THE PIVOTAL ROLE OF CANAL OPERTHE CASE OF THE CANALERO
positions similar to those of the canalero exist. Canal operators seem to play similar key roles yet no systematic review
dirrigation petite et moyenne chelle et grs manuellement ont aussi dvelopp des comptences similaires celle du
IRRIGATION AND DRAINAGE
Irrig. and Drain. (2012)
Published online in Wiley Online Library (wileycanalero. Bien quil ny ait aucun examen systmatique ou aucune analyse comparative de leur activit, les ayguadiers semblenty jouer des rles cls similaires. Ce document est une premire contribution expliquer pourquoi ces agents de terrain peuventeffectuer ces rles importants et continuent de le faire. Copyright 2012 John Wiley & Sons, Ltd.
mots cls: canalero; canal dirrigation; personnel de terrain; gestion de lirrigation; Mexique; distribution deau
INTRODUCTION* Correspondence to: Pieter van der Zaag, UNESCO-IHE, P.O. Box 3015.or comparative analysis of their position exists. This paper makes a rst contribution to explain why such eld levelstaff can perform such signicant roles and continue to do so. Copyright 2012 John Wiley & Sons, Ltd.
key words: canalero; canal irrigation; eld level staff; irrigation management; Mexico; water distribution
Received 25 June 2011; Revised 21 November 2011; Accepted 25 November 2011
Ce document prsente des donnes empiriques sur le rle des ayguadiers, encore appels canaleros, dans un systme dirrigationau Mexique au cours dune priode de deux dcennies. Linfrastructure de canaux ciel ouvert est quipe de portes et de prisesdeau rglables manuellement et manuvres par les canaleros. Pendant la priode dobservation la fois la conguration delinfrastructure et sa gestion ont chang. Le cas du canalero montre comment le personnel de terrain, a priori peu quali, joueun rle important dans la planication et la mise en uvre de la distribution deau. Le canalero apparait comme un acteurincontournable du fonctionnement du systme. Il a cr ses propres champs daction semi-autonomes ainsi quun domaine decomptence dont il drive un certain degr dautorit.Nous avons compil dautres tudes de cas et effectu des entretiens lectroniques avec des experts. Dautres systmesPIETER VAN DER ZAAG1,2* and EDWIN RAP3
1UNESCO-IHE Institute for Water Education, Delft, the Netherlands2Department of Water Resources, Delft University of Technology, Delft, the Netherlands
3Irrigation and Water Engineering Group, Wageningen University, Wageningen, the Netherlands
This paper presents empirical material of the role of canal operators, canaleros, in one irrigation scheme inMexico over a period oftwo decades. The open canal infrastructures are tted with manually operated adjustable gates and intakes. During the period ofobservation, both management set-up and the canal infrastructure changed. The case of the canalero shows how low-ranked eldpersonnel play an important role in scheduling and implementing water distribution. The canalero emerges as a key actor whomakes the system work. Canaleros have created their own semi-autonomous eld of action; an area of competence from whichthey derive a certain degree of authority.The case study ndings are compared with relevant published sources complemented with electronic interviews with
experts. In large- and medium-scale open canal irrigation systems with exible and manually operated irrigation devices,2601 DA, Delft, the Netherlands. E-mail: email@example.com Le rle cl des ayguadiers dans les systmes dirrigation: Le cas ducanalero
Copyright 2012 John Wiley & Sons, Ltd.ATORS IN IRRIGATION SCHEMES:
onlinelibrary.com) DOI: 10.1002/ird.693The role that canal operators play in the water distribution ofirrigation systems around the world is not precisely knownAlthough the name of this eld-level staff varies (canalero.
P. VAN DER ZAAG AND E. RAPin Mexico, ditch tender in the United States, sectoristain Peru, tomero, aguador and regador in Spain, ghafr inSudan, lascar and neerkatti in India, oeloe oeloe inIndonesia, water bailiff in Australia), their roles are similar.They constitute an occupational group of eld workers whomediate the service relation between irrigation organizationsand their water users. Disperse footnotes from colonialrecords to contemporary studies make mention of this actorin everyday irrigation management, yet overall, their roleremains largely invisible and underappreciated. The goalof this study is therefore to visualize and recognize theircontribution to irrigation performance.This study describes in detail the activities of eld
personnel in charge of water distribution in a medium-sized(approximately 10 000 hectares) irrigation scheme inMexico. They are locally known as canaleros: they mustensure that water is distributed from the main canal to thelateral canals and from these to individual farm plots. Theydetermine water distribution programmes and physicallymove the weirs and sluices. In this exible surface irrigationscheme, water distribution is quite complex since waterusers are free to choose the crop they want to cultivate,the date of sowing and time of irrigation. Furthermore,controlling the water ows is generally considered asensitive issue associated with power and political inuence.The canaleros form the frontline of irrigation systemmanagement since they are located closest to the farmers.Despite the fact that this position exists in virtually allnon-automated surface irrigation systems of scale, it hasreceived relatively little attention in the literature. This wasso in 1992 when rst mentioned (Van der Zaag, 1992a;1992b) and remains today a valid observation.What will emerge from the material presented in this
study is that the role of water distributors has not beenradically transformed by new policy, organisational struc-tures and technological innovations. Irrigation ManagementTransfer (IMT), the transition from a government-managedto a user-managed system as well as technological moderni-zation, has not eliminated the crucial role of eld staff in theoperation of irrigation systems. There is continuity in howthey perform their tasks, even when the organisational andtechnological setups in which they operate change. This isindicative of the fundamental nature of their task.Notwithstanding, the literature on irrigation management
has a peculiar blind spot regarding the contribution of eldpersonnel to the performance of irrigation systems. Thiscould be because many studies of irrigation performanceadopt the perspective of the engineers who design irrigationinfrastructure or those that manage it. Not surprisingly, theytend to focus on the role of engineers, government ofcialsand managers in irrigation institutions (Chambers, 1988;Wade, 1982; Uphoff, 1986). The literature on irrigationengineering is particularly geared to reducing the humanCopyright 2012 John Wiley & Sons, Ltd.element, which is frequently viewed as an obstacle toachieving efcient irrigation. Schuurmans et al. (1999) forexample states:
Although operators are able to adapt to unforeseen changes inthe behaviour of the canal, their performance is limited byfatigue and loss of concentration, to compensate for which theyneed to work in shifts. Furthermore, one operator can operate alimited number of control structures at a time. As a result ofthese factors, the costs of operation are relatively high.
