Corallivore assemblage, feeding behavior, and cover of Acropora corals drive fish corallivory on tropical reefs

Le résumé fourni par la source

Abstract Understanding spatial patterns and drivers of corallivory is essential for assessing reef resilience and identifying recovery pathways. Fish corallivory patterns arise from the complex direct and indirect effects of key factors, including food availability (cover of preferred corals), the nature of the corallivore assemblage (richness and biomass), how the corallivores choose to feed (intensity and bout duration), and the local risk profile (predator biomass and habitat structural complexity). While some linkages within this network are well studied, such as the generally positive relationship between coral cover and the abundance of corallivores, the manner and extent to which they can explain overall corallivory rates is less clear. In this study, we used structural equation modeling to provide an integrated assessment of the relative importance of these key elements in both directly and indirectly explaining patterns of total corallivory by fishes (bites in 30 min). We used remote underwater video sampling at four reefs spanning approximately 1800 km across the Torres Strait (TS) and Great Barrier Reef (GBR) in northeast Australia to quantify corallivory. Total corallivory and corallivore biomass were significantly higher in back‐ than fore‐reef habitats, while neither corallivore feeding behavior (intensity or duration) nor richness varied across spatial scales. Due to distinct benthic and corallivore assemblages, separate analyses were conducted for GBR and TS sites. Structural equation modeling showed that among GBR sites, corallivore richness, feeding intensity, and event duration were direct positive drivers of corallivory, while the cover of Acropora corals was both a direct and indirect (via its effect on corallivore richness, feeding intensity, and event duration) positive driver of corallivory. In contrast, TS sites, dominated by large massive corals and near‐monospecific corallivore assemblages, showed no relationship between corallivory and any assemblage or behavioral metrics. Our findings showed little evidence in either area that predation risk (biomass of piscivores) affected corallivory either directly or through changes in feeding behavior.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.