2.4 Statistical Analysis of the Data
The data were subjected to analysis of variance (ANOVA) using SAS [28] with nested design and a general linear
model. For the ANOVA with the nested design in the experiment, three different error structures were conducted to compare mean squares of different sources of variation. The main factor was tested against the main factor error, where the subtreatment was tested against the interaction of replicates (or block) × sub-treatment and the interaction between the main factor and the sub-factor were tested against the sub-factor error [29]. Further analysis was conducted for each individual factor to separate means with the single factor or combination of the factors for the interaction effects using Duncan test at p= 0.05, as needed.