Restricted Relict Viper in the Western Italian Alps

1Tropical Biodiversity Section, MUSE - Museo delle Scienze, Trento Italy; 2School of Science & the Environment, Manchester Metropolitan University, Manchester UK; 3Societa di Scienze Naturali del Verbano Cusio Ossola, Museo di Scienze, Naturali Collegio Mellerio Rosmini, Domodossola Italy; 4Department of Environmental Sciences, Section of Conservation Biology, University of Basel, Basel, Switzerland; 5Karch, Neuchatel,^ Switzerland

A new vertebrate for Europe: the discovery of a range-restricted relict viper in the western Italian Alps

1, 1, 2 3 4,5 SAMUELE GHIELMI *, MICHELE MENEGON *, STUART J. MARSDEN ,LORENZO LADDAGA and SYLVAIN URSENBACHER

Abstract We describe Vipera walser, a new viper species from the north-western Italian Alps. Despite an overall morphological resemblance with Vipera berus, the new species is remarkably distinct genetically from both V. berus and other vipers occurring in western Europe and shows closer affinities to species occurring only in the Caucasus. Morphologically, the new species appear to be more similar to V. berus than to its closest relatives occurring in the Caucasus, but can be readily distinguished in most cases by a combination of meristic features as confirmed by discriminant analysis. The extant population shows a very low genetic variability measured with mitochondrial markers, suggesting that the taxon has suffered a serious population reduction/bottleneck in the past. The species is extremely range-restricted (less than 500 km2) and occurs only in two disjunct sites within the high rainfall valleys of the Alps north of Biella. This new species should be classified as globally ‘endangered’ due to its small and fragmented range, and an inferred population decline. The main near-future threats to the species are habitat changes associated with reduced grazing, along with persecution and collecting.

Key words: Vipers – Vipera berus – Vipera walser – reptile conservation – new species – bPTP species delimitation model – Alps – biogeography – climate change

Introduction on samples collected east of the city of Milan (Kalyabina-Hauf et al. 2004; Ursenbacher et al. 2006a). In this clade, large genetic The identification of species is an essential building block in bio- differentiation was observed suggesting the possibility of several diversity valuations both at global and local scales, and incom- glacial refugia (Ursenbacher et al. 2006a). Moreover, genetic plete knowledge of an area’s diversity may result in suboptimal analyses also considered V. barani Bohme€ & Joger, 1983, and conservation practices (Bickford et al. 2007). The full faunas and V. nikolskii (Vedmederya, et al., 1986) as directly belonging to floras of sites are often not known, meaning that conservation the V. berus complex (Kalyabina-Hauf et al. 2004; Zinenko et al. decisions have to be made based on studies of just a subset of 2010). taxonomic groups (Howard et al. 1998), use of higher tax- The adder was first recorded in the western Italian Alps in onomies (Balmford et al. 1996) or incomplete species lists 1879 by the herpetologist Michele Lessona at ‘Monasterolo’, (Polasky et al. 2000). Recently, the use of genetic markers has west of the city of Torino, a site separated from the main range allowed the identification of morphologically similar species with of Vipera berus by a gap of almost 120 km (Lessona 1879). strong genetic differentiation, generally considered as ‘cryptic During the 1930s, the herpetologist Felice Capra outlined more species’ (Bickford et al. 2007), an approach that allows identifi- precisely the distribution of this Vipera berus population, discov- cation of divergent evolutionary lineages despite subtle, or even ering two populations in the Alps north of the town of Biella. no morphological differences. The addition of these cryptic spe- He found remarkable differences in head lepidosis, especially in cies to site lists has, in some cases, changed quite radically our individuals from the Biella Prealps, from those of Vipera berus assessments of site importance or species-specific conservation from Central and northern Europe (Capra 1954). In 2006, priorities (Bickford et al. 2007). Ghielmi and colleagues published the first record for a new The adder Vipera berus (Linnaeus, 1758) has a wider distribu- Piemonte locality, at Valle Strona in Verbania Province, repre- tion than any other terrestrial snake, occurring between France senting a disjunct population of the species and an expansion of and the Pacific coast of Russia (Saint Girons 1978). Perhaps the known range in Piemonte region (Ghielmi et al. 2006). surprisingly for a species with such a large range, just three sub- In 2005, Ghielmi noticed that while V. berus is associated species have so far been recognized: V. b. berus in the bulk of geographically with one of its main prey items Zootoca vivipara its distribution; V. b. bosniensis Boettger, 1889; and vivipara (Lichtenstein, 1823) across almost all of its European V. b. sachanlinesis Zarevsky, 1917. Studies have confirmed range, the entire known range of Vipera berus in the western genetic differentiation between V. b. bosniensis and V. berus Alps coincided with Z. carniolica Mayer, Bohme,€ Tiedemann & and strongly suggest that morphological differentiation of Bischoff, 2000, formerly considered a subspecies of the common V. b. sachalinensis is related to recent local adaptation (Kalya- lizard, but now treated as a separate species, and the ancestral bina-Hauf et al. 2004; Ursenbacher et al. 2006a). Additionally, oviparous sister species of the widespread Z. vivipara (Surget- the same studies demonstrated the occurrence of one genetic Groba et al. 2002, 2006; Cornetti et al. 2015). This raised the clade in north-eastern Italy, Slovenia and southern Austria, based possibility that small and disjunct populations of V. berus in the western Alps might be differentiated from other V. berus Corresponding author: Michele Menegon ([email protected]) populations. Contributing authors: Samuele Ghielmi ([email protected]), Stuart The aims of this paper are to investigate the genetic and mor- J. Marsden ([email protected]), Lorenzo Laddaga (l.laddaga phological diversity within the adders from Piemonte, evaluate @libero.it), Sylvain Ursenbacher ([email protected]) *These authors contributed equally to this work. the phylogenetic relationships with other adders and European

