Utilizing Remote Sensing Data for Species Distribution Modeling of Birds in Croatia

Accurate information on species distributions and population sizes is essential for effective biodiversity conservation, yet such data are often lacking at national scales. This study addresses this gap by assessing the distribution and abundance of 111 bird species across Croatia, including breeding, wintering, and migratory flyway populations. We combined Species Distribution Models (SDMs) with expert-based population estimates to generate spatially explicit predictions. The modeling framework incorporated high-resolution Earth observation (EO) data and advanced spatial analysis techniques. Environmental variables, such as land cover, were derived from satellite datasets, while climate variables were interpolated from ground measurements and refined using EO-based co-variates. Model calibration and validation were based on species occurrence records and EO-derived predictors. This integrative approach enabled both national-scale population estimates and fine-scale habitat assessments. The results identified critical habitats, population hotspots, and areas likely to experience distribution shifts under changing environmental conditions. By integrating EO data with expert knowledge, this study enhances the robustness of population estimates, particularly where species monitoring data are incomplete. The findings support conservation prioritization, inform land use and resource management, and contribute to long-term biodiversity monitoring. The methodology is scalable and transferable, offering a practical framework for ecological assessments in diverse regions. We integrated expert-based population estimates with species distribution models (SDMs) by applying expert-derived density values to areas of suitable habitat predicted by SDMs. This approach enables spatially explicit population estimates by combining ecological modeling with expert knowledge, which is particularly useful in systems with limited data. Experts provided species-specific density estimates stratified by habitat type, seasonality, behavior, and detectability, aligned with habitat suitability classes derived from SDM outputs.