As climate change continues, climactic extremes are predicted to become more frequent and intense, in some cases resulting in dramatic changes to ecosystems. The effects of climate change on ecosystems will be mediated, in part, by biotic interactions in those ecosystems. However, there is still considerable uncertainty about where and how such biotic interactions will be important in the context of ecosystem disturbance and climactic extremes. Here, I review the role of consumers in seagrass ecosystems and investigate the ecological impacts of an extreme climactic event (marine heat wave) and subsequent widespread seagrass die-off in Shark Bay, Western Australia. Specifically, I compare seagrass cover, shark catch rates, and encounter rates of air breathing fauna in multiple habitat types before and after the seagrass die-off to describe post-disturbance dynamics of the seagrass community, shifts in consumer abundances, and changes in risk-sensitive habitat use patterns by a variety of mesoconsumers at risk of predation from tiger sharks (Galeocerdo cuvier). Finally, I conducted a 16 month field experiment to assess whether xi loss of top predators, and predicted shifts in dugong foraging, could destabilize remaining seagrass. I found that the previously dominant temperate seagrass Amphibolis antarctica is stable, but not increasing. Conversely, an early-successional tropical seagrass, Halodule uninervis, is expanding. Following the die-off, the densities of several consumer species (cormorants, green turtles, sea snakes, and dugongs) declined, while others (Indo-Pacific bottlenose dolphins, loggerhead sea turtles, tiger sharks) remained stable. Stable tiger shark abundances following the seagrass die-off suggest that the seascape of fear remains intact in this system. However, several consumers (dolphins, cormorants) began to use dangerous but profitable seagrass banks more often following seagrass decline, suggesting a relaxation of anti-predator behavior. Experimental results suggest that a loss of tiger sharks would result in a behaviorally mediated trophic cascade (BMTC) in degraded seagrass beds, further destabilizing them and potentially resulting in a phase shift. My work shows that climactic extremes can have strong but variable impacts on ecosystems mediated in part by species identity, and that maintenance of top predator populations may by important to ecological resilience in the face of climate change.