Previous Page Table of Contents Next Page

Participatory action research to improve irrigation operations: Examples from Indonesia and India

D.L. Vermilion and J.D. Brewer, Irrigation Specialists, International Irrigation Management Institute, Colombo, Sri Lanka


This paper reports on how two cases of participatory action research (PAR) involving farmers, agency staff and researchers facilitated a change in irrigation operations and produced significant improvements in performance. The first case is an innovation in dry season rotational irrigation in the Maneungteung Irrigation System in west Java, Indonesia. The second case involves the creation of a farmers' distributary canal committee and the introduction of new operational procedures for enhancing equity and management control in the Paliganj Distributary Canal of the Sone Command in Bihar, India. Involvement of farmers, agency staff and researchers in problem identification, proposing solutions, implementation and evaluation was essential to achieving results. Performance outcomes are presented.


Irrigation systems are complex 'socio-technical systems' (Trist, 1981) which involve interaction between a physical environment, technology, agriculture and often contending interests of multiple stakeholders. According to this view, if planned change does not involve both sound technical analysis and participation of key stakeholders in decision making, the outcome will probably be neither optimal nor capable of mobilizing the needed, support of farmers and agency staff.

Disenchantment with 'blue-print' or 'top-down' models of development was followed in the 1980s by a surge in 'participatory' strategies for rural development (ODI, 1989; Uphoff, 1986). 'Participatory action research' or 'PAR' (Whyte, 1991) is a method for merging both development intervention and research in an effort to involve key stakeholders in a systematic process of change and learning about how to make a socio-technical system more effective and efficient (Korten, 1980). It assumes that the participation of stakeholders is necessary in order to identify optimal solutions, obtain consensus and ensure support for an innovation. Obtaining stakeholder participation requires devising institutional means - such as committees - to bring the stakeholders together to jointly plan and evaluate innovations. In this work, the researchers serve to help organize the work, to provide suggestions for innovations and to monitor and feed back data on the progress of the innovations to the stakeholders.

This paper assesses two cases of participatory action research and compares: 1) how they were implemented; 2) responses of farmers and agency staff; and 3) performance outcomes of the intervention. Reasons are given as to why the experiments were successful despite adverse conditions.


The Maneungteung Irrigation System, which serves 7 611 ha, is the last irrigation diversion along the Cisanggarung River, located only a few kilometres before the river empties into the Java Sea on the northeast coast of west Java, Indonesia. Average annual rainfall is approximately 1 800 mm, concentrated mainly in the wet season between November and May, when virtually all of the design area is irrigated and cultivated. Discharge at the diversion weir drops from 6-12 m3/s for most of the year to 2-5 m3/s between July and October, with river discharge often dropping to 1 m3/s. This allows an amount of diverted water equal to only 20% of canal design capacities during the dry period. The upper area of the system has a slightly undulating terrain with no drainage problems; the lower area, near the Java Sea, is flat and poorly drained. About half of the irrigated area is planted in sugar cane at any time of the year. Other main crops are rice, red onions (shallots), chilies, green beans, mung beans, corn and groundnuts. Average farm sizes are about 0.33 ha.

The system is classified in Indonesia as a 'technical' system, meaning that water flows can be measured and adjusted from the diversion to each tertiary (i.e., field channel) offtake. The most common control structures are romign gates (a vertically adjustable broad-crested weir which theoretically permits simultaneous control and measurement of discharge) at tertiary offtakes, and sliding gates and stop logs along main and secondary canals. Water delivery to each tertiary block is planned according to 15-day estimates of irrigation requirements, which in turn are based on crop water requirements. Under periods of moderate shortage of supply relative to demand, this demand-oriented system is modified by a 'factor k' which reduces target discharges to tertiary blocks proportionate to the ratio of supply to demand.


Between 1987 and 1989 the International Irrigation Management Institute collaborated with the Indonesian Directorate of Irrigation I, the West Java Provincial Irrigation Service and farmer representatives in analysing and testing the potential to improve the performance of dry season rotational irrigation in the Maneungteung Irrigation System. The East Division of the Maneungteung System, which irrigates 4 871 ha, was selected for the PAR experiment. The PAR study team consisted mostly of staff from the national and provincial irrigation agencies, as well as staff from Gajah Mada University and IIMI. Numerous field operations staff and farmers were interviewed individually and in groups about their perceptions of dry season irrigation. Data was collected about the current and previous rotational arrangement, water discharges and cropping intensities around the system. Day and night inspections of all canals down to field channel outlets were normally monitored four times a week during the 1988 and 1989 rotational irrigation periods to compare planned with actual water distribution.

