Canal maintenance: a key to restructuring irrigation management A case o f farmer participation and turnover from Mexico
P I E T S I J B R A N D I J 1 & P I E T E R V A N D E R Z A A G 2'*
I Netherlands Development Organisation, Apartado 68, 1031 Bissau, Guinde Bissau; 2Department of Agricultural Economics and Extension, University of Zimbabwe, P.O. Box MP 167, Mount Pleasant, Harare, Zimbabwe (* request for offprints: Sprengertaan 11, 6703 GA Wateringen, The Netherlands)
Abstract. Farmers becoming involved in canal mainentance is a recent trend in many government- managed irrigation systems. Before being able to assess the pitfalls and perspectives of this trend, it is necessary to examine in detail the issue of canal maintenance itself, an issue which has received relatively little attention. The paper focuses on canal maintenance in an irrigation system in Western Mexico, and finds that canal maintenance differs in several aspects from water distribution. A first difference is that water distribution often results in competition among water users along the same canal, whereas the need for canal maintenance may bring these people together in cooperation, which may help to forestall possible conflicts over water supply. A second important difference is that water distribution is directly productive, whereas canal maintenance involves the reproduction of the canal system, the (often considerable) costs involved being investments. In the case study reviewed, canal maintenance gave rise to certain relationships between different groups of people, and between them and the canal infrastructure, relationships which may not have been uncovered if only water distribution had been studied. It is argued that interventions aimed at handing over management responsibilities to water users would be more effective if such locally specific expressions of cooperation and initiative of the actors involved are taken as a starting point.
Introduct ion
C a n a l m a i n t e n a n c e is a cos t ly a f f a i r . U n d e r the p r e s su re o f s t r uc tu r a l ad ju s t -
m e n t po l ic ies , g o v e r n m e n t s see a poss ib i l i t y t o s ign i f i can t ly r e d u c e o n the i r irr i-
g a t i o n bi l l by h a n d i n g o v e r t h e r e spons ib i l i t y o f c ana l m a i n t e n a n c e to t he w a t e r
users . Th i s p a p e r p re sen t s a case f r o m M e x i c o , w h e r e a h u g e p r o j e c t has s t a r t ed
w h i c h env i sages to ' t u r n o v e r ' m o s t o f t he g o v e r n m e n t m a n a g e d i r r i g a t i o n
sys tems . 1
M a n y o b s e r v e r s h a v e i n t e r p r e t e d t h e t u r n o v e r po l i cy as g o v e r n m e n t ' s a t -
t e m p t to ' sh i f t t he b u r d e n o f m a i n t e n a n c e ' to f a r m e r s , a n d o f t e n o p p o s e it.
H o w e v e r , this j u d g e m e n t seems p r e m a t u r e . As yet no e l a b o r a t e e m p i r i c a l
190
studies on the effects of turnover are available. More disturbing is that very lit- tle has been published about the issue of canal maintenance, and its role in irri- gation management. In our view, possible pitfalls of farmers starting to manage canal cleaning, but also possible perspectives, will only be found if the issue of maintenance itself is examined in detail.
El Operado and its silted canals
This paper is centred around canal maintenance as is practised in one particular irrigation system, namely the Autlän-E1 Grullo irrigation system in Western Mexico, known as El Operado (van der Zaag 1992). 2 It was constructed in the 1950s and serves 8,700 hectares, of which 6,000 ha are currently planted with sugar cane. The canals were completely reconstructed, most of them in con- crete, during 1979-1984. E10perado ' s infrastructure being in quite a good shape, water distribution efficiency ranges around 60%. 3 Water distribution and canal maintenance are performed by personnel of the 'District', the local office of the Ministry of Agriculture and Hydraulic Resources (SARH). The engineer responsible for maintenance is the head of the maintenance depart- ment, who oversees the work of over 60 personnel and has at its disposition a number of machines: two draglines, two hydraulic excavators, two bulldoz- ers, one grader, and some other vehicles and trucks.
