ORGANIZED FARMERS: SPATE IRRIGATION
IN BALOCHISTAN
Frank van Steenbergen[1]
Index:
1. Farmers
and government in spate irrigation in Balochistan
2. Typology
3. Government involvement
4. Farmer management in spate irrigation
- Distribution of spate flows
- Managing sediment transport
- Organizing the maintenance
of diversion structures
- Reactive water rights
- Farmer organization
5. Participatory spate irrigation management
in Balochistan: options
- References
1. Farmers
and government in spate irrigation in Balochistan
This paper discusses the management
of spate irrigation in Balochistan (Pakistan). It discusses
the role of the government, particular in financing construction,
operation and maintenance of the spate systems; it discusses
the role of organized farmers in these functions; and pays
special attention to enabling laws, and their effectiveness,
for enforcing rights to spate water and for formalizing farmer
organizations. It concludes with a perspective on improved
participatory management of spate irrigation systems in Balochistan.
There are a number of reasons
to pay attention to Balochistan, as part of this meeting on
Yemen. The first set of reasons concerns the similarities between
the spate irrigation systems in both areas. Balochistan like
Yemen is an arid region with occasional rainfall events. In
Balochistan as in Yemen plains border highlands, giving rise
to spate systems that vary from small to very large, from very
episodic to almost semi-perennial. Like in Yemen, spate irrigation
in Balochistan provides the economic basis (though often in
combination with other sources of livelihood) for a substantial
- and often poor - part of the rural population. In Balochistan
a third of the irrigated area is under spate; of the remaining
area a substantial part is dependent on groundwater exploitation,
that is no longer sustainable. Again this resembles Yemen.
In these
similarities there are also differences, which make a comparison
worthwhile. There
are differences in the technicalities of spate water control
in Yemen and Balochistan, but a more interesting difference
concerns the role of government in the development and operation
of the spate systems. This has been fairly limited in Balochistan
and by default much of the management is done by farmers. The
role of government and the role of the organized farmers is
discussed in more detail in respectively section 3 and 4 of
this paper. On the basis of that discussion section 5 asks
the question of improved participatory irrigation management
of spate irrigation systems in Balochistan, which I hope is
thought provoking for Yemen as well. First however spate irrigation
in Balochistan has to be put in perspective: by giving a typology
and discussing the marginal nature of the farming systems.
2. Typology (Back
to the top of this page)
Balochistan
is the largest, but least populated (12 person per sqkm)
of four provinces in
Pakistan. Its uniform aridity (nowhere exceeding 400 mm on
average annually, but in many parts as low as 50 mm annually)
makes unirrigated agriculture impossible. According to the
Agricultural Census the area under run-off and spate irrigation
fluctuates between 30,000 to 150,000 ha. The areas under canal
irrigation and minor (including groundwater) irrigation each
are comparable and are 140,000 ha.
Spate irrigation (called sailaba in
Balochistan) is distinct from other rain-dependent production
systems, particularly water harvesting (called khuskaba in
Balochistan), in that the catchment of spate systems is bigger
and water is diverted from river beds, instead of being collected
from hill slopes or in-field. The dividing line
between the smaller spate diversion systems and some of the
larger run off complexes is often arbitrary, however. The farming
systems and water conservation practices are similar, but the
reliability in spate diversion systems is higher.
Spate diversion systems themselves
differ widely. The main parameters are hydrological (catchment
characteristics, rainfall pattern), geographical (location
and level), hydraulic (type of diversion and size of command
area) and sociological (land tenure, social structure and degree
of public intervention). A useful distinction at this point
is the difference between highland and lowland systems (see
also table 1).
Highland systems, found in the
Khurasan Range, on the eastern slopes of the Sulaiman Range,
and the Central Brahui Range in Balochistan, generally have
small catchments. Due to the small size of the catchments the
spates are of short duration, lasting from one hour to one
day usually. They are often difficult to control, because the
slopes are relatively steep in highland areas and because only
course material is available to build diversion structures.
Similarly, the distribution of water within the system tends
to follow the law of gravity more than elaborate distribution
rules.
The lowland
systems, on the other hand, that are common in the vast Kacchi
Plains as well as
in Las Bela and the Kharan basin in Balochistan, have large
catchments and very shallow gradients. The flow lasts several
days and can even be semi-perennial. Usually the flood water
moves slowly. It is often controlled by earthen barrages, but
deflecting spurs and free intakes - similar to those in the
highlands - are found as well. The low velocity flow and its
long duration have resulted in extensive rules on water allocations.
