Arable Farmer-led Group: climate change evidence

Annex D – CXC Measures for the Arable Sector

This annex contains an extract of mitigation measures from the CXC (2020) report that are specific to the arable sector(s).

Growing more grain legumes in rotations

Legumes (e.g. peas, beans, clover) have the special ability to source nitrogen straight from the dinitrogen gas found in the atmosphere, requiring very low (or no) additional nitrogen fertilisers. This is possible due to their symbiotic relationships with bacteria in the soil. They also provide some of this nitrogen to crops which are cultivated with them and also to those which follow them in a rotation (as the above ground residues and roots of the leguminous crops increase the nitrogen content of the soil), reducing the fertiliser requirement for those crops. This measure is about increasing the area of grain legumes in rotations in Scotland.

The description, assumptions and results in the UK MACC report of 2015 (Eory et al. 2015) was used for this measure.

Overview

N fixing crops (legumes) form symbiotic relationships with bacteria in the soil that allows them to fix atmospheric N and use this in place of N provided by synthetic fertilisers. This measure is about increasing the area of grain legumes in arable rotations, thereby reducing N fertiliser use in two ways: by requiring no N fertiliser (so there will be a reduction per ha equivalent to the N fertiliser which would have been applied to the non-leguminous crop that would otherwise have been grown); and by having a residual N fertilising effect so that the crops grown after legumes require less N than when grown after non-legumes (Defra 2011).

Greenhouse gas mitigation summary

Grain legumes are able to fix in excess of 300 kg N ha^-1 y^-1; can supply N to subsequent crops; are valuable as a break crops in arable rotations; and can provide biodiversity benefits (Rees et al. 2014). The abatement achievable is due to the change in crop areas (i.e. replacement of other arable crops with grain legumes in the rotation and applying no fertiliser on them) and a reduction in N fertiliser use of 30 kg ha^-1 on the subsequent crop (Defra 2011).

Table D1. Data from literature on abatement

Abatement	Value	Country	Reference
N use	-0.5 t CO2e ha-1 of soil nitrous oxide emissions	UK	(Moran et al. 2008)
N use	-0.5 t CO2e ha-1 of soil nitrous oxide emissions	UK	(MacLeod et al. 2010)
N use	No fertiliser on the legume, -33 kg N ha-1 on the following crop; i.e. -0.64 t CO2e ha-1 where legumes introduced (not rotation average)	France	(Pellerin et al. 2013)

Costs

We estimated the cost of this measure from the difference of the gross margin in grain legumes (field beans and peas £380 ha^-1, (SAC 2013)) and other crops (weighted average: £809 ha-1, (SAC 2013). The fertiliser savings from the reduced fertilisation of the following crop is accounted for as benefit ( £23.55 ha 1). The net cost is in high contrast with the only data found in the literature, which estimates the net costs as £13.6 ha-1 for the area where legumes are introduced (Pellerin et al. 2013). This estimate consists of savings in fertilisers and their applications, elimination of tillage operation for the following crop, and changes in the gross margins of the rotations.

Current uptake and maximum additional future uptake

The frequency of legumes in the rotation depends on different factors according to the nature of the legume. For example, peas are grown only one year in five due to the need to reduce the risk of disease. This is less of a concern for field beans but these are harvested late and delay sowing, and hence yield, of any subsequent cereal crop. Therefore, in practice, beans are also only likely to be grown once in every 5 years. The inclusion of peas and beans in rotations including oilseed rape is limited to once in every six years, due to disease risk. Peas are unsuitable for 'heavy' soils (effectively clay loam and heavier), while beans are unsuited to light soils (sandy loam and equivalents). Therefore, we limited the applicability of the grain legumes to 1/6 of the total arable crop area in any given year.

In 2016, field beans and peas and peas and beans for human consumption were grown on 3,100 ha and 9,300 ha (0.7% and 1.7% of the arable crop area, respectively) in Scotland. Although we assumed the introduction of Greening measures in the Common Agricultural Policy increases the area where field beans and peas are cultivated to 5% of the arable area, this increase is not included in the future reference scenario, but included in the abatement of this measure.

Assumptions used in the MACC

Table D2. Assumptions used in the modelling

Parameter	Change in value
N fertiliser use on subsequent crop	-30 kg N ha-1
Cost (difference in gross margin between field beans and peas and other crops)	£429 ha-1

Crop varieties with higher nitrogen-use efficiency

Nitrogen fertilisation is essential to achieve current yields of most crops. However, only 49% of the nitrogen applied to and biologically fixed by crops (including grass) is recovered as food and feed in Europe (Westhoek et al. 2015), most of the remaining being lost to the environment as ammonia, nitrate and nitrous oxide, causing multiple environmental problems. Crops need nitrogen for their growth, but due to the nature of biophysical processes they can utilise only part of the nitrogen which is in the soil. Improving the efficiency of crops to utilise the nitrogen fertiliser is therefore key in mitigating emissions as well as reducing the economic loss as unrecovered nitrogen. Plant breeding can contribute to improving the nitrogen-use efficiency. Nitrogen-use efficiency varies between individual plants of the same species, and some of this variation is heritable. Therefore, plants with increased nitrogen-use efficiency can be selected for further breeding. Additionally, radically new cultivars can improve nitrogen-use efficiency and thus reduce GHG emissions or at least the emission intensity of production. For example, perennial wheat can help retain more carbon in the soil as well as reduce fertiliser, pesticide, and fuel use. Nitrogen-fixing cereals, for which three main research streams are ongoing, could, when realised, bring substantial reduction in the nitrogen-fertilisation needs of plants. However, due to data limitations, these mitigation measures only look at possible improvements in existing cultivars.

