Zoological Studies 57: 14 (2018) doi:10.6620/ZS.2018.57-14 Open Access Cranial Variation in the Siberian Weasel Mustela sibirica (Carnivora, Mustelidae) and its Possible Taxonomic Implications Alexei V. Abramov1,*, Andrey Yu. Puzachenko2, and Ryuichi Masuda3 1Zoological Institute, Russian Academy of Sciences, Universitetskaya nab. 1, Saint Petersburg 199034, Russia 2Institute of Geography, Russian Academy of Sciences, Staromonetnyy per., Moscow 109017, Russia. E-mail: [email protected] 3Department of Biological Sciences, Faculty of Science, Hokkaido University, Sapporo 060-0810, Japan. E-mail: [email protected] (Received 3 January 2018; Accepted 10 March 2018; Published 26 March 2018; Communicated by Benny K.K. Chan) Citation: Abramov AV, Puzachenko AY, Masuda R. 2018. Cranial variation in the Siberian weasel Mustela sibirica (Carnivora, Mustelidae) and its possible taxonomic implications. Zool Stud 57:14. doi:10.6620/ZS.2018.57-14. Alexei V. Abramov, Andrey Yu. Puzachenko, and Ryuichi Masuda (2018) Morphometric variation in 23 cranial characters of 555 Siberian weasels (Mustela sibirica) was studied across its whole distribution range. Most of the distribution range in Siberia and China is occupied by medium-sized weasels, whereas the eastern part of the species range - including the Russian Far East, Korea and eastern China - is occupied by the larger form. Specimens from the Pacific islands (Jeju and Tsushima) were morphologically closely related to the western form of M. sibirica than to the neighboring continental weasels. The western form can be treated as nominotypical subspecies M. s. sibirica Pallas, 1773, whereas the eastern form can be treated as M. s. manchurica Brass, 1911. Small-sized weasels from the eastern Himalayan area (Myanmar and southwestern China) form a distinct group within M. sibirica, and they were treated as a subspecies; M. s. moupinensis (Milne- Edwards, 1874). Specimens from the western Himalayas (Kashmir, Nepal and Sikkim) are morphologically distinct from all other populations of Mustela sibirica and can be treated as a separate species Mustela subhemachalana Hodgson, 1837. Key words: Skull variation, Mustela sibirica, Subspecies, Taxonomy. BACKGROUND Mongolia and northwest China) (Heptner et al. 1967; Abramov et al. 2016). It occurs in few Pacific The Siberian weasel Mustela sibirica islands (Jeju, Tsushima and Taiwan) and it has Pallas, 1773 has one of the largest distribution been introduced to the Japanese islands Honshu, ranges in the Palearctic among small mustelid Shikoku and Kyushu (Sasaki 2008). The southern species, with the exception of the stoat Mustela and southwestern limits of its distribution range are erminea Linnaeus, 1758 and least weasel M. poorly known: the records from Laos, Vietnam and nivalis Linnaeus, 1766. Mustela sibirica is widely Thailand are incomplete and unconvincing (see distributed across the East Palearctic, from Eastern Abramov et al. 2016). Europe (Kirov Oblast and Tatarstan in Russia) Geographic variation and intraspecific eastward to the Pacific coast and southward to taxonomy of the Siberian weasel have been the Himalayas (from northern Kashmir eastward to poorly investigated. Intraspecific variation in northern Myanmar); however, the weasel does not morphological characters, such as coat color and occupy a large part of arid Central Asia (southern body size, has led to the description of many taxa *Correspondence: E-mail: [email protected] © 2018 Academia Sinica, Taiwan 1 Zoological Studies 57: 14 (2018) page 2 of 10 of the species or subspecies rank (see Appendix length (PL), maxillary tooth-row length (MxtL), 1). Across a wide species range there are more upper carnassial teeth Pm4 length (Pm4L), greatest than 20 scientific names attributed to M. sibirica, length between oral border of the auditory bulla and 11 of them have recently been recognized as and aboral border of the occipital condyles (BcL), subspecies (Wozencraft 2005). Most subspecific length of the auditory bulla (AbL), zygomatic width names are based on pelage characters only, (ZyW), mastiod width of skull (MW), postorbital however these characters (color intensity, width (PoW), interorbital width (IW), width of expression of white and dark spots on the muzzle, rostrum (RW), greatest palatal width (GpW), width etc.) have high individual and seasonal variations of the auditory bulla (AbW), width of upper molar throughout most of Palearctic range of the species M1 (M1W), cranial height (CH), total length of (Stroganov 1962; Heptner et al. 1967). All previous the mandible (ML), length between the angular morphological and genetic studies on M. sibirica process and infradentale (AL), mandibular tooth- were based on a rather narrow set of samples, row length (MatL), length of lower carnassial teeth and no studies have covered the whole distribution M1 (m1L), height of mandible in the vertical ramus range of this species. (MaH) and minimal palatal width (MpW). See The aim of the present paper is to study Abramov and Puzachenko (2009) for the scheme geographical variation across the Siberian of measurements. weasels’ whole distribution range and use any For damaged skulls with incomplete sets patterns it reveals to elucidate possible taxonomic of measurements, we used the expectation- interpretations. We do this using cranial characters, maximisation (EM) algorithm (Dempster et al. which are consistent across seasons. 