• architectural traits;
  • clonal plants;
  • community diversity;
  • determinants of plant community diversity and structure;
  • growth traits;
  • parameterized model;
  • process from pattern inference;
  • trait manipulation


  1. The degree of dispersion of trait values among species in a community has frequently been used to infer processes of community assembly. However, multiple assembly processes can lead to the same pattern of trait dispersion or the same process can lead to different patterns of dispersion. In particular, competitive processes can lead both to trait overdispersion (if the trait controls niche differentiation and only substantial differences allow coexistence) or to trait underdispersion (if the trait controls position in a competitive hierarchy and only similar values enhance coexistence).
  2. Because different traits are likely to contribute to stabilizing (niche differentiation) and equalizing (competitive hierarchies) mechanisms of coexistence, we compare the role of dispersion of a number of individual traits for species diversity by conducting in silico experiments using field-parameterized spatially explicit models of communities of clonally growing plants. We manipulate both dispersion and means of different traits and examine consequent changes of species diversity in the whole community. We hypothesize that growth traits, which are directly linked to resource acquisition, are likely related to position in competitive hierarchies and thus diversity and trait dispersion will be negatively associated. In contrast, we hypothesize that architectural traits, which control spatial deployment of new plants and are thus less directly linked to resource acquisition, are more likely to be linked to niche differentiation and thus diversity and trait dispersion will be positively associated.
  3. Individual traits differed considerably in effects of trait dispersion on community diversity. Specifically, increasing dispersion in growth traits often decreased diversity, presumably reflecting widening competitive differences and exclusion of weaker competitors, which is consistent with the action of equalizing mechanisms of species coexistence. In contrast, increasing dispersion in architectural traits either increased diversity or had no effect, which possibly indicates niche-based mechanisms of diversity maintenance. Changes of community-wide trait means can have large effects on diversity for a given degree of dispersion, although not usually changing the sign of the relationship.
  4. Synthesis. These results suggest inference of community assembly processes from patterns of trait dispersion without understanding how particular traits function in community assembly may often be misleading. Effects of dispersion of traits that are likely associated with position in a competitive hierarchy are very different from those of traits associated with niche differences.