Maintenance of E10perado involves around 80% of the entire budget for operation and maintenance, and the works include taking care of the storage dam, the irrigation canals, drains and road infrastructure. Canal maintenance is perceived as most critical, since a lack of it directly affects water supply. Dur- ing the months of SeptembeL October and November canal maintenance is a hot issue: during these months the Ayuquila river may still carry a good deal of aguas broncas ('wild waters'). This silt laden water is brown in colour. w a t e t users of E10perado, especially those with sandy soils, or growing maize or sugar cane, use the aguas broncas in the first dry months to give their crops supplementary irrigation. The silt from this water not only enters the main sys- tem, but also the minor canals and fields. While the silt on the fields may be beneficial, in the canals it causes problems. The silt decreases the discharge ca- pacity of the canals. Aquatic plant s attach their roots to the silt and these aquatic plants further decrease the veiocity of watet flow and more silt is
deposited. Of the 232 km of canals, including main canals, laterals and sublaterals (field
canals excluded), approximately 100 km needed to be cleaned during autumn 1988 to prevent problems with watet distribution. Of these 100 km, 58 km were not cleaned at all, and some 700 water users were confronted with delays in receiving their irrigation turns by u p t o a month or more (Sijbrandij 1989). 4
191
This negatively affected agricultural production on 2,900 ha, or 33 % of the to- tal irrigated area. It is estimated that production losses amounted to at least US$ 170,000. 5
The maintenance department cleaned 19 km of canals during the period September-December, mainly by means of a hydraulic excavator. The Depart- ment counted with 43.8 million pesos (approx. US$19,000) for canal cleaning, and at least 50% of it was spent during September-December 1988. 6 The unit price for canal cleaning thus amounts to US$ 0.50 per metre.
On their own initiative, water users cleaned another 23 km of mainly secon- dary and tertiary canals. When expressed in canal length, the farmers did more maintenance than the department (55% as against 45%). The actual amount of silt removed by the farmers, however, was smaller since the canals cleaned by them did not include the main canals. The farmers cleaned the canals by hand, and typically, each water user along a canal would clean a number of canal segments (3 m long). Normally the work could be done by the farmer, sometimes assisted by another worker, within one day. The cost of canal clean- ing by farmers ranged between US$ 0.20 and 0.40 per metre.
The actors: the maintenance department
The facts presented above indicate that the maintenance department fails to comply with its task. It faces four major constraints. The first is a severely res- tricted financial budget. The allocation of money is beyond the control of the District, so that the maintenance department has to try to comply with its task within the financial framework defined by higher authorities. 7 In 1987 only 53°70 of the required funds were allocated to the department. In 1988 the situa- tion was even worse, with less than one third of the funds needed. Secondly, the maintenance department seems unable to increase its execution capacity during the critical period between September and November. Among other things, this is related to the frequent break-down of the machines and the tight bureaucratic rules which the department has to fotlow with respect to the pur- chase of spare parts. A third major constraint is inherent to the organisational set-up within which the department functions: responsibilities over operation and maintenance are separated into two different departments. Lack of perfor- mance by the maintenance department has little repercussion on the depart- ment itself, but it does affect the operation department. A fourth constraint is the lack of accountability of the maintenance department to its clients, i.e. the water users. Farmers cannot let their opinions be known to the maintenance department other than through complaining on an ad hoc basis. No formal contact is established to ensure that the maintenance department is confronted by its own good or bad performance. The water guards, District personnel in
192
charge of water distribution, play an important role in voicing the problems farmers face with maintenance (van der Zaag 1992).
The maintenance department sees the silted canal as a problem which is difficult to solve within the existing institutional environment (the officially al- located budget is not enough, spare-parts of broken down machines are not readily available; workers unwilling to perform the filthy job by hand, etc.). But it appears that some officers within the department are not really willing to solve the problem. The cunning way in which the machinery and the person- nel of the department are used for private projects contrasts with the apparent- ly insurmountable institutional obstacles when regular works have to be per- formed.
The actors: the farmers
For the farmers silted canals imply water shortage, wilting maize, possible financial losses, and arguments among fellow users. Along many silted canals farmers fail to devise an effective solution. They fight over the scarce water and complain to their water guard but are unable to overcome the petty irritations that exist among them. It seems that before a joint initiative can be taken, the attitudes of people towards each other have to change.
In other cases farmers find a solntion to the maintenance problem: some opt for cleaning; others mobilise fellow farmers, go to the District's headquarters and demand maintenance service; yet others directly negotiate a solution with the District, whereby farmers agree to pay part of the costs involved (e.g. costs of fuel, salaries). The different solutions have in common the mobilising of contacts and networks. Such networks enable action to be taken.
