TANK-VILLAGE SYSTEM AS A RESOURCE BASE FOR MULTI-PURPOSE TREES

Proc. of the 5th Regional Workshop on Multi-purpose Trees, (Ed.) HPM Gunasena, University of Peradeniya, Peradeniya, Sri Lanka. 1994 pp 8-19.

TANK-VILLAGE SYSTEM AS A RESOURCE BASE FOR MULTI-PURPOSE TREES P.B. Dharmasena Field Crops Research & Development Institute, Maha Illuppallama, Sri Lanka ABSTRACT Small tank-village in the dry and intermediate zones of Sri Lanka provides ample opportunities for development planners to explore its latent potential of agricultural production, which has never been fully realized. The time-tested three-fold farming system that these communities practised has undergone various evolutionary changes. It is now increasingly realized that the system gradually becomes less productive due to deterioration of physical resources and exhaustion of farmer economy. The important role of multi-purpose trees in addressing the above two matters is discussed in this paper on the basis of research findings so far gathered in relation to the resource utilization in these minor watersheds. The rainfed upland farming in tank catchment areas is found lacking of suitable measures for improving fertility, conserving soil and regulating runoff. Graded hedgerow farming with Nfixing trees (eg. Gliricidia sepium) or strip mulch farming with N-fixing leguminous creepers (eg. Mucuna utilis) were found as best alternatives for maintaining the system sustenance. Water storing efficiency in village tanks can be greatly improved by partial desilting method named as `Cut and fill altering geometry technique', which has been specifically developed for this environment on the basis of recent investigations. This reduces the water surface area to half and allows the other half of fertile moist tank bed for growing selected species such as bamboo (Bambusa spp.), rattan (Calamus spp.), mat grass (Cyperus pangorei) etc. Availability of such species will definitely provide opportunities to initiate cottage industries in these villages. The tank-village home-garden undergoes a very slow process of development because these lands are low resource marginal areas and do experience frequent moisture deficiencies. Introduction of fast growing N-fixing leguminous tree species such as Acacia, Albizzia, Casuarina, Gliricidia etc. will act as a catalyst to accelerate the development process of homegarden vegetation. It is now high time to realize that these innovative concepts must be tried out in order to make productive this risk prone but high potential environment, which was once supposed to be a sustainable farming system in Sri Lanka.

INTRODUCTION People in olden days made living in the dry zone possible by damming small ephemeral rivulets to store water for their needs. The farming practices stemmed from these communities are collectively referred to as the tank-village farming system. As the civilization flourished in these areas, a network of tanks and streams emerged creating a sustainable living environment later recognized as the hydraulic society of Sri Lanka (Leach, 1959). However, this civilization collapsed due to reasons such as foreign interference, disease and pestilence, but some communities could survive in isolation to continue the practice of small tank agriculture (Wijethunge, 1986). At present, high population of farming community and the country's demand for agricultural production have led to exploit the land resource more than once done in past. Thus, the tank-village farming system is changing to be unstable with respect to hydrology, ecology and economy of the system. Tank-village farming system is basically of three-fold: field crop cultivation under rainfed conditions in the well drained upland area; rice cultivation with tank irrigation in the command area of village tanks and a mixed cropping around their dwelling. However, this environment can provide more other opportunities to raise their income level such as animal husbandry, fishery, cottage industry etc. Development of the tank-village system has become a more pressing need than major scheme improvement as now it has been realized that there is a wide gap between the production potential and the present output of these agricultural systems. About 45 percent of the total irrigable area in the country is under village irrigation and it accounts for 23 percent of the national rice production (Thilakasiri, 1986). However, rice yield under minor irrigation is lower by 25 percent compared to major irrigation schemes. This can be more due to seasonal water shortages than any other factors. Land use intensity of the command area of village tanks even in maha season seldom exceeds 50 percent. As rainfall is uncertain (CV = 50 % for maha rainy months) farmers need to wait for cultivation until the tank gains an adequate storage. This leads failing to use most of the seasonal rains for their cultivation (Dharmasena, 1989). The rainfed upland cultivation (traditionally known as chena) in catchments of village tanks produces about 80 percent of Sri Lanka's rainfed grains, pulses and vegetables. However, the average yield of these chena crops during maha season is less than 50 percent of potential yield due to depletion of soil fertility, weed infestation and soil moisture deficiency (Weerakoon, 1992). All these evidences reveal that a) production potential in the tank village system is much higher than once realized in the past, b) it has not been fully exploited and b) hence, it is essential to seek for strategies to maintain this system ecologically balanced and economically sustainable. Further, present farming activities upset the ecological balance of the environment and impoverish the land, therefore, it is high time to test innovative concepts in all farming components to regain what was in the past as sustainable.

