Ammonia oxidisers and their inhibition to reduce nitrogen losses in grazed grassland: a review

ABSTRACT

Pastoral agriculture is a major source for nitrate () contamination in surface and ground waters and for the greenhouse gas emissions in New Zealand. Advances have been made in recent years in understanding the role of different ammonia oxidisers, including ammonia oxidising bacteria (AOB) and ammonia oxidising archaea (AOA) in nitrification, and in developing nitrification inhibitor (NI) mitigation technologies. Results showed that, in the N-rich soil environment under the animal urine patches in grazed grassland, AOB are the dominant microbes responsible for ammonia oxidation whereas AOA play a less important role. A number of laboratory and field studies have demonstrated conclusively that treating grazed pasture soils with a nitrification inhibitor (NI), such as dicyandiamide (DCD), which inhibits the growth and activity of AOB, is an effective means of reducing leaching and emissions.

Professor Hong Di.

KEYWORDS:

• Ammonia oxidising archaea
• ammonia oxidising bacteria
• grazed pastures
• nitrate leaching
• nitrification inhibitor
• nitrous oxide emissions
• soil

Introduction

Nitrate () is a major water contaminant and leaching from livestock production systems is a worldwide environmental problem. The livestock sector is responsible for 18% of global greenhouse gas emissions (CO₂ equivalent) and accounts for 65% of the anthropogenic nitrous oxide (N₂O) emissions (IPCC 2007). Nitrous oxide is a potent greenhouse gas with a long-term global warming potential 265–298 times that of carbon dioxide (IPCC 2007; Myhre et al. 2013). In New Zealand, pastoral agriculture is the dominant land use, where animals graze outdoor pastures throughout the year. In such grazed grassland systems, the dominant source for both leaching and N₂O emissions is animal excreta, particularly urine, deposited by the animal during grazing. As animals graze the pastures, between 70%–90% of the nitrogen (N) ingested is returned to pasture and about 80% of the N is in the urine (Haynes & Williams 1993; Jarvis et al. 1995; Di & Cameron 2002a; Di et al. 2002; Cameron et al. 2013; Selbie et al. 2014a, 2014b). For example, the N loading rate under a dairy cattle urine patch can be as high as 700–1000 kg N ha⁻¹. This high rate of N deposited in a single urination is well in excess of that which can be utilised by plants. The surplus N is prone to leaching after it is converted to by nitrification (e.g. Scholefield et al. 1993; Ledgard et al. 1999; Silva et al. 1999; Di & Cameron 2000, 2002b), or lost as NO, N₂O or N₂ gases, from nitrification or denitrification (Allen et al. 1996; Di & Cameron 2003, 2006; Luo et al. 2007) (Figure 1).

Figure 1. Simplified diagram showing animal urine-N as the main source for N₂O emissions and leaching losses in a grazed grassland, and the point of intervention by a nitrification inhibitor (from Di & Cameron 2016).

Most of the N in the animal urine is urea which, when deposited in the soil, is quickly hydrolised to produce ammonium ():(1) Most of the ammonium () is adsorbed onto the negatively charged soil cation exchange surfaces. Therefore, leaching is usually negligible in most soils with a good cation exchange capacity (CEC).

However, in the soil, is rapidly oxidised, first to hydroxylamine, then to NO₂⁻ and then to by the process known as nitrification:(2) (3) (4)

Nitrate is an anion and is not retained by the negatively charged soil particles. Therefore, when drainage occurs it is leached out of the soil into groundwater or surface waters, contributing to water contamination.

Nitrous oxide is also produced from nitrification as a byproduct (Wrage et al. 2001). In addition to leaching, the produced from nitrification is subject to denitrification, where is reduced step-wise to , NO, N₂O and N₂, with N₂O as an intermediate product:(5)

If a nitrification inhibitor (NI) is applied to the soil to inhibit ammonia oxidation, the first step of the nitrification process, then this can lead to lower concentrations of and lower N₂O emissions from both nitrification and denitrification (Figure 1).

