406 | IAASTD Global Report

elements can provide filters for overland flow of water and sediments and corridors for forest biota, connecting areas with more specific conservation functions (Van Noordwijk et al., 2007). At plot and regional scales, the relationship is more variable because watershed functions not only depend on plot-level land use but also on the spatial organization of trees in a landscape, infiltration, dry-season flow, and other factors. Natural disturbance has a role in maintaining land­scape biodiversity. Options for conserving biodiversity in irrigated agricultural systems include increasing water pro­ductivity and many water management designs and prac­tices that support diverse landscapes, crops and connectiv­ity for plant and animal movement (Molden and Tharme, 2004).
       Traditional irrigation infrastructure development is one avenue for poverty alleviation; significant benefits have been demonstrated through a variety of primary and secondary effects of irrigation system development (Hussain, 2005; Castillo et al., 2007) and management strategies can im­prove equity in irrigation systems and can be complimen­tary to productivity enhancement (Hussain, 2005). As an example, land distribution that results in larger numbers of smaller holding can improve benefit sharing. Appropriate irrigation service charges can ensure adequate spending on operations and maintenance; this supports the poor, who tend to suffer the most when system level maintenance is inadequate. Management and financing options
In order to maintain aquatic ecosystems, managers  are increasingly pressed to maintain agricultural returns with reduced water delivery to irrigation systems. Reducing wa­ter delivered to irrigation requires two actions—a change in agricultural practice combined with a change in water allocation (Molden et al., 2007). Increasing blue water pro­ductivity by reducing water deliveries to agriculture, yet maintaining output, is an important strategy to retain water in aquatic ecosystems, to reallocate supplies, and to help in more precise water management, giving water managers more flexibility to deliver water to where it is needed, when it is needed. Excessive deliveries generate excessive drainage that are hard to control, require energy for pumping, reduce the quality of water and water bodies can provide breeding ground for disease vectors. Moreover, there are high eco­logical benefits in keeping water in rivers.
      There are significant opportunities to improve irrigation water productivity through a combination of field and sys­tem management practices, and policy incentives that raise water productivity, manage salinity and increase yields (e.g., Van Dam et al., 2006). For example, there is substantial scope to reduce water deliveries to irrigation, especially to rice (Bouman et al., 2007). In addition to producing more food, there are ample opportunities in irrigation to generate more value and incur less social and environmental costs.
        Supplemental irrigation, the addition of small amounts of water optimally timed to supplement rain, is probably the best way to increase water productivity of supplies. In Burkina Faso and Kenya, yields were increased from 0.5 to  1.5-2.0 tonnes ha-1 with supplemental irrigation and soil fertility management (Rockström et al., 2003). Yields


can be further increased with deficit irrigation, where wa­ter supplied is less than crop requirements (Zhang, 2003). Increased precision in water management is more capital intensive and therefore particularly relevant to maintain­ing high productivity while decreasing water diversions. In Western Syria, yields increased from 2 to 5 tonnes ha-1 with the timely application of 100 to 200 mm of water (Oweis et al., 2003). It must be noted, however, that precision and deficit irrigation increase risk, and therefore are most ap­propriate under conditions where access to water is assured, and can be carefully managed.
        A key point however, is that increasing productivity of water does not necessarily drive water savings; it may en­courage increased water use because it is more productive (Ahmed et al., 2007). Thus changing allocation policies is also essential to realize reduced diversions of water.
         Reducing deliveries also does not necessarily save water and can have unintended detrimental side effects that can be understood by considering what happens to drainage flows. A common misperception is that because irrigation is typically 40 to 50% efficient at converting irrigation water into evapotranspiration, the focus should be on increasing efficiency and therefore reducing drainage flows (Seckler et al., 2003). Increasing efficiency can be a valuable objective for reducing uptake of water in the system and thus dimin­ishing energy costs of pumping and operation and mainte­nance. However, drainage water plays an important role. Because so much drainage flow is reused downstream, there is actually much less scope in saving water in irrigation than commonly perceived. In fact, in irrigated regions in dry ar­eas it is common to document ratios of evapotranspiration to irrigation plus rain greater than 60% reaching to over 100% when aquifers are mined. These areas include the Gediz basin in Turkey (Droogers and Kite, 1999), Egypt's Nile (Keller and Keller, 1995), Chistian sub-division in Paki­stan and the Bhakra irrigation system (Molden et al., 2000), the Liu Yuan Ku irrigation system (Khan et al, 2006), the Tunuyuan irrigated area in Argentina, the Fayoum in Egypt, and Nilo Coelho in Brazil (Bos, 2004). The perennial veg­etation at Kirindi Oya has been shown to evapotranspire about the same volume of water as rice and generate valu­able ecosystem services; giving a different picture (65% of inflows beneficially depleted) than if paddy rice were consid­ered alone (22% of inflows depleted by rice) (Renaud et al., 2001). In these cases, the problem is not wastage, but that high withdrawals and ET rate reduce drainage and tend to dry up rivers and wetlands, and leave little to downstream use. It is important to consider each case from a basin per­spective, i.e., considering the quality and quantity of water and how drainage flows are used downstream.
       Technologies such as treadle pumps, small diesel pumps, low-cost drip, and low-cost water storage can increase pro­ductivity and incomes for poor farmers (Sauder, 1992; Shah et. al., 2000; Keller et al., 2001; Polak et al., 2004). These approaches provide water at lower unit costs than large scale hydraulic infrastructure, and can be available immediately, without the long delay times of larger scale projects. Inno­vative development and marketing approaches that focus on increasing local private enterprise capacities and market promotion have been credited with successful dissemination