Previous Page Table of Contents Next Page

Chapter 6 - Social and economic issues

Factors influencing farming systems
Strategies to achieve sustainability

As emphasized in the previous chapters, the semi-arid tropics are currently in crisis, not only in terms of current agricultural productivity but also in the prospects for sustainable agricultural development. Recently, rainfed agricultural development has been assigned high priority in the plans of many Asian developing countries and it continues to enjoy an important place on the development agenda of many African countries.

Of the many interacting factors contributing to the current crisis, two are fundamental. First, much of the semi-arid tropics (SATs) is characterized by soil and climate not conducive to agriculture. Secondly, the human populations of the SATs have increased rapidly in recent years, thus intensifying the pressure on agricultural resources.

This chapter aims to analyse the socio-economic aspects of sustainable agricultural development, both in general and in relation to farming systems in particular. The socio-economic analysis focuses on four key areas:

1. In many cases, government policies and international markets have, directly or indirectly, reduced the incentives for agricultural production in the SATs so contributing to stagnation.

2. In general, there has been limited success in developing technologies to improve the productivity of semi-arid agriculture.

3. The limited capacity of the non-agricultural sector to provide alternative employment for the increasing population of semi-arid areas.

4. In the absence of technological breakthroughs or favourable policy incentives, the increasing population pressure on semi-arid agricultural resources has created a crisis in the development of sustainable agriculture.

The discussion that follows relies on three main premises:

1. It is essential to understand the reasons for the current situation in order to devise reasonable, practical strategies for improvement.

2. Farmers must be intimately involved in diagnosis of the problems and in devising improvement strategies.

3. Agricultural technologies and policies (and support systems) are complementary means of improving agricultural productivity and sustainability.

FIGURE 42 - Schematic representation of some farming system determinants (Norman et al. 1982)

Figure 1 represents a farming system as three overlapping sub-systems. Figure 42 taken from Norman et al. (1982) expands the socio-economic dimension. The combination of production processes (crop, livestock and off-farm activities) is the farming system. As illustrated in Figure 1, the environment in which farm households make decisions has biophysical and socio-economic elements.

The biophysical elements, dealt with in previous chapters, determine the physical potential, and constraints on, livestock, tree and crop enterprises. The socio-economic elements include exogenous and endogenous factors.

Exogenous factors are those largely out of the control of the individual household such as: (a) community institutions, including structures, norms and beliefs; (b) support services and policies, related to extension, credit, input distribution systems, markets and land tenure; and (c) non-institutional factors, such as population density, location and infrastructure development.

Endogenous factors, on the other hand, are those that the household manages to some degree, including land, labour, and capital.

These inputs and managerial ability differ for each household to affect the performance of its farming system. The household is at one and the same time a production and a consumption unit (Singh et al. 1986). Farming systems are embedded in multi-level agricultural systems. Two higher systems levels relevant to this discussion are community systems and the national agriculture sector (Dixon 1990). Decisions on agricultural resource management are made at all three levels. These interact. For example, sector policies and programmes influence community decisions. These in turn influence household choices. Influences too can work the other way.

The extent to which a farming system fulfils the household goals depends, amongst other things, on managerial skills and, in most semi-arid areas, considerable luck with the weather and other uncertain environmental elements outside household control (Anderson and Dillon 1992). The dynamic aspect of farming systems should also be noted: current farming systems reflect the cumulative interaction of the biophysical and socio-economic elements over time (Norman et al. 1981; Pingali et al. 1987).

Factors influencing farming systems

Households in the SATs
Communities in the SATs
Effects of support services and policies
Capital and cash
Cropping patterns
Interactions between crops and livestock
Other factors influencing farming systems

The conceptual framework above is a convenient way of structuring the discussion on trends and constraints relating to current farming systems in the SATs. Though the socio-economic or human dimensions are emphasized below, the decisions and actions of households are also conditioned by their unfavourable biophysical environment.

Households in the SATs

Development workers define a rural household as a group of people who live together and eat from the same pot, that is, they share the same hearth. Members are usually related (Ellis 1988). Some points related to this definition with particular reference to the SATs (Norman 1992) are:

1. Semi-arid farm households cultivate mainly using their own labour and only small amounts of capital.

2. Individual household choices are often limited by external forces, including community decisions.

3. Household economies are often in transition from traditional systems to 'partial engagement in markets which tend to function with a high degree of imperfection' (Ellis 1988). Thus, it is not surprising that there is sharing and reciprocity between such households.

4. Households in transition consume a proportion of their produce, which gives them some ability to survive independently of the larger economic system.

The goals of farm households, particularly in unfavourable environments like the SATs, are not always easy to identify. It seems, however, that most wish to earn as much as possible with least effort and risk. Some studies on the goals of farm households and farmers' attitudes to risk and uncertainty have been undertaken in the SATs (Binswanger 1980; Bartlett et al. 1982; Walker and Jodha 1986). There is substantial indirect evidence that farmers like to avert risk by the practice of mixed cropping and crop and livestock diversification. In recent years, it has been increasingly recognized that individual farm households are not usually a simple decision-making unit with a single utility function that represents the joint welfare of its members. This has increased the realization that intra-household relationships influence economic behaviour. Four points to note about intra-household relationships in the SATs (Ellis 1988; Feldstein and Poats 1989) are:

1. Internal relationships become particularly important where there is a high degree of congruence between the production and consumption units. They are also important where some imperfection exists in terms of integration with the factor and product markets.

2. The economic role of individuals within households often differs. For example, men and women may have quite independent income streams.

3. Male and female labour cannot be freely substituted one for the other. Division of labour constrains its seasonal availability, the responsiveness to price changes, and influences the composition and volume of farm output.

4. Increases in household income do not usually benefit all household members equally. Technological innovations tend to disadvantage women relative to men.

Understanding the goals of households and their members is important when designing relevant strategies to improve agricultural productivity and sustainability in the SATs (Matlon and Spencer 1985). Knowledge of their attitudes towards risk and uncertainty and the degree to which they are integrated, or are willing to be integrated into the market economy, is a key input in designing such strategies.

Communities in the SATs

In the SATs, communities often manage common property (and resources) and take mutual steps or to preserve the environment. McCorkle et al. (1987) describe many indigenous farming communities that have successfully managed and sustained the productivity of their rangelands over centuries. Though households in a community vary in wealth, there are often patron-client relationships that provide a safety net for the poorest. On the other hand, individualistic behaviour is often sanctioned, including the adoption of modem technologies. Isolated and inaccessible communities are generally less integrated in the factor and product markets so they tend to be more influenced by community structures, norms, and beliefs. Such influences weaken with increased contact with external institutions, for example, greater dealings with the factor and product markets and with government bureaucracies (Bromley 1989). In this way, sanctions, against the individualistic behaviour of households, crumble. The probability of increased differentiation and exploitation in the society increases, and the potential of community-inspired systems to monitor and control natural resource management declines (Jodha 1990). These trends tend to be reinforced by population increases. Some of the developments associated with these trends are summarized as follows:

1. Poverty is becoming individualized. Haswell (1975) terms it 'institutionalization' of poverty. New power groups in the villages (traders, money lenders), often lack the responsibility of traditional patron-client relationships.

2. There is a trend towards individualization of land tenure (Kohler 1968), though in most areas it is not recognized by law. Usufructuary rights to the use of land are increasingly being rented, purchased, and pledged, particularly where population pressure is becoming increasingly heavy (Raynaut 1976; Smith et al. 1994).

