Nitrogen gaseous emissions from farm effluent application to pastures and mitigation measures to reduce the emissions: a review

Abstract

Potential losses of nitrogen (N) from land application of effluents include ammonia (NH₃) and nitrous oxide (N₂O) emissions. In this review paper, the extent of the NH₃ and N₂O losses resulting from application of effluents to pastoral soils is assessed. Nitrogen losses, as NH₃ and N₂O, from applied effluent to pastoral soil ranged from 1%–66% and <0.1%–6% of the applied N, respectively. The potential mitigation methods include: reducing livestock numbers; lowering N content of the effluent; using N process inhibitors; optimising timing of effluent application; and applying effluent at rates that match plant uptake. It is important to remember that some of these options can result in ‘pollution swapping’ and some of them are in the early stages of development. Future research needs to focus on the overall impact of mitigation options on the whole suite of gaseous emissions and the practicality of those options.

Keywords:

• ammonia
• farm effluent
• manure
• mitigation
• nitrous oxide

Introduction

One of the challenges facing the dairy industry is to optimise a proven effluent management system that can increase nutrient use efficiency and minimise off-farm environmental impacts. Increased animal numbers, greater use of nitrogen (N) fertilisers and higher supplementary feed inputs on dairy farms have markedly changed the volume and types of effluent produced. Different types of effluent (e.g. liquid, solid, fresh or stored manure) with differing concentrations of solids and nutrients are derived from different animal and effluent management systems (Rotz 2004). The effluent produced on farms by off-paddock systems can be classified as farm dairy effluent (FDE), slurries or manures, depending on their solid or dry matter (DM) content (Chadwick et al. 2000; Rotz 2004). Liquid products with a DM content <5%, which are usually derived from the farm dairy milking parlour and associated yards, are termed FDE. Slurries have a DM content of 5%–15% while solid manures contain >15% DM (Longhurst et al. 2012). Both slurries and solid manures are typically derived from animal standoff and housing areas or from effluent solid separation. In this review paper, the general term ‘effluent’ is used to cover all the effluent types. Otherwise, the terms FDE, slurry or manure are used specifically. All these farm effluents contain quantities of valuable plant nutrients that, if correctly applied to land, can improve soil fertility and increase the sustainability of farming systems (Spiehs et al. 2010; Li et al. 2014b).

Environmental concerns from effluent application encompass all aspects of non-point source pollution including air quality (Luo et al. 2008; Gerber et al. 2013). Ammonia volatilisation occurs through the ammonification process and depends on the environmental conditions at the time of effluent application. Nitrous oxide gas is formed in soils through the microbiological processes of nitrification and denitrification. Nitrous oxide production by nitrifying bacteria may arise either during ammonium (NH₄ ⁺) oxidation to nitrite (NO₂ ⁻) or during dissimilatory NO₂ ⁻ reduction when the oxygen (O₂) supply is limited (Saggar et al. 2011; Akiyama et al. 2013; Li et al. 2014b). The complex interactions between soil properties, climatic factors and agricultural practices all affect the extent of these gaseous emissions (Saggar et al. 2013). An understanding of the NH₃ and N₂O emissions from land application of effluents and the development of technologies and strategies to mitigate these emissions is required to reduce nutrient losses and improve nutrient use efficiencies.

This review will investigate the scientific literature regarding NH₃ and N₂O production from effluent application to pastures, the factors controlling these emissions and the major management options available for reducing these losses. The aim is to provide an overview of present knowledge, assess how this information can be used to improve the estimates of emissions from effluent application to pasture and to discuss good practices for reducing these losses.

Ammonia and nitrous oxide emissions from land application of effluent

Ammonia volatilisation

Fresh FDE (containing fresh urine, faeces and washing water) is commonly applied on New Zealand dairy farms. Most urea in effluent is likely to have been hydrolysed to NH₄ ⁺ before application to soil; however, depending on duration of effluent collection and storage, some urea may still remain in applied effluent. Early studies by MacGregor et al. (1979) and O'Toole (1996) suggested that FDE may contain between 3 and 140 mg urea-N L^–1. Urea either in stored or applied FDE is usually rapidly hydrolysed within 1–2 days by urease. This enzyme catalyses the hydrolysis of urea to NH₄ ⁺ and carbamate ions, which decompose to carbon dioxide (CO₂) and NH₃ (Cameron et al. 2013) (Equation 1).

(1)

Ammonia volatilisation occurs as the NH₄ ⁺-N in the applied effluent or soil solution is converted to NH₃ gas, which is subsequently transferred from the soil to the atmosphere (Zaman et al. 2008; Li et al. 2014a).

