Influence of ecological factors on the relationship between mfo induction and fish growth: Bridging the gap using neural networks

Relationships between molecular and biochemical effects of contaminant exposure and organism-level effects are difficult to establish primarily because: (1) the large differences in time-scales of the responses; (2) the complexity of mechanisms involved in such relationships; and (3) the influence of ecological factors on the expression of higher-level responses to stress. In this study we investigated the relationship between mixed function oxidase (MFO) induction and growth in fish as influenced by principal ecological factors in the environment, such as habitat quality, competition and food availability. Because of the complexity and non-linearity of interactions among these factors, we applied a back-propagation neural network approach to investigate this relationship. Inputs into the network included an indicator of contaminant exposure 7-ethoxyresorufin O-deethylase, EROD) and several ecological variables, while growth rate served as the single output variable. Under the field conditions and contaminant exposure regimes observed in this study, EROD in the model simulation contributed minimal marginal ability to predict growth. Indirect effects of contaminants on growth could occur, however, through influences on food chain processes or behavioral dynamics of fish populations.