Résumé
Community Weighted Means (CWM) are valuable tools describing community composition with respect to one given trait. They have been widely used as indicators in global change studies to measure biodiversity responses to environmental perturbations. However, how individual species contribute to such community indicators has hardly been investigated. One of the reasons lies in the absence of a methodological framework relating changes in community dynamics to species-specific population variations. Here, we present a comprehensive framework allowing a finer interpretation of changes in CWM, and we propose a way to compute species contributions to these indicators.We present an analytical framework allowing the quantification of species-specific contributions to changes in the mean (CWM) and the variance (Community Weighted Variance, CWV) of trait distributions in species assemblages monitored through time and space. We apply this approach to a case study investigating the impact of climate change on common bird assemblages in the French Mediterranean area between 2001 and 2012.This approach allows us to identify that a small proportion of the species drive the changes observed at the community level indicator, and allows the identification of those species. Moreover, we show that the species-specific contributions are not homogeneous between taxonomic groups and that migratory species tend to have a higher impact.This novel decomposition and interpretation of Community Weighted Means and Variances (for which specific software package is provided along with this article) sheds new light on the drivers of community modifications in response to environmental changes across time and space. Moreover, it represents a relevant and simple way to assess particular aspects of species-specific responses to environmental changes and it is straightforward to use for widely used ecological data on any species group.