There is significant controversy over whether increases in photosynthesis rates under elevated CO2 (eCO2) can support sustained increases in ecosystem productivity. A key aspect of this question is the role of nitrogen (N) limitation. Free Air Carbon Enrichment (FACE) experiments have yielded contrasting results in different ecosystems, which may be related to N acquisition capabilities associated with mycorrhizal types. Arbuscular mycorrhizal (AM) ecosystems, including the Oak Ridge National Laboratory (ORNL) FACE experiment, generally did not have significant increases in biomass accumulation under eCO2. In contrast, ectomycorrhizal (ECM) ecosystems, including the Duke FACE experiment, did sustain additional growth under eCO2. These processes are generally not included in current earth system models (ESMs). We implemented a new nitrogen cycle model within the Geophysical Fluid Dynamics Laboratory (GFDL) global land model (LM3) that simulates soil microbial activity as well as several different symbiotic nitrogen acquisition strategies. These include an inorganic N “scavenging” strategy analogous to AM associations and an organic N “mining” strategy analogous to ECM associations. We simulated the Duke and ORNL FACE experiments in order to test whether these N acquisition strategies could explain the contrasting forest responses to eCO2 at the two sites. We also ran global simulations using the model.
Results/Conclusions
Model simulations reproduced observed C and N pools in both soils and vegetation at the two FACE sites after spinup and historical simulations. The model was able to reproduce the observed contrast in vegetation responses to elevated CO2 at the two sites. At ORNL, simulated plant growth was strongly N limited under elevated CO2, resulting in little additional biomass accumulation. In contrast, vegetation at the Duke site was able to acquire additional N by stimulating SOM decomposition via the mining strategy associated with its ECM association, and was able to sustain biomass accumulation over the length of the FACE experiment, matching observations. These results suggest that contrasting N acquisition strategies can explain the observed differences in vegetation responses between the sites. Soil C changes under eCO2 were also consistent with observed contrasts between sites. In global simulations, the model successfully simulated spatial patterns of mycorrhizal associations and N limitation. These results highlight the importance of including nitrogen cycle feedbacks between vegetation and soils in order to accurately model ecosystem C and N responses to rising atmospheric CO2 levels.