基于梯度提升回归树的有机污染物生物-沉积物积累因子预测模型
Biota-sediment Accumulation Factor Models of Organic Chemicals in Benthic Invertebrates with Gradient Boosting Regression Tree
-
摘要: 生物-沉积物积累因子(BSAF)是评价底栖无脊椎生物对有机污染物生物积累能力的重要参数,是由化合物、底栖环境与无脊椎生物之间的三相作用决定的。现有模型通常采用线性算法研究化合物BSAF与化合物理化性质的关系,忽略了由于环境-生物-化合物相互作用引发的非线性影响,导致线性模型拟合和预测能力有限。本研究基于理化性质(PCP)和分子指纹(ECFP)描述化合物特征,结合环境样点和生物特征,采用梯度提升回归树(GBRT)的非线性算法,分别构建了底栖生物体内积累因子的GBRT-PCP和GBRT-ECFP预测模型,并与利用岭回归算法构建的线性模型进行比较。结果表明,GBRT模型训练集决定系数(R2)均为0.97,验证集R2为0.82~0.83,表明GBRT模型的拟合优度和预测能力显著优于岭回归模型(训练集和验证集R2分别为0.38~0.56和0.38~0.52)。沉积物有机碳含量对生物-沉积物积累因子的影响呈波动下降趋势,脂质含量呈先波动上升而后下降趋势。GBRT-PCP模型结果表明,化合物疏水性(logKOW)对生物积累影响呈先平稳后上升而后下降趋势,吸附性(logKOC)对生物积累呈波动下降趋势。总体上,具有中等logKOW(6.8~8.2)和中等logKOC(4.4~5.2)的化合物易于积累在生物组织。GBRT-ECFP模型阐明了稠环、芳香环、醚键、C-Br键、联苯键等结构是影响生物积累的关键子结构,该模型基于分子指纹结构可实现对化学品生物积累的高通量预测。本研究建立的模型为化学品生态风险评价和管理决策制定提供理论依据和方法参考。
-
关键词:
- 有机污染物 /
- 底栖无脊椎生物 /
- 生物-沉积物积累因子 /
- 梯度提升回归树
Abstract: Biota-sediment accumulation factor (BSAF) is an essential parameter to assess the bioaccumulation potential of benthic invertebrates for organic chemicals. The bioaccumulation process involves complicated interactions between compounds and environmental sites, and benthic invertebrates. Existing models mostly construct linear models for the relationship between bioaccumulation and physicochemical properties of compounds, neglecting interactions between the three factors mentioned above, resulting with poor goodness-of-fit and predictive ability. Here we developed logBSAF model based on gradient boosting regression tree algorithm (GBRT) with independent variables containing environmental site factors, biological factors, and two distinct compound variable regimes, i.e., physicochemical properties (PCP) and extended connectivity fingerprints (ECFP). In this study, the GBRT-PCP and GBRT-ECFP models of BSAF in benthic invertebrates were constructed, followed by comparisons of nonlinear models based on GBRT algorithm with linear models based on ridge algorithm. The determination coefficients (R2) of GBRT-PCP and GBRT-ECFP models for the training set were 0.97 and 0.82~0.83 for the validation set. Both GBRT models outperformed ridge models in terms of goodness-of-fit and predictive performance, with R2 of 0.38~0.56 for training and 0.38~0.52 for validation set, respectively. The organic carbon of sediments had the effect of fluctuating decline on BSAF. The lipid content of invertebrates showed a tendency for fluctuating increases and subsequent decreases on BSAF. GBRT-PCP model was conducted to identify the interactions between compound hydrophobicity (logKOW) and adsorption potential (logKOC) on BSAF. Results revealed that the logKOW values of compounds showed smooth increases followed by decreases on BSAF. The logKOC values of compounds exhibited fluctuating decreases. The interaction between logKOW and logKOC demonstrated that compounds with intermediatelogKOW(6.8~8.2) and logKOC(4.4~5.2) exhibit enhanced bioavailability. The developed GBRT-PCP model, involving the physicochemical characteristics of compounds as independent variables, could provide quantitative predictions for bioaccumulation of chemicals. Furthermore, substructure analysis of compounds based on GBRT-ECFP model identified the key substructures (e.g., annelated rings, aromatic rings, -O, C-Br bonds, and biphenyl bonds) related to BSAF. The GBRT-ECFP model could support high-throughput prediction performance of chemical bioaccumulation. Based on the GBRT-PCP model and GBRT-ECFP model, it provides benchmarks for the ecological risk assessment and management policy of chemicals. -
