Cambodian polder plan boosts paddy production

The upgrade of the Prey Nup polders in Cambodia meant engaging farmers in a dialogue to manage water resources. Jean-Marie Brun of the French research and technological exchange organisation GRET describes how a new structure has nearly doubled rice production.

Located on the Cambodian coast of the Gulf of Thailand, the Prey Nup polders cover a surface area of 10,500 hectares of paddy fields protected from the intrusion of seawater by nearly 90km of earth dykes. The polder system has been undergoing upgrading since 1998 in a government project, financed by the French Development Agency.

In addition to the “seafront” dykes, there is a network of canals and 36 water gates as well as partition dykes that allow differentiated water management in each of six subdivisions. Each subdivision covers from 600 to 2300 hectares, with almost no internal partitions, and topographical differences of 30-40cm between the highest and lowest plots.

Water is only partially mastered: the system does not have operational impounding reservoirs upstream and only allows excess water to be drained in this wet climate which has an annual average rainfall of over 3200mm. One rice crop is grown annually, during the rainy season, making water management in the polders a true challenge.

The lack of internal partitions within each of the six polders makes it impossible to individualise plot-by-plot management of the water depth. The farmers must therefore work together to define a consensual water management plan and then each farmer must match the crop cycle to this plan.


A farmers’ organisation was created and placed in charge of system management: the Prey Nup Polder Users’ Community (PUC). PUC is made up of 15,000 members who elect one representative per village, a total of 43 village representatives and six polder presidents.

In 2001, at the first Polder Assembly, village representatives and polder presidents met to determine water management master plans. Since then, assemblies have became an annual event and the plan is defined as follows:

  1. Definition of homogenous zones: the polders are mapped using topographical survey data and the representatives’ knowledge of the land.
  2. Description of crop calendars for each zone.
  3. Identification of water levels compatible with cultivation activities, for each zone and time period.
  4. Overall coherency: representation of zones and desired water levels in each transect, by approximately 15-day periods. This stage is crucial, because it is at this time that compromises are made to define a water level that is acceptable to all.

Hands on learning

The polder presidents are in charge of implementing the water management plans. They each have one or two water bailiffs in charge of measuring the depth of water in various places and opening or closing water gates on the president’s orders.

If inconsistencies become apparent when implementing the plans, or if conditions require exceptional measures, the polder presidents can reconvene the Polder Assemblies to revise the management plan. This is a tricky task as the Polder Presidents do not have any references.

The year 2001 was a baptism by fire for the PUC leaders. Climatic extremes meant that production was hit by excessive flooding followed by an end-of-cycle drought.

Initially, the upgrading work had increased water management difficulties in the polders and the first season’s yield was about 1.6 tons of paddy per hectare–nearly identical to yields prior to upgrading. The PUC was faced with a crisis of confidence among its members and was blamed by the local authorities.

However, with the support of the project team, the PUC’s elected officials analysed the reasons for this failure and learned from it. They observed the behaviour of the water levels and learned how to better anticipate changes in water depth.

They also identified the flaws in the technical system itself: the topographical surveys contained errors, which were corrected, and it was found that the partition dykes between polders needed to be raised. Innovations improved monitoring and decision-making.

Along with monitoring the water depth of the canals, the polder presidents included in their working indicators water levels measured in the plots. Water management plans were also communicated better to the users, including a critical analysis session on organised in the Village Assemblies, with all members.

These efforts bore fruit, now the PUC has better mastery of its subject; the harvest was better in 2002 and reached record heights in 2003 (3 tons per hectare on average) as abandoned lands were brought back into use.


The organisational structure in place today seems to have proven its ability to manage water. For all that, the future of the Prey Nup polders is not yet guaranteed and a few important steps still need to be taken.

Technical mastery of water management appears to have been achieved and is now being underpinned by the development of a small salaried staff employed by the PUC. Knowledge and skills are mastered not only by the representatives but also by the wider community. This offsets the risk of an erosion of skills when elected officers are replaced. Financial means are also needed to preserve these skills, to ensure competitive remuneration of staff and elected officers.

Infrastructure maintenance is the primary expense of the PUC. Users’ participation is currently relatively satisfactory and the fee payment rate is approximately 85% with over US$55,000 collected in 2004. However, in the medium term, the budget will need to be increased significantly to ensure maintenance.

Increasing rates and collecting unpaid fees are the two major challenges facing the PUC. The solidity of the support provided by the public authorities will, in this area, be a decisive factor.

Action inspires action. Stay ahead of the curve with sustainability and energy newsletters from edie