【作者】 胡彦兵；

【导师】 汝少国； 马建新；

【作者基本信息】 中国海洋大学 ， 生态学， 2013， 博士

【摘要】 六六六(HCH)、滴滴涕(DDT)和有机锡(OT)是全球近海广泛分布的持久性有机污染物(POPs)，在极微量的水平上就能够对多种海洋生物产生毒性效应，而且能通过海洋食品对人体健康造成损害。目前我国近海HCH、DDT、OT污染现象仍然较为普遍，对其开展生态风险评价可为环境决策提供依据。目前通用的评价方法是将水体或沉积物中污染物的浓度与体外暴露试验的毒性浓度值(或环境基准)进行比较表征风险水平。然而体外暴露没有考虑食物暴露以及污染物的毒物动力学过程对生物蓄积的影响，将低估多种途径联合暴露对海洋生物的风险。本研究监测了烟台金城湾养殖海域水体以及沉积物中HCH、DDT、OT的浓度，并据此构建了逸度食物网模型，估算了食物网不同功能群中生物体内的污染物浓度。采用了多种模型构建了HCH、DDT、OT的生物敏感度分布(SSD)曲线，并估算了安全浓度HC5。在构建逸度食物网模型和SSD模型的基础上评价了该海域HCH、DDT、OT污染对海洋生物的生态风险和水产品消费对人体健康的风险。研究结果表明：(1)表层海水中∑HCH和∑OT(TBT+DBT+MBT)的浓度(OT浓度以Sn计，下同)分别为2.98-14.87和23.88-44.82ng/L，DDT均未检出(<0.032ng/L)；表层沉积物中∑HCH、∑DDT、∑OT的浓度分别为5.52-9.43、4.11-6.72和4.26-4.38ng/g。表层沉积物中γ-HCH、p,p’-DDE的浓度分别为0.64-3.13和1.36-4.02ng/g，高出Macdonald所建立的沉积物质量基准值TEL(γ-HCH，0.32ng/g， p,p’-DDE，2.07ng/g)的概率分别为100%和66.7%；表层海水中TBT的浓度为0.60-2.90ng Sn/L，接近或超过其导致雌性狗岩螺(Nucella lapillus)性畸变的浓度(1-2ng Sn/L)。监测结果初步显示，该海域HCH、DDT和OT残留对于水产品健康具有潜在风险。(2)不同SSD模型对于HCH、DDT、OT毒性数据的拟合效力是不同的，其中对数逻辑斯蒂模型是拟合效果较好的参数模型，但是对于大样本的数据(N>80)其拟合效力明显减弱。非参数的bootstrap以及修正的bootstrap方法均能够较好的吻合原始毒性数据，非常适合于大样本数据SSD曲线的构建，然而这两种方法对原始数据具有很强的依赖性，在采用小样本的数据(N<20)构建SSD曲线时将显著低估污染物对更敏感物种的毒性效应。Bootstrap回归方法能够将参数方法和非参数bootstrap方法的优点结合起来，从而获得较为可靠的HC5值，但对计算机运算性能有较高的要求。(3)构建了逸度食物网模型，模拟了金城湾养殖海域食物网中HCH、DDT、OT的迁移转化过程，并估算了其在13个功能群中生物体内的浓度，预测值与实测值吻合良好。模拟结果表明，部分高营养级的功能群中生物体内∑HCH的含量(2.54-3.24ng/g)，均显著低于更低级的功能群中的平均含量(8.11-26.75ng/g，P<0.01)，不呈现食物网生物放大效应；OT的模拟结果与HCH相似；而DDT随着食物网营养级的升高逐级生物放大，其中较高营养级功能群中∑DDT的平均浓度(22.89-69.45ng/g)，比较低营养级的功能群(0.10-0.37ng/g)高2-3个数量级，该结果与实验室内暴露实验以及野外观测结果的文献报道情况非常吻合。(4)根据该海域食物网中生物体内污染物浓度的模型预测值以及利用生物体内毒性浓度构建的SSD曲线，采用概率方法估算了其对海洋生物的生态风险值。结果表明，按照非保守的估算方法和两种保守的估算方法，HCH的总体风险概率分别为0.029、0.070和0.082，DDT的总体风险概率分别为0.086、0.13和0.20，TBT和DBT的总体风险概率分别为0.09、0.21、0.22和0.06、0.17、0.17。可见HCH对海洋生物的生态风险保守估算值略高于5%的边界管理水平，而DDT、TBT和DBT的非保守估算值和保守估算值生均高于5%，因而对敏感的海洋生物均具有潜在风险。与目前通用(基于水体暴露浓度和水体暴露毒性浓度)的生态风险评价方法所获得结果(HCH为0.0087、0.021、0.020，DDT为0.0013、0.0037、0.0028，TBT为0.04、0.07、0.10，DBT为0.10、0.36、0.15)相比，本方法获得风险值普遍是偏高的，其中的HCH、OT风险值偏差均不超过3倍，而DDT的风险值则高出目前通用方法约30-70倍，表明目前通用方法由于没有考虑到食物暴露因素，低估了DDT这样具有典型食物网生物放大效应的POPs对高营养级海洋生物的生态风险。可见，综合多重暴露途径的、基于生物体内污染物浓度的生态风险评价方法更为可靠。(5)根据生物体内HCH、DDTs和OT浓度的模型预测值，按照EPA的方法分别计算了该海域水产品的消费限量。从OT暴露角度看，该海域水产品的最大允许消费量高于国内海洋食品高消费地区的人均消费量的统计值，不会危害消费者健康；从HCH和DDT暴露非致癌毒性效应的角度看，食用该海域水产品不会危害人体健康，但是从致癌效应的角度看，该海域水产品的最大允许消费量为国内水产品人均消费量的1/5，为海洋食品高消费地区的人均消费量的1/6。本论文首次以逸度食物网模型计算了养殖海域各营养级生物体内的HCH、DDT和OT的浓度，并进行了生态风险评价以及水产品消费对人体健康风险评价。该方法可对尚未有养殖活动的海域开展生态风险评价，并能在开展养殖生产前判断该海域是否适宜水产品养殖，为完善水产品生态和健康风险评价提供了新的技术方法，也为养殖海域水产品安全管理提供了科学依据。更多还原

【Abstract】 Hexachlorocyclohexanes (HCHs), dichlorodiphenyltrichloroethanes (DDTs) andorganotins (OTs) are types of persistent organice pollutants (POPs) extensivelydistributed offshore all around the world, which could cause a series of adverse effectson various marine organisms at very low levels and cause harm to human health viamarine foodstuff. Due to the high levels of the HCH, DDT and organotin residuesoffshore in China, their potential risk to marine organisms and human health shoulddraw attention. Aquatic ecological risk assessment (ERA) is usually based on externalwater-borne exposure concentrations of the target pollutants, which could be definedas an external ERA (EERA). However, the exposure via diet and the toxicity kineticsof target toxicants are not taken into consideration in an EERA, which wouldunderestimate their true risk to organisms in the environment. Therefore, an