ABSTRACT
Variation in gene sequence or gene structure underpins the phenotypic variation essential for developing new crop varieties. In a traditional potato (Solarium tuberosum L.) breeding program, the creation of a new variety begins when crosses are made between parents with desirable and complementary traits. The progeny are then phenotypically evalu ated for 10 to 15 years to determine whether any of them possess suffi cient quality and pest resistance characteristics to warrant being named as new varieties and released. Throughout this process the genetic varia tion responsible for trait variation is, for the most part, not visible and this leads to a wide range of practical difficulties. Some traits, for ex ample, are highly sensitive to environmental variation and hence can only be meaningfully assessed after many years of field evaluation. If the genetic variation responsible for such a trait could be evaluated directly, e.g. with a PCR assay, then judgments about clone performance could be made much earlier. Another practical difficulty arises because many traits are controlled by multiple genes of small or modest effect and so often it is not clear whether two parents share identical or complementary alleles conditioning a genetically complex trait. Were it possible to track all desirable alleles in parental germplasm, crosses could be made in a more informed manner. A third issue concerns gene dosage: the clone of an autotetraploid potato may have zero, one, two, three, or four copies of a given allele. A higher dosage is preferable should a breeder desire most or all progeny to receive the particular allele. While it is possible to evaluate dosage for dominant genes by making test crosses and counting the number of progeny expressing the trait, it would be quicker and
simpler to assess dosage directly by amplifying and characterizing all the alleles present in a given clone, e.g. by pyrosequencing (Rickert et al. 2002).
