There is concern that agricultural production in developing countries will cause environmental threats in the future, as production will have to increase to satisfy the growing demand for food. Intensification leads to high inputs of nutrients in the form of mineral fertilizers and animal feed. Important parts of these inputs leak from the system in the form of nutrient leaching to groundwater and gaseous losses to the atmosphere. Pressure on the existing agricultural land may increase by growing demand for productive land and degradation of the existing agricultural land base. Expansion of agriculture generally leads to massive deforestation.
The study presented in this report concentrated on the interactions between livestock production, crop production and land use. The link between livestock and crop production is through the demand for animal feedstuffs. This report presents long-term scenarios describing these interactions and the possible consequences for crop production and animal waste production. As the world population is expected to stabilize in the second half of the twenty-first century, the scenarios must cover a period of 50-100 years to include the impacts of human population numbers.
Not all environmental consequences can be quantitatively evaluated. World agriculture is currently responsible for more than half of the atmospheric increase of nitrous oxide (N2O), two thirds of the global ammonia (NH3) input into the atmosphere, and 40% of global methane (CH4) emissions. These compounds play important roles in atmospheric chemistry, ozone depletion, aerosol formation and greenhouse warming. Therefore, a number of examples were selected to be worked out in detail, including the emission of ammonia (NH3) and nitrous oxide (N2O) from animal waste and mineral fertilizers, as well as projections of the emission of methane (CH4) from ruminating animals. A number of other environmental effects related to livestock and crop production are discussed in a qualitative way.
Starting from the AT2010 results, and using a population and per caput demand scenario, we have made a projection of regional domestic demand and self-sufficiency for groups of food products. Three scenarios of agricultural production have been compiled: one medium scenario based on the trends of AT2010, one more optimistic (high) scenario where all growth rates, yield and productivity ceilings were taken slightly higher, and a more pessimistic (low) scenario. The low scenario results in a development where more land is required for crop production and more animals are needed to meet the growing demand. In the high scenario the opposite occurs, with a smaller cropland area and fewer animals needed to achieve the same production level.
Combination of optimistic or pessimistic assumptions on crop and livestock production within a region, and combining regional optimistic or pessimistic scenarios may not be realistic. However, juxtaposing the pessimistic and optimistic scenarios provides a range of different views, one of the most important requirements of scenarios formulated by Alcamo et al. (1995). The land use scenarios have been tested against provisional and incomplete data on the irrigation potential and regional estimates of the potential for agricultural expansion. The scenarios should be tested in a geographically referenced model on the basis of the FAO-Agro Ecological Zones (AEZ) approach (e.g. the IMAGE model) to analyse the feasibility of scenarios of the resulting land use.
The scenario development described in this report clearly reveals the linkages between the production of livestock and crops. Apart from the arable land used to support livestock production through feed crops, there are other effects, such as on fertilizer use. The main conclusions from the study follow:
· If the assumptions on increasing land productivity and the population scenario for the period 1990-2025 are realistic, the arable land area in the developing countries may stabilize or even decrease to a level close to the current one. In the medium scenario, i.e. with crop and animal production increasing at current trends or trends that have been predicted by FAO, the area in use for crop production will decrease between 2025 and 2050. This is caused by the simultaneous slowing down of the growth in demand for agricultural products - as determined by population and economic development - and continued possibilities for increasing the productivity of the land. This conclusion is in line with Alexandratos (1995). It should be noted that potential effects of land degradation on the land's productivity and deforestation are not considered.
· The scenarios for irrigated land are based on the trend predicted for 1990-2010, with a slowly decreasing growth rate in the course of time. For the developing countries, including China, this resulted in a 50% increase for the medium scenario. For the Near East and North Africa regions the assumptions on total crop production had to be adjusted to avoid projections that exceed the land and irrigation potential. For the other regions the future irrigated areas do not exceed the estimates of the irrigation potential based on information available in the late 1970s, although the result of the medium scenario that about half of the total increase in irrigated land will occur in South Asia may be unrealistic.
· There may be a growing demand for land either grazing areas or arable land producing animal feedstuffs required to support livestock production. Currently, about 16% of the domestic demand for cereals, 20% starchy foods and 3% oilseeds comes from livestock production in the developing countries. In addition, part of the production is exported and used as animal feed in developed countries. According to the medium scenario the feed use of cereals may increase to 30% of the total demand in 2050, and similar increases may occur for other crops.
Major increases in the demand for livestock products may occur in the Near East and North Africa. As there is not enough productive land or water in this region to increase the feed and food production sufficiently to meet the projected growth in the demand, this may entail much larger feed imports (Appendix 22). This is on the assumption that livestock production will increase on the basis of imported feedstuffs.
The extent of permanent grassland is not changing rapidly at present: there has been even a decrease of about 9 million ha per year in the developing countries over the past three decades. One may doubt the reliability of the estimates of grazing areas, but they are consistent with the tendency towards decreasing reliance on grazing and increasing importance of fodder crops and feed concentrates noted by Alexandratos (1995). If this tendency continues in the future, the land demand induced by livestock production will increasingly come from feed and fodder production.
· Intensive livestock production is more nutrient- and energy-efficient than more extensive production. According to the medium scenario animal excretion of N, P, and K will double in the coming 5-6 decades. In the high scenario, with growth towards more intensive production, the waste production is much lower and even tends to decrease in the period after 2025. In the low scenario, with less growth in animal productivity, the waste production grows much faster than in the other two scenarios. However, it should be stressed that more intensive systems with more confined animals tend to lead to concentration of production. Systems may be increasingly based on the production of feedstuffs elsewhere, creating problems of animal waste disposal.
