Journal of Mountain Science Vol 2 No 2 (2005) - CiteSeerX

27 downloads 942 Views 571KB Size Report
Dec 12, 2006 - been further aggravated by resettlement programs resulting from large ..... favoring the maintenance of the status quo in areas with weak ...
Journal of Mountain Science Vol 3 No 4 (2006): 305~324 http://jms.imde.ac.cn; http://www.imde.ac.cn/journal Article ID: 1672-6316 (2006) 04-0305-20

Participatory and Integrated Research in Mountainous Regions of Thailand and Vietnam: Approaches and Lessons Learned Andreas Neef *, Franz Heidhues, Karl Stahr University of Hohenheim, Stuttgart, Germany *Corresponding author, E-mail: [email protected]

Pittaya Sruamsiri Chiang Mai University, Thailand

With contributions from Peter Elstner, Chapika Sangkapitux, Ulrich Schuler, Chalathon Choocharoen and Ludger Herrmann

Abstract: Participatory and integrated research approaches employed by a long-term ThaiVietnamese-German collaborative research program, ‘The Uplands Program’, that address the vicious circles of resource scarcity, environmental degradation and rural poverty in mountainous regions of northern Thailand and northern Vietnam are discussed in this paper. We present two examples from the Thai component of the research program to show how different disciplines and stakeholders need to cooperate at different scales to make meaningful scientific contributions towards sustainable land use and rural development in mountainous regions. The case of resource conservation in the Thai highlands shows that local and scientific knowledge, conventional surveys and participatory modeling can be creatively combined. Integrated research on the potential of integrating fruit trees and associated technologies into mountain farming systems suggests that natural scientists have to work alongside Received: 26 May 2006 Accepted: 12 September 2006

economists and social scientists to avoid harmful effects of purely technology-driven and productivityenhancing approaches. The success of new technologies cannot be measured solely by adoption rates and yield increases, but also needs to take into account their long-term impact on various groups of farmers and the ecological, economic and social trade-offs that they entail. Technical and institutional innovations need to go hand in hand to provide viable livelihood opportunities for smallholder farmers in mountain watersheds. The major lesson learned from the first six years of our research in the mountains of Thailand and Vietnam is that conventional and participatory approaches are not antagonistic; if scientists from various disciplines and research paradigms are open-minded, the combination of both approaches can produce meaningful results that cater for the needs of both the academic community and local stakeholders in mountain environments. Keywords: Integrated research; participatory approaches; local and scientific knowledge; Thailand; Vietnam

305

Andreas Neef et al.

Introduction Until the late 1980s, the mountainous regions of northern Thailand and northern Vietnam were largely cut off from the rapid economic development in the urban centers such as Bangkok, Chiang Mai and Hanoi, and from the dynamics of the ‘green revolution’ in the traditional rice-growing regions, like the central provinces of Thailand and the Red River and Mekong delta of Vietnam. In recent years, however, mountainous regions and their inhabitants, predominantly ethnic minority groups, have been confronted with profound changes: the increased purchasing power in the urban centers and more developed regions leads to changes in consumption level and structure, reflected in the growing demand for high-value agricultural products, above all the demand for fruits, vegetables and livestock products. At the same time, the pressure on these regions, especially on the erosion-prone hillsides, has been intensified by increasing migration from the overcrowded

valley areas in both countries. In Vietnam this has been further aggravated by resettlement programs resulting from large hydroelectric power dam projects, such as the Hoa Binh and the Son La dams. In Thailand, in particular, urban investors and speculators have acquired land in some of the more favored mountain locations within the last decade and have converted them into tourist resorts and cash crop plantations. Large-scale declaration of national parks, wildlife sanctuaries, watershed conservation zones and forest reserves following a ‘forest without people’ paradigm have caused tremendous reductions in the cultivable area and thus also in livelihood opportunities for highlanders in northern Thailand. The pressure of cultivating in increasingly smaller areas and the rapid change from swidden cultivation and fallow practices to more permanent forms of agriculture has led to vicious circles of land scarcity, resource degradation, rural poverty and food insecurity in many regions of northern Thailand and northern Vietnam (see Figure 1).

Population growth, resettlements

I. The causal factors

Reforestation, dam constructions

II. Non-feasible options

Non-farm opportunities insufficient

Opium substitution programs

Increasing land scarcity

Expansion into forest areas illegal

Reduction of fallow periods in sloping land

predominantly subsistenceoriented regions

III. The vicious circles predominantly market-oriented regions

Insufficient food security and increasing rural poverty

Erosion and nutrient leaching

Declining soil fertility

Reduced land productivity

Excessive input of agrochemicals

Increasing vulnerability of agro-ecosystems and threats to human health

Decline of biodiversity and water quality

Figure 1 Causal factors and vicious circles of land scarcity, environment degradation and rural poverty in mountainous regions of Thailand and Vietnam

306

Journal of Mountain Science Vol 3 No 4 (2006)

Whereas the problems in the more subsistence-oriented regions are largely confined to soil degradation and widespread poverty and food insecurity of the rural population, the excessive application of agrochemicals in more marketoriented regions — often promoted by national and international organizations as they strive towards replacement of opium poppy cultivation — lead to considerable losses of biodiversity and to contamination of surface and groundwater sources with pesticide and fertilizer residues. The on- and off-site effects of inappropriate land use systems pose considerable threats and have raised tensions between upstream and downstream communities. Lowland residents in northern Thailand claim that agriculture on upland slopes has disrupted the hydrological cycle by reducing dry-season flows and leading to higher risks of flash floods, river sedimentation and landslides during the rainy season. The validity of these claims, however, has not yet been substantiated by scientific work. Environmental issues in the highlands of Thailand and Vietnam are often entangled with social and cultural misconceptions. Ethnic minority farmers are widely blamed as the major culprits for the destruction of forests, water overuse and soil degradation, though in reality the causes of the current environmental problems are far more complex. One of the major challenges for research and development efforts in those areas is to contribute to improving the socioeconomic conditions of highland communities while maintaining cultural diversity and minimizing environmental impacts. In the following sections we present the approaches and lessons learned from a long-term Thai-Vietnamese-German collaborative research program that has taken up these challenges.

1

Objectives, Rationale and Concept of the Uplands Program

The Collaborative Research Center ‘Sustainable Land Use and Rural Development in Mountainous Regions of Southeast Asia’ (also called ‘The Uplands Program’) was initiated in July 2000 by the University of Hohenheim, Stuttgart, Germany, in cooperation with four Thai universities and four Vietnamese universities and research organizations. The program is funded by

the Deutsche Forschungsgemeinschaft (German Research Foundation) and co-funded by the National Research Council of Thailand and the Vietnamese Ministry of Science and Technology. It is organized in phases of three years each and may, if successful, be extended to four phases (i.e. until 2012). In its current third phase (July 2006-June 2009) it is composed of a total of 16 sub-projects, ranging from soil science, agronomy, agro-ecology and animal husbandry to economics and social science. 1.1 Objectives of the Uplands Program Primary and long-term objectives of the Uplands Program are the creation of a scientific base for: (1) developing and testing sustainable land use, production and processing systems in ecologically fragile and economically disadvantaged mountainous regions in Mainland Southeast Asia; (2) advancing methods and approaches for analyzing complex agro-ecological systems and their dynamics with special consideration of the interactions between ecological, ethnic diversity and heterogeneous institutional frameworks; (3) developing concepts for rural institutions and policies that can contribute to solving problems of rural poverty and food insecurity in mountainous regions and to improving the resilience of rural households in the dynamic economic environment of Southeast Asia. These objectives are closely connected to each other. In the regions with high population growth and limited agricultural resources, two parallel approaches are important for sustainable land use and rural development: (1) to increase productivity of vital resources while conserving their long-term viability, and (2) to reduce pressure on natural resources by expanding non-farm employment opportunities. The agricultural sector alone, with its limited resource potential in marginal mountainous regions, cannot sustain the growing population. Rural development, with the creation of additional income possibilities, e.g. in the agro-processing sector, is considered a crucial prerequisite for sustainable land use.