The flood rivers in the lowlands have often created their own
plains of fine soils. Due to the soft material and due to low
gradients, the sustainability of the systems is constantly
threatened by excessive silt deposition and subsequently breaches
of the river bank.
The distinction between highland
and lowland systems is not absolute. Lowland systems with small
catchments or in the upper reaches of flood rivers on the plains
share many of the characteristics of the highland systems.
On the other hand, highland systems in more temperate climatic
zones, where precipitation is gentle and spread over a longer
period, conform in some respects to the description of the
lowland systems.
TABLE.1: Typology
of spate irrigation systems
| |
Highland systems
|
Lowland systems
|
|
Catchment
|
Limited
|
Large
|
|
Bed material
|
Stony, coarse
|
Sandy, fine
|
|
Gradients
|
Steep
|
Gentle
|
|
Flow
|
Flash floods
|
Semi-perennial
|
|
Diversion structures
|
Free intakes
|
Also barrages
|
|
Water distribution
|
Simple, 'natural'
|
Complex, 'manipulated'
|
All spate diversion systems in
Balochistan, however, are marked by intrinsic uncertainty in
water supplies and related to this, a marginal agricultural
production system. The uncertainty in water supplies comes
in two shapes. The first is the recurrent uncertainty. Water
availability differs widely between the years: there are either
no floods or several floods. The floods may be too violent
to control and may wipe out diversion structures in one year;
and in the next years the floods may be mild and controllable.
The second element of uncertainty
is the dynamic character of the spate irrigation systems. In
the medium-term the configuration of the spate systems changes:
the bed levels of the spate rivers, the flood channels and
the fields changes and the intake structures need to be adjusted.
As a result of this dynamic nature some areas go out of command
because the flood channel silts up or because it scours out
so much that the flood can no longer be controlled and other
areas become easier to irrigate. In the worst case entire systems
are lost, because the river changes its course.
Spate irrigated areas in general
also support a low value agriculture. The recurrent uncertainty
in water supplies lies at the root. There may be either too
much or too little spate flows. In the first case the spates
may be beyond control, breaking the diversion structure or
the flood channels, before land is irrigated. In the alternative
scenario, the season may not bring any flood or only a very
small flood, that peters out before it irrigates all fields.
A variation on these problems is that the downstream water
users are deprived, because upstream users monopolize the flow.
A further source of insecurity is the additional moisture from
rains at later stages of crop growth, particularly of wheat.
These rains may not come and the crop may be suitable for fodder
only.
The farming systems are dominated
by drought resistant, low yielding sorghum, millet, wheat,
pulses, cotton and oilseeds. Most of the land is under local
cultivars. Even if optimal conditions were to prevail, crop
returns would have difficulty competing with alternate sources
of income.
In addition to the above there
is the sensitivity to crop diseases. In general, in flood irrigated
areas, especially adapted local varieties are grown. The timing
of the flood determines the crop choice and there is little
room to maneuver. This result in monocultures and the impact
of these pests can be dramatic. Yet there are very few substitutes
for the varieties used and agricultural research has concentrated
on perennially irrigated crops. Vulnerability is further increased
as flood irrigation farmers generally lack the financial resources
to apply pesticides.
As a consequence, crop yields
from spate agriculture in Balochistan, as in Yemen, are low.
They typically range between 450-900 kg/hectare for wheat,
360-550 kg/hectare for sorghum, 200-500 kg/hectare for pulses,
360-620 kg/hectare for cotton and 150-350 kg/hectare for oilseeds
(Mu'Allem 1987; IAN MACDONALDS 1987a; 1987b; Shah 1989). A
related phenomenon, peculiar to several flood irrigation areas
is that there are land owners who are 'too poor to farm'. They
lack the draught animals or mechanical traction to prepare
the land and repair the field bunds, and do not have the cash
to buy seeds and may have lost their credibility with money-lenders.
They are often forced to rent out their land to tenants, who
have access to these means of production.
Outmigration is the next step.
With alternative labor markets developing and standards of
living rising, the movement to other areas has accelerated.
This trend has sometimes been self-reinforcing, in particular
in the lowland systems of Kacchi, Las Bela and Kharan, as at
a certain point the number of able-bodied farmers that stayed
behind was insufficient to rebuild the labor-intensive diversion
works. In the absence of bunds on the river, the riverbed regresses,
making the diversion of water even more difficult, causing
further depopulation and ultimately the abandonment of the
spate system, as the critical mass for maintenance is lost.
This process of depopulation can go very fast: when in the
Mirpur area water supplies were reduced because farmers upstream
refused to break their bunds, the area depopulated in a matter
of three years and most of the spate infrastructure became
dysfunctional. In some areas the danger of depopulation is
countered by tying labor to the land. The most important manifestation
is the hereditary tenancy, that is widespread in the Brahui
Mountains. In these areas, tenants were given hereditary land
rights, on the condition that they would perpetually maintain
the field bunds.
