A MODULAR ESTUARINE MESOCOSM

The need to simulate unique features of estuaries and the modes of ent ry of pollutants has led the authors to develop a modular approach to toxicity testing within an estuarine system. The approach incorporates replicate test systems that simulate the major aspects that are uniqu e to estuarine systems. Initial validation of the design simulated a r eference estuary on Wadmalaw Island, South Carolina. Replicate mesocos ms were exposed to target concentrations 0, 2, and 8 mu g/L of azinpho smethyl (AZM), an organophosphate insecticide. Single species toxicity tests were conducted within the systems in conjunction with the ecosy stem simulations. High doses of AZM killed all animal life in the simu lations, whereas lower concentrations showed a dose-response effect. T wenty-four-hour average AZM levels were used to calculate 24-h and 96- h LC50 values for Atlantic silverside minnows (Menidia menidia), mysid shrimp (Mysidopsis bahia), and grass shrimp (Palaemonetes pugio). Mys id shrimp were most sensitive to AZM (96 h LC50 = 0.811 mu g/L), where as grass shrimp were least sensitive (96 h LC50 = 1.34 mu g/L). Dissol ved oxygen was reduced and ammonia concentrations increased from high doses of AZM. Increases in ammonia levels were not reduced by intersti tial microbes, resulting in sustained high (>0.08 mg/L) ammonia concen trations. Lower doses were quickly reduced in the mesocosm. This study demonstrates that ecosystem level responses can be evaluated by the m odular approach. This modular approach can be modified to simulate mos t aquatic ecosystems.