The obtained results could be displayed the effeteness of stepwise regression
analysis to determine the strongest trait/s to produce the high sugar, cane
yields and sugar recovery in sugarcane. The model which has two independent
traits of cane yield (CY) and sugar recovery (SR) resembled to be the fit model
for sugar yield and gave R2 = 1.000 in second and third ratoon crops as well as
over both of them. Also, the model which has two independent traits of sugar
yield (SY) and sugar recovery (SR) resembled to be the fit model for cane yield
and gave R2 = 1.000 in second and third ratoon crops as well as over both of
them. The model included three traits of sucrose (SUC), brix (BR) and pol (PO)
with R2=1.000 in second ratoon or SUC and BR with R2= 0.999 in third and over
both ratoons crops, resembled to be the fit model for sugar recovery. High positive
and significant estimates of correlation coefficients were recorded between
cane yield and each of stalk height, stalk weight, stalk number/fed and sugar
yield in all studied ratoon crops. High and positive significant estimates of correlation
coefficients were recorded between sugar yield and each of stalk height,
stalk weight and stalks number/fed over the two ratoon crops. It is remark result
that the traits of brix, purity %, pol and sugar recovery were correlated each other
in high significant and positive estimates in both ratoon crops, revealing the high
quality traits as important components express high sugar in sugarcane. Pathcoefficient
analysis through the second, third and over both of ratoon crops revealed
that the stalk weight, stalk number/fed and cane yield had the greatest direct
and indirect influences on sugar yield. Moreover, the stalk number and stalk
weight had the greatest direct and indirect influences on cane yield. The path coefficient
analysis exhibited that sucrose contributed the greatest effects on sugar
recovery even direct or indirect effects via pol, brix and purity over the both ratoon
crops.