Implications
Nitrous oxide (N2O) is a potent greenhouse gas and the leading stratospheric ozone-depleting emission. Its concentration continues to increase, predominantly owing to N2O emissions associated with animal production systems. The number of farmed animals will increase to satisfy growing global population demand. Mitigation is required. Tools to identify mitigation options include the stable isotopes of nitrogen (15N). Past and recently developed 15N methodologies provide tools to better understand the sources of N2O, while tracing the substrates responsible for N2O fluxes over space and time through the environment. Integrating the skill sets and efforts of microbiologists, soil and animal scientists will achieve this.
Introduction
Nitrous oxide (N2O) is a greenhouse gas and the dominant stratospheric ozone-depleting substance emitted by humans in the twenty-first century (Ravishankara et al., Reference Ravishankara, Daniel and Portmann2009). Concentrations of N2O have increased since 1800 and continue to do so at 0.26% per annum (Forster et al., Reference Forster, Ramaswamy, Artaxo, Berntsen, Betts, Fahey, Haywood, Lean, Lowe, Myhre, Nganga, Prinn, Raga, Schulz and Van Dorland2007). Anthropogenic emissions (6.7 Tg/year) represent 40% of the total annual (17.7 Tg/year) global emissions of N2O (Forster et al., Reference Forster, Ramaswamy, Artaxo, Berntsen, Betts, Fahey, Haywood, Lean, Lowe, Myhre, Nganga, Prinn, Raga, Schulz and Van Dorland2007). The main anthropogenically derived substrates for N2O production include synthetic and organic fertilisers and excreta; thus, agricultural production of N2O (2.8 Tg/year) dominates anthropogenic emission sources (Denman et al., Reference Denman, Brasseur, Chidthaisong, Ciais, Cox, Dickinson, Hauglustaine, Heinze, Holland, Jacob, Lohmann, Ramachandran, da Silva Dias, Wofsy and Zhang2007; Schils et al., Reference Schils, Eriksen, Ledgard, Vellinga, Kuikman, Luo, Petersen and Velthof2013). Davidson (Reference Davidson2009) showed that the observed increase in tropospheric N2O between 1860 and 2005 could be explained by 2.0% of manure nitrogen (N) and 2.5% of fertiliser N being emitted as N2O over this period. Oenema et al. (Reference Oenema, Wrage, Velthof, van Groenigen, Dolfing and Kuikman2005) partitioned N2O emissions from animal production systems (1.5 Tg/year) into five sources: dung and urine from grazing animals deposited in pastures (41%), indirect sources (27%), animal wastes from housing and storage (19%), application of animal wastes to land (10%) and the burning of dung (3%). The majority of these N2O emissions from agriculture occur as a result of excreta being deposited onto soil where microbial reactions, driven by nitrifiers and denitrifiers, are the major biological pathways for N2O. The Food and Agriculture Organisation of the United Nations predicts the demand for animal products to increase, and thus animal numbers are projected to increase significantly (OECD_FAO, 2012). Given the current trends in tropospheric N2O concentrations and the predicted increases as a result of increasing animal-based agricultural production, it is urgent that a fuller understanding of the sources, processes and management factors that contribute to these agriculturally based N2O emissions is developed. Improved knowledge will lead to mitigation options. It is necessary to differentiate agricultural N inputs from N forms found in the soil and which can also contribute to N2O fluxes. This permits the tracing of a particular N input through the agroecosystem concerned, over time, enabling a full and detailed understanding of the factors affecting N2O fluxes, their duration and magnitude. This article aims to increase awareness of the stable isotopes of N and the vital role they can play in distinguishing N2O sources and fate in animal production systems. Specific reference to methodologies for preparing 15N-enriched samples for analysis and other practical considerations may be found in the literature (Knowles and Blackburn, Reference Knowles and Blackburn1993; Hauck et al., Reference Hauck, Meisinger and Mulvaney1994).
Stable isotopes of nitrogen
The number of protons in an element (equal to the atomic number ‘Z’) is constant, but the number of neutrons (the neutron number ‘Nn’) may vary. Isotopes of a given element differ from one another owing to the number of neutrons they contain. This variation in neutron number does not affect the gross chemical properties of the element. The mass of an element (the sum of Z + Nn) is the superscripted number to the left of the element designation (Kendall and Caldwell, Reference Kendall and Caldwell1998). Thus, for N with an atomic number of 7, the stable isotopes have mass numbers of 14 (14N with seven neutrons) and 15 (15N with eight neutrons). These N isotopes occur naturally in the environment (Sharp, Reference Sharp2007). In air, the natural abundance of 15N is constant with a 15N/14N ratio equal to 1/272 or 0.3676% (Junk and Svec, Reference Junk and Svec1958).
Further terms used when studying the isotopic composition of molecules include the words ‘isotopologue’ and ‘isotopomer’. Isotopologues, as defined by the International Union of Pure and Applied Chemistry, are molecules differing from one another only because of isotopic composition (Sharp, Reference Sharp2007). For example, N2O with mass 44 (14N + 14N +16O) and mass 45 (14N + 15N + 16O) are isotopologues of N2O. However, isotopomers always have the same number of each isotopic atom and as a consequence always have the same mass (Sharp, Reference Sharp2007). For example, the N2O molecules comprising 14N + 15N + 16O and 15N + 14N + 16O are isotopomers of N2O (see below).