In E10perado many farmers became actively involved in canal maintenance, in one way or another. In the following sections we will show that the emerging forms of farmers' cooperation were not short-lived but have a lasting charac- ter. The question arises whether these initiatives had any impact on the formal way in which the E10perado irrigation system is being managed. Towards the end of this paper it will be clear that the informal way in which farmers became involved in canal maintenance indeed affected E10perado ' s management. At the end of 1988 a water users association was formed who in 1989 took charge over maintenance. The case of E10perado, then, enables an elaborate discus- sion on the role of canal maintenance in irrigation management and turnover. This will be done through examining in detail canal maintenance along one canal, 'lateral 2', serving some 450 ha of irrigated fields (Fig. 1). We will first focus on one sublateral canal of lateral 2, the sublateral known as the Limonci- to canal.
4--
193
llano
radial gare
I ~ main eanal
intake
private = ~ l l a t e r a l 2
land°wnersl II ' limoncito canal
- - iiiiii ~ canal ~ ejido Mezquitan
ejido E1 Grullo
ejido Chacaltepec
private landowners
Fig. 1. Lateral 2 and its sublaterals.
The work party: a case of farmers taking the initiative
In September 1987, 10 plots with 50 ha of maize along the Limoncito canat were in trouble. The farmers saw to their awe how the maize had started flower- ing but that the leaves were rolled, indicating water stress. The farmers asked Miguel, their water guard, for an auxiliary irrigation turn with aguas broncas .
They received it, but due to silt on the canal bed, the minor canal could hardly carry a flow sufficient to satisfy one irrigation turn, whereas all 10 maize grow- ers were equally in need of water. Normally, the canal would have a discharge capacity of 180 1/s or three irrigation turns, but now no more than 40 to 50 entered.
Irrigation water enters the intake through an orifice, but will only do so if the water level in the on-going feeder canal is sufficiently backed up. The first 500 m of the canal has a counter gradient; the canal bed slopes upwards instead of down. So, nearly every year the canal becomes badly silted up. In former
194
years different individual farmers had at tempted to clean stretches of the canal. On September 17, 1987, a group of 10 farmers cleaned this canal.
On that day, two farmers were struggling to finish their irrigation turns with very little water. The water only advanced slowly in the furrows, the sandy soils sucking immediately all the water. Both farmers certainly wanted to complete their turns quickly in order to avoid trouble with their compaheros, who were looking on with envy and some with annoyance. But foremost they wanted to save their harvest.
At 8:30 am, Miguel, the water guard, arrived at the inlet of the Limoncito canal, which is part of his daily routine. Chico, who cultivates part of his father 's 10 ha land with maize, sat near the control structure. Chico obviously asked Miguel for irrigation water, as he had done 5 days earlier. Miguel told him the canal could not take more and that he had to wait another four or five days. But Chico was desperate for water and wanted to force his fellow water users to resolve the problem once and for all, and clean the canal. To do this he had first to enrol Miguel in his project. He wanted Miguel to share the responsibility. Miguel had his own view on the silt problem, as he knew before- hand that the maintenance department of the District would be unable or un- willing to clean this canal in the near future. The canal being silted made Miguel's work more difficult, so he also had an interest in the cleaning of the canal.
Chico then took action. He provoked the water guard by lifting the gate of the control structure somewhat. Miguel considered the situation, and decided to support Chico. Chico drew up the gate completely, and as a consequence, discharge into the Limoncito canal dropped to zero. Chico then decided to take the consequence of what had happened. He immediately plunged into the canal. Found struggling with mud would give him some advantage: " H o w could anyone say I am ar rogant?"
I jumped into the canal and started to remove the silt, and soon came one other farmer wanting to irrigate, and joined me. Then, one of the irrigating farmers came, angry with me (bravo conmigo). I told him 'there is no water here, not for me, not for you.' After this, they all helped to clean the canal. We were ten. At noon we had cleaned the worst parts of the canal, and then we started irrigating.
The work party was not a self-evident affair. It hung in the balance and de- pended upon the active decision-making, and the weighing of risks and benefits, of at least three actors. Gradually we see as a trigger effect, how one move provokes another finally resulting in joint action. Only when different actors (with varying motives) are enroled and cooperate, does the work party succeed. I f the felt needs of the actors had not been so great the whole project would probably have died an early death, since attitudes of people towards each other seem to inhibit such joint efforts.
195
Surprisingly, the water users of the Limoncito canal hardly made any refer- ence to the maintenance department of the District, when they were inter- viewed about their initiative. It appeared entirely an affair between farmers and fellowmen.
