DEVELOPMENT OF A MODEL FOR MARKER-ASSISTED SELECTION OF SPECIFIC-GRAVITY IN DIPLOID POTATO ACROSS ENVIRONMENTS

Dry matter content in potato (Solanum tuberosum L.) is an important factor in processing and is estimated by specific gravity. We performed quantitative trait loci (QTL) analysis for this trait in 2x potato in three environments. A population of 110 individuals was derived from the cross of a hybrid of haploid S. tuberosum (2x) and S. chacoense Bitter, with a S. phureja Juz. & Buk. clone. This population was characterized for 10 isozyme loci, 44 restriction fragment length polymorphisms (RFLPs), and 63 random amplified polymorphic DNA (RAPDs). Field trials were conducted in two locations in Michigan in 1990 with three replications, and in one location in 1991 with 90 individuals and two replications. Specific gravity was determined through air weight/(air weight - water weight). QTLs were mapped separately for each location and for the average across environments by one-way analyses of variance for each marker locus. A total of 10 putative QTLs was identified over environments and they were localized on Chromosomes 1, 2, 3, 5, 7, and 11. Numbers and effects of QTLs detected varied across environments. The locus with highest R(2) value per QTL in each environment was chosen to develop multilocus models, which explained 39 to 45% of the phenotypic variation for specific gravity. Each model was also tested with data from the other environments, and in general, the predictive value across environments was weaker. A model based on average specific gravity across environments explained a consistent proportion of the phenotypic variation for specific gravity when tested across environments. Using multiple environment data to develop models may be more valuable for marker-assisted selection in a potato breeding program than models based on single environments.