Small tank heritage of Rajarata

Convocation address of the Rajarata University by Dr. C. R. Panabokke.

I have chosen the title of today’s address for several reasons. Firstly, the small or village tank settlements, I believe, have always been the backbone of the Rajarata civilization from ancient times, and secondly these small tanks find their best expression across the various landscapes of the Rajarata. At the same time, the small tank systems have always occupied a priority place in our national heritage and conscience.

I shall be dealing mainly with the ‘medieval period’ of the small tank heritage from around the 4th or 5th century up to the 12th century. Because of the lack of reliable information that could be gathered covering the ‘ancient’ period prior to the 4th century. I shall not be dealing with that particular period.

Although many well known studies have been carried out in the recent past on the more impressive ancient major or larger irrigation works by several well known scholars such as Parker, Brohier, Nicholas, Paranavitana, and Gunawardena, no studies of similar scope have been reported in respect of the small tank system. From the social science perspective, the well known study at Pul Eliya by Leach stands out as a special land mark.

Wanigaratne (1998) considers that Sri Lanka’s irrigation heritage draws on two main traditions. The ‘Greater Tradition’ or the Mahasammatha follows the construction and management of thousands of small village based tanks systems; and the ‘Lesser Tradition or the ‘Chulasammatha’ that emerges from the construction and management of large storage reservoirs and canal complexes.

Interpretations of ancient texts by several historians, example Pridham, Abeywardena, Siriweera have also provided useful insights in to certain periods. Some of the best accounts of the small tanks systems as they existed during the latter half of the nineteenth century is provided in Ievers’ Manual of the North Central Province published in 1890, and from which I shall be drawing several conclusions in the latter part of my address.

The form and order of small tank systems

A proper scientific study of any kind of natural system begins with a broad characterization and understanding of its form and order. A general glance at the one inch in one mile topographic sheets of the Survey Department shows an apparently scattered and random distribution pattern of small tanks and paddy command areas across the land surface of the North Central Province. However, a closer examination of the natural drainage patterns in relation to the location of the individual tanks reveals a distinct cascade pattern as was identified by Prof. Madduma Bandara in 1983, and it is now referred to as the ‘Small tank cascade system’. Around the same period another son of the soil of the Rajarata, Dr. M. U. A. Tennakoon had been able to came up with the indigenous and traditional understanding of these cascades or the ‘Ellangawas’ which is the folkloristic term for these cascades; and he asserts that this ‘Ellangawa’ or cascade concept had been well entrenched in the minds of the ancient farmers and water resource managers. I am inclined to go along with him in another sense, in that I consider the small tank builders of that period had a profound understanding of their landscapes, landforms and landscape hydrology, or what we in modern science term as ‘geomorphology’ of the different landscapes of this region.

An IFAD funded study for the Anuradhapura Participatory Rural Development Project (PRDP) in 1995 enabled the International Irrigation Management Institute to study the small tank cascade systems in the Anuradhapura district in a wider setting. This was extended to cover the whole of the Rajarata in 1996 under assistance from the Mahaweli Authority of Sri Lanka. As a result of these studies it is now well established that the nine river basins that made by the Rajarata is made up of 50 sub watersheds and 457 Cascades of small tanks. This, in more or less a systematic manner, characterizes the form and order of the small tank system of the Rajarata.

It is within this framework of the location and hydrography of these 457 cascade meso-basins that meaningful analysis and interpretation of the hydrological endowment and carrying capacity of the approximately 3,000 functioning small tanks of the, Rajarata could be made. It is relevant in this connection to refer to a significant statement made as far back as in 1936 by a very distinguished former Director of Irrigation, J. S. Kennedy which states ‘Every village irrigation work has an individuality of its own, and when located on the topo map, the engineer has next to acquire the sense and substance of that individuality’. In other words Kennedy was searching for that elusive form and order in order to better grasp and understand the essential nature of a small tank systems that he was working with.

