Agriculture contributes to over 20% of global anthropogenic greenhouse gas emissions as a result of:

• CO₂ (21%–25% of total CO₂ emissions) from fossil fuels used on farms, but mainly from deforestation and shifting patterns of cultivation;
  
• CH₄ (55%–60% of total CH₄ emissions) from rice paddies, land use change, biomass burning, enteric fermentation, animal wastes;
  
• N₂O (65%–80% of total N₂O emissions) mainly from nitrogenous fertilizers on cultivated soils and animal wastes (OECD, 1998).

Direct emissions of greenhouse gases occur during agricultural production processes from soils and animals and as a result of meeting demands for heat, electricity, and tractor and transport fuels. In addition, indirect N₂O emissions are induced by agricultural activities (Mosier et al, 1998b) and CO₂ also results from the manufacturing of other essential inputs such as machinery, inorganic fertilizers, and agrio-chemicals. Emissions occur at various stages of the production chain and full life cycle analyses are necessary to identify their extent.

In developing countries such as India, emissions mainly arise from ruminant methane, field burning of agricultural residues, and paddy cultivation. Mitigation is difficult to achieve but research into more frequent draining of paddy fields, reduction in the use of nitrogenous fertilizers, and improved diets of cattle is ongoing. Cattle numbers are expected to increase 50% by 2020, which would largely offset any methane avoidance.

As for energy inputs, in many developing countries traditional agriculture still depends on human labour and animal power together with firewood for cooking. Modern agriculture in industrialized countries relies on direct fossil fuel inputs together with embedded energy in fertilizers, and for transport to markets. In the USA each food item purchased has been transported an average of over 2500km (Resources for the Future, 1998) and even further in Europe and Australasia. Recent data for OECD countries suggest the embodied energy in food and drink is 42 GJ per person per year, being 10 times the energy content of the food (Treloar and Fay, 1998).

Increasing energy inputs to meet the growing needs for food and fibre are shown in Figure 3.14. Demand has declined in EITs, increased only slightly in Latin America and Africa in spite of population increases, and increased significantly elsewhere. In developing countries the provision and uptake of “leapfrog” technologies to enable human energy to be replaced by non-fossil fuel energy could be stimulated (Best, 1998).

Table 3.25: Energy use in the agricultural sector in 1995 and annual growth rates in the preceding periods (Price et al., 1998).
	Fuel			Electricity			Primary energy
	Annual growth rate ‘71-‘90	Annual growth rate ’90-‘95	Cons. 1995 (EJ)	Annual growth rate ‘71-‘90	Annual growth rate ’90-‘95	Cons. 1995 (EJ)	Annual growth rate ‘71-‘90	Annual growth rate ’90-‘95	Cons. 1995 (EJ)
OECD Countries	2.3%	1.4%	2.36	1.8%	2.3%	0.20	2.2%	1.6%	2.97
EIT	4.4%	-14.1%	1.04	4.7%	-2.7%	0.22	4.5%	-10.6%	1.71
DCs Asia-Pacific	2.8%	2.4%	1.25	7.9%	8.2%	0.59	4.8%	5.6%	3.03
Rest of the World	3.5%	13.1%	1.05	8.2%	11.6%	0.17	4.7%	12.6%	1.56
World	3.2%	-1.4%	5.70	5.3%	4.6%	1.18	3.8%	0.8%	9.28

The worldwide trend towards energy intensification (GJ/ha) of food and fibre production grown on arable land continues. China, for example, began its “socialism marketing system” recently with the aim of changing agriculture from traditional to more modern production methods (Zhamou and Yanfei, 1998). As a result, total food production on the same land area is projected to rise by around 15% and the standard of living for farmers will be higher but also associated with higher risk. Without greater access to modern energy sources, food and fibre production is unlikely to increase (FAO, 1995). The energizing of the food production chain in terms of quantity and quality is necessary for the attainment of global food security to meet demand for more than one year. To meet the targets of the World Food Summit to reduce the undernourished population to half the current level by 2015, a 4 to 7 fold increase in current commercial energy inputs into agriculture, particularly in developing countries, is anticipated (Best, 1998). In order for agricultural production to be undertaken in a more sustainable manner, one can use husbandry methods and management techniques to minimize the inputs of energy, synthetic fertilizers, and agrio-chemicals on which present industrialized farming methods depend. Any method of reducing these inputs in both developed and developing countries using new technologies must be considered.

Integrated assessment methodologies, which include both direct and indirect energy inputs, have been developed for crops by Kramer et al. (1999) and for milk production by Wells (1999). Both studies analysed the complete production chain up to the “farm gate” and both identified fertilizer inputs as being the major contributor of carbon emissions from the system. For example, manufacturing nitrogenous fertilizers in Germany has specific cumulative energy inputs of around 59MJ/kg of fertilizer (having been reduced by energy efficiency methods from 78MJ/kg in 1970), whereas in the USA they remain at a higher level and have slightly increased (Scholz, 1998).

