In this study of shifting cultivation, it was possible to hold the ethnic group constant in four different contexts and make comparisons similar to those made in the earlier 1979-80 MAB Project (see box). How do the Uma' Jalan modify their shifting cultivation practices in remote, semi-remote, and accessible areas? What processes can
be identified by documenting agroforestry practices over time, in the various contexts? The opportunity to investigate Uma' Jalan participation in a Transmigration project was fortuitous.
The following comparison of population, land/forest use, productivity, and land tenure in the four communities examined in this study demonstrates:
In matters of forest policy, population increase is often taken as a given. The impact of population increase on forests is obvious, and often disastrous. In Kalimantan, there are two important components of population increase: natural increase and immigration.
In Long Segar, the 1990 population of 551 is roughly half of the 1980 figure of 1,052.2 Long Ampung has also decreased slightly, from 486 to 460. Long Segar has spawned three more Kenyah villages. Muara Wahau has an Uma' Jalan population of 164, Tanah Merah of 156 and Karang Umus of roughly 100. Added to the present Long Segar population, this gives a total of 1,070, very close to Long Segar's 1980 population. There is no evidence that any significant number of people has moved to the city or elsewhere. Nor do the population pyramids suggest that a significant number of men have left in search of timber or plantation jobs (see Figures 12, 13, 14). The Uma' Jalan do not appear to be contributing much to East Kalimantan's overall population increase.
The average household size' in Long Segar has decreased fairly steadily from 6.324 in 1980 to 5.35 in 1990.5 In Tanah Merah, average family size was 5.83 in 1986 and 5.20 in 1990, with a gradual downward trend. Those who moved to the Muara Wahau transmigration site have maintained a slightly higher family size, like other transmigrants, with an overall mean of 6.01 and no trend. In Long Ampung, average household size went from 5.89 in 1985 to 6.84 in 1990, with a fairly steady increasing trend. However, in a 1980 census the average household size was 7.03, so over the ten year period household size has gone down there as well.
The sex ratios of the agriculturally active population was calculated for each year and each location. A significant trend was found only in Long Segar, which shows a steady increase over time, from 0.87 in 1980 to 1.27 in 1990 (Figure 15). The increase in the percentage of males active in agriculture is significant when compared to the sex distribution in the total population from the 1990 village census, which shows a very even 279 males to 272 females. This apparently confirms a trend that was observed in 1980, which indicated that women's roles in agriculture were being eroded by "modernizing" influences (Colfer 1983, 1985b).6 This becomes clearer when compared to the more traditional situation in Long Ampung, where the mean ratio over five years is 0.81.
Thus the natural increase of the Dayak population originating from the communities studied was near zero. Apparently, immigration is the principal source of East Kalimantan's population increase, which in turn is one of the principle threats to its forests.
The Uma' Jalan practice a comparatively sustainable form of "forest agriculture". Although the residents of Long Segar have cut extensive areas of old growth, they have not over-used the land and most of their land remains covered with secondary forest of varying ages. Areas they use are not typically invaded by Imperata cylindrica or other "noxious" grasses, suggesting greater sustainability than more intensive systems. In order to support and enhance this comparatively extensive and sustainable system of agroforestry, it is necessary to determine how much land is actually needed to practice it, starting by establishing how much land the Uma' Jalan are currently using.
Over 28 years of Long Segar's habitation, 6054 hectares were cleared for ricefields (including previous year's ricefields and all stages of forest regeneration). 73% of this land, or 4419 ha, was cut from old growth forest in order to insure subsistence and improve the standard of living over the 28 year period. The village thus cleared, on average, 216 hectares per year, of which 158 hectares was old growth. As the years have gone by, a small but increasing proportion of this forest has been commercially logged prior to use.
In Muara Wahau, in three years, the Uma' Jalan have cleared 209 hectares, 95% of which was in old growth, in addition to their original allotment from the Transmigration Project. Much of this land had already been selectively logged. In Tanah Merah, in eight years, 403 hectares of selectively logged forest, 52% of which was old growth, have been cleared. During the past five years in Long Ampung, 883 hectares have been cleared, only 3% of which was old growth, and none of which was commercially logged. Virtually all the forest used for rice cultivation in Long Ampung has been used previously and is in a rotating, long "fallow" system.