ERAbased on biological tissue residue of target toxicants defined as an internal ERA(IERA), was addressed in the present study.The concentrations of HCHs, DDTs and OTs in the water and sediment of theJincheng Bay mariculture area (JBMA) were investigated in the present study, andaccordingly a fugacity-based foodweb bioaccumulation model was developed toestimate the tissue residue levels of these pollutants in various kinds of speciesbelonging to different function groups (FGs) of the JBMA foodweb. Besides, speciessensitivity distribution (SSD) curves were constructed for HCHs, DDTs and OTsemploying multiple models, and the safety levels (HC5) were derived based on thesecurves. By integrating the internal concentration (biological tissue residue levels)generated from the bioaccumulation model with the SSD curves based on internaltoxicity data, an ERA and human health risk assessment (HRA) were performed. Themain results are as follows: (1) The concentrations of the total HCHs (∑HCH),∑OT (ng Sn/L in sea waterand ng Sn/g in sediment), and DDTs (∑DDT) were2.98-14.87,23.88-44.82ng/L,and below the detection limit (<0.032ng/L), respectively, in the surface seawater and5.52-9.43,4.11-6.72, and4.11-6.72ng/g, respectively, in the surface sediment. Thesediment concentrations of γ-HCH and p,p’-DDE were0.64-3.13and1.36-4.02ng/g,respectively, which exceeded the TEL values (0.32ng/g for γ-HCH and2.07ng/g forp,p’-DDE) of the Sediment Quantity Guidelines by the frequency of100%and66.7%,respectively. And the aqueous concentrations of TBT were0.60-2.90ng Sn/L, whichwere comparable or above the critical levels (1-2ng Sn/L) that could deduce imposexfor female Nucella lapillus. The investigation primarily suggested that the HCH, DDTand OT residues were potentially detrimental to marine organisms and the safety ofaquatic products in the study area.(2) The goodness of fit of different SSD models for the toxicity data of HCHs,DDTs and OTs differed, among which the log-logistic models showed the bestgoodness. However, the log-logistic model did not fit well for toxicity with a largesample size as N>80. All the SSD models were well fitted with both the bootstrap andmodified bootstrap approaches for the toxicity data of these compounds, especially forthe toxicity data with a large data size. However, both the two methods depend muchon the original toxicity data, which would underestimate the toxicity to the potentialmore sensitive species not involved in the data assemblage with a small sample size(N<20). The bootstrap regression method could integrate the excellence of both thenon-parametric bootstrap and parametric approaches, with which a reliable HC5couldbe derived. However, this approach costs much more time and a computer with higherspeed is required.(3) A fugacity-based foodweb bioaccumulation model was constructed in thisstudy. Employing this model we stimulated the transportation and transformation ofHCHs, DDTs and OTs in the foodweb of the JMBA, estimated their tissue residuelevels in thirteen FGs. The estimated values were comparable with the observedvalues for validation. According to the simulations, the residue levels of∑HCH(2.54-3.24ng/g) in some FGs of higher level were significantly lower (P<0.01) than those (8.11-26.75ng/g) in FGs of lower level, suggesting no biomagnification in thefoodweb. And similar results were obtained from the simulations of OTs. In contrast,significant biomagnification in the foodweb for DDTs was predicted, with the meantissue residual level ranged from22.89to69.45ng/g in the FGs of higher level, whichwere significantly higher (P<0.01) than those (0.10-0.37ng/g) in the FGs of lowerlevel by2-3folds. These results were highly consistent with those experimented inboth the laboratory and the field, which had been reported in numerous references.