Economic Aspects of the Ancient Irrigation
System of Sri Lanka
Chandana Jayawardana
Abstract
The ancient irrigation system that prevailed in Sri Lanka has been subjected to deep technical
investigations. But such investigations in economic terms have not been evaluated, basically due
to the lack of information. This paper, tries to develop a plausible methodology to generate revenue
from irrigation works, as well as to investigate the potential of such methodology to act as a
regulating measure of the demand for for water. Chronicles and inscriptional evidences were
evaluated to establish the ownership of the irrigation works and types of payments and revenues
associated with each party.
In previous studies, types of revenues were identified as dakapathi, payment to the state in the
form of a tax and the bojakapathi, payment to the supplier of water. The application of this
payment scheme to a reservoir cascade system determined the rived overall revenue equation.
This equation was again analyzed applying the technical restraints and it was proved that water
rates could have been used for the sustainable operation of the cascade system.
1. Introduction
Sri Lanka is well known for its ancient water civilization. The earliest reservoir referred to in the
Mahavamsa, the great chronicle of the history of the island, was the Jayavapi built in the
reign of king Pandukabhaya (377-307BC). The construction of larger scale reservoirs initiated
at the time of King Vasabha (65-109AD) and by 500AD, very advanced hydraulic structures
which would have required sound knowledge of key hydraulic principles pertaining to rainfall,
runoff and storage volumes had been designed and constructed. That trend continued until
the reign of King Parakramabahu I (1153-1186AD). This long development phase had
established a widely spread mosaic system of reservoirs and canals [Fig. 1] regulating the
water flow either extracted from natural streams or received from direct rainfall for agricultural
Chandana Jayawardana Currently works as Project Manager, Cruickshanks Ceylon Pvt. Ltd. B.Sc. (Eng)
Hons, MTech, MIET, MIE (SL), C Eng.
- 128 -
Economic Aspects of the Ancient
Irrigation System of Sri Lanka
and human utility purposes, as well as a means of flood management and soil conservation.
Mendis (2002) referring to the functions of the ancient irrigation works notes, in addition to
the much known hydraulic engineering supremacy, a well established water and soil
conservation ecosystem perspective. Although several studies of the technical aspects of the
ancient irrigation works have been made, economic aspects are not adequately evaluated.
This paper analyzes the potential of the ancient irrigation system as a means of wealth
creation and also analyzes the taxation methodology coupled with ancient water use.
- 129 -
Sri Lankan Journal of Management
Volume 15, Nos. 2 , 3 & 4
2. Ownership of the irrigation works
According to chronicles and inscriptional evidence, the ownership of ancient irrigation works
was vested in either the state, Buddhist monastic establishments, village institutions or
individuals.
All minor and major canals and tanks built by the King were the property of the state, and their
income in the form of taxes on water and fish yield was absorbed by the royal treasury. Most
probably, large irrigation works might have been owned by the state considering the scale of
labour and the materials used for such work. Ven. Walpola Rahula (1993) notes that the
monasteries owned property from about the 2nd century B.C, through royal and private grants
of land and irrigation works. The irrigation works so donated were small tanks but there is at
least one recorded instance of a donation of a major canal, namely the Elahara, to a monastery
some time after its construction. One of the inscriptions at Viharagala refers to the donation of
Upaladonika reservoir by King Vasabha (67-111) to the Ekadorika monastery. Two of the earliest
inscriptions, those at Thonigala and Eriyawewa, refer to grants of reservoirs to monasteries by
parumakas, the provincial rulers. Perera (1949) noted that from the time of King Lajjithissa
(119-109BC) onwards there were many references of grants both of reservoirs and canals to
the Sangha by individuals.
Since the village tank is an important part of the dry zone settlements, tanks constructed with
the communal labour of villagers were owned by village institutions. Samanthapasadika
refers to instances of the construction of reservoirs by villagers and to communally-owned
village reservoirs.
