The Search for the Honduran White Bat ( Ectophylla alba ): Using Cluster Analysis to Identify Bat Species
Wildlife Ecology and Management Elizabeth Grove
Bats vocalize at supersonic frequencies to hunt & navigate (echolocation) and to communicate within social groups (social calls). This makes bats a unique group to detect and monitor using autonomous recording units. Despite this, vocalizations of many bat species are not well studied. Software like Kaleidoscope can automatically identify (Auto-ID) echolocating calls from many insectivorous bats but not echolocating or social calls from fruit bats like the Honduran White Bat ( Ectophylla alba ). Cluster analysis can be used to find calls from other species of bats that are not able to be identified using Auto-ID. This study looks at using cluster analysis to identify species of bats that cannot be Auto-IDed, especially the Honduran White Bat. Bat detectors were placed at several sites in Costa Rica, including spots where there were likely to be Honduran White Bats. Data was run through Auto-ID software and cluster analysis using Kaleidoscope Pro. Clusters of similar calls were compared to sample calls to try to identify species. Using this method could allow more species of bats to be detected and studied.
Advisors: Chris Yahnke, Ph.D.
Subject Area: Wildlife
Type: Oral
34
Jim and Katie Krause CNR Student Research Symposium
Made with FlippingBook - Share PDF online