Wolf feeding ecology in a multi-ungulate system – investigating the effect of individual predator traits and abundance of co-occurring species

Predation patterns by large carnivores are influenced by a combination of physical, behavioural, and environmental factors. With the recolonization of these apex predators reaching vast parts of their former ranges, there is a need to better understand carnivores’ ecology and effect on prey populations in human dominated landscapes. Data on individual traits of large carnivores are difficult to retrieve and even more to associate with individual feeding behaviour. Alongside cluster checks of GPS-collared wolves, the growing field of faecal DNA-based diet analysis in combination with individual genotyping has the potential to increase the feasibility of large-scale analyses of food use related to individual predator traits. However, the validation of prey DNA detection protocols is still lagging behind the methodological advances. In this thesis, the development and empirical validation of a molecular method for prey DNA detection was followed by the analysis of intrinsic and extrinsic factors affecting feeding ecology of wolves (Canis lupus). The method used nanofluidic array technology and a set of 80 multiple species-specific markers to detect DNA of 17 target prey from wolf scats. Through controlled feeding experiments with captive wolves, we estimated method sensitivity and accordingly calibrated the thresholds to reliably define a positive prey detection. The application of this methodology to the Scandinavian wolf population revealed variability at the landscape level in the use of the two main prey species, moose (Alces alces) and roe deer (Capreolus capreolus), indicating a dietary response of wolves to changes in wild ungulates relative abundance. In addition, GPS-data showed how scavenging constituted only a minor proportion of wolves’ feeding behaviour, related to season and with moderate support to bear and human density. By taking advantage of the long-term wolf monitoring, we showed that wolf feeding patterns (i.e. prey use and extent of scavenging) were affected by social status, sex, and level of inbreeding. These innovative patterns underline the relevance of considering predator individual traits when studying carnivore feeding ecology. The development and validation of our molecular method highlighted the overlooked relevance of assessing method sensitivity and including it in the evaluation of optimal thresholds for binary detection of prey species in predator scats. As the molecular method can be easily customized to different ecological settings, it may be further developed and applied to other areas and large carnivores. The knowledge gained in this study has the potential to help understanding the impact of recolonizing wolf populations on prey communities and inform the adaptive management of such predator and prey species living in a landscape highly managed by humans.