Defining the problem

Prior to 1989, rotational irrigation in the East Division of the Maneungteung system was based on dividing the area into seven blocks, with each receiving a water turn for 24 hours on a given day of the week. The Division included 70 tertiary blocks, of which 67 were scheduled to receive water on one day each week. During most of the year continuous irrigation was implemented according to crop water requirements. Each year, rotational irrigation was initiated at about the time supply levels dropped below 60% of system demand, which tended to occur in July or August.

Observations and interviews revealed three main problems with the existing rotational irrigation. First, rotational blocks and relative water demand were very unequal. For example, there were 403 ha in the Friday block and 1331 ha in the Monday block. In the 1988 dry season, 369 ha were planted in the Friday block and 1 107 ha were planted in the Monday block. Although each block had 24-hour turns, the estimated irrigation requirement ranged from 0.25 m3/s for the Friday block to 0.81 m3/s for the Monday block. Inequity was further exacerbated by the Monday block being at the head end and the Friday block being at the tail. Farmers emphasized this problem of equity.

Secondly, there were a large number of gates, often in disparate locations, which had to be monitored and adjusted to implement the rotation. In several locations there were no control structures at boundaries between rotational blocks. Agency staff emphasized the problem of inefficiency.

Thirdly, gate settings were often tampered with and were at odds with the rotational plan, as reported by farmers and observed by the PAR team. Thirty-three day and night time inspections during the 1988 rotational irrigation period documented that 310 of 444 (70%) gate settings observed were according to the plan; 134 settings were either receiving water when they should not or were not when they should. Farmers and agency staff both emphasized the problem of lack of control.

Devising solutions

Through discussions with farmers in the field and meetings with agency staff, the PAR team identified the three main weaknesses of the existing rotational arrangement to be: 1) inequitable and inefficient water distribution; 2) inefficient and cumbersome gate control requirements; and 3) lack of control in implementing the rotation. These communications revealed a common interest between farmers and agency staff in altering the configuration of rotation blocks, changing the timing of rotations and expanding the role of farmers in a new rotational arrangement. Meetings between agency staff and the PAR team produced a range of alternative rotational options and indications of the likely effects of each option on equity, efficiency and management control. Several discussions were held between the PAR team and the irrigation agency, local government, agriculture service officials and farmers about alternative rotational options and criteria for choosing among them. The PAR team identified five options and calculated the implications of each for equity of water delivery, efficiency of operations and control.

A meeting was held prior to the 1989 dry season rotation period which was attended by all involved irrigation agency staff, agriculture service and local government officials and farmer 'representatives' (most of whom were village agricultural officers) from each of the more than 20 villages in the East Division. In the meeting the PAR team described to the group the findings and perceptions about the existing rotational arrangement and then explained the five rotational options. After the agency subdistrict head proposed one of the options, farmer representatives requested some modifications to it, which were accepted, and an agreement was reached. The option was selected on the basis that it would provide the best mix of improvement in equity, efficiency and control through farmer participation. Participants recognized that an attempt to maximize any single performance dimension would seriously compromise another dimension, hence optimization was needed. The village farmer representatives emphasized the importance of improving equity of water delivery per unit of land area (regardless of crops planted). They also proposed: 1) that rotations should be changed at midday instead of midnight (as in the past); 2) that flags should be placed at the head of each secondary canal on the day of its rotation turn (to clarify the arrangement to farmers); and 3) that each village should supply a small group of volunteer farmers to patrol the canals during the night of its rotation turn.

Implementing the innovation

The pilot rotation was implemented from August to October 1989, when water supply dropped well below 60% of demand. The new configuration of gates was managed by agency staff. Village night patrols functioned throughout the rotation period. Farmers on night patrol sometimes reported difficulty in locating some agency staff at night when tampering was discovered. The level of tampering, however, was significantly less than in previous years. The PAR team conducted 42 day and night time inspections of all main and secondary canals and outlets during the rotation period.

