Farmers’ strategies for managing acid upland soils in Southeast Asia: an evolutionary perspective

Agriculture, Ecosystems and Environment 106 (2005) 69–87 www.elsevier.com/locate/agee

Farmers’ strategies for managing acid upland soils in Southeast Asia: an evolutionary perspective R.A. Cramb* School of Natural and Rural Systems Management, University of Queensland, Brisbane, Qld 4072, Australia Received 14 October 2003; received in revised form 29 April 2004; accepted 12 July 2004

Abstract The acid soils of the uplands of Southeast Asia have resisted intensive agricultural use for centuries. In recent decades, however, due to rapid population growth, escalating market demand for agricultural produce, and government policies for land development and settlement, the acid uplands have become the focus of more intensive land-use systems, placing greater demands on farmers and requiring the development and dissemination of improved practices for soil management. In order to develop appropriate soil management technologies and plan effective interventions to facilitate their adoption, it is important to understand the goals and circumstances of farmers in the acid uplands, the range of farming systems they have developed, and the variety of socio-economic factors and trends influencing the evolution of these farming systems. Building on Boserup’s model of agrarian change, an evolutionary framework is developed and applied to five case studies: a long-fallow (shifting) cultivation system in Sarawak, Malaysia; a short-fallow system in South Kalimantan, Indonesia; a continuous cropping system in Bukidnon, Philippines; a tree crop (with intercropping) system in Southern Thailand; a livestock grazing system in Daclac, Vietnam. The framework provides a useful tool to interpret and categorise farmers’ evolving soil management strategies and to plan more effective soil management research and interventions. # 2004 Elsevier B.V. All rights reserved. Keywords: Farming systems; Acid upland soils; Intensification; Commercialisation; Soil management; Southeast Asia

1. Introduction The acid soils of the uplands of Southeast Asia, accounting for 315 million ha or almost two thirds of the total land area of the region, have been brought into intensive agricultural use in comparatively recent times (Von Uexkull and Mutert, 1995). To successive * Tel.: +61 7 3365 2967; fax: +61 7 3365 9016. E-mail address: [email protected].

groups of immigrants and colonisers, the prolific tropical rainforest vegetation, itself a formidable barrier to agricultural development, belied the inherent infertility and fragility of the underlying soils once cleared. Hugh Low, a 19th century botanist and colonial administrator, wrote of insular Southeast Asia: ‘‘Though the vegetation of no country in the world is so luxuriant as that of the Eastern islands, it has been proved by many writers that the soil of some of them is not so fertile as the appearance of the forests

0167-8809/$ – see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.agee.2004.07.011

70

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

would lead the cultivator to expect’’ (Low, 1848: 32). Commenting on the extent of acid tropical soils in Sumatra, Kalimantan, and Papua, where tens of thousands of Javanese and Balinese transmigrants have been settled, Rumawas also remarks that ‘‘. . .the [Indonesian] archipelago as a whole is not as fertile as one is led to believe’’ (Rumawas, 1987: 205), suggesting that the luxuriant natural vegetation has continued to deceive. For centuries indigenous farmers relied on shifting cultivation systems with long-fallow periods to produce food crops for their subsistence from these acid soils. From the late 19th century, European capitalist farmers established perennial tree–crop plantations (notably rubber after 1900), using impoverished labourers from densely populated India, southern China, and Java to clear the forest and work their estates (Drabble, 2000). Smallholders too began to supplement their shifting cultivation plots with perennial tree crops (Cramb, 1988). To a significant extent these tree crops mimicked the function of the original forest in conserving and cycling nutrients and protecting the topsoil. Importantly, the major plantation crops – rubber, oil palm, coffee and other tree crops – are well adapted to acid soil conditions (Myers and De Pauw, 1995). In recent decades, however, due to rapid population growth, escalating market demand for agricultural produce, and government policies for land development and settlement, the acid uplands have become the focus of more intensive land-use systems, placing greater demands on farmers and requiring the development and dissemination of improved practices for soil management (Myers and De Pauw, 1995; Dierolf et al., 2001). These highly weathered acid soils (58% of which are classified as Ultisols or Red-Yellow Podzolic Soils) are generally low in available P and exchangeable bases such as K and Ca, and have low base saturation, low cation exchange capacity (CEC), and high levels of (potentially toxic) Al, Fe and Mn (Pushparajah and Bachik, 1987; Craswell and Pushparajah, 1989; Willett, 1994; Von Uexkull and Mutert, 1995). Following clearing and continuous cultivation with annuals, they rapidly lose organic matter and nutrients and are susceptible to compaction and erosion. To reclaim them for agricultural use chemical inputs are recommended, especially lime and P fertilisers.

Organic amendments, such as farmyard manure, cover crops, green manures, and compost, are also advocated to help ameliorate acid soil conditions and build up depleted soil organic matter (Lefroy et al., 1995; Whitbread and Blair, 1999). In addition, on sloping upland soils, conservation measures such as contour hedgerows of shrub legumes are promoted, both to reduce soil erosion and to provide an additional source of nutrients (Cramb et al., 2000). All these soil management practices have implications for the farm-household economy (Moran, 1987). While there has been considerable technical research in Southeast Asia on the management of acid soils for sustainable crop production, much of it under the auspices of the Australian Centre for International Agricultural Research (ACIAR) and the Acid Soils Network established by the former International Board for Soils Research and Management (IBSRAM) (Craswell and Pushparajah, 1989; Date et al., 1995), adoption of recommended practices by farmers, particularly of organic amendments, has often been disappointing. Increasingly it has been recognised that, in order to develop appropriate soil management technologies and plan effective interventions to facilitate their adoption, it is important to understand the goals and circumstances of farmers in the acid uplands, the range of farming systems they have developed, and, in particular, the socio-economic factors and trends influencing the evolution of these farming systems (Cramb, 2000). In this paper, case studies of five different farming systems in the acid uplands are examined to assess their adaptation over time to the problems associated with acid soils. The case studies are mostly drawn from locations in the vicinity of ACIAR–IBSRAM research sites over the past decade. The case studies are examined within an evolutionary framework, one that builds upon the initial insights of Boserup (1965) in her influential essay on the economics of agrarian change under population growth. Boserup’s analysis of the population-induced intensification of farming systems in a subsistence setting is extended by considering the influence of increasing market access and consequent agricultural commercialisation. The objective is to build and test a framework that can aid research and extension for improved acid soil management.

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

2. An evolutionary framework A basic assumption of farming systems research is that farmers are intentionally rational in the way they manage their farming operations, including their choice of technology (Ruthenberg, 1980). That is, they choose farming technologies in order to further their goals, subject to the constraints imposed by resource availability (land, labour, and capital) and environmental conditions (bio-physical and socioeconomic). This underlying rationality is reflected in the evolution of farming systems as circumstances change. Three broad sets of circumstances are considered here – population density, market access, and government policies. 2.1. Population, resources and land-use intensity Boserup (1965) shows how subsistence farming systems have evolved historically so as to intensify land use as population pressure on the land increases (Table 1). Forest-fallow systems evolve into bushfallow, then short-fallow systems, and finally annual or even multiple cropping. Each stage requires a greater labour input per hectare and per worker, therefore a transition from one stage to the next only occurs when the growth of population makes it necessary to increase total food output from a given area of land. In the course of this evolution, traditionally available technologies are adopted sequentially to suit the requirements of each stage. For example, under shortfallow systems the plough is adopted to cope with the grasses in the fallow that were previously shaded out by forest regrowth. Techniques of fertility management are also related to the fallow system in Boserup’s schema. Whereas forest and bush-fallow systems rely on the ash from the burned fallow vegetation to

71

fertilise the soil, short-fallow cultivators depend almost exclusively on manure from their draught and other domestic animals, while in annual and multiple cropping systems, additional sources of fertilisation are needed, including green manuring, marling, and composts. (The option of purchasing inorganic fertilisers is outside the scope of Boserup’s analysis, but is considered below.) Pingali and Binswanger (1987) confirm and extend Boserup’s analysis, taking into account the characteristics of different soil types and agroclimatic zones, including the problem of soil degradation. For the humid tropics they argue that, historically, as population density increases, farming spreads from the more easily cultivated uplands, where shifting cultivation is traditionally practised, towards the naturally swampy bottomlands which, although they require a greater labour input, provide higher returns to investment in land improvement, especially in the case of irrigated rice cultivation. That is, the bottomlands are the least preferred under low population densities (partly for health reasons) but become the focus of intensification efforts as population density reaches high levels. However, where bottomlands are limited in extent or population growth exceeds their capacity to absorb further labour inputs productively, intensive cultivation is gradually extended into the middle and upper slopes, at the expense of grasslands and forest. In the Philippines, for example, the intensification of the lowlands has indeed begun to level off and the failure of industrialisation to absorb surplus labour has led to a major migration of subsistence farmers into the uplands in recent decades (Garrity et al., 1993; Cramb, 1998). The implication of Pingali and Binswanger’s (1987) analysis is that this is the least preferred option of the farmers concerned because the

Table 1 Boserup’s sequence of farming systems and technologies Farming system

Cropping period

Fallow period

Cultivation technologies

Soil fertility management

Forest-fallow Bush-fallow Short-fallow Annual cropping

1–2 2–6 1–2 6–9

20–25 years 6–8 years 1–2 years 3–6 months

Axe, fire Axe, fire, hoe Plough Plough

Ash fertilizer Ash fertilizer (augmented)a Manure Manure, green manure, compost, marling, forage crop, irrigation

Multiple cropping

10–12 months

0–2 months

Plough, irrigation

a

years years years months

In some bush-fallow systems vegetation for burning is cut and transported from outside the immediate farm area.

