Application of Unmanned Aerial Vehicles and Thermal Imaging Cameras to Conduct Duck Brood Surveys

Abstract

Brood surveys are used to estimate productivity in ducks, but road‐side transects, aerial surveys, and double‐observer ground surveys have likely underestimated productivity. Duck broods are elusive and prefer wetlands with emergent vegetation where they hide at signs of disturbance, making it difficult to get accurate brood counts. Estimates of brood detection probabilities are typically below 50% and variable, which makes biological inferences about abundance tenuous. We conducted a study to evaluate the efficacy of using an unmanned aerial vehicle (UAV) equipped with a thermal imaging camera to survey duck broods in 2 study areas. In Manitoba we located 669 broods with the UAV, compared to 344 detected by double‐observer ground surveys. In Minnesota we detected 225 ducks broods with the UAV, whereas only 105 duck broods were detected by ground observers. Using a Huggins closed‐capture model in program MARK we estimated an average detection probability across both sites of 0.55 (SE = 0.02) with the UAV compared to 0.24 (SE = 0.02) for the ground crews. Although the UAV detected twice as many broods as the ground surveys, detection probability with the UAV was impacted by temperature, humidity, vegetation density, and the criteria we used to determine whether a brood could be classified as resighted. Nevertheless, using a UAV equipped with a thermal imaging camera effectively doubled the number of broods detected compared to traditional methods, and surveys were completed 3 times faster. With advancing drone and camera technology we believe UAV brood counts will become increasingly accurate and provide reliable measures of local duck productivity. © 2021 The Wildlife Society.

Detection Probability and Bias in Machine-Learning-Based Unoccupied Aerial System Non-Breeding Waterfowl Surveys

“…Previous studies using remote sensing found that vegetation cover type, sky condition, species, and survey altitude or GSD can all influence waterfowl detection within a survey area [20,[36][37][38][39]42,43]. Similarly, our results indicate that in UAS imagery, the waterfowl availability to be counted was strongly influenced by the combination of species, sex, vegetation cover type, and GSD.…”

“…Overall, our results show that UASs, aerial imagery, artificial intelligence algorithms, and the use of correction factors may accurately estimate waterfowl abundance within single images across a wide range of environmental and flight conditions. Previous studies evaluating bird detection biases found that vegetation cover type, sky condition, species, and survey altitude or GSD can all influence bird detection within a survey area [20,[36][37][38][39]42,43]. We found that waterfowl availability bias was most influenced by environmental conditions, including vegetation cover type, GSD, and waterfowl characteristics, including species and sex, while perception bias was most influenced by GSD, affecting the probability of detecting birds, and vegetation-cover type, influencing the probability of algorithms generating false positives.…”

“…Advancements in UAS platforms combined with advances in machine learning algorithms may allow managers to monitor waterfowl populations and inform daily management decisions in real time; however, the mixed effects of sunlight, sun position, substrate contrast, and vegetation coverage on waterfowl detection and identification during nonbreeding seasons in UAS surveys require further study [19,28,[37][38][39][40][41]. Although machine learning algorithms have shown promise for quickly detecting avifauna from aerial imagery, certain environmental factors, including vegetation cover and sky conditions that impact availability and detection of avifauna in ocular surveys or by human analysis of the imagery, can also reduce the accuracy of machine-learning algorithms [21,24,26,33,[37][38][39]42,43]. Pre-vious research reported that algorithm-based waterbird detection was influenced by image quality, sun reflections on the water, the presence of bird shadows, vegetation cover type characteristics, and the physical characteristics of the birds themselves, including their coloration, size, and shape [24,31,43,44].…”

“…In a study that identified and detected various species of herons from UAS-collected imagery, vegetation cover types that caused visual occlusion created a greater source of error than individual observer bias [37]. Waterfowl broods were more likely to be detected when located in wetlands with short, sparse vegetation compared to wetlands with dense emergent vegetation, and the detection of nesting waterfowl was significantly lower than the brood detection, likely due to overhead visual occlusion from vegetation over nesting waterfowl [33,38,39]. Cloud cover and amount of sunlight have been shown to not influence the detection of darker-plumaged individuals and decoys, whereas cloud cover had a negative effect on the detection of lighter-colored decoys [37].…”

Comparing methods to estimate feral burro abundance

“…IR sensor and processing technology has progressed and become more widely available since the timing of our surveys (2015–2016); thus, the monetary cost has possibly decreased. Unmanned aerial vehicles, or drones, are a cheaper platform for affixing IR cameras and IR surveys using drones have been widely tested with mixed results (Chrétien et al 2016, Beaver et al 2020, Bushaw et al 2021, Preston et al 2021). Drone flights are currently not practical in most feral horse and burro management areas however, because federal policy requires direct line of sight control of the aircraft (FAA 2021) which is difficult to achieve across these vast areas with rugged topography limiting vehicular access.…”