Fig. 3

Canonical variate analysis (CVA) plots of Fedia graciliflora flower shape data. CVA is corrected for within-group allometry with mock-treated controls (TRV2-E) (N = 30), FgANS (N = 28), FgCYC2A + FgANS (N = 30), and FgCYC2A (N = 19) knockdown treatments compared to untreated (N = 29). 95% confidence ellipses of means for each group is represented (filled ellipses with dashed contours). The outline drawings of flowers show shape changes associated with each CV from the overall average shape (dotted outline and open circles) for CV1 scores of −3 and +4, for CV2 scores of −3 and +2, and for CV3 scores of −2 and +2 (solid black outline with grey background and solid black circles). Note that these outline drawings are an interpolated form of presentation from the actual landmarks based on the thin-plate spline technique that makes it easier to visualise shape changes. This means that the relevant information in this CVA is from the positions of the landmarks, not from the outline drawings