72

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

returns to agricultural investment and intensification in the uplands are low, resulting in a low-yield, lowinput equilibrium. This means that the returns to investment in soil conservation and improvement are also likely to be low, increasing the risk that serious soil degradation will occur. The problem is compounded when tenure in the uplands is insecure and the adjacent grasslands and forestlands are effectively subject to open access, resulting in overgrazing and overcutting, as well as the unregulated upward extension of the cultivation frontier. The evolutionary perspective provided by Boserup (1965) and Pingali and Binswanger (1987) suggests that the adoption of improved soil management technologies by upland farmers will be a function of the degree of intensification of their farming system. Raintree and Warner (1986) analyse the development of agroforestry technology in relation to the intensification of shifting cultivation systems and examine the different agroforestry options that open up at different stages of intensification. Such analysis suggests, for example, that for shifting cultivation systems, research on managed forest-fallows is likely to be more useful than research on more intensive alley cropping or contour hedgerow technologies, which are only likely to be adopted by farmers practising permanent cropping (Garrity et al., 1993). Further analysis of this kind is needed to tailor recommended fertility management practices to the intensity of land use. 2.2. Market access and agricultural commercialisation In addition to population-induced intensification of farming systems, a second major evolutionary trend can be discerned. The growth of market opportunities resulting from general economic growth and improvement in rural infrastructure, and the growing awareness within rural communities of these opportunities, induces the commercialisation of previously subsistence-oriented farming systems (Myint, 1973; Fisk, 1975; Ruthenberg, 1980; Pingali, 1990, 1996). This may reinforce the tendency to more intensive production of traditional food crops in order to produce a marketable surplus (e.g., in the irrigated rice-growing areas of Thailand, Vietnam and the Philippines), or it may result in a shift in farming

systems towards different crop and livestock products to satisfy urban or export demand (e.g., rubber in the uplands of Malaysia, Indonesia, and Southern Thailand; coffee in the Central Highlands of Vietnam; fruit, vegetables and timber in the Philippine uplands; cattle in some of the outer islands of Indonesia). The commercialisation process also affects the use of purchased inputs due to lower farm-gate prices and more widespread availability of seed, inorganic fertilisers, pesticides, farm equipment, and credit (Tomich et al., 1995). Such inputs can ease the constraint imposed by scarce household resources at various stages of intensification. For example, the use of herbicide or machinery can save labour in low intensity (labour-scarce) situations, and high-yielding seeds and fertilisers can economise on land in highintensity (land-scarce) situations (Ruttan and Hayami, 1998). The increased value of commercial production would, in general, be expected to provide additional incentive to invest in soil and land improvement, as well as providing the additional cash needed for such investment. However, the impact of commercialisation on soil management could be positive or negative, depending on the characteristics of the crop and livestock enterprises that are subject to increased commercial demand. For example, whereas an expansion of commercial fruit and timber production may help conserve the soil, an expansion of commercial maize or vegetable production on sloping lands may lead initially to greater erosion (Coxhead and Buenavista, 2001). At the same time, the higher opportunity cost of farm labour resulting from improved access to urban labour markets may actually discourage labour-intensive conservation practices (Pandey and Lapar, 1998), including, for example, the use of composting or farmyard manure, or even single-nutrient as opposed to compound fertilisers. Indeed, farmers often find themselves caught in a double-bind which they express as not having enough time or working capital to undertake recommended farm improvements such as soil conservation measures – the shortage of cash income forces them to work off-farm, where cash-earning opportunities are greater, thus limiting their available time for on-farm improvements (Cramb et al., 2000). Moreover, the markets for the agricultural inputs needed for improved soil and crop management

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

develop unevenly and are often incomplete. For example, the market for high-yielding seed of outpollinating crops such as maize is quick to develop because of the short-term financial gains to suppliers, but the commercial supply of improved planting material for tree crops tends to lag behind. The market for credit is typically confined to seasonal credit for inputs such as fertiliser, secured by a lien over the growing crop, while credit for longer-term investments in soil rehabilitation and improvement or tree– crop establishment is lacking. Thus the low-yield, low-input equilibrium, described above for subsistence production in the acid uplands, may persist even though returns to land improvement in a developing market economy are now high. This is because such improvement requires significant up-front investment in lime, P fertiliser, and soil conservation measures to get the soil to a stable and productive state. Where markets are incomplete, farmers often lack the knowledge, cash, credit, or availability of supplies to undertake this prior investment. Dierolf et al. (2001) describe this in terms of a poverty cycle in the acid uplands in which low yields result in low income, hence a low level of inputs, thus reinforcing low yields. Analysis of such market failures may point to a role for community-based measures or government involvement. 2.3. Agricultural development policies Superimposed on these two major forces for landuse change – population growth leading to intensification of food crop production, and expanding markets leading to commercialisation of farming systems – a range of government policies and programs have affected the evolution of farming systems in the tropical uplands. Whereas pre-colonial agricultural organisation largely took the form of village-based smallholdings, colonial governments adopted land policies that in many cases (e.g., Philippines, Vietnam, Malaya) enabled accumulation of large tracts of more productive and better-located lands by a favoured elite, whether foreign or indigenous (Elson, 1997). This gave rise to a dualistic or bimodal agricultural structure, with a small number of large-scale, capitalintensive, and commercially oriented farms alongside numerous small-scale labour-intensive, and subsistence-oriented farms (Johnston and Kilby, 1975;

73

Tomich et al., 1995). The technical and economic requirements and evolutionary pathways of these two sectors have been markedly different. For example, the plantation sector, with its access to long-term credit and technical expertise, has been able to invest profitably in the development of acid soils through large initial inputs of lime and P fertilisers, overcoming the low-yield, low-income, low-input trap described above (Thomas Fairhurst, pers. commun.). In the post-colonial period dualistic structures have often persisted, notwithstanding a variety of reform programs. Agricultural development strategies in the post-colonial period can be grouped according to whether they tend to reinforce a dualistic agriculture or emphasise widespread development of rural smallholders. Barlow and Jayasuriya (1984) make the useful distinction between focus strategies, which concentrate the state’s resources on a small number of large-scale agricultural entities (including private plantations, state farms, resettlement schemes, and nucleus estate schemes), and dispersal strategies, which spread resources more widely to benefit the mass of smallholder farm families (e.g., through rural feeder roads, research and extension directed to smallholder needs, decentralised rural credit programs, and planting schemes for tree crops). Welldesigned dispersal strategies can help to compensate for incomplete or non-existent markets for the key inputs needed for better soil management. Tomich et al. (1995) present strong arguments, on both efficiency and equity grounds, for the dispersal approach. These arguments are captured in the old Yorkshire farming adage (appropriate for an audience of soil scientists and plant nutritionists): ‘‘Muck be no good except it be spread’’. In recent decades smallholder-based strategies have been adopted to varying degrees in many countries (e.g., the de-collectivisation of agriculture in Vietnam; agricultural extension programs in Indonesia, Malaysia, and Thailand; the Comprehensive Agrarian Reform Program (CARP) in the Philippines). Nevertheless, other government policies have often continued to work against the interests of smallholders in the uplands, such as a protected domestic manufacturing sector (including fertiliser manufacturing), overvalued exchange rates (which make agricultural exports less competitive and imported fertilisers more expensive), high taxes on

74

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

Fig. 1. Hypothesised relationship between population density, market access, and fertility management for acid upland soils.

smallholder exports, inadequate investment in rural infrastructure in upland regions, and expensive, topheavy, often coercive policies for land settlement and development. Hence it is often difficult to assess the net impact of government policies and programs on the evolution of farming systems in the uplands. 2.4. Framework and hypotheses Focusing on the smallholder sector, Fig. 1 depicts in simplified form the hypothesised relationship between population density, market access, and fertility management techniques for acid upland soils. In practice, the response to these two main drivers will also depend on the specific agro-ecological setting (especially soils, slope, and rainfall), as well as local policy settings. However, it is useful to abstract from these influences initially in order to bring out the underlying economic rationale of soil management choices. Considerations of agro-ecology and policy are re-introduced in the case studies that follow. At low population densities and with poor market access (bottom-left sector), farmers typically rely on shifting cultivation to extract a subsistence income from acid upland soils, whether on level or sloping land. The slash-and-burn technique of field preparation reduces acidity and adds a temporary supply of plant-available nutrients, sufficient for a year or two of cropping, after which a long period of forest-fallow is needed to accumulate nutrients in the forest biomass and shade out weed species (Sanchez, 1974, 1994).

