Plots characterizing shape allometry. All plots show log centroid size on the x-axis. (A) Plotted on the y-axis is the overall dentary shape, represented by the regression score. (B) Plotted on the y-axis is PC1. (C) Plotted on the y-axis is PC2. [Color figure can be viewed at wileyonlinelibrary.com]

Contexts in source publication

Context 1

... and group identity were strongly correlated with shape. Group identity showed a stronger association with shape (R 2 = 0.137, p < 0.001) than with size (R 2 = 0.015, p < 0.001). The correlation between overall shape score, indicated as the regression score, and size, was weak but statistically significant (slope = 0.38, R 2 = 0.09, p < 0.001; Fig. 5A). Since PCs 1 and 2 explained a substantial proportion of the shape variation, we also performed a standard linear regression between the first two PC axes and log centroid size, to assess the allometry of the posterior and anterior halves of the mandible, respectively. This relationship between PC1 and size was statistically ...

Context 2

... of the shape variation, we also performed a standard linear regression between the first two PC axes and log centroid size, to assess the allometry of the posterior and anterior halves of the mandible, respectively. This relationship between PC1 and size was statistically significant for positive allometry (slope = 0.37, R 2 = 0.26, p < 0.001; Fig. 5B). Furthermore, the correlation between PC1 and size was nearly identical to that between overall shape score and size (Fig. 5A). The lower statistical power seen in the allometry test on overall shape is probably due to a greater number of shape variables involved, compared with the regression of PC1 on size. In both cases, larger ...

Context 3

... to assess the allometry of the posterior and anterior halves of the mandible, respectively. This relationship between PC1 and size was statistically significant for positive allometry (slope = 0.37, R 2 = 0.26, p < 0.001; Fig. 5B). Furthermore, the correlation between PC1 and size was nearly identical to that between overall shape score and size (Fig. 5A). The lower statistical power seen in the allometry test on overall shape is probably due to a greater number of shape variables involved, compared with the regression of PC1 on size. In both cases, larger sizes were associated with a longer and broader posterior half of the mandible (Fig. 5A, B). There was no visible relationship ...

Context 4

... identical to that between overall shape score and size (Fig. 5A). The lower statistical power seen in the allometry test on overall shape is probably due to a greater number of shape variables involved, compared with the regression of PC1 on size. In both cases, larger sizes were associated with a longer and broader posterior half of the mandible (Fig. 5A, B). There was no visible relationship between PC2 and size (slope = −0.007, R 2 = −0.002, p = 0.78; Fig. ...

Context 5

... test on overall shape is probably due to a greater number of shape variables involved, compared with the regression of PC1 on size. In both cases, larger sizes were associated with a longer and broader posterior half of the mandible (Fig. 5A, B). There was no visible relationship between PC2 and size (slope = −0.007, R 2 = −0.002, p = 0.78; Fig. ...

Context 6

... description, anatomical comparison and morphometric analysis of the specimen indicates that ROMVP 71618 is from a dire wolf, as originally suggested by Churcher (1970). The specimen is morphologically similar to dire wolves from Talara and Rancho La Brea and its size and robustness are congruent with C. dirus (Fig. 5). Furthermore, ROMVP 71618 is probably an aged adult based on heavy tooth wear and loss of the M 1 that appears to have occurred in life. Its advanced age means that discrete diagnostic characters that might have supported taxonomic assignment of C. dirus, such as the number and position of distal cusplets on P 4 (Berta, 1988), have ...