Effects of inter-specific variability on biomass allocation: A hierarchical Bayesian approach

Abstract

The partitioning of total biomass among plant organs (stem, leaf, root, and shoot, respectively) has been a central topic in functional ecology. Many studies referring to inter-specific biomass allocation mainly concentrate on the completely pooled data treating all species involved equivalently. Less work has incorporated species differences into such allocation patterns. In the present paper, we explored the potential effects of species variability on biomass allocation patterns in a hierarchical Bayesian framework, which assigned individual species separate allometric values, drawn from ‘global’ parameters on the community level. Results demonstrated that the inclusion of species identity could significantly decrease the allocation slopes compared to those from the corresponding non-hierarchical models on pooled data, and much less residual variation was left for hierarchical models. Our work exhibits the importance of species variability on biomass partition and the potential impact of sample size on allometric estimates, and indicates that the inclusion of certain complexity such as species difference is a necessary step to better understand the scaling theories in ecology.