Distribution patterns

Although the Rajarata is located wholly within the dry zone of the north central region of this country, its natural environment is by no means uniform or homogeneous across this region as commonly perceived. There is a distinct variation in the annual and seasonal rainfall as well as in the natural hydrology as one proceeds from the eastern part of the region to the western part. Similarly, a distinct variation in landforms, soils and underlying geology can be observed as one proceeds from east to west.

One of the most striking features in the tank distribution pattern in the Rajarata is the decreasing density of small tank cascades as one proceeds from the eastern segment of this region to the western segment. Panabokke 1999 has discussed and explained the underlying reasons for this significant variation in tank density from east to west in the Rajarata region. There is clear evidence that a major portion of the Wilpattu National Park, especially that area outside the Moderagam Aru basin had never been a human habitat even in the ancient period as compared with the Yala National Park.

Evolution and design of early cascade systems

There are hardly any studies that have been carried out to date on the past evolution of these small tanks. Historical references, however, indicate that these small tank systems had been in existence in varying degree since the medieval period.

Indirect and circumstantial evidence points to the fact that it was these small scale irrigation reservoirs or tanks that were the precursors to the larger and more impressive irrigation works that were constructed during the reign of various rulers up to the tenth century AD. From all accounts it is clear that the spread of these small tank irrigation systems would have taken place concurrently with the construction of the major irrigation systems.

However, while the construction of the major large scale irrigation works were controlled and directed by the various kings and the higher echelons of the irrigation bureaucracy of that period, the small tank systems were initiated and constructed by the small village communities by a communal effort throughout this period of approximately fifteen centuries.

There are no recorded scripts or historical documents relating to the methods employed in site selection and the location of small tanks along these inland valleys during the ancient or medieval periods. At best, some fragments of the oral-traditional knowledge handed down from past generations is yet available with some of the village elders now living within these cascades.

The technology of construction and maintenance of these small tanks had been fairly simple and straightforward and certainly within the capability of village labour resources as well as their construction skills. Going by the fact that by the fourth or fifth century AD the design and construction of the major irrigation networks had attained a high level of technical sophistication, it can be reasoned that the technology of small tank design and construction could also have attained a high level of maturity by the same period, or even much before this period.

According to this oral tradition, the initial settlements usually took place in the middle segment of the main valley. The settlement started with a smaller tank which was progressively increased in size by successive generations. Evidence of this was recently seen in some recently excavated cross sections of bunds of some old small tanks that were being rehabilitated by the DAS, particularly where a whole cross section was exposed for placing the new concrete tower sluices in position. One could clearly observe different layers of differently coloured soil material of varying thickness that had been placed in position at different periods, which could not however be adequately dated as regards which periods they were placed.

According to tradition, in selecting the most appropriate location within the main valley, a site that had a low, rounded, rock outcrop that could have been used as a natural spillway essentially determined the location of the main tank bund. This had to be at the terminal end of either the left or right bank extremity of the main tank bund. In all of the small tanks which had a water spread size of more than 50 acres, the spillways were located at these natural low relief rock outcrops. When one studies the detailed geomorphology of this landscape, it is not surprising to observe the high incidence of such rock outcrops of varying shapes, form and relief that are present in this peneplained land surface.

According to Somasiri (1991), the location and distribution of small tanks in the Anuradhapura district of the NCP had been determined, more by social factors and site features rather than by the hydrological properties of the catchment. It is not known at which period a major proliferation of the presently existing small tanks took place within the individual cascades. One fact that is however quite clear, is that not all these smaller or micro tanks, especially in the NCP were for purposes of irrigated paddy. Instead they served a number of other multiple uses including augmentation of the groundwater table in order to keep the domestic well water supply at a minimum level during the protracted dry seasons and as silt trapping tanks (Kulu wewa) during the rainy season.

