in the upper ocean is roughly in equilibrium, as most of the Hg deposited in the ocean is lost via gas escape, and only a small amount of Hg is buried in ocean sediments (Mason and Sheu, 2002;Selin et al., 2008).Therefore, Hg derived from the ocean (i.e., the South China Sea) may have negligible effects on Hg deposition in the XC-03 core.
The effects of climate change on Hg deposition have been demonstrated in various investigations, both on long-term and short-term timescales (Corella et al., 2017;Wang et al., 2021).On a high-resolution time scale, an increase in coarse debris might be related to heavy rainfall triggering intense erosion, which is strong enough to bring coarse debris to the deep water of the lake, and vice versa (Chen et al., 2004).Additionally, previous studies have revealed that increases in the ratios of Al/Ca and Ti/Ca can be used to identify a rise in terrestrial debris entering the lake caused by intense precipitation and vice versa under weak precipitation (Emmanouilidis et al., 2018;Song et al., 2022).As seen in Fig. 3, the elemental ratios and Mz values of the lagoonal sediments have changed significantly over the hundred years, revealing that the precipitation in Hainan Island has experienced more pronounced phase shifts recently.During the period from ~ 1780 to 1830 AD, the levels of Mz decreased slightly, and the values of Al/Ca and Ti/Ca were lower, reflecting a reduction in precipitation in Xincun Lagoon.The enormous swings in Mz and the ratios of each element between 1830 and 1970 AD suggested that the precipitation in the study region changed drastically.Both the Mz values and the ratios of each element have increased tremendously since 1970 AD, which is consistent with the pattern of precipitation instrumental data in Lingshui County during the last 50 yr.The trend of increasing Hg C and Hg AR after 1970 AD coincides with the trend of precipitation.However, Hg C and Hg AR values showed a gradually increasing tendency between 1830 and 1970 AD and displayed stability before ~ 1830 AD, while precipitation exhibited erratic fluctuations during this period.As a result, climate change should have a negligible impact on Hg deposition in the XC-03 core.
As shown in Figs.4a and 4b, the Hg C values showed moderate positive correlations with Al 2 O 3 content (R = 0.51, P < 0.05) and TiO 2 content (R = 0.52, P < 0.05) in the lagoon, Hg concentrations were normalized to lithophile elements Al and Ti as tracers of allochthonous detrital mineral material to further analyze the effect of terrestrial detritus on Hg loading (Boës et al., 2011;Daga et al., 2016).In Fig. 3, the trend of Hg C closely resembles that of Hg/Al and Hg/Ti normaliza-tions, suggesting the debris input from the catchment is an unworthy contribution to the sediment Hg levels (Figs.3h and 3i).Hg retained in sediments is usually strongly absorbed by OM and fine particles; therefore, OM content and sediment grain sizes, to a certain extent can affect the magnitude and timing of the release of Hg from catchments to lakes (Lucotte et al., 1995;Perry et al., 2005;Bookman et al., 2010).To investigate whether there is a direct connection between mercury accumulation and higher OM preservation, Hg concentration was normalized to TOC content.While the Hg C values were strongly correlated with TOC content (R = 0.66, P < 0.05) (Fig. 4c), Hg/TOC normalization revealed several prominent peaks, consistent with Hg C and Hg AR (Fig. 3).This implies that although organic matter has certain adsorption characteristics toward mercury, they are not the primary determinants of mercury deposition.In the XC-03 core, the clay content displayed a negative correlation with Hg C (R = −0.25,P < 0.05) (Fig. 4d), indicating that the absorption of clay is weak.In addition, volcanic eruption is one of the most significant natural sources of atmospheric Hg.However, no volcanic eruptions have been recorded on Hainan Island for more than two centuries, and no mark of Hg peaks ascribed to well-known global volcanic activities (Engstrom et al., 2014;Beal et al., 2015) can be observed in our core.Overall, our results suggest that natural factors have little influence on the accumulation of Hg in Xincun Lagoon.
As a consequence of increased usage and unintentional Hg emissions by anthropogenic activities over time, the prevailing opinion of Hg deposition from diverse profiles involves a 3−5-fold increase, and anthropogenic Hg emissions have produced the majority (73%) of global Hg emissions since preindustrial times (Biester et al., 2007;Yang et al., 2010;Streets et al., 2017).Notably, the Hg records in the XC-03 core agree well with the general trend of global Hg production (Hylander and Meili, 2003;Horowitz et al., 2014), implying that human activities must play a pivotal role in Hg cycling in Xincun Lagoon over the last 200 yr.

