Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Part III - Selected issues: Change and the environment
Introduction
9. Sustaining and expanding the 'green
revolution' in rice
Ecological policies for sustaining high
production in rice: Observations on rice intensification in
Indonesia
10. The problems of upland land management
Pitfalls of the intensification debate
Immediate
needs in upland Java
11. The hazard of fire
A
wider view of the fire hazard
The
need for management
12. Coastal, inshore and marine problems
The south-east Asian response
THE range of environmental and related issues in South-East Asia is very wide; they cannot all be dealt with in one conference, or one book. This part takes up four subjects in some depth; two are essentially agricultural, one is a specific problem that has risen to major significance since 1982, and the other concerns the marine environment. At the conference, each of these issues, and also those treated in Part IV, was given an hour and a half in order to permit time for a full discussion. Some of these matters were covered in depth, and related areas of significance were introduced.
Chapters 9 and 10 take up questions raised above and treat them in much greater depth than in Chapters I and 2. The Green Revolution in rice production began in South-East Asia, at the International Rice Research Institute (IRRI) at Los Baņos in the Philippines. Since the mid-1960s, the development and diffusion of new rice varieties, and of the agro-technology required to make best use of them, have transformed food production in the region. Mainly from this cause, regional rice production has more than doubled between 1970 and 1990.
The 'father' of the rice revolution, the plant-geneticist Te-Tzu Chang, is the author of the first paper in Chapter 9. He does not dwell on achievements, however, but rather discusses the problems that have arisen to threaten the sustainability of these advances, and the management improvements that are required in order to make further progress. He offers a detailed set of prescriptions for better systems of management that could sustain gains to date, extend them to wider areas, and press some way further forward, refraining from any cornucopian prognostications arising from the new advances in biotechnology.
Chang is followed by the anthropologist James Fox, who has had deep involvement with these problems in Java. From this experience, he writes much more than just a discussant comment, but rather a companion paper that expands and further deepens the argument. The two papers in Chapter 9 are a pair, as the presentations on which they are based were a pair in Yogyakarta.
It is urged that much more attention in research and public assistance be given to farmers in the upland, unirrigated areas, and in Chapter 2, Concepcion has already stressed the growing pressures on this land. The question of the upland areas and their management is the specific topic of Allen's Chapter 10. He draws on years of field experience principally in Papua New Guinea, but in this chapter he reviews the often alarmist literature on other areas and finds important contrasts in the degree of unsustainability of upland management. He also stresses the historical recency of most steepland occupation, and the dramatic effect of new roads, new crops and new off-farm jobs in creating a destabilizing, unfamiliar situation to which today's upland farmers are trying to respond.
In discussion, Hardjono takes a more gloomy view from her research in West Java, but comes to the same basic conclusion as that reached by both Chang and Allen: a new set of public policies and attitudes in relation to the hitherto neglected upland areas is urgently required. Rerkasem takes up a question of the meaning of intensification, which was prominent in Allen's verbal presentation; this led to an exchange of practical as well as theoretical significance.
In 1982-3, large forest areas in eastern Borneo were burned and, as has already been noted, there were new fires in 1991. Wirawan has studied the real impact of these fires on forest ecology for several years, adding greatly to what is known about them, and modifying substantially some early statements about the destruction. In Chapter 11, he provides a comprehensive statement on the results of research, his own and that of others, in Kalimantan, and at the same time takes up themes earlier developed by Potter (Chapter 5) and Nicholls (Chapter 7). Some important proposals are made for better control, proposals that might advantageously be heeded in 1992 after the second set of major fires in less than a decade. The two discussants, Kartawinata and Soerianegara, broaden the scope of review to a regional level.
A large part of South-East Asia is sea, much of it quite shallow and, being at low latitude and enclosed, not often disturbed by high winds; higher-energy seas surround the Equatorial archipelago on all sides. The marine food resources are of major importance to the region's population, but the seas include principal international shipping routes, contain important oil and gas fields, and are also the sinks for a growing volume of pollution from the increasingly populous land. In Chapter 12, the only discussion of the marine environment in this book, E. D. Gomez examines this competition, and the increasing interference, and reviews the problems of managing this international environment. Discussing his paper, Soegiarto details some international initiatives, and the participation of United Nations agencies.