There are no records or traditional knowledge available on the hydraulic design and storage capacity of the tanks. It is not known how the key hydrological parameters were determined, especially those relating to the amounts of rainfall, runoff and tank volumes. A body of empirical relationships would have been known at that period, which could have helped in determining the amount of storage that would have been possible within a water body created by an earthen bund of specific dimensions.

It must be remembered, that by the fourth and fifth centuries AD, very advanced hydraulic structures had been designed and constructed which would have required a sound knowledge of some of the key hydrological relationships pertaining to rainfall, runoff and storage volumes. There is no reason to doubt that this body of expert knowledge that was available for major reservoir and sluice-channel construction could have been extended with some modifications for the design and construction of the smaller tank systems as well.

Siriweera (1989) in his review of droughts and famines in the medieval period records that for that whole period between the sixth century AD and the reign of Parakramabahu (1276) there are no references to famines in the chronicles or in literature. It could, thereby, be indirectly inferred that a greater part of the small tank cascades of the Rajarata were in some operational form over this six hundred year period which enabled the settlers of the period to combat affects of drought which are common even today.

Causes for abandonment in the past

Abeysinghe (1982) states that there are several thousand abandoned tanks in the dry zone of this country which must be kept abandoned for one or more reasons that he enumerates as follows, they lack adequate catchment area, they lack suitable soils for agriculture, they lack approach roads, or they are located far away from human habitation or on the borders of nature reserves or wildlife sanctuaries.

In respect of the north central province it is observed that in much of the western segment of this province the abandoned tanks are located in either poorly hydrologically endowed meso-catchments, or in very rocky and gravelly rock-knob-plain landforms. The early settlers had probably no proper appreciation of either the hydrology or the land quality in this region. In contrast, in much of the eastern segment of this province it is observed that there is a sharper relief of the meso land form and that this has contributed to the instability of these tanks. In times of severely cyclonic depressions that occur during the period November-December, high intensity rainfall amounting to over 300 mm per day is experienced. During such storms, the surge of the overland runoff is very high on this type of landscape because of its sharp relief, and this results in a breach of the bunds of tanks unless they are very robustly constructed. This is also the view supported by the local residents of this region.

In contrast, in the central segment of this province there is almost a total absence of abandoned tanks especially in the upper aspects of the sub watersheds where the relief tends to be very subdued. Local residents also state that the small tank cascades located in the upper aspects of these sub-watersheds have been in continuous existence and use since the medieval period.

Adopting a heuristic approach, it could be demonstrated that there are different sets of reasons for the abandonment of small tanks in the different parts of the Rajarata. These have been discussed by Panabokke (1997).

The period of decline 13th to 19th century

Wanigaratne ( 1999) observes that the volume of accretionary capital in labour efforts and material invested in the small tank systems would have been colossal. This capital represented cumulative investment by successive generations of village families in several thousand villages using resources within their reach; and this in turn would have led to the evolution of appropriate institutional mechanisms that helped to maintain some order and stability in the management of these small tanks over hundreds of years.

However their disconnection from the larger irrigation systems that fell into decline from the 12th century onwards also led to a slow entropic decline in the small systems as well, but not to the same extent of a total collapse

While the management bureaucracies of the larger irrigation systems which were under the central government collapsed totally together with the state, the village level tank systems and their management mechanisms survived in varying degree.

Robert Knox (1681 ) gives a graphic description of ‘Every town (village) has one of these ponds (tanks) of which there is a great number, the banks of which are in length above a mile, some less - not all of a size ‘He does not mention sluices and spillways, but he states that they cut a gap in one end of the tank in order to draw water for irrigating the corn (rice).

In the course of field studies conducted in the Kende, Kunchutti, and Ulagalle Korales, I have come across several tank villages in the hydrologically stable and better endowed location which claim an unbroken record of occupation throughout the whole medieval period. Those areas of the Rajarata which have a low percentage of abandoned small tanks could be considered as those which had a better endowed and stable hydrology, which in turn enabled an unbroken continuity of occupation.

