Previous - Next
This is the old United Nations University website. Visit the new site at http://unu.edu
Upland agriculture, 1950-1990: Logging, roads, markets and cash
New Elements in the Pattern of Change
The discussion now turns more specifically to developments over the last 40 years. Upland agricultural systems in the region have continued to be challenged by rapidly growing populations (see Concepcion, Chapter 2). In addition, forests which were not subject to clearing by earlier agricultural expansion are now being logged at a high rate, and logging has lately extended into quite hilly regions. Roads for loggers also provide access for farmers and settlers, both to the logged-over land itself and to distant markets. Demands for cash (to pay for consumer goods, children's education) and for agricultural inputs which have become a necessity in the face of continuing declines in yields, combined with better access to markets, have led to a rapid increase in the cultivation of high-value crops for sale. These crops, such as vegetables, frequently extract a greater cost in terms of erosion and loss of soil nutrients than the 'traditional' upland crops of rice, maize, cassava and other grains and tubers or, in Thailand, opium.
In the lowlands, wet-rice land is reaching its ultimate expansion through the conversion of mangrove swamps to rice land, and the use of tidal irrigation in coastal estuaries. Chang (Chapter 9) makes the point that, following the Green Revolution, the further expansion in area and in yields of irrigated rice is now strictly limited, and that existing lowland systems are under increasing stress from the high costs of maintenance and inputs. Elsewhere, Chang (1987) argues the case for a direct link between increases in rice production in South-East Asia and increases in her population. He goes on to observe that while the growth in production which followed the adoption of the Green Revolution high-yielding varieties (HYVs) of rice has reached a plateau, the populations of Asian countries continue to grow. The clear implication is that, if further production increases from wet-rice areas are limited, the burden of food supply for this rapidly increasing population will fall on upland cultivation systems.
In a study on Indonesia, J. J. Fox (1991: 81-3) comes to similar conclusions about rice production. Although Indonesian rice production (unhusked dry-rice paddy) increased from 15 to 45 million tonnes between 1968 and 1988, demand is estimated to go on increasing at between 2.7 and 3.0 per cent per year, which will require a 34-39 per cent increase in rice production by the year 2000; this is the same increase as occurred between 1968 and 1988. Fox is cautiously optimistic that these challenges can be met, but notes in passing that in Java, some of the better, irrigated land, where the production increases could be expected to occur, is being taken out of rice production in favour of higher-earning crops.
The Link with Timber Production
During 1850 1950, governments in the region attempted to create and protect forest reserves, with varying degrees of determination and success. Over the 1970s and 1980s, however, the general release of control from government to private operators, the logging of forests with modern equipment, such as bulldozers, tractors, timber jinkers and chain saws, and the penetration of roads into huge tracts of previously 'unroaded' land have brought about a new era of forest clearing and agricultural expansion. The deforestation issue has been raised in several previous chapters, particularly by Brookfield (Chapter 1) and Potter (Chapter 5), and the important aspect of fire is discussed by Wirawan (Chapter 11). The link with land degradation has been touched on, but needs to be examined more specifically and considered as a separate issue.
Soil erosion and the degradation of land need not be inevitable outcomes of forest clearing, especially where rapid woody successions colonize areas cleared of forest. In Kalimantan in the late 1980s, for example, the rates of deforestation caused by logging followed by clearing for agriculture and settlement were estimated to be between 600 000 and 1.2 million hectares per year (Potter, 1991). The opened, 'roaded', logged-over and damaged forest is very much easier to enter and clear than closed, undisturbed rain forest. Movement into the areas of large numbers of new settlers, both governmentsponsored and spontaneous migrants, has impacted heavily on the logged-over land.
A similar pattern is found in the Uthai Thani district of Thailand where by 1985 forested land was less than 41 per cent of the total land area and deforestation was continuing. Land settlement by outsiders was 'facilitated in no small part by the clearance of large areas of forest under concession to the state-run Thai Plywood Company.... Construction of roads has also been crucial' (Hirsch, 1990: 56). In northern Thailand, strategic roads built to assist defence against insurgents first brought loggers, and then cultivators, into large areas that, in consequence, rapidly lost their forest cover.