Overview

Nitrogen fertilisation is essential to achieve current yields of most crops. However, only 49% of the nitrogen applied to and biologically fixed by crops (including grass) is recovered as food and feed in Europe (Westhoek et al. 2015), most of the remaining being lost to the environment as ammonia, nitrate and nitrous oxide, causing multiple environmental problems.

Improving the efficiency of crops to utilise the nitrogen fertiliser is therefore key in mitigating emissions as well as reducing the economic loss of unrecovered nitrogen. Nitrogen-use efficiency (NUE) is defined as yield per unit of nitrogen available to the crop (Moll et al. 1982). Barraclough et al. (2010) demonstrated that season and nitrogen input had a significant effect on NUE, but crop variety choice also contributed to NUE variation. It has been proposed that NUE can be improved both via adopting crop, soil and fertiliser management practices and through plant breeding (Barraclough et al. 2010; Hawkesford 2014; Hawkesford 2017; Sylvester-Bradley & Kindred 2009). The latter is possible as NUE varies between plants and some of this variation is linked to phenotypic traits and genotypic markers (Bingham et al. 2012). This variation can be as much as threefold (from 27 to 77 kg DM (kg N)-1), as Barraclough et al. (Barraclough et al. 2010) found in wheat varieties from four different European countries.

Additionally, radically new cultivars can improve NUE and reduce GHG emissions. For example, perennial wheat can help retain more C in the soil as well as reduce fertiliser, pesticide and fuel use (Bell et al. 2008). Nitrogen-fixing cereals (for which three main research streams are ongoing, targeting nodule development, identification of nitrogen-fixing biofertilisers and the introduction of nitrogenase enzyme and pathway into the plant (Beatty & Good 2011)) could, when realised, bring substantial reduction in the nitrogen fertilisation needs of plants.

Breeding for improved NUE can target both the efficiency of nitrogen uptake and nitrogen utilisation in the plant; as these are different physiological processes they are genetically independent, raising the potential for parallel gains (Hawkesford 2014). However, such breeding needs to consider potential trade-offs with other desirable traits; for example, the root system can be modified to increase the uptake of subsoil nitrate, but this adversely affects the uptake of phosphate from the topsoil (Bingham et al. 2012; Ho et al. 2005).

Despite the yield plateau of the last two decades (Knight et al. 2012), most of the experimental studies which have looked at the improvements in NUE of different varieties of the same crop concluded that there has been a continuous improvement in NUE in the past decades. The economics of grain price and fertiliser costs are two potential causes of the yield plateau, resulting in stagnating nitrogen applications in the past two decades for newer varieties which require higher nitrogen rates to manifest their full yield improvement (Knight et al. 2012). This suggests that the improvement might continue as a baseline in the future, and there is scope to accelerate these gains. The assumption in this report is that these improvements can be achieved faster and adopted on larger growing areas, given increased incentives to breeding companies to research and develop and to farmers to adopt such cultivars.

This mitigation measure examines using traditional breeding to improve NUE and considers three major crops in Scotland: wheat, barley and oilseed rape. The measure means cultivating varieties of already common crops in Scotland that have higher NUE than the currently common varieties.

Greenhouse gas mitigation summary

The abatement rate is approximated from an estimate of the NUE or yield improvement, assuming that yields are kept constant and nitrogen application decreases to achieve the same yield. As the genetic gain in breeding is cumulative, the mitigation measure is assumed to have an annually increasing nitrogen-reduction effect (even though new cultivars with improved yields tend to require increasing nitrogen inputs (Foulkes et al. 1998; Knight et al. 2012)).

For wheat and oilseed rape, the gap between the improvements in new cultivars and the realisation of that on farms is 0.013 and 0.012 t ha-1 y-1, respectively, equivalent to 0.2% and 0.4% yield increase annually. The assumed annual nitrogen reduction is therefore 0.2% and 0.4% for these two crops, respectively. The barley annual NUE gain is 1.2%. If we assume that 80% of this gain is realised on farms, there is an additional potential improvement of 0.24% in the NUE. Thus, we assume an annual nitrogen reduction of 0.24%.

Table D3. Data from literature on abatement

Abatement	Value	Country	Reference
Wheat
Yield	+0.063 t ha-1 y-1 (cumulative) of new cultivars (~1%)	UK	(Knight et al. 2012)
Yield	+0.05 t ha-1 y-1 (cumulative) realised on farms	UK	(Knight et al. 2012)
Yield	+0.096 t ha-1 y-1 (cumulative) historically over 20 years (1969-1988)	UK	(Foulkes et al. 1998)
NUE (kg grain N (kg N) -1)	+0.9% y-1 historically over 20 years (1969-1988)	UK	(Foulkes et al. 1998)
Yield	+1% y-1 (cumulative) historically over 75 years (1931-2005)	W Europe	(Bingham et al. 2012)
NUE (kg yield DM (kg N) -1)	+1.2% y-1 (cumulative) historically over 75 years (1931-2005)	W Europe	(Bingham et al. 2012)
Yield	+0.06 t ha-1 y-1 (cumulative) of new cultivars (~2%)	UK	(Knight et al. 2012)
Yield	+0.048 t ha-1 (cumulative) realised on farms	UK	(Knight et al. 2012)

Costs

A price premium might have to be paid for varieties with improved NUE. We assume that other traits of the crops are not going to be adversely affected with the level of improvement set out above; therefore, no costs or benefits beyond the seed price premium and the nitrogen savings are included in the calculations. The seed price premium is estimated to be 10% of the price.

