Quantitative genetics of sugarcane : II. Correlation analysis of continuous characters in relation to hybrid sugarcane breeding.

1. A study has been made of the genetic, clonal and environmental correlations existing among commercially important characters in a sugarcane breeding population of interspecific hybrid origin. 2. In general, there is a close correspondence between genetic correlation coefficients measuring the extent of association at the level of family means, and clonal correlations measuring genotypic relationships at the level of clone means. Greater weight has therefore been given to the numerical estimates of the clonal correlations, in view of their greater accuracy, and the possibility that atypical family groups may unduly influence the genetic correlations. 3. Many significant environmental correlations, generally of a low order of magnitude, have been detected. These have been found to fall naturally into a pattern based on the operation of factors causing differences from plot-to-plot in either growth rate or the process of ripening. 4. A wide variety of genetic relationships cannot be explained in terms of these same two physiological processes. The signs of the clonal correlation coefficients for these particular associations fall very clearly into a pattern identical with that distinguishing the wild species from Saccharum officinarum. 5. The cytological basis for this partial retention of original parent species associations is almost certainly the segregation of whole chromosomes from the wild species as intact units, with little possibility of gene exchange or recombination with chromosomes from S. officinarum. 6. The magnitudes of the observed clonal correlations in general indicate ample scope for independent genetic manipulation of traits from wild and noble canes, provided selection pressure is maintained on all those of commercial importance. However, the interrelationships involving number of stalks per plot, stalk cross-sectional area and fibre percent fresh weight, are somewhat more restrictive. 7. Attention is drawn to the need for a more extensive sampling of the genetic variation available within the wild species S. spontaneum, S. robustum and S. sinense.