3. The traditional systems of communal and reciprocal labour are being replaced by labour paid in kind or, increasingly, in cash, by the job or by the day (Unité d'Evaluation 1978). Even the navetane system, strongly associated with the cultivation of cash crop groundnuts in Gambia and Senegal, is becoming increasingly monetized.

In spite of rapid social changes, traditional hierarchical structures still influence village life so village leaders should be involved in the introduction of agricultural change. It is important, however, that inequalities in living standards within a village are not increased by reinforcing traditional social power with newly-attained economic power.

Unless government makes explicit efforts, common property resources are likely to deteriorate (Jodha 1986) and not be managed for the benefit of the community as a whole. However, it is argued later that community action, control, and regulation need to be increasingly emphasized in strategies to protect common property resources (including grazing land, woodlots, wildlife, water) (FAO 1993) and in controlling degradation and erosion.

Effects of support services and policies

Support services and policies generate incentives for resource management and agricultural production by farm households. An historical perspective is helpful in explaining the type and organization of central institutions in the countries of the SATs.

Policy has had significant influences on semi-arid farming systems in Asia. In China and Mongolia, strong central control led to a surprising uniformity of farming systems until recently. With the introduction of the individual household responsibility in China some 15 years ago, farming systems began to adapt to reflect local resources and other circumstances including market liberalization.

In India, major Government programmes have influenced some farming systems in the SATs. These include the National Watershed Development Programme (Government of India 1990) and various employment guarantee schemes which have propped up farm incomes during drought periods. In a different way, the promotion of well and tank irrigation has re-oriented many rainfed farming systems around small concentrated areas of irrigated land and cash crop production (Walker and Ryan 1990). Another policy change with wide impact has been the liberalization of grain trading between states. The recent coordinated effort to improve oilseed production, under the leadership of the Technology Missions for Oilseeds, has been a classic success for commodity research and development. An interesting byproduct is the adoption of improved crop rotations and cultivation practices for other crops.

In Africa, the different colonial experiences of the Francophone and Anglophone countries have left their mark. Both have much in common, such as commodity-based research programmes, a bias towards research on cash crops during the colonial era (and, to a lesser extent, in the post-colonial era), a general pre-occupation with agricultural productivity issues, and institutionally weak relationships between the social and bio-physical sciences. The differences, however, are more striking. The link, for cash crops, between research and implementation (including extension work) has been much stronger in Francophone countries, where support systems have been strongly coordinated with respect to a individual crops (e.g., groundnuts in Senegal and cotton in Mali).

In general, commodity-based implementation programmes have been less common in Anglophone countries. Consequently, yields of cotton and groundnuts have tended to be larger and there has been greater use of oxen in Francophone countries.

In the 1970s, the popularization of integrated rural development projects embracing both food and cash crops brought about convergence. A major success story of the last 15 years has been the widespread adoption of an improved maize variety in a traditional sorghum and millet area in the wetter part of semi-arid Nigeria (Smith et al. 1994). Similar progress has been witnessed in Kenya. The reasons given for the widespread adoption are: (a) a maize variety that is very responsive to fertilizer; (b) good infrastructural support (good road systems, establishment of integrated agricultural development projects, heavily-subsidized fertilizer, a ready market for the maize); and (c) the fact that it has become a food as well as cash crop for the producers. One of the results (Smith et al. 1994) is that this cash-earning, land-intensive technology has brought widespread use of animal traction, alleviating seasonal labour bottlenecks. Similar secondary effects have also occurred with the adoption of land-intensive technologies for groundnuts and cotton in Senegal and Mali.

Two major objectives have influenced the choice of food policies by African governments in recent decades, particularly in Francophone countries (Stryker 1978), namely, the encouragement of exports of cash-crops to provide much needed foreign exchange and the political desire to provide cheap food for the towns. In general, however, neither approach has met with much success. During the 1980s, exports of cash crops have suffered not only from overvalued exchange rates but also from declining terms of trade on international markets. At the same time, 'cheap food' policies generally acted as a disincentive as far as food production was concerned. Support systems to encourage food production, at least in the African SATs, have rarely been effective.

As a result of increasing pressures for structural adjustments, there has been a trend to allow free market forces to operate. It is likely that the terms of trade for the traditional food crops of SAT Africa (millet, sorghum, and cowpeas) will continue to decline relative to those of the crops that are preferred when incomes go up (rice, wheat, and maize). It seems, therefore, that recent increases in agricultural productivity have occurred in spite of the increasing role of external institutions, and not as a result of their positive influence.

Apart from soil conservation programmes to combat soil erosion and some production-oriented development programmes, particularly in the Francophone countries, relatively few programmes have been implemented with domestic funding to encourage sustainable agricultural development. The Indian National Watershed Development Programme is one exception. Botswana, which is financially prosperous compared to most, is another of the few countries in the SATs with an approved soil conservation strategy. There are, however, even here few links between policy/support systems designed to encourage improvement in agricultural productivity and those designed to ensure sustainable agricultural development. This must be put right if relevant strategies to encourage sustainable agricultural development are to be developed and implemented.

Evidence is emerging from low-income countries that reducing trade restrictions and freeing prices (that is undergoing a structural adjustment process), not only encourages economic growth, but reduces poverty and by doing so, reduces the pressure to over-exploit resources (ADB 1990; De Boer 1993). It is believed, therefore, that freeing the market and reducing market distortions, improves the returns from agricultural activities. This is a key ingredient in implementing effective strategies to address productivity and sustainability simultaneously in low income countries.


Throughout the SATs, many farm households have only usufructuary rights to land. Such land cannot be used as collateral although, as indicated earlier, there are changes towards individual tenure as a result of increased population pressures. This is especially so in Asia and in some African districts with settlement and irrigation schemes or mechanized farms.

In areas with low population pressure, the amount of land farmed has generally been a function of household size (i.e., labour force) and land quality. Traditionally, lighter soils are preferred to heavy ones, though the latter have been cultivated using irrigation during the dry season, particularly in India and to a lesser extent in Africa.

As population density increases, five significant changes are evident:

1. Farm size decreases with the result that the short-term private opportunity cost of leaving land fallow increases. Thus, new ways to maintain soil fertility become more and more important, as the length of fallow and amount of land fallowed have both decreased. However, leaving land fallow does not necessarily ensure that soil fertility is restored. In areas where cash crops are grown; the decline in fertility has been counteracted to some extent through increasing use of inorganic fertilizer.

2. Distribution of land among producers may become an increasing problem. There is some evidence that inequalities are growing, especially in Asia. Such trends have serious implications for the future and, because of the apparent increasing range of farm size, for the differentiation of development strategies to improve agricultural productivity and sustainability.

3. In much of the SATs, farms are being progressively fragmented, particularly where the Maliki law applies in West Africa. Though a degree of fragmentation has some advantages (for example, encouraging equitability in the distribution of land of different qualities, in spreading risk, etc.) it causes problems when mechanization is being introduced or soil conservation strategies requiring cooperation between neighbouring farmers are necessary (World Bank 1992).

4. Poorer quality and heavier land is being brought into cultivation. Poorer land is usually more vulnerable to environmental degradation, and heavier soils are less flexible in use.

5. The opportunity cost of soil moisture is increasing, especially for dry season irrigation. Thus, well irrigation is spreading in the Indian SATs, particularly for cash crops.


Household composition
Total labour inputs

Household composition

A common trend in the SATs is that the traditionally-preferred extended family unit, consisting of more than one married man plus dependents, is breaking up into nuclear or simple units of one married man plus dependents. The underlying reasons revolve around increased contact with the outside world and monetization of the economy. The rate at which this change is taking place depends on several complex interactions. The introduction of cash crops, secular education, increased off-farm employment opportunities, new settlements, and migration may encourage this breakup, although the speed at which it takes place may be tempered by the strength of the traditional hierarchical structure, the ethnic origin of the people concerned, the ownership of cattle and other factors.

