Use of unmanned aerial vehicles (UAVs) for mark-resight nesting population estimation of adult female green sea turtles at Raine Island

Abstract Nester abundance is a key measure of the performance of the world’s largest green turtle rookery at Raine Island, Australia. Abundance surveys have been undertaken in waters adjacent to Raine Island reef using mark-resight counts by surface observer (SO), underwater video (UWV) and unmanned aerial vehicle (UAV) (since 1984, 2013 and 2016 respectively). UAV and UWV may provide more cost-effective and less biased alternatives, but estimates must be comparable with the historical estimates. Here we compare the three methods. The relative likelihood of resighting a marked turtle was significantly higher by SO than the other methods, which led to lower mark-resight population estimates than by UAV or UWV. Most (96%) variation in resighting probabilities was associated with survey period, with comparatively little variation between consecutive days of sampling or time of day. This resulted in preliminary correction factors of 1.53 and 1.73 from SO-UWV and SO-UAV, respectively. However, the SO and UWV estimates were the most similar when turtle densities were the lowest, suggesting that correction factors need to take into account turtle density and that more data are required. We hypothesise that the UAV and UWV methods improved detection rates of marked turtles because they allowed subsequent review and frame-by-frame analysis, thus reducing observer search error. UAVs were the most efficient in terms of survey time, personnel commitment and weather tolerance compared to the SO and UWV methods. This study indicates that using UAVs for in-water mark-resight turtle abundance estimation is an efficient and accurate method that can provide an accurate adjustment for historical abundance estimates. Underwater video may continue to be useful as a backup alternative to UAV surveys.