Fraction of nitrous oxide production in nitrification and its effect on total soil emission: A meta-analysis and global-scale sensitivity analysis using a process-based model

Motoko Inatomi; Tomohiro Hajima; Akihiko Ito

doi:10.1371/journal.pone.0219159

Abstract

Nitrification in terrestrial soils is one of the major processes of emission of nitrous oxide (N₂O), a potent greenhouse gas and stratospheric-ozone-depleting substance. We assessed the fraction of N₂O emission associated with nitrification in soil through a meta-analysis and sensitivity analysis using a process-based model. We corrected observational values of gross nitrification and associated N₂O emission rates from 71 records for various soils in the world spanning from 0.006% to 29.5%. We obtained a median value of 0.14%, and then assessed how the nitrification-associated N₂O emission fraction has been considered in terrestrial nitrogen cycle models. Using a process-based biogeochemical model, we conducted a series of sensitivity analyses for the effects of different values of nitrification-associated N₂O emission fraction on soil N₂O emission. Using an empirical relationship between soil pH and nitrification-associated N₂O emission fraction, the model well simulated global emission patterns (global total in the 2000s, 16.8 Tg N₂O yr^–1). Differences in the nitrification-associated N₂O emission fraction caused differences in total N₂O emission of as much as 2.5 Tg N₂O yr^–1. Therefore, to obtain reliable estimation of soil N₂O emission for nitrogen and climate management, it is important to constrain the parameterization in models by ensuring extensive and accurate observations.

Citation: Inatomi M, Hajima T, Ito A (2019) Fraction of nitrous oxide production in nitrification and its effect on total soil emission: A meta-analysis and global-scale sensitivity analysis using a process-based model. PLoS ONE 14(7): e0219159. https://doi.org/10.1371/journal.pone.0219159

Editor: Jianlin Shen, Chinese Academy of Sciences, CHINA

Received: February 27, 2019; Accepted: June 11, 2019; Published: July 10, 2019

Copyright: © 2019 Inatomi et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the manuscript and its Supporting Information files.

Funding: AI was supported by KAKENHI grant (no. 26281014 and 17H01867) from the Japan Society for Promotion of Science. URL: https://www.jsps.go.jp/. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Nitrous oxide (N₂O) is the third important long-lived greenhouse gas next to carbon dioxide (CO₂) and methane (CH₄) [1] and is the most important substance depleting stratospheric ozone [2]. To reach the overarching mitigation targets of the Paris Agreement [3] we need to suppress the growth of atmospheric N₂O concentration, to which anthropogenic emissions contribute at a level comparable to that from natural sources [4, 5]. Also, assessment and regulation of N₂O emission contribute to management of nitrogen cycle, which is closely related to many issues of human sustainability, such as food production and sanitation [6]. Nevertheless, there remain serious uncertainties in our understanding and predictability of N₂O dynamics.

Terrestrial soils—both natural and agricultural—are a prevailing source of N₂O in the atmosphere [7, 8], but spatial heterogeneity and temporal variability of the N₂O flux make it difficult to quantify broad-scale budgets. Most of the N₂O released from the soil surface is produced by two separate microbial processes, nitrification and denitrification, which differ in terms of active microbes, substrates, and environmental responsiveness [9, 10]. There are still serious knowledge gaps and difficulties in using models to predict soil N₂O emissions in a quantitative manner.

Nitrification by ammonia oxidizers is the primary process of N₂O production in oxic (aerobic) soils and is thought to be more ubiquitous than denitrification, which occurs in anaerobic wet soils. Recent studies have revealed the contributions of different soil microbes, such as ammonia-oxidizing archaea and bacteria, to nitrification [11, 12]. In nitrification, most of the oxidized ammonia is turned into nitrate (NO₃^–) via nitrite (NO₂^–), and a certain (usually small) fraction of nitrogen is released as N₂O. The fraction of nitrification-associated N₂O emission (fN₂O_nit) and its regulation mechanism are important but barely understood. Although fN₂O_nit is critically important to predict soil N₂O emission, a few studies have investigated the responses of fN₂O_nit and the corresponding nitric oxide (NO) emission fraction to soil temperature and moisture conditions [13, 14]. However, observational data and knowledge are still insufficient to evaluate broad-scale emissions, including from a variety of soils. Farquharson (2016) [15] conducted a systematic analysis of fN₂O_nit from agricultural soils in Australia. He found that 0.03% to 1% of nitrogen is released as N₂O associated with nitrification in soils and found no strong relationship with environmental factors such as soil moisture. For a broad range of natural soils, and in other regions, we have found no systematic analysis on fN₂O_nit.

fN₂O_nit should be an important parameter in biogeochemical models that aim to simulate nitrogen cycles and predict N₂O emissions from land. Many terrestrial nitrogen cycle and N₂O emission models have been developed, from simple box-flow models (e.g., the Terrestrial Ecosystem Model [16]) to more mechanistic ones (e.g., the Denitrification Decomposition [DNDC] model [17]). In these models, fN₂O_nit or a similar parameter (i.e., the total N₂O and NO emission fraction) has been determined in a simplified empirical manner. In the well-recognized “hole in the pipe” or “leaky pipe” concept of soil nitrogenous gas emission [18, 19], fN₂O_nit represents the size of the N₂O hole in the nitrification pipe (Fig 1). Note that the nitrification rate (as defined by pipe diameter and flow velocity) varies also with the environmental conditions, so it is possible to regard fN₂O_nit as a constant or as an independent variable that changes with environmental condition. In the latter case, empirical parameterizations have been adopted to determine the N₂O emission fraction, using a limited amount of observational data. As a consequence, these emissions could have a considerable range of bias and error due to the uncertainty of fN₂O_nit values.

Download:

Fig 1. Conceptual diagram of nitrification and the “holes-in-a-pipe” concept.

Ts, soil temperature; WFPS, water-filled pore space.

https://doi.org/10.1371/journal.pone.0219159.g001

Here, we focused on fN₂O_nit from the perspective of the global N₂O budget, aiming at better predictability of emission by biogeochemical models. Our focal research questions are as follows. (1) What is the feasible range of fN₂O_nit in the terrestrial ecosystems? (2) How does fN₂O_nit vary in the field in response to environmental conditions? (3) Can we attain a better parameterization of fN₂O_nit, applicable to global scale, on the basis of present data? First, to clarify the range of variability and the broad-scale trends in these values, we conducted a meta-analysis of the observed values of fN₂O_nit for both natural and agricultural soils. Second, we surveyed how fN₂O_nit was included in current terrestrial N₂O estimation models. Third, we used the results of the meta-analysis and model survey to conduct a series of sensitivity analyses of fN₂O_nit and N₂O emissions by using our biogeochemical model. Finally, we discuss how we can reduce the range of estimation uncertainty in both experimental and modeling studies. Note that this study focuses on nitrification as the first step and that other important processes such as denitrification, nitrifier denitrification, and abiotic production [20] are not explicitly addressed here. We intend to address these other processes in a forthcoming study using a similar approach.

Methods

Meta-analysis

Overview.

A meta-analysis was conducted to reveal the range, frequency, and tendencies of soil fN₂O_nit reported in the literature. The results were reported following the Preferred Reporting Items for Systematic Review and Meta-Analyses protocols (PRISMA) [21] protocol (S1 Table). To obtain information on the general properties of the parameter, we gathered observational values reported from a wide range of studies. By using Web of Science (Thomson Reuters, New York, NY, USA) and Google Scholar (Alphabet, Mountain View, CA, USA), we searched papers and reports that included data on nitrogenous gas exchange and soil biogeochemistry. We used combinations of three terms (each from #1 to #3) in S2 Table to search candidate papers; for example, “nitrous oxide flux” and “nitrification rate” and “soil surface”. No date and time limitations were applied to harvest from the maximum extent of the literature. Also, we examined the reference lists in each paper to find additional literature that did not appear in the web searches.

Study selection.

We carefully selected source data of fN₂O_nit for the meta-analysis, particularly taking into account the consistency between N₂O emission and nitrification rates. First, we removed papers that addressed non-soil N₂O emissions (i.e., from ponds, landfills, animal slurry, etc.). For quantitative consistency, we selected papers including data on gross nitrification (i.e., NH₄⁺ consumption) and associated N₂O emission. Therefore, several papers that reported only net nitrification (i.e., NO₃^– production) rates, potential emission rates, and data under oxygen-free condition, were carefully removed from the meta-analysis. Also, we focused on daily or longer phenomena, so that rates of N₂O emission from the soil surface to the atmosphere could be adequately approximated to N₂O production rates within the soil (i.e., the vertical diffusion time lag was negligible). As a result, papers reporting only instantaneous (i.e., for seconds to minutes) measurements were excluded; in general, these instantaneous measurement data show extremely wide ranges of variability, making a robust analysis difficult. Finally, we selected the source papers by measurement method used in each study, because several methods could give biased values under certain conditions (e.g., DMPP inhibition slows greatly at >25°C [22]). We confirmed that the method-based selection had a small impact on the analysis results. The paper selection was made by two authors and discrepancies were resolved by discussion.

