Residual Level, Prole and Potential Ecological Risk of Organochlorine Pesticides in Subsurface Sediments from Lakeshore Backll Site and Lakebed of an Abandoned Oxidation Pond After Desilting for 15 Years

The occurrence and distribution of 19 organochlorine pesticides (OCPs), together with microbial ester-linked fatty acid methyl ester (EL-FAME) proles were investigated in sediments from an abandoned oxidation pond of Ya-Er lake, China, which had been heavily polluted by hexachlorocyclohexanes (HCHs) and chlorobenzenes in 1980s. Subsurface sediment samples were taken from ve sediment cores along the transect running from lakeshore to lakebed. The concentration of total OCPs ranged from 29.8 to 941.8 ng g -1 dw, with a mean value of 193.3 ng g -1 dw. Hexachlorobenzene (HCB), HCHs and dichlorodiphenyl-trichloroethanes (DDTs) were the three dominant OCP classes, accounting for 26.5%-97.4%, 1.8%-51.5%, and 0.4%-15.5% of the total, respectively. Hot spots of HCB, HCHs and DDTs were detected in 0.9-2.7 m deep layers of the lakeshore, where was one of the dredged sediment backll sites for in-situ remediation of the oxidation pond in 2002-2004. High potential risks of HCHs and HCB were still indicated. Historical industrial input (27.2%), recent agricultural input (14.7%), and persistent residuals (14.3%) were the three major identied sources of OCPs, using Absolute Principal Component Scores-Multiple Linear Regression (APCS-MLR). Redundancy analysis of microbial EL-FAME proles and nine dominant OCPs revealed that the spatial variation in microbial community structure was signicantly corresponded with the OCP composition. This is the rst study highlighted the concern on historical industrial inputs of OCPs in subsurface sediments of the lakeshore disposal zone. The ndings could help to distinguish the articial backll sediments from natural polluted sediments for optimization of further desilting plans.


Introduction
Organochlorine pesticides (OCPs) are synthetic chlorinated compounds that contribute to agriculture production worldwide (Jayaraj et al. 2016). They can enter the aquatic environment through industrial and domestic discharges, agricultural runoff, and atmospheric deposition after long-distance transfer (Bhattacharya et al. 2003; Qu et al. 2019). Owing to their high hydrophobicity and resistance to degradation in the environment, OCPs tend to accumulate in sediments with long half-lives, and in aquatic organisms through food chain, posing adverse effects, i.e., hormone-like effects to biota and human (Tsai 2010;Lyall et al. 2017). Consequently, although the production and usage of OCPs have been strictly limited, OCP pollution in aquatic sediments is still of global concern (Helm et al. 2011 ;Wang et al. 2013; Dirbaba et al. 2018).
Most notably, extremely high levels of OCPs have been reported in sediments surrounding industrial units that synthesize these chemicals, such as Ya-Er lake area in Hubei, central China, Lake Ørsjøen, Southern Norway: McGrath Lake, USA (Brevik et al. 1996;Wu et al. 1997; Anderson et al. 2013). In most cases, sediment remediation projects have been carried out, and one of the most frequent remediation solutions adopted for those heavily polluted sediments is "dig and dump" , involving disposal of the dredged sediments in a con ned disposal facility or on low-lying areas (Clément et al. 2010;Gomes 2013;Liu et al. 2018). There were millions m 3 of dredged sediments produced per project, which might result in severe secondary pollution and uncertain ecological threats to the surrounding environment and human health, if not properly treated. However, there were limited documentation on quantitative assessment of the ecological impairment of the residual pollutants coupled to chemical monitoring in the disposal elds (Zarull et al. 2002). The long-term fate and ecological impacts of OCP residues in dredged sediments deposited into aquatic ecosystem are therefore important topics of study.
Aquatic sediments serve as the reservoir of organic pollutants, thus can be used as natural archives to provide comprehensive information of historical OCP usage records (Lebeuf and Nunes 2005;Yang et al. 2010). On the other hand, OCPs in sediments undergo natural diffusion and attenuation processes, resulting in a secondary distribution pattern, that is more dependent on various factors, such as time, hydrodynamic condition and microbial activity compared to the primary source ( Lv et al. 2020), however, limited data are available regarding the environment fate of OCP residues in various subsurface environments, such as soil-groundwater system (Pan et al. 2019) and lakshore wetland where mixing of lake water and phreatic water occurs in shallow sediments. creating unique biogeochemical conditions for pollutant attenuation and migration.
In addition, microorganisms play a vital role in natural attenuation of organic pollutants, and they are sensitive to the synthetic compounds deposited in aquatic sediments with changes in their composition and activity (DeLorenzo et al. 2001). Several studies have demonstrated the responses of microbial community to different levels of OCPs in contaminated soils and in enhanced remediation treatments (Ali et al. 2019;Sun et al. 2019). Therefore, microbial community analysis could be conducted in line with pollutant analysis, to comprehensively re ect the ecological impacts of residual pollutant stresses in situ (Bissett et al. 2013;Azarbad et al. 2016).
Ya-Er lake is located in the middle and lower reaches of Yangtze River, and is the rst sub-lake of Ya-Er lake surrounding by Gedian chemical plants, wherein once recognized as the hot-spot for industrial OCPs, including hexachlorocyclohexanes (HCHs), chlorobenzenes, HCB, and their notorious derivates polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) (Xu et al. 1994;Wu et al. 1997;. As a countermeasure, the Ya-Er lake had been served as the largest oxidation pond system in Asia, which consisted of a series of ve compartments (400 ha, from Pond No.1 to No.5) to treat around 80,000 tons wastewater daily from 1979 to 1991. In 2003 and 2004, the seriously polluted sludge in the idle oxidation ponds No. 1 to 3 were dredged and pumped into the northern back ll site at Pond No. 1, with the upper layer covered with clay. Today, the back ll site has turned into a reed mash area and for agriculture (Li et al. 2021). The present study has collected the subsurface sediments from Pond No.1 of Ya-Er lake and aims to: (1) determine the concentrations of 19 OCPs in the subsurface layers of the lakeshore back ll site and the lakebed; (2) examine the composition and potential sources of those OCPs; (3) evaluate the ecological impacts of dominant OCPs on sedimentary microbial communities.