The stable isotopes of N provide a unique research tool to elucidate the N2O emission pathways and N2O fate. Scientists may use N compounds that have been enriched in 15N. This means that more 15N has been added to the substrate of interest before its use in the experimental set-up. Such experiments may be described as ‘tracer experiments’ where the extra 15N added, that is, the enrichment, is far in excess of the natural abundance of 15N, thus permitting the scientist to ‘trace’ the flow of 15N through the ecosystem of interest. Alternatively, the scientist may elect to use the 15N already existing in the ecosystem of concern, at natural levels of abundance, and these are termed ‘natural abundance’ experiments. Initially, the use of 15N-enriched tracer will be discussed followed by natural abundance. Fractionation causes changes in the natural abundance of 15N and this is considered below.
15N-enriched tracer studies
As with any experiment, the rationale and aims of a 15N-tracer study need to be clearly thought out. In a tracer experiment, the information gathered from the measurements made includes not only the concentrations of N in the various N pools (e.g. N2O μl/l) but also the level of 15N enrichment (e.g. N2O atom% 15N). The term ‘atom% 15N enrichment’ is an expression that denotes the concentration of 15N as a percentage of the total mass of stable N atoms (14N + 15N) and is calculated numerically as:
$${\rm{atom}}\,\% {{\,}^{{\rm{15}}}} {\rm{N}}\: = \:\frac{{{\rm{No}}{\rm{.}}\,{\rm{of}}{{\,}^{{\rm{15}}}} {\rm{N}}\,{\rm{atoms}}}}{{{\rm{No}}{\rm{.}}\,{\rm{of}}{{\,}^{{\rm{14}}}} {\rm{N}}\:{\rm{ + }}{{\:}^{{\rm{15}}}} {\rm{N}}\,{\rm{atoms}}}}\:\times \:\frac{{{\rm{100}}}}{{\rm{1}}}$$
Although several methods may be used to measure the atom% 15N enrichment of a sample (e.g. Fourier transform infrared (FTIR) determination of 15N2O), the most common method to date has been isotope-ratio mass spectrometry (IRMS). Rather than measuring individual atoms to determine atom% 15N, the mass spectrometer measures the molecules of interest on the basis of their mass, which is a function of their isotopic composition. If the sample is a solid, it must first be combusted to produce N2, and then it can be carried in a He flow through the mass spectrometer. Similarly, dissolved N forms, such as inorganic-N, also need to be converted either to a solid form, and then to a gas (Stark and Hart, Reference Stark and Hart1996). Alternatively, the sample may already be in a gaseous state (Stevens and Laughlin, Reference Stevens and Laughlin1994; Laughlin et al., Reference Laughlin, Stevens and Zhuo1997). A detailed explanation of mass spectrometry is beyond the scope of this paper and the reader is directed to other sources (e.g. Mulvaney, Reference Mulvaney1993; Sharp, Reference Sharp2007). In brief, the gas molecule is introduced into the mass spectrometer's ‘ion source’ where a fraction of the gas molecules are ionised. The positively charged ions are then moved through a magnetic field with the positively charged ions deflected in a circular trajectory, on the basis of their mass to charge ratio (m/z). The charged ions are collected in Faraday cups, thus forming ion currents (I) that are proportional to the quantity of gas. For the N2O molecule, the masses of interest are 44, 45 and 46, and for N2 these masses are 28, 29 and 30. For a given molecular species, the ion currents are used to produce ratios (R). Therefore, for N2, the ratios 29R (29I/28I) and 30R (30I/28I) are derived (Stevens et al., Reference Stevens, Laughlin, Atkins and Prosser1993). For solid samples, the ratios 29R and 30R, which are derived from the combusted materials, can be used in the appropriate equations to determine the atom% 15N enrichment (Mulvaney, Reference Mulvaney1993). Further consideration must also be given to comparing these ratios against standards and determining instrument factors, so that only the true ratio differences between normal and enriched atmospheres are used when deriving gaseous N fluxes (Stevens et al., Reference Stevens, Laughlin, Atkins and Prosser1993; Stevens and Laughlin, Reference Stevens and Laughlin1998). A major focus of many 15N-isotopic studies is to derive N2 and N2O fluxes from the soil nitrate pool. Further detailed explanations of the assumptions, derivations and implementation of the ratios used when determining denitrification of 15N-enriched NO3−, and the respective calculation of 15N-enriched fluxes of N2 and N2O, can be found in the literature (Mulvaney and Boast, Reference Mulvaney and Boast1986; Boast et al., Reference Boast, Mulvaney and Baveye1988; Mosier and Schimel, Reference Mosier and Schimel1993; Mulvaney, Reference Mulvaney1993; Stevens et al., Reference Stevens, Laughlin, Atkins and Prosser1993; Stevens and Laughlin, Reference Stevens and Laughlin1998; Bergsma et al., Reference Bergsma, Ostrom, Emmons and Robertson2001). The ion currents at m/z 44, 45 and 46 are used to calculate the concentration of N2O, in conjunction with a reference gas, whereas its 15N enrichment is calculated from ratios 45R or 46R (Stevens et al., Reference Stevens, Laughlin, Atkins and Prosser1993; Stevens and Laughlin, Reference Stevens and Laughlin1998).
