Abstract
The effectiveness of individual selection versus simultaneous selection on number of millable stalks (NMS), stalk diameter, stalk height, and hand refractometer (HR) brix was investigated in three open-pollinated seedling populations of sugar cane (Saccharum spp.). A randomly-selected seedling population was used as a control. The simultaneous selection index resulted in selected populations with a greater brix yield over individual trait selection and random selection. The large differences in mean performance among selected populations observed at the seedling stage, diminished at the evaluation stage conducted in a settling nursery under normal and restricted irrigated conditions. In general, both individual trait and simultaneous selection produced similar results with the exception of that for HR brix. The repeatability of traits, as measured by correlation coefficients between seedling and settling stages, was low with the exception of that for stalk diameter, which showed high (>0.8) repeatability. Selection at the single-stool seedling stage was considered ineffective because of the low correlation coefficients between seedling and settling stages (repeatability), and low proportion of common selected individuals (which produced better yields than the superior cv. CoH 35) at seedling and settling stages. The reasons for ineffectiveness of selection at the seedling stage are also discussed.
