Drivers of complexity in ecosystem restoration

Abstract

Ecological complexity is linked to emergent properties, such as ecosystem processes and resilience. Enhancing complexity may therefore be a more effective goal for ecosystem restoration than the traditional focus of replicating reference communities, especially under global environmental change. However, ecological complexity is rarely empirically studied in real-world systems, so the potential applications of complexity and the relationships between different aspects of complexity are poorly understood. Here, we examine complexity in the context of restoration within two contrasting ecosystems, 54 calcareous grasslands and 60 broadleaved woodlands. We derive ‘ecological complexity’ indices for multiple ecosystem attributes, including soil bacteria, soil fungi, habitat structure, plants, invertebrates and species networks. Using Structural Equation Models, we quantify the effect of key drivers (including site age and size, amount of grassland/woodland in surrounding landscape, former land-use, establishment method and ongoing management) on complexity and the strength of associations among different aspects of complexity. We find that decisions made at the onset of restoration have a large role in the subsequent development of complexity, including grassland establishment method and former land-use for woodlands. High structural complexity was identified as a positive driver of invertebrate complexity in both habitats, but in general we observed limited relationships between the different complexity aspects. Policy implications. This study shows that ecological complexity is empirically measurable, and its development in restoration can be influenced by key variables, which have mechanistic underpinnings. The foundational role of structural complexity driving other aspects of complexity could guide management efforts to enhance ecological complexity. Our results suggest, however, that ecological complexity is multi-faceted and cannot be reduced to a few simple indicators. Thus, to gain a holistic understanding of how complex an ecosystem is, one needs to measure multiple aspects rather than focusing on, for example, taxonomic diversity of a few taxa. One approach is to consider a syndrome-focussed framework, whereby a set of complexity measures provides a good indication of ecosystems status and resilience.