Review and Analysis of Limitations in Ways to Improve Conventional Potato Breeding

A number of improvements to conventional potato breeding are possible but they all have their limitations which need to be appreciated in designing new breeding programmes. Selection for quantitative traits between crosses before selection within the most promising ones can solve the problem of the inability of intense early-generation selection to affect most economically important traits. It also allows full-sib family selection to be practised but the rate of progress is limited by a low intensity of selection so that continued recurrent selection is required to accumulate small improvements into worthwhile ones. The rate of progress from combined between and within cross selection is limited by the number of vegetative generations required to complete all necessary phenotypic assessments, as well as by the intensities of selection. Both problems could be solved by genomic selection provided adequate selection accuracy can be achieved. However, lack of accurate large-scale phenotyping may limit the use of genomic selection in the immediate future. Marker-assisted selection can be used to stack major genes and QTL alleles of large effect in new cultivars, but the required population sizes will limit the number of unlinked genes going much beyond eight. Site-directed transformation should speed the stacking of transgenes in a new cultivar and hence increase the number of major improvements that can be made by genetic transformation.