Short-term responses of ground-dwelling arthropod communities to habitat loss and fragmentation in experimental mesoscale landscapes.

Habitat loss and fragmentation are main drivers of biodiversity decline. However, disentangling their respective effects remains tricky and contrasting results have emerged in previous studies leading to a call for finely designed landscape-scale experiments that compare different levels of habitat fragmentation at fixed habitat loss, and that for different levels of habitat loss. Here we present early-stage results from MESOLAND, a new landscape-scale experiment designed to monitor the response of ground-dwelling arthropod communities to habitat loss and fragmentation, focusing on their short-term responses. Our experiment took place in a dry grassland in France, with natural stone cover representing the habitat and bare soil with stone cover manually removed being the matrix. We thus created experimental landscapes with nine levels of habitat loss (from 0 to 99%) combined with three levels of fragmentation (low, medium, high). Ground-dwelling arthropod communities were monitored using dry pitfall traps during six non-lethal capture sessions. In the months following habitat removal, we observed a drop in the number of captured arthropods with increasing habitat loss. Humidity-dependant groups such as woodlice and silverfish were affected. We observed no effect of habitat fragmentation at any level of habitat loss and no interactive effect with habitat amount. Early-stage results following the implementation of the MESOLAND experiment indicate that habitat loss has a greater effect than habitat fragmentation on ground-dwelling arthropods communities. We expect detecting further effects in the years to come, as most species will have completed several life cycles. The results of this experiment will contribute to the ongoing debate on habitat loss versus fragmentation, providing essential knowledge for applied spatial conservation.