Significant advances have been made in recent years in understanding the role of different ammonia oxidisers, including ammonia oxidising bacteria (AOB) and ammonia oxidising archaea (AOA), and in developing mitigating technologies based on the use of NIs. This article provides a summary review of progress in these areas, with an emphasis on grazed grassland in New Zealand.

Ammonia oxidisers

The first and rate-limiting step of the nitrification process is the oxidation of NH₃ to NH₂OH (Figure 1). It is performed by the key ammonia monooxygenase (AMO) enzyme, which is encoded by the subunits of amo genes. For over a century it was thought that this process was mostly performed by the chemolithoautotrophic AOB (Purkhold et al. 2000; Kowalchuk & Stephen 2001; Prosser & Nicol 2008). However, the recent discovery of the amoA gene in archaea populations has significantly revised our understanding (Venter et al. 2004; Francis et al. 2005; Könneke et al. 2005; Wuchter et al. 2006; Prosser & Nicol 2008). AOA populations were found to be more abundant than AOB in a range of soils, suggesting a potentially greater role for AOA than AOB in ammonia oxidation (Leininger et al. 2006; He et al. 2007; Chen et al. 2008; Prosser & Nicol 2008).

Recent advances in ammonia oxidiser research, using molecular biology techniques to target the functional amoA gene, indicate that the population abundance and activity of AOB and AOA in ammonia oxidation vary depending on soil and environmental conditions. In a study of intensively grazed dairy pasture soils in New Zealand, Di et al. (2009a) showed that it was AOB, which grew in response to the application of dairy cattle urine (applied at 1000 kg N ha⁻¹), that provided (Figure 2). The AOB population abundance and activity increased significantly following the animal urine application and were significantly inhibited by the application of a NI, dicyandiamide (DCD). In contrast, neither the AOA population abundance nor activity increased following the application of the animal urine (Di et al. 2009a). The nitrification rate was shown to be significantly related to the AOB amoA gene copy numbers but not to that of AOA. In a separate study, Di et al. (2010a) showed that the AOA communities in fact grew in the control treatment (no N applied) rather than in the urine treatment (Figure 3). The AOA abundance was inhibited by the high dose of urine in some of the treatments. Other studies have validated these results (e.g. Offre et al. 2009; Shen et al. 2011; Dai et al. 2013; Robinson et al. 2014a, 2014b; Hill et al. 2015). These results demonstrate that in the high-N soil environment, such as that under the animal urine patch in intensively grazed grassland, it is AOB, particularly Nitrosospira, that are predominantly performing the NH₃ oxidation. AOA may play a more important role in NH₃ oxidation under low fertility or oligotrophic environments.

Figure 2. AOB amoA gene abundance as affected by the application of dairy cattle urine (1000 kg N ha⁻¹) and the nitrification inhibitor DCD in a dairy pasture soil (from Di et al. 2009a).

Figure 3. AOA amoA gene abundance as affected by the application of animal urine (1000 kg N ha⁻¹) and the nitrification inhibitor DCD (Di et al. 2010b).

Soil pH may also affect the growth and activity of AOB and AOA in the soil (Nicol et al. 2008). Under strongly acidic conditions, if AOA growth is not inhibited by high N concentrations, then AOA may become important in NH₃ oxidation (Nicol et al. 2008; Gubry-Rangin et al. 2010; He et al. 2012; Shen et al. 2012; Zhang et al. 2012; Hu et al. 2014; Robinson et al. 2014a).

Soil moisture content can also affect AOB and AOA growth. When the soil moisture was at 60% field capacity, both AOB and AOA growth was restricted (Di et al. 2014). AOB growth was significantly stimulated by the application of animal urine when the soil moisture content was close to, or above, field capacity (Di et al. 2014). However, AOA only grew in the control (no N applied) and was inhibited by the application of animal urine-N. Therefore, in the wet and heavily trampled winter grazing conditions where there is substantial input of animal urine, it is also AOB that are dominant.