3. Government
involvement (Back to the top of this page)
Investment in irrigation in Pakistan
has been dominated by perennial irrigation. Spate irrigation
received relatively less attention, because of the general
lower rates of returns and the difficulty of making it work
technically. The motivation to invest in spate irrigation was
sometimes secondary to the spate systems themselves: the public
investments were justified on the basis of groundwater recharge
or flood protection. Even so 74 permanent structures were constructed
in the past decades in Balochistan, making up one-third of
the portfolio of the Irrigation and Power Department in the
Province. Most of these concerned either earthen, brick or
concrete headworks. The failure rate of spate structures, however,
has been high. An extensive evaluation of 47 schemes, constructed
by the government of Balochistan in the last thirty years,
found that only 34% still functioned satisfactorily (see table
2). The other systems have either become unusable or suffer
from serious operational problems (Groundwater Consult 1991).
TABLE 2 Performance
of government-constructed spate schemes in Balochistan
|
Date of construction
|
Functional
|
Serious operational problem
|
Out of order
|
|
Prior to 1973
|
7
(35%)
|
6
(30%)
|
7
(35%)
|
|
1974 to 1983
|
4
(29%)
|
2
(14%)
|
8
(57%)
|
|
After 1984
|
5
(38%)
|
7
(54%)
|
1
( 8%)
|
|
Total
|
16
(34%)
|
15
(32%)
|
16
(34%)
|
Source:
Groundwater Consult 1991
The overriding
factor behind the high proportion of failures in Balochistan
was the inappropriateness
of the prevailing engineering concept, which was based on controlling
the flow at a single point with heavy civil engineering works
rather than managing the inherently varying flood rivers. The
technical designs for spate systems resembled those for perennial
flows, and did not accommodate the capricious nature of the
spate systems. Some structures were not able to withstand the
force of the violent peak floods. In other cases, headworks
were by-passed by the braiding river. Moreover, the provisions
for sediment transport were generally insufficient and the
intakes silted up (Morton and van Hoeflaken 1994). Trying to
avoid these pitfalls would have required substantial investments
in large headworks, complex silt excluding devices and long
marginal bunds. Though, with these investments, it would have
been possible to control the rivers at a single point, the
low returns to sailaba cultivation ruled against such
high investments.
Besides, in a number of cases
newly constructed spate irrigation systems were not utilized
due to conflicts on the spate water rights[2].
To invest in a single permanent off-take compatible with the
existing 'reactive' rights (further discussed in section 4)
and allocation rules on a river system, is extremely difficult,
since existing practices do not create predictable entitlements,
but at best certain probabilities of irrigation. By improving
the water supply to one area, there is a fair chance that someone
will be worse off somewhere alse at least under some circumstances.
Conflicts are hard to avoid, unless one concentrates on the
tail-end of the system or on small independent systems. However,
even so the changes introduced by the new structures would
have had to be preceeded by intense consultation with the land
owners in different command areas. When the spate systems in
Balochistan were constructed, however no consulatation took
place.
In Balochistan the role of the
Irrigation and Power Department stopped very much after construction.
After care was primarily limited to the posting of linemen
and guards. There was no routine program for repair works,
but repairs were done on an ad hoc basis. In the last five
years, however, the already inadequate budgets for maintenance
by the Irrigation and Power Department have further curtailed.
Precise figures on maintenance requirements and actual expenditures
on spate systems are not available, but the overall estimated
shortfall for the operation and maintenance of irrigation and
drainage structures ranges between 25-75%.
Against these budgets and budget
requirements, there is no dedicated revenue. Water tax, collected
in perennial canal systems by the Revenue Department, is not
levied on spate systems in Balochistan. Farmers in spate irrigated
areas are bound to pay land revenue and other smaller land
based taxes, though. The returns are modest (in the order of
US 1.5/ acre) due to the low per area rates, underassessment
and decreased collection efficiency.
It is fair
to say that the overall lasting impact of the public investment
in new spate irrigation
structures in Balochistan has not been very large. A government
program, that had a far larger impact however, was the bulldozer
subsidy program. Under these subsidy program bulldozers and
frontloaders, that were provided under several aid programs,
were rented out by the Agricultural Engineering Department
for Rs 100-150 per hour (US $ 5-7.5). The allocation of these
quota was done by elected politicians; the political mileage
that was made is testimony to the popularity of the bulldozer
subsidies. The impact of the bulldozer subsidies on the rehabilitation
of the flood irrigation structures has been significant and
the bulldozers were used to rehabilitate and reinforce bunds
and flood channels as well as construct new ones. The bulldozer
subsidies also had an important impact on the social organization
of the flood water users: they removed the need to pool resources
to maintain the structures. The bulldozer subsidies program
however is now in decline. Several of the bulldozer are now
out of order and under the subsidies no provisions were made
for their replacement.