Anthropogenic influences on mercury deposition
The anthropogenic input of Hg in Xincun Lagoon sediments over the past 200 yr is depicted in Fig. 5, indicating that Hg pollution in the XC-03 core has well-known global distribution characteristics and regional differences.The global Hg emission inventory (Hylander and Meili, 2003) revealed that the North American gold and silver rush caused Hg to rise significantly during the 1850s to 1920s, peaking in the 1940s as a result of chemical production for ammunition during the World War II (Fig. 5).Global Hg production achieved its maximum in the 1970s, arising from dominant consumer products such as paint and batteries and in chlor-alkali plants, and then fell.The temporal trend of Hg levels in our present study was in accordance with the global Hg production during the industrial period until the 1970s.Hence, anthropogenic activities such as gold and silver mining, war, and industrial activities (e.g., fuel combustion, iron and steel production, non-ferrous metal smelting, cement production, etc.) are likely to be the critical drivers dominating the Hg accumulation in Xincun Lagoon.
In this study, anthropogenic Hg contamination in the XC-03 core was estimated using the Hg EF and Hg ARA .Here, the average Hg/Al ratio of samples from a depth of 64-53 cm ( ~ 1780-1830 AD) was chosen as the background value, owing to no indication of an obvious fluctuation in Hg and Al concentrations before 1830 AD, as documented in our study and other studies (Fitzgerald et al., 1998;Hermanns and Biester, 2013).Before 1830 AD, most of the Hg EF values were below 1.5 (Fig. 5), and the Hg ARA values remained relatively low and stable, both suggesting a weak human influence on Hg deposition in the lagoon.The Hg EF levels continued to rise rapidly after 1830 AD, reaching widespread high values in the 20th and early 21st centuries.
Hg ARA levels have generally manifested a significant upward trend in Xincun Lagoon since 1830 AD (Fig. 5a), especially in the early 20th century, with two high intervals during the periods around the 1910s and  (Hylander and Meili, 2003).The shadows represent the period of the wars 1950s.The more recent North American gold and silver rush signal (circa the 1850s to the early 1900s) does overlap with many published North American lake and ice core records (Pirrone et al., 1998;Beal et al., 2015), as well as the global Hg production inventory (Hylander and Meili, 2003).Despite the reality that some North American lakes showed no evidence of an apparent upward Hg deposition during that time, a high-resolution ice-core Hg record from the Belukha glacier in Central Asia and Hg deposition records in the ornithogenic sediments from the Yongle Islands of the South China Sea had the North American gold and silver rush signal, bolstering the theory that atmospheric Hg emissions of this period traveled throughout the world (Liu et al., 2012;Eyrikh et al., 2017).During the amalgamation process of extracting North American gold and silver, partial gaseous elemental mercury (Hg(0)) was emitted into the atmosphere (Pirrone et al., 1998), where it can be spread globally and deposited via long-range transport due to the long atmospheric residence time (0.5-1.0 yr) (Lindberg et al., 2007).In the early 20th century, along with the upsurge of China's 'Industrial salvation' ideological trend, the emergence of the modern investment industry boomed, and the investments were mainly in coastal areas (such as Guangdong Province).During this time, heavy government backing for iron and steel production that was melted, roasted, or sintered at high temperatures resulted in most Hg being volatilized into the environment (Sun et al., 2016b;Li, 2022), which may have caused regional mercury contamination.Additionally, the rise of Hg ARA in Xincun Lagoon from the late 1800s to the early 1900s coincides well with that of polycyclic aromatic hydrocarbons (PAHs) in the Pearl River Delta (PRD), which are considered to be unique indicators of human activities in the PRD (Liu et al., 2005).Therefore, an uptick in Hg ARA from 1850 to 1914 AD may be related to the North American gold and silver rush and the 'Industrial salvation' trend.
During wartime, the production and employment of explosives and weapons resulted in massive amounts of Hg being released into the atmosphere, which was then transported and finally deposited all over the world (Horowitz et al., 2014;Sun et al., 2016a;Corella et al., 2017).Our study region was an excellent indicator for some notable regional and local wars because of its proximity to the source area of the wars (Fig. 5).The Hg ARA values for the period 1835 to 1860 AD were marginally higher than those for the background period, which is likely due to the well-known conflicts along the southeast coast of China, i.e., the Opium War I (1839-1842 AD) and Opium War II (1856-1860 AD).Additionally, large amounts of international and domestic wars occurred from 1914 to 1952 AD, accompanied by the second intervals of high Hg ARA in Xincun Lagoon, which may reflect World Wars I (1914I ( -1918 AD) AD) and II (1940-1945 AD), the War of Resistance against Japanese Aggression (1931( -1945 AD) AD), and the Chinese