(go
to KOMUGI Home) (go
to WIS List) (go
to NO.73 Contents)
Number of tillers/plant
revealed high positive correlation with yield indicating that
selection based on tiller number/plant could be more rewarding to
improve the grain yield. Similar results were reported by Larik
(1979). No relationships were found in number of tillers/plant with
number of spikelets/spike, spike length, number of kernels/spike and
kernel weight/spike.
Number of spikelets/spike had high positive correlations with spike
length, number of kernels/spike and kernels weight/spike. These
correlationships show that longer spike bears more spikelets,
resulting increased number and wight of kernels. It was noted that
number of spikelets/spike had positive but not significant
association with yield/plant. These results are not in agreement with
the findings of Kumbhar et al (1983) who reported significant and
positive association between number of spikelets/spike and
yield/plant. Spike length showed strong and positive association with
number of kernels/spike, kernel weight/spike and yield/plant. These
relationships suggest that longer spike produce greater number and
weight of grains per spike and yield, and can be used for selection
criteria. Similar results were reported by Larik (1979), Kumbhar et
al (1983), and Soomro and Qureshi (1987).
Number of kernels/spike had significant and positive association with
weight of kernels/spike and yield/plant. This relation explains that
kernels/spike can be used in breeding purposes. These results are
supported by Kumbhar et al (1983), and Soomro and Qureshi (1987).
A highly positive association was found between kernels weight/spike
and yield/plant. This shows that kernel weight/spike can be used in
selecting a high yielding strain. Similar results were reported by
Larik (1979), Kumbhar et al (1983), and Soomro and Qureshi
(1987).
Multiple correlation and regression
Detail information on the effectiveness of different quantitative
attributes and their contribution towards final yield was obtained by
working out multiple correlation and regression (Table
2). This was
accomplished by assessing the cumulative effect of yield components
on grain yield per plant, i.e., taking grain yield/plant as dependent
character and the other traits which were strongly associated with
yield/plant as independent variables. The multiple correlation
coefficient between yield/plant and other characters was 0.996 which
means that 99.2 per cent of the variation in yield/plant can be
attributed to six independent variables. The calculated F value
817.99 (Table
3) shows that the
multiple regression was highly significant. The t-test
indicated that number of tillers/plant, and kernels weight /plant
contributed significantly towards yield/plant while the other four
variables not. These results are in accordance with the results
obtained by Lark (1979) and Kumbhar et al (1983). From the multiple
regression coefficient carried out in the present study, it may be
inferred that number of tillers/plant and kernel weight/spike are the
effective yield components of the bread wheat.
References
Breggs KG and Aytenfisu A (1980) Crop Sci 20: 350-354.
Kumbhar MB, Larik AS, Hafiz HMI and Rind MJ (1983) Wheat Information
Services 57: 42-45.
Larik AS (1979) Wheat Information Service 50: 36-40.
Soomro S and Qureshi GM (1987) Sindh J Agri Res 6: 82-85.
Steel RGD and Torrie JH (1980) Principles and Procedures of
Statistics. McGraw Hill Book Co; Inc. New York pp. 481.
Worley S Jr, Harmon H, Ramey Jr, Harrel DC and Culp TW (1976) CropSci
16: 30-34.
<--Back
(go
to KOMUGI Home) (go
to WIS List) (go
to NO.73 Contents)