Sample collection
Sediment samples were collected from ve sampling sites with various depths at Ya-Er lake in July 2018. The geographic locations of the sampling sites are shown in Fig. 1. Duplicate sediment cores per site were taken by a portable sampler. Two lakeshore sites (YJF-landward and YJM-inward) were located at the cofferdam area, with an average sampling depth of 2.7 m. In eld, the subsurface sedimentary cores were divided into four sections from top to bottom, that is 0.5-0.9 m, 0.9-1.5 m, 1.5-2.0 m and 2.0-2.7 m, respectively. The other three sampling sites (YJA, YJB and YJC) were at the lakebed of the abandoned oxidation pond No. 1, along the past wastewater ow from the inlet to outlet. The lakebed sediment cores were sectioned at 15-20 cm intervals to a depth of 40-60 cm. For each section, approximately 15 g of undisturbed sub-core sample was taken with a cut-off plastic syringe, placed in an aseptic centrifuge tube, and immediately stored in dry ice for microbial analysis. The remaining parallel samples were mixed thoroughly to form a composite sample for measurements of sediment physicochemical properties and OCPs.

Chemicals
Mixed standard samples of OCPs (including o,p'-DDE, p,p'-DDD, p,p'-DDT, o,p'-DDT, α-HCH, β-HCH, γ-HCH, δ-HCH, HCB, aldrin, dieldrin, endrin, α-endosulfan, β-endosulfan, trans-chlordane, cis-chlordane, methoxychlor, heptachlor, and heptachlor epoxide), recovery indicators (tetrachloroxylene (TCmX) and dechlorobiphenyls (PCB209) ), and the internal standard compound for quantitative analysis (pentachloronitrobenzene (PCNB)) were purchased from Ultra Scienti c, USA. Dichloromethane (DCM), hexane and other organic solvents are chromatography pure. Silica gel and neutral alumina are 100~200 mesh, purchased from Qingdao Haiyang Chemical Co., and Sinopharm Medicine Holding Co., Ltd., China. Silica gel and neutral alumina were extracted in DCM for 48 h, dried at room temperature, and then baked for 12 h at 180℃ and 240℃, respectively. After cooling to room temperature, add 3% deionized water of its mass to reduce the activity, and store it in a drying dish for later use after balancing. Anhydrous sodium sulfate was purchased from Sinopharm Medicine Holding Co., Ltd., China. It was roasted in mu e furnace at 450 ℃ for 4 hours and stored in a sealed container for reserve after drying.

