SIMPLE APPLICATION OF USING RESIDUALS FROM CATCH-CURVE REGRESSIONS TOASSESS YEAR-CLASS STRENGTH IN FISH

Residuals associated with catch-curve regressions can represent variab le recruitment in fish populations. Catch curves are used to estimate steady-state mortality and assume relatively constant recruitment, but this assumption is rarely met. I documented the presence of abundant year classes of largemouth bass (Micropterus salmoides), from earlier sampling and these dominant year classes persisted over time in two re servoirs. I expanded simple linear catch-curve regressions that used a ge (in years) as an independent regressor to multiple regression model s each of which incorporated an additional independent environmental v ariable (ENVIR) that was measured when fish were age 0. The age term i n the regression was proportionally weighted to the sample size at eac h age which deflated the influence of older and rarer fish in the anal ysis. This generalized regression equation: log(e) (NUMBER) = b(0)-b(1 ) (AGE) +/- b(2) (ENVIR); explained variable abundance-at-age (NUMBER) and the environmental term was related to the formation of weak and s trong year classes after accounting for the effects of age. Typically, age will explain the majority (r(2) greater than or equal to 0.5) of the variation in abundance-at-age. For two largemouth bass populations , environmental hydraulic variables were significant (P < 0.10) terms in this equation and explained an additional 12 and 16% of the variati on in number after accounting for the variation explained by age. For data collected in one population 2 yrs after the initial analysis, the same strong and weak year classes persisted, residuals from these cat ch curves were correlated (r = 0.86, P < 0.05, N = 6), and the influen ce of hydrology on year class formation was duplicated. This approach can provide savings in labor and funds as abundance of young fish or r ecruitment indices do not have to be measured each year. (C) 1997 Else vier Science B.V.