The dynamics of community and ecosystem assembly in systems undergoing rapid changes in connectivity are poorly understood. Existing metacommunity paradigms focus on fixed levels of connectivity. However, local communities that have assembled in isolation can become rapidly mixed under events such as flooding or habitat defragmentation. These events are ideal for testing ecological theory concerning 1) the impact of local assembly, 2) community-level interaction and 3) the resilience of both composition and emergent attributes to mixing. A limited body of theory suggests that sudden mixing can result in ?community-level? competition and that trophic structure enhances this effect. I investigated these dynamics using protist microcosms and a completely described master food web involving 48 species of producers, herbivores, omnivores, and predators. Dividing this web into separate species pools, I assembled pairs of random eight-species communities with either high trophic (T) or high competitive complexity (C). I divided 61 pairs (132 total communities) into equal numbers of CC, TC and TT community combinations. After 1.5 months, I mixed these pairs and assembled the mixture for an additional 1.5 months. I recorded composition and abundance, food web properties and trophic structure both at the time of mixing and at the end of the experiment.
Results/Conclusions
Prior to mixing, species coexistence probabilities were significantly greater in competitive communities, but greater in trophic communities after mixing. In mixed communities, overall persistence was greater. However, persistence probabilities for edible algae and certain bacterivores declined significantly in mixture. The proportional contribution of each community to post-mixing extinctions was generally symmetric (equal extinction rates for species added from each paired community), but completely asymmetric in rare cases. This meant most final communities contained a mixture of species from pre-mixing communities. Post-mixing food web connectance and other emergent attributes were more asymmetric compared with pre-mixing communities. These results suggest 1) the effects of mixing depend on the community attribute considered and assembly history and 2) the resilience of attributes depends on species turnover.