Abatement of Agricultural Non-Carbon Dioxide Greenhouse Gas Emissions - Chapter 4 - New Zealand's Greenhouse Gas Emissions and the Contribution from Agriculture

Summary
- Table 4.1: Summary of New Zealand's Greenhouse Gas Emissions
4.1 Estimation of Agricultural Methane Emissions
4.2 The Estimation of Nitrous Oxide Emissions from Agricultural Production Systems
4.3 Uncertainty
4.4 Research to Improve the National Inventory
- Table 4.6: Current Research to Improve the National Inventory
4.5 Conclusions
References

Chapter 4 - New Zealand's Greenhouse Gas Emissions and the Contribution from Agriculture

Summary

In 2000, the official estimates of New Zealand's greenhouse gas emissions showed that methane and nitrous oxide represented 59.6% of total emissions in CO₂ equivalent terms, and emissions from agriculture were 55% of total emissions. Compared to 1990, methane emissions were 6.17% less, a fall largely attributable to the fall in sheep numbers, and nitrous oxide was 6.35% more.

The most important source of agricultural methane is ruminant enteric fermentation (98.7%), and the most important sources of nitrous oxide are faeces and urine deposited by grazing animals on pasture, animal waste and fertiliser nitrogen in soils, and indirect emissions from atmospheric deposition and leaching.

Research is leading to the adoption of revised models for estimating methane and nitrous oxide emissions that take into account animal numbers, animal productivity and New Zealand specific emission rates that are based on empirical data. In the decade 1990 to 2000, productivity gains have been significant, particularly in the dairy and sheep industries. The effect of the revised methodology is to show that the official estimates understate emission rates and that there is a rising linear trend in emission rates. Extrapolation of the curves to 2010 indicates that methane emissions will exceed 1990 levels by 15.7% in gross terms, and nitrous oxide emissions will exceed 1990 levels by 20-30%.

The use of nitrogen fertiliser is increasing at an exponential rate. Emission rates are most sensitive to changes in total animal numbers and to productivity. An emerging policy question is how the trade-off will be made between that national objective of maintaining emissions at 1990 levels and individual farmer's objectives for the performance of their livestock. A number of abatement strategy options may confer productivity advantages as well as reduced emissions.

Current and planned research is directed at:

Extending the range of methane emission measurements from animals.
Improving the measurements of nitrous oxide release from excreta under a variety of soil and seasonal conditions.
Improving the assessment of indirect emissions.
Improving (or selecting) the models that can be used to estimate emissions in the future leading to robust estimates that conform with IPCC good practice.
Developing methodologies that allow broad scale assessments of herbage quality and paddock/farm/locality emissions.

It is particularly important that methods can verify that any abatement strategies adopted in the future are delivering the expected effects.

The New Zealand greenhouse gas emissions that have relevance to this report are summarised in Table 4.1.

Table 4.1: Summary of New Zealand's Greenhouse Gas Emissions

Gas	1990		2000		% Change 1990 to 2000
	CO₂ Equiv (Gg)	% total	CO₂ Equiv (Gg)	% total
CO₂	25 266.88	33.17	30 851.78	40.08	22.10
CH₄	35 390.17	46.47	33 204.84	43.15	-6.17
N₂O	11 898.73	16.26	12 654.41	16.44	6.35
Total	73 161.17	100	76 955.61	100	5.19

Source: National Inventory Report for New Zealand, 1999/2000

Total emissions from agriculture were 43 311 Gg CO₂ equivalents in 1990 (59.2% of total) and 41 940 Gg CO2 equivalents in 2000 (54.5% of total). The 2000 figures represent an increase of 0.5% over 1999, but are 3.17% less than in 1990, due in large part to the reduction in sheep numbers.

The revised 1996 IPCC methodology was used in the preparation of the report. IPCC Good Practice Guidance was followed by subjecting the agricultural sector inventory to quality control and quality assurance procedures with scientific peer review. Changes from the previous National Report included adoption of a June figure for animal population statistics, and the emission factor for field burning of agricultural residues was increased from 0.05 to 0.5.

The report identifies the following planned work that will be incorporated in the next national report:

in the methodology for calculating ruminant methane emissions
review of soil derived N₂O emissions factors.

The report notes that the level of uncertainty in the inventory is very high, but no numerical estimates of the level are possible owing to a lack of the means to quantify uncertainty levels for the non-CO₂ sources, particularly N₂O.

4.1 Estimation of Agricultural Methane Emissions

Agricultural methane emissions arise from three principal sources (5^th National Inventory Report). The relative importance of each source is reflected by the percent of total CH₄ emissions from each source:

Enteric fermentation by ruminants that results in methane release by eructation and flatus (98.7%)
Manure management (1.2%) (the estimate excludes emissions from pig and poultry manure - see below)
Field burning of agricultural residues (0.07%).

