Effects of plant diversity on grassland soil N cycling are greater and more temporally dynamic than effects of elevated carbon dioxide

Background/Question/Methods

In a 13-year grassland experiment, we evaluated the response of soil nitrogen (N) transformations and inorganic N concentrations to elevated atmospheric carbon dioxide (CO₂; +180 ppm), N fertilization (+4 g N m^-2 y^-1), and plant species richness (1 to 16 spp). We focused on soil inorganic N pools, including nitrate and ammonium, and N fluxes, including net N mineralization and net nitrification, because they influence two important functions of terrestrial ecosystems, plant productivity and N retention. We expected treatment effects on soil N dynamics to be temporally dynamic and dependent on how each treatment influenced the amount and nutrient content of plant biomass and soil organic matter.

Results/Conclusions

The effects of elevated CO₂ on soil inorganic N pools and fluxes were small, or insignificant, and relatively static. N fertilization had inconsistent effects on soil N transformations, but increased soil nitrate and ammonium concentrations. Plant species richness had substantial and temporally variable impacts on soil N cycling. Diverse plots initially had lower net N mineralization and nitrification rates than monocultures. After 13 years, these soil N transformations were higher in diverse plots likely because of increased N concentrations in fine root inputs. The most diverse plots had lower soil nitrate concentrations than monocultures and intermediate diversity plots, especially in fertilized plots, apparently due to the greater N uptake capacity (i.e. greater root biomass) of diverse plots. Also, the effects of plant diversity and fertilization on nitrate concentrations were greatest for sampling periods when nitrate was more abundant in all plots. Notably, the first few years of data would not have adequately forecast the effects of fertilization and diversity on soil N cycling in later years. Thus, in grasslands, the dearth of long-term manipulations of plant species richness and N inputs (at ambient or elevated CO₂, or both) is a hindrance to forecasting the state of the soil N cycle and associated ecosystem functions.