Microbial regulation of terrestrial nitrous oxide formation ...

FEMS Microbiology Reviews, fuv021, 39, 2015, 729–749 doi: 10.1093/femsre/fuv021 Advance Access Publication Date: 1 May 2015 Review Article

REVIEW ARTICLE

Microbial regulation of terrestrial nitrous oxide formation: understanding the biological pathways for prediction of emission rates Hang-Wei Hu1,2 , Deli Chen2 and Ji-Zheng He1,2,∗ 1

State Key Laboratory of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China and 2 Faculty of Veterinary and Agricultural Sciences, The University of Melbourne, Parkville Campus, Victoria 3010, Australia ∗ Corresponding author: State Key Laboratory of Urban and Regional Ecology, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing 100085, China. Tel: (+61)-3-90358890; E-mail: [email protected] One sentence summary: This review summarizes the major microbial pathways of soil N2 O production, and key environmental factors modulating their relative contributions, and further proposes to use a combination of state-of-the-art approaches for better source partitioning and incorporation of microbial datasets to achieve better predictive ecosystem models. Editor: Jan Roelof van der Meer

ABSTRACT The continuous increase of the greenhouse gas nitrous oxide (N2 O) in the atmosphere due to increasing anthropogenic nitrogen input in agriculture has become a global concern. In recent years, identification of the microbial assemblages responsible for soil N2 O production has substantially advanced with the development of molecular technologies and the discoveries of novel functional guilds and new types of metabolism. However, few practical tools are available to effectively reduce in situ soil N2 O flux. Combating the negative impacts of increasing N2 O fluxes poses considerable challenges and will be ineffective without successfully incorporating microbially regulated N2 O processes into ecosystem modeling and mitigation strategies. Here, we synthesize the latest knowledge of (i) the key microbial pathways regulating N2 O production and consumption processes in terrestrial ecosystems and the critical environmental factors influencing their occurrence, and (ii) the relative contributions of major biological pathways to soil N2 O emissions by analyzing available natural isotopic signatures of N2 O and by using stable isotope enrichment and inhibition techniques. We argue that it is urgently necessary to incorporate microbial traits into biogeochemical ecosystem modeling in order to increase the estimation reliability of N2 O emissions. We further propose a molecular methodology oriented framework from gene to ecosystem scales for more robust prediction and mitigation of future N2 O emissions. Keywords: nitrous oxide; ammonia oxidation; nitrifier denitrification; heterotrophic denitrification; climate change; modeling

INTRODUCTION: GLOBAL CONCERNS ABOUT THE INCREASING TERRESTRIAL N2 O EMISSIONS One of the major challenges of sustainable ecosystem management is to mitigate the negative effects of global climate

changes caused by steadily increasing atmospheric greenhouse gas (GHG) emissions (Trivedi, Anderson and Singh 2013). N2 O is a potent GHG with ∼300-fold greater warming potential than CO2 on a per molecule basis, and is involved in destruction of the stratospheric ozone layer (Ravishankara, Daniel and Portmann 2009). Globally, soil ecosystems constitute the largest

Received: 28 September 2014; Accepted: 9 April 2015 C FEMS 2015. All rights reserved. For permissions, please e-mail: [email protected]

729

730

FEMS Microbiology Reviews, 2015, Vol. 39, No. 5

Figure 1. Global N2 O emissions from various sources in 1995, 2005 and 2030. ‘Other energy sources’ here includes waste combustion, fugitives from solid fuels, natural gas and oil systems. ‘Other industrial processes sources’ includes metal production, solvent and other product use. ‘Other agricultural sources’ includes field burning of agricultural residues and prescribed burning of savannas. ‘Other waste sources’ includes miscellaneous waste handling processes. Adapted from US EPA, (2012).

source of N2 O emissions (estimated at 6.8 Tg N2 O-N year−1 ), comprising approximately 65% of the total N2 O emitted into the atmosphere, with 4.2 Tg N2 O-N year−1 derived from nitrogen fertilization and indirect emissions, 2.1 Tg N2 O-N year−1 arising from manure management and 0.5 Tg N2 O-N year−1 introduced through biomass burning (IPCC 2007). Other important N2 O sources include ocean, estuaries and freshwater habitats and wastewater treatment plants (Schreiber et al. 2012). In recent years, the excessive use of nitrogen-based fertilizers (ca. 140 Tg N year−1 ), due to the growing food demand of the human population and agricultural expansion, has greatly contributed to the conspicuous elevation in atmospheric N2 O concentrations, from pre-industrial levels of 270 ppbv to current levels approaching 324 ppbv (Galloway et al. 2008). Generally, for every 1000 kg of applied nitrogen fertilizers, it is estimated that around 10–50 kg of nitrogen will be lost as N2 O from soil, and the amounts of N2 O emissions increase exponentially relative to the increasing nitrogen inputs (Shcherbak, Millar and Robertson 2014). Given that farmlands and fertilizer application are predicted to increase by 35–60% before 2030 (IPCC 2007), global N2 O concentrations are likely to continuously rise in the coming decades (Reay et al. 2012), and it is expected that agricultural soils will contribute up to 59% of total N2 O emissions in 2030 (Fig. 1). To enable more effective mitigation strategies to counteract the steady increase in N2 O loadings, it is necessary to better understand the underlying mechanisms leading to soil N2 O formation. A large body of ecological studies has emphasized the central role of soil microbes in regulating the major processes of nitrogen transformations and N2 O emissions (Leininger et al. 2006; Zhang et al. 2010; Baggs 2011; Hu, Xu and He 2014b). The abundance, diversity, structure, physiology and biogeographical patterns of N2 O-producing and -consuming organisms, as well as their interactions have become a focus of microbial ecology (Singh et al. 2010; Barberan et al. 2012). However, most fieldbased and laboratory experiments to date have focused either on estimation and simulation of agricultural N2 O fluxes (Chen et al. 2008a; Reay et al. 2012), temporal and spatial dynamics of N2 O across various ecosystems (Yamulki et al. 2001), or on impact of environmental factors on N2 O fluxes (Chen et al. 2010b; Dai et al. 2013; Nemeth, Wagner-Riddle and Dunfield 2014). Relatively little effort was directed to link microbial pathways, particularly microbial communities, with rates of N2 O fluxes in soil ecosystems (Ma et al. 2008; Cuhel et al. 2010; Singh et al. 2010). On the other hand, emerging biogeochemical ecosystem models at field and regional scales have recognized the great potential of incorporating microbial emission factors and biochemical mechanisms into models to improve the predictive power (Singh et al. 2010; Trivedi, Anderson and Singh 2013), but

are facing difficulties in parameterizing and integrating these microbial ‘codes’ into the modeling efforts, because of the lack of quantitative correlations between microbial data and rates of N2 O fluxes. Therefore, despite their significant importance in regulating N2 O formation, soil microbial activities still remain a ‘black box’ in estimating biogeochemical nitrogen turnover and designing practical mitigation options for N2 O emissions. Accurate identification of these critical microbially mediated N2 Oregulating processes, and directly linking microbial metabolic activity with soil N2 O emissions might be a prerequisite for developing the next generation of microbially oriented ecosystem models. To date, there have been a number of reviews on N2 O production and consumption mechanisms in diverse environments. Most of these reviews, however, have only treated specific microbial pathways (Wrage et al. 2001; Chapuis-Lardy et al. 2007), soil NO transformation (Pilegaard 2013; Medinets et al. 2015), N2 O source partitioning technologies (Baggs 2008) or traditional N2 O simulation models (Chen et al. 2008a), and are primarily focusing on ecosystems other than soils (Schreiber et al. 2012). There is a lack of efforts to incorporate knowledge on the potential contribution of archaea to soil N2 O formation, to differentiate the relative importance of microbial pathways under different environmental conditions by using multidisciplinary approaches, and to improve future N2 O modeling by incorporating microbial dynamics. Therefore, this review attempts to identify the major microbial pathways of N2 O production from soils and the key environmental factors modulating the relative occurrence and significance of functional microbes, to synthesize a combination of state-of-the-art multidisciplinary approaches to effectively discriminate between different microbial pathways, and to propose reasons why these microbial datasets should be urgently included into ecosystem models to better estimate the future terrestrial N2 O budget. This integration will be critical towards a more confident simulation of future GHG emissions from terrestrial ecosystems, and for the development of more targeted mitigation strategies for use in agricultural practices.

OVERVIEW OF THE MAJOR BIOLOGICAL PATHWAYS FOR N2 O EMISSIONS The recent progress in culture-dependent and laboratory microcosm studies, coupled with rapid development of DNA/RNAstable isotope probing (SIP) techniques and high-throughput sequencing approaches, has unraveled the previously unknown involvement of a wide variety of microorganisms in soil N2 O emissions (Baggs 2011; Schreiber et al. 2012 and references

Hu et al.

Figure 2. Simplified schematic representations of the major microbial pathways and microbes for the global N2 O production and nitrogen cycling in soil ecosystems. The multiple pathways include ammonia (hydroxylamine) oxidation and nitrifier denitrification (performed by AOA and AOB), nitrite oxidation (by NOB), heterotrophic denitrification (by heterotrophic bacteria), anammox (by anaerobic ammonia oxidizers) and dissimilatory nitrate reduction to ammonium (DNRA, by unknown microorganisms). Different microbial groups and pathways are indicated clearly by different colors. Adapted from Schreiber et al. (2012), with permission.

therein). As summarized in Fig. 2, nearly all microbes and pathways known to be involved in biogeochemical nitrogen cycling have the potential to catalyze the production of N2 O, and these processes are interrelated and interacting to share intermediates or products. The key multiple pathways of N2 O production and consumption include ammonia (hydroxylamine) oxidation, nitrifier denitrification (Wrage et al. 2001), nitrite oxidation, heterotrophic denitrification, anaerobic ammonium oxidation (anammox) and dissimilatory nitrate reduction to ammonium (DNRA, or nitrate ammonification), with each process modulated by specialized groups of microbial assemblages (Fig. 2). In spite of the complex and multiple routes for N2 O formation, the nitrification-related pathways (including ammonia oxidation and nitrifier denitrification) and heterotrophic denitrification have been established as the most predominant sources of ¨ N2 O emissions from soil ecosystems (Godde and Conrad 1999; Wrage et al. 2005; Zhu et al. 2013; Shcherbak, Millar and Robertson 2014). For instance, nitrification-related pathways have often been assumed to be the principal sources of N2 O in waterlimited soils, while heterotrophic denitrification is largely responsible for soil N2 O emissions at high water contents (Mathieu et al. 2006). In this review, we will mainly focus on the nitrifying and denitrifying communities involved in the processes of N2 O formation, and also briefly introduce other microbial pathways which might be occasionally important in particular cases.

Heterotrophic denitrification Heterotrophic denitrification, as a multistep reaction performed primarily by a variety of bacteria, has been known as a major microbial respiratory process that reduces oxidized mineral forms of N (NO3 − and NO2 − ) to the gaseous products NO, N2 O and N2 under oxygen-limited conditions (Fig. 2; Philippot, Hallin and Schloter 2007). However, heterotrophic denitrification in the presence of O2 has been also reported in physiological studies of pure denitrifier strains isolated from soils and sediments (Patureau et al. 2000), and could even occur in anaerobic microsites of aerated arid or semiarid soils caused by intensive res-