4.1.1 Estimation of methane from enteric fermentation

Methods of measurement of methane are discussed in Chapter 6.

Estimation of enteric methane emissions from ruminants in the National Inventory Report 1999/2000 is based on a Tier 1/Tier 2 hybrid method using the livestock numbers for the June year averaged over three years, multiplied by emission factors derived from a model of ruminant digestion developed by Ulyatt et al (1991). The model utilises climatic regions, land areas and grassland types to estimate the number of animals grazing particular types of forage, animal types (breeding and other) to estimate liveweight gain curves and the intake of digestible dry matter, and a mathematical model of methane production developed by Baldwin et al (1987) to estimate CH₄ synthesis. The elements of the model are illustrated in Figure 4.1.

Figure 4.1 Elements of the Model Used to Estimate Enteric Methane Emissions from Ruminants

Source: Ulyatt et al, 1991

Clark and Ulyatt (2002) have reviewed the 1991 model with the following conclusions:

The climatic zones, North, Central, East and South with their temperature and rainfall descriptors were retained.
The three grassland types (improved, unimproved and tussock) used in the 1991 model were replaced by a single type since studies by Clark (2001) showed that doing so altered emissions by <1%. A weighted average energy concentration for pastures in each climatic zone and for each animal class was calculated and used for all years.
The animal numbers by species were those for the June year, and were based on census data or MAF estimates in those years where census data did not exist. Rolling three year averages were used as required by IPCC 1996 methodology. Actual production data for each year was derived from three sources (MAF, Livestock Improvement Corporation, The Economic Service). Milk yield data was available on a regional basis, but national figures were used for all regions where regional data was not available.
The energy requirements of the classes and sub-classes of livestock were calculated from production figures using the 1991 model methodology with the exception of sheep (see below). The energy requirements and the energy concentrations of forage were used to calculate dry matter intake, and digestible dry matter intake was calculated using climate zone and seasonal data.
Methane emissions were estimated from daily intakes of digestible dry matter and empirical data of methane emissions from New Zealand animals. This approach replaced the mathematical model of rumen function developed by Baldwin et al (1987) that is used in the 1991 emissions model. Clark (2001) found that the methane production values estimated by Ulyatt (1991) that averaged 32g CH₄ per kg digestible dry matter intake exceeded data from a number of New Zealand experimental observations (see Clark & Ulyatt, 2002) by 15-20%.

Ulyatt (Clark & Ulyatt, 2002) has concluded that methane emissions can be predicted from the amount of methane produced per kg of digestible dry matter intake that depended only on pasture quality. However, young sheep (<12 months old) produce less methane than their adult counterparts on the same ration. Data was only available for sheep and lactating dairy cows. For inventory calculations, goats are assumed to be the same as sheep, beef cattle the same as dairy cattle, and deer an average of sheep and cattle.

The main points of difference in the proposed methodology as compared with the 1991 model presently used for National Inventory calculations are:

Replacement of fixed specific emission factors by a model that uses actual animal numbers, actual production data and empirical data on diet quality.
Livestock performance characteristics for each year are based on actual data.
A methane emission factor that is based on New Zealand empirical data.

When the enteric methane inventory is estimated using the revised model, two contradictory trends emerge. When compared to the official estimate of methane emissions for 1990 and 2000, the revised estimates are 26% and 16% lower, a difference largely explained by the lower estimates of methane emitted per kg of digestible dry matter intake. However, the revised estimates based on actual production data rather than a single emission factor show the effect of significant productivity gains in the sector, in particular in the dairy and sheep sectors with the revised 2000 methane emission estimate being about 7% higher than the revised 1990 estimate.

Davison (2000) has summarised the productivity gains in the sheep industry that have been made over the period 1986/87 to 1999/00 (Table 4.2).

Table 4.2: Productivity Gains in the New Zealand Sheep Industry in the Period 1986/87 to 1999/00

	1986/87	1999/00	Change
Total sheep numbers (million)	67.47	46.08	-32%
Number ewes (million)	47.79	32.20	-33%
Lambing %	97.7	114.30	+16.6%
Number slaughtered (million)	31.63	25.55	-19%
Average weight at slaughter (kg)	13.2	16.48	+25%
Total bone-in meat (000 tonnes)	417.6	421.00	+ 0.8%
Average fleece weight	Estimated genetic gain +0.02kg/ sheep/year

Source: Davison, 2000

The decline in sheep numbers is attributable to the conversion of sheep and beef farms to dairying (an estimated 2 500 000 sheep and beef stock units displaced in the past six years (Davison, 2000)) and to forestry. Nevertheless, it is clear that farmers have achieved productivity gains by growing more pasture dry matter on a smaller area of land.