The Uma' Jalan agroforestry system, as currently practiced, appears flexible enough to sustain human life indefinitely. But pressures from the outside require that adaptations be made, and that the amount of land cleared be reduced. The concern with maintaining old growth forest means that steps should be taken to reduce or stop further cutting in old growth. Benign policies, however, should start with land use as it is now, and initiate research to refine the existing system.
Land requirements can be determined by multiplying the average family field size by the period of time necessary for forest and soil rejuvenation. The resulting number of hectares can serve as a guide in establishing a sustainable system which satisfies people's subsistence needs. Concerns about sustaining the timber industry and conserving old growth forests will require additional considerations.
Average field size in Long Segar, a former Resettlement Village which had theoretically been allotted three hectares per family, was 2.35 ha. This was the largest overall average field size of the four study communities. The average Long Segar family used 2.7 ha per year for rice cultivation. Local estimates of required "fallow" time range from 12 to 15 years .7 This amounts to a land requirement of 32 to 40 ha per family under the present system.
The people in the Muara Wahau Transmigration site make the second largest fields, averaging 1.88 ha. More significantly, during the last three years, in which the Uma' Jalan abandoned their allotted Transmigration plots and returned to clearing forest, they cleared the largest fields of the four communities (see Figure 16), cutting the highest proportion of old growth. The environment at Muara Wahau prior to the introduction of intensive logging and agriculture was roughly comparable to that in Long Segar, similarly requiring 32 to 40 ha per family for a sustainable agroforestry system.
Tanah Merah fields are smaller, averaging 1.45 ha. This community is under pressure to reduce its land use by the timber company, other neighbouring groups and the Government. This seems to be motivating people to be more experimental in changing their agroforestry system. They are successfully experimenting with marketing of agricultural produce other than rice, most notably bananas, and actively
planting various tree crops with the intention of securing rights to the land. Threats to land access are more obvious in this area, since land is often bought and sold in nearby Tenggarong and Samarinda; a few people even bought their house/lots in Tanah Merah, a first for these Uma' Jalan. Even under these conditions, however, and assuming that the required fallow period could be shortened to ten years (because of Tanah Merah's better soil quality) and that every family could manage with one ricefield per year, the need is about 15 ha per family.
The smallest field sizes were those in Long Ampung, with an overall average of 1.02 hectares. This is probably due to the lack of a market for any surplus: there is simply no reason to make large fields under normal circumstances. The soil regenerates more quickly in the Apo Kayan, in 8 to 12 years. Recognizing the frequency with which people have two ricefields and so calculating 1.5 ha per family over a ten year rotation, the need again amounts to 15 ha per family.
Thus, only for agricultural sustainability within the existing systems, each family needs between 15 and 40 hectares. The Resettlement and Transmigration programmes typically allot three to five hectares to each family for subsistence. If concerns about stability and sustainability are taken into account, both human and environmental, this is clearly not enough. In addition, the length of time needed for the replenishment of old growth forest remains an issue. Certainly more than 15 years are needed for regrowth of a mature tropical rainforest. Even sustainable timber harvesting requires a longer cycle.' Additional time must be allowed, and therefore land set aside, for sustainable timber harvesting and for maintaining relatively undisturbed old growth forest. This is important for both the sustainability of shifting cultivators' agroforestry system and for international biodiversity concerns.
Another important feature of land use among shifting cultivators is the distance to fields. The farther they must travel, the more inconvenient for the farmers. Such inconvenience represents an opportunity for encouraging more intensive agroforestry use of lands closer to home. The distances travelled from home to field in Long Segar, Muara Wahau and Tanah Merah show a clear trend of increasing distance with the passage of time (Figure 6, p.36). Long Segar shows a gradual lengthening of the distance over the years, from 0 km in 1962 to an estimated average distance of 6.71 in 1990, having reached a high the previous year of 6.94. Long Ampung, the oldest community, has relatively short distances to fields with no trend. This is a particularly significant finding, indicating that its inhabitants are reusing older (and now regenerated) lands in a sustainable agroforestry system.
