Scientific Validation of some Traditional 
Land and Water Management Practices 
under Village Tank Cascade Systems 

Prof. CM. Madduma Bandara 
Emeritus Professor, 

Abstract 

M any recent writings on 
indigenous knowledge 
and practices in Sri 

Lanka appear to be largely 
descriptive than analytical, and 
incl ine to stress the virtues of 
ancient wisdom with a slant 
towards romanticising the past. 
Any attempts towards looking at 
these traditional practices from 
a scientific view point are hard 
to come-by. Some indeed believe 
that, indigenous knowledge 
systems cannot and need not be 
assessed from a scientific 
perspective. However, probing 
into the scientific bases of such 
practices is necessary to ensure 
their continued application as 
well as their very survival. 
The paper attempts to report the 
results of a study undertaken 
recently tn the north western 
part of the Anuradhapura 
District with sites in 
close proximity to Maha 
Willachchiya, Tanthirlmale and 
Medawachchiya. It encompassed 
an examination of a few village 
tank clusters, their cascade 
systems and associated 
environmental characteristics. 
In particular, the attention was 
focused on a few traditional 
water management and land use 
practices such as those 
associated with laterally 
connected tanks (baendi wew or 
aendutu wew), cistern sluices 
(keta or kumba horow), salinity 
interceptor belts (kattakaduwa) 
and on water circulation 
patterns in the cascades. An 
attempt is also made to bring out 
some implications of the findings 
of the study for future planning 
and technical design. 

In t roduc t ion 

Small tanks or reservoirs that store 
water for dry seasons, have always 
formed the life line of village 
economies and human well-being 
in the Dry Zone of Sri Lanka. They 

are of multiple use (for 
irrigation, domestic use and 
livestock) and supports 
aquatic ecosystems and 
human settlements in a 
geo-physical environment 
that would have otherwise 
been left parched and 
desolate. It has been 
demonstrated that, these 
tanks are not isolated 
entities, but often found in 
clusters forming part of a 
hydrologically integrated system 
that is now known as a 'cascade'. A 
'cascade' is defined as a 
"connected series of tanks 
organised within micro- (or meso) 
catchments of the dry zone 
landscape, storing, conveying 
and utilising water from an 
ephemeral rivulet" (Madduma 
Bandara, 1985, Panabokke, 2002) 
(Fig. 1). However, there is no 
evidence to suggest that all these 
tank systems were fully functional 
or operational at all times in 
history. Some tanks remain 
abandoned or in a poor state of 
repair at different times. Similarly, 
not all the tanks in the cascades 
apparently were meant for 
irrigation, but also used as silt 
traps (isweu) for fishing or for the 
benefit of domesticated animals 
and even wild life. In addition, now 
they are being absorbed to the 
natural landscape and has become 
a part of adjacent terrestrial 
systems. 

As a reputed Director of Irrigation 
during the colonial days (Kennedy, 
1936) observed " . .Every Village 
I r r iga t ion work has an 
ind iv idua l i ty o f i t s own, and 
when located on the topographic 
map, t h e e n g i n e e r has ... to 
acqu i r e t h e ' s e n s e and 
subs tance* o f t h a t 
ind iv idua l i ty" . In other words 
Kennedy was searching for that 
elusive 'sense and substance' in 

Sudharma Yatigammana 
Faculty of Science, 

Ganila Paranavithana 
Faculty of Engineering, 

University of Peradeniya. 

order to better grasp and 
understand the essential nature of 
small village tank systems 
(Panabokke, 2002) . The cascade 
concept has obviously advanced 
this philosophical thought further, 
as is evident in its definition itself. 

Method o f Approach 

Five village tank cascade clusters 
in three administrative Divisions of 
Maha Wilachchiya, Nuwaragam 
Palata Central and Medawachchiya 
came within the purview of this 
study (Table 1). 

A reconnaissance field 
investigation undertaken at the 
beginning of the study has revealed 
some features that deserved further 
probing: 

i. Circulation of water within the 
cascade system: both along the 
cascades and across the basins. It 
was obvious that water circulation 
patterns were different in different 
cascades . Lateral connections 
among the tanks (as at 
Ethdathkalla and Pandiggama) as 
well as the longitudinal relations 
along the valleys could be seen at 
several places. 

ii. Along with water-flows, aquatic 
biota -both plant and fish species 
also seemed to circulate in the 
cascades in both upstream (fish) as 
well as downstream directions. 
Agricultural wastes and pollutants 

Economic Review: April/ May 2010 21 



Table 1: Cascade Clusters, Cascades and Tanks 
in the Study 

Cascade Cluster Name of Cascade Total No. of 
Operational 
Tanks 

Parana Halmillewa Parana Halmillewa 14 
Etaweeragollewa 06 
Galegama 04 

Navodagama Navodagama 09 
Sandamal Eliya Sandamal Eliya 12 
Kahagollewa Kahagollewa 06 

M. Katukeliyawa 05 
Wanni Palugollewa 02 

Puwarasankulama Madagalla 02 
Parasangaswewa 02 

Total - 0 5 10 6 2 

Table 2 : Sources of Data and Information 

est. 

