5

SOUTH ASIA

REGIONAL SETTING

Characteristics of the region

South Asia, as defined in this book, comprises eight countries including Afghanistan1. Agriculture accounts for a significant part of GDP throughout the region, and has grown at a remarkable overall rate during the past 30 years as a consequence of the Green Revolution. Nevertheless, the region has a greater number of undernourished and poor than any other developing region, and more than two-thirds of these reside in rural areas.

The South Asia region contains a population of 1344 million people2, more than one quarter of the population of the developing world. Of these, 970 million (72 percent) live in rural areas. Approximately 150 million households, with 751 million people, can be classed as agricultural3. The combination of high population and limited land area (514 million ha), means that the rural population density in South Asia - at 1.89 persons per ha - is higher than in any other developing region. Moreover, the large proportion of inhospitable terrain has led to the concentration of most of the population on less than half of this land area, resulting in severe pressure on natural resources in many places.

The long history of human settlement has resulted in the utilisation of a wide diversity of natural resources for agriculture. In agro-ecological terms4, 20 percent of the region's land consists of steeply sloping hills and mountains containing only five percent of the total population. Nineteen percent is densely populated, humid or moist subhumid lowland containing the bulk (43 percent) of the region's people; while 29 percent is dry subhumid and still heavily populated, as it contains 33 percent of the population. The remaining 32 percent is semiarid and arid lowland supporting only 19 percent of the region's inhabitants. Hill and mountain areas are found in all the countries, but predominate along the southern slopes of the Himalayan range across India, Bhutan, Nepal, Pakistan and Afghanistan. These hill areas have suffered from particularly extensive deforestation and soil erosion.

The humid and moist subhumid agro-ecological zones, which benefit from seasonal monsoon rains and more than 180 growing days per annum, are located in Bangladesh and around the northeastern, eastern and southern fringes of India, and cover the centre, west and south of Sri Lanka. With large areas of alluvial soils and a high proportion of the land under intensive rice cultivation, these areas support a particularly dense population. The dry subhumid areas, characterised by 120 to 179 growing days each year, cover most of the Deccan Plateau in Central India. The northwest of India, most of Pakistan and Afghanistan are semiarid or arid with less than 120 growing days, low population density and large areas of desert. Throughout the region, there are about 74 million ha of forest (14 percent of total land area), 49 million ha of grazing land and about 213 million ha of cultivated land and permanent crops - equivalent to less than 0.16 ha of agricultural land per capita. Freshwater resources are relatively scarce.

Of the 1.2 billion people worldwide living in dollar poverty, over 43 percent are found in South Asia. Of these, the vast majority live in rural areas. Despite improvements in national food security over the last three decades, benefits have not yet reached the entire population of the region and FAO estimates that 254 million people are still undernourished. Indicators of other dimensions of poverty, such as female illiteracy (59 percent), child mortality (89 per 1000 in children under five years), and child malnutrition (51 percent)5, also point to extensive poverty. The rural poor are particularly vulnerable to droughts, floods and other natural disasters. According to IFAD estimates6, about 66 percent of the vulnerable population in India are small farmers and 2 percent are artisanal fishing families. Women are particularly disadvantaged; female-headed farm households have far lower average incomes than equivalent male-headed farm households.

Of the eight countries in the region, only the Maldives and Sri Lanka have achieved middle income status. Average per capita income is low: with a GDP of US$440 per capita. Official development assistance in 1998 amounted to only US$4 per capita (cf. US$21 per capita in Sub-Saharan Africa) and represented only 0.9 percent of regional GNP. Historically, the agricultural sector has generated the surpluses that have supported the growth and development of other sectors of the economy. This process is most advanced in India and Pakistan, where GADP comprises 27 percent7 and 24.6 percent8 respectively of national GDP. The value added from agriculture in 1999 was 28 percent of regional GDP9. The sector employs 59 percent10 of the labour force and generates 16 percent of the value of total exports11.

Major farming systems in South Asia - MAP

Major farming systems in South Asia

For the purposes of this analysis, eleven broad farming systems have been identified. The geographical distribution of nine of the farming systems is indicated in the accompanying Map. The Urban Based and Tree Crop Farming Systems are not mapped. The main characteristics of the major farming systems, including the land area and agricultural population as a proportion of the regional total, principal livelihoods and prevalence of poverty, are shown in Table 5.1. A brief description of each farming system appears in the following paragraphs, and four are analysed in greater depth in subsequent sections.

Rice Farming System

This farming system is dominated by intensive wetland rice cultivation12 by farmers and sharecroppers in fragmented fields with or without irrigation. Of the total system area of 36 million ha, an estimated 22 million ha - or more than 60 percent - is under cultivation. Some 10 million ha, or 43 percent of the cultivated area, is irrigated. Of the total system population of 263 million inhabitants, 130 million are classified as agricultural (17 percent of the regional total). The system is concentrated in Bangladesh and West Bengal, but smaller areas are found in Tamil Nadu and Kerala States of India, and Southern Sri Lanka. The system contains 50 million bovines, used for draft Power, milk and manure, and considerable number of small ruminants. Poor farmers operate extremely small areas, and rely on off-farm income for survival. Poverty is extensive and also quite severe.

Coastal Artisanal Fishing Farming System

In a narrow band along the major part of the coast of Bangladesh and India, and around the Maldives, households supplement artisanal (inshore) fishing with food production - often rice and such cash enterprises as coconuts, livestock and vegetables. The main livelihood is threatened by over-exploitation of the common resource, both locally and by larger well-equipped fishing boats. Total land area is estimated at five million ha with nearly half under cultivation. Coastal land resources are also under pressure from the high population density along the coastline, and from the expansion of modern and capital intensive acquaculture enterprises. One third of the cultivated area, or 0.8 million ha, is under irrigation. Of the total system population of 45 million, about 18 million are classified as agricultural. Off-farm income constitutes a significant source of livelihood, especially for poor households. The management of the system is complex from many points of view - the complexity associated with lying on the interface between marine and terrestial ecosystems is aggravated by the complexity of numerous stakeholders with conflicting objectives. Poverty across the system ranges from moderate to extensive.

Rice-Wheat Farming System

Characterised by a summer paddy crop followed by an irrigated winter wheat crop (and sometimes also a short spring vegetable crop), the Rice-Wheat Farming System forms a broad swathe across Northern Pakistan and India, from the Indus irrigation area in Sindh and Punjab, across the Indo-Gangetic plain to the northeast of Bangladesh. Total system area is 97 million ha with an estimated 62 million ha - more than 60 percent of the land of the system - under cultivation. An estimated 48 million ha, or 78 percent of the cultivated area, is irrigated. The system has a significant level of crop-livestock integration, with an estimated 119 million bovines which produce draft power and milk, as well as manure for composting. Around 73 million small ruminants are kept, principally for meat. Of the total system population of 484 million people, 254 million are classified as agricultural. The Rice and Rice-Wheat Farming Systems together contain 40 percent of the cultivated land in the region and produce the bulk of the marketed foodgrains that feed the cities and urban areas of South Asia.

Highland Mixed Farming System

This farming system, generally intermediate between the rice-wheat plains of the lowlands and the sparsely populated high mountain areas above, extends across the entire length of the Himalayan range, from Afghanistan to the extreme northeast of India, as well as in isolated areas of Kerala and Central Sri Lanka. Major products include cereals, legumes, tubers, vegetables, fodder, fodder trees, orchards and livestock. Total system area is 65 million ha with an estimated 19 million ha - about 29 percent - under cultivation. While most cultivated land is rainfed, an estimated 2.6 million ha, or 14 percent, is irrigated. There are about 45 million bovines and 66 million small ruminants. Of the total human population of 82 million, nearly 53 million are classified as agricultural. The prevalence of poverty, which is aggravated by remoteness and the lack of social services, is rated between moderate and extensive.

Rainfed Mixed Farming System

This predominantly rainfed cropping and livestock farming system occupies the largest area within the sub-continent and, with the exception of a small area in Northern Sri Lanka, is confined entirely to India. Total system area is 147 million ha with an estimated 87 million ha (59 percent) under cultivation. Rice and some wheat are grown, as well as pearl millet and sorghum, a wide variety of pulses and oilseeds, sugarcane, and vegetables and fruit. An estimated 14 million ha, or 16 percent of the cultivated area, is irrigated. There are an estimated 126 million bovines and 64 million small ruminants, which are partially integrated with cropping. Of the total system population of 371 million, 226 million are classified as agricultural. In many instances, relatively small areas are irrigated from reservoirs and in recent decades, tubewells have contributed to an elevated level and stability of cereal production. Vulnerability stems from the substantial climatic and economic variability. Poverty is extensive and its severity increases markedly after droughts.

Dry Rainfed Farming System

Located in a `rain shadow' surrounded by the Rainfed Mixed Farming System in the Western Deccan, this farming system has a higher proportion of irrigation than the moister surrounding areas, allowing it to support a similar range of irrigated and rainfed crops despite the drier climate. Total system area is 18 million ha with an estimated 10 million ha - about 53 percent - under cultivation. An estimated 3.5 million ha, or 36 percent of the cultivated area, is irrigated - and this irrigation is a central determinant of the farming system. Of the total system population of 45 million, nearly 30 million are classified as agricultural. Because of the prevalence of irrigation, vulnerability is somewhat lower than in the neighbouring Rainfed Farming System, and thus the level of poverty is moderate.

Pastoral Farming System

Across the semiarid and arid zones, from Rajasthan in India through Pakistan and Afghanistan, transhumant pastoralists keep mixed herds of livestock. The system includes scattered pockets of irrigation which mitigate the extreme seasonal vulnerability of pastoralists. Total system area covers 55 million ha, and it supports an estimated 12 million cattle and 30 million small ruminants, as well as a significant number of camels. There are scattered areas of cultivation amounting to an estimated 6.8 million ha, of which nearly two thirds (4.6 million ha) is irrigated and sown to rice, wheat and other food and fodder crops. Off-farm income is an important source of livelihood. Of the total system population of 27 million, around 21 million are classified as pastoral or agricultural. In aggregate terms, this farming system is not of great importance, supporting only three percent of the human population and less than 10 percent of the livestock of the region. The level of poverty is moderate to extensive, and is periodically accentuated by droughts.