Ammonia volatilisation patterns and amounts vary greatly. Li et al. (2014a) found that 80% of the total NH₃ volatilisation occurred within the first 5 days of the effluent application, and peaks of the NH₃ fluxes from all types of effluent occurred within 24 hours. Other researchers have found a similar pattern (e.g. Meisinger & Jokela 2000; Sherlock et al. 2002; Panetta et al. 2005; Rochette et al. 2009; Salazar et al. 2014). It has been reported that 1%–66% of applied effluent N is lost as NH₃ (e.g. Sherlock et al. 2002; Petersen & Sommer 2011; Li et al. 2014b), depending on form and DM content of applied effluent, application method, soil and environmental factors.

Nitrous oxide emission

Nitrous oxide gas is formed in soils through the microbiological processes of nitrification (Equation 2) and denitrification (Equation 3) (Schils et al. 2013).

Nitrification is the biological oxidation of soil NH₄ ⁺ to soil NO₂ ⁻ and NO₃ ⁻, producing N₂O as a by-product (Saggar et al. 2013). Denitrification is the stepwise biological reduction of soil NO₃ ⁻ to gaseous N compounds, with N₂O and nitrogen gas (N₂) being the main products. The ratio of N₂O/N₂ produced during denitrification in soil is affected by the presence of NO₃ ⁻, the level of anoxia, pH, temperature, soluble carbon (C) availability and microbial populations (Saggar et al. 2013).

Nitrous oxide emissions are highly dynamic and affected by effluent application time, application method and rainfall or irrigation. Several researchers have found that N₂O emissions peak immediately after application (e.g. Barton & Schipper 2001; Saggar et al. 2004; Bhandral et al. 2009; Sistani et al. 2010; Li et al. 2014b). One possible reason is that liquid from the effluent drives out air trapped in the soil which could contain high levels of N₂O. Schils et al. (2013) suggested that N₂O loss increased immediately after liquid manure injection and was followed by a shift to N₂ emissions. Other studies have shown highly variable patterns of N₂O peak response after effluent application. Whalen et al. (2000) and Sommer & Hutchings (2001) observed that the emissions did not peak until 30 hours after cattle slurry application, although emission peaks were observed much earlier (after 8 hours) during second and third applications. Studies have shown peak emissions 1–7 days after cow manure application (Sherlock et al. 2002). Emissions from applied effluent generally reach the background level within 60 days depending on application rate and climatic conditions (e.g. Luo et al. 2008; Bhandral et al. 2009; Thomsen et al. 2010; Li et al. 2014b). A large range of N₂O emission factors (percentage of applied N lost as N₂O) have been reported after the application of effluent to land, ranging from <0.1%–6% (e.g. Bhandral et al. 2004; Luo et al. 2008; Chadwick et al. 2011; Li et al. 2014b).

Factors affecting ammonia and nitrous oxide emissions

Understanding the factors that affect NH₃ and N₂O emissions will delineate practices to reduce NH₃ and N₂O losses, improve the prediction of these losses, and aid in developing more efficient farm nutrient management plans. Both NH₃ and N₂O emissions are reported to be influenced by effluent characteristics, a number of soil factors and environmental conditions and application managements, as described below. Data found in a search of the literature related to dairy effluent application to pastures were summarised. However, data which we believe have been compromised by variables outside the trial design were excluded.

Effluent characteristics

The effluent DM concentration is an important factor that affects NH₃ emissions from effluent applications (e.g. Smith & Chambers 1995; Li et al. 2014a). The general observation is that as DM increases, the NH₃ losses (percentage of applied N lost as NH₃) increase (Fig. 1). This is likely to be due to the fact that an effluent with higher DM remains on the soil surface, providing increased opportunity for NH₃ contained within the effluent to volatilise to the atmosphere. Conversely, an effluent with lower DM tends to have greater fluidity and, therefore, infiltrates more readily into the soil where NH₄ ⁺ is protected from volatilisation by adsorption on to soil colloids. Another potential reason is that when the effluent infiltrates into the soil, NH₄ ⁺ is nitrified to NO₃ ⁻ more readily due to the higher nitrification rates deeper in the soil than on the surface. However, combined data from the literature do not indicate a significant relationship between NH₃ loss and effluent DM concentration (P > 0.05) (Fig. 1). This is due to other effluent and soil factors that may have influenced the effect of effluent DM on NH₃ losses.

Figure 1 Relationship between ammonia losses from effluent application (% of applied N lost as NH₃) and effluent dry matter content (from Lockyer et al. 1989; Sommer & Olsen 1991; Li et al. 2014a).