307

Andreas Neef et al.

1.2 Sustainability, participation and interdisciplinarity as central concepts 1.2.1 Sustainability The term ‘sustainability’ has become a buzzword in various economic, social and environmental policy arenas during recent years. The international debate on ‘sustainable development’ found its first important milestone in the UN Conference on Environment and Development in Rio de Janeiro in June 1992. In the Agenda 21, “sustainable development” was declared to be the first objective of national and international development and environment policies. One section of article 14 of the Agenda was

devoted to the areas of “sustainable agriculture and rural development”. The sustainability of agricultural production is now a central topic for orientation of future agricultural research in developing and newly industrialized countries. Today, the emphasis of many agricultural research programs is less on maximizing productivity and more on minimizing adverse ecological effects of land use systems and production processes (cf. German in this issue). These efforts can only be sustainable in the long run, however, if economic viability and social acceptability are also ensured, i.e., if a long-term improvement of living conditions is to be achieved for the largest possible stratum of land users (see Figure 2).

Figure 2 The dimensions of land use sustainability

The contribution of the agricultural sciences to improvement of land use sustainability lies especially in two areas: first, agricultural research needs to identify and measure factors, indicators and thresholds determining sustainability. This incorporates both the natural scientific and socio-economic/institutional contexts. The second fundamental contribution of agricultural research is in laying the foundations for developing technological and institutional innovations, which can increase efficiency of resource use and thus enhance sustainability. Beyond that, the Uplands Program makes a third significant contribution to the issue of sustainability by integrating the priorities, worldviews and knowledge of local stakeholders into the research process wherever

308

this is deemed feasible. We believe that sustainability is a concept that needs to be negotiated between different groups in society and that, above all, needs to give a voice to the most marginalized people. This is also stated in the Plan of Implementation worked out at the World Summit on Sustainable Development in Johannesburg, South Africa, in September 2002, which calls for actions to “promote full participation and involvement of mountain communities in decisions that affect them and integrate indigenous knowledge, heritage and values in all development initiatives” (quoted in UNEP 2002: 80). This leads us to the second major concept of our program, the issue of participation.

Journal of Mountain Science Vol 3 No 4 (2006)

1.2.2 Participation Like ‘sustainability’, the concept of ‘participation’ has also moved to the center of both development policy discussion and scientific discourse in international agricultural research. Whereas participatory approaches in development policy and in agricultural extension are well established — and have only recently been subjected to greater scrutiny — their usefulness in agricultural research has always remained controversial. From the political side, participatory approaches in development-oriented international agricultural research are increasingly supported. In the scientific community, however, there is

considerable skepticism regarding agricultural research that is solely geared towards the interests and priorities of farmers and other local stakeholders (see e.g., Neef 2005). Given the controversy surrounding participatory approaches to agricultural research, which has also been prevalent among scientists involved in the Uplands Program, we consider participation both as a research paradigm and as a research topic in itself. Our aim is to optimize participatory approaches depending on research themes, disciplinary context and research phases, rather than maximizing them in the sense of ‘the more participation the better’ (Box 1).

“We understand ‘participation’ as the involvement of all individuals and groups who are directly and indirectly affected by our research activities in the study regions of Thailand and Vietnam. The research process should be designed in a way that is transparent to all people and institutions involved. Thereby, it is intended to increase the sustainability of research activities and improve their results. Furthermore, the status of local stakeholders should be improved. The aim is to transform the relation ‘researcher-passive research recipient’ into ‘researcher-research partner’. In this process, forms and intensity of ‘participation’ can vary according to research topics and different phases of the research program.” Box 1 Definition of participation in the Uplands Program

Hence, we do not regard conventional research and participatory approaches as necessarily antagonistic (cf. Rhoades and Nazarea in this issue). Instead, our aim is to explore the potential and limitations of combining scientific methods with participatory approaches in innovative ways. 1.2.3 Interdisciplinarity Interdisciplinarity is another key feature of the Uplands Program in its effort to integrate the three dimensions of sustainability in the research activi(1)

Sustainable land use

ties in the mountains of Mainland Southeast Asia. The overarching project area A, ‘Participatory research’, was developed to support the crosssectional tasks of ‘participation’ in all research components of the Uplands Program, while also investigating its difficulties and limits and making contributions to the theoretical basis of participatory approaches (Figure 3). Besides project area A, six other project areas (B-G) are divided into two large thematic fields, which are closely connected to each other:

(2)

Sustainable rural development

A: Participatory research D: Integrated production systems B: Soil and water conservation

E: Processing and marketing

G: Farming systems in a regional context

C: Biodiversity in agroecosystems

F: Rural institutions and policy measures

Interdisciplinary, interinstitutional and intercultural cooperation

Figure 3 Thematic fields and project areas (A-G) of the Uplands Program

309

Andreas Neef et al.

(1) Sustainable land use In this thematic field, the research activities concentrate on possibilities for stabilizing land use systems in mountainous regions. This requires an approach which considers upstream and downstream areas as an interconnected system, and thus also examines the external effects of land use change in the highlands on agriculture and the population in the mountain valleys and lowland areas. The research activities in project area B, ‘Soil and water conservation’, focus on the possibilities for efficient use of soil and water resources, and on the impacts of existing and improved land use systems on water quality and on water and soil nutrient balance. Land suitability for different cropping systems and resource conservation measures on erosion-prone hillsides — which should also be profitable in the short run — are identified. In project area C, ‘Biodiversity in agroecosystems’, the impact of different land use practices on agro-ecological systems is analyzed. This includes modeling of land use intensification and land use dynamics. Project area D, ‘Sustainable and integrated production systems’, investigates possibilities for developing economically profitable and ecologically friendly land use systems in mountainous regions. In the more market-oriented study areas in Thailand, the development of perennial cropping systems on the basis of fruit trees is at the center of the research activities. (2) Sustainable rural development In this research theme, possibilities for integrating production systems in mountainous regions into regional economic cycles are investigated. In project area E, ‘Processing and marketing’, technologies appropriate for local processing of agricultural products which possess a high market potential in regional, national and international markets, are developed and tested. Their regional and international market potential is assessed by taking into account the whole chain from production through processing to consumption. Project area F, ‘Rural institutions and policy measures’, analyzes the institutional and political frameworks in disadvantaged mountain areas in order to identify institutions and policy measures that are conducive to enhancing sustainable rural development, taking into account ethnic and socioeconomic diversity. In project area G, ‘Farming

310

systems in a regional context’, the impact of technical and institutional innovations and changed political and economic frameworks on rural farming systems and their resource use are modeled and evaluated with regard to ecological, social and economic sustainability. Special emphasis is given to innovative Multi-Agent Systems that have the potential to link biophysical, agronomic, economic and social processes in an integrated modeling system. The seven project areas of the Uplands Program are further subdivided into research components, so-called subprojects, which are supposed to work together in an integrated, transdisciplinary way (see Figures 5 and 8 below). Two examples of such research activities and their major outcomes are presented below, drawing on the study area in northern Thailand.

