An integrated population modelling workflow for supporting mesopredator management

Expanding populations of mesopredators threaten biodiversity and human health in many ecosystems across the world. Lethal control through harvest is commonly implemented as a mitigation measure, yet the effects of harvest and its interaction with environmental conditions on mesopredator population dynamics have rarely been assessed quantitatively due to data constraints. Recent advances involving integrated populations models (IPMs) have enabled promising alternative approaches for quantitative assessments. Efficient use of multiple datasets, together with the ability to account for bias and uncertainty, make IPMs ideal tools for studying impacts of management actions and environmental conditions on harvested populations for which limited data is available. Here we developed a versatile IPM workflow for studying mesopredator population dynamics under different harvest regimes and applied it to an expanding population of red foxes in Arctic Norway. Our model combined routinely collected data on age, reproductive status, and genetic variation from >3600 harvested red foxes with opportunistic field observations and information published on red foxes elsewhere. This allowed us to quantify population dynamics over a period of 18 years, and to identify the drivers of changes in population growth rates using retrospective (Life Table Response Experiments, LTREs) and prospective (population viability analyses, PVAs) perturbation analyses. We found no long-term trends in population size over the course of our study period, not least due to intense harvest limiting the growth potential of the population. On shorter, year-to-year timescales, however, the numbers of red foxes could change dramatically due to responses of natural mortality and immigration to fluctuations in the availability of rodent prey. Our study highlights the potential of integrated modelling approaches for studying population dynamics even when no structured surveys of living animals are available and illustrates the value of extracting and curating information from harvested animals. Our semi-automated and reproducible modelling workflow is ready to be re-run periodically when new data becomes available for our study population and can easily be transferred and adapted to other harvested species, contributing to the development of cost-effective population analyses that are of high relevance for informing management strategies and mitigating biodiversity loss in practice.