Our understanding of the microbial ecology of AOB and AOA in different soil ecosystems is still limited and further research is needed to improve our knowledge. This knowledge is important for developing management tools, such as NI technologies, for sustainable N management.

Nitrification inhibitors

Although a number of compounds have been shown to have NI properties, only a few have been commercialised, including dicyandiamide (DCD), 3,4-dimethylpyrazole phosphate (DMPP) and nitrapyrin.

DCD inhibits the first-step of the nitrification process, the oxidation of ammonia. In soil studies, it has been shown that an application rate of 8–10 mg DCD kg⁻¹ soil was effective in inhibiting the growth and activity of Nitrosospira (Di & Cameron 2004a; Di et al. 2009a, 2010a). DCD is bacteriostatic and thus does not kill the bacteria. It temporarily inhibits the growth and activity of ammonia oxidisers. DCD is completely degraded in the soil, first to urea and then to CO₂ and NH₃, leaving no other long-term residues (Amberger 1989). Most importantly, DCD is a very safe compound, with an oral LD₅₀ being greater than 30,000 mg kg⁻¹ body weight of female rats (more than 10 times that of table salt at 3000 mg kg⁻¹ body weight), and no adverse reproductive/developmental effects, genotoxicity and carcinogenicity have been reported (OECD 2003). Therefore, the OECD (2003, p. 4) concluded that ‘this chemical is currently of low priority for further work because of its low hazard potential’. Because of the widespread use of DCD in plastics that could be in contact with food, the European Commission has established a tolerable daily intake (TDI) for DCD at 1 mg DCD kg⁻¹ body weight as a food standard guideline (European Commission 1995).

The inhibition action of DCD is also shown to be specific to ammonia oxidisers, and its use in the soil does not significantly affect the soil microbial biomass (Di & Cameron 2004a), methanotrophs (Di et al. 2011) or other agrobacteria (O’Callaghan et al. 2010), or other important enzymatic activities that are important for N cycling (Guo et al. 2013).

DMPP is a relatively new NI and has been through extensive toxicology and ecotoxicology tests (Zerulla et al. 2001). Its oral lethal dose (LD₅₀) in rats ranges from 300 to 2000 mg kg⁻¹ body weight. An application rate of 0.5–1.5 kg ha⁻¹ is sufficient to provide effective inhibition. It has been tested for use in fertiliser formulations or mixing with slurries to mitigate NO₃⁻ leaching and N₂O emissions (e.g. Barth et al. 2001; Pasda et al. 2001; Weiske et al. 2001; Hatch et al. 2005). Its performance in grazed grassland to treat urine patches in a liquid form was found to be comparable with DCD (Di & Cameron 2011, 2012).

Efficacy of nitrification inhibitors to reduce nitrate leaching

Di & Cameron (2002c) showed that treating soil with DCD can reduce NO₃⁻ leaching by an average of 59%. A treatment technology was developed in New Zealand where DCD was applied in a solution form to cover the entire soil surface of a grazed grassland field so that as much of the AOB in the surface soil was treated and inhibited. The DCD was sprayed on to pastures shortly after grazing, before the urine-N was nitrified.

In many parts of New Zealand, leaching mainly occurs over the late autumn, winter and early spring period (e.g. from April to September). A study showed that during winter, when temperatures are low, the DCD effect lasted about 3 months (Di & Cameron 2004a). Therefore two applications of DCD, once in the autumn and once in the spring were sufficient to significantly inhibit AOB and decrease leaching losses (Di & Cameron 2004b; Di et al. 2010b; Podolyan et al. 2014).