2 Selected Results from Integrated and Participatory Research Work in Thailand 2.1 The study sites of the Uplands Program in Thailand The primary study sites of the Uplands Program have been Pang Ma Pha district, Mae Hong Son province, representing more subsistence-oriented mountain areas and the subject of our study on integrating local and scientific soil knowledge (section 2.2.1), and Mae Sa watershed, Mae Rim district, Chiang Mai province, as an example of a strongly market-oriented region, with intensive cultivation of fruits, flowers and vegetables (Figure 4). This watershed has been the target of various socio-economic research processes on such topics as resource tenure security and Multi-Agent Systems modeling, and is also the primary location for our on-farm experiments in the field of horticulture, water- saving irrigation, agro-ecology and integrated pest management. Fruit tree research under more ‘controlled conditions’ has been conducted mainly at the experimental station for mango and longan of Mae Jo University. The technology adoption studies presented in section 2.3.2 have been conducted in various other districts, namely Muang Lamphun, Fang, Doi Tao and Pai, representing a gradient from semi-urban to more remote mountain regions.

Journal of Mountain Science Vol 3 No 4 (2006)

Figure 4 Overview of study areas of the Uplands Program in northern Thailand

2.2 Participatory approaches to soil and water conservation in sensitive highland watersheds Soil and water conservation measures have been actively promoted in the northern Thai hillsides for several decades by national government agencies, NGOs and international development projects. At the end of the 1990s, the global initiative World Overview of Conservation Approaches and Technologies (WOCAT) had revealed 22 technologies and 14 approaches for soil conservation available for extension and implementation in Thailand, ranging from mechanical structures to biological methods (El-Swaify et al. 1999). Due to high implementation costs and slow returns to investment, rates of adoption have remained miserably low, however.

El-Swaify et al. (1999: 20f) concluded that the region “needs substantive investment in more community-based, participatory programs on a ‘landscape’ (ecosystem) basis, backed by appropriate policies and utilizing the best indigenous and scientific knowledge”. In the Uplands Program, we regard soil and water conservation as a socioeconomic-technical issue that needs to be addressed in an integrated approach by different disciplines, such as soil science, economics and rural sociology (Figure 5). Our underlying hypothesis is that low rates of adoption of soil and water conservation measures are the consequence of a whole set of factors, such as insecure resource tenure, lack of reliable data on soil properties and hydrological dynamics, neglect of local knowledge by decision-makers and diverse economic constraints facing farmers.

311

Andreas Neef et al.

Soil Science (Soil Classification and Mapping)

Political Economy (Resource Rights & Policies)

Soil and Water Conservation as a SocioeconomicTechnical Issue

Rural Sociology (Participatory Research)

Agricultural Economics (Multi-Agent System Modeling) Figure 5 Integrated research on soil and water conservation in the Uplands Program

2.2.1 Integrating local and scientific methods for innovative soil mapping One of the major constraints for rational land use decision-making is the lack of detailed soil information on a watershed scale in northern Thailand, particularly on sloping land where traditional soil mapping approaches are arduous and time-consuming. The primary objective of a combined study involving three sub-projects of the Uplands Program, namely ‘Soil Classification and Mapping’, ‘Participatory Research’ and ‘Resource Rights and Policies’, was therefore to identify a more efficient way to scale up soil information to the watershed level. One of the study areas was the Black Lahu village Bor Krai, representing a typical limestone area.1) Information on local soil classification and spatial soil type distribution was collected during a joint survey using a range of participatory methods like group discussions, semi-structured interviews with key informants and participatory mapping. The first step of the knowledge elicitation was to identify farmers with outstanding local knowledge of soil, based on their practical experience. The scientists and a key informant put together a list of eleven male and female farmers whose knowledge of the soil properties of the area was reportedly

above average. In the next step, these farmers were openly asked in a group discussion about the soil types they could distinguish and the parameters for this classification. This approach gave a general overview of local soil knowledge and the results formed the basis for subsequent investigations. To obtain a more detailed and specific local soil classification, farmers were asked individually during field walks about the main soil types and properties. On site, the farmers were interviewed about key soil properties like water infiltration rate, water retention, fertility and suitability for crops (Choocharoen et al. 2005, Schuler et al. 2006). Farmers were also interviewed at their homestead. During these interviews, the collected soil samples were shown and the farmers were asked to identify and name the soils. In a final group discussion, farmers were presented with all soil samples, which they had to sort and rank according to different parameters to cross-check the former information and to obtain a group consensus about the classification method. During this ranking, it turned out that the spatial dimension was very important. Pure soil samples, detached from their environment, were not easy for farmers to distinguish. For this reason, the approach was adapted and farmers were asked to locate their local soils on a topographic map and to describe

1) Similar studies were carried out in granite and sandstone areas to cover the major petrographic units of the northern Thai highlands.

312

Journal of Mountain Science Vol 3 No 4 (2006)

the crop suitability of each soil type (Choocharoen et al. 2005, Schuler et al. 2006). The local soil map was then compared with the scientific soil map and the petrographic map of the village territory. Similar to other studies on local soil knowledge (e.g., Talawar and Rhoades 1998, Barrera-Bassols and Zinck 2003), Lahu farmers in Bor Krai classified soils according to recognizable and easily identifiable properties in the field. Their main criteria for classification were topsoil color, hardness, consistency, suitability for different crops, and yield potential. Farmers distinguished between four main soil types, black, red, orange and yellow. The scientific soil classification according to the FAO’s World Resource Base (WRB) standard came up with eight major soil types, subdivided into further subunits. The spatial, patchy patterns of the soil types on the scientific soil map was completely different from the local soil map, which was not surprising since the WRB classification is based on a mix of physical and chemical soil properties and different diagnostic horizons which contrasts with farmers’ focus on topsoil properties and easily perceivable, morphological soil characteristics. On the other hand, the local soil map showed striking similarities with the petrographic map. One explanation for the strong correlation between the local soil map and the petrographic map was that the parent materials are characterized by different iron minerals which determine their iron content

and thus have a crucial impact on the subsoil color. As most of the cultivated soils in sloping areas of northern Thailand are prone to erosion, the petrographic origin of the soils increasingly becomes a determinant of the soil color and hence correlates strongly with local soil types. In sum, these results suggest new research trajectories in future phases of our research program: to integrate local soil mapping and scientific knowledge (petrographic mapping) for scaling up soil information from the field to the watershed and landscape level (Schuler et al. 2006). In doing so, however, we need to consider the significant variability of local soil knowledge among ethnic groups which has been confirmed by our own studies in other regions of northern Thailand (Schuler et al. forthcoming) and by other authors (e.g., Douangsavanh et al. 2006, for the case of Lao PDR). 2.2.2 Resource rights, tenure security and sustainable resource management In a further integrative step, local soil maps can be overlaid with maps containing information on resource tenure (see example in Figure 6). In the case of Bor Krai we found that most fields were located in the areas with black soils, which had been characterized as the most fertile soil type by local farmers in the local soil mapping exercise, and were thus considered most suitable for upland rice, the crucial staple crop for the local people.

Figure 6 Combination of local soil map and map of land access

Source: Draft by Peter Elstner and Ulrich Schuler

313

Andreas Neef et al.