Liberation War (1946( -1950 AD) AD).Several national and international investigations have documented markedly elevated Hg levels in the 1940s, which were synchronous with the Hg production peak of World War II (Schuster et al., 2002;Horowitz et al., 2014;Liu et al., 2015;Corella et al., 2017), and even some records from surrounding areas (such as coral skeletons in the South China Sea, etc.) have identified war as one of the primary determinants of peaks in Hg levels (Sun et al., 2016a).Between the 1910s and 1950s, prolonged anthropogenic Hg poisoning in the XC-03 core probably resulted from devastating and frequent worldwide wars.Notably, it is possible that the outbreak of the China-Vietnam War in 1979 AD, whose major battlefield was the China-Vietnam border, led to an upsurge in anthropogenic mercury pollution in the early 1980s (Fig. 5).There is evidence that Hg fulminate was extensively employed for igniting munitions, which were historically discarded and dumped at sea (Gosnell et al., 2022), likely in close association with Hg contamination in Xincun Lagoon.As in wars, a large number of weapons and ammunition will be detonated.After detonation, Hg in the explosives would be volatilized and may be combined with emitted particles as particle-bound Hg, discharging into the atmosphere.From the 1950s to 1970s, the Hg ARA values were relatively low to a certain extent due to the economic depression.
The inconsistency trend in Hg ARA with the global Hg inventory after ~ 1975 AD may mostly be the pattern of regional Hg pollution in Xincun Lagoon.Dramatic increases in Hg ARA from the 1970s to 1990s and the maximum peak of the Hg EF occurring in the were possibly the result of increased anthropogenic Hg emissions from intense industrial activities in China and other Asian countries, notably Eastern Asia.According to Pacyna et al. (2006), Hg emissions from Asia accounted for 54% of world anthropogenic atmospheric Hg in 2000, and of all countries, China had the highest anthropogenic Hg emissions, accounting for 28% in 2000.After 1978, the reform and opening-up policy of China was carried out, promoting China to enter a period of rapid industrialization and economic growth, with China's GDP growing at an annual rate of ~ 9.8% from 1978 to 2012 (Tian et al., 2015).Coal combustion as an energy source is the major source of Hg emissions in China, claiming more than 25% of the total worldwide coal production, with high emissions occurring in eastern China (Fu et al., 2012;Huang et al., 2017).It was estimated that at least 2400 t of Hg entered the atmosphere through coal combustion from 1978 to 1995, with Hg emissions at an average rate of 4.8%/yr (Wang et al., 2000).Our study site was close to Guangdong Province, which has made many contributions to atmospheric Hg through coal combustion and nonferrous metal production (Streets et al., 2005;Huang et al., 2017).In fact, since the late 19th century, parallel trends of Hg ARA in the XC-03 core and PAHs (Polycyclic Aromatic Hydrocarbons) in the Pearl River Delta, one of the most densely populated and economically dynamic regions in China (Liu et al., 2005), point to incomplete fossil fuel as an important Hg input to the lagoon.Recently, an evident spike in historical Hg records in response to industrial development after the 1970s was observed not only in our present study but also in other lake, ice core, and marine records (Yang et al., 2010;Duan et al., 2015;Kang et al., 2016).
However, the steep decline in Hg ARA in the late 1990s and early 2000s, corresponding with Hg records from a peat core on the edge of the Tibetan Plateau and the inner shelf of the East China Sea (Shi et al., 2011;Duan et al., 2015), may be explained by the implementation of environmental protection measures and the efficiency improvement of emission reduction equipment in China.For instance, in the mid-late 1990s, regulations to prohibit Artisanal gold smelting and control caustic soda production using a Hg cell electrolysis process were introduced in China (Feng et al., 2009;Zhang et al., 2015).The Hg yield in China in 1999 was down to 195 t, approximately 62% less than that in 1996 (Streets et al., 2005).From the late 2000s, the strikingly enhanced Hg depositional signals may be related to the expedited development of coal combustion, nonferrous metal smelters, and cement production, triggered by the sustained and rapid development of industry and the acceleration of urban construction.Combining the development of industrial activities in Southeast Asia and China, the HYSPLIT backward trajectory simulation confirmed the trail of long-range transported atmospheric Hg pollution from Southeast Asia and China to Xincun Lagoon (Fig. 6).As revealed by the model simulations, in Xincun Lagoon, the summer monsoon delivered air masses mainly from the southwest, with about 65% of air masses coming from Southeast Asia (Fig. 6a).In contrast, air masses over the lagoon in the winter arrived from the northeast, with about 80% of air masses coming from China (Fig. 6b).In summary, the huge spike in Hg after ~ 1970 AD was also likely to have absorbed partial amounts of Hg from industrial activities in Southeast Asia in summer and significant