A series of studies were conducted to determine the optimum rate of DCD that was both effective and cost-effective (Di & Cameron 2002c, 2004a, 2004b, 2005; Di et al. 2009b). These studies showed that DCD applied at 10 kg DCD ha⁻¹ was effective to give a significant reduction in NO₃⁻ leaching when two DCD applications were made between autumn and spring (Figure 4). Molecular microbiology studies have shown that the 10 kg DCD ha⁻¹ application was equivalent to the ED₅₀ where it gave about 50% reduction in AOB growth and in nitrification rate (Guo et al. 2014).

Figure 4. Example breakthrough curve showing the effect of DCD applied at 10 kg DCD ha⁻¹ on concentration from animal urine applied at 1000 kg N ha⁻¹ (from Di et al. 2009b).

In order to reduce the volume of liquid needed, a fine particle spray formulation was developed where DCD was partly dissolved and partly in the form of fine particles (< 100 µn). This fine particle suspension spray was shown to be highly effective in reducing leaching losses (Di & Cameron 2004b, 2005).

The NI technology was not only applicable to grazed dairy pastures where the urine-N loading rate was high (about 700–1000 kg N ha⁻¹), but also to beef cattle and sheep grazing systems where the urine-N loading rates were lower, around 500–700 and 300 kg N ha⁻¹, respectively (Di & Cameron 2007; Moir et al. 2010).

The technology has subsequently been evaluated for its potential to mitigate leaching in a range of soil and environmental conditions in New Zealand and other countries (Di & Cameron 2016). The lysimeter studies which targeted the animal urine patch showed that DCD reduced leaching by an average of about 50% (± 3 SEM). The suction cup and drainage plot studies provided an integrated measurement of the effect of DCD on urine and non-urine patches and showed a reduction of about 30% (± 6 SEM) (Di & Cameron 2016).

Efficacy of nitrification inhibitors to reduce nitrous oxide emissions

The treatment of grazed pasture soil with DCD has an additional benefit of reducing direct N₂O emissions from urine patches (Di & Cameron 2002c, 2003, 2006) (Figure 5). DCD decreased N₂O emissions from both the nitrification and denitrification processes (Di & Cameron 2008), and the reductions in N₂O emissions were largely the result of the inhibition of AOB by DCD in the high-N status urine patch soils (Di et al. 2010b). There was a positive relationship between N₂O emissions and AOB amoA gene abundance, but not with that of the AOA amoA gene abundance (Di et al. 2010b). The efficacy of the DCD NI technology in reducing N₂O emissions in grazed pastures has been assessed extensively under a range of soil and environmental conditions (Di & Cameron 2002c, 2003, 2006, 2008; Di et al. 2007, 2010b; Hoogendoorn et al. 2008; Ledgard et al. 2008, 2014; Smith et al. 2008a; Singh et al. 2009; Qiu et al. 2010; Zaman et al. 2010; de Klein et al. 2011, 2014; Ball et al. 2012; Dai et al. 2013; Cameron et al. 2014; Kim et al. 2014; Hu et al. 2015). These trials showed that DCD decreased N₂O emissions from animal urine-N by an average of 57% (± 2 SEM) (Di & Cameron 2016). Based on these results, the NI technology was successfully incorporated into New Zealand’s greenhouse gas emissions inventory (Clough et al. 2007).

Figure 5. Daily N₂O emissions from a Lismore soil from dairy cattle urine with and without DCD (10 kg DCD ha⁻¹) showing the effect of DCD on N₂O emissions (after Di & Cameron 2006).

The NI technology has also been shown in Canterbury to be effective in winter forage crop grazing systems where the soils are usually wet and heavily trampled by the grazing animal (Ball et al. 2012; Di et al. 2014). DCD applied nearly 3 weeks after winter grazing was still effective in reducing N₂O emissions (Di et al. 2007).

The DCD effect did not diminish after repeated use, as DCD was equally effective in reducing N₂O emissions after 4–5 years of consecutive use (de Klein et al. 2011).