We did not find any correlation between the duration of farmers’ settlement and (1) access to fields with higher soil fertility and (2) distance between fields and homesteads. These findings point to a relatively egalitarian land distribution among the Black Lahu, where old-established families do not seem to control high-fertility plots or fields around the village which are easily accessible. Apart from the academic value of the spatial information on land access, the tenure maps could also serve to enhance land tenure security of the local communities. Most ethnic minority groups in the northern Thai hillsides do not have land titles and are thus faced with the continuous threat of eviction from sensitive watershed areas, protected forest reserves and wildlife sanctuaries. While 99 per cent of the farmers in Bor Krai, which is situated in the Pai Wildlife Sanctuary, consider themselves the legitimate owners of their land, 45 per cent of them stated that their tenure status was insecure. Only 15 per cent believed that they have the right to sell their land. Hence, there is a high demand for local tenure maps that can be used to negotiate land tenure and use rights with government agencies such as the forest department. This, of course, requires that the latter are willing to accept farmers at the ‘negotiating table’. Other findings from the study on resource tenure and management were that farmers in this area do not make long-term investments in soil conservation measures. However, 50 per cent of the Lahu farmers in Bor Krai stated that they would invest more in agriculture if they were granted legal land titles, and 40 per cent of them claimed that they would plant fruit trees. Resource tenure issues were also at the core of another research component in Mae Sa watershed, Chiang Mai province. A study conducted among 211 households in eight villages and supported by the National Research Council of Thailand found that upstream households suffered from both lower land tenure security — due to the location of upstream communities in national parks and forest reserves — and lower water tenure security — as a result of limited water storage and conveyance systems and the imperative to share water with their more powerful downstream peers (Sangkapitux and Neef 2006). Soil conservation measures are considered vital for sustaining agricultural productivity and

314

environmental services in the watershed areas by providing both on-site and off-site benefits. When looking into the effects of various parameters on the propensity of farmers to invest in soil conservation measures, we found that, among the upstream households, farmers who enjoyed a relatively high land tenure security and had access to credit were more likely to invest in soil conservation (Equation 1 in Annex). When the quality of soil was perceived as good or medium by upstream farmers, this had a negative impact on the adoption of soil conservation measures, as the costs of their implementation are quite substantial. In downstream communities, only the type of ownership and use rights had a significant impact on investment in soil conservation measures (Equation 2 in Annex). Results obtained from the multiple regression analysis also showed the tendency that water security negatively affects soil conservation activities in both upstream and downstream communities, although this effect was not statistically significant. Enhanced water security would discourage farmers from engaging in soil conservation since the opportunity costs of the reduction in potential cropping areas by measures such as terracing and planting of grass-strips or hedgerows are considered too high. Another important finding of our studies on resource tenure and resource management in Mae Sa watershed was that farmers consider planting of fruit trees, particularly litchi, as a strategy to improve tenure security, since the National Park authorities acknowledge these practices as being at least more sustainable than growing vegetables and other annual crops (Neef et al. 2004). However, the results of the study should not be interpreted as favoring the maintenance of the status quo in areas with weak enforcement of local people’s land use rights. While the presence of fruit trees and other perennial crops can temporarily enhance tenure security, tree planting does not provide full protection of upland people against interventions from government agencies and encroachment by lowlanders. Several cases are reported in which trees were cut down by park rangers or by angry lowlanders who claimed that the Hmong use too much water for their orchards. There is also the risk that non-productive and older trees are not removed due to the fear of losing land use rights, leading to inefficiencies in resource use.

Journal of Mountain Science Vol 3 No 4 (2006)

Conditional land titles in protected areas for farmers who are willing to engage in long-term agroforestry systems and fruit plantations are often suggested as a possible solution. However, the experiences with granting conditional land rights to individual land managers have been disappointing in Southeast Asia, largely due to the problem of enforcing these conditions (personal communication from David Thomas). A more promising approach could be the allotment of rights to entire communities, who would have to guarantee the compliance of their members with the rules and regulations associated with the grant of those rights. 2.2.3 Integrating biophysical and social science data in Multi-Agent Systems The data gathered by several Thai and German scientists in the field of resource tenure, soil science and land use and integrated into a Geographic Information System (GIS) could be further processed in a participatory type of MultiAgent Systems (MAS) modeling. MAS modeling is an innovative research tool that can be used not only to create virtual societies acting on a given environment, but also has the potential to support negotiation and decision-making processes in multi-stakeholder contexts (e.g., Bousquet et al. 2005, Trebuil et al. 2005). Stakeholder involvement in agent-based modeling and simulation can take various forms: (1) role plays can provide crucial information on behavioral parameters that feed into the model (game theoretical background); (2) stakeholders can validate assumptions of the model and help to refine the model by providing feedback (soft systems validation); (3) stakeholders can be involved as players in a more interactive version of the model where they set their own rules and can change parameters (e.g., Barreteau 2003); and (4) models can be used as a decision-support tool by presenting the consequences of alternative scenarios to the stakeholders. This latter type of participatory MAS modeling was employed in a case study of water allocation between two villages in a subcatchment of the Mae Sa watershed, Chiang Mai province (see map, Figure 4). The two villages (a Hmong village upstream and a northern Thai village downstream) share water from the same creeks and have had a history of water scarcity and water conflicts. A

multi-agent simulation model was developed to analyze the impact of different land use and water management options on water scarcity for the different water users (Figure 7). Three rounds of participatory simulation sessions were organized, during which model improvements and scenarios to be tested for the next session were discussed with the stakeholders (Becu et al. 2006). A big challenge for the scientists was to balance the power differentials and social tensions that became evident in the joint meetings. The location of the upstream Hmong community within the boundaries of the Doi Suthep-Pui National Park — which makes the villagers ‘illegal residents’ in the strict interpretation of the law — and their low social status as ‘hill tribes’ puts them in a vulnerable position where they cannot afford a conflict with their downstream peers, who live mostly outside the national park’s boundaries and enjoy a much higher level of tenure security and political capital. Hence, at one time, separate village meetings were favored over meetings with representatives from both villages, in order to allow both sides to express their views and concerns freely. Our findings suggest that it takes considerable time to reach a common understanding of the model’s interface, its purpose and the differences between scenarios and reality (Becu et al. 2006). However, by participating in various sessions, villagers could gradually capture the complexity of the model and perceive its potential benefits in supporting a dialogue between different stakeholders. An interesting outcome of the sessions was that respondents in both villages finally agreed that upstream villagers were more affected by water scarcity than farmers in the downstream community, which contrasted with initial perceptions regarding upstream villagers exploiting water at the expense of downstream users, but confirmed the results of our studies of water tenure security in the watershed. One of the lessons learned from this pilot study on the potential of participatory simulation tools for enhancing resource conservation and resolving resource conflicts is that the role of the scientists as mediators in such a setting can be crucial. Respondents in the upstream village confirmed during an ex-post assessment that they had perceived us as neutral facilitators, which

315

Andreas Neef et al.

created space for a more balanced dialogue with their downstream peers and enabled them to make their voices heard without having to fear negative repercussions (Becu et al. 2006). One challenge that remains unresolved is the issue of scaling-up

of such approaches. It is evident that such a time-consuming and knowledge-intensive MAS modeling exercise cannot be replicated in each subcatchment of northern Thailand where conflicts over natural resources occur.

Figure 7 Spatial interface of the model (left) and visualization of water scarcity outcomes for different stakeholder groups in the two villages (right)

Month

Season

Year

Upstream fields (Hmong village)

Moderate lack of water

Upstream fields (Thai village)

Serious lack of water

Downstream fields (Thai village) Household water (Hmong village)

Sufficient water

Water company (Thai village) Source: Draft by N. Becu and A. Neef

2.3 Interdisciplinary approach to integrating fruit trees and related innovations into upland agricultural systems Given the susceptibility of sloping lands in mountain watersheds to erosion, the Uplands Program considers agroforestry systems based on fruit trees as viable land use systems at the research sites in northern Thailand and northern Vietnam. In Thailand, in upper watersheds of class 1 and 2 (high elevation, steep slopes) the total area must be kept under forest according to regulations of the forest department (e.g., Thomas 1995). In less erosive uplands (class 3), fruit tree plantations combined with soil conservation measures are allowed, since they fulfill similar functions as forest trees, at least in stabilizing erosion-prone soils. For farmers, fruit trees can serve multiple functions:

316

they provide higher tenure security (see section 3.2), generate relatively high economic returns, are generally less labor-intensive than annual crops, and control soil erosion. Regular pruning of fruit trees can also provide fuelwood, which reduces pressure on adjacent forest areas. Several types of subtropical fruit crops have been selected for more intensive studies in the Uplands Program: litchi (Litchi chinensis, Sonn.) as a typical upland crop for altitudes above 800 m, longan (Dimocarpus longan, Lour.) for the foothills and mountain valleys below 600 m and mango (Mangifera indica, L.) as a crop with a relatively wide range of adaptability to both mountain and lowland conditions. Here again, an interdisciplinary approach is needed to lay the scientific foundations for successfully integrating fruit trees into the complex upland agricultural systems (Figure 8).