FIGURE 1 - Improvement of equity in block areas. Between 1988 and 1989 rotations. East Maneungteung System, West Java

TABLE 1 - Management improvements between 1988 and 1989 rotations, East Maneungteung System, West Java

Irrigable area (ha)


Number of gates


Number of tertiary blocks


Management requirements



% change

Total management inputs




Total required gate operations




Gate supervisions (hrs/week)




a. Gate to be adjusted




b. Gate to be kept closed




Downstream flow must be stopped:




a. Using stop logs




b. Using sliding gates




Performance outcome

Under the new rotation, all tertiary blocks officially received water only once per week. Under the previous rotation the ratio between the area of the largest and smallest rotation blocks was 3.3; under the new rotation it was 1.49, with the smallest block being 564 ha and the largest being 842 ha (see Figure 1). As in previous years, cropping intensities in the upper end, Monday to Thursday blocks were nearly 100%, whereas the two tail-end blocks for Saturday and Sunday had cropping intensities of only 51% and 27%, respectively. The new rotation equalized to a large extent the amount of water delivered per unit area, but given the continuing inequality in cropping intensities between upper and lower areas, the new plan exacerbated inequality of water delivered relative to crop water requirements. Farmer representatives and agency staff accepted this as a step which had to be taken to eventually improve equity in cropping intensity over time.

FIGURE 2 - Sone Command, Bihar, India

Regarding management efficiency and control, the new rotation reduced the total number of gates which had to be closed or opened per week from 219 in 1988 to 166 in 1989, a reduction of 24% (see Table 1). Staff time required for gate supervision was reduced by 15%. From 42 day- and night-time inspections of gate settings throughout the East Division conducted during the experimental rotation in 1989, the PAR documented that 1 099 settings, out of a total of 1 260 observed, were in compliance with the rotational plan. This is an 87% compliance rate, compared to 70% in 1988.


The Paliganj Distributary is part of the Sone Command Irrigation System in south Bihar (Figure 2). The Sone Command diverts water from the river to irrigate a design command area of over 700 000 ha. The Paliganj Distributary diverts water directly from the Patna Canal in the Eastern Sone System. The head gate is 74.4 km from the head of the Patna Canal in the middle reach of the system. The distributary has 40 km of main canal network, including the Paliganj Distributary itself and the Chandos and Bharatpura sub-distributaries. The official gross command area is 14 867 ha with a culturable command area of 12 197 ha. The Paliganj command area lies wholly within the Gangetic Plain. About 1 000 mm of rain falls each year, mostly during the kharif season between June and October. Smaller amounts of rain fall during the rabi season from November to March. The hot season from March to June is generally without rain.

During kharif, the main crop is rice. During rabi, wheat is the main crop but various other crops, including pulses and oilseeds, are also planted. Rabi irrigation is limited to about 1 000 ha. Almost no crops are planted during the hot season, when the canals are closed for maintenance. The Paliganj Distributary serves 76 villages. The villages include about 16 000 households and more than 114 000 people. Almost all the villagers are Hindu. Most villages are made up of mixed castes; no caste predominates in the area, although specific castes dominate some villages.

Water distribution ostensibly follows the satta system, whereby farmers make requests, called satta, to the canal officers for irrigation water. In fact, however, requests are ignored and water is supplied for a fixed area - the 'assured irrigation area' - during kharif and, theoretically, also during rabi. Below each outlet one farmer acts as the sattadar, who takes requests from farmers and manages distribution below the outlet. The irrigation agency is responsible for delivering water to the outlet. Generally, there is one outlet per village. Watercourses are constructed and maintained by villagers. Assessment and collection of water rates is done by the revenue wing of the Irrigation Department with help from the sattadar.

However, due to problems of control and communication, in the Sone Command it is often not clear what is authorized versus unauthorized irrigation. Farmers in the head reaches have tended to take as much water as wanted by obstructing canals to raise water levels. Partly due to shortage of financing and reluctance to enter the area due to farmer discontent about the irrigation service, the presence of agency staff in the area and the recording of actual irrigations has been eratic. The agency tended to assume that all farmers within the command received water and all were assessed water charges, irrespective of whether water was actually delivered.


The Water and Land Management Institute (WALMI) was created in 1983 to train Bihar State irrigation engineers and others in improved irrigation and land management. In 1988, with support from the United States Agency for International Development (US AID), WALMI staff organized an Action Research Programme (ARP) at the Paliganj Distributary to improve the performance of the canal network so as to improve the productivity and profitability of irrigated agriculture in the area. In their first visits to the field in June 1988, the ARP team used vehicles identifiable as belonging to the Water Resources Department (WRD). Farmers reacted with hostility, jeering and threatening the team, as farmers assumed they were WRD staff. The ARP team confined initial activities over the next six months to making hydraulic measurements and observations of the irrigation system, meeting with farmers and collecting agricultural and socioeconomic data.