An interesting factor in the Long Segar land use pattern is the continued, though erratic, involvement of the Government in land use decisions.9 The most distant field site used by Long Segar now, in the Pantun River system, is actively encouraged by the Muara Wahau Transmigration project, which sends extension agents to help the people with tree crop planting. The timber company is also involved, supplying seedlings and additional extension services, in compliance with the government regulation obliging timber companies to provide guidance and help to local communities (Program Bina Desa, or Village Guidance Programme).
Uma' Jalan prefer to be able to conduct their agroforestry activities closer to home. They also maintain that cutting down large trees is difficult, dangerous work which they would gladly avoid. By reducing timber cutting and travel time, projects for intensifying production in a rationalized rotating agroforestry system which builds on their current practices would be particularly welcome. But refinements of their system cannot be attempted without an adequate land base, considerably larger than what is typically specified in resettlement projects at present.
An assessment of existing productivity is necessary for any attempt at improvement of agroforestry systems. The most consistently used measure of productivity comes from rice harvests, but a fair assessment of shifting cultivation cannot stop at foodcrop production. It must include the many other products that shifting cultivators harvest from the surrounding forests.
Rice production at the various study sites differs considerably. Long Ampung had the lowest average yields, 783 kg/ha. Two factors discussed above probably account for this: 1) 74% of their fields are from young secondary forest, one of the least fertile types of land; 10 and 2) the absence of markets gives people little motivation to maximize rice production. Long Ampung remains almost purely a subsistence economy. Muara Wahau produced the next lowest yields, with 818 kg/ha on average. Long Segar averaged 1,170 kg/ha. Tanah Merah had the highest average yields with 1,393 kg/ha. The superior soil in the Jembayan River area may account in part for the better yields, as well as the higher proportion of old secondary forest cut for ricefields. Motivation to sell rice produce is also strong, with easy access both to markets and to consumer goods which can only be purchased with cash.
Farmers carrying Durian seedlings to plant in their rice fields
These figures, however, only account for rice productivity. One of the most important aspects of the Uma' Jalan agroforestry system is the diversity of products that it harvests, hunts and gathers." Appendix B contains a partial listing of forest plants used by the Kenyah, although the variety of products consumed is virtually impossible to quantify. The following is only a brief list of some typical examples of this vast assortment of products.
During the rice production phase, cucumbers, corn and chilies are grown in the ricefield. Mushrooms, ferns, locusts and other animals are harvested for food in the field or its periphery. Longer term crops are also planted (pineapple, sugarcane, cassava) and harvested as needed during the succeeding years. Near the field hut, fruit trees are planted (papaya, jackfruit, bananas), which also continue to bear for years to come. Such fruit trees can even be successfully protected when reburning an area for a new field.
As the forest begins to grow up on the field, the planted fruit trees are supplemented by wild fruits that take root. Various other products, like rattan, bamboo, palms and pandanus can be planted, tended, or simply collected.
Many kinds of wood are used by local people for house and fieldhut construction, boats and paddles, roofing shingles, musical instruments and firewood. As the forest matures, a greater variety of animal life begins to inhabit it. These animals provide important protein in the people's diets, as well as various ornamental and often useful objects (teeth, skin, fur, feathers, beaks). The continual ground cover also ensures reasonable water quality that in turn sustains fish, snails, shrimp and other freshwater products in the rivers.
Mechanized logging has a marked negative effect on these nonfieldcrop aspects of the Uma' Jalan agroforestry system. Besides the direct damage to the forest incurred by the logging process (opening of the canopy, soil erosion and compaction from heavy equipment, reduced water quality in rivers, damage to surrounding trees, etc.), it is believed that logging makes a humid tropical rainforest drier and hence more susceptible to forest fires. The absence of forest fires during 1972 drought and the devastating ones during the 1982 drought could be related to the increase in logging activity in the area.
