Effects of Density-Dependent and Climatic Factors on American shad, Alosa sapidissima, Recruitment: A Predictive Approach

Abstract

We developed environment-dependent stock–recruitment models for American shad, Alosa sapidissima, in the Connecticut River to forecast recruitment variability and measure density-dependent effects. These models were fitted to the 1966–78 stock–recruitment estimates and to May and June river flow, water temperature, and rainfall data shown previously to affect American shad year-class strength. We also attempted to validate the models by forecasting the 1979–84 year-classes based on juvenile indices, parent stock size, and hydrographic data for these years. The stock–recruitment models without environmental factors explained less than 3% of the recruitment variability, and none of the density-dependent exponents were statistically different from 0. The predictive capability of the Ricker stock–recruitment model improved dramatically (r² = 0.90) when combined with mean May flows, June flows, and the number of American shad lifted over the Holyoke Dam. The density-dependent exponents of these multiple regression models were highly significant, indicating that density-dependent processes are hidden by climatically induced variability in recruitment. Two environment-dependent stock–recruitment models predicted 80–90% of the American shad recruitment variability from 1979–84.