731

piration (Abed et al. 2013) and in marine sediments (Gao et al. 2010). The sequential processes of bacterial denitrification (NO3 − → NO2 − → NO → N2 O → N2 ) are regulated by divergent reductases encoded by distinct functional genes (Table S1, Supporting Information). For instance, the first step (NO3 − → NO2 − ) is catalyzed by the narG or napA genes encoding the nitrate reductase; the second step (NO2 − → NO) is catalyzed by the nirK or nirS genes encoding two entirely different types of nitrite reductase; the third step leading to N2 O formation (NO → N2 O) is mediated by the cnorB or qnorB genes encoding the nitric oxide reductase and the last step, reduction of N2 O (N2 O → N2 ) by the nosZ gene encoding the nitrous oxide reductase, is the only known microbial process which could reduce N2 O to N2 in the biosphere (Philippot, Hallin and Schloter 2007; Jones et al. 2013). There is evidence that 30–80% of the N2 O produced from deeper soil layers may be reduced to N2 before diffusion into the atmosphere (Clough, Sherlock and Rolston 2005). In addition to bacteria, fungi could also play vital roles as key producers of N2 O via heterotrophic denitrification in a wide variety of soils (Thamdrup 2012), particularly in semi-arid regions (Crenshaw et al. 2008), tropical arable peat (Yanai et al. 2007) and forest and grassland ecosystems (Laughlin and Stevens 2002; PrendergastMiller, Baggs and Johnson 2011). The fungal denitrification system comprises a copper-containing nitrite reductase together with a cytochrome P450 nitric oxide reductase to reduce nitrite to N2 O (Shoun et al. 2012). The primary product of fungal denitrification is N2 O, because fungi generally lack the nosZ gene to further reduce N2 O to N2 (Sutka et al. 2008; Baggs 2011; Philippot et al. 2011), but their in situ contribution to N2 O has yet to be directly measured. Among the denitrifying genes, the nirS, nirK and nosZ genes have received much more scientific interest relative to other denitrifying genes (e.g. napA, narG and cnorB) in environmental investigations and laboratory microcosms, and their abundance, structure, expression and metabolic activity could serve as potential indicators for denitrification-derived N2 O fluxes in soils (Morales, Cosart and Holben 2010). Soils with high ratios of (nirK + nirS)/nosZ gene copies might be associated with a high capacity for N2 O production and thus high levels of N2 O emissions, or vice versa. However, it should be noted that the copper-containing nitrite reductase (encoded by nirK) and the heme-containing cytochrome cd1 nitrite reductase (encoded by nirS) are mutually exclusive and have never been found in the same cell (Zumft 1997). Nearly one-third of nirS- or nirK-containing denitrifiers, such as Agrobacterium tumefaciens and some strains within the genus Thauera, lack the nosZ gene (Philippot et al. 2011; Bakken et al. 2012), and therefore these nosZ-lacking microbes do not hold the genetic capacity to reduce N2 O. Molecular investigations also confirmed a much lower abundance of nosZ relative to other denitrifying genes (Bru et al. 2011), and the copy numbers of nirS and nirK can exceed that of nosZ by an order of magnitude in various soil environments (Philippot et al. 2011). Considering that the nosZ gene is the only known biotic sink for N2 O, its quantities and activity could represent an independent indicator for the measured N2 O concentration above soil surface. For example, N2 O fluxes and ratios of N2 O/(N2 + N2 O) were inversely correlated with the abundance and transcript copy numbers of nosZ in field-scale and microcosm studies (Philippot et al. 2009; Harter et al. 2013), and were regulated by community structures of the nosZ-containing denitrifiers in agricultural fields (Cavigelli and Robertson 2000). Phylogenetic analysis suggested that the denitrifying organisms belonging to the Bacteroidetes, Gemmatimonadetes and Delta-proteobacteria constitute a

732


significant proportion of the N2 O-reducing (i.e. nosZ-containing) bacteria in soil ecosystems (Jones et al. 2013). Given that N2 O is either an intermediate or a terminal product of heterotrophic denitrification, soil could be either a source or a sink of N2 O depending on the relative metabolic activity of the N2 O-producing and -reducing enzymes. Previous studies suggested that heterotrophic denitrification can produce large amounts of N2 O in soils with low pH and low O2 levels (Anderson et al. 1993; Wrage et al. 2001), whereas net consumption of N2 O was frequently reported in grasslands and forests containing low mineral nitrogen concentrations (Chapuis-Lardy et al. 2007) and high carbon availability (Kool et al. 2010). In line with these findings, expression of the nosZ gene was observed to be inhibited by O2 and low pH in laboratory microcosms and batch cultures (Bergaust et al. 2010; Schreiber et al. 2012), and was more sensitive to O2 compared with expression of other denitrifying genes (e.g. nirS, nirK, cnorB and narG) (Chapuis-Lardy et al. 2007). More inhibition of the nosZ gene relative to other denitrifying genes will lead to accumulation of N2 O in the soil matrix, and therefore oxygen-limited or acidic soils might represent a hotspot for N2 O production from heterotrophic denitrification, whereas aerobic, low-nitrogen and high-carbon conditions are expected to favor N2 O consumption. However, the degree to which different denitrifying genes are governed by multiple environmental factors remains unknown, and a consistent quantitative or empirical relationship between activity of denitrifying genes and rates of N2 O emission is still missing. Discrimination between gross N2 O production and gross N2 O consumption processes via 15 N isotope tracer techniques merits further study, in order to simultaneously delineate the production activity of N2 O (by specifically targeting nirK and nirS) and the consumption activity (by specifically targeting nosZ).

Nitrification-related pathways (including ammonia oxidation and nitrifier denitrification) Nitrification is the aerobic oxidation of ammonia to nitrate via nitrite (NH3 → NH2 OH/HNO → NO2 − → NO3 − ), with each step performed by a specialized group of prokaryotes. The first step (NH3 → NH2 OH/HNO → NO2 − ), ammonia oxidation, catalyzed by the amoA gene encoding the ammonia monooxygenase (AMO) (Table S1, Supporting Information), is known to be performed by two distinct types of microbes: ammonia-oxidizing bacteria (AOB), belonging to two monophyletic groups within β- or γ -proteobacteria (Purkhold et al. 2000), and ammonia-oxidizing archaea (AOA), affiliated within the newly described Thaumarchaeota phylum (Brochier-Armanet et al. 2008). The second step (NO2 − → NO3 − ) is regulated by the nxrB gene encoding the nitrite oxidoreductase within nitrite-oxidizing bacteria (NOB) (Freitag, Rudert and Bock 1987). Ammonia oxidation is believed to be the rate-limiting step for the whole nitrification processes (Kowalchuk and Stephen 2001), and thus is critical for production of the nitrification-originated N2 O. It has been estimated that ammonia oxidation could contribute up to 80% of soil N2 O emissions, depending on particular soil ecosystem types, tem¨ perature and moisture contents (Godde and Conrad 1999). In a classical view, N2 O is produced as a byproduct of nitrification via chemical decomposition of hydroxylamine (NH2 OH), nitroxyl hydride (HNO) or NO2 − (intermediates or end products of ammonia oxidation) with organic and inorganic compounds at low pH (80–90% WFPS (Braker and Conrad 2011; Huang et al. 2014). However, despite the negative correlations between WFPS and O2 levels, the effect of WFPS on the relative contributions of nitrification and denitrification to N2 O emissions is much more complex, and cannot be as clearly depicted as that for O2 (Fig. 3). Soil water content not only determines the availability of O2 , but also influences diffusion and transport of nutrients within the soil matrix and the metabolic activity of microbial cells (Hu et al. 2014a), which could confound the relationships between WFPS and rates of N2 O emissions.

735

Soil pH A substantial number of ecological studies have demonstrated that abundance and structure of nitrogen-cycling genes, and rates of nitrification and denitrification are strongly regulated by soil pH (He et al. 2007, Shen et al. 2008; Liu et al. 2010; Bakken et al. 2012; Hu et al. 2013), and product ratios of N2 O/(N2 + N2 O) have a significantly negative relationship with soil pH within the normal range from pH 5 to 8 in agricultural soils (Chapuis-Lardy et al. 2007; Bakken et al. 2012). Global meta-analysis of field experiments has also revealed that the amounts of N2 O substantially increase in soils with lower pH values (Shcherbak, Millar and Robertson 2014). By contrast, under alkaline conditions, N2 is more favored as the end product of denitrification and thus less accumulation of N2 O will be observed (Richardson et al. 2009). Considering that acidic soils occupy around 30% of the Earth’s ice-free lands, it was suggested that liming of slightly acidic soils might be a way to alleviate global N2 O emissions (Bakken et al. 2012), but this needs rigorous verifications in field experiments. Possible mechanisms for positive effects of acidity on N2 O emissions have been attributed to the inhibition of the nitrous oxide reductase through affecting the enzyme assembly in the periplasm (which is the location of the most functional enzymes) under acidic conditions (Wrage et al. 2001; Bakken et al. 2012), or due to different magnitudes of changes in different microbial pathways (Fig. 3). For example, regression analysis of 12 soil microcosms with isotope tracing approaches found that soil pH was a significant predictor of the relative contributions of biological pathways, with nitrifier denitrification being positively related to pH, and heterotrophic denitrification decreased with increasing pH (Fig. 3; Kool et al. 2010). These findings are supported by observations that the reductases for nitrate, nitrite and nitric oxide are more active at pH < 7 (Richardson et al. 2009), and expression of the nosZ gene was observed to be inhibited by low pH in laboratory microcosms and batch cultures (Bergaust et al. 2010; Bakken et al. 2012). Moreover, N2 O-producing activity of denitrifier strains was also found to increase at low pH, with more N2 O produced under acidic than under alkaline pH (Mothapo et al. 2013). There have been numerous similar observations of pH effects on denitrification and the denitrification-derived N2 O production, indicating a generality of the phenomenon in soils (Thomson et al. 2012 and references therein). Contributions of ammonia oxidation to N2 O emissions as affected by soil pH have not been reported, but by compiling data from more than 100 isotope dilution studies conducted in a broad range of terrestrial ecosystems, a slightly negative correlation between gross nitrification rates and soil pH was observed (Booth, Stark and Rastetter 2005). If nitrification rates are directly related to rates of N2 O emission as observed in wastewater treatment plants (Ni et al. 2013a), a decreasing tendency of ammonia oxidation contribution to N2 O along the increasing soil pH gradient can be postulated (Fig. 3). Moreover, growing evidence from laboratory incubations and environmental investigations ascribes nitrification activity in acidic soils mainly to AOA (Zhang et al. 2012; Hu et al. 2014a; Hu, Xu and He 2014b), and a recent physiological study found that, among the seven examined ammonia oxidizers, the AOA strains cultivated from acidic soils had the highest N2 O production rates (Jung et al. 2014). Therefore, it is reasonable to hypothesize that in strongly acidic soils, AOA may substantially contribute to N2 O production via the ammonia oxidation pathway, due to their high affinity for ammonia substrates in ammonia-poor acidic soils (He, Hu and Zhang 2012). The increasing trend of nitrifier denitrification with the increasing soil pH requires more in-depth

736


examination by performing isotope tracing approaches, and a mechanistic explanation is lacking due to the poor understanding of the nitrifier denitrification pathway for both AOA and AOB. Apart from the above biological reactions, chemical production of N2 O from inorganic nitrogen compounds, i.e. chemodenitrification, was also reported to be favored in acidic soils (pH < 5) with high nitrogen fertilizer inputs (Fig. 3; Van Cleemput 1998; Braker and Conrad 2011), but in soils with high pH values, the N2 O derived from chemo-denitrification constituted only 0.1–1.3% of total N2 O production (Zhu et al. 2013).

Other factors Beyond the significant impact of O2 , WFPS and pH, other factors such as fertilizer types, soil types and climatic scenarios have also been reported to strongly affect N2 O emissions in some cases. For example, meta-analysis of field experiments and modeling suggested a greater loss of N2 O from urea than from (NH4 )2 SO4 fertilizers treatments (Bouwman, Boumans and Batjes 2002), which was supported by similar observations in laboratory incubations of loam and sandy loam soils (Zhu et al. 2013). As for the soil types, nitrification-related N2 O production was found to be significantly higher in grassland and arable soils than in forest soils (Kool et al. 2010). Nitrifier denitrification might be a major N2 O contributor in semi-arid and Arctic soils, whereas heterotrophic denitrification dominated N2 O production in temperate or forest soils with high organic matter (Sanchez-Martin et al. 2008). In addition, studies on responses of soil N2 O emissions and microorganisms to future climatic scenarios are also emerging. Increased N2 O emissions have been reported in soils with excessive reactive nitrogen under elevated CO2 (Baggs et al. 2003), and in field/microcosm measurements under simulated warming and elevated CO2 (Elberling, Christiansen and Hansen 2010; Van Groeningen, Osenberg and Hungate 2011; Reay et al. 2012). Although several studies have ascribed the shifts in N2 O fluxes and nitrogen transformations under elevated CO2 and warming to changes in AOB abundance (Horz et al. 2004) or community structures (Avrahami, Liesack and Conrad 2003), the detailed microbial basis remains less well documented (Singh et al. 2010). Overall, all of these factors confound the relationships between microbial community and surface fluxes of N2 O, and add to the difficulty to predict their shifts based on a single environmental factor. More research is ultimately needed to elucidate how soil variables interact to control N2 O emissions from complex soil environments.

TECHNIQUES AND APPROACHES TO DIFFERENTIATE MICROBIAL PATHWAYS OF N2 O IN SOILS Because multiple N2 O-producing and -reducing pathways are simultaneously involved in different micro-environments in the same soil, it remains a great challenge to allocate their relative contributions based merely on basic soil characteristics. Attempts have been made to link rates of ammonia oxidation, heterotrophic denitrification and soil N2 O fluxes with the abundance, community composition, and expression of the key nitrogen-cycling functional genes like amoA, hao, nirK, nirS, narG, norB, napA and nosZ (Table S1, Supporting Information) in various ecosystems by using molecular methodologies (Philippot et al. 2002; Balser and Firestone 2005; Ma et al. 2008; Avrahami and Bohannan 2009; Dai et al. 2013). For instance, the distribution of nosZ-containing denitrifiers in grassland fields was found

to significantly correlate with potential N2 O emissions and the ratio of N2 O/(N2 O + N2 ) (Philippot et al. 2009), which was also strongly linked with the gene copy numbers of nirS, napA and narG in another grassland investigation (Cuhel et al. 2010). Community diversity of ammonia oxidizers can have strong relationships with N2 O emission rates, which was also tightly related to the ammonia-oxidizing community compositions (Avrahami and Bohannan 2009). However, correlative evidence and genetic potential cannot fully explain the metabolic activity of multiple N2 O-relevant functional microbes unless combined with powerful tools to resolve the relative importance of microbial sources (Huang et al. 2014). Therefore, a multidisciplinary approach by appropriately combining different source partitioning techniques with molecular approaches is required to resolve the question of the relative importance of various sources and sinks for soil-emitted N2 O under diverse scenarios.