The Uma' Jalan, in the past, have been able to turn to the non-rice components of their agroforestry system for subsistence in times of natural disaster. They hunt, fish, collect produce, and depend on cassava or other perennial crops less affected by the particular disaster. But continued logging (as well as the other misuses of the rainforest) seems likely to adversely affect the viability of the overall system by provoking a greater dependence on annual crop production than is warranted or even possible.
Future attempts to manage these forests should focus on these "subsidiary" forest products, many of which have a long history as export products.12 Enhancement and refinement of this integrated agroforestry system should play a major role in forest planning and management. Such an integrated approach can serve to protect the forests, enhance people's incomes and increase national productivity, more effectively than current fragmented efforts.
A cultivator's willingness to manage land in sustainable ways is closely linked to land tenure, and increased tenure security is as necessary in East Kalimantan as in other areas of Indonesia. Since people rarely have written title to their land, a better understanding of traditional land tenure systems is important. Traditional (or adat) rights to land are formally recognized, although subject to varying interpretations, within Indonesia's Basic Regulations on Agrarian Principles (originally formulated in 1960, see Agraria 1976).
Efficient management of Indonesia's forest will require that this formal recognition of indigenous people's claims on their land be put into practice (see also Mitchel et al 1990).
Thus an understanding of both Indonesian law and traditional land tenure is necessary in order to make viable recommendations for agroforestry development. Given the Indonesian population distribution, many of the relevant laws were written primarily with the Javanese context in mind. Article 5 defines rights of ownership (hak milik) in the following way:
The Agrarian law which applies to the earth, water and air space is Adat-Law [traditional law] insofar as it is not in conflict with the National and State interests based on the unity of the nation...
Marker indicating land ownership
A first problem with applying Indonesian law, as it now stands, to shifting cultivators, is illustrated in Article 7:
In order not to harm the public interest, excessive ownership and control of land are not permitted.
The definition of "excessive ownership" in fertile, densely populated Java, where land ownership is often measured in 10m2 units (ares), is clearly quite different from that in a sparsely populated, unfertile tropical rainforest area like East Kalimantan. Similarly Article 10, Paragraph (1), written to minimize problems of absentee landlords, ignores the forest fallow periods necessary to maintain soil fertility and productivity in East Kalimantan (see also Ngo 1990):
Every person and every corporate body having a certain right on agricultural land, is in principle obliged to cultivate or to exploit it actively by himself...
Ambiguity can derive from the fact that shifting cultivation is actually an agroforestry system, neither fully agriculture nor fully forestry. Article 45, though it does expressly state people's right to collect under certain conditions, ignores the integral role which forest product collection plays in indigenous shifting cultivation systems:
(1) The right of opening up land and of collecting forest products may only be possessed by Indonesian citizens and is regulated by Government Regulations.
(2) Using the right to collect forest products legally does not naturally imply ownership with regard to that land.
Contradictory interpretations of the law are easy in the case of shifting cultivation. Article 15, for instance, states:
The cultivation of land, including the increase of its fertility as well as the prevention of its damage is the duty of every person, corporation or organization having legal relations with mentioned land, with due consideration to the economically weak party,
But Article 27 maintains that:
The right of ownership is annulled if: a. the land falls back to the state: [...] (3) because the land is lying fallow. b. the land is destroyed.
The fertility-enhancing effects of forest fallows in the shifting cultivation system would be difficult to dispute. But it seems that this activity would also justify the elimination of ownership rights to the land lying fallow. While the word "destroyed" is not very clear in this context, it could be argued that transmigration and some timber or mining activities are destroying the land by depleting it, while integral shifting cultivators who let their lands lie fallow are in fact protecting it.
Rights of use (hak ulayat), as opposed to rights of ownership, are also recognized and regulated by the law (particularly Articles 41 and 43), and could be used as a basis for formulating policies more consistent with shifting cultivation. But it has been argued that these Articles are currently used by officials as a substitute for granting rights of ownership, weakening the claims of indigenous groups (Dove
1987).