An emerging canal organisation
Lateral 2, the canal feeding the Limoncito canal, has one pressing problem: its intake is too small. The water guard had informed the operation department various times of the problem over a period of three years. It was up to the main- tenance department to enlarge the intake, since the department is also responsi- ble for improvement of the infrastructure. Nothing happened. Water shortage resulting f rom the small intake created conflicts among water users, and finally Miguel (the water guard) succeeded in convincing some water users of the exact cause of the problem. From that moment (during August-October 1988) things began to move. Miguel summarises the story:
In lateral 2 they had this problem of the small intake. I also had problems there. There was never enough water. What did the water users do? They assembled. They went to the District: Sirs, are you going to clean our canal? No. And repair the road? No. And the intake? No! So they went to the sugar cane organisation who lent them money. They constructed a new intake, cleaned the canal, repaired the roads.
It was quite a complex project which involved different departments of the Dis- trict (who eventually did not cooperate), the Grande Irrigación office (who provided the technical drawings), the CNC sugar cane growers' association (who provided through the sugar factory a loan of 19 million pesos or US$ 8,000), a construction firm, the watet guard and, of course, the water users. Note that the water users did not form a homogenous group, but included farmers f rom three different ejidos with less than 10 ha each and a few rich en- trepreneurial farmers with over 20 ha. 8 It was all organised by Don Enrique, one of the leading persons along this canal.
At the beginning of July the waters users of lateral 2 held a meeting. The venue was in the field. Some 30 water users were present. Two engineers f rom Grande Irrigación (the local SARH office in charge of the rehabilitation project of E1 Operado) showed up, and it was decided to go to the intake of lateral 2. The engineers and Dort Enrique discussed whether a bigger pipe should replace the current one, or that a rectangular concrete structure should be cast in place. The engineers promised to help the watet users with a technical drawing. Don Enrique was angry with the head of the maintenance department who failed to attend. He told one of the authors that he could do the job without hirn. He looked determined.
196
At the beginning of September a construction firm started the work on the intake. Two weeks later it had been completed. The pipe had been replaced by quite a big rectangular concrete strncture. The new construction looked profes- sional, not the botched job the maintenance department perhaps expected from these farmers. On September 26, Miguel, the water guard, tested the new intake: more water entered the lateral without having to back up the watet in the main canal. That same day water users from the Llano canal also cleaned their canal, organised in two working groups (one for each ejido).
Miguel was pleased. While in other parts in E 1 0 p e r a d o problems were mounting, with this new intake and the cleaned canal, water distribution here would be easy. And if this was not enough, two weeks later yet another problem was solved: on the initiative of Don Alonso, farmer leader in E1 Limoncito, water users demolished the Limoncito intake, and enlarged it. Now more water entered the canal with the wrong gradient.
The snccess of infrastructural improvement in lateral 2 by its water users is impressive. The maintenance department was set aside, and farmers moved themselves from the back to the front stage. They rightly say 'we arrange out affairs amongst ourselves, without those engineers in their offices', and 'we simply say that it is now our canal'. There is some continuity between the story of the Limoncito canal and that of the entire lateral 2. Initiative after initiative is taken, like dominoes tumbling. The former experience of the Limoncito farmers of j oint action (cleaning their canal in 1987), together with the impor- tant experiences of the rehabilitation of the intake of lateral 2, made it fairly easy for the Limoncito farmers to decide to rehabilitate their intake of the Limoncito canal as weil. To the farmers of the Llano canal, cleaning their canal now seemed a relatively small thing to do, and they did it without much fuss and in a clearly defined way, each ejido doing their share, and each water user cleaning 5 canal segments per hectare. All the water users profited from these initiatives, especially since the water guard was willing to give them irrigation turns more frequently than officially allowed. 9
The rehabilitation of the lateral 2 shows that in that area watet users seem to have a fair degree of self management. The setting in which the farmers meet (the field) ensures that it is their project, defined by those concepts and words they prefer and not of the District engineers. The majority of farmers, however, fail to press for proper decision-making procedures and to urge for an adequate account of expenditures. Don Enrique monopolised information, procedures were ill-defined, and the project itself was not well delineated (a road repair was later included). Too much was left to the discretion of the one who presented himself as leader. 1° It is not necessarily self-evident that the water users of lateral 2 will support the next initiative of a joint project. However, if they choose to do so, it may well be possible that among the farm- ers some will demand better formal arrangements. Smaller works, such as the cleaning of canals, already seem to be bettet organised.