9. Sustaining and expanding the 'green revolution' in rice
An overview of the 'green revolution' in
south-east Asia, 1970-1989
Problems of the green revolution
Areas
for future endeavour
TE-TZU CHANG
An overview of the 'green revolution' in south-east Asia, 1970-1989
THE 'Green Revolution' in rice over the 1970s and 1980s was based mainly on the use of irrigation water, high-yielding varieties (HYVs), chemical fertilizers, and the control of insect pests by chemicals, multiple monoculture of the rice crop and various forms of government support. The combined impact of these components on rice production was most obvious in South and South-East Asia, mainland China and South Korea.
In South-East Asia, the increase in grain production was greatest in those areas where water could be regulated or used for irrigation, the rate of HYV adoption was high, quick acting fertilizers were liberally used, major insect pests were controlled by chemicals and/or varietal resistance, and attractive incentives were provided by government subsidy or price support. Growing two or more crops of rice on the same land further added to production (Chang, 1979). However, the unprecedented rise in grain yield was confined to the more favourable production environments, under irrigated and rainfed-wetland cultures, and in areas without political or military strife. Patterns of change in production vary markedly from one country to another (Barker, Herdt and Rose, 1985; Hsieh, Flinn and Amerasinghe, 1982).
Even within a favoured rice-growing region, rice yields per crop have been variable and affected by a range of factors. Climatic factors, including drought, flood, or the alternation of the two, adversely impact on grain yield over a wide region; for example, 1972 was a flood year while 1982 and 1987 were drought years. Grain yields are generally higher in the dry season than in the wet season, largely due to the difference in solar radiation intensity.
Disease and insect outbreaks have been important. Irrigation coupled with heavy nitrogen fertilizer use, and continuous monoculture of one rice variety, have together led to heavier and more widespread infestations than before. The most destructive pests were the brown planthopper and the tungro virus disease.
Variations in the supply of fertilizers at equitable costs have had notable effect. Oil crises in the 1970s, and a rise in fertilizer price since the mid-1980s, have had dampening effects on rice production (Chapman and Barker, 1987). This is the case in the Philippines where rising fertilizer costs and a stagnant rice price have resulted in production drops (David, 1988).
There are also other and unquantifiable elements. Government subsidies or pricesupport programmes have been a major factor in stimulating or depressing rice production. J. J. Fox (1991) and David (1988) have documented these effects for Indonesia and the Philippines, respectively. Moreover, rice farms vary greatly in area, tenure status, technology adoption, farmers' experience, capital availability and other production factors (see Barker, Herdt and Rose, 1985). With the persistence of both biological and socio-economic constraints, on-farm yields continue to lag behind those potentially attainable (K. A. Gomez, 1977). Ineffective transfer of improved technology from the researchers to the farmers accounts for part of this gap.
Historical accounts of the Green Revolution in rice have been given by Chandler (1968), Chang (1987, 1988), Dalrymple (1986) and the International Rice Research Institute (IRRI, 1972). Herdt and Capule (1983) and Dalrymple (1986) detailed the spread and distribution of HYVs in different countries. By 1987, the HYV area in SouthEast Asia exceeded 18 million hectares (David, 1991). The contribution of specific factors (irrigation, HYV, fertilizer and others) to rice production increases in Indonesia and the Philippines during 1965-80 was studied by Herdt and Capule (1983).
Rice production in the region rose from 53.5 million tonnes in 1970 to 112.4 million tonnes in 1990. The rate of increase was 28.7 per cent during 1968-79 and 34.8 per cent during 1980 9; the corresponding yield increases were 21.4 and 23.9 per cent. Within South-East Asian countries, new technology has had little impact in Cambodia and Malaysia. Grain yield per hectare showed a small increase in Malaysia, while Cambodia experienced a distinct drop in most years due to continued military strife. Laos made a notable gain in both production and yield. Myanmar had one of the largest gains in grain yield (82 per cent in 1984); however, the three top-ranking varieties are not the HYVs (Win and Win, 1990). There was an 84 per cent yield increase in Indonesia, largely due to the new technology and government support. The Philippines and Vietnam followed with 55 and 51 per cent increases, respectively. Thailand's rice production grew by 47 per cent, but due more to increases in planted area (a peak of 30 per cent in 1985) than in grain yield which was nil (IRRI, 1988b). Production and yield changes in eight countries of the region are shown in Figures 9.1-9.2.