In Indonesian Forestry Law, Law No. 5/1967 specifies the Basic Rules of Forestry. Combined with the Agrarian Law recognizing traditional land rights, these rules would seem to support the rights of shifting cultivators to their land (assuming that the rights granted under adat law are considered "proprietary rights"). Article 2 states:
Depending upon the ownership, the [Forestry] Minister shall declare a forest as:
(1) "State Forest", which is a forest region or forest growing on a piece of land not covered by any proprietary rights.
(2) "Proprietary Forest", which is a forest growing on a piece of land covered by proprietary rights.
Article 11, Paragraph (1) expands thus:
The management of Proprietary Forests shall be carried out by their owners under the guidance of the Minister...
The rights of Indonesian citizens to collect forest products (Article 14 (4)), hunt wild animals (Article 16), and participate in protecting the forests (Article 15 (3)), in compliance with governmental regulations, are expressly granted.
However, the Basic Rules of Forestry also contains some inconsistencies and ambiguities.14 Article 5, Paragraph (1) specifically states that:
All forests within the territory of the Republic of Indonesia, including the natural resources they contain, are taken charge of by the State.
Article 17 states:
The exercise of communal and individual rights to exploit or benefit from forests based on some or other existing legal regulations [e.g. the Agrarian Law?], shall not infringe upon the achievement of the aims stated in this Law.
Although agricultural and forestry laws are most germane to shifting cultivation issues, agencies concerned with mining, the environment or resettlement also establish procedures and laws which have an impact on people living in forests." The situation of Long Segar has been examined from the legal standpoint (Moniaga 1990), and it was found that formal involvement of Long Segar residents in the Resettle ment Programme amounts to a tacit acknowledgement of their right to be there. Three separate regulations require registration of land by owners, but compliance in Long Segar would be legally impossible: such registration must be approved by a village head recognized by the Government, but Long Segar does not have an administrative village status and thus has had no legal village head with this authority. 16. In 1983, a Consensus Forest Use Plan (TGHK) was drawn up which officially included Long Segar in an area demarcated as "conversion" forest (available for conversion to other uses, usually agriculture).
Two crucial aspects of Indonesian law are 1) the nature of adat, or traditional law, and 2) whether or not adat law is still recognized locally. The most fundamental feature of traditional Uma' Jalan land tenure is the cutting of old growth forest. This act, in theory, assures the cutter, his household and those who inherit from him the right to re-use the land at their discretion. Nothing need be planted on the land to maintain the cutter's (and his descendants of both sexes') rights to it. Uma' Jalan in other locations can still describe the forest areas that they consider their own in Long Ampung (current views of Long Ampung residents discussed in the section "A Return to the Longhouse of the Tarsier," Ch. 2, p.60). This ownership, however, is quite theoretical, since no one has returned after moving away to make ricefields on Long Ampung land. The people of Long Ampung readily acknowledge that they use others' land freely now. This is consistent with the view that in Bornean land tenure residual rights remain in communities."
In areas nearer to the city, where land is normally bought and sold, this practice might be expected to change." In Long Segar and Tanah Merah, there is confusion about land tenure. Although villagers still recognize their own traditional land tenure system, they realize that the Government often does not. People still cut old growth forest in the hope that this will establish rights to land, but they feel no certainty that this is the case. The fact that both communities are situated in timber concessions complicates traditional rights. Unsure of their legal rights, villagers realize that the concession holders to have important, even preeminent rights to this land. As time goes by and timber resources become scarcer and more valuable, conflicts between the companies and local populations will increase (quite an uneven competition; see Dove 1987; Moniaga 1990; Ngo 1990).
