Functional trait patterns in grassland communities, and the importance of scale

For a long time, ecologists have worked on gaining insight into the processes that govern the assembly of natural plant communities. Plant trait-based approaches have great potential to improve our understanding of community assembly and assessment of ecosystem services. In many trait-based studies, trait-for-species substitutions are used by assigning mean trait values to each species in a community, which means that only between-species trait variability (i.e. interspecific trait variability) is considered. As there is growing evidence of the importance of within-species trait variability (i.e. intraspecific trait variability), this work is dedicated to studying how the applicability of these trait-based approaches is constrained by intraspecific variability and scale dependence. It could be shown that both interspecific trait variability and intraspecific trait variability indeed determine the trait patterns among habitats, communities and species. As a consequence, neglecting either type of trait variability by relying on species potential trait values derived from a much larger scale than the processes studied can lead to misleading conclusions, also on the community level. The benefits of using species mean trait values derived from large databases for a trait-based study will strongly depend on the level and scale of the question. In summary, there is a need to clearly differentiate between realized and potential traits on all levels in trait-based studies.