Search

Landscape dynamics are driven by complex interactions among ecological, social, economic, and policy factors. In conservation areas, these factors are usually related to an increasing number of diverse land resource managers and users. Land use conflicts occur frequently because they have different interest, objectives and perceptions of landscape resources and their use. Sharing these different, but all legitimate, perceptions of the landscape and its use among concerned stakeholders is a pre-requisite for better collective land management, particularly in conservation areas. This research is using the integrative companion modeling approach to co-construct an agent-based model representing the dynamic interactions between vegetation dynamics, reforestation efforts, and livestock grazing in the upper watershed of Nan province, northern Thailand. The paper focuses on the participatory modeling process implemented with local stakeholders at this site. Three main investigation tools were used to exchange and gather knowledge on ecological and human decision making processes: field surveys (land use history and analysis of vegetation dynamics at the landscape level), farmers' interviews (analysis of individual decision making and determining factors across different farm types), and institutional analysis in relation with changes in land use policy and related state interventions. This knowledge was first assembled in simple gaming exercises presented to local herders and foresters to further enrich and validate the researchers' understanding of key interactions regulating vegetation and land use dynamics. The outputs of these collaborative modeling activities were used to design a hybrid agentbased hybrid simulator blending a role-playing game and a computer program developed under the CORMAS platform. This simulation tool, representing the complex human and ecological interactions at the landscape level, allowed stakeholders to criticize and improve this comprehensiv

Sustainable management of renewable resources is often complicated by diversity and dynamic nature of the ecological and socioeconomic systems involved. As these system dynamics and interactions are highly complex and frequently unpredictable, there is a need to opt for transdisciplinary research addressing adaptive and integrated renewable resource management. Companion Modeling (ComMod) is a multi-agent systems (MAS)-based approach relying on synergistic effects between Role-Playing Games (RPG) and Agent-Based Models (ABM) to facilitate information sharing, collective learning, and exchange of perceptions to support negotiation, facilitate collective decision-making, and to strengthen adaptive resource management capacity. Iterative and adaptive sequences of field work and modeling activities allow inclusive and interactive participation of stakeholders during design, implementation, calibration, and validation steps of the models, as well as joint use to explore possible future scenarios. ComMod was implemented in a study of a conflict between two ethnic communities and a newly proposed national park in Northern Thailand. Deforestation, biodiversity conservation, and livelihoods were key issues discussed during RPG sessions, and subsequently represented in an ABM simulator. Consequently, local stakeholders learned about agro-ecological and socioeconomic dynamics and gained an increased awareness of these key issues. Mutual understanding was improved, and the importance of collaborative discussion, essential to negotiation and decision-making, became obvious. Finally, this Northern Thailand experience has shown that collaborative interactions between researchers and local stakeholders mediated by ComMod tools were supportive of improved communication among the conflicting parties and joint learning for adaptive and integrated sustainable management of renewable resources.

How can simple interactive models catalyze collective learning and action about local water management? This paper aims at providing an answer to this question by drawing on a Companion Modeling (ComMod) experiment on irrigation water sharing in a highland community of Northern Thailand. In a ComMod process, simulation models integrating different stakeholders¿ points of view on the problem at stake are developed and used as communication platforms to facilitate the collective exploration and assessment of various possible future scenarios. In this case study, the ComMod process combined a preliminary diagnostic-analysis of the heterogeneous socio-political context, a Role-Playing Game (RPG) and an associated simple Agent-Based Model (ABM). An ABM was used to run simulations to stimulate a plenary debate and, later on, to facilitate discussions within small homogeneous groups of farmers. The various effects of the process on the participants in term of learning, communication, behavior change, and new practices were evaluated through series of individual interviews. This ComMod process stimulated individual and collective learning and coordination among multiple stakeholders exploring pathways to solve their common irrigation water use problem. We show that in participatory modeling, simple models can be useful to mediate water use conflicts and accommodate multiple interests among stakeholders. To do so the participatory aspects of the modeling and simulation process must be carefully managed. In particular, much attention needs to be paid to the initial socio-political context and its power inequities to ensure the genuine involvement of all concerned stakeholders, including the usually voiceless and resource-poor ones.

