Global food production at risk

Bo Döös discusses the processes of environmental degradation that could affect global food production. The author has been Professor of Meteorology at the University of Stockholm, Director of the Global Atmospheric Research Programme and Director of the World Climate Programme. He is now a consulting expert on global environmental issues.

To what extent is ongoing and expected environmental degradation likely to affect the global food production system?

Will the future capacity of the global food production system be sufficient to meet the demand of the rapidly growing world population?

Is there a risk that due to acute food shortages in many less-developed countries within the next decades large-scale migrations will be set in motion?

Clearly, it is not possible to answer these questions with any accuracy. Nevertheless, this does not imply that attempts should not be made.

The first question is undoubtedly the simplest one. Even if our knowledge about the state of the global environment is limited, we can make a reasonably reliable estimate of how human activities are currently affecting the capacity of the global food production system.

The second question is much more problematic. Basically, the difficulties we are confronting are of two kinds. One kind of difficulty is that both the capacity of the food support system and the demands on this system are dependent on several factors which have very limited predictability. Examples of such factors are the rate and geographic distribution of social and economic development and the willingness of the nations to take action aimed at reducing the environmental degradation.

The other kind of difficulty is our limited knowledge of the various physical, chemical, and biological processes involved, and of the numerous feedback mechanisms which are at play within the global environment.

A reliable answer to the second question is a necessary but not sufficient condition for answering the third question. The risk of the occurrence of large-scale migration is usually a result of several interactive factors involving, for example, political stability, lack of religious freedom, unemployment and insufficient food purchasing power.

This article will not deal with the migration problem in all its complexities, but discusses the probabilities that insufficient availability of food in certain regions of the less-developed world may be a major contributing factor in triggering mass migrations that are mainly directed to more-developed countries.

There are many factors that have a positive or negative influence on the future capacity of the global food support system. In view of the difficulties in making reliable estimates of the magnitude of these factors, the approach taken here does not induce the most accurate forecast of global food production. Instead, the aim has been to estimate what the future availability of food would be under comparatively favourable circumstances.

This preliminary analysis considers cereal production alone and deliberately:

• underestimates the expected losses caused by various stresses on the system;
• overestimates the possible gains which can be achieved; and
• does not overstate the future demands on the system.

If such a projection indicates that serious insufficiencies are likely to occur in the near future, this would provide valuable guidance about the magnitude and urgency of the required response actions. In fact, even if global food production equalled future demand, a problem still exists due to the fact that possible surpluses of food in some more-developed countries cannot be expected to be readily available in countries with insufficient food production because of their low purchasing power.

Agricultural production is affected in numerous ways by human-induced degradation of the environment: by climate change; acidification; toxification; water and wind erosion; salinization and water shortage; nutrient loss; increased uv-radiation; and the changing incidence of pests and disease. The analysis starts by considering the effect of these processes on global grain production to the year 2025. The results are summarized in Figure 2 below.

The first factor considered is climate change. Given estimates of emissions and atmospheric concentrations of the major greenhouse gases, current climate models can project temperature on a continental scale with reasonable accuracy. Predictions on a sub-continental scale cannot, however, be considered reliable, and precipitation changes are particularly uncertain.

Nevertheless, some conclusions about the possible impacts of a climate change on agriculture have been drawn. For example, the Second World Climate Conference Task Group on Agriculture concluded that the overall effects of climate change and the CO2 fertilization effect on net global food production potential could range from +10% to -20% by about the mid-21st century. For the purposes of this analysis, no net change is assumed as the projections are so uncertain.

Considerable efforts have been made to estimate the magnitude of the different types of human-induced land degradation. Although these estimates do exhibit large differences there is a unanimous opinion that the degradation is severe in many regions, particularly in the developing world, and that it is increasing at an alarming rate.

The rate of soil erosion caused by water and wind exceeds by far its renewal rate of around 5 cm in 100 years — in temperate regions by a factor of 10 to 20 and in the tropics by twice as much. The estimate of the area that becomes unfit for agriculture each year, 5-6 million hectares (ha), according to the Food and Agriculture Organization (FAO) is used as the basis for the calculations here. It is assumed that half this area is used for grain production and that the land is already marginal (i.e. at 2 tons/ha, yields are already depressed below the global mean value of 2.56 tons/ha).