In the absence
of a pervasive role of the government, the spate irrigation
systems in Balochistan
are by and large farmer-managed. Some of these systems in fact
represent some of the world's largest farmer-management systems.
The most spectacular system is the series of earthen bunds
on the Nari River in the Kacchi Plains, irrigating over 50,000
ha. Management of the spate systems requires arrangements for
the internal water distributions, the management of the sediment
transport, the organization for maintenance, reinforced by
water rights. These various management functions are discussed
in the subsequent section.
4. Farmer
management in spate irrigation (Back
to the top of this page)
The three main function in the
management of spate irrigation systems are:
¨ the
distribution of spate water
¨ the
management of silt and scour processes in the flood channels
¨ the
maintenance and rehabilitation of diversions structures.
This section describes how farmers
manage these functions as well as the type of water rights
and the type of farmer organizations in place.
Distribution of spate flows (Back
to the top of this page)
The risks
in flood irrigated agriculture are high, but the probability
of being irrigated or not is
not equally distributed throughout the commnad area of the
flood irrigation systems. Within the area served by one flood
river and within the command area of one off-take, there may
be land with high, medium and low probability of irrigation.
This probability depends on the location and level of the command
area along the flood river. Often upstream land and low lying
land takes precedence and in Balochistan one will typically
find a flood channel with a series of small and big obstructions,
forcing the water to irrigate sub command areas in stages from
head to tail. Much depends, however, on the ability to control
the floods: where the flood channel is deep and its bed is
rocky and steep, management and control of flood water is
far more difficult. Obstructions will not stand up in high
floods and a 'free distribution' will be in place, where water
finds its own way through the command area, with low-lying
areas having the largest probability of irrigation4.
Yet if floods are moderate, a sequence of water deliveries
from head to tail can be practized.
The point
to be made is that unlike perennial systems the distribution
of the flood water is not
entirely in control of farmers. Nevertheless in most spate
systems in Balochistan there are water allocation rules in
place that modify the probabilities of irrigation. Between
off-takes, there may be rules on when to close an intake or
break a barrage. Within the systems, an important allocation
rule is the right to a second irrigation. It is either at the
discretion of an upstream land owner to divert water for a
second time, or alternatively he can only avail of it, after
all tail-enders have been served. The latter rule clearly spreads
the probability more evenly[3]. The spectrum of water allocation rules is given in box 1. These
allocation rules reduce - to some extent- the antagonism and
mitigate the inequity. This is important, as farmers are dependent
on one another in the maintenance of the works and too much
unfairness would be dysfunctional.
Box 1: Water allocation rules
in spate irrigation
¨ the construction
of diversion structures and obligations to break them;
¨ the rights
of some land tracts against others to be irrigated, when
spate flows are small;
¨ the sequence
in which lands are irrigated and the proportions to go
in the different flood
channels;
¨ normative
rules on water usage, like the entitlement to a second
turn and the depth of water
to be applied; and
¨ agreements
on the disposal of the high and unusable floods.
Managing sediment transport (Back
to the top of this page)
Scour and siltation are part and
parcel of spate irrigation systems in Balochistan. Rivers in
spate lift and deposit huge quantities of sediment. As a result
there is constant change in bed levels, resulting in changes
in bed levels and water distribution. The impact of these processes
differs between the various systems. It depends on the amount
and composition of the sediment load that a river carries,
which depends on the rainfall pattern and the characteristics
of the catchment area; its geology, morphology and vegetation
cover. Farmers are usually able to identify the origin of a
flood by the type of sediment that is transported by it. The
degree of siltation and scour also depends on the local topography
and the type of material. In spate irrigated areas with low
gradients, as are found on the plains, a river is always in
danger of choking itself with its own silt deposits and finding
another way. Moreover, in the fine sandy deposits of the plains,
the scouring of the riverbed is a larger danger than it is
in the rocky and pebbly river beds of the highlands of the
Balochistan. As a result, the lowland flood irrigation systems
are particularly dynamic.