Data extraction.

Data on fN₂O_nit values and associated properties were extracted: soil pH, solvent of pH, soil temperature, soil moisture content with units, soil texture, clay / silt / sand composition, latitude, longitude, land-cover type, and soil-type classification. Few papers provide values of fN₂O_nit directly, and therefore, if applicable, we calculated the values from gross nitrification and associated N₂O emission rates measured under the same condition.

Data analyses.

A statistical software R [23] was used to calculate the statistical metrics for the records: i.e., mean, median, maximum, minimum, standard deviation [σ], kurtosis, skewness, and quartiles. A few extreme values can, in most cases harmfully, affect the results of statistical metrics. To assess the influences of outliers, these statistical metrics were also calculated after removing top and bottom outliers (10% from all the records). Note that we used both datasets with and without outlier values in the following analyses and model simulations. Furthermore, to reduce the size effect of different sample numbers (i.e., weights) among papers, these metrics were also calculated using the mean values for each paper.

Parameterization of fN₂O_nit in other models

We then assessed how fN₂O_nit has been parameterized in other terrestrial nitrogen models and evaluated the influence on N₂O emission estimation. According to a review by Frolking et al. (1998) [24], former models have adopted different constant fN₂O_nit values, namely 0.5% in the ExpertN model and 2% in the CENTURY model. These values were examined in the sensitivity simulations mentioned below. In later models, fN₂O_nit was parameterized as a function of environmental conditions in different manners. The modified DNDC model [25] (their Table 3, Eq 8) adopted the following parameterization: (1) (2) where Ft is a scholar function of soil temperature (Ts,°C) and WFPS is water-filled pore space (fraction). This parameterization gives a peak value, about 0.06%, at about 35°C under saturated soil water conditions (Fig 2A). The Dynamic Land Ecosystem Model [26] (DLEM) parameterizes fN₂O_nit as a function of WFPS (%): (3)

Download:

Fig 2. Relationship between the fraction of nitrification-associated N₂O emission (fN₂O_nit) assumed in the models.

(a) DNDC and (b) DLEM. WFPS, water-filled pore space.

https://doi.org/10.1371/journal.pone.0219159.g002

In this parameterization, the fN₂O_nit value increases with increasing soil water content (Fig 2B) and does not exceed 0.1%. The Community Land Model [27] and O-CN model [28] adopted a version of the DNDC parameterization. In the recent paper on N₂O model intercomparison [29], an elaborate table summarizes how nitrification and N₂O emission are formulated in contemporary land nitrogen models.

Description of the N₂O simulation model

To assess the range of estimations associated with variations in fN₂O_nit, we conducted a series of simulations using a process-based model, namely, Vegetation Integrative SImulator for Trace gases (VISIT [30, 31]). This model was selected, because it has an intermediate complexity among the global N₂O models and gave moderate results in the model intercomparison project [29]. Also, the model was used for a regional evaluation of soil N₂O emission in East Asia, one of the highly human-influenced regions, demonstrating the credibility for broad-scale applications [32]. Briefly, the model consists of water, carbon, and nitrogen cycling schemes for terrestrial ecosystems and is aimed at assessing atmosphere–ecosystem exchange of greenhouse gases and trace gases. The nitrogen cycle is fully included, from inputs (atmospheric deposition, biological fixation, and fertilizer input) to outputs (leaching, ammonia volatilization, and nitrogenous gas emissions through nitrification and denitrification). Intra-ecosystem dynamics of nitrogen among plant, soil, and microbe is simulated in an explicit manner. The model was validated by comparing various biogeochemical aspects with observational data [33–35]. In the VISIT model, fN₂O_nit is assumed to be a universal, constant value (1% of gross nitrification).

A brief description of the method used to estimate fN₂O_nit is given below. In the VISIT model, soil N₂O production through the nitrogen cycle is conceptualized by using the “hole in the pipe” scheme (see Fig 1). Gross nitrification (R_nit) and associated N₂O emission (N₂O_nit) are related as follows: (4)

Nitrification rate and its environmental dependencies were derived from the NGAS scheme developed by Parton et al. (1996) [36], as follows: (5) where p_WFPS-nit, p_pH, and p_Ts denote the environmental scalar functions derived from observations (see ref. [36]) of water-filled pore space (WFPS), soil pH, and soil temperature (Ts), respectively. K denotes the coefficient of N turnover, taking values from 3.5 of natural soils to 12.0 of agricultural soils. F_max is the maximum nitrification-associated gas flux coefficient and NH₄⁺ is the soil ammonium content. WFPS and NH₄⁺ were simulated by VISIT and so varies temporally and spatially.

Global simulations by using the VISIT model were conducted with the common protocol and initial and boundary conditions. Namely, they were conducted at a spatial resolution of 0.5° x 0.5° for latitude and longitude, during the period from January 1901 to December 2016. Historical climate data from CRU TS3.25 [37] (temperature, precipitation, vapor pressure, and cloudiness) and land-use data [38] were used to drive the model. Historical changes in atmospheric nitrogen deposition were derived from Galloway et al. (2004) [39], and in croplands, input of nitrogen fertilizer was determined on a country-basis by using FAOSTAT (http://www.fao.org/faostat). The amount of national fertilizer use was divided by total cropland area and allocated to each grid cell. For each grid, a spin-up calculation was conducted for 300 to 2000 years, depending on case, under stationary conditions until a stable-state carbon budget was reached, before starting the historical experiment.

Sensitivity simulations

All sensitivity simulations were conducted by VISIT using the common forcing dataset and protocols; only the fN₂O_nit value was changed. First, to simply assess the sensitivity of N₂O emission estimation to fN₂O_nit, we halved (i.e., 0.5%) and doubled (2%) the parameter value (originally 1% in VISIT) and compared the results. We then changed the fN₂O_nit values to those obtained by the meta-analysis mentioned above, that is, mean and median values for all data and several subsets. In these simulations, constant fN₂O_nit values were applied to all grids. Second, the fN₂O_nit value was replaced by those of DNDC and DLEM described above, using the same temperature and moisture conditions. Third, finally, we derived an empirical relationship between soil pH and fN₂O_nit value from the meta-analysis data. Such a relationship was shown in laboratory studies [40] and previous meta-analysis [15], but has not been examined by models at the global scale. Here, we used the global soil pH map (S1 Fig) produced by the Global Soil Data Task of the International Geosphere-Biosphere Programme [41].

Results and discussion

We obtained 71 records from 13 studies in the published literature (Table 1; Fig 3), covering a wide range of different ecosystems and soil texture types; see S3 Table for the data extracted. Although no date and time limitation were applied, the data were obtained from 1985 to 2013. Although a large number of papers addressed the nitrification and N₂O emission (2184 papers), we found that only a small number of papers (35 papers) contain the data on gross nitrification rate for a sufficiently long period. Other 2149 papers were rejected, although they contained partial data on nitrification-associated N₂O emission. Furthermore, many of the measurements (22 out of 35) were made using problematic methods or conditions and so removed. As a result of data selection, measurements in the literature used in this study were made mainly by using two methods of soil incubation: the inhibitor (C₂H₂, N-serve, and NaClO₃) treatment and the stable carbon isotope (¹⁵N labelling) method [10, 42]. Most of the measurements were conducted in the laboratory: only one study was done in the field. The record number is not so abundant, but the dataset covers a wide variety of ecosystems and soils such as forest, grassland, and cropland. Therefore, we used this dataset for following analyses and model assessments.

Download:

Fig 3. PRISMA flow diagram.

PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analysis.

https://doi.org/10.1371/journal.pone.0219159.g003

Download:

Table 1. List derived from literature search of methods to measure nitrification-associated N₂O flux.

https://doi.org/10.1371/journal.pone.0219159.t001

The observational fN₂O_nit values (n = 71) were distributed widely from 0.006% to 29.4%, with a mean of 1.92% and a median of 0.19% (first column of Table 2). The high mean value was likely attributable to the presence of a few anomalous values in the dataset. A histogram of the observed fN₂O_nit values (Fig 4) showed a concentrated and skewed distribution with a clear peak around the median value. Additionally, with removal of 10% outliers (top 4 and bottom 3 records, second column of Table 2) the mean (0.43%) and median (0.14%) values became lower. Notably, standard deviation narrowed greatly after the removal of outliers, and both the maximum value and minimum value were obtained when the C₂H₂ inhibition method was used. When publication-based data (i.e., the means of the values reported in each paper) were used, the mean and median value became higher (5.17% and 0.57%, respectively; third column of Table 2).

