Abstract
Passive acoustic detectors are widely used for monitoring distribution of cetaceans. Autonomous visual methods are less frequently employed; they are limited to detections during daylight and good weather, but offer potential advantages due to certainty of species' identification and longevity of deployment. To compare performance of acoustic and visual methods, temporal changes in distribution of southern right whales, Eubalaena australis, were monitored in the remote sub-Antarctic Auckland Islands (50 degrees 31' S, 166 degrees 16' E). A time-lapse camera was deployed for 2 years from August 2010, taking an image every 70 min. The presence of whales was scored for each image taken during daylight (n = 8295). A passive acoustic recorder was deployed in August 2011, recording for 3.75 min every hour until battery life expired after 10 months. Each recording (n = 6978) was aurally reviewed. Both methods revealed a similar seasonal distribution; maximum detection rates were in the austral winter and no whales were detected in January or February. However, at the peak of right whale occurrence, the proportion of samples with whales detected was much higher for the acoustic recorder (93.9% in August) than the time-lapse camera (14.8%). A generalised additive model fitted to the visual data revealed significant effects of sea state and visibility. Acoustic detection rates were higher, probably because detection range is greater, and less affected by weather. The solar-powered time-lapse camera system, however, functioned effectively for much longer. We discuss the relative merits of visual and acoustic detectors and attempt to draw conclusions about their efficacy for different focal species and monitoring locations.