Salinization is, at present, mainly a problem confined to irrigated areas, and its contribution to the degradation of agricultural land is considerably less than soil erosion. In future, however, in connection with the expected rise in sea level and the fact that a considerable part of the arable land is along low-lying coastal areas, salinization is likely to be an increasingly important problem, particularly in developing countries.

According to the FAO, about 2 million ha of land is lost each year due to salinization. It is assumed that half this area is under grain production but that, at 3 tons/ha, yields are relatively high as the land is irrigated and often double-cropped.

There are several kinds of chemical and biotic processes caused by human activities which are expected to contribute significantly to the reduction of agricultural production, either by causing damage to the soil or to crops.

One example of chemical processes is soil acidification. This can cause loss of mineral nutrients and increased concentrations of potentially toxic elements. Ground-level ozone is increasing in concentration and ultraviolet radiation is increasing in intensity, both of which contribute to crop damage. Furthermore, climate warming will inevitably lead to increases in plant pests and diseases.

The loss of production caused by these processes is difficult to evaluate. Lester Brown of the Worldwatch Institute reckons they currently represent an annual loss of about 4 million tons of grain. Given increasing emissions and that more extensive areas will be affected, this estimate could well be substantially exceeded in the near future.

What other influences may affect global food production? Although agricultural land constitutes only a small fraction of the global land area, there is a very limited potential for expansion. Significant potential reserves exist only in sub-Saharan Africa and South America. The best land is, however, already taken. Most of the reserves are more suitable for use as permanent pastures or as forest, the demand for which is also growing. Consequently, the area per capita for agriculture is bound to decline.

The space required for urban and industrial developments and transportation networks will decrease the area available for grain production in the future. According to the Agenda 21 estimates, between 0.1 and 0.25 ha of land per capita is required for development. The estimate of land lost to development presented here assumes the lower end of this range and the present rate of population growth of 100 million a year. Half the area is taken from arable land and, at 3 tons/ha, a higher yield than the global mean is assumed since the lost land is often of above-average productivity.

Agricultural land is, however, expected to be gained at the expense of forest, continuing past trends. As the forest resource is limited, it can be predicted that deforestation rates will decrease over the period to 2025. A value of 10 million ha/yr is assumed, of which a quarter is converted to grain production at a low yield of 2 tons/ha.

When estimating the net change in future grain production, it is also necessary to consider the many possibilities that exist to improve the efficiency of the global food support system with regard to increasing the yield from existing arable land and making the food products available where they are needed.

The most significant enhancement of grain yields could be achieved through increased fertilizer use. It has been estimated that a gain of 21 million tons a year might be possible. Greater emphasis on biological approaches to soil management would be needed, alongside reduced use of mineral fertilizers because of their adverse environmental consequences.

It is assumed that the area under irrigation will continue to increase at its present rate of about 2 million ha/yr though this may be over-optimistic in view of competing demands for water for industrial, domestic and municipal purposes and the impact of climate change.

It is extremely difficult to estimate the gain resulting from biotechnology. It is unlikely that there will be a significant increase over the next one to two decades but over a longer period, say by 2025, the gain in grain production could be as much as 3 million tons a year.

There are many ways in which the efficiency of the global food support system could be improved by raising the quality of management and thereby reducing the extensive losses which occur in all the phases of the food support system. Although it is difficult to make a realistic estimate of the rate at which such losses could be reduced, it is considered that a gain of approximately 5 million tons a year could be achieved.

The following points can be made on the basis of this analysis.

• Over the period 1990-2025, the average value of the expected increase in grain production is estimated to be 37 million tons a year. This figure is likely to be constant over the period to 2025.
• Over the same period, the average loss in grain production is estimated to be 19 million tons a year. The losses will be greater towards the end of the period.
• The average net gain over the period is 18 million tons a year. This is about 1% of present annual grain production. Due to the increase in losses over time, the net gain will diminish in time and by the end of the period will probably be much lower than this value.

