Metacommunity resilience against simulated gradients of wildfire: disturbance intensity and species dispersal ability determine landscape recover capacity.

Disturbances, such as wildfires, are one of the main drivers of biodiversity dynamics, and their frequency and intensity are expected to increase due to global change. Such disturbances generate a mosaic of affected and unaffected patches that change landscape structure and, consequently, metacommunity networks. Therefore, to fully understand the consequences of such disturbances, a landscape perspective is required. In 2012, a wildfire burned 13 000 hectares in the NE Iberian Peninsula affecting a pond macroinvertebrate metacommunity. Communities were highly resilient to this disturbance, recovering after one hydroperiod. Their resilience was related to dispersal, being lower in species with weak dispersal abilities than those with strong dispersal abilities. This suggested that the metacommunity network played a major role in defining system resilience. In this context, we introduced a theoretical analysis based on this network in which we evaluated metacommunity resilience across several gradients of disturbance size and intensity incorporating species dispersal ability. Our study supports the empirical observation of a highly resilient metacommunity but also reveals that increased disturbance regimes might lead to a collapse of this resilience. Disturbance size and intensity interacted to determine the community recovery rate, which was high when both variables were low. Nevertheless, the transition from high to low resilience was sharp and depended on species dispersal. Diversity recovery was mostly driven by disturbance intensity, abruptly collapsing with its rise. This response highlighted the qualitative difference in the effect of size and intensity. These results not only illustrate the mechanisms shaping the studied metacommunity but also more generally stress the strong role of metacommunity mechanisms and landscape structure in biodiversity resilience. Finally, this study highlights the importance of using theoretical approaches rooted in empirical data to determine metacommunity dynamics and the need to preserve and build connected and heterogeneous landscapes to address future disturbance scenarios. [ABSTRACT FROM AUTHOR]