Download:

Fig 4. Histogram of the N₂O emission associated with nitrification, obtained by a meta-analysis of 71 observations.

(a) All data, (b) data of 0–1%, and (c) data of 1–10%.

https://doi.org/10.1371/journal.pone.0219159.g004

Download:

Table 2. Summary of statistics on measured nitrification N₂O emission ratio.

https://doi.org/10.1371/journal.pone.0219159.t002

In comparison with the study of Australian agricultural soils by Farquharson (2016), typical values of fN₂O_nit obtained by our meta-analysis seem comparable. That study found that fN₂O_nit values varied from 0.03% to 1%, with a typical value of 0.2%, that falls between the mean and median values obtained in the present study for all data. When removing outlier values, our results of mean and median became even closer to the result of Farquharson (2016). Namely, we examined anomalous values removed from the present meta-analysis, such as zero to negative and extremely high, 100% values. Several these values were obtained by Ambus (1998) [53] in the only study that conducted C₂H₂ treatment in the field; this probably led to a larger fluctuation in values than in the laboratory studies. Negative values of fN₂O_nit are attributable to net N₂O uptakes, which are sometimes observed but are usually small [54]. The fN₂O_nit values used by emission models assuming a constant N₂O emission fraction (0.5% to 2.0%) fell within the range of observed values. For example, the constant fN₂O_nit value used in the original VISIT model, 1%, did not differ significantly from the mean value of the observed dataset (Student’s t-test, t = 1.36, p = 0.177). However, it should be noted that the median observed value (0.19%) was much lower than the model-assumed value.

The fN₂O_nit values used in the DNDC and DLEM models including the environmental variability of the N₂O emission fraction were generally lower than those used in the models assuming constant values. As shown in Fig 2, these models assume the maximum fN₂O_nit values of 0.06 to 0.1%. However, these fN₂O_nit values were still within the range of observed values. We examined the spatial distribution of mean fN₂O_nit values estimated by the DNDC (Fig 5A) and DLEM (Fig 5B). When the DNDC parameterization was used, higher fN₂O_nit values were estimated mainly in the humid tropics. In contrast, when the DLEM parameterization was used, higher fN₂O_nit values were obtained in humid temperate to boreal regions such as Europe, eastern and western North America, and the Tibetan Plateau. Note again that a common soil temperature and moisture (after the VISIT simulation) were used in this comparison, and the differences among the results were caused exclusively by differences in the fN₂O_nit parameterizations. The difference of fN₂O_nit indicated here may account for a part of outcomes of the N₂O model intercomparison project [29]. The project showed that the existing models differ in global soil N₂O emission by about 20%, and our study implies that fN₂O_nit is one of the key parameters to reduce the estimation uncertainty.

Download:

Fig 5. Distributions of estimated fraction of nitrification-associated N₂O emission, fN₂O_nit.

(a) DNDC and (b) DLEM parameterizations.

https://doi.org/10.1371/journal.pone.0219159.g005

Sensitivity of N₂O flux to fN₂O_nit

The default VISIT model with a constant fN₂O_nit of 1% estimated total N₂O emission from terrestrial soils as 15.47 ± 0.52, 16.32 ± 0.98, and 17.03 ± 0.73 Tg N₂O yr^–1 (average ± s.d. of interannual variability) in the 1980s, 1990s, and 2000s, respectively (Fig 6). When converted into nitrogen weight (multiplied by 28/44), these values correspond to 9.85 to 10.84 Tg N yr^–1. These estimates are close to previous estimations: e.g., 11.1 Tg N yr^–1 by IPCC (2013) [1] and 11.4 Tg N yr^–1 by Syakila and Kroeze (2011) [55] for emissions from natural vegetation, agriculture, and deposition on land. In the 2000s simulation, 16.5% of total N₂O emission was from nitrification and 84% was from denitrification. About 32% and 68% of N₂O emission occurred in agricultural and natural ecosystem soils, respectively. During the simulation period, total N₂O emission increased from 12.11 Tg N₂O yr^–1 in 1901 to 18.60 Tg N₂O yr^–1 in 2016 as a result of the increase of fertilizer and deposition inputs. See our resent study [32] on the temporal change and its driver of the regional N₂O emissions.

Download:

Fig 6. Sensitivity analysis of global N₂O emission to the fraction of nitrification-associated N₂O emission, fN₂O_nit.

(a) Fixed 1%, (b) soil pH-based parameterization, and (c) median of meta-analysis records (0.139%, outliers removed). Each N₂O flux was estimated by using the VISIT model. Decadal mean values for the 1980s, 1990s, and 2000s are shown.

https://doi.org/10.1371/journal.pone.0219159.g006

The sensitivity analysis using different values of fN₂O_nit (constant) indicated that the simulated total N₂O emission was sensitive to the assumed emission fraction. When fN₂O_nit = 0.5% was used, the total nitrification-associated N₂O emission was reduced to 1.59 Tg N₂O yr^–1 (–43% in comparison with fN₂O_nit = 1% case) in the 2000s. Because of the smaller nitrogen loss by nitrification in these cases, N₂O emission from denitrification increased slightly because of the use of excess inorganic nitrogen in the soils. As a result of compensation, total N₂O emission was only slightly affected (–4.3%) by the halved fN₂O_nit value. When fN₂O_nit = 2% was used, total nitrification-associated N₂O emission increased to 4.4 Tg N₂O yr^–1 (+56.6%). The asymmetric sensitivity of nitrification-associated N₂O emission to the change in fN₂O_nit value is attributable to alteration of the nitrogen stock in the soils and the non-linear response of N₂O emission to nitrogen availability [36]. Finally, when using the parameterizations of DNDC and DLEM models, lower total N₂O emissions were estimated (15.0 Tg N₂O yr^–1) with lower contribution of nitrification-associated emission due to the generally low value of fN₂O_nit (data not shown).

When the median fN₂O_nit value of the meta-analysis (0.14%) was used in the VISIT simulation, the total N₂O emission was estimated as 15.4 Tg N₂O yr^–1 in the 2000s; nitrification-associated N₂O emission was largely reduced to 0.50 Tg N₂O yr^–1. In contrast, when the mean value of all data (fN₂O_nit = 1.92%) was used, higher rates of total and nitrification-associated N₂O emission (17.5 and 4.3 Tg N₂O yr^–1) were estimated. Therefore, selection of representative fN₂O_nit value can affect the simulation result by as much as 2.1 Tg N₂O yr^–1 at the global scale. When including the difference in model parameterizations, the uncertainty becomes even larger to 2.5 Tg N₂O yr^–1.

These results confirmed that the estimated N₂O emission was sensitive to the assumed fN₂O_nit value, which was poorly constrained in the present models and varied with selection of the metric from the observational data. Although observations implied that the fN₂O_nit can be variable in response to environmental conditions such as temperature and moisture, the scarcity of observational evidence has prevented us to use a standard parameterization and permitted us to assume constant values. Apparently, additional constraints and new parameterizations are required to obtain a reliable N₂O budget and its flow components. Observational data and insights are accumulating with support of technical developments such as isotopic tracers, but it would take decades to obtain a comprehensive dataset with enough coverage. Next, we made an attempt to develop a new parameterization of fN₂O_nit applicable at the global scale.

Application of pH-based parameterization

In 54 records of the meta-analysis data, soil pH condition was included, allowing us to relate with fN₂O_nit (Fig 7A). It was found that fN₂O_nit takes higher values at acidic soil conditions with pH below 5 and lower values under neutral to alkaline soil conditions. We obtained a regression curve using exponential function, which gives slightly higher fN₂O_nit values in comparison with the equation of Martikainen (1985) [40]. Using the soil pH map and the regression curve, global distribution of pH-based fN₂O_nit was obtained (Fig 7B). As expected from the pH pattern, boreal conifer forest soils and humid tropical soils show higher fN₂O_nit values. High fN₂O_nit in humid tropics estimated by the present study is consistent with those by DNDC and DLEM parameterizations, while the three maps differ largely in temperate regions. When using the pH-based parameterization, total N₂O emission in the 2000s was estimated as 16.8 Tg N₂O yr^–1 (25.2% by nitrification and 74.8% by denitrification). Global distribution of soil N₂O emission was reasonably simulated (Fig 8; see S2 Fig for seasonal change), in comparison with those obtained by atmospheric inversion studies [56]. For example, high emissions from temperate croplands and tropical forests were well captured.