Sparse (Arid) Farming System

The land area of the system is estimated at 57 million ha, supporting an estimated 16 million bovines and 29 million small ruminants. About 1.7 million ha is cultivated, practically all under irrigation. There are some scattered irrigation settlements in the arid areas; in most cases used by pastoralists to supplement their livelihoods. The rest of these areas are utilised for opportunistic grazing where water is available for livestock. The human population is 23 million, of which 9.6 million are classified as pastoral or agricultural. There is a gradual transition from the Pastoral System to this system, which has moderate to extensive poverty that is often severe after droughts.

Sparse (Mountain) Farming System

This system lies at altitudes above 3000 metres along the mid level and upper slopes of the Himalayan Range and occupies an estimated area of 34 million ha with a population of 3 million people, of whom 2.8 million are classified as agricultural. A number of small settlements depend on potatoes and buckwheat, plus cattle and yak herds. Cultivated area is 1.9 million ha, or only five percent of total area, and only around 10 percent is irrigated. There are an estimated 10 million cattle and yaks, and nine million sheep and goats. During the summer, herders graze cattle and yak on the higher slopes. Generally, household incomes are supplemented by seasonal migration and in some cases by trade, mountaineering and tourism. Poverty tends to be moderate overall, although more prevalent in remote areas.

Tree Crop Farming System

This scattered farming system comprises plantation companies and smallholders producing substantial areas of tea, rubber, coconuts and other tree crops. It is estimated that the system covers three million ha of land, with some 1.2 million ha of annual and permanent cropland. Concentrations of this system are found in the lowlands of Sri Lanka (especially coconuts), Kerala in India (including spices), and the upland areas of India, Nepal, Bangladesh and Sri Lanka (tea estates). The estimated agricultural population is seven million. Poverty is moderate and largely confined to agricultural workers. Given the dispersed nature of this system, it is not distinguished on the map.

Urban Based Farming System

In most large towns and cities in the region the intensive production of perishable high-value commodities - such as milk and fresh vegetables - has expanded. These are generally commercial systems with high levels of external inputs and with effective links to the surrounding rural areas for stock feed and fodder. The system has an agricultural population of 11 million, and contains around 12 million head of bovines (cattle and buffaloes).

Region-wide trends in South Asia

Population, hunger and poverty

The region's 1999 population of about 1344 million13 is expected to increase by approximately 1.4 percent per annum to 1650 million in 2015. Thereafter, population growth is expected to slow to about 1 percent per annum to reach 1920 million by 2030. The proportion of the total population living in cities (presently 28 percent14) has increased markedly over the last four decades and is expected to continue to expand, reaching 53 percent in 2030.

A key factor affecting the pace and direction of change will be the increase in human mobility - reflected in rapid urbanisation and more off-farm employment, especially for adult males. In some cases this mobility will lead to the adoption of more extensive, low labour systems of production - such as paddy cultivation rather than cotton cultivation - so as to provide the best returns per labour day for periods that can be fitted-in with off-farm employment. It is also likely to be reflected in an improved asset position on farms; more mechanisation, and also an increase in the area of land per operator. As greater numbers of adult males migrate seasonally and semi-permanently in search of alternative livelihoods, it is anticipated that women farmers will assume greater responsibilities for the management and operation of farms.

The absolute number of undernourished people in the region, some 284 million in 1995-199715, has remained stubbornly high, as shown in Figure 1.2 in Chapter 1, but is projected to decline dramatically - to 82 million by 2030. This represents a decline from 23 percent of the population to only four percent. Average food intake was estimated at 2424 Kcal per person per day in 1995-1997 and is expected to increase to 2790 Kcal by 2015 and to reach 3040 Kcal by 2030. The quality of the diet is also expected to improve, with consumption per head of both meat and dairy products forecast to double over the period from 1995-1997 to 2030.

Poverty levels are high throughout the region, although lower in Sri Lanka than elsewhere. In India, some 44 percent of the population live on less than US$1 per day, and 86 percent on less than US$216. Figures are slightly lower, but comparable, for Nepal, Pakistan and Bangladesh. Although figures for Afghanistan are not available, it is estimated that approximately 530 million people within South Asia live on less than US$1 per day. According to World Bank projections, those in `dollar' poverty should decline from the current level of 40 percent to between 18-25 percent by 2030, depending on the growth scenario selected.

Natural resources and climate

The area of land under annual cultivation and permanent crops in the region is estimated at 213 million ha (49 percent of total land area) and is expected to show only a marginal increase by 2030. Additional land area would come principally from grazing and forest land, and would therefore involve significant investments, as well as some foregone production. Estimated increases in development of cultivated land would be greater but for losses to buildings and civil works in existing urban centres17.

Soil erosion in the highland and rainfed watersheds is widespread and is being aggravated by heavy grazing and the loss of nutrients from cropland. Overgrazing has also led to soil degradation in large areas of the Mixed and Pastoral Farming Systems. Severe soil management problems also confront the intensive Rice and Rice-Wheat Farming Systems where yields are increasing more slowly, or even stagnating in some high intensity areas. This is attributed by some analysts to deteriorating soil physical condition, declining organic matter and unbalanced fertilisation - with over-reliance on nitrogen fertilisers, relative neglect of other major and micro-nutrients, and declining use of animal manure18.

The overall scarcity of water resources in the region, and their geographic distribution, has shaped the development of South Asia's farming systems. The region's irrigated land area is forecast to grow slowly, from 85 million ha to 95 million (44 percent of cultivated land) in 2030. The high cost of developing new, environmentally sustainable, irrigation systems - and the reluctance of donors to finance large irrigation projects - may lead governments to place greater emphasis on modernising existing irrigation schemes and the improvement of water use efficiency19. This will involve participatory management or transfer of ownership to users, improved design of operating systems, better drainage and cost recovery. Considerable potential remains for improving water management through better rainfall conservation and use; further development of water storage; and conjunctive use of groundwater. Increased on-farm water conservation can be achieved through conservation agriculture; including mulching, bunding and windbreaks.

Shortages of groundwater in some localities have already started limiting the use of tubewells within the intensive Rice and Rice-Wheat Farming Systems20. Such shortages are bound to worsen over the next three decades. It has been estimated that, if present rates of increase of overdraft continue, as much as one-third of irrigation blocks may be over exploited by 202021. The Rainfed Mixed and Pastoral Farming Systems will also be constrained by the limited availability of both ground and surface water for crop and livestock production. There will be strongly increasing demand for water from urban and industrial users during the coming decades. As a result, local water markets are likely to develop, with agriculture at a disadvantage except in the case of high value crops. As labour costs rise, it is likely that irrigation water will increasingly be used to replace labour in land preparation and weed control in rice. Increasing pressure on water resources also seems to have given rise to problems of declining quality of drinking water - including widespread arsenic poisoning in Bangladesh.

Due to high population densities, forest resources are under threat in all countries of the region. Although FAO data indicate a slight increase of forest and woodland cover throughout the region, to 85 million ha in 1994, there is evidence of deforestation occurring during the 1990s at the rate of 0.13 million ha (0.2 percent per annum). The pressures causing deforestation are expected to increase still further, particularly in the Highland Mixed Farming System - little forest remains in most other farming systems of the region. In many instances, deforestation results from increasing pressure to obtain agricultural land, but conflicts of interest between individuals and forest departments have also contributed. The growing realisation that forest communities have to be fully involved in forest management, and that forests need to be managed for the benefit of such communities, has done much to overcome the latter factor.

Climate change is projected to have a significant negative impact on agricultural production within the region. The frequency of storms which batter the coastal areas of Bangladesh and Eastern India is expected to increase. Not only will the variability of climate cause fluctuations in yield, but average yields of rainfed crops could drop by a substantial amount22.

Science and technology

Agricultural research in South Asia has been strengthened dramatically over the past 40 years through reorganisation of the National Agricultural Research Systems (NARS); with the establishment of central co-ordinating bodies23, decentralisation into regional research centres, greatly enhanced manpower and increased investment. The NARS in the region have benefited greatly from the strong linkages and networks they have established with International Agricultural Research Centers (IARCs), such as IRRI, CIMMYT and ICRISAT, building on their participation in the development of green revolution technology. More recently, the many NARS have shifted their emphasis to meeting the post-green-revolution challenges of stagnant yields and resource management, by developing technologies for resource-poor farmers and farmers in sub-optimal crop environments.

Increasing demand, changing consumer preferences and degrading resources within the region will continue to pose new challenges to these agricultural research systems, and increasingly complex and diversified technologies will be required to meet these challenges. The high returns to agricultural research in marginal lands24, the growing role of women in farm household decision making and the emergence of new areas of integrative research (agroforestry, biotechnology, etc.), will require the development of new skills and inter-disciplinary research efforts. Although private-sector research will make increasing contributions to the development of new technologies for improving the profitability of commercial ventures, the focus of public-funded agricultural research is likely to remain the resource-poor small farmer.

Returns from past agricultural research in the region (e.g. the HYV-fertiliser-irrigation technology) seem to have been concentrated in certain systems25, such as the Rice and Rice-Wheat Farming Systems. Emerging problems require a better mixture of upstream (basic or strategic) and downstream (adaptive and participatory, farm-level) research. A major issue is that much of the present agricultural research effort fails to address many topics of most interest to farmers, e.g. improved low-input crop husbandry practices and minimised production risks.

Use of inorganic fertilisers in the South Asia region has expanded rapidly in recent decades; from three kg of plant nutrients per ha in 1970 to 79 kg/ha in the mid-1990s. The rate of consumption is expected to continue to increase, albeit more slowly. Application rates are high in irrigated, intensive farming systems compared to dry and risky production environments. In some of the more remote areas, such as the Sparse (Mountain) Farming System, high transport costs can easily exceed the material cost.

