Predicting Progeny Performance in Sugarcane from Coefficient of Parentage, AFLP, TRAP and Euclidean Derived Parental Genetic Relationships.

Knowledge of genetic diversity/similarity among parents could be useful in planning crosses in a commercial breeding program.  Genetic similarity (GS) among nine sugarcane parental genotypes was estimated using target region amplification polymorphism (TRAP), amplified fragment length polymorphism (AFLP) markers, coefficient of parentage (f) and euclidean (E) methods.  The TRAP-, AFLP-, f- and E-estimated GS’s relationship to progeny performance was assessed among five bi-parental crosses involving nine parents.   Thirty randomly selected clones from each of the five families were evaluated in a randomized complete block design field trial with two replications during the years 2004 and 2006.  Phenotypic data for stalk height, stalk count, stalk diameter, cane yield and theoretical recoverable sugars (TRS) were collected from which genetic parameters namely family means, genetic variances (σ²_g), average mid parental heterosis (AMPH) and percent heterotic clones (PHC) values were calculated.  The ANOVA results indicated that significant differences existed among the clones within families for all the traits.  The results also indicated that heterosis for TRS could be predicted by family means and σ²_g.  AFLP-GS could be used as a good predictor of genetic parameters for most of the traits including TRS.  The TRAP-, f- and E-GS could also be used as a good predictor for TRS.  From all the four GS estimates, it was found that divergent crosses have to made to achieve maximum heterosis for the trait TRS however the same could not ascertained for other traits.