Líffræðifélag Íslands - biologia.is
Líffræðiráðstefnan 2025
Erindi/veggspjald / Talk/poster V78
Höfundar / Authors: Lieke Ponsioen (1), Kalina H. Kapralova (2), Fredrik Holm (1), Benjamin D. Hennig (1)
Starfsvettvangur / Affiliations: (1) University of Iceland, (2) Keldur
Kynnir / Presenter: Lieke Ponsioen
Salmonids are particularly vulnerable during their embryonic development, yet monitoring of their spawning habitats remains infrequent and typically depends on manual redd (nest) counts. This traditional approach is often subject to significant sampling errors, leading to potential over- or underestimations. However, because salmonid spawning habitat in shallow water areas can be distinguished by their visible reflection, they can be identified using standard unmanned aerial vehicles (UAVs, better known as drones). Here, we developed a standardised, low-cost method for detecting salmonid spawning habitat using a semi-automated classification workflow. RGB imagery was collected via UAVs from two contrasting Icelandic lakes, and a supervised classification approach was applied. Six endmember classes were identified for each site with high classification accuracy. Most importantly, producer’s and user’s accuracy for classifying spawning redds was >90% after applying post-classification improvements for both study areas. Our results present a novel, efficient alternative to traditional redd counting, overcoming key limitations such as observer bias and limited temporal and spatial coverage. This method enables large-scale data collection and analysis in a cost- and time-effective manner, providing a valuable tool for long-term monitoring of salmonid spawning habitats.