Trait–environment relationships depend on species life history

Abstract

AbstractQuestionPredicting the functional response of vegetation to environmental variations remains highly challenging. A first reason is that trait–environment relationships result from joint intraspecific and interspecific responses. A second reason is that most plant communities host sets of species that are expected to display different environmental responses according to their ecophysiological strategies.LocationMediterranean rangeland, Larzac Plateau, France (“La Fage” INRAE experimental station: 43°55′ N, 3°05′ E).MethodsIn a Mediterranean rangeland, we tested whether changes in trait values in response to management intensification (increased resource availability and grazing intensity) differ between annuals and perennials that coexist locally. To capture the multiple facets of plant phenotypes, we studied nine traits related to carbon acquisition and conservation, plant size, water‐use efficiency, phenology and reproduction.ResultsFor most traits, we evidenced a significant interaction between environment and life history. In particular, changes in trait values were higher in perennials for leaf traits related to tissue density and growth. These differences were explained by: (a) a lower species turnover in annuals, and (b) the presence of species with more distinct trait values across environments in perennials. For most traits, these changes were not accompanied by differential intraspecific variations between annuals and perennials, which invalidates previous theoretical predictions. In addition, the contribution of intraspecific trait variation to among‐communities trait variance was higher for annuals than perennials for most traits.ConclusionAltogether, our findings highlight that changes in trait values with the environment can depend on functional groups (here, life history), which can further critically impact the assessment of the functional response of communities. They challenge the mean field approaches to communities in which trait aggregation is most often blindly applied, without considering potential different trajectories in response to environmental changes at various nested organizational levels.