The average rate of deforestation over 1976-80 for the whole of Indonesia was estimated at 550 000 hectares per year, compared to 325 000 hectares in Thailand, 230 000 hectares in Malaysia, 120 000 hectares in Laos and 100 000 hectares in the Philippines (Ooi, 1987: 7). The main causes are stated to be shifting cultivation, unorganized and spontaneous encroachment on forest land, squatting, migration of landless and displaced lowlanders into upland forested areas, refugee encroachments and government-sponsored land settlement, transmigration in Indonesia and state-sponsored agricultural development schemes in Malaysia in particular.
Linkage with External Areas
Improvements in communications, particularly roads, has also been an important influence in non-logging areas. The Gunung Kidul district south of Yogyakarta, in Java, was already largely deforested by 1904 when remaining teak forests were taken over and managed by the colonial forest service. This limestone area was connected to the outside world in the 1930s by a dry-weather bullock track and, although the roads were later improved, in 1949 it was still possible for Indonesian nationalist forces to take shelter in the area after they had been driven from Yogyakarta by the Dutch. By 1989, an asphalt road ran through the whole area and many hamlets were connected to this road by allweather, gravel roads. Minibuses travelled frequently along the main road, and villageowned vehicles were reasonably common. Villagers could travel to Yogyakarta and back in a day and to the district centre and market place in less than an hour (Nibbering, 1991b: 71-2). They now have much-improved access to off-farm employment, the city and regional markets, cheaper farm inputs, information and primary and secondary education for their children.
The most dramatic change in conditions of access has been in the populous central highlands of Papua New Guinea, which until the 1950s had no road to the coast. Now a network of several thousand kilometres of roads links all densely populated areas to the coast at Lae and Madang, bearing a heavy traffic of buses, minibuses and trucks. In more accessible areas, there have been major changes in land use, incorporating a range of cash crops. Coffee from this region, only a pioneer crop in the 1950s, is now a leading export. With the development of both urban and rural markets, trade in foodstuffs has become important. There is extensive movement of people, leading to an actual reduction in the population of some marginal areas, despite continued high rates of natural increase. The conditions of production and consumption, even in areas distant from the roads, have been transformed.
Consequences of Changes in Crops
A major trend, which distinguishes the 1970s and 1980s from the previous 150 years, is the partial switch from subsistence staples to high-value cash crops, particularly vegetables. As shall be described below, declining soil fertility and subsequent declining yields have eventually forced farmers to purchase off-farm fertility-maintenance inputs-inorganic fertilizers in particular-to maintain yields. This alone requires farmers to have access to cash. With better access to markets, burgeoning demand for vegetables in towns and cities, a greater awareness among upland parents of the value of education for their children, and an increased desire for manufactured consumer goods, the switch to high-value crops and the partial substitution of subsistence staples for purchased foods is an economically rational and socially desirable action for farmers. However, it is not always ecologically sensible.
First, despite high and increasing application rates of inorganic fertilizers, declining soil fertility remains a feature of many areas of vegetable growing; it is reported, for example, in West Java and the Kundasang area of Sabah (Imam Ali, personal communication; Hardjono, 1991). In addition, many farmers find that, because of poor management or the quality of their land, they cannot afford the costs of the inputs and, as a consequence, they enter a downward spiral of decreasing yields and an increasing inability to purchase the required fertilizers to improve yields.
Secondly, soil erosion is a serious problem in many vegetable-growing areas. Annual cropping increases the exposure of soils to rainfall, and farmers are reluctant to terrace because of the costs and the technical difficulty. At Kundasang, migrant ethnic Chinese farmers view vegetable growing as a means of raising capital for onward investment in non-farming activities, and they appear to have few concerns for the long-term sustainability of the enterprise. Topsoil is sometimes deliberately scraped off plots to reduce weed infestation, and crops are maintained in the subsoil by inorganic fertilizers.
The unsafe use of pesticides (many of which are banned elsewhere in the world), often used at rates well above those recommended, is common. Indigenous Kadazan landowners in the hills surrounding Kundasang are adopting some of these practices, in particular the use of spray irrigation and fertilizers and pesticides. However, most cannot achieve the production levels of the immigrants, nor are they as successful at judging the market and producing the highest-value crops (Imam Ali, personal communication). They are thereby putting the land, in which they have a long-term interest, at risk.
Where farmers cannot afford the costs of off-farm inputs, they may eventually be forced out of farming altogether by indebtedness. Poorer farmers in the Uthai Thani district in Thailand, confronted with declining yields, increasing soil compaction, weeds and pests, seek labouring opportunities from wealthier farmers in order to buy fertilizers, pesticides and tractor time. Some farmers are forced to lease their land to wealthier villagers and seek off-farm employment in towns, in order to pay off debts, and because they lack the capital to farm the land themselves (Hirsch, 1990: 74).