The dairy industry presents a similar picture (Clark & Ulyatt, 2002). Milking cow numbers have increased by 887 864 and other dairy animals by 249 555 over the decade 1990 to 2000. Milk yields per cow have increased by an average 747 litres and cow weights by 34 kg. Thus the increase in methane emissions from the national dairy herd are in part due to the increase in number that would be taken account of in the official estimates, and in part by the increase in liveweight and milk production that the specific emission factor understates.

Rys (2002) has also summarised the productivity gains made by the sheep, beef cattle and dairy cattle industries over the period 1990 to 2000.

The net effect is that if the revised model to estimate methane emissions is adopted as the official inventory tool, it will show an upward trend in total enteric methane emissions over the period 1990 to 2000. Clark and Ulyatt (2002) found that the revised methane emissions per animal for dairy and beef cattle sheep and deer over the period 1990 to 2000 trend fitted a linear relationship (goats were assumed to have constant emissions over the period), and used this to estimate 2010 emissions by extrapolation. The predicted total and per animal methane emissions are shown in Table 4.3.

Table 4.3: Predicted Total and Per Animal Methane Emissions for 2010 Compared with 1990 Levels

	Total Methane (Gg)	Methane per Animal (kg/year)	Change in Emissions since 1990 (Gg)	Forecast % Change since 1990
Dairy Cattle	446.14	81.9	+220.30	+97
Beef Cattle	237.47	55.7	+7.33	+3
Sheep	483.52	13.3	-130.27	-21
Deer	101.23	24.2	+82.95	+553
Goats	1.75	9.4	-7.9	-72
TOTAL	1270.11		+172.41	+15.7

Source: Clark & Ulyatt, 2002

Predictions of changes in levels are sensitive to two factors, changes in animal numbers and changes in productivity (in particular changes in liveweight and in the weight of product). For example, the predicted emissions from dairy cattle assume an increase of 2 000 000 head of cattle and an increase in methane emission per head of 23% over 1990 figures. By 2000, the increase in dairy cattle numbers over 1990 was 888 000, and it is predicted that continuing conversion of sheep and beef farms to dairy farms will enable increases in dairy cattle numbers of 10 000 per year (Davison, pers comm). Whether this rate will be achieved will depend on the availability and cost of land and the market prices for dairy products. Bodeker (pers comm) considers that the rate of expansion of dairying on to sheep and beef farms will slow, but there is considerable unrealised productive capacity on the newly established farms.

The increased productivity of sheep is compensated for by a predicted continuing decline in sheep numbers as the result of conversion of sheep and beef farms to dairying or to forestry (predicted plantings are 20 000 to 60 000 per year). Deer numbers are predicted to continue to grow displacing sheep and beef cattle, but the rate will depend on market price differentials. Methane emissions per head are also expected to grow. However, actual measurement of methane emissions being undertaken at present will have an unknown impact on future per head estimates.

Clark and Ulyatt (2002) examined the sensitivity of methane emission estimates for 2010 to changes in the levels of animal performance, total animal numbers, and changes in the species mix in animal population estimates. Varying performance increases between 10 and 30%, and using 2010 population estimates, showed that total methane emissions could range from 1 127.71 to 1 322.99 Gg methane, 2.7% to 20.5% higher than 1990 levels. Since a 10% production increase has been obtained in the dairy and sheep industries in the 1990-2000 decade, and 30% is a likely upper limit, a more probable range of production increases of 15 to 25% increase would result in methane emission increases of 7 to 16% above 1990 levels. Methane emission rates were sensitive to changes in total animal numbers, but were relatively insensitive to varying the species mix in population numbers (for example reducing the estimated 2010 numbers of dairy cattle and deer and increasing the sheep numbers by a corresponding amount).

4.1.2 Methane emissions from animal manure management

Methane emissions from grazing animals is estimated by multiplying the estimated faecal dry matter output for dairy and beef cattle, sheep, goats and deer (Joblin & Waghorn, 1994) by emission factors derived from the measurement of the methane output of sheep faeces incubated at 37°C for 20 days. This experimental treatment measures the maximum rate of methane production rather than an actual rate at ambient temperatures. Thus the method probably overestimates actual methane emissions from this source.

It is thought that the faeces of most species shed on pasture decompose aerobically (National Inventory Report 1999/2000). However the form of bovine faeces may facilitate some anaerobic fermentation and the production of methane (Tate, pers comm). Further research is planned to clarify emission rates from the faeces of all animal types.

