Phylogenetic diversity metrics can discern the relative contributions of ecological and evolutionary processes associated with the assembly of plant communities. However, the magnitude of the potential variation associated with phylogenetic methodologies, and its effect on estimates of phylogenetic diversity, remains poorly understood. Here, we assess how sources of variation associated with estimates of phylogenetic diversity can potentially affect our understanding of plant community structure for a series of temperate forest plots in China.

In total, 20 forest plots, comprising of 274 woody species and 581 herbaceous species, were surveyed and sampled along an elevational gradient of 2800 m on Taibai Mountain, China. We used multi-model inference to search for the most parsimonious relationship between estimates of phylogenetic diversity and each of four predictors (i.e. type of phylogenetic reconstruction method, phylogenetic diversity metric, woody or herbaceous growth form and elevation), and their pairwise interactions.

There was no significant difference in patterns of phylogenetic diversity when using synthesis-based vs. molecular-based phylogenetic methods. Results showed that elevation, the type of phylogenetic diversity metric, growth form and their interactions, accounted for >44% of the variance in our estimates of phylogenetic diversity. In general, phylogenetic diversity decreased with increasing elevation; however, the trend was weaker for herbaceous plants than for woody plants. Moreover, the three phylogenetic diversity metrics showed consistent patterns (i.e. clustered) across the elevational gradient for woody plants. For herbaceous plants, the mean pairwise distance showed a random distribution over the gradient. These results suggest that a better understanding of temperate forest community structure can be obtained when estimates of phylogenetic diversity include methodological and environmental sources of variation.