A new endemic species of the subgenus Mus (Rodentia, Mammalia) on the Island of Cyprus Author Cucchi, T. Author Orth, A. Author Auffray, J. - C. Author Renaud, S. Author Fabre, L. Author Catalan, J. Author Hadjisterkotis, E. Author Bonhomme, F. Author Vigne, J. - D. text Zootaxa 2006 2005-06-23 1241 1 36 journal article 10.5281/zenodo.172873 afeb6c36-4a31-4df2-96c2-ff538d37f726 1175­5326 172873 Description of Mus cypriacus sp. n. Mammalia, Rodentia , Muridae , Murinae Genus Mus Linneaus, 1758 Mus cypriacus sp. nov. Holotype Number 2005­ 988 in the collection of Vertébrés, Mammifères et Oiseaux , of the Muséum National d’Histoire Naturelle , Paris, France ( MNHN ). Adult female (skull and body in alcohol), age class 5 (6­8 months) (sensu Lidicker 1966 ); collected on the 22nd of April 2004 , by Josette Catalan, Annie Orth and Laurent Fabre at Alassa village of Lemesos (Limassol) district of Cyprus (34°46 N, 32°55 EW). Head and body: 84 mm , tail: 72 mm , foot: 18 mm Sampling of soft parts: muscles for DNA analysis and digestive tractus Karyotype (2N=40) Paratype 1 Number 2005­ 989 in the collection of Vertébrés, Mammifères et Oiseaux , of the Muséum National d’Histoire Naturelle , Paris, France ( MNHN ). Adult female (skull and body in alcohol), age class 4 (4­6 months) (sensu Lidicker (1966)) , collected on the 26th April 2004 by Jean­Christophe Auffray and François Bonhomme at Pachna (Lemesos district), Cyprus (34°47 N, 32°47 EW) Head and body: 79 mm , tail: 75 mm , foot: 17 mm Sampling of soft parts: liver for DNA analysis and digestive tract. Karyotype (2N=40). Paratype 2 Number 2005­ 990 in the collection of Vertébrés, Mammifères et Oiseaux , of the Muséum National d’Histoire Naturelle, Paris, France ( MNHN ); Adult female (skull and body in alcohol), age class 5 (6­8 months) (sensu Lidicker (1966)) , collected on the 26th of April 2004 , by Josette Catalan, Annie Orth, Jean­Christophe Auffray, François Bonhomme, Thomas Cucchi, Laurent Fabre and Eleftherios Hadjisterkotis at Pera Pedi (Lemesos district), Cyprus (33°44 N, 32°04 EW) Head and body: 81 mm , tail: 74 mm , foot: 18 mm Sampling of soft parts: liver for DNA analysis and digestive tractus Karyotype (2N=40) Type locality Alassa (34°46 N, 32°55 EW) is located at 322 meters a.s.l. in the Troodos region. The specimens were live­trapped in abandoned cultivation terraces colonised by Thorny Broom ( Calycotome villosa ), Thorny Gorse ( Genista sphacelata ), Rock­Rose ( Cistus creticus and C. salviifolius ), Asphodel ( Asphodelus aestivus ) and shrubs such as Mastic trees, or Lentisks ( Pistacia lentiscus ) and grasses (Avena sp. oats). Pachna (34°47 N, 32°47 EW) is located at 655 meters a.s.l. in the Troodos region. The specimens were live­trapped in ancient Babylonian­type terraces and vineyard ( Vitis vinifera ) with low walls which favour the growth of Rock­roses, Terebinth ( Pistacia terebinthus ), Almond trees ( Prunus dulcis ) and some rosaceae nearby, Prickly burnet ( Sarcopoterium spinosum ) and Mediterranean Hawthorn ( Crataegus azarolus ). Pera Pedi (33°44 N, 32°04 EW) is located at 849 meters a.s.l. in the Troodos region. The specimens were live­trapped at the edge of a vineyard mixed with Syrian Pears ( Pyrus syriaca ), bordering with an abandoned vineyard invaded by Thorny Broom and Thorny Gorse. On the uncultivated terraced border there were Mastic trees, Terebinths and Almond Trees ( Prunus dulcis = Amygdalus communis ). Other material 1 young adult male (skull and body in alcohol) of the first generation from the breeding colony at Montpellier University in the collection of the Slovene Museum of Natural History, Ljubljana, Slovenia . Sampling of soft parts (liver for DNA analysis and digestive tract is available. 