Cartographic Maps and Field Sketches: 

• ABMP* Maps - 1:10,000 
• Topographic Map 1:50,000 
• One Inch Topographic Map : 1:63.360 

• Satellite Images : Google 2009 (Image date 2003) 
• Aerial Photographs (1956,1992) 
• Land Use Maps of cascades prepared by the District 

Land Use Division 
• Field sketch maps 

Water quality investigations 
Engineering level measurements 

* ABMP: Agricultural Base Mapping Project 

Jalagilma areas 
were incorporated 
in some 
r e h a b i l i t a t i o n 
processes. 

Fig . 1 : (a) B a n d a r a R a t m a l e Cascade nea r 
Mihintale (After Madduma Bandara, 1985) , and 
(b) & (c) Impressions o f a Cascade System (After 
Ratnatunge, 1 9 7 0 ) 

also seemed to circulate along the 
cascades. 

iii. Crop calendars and the 
sequence of agricultural activities 
in each tank in the cascade 
appeared to have some relationship 
with water availability. 

iv. New development efforts, as well 
as emerging disputes, revealed the 
need for review of several prevailing 
legal regimes related to land and 
water at village and cascade levels. 

v. Certain traditional cascade 
management pract ices such as 
protecting or re-introducing 
kattakaduwa and demarcating 

vi. Sociological 
patterns of village 
settlements in 
different parts of 

the cascade also seemed to impinge 
on any attempts towards cascade-
wide institutional developments. 

vii. New agri-roads and other 
infrastructure, appeared to address 
the emerging needs with the advent 
of modern agricultural technologies 
such as combined harvesters 
(Sunami) 

S o u r c e s and Nature o f Data 
Col lec ted 

Some details of data sources used 
are given in Table 2. 

Field visits for engineering level 
measurements were undertaken on 

2 / 1 2 / 2 0 0 9 (Kadawath Rambewa) 
Katukeliyawa ( 4 / 1 2 / 2 0 0 9 ) Parana 
Halmillewa ( 1 2 / 1 2 / 2 0 0 9 ) Maha 
Katukeliyawa ( 2 9 / 1 2 / 2 0 0 9 ) and 
Ethdathkalla ( 1 5 / 0 1 / 2 0 1 0 ) . 
Similarly, field visits for water 
quality measurements took place 
from 12 to 13 December and from 19 
to 20 December 2009 , at which 
water samples were collected from 
tanks and kattakaduwas sites below 
the tank bund. Therefore, the period 
of field visits fell within the wet 
season in the area. 

Tools and Methods 

Some issues identified during the 
reconnaissance survey and 
appeared amenable to scientific 
assessment, were examined 
further in each cascade system with 
the help of: 

• Available land . use and 
topographical maps at different 
scales 

• Aerial photographs and satellite 
images where feasible 

22 Economic Review: April/ May 2010 — 



tanks at Paranahalmillewa cascade 
(excluding Galegama but including 
Etaweeragollewa) came up to 36. 

Galegama tank in the northern part 
of the cluster flows into a different 
cascade where water flows from 
Galegama, down to Kongollewa, 
Lciugaswewa, and finally to Aluth 
Halmillewa before joining Boo Oya 
(Fig. 5). Etaweera Gollewa is a small 
cascade with two main tanks where 
Pahala Tammennawa form the 
lower tank from which water drains 
into drainage channel in the main 
axis of the cascade. A satellite 
image (Google 2 0 0 8 ) shows the 
location of the lower part of the 
cascade cluster where some of the 
selected tanks are situated. 

Navodagama cascade 

This is single cascade system with 
five main tanks, namely, 
Navodagama, Wanni Helambawewa, 
Maha Helambawewa, Ethdath kalla 
and Millawetiya, located in the 
Wilachchiya Division. It also has 
at least 7 abandoned tanks, out of 
which a few have been recently 
renovated. The largest of these 
abandoned tanks is 
Muththankulama located at the 
down end of the cascade that drains 
directly into the Wilachchiya Major 
Irrigation Tank. In the historical 
past, Muththankulama, would have 
functioned as the biggest tank in 
the Navodagama cascade. Today, it 
cannot be renovated, because 
people have settled in its tank-bed 
and a part of its old paddy tract if 
asweddumised gets inundated by 
the water spread of Wilachchiya 
Major Irrigation tank downstream. 

• A household survey was 
undertaken in all selected tank 
clusters using a questionnaire 

• A few Case studies to understand 
lateral flows, and spill level 
measurements 

• A few scientific investigations of 
water quality undertaken in 
selected cascades 

• Some laboratory analyses of 
collected water samples undertaken 
to determine a few selected 
chemical properties such as 
conductivity, salinity and pH. 

