From Rub Tree Prediction to Targeted Genetic Sampling in Brown Bears: Linking Scent-Marking Ecology and Spatial Modelling

Scent marking has been discussed as an important component of communication in brown bears (Ursus arctos Linnaeus, 1758). However, the environmental factors influencing the occurrence of rub trees and their value for non-invasive genetic sampling remain poorly understood. This study examined the patterns of rub tree occurrence in the eastern High Tatra Mountains (Slovakia) at two spatial scales. At the tree scale, paired-design generalized linear mixed models showed that rub trees were more frequently recorded on large-diameter coniferous trees, indicating an association with visually prominent and chemically suitable substrates. At the landscape scale, logistic regression models revealed that the probability of rub tree occurrence increased with elevation and distance from human settlements, identifying high-elevation forests as areas of higher predicted rub tree occurrence. The best-supported model was used to produce a predictive map of rub tree occurrence across the study area. We also evaluated whether rub trees are reliable sources of biological material for non-invasive sampling. Hair collected during repeated field visits provided DNA suitable for genotyping and individual identification. Overall, the results show that rub trees exhibit non-random spatial patterns and represent effective focal points for systematic genetic sampling, linking patterns of rub tree occurrence to the spatial targeting of non-invasive genetic sampling in mountain landscapes.