Detectability and visibility biases associated with using a consumer‐grade unmanned aircraft to survey nesting colonial waterbirds. Issue 3 (1st August 2018)

• Record Type: Journal Article
• Authors: Barr, Jared R.; Green, M. Clay; DeMaso, Stephen J.; Hardy, Thomas B.
• Is Part Of: Journal of field ornithology. Volume 89:Issue 3(2018)
• Page Start: 242
• Page End: 257
• Publication Date: 2018-08-01
• Subjects: aerial system -- bird count -- detection probability -- drone -- GLMM -- UAS -- UAV; Birds -- Periodicals; Ornithology -- Periodicals; Ornithologie -- Périodiques; Oiseaux -- Baguage -- Périodiques; Oiseaux -- Périodiques; 598
• DOI: 10.1111/jofo.12258
• Languages: English
• ISSNs: 0273-8570

Abstract

Abstract: Surveys of colonial‐nesting waterbirds are needed to assess population trends and gain insight into the health of wetland ecosystems. Use of unmanned aerial systems (UAS) for such surveys has increased over the past decade, but possible sources of bias in surveys conducted with UAS have not been examined. We examined possible visibility biases associated with using a UAS to survey waterbird colonies in cypress‐tupelo watersheds and coastal island habitats in Texas in 2016. We used known numbers of four waterbird decoy types, including Black Skimmers ( Rynchops niger ), terns, and white‐ and dark‐plumaged herons, to estimate their detectability in each habitat. Six observers independently counted decoys from aerial imagery mosaics taken with a consumer‐grade, off‐the‐shelf quadcopter drone. We used generalized linear mixed‐effects models to estimate detection probabilities of each decoy type. Black Skimmers at the coastal island had a detection probability of just 53%. Detectability of both white‐ and dark‐plumaged herons was lower in the canopied cypress‐tupelo habitat than the coastal island. In addition, cloud cover > 50% further reduced detectability of white heron decoys in cypress‐tupelo habitat. Use of the double‐count method yielded biased‐low abundance estimates for white‐ and dark‐plumaged herons in canopied sites, suggesting that habitat differences were a greater source of bias than observer error. Black Skimmers were the only decoy type to be imperfectly detected at the coastal island, a surprising result given the stark contrast of their plumage with their sand and shell nesting substrate. Our results indicate that UAS‐derived photographic surveys are prone to low detection probabilities at sites where vegetation occludes nests. In habitats without canopy, however, UAS surveys show promise for obtaining accurate counts of terns, white herons, and dark herons.