1977) to estimate the missing values. The missing data were estimated for males and females separately. In all cases, the missing completely MATERIALS AND METHODS at random hypothesis was accepted. EM uses an iterative process to estimate the means, We have examined 555 (including 416 covariance matrix and correlation of measures with intact) adult M. sibirica skulls (450 males and missing values. Grubbs’ two-sided test (Stefansky 105 females) across the species’ distribution 1972) for revealing outliers in new variables was range. The specimens were separated into 33 used. Statistical differences between the measures geographic samples (Appendix 2). Age classes (average, variance and type of distribution) were were defined by scoring morphological features tested by comparing the full data set including of skull structures (the development of crests, the the missing values filled in by the EM method and obliteration of sutures, tooth wear and dentition) the initial data set. No statistical differences were used in mustelids (Buchalczyk and Ruprecht detected between the two data sets. We used 1977). estimated values for multivariate analysis only. The studied specimens are deposited in As a high degree of sexual size dimorphism the Zoological Institute of the Russian Academy (SSD) was observed in Siberian weasel skulls of Sciences (Saint Petersburg, Russia), the (Abramov and Puzachenko 2009 2012), we Zoological Museum of Moscow State University investigated the male and female samples (Moscow, Russia), the Institute of Animal separately. Systematics and Ecology of the Siberian Branch Before the multivariate statistical analysis, all of the Russian Academy of Sciences (Novosibirsk, the measures were standardised to exclude the Russia), B.M. Zhitkov Russian Research Institute impact of the ‘scale’ on different measurements of Game Management and Fur Farming (Kirov, according to the following transformation: Russia), the Natural History Museum (London, xi - xmin xi = , where xi is the standardised UK), National Science Museum (Tokyo, Japan), xmax - xmin Swedish Museum of Natural History (Stockholm, variable, xi , xmin, and xmax are observed, minimum, Sweden), Shanghai Natural History Museum and maximum values of the i-th variable, (Shanghai, China) and the Institute of Zoology and respectively. Genofunds of Animals (Almaty, Kazakhstan). The square dissimilarity matrix contained Twenty-three measurements were taken for the Euclidean distances matrix and the similarity each skull using digital calipers to the nearest matrix of Kendall’s tau-b (corrected for ties) 0.1 mm: condylobasal length (CbL), neurocranium measure of association (Kendall 1975); these length (NcL), viscerocranium length (VcL), palatal © 2018 Academia Sinica, Taiwan Zoological Studies 57: 14 (2018) page 3 of 10 were calculated among all pairs of specimens. of the SHM model (K1) in males mostly correlated Two multidimensional morphological spaces with the relative length of the viscerocranial part of (morphospace) were constructed using the the skull. The first dimension E1 in the female SZM Nonmetric Multidimensional Scaling (NMDS) model was equal to the E1 in the male model, technique (Davison and Jones 1983; NCSS while the equivalent of the K1 dimension of the 2007 (Hintze 2007)) based on two types of male SHM model was K2 in the female model. morphological distance matrices (see more details The K1 dimension had a negative correlation in on our approaches to multivariate statistical females with the relative dimensions of teeth. analyses in previous publications: Abramov et al. The relative variance components in the 2009; Baryshnikov and Puzachenko 2011 2012; variance of the SZM and SHM coordinates Abramov and Puzachenko 2012 2016). These associated with the geographical “factor” were no morphospaces were constructed to provide a more than 60% and 47% for the dimensions E1 of compact representation of the morphological SZM models, for males and females, and 37-36% distances between skulls. The Euclidean for the dimensions K1 of SHM models. This means metric, as a simple geometric distance