The amount of mineral-N contained within different effluents depends on the type of effluent, storage duration and how it has been managed during this time (Rotz 2004). The rate and amount of NH₃ loss can be directly related to concentrations of NH₄ ⁺-N in the applied effluent (e.g. Rochette et al. 2009; Li et al. 2014b). Analysis of data from the literature reveals that there is a significant exponential relationship between NH₃ loss and effluent NH₄ ⁺-N concentration in effluents (R² = 0.515, P < 0.05) (Fig. 2).

Figure 2 Relationship between ammonia losses from effluent application (% of applied N lost as NH₃) and effluent NH₄ ⁺-N concentration (from Beauchamp et al. 1982; Lockyer et al. 1989; Sommer & Olesen 1991; Li et al. 2014a).

Some other factors, such as pH and effluent C concentration, also affect NH₃ emissions. Meisinger & Jokela (2000) found that lowering the effluent pH below 7 before it is applied can decrease NH₃ volatilisation. This is because the equilibrium between NH₃ and NH₄ ⁺ is affected by pH, with more N in the gaseous NH₃ form at higher pH. There is little information in the literature describing the effect of effluent C on NH₃ emissions. However, it is likely that a high effluent C concentration would lead to immobilisation of N and therefore less availability of NH₃ for volatilisation. This effect may be obscured by other factors such as the fact that a high C effluent is also likely to have a higher NH₄ ⁺ content.

Effluent characteristics, such as C, N and DM, can all affect N₂O emissions and, potentially, emission factors. As discussed in the following paragraphs, data can be found in the literature that imply relationships between N₂O emission factors and individual effluent components (Figs 3–5); however, there are no reports of studies where one factor in the applied effluent is altered, but all other factors remain constant. Therefore, these relationships need to be interpreted with caution.

Addition of effluent C to soil not only stimulates microbial growth and respiration but also provides the organic C needed by the soil denitrifiers (Cameron et al. 2013). Pelster et al. (2012) suggested that the greater C content of dairy effluents is the reason for the larger N₂O emissions when compared with pig manure. Barton & Schipper (2001) also suggested that higher N₂O emissions from dairy effluent than from inorganic N fertiliser were due to enhanced denitrification activity by increasing C availability and/or by decreasing soil aeration following increased respiration. Analysis of the data from the literature suggests that there is a significant exponential relationship between N₂O emission factors and effluent C concentration (R² = 0.082, P < 0.05) (Fig. 3). The increase may not be entirely due to increasing C content, as there is typically a corresponding increase in total N. Thus, the increase in emission factors may be due to other factors associated with the effluent C increase.

Figure 3 Relationship between nitrous oxide emission factors from effluent application (% of applied N lost as N₂O) and effluent C concentration (from Chadwick et al. 2000; Velthof et al. 2003; Amon et al. 2006; Bhandral et al. 2007; Luo et al. 2008; Rochette et al. 2008; Li et al. 2014b).

The availability of NO₃ ⁻ for denitrifying bacteria is the first step in biological denitrification, and the denitrification rate and N₂O emission would therefore be expected to depend on effluent N concentration. The trend is that higher effluent total N concentrations would lead to higher N₂O emissions. An increase in N₂O emissions due to an increase in N concentration (and therefore an increase in N load on to soil) may not affect N₂O emission factors. However, analysis of the data from the literature implies that there is a significant exponential relationship between N₂O emission factors and effluent total N concentrations (R² = 0.164, P < 0.001) (Fig. 4). This increase in N concentration is likely to correspond with increases in other components which could also affect emission factors.

Other effluent properties, such as DM content, could also affect the value of the N₂O emission factor. Data from the literature suggest that higher effluent DM content generally leads to higher N₂O emission factors (R² = 0.226, P < 0.05) (Fig. 5). These effects may be obscured by other factors such as the fact that the effluent with higher DM content is also likely to have a higher C and N concentration. Effluent with a larger amount of fine solids can block soil pores, enhancing anaerobic conditions and N₂O emissions, but solid manures generally have lower N contents and result in lower N₂O emission than slurries (Chadwick et al. 2011). Effluents with a lower DM content infiltrate the soil more, which may also affect N₂O emissions and complicate the relationship between DM and N₂O emission factor.

Figure 4 Relationship between nitrous oxide emission factors from effluent application (% of applied N lost as N₂O) and effluent total N concentration (from Velthof et al. 2003; Amon et al. 2006; Bhandral et al. 2007; Luo et al. 2008; Rochette et al. 2008; Li et al. 2014b).

