Head shape variation in eastern and western Montpellier snakes

Headnote

Abstract. The Montpellier snake Malpolon monspessulanus is a wide-ranging species that inhabits Western and Eastern Europe, North Africa and Middle East. Four clades have been recognised as two species, M. insignitus and M. monspessulanus, each with two subspecies. Clades have been substantially identified on the basis of molecular data, pholidosis and colouration, while morphometric traits have been ignored. We compared head shape of 54 specimens belonging to three out of the four clades (M. insignitus insignitus, M. i. fuscus, and M. monspessulanus monspessulanus) by means of geometric morphometrics. We found a significant differentiation: the supraocular and frontal area showed the largest amount of variation, being respectively much thinner in M. i. insignitus, a bit less thin in M. i. fuscus and definitely wider in M. m. monspessulanus. Our findings are fully in agreement with the genetic studies and phylogeny explains more than 20% of the observed variation, supporting the taxonomic distinction inside the genus Malpolon. The functional and/or adaptive meaning of the observed differences is not clear, but it seems unlikely that it may be related to diet. Combining morphological data with phylogeography and environmental features, we formulated an explanatory hypothesis that allowed a precise and testable prediction.

Keywords. Taxonomy, phylogeography, Psammophiinae, Malpolon, Mediterranean, Geometric morphometrics, nonparametric MANOVA, variation partitioning.

INTRODUCTION

The genus Malpolon Fitzinger 1826 belongs to the subfamily Psammophiinae (family Lamprophiidae; Pyron et al., 2013), a taxonomic group of snakes which includes 49 species (Uetz and Hosek, 2013) with an Afro-asiatic distribution (Sindaco et al., 2013; Uetz and Hosek, 2013). The genus currently counts two Mediterranean species (Böhme and De Pury, 2011), which show allopatric geographic distributions (Fig. 1): Malpolon monspessulanus Hermann 1804 (western Montpellier snake), which ranges from Liguria (W Italy) through France, Iberian Peninsula, Morocco, and Western Sahara (Fahd and Pleguezuelos, 2001; Geniez et al., 2006; Sindaco et al., 2013); M. insignitus Geoffroy Sant-Hilaire 1827 (eastern Montpellier snake), which spreads from eastern Morocco eastward around the Mediterranean Sea, reaching Anatolia and Balkans, Middle East, Iraq and Iran, and also Caucasia and Transcaucasia (Sindaco et al., 2013). Two subspecies are currently recognized within each species (Fig. 1, see also Sindaco et al., 2013; Uetz and Hosek, 2013): M. insignitus insignitus Geoffroy Sant-Hilaire 1827, which occurs in the eastern coast of Mediterranean Africa (from eastern Morocco to Egypt) and Middle East (De Haan, 1999; Carranza et al., 2006), and M. i. fuscus Fleischmann 1831, which occupies the remnant part of the areal (Balkans, Anatolia, Caucasia and Transcaucasia, Iraq and Iran; Fig. 1; De Haan, 1999); M. monspessulanus monspessulanus Hermann 1804, which ranges from eastern Morocco to South-western Europe, and M. m. saharatlanticus Geniez, Cluchier and De Haan 2006, a vicariant of the nominal subspecies, in western Morocco and Western Sahara (Geniez et al., 2006).

Specific and subspecific distinctions are based on molecular data and phenotypic traits. The phylogenetic reconstruction using mitochondrial DNA (cyt b and 12S rRNA; Carranza et al., 2006) revealed that i) the genetic distance between the eastern (M. insignitus) and western (M. monspessulanus) clades justifies their raising to the status of distinct species (Carranza et al., 2006), ii) the occurrence in the eastern clade of a greater genetic diversity, which suggests maintaining the subspecies M. i. fuscus, and iii) the genetic variability of the western clade is far lower, even comparing Moroccan populations to European ones, supporting the hypothesis that current European populations are descendant of a recent recolonization from Morocco through Gibraltar (80,000- 170,000 years ago; Carranza et al., 2006). Since molecular investigations did not cover the westernmost part of the M. monspessulanus range in Africa, they failed to detect the subspecies M. m. saharatlanticus, whose description is then based only on morphology (Geniez et al., 2006). Nevertheless, as noticed by Geniez et al. (2006), a partial molecular support to its validity comes from the sole specimen from the northern limit of M. m. saharatlanticus (from Essaouria, SW Morocco; Fig. 1; Carranza et al., 2006), which has been included in the genetic analysis, in fact this specimen partly diverges from M. m. monspessulanus, suggesting a possible genetic differentiation also within the western clade.

