RESUMO
Worldwide, many coral reefs are at risk of shifting to degraded algal-dominated states, due to compromised ecological conditions. Functional diversity of herbivorous fishes maintains coral reef health and promotes reef resilience to disturbances. Given previous evidence, it appears the functional roles of herbivorous fishes differ across geographical locations, indicating a need for further assessment of macroalgal consumption by herbivorous fishes. We assessed functional diversity by examining foraging behavior of herbivorous fish species on macroalgae on a fringing reef in Moorea, French Polynesia. We video-recorded choice experiments containing seven common macroalgae and used Strauss' linear resource selection index to determine macroalgal selectivities. We used cluster analysis to identify any distinct groups within herbivorous fish species, given the macroalgal species they targeted, and fitted generalized linear mixed-effects models to identify factors that best predicted the number of bites taken on macroalgae. Seven species from 3 fish families/tribes took a total of 956 bites. Fish species differed in their selectivity with some species (Naso lituratus, N. unicornis, Calatomus carolinus) strongly preferring one or two macroalgal species, while other fish species (Acanthurus nigrofuscus, Ctenochaetus striatus, Chlorurus sordidus, Balistapus undulatus) were less selective. This resulted in fish species forming two clusters. Only 3 of 7 macroalgae were preferred by any fish species, with two fish species both preferring the same two macroalgae. The limited differences in fish species' preferences for different macroalgae suggests limited functional complementarity. Two models (macroalgal species identity+fish functional group, macroalgal species identity+fish species) best predicted the number of bites taken on macroalgae compared to models incorporating only a single explanatory factor or fish family. In the context of this Moorean fringing reef, there is greater functional redundancy than complementarity of herbivorous fishes consuming macroalgae, and the fishes grouped together according to their relative selectivity. We observed fish species that are not classified as browsers consuming macroalgae, suggesting diets of herbivorous fishes may be broader than previously thought. Finally, we observed macroalgal selectivities and consumption that differed from previous studies for the same fish species. Our results contribute to the understanding of functional diversity of herbivorous fishes across coral reefs, and also highlight the need for additional research to further elucidate the role of context and functional diversity of herbivorous fishes consuming macroalgae on coral reefs.
RESUMO
Herbivory assays are a valuable tool used by ecologists to understand many of the patterns and processes affecting herbivory, a widely recognized driving force in marine communities. However, methods vary substantially among studies in both design and operation, and the effect of these differences has yet to be evaluated. We assessed the effects of several key components of assay design on estimates of herbivory to offer four recommendations. First, we found assays out-planted on sequential days in both predictable and random locations within a 60m2 site experienced temporal increases in herbivory by an increasingly diverse assemblage of fishes. Thus, we strongly advise against placing herbivory assays in the same site over a series of days. Second, we found while the amount of biomass consumed in assays was density dependent, the percent loss was not. Thus, we recommend researchers report percent consumption because this metric is robust to differences in biomass offered and will facilitate comparisons across studies. Third, we found associational effects, where proximity of species of differing palatabilities impacted estimates of herbivory rate on one or both species, but these impacts were not consistent across species or sites. Thus, we recommend the effect of association be directly tested for multi specie herbivory assays. Fourth, we found no effect of attachment method on estimates of herbivory rate and recommend researchers continue to use the attachment method in which they are most confident. We hope our experimental results prove useful in the future when designing, conducting, and interpreting herbivory assays.
