Premium
Uses of Probabilistic Exposure Models in Ecological Risk Assessments of Contaminated Sites
Author(s) -
Macintosh David L.,
Suter Glenn W.,
Hoffman F. Owen
Publication year - 1994
Publication title -
risk analysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.972
H-Index - 130
eISSN - 1539-6924
pISSN - 0272-4332
DOI - 10.1111/j.1539-6924.1994.tb00259.x
Subject(s) - food web , variance (accounting) , environmental science , probabilistic logic , heron , point estimation , range (aeronautics) , ecology , exposure assessment , monte carlo method , statistics , econometrics , mathematics , biology , ecosystem , engineering , accounting , aerospace engineering , business
Food web models have two uses in assessments of environmental contaminants. First, they are used to determine whether remediation is needed by estimating exposure of end‐point species and subsequent effects. Second, they are used to establish cleanup goals by estimating concentrations of contaminants in ambient media that will not cause significant effects. This paper demonstrates how achievement of these goals can be enhanced by the use of stochastic food web models. The models simulate the dynamics of PCBs and mercury in the food webs of mink and great blue herons. All parameters of the models are treated as having knowledge uncertainty, due to imperfect knowledge of the actual parameter values for the site, chemicals, and species of interest. This uncertainty is an indicator of the potential value of additional measurements. In addition, those parameters that are responsible for variance among individual organisms are assigned stochastic uncertainty. This uncertainty indicates the range of body burdens that are expected when the end‐point species are monitored. These two types of uncertainty are separately accounted for in Monte Carlo simulations of the models. Preliminary monitoring results indicate that the models give reasonably good estimates of heron egg and nestling body burdens and of variance among individuals.