Search

Trophic cascades caused by a reduction in predators of sea urchins have been reported in Indian Ocean and Caribbean coral reefs. Previous studies have been constrained by their site-specific nature and limited spatial replication, which has produced site and species-specific understanding that can potentially preclude larger community-organization nuances and generalizations. In this study, we aimed to evaluate the extent and variability of the cascade community in response to fishing across ~23° of latitude and longitude in coral reefs in the southwestern Indian Ocean. The taxonomic composition of predators of sea urchins, the sea urchin community itself, and potential effects of changing grazer abundance on the calcifying benthic organisms were studied in 171 unique coral reef sites. We found that geography and habitat were less important than the predator-prey relationships. There were seven sea urchin community clusters that aligned with a gradient of declining fishable biomass and the abundance of a key predator, the orange-lined triggerfish (Balistapus undulatus). The orange-lined triggerfish dominated where sea urchin numbers and diversity were low but the relative abundance of wrasses and emperors increased where sea urchin numbers were high. Two-thirds of the study sites had high sea urchin biomass (>2,300 kg/ha) and could be dominated by four different sea urchin species, Echinothrix diadema, Diadema savignyi, D. setosum, and Echinometra mathaei, depending on the community of sea urchin predators, geographic location, and water depth. One-third of the sites had low sea urchin biomass and diversity and were typified by high fish biomass, predators of sea urchins, and herbivore abundance, representing lightly fished communities with generally higher cover of calcifying algae. Calcifying algal cover was associated with low urchin abundance where as noncalcifying fleshy algal cover was not clearly associated with herbivore abundance. Fishing of the orange-lined triggerfish, an uncommon, slow-growing by-catch species with little monetary value drives the cascade and other predators appear unable to replace its ecological role in the presence of fishing. This suggests that restrictions on the catch of this species could increase the calcification service of coral reefs on a broad scale., (© 2016 by the Ecological Society of America.)

Removal of predators can have strong indirect effects on primary producers through trophic cascades. Crustose coralline algae (CCA) are major primary producers worldwide that may be influenced by predator removal through changes in grazer composition and biomass. CCA have been most widely studied in Caribbean and temperate reefs, where cover increases with increasing grazer biomass due to removal of competitive fleshy algae. However, each of these systems has one dominant grazer type, herbivorous fishes or sea urchins, which may not be functionally equivalent. Where fishes and sea urchins co-occur, fishing can result in a phase shift in the grazing community with subsequent effects on CCA and other substrata. Kenyan reefs have herbivorous fishes and sea urchins, providing an opportunity to determine the relative impacts of each grazer type and evaluate potential human-induced trophic cascades. We hypothesized that fish benefit CCA, abundant sea urchins erode CCA, and that fishing indirectly reduces CCA cover by removing sea urchin predators. We used closures and fished reefs as a large-scale, long-term natural experiment to assess how fishing and resultant changes in communities affect CCA abundance. We used a short-term caging experiment to directly test the effects of grazing on CCA. CCA cover declined with increasing fish and sea urchin abundance, but the negative impact of sea urchin grazing was much stronger than that of fishes. Abundant sea urchins reduced the CCA growth rate to almost zero and prevented CCA accumulation. A warming event (El Niño Southern Oscillation, ENSO) occurred during the 18-year study and had a strong but short-term positive effect on CCA cover. However, the effect of the ENSO on CCA was lower in magnitude than the effect of sea urchin grazing. We compare our results with worldwide literature on bioerosion by fishes and sea urchins. Grazer influence depends on whether benefits of fleshy algae removal outweigh costs of grazer-induced bioerosion. However, the cost-benefit ratio for CCA appears to change with grazer type, grazer abundance, and environment. In Kenya, predator removal leads to a trophic cascade that is expected to reduce net calcification of reefs and therefore reduce reef stability, growth, and resilience.

This study examined trends along a gradient of fishing intensity in an artisanal coral reef fishery over a 10-year period along 75 km of Kenya's most populated coastline. As predicted by Malthusian scenarios, catch per unit effort (CPUE), mean trophic level, the functional diversity of fished taxa, and the diversity of gear declined, while total annual catch and catch variability increased along the fishing pressure gradient. The fishery was able to sustain high (approximately 16 Mg x km(-2) x yr(-1)) but variable yields at high fishing pressure due to the dominance of a few productive herbivorous fish species in the catch. The effect of two separate management strategies to overcome this Malthusian pattern was investigated: fisheries area closure and elimination of the dominant and most "competitive" gear. We found that sites within 5 km of the enforced closure showed significantly lower total catch and CPUE, but increased yield stability and trophic level of catch than predicted by regression models normalized for fishing effort. Sites that had excluded illegal beach seine use through active gear management exhibited increased total catch and CPUE. There was a strong interaction between closure and gear management, which indicates that, for closures to be effective at increasing catch, there must be simultaneous efforts at gear management around the periphery of the closures. We propose that Malthusian effects are responsible for the variation in gear and catch and that active management through reduced effort and reductions in the most competitive gear have the greatest potential to increase the functional and trophic diversity and per-person productivity.

Identifying the rates of recovery of fish in no-take areas is fundamental to designing protected area networks, managing fisheries, estimating yields, identifying ecological interactions, and informing stakeholders about the outcomes of this management. Here we study the recovery of coral reef fishes through 37 years of protection using a space-for-time chronosequence of four marine national parks in Kenya. Using AIC model selection techniques, we assessed recovery trends using five ecologically meaningful production models: asymptotic, Ricker, logistic, linear, and exponential. There were clear recovery trends with time for species richness, total and size class density, and wet masses at the level of the taxonomic family. Species richness recovered rapidly to an asymptote at 10 years. The two main herbivorous families displayed differing responses to protection, scarids recovering rapidly, but then exhibiting some decline while acanthurids recovered more slowly and steadily throughout the study. Recovery of the two invertebrate-eating groups suggested competitive interactions over resources, with the labrids recovering more rapidly before a decline and the balistids demonstrating a slower logistic recovery. Remaining families displayed differing trends with time, with a general pattern of decline in smaller size classes or small-bodied species after an initial recovery, which suggests that some species- and size-related competitive and predatory control occurs in older closures. There appears to be an ecological succession of dominance with an initial rapid rise in labrids and scarids, followed by a slower rise in balistids and acanthurids, an associated decline in sea urchins, and an ultimate dominance in calcifying algae. Our results indicate that the unfished "equilibrium" biomass of the fish assemblage > 10 cm is 1100-1200 kg/ha, but these small parks (< 10 km2) are likely to underestimate pre-human influence values due to edge effects and the rarity of taxa with large area requirement, such as apex predators, including sharks.