Implications of such a trend in West Africa are:

Fields farmed by a household are traditionally divided into common and individual fields. The common fields, controlled by the head of the household, provided food for all members of the household. An increasing proportion of the fields are coming under the control of other individuals in the household. The obligation of household members to work the common fields is decreasing, and the assurance of food from the household farm to meet subsistence needs no longer exists. Increased individualization of fields and the need for cash to pay taxes have both encouraged the growth of market crops Decisions are increasingly made by individuals within sub-households rather than by the extended-household head. This creates problems in introducing improved technology, especially if an extension or institutional credit programme is involved, because such programmes tend to be directed at household heads.

In many areas, the break-up of families is resulting in smaller farms, increased fragmentation of fields and younger, less-experienced household heads. Dependant-per-worker ratios are commonly increasing, giving poorer net worth and cash liquidity levels. Such trends raise questions about the appropriateness of the use of certain types of technology, for example oxen and the question of cattle ownership. Poorer liquidity and net worth are likely to make the purchase of cattle more difficult, and ownership could entail management by herders because of labour limitations.

In Southern Africa, for example Botswana, fragmentation of farm households has taken place and up to 30% of the households engaged in cropping are headed by women. Where senior males are linked with such households, they spend much of their time at the cattle posts, which are generally distant from the cropping areas or, more recently, increasingly work full-time in towns (Heinrich et al. 1990).

Similar trends are evident in marginal, dry, mountain areas, for example, in the hills of Pakistan where up to 70% of household income may be derived from seasonal migration to the towns. Such trends also occur on desert fringes, for example in the Sindh, Pakistan, and in Rajasthan, India. The break-up of the extended family increases the vulnerability of individuals to droughts or other adverse setbacks to the farming system.

Total labour inputs

Farm work in the SATs revolves around crops, livestock, and off- farm enterprises. Because of the seasonably of agriculture, in many areas where population densities are relatively low, labour rather than land is the greater constraint to expansion of production (Unite d'Evaluation 1978; Heinrich et al. 1990). In Asia, in contrast, land has been relatively scarce in the SATs for at least a generation, so farm size tends to be determined by land availability rather than seasonal labour peaks.

The major labour input on the household farm tends to be provided by household members. In West Africa, these are often male adults in contrast to Southern Africa and Asia. Reciprocal and communal labour have slowly given way to a significant level of contract and wage labour.

The total annual work done by household members often appears to be rather small mainly because of seasonally of cropping. Typically, the coefficient of variation for monthly labour inputs increases as one moves into the drier parts of the SATs.


Although the allocation of labour to crops is particularly seasonal, household livelihood strategies tend to spread labour demand over the year. For example, during the dry season, livestock absorbs considerable labour for watering and grazing, and off-farm work is emphasized. Attempts to increase the productive use of labour during the dry season include cultivation using residual moisture or irrigation; and the traditional response of short-season migration. In parts of Southern Africa, for example, Botswana, where rainfall amounts and its dependability both between and within years tends to be unreliable, less effort is made to synchronize seasonal activities. The diverse activities related to crops, livestock and off-farm employment tend to be undertaken independently.

Seasonal labour bottlenecks are characteristic of the SATs and can critically influence the level of agricultural activity of a household throughout the rest of the year. They are influenced by several factors including: the significance of timeliness in operations, the length of the growing season (the shorter it is, the more peaked its labour activity), the type of technology employed, and the power source used. At risk of oversimplification, the following generalizations appear to apply in a West African context:

· Where only hand labour is available, weeding is often considered to be the most demanding operation.

· Introducing improved technology (for example improved seeds and fertilizer) without changing the power source shifts the bottleneck from weeding to harvesting the increased yields.

· A change from hand to animal power, using indigenous technology and ridging equipment, accentuates the weeding bottleneck, and, under certain conditions, the harvesting bottleneck.

· Combining animal power with ridging, planting, and weeding equipment - together with improved land-intensive technology - tends to accentuate the harvesting bottleneck even further.

Farmers use various strategies to alleviate labour bottlenecks. Some well-established ones are: (a) working more days and longer hours per day on farm work at busy periods; (b) reducing time spent on off-farm work; (c) using more women and child labour; (d) hiring labour - though of limited potential, this tends to increase as population density rises (Smith et al. 1994); and (e) growing crops in mixtures.

Unlike India (Walker and Ryan 1990), the African SATs tend, generally, not to have landless labourers in rural areas. The opportunity cost of hired labour, which comes from other farming households, is therefore high. Such labour is often offered because cash liquidity levels are low (Matlon 1977). Thus the negative effects of labour bottlenecks in Africa are probably greater than in India. In India the effects may be positive because bottlenecks provide employment for landless labourers.

Turning to Southern Africa, where animal draught (cattle or donkeys) is traditional, timeliness of ploughing and planting is critical in ensuring operations are undertaken under good soil moisture conditions. This is important in ensuring good stand establishment. Row planting combined with inter-row cultivation using animals is a major advance from the broadcast and hand weeding system traditionally practised (Heinrich et al. 1990).

Seasonal labour bottlenecks have implications when developing strategies to improve agricultural productivity and sustainability. Too often, strategies aimed simply to maximize production per unit area or to increase the area cultivated have led to development of inappropriate technology, that accentuates seasonal bottlenecks (Kafando 1972). Use of animal power to alleviate specific bottlenecks has been most common, but chemical methods (herbicides) and biological improvements (including new varieties and practices) are still in their infancy. The importance of improving labour productivity at busy periods is illustrated by the fact that the marginal productivity of labour at such times can be as much as three to four times higher than the wage rate (Norman 1970). Thus, although water availability and soil fertility may be the major physical constraints on soil productivity and sustainability throughout the SATs, strategies to improve them are unlikely to succeed unless they are compatible with increasing the productivity of labour at times of seasonal bottlenecks.

Capital and cash

Apart from livestock, the capital of small farm households consists largely of goods produced by them directly, such as hand tools, grain stores, etc. Consequently, capital tends to be small. As population density and land use intensity increase, capitalization tends to increase. With the introduction of improved technology, capital tends to increase further and change in character, enabling farmers to buy items such as inorganic fertilizer and animal equipment etc.

In Asia, cash has been most needed to purchase foodgrain in the gap before the next harvest, and for social expenditure on marriage and festivals. This contrasts with West Africa, where most explicit farm expenditures are for non-household labour. When agricultural activity is approaching its peak (June to September) cash resources are at their lowest. Introduction of improved technology is likely, initially at least, to exacerbate this. Traditional sources of credit, often obtained to buy food pre-harvest, commonly incur high implicit, if not explicit, interest rates (Ellis 1988).

Institutional credit tends to be repaid quickly only where programmes are carefully coordinated with other external institutions and support systems, particularly input distribution and product marketing. Certainly, this is true of the introduction of draught oxen systems in West Africa to help grow cash crops (e.g., groundnuts in Senegal and the Gambia, cotton in Mali Sud and the Gombe area of Nigeria, and maize, also in Nigeria).

In India and to some extent in Southern Africa, oxen are an integral part of subsistence farming. In West Africa, however, it is difficult to envisage millet and sorghum, which are mainly food crops, justifying the costs of oxen and equipment. Oxen have been tried in areas without a cash crop but pricing policies, particularly in Francophone countries, together with unimproved technology, have generally made such efforts futile. As indicated earlier, maize grown in northern Nigeria both as a cash crop and for household consumption, and oilseed production in India, are exceptions.