Download:

Fig 7. Parameterization of nitrification-associated N₂O emission fraction, fN₂O_nit as a function of soil pH.

(a) Relationships in the meta-analysis data. Orange dashed curve is obtained by Gauss-Newton non-linear regression of an exponential function: fN₂O_nit = 47.59 exp(–1.345 · pH) (R² = 0.557). Grey curve shows an empirical function by Martikainen (1985) for reference. (b) Global distribution of fN₂O_nit estimated using the regression curve and soil pH map (S1 Fig).

https://doi.org/10.1371/journal.pone.0219159.g007

Download:

Fig 8. Simulated distribution of N₂O emission using VISIT model with the pH-based fN₂O_nit parameterization.

(a) Global map for the 2000s and (b) frequency distribution of N₂O emission intensity; note the log scale of y-axis.

https://doi.org/10.1371/journal.pone.0219159.g008

Impacts on global N₂O budget

The simulated N₂O emission account for important features of the global budget. For example, interannual variability in the total N₂O emission was comparable with the atmospheric growth rate especially after 1990, i.e. the period when ample observational data became available (Fig 9). The decline after the Mt. Pinatubo eruption in 1991 and following increase were well captured, implying the major impact of soil emission on the atmospheric N₂O variability in recent decades. As clearly shown in the relationship between nitrogen input and N₂O emission (Fig 10A), the historical increase of N₂O emission in recent decades is mainly attributable to increased land N inputs by atmospheric deposition and fertilizer use. The slope, so-called emission factor, was estimated as 1.75%. This is a bit higher than the typical emission factor value of IPCC guideline [57], 1%, but note that the present estimate includes the effects of climate and land-use changes. In the model simulation, N₂O emissions from nitrification and denitrification increased in parallel, as shown by the linear relationship between the two emissions (Fig 10B). Validating the N₂O production scheme at broad scales is difficult even by comparing with inversion studies. In forthcoming studies, appropriate observations of N₂O isotopomers may provide supporting evidences [58, 59].

Download:

Fig 9. Time-series of total soil N₂O emissions estimated using VISIT model with the pH-based fN₂O_nit parameterization.

Observed global mean growth rate of atmospheric N₂O by the World Data Center for Greenhouse Gases (https://ds.data.jma.go.jp/gmd/wdcgg/wdcgg.html) are shown for reference.

https://doi.org/10.1371/journal.pone.0219159.g009

Download:

Fig 10. Relationships in the simulated global nitrogen budget by VISIT model with the pH-based parameterization of fN₂O_nit.

(a) Total N₂O emission related to nitrogen input by biological fixation, atmospheric deposition, and fertilizer. Orange dashed line shows liner regression: N₂O emission = 0.0175 N-input + 5.94 (R² = 0.804). (b) Relationship between nitrification- and denitrification-associated N₂O emission: N₂O-denitrification = 3.57 N₂O-nitrification– 2.33 (R² = 0.794). Simulation data from 1901 to 2015 were used.

https://doi.org/10.1371/journal.pone.0219159.g010

Our meta-analysis and model simulations suggest the importance and uncertainty of fN₂O_nit values in the evaluation of global N₂O budget. Variability in a single parameter could cause a difference in total N₂O emission of as much as 2.5 Tg N₂O yr^–1 (in the 2000s, 15.0 to 17.5 Tg N₂O yr^–1)—equivalent to the variability of 0.2 Pg CO₂-C yr^–1 (based on a global warming potential of 298 for N₂O with 100-yr horizon [1]). The magnitude of the estimated N₂O emission increase from the beginning to the end of the simulation was about 3.9 Tg N₂O yr^–1 in the simulations with different fN₂O_nit values (comparable with a simulation by O-CN model [60] and a global synthesis [55]). Because the global nitrogen cycle would be further perturbed by human activities and climate change [5], the uncertainty in the present models can be a critical limiting factor for environmental management. Although global N₂O budget may be constrained by using atmospheric observational data to some extent, in-depth understanding of flow components and their environmental regulations is essential to conduct effective nitrogen and climate managements.

Concluding remarks

To our knowledge, this is the first study to have focused on fN₂O_nit in a comprehensive manner. We should pay attention to the fact that this study used a limited number of observational data and extrapolated them to the global scale. Nevertheless, the dataset covering a variety of ecosystems and soils and the process-based model assessment gave us clues to better understanding of N₂O cycle. Our findings gives an explanation for the results of the N₂O model intercomparison project, which shows 20% of global soil N₂O emission difference among terrestrial models [61]. In our analysis, selection of fN₂O_nit affected the estimation of global N₂O emission by about 15% (2.5/16 Tg N₂O yr^–1). Because of the scarcity of reliable observational data, our meta-analysis did not give a conclusive value or equation for fN₂O_nit. Most of observed values were low (<1%) but neglecting high values may result in underestimation of nitrification-associated N₂O emission at broad scales. We found a potential and representative range of fN₂O_nit values and a useful pH-based empirical model covering both natural and agricultural soils. This result should encourage extensive observations of nitrification-associated N₂O emission and fN₂O_nit by using a standardized protocol especially in the field. Although the present meta-analysis showed that the majority of empirical data were obtained by the C₂H₂ inhibition or isotopic tracer methods, further discussions on effective measurement strategy (e.g., spatial and temporal coverage and representativeness) are required for field and model researchers to improve model accuracy. Because of extreme complexity of the soil biogeochemical processes, it is inevitable to use simplified schemes like the ‘hole-in-a-pipe’ concept and bulk parameters like fN₂O_nit to conduct simulations at broad scales. Because N₂O has a high global warming potential, a small difference in estimated N₂O emission can considerably influence the total greenhouse gas budget, as shown by our sensitivity simulations. To develop a better parameterization of fN₂O_nit and other, related properties, further observations—especially in the field—and process studies of the nitrogen cycle and greenhouse gas emissions are critically important.

Supporting information

S1 Table. PRISMA 2009 check list.

https://doi.org/10.1371/journal.pone.0219159.s001

(DOC)

S2 Table. List of search terms.

https://doi.org/10.1371/journal.pone.0219159.s002

(XLSX)

S3 Table. Data extracted from the literature and used by the meta-analysis.

https://doi.org/10.1371/journal.pone.0219159.s003

(XLSX)

S1 Fig. Map of soil pH (Global Soil Data Task, International Geosphere-Biosphere Program).

https://doi.org/10.1371/journal.pone.0219159.s004

(TIF)

S2 Fig. Seasonal change in total N₂O emission simulated by the VISIT model.

(a) Norther winter (DJF: December, January, and February), (b) northern spring (MAM: March, April, and May), (c) northern summer (JJA: June, July, and August), and (d) northern autumn (SON: September, October, and November).

https://doi.org/10.1371/journal.pone.0219159.s005

(TIF)

