Automated signal recognition as a useful tool for monitoring little-studied species: The case of the Band-tailed Nighthawk

Abstract

Passive acoustic monitoring, when coupled with automated signal recognition software, is a useful technique for monitoring vocally active taxa. In this study, we evaluated the utility of automated signal recognition to gain insights into the ecology of little-studied species. For this purpose, we selected an avian family, Caprimulgidae (nightjars), composed of cryptic and nocturnal species, and focused the study on a Neotropical wetland, the Brazilian Pantanal. We reviewed the number of publications, observations, and recordings available for each nightjar inhabiting the Brazilian Pantanal and the Band-tailed Nighthawk (Nyctiprogne leucopyga) was identified as the species with the least information available. We employed automated signal recognition software to study the vocal behavior of this species over a complete annual cycle in the Pantanal. Previous knowledge about the ecology of this species is based on general descriptions and anecdotal observations. Our findings corroborate that the Nighthawk is a resident species of the Brazilian Pantanal, and according to seasonal changes in vocal activity, the breeding season extends from July to October. The breeding period starts at the end of the dry season (July–August), and the nesting period may occur at the beginning of the wet season and following the first rains, which is a period of maximum insect food availability. The vocal activity of the Nighthawk was restricted to the nocturnal period and was maximum at dusk. That preference for dusk is in disagreement with the pattern described for the other four nightjars in the study area, which highlights the importance of performing species-specific studies and avoiding drawing any conclusions about the activity pattern of a species based on the genus or family to which it belongs to. Automated signal recognition software was able to detect over three quarters of the songs annotated by a human on a subset of sound recordings, therefore proving its utility for monitoring the Band-tailed Nighthawk.