Cropping patterns

A wide range of crops can be grown in the SATs despite the physical and biological limitations. The crops actually grown reflect socio-economic circumstances, both exogenous and endogenous, past and present. Such circumstances need to be understood to appreciate the dynamic, evolving nature of agriculture in the SATs. Understanding can help predict the future (Pingali et al. 1987).

Crop and livestock husbandries have evolved over generations and are often adapted to the environment. Relatively recent, accelerated population increases have tended to upset the process of gradual adaptation. This highlights the need to develop relevant improved technologies to increase productivity and ensure sustainability. It is apparent that many traditional practices can be used as building blocks to develop improved farming systems. Three such practices are particularly important in many parts of the SATs:

1. The use of mounds etc. Whether crops are grown on ridges, mounds, or on the flat depends on many factors, including rainfall, availability of organic matter, culture, soil type, and type of power used (animal, hand, etc.).

2. The 'ring' cultivation system. This system, which used to be popular in West Africa, involves the permanent cultivation of some fields, usually near the compound, where fertility is maintained by manuring. Fields farther away are cultivated for a few years, after which soil fertility is restored by fallowing (Marchal 1977). Increasing pressure on land is leading to a higher proportion of permanently cultivated fields, and the remaining outer fields are being left fallow for progressively shorter periods. There is evidence that the total amount of manure applied increases as the proportion of permanently cultivated land rises.

3. The use of mixed cropping. Crops are grown in mixtures in many parts of the SATs. Norman (1974) shows that though yields of individual crops are often depressed when grown in mixtures, this is more than offset by other crops in the mixture, resulting in a higher return (value) per hectare. In spite of higher total labour inputs, the returns over the year per work-hour and, to a greater extent, per work-hour during the bottleneck period, are usually higher for crop mixtures than for single crops. Crops grown in mixtures at existing technological levels tend to be more profitable, whether land or labour is more limiting. They are also more dependable (Abalu 1976) and are a form of rotation.

In general, traditional agricultural practices have been neglected in the development and dissemination of improved technology. The major programmes undertaken in Asia have concentrated on multi-environment on-farm testing of technologies first developed on research stations. Small-scale applications by non-government organizations are an exception to this generalization.

In African Francophone countries, where cash crops for export have been introduced and their yields substantially increased, much of the crop is grown in pure stands. This is partly because the technology was developed for sole stands and partly because of the success of the external support systems that encourage growing of these crops according to official recommendations. Where improved technology has not been adopted and yields are less improved, such cash crops are often still grown in mixtures, for example as in Nigeria. The practice of mixed cropping still dominates food crops. In India, high-yielding varieties increase the significance of single cropping (Jodha 1977). Only relatively recently has the potential of mixed cropping using improved technology been demonstrated. This could have important implications for ensuring sustainability. It has also been demonstrated that, though the number of crop mixtures decreases with the introduction of animal traction, mixed cropping is not incompatible with animal traction (Unite d'Evaluation 1978).

Increased emphasis on the development of improved technologies for mixed crops appears to be justified. This is particularly so in districts where they are still dominant and in areas where the potential for sequential cropping is limited by the shortness of the rainy season. Complementarity of cropping enterprises will be enhanced when one or more of the following characteristics outweigh the negative competitive effects between species: (a) different growth cycles, (b) different rooting habits, (c) symbiosis between species, (d) compatible labour requirements. Mixed cropping is most effective when the products have varied or multiple uses for both human and animal consumption. For example, in some areas the crop residues from cowpeas (used to feed livestock) may be a more important product than the grain, which is used for human consumption.

Interactions between crops and livestock

Animals as a source of power

Livestock are often an under-estimated part of the farming system. Whilst one-third or more of household income in the SATs may come from livestock they also have multiple uses: (a) they are a form of saving and investment (McIntyre et al. 1992); (b) a source of meat, milk, manure, fuel and other by-products; and (c) a source of draught power.

Ownership of livestock, other than cattle, tends to be widely dispersed, both between and within households. Cattle ownership, on the other hand, tends to be unevenly distributed, being concentrated in wealthier, often more influential, households. In West Africa and, in places, in Southern Africa and the Middle East, cattle ownership and management are often separate. In West Africa management is in the hands of nomadic herders, usually the Fulbe, who also own cattle in their own right.

The potential benefits of some degree of integration between crops and livestock have been recognized traditionally in the SATs. Integration of crops and livestock can, in theory, lead to more efficient use of land unsuitable for crop production. It can provide a use for crop residues and by-products, provide manure, and be a source of income, savings, and investment. In West Africa, this is so even where livestock are owned and managed by nomadic herders, and crops are grown by sedentary farmers. Such symbiosis developed in areas with relatively favourable land/labour relationships.

Increases in population density, however, have forced and are forcing changes in traditional relationships. The diminishing availability of land is resulting in conflicts between herders and farmers, and conflicts about whether resources such as labour and capital should be devoted to products for human consumption or to animal production. There is concern too about declining soil fertility. It is one of the paradoxes of the ever-decreasing land to labour ratio that increasing conflicts between crop or animal production inhibit the benefits of livestock manure in preventing the decline of soil fertility.

Unfortunately, most of the land newly taken into cultivation has been prime pasture land, thus depriving livestock of a relatively sure supply of better quality fodder. This increasingly forces livestock owners to rely upon degraded common pasture (termed 'wasteland' in India) for grazing. Because of overgrazing, such common resources become heavily degraded.

The currently-developing competitive relationships need to be reversed and symbiotic relationships re-established if ecological stability is to be achieved. In the last 15 to 20 years, a major spontaneous trend has developed in the middle belt of Nigeria for many of the nomadic Fulbe to settle with their animals. Here population densities are lower than farther north. Progressive tree cutting has resulted in the southward movement of the zone in which tsetse fly become a problem for livestock.

Livestock have a key role in ecological sustainability in the SATs in the maintenance of soil fertility. The role of manure in the 'ring' system is described above. Its application increases crop yield and improves soil quality. Within the SATs, its use is quite varied and its marginal value product appears to increase with increasing population density (McIntyre et al. 1992). Significantly, in the most densely populated areas where little forest is left, such as the Indian SATs, manure is collected and much is burnt as fuel. One of the drawbacks of manure as fertilizer is its bulk in relation to nutrient content.

It can incur high handling costs even within a farm.

The demand for livestock feed and fodder is substantial in the SATs, but biomass production is small and unreliable. Sown pastures are not economic, but improved use of crop residues appears a possibility. As population density increases, the value of crop residues for use as fuel rises. In the case of fresh milk production, treatment of crop residues to improve digestibility appears economic. In other cases, residue treatment is not attractive. The only successful instance of urea treatment of straw for livestock fattening is in a higher rainfall zone in China.

A special case deserves mention. In Syria, traditional links between nomadic herders and sedentary farmers provides all-important manure for the latter's fields. Here the value of manure outweighs that of the crop residues removed by grazing.

