Land Use, Nature Conservation and the Stability of Rainforest Margins in Southeast Asia

The purpose of this paper is to summarise different research results about the impact of macro-level factors and “extra-sectoral” policies on tropical forest cover. Specifically, we are interested in the forest margins — i.e. the spatial transition zone between tropical forests and converted land uses. What are the policy factors that accelerate frontier expansion, and which ones tend to slow it down? The term “extra-sectoral” refers to all the things that happen outside of forests and forestry, yet nevertheless have a significant effect on forests. For instance, how do changes in international trade and a country’s balance of payment affect deforestation? What does it mean for pressures on forests that a country drastically devalues its currency? What is the role of population growth?

If Nancy Peluso’s “Rich Forests, Poor People” may have been a valid observation until 1992, now ten years later (2002) the title is definitely no longer appropriate, not even for larger Outer Islands such as Sumatera, Kalimantan or Sulawesi and Papua. The situation is becoming extremely cumbersome for millions of shifting cultivators living in or around forest areas. A more recent mapping of the forest cover of Indonesia by the Ministry of Forestry (MOF) has revealed that the rate of deforestation in Indonesia approximately doubled between 1985 and 1997, from less than 1.0 million ha to at least 1.7 million ha each year (Holmes 2002). The mapping predates the widespread forest fires of 1997–98 and the extensive illegal logging that followed the political crisis of 1998. The analysis of the MOF data concludes that over 20 million ha of forest cover have been lost over a twelve year period including 6.7 million ha in Sumatera and 8.5 million ha in Kalimantan. This amounts to an average annual rate of 1.67 million ha, nation wide-roughly 600 ha per day or 190 ha per hour. The rate in the three islands of Sumatera, Kalimantan and Sulawesi is 1.45 million ha per year. The rate of forest loss between 1985 – 1997, which has been quoted from the Holmes’ report shows the disastrous decline of Indonesia’s tropical rain forest (Annex.1 & 2). In central Sulawesi where also the Lore Lindu is located, the forest area constituted 72.3% of total area in 1985. However, the forest area declined to 56.7% in 1997, which means 15.6% in 12 years or more than one percent per annum, equivalent to an average of 35.000 ha.

The use and management of natural resources in the rainforest margins of Central Sulawesi has become a focus of national and international research very recently. That is because the area was always located in a political and economic periphery and the pressure on the rainforest resources had been relatively low. But nowadays the awareness of resource management of tropical rainforests has risen. Deforestation has got worldwide attention mainly because of its climatic relevance. Also a growing population pressure in Indonesia accompanied by an actual political and economic transformation process has intensified the impacts on the resources in Central Sulawesi. Especially during the last two decades the migration into the region has increased and strengthened the pressure on the access to land. But the foundations for the cultural landscape happened much earlier. These historically developed patterns of population and settlement structure lead to different attitudes of today’s resource management and effects on land use in the region. The reflection of the processes, which show the way to a local differentiation of resource management around the Lore Lindu National Park, is stressed in this paper and will be exemplified with micro studies at the village level. The paper focuses on the external and internal impacts on the shape of the cultural landscape during different historical periods of the last century and reflects the demographic development as well as the land use changes in the region.

Currently the push toward the frontier areas, which until twenty years ago were still largely untouched by commercial agriculture, has taken a massive form. In most cases the process has resulted in an agrarian differentiation that is not in favor of the local or indigenous communities. Even worse, in many cases the local communities, it seems, do not have the institutions and social organizations to defend their own interest, and their environment on which they partly depend. Cornered by the influx of land hungry migrants and by the shortage of land, many local communities start to demand rights on their ancestors’ land, large parts of which are under the category of state forestland, protection forest or even as National Park. In the district North Lore, Central Sulawesi, on the borders of the Lore Lindu National Park, local discontent toward the one sided decision regarding borders of the national park has resulted in activities in mapping the ancestors’ land and the establishment of forest zones where local people have use-rights. However, there are still no institutions and organizations able to manage this land. Meanwhile, unregulated extraction of timber and the expansion of agriculture land into the forest and the National Park keep going on.

