Understanding the drivers and consequences of global environmental change is crucial to inform predictions of effects on ecosystems. We used the mammal community of Białowieża Forest, the last lowland near-primeval forest in temperate Europe, as a sentinel of global change. We analyzed changes in stable carbon (δ¹³C) and nitrogen (δ¹⁵N) isotope values of hair in 687 specimens from 50 mammal species across seven decades (1946–2011). We classified mammals into four taxonomic-dietary groups (herbivores, carnivores, insectivores, bats). We found a significant negative trend in hair δ¹⁵N for the mammal community, particularly strong for herbivores. This trend is consistent with temporal patterns in nitrogen deposition from (¹⁵N depleted) industrial fertilizers and fossil fuel emissions. It is also in line with global-scale declines in δ¹⁵N  reported in forests and other unfertilized, non-urban terrestrial ecosystems and with local decreases in N foliar concentrations. The global depletion of ¹³C content in atmospheric CO₂ due to fossil fuel burning (Suess effect) was detected in all groups. After correcting for this effect, the hair δ¹³C trend became non-significant for both community and groups, except for bats, which showed a strong decline in δ¹³C. This could be related to an increase in the relative abundance of freshwater insects taken by bats or increased use of methane-derived carbon in food webs used by bats. This work is the first broad-scale and long-term mammal isotope ecology study in a near-primeval forest in temperate Europe. Mammal communities from natural forests represent a unique benchmark in global change research; investigating their isotopic temporal variation can help identify patterns and early detections of ecosystem changes and provide more comprehensive and integrative assessments than single-species approaches.

The dataset contains the δ¹³C and δ¹⁵N stable isotope values for 687 hair samples belonging to 50 mammal species and collected between 1946 and 2011 in Białowieża Forest (Poland). The δ¹³C values corrected for the Suess effect are also included (see the article for explanations on the formula applied). The dataset also includes the species name (both common and scientific) and the year of collection.