Advancing understanding of ecosystem functioning through food web modelling
Ecological communities can be described as networks, with species as nodes and interactions as links. These networks are the backbone of ecosystems all over world and provide many amenities that our society benefit from. Most ecosystem functions and related services provided by such networks involve interactions among organisms at different trophic levels, e.g. pollination and biological pest control. Our mechanistic understanding of such ecosystem functioning, that depend on species interactions, in multi-trophic communities is poor however, and research has been limited to experiments on small study systems, or to statistical descriptions in larger communities. Therefore, approaches that can overcome the existing trade-off between mechanistic insight and ecological realism are needed. Dynamic food web modeling of predator-prey communities potentially allows for this, thanks to recent advances in body size – based parameterization of food web models. This approach needs to be developed further, however, to become a quantitative and predictive tool useful for understanding impacts of anthropogenic stress on community structure and stability, and how ecosystems might be managed to conserve the delivery of ecosystem services. In this project we aim to contribute to this goal by studying the role of body size and other species traits for accurately predicting trophic interaction strengths between species. These are crucial parameters of food web models and need to be established in order to predict the outcome of population dynamics. More specifically, we will formulate dynamic models of real tritrophic communities studied in parallel in controlled cage experiments at SLU, Uppsala, and aim to reproduce the observed population dynamics with the models. We will use a food web model which initially mechanistically relates predation rates (interaction strengths) to body size and abundance only, and then develop this by successively including effects of other species traits. This will be done in an iterative process where empirical observations are used to formulate model hypotheses and then testing them against the observed experimental dynamics. We anticipate that this will significantly advance existing approaches to parameterize dynamic food web models and subsequent analyses of the model should provide mechanistic insight into the function of predation and the service of biological control provided by this function.