In addition to the genetic differences, a set of phenotypic traits allows discrimination among the species and subspecies of the genus (Table 1; Szyndlar, 1988; De Haan, 1999; Geniez et al., 2006). These traits concern pholidosis (i.e., scales arrangement), colour pattern and skull morphology, but, interestingly, none of them involves morphometric features. Actually, as far as we know, the only attempt to differentiate members of the genus Malpolon through morphometric measures was carried out by Geniez et al. (2006). Their study considers three out of the four Malpolon clades (M. i. fuscus was excluded) and, besides colouration and pholidosis, it is focused on the head shape, represented by four linear measures (see Table 1 in Geniez et al., 2006). Even though their analysis has been able to discriminate the three clades, the contribution of the morphometric characters appears to be negligible, with little differences in head shape among clades (at least for the measured traits). Despite Geniez et al. (2006) did not address explicitly this issue, the analysis of the data presented in their article supports this deduction because: i) all the linear measures are highly correlated to each other and to snout-to-vent length (see the loadings of the first principal component in table 5 of Geniez et al., 2006); ii) PC1 axis is a surrogate of the specimen's size, thus clades differentiation along PC1 represents the linear size difference among subsamples (directly appreciable from table 2 in Geniez et al., 2006); iii) PC2 axis is weakly correlated to morphometric measures; iv) coherently with the points above, the authors do not use morphometrics to explain the observed differentiation (see Table 6 and discussion in Geniez et al., 2006). In other words, the analysis, by merging size with shape, does not allow detecting the actual occurrence of morphometric differentiation, therefore a study explicitly addressing this issue and comparing Malpolon clades from a morphometric perspective is still lacking.

In order to start filling the gap, we decided to focus on the head for at least two reasons. Firstly, the snake head is thought to be subject to strong ecological and evolutionary constraints (Shine, 1991; Bonnet et al., 2001; Aubret et al., 2004; Manier, 2004; Vincent et al., 2006; Gentilli et al., 2009; Hampton, 2011; Llorente et al., 2012; Henderson et al., 2013), so the different evolutionary histories undergone by each clade may have lefttraces on the head shape (Shine, 1991; Grudzien et al., 1992; Forsman and Shine, 1997; Hibbitts and Fitzgerald, 2005; Vincent et al., 2006a; Herrel et al., 2008). Secondly, the genus Malpolon is characterised by a peculiar head shape: the combination of large eyes and oblique supraocular scales gives this snake a typical "grim facial expression" (Kreiner, 2007). This trait is shared with other species of Psammophiinae (even if not so accentuated; Coborn, 1991), suggesting its possible genetic base and thus a potential systematic value. In this scenario, our study aims to answer three main questions: i) does head shape vary among the Malpolon clades? ii) Is the morphometric variation in agreement with the phylogenetic relationship among clades? iii) Is there any relationship between geographic pattern of head shape variation and environmental features?

MATERIALS AND METHODS

Samples and geometric morphometrics

We analysed 54 specimens (36 males and 18 females) obtained from the collections preserved in the Natural History Museums of Milan (MSNM), Florence ("La Specola", MZUF) and Genoa ("Giacomo Doria", MSNG; Table 2; Table S1). Only adult males and females were considered (age was assessed following Feriche et al., 2008) and specimens from Lampedusa and Cyprus were excluded in order to avoid a possible "island" effect on the size and shape of the head (Bobak, 2006). Unfortunately, all collections lacked specimens of the subspecies M. m. saharatlanticus, which, consequently, was not included in our comparison. Each specimen was assigned to one of the remnant three subspecies on a geographic base (Fig. 1; De Haan, 1999; Carranza et al., 2006; Geniez et al., 2006; Sindaco et al., 2013). The species identification was confirmed by pholidosis and colouration, whenever possible.