Using 15N-enriched substrates in tracer studies
A potential artefact when using 15N-tracer studies is that the application of a relatively large rate of N may unnaturally perturb the system under investigation. Fortunately, in agroecosystems, this is less of an issue, as substrates containing 15N, used in tracer studies to follow N2O and N2 emissions, generally consist of N forms such as fertilisers, excreta (dung and urine) and plant residues, which by their very nature perturb the system. Therefore, these types of N substrates are potentially ideal for use in 15N-tracer studies, if suitable enrichment of the substrate can be undertaken.
Once the experiment has been conceived, the experimental treatments must be refined. The first step is to consider the rate of N required in any given treatment. The rationale for this is similar to any non-15N experimentation where an N substrate is being added. Thought then needs to go into determining which of the various N pools the 15N tracer will be measured in, how often the N pools will be measured and what will be the period of the experiment. Naturally, the type of N substrate being applied will also have a bearing on the N pools being measured and their frequency of measurement. If the 15N tracer is to be followed in multiple pools (e.g. soil, plant and gases) over time, dilution of the 15N tracer may occur immediately in the soil, because of antecedent N, or at a slower rate because of other N inputs resulting from management and/or soil-N mineralisation. The 15N enrichment in the N pool of interest may also be diluted because of the pool of background 14N being significantly larger than the 15N pool evolving (e.g. 15N2 evolving into ambient air). Alternatively, significant loss of the 15N tracer may occur early in the experiment. For example, if the experiment aims to follow the contribution of 15N-enriched urea to an N2O flux, then allowance needs to be made for a significant fraction of the urea fertiliser (∼20%), and the 15N embodied in it, to be potentially lost within hours of the experiment, starting as a result of ammonia volatilisation. Thus, the N pool(s) of interest, dilution and early loss of 15N from the experimental system need to be considered when deciding on the level of 15N enrichment to use.
Finally, the last factor to consider when determining what level of 15N enrichment to use is the sensitivity of the mass spectrometer (Stevens et al., Reference Stevens, Laughlin, Atkins and Prosser1993). The more sensitive the mass spectrometer is, the lower the required 15N enrichment, all things being equal. The experimental set-up and environmental conditions may also determine the level of enrichment to use. When measuring gas fluxes from soils with headspace chambers, the sensitivity increases with high gas fluxes and smaller headspaces (large surface area to volume ratios). Stevens et al. (Reference Stevens, Laughlin, Atkins and Prosser1993) provide an excellent example of how to determine sensitivity for a mass spectrometer. As 15N enrichment and gas fluxes decrease, the coefficient of variation will increase and data quality will suffer (Stevens et al., Reference Stevens, Laughlin, Atkins and Prosser1993). For example, using an enclosure time of 2 h with an enclosure volume: surface area ratio of 5 : 1, and a NO3−-15N enrichment of 60 atom% Stevens and Laughlin (Reference Stevens and Laughlin1998) reported a limit of detection for N2 fluxes of 3.5 g N2-N/ha per day. A point not always recognised by researchers using 15N enrichment to measure N2 fluxes is the requirement to be able to measure both the 29R and 30R ratio. In order to do this, there must be both 14N and 15N present. Thus, starting an experiment with a substrate that is 98 atom% 15N enriched (a commonly available 15N enrichment), and with little or no potential for 15N dilution, reduces the chances of measuring robust 29R and 30R ratios.
Fractionation results from the differential responses of stable isotopes in either kinetic reactions, where lighter isotopes tend to react faster, or exchange reactions, where heavy isotopes concentrate where bonds are strongest (Fry, Reference Fry2006). The result is that products have a lighter isotopic composition owing to the fractionation process. The degree of fractionation can be calculated as a fractionation factor. Detailed discussion of fractionation factors is beyond the scope of this article and the reader is directed to other articles (Fry, Reference Fry2006; Sharp, Reference Sharp2007). It has been experimentally shown that the isotopic fractionation during the production of N2O via denitrification may vary with the 15N enrichment of the nitrate substrate; however, the effect of such isotopic fractionation was shown to be negligible if the enrichment of the substrate was greater than 0.6 atom% 15N (Mathieu et al., Reference Mathieu, Lévêque, Hénault, Ambus, Milloux and Andreux2007). It has also been shown that isotopic fractionation does not bias quantifications of gross N transformations in modelling studies when isotopic 15N enrichment is used (Rütting, Reference Rütting2012). Thus, the use of high levels of 15N enrichment avoids any bias owing to isotopic fractionation.
