|South-East Asia's Environmental Future: The Search for Sustainability (UNU, 1993, 422 pages)|
|Part I - The driving forces of change|
|1. The dimensions of environmental change and management in the south-east Asian region|
The trends reviewed cover the limited span of the last 1(}15 years, and mainly used are some very imperfect international data sources which conceal as much as they reveal. The principal sources used are those collated by the Food and Agriculture Organization (FAO) of the United Nations, supplemented by a few others (various issues of the FAO Production Yearbook, FAO Trade Yearhook, Far East and Australasia Yearbook and World Development Report). The notorious inadequacy of these sources can be offset to some degree by use of national data, but this is done only here and there. Since the immediately following chapters deal with population, urbanization and industrialization, it is appropriate here to review changes in land use and agricultural production since the late 1970s.
Notwithstanding the rapid growth of cities and industry, and the large but inadequately quantified expansion of off-farm employment in rural areas, agriculture continues to employ the most people in all the larger countries of the region and will continue to do so into the twenty-first century. The agricultural share of gross domestic product (GDP) is, however, declining in all regional countries, but at very different rates (Table 1.1). The decline is particularly rapid in Thailand where mineral and industrial production increased by over 60 per cent between 1985 and 1989. In Malaysia, the industrial sector gained 20 points in its percentage share of GDP between 1960 and 1988, and is now well ahead of agriculture; industrial exports now exceed agricultural exports. However, in spite of a remarkable growth in the electrical and electronics industries such that they have become principal generators of national exports (Bank Negara Malaysia, 1991), a major part of Malaysian manufacturing continues to be agro-based, wood-based or petroleum-based (Osman Rani and Haflah Piei, 1990). This is even more true of Indonesia (Hill, 1990). With the exception of Singapore, the economic health of the South-East Asian countries continues even in the early 1990s to depend primarily on production from their own resource.s-minerals, timber and agriculture; despite the industrial deepening now taking place, resource use remains basic.
TABLE 1.1 Share of Agriculture in GDP in Selected Countries in South-East Asia, 1978-1988 (per cent)
Sources: Far East and Australasia "Annual).
a Figures from Far Eastern Economic Review Year hook (various years).
n.a. = Not available.
Aggregated over the whole region, the changes in land use over this short period have not therefore been of a dramatic order (Figure 1.2), but none the less the proportion of the total regional land area stated to be under forest and woodland shrank from 61 per cent in 1973 to 55 per cent in 1988, with a trend suggesting that it may fall below 50 per cent by, or soon after, the year 2000. The fact of change is certainly much greater than this, for official data often maintain the area of forest without adjustment for several years, until new survey data become available. At regional level, there is little in the way of reliable data on the areas of forest that have been logged, though there are good surveys in Peninsular Malaysia, and inventory data will soon come to hand from much of Indonesia and some other areas. Much of the forest cover given as such in the national data has, from the air at low altitudes, the appearance of the back of a mangy dog. Even data based on satellite imagery do not distinguish at all well between undisturbed primary forest, logged or 'managed' forest and secondary forest areas, some of which have multiuse functions, and one study has bluntly suggested that such data are full of 'ambiguities and impossibilities' (Blasco and Achard, 1990). Moreover, the infrequency of cloud-free passes over the low-latitude Asian Tropics means that years of imagery may be required to produce one static map of a given large area such as Sumatra, Papua New Guinea or Borneo. Until, or unless, these problems can be resolved and the results compared through time, for purposes of regional generalization one has either to rely (with reservations) on data such as these or else use estimates of deforestation rates that on a global scale range from 70 000 to 200 000 square kilometres per year (FAO/UNEP, 1981; Myers, 1986; Train, 1988;Tyler, 1990). It is more than probable that much less than 33 per cent of the regional forest area-that is less than 15 or 16 per cent of the whole regional land area-still remains largely free from substantial ongoing human interference; ongoing because the amount of truly undisturbed primary forest is probably considerably smaller than is often stated in the modern literature.
Unlike what has happened in the American Tropics, however, there is no expansion in the very small area of permanent pasture, consistent at 3.3 per cent of the whole. Land under crop, both annual and permanent, has expanded quite slowly from 14 to 16 per cent, though the official data seem to lag behind observable reality. More of the statistically recognized expansion has been in a rag-bag category termed 'other land use' now occupying 26 per cent. Of this category, a high proportion must certainly be in land taken up for human settlement, but some degraded land and national parks must also be included. Moreover, it is clear that land is classified differently in some countries. Over 90 per cent of urbanized Singapore is in this category, and this is credible. Less explicably, however, 50 per cent of Vietnam and Brunei, and about 33 per cent of Thailand, the Philippines and Myanmar are under 'other land use'.