References

1. Intergovernmental Panel on Climate Change (IPCC). Climate Change 2013: The Physical Science Basis. 2013:996.
- View Article
- Google Scholar
2. Ravishankara AR, Daniel JS, Portmann RW. Nitrous oxide (N₂O): The dominant ozone-depleting substance emitted in the 21st century. Science 2009; 326:123–5. pmid:19713491
- View Article
- PubMed/NCBI
- Google Scholar
3. Rogelj J, den Elzen M, Höhne N, Fransen T, Fekete H, Winkler H, et al. Paris Agreement climate proposals need a boost to keep warming well below 2°C. Nature 2016; 534: 631–9. pmid:27357792
- View Article
- PubMed/NCBI
- Google Scholar
4. Bouwman AF, Fung I, Matthews E, John J. Global analysis of the potential for N₂O production in natural soils. Global Biogeochem Cycles 1993; 7(3): 557–97.
- View Article
- Google Scholar
5. Galloway JN, Townsend AR, Erisman JW, Bekunda M, Cai Z, Freney JR, et al. Transformation of the nitrogen cycle: recent trends, questions, and potential solutions. Science 2008; 320: 889–92. https://doi.org/10.1126/science.1136674. pmid:18487183
- View Article
- PubMed/NCBI
- Google Scholar
6. Davidson EA. The contribution of manure and fertilizer nitrogen to atmospheric nitrous oxide since 1860. Nature Geoscience 2009; 2: 659–62. https://doi.org/10.1038/NGEO608
- View Article
- Google Scholar
7. Gruber N, Galloway JN. An Earth-system perspective of the global nitrogen cycle. Nature 2008; 451: 293–6. pmid:18202647
- View Article
- PubMed/NCBI
- Google Scholar
8. Tian H, Lu C, Ciais P, Michalak AM, Canadell JG, Saikawa E, et al. The terrestrial biosphere as a net source of greenhouse gases to the atmosphere. Nature 2016; 531: 225–8. pmid:26961656
- View Article
- PubMed/NCBI
- Google Scholar
9. Barnard R, Leadley PW, Hungate BA. Global change, nitrification, and denitrification: A review. Global Biogeochem Cycles 2005; 19(GB1007). https://doi.org/10.1029/2004GB002282GB1007
- View Article
- Google Scholar
10. Butterbach-Bahl K, Baggs EM, Dannenmann M, Kiese R, Zechmeister-Boltenstern S. Nitrous oxide emissions from soils: how well do we understand the processes and their controls. Phil Trans Royal Soc 2013; 2013 (20130122). https://doi.org/10.1098/rstb.2013.0122
- View Article
- Google Scholar
11. Bollmann A, Conrad R. Influence of O₂ availability on NO and N₂O release by nitrification and denitrification in soils. Global Change Biol 1998; 4: 387–96.
- View Article
- Google Scholar
12. Hink L, Nocol GW, Prosser JI. Archaea produce lower yields of N₂O than bacteria during aerobic ammonia oxidation in soil. Environmental Microbiology 2017; 19(12): 4829–37. pmid:26971439
- View Article
- PubMed/NCBI
- Google Scholar
13. Stark JM. Modeling the temperature response of nitrification. Biogeochemistry 1996; 35: 433–45.
- View Article
- Google Scholar
14. Bateman EJ, Baggs EM. Contributions of nitrification and denitrification to N₂O emissions from soils at different water-filled pore space. Biol Fertil Soils 2005; 41: 379–88.
- View Article
- Google Scholar
15. Farquharson R. Nitrification rates and associated nitrous oxide emissions from agricultural soils–a synopsis. Soil Research 2016; 54: 469–80. https://doi.org/10.1071/SR15304
- View Article
- Google Scholar
16. Raich JW, Rastetter EB, Melillo JM, Kicklighter DW, Grace AL, Moore III B, et al. Potential net primary productivity in South America: application of a global model. Ecol Appl 1991; 1: 399–429. pmid:27755669
- View Article
- PubMed/NCBI
- Google Scholar
17. Li C, Frolking S, Frolking TA. A model of nitrous oxide evolution from soil driven by rainfall events: 1. model structure and sensitivity. J Geophys Res 1992; 97: 9759–76.
- View Article
- Google Scholar
18. Firestone MK, Davidson EA. Microbiological basis of NO and N₂O production and consumption in soil. In: Andreae MO, Schimel DS, editors. Exchange of Trace Gases between Terrestrial Ecosystems and the Atmosphere: John Wiley and Sons Ltd; 1989. p. 7–21.
19. Davidson EA, Keller M, Erickson HE, Vershot LV, Veldkamp E. Testing a conceptual model of soil emissions of nitrous and nitric oxides. BioScience 2000; 50(8): 667–80.
- View Article
- Google Scholar
20. Zhu X, Burger M, Doane TA, Horwath WR. Ammonia oxidation pathways and nitrifier denitrification are significant sources of N₂O and NO under low oxygen availability. Proc Nat Acad Sci USA 2013; 110(16): 6328–33. pmid:23576736
- View Article
- PubMed/NCBI
- Google Scholar
21. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic review and meta-analyses: PRISMA statement. BMJ 2009; 339: b2535. pmid:19622551
- View Article
- PubMed/NCBI
- Google Scholar
22. Chen D, Suter HC, Islam A, Edis R. Influence of nitrification inhibitors on nitrification and nitrous oxide (N₂O) emission from a clay loam soil fertilized with urea. Soil Biol Biochem 2010; 42: 660–4. https://doi.org/10.1016/j.soilbio.2009.12.014
- View Article
- Google Scholar
23. R-Core-Team. R version 3.4.2. 2017.
24. Frolking SE, Mosier AR, Ojima DS, Li C, Parton WJ, Potter CS, et al. Comparison of N₂O emissions from soils at three temperate agricultural sites: simulations of year-round measurements by four models. Nutr Cycl Agroecosys 1998; 52: 77–105.
- View Article
- Google Scholar
25. Li C, Aber J, Stange F, Butterbach-Bahl K, Papen H. A process-oriented model of N₂O and NO emissions from forest soils: 1. model development. J Geophys Res 2000; 105: 4369–84.
- View Article
- Google Scholar
26. Tian H, Xu X, Liu M, Ren W, Zhang C, Chen G, et al. Spatial and temporal patterns of CH₄ and N₂O fluxes in terrestrial ecosystems of North America during 1979–2008: application of a global biogeochemistry model. Biogeosci. 2010; 7: 2673–94. https://doi.org/10.5194/bg-7-2673-2010
- View Article
- Google Scholar
27. Saikawa E, Schlosser CA, Prinn RG. Global modeling of soil nitrous oxide emissions from natural processes. Global Biogeochem Cycles 2013; 27: 972–89. https://doi.org/10.1002/gbc.20087
- View Article
- Google Scholar
28. Zaehle S, Friend AD. Carbon and nitrogen cycle dynamics in the O-CN land surface model: 1. Model description, site-scale evaluation, and sensitivity to parameter estimates. Global Biogeochem Cycles 2010; 24 (GB1005). https://doi.org/10.1029/2009GB003521.
- View Article
- Google Scholar
29. Tian H, Yang J, Lu C, Xu R, Canadell JG, Jackson RB, et al. The global N₂O Model Intercomparison Project. Bull Am Meteorol Soc 2018; 99(6): 1231–51. https://doi.org/10.1175/BAMS-D-17-0212.1
- View Article
- Google Scholar
30. Inatomi M, Ito A, Ishijima K, Murayama S. Greenhouse gas budget of a cool temperate deciduous broadleaved forest in Japan estimated using a process-based model. Ecosystems 2010; 13: 472–83. https://doi.org/10.1007/s10021-010-9332-7
- View Article
- Google Scholar
31. Ito A, Inatomi M. Use of a process-based model for assessing the methane budgets of global terrestrial ecosystems and evaluation of uncertainty. Biogeosci 2012; 9: 759–73. https://doi.org/10.5194/bg-9-759-2012
- View Article
- Google Scholar
32. Ito A, Nishina K, Ishijima K, Hashimoto S, Inatomi M. Emissions of nitrous oxide (N₂O) from soil surfaces and their historical changes in East Asia: a model-based assessment. Prog Earth Planet Sci 2018; 5(55): https://doi.org/10.1186/s40645-018-0215-4
- View Article
- Google Scholar
33. Ito A, Oikawa T. A simulation model of the carbon cycle in land ecosystems (Sim-CYCLE): A description based on dry-matter production theory and plot-scale validation. Ecol Model 2002; 151: 147–79.
- View Article
- Google Scholar
34. Ito A. The regional carbon budget of East Asia simulated with a terrestrial ecosystem model and validated using AsiaFlux data. Agr For Meteorol 2008; 148: 738–47. https://doi.org/10.1016/j.agrformet.2007.12.007
- View Article
- Google Scholar
35. Ito A, Nishina K, Reyer CPO, François L, Henrot A-J, Munhoven G, et al. Photosynthetic productivity and its efficiencies in ISIMIP2a biome models: benchmarking for impact assessment studies. Env Res Lett 2017; 12 (085001). https://doi.org/10.1088/1748-9326/aa7a19
- View Article
- Google Scholar
36. Parton WJ, Mosier AR, Ojima DS, Valentine DW, Schimel DS, Weier K, et al. Generalized model for N₂ and N₂O production from nitrification and denitrification. Global Biogeochem Cycles 1996; 10: 401–12.
- View Article
- Google Scholar
37. Harris I, Jones PD, Osborn TJ, Lister DH. Updated high-resolution grids of monthly climatic observations–the CRU TS3.10 Dataset. Int J Climatol 2014; 34: 623–42. https://doi.org/10.1002/joc.3711
- View Article
- Google Scholar
38. Hurtt GC, Chini LP, Frolking S, Betts RA, Feddema J, Fischer G, et al. Harmonization of land-use scenarios for the period 1500–2100: 600 years of global gridded annual land-use transitions, wood harvest, and resulting secondary lands. Clim Chan 2011; 109: 117–61. https://doi.org/10.1007/s10584-011-0153-2
- View Article
- Google Scholar
39. Galloway JN, Dentener FJ, Capone DG, Boyer EW, Howarth RW, Seitzinger SP, et al. Nitrogen cycles: past, present, and future. Biogeochem 2004; 70: 153–226.
- View Article
- Google Scholar
40. Martikainen PJ. Nitrous oxide emission associated with autotrophic ammonium oxidation in acid coniferous forest soil. Appl Env Microbiol 1985; 50(6): 1519–25.
- View Article
- Google Scholar
41. IGBP-DIS. Global Soil Data Products CD-ROM. Oak Ridge, Tennessee, U.S.A.: Oak Ridge National Laboratory, 2000.
42. Mosier AR, Duxbury JM, Freney JR, Heinemeyer O, Minami K. Nitrous oxide emissions from agricultural fields: Assessment, measurement and mitigation. Plant Soil 1996; 181: 95–108.
- View Article
- Google Scholar
43. Ambus P. Relationship between gross nitrogen cycling and nitrous oxide emission in grass-clover pasture. Nutr Cycl Agroecosys 2005; 72: 189–99. https://doi.org/10.1007/s10705-005-1269-4
- View Article
- Google Scholar
44. Carter MS. Contribution of nitrification and denitrification to N₂O emissions from urine patches. Soil Biol Biochem 2007; 39: 2091–102. https://doi.org/10.1016/j.soilbio.2007.03.013
- View Article
- Google Scholar
45. Garrido F, Hénault C, Gaillard H, Pérez S, Germon JC. N₂O and NO emissions by agricultural soils with low hydraulic potentials. Soil Biol Biochem 2002; 34: 559–75.
- View Article
- Google Scholar
46. Khalil K, Mary B, Renault P. Nitrous oxide production by nitrification and denitrification in soil aggregates as affected by O₂ concentration. Soil Biol Biochem 2004; 36: 687–99.
- View Article
- Google Scholar
47. Klemedtsson L, Svensson BH, Rosswall T. A method of selective inhibition to distinguish between nitrification and denitrification as sources of nitrous oxide in soil. Biol Fertil Soils 1988; 6: 112–9.
- View Article
- Google Scholar
48. Maag M, Vinther FP. Nitrous oxide emission by nitrification and denitrification in different soil types and at different soil moisture contents and temperatures. Appl Soil Ecol 1996; 4: 5–14.
- View Article
- Google Scholar
49. Mathieu O, Hénault C, Lévêque J, Baujard E, Milloux M-J, Andreux MF. Quantifying the contribution of nitrification and denitrification to the nitrous oxide flux using ¹⁵N tracers. Env Pollut 2006; 144: 933–40. https://doi.org/10.1016/j.envpol.2006.02.005
- View Article
- Google Scholar
50. Mørkved PT, Dörsch P, Bakken LR. The N₂O product ratio of nitrification and its dependence on long-term changes in soil pH. Soil Biol Biochem 2007; 39: 2048–57.
- View Article
- Google Scholar
51. Mørkved PT, Dörsch P, Henriksen TM, Bakken LR. N₂O emissions and product ratios of nitrification and denitrification as affected by freezing and thawing. Soil Biol Biochem 2006; 38: 3411–20. https://doi.org/10.1016/j.soilbio.2006.05.015
- View Article
- Google Scholar
52. Tortoso AC, Hutchinson GL. Contributions of autotrophic and heterotrophic nitrifiers to soil NO and N₂O emissions. Appl Env Microbiol 1990; 56(6): 1799–805.
- View Article
- Google Scholar
53. Ambus P. Nitrous oxide production by denitrification and nitrification in temperate forest, grassland and agricultural soils. Eur J Soil Sci 1998; 49: 495–502.
- View Article
- Google Scholar
54. Chapuis-Lardy L, Wrage N, Metay A, Chotte J-L, Bernoux M. Soils, a sink for N₂O? A review. Global Change Biol 2007; 13: 1–17.
- View Article
- Google Scholar
55. Syakila A, Kroeze C. The global nitrous oxide budget revisited. Greenhouse Gas Measur Manag 2010; 1: 17–26. https://doi.org/10.3763/ghgmm.2010.0007
- View Article
- Google Scholar
56. Thompson RL, Patra PK, Ishijima K, Saikawa E, Corazza M, Karstens U, et al. TransCom N₂O model inter-comparison–Part 1: Assessing the influence of transport and surface fluxes on tropospheric N₂O variability. Atm Chem Phys 2014; 14: 4349–68. https://doi.org/10.5194/acp-14-4349-2014
- View Article
- Google Scholar
57. Intergovernmental Panel on Climate Change (IPCC). 2006 IPCC Guidance for National Greenhouse Gas Inventories: Cambridge University Press; 2006.
58. Yoshida N, Toyoda S. Constraining the atmospheric N₂O budget from intramolecular site preference in N₂O isotopomers. Nature 2000; 405: 330–4. pmid:10830958
- View Article
- PubMed/NCBI
- Google Scholar
59. Park S, Pérez T, Boering KA, Trumbore SE, Gil J, Marquina S, et al. Can N₂O stable isotopes and isotopomers be useful tools to characterize sources and microbial pathways of N₂O production and consumption in tropical soils? Global Biogeochem Cycles 2011; 25(GB1001). https://doi.org/10.1029/2009GB003615
- View Article
- Google Scholar
60. Zaehle S, Ciais P, Friend AD, Prieur V. Carbon benefits of anthropogenic reactive nitrogen offset by nitrous oxide emissions. Nature Geosci. 2011; 4: 601–5. https://doi.org/10.1038/NGEO1207
- View Article
- Google Scholar
61. Tian H, Yang J, Xu R, Lu C, Canadell JG, Davidson EA, et al. Global soil nitrous oxide emissions since the pre-industrial era estimated by an ensemble of terrestrial biosphere models: Magnitude, attribution and uncertainty. Global Change Biol 2019; 25: 640–59. https://doi.org/10.1111/gcb.14514
- View Article
- Google Scholar