This Chapter focuses on change of ethnic composition, which is one aspect of demographic structure, and its effects on the unity of communities. This Chapter is a component of ongoing research on the impact of spontaneous migration on the cultural landscape and agrarian structure in two forest-margin villages in Central Sulawesi. The research is part of a multi-disciplinary research program on the “Stability of Rainforest Margin in Indonesia” and focuses particularly on the margins of the Lore Lindu National Park (LLNP) in Central Sulawesi. The Institut Pertanian Bogor, Universitas Tadulako (Indonesia), University of Göttingen, and University of Kassel (Germany) execute the program.

Since agrarian structure refers both the composition of agrarian resources and the social relationship regarding distribution of access to such resources, it clearly assumes the articulation of a social formation, which is “a multiplex entity in which several modes of production co-exist, one of which, however, is dominant” (Worsley 1988). Referring to the elements of the mode of production, namely forces and relations of production (Worsley 1988; Taylor 1989), agrarian resources are connected with the forces of production, particularly the means of production, and social relationship is connected with the relations of production. Hypothetically, then, it can be said that any change in the structure of social formation implies a change in agrarian structure.

‘Networks’ and ‘networking’, as Haverkort notes, is ‘nothing new under the sun’, and networking between farmers is ‘as old as farming itself’. However, in recent years new forms of farmer networks are arising partly due to new political space in some countries, such as Thailand, and partly due to government bureaucracies that have largely failed to deliver adequate services. In Northern Thailand, for example, ethnic communities came under increasing pressure through new government legislation banning traditional forms of agricultural production in the name of conservation. Here, traditional forms of practices, information exchange and cooperation are revived and linked with new forms and opportunities. They grew out of the need to fight for the protection of farmers’ land, for conflict resolution over resource use among themselves or vis-à-vis other, often powerful and well-connected interest groups, for civil rights (e.g. citizenship), for protecting their community from drug trafficking and abuses, and for getting access to relevant information, new technologies, or development funds.

In recent years, devolution of forest management to local communities has become a major policy trend in developing countries. The term devolution is used here to refer to the transfer of responsibility and authority over natural resources from the state to non-governmental bodies at the local level (Meinzen-Dick and Knox 2001, p. 42). Devolution policies aim to address institutional problems that have been identified as major reasons behind the degradation and misuse of forest resources in developing countries such as state property and centralized management of forest resources, corruption in the forestry administration, lack of effective monitoring, and enforcement and deficient incentives for the local communities (McCarthy 2000a; Ligon and Narain 1999; Wibowo and Byron 1999). Different mechanisms have been identified in the relevant literature on the subject by which devolution can lead to a more sustainable forest management in terms of equity, efficiency, and environmental sustainability: the creation of incentives by a fair and democratic distribution of benefits; the creation of accountability; the reduction of transaction costs; the mobilization of local knowledge; the strengthening of local institutions for sustainable resource management; and — in view of a low state capacity — the limitation of the role of the state to the provision of enabling frame conditions and the protection of public interests (compare Ribot 2002, Meinzen-Dick et al. 2001, Birner and Wittmer 2000; World Bank 1997). However, the empirical evidence on the effects of devolution in the forestry sector has been mixed (Banerjee 1997, Meinzen-Dick et al. 2001, Ribot 2002), which implies a need for more theoretical and empirical research on devolution.

Enhancing agricultural technology is an important policy tool for improving livelihoods and stimulate economic growth in rural areas of developing countries. It is however not completely understood whether technical progress and agricultural intensification would reduce or aggravate pressure on forests (Angelsen et al, 1999). Does technical improvement diminish forest clearing by reducing the area farmers need to make a living? Or does technical progress lead to increased agricultural expansion and deforestation by rendering agriculture more profitable? The question of a ‘forest saving’ or a ‘forest clearing’ effect of technical progress is extremely relevant within the framework of interactions between environmental and socio-economic goals. Policy debates have been dominated by the assumption of a win-win relation between technical progress and forest conservation (Angelsen and Kaimowitz 2001).