The aim of this analysis has been to project what the situation would be under comparatively favourable circumstances. Even with such a modest ambition, there is no doubt that the results obtained are questionable, not least because of the lack of information. However, two conclusions can probably be drawn with some certainty:

• unless steps are taken to reduce environmental degradation and at the same time make more systematic use of existing opportunities to increase grain production, it is likely that the present net gain of about 1-2% a year will decrease to about 0.5% a year within the next few decades; and
• comparing the estimated rate of change in world grain production with the rate of change in world population — 1.73% a year at present to 1% a year in 2025 according to the UN medium variant projection — reveals a widening gap between the expected demand and production of grain.

The overall conclusion from this analysis is that, even with a very optimistic view of the future increase in grain production, production per capita will decrease over the next three to four decades, falling from the present 340 kg a year to well below 300 kg.

Finally, the potential consequences of this projection of world grain production for large-scale migration are considered. Comparatively large-scale migrations have occurred in the past due to various causes. In some instances, they have been initiated through unfortunate circumstances such as repeated harvest failures and agricultural crises or political instabilities. In other cases, the causes have been more complex, and have involved several interactive driving forces of a socio-economic and political nature.

One particular type of cause-effect chain is relevant to the present discussion. Here, the major driving force is rapid population increase in combination with limited capacity for food production. This can lead to environmental degradation and a further reduction in food availability and, in turn, may generate social unrest and migration and thereby cause national and international conflicts. The process may also be initiated at other stages and be further amplified through feedback processes.

A conclusive answer can clearly not be given regarding the risk of large-scale migrations due to food shortages within the next few decades. However, the very pessimistic prospects for increasing food production at the same rate as world population growth and the fact that this is particularly pronounced in certain large, sub-continental regions provide a very definite indication that they are likely to occur.

The problem we are concerned with is thus not so much to judge whether or not there will be mass migrations, but rather to examine the possibilities of predicting where and when they are likely to occur and what their magnitude might be. At present, no models exist which could provide reliable answers to such detailed questions and, in view of the many interactive factors involved, it is not likely that the situation will change in the near future. We have to accept that only very tentative and simplified indications can be given.

• Regions from which major outflows can be expected are those where rapid population growth coincides with low food production and insufficient food purchasing power.
• An additional criterion in identifying vulnerable regions is the expected scarcity of water. Vulnerability to food shortage and emigration could be further enhanced by water shortage, particularly in North and West Africa and in western Asia.
• Predictions of numbers of undernourished people can be made with some degree of accuracy. However, estimates of the outflow of people from the regions which suffer most severely are much more difficult to make.
• Extensive migrations may also be caused by sea-level rise which will affect low-lying agricultural coastal areas. Particularly sensitive will be areas in East Asia where much of the fertile land is near sea level.
• The trajectories of migration are expected in general to be from less-developed countries to more-developed countries. Migration in Africa, which has mainly been of a short-range nature, can in the future be expected to be directed towards Europe and to be of massive dimensions. Migrations from western and southern Asia are also likely to be mainly towards Europe.

What is most difficult to predict are the possibilities the migrants may have of reaching their destinations. The more-developed countries can be expected to be increasingly determined to resist immigrations.

Despite the fact that the consequences of an ongoing degradation of the global environment have been somewhat underestimated in this analysis, the results strongly suggest a decline in the capacity of the global food production system to feed the growing world population. This in turn may lead to mass migrations directed from the less-developed to the more-developed countries. Even if such a projection cannot be considered to be perfectly convincing, the very fact that such a development cannot be excluded indicates the need to carry out a more detailed assessment.

Of course, the need for further research does not imply that our existing knowledge cannot form the basis for decisive action to reduce environmental degradation, to improve land productivity and for the development of a global strategy aimed at reducing the number of undernourished people. This action could significantly reduce the risk of an escalation in the occurrence of famines and migration driven by hunger whatever the future holds in store.

Further information

Bo R Döös, Jordangasse 7/13, A-1010 Vienna, Austria. Fax: 43-1-5350188.

Acknowledgements

This article is extracted from a longer account, "Environmental degradation, global food production and risk for large-scale migrations," published in the March 1994 issue of Ambio.