In effect, M. U. A. Tennakoon’s, ‘Wew-Bendi - Rajje’ refers to this area which he identifies as the more hydrologically stable areas of the Rajarata and which encomposses much of the ancient and present Kende Korales and the Rambewa - Ranpathvila areas.

The nineteenth century

One of the best accounts of the state of small village tank systems as they existed in the latter half of the nineteenth century can be found in Chapter eleven, Irrigation, pages 132-169 of R. W. Ievers Manual of the North Central Province (1890). Two statements by Ievers shows his profound understanding of the Rajarata environment. These are

1. ‘It may be broadly stated that without artificial irrigation and storage of water, human existence in the North Central Province would be impossible’.

2. ‘As the North Central Province, although apparently flat, is in reality undulating, the ancient tank builders took advantage of this conformation to make chairs of tanks in the valleys’.

It is stated that in 1855 the total number of tanks in Nuwarakalawiya was around 2,000. Of these 1,514 tanks were with few exceptions, regularly cultivated. In the year 1873, the Nuwarakalawiya register shows a total of 2,87? tanks with 1,500 of these abandoned or uninhabited.

It also states that between 1815 and 1874, no government assistance was given to the people to restore or repair their sluices or tanks, and they were left to their own devices. On the top of this the British administration abolished the ‘Rajakariya’ system in 1832 on grounds that it was some form of slavery. No alternative system was introduced for the repair and maintenance of tanks by the community, and many minor irrigation works fell into neglect and general decay.

According to I. K. Weerawardena (1986), scripts and pillar inscription of the 10th century prove the point that farmers had to adhere to certain laws laid down by the king or regional chieftains in relation to the repair, maintenance and management of small irrigation systems. The adherence to these laws over many generations resulted in the birth of customs and traditions, which gave the irrigation systems a discipline which continued up to British times. The ancient Rajakariya system was a compulsory personal labour obligation that helped to guarantee the maintenance of these small irrigation systemic over several centuries.

After forty years of neglect, Governor William Gregory in 1873 created the North Central Province for the expressed purpose of giving long-needed relief for restoration of village tanks. Mr. J. F. Dickson the Government Agent NCP at that time proceeded to give effect to the new policy of Governor Gregory, and by 1879 he had successfully effected ‘astonishing improvements; according to Ievers. Experiments were made during 1888 for producing a cheap and efficient village tank sluice. By 1890 a total of 958 village tank had been provided with durable cement pipe sluices. This sluicing of tanks alone resulted in a remarkable increase in paddy production in the Rajarata.

Bertram Bastianpillai (1967) deals very completely and analytically with this period of British rule from 1870 to 1890 in his ‘Revival of Irrigation Enterprises’ in the Journal of Historical and Social Studies.

By the turn of the century, almost all small village tanks had been supplied with a durable sluice which helped to conserve tank water supply, and also to do away with the wasteful practice that Knox had described of ‘cutting a gap in one end of the bank to draw water little-by-little for watering their corn’.

What was to follow in the subsequent century has been adequately documented by several writers over the last fifty years. Aheeyar (2000) describes in summary, the forms of minor tank irrigation institutions that existed during the Pre-colonial Era, the Colonial Era, and the Post Independence Period. It is pointed out that the country’s customary water laws had evolved over the years for collective operation, and because water was the scarce resource, there were more tenurial concerns towards water than land.


There is now an increasing awareness of the multiple functions and importance of the diverse benefits of the presently functioning small tank, or village irrigation systems. It is also now well recognized that since these small tanks constitute a very important part of the rural landscape and its eco-systems, there is a strong rationale for ensuring the sustainability of these village tank settlements for economic, social and environmental reasons.

I have given you only some selected aspects of the heritage of these small village tank systems. Being located in the very heartland of the Rajarata, your University has all the opportunities to explore and study the many aspects of these cascade systems which I have not been able to identify and address in the course of my own studies conducted in the recent years.