A comparison of phylogenetic and species beta diversity measures describing vegetation assemblages along an elevation gradient

Prepared by Tammy L. Elliott & T. Jonathan Davies

An alder thicket plant community on Mount Irony, Labrador, Canada. The alder thicket plant community transitions into a boreal forest community at lower elevations. Photo credit: Tammy Elliott.

Plant ecologists have long been interested in describing and delineating plant communities. Until about a decade ago, data based on environmental site characteristics and species diversity were often used to meet these objectives. More recently, data derived from the evolutionary relationships among species taken from phylogenetic trees have also been used to characterize plant communities.

A common assumption in plant community ecology is that closely-related species are more ecologically similar and should therefore share similar growing requirements. This assumption is based on the idea that the evolutionary relationships among species might capture the environmental (abiotic) growth preferences of species. If this assumption is valid, then including phylogenetic information into calculations differentiating plant communities should improve the detection of differences among communities.

In this project, our goals were to 1) compare environmental site differences among vegetation plots to differences among plots based on species composition and the evolutionary relationships of species; 2) examine if plant communities delineated using the evolutionary relationships among species more closely resemble those based on abiotic differences compared to those based on differences among species composition; and 3) evaluate whether species with similar evolutionary histories co-occur in similar vegetation communities.

To address these goals, a sampling grid with 176 1-m² plots was created on Mount Irony, in subarctic Canada. In each plot, we measured the percentage cover of conifers, ferns and flowering plants. We also measured elevation, slope, sky visible and depth to impermeable layer for each of the 176 plots. These measurements were used to calculate a measure that compares the abiotic differences among plots. We then recorded differences in plant species composition and the evolutionary differences of the species among plots.

Our first analysis examined whether environmental, species and evolutionary differences among plots showed similar patterns. For the most part, the dissimilarity patterns were similar for the three different measures: plots that were dissimilar in their environmental dimensions were also dissimilar in species composition and in the evolutionary relationships of species.

We then used these dissimilarity values to group plots into clusters. Based on our initial predictions, we expected plots clustered on the evolutionary histories of the plant species within them to more closely resemble the clustering patterns based on environmental attributes, because we hypothesized that the evolutionary relationships among species might capture the environmental growth preferences of species. However, neither species composition nor evolutionary relationships among species grouped plots into clusters similar to those based on the environmental variables. In addition, we found no evidence that species with similar evolutionary histories co-occurred in similar vegetation communities. 

In summary, although we found that plot differences in environment, species composition and evolutionary relationships among species were similar, this did not transfer into how plots clustered. In addition, we did not find any evidence that species within clusters were more evolutionarily related than expected by chance. Our results suggest that adding phylogenetic data to local scale analyses might have only limited value.

A tundra plant community on Mount Irony, Labrador, Canada. Photo credit: Tammy Elliott.

This is a plain language summary for the paper of Elliott & Davies published in the Journal of Vegetation Science (https://doi.org/10.1111/jvs.12700).