J Zool Syst Evol Res (2016) 54(3), 161--173 162 GHIELMI,MENEGON,MARSDEN,LADDAGA and URSENBACHER vipers, formalize the description of this new taxon and make rec- (V. ammodytes ammodytes (Linnaeus, 1758), V. aspis francisciredi Lau- ommendations for further research and management aimed at its renti, 1768, V. ursinii ursinii (Bonaparte, 1835) – all from Italy; conservation. V. darevskii Vedmederja, Orlov & Tuniyev, 1986 – Turkey; V. eriwanensis Reuss, 1933 – Iran; V. kaznakovi Nikolsky, 1909 – Georgia, Gen- Bank: FR727103, FR727034; V. kaznakovi Nikolsky, 1909 – Russia, Materials and methods GenBank: KC176736; V. dinniki Nikolsky, 1913 – Georgia, GenBank: KC176731, AJ275773; V. u. graeca Nilson & Andren, 1988 – Greece, Between 2011 and 2014, 31 individuals of a viper attributed to Vipera GenBank: FR727087, FR727018; V. anatolica Eiselt &Baran, 1970 – berus were found in Piemonte (hereafter called the adder from Piemonte), Turkey, GenBank: KC316113; V. seoanei Lataste, 1879 – Spain, Gen- in the western Italian Alps. Morphological analyses were carried out on Bank: AJ275782, DQ186030, FR727035, DQ185984; DQ185938; and an additional 16 specimens currently deposited in the Natural History Macrovipera lebetina (Linnaeus, 1758) – Turkey), most obtained from Museum of Genova (14 specimens), in the Natural History Museum of GenBank, were used as outgroups. Sequences were aligned by eye. The Torino (one specimen) and in the Insubric Civic Museum of Natural His- appropriate model of sequence evolution was determined using the pro- tory at Clivio and Induno Olona (one specimen). For each individual cap- gram JMODELTEST v2.1.5 (Darriba et al. 2012). The chosen model was tured, geographic coordinates and altitude were recorded with a GPS. applied to the data matrix in order to produce maximum-likelihood (ML) Photographs were taken in order to document body and head lepidosis. estimates using PHYML v3.0 (Guindon and Gascuel 2003). Maximum par- For most individuals, non-invasive DNA samples were taken by cutting simony (MP) and neighbour joining (NJ) analyses were performed using the distal margin of a ventral scale with a sterile scalpel blade or by MEGA v6.06 (Tamura et al., 2007) with the model suggested by JModelT- using buccal swabs (as detailed in Beebee 2008). All sampled individuals est. Robustness of the trees was assessed through bootstrap resampling were immediately released at the exact place of capture. Three additional with 1000 repetitions. In addition, Bayesian inference analysis was done genetic samples were collected in 2002 and 2003 during a previous sur- with the software MRBAYES v3.12 (Huelsenbeck & Ronquist 2001) using vey (Ghielmi et al. 2006). the GTR+I+G model of substitution. The analysis was run with four Ventral, subcaudal and midbody scale rows were counted using stan- chains of 5 9 106 generations, and sampling was performed every 100 dard techniques (Dowling 1951; Saint Girons 1978). Total length, snout- generations. The first 10% of trees was discarded (burn-in) after the con- vent length (SVL) and tail length were recorded to the nearest millimetre. trol that the runs were stable using Tracer v1.6 (Rambaut et al. 2014), The following scales were counted (scales located on head sides were and the analyses done four times to avoid local optima (see Huelsenbeck counted on both sides): rostral, loreals (defined as the scales between & Imennov 2002). The genetic p-distances between the different species nasal, upper labials and perioculars), perioculars, apicals, frontal (when were generated with MEGA. fragmented, the number of fragments was recorded), parietals (when frag- A Bayesian implementation of the Poisson tree processes (bPTP; mented, the number of fragments was recorded) and subocular (following Zhang et al. 2013) model was accessed through the web interface avail- the method proposed by Vacher & Geniez, 2010). Comparative scale able at http://species.h-its.org/ptp/. PTP is a single-locus species delimita- counts and character states were based on a sample of 48 individuals of tion method using only nucleotide substitution information, implementing the adder population from Piemonte and 135 Vipera berus from the a model assuming gene tree branch lengths generated by two independent alpine region and belonging to the Italian clade (see Ursenbacher et al. Poisson process classes (within- and among-species substitution events). 2006a,b). This represents the geographically closest population and the The bPTP analysis was run using 100 000 MCMC generations, with a one showing some convergent features in terms of lepidosis, compared to thinning of 100 and burn-in of 0.1. The bPTP model can be used to deli- the more northerly V. berus populations. Statistical comparisons were mit phylogenetic species in a similar way to the popular and widely used conducted using Student’s t-test (if sample variances equal), Welch’s general mixed Yule coalescent (GMYC) approach (Pons et al. 2006), but t-test (unequal variances) and Wilcoxon’s test. Multivariate comparisons without the requirement for an ultrametric tree (Zhang et al. 2013), elimi- of selected morphological characters between populations of V. berus and nating the possible errors due to time calibration. the adder from Piemonte were conducted using a permutational multivari- Additionally, two nuclear protein-coding loci (BTB and CNC homol- ate analysis of variance (PERMANOVA; Anderson 2001) based on the Bray– ogy 1 – BACH1; Townsend et al. 2008 and recombination activating Curtis dissimilarity measure with 1000 random permutations. To repre- gene 1 – RAG1; Townsend et al. 2004) were used to investigate the sent graphically the differences between the two species, non-metric mul- genetic variability within the nuclear genome. PCRs and sequencing were tidimensional scaling (nMDS) was used. Discriminant analysis was used fi conducted with the same protocol as for the mtDNA, with, respectively, for visually con rming or rejecting the hypothesis that the two species the 6primers F_Nik_Bach1 and R1_Bach1 (Stumpel€ 2012), and the pri- are morphologically distinct. Analyses were carried out separately on mers Rag1_F1 and Rag1_R1 (Stumpel€ 2012), and the PCR amplification males and females. For the nMDS, the following variables were consid- cycles following Stumpel€ (2012). For this aspect, four individuals were ered: subcaudals, crown scales, apicals, perioculars, parietals and loreals investigated: one individual from Piemonte and one sample each of the (only on the right side because the adder from Piemonte showed a much following species: V. berus, V. a. francisciredi and V. a. ammodytes higher degree of asymmetry on loreal scales count, compared to other (similar individuals as for the mtDNA analysis). Sequences were V. berus). The analysis was carried out on the adder from Piemonte and deposited in GenBank nDNA no: KX357727-KX357730; KX357762- three groups of V. berus from the Italian clade having the same number fi KX357765. The relationship between the different sequences was of samples, in order to evaluate intraspeci c variability. Analyses and displayed using TCS 1.21 (Clement et al. 2000) with a parsimony tests were computed with R v3.1.2 (R Core Team, 2014) and PAST3 connection set up to 90% for BACH1 and up to 95% for RAG1. Genetic (Hammer et al. 2001). differentiation was evaluated with p-distance using MEGA. DNA was extracted using a QIAamp DNA Mini Kit (Qiagen, Hilden, Given the extinction risk of the taxon herein described, we have Germany) and several portions of the mitochondrial DNA were amplified. fi decided not to collect any individuals from the wild and to use the speci- A portion of the cytochrome b (cytb; 927 bp) was ampli ed for all sam- mens deposited in the Natural History Museum ‘Giacomo Doria’ of Gen- ples following Ursenbacher et al. (2006a,b), whereas portions of the 16S ova as type specimens, while tissues samples used for DNA analysis are ribosomal RNA (16S; 473 bp), the mitochondrially encoded NADH dehy- deposited at the Muse, the Science Museum of Trento, Italy. drogenase 4 (ND4; 796 bp) and the control region (CR; 1058 bp) were amplified following, respectively, Lenk et al. (2001), Arevalo et al. (1994) and Ursenbacher et al. (2006a,b) for each cytb haplotype and each Results region (six individuals). PCRs were conducted in 25 ll volumes with 1 2 ll of DNA template, 1xPCR buffer (Qiagen), 2 mg mlÀ of Q solution Phylogeny (Qiagen), 2 mM of MgCl2, 0.2 mM dNTPs, 0.5 lM of each primer and The best model selected by JModelTest was GTR+G (freq 0.5 units of Taq polymerase (Qiagen). Successfully amplified fragments = = = = were sequenced by Macrogen Inc (Seoul, South Korea). Sequences were A 0.3065 freq C 0.2830; freq G 0.1187; freq T 0.2919; deposited in GenBank mtDNA no: KX357717-KX357726; KX357731- R(a) = 3.68; R(b) = 47.2; R(c) = 7.22; R(d) = 5.71; R KX357761; KX357766-KX357773. Additionally, sequences of the (e) = 33.7; R(f) = 1.00; gamma shape = 0.210). All genetic different Vipera berus genetic clades (individuals It1, Se2 and Ch2 in reconstruction methods demonstrated that the adder from Pie- Ursenbacher et al. 2006a,b), as well as several other viper species monte region belongs to a completely isolated cluster, which is J Zool Syst Evol Res (2016) 54(3), 161--173 © 2016 Blackwell Verlag GmbH Description of Vipera walser 163 not related to any V. berus samples (Fig. 1). The reconstruction and the other vipers used in this study are 3.97% with V. eriwa- for each gene separately analysed also confirmed this conclusion nensis and V. ursinii ursinii, 4.24% with V. darevskii, 5.36% (Figure S1). Indeed, these snakes are genetically more closely with the different clades of V. berus and 14.8% with V. am- related to the V. ursinii complex than to the V. berus complex modytes (Table 1). The Bayesian implementation of the Poisson (Table 1). The adder from Piemonte appears more closely related tree processes (bPTP) species delimitation model supports the to the cluster regrouping V. dinniki, V. kaznakovi (from Georgia) status of full species for the adder population of Piemonte (see and V. darevskii even if the bootstrap support is limited. It is Fig. 1). thus likely that the split between V. ursinii, V. darevskii–V. kaz- Examination of the nuclear genes demonstrated very different nakovi and the adder from the Piemonte occurred during a simi- sequences between the different analysed species (Fig. 2). The lar period. The p-distances between the adder from the Piemonte genetic p-distance between the adder from Piemonte and V. berus is 0.125% for BACH1 and 0.302% for RAG1; moreover, no identical alleles have been detected between the adder from Piemonte and the other analysed species.

