Tomaž Skrbinšek

Reliable estimates of brown bear abundance are essential for effective conservation management. Non-invasive genetic monitoring provides robust and precise census estimates, but it is costly, logistically demanding, and often delayed due to extensive laboratory processing.
In Slovenia, the first national genetic survey was conducted in 2007 and repeated at eight-year intervals (2015, 2023) in accordance with the national monitoring strategy. Because management requires annual, up-to-date information, we developed demographic models calibrated with genetic estimates and validated using independent data (e.g., mortality records and monitoring at permanent counting sites). Models are updated annually with new mortality and demographic data and recalibrated after each genetic survey.
Simulations showed that models calibrated in 2007 and 2015 would deviate by up to 7% over an eight-year projection period. The 2023 genetic survey enabled the first empirical blind validation: eight years after calibration, model predictions differed from the genetic estimate by only 2.7%, with a mean deviation below 1% (≈8 individuals). For management purposes, substantially larger deviations (up to ~20%) would remain acceptable, confirming both the reliability of the modeling framework and the adequacy of the eight-year genetic interval.
Demographic modeling is approximately 100 times less costly than genetic censusing and provides fully up-to-date annual estimates, whereas genetic results typically reflect population status with a 1.5–2-year delay. Its main limitation is the need for periodic calibration against reliable empirical estimates. The key strength thus lies in integrating both approaches: periodic genetic surveys combined with annual, model-based estimation. We believe that developed methods and framework is transferable to other brown bear populations and wildlife species, provided that reliable long-term mortality data are available.