Interactions between plant community composition and precipitation alter diversity effects on plant productivity

Laura Y. Podzikowski and James D. Bever, Ecology and Evolutionary Biology, University of Kansas, Lawrence, KS, Laura Y. Podzikowski, Kansas Biological Survey, University of Kansas, Lawrence, KS, Bryan L. Foster, Department of Ecology and Evolutionary Biology and Kansas Biological Survey, University of Kansas, Lawrence, KS, Kristen Mecke, Department of Ecology and Evolutionary Biology, University of Tennessee, Knoxville, TN

Background/Question/Methods
There is overwhelming evidence of a positive and saturating relationship between species richness and plant productivity. However, what environmental and biological conditions—such as altered precipitation and changes in species composition—influence when the biodiversity-productivity relationship generates the greatest over-yielding remains poorly understood. In a grassland biodiversity experiment established in spring 2018 at the University of Kansas field station (Jefferson Co., KS) we test the effects of plant diversity, plant community composition, and precipitation on plant productivity. We manipulated plant diversity, selecting six plant species each from three families (Poacea, Asteracae, and Fabaceae) commonly represented in native tallgrass prairies. We manipulated species richness by randomly selecting 1, 2, 3, or 6 species. To increase functional and phylogenetic dissimilarity within communities we assembled mixtures (2, 3, and 6 species) by randomly selecting species from one family or multiple families. Precipitation is altered to 50% and 150% of ambient conditions by rainfall manipulation shelters. To quantify productivity, we assessed plant cover monthly throughout 2019-2020. We partition the net biodiversity effects (NE) into the complementary effect (CE, when species productivity is higher than expected based on monoculture yields) and the selection effect (SE, when productivity is driven by a dominant species).
Results/Conclusions
In 2019, we observed positive saturating effects of richness on CE, SE, and NE; however, differences between treatments had yet to manifest. In 2020, we observed main effects of precipitation on CE and NE and interactions between treatments on SE. Under increased precipitation CE and NE increased. Under increased precipitation, the relationship between SE and richness was weakest in plots with plants from multiple families and strongest with plants from one family. In low precipitation, SE increased with diversity independent of community composition. However, a high amount of variation in SE remained unexplained. Including plant family in our model improved fits for SE both years, while it did not improve models of CE. The relationship between SE and richness was steepest in Fabaceae mixtures both years. In 2020, we observed a negative relationship between SE and richness in Asteraceae. However, on average SE was lower with increased precipitation. Divergent trends between SE and CE in 2020 rendered the relationship between plant productivity and precipitation undetectable. These findings suggest the interactions between community composition and water stress are complex and may alter when a single species dominates mixtures—a pattern which has implications for system stability and resilience.