Journal of Mountain Science Vol 3 No 4 (2006)

Horticulture (Crop Physiology)

Agro-Ecology (Integrated Pest Management)

Food Technology (Value-added Fruit Products)

Plant Nutrition (Micro-nutrients)

Integrating fruit trees into upland farming systems

Marketing (Consumer Preferences, Free-Trade Agreements)

Engineering (Water-saving Irrigation)

Rural Sociology (Innovation Adoption)

Farm Economics (Farming Systems Research)

Figure 8 Integrated research on off-season fruit production in the Uplands Program: Research components, scales and dimensions of sustainability

In the following section, we present the different integration concepts in terms of the three dimensions of sustainability, namely ecological, social and economic (cf. Figure 2 in section 1). 2.3.1 Integration of disciplines at the field and landscape level: the ecological dimension For more than six years horticultural research in the Uplands Program has addressed the problem of alternate bearing — which means that a year with high yields is usually followed by a year with drastically reduced yields. Since flower induction is a very complex process involving an interplay of climate variables, plant physiological parameters, and diverse stress factors, the phenomenon of alternate bearing has not been fully understood to date, although scientists have made significant progress in recent years (e.g. Bangerth in press, Naphrom et al. 2004). The accidental discovery by a farmer in northern Thailand, however, that KClO3 has a positive effect on flower induction in longan2) has triggered a ‘revolution’ in the longan sector. Reducing alternate bearing of fruit trees has

been very successful with longan (by applying potassium chlorate - KClO3) and, to a lesser extent, also with mango (using paclobutrazol). This enabled investigation of the regulatory mechanisms of flowering, such as acquisition and distribution processes of external compounds and enzymatic and hormonal activities. These two species are likely to be the first tropical fruit crops in which flowering behavior can be more or less reliably manipulated. These flower-induction technologies can be used (1) to ensure a high and regular yield over many years within the ‘traditional’ harvest season, and (2) to produce off-season fruits that can fetch higher prices in local and international markets. Application of KClO3 produces such reliable results in longan trees that harvesting all year round is now possible (Manochai et al. 2005). However, scientists at Mae Jo University have also found that the continuous use of high doses of KClO3 can have serious environmental consequences, such as accumulation in the soil and a negative impact on soil micro-organisms (personal communication from Somchai Ongprasert). Similarly, the growth

2) The farmer had observed that, after he had rinsed the residues from firecracker production (containing KClO3) next to a longan tree, this tree started to blossom out of season.

317

Andreas Neef et al.

retardant paclobutrazol used for flower induction in mango has adverse effects on the environment through its long-term persistence in soils (Neidhart et al. 2006a). Mango exports have recently come under increasing scrutiny by countries such as Japan where agrochemical residues are strictly regulated and high standards of food safety apply. Hence, while flower-induction technologies (FIT) have enhanced understanding of the complicated physiological processes involved in the flowering of tropical fruits, it is important to develop environmentally friendly flower manipulation methods that will reduce the alternate bearing phenomenon and enhance off-season fruit production using less of such potentially hazardous agrochemicals or none at all. A promising alternative is the technique of partial-root zone drying, which is a water-saving irrigation method that puts the fruit tree under temporary water stress (e.g. Spreer et al. 2006) and has also been found to play a role in inducing flowering through its effect on plant hormone regimes. Further research by the agricultural engineering group of the Uplands Program is needed to confirm whether this technique can be applied as a viable and environmentally friendly method to manipulate flowering in subtropical fruit trees. A study on micronutrient deficiencies in fruit trees conducted by Thai and German agronomists in the Uplands Program under controlled conditions in the greenhouse, where boron- and zinc-deficient plants started to flower unexpectedly, has also suggested that micronutrient supply is an important factor in flower induction (Roygrong et al. 2006). The question is whether off-season flowering can be induced through short-term regulation of the plant’s nutritional status. However, the authors state that the benefits of flower induction using this kind of stress may be outweighed by the negative effects of micronutrient deficiencies, such as growth depression and necrosis of young leaves (Roygrong et al. 2006). Carrying out experiments in the greenhouse and on research stations is often a lengthy process. Traditionally, agricultural scientists have tended to test certain technologies during several seasons under so-called ‘controlled conditions’ before they are released for on-farm testing. Real-life problems, however, such as the phenomenon of alternate bearing in fruit trees, or excessive use of scarce

318

water resources for irrigation, cannot afford to wait years for eventual solutions. It has therefore been the policy of the Uplands Program to conduct on-station and on-farm trials simultaneously to save time and allow for synergies between ‘controlled’ experiments and more ‘realistic’ farmer-field trials under the heterogeneous conditions of the mountainous regions of Thailand. Through intensive interactions with fruit growers in the highlands, it has also been found that off-season fruit production is a natural phenomenon in some areas with particular micro-climatic conditions. Since cooler temperatures are a major factor for flower induction (e.g. Naphrom et al. for mango, Chattrakul 2005 for litchi), current research in the Uplands Program focuses on the potential of cultivating fruit trees at higher altitudes in order to stimulate natural off-season flower induction and reduce the amount of agrochemicals currently used in litchi, longan and mango production. Farmer-managed on-farm experiments are initiated to determine the agroecological and economic niches for such cropping practices. Reduction of agrochemicals, namely pesticides, has also been the focus of another on-farm research component of the Uplands Program focusing on balancing pests and natural enemies in litchi orchards by different strategies for management of undergrowth vegetation. The results from this study suggest that spatially and temporally stochastic removal or disturbance of the undergrowth vegetation can “generate patch-dynamics of distinct successional stages of vegetation” (Euler 2006) that provide suitable ecological conditions for enhanced biodiversity and a balance of pests and beneficial insects. Evidence also suggests that landscape elements, such as forests, surrounding litchi orchards in mountain watersheds play an important role as (re)colonization sources for both beneficials and certain pest insects (Euler 2006). Off-season production of litchi, longan and mango certainly adds further to the complexity of pest management. Farmers and extension officers in Lamphun province reported that the temporal occurrence of pests in their longan orchards has increased since the introduction of off-season technology in the late 1990s because the major pests benefit from a continuous ‘feed supply’ throughout the year.

Journal of Mountain Science Vol 3 No 4 (2006)

The examples presented above show the complexity of factors that need to be considered in enhancing productivity of subtropical fruit trees, both in-season and off-season, using ecologically sustainable technologies. This calls for close cooperation between horticulture, plant physiology, plant nutrition, agro-ecology and agricultural engineering at the field and landscape scale. The natural science and technical projects, however, cannot work in isolation from economics and other social sciences, when it comes to farmers’ adoption and when we move along the value chain from production to processing, marketing and consumption. 2.3.2 Adaptation and adoption strategies of fruit growers: a social science perspective In recent years the classical ‘adoption and diffusion of innovations’ perspective, which became particularly prominent as a result of the work of Rogers (1962, 2003), has come under increasing scrutiny. The criticism has focused both on its theoretical shortcomings and on the intervention practices it has fostered. Specific concerns have been raised regarding the uncritical, pro-innovation stance of its proponents, the underlying linear model of the innovation process, a biased perception of innovativeness, and the ‘mechanical’ nature of innovation diffusion. This simplistic perspective is increasingly being replaced by a view of ‘innovations’ as multidimensional, collective and non-neutral. Innovation processes from this perspective are based on interaction, communication, conflict and negotiation between various individual and institutional actors involved, rather than on the transfer of a fixed technological or institutional package (cf. Leeuwis 2004). The introduction of new technologies in longan, mango and litchi production systems has been a focus of social scientists and economists in the Uplands Program as shown in a study of the introduction of flower induction technologies (FIT) in longan (by application of KClO3) and in litchi (by girdling, i.e. the removal of a strip of bark tissue) conducted by Neef in 2002 and 2004. For each crop, three communities distinguished according to their closeness to urban centers (semi-urban, rural, remote) in several districts of northern Thailand

