Harnessing fishery-independent indicators to aid management of data-poor fisheries: weighing habitat and fishing effects

State indicators, e.g., mean size and trophic level of the fish assemblage, can provide important insights into the effects of fishing on ecosystems and the resource potential of the fishery. On coral reefs, few studies have examined the relative effects of fishing and other drivers, such as habitat, on these indicators. In light of habitat heterogeneity and increasing habitat degradation, this lack of understanding limits the usefulness of indicators for monitoring the effect of fishery management actions. Identifying thresholds or nonlinearities in relationships between fishing pressure and state indicators has been suggested as a basis for biomass-based targets to support management efforts in low research capacity contexts. Using data collected in Seychelles, we examined the relative influences of fishable biomass (proxy for fishing pressure) and the benthic habitat on fisheries-independent indicators characterizing attributes of the fish community important for fisheries production. We characterized the driver-indicator relationships, and compared local-scale relationships for Seychelles with large-scale relationships published for the Indian Ocean. We found that both habitat and fishing pressure influenced indicators, but habitat effects were particularly strong. This knowledge provides managers with the capacity to implement a diverse array of complementary management actions targeting these drivers. A number of the Seychelles scale driver-indicator relationships were linear, suggesting gradual changes in indicators in response to changes to drivers. This contrasted with relationships published for the Indian Ocean, which were characterized by thresholds below which exploitation is likely to have significant detrimental effects on the functioning of important ecosystem processes. These scale-specific differences are likely driven by the narrower range of fishing pressures found in Seychelles. Importantly, it indicates that, although biomass-based targets derived from large-scale relationships may provide a useful starting point for setting management targets, the local context must be considered.