Spedding (1992) expounded the view that if abundant renewable energy supplies were to become available, then energy would be the only important natural resource since all other natural resources could be generated and all waste streams neutralized. In theory even soil could be considered to be a dispensable resource since crops could be grown hydroponically in nutrient solutions, though in practice this would not be feasible. Agricultural industries can contribute to a more sustainable energy future by providing biomass products. Surplus crop and animal waste products (where not used for soil amendments or fertilizers) can be used as bioenergy sources. Growing crops for energy is well understood, though usually only economically viable where some form of government incentive exists or the environmental benefits are fully recognized (Sims, 1997). Possible conflicts of land use for sustainable food production, soil nutrient depletion, water availability, and biodiversity need to be addressed.

Farming, fishing, and forestry continue to grow in energy intensity to meet the ever-increasing global demands for food and fibre. The present challenge is to offset this trend by introducing more efficient production methods and greater adoption of new technologies and practices. Whilst reducing energy intensity, agriculture must also become more sustainable in terms of reduced nutrient inputs, lower environmental impacts, and with zero depletion of the world’s natural resources such as fish and topsoil. This can only be successfully achieved if practical support is received from primary producers, and this will only occur if other benefits are perceived (Section 3.6.4.5).

As to methane and nitrous oxide, accurate measurement of these anthropogenic greenhouses gas emissions poses challenges as they arise from diffuse sources and wide ranges are quoted. Methane arises from conversion of tropical rainforests to pasture (120-480MtC/yr); rice paddies (120-600MtC/yr); ruminants (390-600MtC/yr); and animal wastes (60-160MtC/yr). Nitrous oxide mainly arises from use of mineral nitrogenous fertilizers (140-200MtC/yr); use of organic fertilizers (140-200 MtC/yr); and deforestation by burning and subsequent cultivation (200-260MtC/yr) (Ahlgrimm, 1998). Another estimate for total agricultural N₂O emissions exceeded 840MtC_eq/yr (6.3 TgN/yr) but also included manure storage, animal droppings on pasture (Oenema et al., 1997), and cultivation of organic soils (Kroeze and Mosier, 1999). OECD regional sources give lower estimates (Table 3.26). Thus there remains a high degree of uncertainty concerning the actual levels of N₂O emissions, since many countries have not yet adopted the IPCC revised 1996 guidelines for national greenhouse gas inventories for emissions from agricultural systems.

Table 3.26: Major sources of methane and nitrous oxide by region in 1995 (MtCeq/yr). (Adapted from OECD, 1998)
Source	Canada	USA	Europe	Japan	EIT	Oceania	World
CH4 animals	2.9	23.6	39.5	1.2	34.6	21.0	438.1
CH4 animal wastes	0.8	11.9	18.7	2.5	12.3	1.6	84.0
N2O fertilizer	2.4	21.0	15.0	0.7	18.9	4.1	112.4
N2O animal wastes	0.8	4.8	8.2	0.6	7.2	1.6	85.6

Strategies for reducing methane emissions from paddy rice and ruminant animals are being evaluated (Yagi et al., 1997), as are techniques to reduce N₂O emissions by better treatment of wastes, improved pasture and animal management, and improved use of nitrogenous fertilizers. Reducing N₂O emissions has to be achieved in areas of intensive agriculture by reducing the N surplus of the system. Improvements in modelling nitrogen and carbon fluxes for agricultural ecosystems have recently been developed (see, for example, Li, 1998, 1999) and applied on the county level for the USA (Li 1995, Li et al. 1996) and for China (Li et al., 1999). By considering the specific interaction between agricultural management with climate and soil conditions, the model simulations have demonstrated large potentials for mitigating N₂O and other greenhouse gas emissions by changing management practices. These include adjusting fertilizer use in poor or rich soils; altering timing of fertilizer or manure applications based on rainfalls; and altering timing and depth of tillage. Based on the modelled results for the USA and China, the most effective way to reduce agricultural N₂O emissions and to ensure adequate crop yields is to optimize fertilizer use in arable soils, particularly those which contain soil organic carbon greater than 3%. However, farmers will need to first accept this management practice if it is to be implemented. A full discussion on the complexities of soil carbon is provided in the Special Report on Land Use, Land Use Change, and Forestry (IPCC, 2000).

There is debate whether such process-based, spatially and temporally integrated models are ready to be used for country inventories and would be better than the IPCC methodology (Frolking, 1998). Considerable uncertainty remains as a result of the sensitivity of underlying assumptions (as discussed at the European Federation of Clean Air and Environmental Protection Associations’ Second International Symposium on Non-CO₂ Greenhouse Gases, Noordwijkerhout, Netherlands, 8-10 September, 1999). Validation by independent atmospheric budget measurements is needed.

Land clearance activities are covered in Chapter 4; the transportation of products from the “farm gate” to the market or processing plant in Section 3.4; and processing of the agricultural, horticultural, forest, or fish products in Section 3.5.

IPCC Homepage