Farmers,
however, are not passive actors in these scour and siltation
processes. They actively
manipulate land formation. They may deepen the headreach of
a flood channel, in order to attract a larger flood that will
further scour out the channel. If a flood river breaks its
banks, farmers may close the breaches, if it deflects water
away from their land or on other occasions, they will leave
the breaches intact, so that these will act as escapes, creaming
off the peaks of the very high floods and maintaining the flow
at their own system at a manageable level. In other cases farmers
will manipulate the siltation process to force the river bed
to purposely silt up. The latter is in practice where the river
has become uncontrollable, because its bed may has become to
deep or to steep. The remedy is to built a strong permanent
bund across the river and force the river to deposit its sediment
load upstream of the bund[4].
In Balochistan
these land formation processes are managed by farmers. This
is unlike the spate
irrigation systems on the eastern slope of the Suleman Range
(DI Khan and DG Khan) in NWFP and Punjab Province, where the
civil administration actively intervenes in instructing farmers
to plug breaches and to connect flood channels. During the
colonial times the rights to the spate flows were registered
and since then disputes are resolved through a special functionary.
In Balochistan there is only limited involvement by the government in
supervising the timely breaking of the bunds in the main Nara
system as well as in the major flood channels. This is done
through the office of the so-called teshildar gandajat, a left-over
from the period when the Kacchi Plains were ruled by the Khan
of Khalat. This supervision has lost much of its force now
and is now often limited to conflict resolution.
Organizing the maintenance
of diversion structures (Back
to the top of this page)
Maintenance arrangements have
to cope with the inequities between areas with low and hogh
probability of irrigation. Moreover, in spate systems returns
for some farmers (in areas with a verdy low probability) are
so low, that their contribution to the maintenance of the systems
is hardly matched by the marginal benefit they derive from
it. This is unlike perennial irrigation systems, where the
value of water is almost always far in excess of the maintenance
efforts;
Notwithstanding
these difficulties, in
many spate irrigation systems maintenance and rehabilitation
is vital: without it there is simply no irrigation in a subsequent
year. This is unlike perennial system, where often more neglect
is tolerable. This places a high demand on the organization
of farmers, complicated by the unavoidable inequity in water
supplies.
In response
to these challenges in almost all spate systems clear rules
have developed on the
contribution of individual farmers. There are broadly two types
of systems in this regard. The first category of systems are
the 'regulated' systems. In a regulated system the area that
is entitled to spate irrigation is clearly demarcated. The
question is how do farmers within these areas contribute, given
the fact that some areas stand a far better chance of being
irrigated than others. There are two extremes:
¨ only
farmers that had their land irrigated contribute to the maintenance
of the system.
¨ all
farmers contribute in proportion to the land that is entitled
to spate
flows, irrespective of the probability of the spate flow;
The first option may seem fair,
but does not work in practice and also does not occur in Balochistan.
It does not work, because it provides too narrow and unpredictable
of basis for the maintenance work. Who remembers which land
was irrigated the previous year? What if the diversion structure
is destroyed before anyone watered his fields?
The second
extreme on the other hand is straightforward: every one contributes
to the land
in the designated command area. However, it puts a disproportional
burden on farmers in areas with a low probability. One often
finds therefore in systems, where maintenance is organized
on this principle, that water distribution rules are in place,
that favor the low probability areas (such as no second irrigation
and the following of strict sequence of water turns). Or,
particularly in small systems, land tenure may be such that
farmers have land both of high and low probability. A different
route is that of dispensations in the labor or cost contributions
for deprived farmers. One such dispensation is that persons,
who have bullocks or tractors are expected to take these along
during the maintenance works, whereas other farmers only provide
labor. Bullock or tractor owning farmers are ofthe those, that
have a disproportianal part of their land under cultivation
and in this way fairness is restored. In larger systems maintenance
is often done by labor contributions, but financed out of a
special water tax (gham). Often different rates are assessed
on different parts of the command area. For areas with a low
probability a lower tax is assessed. However, such compensation
rules are not in place everywhere.
The opposite
to the regulated systems, described above, are the voluntary
systems. In these
systems the area that is irrigated from spate flows is not
predetermined. The rules are essentially based on a 'take it
or leave it'. If one contributes to the repair of the structures,
one is entitled to the flow. If one does not contribute, the
field intake is closed. These voluntary systems can be very
unfair on low probability land owners, as there is no dispensation.
In some cases in fact, the tail enders contribute more per
acre. Where deflector spurs are rebuilt a number of times,
one may find tail enders contributing to several rehabilitations
before they get the water.
Reactive water rights (Back
to the top of this page)
There is
a categorical difference in the water rights in spate diversion
systems and perennial
systems (Varisco 1983). Whereas in perennial systems in Balochistan,
individual rights are often sharply defined in fixed proportions
of the flow and the allowed usage time, water rights in flood
systems are 'reactive'. They cope not only with the unknown
proportions of the next flood9,
but also with the medium-term changes in the river morphology,
due to scour, siltation and change of river course. Water distributions
in the flood water irrigation systems are based on allocation
rules rather than alienable property. They describe was is 'done' and
what is 'not done'. The spectrum of possible rules is given in
box 1.