Within the region, there is sharp differentiation in trends, and even within a country. The slow change in the total forest area is seen to be dominated by the reported constancy of forest cover in Papua New Guinea, Myanmar and Cambodia. In Papua New Guinea, Cambodia, Indonesia, Malaysia and Laos, more than 50 per cent of the land was recorded as under forest in 1988. Even on the basis of the FAO data, there have been steep reductions over the 15-year period in Malaysia, the Philippines, Thailand, Laos and Vietnam, while other studies record much greater losses. Thus, in the Philippines, about 70 per cent of the country was forested at the end of the Spanish period and 57 per cent in 1934, but by 1969 the forest area had fallen to only about 35 per cent and by 1988 to 21.5 per cent (Bautista, 1990). Timber production, which peaked between 1967 and 1975, is now only 33 per cent of the former level (Boado, 1988). Moreover, only 15 per cent of the remaining forested area was of the valuable 'old-growth dipterocarp', once the dominant type but which in 1988 occupied only about 20 per cent of its 1969 area (Bautista, 1990). In Vietnam, a reduction of from 48 to 24 per cent of the country under forest was recorded between 1943 and 1983 (Le Trong Cuc, 1989). The recent decline in Thailand has been even more dramatic, the forested area falling from 53 per cent of the country in 1961 to only 29 per cent in 1985 (Arbhabhirama et al., 1988). All these sources suggest more rapid loss than does Figure 1.3.
A comparison of the forest cover of Peninsular Malaysia between 1972 and 1981 (Figures 1.4a, 1.4b) shows clearly the scale of transformation that has taken place. The peninsula's forests have been greatly reduced by large-scale land-settlement schemes in this period, as well as by selective logging, which has extended over huge additional areas in that one decade, making deep inroads into the largest remaining block of primary forest in the peninsula. Further encroachment took place between 1981 and 1991, especially in the north and east, but may now have reached its limits, as land development by the principal government agency was halted in that year, costs having exceeded the returns from new settlements achieved (Government of Malaysia, 1991 a). Data of comparable quality are also available for a few other areas of South-East Asia, including a part of eastern Borneo, where perhaps the best use of remote-sensing data so far achieved in the region has been undertaken (Malingreau, Tucker and Laporte, 1989). For most areas, however, only much more generalized maps are available.
It has already been noted that the international data are variable in their information on the fate of wholly deforested land. However, some comparative trends of significance are revealed in broad terms. Arable land has increased sharply only in Thailand, where it has risen from 31 to 39 per cent of the whole (Figure 1.5). The small area under permanent crops in Vietnam has almost doubled, but the arable area has declined since 1978. The small increase in Malaysia is not easily explicable, since in Peninsular Malaysia most modern clearance has been for agriculture, so that by the official landuse surveys the agricultural area of the peninsula increased from 21 to 35 per cent of the whole between 1966 and 1982, an increase of 68 per cent in 16 years (Brookfield, 1993; Brookfield and Byron, 1990). The definition of arable land should include temporary fallow, hence giving great scope for variable interpretation between countries and through time; this problem has to be borne in mind in all measures based on the stated arable area. The problem is particularly acute in data for Papua New Guinea, which is shown as having only 0.07 per cent of its area in arable land. Yet survey data derived from air photography in the 1970s show 0.5 per cent in current arable, and if all fallow land-use types are added, together with permanent crops, the proportion rises to 6.6 per cent. An alternative basis of estimating mean per capita use of land plus fallow on a mean 10-year cycle, and adding permanent crops, would yield a figure of 5.6 per cent (R. L. Hide, personal communication).
Despite these data problems, it is probably correct that the actual expansion of arable land in most parts of the region has been relatively small in the last 15 years. This has significance in relation to the rapid population growth discussed in Chapter 2. During this same period, the food needs of growing numbers of people have been met, and without reliance on large volumes of imported food. In important lowland areas, great improvements in yield have been achieved, and there has been significant intensification of cropping. However, this is not true everywhere, and it is not true in most of the many upland parts of South-East Asia. A shortening of fallow periods has been noted in several of these upland areas, with potentially serious consequences for sustainability of production and environment.