Current uptake and maximum additional future uptake

The measure is in theory applicable to all crops, although here we considered only three major crops: wheat, barley, and oilseed rape. The current NUE of the common cultivars is regarded as the baseline, and thus the current uptake of this measure is assumed to be zero.

Summary of assumptions used in the MACC

Table D4. Assumptions used in the modelling

Parameter	Change in value
N application	-0.13% annually (cumulative)
Crop yield	No change
Seed cost	+10%

Intercropping

Intercropping is the spatially and temporally coexistence of two or more arable crops. Typically, one of the crops is a grain legume, and therefore biologically fixes nitrogen. Hence, there will be a reduction in the quantity of fertiliser applied per hectare. In addition, there is the potential for some of the fixed nitrogen to be transferred to the other crop, further reducing the requirement for inorganic nitrogen. However, the nitrogen concentration of legumes is higher than non-legume crops, thus the emissions from the residues will be increased. In the UK, the grain legume is typically pea or faba bean and the cereal is spring oats, spring barley or spring wheat. Although the technology is being developed to separate the cereal from the grain legume, intercrops are usually used as feed for ruminants or monogastrics. However, there will be the need to adjust the ration depending on the protein content of the actual harvested crop. It is assumed that the yield of the intercrop is similar to that of the sole cereal crop.

Overview

Intercropping is the spatially and temporally coexistence of two or more arable crops. Typically, one of the crops is a grain legume, and therefore biologically fixes nitrogen. Hence, there will be a reduction in the quantity of fertiliser applied per hectare. In addition, there is the potential for some of the fixed nitrogen to be transferred to the other crop, further reducing the requirement for inorganic nitrogen. However, the nitrogen concentration of legumes is higher than non-legume crops; thus the emissions from the residues will be increased. In the UK, the grain legume is typically pea or faba bean and the cereal is spring oats, spring barley or spring wheat. Although the technology is being developed to separate the cereal from the grain legume, intercrops are usually used as feed for ruminants or monogastrics. However, there will be the need to adjust the ration depending on the protein content of the actual harvested crop. It is assumed that the yield of the intercrop is similar to that of the sole cereal crop.

Greenhouse gas mitigation summary

The mitigation arises due to the reduction in inorganic fertiliser applications. In addition, there is the potential for a reduction in fuel use as there will be a reduction in the number of tractor passes due to a reduction in the number of fertiliser applications. There will be an increase in the nitrous oxide emissions from the residues due to the higher nitrogen concentration of the legume relative to the cereal.

As a result of the legume component of the intercrop, it is assumed that the inorganic fertiliser input is approximately halved (SAC 2018). It is assumed that this measure is only applicable to the spring barley and oat area that is used for feed. Based on spring barley and spring oat area, and IPCC (2006), it is assumed that the N contained in the residue will increase by 40%.

Costs

Pea seed is approximately 50% more expensive than barley seed (SAC 2018). Assuming a replacement rate seed mixture, the costs of the seeds will be three times higher than for a pure barley crop.

Current uptake and maximum additional future uptake

The measure is applicable to the area sown for feed production. The crop can be harvested as either whole crop silage or grain. Based on reported figures (Scottish Government 2018b), the tonnage used for feed is 50% of the barley crop, which equates to 43% of the spring barley area. In the case of oats, 30% was used for feed. Therefore, the applicability is 43% of winter and spring oat, and 100% of spring non-malting barley.

There is increasing interest in growing intercrops. However, there are no reported figures on current uptake. It has been assumed that this measure is applicable to intercrops that are grown for feed. As the technology improves, there is the potential for the grain to be separated and therefore used for human consumption. However, for this to be common practice, there is the need for the market to accept products that have been grown as intercrops as opposed to pure stands.

Assumptions used in the MACC

Table D5. Assumptions used in the modelling

Parameter	Change in value
N application rate	-50%
Residue returns (N)	+40%
Energy CO2	Reduction in 1 tractor pass: -1.57 l ha-1 (@2.594 kg CO2e l-1)
Seed costs	+200% (+£150 ha-1)

Nitrification and urease inhibitors

Nitrous oxide emissions from soils are a result of bacteria transforming nitrogen compounds (such as those applied as fertilisers) and in the meantime releasing some of the nitrogen as nitrous oxide. One of these processes, nitrification, can be slowed down by certain chemical compounds (like dicyandiamide (DCD), 3,4-dimethyl pyrazole phosphate (DMPP) and nitrapyrin), which depress the activity of nitrifying bacteria. This, in turn, improves the availability of the nitrogen fertiliser for the plants and reduces both nitrous oxide emissions and nitrate leaching (however, in some cases it can increase ammonia and hence indirect nitrous oxide emissions). Furthermore, a large proportion of the nitrogen in urea-based fertilisers gets transformed into ammonia and, due to the urease enzyme, released by soil bacteria. This leads not only to ammonia (and indirect nitrous oxide) emissions but reduces the nitrogen plants can utilise. Urease inhibitors delay urea hydrolysis to ammonia, reducing ammonia emissions. Using urea in combination with urease inhibitors and nitrification inhibitors can further reduce nitrous oxide emissions. Nitrification and urease inhibitors can be injected into the soil together with liquid fertilisers; applied as a coating on granular fertilisers; and mixed into slurry before application. They can also be spread after grazing to reduce emissions from urine.