The National Inventory Report does not include emissions from pig and poultry faeces as no New Zealand specific emission factors are available. No methane emissions from dairy farm liquid manure disposal systems are reported.

4.1.3 Methane emissions from the field burning of agricultural residues

While the National Inventory Report 1999/2000 has changed the "fraction [of] land burned" from 0.05 to 0.5%, burning is not a significant source of trace gases because of the relatively small area devoted to grain crops (1.4% of total farm land), and burning of residues is not a common practice. The burning of tussock grasslands, once a common practice, is now actively discouraged for soil conservation and biodiversity reasons.

4.2 The Estimation of Nitrous Oxide Emissions from Agricultural Production Systems

The National Inventory is constructed using the six alternative animal waste management systems described in the IPCC guidelines. The elements that make up the National Inventory are illustrated in Figures 4.2 and 4.3. The keys to the figures indicate where specific New Zealand partitioning and emission factors are employed. IPCC guidelines and factors for Oceania are used for all other calculations.

The contributions that the various elements make to the nitrous oxide emission profile are summarised in Table 4.4. As would be expected in a country where 88% of agricultural land is devoted to pastoral farming, and where intensive swine and poultry farms make relatively minor contributors of animal waste, the principal sources of nitrous oxide are: emissions from urinary and faecal nitrogen deposited by grazing animals and nitrogenous fertilisers through direct emissions from agricultural soils; and indirect emissions arising from ammonia volatilisation and atmospheric deposition and leaching of nitrate to ground waters, rivers and to the sea. Other animal waste management systems and burning make only minor contributions. The relative inefficiency of use of plant nitrogen by grazing animals is a feature of the New Zealand pastoral system. Haynes and Williams (1993) reported N retention values of 16% on a typical dairy farm and 8% on a typical hill country sheep farm.

Figure 4.2: Sources of Direct Nitrous Oxide Emissions Estimated in the National Inventory

Figure 4.2: Sources of Direct Nitrous Oxide Emissions Estimated in the National Inventory

Source: National Inventory Report for New Zedaland, 1999/2000

Key to Figure 4.2

Fractions

Animal waste (faeces and urine) deposited on pasture by grazing animals. This is made up of the entire nitrogen excretion from non-dairy cattle, sheep, goats and deer, 89% of dairy cattle output and 3% of swine output. The values for nitrogen excretion by dairy cattle, non-dairy cattle, sheep, goats and deer are New Zealand specific. Output calculations are based on McCrae and Ulyatt (1974) for sheep, Ulyatt (pers comm) for cattle, goats and deer. Oceania default values (IPCC, 1997) are used for pigs and poultry.
A mixture of gases is emitted through burning. IPCC values are used to calculate the amount of each gas produced. The fraction of land burned was increased from 0.05 to 0.5 for the 5^th National Inventory.
The area of cultivated organic soils is an estimate since no data is collected. A review by Kelliher et al (2002) indicates that the area of `cultivated' organic soils is larger than the currently used figure, and that the revised IPCC default emission rate of 8 kg N/ha/year be used in the future.
Includes nitrogen from the residues of non-nitrogen fixing crops (FRAC_NCR0), residues of nitrogen from nitrogen-fixing crops (FRAC_NCRBF), but excludes the fraction of the crop residue removed from the field (FRAC_R) and the fraction of residue burned in the field (FRAC_BURN).
Crop biomass multiplied by the fraction of nitrogen in the crop (FRAC_NCRBF).
Nitrogen content of applied synthetic fertiliser, but excludes the fractions that are volatilised (FRAC_GASF) and leached (FRAC_LEACH),
Excreta nitrogen is allocated proportionately to anaerobic lagoons, solid storage and dry lot and other systems. The nitrogen contribution from dairy cattle is based on Ulyatt (pers comm). The % nitrogen allocated to anaerobic lagoons is based on Haynes and Williams (1993). The amount spread on pasture and the amount deposited by grazing animals is considered under "soils".

Emission factors

The emission factor (EF₃PR&P) for nitrous oxide emitted from animal waste deposited on pasture is New Zealand specific based on Carran et al (1995) and Sherlock et al (1995).
The emission rate (EF₂) is the IPCC default rate of 5 kg nitrous oxide N/ha/year. IPCC good practice guidance has updated this rate to 8 kg nitrous oxide N/ha/year since the 5^th National Inventory was completed.
The emission factor (EF₁) is the IPCC rate of 0.0125 kg nitrous oxide N/kg N.
The IPCC emission factors are applied to the proportions of nitrogen in anaerobic lagoons, solid storage and drylot, and other storage systems.