Extraction and instrument analysis of OCPs in sediment
The OCPs were extracted from the freeze-dried and screened (100 mesh) sediment samples and analyzed according to USEPA approved technique under Method 8080A (Huang et al. 2018). In brief, 10.000 g of sediment sample was spiked with 20 ng recovery indicators (including TCmX and PCB209) and mixed with anhydrous sodium sulfate. The sample was extracted with DCM in Soxhlet extractor for 24 h. At the same time, activated copper granules were added to the collection ask to remove elemental sulfur. The sediment extract was concentrated to 2 mL by rotary evaporation, then puri ed and separated by a chromatography column containing deactivated silica gel/alumina (V/V= 2:1), eluted with 30 mL of DCM/n-hexane (V/V= 2:3). The eluate solution was further concentrated to 0.5 mL by rotary evaporation and transferred to a 2 mL cell bottle. It was blown to 200 μL with soft, high-purity nitrogen. Finally, 20 ng (4 μL of 5 mg L -1 ) of internal standard PCNB was added and stored at -20℃, prior to instrumental analysis.

Microbial community analysis
Microbial community structure was characterized by ester-linked fatty acid methyl ester (EL-FAME) analysis (Schutter and Dick 2000). The extraction and quanti cation procedures of EL-FAMEs were mainly performed according to the previous report (Wang and Tam 2012). Quanti cation of FAMEs was carried out using a Thermo Scienti c Trace 1300 Gas Chromatograph and ISQ LT Mass Spectrometer (GC-MS) equipped with a 30 m DB-5MS capillary column. The injector temperature was 250 ℃; in splitless mode; with helium carrier gas ow velocity of 0.8 mL min -1 . The oven temperature was initially set at 50 °C (held for 2 min), increased at 50 °C min −1 to 150 °C (held for 2 min), then at 2.5 °C min −1 to 195 °C (held for 3 min) and a further increase at 2.5 °C min −1 to 240 °C (held for 5 min). The whole program lasted for 50 min. The MS detector was operated in electron-impact ionization (EI) in selective ion monitoring (SIM) mode. The quadrupole and ion-source temperatures were held at 230 °C and 280 °C, respectively. Fatty acid peaks were identi ed using Bacterial Acid Methyl Esters CP Mixture (Matreya LLC.). A standard fatty acid nomenclature, e.g., A:BωC, was adopted (Vestal and White 1989).

Statistical analysis
A parametric one-way analysis of variance (ANOVA), followed by Tukey's multiple comparison test, was used to reveal any signi cant differences in OCP concentrations among sampling sites at P≤0.05. A principal component analysis (PCA) using Varimax rotation was adopted to identify the potential sources of OCPs, that responsible for the variation in their distribution pro les ( Furthermore, a constrained ordination technique, redundancy analysis (RDA), was used to examine the relationships between the dominant OCP pollutants and microbial FA species by CANOCO 5.0 from Microcomputer Power (Ithaca, NY, USA).
The detection rates of the 19 OCPs varied from 44.4%-100.0% (Table 1). Eight OCPs were detected in all the samples, namely HCB, α-, β-, and δ-HCH isomers, p,p'-DDE, trans-chlordane (TC), cis-chlordane (CC) and heptachlor-epoxide. It was noted that p,p'-DDT exhibited the lowest detection rate, and was mainly detected in the lakeshore samples. Likewise, o,p'-DDT and p,p'-DDD had higher detection rates in the lakeshore compared to the lakebed samples.
The coe cients of variation (CV) of HCB, α-HCH, γ-HCH, all DDTs except p,p'-DDE were all over 140%, indicating great spatial heterogeneity among samples. From Fig. 2, there were much greater site-speci c variations between the lakeshore sites (YJF and YJM) than those in the lakebed sites (YJA&B&C or YJS). YJM exhibited the hotspots of HCB, with an average value of 495.2 ng g -1 , signi cantly higher than those in YJF (75.4 ng g -1 ) and YJS (76.6 ng g -1 ) (P<0.01). Moreover, the residual levels of HCB greatly increased with depth by an order of magnitude from M1 to M4, and nearly doubled from F1 to F3. By contrast, YJF harbored the highest residual levels of HCHs and DDTs (27.9 and 11.2 ng g -1 ), followed by YJM (15.0 and 4.9 ng g -1 ) and YJS (9.9 and 2.2 ng g -1 ). The concentration of HCHs and DDTs generally reached their peak values in the 0.9-2.0 m deep layers of the lakeshore sites (i.e., F2, F3 and M3). Besides, the sitespeci c variations for the other OCPs were relatively small and insigni cant.