197
The various initiatives taken by farmers in lateral 2 have left their mark on the landscape, and have produced new points of reference, around which new initiatives may grow. In a sense, then, these initiatives have a lasting character.
Canal maintenance: an empty arena?
Problems encountered with canal maintenance point out that the relationship between the District and the water users is strained. The gap that exists between the farmers and District personnel (a gap which with respect to watet distribu- tion is effectively bridged by the water guard) has gradually widened with respect to maintenance. It is only because farmers take initiatives themselves that the long-term viability of E10perado is not at stake. Despite the fact that farmers' formal role is nil, canal maintenance appears to involve an arena in which a variety of people is negotiating and taking action. This conclusion may have, and in our view should have, implications when government hands over the formal responsibility over maintenance to water users. This is what hap- pened in E10perado in 1989, which will be discussed in the following section.
Let us first summarise the most important issues which emerged when farm- ers became involved in canal maintenance in E10perado , and review possible implications for formally established farmer organisations in charge of main- tenance. 11 Four major issues can be singled out: social issues at canal level; or- ganisational issues; the role of government institutions and their personnel; and control over resources.
Social issues at canal level
Farmers taking initiative with respect to the silted canals drew upon social net- works that exist along the canals. However, group action is not a self-evident affair, since people tend to distrust each other. Furthermore, along many canals more formal boundaries exist: that between ejidos and between ejidatarios and private landowners. The entrepreneurial farmers may be reluc- tant to join hands with the peasant farmers. As we have seen, these inhibiting factors may be overcome when problems are really serious.
Organisational issues
The farmers met and resolved problems nearly always in a field setting, that is in their own domain, and not in a formal ejido setting, nor in a setting of the District. For farmers to meet in a setting which is entirely 'theirs' helps de-
198
fine projects in their concepts and words, and thus remain 'theirs'. It may fur- thermore ensure that such projects are defined in practical terms, and are not drawn into the private projects and politicking practices of both ejido leaders and District engineers.
Former experiences with tackling shared problems through joint actions helped to find solutions for current problems. The presence of farmer leaders who are accepted as such by their fellowmen, and who date to 'stick their neck out ' may be decisive. These leaders have to act cautiously, otherwise they face the risk of being criticised. It was therefore that Chico humbly jumped into the canal. This also raises the issue of leaders being accountable to the wider group. This was what Don Enrique refused to do, which created quite some animosity among many farmers of lateral 2, not in the least because so much money was involved in the project.
The role o f government institutions
Obviously, in case a farmer leader is acquainted personally to District en- gineers, solutions can be found more easily. But access to the highest District engineers can also be forced if a large number of water users together go to the District and ostensibly sit and wait in front of the chief engineer's office. Fur- thermore, the role of the water guard emerged as important: explaining to farmers the root of the problem at the maintenance department, informing them about actions undertaken by farmers along other canals, and offering al- ternative solutions.
Control over resources: investment and ownership
The joint undertakings of farmers involved many resources: investing labour, money, social capital, and skills in cleaning canals or rehabilitating an intake structure. As a consequence the farmers' definition of ownership over that in- frastructure, which has always been rather vague, changes. Viewed from the perspective of the water users, it shifts from being 'the government 's ' to 'ours ' (Coward 1986).
In out view the four issues elaborated here were not only relevant when farm- ers 'spontaneously' got involved in canal maintenance, but will also be so when farmers will get formal responsibility over maintenance.
Maintenance: a government's gift?
In the beginning of 1988 some farmers urged the chief engineer of the District to hand over the maintenance task to the users. They knew that it would be
199
necessary to create a water users organisation, namely an asociación de usuarios, an organisational entity recognised by the Federal Water Law. This association would then receive the full responsibility for maintenance. It is im- portant to note here, that simultaneously to this initiative of farmers, there were some higher engineers within the SARH Ministry who also realised that by handing over certain activities to farmers, the financial crisis which the Mex- ican government faced during these years could be alleviated. 12 In addition, during the electoral campaign (first half of 1988) of the 'official ' candidate and future Mexican president, many references were made to such an incipient policy.
At the end of 1988, indeed the asociacißn de usuarios (water users associa- tion) was formed. All ejidos and both private landowners' organisations elect- ed delegates. These delegates constituted the association's general assembly, which elected the executive board. At the beginning of 1989, the SARH Minis- try and the watet users association signed an agreement whereby the SARH would transfer all machinery to the association. In May 1989, the association started to clean the canals. Within a period of one year four practices changed, which previously blocked a proper maintenance system.
