COS 123-5 - Reduction in top-down control results in dissimilarity in community composition and alternative community paths leading to coral- or algal-dominance on a coral reef

Abstract: Top-down (e.g. herbivory and predation) and bottom-up (e.g. nutrients) processes both structure ecological communities, but their relative importance in driving the community assembly of primary producers is debated. In many marine systems, severe disturbances such as marine heatwaves and cyclones can trigger mass die-offs of primary producers. Many of these ecosystems are also experiencing human-induced changes to top-down and bottom-up forcing through fisheries and nutrient pollution, thus, making community re-assembly after disturbance unpredictable. Here we experimentally test the relative importance of top-down and bottom-up processes in driving community development after a simulated disturbance on a coral reef using a factorial experiment that manipulated herbivory and nutrient enrichment for four years.





We conducted our experiment on a coral reef in Moorea, French Polynesia starting in 2018, at which time corals and macroalgae accounted for ~70% and ~2% of benthic cover, respectively. Our experiment consisted of eight 30 m²plots in which we removed all live corals and fleshy macroalgae to mimic the effects of a severe cyclone. Within each plot, we then nested 4 different herbivore exclosures (~1 m² each) with different size openings that allowed different size herbivorous fishes access to the benthos, creating four levels of herbivory (very low, low, medium, high). Each of the 30 m² plots was then assigned to either ambient or enriched nutrient conditions. Thus, each combination of herbivory and nutrients had n=4 for replication. To monitor community development, we photographed plots 3x annually and created high-resolution digital photomosaics that we then used to quantify the abundance of benthic organisms.




Overall, herbivory strongly affected community development, whereas nutrients had a weaker, but significant, effect. After four years, herbivory was inversely related to beta-diversity, where plots with high herbivory had low beta-diversity and plots with low levels of herbivory had high beta-diversity. This dissimilarity in community composition in low- and very low- herbivory plots was the result of plots becoming rapidly dominated by either coral or macroalgae. In contrast, community development in plots with higher levels of herbivory was slow but predictable, with all plots slowly increasing in coral cover and remaining relatively free of macroalgae. Importantly, this result suggests that reduced top-down pressure can result in alternate community paths that lead to either coral- or algal- dominance under the same set of environmental conditions. We conclude that herbivory strongly influences both the rate and predictability of community development in these ecosystems.