P. Y. P. Tai

Repeatability between two intermediate sugarcane genotype selection stages in Florida.

Improved yield and disease resistance on sand soils are priorities of the Canal Point (CP) sugarcane (Saccharum spp.) breeding and selection program. Analyses of historical phenotypic data can provide helpful information in guiding selection strategies to meet these priorities. Correlation analysis was used to examine repeatability of phenotypic data used to advance genotypes from an unreplicated single location clonal crop test (stage II) to the subsequent stage (stage III; two replicate, four location clonal crop experiment). Correlations between data for four traits measured in stage II and the corresponding data pooled across soil types for the same genotypes in stage III varied across 23 series of the CP program. Generally, when correlations were statistically significant (P < 0.05), correlation values were low (means; theoretical recoverable sucrose (TRS) r = 0.40, cane yield r = 0.27, and economic index r = 0.23). Similar trends were evident for correlations between data from stage II and stage III on muck soil and stage II and stage III on sand soil across 10 series of the CP program. A 10% reduction in the number of genotypes advanced to stage III over that period would have meant losing only 1 and 13 genotypes that had commercial potential on muck and sand soils, respectively (n = 1278). Correlations between the phenotypic data were significant only for stage III comparisons between TRS and cane yield, which were negatively associated on either soil type. These results indicate that changes in the advancement strategy from stage II are not required as advancing approximately 135 genotypes identifies almost all genotypes with the genetic potential to yield well on muck or sand soils in stage III. Increasing genotypes in stages prior to stage III and changing crossing strategies to improve identification of disease-resistant, high-yielding genotypes for sand soils is recommended.