Reef fish assemblages are structured by many factors like depth, habitat and zonation’s in marine protected areas (MPAs). These fish assemblages can be determined by different sampling methods and design. For instance, stereo Baited Remote Underwater Video (stereo-BRUV) allows for monitoring fishes at various depths (>40m) beyond the reach of conventianol diver-based methodologies. This is relevant in a way that fish diversity, biomass and species richness changes with depth. Furthermore, the spatial distribution and composition in fish assemblages can have a strong correlation with fine-scale habitat differences. However, fine-scale measurerements of habitats are rarely available for marine research, particularly for deeper (>40m) marine environments. Stereo-BRUV studies lacking this information prior of sampling reef fish assemblages had to rely upon hierarchical classifications of habitat. One of the aims of this study was to examine which habitat classification was most sensitive to the detection of small-scale changes in fish assemblages by sampling different habitats and depths (e.g, 15, 50 and 100 meters) in the Saba Marine Park (SMP) area. We compared three categorization methods of habitat commonly used in literature 1) two scale on “habitat”; low (sand) and high (reef) relief (Colton et al. 2010); 2) three scale on “relief”; low, medium and high relief (Watson et al. 2005) and 3) fine-scale consisting of 6 types of “habitat complexity” (Polunin et al. 1993). The “habitat complexity” category method produced less variance among fish populations’ mean within each habitat. This increased the probability of finding a significant difference between two fish populations’ mean, as smaller variations lowered the possibility of overlapping standard deviations. For this reason we found a significant interaction effect of zonations (fishing vs. no-fishing zone) within the habitat complexity category on fish biomass, density and species richness, in all depths within the SMP boundary (15 and 50m). This effect was not found when using the habitat or relief category. However, the differences among the habitat category methods were less significant on the statistical power of detecting those changes in fish biomass, density and species richness. Overall, habitat characteristics, such as sand bottom or low complexity in substrate structure, were associated with lower values of fish biomass, density and species richness. These values increased gradually with habitats containing more complex, reef-based structures. From the shallow (15m) to deeper (50 and 100m) areas the habitat complexity in terms of reef structures significantly declined. Along this depth gradient, the structure of reef fish assemblages changed from higher densities of herbivorous species at 15 meter depth and higher carnivorous species richness and densities found at 50 and 100 meters depth. Some species (within families of Lutjanidae and Serranidae), also important to fisheries, were distributed over the full depth range. Moreover, stereo-BRUV detected high densities of larger predatory species (Carcharhinus perezii, Ginglymostoma cirratum), especially at depths of 50 meters. The changes found in fish assemblages were less determined by the effect of the no-fishing zone. On the contrary, the mean fish biomass and density were higher in the zone without protection from fisheries, indicating that fishing pressure was low in the SMP. In conclusion, depth and finer-scale habitat complexity were the main drivers that structure reef fish assemblages. These results indicate that the chosen categories of habitat and depth have a significant effect on studying reef fish assemblages across different zones in the SMP.
The parrotfish Sparisoma viride is an abundant and ecologically important member of the tropical NW Atlantic reef fish fauna. Sagittal otoliths of 417 individuals were analysed to estimate age-based demographic variables at 4 localities (Lee Stocking Island, Barbados, Los Roques Archipelago and the San Blas Archipelago) spanning 14° of latitude. The sampling localities ranged from an area protected from trap- and net-based reef fisheries (Los Roques) to an area supporting a dense human population and sustained trapping and spearing for reef fishes including S. viride (Barbados). Examination of sectioned sagittal otoliths from each locality revealed regular increments in the sagittal matrix. A preliminary validation at San Blas was consistent with these increments being annual check marks. These increments provided estimates of age structure, maximum longevities and mortality rates for the 4 study populations of S. viride. Von Bertalanffy growth functions fitted to each size-at-age plot generated similar growth curves from 3 of these 4 localities. The exception was Lee Stocking, where fish grew faster and reached a substantially larger size than those from the other 3 localities. Further analysis of the growth curves demonstrated that the differences between Lee Stocking and the other localities were attributable to more rapid growth over the first 4 yr of life. Age-based growth curves derived from the Los Roques population were very similar to a size-based curve generated by an independent study on S. viride carried out in Bonaire, adjacent to Los Roques. Maximum longevities for all 4 of our populations varied from 7 to 9 yr. Mortality rates generated from catch curve analysis were also similar among localities and suggest that maximum life spans do not exceed 12 yr. This result differs from that obtained at Bonaire, where repeated censuses of tagged fish suggest 30 yr maximum longevity. Abundances of S. viride varied 3-fold among localities, being highest at Los Roques (protected from reef fishing), lowest at Barbados (high fishing) and Lee Stocking (low fishing). Thus our age-based study suggests that S. viride is a relatively short-lived fish with consistent demographic parameters over a range of localities, latitudes and fishing intensities.