UBON RICE VARIETIES Ubon Ratchathani Province Lower northeast Thailand

1
UBON RICE VARIETIESUbon Ratchathani
ProvinceLower northeast Thailand

• A role-playing game to understand how farmers
  manage rice varieties, and select rice seed
  suppliers
• Chirawat VEJPAS, François Bousquet Guy Trébuil
• Faculty of agriculture, Ubon Ratchathani
  University
• CU-CIRAD ComMod Project, Chulalongkorn
  University, Bangkok, Thaïlande

2
The context

• The successful dissemination of few premium
  quality aromatic rainfed lowland rice (RLR)
  varieties (KDML 105, RD 6 RD 15 since 1959,
  1977 1978 respectively) led to a loss of rice
  biodiversity over the last 25 years.
• But more recently released recommended varieties
  are not being adopted by farmers.
• A very centralized RLR seed supply system
  controlled by public agencies provides only seed
  of few recommended varieties. To grow other
  cultivars, farmers depend on seed exchanges among
  themselves.

3
The context (2)

• There is an inadequate supply of quality seeds
  leading to the recent emergence of several new
  seed supply agents (private companies traders,
  community seed centers, etc.) with government
  support.
• Three different types of RLR fields (upper,
  medium lower paddies) can be distinguished in
  the small rainfed catchments cropped by farmers.
• The diversity of farmers strategies in RLR
  production depends on ethnicity (Lao vs Khmer),
  farm size (2-15 ha), labour availability,
  location/accessibility, market integration, etc.

4
Objectives

• To understand farmers decision-making processes
  regarding
• The choice of RLR varieties,
• The allocation of these various varieties to
  different types of paddyfields along the
  toposequence,
• Their choice of seed sources / suppliers.
• To compare farmers decision-making processes
  across a range of different agricultural
  situations in Ubon Ratchathani Province according
  to
• Access to agricultural services,
• Ethnicity,
• Availability of specific / niche markets for
  paddy.

5
Conceptualisation of the game (1)

• A series of meetings was held among researchers
  from concerned institutions (OARD, UBU, RRI) to
  construct a shared conceptual model by using
  Unified Modelling Language (UML) diagrams.
• Knowledge gaps were filled by data from a
  specific farm survey (258 farmers in 25
  districts) interviews of key informants.
• Model verification, calibration validation
  implemented through
• The construction and use of a role-playing game
  (RPG),
• Individual interviews with players and group
  discussions after gaming sessions to analyze the
  linkages between the game and farmers real
  circumstances.

6
Conceptualisation of the game (2)

• Conceptual model for the choice allocation of
  RLR varieties

7
Conceptualisation of the game (3)

• Conceptual model for the choice of seed sources /
  suppliers

8
Conceptualisation of the game (4) Translation of
the conceptual models into a RPG to facilitate
sharing, discussion, and their improvement

• Design of the game

9
Main features of the RPG to be described

• Players other participants
• The gaming room
• Game supporting features
• Duration / Time management

10
The players other participants

• Around 12 growers with different farm size,
  coming from villages with different access to ag.
  services. Each player manages an amount of land
  similar to what he has in reality.
• The players were selected by the research team
  based on farm survey data to represent
  contrasting situations a balance between male
  female players. Players received free transport
  per diem (the official daily minimum wage).
• Two gaming sessions were held at different
  locations
• September 2003 northern Ubon, 12 Thai-Lao
  players from 2 villages close to the rice
  research center (RRC) and 2 more distant ones
• January 2004 southern Ubon, 13 Thai-Lao and
  Thai-Khmer players from 3 villages far from
  government seed production facilities.
• Information on social relations among selected
  players obtained during interviews before the
  gaming sessions.
• Researchers manage market desks paddy seed
  sales. Other members of the research team played
  the roles of moderators, assistants, and game
  observers.

11
The gaming room

• Two boards (6 players per board) representing
  different agricultural settings / farm sizes are
  played simultaneously
• A public bulletin showing the list of available
  varieties and sources of seeds
• Two market desks (registration of transactions)

12
The gaming room the 3D board

• Different farms on the board
• number and location of fields,
• 6 players/farms per board

13
Game support features post it stickers

• Three different types of stickers were used
• One color per variety
• One color per source of seed
• One color per planting method (Transplanting or
  Direct sown)
• Stickers affixed to each cropped field, every
  year/round of play.

