2364 Environmental Toxicology and Chemistry, Vol. 19, No. 9, pp. 2364–2371, 2000 Printed in the USA 0730-7268/00 $9.00 + .00 PREDICTION OF POPULATION-LEVEL RESPONSE FROM MYSID TOXICITY TEST DATA USING POPULATION MODELING TECHNIQUES ANNE KUHN,*² W AYNE R. MUNNS,JR.,² S HERRY POUCHER,‡ DENISE CHAMPLIN,² and S UZANNE LUSSIER² ²U.S. Environmental Protection Agency, Atlantic Ecology Division, 27 Tarzwell Drive, Narragansett, Rhode Island 02282 ‡Science Applications International Corporation, 221 Third Street, Newport, Rhode Island 02840, USA ( Received 23 April 1999; Accepted 13 January 2000) Abstract—Acute and chronic bioassay statistics are used to evaluate the toxicity and risks of chemical stressors to the mysid shrimp Americamysis bahia (formerly Mysidopsis bahia). These include LC50 values from acute tests, chronic values (the geometric mean of the no-observed-effect concentration and the lowest-observed-effect concentration from 7-d and life-cycle tests), and U.S. Environmental Protection Agency water quality criterion continuous concentration (CCC). Because these statistics are generated from responses of individual organisms, the relationships of these statistics to significant effects at higher levels of ecological organization are unknown. This study was conducted to evaluate the quantitative relationships between toxicity test statistics and a concentration-based statistic derived from exposure–response models relating projected population growth rate to stressor con- centration. This statistic, C*, describes the concentration above which mysid populations are projected to decline in abundance as determined using population modeling techniques. An analysis of responses of A. bahia to 10 metals, nine organic compounds, and ammonia surprisingly indicated the acute LC50 to be the best predictor of C*, followed by the chronic value from life-cycle tests, which predicted population-level response almost equally as well. The chronic value for the 7-d test was less predictive of population-level effects. The CCC was lower than C* for 94% of the compounds evaluated, indicating the criterion value to be protective of population-level effects for A. bahia, as intended. Keywords—Americamysis Mysidopsis Population model Exposure–response models Toxicity tests INTRODUCTION The acute and chronic responses of the mysid shrimp Amer- icamysis bahia (formerly Mysidopsis bahia) are used by the U.S. Environmental Protection Agency (U.S. EPA) in a num- ber of regulatory and research programs addressing marine water quality [1–8]. Bioassays established with this species include a 96-h acute mortality test [9]; a 7-d survival, growth, and fecundity test; and a 28-d life-cycle test with survival, growth, and reproduction endpoints. The 7-d test was devel- oped as a rapid chronic test by the U.S. EPA to evaluate the toxicity of complex effluents and associated marine receiving waters for National Pollutant Discharge Elimination System permits [10]. The 28-d life-cycle test was designed to evaluate chronic toxicity through the first brood release of the life cycle of A. bahia [11]. The standard concentration-based endpoint statistics cal- culated from data generated in these assays include concen- trations lethal for 50% of the test organisms (LC50s) from the acute test and chronic values from both the 7-d and life-cycle tests [9–11] (Table 1). A chronic value is defined as the geo- metric mean of the no-observed-effect concentration (NOEC) and the lowest-observed-effect concentration (LOEC). These statistics are designed to indicate the potential of chemical stressors to elicit biological effects in natural systems. The acute and chronic data are also used to calculate criterion continuous concentrations (CCCs) in development of U.S. * To whom correspondence may be addressed (kuhn.anne@epa.gov). This manuscript has been reviewed by the U.S. Environmental Protection Agency, Atlantic Ecology Division, Narragansett, Rhode Island. Approval does not signify that the contents necessarily reflect the views and policies of the Agency. This paper is AED contribution number NHEERL-NAR-2037. EPA water quality criteria [12]. The water quality CCC is an estimate of the threshold concentration producing an unac- ceptable biological effect and is intended to provide protection to aquatic species from long-term chronic exposure to chemical stress. A long-standing interest has existed in the relation between laboratory toxicity tests results and risks to natural ecological systems [13–15] and in the statistical relations among end- points [16,17]. The central question is whether test endpoints that focus upon the responses of individuals can be extrapo- lated to evaluate the long-term risks of contamination at higher levels of ecological organization (populations, communities, and ecosystems). Several studies have been conducted to as- sess population-level effects of environmental contaminants on aquatic animals [18–38]. Unlike statistical analyses of in- dividual response endpoints from a typical bioassay, these studies provide evidence of contaminant effects on the de- mographic responses of populations. The primary objective of this study was to evaluate the ecological relevance of mysid bioassays by comparing statis- tics from standard toxicity tests with a statistic calculated from population modeling. This latter statistic, C*, is derived from exposure–response models relating population growth rate to contaminant concentration, and describes the concentration predicted to result in zero population growth. This calculation of a projected ‘‘threshold’’ concentration for mysid popula- tions is based on an assumption that a model-derived popu- lation-level statistic is more ecologically relevant than a sta- tistic based upon responses of individual mysids. The com- parison of C* with the CCC for a given contaminant also permits evaluation of the degree of protection afforded by the CCC. Another objective was to identify the 28-d chronic test