Genetic variability within the adder population from Piemonte The analysis of the complete cytb of 23 samples produced three different haplotypes (maximum two mutations, 0.26% differences). The analyses of the 16S, cytb, CR and ND4 were conducted on six individuals regrouping the three different haplotypes and all regions and produced a combined sequence of 3254 bp with a maximal divergence of 2 bp (0.0092%) within all adders from Piemonte. From the six analysed individuals, four distinct haplotypes were found, and three individuals share the same combined haplotype.

Morphology

The PERMANOVA of a set of morphological features, namely subcaudals, crown scales, apicals, perioculars and parietals, revealed signiﬁcant differences between the two species (F = 11.75, p = 0.0001 in females and F = 4.35, p = 0.0002 in males; see Fig. 3 and Table S1). Discriminant analysis correctly classiﬁed 94% and 88% of females and males, respectively. Hence, there Fig. 1. Phylogenetic tree based on the Maximum-Likelihood reconstruc- are clear morphological differences between individuals of the tion (calculated with PHYML v3.0; Guindon and Gascuel 2003) using two species (Fig. 4). 3254 bp of the mitochondrial DNA showing position of Vipera walser, and relationships with other Vipera species. Values of bootstrap supports are showed for neighbour joining (top left), maximum parsimony (top right), maximum-likelihood (bottom left) and Bayesian inferences (bottom right). Support values for the bPTP species delimitation model are shown in parenthesis right of the species epithet.

