Anthropogenic Impact and Ecological Risk Assessment of Thallium and Cobalt in Poyang Lake Using the Geochemical Baseline

: As the largest freshwater lake in China, water environmental pollution in Poyang Lake is increasingly attracting attention. However, to date, the occurrence of seldom-monitored trace elements (SMTEs) (such as thallium (Tl) and cobalt (Co)) and their potential ecological inﬂuences in the Poyang Lake area remain unclear. Here, geochemical baseline concentrations (GBCs) were employed to investigate anthropogenic inputs of these two metals (Tl and Co) and were used to assess the potential ecological risk in surface sediments of Poyang Lake. The results showed that the mean concentrations of Tl and Co were 1.11 mg/kg and 15.42 mg/kg, respectively, slightly higher than background values (BVs) in local soil. Spatially, the Co concentrations in surface sediments of Poyang Lake varied; Co concentrations in the west of the study area were signiﬁcantly lower than in the middle of the lake. In contrast, no signiﬁcant spatial variation of Tl was found in studied areas of Poyang Lake. Average GBCs were similar to the mean measured concentrations of Tl and Co in the lake, and both were higher than the BVs in local soil. In anthropogenic source analysis based on the GBCs, 47.22% samples and 41.67% samples were inﬂuenced by human activity for Tl and Co, respectively; the mean anthropogenic contribution rate was 5.90% for Tl (0.43 to 16.22%), and 13.01% for Co (0.82 to 31.59%). These results indicated that Tl and Co mainly originate from natural processes. However, the anthropogenic contribution should not be ignored. Moreover, the geo-accumulation index ( I geo ) and potential ecological parameter ( E r ) generally indicated no contamination and a low potential risk for Tl and Co in Poyang Lake sediments, which used the GBCs as BVs. The results of this study suggest that GBCs are more scientiﬁcally robust and practical as a reference than BVs. They can also be used when BVs are lacking.


Introduction
Due to their persistence, inherent toxicity, and non-degradability, the contamination of the aquatic environment with trace metals is highly concerning [1,2]. While previous studies have tended to focus on the commonly monitored trace elements (e.g., Cu, Hg, Pb, As, Cd, Cr, Mn, and Zn) [3,4]. Comparatively, 'seldom-monitored trace elements (SMTEs)', such as thallium (Tl) and cobalt (Co), have caused less concern, especially for sediments. However, these two metals also have potential health effects through accumulation in the food chain [5,6]. Tl and Co are dispersed metal elements that seldom occur in the form of independent minerals; they are commonly found in sulfur mines, the GBCs of Tl and Co in the surface sediments of Poyang Lake; (3) to analyze anthropogenic contributions based on GBCs; and (4) to evaluate the contamination state and potential risks of Tl and Co in the study area using the I geo , PI, and E r . The results of this study provide an effective reference of geochemical patterns for Tl and Co in the surface sediments of Poyang Lake, and give new insights that will help to protect water quality.

Study Area
Poyang Lake (28.400 • -29.767 • N, 115.817 • -116.767 • E), with a drainage area of 162,000 km 2 , represents 97% of Jiangxi province [33]. Figure 1 shows the study area, with the location of sampling sites, choosing WG84 as the geographic coordinate system. The area experiences a subtropical humid monsoon climate, and annual mean temperatures are from 16.5 to 17.8 • C. The rainy season is from April to September, and mean annual precipitation is 1570 mm.
Water 2018, 10, x FOR PEER REVIEW 3 of 14 establish the GBCs of Tl and Co in the surface sediments of Poyang Lake; (3) to analyze anthropogenic contributions based on GBCs; and (4) to evaluate the contamination state and potential risks of Tl and Co in the study area using the Igeo, PI, and Er. The results of this study provide an effective reference of geochemical patterns for Tl and Co in the surface sediments of Poyang Lake, and give new insights that will help to protect water quality.

Study Area
Poyang Lake (28.400°-29.767° N, 115.817°-116.767° E), with a drainage area of 162,000 km 2 , represents 97% of Jiangxi province [33]. Figure 1 shows the study area, with the location of sampling sites, choosing WG84 as the geographic coordinate system. The area experiences a subtropical humid monsoon climate, and annual mean temperatures are from 16.5 to 17.8 °C. The rainy season is from April to September, and mean annual precipitation is 1570 mm.