Others have suggested that canal operators are rent-seekers: they ask and receive bribes for preferences in waterallocation and distribution (Wade, 1982; Huppert andWolff, 2002; Rinaudo, 2002). The engineering solution iseither to simplify the hardware (xed distribution works asproposed by Horst, 1990) or to add software to the hardware(automated water control as proposed by many includingPlusquellec et al., 1994; Clemmens et al., 2005).This paper takes a different view and argues that human
competence is fundamental to making exible irrigationtechnology work. Canaleros are neither fully constrainedby the technical and organisational structures in which theyoperate nor do they simply execute managerial orders.Rather, they actively shape their working environments.The dialectic relation between canaleros and the socio-technical environment in which they operate makes it interest-ing and relevant while investigating their role over time. Thispaper is therefore a longitudinal case study of one irrigationscheme in Mexico located in the valley of Autln-El Grulloin western Mexico (Van der Zaag 1992a; Rap, 1993, 2004;Long, 2001; Schippers, 2009). The empirical observationsthat form the basis of this paper were made during threeperiods: 19871989, 19921993 and 2009. This allowedfor a unique analysis that spanned 22 years of continuityand change in the professional performance of irrigationpersonnel against the backdrop of institutional reform andinfrastructural modernisation. Research methods such asparticipant observation, discharge measurements, eldexperiments, interviews and document analysis made itpossible to systematically follow one irrigation system overmore than two decades and to compare, conrm and contrastoriginal observations with more recent ndings. In addition,we refer to studies performed in other areas of the world andconducted electronic interviews with 14 irrigation experts toassess the wider occurrence of the phenomenon under study.In the following section, we characterize the specic type
of irrigation system where we observed the canalerospractices. Sections 3 and 4 describe the work routine ofone particular canalero, Miguel, and the strategies hedevised during the 19871989 period (Van der Zaag,1992a). Section 5 investigates the extent that the role ofthe canaleros evolved in the 1990s (Rap, 1993, 2004) andmore recently in 2009 (Schippers, 2009) as a result ofIrrig. and Drain. (2012)
the time of the rst research (19871989), Miguel was in
ROLE OF CANAL OPERATORS IN IRRIGATION SCHEMEShis early forties. He was born in a small village in the valleyof Autln-El Grullo, the second child in a family of 12. Thefamily was poor and did not own land. Having followedinstitutional and technological changes. In Section 6, wepresent discussion and conclusions.
THE AUTLN-EL GRULLO IRRIGATIONSYSTEM IN THE 1980S
Constructed in the 1950s, the Autln-El Grullo irrigationdistrict is a medium-scale irrigation system that receiveswater from an upstream reservoir. It was designed andconstructed by a competent Mexican hydraulic engineeringconsortium. The scheme consists of a network of opencanals and exible division structures, with the main systembased on the principle of downstream control. In 1987, theAutln-El Grullo system irrigated 8,700 ha of which 6,000ha were planted with sugar cane. Water distributionefciency, dened as the volume of irrigation waterreaching the elds in proportion to the volume entering thesystems head works, was estimated at around 60%. Untilthe late 1980s, water distribution and canal maintenancewere performed by the district: the local ofce of theMinistry of Agriculture and Hydraulic Resources (SARH).In 1990, irrigation management was transferred to the localWater Users Association (WUA) and some of the eldpersonnel was employed by the new management entity(Rap, 1993).The head of the district operation department is responsi-
ble for water distribution and is supported by staff at thedistrict ofces. The lowest level staff in the department arethe six canal guards or canaleros, the eld personnel whodistribute the irrigation water. They ensure that water isdiverted from the river into the main canals, then from themain into the lateral and sub-lateral canals, and from thereto individual farm plots. The canaleros physically movethe manually operated and adjustable weirs, sluices andvertical sliding gates. They elaborate rules of thumb as tohow water ows change after lowering a particular gate bya certain number of screw threads. They work out (in theirheads, not on paper) the water distribution programmes.Canaleros form the frontline of the district since theycommunicate every day with both farmers and districtengineers. Farmers put forward requests for irrigation turnsto their canalero and if the requests are deemed reasonable,the farmers will receive water normally within one to vedays. Farmers themselves are not supposed to move gates.
The daily work routine of a canalero
To appreciate the position the canalero nds himself in, thissection describes the practices of one canalero, Miguel. AtCopyright 2012 John Wiley & Sons, Ltd.four years of primary school, Miguel was sent at 13 to anaunt in Guadalajara to work in the shoe industry. Hereturned home at the age of 17. His father, who had workedas a labourer in the construction of the irrigation systemduring the 1950s, eventually managed to arrange him apermanent job in the maintenance brigade. Miguel workedfor two years with his machete cutting weeds in the canals.Then, in 1970, he was asked to become a canalero. In1972, he married Celia and they had three children. Hebought a house from his brother-in-law, and with the sevencows Miguel had accumulated over the years, he bought asmall shop. In 1981, Miguel was able to buy a second-handpick-up truck from the melon prots produced on four ha ofshare-cropped land.In 1987, Miguel is responsible for an area of 1,700
hectares divided into 500 elds to which he has to allocateirrigation water. In total, he deals with 350 water users.He controls the gates over 50 kilometres of lined canals.There are 30 gates and main intakes along his 18 km stretchof main canal. He is responsible not only for his ownwater ow (700 to 2,000 litres per second) but also the owsdestined for the downstream zones of Gabriel and Pedro.Normally, between 2,000 and 5,000 lps ow through hiszone. This creates an extra burden to his task: his colleaguescan easily accuse him of using part of their water. Thismeans he has to be careful in adjusting the gates in the mainchannel. Similarly, it is also a great advantage: an upstreamzone is never short of water. During the irrigation season,Miguel works 60 hours per week. Each day he drives hismotorbike between 50 and 80 kilometres through his zone.The major task of his work is to check that all the irrigationturns are running well, adjust gates and talk to the farmers ortheir labourers (Figure 1).In addition to his role in distributing water ows, the
canalero also plays a role in managing information owsin the eld and with the ofce. When he is driving throughhis zone observing cautiously all the elds, sluices andgates, farmers have little difculty in signalling him.Arriving at a plot where someone is irrigating, Miguel willpause. In seconds, he can observe the situation and judgewhether it is to his expectation. He will often chat with thefarmer or labourer irrigating. This chat is partly aboutirrigation matters but invariably other subjects are discussed.This is useful because the canalero needs a lot of informa-tion: how his farmer is doing, what is happening on this orthat plot, who is the new share-cropper of such and such aplot, etc. Farmers need information too: current prices ofagricultural products, the names of buyers, new varietiesof crops, agro-chemical products and amounts of fertilizerto apply.Another recurring theme of these conversations with
farmers is the suspension of sugar cane plots. Four to eightweeks before harvesting, sugar cane plots have to dryIrrig. and Drain. (2012)
P. VAN DER ZAAG AND E. RAP(this enhances sugar content but it is necessary because ofthe lorries entering the eld) and irrigation services aresuspended. Repeatedly, Miguel urges farmers irrigating nearthese suspended plots to irrigate with care. Farmers some-times complain when they are denied an irrigation turn untilan adjoining sugar cane plot has been harvested; this is donein cases where the eld canal leaks. This is a commonsource of irritation for farmers who grow crops other thansugar cane such as maize that needs water more frequently.They do not prot from the suspension. While performinghis daily rounds, Miguel cannot take too long to solve suchgrievances. He is under pressure to continue. At 2:30 P.M,he has to be present at the daily meeting of canaleros inthe district ofce. In addition, he may have other things todo such as tend to his own crops (most canaleros share-cropsome land) or some other business or return home tohave lunch.Apart from the 15 to 25 water users irrigating, Miguel will
also encounter perhaps 10 to 20 farmers on his daily roundswho request an irrigation turn. He is often called as he drivesby. The moment he sees who the farmer is, he is alreadyprocessing the relevant information: that is farmer A with
Figure 1. Miguels zone of approx. 1,700 he
Copyright 2012 John Wiley & Sons, Ltd.plot number B lying at canal C with crop D, which had itslast irrigation turn on date E. When the farmer then asksfor an irrigation turn, Miguel has already processed thedata and will respond within seconds by saying: no, youirrigated your plot two weeks ago, you have to wait anotherweek; or, you will have to wait until farmer F has nished,and F will start irrigating when G has nished, so your turnwill be Friday at 6 pm. Sometimes he will take hisnotebook from his small rucksack in which he writes thename of the farmer, the acreage, the crop and the dates ofthe irrigation turns received for each plot. When the farmerdoes not agree with a decision, Miguel refers to his book.He will look up the plot and show the farmer the date ofthe last irrigation turn and insist that he still has to wait.Miguel, however, knows most of the necessary informationby heart.It is often the canalero who looks for a particular farmer.
