Fig. 6.

Exogenous GA treatment most strongly affects floral organ patterning in whorl three (stamens). a Mean frequencies of deviations in floral organ number for each floral organ type under control growth conditions (black) and exogenous GA treatment (white) averaged across all genotypes. Values shown are the mean of 352 independent flowers ± S.E. Asterisks indicate a significant difference between control and GA-treated conditions (p < 0.05). Comparisons were made using LSD values at a significance threshold of 5% (see Online Resource 5). b Plotted means of the frequencies of deviations in stamen numbers for long and short stamens under control growth conditions (black) and exogenous GA treatment (white) for flower positions 1–10 and 15, averaged across all genotypes. Values shown are the mean of 32 independent flowers ± S.E. At the level of long and short stamens, both GA treatment and flower position have significant (p < 0.05) effects on the frequency of deviations, but there was no significant interaction between the two (long: p = 0.118, short: p = 0.200; Table 3). c Mean frequencies of organ fusion events for each floral organ type under control growth conditions (black) and exogenous GA treatment (white) averaged across all genotypes. Values shown are the mean of 352 independent flowers ± S.E. Asterisks indicate a significant difference between control and GA-treated conditions (p < 0.05). Comparisons were made using LSD values at a significance threshold of 5% (see Online Resource 5). d–f Examples of organ fusion in stamens; white arrows mark the point of fusion/splitting between two organs. g Example of stamen homeosis, with conversion of half the organ to petal identity. This phenotype was observed only in short stamens