The Lore Lindu National Park has received much attention nationally and internationally during the past few decades. Maintaining existing levels of biodiversity has become an increasingly challenging task involving decision-makers and stakeholders at different levels of aggregation. A quarter of all mammal species of the earth are facing extinction within the next 30 years and one in eight bird species (12%) within the next 100 years (Hilton-Taylor, C., 2000; BirdLife International, 2000). The three factors that are responsible for the decrease in biodiversity are habitat loss, overexploitation and the introduction of alien species (Primack 2000). Sulawesi has an outstandingly unique flora and fauna (Stattersfield et al. 1998, Myers et al. 2000) and the Lore Lindu National Park is a key protected area for the conservation of Sulawesi’s biodiversity since it is one of only two protected areas in the Central Province (Coates et al. 1997). But the loss of biodiversity is closely linked to other problems and much more is at stake. Policy makers, scientists and the public are increasingly concerned about tropical deforestation and its negative consequences such as climate change, reduced timber supply, flooding, silting, and soil degradation (Kaimowitz and Angelsen 1998). In the case of Central Sulawesi, a combination of numerous factors contribute towards a trend of decreasing biodiversity. Primary forest lands are rapidly being converted to agricultural lands and practices such as slash and burn have increased this trend. The conversion of primary forest for the cultivation of perennial crops, mainly cocoa has been taking place for a long time. Farmers may very well be aware of the negative long term consequences of encroaching on the national park and logging of primary forest trees, but under the existing socio-economic circumstances, often short term private benefits will prevail over long term social costs. A central question in this connection would be:

To what extent is it really necessary for households to use land that is situated in the national park or convert land that is still covered by primary forest to agricultural land?

Research on tropical forests started many decades ago. According to Whitmore (1991) it began with taxonomy and floristic and structural inventories and was followed by classification of forest communities. Subsequent successional relationships between these communities generated special attention. Then investigation into ecophysiology became a very active field. Fluxes of water and nutrients were measured. So step-by-step basic knowledge about the functioning of tropical forests was compiled. And especially foresters contributed to this knowledge concerning production and regeneration issues. To date many scientists agree upon the necessity of a holistic approach, an ecosystem approach for understanding the multidimensional complexity of tropical forests in time and space, and they also agree upon the many “white spaces” left to investigate, especially functional relationships between plants and animals and the stability of these tropical ecosystems.

Biodiversity brings many benefits to humans, including direct ones, such as the provision of food, shelter and clothing. Many of these products are traded in the market, and have a recognized economic value. Others are subsistence products, but critical to the livelihood needs of people living in rural areas of developing countries, such as wild foods and bush meat, products that normally fall outside conventional economic analysis. There are also indirect ecological benefits and non-consumptive uses such as recreation (Stocking et al. 1995).

Shifting cultivation is blamed for 50% of all deforestation in Southeast Asia (FAO 1993). Clearly it has had, and will continue to have, a profound impact on the structure and functioning of rainforests. The most obvious changes that occur are a reduction in tree stature and an increase in the grain size of the landscape. These changes are a consequence of an increase in disturbance frequency (return interval shifts from 100s to l0s of years) and spatial scale of disturbance (patch size increases from 10s or 100s to 10,000s of square meters). Less obvious are the longterm shifts in species composition, species diversity, and forest structure that accompany this altered rainforest disturbance regime. Few have documented longterm responses to shifting cultivation, and fewer have taken the next step to determine the consequences of altering forest structure and organization for critical ecosystem functions such as nutrient cycling and maintenance of soil fertility. Understanding changes in biodiversity and ecosystem function within the context of a dynamic human response to economic and ecological drivers is more complex still. The focus of this paper is to address these issues, in answering two questions about rainforest conversion to shifting cultivation:

Indonesia is a country of megadiversity and ranks 5th on the list of the world’s richest countries in terms of biological diversity. The rich diversity of Indonesia is evidenced by the rain forests of Lore Lindu National Park (229,177 ha), Central Sulawesi, which may contain about 150 species of trees (> 10 cm dbh) in one hectare at ca. 1000 m elevation (Kessler et al. submitted). More than 2000 species of woody plants have been recorded from Sulawesi in a recent census (Keßler et al. 2002).