Figure 5 Relationship between nitrous oxide emission factors from effluent application (% of applied N lost as N₂O) and effluent dry matter content (from Chadwick et al. 2000; Velthof et al. 2003; Amon et al. 2006; Rochette et al. 2008; Li et al. 2014b).

Application method

Application method may also affect gaseous N losses. There are four broad categories of land application methods: irrigation of effluent; broadcast spreading; band spreading; and injection into the soil (Rotz 2004).

Meisinger & Jokela (2000) suggest that during the irrigation process, 15%–43% of the total N applied was emitted as NH₃ or otherwise made airborne before contacting the plant or soil surface. Ammonia volatilisation then continues from the field surface, causing an additional loss of up to 35% of total applied N. The range of losses reflects the difference in ambient temperature, precipitation and other weather conditions. If soil conditions are wet following irrigation then rapid incorporation of the effluent is slowed, which may lead to greater NH₃ losses (Rodhe et al. 2006).

A wide range of N₂O emissions have been found after irrigation with effluent. Luo et al. (2008) found that irrigation with FDE generally increased N₂O emissions, and overall N₂O emissions from effluent-derived N ranged from 0.01%–4.93% depending on irrigation time, soil water-filled pore space (WFPS) and grazing history. Bhandral et al. (2004) applied dairy effluent by irrigation to a poorly drained soil and observed that N₂O emissions increased immediately after the application of the dairy effluent, and the total N₂O emitted from effluent application ranged from 2%–5.8% of the total N added in the effluent. This suggests that the soil conditions (including moisture, surplus N and bulk density), manure characteristics and the interaction between the soil and effluent affect the emission of N₂O from soils after irrigation with effluent.

Broadcast spreading throws effluents from the side or the back of a storage vehicle (i.e. slurry wagons and muck spreaders). This method is able to deal with effluents of different solid contents through the use of different storage and spreading units. The efficiency of broadcast spreading techniques for reducing NH₃ losses is variable and again depends on the effluent, soil and weather conditions (Thomsen et al. 2010). In general, less than 1% of the total N applied is lost as NH₃ during the spreading operation (Meisinger & Jokela 2000). The low loss is probably due to the lack of airborne fine particles as a result of this application technique. Compared with application to bare soil, broadcast spreading on grassland or heavy crop residue may increase NH₃ loss by 30%–50%. This could be due to the grass or crop residue serving as a barrier and preventing much of the effluent from making contact with the soil. Also, effluent adhering to the grass could create a larger surface area for NH₃ volatilisation, as suggested by Meisinger & Jokela (2000). Limited data show that broadcast spreading of effluent results in N₂O emissions ranging from 0.07%–1.69% of the total effluent N applied (Misselbrook et al. 2002; Webb & Misselbrook 2004; Thorman et al. 2007). Band spreading takes place using a trailing shoe system, where the trailing shoe is attached to either the slurry tanker or to the storage facility via an umbilical hose. A number of techniques have been used to apply effluent in bands in Europe (Sistani et al. 2010; Thomsen et al. 2010). This technique is primarily used on grassland or other established crops. When bands are well formed and maintained on the surface, NH₃ loss can be reduced by 30%–70% compared with broadcast spreading on the same crop (Rotz 2004). Thomsen et al. (2010) found N₂O emission factors from banded application of effluent of 0.5%.

Effluent injection is the direct injection of effluent into soil, which can occur at either deep or shallow depths. A large portion of the NH₃ emission occurs within a few hours of spreading and rapid incorporation of manure/slurry into the soil can be a method to reduce NH₃ loss (Rodhe et al. 2006). However, this method would be of limited use for application on grazed permanent pastures. Cultivation for pasture renewal can be an opportunity for incorporation of manure/slurry into the soil. Incorporation of effluent may also affect N₂O emissions, but the results vary. Higher N₂O emissions have been shown to occur in some studies (Misselbrook et al. 2002; Wulf et al. 2002; Velthof et al. 2003; Dell et al. 2011); however, this is not always the case (Vallejo et al. 2005). Some research has shown that in different years of the same study (e.g. Rubaek et al. 1996; Sistani et al. 2010), injection of effluent can increase, decrease or have no effect on N₂O emission compared with surface broadcasting. This suggests that when conditions are favourable for denitrification, effluent injection can result in increased N₂O emissions compared with surface broadcasting. But under other conditions, effluent injection has the potential to reduce both NH₃ and N₂O emissions.

Soil factors

Soil conditions, such as moisture content, temperature, C and N concentrations, cation exchange capacity (CEC), pH and plant or residue cover, affect NH₃ and N₂O losses.

