Recent developments in estimating fishing and natural mortality and tag reporting rate of lobsters using multi-year tagging models

Abstract

From May 1992 to May 1995, tagging of southern rock lobster, Jasus edwardsii, was undertaken in May, February and September in northwestern Tasmania, Australia. Previous modelling of these data using multi-year tagging models produced estimates of annual fishing mortality and tag reporting rates with low relative standard errors and a very low natural mortality estimate (0.0) with high relative standard error. In this model, natural mortality was partitioned within a year by the duration of the time periods and fishing mortality was partitioned by the proportion of fishing effort expended during each period. We altered the model to take advantage of quantitative information on the amount of fishing effort. In the new models, fishing mortality ( F) was replaced by catchability ( q) multiplied by fishing effort ( e) (i.e., F= qe). The most parsimonious model was based on constant catchability over years with unequal catchability among periods within a year. A model with separate catchabilities for each sex did not improve the model. Annual fishing mortality and tag reporting rate estimates had low relative standard errors and were similar to the previous model. The natural mortality estimate (0.12 per year) was higher and more precise than in the previous model and similar to previously published values. Based on Akaike’s information criterion (AIC), we found the model based on effort (i.e., F= qe) to be superior to previous models that used effort information merely to apportion fishing mortality to periods of the year. Comparison of the residuals between the two models demonstrated that the inclusion of effort in the model reduced problems with patterns in the residuals.