Effect of nitrification inhibitor on pasture production

The treatment of grazed pasture soil with a nitrification inhibitor such as DCD can also increase N use efficiency and pasture production (e.g. Di & Cameron 2002c, 2004b, 2005, 2007; Moir et al. 2007, 2010; Carey et al. 2012). Di & Cameron (2007) reported an average increase in N use efficiency of 32% and an average herbage dry matter yield increase of 25% when DCD was applied to urine patches. Moir et al. (2007) reported a 21% increase in pasture dry matter following DCD application to an intensively grazed dairy pasture soil. Similarly, Carey et al. (2012) reported a 19% increase in dry matter yield from 132 split paddock trials on dairy farms across New Zealand.

However, reported pasture responses to DCD application were more variable than the reported effects on leaching and N₂O emissions (e.g. Monaghan et al. 2009; O’Connor et al. 2012; Cameron et al. 2014; de Klein et al. 2014; Kim et al. 2014; Ledgard et al. 2014). This is not surprising in view of the fact that pasture growth can be affected by many factors, including the supply of N and other nutrients, soil moisture status, temperature and other soil physical conditions. In addition, the random and uneven distribution of animal urine patches also complicates the pasture measurements (Cameron et al. 2014).

Other benefits of nitrification inhibitor use

The reduction in leaching as a result of NI application to grazed pastures has also been shown to lead to lower leaching losses of counter cations (Di & Cameron 2004b, 2004c, 2005). As shown in Figure 6, the Ca²⁺, Mg²⁺ and K⁺ leaching losses were positively correlated to concentration. Due to decreased leaching by the use of DCD, Ca²⁺, Mg²⁺ and K⁺ losses were correspondingly decreased by 50%, 52% and 65%, respectively (Di & Cameron 2004c). The reduction in leaching of these valuable macronutrients would reduce the need for application of these cation fertilisers where DCD is applied.

Figure 6. Relationship between the concentrations of and cations (Ca²⁺, Mg²⁺ and K⁺) in the leachate from the lysimeters (from Di & Cameron 2005).

The NI technology can reduce the risk of animal poisoning by high concentrations in pasture. Nitrate concentrations in pasture in the range of 1500–2000 mg dry matter are considered risky for the grazing animal (Burgemeister 2003). The concentration can be particularly high in the urine patch areas of pastures due to the high N-loading rate. The application of DCD can decrease concentrations in pasture to below 800 mg dry matter (safe for animals) (Moir et al. 2012). Therefore, DCD can increase pasture yield and improve pasture quality by decreasing concentration in the pasture, without compromising the other pasture-quality indicators (Moir et al. 2007).

Factors that affect efficacy of nitrification inhibitors

Nitrification inhibitor properties

In addition to DCD, DMPP has also been evaluated for its efficacy. Di & Cameron (2011) found that DCD and DMPP, when applied at the same rate of 10 kg ha⁻¹, both in liquid formulations, were equally effective in inhibiting the growth of AOB. The cost of DMPP was significantly higher than that of DCD, so lower rates of DMPP were also assessed. It was shown that DMPP applied at the rate of 1 kg ha⁻¹ was as effective at reducing leaching as DCD applied at 10 kg ha⁻¹ (Di & Cameron 2012). The 1 kg DMPP ha⁻¹ rate is close to the recommended rate for fertiliser formulations (Zerulla et al. 2001).