Overview

Nitrification inhibitors depress the activity of nitrifying bacteria, improving the nitrogen fertiliser's plant availability and reducing nitrous oxide emissions and also nitrate leaching in high rainfall circumstances (Akiyama et al. 2010), although in some cases they can increase ammonia (and hence indirect nitrous oxide) emissions (Lam et al. 2017). Various compounds have been identified as nitrification inhibitors; probably the most widely studied are dicyandiamide (DCD), 3,4-dimethyl pyrazole phosphate (DMPP) and nitrapyrin.

Furthermore, urea-based fertilisers have a high rate of ammonia volatilisation when applied to soils, due to the urease enzyme released by soil bacteria. This leads not only to ammonia (and indirect nitrous oxide) emissions, but reduces the N plants can utilise. Urease inhibitors delay urea hydrolysis to ammonia, reducing ammonia emissions (Harty et al. 2016). Using urea in combination with urease inhibitors and nitrification inhibitors can therefore further reduce nitrous oxide emissions.

Nitrification and urease inhibitors can be injected into the soil together with liquid fertilisers, applied as a coating on granular fertilisers and mixed into slurry before application. They can also be spread after grazing to reduce emissions from urine.

In our analysis, we considered the application of nitrification inhibitors with ammonium nitrate fertiliser, and nitrification and urease inhibitors with urea applications. We expressed the effect as a change in the soil nitrous oxide emission factor.

Greenhouse gas mitigation summary

The effectiveness of nitrification inhibitors in reducing nitrous oxide emissions and nitrogen leaching depend on a variety of factors. In a meta-analysis of 113 datasets of field experiments Akiyama et al. (2010) found that the nitrous oxide reduction effect depended on the type of nitrification inhibitor and land use type. The effect also depends on the type of fertiliser used (Misselbrook et al. 2014) and on environmental conditions at the site (Cardenas et al. 2019).

UK experiments showed variable results. In fertiliser experiments by Misselbrook et al. (2014) across six sites (including arable and grassland fields), nitrous oxide emissions from ammonium nitrate were significantly reduced at two sites (average effect -43%), while nitrous oxide emissions from urea treatment were significantly reduced at four sites (average effect -54%). The mean nitrous oxide emission reduction across the six experiments was 38% and 64% for DCD applied with ammonium nitrate and urea, respectively. There was no significant effect of DCD on ammonia emissions, apart from at one site. Nor was yield significantly affected either in all but one case (where it was reduced by 20%).

Cattle urine experiments by the same authors showed significant reduction in three out of four cases, with a mean effect of -70%. Ammonia emissions and grass yields were not significantly affected. Slurry experiments did not reveal any significant effect, as variability amongst the replicates was very high (Misselbrook et al. 2014).

Grassland experiments in the UK with ammonium nitrate and urea fertiliser showed mixed results too. Cardenas et al. (2019) found that DCD increased the nitrous oxide emission factor at one site significantly (by 20%), decreased it at another site significantly (by 52%), and had no significant effect at a further three sites. When DCD was applied with urea the nitrous oxide emission factor changed significantly at only one site (-94%). However, applying urea instead of ammonium fertiliser reduced the nitrous oxide emission factor by 49%, and using urea combined with DCD resulted in a 85% reduction in the nitrous oxide emission factor compared to using ammonium nitrate only. Yield changes were not significant in any case.

Experiments at two permanent grassland sites in Ireland showed that urea applied with a combination of urease and nitrification inhibitor reduced nitrous oxide emissions by 56% (Harty et al. 2016).

Table D6. Data from literature on abatement

Abatement	Value	Country	Reference
Nitrous oxide emissions	Average: -38% (95% confidence interval: ‑44% to ‑31%) DCD: ‑30% (95% confidence interval: ‑36% to ‑26%) nitrapyrin: ‑50% (95% confidence interval: ‑55% to ‑30%) DMPP: ‑50% (95% confidence interval: ‑55% to ‑42%)	Across the world	(Akiyama et al. 2010) meta-analysis
Nitrous oxide emission factor	DCD with ammonium nitrate: -38% DCD with urea: -64% DCD with cattle urine: -70%	UK, grass and arable	(Misselbrook et al. 2014) - experiments
Nitrous oxide emission factor	DCD with ammonium nitrate: -19% DCD with urea: -66%	UK, grass	(Cardenas et al. 2019) - experiments
Nitrous oxide emission factor	DCD and NBPT with urea: -56%	Ireland, grass	(Harty et al. 2016) - experiments

Costs

Agrotain® Plus, which is a combined urease and nitrification inhibitor, costs around £0.1 (kg N)^-1, derived from information posted on agricultural forums (precise price information was not publicly available). This value was used for both the nitrification and urease inhibitor application.

Current uptake and maximum additional future uptake

The current uptake is assumed to be 0%.

Assumptions used in the MACC

Table D7. Assumptions used in the modelling

Parameter	Change in value
Ammonium nitrate EF1 change	-25%
Ammonium nitrate EF1 change	-50%
Fertiliser cost change	£0.1 kg N-1

Soil pH management

One of the important properties of agricultural soils is their pH, in other words their acidity level. The optimal soil pH is between pH 5.6 and pH 6.2, depending on the soil type. Soil surveys in Scotland show that many areas have soils that are too acidic. The acidity of these soils compromises crop growth, reducing the yield, and increases the proportion of nitrogen fertiliser which is converted to nitrous oxide and emitted from the soil. Soil pH control is, therefore, a practice which can increase the yield and reduce nitrous oxide emissions at the same time. It involves applying lime to the soil, when and where needed, and usually testing the soils every four to five years.

