Senior Agricultural Economist, Asia Region, World Bank
The popular image of “desertification” is of a world being engulfed by advancing sand at an alarming rate. However, the evidence suggests that, while there are parts of the arid and semi-arid areas of the globe where such events occur, advancing sand is not the main problem. The main problem is the far less dramatic, but much more widespread, deterioration of soil and vegetation in areas that are often one zone away from the desert fringe itself, and where population pressure has built up faster than agricultural technologies have been able to adapt. It is uncommon for such areas to be overcome by shifting sands, as indeed it is quite uncommon for the deserts themselves to feature active dune systems.
The distorted popular image of desertification, while generally not accepted by serious scientists, has contributed to excessive attention by public administrators on generally uneconomic interventions aimed at stopping mobile sands. This sometimes has been at the expense of more economic efforts to solve the broader problem of land degradation in dry areas. Such efforts to counter degradation increasingly feature a wide range of measures, including policy change, technology improvement, and local participation. The need in developing viable projects is to understand the underlying causes of dryland problems rather than their superficial symptoms and, particularly for the low-income countries, to allocate scarce resources efficiently across the full range of agroeconomic zones.
Desertification can be defined as a loss of land resilience in dryland areas caused, at least partly, by man. (Resilience is the ability to bounce back to previous levels of production when better rainfall returns.) It is extremely hard to measure, because in dry areas, periods of both below average and above average rainfall can persist for as much as 10 to 15 years. The “green” boundary of the Sahel, as seen by remote sensing satellites, pulsates. It advances and retreats in any one location by as much as 200 kilometers between years. To pick up a trend of average desert advancement, such as the 7 kilometers per year often quoted for the Sahel, would take about 30 to 40 years of good data. A number of the earlier reports of irreversible desertification in the Sahel are now believed by some to have been unduly influenced by the fact that the ground observations were made in only a few locations and toward the end of a long dry period. Selecting the base year for comparison is particularly difficult; these areas may never reach a condition of “equilibrium.” For example, parts of the Simpson Desert in Australia saw an increase in the growth of bush vegetation in the mid-1970s due to an unusual sequence of high rainfall. Such bush, having become established, is likely to mature, age, and die over a period of 70 years, creating an “abnormal” vegetation for over half a century. What should be defined as “normal” under such circumstances?
The true causes of land degradation and the loss of resilience in dryland areas are often elusive. Clearly, too many people and live-stock extracting nutrients from too little land is one factor in reducing the important protection of vegetative ground cover, but usually it is a far more complex story than that. Often it is highly location specific. The challenge is to reach past the superficial “trees gone ergo plant trees” syndrome and focus on the opportunities for improved production and resilience, rather than on the negative approach that concentrates on stopping the advance of deserts, real or imagined.
A fundamental question for a poor country is the level of investment that can be justified in dryland areas. Probably, on efficiency grounds, only rather modest levels of investment can be justified. The maximum incremental benefits achievable from investments in dryland areas are generally quite small and also highly variable relative to areas with higher rainfall. A perfectly valid dryland strategy in some countries could even be to concentrate most investments in the less dry areas and rely on growth in those areas to reduce the need for reliance on the more marginal areas, which are poorly suited to productive activities. Some middle-income countries, for example, Tunisia, have already seen a reduction in the population pressure in arid areas due to growth of other sectors, although initially, this can lead to increased degradation, as intensive agriculture structures, such as terraces, break down.
There might be a case for substantial investment in dryland areas on equity grounds, but the correlation between poverty and aridity is not clear. In fact, in the Sahel, one study suggests that people in the driest zone have greater resilience to drought through various adaptive strategies than those in the less dry zone.
The sensible investment approach, which the Bank is currently following, is to commit modest, but important, amounts of financing to foster the creation of a wide range of economic opportunities in drylands. The three keys to progress, reflected in the new strategy to counter land degradation, are technology, policy, and the participation of affected persons.
Technology. Lack of technology remains a serious problem in areas with less than 800 millimeters of rainfall a year. In these areas, a rural family needs a wide range of agricultural technologies that it can adopt fairly flexibly. The reality is, however, that as one passes down the climatic gradient from land with higher potential to dryland areas, alternative technologies are less likely to be available in drier areas. Available economic options are reduced further by high transport and other costs that are a rational response to the low productivity of dryland areas. Also, a package of technologies that may be suitable in one year may turn out to be wrong the next under a different rainfall pattern.
