Autonomous sensing methods are transforming the scale and resolution of ecological data collection. In particular, autonomous acoustic recording shows promise for surveying sound-producing wildlife such as birds, bats, anurans, insects, and canids. Yet until recently, expensive recording technology and time-consuming data analysis limited the adoption of these methods at scale. To harness the full potential of acoustic monitoring, ecologists are combining new low-cost autonomous recorders with machine learning algorithms that rapidly identify the species vocalizing in each recording.
This webinar will introduce the principles of conducting these autonomous acoustic surveys. We will describe a framework for implementing acoustic monitoring at large scales using an open-source, inexpensive acoustic recorder, the AudioMoth. We will discuss how autonomous recorders can be used to survey animals in a variety of challenging scenarios, including monitoring remote locations, assessing species presence across large areas, and searching for rare, disturbance-sensitive, or nocturnal animals. Finally, we will describe how these acoustic data can be evaluated using machine learning and other analysis techniques.
Tessa Rhinehart is a Research Programmer in the Kitzes Lab at the University of Pittsburgh. She develops machine learning and statistical methods to survey animals on large scales.