Sample Collection
A total of 36 surface sediment samples were collected from Poyang Lake in July 2016; sampling sites were arranged from the western side of the lake to the center of the lake (Figure 1). The sampling area extended from 116.050 • to 116.416 • E, with the dividing line between the western and central lake at 116.183 • E. Sampling sites were numbered W01 to W20 (western lake) and M01 to M16 (central lake). Surface sediment sampling (at depths of between 0 and 10 cm) was carried out using a bottom sampler at each location. Samples were sealed in clean polyethylene bags, which were soaked in nitric acid for two days and rinsed with deionized water in advance, then samples were stored at low temperature and transported to the laboratory immediately. All collected sediment samples were freeze-dried, crushed, and ground in an agate mortar. Stones and plant fragments were removed by passing the samples through an acid-cleaned 65-mesh nylon sieve for chemical analysis.

Chemical Analysis and Quality Control
Tl and Co concentrations in sediments were measured using an established method [34]. In brief, 40 mg of each dry sediment sample was weighed; this was then digested using a strong acid solution (HNO 3 + H 2 O 2 + HF), and finally diluted with ultrapure water to 80 g. Total Tl and Co concentrations were measured by an inductively coupled plasma mass spectrometer (ICP-MS; PerkinElmer Elan DRC-e, USA) with a detection limit of 0.05 mg/kg. All chemical treatments were performed in an ultraclean laboratory, and adequate quality control measures were followed in all aspects of the study. Quality control was assured by blanks, standard reference materials and duplicates; Chinese stream sediments (GSD-10, GSD-1a) were employed to ensure the accuracy of the analytical procedures. The recovery of standard reference materials was 107-114% for Tl, and 100-105% for Co.

Geochemical Baseline Concentrations (GBCs)
The concept of GBCs was first put forward in the 1970s. Generally, GBCs are representative of element concentrations that to some extent reflect human activities; they differ from BVs in sediments or soil [35]. In this study, the GBCs for Tl and Co in the surface sediments of Poyang Lake were established using a normalization method [32]. This method uses correlation analysis between a conservative reference element (inert element) and the study element. In this study, lithium (Li) was recognized as the reference element owing to its stability and the significant liner correlation with Tl and Co. The GBCs of Tl and Co in the surface sediments of Poyang Lake were established using the following equation: where C m is the concentration of Tl or Co (mg/kg), C N is the concentration of Li (mg/kg), and a and b are regression constants. Data outside the 95% confidence limit were removed through regression analysis, indicating that they were influenced by anthropogenic pollution. Data within the 95% limit were employed to obtain new regression constants (c and d). Finally, the GBCs of Tl and Co were obtained from the new regression line, and the GBCs were averaged as a reference to identify surface sediment contamination in Poyang Lake. The GBCs were calculated as follows: where B m is the GBC of Tl or Co (mg/kg) in the study area, C N is the normalized average of Li (mg/kg), and c and d are the new regression constants.

Geo-Accumulation Index (I geo )
The geo-accumulation index is a quantitative evaluation method for sediments (especially sediments in water environments) [36]; it is widely employed to quantitatively assess the degree of contamination [37,38]. The equation for I geo is defined as follows: where C n is the measured concentration of an element in the sediment (mg/kg), B n represents the background concentration (mg/kg), and 1.5 corresponds to possible variation of in crustal contribution, mainly by lithogenic effects. A positive I geo generally indicates an anthropogenic contribution [29]. Seven classes of I geo were classified as follows: uncontaminated (I geo < 0); uncontaminated to moderately contaminated (0 ≤ I geo < 1), moderately contaminated (1 ≤ I geo < 2), moderately to strongly contaminated (2 ≤ I geo <3), strongly contaminated (3 ≤ I geo < 4), strongly to extremely contaminated (4 ≤ I geo <5), and extremely contaminated (I geo ≥ 5).

Potential Ecological Risk Parameter (E r )
Single pollution indices (PI) are useful tools for determining contamination levels [27]. In this study, we used the potential ecological parameter (E r ), which combines the concentrations of metals with their ecological and toxicological effects [39], to assess the potential ecological toxicity of Tl and Co in the surface sediments of Poyang Lake. RI can be calculated as the sum of the individual potential risk factors (E r ), which were given by the following equations: where PI denotes the contamination level for a certain metal, C x is the measured concentration of the metal (Tl and Co), C 0 indicates the BV of metals in the sediments. E r demonstrates the potential ecological risk of the element. T r is the toxic response factor for the study element. In this study, the T r of Tl and Co were defined as 10 and 5, respectively, based on previous studies [22,40].

