| In the absence of large mammalian predators, most ungulate populations studied has increased rapidly and are probably capable of overexploiting forage and subsequently becoming limited by food shortages. Food supply is therefore often mentioned as the primary factor limiting large herbivore population growth rate and bottom-up control is generally suggested to be the universal primary standard. However, bottom-up control can be overridden by secondary processes i.e. top-down processes from predators as an example. There are several examples of to-down processes by predation with devastating effects are introductions of non-native predators that has caused the extinction of species and whole community assemblages. However, examples from natural systems with native predator-prey communities are rare.
This study focuses on an almost 40 year long time-series collected by me and my predecessor at Grimsö Wildlife Research Station (Dep. of Ecology, SLU). A unique data set, based on 458 known-aged, radio collared roe deer followed throughout their life, with known mortality causes. This population has experienced four risk regimes during that time: one predator-free regime during the peak years of the sarcoptic mange that severely limited the red fox population and reduced predation on neonate fawns for a few years, and three increasingly complex predator communities successively adding fox, lynx and finally during the last eight years also wolf.
Presently, reproductive lynx and wolf populations are found almost throughout the Scandinavian Peninsula and their influence on prey populations is a highly debated societal issue. To bring facts into that debate, even if only from one population, might in fact balance the arguments and assist the conservation of these charismatic large carnivores.
To assess the impact on the roe deer population growth rate of the successively increasing complexity in the carnivore community, we will use the latest tools in terms of survival analysis. Firstly, the Non-Parametric Cumulative Incidence Function Estimator that helps testing the hypothesis of additive vs. compensatory predation-related mortalities. Secondly, by extending the competing-risk framework, we will investigate the influence of several variables (sex, age, winter severity, etc.) for each predation risk and finally, will Cox Proportional Hazards be implemented to detect the main drivers of prey selection.
The last step will be to parameterize a Matrix-Projection Model with survival rates obtained from the analysis and reproductive rates estimated in the same population, in order to project roe deer population growth rate (λ) within each predator community.
This work will in parts be performed by a highly competent MSc-student and I feel confident that by a successful outcome of this proposal we will be able to finalize all the analysis not included in the MSc and also a draft for submission to a high ranked journal within the month of salary we here ask for. |