Choice of parents , number of crosses , and population size for participatory and classical plant breeding

Variability in breeding programmes for self-fertilised crops is created by crossing, and if the value of the crosses cannot be predicted in advance then this uncertainty means that a strategy of making many crosses is appropriate. However, since there is a limit to the capacity of a breeding programme, the more crosses that are made the smaller the size of each cross (F2 population size or numbers of progeny) and the lower the probability of recovering good genotypes from a favourable cross. Published theory on the optimum number of crosses in a plant breeding programme usually assumes that the value of all crosses are equal and hence overestimates the number of crosses required in a breeding programme. When the optimum size of a population in a favourable cross is considered theory predicts that huge populations are required. If linkage between traits is also considered this population size is even larger. Hence, in an inbreeding crop, one possible strategy is to select a small number of crosses that are considered favourable and produce large populations from them to increase the probability of recovering superior genotypes. This strategy is ideally suited to the particular constraints and advantages of participatory plant breeding (PPB). Such an approach although not essential in classical breeding (CB) may still be advantageous. When a breeding programme is based on few crosses, the choice of parents is critical and farmer participatory methods are highly effective in narrowing the choice. In the participatory plant breeding of self pollinated crops using a lowcross-number approach modified bulk population breeding methods, and recurrent selection are desirable strategies.