Spatial networks of habitats, populations, and communities: connecting approaches to keep cutting edges

Paul Savary 

Purpose of review: Spatial networks are extensively used in ecology to represent exchanges among landscape features (e.g., habitat patches, river segments) or biological entities (e.g., individuals, populations, communities). I reviewed the literature produced in the past 25 years using these networks. Distinct types of spatial networks have emerged in several subfields of ecology. I aimed to assess whether this gave rise to disconnected research silos or, in contrast, whether methodological similarities generated bridges to connect theoretical frameworks.
Recent findings: I reviewed 679 papers using eight types of spatial networks. Habitat networks were the most used, usually for connectivity assessments with conservation-oriented purposes. In contrast, studies using metapopulation, metacommunity, or river networks were the most embedded in theoretical ecology. Population genetic networks were essentially used in landscape genetics, whereas dispersal networks and spatial networks found more diverse uses. Finally, meta-networks combining several of the above have more recently favored the integration of these typically disconnected approaches.
Summary: The lack of connection among research branches mobilizing spatial networks mainly stems from an opposition between applied and theoretical objectives, further reinforced by the differentiation of journal scopes. This divergence can create a mismatch between recent theoretical advances and current methodological designs, possibly affecting the tested predictions and result interpretations. Yet, the diversity of spatial networks is also beneficial. Provided it is acknowledged properly, future works could advantageously build upon existing frameworks for cutting-edge research, as exemplified by recent works on meta-networks.