Animals as a source of power

Animals are a traditional source of power in the farming systems of the Asian SATs. Oxen are used as cart and plough animals throughout most of southern Asia, but many small farmers do not own oxen and some are forced to exchange human labour for that of oxen. In contrast, animal traction has been used in the West African SATs for little more than 70 years. Introduction of animal traction can help increase the efficiency and productivity of human resources by the use of equipment designed to maximize the effectiveness of labour during seasonal bottlenecks. The standard idea that draught power is best used to increase the area cultivated (extensification) has been replaced, particularly in Francophone countries, by the concept of its use in intensification to increase soil productivity through manure application, deep ploughing, burial of crop residues, etc. Donkeys and horses as well as oxen are used in the SATs, particularly on lighter soils. In West Africa, animal traction is now closely linked with commercialized cropping. Its successful introduction was often linked to a particular cash crop and the use of improved technologies that give high yields per hectare. This provided the revenue to pay for the equipment and sometimes the animals. It was often complemented by a strong support system embracing an input distribution network, institutional credit, appropriate extension services, and a market for the product.

Many problems have been encountered on the way to the successful adoption of animal traction in West Africa (Faye and Niang 1977). They include: (a) shortage of trained animals and operators, especially for inter-row cultivation; (b) the weakness of draught animals caused by lack of supplementary feed; (c) use of inappropriate equipment; (d) inadequate facilities for repair and servicing of equipment; (e) the non-availability of suitable equipment; (f) under-use of animals during the year as a whole; (g) fragmented holdings that reduce work efficiency; (h) damage to equipment from the large numbers of tree stumps in the fields; and (i) the lack of finance to help farmers hire draught animals. There are two further major problems relating to animal traction in West Africa:

· Prices for cash crops have often increased more slowly than prices of animals and equipment. This has slowed the adoption of draught animals and reduced beneficial interactions between crop production and livestock. Interactions can develop between farmers who own oxen and equipment and those who do not. Households who own draught animals can plough for those who do not and the latter can pay with labour. The potential for exploitation is obvious, especially if such labour is demanded at times when its opportunity cost is high. This will always occur to some degree, but it is most likely when relatively few households own animals. Interestingly, this does not appear to have been a problem in Botswana, where draught animals are traditional. However, those households not owning traction tend to plant late and be less timely in relation to soil moisture. Consequently, their yields tend to be lower.

· Incorporation of residues by deep ploughing after harvest, the cornerstone of past intensification in Francophone countries, particularly Senegal, has not been particularly successful (Hopkins 1974). Although land intensification technologies enabled the adoption of animal traction, many farmers see the use of animals more as a means of extensification. The central problem, however, is that deep ploughing is both time-consuming and power-intensive, while the period available after harvest when the soil is suitably moist is too short for the operation with animal draught.

Other factors influencing farming systems

Although many farm households in the SATs have undoubtedly improved their standard of living over the last 50 years, progress has not been as great as desired. Indeed, there is reason to be seriously concerned about the future. With increased accessibility to the outside world, there have been three trends with important implications when considering future action:

· Though there has been significant migration to the towns, it has been insufficient to check or reduce increasing population pressure of less favoured rural areas in the SATs.

· Because of progressive absorption into the market economy and the movement towards individual rather than shared poverty (i.e., central government rather than community control and responsibility), many farm households have become more vulnerable to drought. Consequently, annual variations in living standards have become more marked, except where incomes have been raised substantially above subsistence level.

· Another trend, also related to increased monetization and individualization of poverty, is increased differentiation in living standards within many communities. A long-standing feature of subsistence farming in many parts of the SATs, particularly West Africa, is the seasonal variation in living standards known as 'hungry gap' (soudure). This is now more open to being exploited; food is often least available when the demands of the agricultural cycle are highest. Two important implications of seasonal hunger are:

- Increased labour effort, except perhaps through changing the power base, is unlikely for many farm households during peak labour demand periods without an improvement of their nutritional levels.

- The hungry gap affects the more disadvantaged households and members of society worst. With the changes from shared poverty and social power to increased individualization and economic power, poor households are becoming more vulnerable to exploitation. Another change with equally severe long-run consequences is for poor farmers to obtain credit before harvest at high rates of interest (Dubois 1975). In the Indian SATs, unlike West Africa, localized shortages do not translate into sharply rising food prices (Walker and Ryan 1990) because of the well integrated nature and the large size of the economy combined with the fact that dry periods tend to be geographically localized.

Hill (1972) and Matlon (1977), referring to the SATs of West Africa, suggest that households have always had a degree of heterogeneity, but laws of inheritance, a relatively egalitarian land tenure system, the availability of surplus land, traditional hand-powered technologies, and community-minded ruling elites led to fairly-even income distribution. Matlon also presents evidence supporting the view that the degree of equality is inversely correlated with the degree of involvement in cash markets and also with village size and population pressure. The same study also provides some evidence that incomes derived from farming, in semi-arid Nigeria, are less variable than off-farm sources of income, though analysis of all sources of income revealed greater equality in income distribution.

The proportion of income derived from farming tends to be higher for poor households than for wealthier ones, though absolute levels are much lower. At the same time lower-income farmers earn a higher proportion of their income as farm labourers for others than higher-income farmers. This is presumably because the poor needed cash income to overcome seasonal problems, and hired farm work is available when their need for additional income is greatest. However, the higher income households tended to participate in more remunerative off-farm work that required some capital (e.g., trading), thus further differentiating themselves from the poor.

The trend towards increasing inequalities in income at village level and the associated seasonal hunger that often arises is obviously of major concern. Is this avoidable? The answer is yes. Up to now technology development has not recognized the heterogeneity of farmers and support systems have been geared towards the better endowed or more influential farmers. However, technology alone cannot be expected to solve all the problems of income distribution; the design and implementation of suitable policies will play a critical role. In the Indian SATs, the dense population and legislative threats to impose land ceilings seem to have reduced farm size and encouraged intensive use of labour and land, no matter what the farm size (Walker and Ryan 1990). On this basis, the same types of technology seem to be relevant for all farmers in the SATs, though some differentiation might be desirable when other factors are taken into consideration (for example, differentiation in terms of accessibility to support systems, cash flows, etc.).

The challenges of the SATs include the general one of identifying sustainable ways to increase agricultural productivity but also the need to recognize the heterogeneity of the farming population when developing strategies. This will enable the livelihoods of all to be improved and become sustainable in the long run. In doing so, it is important to recognize limitations other than those relating to resources, that prevent many farm households from maximizing their technical efficiency. Failure to do so will exacerbate the situation and further inequalities will develop.

Strategies to achieve sustainability

Background - Policies and institutions
Role of livestock
Productivity with sustainability
Agricultural productivity
Ecological sustainability
Need for a coordinated approach

Background - Policies and institutions

It is generally recognized that environmental degradation in high-income countries usually results from wealth, over-development, and waste generation exceeding the assimilative capacity of the environment. In low income countries, however, degradation is often associated with poverty and is caused by depletion of the resource base (Altieri 1988; Mellor 1988; World Bank 1992).

As indicated above and by many others (Oram 1988), farmers in the SATs are compelled to modify traditionally sustainable practices in ways that are environmentally damaging. They do this trying to maintain their short-term standards of living, which are frequently little above the survival level.

In the SATs, rapid population growth, combined with the fact that many people still live in rural areas, creates enormous, ever-increasing pressures on an already stressed ecological environment. The challenge of attaining and maintaining a sustainable agriculture cannot be met without addressing several fundamental issues, which go way beyond the relatively narrow focus of the present discussion. These issues involve the governments of SAT countries in political commitments through the vigorous support of appropriate programmes. These are well articulated in a recent document FAO (1993). They include programmes designed to cut down population growth rates and to create employment opportunities outside the agricultural sector. Whatever happens, given the current population growth and the relatively low absorptive capacity of the non-agricultural sector, it is likely that more rather than fewer people will be trying to derive a living from agriculture for years to come.

