Nitrogen limitation on land: how can it occur in Earth system models?

Abstract

The representation of the nitrogen (N) cycle in Earth system models (ESMs) is strongly motivated by the constraint N poses on the sequestration of anthropogenic carbon (C). Models typically implement a stoichiometric relationship between C and N in which external supply and assimilation by organisms are adjusted to maintain their internal stoichiometry. N limitation of primary productivity thus occurs if the N supply from uptake and fixation cannot keep up with the construction of tissues allowed by C assimilation. This basic approach, however, presents considerable challenges in how to faithfully represent N limitation. Here, we review how N limitation is currently implemented and evaluated in ESMs and highlight challenges and opportunities in their future development. At or near steady state, N limitation is governed by the magnitude of losses from the plant-unavailable pool vs. N fixation and there are considerable differences in how models treat both processes. In nonsteady-state systems, the accumulation of N in pools with slow turnover rates reduces N available for plant uptake and can be challenging to represent when initializing ESM simulations. Transactional N limitation occurs when N is incorporated into various vegetation and soil pools and becomes available to plants only after it is mineralized, the dynamics of which depends on how ESMs represent decomposition processes in soils. Other challenges for ESMs emerge when considering seasonal to interannual climatic oscillations as they create asynchronies between C and N demand, leading to transient alternations between N surplus and deficit. Proper evaluation of N dynamics in ESMs requires conceptual understanding of the main levers that trigger N limitation, and we highlight key measurements and observations that can help constrain these levers. Two of the biggest challenges are the mechanistic representation of plant controls on N availability and turnover, including N fixation and organic matter decomposition processes.

Variability of projected terrestrial biosphere responses to elevated levels of atmospheric CO<sub>2</sub> due to uncertainty in biological nitrogen fixation

“…In their current state, NDT and NDS are very sensitive to instantaneous shifts in plant N demand. It was suggested before that, without perturbation, the degree of modelled N limitation is controlled by the magnitudes of BNF and N losses (Thomas et al, 2015). We did not generally find that NDT and NDS predicted higher BNF than other approaches in regions with high N losses.…”

“…Instantaneous BNF is calculated as a saturating function of NPP, ensuring that the fixation rate does not increase strongly when NPP is high. This BNF representation was previously used in the CLM (Thornton et al, 2007) and Jena Scheme for Biosphere-Atmosphere Coupling in Hamburg (JSBACH; Goll et al, 2012) models.…”

“…These C-N models predict that ecosystem N availability attenuates the responses of the terrestrial C cycle to eCO 2 and climate change, thereby altering the C-cycle-related biosphere-climate feedbacks (Thornton et al, 2007;Sokolov et al, 2008;Zaehle et al, 2010b;Arora et al, 2013;Smith et al, 2014;Zhang et al, 2014). Furthermore, atmospheric CO 2 and climate change modulate the terrestrial source of the greenhouse gas N 2 O, potentially providing an additional feedback to the climate system Zaehle, 2013).…”

“…This approach was originally taken with the awareness that it largely ignored the biogeochemistry of BNF, and it was thus applied as a (time-invariant) climatology to drive N cycle models (Zaehle et al, 2010b), but also applied as a dynamic-process representation (Yang et al, 2009;Wania et al, 2012;Smith et al, 2014). Cleveland et al (1999) also presented a second, considerably weaker correlation of BNF with net primary productivity (NPP), which was subsequently applied in TBMs as well (Thornton et al, 2007;Goll et al, 2012).…”

Partitioning direct and indirect effects reveals the response of water-limited ecosystems to elevated CO2

“…Although it is incontrovertible that these factors play important roles in regulating ecosystem response to eCO 2 (e.g., refs. 41-44), they remain poorly simulated by current terrestrial biosphere models (45)(46)(47)(48). Therefore, we excluded such effects to avoid introducing further levels of uncertainty related to the specific model structure.…”