Table 1. Genetic distance (p-distance) between Vipera walser sp. no. and the different taxa of European vipers calculated on 3254 bp of the mtDNA (see Material and methods for more details on the genes analysed)

Species p-distance (%)

V. u. ursinii 3.97 V. eriwanensis 3.97 V. darevskii 4.24 V. dinniki 4.76 V. kaznakovi (Russia) 5.34 V. berus 5.36 V. u. graeca 5.60 V. kaznakovi (Georgia) 5.62 V. seoanei 5.82 V. anatolica 6.76 Fig. 2. Genetic relationship of the two nuclear genes analysed (BACH1 V. aspis 8.73 V. ammodytes 10.06 and RAG1) between the four analysed species: V. ammodytes, V. aspis, M. lebetina 13.46 V. berus (Italian clade) and V. walser analysed with TCS 1.21 (Clement et al. 2000). Every mutation is represented by a circle.

Fig. 3. Results of the non-metric multidimensional scaling (nMDS, Bray–Curtis with similarity index), for the females (left) and males (right) conducted separately. The following variables were considered: subcaudals, crown scales, apicals, perioculars, parietals and loreals (only on the right side because V. walser show a much higher degree of asymmetry on loreal scales count, compared to V. berus). The analysis was carried out on V. walser and three groups of V. berus having the same number of samples, in order to evaluate intraspeciﬁc variability. V. walser are in red. The graphs show V. walser to be well differentiated in respect to the three groups of V. berus, which are mostly overlapping.

Fig. 4. Discriminant analysis, based on the six meristic variables, correctly classiﬁes 94% of females (left) and 88% of males. The strongest discrimi- natory variables are crown scales and parietals. Blue = V. berus; red = V. walser

Paratypes Taxonomy One adult male: MSNG33638M collected at Monte Rosso del Vipera walser Ghielmi, Menegon, Marsden, Laddaga & Ursen- Croso, on 30 August 1933. One juvenile male: MSNG33637B bacher sp. nov. (Figs 1–4). and one subadult male: MSNG30818C collected at Alpe Finestre by Felice Capra, respectively, on 28 July 1930 and 15 August Holotype 1928. One adult female: MSNG30818A, one subadult female: Adult female: MSNG34485, collected in S. Giovanni d’Andorno, MSNG30818B, and two juvenile females: MSNG33637C and on the road to Oropa in the Biella prealps, at about 1300 m a.s.l. MSNG33637D collected by Felice Capra at Alpe Finestre by A. Rosazza in the summer of 1930 (Fig. 5). between August 1928 and August 1939. One juvenile female: J Zool Syst Evol Res (2016) 54(3), 161--173 © 2016 Blackwell Verlag GmbH Description of Vipera walser 165

Fig. 5. Habitus of the holotype of Vipera walser sp. nov.

Fig. 6. Pattern variation in adult male (left) and adult female (right) of Vipera walser sp. nov.

MSNG30286 collected by F. Capra at Monte Rosso del Croso Type locality on 12 September 1934; one adult female MSNG33637A col- San Giovanni d’Andorno, strada per Oropa at 1300 m a.s.l. in lected by F. Capra at Alpe le Piane on 5 August 1937; one adult the Alps north of town of Biella, a subrange of the Pennine female MSNG41663 collected by A. Margiocco at Piedicavallo Alps, north-western Italy. in September 1967. J Zool Syst Evol Res (2016) 54(3), 161--173 © 2016 Blackwell Verlag GmbH olSs vlRs(2016) Res Evol © Syst Zool J G 166 06BakelVra GmbH Verlag Blackwell 2016 54

3,161--173 (3), Table 2. Mean sizes, general and head scalation of Vipera walser sp. nov. and other related species, with standard deviations and minimal/maximal values, when provided. Origin of the data: (1) this study; (2) Ursenbacher et al. 2005; (3) Scali and Gentilli 1998; (4) Joger and Stumpel€ 2005; (5) Nilson et al. 1995; (6) Orlov and Tuniyev 1990; (7) Geniez and Teynie, 2005; (8) Goc€ßmen et al. 2014; (9) Avcy et al. 20- 10. The number of males and females is indicated except for the last two species, where data have been gathered from studies and the information was not available.