In some cases there is considerable wastage of nutrients - particularly nitrogen in irrigated and wetland cultivation - due to low efficiency in the use of fertiliser. Severe losses by volatilisation occur in Pakistan, which is a large consumer of nitrogen fertiliser. The future rate of increase in total fertiliser use is expected to decline, although the rate of decline will depend to some extent on progress in raising the efficiency of fertiliser use. Nitrogen losses from volatilisation, leaching and runoff could, therefore, remain constant or even diminish.

At present, use of fertilisers for paddy is constrained by low rice prices. The use of fertilisers in the coming decades will be influenced by fertiliser prices (likely to increase in real terms), commodity prices (likely to decrease), use of organic fertilisers (likely to increase) and fertiliser use efficiency (also likely to increase). The possibility of significant contributions from the use of Biological Nitrogen Fixation (BNF) cannot be discounted26. In parallel with the increase in fertiliser use, the demand for agricultural chemicals has grown, although moderated somewhat by the spread of IPM.

Paddy rice has traditionally been the most important single crop in South Asia and it has increased in importance over the last 30 years - principally as a result of the green revolution - with yields increasing by an average of almost two percent per annum over the period 1970 to 2000. Rice area has also increased but more slowly (see Table 5.2), resulting in production of 2.5 percent over the last 30 years, to 184 million tons. Production is forecast to further increase in the period to 2030. Wheat production has shown by far the strongest growth among the cereals in recent decades with increases in both yields (nearly three percent per annum) and land area (1.4 percent per annum) to achieve an overall increase in production of more than 250 percent to almost 100 million tons in 2000. This growth is expected to continue in the period to 2030. By contrast, coarse grains such as millet and sorghum, have seen declines in harvested areas in the last three decades, and only small increases in yields have avoided absolute declines in outputs of these crops.

Fruits and vegetables have equalled wheat production trends in recent years, showing that diversification of traditional cultivation patterns is already well underway. Harvested areas have increased faster than for any other crops. The modest increases projected in irrigated area, and improvements to established irrigation schemes, are likely to encourage even greater diversity of cropping in the future. When combined with the growing demand from urban and international consumers, as well as improved marketing and processing channels, additional strong growth in fruits, vegetables and other non-traditional products is anticipated.

Cultural factors strongly affect livestock production and consumption throughout the region. Cattle numbers have shown only moderate growth from some 230 million head in 1970 to 277 million head in 2000 (see Table 5.3); an annual growth rate of about 0.6 percent over the entire period. Buffalo numbers have increased faster, by 1.9 percent per annum over the same period, although the rate of growth has declined during the past decade. Similar rates of growth have been seen for small ruminants, which totalled 321 million in 2000. The largest increases in livestock numbers and production have occurred in poultry. Intensive, large-scale production to meet urban demand has already developed in most countries of the region. In the period 1970-2000, the rate of increase in number of poultry was 3.8 percent per annum.

For meat, milk and eggs, output has grown more rapidly than livestock populations, suggesting that production efficiency has improved in recent decades. Buffalo and cattle numbers have increased by an average of less than one percent per annum, and small ruminants by just over two percent, while dairy output has risen by over four percent. With increased incomes, meat consumption (particularly poultry meat and eggs, sheep and goat meat), and demand for dairy products, are expected to continue their significant expansion. However, the large ruminant population is expected to stabilise or even decline as tractors replace both draught buffalo and oxen. The management of the remaining animals will increasingly intensify and focus on productivity - probably more than compensating for this decline.

Trade liberalisation and market development

The recent trend towards increased market liberalisation is expected to continue over the next decades. Commercial farmers may be exposed to declining terms of trade, especially for cereals in international markets. Significant diversification is expected in almost all farming systems; partially in response to intensified external competition for previously protected domestic markets in the case of basic staples such as rice, and partially as a result of increasing export opportunities. South Asia could build upon its dominant world position in a few niche markets - such as that for mango and cashew products - and develop a wide range of competitive fruits, spices, colorants and other tropical products. Given the large number of expatriate South Asians living in the industrialised countries, and the popularity of regional cuisine, it is also to be expected that considerable growth will occur in processed foods. This tendency is likely to accelerate as advances in packaging and transport technology allow fresh breads, curries and other perishable products to be delivered to western markets at economical prices.

Although rural populations in South Asia may provide only limited markets for imported foodstuffs in the medium term, the large and expanding urban markets will become increasingly important consumers of both regional and extra-regional produce in the coming decades. Urban-based agriculture will expand and intensify in response to this demand. Dairying is expected to grow still further, as will processed dairy products; while poultry, lamb and goat consumption will expand rapidly. Vegetable and vegetable oil production will also continue to grow strongly. As international labour markets develop, remittances from overseas work will increase and a significant proportion will flow into agriculture and rural investment in general. These cash incomes are an important source of livelihoods in many marginal rural areas - for example in the Highland Mixed Farming System - and often finance farm improvements.

Policies, institutions and public goods

A large proportion of the increase in foodgrain production during the Indian Green Revolution occurred in the 10 percent of districts with adequate local infrastructure - especially for water management, transport, and electricity for tubewells. On the other hand, agricultural development in many areas of the region has been constrained by a lack of infrastructure. In particular, the shortage of roads in remote and sparsely populated areas pushes up transport costs for both inputs and marketed produce, while the lack of health and educational services reduces labour productivity27.

Most countries in the region have policies that favour urban areas and the manufacturing sector rather than the rural and agricultural sectors. Intervening in the market to maintain food prices artificially within reach of the increasing, politically articulate urban population may also aggravate already low internal cereal prices, thus exacerbating the bias against rural communities. Because the majority of the poor are located in rural areas, poverty reduction efforts should be targeted towards increased agricultural income.

Decentralisation and local institutional performance will be key issues in the development of most farming systems. Line departments have already given up some of their functions to the private sector. They have also begun to hand over others to empowered rural communities and this trend will become more pronounced in the future. However, this transfer may not take place very rapidly and the quality of its implementation will require monitoring, given the close relationship between rural community constituencies and politicians. Another far-reaching change that is occurring in association with decentralisation, is the growing role of women in panchayat (local council) and district-level decision making. There is also an increase in public-private partnerships for agricultural development, which will have far reaching effects on deciding agricultural research priorities.

Local institutions in some parts of the region (e.g. India) are being mandated to play a larger role in land administration, but often lack the capacity to discharge such functions properly. In general, land reforms in the region have met with limited success in improving access to agricultural resources. Land ceilings have had some limited impact, but larger farmers have often found ways to maintain their holdings. The implementation of these programmes is rendered difficult by the inadequate state of land records in some areas.

Information and human capital

By 2030, the majority of farm households will be literate and numerate, and nearly all younger people - including women - are expected to have a basic school education. However, continuing investment in rural education is required not only to support this transformation, but also to address two other key human resource needs: (i) equipping workers with the skills necessary to transfer from agriculture to the off-farm economy; and (ii) ensuring that those remaining in agriculture possess the skills needed to manage the emerging knowledge-intensive farming systems. Although total labour requirements in agriculture are not likely to increase, improved skills are required to raise efficiency and productivity. Such local increases of human capital would underpin the development of small-scale local rural industry, perhaps along the lines of some of the more successful examples of rural industry found in China.

The shift to commercial, knowledge-intensive, farming systems requires a greatly improved flow of information to farmers, as well as support services concerned with new technologies and market information. This is emerging as a key issue and is a promising area for public-private partnerships. In some areas cellular telephone communications are revolutionising the availability of market information, for example, in Bangladesh. A number of computer-based systems for farm extension have shown promise, such as the touch-screen systems tested by the Swaminathan Research Foundation. The improvement in the information flow across the government-NGO interface is well documented28.

It is widely recognised that an information revolution will occur in rural South Asia, although its full effects cannot be predicted. However, without doubt, the wide availability of market information will substantially improve the efficiency of the system and increase access by smallholders. It may also be that widespread dissemination of technical information through multiple channels will lead to a redefinition of the role of agricultural extension.

Selection of farming systems for analysis

Four of the farming systems outlined in the previous section have been selected for analysis, using criteria based upon agricultural population, the incidence of poverty, and apparent potential for agricultural growth and poverty reduction in the coming 30 years:

• Rice Farming System;
• Rice-Wheat Farming System;
• Rainfed Mixed Farming System; and
• Highland Mixed Farming System.

Most of the poor inhabitants in the region live within these four farming systems, which also contribute more than three-quarters of the regional GADP. They are described in more detail in the following sections.

RICE FARMING SYSTEM

Characteristics of the system

The Rice Farming System covers about seven percent of the land in the region (see Box 5.1) and encompasses much of the old and well-developed wetland rice areas, particularly Southern Bangladesh and South Punjab, but also including coastal zones of Tamil Nadu and Kerala and the wet zone and irrigated parts of the dry zone of Sri Lanka. Land is generally farmed by owner operators or sharecroppers. Farm size tends to be bi-modal, with a large number of tenants and sharecroppers with small areas of the order of 0.3 to 1 ha (see Box 5.2), and a few medium to large owner-operated farms that could be from 4 to 10 ha or more. Rice is invariably grown in the wet season. In the dry season a second rice crop, or another less water-demanding crop (e.g. coarse grains, oil seeds, legumes, vegetables), will be grown.

In most cases, the system is supported by supplementary irrigation in the monsoon season and full irrigation in the dry season. Because of the limited resource base of the farming system and the fact that these areas are generally close to urban areas, off-farm employment is common. The strategic importance of rice as a food staple and the generally easy access have resulted in well developed extension services. With virtually continuous paddy cropping there are limited fodder resources except for paddy straw to support ruminants, which generally include milk and draught animals - buffalo, or oxen in the drier areas, where these have not already been replaced by tractors. Nevertheless, this farming system contains a significant proportion of the region's cattle and dairy animals, the latter principally because of proximity to large urban centres.

