Revealing perturbation responses with limited observations of biological communities

Restrictions in empirical research of biological communities have limited our understanding the combined influence environmental variability and system structure on community composition. Spatial patterns composition less accessible systems, such as marine benthos, can often not be explained by many factors beyond direct impact environment members. We present a method that combines commonly collected data with analyses qualitative mathematical models, to assess only impacts variability, but also propagation through complex interaction networks. Transformed spatial describe members’ observed similarity response an external input. The output models describes predicted input entering any its variables. A statistically significant agreement between similarities indicates respective variable likely gateway for into system. is applied benthic macroinvertebrate Rance estuary (Brittany, France). Organisms identified gateways traits agree their documented spatially temporally structured variability. suggest use this novel framework more comprehensive identification drivers change, including members cascades environmentally driven change structure.