The novel framework of phylogenetic and trait-based community structure: distinct trait roles
Since Webb et al. (2002) raised the framework of phylogenetic community structure, mechanistic understanding on community assembly has experienced remarkable progress in the last decade. However, phylogenetic approaches highly rely on phylogenetic relatedness being a comprehensive proxy of ecological similarity (Swenson 2013; Losos 2008), which implies the necessity of phylogenetic-trait intergration in revealing mechanism underlying community assembly.
    Distinct functional role and contribution of traits in assembling species has been mentioned in earlier empirical and theoretical studies (Kraft & Ackerly 2010; Swenson & Enquist 2009; Grime 2006; Thompsonet al. 1996). Swenson & Enquist (2009) and Grime (2006) have detected that plant traits related to productivity often appeared clustered in local communities, while disturbance and regeneration related traits displayed locally over-dispersed. Kraft & Ackerly (2010) have ever predicted that functional traits related to resource use and environmental tolerance were more likely to answer habitat filtering, resource use strategy traits were more sensitive to interspecific competitive exclusion and niche differentiation, physical and chemical defense traits displayed a major role in responding to enemy-mediated negative density dependence. These information together have presented an apparent signal that part of traits are more important (predominant traits) in responding to a certain ecological process, and the others performed relatively less sensitive (assistant traits) response to assemblying process. Absolutely, either predominant or assistant traits are indispensable niche components at both species and community levels. Actually, it appears extraordinarily challenge to decide which functional attributes display a more important role than the others, as they work as an alliance in most of cases (Zukswert & Prescott 2017; Grassein et al. 2014).
    Based on prior understanding on community structure, we rethink over community assembly process by taking account of distinct functional role (predominant and assistant role) of phenotype. Through simulating assembly process in simple habitat, we found that phylogenetic dispersion was accordant with the functional dispersion of conserved predominant trait, and opposed to the structure of convergent predominant trait. This result of predominant traits could also occur in assistant trait, especially when assistant trait presented similar ecological and evolutionary chacteristics (Table 1). But in most instances, the dispersion of assistant traits appeared unpredictable in simple habitat. By comparison, phylogenetic and trait dispersion in complex habitat appear extremely complicated, as community structure is resulting from similar or/and distinct assembly processes acting on conserved or/and convergent functional attributes. Therefore, it seems impossible to predict the functional role of a certain trait in structuring community within a complex habitat. These patterns have also provided two reasonable explanations for the incongruent dispersion between conserved trait and phylgoenetic relatedness (Du et al.2017; Yang et al. 2014; Swenson & Enquist 2009): the trait might act an assistant role organizing community in simple habitat; or the community is defined upon a heterogeneous habitat.