Robustness of mutualistic networks under phenological change and habitat destruction
Global changes, such as climate change and habitat destruction, threaten biodiversity and ecosystem functioning. Climate change can alter species phenologies, weakening species interactions. Meanwhile, habitat destruction can reduce biodiversity and population viability. The effects of these factors have been studied separately in mutualistic networks and other ecological networks. However, we still know very little about their potential effects on the diversity and structure of mutualistic networks when both factors act simultaneously. We developed a spatially explicit, stochastic, metacommunity model, to explore the effects of habitat destruction and phenological shifts on the diversity and structure of mutualistic networks. In our model all species are obligate mutualists. The presence of a species in a particular site depends on the strengths of its mutualistic interactions in the local network, as well as on the colonization from other sites. The strength of a mutualism depends on the temporal overlap between the species along a year, which is taken from empirical data at different times. This allows us to asses the consequences of realistic changes in phenology.
Habitat destruction causes the local loss of interactions, which in turn leads to a uniform decline in local diversities with the number of sites destroyed. Meanwhile, global diversity remains high until non-destroyed, species poor, sparsely distributed sites, can no longer act as dispersal sources, and the whole metacommunity suddenly collapses at a critical level of habitat loss. Given an external pool of species, the recovery of the metacommunity is hampered by hysteresis, i.e. the number of sites to be re-created must the higher than the number destroyed. Small changes in phenology can cause some increase in network connectance, thanks to the appearance of a few novel interactions involving generalist species, helping to cope with habitat destruction. However, moderate to large changes tend to destroy or weaken many interactions, making metacommunities vulnerable to collapse with lower levels of habitat loss. Overall, we conclude that habitat destruction and phenological change act synergistically, i.e. their combined effects are larger than the sum of their separate effects. Thus we predict, that the combined detrimental effects of reduced phenological and spatial overlap between mutualists, such as plants and pollinators, can be more damaging than expected.