This technique is suited to a situation in which labour is scarce relative to land, there is little or no cash flow, and the cost of purchased inputs is high. At high population densities and with poor market access (bottom-right sector) it is hypothesised that farmers more readily adopt Boserupian soil management practices, relying on large ruminant livestock not only for draught power but as a source of manure. That is, by incorporating livestock and utilising crop residues and other sources of biomass, farmers draw mainly on an endogenous supply of nutrients and ameliorants. These techniques are suited to a situation in which labour is abundant relative to land, but cash flow is still a constraint and purchased inputs are costly and difficult to obtain. At low population densities and with good market access (upper-left sector), commercial crops and livestock become viable, often in combination with traditional shifting cultivation systems. The commercial crops include tree crops such as rubber and coffee that are relatively tolerant of acid soil conditions. The addition of commercial activities generates the cash flow that permits the purchase of exogenous nutrients, including purchased lime, NPK fertilisers, and trace elements, as well as labour-saving inputs, in particular herbicide. It is hypothesised that this reliance on exogenous nutrients develops where labour is still scarce relative to land but the ability to access purchased inputs has been enhanced. At high population densities and with good market access (upper-right sector) integrated use of endogenous and exogenous nutrients, such as recommended by soil scientists (Whitbread and Blair, 1999; Gruhn et al., 2000), becomes feasible. The relative abundance of labour facilitates the use of labourintensive fertility management techniques, such as composting or the collection and spread of farmyard manure, while the high productivity and reduced cost of purchased fertilisers encourages their use. The possibility of purchasing organic fertilisers such as chicken manure from commercial suppliers also increases the chances of integrated soil fertility management. However, note that, for a given, relatively high population density, further improvement in market access can induce a shift in the farming system from one based on integrated nutrient use to one relying more on exogenous nutrients (i.e., crossing the diagonal boundary between the upper-

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

75

Table 2 Characteristics of case study sites Location

Sarawak, Malaysia

South Kalimantan, Indonesia

Bukidnon, Philippines

Southern Thailand

Daclac, Vietnam

Latitude Altitude (m) Topography

28N 30–350 Steeply sloping

48S <100 Undulating

Rainfall (mm) Dry months (<50 mm) Population density (persons/km2) Market access

3000 0 15–20 Poor

2500 2–3 160 Intermediate

88N 500 Undulating to sloping 2500 2–3 200 Good

78N <100 Undulating to sloping 2400 0 115 Good

128N 500 Undulating to sloping 1900 4 50 Intermediate

Note: The soils at all sites are ultisols of pH 4.5–5.5 and low to moderate fertility.

right and upper-left sectors), due in part to the increased opportunity cost of farm labour. The following case studies test the utility of the above framework by examining the ways in which smallholder farming systems have been adapted to the combined influences of population growth and movement (spontaneous and planned) and a developing market economy, taking into account the influence of local agro-ecological conditions and government policies towards smallholders. The case studies include a long-fallow system in Sarawak, Malaysia; a short-fallow system in South Kalimantan, Indonesia; a continuous cropping system in Bukidnon, the Philippines; a tree–crop system with intercropping in Southern Thailand; a livestock grazing system in Daclac, Vietnam (Table 2). Though a comparison between the case studies broadly illustrates the effects of differing population densities and access to markets, in line with the evolutionary model outlined above, the main emphasis in what follows is on the particular circumstances influencing the choice of evolutionary pathways within each case.

3. Case studies 3.1. Long-fallow (shifting) cultivation in Sarawak, Malaysia Iban shifting cultivators began settling in Sarawak as early as the 16th century. Having progressively cleared the rainforest for hill rice cultivation they allowed the land to revert to secondary forest which they re-cleared generation after generation in a stable forest-fallow cycle. This farming system was sustain-

able, providing the cultivators with their subsistence needs and, in good years, a small surplus for trade, until population growth in recent decades began to place the system under increasing pressure (Cramb, 1988). A long-term study was conducted of 46 farmhouseholds in two Iban villages in the Saribas District – Batu Lintang and Nanga Tapih – beginning in the 1979–1980 farming year and continuing through the 1980s and 1990s (Cramb, 1984, 1989, 1991, 1993). (This case study had no connection with ACIAR– IBSRAM research sites.) In 1980 the population density in the villages was 14 and 18 persons/km2 respectively, representing the lower end of the population density spectrum considered in this paper. The villages were remote from the main market centre of Kuching, which was distant 250 km by road. However, Batu Lintang had been linked to the main trunk road since 1980, and though residents of Nanga Tapih still relied on river transport, their journey to the nearest road had been substantially reduced. The study villages were located approximately 28N and 1128E. The physical environment of the villages has been described by Sibat (1980). The terrain consists of moderately to very steep dissected hills and mountains, with elevation ranging from 30 to 350 m a.s.l. The soils are mainly Red-Yellow Podzolic Soils of the Merit family – shallow, highly leached, acidic, red to yellow clay-loams and clays of low fertility. (According to Tie (1982): 86, the Merit Family largely corresponds to Typic Paleudults in the USDA Soil Taxonomy.) The climate is equatorial, with rainfall averaging over 3000 mm. Rainfall peaks in December–January during the north-east monsoon and tails off in mid-year, but there is no true dry season as found for example in South Kalimantan (see below), hence

76

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

escaped fires are rare and limited in extent. The natural vegetation of the area is Lowland Mixed Dipterocarp Forest. However, all but very small pockets of the original forest have been felled, giving rise to a patchwork of rice farms, rubber and pepper gardens, and various stages of secondary forest regrowth. The postwar growth in population in the two villages had led all farmers to shorten the fallow period for hill rice farming, but not below a minimum of about 7 or 8 years, below which farmers considered the land could not recover sufficiently to make cropping worthwhile. Hence in the 1960s and 1970s, when the prices of cash crops (rubber and pepper) were low, most farmers responded by borrowing or renting bunded, rainfed rice-land in the lower portion of the Saribas basin, a day’s journey or more by river from their own villages. From the late-1970s, due to an upsurge in pepper prices, they returned to their upland environment and concentrated their efforts on perennial cash crops, particularly pepper, for which they received technical and material support for crop establishment from the Department of Agriculture through a planting grant. From this point, pepper cultivation gained in importance while shifting cultivation of rice declined. In 1979–1980, 44 farmers from the two villages were surveyed throughout the farm year. The shifting cultivation system involved the alternation of brief (mostly 1 year) periods of cropping with longer (on average 10-year) periods of fallow. Small fields of 1– 3 ha were cleared, using bush-knife and axe, and the vegetation burned, thereby clearing a surface for planting while ameliorating the acid soil conditions and providing a fertilising effect. (Andriesse and Schelhaas (1987), in a study of the fertilising effects of burning in a similar shifting cultivation system in Sarawak, found large increases in available P and in both exchangeable and extractable Ca, Mg and K in the topsoil, accompanied by a corresponding increase in pH.) If the burn was poor the remaining debris was often stacked in heaps and re-burned. With no further field preparation, the rice was direct-seeded using a dibble-stick. Maize, cassava, and numerous vegetables were intercropped with the rice. A small dose of di-ammonium phosphate (DAP), sold at a subsidised price by the government for seed dressing, was in fact usually applied as a top dressing at an average rate of 16 kg/ha (though farmers only

applied it to the less fertile parts of their farm). One round of manual weeding was performed, frequently supplemented by the spraying of paraquat (which reduced the overall labour requirement for weeding by 40%). Pest and disease control was sporadic and largely ineffective. Harvesting was done with a small hand-held blade and post-harvest operations were performed manually. The average labour input in 1979–1980 was 209 days/ha (a third of which was for weeding). The mean yield of unhusked rice was 525 kg/ha, reflecting a generally poor season. Over six seasons the mean yield ranged from about 500 to 1000 kg/ha. On average, survey households produced 690 kg of unhusked rice, which was 72% of estimated requirements. In 1979–1980 pepper was cultivated in small gardens of 0.2–0.3 ha, with a labour input of around 900 days/ha and intensive use of fertiliser (over 1000 kg/ha of compound NPK fertilisers), including the application of dolomitic lime at planting to raise soil pH. There was no use of recommended leguminous cover crops. Expenditure on fertiliser and agrochemicals for pepper absorbed 10–20% of total farm cash receipts. The yield of black pepper was 2000–3000 kg/ha, giving a gross margin per hectare 10 times that of hill rice and a gross margin per day twice that of hill rice. At Batu Lintang pepper absorbed 26% of household labour and contributed 29% of gross farm income; at Nanga Tapih, where shifting cultivation had been under more pressure, pepper absorbed 35% of household labour and contributed 61% of gross farm income. Essentially, farmers in these two villages had become intensive, commercial pepper growers, relying solely on purchased inputs for fertility management, with shifting cultivation a ‘‘spare-time’’ activity fitted into the slack periods for pepper. A recent field visit (June 2001) indicated that this trend had continued through successive pepper booms and busts, with most farmers no longer practising shifting cultivation at all. 3.2. Short-fallow cultivation in South Kalimantan, Indonesia The Indonesian Government’s transmigration program resettled farm families from the densely