2

MULTIFORMITY OF RESOURCES Land and water resources in the context of tank-village farming system are of multifunctional and multi-purpose. This multiformity can be used as a base for developing strategies to improve the economic status of the community. There are seven agricultural opportunities found in the tank-village environment: 1) rainfed upland farming; 2) agrobased forestry; 3) livestock development; 4) command area cultivation; 5) homestead farming; 6) agro-based industries, and 7) fishery. Hydrological investigations carried out in tank-village watersheds showed that the rainfall is temporally stored in three different forms namely soil moisture, surface water (tank) and ground water. About 60 percent of the annual rainfall is trapped in the soil continuously transmitting back to atmosphere, 30 percent stored in the tank, and the balance percolates down to replenish the ground water reserve. Fig. 1 illustrates the utilization of three sources of water for seven agricultural opportunities. Fig. 1. Multi-functional water use pattern in the tank-village system.

SOIL MOISTURE 60-65 %

HO M

EG AR DE N

WATER FO% 100 SURFACE WATER 25-30 % LO WL AN D

GROUNDWATER 5-10 %

COTTAGE INDUSTRIES

OCK LIVEST

FO

RE ST RY

RAINFED

Y ER H C FI

Minor watersheds in the dry zone are a composite of three main land categories. They are: a. The upland, which is covered by forest or fallow vegetation or used for rainfed cultivation. b. The lowland, where tank irrigated farming or dug well cultivation is practised and c. The tank, which covers a part of the lowland to store water for irrigation and domestic use. About 70 percent of the extent in a minor watershed is upland area and the balance is approximately shared in equal extents by the tank and the lowland. The link between

3

land resource and agricultural opportunities is shown in Fig. 2. The distribution of land and water resources in these watersheds as mentioned above envisages the importance of paying more attention to upland area and focusing more agricultural opportunities to utilize soil moisture compared to other land and water sources. Fig. 2. Agricultural opportunities of land resources in tank-village system.

UPLAND 70 %

FO OCK LIVEST

HO M

EG AR DE N

LAND FO% 100 LOWLAND 15 %

LO WL AN D

TANK 15 %

COTTAGE INDUSTRIES

RE ST RY

RAINFED

Y ER H C FI

In finding next phase of agricultural development in the dry zone Panabokke (1992) stresses the importance of considering both the rainfed upland component as well as the small tank irrigated component as one integrated whole, including agro-forestry for the tank-village farming system. This brings us the message that the essence of sustainable farming must be derived from the concepts of integration of farming and inclusion of agro-forestry in planning future strategies of tank-village farming. Thus, the objective of this paper is to discuss the role of multi-purpose trees in formulating strategies to improve the tank-village farming system as one integrated whole.

MPTS IN RAINFED UPLAND The success of rainfed upland farming in the dry zone is constrained by 4 main factors (Somasiri et al, 1990). 1. Deterioration of the surface soil structure and texture affecting soil tilth and moisture status. 2. Difficulty in land preparation due to resurgence of obnoxious weeds. 3. Depletion of nutrient reserves in the soil. 4. Timeliness of land preparation to establish crops with the on-set of rains, because both low and high soil moisture status make soil condition unsuitable for working.

