Can owls be used to monitor the impacts of urbanisation? A cautionary tale of variable detection

Context Due to their important ecological roles, predators are increasingly being suggested as targets for biodiversity studies investigating how they respond to landscape change and transformation. But there is limited literature investigating our capacity to accurately monitor changes in their occupancy. Aims To test the efficacy of playback surveys for monitoring owls as a basis for investigating change in owl occupancy over time. We ask whether playback is an effective tool, and whether it can be optimised to improve its utility. Methods Using the urban–forest interface of Melbourne, Australia, as a case study, we used playback techniques to survey for the presence of three owl species: the powerful owl (Ninox strenua); southern boobook (Ninox boobook); and eastern barn owl (Tyto javanica). Sites were repeat surveyed at least 16 times throughout the year and occupancy models were developed to establish how season and temperature influence nightly detection probabilities of owls. Key results All three species of owl were detected through playback survey approaches, but the detection probabilities varied greatly between species and across seasons and temperature conditions. Eastern barn owls are poor candidates for playback surveys due to their low detection probabilities. The southern boobook and powerful owl are responsive to playback, but detection probabilities are influenced by season and/or temperature conditions. To optimise survey approaches, southern boobooks should be surveyed during spring and summer and the powerful owl should be surveyed on nights where the minimum temperature is near 20°C. Conclusions Although there is considerable interest in using predators such as owls to monitor biodiversity impacts associated with landscape change, poor detection rates can limit their utility. However, optimising survey approaches that consider shifting detection probabilities under different conditions such as time of year or temperature may improve the utility of predators as surrogates in biodiversity monitoring. Implications Optimising survey approaches for owls considerably reduces the window of opportunity in which to conduct surveys. To counter this, the intensity of survey effort needs to be increased during key periods. The use of highly trained citizen science teams may be one effective way of delivering such an approach.