COS 90-5 - Real-time thermal imagery from an unmanned aerial vehicle locates nests of a grassland songbird at rates similar to traditional methods

Thursday, August 15, 2019: 2:50 PM
L007/008, Kentucky International Convention Center
Darren S Proppe1, Chace N Scholten2, Alexandra J Kamphuis2, Kaitlyn J Vredevoogd2, Kristin G Lee-Strydhorst2, Jenna L Atma2, Connor B Shea2 and Olivia N Lamberg2, (1)Biology, Calvin College, Grand Rapids, MI, (2)Calvin College
Background/Question/Methods

Monitoring songbird reproductive output is a well-established method for assessing population persistence, but reproductive studies in grassland songbirds often suffer from low sample sizes and high labor costs due to the difficulty of locating nests. Physical disturbance to nesting habitats during nest surveys can also be problematic. However, technological advances in unmanned aerial vehicles (UAVs or drones) might enhance our ability to locate these cryptic nests. We compared the effectiveness of locating grassland songbird nests with a thermal camera mounted on a UAV versus more traditional targeted nest searches, which included behavioral observation and intentional flushing. We used a paired experimental design with two rotating research teams to survey twenty field sparrow (Spizella pusilla)territories with known active nests. Each search lasted for up to two hours and the two search methods occurred on consecutive days. Research teams were blind to the nest location and the results from the alternate search method. For each search, we recorded the outcome and time it took to locate the nest.

Results/Conclusions

Early morning UAV searches located nests 35% faster on average than traditional methods. However, search times within each method were variable and the results from a survival-style analysis indicated that the efficiency between the thermally equipped drone and more traditional methods did not differ statistically. Although UAV-assisted searches did not increase efficiency, they did reduce the need to traverse and trample vegetation within the territory since the aircraft could be piloted from the territory edge. UAV-assisted searches may be less invasive than traditional methods, although further study is needed to assess how UAV presence impacts avian stress responses. Nonetheless, our results indicate that UAV-assisted nest searches can replace traditional, targeted nest searching techniques in a ground nesting grassland bird. Continued advances in UAV and thermal technology are likely to increase the efficiency of UAV-assisted nest searches in the near future, and may eventually remove the need for humans to enter target territories when monitoring nest success.