Variability of quantitative characters affecting yield and selection in sugarcane (saccharum SP.)

Studies were undertaken on the variability of quantitative characters in eleven sugarcane (Saccharum sp.) clones. Plant height, and leaf length and width, were measured at 3, 7, 8 and 11 months after planting: stalks were counted at 3, 6, 7 and 8 months; brix and stalk diameters were measured at 7, 8, 9, 10, and 11 months; stalk length, cane weight, pol, fiber, purity, sucrose percent fresh weight, cane yield per hectare and sucrose yield per hectare were obtained at harvest 11 months after planting. At each stage of growth tested there were significant varietal differences for all characters tested. Clones C0285, N00310, NC0376, C0718 and C0312 had high cane yield, while CP44-101, NC0310, NC0376 and C0718 proved to be good in juice quality. C0349, and P0J2878 were the poorest in every respect. Correlations of characters among clones indicated that stalk number per plot was the most important cane yield component (r= 0.89**) followed by stalk length (r= 0.66*). Cane thickness was negatively correlated to stalk number per plot (r= -0.86**) and cane yield (r= -0.62°), indicating a compromise must be reached between the two goals of thick canes and high cane yield. While cane yield (tons per hectare) was highly correlated to sucrose yield (r= 0.92**), it was uncorrelated to sucrose percent fresh weight, hence both characters can be selected for simultaneously to achieve high sucrose yield. The correlation between sucrose percent fresh weight and sucrose yield (tons per hectare) was non-significant (r= 0.47). of the leaf characters studied only leaf width was significantly related to cane yield (r= -0.61*) and sucrose yield (r= -0.65*), and hence could be a useful character in sugarcane selection. The correlations between stalk number per plot at 7 months and cane yield and sucrose yield at harvest were high (r= 0.97**, and r= 0.72*). Selection before maturity is attained could be useful, especially in those areas where canes take a long time to mature since, in so doing the time taken between stages of selection would be considerably reduced. Under the narrow range of environments tested most varieties responded similarly to differences in environment. However, a few varieties, especially C0285, C0312 and NC0310 behaved quite differently from the rest indicating the existence of a genotype by environment interaction. This indicates testing in different environments e.g. locations, would enable the breader to identify varieties adapted to different environmental conditions. A covariance analysis indicated the extreme importance of randomizing and replicating the wett sources for each clone if comparisons among genotypes are to be valid. For this reason the above conclusions must be regarded as tentative. It also suggested that different clobes may have different optimum planting rates.