La modélisation d'accompagnement a été appliquée à la gestion de deux systèmes semenciers céréaliers : le sorgho au Mali et le riz inondé de bas-fond en Thaïlande. Le principe de cette démarche participative est d'impliquer chercheurs, décideurs, agents du développement et acteurs locaux dans la construction, la discussion et la critique de modèles de leur système semencier. Au Mali, nous nous sommes intéressés à la dynamique de la diversité variétale à l'échelle du village. Sur la base d'une compréhension partagée des choix des paysans et des mécanismes d'échanges, nous avons pu définir et simuler les règles de fonctionnement de structures locales de gestion des semences. En Thaïlande, nous avons travaillé à l'échelle provinciale sur le système semencier formel. La démarche de modélisation d'accompagnement a abouti à une représentation de la circulation des semences et des informations entre les différents acteurs du système. Dans les deux cas, au-delà de la production collective de modèles du système semencier, le processus a permis d'initier des dynamiques de réflexion collective et de partenariat plus équilibré entre les acteurs du système.

Water is used and managed by stakeholders at different levels for diverse objectives, therefore understanding decision making and supporting coordination is crucial in achieving resilient water management. Companion Modeling (ComMod) is an interactive process facilitated by evolutionary models for knowledge generation and exchange, and for supporting collective decision-making. Role-playing games and computerized multi-agent simulations for focused group debates are complementary tools combined in a ComMod cycle and used at the field workshops. Agent-based modeling is used to understand how different processes in direct competition are coordinated, and to mediate the collective search for acceptable solutions to conflicting parties facilitated through exchanges. This paper compares the process of agent-based modeling applied in eight case studies with diverse natural and socio-economic conditions and different resource management problems in Bhutan, Thailand and Vietnam to show the creative thinking in developing and applying flexible ComMod modeling tools and provide lessons for their use in other situations.

In northeast Thailand, 85% of the farmers are smallholders who are unable to meet their basic needs from agricultural production only. These tiny farms survive thanks to non-farm income, which faces increased difficulties as other economic sectors ran out of steam during the recent economic crisis of the late 1990s. In this context, farmers have to rely more on their agricultural production activity and income. But how can this be made possible in a region well-known for its very constraining soil and climatic conditions? To answer this question, and to examine the whole complexity of agricultural development issues, this article proposes an analysis of recent agrarian transformations and an understanding of farmers' current practices and strategies. A diagnostic analysis of a village agrarian system located in non Kaen Province in upper northeast Thailand was carried out in 2002 by using a combination of field observations, interviews with key witnesses of the local history, and a farm survey of 26 diverse households. The recent history illustrates that farmers have shown a high adaptability to rapid changes in their economic environment, such as changing market demand for agricultural products and labour. The differentiation among farming households has led to different types of farmers with different resources, practices, and strategies. Families belonging to the most frequent type have a very tiny holding and their members are permanently or seasonally involved in unskilled off-farm activities. Their low and unstable total cash income is often insufficient to meet the family's basic needs and they become deeply indebted. As efforts to improve water availability are being made, many Isarn families could improve their living conditions by diversifying their agricultural production systems with more small-scale fruit, vegetable, fish, or livestock production to improve household food security and cash income, eventually combined with non-agricultural activities. Bu

