MICHAEL D. MURPHY, Florida Marine Research Institute, Florida Department of Environmental Protection, 100 Eighth Avenue SE, St. Petersburg, Florida 33701-5095, USA

When age-frequency data are insufficient for fisheries scientists to estimate year- or age-specific mortality, they often pool those data to provide a single estimate of mortality for all fully recruited age groups. The accuracy of a pooled estimate depends largely on whether or not the sampled population is in a steady state, i.e., a state in which the rates of recruitment and mortality are relatively constant. Although both the maximum likelihood (ML) and least square (LS) estimators can provide very accurate estimates of mortality for exact steady-state age frequencies, the effect of random variation within the age frequency has not been investigated. In this study, I evaluated the effect of sample size and mortality rate on the accuracy and precision of Chapman-Robson maximum likelihood estimators and the least squares "catch curve" estimator. Using simulation modeling, the LS estimator is negatively biased, especially when small, random samples are drawn from a steady-state population whereas the ML estimators are not. If the sample age-structure is truncated using some minimum abundance criterion, bias in the LS estimator is reduced. Application of the two estimation techniques to preliminary age composition data for adult red drum Sciaenops ocellatus from the eastern Gulf of Mexico appear to confirm the findings of the simulations.