Overview

The acidity of soils (soil pH) plays a major role in regulating the chemistry and fertility of soils and depends upon the net balance of a wide range of chemical and biological processes. Good management of soil acidity is essential to optimise crop productivity. Most crop plants are more productive in a range of pH between 5.5 to 7.0. Outside of this range productivity decreases and the utilisation of nutrients added – including nitrogen fertilisers – becomes less efficient. There is a range of indirect ways in which pH influences GHG emissions, making pH management an important tool in GHG mitigation.

Soil pH plays an important role in regulating and modifying nitrous oxide emissions. In more acid soils, there is a higher ratio of N₂O:dinitrogen^[38] emission because the N₂O reductase enzyme which converts N₂O to dinitrogen is inhibited (Liu et al. 2014). Thus, in soils that have a tendency to produce N₂O by denitrification, more acid conditions are likely to lead to a higher N₂O emission rates (Simek et al. 1999; Goulding 2016; Zhu et al. 2019). Because soil acidity can also reduce crop growth, maintaining soil pH at an appropriate level is considered important for both the optimisation of crop production and efficient use of fertiliser inputs (Goulding 2016). Lower crop biomass and higher nitrous oxide emissions in acid conditions can lead to a large increase in the quantity of Nitrous Oxide produced per unit of crop product) (nitrous oxide emission intensity).

Evidence suggests that lime application may modify soil microbial communities (Goulding 2016) and increase organic matter inputs (Fornara et al. 2011; Jokubauskaite et al. 2016) with the effect of increasing soil carbon stocks (SOC) (Li et al. 2018, Fornara et al. 2011).

Managing soil pH involves gathering information on the current status of the soil (e.g. via soil sampling and analysis) and the application of lime on land which is below the optimal pH for crop or grass growth. Optimal pH varies depending on the land use, type of crop grown, and soil type. Required lime application rates to optimise pH vary depending on soil type and on the difference between the existing soil pH and the target pH. Usually it is sufficient to repeat this process in every four years.

Greenhouse gas mitigation summary

Changes in nitrous oxide emissions following lime application result from changes to the nitrification and denitrification processes. These effects are context specific, with variable relationships between pH and the proportion of applied nitrogen emitted as nitrous oxide (Skiba et al. 1998; Russenes et al. 2016). However, since liming increases soil nutrient availability (ALA 2011; Goulding 2016), requirement for nitrogen application is likely to decrease, or the same nitrogen fertilisation rate would result in increased yield, i.e. a net reduction in nitrous oxide emission intensity.

SOC content is likely to increase where pH is raised – again, a complex and context specific response (Li et al. 2018). In grassland, Fornara et al. (2011) reported substantial increases in grassland SOC for limed treatments, both in fertilised and unfertilised swards. For cropland, Tu et al. (2018) reported a positive correlation between pH and SOC. Based on the aforementioned papers and work in SRUC here we assume that an increase of 1 pH unit in the range pH 4-7 corresponds to an increase in SOC concentration of 0.82-1.97 g kg^-1. At a typical soil bulk density of 1.1 g cm^-3, and assuming pH impact to 20 cm depth (Goulding 2016) this roughly equates to an increase of 1.8-4.3 t C ha^-1. Assuming a 20-year stabilisation period (de Klein et al. 2006), this equates to a sequestration rate of 330-788 kg CO₂e ha^-1 year^-1. Note that this is a broad extrapolation based on site-specific data and should be taken as an indication only. To provide a conservative estimate we assume 300 kg CO₂e ha^-1 year^-1 C sequestration in this work.

Direct CO₂ emissions from lime application means that lime can be (though is not necessarily) a net source of CO₂ (Hamilton et al., 2007). The relevant IPCC Guidelines for National greenhouse gas Reporting (de Klein et al., 2006) assume lime to be a CO₂ source, with an estimate of 0.0625—0.125 kg CO₂ kg lime^-1. This emission factor is directly related to the mass fraction of C in lime (CaCO₃), with the maximum emission assuming release of all molecular C to the atmosphere as CO₂ (de Klein et al., 2006; Fornara et al., 2011). This contrasts with the findings of Hamilton et al. (2007), who show that whilst lime can be a source of CO₂, it is more often a net sink. Fornara et al. (2011) also show that lime can be a C sink; the authors identify two pathways by which this can be the case. Lime may either a) increase carbonic acid (HCO₃^-) concentrations in soil water, sequestering 25-50% of lime C, or b) contribute to the movement of existing soil C from labile to humified pools, increasing its net storage time in the soil.

Emissions associated with lime extraction (embedded emissions) have been estimated at 0.074 kg CO₂e kg lime^-1 (range 0.054—0.089 kg CO₂e kg lime^-1) (Kool et al. 2012).

Costs

The costs of lime application include purchase of lime, spreading and soil analysis. It is recommended that farms apply lime at three to six year intervals depending on results of soil analyses (SRUC 2014). The financial benefits of soil pH management consist of the additional income from yield increase.

Applicability and uptake

The Scottish Government (2018a) reports that 64% and 30% of farms carried out pH testing on arable and grazing land respectively in 2016. A recent survey of over 1,000 fields of grassland (Ayrshire, Water of Coyle) and arable land (Perth, East Pow), showed that 57% of grassland soils and 34% of arable soils had low or very low pH values (SRUC 2018). This is consistent with UK data indicating that between 31% and 49% of arable and grassland soils have suboptimal pH (PAAG 2016). The applicability of the measure is assumed to be 50% on fertilised grasslands and 30% on croplands.