Statistical Analysis
In this study, standard statistical analyses (mean, median, standard deviation) were carried out to describe the measured concentrations and GBCs of Tl and Co. The ArcGIS 10.1 software package was used for spatial mapping. SPSS 17.0, Origin Pro 2017, and Microsoft software packages were combined to process analytical data.

Concentrations and Spatial Variability of Tl and Co in the Surface Sediments
The concentrations of Tl and Co in the surface sediments of Poyang Lake are shown in Table 1 and Figure 2. The mean concentrations of Tl and Co in the sediments were 1.11 ± 0.15 mg/kg and 15.42 ± 4.09 mg/kg, respectively. The mean concentration of Tl was more than twice the BVs in the Yangtze River sediments [41]. Meanwhile, the mean Tl concentrations already exceeded the BVs in the soil of Nanchang (0.80 mg/kg) [42] and Jiangxi Province (0.88 mg/kg) [43]. The mean concentration of Co was higher than the BVs in Chinese sediments [41], but slightly lower than the BVs in Yangtze River sediments [41]. Meanwhile, mean Co concentrations also exceeded BVs in the soil of Nanchang (13.00 mg/kg) [42] and Jiangxi Province (11.50 mg/kg) [43]. In general, about 97.22% of samples for Tl and 80.56% of samples for Co exceeded the BVs in Nanchang soil, indicating the accumulation of these two metals in the sediments. Compared with previous studies [3,4], the concentrations of Tl and Co in Poyang Lake were found to have increased slightly ( Table 2), suggesting that they were influenced by anthropogenic activities in the Poyang Lake basin. This is consistent with previous studies showing how anthropogenic activities [31], such as hydro-engineering works [44], shipping, mining, and agriculture, affect metal concentrations in Poyang Lake.

Establishment of Geochemical Baselines for Tl and Co
The results of correlation analysis revealed that the concentration of Tl and Co had significant positive relationships with the concentration of Li (p < 0.0001); the correlation coefficient for the linear regression was 0.7318 for Tl and 0.7582 for Co (Figure 3a,b). We removed all points that were beyond the 95% confidence band. Based on the remaining data, geochemical baselines were developed (Figure 3c,d) and the correlation coefficients of Tl and Co increased to 0.9781 and 0.9815, respectively.
As a result, the GBC ranges of Tl and Co in the surface sediments of Poyang Lake were 0.83-1.26 mg/kg and 6.95-19.73 mg/kg, respectively, and the mean GBC values were 1.12 ± 0.10 mg/kg and 15.44 ± 3.08 mg/kg, respectively. It was found that the GBC of Tl in the surface sediments of  To gain a comprehensive understanding of the Tl and Co statuses of the surface sediments in this study, a comparison of the Tl and Co concentrations with those in other regions of China was performed ( Table 2). The mean concentration of Tl in this study was higher than those reported at Chao Lake, Three Gorges Reservoir, and Laizhou Bay, but lower than those in the Xiang Jiang River, Hengyang section, and Beijiang River [19,[45][46][47][48]. Meanwhile, the mean concentration of Co in our study was also higher than those of the Beijiang River and Chao Lake, but was significantly lower than those in the Xiang Jiang River, Hengyang section, and Three Gorges Reservoir [19,[45][46][47]. These differences in concentrations of Tl and Co may not reflect actual pollution levels, owing to existing spatial heterogeneity [29]. Therefore, the reference used for metals must be considered when assessing metal contamination in sediments.
Spatially, the distribution of Tl concentrations was different from that of Co (Figure 2a,b). The concentrations of Tl in the surface sediments were not significantly different between the western and middle areas of Poyang Lake. Tl concentrations were highest at M07 and lowest at M16. While, larger variations of Co concentrations were observed. The Co concentrations in the surface sediments of the west lake were slightly lower compared with those in the middle of the lake; Co concentrations were highest at M02 and lowest at W19. The results indicate that Tl and Co have different spatial patterns, perhaps demonstrating that they were influenced by varying exogenous inputs.

