success, the CGIAR has estimated that 30 years of agricultural research on seven major crops and three livestock products has improved yield gains so much that, had this gain not occurred, an additional 170-340 million ha of forests and grasslands would have been needed for production (Nelson and Maredia, 1999; FAO, 2003a). Other estimates of forestalled conversion of habitat to agricultural use are as high as 970 million ha (Golkany, 1999). A cost/benefit analysis by ACIAR (Raitzer and Lindner, 2005) found that research projects involving forestry/agroforestry had the greatest benefits (42.9%). Increases in total factor productivity, which contribute to increased output, are always associated with investment in research (Pingali and Heisey, 1999; McNeely and Scherr, 2003). These studies pay less attention to the social and institutional distribution of impacts or to noneconomic benefits.

3.2.3.1.2 Assessment of livelihood impacts of AKST Systematic and detailed impact assessments of AKST's contribution to livelihood improvement and the sustainability of livelihoods over time are generally lacking. A livelihood is said to be sustainable "when it can cope with and recover from shocks and maintain or enhance its capabilities and assets both now and in the future, while not undermining the natural resource base" (Carney, 1998). Indirect impacts of AKST in relation to ownership of assets, employment on and off farm, vulnerability, gender roles, labor requirements, food prices, nutrition and capacity for collective action have been less thoroughly researched than the financial and economic impacts (Hazell and Haddad, 2001; Meinzen-Dick et al., 2004), although recent impact assessments of Participatory Methods have more comprehensively addressed these issues. Comparative case studies of livelihood change incorporating qualitative dimensions and complementing other methods have begun to document the noneconomic impacts of AKST. (www.prgaprogram.org/ modules.php?op=modload&name=Web_Links&file=index &req=viewlink&cid=133&min=0&orderby=titleA& show=10).

Livelihoods approaches have usefully contributed to conceptual and methodological innovations.

The concept of "sustainable livelihoods" is both an AKST product and a tool, which facilitates the analysis of livelihood status and changes and the understanding of ex ante and ex post impacts. The livelihoods framework considers livelihoods as comprising the capabilities, assets and activities required for a means of living. This is a broader and more holistic view than just equating "livelihood" with income or employment (Booth et al., 1998). It links the notion of sustaining the means of living with the principle of environmental sustainability (Carney, 1998).

     The elements of the livelihoods framework include the assets that people use and combine to make a living, the factors which cause vulnerability; the policies, institutions and processes which affect the environment for livelihoods; the livelihood strategies followed and the outcomes. The livelihoods framework has been used to assist situational

analysis for research and development planning and to assess specific institutional, policy and technology and rural development options prior to intervention (Ashley and Carney, 1999; Shackleton et al., 2003; OECD, 2006b). More recently it has been used to assist evaluation of outcomes and impacts (Ashley and Hussein, 2000; Adato and Meinzen-Dick, 2003; Meinzen-Dick et al., 2004; Adato and Meinzen-Dick, 2007) and has complemented more macrolevel economic impact assessments. Livelihoods analysis has been further assisted by the development and refinement of participatory tools for poverty and situational analysis, especially in the context of improving client orientation and gender relevance of agricultural research and development (World Bank, 1998). Recently, the framework has helped to identify principles and processes critical to achieving sustainable livelihoods, and to understand the complexities associated with partnerships to promote local empowerment, resiliency and diversification (Butler and Mazur, 2007). Its limitations include the absence of integration of dimensions of power, the unspecified nature of "institutions and processes" which require further elaboration of knowledge, culture and innovation and the need for further tools to understand the dynamics of livelihood changes.

3.2.3.2 The contribution of AKST to livelihoods improvement

The improvement of livelihoods depends on the accessibility of the products of AKST. This depends on the factors influencing uptake, the distribution of benefits of specific technologies and their impacts. Particular attention is paid to impacts on overall levels of poverty and economic status, human health; natural and physical assets, social relationships, and vulnerability.

3.2.3.2.1 AKST and poverty

Some gains have been made in the reduction of poverty, but the contribution of AKST to increasing agricultural production and agriculture based incomes has been very different in different regions, agroecologies and for different groups of people.

AKST and agricultural transformation have had an important influence on the economic and social situation of many countries. Poverty is a serious global problem with 3 billion people (2.1 billion are rural poor) earning less than the purchasing power equivalent of US$2/day. The impacts of AKST are location specific and depend on complex interacting factors. Between 1990 and 2002, the proportion of people living in extreme poverty fell more rapidly in much of Asia compared with Africa, Latin America and the Caribbean (UN, 2006a); while in central and eastern Europe, the poverty rates increased. In sub-Saharan Africa, although there was a small decline in the rate of poverty, the number of people living in extreme poverty increased by 140 million. Poor countries (especially in SSA) have gained proportionately less than some richer countries (USA and Europe). Similarly, major benefits have escaped marginal agroecological regions (rain-fed dryland areas) and marginalized