Risk assessment of OCPs in the subsurface sediments
Two widely used sediment quality guidelines, recommended by US National Oceanic and Atmospheric Administration (NOAA 1999) and Canadian Council of Ministers of the Environment (CCME 2002), were applied to evaluate the possible ecotoxicological risk of OCPs in the study sites (Table 2). Both guidelines showed the highest risk for β-HCH, as all the sampling sites had concentrations above its ERM and PEL values. High risks of HCB, α-, γ-, and δ-HCHs were also indicated, with 94%, 83%, 76% and 83% of the sites above their respective ERL values, respectively. Notably, only YJB in the centre of the pond showed little risk of HCB and HCHs. As for DDTs, all the sampling sites had the concentrations below their ERM values, and only 6% of the sampling sites (at either F2 or F3) had the concentrations above PELs for o,p'-DDT, p,p'-DDT and p,p'-DDE. 39%, 6% and 50% of the sampling sites had the concentrations between TELs and PELs for p,p'-DDT, p,p'-DDD and p,p'-DDE, indicating the site-speci c moderate risks at the lakeshore. By contrast, 78% and 94% of the sampling sites had the concentrations below TELs for o,p'-DDT and p,p'-DDD. As for those minor OCPs, 100%, 78%, 100%, 100% and 100% of the sampling sites had the concentrations below TELs for heptachlor, heptachlor-epoxide, ΣChlordane, dieldrin and endrin, which exhibited low possibilities to have an adverse effect on sediment dwelling organisms. However, 61%, 78% and 67% of the sampling sites had the concentrations between ERLs and ERMs for ΣChlordane, dieldrin and endrin, which need further attention.
In addition, the ratios of TC to CC were mostly less than 0.76, indicating the historical inputs of technical chlordance (Li et al. 2006), except for F3 (1.3), M3 (2.1) and M2 (0.8). The ratios of α-to β-Endosulfan were far less than 2.33 (Rice et al. 1997), suggesting no new input for endosulfans in the study site.

PCA analysis for source identi cation
Principal component analysis (PCA) was performed in this study to further identify the OCP sources. Five principal components (PCs) with eigenvalues over 1 were extracted, accounting for 80.8% of the total variance (Table 3). PC1 was predominately weighted by three HCH isomers (δ-, α-and γ-HCH), p,p'-DDE, its parent p,p'-DDT, and trans-chlordane, and accounted for 27.6% of the total variance. HCHs were frequently reported as the main products of the previous local chemical plant from 1962-1987 (Xu et al. 1994). These compounds were representative dominant OCPs detected below the 1st subsurface layers of lakeshore sediments, especially at F3 and F2, but not at F4 (Fig. 4), indicating that they were from the historically polluted dredged sediments but did not extend to the 4th layer at YJF. PC2 was predominately composed by heptachlor-epoxide, its parent heptachlor, aldrin, methoxychlor and β-endosulfan, and accounted for 23.4% of the total variance. This cluster of OCPs were associated with their recent usage as alternative of DDTs dated after 1990s in agricultural activities (Niu et al. 2016). They were found dominant in the 1st subsurface layers at F1, M1, A1, and at F4 (where was once the surface layer before being covered by the dredged sediments), and decreased in the order of YJF>YJS>YJM. PC3 was heavily weighted by p,p'-DDD, o,p'-DDT, followed by cis-chlordane and α-endosulfan, responsible for 12.3% of the total variance. This pro le was indicative of usage of dicofol, chlordane and endosulfan (Qiu et al. 2005), and most typical at F2 besides those shallow 1st subsurface layers, when comparing with the contribution pattern of PC2. PC4 was responsible for 9.5% of the total variance. This factor was positively weighted by HCB and endrin, and negatively by β-HCH. HCB and β-HCH were among the most persistent OCPs in the environment (Willett et al. 1998), with endrin showing similar source. Notably, HCB was largely retained in M2-M4 at YJM, where β-HCH had the lowest levels and relative proportions. PC5 was weighted by deildrin, and responsible for 8.1% of the total variance.

