Phylogenetic diversity–ecosystem function relationships are insensitive to phylogenetic edge lengths

SummaryFor experiments that link manipulated biodiversity to ecosystem function ( EF ), phylogenetic diversity ( PD ) has been an especially powerful form of biodiversity that explains variation in EF . PD represents the total amount of evolutionary history or genetic changes represented by a suite of species and potentially the total accumulation of species niche and functional differences. Analyses often use PD in linear models and assume that ecological differences are proportional to phylogenetic distances. Yet, it is unclear whether alternative models of evolutionary change would improve both the statistical fit and conceptual understanding of how PD influences EF . Here, I use a PD – EF relationship and systematically alter models of evolution including changes in the rate of evolution from phylogenetic root–to‐tip and constancy of evolutionary rate across clades. I also compare the PD – EF relationship to several randomization procedures that sequentially remove aspects of the observed phylogeny. I show that changing edge lengths with evolutionary models does not strongly affect PD – EF relationships. Moreover, the observed relationship was not substantially different than the explanation provided by a phylogenetic tree with edge lengths randomized, but was significantly better than randomizations that affected the topology of the phylogenetic trees. These results reveal that changes to the edge lengths have little effect on PD – EF relationships, and it is the topology that really matters. Further, placing species in their correct phylogenetic positions is much more important than developing better estimates of edge lengths.