Lack of germplasm resources is one of the major factors restricting corn breeding in China. In order to broaden the genetic basis of corn breeding, many maize breeders have done researches with different populations. In the present study, 45 combinations derived from crosses of 5 synthesized corn populations and 9 testers were used with incomplete diallel cross design, to evaluate the major characteristics and combining ability of the 5 populations. The results showed that variances in the 5 populations were significant in all the observed traits except pollening stage, silking stage, green leaves at dough stage, kernels/row, percentage of kernel to cob and kernel depth (Table 1). Table 2 showed that plant height and ear height were higher in GP-1 and GP-2; kernels/row, ear weight, kernel weight/ear and yield per plant were higher in GP-3 and GP-5; plant height and ear height were lower in GP-4. Within the 5 synthesized corn populations, comparatively greater genetic variation was detected for plant height and ear height in GP-4, for ear length in GP-1, for rows/ear, ear weight, kernel weight/ear and yield per plant in GP-2, and for 100-kernel weight in GP-5. There was a smaller genetic variation in GP-3. Normal distribution was observed in yield per plant within the five populations. The frequencies of plants with an average yield per plant higher than 160g, that of control, in the five populations were different, which in GP-1, GP-2, GP-3, GP-4 and GP-5 were 28%, 26%, 31%, 14% and 26%, respectively.
Analysis of combing ability for the traits with significant variations over two seasons showed that all the major traits of GP-5 and GP-3 were higher in GCA, while only ear length, plant height or ear height were detected to be prominence in the other 3 populations (Table 5). There was a big difference in numbers of combinations with significant SCA effects on these traits. GP-1 was involved in combinations with both maximum positive and maximum negative SCA effects in ear length. Most combinations of GP-2 and GP-5 showed maximum positive and maximum negative SCA effects in the other three traits (Table 6). Among the 11 combinations superior to the reference in yield per plant (Table 7), when ES40, R08 and 48-2 were used as one of the parents, the possibility to get hybrids of high yield per plant was great in GP-5, and then in GP-4 and GP-3.
It is concluded that GP-3 and GP-5 possess an appropriate plant height and ear height, with higher yield traits, better general combing ability, greater frequencies of elite individual, and high-yielding combination, compared with the reference cultivar. They exhibit a better potentiality. GP-4 is lower in mean values of the major yield traits and lower in positive GCA effect, but possess a lower plant height and ear height, a higher high-yielding combination frequency than the reference cultivar. It also shows a good potentiality for breeding when improvement is made for GCA and SCA. GP-1 and GP-2 show higher plant height and ear height, a negative GCA in major traits and a lower high-yielding combination frequency compared with the reference cultivar, so they are of less breeding potentiality.