The outcome of increased commercialization and monetization of upland farming economies has not everywhere been negative. In his detailed study of agricultural change and environmental degradation in the Gunung Kidul district in Java, Nibbering (1991b) found that increasing use of off-farm inputs and improved access to markets, greater offfarm wage-earning opportunities, and greater political security have brought about improved social, economic and environmental conditions in this district. He argues that, during the 1950s and 1960s as pressure on land increased, farmers' returns to labour declined steadily because fields could not be cultivated further without the costly construction of terraces, and without working increasingly marginal land which had to be brought into production at high cost. In the 1970s, when farmers gained access to fertilizers and pesticides and changed cropping techniques, this situation changed. As returns to labour increased, farmers began to invest more in ecologically sound practices, particularly in tree planting, which supplements farm incomes through the sale of timber, but which also greatly improves the local agricultural environment and increases the supply of fuelwood.
The environmental consequences of upland agricultural expansion: Sustainability and unsustainability
The evidence on sustainability of these developing upland agricultural systems is contradictory. In some areas, improved access and incorporation into the market economy have apparently brought about the rehabilitation of previously degraded areas, while in other locations, the 'mining' of forest and soil resources has continued unabated. Most researchers in upland ecology concentrate on soil erosion in already cleared land as an even more serious ecological threat than deforestation. Shifting cultivation has long been viewed, particularly from within the region, as the major problem. But it must be recognized that in many of the areas discussed-cleared and settled by or before the 1930s and 1940s-shifting cultivation has long since given way to more permanent agriculture. It is continuing pressure on this land which is giving rise to concern.
The great expansion of agriculture from the lowlands into the uplands, which occurred throughout South-East Asia from the 1800s onwards, resulted in widespread degradation of the land. Across the whole region, the outcomes were similar. The conversion of forest to grassland, soil erosion and declining soil fertility led to the move away from shifting cultivation in the direction of permanent cultivation. The ability of watersheds to retain water from rainfall has declined, and rapid runoff has resulted in severe flooding damaging lowland wet-rice systems.
In almost every case where the sequence is known, or can be reconstructed, the earliest upland farmers practiced forms of shifting cultivation in which soil fertility was maintained by a natural regrowth fallow, and soil erosion was minimized because most of the land was protected by a tree and scrub cover. As more migrants arrived, and populations continued to expand, all cultivable and accessible land was occupied, fallow lengths decreased, and large, previously forested areas were converted into scrub and grassland. This situation had been reached in Java by 1920 (Booth, 1988: 100). Cropping of the same tract of land was more frequent, and complete tillage became common, together with the application of animal manure to maintain soil fertility (Palte, 1989). In the Philippines, even in areas where shifting cultivation had been practiced for hundreds of years, fallow lengths declined and there was a 'trend towards a cultivation cycle based on the annual cropping calendar of the dominant subsistence crop' (Cruz, 1986b).
In a number of areas in the highlands of Papua New Guinea, permanent sweet potato cropping systems based on complete tillage and composting have been developed since the conversion of forest to grassland. At their most elaborate, these systems involve throwing the soil into large mounds within which plant matter imported from nearby fallow, as well as from the garden site, is composted. Under this system, and others involving complete tillage of the surface, some significant areas are now cultivated permanently, with only short fallow gaps of, at most, a few months between each crop.
These technologies were developed on flat-to-rolling land and on alluvial plains, where they appear sustainable; in the composted-mound system, a high rate of soil formation more than offsets any loss from erosion. However, these intensive systems have been extended on to the hill slopes and valley sides as populations have increased. Here, they give rise to rapid loss of topsoil and accompanying rapid declines in soil fertility. On volcanic-ash soils on steep slopes in the Tari basin, sweet potato yields fell from 8 to 2 tonnes per hectare in less than 40 years, whereas on alluvial plains and recent colluvial soils, yields can be maintained at 12 tonnes per hectare for over 50 years (Wood, 1984). The outcome is the abandonment of fields on the slopes, the establishment of a tall grass disclimax maintained by irregular burning, and the continued clearing of forests higher on the valley sides.
