|Abstract or Summary
- Four winter and four spring barley cultivars along with their F₁
hybrids were grown on the Hyslop Agronomy Farm near Corvallis, Oregon,
a high rainfall site (over 1000 mm annually) located in the Willamette
Valley. The experiment consisted of four replications. Ten seeds of
each F₁ and parents were seeded in a three meter row. Plants were
spaced 30 cm between and within the rows.
Investigations were made to evaluate the presence of hybrid
vigor when hybridizing winter and spring barley. Gene action estimates
contributing to yield and its components were obtained. Additional
information concerning the inheritance and the association of
12 agronomic traits and yield were also evaluated.
The data were analyzed by using the factorial technique,
analysis of variance, correlations, path-coefficient, parent-progeny
regressions, and by polynomial and multiple regressions.
The traits measured were: (1) total yield per plant, (2) kernel
weight, (3) tiller number per plant, (4) kernel number per spike, (5) heading date, (6) plant height, (7) flag leaf area, (8) flag
leaf width, (9) flag leaf length, (10) culm diameter, (11) head
length, and (12) head extrusion.
Heterobeltiosis was observed in all of the hybrids except in the
cross between Adair and Short Wocus. The absence of hybrid vigor in
this cross may be attributed to the fact that these two parents are
Combining ability analysis as well as narrow sense heritability
estimates indicated that additive gene action is more predominant
in the expression of heading date, plant height, tiller number, head
length, culm diameter, flag leaf width and kernel weight. Heterobeltiosis
and high SCA estimates were observed in most of the hybrids
for yield per plant and kernel number per spike. This implies that
most of the genetic variation associated with these two traits
resulted from non-additive gene action. The high SCA estimate
observed for head extrusion indicates that this trait appeared to be
controlled by non-additive gene action. Both additive and nonadditive
gene action are important for the expression of flag leaf
area and flag leaf length.
When the 12 traits were correlated and the r values partitioned into
direct and indirect effects it was observed that tiller number, kernel
weight and kernel number per spike exerted the highest direct effect
on yield. Plant height had a small positive direct effect, but an
indirect effect on yield through kernel weight. It was further
observed that a negative association exists between tiller number and kernel weight. This indicates that, in this barley population, it
would be difficult to select for a plant containing large grain type
and short stature, but it may be easy to select short plants with a
relatively large number of tillers and high kernel number per spike.
Path coefficient analysis indicated further that flag leaf area has
a negative direct effect on yield and the remaining six traits have
a negligible or no direct effect on yield, indicating that these
traits can not be used as selection criteria.
More genetic variability for tiller number and kernel number
per spike was noted in the winter barley while more genetic variability
for kernel weight was found in the spring barley.
The results of this study suggest that increases in yield could
be achieved by selecting plants with relatively high tiller number
and high kernel weight. Selection for a high kernel number per spike
can be accomplished by selecting for large culms and longer peduncle.
By this procedure the breeder can take advantage of the additive
genetic variance associated with tiller number and kernel weight and
by the positive correlation that exists between culm size and peduncle
length with kernel number per spike.