Context-Dependent Diversity-Effects of Seaweed Consumption on Coral Reefs in Kenya

Abstract

Consumers and prey diversity, their interactions, and subsequent effects on ecosystem function are important for ecological processes but not well understood in high diversity ecosystems such as coral reefs. Consequently, we tested the potential for diversity-effects with a series of surveys and experiments evaluating the influence of browsing herbivores on macroalgae in Kenya's fringing reef ecosystem. We surveyed sites and undertook experiments in reefs subject to three levels of human fishing influence: open access fished reefs, small and recently established community-managed marine reserves, and larger, older government-managed marine reserves. Older marine reserves had a greater overall diversity of herbivores and browsers but this was not clearly associated with reduced macroalgal diversity or abundance. Experiments studying succession on hard substrata also found no effects of consumer diversity. Instead, overall browser abundance of either sea urchins or fishes was correlated with declines in macroalgal cover. An exception was that the absence of a key fish browser genus, Naso, which was correlated with the persistence of Sargassum in a marine reserve. Algal selectivity assays showed that macroalgae were consumed at variable rates, a product of strong species-specific feeding and low overlap in the selectivity of browsing fishes. We conclude that the effects of browser and herbivore diversity are less than the influences of key species, whose impacts emerge in different contexts that are influenced by fisheries management. Consequently, identifying key herbivore species and managing to protect them may assist protecting reef functions.

Fig 1. Mean (± SE) (a) biomass of herbivorous fishes (kg ha-1), (b) biomass of sea urchins (kg ha-1), and (c) benthic composition (% cover).

Letters represent homogenous subgroups.

Fig 2. Non-metric multidimensional scaling analysis (MDS) of herbivorous fishes and sea urchin species abundance estimates at the six study sties.

Shapes represent fisheries management regime. Points that are closer together in ordination space are more similar in terms of community composition, and individual browsing fishes (black) and sea urchins (grey) are overlaid to visualize dominant species.

Fig 3. Relationships between (a) herbivore and macroalgal diversity (Shannon’s H) (x ± SE), (b) herbivore diversity and macroalgal abundance (% cover), (c) herbivorous fish biomass (kg ha^-1) and macroalgal abundance, and (d) sea urchin biomass and macroalgal abundance at each of the six study sites.

Shapes represent fisheries management at each site.

Fig 4. Time series of the mean abundance (% cover ± SE) of total macroalgae on the experimental substrata at the six study sites over the ~90 day experimental period, presented by the three fisheries management categories.

Algae grew on plates for over one year before removing cages and exposing them to herbivory. Note different scales on y-axes. Asterisks indicate significant differences in abundance from the beginning of the experiment to the final sampling event at each site (d.f. = 1). ***p<0.001, **p<0.01, *p<0.05.

Fig 5. Non-metric multidimensional scaling analysis (MDS) of successional trajectories showing similarity of algal communities in the six study sites on the experimental plates after cages were removed to the final sampling event (~90 days).

Fisheries management regimes indicated by different symbols next to the initial algal communities on day 0 and the final 90 day samples are the enlarged data points. Arrows indicate direction of change over time and are scaled to the magnitude of change in the algal community. Distance matrices were analyzed separately for each site using permutational analysis of variance (PERMANOVA; n = 999 permutations; Stress = 0.11).

Fig 6. Mass of macroalgae removed by herbivorous fishes (% 24h^-1; x ± SE, n = 13) for 6 taxa of algae used in selectivity assay experiments within the Mradi community closure.

Letters indicate homogeneous subgroups.

Fig 7. Bite rate of browsing fishes (h^-1; x ± SE, n = 13) on experimental selectivity assays of macroalgae in Mradi (in the channel) community closure.

Letters indicate homogeneous subgroups. Note scale differences between y-axes.