Usually this happens when the canalero has worked out anirrigation schedule for a certain lateral canal. Perhaps, forexample, farmer H has nearly nished his turn and farmerI has already applied for his, but Miguel judges that afterH, it would be easier if farmer J followed. He will try to
ctares (source: Van der Zaag, 1992a)
Irrig. and Drain. (2012)
water rst, which he has to pass on to the downstream
so within a period of perhaps one week he has to re-establish
ROLE OF CANAL OPERATORS IN IRRIGATION SCHEMEScanalero, adjusting all the gates in his part of the main canal.The downstream canalero will get his extra water two tofour days after his request. He can now implement theadditional irrigation turn and adjust the inlet from the mainto the lateral canals and other sluices in the laterals andsub-laterals.locate farmer J and inform him to irrigate on a particularday. Sometimes he cannot nd the farmer (those who growsugar cane often do not show up in the eld for a week ormore). This is one of the things that irritates Miguel. Heconsiders them lazy and disrespects their taking little interestin their crop.
Programming a canal
The canalero works out a water distribution schedule in anyone canal based on the differences in water demand eld-by-eld, dependence on crop and soil, optimal distributionpatterns, dependence on the canal system and farmer requests.The pattern evolves during the irrigation season. Waterdemand gradually increases from November to Marchbecause of climatic conditions and crop stage and thendecreases again. The canalero thus builds up an increasinglycomplex water schedule along each canal in his zone. Oncea basic irrigation schedule has been established at thebeginning of the irrigation season, the canalero can add anextra irrigation turn without reconsidering the whole irrigationplan. From a simple core pattern, a complex schedule ofirrigation turns evolves that is well structured and accountsfor the differences in water need that exist from plot to plot.Take for instance a large lateral canal irrigating 600 ha. At
any one time, the canalero may have six irrigation turns`working. With an average plot size of 3.5 ha and a waterow at eld level of 70 lps, a normal irrigation turn will takebetween 21 to 32 hours. This means that every day or so thecanalero has to ensure that six other farmers are ready totake over these turns. When he sees he is falling behindschedule, that is, the six turns working are not enough toensure that each plot gets its water in time, he will decidethat extra water is needed in the lateral canal. At rst, hemay try to avoid this since it requires considerable efforton his part. If the situation in the lateral gets worse, he willbe forced to programme an additional turn.First, he will have to consider whether it is possible to
take an irrigation turn from another canal in his zone. If thisis not possible, as during January and March when waterdemand is on the increase, he will need more water enteringhis zone. Therefore, he will inform the engineer at the dailymeeting of canaleros that he needs extra water. When hisrequest is granted and the increase in discharge reaches themain canal, the canaleros cooperate to send this extra waterto the right zone. The upstream canalero receives the extraCopyright 2012 John Wiley & Sons, Ltd.the complex irrigation order without having a `core uponwhich to slowly reconstruct the schedule. For him, this willbe the most critical week of the year.The canalero has to work in a systematic manner to cope
with his task. He does not, however, make elaborateprogrammes before entering the eld. He structures thewater distribution pattern in a practical immediate way. It isa tough schedule to develop and his freedom of action islimited. Nevertheless, it is this domain of programming,adjusting gates and estimating water ows from which hederives his competence. He is the only one who has the properlocal or situated knowledge and experience to do the job. Hisdemanding task thus produces specic knowledge, which inturn gives him a certain degree of authority and freedom indecision making. The canalero not only needs this freedomof action to be able to do his job properly, he may also use itfor his own benet. This may vary from favouring farmerfriends to irrigating his own share-cropped plot morefrequently than is formally permitted. We present thefollowing case to demonstrate the intimate knowledge thatcanaleros have of canal infrastructure and their competencein controlling hydraulic ows.
THE EXPERT AND THE ENGINEER
One afternoon, Miguel invited a researcher (the author:Pieter van der Zaag) to review a situation that botheredhim. We arrived at the intake of Lateral 5 (located 5kilometres downstream of the intake of the main canal) thatserves 500 hectares (Figure 2). Here, Miguel had establisheda represo in the main canal; that is, by lowering the radialgate immediately downstream of the off-take, he had backedup the water in the main canal. In fact, he had already usedsome of the freeboard, the water standing less than 15centimetres below the canal crest. Miguel conrmed thatthis was the maximum head possible and thus no more watercould possibly enter the lateral canal that had an intake pipeThe canalero allows an extra irrigation turn (somewherebetween 70 and 100 lps) to be superimposed upon existingwater ows, which is not so difcult to implement. He doesnot reconsider the ow pattern. At each gate, he considersthe extra water to be passed on and assesses the adjustmentshe has to make in terms of the number of screw threads orturns of the hand wheel operating the gate. There are twosituations that may severely disrupt the established waterschedule: interference by a superior (see further) or suddenrains during the dry season. If enough rain has fallen (morethan 3050 mm), the immediate result is that all water usersstop irrigating. This may indeed be welcomed by thecanalero. However, he fears the consequences. In two tothree weeks, all water users will urgently want to irrigate,Irrig. and Drain. (2012)
; SL =
P. VAN DER ZAAG AND E. RAPthat was too small (18 inch). That morning, the water gaugemeasured that 559 lps was entering the lateral. But, Miguelsaid, it is not enough. I am running behind schedule. Beforenishing a complete round of irrigation turns, farmers arealready requesting the next.It was April and the period of peak water demand. The year
before, Miguel had managed intake 5. However, during thisirrigation season, 30 hectares of previously unirrigated landhad been added, of which 20 hectares were sugar cane andanother 30 hectares had changed from annual crops to sugarcane. Water demand had thus increased at least 10%compared to the previous year. We discussed the possibilitiesfor increasing the discharge. The researcher said that the onlyway to augment the ow was to further back up the water inthe main canal. This would only slightly increase the waterow into the lateral (perhaps a 5% increase, the inlet being apipe) but any increase would help. Miguel dismissed the
Figure 2. lateral 5 canal with irrigated eldsresearchers suggestion outright: the lateral canal wouldreceive hardly any more water and further backing up themain canal would disturb the main system and temporarilydecrease the water owing into the zone of the next canalero(Gabriel) at kilometre 18. According to Miguel, it would takemore than 12 hours before the water ow would stabilise. Theresearcher, however, calculated that lowering the gate to backup water in the main canal at 6 pm would only be felt by thedownstream canalero hours later when it would be night.The following morning, the situation would be normalisedand because the canalero would not visit his zone during thenight, he would not even notice any change. Miguel did notaccept the outcome. He insisted that by lowering the gate inthe afternoon at kilometre 5, the water gauge would registera considerable decrease of discharge in Gabriels zone thenext morning at 9 A.M. and that this would cause problems.Miguel, however, could not explain why an increase of 10centimetres in the water level would take such a long time to
Copyright 2012 John Wiley & Sons, Ltd.normalise at kilometre 18. Therefore, the discussion endedand they went home.Late afternoon, however, Miguel lowered the gate at
kilometre 5 so that the water in the main canal increasedby another 10 centimetres: the water standing level withthe road alongside the canal. At 9 P.M., he sat at home butfelt uneasy. Would too much water enter into the lateralcausing the lateral canal to overow? Would the main canalovertop? He went on his motorbike to check but everythingwas OK. The next morning, Miguel was waiting for thewater gauge at the intake of lateral 5. He was now convincedthat backing up the water in the main had decreased thedischarge in Gabriels zone considerably. He wanted badlyto draw the gate up again to undo the changes made.However, he also wanted the water gauge to measure thedischarge into the lateral to see whether if it had increasedsignicantly. When the water gauge nished measuring
sub-lateral (source: Van der Zaag, 1992a)(with a propeller meter), Miguel quickly drew up the gate.The reading of the water gauge showed that there wasno increase in the water running into the lateral (reading557 lps) and the researchers suggestion proved useless.Miguel felt uneasy; he expected trouble with Gabriel duringthe canaleros meeting that afternoon.When the meeting started, Gabriel immediately accused
Miguel of stealing water from him (200 lps). Then, thedischarge readings were read and Gabriel had been right.The discharge in his zone had decreased by 263 lps whereasin Miguels zone, it had increased by 211 lps. The wave(the decrease in discharge) had been exactly at the spot wherethe water gauge measured Gabriels water attributing the lostwater to Miguels zone. The following day, the wave hadreached Pedros zone downstream of Gabriel while thereadings for Miguel had normalised. Now Gabriel had to bearthe brunt: it now appeared that he had taken water from Pedro.The next day, the situation persisted (Table I).