15 skulls ( 10 males and 5 females ) in the collection of the ISE­M (UMR5554), Université Montpellier 2, Montpellier, France Etymology Latin: “from Cyprus ” Vernacular name: “the Cyprus mouse” Distribution and habitat Among the 15 trapping localities sampled in different biotopes on the southern part of the island, Mus cypriacus was captured in eight of them. If further ecological studies and trapping campaigns are required in order to establish the range of the new species, it is possible to identify some trends in its distribution and habitat. Mus cypriacus was found mainly in the Troodos region between 300 and 900 meters a.s.l. Its habitat comprises abandoned cultivation terraces with vineyard, grassy fields, and bushes such as Mastic trees, Terebinths Thorny Broom and Thorny Gorse. In the altitudinal range of 100­150 m , Mus cypriacus can be syntopic with the house mouse (M. m. domesticus ) in forested riverine areas. At low altitudes (less than 100 meters a.s.l.) Mus cypriacus seems absent from areas with strong anthropogenic pressure such as the overexploited agricultural fields of the Mesaoria (central plain of the island), human dwellings and farms, orchards (orange groves) where the house mouse is almost exclusively dominant (Cucchi 2005). Description of external and cranial characters Mus cypriacus ( Figure 3 ) is a typical mouse with clear agouti fur on the back and a whitish belly. The underpants, hands and feet are pinkish. The ears are large and elongated with white hairs behind them. The eyes are large and protruding. The tail is long compared to the head and body length. The mammary formula is 3 + 2 = 10. FIGURE 3: Adult live specimen. The skull ( Figure 4 ) is large but typical of the subgenus Mus . To illustrate the general morphology of the skull we have figured the different views of the Holotype and the paratype 1 because the holotype skull is incomplete (dislocation of the interparietal and the right tympanic bulla during measurements). FIGURE 4: Dorsal, ventral and lateral views of the holotype (n° 2005­988) and paratype 1 (n° 2005­989) of Mus cypriacus . Mus cypriacus presents a shallow and angulated skull with a very robust general appearance. In ventral and dorsal views, the zygomatic arch is very angular, wide and thick and provides a large ocular orbit. The anterior part of the malar process is large. The squamosal process is well developed with a strong embranchment on the skull box. In the dorsal view, the fronto­parietal ridge is angular and salient. The coronal suture is hemi­circular. A crest marks the parietal/interparietal suture. The external occipital crest is also strongly pronounced. In dorsal and ventral views the muscular insertion of the anterior superficial masseter just posterior to the suture of the maxilla and premaxilla is proeminent. In the lateral view, according to the criteria of Marshall & Sage (1981) , the suture between the ventral wing of the parietal and the squamosal is simple, regular and gently curved. The post tympanic hook of the squamosal bone is large and projected in a posterior direction. The anterior profile of the zygomatic plate is straight. The anterior part of the first lower molar (m1) of Mus cypriacus is tetralobed according to the description of Darviche & Orsini (1982) because of a developed and distinct labial anteroconid (tE according to Michaux (1971) . Figure 5: A: Minimum evolution tree based on the analysis of the D­Loop (926 sites). B: same phylogenetic inference on 761 bp of ABPa intron 2. Only the bootstrap values of each node superior to 50 are indicated. Comparisons with the other circum­Mediterranean species Mitochondrial DNA and nuclear genes The trees presented in figure 5A and 5B confirm the results reported in Bonhomme et al . (2004) , with the difference that Mus spretus has been added as a supplementary external group. In figure 5A , bootstrap values show a good support for the new species as a sister taxon of the eastern Mediterranean short­tailed mouse M. macedonicus (including its two subspecies M. macedonicus macedonicus and M. macedonicus spretoides which are the closest mainland relatives), not very far after the divergence of Mus spicilegus . In figure 5B however, the ABPa intron 2 sequences tend to show a higher closeness of cypriacus with spicilegus , although this is loosely supported (54% bootstrap). Based on the D­loop net % divergence between species ( musculus s.lato / spretus =11.04 %, cypriacus / macedonicus s.lato = 4.17%, K2P model with gamma correction as for the phylogenetic inference), we can estimate a divergence time for Mus cypriacus , using an external calibration