To avoid genetic background differences, our study used three populations divergently selected for one- and two-eared types.
The following six genotypes, including one- and two-eared subpopulations were derived from three original populations: (i) [MoSQA(S7-H)C8 x Georgia Cow Corn](H-1E)C8 (ACC-1E), (ii) [MoSQA(S7-H)C8 x Georgia Cow Corn](H-2E)C8 (ACC-2E), (iii) [MoSQB(S8-H)C8 x Georgia Cow Corn](H-1E)C8 (BCC-1E), (iv) [MoSQB(S8-H)C8 x Georgia Cow Corn](H-2E)C8 (BCC-2E), (v) SI171(H-1E)C8 (SI171-1E), and (vi) SI171(H-2E)C8 (SI171-2E).
Phenotypic mass selection for one- and two-eared types was subsequently performed in each population for eight cycles.
The sum of squares due to entries with five degrees of freedom was partitioned into two degrees of freedom for a comparison among the three populations and one degree of freedom each for comparisons between the one- and two-eared selections within the ACC, BCC, and SI171 populations.
Differences between one- and two-eared subpopulations of ACC (ACC-1E vs.
Pairwise comparisons of all characters for the one- and two-eared selections of ACC, BCC, and SI171 averaged over all N levels and plant densities showed that stalk crushing strength was significantly higher for the one-eared subpopulation than for the two-eared subpopulation in ACC, but the result was opposite for the BCC population (Table 3).