4

It must also be emphasized that any intervention in the upland areas will definitely have an effect on the village tank system and the associated settlements. Soil and moisture conservation is the key factor in improving the rainfed upland farming as the soil erosion has been the main cause for most of the above constraints. An assessment was made by the author in a previous study to assess the erosion controlling ability of different farming practices, and the results are summarized in Table 1. Since farming practices were tested on different land slopes, the soil loss data could not be directly compared in this study. Thus, the Universal Soil Loss Equation was used after modifying it to obtain an index (Ec), which is independent of rainfall erosivity and land slope (Dharmasena, 1992). Table 1. Soil loss under different farming methods (At Regional Agricultural Research Centre, Maha Illuppallama during Maha 1989/90 season). -----------------------------------------------------------------------------------------------------Farming Soil loss Ec Percent practice (t/ha) protection -----------------------------------------------------------------------------------------------------Chena farming 14.99 0.435 Plough farming 12.42 0.432 1 Terrace farming 8.33 0.260 40 Strip mulch farming 1.99 0.087 80 Graded hedgerow farming 1.28 0.061 86 -----------------------------------------------------------------------------------------------------Ec = Erosion coefficient. Results indicated that chena and ploughed lands had no considerable protection against soil erosion. A certain proportion of erosion could be prevented by terracing the land but not up to a satisfactory level. The two farming methods named as strip mulch and graded hedgerow had a distinct ability for conserving the soil, hence they are further described below. Graded hedgerow Gliricidia sepium is used as the hedgerow and planted on rows with a slight gradient (hence named as graded hedgerow) to facilitate the diversion of excess runoff into a protected waterway. Gliricidia is planted at 4 m spacing with 0.5 m within row distance. Small soil ridges are formed along the hedgerows with a small drain on the upper side. In addition a grass strip of 2 m is established in place of each fifth hedgerow at 20 m intervals where a line of fruit crops can be grown. Strip mulch A permanent strip of live mulch with Centrosema pubescen or Mucuna utilis is maintained at 5 m intervals and in 0.75 m wide rows across the slope. Weed and crop residues are also heaped on the same strip expecting to form a runoff filter reinforced by the live mulch. Excess part of this trash barrier can be spread over the cropped land as mulch. The selected farming methods must not only control soil erosion but also should regulate runoff, because rainfed upland needs more absorption of water in relatively dry months

5

and tank needs more runoff in relatively wet months. For this situation, rainfall-runoff relationship must have a high gradient with a wide intercept. Rainfall-runoff relationships found for different farming methods are shown in Fig. 3. The traditional practice in rainfed upland cultivation known as chena farming produces the highest runoff yield. Runoff threshold (defined as the amount of rainfall above which runoff is generated) is relatively low therefore, the chena farming practice does not encourage wetting the soil adequately before runoff commences. This is not a favourable situation for tank catchments. Nevertheless, chena farming has no protection against soil erosion hence it cannot be recommended.

400

300

CHENA GRADED HEDGEROW

200 STRIP MULCH PLOUGHED 100 TERRACED 0 100

200

300

400

500

Fig. 3. Rainfall-runoff relationships and runoff thresholds for different farming methods. Runoff threshold value and gradient of the rainfall-runoff relationships of hedgerow and strip-mulch farming practices are higher compared to that of plough and terrace farming practices. This leads more water absorption in dry months and more runoff in wet months with these practices qualifying them ideal for tank catchments, along with their potential for withholding the soil movement. This achievement is primarily due to reasons such as leguminous trees and creepers improve the soil physical properties, reduce the soil temperature, act as a mulch to protect the soil from erosion, and manage to generate non-erosive runoff.

MPTS IN TANK VICINITY Storage capacity of village tanks has been reduced by 20-30 % due to sedimentation of soils eroded from catchment areas. This has been accelerated during recent past due to increasing extent of chena cultivation (Dharmasena, 1993a). Recent tank renovation activities have led to increase the storage capacity by raising the dam height and spill

6

level without desilting, and consequently the water spread area in village tanks have expanded to a great extent. In most cases, such expansion caused inundation of the paddy land of the upper tank. In a study carried out by the author it was found that water loss from the tank has a good relationship with area/depth ratio of the water body (Fig. 4). It clearly shows that a shallowly spread water body causes more losses than a deep tank with a same storage capacity. The `cut and fill altering geometry technique' has been proposed on the basis of the above fact in order to form a deep storage without changing the capacity but with protective measures against sedimentation (Dharmasena, 1993a). Fig. 5 illustrates the desilting technique with the protective `eyebrow' bund on the tank bed which regulates the runoff flow into tank. Another important aspect of this technique is that about half of the extent occupied by the tank would become a free land which is fertile, and moist under the influence of ground water. Organic matter content of this soil is in the range of 5 - 8 % (Dharmasena, 1992). This land must be kept away from annual cultivation but may be utilized for a perennial type of vegetation. In a cottage industry improvement programme, this land may best be utilized to grow bamboo (Bambusa spp), rattan (Calamus spp), mat grass (Cyperus pangorei), Vetakeyya (Pandanas spp) etc. It is important to note that these plant species are seldom found in the tank village system at present as they need moist soil environment throughout the year. 100.0