point. This point is given by the study of Chevret et al. , (2003) , who set a divergence of M. spretus and M. musculus s.lato at 1.4 Myr ago, considering a Mus / Rattus divergence at 12 Myrs. Thus the cypriacus / macedonicus ­spretoides divergence time is estimated at: 1.4 x 4.17/11.04 = 0.53 Myrs. Additionnally the Øp53 pseudogene is present in all the cypriacus samples, as in macedonicus­spretoides, spicilegus and domesticus . It is absent in spretus and musculus ( Ohtsuka, et al. , 1996 ; Tanooka, et al. , 2001 ). As to the Zfy­2 Y chromosome gene, the ten cypriacus males tested showed a musculus pattern as in macedonicus­spretoides , spicilegus and spretus , but unlike domesticus ( Tucker, et al. , 1989 ). Chromosomal data All the individuals analysed presented a karyotype composed of 40 acrocentric chromosomes (2n = 40 ­ NF = 40), which is a diagnostic character of the whole subgenus Mus ( Boursot et al. 1993 ) . Chromosomal identification by G­banding revealed that Mus cypriacus presents the standard band pattern of the subgenus Mus described by Cowell (1984) . The Y chromosome is as large as the one in Mus m. domesticus from Cyprus , as well as in the other subspecies of M. musculus ( M. m. domesticus , M. m. musculus , M. m. castaneus ). This large Y chromosome contrasts with the small Ys described for wild European mice Mus macedonicus ( Israel : ( Ivanitskaya, et al. , 1996 ); Greece and Bulgaria : J. Britton­ Davidian pers. com.), Mus spicilegus (southern USSR: ( Bulatova & Kotenkova., 1990 ); Bulgaria : J. Britton­Davidian pers. com), Mus spretus ( Spain and France : J. Britton­Davidian pers. com.), as well as in two Asian mice Mus booduga ( Sen & Sharma, 1983 ) and M. cervicolor ( Markvong, et al. , 1973 ). External measurements and cranio­dental morphometry Morphometric traits The range of morphometric variables used by Orsini et al. (1983) is summarized in table 3. The body size range of Mus cypriacus (75­91) falls within the one of all the Mediterranean species but its Tail length and HB/T index mean values (74.15 and 1.10, respectively) fall within the ranges of the house mouse (M. m. domesticus ) (69­88 and 0.90­1.20, respectively). Mus cypriacus zygomatic index (0.86) and lower tooth row length (3.39) are slightly larger than that of wild Mediterranean relatives (0.60­082 and 3.01­3.19, respectively). Mus cypriacus and M. m. domesticus present the same relative tail size ( P: NS) ( Table 4 ). Mus cypriacus shares similar values of ZI with wild species ( P: NS) but has a significantly longer tooth row (LTR) ( P ***) than all its relatives ( Table 4 ). TABLE 3. External (HB/T) and cranial (ZI, LTR) measurements of specimens of the type­series, and of samples of the four Mediterranean sub­species (all genetically typed from the ISEM collections). HB, T, HB/T, ZI and LTR are respectively the length of the head and body, tail, the head and body length/tail length, the zygomatic index and the length of the lower tooth row. The values are displayed in millimetres. For all species, mean (M), range and sample size (n) are given.
HB T HB/T ZI LTR
M, (range), n M, (range), n M, (range), n M, (range), n M, (range), n
Holotype n° 2005­988 84 72 1.16 0.89 3.53
Paratype 1 n° 2005­989 79 75 1.05 0.69 3.39
Paratype 2 n° 2005­990 81 74 1.09 0.69 3.43
Mus cypriacus 81.6 (75­91) 18 74.5 (63­87) 18 1.10 (0.99­1.19) 18 0.86 (0.60­1.20) 11 3.39 (3.30­3.51) 13
M. m. macedonicus 76.4 (66­92) 18 66.3 (61­72) 18 1.16 (0.99­1.51) 18 0.82 (0.65­1.04) 19 3.19 (3.08­3.51) 14
M. m. spretoides 79.9 (69­94) 15 62.5 (54­71) 15 1.28 (1.17­1.47) 15 0.84 (0.67­1.09) 7 3.16 (3.07­3.26) 6
M. spretus 84 (70­94) 15 56.4 (49­75) 15 1.51 .97­1.92) 15 0.80 (0.57­0.96) 11 3.17 (2.95­3.37) 10
M. m. domesticus 82.5 (62­95) 27 77.9 (69­88) 27 1.06 (0.90­1.20) 27 0.60 (0.46­0.78) 13 3.01 (2.86­3.23) 13
TABLE 4. ANOVA and matrix of pairwise multiple comparisons (Tuckey HSD) on External body (HB, T, HB/T) and cranial (ZI, LTR) measurements of Mus cypriacus specimens (type excluded) and samples of Mediterranean species. Above the diagonal is displayed the matrix of pairwise mean differences. Below the diagonal is displayed the matrix of pairwise comparison probabilities.
