Seasonal drivers of reported human-wild boar incidents in a human-dominated biodiversity hotspot

Human-wild boar conflict incidents (HWBC) are increasingly reported in human-dominated landscapes, yet their seasonal drivers remain poorly quantified at broad scales in Türkiye. Using 105 summer and 152 winter reported incidents collected from media and online sources between 2006 and 2025, this study examines how land use, topography, and human presence shape the spatial patterns of reported HWBC incidents across the country within a resource selection function framework. Incident locations were contrasted with 1000 road-biased background points, and environmental predictors (reclassified CORINE land-cover classes, elevation, terrain ruggedness, and settlement cover) were summarized in 5-km buffers and analysed with season-specific logistic regression models, multi-model inference, and Kruskal-Wallis/Wilcoxon tests; temporal variation among pre-COVID, lockdown, and post-lockdown periods was evaluated with a chi-square test. In both seasons, reported encounters clustered at low elevations in settlement-rich landscapes, with urban cover strongly increasing conflict probability and elevation consistently reducing it, while cropland, forest, and grass-sparse vegetation showed negative associations, indicating that conflicts are more likely where natural or semi-natural cover is relatively scarce within human-dominated mosaics. Seasonal contrasts emerged for orchards and ruggedness: orchard cover and terrain ruggedness increased conflict probability in winter but had weak or no effects in summer, suggesting that winter foraging on tree crops and reliance on rough terrain as security cover amplify conflict risk in lowland peri-urban and agricultural areas. Incident frequencies differed significantly among the pre-COVID, lockdown, and post-lockdown periods, implying that changes in human mobility, traffic, and reporting effort during the pandemic contributed to observed temporal patterns in HWBC. Because the response variable is based on media- and online-reported incidents rather than systematic damage assessments, the patterns described reflect spatial distributions of reported encounters and complaints rather than a complete census of all conflict outcomes.