An automated work-flow for pinniped surveys: A new tool for monitoring population dynamics

• DOI: 10.3389/fevo.2022.905309
• Publication Date: 2022-08-11T11:31:26Z
• Authors: Eduardo Infantes; Daire Carroll; Willian T. A. F. ...; Tero Härkönen; Scott V. Edwards; Karin C. Harding
• Keywords: Aerial Survey; Drone; Harbor seal

Detecting changes in population trends depends on the accuracy of estimated mean growth rates and thus quality input data. However, monitoring wildlife populations poses economic logistic challenges especially complex remote habitats. Declines can remain undetected for years unless effective techniques are developed, guiding appropriate management actions. We developed an automated survey workflow using unmanned aerial vehicles (drones) to quantify number size individual animals, well-studied Scandinavian harbour seal ( Phoca vitulina ) as a model species. compared ground-based counts telescopes with manual flights, zoom photo/video, pre-programmed flights producing orthomosaic photo maps. used machine learning identify count both pups older seals we present new method measuring body automatically. evaluate population’s reproductive success drone data, historical predictions from Leslie matrix model. The most accurate time-efficient results were achieved by performing where identified their sizes measured detector was 95–97% classification error 4.6 ± 2.9 3.1 2.1 during good light conditions. There clear distinction between breeding time. 320 season 2021 drone, which is well beyond expected number, based data pup production. high confirms earlier indications deteriorating rate this important colony. show that drones powerful tools inaccessible areas be assess annual recruitment seasonal variations condition.