Summary of assumptions used in the MACC

Table D8. Assumptions used in the modelling

Parameter	Change in value
Yield change	+6.22% (crops and grass)
EF1 change	-3%
C sequestration	300 kg CO2e ha-1 y-1
Lime cost	£111 ha-1 (3.7 t ha-1 lime @ £30 t-1) in every 4 years
Lime spreading cost	£10.16 ha-1 in every 4 years
Soil analysis cost	£20 ha-1 in every 4 years

Slurry injection and bandspreading of slurry

Livestock slurry can be applied to the soil using a range of techniques. The most common approach in Scotland is to use a low trajectory splash plate (broadcast). Bandspreading and injection are practical alternative methods that can be used for spreading slurry to cropland and grassland. In bandspreading, a series of parallel pipes connected to a slurry tank apply the slurry in discreet bands directly on the surface (trailing shoe / trailing hose), while in injection the slurry is placed into slits cut by machinery. These techniques reduce odour and crop contamination. They also mitigate ammonia emissions and subsequent nitrous oxide emissions (from the transformation of ammonia into nitrous oxide). However, due to the increased pool of ammonium nitrogen in the soil and the changes in the soil conditions, direct emissions of nitrous oxide can increase.

Overview

Livestock slurry can be applied to the soil using a range of techniques. The most common approach in Scotland is to use a low-trajectory splash plate (broadcast), which accounts for over 68% of applications (Scottish Government 2016). Bandspreading and injection are practical alternative methods that can be used for spreading slurry to cropland and grassland.

Compared to broadcast, these techniques spread the slurry more evenly, reducing odour and crop contamination (Thorman 2011). Bandspreading and injection have also been shown to reduce ammonia emissions (Hafner et al. 2019), and therefore the associated indirect GHG emissions. However, due to the increased pool of ammonium-nitrogen in the soil and the changes in the soil conditions, emissions of nitrous oxide can increase (Thorman 2011).

The measure entails switching from applying slurry via a splash plate (broadcast) to: (a) band spreading in which a series of parallel pipes connected to a slurry tank applies the slurry in discreet bands on the grass surface (trailing shoe / trailing hose) or (b) injection of slurry below the soil surface.

Greenhouse gas mitigation summary

Trailing shoe and injection technology can dramatically reduce ammonia emissions (Defra 2007, Hafner et al. 2019). The semi-empirical model of Hafner et al. (2019) using European data predicted a reduction relative to broadcast of 63% by injection; this compares with a reduction of 48% observed in a UK study (Defra 2007). The model predictions for trailing shoe were 33% lower than for broadcast (Hafner et al. 2019). However, in a UK study, trailing shoe / hose only had a significant impact on ammonia emissions in three of fourteen experiments with a mean reduction of 12% (Defra 2007).

The lack of effect was explained by the fact that the slurry did not stay in the band and therefore did not rapidly infiltrate the soil. The results of the Defra Greenhouse Gas Platform Project^[39] revealed that trailing hose resulted in a reduction in emissions in spring, but not in autumn. The leaching losses associated with autumn applications of slurry were higher for trailing shoes than for broadcast. In terms of direct nitrous oxide emissions, the effect of application method was variable with either no effect or increases in emissions being reported (Bourdin et al. 2014, Chadwick et al. 2011, Defra 2007). The results of the Defra study (2007) showed no consistent impact on N use efficiency of the slurry, and thus there is no consistent effect on reducing the requirement for the associated inorganic fertiliser inputs.

Compared to broadcast, the energy cost of trailing hose and injection are higher. The Farmscoper tool (ADAS 2017) assumes that the significant reductions on ammonia volatilisation will be offset by increases in energy required to power the equipment for both band spreading and injection, along with mixed effects on nitrate leaching and direct nitrous oxide (Table D9 and D10).

Table D9. Effect on pollutant flows of Farmscoper measure 70: Using slurry trailing hose / trailing shoe application techniques ( ADAS 2017)

	Ammonia	Energy use	Nitrate	Nitrous oxide
	Grass	Arable	Arable/Grass	Arable/Grass
Pathway(s)	Gaseous	Gaseous		All flows
Effect	-50%	-25%	+50%	+10%

Table D10. Effect on pollutant flows of Farmscoper measure 71: Use slurry injection application techniques ( ADAS 2017)

	Ammonia	Energy use	Nitrate	Nitrate	Nitrous oxide	Nitrous oxide
	Grass	Arable/Grass	Grass	Grass	Grass	Grass
Pathway(s)	Gaseous		Runoff/ Preferential	Leaching	Runoff/ Preferential	Leaching
Effect	-80%	+100%	-50%	+25%	-50%	+25%

From the evidence detailed above, it is assumed in this work that the volatilisation (FRACgas) is reduced by 48% and 12% for injection and trailing hose, respectively. Thus, the revised FracGas values are 0.1 for injection and 0.18 for trailing hose. It is assumed that the increase in fuel use offsets the reduction in nitrous oxide emissions and thus neither is changed in the model.

Table D11: Data from literature on abatement

Abatement	Value	Country	Reference
FracGas	Injection: 0.1 Trailing shoe / hose: 0.18	UK	(Defra 2007)
Energy CO2	No change (reduction in nitrous oxide is offset increase in CO2)	UK	(ADAS 2017)

Costs

This measure involves the purchase of equipment for band spreading or injection and higher operating costs associated with increased fuel use, particularly for injection systems. Cost estimates are given in Table D12.