Many studies focus on the impact of forest disturbance on tropical diversity (e.g. Chung and Maryati 1996; Dahaban et al. 1996; Eggleton et al. 1997; Hamer et al. 1997; Intachat et al. 1997, 1999a; Johns 1992; Lambert 1992; Lawton et al. 1998; Liow et al. 2001; Watt et al. 1997; Willott 1999; Willott et al. 2000). In contrast, agricultural ecosystems and forest plantations have been taken into account only exceptionally (e.g. Chey et al. 1997; Davis et al. 2000; Holloway 1998; Holloway et al. 1992; Intachat et al. 1999b). Published studies of agroforests do not provide comparative data from pristine forests to evaluate the relative importance of agroforestry for maintaining tropical biodiversity (e.g. Klein et al. 2002a, 2002b; Perfecto and Snelling 1995; Perfecto and Vandermeer 1996). Secondary forests and certain land-use systems may be important in preserving at least a fraction of the tropical biodiversity (Hughes et al. 2002; Lugo 1992) of which insects are a major part (Hammond 1992; Stork 1988).

Sulawesi, one of the largest islands of the Malay Archipelago, provides living space for a unique fauna and flora. The island is characterized by vast numbers of endemic plant and animal species originating from the Australian as well as the Oriental region. The destruction and degradation of rain forest habitats pose serious threats to wildlife around the tropical world. On Sulawesi, recent population growth, resettlements, and social unrest have severely added to this problem, hence many of the island’s endemic species face a highly unknown future.

According to Fooden (1969, 1980) there are 19 species in the genus Macaca, of which seven occur endemically on Sulawesi, an area covering less than 2% of the total generic range (Albrecht 1978 cited in Bynum et al. 1997). Although there are already several field studies on Macaca nigra, Macaca nigrescen, and Macaca maura (MacKinnon 1980, Sugardjito et al. 1989, Supriatna 1991, Kohlhaas 1993, Kinnaird and O’Brien 1995, 1996, Reed et al. 1997, Rosenbaum et al. 1998, Okamoto and Matsumura 2002), only little has been done on the behaviour and ecology of the Tonkean macaque Macaca tonkeana. The Tonkean macaque (Macaca tonkeana) is a Central Sulawesi endemic and considered at lower risk/near threatened (Lr/nt) by IUCN categories (Hilton-Taylor 2000). Sulawesi Macaques are found in lowland and hill forests but only uncommon at elevations above 1500 m (Sarasin & Sarasin 1905, cited in Whitten, 1987). For Macaca tonkeana, the primary conservation issue at this time appears to be the ability of populations to persist in highland protected areas, since habitat encroachment, hunting and eradication as crop raiders appear to pose problems at moderate elevations (Bynum et al. 1999). The Lore Lindu National Park in Central Sulawesi is located in the centre of the species’ range and comprises generally mountainous terrain with over 90% of the park area located above 1000 m (Wirawan 1981).

Tropical deforestation and forest fragmentation are probably the most serious threats to biodiversity (see Turner 1996) and it has been theoretically stated that even the largest protected areas in the tropics might be too small to sustain populations of all species of the original system (Terborgh 1999). But species loss in forest fragments is a complex process and appears often only after considerable time lags, especially in vertebrates (Brooks et al. 1999b). Therefore, empirical evidence for such extinctions can only be obtained from areas with a long deforestation history and long-known faunal composition (e.g. van Balen 1999). Such empirical data are scarce but are essential in order to convince land use managers of the long-term effects of forest loss on biodiversity. Species area models, however, are a valuable tool in the prediction of tropical vertebrate species loss (van Balen 1999; Brooks et al. 1997, 1999a, c, 2002; Cowlishaw 1999).

The ongoing debate over the causes of deforestation in Indonesia views smallholder production and their growing number as a main cause of deforestation in Indonesia. With the common terminology ‘smallholder conversion’ a broad range of conversion strategies like the shifting cultivation-forest pioneer continuum, smallholder tree crop production and transmigration activities is summarized (Sunderlin and Resosudarmo, 1996). However, after ‘smallholder conversion’ the created landscape along the rainforest margin area is usually characterized by a patchwork of different land use types in ever smaller patches undergoing a gradual change from forest dominated areas via annual crops to perennial plantation interspersed by secondary forest, pasture and annual crops. Since the Indonesian economic crisis in 1997 land clearing increased dramatically for establishing export tree crops. Indonesia is now the third largest producer of cocoa (Sunderlin, 2000, Maertens et al, 2002) with Central Sulawesi as one of the main Indonesian production areas. The question then arises: to what extent do these conversion activities affect the hydrological behaviour of the affected areas ?