Natural abundance isotopic signatures The N2 O isotopic species occur in natural environments with three types: 14 N15 N16 O (0.37%), 15 N14 N16 O (0.37%) and 14 N14 N16 O (>99%). The intramolecular distribution of nitrogen isotopes in the internal (α) and external (β) positions of the linear N2 O (Nβ Nα -O) molecule (e.g. 14 N15 NO versus 15 N14 NO) is defined as site preference (SP = δ 15 Nα − δ 15 Nβ ) (Toyoda et al. 2005). The SP values remain constant during the course of N2 O production, and was presumed to be determined by the reaction steps leading to N2 O formation (Toyoda et al. 2005; Sutka et al. 2006) and the relative importance of N2 O production and consumption processes (Jinuntuya-Nortman et al. 2008), and to be independent of the isotopologue characteristics of the substrates (Sutka et al. 2006; Jung et al. 2014). This powerful tool opens up the avenues to distinguish between different microbial processes in a noninvasive way, and has been widely used to quantitatively interpret N2 O sources (Mathieu et al. 2006; Well et al. 2008; Kato et al. 2013). In order to evaluate the effectiveness of isotopic signatures in source allocation, we carried out a detailed comparative analysis of the currently available data to explore the processspecific isotope effects (Fig. 4). This comparison cautions against oversimplification of the discrepancies between batch cultures, microcosms and field observations, because soil ammonia concentrations, dissolved oxygen levels, pH and microbial cell density are variable across the studies (Fig. 4, Tables S2 and S3, Supporting Information). Fig. 4 illustrates that N2 O produced from different pathways of enriched or pure cultures spanned a wide spectrum of the SP values. For example, the SP values for AOA ammonia oxidation ranging from +13.1 to +34.0 are comparable to those for AOB ammonia oxidation ranging from +14.9 to +36.3. The SP for methanotrophic ammonia oxidation showed slightly higher values ranging from +30.8 to +35.6, but still fell within the ranges of AOA and AOB ammonia oxidation (Fig. 4). By contrast, the SP for AOB nitrifier denitrification had negative values ranging from −19.9 to +0.1, but the SP values for AOA nitrifier denitrification are not available yet due to the supposition of AOA’s inability to denitrify (Jung et al. 2014). It is notable that the SP values for fungal denitrification varied from +22.8 to +40.0, which overlap with AOA and AOB ammonia oxidation. Bacterial denitrification showed significantly lower SP values (ranging from −7.5 to +27.7) compared with fungal denitrification, but these values are difficult to distinguish from AOB nitrifier denitrification. Therefore, the SP tools are useful in discriminating N2 O produced by ammonia oxidation and by nitrifer denitrifiation in soils where N2 O emissions are dominated by nitrification,

Hu et al.

737

Figure 4. Isotopic signatures (including site preference, δ 15 Nbulk -N2 O, and δ 18 O-H2 O) of N2 O from pure or enriched cultures of AOA, AOB, methane oxidizers, bacterial denitrifiers, and fungal denitrifiers and from soil microcosms and in situ field measurements. Detailed information about these data and the relevant references are listed in Tables S2 and S3 (Supporting Information). (Abbreviations: AO, ammonia oxidation; ND, nitrifier denitrification; AOA, ammonia-oxidizing archaea; AOB, ammonia-oxidizing bacteria).

but are ineffective when fungal denitrification is co-occurring (Fig. 4). In this case, the fungal inhibitor cycloheximide and the bacterial inhibitor streptomycin might be selected to differentiate between N2 O produced by bacteria and fungi (Sutka et al. 2008). In parallel with these studies, efforts have also been directed to capture the isotopomeric compositions of N2 O from a variety of soil environments (Fig. 4 and Table S3, Supporting Information), which could be compared with those from enriched or pure cultures of microorganisms to infer the dominant pathways. For example, the SP values of N2 O from alpine meadow and shrub soils were estimated as +33.7 and +30.1, respectively, which are close to the SP values reported for ammonia oxidation and fungal denitrification (Kato et al. 2013). As shown in Fig. 4, most of the reported SP values from soils have a wide span of positive values from +2.2 to +41.8, except for one temperate grassland soil with −16.8, and therefore ammonia oxidation or fungal/bacterial denitrification might be the dominant N2 O pathways in these soils. Moreover, the SP values from Pacific waters ranged from +8 to +10 (Sutka et al. 2004), which was attributed to a predominant N2 O source from AOA ammonia oxidation (Santoro et al. 2011), while N2 O emitted from irrigated soils had SP values ranging from −5.4 to +0.1, pointing to a dominant role of nitrifier denitrification or bacterial denitrification (Huang et al. 2014). Therefore, site preference might be a useful tool in some cases, but not a strictly unique parameter as previously reported, because our synthesis indicates obvious overlapping of the SP values between different cultures and pathways (Fig. 4 and Table S2, Supporting Information). In addition to the SP values, other isotopic signatures including δ 15 Nbulk -N2 O and δ 18 O-N2 O, are also emerging indicators used for N2 O source allocation (Jung et al. 2014). However, unlike the SP values, the δ 15 Nbulk -N2 O and δ 18 O-N2 O values are generally regulated by (1) isotopic signatures of the substrates (e.g. NH4 + , NO2 − and NO3 − ); (2) isotopic fractionation characteristics of individual microbial pathways; (3) isotope exchange between different moieties in the medium (e.g. oxygen atom exchange between O2 , H2 O and intermediates of nitrification and denitrification) (Toyoda et al. 2005) and (4) N2 O consumption pro-

cesses. As shown in Fig. 4, the δ 15 Nbulk -N2 O values span a wide range, and overlap between AOA ammonia oxidation (−35.5– +8.7), AOB ammonia oxidation (−28.6–+0.3), AOB nitrifier denitrification (−34–+56.9), methanotrophic ammonia oxidation (+0.8–+3.4), bacterial denitrification (−37.2–+36.7) and fungal denitrification (−20.0–−2.6). The δ 15 Nbulk -N2 O values obtained from soils varied broadly from −91.6 in a tropical forest soil to −3.4 in a temperate urea-amended grassland soil (Fig. 4), while positive values of +5–+10 were reported from surface ocean and troposphere (Jung et al. 2014). δ 15 Nbulk -N2 O values lower than −40 do not overlap with any values from AOA, AOB or denitrifiers (Fig. 4), indicating that some unknown microorganisms or processes are responsible for emitting N2 O in these soils, or the ranges of isotopic signatures for existing pathways are not fully explored. δ 18 O-N2 O values from the enriched and pure cultures overlapped in a narrow range from +15.6 to +57.3, except for those from AOB nitrifier denitrification with much lower values ranging from −8.4 to +10.8 (Fig. 4). As a result, δ 18 O-N2 O values could provide additional information to aid the distinction of nitrifier denitrification from other processes. In natural environment, the δ 18 O-N2 O values varied from +9.6 in tropical forest soils to +48.4 in temperate arable soils (Fig. 4), and those from the surface oceans to troposphere ranged between +20 and +40 (Jung et al. 2014). It is hard to ascertain the microbial N2 O sources based merely on these δ 18 O-N2 O values. It was previously thought that N2 O produced from nitrification is more depleted (more negative values) in δ 15 Nbulk -N2 O and δ 18 O-N2 O compared with that produced from heterotrophic denitrification (Baggs 2008), but these results are questioned by our analysis which suggested difficulties in using them alone for source differentiation. These natural abundance isotopic signatures are not as process dependent as previously expected (Frame and Casciotti 2010), and are not constant or sufficiently differentiable to resolve microbial processes (Park et al. 2011). Comparative analysis of these isotopic signatures, such as SP versus δ 18 O-N2 O, or SP versus δ 15 Nbulk -N2 O, might be a powerful way to resolve the different processes (Fig. 5), but can only indicate whether a process is occurring or not (Baggs 2008). Quantifying their

738


Figure 5. Joint use of site preference versus δ 18 O-H2 O, or site preference versus δ 15 Nbulk -N2 O to infer sources of soil N2 O. The boxes indicate the expected ranges for different microbial pathways based on the available data from enriched or pure cultures of nitrifiers and denitrifiers in Table S2 (Supporting Information). The open triangles are examples of isotopic signatures from soil microcosms and in situ field measurements (Table S3, Supporting Information). If the measured isotopic signature locates in only one box, the soil N2 O production is dominated by the corresponding pathway; if in the overlapped area of two or three boxes, multiple pathways are occurring simultaneously; if outside of all the boxes, some unknown pathways might function or the range of isotopic signatures for existing pathways are not fully explored. More in situ field measurements by employing inhibitors or substrate omissions for specific pathways are required to verify that these natural abundance isotopic signatures are caused by particular pathways.

accurate contributions is largely restricted by the lack of isotopic signatures from representative pure cultures and different ecosystems, experimental conditions and climatic types. Moreover, soil microbial communities are far more diverse than the characterized pure cultures, which add difficulties to extrapolate the laboratory findings to in situ soil microorganisms with the same functional traits.

15

N, 18 O and 13 C isotope enrichment techniques

In addition to the natural isotopic signatures, stable isotope enrichment approaches have been also developed to identify N2 O sources following application of 15 N-labeled fertilizers in short-term experiments. Generally, denitrification-derived N2 O is quantified following supply of 15 N-NO3 − , while nitrificationderived N2 O following supply of 15 N-NH4 + (Baggs 2008). The reduction of N2 O to N2 can also be quantified by determining 15 N in N2 after supply of 15 N-NO3 − (Stevens and Laughlin 1998). For example, applications of 14 NH4 15 NO3 and 15 NH4 14 NO3 have been used to determine the relative contributions of nitrification and denitrification to N2 O production (Baggs and Blum 2004). Enrichment of 15 N at the internal (α) position of the N2 O molecule is thought to be from nitrification, and the N2 O from denitrification has a lower SP value (∼0) (Santoro et al. 2011). While the single 15 N-isotopic enrichment technique could differentiate N2 O production between nitrification and denitrification, it does not enable discrimination of the N2 O from nitrifier denitrification or N2 O as a byproduct of ammonia oxidation (Wrage et al. 2005), particularly under oxygen-limiting conditions, where ammonia oxidation, nitrifier denitrification and heterotrophic denitrification are occurring simultaneously. A dual labeling technique, including application of 15 N- labeled NH4 + or NO3 − and 18 O-labeled H2 O followed by quantification of 15 N and 18 O signatures (Wrage et al. 2005; Baggs 2008), has been applied to distinguish between nitrifier denitrification and ammonia oxidation in mixed population systems, assuming that oxygen sources acquired during different nitrification steps are also different. As shown in Fig. 6, for the nitrifier denitrification pathway, the first step of oxidizing ammonia to NH2 OH/HNO incorporates one oxygen atom from

Figure 6. Schematic representation of the different oxygen atoms incorporated during the predominant microbial pathways (including ammonia oxidation, nitrifier denitrification and heterotrophic denitrification) of N2 O production in soils. Adapted from Huang et al. (2014), with permission.

O2 , and the oxidation of NH2 OH/HNO to NO2 − incorporates a second oxygen atom from H2 O, leading to a 50% dependence of δ 18 O-N2 O on δ 18 O-H2 O assuming no oxygen exchange between NO2 − and H2 O (Santoro et al. 2011). However, theoretically 100% of oxygen originates from O2 for the ammonia oxidation pathway. For the heterotrophic denitrification pathway, one-third of oxygen originates from O2 and two-thirds are from H2 O (Fig. 6). However, it has been recognized that the potential oxygen exchange between H2 O and intermediates during N2 O production could complicate the data interpretation (Wrage et al. 2005). This dual labeling approach was further refined by considering the 18 O-NO3 − to quantify the exchange of oxygen atoms between nitrogen oxides and H2 O during nitrification and denitrification in soils (Kool et al. 2010). Although the 15 N and 18 O enrichment techniques and natural isotope signatures have shown advantages in partitioning the microbial N2 O pathways, they proved powerless in providing critical information regarding the metabolic activity of the

Hu et al.

functional microorganisms. In fact, soil nitrogen cycling and N2 O emissions are maintained and regulated by the functionally active assemblages of microorganisms (Hu et al. 2014a), but in natural soil environments up to 80% of microbial species are expected to be metabolically inactive or in a dormant state (Lennon and Jones 2011). Designing more targeted mitigation strategies therefore relies on partitioning the relative contributions of microbial pathways together with an understanding of the functionally active microbes directly contributing to the fundamental processes. These efforts could be substantiated by combining 15 N-18 O- enrichment techniques with 13 C-CO2 DNA/RNA SIP which may allow the functionally active players involved in ammonia oxidation to be identified (Zhang et al. 2012; Hu et al. 2014a) possibly by indirectly linking them with N2 O production. However, the heterotrophic growth of ammonia oxidizers could not be detected by DNA-SIP (Pratscher, Dumont and Conrad 2011), activity of N2 O production is not necessarily synonymous with cellular growth rates (Shaw et al. 2006) and 13 C-DNA/RNASIP to detect active denitrifiers is limited by the assimilation of the labeled 13 C by other respiratory processes in denitrifiers (Butterbach-Bahl et al. 2013). Moreover, while the recent application of nano-scale secondary ion mass spectrometry (NanoSIMS) has demonstrated the potential for spatially tracking 15 Nenriched bacteria within the soil matrix (Herrmann et al. 2007), it does not allow detection of N2 O production in single cells, because N2 O can easily escape from the cells. In order to determine the spatial distribution of different N2 O-producing microorganisms in single cells across soil profiles, it is desirable to combine Nano-SIMS, 15 N, 18 O and 13 C isotope enrichment techniques with fluorescence-based methodologies (e.g. dyes changing fluorescence after reaction with produced N2 O).