(namely Muang Lamphun, Chom Thong and Doi Tao for longan, and Mae Rim, Pai and Fang for litchi) were purposively selected, and all fruit growers in each community were interviewed by means of a standardized questionnaire. The findings suggest that adoption of flower induction technology by KClO3 in longan was extremely rapid. From its accidental discovery in 1998 until the time of the first survey on longan in 2002, 92.3 per cent of the interviewed farmers in the three districts had already adopted the technology. When we interviewed the same farmers two years later (2004), all farmers had applied KClO3 in their longan orchards at least once since the introduction of this technology. Farmer-to-farmer exchange was the most important factor in the diffusion and adoption of the new technology among longan farmers, followed by the agricultural extension service and traders/shopkeepers (Figure 9). Other farmers (relatives, friends, neighbors) were also the main source of information for litchi growers on the girdling technique to stimulate flower induction. While the extension service organized regular training workshops for Thai litchi producers in the rural region (Fang district), the Hmong litchi growers in the peri-urban area (Mae Sa watershed, Mae Rim district) and in the remote region (Pai district) were neglected by agricultural extension officers, reflecting the strong social disparities in Thai-Hmong relations. Table 1 shows the adoption behavior of longan and litchi farmers with regard to Flower Induction Technology (FIT) in different regions. Among the longan farmers, the majority in the semi-urban and rural region initially tried out the new technology only on some trees before adopting FIT in the whole orchard. Longan farmers in the remote region were more risk-averse; half of them observed the experiments of their friends and neighbors before starting to adopt. A multinomial logit regression analysis identified sufficient availability of labor, farm and market experience of longan growers, income, and farm size as the most determining factors of FIT adoption. Large farm owners were able to adopt KClO3 in selected parts of their orchards, which enabled them to use the technology for producing off-season fruits. Smallholders, in contrast, mostly used the FIT to increase the yields in the traditional harvest season (June-August), when prices are generally very low.

319

Andreas Neef et al.

Therefore, the major beneficiaries of this technology are the larger producers, who are thus able to capture the windfall profit, while the small fruit growers are increasingly trapped in the price squeeze: they have to spend more on input supply (KClO3) and labor, while suffering from declining prices due to the oversupply of longan in the main

90

harvest season. This phenomenon has been described by some scholars as the ‘agricultural treadmill’ (e.g. Röling 2003). Hence, what may be regarded as a huge success in terms of technology adoption has a differential economic and social impact on the adopters.

Figure 9 Sources of information influencing adoption behavior of Flower Induction Technologies (FIT) by longan farmers in three regions (n = 57) Peri-Urban

80

Rural

Remote

70

in percent

60 50 40 30 20 10

rc e

s

ks

ou th e O

Le

rS

ts /B af le

ia m ed M as

s

rs ad e Tr

oo

(R ad io )

s er ke ep op /S h

ra lE ul tu Ag ric

O th e

rF

xt

ar m

en s

er s

io n

0

Source: Field survey 2005 Table 1 Adoption behavior of farmers applying Flower Induction Technologies (FIT) in longan (application of KClO3) and in litchi (girdling) in different regions Study regions

Longan-producing regions (Adopting farmers; n = 60)

Litchi-producing regions (Adopting farmers: n = 60)

Semi-urban (Thai)

Rural (Thai)

Remote (Thai)

Average

Semi-urban (Hmong)

Rural (Thai)

Remote (Hmong)

Average

5.0

25.0

5.0

11.7

45.0

25.0

66.7

40.8

2. Use FIT on some trees prior to full adoption

55.0

40.0

10.0

35.0

50.0

20.0

33.3

34.7

3. Observe the experiments of other farmers first

15.0

20.0

50.0

28.3

5.0

55.0

0.0

24.5

25.0

15.0

35.0

25.0

0.0

0.0

0.0

0.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

100.0

Adoption behavior 1. Use FIT in the whole orchard immediately

Combination of adoption strategies 1.-3. Total

Source: Field survey 2002

320

Journal of Mountain Science Vol 3 No 4 (2006)

76.8 per cent of the longan farmers interviewed in 2004 stated they have experience with producing off-season longan, and 85.7 per cent of these farmers reported that they could fetch higher prices. However, the share of off-season longan still stands very low at an estimated 10 per cent of the total annual longan production (personal communication from Somchai Ongprasert). Aside from small orchard sizes, the major reasons given by the non-adopters for not producing off-season longan were (1) concerns about reduced fruit quality, (2) skepticism as to whether the technology would work, and (3) concerns about increased occurrence of pests and diseases. The quality concerns were confirmed by longan farmers with off-season experience, since 46.5 per cent of them stated that they harvested smaller fruits, while 14.0 per cent claimed that the fruit quality was reduced in terms of color, taste or sweetness. The phenomenon of smaller fruits in off-season production is an effect of excessive fruit set due to KClO3 application, a problem that is currently being addressed in the third phase by horticultural scientists in the Uplands Program. Off-season production of litchi still cannot be reliably induced by FIT; the practice of girdling can only support the effect of cool temperature on flowering behavior and bring the harvest forward to a slightly earlier date (Chattrakul 2005). As regards FIT adoption in litchi, an interesting finding was that a much higher share of litchi farmers in the semi-urban and remote regions of the northern Thai highlands, all of them belonging to the Hmong ethnic minority, adopted the girdling technology in the whole orchard compared to their Thai peers in the rural area (Table 1). These results contrast sharply with the common view of ethnic minority highlanders as being more risk-averse and less innovative than the Thai majority rural population. The innovativeness of Hmong farmers was also confirmed by a study by Uplands Program economists, who found extremely rapid adoption processes for a range of innovations as diverse as new pruning techniques for litchi, irrigation sprinklers, and cell phones (Schreinemachers et al. 2006). By employing social network analysis, the study also showed the varying impact of communication, interpersonal networks and social capital on innovation adoption, depending on the attributes of the innovation in question.

2.3.3 Considering the value chain from production to processing and marketing: the economic dimension As the study on adoption behavior has shown, the integration of fruit trees and related technical innovations into upland farming systems is not as straightforward and simple as it appears. There are significant trade-offs to be considered when taking into account the economic dimension of sustainability. Widespread monocropping of litchi and tangerine in upland areas has not only had adverse effects on the environment, it has also made fruit growers much more vulnerable to price fluctuations. The Free Trade Agreement for fruits and vegetables concluded between Thailand and China in October 2004 has benefited the longan farmers — for whom China is an important market — to a certain extent, but has caused an influx of temperate and subtropical fruits, such as apples, peach and litchi, from Southwest China, and this has drastically reduced prices for locally produced fruits. On the other hand, Thai exports to China are facing a number of trade barriers, both tariff types (such as provincial taxes) and non-tariff types (such as regulations on chemical residues in fruits), as a study by the Thai marketing research group of the Uplands Programs found (Daroonpate and Sirisupluxuna 2006). Thai economists in the Uplands Program have employed regional modeling approaches based on multi-goal programming, which optimize farm management under consideration of the three major sustainability dimensions. The major recommendations from the study in different regions of northern Thailand were that upland farmers in market-oriented regions should diversify their cropping patterns in order to be more resilient to market dynamics in an increasingly globalized environment (Praneetvatakul and Sirijinda 2006). Diversification is also at the center of the Uplands Program’s research in the field of fruit processing. It aims at the use of various quality types and size classes of raw fruits and at the reduction of post-harvest losses resulting from seasonal overproduction and restricted access to markets. Research by Thai and German food scientists found that the wide diversity of mango cultivars grown in the northern Thai uplands offers interesting and novel options for product differentiation, e.g. mango purees, rich in pro-