Conclusions
The historical change in Hg deposition in Xincun Lagoon in South China has been revealed over the past 200 yr.Our results suggested that climatic environment, catchment erosion, organic matter, and clay were not controlling factors for Hg accumulations in Xincun Lagoon over the last 200 yr, while the increasing anthropogenic activities have heavily influenced the Hg pollution in the study region.Since ~ 1830 AD, the Hg C , Hg AR , and Hg ARA values in Xincun Lagoon have been on the rise, which was temporally consistent with the global Hg production until ~ 1975 AD, reflecting the growing influence of human activities on Hg deposition.
From the 1830s to 1850s and from the 1910s to 1950s, high Hg pollution was dominated by domestic and foreign wars.Moreover, the Hg records in the study were increasing steeply after ~ 1975 AD, attributed to the China-Vietnam War and industrial activities in China and Southeast Asian countries.This is reinforced by the result of the HYSPLIT backward trajectory simulation, which revealed that atmospheric Hg comes mainly from Southeast Asia in the summer and from China in the winter.Thus, Hg records in our study present a pattern of elevated mercury contamination before ~ 1975 AD that was largely driven by regional and global wars, with a further rise due to regional industrial development and urban construction after ~ 1975 AD.

Fig. 1
Fig. 1 The location of sample site in Xincun Lagoon in Lingshui County, southeast Hainan Island, China and schematic of the Asian monsoon system Bq/kg) Activity / (Bq/kg)

Fig. 3
Fig.3Variations of multi-proxies in the XC-03 core in Xincun Lagoon in Lingshui County, southeast Hainan Island, China.a. Hg concentration (Hg C ); b.Hg accumulation rate (Hg AR ); c.Al 2 O 3 content; d.TiO 2 content; e. CaO content; f.Al/Ca; g.Ti/Ca; h.Hg/Al; i. Hg/Ti; j. total organic carbon; k.Hg/TOC; l. dry bulk density (DD); m. clay content; n. mean grain size (Mz).The trends of Al/Ca and Ti/Ca were filtered using a 5-point moving average (grey lines).The dotted lines represent the averages of Al/Ca and Ti/Ca, respectively

Fig. 4
Fig. 4 Correlations of Hg C with multi-proxies in the XC-03 core in Xincun Lagoon in Lingshui County, southeast Hainan Island, China.a. Hg C with Al 2 O 3 content; b.Hg C with TiO 2 content; c.Hg C with total organic carbon (TOC) content; d.Hg C with clay content

Fig. 5
Fig.5Variations of proxies in the XC-03 core and global mercury production.a. Anthropogenic Hg accumulation rates (Hg ARA ; this study); b.Hg enrichment factors (Hg EF ; this study); c. global mercury production from 1778 to 1997(Hylander and Meili, 2003).The shadows represent the period of the wars

Fig. 6
Fig. 6 Air mass backward trajectories in Xincun Lagoon in Lingshui County, southeast Hainan Island, China.a. Summer; b.Winter.The percentage values in the figure refer to the contribution of the different air mass backward trajectories.More details about the simulations can be found in section 2.6 The total 210 Pb ( 210 Pb tot ), excess 210 Pb ( 210 Pb ex ), and 226 Ra specific activity vertical distributions of the XC-03 core are shown in Fig. 2. In general, the specific activities of 210 Pb tot and 210 Pb ex decreased with increasing depth, and the activity of 226 Ra also fluctuated.Since 210 Pb tot , 210 Pb ex , and 226 Ra were balanced at 35 cm depth (Figs.2a and 2b), the chronology was constructed using the constant rate of supply (CRS) model for the upper 35 cm, ranging from ~ 1909 to 2015 AD.The age at 20 cm from the top yielded by the CRS model was consistent with the AMS 14 C date (measured data: 101.51 ± 0.38 pMC; Lab ID: Beta-529986) at this depth with a calibrated 14 C age of 1954−1956 AD, confirming the accuracy and reliability of 210 Pb dating.