Inhibition techniques The use of nitrification inhibitors together with nitrogen fertilizers has shown great potential in improving fertilizer efficiency and to reduce N2 O emissions and nitrate leaching in agroecosystems (Magalhaes, Chalk and Strong 1984; Di et al. 2010; Liu, Wang and Zheng 2013). Appropriate use of nitrification inhibitors could also serve as a powerful tool to differentiate biological N2 O pathways in laboratory experiments. Low levels (10– 100 Pa) of acetylene (C2 H2 ) can effectively inhibit nitrification, and therefore were thought to be able to eliminate N2 O production from both ammonia oxidation and nitrifier denitrification (Zhu et al. 2013). Thus N2 O produced in the low-level C2 H2 added treatments was predicted to be from only heterotrophic denitrification. The differentiation of N2 O pathways between ammonia oxidation and nitrifier denitrification could be further achieved by using O2 as an inhibitor of nitrifier denitrification (Wrage et al. 2004), but it should be noted that high levels of O2 cannot only inhibit nitrifier denitrification, but can also increase rates of N2 O production from ammonia oxidation (Fig. 3). Notably, C2 H2 inhibition of N2 O reductases has been also reported for denitrifying microorganisms (Toyoda et al. 2005). High concentrations of C2 H2 (>10 kPa) could inhibit both nitrification and the last step of heterotrophic denitrification (the conversion of N2 O to N2 ) and thus contribute to N2 O accumulation (Klemedtsson et al. 1988; Abed et al. 2013). The N2 O measured in the high-level C2 H2 added treatments (>10 kPa) can be defined as the gross N2 O production from heterotrophic denitrification. Therefore, care is needed to select appropriate levels of C2 H2 to avoid biased estimation against any particular microbial pathway. Although inhibitors are discriminative over pathways of nitrification and denitrification, it remains unclear whether they

739

can differentiate the relative contributions of AOA and AOB to N2 O (Chen et al. 2008a). AOA and AOB are divergent in many biochemical and generic features, such as membrane structures, cell size, amo gene structures and pathways of ammonia oxidation and carbon fixation (as reviewed in He, Hu and Zhang 2012). These differences are thought to affect their relative sensitivity to inhibitors, and thus many studies have tried to find specific inhibitors exclusively targeting AOA or AOB. However, it was reported that AOA were more strongly inhibited by dicyandiamide (DCD) compared with AOB in acidic soils (Zhang et al. 2012), while DCD had more effective inhibition on AOB in nitrogenrich grassland soils (Di et al. 2009; Dai et al. 2013). C2 H2 was also shown to be a non-selective nitrification inhibitor, and it can impede growth of AOA or AOB depending on which group is more functionally dominant in nitrification (Jia and Conrad 2009; Offre, Prosser and Nicol 2009). Some recently tested inhibitors in batch cultures included DCD, allylthiourea, amidinothiourea, nitrapyrin, antibiotic sulfathiazole and NO-scavenger carboxyPTIO with a broad range of concentrations (Shen et al. 2013). Apart from nitrapyrin, the other five inhibitors showed contrasting half maximal effective inhibitory concentrations between representative strains of soil AOA and AOB. For example, the inhibitor allylthiourea, targeting at reducing the AMO turnover rates, could markedly inhibit ammonia oxidation of AOB, but nearly 1000 times higher concentrations were needed to inhibit AOA; DCD inhibited growth of AOB at a concentration of 10 times lower that that effective on AOA; the ammonia oxidation rates of AOB were completely hampered by carboxy-PTIO at 52 μM, but even 200 μM carboxy-PTIO did not have strong impact on AOA (Shen et al. 2013); sulfonamide had no obvious inhibitory effect on AOA at doses which effectively inhibited growth of AOB (Schauss et al. 2009). Therefore, although no exclusive nitrification inhibitor has yet been found, a promising approach may be to choose an appropriate concentration of inhibitor which can completely inhibit AOA or AOB, but have little effect on the other. However, more strains of AOA and AOB need to be tested with these nitrification inhibitors in order to be able to generalize, and the efficacy should be tested in situ in more diverse soil types. In field studies and agricultural practices, the effects of the widely used nitrification inhibitors DCD and 3,4dimethylpyrazol phosphate (DMPP) on soil N2 O emissions are highly variable across soil types. For instance, application of DCD could reduce the total N2 O emissions by 65% in acidic soils, while its effects in alkaline or neutral soils were not obvious (Robinson et al. 2014). In a wheat–maize rotation, DCD and DMPP reduced the annual N2 O emissions by 35 and 38%, respectively (Liu, Wang and Zheng 2013), and reduced the cumulative N2 O emissions by 97 and 99% in a calcareous fluvo-aquic soil (Huang et al. 2014). For comparable or better inhibition of soil N2 O emissions, significantly lesser amounts of DMPP than DCD are used, and a meta-analysis found that application rates of DMPP and DCD ranged from 0.5–5 and 7–30 kg ha−1 , respectively, in agroecosystems (Liu, Wang and Zheng 2013). These commercial nitrification inhibitors could be used for partitioning N2 O sources under in situ field conditions, and their different effects on nitrifiers could help in designing low-emission agricultural guidelines for selection of nitrification inhibitors suitable for particular soil types. Although field application of C2 H2 produced from CaH2 granules was sometimes reported (Klemedtsson and Mosier 1994), this approach was widely criticized due to field decomposition of C2 H2 , utilization of C2 H2 for denitrification, acetylenecatalyzed oxidation of NO (Murray and Knowles 2003) and inadequate diffusion in water-saturated and fine-textured soils (Watts and Seitzinger 2000).

740


IMPROVING THE PERFORMANCE OF BIOGEOCHEMICAL N2 O MODELS BY INCORPORATING MICROBIAL DATA Current biogeochemical N2 O models and their limitations Computer simulation models, which can integrate a suite of climate and soil variables, and various interacting nitrogen transformation processes, are important for quantitative assessment of N2 O emissions (Del Grosso et al. 2006). To date, numerous biogeochemical models have been developed to predict nitrogen dynamics and N2 O fluxes under various climatic and landmanagement scenarios (as reviewed in Chen et al. 2008a), and generally been classified into three categories: laboratory, field and regional/global scales. At the global scale, the bottom-up models which are dependent on the extrapolation from individual chamber measurements to larger regions (Griffis et al. 2013), and the top-down models which are based on changes in atmospheric N2 O levels over time that are assigned to changes in anthropogenic activities known to influence N2 O fluxes (Davidson 2009), are in broad agreement (Shcherbak, Millar and Robertson 2014). The laboratory scale models could explicitly simulate diffusion of N2 O gases through soil profiles and aggregates from the denitrification pathway (Leffelaar and Wessel 1988; Arah and Smith 1989), but the N2 O released from soil nitrification was not simulated, which restricted their applicability to site-specific incubation experiments under anaerobic conditions. There is a large inconsistency in estimates of N2 O emissions at field and regional scales in diverse circumstances (Reay et al. 2012). Several conventional process-based field-scale N2 O simulation models with their main emission factors and input data, which have been the most widely used, are briefly described below, because most of the N2 O simulation applications were carried out at this scale, and this scale can bridge the knowledge gap between the laboratory and global level. (1) The NGAS-DAYCENT model captures daily fluxes of NO, N2 O and N2 from soils with finer spatiotemporal resolution, and simulates the nitrification-derived N2 O using a fixed fraction of the nitrification rate as a function of soil ammonium, moisture, pH, temperature and soil texture (Parton et al. 1996). The sub-model of DAYCENT simulates the denitrification-derived N2 O as a function of soil nitrate, moisture, labile carbon availability and soil physical properties (Del Grosso et al. 2000). Input data for DAYCENT includes site-specific soil properties, daily climate variables and historical and current land use practices. (2) The PnET-N-DNDC (Photosynthesis and evapotranspiration-nitrification-denitrification and decomposition) model was specifically developed to estimate daily N2 O fluxes through the nitrification and denitrification pathways from agricultural ecosystems (Li 2000; ButterbachBahl et al. 2001). This model requires detailed data on soil properties, vegetation, climatic information, atmospheric nitrogen inputs and land-use and land-management types, with special emphasis on the mechanistic description of nitrification and denitrification based on soil environmental variables. (3) The WNMM (water and nitrogen management model) is a spatially referenced biophysical model, coupled with a geographic information system, to simulate the key processes of carbon and nitrogen dynamics in intensive cropping systems (Li et al. 2008). This model estimates nitrificationderived N2 O as a function of nitrification rates and WFPS,

while it simulates denitrification-derived N2 O as a function of soil temperature, moisture and organic carbon contents. The ratio of N2 O to N2 produced is set to be fully controlled by the soil water saturation status (Li et al. 2008), rather than by the ratio of the (nirK + nirS)/nosZ gene abundances. (4) The Expert-N model describes the daily dynamics of water, carbon and nitrogen in soil-plant-atmosphere systems (Kaharabata et al. 2003). N2 O production from nitrification is simulated using a fixed fraction of nitrified ammonium, while denitrification-derived N2 O is regulated by the half saturation kinetics of nitrate contents. Input data for Expert-N include detailed soil properties segmented into different soil horizons, detailed crop properties during the growing season, daily meteorological data and historical information of soil carbon and nitrogen usage (Engel and Priesack 1993). (5) The CERES-NOE is a relatively simple model for production and consumption of N2 O through nitrification and denitrification in agricultural soils (Henault et al. 2005). CERES-NOE predicts denitrification-derived N2 O from potential denitrification rates as a function of soil water content, nitrate and temperature, while nitrification-derived N2 O is modeled from nitrification rates related to soil water and temperature. This model requires detailed site-specific parameters, such as potential denitrification/nitrification rates, soil moisture and proportions of nitrified and denitrified nitrogen emitted as N2 O. Comparative studies on four field-scale models (DAYCENT, Expert-N, DNDC and the daily version of NASA CASA) across five temperate agricultural sites found quite different simulations in terms of N2 O emissions (Frolking et al. 1998). DAYCENT overestimated annual N2 O emissions in nitrogen-rich pastures by more than 300%, while nitrification was grossly underestimated (Stehfest and Muller 2004). The FASSET model could well simulate the daily N2 O fluxes from soils, but could not predict most of the large measured daily N2 O peaks (Chatskikh et al. 2005). Comparison of three gas modules from loam-textured arable soils also found consistently different performances in simulating N2 O emissions between WNMM with DAYCENT and DNDC (Li et al. 2005). The inconsistency in these N2 O simulation models might be attributed to the oversimplification of the microbial processes of N2 O production in most of the biogeochemical ecosystem models, and the assumption that all soils would have the same microbial community phenotypes (Chen et al. 2008a; Bakken et al. 2012). For example, nitrification-derived N2 O is estimated using a fixed fraction of soil nitrification rates: DAYCENT used 2%, while DNDC used 0.25%, Expert-N used 0.5% and WNMM used 0.1–0.5% to estimate the contribution of nitrification to N2 O emissions (Chen et al. 2008a). The denitrification derived N2 O is predicted from soil moisture content and a default N2 O/(N2 O + N2 ) ratio (Butterbach-Bahl et al. 2013), while nitrogen fixation of leguminous crops has a default emission factor of 1.25% of the fixed nitrogen (Itakura et al. 2012). The regional scale NASA-CASA model adopted a default value of 2% of total mineralized nitrogen as the gaseous losses of N2 O, but this results in overestimation of emissions in most agroecosystems (as reviewed in Chen et al. 2008a). The International Panel on Climate Change (IPCC) methodology utilizes a singular emission factor of 1% of the added fertilizer nitrogen to produce the global inventory of agricultural N2 O emissions (IPCC 2007). However, as discussed previously, the nitrification-related and denitrification pathways for N2 O production vary greatly across soils with different soil conditions (Fig. 3), and meta-analysis revealed

Hu et al.