Problems of the green revolution
The Green Revolution in rice was certainly a much welcomed development during the late 1960s and the early 1970s when the spectre of region-wide food shortage had been forecast. It has staved off the feared crisis, but various developments associated with it have given rise to a range of new problems. Most of these problems can be attributed to the inappropriate use of new technology, but others were neglected by decision-makers (Chang, 1988).
Water Use
Costly irrigation projects have often failed to attain the expected effectiveness and efficiency of usage due to poor management. Both technical problems and human factors are involved in the disappointing result. Rising construction costs of new projects, less suitable sites for prospective projects and low returns from past efforts are important factors leading to a drop in new irrigation works (Brown, 1989; Chang, 1988). The annual rate of increase in the irrigated land area in South-East Asia has dropped from the 3-4 per cent average of 1970 85 to 1.5 per cent during 1985-8 (David, 1991). Prospects for a new expansion in irrigated land are therefore limited (IRRI, 1989; Levine et al., 1988). Moreover, nearly all irrigation works are adversely affected by silting, salination and erratic weather. Distribution systems are generally ineffective and wasteful, while water will undoubtedly become more costly. Meanwhile, irrigated land areas near urban centres are contracting at alarming rates.
Use of poor-quality underground water and excessive pumping have resulted in rising soil salinity or alkalinity which also leads to zinc deficiency. Decline in soil productivity due to the changes in soil properties is becoming more widespread (De Datta et al., 1979; Pingali, Moya and Velasco, 1990). Partly due to prohibitive costs, drainage works are generally neglected. Floods are another yield-destabilizing factor in many parts of mainland South-East Asia. In waterlogged soils, standing water of shallow depth also constrains yield.
Pest Infestation
Continuous planting of a few genetically related and similar HYVs, often under double or triple cropping over a wide area, has led to the appearance of new biotypes in variable insect pests or pathogens and to the repeated breakdown of varietal resistance of the vertical type, causing large yield fluctuations. Pest incidence was aggravated by staggered planting dates within a region (Chang, I 988). Such repeated pest infestations by the brown planthopper, causing heavy damage to rice crops, have occurred in Indonesia, the Philippines and Vietnam (see Dyck and Thomas, 1979), and the phenomenon has been aptly described as a 'boom and bust cycle' (R. A. Robinson, 1976). Insect pests with a short life span, with inherently large variability in field population, such as the brown planthopper, can quickly react to a change in the resistance gene of the new cultivar; they do so by shifting their population structure to a different biotype that can overcome the defence mechanism (mainly of non-preference by the insect) of the newly bred cultivar. When grown under double or triple cropping, the effective life span of a resistant variety is reduced to 3-5 years. Sequential release of resistant varieties with the same lineage of descent, such as IR36, IR42, IR64 and IR70, are prone to favour rapid shifts in insect biotypes (Chang, 1984, 1988; Saxena and Barrion, 1985). A large number of the HYVs not only have the same semi-dwarfing gene (sd 1 ) but also the Cina cytoplasm, a combination that renders them more vulnerable to serious epidemics (Chang, 1984; Hargrove, Cabanilla and Coffman, 1988).
Indiscriminate use of wide-spectrum insecticides reduces the natural enemies of the rice pests and has led to the resurgence of the brown planthopper following repeated applications of the chemicals (Heinrichs and Mochida, 1984). Inappropriate application methods also raise production costs without attaining effective control, while excessive chemical residues harm the ecosystem.
Efficiency of Fertilizer Use
Most rice farmers have yet to obtain full returns from their inputs of chemical fertilizers, especially nitrogen, because of improper methods of application, principally the use of top dressing only, by delayed application (often due to delayed arrival), by poor weed control or ineffective water management, leading to leaching, volatilization or runoff. The efficiency of nitrogen fertilizers is less than 40 per cent (Craswell and Vlek, 1979; De Datta, Magnaye and Moomaw, 1968).
Restricted use of phosphorus, potassium fertilizers, other elements in the 'minor' category (calcium, sulphur, iron, silica, boron) and organic fertilizers, where nitrogen is the sole additive, may lead to an imbalance in soil nutrients. Excessive nitrogen use in some instances may have adverse environmental effects.
The 'Yield Plateau'
A record crop yield per season of 11 tonnes per hectare was established by planting IR8 and IR24 in the late 1960s (see Chang, 1988). Later releases have not surpassed the apparent yield ceiling, although the grain yield per hectare per day has been raised by 40 50 per cent in the earlier-maturing varieties. Enriching the air with carbon dioxide (CO2) could increase grain yield by 10 per cent (Yoshida, 1981). Hybrid rice may add 15-20 per cent to the yield, but the technology is beyond the reach of rice farmers in South-East Asia.
