Kovaleva V.Y., Efimov V.M., Moroldoev I.V., Litvinov Y.N. 2026. Craniometry of mountain voles of the genus Alticola (Rodentia: Arvicolinae): Orthogonal decomposition of multivariate variability into intra- and interspecific components // Russian J. Theriol. Vol.25. No.1. P.32–43 [in English].
Vera Y. Kovaleva [v_kov_65@mail.ru], Institute of Systematics and Ecology of Animals of SB RAS, Novosibirsk 630091, Russia; Vadim M. Efimov [vmefimov2@yandex.ru], Federal Research Center Institute of Cytology and Genetics Siberian Branch of the RAS, Novosibirsk 630090, Russia; Igor V. Moroldoev [igmor@list.ru], Institute of Systematics and Ecology of Animals of SB RAS, Novosibirsk 630091, Russia; Yuri N. Litvinov [lyun13@yandex.ru], Institute of Systematics and Ecology of Animals of SB RAS, Novosibirsk 630091, Russia.
doi: 10.15298/rusjtheriol.25.1.05
ABSTRACT. In the analysis of quantitative traits within a genus, distinguishing between intra- and interspecific variability is crucial. This study aimed to separate and compare these two components of diversity in the skull morphology of mountain voles. Using Fisher’s method, we performed an orthogonal decomposition of a craniometric data matrix from 12 species and subspecies of the genus Alticola (311 specimens, 15 traits). We created two statistically independent matrices: one representing interspecific variation (differences between species’ weighted centroids) and another representing pooled intraspecific variation (the combined within-group variability). The total variance of the original data was partitioned entirely between these two new matrices. The interspecific component accounted for 44.32% of the total variance, while the intraspecific component accounted for the remaining 55.68%. For each matrix, we computed principal components, which were interpreted as structural-functional modules. The analysis revealed that the first principal component in all cases represented general size-age variability, while the second and third components captured shape variation in the skull and mandible. However, the overall modular structure differed significantly between the interspecific and intraspecific levels. The main finding was that the integration of traits from the facial and cerebral parts of the skull, and mandible was only observed in the interspecific matrix. This suggests that the phenotypic patterns driving divergence between species are distinct from the patterns of variation within species. We propose that this difference may be due to differing rates and mechanisms of epigenetic and genetic restructuring of the phenotype operating at these separate evolutionary levels.
KEY WORDS: principal component analysis, size and shape, morphological integration and modularity, facial and cerebral parts of the skull, genus Alticola.