Vipera Vipera berus Vipera berus Vipera berus Vipera berus Vipera darevskii recalculated Vipera walser sp. nov.(1) Vipera berus (Italian clade)(1) (Italian clade)(2) (Po Plain)(3) (Northern clade)(2) bosniensis(4) kaznakovi(4, 5, 6) from (6, 7, 8 and 9)

Female (31) Male (17) Female (70) Male (65) Female (35) Male (28) Female (13) Male (11) Female (54) Male (47) Female Male Female Male Female Male

Ventral scales 148.50 3.45 143.33 3.28 145.13 4.96 142.68 4.49 145.77 3.23 142.43 3.73 145.9 1.3 142.1 1.1 147.15 3.55 144.11 3.54 144.90 2.81 141.75 3.19 136.2 2.60 135.0 1.81 136.5 4.4 134.6 3.5 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ 141–156 138–149 131–165 131–158 137–153 134–149 140–154 136–148 139–155 135–152 139–159 136–149 130–139 133–139 132–144 129–136 Subcaudal 27.96 2.41 35.06 2.41 28.39 3.28 35.05 3.27 27.63 2.13 34.36 2.39 33.8 1.1 41.5 1.1 30.50 3.59 36.82 2.71 36.87 2.71 28.60 2.18 28.4 1.69 33.6 2.80 28.9 2.76 33.4 3.68 Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ scales 23–32 30–38 22–41 27–43 23–34 29–39 29–42 35–47 21–39 30–42 32–42 24–32 26–32 23–41 25–33 29–38 Loreals (both 9.45 2.16 9.19 1.72 7.29 2.29 6.12 2.18 6.71 2.47 5.89 2.60 5.54 2.40 4.47 1.99 11.06 3.13 Æ Æ Æ Æ Æ Æ Æ Æ Æ side) Perioculars 20.00 1.51 19.50 1.83 18.20 1.69 18.55 1.94 18–22 20.0 1.98 18.04 1.55 Æ Æ Æ Æ Æ Æ (both side) Apicals 2.29 0.74 1.94 0.43 2.00 0.24 1.95 0.21 1.50 0.54 1.57 0.5 Æ Æ Æ Æ Æ Æ Crown scales 18.07 4.41 16.00 2.30 13.70 3.51 12.26 3.27 14.94 3.79 7.5 2.4 HIELMI Æ Æ Æ Æ Æ Æ Subocular 1.55 0.30 1.50 0.16 1.14 0.31 1.09 0.23 Æ Æ Æ Æ ,M ranks

Total length 455.56 167.1 386.00 50.83 491.83 71.79 451.88 71.25 420.0 37.1 437 21.3 479.4 45.8 466.4 40.4 382.1 46.7 376.8 32.0 ENEGON Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ (mm) Tail length 43.42 17.69 50.83 14.16 52.27 8.75 60.02 10.19 56.6 5.7 76.3 5.0 52.0 6.9 64.0 5.7 45.1 6.3 55.1 5.3

Æ Æ Æ Æ Æ Æ Æ Æ Æ Æ ,M (mm)

% Tail 9.90% 0.70% 12.8% 1.09% 10.7% 1.25% 13.3% 1.54% 11.9% 14.9% 10.8% 13.7% 11.8% 1.10% 14.7% 1.71% ARSDEN Æ Æ Æ Æ ,L ADDAGA n U and RSENBACHER Description of Vipera walser 167

Fig. 7. Variation in head scalation in adult female (upper four photographs) and adult male (lower four photographs) of Vipera walser sp. nov.