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

populated islands of Java and Bali to the more sparsely populated outer islands (Babcock, 1986). Many of the resettlement sites have problem soils, placing severe constraints on farm productivity and requiring settlers to adapt their farming practices (Rumawas, 1987). Transmigration settlements have been established on acid soils in various parts of South Kalimantan, where the indigenous Banjarese farmers practised shifting cultivation. South Kalimantan has abundant rainfall (2000–3000 mm annually) but a marked dry season of 2–3 months, resulting in seasonal drought conditions and the spread of fires, which encourage and maintain grasslands once the forest has been cleared. There are an estimated 960,000 ha of Imperata cylindrica grasslands in the province. The population in the province grew at an annual rate 1.9% in the 1980s but in Tanah Laut, the southernmost district, the growth rate was 4% due to the number of transmigration settlements established there in the previous three decades. The population density in the province as a whole was 65 persons/km2 in 1988, but in the transmigration villages it ranged from 150 to 250 persons/ km2. A survey was conducted in 1993 of 45 farm households in the Batutungku Transmigration Area in Tanah Laut, including 15 from each of three successively established transmigration villages (Masyhuri and Cramb, 1995). The location of the area is roughly 48S and 1158E, about 65 km by road to the southeast of the provincial capital Banjarmasin. The soil at the site is a Typic Paleudult with a pH of around 4.3. The terrain is undulating. The population density in the transmigration area was estimated to be 255 persons/km2 at the time of initial settlement but had fallen to 158 persons/km2 at the time of the survey. As in many transmigration sites, settlers were allocated 2.0 ha, comprising 0.25 ha of home-lot, 0.75 ha of cleared farmland (termed Dryland I) and 1.0 ha of uncleared farmland (termed Dryland II). All the allocated land was initially Imperata grassland of similar fertility; Dryland I, however, was closer to the village settlement than Dryland II. Many of the settlers had subsequently left the scheme, enabling those who remained to acquire more land. The average area owned had increased by more than 50%, the increase being greater the longer the period of settlement. On average settlers used 1.9 ha in the wet season and 1.7 ha in the dry season. That around 45% of the land

77

was unutilised was due to the increase in farm size since resettlement. The enlarged area could not all be farmed with the available labour. In any case, farmers preferred to fallow their farming areas to help maintain soil fertility and control insect pests. Typically, a given area would be cropped continuously for 2–3 years before being rested for a similar period. Dryland I tended to be used more intensively than Dryland II because of its closer proximity to the village. The major cropping pattern involved two successive crops of maize. Groundnut, upland rice, and cassava were also cropped in varying combinations. Some farmers cultivated roselle, an annual fibre crop. Tree crops and other perennials (coconut, clove, jackfruit, banana, bamboo, etc.) were also planted, mainly in the home garden but also to some extent in Dryland II. In addition to cropping activities, most farmers had cattle that were grazed on the Imperata grasslands as well as being stall-fed with grass and crop residues. The normal cultivation practice for maize was to burn the Imperata grass, then plough and harrow, using a pair of draught animals. Fertiliser inputs for maize averaged 125 kg/ha of urea, 130 kg/ha of triple super phosphate (TSP), and 75 kg/ha of KCl, supplemented with 840 kg/ha of farmyard manure (1340 kg/ha in the wet season and 345 kg/ha in the dry season). There was no use of Imperata compost (one of the experimental practices examined by soil scientists in the area). The yield of maize averaged 2600 kg/ha. Farm income accounted for 75% of total household income (the rest came mainly from nonfarm wage work). Farm income was evenly divided between crop and livestock income. On average, maize was the major source of crop income. Thus the Javanese transmigrants, with limited labour and capital, had developed a compromise between the intensive cultivation practices of their place of origin and the extensive, shifting cultivation practices of the local population. The outcome was a short-fallow system that, as with shifting cultivation, made use of fire to help prepare the soil for cropping, but was crucially dependent on cattle for draught, dung and dollars – i.e., for cultivation (to deal with the grassy weed problem); to supply manure to the cultivated fields; as a direct source of income. The system also depended on the generation of enough cash to purchase inorganic fertilisers, in order to

78

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

escape the low-yield, low-input equilibrium characteristic of the acid uplands. According to Azwar (pers. commun., July 2001) the increased price of maize in recent years had reinforced this trend, leading to the expansion of maize production based on the use of hybrid seed and high inputs of purchased fertiliser. 3.3. Continuous cultivation in Bukidnon, Philippines The Bukidnon plateau in northern Mindanao was until recent decades a ‘‘spacious plateau of rolling grasslands’’ with few inhabitants (Edgerton, 1982). Historically, the ‘‘people of the mountains’’, the Bukidnon, practised shifting cultivation, in the process gradually pushing back the forest margins and creating the grasslands, which remained sparsely populated. In the 1920s and 1930s, Cebuanos and other lowland farmers moved south onto the plateau in increasing numbers until they became the dominant group (Edgerton, 1982). They practised maize cultivation using their traditional plough technology, at first taking up the flatter, more accessible land, but progressively moving onto sloping lands as well. Rainfall in Bukidnon averages 2540 mm, fairly evenly distributed throughout the year but with a short dry season lasting 2–3 months in the northern part of the province (Garcia et al., 1995). The terrain is undulating in the central plateau, phasing into steep mountains to the west and east. Over 80% of the province lies above 500 m. Most of the soils of the province are highly weathered, deep, reddish-brown to yellowish-brown clays. They are moderately fertile because of their largely igneous origin but crop production is limited by their moderate-to-strong acidity and P deficiency (DENR, 1993). The population density in 1990 averaged 102 persons/km2 (Garcia et al., 1995), but was closer to 200 persons/ km2 in the vicinity of the Sayre Highway, the major north–south route through the province. A survey was conducted in 1991 of 30 maize farmers in the village of San Jose near the provincial capital, Malaybalay (Mugot and Cramb, 1995). The area occupied by San Jose (approximately 88N and 1258E) straddles the Sayre Highway and includes both level and sloping uplands – 15 farmers were selected from each land type. The soils close to the highway are classified as Andic Haplohumults with a pH of 4.6–4.9

(Duque et al., 1995). Survey respondents had resided in San Jose for an average of 19 years; 93% had originated outside Bukidnon, reflecting the long-term southward migration described above. Farms were located up to 6 km from the highway, giving them reasonable market access to Malaybalay. Farm size averaged 2.4 ha and ranged from 1 to 5 ha. Almost all farmers (93%) had some form of title to their land, mostly through the Comprehensive Agrarian Reform Program (CARP). Maize was the dominant land use, though some had cultivated rice, mungbean or cowpea and some had land under sugarcane, coffee, cocoa, or bananas. Most earned occasional income from livestock sales (though cattle numbers were lower than in South Kalimantan) and most had a source of off-farm income, mainly wage-work on other farms. However, on average 72% of household receipts came from the sale of maize. Farmers generally planted two maize crops, the second being vulnerable to drought. At the start of the first crop season over half the farmers slashed their fields to clear them before ploughing, and 17% burned the slashed vegetation. Tillage typically involved ploughing, harrowing, and furrowing, using animal power (63%), tractor power (21%) or both. Most farmers with sloping land had adopted some form of contour farming due to a university extension project. Farmers were evenly divided between the use of a local white maize variety and hybrid maize, though the incidence of hybrid maize was higher among farmers with level land closer to the highway (80%) and was increasing. All farmers used inorganic fertilisers on their maize fields, including urea and compound NPK fertilisers. No use of organic fertilisers was reported. The average dose of N per season was around 60 kg/ha (similar to the rate reported in the South Kalimantan survey). Farms on level land planted with hybrid maize received double the fertiliser dose of farms on sloping land planted with local varieties. Sixty percent of farmers had applied lime to their maize fields within the previous 5 years (mostly 2–3 years previously) at an average rate of 1800 kg/ha. The proportion applying lime was higher for farmers with level land (87%) than for those with sloping land (33%). In the latter category, all those who did not apply lime said this was because the land had been recently opened up. All farmers left the maize stover in the field after

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

harvest, to be incorporated at the start of the next season. The maize cobs were mostly left at the shelling station or in the field, but 37% used them for household fuel. Composting was not practised. The output of maize in 1990 averaged 7000 kg per household, comprising 4200 kg from the first crop and 2800 kg from the second. The average yield was 1800 kg/ha, ranging from 2200 kg/ha from first season crops on level land to 1400 kg/ha from second season crops on sloping land. Most of the maize produced was intended for sale, though maize was also the major staple for households in the survey. Farmers identified a number of problems affecting the profitability of their maize enterprise, including the high cost of hybrid seed and fertiliser (73%), low maize prices at harvest (63%), and lack of capital to purchase inputs (17%). Thus economic problems tended to dominate soil management problems in this intensive, highly commercialised farming system. Most farmers saw the way to improve production in the future as the adoption of the commercial, seedand-fertiliser technology, including in some cases a greater use of lime. Apart from the virtually automatic incorporation of maize stover, technologies involving the use of on-farm organic fertilisers and amendments were not part of their plan. 3.4. Tree crop cultivation and intercropping in Southern Thailand Acid upland soils occupy over 3.5 million ha or 50% of the total area of Southern Thailand (Suthitipradit et al., 1995). The terrain is mountainous, phasing into low hills and undulating terraces and a narrow coastal plain. Rainfall averages 2400 mm and occurs throughout the year (Arbhabhirama et al., 1988). The natural vegetation is tropical rainforest, much of which has been cleared for tree crops. The population density was around 115 persons/km2 in the mid-1990s but varied widely between districts. Rubber, the major cash crop for the region, is tolerant of acid soil conditions and responds well to a balanced supply of nutrients (Pushparajah and Bachik, 1987), thus providing an economic alternative to annual crops for farmers in this environment. Rubber planting began in the largely Malay provinces of Southern Thailand early in the 20th century (Dixon, 1991: 89) and in 1990 rubber occupied 1.55 million ha, repre-