Percent water loss

90.0

y = 59.471x-1.3351 R2 = 0.786

80.0 70.0 60.0 50.0 40.0 30.0 0.7

0.9

1.1

1.3

1.5

Capacity/area (m) Fig. 4. Effect of tank geometry on water loss. MPTS IN HOMESTEAD In old settlements the hamlet is located in imperfectly drained land adjoining to paddy tract below the tank bund. Due to the influence of ground water these home-gardens are covered with a well developed perennial vegetation which consist of mango, coconut, jak, etc. The expansion of hamlet area has taken place along roadsides on lands degraded due to chena cultivation. These new home gardens undergo a very slow process of development because lands are low resource marginal areas and do experience frequent

7

moisture deficiencies. Even under such circumstances 75 percent of the home garden plant species are of multipurpose (Dharmasena, 1993b). A home garden survey indicated that most possible reason for less availability of introduced tree species such as coconut, orange, lime breadfruit etc. in these home garden is the dry environment which does not favour the establishment of such crops. Thus, rapid formation of a favourable microclimate in the homestead is a prerequisite for establishment of most introduced tree species. At present, this role is played by large canopy wild species, which are tolerant to long dry periods (Dharmasena, 1993b). The development process of the home garden can be accelerated by planting fast growing tree species instead of wild species with slow growth. Introduction of fast growing N-fixing shade tree species such as Acacia spp., Albizzia spp., Casuarina spp., Gliricidia spp., Eucalyptus spp. etc. to the homegarden will create an early shady environment in which the introduced plant species would be survived. These trees will also help to enrich fertility of the soil which has been degraded due to continuous cultivation.

Tank catchment Natural streams Old tank bed Removed sediment new tank bed Tank bund

kattakaduwa

Fig. 5. Reduction of tank bed area due to desilting. Although the tank-village farming community makes a considerable contribution to the national agricultural production, these farm families live at a very low standard of living. Returns from their farming efforts do not satisfy them by any means. Improvement on resource base, introduction of improved farming methods and other agricultural and nonagricultural income generation opportunities and external support for initial investment would be the main strategies in any development effort of these rural communities. The importance of multi-purpose trees in strategic planning of this poorly managed production system can neither be ignored nor be confined to a specific farming component of the system. Thus, it is now high time to realize that the conceptual `MPTS approaches' discussed in this paper must be tried out in order to achieve the sustainability of this risk prone but high potential farming environment existing in the dry zone of Sri Lanka.

8

REFERENCES Dharmasena, P.B. (1986). Optimum utilization of the storage in village tanks. Trop. Agric. Vol. 145 : 1 - 11. Dharmasena, P.B. (1992). Soil erosion control measures for rainfed farming in the dry zone of Sri Lanka. Ph.D Thesis (unpublished), University of Peradeniya, Peradeniya. p 258. Dharmasena, P.B. (1993a). Water resource utilization in small tank watersheds. Paper presented at the symposium on `Tank-based Agriculture; Problems and Prospects' organized by SLAAS held on 20th August, 1993 at ISTI, Maha Illuppallama. Dharmasena, P.B. (1993b). Man-environment interaction in tank-village homegardens; the trends in vegetation. Proc. Fourth Regional Workshop on MPTS, kandy, Sri Lanka, March 12 - 14, 1993. p 90 -101. Leach, E.R. (1959). Hydraulic society in Ceylon. Past and Present, 15, 2 -26. Somasiri, S., J. Handawala, W.L. Weerakoon, P.B. Dharmasena and S.N. Jayawardena.(1990). Rainfed Upland Farming for the Dry Zone. Agro-technical Information Bulletin. Department of Agriculture, Peradeniya. Thilakasiri, S.L. (1986). Village irrigation. Economic Review 2 : 3 - 15. Weerakoon, W.L. (1992). Sustainability in rainfed upland agriculture in the dry zone of Sri Lanka. Proc. SLAAS Symposium on Rainfed Agriculture : p 63 - 66. Wijethunge, D.R. (1986). The Wew-sabha and people's participation in small irrigation systems. Proc. Workshop on Participatory Management in Sri Lanka's Irrigation Schemes - organized by IIMI. p 148-159.

9