Establishment of Geochemical Baselines for Tl and Co
The results of correlation analysis revealed that the concentration of Tl and Co had significant positive relationships with the concentration of Li (p < 0.0001); the correlation coefficient for the linear regression was 0.7318 for Tl and 0.7582 for Co (Figure 3a,b). We removed all points that were beyond the 95% confidence band. Based on the remaining data, geochemical baselines were developed (Figure 3c,d) and the correlation coefficients of Tl and Co increased to 0.9781 and 0.9815, respectively.
As a result, the GBC ranges of Tl and Co in the surface sediments of Poyang Lake were 0.83-1.26 mg/kg and 6.95-19.73 mg/kg, respectively, and the mean GBC values were 1.12 ± 0.10 mg/kg and 15.44 ± 3.08 mg/kg, respectively. It was found that the GBC of Tl in the surface sediments of Poyang Lake was markedly higher than the BVs in the soil of Nanchang (0.80 mg/kg) [42], Jiang Xi Province (0.88 mg/kg) [43], and Yangtze River sediments (0.49 mg/kg) [41]. The GBC of Co also exceeded the BVs in the soil of Nanchang (13.00 mg/kg) [40] and Jiang Xi Province (11.50 mg/kg) [43], but was lower than the BV in the Yangtze River sediments (17.00 mg/kg) [41]. The mean GBCs very roughly approximated the mean concentrations of Tl and Co in Poyang Lake; therefore, we believe that the established GBCs are reasonable [29] and could be employed for the further pollution assessment of Tl and Co in Poyang Lake sediments.  To better understand the impact of anthropogenic sources, anthropogenic contribution rates were calculated based on the GBCs. Using the established linear regression equation, baseline concentration values of Tl and Co were determined for each sampling site in order to determine the differences between measured concentrations and baseline concentrations in the surface sediments of Poyang Lake. The GBCs of Tl and Co reflect the natural concentration status of the study area. If a measured concentration is higher than the baseline concentration, the metal has been influenced by anthropogenic inputs. Conversely, if the measured concentration is lower than the baseline concentration, variability is from a natural source. Thereby, the anthropogenic contribution of each sampling site can be obtained using the equations: Contribution rate (%) = (Cm − Pm)/Cm × 100% (8) where Pm is the GBC at each sampling site, CN represents the measured concentration of Li in the sediments, c and d are constants in Equation (7), and Cm is the measured concentration of Tl or Co in the surface sediments. Our calculations showed that 47.22% of the differences between the measured concentrations and GBCs were positive for Tl (Figure 4). The anthropogenic contributions of Tl varied between 0.43% and 16.22%, with a mean value of 5.90%. The maximum value was 16.22%, which was observed at site M10 (Figure 4). For Co, 41.67% of samples were impacted by anthropogenic sources to varying degrees (Figure 4). While the anthropogenic contribution of Co ranged from 0.82 to 31.59%, the mean value was 13.01%; the maximum was observed at site W09 ( Figure 4). As shown in Figure 2 and Figure 4, the maximum anthropogenic contributions were not found at the highest  To better understand the impact of anthropogenic sources, anthropogenic contribution rates were calculated based on the GBCs. Using the established linear regression equation, baseline concentration values of Tl and Co were determined for each sampling site in order to determine the differences between measured concentrations and baseline concentrations in the surface sediments of Poyang Lake. The GBCs of Tl and Co reflect the natural concentration status of the study area. If a measured concentration is higher than the baseline concentration, the metal has been influenced by anthropogenic inputs. Conversely, if the measured concentration is lower than the baseline concentration, variability is from a natural source. Thereby, the anthropogenic contribution of each sampling site can be obtained using the equations: Contribution rate (%) = (C m − P m )/C m × 100% where P m is the GBC at each sampling site, C N represents the measured concentration of Li in the sediments, c and d are constants in Equation (7), and C m is the measured concentration of Tl or Co in the surface sediments. Our calculations showed that 47.22% of the differences between the measured concentrations and GBCs were positive for Tl (Figure 4). The anthropogenic contributions of Tl varied between 0.43% and 16.22%, with a mean value of 5.90%. The maximum value was 16.22%, which was observed at site M10 ( Figure 4). For Co, 41.67% of samples were impacted by anthropogenic sources to varying degrees ( Figure 4). While the anthropogenic contribution of Co ranged from 0.82 to 31.59%, the mean value was 13.01%; the maximum was observed at site W09 ( Figure 4). As shown in Figures 2 and 4, the maximum anthropogenic contributions were not found at the highest concentrations of Tl and Co. This confirms that using GBCs to distinguish anthropogenic input and natural sources has practical significance.
Water 2018, 10, x FOR PEER REVIEW 9 of 14 concentrations of Tl and Co. This confirms that using GBCs to distinguish anthropogenic input and natural sources has practical significance. From the above, the anthropogenic influences for both Tl and Co were relatively minor, but differed from one another. The anthropogenic contribution differed depending on sample location. Generally, the proportion of sampling sites influenced by anthropogenic sources was fewer than 50% for Tl and Co. The results show that the increase of Tl and Co in the surface sediments of Poyang Lake can mainly be attributed to natural factors; however, the anthropogenic influences cannot be neglected.

