A Comparative Analysis of Surplus Production and Analytical Models for Assessing Kapenta (Limnothrissa miodon) Stock in Lake Kariba

Abstract

Surplus production (SP) models, a Maximum Entropy model (ME) and an Analytical Model are analyzed for fishery stock assessment for Kapenta (Limnothrissa miodon) in Lake Kariba. Five traditional surplus production models; Schaefer, Schnute, Walters and Hilborn, Fox and Clarke, Yoshimoto and Pooley (CYP) models are tested for effort and catch data of Kapenta which occupies about 80% of fish landings in Lake Kariba. The CYP model is the only model that has a high goodness of fit and is statistically significant but the model showed a lower Mean Square Error than the Walters and Hilborn model. The ME model was a better estimate of catch than the Walters and Hilborn Model and estimated annual biomass for the fishery. The Analytical Model showed a more conservative Allowable Biological Catch (ABC) than the ME model and the estimated fishing mortality (F) for 2009 to be lower than that of 1982 as well as the fishing mortality at ABC suggesting that there is a small room to increase fishing intensity. Both the Analytical Model and the ME model estimated comparable estimates for 1982 stock biomass.