An evaluation of acoustic field recorders paired with automated call recognition as a monitoring tool for the Mallee Emu-wren Stipiturus mallee

Abstract

Advances in technology are changing the way that ecological monitoring is carried out, especially for those species with ecological characteristics that have traditionally made monitoring difficult. Autonomous acoustic recorders coupled with automated signal detection software is one such approach where technological advances are delivering rapid improvements in the passive monitoring of vocal fauna. Here we characterise the three common call types of the endangered Mallee Emu-wren Stipiturus mallee and present a signal detection template, or call recogniser, for the species. We evaluate the performance of this tool against an independent dataset of field recordings containing Mallee Emu-wren vocalisations. The recogniser performed well with mean precision and recall metrics ranging between 0.55–0.97 and 0.70–0.95, respectively, depending on user parameters. This tool is widely applicable in the ongoing conservation of the Mallee Emu-wren, particularly as a low-cost method for post-release monitoring following a future Mallee Emu-wren translocation.