Marine mammals feed at a variety of trophic levels, occur from freshwater to openocean ecosystems and are found across virtually all latitudes. Due to their high historical, and sometimes present-day, abundances, capability for large-scale movements and high metabolic rates, they have the potential to affect the structure and function of ecosystems through a variety of mechanisms over both ecological and evolutionary time. Usually, the effects of marine mammals on ecosystems are explicitly or implicitly considered to occur through their ability to remove prey through direct predation. Recent empirical studies and a rich theoretical framework, however, demonstrate that marine mammals can affect ecosystems through more diverse pathways, including those that are driven by marine mammal behaviour. Thus, non-consumptive effects of and on marine mammals may be critical in shaping their ecological importance. Nonconsumptive effects may include risk effects, whereby predators induce costly changes to prey behaviour that impact prey population sizes or the magnitude and spatiotemporal patterns of prey impacts on communities (e.g. behaviour-mediated trophic cascades). Changes in the abundance of large apex predators (both marine mammals and sharks) and the introduction of perceived and real risks (human disturbance) may also affect behaviours of marine mammals and their prey that cascade to the wider ecosystem; the conditions under which such cascading effects might be most important, however, remain poorly understood. Other behaviour-driven ecological roles of marine mammals may include foraging tactics that facilitate the foraging of other species (especially seabirds), translocating nutrients and linking the dynamics of spatially distinct food webs.

Predators elicit prey spatial shifts that may influence prey resources. The nature of these indirect effects is difficult to predict, however, in part because the means by which prey differentiate safe from dangerous space are poorly understood. Prey often avoid their predators, but they may instead favor predator-rich areas that facilitate escape, or discourage attack, when a predator is encountered. We investigated how olive-headed sea snakes Disteria major index their risk of tiger shark Galeocerdo cuvier predation over seagrass bank microhabitats (edges, interiors) in Shark Bay, Australia. D. major is equally likely to escape sharks in both microhabitats, so we expected to observe avoidance of predator-rich space. Supporting our prediction, snakes used microhabitats roughly in proportion to food supply when sharks were scarce and avoided edges, which are preferred by sharks, when sharks were abundant. Thus, D. major appears to measure danger across seagrass banks using variability in predator density and to seek low-encounter microhabitats when antipredator investment is needed. Our results suggest that the influence of predators on sea snakes is underappreciated and, in the context of previous work, that sympatric prey species sharing predators may show opposite spatial shifts when threatened, potentially leading to different predator indirect effects. KEY WORDS: Anti-predator behavior · Predation risk · Predator indirect effects · Prey escape tactics · Tiger shark Resale or republication not permitted without written consent of the publisherMar Ecol Prog Ser 387: [287][288][289][290][291][292][293] 2009 tiate safe and dangerous areas can in fact rely on information pertaining to one or both of these components of risk. In situations where the probability of capture is spatially uniform, we would expect safety-conscious prey individuals to select locations where predators are scarce and encounter rates are relatively low when there is need for anti-predator investment (i.e. when predators are present in the environment; Lima 1992).In Shark Bay, Disteria major forages diurnally for small teleosts over shallow, offshore seagrass banks (1.5 to 4.5 m in depth) that are surrounded by largely unvegetated deeper waters (6 to 12 m; Kerford 2005). The fishes on which D. major subsists are evenly distributed over these seagrass banks (based on species counts in fish traps, see Heithaus & Dill 2006), but predation risk is not. Instead, habitat use patterns of tiger sharks, the only major local sea snake predator (Heithaus 2001), indicate that they are more likely to be encountered along the periphery of seagrass banks (edge microhabitats) than at their center (interior microhabitats; Heithaus et al. 2006). Following encounters with predators over seagrass banks, D. major escapes into the vegetation (A. Wirsing pers. obs.), which provides a structural refuge (Kerford 2005, Kerford et al. 2008). Because seagrass is equally available and accessible in both microhabitats (Wirsing et al. 2007), D. major shoul...

No abstract

Sirenians (manatees Trichechus spp. and the dugong Dugong dugon) are large-bodied and almost exclusively herbivorous aquatic mammals occurring in coastal and freshwater habitats of most tropical and subtropical regions. Although sirenian ecology, habitat use, and abundance have been investigated across their range, little is known about the roles and impacts of these megaherbivores outside of the trophic relationships of 2 relatively well-studied species (T. manatus, D. dugon). This knowledge gap limits our understanding of how sirenians affect communities and ecosystem dynamics. We review the literature on the ecological roles and importance of sirenians in aquatic ecosystems, including the extinct Steller’s sea cow Hydrodamalis gigas. Our review reveals that sirenian herbivory and disturbance can markedly affect species biomass, productivity, and composition in macrophyte communities, thereby indirectly influencing other community members (e.g. invertebrates) and carbon storage in macrophyte biomass and sediments. Sirenians may also couple disparate ecosystems via their movements, compete with other herbivores, and serve as prey for, and be responsive to, predators, including humans. We also identify key knowledge gaps, including (1) the ecology of Amazonian T. inunguis and African T. senegalensis manatees, (2) experimental investigations of the magnitude of trophic impacts and their potential to cascade to other species and carbon dynamics, (3) the roles and impacts of sirenians as omnivores, and (4) the non-consumptive effects of predators on sirenians other than dugongs. Finally, we highlight crucial avenues for future sirenian research, including changes to their ecology in the face of ongoing seagrass declines and climate change.