First Paragraph
Cetaceans are increasingly being included as top trophic-level predators in models of ecosystem dynamics (Baumgartner and Mate, 2003; Tynan, 2004; Redfern et al., 2006). Traditional visual survey methods for cetaceans detect only a fraction of the animals present, both because visual observers can see them only during the very short period when they are at the surface, and because visual surveys can be undertaken only during daylight hours in relatively good weather (Mellinger and Barlow, 2003). Perhaps more importantly, visual survey results can be highly variable, due both to clumping of cetaceans into large groups and to their relatively limited spatial and temporal scales. Surveys are typically performed using a small number of observation points—one or a few vessels—for a few weeks to a few months of the year.