ANOVA HB
F 3.866 df=4 **
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus M. m. domesticus ­­­ * NS NS NS M. m. macedonicus ­6.093 ­­­ NS NS NS M. m. spretoides ­2.615 3.478 ­­­ NS NS M. cypriacus ­0.870 5.222 1.744 ­­­ NS M. spretus 1.452 7.544 4.067 2.322 ­­­
ANOVA T
F 51.777 df=4 ***
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus M. m. domesticus ­­­ *** *** NS *** M. m. macedonicus ­11.574 ­­­ NS *** *** M. m. spretoides ­15.385 ­3.811 ­­­ *** * M. cypriacus ­3.352 8.222 12.033 ­­­ *** M. spretus ­21.452 ­9.878 ­6.067 ­18.1 ­­­
ANOVA HB/T
F 46.182 df=4 ***
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus M. m. domesticus ­­­ * *** NS *** M. m. macedonicus 0.095 ­­­ ** NS *** M. m. spretoides 0.221 0.126 ­­­ *** *** M. cypriacus 0.038 ­0.057 ­0.183 ­­­ *** M. spretus 0.446 0.351 0.226 0.408 ­­­
ANOVA ZI
F 8.445 df=4 ***
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus M. m. domesticus ­­­ *** *** *** *** M. m. macedonicus 0.219 ­­­ NS NS NS M. m. spretoides 0.234 0.015 ­­­ NS NS M. cypriacus 0.253 0.033 0.018 ­­­ NS M. spretus 0.193 ­0.027 ­0.042 ­0.6 ­­­
ANOVA LTR
F 29.268 df=4 ***
Tuckey HSD M. m. domesticus M. m. domesticus ­­­ M. m. macedonicus 0.183 M. m. spretoides 0.149 M. cypriacus 0.384 M. spretus 0.158
......continued on the next page
M. m. macedonicus *** ­­­ ­0.034 0.201 ­0.025
M. m. spretoides ** NS ­­­ 0.236 0.01
M. cypriacus *** *** *** ­­­ ­0.226
M. spretus *** NS NS *** ­­­
Geometric morphometrics of cranial and dental characters Sexual dimorphism on size and shape Sexual dimorphism has generally no or little influence on the size and shape of cranial and dental traits in rodent species (Auffray, et al. , 1996; Renaud, 2005 ). Our results are in agreement with these views. Sexual dimorphism does not appear to be sufficient to interfere with the size and shape differentiation of skulls, teeth and mandibles within the different species ( Table 5 ). A test focusing on Mus cypriacus confirmed the absence of significant sexual dimorphism on size and shape of cranial and dental features ( Table 6 ). Therefore, the animals were pooled independently of their sex in the subsequent statistical analyses. TABLE 5. MANOVA of sex and species for the dental and cranial shapes of Mus cypriacus (types excluded) and Mediterranean species.
m1
Effect Wilks F df1 df2 P
Sex 0.5181 1.2133 1 52 0.3059
Species 0.004 4.463 4 56 ***
M1
Effect Wilks F df1 df2 P
Sex 0.5915 0.6907 1 63 0.8498
Species 0.0016 4.3135 4 56 ***
Mandible
Effect Wilks F df1 df2 P
Sex 0.7165 0.4946 1 53 0.9597
Species 0.0026 4.5676 4 54 ***
Skull (ventral)
Effect Wilks F df1 df2 P
Sex 0.3525 1.0804 1 53 0.4377
Species 0.0001 5.0239 4 52 ***
Skull (dorsal)
Effect Wilks F df1 df2 P
Sex 0.458 0.7397 1 52 0.782
Species 0.001 3.142 4 52 ***
TABLE 6. MANOVA of intraspecific sexual dimorphism for cranial and dental shape of Mus cypriacus (types excluded).