a nitrogen-induced emission factor ranging from 1.43 to 1.90 in different terrestrial ecosystems (Bouwman, Boumans and Batjes 2002). The fraction of N2 O emission to total nitrification is also influenced by environmental variables such as moisture and temperature, from around 0.03% at 5◦ C and 40% WFPS to 0.12% at 25◦ C and 60% WFPS (Chen et al. 2010b). Therefore, adoption of fixed ratios of relative contributions and emission factors in modeling will be definitely biased against particular soil types, and cannot account for the complex interplay of numerous microbial processes and the spatial-temporal variability of measured N2 O emissions owing to the regional variations in climate, ecosystems and land management. The common components of the process-based N2 O simulation models include soil-air, atmosphere and climate interactions, plant growth, carbon and nitrogen cycling (nitrification and denitrification) processes and land use management (as reviewed in Langeveld and Leffelaar 1996). Despite the increasing importance of microbially-mediated processes in N2 O emissions recognized by modeling efforts (Chen et al. 2008a), microbial populations have not been used as a major controlling factor for N2 O emissions in the majority of existing models (Bakken et al. 2012). The general absence of microbial data in modeling efforts is due to the previous assumption of minor effects of microbial community structures on large-scale models (Schimel 1995). However, the ever-increasing accumulation of next-generation sequencing data facilitates our ability to measure the enormous microbial diversity and the highly spatiotemporal dynamics of soil microbes (Lauber et al. 2009; Shade et al. 2013), and to characterize and predict the response of microbes to environmental parameters (Fierer et al. 2011). In the light of the strong impacts of the abundance, diversity, community structures and activity of nitrogen-cycling microorganisms on soil N2 O fluxes (see examples such as Ma et al. 2008; Avrahami and Bohannan 2009; Morales, Cosart and Holben 2010; Braker and Conrad 2011; Philippot et al. 2011; Santoro et al. 2011; Jung et al. 2014; Nemeth, Wagner-Riddle and Dunfield 2014), there is an urgent need to develop a new soilmicrobial-N2 O emission module for use in the conventional biogeochemical models (Wallenstein and Hall 2012). It has been also argued that ecological functions, like the production and consumption of N2 O by specialized groups of functional organisms, are more sensitive to changes in microbial community structure, and thus parameterization and incorporation of microbial data into models will have a great potential to improve their predictive power (Singh et al. 2010; Nazaries et al. 2013). In fact, there have been attempts at integrating microbial traits and function into ecosystem models at local and regional scales (Treseder et al. 2012). For instance, incorporation of microbial diversity was pioneered in a small-scale decomposition model based on the succession of three functional lineages of microbes with contrasting enzymatic capacities (Moorhead and Sinsabaugh 2006). Microbial dynamics have been considered in recent modeling efforts for nitrogen deposition (Gerber et al. 2010), and embedded as a new module to improve the performance of the Community Land Model for simulating soil carbon cycling on a global scale (Wieder, Bonan and Allison 2013). By incorporating the dormant and active microbial parameters in the microbial enzyme-mediated decomposition (MEND) model, estimates of microbial biomass carbon decomposition could be improved by 21–71% without accounting for the metabolic activity of the microbes (Wang et al. 2014). As far as we know, the abundance and community structures of microbes characterized by advanced molecular biology approaches have never been parameterized in ecosystem models, and modeling efforts have not kept pace

741

with the rapid advances in the microbial ecology of N2 O relevant microorganisms (Bakken et al. 2012).

A methodological framework from genes to ecosystem modeling of N2 O emissions In view of the inconsistencies in the estimation of N2 O emissions at field and regional scales, future modeling efforts should attempt to represent the direct control of microbial populations over N2 O rates and the dynamics of functional microorganisms in modeling parameters. We acknowledge that generating a set of ‘microbial N2 O emission factors’ across a wide coverage of climatic scenarios, land management practices and land use types is remarkably challenging, and such effort will considerably increase the number of parameters and model complexity. However, it is necessary to identify some simple microbial parameters which reflect the dominant processes of N2 O production and consumption. Up-scaling of microbial data to inform ecosystem decision making will rely on direct linkage between microbial communities and soil N2 O emissions (Fig. 7), and will also benefit from close collaborations between microbial ecologists, biogeochemists, agronomists, soil scientists and modelers. (1) At the cellular and genetic scales (Fig. 7), studies should be devoted to unravelling the genetic, phylogenetic and physiological characteristics of the currently available enriched or pure cultures of AOA, AOB and denitrifiers, and to determine their specific N2 O production rates and isotopic signatures of different pathways for each strain. It is also highly desirable to direct future efforts to cultivate more strains spanning all the major functional lineages of the N2 Oproducing and -consuming microorganisms. These studies could provide fundamental information for species-specific N2 O production rates of the major N2 O regulating groups. Meanwhile, a more mechanistic understanding of the exact biochemical models of the NH4 + -dependent N2 O formation by AOA and AOB is required, and the exact inhibition mechanisms of N2 O reductase (e.g. by O2 , pH and NO) need to be examined across divergent strains of denitrifiers. This will serve as the theoretical basis for constructing microbial modules in laboratory or field scale models. The importance of incorporating comprehensive microbial mechanisms of N2 O production to improve process descriptions and modeling confidence has been exemplified in existing mathematical models in wastewater treatment plants (Ni et al. 2013a, b). (2) Nitrogen-cycling microbes live in complex communities and closely interact with each other to produce and reduce soil N2 O emissions (Butterbach-Bahl et al. 2013). If one step is affected, other steps can be directly or indirectly influenced through metabolic networks due to lack of intermediates or substrates for the subsequent microbes. There is evidence for a strong and positive co-occurrence pattern between AOA and NOB in grassland soils, indicating that they may occupy the same niche space for carrying out nitrification (Daebeler et al. 2014). Therefore, diverse N2 O pathways are not only co-occurring in soils but are also mutually impacted, which necessitates a systems biology based approach (Bissett et al. 2013) to improve the mechanistic understanding of microbial regulation of N2 O emissions. At the community level, shotgun Illumina sequencing could provide comprehensive information for the genetic inventory of known and novel N2 O-relevant genes, and provide data for constructing metabolic network

742


Figure 7. A proposed methodological framework from genes to ecosystem modeling for better understanding of the microbially mediated soil N2 O production. Some examples of state-of-the-art approaches required to provide critical evidence for N2 O emissions at different scales are also listed beside the figure.

models (Barberan et al. 2012). Establishing these genomic networks will provide mechanistic descriptions of interactions between N2 O-relevant groups, and assess how the interactions at the community level will influence soil N2 O emissions. (3) In laboratory microcosms, quantitative response curves for rates of soil N2 O emissions as a function of changes in microbial communities (including abundance,

diversity, structures, expression, interactions and physiological capacities) and stoichiometric or kinetic parameter values related to different N2 O-generating microorganisms should be more accurately quantified under strictly controlled conditions (e.g. different levels of soil pH and water contents). The omics-based high-throughput approaches (e.g. metatranscriptomics, metaproteomics and metabolomics) could allow detailed information about the

Hu et al.

taxonomic, physiological and functional microbial properties impacting terrestrial N2 O emissions to be demonstrated, and might permit direct linkage with rates of N2 O production (Trivedi, Anderson and Singh 2013). However, results from these approaches can be misleading if N2 O production comes from inhibition of the N2 O reductase, but can inform us about the potentially relevant groups that should be studied under controlled laboratory conditions. Alternatively, there is clear niche separation between N2 O regulating organisms, where nitrification in alkaline or neutral soils and nitrogen-rich soils is dominated by AOB (Di et al. 2009; Shen, Xu and He 2014), whereas nitrification in acidic soils and nutrient-deficient soils is dominated by AOA (Zhang et al. 2012; Hu, Xu and He 2014b). Therefore, rates of soil N2 O emission via the nitrification-related pathways might be estimated based on the abundance and cellular N2 O rates of the dominant ammonia oxidizers in particular soils. These laboratory experimental data provide opportunities to parameterize all the known biochemical reactions at finer scales, but precise estimation and calibration of microbial parameters awaits examinations across a range of ecosystems. (4) Under field conditions, long-term and high-frequency monitoring of critical N2 O-relevant biomarkers (e.g. amoA, napA, nirK, nirS, narG and nosZ by using quantitative PCR and nextgeneration sequencing methodologies) linked with N2 O flux measurements (e.g. by application of automated chamber systems, N2 O microelectrodes and open-path Fourier transform infrared (OP-FTIR) (Bai et al. 2014) will be essential to account for the temporal dynamics of microbes, to obtain robust field datasets for different ecosystem types and to improve the simulation performance in N2 O modeling (Nazaries et al. 2013). New high-sensitivity instruments, such as the quantum cascade laser absorption spectrometer which is able to provide continuous observations of N2 O fluxes over long periods at realistic scales (Eugster et al. 2007) and with the potential to measure isotopic signatures of N2 O, will provide not only more accurate fluxes but also information on sources of N2 O. An example of such a monitoring program covering a large variety of ecosystems is the Australian National Agricultural Nitrous Oxide Research Program. Moreover, real-time measurements of the isotopic signatures (SP, 15 N and 18 O) of N2 O (by application of highresolution quantum cascade laser absorption spectroscopy with a greater sample throughput) are also highly recommended (Mohn et al. 2012), and these data could be compared with the isotopic data from culture-dependent studies to facilitate adequate interpretations of different microbial pathways over the long term. These efforts should be coupled with identifying sensitive indicator genes directly impacting N2 O fluxes, thereby reducing the explicit microbial parameters into a small set for use in N2 O models. (5) Previous investigations have found that large-scale distribution patterns of nitrogen-cycling genes are highly dependent on soil properties (Bru et al. 2011), and in particular, abundance, diversity and structures of AOA and AOB could be largely predicted by soil pH, with distinct phylotypes adapted to growing under different pH values (Hu et al. 2013). Therefore, it is important to systematically survey the biogeographic distribution of the N2 Orelevant indicator genes shaped by biotic and abiotic factors in various soil ecosystems at regional, national and global scales. Mapping these key N2 O-regulated microorganisms in combination with Geographic Information Systems and

743

satellite remote-sensing data, together with the speciesspecific capacity of N2 O production, will help with the provision of critical information for modeling large-scale N2 O emissions. (6) The final step involves integration and parameterization of the indicator genes into N2 O models, and further validation and optimization of these new N2 O emission modules against long-term N2 O emission data from diverse field studies and soil incubations. The key potential indicator genes, which have been widely reported with soil nitrification/denitrification rates and N2 O fluxes (see examples such as Ma et al. 2008; Avrahami and Bohannan 2009; Philippot et al. 2009; Di et al. 2010; Morales, Cosart and Holben 2010; Philippot et al. 2011; Harter et al. 2013; Nemeth, Wagner-Riddle and Dunfield 2014; Robinson et al. 2014), could be promising to be parameterized into future N2 O models. We propose that abundance of AOA and AOB amoA genes together with their specific N2 O production rates could be incorporated to infer the contribution from the nitrification pathway, while abundances of nirK, nirS and nosZ and ratios of (nirK + nirS)/nosZ) together with specific N2 O production/consumption rates could be used to simulate the contribution from the denitrification pathway and the ratios of N2 O/(N2 O + N2 ). Continent-scale datasets, for example, from the National Ecological Observatory Network and Long-Term Ecological Research sites in the United States will be highly desirable to validate the mechanistic equations in the new-generation microbially-based N2 O models. In this way, the microbial mechanisms for soil N2 O emission might be practically represented in ecosystem models, which should be rigorously compared across models to quantify the benefit of incorporating microbial diversity, function and evolution (Todd-Brown et al. 2012).

CONCLUDING REMARKS Globally, very few mitigation strategies are available to substantially reduce soil N2 O emissions, apart from slowing down nitrification by amendment of nitrification inhibitors (Chen et al. 2008a) and reducing the inputs of anthropogenic reactive nitrogen (Bakken et al. 2012). Large-scale N2 O mitigation options in agricultural practices call for an improvement in nitrogenuse efficiency through using slow- or controlled release fertilizers or fertilizers combined with urease or nitrification inhibitors (Di et al. 2010), plant breeding or engineering crop plants (Thomson et al. 2012), matching soil available nitrogen pool and crop nitrogen demand (Gentile et al. 2008), optimizing fertilizer placement and timing (Reay et al. 2012; Shcherbak, Millar and Robertson 2014), and improving land management to reduce anaerobic conditions and denitrification rates (Singh et al. 2010). However, the success of such strategies will rely on in-depth understanding of the physiology and regulatory biology of the key N2 O-producing and -reducing microorganisms (particularly nitrifiers and denitrifiers), and on efforts to eliminate N2 O production and/or to promote N2 O consumption at the microbial community level. The transcription of the key functional genes involved in N2 O production is regulated by a network of transcriptional and ancillary regulators (Zumft 1997), and understanding how they respond to a series of intra- and extracellular signals will be critical for the successful microbe-targeted options. One such example is that soil biochar amendment was recently demonstrated to be a potential mitigation option to reduce soil N2 O emissions by enhancing the abundance and

744


expression of the bacterial N2 O reductase and promoting the reduction of N2 O to N2 (Harter et al. 2013). Meanwhile, N2 O mitigation will also benefit from progress in discovering new microbes capable of reducing N2 O. For example, the denitrifier phenotype of Paracoccus denitrificans in batch cultures demonstrates an outstanding performance of reducing NO and N2 O all the way to N2 (Bakken et al. 2012), inoculation with nosZ-containing Bradyrhizobium japonicum effectively reduced N2 O emissions from soybean root systems in pot experiments (Itakura et al. 2012), and the presence of arbuscular mycorrhizal fungi induced a reduction of 34–42% in N2 O emissions (Bender et al. 2014); if their capacity could be exemplified in soils, they will be novel approaches to combat N2 O release. Overall, considering the principal roles of soil microorganisms in all the processes of N2 O production and consumption, we propose that exploring the functional genes and enzymes, as well as their regulatory mechanisms, should be central to any future strategy for controlling N2 O emissions from soil ecosystems. In addition, although great progress has been made, most of the challenges in modeling N2 O as summarized by Chen et al. (2008a) remains: lack of long-term largescale measurement of N2 O emissions to separate the N2 O contribution between nitrification and denitrification, incapability to partition N2 O and N2 in denitrification, and poor understanding of the interaction of nitrification inhibitors and soil properties. The incorporation of microbial processes into the biogeochemical and agroecosystem models is urgently needed to improve the accuracy of simulating N2 O emissions and to identify more effective and novel mitigation measures.

