s:3281:"%T Modifying forestry and agroforestry to increase water productivity in the semi-arid tropics 
%A Ong C K 
%A Black C R 
%A Muthuri, C.W. 
%X The need to increase water productivity is a growing global concern as the World Commission onWater has estimated that demand for water will increase by c. 50% over the next 30 years and approximately half of the world’s population will experience conditions of severe water stress by2025. Three-quarters of African countries are expected to experience unstable water supplies, whereby small decreases in rainfall induce much larger reductions in streamflow. Vital water catchments have been lost or are being degraded, creating concerns about the loss of hydrological functions and increasing competition for scarce water resources between agriculture, urban centres, industry and wildlife. The challenge is to improve water productivity at the landscape or river basin level, especially for the rapidly growing populations in many developing countries. Water productivity is defined as the value or benefit derived from the use of water. In this review, we describe recent findings regarding the hydrology of forests and agroforestry systems and indicate how modifications to tree-based systems may increase water productivity in the semi-arid tropics. Throughout the tropics, reforestation using eucalyptus has been the most significant approach for modifying the water productivity of forestry and agroforestry systems. Fortunately, the ‘eucalyptus dilemma’ of providing local benefit at the expense of decreased stream flow for downstream users has been well articulated in many countries where evidence-based research has changed water policy and discouraged further planting of eucalyptus on water catchments. In East Africa, the most popular replacement for eucalyptus has been another exotic from Australia,Grevillea robusta, which has become the species of choice for farmers in the sub-humid highlands.However, attempts to introduce this species to semi-arid areas have been disappointing due to its evergreen canopy and consequently high demand for water. Current attempts to use deciduous species such as Melia volkensii and Paulownia fortunei in such areas are much more encouraging, although the long-term implications are still unknown. Agroforestry has the potential to improve water productivity in two ways. Trees can increase the quantity of water used on-farm for tree or crop transpiration and may also improve the productivity of the water that is used by increasing the biomass of trees or crops produced per unit of water used. Plot-level evidence shows that improvements in water productivity resulting from modifications to the microclimate experienced by crops may be limited. Instead, evidence from semi-arid areas of India and Kenya has shown that the greater productivity of agroforestry systems is primarily due to the greater quantity of water used. Further research is needed to examine the impact of the increased water use on drainage and base flow at the landscape level. Finally, we describe some of the technical approaches, which may be used to improve water productivity based on differences in tree phenology and the challenges facing smallholders in areas of growing water scarcity. 
";