Irrig. and Drain. (2012)
Source: data from Distrito de Desarrollo Rural, El Grullo
ROLE OF CANAL OPERATORS IN IRRIGATION SCHEMESFor the researcher, this was an embarrassing experiencebecause Engineer Corona, head of the operation department,had also started to question Miguel. How could this havehappened? Miguel then explained that it was an experimentto try to get more water into this lateral canal because therewas a severe water shortage, and this in turn was causedby the expansion of the area under sugar cane (implicitlycriticising his superior for having allowed this). The calcula-tions of the researcher (an engineer) were incorrect. He hadseverely overestimated the velocity of the water wave whereasthe canalero had made a much more realistic estimate.This case shows that the canalero had detailed and
relevant knowledge about the canal infrastructure and waterows that enabled him to execute his role of waterdistributor competently. Miguel simply `knows how theinfrastructure and water ows react to certain manipulationsof the gates. His knowledge about the infrastructure is bothsituation-specic and implicit. He knows how the canals inhis area react to particular manipulation but might notknow it for another situation. He could not convince the re-searcher. Yet, his knowledge proved to be far more effectivethan the formulae and assumptions of the researcher. It isimportant to note that the canaleros competence in con-necting the physical ows and elements in water distributionwere intertwined with his role of mediating between theTable I. Consequences of slightly lowering a gate at Km 5 in theright main canal at 6 P.M. on April 6, 1988. Discharge in lpsmeasured between 9 and 10 A.M
Zone 4 Zone 5 Zone 6 Lateral 5
day (Miguel) (Gabriel) (Pedro) (Zone 4)
April 6 2003 1367 908 559April 7 2214 1104 919 557April 8 1971 1568 754 569April 9 1941 1528 812 N/Airrigation managers and water users.
THE CANALEROS DILEMMAAND THE SHAPINGOF OPERATIONAL ARRANGEMENTS
The water is not waiting for you, the canaleros say. Whilethey have a continuous water ow running down the canals,the canaleros are bridging both farmer demands and(infra) structural constraints while seeming to be workingaccording to the rules. They have to have a good memory,an agile mind and act diplomatically with water users. Thatis the only way to solve this complex equation in a quickand simple way. Miguel:
Copyright 2012 John Wiley & Sons, Ltd.Its difcult. When I have problems in the irrigation season, Ifeel bad. At night I am thinking: what do I have to do thefollowing day? How will I tackle that problem? It is usuallythese two things: not to have trouble with the district, nor troublewith the people. And when the district orders me to do some-thing, then I have to do it. The moment it really becomes toughis when I have to explain to the farmers: why this, why that, inorder to keep me on good terms with the people. [17/8/87]
And 18 months later:
The most important thing for me is to have the water usersatised, to have the farmer contented. Unfortunately, some-times it is not possible to solve this problem. There are twomagnets and you stand in between them, that is, your bossesand the water user are dragging you in opposite directions. Thus,you have your xed water ow, so if I should give in to therequest of the water users, then I would not comply with theworking norms, since I would have to take water from thedownstream canaleros, and I would have trouble with them.[12/4/89]
The situation a canalero nds himself in can be calledthe canaleros dilemma: being caught between the twodifferent demanding worlds of engineers and farmers. Itcalls for constant choices that are often mutually incompat-ible. How does Miguel handle such situations? We nextconsider a potentially conictive situation with water usersthat Miguel had to handle carefully. This implies thatbesides water distribution and information exchange, thecanalero plays a role in conict resolution.
Seeing is understanding: the case of theincredulous farmers
The researcher met Miguel and again visited the elds oflateral 5 where there were problems because of the limitedcapacity of the inlet. Miguel looked tired. He recountedthe discussions he had with the farmers. He said he hadanticipated problems and that two years ago he had toldhissuperior that he should press the maintenance departmentto construct a new inlet with a diameter of 24 inches(compared to the actual 18 inch pipe) but nothing was done.At the beginning of the current irrigation season, the districtengineer had allowed an additional 50 hectares of sugar caneto be planted along this canal. Miguel had been angry andhad told the engineer that he would not be responsible forthe problems this would cause. Miguel concluded: now itis me who is facing the problems, adding we are like bullghters: we ght the bulls and the bosses are way up inthe stand yelling ol, ol.That morning, Miguel came across two water users from
the tail of the canal who had accused him of denying themwater. He related how he had tried to explain the situationto them, that because of the small inlet, no more water couldenter the lateral. But they did not believe me, he sighed,Irrig. and Drain. (2012)
To summarise Van der Zaags ndings (1992a) presentedabove, we distinguished three intermediary roles of the
P. VAN DER ZAAG AND E. RAPcanalero: water distribution, information exchange andconict resolution. Two minor roles are administration andmaintenance. In these different roles, canaleros are keypassage points for water and information ows; they mediatebetween different infrastructural and organizational levelsand connect and mobilize the water network (Van der Zaaget al., 2001) that consists of people, things and ows involvedin irrigation. A situational knowledge of the complexphysical, social and organizational processes allows them tocompetently distribute water and deal with the conicts anddilemmas that emerge in their daily practice.because they had seen that the lateral canal upstream wascompletely lled with water and yet I did not give themwater. They thought Miguel was favouring other waterusers. I tried to explain to them that the lateral was indeedfull of water because I had set the water level very high inorder to feed sub-lateral 1 (Figure 2), but that all that waterwas dead water. He could not convince them, so hedecided that the only way they would understand was toshow in practice that there was not much water. Therefore,he took them to the spot where the sub-lateral branches offfrom the lateral.
