The phylogenetic distribution of functional diversity across a worldwide forest dynamics plot network

Background/Question/Methods

Phylogenetic and functional diversity are routinely quantified in community ecology to understand species co-occurrence. Such analyses typically also quantify phylogenetic signal in trait data for the species in the assemblage. While this work is interesting, the integration of phylogenetic information with trait data in the context of species assemblages will be most informative on larger spatial scales with denser taxonomic sampling. In other words, it is nearly impossible to make robust inferences about the evolution of traits as it relates to community assembly using local scale community inventories. This has slowed progress towards an interesting synthesis of community ecology and evolutionary history and the phylogenetic and functional dimensions of biodiversity. Here we scale out the community phylogenetics research program to trace trait evolution on large molecular phylogenies containing the species from local assemblages around the world while placing this information back into the context of local scale assemblages. The work uses a large molecular phylogeny and detailed trait information for the species found in 14 Center for Tropical Forest Science forest plots located in temperate and tropical forests in Asia and the Americas.  We first identify nodes that are hot spots and cold spots of functional diversity within the large phylogeny. Next, we quantify whether these nodes contain species that are typically over-represented individual forest plots indicating that a locally dominate lineage is spectacularly functionally diverse or poor. Finally, we quantify whether these lineages tend to have species that co-occur on very fine scales. 

Results/Conclusions

The results show that there are very few nodes within the phylogeny that have a higher than expected degree of functional diversity and that these nodes tend to be terminally distributed. Conversely, there are many nodes with lower than expected functional diversity and these nodes are distributed throughout the phylogeny. Next, we show that lineages that are over-represented in individual forest dynamics plots often are lineages that have lower than expected functional diversity. In other words, species from the dominant lineages in forest plots tend to be functionally similar. Lastly, we show that the species from these dominant lineages that tend to be functionally similar often do not co-occur on local scales. The presentation will end with an argument for transforming community phylogenetics from a field more reliant on using phylogenies as an independent variable for analysis towards a field where densely sampled phylogenies are used as backbone information upon which ecological data can be hung.