79

senting around 50% of cultivated land in Southern Thailand and 89% of Thailand’s total rubber area. Songkhla Province was the major rubber-producing region (Kiripat, 1991; Rubber Research Institute, 1992). Unlike in Malaysia where large estates and land development schemes dominated the rubber industry until recent decades (Drabble, 2000), most rubber growers in Southern Thailand are smallholders; about 95% have rubber holdings of less than 8 ha (Southern Studies Group, 1982). Though many rubber smallholders undertake other agricultural activities such as rice cultivation and the rearing of livestock, they are mainly dependent on rubber, which provides a low and unstable income. However, during the first 3 years of rubber establishment farmers have the opportunity to intercrop the young rubber trees with food crops, whether for their own consumption or for sale, thus supplementing their otherwise reduced income. Extensive areas of rubber are being replanted with support from the government’s replanting scheme, and some new areas are being planted. Many rubber smallholders practise intercropping with young rubber, using a range of crops and soil management practices. The task of managing acid soils within this context is somewhat different to the situation of shortfallow or continuous cultivation of field crops described above. A survey of 60 rubber farmers in three villages was undertaken in 1993 (Masae and Cramb, 1995). Two villages in Songkhla Province were selected to represent differing degrees of access to the main urban centre, Hat Yai. The first, Khuan Lang, was a peri-urban rubber-growing village, 5 km from Hat Yai, while the second, Pluk Nooh, was a relatively remote village, 63 km from Hat Yai. The third village, Thung Naree in Pattalung Province, though of intermediate distance from Hat Yai, was selected because it was well known for growing pineapple as an intercrop, its produce being marketed widely throughout the southern part of Southern Thailand and northern Malaysia. Respondents in each village were randomly selected from the population of rubber farmers who had replanted at least some of their rubber land within the previous 5 years. Most respondents (87%) regarded rubber production as their primary occupation, though in Thung Naree 40% were primarily rice farmers with rubber as their secondary occupation. Khuan Lang households,

80

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

with better access to Hat Yai, had more members involved in non-farm work than the other two villages. The average landholding size was 6.8 ha, of which 5.6 ha were under rubber. In each case some of this area was planted to young rubber (i.e., aged 3–5 years), the average being 2.5 ha or about 45% of the total rubber area. Almost half of the area under young rubber was intercropped. For those households who practised intercropping, the effective area under intercrops averaged 1.5 ha. Those who planted intercrops had on average 50% larger land holdings and 30% more land under young rubber than those who did not. Among farmers who did not grow any intercrops, the most frequent reason given (58%) was lack of time and household labour. Relatedly, in some cases their rubber plots were less accessible. Five different kinds of crop were grown as intercrops – upland rice, pineapple, maize, banana, and vegetables. Upland rice was the most common intercrop. The incidence was higher in Pluk Nooh, the more remote village (92%), compared with 40% in Khuan Lang and only 8% in Thung Naree, the pineapple village. Upland rice was widely grown in the past since it was the staple crop and many rubber growers did not have enough rice to meet their household’s needs. However, there had been a trend towards cash crops as market access improved. Pineapple ranked as the second most important intercrop, primarily due to the inclusion of Thung Naree in the survey. Maize and banana were grown by a few respondents in each village. Though vegetables of some kind were grown in all three villages, the highest incidence was in Khuan Lang because of the nearby market in Hat Yai. Various cropping patterns were practised. Upland rice, the most common crop, was grown once, twice or three times within the 3-year intercropping period. It was also grown in combination with other crops, especially maize, in different sequences. Maize was also grown with several kinds of vegetable and for up to 2 years. The perennials, banana and pineapple, were naturally grown as single intercrops throughout 3 years of intercropping, without any replanting. Intercroppers cited three reasons for their choice of crops: (1) they could be used for home consumption (e.g., upland rice); (2) they could be grown easily and needed little care (e.g., banana); (3) they could be sold easily in local markets (e.g., pineapple). Maize and

vegetables were grown for both home consumption and for sale. Most plots of young rubber included in the survey had been replanted rather than newly planted. Before replanting, the old rubber trees and shrubs were cleared using a tractor. The trees were sold to sawmills to be made into furniture and other products, such as wooden boxes. The remaining vegetation was commonly burned (78% of respondents), for two reasons. First, farmers believed that burning could enhance soil fertility for both the rubber seedlings and the intercrops. The second, more frequently cited reason was to speed the land-clearing operation in order to meet the planting deadline given by the government’s rubber replanting agency. Direct application of compound (NPK) fertiliser to intercrops was practised by 80% of the intercropping households, most of the remainder relying on the fertiliser applied to the young rubber. All the Thung Naree households applied fertiliser to their pineapple crop. The fertiliser rate varied from crop to crop, averaging about two 50 kg bags/ha for rice and maize, four bags per hectare for vegetables, and nine bags per hectare for pineapple. None of the respondents reported applying any lime; indeed, few of them realised that lime was useful for reducing soil acidity. Only two households used farmyard manure; no other form of organic fertiliser was used. Weeding was common only among pineapple and vegetable growers. Only vegetable growers practised pest control. Cumulative gross returns of over THB 10,000 (USD 220) per hectare were obtained for the rice– rice–rice cropping pattern (notwithstanding sharply diminishing yields with each successive year), the rice–maize pattern, the mixed vegetable pattern, and the continuous banana pattern. In contrast, pineapple intercropping generated a cumulative gross income of over THB 280,000 (USD 6000) per hectare, though this required an annual cash outlay of about THB 41,000 (USD 900) per hectare. Intercropping in the study area was thus one component of a commercially oriented tree–crop system. The strategy for soil management needs to be seen in this context. Farmers relied on increasing levels of purchased nutrients, the level depending on the commercial return to the intercrops. (The use of organic residues or a green manure crop was not encountered.) The three villages represented different

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

stages of intercropping development. Pluk Nooh was relatively remote and hence tended to follow the traditional practice, growing crops mainly used for home consumption, such as rice and banana, with a low level of purchased inputs. Khuan Lang was close to the expanding urban market of Hat Yai and focused more on cash crops such as maize and vegetables, with a higher level of inputs. Thung Naree represented the highest degree of commercialisation with its emphasis on intensive pineapple production for a regional market. 3.5. Livestock grazing in Daclac Province, Vietnam Acid upland soils account for 5.6 million ha or 95% of the Central Highlands of Vietnam (Nguyen and Thai, 1995). Two thirds of this land (3.5 million ha) is used for shifting cultivation by Montagnard minority groups. Extensive Imperata cylindrica grasslands have been created due to a combination of shifting cultivation, dry season fires, and the defoliation of forests during the war with the US. After 1975, farmers of the majority Kinh group from the Red River delta in northern Vietnam were settled in the region. The population density of the Central Highlands region was 50 persons/km2 in 1992 (Tran, 1998: 2). The post-1975 campaign to establish producer cooperatives resulted in membership of 65,000 in the Central Highlands by 1980, or 30% of farm households (Tran, 1998: 24–25). State farms were also established. However, following economic reforms initiated in 1986, farm households have increasingly been able to operate on their own account, utilising land allocated to them by the commune on a long-term basis. Farmers graze their own small herds of cattle (mostly around five head) on the native grasslands, and cultivate a range of field crops (lowland rice, maize, cassava, sweet potato) and tree crops (coffee, fruit trees, and timber species). M’Drak, the site of this case study, is located in Daclac Province in the eastern Central Highlands at 128N and 1098E and an altitude of around 500 m. M’Drak has an annual rainfall of 1890 mm with a marked dry season (<50 mm per month) of around four months. Soils are of pH 5.0–5.5 and low-medium fertility. The landscape comprises small areas of home gardens and bunded rice-fields with extensive areas of