Differential diagnosis fragmentation of parietals and number of apicals). Identiﬁcation Vipera walser sp. nov. is generally similar to the species of the based solely on observation of external morphology is less obvi- subgenus Pelias and can be confused with V. berus, which ous if individuals of V. berus from southern Alps are considered. co-occurs on the Alps in allopatry (Fig. 6, Table 2). The species Despite this, discriminant analysis correctly identiﬁed individuals differs in a generalized higher count of cephalic scales, in partic- to species in 94% of females and 88% of males, based on a set ular the ones listed below (V. berus in parentheses): higher num- of analysed characters (see Figs 2 and 3). The mean p-distance, ber of crown scales: 7–30, mean 17.4 (versus 4–22, mean 13.0); based on a combined dataset of about 3000 base pairs of mito- loreals: 4–15, mean 9.36 (versus 2–12, mean 6.72); and, to a les- chondrial genes, between V. berus and V. walser is 5.36%. ser extent, perioculars: 16–23, mean 19.8 (versus 13–23, mean Based on our current knowledge of its distribution, Vipera wal- 18.4) (see Table 2). V. walser, in contrast to V. berus, also ser is restricted to the Alps north of town of Biella, a subrange shows a marked tendency towards fragmentation of the cephalic of the Pennine Alps, west of the river Ticino, north-western Italy large shields: the parietal scales are often completely broken (Fig. 8). down into several smaller scales: 2–14, mean 6.3 (versus 2–10, The differences in cephalic scale count between Vipera walser mean 2.4; see also Fig. 7). Less commonly, also the frontal scale and V. berus are shown in Table 2: Crown scales (females: is fragmented into smaller scales. Some individuals exhibit a dor- t45,49 = 4.81, p < 0.0001; males: t28,71 = 5.20, p < 0.0001); lore- sum of the head covered in small, irregular scales, like in als (females: t94,59 = 7.52, p < 0.0001; males: t62,67 = 4.43, V. aspis. V. walser has between 1.5 and 2 rows of subocular p < 0.0001); and, inÀ females only, perioculars (female:À scales on both sides of the head in 85% of the analysed speci- t = 5.33, p < 0.0001; males: t = 0.16, p = 0.87) and 64,16 17,25 À mens (V. berus has typically one row of suboculars, with the apicals (females: t32,86 = 2.14, p = 0.04; males: t18,08 = 0.12, exception of some populations in the southern Alps). The dorsal p = 0.91); the number of scales between the eyes and theÀ supral- zigzag is often broken down into separate bars as in Vipera aspis abials are higher (females: t66,40 = 5.85, p < 0.0001; males: (Linnaeus, 1758) or Vipera berus bosniensis (see Fig. 6). Despite t37,93 = 7.90, p < 0.0001). the lack of a strictly diagnostic morphological character, V. walser can be readily distinguished from populations of Paratype variations V. berus from Central and northern Europe by a combination of Details and meristics for the analysed individuals, including the several characters (e.g. the number of subocular scales, type series, are summarized in Table 3. J Zool Syst Evol Res (2016) 54(3), 161--173 © 2016 Blackwell Verlag GmbH 168 GHIELMI,MENEGON,MARSDEN,LADDAGA and URSENBACHER

Fig. 8. Currently known extent of occurrence of Vipera walser sp. nov. (in blue) and V. berus (in red) in north western Italy

Description of the holotype Morphological Character Congruence (MTMC) approach has Adult female conserved in 70% EtOH in rather good condition, been formalized by Miralles and Vences (2013) and represents with the body slightly swollen probably due to preservation. the most common practice in zootaxonomic studies, combining Snout-vent length (SVL) 515.2 mm, tail 55.0 mm, ratio of tail evidence from DNA sequences and morphological data. Integra- proportion (TL/SVL) 0.107. Two apical scales in contact with tive taxonomy has been also proposed as a framework to bring the rostral. Head oval shaped, wider in the temporal region, neck together conceptual and methodological developments aimed to not very distinct, snout rounded. Frontal single, and larger than describe, classify and name new taxa (Padial et al. 2010). The any other scale on head, five parietals. Rostral slightly higher integration by congruence approach of integrative taxonomy fol- than broader; nasal roundish, nostril circular and approximately lows the principle that different lines of evidence should be com- in the centre of the nasal; one internasal on left side of the head bined to delimit species, such as genetic (mtDNA and nuclear), and two on right side; perioculars 11–10; two rows of suboculars morphological, distributional and ecological data. The genetic on both sides of the head; circumoculars separated from nasals differentiation between V. walser and V. berus, both on mito- by six and five loreal scales, respectively, on right and left side; chondrial and nuclear DNA, is beyond known values between supralabials 9–9, the fourth and the fifth below the eye; 147 ven- well-established species within the same subgenus. The status of trals; 31 divided subcaudals (excluding spine); anal entire; 21 full species is further confirmed by the bPTP analysis and as a scale rows at midbody. Dorsum is brown in colour with a contin- morphological line of evidence by the discriminant analysis. Fur- uous and regular darker brown zigzag. Head is reddish-brown thermore, there is no evidence of introgression from, for exam- with scattered, faint darker markings, and a more obvious ple, V. berus as confirmed by the numerous individuals analysed inverted V-shaped ornamentation just before the neck. Labials for mtDNA, and the strong difference between these two species are paler with black markings bordering the edges. A wide black on the two nuclear genes sequenced. The species, within the band is present on both sides of the head between the postocu- alpine context, inhabits an ecologically peculiar area, character- lars and the neck. Ventrals are black, with white, scattered speck- ized by some the highest rainfall of the whole alpine region ling along the lower margin of the scales and, more consistently, (Mercalli et al. 2008). on both scale extremes by the first row of dorsals. The discovery of the V. walser lineage was particularly unexpected, especially in this biologically well-known and densely Etymology sampled region of western Europe. The species shows closer Vipera walser sp. nov. is named after, and dedicated to, the Wal- genetic affinities with, on one hand, V. darevskii and V. kaz- ser people with whom it shares an extraordinary beautiful and nakovi, species occurring in the Caucasus and, on the other, with wild area of the south-western Alps. the V. ursinii complex (see Table 1), than with the V. berus complex. Limited phylogenetic support suggests a simultaneous split between V. ursinii complex, V. kaznakovi (Georgia) com- Discussion plex and V. walser (possible trichotomy). Moreover, the ML Delineating species boundaries correctly is crucial for the discov- phylogenetic reconstruction regrouped V. walser with the V. kaz- ery of life’s diversity because it determines whether or not differ- nakovi (Georgia) complex, whereas the genetic distance dis- ent individual organisms are members of the same entity (Dayrat played more affinities with the V. ursinii complex. 2005). Most evolutionary biologists now agree that species are Until now, it was believed that western Europe was colonized separately evolving lineages of populations or meta-populations, from the Pelias subgenus only by V. berus (including V. seoanei with disagreements remaining only about where along the diver- Lataste, 1879, restricted to the Iberian peninsula), and the gence continuum separate lineages should be recognized as dis- V. ursinii group, which occupy distinct habitats (cold forest for tinct species (Padial et al. 2010). The Mitochondrial Tree V. berus and steppe areas for V. ursinii; Saint Girons 1978). The J Zool Syst Evol Res (2016) 54(3), 161--173 © 2016 Blackwell Verlag GmbH Table 3. Details of the morphological measurements of the investigated individuals of V. walser sp. nov. walser Vipera of Description

