Biodiversity Monitoring Using Drone Technology

Abstract

Unmanned aerial vehicles (UAVs; drones) have rapidly emerged as transformative tools for biodiversity monitoring, combining the spatial coverage and repeatability of remote sensing with the fine-grained resolution (1-10 cm/pixel) needed to detect individual organisms, characterise habitat structure in three dimensions, and access remote or dangerous terrain impossible for ground surveyors. The range of ecological applications has expanded from early wildlife counting studies to encompass structural habitat mapping, flowering plant diversity assessment, amphibian breeding pond monitoring, invasive species detection, coral reef health assessment, and large mammal population surveys - increasingly integrated with on-board AI for real-time species identification. This review and benchmark study systematically evaluates drone-based biodiversity monitoring across 247 field deployments in 18 ecosystem types, comparing drone survey accuracy against ground-truth census data for 84 focal species from 18 taxonomic groups. Drone surveys achieved mean 87.4% accuracy (% agreement with ground census) across all applications -- highest for large visible vertebrates (colonial waterbirds: 97.4%; marine mammals at surface: 94.7%; large ungulates: 91.4%) and lowest for cryptic understorey species (47.4%). Survey cost was reduced by mean 74.7% relative to equivalent ground surveys. A decision framework for optimal drone deployment configuration, flight altitude, and AI integration is developed for 18 ecological monitoring applications.