Defining the problem

Over time, familiarity with the researchers caused the hostility to dissipate. A meeting of leading farmers from villages in the lower reaches of the Paliganj Distributary was organized by the ARP team on 17 March 1989. At this meeting, farmers expressed anger at the WRD and the Sone Command Area Development Agency (SCADA).1

1 The Sone Command Area Development Agency works with villages on watercourse improvement, land consolidation, soil testing, and other farm and village-level activities.

Farmers and the PAR team found many water distribution problems, including, among others: unreliable supply at the head of the Paliganj Distributary, unavailability of water when required and excess water when not needed, poor condition of the canals and structures, including broken and missing regulators, open cuts in the canal banks to take water, failure of agency officers to visit the area; no channels for communication between farmers and officers; and lack of a system for inter-village cooperation.

Devising solutions

During 1988 and 1989 several meetings were held between the PAR team and farmer representatives from the different villages along the canal, which led to the formation of a Farmers' Committee which represented 20 villages. Farmers and the PAR team identified three problems for priority action:

· 'unattended' operation of the distributary canal,
· lack of farmer involvement in irrigation management, and
· lack of communication between users and system managers.

In a meeting in September 1989, the PAR team facilitated organization of the Farmers' Ad Hoc Canal Operation Committee which consisted of farmer representatives from 20 villages. Its initial objective was to improve water distribution along the distributary canal. Farmers felt that this was the level at which most of the problem with water distribution occurred and that until this was improved, there was little incentive for farmers to organize and invest in irrigation management at the field channel level. Formation of a distributary canal committee preceded formation of village irrigation committees (which were based on outlets). Shortly after the September meeting, at the request of the WALMI, the engineer-in-chief of the Water Resources Department issued a formal order for handing over operation of the system to the Committee. Group 'walk-throughs' along the distributary canal were made by Committee members to identify problems with obstructions and unwanted water tapping. Upper-end farmers reported being impressed for the first time by the severity of drought in the tail end.

Between 1989 and 1991 the PAR team met often with the Committee, advising it on particular issues, discussing technical problems and providing specially prepared technical information in Hindi. After the initial focus on operating the distributary canal, the Committee became involved in maintenance, organizational issues, and resource mobilization. WALMI staff also organized two-day training sessions on various aspects of irrigated agriculture for village irrigation committees. Members of the Committee were taken on tours to observe irrigation in other states in India. Between 1988 and 1991 the Committee began implementing new arrangements for patrolling and removal of canal obstructions and tapping, rotational irrigation, maintenance and raising a fee of 1 kg of grain per bigha (1/4 acre) of land irrigated.

In April 1991 a two-day 'farmers' camp' was held with 200 members of the 20 village irrigation committees, involving formal training, speeches and small group discussions about various agricultural and irrigation problems. In September 1993 another farmers' camp was held which was attended by over 500 farmers and many officers of the state Ministry of Water Resource Development. In this meeting the Minister of WRD agreed to turn over management authority of the distributary canal to what was then called the Paliganj Distributary Farmers' Committee.

Implementing innovations in canal operations

By early October 1989 farmers were removing canal obstructions and informing farmers of a new collective plan for improving operations. Water began reaching the tail end for the first time in ten years. For the 1990 kharif season, the Farmers' Committee adopted a simple operations plan developed with the help of the ARP team. The plan rotated water among five blocks - head, middle, tail, and Chandos and Bharatpura sub-distributaries - within the standard 10 day rotation for the whole distributary. WRD agreed to carry out the plan. During kharif seasons in 1991, 1992 and 1993, the Farmers' Committee, the ARP team, and the Water Resources Department implemented a more sophisticated 15-day rotation to further enhance equity by area. This was hampered somewhat, however, by the overall condition of the irrigation canals. The ARP team developed a computer program to simplify the necessary calculations. Delivery to the tail portions, although improved, is still not satisfactory to the farmers. During the winter rabi season, water demand is only 60% of that during kharif. Because canal capacities are sufficient to meet demand, water shortage is not a major problem, so farmers decided that no operational plan was needed during the rabi season.