The process of commercialisation-diversification in the highlands of upper northern Thailand and the accompanying dismissal of self-subsistence are documented based on the findings from seven case studies carried out in different agricultural and social situations during the past decade. The characteristics of the key driving forces powering this agrarian transition such as rapid economic growth, decrease in the share of labour employed in the agriculture, urbanization and changes in food consumption patterns, and improved communication infrastructures, are presented in the Thai context. The environmental impact of these profound agrarian transformations on the degradation of key renewable resources, particularly soil erosion, is assessed. Their socio-economic consequences on an extensive differentiation among farming households and equity issues are also discussed. Finally the authors draw several lessons from this Thai experience that illustrate the very strong adaptive capacity of small highland farmers. They could be useful in similar agro-ecological zones of neighbouring countries that are presently experiencing the same kind of agricultural transition in the Montane Mainland Southeast Asia ecoregion. Particularly, the article underlines the need for more holistic and integrated approaches to agricultural development and the management of renewable resources in highland agro-ecosystems to alleviate poverty while conserving the resource base.

Dans les basses terres du nord-est de la Thaïlande, les conditions bioclimatiques sont particulièrement difficiles: répartition irrégulière des précipitations, sols à texture grossière. La productivité agricole y est faible et la région est la plus pauvre du pays. Les parcelles de riz de bas-fond occupent l'essentiel des terres agricoles et les millions de paysans de la région doivent faire face à des problèmes de gestion de l'eau lors des fréquentes sécheresses ou inondations. Pour réduire les risques à la fois écologiques et économiques, les paysans mettent en oeuvre diverses stratégies. La migration est, dans ce cadre, considérée selon les points de vue comme une solution ou une contrainte. Les systèmes Multi-agent (MAS) sont utilisés pour explorer des interactions entre les dynamiques écologiques et sociales dans des systèmes complexes en créant des sociétés virtuelles partageant un même environnement et ses ressources. Nous emploierons un MAS dans une approche de modélisation d'accompagnement (ComMod) afin de mieux comprendre et de modéliser les processus de décision des exploitants concernant la gestion des ressources (terre et eau) au niveau de la parcelle, de l'exploitation, et de la répartition des eaux communautaires. Cette communication présente les caractéristiques originales et la dynamique de l'agriculture de la communauté de Lam Dome Yai dans la province d'Ubon Ratchathani. Les principaux processus de décision des exploitants vis-à-vis de l'utilisation de la terre, de l'eau ou de la migration sont expliqués. Des enquêtes d'exploitation et des Gis sont utilisés pour analyser les changements spatiaux et temporels, et définir une interface spatiale adaptée pour le futur modèle MAS. La mise en évidence des processus de décision au niveau de ménage sert à définir le comportement social des agents représentés dans le modèle. L'analyse diagnostique préliminaire montre que l'impact des migrations de travail sur la gestion des ressources renouvelables est plus

Thanks to recent advances in the field of distributed artificial intelligence, agent-based models (ABM) can now be used to run simulations of social phenomena based on their computerized representations, and to apply experimental methods in social sciences (Axelrod 1997, Gilbert and Troitzsch 1999, Jager 2000). In the field of renewable resource management and environmental sciences, several ABM simulation platforms offer the possibility to explore interactions between social and ecological dynamics (Costanza and Ruth 1998, Bousquet et al. 1998, Lansing 2002). In these complex systems, the social and economic dynamics can be viewed as a set of interactions among heterogeneous agents, generating aggregate phenomena that are different from the behaviour of groups of average individuals considered in classical economic thinking (Rouchier and Bousquet 1998). Such a view was adopted in the research presented here. The agent-based model presented in this paper was built to explore the interrelated roles of formal and informal credit in a socially heterogeneous community of small farmers exploiting a highland catchment of mountainous upper northern Thailand. Formal credit corresponds to institutionalized credit funds whereas informal credit is seen as loans settled among villagers, either without interests within networks of acquaintances, or with high interest rates when loan sharks are involved. An original characteristic of the companion modelling approach (Bousquet and Trébuil 2005, http://commod.org) and the simulation process adopted in this case study is the co-construction of the model with the farmers and the use of simulations with them in their village. The objective was to facilitate collective decision-making regarding the adaptation of the local rules for the allocation of rural credit to allow a more equitable and extensive process of expansion of perennial crops (Barnaud et al. 2005). Following a description of the methodology and tools used in this experim

