Ugo Bastolla
Bioinformatics Unit, Universidad Autonoma de Madrid, Spain

Which properties of ecosystems favour their stability against environmental perturbations and help to maintain biodiversity? We address this question considering model ecosysytems of plants and pollinators that combine intra-group competition and inter-group mutualism, and adopting as a tool the structural stability of the system against global perturbations of dynamical parameters. We show that structural stability can be analytically predicted through two quantities, the effective interspecific competition and the propagation of perturbations. The mutualistic network architecture and parameters affect these control variables in a complex way that rationalize previous contradictory results. In particular, mutualism reduces the effective competition only when the direct competition between species in the same group is weak, and the propagation of environmental perturbations is reduced in highly connected networks, a mechanism reminiscent of MacArthur's proposal that ecosystem complexity enhances stability. Finally, we present an analysis of co-occurrence patterns in environmental samples of bacterial OTUs (Operational Taxonomic Units). Comparison with a recently proposed null-model of absence of interactions, that we modified to take into account environmental preferences, show a larger number of significant co-occurrences (aggregations) than exclusions (segregations). Aggregations are little likely to be explained by environmental preferences not considered in the null model, and many of them correspond to known examples of synergistic interactions. Cosmopolitan bacteria found in many different environments are more prone to aggregation, suggesting that part of their aggregations reflect ecological interactions that favour their remarkable cosmpolitism. Aggregations are more frequent for phylogenetically related species, and we discuss their possible role in bacterial speciation at the light of the recently proposed Black Queen Hypothesis that addresses the evolution of metabolic dependencies between bacteria.

seminar schedule