-
Zhen X M, Tang J H, Xie Z Y, et al. Polybrominated diphenyl ethers (PBDEs) and alternative brominated flame retardants (aBFRs) in sediments from four bays of the Yellow Sea, North China[J]. Environmental Pollution, 2016, 213:386-394 王国光, 刘巧灵, 冯丽娟, 等. 黄海、东海及其邻近海域沉积物中的典型持久性有机污染物[J]. 中国科学:化学, 2017, 47(11):1284-1297 Wang G G, Liu Q L, Feng L J, et al. Typical persistent organic pollutants (POPs) in sediments from the Yellow and East China Seas and adjacent coastal areas, China[J]. Scientia Sinica Chimica, 2017, 47(11):1284-1297(in Chinese)
池仕运, 王瑞, 魏秘, 等. 金沙江上中段大型底栖无脊椎动物群落结构特征和多样性分析[J]. 生态学报, 2022, 42(21):8723-8738 Chi S Y, Wang R, Wei M, et al. Community structure and diversity of macroinvertebrates in the upper and middle reaches of Jinsha River based on the monitoring data from 2010-2019[J]. Acta Ecologica Sinica, 2022, 42(21):8723-8738(in Chinese)
Guareschi S, Wood P J. Taxonomic changes and non-native species:An overview of constraints and new challenges for macroinvertebrate-based indices calculation in river ecosystems[J]. Science of the Total Environment, 2019, 660:40-46 Resh V H, Norris R H, Barbour M T. Design and implementation of rapid assessment approaches for water resource monitoring using benthic macroinvertebrates[J]. Austral Ecology, 1995, 20(1):108-121 陈凯, 于海燕, 张汲伟, 等. 基于底栖动物预测模型构建生物完整性指数评价河流健康[J]. 应用生态学报, 2017, 28(6):1993-2002 Chen K, Yu H Y, Zhang J W, et al. Predictive model based multimetric index of macroinvertebrates for river health assessment[J]. Chinese Journal of Applied Ecology, 2017, 28(6):1993-2002(in Chinese)
Buchwalter D B, Jenkins J J, Curtis L R. Respiratory strategy is a major determinant of [3H]water and[14C] chlorpyrifos uptake in aquatic insects[J]. Canadian Journal of Fisheries and Aquatic Sciences, 2002, 59(8):1315-1322 Ratier A, Lopes C, Labadie P, et al. A Bayesian framework for estimating parameters of a generic toxicokinetic model for the bioaccumulation of organic chemicals by benthic invertebrates:Proof of concept with PCB153 and two freshwater species[J]. Ecotoxicology and Environmental Safety, 2019, 180:33-42 Conrad A U, Comber S D, Simkiss K. Pyrene bioavailability; effect of sediment-chemical contact time on routes of uptake in an oligochaete worm[J]. Chemosphere, 2002, 49(5):447-454 Gewurtz S B, Lazar R, Douglas Haffner G. Comparison of polycyclic aromatic hydrocarbon and polychlorinated biphenyl dynamics in benthic invertebrates of Lake Erie, USA[J]. Environmental Toxicology and Chemistry, 2000, 19(12):2943-2950 Hope B, Scatolini S, Titus E, et al. Distribution patterns of polychlorinated biphenyl congeners in water, sediment and biota from Midway Atoll (North Pacific Ocean)[J]. Marine Pollution Bulletin, 1997, 34(7):548-563 Thomann R V, Komlos J. Model of biota-sediment accumulation factor for polycyclic aromatic hydrocarbons[J]. Environmental Toxicology and Chemistry, 1999, 18(5):1060-1068 Epplett T D, Gewurtz