The field research in the Riam Kanan valley was carried out in 1980 and 1982, with kind assistance from the South Kalimantan provincial government and department of forestry. The research and subsequent data analysis and writing were variously supported by the FAO/UNDP, the Rockefeller Foundation,/the Ford Foundation, and the East-West Center (EAPI). The author is also grateful to the Gadjah Mada University Environmental Studies Center for institutional support during this period. Finally the author wishes to thank B.S. Eko Prakoso for his assistance in drawing Figs. 1 and 2, Samuel Fujisaka and Laurence L. Hamilton for their helpful comments on an earlier draft of the paper. Earlier reports on this research were presented at FAO workshops in Yogyakarta on 25 June 1980, and in Manila on 15–18 October 1980. Other publications on this research include the following by Dove, Michael R.: “Masalah Ekologi-Mikro dan Pembangunan-Makro: Suatu Kasus di Kalimantan Selatan”, Agro-Ekonomika XII, 14 (1981): 25–40Google Scholar; “Symbiotic Relationships Between Human Populations and Imperata cylindrica: The question of Ecosystemic Succession and Preservation in South Kalimantan”, in Conservation Inputs from Life Sciences, ed. Nurdin, M. et al. (Bangi: Universitas Kebangsaan Malaysia, 1981), pp. 187–200Google Scholar; “Theories of Swidden Agriculture and the Political Economy of Ignorance”, Agroforestry System 1 (1983): 85–99CrossRefGoogle Scholar; and “The Practical Reason of Weeds in Indonesia”, Human Ecology in press. The author alone is responsible for the analysis presented in this paper.

¹ This sort of vegetation gradient is common in the tropics. For an African example see Sanchez, P.A., Properties and Management of Soils in the Tropics (New York: John Wiley and Sons, 1976), pp. 387–88Google Scholar.

² Musa, Syarkani et al. , “Ekologi Kawasan Waduk Riam Kanan, Kalimantan Selatan” (Paper read in seminar at the Regional Planning and Development office,Banjarmasin,16 March 1983), p. 5Google Scholar.

³ Ibid., pp. 5, 7–8.

⁴ Ibid., p. 3.

⁵ The reservoir apparently has had other impact as well. The villagers of Rantau Balai claim that the duration of flooding is now too long for wet rice to be grown along the banks of the Riam Kanan, whereas this was not the case before the dam was constructed.

⁶ For Borneo, see Dove, Michael R., Swidden Agriculture in Indonesia: The Subsistence Strategies of the Kalimantan Kantu (Berlin: Mouton, forthcoming)CrossRefGoogle Scholar; Drake, R.A., “The Material Provisioning of Mualang Society in Hinterland Kalimantan Barat, Indonesia” (Ph.D. diss., Michigan State University, East Lansing, 1982)Google Scholar; Freeman, Derek, Iban Agriculture: A Report on the Shifting Cultivation of Hill Rice by the Iban of Sarawak (London: Her Majesty's Stationery Office, 1955)Google Scholar; Geddes, W.R., The Land Dyaks of Sarawak (London : Her Majesty's Stationery Office, 1954)Google Scholar; and Padoch, Christine, Migration and Its Alternatives Among the Iban of Sarawak (Leiden: Royal Institute of Linguistics and Anthropology, 1982)CrossRefGoogle Scholar. For mainland Southeast Asia, see Condominas, Georges, We have Eaten the Forest (New York: Farrer, Strauss, Giroux, 1977)Google Scholar; and Izikowitz, Karl G., Lamet: Hill Peasants in French Indochina (Gothenburg: Etnografiska Museet, 1951)Google Scholar.

⁷ Cf. Freeman, Iban Agriculture, ch. 6.

⁸ Worldwide, government planned agricultural systems in resettlements have been notoriously unsuccessful. For example, see Scudder, Thayer, “The Human Ecology of Big Projects: River Basin Development and Resettlement”, Annual Review of Anthropology 2 (1973): 45–55Google Scholar.