m 1
Effect Wilks F df P
sex 0.053 11 11 0.5507 NS
M1
Effect Wilks F df P
sex 0.029 2.617 13 0.4528 NS
......continued on the next page TABLE 6 (continued)
Mandible
Effect Wilks F df P
sex 0.092 0.757 13 0.729 NS
Skull dorsal
Effect Wilks F df P
sex 0.221 3.679 10 0.3884 NS
Skull ventral
Effect Wilks F df P
sex 0.047 1.689 11 0.5435 NS
Interspecific size difference All morphological features (skull, mandible and molars) show differences in size among species ( Table 7 ). Mus cypriacus is significantly larger than its Mediterranean relatives. Interspecific shape difference Cranial characters Significant interspecific morphological differentiation was found for both skulls (dorsal and ventral views, MANOVA on Procrustes residuals: P <0.001) and mandibles (MANOVA on Fourier coefficient: P <0.001) ( Table 5 ). The pattern of skull differentiation is shown on figure 6. The morphological variation of the specimens was plotted in a bidimensional space combining a size estimator (CS) on the vertical axis and a synthetic shape variable on the horizontal one. This synthetic shape axis corresponds to the first axis of the landmark analysis (PC1) for each view. The combination of shape (PC1) and size (CS) information allows a clear separation of Mus cypriacus whereas the skull configurations of the other species display an important overlap (fig. 6AB). On the dorsal view, Mus cypriacus tends to segregate towards positive values on the shape axis, being opposed to domesticus . The main difference lies in the squamosal process (LMs 9, 10, 18) which is much more developed in Mus cypriacus than in the other species (fig. 6 A). Mus cypriacus is also characterised by a wide and anteriorly displaced zygomatic plate (LM 11), a wide anterior part of the malar process (LMs 14, 15), a longer nasal bone and a narrower parietal on the median axis (LMs 3, 4). In the ventral view, cypriacus specimens are tightly clustered on the shape axis towards negative PC1 values. There is an overlap, however, with the large range of variation of spretus . In the ventral view ( Fig. 6 B), Mus cypriacus differs mainly because of its relatively longer upper tooth row (LMs 9,10) and nasal bone (LMs 1,2) which enlarge the anterior part of its skull. The posterior part of the skull of Mus cypriacus shows a reduction of the occipital foramen (LMs 7,8), and a posteriorly displaced external acoustic meatus (LMs 16,17,19). The pattern of differentiation observed for the mandible (fig. 7AB) shows that Mus cypriacus emerges as a well separate group whereas the other species, except domesticus segregating along CA2, share similar outline shapes leading to an important overlap on the canonical plane. For mandibles the differences are difficult to describe, but we can observe that the divergence of Mus cypriacus lies mainly on the height of the corpus as well as on the convexity of the tooth row edge, probably because the molar alveolus bears larger teeth (See table 4). The angle of the incisor alveolus and the development and position of the coronoid process are also probably involved in the divergence of Mus cypriacus from other Mediterranean species. Generally, Mus cypriacus displays a more robust mandible than other Mediterranean species. Dental characters Similarly, molars display highly significant interspecific divergences (MANOVA on Fourier coefficients: P <0.001) ( Table 5 ). The pattern of differentiation of the first lower molar displayed on the first canonical plane ( Fig. 7 C), represents up to 81 % of the among­group variance, and shows that the different species are distinguished from one another without overlap. Mus spretus occupies a central position whereas domesticus segregates towards positive CA1 values. Both macedonicus subspecies show a similar range of negative scores along CA1. These two taxa segregate along CA2, however. This axis also isolates cypriacus , characterised by negative scores along CA2. The mean reconstructed outline corresponding to each taxon allows us to interpret this general pattern in terms of difference in tooth shape. Mus cypriacus is characterised by a wide molar with a well­developed lingual anteroconid (tD) that causes a general squared shape to the m1. TABLE 7. ANOVA and matrix of pairwise multiple comparisons (Tuckey HSD) on cranial and dental size of specimens of Mus cypriacus (type excluded) and samples of Mediterranean species. Above the diagonal: the matrix of pairwise means differences. Below the diagonal: the matrix of pairwise comparison probabilities.