Since late 2002, such a ComMod process is being tested in an Akha village of upper northern Thailand to promote collective watershed management, and particularly to examine the interactions between soil and water conservation, agricultural diversification, and social equity. Following a presentation of the context & problem, and of the original characteristics of the ComMod approach, this article describe its concrete use with Akha villagers. The article explains the way a shared representation of the watershed and its dynamics is achieved, and the way it supports collective learning on land management. The flexibility and adaptive characteristics of the gaming and simulation tools used with the people will be highlighted as they fit with the evolving nature of coordination processes in adaptive NRM. Finally, preliminary lessons from this ComMod experiment are presented as well as new perspectives for improving this approach.

This introductory chapter recalls the origins of this publication at the interface between the personal interest of several colleagues from different Southeast Asian countries and a growing interest in methodological innovation in the field of integrated natural resource management (INRM) in the Consultative Group on International Agricultural Research. The historical development of the so-called "companion modeling" (ComMod) approach relying on the use of multi-agent systems (MAS) for INRM is also described, and its main principles and objectives are defined: to develop simulation models integrating various stakeholders' points of view and to use them within the context of platforms for collective learning. The ComMod methodology used to facilitate such a process in INRM is presented, with an emphasis on the combination of key tools used with stakeholders, such as conceptual models, MAS, and role-playing games. A final section introduces the diversity of the Asian experiences presented in this book and its content.

Rainfed lowland rice varietal and seed management involves a complex system dealing with problems such as variety adoption, biodiversity, and the supply of good-quality seed. Participatory modeling of rainfed lowland rice varietal and seed management in lower northeast Thailand has been carried out to better understand the seed system and identify problems. Conceptual modeling was done through interinstitutional research team meetings, stakeholder analysis, surveying stratified random sampling of farmers and seed supply agents in Ubon Ratchathani Province, and conducting role-playing games (RPGs). The seed system was divided into three subsystems: farmers' decision-making related to rice varieties, farmers' management of seeds, and the whole existing seed supply system. A first RPG representing the first two subsystems was used at two different locations with 25 farmers. The initial findings from the RPG helped to validate and improve the conceptual model and provide a common understanding of farmers' rice varietal and seed management. Problems of limited access to or sharing of information about varieties and seed, the need for early-maturing varieties, and the scarcity of good-quality seeds were identified. A second RPG will deal with the whole seed supply system. A more comprehensive analysis of the RPG results with those of the farm survey will be done to improve the conceptual models, together with developing a multi-agent model representing the whole rainfed lowland rice seed system.

IRRI, the Center for International Cooperation in Agricultural Research for Development (CIRAD), and other CGIAR centers are working constantly to develop innovative methodologies based on new approaches and paradigm shifts. This publication presents breakthroughs in research on integrated natural resource management (INRM) - spatial modeling and adaptive management of renewable resources are key findings - based on a 2001-2004 companion modeling project in Thailand. It recommends that INRM projects satisfy the following minimum set of criteria: - Be defined in a collaborative and equitable manner with all relevant stakeholders and partners. - Generate new knowledge as international/regional public goods based on both indigenous knowledge and modern science. - Effectively communicate and disseminate results and conclusions to all stakeholders. - Reform and strengthen institutions from local to policy levels.

