Table 1.
Statistical significance of the interaction between In(CS) and age-by-sex groups.
Table 2.
Descriptive and analytical statistics for In(CS).
Table 3.
Pairwise differences in mean shapes among age levels.
Fig 1.
PC plot of Procrustes shape coordinates for frontal facial landmark configurations.
Points represent mean PC score for the age-by-sex groups along the first two PCs. Black lines are two dimensional representations of the phenotypic growth trajectories that exist in the multidimensional shape space. The relative amount of shape variation explained by the first two PCs are shown in percentage. TPS grids along the axes illustrate shape difference between configurations with low/high PC scores and the grand mean shape, with ×2.5 magnification of the real shape difference. Mean shape for each age×sex group is illustrated relative to the grand mean shape using TPS grids next to the corresponding point, with ×4 magnification.
Fig 2.
PC plot of Procrustes shape coordinates for frontal facial landmark configurations after size standardization.
Points represent mean PC score for the age-by-sex groups along the first two PCs after size standardization. Black lines are two dimensional representations of the phenotypic growth trajectories that exist in the multidimensional shape space. The relative amount of shape variation explained by the first two PCs are shown in percentage. TPS grids along the axes illustrate shape difference between configurations with low/high PC scores and the grand mean shape, with ×2.5 magnification of the real shape difference. Mean shape for each age×sex group is illustrated relative to the grand mean shape using TPS grids next to the corresponding point, with × 4 magnification.
Fig 3.
Visualization of developmental shape changes of facial morphology.
Facial shapes at three age levels are superimposed to facilitate visualization, with ×4 magnification. (A) Shape development of frontal facial morphology before removing allometry; (B) shape development of frontal facial morphology after removing allometry; (C) shape development of lateral facial morphology before removing allometry.
Table 4.
Shape variance at each age level and pairwise comparisons among age levels.
Table 5.
Differences in phenotypic trajectory attributes between sexes.
Table 6.
Shape difference between sexes at each age level and pairwise comparisons among age levels.
Fig 4.
PC plot of Procrustes shape coordinates for lateral facial landmark configurations.
Points represent mean PC score for the age-by-sex groups along the first two PCs. Black lines are two dimensional representations of the phenotypic growth trajectories that exist in the multidimensional shape space. The relative amount of shape variation explained by the first two PCs are shown in percentage. TPS grids along the axes illustrate shape difference between configurations with low/high PC scores and the grand mean shape, with ×2.5 magnification of the real shape difference. Mean shape for each age×sex group is illustrated relative to the grand mean shape using TPS grids next to the corresponding point, with ×4 magnification.
Table 7.
Multivariate regression of lateral facial shape on In(CS) for each age-by-sex group.
Fig 5.
Allometric trajectories for each age-by-sex group in the first two PCs.
Lateral facial shapes were estimated at the smallest and largest In(CS) of each age-by-sex group. Lines extending from the predicted mean shape at the smallest In(CS) (open symbols) to the predicted mean shape at the largest In(CS) (solid symbols) are two dimensional representations of the static allometric trajectories that exist in the multidimensional shape space. The relative amount of shape variation explained by the first two PCs are shown as percentages. TPS grids along the axes illustrate shape differences between individuals with low/high scores along either PC and the grand mean shape, with ×2.5 magnification.
Fig 6.
Predicted developmental shape changes of lateral facial morphology.
TPS grids represented facial shape at the younger age (reference shape) warped to the facial shape at the older age (target shape), with ×3 magnification.
Table 8.
Pairwise differences in rate and direction of changes in lateral facial shape with changing In(CS) within- and between-sexes among all age-by-sex groups.