SUPPLEMENTARY DATA Supplementary data is available at FEMSRE online.

ACKNOWLEDGEMENTS We thank Dr PM Chalk (The University of Melbourne) and the two anonymous reviewers for their valuable comments on our manuscript.

FUNDING This work was financially supported by the Chinese Academy of Sciences (XDB15020200) and the Natural Science Foundation of China (41230857, 41025004), filling the Research Gap Program of Department of Agriculture/GRDC (1202.006), Australian Research Council (DE150100870) and MLA (B.FLT.0148) of Australia. Conflict of interest. None declared.

REFERENCES Abed RM, Lam P, de Beer D, et al. High rates of denitrification and nitrous oxide emission in arid biological soil crusts from the Sultanate of Oman. ISME J 2013;7:1862–75. Anderson IC, Poth M, Homstead J, et al. A comparison of NO and N2 O production by the autotrophic nitrifier Nitrosomonas europaea and the heterotrophic nitrifier Alcaligenes faecalis. Appl Environ Microb 1993;59:3525–33. Arah JRM, Smith KA. Steady-state denitrification in aggregated soils: a mathematical model. J Soil Sci 1989;40:139–49. Avrahami S, Bohannan BJM. N2 O emission rates in a California meadow soil are influenced by fertilizer level, soil moisture and the community structure of ammonia-oxidizing bacteria. Global Change Biol 2009;15:643–55.

Avrahami S, Liesack W, Conrad R. Effects of temperature and fertilizer on activity and community structure of soil ammonia oxidizers. Environ Microbiol 2003;5:691–705. Baggs EM. A review of stable isotope techniques for N2 O source partitioning in soils: recent progress, remaining challenges and future considerations. Rapid Commun Mass Sp 2008;22:1664–72. Baggs EM. Soil microbial sources of nitrous oxide: recent advances in knowledge, emerging challenges and future direction. Curr Opin Environ Sust 2011;3:321–7. Baggs EM, Blum H. CH4 oxidation and emissions of CH4 and N2 O from Lolium perenne swards under elevated atmospheric CO2 . Soil Biol Biochem 2004;36:713–23. Baggs EM, Richter M, Cadisch G, et al. Denitrification in grass swards is increased under elevated atmospheric CO2 . Soil Biol Biochem 2003;35:729–32. Bai M, Suter H, Lam SK, et al. Use of open-path FTIR and inverse dispersion technique to quantify gaseous nitrogen loss from an intensive vegetable production site. Atmos Environ 2014;94:687–91. Bakken LR, Bergaust L, Liu BB, et al. Regulation of denitrification at the cellular level: a clue to the understanding of N2 O emissions from soils. Philos T Roy Soc B 2012;367:1226–34. Balser TC, Firestone MFK. Linking microbial community composition and soil processes in a California annual grassland and mixed-conifer forest. Biogeochemistry 2005;73: 395–415. Barberan A, Bates ST, Oasamayor E, et al. Using network analysis to explore co-occurrence patterns in soil microbial communities. ISME J 2012;6:343–51. Bartossek R, Nicol GW, Lanzen A, et al. Homologues of nitrite reductases in ammonia-oxidizing archaea: diversity and genomic context. Environ Microbiol 2010;12:1075–88. Beaumont HJE, Hommes NG, Sayavedra-Soto LA, et al. Nitrite reductase of Nitrosomonas europaea is not essential for production of gaseous nitrogen oxides and confers tolerance to nitrite. J Bacteriol 2002;184:2557–60. Beaumont HJE, Lens SI, Reijnders WNM, et al. Expression of nitrite reductase in Nitrosomonas europaea involves NsrR, a novel nitrite-sensitive transcription repressor. Mol Microbiol 2004;54:148–56. Bender M, Conrad R. Kinetics of CH4 oxidation in oxic soils exposed to ambient air or high mixing ratios. FEMS Microbiol Ecol 1992;101:261–70. Bender SF, Plantenga F, Neftel A, et al. Symbiotic relationships between soil fungi and plants reduce N2 O emissions from soil. ISME J 2014;8:1336–45. Bergaust L, Mao Y, Bakken LR, et al. Denitrification response patterns during the transition to anoxic respiration and posttranscriptional effects of suboptimal pH on nitrous oxide reductase in Paracoccus denitrificans. Appl Environ Microb 2010;76:6387–96. Bissett A, Brown MV, Siciliano SD, et al. Microbial community responses to anthropogenically induced environmental change: towards a systems approach. Ecol Lett 2013;16:128– 39. Booth MS, Stark JM, Rastetter E. Controls on nitrogen cycling in terrestrial ecosystems: a synthetic analysis of literature data. Ecol Monogr 2005;75:139–57. Bouwman AF, Boumans LJM, Batjes NH. Modeling global annual N2 O and NO emissions from fertilized fields. Global Biogeochem Cy 2002;16:13. Bowatte S, Newton PCD, Brock S, et al. Bacteria on leaves: a previously unrecognized source of N2 O in grazed pastures. ISME J 2014;9:265–7.

Hu et al.

Braker G, Conrad R. Diversity, structure, and size of N2 Oproducing microbial communities in soils – what matters for their functioning? Adv Appl Microbiol 2011;75:33–70. Brochier-Armanet C, Boussau B, Gribaldo S, et al. Mesophilic crenarchaeota: proposal for a third archaeal phylum, the thaumarchaeota. Nat Rev Microbiol 2008;6:245–52. Bru D, Ramette A, Saby NPA, et al. Determinants of the distribution of nitrogen-cycling microbial communities at the landscape scale. ISME J 2011;5:532–42. Butterbach-Bahl K, Baggs EM, Dannenmann M, et al. Nitrous oxide emissions from soils: how well do we understand the processes and their controls? Philos T Roy Soc B 2013;368:20130122. Butterbach-Bahl K, Stange F, Papen H, et al. Regional inventory of nitric oxide and nitrous oxide emissions for forest soils of Southeast Germany using the biogeochemical model PnETN-DNDC. J Geophys Res 2001;106:34155–65. Cavigelli MA, Robertson GP. The functional significance of denitrifier community composition in a terrestrial ecosystem. Ecology 2000;81:1402–14. Chapuis-Lardy L, Wrage N, Metay A, et al. Soils, a sink for N2 O? a review. Global Change Biol 2007;13:1–17. Chatskikh D, Olesen J, Berntsen J, et al. Simulation of effects of soils, climate and management on N2 O emission from grasslands. Biogeochemistry 2005;76:395–419. Chen DL, Chalk PM, Freney JR. Nitrogen transformations in a flooded soil in the presence and absence of rice plants: 2. Denitrification. Nutr Cycl Agroecosys 1998;51:269–79. Chen DL, Li Y, Grace P, et al. N2 O emissions from agricultural lands: a synthesis of simulation approaches. Plant Soil 2008a;309:169–89. Chen DL, Suter HC, Islam A, et al. Influence of nitrification inhibitors on nitrification and nitrous oxide emission from a clay loam soil fertilized with urea. Soil Biol Biochem 2010b;42:660–4. Chen DL, Suter HC, Islam A, et al. Prospects of improving efficiency of fertilizer nitrogen in Australian agriculture: a review of enhanced efficiency fertilizers. Soil Res 2008a;46:289– 301. Clough TJ, Sherlock R, Rolston DE. A review of the movement and fate of N2 O in the subsoil. Nutr Cycl Agroecosys 2005;72:3–11. Cole JA, Brown CM. Nitrite reduction to ammonia by fermentative bacteria: a short circuit in the biological nitrogen cycle. FEMS Microbiol Lett 1980;7:65–72. Corker H, Poole RK. Nitric oxide formation by Escherichia coli dependence on nitrite reductase, the no-sensing regulator Fnr, and flavohemoglobin Hmp. J Biol Chem 2003;278:31584–92. Crenshaw CL, Lauber C, Sinsabaugh RL, et al. Fungal control of nitrous oxide production in semi-arid grassland. Biogeochemistry 2008;87:17–27. Cuhel J, Simek M, Laughlin RJ, et al. Insights into the effect of soil pH on N2 O and N2 emissions and denitrifier community size and activity. Appl Environ Microb 2010;76:1870–8. Daebeler A, Bodelier PLE, Zheng Y, et al. Interactions between Thaumarchaea, Nitrospira and methanotrophs modulate autotrophic nitrification in volcanic grassland soil. ISME J 2014;8:2397–410. Dai Y, Di HJ, Cameron KC, et al. Effects of nitrogen application rate and a nitrification inhibitor dicyandiamide on ammonia oxidizers and N2 O emissions in a grazed pasture soil. Sci Total Environ 2013;465:125–35. Davidson EA. The contribution of manure and fertilizer nitrogen to atmospheric nitrous oxide since 1860. Nat Geosci 2009;2:659–62.

745

Del Grosso SJ, Parton WJ, Mosier AR, et al. General model for N2 O and N2 gas emissions from soils due to dentrification. Global Biogeochem Cy 2000;14:1045–60. Del Grosso SJ, Parton WJ, Mosier AR, et al. DAYCENT Nationalscale simulations of nitrous oxide emissions from cropped soils in the United States. J Environ Qual 2006;35:1450–60. Di HJ, Cameron KC, Shen JP, et al. Nitrification driven by bacteria and not archaea in nitrogen-rich grassland soils. Nat Geosci 2009;2:621–4. Di HJ, Cameron KC, Sherlock RR, et al. Nitrous oxide emissions from grazed grassland as affected by a nitrification inhibitor, dicyandiamide, and relationships with ammonia-oxidizing bacteria and archaea. J Soil Sediment 2010;10:943–54. Elberling B, Christiansen HH, Hansen BU. High nitrous oxide production from thawing permafrost. Nat Geosci 2010;3:332–5. Engel T, Priesack E. Expert-N, a building block system of nitrogen models as resource for advice, research, water management and policy. In: Eijsackers HJP, Hamers T (eds). Integrated Soil and Sediment Research: A Basis for Proper Protection. Dordrecht, Niederlande: Kluwer Academic Publishers, 1993, 503–7. Eugster W, Zeyer K, Zeeman M, et al. Methodical study of nitrous oxide eddy covariance measurements using quantum cascade laser spectrometery over a Swiss forest. Biogeosciences 2007;4:927–39. Fierer N, Lauber CL, Ramirez KS, et al. Comparative metagenomic, phylogenetic and physiological analyses of soil microbial communities across nitrogen gradients. ISME J 2011;6:1007–17. Frame CH, Casciotti KL. Biochemical controls and isotopic signatures of nitrous oxide production by a marine ammoniaoxidizing bacterium. Biogeosciences 2010;7:2695–709. Freitag A, Rudert M, Bock E. Growth of nitrobacter by dissimilatoric nitrate reduction. FEMS Microbiol Lett 1987;48:105–9. Frolking SE, Mosier AR, Ojima DS, et al. Comparison of N2 O emissions from soils at three temperate agricultural sites: simulations of year-round measurements by four models. Nutr Cycl Agroecosys 1998;52:77–105. Galloway JN, Townsend AR, Erisman JW, et al. Transformation of the nitrogen cycle: recent trends, questions, and potential solutions. Science 2008;320:889–92. Gao H, Schreiber F, Collins G, et al. Aerobic denitrification in permeable Wadden Sea sediments. ISME J 2010;4:417–26. Gentile R, Vanlauwe B, Chivenge P, et al. Interactive effects from combining fertilizer and organic residue inputs on nitrogen transformations. Soil Biol Biochem 2008;40:2375–84. Gerber S, Hedin LO, Oppenheimer M, et al. Nitrogen cycling and feedbacks in a global dynamic land model. Global Biogeochem Cy 2010;24:GB1001. ¨ Godde M, Conrad R. Immediate and adaptational temperature effects on nitric oxide production and nitrous oxide release from nitrification and denitrification in two soils. Biol Fert Soils 1999;30:33–40. Griffis TJ, Lee X, Baker JM, et al. Reconciling the differences between top-down and bottom-up estimates of nitrous oxide emissions for the US Corn Belt. Global Biogeochem Cy 2013;27:746–54. Harter J, Krause HM, Schuettler S, et al. Linking N2 O emissions from biochar-amended soil to the structure and function of the N-cycling microbial community. ISME J 2013;8:660–74. He JZ, Hu HW, Zhang LM. Current insights into the autotrophic thaumarchaeal ammonia oxidation in acidic soils. Soil Biol Biochem 2012;55:146–54. He JZ, Shen JP, Zhang LM, et al. Quantitative analyses of the abundance and composition of ammonia-oxidizing

