Statistical analysis.
The Bartlett’s test was used to test for homoscedasticity, following
which data were subjected to analysis of variance (ANOVA). A two-way
ANOVA with “shading level” and “harvest time” as fixed factors was
used for STL profile, while a three-way ANOVA model “2 shading level ×
2 harvest time × 4 extract concentration” was employed to evaluate the
allelopathic activity in both wheat coleoptile and weed phytotoxicity
bioassays. Means of STL data were separated by the
Student–Neuman–Keuls test only when the F -test was significant
at the 0.05 probability level. Wheat elongation data are expressed as
percentage difference from the control. Zero ppm was also included in
the statistical analysis.