Table D12. Financial costs and benefits of the measure

Costs/savings	Value ('-' sign for savings)	Reference
Trailing shoe / trailing hose	£0.91/m3 slurry	(ADAS 2017)
Slurry injection	£2.38/m3 slurry	(ADAS 2017)

Applicability

In the UK, 43% of the organic manures applied is cattle or pig farm-yard manure (FYM), and 44% is cattle or pig slurry (Defra 2019). Therefore, it is assumed that the measure is applicable to 50% of organic manure applications. Slurry is applied to 2.2 % of the winter sown crops, 5.6% of the spring sown crops and 25.6% of grassland (Defra 2019). It has been assumed that the slurry applied to cropland is incorporated at the time of application. Therefore, this measure is regarded as only being applicable to grassland.

Current uptake and maximum additional future uptake

In Scotland in 2016, 17.2 million tonnes of FYM or slurry was applied (Scottish Government 2016). In 2016, 4.8 million tonnes (28%) was bandspread, and 0.6 million tonnes (3.5%) was injected. In addition, 0.4 million tonnes were broadcast and ploughed in within 4 hours (2.3%). These technologies reduce ammonia emissions; therefore, the reduction in indirect emissions will not apply to this (we assume that these technologies are used on cropland). Based on UK figures, we estimated that 50% of the 17.2 million tonnes is slurry and the remainder FYM (Defra 2019). Thus, as ammonia emission mitigation technologies are already applied to 5.8 million tonnes (bandspread, injected and ploughed in within 4 hours), the potential for uptake is 2.8 million tonnes.

Assumptions used in the MACC

Table D13. Assumptions used in the modelling

Parameter	Change in value
FracGas	Injection – 0.1 Trailing shoe / hose 0.18
Energy CO2	Reduction in nitrous oxide is offset by increase in CO2
Cost of trailing shoe / trailing hose application	£0.91 m-3 slurry
Cost slurry injection	£2.38 m-3 slurry

Variable rate nitrogen and lime application

Crop-growing conditions are not uniform within a field; while some parts of the field give high yields, in other parts the crops do not perform well. Differences in soil structure, acidity (pH), nutrient content, among other things, can cause such variation. If the variation is considerable, tailored field operations can save resources and enhance the yield. High-resolution sensors, mapping, and decision-making computer systems and variable-rate spreading technologies applied together (in other words precision farming technologies) are capable of varying the rate of inputs applied to soils within one square metre. Variable-rate nitrogen fertiliser application can reduce GHG emissions and GHG emission intensity as these types of fertiliser result in high or equal yield while using the same or less input.

The five main ways they can affect GHG emissions are: increasing yield, reducing nitrogen fertiliser application, reducing tillage and thus increasing soil carbon sequestration, reducing fuel consumption, and reducing other inputs to field operations (impacting off-farm emissions). Maintaining soil pH at an appropriate level is considered important for both maximising crop production and efficient use of fertiliser. Lower crop biomass and higher nitrous oxide emissions in acid conditions can lead to a large increase in emission intensity nitrous oxide (the quantity of nitrous oxide produced per amount of crop produced). Precision lime application takes account of the often large gradients in pH within fields, applying lime with variable rate applicators on a spatial basis according to the lime required to bring soil up to a target pH.

Overview

Nitrous oxide emissions arising from the use of synthetic nitrogen fertilisers can be reduced by more targeted use, supported by a better understanding of spatial heterogeneity in field conditions, linked to technology capable of delivering variable rate fertiliser applications. Precision agriculture technologies (PATs) allow the farmers to consider the field as a heterogeneous entity and apply selective management, potentially increasing efficiency (Aubert et al. 2012). Schwartz et al. (2010) categorised PATs into guidance, recording and reacting technologies.

Guidance technologies (e.g. controlled traffic farming, machine guidance) help to make machinery movement more precise within and between the fields. Recording technologies (e.g. soil mapping, canopy sensing) collect information from the field (including the soil and crops) before, during or after the growing period. Recorded data, in turn, can be integrated to support the use of variable rate nitrogen applications. This can take into account not only in-field variation, but the temporal aspect if in-season information is collected (Diacono et al. 2013). The technology is rapidly developing, and under the H2020 EU research funding scheme there have been more than a dozen projects in recent years working on technological and infrastructure development for precision solutions across farming systems.

Machine guidance technologies are systems that pilot machinery using GPS in order to reduce overlaps of and avoid gaps between passes. At the entry level a GPS receiver mounted on the machinery and a lightbar or an on-board display providing driving direction is needed; with such systems ±40 cm accuracy can be achieved. More advanced solutions, with accuracy up to ±2 cm, use auto-guidance systems (auto-steering) integrated in the tractor's hydraulics and directly control steering. Machine guidance is a prerequisite for VRNT, but could be used in itself (Barnes et al. 2017a).

Variable rate nitrogen technology (VRNT) makes it possible to adjust the application rate to match fertiliser need better in that precise location within the field. Using a digital map or real-time sensors, a decision tool calculates the N needs of the plants and transfers that information to a controller, which adjusts the spreading rate (Barnes et al. 2017a). VRNT applications in crop and grass production can reduce GHG emissions and their intensity as they result in high or equal yield while using the same or less input. The five main ways they can affect GHG emissions are summarised by Balafoutis et al. (2017): increasing yield with while reducing N fertiliser application; reducing tillage and thus increasing soil C sequestration; reducing fuel consumption; and reducing other inputs to field operations (impacting off-farm emissions).

Current commercially available VRNTs adjust N rates on the basis of canopy reflectance measurements using software to model the link to crop N requirement. However, new research is being undertaken on the underlying causes of variable reflectance (and N recovery), which is likely to be soil related. The outcome of this research is likely to lead to new approaches to precision management within the next 5-10 years.