Most farmers know the value of using fertiliser and improved seed. Nevertheless, uptake of improved varieties has not always been enthusiastic. This is largely due to lack of high quality certified seeds, the poorer taste of some new varieties, their lower tolerance to early or late transplanting - and the sometimes marginal increase in yield compared to local varieties. Typically, farm households must supplement their food production with off-farm income in order to ensure food security.

Considerable diversity exists within the system, with a transition to the Rice Wheat System in the northern areas of Bangladesh and in Bihar State, while in more southern parts of this zone, aquaculture is widely practised and households tend to keep more cattle - including dairy cattle - and more poultry. The south-eastern areas of the system in Tamil Nadu and the southwestern parts in Kerala have less risk of cyclones and floods and tree crops are more common.

Despite increasing diversification in recent years, households still tend to depend on a limited crop income that is vulnerable to fluctuating yields, or even complete failure in flood and cyclone prone areas. The major source of vulnerability, however, is failure to find work in the vicinity, or inability to work because of sickness - a frequent occurrence. There are few traditional mechanisms to which the poor can turn in time of need; access to government safety nets, such as employment guarantee schemes, is often inadequate. In this respect, poor small farmers are just as vulnerable as landless rural workers.

Trends and issues in Rice System

During the coming 30 years, land availability per capita is expected to decline because of population increase and the loss of land to urban expansion. This loss may be aggravated by land degradation, including soil erosion and salinisation. Moreover, increasing water scarcity in agriculture is expected, as demand for water grows in domestic and industrial sectors. However, some market-driven land consolidation can also be expected, together with gradual reduction in sharecropping. Education levels for both men and women are expected to increase substantially. Widespread adoption of hybrid rice - both existing and future new varieties - is expected, except in deepwater and other unfavourable rice ecologies.

The declining terms of trade for rice, and increasing costs of production, will induce diversification of production to include inter alia, dairying and aquaculture. Intensive pressure on land will limit the availability of fodder and lead to a decline in the population of buffaloes. Draught buffalo power will be replaced by two-wheel tractors, which will also be widely used for local transportation. Increased labour costs will favour mechanisation of other operations, such as sowing, weeding and threshing. Given the declining profitability of rice, only a modest increase in external input use is expected - with no significant increase in the riskier flood-prone areas. The gradual decline in rice profitability, and increasing population pressure, will force many male household members to seek an increasing part of their income from off-farm sources. In these cases, women will shoulder an increasing burden of farm work. Overall, there may be a gradual improvement in household food security and a modest reduction in poverty.

Some improvement is expected in public infrastructure and other aspects of the socio-institutional environment affecting agricultural production. In particular, transport, educational and health facilities will improve. These will be associated with significant decentralisation of decision-making to the district and the local panchayat levels. At the local level, farmers' organisations are also likely to grow stronger. The prevailing feminisation of poverty in large areas of South Asia will most probably begin to be reduced. With the enforcement of the one-third quota for women in local office, local governance will increasingly reflect the needs of women, who are currently disproportionately represented among the very poor.

As public sector research and extension budgets decline during this period, greater emphasis will be placed on the private sector, farmers' organisations, and upon individual farmer experimentation and technology sharing. The enhanced role of farmers and the private sector in advisory services will tend to improve the relevance of extension messages. The success of the Grameen Bank in Bangladesh has generated a plethora of micro-finance activities throughout the region. There is likely to be an increase in seasonal and permanent out-migration, and a slow growth of the off-farm rural economy.

The main issues driving the evolution of the system are the currently low and declining paddy prices combined with increasing labour costs. These make it increasingly unattractive to use high rates of fertiliser application and consequently inhibit the growth of rice productivity. Low domestic paddy prices are largely a reflection of declining world prices. In some instances, however, this trend is exacerbated by government attempts to keep rice prices low in order to satisfy urban populations - using price controls and monopoly purchases with countervailing subsidies on fertilisers, irrigation and other inputs30.

Priorities for Rice System

In order to make substantial progress in reducing poverty in this system, the following household strategies to escape poverty need to be supported in the following order of importance: diversification; increased off-farm income; intensification of existing production patterns; and exit from agriculture. While intensification and diversification in this farming system require some attention to natural resources - soil management and flooding in particular - the major problem is the low and progressively declining producer price for rice resulting from unfavourable terms of trade. This can only be countered in the long run by using more efficient production practices, thereby increasing the comparative and competitive advantage of local rice production. An effective programme of research and extension to improve labour and factor productivity is needed to improve farmers' production practices. In order to maintain adequate incentives to farmers it will also be necessary to avoid urban bias in trade and economic policies.

The productivity of the Rice Farming System can be increased through intensification, although with a different emphasis to that of the earlier variety-chemical based intensification. Despite major investments in irrigation infrastructure in the past, many schemes still perform poorly. Upgrading these systems to improve irrigation security, plus the introduction of tubewells for conjunctive use of groundwater, generally results in significant increases in production and farm incomes. Water user associations are an effective means of improving water use efficiency31. Box 5.3 shows that small adjustments in policies and technology can unleash major investments in the rural sector which, in turn, increase productivity and reduce poverty.

Other areas in which interventions can increase the productivity of this farming system are research and extension; farmer training (particularly in farm management, IPM32 and other crop production practices); mechanisation of operations such as weeding and threshing; land consolidation and leasing mechanisms to increase operated field sizes and, over time, operated farm sizes; and, improved post-harvest handling and milling. Box 5.4 shows the potential for joint intensification and diversification of irrigated rice systems in Bangladesh.

Where systems are already quite intensive, further opportunities for improving productivity through diversification appear to involve the incorporation of income-generating enterprises such as dairying, aquaculture, horticulture and local value-added processing33. Where the risk of flooding is low, aquaculture can provide additional income in inter-harvest periods. Duck or chicken raising provides other possibilities; although in order to be adopted all of these potential enterprises need to show better returns to labour than the operator could get from off-farm income-generating activities. Their widespread adoption would also need some strengthening of the relevant support services.

Whether farmers chose crop intensification strategies or crop diversification strategies in a specific area depends not only on the individual circumstances of the farm household, but also on the availability of appropriate technologies; this in turn is linked closely to the effectiveness of agricultural research related to the zone.

The system can also be improved by diversification into vegetable or other cash crops where there is nearby urban demand - as demonstrated by the expansion of horticultural production around Bogra in Bangladesh35. A combination of these practices can, in favourable circumstances, lead to highly productive and profitable integrated, intensive farming systems. Similarly, dairying has been a traditional engine of growth for smallholder farming systems in South Asia. One well-known dairy scheme is the Milk Vita programme in Bangladesh (see Box 5.5).

RICE-WHEAT FARMING SYSTEM

Characteristics of the system

The Rice-Wheat Farming System covers 19 percent of the land in the region (see Box 5.6), and occupies a broad swathe of land from Northern Pakistan, through the Indo-Gangetic plain including the Terai of Nepal, and the Gangetic plain in Uttar Pradesh, Bihar and West Bengal, to finish in Northwest Bangladesh. The farming system is characterised by a `summer' (monsoon-season) wetland rice crop and a `winter' (cool, dry season) wheat crop - sometimes followed by a short `spring' vegetable crop) - see Box 5.7.

Poverty and household food insecurity are widespread, principally among landless agricultural workers and sharecroppers. In some areas cotton is part of the system, but its cultivation is constrained by its high labour requirements and the long duration of the crop. Commonly, some form of irrigation supports the system. It contains large populations of cattle, sheep and goats. Livestock are more common than in the Rice Farming System, although often they may not be strictly a part of the farm system but consist of large cattle herds owned by landlords or businessmen. These cattle graze the stubble after harvest and return to wastelands or uplands during the crop season. Substantial heterogeneity exists within the farming system chiefly in terms of access to land and water resources.

With the development of improved varieties of both rice and wheat, and the use of irrigation and fertiliser, the Rice-Wheat Farming System has shown remarkable increases in production. However, in recent years the declining or stagnant yields and factor productivity of the system have given cause for concern and are the subject of an on-going research programme by the Rice-Wheat Consortium for the Indo-Gangetic Plains.

The western parts of the system in Pakistan tend to have fewer animals and are more mechanised. In these irrigated areas, away from major flood-prone districts, vulnerability is associated with price variation, crop pests and inability to earn sufficient off-farm income. It is not an easy system to manage. In particular, wheat sowing must follow immediately after the rice harvest if subsequent wheat yields are to be satisfactory. Over time, the system has expanded into areas where groundwater is not so easily accessible and because irrigation supply is unreliable, farmers have to transplant at the onset of the monsoon. In order to maintain flexibility they continue using traditional varieties, but these are slow maturing. Consequently, all too often, wheat is sown late, thereby depressing yield as high temperatures affect the plants while they are setting seed.

Trends and issues in Rice-Wheat System

Average farm size will continue to grow gradually due to permanent and seasonal out-migration of youth, although the bimodal farm size pattern is expected to be accentuated in the coming decade. Later, it is expected that the number of sharecroppers and small farms will diminish. Land close to urban centres will continue to increase rapidly in value, while some observers expect an increase in absentee ownership. The practice of sharecropping can be shown to have some advantages, but has the big disadvantage of perpetuating a traditional and largely immutable operating system within which innovation is difficult or impossible to promote.

The decline in soil productivity, due inter alia to excessive reliance on unbalanced application of mineral fertiliser, is likely to continue depressing wheat and rice yields for some time to come. No definite answers have yet emerged from research in this area37. Similarly, the build-up of soil salinity and sodicity in the dry, western areas of this farming system - caused by poor water management - will also continue until appropriate steps are taken to improve water control at farm level. Reversing these trends will depend on the success of future research and extension, as well as on policy decisions favouring balanced fertiliser use and efficient utilisation of water.

