The structure of ecological networks

Ecological networks describe who interacts with whom in ecosystems — which preys are eaten by which predator, which plants being pollinated by which insect or which hosts getting infected by which parasitoid. The summarized pattern of all interactions between species forms a complex network where the nodes represent the species and the edges represent the interaction links. The network perspective has been shown to be extremely useful in analyzing the functioning of a wide variety of systems of interacting objects covering many domains of active research, including ecology. The network approach allows us to study both large scale statistical properties of the whole network, as well as properties of individual species and interactions and how they joint contribute to structure, function and stability of ecosystems.

In my research I am interested in which are the factors driving species interactions — what determines the presence or absence of interactions? Species interactions are determined by traits: a species will only interact with another species if they have matching traits, for example the right body size for the right size of mouth or the right proboscis length for the right corolla depth or live at the same depth in the water column. Rather than a single trait, multiple traits may be required to determine an interaction, thereby giving rise to a ”niche-space”. I am interested in identifying the most important traits and understand how this contributed to the overall network structure, population dynamics and ecosystem functioning. I work with empirical data from a wide variety of ecological networks in collaboration with both empiricists and other theoreticians.

Species extinctions

Extinction of species in an ecosystem can have extensive consequences on the remaining community with possibility of a cascade of secondary extinctions. At present, there are mainly two directions of research analyzing species extinctions and their effect on the rate of secondary extinctions. The first direction is purely topological and a species goes secondarily extinct only when it loses all of its prey. The second direction use models where population dynamics can be tracked. These dynamical models make it possibly to capture also other scenarios than the pure topological, such as top-down effects. In my research I use both approaches. However, I also work on developing an intermediate approach using Bayesian Networks (also named belief networks) which is a graphical model based on probabilistic relationships.

Ecological networks in a spatial perspective

The structure and the dynamic behavior of ecosystems are affected by both local and regional processes, such as population dynamics and species dispersal. I am interested in how the spatial scale we are focusing on have implications for both ecosystem structure and dynamics. I work on models that i) analyzes how much of the network structure that cab be assigned to depend on spatial versus trophic factors and ii) analyze the effects of spatial parameters (for example patch distribution and species dispersal) on population and metacommunity dynamics.