Major improvements in yields of cereal crops-overwhelmingly rice- have been created by the Green Revolution, so that growth of production has ceased to depend mainly on increases in agricultural area. Multiple cropping has also expanded and, because of it, the harvested area has enlarged more rapidly than the arable area. The area harvested for cereals (Figure 1.6) exhibits a marked upward trend over the 12 years 1978-89, with an increase of 13 per cent; production has meanwhile increased by 53 per cent. However, technology has not rendered output immune to the vagaries in rainfall. The dry-season harvested rice area in Central Thailand fell by almost 50 per cent from the wet year 1979 to the drought year 1980. Aggregated over the region, the trend in area harvested exhibits principal breaks in its upward trend in 1982 and 1987. These were years of drought related to the El Niño Southern Oscillation (ENSO) in large and important parts of the South-East Asian region, and the data show the significance of these episodic events, discussed by Nicholls in Chapter 7.
There are some sharp contrasts between national trends, which were in some cases influenced by political events. Thus, in Cambodia both area harvested and production fell by about 40 per cent between 1978 and 1979, and in Myanmar production as well as area harvested have stagnated through most of the 1980s. Philippine production increased sharply between 1983 and 1986, but the progress has not been sustained.
Over the region as a whole, the area harvested for rice increased by 14 per cent between 1978 and 1989 with breaks in growth only in 1982 and 1987; production increased by 52 per cent, with a break in growth only in 1987. Thailand and Indonesia together accounted for 69 per cent of regional production in 1989. Whereas Thai area and production expanded respectively by 24 and 22 per cent over the 12 years, or by 35 per cent from 1979, Indonesian area was up by only 15 per cent, but production by 69 per cent (Figure 1.7). In the early 1970s, farmers in Central Thailand required five times the land needed in an intensively farmed pan of Central Java to obtain an equal crop (Barker and Anden, 1975). By 1989, the mean yield for all Indonesia was more than twice that of Thailand. During this period, Indonesia changed from being the world's largest importer of rice to being self-sufficient. In Thailand, there was much less adoption of new varieties and associated methods, in part because of inadequate control over irrigation and rain-water in both wet and dry seasons, and in part because of less intensive farming practices on larger holdings. These problems were recognized early, but have persisted (Arbhabhirama et al., 1988; Sris-wasdilek, Adulavidhaya and Isvilanonda, 1975). In terms of future management, they have important implications.
There was also a spectacular increase in rice productivity in Vietnam, though from a lower base in terms of yield; a production rise of 80 per cent was achieved against an area increase of only 7 per cent. The Philippines achieved 31 per cent growth in rice production with only 13 per cent increase in the area harvested. Malaysia, by contrast, expanded production by only 13 per cent, while the area harvested rose by 9 per cent; the best year was 1985, since when both area and production have declined. The ricefarming sector in Malaysia has consistently remained an area of relative poverty in a rapidly industrializing country, and over 900 square kilometres of irrigated land have gone out of production, partly due to hydrological problems, and partly because labour has been transferred to other parts of the economy (New Straits Times, 4 January 1991).
TABLE 1.2 FAO Index of Per Capita Agricultural Production by Country a in South-East Asia. 1979-1990 ( 1979-81 = 100)
The contrasts in agricultural change may be summarized by using the FAO index of all per capita agricultural productivity over 1978-89 (Table 1.2). The most remarkable growth is in Cambodia from its very low base, but this does not yet reflect high productivity. Among the larger countries, Indonesia and Vietnam exhibit parallel growth, and draw well ahead of Thailand; since 1984, Thailand has had a significantly lower share of GDP in agriculture than the other large regional countries. The per capita agricultural productivity of Papua New Guinea has stagnated while its mineral production has increased so as to completely dominate the national economy. The total productivity of Philippine agriculture has been almost stagnant or in decline since 1981 and that of Myanmar has dropped sharply since the mid-1980s. Especially in the latter half of the 1980s, however, the leader in real growth is Malaysia with its high dependence on export tree crops.
Exports of agricultural, forestry and fishing products distinguish the export-based rural economies of Malaysia, Thailand and Indonesia from the other regional countries. Deleting the re-exports of Singapore from the regional total, these three countries produced 79 per cent of all greater South-East Asia's non-mineral primary exports in 1978 and 87 per cent in 1990. Among other countries, only Vietnam shows a significant increase; the value of exports from the Philippines declined from 1980 to 1986, and there has since been only a small recovery. Myanmar and Papua New Guinea have also lately exported at below peak levels attained some years ago.