Migration of people to less populated areas has limited potential. Even in India, with its well-integrated economy embracing several ecological zones, there has been less migration from poor areas than expected (Walker and Ryan 1990). Where migration has occurred it has sometimes been a very mixed blessing, particularly where movement has been into even more fragile ecosystems (e.g., more arid areas of Niger) rather than wetter areas (e.g., sub-humid zone of Nigeria).

Even assuming that population expansion in rural areas can be tightly controlled, the challenges of ensuring crop sustainability are formidable. If the issues are to be successfully addressed, it will be necessary to widen the focus from simple crop sustainability to a concept of sustainable livelihoods for farm households. As indicated earlier, such households derive their living from a combination of growing crops, keeping livestock and off-farm, income-earning activities. Failure to consider the current and potential, positive and negative, relationships between these three types of activities could seriously limit both the choice of strategies that can be used and the chances of ensuring crop sustainability.

All the issues cannot be discussed here, but off-farm employment could play a key role in solving cash flow problems of farm households, and also provide income independent of the natural resource base. Obviously, the more reliable such income, the greater will be the potential for relieving the pressure on the natural resources and the greater the chances of achieving crop sustainability. Indeed, there is some agreement with Walker and Ryan (1990) that the demand for off-farm labour must accelerate. Unfortunately, government incentives and programmes to encourage off-farm employment outside the urban areas are rare. This is particularly true for employment in the informal sector, in which most of these jobs are found (Liedholm and Mead 1993). Studies show that the potential for stimulating employment in the informal sector is substantial, especially where agriculture is partly commercialized, so two-way links to the non-agricultural sector are already established. There is a need for investment of public funds in training, etc. This needs greater attention by public authorities not only to generate employment but, by doing so, to help relieve pressure on the natural resource base. The difficulties of absorbing any population increases into the non-agricultural sector are illustrated by the relationship, R = P/N where: R represents the rate at which the jobs need to expand to absorb all increases in population, P is the rate of growth in the population/labour force, and N is the proportion of the labour force in the non-agricultural sector. In many SAT countries, P is often about 3% and N often only 0.25, indicating that R is about 12% - an almost impossible task.

In conclusion, two basic points are reiterated, which are the foundation for addressing productivity and sustainability issues:

· Markets should be liberalized allowing market signals to guide allocation of resources and product mixes. This will allow opportunities for comparative advantage to be exploited, thereby encouraging more efficient use of resources and national goals of food self-sufficiency to be replaced by goals of food security at household and national levels. Subsidies need to be given sparingly and selectively and to be designed to bring about congruence between production and sustainability. It is recognized that this is not necessarily the most desirable approach, but until full-resource, cost accounting systems (De Boer 1993) can be implemented, there are few alternatives.

· Ways are needed to re-establish a degree of community responsibility, control, monitoring, and regulation of natural resources. These should include soil conservation strategies requiring community action so that the community as a whole benefits (FAO 1993).

Role of livestock

Although it has been suggested that livestock should be discouraged and, if possible, eliminated from areas in the SATs, particularly those densely populated, it is obvious that such a strategy may not be feasible or indeed rational. Reducing livestock numbers or keeping them penned was advocated on page 77. One implication behind such proposals is that livestock have caused many problems. Others argue that rapid increases in human population and crop cultivation are the major causes (Winrock International 1992).

Many of the inhabitants of the SATs are traditional livestock owners. Coercion to dispose of livestock is likely to meet with resistance, and incentives to do so are beyond the resources of many governments. After all, even in densely populated India, livestock remain an important component of most farming systems.

As land use intensifies and labour requirements rise, animal traction (usually oxen but sometimes donkeys) is required to deal with seasonal bottlenecks (Pingali et al. 1987) This in spite of the increased use of temporary hired labour, sometimes originating from other areas (Smith et al. 1994). The alternative use of tractors is unlikely to be economic, given the rapidly increasing population densities (and smaller farms) and the time-bound nature of some of the tasks in dryland agriculture. This inhibits the development of profitable contract hire or rental markets for expensive tractors (Pingali et al. 1987; Walker and Ryan 1990).

Livestock husbandry is one sustainable use of land that is unsuitable for crop cultivation. Such use depends on improving the pasture quality, while allowing some of its fertility to be transferred to cultivated land through the livestock's dung. Incentive schemes for improved livestock husbandry have generally failed, either because they required fundamental changes in the way livestock owners lived and/or because the incentives were insufficient. As an example, in Northern Nigeria in the 1960s, the nomadic Fulbe were encouraged to operate in grazing reserves, and a market was provided for the milk their animals produced.

As implied above, there is, in terms of resources available to individual households, a degree of complementarity or symbiosis between crops and livestock, which would be lost if livestock were eliminated. The challenge is to try to identify attractive ways that maintain complementarity whilst allowing fundamental changes be determined by market forces. For example, in Nigeria during the last 15 to 20 years, increasing conflicts between nomadic or semi-nomadic livestock owners and sedentary cultivators, have arisen as a result of increasing population densities. This has brought about economic incentives for many Fulbe to settle in the southern part of the semi-arid zone and in the sub-humid zone. Given this situation, it is felt that more intensive ways should be sought to keep livestock through the use of legumes (both grasses and trees), fodder banks, alley farming, and living hedges. Such systems are likely to be attractive, if a secure market can be developed for a product that can be regularly and readily marketed (e.g., milk for urban markets).

Similarly, in India, cattle are not only a key component of the farming system but also have spiritual significance, so development cannot proceed without fall consideration of their multi-faceted role. It is believed that livestock have a future in the SATs. There is, however, a need to identify and implement strategies that will:

· maintain the complementarity between crops and livestock in productive ways that help attain long-term sustainability. Given the human population trends in the SATs, these will undoubtedly evolve in the direction of more integrated, mixed crop/livestock systems (Winrock International 1992);

· encourage both the increased commercialization of livestock husbandry and the adoption of intensification. The off-take rates for meat animals, for example, could be increased. Currently these are often as low as 8% compared with commercial levels of 15%. In other words, market forces may be used to encourage the necessary adjustments.

Productivity with sustainability

Until recently, in most low-income countries, priority has been given to increasing agricultural productivity, in particular that of foodgrain. Lip service has been paid to issues of ecological sustainability, but only recently has this become the focus of donor agency support. The challenge in the SATs is to increase agricultural productivity whilst ensuring ecological, economic and social sustainability.

Governments are preoccupied, because of the poor state of their economies, with the short-term problems of increasing production and do not have the resources or security to worry about long-term sustainability. As indicated earlier, policies often have not even encouraged improvements in agricultural productivity.

The poorer and closer to subsistence level farmers are, the greater the likelihood that their felt needs are those requiring fulfilment in the short term (e.g., producing enough food to survive until next year). Davies et al. (1991) and Frankenberger and Hutchinson (1991) have stressed the critical nature of the relationship between food security and sustainable resource management. Only relatively rarely, where soil erosion is very severe and threatens immediate survival, will farmers be inclined to implement soil conservation measures, without immediate payoff in terms of income (Norman and Douglas 1994).

Agricultural productivity

As will be apparent to anyone familiar with the SATs, adoption of improved technologies resulting in increased output per person as land has become more limiting has been relatively rare. Population-induced land intensification without appropriate technologies to raise productivity substantially is unlikely to increase output per worker or food available per head (Smith et al. 1994).

The development of relevant improved technology is complicated not only by the relatively poor and heterogeneous environment in which the farmers operate (Walker and Ryan 1990), but also by social and economic factors. Figure 43 illustrates the difficult task of developing improved technologies in various parts of the West African savannah. The schematic diagram shows the relationships of five interwoven variables: household goals, market and support system development, population density, market opportunities, and primary technology development requirements.