For each specimen we recorded sex and snout-to-vent length (SVL) and we took a picture of the dorsal view of the head, using a Nikon D50 Professional camera at 6.1 megapixel resolution equipped with a Nikkor 60 mm AF-S Micro lens, at a fixed distance of 18 cm. Photos were subsequently used to perform the analysis of the shape by means of geometric morphometrics. Geometric morphometrics is a well-established technique (Claude, 2008; Kaliotzopoulou, 2011; Adams et al., 2012) that effectively allows removing non-shape information from geometric structures (Bookstein, 1997; Dryden and Mardia, 1998) and has been proved to be more powerful than traditional morphometrics (Rohlf, 2000). Fourteen two-dimensional landmarks and eight semi-landmarks (Fig. 2; Table S2) were digitized for each specimen using TpsDig2 (Rohlf, 2010; available at: http://life.bio.sunysb.edu/morph/), which resulted in 22 pairs of coordinates that together described head shape. These configurations were superimposed using a general procrustes analysis (GPA; Rohlf and Slice, 1990) in order to remove information concerning size and orientation and to preserve only information related to shape (Dryden and Mardia, 1998; Claude, 2008). Since head is a bilaterally symmetric structure and we were not interested in the quantification of the deviation from this condition, we followed the procedure suggested by Klingenberg et al. (2002) to remove left-right asymmetry from the global shape. The procedure involves three steps: 1) generating a reflected copy of each configuration; 2) superimposing the combined collection of original and mirrored configurations; 3) averaging the coordinates of corresponding landmarks in the original and mirrored configuration of each specimen in order to obtain a perfectly symmetric configuration. We specifically implemented the first and third step in R (version 3.0.1; R core Team, 2013), while general procrustes superimposition (second step) was performed by the "gpagen" function of the "geomorph" R-package (Adams and Otarola-Castillo, 2013).

Head shape variation among clades

Differences in head shapes were analysed by means of a non-parametric distance-based MANOVA (hereafter NPMANOVA, Anderson, 2001; McArdle and Anderson, 2001). This method substitutes the usual raw data matrix of the dependent variables of MANOVA with a distance matrix computed between each pair of observations; hence it decomposes the sum of the square pairwise distances in an ANOVA-like way (see Anderson, 2001 for details). The technique allows covariates and interactions and the P-values of the model terms are assessed via permutation of the observations (Anderson, 2001). We applied the NP-MANOVA to the matrix of the pairwise Procrustes distances (Claude, 2008) computed on the superimposed configurations, using "clade" (three levels factor: M. m. monspessulanus, M. i. insignitus, M. i. fuscus), "SVL" (covariate) and their interaction as independent variable. SVL was included to take into account a possible allometric effect and it was log-transformed according to the non-linearity of the general allometric function (Gayon, 2000; Claude, 2008). The variable "sex" was excluded from the model because of the low and unbalanced number of females in our sample (we had only two adult females for M. i. fuscus; Table 2). Since "sex" could affect head shape, we tried to ensure the reliability of the results by assessing the coherence of the outcomes of a model that considers only males (MALE) with those of a model that considers both sexes (BOTH). Both models were based on the same set of independent variables, i.e., clade, SVL and their interaction. MALE tests if head shape of males varies among the clades, taking into account allometry; BOTH should allow a generalization of MALE results under the a priori assumption that no sexual dimorphism occurs in head shape. To give some support to this assumption, we tested the occurrence of intersexual difference in head shape in the M. m. monspessulanus clade, where males and females were balanced and numerous enough (Table 2). NP-MANOVA was performed with the function "adonis" of the "vegan" R package (Oksanen et al., 2013), and we tested the assumption of homogeneity of dispersions among clades (Anderson, 2001, 2006; Warton et al., 2012) with the functions "betadisper" and "permutest" (Anderson, 2006; Borcard et al., 2011) of the "vegan" R package obtaining always a non-significant difference among the three groups (P > 0.05, no. of permutations = 4999).

Relationship between head shape and phylogeny

The relationship between phylogeny and morphology was assessed by the combination of two analyses. Firstly, we used the partitioning of the determination coefficient (the so-called "variation partitioning"; Peres-Neto et al., 2006) of MALE and BOTH models to disentangle the pure contribution of the variables "clade" and "SVL" from their shared effects. Secondly, we compared the topology of the phylogenetic tree linking the analysed clades with that of an analogous tree built on the morphometric distances among the mean shapes predicted by the models for each clade. Indeed, if head shape had a partial genetic base, we would expect that the predicted forms for the two subspecies of M. insignitus should resemble each other more than with the predicted shape for M. m. monspessulanus. Since we had only three clades, we lacked the power to test statistically our expectation (only three topologies are possible with three groups), so we just assessed the coherence of the trees. Partition variation was performed with the function "varpart" of the vegan R package (Oksanen et al., 2013), while the morphometric tree was generated with the base function "hclust" in R (R core Team, 2013).

Relationship between head shape and environment

To characterize the areas inhabited by each clade we used the map by Sindaco et al. (2013) and extracted from it the set of one-degree cells where Malpolon has been reported to occur. Each cell was assigned to a clade on the basis of its geographic position (De Haan, 1999; Geniez et al., 2006; Carranza et al., 2006). Then, for each cell, we computed the mean value of two climatic variables, mean temperature of the coldest month (MTCM) and annual precipitation (AP), both obtained from www.worldclim.org (Hijmans et al., 2005) at 10 arc-minutes resolution. We chose these variables to take into account the thermophily of Montpellier snakes (Ottonello et al., 2006; Moreno- Rueda et al., 2009) and to capture climate-habitat relationship at this spatial resolution via correlation (McGill, 2010; Mangiacotti et al., 2013). The two variables were weakly correlated (Spearman correlation coefficient: rs = 0.22), indicating their ability to load different kind of environmental information.

