tourism has vigorously helped to promote them. In South Tunisia for example, several water-harvesting techniques given up in the 1960s have been reestablished to produce fresh fruits and vegetables for hotels and ecotourism.

In poorer regions, the productivity of land and water in rainfed areas is greatly enhanced by harvesting water. Marginal lands with annual rainfall of less than 300 mm a year can be cultivated if limited but controlled additional water is made available (Oweis et al., 1999; Rodriguez, 1996; Rodriguez et al., 1996). In many instances, appropriate water-harvesting techniques can provide an incremental water supply.

Modernization and diffusion of these ancient technologies have to be sought to increase agricultural productivity and provide a sustained economic base. The choice of technique depends on the rainfall and its distribution, land topography, soil type and depth and local socioeconomic factors, so these systems tend to be very site specific. The water-harvesting methods strongly depend on local conditions and include widely differing practices-bunding, pitting, microcatchments and harvesting flood- and groundwater (Critchley et al., 1992; Prinz, 1996).

Rainwater harvesting areas are not well mapped and few statistics are available nationally or regionally. Several experiences are quoted in specialized literature but little information is given about their importance in concerned areas, benefits to people and economic return. AQUASTAT FAO databases for CWANA are available only for Egypt (133,000 ha), Iran (40,000 ha), Lebanon (500 ha) and Tunisia (898,000 ha).

Work in Tunisia may be divided broadly into two types: description and rehabilitation of indigenous systems and the large-scale technical development program of the Departement de la Conservation des Eaux et du Sol. This program is one of the few in CWANA that integrates soil and water conservation into hydrological priorities (Selmi, 1994). As well as constructing bunds and terraces for conservation, it includes building small dams on watercourses high in the catchments of major rivers. Purposes include flood control, recharge of shallow groundwater for irrigation and reduction of siltation of major dams supplying domestic and industrial needs. Among the spinoffs is that hill farmers will have water from small dams to use in supplemental irrigation. This program seeks to conserve soil and water by focusing primarily on engineering works. Unfortunately, the socioeconomic problems and different options for land users are essentially neglected.

Indigenous systems in Tunisia have recently been described in two monographs, by Ennabli (1993) and Alaya et al. (1993). The former provides detailed descriptions of nearly 30 traditional systems for capturing and using water in the dry areas of Tunisia. The water interception, concentration, conveyance and storage techniques reported (many still in use) illustrate the wealth of ingenuity in human adaptation to dry environments. Alaya et al. (1993) focus on tabia, the earthen bunds widely and variously used in Tunisia to intercept and redirect runoff water to crops and trees. Though primarily an implementation manual, this book is also rich in descriptions of traditional practices.

The meskat system, using tabias to support olive plantations, covers about 300,000 ha in central Tunisia (Prinz,

1994). It comprises catchments of about 500 m2 surrounded by tabia and spillways to control runoff flow into bunded plots of trees. This is a successful system, but according to Reij et al. (1988) it suffers heavily from increasing land pressure, resulting in a decrease of catchment areas and leading to lower efficiency.

The jessour system is based upon cultivating sediments built up behind large tabia, often stone-reinforced and with stone spillways, constructed in water cascading down narrow mountain valleys in southern Tunisia. Akrimi et al. (1993), from the Institut des Regions Arides (IRA) near Medinine, reported a multidisciplinary study involving jessour cultivators in the Matmata Mountains. Maintaining the tabia and spillways is a major problem in some areas, due partly to the degree of outmigration. It is Tunisian government policy to assist with jessour rehabilitation. Proposals for further research by the same IRA team note the launching of major development schemes for soil conservation and rainwater harvesting, but comment that community participation has been weak because the schemes have failed to take account of local traditions and existing production systems.

Development schemes in Jordan involve building earth dams that divert runoff to improve pastures and bunds to conserve soil and moisture on steep land. Research was started in 1987 by the University of Jordan to explore the development potential, particularly the water-harvesting potential of a 70-km2 catchment under low rainfall (100-250 mm per annum) east of Amman (Taimeh, 1988). Irrigation from wadi flows trapped by earthen dams and microplots supporting fruit trees are two techniques that have shown promise, both socioeconomically and technically. Currently data collected on this catchment are used to develop a coupled prediction-optimization model for harvesting, storing and using water in similar dry areas of Jordan and elsewhere (Sarraf and Taimeh, 1994).

Other ongoing regional activities include a relatively large development project, with an included research component, in a steppe in southern Syria using the integrated management of soil, water and vegetation (Rashed, 1993). The project uses water supplied by various harvesting techniques and a limited groundwater supply to enhance production, particularly of forage crops and shrubs.

In Yemen, a major research focus is to conserve the ancient terrace system, parts of which have fallen into disrepair following socioeconomic changes. The terraces are not just to conserve soil and water but also to control water, including harvesting water for human consumption, flowing from the high, often degraded pasture lands and to protect the lowlying intensively cultivated banks of the main wadis and the flood irrigation systems. A new multidisciplinary project with a participatory approach addresses the socioeconomic, institutional and policy issues that are involved (Mouhred, 1994).

The rainfed coastal areas of Egypt have received considerable R&D attention over recent decades. Initially the aim was to settle the Bedouin population. Projects were undertaken to rehabilitate degraded rangeland and increase use of runoff, through terracing wadis, similar to Tunisian jessours and enhancing indigenous runoff farming systems (Perrier, 1986). More recently, the coastal areas have come to be seen as another small but potentially productive national agricultural resource and emphasis has shifted toward more