The Government also considers shifting cultivation lands lying fallow (in varying stages of forest succession) to be unused. Land that is not used "reverts" to the State. Thus in these communities (as well as in Muara Wahau) people have begun to plant tree crops, with the explicit purpose of establishing more secure claims to land. Tree tenure is a related and complex issue in the traditional system.19 Planting a tree (or other crop), in theory, gives one ownership of that tree or plant, regardless of who owns the land, though one cannot plant more than perhaps 10 trees on someone else's land. One's rights to what one plants are not absolute (see box). Anyone is allowed pick someone else's fruit to satisfy hunger on the spot; but one is not permitted to take it home, sell it, or destroy it without special permission. Similarly plants from people's forested fallow lands which have useful parts (e.g., banana leaves, medicinal plants, bamboo, pandanus) belong to the landowner.
One is expected to ask permission to harvest plant parts or fruits from other people's land for private use, though permission is normally granted. Yet if there is a community function, like the rice harvest party (uman undat), the community is free to go and collect what is needed for the function from anyone's land.
Although the Kenyah system involves use of "fallow land" (or forest in the process of regeneration) for a variety of useful products, this aspect of their system is often not sufficiently understood by government authorities. For example, to be considered "in use" by officials, tree crops must be planted in rows, and preferably as a monocrop. Some Kenyah now realize this, and are encouraging the planting of their trees in rows so that the authorities will acknowledge that the land is being used (and thus is owned).
This kind of change is likely to result in less sustainability, stability and productivity than the traditional system .22 An improved (or "rationalized") version of the complex, mixed cropping that characterizes traditional Kenyah agroforestry patterns is much more likely to achieve greater overall productivity, stability and sustainability. Acknowledgement and formal clarification of shifting cultivators' claims to land is an important prerequisite for environmentally suitable forest management which incorporates the knowledge, abilities and rights of the indigenous people.
A variety of attempts have been made to improve the standard of living of tropical forest dwellers and to sustain tropical rainforests as a source of national income. Such attempts have often suffered from a lack of information, as well as a plethora of inaccurate assumptions and beliefs, such as the view that integral shifting cultivators are only marauders, rather than managers of their forests (see the "Ten Common Beliefs" in the Introduction, p.5). But international interest in maintaining the biodiversity that characterizes tropical rainforests, and national interests related to timber and mineral extraction, transmigration and agricultural production, are such that much more is at issue than the agricultural sustainability of indigenous shifting cultivation. The development of policies that will protect the human and biological diversity that characterizes such forests will require the following five important policy initiatives.
A first priority must be the containment and regulation of the timber companies which now have a virtually free reign in cutting the forests of Indonesia. Enforcement of the existing contracts between the Ministry of Forestry and the timber companies is imperative if Kalimantan's forests are to continue to support people, contribute to national income and serve as a global gene pool. The Ministry of Forestry estimated in 1990 that only about 4% of the timber concessionnaires complied with the requirements of their contracts with the government.
Of Indonesia's total timber production, 40% comes from East Kalimantan (Sakuntaladewi and Amblani 1989). In East Kalimantan alone, 9.4 million hectares (94,000 km2) are within timber concessions (Beukeboom 1989). The Government of Indonesia has classified East Kalimantan's 178,989 km2 of forests in the following way: 41,241 km2 are "regular production" forest, 52,940 km2 "limited production" forest, and 38,675 km2 "conversion" forest. Another 46,000 km2 are classified as "conservation" or "protection" forest. Production forest is the main category in which the timber industry is expected to operate. Limited production forest, for varying reasons, is less appropriate for logging, and thus has special, more restrictive regulations applying to logging operations. Conversion forest has been identified as appropriate for conversion to agricultural uses (including plantation agriculture, transmigration, and, in practice, shifting cultivation). Conservation and protection forests include nature reserves and upper watershed areas. Although gov ernmental regulations for use of timber resources are adequate, sometimes quite good, there is virtually no enforcement and little coordination. Areas classified as conservation or protection forest, for instance, are sometimes granted to timber concessionaires for logging, and often used for shifting cultivation.
Although shifting cultivation is reported in the same document to account for only 6,643 km-', this is a meaningless figure since the forest fallow areas are not included. Even were such components to be included, these areas cannot easily be differentiated from forests in succession for other reasons (natural or human), so that figures on the extent of this kind of shifting cultivation are often inaccurate. The above figures do, however, reflect the significant difference in scale between the activities of the timber industry and those of shifting cultivators in East Kalimantan.