The environmental difference among the areas inhabited by the four Malpolon clades was assessed by a NP-MANOVA: the pairwise Euclidean distances of the scaled (standardized) values of the climatic variables of each occurrence cell (Legendre and Legendre, 1998) were used as the dependent matrix and "clades" (four levels-factor) as the independent variable. Given the low number of clades, the analysis of the relationship between morphometry and environment was just qualitative.

RESULTS

Head shape variation among clades

The difference in head shape among the three Malpolon clades was highly significant for both MALE and BOTH models (Table 3). No sexual dimorphism in head shape was detected in M. m. monspessulanus ("sex": F1,20 = 0.90, P > 0.05, no. of permutations = 9999).

The visual comparison of the predicted forms for each group allowed recognizing two main patterns of variation, loaded by different regions of the head (Fig. 3A; Fig. S1). The most evident one regarded the supraocular area and the external margin of the parietal scales, which showed a clear differentiation between the two species (M. i. insignitus and M. i. fuscus on one side and M. m. monspessulanus on the other side; Fig. 3B, 3C). In particular the two subspecies of M. insignitus had: i) a less flared frontal scale; ii) thinner and longer supraocular scales, which determine a more acute silhouette of the anterior part of the head; iii) a more arched margin between the frontal and supraocular scales; iv) a less curved posterolateral margin of the parietal scales. The difference between the subspecies of M. insignitus were less evident and were loaded by the prefrontal and pari- etal regions: the intersection point between prefrontal and frontal scales as well as the intersection between the posterior margin of the parietals were shifted to the front in M. i. fuscus with respect to M. i. insignitus (Fig. 3D).

The NP-MANOVA also detected a significant allometric effect (Table 3), which was shared by all clades (the interaction term was indeed not significant). Allometry seemed to act mainly on the prefrontal scales, leading to a relative elongation of the scales with increasing size (Fig. 4). Correspondingly, the anterior margin of the frontal scale and the posterior margin between the parietals became more acute.

Relationship between head shape and phylogeny

The partition of the determination coefficient assigned a remarkable part of variation to the factor "clade" (more than 20% in both models; Fig. 5A), while the effect of size appeared negligible (no more than 1%). The overlap between the two variables was null in both models (Fig. 5A), indicating that the pure effect of the phylogenetic factor weighted considerably more than allometry. The morphometric tree obtained from the mean predicted shapes showed that the two subspecies of M. insignitus were more similar to each other than to M. m. monspessulanus (Fig. 5B), i.e., the morphometric relationship among clades predicted by the factor "clade" matched the phylogenetic one.

Relationship between head shape and environment

The environmental features of the occurrence cells are significantly different among clades (F3,349 = 60.28, P . 0.0001, no. of permutations = 9999; Fig. 6A). In par- ticular, M. m. monspessulanus and M. i. fuscus inhabit relatively colder and rainier areas than the couple M. m. saharatlanticus - M. i. insignitus, which, conversely, share a drier and warmer environment (Fig. 6A). The analysis of the climatic distance confirmed this crossed resemblance between the least related clades (Fig. 6B).

DISCUSSION

The species and subspecies of the genus Malpolon have been substantially identified on the basis of molecular data and, to a less extent, on pholidosis and colouration pattern (De Haan, 1999; Carranza et al., 2006; Geniez et al., 2006). On the other hand, morphometric traits have been practically ignored, even though such information can be very helpful, not only for taxonomic purposes, but also in order to shed light on the evolutionary mechanisms that had driven the differentiation of Malpolon lineages. With this in mind, we looked for the occurrence of a morphometric diversification in three out of the four recognised clades of the Montpellier snakes, by focusing on head shape because of its potential genetic, evolutionary and adaptive meaning (Shine, 1991; Forsman and Shine, 1997; Bonnet et al., 2001; Manier, 2004; Vincent et al., 2006a; Herrel et al., 2008; Gentilli et al., 2009; Hampton, 2011; Llorente et al., 2012; Henderson et al., 2013). We found three important results, i.e., i) head shape varies significantly among the three clades; ii) the differences are fully in agreement with the previous phylogenetic analyses and phylogeny explains a relevant part of the observed shape variation; iii) the tree obtained from climatic data does not match that from phylogeny.