Labelling of plant materials/fertiliser/ruminant excreta with 15N
The 15N content of soil inorganic-N, fertilisers, excreta, plants and gases can all be enriched in order to follow the fate of these compounds in the environment. The degree of enrichment required depends on the material being enriched and the proposed nature of the ensuing experiment. Labelling of dairy cow manure or slurry is generally performed by feeding animals herbage that has been fertilised with 15N-enriched fertilisers or by feeding 15N-enriched urea. Powell et al. (Reference Powell, Wu, Kelling, Cusick and Munoz2004) describe in detail a forage method where feeding of 15N-enriched forage resulted in an increase in the 15N enrichment of the urine, the endogenous N (microbes and microbial products from the rumen, intestine and hind gut plus digestive tract N) and the undigested feed N. Silage components had 15N enrichments of 1.17 to 6.44 and the resulting manure was ∼2.5 atom% 15N. They also describe a urea method where 15N-enriched urea was fed to ruminally fistulated cows that resulted in 15N enrichment of the urine and the endogenous N but not the undigested feed N, as no 15N-enriched feed was provided (Powell et al., Reference Powell, Wu, Kelling, Cusick and Munoz2004). Single dosing with 5 atom% 15N urea produced urine ∼1.25 atom% 15N, whereas repeated 50 g doses at regular intervals produced urine ∼1.25 to 2.15 atom% 15N. Of particular interest in a study by Powell et al. (Reference Powell, Wu, Kelling, Cusick and Munoz2004) are the results showing the time that elapses between administering the 15N-enriched forage or urea and the observed peak in 15N enrichment of the excreta. Lampe et al. (Reference Lampe, Dittert, Sattelmacher, Wachendorf, Loges and Taube2006) produced 15N-enriched (0.72 atom%) slurry by feeding steers 15N-enriched hay and maize silage. Yue et al. (Reference Yue, Liao, Ji, Zhang, Zuo and Rong2012) generated a more highly 15N-enriched manure by letting the animal's gut empty out for 2 days and then feeding mixed silage (26.6 atom% 15N), resulting in cattle manure with 8.0 atom% 15N. Manures of other animal species (sheep and pigs) have been similarly labelled (Sorensen and Thomsen, Reference Sorensen and Thomsen2005; Bosshard et al., Reference Bosshard, Oberson, Leinweber, Jandl, Knicker, Wettstein, Kreuzer and Frossard2011).
The choice of method for 15N-labelling excreta depends on the intention of the experiment. Obviously, 15N labelling of animal excreta is expensive and labour intensive, with due attention needed to be given to animal ethics. However, it provides a genuine urine or faeces matrix that is 15N labelled. In the case of urine, it avoids the researcher having to choose between synthetic urine mixtures where potential bias in N2O fluxes may occur as a result (Kool et al., Reference Kool, Hoffland, Abrahamse and van Groenigen2006), or having to further dope collected unenriched urine with 15N urea to generate 15N enrichment in the collected urine. As seen from these results (Powell et al., Reference Powell, Wu, Kelling, Cusick and Munoz2004), the degree of 15N enrichment is too low to follow evolution of N2 fluxes, which is ideally 40 to 60 atom% 15N, but it is sufficient to allow the detection of 15N-enriched N2O to partition the N2O sources. An example of a study that used fresh urine labelled with 15N urea is that of Taghizadeh-Toosi et al. (Reference Taghizadeh-Toosi, Clough, Condron, Sherlock, Anderson and Craigie2011) who, upon applying the 15N-enriched urine to pasture, were able to partition N2O sources and show the presence of biochar-mitigated urine-derived N2O emissions, with the contribution of urine to N2O emissions lower in the presence of biochar.
The recent work by Jost et al. (Reference Jost, Joergensen and Sundrum2013) is highly relevant when choosing a method to look at manure contributions to N2O emissions. Jost et al. (Reference Jost, Joergensen and Sundrum2013) found that total N2O emissions were correlated with faecal microbial biomass N, showing the significance of including the endogenous N component. This implies that the forage method would be best, if the rationale for the experiment was to determine N fluxes from faeces (e.g. N2O). Others have collected manure and only labelled inorganic-N pools of the manure by adding 15N-enriched inorganic-N salts, for example, (Paul and Beauchamp, Reference Paul and Beauchamp1995) to trace N cycling of the inorganic-N.