A second
domain regulated by reactive water rights concerns the changes
in the lay-out of
the systems, as a result of natural or man-manipulated siltation
or scour. Water rights may accept the changed circumstances
or may try to restore prior access to the water. This also
depends on the nature of the change.
In most - especially smaller -
systems in Balochistan, water rights are enforced indigenously,
by the farmers of the flood irrigation systems. Where conflicts
occur, outside political or administrative support may be mobilized,
but essentially the distribution of water and repair of irrigation
infrastructure is not a government responsibility. Its involvement
is limited to extend help on an incremental basis, at varying
degrees of intensity, keeping in line with political principles
of patronage, and divide and rule.
In fact, the only time when a
government was directly involved in the management of spate
irrigation systems in Balochistan was during the latter period
of Kalat State, when the native ruler appointed the so-called teshildar
ghandajat, who was to supervise the breaking of the different
barrages in the Nari River at a specified time. As discussed,
the background of this practice was that otherwise the ruler's
own land at the tail of the system might not be inundated.
Yet this regime of regulated flood management was exogenous
to the resource, and disintegrated when the political constellation
changed. After Kalat State joined the newly formed republic
of Pakistan in 1948, the former administration of the water
rights was dismantled, and the upstream landowners refused
to break their barrages after the previously specified time.
In the ensuing vacuum, water rights were often determined by
the relative strengths of the communities along the flood river
rather than by formal regulation. In contrast, in the flood
irrigated areas of Punjab Province of Pakistan, government
involvement has been more persistent. In the colonial period
the British administration, in order to safeguard the land
revenue from the flood irrigated areas of Punjab, directly
managed the use of flood water, and organized the repair of
bunds and flood channels by corvee labor and posted watchmen,
who were to break the barrages in time (Bolton 1908). The system
has slackened somewhat, but still the District Collector supervises
the water distribution and timely breaking of earthen bunds
in this area (PARC/UNEP/ESCAP 1994).
Farmer organization (Back
to the top of this page)
Who organizes the maintenance
work and who supervises the water distribution in the spate
system in Balochistan? In general, given the magnitude of the
task, the work is organized in a surprisingly informal way.
A number of influential farmers take the initiative. These
are often farmers with large holdings, who stand to gain most
from the reconstruction of the structures and also have most
resources to mobilize fellow-farmers. The maintenance is done
in a campaign mode, with a series of days planned for joint
labor. In some of the large spate systems on the Kacchi Plains,
a water tax (gham) is still collected through a network of
local leaders. Alternatively the influential farmers try to
arrange the subsidized bulldozers to their areas. In fact the
arrival of the subsidized bulldozers has removed the need for
structured farmer organization. In summary there are organized
farmers and clear rules, as described above, but no farmer
organization as such.
In some of
the larger lowland systems specialist staff is deployed.
They are of two types.
In the Kacchi Plains designated 'engineers' supervise the construction
of the large earthen bunds. During the flood season they are
engaged to patrol the safety of the bund. They are paid either
in kind or from the gham. This office of 'engineer', however, is
not common in all the spate systems, not even in the Kacchi
Plains.
A second
type of dedicated staff are the 'sepoys' (literally: policemen) engaged in a limited
number of spate systems in Las Bela. Their role is to mobilize,
if necessary by force, farmers to contribute tot the reconstruction
of the diversion structures. One finds the sepoys particularly
in systems based on deflector spurs, where the burden on tail-end
farmers (and hence the reluctance to contribute) relatively
large. The root of this position is the pre-partition time,
when the native ruler organized the construction of the diversion
structures with forced labor. His interest at that time was
the land revenue that he received from a well-functioning spate
system. After the dissolution of the princely state, farmers
valued the role of the sepoys and continued their engagement.
Again, however, the position of sepoy is far from uniform in
spate systems.
Remarkably in none of the spate
system dedicated persons are engaged to supervise the distribution
of water, unlike the perennial systems in Balochistan, where
often water bailiffs (called mir-i-aab, rais or sharistra)
supervised the irrigation turns of the individual farmers[5].
No such position is in place in spate systems and instead social
control by individual farmers regulates the water distribution.
The reason is probably that it is difficult to do it otherwise:
when the spate comes, so many areas are irrigated in a very
short time that it is impossible for a specialist water bailiff
to patrol.
