Emergent Encoding of Species Interactions in Spatial Metapopulation Networks

Species coexistence in spatially structured ecosystems depends on the interaction between local processes, such as competition and mortality, and regional processes, such as dispersal. Metacommunity models describe these systems as local communities connected by dispersal networks, allowing the effects of spatial structure on species persistence and coexistence to be investigated. This thesis studies a competitive metacommunity in which species disperse between habitat patches and compete for a finite amount of available space. In homogeneous environments, the pathogen-free model generally leads to monodominance, while coexistence is possible only under fine-tuned conditions. We investigate how this result is modified by species-specific density-dependent mortality, interpreted as the effect of host-specific pathogens. By acting more strongly on locally abundant species, this mechanism can limit competitive dominance. When pathogen dynamics are fast relative to host demography, it gives rise to an effective quadratic mortality term in the metacommunity dynamics. We show analytically that this additional mortality allows stable coexistence within a finite region of parameter space. As pathogen pressure increases, competitively dominant species are increasingly limited, allowing previously excluded species to establish and increasing species richness. However, the same mortality mechanism also reduces population abundances and, when sufficiently strong, suppresses even the species that initially benefit from reduced competition. We derive the critical pathogen strengths associated with these transitions. This reveals a coexistence--abundance trade-off: pathogen-mediated self-limitation can promote species richness while reducing total community abundance. The analysis is extended to heterogeneous habitats through a perturbative mean-field approach. Numerical results show that pathogens generally promote coexistence, reduce total abundance and smooth spatial differences, except for localization jumps associated with newly invading species. Finally, simulations using effective kernels constructed from explicit dispersal networks confirm that the main coexistence--abundance trade-off persists beyond the mean-field approximation. Overall, these results identify species-specific natural enemies as a stabilizing mechanism complementary to environmental heterogeneity.