There has been a rapid expansion of tubewells in some areas such as Western and Central Uttar Pradesh, resulting in declining water tables. Efforts to regulate groundwater abstraction have had limited effect and with millions of private tubewell owners, the task of conventional regulation through licensing and monitoring has proved insurmountable. It is clear from recent evidence gathered in Gujarat and Rajasthan38 that as the exploitable aquifers are depleted and degraded, it is likely that some groundwater-dependant irrigation farmers will have to convert, wholly or partially, to rainfed cropping. There will be increasing levels of mechanisation, especially for primary tillage, which is already noticeable in the Pakistan Punjab where most draught oxen have disappeared. The typical enterprise pattern is expected to shift towards more dairy, horticulture and feed grains. Because of the availability of feed grains and crop by-products, plus the proximity of major urban markets, this farming system is a likely location for the expansion of specialised large-scale, industrial poultry production.

Overall, most farmers are expected to achieve both household food security and increased farm household incomes by the year 2030. There may also be a gradual decline in poverty among the landless as wage rates rise - even if opportunities for agricultural employment diminish. Some improvement in infrastructure is probable; particularly better roads. Decentralisation will bring decision making closer to the farm, and in the process women will have a greater role in local governance. It is expected that farmers' organisations will be strengthened. Funding of public-sector research and extension will decline, and the private sector and farmers' organisations may play a bigger role in experimentation and advisory services; which may improve the efficiency of the dissemination of technical information. Modest growth of the off-farm rural economy is expected.

Priorities for Rice-Wheat System

In order to make substantial progress in reducing poverty in this system, the following household strategies to escape poverty need to be supported (in order of importance): diversification; increased off-farm income; intensification of existing production patterns; increased farm size; and exit from agriculture. There are a number of worthwhile options for supporting these poverty pathways that will also have direct impacts on agricultural growth, although the opportunity cost of labour will be a critical determinant of the viability of interventions.

One high priority is to tackle the resource conservation issues; such as declining soil fertility, development of salinity and sodicity problems on irrigated land in western areas, and groundwater depletion in zones irrigated by tubewells. Research is being conducted to develop technologies that can improve the level of soil fertility, which has fallen as a result of the continuous, intensive cereal production practised since the Green Revolution started. However, governments sometimes pursue policies that work at cross-purposes with such research efforts. For example, the continuance of heavy subsidies on urea fertiliser in India, while at the same time deregulating the prices of P and K fertiliser, is causing an imbalance in fertiliser use among farmers - particularly resource-poor farmers practising the rice-wheat rotation. As a result, there is continuous mining of P and K nutrients in the soil causing long-term damage to soil productivity.

As mentioned above, an important factor contributing to the development of salinity and sodicity in irrigated areas of this farming system is the inefficient use of water, particularly at farm level. The highly undervalued price of water from canal systems and the heavy subsidies - up to 100 percent on shallow tubewells, 25 to 50 percent on pumps, plus varying rates on electricity for pumping - are all incentives for excessive use of water by farmers, with consequent waterlogging. Depletion of water tables by indiscriminate sinking of tubewells has also been encouraged by these same subsidies, as well as by the absence of appropriate monitoring and regulatory mechanisms. While waiting for governments to deal directly with these policy issues, another option would be to improve moisture conservation by employing a range of techniques such as zero tillage, use of plastic and other mulches, and planting of windbreaks. Improvement of irrigation security would also materially improve the system, by allowing the adoption of short-duration rice varieties, more timely wheat sowing and production of spring vegetables.

Many of these resource management technologies are scale-neutral and thus poor smallholders - although perhaps not sharecroppers - will gain from their adoption. The need for close co-ordination between water management, fertiliser use, pest control and other husbandry practices, has resulted in the development of Integrated Crop Management (ICM) systems39. The dissemination of ICM practices to farmers could be a good option for increasing rice yields and reducing the cost of rice production in South Asia. In some countries the active participation of women in these practices will also be a challenge. Another approach that would be appropriate, especially in rainfed areas susceptible to erosion or where labour costs are high, is Conservation Agriculture (CA), in which minimum tillage is practised to obtain better moisture and soil conservation. The objective should be to let farmers choose those approaches and practices that best suit their conditions and circumstances.

Good opportunities also exist for better integration of ruminant livestock into smallholder farming systems; with profitable conversion of crop by-products, increased use of manure, and possibilities of regular cash income from dairying. The treatment of straw, using urea or other chemicals to increase its value to livestock, is likely to be adopted more widely. Such labour-intensive expansion of livestock production on smallholder farming systems will increase output and reduce poverty. Large-scale industrial production of animals, especially dairy and poultry, will also expand. The expansion of livestock enterprises will stimulate feed production and the development of a feed industry. It will be desirable to site industrial livestock production units so as to minimise feed transport costs and to contain environmental problems involved in processing. Since industrial livestock production will boost output but not reduce poverty to any significant degree, the challenge will be to promote linkages - between industrial livestock production and smallholder livestock production - which benefit poor farmers. Investments in industrial livestock production in this zone are likely to be short term, since in the medium and longer term the locus of feed production and industrial livestock production may shift to the Rainfed Mixed Farming System, where land values and population densities are lower.

The system can easily incorporate many other forms of diversification, including the introduction of fruit trees or other cash crops in situations where land is owner occupied - but this would normally not be possible under sharecropping arrangements. Diversification will generally require investments in marketing, transport infrastructure, research, extension and other support services. Although governments may continue to provide certain categories of seed and planting material, it is expected that farmer co-operatives and the private sector would progressively take over this activity. Similarly, the demand for technical and market information could be addressed through public-private partnerships.

HIGHLAND MIXED FARMING SYSTEM

Characteristics of the system

The Highland Mixed Farming System incorporates the cultivation of a range of cereals, legumes, tubers, fodder, fodder trees and livestock (see Box 5.8). It stretches in a narrow band across upland areas of Pakistan, India, Nepal, and Bangladesh, occupying the area between the plains below and the generally uninhabited mountain areas above. There is also a small area in Central Sri Lanka. In total, the system covers about 13 percent of the land area of the region. In the more remote areas, where mineral fertilisers are very costly or unavailable and distance to markets is prohibitive, the flow of nutrients to cultivated land is generally only from grazing or cutting of fodder trees from woodland areas at higher altitude. However, in more accessible areas, such as the central highlands of Sri Lanka, the opportunity for successful vegetable production, particularly potatoes, can result in a highly intensive system of commercial production. Similarly, in the hills of Himachal Pradesh there has been a large expansion in horticulture, particularly apple orchards. Generally, ruminant livestock are an important part of the system as they provide draught power, milk, manure and cash income. A typical farm household is described in Box 5.9.

In most cases the existing communities have lived in these areas for thousands of years, often as tribal societies that originally practised shifting cultivation. In fact, two thirds of the global indigenous population live in Asia, many of whom reside within the Highland Mixed Farming System. Most, however, have now adopted more settled agriculture. In some areas, these original communities have been joined by people from lower down the slope, pushed up by population pressure and this has led to ethnic tensions in certain cases. In some instances there are also cases of land speculation where forest areas are cleared by `farmers' with the support of a local landowner or businessman. In most instances farmers in these areas do not have land titles and some communities may well be living inside the boundary of the forest as recognised by the forest department. The most important issue overall is the ever-increasing population pressure, caused both by reduced mortality rates and by immigration by `lowlanders', which can make the already fragile system unsustainable.

Trends and issues in Highland Mixed System

For many of these upland areas, particularly where there has been little development assistance, the trend has been one of a general deterioration in incomes, living standards and the overall environment. A reduction of forest cover on the upper slopes, unimproved agriculture and continuous cultivation on steep slopes - often with thin or poor soils - have led to impoverishment of communities that now barely eke out an existence. They have virtually no cash income and women have to walk progressively further to secure their water and fuelwood needs. Soil and water conservation under these conditions is generally poor, and erosion and the decline of soil fertility can pose grave threats to household survival.

Generally, there are few local opportunities for young people, who often leave to find work elsewhere leaving behind an ageing resident population. The difficulties experienced by shifting cultivators in adapting to sedentary cultivation has also been a factor. As in some other systems, the situation has not been helped by the general insecurity of land tenure - most families having settled either on common land or within forest boundaries. In some areas, social cohesion is absent or very poorly developed -each family becoming increasingly a law unto itself, with none of the normal hierarchy of village councils or headmen. This lack of social capital makes it difficult to resolve individual land ownership disputes or to obtain agreement on the use and management of common lands. Both of these are pre-requisites for development, being essential for the introduction of improved technology, land use guidelines, and the application of soil and water conservation methods. Several other factors aggravate the lack of social capital, including a low level of education and the paucity of communications facilities.

There are, however, a few promising developments, which are worthy of note. Pockets of cash crop or livestock production have sprung up and have provided the basis for the improvement of the whole farm systems. Examples include apples in Himanchal Pradesh, citrus in Eastern Nepal, virus-free potato planting material in Pakistan and vegetable seed in Nepal. In these examples, production was linked to well-organised collection and marketing, often with private sector involvement. Sound technical and marketing information underpinned all cases. Road access has been a determining factor in permitting the expansion of milk production, but is of less importance for non-perishable products such as vegetable seed.

Priorities for Highland Mixed System

In order to make substantial progress in reducing poverty in this system, the following household strategies to escape poverty need to be supported (in order of importance): exit from agriculture; diversification; increased off-farm income; and intensification of existing production patterns. Emigration can be facilitated by measures to improve rural and urban labour market function and the provision of basic education and vocational skills, focused on the poorer parts of the system.

With regard to diversification and intensification, lessons drawn from the few pockets of successful development provide the directions for strategic interventions (see Box 5.10). In scattered areas with reasonable market access or strengthened social capital, for example along the milk collection routes outside major Hill cities such as Kathmandu, intensification of production is evident. To the northeast of Kathmandu, the production of fresh vegetables has expanded and the resulting cash income has led to increased input use on other crops. In some areas, the extensive planting of farm trees for fodder and timber - substituting for the loss of access to forest resources - is increasing the natural capital of the farms and reinforcing the basis for nutrient recycling through crop-livestock integration40. Other examples have shown how community empowerment improves resource management and incomes; even those of the poorest within a village. This is achieved through communal management of common resources, and group action to increase production through the acquisition of improved germplasm and inputs, plus better marketing41.