Cascades In the Survey 

Some 5 tank clusters' in 3 different 
Divisions geographically separated 
from each other, came within the 
purview of the study. Each cluster 
is comprised of a number of 
cascades and each cascade has a 
number of tanks. A list of these 
'tank clusters' and 'cascades' is 
given in Table 1. A brief description 
of each cascade cluster is 
presented below: 

Parana Halmil lewa c a s c a d e 
c l u s t e r 

As indicated in Fig. 5 below, this 
cluster is located in the 
Medawachchiya Division along the 
Kebitigollewa Road. This comprises 
of at least 3 distinct cascades , 
namely, Parana Halmillewa, 
Galegama and Etaweeragollewa. 
Kadawath Rambewa is considered 
as a single tank draining directly 
into the main stream below 
Peddegama Tank, The upper-most 
tank in the cascade is 
Indigahawewa belonging to 
Pandiggama systems at the top 
end of the cascade. From there, 
water flows into Pandiggama 
Kuda wewa, and then to 
Ethakada, and from there to 
Tikiri Kumbugollewa, 
Peddegama and finally to 
Parana Halmillewa located at 
the lowest end of the cascade. 
Altogether, a total of 35 tanks 
both operational as well as 
abandoned were recorded in the 
baseline survey completed in 
April 2009 (Jayakumara et al., 
2 0 0 9 ) . With the Gallehewa 
small tank that we found in the 
upper reaches of Pandiggama 
Kudawewa, the total number of Fig. 2 : Tanks in the Sandamal Ellya Cascade 

Older maps indicate that even 
Ethdathkalla and Millawetiya were 
abandoned tanks that have been 
renovated in recent times. One of 
the most interesting features of the 
cascade is the existence of three 
parallel connected tanks, namely, 
Millawetiya, Ethdathkalla and 
Maha Helamba wewa. These are not 
connected now as in the past (as 
depicted on old topographic maps), 
because they have been renovated 
as single tanks. However, the 
wisdom of disintegrating the 
ancient system remains 
questionable. 

Sandamal Bliya cascade 

This is a cascade covering a large 
area with numerous abandoned 
tanks, situated close to Wilpattu 
National Park, and mostly in the 
Miocene Limestone area (Fig. 2). 
Plan Sri Lanka has renovated 6 
village tanks within this cascade, 
namely, Dalukpotana, Pahala 
Gonewa, Maha Viharagama, 
Manewa, Kuda Kumbuk Gollewa 
and Helambagaswewa. Some of 
these tanks do not appear even in 
recent Topographic Maps at the 
Scale of 1:50,000, and it appears 
that Plan Sri Lanka has responded 
to requests from respective village 
communities in choosing them for 
rehabilitation. This cascade also 
drains into Moderagam Aru joining 
it about 2km below the bund of the 
Maha Wilachchiya Major Tank. 

Kahagollewa cascade cluster 

There are three cascades in the 
Kahagollewa cluster, namely 
Kahagollewa, Katukeliyawa and 

Rehabilitated Tanks 

Existing Tanks 
| ^ Paddyfields 

Roads 
-~—~ Streams 

Cascade Boundary 
• Spot heights 

Economic Review; April/ May 2010 23 



PIg.3: Levelling points a t three tank systems with lateral relations 

both ways (e.g., Pahala 
Parasangaswewa). In some 
tanks people had to use nets 
across spillways to prevent 
such movement (e.g., 
Katukeliyawa). This brings out 
the need to take the whole 
cascade into account in 
introducing fish fingerlings. It 
also shows the need to re­
design spill ways to facilitate 
fish movement through some 
form of fish-ladders'. 

Wanni Palugollewa. This area is 
predominantly a Muslim settlement 
except for a few Sinhala Villages 
such as Wanni Palugollewa and 
Kokpetiyawa. Wanni Palugollewa 
has one operational tank with an 
elongated paddy tract. As could be 
surmised from cascade 
considerations, somewhere near 
the middle of this paddy tract, there 
are ruins of an old tank bund 
indicating the existence of another 
small tank in the historical past. 

Puwarasankulama cascades 

A few minor tanks were developed 
by Plan Sri Lanka near the 
medium-sized tanks of 
Puwarasankulama and 
Paniyankadawala, falling within two 
micro-catchments , namely, the 
Madagalla and Parasangaswewa 
cascades. Tanks rehabilitated were 
Madagalla and Elapathgama in the 
former and Pahala 
Parasangaswewa in the latter. 
These were indeed 
abandoned small tanks 
before taking up for 
rehabilitation. Despite 
their small size, they 
display some 
characteristics common to 
rehabilitated cascades 
elsewhere. 

