Akademska digitalna zbirka SLovenije - logo
E-viri
Celotno besedilo
Recenzirano Odprti dostop
  • Loss of habitat and connect...
    Thompson, Patrick L.; Rayfield, Bronwyn; Gonzalez, Andrew

    Ecography (Copenhagen), January 2017, Letnik: 40, Številka: 1
    Journal Article

    Habitat loss fragments metacommunities, altering the movement of species between previously connected habitat patches. The consequences of habitat loss for ecosystem functioning depend, in part, on how these changes in connectivity alter the spatial insurance effects of biodiversity. Spatial insurance is the maintenance of biodiversity and stable ecosystem functioning in changing environments that occurs when species are able to move between local habitat patches in order to track conditions to which they are adapted. Spatial insurance requires a combination of species sorting dynamics, which allow species to disperse to habitats where they are productive, and mass effect dynamics, where dispersal allows species to persist in marginal habitats where environmental conditions do not support growth. Here we use a spatially explicit metacommunity model to show that the relative contribution of species sorting and mass effects to spatial insurance changes with the rate of dispersal. We then simulate different sequences of habitat loss by removing habitat patches based on their betweenness centrality (the degree to which a patch serves as a connection between other patches in the metacommunity). We demonstrate that the sequence of habitat loss has a large, non‐linear impact on diversity, ecosystem functioning and stability. Spatial insurance is lost because habitat fragmentation impedes species sorting, while promoting mass effects and dispersal limitation. We find that species sorting dynamics, and thus spatial insurance, are most robust to the removal of habitat patches with low betweenness centrality. These findings advance our understanding of how habitat connectivity facilitates the maintenance of biodiversity and ecosystem functioning, and may prove useful for the design of habitat networks.