(4) The potential ecological risk values of HCHs, DDTs and OTs to marineorganisms were estimated by integrating their biological tissue residue levels based onthe bioaccumulation model with the SSDs based on internal toxicity data. The overallrisk probabilities of∑HCH,∑DDT, TBT and DBT were0.029,0.086,0.09and0.06,respectively on non-conservative estimation basis, and0.070,0.13,0.21and0.017,respectively on one conservative estimation basis. And the results of anotherconservative estimation were similar to those of the former conservative estimation.These results showed that the risk level of∑HCH was higher than0.05based on theconservative estimations, a critical level for management control, whereas those of∑DDT, TBT and DBT were all higher than0.05based on both the conservative andnon-conservative estimations, suggesting potential detrimental effects on marineorganisms. Besides, based on external toxicity data, the overall risk was0.0087,0.0013,0.04, and0.01for∑HCH,∑DDT, TBT, and DBT, respectively evaluated bynon-conservative estimation, and0.021,0.0037,0.07, and0.36evaluated by the firstconservative estimation, similar to that evaluated by the second conservativeestimation. It is clear that slight differences of the risk values between the IERA andEERA were observed for both HCHs and OTs, with a factor of less than4, while therisk values based on IERA were much higher than those based on EERA byapproximately30to70folds. The comparison revealed that underestimate of trueecological risk would be performed for the POPs which presenting biomagnificationeffects such as DDTs, while an IERA based on an aggregate exposure viamultipahway including the food exposure was more reliable.(5) According to the EPA approaches, the consumption limits of the marine food in the JMBA were calculated based on the biological tissue residue levels of HCHs,DDTs and OTs. The maximum allowable consumption rate (CR) of marine food washigher than the domestic average consumption rate, suggesting little or no risk tohuman health from the exposure of OTs through marine food of this area. And similarresults were concluded for noncancer health effects from the exposure of HCHs andDDTs. However, considering the cancer health effects from the exposure of HCHsand DDTs, the CR was just1/5and1/6of the average consumption rate in China andthe average consumption rate in a typical coastal area with high consumption rate ofmarine food, respectively.In this study, the biological residue levels of HCHs, DDTs and OTs wereestimated via the fugacity-based foodweb bioaccumulation model, which wereemployed for ERA and HRA for the first time. This model could be applied toconduct an ERA in marine areas to determine whether the area was available foraquaculture anterior to the aquaculture performance, which could also serve as a basisfor the safety management in mariculture sea areas.更多还原