Environment–isotope associations are consistent across temporal scales but vary across regions and insectivorous bird species

Species distributions and fitness are shaped by environmental conditions. However, assessments of animal–environment associations often rely on occurrence data that are constrained by uneven sampling effort, detectability and geographic biases, and pose particular challenges for migratory species. Stable carbon isotope values (δ13C) have thus been proposed to overcome some of these limitations because δ13C covary with environmental variables and may serve as an indirect proxy for the conditions animals experienced during tissue formation. Despite widespread use of δ13C in ecology, environment–isotope relationships across species and across geographic and temporal scales remain poorly tested. To test within-site repeatability of δ13C and whether environment–isotope associations remain consistent across species, space and time, we analyzed δ13C from 781 locally-formed blood plasma samples of five bird species collected in Nigeria (tropical West Africa) and Czechia (temperate Central Europe). We linked isotope values to remotely sensed environmental conditions gathered for two time periods across 161 sampling sites to quantify environment–isotope associations and their spatiotemporal variation. Carbon isotope values showed moderate within-site repeatability (ICC: 0.57–0.71) across species, yet values at some sites varied as widely as those across entire sampling regions. Isotopic variation was weakly associated with precipitation, surface temperature, Normalized Difference Vegetation Index, and woody vegetation cover, explaining up to 3.1% of variation, whereas differences among sites explained up to 19.0% in Czechia. However, environment–isotope associations were consistent between 2-month and 1-year periods but varied strongly in direction and magnitude between Nigeria and Czechia and among species, suggesting that trophic integration and local food-web complexity reshape baseline carbon signals as they move through local food webs to insectivorous birds, thereby weakening broad-scale environment–isotope relationships. Our results demonstrate that rather than a transferable proxy of environmental conditions, δ13C should be interpreted as an integrated measure of realized isotopic niche and needs careful validation for the focal tissue, species, region and timescale. Future studies should expand species coverage, capture the full range of isotopic and environmental gradients, and combine multi-isotope and other approaches to identify when and where environment–isotope relationships can be applied broadly.