The experimental data were subjected to an analysis of variance (ANOVA) for a completely random design using a statistical analysis system (SAS Institute, Inc., Cary, NC, 2000). Duncan’s multiple range tests were used to determine the
difference among means at the level of 0.05.