Likewise, individuals also owned small reservoirs and canals. Siriweera (2004) identifies the
term vapi-hamika found in the Brahmi inscriptions from the third century B.C. onwards as
denoting a ‘tank-owner’. The Thimbirivava inscription of the third century A.D. refers to a
reservoir owned by a family, tumaha kula sataka. As mentioned earlier, the grants of reservoirs
and canals to the Sanga by individuals because those works were owned by those individuals.
The Kahambiliyavava inscription of Vikramabahu I (1111-1132) indicates that irrigation works
owned by individuals continued to exist until the end of the period of the dry zone civilization.
From the above description, it is evident that irrigation works were owned by several institutions,
depending on the use of funds for those enterprises. This ownership gave the right to earn
revenue from the reservoirs as well as from canals, basically through supply of water and
fishery.
3. Revenue generation from irrigation works
Considering the massive material and labour resource inputs required for the construction as
well as maintenance of irrigation works, there should be adequate income generated from
those enterprises. Codrington (1937) notes that the responsibility of the state for infrastructure
development required some form of state taxation, in order to fund those endeavours.
Evidence of such taxation has also been noted by scholars who studied ancient inscriptions.
Two terms have been exclusively used in inscriptions to denote the forms of revenue
- 130 -
Economic Aspects of the Ancient
Irrigation System of Sri Lanka
generated from irrigation work, namely, dakapati and bojakapati. They have been used
singularly or collectively depending on the context in which they were used. Slab inscriptions,
such as Angunukola-vihara inscription, Molahitiyavelegala inscription and Galvewa inscription
mention the term dakapati in the singular.
There were several slab inscriptions highlighting the application of bojakapathi such as
Galkovila inscription of Bhathika Thissa (140-164AD) and Ruvanveli dagaba inscription of
Gajabahu (112-134AD). Some slab inscriptions mention both dakapati and bojakapati such
as the Nagirikanda rock inscription.
Some scholarly work on the derivation and the interpretation of these terms dakapathi and
bojakapathi has been done by Paranavitana, Codrington (1937) and Perera (1949). The
conclusion is that dakapati meant payments to the owner of the water resource while
bojakapati meant payments due to the state.
There was also a well organized system of distribution of water to ensure an equal share for all
cultivators during periods of scarcity. Codrington (1937) identifies three types of paddy fields
categorized for the purpose of taxation, from the 12th century AD, naming utthe, meda and
pesse, in decreasing order of soil fertility. As mentioned in the commentary
Vissuddhimargasannaya, the farmer’s share of water was called diyamura and for this he had
to pay a stipulated amount to the water supplier. In the early Anuradhapura period, this was
called dakapati and in the late Anuradhapura period, diyabedum.
From the above discussion, it is evident that there were several economic transactions
associated with the ownership of irrigation works and use of water. Owners of a single
component of a cascade could earn by supplying water or fish yields to the down stream
users. At the same time they had to pay for water received from upstream suppliers. These
transactions could be analyzed by using a model cascade as shown in Fig. 2, consisting of N
number of suppliers (S, the owners of the reservoirs, N number of end users and U, end users
like farmers and other utilities).
- 131 -
Sri Lankan Journal of Management
Volume 15, Nos. 2 , 3 & 4
It should be noted that a supplier down stream is also a user with upstream. Therefore, his
earnings would consist of receipts from the next supplier and the user while his expenses
would consist of payments to the previous supplier and payments to the state. The difference
will form the savings. Considering supplier Si , this economic relationship could be illustrated
as eq 1.
PSi+1 + PUi = PSi + TSi + SSi
[eq.1]
Where,
PSi+1
: Revenue received by Si from the next supplier
PUi
: Revenue received by Si from the next end user
PSi
: Payment by Si to the previous supplier
TSi
: Payment by Si to state
SSi
: Savings by Si
Applying eq. 1 from S1 to SN,
PS2 + PU1 = PS1 + TS1 + SS1
PS3 + PU2 = PS2 + TS2 + SS2
…………………………………………………..
