Classic theory set the stage for decades of work examining the relationship between diversity and stability that continues today. An emerging theme from modern diversity/stability research is that non-random structure in real communities imparts stability. Two structural components receiving extensive attention are 1) allometric scaling of interaction rates, where larger species eat smaller ones and experience other rates dependent on body-size, and 2) dispersal among locally distinct communities. It is also known that movements of top predators can couple and stabilize local food webs. However, the full effects of dispersal variation among species on food web stability are unclear, especially when other structuring elements are related to body size. We examined how body-size scaling of dispersal rates influenced stability in model trophic metacommunities. Realistic food web topologies were paired with interaction strengths parameterized in terms of scaled biomass fluxes and non-linear function shape parameters using a generalized modeling framework. Interaction strengths followed standard allometric relationships. Communities varied spatially and were connected by dispersal to form trophic metacommunities whose stability could be examined using linear stability analysis. We asked whether body-size scaling of dispersal rates influenced metacommunity stability relative to dispersal that was either uniform or randomly varying among species.
Results/Conclusions
Linear stability varied dramatically over the million-plus food webs examined, depending on flux-scale parameters, shape parameters, and dispersal rates. We found that body-size scaling in dispersal generally had a strong stabilizing effect on trophic metacommunities, particularly when dispersal was highest in large-bodied top predators. Cases where species had random (non-allometric) variation in dispersal rates or other body-size/dispersal relationships showed less consistent stabilizing effects. Scale and shape parameters had effects on local stability of food webs that were consistent with previous studies. These locally stable food webs provided stability to the metacommunity when linked to unstable webs, and this effect was greatly enhanced when dispersal scaled positively with body-size. Our results reinforce results previous modeling, namely that stability is promoted by 1) spatial heterogeneity in local food web interactions and 2) coupling of local food webs by top predators. Our results show that realistic patterns of dispersal rate variation among individual species enhance these effects. Our ongoing work is expanding to include competition/colonization trade-offs among trophically similar species and to different types of spatial network structures. We conclude that incorporating realistic patterns of dispersal variation is a crucial next step in understanding metacommunity stability.