I opened the sluice that had backed up the water, and after a fewminutes they could see that of all that water only a small streamwas left. I nally convinced them, and they understood thatthere was in fact very little water, because of the intake of lateral5. [12/4/89]
A few days later when the researcher was talking to somefarmers of that canal, it appeared that the two water usershad been able to explain to other farmers why there wasinsufcient water running into their area. Despite a toughschedule that day, Miguel decided to take the time tocarefully explain the situation to the two complaining waterusers. Miguel could have adopted another attitude: Noticethe things you have to do! If I were a son-of-a-bitch, I wouldhave said: dont make trouble, I am giving the orders here,and I dont care what you think. But Miguel had learnedthat involving farmers by explaining in detail was the bestway to avoid clashes. Taking the time to explain thesituation to them was an investment. He anticipated thatthey would explain the issues to their fellow farmers andthat they would have less trouble in convincing them thanhe. This attitude avoided an escalation of the problem. Thisexample shows that Miguel solved the canaleros dilemmain a particular way. In the case of the incredulous farmers,he involved those who accused him of giving preferenceto others. Miguel made choices and thus consciouslyshaped his working arrangements and actively gavedirection to the outcome of his work.Copyright 2012 John Wiley & Sons, Ltd.CONTINUITY OF THE CANALEROS ROLEIN TIMES OF INSTITUTIONAL AND
The way in which the roles and work arrangements of thecanalero in this and other Mexican irrigation systems haveevolved is related to several signicant changes in irrigationmanagement in the 1990s and 2000s. During the 1990s,Irrigation Management Transfer (IMT) reduced the role ofgovernment and bureaucratic staff in irrigation management.As part of the neoliberal reforms in Mexico during theSalinas administration (19891994), some 2.5 million ha ofgovernment irrigation districts were transferred to the WaterUsers Associations (WUAs) (Comisin Nacional del Agua(CNA) (1994). The main objective of the IMT policy was toreduce public expenditure on irrigation by creating nanciallyself-sufcient WUAs that would pay the full operating costsof the irrigation districts (Gorriz et al., 1995; Johnson,1997). For this purpose, the irrigation fees were signicantlyincreased. In 1990, the Autln-El Grullo district was one ofthe rst Mexican systems where irrigation managementresponsibilities were transferred to the local WUA. Theirrigation system subsequently became a pilot project topromote the success of IMT policy to water users from allover the country and the Mexican policy model to foreignvisitors (Rap, 2006).The IMT policy promised to radically improve the
performance and accountability of irrigation managementthrough the participation of water users at all levels. Theexecutive board of the Autlan-El Grullo WUA that waselected by water users introduced direct user participation inthe operation of the irrigation infrastructure. The WUAsregulations of 1991 stated that gate leaders who were electedby the water users at every tertiary canal became responsiblefor the actual water distribution according to a rol de riego(xed order of irrigation turns). TheWUAs regulations stated:
All this resulted in more participation of the users, which at thesame time has brought an end to frictions and conicts betweenthem, applying irrigations turns correctly and in an orderlymanner.
This would seriously reduce the role of canaleros andhence their numbers. This was in line with an informalpolicy objective to reduce the inuence of corruptunionized eld staff according to some policy makers(Rap, 2004). However, this formal structure of user partici-pation was not sustained in everyday practice as was foundby a follow-up study in 1993 (Rap, 1993) and severalsubsequent visits during the following decade. To beginwith, the gate leaders were not paid for assuming theresponsibility of distributing water at a canal. Observationsand interviews demonstrated that this was a stressfulposition that often entailed tensions with fellow water usersIrrig. and Drain. (2012)
The modernization project
ROLE OF CANAL OPERATORS IN IRRIGATION SCHEMESwho wanted their turn earlier or longer. Further, it requireda competence to deal with unexpected situations andvariations from an earlier agreed order in water distribution.These less visible aspects of water distribution and conictmanagement were largely overlooked in the new institu-tional architecture after IMT. Whereas the water usersgenerally accepted the canaleros mediating role andauthority in such controversial situations, they did not nec-essarily accept the same from fellow farmers. In addition,the gate leaders remained dependent on the canalerosdecisions on water distribution and their exibility inadapting the xed order of irrigation turns at a canalaccording to changing circumstances. In the end, mostcanaleros in the Autln-El Grullo irrigation system decidedto work around the gate leaders, undercutting their authority.The canaleros of the Autln-El Grullo system thus
managed to reproduce and continue their central role inwater distribution. The rst general manager of the Autln-ElGrullo WUA contracted two experienced canaleros whohad worked for the irrigation district. He offered themto work, one on each river bank, for a wage much better thanunder government contract. The operations manager of theWUA who assumed responsibility for water distributionwas a young engineer who started out as an apprentice. Hewas introduced to the irrigation system by joining thecanaleros in the eld. This training period exposed him tothe daily irrigation practice and triggered his respect forthe work of the canaleros and their competent performancein water distribution. The two experienced canalerossubsequently trained two young men to become newcanaleros. The roughly 10 000 hectare now irrigated cameto be operated by these four canaleros in comparison tothe six that were active before the transfer. Hence, thenumber of canaleros was curtailed but not their role in waterdistribution. Canaleros maintained their central role inwater distribution and reproduced their specic situationalknowledge of how to deal with the water network.Despite the IMT and the concomitant institutional
changes, the canaleros continued to play crucial intermedi-ary roles in water distribution, information brokerageand conict resolution. An additional role that becameincreasingly strategic under the WUA was fee administra-tion. The WUA negotiated a nancial agreement to let thelocal sugar renery pre-nance the irrigation fee ofsugarcane growers at the beginning of each irrigationseason. This was crucial for the nancial basis of theorganization since it secured more than 70% of annualrevenue. However, the actual payment of the correct feesrequired bringing together disparate data on plot, crop,irrigated surface, land owner and water user. This reinforcedthe canalero as an obligatory passage point by being ableto link the administration of the sugar renery and the WUAwith detailed knowledge of the eld situation. Thus, evenCopyright 2012 John Wiley & Sons, Ltd.Since 2000, the infrastructure of parts of the Autln-ElGrullo irrigation system is technied, a particular form ofirrigation modernization whereby the open canal infrastruc-ture at the secondary and tertiary levels is replaced by anunderground pipe network (Schippers, 2009). This pipenetwork has an intake box at the primary canal from wherethe water is diverted into pipes of gradually diminishingdiameters as the network extends into the lower lying areas.For every four hectares, a eld intake (hydrante) services aseries of eld level pipes (multicompuertas) that distributethe water over the furrows.The Mexican government co-nanced these moderniza-
tion programs to induce water savings and improve wateruse efciency. Although it was not evident what the directbenets of these measures were in this relatively waterabundant system, farmers were generally content since thissubsidised technology signicantly reduced labour inputand simplied eld irrigation. The following scene selectedfrom detailed observations in 2009 illustrates severalissues regarding the implications of these technologicalinnovations for the canaleros role in water distribution(Schippers, 2009).Chico, the local canalero, drives his car into the area of two
