The non-habitat landcover or matrix of a landscape is known to alter species’ responses to habitat loss and fragmentation. Local processes, resulting from the matrix directly surrounding a patch, affect populations through edge contrasts and spillover of both abiotic and biotic factors. Regional processes from the landscape matrix affect populations during dispersal through differences in movement resistance or mortality. However, the relative magnitude of these matrix effects operating at two different scales in the landscape is unknown. Interaction between these scales may be driving discrepancies in results of matrix studies, and if so, understanding this relationship is vital to predicting species’ responses to landscape changes. In order to isolate the matrix effects from other landscape processes, we conducted a large-scale experiment across 12 replicate landscapes in a paired design, using the Opuntia cactus and cactus arthropod system to model the response of fragmented populations to matrix change. We manipulated both the local and regional matrix compositions and followed cactus arthropod populations across two years and five generations.
Results/Conclusions
We found that populations responded more intensely to changes in regional matrix conditions. The local-scale matrix effects interacted strongly with these regional effects and altered populations’ responses to local matrix composition. Our results allowed us to disentangle these multiscale processes and explain contradictory results found previously in correlative studies of matrix effects across landscapes.
