Biological nitrogen fixation can fuel CO 2 sequestration by forests but can also stimulate soil emissions of nitrous oxide (N 2 O), a potent greenhouse gas. Here we use a theoretical model to suggest that symbiotic nitrogen-fixing trees could either mitigate (CO 2 sequestration outweighs soil N 2 O emissions) or exacerbate (vice versa) climate change relative to non-fixing trees, depending on their nitrogen fixation strategy (the degree to which they regulate nitrogen fixation to balance nitrogen supply and demand) and on nitrogen deposition. The model posits that nitrogen-fixing trees could exacerbate climate change globally relative to non-fixing trees by the radiative equivalent of 0.77 Pg C yr −1 under nitrogen deposition rates projected for 2030. This value is highly uncertain, but its magnitude suggests that this subject requires further study and that improving the representation of biological nitrogen fixation in climate models could substantially decrease estimates of the extent to which forests will mitigate climate change.

Symbiotic nitrogen fixation (SNF) is a key ecological process whose impact depends on the strategy of SNF regulation—the degree to which rates of SNF change in response to limitation by N versus other resources. SNF that is obligate or exhibits incomplete downregulation can result in excess N fixation, whereas a facultative SNF strategy does not. We hypothesized that tree‐based SNF strategies differed by latitude (tropical vs. temperate) and symbiotic type (actinorhizal vs. rhizobial). Specifically, we expected tropical rhizobial symbioses to display strongly facultative SNF as an explanation of their success in low‐latitude forests. In this study we used 15N isotope dilution field experiments in New York, Oregon, and Hawaii to determine SNF strategies in six N‐fixing tree symbioses. Nitrogen fertilization with +10 and +15 g N m−2 year−1 for 4–5 years alleviated N limitation in all taxa, paving the way to determine SNF strategies. Contrary to our hypothesis, all six of the symbioses we studied sustained SNF even at high N. Robinia pseudoacacia (temperate rhizobial) fixed 91% of its N (%Ndfa) in controls, compared to 64% and 59% in the +10 and +15 g N m−2 year−1 treatments. For Alnus rubra (temperate actinorhizal), %Ndfa was 95%, 70%, and 60%. For the tropical species, %Ndfa was 86%, 80%, and 82% for Gliricidia sepium (rhizobial); 79%, 69%, and 67% for Casuarina equisetifolia (actinorhizal); 91%, 42%, and 67% for Acacia koa (rhizobial); and 60%, 51%, and 19% for Morella faya (actinorhizal). Fertilization with phosphorus did not stimulate tree growth or SNF. These results suggest that the latitudinal abundance distribution of N‐fixing trees is not caused by a shift in SNF strategy. They also help explain the excess N in many forests where N fixers are common.

Nitrogen (N)‐fixing trees fulfil a unique and important biogeochemical role in forests through their ability to convert atmospheric N2 gas to plant‐available N. Due to their high N fixation rates, it is often assumed that N‐fixing trees facilitate neighbouring trees and enhance forest growth. This assumption is supported by some local studies but contradicted by others, leaving the overall effect of N‐fixing trees on forest growth unresolved. Here we use the US Forest Service's Forest Inventory and Analysis database to evaluate the effects of N‐fixing trees on plot‐scale basal area change and individual‐scale neighbouring tree demography across the coterminous US. First we discuss the average trends. At the plot and individual scales, N‐fixing trees do not affect the relative growth rates of neighbouring trees, but they facilitate recruitment and inhibit survival rates, suggesting that they are drivers of tree turnover in the coterminous US. At the plot scale, N‐fixing trees facilitate the basal area change of non‐fixing neighbours. In addition to the average trends, there is wide variation in the effect of N‐fixing trees on forest growth, ranging from strong facilitation to strong inhibition. This variation does not show a clear geographical pattern, though it does vary with certain local factors. N‐fixing trees facilitate forest growth when they are likely to be less competitive: under high N deposition and high soil moisture or when neighbouring trees occupy different niches (e.g. high foliar C:N trees and non‐fixing trees). Synthesis. N‐fixing trees have highly variable effects on forest growth and neighbour demographics across the coterminous US. This suggests that the effect of N‐fixing trees on forest development and carbon storage depends on local factors, which may help reconcile the conflicting results found in previous localized studies on the effect of N‐fixing trees on forest growth.