There were hardly any 
permanent settlements 
under these tanks earlier. 
After rehabilitation, 
second generation settlers 
from neighbouring 
settlements of 
Puwarasankulama and 
Paniyankadawala received 
land and began cultivation 
under the tanks. 

T a n k s Affec ted by Mate r i a l 
R e c e i v e d from Upst ream 
Catchment Areas 

As part of the household survey 
conducted, people were asked 
whether their tanks are affected by 
upstream flows of fish, other fauna 
and flora as well as agricultural 
wastes. The responses received are 
summarised in Table 3. 

As could be seen, nearly 30% of the 
respondents reported that such 
flows affects their tanks in a 
significant way. The most affected 
tanks appear to be Navodagama, 
Parana Halmillewa and Kuda 
Kumbugollewa. One common 
feature in the last two tanks is 
their location in the lower parts of 
the respective cascades . 
Sometimes, the effects are not 
necessarily negative. At least in a 
few cases, it has been reported that 
fish introduced to upstream or even 
downstream tanks, ended up in a 
different tank, because they migrate 

Trans-migration of aquatic flora and 
fauna and agricultural wastes 
appears to produce more negative 
than positive results. Colonisation 
of tanks with unprecedented 
volumes and varieties of aquatic 
plants such as lotus, 'minneri lavan' 
and even the traditional 'ramba'may 
be considered as signs of 
eutrophication. Of course, lotus -
an edible plant with flowers used 
for worship, is useful if it does not 
reduce tank water capacities 
significantly. However, its 
unprecedented profuse growth in 
recent years is exceptionally 
noticeable. Similarly, 'ramba' as 
seen at Pandiggama is a good 
elephant feed. On the other hand, 
'minneri lawzn'not only reduces tank 
capacity as it forms a floating mat', 
but also discourages people from 
bathing in the tanks due to its itchy 
effect on the skin. At Tikiri 
Kumbugollewa, people were seen 
collecting them and burning large 
heaps on the tank bed. 

Table 3 : Tanks reported as receiving significant quantities of fish, 
aquatic plants, and agricultural wastes from upstream areas 

Cascade Cluster Cluster-
level 
responses 

Tanks reported 
receiving 

Tanks reported Receiving 
fish, wastes,etc. from 
upstream areas 

Parana Halmillewa 21 Parana Halmillewa 08 
Hinguruwewa 04 
Pahala Tammennewa 04 
Kudawewa 03 
Galegama 01 

Navodagama 15 Navodagama 14 
Sandamal Eliya 20 Kuda Kumbkgollewa 08 

Pahala Gonewa 05 
MahaViharagama 03 
Helambagaswewa 03 
Dalukpotana 01 

Kahagollewa 01 Kahagollewa 00 
Katukeliyawa 01 
Wanni Palugollewa 00 

Puwarasankulama 05 Madagallewa 02 
Parasangaswewa 02 

62 (29.8%) 59 (28.36%) 

Economic Review: April/ May 2010 — 



Role and R e l a t i o n s h i p s o f 
Latera l Tanks [baendi wew or 
aenduthu wew) in the Cascade 

Water circulation in a cascade is 
not only longitudinal along the 
cascade from upstream to 
downstream, but also it happens 
sometimes between parallel lateral 
tanks. Brohier (1937) has discussed 
the inter-connectedness of larger 
ancient reservoirs. Nevertheless, 
this aspect has not received 
adequate attention in tank 
rehabilitation and cascade studies 
in the past. It is undoubtedly an 
element of our ancient tank 
irrigation technology that reflects 
some enlightened attempts towards 
optimising the use of scarce water 
resources in small catchments in 
the Dry Zone. In our study areas, 
we have come across several 
places where this technology is still 
in pract ice or there is some 
evidence on ground that it has been 
used in the past. These include 
Ethdathkalla, Millawetiya and Maha 
Helambawewa tank system, 
Aliyawetuna wewa, Loku 
Katukeliyawa and the Pandiggama 
subsystems. At Sandamal Eliya 
cascade too Mahaviharagama and 
Kaluvila tank systems display some 
lateral relations. We have 
undertaken some engineering level 
measurements in a few such places 
to ascertain the possible lateral 
water flows and their results are 
given in Fig.3. 

E t h d a t h k a l l a 
system 

l a t e r a l t a n k 

fbutlet from 
Jjhalagahawewa 

~~l r i 2 5 m m n 

I S p l " ° f [ inlet toKudawewa " ] 
Kudawewa ! • 

' — J , _ _ _ _ _ _ _ — — -\ 
• ___2_9QTrn_j 

rsp¥of 1 

, Mahawewa, 

Fig. 4 : Imperceptible level differences in the Pandiggama lateral 
tank system 

Fig. 5(a): Electrical conductivity variations along the Paranahalmillewa 
Cascade; Figures in black inside the tanks indicate salinity of tank water; 
Those below the bunds indicate salinity in Kattakaduwa areas 