Contribution of pollution sources
Based on the pollution sources identi ed by PCA, an APCS-MLR model was employed to establish the function relationship between pollution sources and concentrations of each OCP variable, for calculation of the source contributions. The linear relationships between the predicted and the observed values of each OCP showed good consistency, with R 2 values ranging from 0.65 to 0.94. Results of source apportionment with APCS-MLR model are shown in Fig. 5. Average contributions of different pollution sources of OCPs in the subsurface sediments of Ya-Er lake were 27.2%, 14.7%, 0.9%, 14.3% and 3.1% for historical industrial input (S1), recent agricultural input (S2), historical mixture input (S3), persistent residuals (S4) and diedrin (S5), respectively. Apart of these sources, the average contribution of unidenti ed source (B) was highest, up to 39.9%, which was much higher in most of the lakebed sediments compared to the lakeshore ones (Fig. 5A). Notably, the contributions of this unidenti ed source varied greatly among OCPs, which were polarized between the more easily-degraded or transformed OCPs like o,p'-DDT, trans-chlordane and aldrin (<5%) and those more persistent ones including p,p'-DDD, cis-chlordane, deidrin, β-HCH and HCB (42%-76%) (Fig. 5B). Thus, the unidenti ed source could be interpreted as source of biodegradation.

Ecological impacts of OCPs on microbial community
Microbial EL-FAME pro le of the 18 sediment samples (with duplicates or triplicates) was examined to explore the ecological impacts of OCPs on microbial community structure. The sample scatter plot based on PCA analysis of the microbial EL-FAME pro le resembled OCP distribution pattern, with the rst two PCs explained 71.9% of the total variance. The differences of microbial community structure among lakeshore samples from YJF and YJM were much greater than those among lakebed samples, in line with the detected gradients in OCP levels. Redundancy analysis (RDA) further con rmed a signi cant relationship between the EL-FAME pro le and nine dominant OCPs, including HCB, four HCH isomers and four DDTs. The rst and all canonical axes contributed to explain 52.0 and 78.4% of the total variance in microbial species data, with F-values of 8.7 (P = 0.014) and 3.2 (P = 0.008), respectively. From the RDA plot of environmental variables, p,p'-DDT, α-HCH, and δ-HCH, were highly, positively correlated with the rst axis, followed by p,p'-DDE, HCB and γ-HCH (Table 4 and Fig. 6). They all showed signi cant simple effects on microbial community structure (P<0.05, when individually tested). However, p,p'-DDT and HCB were selected in the rst and second place to explain 39.1% and 15.1% of the total variance, respectively, after the conditional effects of the three HCH isomers and p,p'-DDE decreased remarkably, indicating strong mutual correlations among these OCPs of the same origin-S1. Meanwhile, o,p'-DDT, also positively correlated with the second axis, was identi ed as the third important environmental factor affecting microbial community structure (8.8%, P = 0.01). These results indicate HCB and o,p'-DDT had posed independent in uences on microbial community composition in addition to p,p'-DDT.
The distribution of microbial FAME biomarkers was highly correlated to the OCP level and composition. Hydroxyl fatty acids (FAs), indicative of Pseudomonas, Actinomycetales, and Mycobacterium (Zelles 1999), were only enriched in those heavily-polluted dredged sediments at lakeshore sites (F2&F3, M2-M4) (Fig. 6A&B). In particular, enriched 18:2ω9, monounsaturated fatty acid (MUFA) of 18:1ω9c, and cyclopropane FA of cy19:0, were detected at F2 with the highest level of DDTs, suggesting a variety of microbial functional groups, i.e., fungi, aerobic and anaerobic gram-negative bacteria were actively involved to degrade DDTs at this site. High ratios of cis/trans 18:1ω9 and MUFA/Branch corresponded to the high levels of o,p'-DDT and p,p'-DDD (Fig. 6C). By contrast, branched FAs (i15:0, a15:0, i16:0 and i17:0) indicative of gram-positive bacteria were dominant in the lightly polluted lakebed sediments, but with high proportions of refractory OCPs like β-HCH.