PSi+1 + PUi = PSi + TSi + SSi
…………………………………………………..
PSN + PUN-1= PSN-1 + TSN-1 + SSN-1
PUN
= PSN + TSN + SSN
Taking the summation of the above set of equations,
N
N
6PSi + 6 PUi =
i=2
i=1
N
N
i=1
i=1
6 PSi +
N
6TSi +
6SSi
i=1
and by simplifying,
N
6PUi = PS1 +
i=1
N
6 TSi +
i=1
N
6 SSi
i=1
Rewriting this equation, PU = PS1 + TS + SS
Where,
N
PU = 6 PUi : Total payment by end users belong to this cascade
i=1
PS1
:
Payment by S1 to the state
- 132 -
[eq.2]
Economic Aspects of the Ancient
Irrigation System of Sri Lanka
N
TS = 6TSi : Total earnings by state from this cascade
i=1
N
SS = 6 SSi : Total saving by owners of the components of cascade
i=1
The capacity of the first reservoir in the cascade should be large enough to feed all the
reservoirs down stream. Therefore, that would be a state owned property and no payment
will be applicable, making PS1=0 and the above equation will become,
PU = TS + SS
[eq.3]
The above economic analysis could be merged with technical perspectives also by considering
the following conditions:
1.
The amount of water available to the cascade is a fixed quantity at any given
time
2.
Multiple rates applied for taxation depending on the retention capacity and the
function of the reservoirs.
The main principle behind condition (i) is that the development of irrigation structures should
ensure that rate of withdrawal from the water source should not exceed the rate at which the
source is renewed. Further, the withdrawal should not jeopardize the biodiversity of the
water source.
Condition (ii) could be verified by the fact that there were several types of reservoirs available
in a cascade. Arumugam (1957) identifying the importance of small scale village tanks notes
that the village reservoir is the provider of all the material needs of village life. Thus when the
reservoir breaches the village migrates to a different location. Panabokka, et al (2002), analyzing
technical data of several cascades in detail, note the different water spread areas of reservoirs
varying from more than 80 ha to less than 10 ha. Based on the above evidence it could be
assumed that a cascade consisted of different scales of reservoirs. The ratio of total reservoir
capacity to the command area is another parameter used to calculate the difference in the
size of reservoirs. Tennakoon (1986), highlighting the functions of small reservoirs, notes that
apart from irrigating cultivation, they provided multiple uses including augmentation of the
ground water table in order to keep the domestic well water supply at a minimum level
during the protracted dry season. Another function of some reservoirs was sill tapping, which
was known as kuluwewa, during the rainy season.
There is no direct evidence to conclude that different rates of taxation were imposed for
different categories of reservoirs. But Codrington (1937) quotes several examples from ancient
Hindu legal sources to prove that different tax rates applied to farmers depending on the
means of irrigation used for cultivation. Such methodology would have been well extended
to water supply also as it was not fair to apply a flat rate of taxation, irrespective of the size and
- 133 -
Sri Lankan Journal of Management
Volume 15, Nos. 2 , 3 & 4
function of the reservoir. Further, only such discrimination would allow maintaining the
controllability of the water flow through the cascade.
4. Regulatory measures
In the light of the above discussion the terms of eq. 3 could be expressed as follows ,assuming
that only two categories of reservoirs were available in the cascade under consideration.
PS = p[Q1+Q2] and TS = r1.Q1 + r2.Q2
Where,
p
: average rate at which end users are paying
r1
: rate at which category 1 reservoir owners are paying
r2
Q1
Q2
: rate at which category 2 reservoir owners are paying
: water availability for category 1 reservoirs
: water availability for category 2 reservoirs
And hence
p [Q1+Q2] = r1.Q1 + r2.Q2 + SS
Simplifying,
[p-r1].Q1
= - [p-r2].Q2 + SS
Q1
= - [p-r2]/ [p-r1].Q2 + 1/ [p-r1].SS and denoted by f1 [Q1, Q2]
[eq.4]
Here p should be greater than r1 and r2, for a positive gain for reservoir owners. If the total
savings from the cascade is s constant, eq.4 is in the form of a straight line in Q1-Q2 plane.