S, Lazar R, et al. Seasonal dynamics of PCBs in the plankton of Lake Erie[J]. Journal of Great Lakes Research, 2000, 26(1):65-73 Wang Z, Ma X D, Lin Z S, et al. Congener specific distributions of polybrominated diphenyl ethers (PBDEs) in sediment and mussel (Mytilus edulis) of the Bo Sea, China[J]. Chemosphere, 2009, 74(7):896-901 Prosser R S, Mahon K, Sibley P K, et al. Bioaccumulation of perfluorinated carboxylates and sulfonates and polychlorinated biphenyls in laboratory-cultured Hexagenia spp., Lumbriculus variegatus and Pimephales promelas from field-collected sediments[J]. Science of the Total Environment, 2016, 543:715-726 Arnot J A, Gobas F A. A review of bioconcentration factor (BCF) and bioaccumulation factor (BAF) assessments for organic chemicals in aquatic organisms[J]. Environmental Reviews, 2006, 14(4):257-297 Mukai Y, Goto A, Tashiro Y, et al. Coastal biomonitoring survey on persistent organic pollutants using oysters (Saccostrea mordax) from Okinawa, Japan:Geographical distribution and polystyrene foam as a potential source of hexabromocyclododecanes[J]. Science of the Total Environment, 2020, 739:140049 Liu W X, He W, Wu J Y, et al. Residues, bioaccumulations and biomagnification of perfluoroalkyl acids (PFAAs) in aquatic animals from Lake Chaohu, China[J]. Environmental Pollution, 2018, 240:607-614 American Chemical Society. Chemical abstracts service[EB/OL].[2023-01-10]. https://www.cas.org/zh-hans/cas-data/cas-registry Di Toro D M, Zarba C S, Hansen D J, et al. Technical basis for establishing sediment quality criteria for nonionic organic chemicals using equilibrium partitioning[J]. Environmental Toxicology and Chemistry, 1991, 10(12):1541-1583 Wang X L, Zhong W J, Xiao B W, et al. Bioavailability and biomagnification of organophosphate esters in the food web of Taihu Lake, China:Impacts of chemical properties and metabolism[J]. Environment International, 2019, 125:25-32 Yoon S J, Hong S, Kim T, et al. Occurrence and bioaccumulation of persistent toxic substances in sediments and biota from intertidal zone of Abu Ali Island, Arabian Gulf[J]. Marine Pollution Bulletin, 2019, 144:243-252 Sun R X, Luo X J, Tang B, et al. Bioaccumulation of short chain chlorinated paraffins in a typical freshwater food web contaminated by e-waste in South China:Bioaccumulation factors, tissue distribution, and trophic transfer[J]. Environmental Pollution, 2017, 222:165-174 Ma X D, Zhang H J, Wang Z, et al. Bioaccumulation and trophic transfer of short chain chlorinated paraffins in a marine food web from Liaodong Bay, North China[J]. Environmental Science & Technology, 2014, 48(10):5964-5971 Arnot J A, Gobas F. A food web bioaccumulation model for organic chemicals in aquatic ecosystems[J]. Environmental Toxicology and Chemistry, 2004, 23(10):2343-2355 Goto A, Tue N M, Isobe T, et al. Nontarget and target screening of organohalogen compounds in mussels and sediment from Hiroshima Bay, Japan:Occurrence of novel bioaccumulative substances[J]. Environmental Science & Technology, 2020, 54(9):5480-5488 La Guardia M J, Hale R C, Harvey E, et al. In situ accumulation of HBCD, PBDEs, and several alternative flame-retardants in the bivalve (Corbicula fluminea) and gastropod (Elimia proxima)[J]. Environmental