5. Participatory
spate irrigation management in Balochistan: options (Back
to the top of this page)
As most of
the spate systems in Balochistan are de facto farmer-managed,
the scope for irrigation
management transfer in Balochistan is limited to probably 25
functional structures. The question however is whether in all
these 25 cases the 'pucca' structures are what farmers would
rebuild, if flood damage would require so. It is more likely
that farmers would prefer a menu of earth moving interventions
to restore their diversion structures: the operation of heavy
tractors and bulldozers is technically less demanding and can
be undertaken at lower cost. The conclusion could be that several
of the spate works will last as long they will last, but once
gone will not be restored in the same shape under farmer management.
Leaving aside
the management of the spate systems, constructed earlier
with public funds, there
is however a large scope for improved participatory management
of the spate systems of Balochistan. The improvements concern
the type of organizations and the type of spate water management.
Most of the systems are managed in a rather passive way, with
campaigns to rally contribution for O&M (if already bulldozers
have not been secured in sufficient quantity) and rules on
what to do and what not do in water distribution, but no local
planning.
Strengthening of farmer organizations
and assigning them planning and active management functions
should be done in the light of a river management approach.
The flood management system in D.I. Khan District of Punjab
in Pakistan may serve as an example (Bolton 1908). Under this
approach rather than controlling the river, one tries to manage
it in its entirety. This requires a more comprehensive look
at the river system and the land formation processes. Examples
of interventions are; controlling breaches in the river banks,
by making bunds in the newly formed spillways, if necessary
in stages; forcing a degrading river to silt up by blocking
it with a permanent bund, even if it means that the river may
eventually find a new route away from the bund; excavating
new flood channels to link up braided streams. These interventions
mainly involve earthmoving and resemble existing farmers techniques
of manipulating siltation and scour. This 'management' approach
is less capital-intensive and has a far higher chance of creating
sustainable flood irrigation systems and as such a superior
alternative to past approach of building heavy headworks at
single points.
Wherever
possible, farmers should continue to be the prime implementers
under the river management
approach. When subsidized government bulldozers continue to
be made available in Balochistan, a river-level management
approach allows one to use them strategically, instead of letting
bulldozer allocations be solely determined by political opportunism
and individual demands. Two large spate irrigation systems
were completed in the Kacchi Plains (with the help of NGO's)
on this formula, from which valuable lessons can be learned.
The first dam is the Rehanzai Bund using water form the Bolan
River and an off-shoot of the Nari River. This two kilometer
wide earthen dam was made by farmers organized in village organizations
and has been very successful in rehabilitating area, that had
not be cultivated because the river has braided. Part of the
success was at the detriment of down stream water users, who
despite formal assurance were deprived of water, as farmers
on one of the flood channel of Rehanzai refused to break their
bunds. This underlines the need of a river basin approach.
The second example is the Sonwa dam, built at the tail of the
Nari, planned and implemented by the farmers on the three main
flood channels benefiting from the three kilometer wide bund.
Unfortunately the bund broke after two very productive years
and was not rebuilt by the farmers, who instead are using part
of it. This example also shows that one should not expect miracles
from farmer organizations.
While the techniques are not new,
the most important crucial element in a river basin management
approach is the appropriate organizational framework. Since
the interventions involve changes that affect several intakes
simultaneously, an institutional structure is required that
supersedes the interests of the land owners on a single flood
channels only and avoids decisions that are taken by a show
of strength only. The challenge is to have a supra-local organization
with a permanent character, since - although it may be intermittent
- management never stops and the build-up of knowledge on the
river's behavior is essential.
The new Balochistan Irrigation
and Drainage Act, accepted in 1997 has the provisions for this.
The Act was prepared as part of the institutional reform in
the water sector, that is meant to create financial autonomy.
One important step in this regard is the transformation of
the Irrigation and Power Department into an autonomous Irrigation
and Drainage Authority with responsibilities for surface and
groundwater management. The Authority can under the Act delegate
its responsibilities in water management and revenue collection
to Farmer Organizations. These Farmer Organizations are formed
on the basis of hydraulic units and can federate. At present
the hydraulic units specified are distributaries and minors
of large canal systems. This should be corrected. More work
needs to be done on the Act: the legal status of the farmer
organizations and the procedures for registering them are not
yet worked out. The first step however is to put the Act in
its current shape in practice, which is not yet done.
A second
legal reform that is overdue is the codification of spate
water rights. Water rights
are only recorded on two flood rivers, i.e. the Porali and
the Nari. The competition for water will only increase and
with it the risk of depriving downstream water users. The codification
on the Porali has led to a conflict-free distribution of water
on this river. In case of the Nari, the rules on the breaking
of the bunds is not always enforced. Even so, without these
rules the chaos would have far larger.
