Generalist plants are more competitive and more functionally similar to each other than specialist plants: insights from network analyses

Abstract

AbstractAimEcological specialization is defined by the variety of environments species occupy. Identifying the mechanisms that influence specialization is critical to understand patterns of species coexistence and biodiversity. However, the functional attributes that result in specialization are still unknown. Similarly, there is contrasting evidence between the degree of specialization and the local abundance of species. We investigated whether specialist and generalist plant species (a) are associated with distinct functional profiles, using core plant functional traits and strategies, (b) show similar functional variation and (c) perform at the local scale.LocationFrance.TaxonHerbaceous plants.MethodsWe analysed the structure of a bipartite network that includes the occurrences of ~2,900 plant species at ~90,000 sites to identify ecologically consistent sets of species and sites (i.e. ‘modules’). This innovative approach then enabled us to define a metric of specialization, by quantifying occurrences of species at sites that belong to one or several modules. We used functional traits related to resource acquisition, competition for light and dispersal ability, as well as indices of competitive, stress tolerance and ruderal strategies.ResultsWe identified five major modules in the bipartite network related to different environments and composed of species with differing functional attributes. Specialist species were less competitive and shorter, and had higher stress tolerance and stronger resource conservation, while generalist species were taller. Generalists were also more similar to each other than specialists. In addition, specialists had higher local abundances and occurred in communities with plants of similar height.Main conclusionsWe found distinctive functional signatures of specialist and generalist species in grassland communities across diverse environments at regional and community scales. By testing classic macro‐ecological hypotheses, network metrics can benefit community ecology by identifying distinct ecological units at a large scale and quantifying the links developed by species.