ANOVA mandible Ha1 area
F 20.7 df=4 ***
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus M. m. domesticus ­­­ *** NS *** NS M. m. macedonicus ­0.286 ­­­ ** * *** M. m. spretoides 0.03 ­0.315 ­­­ *** NS M. cypriacus ­0.524 ­0.238 ­0.554 ­­­ *** M. spretus ­0.052 ­0.338 ­0.022 ­0.576 ­­­
ANOVA m1 Ha1 area
F 8.814 df=4 ***
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus M. m. domesticus ­­­ NS NS *** NS M. m. macedonicus 0.029 ­­­ NS * NS M. m. spretoides 0.02 ­0.009 ­­­ * NS M. cypriacus 0.058 0.03 0.038 ­­­ * M. spretus 0.027 ­0.002 0.007 ­0.031 ­­­
ANOVA M1 Ha1 area
F 8.911 df=4 ***
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus M. m. domesticus ­­­ NS NS ** *** M. m. macedonicus 0.004 ­­­ NS ** NS M. m. spretoides ­0.029 ­0.033 ­­­ *** NS M. cypriacus ­0.042 ­0.039 ­0.072 ­­­ NS M. spretus 0.001 ­0.003 0.03 ­0.041 ­­­
ANOVA Skull dorsal centroid size
F 21.805 df=4 ***
Tuckey HSD M. m. domesticus M. m. domesticus ­­­ M. m. macedonicus 33.51 M. m. spretoides 6.138 M. cypriacus 72.201 M. spretus ­29.432
M. m. macedonicus M. m. spretoides M. cypriacus M. spretus NS NS *** NS ­­­ NS ** *** ­27.372 ­­­ *** * 38.692 66.064 ­­­ *** ­62.942 ­35.57 ­101.633 ­­­
.................................to be continued Table 7 (continued)
ANOVA Skull ventral centroid size
F 21.124 df=4 ***
Tuckey HSD M. m. domesticus M. m. macedonicus M. m. spretoides M. cypriacus M. spretus
M. m. domesticus ­­­ 27.752 ­3.456 71.722 ­26.116
M. m. macedonicus NS ­­­ ­31.208 43.97 ­53.868
M. m. spretoides NS NS ­­­ 75.178 ­22.66
M. cypriacus *** ** *** ­­­ ­97.838
M. spretus NS *** NS *** ­­­
The pattern of differentiation of the first upper molar shape ( Fig. 7 D) is slightly different. Mus spretus is once again in a rather central position but Mus cypriacus is separated from the other taxa more clearly than for the m1, clustering towards positive values of CA1. The second axis clearly separates M. m. domesticus (positive CA2 values), from M. macedonicus macedonicus . According to mean outlines, the M1 of Mus cypriacus is wider and its postero­labial part is more convex, apparently because of a more developed t8. Furthermore, Mus cypriacus mainly differs from relative Mediterranean species by a more developed t1 and t6. Allometric variation A linear regression of the synthetic shape variable (PC1 for skull, CA1 for the molars and mandibles) on size (Log­transformed CS for skull, Ha1 area for mandibles and molars) has been performed in order to investigate possible patterns of allometric variation. The shape of the skull in both dorsal ( n = 57, r = 0.548, r2 = 0.301, P <0.0001) and ventral views ( n =57, r = 0.464, r2 = 0.183, P <0.001) displays highly significant allometric variation. Among the cranial characters, the shape of the mandible is also strongly related to size ( n = 59, r = 0.682, r2 = 0.465, P <0.0001). The m1 shape is not affected by allometric variation ( n = 61, r = 0.230, r2 = 0.053, P = 0.075) whereas the one for M1 ( n = 70, r = 0.440, r2 = 0.183, P <0.001) displays significant ones. The general pattern is thus of a relationship between size and shape differentiation of Mus cypriacus . Nevertheless, these results must be considered with caution. The percentage of morphological variance explained is low in most cases, and allometric patterns cannot be recognised at an intraspecific level, or even among the mainland Mediterranean species. The significant relationships may result from the important differentiation of Mus cypriacus in both size and shape. Being the largest of the species considered, and displaying the most divergent shape in most of the cases, the co­occurrence of extreme values on both axes can lead to a significant relationship without biological meaning.