81

native grasslands on the surrounding low hills. The area of upland crops and tree crops has been slowly expanding near houses and villages (Ibrahim et al., 2000). A study to monitor and evaluate the impacts of introduced forages was conducted in Cu’kroa Commune, M’Drak, in 2000, using a range of conventional and participatory techniques (Cramb and Purcell, 2001; Ho, 2002; Cramb et al., 2004). Some key findings of that study are reported here. Cu’kroa Commune comprised six villages with a total of 333 households (a mean of 55 per village) and a population of 1544 (a mean of 4.7 per household, including 2.9 workers and 1.8 dependent children). The livestock population included 1350 cattle (a mean of 4.0 and a range of 0–47 per household), as well as smaller numbers of buffalo, pigs, poultry, and fishponds. Farmers mostly grazed their cattle on common grasslands under commune administration, in some cases up to 3–5 km from their homesteads. In focus groups, farmers ranked their economic activities in order of importance (not necessarily correlated with the extent of land use). The first six activities were: rice, livestock, coffee, fishponds, labouring, and selling of manure (there was a demand for manure for coffee, grown within the commune and extensively in the adjacent state farm). The cultivation of root crops (cassava, sweet potato) was also mentioned by most participants but on average received a low score for importance, perhaps because of the low status of these crops. Focus groups also listed and scored their major farming problems. The most frequently mentioned problems were ranked in importance as follows: lack of capital; poor supply of farm inputs and equipment; poor access to services; soil acidity and infertility; difficult climate; lack of management knowledge; poor access to fertiliser; soil erosion; low crop yield. It can be seen that lack of capital stood out as a major constraint and that this was related to a cluster of problems related to soil and water management and the supply of inputs and knowledge needed to deal with these problems. Farmers also listed and scored the impacts of the introduced forages with which they had been experimenting. The most frequently mentioned impacts were ranked as follows: better quality feed; better reproductive performance; fatter animals;

82

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

increased quantity of feed; increased livestock numbers; a saving of labour; erosion control; production of more manure. In focus group discussions, farmers developed a wealth ranking system – ‘‘better-off’’, ‘‘average’’, and ‘‘poor’’ – and allocated all households in each village to a category. The combined result for the commune was that 15% of households were assigned to the better-off category, 38% to the average category, and 47% to the poor category. Interviews were conducted with five households in each of these categories. In summary, all categories of household had a small area (around 0.2 ha) of bunded rice that was an important and intensively managed component of the farming system. However, moving from ‘‘poor’’ to ‘‘betteroff’’ households, there was an increase in the size and physical capacity of the household workforce, the size of the landholding, the number of cattle (and other livestock), the use of farmyard manure, the use of purchased inputs (fertiliser, herbicide, insecticide, concentrates), and the importance of tree crops (coffee, fruit trees, timber trees). There was a decline in the importance of cassava, the extent of off-farm labouring, and the number of months for which there was reported to be a shortage of cash. Thus ‘‘average’’ and ‘‘better-off’’ households, largely because of their cattle-grazing activity, were able to manage their acid soils more effectively than ‘‘poor’’ households through the addition of a combination of organic (farmyard manure) and purchased fertilisers (urea, phosphates, NPK, and lime) to both field crops and tree crops, especially coffee. The addition of managed forages had the potential to augment this process by increasing the production of both manure and cash income, hence it was an attractive option to farmers, though still only on a small scale.

4. Discussion The case studies do not represent stages in a single evolutionary sequence. Rather each case indicates a mode of adaptation to a specific set of opportunities and constraints, as well as a particular evolutionary pathway for that setting. Nevertheless, when interpreted in terms of the conceptual framework summarised in Fig. 1, there are some general tendencies

that emerge from a comparison between the case studies and some lessons for research and extension. The sparsely populated, relatively remote, and steeply sloping acid uplands of Sarawak have continued to be utilised by the Iban and other Dayak tribes for shifting cultivation, long after farming systems in other regions of Southeast Asia have been intensified. Though the system has been sustainable for several centuries, recent population growth has forced farmers to reduce fallow periods to their lower limit in this environment. Further intensification in line with Boserup’s (1965) sequence was not feasible, given the high rainfall, steep slopes, and shallow, infertile soils. Instead farmers had two adaptive options – movement into a lowland environment to cultivate rainfed, bunded rice for subsistence (conforming to Pingali and Binswanger’s (1987) spatial intensification sequence); or a change in the upland farming system to focus on intensively managed perennial cash crops, involving the purchase of lime and inorganic NPK fertilisers (following a pathway of commercialisation). Having tried the first strategy in the 1960s and early 1970s, by the mid-1970s, with improved transport and market infrastructure, a boom in pepper prices, and technical and material support from the government, farmers eventually opted for the second strategy, while maintaining shifting cultivation of hill rice on a reduced scale as a spare-time activity. This corresponds to a move from the lower-left to upper-left sector in Fig. 1. Javanese transmigrants in South Kalimantan have adapted their land-use practices to their new environment. Rather than practise continuous cropping, as envisaged by the transmigration authorities, they have developed a short-fallow system of cultivation, using fire and plough to prepare the land and allowing it to revert to a grass fallow for a period roughly equal to the cropping period. This has been made possible by the acquisition of additional land. In accordance with Boserup’s (1965) schema, farmers relied on the addition of manure from their draught animals, both directly by grazing their animals on the fallowed fields, and indirectly by transporting farmyard manure from the home-lot to the cultivated fields. However, labour-intensive composts or green manure crops were not adopted. Instead farmers supplemented farmyard manure with moderate levels of inorganic fertilisers, reflecting the degree to which the farming system had

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

become commercially oriented. Rising maize prices and improved market access in recent years had increased the cash flow and reinforced the trend towards the use of purchased inputs for soil fertility management. Thus the farming system in the region had moved successively from the lower-left sector in Fig. 1 (the pre-transmigration system of shifting cultivation), through the lower-right sector, into the upper-right sector, and was moving progressively closer to the upper-left sector. In Bukidnon, with less land available, many farmers had adopted an intensive, commercial maize farming system, particularly those with relatively level land close to transportation routes. The use of moderate to high levels of inorganic fertiliser and lime was widespread and most farmers had adopted hybrid maize. Many farmers with sloping land had also adopted soil conservation practices. The role of ruminant livestock in the farming system had begun to decline, with over a third of farmers using tractors for some or all of their tillage operations. Though crop residues (maize stover) were retained in the field and incorporated during ploughing, no farmer had introduced a legume into the crop rotation nor adopted the use of compost or green manure. The use of farmyard manure was minimal. Nevertheless, most farmers reported increased yields and income from maize over the past decade as a result of the new seedand-fertiliser technology. Thus farmers were not following the labour-intensive path of soil fertility management using endogenous resources but were relying instead on the supply of exogenous inputs, made possible by the general economic development of the region. In terms of Fig. 1, the farming system had initially moved from the lower-left sector (the shifting cultivation system practised by the indigenous Bukidnon) to the upper-right sector, but with commercialisation had since moved firmly into the upper-left sector. Rubber trees, being well adapted to acid soils, have provided smallholders in Thailand, Malaysia, and Indonesia with an important source of cash income. Intercropping of young rubber trees with food crops has been practised for a long time among Malay rubber smallholders in Southern Thailand as a way of utilising spare land and labour to supplement household income. Initially, the practice used traditional techniques and was mainly for subsistence purposes.

83

However, with improved market access there has been an evolution towards commercial intercropping, with greater use of purchased inputs, particularly inorganic fertiliser. The place of intercropping in the overall farming system – as an adjunct to clearing and replanting commercial rubber smallholdings – affected the soil management practices adopted. The cleared rubber plots were routinely burned to facilitate rapid replanting, in the process helping to ameliorate the otherwise limiting acid soil conditions and facilitate the cultivation of food crops. The productivity of the soil declined rapidly, as evidenced by the sharp decline in the yield of crops such as rice in the second and third cropping. However, from the farmers’ point of view, there was no necessity for intercrop yields to be maintained, as within 3 years or so the land would revert to a pure stand of rubber (in contrast to the situation of maize farmers in Bukidnon). Where the commercial returns to intercropping were sufficiently high, as in the case of pineapple, farmers were willing to invest in the extra inputs (fertiliser, hired labour) needed to maintain productivity throughout the intercropping period. For farmers to adopt modified cropping systems and soil management practices, in line with the Boserup model, they would need to be convinced that any additional labour and input requirements were minimised or amply rewarded with assured market returns. Hence labourious practices involving the utilisation of organic residues or the substitution for a marketable crop of a green manure crop were unattractive. In terms of Fig. 1, the farming system had moved from the lower-left sector prior to the advent of rubber (shifting cultivation of rice) to the upper-left sector (commercial rubber production, with or without upland rice). Within this larger system, intercrop practices varied from the lower-left sector for remote villages growing upland rice in a quasi long-fallow system (with rubber instead of secondary forest as the ‘‘fallow’’ vegetation) to the upper-left sector for commercial vegetable and pineapple farmers. Immigrant Kinh farmers in the still relatively sparsely populated and poorly serviced grasslands of Daclac Province in Vietnam depend mainly on cattle production and have chosen crops such as cassava, rice, and coffee which can grow well in acid soils. Poorer farmers, with less land and labour, concentrated on food crops such as cassava and rice,

84

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

non-ruminant livestock (pigs and poultry), and fish, using low levels of inputs. Average and better-off farmers had more land and labour and, in addition to rice cultivation for subsistence, had coffee and cattle to generate cash income with which to purchase inputs of fertiliser and lime. Cattle played a key role in land utilisation and soil management for most farmers in this environment. Cattle utilised the extensive Imperata grasslands at low cost and contributed farmyard manure to food crops and the emerging cash crop, coffee. The sale of cattle (and manure) eased the capital constraint, identified by farmers as their most important problem, thus facilitating the purchase of inputs for their home gardens, upland crops, and coffee plots, and reducing the need to work off-farm. The introduction of new forage species has helped supplement the feed supply for cattle, thus complementing the beneficial effects of cattle on acid soil management. In terms of Fig. 1, the farming system has again moved from a starting point in the lower-left sector (the shifting cultivation system of the indigenous Montagnard farmers) into the upper-right sector.

