Active reallocation of food-web interactions under environmental change

Background/Question/Methods

Degradation of natural habitats is a great threat to global biodiversity. Extinctions of component species and changes in species diversity have been widely documented, but studies on the effects of habitat change on interacting networks of species remain rare. In previous work, it has been shown that the architecture of species interaction networks can change dramatically following anthropogenic habitat modification. Similar changes are increasingly being documented in other food-web studies. Despite their importance for ecosystem functions and services, the mechanisms underlying these changes remain unknown. Furthermore, our ability to understand how food-web architecture will change following human modification of natural habitats has been held back by the absence of predictive models. Here, I describe analysis on host-parasitoid food webs collected in five contrasting locations, each location having a different gradient of modification between sampled habitats. I also present the first model of feeding reallocation that accurately generates the quantitative food-web structures found in impacted environments.

Results/Conclusions

I show that consumers in more altered habitats demonstrated resource selection in ways that cannot be accounted for by the availability and distribution of their hosts alone. In particular, we observed greater than expected consumer specialisation (50% of instances), and more generalist consumers tended to engage in greater levels of over-specialisation. When considering 4 quantitative food web metrics (weighted connectance, interaction diversity, generality and vulnerability), our model food web predictions were almost always within 5% of empirically observed values. The model is both simple and general, and suggests that increasingly severe gradients of habitat modification lead to increasingly complex consumer behaviour. While focusing on one of the multiple drivers of environmental change, our results are of broad relevance to understanding and predicting species invasions and their consequences, and for sustainable agriculture in regions that are heavily dependent on ecosystem services--such as biological control--that are associated with species interactions. Identifying general rules about the re-organisation of food webs has important implications for predicting the community-wide consequences of global environmental change.