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

Abstract

AbstractQuestionEcologists have long been interested in the delineation and description of plant communities but have only recently included phylogenetic data into these analyses. Here, we assess whether species‐based dissimilarities (beta diversity, BD) and more recent phylogenetic beta diversity (PhBD) measures are correlated with dissimilarities among sites based on abiotic variables. Additionally, we examine whether BD and PhBD measures aggregate sites into clusters that reflect their environmental attributes. Assuming phylogenetic conservatism in abiotic niche preferences, we predict PhBD dissimilarity matrices will correlate to those based on abiotic site variables, and that clusters determined by PhBD will more closely match to assemblages clustered by abiotic environment than will clusters determined by species BD.LocationMount Irony, Labrador in the Eastern Canadian subarctic.MethodsWe combine vascular plant co‐occurrence data collected from an elevation gradient on Mount Irony with information on phylogenetic relatedness to compare site dissimilarities based on abiotic variables with those estimated on measures of BD and PhBD. We also examine whether clusters based on BD and PhBD resemble clusters based on abiotic variables, and investigate whether similarly clustered sites are composed of species with similar evolutionary histories.ResultsWe found significant correlations among the dissimilarity matrices and ordinations for the abiotic variables, BD and PhBD; however, neither BD nor PhBD aggregated assemblages into clusters that reflected their environmental differences. Further, we found no evidence that species within clusters were any more closely related than expected by chance. We observed similar patterns when communities were defined by all vascular plants and only angiosperms.ConclusionsThe correlations among site dissimilarities based on abiotic variables, BD and PhBD suggest environmental filtering; however, sites clustered by BD and PhBD did not resemble those clustered by abiotic variables, indicating that the added value of phylogenetic data for local scale analyses might be limited.