Higher soil pH generally leads to enhanced NH₃ emissions by affecting the equilibrium between NH₄ ⁺ and NH₃ (Petersen & Sommer 2011; Salim et al. 2014). The CEC of soils influences the NH₃ concentration through the reaction of NH₄ ⁺ ions with the negatively charged cation exchange sites. Hence, soils with low CECs are more prone to NH₃ volatilisation than soils with high CECs. There is little information regarding the relationship between NH₃ emissions from effluent application and CEC; however, Cameron et al. (2013) reported that there was an inverse relationship between NH₃ volatilisation from applied urine and the CEC of a number of soils. It has been suggested that the higher CEC of clay soils can reduce N availability through increased adsorption of NH₄ ⁺ compared with sandy soils (Jarecki et al. 2008).

Soil physical conditions, such as soil moisture, at the time of application are also very important in terms of NH₃ volatilisation. Sommer & Olsen (1991) reported a two- to three-fold increase in NH₃ emissions from moistened soils compared with those in an air-dry condition and they ascribed the increase to a lower absorption of the liquid fraction into the wetter soils.

Soil pH affects the rates of nitrification and denitrification and therefore the rates of N₂O production (van der Weerden et al. 1999; Cameron et al. 2013). It has been suggested that most denitrifying bacteria grow best in near neutral pH conditions (pH 6–8) and account for significant N₂O emissions. Under acidic conditions, availability of organic C and N mineralisation is lower which results in a decrease in the soil denitrifier community. Hence, overall denitrification rate is higher in neutral or slightly alkaline soils than acidic soils (Thangarajan et al. 2013). The ratio of N₂O/N₂ produced is also affected by soil pH. Rochester (2003) found that the proportion of N₂O increases as pH decreases, with N₂O frequently appearing as the dominant product in acidic soil; approximately the same amounts of each gas are emitted from acidic soils (pH <6.0), whereas for the alkaline grey clays (pH 8.3–8.5) the ratio of N₂O/N₂ is about 0.024. Zaman et al. (2013) also suggested that increasing soil pH above 6 may offer a mechanism to mitigate N₂O emissions by shifting the balance between N₂O and N₂.

Nitrate concentrations have been observed to influence N₂O in the gaseous products of denitrification. It has been found that N₂O emissions increase after the addition of NO₃ ⁻, as the denitrification rate and N₂O emissions depend on NO₃ ⁻ concentrations (Saggar et al. 2004). Blackmer & Bremner (1978) found that nitrate concentrations as low as 50 mg kg⁻¹ in soil inhibit N₂O reduction to N₂. Therefore, at low NO₃ ⁻ concentrations, N₂ is the predominant product and at high NO₃ ⁻ concentrations, N₂O often predominates (Saggar et al. 2013).

Soil WFPS can reflect seasonal changes in temperature and rainfall, and therefore could also affect emissions of N₂O.When WFPS is greater than 55%–60%, there is a sharp increase in N₂O emissions (e.g. Luo et al. 2008; van der Weerden et al. 2014), which may result in relatively high emission factors.

Animal treading leads to soil compaction and has been shown to be an important factor affecting N₂O emissions in grazed pastures (Menneer et al. 2005). For example, Bhandral et al. (2003) observed that soil compaction resulted in a high N₂O emission flux from effluent application, with about 10% of the N applied emitted from compacted soil, whereas uncompacted soil resulted in a loss of only 0.5% of N applied.

Environmental factors

The dominant environmental factors influencing NH₃ and N₂O losses from land-applied effluent include wind speed, rainfall and temperature.

The partial pressure of NH₃ in the atmosphere affects the release of NH₃ gas from the soil to the atmosphere. Hence, temperature and wind speed affect NH₃ volatilisation through their effect on partial pressure close to the soil surface (Meisinger & Jokela 2000). Bouwman et al. (2002) observed that high temperature increases the relative proportion of NH₃ to NH₄ ⁺ present, decreases the solubility of NH₃ in water, and increases the diffusion of NH₃ away from the air-soil interface. Moyo et al. (1989) found that urease activity increased with an increase in temperature from 2–45 °C, and peaked at 37 °C. Furthermore, urea hydrolysis, a function of microbial activity, is also affected by temperature. When the temperature is reduced to a point where microbial activity ceases, the production of NH₄ ⁺ ions and the subsequent loss of NH₃ should stop, i.e. below −4 °C (Meisinger & Jokela 2000). However, Misselbrook et al. (2005) found no relationship between temperature and NH₃ emissions. They concluded that other factors were affecting NH₃ emissions such as soil moisture, humidity, wind speed and rainfall, which may have negated the expected influence of increased temperature.