746


bacteria and ammonia-oxidizing archaea of a Chinese upland red soil under long-term fertilization practices. Environ Microbiol 2007;9:2364–74. Henault C, Bizouard F, Laville P, et al. Predicting in situ soil N2 O emissions using NOE algorithm and soil database. Glob Change Biol 2005;11:115–27. Herrmann AM, Clode PL, Fletcher IR, et al. A novel method for the study of the biophysical interface in soils using nanoscale secondary ion mass spectrometry. Rapid Commun Mass Sp 2007;21:29–34. Hofstra N, Bouwman AF. Denitrification in agricultural soils: Summarizing published data and estimating global annual rates. Nutr Cycl Agroecosys 2005;72:267–78. Horz HP, Barbrook A, Field CB, et al. Ammonia-oxidizing bacteria respond to multifactorial global change. P Natl Acad Sci USA 2004;101:15136–41. Hu HW, Macdonald CA, Trivedi P, et al. Water addition regulates the metabolic activity of ammonia oxidizers responding to environmental perturbations in dry sub-humid ecosystems. Environ Microbiol 2014a;17:444–61. Hu HW, Xu ZH, He JZ. Ammonia-oxidizing archaea play a predominant role in acid soil nitrification. Adv Agron 2014b;125:261–302. Hu HW, Zhang LM, Dai Y, et al. pH-dependent distribution of soil ammonia oxidizers across a large geographical scale as revealed by high-throughput pyrosequencing. J Soil Sediment 2013;13:1439–49. Huang T, Gao B, Hu XK, et al. Ammonia-oxidation as an engine to generate nitrous oxide in an intensively managed calcareous Fluvo-aquic soil. Sci Rep 2014;4:3950. IPCC. Climate change 2007: the physical science basis. In: Solomon S, Qin D, Manning M, et al. (eds). Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press, 2007. Itakura M, Uchida Y, Akiyama H, et al. Mitigation of nitrous oxide emissions from soils by Bradyrhizobium japonicum inoculation. Nat Clim Change 2012;3:208–12. Jia ZJ, Conrad R. Bacteria rather than archaea dominate microbial ammonia oxidation in an agricultural soil. Environ Microbiol 2009;11:1658–71. Jinuntuya-Nortman M, Sutka RL, Ostrom PH, et al. Isotopologue fractionation during microbial reduction of N2 O within soil mesocosms as a function of water-filled pore space. Soil Biol Biochem 2008;40:2273–80. Jones CM, Graf DRH, Bru D, et al. The unaccounted yet abundant nitrous oxide-reducing microbial community: a potential nitrous oxide sink. ISME J 2013;7:417–26. Jung MY, Park SJ, Min D, et al. Enrichment and characterization of an autotrophic ammonia-oxidizing archaeon of mesophilic crenarchaeal group I.1a from an agricultural soil. Appl Environ Microb 2011;77:8635–47. Jung MY, Well R, Min D, et al. Isotopic signatures of N2 O produced by ammonia-oxidizing archaea from soils. ISME J 2014;8:1115–25. Kaharabata SK, Drury CF, Priesack E, et al. Comparing measured and Expert-N predicted N2 O emissions from conventional till and no till corn treatments. Nutr Cycl Agroecosys 2003;66: 107–18. Kartal B, Almeida NM, Maalcke WJ, et al. How to make a living from anaerobic ammonium oxidation. FEMS Microbiol Rev 2013;37:428–61. Kartal B, Maalcke WJ, de Almeida NM, et al. Molecular mechanism of anaerobic ammonium oxidation. Nature 2011;479:127–30.

Kato T, Toyoda S, Yoshida N, et al. Isotopomer and isotopologue signatures of N2 O produced in alpine ecosystems on the Qinghai-Tibetan plateau. Rapid Commun Mass Sp 2013;27:1517–26. Kim JG, Jung MY, Park SJ, et al. Cultivation of a highly enriched ammonia-oxidizing archaeon of tharmarchaeotal group I.1b from an agricultural soil. Environ Microbiol 2012;14: 1528–43. Klemedtsson L, Mosier AR. Effect of long-term field exposure of soil to acetylene on nitrification, denitrification and acetylene consumption. Biol Fert Soils 1994;18:42–8. Klemedtsson L, Svensson BH, Rosswall T. A method of selective inhibition to distinguish between nitrification and denitrification as sources of nitrous oxide in soils. Biol Fert Soils 1988;6:112–9. Kool DM, Dolfing J, Wrage N, et al. Nitrifier denitrification as a distinct and significant source of nitrous oxide from soil. Soil Biol Biochem 2011;43:174–8. Kool DM, Wrage N, Zechmeister-Boltenstern S, et al. Nitrifier denitrification can be a source of N2 O from soil: A revised approach to the dual-isotope labelling method. Eur J Soil Sci 2010;61:759–72. Kowalchuk GA, Stephen JR. Ammonia-oxidizing bacteria: a model for molecular microbial ecology. Annu Rev Microbiol 2001;55:485–529. Kuiper I, De Deyn GB, Thakur MP, et al. Soil invertebrate fauna affect N2 O emissions from soil. Global Change Biol 2013;19:2814– 25. Langeveld CA, Leffelaar PA. Approaches in Field Scale Modeling of Nitrous Oxide Emission from Grassland Soils. Braunschweig: Technische Universitat Braunsweig and FAL Volkenrode, 1996. Lauber C, Hamady M, Knight R, et al. Pyrosequencing-based assessment of soil pH as a predictor of soil bacterial community structure at the continental scale. Appl Environ Microb 2009;75:5111–20. Laughlin RJ, Stevens RJ. Evidence for fungal dominance of denitrification and codenitrification in a grassland soil. Soil Sci Soc Am J 2002;66:1540–8. Leffelaar PA, Wessel WW. Denitrification in a homogeneous closed system: experiment and simulation. Soil Sci 1988;146:335–49. Leininger S, Urich T, Schloter M, et al. Archaea predominate among ammonia-oxidizing prokaryotes in soils. Nature 2006;442:806–9. Lennon JT, Jones SE. Microbial seed banks: the ecological and evolutionary implications of dormancy. Nat Rev Microbiol 2011;9:119–30. Li CS. Modeling trace gas emissions from agricultural ecosystems. Nutr Cycl Agroecosys 2000;58:259–76. Li Y, Chen DL, Barker-Reid F, et al. Simulation of N2 O emissions from rain-fed wheat and the impact of climate variation in southeastern Australia. Plant Soil 2008;309:239–51. Li Y, Chen DL, Zhang YM, et al. Comparison of three modeling approaches for simulating denitrification and nitrous oxide emissions from loam-textured arable soils. Global Biogeochem Cy 2005;19:1–15. Lipschultz F, Zafiriou OC, Wofsy SC, et al. Production of NO and N2 O by soil nitrifying bacteria. Nature 1981;294:641–3. Liu B, Morkved PT, Frostegard A, et al. Denitrification gene pools, transcription and kinetics of NO, N2 O, and N2 production as affected by soil pH. FEMS Microbiol Ecol 2010;72:407–17. Liu C, Wang K, Zheng X. Effects of nitrification inhibitors (DCD and DMPP) on nitrous oxide emission, crop yield, and nitrogen uptake in a wheat-maize cropping system. Biogeosciences 2013;10:2427–37.

Hu et al.

Loscher CR, Kock A, Koenneke M, et al. Production of oceanic nitrous oxide by ammonia-oxidizing archaea. Biogeosciences 2012;9:2095–122. Lund MB, Smith JM, Francis CA. Diversity, abundance and expression of nitrite reductase (nirK)-like genes in marine thaumarchaea. ISME J 2012;6:1966–77. Ma WK, Bedard-Haughn A, Siciliano SD, et al. Relationship between nitrifier and denitrifier community composition and abundance in predicting nitrous oxide composition from ephemeral wetland soils. Soil Biol Biochem 2008;40: 1114–23. Magalhaes AMT, Chalk PM, Strong WM. Effect of nitrapyrin on nitrous oxide emission from fallow soils fertilized with anhydrous ammonia. Fert Res 1984;5:411–21. Mathieu O, Henault C, Leveque J, et al. Quantifying the contribution of nitrification and denitrification to the nitrous oxide flux using 15 N tracers. Environ Pollut 2006;144:933–40. Medinets S, Skiba U, Rennenberg H, et al. A review of soil NO transformation: associated processes and possible physiological significance on organisms. Soil Biol Biochem 2015;80:92–117. Mohn J, Tuzson B, Manninen A, et al. Site selective real-time measurements of atmospheric N2 O isotopomers by laser spectroscopy. Atmos Meas Tech 2012;5:1601–9. Moorhead DL, Sinsabaugh RL. A theoretical model of litter decay and microbial interaction. Ecol Monogr 2006;76:151–74. Morales SE, Cosart T, Holben WE. Bacterial gene abundances as indicators of greenhouse gas emission in soils. ISME J 2010;4:799–808. Morley N, Baggs EM, Dorsch P, et al. Production of NO, N2 O and N2 by extracted soil bacteria, regulation by NO2 − and O2 concentrations. FEMS Microbiol Ecol 2008;65:102–12. Mothapo NV, Chen H, Cubeta MA, et al. Nitrous oxide producing activity of diverse fungi from distinct agroecosystems. Soil Biol Biochem 2013;66:94–101. Murray RE, Knowles R. Production of NO and N2 O in the presence and absence of C2 H2 by soil slurries and batch cultures of denitrifying bacteria. Soil Biol Biochem 2003;35:1115–22. Nazaries L, Murrell JC, Millard P, et al. Methane, microbes and models: fundamental understanding of the soil methane cycle for future predictions. Environ Microbiol 2013;15:2395–417. Nebert LD, Bloem J, Lubbers IM, et al. Association of earthwormdenitrifier interactions with increased emission of nitrous oxide from soil mesocosms amended with crop residue. Appl Environ Microb 2011;77:4097–104. Nemeth DD, Wagner-Riddle C, Dunfield KE. Abundance and gene expression in nitrifier and denitrifier communities associated with a field scale spring thaw N2 O flux event. Soil Biol Biochem 2014;73:1–9. Ni BJ, Ye L, Law Y, et al. Mathematical modeling of nitrous oxide (N2 O) emissions from full-scale wastewater treatment plants. Environ Sci Technol 2013a;47:7795–803. Ni BJ, Yuan Z, Chandran K, et al. Evaluating four mathematical models for nitrous oxide production by autotrophic ammonia-oxidizing bacteria. Biotechnol Bioeng 2013b;110:153–63. Offre P, Prosser JI, Nicol GW. Growth of ammonia-oxidizing archaea in soil microcosms is inhibited by acetylene. FEMS Microbiol Ecol 2009;70:99–108. Park S, Perez T, Boering KA, et al. Can N2 O stable isotopes and isotopomers be useful tools to characterize sources and microbial pathways of N2 O production and consumption in tropical soils? Global Biogeochem Cy 2011;25:GB1001.