The largest difference in head shape occurs between the western and eastern clades and is loaded by the frontal and supraocular regions: the frontal scale is more flared in M. insignitus than in M. monspessulanus and, contemporarily, the supraocular scales are larger and with a continuous joint with the parietals. The two subspecies of the eastern species are less differentiated between each other, with M. i. fuscus showing a little enlargement of the pre-parietal zone and an anterior shiftalong the mid-line of the intersection point between the frontal and prefrontal scales as well as the posterior intersection between the parietal scales. All these differences cannot be ascribed neither to a combination of allometric effect nor to an inhomogeneous distribution of size classes in our sample (Table 2) because SVL is not statistically different among clades (F2,33 = 1.42; P > 0.05), allometry has an overall low effect (Fig. 5A), and it does not involve the same regions of clade (allometry involved mainly the prefrontal region, whereas clades were loaded by the frontal, supraocular and partly parietal one; compare Fig. 3 and Fig. 4). Even when some overlapping effect of clade and size occurs (eastern clades, where specimens have smaller SVL; Table 2), shape changes due to clade are opposite to what could be expected on the basis of size differences between samples (Fig. 3D and Fig. 4).

Our results suggest that head shape patterns of Montpellier snakes match the phylogenetic reconstruction based on genetic data. A similar result has been already obtained for other snakes (e.g., Vipera aspis, Gentilli et al., 2009; Coronella austriaca, Llorente et al., 2012), but we found that this effect seems more pronounced in the genus Malpolon. Indeed, the differences in head shapes showed in Fig. 3A were not amplified, contrary to what was done by both Gentilli et al. (2009) and Llorente et al. (2012), supporting the intensity of the "clade" effect. Therefore, we can reliably state that our results corroborate the distinction of the two species (M. monspessulanus and M. insignitus) and also the recognition of two subspecies inside the insignitus clade, as suggested by both molecular analysis (Carranza et al., 2006) and pholidosis (De Haan, 1999).

The functional and/or adaptive meaning of the observed shape differences is not clear. Herpetologists have usually related head shape modification in snakes to diet and prey size (Shine, 1991; Grudzien et al., 1992; Bon- net et al., 2001; King, 2002; Aubret et al., 2004; Boback, 2006; Vincent et al., 2006a, b; Hampton, 2011), but this idea does not seem to fit the Malpolon case because of two main reasons: firstly, the head features associated with prey size are mainly related to head width and snout length/ width (Shine, 1991; Bonnet et al., 2001; Hampton, 2011); these traits are expected to mould the shape of the parietal scales, while our results focused mainly on the pre-parietal area. Secondly, the (scanty) data about Malpolon diet do not highlight any sort of difference between eastern and western populations and describe Montpellier snake as an opportunistic predator of items of various sizes (De Haan, 1999; Ottonello et al., 2006; Plezueguelos et al., 2009). The occurrence of a systematic difference in the size of available prey is a prerequisite for a diet-shape relationship (Aubret et al., 2004). Since it does not seem so, we do not think feeding is the major key to understand the observed pattern of head shape variation in Montpellier snakes, even though further studies are needed to exclude the diet divergence hypothesis.

An alternative hypothesis to explain the observed pattern of head shape variation may come from the combination of the morphological data with the phylogeography of the genus and with the environmental features of the geographical areas occupied by each clade of Malpolon. Each clade inhabits climatically and topographically different areas, suggesting that these environmental factors could be among the causal factors for the observed morphological variation. More interestingly, the comparison identifies two main climatic groups: M. i. fuscus and M. m. monspessulanus on the one hand, M. i. insignitus and M. m. saharatlanticus on the other hand. If we compare the mean head shape of the three analysed clades (Fig. 3), we notice that the enlargement of frontal and supraocular scales of M. i. fuscus with respect to M. i. insignitus seems to shiftthe insignitus head to the monspessulanus form (Fig. 5B). In the light of the phylogeographic reconstruction by Carranza et al. (2006), the two above considerations allow formulating a testable hypothesis: after the speciation event that led to the distinction of the eastern and western Montpellier snake, M. insignitus dispersed eastward along the Mediterranean coast of North Africa, penetrated into Middle-East and, through Turkey into Balkans and Caucasia. Here it encountered different environmental conditions, which have induced the modification of the head shape (this is the hypothetical assumption: the environment-shape causal relationship). Since the conditions of this region partially match those occurring in the western part of the Malpolon range, M. m. monspessulanus and M. i. fuscus may have shown a certain degree of evolutionary convergence in head shape. Even if this interpretation is quite speculative, it has the advantage of giving a symmetric prediction for the other two clades: under the same assumption head shape of M. m. saharatlanticus should slightly differ from that of the nominal subspecies in the direction that tends towards M. i. insignitus.

