aquaculture systems or integrated crop/livestock systems. While the greatest potential increases in yields and water productivity are in rainfed areas in developing countries, where many of the world's poorest rural people live, equally important is improved management of large dams and ir­rigation systems to maintain aquatic ecosystems.

9. The potential benefits and risks of bioenergy are strongly dependent on particular local circumstances.
Research is needed on better understanding these effects and improving technologies. Expansion of biofuel production from agricultural crops (1st generation) may in certain cases promote incomes and job creation, but negative effects on poverty (e.g., rising food prices, marginalization of small-scale farmers) and the environment (e.g., water depletion, deforestation) may outweigh these benefits and thus need to be carefully assessed. Small-scale biofuels and bio-oils could offer livelihood opportunities, especially in remote regions and countries where high transport costs impede agricultural trade and energy imports. There is also considerable poten­tial for expanding the use of digesters (e.g., from livestock manure), gasifiers and direct combustion devices to generate electricity, especially in off-grid areas and in cogeneration mode on site of biomass wastes generating industries (e.g., rice, sugar and paper mills). The next generation of liquid biofuels (cellulosic ethanol and biomass-to-liquids technol­ogies) holds promise to mitigate many of the concerns about 1st generation biofuels but it is not clear when these tech­nologies may become commercially available. Moreover, considerable capital costs, large economies of scale, a high degree of technological sophistication and intellectual prop­erty rights issues make it unlikely that these technologies will be adopted widely in many developing countries in the next decades. Research and investments are needed to ac­celerate the development of these technologies and explore their potential and risks in developing countries.

6.1 Improving Productivity and Sustainability of Crop Systems

6.1.1 Small-scale, diversified farming systems
Considerable potential exists to improve livelihoods and reduce the environmental impacts of farming by applying existing AKST in smarter ways to optimize cropping and livestock systems, especially in developing countries.      Small-scale diversified farming is responsible for the lion's share of agriculture globally. While productivity in­creases may be achieved faster in high input, large scale, specialized farming systems, greatest scope for improving livelihood and equity exist in small-scale, diversified pro­duction systems in developing countries. This small-scale farming sector is highly dynamic, and has been responding readily to changes in natural and socioeconomic circum­stances through shifts in their production portfolio, and spe­cifically to increased demand by increasing aggregate farm output (Toumlin and Guèye, 2003).      Small-scale farmers maximize return on land, make efficient decisions, innovate continuously and cause less damage to the environment than large farms (Ashley and Maxwell, 2001). Yet they have lower labor productivity and are less efficient in procuring inputs and in marketing, es-

pecially in the face of new requirements regarding produce quality. Land productivity of small-scale farms was found to be considerably higher than in large ones in a comparison across six low-income countries (IFAD, 2001).      AKST investments in small-scale, diversified farming have the potential to address poverty and equity (especially if emphasis is put on income-generation, value-adding and participation in value chains), improve nutrition (both in terms of quantity and quality through a diversified produc­tion portfolio) and conserve agrobiodiversity. In small-scale farming, AKST can build on rich local knowledge. Un­derstanding the agroecology of these systems will be key to optimizing them. The challenges will be to: (1) to come up with innovations that are both economically viable and ecologically sustainable (that conserve the natural resource base of agricultural and non-agricultural ecosystems); (2) develop affordable approaches that integrate local, farmer-based innovation systems with formal research; (3) respond to social changes such as the feminization of agriculture and the reduction of the agricultural work force in general by pandemics and the exodus of the young with the profound implications for decision making and labor availability. Small-scale farming is increasingly becoming a part-time ac­tivity, as households diversify into off-farm activities (Ash­ley and Maxwell, 2001) and AKST will be more efficient, if this is taken into account when developing technologies and strategies for this target group.

6.1.1.1 Research options for improved productivity
To solve the complex, interlinked problems of small farmers in diverse circumstances, researchers will have to make each time a conscious effort to develop a range of options. There will be hardly any "one-size-fits-all" solutions (Franzel et al., 2004; Stoop and Hart, 2006). It is questionable if AKST will have the capacity to respond to the multiple needs of small-scale diversified farming systems (Table 6-1, 6-2).      AKST options that combine short-term productivity ben­efits for farmers with long-term preservation of the resource base for agriculture (Douthwaite et al., 2002; Welches and Cherrett, 2002) are likely to be most successful. In small-scale, diversified farming systems, suitable technologies are typically highly site-specific (Stoop and Hart, 2006) and sys­tems improvements need to be developed locally, in response to diverse contexts.

Integrated, multifactor innovations. In the past, a distinc­tion was made between stepwise improvements of indi­vidual elements of farming systems and "new farming sys­tems design". Stepwise improvement has had more impact (Mettrick, 1993), as it can easily build on local knowledge. Recently, successful innovations of a more complex nature were developed, often by farming communities or with strong involvement of farmers. Examples include success cases of Integrated Pest Management (see 6.4.3) as well al­ternative ways of land management such as the herbicide-based no-till systems of South America (Ekboir, 2003), the mechanized chop-and-mulch system in Brazil (Denich et al., 2004) or the Quesungual slash-and-mulch systems in Hon­duras (FAO, 2005).      In the future, research addressing single problems will probably become less relevant, as the respective opportuni-