THE DYNAMICS OF EXPLOITED GROUPERS (SERRANIDAE): AN INVESTIGATION OF THE PROTOGYNOUS HERMAPHRODITIC REPRODUCTIVE STRATEGY

Groupers (Serranidae) are an important target of reef fisheries throughout tropical and subtropical regions of the world. Two potential consequences for management of the biological environment of groupers are potential effects of: (1) interspecific interaction and (2) protogynous hermaphroditism on population dynamics. The latter problem was investigated through simulation. Results indicated that protogynous hermaphroditism may be a superior reproductive strategy to gonochorism if sperm is not limiting in the population. If high proportions of eggs go unfertilized as fishing pressure increases, reproductive success of protogynous populations may become less than their gonochoristic counterparts.;Similarly fitting the production model to species-aggregated yield data from a simulated multispecies grouper fishery resulted in accurate estimation of Y(,max). Estimated f(,opt), however, was influenced heavily by composition of the catch. The advised f(,opt) resulted in under- and overfishing of less dominant species in the complex. The approach was applied to empirical data from a multispecies grouper fishery in Puerto Rico. A model stimulating interspecific interaction is suggested.;The utility of management advice from yield per recruit and stock production models applied to populations with the parameters of groupers was investigated. Potential reproductive strategy-specific errors were identified in yield-per-recruit (YPR) models. Also, the Beverton and Holt model overestimated the F resulting in maximum YPR. Management based on these predictions is subject to type I (non-conservative) error for controlling fishing mortality and type II (conservative) error from the perspective of minimum size regulations. Production models overestimated sustainable effort for both protogynous and gonochoristic populations, but the error was relatively greater for sperm-limited hermaphrodites. Thus management error based on these predictions is type I. Predictions of effort required to harvest maximum yield (f(,opt)), maximum yield (Y(,max)), and maximum yield per effort (U(,max)) from equilibrium approximation fits of the generalized production model were accurate.