A further and complementary consideration is that the region of the head which shows the greatest variability involves the typical trait of the genus, i.e., the "grim facial expression" (Kreiner, 2007): this suggests that the key to the interpretation of the observed head shape variation may lie in the functional meaning of this trait. Even if no information is available about it (as far as we know), some indications may come from the hunting style of the Montpellier snakes: preys are spotted by sight and the first third of the trunk is often uplifted in order to have a better view of the surroundings (Arnold and Burton, 1985; De Haan, 1999; Kreiner, 2007). Such a tactic justifies the great development of the eyes and, consequently, the expansion of the supraoculars and the parallel reduction of the frontal. It is not improbable that also the degree of binocular vision could have taken advantage from this particular configuration, improving hunting performance (Hibbitts and Fitzgerald, 2005). So, different environmental characteristics may have required slightly different eyes-supraoculars-frontal packages and the observed among-clades differences may be the result of this adaptation.

In conclusion our study gives strong support to the taxonomic and phylogenetic relationships among three out of the four currently recognized clades inside the genus Malpolon; it highlights the potential importance of head shape in the reconstruction of the evolutionary history of the Montpellier snakes; it shows how geometric morphometrics can be a powerful instrument to validate molecular phylogeny and to investigate the underlying causal mechanisms (Rohlf, 2002; Kaliontzopoulou, 2011).

ACKNOWLEDGEMENTS

We would like to thank Giuliano Doria, and the technicians of the Natural History Museum of Genoa, and Annamaria Nistri for the opportunity of photographing and measuring specimens from the collections of MSNG and MZUF, respectively. We would also thank Manuela Springolo for the unforgettable sweets which greatly helped doing the analyses and all the other occasional collaborators of the MiPiPa herp-group. We thank also two anonymous reviewers for improving the ms with their constructive comments. Supplementary materials associated with this article can be found at http://www-3. unipv.it/webshi/appendix/welcome.htm.

References

REFERENCES

Adams, D.C., Otarola-Castillo, E. (2013): Geomorph: an R package for the collection and analysis of geometric morphometric shape data. Met. Ecol. Evol. 4: 393-399.

Adams, D.C., Rohlf F.J., Slice, D.E. (2012): A field comes of age: geometric morphometrics in the 21st century. Hystrix 24: 7-14.

Anderson, M.J. (2001): A new method for non-parametric multivariate analysis of variance. Aus. Ecol. 26: 32-46.

Anderson, M.J. (2006): Distance-based tests for homogeneity of multivariate dispersions. Biometrics 62: 245-253.

Arnold, E.N., Burton, J.A. (1985). Guida dei rettili e degli anfibi d'Europa. Franco Muzzio & C., Padova.

Aubret, F., Shine, R., Bonnet, X. (2004): Adaptive developmental plasticity in snakes. Nature 431: 261-262.

Boback, S.M. (2006): A morphometric comparison of island and mainland boas (Boa constrictor) in Belize. Copeia 2006: 261-267.

Böhme, W., De Pury, S. (2011): A note on the generic allocation of Coluber moilensis Reuss 1834 (Serpentes: Psammophiidae). Salamandra 47: 120-123.

Bonnet, X., Shine, R., Naulleau, G., Thiburce, C. (2001): Plastic vipers: influence of food intake on the size and shape of Gaboon vipers (Bitis gabonica). J. Zool. 255: 341-351.

Bookstein, F.L. (1997): Morphometric tools for landmark data: geometry and biology. Cambridge University Press, Cambridge.

Borcard, D., Gillet, F., Legendre, P. (2011): Numerical ecology with R. Springer, New York.

Carranza, S., Arnold, E.N., Pleguezuelos, J.M., (2006): Phylogeny, biogeography, and evolution of two Mediterranean snakes, Malpolon monspessulanus and Hemorrhois hippocrepis (Squamata, Colubridae), using mtDNA sequences. Mol. Phylogenet. Evol. 40: 532-546.

Carranza, S., Arnold E.N., Geniez, Ph., Roca, J., Mateo, J.A. (2008): Radiation, multiple dispersal and parallelism in the skinks, Chalcides and Sphenops (Squamata: Scincidae), with comments on Scincus and Scincopus and the age of the Sahara Desert. Mol. Phylogenet. Evol. 46: 1071-1094.