Studies using 15N-labelled manures have been used to trace manure effects on N cycling in soils and agronomic effects (Berntsen et al., Reference Berntsen, Petersen, Sorensen and Olesen2007; Bosshard et al., Reference Bosshard, Sorensen, Frossard, Dubois, Mader, Nanzer and Oberson2009; Bosshard et al., Reference Bosshard, Oberson, Leinweber, Jandl, Knicker, Wettstein, Kreuzer and Frossard2011), but the number of studies that have included measures of N2O and its enrichment following manure, slurry or digestate applications are relatively few (Hauck et al., Reference Hauck, Meisinger and Mulvaney1994; Dittert et al., Reference Dittert, Bol, King, Chadwick and Hatch2001; Lampe et al., Reference Lampe, Dittert, Sattelmacher, Wachendorf, Loges and Taube2006; Schouten et al., Reference Schouten, van Groenigen, Oenema and Cayuela2012). The study by Dittert et al. (Reference Dittert, Bol, King, Chadwick and Hatch2001) is a good example of a study where 15N tracing was used to demonstrate the potential of a nitrification inhibitor (3,4-dimethylpyrazole phosphate) to reduce N2O emissions from slurry injected into soil. In this instance, the dairy slurry was 15N enriched and the isotopic composition of the soil and N2O pools were monitored. Besides having lower N2O emissions, the 15N enrichment of the N2O emitted was lower in the nitrification inhibitor treatment, indicating that less N2O was derived from the slurry in this treatment. Further such studies are urgently required to assess management and mitigation strategies for reducing N2O emissions. For example, the call to reduce excess N in the diet of the ruminants or the manipulation of the ruminant diet with feeds varying in C : N ratios will have implications for N2O fluxes from faeces that should be ascertained.
In comparison with generating 15N-labelled excreta, the production of 15N-enriched plant residues is relatively straightforward. Plants are grown with appropriate nutrition and water (avoiding leaching events), generally in a sandy matrix to avoid 14N mineralisation from the soil organic-N pool diluting the 15N pool, with N nutrition provided by adding a 15N-enriched fertiliser. Again urea is best avoided so that 15N is not lost unnecessarily owing to NH3 volatilisation. Once the plant is at the required stage of growth, it may be harvested and used in 15N-tracer experiments. Previous results using 15N to apportion N2O sources have shown that emissions from plant residue applications can be short-lived (Frimpong and Baggs, Reference Frimpong and Baggs2010; Frimpong et al., Reference Frimpong, Yawson, Baggs and Agyarko2011). Ruminant grazing of pasture and forage crops causes fresh litterfall, as animals fail to ingest all harvested herbage (Lodge et al., Reference Lodge, King and Harden2006; Campanella and Bisigato, Reference Campanella and Bisigato2010; Pal et al., Reference Pal, Clough, Kelliher, van Koten and Sherlock2012). One study, replicating a grazing-induced litterfall event, used 15N tracer to show that fresh litter deposition contributed to the N2O flux (1% of N applied) from the soil surface and enriched the soil inorganic-N pool (Pal et al., Reference Pal, Clough, Kelliher and Sherlock2013). Experiments have also been conducted using 15N-enriched N2 to study the fate of biologically fixed N2. For example, Carter and Ambus (Reference Carter and Ambus2006) showed that easily degradable clover residues (Trifolium sp.) made a minor contribution to N2O fluxes. Other studies have shown that the dynamics of N2O emissions derived from 15N-labelled residues are impacted upon by earthworms (Giannopoulos et al., Reference Giannopoulos, van Groenigen and Pulleman2011). Modelling of 15N studies conducted by Delgado et al. (Reference Delgado, Del Grosso and Ogle2010) suggest that residues should not be treated in the same way as fertilisers in terms of N2O emissions, and they call for more residue studies to examine N2O losses from agroecosystems. This will be done best by using 15N tracing.
The study of fertiliser applications using 15N tracer is perhaps the easiest of the substrates to deal with, as it requires little preparation other than perhaps diluting the acquired 15N-enriched isotope fertiliser to a level of enrichment suitable for the experimental objectives. One consideration is the form to apply the fertiliser in. The easiest but perhaps the least conventional way to do this is to water a fertiliser solution onto trial plots. But again the experimental design and objectives need to be considered.
A role for 15N-enriched N2O
It is well recognised that N2O may be consumed in the soil profile by denitrifiers (Chapuis-Lardy et al., Reference Chapuis-Lardy, Wrage, Metay, Chotte and Bernoux2007). The use of 15N-enriched N2O is an underutilised tool that can increase our understanding of N2O production and fate in agroecosystems. Again, owing to its very nature, the 15N-stable isotope acts as a tracer for the N2O molecule when it is added to the soil. If N2O is sufficiently enriched in 15N and of high enough concentration, it is theoretically possible to observe N2 production, but the large N2 background makes it impractical. However, using soil columns, Clough et al. (Reference Clough, Kelliher, Wang and Sherlock2006) demonstrated that the addition of 15N-enriched N2O, along with an inert tracer gas SF6, could be used to calculate an N2O sink (consumption plus absorption by water), whereas the corresponding decrease in the 15N enrichment between successive soil depths enabled N2O production in the soil profile to be calculated simultaneously, as the N2O diffused through the soil. One reason for the low uptake of 15N-tracer studies directly using N2O is the cost of commercially available 15N-enriched N2O. However, 15N-enriched N2O can be made and collected on a small scale in the laboratory by gently and carefully heating small quantities of ammonium nitrate using an oil bath or muffle furnace (Friedman and Bigeleisen, Reference Friedman and Bigeleisen1950). Further studies with 15N-enriched N2O are required to increase our understanding of the factors affecting N2O : N2 ratios in soils (e.g. carbon supply and soil pH) to enable the design of N2O flux mitigation strategies focused on soil and manure management.