The critical component in the strategy for development in the hill areas must be to arrest land degradation through soil and water conservation, as part of an overall programme to improve living standards - a strategy that could be termed conservation by stealth. Since traditional engineering techniques for soil conservation are expensive and often not sustainable, a variety of biological integrative activities is required. This could involve a range of measures including: integrated crop and livestock husbandry; agro-forestry; introduction of perennial cash crops such as coffee or fruit trees (where this is feasible and markets are available); planting of fodder crops; stall feeding of livestock; community managed animal health services and, provision of drinking water. The outcomes of these measures can be further enhanced by under-sowing wetland rice with leguminous fodder species such as Egyptian clover (Trifolium alexandrianum) before the basin dries out. This would improve livestock feed supply and also provide green manure for the following crop.

Since remoteness and high transport costs usually make the use of mineral fertilisers difficult or impractical, development of livestock enterprises such as dairying or goat production - with emphasis on improving nutrition, stall feeding and collection of dung - appears to be a good entry point in attacking poverty in upland areas. Where this can be accompanied by introduction of a cash enterprise, such as orchard fruit or vegetables, the chances of a significant reduction in poverty are much improved. Finding suitable cash crops for the more remote areas may, however, be difficult. Nonetheless, even in remoter areas there may be opportunities for production of high-value, low-bulk commodities such as vegetable seeds, spices or medicinal plants.

All components of the programme need to be implemented in a participatory manner with the full involvement of the local community and joint support of government and NGO partners. A supportive policy environment must exist or be created, as this will facilitate local decision-making and allow local development and management of resources - especially in view of the difficulty of communicating with central or even district authorities. Devolution of decision-making authority will also help to build social capital, which is sometimes lacking in hill-dwelling peoples. Participatory research and extension services will be needed to support households and communities during the adoption and adaptation period. Projects aimed at development of upland watersheds now stress the importance of full community participation in all aspects of design and implementation. Such projects also include helping local communities to develop the ability to take charge of their own affairs and to manage the land and water resources in their area as one of their objectives, while still being able to request technical advice from the respective government agencies.

In some instances, rural development interventions must be broader still, in order to include education and health clinics. Establishment of high-quality, vocational training schools will foster the remittance economy that often constitutes an important source of rural household income and helps to reduce land fragmentation and maintain viable farm sizes.

RAINFED MIXED FARMING SYSTEM

Characteristics of the system

The Rainfed Mixed Farming System covers nearly 30 percent of the land in the region (see Box 5.11), almost entirely within Central and Southern India, although there is a small area in Northern Sri Lanka. The system is not supported by any large irrigation system, but in many instances relatively small areas irrigated from tanks reduce vulnerability to drought and permit dry season cropping. This traditional tank-based supplemental system has been further enlarged in recent decades by the use of tubewells.

Overall, however, being mostly dependent on rainfall, the system faces relatively high levels of risk, and introducing new technology is therefore difficult. Crops grown within the system include wheat, barley, vegetables and fodder crops in the cooler northern areas, while maize, sorghum, finger millet, vegetables, chickpea, pigeon pea, green gram, black gram and groundnuts are more common in the warmer climates of Southern India. Smaller areas of soybean, rapeseed, chilli, onions and sesame are grown mainly as cash crops. Double cropping is possible only where irrigation is available. In Southern India and the Dry Zone of Sri Lanka, where land holdings are smaller, farmers prefer to grow wetland rice on any irrigated land available, as well as in the poorly drained valley bottoms during the wet season. The above-mentioned coarse grains, pulses and oilseeds are grown on the upper parts of the landscape. Some fruit trees such as mango are grown in home gardens. Box 5.12 describes a typical household in this system.

Livestock are an important part of the farming system - which supports the largest share of cattle, sheep and goats in the region - and they usually provide the major part of the farm family's cash income, particularly through sales of adult animals or young stock, since most areas are too remote for commercial milk production. As infrastructure improves, however, dairy production is becoming more important in some areas.

While the total area covered by the system is larger than the preceding Rice-Wheat Farming System, the heavy reliance on rainfed agriculture imposes a much lower population density - about half that of the latter system. A large proportion of the rural population in the Rainfed-Mixed Farming System lives in chronic poverty, aggravated by periodic drought-induced severe transient poverty (see Box 5.13). Agriculture is oriented towards subsistence; while most areas are poorly served by infrastructure and services, and are remote from markets. Agricultural extension services in these areas are typically weak, farmers mostly use traditional technology with a strong bias towards risk avoidance. Land tenure is often an issue and farmers may not have sufficiently clear titles to their land to be able to use it as collateral for obtaining institutional credit.

Trends and issues in Rainfed Mixed System

Enterprise patterns have been changing fairly rapidly, despite only moderate market access. Vegetable and dairy production have expanded where tank irrigation is available, driven by market forces. There has also been a boom in oilseeds, fuelled by significant production subsidies43 and this growth is likely to continue. It is expected that there will be increasing scarcity of fresh water resources as agricultural and urban demands expand. Land degradation, including soil fertility decline, is expected to intensify. Food crop production will increase and the use of hybrid sorghums and millets will become more widespread. Soybean and mung beans may, to a significant degree, replace traditional pulses. The system is expected to become more commercial, with a modest increase in the use of external inputs and further mechanisation. Livestock productivity is expected to increase through the spread of stall-feeding and in response to better market access. Although household food security will improve, there will still be food deficits in drought years.

With regard to the external environment, some improvement of transport infrastructure and social services is expected. Government agencies will decentralise to a significant degree and the role of women in local decision making may be strengthened in some countries. Whilst the scope of public sector research and extension will contract, there will be a greater role for farmer organisations in the provision of agricultural services. Limited expansion of the off-farm rural economy is expected.

Priorities for Rainfed Mixed System

In order to make substantial progress in reducing poverty in this system, the following household strategies to escape poverty need to be supported (in order of importance): diversification; then intensification, increased off-farm income and exit from agriculture (all of equal importance) and finally increased farm size.

Little can be done to significantly reduce poverty within the Rainfed Mixed Farming System without increasing the overall water security of the farm household, which not only underpins intensification and diversification, but also greatly improves the quality of life of the household. In the past, this has generally entailed improving the availability of water for drinking and irrigation by repairing existing tanks, improving diversion works from streams and increasing the scale of water harvesting, as well as acquiring more low-lift pumps and tubewells. These improvements will require social mobilisation and participatory planning if they are to be sustainable.

However, in future the emphasis must shift to the maximisation of moisture and soil conservation for increased production. Conservation agriculture will also reduce the impact of climatic risk. The approach includes measures such as timely cultivation, minimum tillage, rapid seeding with early-maturing varieties, mulching and - where the low opportunity cost of labour in the dry season and the nature of the soil make it an option - bunding and tied ridging. In this area new technology is being adopted slowly at best. More sophisticated methods, such as the use of plastic film for water conservation, are absent altogether. Where adequate and secure drinking water is available, one good entry point is the improvement in livestock production through planting of fodder grasses and stall feeding. This in turn can lead to a build-up of fertility, better rainfed crops and more arable by-products for animal feed. Box 5.14 outlines the potential of water harvesting which is discussed further in the accompanying case study.

Since the farming system depends for its survival on rainfall and possibly groundwater, appropriate measures are required in the upper catchments to make sure that tree cover is adequate to reduce run-off and to encourage infiltration and percolation that can recharge local aquifers. Improvements to the system will require inputs from a spectrum of disciplines, including irrigation engineering, forestry, fodder production, livestock husbandry and horticulture, in addition to traditional arable crop agronomy. The introduction of horticulture and fruit crops has historically been constrained by remoteness from markets. This can be overcome in areas where there is some comparative advantage for fruit and vegetable production, if private sector interest in establishing processing plants can be generated.

Some research stations claim to have developed packages of practices suitable for this rainfed farming system. In general, adoption of these recommendations is disappointingly low, but this can be explained by farmer's aversion to risk, or by shortage of labour or other resources to implement improvements. Successful projects or programmes of assistance need to have a wider focus, rather than merely being concentrated on agriculture-related measures. Provision of reliable drinking water is generally the first priority need in these areas. Other social services including health clinics and primary schools, are also normally lacking.

The development of markets is less important than in the Rice-Wheat Farming System. However, access by poor households to land, water and forest resources will be a critical issue. Similarly, access to sources of information is important for the intensification and diversification of these systems. Conversely, the expansion of information dissemination systems will be important and is expected to accelerate the diversification of part of the system towards high value enterprises - after food needs are met.

STRATEGIC PRIORITIES FOR SOUTH ASIA

South Asia's total agricultural population is second only to East Asia and the Pacific, yet it is experiencing a much higher population growth rate. Compared with several other regions, South Asia has relatively little high potential land with fertile soils plus good irrigation or rainfall. The region has a long history of intensive agriculture, which has led to substantial resource degradation in some locations. During the past 30 years - against all odds - the research and agricultural support services of the system have generated growth in food production in excess of population growth, and reduced the proportion of people living in poverty. The opening of the region's economies has had a profound effect on farming system dynamics in the past decade. Farming systems in the region are likely to evolve rapidly during the coming 30 years, as technologies, institutions and markets change, and the pace of change is expected to accelerate.

Available projections suggest some improvement in indicators related to hunger and poverty in coming years, albeit at a slower rate than those implied by the international development goals for halving of hunger and poverty45. In order to meet these goals, increased effort by public and private organisations and farm communities is required. Table 5.4 indicates the potential for growth and poverty reduction within each of the farming systems in the region. Although the Tree Crop System, with its focus on relatively high value products with considerable scope for productivity gains, has perhaps the highest potential for growth, major gains in poverty reduction are more likely to come from the Rice-Wheat System where poverty is widespread and severe.