The first practice refers to the way in which money was collected for main- tenance. Formerly it was the government's tax revenue office which collected the water fees f rom the farmers; and eventually the District was allocated a budget for operation and maintenance. With the watet users association in place, it could now collect the fees directly from the watet users, and immedi- ately use it according to the plan, or bank it and receive interest from it.
A second practice which changed involved the manner in which the level of the water fee was established. From now on, the fee corresponded with the re- quired works planned for. So, during May 1989 the maintenance needs were inventoried in detail. The board gave the association's delegates the responsi- bility to define maintenance needs along the canals of their constituencies (the 17 ejidos and two private landowner organisations). Then, the board, aided by the District engineers, calculated the detailed costs of the required work. New watet fees were fixed on the basis of the required budget. The required work, the budget and the proposed new water fees were presented to the assembly of delegates. The delegates were informed how the money would be spent, and although they thought the new fees were high, they wanted all canals to be cleaned, and accepted the fees (for example, the fee for sugar cane was fixed on 70,000 pesos/ha, approx. US$ 36; an increase of 240%). This meant a fun- damental break with the former practice of establishing the annual watet fee. Now, farmer delegates established the watet fees on the basis of detailed cost estimates. If they wanted to lower the fees, they had to specify which canals or drains to leave outside the cleaning programme. Furthermore, the delegates had access to information with which the association's expenditure could be checked at the end of the season.
200
The water users who wanted maintenance work to be carried out on a partic- ular canal had to report this to his/her delegate, who then submitted it to the association's board, by filling in a requisition form. Any watet user going directly to a District engineer (as was the normal practice) was told to refer to their delegate. This responsibility boosted the delegates' morale and put them into a position whereby both the association's board and the watet users started to take them into account. The delegates not only took note of maintenance needs, but also monitored whether the work was done properly. By introducing a procedure for making requests and by installing a control mechanism, the in- formal way of handling maintenance work was abandoned, and a third prac- tice was altered.
At the end of the year many farmers as well as the District employed watet guards concluded that canal maintenance was performed satisfactorily. The as- sociation had cleaned more canals in less time, and with only half the budget (in real terms) than the maintenance department of the District had done the previous year. In January 1990 the association presented its balance sheet for the period between May to December 1989. It had collected over 300 million pesos (more than US$100,000). It had spent 200 million (125 million on canal maintenance and 75 million on office equipment), and had built up an acquisi- tion and machine maintenance fund of over 100 million (approx. US$ 36,000). Part of it was used to acquire two pick-up trucks and two second-hand tipper lorries. It appeared that through cleaning part of the canals by hand, a saving of 20 million pesos was achieved. The balance sheet, furthermore, revealed that the hiring out of machines beyond E10perado for private persons and institu- tions had generated some 50 million pesos of extra income (US$18,000). Here a fourth practice had changed: machines were no longer used solely as personal assets.
Linking what had been separated
Despite the positive change of four major practices, there were still some rea- sons to remain cautious. The most important reason being the weak relation- ships between the different levels within the new organisation. Although most were satisfied with the canal cleaning during 1989, farmers still feared that wi- thin a year or two the board, the delegates and the District would return to their former practices. The farmers' scepticism was strengthened by the way they were informed. The delegates hardly involved their water users in discussions concerned with basic decisions. Instead they informed the water users after- wards, if they informed them at all. The delegates themselves were duly in- formed by the board but hardly any discussion took place during the assem- blies. These assembly meetings continued to be a unidirectional information
201
flow from the board and District engineers to the delegates with no clear
decision-making procedures to follow. The association was a formal organisation, and as yet it did not appeal to,
or connect up with another set of existing practices of farmer initiatives, and cooperation at field level. This was partly because the basis of the organisation was formed by the different ejido organisations whose boundaries were differ- ent from the hydraulic canal boundaries.