In recent years, Thailand has faced serious water problems, such as water shortages in the dry season, flooding during the rainy season and pollution with agrochemicals and industrial waste. Agriculture is the main user of available water resources and accounts for 70% of the water demand. Whereas traditionally irrigation in Thailand has been supplementary during the rainy season, water is now increasingly used for irrigation during the dry season. However, other sectors are continuously increasing their share in the withdrawal of water resources. The management of water resources is characterized by institutional and legal pluralism, responsibilities being shared among as much as eight different ministries, divided into more than 30 ministerial departments (Sethaputra et al. 2001).

A viable practice to overcome low availability of P in soils could be the intercropping of P efficient species with inefficient ones, or rotation systems in which crop residues of P efficient species are incorporated into the soil. Efficient plants can alter the chemical mobility and bioavailability of inorganic P in the rhizosphere through their uptake and exudation activities, which modify vital processes governing the P concentration in soil solutions such as adsorption-desorption, through pH changes, organic ligands secretion, affecting concentrations of metal cations which fix P such as Ca, Fe and Al (Gardner et al. 1983, Hinsinger, 2001).

Throughout the developing world, many development projects promote various technologies to address the decline in soil fertility. The majority of them rely on research results which are often not reproducible in different agro-ecological zones because of differences in soil properties, rainfall regime, economic framework, availability of fertilizers and other inputs, and management practices. Limited economic resources, coupled with the focus on extension instead on research and the short life span of development projects are major reasons not to conduct intensive adaptation trials for potential innovations before promoting them to farmers. Consequently, many extension projects have failed because they have not been able to address the real problems adequately in different project environments with their own specific characteristics. Therefore, there is a need to apply suitable methodologies to quickly assess the potential of technologies to be introduced under specific local conditions and to assess their risk of failure.

Since the introduction of rubber at the turn of the 20^th century smallholders have developed an original complex agroforestry system called jungle rubber, in which non selected young rubber trees (seedlings) are managed extensively alongside secondary forest re-growth.

In many parts of Indonesia, farmers are traditionally managing forest gardens, some of which have already been researched in detail, for example in the description of the “Lembos” of East Kalimantan by Sardjono (1990). The rattan gardens of the Dayak in East- and Central Kalimantan have been researched by Arifin (1995), Kraienhorst (1990) and Weidelt (1990, 1996). Sundawati (1993) has presented the “Tembawangs” of West Kalimantan and, most recently, research has been conducted in forest gardens in Central Maluku (Kaya et al. 2002).

Landscape ecology is based on the analysis of the linkages between spatial pattern and ecological processes within a landscape (Turner, 1989). Three landscape characteristics are important to the linkage between pattern and process: structure, function and change (Forman and Godron 1986). “Structure” refers to the spatial relationships between components of a landscape, that is, the distribution of energy, materials, and species in relation to the size, shape, numbers, kinds and configuration of components. “Function” refers to the interactions between the spatial elements, such as the flow of energy, materials, and organisms among the components of the landscape. “Change” refers to alteration in the structure and function of the landscape mosaic through time. Understanding the structure, function and change of landscapes helps to determine the relationship between landscape pattern and ecological process and has been the main research item in many studies of ecological systems.

Land use change, the physical change in land cover caused by human activities such as agriculture and silviculture, is a common phenomenon associated with population growth, market development, technical and institutional innovation, and related rural development policy. Changes in land use can have various consequences on economic growth, the level and distribution of income, and on natural resources such as biodiversity, ecosystems, water, and soils. Land use change leads to change in — and is influenced by — socio-economic indicators such as agricultural productivity, wealth, and human capital. A better understanding of the complex interactions of these changes over time can enable decision makers at various policy levels to design and implement regionally adapted policy interventions which stimulate benefits and counteract negative consequences of land use change by considering the trade-offs among economic, environmental, and social objectives in the process of sustainable rural development.

The contributions in this reader document the result of two years of research in an intercultural context. It has been managed by joint German-Indonesian planning and steering committees, and it has been conducted in close co-operation between more than 100 German (including a few other European nationals) and Indonesian scholars and researchers, both parts being equally represented in numbers.