lower costs than conventional programs and on their ability to produce higher genetic gains per year (e.g., Ceccarelli et al., 2001a, 2003; Smith et al., 2001; Witcombe et al., 2001; Virk et al., 2003, 2005). However, participatory research projects (composed of both formal and informal actors) have also led to the spread of socially responsible, technical innovations and important policy changes (Joshi et al., 2007). These innovations have been shown to improve the welfare of the poor and socially excluded. One of the best examples is a 1997 client-oriented participatory crop improvement (PCI) project in Nepal in which there was formal recognition that informal R&D processes were taking place,

and a move to encourage those processes (Biggs, 2006). This led to changes in National Varietal Release Procedures and to more effective collaboration between different actors. Informal developments were essentially legitimized and supported. Nevertheless, the benefits of farmer participation may not be universal, and adoption of participatory methods has not been as high as expected, notably because of methodological limitations to upscaling (Witcombe et al., 2005).

The quality issue. In developed countries, changes in the consumers' preferences have pushed the labeling of the