Assessment of Potential Ecological Risk by Tl and Co
In the assessment of potential ecological risk, we employed the mean GBCs of Tl and Co instead of BVs for several reasons: (1) GBCs consider the natural variations of a metal, but can also distinguish anthropogenic contributions from exogenous inputs; (2) the BVs of Tl and Co in the sediments of Poyang Lake were unavailable; (3) assessments of lake sediment state with BVs of the soil or stream sediments have been shown to be imprecise. Three different quantitative indices (Igeo, PI, and Er) were used to evaluate the potential ecological risk of Tl and Co in Poyang Lake sediments. Table 3 summarizes the calculated Igeo values for Tl and Co. Based on the GBCs, negative Igeo values were observed for Tl and Co, except for a very few sampling sites. The results revealed that these two metals did not cause pollution in the surface sediments of Poyang Lake. To be specific, all Igeo values for Tl were less than zero, ranging from −1.20 to −0.25, and with a mean of −0.61. Therefore, the surface sediments of Poyang Lake were uncontaminated by Tl. For Co, almost all Igeo values were also less than zero, except for that at M02; the Igeo values of Co varied between −2.04 and 0.02, with a mean of −0.64. In general, this also suggests that the sediments of Poyang Lake were uncontaminated by Co, although site M02 was found to be uncontaminated to moderately contaminated. The large spatial variation of Co suggested that its distributions could be influenced by anthropogenic activities. From the above, the anthropogenic influences for both Tl and Co were relatively minor, but differed from one another. The anthropogenic contribution differed depending on sample location. Generally, the proportion of sampling sites influenced by anthropogenic sources was fewer than 50% for Tl and Co. The results show that the increase of Tl and Co in the surface sediments of Poyang Lake can mainly be attributed to natural factors; however, the anthropogenic influences cannot be neglected.

Assessment of Potential Ecological Risk by Tl and Co
In the assessment of potential ecological risk, we employed the mean GBCs of Tl and Co instead of BVs for several reasons: (1) GBCs consider the natural variations of a metal, but can also distinguish anthropogenic contributions from exogenous inputs; (2) the BVs of Tl and Co in the sediments of Poyang Lake were unavailable; (3) assessments of lake sediment state with BVs of the soil or stream sediments have been shown to be imprecise. Three different quantitative indices (I geo , PI, and E r ) were used to evaluate the potential ecological risk of Tl and Co in Poyang Lake sediments. Table 3 summarizes the calculated I geo values for Tl and Co. Based on the GBCs, negative I geo values were observed for Tl and Co, except for a very few sampling sites. The results revealed that these two metals did not cause pollution in the surface sediments of Poyang Lake. To be specific, all I geo values for Tl were less than zero, ranging from −1.20 to −0.25, and with a mean of −0.61. Therefore, the surface sediments of Poyang Lake were uncontaminated by Tl. For Co, almost all I geo values were also less than zero, except for that at M02; the I geo values of Co varied between −2.04 and 0.02, with a mean of −0.64. In general, this also suggests that the sediments of Poyang Lake were uncontaminated by Co, although site M02 was found to be uncontaminated to moderately contaminated. The large spatial variation of Co suggested that its distributions could be influenced by anthropogenic activities.

Potential Ecological Risk
The PI values for Tl and Co are shown in Table 3. For Tl, PI varied between 0.65 and 1.26, with a mean value of 0.99; the results suggest that 47.22% of sampling sites had minor Tl enrichment, and 52.78% had no contamination. The PI of Co varied between 0.36 and 1.52, with a mean value of 1.00. The results suggest that the study area has minor Co pollution. The distribution of PI values for Tl and Co is presented in Figure 5; contamination levels of Tl and Co were marginally different. The PI of Co in the surface sediments of Poyang Lake showed greater fluctuation, indicating that Co is more susceptible to anthropogenic discharge. The potential ecological risk of metals is not only related to PI, but is also dependent on the toxic response factor (Tr). Considering the different toxicities of the contaminants to organisms, Er and RI were adopted to comprehensively assess the potential ecological risk resulting from Tl and Co. According to the Häkanson ecological risk assessment index, all sampling sites of Tl and Co exhibited a significantly low potential ecological risk. The results of Er ranged from 6.54 to 12.59 for Tl, with a mean value of 9.89; the maximum value was recorded at site M07. For Co, Er varied between 1.82 and 7.60, with a mean value was 4.99; the maximum value was recorded at site M02. All Er values for Tl and Co were below 40.00, and RI values were far less than 150. In consequence, all surface sediment sampling sites had a low potential ecological risk from Tl and Co.