As a result of differential growth of the population of the different livestock species and the increasing efficiency of N usage by the animals, the increase in NH3 and N2O emissions is less than would be expected on the basis of growth of total livestock production. However, in some regions, particularly in Asia, the projected growth of production and intensification is rapid and concentration of ammonia emissions may lead to adverse environmental effects such as soil acidification.
· Intensive crop production is more nutrient-efficient than more extensive crop production. The scenarios of fertilizer use show impressive increases, on the average to NPK levels now prevailing in Europe. At present there are large losses of nitrogen from mineral fertilizers. NH3 volatilization to the atmosphere amounts to almost 20% of the total mineral N fertilizer use in developing countries. Because the emissions are related mainly to the type of fertilizer, the large N losses can be avoided. If the loss rates are assumed to decrease to current levels in developed countries (5%), the NH3 emission in developing countries projected for 2025, based on the fertilizer scenario, may be lower than the current emission. According to the medium scenario, N2O emission from mineral fertilizers will increase by a factor of 3 in the coming 5-6 decades.
· Further environmental aspects of the growing fertilizer use have not been assessed. However, leaching and contamination of groundwater by fertilizers and agrochemicals may increase, in particular, in intensive rice growing areas where percolation rates and associated nutrient losses are high. Groundwater contamination as observed at present in Europe (RIVM/RIZA, 1991) may also become a problem in developing countries with intensification of agriculture.
· According to the medium scenario, the methane emissions from enteric fermentation will double in the period 1990-2050. Without major changes in the different waste management systems, the CH4 from animal waste will remain an unimportant global source. However, with increasing intensity and concentration of production, there may be growing disposal problems with more waste storage in lagoons, in liquid form or as slurry; the associated CH4 emission may then become a major global source.
· The CH4 emissions from rice fields may stabilize if emission rates per unit area do not change. However, if the emission rates are proportional to total biomass production, the global CH4 emission from rice paddies may increase further in the coming decades.
By looking into the interactions between the different production systems, the study has helped to lay bare a number of important knowledge gaps and this has resulted in the following recommendations for future studies:
There are few studies on the effect of land degradation on productivity in developing countries. This information is crucial for studies such as this one. If yields are negatively influenced by degradation, the crop production scenarios may not be realized, leading to effects on the arable land areas. Further detailed analysis of the results of the Forest Resources Assessment Project may provide answers to questions about the importance of loss of productivity in shifting cultivation as a driving force of deforestation and other land use changes. In addition, in most developing countries in both dry and humid climates many changes in the state of forest resources are caused by pastoral uses of forests and woodlands with no or insufficient management.
Worldwide grazing areas are known from FAO country estimates of permanent pastures. To complement the understanding of land use dynamics worldwide, geographic information on the extent, productivity and the intensity of use of pastures and arable lands is urgently required. It is interesting to study the possible evolution of forest conversion for livestock production. This seems to be a process occurring in many countries, particularly in the Amazon Basin. This information will also help to quantify the contribution of grazing to animal nutrition.
The most recent information on the irrigation potential is the estimate for Africa and for other regions from FAO (1984). In the near future analysis of the most recent data on fresh-water resources will yield revised estimates for all developing countries.
Because additional uncertainties are associated with possible effects of climatic change, it is more difficult to forecast irrigation potentials. River discharges are extremely sensitive to minor changes in annual rainfall and seasonal distribution patterns. In addition, the seasonal water usage depends very much on the cropping patterns, which may change as a result of adaptation to climate change. A possible way to study future irrigation potentials may be through scenario analysis of climate change and adaptation.
Estimates of the proportions of the various crops used as animal feed are derived indirectly from the supply-utilization accounts. Direct estimates of feed use from the Livestock Environment Study may lead to different estimates of the feed intensity for the various regions. Combination of the direct and indirect approaches may lead to more reliable estimates of feed intensities. Agricultural products that do not enter the market are neither recorded in the FAO statistics, nor in those of many individual countries. Therefore, no data are available on the production of fodder crops or on the extents of land needed to produce them. Information on the use of agricultural residues is very scarce as well. Crop residues may play an important role in animal nutrition, but they may also be burnt. During the burning many polluting compounds are released into the atmosphere. More complete knowledge on the direct use of crop products as animal feed, and the role of fodder's and crop residues would be a major contribution to understanding the interactions between livestock and crop production.
It will be useful to repeat the assessment of fertilizer intensity on the basis of country data. The greatest uncertainty of the model presented in this study is the arbitrarily chosen maximum fertilizer intensity. In addition, the functions used do not describe the accumulation of soil stocks. For example, the required phosphorus inputs may decrease with continued fertilization (Van Duivenbooden, 1995). An alternative to developing scenarios or making projections on fertilizer use is based on yield response functions. Response functions cannot be developed on the basis of total country fertilizer use. Country and crop-specific data on the percentage of the area that is actually fertilized and the fertilizer application rates are needed. A first attempt has been made by FAO/IFA/IFDC (1994), with reported data for 81 countries. Such information should be linked with data on nutrient inputs from animal wastes, crop residues and biological nitrogen fixation, which are important contributors to plant nutrition. These need to be quantified to better understand the observed changes in the fertilizer intensity and to assess strategies to reduce avoidable losses and increase fertilizer recovery.
ABATEMENT OF POLLUTION
No abatement strategies have been taken into account in this study, except for the assumption that higher animal productivity leads to lower CH4 production from enteric fermentation and less waste production per unit product. Strategies to avoid environmental pollution effects from fertilizer use include the promotion of slow-release fertilizers that may increase the N-use efficiency and decrease NH3 volatilization and N2 emission. Similarly, incorporation of animal manure prevents NH3 volatilization and improves the N recovery rate.