Unity makes strength: clustered extinctions have stronger, longer-lasting effects on metacommunities dynamics

In this article, Saade et al. (2021) investigate how the rate of local extinctions and their spatial distribution affect recolonization dynamics in metacommunities. They use an elegant combination of microcosm experiments with metacommunities of freshwater ciliates and mathematical modelling mirroring their experimental system. Their main findings are (i) that local patch extinctions increase both local (α-) and inter-patch (β-) diversity in a transient way during the recolonization process, (ii) that these effects depend more on the spatial distribution of extinctions (dispersed or clustered) than on their amount, and (iii) that they may spread regionally. Microcosm experiments are already quite cool just by themselves and have contributed largely to conceptual advances in community ecology (see Fraser and Keddy 1997, or Jessup et al. 2004 for reviews on this topic), but they are here exploited to a whole further level by the fitting of a metapopulation dynamics model. The model allows both to identify the underlying mechanisms most likely to generate the patterns observed (here, competitive interactions) and to assess the robustness of these patterns when considering larger spatial or temporal scales. This release of experimental limitations allows here for the analysis of quantitative metrics of spatial structure, like the distance to the closest patch, which gives an interesting insight into the functional basis of the effect of the spatial distribution of extinctions.