Figure 4.3: Indirect Nitrous Oxide Emissions

Source: National Inventory Report for New Zealand, 1999/2000

Key to Figure 4.3

Nitrogen applied to soils as synthetic fertiliser (N_FERT) and animal wastes (N_AWMS) may be lost to the atmosphere as ammonia and nitric oxides, or lost by leaching or run-off to water bodies.

Pathway 1 represents the fractions of fertiliser nitrogen and animal waste nitrogen that are volatilised, and the proportion that is converted to nitrous oxide.

Pathway 2 represents the proportion of leached nitrogen that is emitted as nitrous oxide from water bodies. A New Zealand specific emission factor based on data from Ledgard et al (1996) and Heng et al (1991) is used.

Table 4.4: Nitrous Oxide Emissions from Agricultural Production in 2000

The ongoing development of a robust, credible inventory based on IPCC Good Practice and Tier 2 methodology remains a high priority. Further, the bases for measuring and estimating the emissions of methane and nitrous oxide need to be sufficiently sensitive to demonstrate the impact of the introduction of abatement measures. There are a number of matters related to the calculation of the National Inventory that are the subject of current investigation. The research programme to support the development of the inventory is discussed in Section 5.4.

4.2.1 The effect of animal numbers and productivity

For faeces and urine deposited on pasture, nitrous oxide emissions are calculated from animal numbers*kg nitrogen excreted per animal*emission factor. The animal numbers are derived from national statistics, nitrogen excretion is based on a limited number of observations, and a single emission factor is used (de Klein et al, 2001; Ledgard, 2002).

An alternative approach is to calculate nitrogen excretion as the difference between nitrogen intake and nitrogen incorporated in the production of meat, milk, fibre etc using a nutrient balance model (OVERSEER) that employs regional livestock population statistics, regional animal production data and regional estimates of pasture dry matter production and nitrogen concentration (de Klein et al, 2001; Ledgard, 2002). When the inventory is calculated using these parameters, nitrogen excretion rates for 1990, 1999 and 2010 (estimated) were 33%, 51% and 58% higher than the official estimates. When the emission factor is applied, nitrous oxide emissions were estimated in 1999 to be 10% higher than 1990 levels and are projected to be 20-30% higher in 2010, whereas official estimates indicate a lesser increase (8%). The upward trend in emissions parallels the observations on methane emissions discussed above and the explanation is identical.

The OVERSEER model is widely used in New Zealand, both as a research tool and as a feed budgeting tool by farmers and their consultants (Ledgard, 2002). Currently, it accommodates environmental nitrogen losses and will be updated in 2003 to include greenhouse gas emissions.

4.2.2 Specific nitrous oxide emission factor

Since 1997, New Zealand has used a single country-specific emission factor to estimate nitrous oxide emissions from excreta (EF_3PR&P) (Sherlock et al, 1997). The question is raised as to whether a single factor can be used to represent all excreta emissions in all regions and soil types, given the extent of variability in agricultural practices and husbandry. de Klein et al (2001) examined nitrous oxide emissions from cattle urine applied to four soils in three different regions and found emissions ranged from <0.5% to 2.6% of the urine nitrogen applied, the difference being largely attributable to soil drainage class.

When the observed rates were applied to national soil drainage class distribution data and a weighted average EF calculated, the result, 0.94%, was close to the EF currently used (1%).

Saggar (2002) is developing a denitrification-decomposition model (NZ-DNDC model) to simulate emissions from soils that takes account of climate, drainage type, grazing and excretal nitrogen inputs.

Nonetheless, there are limited observations on the similarities and differences between cattle and sheep urine (Whitehead, 1995; Barton et al, 2000), a lack of data on the partitioning of nitrogen excreted between faeces and urine, and limited data on the rate of emission from urine as compared to faeces (Yamulki et al, 2000). While it is clear that emission rates are higher from poorly drained soils, it is not yet clear how much influence, if any, the form of nitrogen application has on emission rates.

Most of the observations to date have been made on flat, relatively fertile, land. There is a lack of data on emissions in hill country that is more typical of the land grazed by sheep, beef cattle and deer. In general, this land would be expected to be less fertile with a much tighter nitrogen cycle.

4.2.3 Use of nitrogen fertiliser

While the fixation of nitrogen by pasture legumes remains the most important means of supplying nitrogen to New Zealand soils, the use of synthetic nitrogen fertiliser has been growing steadily over the past decade (Table 4.5).

