Carlos Bautista, Natalia Bragalanti, Paolo Ciucci, Manuel Díaz-Fernández, Néstor Fernández, Claudio Groff, Djuro Huber, Klemen Jerina, Alexandros A. Karamanlidis, Tobias Kummerle, Javier Naves, Julian Oeser, Santiago Palazón, Mihai Pop, Eloy Revilla, Matija Stergar, Nuria Selva

Brown bear damage represents a major conservation and wildlife management challenge across Europe, yet anticipating when such events are more likely to occur remains difficult. Brown bears rely on seasonal forage availability, which in turn might be affected by transient climate shifts (e.g. warmer climate) and changing meteorological extremes (e.g. more extreme drought years). Both may expose bears to nutritional stress during critical periods, forcing them to seek alternative, anthropogenic food sources, thus increasing damage risk. Here, we present a continental-scale analysis examining how year-to-year climate variability influences annual brown bear damage across Europe. We compiled a harmonized dataset of >35,000 compensated damage claims collected over ~21 years across 12 bear management units in temperate Europe. The dataset includes high precision damage locations (~65% GPS-based) and information on damage type (e.g. damage to apiculture, crops and livestock). For each damage event, we extracted two proxies capturing climate variability: meteorological variables describing temperature and precipitation conditions derived from the CHELSA v2.1 dataset, and vegetation phenology metrics derived from MODIS satellite data. We used a Bayesian multi-level regression framework to test specific hypotheses about how climate variability may influence damage occurrence, including whether (i) cold and low-productivity springs and (ii) warm and dry summers, which are associated with reduced or delayed resource availability, and (iii) warm autumn conditions, which are associated with prolonged foraging activity, are linked to increased damage occurrence. We also investigated whether cumulative seasonal effects further amplify damage. Our modelling framework accounts for spatial structure, temporal autocorrelation, and potential lagged effects, with the aim of providing a basis for anticipating damage risk and supporting management decisions under future climate scenarios.