321

Andreas Neef et al.

vitamin A, combined with pectin and polyphenol recovery from peels (Neidhart et al. 2006b). Post-harvest scientists in the Uplands Program have also identified methods to optimize fruit drying technologies for longan and litchi that can be adopted by small and medium-sized enterprises or farmers’ cooperatives. In order to enable small-scale farmers to reap the benefits of the technological innovations in fruit production and processing, however, new institutional arrangements need to be put in place. Until now, the many thousands of small-scale longan, litchi and mango growers in northern Thailand have had little influence on determining prices or adding value to their products. More than 50 per cent of the fresh longan is traded via middlemen (local and district assemblers), for instance, and only a small fraction is sold by farmers directly to consumers or retailers (Isvilanonda in press). Contract farming arrangements with big agro-processing companies, e.g. for canning, provide farmers with a certain security, but at the expense of relatively low prices. Processing and marketing cooperatives are virtually non-existent in the northern Thai highlands. Future studies need to focus on the potential for collective action and enhancing social capital among mountain farming communities to make them play a more active role in the postharvest/marketing sector rather than just being suppliers of agricultural raw materials.

3 Discussion and Conclusions The examples of participatory approaches to soil and water conservation provided here have shown the value of integrating local knowledge and local perspectives into research approaches aimed at sustainable resource use in mountainous regions. Conventional soil surveys could be blended with local soil knowledge which provides more relevant soil information for local stakeholders and opens up new opportunities for scaling-up soil data to the watershed and landscape level. Classical farm surveys on land and water tenure combined with biophysical data and GIS information can be integrated into participatory Multi-Agent Systems models, through which scientists and local stakeholders can jointly work out scenarios for

322

improved resource allocation and water conservation. A welcome ‘by-product’ of the rather extractive research on resource tenure was the map of land access and customary ownership that villagers can potentially use to negotiate their land rights with government agencies in protected areas. Our integrated research on off-season production of subtropical fruit trees and its introduction into upland farming systems shows that the research mandate of agricultural scientists does not end with the ‘successful’ introduction of a technological innovation. On the contrary: new research questions emerge due to the differential impact on different groups of farmers and the significant trade-offs between the ecological, social and economic sustainability dimensions of most productivity-enhancing technologies. To avoid the harmful ‘agricultural treadmill’ effect that benefits mainly large farmers and the food industry rather than smallholder agriculture, technical and institutional innovation processes have to go hand in hand. Integrated research in mountainous regions, in particular, needs sustained and longterm commitment on the part of scientists, local stakeholders, policy-makers, and funding organizations. This program has been fortunate enough to obtain funding for a long-term period (nine years) and it might possibly be extended to a total of 12 years, which is a notable exception in national and international agricultural research funding. The major lesson from the examples provided in this paper is that conventional and participatory research approaches can not only combine well, they can support each other and be ‘enriching’ for both scientists and local stakeholders (cf. Rhoades and Nazarea in this issue), in contrast to the gulf that is often alleged to exist between classical, cutting-edge scientists and participatory, grassroots researchers. In truly transdisciplinary research towards sustainable land use, resilient livelihoods and healthy ecosystems in mountainous regions, we need to cross the boundaries of our own disciplines and integrate different knowledge domains.

Acknowledgements We would like to thank Sukit Kanjina, Pakakrong Makpun, Rattana Kouy-Charoenpanit,

Journal of Mountain Science Vol 3 No 4 (2006)

Patcharin Inthanu, Key Nanthasen, Wipawadee Saepueng, Teeka Yothapakdee, for their help in data collection, data entry and analysis. We are indebted to David Thomas for his useful comments on an earlier draft of this paper.

The financial support of the Deutsche Forschungsgemeinschaft (German Research Foundation) and the co-funding of the National Research Council of Thailand are gratefully acknowledged.

References

Hegele M., Naphrom D., Manochai P., Chattrakul A., Sruamsiri P. and Bangerth F. 2004. Effect of leaf age on the response of flower induction and related hormonal changes in Longan trees after KClO3 treatment. Acta Horticulturae 653: 41-49. Isvilanonda S. in press. Fresh longan marketing and reference market: a case of longan grown in northern Thailand. In: Heidhues F., Herrmann L., Neef A., Neidhart S., Sruamsiri P., Dao Chau Thu and Valle Zárate A. (eds.), Sustainable land use in mountainous regions of Southeast Asia: Meeting the challenges of ecological, socio-economic and cultural diversity. Springer-Verlag, Berlin, Heidelberg, New York, London, Paris and Tokyo. Leeuwis C. 2004. Communication for Rural Innovation. Oxford: Blackwell Publishing. Manochai P., Sruamsiri P., Wiriya-alongkorn W., Naphrom D., Hegele M. and Bangerth F. 2005. Year around off season flower induction in longan (Dimocarpus longan, Lour.) trees by KClO3 applications: potentials and problems. Scientia Horticulturae 104: 379-390. Naphrom D., Sruamsiri P., Hegele M., Boonplod N., Bangerth F. and Manochai P. 2004. Hormonal changes in various tissues of mango trees during flower induction following cold temperature. Acta Horticulturae 645: 453-457. Neef A. 2005. Participatory approaches for sustainable land use in Southeast Asia: an overview. In: Neef A. (ed.), Participatory approaches for sustainable land use in Southeast Asia. Bangkok, Thailand: White Lotus Press, pp. 3-32. Neef A., Sirisupluxuna P., Sangkapitux C. and Heidhues F. 2004. Tenure security, long-term investment and nature conservation – Getting causalities and institutions right. Proceedings of the 85th Seminar of the European Association of Agricultural Economists ‘Agricultural Development and Rural Poverty under Globalization: Asymmetric Processes and Differentiated Outcomes’, Florence, Italy, 8-11 September 2004. Neidhart S., Jaradrattanapaiboon A., Reintjes K., Jöns B., Leitenberger M., Ingwersen J., Kahl G., Sruamsiri P., Streck T. and Carle R. 2006a. Which risks do result from the application of paclobutrazol in off-season mango production regarding residues in fruit and soil? First results of a long-term field study in northern Thailand. Poster presented at the International Symposium ‘Towards Sustainable Livelihoods and Ecosystems in Mountainous Regions’, Chiang Mai, Thailand, 7-9 March 2006 (abstract on CD-rom). Neidhart S., Vásquez-Caicedo A.L., Sirisakulwat S., Schilling S., Sruamsiri P. and Carle R. 2006b. Integrated mango processing into pulp products rich in pro-vitamin A and value-adding by-products. Proceedings of the International Symposium ‘Towards Sustainable Livelihoods and Ecosystems in Mountainous Regions’, Chiang Mai, Thailand, 7-9 March 2006 (on CD-rom). Praneetvatakul S. and Sirijinda A. 2006. Modeling sustainability of highland agricultural systems in northern Thailand. Proceedings of the International Symposium ‘Towards Sustainable Livelihoods and Ecosystems in Mountainous Regions’, Chiang Mai, Thailand, 7-9 March 2006 (on CD-rom). Röling N. 2003. From causes to reasons: the human dimension of agricultural sustainability. International Journal of