[ref1] 1. Intergovernmental Panel on Climate Change (IPCC). Climate Change 2013: The Physical Science Basis. 2013:996.
View Article
Google Scholar

[2] View Article

[3] Google Scholar

[ref2] 2. Ravishankara AR, Daniel JS, Portmann RW. Nitrous oxide (N₂O): The dominant ozone-depleting substance emitted in the 21st century. Science 2009; 326:123–5. pmid:19713491
View Article
PubMed/NCBI
Google Scholar

[5] View Article

[6] PubMed/NCBI

[7] Google Scholar

[ref3] 3. Rogelj J, den Elzen M, Höhne N, Fransen T, Fekete H, Winkler H, et al. Paris Agreement climate proposals need a boost to keep warming well below 2°C. Nature 2016; 534: 631–9. pmid:27357792
View Article
PubMed/NCBI
Google Scholar

[9] View Article

[10] PubMed/NCBI

[11] Google Scholar

[ref4] 4. Bouwman AF, Fung I, Matthews E, John J. Global analysis of the potential for N₂O production in natural soils. Global Biogeochem Cycles 1993; 7(3): 557–97.
View Article
Google Scholar

[13] View Article

[14] Google Scholar

[ref5] 5. Galloway JN, Townsend AR, Erisman JW, Bekunda M, Cai Z, Freney JR, et al. Transformation of the nitrogen cycle: recent trends, questions, and potential solutions. Science 2008; 320: 889–92. https://doi.org/10.1126/science.1136674. pmid:18487183
View Article
PubMed/NCBI
Google Scholar

[16] View Article

[17] PubMed/NCBI

[18] Google Scholar

[ref6] 6. Davidson EA. The contribution of manure and fertilizer nitrogen to atmospheric nitrous oxide since 1860. Nature Geoscience 2009; 2: 659–62. https://doi.org/10.1038/NGEO608
View Article
Google Scholar

[20] View Article

[21] Google Scholar

[ref7] 7. Gruber N, Galloway JN. An Earth-system perspective of the global nitrogen cycle. Nature 2008; 451: 293–6. pmid:18202647
View Article
PubMed/NCBI
Google Scholar

[23] View Article

[24] PubMed/NCBI

[25] Google Scholar

[ref8] 8. Tian H, Lu C, Ciais P, Michalak AM, Canadell JG, Saikawa E, et al. The terrestrial biosphere as a net source of greenhouse gases to the atmosphere. Nature 2016; 531: 225–8. pmid:26961656
View Article
PubMed/NCBI
Google Scholar

[27] View Article

[28] PubMed/NCBI

[29] Google Scholar

[ref9] 9. Barnard R, Leadley PW, Hungate BA. Global change, nitrification, and denitrification: A review. Global Biogeochem Cycles 2005; 19(GB1007). https://doi.org/10.1029/2004GB002282GB1007
View Article
Google Scholar