Table 4.5: Nitrogen Fertiliser Use 1988 to 2002

YEAR	N (tonnes)
1988/89	51 663
1989/90	59 265
1990/91	61 694
1991/92	70 122
1992/93	104 095
1993/94	124 131
1994/95	151 263
1995/96	153 780
1996/97	143 295
1997/98	155 467
1998/99	166 819
1999/2000	189 096
2000/01	248 000
2001/02	279 148

Source: Furness, pers comm, 2002

Furness (pers comm) noted that some caution should be exercised in using these figures, because in some years estimates have been used in the absence of actual figures.

There are a number of reasons for the increased use, for example, increased returns for products, tactical use to overcome seasonal feed shortages, and loss of clover in pasture due to clover root weevil. There is some evidence indicating that it is being used more strategically to ensure a steady supply of forage, and that this accounts for the sharper increase in use in the last three years. Whether the trend will continue, and how much it might be modified by the economics of farming, is difficult to determine. Ledgard (2002) has shown that nitrous oxide emissions from nitrogen fertiliser in 2000 were 117% higher than the 1990 levels and, assuming a linear extrapolation of the rate of increase over the past three years, are projected to be 178% higher in 2010.

The extent of use of nitrogen fertiliser in the future represents an area of uncertainty in estimating future nitrous oxide emissions.

4.3 Uncertainty

The National Inventory 1999/2000 notes:

"The current level of uncertainty in the inventory is very high, although a numerical value for this submission has not been included as there is currently no accurate method of assessing the uncertainty of several of the non-CO₂ key sources - particularly that of nitrous oxide."

Animal numbers used in calculations are based on census and sample data, and do not contribute significantly to uncertainty. The uncertainty of methane emissions estimated according to the Ulyatt (1991) model as used to compile the National Inventory is considered to be about ±20% (National Inventory). The revised method of estimating emissions (Clark & Ulyatt, 2002) is believed to give a more accurate picture of emissions from grazing animals, but is not necessarily less uncertain because of the variation inherent in the biological processes being measured.

The National Inventory does not assign a numerical value to the uncertainty of nitrous oxide emission estimates for the following reasons:

Some of the input data is highly uncertain.
There is inconsistency in the source data for fertiliser use statistics.
Some of the New Zealand specific emission factors are based on limited data - for example, the leaching fraction estimate is based on studies of high nitrogen input land and should be regarded as an upper limit rather than an average value.
The area of organic soils under cultivation is not accurately known.
The nitrogen in New Zealand soils is mostly derived from biological processes - the decay of nitrogen-fixing pasture legumes and animal excreta - and can only be estimated by indirect means.
Some IPCC guidelines and default values do not represent New Zealand conditions or practices.

In their study of the revised method of estimating nitrous oxide emissions, de Klein et al (2001) found that the emission factor for nitrogen from faeces and urine (EF_3PR&P) was most highly correlated with empirical emission rates accounting for 80% of the uncertainty. The emission factor for leachate nitrogen (EF₅) was the next most influential variable accounting for 95% of the uncertainty. New Zealand's N₂O emission rate has a positively skewed distribution reflecting the skewed distribution of soil drainage classes. They are 95% confident that the emission rate is between 23 and 81 Gg per year (mean=44.3 Gg, mode=40.6 Gg).

The adoption of the revised model for estimating enteric methane, and the OVERSEER model for estimating nitrogen output by grazing ruminants, needs to be preceded by completion of the work and its publication in a peer reviewed journal. Clark and Ulyatt (2002) identified the need for methane emission measurement from a wider cross-section of the livestock sector in order to verify and expand the relationship of emissions to herbage quality, feed intake, animal, age and production characteristics. However, the inherent variability of the biological processes that lead to the formation of methane is likely to limit the degree to which measurement error can be reduced. This suggests that measurement of the scale of uncertainty should be a part of research to improve the accuracy of measurements or development of estimation models. de Klein et al (2001) have made such uncertainty estimates in their examination of the emission factor for nitrous oxide from urine.

4.4 Research to Improve the National Inventory

As the figures in Table 3.2 show, there has been a substantial increase in effort to improve the estimates of methane and nitrous oxide emissions from agriculture and this effort will continue (Rys, pers comm). The objectives of the research programme are to develop New Zealand specific Tier 2 emission estimates wherever appropriate, or verify that IPCC default values provide a satisfactory estimate:

Develop robust models for estimation of emissions that reflect New Zealand farming conditions and systems

Reduce the level of uncertainty in emission estimates as far as is practical

Provide for estimating changes in emission rates as the result of the adoption of mitigation strategies.

The funding agencies, FRST, MAF and MfE are assisted in identifying and setting priorities by inter-institutional expert groups (`Methanet' and `N_ZOnet'). The groups also foster collaborative research.

The current research is summarised in Table 4.6.