Wind speed is positively associated with NH₃ emissions by increasing the mass transfer and air exchange between the soil surface and the atmosphere (Misselbrook et al. 2005). Sommer & Hutchings (2001) found that there was a linear relationship between wind speeds up to about 2.5 m s⁻¹ and NH₃ volatilisation. The greatest effect of wind speed is at the beginning of an effluent application, before drying and surface depletion of NH₄ ⁺-N occur. Emissions from surface-applied manure may also be reduced by the formation of a surface crust which acts as a physical barrier to NH₃ emissions (Meisinger & Jokela 2000).

Crust formation is inhibited by regular wetting (Misselbrook et al. 2005), thereby increasing NH₃ emissions. Rainfall after an effluent application can reduce NH₃ emissions by transporting NH₄ ⁺ from the surface into the soil profile, therefore reducing the N content of the effluent on the soil surface (Misselbrook et al. 2005; Webb et al. 2012). Pain et al. (1997) also reported reductions in NH₃ emission of about one-third from an 18 mm rainfall event after effluent application. However, Sommer & Hutchings (2001) noted an increase in NH₃ volatilisation rate immediately following rainfall events several days after effluent application, perhaps due to rewetting and subsequent redrying of the solid manures.

Temperature varies temporally and can affect N₂O emissions (Schils et al. 2013). De Klein & Van Logtestijn (1996) found that with an increase in temperature from 10 °C to 20 °C, the denitrification rate increased about 10-fold in non-irrigated plots and three-fold in irrigated plots. Luo et al. (2008) noticed that there were highly variable patterns of N₂O flux response following effluent application during different seasons, which were considered to be caused by the differences in temperature in autumn and winter.

Enhanced N₂O emissions following rainfall are commonly reported for agro-ecosystems and have been generally attributed to the dual effects of bringing applied substrates into contact with soil microorganisms and a reduction in soil O₂ levels (Luo et al. 2013). Saggar et al. (2004) suggested that the highest N₂O emissions or denitrification rates in temperate grassland soils were obtained following rainfall or irrigation events. It has been suggested that more N₂O is produced at lower temperatures and that the N₂O:N₂ product ratio is increased as the activation energy of N₂O reduction is higher than the activation energy of N₂O production (Holtan-Hartwig et al. 2002).

Management practices to control ammonia and nitrous oxide emissions

Effluent contains valuable organic matter and nutrients that, if applied to pasture, can help to rebuild soil texture and biological activity and provide nutrients required for growth of the next crop (Luo et al. 2008). However, environmental concerns from applying effluent encompass nitrogen gaseous losses which could reduce effluent fertiliser value and add to environmental degradation. There are a number of possible methods that can be used to reduce NH₃ and N₂O emissions from effluent application to pastures, but the key option is to improve overall N efficiency (Saggar et al. 2013). For example, soil management to minimise risk, optimum use of effluent, use of soil N process inhibitors, use of supplementary low-N feed and diet manipulation. Some of these options are currently in use by farmers, while others require further research and development (Table 1). Some strategies, such as reducing livestock numbers, lowering the amount of N, avoiding excessive manure application and using N process inhibitors can reduce both NH₃ and N₂O emissions. However, other strategies, such as timing of effluent application, mixing with soil and effluent pre-treatment, can result in ‘pollution swapping’ between NH₃ and N₂O emissions.

Table 1 Potential mitigation methods for NH₃ and N₂O from effluent application in pasture.

Factors affecting N2O and NH3 emission	Management	Strategies
Animal management	Reduce livestock numbers	Strategic selection of animal type or animal breeding
Feed intake	Reduce N excretion	Decrease the crude protein content of the diet; use breeds of livestock, feed formations or other practices that improve animal performance
Soil and effluent management	Inhibitors	Nitrification and/or urease inhibitors
	Optimise timing of effluent application	Climate and soil factors (under low soil moisture and cool season); seasonal patterns of plant growth
	Avoid excessive effluent application	Match effluent application to crop uptake requirements
	Proper application methods	Tillage or irrigation immediately after application; injection techniques (rapid incorporation), where possible; band spreading with a trailing hose or similar devices
	Improve effluent treatment	Pre-treatment of effluent (reduce pH and C)

Reduce N excretion

The simplest approach to mitigate NH₃ and N₂O emissions from effluent application to pasture is to reduce animal livestock numbers, and hence the amount of effluent generated. Therefore, reducing meat and animal fat consumption in the industrial world would contribute to livestock numbers and animal waste reduction, and also could decrease the gaseous N emissions. However, increases in human population would lead to increased animal protein consumption, so reducing numbers of animals to lower gaseous N emissions is likely to be unacceptable, unless there are significant increases in per-animal productivity. Strategic selection of animal type or animal breeding would help to reduce NH₃ and N₂O emissions. For example, it is suggested that increasing milk production efficiency in dairy cows will offer the potential to reduce the amount of N that ends up in excreta, due to more N being used for milk formation than maintenance (Satter et al. 2002). Similarly, potential benefits might be achieved by growing meat-producing animals to their finishing weight more quickly and thereby reducing total intake and N excretion (de Klein & Eckard 2008).

