Search USGSSearch

Geology topics

C.G. Severn

Publications and source records attributed to C.G. Severn.

2 recordsLinked to original sources

Evaluation of numerical sediment quality targets for the St. Louis River Area of Concern

Numerical sediment quality targets (SQTs) for the protection of sediment-dwelling organisms have been established for the St. Louis River Area of Concern (AOC), 1 of 42 current AOCs in the Great Lakes basin. The two types of SQTs were established primarily from consensus-based sediment quality guidelines. Level I SQTs are intended to identify contaminant concentrations below which harmful effects on sediment-dwelling organisms are unlikely to be observed. Level II SQTs are intended to identify contaminant concentrations above which harmful effects on sediment-dwelling organisms are likely to be observed. The predictive ability of the numerical SQTs was evaluated using the matching sediment chemistry and toxicity data set for the St. Louis River AOC. This evaluation involved determination of the incidence of toxicity to amphipods (Hyalella azteca) and midges (Chironomus tentans) within five ranges of Level II SQT quotients (i.e., mean probable effect concentration quotients [PEC-Qs]). The incidence of toxicity was determined based on the results of 10-day toxicity tests with amphipods (endpoints: survival and growth) and 10-day toxicity tests with midges (endpoints: survival and growth). For both toxicity tests, the incidence of toxicity increased as the mean PEC-Q ranges increased. The incidence of toxicity observed in these tests was also compared to that for other geographic areas in the Great Lakes region and in North America for 10- to 14-day amphipod (H. azteca) and 10- to 14-day midge (C. tentans or C. riparius) toxicity tests. In general, the predictive ability of the mean PEC-Qs was similar across geographic areas. The results of these predictive ability evaluations indicate that collectively the mean PEC-Qs provide a reliable basis for classifying sediments as toxic or not toxic in the St. Louis River AOC, in the larger geographic areas of the Great Lakes, and elsewhere in North America.

Archives of Environmental Contamination and Toxico

Evaluating sediment chemistry and toxicity data using logistic regression modeling

This paper describes the use of logistic-regression modeling for evaluating matching sediment chemistry and toxicity data. Contaminant- specific logistic models were used to estimate the percentage of samples expected to be toxic at a given concentration. These models enable users to select the probability of effects of concern corresponding to their specific assessment or management objective or to estimate the probability of observing specific biological effects at any contaminant concentration. The models were developed using a large database (n = 2,524) of matching saltwater sediment chemistry and toxicity data for field-collected samples compiled from a number of different sources and geographic areas. The models for seven chemicals selected as examples showed a wide range in goodness of fit, reflecting high variability in toxicity at low concentrations and limited data on toxicity at higher concentrations for some chemicals. The models for individual test endpoints (e.g., amphipod mortality) provided a better fit to the data than the models based on all endpoints combined. A comparison of the relative sensitivity of two amphipod species to specific contaminants illustrated an important application of the logistic model approach.

Environmental Toxicology and Chemistry