We present two models to monitor the health of ecosystems by assessing hazard from a persistent organic compound to a top predator species. Our diet model predicts the dietary exposure of American Mink (Neovison vison) to PCB toxic equivalents (TEQ) by combining concentrations in their prey using weighted average proportions consistent with literature-based mink diets. Our bioaccumulation model predicts the dietary exposure of mink to PCB TEQ based on each congener’s total concentration in water (dissolved plus particulate fractions), the octanal/water partition coefficient (log K_ow) of the compound, and the trophic levels of prey taxa. Both models predict mink dietary concentrations which can be directly compared with each other and with lowest observable adverse effects concentrations (LOAECs) to assess chronic and acute hazards of PCB TEQ to mink. By our choice of certain parameters in the bioaccumulation model, we forced it to match the diet model within less than 5% for Eighteenmile Creek in western New York State. When the two models were used for a similar creek about 25 km away, the differences in their predictions were of the same magnitude.

我们提出了两个模型，通过评估持久性有机化合物对顶级捕食者物种的危害来监测生态系统的健康。我们的饮食模型通过使用与基于文献的水貂饮食一致的加权平均比例结合猎物中的浓度来预测美国水貂 ( Neovison vison ) 对 PCB 毒性当量 (TEQ) 的饮食暴露。我们的生物累积模型根据每种同源物在水中的总浓度（溶解加颗粒部分）、辛醛/水分配系数（log K _ow) 的化合物，以及猎物类群的营养水平。两种模型都可以预测水貂日粮浓度，这些浓度可以直接相互比较，并具有最低的可观察不良反应浓度 (LOAEC)，以评估 PCB TEQ 对水貂的慢性和急性危害。通过我们在生物蓄积模型中选择某些参数，我们迫使它与纽约州西部 18 英里溪的饮食模型的匹配度低于 5%。当这两个模型用于大约 25 公里外的一条类似小溪时，它们的预测差异是相同的。