TABLE 2 - Performance of the Paliganj Distributary during kharif, 1988-1992 (percentage of kharif water delivery at Paliganj head regulator)

Canal reach

% area






Paliganj Distributary - Head







Chandos Sub-distributary







Paliganj Distributary - Middle







Bharatpura Sub-distributary







Paliganj Distributary - Tail







Performance outcome

Between 1988 and 1993, the equity of kharif season water distribution among the five blocks of the Paliganj Distributary improved significantly. Tail areas now receive a significantly higher fraction of the water. The tail-end section of Paliganj Distributary Canal serves 30.7% of the irrigable area. Before the action research, this portion of the canal received only 10 to 12% of the water diverted into the canal. For three consecutive years after the Farmers' Committee became involved in managing the canal, 18% of the water reached the tail portion (Table 2). Water flows into the lower Bharatpura sub-distributary canal increased from about 4 to 12.8% of the water. This was accompanied by a relative decline in water diverted into head and middle-section blocks. Irrigated area increased significantly over the same period. Between 1990 and 1993, the area irrigated by the Paliganj Distributary Canal during the kharif season rose from 3 614 ha in 1990 to 4 350 ha in 1992 and 1993.

Farmers reported in the camps and other meetings that, in their view, the most important results of the PAR programme were: 1) improvement of relations between farmers and WRD officers; 2) some unity among farmers about solving the water problem; and 3) upstream farmers becoming involved in helping to solve the problems of tail-end farmers. Farmers recommended to the Ministry that: 1) the Paliganj Distributary Farmers' Committee be formally registered under the Societies Registration Act of 1860; 2) there should be a formal agreement for participatory management between the Committee and the Ministry; and 3) the Ministry should remove all rules and regulations which constrain formation of distributary canal committees throughout the Sone Command. These recommendations have yet to be formally accepted, although as of 1994 the final two recommendations were under active discussion.


Both cases of participatory action research had partial success in difficult environments. In Indonesia, the agency did not have a tradition of participatory planning and implementation. Farmers noted that agency staff often received payments for unofficial deliveries during dry season and that it tended to discourage formation of strong farmer associations. The Paliganj area in Bihar is designated as a 'socially disturbed' area, with frequent conflict between caste and landed and landless groups. And yet, despite the lack of either a clear policy or strong political support for the experiments, they were accepted by both the agencies and farmers. Lack of resistance to the interventions can be explained by the fact that both experiments did not seriously threaten, replace or make redundant the role of the agencies (as full management turnover might do), but augmented agency capacity through farmer participation. The visibility of the experiments probably also helped prevent resistance.

Organizing farmers to take part in the innovations was a key element in both cases. Farmer organizations that could select representatives to work with the agency officials and researchers were essential. Through discussions among farmers, agency personnel and researchers, farmers took an active part in evaluating the innovations, often making use of data supplied by the researches as well as their own experiences. Contrary to conventional views about organizing farmers from the bottom up (Uphoff, 1992), farmer and agency staff in both experiments recognized the need to organize and initiate action at the supra-local or distributary canal level, where potential to improve water distribution was judged to be the greatest.

Past efforts to organize water users associations have sometimes faltered with community organizers who had no background in irrigation and had little clout with the irrigation agency. However, PAR teams in both the cases reported above included both agency staff and researchers trained in irrigation and agricultural engineering. They were able to provide the necessary technical answers during negotiations and had credibility with both the farmers and the agency.


Korten, D. 1980. Community Organization and Rural Development: A Learning Process Approach. Public Administration Review. Vol. 40 (September-October), No. 5.

Overseas Development Institute [ODI], Agricultural Administration Network. 1989. 340 Abstracts on Farmer Participatory Research. London: ODI.

Trist, E. 1981. The Evolution of Socio-Technical Systems: A Conceptual Framework and Action Research program. Toronto: Ontario Ministry of Labour.

Uphoff, N. 1992. Learning from Gal Oya: Possibilities for Participatory Development and Post-Newtonian Social Science. Ithaca, New York: Cornell University Press.

Uphoff, N. 1986. Improving International Irrigation Management with Farmer Participation: Getting the Process Right. Studies in Water Policy and Management No. 11. Boulder and London: Westview Press.

Whyte, W.F. (ed.). 1991. Participatory Action Research. Newbury Park, California; London and New Delhi. Sage Publications.


Professor L.P. Srivastava, Water and Land Management Institute, Patna, was principal investigator in India. Mr. H. Murray-Rust, Lecturer at Peshawar University, Pakistan, and Mr. Soedarmanto of Gajah Mada University were key researchers for the Indonesia study.

Previous Page Top of Page Next Page