Claude, J. (2008): Morphometrics with R. Springer, New York.

Coborn, J. (1991): The atlas of snakes of the world. T.F.H. Publications Inc. Neptune, N.J.

De Haan, C.C. (1999): Die Europäische Eidechsennatter Malpolon monspessulanus (Hermann, 1804). In: Handbuch der Reptilien und Amphibien Europas, Vol. 3/IIA, Serpentes II: Colubridae 2, pp. 661-756. Bohme, W. Ed., Aula-Verlag, Wiesbaden.

Dryden, I.L., Mardia, K.V. (1998): Statistical shape analysis. John Wiley & Sons, New York.

Fahd, S., Pleguezuelos, J.M. (2001): Los reptiles del Rif (Norte de Marruecos), II: anfisbenios y ofidios. Comentarios sobre la biogeografía del grupo. Rev. Esp. Herp. 15: 13-36.

Feriche, M., Pleguezuelos J., M., Santos, X. (2008): Reproductive ecology of the Montpellier snake, Malpolon monspessulanus (Colubridae), and comparison with other sympatric colubrids in the Iberian Peninsula. Copeia 2008: 279-285.

Forsman, A., Shine R. (1997): Rejection of non-adaptive hypotheses for intraspecific variation in trophic morphology in gape-limited predators. Biol. J. Linn. Soc. 62: 209-223.

Geniez, P., Cluchier, A., De Haan, C.C. (2006): A multivariate analysis of the morphology of the colubrid snake Malpolon monspessulanus in Morocco and Western Sahara: biogeographic and systematic implications. Salamandra 42: 65-82.

Gentilli, A., Cardini, A., Fontaneto, D., Zuffi., M.A.L. (2009): The phylogenetic signal in cranial morphology of Vipera aspis: a contribution from geometric morphometrics. Herpetol. J. 19: 69-77.

Grudzien, T.A., Huebner, B.J., Cvetkovic, A., Joswiak, G.R. (1992): Multivariate analysis of head shape in Thamnophis s. sirtalis (Serpentes: Colubridae) among island and mainland populations from northeastern lake Michigan. Am. Midl. Nat. 127: 339-347.

Gayon, J. (2000): History of the concept of allometry. Amer. Zool. 40: 748-758.

Hampton, P.M. (2011): Comparison of cranial form and function in association with diet in natricine snakes. J. Morphol. 272: 1435-1443.

Henderson, R.W., Pauers, M.J., Colston, T.J. (2013): On the congruence of morphology, trophic ecology, and phylogeny in Neotropical treeboas (Squamata: Boidae: Corallus). Biol. J. Linn. Soc. 109: 466-475.

Herrel, A., Vincent, S.E., Alfaro, M.E., Van Wassenbergh, S., Vanhooydonck, B., Irschick, D.J. (2008): Morphological convergence as a consequence of extreme functional demands: examples from the feeding system of natricine snakes. J. Evol. Biol. 21: 1438-1448.

Hibbitts, T.J., Fitzgerald, L.A. (2005): Morphological and ecological convergence in two natricine snakes. Biol. J. Linn. Soc. 85: 363-371.

Hijmans, R.J., Cameron, S.E., Parra, J.L., Jones, P.G., Jarvis, A. (2005): Very high resolution interpolated climate surfaces for global land areas. Int. J. Climatol. 25: 1965-1978.

Kaliontzopoulou, A. (2011): Geometric morphometrics in herpetology: modern tools for enhancing the study of morphological variation in amphibian and reptiles. Basic Appl. Herpetol. 25: 5-32.

Kreiner, G. (2007): The snakes of Europe. Edition Chimaira, Frankfurt am Main.

Klingenberg, C.P., Barluenga, M., Meyer, A. (2002): Shape analysis of symmetric structures: quantifying variation among individuals and asymmetry. Evolution 56: 1909-1920.

Kornilios, P., Kyriazi, P., Poulakakis, N., Kumlutas, Y., Ilgaz, Ç., Mylonas, M., Lymberakis, P. (2010): Phylogeography of the ocellated skink Chalcides ocellatus (Squamata, Scincidae), with the use of mtDNA sequences: A hitch-hiker's guide to the Mediterranean. Mol. Phylogenet. Evol. 54: 445-456.

Legendre, P., Legendre, L. (1998): Numerical ecology. Elsevier Science B.V., Amsterdam.

Llorente, G.A., Vidal-García, M., Garriga, N., Carranza, S., Pleguezuelos, J.M., Santos, X. (2012): Lessons from a complex biogeographical scenario: morphological characters match mitochondrial lineages within Iberian Coronella austriaca (Reptilia: Colubridae). Biol. J. Linn. Soc. 106: 210-223.

