Butterflyfish community composition
The overall abundance of butterflyfishes did not change significantly over the five surveyed years (mixed effects model, fixed effect: year, random effect: site, χ2(3) = 2.35, P = 0.5, Figure 2a). There was a significant but weak effect of year on community structure (PERMANOVA F(3,73) = 1.7, R2 = 0.06, P = 0.03) accounting for 6% of the variation in composition. This result was driven by a significant difference between the 2016 and 2022 communities only (Figure 2c, pairwise adonis, F = 2.64, R2 = 0.07, P = 0.02). While these results show an overall shift in the butterflyfish community, this was driven by rare species. Therefore, we also tested whether the community shifted significantly when we included only the five most abundant species (C. citrinellus, C. lunulatus, C. vagabundus, C. unimaculatus, C. rafflesia ). These species represent at least 70% of the total butterflyfish abundance in each year (2016: 82%, 2017: 75%, 2018: 81%, 2022: 72%), include each feeding guild (facultative corallivore, obligate corallivore, and invertivore respectively), and are responsible for a majority of aggressive interactions. There was no significant difference in the abundance of these dominant species following the bleaching event, suggesting that while the overall community shifted slightly, the majority of species involved in aggressive encounters remained at the same level of abundance (PERMANOVA F(3,70) = 1.8, R2 = 0.06, P = 0.06).
The relative abundance of feeding guilds (facultative, obligate coral feeders, and invertivores) did not differ significantly between the four surveyed years (mixed effects model, fixed effects: year and feeding guild interaction; random effect: location, χ2(9) = 8.9, P = 0.4, Figure 2b).