Consequences can be measured in terms of human welfare. Everyone participates in a social system in which status and security depend on inflationary ceremonial exchanges (Allen and Crittenden, 1987). In order to keep up with their competitive neighbours on the better land, men on less capable soil must push their land or themselves harder-their women in particular. The effect is apparent in the author's analysis of the birth weights of more than 2,900 children born in the Tari basin. There is a statistically significant 150-gram difference between the weight of children born in the best area and those born on the poorer land.
It is difficult to quantify the meaning of these changes in terms of soil erosion, or soilfertility decline, across the whole of South-East Asia. It must be assumed that, on land which was not adequately terraced, severe soil erosion occurred when permanent cultivation systems first became established, in the absence of terracing or other soilfertility maintenance techniques. But, as Nibbering (1991b) explains for Indonesia, the conditions of soil type, slope, rainfall intensity, vegetation cover, and cropping and cultivation practices vary so much from place to place that soil losses may differ by orders of magnitude between areas in relatively close proximity. He concludes, very sensibly, that it is probably an unrewarding task to seek to establish erosion rates, and even if it were possible to do so within acceptable levels of confidence, little would have been achieved towards solving the problem of the erosion itself.
Migration and Resettlement
Throughout the region, numerous attempts have been made to address the problems created by the expansion of upland agriculture in the 1800s and its modern manifestations. Perhaps the best known attempt to reduce the pressure on land in Java-the Indonesian transmigration project which drew many of its recruits from upland areas-is now widely seen as an environmental threat in itself, and as a major contributor to the deforestation taking place in the outer islands. Transmigration settlers are poorly cared for, relative to, for example, Malaysian FELDA settlers; in Kalimantan, they have been required to be largely self-sufficient in food, a factor which has led them into annual cropping on land which has been placed in a capability class for tree crops only.
A good deal of resettlement from upland areas has been of a more spontaneous kind, and it has generally been to lowland areas, especially to the towns and cities. Although a proportion of those who have moved from the highlands of Papua New Guinea now live in planned land-development areas where they cultivate tree crops, a larger number has settled in and around the urban areas. The same has happened in the Philippines and in Java. Within the uplands, there is migration from more remote areas to regions where cash cropping is more rewarding, and where off-farm employment is readily available. There is little data on the basis of which these movements might be quantified, but there can be no doubt concerning their substantial volume.
Social Forestry
Programmes which directly address the problems of deforestation, and the rehabilitation of degraded land, are the social-forestry schemes which are found in a number of SouthEast Asian countries. In the Philippines, social-forestry projects have been concerned with environmental issues, but they are also explicitly designed to improve living conditions, provide employment, augment incomes and make rural communities more self-reliant (Aguilar, 1986). In contrast to previous attempts to control upland forest clearing, which relied unsuccessfully on exclusion of people from the forests, coercion and punishment, the newer projects allow people to remain in the areas under a form of 'managed occupancy'.
Although it is difficult to generalize about the success of these projects because the outcomes have been so variable, it is apparent that, after enthusiastic beginnings, reforestation and terracing targets have rarely been met. The major difficulties facing such projects appear to be the quality of government extension agents, their rapid turnover and a continuing tendency for them to pressure farmers into participation. In one scheme in the Philippines, increasing coercion resulted in extensive burning after a number of years of declining frequency of forest fires (Aguilar, 1986).
Reforestation, or regreening, has a long history in Indonesia. The colonial government recommended planting leucaena on deforested hillsides as a precursor to terracing (Nibbering, 1991b: 169). The independent Indonesian government's first National Regreening Week was held in 1959. Soemarwoto (1991) observes that the outcome of regreening programmes has been variable, but the rate of deforestation is clearly still exceeding the rate of regreening. He also observes that tree planting alone does not necessarily reduce erosion and may even interfere with soil-water conditions.
This chapter has attempted the impossible: to review a complex topic across a highly varied region without becoming overly superficial. Nevertheless, it is clear that a major transformation of upland environments has occurred in South-East Asia since around 1850 and that it continues in the early 1990s at a rapid and unsustainable pace. The transformation has arisen in important measure from large population increases, which have been associated with increased food production from lowland wet-rice systems, and improved health and living conditions. lt has further involved major movements of people from the lowlands into the uplands, the clearing of forests from a vast area of land and the intensification of agricultural systems. Later, improvement in transport and communication networks. access to markets and the commercialization of agriculture, including the increased use of purchased fertilizers and pesticides, have generated new conditions that impact upon different areas in different ways.
