Mn. Uddin et al., RELATIONSHIPS AMONG ATTRIBUTES OF PURE LINES OF BREAD WHEAT DERIVED FROM HETEROTIC F1 HYBRIDS, Australian Journal of Agricultural Research, 45(3), 1994, pp. 601-611
Measurements were made of the number of spikes per unit area, the numb
er of kernels per spike, kernel weight and plant weight in four hetero
tic hand-made hybrids and the commerical hybrid Comet, together with t
he parents of the hand-made hybrids, and 75 pure lines derived from ea
ch hybrid by single seed descent. Kernel weight exhibited the stronges
t heterosis, while the incidence of positive heterosis for the remaini
ng yield components varied between crosses. Mid-parent heterosis for p
lant height was significant for all hand-made hybrids. In both hybrids
and pure lines, high yield was associated with intermediate kernel we
ight and intermediate plant height, but for each trait higher values w
ere observed in the hybrids and pure lines compared with their parents
. In terms of both plant height and the major components of yield, viz
. kernel number per unit area and kernel weight, the highest yielding
pure lines and the hybrids were similar, demonstrating that it is poss
ible to recover homozygous genotypes with the same basic characteristi
cs as the heterotic hybrids from which they were derived. The absence
of any consistent pattern on how the best lines obtained their yield s
uggests that selection for yield based on its components would be unre
warding, except for selection against extremes of kernel weight and pl
ant height.