The technical restraint that the total amount of water which could be released to the cascade
from the natural water stream is a constant, say k, could be mathematically expressed as,
Q1 + Q2 = k
Q1 = -Q2 + k, and denoted by f2 [Q1, Q2]
[eq.5]
By plotting both equations in a single Q1-Q2 plane, the operating point of Q1 and Q2 could be
found for the given values of p, r1, r2 and SS. These are marked as Q1A and Q2A in Fig. 3. By
changing the values of these parameters, the inclinations of graph f1 [Q1, Q2] could be changed.
Or else by changing the total amount of water availability for this cascade, the graph f2 [Q1, Q2]
could be shifted to the left or right. Both these changes give new operating points, which
imply that the demand for water at any given time, by different categories of reservoirs could
be controlled by varying parameters p, r1 and r2 or byvarying the availability of water. That kind
of control measure would be essential for the sustainability of a cascade system, which may
otherwise lead to saving maximization affecting the water balance of the system.
- 134 -
Economic Aspects of the Ancient
Irrigation System of Sri Lanka
5. Conclusion
Based on the above discussion it could be concluded that the ancient irrigation system in Sri
Lanka possessed a sound economic footing, in addition to the much explored technical
background. Interested parties might have invested in irrigation enterprises and sometimes
diverted the returns of their investments to the maintenance and upliftment of Buddhist
monasteries. Thinking of the massive numbers of monks residing in monasteries like Mahavihara,
Abhayagiri, and Chethiyagiri, which is known as Mihintalaya today, etc., there should be a
continuous fund flow for their susceptibility and the revenue from the reservoirs might have
fulfilled that requirement.
Returns on those investments were subjected to a regulatory process. The regulating bodies
might have used the water rates to decide the capacity of reservoirs and canals. This was an
essential feature in an irrigation system like the one that prevailed in ancient Sri Lanka, in
which many reservoirs and canals were interconnected in the form of cascades. The personal
interests of the owners may not be aligned with the optimum operation of cascades. If
individuals were allowed to construct irrigation structures as they wished, such a strategy
would adversely affect the sustainability of an interconnected cascade. Such a phenomenon
could be prevented by manipulating water rates by the regulating bodies.
- 135 -
Sri Lankan Journal of Management
Volume 15, Nos. 2 , 3 & 4
References:
Arumugam, S.,(1952). Development of Village Irrigation Works, Transaction for 1957, The
Institute of Engineers, Ceylon, Colombo.
Codrington, H.W.,(1980). Ancient Land Tenure Revenue in Ceylon, 1937, Sinhala translation by
Department of Education Publications, Colombo.
Wilhelm Geiger., Mahavamsa Tr. P.T.S, London, 1961, reprinted by Buddhist Cultural Centre,
Colombo.
Mendis, D.L.O., Water Heritage of Sri Lanka, (2002) Sri Lanka Pugwash Group, Colombo, 2002
Panabokka, C.R., Sakthivadivel, R., Weerasingha, A.D., (2002). Small Tanks in Sri Lanka: Evolution,
Present Status and Issues, International Water Management Institute, Colombo.
Perera, L.S., The Institution of Ancient Ceylon from Inscriptions, Vol. I, reprinted by International
Centre for Ethnic Studies, 2001
Siriweera, W.I., (2004). History of Sri Lanka: from earliest times up to the sixteenth century,
Dayawansa Jayakody and Co., Colombo.
Tennakoon, M.U.A., (2004). Tanks are not Mono Functional: They are Multi Functional, Small
Tank Settlements in Sri Lanka, Proceedings of a Symposium, Colombo.
Walpola, Rahula., (1993). History of Buddhism in Ceylon, Buddhist Cultural Centre, Colombo.
- 136 -