⁹ See Weinstock, Joseph A., “Rattan: Ecological Balance in a Rainforest Swidden”, Economic Botany 37, 1 (1983): 58–68CrossRefGoogle Scholar.

¹⁰ Musa et al., “Ekologi Kawasan Waduk Riam Kanan”, p. 11.

¹¹ Sanchez, , Properties and Management of Soils in the Tropics, p. 363Google Scholar.

¹² Ibid., pp. 364–65.

¹³ A second reason for planting these tree crops is to strengthen their rights to fallowed swidden land in the eyes of the government. See Dove, Michael R., “Theories of Swidden Agriculture, and the Political Economy of Ignorance”, Agroforestry Systems 1 (1983): 85–99 (esp. 87)CrossRefGoogle Scholar.

¹⁴ See Hamilton, Lawrence S., Tropical Forested Watersheds: Hydrologic and Soils Response to Major Uses or Conversions (Boulder: Westview Press, 1983), p. 19Google Scholar.

¹⁵ Hamilton, , Tropical Forested Watersheds, p. 18Google Scholar. See also Sanchez, , Properties and Management of Soils in the Tropics, pp. 120, 173, 361, 363, 368, 371Google ScholarPubMed; and Gibson, T., “Toward a Stable Low-Input Highland Agricultural System: Ley Fanning in Imperata cylindrica Grasslands of Northern Thailand”, Mountain Research and Development 3, 4 (1983): 378–85CrossRefGoogle Scholar.

¹⁶ Hamilton, , Tropical Forested Watersheds, pp. 16, 21Google Scholar.

¹⁷ Conklin, Harold C., Hanunóo Agriculture: A Report on an Integrated System of Shifting Cultivation in the Philippines (Rome: FAO, 1957), p. 2Google Scholar.

¹⁸ Cf. Vayda, Andrew P., “Research in East Kalimantan on Interactions Between People and Forests: A Preliminary Report”, Borneo Research Bulletin 13, 1 (1981): 3–15Google Scholar.

¹⁹ See Hamilton, , Tropical Forested Watersheds, p. 21Google Scholar; and Sanchez, , Properties and Management of Soils in the Tropics, p. 120Google Scholar.

²⁰ Hamilton, , Tropical Forested Watersheds, pp. 127–28Google Scholar.

²¹ In South Kalimantan (as elsewhere in Indonesia), the principal tool in government policy on the perceived problem of swidden cultivation is resettlement of forest dwelling peoples, meaning the integral swidden cultivators. The partial swidden cultivators are consequently left relatively alone.

²² See Low, Hugh, Sarawak: Its Inhabitants and Productions (London: Richard R. Bentley, 1848), p. 91Google Scholar; and Medway, Lord, Mammals of Borneo (Kuala Lumpur: Malaysian Branch of the Royal Asiatic Society, 1977), p. 150Google Scholar.

²³ See Richards, P. W., The Tropical Rain Forest: An Ecological Study (Cambridge: Cambridge University Press, 1952), p. 335Google Scholar.

²⁴ Indirect evidence of this concentration of population is the late nineteenth-century migration of large numbers of Banjarese to new agricultural lands in peninsular Malaya. See Hill, R.D., Rice in Malaya: A Study in Historical Geography (Kuala Lumpur: Oxford University Press, 1977), pp. 101, 180Google Scholar. On the analysis of contemporary grasslands as evidence for historic concentration and location of human populations, see Robbins, R.G., “Correlation of Plant Patterns and Population Migration into the Australian New Guinea Highlands”, in Plants and the Migrations of Pacific Peoples, ed. Barrau, Jacques (Honolulu: Bishop Museum Press, 1963), pp. 45–59Google Scholar.

²⁵ See Ras, J.J., Hikajat Bandjar: A Study in Malay Historiography (The Hague: Martinus Nijhoff, 1968), pp. 182–200Google Scholar.