Nitrous oxide accounts for a significant share of the GHG emissions from arable and grassland systems. Soil pH plays an important role in regulating and modifying these nitrous oxide emissions. In more acid soils, there is a higher ratio of nitrous oxide: N₂ emission from denitrification because the nitrous oxide reductase enzyme which converts nitrous oxide to N₂ is inhibited (Liu et al. 2014; Zhu et al. 2019).

Thus, in soils that have a tendency to produce nitrous oxide by denitrification, more acid conditions are likely to lead to higher nitrous oxide emission rates. Because soil acidity can also reduce crop growth, maintaining soil pH at an appropriate level is considered important for both the optimisation of crop production and efficient use of fertiliser inputs (Goulding 2016). Lower crop biomass and higher nitrous oxide emissions in acid conditions can lead to a large increase in emission intensity nitrous oxide (the quantity of nitrous oxide produced per unit of crop). New precision approaches to lime application take account of the often large gradients in pH within fields, applying lime with variable rate applicators on a spatial basis according to the lime required to bring soil up to a target pH. Although this management approach is specifically designed to optimise crop growth through pH management, it is likely that there will be co-benefits in terms of nitrous oxide emission given the sensitivity of emissions to pH. Preliminary measurements highlight the increased emissions of nitrous oxide in the more acidic areas of grassland (figure below). Work is currently underway at SRUC in the UK in partnership with other European countries and AgResearch in NZ to test this hypothesis using conventional and variable rate lime applications on grassland soils, followed by subsequent measurements of nitrous oxide emission during the growing season (http://eragas.eu/research-projects/magge-ph).

Spatial heterogeneity of pH and associated nitrous oxide emissions from a Scottish grassland soil to a depth of 20 cm measured on a 10 by 10 m grid (each square on this map) at the Easter Bush field site in SE Scotland. Measurements of pH provided by Soil Essentials.

Variable-rate lime applications may therefore provide an opportunity to optimise productivity while reducing GHG emissions. The technology is becoming widely available, and although uptake is currently low, it is likely there will be increased adoption of precision liming over the next 10 years.

The measure would require farmers to use machine guidance systems as well as VRNT and variable rate lime application for their arable and temporary grassland field operations, either buying the system, or using contractors for fieldwork who use these technologies. In line with our previous estimates (Eory et al. 2015), we assumed the implementation of a medium accuracy system, capable of 10 cm-accuracy auto-steering and including yield mapping and variable-rate nitrogen application.

Greenhouse gas mitigation summary

As the variety of possible VRNT system specifications is large, and measurements of environmental effects are relatively sparse, currently it is not possible to derive robust quantitative information on the GHG effects of this technology. Eory et al. (2015) derived a central estimate from international studies of 20% N reduction in application, assuming no effect on the yield. Experimental evidence on the N fertiliser use and yield effect shows a large variation, between -57% and +1% and -2% to 10%, respectively. However, from a commercial perspective, it is most likely that a grower would choose to use the same amount of fertiliser N and obtain a high yield when using this technology. Barnes et al. (2017b) found that most potato and wheat farmers in the UK reported a -5% - +5% effect of the technology on N fertiliser and fuel use, and a 5-10% increase in wheat yield. From this information, the abatement here is assumed to consist of a 5% decrease in N use and a 7.5% increase in yield.

Experimental evidence suggests modest yield responses and emission reductions in response to lime. The magnitude of the response depends on the baseline. Liming on acid soils is considered to be a part of good agricultural practice; However recent surveys in Scotland have indicated that 63% of soils have a pH of 6.25 or below and 13% of soils have a pH of 5.5 or below (Edwards et al. 2015). These values lie below the optimum for many crops and are likely to require lime addition to ensure improved crop production.

Costs

We derived the net costs for an average size (120 ha) farm, considering: capital investment in equipment (auto-steer: £5,000 every 5 years; yield monitor: £5,000 every 15 years); maintenance of the equipment (5% of capital cost); signal costs (annual £250); training (£500 every 5 years); and changes in fuel and fertiliser costs and income. We assumed that the costs, calculated at an area basis, would be the same on smaller farms as it is possible to hire contractors to apply VRNT.

The additional cost of the variable rate liming was estimated by adding the cost of soil mapping; £120 ha^-1 (Soil Essentials pers. comm.).

Current uptake and maximum additional future uptake

The measure is applicable on all conventional (fertilised) arable and improved grasslands (i.e. grassland which is fertilised) which needs pH management. To reflect constraints we assumed that it is applicable on 30% of cropland and 20% of grassland. Current adoption of VRNT is around 8% across the UK (Barnes et al. 2017b). However, uptake is probably rather smaller in Scotland (~5%), given the smaller arable farm size. The current uptake of variable rate nitrogen and lime application is estimated to be negligible (0%).

Assumptions used in the MACC

Table D14. Assumptions used in the modelling

Parameter	Change in value
Yield	+7.5%
Synthetic N application rate	-5%
C sequestration	300 kg CO2e ha-1 y-1
Fuel CO2	-3%
Emissions from the use of lime	215.70 kg CO2e ha-1 y-1
EF1 change	-3%
Lime cost	£111 ha-1 (3.7 t ha-1 lime @ £30 t-1) in every 4 years
Lime spreading cost	£10.16 ha-1 in every 4 years
Soil analysis and map cost	£120 ha1 in every 4 years
Training	£500 in every 5 years
Auto-steer	£5,000 in every 5 years
Yield monitor	£5,000 in every 15 years
Signal cost	£250 y-1
Maintenance	5% of capital cost
Change in field operation costs from reduced overlaps	-3%