5. Conclusion The evolution of farming systems in the acid uplands of Southeast Asia has been subject to two major influences – population growth and movement, leading to the intensification of land use, and the opening of the uplands to domestic and international markets for food and industrial crops, leading to the commercialisation of previously subsistence farming systems. These underlying trends have been variously accelerated and retarded by a range of agrarian policies and programs. Farmers’ strategies for dealing with the variety of technical problems associated with acid upland soils have been a response both to the specific bio-physical constraints in each location (not just acid soil infertility but the steepness of the terrain, climatic constraints, etc.) and the socio-economic constraints resulting from population pressure, market access, and government policies and institutions. The case studies tend to confirm the rationality assumption of farming systems research, namely that farmers, in pursuing their goals, are continually adapting their farming practices – in this context, management practices for acid upland soils – to

changing resource endowments and bio-physical and socio-economic conditions. Though in each of the case studies the evolutionary tendencies described by Boserup (1965) and Pingali and Binswanger (1987) were evident, namely, the substitution of ever more intensive soil management practices as population pressure on the land increased, it seems clear that the parallel opening up of commercial opportunities has provided an alternative pathway for most farmers. Production of food crops, tree crops, and livestock for domestic and export markets has generated the cash flow needed to augment traditional soil and crop management practices with substantial use of purchased inputs, not only of fertiliser but improved planting materials, herbicides, and machinery. Much of this development has occurred through the autonomous activity of farmers, rural traders, and agribusiness supply firms. The framework introduced in this paper and outlined in Fig. 1 helps to capture these alternative pathways. The utility of the framework is in helping to think strategically about acid soils research. Designing soil management practices on their bio-physical merits alone can waste scarce research resources. There is a need to match research effort to the evolutionary pathway of a farming system. For example, designing labour-intensive or financially expensive fertility management practices for shifting cultivation systems (lower-left sector) is unlikely to be warranted. In situations of low-medium population density, with labour the limiting factor, research on highly labour-intensive practices (such as the handcutting and composting of Imperata cylindrica in the Indonesian case) is also likely to be inappropriate. Conversely, in situations of relatively high population density (such as the Philippines case), use of scarce land for forage crops or green manures, rather than for food crops, would likely be resisted by farmers (cf. Fujisaka, 1992; Sumberg, 2002). Adding in the market access dimension in Fig. 1 shows that, at moderate to high population densities (as in the Indonesian and Philippines cases), interest in exogenous inputs can increase as improved market access increases cash flows, reduces the cost of purchased inputs, and increases the opportunity cost of farm labour. In fact, the case studies indicate a tendency for farmers to treat exogenous and endogenous nutrient sources as substitutes rather than

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

complements (cf. Omamo et al., 2002), suggesting that the upper-right sector may in fact be a rather narrow transitional zone as commercialisation exerts increasing influence. Hence, if integrated nutrient management is the desired end, there is a need to find sources of organic nutrients that economise on the use of land in land-scarce situations, and economise on labour where improved market access is raising the opportunity cost of farm labour. The development of input markets for organic nutrients also needs further investigation (Place et al., 2003). More generally, as an IFPRI/FAO workshop in 1995 concluded, there is a need for ‘‘prioritised and strategic problem-solving agricultural research that is related to plant nutrition management and the incorporation of mineral and organic sources of plant nutrients in the soil’’, and for greater use of ‘‘participatory and farmer-adapted approaches to technology development’’ (Gruhn et al., 2000: 24). Such a strategic and participatory approach has become more widespread in the years since the 1995 workshop, thus increasing the likelihood of improved management practices for acid upland soils in Southeast Asia that can generate the higher cash returns and labour-use efficiencies farmers increasingly seek, while contributing to the longer-term maintenance of the soil resource.

Acknowledgements An earlier version of this paper was presented at a conference on ‘‘Management of Infertile Acid Upland Soils of the Tropics and Subtropics: Research to Implementation’’ held at Kasetsart University, Bangkok, 11–13 June 2001. I am grateful to Pax Blamey, Colin Brown, Peter Kerridge, Eberhard Bruenig, the editor and two anonymous referees for helpful comments. Funding for this research was provided in part by the Australian Research Council and the Australian Centre for International Agricultural Research. References Andriesse, J.P., Schelhaas, R.M., 1987. A monitoring study on nutrient cycles in soils used for shifting cultivation under various

85

climatic conditions in tropical Asia. III. The effects of land clearing through burning on fertility level. Agric. Ecosys. Environ. 19, 311–332. Arbhabhirama, A., Phantumvanit, D., Elkington, J., Ingkasuwan, P., 1988. Thailand: Natural Resources Profile. Oxford University Press, Singapore. Babcock, T., 1986. Transmigration: the regional impact of a miracle cure. In: MacAndrews, C. (Ed.), Central Government and Local Development in Indonesia. Oxford University Press, Singapore. Barlow, C., Jayasuriya, S.K., 1984. Problems of investment for technological advance: the case of Indonesian rubber smallholders. J. Agric. Econ. 35, 85–95. Boserup, E., 1965. The Conditions of Agricultural Growth: The Economics of Agrarian Change Under Population Pressure. Earthscan, London. Coxhead, I., Buenavista, G. (Eds.), 2001. Seeking Sustainability: Challenges of Agricultural Development and Environmental Management in a Philippine Watershed. Philippine Council for Agriculture, Forestry and Natural Resources Research and Development (PCARRD), Los Banos. Cramb, R.A., 1984. Hill Rice Farming in the Upper Saribas Area. Farm Management Report No. 6. National Extension Project, Department of Agriculture, Kuching, Sarawak. Cramb, R.A., 1988. The commercialisation of Iban agriculture. In: Cramb, R.A., Reece, R.H.W. (Eds.), Development in Sarawak: Historical and Contemporary Perspectives. Monash Papers on Southeast Asia No. 17, Centre of Southeast Asian Studies, Monash University, Melbourne, pp. 105–134. Cramb, R.A., 1989. The use and productivity of labour in shifting cultivation: an East Malaysian case study. Agric. Syst. 29, 97– 115. Cramb, R.A., 1991. The changing agricultural economy of the Saribas Iban. In: Said, Mhd. Ikmal, Saravanamuttu, John (Eds.), Images of Malaysia, Persatuan Sains Sosial Malaysia, Kuala Lumpur, pp. 8–36. Cramb, R.A., 1993. Shifting cultivation and sustainable agriculture in East Malaysia: a longitudinal case study. Agric. Syst. 42, 209– 226. Cramb, R.A., 1998. Environment and development in the Philippine uplands: the problem of agricultural land degradation. Asian Stud. Rev. 22, 289–308. Cramb, R.A., 2000. Processes influencing the successful adoption of new technologies by smallholders. In: Stur, W.W., Horne, P.M., Hacker, J.B., Kerridge, P.C. (Eds.), Working with Farmers: The Key to Adoption of Forage Technologies. ACIAR Proceedings No. 95, Australian Centre for International Agricultural Research, Canberra, pp. 11–22. Cramb, R.A., Garcia, J.N.M., Gerrits, R.V., Saguiguit, G.C., 2000. Adoption of soil conservation technologies in the case study sites. In: Cramb, R.A. (Ed.), Soil Conservation Technologies for Smallholder Farming Systems in the Philippine Uplands: A Socioeconomic Evaluation. ACIAR Monograph No. 78, Australian Centre for International Agricultural Research, Canberra, pp. 72–118. Cramb, R.A., Purcell, T., 2001. How to Monitor and Evaluate Impacts of Participatory Research Projects: A Case Study of