lateral canals that are currently serviced by a pipe network.We encounter a small group of men alongside one of theirafter IMT, the canaleros dilemma, of being caughtbetween two different demanding worlds of engineers andfarmers still applied.A factor that facilitated the persistence of the canaleros
eld of action in this system was the relative abundance ofavailable irrigation water. Besides the Tacotan-dam(148 million m3 storage capacity) in January 1992, the newTrigomil-dam (324 million m3) was put in operation. Wateravailability increased, which obviously eased the relationsbetween irrigators and between irrigators and the canaleros.The increased water availability also led to an increase inirrigated area over the years from 8,700 ha in 1987, 10 000ha in 1997 to 11 500 ha in 2009, of which between65%-75% was sugar cane. After transfer, the number ofcanaleros initially dropped from six to two in 1990 and to fourin 1993 but bounced back to seven in 2009 because of theexpansion of the irrigated surface. The reduced number ofstaff in the operations department initially implied a majorincrease in work pressure for the canaleros. Their servinglarger command areas may have been partially facilitated byfaster information ows (apart from the existing radio system,they now could also rely on cellular phones), increased (orrather more comfortable) mobility (from motor cycle toVolkswagen Beetle, to pick-up truck) and better payment(but exible contracts and no union afliation).Irrig. and Drain. (2012)
plots. They signal Chico to stop. Two men approach the car.From their gestures and expression on their faces, Chicoconcludes they are angry. They start talking to Chico abouta fellow farmer who refuses to stop irrigating. He had toldthem that there was little water pressure reducing the waterow into his eld. One of the men says he desperately needswater at his plot because he replanted some cane crops. Chicolistens and then suggests that the pipe or eld intake may beblocked. He drives the car a little further and systematicallychecks one-by-one which part of the pipe is clogged. Helocates the spot where the line is clogged that is causing thelow ow. Then he notices that the iron mesh that preventsgarbage from entering the pipeline has been pulled up. Heswears and pushes it back into its proper position.As we return to the same area the next day, we drive to
problem by systematically manipulating the intakes. Hisproactive attitude underlines the canaleros approach identi-ed earlier: his efforts to understand and solve the physicalproblem underlying the conict, the manner in which hecommunicates with the conicting parties and the way hereprimands them for being part in the problem. To asserthis authority as a canalero, he visually demonstrates hiscompetence in solving the problem and thereby the conict.
DISCUSSION AND CONCLUSIONS
This paper has shown the important role that canaleros playand claim as intermediaries in the Autln-El Grullo irriga-tion system. In their daily work, the canaleros mediate
P. VAN DER ZAAG AND E. RAPthe same group of irrigators. Chico gets out of his car andgreets them. A man who made trouble yesterday seemscontent now with the water he is receiving while stillblaming others of irrigating far too long. Chico cautions thatthe eld intakes can be blocked sometimes, especially whenthe iron mesh in the intake box has been removed. Heexplains that he has requested for a weir to be constructedjust downstream of the pipe intake that would allow themto get more water. When we drive away, he comments tothe researcher: Did you notice how easy they were withme now? There is water now and they are quiet.This scene highlights the intermediary role that canaleros
continue to play in water distribution and particularly illus-trates their role in conict management even if the irrigationtechnology changes. The particular physical characteristicsof the new technology accentuate the canaleros technicaland social competence. The sub-surface pipe network is notdirectly visible to the farmers in the same way that the eldcanals are, which can complicate situations when it createsmisunderstandings between irrigators. Chico consciously triesto work out a solution for the potential conict with thefarmers in the eld. He actively tracks down the cause of the
Figure 3. IntermediarCopyright 2012 John Wiley & Sons, Ltd.tion scheme
Relation with the water users
ition of the canalerobetween different hydraulic and infrastructural levels toensure that water is distributed from the main to the lateralcanals and applied to individual plots. As an occupationalgroup, they also actively create linkages between waterusers and the irrigation organization and connect differentadministrative realities with the eld. Their professionalcompetence forms the middle ground upon which all thislinking is produced. By acting as intermediaries in waterdistribution, information exchange or conict resolution,the canaleros acquire specic skills and situational knowl-edge to deal with the diverse social, technological and orga-nisational arrangements involved in irrigation (Figure 3).The canaleros competence is directly related to the
type of infrastructure he has to work with and is exhibitedwhile operating the system. The typical exible canalinfrastructure makes the position of the canalero crucial.The irrigation infrastructure of the Autln-El Grullo systemis characterised by manually adjustable sliding gates andintakes. It is exible in that it can meet varying demandsfor irrigation water from water users. However, it is onlya potential exibility: the structures have to be operatedproperly for the system to become exible.
Relation with the irrigation Irrig. and Drain. (2012)
ROLE OF CANAL OPERATORS IN IRRIGATION SCHEMESDespite their low rank, eld personnel do not simplyexecute managerial guidelines but adapt them to the varyingneeds, constraints and contradicting demands experienced ateld level. In describing the daily routines of the canaleros,we emphasized their procient and situation-specicknowledge of the physical infrastructure, canals, waterows, elds, crops, farmers and irrigators: in short, thewater network in the section where they are working. Thecanalero derives competence and commands respect fromhis experience and knowledge of programming, adjustinggates and estimating water ows, but also from resolvingconicts between water users. Since their competence islocated within such diverse and complex water networkswith constantly uctuating pressures and ows, simpleinstructions, xed rotation schedules and standard bureau-cratic rules are usually not appropriate. Consequently, theyhave considerable discretion in dealing with these situations.In this sense, the canaleros as front-line workers are therural counterpart of the street-level bureaucrat but workin a less bureaucratized workspace (Lipsky, 1980).This longitudinal study observed how policy and techno-
logical transformations have changed the roles, technologiesand work arrangements of the canaleros in the last decades(Barley, 1990). After IMT, more emphasis has been placedon their role in fee administration and they have assumeda role in maintenance of the new pipe networks introducedby modernization. Further, new technologies that increasemobility and enhance communication have made it possiblefor them to serve larger operational areas. In addition, theirworking conditions have changed: improved wages andexible contracts. Nevertheless, diachronic observations indifferent periods reveal that the canaleros continue tomediate the service relation between the irrigation organisa-tion and water users in the Autln-El Grullo irrigationsystem. Engineers, farmers and eld personnel expressedthe view that the canaleros have been a constant factor inthe irrigation system. Our hypothesis based on this study isthat the canaleros have ensured continuity in the face ofchange because of their competencies to connect differenttechnological, organizational and administrative levels andpractices, which are essential to making the irrigationsystem work. The canaleros know how to deal with thedilemma of contradictory demands between managers andwater users and therefore constitute an obligatory passagepoint in everyday irrigation management (Latour, 1987).The irrigation literature has given little attention to