Production Costs and Government Support
The return from nitrogen use in relation to the rice price has been continuously eroding from I (N): 10 (rice) in 1972 to 1: 5 in the mid-1980s (Chapman and Barker, 1987). Continued rising costs of labour and chemicals, especially nitrogen fertilizers, have markedly increased the cost of production inputs all over the region. Wetland rice, especially when it is manually transplanted, will remain a labour-intensive crop. The real price of rice has been steadily declining on the world market since 1976 (David,1991) and will continue to do so (World Bank, 1990c) because of the cheaper price of wheat. Unless government support continues or the output/input ratio is raised, the incentive for increased yield quickly disappears. The Philippines and Indonesia have experienced difficulties in maintaining a balanced rice economy after attaining self-sufficiency. Massive and continued government support became burdensome. In contrast, continued production increases in Myanmar, Thailand and Vietnam were sustained by expanding export markets.
While the main beneficiaries of the Green Revolution are the HYV adopters, the rice consumers and the government, most urban consumers are not aware of the farmer's arduous role in increasing food production. As a result, rice farmers often lack public support when they face problems beyond their ability to cope. In most areas, rice farmers have yet to organize themselves into community-wide organizations such as rural cooperatives, irrigation associations and seed producers' associations.
Change in Farming Systems
Wherever higher rice yields can be obtained with the new technology, farmers tend to replace other crops in the production system, often a legume, by planting another crop of rice. Similarly, the area planted with a green manuring crop in the winter season has also declined in the 1970s and 1980s. These changes may affect human nutrition in terms of plant-protein supply. A legume crop in the winter season not only helps to sustain soil productivity but also interrupts the continuous presence of rice plants in the field-one means of reducing rice pest populations. Although no serious malnutrition has been reported from HYV areas' milled rice alone, containing sufficient protein and essential amino acids to support a labouring adult, is inadequate to meet the needs of a growing child (Hegsted, 1969).
Equity among Rice Farmers
During the 1970s, socio-economic scientists were concerned about the income gap between large and small rice farmers who adopted the new technology. After years of extensive research and surveys, such a fear proved to be largely unfounded (Anderson et al., 1985; Herdt and Capule, 1983; Kikuchi and Hayami, 1982). On the other hand, the gap between farmers in favoured (irrigated and well-watered rainfed-wetland) areas and subsistence fanners in rainfed areas, especially the dryland and deepwater environments, continues to widen. The subsistence farmers have yet to benefit from modern technology (IRRI, 1989).
Another little appreciated fact is that rice can support more people per hectare of land than other cereals (Lu and Chang, 1980); hence, human population growth usually accompanies the adoption of irrigated rice (Chang, 1987; Hanks, 1972). The positive association between rice-production increases and population growth can be readily gleaned from available statistical data. While the adoption of HYVs has levelled off in the major rice-growing nations of tropical Asia, human expansion continues at an unabated pace (Brown, 1989; Concepcion, Chapter 2).
With a view to stimulating discussion, the following areas for future endeavour are proposed:
1. Water use must be made more efficient and less detrimental to soil productivity-irrigation projects in South-East Asian nations are rapidly reaching affordable limits. Conserving the existing resources and maximizing water-use efficiency for rice and other crops in the rotation system are of paramount importance to sustained land productivity. Rice is very much a water-consuming plant (King, 1966). Wasteful irrigation practices must be curbed. Varietal differences in water-use efficiency should be explored and implemented. Meanwhile, the use of gravity-flow or underground water of high salt content must be controlled in order to check salination.
2. The restoration of biotic diversity in major cultivars is an important area of concern. Host resistance can be made more effective and durable only when the genetic diversity in commercial cultivars is reinstated, resistance genes are appropriately deployed for different seasons and/or areas and in relation to the prevalent races of the rice pest. More stable forms of pest resistance need to be developed. In addition, varietal mixtures, multiline varieties and intercropping will help to slow down population changes in a pest. New resistance genes may be exploited in distantly related grasses.