14
Game support features record keeping

• One note pad per farmer and per year to record
• Selected varieties per field,
• Choice of crop establishment technique
• Amount source of seed purchased,
• Amount of paddy sold on the market,
• Amount of paddy kept for consumption production
  of own seed
• Note pad given to the market desk at beginning
  and at the end of each year/round of play
• ? data entered in an Excel file to estimate
  expenses incomes (close up of Excel file)

15
Duration / Time management
At market desk Sell paddy
4 steps per round of play
At the board Harvest rice Keep own seed,
Consumption / sale

• One round of play one crop year 30 to 60
  minutes.
• 6 crop years played in succession in one day in
  Sept 2003 4 years in Jan. 2004 to observe
  trends rythm of seed renewal.
• Gaming workshop over two days in Sept 03 D1
  gaming session discussion after two
  years/modifications of rules ½ D2 individual
  interviews. Only one day for both activities in
  Jan 2004.

16
Role of the moderators rule management

• 2 Thai moderators they introduce the
  participants, they spell out the objectives of
  the session/workshop, and explain the game rules
  proposed by the research team to the players in
  the local language.
• They assist the players at the board and at the
  market desks and they make decisions in case of
  unexpected behaviour
• After two rounds of play/ two crop years they
  encourage the players to express their opinions
  about the game, its rules to propose suitable
  modifications (modifications requested by the
  players were dealing with the price of seed,
  sales of glutinous rice, possibility to grow more
  than one variety per field, etc.)
• They assess the feasibility of the proposed
  changes and insert them in the game if practical.
• Following more rounds of play, they propose more
  changes to test the players responses to factors
  such an increase in family size, better
  availability of RLR seed, the introduction of a
  new variety.

17
Role of assistants observers

• Assistants help the players
• 3 at each board to affix stickers to fields,
  prepare the note pads bring them to the market
  desks.
• 2 at the market desks to register the players
  decisions and to settle their accounts.
• 4 observers monitor the individual behaviour of
  the players their interactions (2 of them
  record sequences by using digital cameras)
• Only few interactions were recorded during the 2
  gaming sessions. No exchange of seeds, no
  counseling or imitation occurred.

18
Indicators for the analysis of the game

• The analysis emphasizes the understanding of the
  dynamics observed during the gaming session
  less attention is paid to the explanation of
  quantitative results.
• Key indicators for the analysis of a gaming
  sessions
• Balance between glutinous and non glutinous RLR
  varieties depending on ethnicity, family
  size/needs, farm size farmer strategy
• Rythm of seed renewal for each variety, depending
  on access to suppliers, seed quality,
  farmerstrategy
• Location of the different varieties along the
  slope depending on the availability of
  upper/medium/lower paddies, family needs farmer
  strategy
• Changes in seed sources / suppliers during the
  game.

19
UML diagram after the game
20
Participatory analysis of the game

• Collective debriefing after 2 or 3 rounds of
  play/crop years at the end of the gaming
  sessions
• Comments from the players on the game structure,
  features and propositions of new features and/or
  rules to be introduced and tested.
• Comments on the evolution and management of the
  session at the end.
• Individual interviews the day after the gaming
  session (or just after it in January 2004)
  focusing on
• Similarities differences between the game (its
  structure, features rules) and farmers actual
  circumstances
• Similarities differences between farmers
  actions during the game and their actual
  practices
• Making farmers decisions during the game more
  explicit, particularly in the case of unexpected
  ones
• The usefulness of the game for the players and
  possibly other stakeholders.

21
Validation

• Following the introduction of changes requested
  by the players, they agreed that the game was
  representing their actual cropping seed systems
  at the farm level
• The players introduce many aspects from their
  actual farming context, strategies practices
  into the game (more than requested by the game
  rules and features)
• But many players ignored that there was such an
  extensive number of seed suppliers, information
  sharing among farmers is poor
• The game confirmed that individual decisions
  dominates in the management of cropping seed
  systems at farm level.
• It also confirmed the priority given to the
  production of glutinous rice to meet family needs
  on the Thai-Lao farms
• Major factors influencing the choice of seed
  suppliers are seed quality and accessibility
  (distance social network).

22
Perspectives

• Associate this RPG with another one representing
  the functionning of the RLR seed supply system at
  the provincial level (see the description of the
   Ubon rice seeds  RPG).
• Following their validations by stakeholders,
  integrate both RPG into a single multi-agent
  model to simulate possible scenarios of changes
  in the provincial RLR seed system identified by
  the concerned stakeholders.
• Test the use of machine learning techniques to
  analyse the results of the gaming sessions in a
  more objective and rapid way (in collaboration
  with the Social Simulation Laboratory at Kyoto
  university).