Comparative life histories and stock assessments of rockcods (family Serranidae) from the east coast of South Africa.
The family Serranidae is a diverse group of fishes, of which the genus Epinephelus (rockcods or groupers) is the largest. Serranids are commonly caught in reef fisheries in tropical and warm-temperate latitudes, and are targeted because of their tasty flesh and high value. In South Africa, epinepheline serranids mainly occur in hook and line fisheries in the province of KwaZulu-Natal. Mostserranids are caught by the boat-based (skiboat) fishery, and the commonestspecies are the endemic catface rockcod (Epinephelus andersoni), thehalfmoon rockcod (E. rivulatus), the yellowbelly rockcod (E. marginatus) and the endemic white-edge rockcod (E. albomarginatus). Although serranids contribute about ten percent to catches by the commercial and recreational skiboat sectors in KwaZulu-Natal, representing an estimated total catch of around 200 mt per year, little is known about these fishes in South Africa. From the mid-1980s to the mid-1990s, the mean lengths of E. marginatus and E. albomarginatus in the region declined significantly. Over this period, lengths of E. andersoni remained the same, while those of E. rivulatus increased. Lengths of E. marginatus and E. albomarginatus from Mozambique, where fishing effort was low at the time of sampling, were significantly greater than in KwaZulu-Natal. Monthly biological data were mostly collected from commercial skiboat catches on the northern and southern coast of KwaZulu-Natal. Additional data for E. marginatus and E. albomarginatus were also collected irregularly from commercial catches made in Mozambique. Unless the fish had ripe ovaries, all gonads had to be sectioned to establish sex and stage. Histology revealed that all gonads had a female-like appearance, with lamellae and a central lumen. In E. andersoni, there was a complete overlap of male and female length frequencies, and their meanlengths were not significantly different. Some males and inactive bisexuals were both smaller and younger than the female size and age at first maturity. Together with the occurrence of mature bisexual fish (transitionals), these observations indicate that males are derived from immature or mature females, hence this species is a diandric protogynous hermaphrodite. The other three species exhibit typical signs of monandric protogynous hermaphroditism. Males and females had significantly different mean lengths, and age and length frequencies by sex werebimodal. Transitional individuals were recorded in E. rivulatus. E. andersoni and E. rivulatus matured at small sizes and early ages relative to E. marginatus and E. albomarginatus. Ripe ovaries were much larger than ripe testes in all four species. E. andersoni, E. marginatus and E. albomarginatus spawned in spring and summer, while E. rivulatus spawned in winter and spring. There were no indications of spawning in E. andersoni in the southern sampling region, and few ripe individuals of E. albomarginatus were encountered in KwaZulu-Natal samples. Size at maturity of this species was much smaller in Mozambique samples. Large, reproductively inactive individuals of E. andersoni were frequently observed in the spawning season. The lack of reproductive activity of E. andersoni and E. albomarginatus in KwaZulu-Natal may be because this area represents thesouthernmost limit of the distribution of these species. Ageing of the four species was undertaken using sectioned otoliths. Age validation was undertaken by a combination of tetracycline marking in captive fishes, and analysis of the marginal zone in otoliths. All four species are relatively long-lived, although estimates of maximum age may be under-estimated because of long-term harvesting. In all four species, fish from the southern sampling region were larger than fish from the northern region at the same age. Only in the case of E. rivulatus were these significant enough to warrant the fitting of two growth curves to the northern and southern populations. Males in all four species tended to be larger than females at the same age, suggesting that there may be a growth spurt following sex change, or that faster-growing females changed sex. A logistic growth curve was fitted to the age-length data for E. andersoni, while von Bertalanffy curves produced the best fit for the other species. Based on the rates at which L∞ attained in these four species, E. marginatus and E. albomarginatus are slow-growing species, while E. andersoni and, particularly, E. rivulatus arefaster growing. Rates of total mortality and natural mortality were estimated using length-converted catch curves and the Rikhter and Efanov equation, respectively. Stock assessments undertaken by yield per recruit and spawner biomass per recruit analyses indicate that E. andersoni in KwaZulu-Natal is currently optimally exploited, while E. rivulatus is lightly exploited. Both E. marginatus and E. albomarginatus are over-exploited. The potential problems in applying standard per recruit models to species with complex life histories are discussed. Support for the reduced stock status of the latter two species is provided by the observed changes in lengths of these species over a ten-year period, and their relatively small size in KwaZulu-Natal compared to the lightly-fished Mozambique populations. Local fishers in KwaZulu-Natal have also reported declines in sizes and reduced catches of these two species.The life history styles and other features of the four species are compared and discussed with reference to the resilience of these species to harvesting. Two of the species (E. marginatus and E. albomarginatus) are monandric protogynous hermaphrodites, which grow slowly, mature late and attain large sizes. E.andersoni and E. rivulatus grow faster, mature earlier and are smaller species. The normally deleterious effects of fishing on sex-changing species are not manifested in these two species, possibly because E. rivulatus is so small, that males are not selectively removed. In contrast, E. andersoni is a diandric protogynous hermaphrodite, and hence, does not rely on sex-change as a source of males. The current management methods for serranids in KwaZulu-Natal are presented, and suggestions for future approaches are discussed.