Skip to main content

Assay on an almost optimal and simple selection method for balancing gain and genetic diveristy. The spm 3,2,1 strategy


Research objectives and content Our purpose is to assay a simpler selection method than those based on optimization techniques for balancing genetic gain and genetic diversity. It should be simple, robust and
breeding. This simpler sub-optimal scheme is based on weighted selection method, in which every family contribution in the next generation are assigned based on the family performance. Such a scheme, the so called Single Pair Mating 3,2,1 strategy, would have the following characteristics: I) Random allocation of mates, once mate per individual in order to obtain only full sibs families. 2) Division of breeding population into three family groups, one with the best, one with intermediary and one with the worst. From every family belonging to the first group are taken three offspring, two from every family in the second group and only one from the third group. 3) The number of families in the different groups will be chosen in such a manner that the proportion of the best families will be the same as that for the worst families, in order to keep constant the size of the breeding population along the generations. This new strategy will be compared with both an optimal selection algorithm and a simpler alternative with a still simpler alternative where all the contributions from the parents is constrained to a certain value (within family selection). This comparison will be made as a gain versus diversity diagram and will be study as a function of selection design variables. As genetic diversity measure will be used status number, fully developed in the proposed host institution. From the results it will be known how close to optimum 3,2,1 is and how this depends on the circumstances. Training content (objective, benefit and expected impact) By means of optimized selection methods it is possible to obtain the best solutions for specific schemes, but they require much time by highly trained people for implementation to every particular case. Given that, in many circumstances the required people is a severely limiting factor, as it happens in less-favoured regions, therefore, it would be necessary to develop simpler strategies, like the present proposal. Despite simpler methods give small and predictable loss of final profit, their possible mistakes are less likely to occur and more likely to be detected if they occur. This method should provide reasonable profits close to optimum, thus is, more genetic gain than the most conservative within family selection, and also better maintenance of genetic diversity than with random mating designs. Such a strategy would provide great benefit not only in developed breeding systems. but also especially in those underdeveloped ones in the less-favoured regions. Links with industry / industrial relevance (22)


Aristotle University of Thessaloniki
Analytical Chemistry Lab
54006 Thessaloniki