Table 4.6: Current Research to Improve the National Inventory

	2001/02	2002/03
Nitrous oxide	Determination of the N₂O emission factor from animal urine following application in summer in three regions of NZ (Lincoln University, AgResearch, Landcare)	Determination of the N₂O emission factor from dung and urine following application in spring in three regions of NZ
	Indirect N₂O emissions from leached nitrogen (Crop & Food, AgResearch)	Partitioning of nitrogen excretion by sheep, beef and dairy cattle between dung and urine
	Evaluation of process based models (Crop & Food, AgResearch, Landcare)	Upscaling national N₂O estimates: method evaluation
	Improved estimates of nitrogen intake by grazing animals (AgResearch)	Recalculation of the NZ N₂O inventory using all new emission factors gathered to date and expert judgement for the remainder
		Quantification of N₂O emissions from soils using ¹⁵N isotopomer analysis
Methane	Recalculate the 1990-2000 methane inventory using a modified version of the Ulyatt et al (1991) model and update emission predictions for 2010 (AgResearch)	Methane emissions from growing dairy heifers using SF₆ technique
	Tier 2 methane inventory development (AgResearch)	Methane emissions from grazing deer using SF₆ technique
	Quantify methane emissions from dung deposition on pastures by grazing sheep (AgResearch and others)	Derive generalised relationships between methane emissions from grazing ruminants and feed/animal characteristics
	Develop a methodology to obtain spatial and temporal estimates of herbage quality using satellite imagery (Landcare, AgResearch)	Spatial and temporal estimates of herbage quality
	Paddock and farm-scale N₂O and CH₄ emissions measurement (Landcare)	Re-calculation of the NZ CH₄ emissions inventory using IPCC Tier 2 approach and provision of a prototype computerised method for inventory calculations
	Novel field-based method for measuring N₂O and CH₄ fluxes (Landcare)

Source: MAF; FRST

4.5 Conclusions

It is clear that grazing ruminants are responsible directly or indirectly for almost all of the methane and most of the nitrous oxide emissions from agriculture. The work of Clark and Ulyatt (2002) and Ledgard (2002) demonstrates the sensitivity of the emissions estimates, not only to animal numbers, but also to the effect of productivity gains. These gains have been significant in the dairy and sheep industries in the past decade. Revised inventory calculations, which take both animal numbers and productivity gains into account, demonstrate an upward trend in emissions that is in contrast to the trend demonstrated by the present method of making inventory calculations.

The adoption of abatement strategies, which yield both a reduced emission rate and a productivity gain, such as greater retention of carbon and nitrogen as animal tissue or product, will pose an interesting policy question as to the relative values of the increased production and the cost of emissions and who captures the benefits of productivity gains or emission reductions.

To illustrate this point, the productivity and methane output changes for an `average' dairy cow and an `average' ewe were compared based on the production data of Rys (2002), methane output data of Clark and Ulyatt (2002), and nitrogen excretion data of Ledgard et al (2002).

The difference in milk solids produced per cow between 1990 and 2000 is 68 kg, which at the 2002 announced price of $3.60/kg is valued at $244.80. The difference in methane output per cow is 8.5 kg methane/year or 1785 kg CO₂equivalent/year. The difference in N₂O output is 294g/year or 91.14 kg CO₂ equivalent. At $25/tonne these emissions are valued at $6.74.

As compared to the 1990 ewe that produced 0.98 lambs at an average slaughter weight of 13.2 kg, the 2000 ewe produced 1.14 lambs at an average slaughter weight of 16.48 kg, that is, 5.85 kg more lamb meat valued at $21.90 ($3.75/kg). She also produced 0.5 kg more wool valued at $2.25 ($4.50/kg). The difference in methane output was 1.4kg/year (29.4 kg CO₂ equivalent), and the difference in N₂O production was 45g/year (13.95 kg CO₂ equivalent). Thus the increment in her emissions was valued at $1.08 at $25/tonne.

References

Baldwin RL, Thornley JHM & Beever DE (1987): Metabolism of the lactating cow. II - Digestive elements of a mechanistic model. Journal of Dairy Science, 54: 107-131.

Barton L, de Klein CAM & Sherlock RR (2000): N₂O Emissions from Livestock Urine Applied to Pasture. Report for MfE and MAF. Wellington, New Zealand. 3p.

Bodeker P (2002): personal communication.

Carran RA, Theobold PW & Evans JP (1995): Emissions of nitrous oxide from some grazed pasture soils in New Zealand. Australian Journal of Soil Research, 33: 341-352.

Clark H (2001): Ruminant Methane Emissions: A Review of the Methodology used for National Inventory Estimations. A report prepared for MAF by AgResearch Ltd.

