Predicting progeny performance and genome analysis in sugarcane (Saccharum spp)

Genetic diversity/similarity (GS) was estimated among nine sugarcane parental genotypes using target region amplification polymorphism (TRAP), amplified fragment length polymorphism (AFLP) markers and coefficient of parentage (f). Its relationship to progeny performance was assessed among five bi-parental crosses involving the nine parents. Phenotypic data for stalk-height, -count, -diameter, cane yield and theoretical recoverable sugars (TRS) were collected from which genetic parameters (family means, genetic variances, average mid parental heterosis, percent heterotic clones and mid parental values) were calculated. For TRS, families with higher means and variances had a greater proportion of heterotic clones. AFLP-GS was found to be a good predictor of genetic parameters for most of the traits and TRAP-GS (and f-GS) could be used as a good predictor for TRS. Framework linkage maps of Saccharum officinarum \u27Louisiana Striped\u27 and S. spontaneum \u27SES 147B\u27 were constructed using AFLP, sequence related amplified polymorphism (SRAP) and TRAP markers: the S. officinarum map comprised of 146 linked markers spanning 49 linkage groups (LG) and the S. spontaneum map comprised of 121 linked markers spanning 45 LG. Compared to AFLP, SRAP and TRAP markers appear less efficient for linkage mapping in sugarcane, a complex polyploid. The Saccharum interspecific F1 mapping population was evaluated for Brix(B), pol(P) and sucrose(S) at the early(E) and late(L) plant growing seasons in two years, 2004(04) and 2005(05). Conventional quantitative trait loci (QTL) analysis identified markers associated with these traits. In S. officinarum, 50 QTLs were identified with LOD scores ranging from 2.51 to 7.64, explaining from 15.9% (04LP) to 47.8% (04EB) of phenotypic variation. In S. spontaneum, 26 QTLs were identified with LOD scores ranging from 2.69 to 7.51, explaining from 6.5% (04LP) to 43.5% (04LB) of phenotypic variation. Thirty-four digenic interactions were observed in S. officinarum and four in S. spontaneum. Several SRAP and TRAP markers were found to be associated with traits indicating their potential usefulness in QTL tagging. A non-parametric approach, discriminant analysis (DA), also identified several markers which were either similar or localized to the same genomic regions as identified by QTL analysis validating DA as a viable option to identify marker-trait associations. Markers identified in this study in both parents would serve in marker assisted introgression breeding for sugarcane improvement