Prey state and prey density mediate antipredator responses that can shift community structure and alter ecosystem processes. For example, well-nourished prey at low densities (i.e., prey with higher per capita predation risk) should respond strongly to predators. Although prey state and density often co-vary across habitats, it is unclear if prey responses to predator cues are habitat-specific. We used mesocosms to compare the habitat-specific responses of purple sea urchins (Strongylocentrotus purpuratus) to waterborne cues from predatory lobsters (Panulirus interruptus). We predicted that urchins from kelp forests (i.e., in well-nourished condition) tested at low densities typically observed in this habitat would respond more strongly to predation risk than barren urchins (i.e., in less nourished condition) tested at high densities typically observed in this habitat. Indeed, when tested at densities associated with respective habitats, urchins from forests, but not barrens, reduced kelp grazing by 69% when exposed to lobster risk cues. Barren urchins that were unresponsive to predator cues at natural, high densities suddenly responded strongly to lobster cues when conspecific densities were reduced. Strong responses of low densities of barren urchins persisted across feeding history (i.e. 0-64 days of starvation). This suggests that barren urchins can respond to predators but typically do not because of high conspecific densities. Because high densities of urchins in barrens should weaken the non-consumptive effects of lobsters, urchins in these habitats may continue to graze in the presence of predators thereby providing a feedback that maintains urchin barrens.