Under such variable climatic conditions even research itself becomes difficult because of the changing impact on experiments of differing rainfall patterns. It, therefore, becomes particularly important in drylands to offer alternatives in the form of a menu of approaches for the farmer to select from depending on the way the season progresses. Some work in West Africa has been done on seasonal rainfall forecasting on the basis of rain patterns at the beginning of the season to help in the formation of such technology menus.
Perhaps the most promising technology in areas that produce rainfed crops is the use of low-cost moisture conservation techniques that combine contour farming, leaving a ridged pattern on the soil surface, with dense rows of vetiver grass planted on the contour or, alternatively, rows of stones. The aim is to slow surface water movement and, through improved percolation, increase the moisture available to the plant. The World Bank has been in the forefront of a shift toward such low-cost techniques for soil and moisture conservation. In pastoral areas we have rather little to offer to substantially raise productivity within the traditional grazing system. However, there are some intriguing hypotheses about the positive impact on rangeland condition and productivity of very intensive management systems, which are being followed by a number of ranchers in the United States. These systems alternate short grazing periods for large numbers of cattle in fenced paddocks with longer, but variable recovery periods, and include careful, almost daily, monitoring of forage. Such systems, it is claimed, improve sustainability because hoof action disturbs the soil and encourages re-seeding, and because heavy, short duration grazing cleans up dead matter that can inhibit further growth. They also ensure that recovery periods are precisely matched to the changing rates of seasonal growth. Such intensive management is a tall order in a situation that involves communally owned range-lands; for example, the Bank-financed Kenya Baringo Semi-Arid Areas Project found it impractical to initiate this approach.
The biggest challenge with technologies in dry areas—apart from the obvious one, that plants do not grow without water—is that the Bank and other development agencies are trying to assist families who not only cannot afford to place a high value on anything beyond the short term because of their poverty but are also highly risk averse in an already highly risky environment. Their problems are exacerbated by a low level of services, such as extension and rural infrastructure, because these families live in low population density, low productivity areas, and this raises their costs, and hence investment risks, still further. Nevertheless, the right technology at the right level of investment can give good returns. A project in Burkino Faso supported by OX-FAM, aimed at developing a contour cultivation system using stone “bunds” or small retaining walls placed on the contour, yielded an economic rate of return of over 30 percent.
Policies. Certain types of agricultural policies may be extremely detrimental to sustainable dryland management. For example, in the Sudan, large-scale mechanized farmers have, for many years, been receiving a tractor fuel allocation on a cultivated hectare basis at a very favorable exchange rate. Therefore, land that is totally exhausted from many years of cropping often does not get rested soon enough. Provided some sort of rough cultivation can be shown to the inspector, a fuel allocation can be obtained and surplus fuel can be sold on the unofficial market at a good profit.
“Getting the prices right” may not always help sustainability in dryland areas. In the Sudan, it is unclear whether a higher price for sorghum would reduce water and wind erosion. It would raise the incremental benefits from a given level of investment in soil conservation, but it would also encourage both expansion of cultivation into more marginal areas and increase the gap between fallow periods thus depleting the soil. Another important policy measure is land legislation. In pastoral areas there may be a need for legislation to assign more secure legal land rights to groups of people who have been able, in the past, to manage rangelands without too much irreversible degradation, but who now are under increasing threat from outsiders. In many such areas, “land grabs” by outsiders are proceeding at an alarming rate.
Participation. Involving local people in designing strategies is perhaps the key to developing a successful and sustainable project. This has been an important element in Bank projects in West Africa, in the OXFAM Yatenga Project and other Bank projects in Burkina Faso, and in the Kenya Machakos Soil Conservation Project, supported over the years by a number of different donors. The increasingly well-developed procedures of Rapid Rural Appraisal provide a useful framework for establishing such participatory approaches. This involves local populations in the early discussions and design of projects. However, it is important to retain within that framework some rigorous analysis of the technical options. This approach has been fostered by the International Institute for Environment and Development, the Ford Foundation, and the Aga Khan Foundation, and has been refined and adapted in the field by local nongovernmental organizations, such as MYRADA in India.
Finally, if there is one underlying principle in designing interventions in dryland areas facing land degradation, it is that projects should pursue the maintenance of resilience in both the land resource and the people. This is perhaps hardly surprising since it is precisely such resilience that plants and animals of arid and semi-arid areas have shown by living in such harsh climates. But it is worth remembering also that migration has not been spurned by animals, or even plants, as one option in a survival strategy.