et al., 2002), are more viable options. Soil erosion control can be costly and hence difficult to implement in developing countries (Wheaton and Monke, 2001). Governments can help by providing technical advice, economic incentives and public education programs (Warkentin, 2001). Land care schemes have been successfully adopted in several countries, and are effective in promoting "land literacy" and good ag­ricultural practices, including leys and crop rotations and growing cover crops (Lal, 2001).

6.6.2.2 Options for low fertility lands
Agroforestry. In tropical areas, low fertility is often found in deforested areas, where critical topsoil has washed away. The replacement of traditional slash and burn cultivation by more diversified production systems based on forest prod­ucts, orchard products, and forages and food products (Bar­rett et al., 2001; Ponsioen et al., 2006; Smaling and Dixon, 2006) and applying agroecological principles creatively (Al-tieri, 2002; Dalgaard et al., 2003) can improve soil fertility.
       The adoption of agroforestry can maintain land pro­ductivity, decrease land degradation and improve rural peo­ple's livelihood (Albrecht and Kandji, 2003; Oelberman et al., 2004; Schroth et al., 2004; Reyes et al., 2005; Jiambo, 2006; Rasul and Thapa, 2006). At the landscape scale, the spatial organization of tree and forest landscape elements can provide filters for overland flow of water and sediments as well as corridors for forest biota, connecting areas with more specific conservation functions. 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 (Van Noordwijk et al., 2007).
        Consecutive nutrient exports may lead to extremely low K and P levels (Alfaia et al., 2004), e.g., decreased N and P availability with alley cropping (Radersma et al., 2004). Some crops, e.g., sugarcane (Saccharum officinarum) seem to be unsuitable for agroforestry (Pinto et al., 2005). Eco­logical agriculture could become an alternative if market distortions created by subsidies were removed, financial benefits were provided to resource-conserving farmers, and extension, credit, research were available (Rasul and Thapa, 2003). The adoption of integrated soil fertility management strategies at the farm and landscape scale requires consensus building activities (Barrios et al., 2006). However, promot­ing and supporting participatory technologies have limited impact when they are not grounded in participatory policy development and implementation (Desbiez et al., 2004; De Jager, 2005). Labor-intenseive ecoagriculture will not suc­ceed unless farmers and the agricultural sector have higher total factor productivity including total labor productivity (Buck et al., 2004).

Soil water conservation and storage. The adoption of con­servation agriculture is key to increasing water storage in marginal lands, and in most places suitable equipment is available (hand, animal-drawn, or tractor-drawn) for re­source-poor farmers (Bolliger et al., 2006). Adoption of conservation agriculture also reduces soil erosion losses, (den Biggelaar et al., 2003) decreases siltation and pollu­tion of water bodies, and has benefits for human health and

biodiversity. Efforts to promote soil water conservation and storage will need to address site-specific conditions (Knowler and Bradshaw, 2007). Widespread implementation will re­quire integration into institutions, incentive structures, and education (Molden et al., 2007) and extension outreach.
         Methods to be considered include (1) conservation agri­culture, including the use of water-efficient crops; (2) supple­mental irrigation in rainfed areas; and (3) water harvesting in drier environments (Goel and Kumar, 2005; Hatibu et al., 2006; Oweis and Hachum, 2006).

Soil amendments. Municipal waste materials, composted or uncomposted (such as leaves and grass clippings, sludges, etc.), can be valuable soil amendments for farms near cit­ies or towns and are inexpensive if transport costs are low (Smith 1996; Kashmanian et al., 2000). Municipal sludges can be also applied to cropland provided they possess the qualities needed by their potential users and do not possess toxins or heavy metals, such as nickel or cadmium (Smith, 1996). Other developments such as N-fixation by non-le­gume crops (e.g., Azospyrilllum), P solubilizing bacteria, and mycorrhizal associations in tropical cropping systems are expected to result from future biotechnology investiga­tions (Cardoso and Kuyper, 2006).
        The high risk of crop failure from insufficient soil mois­ture hinders investments in soil fertility and tilth, which in turn diminishes the potential of soils to capture and retain water, therefore increasing the vulnerability to drought. A challenge for AKST will therefore be how to couple incre­mental improvements in crop water relations with low-cost investments to replenish soil fertility in order to break this cycle (Rockström, 2004; Sanchez, 2005). More widespread use of practices like green manuring, composting, farmyard manure management, and use of agricultural by-products and residues can guide decision-making.

6.6.3 Sustainable use of water resources to meet on-farm food and fiber demands
A major challenge over the next 50 years will be to meet food and fiber demand with minimal increases in the amount of water diverted to agriculture. Aquatic ecosystems and peo­ple whose livelihoods depend on them are likely to be the biggest losers as more and more fresh water is diverted to agriculture on a global scale.
        AKST can provide options for improving water man­agement in agriculture that can address the growing prob­lem of water scarcity, ecosystem sustainability and poverty alleviation. Chapters 4 and 5 present projections concerning the land and water required at the global level to produce enough food to feed the world in 2050. These include reli­ance on various options including intensification and expan­sion of rainfed and irrigated agriculture and trade as entry points to reduce the need to expand water and land diverted to agricultural production. In an optimistic rainfed scenario, reaching 80% maximum obtainable yields, while relying on minimal increases in irrigated production, the total cropped area would have to increase by 7%, and the total increase in water use would be 30%, with direct water withdrawals increasing by only 19%. In contrast, focusing on irrigation first could contribute 55% of the total value of food supply by 2050. But that expansion of irrigation would require 40%