A second priority is to grant the indigenous people of Kalimantan unambiguous and comparatively long-term access to significant areas of land. As seen, for maintenance of a reasonable but somewhat reduced subsistence base, cultivators need between 15 and 40 hectares per family, depending on the forest fallow period required in that area. For continued maintenance of significant areas of old growth forest, 100-200 hectares per family are necessary. Such rights, in either case, must be tied to a planning process that will restrict the cutting of old growth forest. Communities should maintain linked community forest reserves, with buffer zones, to preserve community access to minor forest products and habitat for forest animals. Such reserves must also have legal status and be under largely local management .24 The legalization of local people's claims to their land is preferable also because people typically behave in ways they perceive as profitable for themselves and their families .25 Assuring them their rights to land is the first step toward giving them back their stake in forest conservation.26
Georgia Pacific truck
hauling timber in the
Long Segar area
The potential adverse effects of shifting cultivation are usually recognized but often vastly exaggerated, while the potential for a positive contribution from communities of shifting cultivators is completely ignored. Forest dwelling people have a vast store of indigenous knowledge useful in conservation and development.27 The biodiversity and valuable timber reserves that characterize these forests grew up in part under the management of shifting cultivators.
Research and development activities that link modern scientific methods with the forest knowledge and management strategies of indigenous cultivators need to be expanded. There have been decades of largely fruitless attempts to adapt sedentary food crop agriculture to humid tropical rainforests. These methods have usually proved too risky, too expensive, and too environmentally costly, or, put another way, unstable, unproductive, and unsustainable. Shifting cultivation offers a possible model that is only beginning to be tried seriously and scientifically. Two early efforts are: enrichment planting, building on local people's knowledge (discussed in Leighton and Peart 1990); and the forest management system used in the Palcazu forest of Peru, the key feature of which is the rotation of long, narrow (30-40m) strips through a production stand of timber, devised from a combination of indigenous and scientific knowledge (described in Hartshorn 1989, 1990: Hartshorn et al 1987). Collaborative approaches such as those used in Farming Systems Research and Development could be expanded to incorporate a more meaningful ecological component (Colfer 1991b; see also Poffenberger's 1990 collection).
Another critical endeavour for forest maintenance must be more active efforts to stabilize population. The availability of birth control for local people is important, since the need for land and forest typically increases with increasing population. Even more important is the need to revise programs which increase the population through transmigration. East Kalimantan, which had a total population of 1.5 million in 1985, had absorbed 173,452 government-sponsored transmigrants by 1989, a population increase of 11 % (Sakuntaladewi and Amblani 1989), and an undetermined but relatively large number of spontaneous transmigrants. The soils that sustain tropical rainforests have low fertility and cannot support dense populations practicing settled agriculture. Transmigration, though well-intentioned, tends export Java's poverty to the Outer Islands.
Urging the Indonesian government to implement more environmentally sensitive policies is laudable. But it is not enough. Two additional facts must be remembered: 1) Indonesia needs the revenues from its natural resources for development activities; and 2) the entire world benefits from maintaining the cultural and biological diversity of Indonesia's tropical rainforests. Effective mechanisms must be found for transferring funds from wealthier countries, as well as from national sources, to Indonesian Government efforts to maintain these precious resources.
Immediate action on these issues by the Indonesian Government and the community of nations is of critical importance for the maintenance of any significant amount of humid tropical rainforest. The biodiversity characterizing these environments is a precious and irreplaceable resource, but a less widely valued feature of tropical rainforests lies in the people who inhabit them. The human systems that have emerged in tropical rainforests, just like the plants and animals, are part of an irreplaceable and precious global heritage. The loss of such human systems, and their associated indigenous knowledge, would indeed be tragic. Indigenous management systems can and should serve as a foundation in the development of new agroforestry systems for the humid tropics, systems that will be sustainable under our changing global conditions.