Discussion
In this study, the pollution feature of 19 OCPs in subsurface sediments from the abandoned oxidation pond of Ya-Er lake was assessed over a decade after its dredging treatment. HCB and HCHs from the historical industrial discharge were still served as the two dominant OCPs in this pond. Although "hot spots" of HCB and HCHs mostly detected in the lakeshore sediment cores, the residual concentrations of HCB in lake subsurface sediments were around 1/1000 to 1/50 of the historical data recorded in 1991-1994, but the maximum and mean concentrations of ΣHCHs were just a fraction of the reported values at Pond No. 1-3 before dredging (Table 5). These ndings suggest that HCB and HCHs in the . This distinct discrepancy in OCP composition between the two lakeshore sites indicated weak exchange capability of the lipophilic OCPs adsorbed on sediment particles which had low mobility in the phreatic water through lateral dispersion. Nevertheless, the pro le of HCHs, DDTs and chlordanes in the shallow subsurface layer at lakebed site (A1) closest to the lakeshore more closely resembled those in the deep subsurface layers at inward lakeshore site (M2-M4) than the other lakebed sites (Fig. 3). This might be attributed by the potential interaction between lake water and phreatic water in the lake hyporheic zone when seasonal ood and storm run off occurred (Hester et al. 2013). It is thus necessary to combine hydrogeolocial survey and modeling to elaborate this process and quantify the exchange ux for further risk assessment.
The multivariate correlation analysis of the microbial and OCP data helps to directly evaluate the ecological impacts of OCPs in lake subsurface sediments. The most in uential OCP variables, including p,p'-DDT, α-HCH, δ-HCH, and p,p'-DDE, were generally more easily-degraded by a variety of microorganisms including both bacteria and fungi (Subba-Rao et al. 1985;Concha-Graña et al. 2006). So they might be more impactful to induce the shifts of microbial communities than those more persistent ones like β-HCH and HCB (Brahushi et al. 2004). Previous studies on aerobic OCP-degrading bacteria have shown that many gram-negative bacterial genera of Alcaligenes, Sphingobacterium, and Sphingomonas were capable of degrading DDT and HCH, while only one gram-positive bacterial genus of Nocardioides was reviewed involving the degradation of HCB (Nagata et al. 2016). The indication of potential microbial functional groups by FAME biomarkers showed a consistent result regarding the dominant OCP distribution. Moreover, the shift of dominant microbial groups could be used as good indicator to monitor the natural bio-attenuation process, during which microbial specialists capable of utilizing different types of organic compounds took the predominance in turn, in the order of their easiness of degradation. So the enriched gram-negative bacteria at F2 would be actively involved in the degradation of DDTs, while the microbial communities dominant with gram-positive bacteria were restored or screened out after those readily-degradable carbon sources were exhausting, similar to the microbial responses after oil spills (Kimes et al. 2014). However, as FAME biomarkers could not identify microbial functional groups at species level, further studies based on genotypic biomarkers, such as 16S rRNA and dehalogenase genes, would be employed to provide insights on the key microbial species involved and their role in bio-attenuation of OCPs under natural environmental conditions.

Conclusion
The present study indicates the OCP contamination has been mitigated greatly in the Ya-Er lake 15 years after dredging treatment. The concentrations of HCHs and DDTs were comparable to and/or lower to the values reported by previous studies on freshwater lakes. But a higher level of HCB was still found than the few reported values, especially in the hot spots distributed in 0.9-2.7 m deep layers in the lakeshore dump site. Source appointment analysis con rmed the main contributions of historical industrial input and its residuals at the lakeshore back ll site YJF and YJM, respectively. The relationship between OCP pro le and microbial community structure was signi cant, with DDTs and HCHs being identi ed as more important OCP variables than HCB in affecting microbial communities in lake sediments. The other widely-concerned POPs like PCDD/Fs and PCBs, and concomitant heavy metals like mercury, lead and arsenic, were not determined in this study. It thus required further investigation regarding their combined effects on microbial community structure and diversity, as well as on other aquatic biota, for a comprehensive assessment of the ecological impacts of OCPs on-site, thus providing case-speci c remediation strategies in this area.    Location of the Ya-Er lake and the sampling sites at Pond No.1 Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors. Source contributions of OCPs to each site (A) and to each OCP variable (B) by APCS-MLR model in subsurface sediments of Ya-Er lake. S1 indicates historical industrial input; S2 indicates recent agricultural input; S3 indicates historical mixture input; S4 indicates persistent residuals of HCB and β-HCH; S5 is Diedrin; B indicates unidenti ed source.

Figure 6
Sample score plot (A), loading plot of fatty acids (B) and biplot of OCPs-FAME biomarkers (C) based on the RDA analysis of EL-FAMEs pro les and the nine dominant OCPs in sediment core samples from Ya-Er lake. The six FAME biomarkers (G+: gram-positive bacteria; G-: gram-negative bacteria; MUFA/Branch: monounsaturated fatty acids/branched fatty acids; Cis/trans: 18:1ω9cis/18:1ω9trans;