Introducing desirable patches: A nuclear option for ecosystem restoration?

Maarten B. Eppinga, Department of Geography, University of Zürich, Zürich, Switzerland, Theo Michaels, Department of Ecology and Evolutionary Biology, University of Kansas, Lawrence, KS, Maria J. Santos, Geography, University of Zurich, Zurich, Switzerland and James D. Bever, Ecology and Evolutionary Biology, University of Kansas, Lawrence, KS

Background/Question/Methods:
Critical transitions between alternate stable states provide an explanation for rapid ecosystem degradation, yielding important implications for ecosystem conservation and restoration. However, applications of the alternative stable states framework may be impeded by a mismatch between patch-scale driving processes and landscape-scale emergent system transitions. Recent advances in ecology have identified three types of spatial mechanisms linking patch-scale processes with landscape-scale dynamics: i) auto-catalytic nucleation; ii) directed dispersal; and iii) resource concentration. Until now, these mechanisms have been studied separately, using conceptual and theoretical frameworks that differ in the level of detail considered and the formalisms used. As a result, we currently lack a systematic assessment procedure to identify the potential of each type of patch-scale driving process for restoring a particular ecosystem of interest.
Here we present a spatially explicit modelling framework with sufficient flexibility to consider all three types of patch-scale driving processes. We utilize this framework to establish for each process how its occurrence is reflected in the relationship between patch size and patch dynamics, how this relationship is mediated by patch geometry, and how the resulting patch dynamics depend on the relative position of the patch within the landscape.

Results/Conclusions:
We found that each of the three patch-scale driving processes was characterized by a unique set of relationships between patch size, geometry and position in the landscape on the one hand, and the resulting patch dynamics on the other hand. For example, auto-catalytic nucleation was the only process that exhibited accelerating rates of expansion with increasing patch size, and was not affected by a patch’ relative position in the landscape. As a result, the three mechanisms also yielded different trade-offs between introduced patch size and the restoration benefits obtained. Thus, knowing the relative importance of each type of patch-scale driving process is an important prerequisite for optimally leveraging local dynamics to reach landscape-scale restoration targets.
Our work synthesizes recent insights in the ways in which the introduction of desirable patches can be used to accelerate ecosystem restoration processes. The presented findings also suggest a set of diagnostic experiments to assess the relative importance of each type of patch-scale driving process within a particular ecosystem, to be tested in future empirical studies.