747

Parton WJ, Mosier AR, Ojima DS, et al. Generalized model for N2 and N2 O production from nitrification and denitrification. Global Biogeochem Cy 1996;10:401–12. Patureau D, Zumstein E, Delgenes JP, et al. Aerobic denitrifiers isolated from diverse natural and managed ecosystems. Microb Ecol 2000;39:45–152. Philippot L, Andert J, Jones CM, et al. Importance of denitrifiers lacking the genes encoding the nitrous oxide reductase for N2 O emissions from soil. Global Change Biol 2011;17:1497–504. Philippot L, Cuhel J, Saby NPA, et al. Mapping field-scale spatial patterns of size and activity of the denitrifier community. Environ Microbiol 2009;11:1518–26. Philippot L, Hallin S, Schloter M. Ecology of denitrifying prokaryotes in agricultural soil. Adv Agron 2007;96:249–305. Philippot L, Piutti S, Martin-Laurent F, et al., Molecular analysis of the nitrate-reducing community from unplanted and maize-planted soils. Appl Environ Microb 2002;68:6121–8. Pilegaard K. Processes regulating nitric oxide emissions from soils. Philos T Roy Soc B 2013;368:20130126. Pratscher J, Dumont MG, Conrad R. Ammonia oxidation coupled to CO2 fixation by archaea and bacteria in an agricultural soil. P Natl Acad Sci USA 2011;108:4170–5. Prendergast-Miller MT, Baggs EM, Johnson DJ. Nitrous oxide production by the ecotomycorrhizal fungi Paxillus involutus and Tylospora fibrillosa. FEMS Microbiol Lett 2011;316:31–5. Purkhold U, Pommerening-Roser A, Juretschko S, et al. Phylogeny of all recognized species of ammonia oxidizers based on comparative 16S rRNA and amoA sequence analysis: implications for molecular diversity surveys. Appl Environ Microb 2000;66:5368–82. Ravishankara AR, Daniel JS, Portmann RW. Nitrous oxide (N2 O): the dominant ozone-depleting substance emitted in the 21st century. Science 2009;326:123–5. Reay DS, Davidson EA, Smith KA, et al. Global agriculture and nitrous oxide emissions. Nat Clim Change 2012;2:410–6. Remde A, Conrad R. Production of nitric oxide in Nitrosomonas europaea by reduction of nitrite. Arch Microbiol 1990;154:187– 91. Richardson D, Felgate H, Watmough N, et al. Mitigating release of the potent greenhouse gas N2 O from the nitrogen cycle – could enzymic regulation hold the key? Trends Biotechnol 2009;27:388–97. Ritchie GAF, Nicholas DJD. Identification of the sources of nitrous oxide produced by oxidative and reductive processes in Nitrosomonas europaea. Biochem J 1972;126:1181–91. Robinson A, Di HJ, Cameron KC, et al. The effect of soil pH and dicyandiamide (DCD) on N2 O emissions and ammonia oxidizer abundance in a stimulated grazed pasture soil. J Soil Sediment 2014;14:1434–44. Rutting T, Boeckx P, Muller C, et al. Assessment of the ¨ ¨ importance of dissimilatory nitrate reduction to ammonium for the terrestrial nitrogen cycle. Biogeosciences 2011;8: 1779–91. Sanchez-Martin L, Vallejo A, Dick J, et al. The influence of solube carbon and fertilizer nitrogen on nitric oxide and nitrous oxide emissions from two contrasting agricultural soils. Soil Biol Biochem 2008;40:142–51. Sanford RA, Wagner DD, Wu Q, et al. Unexpected nondenitrifier nitrous oxide reductase gene diversity and abundance in soils. P Natl Acad Sci USA 2012;109:19709–14. Santoro AE, Buchwald C, McIlvin MR, et al. Isotopic signature of N2 O produced by marine ammonia-oxidizing archaea. Science 2011;333:1282–5.

748


Schauss K, Focks A, Leininger S, et al. Dynamics and functional relevance of ammonia-oxidizing archaea in two agricultural soils. Environ Microbiol 2009;11:446–56. Schimel J. Arctic and Alpine Biodiversity: Patterns, Causes and Ecosystem Consequences. Berlin: Springer, 1995. Schmidt CS, Richardson DJ, Baggs EM. Constraining the conditions conductive to dissimilatory nitrate reduction to ammonium in soil. Soil Biol Biochem 2011;7:1607–11. Schreiber F, Wunderlin P, Udert KM, et al. Nitric oxide and nitrous oxide turnover in natural and engineered microbial communities: biological pathways, chemical reactions, and novel technologies. Front Microbiol 2012;3:372. Shade A, Caporaso JG, Handelsman J, et al. A meta-analysis of changes in bacterial and archaeal communities with time. ISME J 2013;7:1493–506. Shaw LJ, Nicol GW, Smith Z, et al. Nitrosospira spp. can produce nitrous oxide via a nitrifier denitrification pathway. Environ Microbiol 2006;8:214–22. Shcherbak L, Millar N, Robertson GP. Global metaanalysis of the nonlinear response of soil nitrous oxide (N2 O) emissions to fertilizer nitrogen. P Natl Acad Sci USA 2014;111:9199–204. Shen JP, Xu ZH, He JZ. Frontiers in the microbial processes of ammonia oxidation in soils and sediments. J Soil Sediment 2014;14:1023–9. Shen JP, Zhang LM, Zhu YG, et al. Abundance and composition of ammonia-oxidizing bacteria and ammonia-oxidizing archaea communities of an alkaline sandy loam. Environ Microbiol 2008;10:1601–11. Shen T, Stieglmeier M, Dai J, et al. Responses of the terrestrial ammonia-oxidizing archaeon Ca. Nitrososphaera viennensis and the ammonia-oxidizing bacterium Nitrosospira multiformis to nitrification inhibitors. FEMS Microbiol Lett 2013;344:121–9. Shoun H, Fushinobu S, Jiang L, et al. Fungal denitrification and nitric oxide reductase cytochrome P450nor. Philos T Roy Soc B 2012;367:1186–94. Singh BK, Bardgett RD, Smith P, et al. Microorganisms and climate change: terrestrial feedbacks and mitigation options. Nat Rev Microbiol 2010;8:779–90. Silver WL, Herman DJ, Firestone MK. Dissimilatory nitrate reduction to ammonium in upland tropical forest soils. Ecology 2001;8:2410–6. Simon J, Einsle O, Kroneck PMH, et al. The unprecedented nos gene cluster of Wolinella succinogenes encodes a novel respiratory electron transfer pathway to cytochrome c nitrous oxide reductase. FEBS Lett 2004;569:7–12. Spang A, Poehlein A, Offre P, et al. The genome of the ammoniaoxidizing Candidatus Nitrososphaera gargensis: insights into metabolic versatility and environmental adaptation. Environ Microbiol 2012;14:3122–45. Starkenburg SR, Arp DJ, Bottomley PJ. Expression of a putative nitrite reductase and the reverse inhibition of nitritedependent respiration by nitric oxide in Nitrobacter winogradskyi Nb-225. Environ Microbiol 2008;10:3036–42. Stehfest E, Muller C. Simulation of N2 O emissions from a urineaffected pasture in New Zealand with the ecosystem model DayCent. J Geophys Res 2004;109:D03109. Stevens RJ, Laughlin RJ. Measurement of nitrous oxide and dinitrogen emissions from agricultural soils. Nutr Cycl Agroecosys 1998;52:131–9. Stieglmeier M, Mooshammer M, Kitzler B, et al. Aerobic nitrous oxide production through N-nitrosating hybrid formation in ammonia-oxidizing archaea. ISME J 2014;8:1135–46.

Strous M, Pelletier E, Mangenot S, et al. Deciphering the evolution and metabolism of an anammox bacterium from a community genome. Nature 2006;440:790–4. Sutka RL, Adams GC, Ostrom NE, et al. Isotopologue fractionation during N2 O production by fungal denitrification. Rapid Commun Mass Sp 2008;22:3989–96. Sutka RL, Ostrom NE, Ostrom PH, et al. Nitrogen isotopomer site preference of N2 O produced by Nitrosomonas europaea and Methylococcus capsulatus Bath. Rapid Commun Mass Sp 2004;18:1411–2. Sutka RL, Ostrom NE, Ostrom PH, et al. Distinguishing nitrous oxide production from nitrification and denitrification on the basis of isotopomer abundances. Appl Environ Microb 2006;72:638–44. Thamdrup B. New pathways and processes in the global nitrogen cycle. Annu Rev Ecol Evol S 2012;43:407–28. Thomson AJ, Fiannopoulos G, Pretty J, et al. Biological sources and sinks of nitrous oxide and strategies to mitigate emissions. Philos T Roy Soc B 2012;367:1157–68. Tiedje JM, Sexstone AJ, Myrold DD, et al. Denitrification: ecological niches, competition and survival. Anton Leeuw 1982;48:569–83. Todd-Brown KEO, Hopkins FM, Kivilin SN, et al. A framework for representing microbial decomposition in coupled climate models. Biogeochemistry 2012;109:19–33. Tourna M, Freitag TE, Nicol GW, et al. Growth, activity and temperature responses of ammonia-oxidizing archaea and bacteria in soil microcosms. Environ Microbiol 2008;10:1357–64. Tourna M, Stieglmeier M, Spang A, et al. Nitrososphaera viennensis, an ammonia oxidizing archaeon from soil. P Natl Acad Sci USA 2011;108:8420–5. Toyoda S, Mutobe H, Yamagishi H, et al. Fractionation of N2 O isotopomers during production by denitrifier. Soil Biol Biochem 2005;37:1535–45. Treseder KK, Balser TC, Bradford MA, et al. Integrating microbial ecology into ecosystem models: challenges and priorities. Biogeochemistry 2012;109:7–18. Trivedi P, Anderson IC, Singh BK. Microbial modulators of soil carbon storage: integrating genomic and metabolic knowledge for global prediction. Trends Microbiol 2013;21: 641–51. Turner D, Chen D, Galbally I, et al. Spatial variability of nitrous oxide emissions from irrigated dairy pasture. Plant Soil 2008;309:77–88. US EPA. Global Anthropogenic Non-CO2 Greenhouse Gas Emissions: 1990–2030. Washington: U.S. Environmental Protection Agency, 2012. Vajrala N, Martens-Habbena W, Sayavedra-Soto LA, et al. Hydroxylamine as an intermediate in ammonia oxidation by globally abundant marine archaea. P Natl Acad Sci USA 2013;110:1006–11. Van Cleemput O. Subsoils: chemo-and biological denitrification, N2 O, and N2 emissions. Nutr Cycl Agroecosys 1998;52:187–94. Van Groeningen KJ, Osenberg CW, Hungate BA. Increase soil emissions of potent greenhouse gases under increased atmospheric CO2 . Nature 2011;475:214–6. van Wonderen JH, Burlat B, Richardson DJ, et al. The nitric oxide reductase activity of cytochrome c nitrite reductase from Escherichia coli. J Biol Chem 2008;283:9587–94. Walker CB, de la Torre JR, Klotz MG, et al. Nitrosopumilus maritimus genome reveals unique mechanisms for nitrification and autotrophy in globally distributed marine crenarchaea. P Natl Acad Sci USA 2010;107:8818–23.

Hu et al.

Wallenstein MD, Hall EK. A trait-based framework for predicting when and where microbial adaptation to climate change will affect ecosystem functioning. Biogeochemistry 2012;109: 35–47. Wang G, Jagadamma S, Mayes MA, et al. Microbial dormancy improves development and experimental validation of ecosystem model. ISME J 2014;9:226–37. Watts SH, Seitzinger SP. Denitrification rates in organic and mineral soils from riparian sites: a comparison of N2 flux and acetylene inhibition methods. Soil Biol Biochem 2000;32: 1383–92. Webster EA, Hopkins DW. Contributions from different microbial processes to N2 O emission from soil under different moisture regimes. Biol Fert Soils 1996;22:331–5. Well R, Flessa H, Lu X, et al. Isotopologue ratios of N2 O emitted from microcosms with NH4 + fertilized arable soils under conditions favoring nitrification. Soil Biol Biochem 2008;40:2416–26. Wieder WR, Bonan GB, Allison SD. Global soil carbon projections are improved by modeling microbial processes. Nat Clim Change 2013;3:909–12. Wrage N, van Groenigen JW, Oenema O, et al. A novel dualisotope labelling method for distinguishing between soil sources of N2 O. Rapid Commun Mass Sp 2005;19:3298–306. Wrage N, Velthof GL, Oenema O, et al. Acetylene and oxygen as inhibitors of nitrous oxide production in Nitrosomonas europaea and Nitrosospira briensis: a cautionary tale. FEMS Microbiol Ecol 2004;47:13–8. Wrage N, Velthof GL, van Beusichem ML, et al. Role of nitrifier

749

denitrification in the production of nitrous oxide. Soil Biol Biochem 2001;33:1723–32. Wunderlin P, Mohn J, Joss A, et al. Mechanism of N2 O production in biological wastewater treatment under nitrifying and denitrifying conditions. Water Res 2012;46:1027–37. Yamulki S, Toyoda S, Yoshida N, et al. Diurnal fluxes and the isotopomerratios of N2 O in a temperate grassland following urine amendment. Rapid Commun Mass Sp 2001;15:1263–9. Yanai Y, Toyota K, Morishita T, et al. Fungal N2 O production in an arable peat soil in central Kalimantan, Indonesia. Soil Sci Plant Nutr 2007;53:806–11. Yin SX, Chen D, Chen LM, et al. Dissimilatory nitrate reduction to ammonium and responsible microorganisms in two Chinese and Australian paddy soils. Soil Biol Biochem 2002;34:1131–7. Yu R, Chandran K. Strategies of Nitrosomonas europaea 19718 to counter low dissolved oxygen and high nitrite concentrations. BMC Microbiol 2010;10:70. Zhang LM, Hu HW, Shen JP, et al. Ammonia-oxidizing archaea have more important role than ammonia-oxidizing bacteria in ammonia oxidation of strongly acidic soils. ISME J 2012;6:1032–45. Zhang LM, Offre PR, He JZ, et al. Autotrophic ammonia oxidation by soil thaumarchaea. P Natl Acad Sci USA 2010;107:17240–5. Zhu X, Burger M, Doane TA, et al. Ammonia oxidation pathways and nitrifier denitrification are significant sources of N2 O and NO under low oxygen availability. P Natl Acad Sci USA 2013;110:6328–33. Zumft WG. Cell biology and molecular basis of denitrification. Microbiol Mol Biol R 1997;61:533–616.