Reef sharks are experiencing rapid declines in abundance due to overfishing. Recent studies suggest that a high biomass of sharks promotes a trophic structure within a fish community very different to that on a reef where sharks are in low numbers. Although this implies that reef sharks have an important top-down role in structuring communities, declines in sharks can occur concomitantly with other anthropogenic (habitat loss) and natural processes (hurricanes, coral bleaching, disease, etc). Thus, the effects of shark removal may be confounded with other natural events that restructure fish communities from a top-down or bottom-up perspective.
Here, we use the results of two monitoring programs, long-term (decadal) and large-scale (100s-1000s km), to disentangle the relative effects of top-down and bottom-up processes in structuring fish communities on coral reefs throughout the Indo-Pacific. Firstly, we compared the trophic structure of fish communities at isolated atolls (i.e. local spatial scale) on the north-west coast of Western Australia using a 16-year data set of species abundances and benthic habitat information. We then used a dataset of fish abundances and benthic habitat information recorded from coral reefs of 17 countries (i.e. broad spatial scale) across the south-central Pacific (n= 646).
Results/Conclusions
For the local scale comparison, one set of these atoll reefs is subject to a small-scale fishery that targets sharks (Scott Reef) and the other is protected from all fishing (Rowley Shoals). We demonstrate that shark abundance contributes to the structure of reef fish communities by driving mesopredator release and a trophic cascade at fished reefs. Further, we show that these community responses are a function of a synergistic response to changes in both top-down and bottom-up processes.
Using geographically-weighted regression (for our broad spatial scale study) we demonstrate that the distribution of top-order predators was negatively related to human activity and that this relationship (amongst a suite of habitat, demographic and environmental variables) was spatially variable. Using k-means clustering, we reveal that there are distinct regions where human activity has a similar impact on reef shark distribution. Finally, within these regions we investigate the significance and contribution of covariates to shark distribution and investigate their subsequent impact on the reef fish community using structural equation modeling. Results suggest that there is an evident interplay between sharks and human activity that impacts reef fish community structure throughout the Pacific.