Integrated watershed management implies a collective management of the land reconciling ecological dynamics and social processes to ensure a viable and equitable use of renewable resources and to mitigate conflicts. Based on the integration of existing knowledge from different sources and disciplines, this chapter describes the construction of a spatially explicit multi-agent model to analyze the poorly understood interaction between the risk of land degradation and crop diversification and agricultural commercialization of heterogeneous household-based farming systems in a highland Akha village catchment of up-per northern Thailand. In this region, cash cropping on sloping land is commonly blamed by lowlanders for aggravating land degradation. But on-farm agronomic surveys led to the hypothesis that this interaction is far more complex and could be further examined by using an integrative model to explore simultaneously the interaction between the agronomic and socioeconomic components of the system. The simulated behavior of the different model entities is based on previous field observations and measurements. This agronomic simulator represents actual farmers' cropping practices at the field level under different slope and climatic conditions. The social dynamics are taken into account through the representation of three main types of households identified through a farm survey. They have contrasting historical backgrounds, are managing different amounts of resources, and correspond to a gradient of integration into commercial agriculture. Because key agro-ecological and socio-economic processes need to be simulated at different pertinent scales, this multi-agent model is loosely linked to a geographic information system (GIS) displaying the distribution of three complementary spatial entities in the catchment. Following a presentation of the selected integrative modeling approach, the model conceptualization, its architecture, and modeling sequences are describe

In this new century, there is an urgent need to integrate and organize knowledge into suitable frameworks to examine essential problems with the people involved in solving them. Recent advances in computer science, particularly distributed artificial intelligence and multi-agent systems (MAS), are creating a strong interest in using this new knowledge and technologies for various applications to better deal with the increasing complexity of our fast-changing world, particularly for studying interactions between societies and their environment. By emphasizing the importance of interactions and points of view, the MAS way of thinking can facilitate high-level interdisciplinary training and collaborative research among scientists working in ecology and social sciences to examine complex problems in the field of integrated natural resource management (INRM). This paper describes how a recent project based on a series of short courses in the field of MAS, social sciences, and INRM at three different universities in Thailand tried to transfer European expertise and research results to an Asian audience of graduate and postgraduate students and young researchers interested in innovative and action-research-oriented interdisciplinary approaches. The course structure, organization, and contents are described and assessed. The course participants are characterized and their opinions are used to evaluate the strengths and weaknesses of this very interdisciplinary training program. The first sustainable outputs and key preliminary lessons learned from this innovative collective learning experience are presented. In conclusion, the authors suggest ways to support the emergence of a regional network of "MAS for INRM" practitioners in Southeast Asia to build on the dynamics begun by this project and serve the need for such interdisciplinary training across Southeast Asia.

Growing rice is the main activity in northeast Thailand. Unpredictable droughts and coarse-textured soil are the principal constraints usually cited to explain the low yields and economic poverty of this region. Past studies tried to improve the drought tolerance of rice varieties and hydro-logical functioning at the field level. How water is used at the farm level remains largely unknown. Consequently, it is relevant to understand the interactions between the water-resource and water-use dynamics. This article proposes to develop a simulation tool based on multi-agent systems to explore adaptations of the rice cropping pattern to rain-fall variability. An environment containing the main biophysical entities involved in decision-making rules for water use is modeled and its hydrological functioning is verified. Preliminary simulations are run to illustrate the model capacities. The next step will consist of adding autonomous agents to simulate scenarios in which farmer agents may cooperate to use water.

A key challenge in integrated natural resource management (INRM) consists in reconciling ecological and socio-economic dynamics for a collective, sustainable, and equitable management of the land. Diverse stakeholders have different legitimate points of view and representations of the system they exploit and its management is better seen as a collective learning process. We have tested a "companion modeling" (COMMOD) approach combining multi-agent systems (MAS) models and role-playing games (RPG) to examine concrete INRM problems with stakeholders. This participatory process consists in repetitive back and forth steps between the model and the field to facilitate dialogue, shared learning, and collective decision-making through interdisciplinary and action-oriented research. The methodological features of this approach are de-scribed and two applications in rural Thailand are presented. Among the lessons learnt, we argue that COMMOD users need to distinguish between two specific contexts: (i) The production of new knowledge on a given complex system and (ii) Supporting evolving, iterative, and continuous collective decision-making processes. Such MAS-based and bottom-up COMMOD approach offers a great potential where land management policy frameworks encourage the decentralized management of natural re-sources.