Clark H, de Klein CAM & Newton P (2001): Potential Management Practices and Technologies to Reduce Nitrous Oxide, Methane and Carbon Dioxide Emissions from New Zealand Agriculture. Report for MAF, September 2001. 81p.

Clark H & Ulyatt MJ (2002): A Recalculation of Enteric Methane Emissions from New Zealand Ruminants 1990-2000 with Updated Emission Predictions for 2010. Report prepared for the MAF by AgResearch Ltd.

Davison RM (2000): Genetic Improvement: Importance to Sheep and Beef Farming. New Zealand Society of Animal Production Conference, Hamilton.

Davison RM (2002): personal communication.

de Klein CAM, Sherlock RR, Ledgard SF, Barton L, Kelliher FM, Walcroft AS & Rys G (2002): Nitrous oxide emissions from New Zealand agriculture: research to refine the National Inventory. New Zealand Journal of Agricultural Research, 44: (in press).

Furness H (2002): personal communication.

Haynes RJ & Williams PH (1993): Nutrient cycling and soil fertility in the grazed pasture ecosystem. Advances in Agronomy, 49: 119-199.

Heng LK, White RE, Bolan NS & Scotter DR (1991): Leaching losses of major nutrients from a mole-drained soil under pasture. New Zealand Journal of Agricultural Research, 34: 325-334.

IPCC (1997): Revised 1996 IPCC Guidelines for National Greenhouse Inventories: Volume 3, Reference Manual. Paris, France. IPCC/OECD/IEA.

IPCC (2000): Good Practice Guidance and Uncertainty Management in National Greenhouse Gas Inventories. Paris, France. IGES/OECD.

Joblin KN & Waghorn GC (1994): Estimates of Methane Production from Animal Waste Deposited on New Zealand Pastures. Report for MAF and MfE prepared by AgResearch Ltd.

Ledgard SF, Clark H & de Klein CAM (2002): Improved Estimation of Nitrous Oxide Emissions Using Upgraded Pasture Intake Data for Grazing Animals. Report for MAF prepared by AgResearch Ltd.

Ledgard SF, Clark DA, Sprosen MS, Brier GJ & Nemaia EKK (1996): Nitrogen losses from grazed dairy pasture as affected by nitrogen fertiliser. Proceedings of the New Zealand Grassland Association, 57: 21-25.

McCrae JC & Ulyatt MJ (1974): Quantitative digestion of fresh herbage by sheep: the sites of digestion of some nitrogenous constituents. Journal of Agricultural Science, 82: 309-319.

Muller C, Sherlock RR & Williams PH (1995): Direct field measurements of nitrous oxide emissions from urine affected and urine unaffected pasture in Canterbury. In Proceedings of the Workshop on Fertiliser Requirements of Grazed Pasture and Field Crops: Macro and Micronutrients: 243-347. (Eds) Currie LD & Loganathan P. Occasional Report No. 8, ISSN 0112-9902. Massey University, Palmerston North.

New Zealand Climate Change Project (2002): Fifth National Inventory Report for New Zealand: Greenhouse Gas Inventory 1990-2000 (www.climatechange.govt.nz).

Rys GJ (2002): Climate Change: The Global Problem Requiring a Global Response. New Zealand Society for Animal Production Conference, Hamilton.

Saggar S (2002): Tracing greenhouse gases in dairy pastures. Discovery, Landcare Research, Issue 2, 2002, ISSN 1175-7329.

Sherlock RR, Muller C, Hendriksen SD, Barringer JRF & Cameron KC (1997): Methodology for Assessing Nitrous Oxide and Methane Fluxes from Agricultural Soils. Lincoln Soil Quality Research Centre, Lincoln University, New Zealand. 137p.

Tate Kevin (2002): personal communication.

Ulyatt MJ (1991): personal communication

Ulyatt MJ, Betteridge K, Costall D, Knapp J & Baldwin RL (1991): Methane Production by New Zealand Ruminants. Report prepared for MfE by DSIR Grasslands, Palmerston North, New Zealand

Whitehead DC (1995): Grassland Nitrogen. CAB International, Wallingford, UK.

Yamulki S, Wolf I, Bol R, Grant B, Brumme R,Veldkamp E & Jarvis SC (2000): Effects of dung and urine amendments on the isotopic content of N₂O released from grasslands. Rapid Communications in Mass Spectrometry, 14: 1356-1360.

Contact for Enquiries

MAF Information Services
Pastoral House
25 The Terrace
PO Box 2526
Wellington, NEW ZEALAND

Fax: +64 4 894 0721
Contact this person