Older one-inch topographic maps, 
clearly indicate that all three tanks 
in the system, namely, 
Ethdathkalla, Maha Helambawea 
and Millawetiya were connected 
through a continuation of the same 
tank bund. Level points taken by 
us at critical points indicate a flow 
direction towards Ethdathkalla 
from the two lateral tanks on either 
side, indicating water deficits at 
Ethdathkalla. However, in view of 
the fact that, all tank bunds were 
obviously built on the same 
contour, and the almost 
imperceptible level differences at 
respective spills, it would have also 

Fig. 5(b) : Electr ical conductivity variations along the 
Navodagama Cascade 

25 Economic Review: April/ May 2 0 1 0 



Fig. 5(c) : Elec t r ica l conductivi ty variat ions 
along the Kahagollewa Cascade 

been possible to generate reverse 
flows, when Ethdathkal la received 
excess ive water from high ra ins . 
The c a n a l t h a t c o n n e c t e d 
Millawetiya and E t h d a t h kal la is 
sti l l s e e n in r u i n s n e a r l a t t e r ' s 
recently constructed spill. This old 
lateral c a n a l between Millawetiya 
and Ethdathaklle tanks, possess a 
level difference of only 5 0 m m (2 
inches) and could be considered 
a s a level cross ing between these 
two t a n k s . Maha He lambawewa 
tank is located at a slightly higher 
elevation accord ing to the levels 
obtained, causing water to flow form 
it to E t h d a t h k a l l e along an old 
l a t e r a l l ink b e t w e e n t h o s e to 
tanks. It is likely that, not only the 
existence of such a link canal was 
not recognised by the engineers , 
b u t a l s o they h a v e effectively 
destroyed it in order to build the 
new spill-way. 

Aliyawetunuwewa 
tank system 

Aliyawetunuwewa 
is a small t a n k 
on the head of a 
c a s c a d e from 
which w a t e r 
flows down to 
Ihala and Pahala 
K a h a g o l l e w a 
t a n k s 
respectively. On 
the left bank of 
the t a n k , t h e r e 
is a feeder canal 
from a s m a l l 
fores t t a n k 
n e a r b y . T h i s 
was renovated by 
the elders of the 
vi l lage s o m e 
y e a r s a g o , b u t 
w a s not 

c o n s i d e r e d in 
t h e 
rehabilitation of 
Aliy awetu nu we wa 
t a n k . On t h e 
right bank also, 

there are t races of a forest canal , 
t h a t b r o u g h t o v e r l a n d flows of 
w a t e r from the r a i n s in u p p e r 
ca tchment areas . 

Pandiggama tank system 

Pandiggama is inhabited by people 
who cal l t h e m s e l v e s Wanniye 
Minissu, and take pride in the fact 
that , they belong to the original 
i n h a b i t a n t s of the I s l a n d who 
descend from Kuveni, the queen of 

Katukeliyawa tank system 

The lateral flow between Loku and 
Kuda Katukel iyawa t a n k s is still 
o p e r a t i o n a l d u r i n g t h e r a i n y 
seasons. Here, the flow is from Loku 
Katukeliyawa to Kuda Katukeliyawa 
that shows some water deficits. 

Fig. 6 (a): Plan o f Kumba 
Horowwa 

ancient Sri Lanka. Pandiggama is 
served by a cluster of 7 small tanks. 
The bunds of three bigger tanks , 
namely , Mahawewa, K u d a Wewa 
and Thalgaswewa are located on the 
s a m e contour . These three t a n k s 
were c o n n e c t e d l a t era l l y in the 
h i s t o r i c a l p a s t . W a t e r flow 
direction was from Thalgaswewa to 
Kuda Wewa and then to Mahawewa. 
About 2 0 years ago, Mahawewa and 
Kudawewa were r e c o n n e c t e d by 
restoring the old link. The villagers 
requested the Plan to connect the 
other two tanks (Thalgaswewa) a s 
well, but this was not taken up due 
to d e a r t h of f u n d s . T h e level 
measurements we made at critical 
points indicate that, the spill levels 
of t h e s e t h r e e t a n k s h a v e 
surprisingly small level differences 
in e levat ion of l e ss t h a n 1 5 c m . 
T h e s e m i n o r level d i f f erences 
between tanks lateral canals were 
poss ib ly m a i n t a i n e d al lowing 
excess water from one tank to be 
shared with other parallel tanks at 
the same elevation. Here too, it is 
possible that there could have been 
reverse lateral flows, depending on 
the intensity of rains that brought 
water to the tanks. 

F i g u r e 4 i l l u s t r a t e s t h e level 
difference of the old canal , from 
T h a l a g a h a w e w a to Kudawewa a s 
1 2 5 m m (5 i n c h e s ) . F u r t h e r , it 
indicates tha t the level difference 
between the spill of Kudawewa to 
spill of Mahawewa is only 9 0 m m 
(nearly 4 inches). 