86

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87

the Forages for Smallholders Project. CIAT Working Document No. 185, International Centre for Tropical Agriculture, Cali. Cramb, R.A., Purcell, T., Ho, T.C.S., 2004. Participatory assessment of rural livelihoods in the Central Highlands of Vietnam. Agric. Syst. 81, 255–272. Craswell, E.T., Pushparajah, E. (Eds.), 1989. Management of Acid Soils in the Humid Tropics of Asia. Australian Centre for International Agricultural Research and International Board for Soils Research and Management, Canberra. Date, R.A., Grundon, N.J., Rayment, G.E., Protert, M.E. (Eds.), 1995. Plant-Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht. DENR, 1993. Ecological Profile of Bukidnon. Department of Environment and Natural Resources, Region X, Cagayan de Oro City. Dierolf, T., Fairhurst, T., Mutert, E., 2001. Soil fertility kit: a toolkit for acid. Upland Soil Fertility Management in Southeast Asia, Potash and Phosphate Institute, East and Southeast Asia Program, Singapore. Dixon, C., 1991. South East Asia in the World-Economy. Cambridge University Press, Cambridge. Drabble, J., 2000. An Economic History of Malaysia. Macmillan, London. Duque, C.M., Cagmat, R.B., Daquiado, N.P., Maglinao, A.R., 1995. Management of acid soils for sustainable food crop production in the Philippines. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 757–760. Edgerton, R.K., 1982. Frontier society on the Bukidnon plateau 1870–1941. In: McCoy, A.W., de Jesus, E.C. (Eds.), Philippine Social History: Global Trade and Local Transformations. Ateneo de Manila University Press, Manila, pp. 361–390. Elson, R., 1997. The End of the Peasantry in Southeast Asia. Macmillan, London. Fisk, E.K., 1975. The response of non-monetary production units to contact with the exchange economy. In: Reynolds, L.G. (Ed.), Agriculture in Development Theory. Yale University Press, New Haven, pp. 53–83. Fujisaka, S., 1992. Farmer knowledge and sustainability in ricefarming systems: blending science and indigenous innovation.. In: Moock, J.L., Rhoades, R.E. (Eds.), Diversity, Farmer Knowledge and Sustainability. Cornell University Press, Ithaca. Garcia, J.N.M., Gerrits, R.V., Cramb, R.A., Saguiguit, G.C., 1995. Soil Conservation in an Upland Farming System in Bukidnon: A Socio-Economic Survey. SEARCA-UQ Uplands Research Project, Survey Report No. 3. SEARCA, Los Banos, Philippines. Garrity, D.P., Kummer, D.M., Guiang, E.S., 1993. The Philippines. In National Research Council, Sustainable Agriculture and the Environment in the Humid Tropics. National Academy Press, Washington, DC. Gruhn, P., Goletti, F., Yudelman, M., 2000. Integrated Nutrient Management, Soil Fertility, and Sustainable Agriculture: Current Issues and Future Challenges. Food, Agriculture, and the Environment Discussion Paper No. 32. International Food Policy Research Institute, Washington, DC. Ho, T.C.S., 2002. Sustainable Rural Livelihoods: A Case Study of Curoa Commune in the Central Highlands of Vietnam. MNatResEcon Research Project. University of Queensland, Brisbane.

Ibrahim, Truong, T.K., Heriyanto, 2000. Forage technologies for smallholders in grassland areas. In: Stur, W.W., Horne, P.M., Hacker, J.B., Kerridge, P.C. (Eds.), Working with Farmers: The Key to Adoption of Forage Technologies, ACIAR Proceedings No. 95. Australian Centre for International Agricultural Research, Canberra, pp. 277–280. Johnston, B.F., Kilby, P., 1975. Agriculture and Structural Transformation. Oxford University Press, New York. Kiripat, S., 1991. An Economic Analysis of the Growth of the Rubber Industry in Thailand, 1970-1988. PhD Thesis. University of Western Australia, Perth. Lefroy, R.D.B., Blair, G.J., Craswell, E.T. (Eds.), 1995. Soil Organic Matter Management for Sustainable Agriculture. ACIAR Proceedings No. 56, Australian Centre for International Agricultural Research, Canberra. Low, H., 1848. Sarawak: Its Inhabitants and Productions. Richard Bentley, London. Masae, A., Cramb, R.A., 1995. Socio-economic aspects of rubber intercropping on acid soils in Southern Thailand. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 679–684. Masyhuri, Cramb, R.A., 1995. A socio-economic assessment of land-use practices in a transmigration settlement on acid soils in South Kalimantan, Indonesia. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 685–688. Moran, E.F., 1987. Socioeconomic considerations in acid tropical soils research. In IBSRAM, Management of Acid Tropical Soils for Sustainable Agriculture, International Board for Soils Research and Management, Bangkok, pp. 224–227. Mugot, I.O., Cramb, R.A., 1995. A socio-economic assessment of smallholder maize farming on acid soils in Northern Mindanao, the Philippines. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 671–677. Myers, R.J.K., De Pauw, E., 1995. Strategies for the management of soil acidity. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 729–741. Myint, H., 1973. The Economics of the Developing Countries, 4th ed. Hutchinson, London. Nguyen, T.S., Thai, P., 1995. Acid upland soils in Vietnam and their management for agriculture. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 771–774. Omamo, S.W., Williams, J.C., Obare, G.A., Ndiwa, N.N., 2002. Soil fertility management on small farms in Africa: evidence from Nakuru District, Kenya. Food Policy 27, 159–170. Pandey, S., Lapar, M.A.L., 1998. A microeconomic analysis of adoption of contour hedgerows in the Philippine uplands. In: Penning de Vries, F.W.T., Agus, F., Kerr, J. (Eds.), Soil Erosion at Multiple Scales: Principles and Methods for Assessing Causes and Impacts. CABI Publishing, Wallingford, pp. 83–98.

R.A. Cramb / Agriculture, Ecosystems and Environment 106 (2005) 69–87 Pingali, P.L., 1990. Institutional and environmental constraints to agricultural intensification. In: McNicoll, G., Cain, M. (Eds.), Rural Development and Population: Institutions and Policy. Oxford University Press, New York, pp. 243–260. Pingali, P.L., 1996. Agricultural commercialisation and farmer product choices the case of diversification out of rice. In: Pingali, P.L., Paris, T.R. (Eds.), Competition and Conflict in Asian Agricultural Resource Management. International Rice Research Institute, Los Banos, pp. 62–89. Pingali, P.L., Binswanger, H.P., 1987. Population density and agricultural intensification: a study of the evolution of technologies in tropical agriculture. In: Johnson, D.G., Lee, R.D. (Eds.), Population Growth and Economic Development: Issues and Evidence. University of Wisconsin Press, Madison, pp. 27–56. Place, F., Barrett, C.B., Freeman, H.A., Ramisch, J.J., Vanlauwe, B., 2003. Prospects for integrated soil fertility management using organic and inorganic inputs evidence from smallholder African agricultural systems. Food Policy 28, 365–378. Pushparajah, E., Bachik, A.T., 1987. Management of acid tropical soils in Southeast Asia. Management of Acid Tropical Soils for Sustainable Agriculture, International Board for Soils Research and Management, Bangkok. Raintree, J.B., Warner, K., 1986. Agroforestry pathways for the intensification of shifting cultivation. Agrofor. Syst. 4, 39–54. Rubber Research Institute, 1992. Para Rubber Information. Rubber Research Institute, Thailand, Bangkok. Rumawas, F., 1987. Reclamation of degraded acid tropical soils in Indonesia. Management of Acid Tropical Soils for Sustainable Agriculture, International Board for Soils Research and Management, Bangkok. Ruthenberg, H., 1980. Farming Systems in the Tropics, 3rd ed. Clarendon Press, Oxford. Ruttan, V.W., Hayami, Y., 1998. Induced innovation model of agricultural development. In: Eicher, C.K., Staatz, J.M. (Eds.), International Agricultural Development, 3rd ed. Johns Hopkins University Press, Baltimore, pp. 163–178. Sanchez, P.A., 1974. Properties and Management of Soils in the Tropics. Wiley, New York.

87

Sanchez, P.A., 1994. Alternatives to slash and burn: a pragmatic approach for mitigating tropical deforestation. In: Anderson, J.R. (Ed.), Agricultural Technology: Policy Issues for the International Community. CABI Publishing, Wallingford, pp. 451– 479. Sibat, P., 1980. Report on a Semi-Detailed Soil Survey of the Nanga Spak Area. Report No. 218. Soil Survey Division, Department of Agriculture, Kuching, Sarawak. Southern Studies Group, 1982. Para Rubber Crisis in Southern Thailand. Punmitr Paritat, Bangkok. Sumberg, J., 2002. The logic of fodder legumes in Africa. Food Policy 27, 285–300. Suthitipradit, S., Nualsri, L., Sophanodora, P., Limchitti, L., Kungpisdan, 1995. Management of acid soils for opportunity food crop production in Southern Thailand. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 761–766. Tie, Y.L., 1982. Soil Classification in Sarawak. Technical Paper No. 6. Soils Division, Department of Agriculture, Kuching, Sarawak. Tomich, T.P., Kilby, P., Johnston, B.F., 1995. Transforming Agrarian Economies: Opportunities Seized, Opportunities Missed. Cornell University Press, Ithaca. Tran, T.Q., 1998. Vietnam’s Agriculture: The Challenges and Achievements. Institute of Southeast Asian Studies, Singapore. Von Uexkull, H.R., Mutert, E., 1995. Global extent, development and economic impact of acid soils. In: Date, R.A., Grundon, N.J., Rayment, G.E., Probert, M.E. (Eds.), Plant–Soil Interactions at Low pH: Principles and Management. Kluwer, Dordrecht, pp. 5–19. Whitbread, A.M., Blair, G.J. (Eds.), 1999. Integrated Nutrient Management in Farming Systems in Southeast Asia and Australia. ACIAR Proceedings No. 93, Australian Centre for International Agricultural Research, Canberra. Willett, I., 1994. Physical and chemical constraints to sustainable soil use under rainfed conditions in the humid tropics of Southeast Asia. In: Syers, J.K., Rimmer, D.L. (Eds.), Soil Science and Sustainable Land Management in the Tropics. CABI Publishing, Wallingford, pp. 235–247.