References: (Back
to the top of this page)
Bolton, H.N. (1908). A report
on irrigation in Dera Ismail Khan. Reprinted in: BARD Rod-kohi
agricultural problems and prospects symposium, November 27-29
1989. Islamabad: Pakistan Agricultural Research Council,
16-37.
Brunner, U. and Haefner, H. (1986).
The successful floodwater farming system of the Sabeans, Yemen
Arab Republic. In: Applied Geography, 6, pp. 77-86.
Buzdar, N,, Nagy, J.G., Farid
Shabir, G., Keatinge, J.D.H. and Khalid Mahmood (1989). Rainfed
agriculture in highland Balochistan: a farming systems perspective.
Research Report 54. Quetta: Arid Zone Research Institute.
Camacho, R.F. (1987) Traditional
spate irrigation and wadi development schemes. In: FAO/UNDP Spate
irrigation: proceedings of the subregional expert consultation
on wadi development for agriculture in natural Yemen, 6-10
December 1987.
Rome: FAO, 60-72.
Girgirah, A.A., Maktari, M.S.,
Sattar, H.A., Mohammed, M.F., Abbas, H.H. and H.M. Shouhibi
(1987). Wadi development for agriculture in PDR Yemen. In:
FAO/UNDP Spate irrigation: proceedings of the subregional
expert consultation on wadi development for agriculture in
natural Yemen, 6-10 December 1987.
Rome: FAO, 9-26.
GROUNDWATER CONSULT (1991). Balochistan
flood water irrigation systems. Islamabad: Royal Netherlands
Embassy.
Khan, Aminullah (1979). Hill-torrent
(rod kohi) control scheme for the Daman tract of D.I. Khan
District. In: National seminar on land and water resources
development of barani areas. Lahore: CEWRE, 80-87.
Khan, Amir Nawaz (1987). Spate
irrigation in Pakistan. In: FAO/UNDP Spate irrigation: proceedings
of the subregional expert consultation on wadi development
for agriculture in natural Yemen, 6-10 December 1987. Rome:
FAO, 167-170.
Morton, J, and van Hoeflaken,
H. (1994). Some findings of a survey of flood irrigation
schemes in Balochistan, Pakistan. IMN Network Paper 31.
London: ODI, 1-9.
IAN MACDONALDS AND ASSOCIATES
(1987). Diagnostic study of farming systems in Baluchistan:
Kacchi. Quetta: BAERP.
IAN MACDONALDS AND ASSOCIATES
(1987). Diagnostic study of farming systems in Baluchistan:
Las Bela. Quetta: BAERP.
Makin, M.J. (ed) (1977) Yemen
Arab Republic: Montane Plain and Wadi Rima Project: a land
and water resources survey. London: Ministery of Overseas
Development.
Maktari, A. (1971). Water rights
and irrigation practice in Lahj. Cambridge: Cambridge
University Press.
Mu'Allem, S.A. (1987). Crop production
under spate irrigation in the coastal area of PDRY. In: FAO/UNDP Spate
irrigation: proceedings of the subregional expert consultation
on wadi development for agriculture in natural Yemen, 6-10
December 1987. Rome: FAO, 167-170.
Mundy, M. (1985). Agricultural
development in the Yemeni Tihama: the past ten years. In: B.R.
Pridham Economy, society and culture in contemporary Yemen.
Exeter: Centre for Arab Gulf Studies, University of Exeter,
22-40.
Munir, M. and Muhammed Akram Kahlown
(1988). Sailaba irrigation in Baluchistan: bench mark study.
Lahore: WAPDA Planning Directorate (South).
NESPAK (1993) Feasibility studies
for development of rod kohi schemes in DI Khan. Lahore: NESPAK.
Serjeant, R.B. (1964). Some irrigation
systems in Hadramawt. In: Bulletin of the School of Oriental
and Africa Studies, 27, 33-76.
Shah, S.H. (1989). Cropping patterns
in rod-kohi agriculture in D.I. Khan. In: BARD Rod-kohi
agricultural problems and prospects symposium, November 27-29
1989. Islamabad: Pakistan Agricultural Research Council,
115-121.
Varisco, D.M. (1983). Sayl and
ghayl: the ecology of water allocation in Yemen. Human Ecology,
11, 4, 365-383.
Wildenhahn, E. (1985). Traditional
irrigation systems in the Southwest of Saudi Arabia. In: J.F.
Mock. Traditional irrigation systems and potential for their
improvement. Hamburg/Berlin: Verlag Paul Parey:119-134.