Bangerth F. in press. Flower induction in perennial fruit trees: Still an enigma? Acta Horticulturae. Barrera-Bassols N. and Zink J.A. 2003. Ethnopedology: a worldwide view on the soil knowledge of local people. Geoderma 111: 171-195. Barreteau O. 2003. The joint use of role-playing games and models regarding negotiation processes: characterization of associations. Journal of Artificial Societies and Social Simulation 6(2): Online: http://jasss.soc.surrey.ac.uk/6/2 /3 .html. Becu N., Neef A., Sangkapitux C., Kanjina S. and Schreinemachers P. 2006. Participatory Multi-Agent System simulation to support collective decision-making in water management: potential and limits of stakeholder integration. Paper presented at the Symposium ‘Integrating Participatory Approaches into Economic Modeling, Assessment and Valuation for Agricultural Development and Natural Resource Management’ organized at the Triennial Conference of the International Association of Agricultural Economists (IAAE), Gold Coast, Australia, August 12-18, 2006. Bousquet F., Trébuil G., Boisseau S., Baron C., d’Aquino P. and Castella J.-C. 2005. Knowledge integration for participatory land management: The use of multi-agent simulations and a companion modelling approach. In: Neef A. (ed.), Participatory approaches for sustainable land use in Southeast Asia. Bangkok: White Lotus. Pp. 291-310. Chattrakul A. 2005. Mechanism of physiological response of litchi when flowering under low temperature condition. PhD thesis, Chiang Mai University, Chiang Mai, Thailand. Choocharoen C., Schuler U., Elstner P. and Neef A. 2005. Blending local and scientific knowledge for soil classification and soil mapping: A case study from a Black Lahu village in Mae Hong Son province in northern Thailand. Paper presented at the 4th International MMSEA Conference “Sustainable Use of Natural Resources and Poverty Dialogue in Mainland Montane South-East Asia” in Sa Pa, Vietnam, 16-19 May 2005. Daroonpate A. and Sirisupluxuna P. 2006. Domestic Fruit Consumption and Fruit Exports of Thailand. Final Report submitted to National Research Council of Thailand (in Thai). Douangsavanh L., Manivong V., Polthanee A., Katawatin R. and Inoue Y. 2006. Indigenous knowledge of soil classification of ethnic groups in Luang Prabang Province of the Lao PDR. Journal of Mountain Science 3(3): 247-258. El-Swaify S. with an international group of contributors 1999. Sustaining the global farm – Strategic issues, principles and approaches. Honolulu, Hawaii, USA: International Soil Conservation Organization (ISCO) and the Department of Agronomy and Social Science, University of Hawaii at Manao. Euler D., Martin K., Sauerborn J. and Hengsawad V. 2006. Management and biodiversity of litchi agroecosystems in northern Thailand: Towards sustainability on field and landscape scale. Proceedings of the International Symposium ‘Towards Sustainable Livelihoods and Ecosystems in Mountainous Regions’, Chiang Mai, Thailand, 7-9 March 2006 (on CD-rom).

323

Andreas Neef et al.

Agricultural Sustainability 1(1): 73-88. Rogers E.M. 1962. Diffusion of innovations. New York: The Free Press. Rogers E.M. 2003. Diffusion of innovations. Fifth Edition. New York: The Free Press. Roygrong S. 2006. Role of boron and zinc in flower induction of lychee (Litchi chinensis, Sonn.). Proceedings of the International Symposium ‘Towards Sustainable Livelihoods and Ecosystems in Mountainous Regions’, Chiang Mai, Thailand, 7-9 March 2006 (on CD-rom). Sangkapitux C. 2006. Tenure and Valuation of Water Resources as Issues for Sustainable Resource Management: A case Study of Mae Sa Watershed, Mae Rim District, Chiang Mai, Thailand. Final report submitted to the National Research Council of Thailand (in Thai). Sangkapitux C. and Neef A. 2006. Assessing water tenure security and livelihoods of highland people in northern Thailand. Quarterly Journal of International Agriculture 45: 375-394. Schreinemachers P., Tipraqsa P. and Berger T. 2006. Assessing innovations and sustainability strategies with multi-agent systems. Proceedings of the International Symposium ‘Towards Sustainable Livelihoods and Ecosystems in Mountainous Regions’, Chiang Mai, Thailand, 7-9 March 2006 (on CD-rom). Schuler U., Choocharoen C., Elstner P., Neef A., Stahr K., Zarei M. and Herrmann L. 2006a. Soil mapping for land-use planning in a karst area of N Thailand with due consideration of local knowledge. Journal of Plant Nutrition and Soil Science, 169: 444-452. Schuler U. Choocharoen C., Weiss A., Herrmann L., Neef A. and Stahr K. 2006b. Elicitation of local soil knowledge in northern Thailand and consequences for land use decision-making.

Poster presented at the Second International Conference on ‘Sustainable Sloping Lands and Watershed Management: Linking Research to Strengthen Upland Policies and Practices’. 12-15 December 2006, Luang Prabang, Lao PDR. Spreer W., Ongprasert S., Wiriya-Alongkorn W. and Müller J. 2006. Stomatal resistance for monitoring drought stress under deficit irrigation applied to litchi (Litchi chinensis, Sonn.), mango (Mangifera indica, L.) and longan (Dimocarpus longan, Lour.) in northern Thailand. Proceedings of the International Symposium ‘Towards Sustainable Livelihoods and Ecosystems in Mountainous Regions’, Chiang Mai, Thailand, 7-9 March 2006 (on CD-rom). Talawar S. and Rhoades R.E. 1998. Scientific and local classification and management of soils. Agriculture and Human Values 15: 3-14. Thomas D. 1995. Opportunities and limitations for agroforestry systems in the highlands of North Thailand. In: Turkelboom F., van Look-Rothschild K. and van Keer K. (eds.), Highland Farming: Soil and the Future. Proceedings, December 21-22, 1995, Chiangmai, Thailand: Mae Jo University and Catholic University of Leuven. Pp. 126-160. Trébuil G., Bousquet F., Ekasingh B., Baron C. and Le Page C. 2005. A multi-agent model linked to a GIS to explore the relationship between crop diversification and the risk of land degradation in northern Thai highlands. In: Bousquet F., Trébuil G. and Hardy B. (eds.), Companion modelling and Multi-Agent Systems for integrated natural resource management in Asia. Los Baños, The Philippines: International Rice Research Institute. Pp. 167-190. UNEP 2002. Mountain Watch. Cambridge, UK: United Nations Environmental Program (UNEP) World Conservation Monitoring Centre.

Annex Multiple regression analysis of soil conservation investments (SCI) in upstream and downstream communities of Mae Sa watershed, Mae Rim district SCIUP = 0.215 + 0.332LT – 0.059Q_good – 0.073Q_medium + 2.068Wealth – 0.021WS – 0.06C (4.121)*

(8.478)*

(-1.773)**

R 2 = 0.366

R

2

= 0.342

(-2.150)*

(0.688)

(-1.495)

Durbin – Watson = 1.684

n = 167

SCIDW = 0.197 + 0.242LT - 0.054Q_good – 0.072Q_medium + 3.367Wealth– 0.055WS + 0.044C (1.726)** (4.655)*

R 2 = 0.380

(-0.642)

R

2

= 0.290

(-0.831)

(0.079)

Durbin – Watson = 1.830

Legend: SCIUP = Soil conservation investment in upstream communities SCIDW = Soil conservation investment in downstream communities LT = Land tenure expressed in type of rights Q_good = Soil quality perceived as good Q_medium = Soil quality perceived as medium Wealth = Wealth expressed in physical assets WS = Water security C = Access to credit ( ) = t-statistics of variable coefficient n = number of households in the sample * = significance at 90% confidence level, ** = significance at 95% confidence level Source: Sangkapitux 2006

324

(1)

(-2.093)*

(-1.407)

(1.170) n = 48

(2)