Comparison of Real-Time and Post-Cruise Acoustic Species Identification of Dolphin Whistles Using ROCCA (Real-Time Odontocete Call Classification Algorithm)

摘要

Southwest Fisheries Science Center (SWFSC) has been conducting shipboard visual line-transect surveys of marine mammals since the 1970s (Kinzey et al., 2000). Passive acoustic detection methods were incorporated in a systematic manner in 2000 (Rankin et al., 2008a). Passive acoustic detection of marine mammals is not limited by light, sea, and weather conditions and in some conditions, such as low light and inclement weather, can be more effective than visual methods (Thomas et al., 1986). Passive acoustic detection methods have been shown to improve the accuracy of cetacean abundance estimation for some species when coupled with visual detection (Barlow and Taylor, 2005; Barlow and Rankin, 2007). In addition, passive acoustic detection provides valuable information about marine mammals missed by visual observers during surveys (Rankin et al., 2008a). The ability of passive acoustic methods to aid in population surveys depends on the ability to accurately determine species identity from acoustic detections.