Several mechanisms can explain the formation of spatial patterns of vegetation, particularly in systems in which resource scarcity prevents full plant cover, such as drylands. The physical-geomorphological template determines where plants can and cannot grow. But, within plant-amenable soil patches vegetation patterns can emerge either from self-organization, due to plant-plant-resource interactions, or as a result of plant-animal-resource interactions. We therefore ask: (1) What are the conditions necessary for self-organized vs. animal-mediated vegetation patterns? (2) Can different scale-dependent vegetation patterns coexist within the same landscape? And (3) What are the ecological consequences of these patterns for biodiversity, ecosystem function and resilience. We used a combined soil-plant-animal model to evaluate the coexistence of vegetation pattern formation as a result of self-organization of both plant and social insect colonies, validated using imagery data from various field sites.
Our analyses highlight the importance of spatial scales for understanding vegetation patchiness. First, self-organized plant-plant-water interactions are relevant only in soil patches that are large enough to sustain hundreds of interacting individuals. Second, the typical size of vegetation patches in the system usually reflects the mechanism that explains the formation of vegetation patchiness. For example, small-scale vegetation patterns are caused by belowground competition-facilitation for water, and therefore reflect the scale of plant root systems and water redistribution; whereas, evenly distributed large-scale vegetation spots (or gaps) can be caused by the activity of subterranean ecosystem engineers (e.g., termites and ants), and therefore reflect the scale of foraging and territoriality in these animal colonies. The model and field data demonstrate that intraspecific competition between territorial animals can generate the large-scale hexagonal regularity of these patterns. Furthermore, we found that coexistence of multi-scale patterns contributes to dryland ecosystem functions such as enhanced resistance to and recovery from drought. Overall, our findings emphasize the potential for co-occurrence and complementarity among distinct patterning mechanisms in generating multi-scale regularity. The potential influence of multi-scale vegetation patterns on biodiversity, as a result of different species filtering processes, remains to be tested.