[31] View Article

[32] Google Scholar

[ref10] 10. Butterbach-Bahl K, Baggs EM, Dannenmann M, Kiese R, Zechmeister-Boltenstern S. Nitrous oxide emissions from soils: how well do we understand the processes and their controls. Phil Trans Royal Soc 2013; 2013 (20130122). https://doi.org/10.1098/rstb.2013.0122
View Article
Google Scholar

[34] View Article

[35] Google Scholar

[ref11] 11. Bollmann A, Conrad R. Influence of O₂ availability on NO and N₂O release by nitrification and denitrification in soils. Global Change Biol 1998; 4: 387–96.
View Article
Google Scholar

[37] View Article

[38] Google Scholar

[ref12] 12. Hink L, Nocol GW, Prosser JI. Archaea produce lower yields of N₂O than bacteria during aerobic ammonia oxidation in soil. Environmental Microbiology 2017; 19(12): 4829–37. pmid:26971439
View Article
PubMed/NCBI
Google Scholar

[40] View Article

[41] PubMed/NCBI

[42] Google Scholar

[ref13] 13. Stark JM. Modeling the temperature response of nitrification. Biogeochemistry 1996; 35: 433–45.
View Article
Google Scholar

[44] View Article

[45] Google Scholar

[ref14] 14. Bateman EJ, Baggs EM. Contributions of nitrification and denitrification to N₂O emissions from soils at different water-filled pore space. Biol Fertil Soils 2005; 41: 379–88.
View Article
Google Scholar

[47] View Article

[48] Google Scholar

[ref15] 15. Farquharson R. Nitrification rates and associated nitrous oxide emissions from agricultural soils–a synopsis. Soil Research 2016; 54: 469–80. https://doi.org/10.1071/SR15304
View Article
Google Scholar

[50] View Article

[51] Google Scholar

[ref16] 16. Raich JW, Rastetter EB, Melillo JM, Kicklighter DW, Grace AL, Moore III B, et al. Potential net primary productivity in South America: application of a global model. Ecol Appl 1991; 1: 399–429. pmid:27755669
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref17] 17. Li C, Frolking S, Frolking TA. A model of nitrous oxide evolution from soil driven by rainfall events: 1. model structure and sensitivity. J Geophys Res 1992; 97: 9759–76.
View Article
Google Scholar

[57] View Article

[58] Google Scholar

[ref18] 18. Firestone MK, Davidson EA. Microbiological basis of NO and N₂O production and consumption in soil. In: Andreae MO, Schimel DS, editors. Exchange of Trace Gases between Terrestrial Ecosystems and the Atmosphere: John Wiley and Sons Ltd; 1989. p. 7–21.

[ref19] 19. Davidson EA, Keller M, Erickson HE, Vershot LV, Veldkamp E. Testing a conceptual model of soil emissions of nitrous and nitric oxides. BioScience 2000; 50(8): 667–80.
View Article
Google Scholar

[61] View Article

[62] Google Scholar

[ref20] 20. Zhu X, Burger M, Doane TA, Horwath WR. Ammonia oxidation pathways and nitrifier denitrification are significant sources of N₂O and NO under low oxygen availability. Proc Nat Acad Sci USA 2013; 110(16): 6328–33. pmid:23576736
View Article
PubMed/NCBI
Google Scholar

[64] View Article

[65] PubMed/NCBI

[66] Google Scholar

[ref21] 21. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic review and meta-analyses: PRISMA statement. BMJ 2009; 339: b2535. pmid:19622551
View Article
PubMed/NCBI
Google Scholar

[68] View Article

[69] PubMed/NCBI

[70] Google Scholar

[ref22] 22. Chen D, Suter HC, Islam A, Edis R. Influence of nitrification inhibitors on nitrification and nitrous oxide (N₂O) emission from a clay loam soil fertilized with urea. Soil Biol Biochem 2010; 42: 660–4. https://doi.org/10.1016/j.soilbio.2009.12.014
View Article
Google Scholar

[72] View Article

[73] Google Scholar

[ref23] 23. R-Core-Team. R version 3.4.2. 2017.

[ref24] 24. Frolking SE, Mosier AR, Ojima DS, Li C, Parton WJ, Potter CS, et al. Comparison of N₂O emissions from soils at three temperate agricultural sites: simulations of year-round measurements by four models. Nutr Cycl Agroecosys 1998; 52: 77–105.
View Article
Google Scholar

[76] View Article

[77] Google Scholar

[ref25] 25. Li C, Aber J, Stange F, Butterbach-Bahl K, Papen H. A process-oriented model of N₂O and NO emissions from forest soils: 1. model development. J Geophys Res 2000; 105: 4369–84.
View Article
Google Scholar

[79] View Article

[80] Google Scholar

[ref26] 26. Tian H, Xu X, Liu M, Ren W, Zhang C, Chen G, et al. Spatial and temporal patterns of CH₄ and N₂O fluxes in terrestrial ecosystems of North America during 1979–2008: application of a global biogeochemistry model. Biogeosci. 2010; 7: 2673–94. https://doi.org/10.5194/bg-7-2673-2010
View Article
Google Scholar

[82] View Article

[83] Google Scholar

[ref27] 27. Saikawa E, Schlosser CA, Prinn RG. Global modeling of soil nitrous oxide emissions from natural processes. Global Biogeochem Cycles 2013; 27: 972–89. https://doi.org/10.1002/gbc.20087
View Article
Google Scholar

[85] View Article

[86] Google Scholar

[ref28] 28. Zaehle S, Friend AD. Carbon and nitrogen cycle dynamics in the O-CN land surface model: 1. Model description, site-scale evaluation, and sensitivity to parameter estimates. Global Biogeochem Cycles 2010; 24 (GB1005). https://doi.org/10.1029/2009GB003521.
View Article
Google Scholar

[88] View Article

[89] Google Scholar

[ref29] 29. Tian H, Yang J, Lu C, Xu R, Canadell JG, Jackson RB, et al. The global N₂O Model Intercomparison Project. Bull Am Meteorol Soc 2018; 99(6): 1231–51. https://doi.org/10.1175/BAMS-D-17-0212.1
View Article
Google Scholar

[91] View Article

[92] Google Scholar

[ref30] 30. Inatomi M, Ito A, Ishijima K, Murayama S. Greenhouse gas budget of a cool temperate deciduous broadleaved forest in Japan estimated using a process-based model. Ecosystems 2010; 13: 472–83. https://doi.org/10.1007/s10021-010-9332-7
View Article
Google Scholar

[94] View Article

[95] Google Scholar

[ref31] 31. Ito A, Inatomi M. Use of a process-based model for assessing the methane budgets of global terrestrial ecosystems and evaluation of uncertainty. Biogeosci 2012; 9: 759–73. https://doi.org/10.5194/bg-9-759-2012
View Article
Google Scholar

[97] View Article

[98] Google Scholar

[ref32] 32. Ito A, Nishina K, Ishijima K, Hashimoto S, Inatomi M. Emissions of nitrous oxide (N₂O) from soil surfaces and their historical changes in East Asia: a model-based assessment. Prog Earth Planet Sci 2018; 5(55): https://doi.org/10.1186/s40645-018-0215-4
View Article
Google Scholar

[100] View Article

[101] Google Scholar

[ref33] 33. Ito A, Oikawa T. A simulation model of the carbon cycle in land ecosystems (Sim-CYCLE): A description based on dry-matter production theory and plot-scale validation. Ecol Model 2002; 151: 147–79.
View Article
Google Scholar

[103] View Article

[104] Google Scholar

[ref34] 34. Ito A. The regional carbon budget of East Asia simulated with a terrestrial ecosystem model and validated using AsiaFlux data. Agr For Meteorol 2008; 148: 738–47. https://doi.org/10.1016/j.agrformet.2007.12.007
View Article
Google Scholar

[106] View Article

[107] Google Scholar

[ref35] 35. Ito A, Nishina K, Reyer CPO, François L, Henrot A-J, Munhoven G, et al. Photosynthetic productivity and its efficiencies in ISIMIP2a biome models: benchmarking for impact assessment studies. Env Res Lett 2017; 12 (085001). https://doi.org/10.1088/1748-9326/aa7a19
View Article
Google Scholar

[109] View Article

[110] Google Scholar

[ref36] 36. Parton WJ, Mosier AR, Ojima DS, Valentine DW, Schimel DS, Weier K, et al. Generalized model for N₂ and N₂O production from nitrification and denitrification. Global Biogeochem Cycles 1996; 10: 401–12.
View Article
Google Scholar

[112] View Article

[113] Google Scholar

[ref37] 37. Harris I, Jones PD, Osborn TJ, Lister DH. Updated high-resolution grids of monthly climatic observations–the CRU TS3.10 Dataset. Int J Climatol 2014; 34: 623–42. https://doi.org/10.1002/joc.3711
View Article
Google Scholar