3. Chemical fertilizers should be used more efficiently in combination with other plant nutrients or soil ameliorants. The efficiency of nitrogen fertilizers can be raised significantly by improved formulation, deep placement in the soil, proper water management and weed control, and more refined timing in top dressing. The addition of other mineral elements will help to remedy nutrient deficiencies or imbalances. Supplementing chemical fertilizers with organic matter promotes a productive soil structure in the long run, and biological nitrogen fixation on subsistence farms will boost production and reduce costs.
4. Innovations from biotechnology have to be widely implemented. In addition to the popular focus on genetical engineering by molecular manipulations, equally fertile areas of application exist in the biological control of insects and diseases and the enhancement of biological nitrogen fixation by soil microbes. Research by multidisciplinary teams of scientists will open up new fields for crop production by diverse means. Concurrent developments in cellular and molecular biology have already demonstrated promise in tapping useful genes in novel gerrnplasm of distant relatives, especially when the rice plant is well adapted to tissue culture and cellular manipulation (Chang and Li, 1991).
The rice plant can also increase its tolerance to waterlogged conditions and provide potential for root-zone nitrogen fixation to other dryland cereals by virtue of its aerenchyma tissues in the stems and roots-a feature that may grow in importance as the global warming trend leads to sea-level rises and flooding in coastal or lowlying areas (Chang and Vaughan, 1991).
Rice is a C3 plant and, therefore, relatively poor in both photosynthetic efficiency and water use compared to C4 plants such as maize, sorghum and sugar-cane. Biotechnological advances may improve the efficiency of rice in these aspects through wide hybridization and gene transfer by tapping the gene-pool in C4 grasses (Chang and Vaughan, 1991). In the more immediate future, improving the harvest index of rice varieties from 0.5 to 0.6 appears to be an attainable goal in raising grain production in proportion to the total biomass of the plant. A yield of up to 15 tonnes per hectare for irrigated rice in the Tropics is a plausible target (Akita, 1989).
5. Improved production technology plays a vital role in lowering costs. The need to improve the output/input ratio in rice production can be met by developing innovations that will reduce labour use and time lag in growing the labour-intensive rice crop. Better soil tillage tools, seeders, fertilizer applicators, transplanters, weeders, reapers and grain threshers are in various stages of development. As the movement of rural males to urban areas will continue and expand, an increasing proportion of the on-farm chores is being handled by women and children. New machinery and tools should be gender neutral so that they will ease the manual chores rather than displace the workers; hence, small family-sized machinery will win initial acceptance, while larger machinery run on a community-wide co-operative basis will be more efficient.
6. Improved farming practices incorporating traditional agrosystem conservation methods and cropping patterns can be promoted. Crop diversification effectively reduces pest epidemics, increases plant-protein supply and helps to sustain land productivity. A legume, or green manuring crop, would grow during the winter months in many areas, while wetland-rice cultivation and aquaculture could be complementary. In hilly areas, practices such as interplanting of different crops, hedge-row planting and terracing will reduce soil erosion. Agroforestry is a new promising venture for such areas and diversified farming generally boosts and stabilizes farm income.
7. Increased aid to subsistence farmers in rainfed areas requires effective action programmes. Dryland, deepwater and tidal swamps occupy about 22 per cent of the total rice land area in South-East Asia. These farmers have not benefited from the modern technology that launched the Green Revolution or received sufficient public assistance. Their concentration in environmentally fragile areas poses a great threat to the ecosystem of the adjacent regions, especially those downstream. Such a vulnerable situation exists particularly in the case of upland (dryland) rice in hilly areas. Variable water supply, low soil productivity and meagre production inputs are common denominators in all three types of rice culture. The indispensable varietal improvement efforts should be reinforced by affordable production packages that will sustain land productivity.
In upland areas, soil erosion can be reduced by zero tillage, interplanting rice with deep-rooted plants, mulching the soil surface, and contour planting or terracing. Soil nutrient supply can be augmented by crop-diversification schemes including legumes, and nutrient recycling expanded by crop-livestock integration. The upland rice plant can be modified to develop more extensive roots and a greater competitiveness with weeds.
In deepwater and tidal swamp ecosystems, rice varieties with increased tolerance to variable water depth and adverse soil factors are needed so that more stable crop production, rather than unrealistic yield increases, can be attained. Ratooning-allowing new shoots to grow from basal stems after cropping-and green manuring would raise cropping intensity, while fish production could be incorporated into the farming system. Pest control should rely more on genetic resistance and biological control rather than chemical control. Among the different rice cultures in South-East Asia, the deepwater and tidal wetlands offer greater potential for expanded rice cultivation.