Table 5.4 also indicates the relative importance of five household strategies for escape from poverty. Clearly, given their importance across many different systems, strong support will be needed both for small farm diversification and for growth in employment opportunities in the off-farm economy. Measures that assist farm households to leave agriculture will be an important third priority, and would need to include improved rural education and vocational skills training. There is also substantial potential for poverty reduction by means of intensification of existing production patterns, largely through improved water management and adoption of improved technologies. Given the pressure on land, there is only limited opportunity for poverty reduction through the expansion of the farm or herd size of poor households. The following sections summarise the strategic priorities for actions, under the areas of policies, markets, information, technologies and natural resources.

Policies, institutions and public goods

Most countries in the region have policies that effectively favour urban areas and the manufacturing sector rather than rural areas and the agricultural sector, i.e. there are national trade and price distortions with a negative impact on the commercialisation of farming. For example, the need to maintain food prices within the reach of the increasing, politically articulate, urban population causes some governments to keep inter-harvest food prices artificially low by releasing stocks of grain as prices rise; thus adversely affecting farmgate prices. Because the majority of the poor are located in rural areas, poverty reduction efforts should be targeted towards increased farm household income. Moreover, successful rural poverty reduction has a positive impact on urban poverty reduction, but the converse is not true. There is a substantial advocacy challenge in this area, for which coalitions of civil society organisations are well placed to assist government.

Decentralisation and the strengthened performance of local institutions will be key strategic priorities for the development of most farming systems. Line departments in most countries have already given up some of their functions to the private sector, and plan to hand-over other functions to empowered communities. This trend is likely to become more pronounced in the future, but may require careful monitoring to ensure effective functioning, given the close relationship between rural community constituencies and politicians. Rural development will ultimately depend heavily on the effectiveness of local institutions, including community institutions. For this reason performance-based incentive schemes will be desirable. Public resources should be invested in the capacity building of local organisations, at both community and district levels. Another far-reaching change which is occurring in association with decentralisation and which needs to be encouraged, is the growing role of women in local level decision making, for example, panchayats in India. The growth of public-private partnerships for agricultural development should also be promoted, as they will have far-reaching effects on determining agricultural research priorities.

A high proportion of the increase in foodgrain production during the Indian Green Revolution occurred in districts with good irrigation facilities and adequate local transport. Agricultural development in many other areas of the region, however, has been constrained by a lack of infrastructure. In particular, the shortage of roads in remote and sparsely populated areas pushes up input and product transport costs46 and the lack of health and educational services reduce labour productivity47. Investments in roads and educational services should, therefore, be an essential ingredient of a strategy for increasing agricultural production and rural development.

Traditionally, irrigation water was available without cost in many areas - and was often seen as a way of supporting farmers and keeping down food prices. However, significant improvement in water management will only be possible if realistic water charges are introduced. In the case of surface water systems, better arrangements are required to provide an equitable balance between the benefits accruing to downstream users of irrigation water and electricity, and the well-being of the communities living in the upper catchment above and around the dam.

A double benefit, to growth and to poverty reduction, could result from an overhaul of obsolete land tenure policies and regulation. Increased operated size of fields and, over time, of farms will facilitate improved resource management and also intensification and diversification. In addition, open access resource regimes need to be converted to managed common poverty situations, within an effective legal framework.

The linkages between proper broad-based agricultural productivity growth and off-farm rural economic growth are well established48. However, less attention has been given to the integration of farm and off-farm economies, and in particular the types of developments which are simultaneously mutually reinforcing and which foster poverty reduction.

Trade liberalisation and market development

The three significant global issues that affect this region are: (i) gaining access to developed country markets in OECD countries; (ii) declining terms of trade for agricultural commodities; and, (iii) the continuation of subsidies for agriculture in developed countries. The terms of trade for rice, wheat and other foodgrains have been declining and the ongoing opening of economies exposes farmers to these forces, especially in the intensive Rice and Rice-Wheat Farming Systems. While some producers in these intensive systems may have the potential to remain internationally competitive, additional market development will inevitably be needed even in these favourable areas. Generally, market-related priorities within intensive systems need to address two issues: (i) reducing risk associated with monoculture activities by promoting enterprise diversification and value adding local processing; and (ii) the development of alternative income sources for landless and sub-marginal producers. There are many minor products that also generate income for poor households through the market; of which dried manure cakes and fodder are two important examples.

Producers in other systems with less potential in traditional products will need to develop market strategies that exploit alternative advantages. In relatively high poverty areas, such as the Highland Mixed Farming System, labour intensive activities such as horticulture and dairy have provided successful diversification pathways, but only if market access is sufficiently developed. These systems may thus require a particular focus on market-related infrastructure development, such as access roads and market facilities. However, less successful and more isolated households will leave the system, leading to a need for training and other exit support services.

Notwithstanding the attention being given to global issues, functioning local commodity markets and price information systems are of more direct interest to most smallholders. The experience of the recent past has demonstrated that market-led growth does not necessarily lead to benefits for the poor unless strong institutions are in place. Considerable investment is needed in these areas, and especially in the more remote farming systems such as the Highland Mixed and Pastoral Farming Systems. Similarly, further development of rural financial services, including micro-finance and linkages to mainstream banking, are needed for the full potential of farming systems to be realised.

Information and human capital

Investment in rural education is required for two principal reasons: (i) to equip workers with the skills to transfer from agriculture to the off-farm economy (whether locally or through permanent emigration); and (ii) to equip those remaining in agriculture to manage the emerging, knowledge-intensive, farming systems. Although total labour requirements in agriculture are not likely to increase, improved skills are required to increase efficiency and productivity. These local increases of human capital are needed in order to underpin diversification towards high value skill intensive enterprises as well as the development of small-scale local rural industry.

In addition, further investment in the training of professionals for research and agricultural support services is required. There emphasis should be placed on two aspects. First, mature and older professionals need to be updated in modern approaches and skills for agricultural development. Second, some advanced tertiary training institutions should incorporate soft systems approaches to agricultural education.

The shift to commercial knowledge in intensive farming systems requires a greatly improved flow of information concerning technologies and markets, to both farmers and support services. This emerges as a key issue and is a promising area for public-private partnerships. Publicly funded and managed information systems - particularly those providing market data - have generally proven to be of only limited value in strengthening the position of small farmers, while incurring substantial long-term operating costs. Technologies have been demonstrated which enable effective village level access to computer, or Internet based, information. A high priority for investment in the coming 30 years would be empowering small farmers to access Internet based information on markets, services and technologies.

Science and technology

Notwithstanding the existence of a range of technologies that are ready for adaptation and adoption by farmers, there is a continuing need for research efforts that respond to farmers' needs and to a number of strategic priority production problems; including soil degradation in the intensively cultivated systems, continuing germplasm improvement in food crops and, perhaps of greatest importance in the long run, effective biological nitrogen fixation (BNF) to boost cereal productivity through rotations with legume crops. There is already substantial capacity and a number of proven successes in the larger National Agricultural Research Systems (NARS) to exploit modern research techniques including: the potential of biotechnology, e.g. in the development of new genotypes possessing high yield potential and resistance to biotic stresses. Other potentials lie in the field of animal reproduction; recycling technologies such as biogas49; and value added through new agro-processing technologies.

The germplasm banks held by the NARS and the Consultative Group on International Agricultural Research (CGIAR), the latter under FAO custodianship, will be an important resource for public sector research and are also being utilised in private sector plant breeding efforts. Investments in support of agricultural research and extension - including livestock, forestry and fishery - will pay off handsomely if carried out using a problem oriented, farmer-centred multi-disciplinary approach. Two priority foci would be: (i) maintaining and expanding public research capacity, while at the same time ensuring its responsiveness to farmer needs; and (ii) concurrently building partnerships between the public sector, the private sector and farmer organisations. Demand-driven research structures will therefore be essential in the future.

In many countries in the region, the utilisation of inorganic fertiliser shows considerable scope for efficiency improvement. In the short run, priority should be given to better on-farm integrated nutrient management, combining inorganic and organic nutrient sources, and economic incentives for balanced fertiliser use - with reduced greenhouse gas emissions as an added advantage. These improvements should be combined with technological improvements, such as more cost-effective slow-release formulations, in order to reduce losses of plant nutrients. In the long run, breakthroughs in BNF, if they occur, could have a major impact on agriculture.

It is conventional wisdom that vigorous agricultural growth has to be established in order to prime the off-farm rural economy, and that both these sub-sectors are interdependent through consumption, production and labour market linkages. However, agricultural policy and development planning will need to take account of the increasing degree to which the roles are reversed. In many cases, off-farm employment is priming the rural farm economy and the growth of agriculture depends on investments derived from off-farm sources, including remittances. The typically lower return per labour day in agriculture compared to off-farm employment, as well as the seasonality of such off-farm employment, have important implications for labour availability on the farm - particularly at times of peak labour demand. Socio-economic research is needed concerning the impact of this phenomenon on the farm enterprise, as well as to identify the means by which farmer can adjust to the demands of off-farm employment. The policy implications for government also need to be better understood, e.g. the need to facilitate the provision of mechanised land preparation and harvesting services by farmer companies or the private sector.

Natural resources and climate

Natural resources will come under increased pressure, as a result of the growing demand for additional food and other agricultural produce to meet the needs of the expanding population. Population pressure is likely to be more intense in those farming systems already under severe strain, viz the Rice, Rice-Wheat and Highland Mixed Farming Systems. Thus, sustainable natural resource management depends on measures that limit increases in population pressure on the natural resource base, especially in marginal areas.

The little forest left in the region will come under ever-greater pressure during the coming 30 years. In most countries of the region forest departments have been slow to change their approach to forest management. This lack of progress has occurred despite ample evidence that sustainable productivity increases can be achieved by a joint system of forest management, which allows benefits to be enjoyed by forest dwelling communities. It seems evident that high priority should be given to re-orienting forestry development policy, and to changing departmental regulations to facilitate the establishment of joint forest management systems. These changes would increasingly empower local communities, while at the same time resulting in more efficient production of forest products. A further change needed in support of such development will be the re-definition of forest boundaries, leading to the development of optimal management systems for joint forest management areas and conventional forest departmental management areas. Such measures should be complemented by incentives for agroforestry, including fodder and commercial species, and community woodlots, for fuelwood and construction timber.

