Evaluation of full-sib sugarcane families for cane yield potential through Random Coefficient Model (RCM) analysis

Abstract

Identification of promising genotypes in sugarcane (Saccharum spp. hybrids) breeding is achieved by multi-stage evaluation initiated by evaluation of thousands of seedlings derived from true seed and then vegetatively propagated. Family selection for cane yield in ground nursery is done through weighing cane yield / plot, which is laborious and expensive. In recent times, various statistical tools have been applied for overcoming the constraints in family selection at single stool stage. A population of 1,800 seedlings derived from fifteen bi-parental crosses involving Indian Co canes and Erianthus introgressed hybrids were evaluated in seedling stage. A subset comprised of 450 clones selected from the single stool stage was evaluated in the first clonal trial. Random Coefficient Model (RCM)was employed to predict the promising families contributing a greater number of high yielding clones using the cane yield data recorded in the early stages of selection viz., single stool and first clonal crop stages. The hybrids of the five elite parental combinations viz., Co 86011 x CoT 8201, Co 8371 x Co 86011, Co 86002 x CoC 671, Co 62198 x Co 8353 and GU04(28) E02 self x Co 0314 identified through RCM analysis showed an improvement of 41.86 % higher cane yield than the clones in the average families (Co 85002 x Co 8209, CoM 0265 x Co 0314, Co 8371 x Co 94008, Co 11015 x Co 775 andGU04 (50) RE6 x Co 94008). The positive slope values estimated by the RCM method for the superior crosses highly correlated with the proportion of elite cane yielding clones in the plant crop trial.