A reduction in the amount of effluent N applied per hectare of land would help to minimise the potential risks. However, the area required for effluent treatment would then be increased. Decreasing the N content of the effluent (by decreasing N excretion) can offer potential to reduce subsequent N losses. It is suggested that avoiding excessive N in the diet or adding bacteria would keep feeding rations with a better N balance, thereby reducing the N content of the effluent. Luo et al. (2008) found that use of these practices could reduce N excretion by up to 25% in animal operations. There was a 20% reduction in estimated N excretion from urine and faeces by reducing the protein content of supplemental feed (Montes et al. 2013). Nitrogen excretion reduction might be directly achieved by using different breeds of livestock, feed formations or other practices that improve animal performance and thereby the product yield per unit of effluent N (Luo et al. 2008). Although animals cannot avoid excreting N, farmers can reduce the N excretion by decreasing the crude protein (CP) content of the diet (Hristov et al. 2011). Emissions from land-applied manure are further reduced because low-CP diets produce manure with a slower mineralisation rate of N (Powell et al. 2011). It has been suggested that addition of available C to soil would increase denitrification and also the ratio of N₂:N₂O (Oenema et al. 2014). Therefore, higher C in the low-CP diet would reduce the associated N₂O production of denitrification (Saggar et al. 2013).

Inhibitors

Technologies employing urease and nitrification inhibitors have been used as effective mitigation strategies to control N losses from urea and urine application to soil by acting on the N processes of urea hydrolysis and nitrification, respectively (Cameron et al. 2013; Montes et al. 2013). This concept has also been tested on effluent. As part of a proof-of-concept study, Li et al. (2014a) found that the addition of N-(n-butyl) thiophosphoric triamide (NBTPT, commercially named Agrotain) to freshly mixed dung and urine reduced NH₃ volatilisation from 15% to 8% of N following application to pasture. However, whether NBTPT should be used depends on the amount of urea in the effluent. The potential will be limited when there is only a small amount of urea in applied effluent, as most of the urea in urine would have already been transformed to NH₄ ⁺ soon after the effluent was collected. Studies have also found that the application of dicyandiamide (DCD) to effluent can reduce N₂O emission by about 10% to 80% (Vallejo et al. 2005; Li et al. 2014b). These inhibitor technologies require further testing under a range of soil and environmental conditions and their practicality needs to be explored.

Optimise timing of effluent application

In order to improve the efficiency of effluent N use by plants, farmers should consider climate and soil factors (e.g. rainfall and soil drainage) at the time of application and seasonal patterns of plant growth. Houlbrooke et al. (2004) suggested that a reduction in the use of effluent under wet winter and spring conditions could potentially reduce N₂O emissions and N leaching. Luo et al. (2008) also found that strategic application of effluent to pastures under dry soil-moisture status could reduce N₂O emissions by up to 96%. Delaying effluent application until after grazing events could further reduce N₂O emissions by reducing the level of surplus mineral N in soil. Therefore, timing FDE application to occur when soil conditions favour quick infiltration of effluent can reduce NH₃ and N₂O emissions. Thorman et al. (2007) suggested that effluent be applied to an actively growing pasture in spring. The best strategy to reduce N losses is to apply the right amount of effluent close to the time the nutrients are needed by the crop, as suggested by Saggar et al. (2013). On a daily time scale, effluent could potentially be applied in the late afternoon or evening to take advantage of the marked diurnal trend in NH₃ losses, which consist of high daytime losses and lower losses at night (Sommer & Hutchings 2001).

Match effluent application rate to pasture requirements

Application rate and effluent DM content can be manipulated to reduce NH₃ and N₂O losses. Land subjected to high effluent applications can generally emit higher N₂O because the effluent adds N and available C which promote denitrification (Saggar et al. 2004). Moreover, if best management practices are not followed, effluent is often applied to land as a means of disposal, so rates can be excessive. Optimum levels of effluent N, however, will depend on the particular production system concerned. In farm systems, any effluent N applications should be adjusted to take into account the N loading of the applied effluent and to match pasture requirements (Luo et al. 2010). In this regard, use of a nutrient budget model, such as Overseer^® (Wheeler et al. 2008), is a useful tool for assessing the sustainability of N flows within farm systems and identifying opportunities for reducing NH₃ and N₂O emissions.

