ABSTRACT
Client-oriented breeding (COB) uses participatory methods to increase client engagement. To target the rice-breeding programme precisely and more easily involve farmers, we used a few crosses that allowed large segregating populations. This ‘smart-cross’ approach, which applies to participatory as well as conventional breeding programmes, predicts parental value from information on, e.g. disease resistance and adaptation to stress. Often one parent is a farmers’ preferred variety and the other is chosen for its complementarity. Adopting a few-cross approach enables improved client orientation and greater farmer participation. We extended existing models on the optimal number of crosses by assuming that the probability of success of crosses declines as more are made. Fewer crosses always proved more efficient – in rice breeding in India and Nepal, we obtained very high success rates with few crosses, in terms of the number of varieties released officially or preferred by farmers. However, institutionalization of low volume crossing requires the replacement of high-volume crossing. We discuss how the latter is a social construct that makes it hard to change. Lack of institutionalization is a major limitation in the sustainability of COB, along with constraints to seed supply that must rely on project interventions when the formal seed-supply system is not effective.
