technique dating back 4,000–5,000 years. It is currently under revival in response to the escalating water scarcity (Falkenmark et al., 2001). Harvested precipitation, i.e., collected runoff water, may be either diverted directly to the cropped area during the rainfall event (“runoff farming”) or may be collected for irrigation or other purposes such as domestic use or livestock watering (Oweis et al., 1999). Runoff farming and using stored water for irrigation may be practiced at micro, meso and macro scale, and numerous technologies have been developed according to specific environmental and sociocultural conditions (see Critchley et al., 1991; Agarwal and Narain, 1997; Prinz et al., 2000; Prinz, 2002; Prinz and Malik, 2002; Mahnot et al., 2003; Oweis et al., 2004; CSE, 2006). Other water-harvesting techniques include floodwater harvesting, fog and dew harvesting, and groundwater harvesting by qanats, underground dams or special wells.

Water harvesting may not only tap unused water resources and thereby increase crop productivity and minimize the risk of crop failure in dry areas, it may also allow producing crops in environments where cropping is not feasible without such technologies. Furthermore, water harvesting may facilitate forestation or reforestation, fruit tree planting or agroforestry, and protect land from degradation and desertification (Prinz, 2002). Groundwater recharge for more sustainable use by different sectors represents another important benefit of water harvesting.

However, it is not clear if widespread use of water-harvesting technologies is achievable, since construction and maintenance costs, particularly the labor costs, are generally important. Furthermore, many water-harvesting projects require collective action at the community or watershed level, and land lost for catchment areas represents an opportunity cost that may deter small-scale farmers in land-scarce areas from adopting water-harvesting technologies (Rosegrant et al., 2002). Since certain technologies may require inputs that are too expensive for some farmers to supply, some intervention of state authorities may be needed (Prinz, 2002).

As rainwater harvesting should be an integral component of a farming system, a systems approach has to be followed and water-harvesting technologies should be combined with other improved management practices such as adequate fertilization, pest management, improved varieties, crop rotations, and efficient irrigation techniques. Applying remote sensing data and hydrological models at the watershed level may not only facilitate the identification of suitable water-harvesting sites and technologies but also help prevent problems between upstream and downstream water users and allow for supplying sufficient quantities of water for natural flora and fauna.

Since water harvesting operates at both the household or farm scale and the community or watershed scale, farming systems research must consider institutional and land-tenure issues, in which traditional and formal institutions may play a crucial role. Little research has been carried out in this respect, and thus AKST still faces important biophysical and socioeconomic knowledge gaps with regard to water harvesting. Extension and irrigation staff require more knowledge about water-harvesting techniques and the associated socioeconomic implications to achieve the potential gains in crop yields from water harvesting in combination with

supplemental irrigation and improved farm management practices (Falkenmark et al., 2001). Since water-harvesting technologies originated in CWANA, a wealth of indigenous knowledge exists in the region that can be used to develop new practices and improve the efficiency of systems still in use today. For widespread adoption of such technologies, however, land-tenure systems will have to accommodate ownership or long-term use rights, so that farmers will be willing to invest in water-harvesting systems. Policies should encourage the required inputs for construction and maintenance.

Use of unconventional water resources. Rather than seeking pristine new water sources, a wide range of alternative water supplies will increasingly be used to meet demands. Reclaimed water, gray water, fog collection, recycled water, brackish water, saltwater, or desalinated water may all be considered usable for particular needs, and in fact may have environmental, economic or political advantages. Reclaimed water such as treated wastewater can be used to recharge groundwater aquifers, supply industry, irrigate certain crops, or augment potable supplies (Gleick, 2000).

However, using unconventional water resources may pose its own problems. Treated wastewater used in agriculture might entail health hazards and water-quality problems, requiring regulations regarding its treatment and reuse. Such regulations will particularly have to cover the responsibility of water polluters in treating their wastewater to make it safe to use (e.g., in agriculture) or to discharge into the environment. More training for farmers, water users and crop consumers will be needed to address issues related to health and water quality aspects.

Groundwater recharge. Groundwater resources are being overexploited in most CWANA countries (FAO, 1997; Aquastat, 2006). This is due to overpumping and also to reduced recharge related to diminishing infiltration rates caused by expansion of urban areas, inadequate land management, and climatic changes (Morris et al., 2003). Watertable elevations are dropping and seawater intrusion is becoming a common problem in many CWANA countries.

Maintaining and increasing aquifer recharge may counterbalance increased exploitation to a certain extent. Foremost, it is important to enhance natural recharge by adequate land management, i.e., by reducing runoff of precipitation. This may not only increase aquifer recharge but will also allow storing a greater part of the scarce precipitation in the soil for crops to use and will reduce erosion. The high evaporation rates in the CWANA region make groundwater storage particularly advantageous (UNEP/IETC, 2001).

Artificial recharge of groundwater aquifers is also aviable option. Artificial recharge can be achieved through surface spreading and preventing runoff or by direct well injection into the groundwater. Sources of recharge water may include precipitation and storm runoff, trapped from cultivated and uncultivated land, from urban areas, surface water, leaks in water supply systems, overirrigation, or treated wastewater (Morris et al., 2003; NEIWPCC, 2003). Artificial recharge may require temporary storage structures and water treatment in sedimentation tanks to improve water quality, particularly for treated wastewater. As a rule the quality of water recharged into an aquifer should be at least