Review: (Dung et al., ????) Agent-based modeling and simulation of integrated rice-shrimp farming in Bac Lieu Province, Mekong Delta, Vietnam

Agent-based modeling and simulation of integrated rice-shrimp farming in Bac Lieu Province, Mekong Delta, Vietnam

Le Canh Dung1, Chu Thai Hoanh2, Christophe Le Page3, and Nantana Gajaseni4
1Mekong Delta Development Research Institute, Can Tho University, Vietnam (lcdung@ctu.edu.vn) Currently a PhD student at Faculty of Science, Chulalongkorn University, Thailand
2International Water Management Institute (IWMI), Regional Office for Southeast Asia, Penang, Malaysia
3Centre de coopération internationale en recherche agronomique pour le development (CIRAD), CUCIRAD Unit, Faculty of Science, Chulalongkorn University, Thailand
4Faculty of Science, Chulalongkorn University, Thailand

Main Objective
This study aims to integrate and share stakeholder knowledge about rice-shrimp farming in Bac Lieu Province, Vietnam. The culmination of this knowledge will accelerate collective improvements in the design and management of rice-shrimp production systems (by comparing product mixture?). The research objectives are two-fold: “(1) to understand better the land-use decisions based on interactions among water quality, risk perception, and market price factors; (2) to provide a research process for analyzing decision-making in land-use that can be used for modeling and policy analysis.”

Key Methods
The key method is companion modeling (ComMod). It has two parts. The first is a Role-Playing Game (RPG) between farmers from two villages in Bac Lieu who recreate their famring decisions to reveal diverse water and land-use strategies. The second part is a generic Agent-Based Model (ABM) that used a subset of individual farming decisions to parameterize integrated rice-shrimp farming based on water salinity. The ABM was used using Common-Pool Resource and Multi-Agent System (CORMAS) platform (Le Page et al., 2000). ABM was used to analyze interaction among biophysical and socioeconomic factors that affect land-use decisions. This model accounted for (1) production cost, (2) price of product return and (3) risk of shrimp disease. Currently only one agent is analyzed. The model will become more complex as sluice operator and marketing intermediary agents risk perceptions and economic conditions are added. The model shows spatial setting. More complex model can be created to select production systems at two extremes of canal with different salinity conditions.

The model is split in to three modules: (1) Water (Space- Canal, Village, Farm, Plot), (2) Production (Crop- Rice, Shrimp, Crab, Fish), (3) Social (Household- 3 key operation methods of the farm are: (a) initiate household income status, (b) produce crops, and to (c) harvest products). It is not clear how these modules are embedded in the ABM.

Important Results and Conclusions
This iteration of the model was able to mimic the RPG production change due to water salinity. This will help gain credibility of the model in order to gather more data.

Assumptions
Use of this model assumes that ALL key stakeholders of the common resource (the canal in this instace) are involved in RPG. It also assumes that all stakeholders WANT to share experiences in land-use decisions in order to strengthen the adaptive management capacity of the local community.

Opinion
I am unsure of where this paper was published or how many times it has been cited. Given the terseness of its 5 pages the references list is short and appears to support the model that was built. The first research objective was partially met since land-decision used based on water salinity. It is unclear whether additional scenarios based on sluice operator decisions, risk and economic conditions were actually simulated and assessed. Therefore I am left unaware of scenarios related to risk perception and market price factors. The beginnings of the second research objective was met since the expansion of this model could provide a research process of analyzing decision-making in land-use that can be used for modeling policy analysis.

Impact on Food Distribution Environmental and Nutritional Impacts in Hawaii
This type of modeling could be used to test farm production reaction to various environmental factors. The methodology of the RPG would have to be representative of generalized groupings of operations in Hawaii since it is not practical to interview all producers.