Linking animal movement and remote sensing – mapping resource suitability from a remote sensing perspective

摘要: Optical remote sensing is an important tool in the study of animal behavior providing ecologists with means to understand species–environment interactions combination movement data. However, differences spatial and temporal resolution between data limit their direct assimilation. In this context, we built a data-driven framework map resource suitability that addresses these as well limitations satellite imagery. It combines seasonal composites multiyear surface reflectances optimized presence absence samples acquired within cross-validation modeling scheme. Moreover, it responds dynamic, site-specific environmental conditions making applicable contrasting landscapes. We tested using five populations White Storks (Ciconia ciconia) model related foraging achieving accuracies from 0.40 0.94 for presences 0.66 0.93 absences. These results were influenced by composition indicated lower associated higher day relation target dates. Additionally, population selection our marked negative relationship variability samples. Our approach spatially splits training validation. As result, when represent different unique resources, face bias during Despite inaccuracies, offers basis analyze interactions. standardizes site-dependent behavioral characteristics, can be used comparison intra- interspecies requirements improves analysis along migratory paths. due its sensitivity selection, contribute toward better understanding species requirements.