Importance of water consumption by perennial vegetation in irrigated areas of the humid tropics: evidence from Sri Lanka

In tropical, monsoon climates of South-East Asia, irrigation facilities supplement rain in the wet season and enable crops to be cultivated during the dry season. In the Dry Zone of Sri Lanka, 70% of the average annual rainfall of 1000 mm falls in a 3 month period. During the dry season, reference evapotranspiration has less rainfall - about 700 mm, indicating that much additional supply is meant to support crops, mainly paddy. In this climatic context, irrigation has dramatically changed the local environment, creating ecosystems quite similar to that of the wet zone to flourish. In these systems, recharge of shallow groundwater by percolation from irrigated fields, canals, and tanks, has provided a continuous supply of water for natural vegetation and homestead gardens. Much of the water used by this non-crop vegetation is beneficial. Growth of fruit and coconut trees can be quite profitable, while other trees enhance the environment. In 1998, IWMI performed a comprehensive water balance in the command area of the Kirindi Oya irrigation scheme, Sri Lanka, based on surface flow measurements, rainfall data, and estimation of crop water requirements. This water balance showed that evaporation consumed 78% of the total amount of water available for use. The amount of evaporation is split into process depletion (crops for 28%), direct evaporation from tanks (7%), inter-seasonal fallow (10%) and from non-crop vegetation for 55%. The main conclusion from this study is that perennial vegetation as the main component of noncrop vegetation, is a significant consideration in tropical humid environments in planning, management and performance assessment. Designers. managers, and researchers need to specifically incorporate the evaluation of evaporation by non-crop vegetation and perennialvegetation in their approach of water requirements. Further investigation is needed to estimate water consumption by land cover type to assess their respective beneficial use. (C) 2001 Elsevier Science B.V. All rights reserved.