Background/Question/Methods The land nitrogen cycle is fundamental for understanding plant production, ecosystem carbon exchange, and how terrestrial ecosystems respond to climate change. While we have considerable knowledge of how nitrogen cycles and interacts with carbon, phosphorus and other resources across land biomes globally, it is critical to work towards abstracting this empirical knowledge into dynamical models that can be evaluated in the context of the larger coupled earth climate-biogeochemical system. Such abstractions have proven difficult, however, and have often relied on prescribed rather than dynamical representations of key processes and feedbacks. We here present a prognostic terrestrial nitrogen cycle designed to capture essential processes and feedbacks of the land biosphere, and it's coupling to both carbon and climate. Our model builds upon the Princeton NOAA Geophysical Fluid Dynamics Laboratory LM3V land model. It offers a unique opportunity to evaluate key processes and feedbacks that can act to govern the response of land ecosystems to regional and global changes in climate, atmospheric carbon dioxide, and land use. We here specifically address the question of how the land biosphere responds to increasing CO2 and climate change, with the goal of identifying essential feedbacks that need further attention. Our methods and approach include critically testing the model predictions against empirical observations of patterns of N cycling globally, as well as the ability of the model to resolve observed trends in the global atmospheric and ocean carbon cycles.

Results/Conclusions Our model captures many essential characteristics of carbon-nitrogen interactions, and is capable of broadly recreating spatial and temporal variations in global nitrogen and carbon dynamics. The introduced nitrogen dynamics allows us to capture empirically observed responses in net primary production to step changes in carbon dioxide, while a carbon-only model fails to do this. Consistent with theories of successional dynamics, we find that physical disturbance induces strong carbon-nitrogen feedbacks, caused by intermittent nitrogen loss and subsequent nitrogen limitation. In contrast, carbon-nitrogen interactions are weak when the coupled model system approaches equilibrium. Thus, at steady-state many simulated features of the global carbon cycle, such as primary productivity, carbon inventories, and biogeography are similar to simulations that do not include nitrogen feedbacks. A surprising and potentially important finding is the possibility of a change in the terrestrial carbon cycle over the last twenty years. We discuss the evidence and implications of such a change.