Can the functional consequences of species extinction be predicted by evolutionary history?

The past decade has been one of contentious debate about how species extinctions affect the productivity of ecosystems.  This contention has been fueled by the results of many experiments that have manipulated species diversity directly and found that productivity increases as a function of species richness, but often the productivity of a diverse community is no more productive than the single most productive species.  These results contradict a common idea that species need to partition resources in space or time to coexist, and which should cause a diverse community to be more productive than its ‘best’ species.  Here we argue that this conflict may be partly resolved by considering the evolutionary history of the organisms used in experiments.  We assembled the data from 11 large-scale manipulations of grassland plant diversity.  For the collection of 176 angiosperm species used in these studies, we utilized super-matrix phylogenetic analyses based on all available data from GenBank, followed by mixed model Bayesian-MCMC estimation of the posterior probability distribution of phylogenetic tree space. We used this tree to ask whether evolutionary divergence among species could predict the impacts of plant species richness on plant production in >1600 pots and field plots that were seeded with two to 32-species polycultures. Our analyses show that the evolutionary history of angiosperms does indeed help predict the impacts of plant diversity on biomass production, revealing the divergence necessary to overcome niche conservatism.  Thus, phylogenetic information offers a deeper understanding of the functional consequences of extinction.