A December 1990 seminar on technologies for sustainable agriculture in upland SouthEast Asia included contributions on economics, land tenure, farming systems, agronomy and soil management, but no clearly discernible overall solution was offered (Blair and Lefroy, 1991). Following a detailed study of deforestation in the Philippines, Kummer (1990b) concludes that the history of forest use in individual countries is important, and population increase is not a critical explanation of recent deforestation, but elite control over forest land and corruption is extremely important, at least in the Philippines. Whatever the causes, the solutions to deforestation are not amenable to technical solutions. As Kummer argues, the fundamental issue is who has the right to use the forest resource.
Soemarwoto (1991), who has long pondered these problems, has put forward a proposition to reverse some of these consequences. In general terms, he argues that past solutions, which have had at their core land rehabilitation and reforestation, have demonstrably failed. Therefore, solutions are required which give priority to the needs of people and leave the environment to rehabilitate itself, once the people can be persuaded to leave it alone for a while. Soemarwoto argues that, in order to do this, the lives of the rural poor must be improved. Their agricultural systems will have to be intensified even further in order to reduce the amount of land necessary for support. Off-farm employment will have to be increased, training for off-famm work improved, a reduction in the rate of population increased through family planning, and improved and fairer marketing schemes instituted. Of the cases reviewed in this chapter, Nibbering's offers the greatest support to Soemarwoto's proposition. The people of the Gunung Kidul district began their own land rehabilitation when their social, economic and political conditions improved.
What remains to be done therefore is to convince political leaders, bureaucrats and South-East Asia's burgeoning middle classes that their futures are inextricably linked to the lives of rural people and to the upland environments in which these people live.
Pitfalls of the intensification debate
KANOK RERKASEM
THE argument that traditional agricultural systems, such as shifting cultivation, are energy efficient needs to be handled with caution. The available data are highly variable, depending on attributes. In New Guinea, for example, Rappaport (1971) reported an energy output/input ratio for Tsembaga slash-and-bum agriculture of about 16: 1, and the highest value under the most favourable conditions of 20: 1. However, recalculation of the same data, with inclusion of the energy used in burning the field, drops the ratio to 0.11: 1, thus indicating that shifting agriculture is just as dependent on an energy subsidy to achieve high production with low human labour input as is modem mechanized agriculture (Rambo, 1981).
Moreover, high biodiversity (characteristic of shifting agriculture) is not necessarily less vulnerable to changes in physical conditions (May, 1974). Time lags in a system with patchy distribution of pests can contribute to the destabilization of the ecosystem (Conway, 1983).
The object of intensification is to increase the productivity of an agricultural system. This may take several forms, including application of herbicides and pesticides, addition of fertilizers, introduction of new varieties, mechanical tillage, continuous cropping or an alternation of crops (Conway, 1983; Trenbath, Conway and Craig, 1990). However, the consequences of intensification may also take different forms. The intensified system may not behave as anticipated. Pesticide resistance is an example. Other internal and external elements and processes may also be affected in undesirable ways. Intensification should be considered in terms of its effects on system properties. These include productivity, sustainability. measures of social equity, autonomy and solidarity (see, for example, Conway, 1986; Conway and Barbier, 1990; Marten and Rambo, 1988). The trade-off between effects on these properties may offer guidelines on management of intensified systems, of assistance in the problems now faced by man.
Immediate needs in upland Java
JOAN HARDJONO
ALLEN'S Chapter 10 is closer to the real problems of upland farmers in South-East Asia than was the somewhat detached overview of intensification he presented in Yogyakarta. He is certainly right when he states that new approaches and attitudes are needed, but the problem lies in waiting for these to be developed. As Indonesians say, nasi sudah jadi bubur (the rice is already overcooked). A great deal of environmental damage has already been done. The question concerns steps that can be taken now. This comment relates specifically to Java, and especially to West Java, where the author has undertaken a study of upland farming problems (Hardjono, 1991).
Java in the early 1990s has more than 100 million people. At the same time, it also has some very serious environmental problems, most of which are related to land use. With steady population growth, wet-rice cultivation has been unable to support the increase in numbers, despite ever-greater intensification. The result has been the extensification of agriculture into upland areas. It is here, in the uplands, that the question of sustainability is perhaps at its most pressing.