Production constraints in rainfed areas are more location- and season-specific than in the irrigated areas. Rainfed rice cultures also suffer from a lack of experiment stations and trained workers to handle the complicated production problems in their areas. Therefore, on-farm testing will assume a greater importance in research and development before the improved technology is passed on to the growers. Rice farmers themselves need to be closely involved in the testing process, more as partners than mere recipients.
Above and beyond the above scientific endeavours, greater inputs into the politico/socio-economic aspects of agricultural reconstruction are imperative to meeting the future demand for rice. National policy makers must exercise strong and clear political will, in helping the beleaguered rice farmers in obtaining equitable returns from rice cultivation. The close association between a balanced national rice policy and steady growth in rice production in a given country is apparent in each of the South-East Asian countries (Parker, Herdt and Rose, 1985; David, 1988; J. J. Fox, Chapter 9A; IFPRI, 1977) and also indicated in Figures 9.1 and 9.2. On the other hand, government policies need to be periodically reviewed and modified in the light of changing circumstances. Support from the general public will add weight to the farmers' voice in the decisionmaking process (Chang, 1988).
Rice farmers themselves need to be more strongly organized in production and marketing activities. There is ample evidence to show the increased effectiveness of organized farm communities in dealing with irrigation works, integrated-pest management practices, use of modern farm machinery, and communal use of seedling nurseries, grain combines, grain driers, pure-seed nurseries and mechanized transplanters (J. J. Fox, 1991; Shen, 1974). Grain godowns and rice mills can be set up on a community basis in anticipation of increased production. South-East Asian countries can learn from the successful experience of Taiwan in the early 1970s (Shen, 1974).
Technology transfer from the rice researchers to the growers has been a weak link in raising on-farm yields even after the improved technology is proven sound and workable. Many extension workers in the region lack farming expertise and experience which are essential in winning the farmers" confidence. Their numbers are insufficient for the widely scattered farms, although the International Rice Research Institute (IRRI) has devoted many resources to the training of extension workers in rice production (IRRI, 1985). Again, Taiwan's experience shows the importance of vocational schools in producing innovative farmers and production specialists. Most countries in the region do not provide sufficient educational opportunities for promising farm youth, who wish to pursue a career in agriculture.
In the developing countries, their limited national research resources could be more fully used if nations facing common problems formed a research consortium to share gerrnplasm, research findings and expertise. Such networks for rainfed-wetland, deepwater, tidal-swamp and dryland rice were initiated in tropical Asia during the 1980s, with the IRRI serving as the co-ordinating agency.
In conclusion, all sectors of society must collaborate in conserving a productive and sustainable rice ecosystem which is vital to future food security in the region. Rice farmers need increased help in order to continue their lifelong service to mankind. It will be an unprecedented challenge for all concerned to maintain rice production increase at the annual rate of 2.6 per cent over the 1990s (World Food Council, 1991) in order to cope with the projected human population expansion.
Introduction
An outline of Indonesia's achievements in the
green revolution
The problems of expanding production in
Indonesia
Assessing the yield potential of the green
revolution
Diminishing biotic diversity and increasing
vulnerability
Facing the problems of sustainability
Rice production and sustainability in the 1990s
Editorial
comment
JAMES J. FOX
TE-TZU CHANG can genuinely be described as one of the founders of the Green Revolution in rice. He was a member of the original plant-breeding team at the International Rice Research Institute (IRRI) that crossed a semi-dwarf variety of rice from Taiwan with a long-stemmed vigorous strain from Indonesia and, by bringing together these two distinct forms of rice from different parts of Asia, succeeded in producing the first high-yielding variety (HYV). It is an exceptional honour to be on the same programme with him and to be asked to comment on his contribution.
His chapter succinctly points to a whole range of issues that are directly related to sustaining and expanding the Green Revolution in rice. The comments here will focus on a number of his most important points that relate to the author's own research on the development of rice intensification in Indonesia.
Chang reports an 84 per cent yield increase in Indonesia since 1970 which can, to a large extent, be attributed to the adoption of new rice technology. This is the highest yield increase for any of the countries of South-East Asia. Comparison of Indonesia's 1990 yields with the rest of South-East Asia is particularly instructive. The average yield in a majority of the countries in the region (Laos, Malaysia, Myanmar, the Philippines and Vietnam) stands at about 3 tonnes of rough rice per hectare. As Chang indicates, these yields have been achieved in various ways. The Philippines, Vietnam and Laos show steady upward growth in yields from the time of the adoption of the new technology.