Manier, M.K. (2004): Geographic variation in the longnosed snake, Rhinocheilus lecontei (Colubridae): beyond the subspecies debate. Biol. J. Linn. Soc. 83: 65-85.

Mangiacotti, M., Scali, S., Sacchi, R., Bassu, L., Nulchis, V., Corti, C. (2013): Assessing the spatial scale effect of anthropogenic factors on species distribution. PLOS ONE 8: e67573. doi:10.1371/journal. pone.0067573

McArdle, B.H., Anderson, M.J. (2001): Fitting multivariate models to community data: A comment on distancebased redundancy analysis. Ecology 82: 290-297.

McGill, B.J. (2010): Matters of scale. Science 328: 575.

Moreno-Rueda, G., Pleguezuelos, J.M., Alaminos, E. (2009): Climate warming and activity period extension in the Mediterranean snake Malpolon monspessulanus. Climatic Change 92: 235-242.

Oksanen, J., Blanchet, F.G., Kindt, R., Legendre, P., Minchin, P.R., O'Hara, R.B., Simpson, G.L., Solymos, P., Stevens, M.H.H., Wagner, H. (2013): Vegan: community ecology package. R package version 2.0-7. http://CRAN.R-project.org/package=vegan.

Ottonello, D., Oneto, F., Salvidio, S., Lamagni, L. (2006): Il colubro lacertino, Malpolon monspessulanus, nell'Italia Nord Occidentale: distribuzione, dieta e morfometria. In: Atti del 6° Congresso Nazionale della Societas Herpetologica Italica (Roma, 27.IX-1. X.2006), pp. 173-181. Bologna, M.A., Capula, M., Carpaneto, G.M., Luiselli, L., Marangoni, C., Venchi, A., Eds, Edizioni Belvedere, Latina.

Peres-Neto, P.R., Legendre, P., Dray, S., Borcard, D. (2006): Variation partitioning of species data matrices: estimation and comparison of fractions. Ecology 87: 2614-2625.

Pleguezuelos, J.M. (2009). Culebra bastarda - Malpolon monspessulanus. In: Enciclopedia Virtual de los Vertebrados Españoles. Salvador, A., Marco, A., Eds. Museo Nacional de Ciencias Naturales, Madrid (http://www. vertebradosibericos.org).

Pyron, R.A., Burbrink, F.T., Wiens, J.J. (2013): A phylogeny and revised classification of Squamata, including 4161 species of lizards and snakes. BMC Evol. Biol. 13: 93.

Quinn, G.P., Keough, M.J. (2002): Experimental design and data analysis for biologists. Cambridge University Press, New York.

R Core Team (2013): R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/.

Rohlf, F.J., Slice, D.E. (1990): Extensions of the Procrustes method for the optimal superimposition of landmarks. Syst. Zool. 39: 40-59.

Rohlf, F.J. (2000): Statistical power comparisons among alternative morphometric methods. Am. J. Phys. Anthropol. 111: 463-478.

Rohlf, F.J. (2010): tpsDig, digitize landmarks and outlines, version 2.16. Department of Ecology and Evolution, State University of New York at Stony Brook.

Sindaco, R., Venchi, A., Grieco, C. (2013): The Reptiles of the Western Palearctic. 2. Annotated checklist and distributional atlas of the snakes of Europe, North Africa, Middle East and Central Asia, with an update to the Vol. 1. Edizioni Belvedere, Latina.

Shine, R. (1991): Why do large snakes eat larger prey items? Funct. Ecol. 5: 493-502.

Szyndlar, Z. (1988): Two new extinct species of the genera Malpolon and Vipera (Reptilia, Serpentes) from the Pliocene of Layna (Spain). Acta Zool. Cracov. 31: 687-706.

Uetz, P., Hosek, J. (2013): The Reptile Database, http:// www.reptile-database.org [accessed on January, 3rd 2014].

Vincent, S.E., Dang, P.D., Herrel, A., Kley, N.J. (2006a): Morphological integration and adaptation in the snake feeding system: a comparative phylogenetic study. J. Evol. Biol. 19: 1545-1554.

Vincent, S.E., Moon, B.R., Shine, R., Herrel, A. (2006b): The functional meaning of "prey size" in water snakes (Nerodia fasciata, Colubridae). Oecologia 147: 204-211.

Warton, D.I., Wright, T.W., Wang, Y. (2012): Distancebased multivariate analyses confound location and dispersion effects. Met. Ecol. Evol. 3: 89-101.