Evaluating parents for cane yield in sugarcane breeding using best linear unbiased prediction analysis of progeny data derived from family plots

Abstract

The success of sugarcane breeding is determined by the choice of parents, crosses made and progeny testing. The objectives of this study were to use best linear unbiased prediction (BLUP) to evaluate parents using progeny data obtained from family plots and to determine the proportion of elite parents within the Midlands breeding populations. Data for stalk number, height and diameter, collected from 20 seedling progenies per family plot in each of the three replications, were used to estimate cane yield. Humic soil populations were established at the Bruyns Hill research station, whereas sandy soil populations were planted at the Glenside research station. BLUP analysis was performed using the Statistical Analysis System. Using BLUP estimates, females (82H0397, 96H0259, 98B0460, 98S0290, 98H0590, 98S1362 and N52) and males (02S0639, 82H0397, 97B0272, 98B0460, 98B1889 and N52) produced progenies with significantly (p < 0.05) higher cane yield in humic and sandy soil populations. Humic soil populations produced a higher proportion of elite parents (28.9%) compared with the sandy soil populations (22.5%) probably due to longer cycles of recurrent breeding and selection. Generally, the proportion of elite parents is still low for both humic and sandy soil populations, suggesting that intensive parent evaluation and development is required.