these canal operators. It tends to focus on the role ofirrigation engineers, bureaucrats and managers in theirhydraulic institutions (Chambers, 1988; Uphoff, 1986;Wade, 1982). The irrigation engineering literature isgeared towards reducing or downplaying the humanelement. The expectation is that xed structures or pipedsystems will remove unpredictable human behaviour andCopyright 2012 John Wiley & Sons, Ltd.rent seeking will disappear when irrigation systems areautomated.In addition, the literature on IMT has largely disregarded
the changing or constant roles of these eld workers duringand after the reform process. The IMT policy discourseassumed a radical change in the daily management ofirrigation systems by introducing user participation atdifferent levels. However, this longitudinal study shows thatin spite of these changes, there was a signicant degree ofcontinuity in the manual operation of the infrastructure andwater distribution in the eld. That continuity was providedby the occupational group of canaleros presently underWUA management. Water users nd it convenient toexternalize the stressful and conictive task of waterdistribution and irrigation managers need to keep some localcontrol over what happens in an open canal system andcanaleros know and need their job.The ndings of this study have a relevance beyond the
particular case studied because they reverberate withobservations from many other parts of the world. In theliterature dealing with government managed irrigationsystems in Mexico, various authors briey refer to the peoplecalled canalero or fontanero (Finkler, 1974; Hunt and Hunt,1974; Finkler, 1978; Mares, 1980; Martnez Vzquez,1980). Authors suggest that canaleros usually take bribes,which warns us about glorifying the role of the canalero.Finkler stressed that water users have to maintain a goodrelationship with the canalero, otherwise they might facewater shortages. Many water users consciously invest in thisrelationship. Martnez Vzquez is the only scholar who statedthat district engineers manipulate canaleros to their ownbenet. However, none of these authors have remarked onthe productive nature of the canaleros work, which emergesfrom the evidence presented in this paper.A PhD study on the Left Bank irrigation district of the
River Santiago in Nayarit and several other systems inWestern Mexico conrmed that after transfer, the canaleroscontinued to form an occupational community with a sharedbody of practices; a culture and an understanding ofirrigation management that is rooted in past practice (Rap,2004). A difference with Autln-El Grullo was that this grouphad become more fragmented and divided in contrast tothe close group that we depicted here. Under thenew management regime, labour relations and workingconditions have signicantly deteriorated for the canaleros.As they are no longer protected by unions and xedcontracts, they are subject to increased job insecurity andnancial dependency on management. This has created thepossibility for management to divide and rule over the groupof canaleros and to let patron-client relations and politicalinterests gain more inuence over their employment andwork. As a consequence, there is more circulation andless continuity within this occupational group (Rap, 2007).Irrig. and Drain. (2012)
P. VAN DER ZAAG AND E. RAPThe only other elaborate discussion of the position of thecanalero we know of is provided by Ter Hofstede and VanSantbrink (1979) who discussed the role of the oeloe oeloeor water distributor in irrigation systems in Indonesia duringcolonial rule. The oeloe oeloe originally formed part of thevillage council. Through interventions by the colonialpower, the oeloe oeloe gradually became an employeeof the government. The oeloe oeloe emerged as a keyintermediary, cunningly used by the colonial governmentto extend control over irrigation affairs into the tertiaryunits. The oeloe oeloe in Indonesia must have found himselfin a similar situation to the canalero in Mexico. They bothfaced the canaleros dilemma.Vos (2005) and Vos and Vincent (2011) pointed out how
eld operators skills and experience of operating undershotsliding gates contributed to good water delivery performanceand volumetric water control in a large scale open canalirrigation system at the northern coast of Peru. Vos (2005)also referred to other studies from Argentina, Arizona andSri Lanka that equally demonstrated the ability of operatorsto control water ows and get required discharges and waterlevels by changing gate settings, sometimes even better thana computer (Murray-Rust and Snellen, 1993; Clemmenset al., 1994; Sloan, 1997; Godaliyadda et al., 1999).In the Valencian irrigation system, the Acequa Real del
Jucar (Royal Canal) constructed by the Spanish King inthe 13th Century during the Reconquest of Spain, the surfaceirrigation was carried out for generations by an operationalteam from the region led by the acequiero mayor. At theeld level, the regadores were responsible for waterdistribution from the tertiary and lower-level canals to thetypical Valencian huertas (plantations) with orange trees.They were recognizable by their leather footwear suitablefor entering elds with manta irrigation (ood irrigation).The modernization of the infrastructure and the introductionof a pressurized pipe system and ferti-irrigation (drip) ateld level implies that slowly, this operational hierarchy isbeing replaced in modernized sectors by more formallyeducated engineering staff. Some of the former celadoresand regadores were able to make that transition withadditional training (Gonbour et al., 2010).Preliminary studies in Zimbabwe and Sudan have also
suggested the broader relevance of these ndings for parts ofAfrica (Manzungu, 1999; Woldegebriel, 2011). An interestingnding was that the disappearance of canal guards and re-placement by water user associations in the Gezira schemeled to a chaotic situation in water distribution after transfer.This forced the security companies that weremade responsibleby the regime for operation of the system to employ severallocal ghafrs again (pers. comm. Yasir Mohamed, 31 May2010 and Alex Bolding, 6 June 2010; Woldegebriel, 2011).To assess the wider occurrence of the phenomenon under
study, we carried out electronic interviews with irrigationCopyright 2012 John Wiley & Sons, Ltd.experts who conrmed the relevance of the role of eld staffwho represent the frontline of irrigation organisations inmany parts of the world. Different cases and names werereported from various corners of the world: Colombia,Chile, India, Tanzania, Malawi, etc. Like the canalero, ditchtenders in other parts of the world are often involved inwater distribution, information exchange (communication),conict management, maintenance and fee collection(resource mobilisation), although a signicant difference inroles exists. This indicates the need for further research toestablish how widespread and specic the role of similareld staff is and to what historical, socioeconomic, agro-ecological, infrastructural and management conditions it istied to. Research, for example, could show in whichtechnological contexts eld staff play proactive roles. Thepresent Mexican case study indicates that the role of thisoccupational group is associated with exible and manuallyoperated irrigation devices of large and medium scale opencanal infrastructures (pers. comm. Herv Plusquellec,10 May 2010). This type of eld staff does not seem to existin many automated systems in the USA and Australia (pers.comm. Robert Chambers, 17 May 2010) and different partsof Europe, but we have also seen that modernization ofirrigation systems does not necessarily imply the disappear-ance of this staff. The fact that a similar function exists inxed or proportional systems such as the Warabandi systemin Pakistan and Northern India (pers. comm. Flip Wester, 10May 2010) indicates that we have to look more widely.A central but little studied issue is the gender component
in this type of work. The vast majority of cases that wereviewed involve males; however, exceptions do exist.Masculinity seems to play a major role in the professionalidentity and performance of these water operators (Rap,2004; Zwarteveen, 2008). Placed within a wider context, thisstudy suggests that a comparative and longitudinal researchproject is needed that poses the following research question:
What are the roles and responsibilities of eld-level personnel inirrigation water management under different agro-ecological,infrastructural and institutional conditions, how are they associatedwith the emergence of different (gendered) forms of power andauthority and how have they evolved over time and space?
