Calotriton arnoldi, Carranza & Amat, 2005

García-Salmerón, Alejandro, Muñoz-Muñoz, Francesc, Soler-Membrives, Anna & Valbuena-Ureña, Emilio, 2025, Head form variation in mountain newts of the genus Calotriton (Amphibia, Salamandridae) in a phylogeographic context, Zoologica Scripta 54 (4), pp. 473-486 : 477

2.4 | Head form differences between C. arnoldi and C. asper

Differences in size and shape were assessed with analyses of variance using residual randomization procedure implemented in the procD.lm function ( Collyer & Adams, 2023). A Type I ANOVA with the logarithm of SVL (logSVL), sex and species as predictors, and the logarithm of CS (logCS) as a dependent variable was conducted to detect differences in head size among species when controlling for body size and sexual dimorphism. Since considerable size variation was observed among individuals and populations, the logarithms of SVL (logSVL) and CS (logCS) were used in all analyses ( Klingenberg, 2022). The effects of species, sex, and size on head shape were also tested with a Type I ANOVA in which symmetric shape coordinates were entered as dependent variables. In both analyses, significance was assessed using permutation tests with 1000 iterations. Since a significant dependence of shape on size was detected (see Section 3), subsequent analyses were conducted twice, one with total shape variation (i.e., not corrected by size) and the other with size-corrected data (i.e., using the residuals obtained in the multivariate regression). However, only the results derived from total shape variation are shown because a small percentage of shape variation was explained by size, and the use of size-corrected data provided very similar results. To examine the axes of greater shape variation a principal component analysis (PCA) was conducted with the covariance matrix of total shape variation ( Jolliffe, 1986; Klingenberg et al., 2002).