Writing of Papua New Guinea, Allen mentions problems created by socially determined production surpluses. Many of Java's upland farmers have incomes that place them not very tar beyond the fulfilment of physiologically minimum requirements. In endeavouring to meet these needs, they have unfortunately resorted to land-management practices that, on the whole, have negative environmental consequences. Most farmers in upland Java cultivate dry arable crops, though in some areas agriculture takes the form of relatively intensive vegetable growing which, at least in the short run, is more profitable. Eventually, however, both forms of upland agriculture lead to land degradation when they are practiced on steep slopes. This is reflected in declining soil fertility and in erosion, which has implications for those who live in the lowlands.
Farmers in Java are no longer able to move to new areas, yet their soil-management techniques are inadequate to ensure sustainable use of the land. A shortage of land prevents them from practicing fallowing, and the majority cannot afford to buy inputs to maintain productivity as well as their own basic income levels. This circle of poverty is very difficult to break. yet it must be done in the interests of the environment.
Population reduction in the uplands is one obvious course. Those who are cautiously optimistic about future trends in the Indonesian economy see the movement of labour out of low-income occupations in the uplands into nonagricultural employment as the answer to pressure on land resources. Unfortunately, given present trends in non-agricultural labour absorption, it will be many years before this transfer takes place on a sufficient scale to have meaningful effect on upland populations. There is very little opportunity for non-farm work in the uplands themselves, while industrial expansion in and around major urban centres tends increasingly to be capital-intensive, rather than labour-intensive, in nature.
Similarly, government-sponsored transmigration from Java, while it has provided land for large numbers of upland farm families in the 1970s and 1980s, cannot be expected to continue as a means of population reduction. The many reasons include environmental deterioration in the dry-farming regions outside Java. Obviously, only family planning can reduce population pressure in the uplands as in other areas. Its importance cannot be sufficiently stressed, but it will be many decades before the effects can be felt. In the meantime, the starting point towards more sustainable upland agriculture must lie in government attitudes and policies.
The word 'programmes' has been deliberately avoided because there have been many, and there is little to show for them. Moreover, the government has attempted to find one system or model that can be applied everywhere, not only in upland Java but in dryfarming regions throughout Indonesia. The Bimas programme, developed as the vehicle for Indonesia's rice intensification in the 1970s, was essentially a one-package programme for all growers of irrigated rice. Although there are many wet-rice ecosystems, basically one approach was offered to all. In the uplands, the differences from region to region are far greater. Yet dry-crop programmes, such as that for maize improvement, and later the attempts to develop soya bean cultivation, have tended to ignore biophysical variations. They have adopted an approach under which the same extension services and inputs are offered to all participants. Furthermore, they have usually been stimulated by market considerations, rather than by the suitability of the land to the crops involved. Such an approach has achieved, predictably, only limited success.
Programmes involving reforestation and regreening have likewise tended to be unsuccessful; they have not taken into account the immediate needs of individual farmers. With access to only a smallholding, the farmer's main concern is with food production for his own household. Attempts to develop cash crops, such as tea, have run into the same constraint. Given the limited land available per household, the returns are too low to persuade the farmer to give up food-crop production.
Official attitudes therefore need to be completely readjusted. This readjustment would involve an end to present priorities in favour of irrigated rice farming rather than upland cultivation. Application of single solutions intended to deal with low incomes and environmental problems everywhere at the same time should cease. Solutions that are location-specific in their approach must be sought. The Indonesian government has had the political will to achieve self-sufficiency in rice production. Given the same political commitment, there could be similar achievement in regard to upland agriculture, where environmental damage makes the need even more urgent.
Rerkasem's comment and, to some extent, the one by Hardjono relate to Allen's presentation in Yogyakarta, which he did not intend for this book. In it, Allen began with the observation that upland farmers, like lowland farmers, have usually intensified their cultivation through time, and he explained this as a consequence of 'increased population, competitive behaviour, fear and loathing of "other", and a desire for status', to which desire for material goods is now added. Following and elaborating a line of analysis developed by Odum ( 1975), Allen argued that this led not only to the degradation that is widely observed and reported but also to a decline in the energy efficiency of farming. This latter assertion led to a productive exchange of views, part of which is reported above.