The authors welcome suggestions and contributions inthis eld for such a research project. The point of thiselaborate empirical case study of the canalero is thereforenot to generalize these ndings to all irrigation systemsor to glorify the role of the canalero in the operation ofirrigation systems. It is rather to create an understanding ofwhy these actors perform such an important role in manyirrigation contexts, continue to do so over several decades,and to visualize and recognize their contribution to irrigationperformance given the lack of appreciation of their role inthe relevant literature.Irrig. and Drain. (2012)
We would like to thank the following persons for sharing
Johnson III SH. 1997. Irrigation management transfer: decentralizing public
ROLE OF CANAL OPERATORS IN IRRIGATION SCHEMESirrigation in Mexico. Water International 22(3): 159167.Latour B. 1987. Science in action, how to follow scientists and engineers
through society. Open University Press: Milton Keynes.Lipsky M. 1980. Street-level bureaucracy: dilemmas of the individual in
public service. Russell Sage: New Haven.Long N. 2001. Development sociology: actor perspectives. Routledge: London.Manzungu E. 1999. Strategies of smallholder irrigation management inZimbabwe. PhD Dissertation. Wageningen University: Wageningen.their insights: Maarten van Bentum, Bert Bruins, AlexBolding, Robert Chambers, Maurits Ertsen, Bruce Lankford,Ruth Meinzen-Dick, Yasir Mohamed, Eric Ofosu, HervPlusquellec, Diana Suhardiman, Shilp Verma, Gerrit vanVuuren and Flip Wester. We have beneted from thecritical and constructive comments and suggestions by twoanonymous reviewers, for which we are grateful.
Barley SR. 1990. Images of Imaging: Notes of doing longitudinal eldwork. Oranization Science 1(3): 220247.
Chambers R. 1988. Managing canal irrigation. Cambridge UniversityPress: Cambridge.
Clemmens AJ, Bautista E, Wahlin BT, Strand RJ. 2005. Simulation ofautomatic canal control systems. Journal of Irrigation and DrainageEngineering 131(4): 324335.
Clemmens AJ, Sloan G, Schuurmans J. 1994. Canal-control needs: exam-ple. Journal of Irrigation and Drainage Engineering 120(6): 10671085.
Comisin Nacional del Agua (CNA). 1994. Transferencia de los Distritosde Riego en Mxico. Comisin Nacional del Agua: Mexico D.F..
Finkler K. 1974. Estudio comparativo de la economa de dos comunidades deMxico; el papel de la irrigacin. InstitutoNacional Indigenista:MexicoD.F..
Finkler K. 1978. From sharecroppers to entrepreneurs: peasant householdproduction strategies under the ejido system of Mexico. EconomicDevelopment and Cultural Change 27: 103120.
Godaliyadda GGA, Renault D, Hemakumara HM, Makin IW. 1999.Strategies to improve manual operation of irrigation systems in Sri Lanka.Irrigation and Drainage Systems 13(1): 3355.
GonbourM, Habert L, Muhlig I, SinghM, Sonneveld E. 2010. Surface to dripirrigation the consequences for irrigation eld staff: changing watermanagement and organisation in the Acequia Real del Jcar irrigationsystem. Unpublished MSc report. Wageningen University: Wageningen.
Gorriz CM, Subramanian A, Simas J. 1995. Irrigation ManagementTransfer in Mexico. Process and Progress. World Bank Technical PaperNo. 292. The World Bank: Washington, DC.
Horst L. 1990. Interactions between technical infrastructure andmanagement. ODI/IIMI Irrigation Management Network Paper 90/3b.IIMI: London: ODI / Colombo.
Hunt E, Hunt RC. 1974. Irrigation, conict, and politics, a Mexican case. InDowning T, Gibson M (eds). Irrigations impact on society. Universityof Arizona Press: Tucson; 129157 pp.
Huppert W, Wolff B. 2002. Principal-agent problems in irrigation - invitingrent seeking and corruption. Quarterly Journal of InternationalAgriculture 41(12): 99118.Copyright 2012 John Wiley & Sons, Ltd.Boege E, Calco P, et al., Caciquismo y poder poltico en el Mxico rural.Siglo XXI Editores: Mxico D.F.; 148194 pp.
Murray-Rust DH, Snellen WB. 1993. Irrigation system performance:assessment and diagnosis. IIMI: Colombo / Wageningen: ILRI / Delft: IHE.
Plusquellec H, Burt C, Wolter H. 1994.Water control in irrigation systems.World Bank Technical Paper no 242. World Bank: Washington DC.
Rap E. 1993. Unpublished MSc thesis. Wageningen University: Wageningen.Rap E. 2004. The success of a policy model: Irrigation management transfer inMexico. PhD dissertation. Wageningen University: Wageningen.
Rap E. 2006. The success of a policy model: Irrigation management transferin Mexico. Journal of Development Studies 42(8): 13011324.
Rap E. 2007. Cultural performance, resource ows and passion in politics:A situational analysis of an election rally in Western Mexico. Journal ofLatin American Studies 39(3): 595625.
Rinaudo J-R. 2002. Corruption and allocation of water: the case of publicirrigation in Pakistan. Water Policy 4(5): 405422.
Schippers J. 2009. Under pressure: continuity and change in irrigationwater management in Western Mexico. Unpublished MSc thesis.Wageningen University: Wageningen.
Schuurmans J, Clemmens AJ, Dijkstra S, Hof A, Brouwer R. 1999.Modeling of irrigation and drainage canals for controller design. Journalof Irrigation and Drainage Engineering 125(6): 338344.
Sloan G. 1997. Evolution from manual canal control to remote canalcontrol. Proceedings of the International Workshop on Regulation ofIrrigation Canals: State of the Art of Research and Applications,April 1997. Marrakech, Morocco; 203206 pp.
Ter Hofstede K, Van Santbrink J. 1979. Neerlands Indi, koloniaal water-beheer. MSc thesis. Wageningen Agricultural University: Wageningen.
Uphoff N. 1986. Getting the process right: farmer organization and par-ticipation in irrigation water management. Unpublished paper. WaterManagement Synthesis II Project. Cornell University: Ithaca.
Van der Zaag P. 1992a. Chicanery at the canal; changing practice inirrigation management in Western Mexico. Latin America Studies 65.Cedla: Amsterdam.
Van der Zaag P. 1992b. Users, operators and hydraulic structures: a case ofirrigation management in Western Mexico. Irrigation ManagementNetwork Paper 19. ODI: London.
Van der Zaag P, Bolding A, Manzungu E. 2001. Water-networks andthe actor: the case of the Save River Catchment, Zimbabwe. In Hebinck P,Verschoor G (eds). Resonances and dissonances in development; actors,networks and cultural repertoires. Van Gorcum: Assen; 257279 pp.
Vos J. 2005. Understanding water delivery performance in a large-scaleirrigation system in Peru. Irrigation and Drainage 54: 6778.
Vos J, Vincent L. 2011. Volumetric water control in a large-scale opencanal irrigation system with many smallholders: the case of Chancay-Lambayeque in Peru. Agricultural Water Management 98: 705714.
Wade R. 1982. The system of administrative and political corruption:canal irrigation in South India. Journal of Development Studies 18(3):287328.
Woldegebriel E. 2011. Irrigation management transfer in the GeziraScheme, Sudan: A case study on farmers operation and maintenancestrategies in Tuweir minor canal. MSc thesis. Wageningen University:Wageningen.
Zwarteveen M. 2008. Men, masculinities and water powers in irrigation.Water Alternatives 1(1): 111130.Mares DR. 1980. Articulacin nacional-regional en el desarrollo rural: lairrigacin. Amrica Indgena 3(XL) 3: 471497.Martnez Vzquez VR. 1980. Despojo y manipulacin campesina: historia
y estructura de dos cacicazgos del Valle del Mezquital. In Bartra R,Irrig. and Drain. (2012)