Plant–plant interactions scale up to produce vegetation spatial patterns: the influence of long‐ and short‐term process

Vegetation spatial patterns emerge in response to feedback interactions between organisms and their environment, because of the redistribution of water and nutrients around the plant canopy or as a consequence of facilitation/competition interactions at the plant level, even in the absence of pre‐existing substratum heterogeneities. It has been suggested that changes in vegetation spatial patterns are a signal of transition shift in ecosystems. Understanding the factors that lead to aggregated spatial patterns and control the transition to random distributions requires that environmental and species information is taken into account. In this study, we investigated the relative contributions of aridity (a long‐term process), to which vegetation is adapted, and the area covered by bare soil (short‐term process) to plant–plant associations and their contribution to aggregated spatial patterns. The study was conducted in a gradient of aridity ranging from that in subalpine grassland habitats in the Pyrenees and Sierra Nevada mountains to that in the semiarid steppes of Cabo de Gata and the middle Ebro Valley in Spain. We compared sites that differed in aridity and a geophysical feature (north‐ vs. south‐facing slope). We observed that the relative contribution of aridity and bare soil to plant–plant facilitation and vegetation aggregation differed in subalpine and semiarid areas. Facilitation in subalpine habitats had a marked effect on aggregated spatial patterns, while aridity contributed to disruption of these patterns. Conversely, in semiarid habitats, the disruption of aggregated patterns was mainly promoted by an increase in bare soil area rather than in aridity. In semiarid habitats, the higher level of stress on south‐facing slopes increased facilitation interactions relative to north‐facing slopes, although this did not enhance the persistence of aggregated spatial patterns. We conclude that the use of aggregated spatial patterns as an indicator of ecosystem shift must distinguish and separately take account of long‐term processes to which vegetation adapt, and short‐term process.