The discussion focused on aspects of the energy-subsidy question in relation to intensification. Rerkasem's argument was supported by Soemarwoto, who added that whether or not it is appropriate to take account of the energy used in burning the forest in shifting cultivation, it is surely appropriate to take account of the solar energy consumed in producing the fallow. If this energy is included, then shifting cultivation uses a multiple of the solar energy used in continuous cultivation, dependent on the number of months and years of fallow. This has practical as well as merely theoretical significance, since a large part of the defence of shifting cultivation as a system rests on its allegedly high ratio of energy output to energy input, taking account only of labour and imported energy on the input side. If imported energy from fossil fuels (which were produced by solar energy) is included, then current solar energy inputs should also be taken into our calculations.
James Fox then pointed out that if imported energy is fully accounted for in evaluating the Green Revolution, then its successes are not so momentous. having coincided with, and relied on, a massive petroleum boom in Indonesia. Nitrogen is the fuel of the Green Revolution. and some of this energy also goes to the uplands. Soegiarto supported this view, and he questioned the sustainability of present forms of intensive agriculture. He added that much land is being taken out of agricultural use, and that there is need for a better allocation of land, both upland and lowland, to optimal uses. Chang then asked how man can manage in the future. Neither modern European nor Asian agriculture is sustainable when all the side-effects of chemicalization, including increased presence of nitrates in drinking water, are taken into account.
Introduction
Losses and impairments due to fire
Factors promoting the spread of fire
Reducing the incidence of large-scale fire
NENGAH WIRAWAN
FIRE has been a very important tool for mankind since time immemorial. Although some anthropologists believe that the use of fire began with Homo erectus more than a million years ago, solid evidence for the controlled use of fire by hominids can only be traced as far back as the ancestors of the Neanderthals, the early form of Homo sapiens, who lived in Europe about 200,000 years ago (Benditt, 1989).
Aside from its domestic use, fire has also become an integral, natural part of many ecosystems (Bormann and Likens, 1979; Mueller-Dombois, 1981; Newsome, 1985; Spurr and Barnes, 1980). While natural and anthropogenic fires are considered to have changed the Holocene vegetation of North America (Russell, 1983), fire is certainly not new to the wet tropical rain forests. Studies in the headwaters of the Amazon (Sanford et al., 1985) have found abundant evidence of charcoal up to 6,000 years old in the rain forest soils of the Upper Rio Negro in Venezuela. Charcoal deposits about 500 years old were also found in the colluvial soils in several locations in the undisturbed and unburned lowland rain forest in the Kutai National Park in East Kalimantan, Indonesia (Shimokawa, 1988). A report by Endert (1927)-which described extensive fires in the coastal area along the Bengalon River, East Kalimantan in 1915 following a long drought period in 1914-and further findings (Goldammer and Seibert, 1989) of charcoal deposits of up to 17,500 years old from neighbouring locations suggest that fires have repeatedly broken out in the lowland dipterocarp rain forest of East Kalimantan since the late Pleistocene.
During the early 1980s to early 1990s, and especially in the long drought period of 1982-3, extensive fires swept through natural forests in many parts of the world. While the ravaging fires in Australia and Europe received wide media coverage, I million hectares of larch forests were wiped out on the Tahsing-an-ling mountain in northeastern China (H. Tagawa, personal communication).
Similar fires also occurred in Sumatra, Borneo and a number of other places in SouthEast Asia, with the worst and most extensive ones in Sabah and East Kalimantan. In Sabah, fire burned at least 950000 hectares of logged and unlogged dryland forests (Beaman et al., 1985). In East Kalimantan, it burned 2 717 000 hectares of swamp and dryland forests or 3 193 000 hectares including settlement and agricultural land areas (Schindele, Thoma and Panzer, 1989). The East Kalimantan areas were again struck by drought and fire in 1987, though less intense and also smaller in scale. There were further outbreaks, larger than in 1987, in 1991.
Furthermore, in the United States during 1990 alone, 10 000 of the 300 000-hectare Yosemite National Park (USA Today, 18 August 1990) and nearly half of the 880000 hectares of the Yellowstone National Park in the Rocky Mountains (Cross, 1990) was burned out. These occurrences highlight the importance of fire as a factor that affects the world's resources and environments.
This chapter is an attempt to assess the role of fire in the loss or impairment of the natural resources and environments of South-East Asia. Ignoring domestic fires, and those that often happen in settlement or industrial areas, only wildfires that occur in wildland areas, including swidden agricultural fields, grassland, scrub and forest areas, are discussed. The great forest blaze in East Kalimantan in 1982-3 will be used as a reference.