Fig. 6 lb): A Functioning Keta 
Horowwa 

26 Economic Review: Apr i l / May 2010 



When global climatic changes and 
their local impacts are considered, 
the above-described lateral flow 
tank systems may prove valuable 
for rational water management in 
the future. This is particularly in 
view of the extremes patterns of 
localised rainfall (referred to as 
salaka wessa by some villagers) and 
dry spells that are likely to happen 
now with increased frequency. 

Water Quality Variations in the 
Cascades 

An attempt has been made to 
understand water quality variations 
along the cascades in order to 
understand their spatial patterns. 
It must be noted that, this is only 
a reconnaissance study with an 
extremely limited sampling frame 
and a limited number of selected 
parameters felt relevant for 
evaluation. A comprehensive water 
quality survey was beyond our 
purview in terms of available time 
and resources at our disposal. 
However, we suppose, such a 
comprehensive survey is a dire need 
in these times of high agro-
chemical use. 

The few cascades we selected for 
this exercise included, 
Navodagama, Kahagollewa and 
Parana Halmillewa. Water samples 
were collected and analysed in 
December 2009 at the height of the 
rainy season. Therefore, we 
presume that actual readings 
during dry weather conditions may 
prove different. Some results of our 
analyses are depicted on maps 
(Fig. 5. a, b, & c). 

V a r i a t i o n s in e l e c t r i c a l 
conduc t iv i ty 

As could be seen in Fig. 5a , 
Electrical Conductivity (EC) values 
of tank water (marked in bold black) 
when plotted on a cascade map, 
shows a definite variation from top 
end to the bottom end of the 
cascade. Thus geographical 
distribution pattern of EC revealed 
3 major features hitherto little 
known to science. These are: 

i. Increasing EC in tanks along the 
cascade in a downstream sequence 

ii. EC levels below the tank bund 
in Kattakaduwa (interceptor belts) 
area were several times higher 

iii. EC at Kattakaduwa also showed 
a definite increase in a downstream 
direction 

Implications of high EC which is an 
indicator of amount of ions, for 
tank rehabilitation are interesting 
and challenging. The decision made 
by some tank rehabilitating 
agencies such as Plan Sri Lanka to 
re-introduce Kattakaduwa, seem to 
be scientifically validated by these 
findings. Rice, being a salt sensitive 
crop, must be protected from even 
slightly saline waters to maintain 
high levels of productivity. It also 
reveals the ancient wisdom of 
leaving a belt of land below the 
tank bund, to absorb salinity 
appropriately called Kattakaduwa 
implying an area affected by salinity. 

The observed tendency for salinity 
concentration in the kattakaduwa 
may be explained in terms of 
seepage from dead-storage areas 
inside the tanks, where there is a 
salinity build-up due to the vertical 
salinity gradient found in most 
tanks (Schiemer, 1983). Seepage 
lines normally tend to bend under 
the weight of tank bunds and 
perhaps reappear in Kattakaduwa 
areas. 

The salinity distribution patterns 
were consistently similar in all the 
three cascades that we have 
selected, namely, Parana 
Halmillewa, Navodagama and 
Kahagollewa, that are 
geographically separated from each 
other (Fig. 5a, 5b & 5c). Therefore, 
we are inclined to conclude that, 
in general, this may be a recurrent 
phenomenon in almost all cascade 
systems. 

It may be noted that, most modern 
mechanical sluices have been, 
perhaps designed to convey bottom 
saline water from tanks, for 
convenience, without a proper 
understanding the above fact. On 
the other hand, traditional 'keta' or 
'kumba horowwa' (Fig. 6a & 6b) 
carried the top surface water of a 
tank which is least affected by 
salinity. However, almost all Keta 
and 'kumba horowwas', have been 
destroyed and replaced by modern 
irrigation engineering in small 
tanks. Therefore, there is a 
challenge to develop a new sluice 
design that can combine the best 
features of both old and the new. 

Increasing salinity normally 
increases stress on seed banks of 
aquatic plants, and weeds. 
(Schiemer, F. (1983).Yatigammana, 
2004). As a result, it may decrease 
the species richness of aquatic 
communities, resulting in some 
loss of biodiversity. The possible 
effect of higher levels of salinity on 
aquatic fauna may be more complex 
and remains less understood. 
Therefore, it would be necessary to 
make further inquiries to ascertain 
the actual impacts of salinity on fish 
breeding and aquaculture. Does the 
increasing levels of salinity make 
lower tanks in the cascade less 
suitable for aquaculture? Does it 
require a more circumspect choice 
of fingerling species that can adapt 
to salinity variations? In response 
to these questions, salt tolerant 
fish species such as Etroptus 
suratensis that are found in lagoons 
as well as reservoirs of Sri Lanka 
could prove to be the better adapted 
species in Kattakaduwa as well as 
the reservoirs, especially towards 
the lower end of cascades . 
Therefore, it may be suggested that, 
when introducing fish into 
reservoirs, the salinity levels of the 
systems must be taken into 
consideration to get better yields. 
Thus, it will facilitate the maximum 
use of high saline waters for the 
aquaculture. 