Modelling N transformations, N2O and N2 emissions using 15N-tracer studies
The first models developed using 15N-tracer data focused on determining gross production and consumption of mineral N, on the basis of the exchange between organic and mineral N (Kirkham and Bartholomew, Reference Kirkham and Bartholomew1954).These early models were simple enough to allow the development of straightforward analytical solutions. However, process-specific gross N rates including production of an N species (e.g. NO3−) from various sources can only be quantified with models that are based on more realistic N-transformation concepts, such as those developed by Myrold and Tiedje (Reference Myrold and Tiedje1986) and Barraclough and Puri (Reference Barraclough and Puri1995). The set of simultaneous equations developed is solved using numerical integration with parameters in these models determined by suitable parameter optimisation routines (Mary et al., Reference Mary, Recous and Robin1998). Further developments of more realistic and arguably more complex analysis models utilise parameter optimisation routines that can handle large numbers of parameters, such as Markov Chain Monte Carlo techniques (Muller et al., Reference Muller, Rutting, Kattge, Laughlin and Stevens2007; Ruetting and Mueller, Reference Ruetting and Mueller2007). Currently, N2O fluxes can be described by such models so that source partitioning and rates of N2O consumption/production can be derived (Abbasi and Muller, Reference Abbasi and Muller2011), and the microbial processes responsible for the observed 15N dynamics (e.g. autotrophic v. heterotrophic nitrification or denitrification) can also be determined (Stange and Dohling, Reference Stange and Dohling2005). Useful future developments in these models would be the use of longer time periods and the use of 15N-labelled substrate pools to realistically mimic excreta or slurry deposition. To advance our understanding of N-transformation processes related to various soil organic N pools, there needs to be more utilisation of 15N labelling in experiments, where various organic N and mineral N pools are 15N labelled. Although the costs of these experiments (15N label and the analysis costs) are relatively high, the data from such studies are essential, if we are to fully understand the role of soil N-transformation processes that produce N2O from animal excreta. Studies with 15N-enriched substrates focusing on inputs other than inorganic N are beginning to appear and provide insights into the effects of substrate additions on gross N dynamics. For example, Nelissen et al. (Reference Nelissen, Rütting, Huygen, Staelens, Ruysschaerta and Boeckx2012) modelled soil mineral N dynamics following the application of 15N-labelled biochar to a soil. With the advances in analytical techniques and analysis models (15N-tracing models), it is now possible to analyse complex system dynamics. It is mainly the costs associated with 15N-tracing studies that may prevent further large-scale experimental work. Furthermore, analytical challenges are still to be solved, such as the development of reliable field methods to quantify for instance the N2/N2O ratios.
Utilising 15N-enriched tracers to understand microbial contributions to N2O dynamics
The predominant biological processes in soils forming N2O predominantly include nitrification, nitrifier denitrification, nitrification-coupled denitrification and denitrification (Wrage et al., Reference Wrage, Velthof, van Beusichem and Oenema2001; Kool et al., Reference Kool, Dolfing, Wrage and Van Groenigen2011b). The use of 15N on its own cannot differentiate between the inorganic-N sources contributing to N2O fluxes or determine the significance of individual processes to soil-derived N2O emissions. Thus, Wrage et al. (Reference Wrage, van Groenigen, Oenema and Baggs2005) devised a novel dual isotope method (15N and 18O (oxygen)) to assign N2O production to these processes. The method assumed (a) no preferential removal of 18O or 16O during nitrifier denitrification or denitrification, (b) the 18O signature of the applied 18O-labelled water would remain constant over the experimental period and (c) exchange of O between H218O and NO3− would be negligible. Following the application of N to a silt loam soil at 50% water-filled pore space, the assumptions were validated. Wrage et al. (Reference Wrage, van Groenigen, Oenema and Baggs2005) showed that nitrifier denitrification is a significant source of N2O in soil. However, the assumption that there was negligible exchange of O between H218O and NO3− was subsequently proven to be violated (Kool et al., Reference Kool, Wrage, Oenema, Dolfing and Van Groenigen2007; Kool et al., Reference Kool, Muller, Wrage, Oenema and Van Groenigen2009a and Reference Kool, Wrage, Oenema, Harris and Van Groenigen2009b). Thus, the dual isotope method was revised by introducing an additional 18O-labelled NO3− treatment so that O exchange during denitrification could be accounted for (Kool et al., 2010 and Reference Kool, Wrage, Zechmeister-Boltenstern, Pfeffer, Brus, Oenema and Van Groenigen2011a), and it was subsequently shown that nitrifier denitrification made a significant contribution to the N2O fluxes in a number of soils examined. This method holds great promise for furthering our understanding of the role of biological processes in producing N2O and needs to be applied across a wider range of soils and agroecosystems, in particular.
