Functional and phylogenetic diversity of plant communities differently affect the structure of flower-visitor interactions and reveal convergences in floral traits

Community ecology has moved from descriptive studies to more mechanistic approaches asking questions about causes and consequences of community composition and interactions between species. Many ecological processes are shaped by the presence or absence of functional groups, not necessarily species. Thus, the diversity of functional traits, i.e. their interspecific variation, is a key feature of plant communities with consequences on other trophic levels. In a simulation study based on a quantitative flower-visitor network and quantitative measurements of flower traits, we tested how the functional FDiv and phylogenetic diversity PDiv of plant communities affect animal species richness and diversity as well as network properties. Within the limitations of the assumption that plants maintain the qualitative and quantitative interactions with animals in subsampled communities, we found that functionally diverse plant communities support a large number of animal species (not necessarily animal diversity). Additionally, the network structure was more complementarily specialized (higher -values) and comprised a larger number of unrealized links (low connectance) and thus a higher partitioning of resources among consumers in functionally diverse plant communities than in communities with a lower FDiv. For the phylogenetic diversity PDiv of plant communities we found contrasting effects, which may be explained by divergences or convergences of functional traits. Our results support the notion that functionally diverse plant communities offer a large number of niches that can be occupied by a larger number of flower visiting species specialized to a specific set of flower traits. Thus, functional flower traits serve as barriers that exclude some flower visitors but also as attractive features that facilitate interactions with other animal species. Our study fosters a trait-based definition of niches and functional groups and may stimulate field studies testing the predictions of this simulation.