[115] View Article

[116] Google Scholar

[ref38] 38. Hurtt GC, Chini LP, Frolking S, Betts RA, Feddema J, Fischer G, et al. Harmonization of land-use scenarios for the period 1500–2100: 600 years of global gridded annual land-use transitions, wood harvest, and resulting secondary lands. Clim Chan 2011; 109: 117–61. https://doi.org/10.1007/s10584-011-0153-2
View Article
Google Scholar

[118] View Article

[119] Google Scholar

[ref39] 39. Galloway JN, Dentener FJ, Capone DG, Boyer EW, Howarth RW, Seitzinger SP, et al. Nitrogen cycles: past, present, and future. Biogeochem 2004; 70: 153–226.
View Article
Google Scholar

[121] View Article

[122] Google Scholar

[ref40] 40. Martikainen PJ. Nitrous oxide emission associated with autotrophic ammonium oxidation in acid coniferous forest soil. Appl Env Microbiol 1985; 50(6): 1519–25.
View Article
Google Scholar

[124] View Article

[125] Google Scholar

[ref41] 41. IGBP-DIS. Global Soil Data Products CD-ROM. Oak Ridge, Tennessee, U.S.A.: Oak Ridge National Laboratory, 2000.

[ref42] 42. Mosier AR, Duxbury JM, Freney JR, Heinemeyer O, Minami K. Nitrous oxide emissions from agricultural fields: Assessment, measurement and mitigation. Plant Soil 1996; 181: 95–108.
View Article
Google Scholar

[128] View Article

[129] Google Scholar

[ref43] 43. Ambus P. Relationship between gross nitrogen cycling and nitrous oxide emission in grass-clover pasture. Nutr Cycl Agroecosys 2005; 72: 189–99. https://doi.org/10.1007/s10705-005-1269-4
View Article
Google Scholar

[131] View Article

[132] Google Scholar

[ref44] 44. Carter MS. Contribution of nitrification and denitrification to N₂O emissions from urine patches. Soil Biol Biochem 2007; 39: 2091–102. https://doi.org/10.1016/j.soilbio.2007.03.013
View Article
Google Scholar

[134] View Article

[135] Google Scholar

[ref45] 45. Garrido F, Hénault C, Gaillard H, Pérez S, Germon JC. N₂O and NO emissions by agricultural soils with low hydraulic potentials. Soil Biol Biochem 2002; 34: 559–75.
View Article
Google Scholar

[137] View Article

[138] Google Scholar

[ref46] 46. Khalil K, Mary B, Renault P. Nitrous oxide production by nitrification and denitrification in soil aggregates as affected by O₂ concentration. Soil Biol Biochem 2004; 36: 687–99.
View Article
Google Scholar

[140] View Article

[141] Google Scholar

[ref47] 47. Klemedtsson L, Svensson BH, Rosswall T. A method of selective inhibition to distinguish between nitrification and denitrification as sources of nitrous oxide in soil. Biol Fertil Soils 1988; 6: 112–9.
View Article
Google Scholar

[143] View Article

[144] Google Scholar

[ref48] 48. Maag M, Vinther FP. Nitrous oxide emission by nitrification and denitrification in different soil types and at different soil moisture contents and temperatures. Appl Soil Ecol 1996; 4: 5–14.
View Article
Google Scholar

[146] View Article

[147] Google Scholar

[ref49] 49. Mathieu O, Hénault C, Lévêque J, Baujard E, Milloux M-J, Andreux MF. Quantifying the contribution of nitrification and denitrification to the nitrous oxide flux using ¹⁵N tracers. Env Pollut 2006; 144: 933–40. https://doi.org/10.1016/j.envpol.2006.02.005
View Article
Google Scholar

[149] View Article

[150] Google Scholar

[ref50] 50. Mørkved PT, Dörsch P, Bakken LR. The N₂O product ratio of nitrification and its dependence on long-term changes in soil pH. Soil Biol Biochem 2007; 39: 2048–57.
View Article
Google Scholar

[152] View Article

[153] Google Scholar

[ref51] 51. Mørkved PT, Dörsch P, Henriksen TM, Bakken LR. N₂O emissions and product ratios of nitrification and denitrification as affected by freezing and thawing. Soil Biol Biochem 2006; 38: 3411–20. https://doi.org/10.1016/j.soilbio.2006.05.015
View Article
Google Scholar

[155] View Article

[156] Google Scholar

[ref52] 52. Tortoso AC, Hutchinson GL. Contributions of autotrophic and heterotrophic nitrifiers to soil NO and N₂O emissions. Appl Env Microbiol 1990; 56(6): 1799–805.
View Article
Google Scholar

[158] View Article

[159] Google Scholar

[ref53] 53. Ambus P. Nitrous oxide production by denitrification and nitrification in temperate forest, grassland and agricultural soils. Eur J Soil Sci 1998; 49: 495–502.
View Article
Google Scholar

[161] View Article

[162] Google Scholar

[ref54] 54. Chapuis-Lardy L, Wrage N, Metay A, Chotte J-L, Bernoux M. Soils, a sink for N₂O? A review. Global Change Biol 2007; 13: 1–17.
View Article
Google Scholar

[164] View Article

[165] Google Scholar

[ref55] 55. Syakila A, Kroeze C. The global nitrous oxide budget revisited. Greenhouse Gas Measur Manag 2010; 1: 17–26. https://doi.org/10.3763/ghgmm.2010.0007
View Article
Google Scholar

[167] View Article

[168] Google Scholar

[ref56] 56. Thompson RL, Patra PK, Ishijima K, Saikawa E, Corazza M, Karstens U, et al. TransCom N₂O model inter-comparison–Part 1: Assessing the influence of transport and surface fluxes on tropospheric N₂O variability. Atm Chem Phys 2014; 14: 4349–68. https://doi.org/10.5194/acp-14-4349-2014
View Article
Google Scholar

[170] View Article

[171] Google Scholar

[ref57] 57. Intergovernmental Panel on Climate Change (IPCC). 2006 IPCC Guidance for National Greenhouse Gas Inventories: Cambridge University Press; 2006.

[ref58] 58. Yoshida N, Toyoda S. Constraining the atmospheric N₂O budget from intramolecular site preference in N₂O isotopomers. Nature 2000; 405: 330–4. pmid:10830958
View Article
PubMed/NCBI
Google Scholar

[174] View Article

[175] PubMed/NCBI

[176] Google Scholar

[ref59] 59. Park S, Pérez T, Boering KA, Trumbore SE, Gil J, Marquina S, et al. Can N₂O stable isotopes and isotopomers be useful tools to characterize sources and microbial pathways of N₂O production and consumption in tropical soils? Global Biogeochem Cycles 2011; 25(GB1001). https://doi.org/10.1029/2009GB003615
View Article
Google Scholar

[178] View Article

[179] Google Scholar

[ref60] 60. Zaehle S, Ciais P, Friend AD, Prieur V. Carbon benefits of anthropogenic reactive nitrogen offset by nitrous oxide emissions. Nature Geosci. 2011; 4: 601–5. https://doi.org/10.1038/NGEO1207
View Article
Google Scholar

[181] View Article

[182] Google Scholar

[ref61] 61. Tian H, Yang J, Xu R, Lu C, Canadell JG, Davidson EA, et al. Global soil nitrous oxide emissions since the pre-industrial era estimated by an ensemble of terrestrial biosphere models: Magnitude, attribution and uncertainty. Global Change Biol 2019; 25: 640–59. https://doi.org/10.1111/gcb.14514
View Article
Google Scholar

[184] View Article

[185] Google Scholar

Figures

Abstract

Introduction

Methods

Meta-analysis

Overview.

Study selection.

Data extraction.

Data analyses.

Parameterization of fN2Onit in other models

Description of the N2O simulation model

Sensitivity simulations

Results and discussion

Sensitivity of N2O flux to fN2Onit

Application of pH-based parameterization

Impacts on global N2O budget

Concluding remarks

Supporting information

S1 Table. PRISMA 2009 check list.

S2 Table. List of search terms.

S3 Table. Data extracted from the literature and used by the meta-analysis.

S1 Fig. Map of soil pH (Global Soil Data Task, International Geosphere-Biosphere Program).

S2 Fig. Seasonal change in total N2O emission simulated by the VISIT model.

References

Parameterization of fN₂O_nit in other models

Description of the N₂O simulation model

Sensitivity of N₂O flux to fN₂O_nit

Impacts on global N₂O budget

S2 Fig. Seasonal change in total N₂O emission simulated by the VISIT model.