²⁶ See Low, , Sarawak, p. 91Google Scholar, for his mid-19th century account of a distinctive method of hunting deer in open country around Banjarmasin; and see Medway, , Mammals of Borneo, p. 150Google Scholar, for another mid-19th century report on huge herds of Cervus timorensis on the grassy plains to the west of the Meratus Mountains. I have no historical records on herding, but contemporary Banjarese affirm that even before the introduction of the plough, it was the custom of their people to keep both cattle and water buffalo.

²⁷ See Dove, Michael R., “Man, Land and Game in Sumbawa: Some Observations on Agrarian Ecology and Development Policy in Eastern Indonesia”, Singapore Journal of Tropical Geography 5, 2 (1984): 112–24 (esp. 116–19)CrossRefGoogle Scholar; Seavoy, Ronald E., “The Origin of Tropical Grasslands in Kalimantan, Indonesia”, Journal of Tropical Geography 40 (1975): 48–52Google Scholar; and Sherman, Georgè, “What ‘Green Desert’? The Ecology of Batak Grassland Farming”, Indonesia 29 (1980): 113–48 (esp. 118–19)CrossRefGoogle Scholar.

²⁸ A decrease in forest area, caused by increasing population density and decreasing forest swidden fallows, is widely regarded as the most common stimulus for the development of grassland agriculture. See Boserup, Ester, The Conditions of Agricultural Growth: The Economics of Agrarian Change Under Population Pressure (Chicago: Aldine, 1966)Google Scholar.

²⁹ An account of the use of hoes to cultivate Imperata lands in this region almost fifty years ago is provided by Gerlach, J. C., “Bevolkingsmethoden van Ontginning van Alang-Alang-Terreinen in de Zuider en Oosterafdeeling van Borneo”, Landbouw 14 (1938): 446–50Google Scholar.

³⁰ In a study carried out in the adjoining valley of the Riam Kiwa, it was found that widows who lack the capital to cultivate the grasslands by plough, or the strength to clear the forest by hand, are forced to cultivate the grassland by hand, using a hoe. See Potter, Lesley M., “From Forest to Grassland: Resource Allocation, Intensification and Farming System Change in South Kalimantan, Indonesia” (Paper read at a meeting of the Asian Studies Association of Australia,Adelaide,13–19 May 1984), p. 6nGoogle Scholar. Among the Banjarese of Rantau Balai, widows generally farm not grassland but forest, by buying already cleared forest plots from village men.

³¹ A handful of Madurese immigrants in the Riam Kanan valley still sell their services as expert ploughmen and cattle handlers.

³² Krisnawati Suryanata, “Perkembangan Agroekosistem di Daerah Transmigrasi: Studi Kasus Desa Pantai Cabe di Kalimantan Selatan” (Report prepared for the Indonesian Department of Transmigration, 1984), pp. 45–47, calculated a maximum cultivation period of five years in a grassland area to the north of the Riam Kanan. During these five years, the land is successively planted in long beans, dry rice, maize, and cassava.

³³ Cf. Fulcher, Mary B., “Dayak and Transmigration Communities in East Kalimantan”, Borneo Research Bulletin 14, 1 (1982): 14–24 (esp. 20n)Google Scholar.

³⁴ Suryanata, “Perkembangan Agroekosistem di Daerah Transmigrasi”, pp. 48–49.

³⁵ Cf. Holm, LeRoy G. et al. , The World's Worst Weeds: Distribution and Biology (Honolulu: The University Press of Hawaii, for the East-West Center, 1977), p. 64Google Scholar; Sherman, D. George, “The Culture-Bound Notion of ‘Soil Fertility’: On Interpreting Non-Western Criteria of Selecting Land for Cultivation”, Studies in Third World Societies 14 (1980): 487–511 (esp. 497–98)Google Scholar; Sherman, “What ‘Green Desert’?”, p. 132; and Suryanata, “Perkembangan Agroekosistim di Daerah Transmigrasi”, p. 47.