In September 1989, the association, having acquired some experience with maintenance, felt confident to also participate in water distribution. In discus- sions with the association's board, the District decided to post a District exten- sion worker to each watet guard, and along each canal elected farmers were to supervise watet distribution, the so-called jefes de toma, or gate leaders. A group of extension officers worked intensively during September through to November. Ejido meetings were organised and along each canal farmers were elected gate leaders. At the end of November this new organisational structure for watet distribution was formally instituted in a general assembly of the as- sociation, whereby not only the delegates were present, but also all elected gare leaders, and E10perado ' s six watet guards. During January to March 1990 the new organisational set-up was successfnlly implemented on the left bank of E1
Operado. By establishing a new position (the gate leaders), the ässociation had estab-
lished a crucial nexus with the field level. Apart from the formally elected delegates who link the board (and District engineers) with the ejido, the jefes de toma now formed the link between the association and the field. This second link made monitoring the field situation possible. Field staff (canaleros), board members and the delegates of the water users could be held responsible for their actions by the water users. Moreover, a canal-based form of farmer representa- tion linked the association's formal organisation to other existing expressions of cooperation found at canal level.
Canal maintenance and turnover: a conc lus ion
The case study presented in this paper revealed that canal maintenance in E1 Operado improved considerably; more canals have been cleaned for less money, a maintenance fund was established, and active participation of water users increased. In our view, a number of changes in the system's management and in the organisational structure are responsible for this improvement. First, a direct link between means and ends was created, that is, between water fees and clean canals. Second, project staff increasingly became accountable to their clients instead of towards higher authorities. Third, the ejido, the basic organisational entity upon which the organisational structure of the irrigation
202
system relied, gradually was replaced by a canal-based system of representa- tion. Fourth, the gap between water distribution and canal maintenance was bridged in the field by creating the position of gate leader, and in the office by bringing the responsibility over water distribution and canal maintenance together in one department.
A new organisational structure linking means, responsibilities, communica- tion channels and accountability issues, has made management in E1 Operado more effective. It should be noted, however, that the changes mentioned above took place gradually and informally. These changes had not been planned in advance by any of the parties involved, but rather were an outcome of an on- going process of negotiation and struggle between groups of people with differ- ent interests. In this government-managed irrigation system, canal main- tenance was not an empty arena, void of farmers' initiatives.
What, then, can be concluded from the case study presented in this paper? The first conclusion is that an irrigation system creates its own organisational culture of formal and informal ways of coping with an infrastructure and of dealing with each other. The restructuring of an irrigation system's organisa- tion should be based on these existing expressions of cooperation and initiative of the actors involved. This implies that for each irrigation system a tailor- made intervention strategy should be developed.
Another conclusion is that when compared to the analysis of water distribu- tion, analysing canal maintenance reveals another set of characteristics of a system's management. First, canal maintenance involves considerably more resources (labour, capital and machines) than water distribution. Second, proper canal maintenance is a condition for adequate watet supply, whereas water supply only indirectly influences the need for maintenance. Third, in contrast to water supply, which is immediately productive and entirely directed to short-term goals, canal maintenance involves the reproduction of the canal system and thus has a longer term perspective too. The costs involved in canal maintenance can thus be considered investments. Fourth, both canal main- tenance and water distribution have a direct bearing on a system's organisa- tion; as watet distribution often results in competition among water users along the same canal, the need for canal maintenance may bring these people together in cooperation, and this may help to forestall possible conflicts over water supply. 13
Seen in this light, the analysis of canal maintenance reveals another level to the relationships between different groups of people involved, and to their rela- tionship vis-ä-vis the canal infrastructure. The study of canal maintenance, therefore, has the promise to reveal new insights into the strengths and weak- nesses of a system's management. TM Our third conclusion is, that given the re- cent trend of turning over management responsibilities to water users, these in- sights create an opening to improving this type of interventions; thereby
203
making canal maintenance a key to restructuring irrigation management. It is therefore surprising that in much of the literature on irrigation management, canal maintenance receives only secondary attention.15
Notes
1. The project has a budget of 1.2 billion US dollars, half of which is f inanced by a World Bank
loan. Rehabilitation and other construction works represent 77% of the entire budget, whereas
1.5% is desfined to institutional development and support of water users organisations.
(World Bank 1992: 12) 2. The material upon which this paper is based was collected during fieldwork from 1987 to 1990.
3. Wate t distribution efficiency is here defined as the volume of irrigation water reaching the
fields in proport ion to the volume entering the system's head works. 4. A similar situation occurred in au tumn 1987, when the sediments in 5 secondary canais created
problems for 300 water users (1,500 ha). The watet users managed to solve these problems in
different ways. An unknown number of other canais were also silted up in that year.
5. Yield reductions were not measured, but a figure o f 5°70 seems conservative, or an estimated
financial loss o f between US$ 30 and 100 per hectare, depending on the crop.