The complexity of identifying appropriate paths for the improvement of agricultural productivity is reinforced by the following observations:

· Population density affects the emphasis. In areas of sparse population (Areas 1 and 4), labour-saving strategies are more significant, but in densely settled areas (Areas 3 and 6), yield-increasing strategies are required. At intermediate densities (Areas 2 and 5), both technological options must be taken into account.

· Market system development, including the development of a good road and transport system, permits the traditionally important goal of food self-sufficiency to become at least partly diluted in favour of a more commercialized agriculture that involves entering the market place (i.e., food security). In general, however, as is shown by the history of market development in the West African SATs, the markets for improved inputs and input-related services lag behind those for the products. The introduction of improved crop technologies can be slowed in areas where markets for inputs and services are still relatively poorly developed (Areas 1 and 3). Historically, market-system development, particularly on the input side, has been concentrated where rainfall is high enough for cash crops to be grown (Areas 4 and 6). Recently, however, there has been considerable success with maize and oilseeds in the SATs, These initially were cash crops sold domestically to consumers further south but they have now also become a food crop for the producers themselves (Smith et al. 1994).

Though the use of oxen to replace human labour is an obvious attraction, the cost affects their potential in the West African savanna. Where new inputs are part of a technology package, their ability to replace other inputs, or to complement them, must be considered. Because oxen or donkeys have not been used in the traditional food production sector in the West African SATs, labour-saving technologies using animal traction have worked better where there is adequate rainfall for cash crops (Areas 4 and 5) rather than where market system development is poor (Areas 1 and 2). They have also worked better when combined with yield-increasing technologies intrinsically more relevant to locations like Area 6.

FIGURE 43 - Schematic breakdown of relationship between population density and market system development (Source: Norman et al. 1982)

Input delivery systems and input-related services are likely to be more relevant where land is a constraint and land is intensively used. Yield-increasing techniques, including the use of improved seed, fertilizers, and pesticides, make it easier for scientists to develop improved technologies suited to districts like Area 6 than for those like Area 4.

With scientists' current orientation, prospects are not good for developing technologies to benefit farm households in locations such as Areas 1 to 3, where market systems are generally poorly developed. Though it would not increase productivity spectacularly, some progress could be made in such areas by scientists changing their orientation from modifying the environment to fit the plant to modifying the plant to fit the environment.

The greatest challenges lie in Areas 2 and 3 where both marketing systems are poorly developed and the unexploited carrying capacity of the land is small compared with Areas 5 and 6. Although this discussion focuses on the interdependence of support systems (i.e., market-structure development) and improved technologies, there are also interdependencies within villages between the two types of factors. From the communities' point of view, designing and implementing relevant strategies that help all farm households are obviously desirable. Such strategies should include both improved technologies and support systems.

Heterogeneity within the villages must be recognized when designing such strategies. The challenge is to find ways to help disadvantaged farm households. It is easy to design improved technologies suitable for large-scale farmers only, but it is almost impossible to do so just for small-scale ones. Where support systems are limited or there is a hierarchical village structure that causes problems of accessibility, the probability of differential access is greater than elsewhere. The need here is to design a cost-efficient support system that will ensure fair access but not alienate the village leadership.

The above indicates broadly that the relevant improved technologies actually chosen will be determined by the interaction of several variables, socio-economic and physical. In contrast to this differentiation it is important to note that the physical characteristics and constraints that define the SATs provide a degree of commonality to the general approach for developing appropriate technologies. The rainfall patterns and soil heterogeneity pose five important implications (Heinrich et al. 1990):

1. Farmers are likely to adjust their farming plans as the cropping season unfolds. This means that, by using a decision-tree strategy, the area planted, the crops grown, and the practices used can be very different from those originally planned. The implication is obvious: researchers need to mimic such sequential decision-making by developing options from which the farmer can choose, depending on the seasonal circumstances.

2. The heterogeneity of a farmer's soils (and, for that matter, of the resource base) implies the need to target technologies to varied situations to make them more relevant. Such a location-specific approach demands greater emphasis on farmer-based work and farming systems research than is necessary in more uniform, favourable environments.

3. Because of the uncertain and heterogeneous characteristics of the SATs, researchers need to develop considerable amounts of conditional information indicating what should be done 'if this or that happens'.

4. Because of the harsh and heterogeneous nature of the SATs, most technologies are likely to give incremental changes rather than marked jumps in productivity. Unlike the more favourable environments of the Green Revolution areas, much more attention needs to paid in the SATs to the lower rungs on the technology ladder. This is because these often relate to ensuring better timeliness in operations to maximize the return from limited soil moisture. As a result, they require relatively large changes in the farming systems for their implementation. Such 'lumpy' changes are much less likely to be adopted, without substantial cajoling and support, than the 'divisible' types of technology (e.g., improved seed, fertilizer) associated with the Green Revolution. The latter land-intensive technologies only start to become relevant after the 'lumpy' changes have been adopted.

5. Given the environment of the SATs and the precarious level of living of most farmers, it is likely that a high premium will be placed on stability of return (e.g., a certain level of food grain production each year to meet household needs). Most farming households in the SATs are likely to interpret this in terms of self-sufficiency (producing the food themselves) rather than in terms of food self-security achieved by earning enough to ensure adequate food for the household. In the Indian SATs, variation in supplies may result from annual variations in the area planted because of differences in the number of planting rains, rather than from variations in yield per hectare (Walker and Ryan 1990). Although this kind of factor is recognized, researchers still need to concentrate on technologies that ensure reliable yields per hectare, even though they do not give maximum yields in some years. This may be justified because: (a) farmers often articulate stability of yield as an issue; (b) as population densities increase, land will become more limiting, so the variation in area planted is likely to decrease; (c) the marked variation in rainfall pattern in some of the harsher parts of the SATs (e.g., Botswana) means that farmers are likely to suffer wide annual variations in area planted and in yields per hectare.

Ecological sustainability

It is more challenging to attain and maintain ecological sustainability in regions where it is difficult to increase agricultural productivity (e.g., Area 3) than in areas where agriculture has become to some degree commercialized (e.g., Area 6) and where some reliance can be placed on profitable new technologies, market access and external inputs. Even the more favourable areas face problems because:

· appropriate new technologies cannot be devised and introduced fast enough to match ever-increasing population densities, both from local fertility and lowered mortality, and from migration from less favoured areas;

· the trend towards free market economies and reduction of fertilizer subsidies (e.g., more than 80% in Nigeria), credit, etc., which will make external inputs less attractive though perhaps more readily available. One point of concern is that, over the next 20 years or so, increasing scarcity of phosphatic fertilizer and fossil fuels may cause rising prices;

· the continuation of commercially-oriented farming systems is likely to encourage specialization as in some of the maize-growing areas of northern Nigeria where cereal crops are emphasized at the expense of legumes (Smith et al. 1994). In India too, increased integration of the economy, aided by improvements in marketing and infrastructure, has resulted in increased specialization in agricultural activities based on comparative advantage (Walker and Ryan 1990).

Need for a coordinated approach

Preventive measures
Corrective measures
Policy measures

The separate measures generally used to increase productivity and ecological sustainability have not been particularly successful. For example, the approach to problems of soil erosion by soil conservation specialists has often been expensive and fruitless. A new perspective is needed. Three broad kinds of measures need to be implemented to bring congruence to agricultural productivity and sustainability in an acceptable, resource-efficient way (Norman and Douglas 1994). These are preventive, corrective and policy measures. The following discussion focuses the approach on land degradation, but it could equally be applied to other aspects of sustainability.

