Can modularity buffer cascading effects in mutualistic networks?

Background/Question/Methods

Species are interconnected in the communities through ecological interactions, such as mutualisms. In these webs of interacting species, changes in one species can propagate throughout the community, leading to cascading effects. In this study, we investigate if modules (semi-independent groups of highly interacting species) within mutualistic networks can act as buffers against the spreading of cascading effects. We also investigate the hypothesis that species with greater number of interactions would propagate perturbations faster than species with smaller number of interactions. Our database was composed of 28 empirical mutualistic networks for which we had information on the interaction frequencies between pairs of species. Each species was assigned to a module through an optimization method and we also recorded the number of interactions of each species. For each network we performed 1000 numerical simulations in which we randomly perturbed one species and recorded which species started the perturbation and the time it took for the perturbation to reach all network species (spreading time). We then calculated mean spreading time of all species within a module and of all species with the same number of interactions.  

Results/Conclusions

Perturbation's spreading time and the number of interactions of a species were not correlated. In addition, we did not encounter a relationship between modules and perturbation' spreading time. Thus, our results suggest that modules do not buffer perturbation spreading. In theoretical work and in studies with food webs it has been argued that the presence of modules is related to the web' stability through the reduction of perturbation spreading. However, our results indicate that for mutualistic webs this might not be true, probably due to the higher connectivity between modules in many mutualisms.  Furthermore, since modules are usually organized around a few hub species, future work should focus on these hub species and evaluate their role in the cascading effects.