Potential for aquaculture in community-managed irrigation systems of the dry-zone, Sri Lanka : impacts on livelihoods of the poor

Rainfed areas in the Dry-Zone of Sri Lanka are characteristic of extensive marginal agro-ecosystems known as the semi-arid tropics (SAT) populated by poor farming communities. In the Dry-Zone and elsewhere, the traditional response to seasonal water scarcity was to construct rainfall-harvesting devices known as ‘tanks’; created by building earthen dykes across ephemeral streams in undulating terrain. Most are held in common ownership by adjacent communities, who use them for multiple functions including irrigation, bathing and fishing. Storage efficiency is enhanced by arranging tanks in cascading sequence within watersheds so that drainage waters can be re-used. The aim of this study was to evolve improved collective strategies for the management of seasonal water bodies (focussing on aquatic production) in order to reduce the vulnerability of the poorest groups. Understanding of these complex systems requires a holistic approach which integrates hydrological, biological and socio-economic factors on a suitable (watershed) scale. Work commenced with a comprehensive situation analysis, culminating with the formulation of a participatory research agenda for action research based on low-input stocking enhancements. Village livelihoods have traditionally revolved around paddy cultivation as the primary tank function; however, in recent times, water-use strategies have responded to a range of demographic, economic and environmental pressures with implications for the sustainable management of natural resources, especially living aquatic organisms. Natural fish production in the most seasonal tanks relies on intermittent spill-events which link successive tanks; these provide migration routes which permit recruitment of stocks from lower perennial tanks. Rehabilitation initiatives that increase the storage / irrigation capacity of tanks or poorly designed surplus weirs that impede migration have negative impacts on fisheries, though they are rarely considered by planners. The fundamental concept of the purana complex (PC) as the smallest logical sub-component of the watershed for intervention is introduced. Within PC boundaries discrete community groups bound by ties of kinship and caste, control access to private and commonly held natural resources. PCs in the uppermost reaches of watersheds are distinguishable by the highly seasonal nature of their tanks and poor physical infrastructure relative to lower watershed communities. Such areas are also often buffer zones between as yet uninhabited hinterlands and settled areas where cultivation potentials are further restricted due to wild animal incursions. Consequently, these groups exhibit the greatest dependence on exploitation of the natural resource base. This often includes less seasonal tanks in lower PCs where fisheries are of less significance to local livelihoods. Such low-level ‘poaching’ is generally well tolerated, but potential for conflict exists where development efforts restrict hitherto free access to these resources. These findings were the basis for two phases of action research which involved the stocking of ten tanks belonging to seven communities in North West Province (2000-2001). Phase 1 trials encompassed a range of social and physical and settings from lower to upper watershed. Results indicated that the use of costly hatchery-produced seed was unlikely to be sustainable given (1) a background of highly erratic natural production (2) uncertain returns to individual effort and (3) a low priority accorded to fish production from village tanks given the availability of low-cost commercial production from perennial reservoirs. The second phase was restricted to low-caste communities in upper watershed areas and relied entirely on wild-fish stocks captured from perennial reservoirs lower in the watershed. Also emphasis was on intermittent ‘staggered’ harvesting using hook and line gears rather than the single intensive ‘collective harvests’ adopted in phase 1 trials. High yield potentials were demonstrated in the smallest tanks (