Temperature

Nitrification inhibitors such as DCD and DMPP are biodegradable in the soil and their longevity (as indicated by half-lives) and efficacy are temperature dependent. At 8°C (simulating winter temperatures in many parts of New Zealand) the half-life of DCD was about 110 days (Di & Cameron 2004a). However, when the temperature was at 20°C (simulating summer soil temperatures in New Zealand) the half-life was decreased to about 20 days. Gillingham et al. (2012) summarised results from a series of field trials across New Zealand and found that the residence time ranged from 42 to 84 days in the North Island (warmer) and 40 to 160 days in the South Island (cooler). Qiu et al. (2010) reported that the amount of N₂O emitted (12.7 kg N₂O–N ha⁻¹) during the winter season was 1.6 times that from the summer season (7.8 kg N₂O–N ha⁻¹) and DCD decreased the N₂O emissions by 69% and 40% for the winter and summer season, respectively. These results demonstrated the higher N₂O emissions potential during the wet and cold winter period compared with the warm and dry summer period, and the higher efficacy of DCD during the winter compared with summer.

Soil properties

The reduction of leaching by NIs relies on the soil exchange complex to retain from the animal urine or N fertiliser. Therefore an adequate cation exchange capacity (CEC) of the soil is important. However, Cameron et al. (2007) reported that the use of DCD to treat animal urine patches was able to reduce leaching by 23%–32% even in a very free-draining pumice soil with low CEC. Soil organic matter becomes important in increasing the CEC and retaining NH₄⁺ in course-textured soils.

Soil compaction due to animal trampling in grazed pastures did not significantly affect the efficacy of DCD in reducing N₂O emissions (Ball et al. 2012) or leaching (Hill et al. 2015). DCD was also found to be highly effective in decreasing N₂O emissions in extremely wet soils, such as at 130% field capacity (Di et al. 2014).

Rainfall

DCD is water soluble and therefore may move through the soil under very high rainfall conditions (Shepherd et al. 2012). However, a study by Di et al. (2009b) found that DCD reduced NO₃⁻ leaching by 44%–71% under two contrasting rainfall conditions (1260 mm vs 2145 mm) and the efficacy of DCD was not significantly affected by the different rainfall inputs. The importance of rainfall in affecting the efficacy of DCD may also depend on the soil drainage status, and the timing of rainfall in relation to DCD application.

Timing of application

To maximise effectiveness, it was found that the NI was most effective if applied shortly after grazing (i.e. when the pastures are short and the urine patches are fresh; and thus before significant nitrification has taken place). It is important to apply the NI when the pasture is short so that the NI can reach the soil. If, for practical reasons, the NI needs to be applied to tall pasture, it is important to ensure that there is sufficient rain or irrigation water (about 10 mm) to wash the NI off the foliage into the soil.

Concluding remarks

Research has clearly shown that AOB and AOA prefer different soil and nitrogen environments to grow and function. In the high-N soil environment of grazed grassland, it is the AOB that are the dominant players in performing ammonia oxidation, whereas AOA may play a role in low-N or strongly acidic soils. Research spanning more than a decade has conclusively demonstrated that treating grazed pasture soils with a NI is an effective technology for NO₃⁻ leaching and N₂O emissions in grazed pastures. The large body of scientific evidence from a number of laboratory and field studies have proven the consistency and reliability of the technology across a wide range of soil and environmental conditions. The NI technology is one of the few environmental technologies that also provide an economic benefit by increasing the N use efficiency and pasture dry matter yield. The most widely evaluated NI, DCD, is one of the safest compounds, and has been shown to pose little or no health risks to either grazing animals or humans (OECD 2003). However, DCD is currently not in commercial use in New Zealand, awaiting the establishment of a standard for food in the Codex Alimentarius by the UN Food and Agriculture Organization (FAO). In view of the huge environmental benefits of this technology, and the low toxicity of DCD, it would seem a logical step to list DCD in the Codex Alimentarius so that these benefits can be captured.

Future research should be directed to:

• The search for new NIs so that a range of such identified compounds may be designed to suit specific soils, environments or production systems.

• Identification of AMO structure to aid the search for new NIs.

• Understanding the variations of pasture response to NI applications.

• Synthesis of research data into process- or management-based models.

Acknowledgements

We would like to thank Dr Barbara Brown for technical support with this review.

Disclosure statement

No potential conflict of interest was reported by the authors.