|
[9001927.] رقم البحث : 9001927 - |
|
Correlation, Stepwise Regression and Path-Coefficient Analyses in New Crosses of Sugarcane / الارتباط و الانحدار التدريجي و معامل المرور في هجن جديدة من قصب السكر |
| |
قطاع الدراسات الزراعية / |
| تخصص البحث : |
| |
Egyptian sٍugar journal / / vol.10 - 5/7/2018 |
| |
اسماء فتحى محمود - مؤلف رئيسي |
| |
A.Abo-Elwafa |
| |
sugar technology
sugarcane |
| |
Analysis of variance revealed significant differences among evaluated sugarcane crosses for stalk length and diameter, stalks number/stool, stool weight, mean stalk weight, pith and brix. The differences among male half-sib families and females within male were highly significant for those traits denoting the presence of additive and non-additive genetic variance. The obtained results illustrate the different genetic make-up among crosses. Consequently, it provides good materials for further selection for desirable genotypes.
Highly positive significant correlation coefficient values were recorded between stool weight and each of stalk length and stalks number/ stool in plant cane and first ratoon crops, respectively. Also, this trend could be found between pith and each of stalks number/ stool and stalk weight in first ratoon crop. Moreover, positive significant correlation coefficient estimates found between stalk weight and each of stalk length and stalk diameter in the first ratoon crop. Non-significant positive and weak correlation values were obtained between brix and stalk weight as well as stalk diameter in plant cane and first ratoon crops, respectively. Otherwise, brix was negatively correlated with each of stalk length, stalks number/ stool, stool weight and pith, respectively.
It is clear from stepwise regression that the first model includes only stalks number/stool and gave high R2 of 0.452 and 0.863 in plant cane and first ratoon crops, respectively. Moreover, the fitted model no. 2 includes stalks number/stool and stalk diameter with R2 close to unity (=0.989) in first ratoon crop. Consequently, this model is superior one and could be used to determine those two traits for selection to high stalk yield in sugarcane. Meanwhile, the high R2 of 0.956 in plant cane crop resulted in model 4 which includes stalks number/stool, stalk length, stalk diameter and pith. This different finding comparing to the first ratoon crop may be due to different canopy and environmental interaction in both cane crops. The expected sugar yields for all fitted stepwise models were insignificant different relative to the actual sugar yield in the two successive crops as revealed by t-test which tend to be zero in some models. Moreover, the correlation coefficients between expected and actual yield of stalk weight were positive, very high and close to the unity in the first ratoon crop. These results display the effectiveness of stepwise regression analysis to determine the strongest trait(s) to in contribution to high stalk yield from cane components in sugarcane.
Also, the path-coefficient analysis exhibited two remarks should be taken an interest issue i.e. a) the direct and indirect effects of stalks numbers/stool were increased from plant cane crop to first ratoon crop. This result was coupled with b) the residual factors (1-R2) were decreased from plant cane crop to first ratoon crop. These results indicate that the genes controlling the stalks number/stool and stalk weight in the current path analysis tend to the maximum genetic expression and could be generally a function of cane stool weight. Consequently, the selection should be directed to both of them to express maximum cane yield of sugarcane. |
| |
 Download Paper |
تسجيل دخول 