³⁶ Conklin, , Hanunóo Agriculture, p. 1Google Scholar.

³⁷ Elsewhere in Indonesia, see Ormeling, F. H., The Timor Problem: A Geographical Interpretation of an Underdeveloped Island (The Hague: Martinus Nijhoff, 1956)Google Scholar; and Sherman, “What ‘Green Desert’?”. Elsewhere in the tropics, see Clarke, W.C., “From Extensive to Intensive Shifting Cultivation: A Succession from New Guinea”, Ethnology 5 (1966): 347–59CrossRefGoogle Scholar; and Netting, Robert McC., Hill Farmers of Nigeria: Cultural Ecology of the Kofyar of the Jos Plateau (Seattle: University of Washington Press, 1968)Google Scholar.

³⁸ This is widely misunderstood by workers in other fields, but it has long been known to specialists. As Holm et al. have written in The World's Worst Weeds, “Studies of the nutritive value of I. cylindrica and of its usefulness as a tropical fodder grass date back almost a century” (p. 69).

³⁹ Soewardi, Bedjo and Sastradipradja, Djoko, “Alang-Alang (Imperata Cylindrica [L.] Beauv.) and Animal Husbandry”, in Proceedings of Biotrop Workshop on Alang-Alang (Bogor: BIOTROP, 1980), pp. 157–78 (esp. 165)Google Scholar.

⁴⁰ On the necessity for periodic burning, see Holm, et al. , The World's Worst Weeds, p. 69Google Scholar; Sanchez, , Properties and Management of Soils in the Tropics, p. 538Google ScholarPubMed; and Boulière, Francois and Hadley, Malcolm, “The Ecology of Tropical Savannas”, Annual Review of Ecology and Systematics 1 (1970): 125–52CrossRefGoogle Scholar.

⁴¹ Andrews, Alan C., “Imperata Cylindrica in the Highlands of Northern Thailand: Its Productivity and Status as a Weed”, Mountain Research and Development 3, 4 (1983): 386–88CrossRefGoogle Scholar, similarly concludes that “It [Imperata] is still the basis of the cattle industry that exists in the [Thai] highlands today.”

⁴² For Indonesia, see Seavoy, “The Origin of Tropical Grasslands in Kalimantan”, pp. 49–51; Sherman, “What ‘Green Desert’?”, pp. 118–19; and Soewardi, and Sastradipradja, , “Alang-Alang (Imperata cylindrica [L.] Beauv.) and Animal Husbandry”, p. 158Google Scholar. For elsewhere in the region, see Gibson, “Toward a Stable Low-Input Highland Agricultural System”, p. 381; and Holmes, J.H.G. et al. , “The Use of Imperata Cylindrica (L.) Beauv. by Grazing Cattle in Papua-New Guinea”, in Proceedings of Biotrop Workshop on Alang-Alang (Bogor: BIOTROP, 1980), pp. 179–91Google Scholar.

⁴³ See Dove, , “Man, Land and Game in Sumbawa”, pp. 115–19Google Scholar.

⁴⁴ See Holm, et al. , The World's Worst Weeds, p. 70Google Scholar.

⁴⁵ Cf. Sherman, , “What ‘Green Desert’?”, p. 140Google Scholar.

⁴⁶ Musa, et al. , “Ekologi Kawasan Waduk Riam Kanan”, pp. 4, 6Google Scholar.

⁴⁷ This list of reforestation objectives is based on interviews with government officials and also on Musa et al., “Ekologi Kawasan Waduk Riam Kanan”, pp. 19–20.

⁴⁸ Ibid., p. 4.

⁴⁹ For example, in 1980 government contractors were willing to pay only 500 rupiah/day to recruit local labour, whereas the local villagers could earn 1000 rupiah/day (plus meals) working in one another's fields, or up to 2000 rupiah/day gathering rattan in the forest.

