INFLUENCE OF SHELTERBELT TYPE ON POTENTIAL EVAPORATION IN AN ARID ENVIRONMENT

The results from three studies in central Tunisia are presented, the f irst and second in November-December of two consecutive years and the third in April-May. The spatial variation in reductions of potential e vaporation behind the shelterbelts, measured with the Andersson evapor imeter, is reported. The types of shelterbelt systems were a mechanica l shelter of porous nylon netting on the open plain, a cactus Opuntia ficus indica shelter with very low porosity, a nursery with an Acacia cyanophylla shelter within a system of rows of tall Eucalyptus camaldu lensis situated 40 to >170 m from the shelter, and an orchard of dispe rsed young fruit trees sheltered by young plantations of Acacia cyanop hylla and Casuarina equisetifolia shelterbelts. The mechanical shelter reduced potential evaporation similarly in the three studies. The cac tus shelter was inefficient owing to its low porosity. In the nursery the tall eucalyptus trees reduced potential evaporation at distances > 20H (H is distance in terms of height of the shelter) behind the acaci a shelters. In the orchard the fruit trees had an effect by increasing the surface roughness and contributing by transpiration to altered ev aporative demands. Furthermore, seasonal differences in the relative r eductions of potential evaporation were found, especially in the orcha rd. The relative reduction in daily potential evaporation was strongly correlated to relative reduction in daily mean wind speed, principall y because the soil in all systems was kept bare. The Andersson evapori meter was shown to be practical and convenient for these types of stud ies.