One of the most intriguing and exciting isotopic developments in recent years has been that of stable isotope probing (SIP) of nucleic acids. The method relies on microorganisms assimilating significant quantities of the isotope concerned. This has been used successfully to trace uptake of 13C-labelled compounds into the DNA or RNA of soil microorganism (Radajewski et al., Reference Radajewski, McDonald and Murrell2003). Previously uncultivated N2 fixers, which assimilate N, have been identified using this method with 15N (Buckley et al., Reference Buckley, Huangyutitham, Hsu and Nelson2007). However, the direct use of SIP to identify dissimilatory organisms and/or conditions that promote N2O production/consumption has generally been limited to conditions where denitrification is optimal and where 13C compounds are dosed to identify organisms operating in the denitrifying conditions. For example, a study by Ishii et al. (Reference Ishii, Ohno, Tsuboi, Otsuka and Senoo2011) supplied 13C-labelled succinate with and without N2O to determine what microbes were undertaking N2O consumption in rice paddy soils. This examination of denitrifiers by 13C-proxy, under denitrifying conditions, in the absence or presence of substrates has merit, but it should also utilise 15N-labelled N substrate to further strengthen the findings (fate or change in 15N substrate would provide information on denitrification activity), and with emphasis placed on RNA–SIP under such conditions, as RNA provides information on active microorganisms. It is not yet understood how individual nitrification or denitrification genotypes affect N2O production (Braker and Conrad, Reference Braker and Conrad2011). This methodology holds much promise and needs to be applied widely to excretal and fertiliser inputs in agroecosystems, so we can determine the key microbes and their function as it relates to nitrification and nitrification processes.
Natural abundance studies
Natural abundance studies utilise the naturally occurring isotopic composition of the molecule in question and report the abundance of the atom concerned in delta notation (δ) in units of ‰:
$$\delta x\: = \:\left( {\frac{{{{R}_{{\rm{sample}}}}}}{{{{R}_{{\rm{standard}}}}}}\:{\rm{ - }}\:{\rm{1}}} \right)\:\times \:{\rm{1000}}$$
where δx is the value of the heavy isotope in the sample (R sample) relative to a standard (R standard), where the international standard for δ 15N is N2 in air. The measurement of molecules at natural abundance may also be carried out using IRMS (see above), infrared or laser spectroscopy (as noted below).
An area where natural abundance of N isotopes has come into play with respect to N2O is in the field of isotopomers. The N2O molecule is linear and when 16O is the oxygen isotope in the molecule it takes the form 14N15N16O, 15N14N16O or 14N14N16O. The intramolecular distribution of 15N at the central (α) or end (β) positions of the molecule are assessed by studying the abundance of 15N in the molecule. Using δ-notation, δ 15N denotes the difference in the 15N/14N ratio with respect to a standard, usually atmospheric N2 (Mohn et al., Reference Mohn, Tuzson, Manninen, Yoshida, Toyoda, Brand and Emmenegger2012). The relative difference in the ratios of 14N15N16O to 14N14N16O and 15N14N16O to 14N14N16O are denoted δ 15Nα and δ 15Nβ, respectively, whereas bulk value, δ 15Nbulk = (δ 15Nα + δ 15Nβ)/2 (Mohn et al., Reference Mohn, Tuzson, Manninen, Yoshida, Toyoda, Brand and Emmenegger2012). The value for δ 15Nbulk in the troposphere is reported to range from 6.3 to 6.7‰ depending on location and time of sampling (Mohn et al., Reference Mohn, Tuzson, Manninen, Yoshida, Toyoda, Brand and Emmenegger2012). Another important piece of data able to be derived from the isotopomer measurement is the site preference (SP = δ 15Nα – δ 15Nβ) of the N2O molecule. This is deemed independent of the isotopic composition of the substrate the N2O molecule derives from and supplies process information (Mohn et al., Reference Mohn, Tuzson, Manninen, Yoshida, Toyoda, Brand and Emmenegger2012).
Before development of instrumentation for measuring δ 15Nα and δ 15Nβ, only the average 15N value of the N2O molecule was determined (δ 15Nbulk). Isotopomers of N2O can now be measured using FTIR spectroscopy (Griffith et al., Reference Griffith, Parkes, Haverd, Paton-Walsh and Wilson2009), tuneable diode laser absorption spectroscopy (Pattey et al., Reference Pattey, Strachan, Desjardins, Edwards, Dow and MacPherson2006), gas chromatography-IRMS (Toyoda and Yoshida, Reference Toyoda and Yoshida1999; Rockmann et al., Reference Rockmann, Kaiser, Brenninkmeijer and Brand2003; Kaiser et al., Reference Kaiser, Park, Boering, Brenninkmeijer, Hilkert and Rockmann2004) and more recently quantum cascade laser absorption spectroscopy cavity-ringdown (QCLAS) spectroscopy utilising mid-infrared lasers (Mohn et al., Reference Mohn, Tuzson, Manninen, Yoshida, Toyoda, Brand and Emmenegger2012).