Conclusions
This study presents the first analysis of Tl and Co in the surface sediments of Poyang Lake based on GBCs. The contents of Co were apparently higher than Tl; mean concentrations were 1.11 mg/kg for Tl and 15.42 mg/kg for Co. Spatially, Tl and Co showed different distribution patterns. For Tl, there was weak spatial variation across the study area. In contrast, Co concentrations in the surface sediments of Poyang Lake were significantly higher in the western lake than in the middle of the lake.
The GBCs of Tl and Co in Poyang Lake were 0.83-1.26 mg/kg and 6.95-19.73 mg/kg, respectively. Mean GBCs were similar to mean measured concentrations of Tl and Co in the lake, and both were higher than the BVs in local soils. For anthropogenic source analysis, we employed the GBCs; the mean contribution rate was 5.90% for Tl (ranging from 0.43 to 16.22%), and the mean contribution rate for Co was 13.01% (ranging from 0.82 to 31.59%). The results suggest that the Tl and Co in Poyang Lake mainly originate from natural sources, with only minor anthropogenic inputs.
Based on the GBCs established for Tl and Co in the surface sediments of Poyang Lake, the Igeo, PI, and Er were employed to estimate potential ecological risk. The contamination levels of Tl and Co in the surface sediments were found to be not polluted based on the Igeo, except for one sampling site. For PI, 47.22% of the sampling sites showed slight enrichment in Tl, and the whole study area showed slight enrichment in Co (mean PI value of 1.00). Based on Er, the potential ecological risk of Tl and Co was low. The use of GBCs was found to be a practical method for accurately assessing metal contamination in surface sediments and for creating rational strategies in water environment protection.   The potential ecological risk of metals is not only related to PI, but is also dependent on the toxic response factor (T r ). Considering the different toxicities of the contaminants to organisms, E r and RI were adopted to comprehensively assess the potential ecological risk resulting from Tl and Co. According to the Häkanson ecological risk assessment index, all sampling sites of Tl and Co exhibited a significantly low potential ecological risk. The results of E r ranged from 6.54 to 12.59 for Tl, with a mean value of 9.89; the maximum value was recorded at site M07. For Co, E r varied between 1.82 and 7.60, with a mean value was 4.99; the maximum value was recorded at site M02. All E r values for Tl and Co were below 40.00, and RI values were far less than 150. In consequence, all surface sediment sampling sites had a low potential ecological risk from Tl and Co.

Conclusions
This study presents the first analysis of Tl and Co in the surface sediments of Poyang Lake based on GBCs. The contents of Co were apparently higher than Tl; mean concentrations were 1.11 mg/kg for Tl and 15.42 mg/kg for Co. Spatially, Tl and Co showed different distribution patterns. For Tl, there was weak spatial variation across the study area. In contrast, Co concentrations in the surface sediments of Poyang Lake were significantly higher in the western lake than in the middle of the lake.
The GBCs of Tl and Co in Poyang Lake were 0.83-1.26 mg/kg and 6.95-19.73 mg/kg, respectively. Mean GBCs were similar to mean measured concentrations of Tl and Co in the lake, and both were higher than the BVs in local soils. For anthropogenic source analysis, we employed the GBCs; the mean contribution rate was 5.90% for Tl (ranging from 0.43 to 16.22%), and the mean contribution rate for Co was 13.01% (ranging from 0.82 to 31.59%). The results suggest that the Tl and Co in Poyang Lake mainly originate from natural sources, with only minor anthropogenic inputs.
Based on the GBCs established for Tl and Co in the surface sediments of Poyang Lake, the I geo , PI, and E r were employed to estimate potential ecological risk. The contamination levels of Tl and Co in the surface sediments were found to be not polluted based on the I geo , except for one sampling site. For PI, 47.22% of the sampling sites showed slight enrichment in Tl, and the whole study area showed slight enrichment in Co (mean PI value of 1.00). Based on E r , the potential ecological risk of Tl and Co was low. The use of GBCs was found to be a practical method for accurately assessing metal contamination in surface sediments and for creating rational strategies in water environment protection.