Spatial pat terns o f other water 
quality parameters 

Apart from EC and pH 
measurements , several other 
chemical parameters have also 
been determined from the samples. 
These included, Suphate, 
Ammonia, Nitrite, Nitrate, 
Phosphate, Dissolved 
Phosphorous, and Dissolved 
Oxygen. In addition, bio-indicators 
among which the zooplankton and 
phytoplankton groups also 
identified. The distribution patterns 
of some of these measured 
environmental variables indicate 
the following features: 

i. In Parana Halmillewa Cascade, 
pH increases significantly 
downstream from Tikiri 
Kumbukgollewa to Parana 
Hallmillewa 

ii. Both Phosphates and Sulphates 
seem to decrease downstream in 
the Navodagama cascade " 

Economic Review: April/ May 2 0 1 0 27 



rii. In t h e P a r a n a Halmi l l ewa 
C a s c a d e , Sulphate increases from 
Indigahawewa to E t h a k a d a T a n k s 
and marks a slight drop thereafter; 
However , f rom P e d d e g a m a to 
P a r a n a Halmillawe again there is 
an increase . 

iv. A m m o n i a too i n c r e a s e s 
downs tream a s S u l p h a t e s in the 
Navodagama Cascade 

v. At Kahagollewa C a s c a d e , there 
is definite increase in Nitrite levels 
downstream from Aliyawetunuwea 
to Pahala Kahagollewa 

vi. In Parana Halmillewa Cascade , 
Nitrite levels s e e m to behave in 
the same manner a s Sulphates a s 
described under section ii above 

vii. In the Navodagama C a s c a d e , 
N i t r i t e s a p p e a r s to d e c r e a s e 
d o w n s t r e a m a n d b e h a v e in the 
same ways a s Sulphates 

viii. Z o o p l a n k t o n s s e e m to 
d e c r e a s e s d o w n s t r e a m in the 
Kahagollewa Cascade 

ix. P h y t o - p l a n k t o n s s e e m to 
d e c r e a s e d o w n s t r e a m at P a r a n a 
Ha lmi l l ewa C a s c a d e from 
I n d i g a h a w e w a to Tikiri 
Kumbukgollewa. 

The above resul t s may be treated 
only a s pre l iminary findings for 
r e a s o n s given e a r l i e r . However , 
their geographical variations along 
the c a s c a d e s due to t h e i r 
s i g n i f i c a n c e , d e s e r v e s o m e 
i n t e r p r e t a t i o n . In t h e P a r a n a 
Ha lmi l l ewa C a s c a d e , E t h a k a d a 
T a n k (a medium-s i zed irr igat ion 
t a n k ) a p p e a r s to b r e a k t h e 
sequence of variations in Sulphates 
and Nitrates. This may perhaps be 
due to the uses of these nutrients 
by a b u n d a n t m a c r o - p h y t e s a n d 
phyto-planktons of the system. 

In general , a l m o s t all the above 
c h e m i c a l p a r a m e t e r s a p p e a r to 
d i sp lay s o m e i n c r e a s e s 
downstream with the exception of 
N a v o d a g a m a w h e r e s o m e loca l 
f a c t o r s m a y be a t p lay . T h e 
downstream increases in Sulphates 
and Nitrites which a r e normal ly 
a s s o c i a t e d with h e a v y a g r o -
chemica l u s e s in rice fields, may 
accumulate and conveyed through 
r u n n i n g w a t e r , r e s u l t i n g in the 
d i s t r i b u t i o n p a t t e r n s d e s c r i b e d 
above. In addit ion, in re servo ir s 
t h a t a p p e a r to have less m a c r o -
phytes within the l i t toral region 
where the absorbance of nutrients 

by b i o t a c o u l d be lower h a v e 
different properties of water quality. 

Some Implications of the 
Results of Study 

T h e i m p l i c a t i o n s of t h e above 
findings on tank construction and 
rehabilitation planning may include 
the following: 

a) T a n k s a t the lower end of the 
c a s c a d e s d e s e r v e m o r e r e g u l a r 
vigilance and monitoring of impacts 

b) Aquaculture development may 
have to take these findings in their 
p lans to in troduce fingerlings of 
different fish spec i e s with m o r e 
sa l in i ty t o l e r a n t o n e s in 
downstream tanks of the cascades 

c) The re lat ionships between the 
above findings with weed growth 
and their control in tank systems 
deserves further investigation. 