6. Data on actual expenditures of the department could not be obtained. 7. The budget is finally approved and allocated by the central office o f the SARH ministry in
Mexico City. No provisions are made for the depreciation of this budget through the irrigation
season because of inflation (in 1987/88 running between 10 and 15070 per month) .
8. An ejido is an agrarian communi ty which corporately holds lands, received as a graut f rom
the government mainly during the 1930s and 1940s. Its members are called ejidatarios. Ejidatarios may not seil their ejido land, and they depend on the ejido organisation for many
services, such as agricultural credit.
9. More frequent turns are important for many of the plots along lateral 2, which have coarse-
textured soils. 10. Farmers are weil aware o f this, but to elect another leader is difficult (1) becanse as yet the
canal organisation has no formal structure, no t are regular meefings held; the position ofjefe depotrero is only formally recognised in the ejido, which is quite a different setting; (2) some-
one who proposes that Don Enrique has to resign, himself risks becoming the subject of back-
biting.
11. This summary is not only based on the cases presented in this article but also on detailed analy-
sis of farmers ' mobilisations along other canals during 1987 and 1988.
12. Note that the government was heavily subsidising the irrigation systems (Olaiz y Pérez et al.
1986).
13. Martin (1986) found for several farmer managed irrigation systems in Nepal that joint canal maintenance forestalled eonflicts in water distribution.
14. Including canal maintenance as a topic in so-called rapid appraisal missions seems self-evident,
as Yoder & Martin (1985) do. Chambers & Carruthers (1986), however, do not refer to main- tenance as an important diagnosing indicator.
15. A recent literature survey on irrigation water management (3urriëns & De Jong 1989) does not
dedicate one paragraph specifically to canal maintenance. The same holds for Robert Chambers ' work on canal irrigation on South Asia (Chambers 1988), or Uphof f ' s work on irri-
gation bureaucracies (Uphoff 1991). Welcome exceptions are Bottrall (1981) who includes maintenance in his survey of four irrigation systems; and Merrey (1986) who presents a case
study of water course improvement (a speeial case of canal maintenance) to reveal cultural pat-
terns which underly irrigation practice in a village in Pakistan.
204
References
Bottrall A.F. 1981. Comparative study of the management and organization of irrigation proj ects. World Bank Staff Working Paper no. 458. World Bank, Washington
Chambers R. 1988. Managing Canal Irrigation. Cambridge University Press Chambers R. & Carruthers I. 1986. Rapid appraisal to improve canal irrigation performance: ex-
perience and options. IIMI Research Paper No. 3. IIMI, Digana Coward Jr. W.E. 1986. State and locality in Asian irrigation development: the property factor.
In: Nobe K.C. & Sampath R.K. (eds.), Irrigation management in developing countries: current issues and approaches. Studies in Watet Policy and Management, No. 8 (pp. 491-508). West- view Press, Boulder.
Jurriëns R. & de Jong K. 1989. Irrigation water management, a literature survey. Working party Irrigation and Development. Wageningen Agricultural University.
Martin E.D. 1986. Resource mobilization, water allocation, and farmer organization in hill irriga- tion systems in Nepal. PhD. thesis. Cornell University, Ithaca
Merrey D.J. 1986. Reorganizing irrigation: local level management in the Punjab (Pakistan). IIMI Research Paper no. 4. IIMI, Digana
Olaiz y Pérez A., Ortega Gil G.E. & Ortiz Rendón G.A. 1986. Reformas a la legislación fiscal en materia de aguas. Ingenieria Hidräulica en México (enero-abril de 1986) 65-71.
Sijbrandij P. 1989. Irrigation and human organization, searching for the meaning of maintenance in a government-managed irrigation system in Western Mexico. MSc. thesis. Dept. of Irrigation and Soll and Water Conservation, Wageningen Agricultural University
Uphoff N. (with Ramamurthy P. & Steiner R.). 1991. Managing Irrigation Bureaucracies: Analyz- ing and Improving the Performance of Bureaucracies. Sage, New Delhi
World Bank. 1992. Development Bank Business Report. World Bank, Washington Yoder R. & Martin E. 1985. Identification and utilization of farmer resources in irrigation de-
velopment: a guide for rapid appraisal. ODI Irrigation Management Network Paper 12c Zaag van der P. 1992. Chicanery at the canal; changing practice in irrigation management in
Western Mexico. Latin America Studies, 65. Cedla, Amsterdam
