AQUATOX: Modeling Fate of Toxic Organics in the Galveston Bay Ecosystem

AQUATOX is a process-based, time-varying, fate and effects simulation model that integrates aquatic ecology, chemical dynamics, bioaccumulation, and ecotoxicology. It can be used to predict the environmental fate and both direct and indirect simultaneous effects of nutrients, sediments, and up to 20 toxic chemicals in aquatic ecosystems. It provides Latin hypercube uncertainty analysis and nominal sensitivity analysis for any and all loadings and chemical and biotic parameters. Results are given in tabular and graphical forms, including concentrations, rates, mass balances, probabilistic risk graphs, and tornado diagrams (for sensitivity analysis). The model has been peer reviewed and released for use by the U.S. Environmental Protection Agency. The model also has been favorably reviewed in the open literature. One review stated: "AQUATOX fully closes the loop between eutrophication, contaminant fate and effects and, as such, is the most complete model described in the literature." Release 2.2 can represent as few or as many biotic groups as desired (up to 34 algal, macrophyte, invertebrate, and fish groups plus up to 15 age classes of one game fish). It has been validated for a variety of environments including ponds, lakes, reservoirs, small streams, and rivers. The model is a part of the BASINS modeling system and has linkages to the HSPF and SWAT runoff models to represent an integrated watershed. Recently an estuarine mode was incorporated in Release 3, which is in beta test. Application to Galveston Bay, Texas, exemplifies the analysis of bioaccumulation of toxicants in commercial fish and sea gulls. Representative published ecosystem data from Galveston Bay were used to calibrate the estuarine model. Because the model is intended to be exploratory in nature, and Galveston Bay was used only as an illustrative estuary, the objective was to obtain approximate, reasonable behavior without spending significant time on the calibration. Published PCB data from Barnstable Harbor, Massachusetts, were used to verify the generality of the estuarine ecosystem bioaccumulation model. The observed concentrations of total PCBs in the water and bottom sediments in the Massachusetts site were set as constant values in a simulation of Galveston Bay. The predicted concentrations in the various biotic compartments at the end of the simulation were similar to those observed in Barnstable Harbor, even though the sites are different. The model then was used to analyze the biomagnification potential of PFOS, a persistent surfactant, in Galveston Bay. The PFOS concentration was kept constant at 1 microgram/L in the water column. Over the course of the year's simulation, PFOS reached an approximate steady state in the various biotic compartments. Catfish, sea bass, and redfish have similar predicted body burdens, according to the simulations. Shrimp and oysters have somewhat lower predicted body burdens. The concentration in gulls reflects the weighted exposure in their various prey species, including dead organisms. In another simulation experiment a nominal concentration of 1 Ýg/L of PFOS was used as an initial condition with no additional loadings of PFOS to the estuary. Initially the PFOS was shown to mix into the lower layer. However, over a year the normal flushing of the Bay was predicted to remove 94% of the PFOS through washout. Of the remaining PFOS, 9% was predicted to be in fish and 5% in invertebrates.