The management of water resources has transnational, national and local dimensions. Equitable agreements on sharing of water among states facilitate optimal water use. At national and local levels, conservation of water resources in cultivated areas will require the establishment of incentives for efficient use and mechanisms for the monitoring and regulation of water use - notably of groundwater levels and water quality. Water resources can be further developed through traditional means, such as construction or rehabilitation of tanks and tubewells, and improved water management.

Significant improvements are also likely to result from the introduction of realistic water service charges. This will be a question of `when' rather than `if' since the fiscal burden of running irrigation schemes can no longer be borne by governments in the region. Such improvements will include a wide range of improved management techniques, including laser levelling of irrigation basins, and both sprinkler and trickle irrigation. The greatest contribution in the future is expected from soil moisture conservation by means of mulching, bunding, relay cropping, windbreaks, plus general purpose tree planting, and, most importantly of all, water harvesting. This will include effective techniques for moisture conservation at farm level, through adjustments in cropping patterns, mulching, minimum tillage and relay cropping. Water harvesting has the potential to double the productivity of semiarid and subhumid areas50.

Soil degradation is a critical issue in most of the farming systems of the region. Priority needs to be given to on-going and new participatory research and extension programmes that address these problems, such as those of the Rice-Wheat Consortium for the Indo-Gangetic Plains which has already led to spectacular expansion of zero-till wheat in the Rice-Wheat Farming System. A number of successful participatory, community-based pilot watershed management projects in the Rainfed Mixed and Highland Mixed Farming Systems have provided the methodology and experience to improve the management of the large number of critical watersheds in the region51.

Governments are likely to continue, quite correctly, to give greatest attention to the relatively high potential, well-watered cereal producing areas of the Rice and the Rice-Wheat Farming Systems in order to ensure an adequate level of self sufficiency in key staples at acceptable prices. The main technical strategy will be to improve the efficiency of production. The potential is already evident in areas, where relatively cheap tubewells have provided a timely supply of water at transplanting - which the state-managed surface irrigation schemes often cannot guarantee - which boost the yields of rice and the following crop. Very significant changes can be anticipated as realistic water chargers are introduced. These will involve a wide range of water management techniques including laser levelling of irrigation basins, sprinkler and trickle irrigation; and the introduction of greater moisture conservation, including zero tillage, mulching, windbreaks and relay cropping.

The high potential areas will also need considerable attention to better soil management. In the Rice-Wheat Farming System, where intensive cultivation and unbalanced fertiliser applications have led to a deterioration in soil structure and fertility, solutions are now being developed to tackle the immediate problem. However, it will be important to develop a range of technologies for sound soil management under highly-intensive, continuous, irrigated cultivation. The need for precision management of fertiliser nutrients is a closely related issue, not only from the point of view of efficiency in use for grain formation, but also to minimise groundwater pollution. Improved pest management will also be necessary, and it is expected that this will be biologically oriented for most crops and, increasingly, for livestock. While these areas will remain quite highly specialised in cereal production, there will be a degree of diversification into high value enterprises - including orchard crops and livestock. Household livelihoods will also diversify, and an increasing share of off-farm income is expected for smallholder households.

The low potential areas, such as the Rainfed Mixed, Highland Mixed and Pastoral Farming Systems, are intrinsically less able than the lowland areas to provide production responses in food crops. Sustainable resource management with water harvesting is required to underpin diversification into high-value produce as market access spreads. The severe poverty prevalent in these areas will probably lead to continuing seasonal and permanent migration, thereby limiting the increase of population pressure on the resource base.

However, there are some significant possibilities for development. A community-based micro-watershed approach, such as that which is succeeding in parts of India, would form a sound basis for the development of the Rainfed Mixed Farming System. Conservation agriculture should be introduced; including the greater integration of livestock and trees into the farming systems - with fodder trees, catch crops and green manure crops. Such programmes would include the introduction of cash enterprises, such as dairying or orchard crops, wherever market access is feasible. Greater attention to soil and water conservation should also form a vital element in such programmes; including zero tillage, mulching, relay cropping, windbreaks and on-farm tree planting. Development of sound and sustainable agroforestry and forestry operations, with participatory management for much of the forest area, will be important in maintaining a good forest cover in the upper catchments, as well as on the large areas of wasteland that currently exist in the region.

Conclusions

Agricultural development will remain an important component of poverty reduction programmes for the foreseeable future in South Asia. There are some major resource degradation challenges, as well as linkages to the off-farm rural economy, to be taken into account. The main source of reduction of hunger and poverty would be diversification to high value enterprises, including local processing. Increased off-farm income and intensification of existing production patterns are next in importance, followed by exit from agriculture. Increased farm size is expected to be of lesser importance. Four broad strategic initiatives are proposed:

Improved water resource management. Improved water management is essential to support the intensification and diversification of production and to reduce resource depletion, for both surface and underground water schemes. Components include: efficient technologies; conjunctive use; water charges and other regulatory measures; water users' associations; and watershed protection.

Strengthened resource user groups. Strengthening resource user groups is one way to redress the extensive land and water degradation in plains and hills, and protect watershed resources. Components include: resource management groups for watershed management in hill and mountain areas; range management groups in pastoral areas; and policies to encourage effective common property resource management.

Re-oriented agricultural services. The re-orientation of agricultural research, education, information and extension systems to involve farmers fully will underpin the drives for intensification and enterprise diversification and promote sustainable resource management. Components include: models for joint public-private service provision; pluralisitic advisory services; Internet based delivery of market and technical information to small farmers; and the incorporation in higher education systems of interdisciplinary learning and approaches.

Improved rural infrastructure. Returns to transport and health investments are high and beneficial to the poor, especially in low potential and highland areas in the region. Components include: roads; drinking water; schools; health facilities; and effective models for private sector participation.

1 See Annex 3 for the countries in the region.

2 FAOSTAT.

3 FAOSTAT.

4 See Annex 5 for an explanation of agro-ecological zones.

5 World Bank 2000a.

6 Jazairy et al 1992.

7 Government of India 2000.

8 Government of Pakistan 1998.

9 World Bank 2000a.

10 FAOSTAT.

11 FAO 1998b.

12 Intensive is used here to indicate the growing of two or more crops of paddy during the year.

13 FAOSTAT.

14 World Bank 2000a.

15 FAO 2000a. Except where specifically noted, quantitative information in this section is derived from FAO (2000a).

16 World Bank 2000b.

17 It is possible that average farm size will increase somewhat by 2030, as a result of continued migration of rural population to urban areas and because of economies of scale, particularly through mechanisation.

18 Tran and Nguyen 2001.

19 A recent World Bank supported water resources sector review calls for a shift of emphasis from development to management.

20 Even though groundwater resources in these farming systems are replenished annually, unlike other regions.

21 Government of India and World Bank 2001.

22 Fischer et al 2001.

23 The pro-active leadership of these Councils is demonstrated by their long-term strategic planning.

24 Fan et al 2000.

25 Anderson 1994.

26 FAO 2001.

27 While attention to the HIV/AIDS-agriculture linkages has focused on Africa, it is conceivable that AIDS will also cause major suffering and affect agricultural productivity during the coming 30 years elsewhere, including rural South Asia.

28 Farrington and Thiele 1993.

30 It is estimated that subsidies and transfers account for nearly 40 percent of Government of India expenditure (World Bank 2000a).

31 In Gal Oya and Mahaweli schemes in Sri Lanka, some 500000 farmers have organised into 33000 water user associations, with improved cropping intensity and reduced irrigation problems.

32 In excess of 80000 farmers have been trained in IPM and related improvements in rice, cotton, sugarcane and oilseeds.

33 These opportunities related to specialised market opportunities or unused farm resources, such as fish production in paddies after pesticide use is discontinued, can be termed `niche' or `patch' improvement of systems (Pretty 2000).

34 Hoque 2001.

35 As witnessed during the implementation of Government of Bangladesh/FAO/ADB Horticultural Development Project.

36 Dugdill and Bennett 2001.

37 However, some promising minimum and zero till resource conservation technologies have been tested in the field, and are currently spreading rapidly - it is estimated that 100000 acres will be under these improved methods this year in Haryana (Hobbs, pers. comm).

38 Moench 2001.

39 ICM has been effectively applied to increase rice yield and productivity in Australia, Egypt, the Republic of Korea and the FAO project BGD/89/045, Thana Cereal Technology Transfer and Identification in Bangladesh.

40 Several technology packages have been successfully tested, including those from Pakribas and Lumle in Nepal, and also permaculture in Jajarkot, Nepal. The development of apple industry in Himanchal Pradesh was often cited as a major development success, but the industry has recently come under pressure from apples imported from East Asia.

41 As demonstrated in the FAO/UNDP Farm level Applied Resource Management Programme.

42 Kiff and Pound 2001.

43 Gulati and Kelley 1999.

44 Abstracted from Case Study 4, Annex 1.

45 FAO 2000a.

46 Fan et al (2001) report higher rates of return for investment in roads and research in rainfed areas than in irrigated areas in India. Notably, their poverty reduction potential was also higher.

47 While attention to the HIV/AIDS-agriculture linkages has focused on Sub-Saharan Africa, it is conceivable that AIDS will also cause major suffering and affect agricultural productivity during the coming 30 years in rural South Asia.

48 See Chapter 1.

49 The benefits of biogas could be extended to high altitudes if cold tolerant digesters could be developed, which could have significant positive impacts on reduced deforestation, reduced women's work and improved recycling of nutrients to crop production.

50 Agarwal, pers. comm.

51 The approach of improved watershed management combined with income generation (which was effectively applied in Tunisia, as described in Box 3.6 of Chapter 3) was pioneered through a global network which included one successful pilot project in the Highland Mixed Farming System in Nepal. Some aspects of this approach could enrich the existing participatory approaches to watershed management in the region.