Foraging theory posits that isolation from refuge habitat within a landscape increases perceived predation risk and, thus, suppresses the foraging behavior of prey species. However, these effects may depend fundamentally on resource availability, which could affect prey boldness and can change considerably through bottom-up processes. We conducted a field survey and experiment in a coral reef to test the effects of isolation from refuge habitat (i.e., reef structure) on herbivory by reef fishes and whether these effects depend on resource density. By fitting continuous-time, pure death Markov processes to our data, we found that at both the local and landscape scale distance from refuge habitat reduced herbivory in attractive resource patches of palatable benthic algae. However, our field experiment revealed that higher initial resource densities weakened negative effects of distance from refuge habitat on herbivory. Furthermore, we observed higher bite rates and greater total lengths of herbivorous fishes with greater distance from refuge habitat-responses consistent with higher perceived predation risk. Our results suggest that while the loss or fragmentation of refuge habitat reduces consumer control of resources, greater resource densities can partially counteract this effect by altering landscapes of fear of consumer species. Our findings emphasize the importance of considering the spatial context of species interactions that structure communities.

Six new records of reef fish and a hybrid species are reported from Rocas Atoll, off Rio Grande do Norte State, northeast Brazil. Data on species abundance, distribution of the recorded species in Brazilian oceanic islands, trophic structure, and the occurrence of the hybrid species “Menephorus dubius” are provided.

Environmental stressors often interact, but most studies of multiple stressors have focused on combinations of abiotic stressors. Here we examined the potential interaction between a biotic stressor, the vermetid snail Ceraesignum maximum , and an abiotic stressor, high sedimentation, on the growth of reef-building corals. In a field experiment, we subjected juvenile massive Porites corals to four treatments: (i) neither stressor, (ii) sedimentation, (iii) vermetids or (iv) both stressors. Unexpectedly, we found no effect of either stressor in isolation, but a significant decrease in coral growth in the presence of both stressors. Additionally, seven times more sediment remained on corals in the presence (versus absence) of vermetids, likely owing to adhesion of sediments to corals via vermetid mucus. Thus, vermetid snails and high sedimentation can interact to drive deleterious effects on reef-building corals. More generally, our study illustrates that environmental factors can combine to have negative interactive effects even when individual effects are not detectable. Such ‘ecological surprises' may be easily overlooked, leading to environmental degradation that cannot be anticipated through the study of isolated factors.

Death and life: Muricid snails consume the vermetid gastropod, Dendropoma maximum, and use empty shells for reproduction Dendropoma maximum, a large sessile vermetid gastropod, is abundant on shallow coral reefs and can reduce the growth and survival of corals (Shima et al. 2010). Despite high densities and detrimental effects, little is known about vermetid ecology, particularly their interactions with predators. In the austral winter, we observed two species of muricid gastropods, Mancinella armigera (Link, 1807) and Menathais tuberosa (Röd-ing, 1798; Fig. 1a) feeding on the vermetid gastropod, D. maximum, in the shallow lagoon of Moorea, French Polynesia (17.28°525-645¢S, 149.47°335-535¢W). Until now, vermetid predation had not been documented in French Polynesia, although M. amigera had been reported to consume vermetids in other areas of the South Pacific (Taylor 1976). On five occasions, we observed M. armigera with its proboscis inserted into a shell of D. maximum. When the muricid was removed, a blue substance was observed in the empty vermetid shell tube (Fig. 1b), along with the operculum of the D. maximum. In addition to these field observations, we placed live vermetids and muricids in lab aquaria together and observed an additional five predation events. In the field, we found a greater number of D. maximum recently killed (Fig. 1c) within 1 m of a M. amigera than in nearby (within 3 m) control plots lacking M. amigera (t 4 = -3.06, p = 0.02, 4 ± 2.3; vs. 0.2 ± 0.4, mean ± 95 % confidence interval, n = 5), suggesting that the predator may represent a significant source of mortality for vermetids.During our surveys, we also discovered egg masses from muricids deposited into vermetid shells (Fig. 1d). In one instance, we observed recently deposited egg cases when removing a M. armigera from atop a D. maximum. In all other instances, the egg cases were found near pairs of M. armigera, suggesting that those snails deposited the eggs. The use of shells of recently killed vermetids for reproduction has not been reported previously. This interaction represents a novel use of a prey resource, in which the consumption of a prey also facilitates the predator's reproduction by creating a new microhabitat for eggs.Acknowledgments We thank G. Paulay for helpful discussion, the Richard B. Gump Field Station for facilitating our work, and the National Science Foundation for funding (OCE-1130359). References