This same technology has not, however, had much impact in other South-East Asian nations. Thus, except for a short-lived spurt in the early 1980s, Malaysia's yields have remained relatively unchanged since the 1970s. (1) A similar plateau has been maintained in Myanmar since the 1980s. In the 1970s and 1980s, Thailand's yields also held steady, but at an average level of only 2 tonnes per hectare. Cambodia's situation is even less satisfactory where average harvests have returned just over 1.3 tonnes per hectare. Cambodia's case is all the more disconcerting in that the Khmer Rouge's ill-conceived and enormously destructive effort at revolutionary collectivization was carried out as an express attempt 'to achieve three tons of rice per hectare' (Chandler, Kiernan and Boua, 1988).
Indonesia stands out in all of South-East Asia by achieving an increase in average yields to almost 4.5 tonnes per hectare. This achievement involved a national commitment to the effective and appropriate use of the available technology. Such a goal was not reached easily, nor was it allowed to falter despite major set-backs. The initial technology itself proved to be 'inappropriate' in certain crucial features, and in the course of the development of this technology, the nature and understancling of the Green Revolution had to be substantially modified.
An outline of Indonesia's achievements in the green revolution
If one focuses too narrowly on agriculture, it is possible to overlook the fundamental changes that were required for the effective adoption of the new technology in a large nation such as Indonesia. To achieve its goals, Indonesia had, first of all, to establish an agricultural support system consisting of: (i) an entire network of seed-production facilities to produce HYVs of rice; (ii) research facilities with experimental stations throughout the country to provide the knowledge necessary to adapt the technology to local conditions and to monitor its development; (iii) an extension service with the capacity to instruct and inform farmers in new methods of agriculture; and (iv) an administrative bureaucracy able to direct national policy to local areas.
Indonesia, however, had to do far more than just provide an agricultural infrastructure. It also had to make major investments to: (i) improve its irrigation systems; (ii) develop a massive fertilizer industry to support agricultural production; and (iii) create a national transport and storage network for distributing agricultural inputs, such as fertilizers, and a further network of storage facilities for rice and other agricultural products.
Furthermore, it was necessary to establish an entire rural banking system to channel credit to rice farmers to assist them in their initial adoption of the new inputs. In addition, a network of local co-operatives was created to support farmers. Finally, but no less important, Indonesia had to establish and coordinate a management and logistic planning authority with the capacity to oversee and adjust a pricing policy that would maintain a favourable relationship between rice and fertilizer to foster production. (2)
That Indonesia managed to achieve its principal goals during the course of two decades is a well-known story. (For a recent account, see J. J. Fox,1991.) In the late 1960s and in the 1970s, Indonesia was the largest rice-importing nation on the world market. By 1985, it had achieved technical sell:sufficiency in rice, had reserve rice stocks of approximately 2 million tonnes and, though not exporting rice, was involved in 'lending' rice to Vietnam and the Philippines. At the start of the rice intensification programme, production averaged around 11 million tonnes of milled rice. By the mid-1980s, it had more than doubled to over 25 million tonnes and by 1989, it had reached 30 million tonnes. By the late 1980s, Indonesia had become a major producer (and exporter) of urea fertilizer, capable of an annual production of over 5 million tonnes of urea. It was also producing over I million tonnes of triple super-phosphate and more than 650 million tonnes of ammonium sulphate. Irrigation systems, particularly in Java, had undergone substantial rehabilitation and improvement. The country also had, as a result of its rice programme, a network of over 5,000 rural banks that had begun by providing subsidized credit for rice; and by the 1980s, they had developed into institutions offering both general rural credit and attractive rural savings plans. Furthermore, most indicators showed a marked improvement in the rural standard of living.
Thus, as Indonesia entered the new decade of the 1990s, it could claim enormous success in achieving its goal of increasing rice production, in modernizing its agricultural capacities, and in improving the welfare of its population. The problems faced appeared to be those created by success. Yet at no time in the previous decades was there a greater concern with issues of sustainability and a greater uneasiness about the possibilities of future production increase. There were several reasons for this situation which need to be considered in relation to past practices.