⁵⁰ Cf. Sherman, “What ‘Green Desert’?”, p. 123.

⁵¹ This sentiment is echoed by Musa et al., “Ekologi Kawasan Waduk Riam Kanan”, p. 17.

⁵² Although only nominal Muslims, the Banjarese appear to neither eat nor hunt wild pigs, as a result of which the pigs are ubiquitous in the valley's forested areas.

⁵³ E.g., see Holm, et al. , The World's Worst Weeds, pp. 66–68Google Scholar.

⁵⁴ Similarly among Ibanic groups in West Kalimantan, rubber trees are planted in old pepper (Piper nigrum) gardens that have become dominated by Imperata, for the explicit purpose of reclaiming the land from this succession.

⁵⁵ This assumes that one measured liter of unmilled grain weighs an average of 0.62 kg. Potter, “From Forest to Grassland”, Table 1, presents harvest figures that also suggest a generally higher productivity for forest fields in this region. The grassland field yields are comparable to those in Sherman, “What ‘Green Desert’?”, p. 131.

⁵⁶ Ibid., pp. 133–34, reports that based on his survey of the literature, most cultivators of grassland also cultivate some other type of land as well. In Indonesia such “mixed” system of agriculture is common. Compared to nonmixed systems, it is ecologically more stable and permit greater intensification of labour. See Dove, “Man, Land and Game in Sumbawa”, p. 120.

⁵⁷ This assumes an average yearly harvest of 1550 kilograms of rice, and it assumes a market value for rice of 16 cents U.S./kg., based on the prevailing price of 100 rupiah per kilogram in 1980 in the Riam Kanan valley, and the then prevailing exchange rate of 630 rupiah to the U.S. dollar.

⁵⁸ See Banijbatana, Dusit, “Forest Policy in Northern Thailand”, in Farmers in the Forest: Economic Development and Marginal Agriculture in Northern Thailand, ed. Kunstadter, Peter et al. (Honolulu: East-West Center, 1978), pp. 54–60 (esp. 58)Google Scholar.

⁵⁹ Cf. Peter Kunstadter, “Alternatives for the Development of Upland Areas”, in ibid., pp. 289–308 (esp. 298).

⁶⁰ Neither did the author observe any significant damage done to standing forest by the burning of cleared and dried forest swiddens, despite the much greater biomass and hence amount of combustible material involved. As noted earlier in the text, the climate of the Riam Kanan valley makes the Banjarese worry more about fires that do not burn well enough, than fires that burn too well or travel too far. Cf. Dove, , Swidden Agriculture in Indonesia, pp. 132, 137Google Scholar.

⁶¹ Holm, et al. , The World's Worst Weeds, p. 70Google Scholar. See also Hamilton, , Tropical Forested Watersheds, pp. 76, 79, 129Google Scholar; and Soepardi, Goeswono, “Alang-Alang (Imperata Cylindrica [L.] Beauv.) and Soil Fertility”, in Proceedings of Biotrop Workshop on Alang-Alang (Bogor: BIOTROP, 1980), pp. 57–69 (esp. 63)Google Scholar.

⁶² For example, the stocking rate (of cattle and water buffalo) in the grassland around Rantau Balai was estimated in 1980 to be just 32 head/km².

⁶³ In fact, even some natural forests fail to check erosion as well as Imperata grasslands. See Gibson, “Toward a Stable Low-Input Highland Agricultural System”, p. 379.

⁶⁴ Hamilton, , Tropical Forested Watersheds, p. 110Google Scholar.

⁶⁵ Ibid., p. 123, and passim. Specifically regarding Imperata grasslands, see Gibson, “Toward a Stable Low-Input Highland Agricultural System”, pp. 379–81. The higher water yields expected from grassland can be expected from cropland as well. See Hamilton, , Tropical Forested Watersheds, pp. 93, 102Google Scholar.