It is well recognised that N2O isotopomers can be used to constrain the atmospheric N2O budget and they confirm that the increase in atmospheric N2O is a result of anthropogenic perturbation of the N cycle (Yoshida and Toyoda, Reference Yoshida and Toyoda2000; Park et al., Reference Park, Croteau, Boering, Etheridge, Ferretti, Fraser, Kim, Krummel, Langenfelds, van Ommen, Steele and Trudinger2012). Initial results examining nitrification and denitrification processes showed that different groups of organisms produced differing isotopomer signatures, and that SP values of 33‰ and ∼0‰ were characteristic of nitrification and denitrification, respectively (Sutka et al., Reference Sutka, Ostrom, Ostrom, Gandhi and Breznak2003; Sutka et al., Reference Sutka, Ostrom, Ostrom, Breznak, Gandhi, Pitt and Li2006).
Isotopomer science is in its infancy in the context of examining N2O fluxes and sources from agroecosystems. The isotopomer analyses of N2O have been applied to examine the effects of various treatments on N2O production and consumption including: biogas residue application to soil (Koster et al., Reference Koster, Cardenas, Senbayram, Bol, Well, Butler, Muhling and Dittert2011), cropping soils receiving organic and synthetic fertilisers (Toyoda et al., Reference Toyoda, Yano, Nishimura, Akiyama, Hayakawa, Koba, Sudo, Yagi, Makabe, Tobari, Ogawa, Ohkouchi, Yamada and Yoshida2011), comparisons of tropical forest and cropping soils (Park et al., Reference Park, Pérez, Boering, Trumbore, Gil, Marquina and Tyler2011), soil moisture conditions (Well et al., Reference Well, Kurganova, de Gerenyu and Flessa2006; Jinuntuya-Nortman et al., Reference Jinuntuya-Nortman, Sutka, Ostrom, Gandhi and Ostrom2008; Bergstermann et al., Reference Bergstermann, Cardenas, Bol, Gilliam, Goulding, Meijide, Scholefield, Vallejo and Well2011), composting (Maeda et al., Reference Maeda, Toyoda, Shimojima, Osada, Hanajima, Morioka and Yoshida2010), stimulated soil denitrification using glucose (Meijide et al., Reference Meijide, Cardenas, Bol, Bergstermann, Goulding, Well, Vallejo and Scholefield2010), cultivation of temperate grassland (Ostrom et al., Reference Ostrom, Sutka, Ostrom, Grandy, Huizinga, Gandhi, von Fischer and Robertson2010), microbial processes (Bol et al., Reference Bol, Toyoda, Yamulki, Hawkins, Cardenas and Yoshida2003; Toyoda et al., Reference Toyoda, Mutobe, Yanmagishi, Yoshida and Tanji2005; Perez et al., Reference Perez, Garcia-Montiel, Trumbore, Tyler, De Camargo, Moreira, Piccolo and Cerri2006; Sutka et al., Reference Sutka, Ostrom, Ostrom, Breznak, Gandhi, Pitt and Li2006; Well et al., Reference Well, Flessa, Xing, Ju and Romheld2008), the effect of ruminant diet on subsequent slurry N2O fluxes (Cardenas et al., Reference Cardenas, Chadwick, Scholefield, Fychan, Marley, Jones, Bol, Well and Vallejo2007) and ruminant urine (Yamulki et al., Reference Yamulki, Toyoda, Yoshida, Veldkamp, Grant and Bol2001). Most of these early studies were of short duration and had limited temporal sampling. The study by Park's et al. (Reference Park, Pérez, Boering, Trumbore, Gil, Marquina and Tyler2011) suggested that the δ 15Nbulk data could be used for distinguishing N2O fluxes from fertilised and natural ‘background’ fluxes, and that the SP δ 15N results could be used to differentiate between consumption and production of N2O by microbial pathways. Enticingly, the use of QCLAS has been shown to be capable of continuous analysis of N2O isotopomers with identification of N2O source processes possible (Mohn et al., Reference Mohn, Tuzson, Manninen, Yoshida, Toyoda, Brand and Emmenegger2012). This technology also needs to be deployed across agroecosystems to help understand not only excreta sources of N2O, and temporal dynamics, but also to explore and demonstrate the success of mitigation options.
Conclusion
The impending increase in tropospheric N2O emissions as a result of existing and projected increases in anthropogenic animal production systems demands mitigation options. These can only be implemented if N2O emission sources and their temporal dynamics can be traced in conjunction with N2O fate. The stable isotopes of N and associated methodologies provide the tools to achieve this tracing. More 15N-tracer studies are needed to ascertain soil and excreta contributions to N2O dynamics. Relatively new research fronts using SIP and isotopomers of N2O offer exciting potential as diagnostic tools to evaluate effects and mitigation success. Collaborations between microbiologists, animal production specialists and soil scientists will bring much needed synergies to address the N2O issue.
Acknowledgements
This paper was published as part of a supplement to animal, publication of which was supported by the Greenhouse Gases & Animal Agriculture Conference 2013. The papers included in this supplement were invited by the Guest Editors and have undergone the standard journal formal review process. They may be cited. The Guest Editors appointed to this supplement are R. J. Dewhurst, D. R. Chadwick, E. Charmley, N. M. Holden, D. A. Kenny, G. Lanigan, D. Moran, C. J. Newbold, P. O'Kiely, and T. Yan. The Guest Editors declare no conflict of interest.