The problems of expanding production in Indonesia
Each year from 1967 to 1989-with two exceptions, 1972 and 1975-saw an increase both in production and yields in Indonesia. In 1990, however, production steadied and then, in the dry El Niņo year of 1991, actually declined slightly. During two previous El Niņo episodes in the 1980s, there were continued increases in both production and yield, despite the lack of rain. The question whether Indonesia was reaching some son of yield plateau was asked.
Indonesia's goal in its rice intensification programme was to achieve maximum production; but initially she felt it necessary to attain self-sufficiency. Once selfsufficiency was reached, production increases were still needed, not just to match population growth but also to meet the growing demand for rice from an increasingly prosperous population. A strategy of maximum production meant laying stress on the most favourable growing areas with ready access to high inputs. Begun in 1987, the Indonesian government's programme (known as Supra Insus) was a high]y regimented plan directed towards this end. (3) The favoured areas in Indonesia were the four provinces of Java and a number of other important rice-growing provinces: Bali, South Sulawesi, North Sumatra, West Sumatra, South Sumatra, West Nusa Tenggara and South Kalimantan. Thus, 11 out of 27 provinces were the mainstays of Indonesian rice production.
The island of Java has consistently contributed about 63 per cent to total production since the beginning of Indonesia's rice intensification programme (J. J. Fox, 1992). The key to this increased output on Java was not the opening up of new cultivation areas but the more effective utilization of existing land- higher cropping intensity and improved yields from established irrigated rice fields (sawah). Elsewhere in the country, production was less dependent on increasing intensity and yields, and more on the opening up of new growing areas. Yield figures highlight this difference between Java and the other islands. Java's average yields rose from 2.58 tonnes per hectare in 1968 to 4.98 tonnes per hectare by 1989, whereas yields elsewhere only rose from 2.15 tonnes per hectare in 1968 to 3.52 tonnes per hectare by 1989 (J. J. Fox, 1991: 80). Thus, during the period of its intensification programme, Indonesia experienced a widening gap in yields between Java and the rest of the country. Moving into the 1990s, Indonesia faced two questions: whether Java's yields would continue to increase, and whether the other islands would begin to close the gap with Java in yields, following its pattern of intensification.
The context for these questions was set by developments in Java itself as rapid industrialization, the growth of urban centres, and the demand for more residential and industrial land exerted pressure to take prime agricultural land out of production for other uses. Although the government took measures to shift more sugar-cane planting from lowland sawah to upland fields in the late 1980s and, in the longer term, to shift its cultivation from Java to Sumatra, other high-value crops began to be planted more extensively on sawah, thus further reducing the availability of this land for the production of rice.
Assessing the yield potential of the green revolution
Chang notes that the crop yield per season of 11 tonnes per hectare was set by two of the earliest HYVs, IR8 and IR24. Later varieties have not delivered similar yields, although some, with shorter maturing times, have surpassed these levels calculated on a yield per hectare per day basis. A yield of 11 tonnes per hectare, even though under ideal trial conditions, is significantly higher than the average of either 3, 4 or even 5 tonnes per hectare reported for other parts of South-East Asia. The evidence would seem to indicate that there are no inherent biotic constraints to prevent rice yields from continuing to increase from their current levels. Although there still remain various uncertainties about the use of hybrid technology for rice, especially in the Tropics, development of such hybrid strains may also offer the possibility of further yield increases in the future.
If there is one lesson that farmers in Indonesia have learned-if indeed they ever had to learn what they already knew-it is that quality seed is crucial for high production. Despite serious set-backs shortly after the HYVs were introduced, these varieties ultimately proved their worth to farmers in those growing environments where they were indeed suitable. Overzealous government initiatives to foist these varieties into environments where they were clearly unsuited only undermined the credibility of government efforts.
It takes a number of successive generations to breed a new variety of rice in sufficient quantities to be distributed to farmers for planting in their fields. Early in its intensification programme, in addition to establishing government nurseries, Indonesia introduced a colour-coded certification system to identify for farmers the quality and 'generation' of the seed that was being sold. Generally-or at least, ideally-farmers, particularly those in Java, would then use this seed and its progeny for three successive seasons before obtaining 'new generation' seed from a nursery source. In pursuit of higher yields, the government endeavoured, in the late 1980s, to provide and to foster the planting of 'first-generation' seed for each planting season. By this time, however, the range of HYVs available to farmers was limited to a tiny fraction of the total number that had been bred for introduction. The biotic diversity of rice in Indonesia, as indeed throughout South-East Asia, had been greatly diminished.