⁶⁶ Hamilton, , Tropical Forested Watersheds, p. 70Google Scholar.

⁶⁷ Ibid., p. 119.

⁶⁸ Musa, et al. , “Ekologi Kawasan Waduk Riam Kanan”, p. 9Google Scholar.

⁶⁹ Cf. Dove, , “Man, Land, and Game in Sumbawa”, p. 119Google Scholar.

⁷⁰ Musa, , “Ekologi Kawasan Waduk Riam Kanan”, p. 4Google Scholar.

⁷¹ Ibid., pp. 4, 17.

⁷² The valley was still populated as of early 1985.

⁷³ Hamilton, , Tropical Forested Watersheds, p. 131Google Scholar.

⁷⁴ Barlow, Colin, The Natural Rubber Industry: Its Development, Technology, and Economy in Malaysia (Kuala Lumpur: Oxford University Press, 1978)Google Scholar, Table 7.16, estimates an average gross yield of 1193 kilograms per hectare per year from high yielding rubber. At the 1980 market price of rp. 500/kg. for smoked rubber in the Riam Kanan valley, this represents a return of $8,521 in nine years (again based on 1980 exchange rates), compared with the earlier calculated $338 for a pine plantation over the same period of time.

⁷⁵ Such conflicts of interest (e.g., between rural peasants and urban industrialists) have been identified as a pervasive source of difficulties in river basin development in Africa. See Scudder, Thayer, “River Basin Development and Local Initiative in Africa”, in Human Ecology in Savanna Environments, ed. Harris, David R. (London: Academic Press, 1980), pp. 383–405Google Scholar.

⁷⁶ The dean of Indonesian scholars of swidden agriculture has recently concluded as well that “No resettlement is needed” in order to develop swidden agriculturalists. See Hardjosoedirdjo, Soedarwono, “Institutional Aspects of Improvement in Shifting Cultivation”, in Improved Production Systems as an Alternative to Shifting Cultivation (Rome: FAO Soils Bulletin No. 53, 1984), pp. 109–115 (esp. 114)Google Scholar. The same conclusion is reached by Hamilton, , Tropical Forested Watersheds, p. 21Google Scholar.

⁷⁷ See Chang, Jen-Hu, “The Agricultural Potential of the Humid Tropics”, The Geographical Review LVIII, 3 (1968): 335–61Google Scholar.

⁷⁸ See Suryanata, , “Perkembangan Agroekosistem di Daerah Transmigrasi”, p. 99Google Scholar.

⁷⁹ See Andrews, , “Imperata Cylindrica in the Highlands of Northern Thailand”, p. 387Google Scholar.

⁸⁰ See Holmes, et al. , “The Use of Imperata Cylindrica (L.) Beauv. by Grazing Cattle in Papua-New Guinea”, pp. 179, 190–91Google Scholar.

⁸¹ This inevitably leads to some rather schizophrenic discussions of the Imperata “problem”. For example, Soerjani, M. et al. , “Imperata Research and Management in Indonesia”, Mountain Research and Development 3, 4 (1983): 397–404CrossRefGoogle Scholar, write that Imperata grasslands are a “wasteland”, then they admit that the nutritive value of Imperata as fodder “appears not to be inferior to that of napier grass”. They state both that the soil under Imperata “is considered infertile”, and also that Imperata “may occur in all kinds of soil”. They write that “It has poor value in reducing runoff and adds to the risk of erosion and flooding”, and also that it “can be useful as a ground cover to preserve soil condition, prevent soil erosion, and consolidate dikes and dams”. Finally, they dramatically conclude that “The rehabilitation of alang-alang dominated areas has, therefore, become the first priority in agricultural management in Indonesia”, while elsewhere they note that some Imperata grasslands are actively cultivated (for thatching and roofing), others are converted to agricultural use with “satisfactory results” by simple tillage, and finally, some revert to secondary forest on their own, by “natural succession”.