Species richness and phylogenetic diversity as predictors of productivity across a diversity manipulation experiment

Biodiversity losses on a global scale adversely impact terrestrial ecosystem services. With declining plant diversity, plant productivity decreases, thereby diminishing the system's capacity to sequester carbon (C), and the magnitude of this effect can rival and amplify climate change. However, the mechanisms driving productivity in diverse communities remain contested. In a large biodiversity manipulation experiment, we test whether species richness, plant composition (phylogenetic dispersion), or levels of precipitation predict changes in plant cover, net primary production (NPP), and plant water use efficiency (WUE). In experiments established in spring 2018 at the University of Kansas field station (Jefferson Co., KS), we manipulated plant diversity, selecting six plant species each from three families (Poacea, Asteracae, and Fabaceae) that are commonly represented in native tallgrass prairies. From each family, we manipulated species richness by randomly selecting 1, 2, 3, or 6 species. To increase dissimilarity within the community (phylogenetic dispersion), we also created plots by randomly selecting species from different families to assemble communities of 2, 3, and 6 species. Precipitation in the plots has been manipulated through the installation of rainfall manipulation shelters to 50% and 150% of ambient precipitation. Overall, treatment combinations are replicated 12 to 36 times, totaling 240-1.5x1.5m plots. In 2019, year two, we observe an increase in percent cover with increasing species richness, suggesting we have establishment and are already seeing diversity effects in the plots. However, differences between cover and phylogenetic dispersion have yet to manifest. NPP did not differ on average between monocultures and 3 species mixed assemblages. Similarly, in the 3 species mixed plots we saw no differences between NPP in over and under-dispersed plots. We observed no differences in water use efficiencies of plants with varied richness or phylogenetic dispersion treatments. Water use efficiency of plants increased, however, with NPP. We also observed a downward trend in water use efficiencies in plots receiving 150% ambient rainfall treatments, suggesting plants in year two are responding to changes in water availability.