Preventive measures

These aim to prevent loss of soil productivity by developing and disseminating technologies that yield short-term production benefits and at the same time at least maintain the long-term productivity of the land. This approach involves: (a) the screening of all potential technologies to ensure, as far as possible, that they will have no negative environmental impact; and (b) the development of technologies that have a positive production impact and are also likely to enhance soil productivity in the long run. This implies changes in the types of research to be encouraged.

Much greater emphasis is needed on productive legumes, for human and livestock consumption and for mulching and green manure. Though there have been increases in rainfed production of oilseeds in India, there has not been a general Green Revolution in legume husbandry.

Increased emphasis needs to be placed on technologies with small external inputs. Good examples would be the development of cultivars that improve the efficiencies of the use of soil nutrients and soil moisture, or cultivars that are striga resistant.

Greater attention needs to be paid to constructive exploitation of the biological interactions in traditional farming systems. This implies favourable consideration of biodiversity, nutrient recycling, the role of mixed cropping in reducing soil erosion (Odemerho and Avwunudiogba 1993), alley farming, tree legumes, wind breaks, etc.

Exploitation of biological interactions to attain ecological sustainability is critically important where commercialization of agriculture is very limited. It could also have a key role in reducing reliance on external inputs in districts where agriculture is already highly commercialized.

All this implies changes in the approach to research and extension work:

· Research. The conventional reductionist approach to developing improved technologies needs to be complemented by a more complex participatory systems approach. This is necessary because of the required change in direction from a commodity- to a production-systems or farming systems approach.

· Extension work. Greater emphasis is required on information-based technologies (Hudgens 1992) rather than on material-input technologies. This change from a commodity to a production system emphasis implies a need for re-organized extension services that interact effectively with farmers (Hudgens 1992) or for greater use of farmers as surrogates for extension agents.

Corrective measures

These are mainly physical in nature and have been part of the traditional approach to soil conservation. They place emphasis on physical structures to check further erosion once it has developed or in some case to avoid erosion on land just being opened up to cultivation. Appropriate solutions need to be based on an understanding of the causes of the soil erosion, rather than on the symptoms.

Policy measures

Preventative and corrective measures are not mutually exclusive; indeed they overlap. The thrust of overall policy can be used to coordinate the best approach. Policies can be devised that prevent losses in soil productivity or erosion arising in the first place and also help to check them, once they have developed. This involves designing and implementing policies that: (a) eliminate possible conflicts between the short-term aims of production and measures designed to encourage long-term sustainability (i.e., conservation); (b) use incentives to encourage adoption of strategies that conserve the environment for use by future generations.

The first approach is obviously preferable. Using appropriate technology, it is likely to be more effective in maintaining soil fertility and in promoting good land-use management. The second approach, on the other hand, is likely to be more effective in rehabilitation programmes. A combination of strategies is most likely to maximize the effectiveness of a soil conservation programme.

It is critical for success that programmes designed to sustain or improve soil productivity exploit the complementary relationship between technology and policy. For example, policy measures relating to soil conservation can be influential in encouraging or discouraging the adoption of technologies developed as preventative measures. Heavily-subsidized chemical inputs, for example, may discourage the adoption of other technologies that rely less on chemicals and have a neutral or positive rather than negative impact on ecological sustainability. Implementation of policies that encourage greater community responsibility and control of natural resources could be used constructively in encouraging sustainability. The increasing individualization of tenure occurring in many areas could also be used in this way.

The policies can be used in two other ways to encourage farmers to adopt strategies encouraging ecological sustainability:

· Subsidies for soil conservation measures may make them more attractive to farm households. They are likely to be most applicable in the promotion of corrective measures which usually emphasize physical structures. They are not likely to be very attractive to farm households close to survival level who do not see degradation or erosion as immediate threats. Soil conservation itself offers little in the way of immediate benefits though it may lead to benefits in the future. Finally, severe limits to government budgets are likely to preclude subsidies to strategies that pay off only at some time in the future.

· Cross-compliance policies can be designed to encourage production and conservation simultaneously. This carrot-and-stick approach (Napier 1987) fosters the notion that, if something is taken out of the land to encourage production, something then needs to be put in to sustain productivity in the future. It requires that the farmers participate in a specific conservation practice, if they are to benefit from programmes designed to stimulate production. In the USA in recent years, production and conservation policies have been increasingly linked. For example, government subsidized loans are available to purchase inputs only if the farmer pursues certain conservation strategies. Unfortunately, such policies appear not to be applied currently in low-income countries (Anderson and Thampapillai 1990). Cross-compliance policies may be difficult to put in place for farmers operating near the survival level. On a related point, legislation is often used in high-income countries to control or prevent damaging practices such as overuse of chemicals causing pollution of water supplies. In low-income countries, however, short-run pre-occupations, such as ensuring enough food for next year predominate. Only limited resources are therefore available for enforcement of any approved legislation and many households are unable to pay penalties for any violations. Thus, it is unlikely legislation will be very effective. Consequently, using market forces as far as possible to encourage sustainability is favoured rather than instituting expensive and unenforceable regulations. This could, for example, involve the elimination of overvalued exchange rates and/or the removal of subsidies on credit for the purchase of fertilizer, pesticides, etc.

The combined objectives of all these strategies should be to bring convergence between the private short-term interests of farm households wishing to achieve an adequate current standard of living and the long-term interests of society in maintaining the environment for future generations. The carrot-and-stick approach is likely to be more practicable than encouraging the adoption of conservation strategies through direct soil conservation subsidies. Obviously, if possible, the preventative approach is the least costly and most attractive.


The central theme of this chapter is that conservation strategies will not be successful in the SATs, unless they are closely linked with strategies for increasing agricultural productivity and profitability. This link is necessary for two reasons. Firstly, conservation innovations must be attractive to farmers; and secondly because of the complex interactions of the bio-physical and socio-economic determinants of sustainability (CGIAR 1989; Smith 1993).

SAT development also requires, amongst other things, the efficient operation of market forces and the development of productive, more-sustainable technologies. Technological 'fixes' to ensure long-term sustainability are unlikely. Even in high-income countries, where the standard of living of most farmers is much better, non-market incentives (subsidies) are still applied in a search for sustainability. Any proposed market and profit-oriented strategies need to be supplemented with natural-resource management components that are compatible with long-term sustainability (Hudgens 1992).

For these measures to be effective, development must be assigned a high priority by governments (FAO 1993). Moreover, it is believed that the limited financial resources available in low-income countries will need to be supplemented with resources from donor countries over a period of years. Donor countries have been primarily responsible for the current emphasis on sustainable issues, which is but the latest in a succession of approaches, also mostly promulgated by donors (Ruttan 1991). They should now be encouraged to support the changes in agricultural research and development institutions (Eicher 1990) and the subsidy of policies that encourage sustainability described above.

To conclude, there is no quick fix to improve the situation in the SATs. On the contrary, successful development requires the implementation of a series of well-integrated strategies formulated through interactive consultation of all involved in agricultural development including farmers, researchers, planners, extension agencies, non-governmental organizations and commercial firms. This, together with the very heterogeneity of the farming systems in the SATs and the limited resources available for research and development, encourages strong support for either an interdisciplinary approach involving farmer participation (see Chapter 3, section Field indicators of biological and nutritional problems and Chapter 4, section African examples), or a farming systems approach (FAO 1989a, 1990) to addressing agricultural productivity and sustainability problems in the SATs.

Previous Page Top of Page Next Page