Science & Technology, 2012, 46(11):5798-5805 Wang F E, Wang F X, Yang H R, et al. Ecological risk assessment based on soil adsorption capacity for heavy metals in Taihu Basin, China[J]. Environmental Pollution, 2023, 316:120608 McCoubrey L E, Thomaidou S, Elbadawi M, et al. Machine learning predicts drug metabolism and bioaccumulation by intestinal microbiota[J]. Pharmaceutics, 2021, 13(12):2001 Hu B F, Xue J, Zhou Y, et al. Modelling bioaccumulation of heavy metals in soil-crop ecosystems and identifying its controlling factors using machine learning[J]. Environmental Pollution, 2020, 262:114308 Gao F, Shen Y K, Sallach J B, et al. Direct prediction of bioaccumulation of organic contaminants in plant roots from soils with machine learning models based on molecular structures[J]. Environmental Science & Technology, 2021, 55(24):16358-16368 Belfroid A C, Sijm D T H M, Van Gestel C A M. Bioavailability and toxicokinetics of hydrophobic aromatic compounds in benthic and terrestrial invertebrates[J]. Environmental Reviews, 1996, 4(4):276-299 United States Environmental Protection Agency (US EPA), Office of Pollution Prevention and Toxics. Exposure Assessment Tools and Models, Estimation Program Interface (EPI) Suite Ver 4.1[CP/OL]. https://www.epa.gov/tsca-screening-tools/epi-suitetm-estimation-program-interface Rogers D, Hahn M. Extended-connectivity fingerprints[J]. Journal of Chemical Information and Modeling, 2010, 50(5):742-754 Friedman J H. Greedy function approximation:A gradient boosting machine[J]. The Annals of Statistics, 2001, 29(5):1189-1232 Maruya K A, Lee R F. Biota-sediment accumulation and trophic transfer factors for extremely hydrophobic polychlorinated biphenyls[J]. Environmental Toxicology and Chemistry, 1998, 17(12):2463-2469 Cornelissen G, Gustafsson O, Bucheli T D, et al. Extensive sorption of organic compounds to black carbon, coal, and kerogen in sediments and soils:Mechanisms and consequences for distribution, bioaccumulation, and biodegradation[J]. Environmental Science & Technology, 2005, 39(18):6881-6895 Wang R B, Li X M, Xu J H, et al. Bioavailability for organic chemical bioaccumulation follows the power law[J]. Environmental Pollution, 2021, 288:117716 李胜利, 易茂红, 陈凯, 等. 渭河底栖动物性状和功能对空间尺度环境变量响应的生态区差异性[J]. 生态学报, 2018, 38(7):2566-2578 Li S L, Yi M H, Chen K, et al. Differences in responses of macroinvertebrate traits and functional diversity to environmental variables at different spatial scales between ecoregions in the Wei River Basin, China[J]. Acta Ecologica Sinica, 2018, 38(7):2566-2578(in Chinese)
Pramanik S, Roy K. Modeling bioconcentration factor (BCF) using mechanistically interpretable descriptors computed from open source tool "PaDEL-Descriptor"[J]. Environmental Science and Pollution Research, 2014, 21(4):2955-2965 Kanaly R A, Harayama S. Biodegradation of high-molecular-weight polycyclic aromatic hydrocarbons by bacteria[J]. Journal of Bacteriology, 2000, 182(8):2059-2067 Nakata H, Murata S, Filatreau J. Occurrence and concentrations of benzotriazole UV stabilizers in marine organisms and sediments from the Ariake Sea, Japan[J]. Environmental Science & Technology, 2009, 43(18):6920-6926 -
点击查看大图
计量
- 文章访问数: 1666
- HTML全文浏览数: 1666
- PDF下载数: 129
- 施引文献: 0
下载:
