Using community phylogenetics to assess changes in plant assemblages since the last glacial maximum (21ka)

Background/Question/Methods

Paleoecological datasets are useful for studying changes in species distribution and community structure over extended periods of climate change. Increasingly, genetic analyses are being integrated with community ecology and are providing new methods to explore processes governing community assembly. From a phylogenetic point of view, communities can be clustered, neutral, or over-dispersed, with biotic interactions being recognized as one of the main factors driving these differences in structure across space. Analyzing temporal changes in patterns of phylogenetic structure could provide deeper insights into the role of biotic interactions in shaping species distributions and assemblages, and potentially reveal how biotic interactions change through time in response to climate change. Here, we present the spatial and temporal patterns of phylogenetic community structure in plants (as recorded in the fossil pollen record) in eastern North America over the last 21 ka. Phylogenetic structure was calculated using two metrics, namely Net Relatedness Index (NRI) and Nearest Taxon Index (NTI). 

Results/Conclusions

Across space, the two metrics of phylogenetic structure showed similar latitudinal gradients but contrasting interpretations, with NTI indicating clustering where NRI indicates over-dispersion and vice versa. We attribute these contrasting patterns to differences in the metrics; NRI encapsulates deeper phylogenetic relationships while NTI focuses on the tips of the phylogeny. These contrasting patterns disappeared after taking the deep split between angiosperms and gymnosperms into account. Moving from north to south, both NTI and NRI values show that plant assemblages change from being clustered to neutral to over-dispersed towards the southernmost latitudes. From a temporal perspective, these values suggest that communities in recently deglaciated areas tended to be clustered and became over-dispersed as time advanced. Additionally, these patterns are consistent through time. Although it is traditionally inferred that competition results in an over-dispersed community and habitat filtering results in a clustered community, our results do not distinguish which factors are shaping assemblage structure. In order to infer the relative role of competition and habitat filtering we are complementing phylogenetic approaches with plant trait/functional data.