That may hold globally, but most of the world's poor, especially in Africa, produce food under rainfed conditions (see Table 3.3 in Chapter 3). Much more can be done to improve water use in arid and semi-arid regions. Water use efficiency under these conditions could be 3 to 4 times higher than current values (Bindraban et al., 1999). These assessments are confirmed by field observations. Comparable gaps of three-to four-fold between actual and attainable yields have been reported from semi-arid regions in the Mediterranean (French and Schultz, 1984); Eastern Africa (Smaling et al., 1992); Sub-Saharan Africa (Rockstrom, 2001); Sahel (Breman et al., 2001); Southern India (Ahlawat and Rana, 1998) and Western China (Li et al., 2001).
There is increasing evidence from Asia that research and development (R&D) investments in rainfed areas offer win-win outcomes, in terms of both productivity growth and reductions in poverty, far in excess of similar investments in irrigated agriculture (Fan et al, 2000a; 2000b). Yield gaps in rainfed areas are often higher than in irrigated areas, and hence the return to further research and development and infrastructure investments can be higher. While irrigated areas have traditionally had higher adoption rates of modern varieties of crops than rainfed areas, there is accumulating evidence that rainfed areas in Sub-Saharan Africa have average adoption rates that are now approaching those of irrigated areas in Asia in the 1980s (Evenson and Gollin, 2001).
Drought risk, however, impedes investments, causing production to stagnate at subsistence levels with low water-use efficiencies. Climate change is expected to further exacerbate these risks. Resolving water scarcity problems requires an integrated water resource management approach (Box 4.7). The understanding of water cycles and related linkages between societal sectors is weak. Conflicting goals remain unresolved and fundamental trade-offs are not made explicit. The conventional, compartmentalized supply-oriented approach cannot cope with aspects of linkages between water, land-use and ecosystem demand in the context of socio-economic development and environmental sustainability (Ait Kadi, 2004).
A supply management strategy and a more rigorous demand management strategy (involving comprehensive reforms and actions to better use existing supplies) are both needed to avert water scarcity that impedes agricultural development. The sustainable use of water resources calls for an enabling political, legal and institutional environment to transcend traditional boundaries between sectors and involve a variety of users and stakeholders using a catchment approach. With agriculture being by far the largest user of water, improving water-use efficiency will remain a key dimension in resolving water scarcity problems. Issues of poor utilization, deteriorating quality and shortages can be addressed, and cross-boundary issues should be resolved.
