Linking noninvasive genetic sampling and traditional monitoring to aid management of a trans‐border carnivore population

Abstract

Noninvasive genetic sampling has been embraced by wildlife managers and ecologists, especially those charged with monitoring rare and elusive species over large areas. Challenges arise when desired population measures are not directly attainable from genetic data and when monitoring targets trans-border populations. Norwegian management authorities count individual brown bears (Ursus arctos) using noninvasive genetic sampling but express management goals in the annual number of bear reproductions (females that produce cubs), a measure that is not directly available from genetic data. We combine noninvasive genetic sampling data with information obtained from a long-term intensive monitoring study in neighboring Sweden to estimate the number of annual reproductions by females detected within Norway. Most female brown bears in Norway occur near the border with neighboring countries (Sweden, Finland, and Russia) and their potential reproduction can therefore only partially be credited to Norway. Our model includes a simulation-based method that corrects census data to account for this. We estimated that 4.3 and 5.7 reproductions can be credited to females detected with noninvasive genetic sampling in Norway in 2008 and 2009, respectively. These numbers fall substantially short of the national target (15 annual reproductions). Ignoring the potential for home ranges to extend beyond Norway's borders leads to an increase in the estimate of the number of reproductions by -30%. Our study shows that combining noninvasive genetic sampling with information obtained from traditional intensive/invasive monitoring can help answer contemporary management questions in the currency desired by managers and policy makers. Furthermore, combining methodologies and thereby accounting for space use increases the accuracy of the information on which decisions are based. It is important that the information derived from multiple approaches is applicable to the same focal population and that predictions are cross-validated. When monitoring and management are constrained to administrative units, census data should be adjusted by discounting portions of individual space utilization that extend beyond the focal jurisdiction. Our simulation-based approach for making such an adjustment may be useful in other situations where management authorities target portions of trans-border populations.