d) The u s e of a g r o - c h e m i c a l s by 
farmers in the upper cascade areas 
m a y need to be m o n i t o r e d a n d 
controlled in order to protec t the 
tank sys tems at the lower end of 
c a s c a d e s . Possibilities of planting 
pollutant absorbing grass varieties 
or s e d g e s a long t h e d r a i n a g e 
channels (or Kiul Ela) , may prove 
helpful in controlling the inflow of 
a g r o - c h e m i c a l s to r e s e r v o i r s a t 
least to some extent. Development 
of engineered wetlands on drainage 
p a t h s and c a n a l s in the c a s c a d e 
s y s t e m . T h o s e w e t l a n d s to be 
designed a s free surface wetlands. 
Cattail species (Taipha species) is 
proposed a s the plant s ince it is 
suitable to the dry zone. 

e) Any health impacts of changing 
w a t e r qua l i ty in, t h e t a n k s for 
h u m a n s (for bathing and washing) 
a s well a s to domesticated animals 
(particularly to buffaloes, cattle and 
goats) and wild animals and aquatic 
b irds a n d fish where n e c e s s a r y , 
need to be monitored. 

f) Most modern mechanical sluices 
have been , p e r h a p s des igned to 
convey bottom water from tanks, 
for convenience, oblivious to their 
salinity implications. On the other 
hand , traditional 'keta' or 'kumba 
horowwa' carr ied the top surface 
w a t e r of a t a n k w h i c h is l e a s t 
a f fec ted by sa l in i ty . However , 
almost all 'kumba horowwas have 
been des troyed a n d rep laced by 
modern irrigation engineering in 
small tanks . Therefore, there is a 
challenge to develop a new sluice 
design that can combine the best 
features of both old and the new. 

References: 
Brohier R. L. (1937), Inter-Relation of 
groups of ancient reservoirs and 
channels in Ceylon. Journal of Royal 
Asiatic Society, Ceylon Branch, 34 
(90), pp 6 4 - 8 5 . 

J a y a k u m a r a , S.K., Nianthi, R. 
Madduma Bandara , C M . ( 2 0 0 9 ) A 
Basel ine Survey of the P a r a n a 
Halmillewa Cascade in the 
Medawachchiya Division of the 
Anuradhapura District of Sri Lanka. 
Plan Sri Lanka (Unpublished Report) 

Kennedy, J . S . (1936). Administrative 
Reports of the Department of 
Irrigation, Colombo. 

Madduma Bandara (1985). Catchment 
Ecosystems and Village Tanks in the 
Dry Zone of Sri Lanka. In Strategies 
for River Bas in Development. ED 
Lundqvist, J . , Lohm, U. Falkanmark, 
M.). Geojournal Library, Dortrecht, 
Reidel Publishing Company. 2 6 5 -
277pp . 

Panabokke, C.R., Shakthivadivel.R, 
Weerasinghe A.D (2002).Small Tanks 
in Sri Lanka, Evolution, Present 
Status and Issues. IWMI, Colombo, 
73pp 

Ratnatunge, P.U. (1970). Unpublished 
Reports of the Freedom from Hunger 
Campaign Board, Colombo 

Schiemer, F. ( 1 9 8 3 ) . Limnology of 
Parakrama Samudra Sri Lanka. A case 
study of an ancient man-made lake 
in the tropics. W.Junk, The Hague. 
Dev. In Hydrobiol.12: 236 pp. 

Yat igammana , S.K. ( 2 0 0 4 ) . 
Development and application of 
paleoecological approaches to study 
the impacts of anthropogenic activities 
on reservoirs in Sri Lanka. Ph.D 
Theses, Queens University, Kingston, 
Canada. 233pp. 

Acknowledgements 
The a u t h o r s wish to record their 
g r a t i t u d e to Plan Sri L a n k a for 
providing a fascinating opportunity to 
study the ancient village tank systems 
that formed and essential component 
of symbolic cultural trinity, namely 
the wewa, we la, and dagaba. They are 
also thankful to Mr. Nayana Wijetileka 
(Open University) who assisted us in 
the field and creating GIS maps, and 
Ms. Buddhika Perera (Institute of 
F u n d a m e n t a l S tud ie s ) for her 
a s s i s t a n c e in ana lys ing water 
samples. 

Contact Details of Authors: 
C M . Madduma B a n d a r a , Ph. D 
(Cambridge) E m e r i t u s Professor, 
University of Peradeniya; 6 0 0 / 3 1 , 
Airport Road, New Town, 
A n u r a d h a p u r a . E-mail: 
madduhandigyahon rnm 

Dr. S.K. Y a t i g a m m a n a , Ph.D 
(Queens, Canada), Senior Lecturer, 
Department of Zoology, University of 
Peradeniya, Sudharma_y@yahoo.com 

Mr. G.N. Paranavithana, B Sc. (Eng) 
University of Peradeniya. 
GanilaO l@gmail.com n 

28 Economic Review: April/ May 2010 — 

mailto:Sudharma_y@yahoo.com
mailto:l@gmail.com