Sensitivity of maximum sustainable harvest rates to intra-specific life history variability of lake trout (Salvelinus namaycush) and walleye (Sander vitreus)

Using lake trout (Salvelinus namaycush) and walleye (Sander vitreus) we examined the extent of error that is created when non-population-specific or non-sex-specific data are used to develop fishery sustainability models. To put biases in perspective we first compared relative differences in sustainable harvest rate estimates of fast and slow growing populations to the mean difference between species. Few estimates of early mortality exist for either species, and a sensitivity analysis indicated that sustainable harvest rate estimates varied by more than 90% with those available. To evaluate biases created by predicting life history traits we estimated maturation age from asymptotic length in populations whose life history parameters had been measured directly. Incorporating species-specific information produced maturation ages that were closer to those measured (mean difference 22%) than did the estimates from inter-specific relationships (mean difference 42%). This resulted in sustainability estimates from species-specific predictions being closer to those estimated using actual maturation data. In evaluating problems associated with using non-sex-specific data we found that estimates of maximum sustainable harvest rates increased when female-specific life history data were used (could only be done for walleye) instead of combined-sex data (mean change in the two populations was 7%). Analyses indicate that it is desirable to collect life history data on specific populations of lake trout and walleye. If this is not possible, estimates of maximum sustainable harvest rates are most influenced by variation in estimates of early mortality, followed by the taxonomic resolution used to estimate co-variation among life history traits, and finally the use of combined-sex versus female-specific data.