Ecological networks in motion: micro- and macroscopic variability across scales

There has been an intense focus on the response of species to environmental changes, and more recently, the interactions of species have been examined in a similar way in order to understand the stability of entire communities and networks of interacting species. As a consequence, ecological networks have been placed in spatial and temporal contexts in order to reveal what may drive network variability. Understanding the spatial and temporal variability of ecological networks, and in particular the underlying forces facilitating changes, seems pertinent in our attempts to understand and anticipate how ecological networks may vary and respond to future environmental scenarios. Network variability has been studied at widely differing temporal and spatial scales. For example, studies exploring temporal variability ranges from within-season comparisons to comparisons over vast geological time spans, and the spatial extent ranges from the scale of a single pond to global analyses. Here, we highlight the outcomes from such studies and emphasize the identified mechanisms driving spatio-temporal variability in ecological networks. Specifically, we describe how ecological networks vary over different temporal (years, centuries and millennia) and spatial (local, regional and global) scales, discuss how this variability is monitored and identify potential future directions. Present knowledge allows some tentative generalizations. First, ecological networks tend to exhibit considerable spatial and temporal stability in several macroscopic features (e.g. connectance, nestedness), but studies also show that macroscopic features may change, for example, in relation to mass extinction or steep environmental gradients. Secondly, microscopic features (e.g. individual specialization levels, species roles and partner affiliations), albeit less studied, seem to show strong variability, and in several cases, microscopic instability co-occurs with macroscopic stability. We therefore recommend a stronger focus on this macro–micro interplay and list ideas (e.g. temporal species centrality measures and interaction phenologies), towards expanding the microscopic toolbox of network ecologists. A lay summary is available for this article.