Optimal application methods

Injection of effluent or rapid incorporation within a few hours of application, which minimises contact of the effluent with air, tends to reduce NH₃ emissions (Webb et al. 2010). However, the potential of this measure is limited for grazed pastures. As found by Mkhabela et al. (2008), water irrigation immediately following effluent application can reduce gaseous NH₃ volatilisation. Incorporation of effluent into the soil can lead to N ‘pollution swapping’ with associated increased direct N₂O emissions and leaching from the soil (Brink et al. 2001). In addition, it can increase CO₂ emissions because more fuel is required by tractors for effluent incorporation (Hansen et al. 2003). Band spreading with a trailing shoe or similar device can provide an alternative option for an established crop or grassland where effluent incorporation is not possible (Webb et al. 2010).

Pre-treatment of effluent

Decreasing the effluent pH below 7 before it is applied can reduce NH₃ emissions (Petersen & Sommer 2011). Ndegwa et al. (2011) suggested that the addition of nitric or sulphuric acid to the effluent (dropping effluent pH to 6.5) reduced NH₃ loss by up to 75%. Using alum or ferrous sulphate can also have an acidifying effect on effluent, which may reduce N loss. Effluents with reduced C and acidification of effluent have the potential for further reducing N₂O emissions following land application (Fangueiro et al. 2010). Results from previous studies on the effect of pre-treatment of effluent and application method on NH₃ and N₂O emissions after effluent application suggested that emissions after digestion and acidification of effluent were significantly lower (P < 0.05) than those from untreated effluent (Wulf et al. 2002; Bhandral et al. 2009).

Interactions of NH₃ and N₂O mitigation strategies

As discussed above, there is a range of mitigation options that can be used to reduce N losses before and during effluent application. However, some of these practices may result in ‘pollution swapping’. For example, Brink et al. (2001) found that NH₃ abatement may have an adverse effect on N₂O emissions, while abatement of N₂O has no effect on NH₃ volatilisation. The use of effluent injection techniques generally decreases NH₃ emissions, but it can increase N₂O emissions (Flessa & Beese 2000; Wulf et al. 2002). Similarly, anaerobic digestion may reduce N₂O emissions from the land application of effluents, but such effluent treatment may enhance NH₃ emissions resulting from a higher effluent pH (Saggar et al. 2004). Previous studies (e.g. Zaman et al. 2013; Li et al. 2014b) have shown that DCD can reduce N₂O emissions from urine and effluent applications in pastoral soils. However, DCD application could potentially increase NH₃ emissions (Schils et al. 2013). The measures to reduce gaseous emissions of both NH₃ and N₂O from land application of farm effluents would therefore require some trade-offs between these two gases.

Conclusion

Ammonia and N₂O production and emissions are due to a complex combination of physical, chemical and biological processes. Studies on the general magnitude and pattern of field emissions have enhanced our understanding of gaseous emissions processes. Ammonia volatilisation is regulated by NH₄ ⁺ concentration, effluent DM and pH, soil pH, CEC, soil moisture, temperature and wind velocity. Nitrous oxide emission is regulated by C supply, NO₃ ⁻-N concentration, aeration status, pH and temperature. Agricultural management practices and weather conditions can also regulate NH₃ and N₂O emissions in the field by affecting the basic factors that influence these gaseous losses. There is a range of mitigation options that can be used to reduce N losses before and during effluent application, including: reducing livestock numbers; lowering the amount of N; using N process inhibitors; optimising timing of effluent application; matching effluent application rates to crop requirements; and improving effluent treatment before application. Some of these options can reduce both NH₃ and N₂O emissions, such as reducing livestock numbers; lowering the amount of N; avoiding excessive manure application; and using N process inhibitors. It is important to remember, however, that some of these options, such as timing of effluent application, can result in ‘pollution swapping’ between these two gases and some of these options, such as inhibitor use, are in the early stages of development. Future research needs to focus on the overall impact of mitigation options on the whole suite of gaseous emissions and the practicality of those options.

Acknowledgements

The authors gratefully acknowledge funding from Ballance Agri-Nutrients Limited and the New Zealand Government through the Livestock Emissions & Abatement Research Network (LEARN) scholarship programme. The authors also acknowledge funding for Taishan Scholars Programme for overseas experts. Any view or opinion expressed does not necessarily represent the view of the funding organisations.