Accounting for effort when comparing tropical fisheries in lakes, river-flood-plains, and lagoons

Data from 59 tropical fisheries which were artisanal and multispecies in character indicated that fishing effort was a major determinant of yield. Effort was defined as number of full‐ and part‐time fishermen. The largest subset, which comprised 31 African lake fisheries, conformed to a function related to the Schaefer model. Data from different lakes were thus treated as samples from a hypothetical multispecies fishery. The model explained 75% of the variance in log(yield). The residuals were correlated ( P < 0.05) with log(morphoedaphic index). The effort model and the morphoedaphic index jointly explained 80% of the variance in log(yield). A function to account for the nonlinear effect of fishing effort was proposed for sets of fisheries conforming to the model. Data from river‐floodplain and lagoon fisheries were very similar and were best described by the same effort model. Differences in the model parameters for these fisheries compared to those of lakes suggested differences in effective effort. A “precision improvement index” was proposed to compare the predictive value of models with different predictors and transformations. A comparison of the predictability of yields revealed that, despite variation in socioeconomics, commercialization, and biological productivity between fisheries, this crude effort measurement was more important than indices of productivity for these sets of fisheries. Future progress in predicting and understanding fisheries yields depends on the synchronized collection of variables that relate to the fishery and to biological productivity in numerous comparable systems.