Despite its pivotal feedback to carbon cycling, representing the dynamic response of vegetation to nitrogen limitation is a key challenge for simulating the terrestrial carbon sink with land models. Here, we explore a representation of this dynamic response of vegetation to nitrogen limitation with a novel representation of biological nitrogen fixation and nitrogen cycling in the Canadian Land Surface Scheme Including Biogeochemical Cycles. First, we assess how incorporating the dynamic response of vegetation to nitrogen limitation via biological nitrogen fixation influences the response to CO2 and nitrogen fertilization experiments, comparing simulations against observation‐based estimates from meta‐analyses. This evaluates whether the underlying mechanisms are realistically represented. Second, we assess how incorporating the dynamic response of vegetation to nitrogen limitation via biological nitrogen fixation affects simulated terrestrial carbon sequestration over the 20th and early 21st century, examining the effects of global change drivers (CO2, nitrogen deposition, climate, and land‐use change) acting both individually and concurrently. Including nitrogen cycling reduces the terrestrial carbon sink driven by elevated CO2 over the historical period. Representing the dynamic response of vegetation to nitrogen limitation via biological nitrogen fixation increases the estimate of the present‐day terrestrial carbon sink by 0.2 Pg C yr−1 (because elevated CO2 intensifies nitrogen limitation, which drives the upregulation of biological nitrogen fixation, alleviating nitrogen limitation). Our results highlight the importance of the dynamic response of vegetation to nitrogen limitation for realistically projecting the future terrestrial carbon sink under global change with land models.

Nitrogen-fixing trees are an important nitrogen source to terrestrial ecosystems. While they can fuel primary production and drive carbon dioxide sequestration, they can also potentially stimulate soil emissions of nitrous oxide, a potent greenhouse gas. However, studies on the influence of nitrogen-fixing trees on soil nitrous oxide emissions have not been quantitatively synthesised.Here, we show in a meta-analysis that nitrogen-fixing trees more than double soil nitrous oxide emissions relative to non-fixing trees and soils. If planted in reforestation projects at the global scale, nitrogen-fixing trees could increase global soil nitrous oxide emissions from natural terrestrial ecosystems by up to 4.1%, offsetting climate change mitigation via reforestation by up to 4.4%.

Abstract. Representing biological nitrogen fixation (BNF) is an important challenge for coupled carbon (C) and nitrogen (N) land models. Initial representations of BNF in land models applied simplified phenomenological relationships. More recent representations of BNF are mechanistic and include the dynamic response of symbiotic BNF to N limitation of plant growth. However, they generally do not include the competitive dynamics between N-fixing and non-fixing plants, which is a key ecological mechanism that determines ecosystem-scale symbiotic BNF. Furthermore, asymbiotic BNF is generally not included in land models. Here, we present LM4.1-BNF, a novel representation of BNF (asymbiotic and symbiotic) and an updated representation of N cycling in the Geophysical Fluid Dynamics Laboratory Land Model 4.1 (LM4.1). LM4.1-BNF incorporates a mechanistic representation of asymbiotic BNF by soil microbes, a representation of the competitive dynamics between N-fixing and non-fixing plants, and distinct asymbiotic and symbiotic BNF temperature responses derived from corresponding observations. LM4.1-BNF makes reasonable estimations of major carbon (C) and N pools and fluxes and their temporal dynamics, in comparison to the previous version of LM4.1 with N cycling (LM3-SNAP) and to previous representations of BNF in land models generally (phenomenological representations and those without competitive dynamics between N-fixing and non-fixing plants and/or asymbiotic BNF) at a temperate forest site. LM4.1-BNF effectively reproduces asymbiotic BNF rate (13 kgNha-1yr-1) in comparison to observations (11 kgNha-1yr-1). LM4.1-BNF effectively reproduces the temporal dynamics of symbiotic BNF rate: LM4.1-BNF simulates a symbiotic BNF pulse in early succession that reaches 73 kgNha-1yr-1 at 15 years and then declines to ∼0 kgNha-1yr-1 at 300 years, similarly to observed symbiotic BNF, which reaches 75 kgNha-1yr-1 at 17 years and then declines to ∼0 kgNha-1yr-1 in late successional forests. As such, LM4.1-BNF can be applied to project the dynamic response of vegetation to N limitation of plant growth and the degree to which this will constrain the terrestrial C sink under elevated atmospheric CO2 concentration and other global change factors.

No abstract

The strength of CO2 fertilisation is a major uncertainty across terrestrial biosphere models (TBMs) and is suggested to be overestimated without a representation of nitrogen (N) limitation. Here, we compare TBM projections with and without coupled C and N cycling over alternative future scenarios (the Shared Socioeconomic Pathways) to examine how representing N cycling influences CO2 fertilisation as well as the effects of a comprehensive group of physical and socioeconomic global change drivers. Because elevated N deposition and N mineralisation (driven by elevated temperature) have stimulated terrestrial C sequestration over the historical period, a TBM without N cycling must exaggerate the strength of CO2 fertilisation to compensate for these unrepresented N processes and to reproduce the historical terrestrial C sink. As a result, it cannot reliably project the future terrestrial C sink, overestimating CO2 fertilisation as CO2 increases faster than N deposition and temperature in future scenarios.