Total Tail Perioculars Parietals length length Crown Loreals (sum left+ Suboculars (sum left+ ID ID locality Age (in mm) (in mm) scales Rostral (mean) right) (mean) Apicals Frontal right) Subcaudals Ventrals MSR

M1 v.stronam1 ad 14 1 4 20 1,5 2 1 3 36 M2 v.stronam2 ad 18 1 4,5 21 1,5 2 2 3 36 M3 v.elvom1 ad 10 1 4,5 16 1,5 2 1 9 34 M4 v.olocchiam1 ad 14 1 5,5 22 1,25 2 1 2 38 M5 v.mastallonem1 ad 17 2 4 20 1,25 2 1 6 31 M6 v.dolcam1 juv 20 1 5,5 21 2 1 1 3 38 M7 a.meggianam1 ad 410 61 17 2 5,5 21 1,5 2 1 4 38 142 22 M8 oropam1 ad 5 3 7 M9 v.elvom2 (MRSN)(MZUT R 2069) ad 16 1 1,5 2 1 7 36 M10 m.rossodelcrosom1 (MSNG33638M) ad 481 61 15 1 2,5 18 1,5 1 1 2 33 149 21 M11 cimarascam1 (MSNG32286) ad 480 60 19 1,5 2 1 2 35 143 21 M12 a.ﬁnestrem1 (MSNG30818C) subad. 306 40 15 1 3,5 16 1,5 2 1 2 37 143 21 M13 a.ﬁnestrem2 (MSNG33637B) juv 224 27 16 1 4,5 19 1,5 2 1 2 36 147 21 M14 sesseram1 (MISN N° cat.2) ad 17 1 5,5 19 1,5 2 1 6 36 138 21 M15 v.stronam3 ad 505 16 1 4,5 18 1,5 2 1 2 33 141 19 M16 v.stronam4 juv 210 17 1 5 22 1,5 2 1 4 34 142 21 M17 v.riobachm1 ad 472 56 16 1 5,5 20 1,5 2 1 4 30 (28–32) 145 21 F1 v.chiobbiaf1 juv 21 1 6 21 1,5 2 1 5 31 F2 v.stronaf1 ad 17 1 4,5 21 1,5 2 1 9 26 F3 v.stronaf2 ad 20 1 5,5 21 2 3 1 3 27 F4 v.masttallonef1 juv 19 1 5,5 19 1,625 2 1 4 30 F5 v.stronaf3 subad. 26 1 5 23 1,5 2 1 10 27 F6 v.stronaf4 ad 25 1 7,5 22 2 4 1 13 27 F7 v.dolcaf1 ad 16 1 6 19 1,5 1 1 12 32 F8 v.mastallonef2 juv 16 1 4 20 1,5 3 1 8 27 F9 v.mastallonef3 ad 18 1 4 20 1,5 2 1 6 29 F10 v.elvof1 ad 610 65 18 1 5 18 1,5 3 2 14 32 148 21 F11 v.vognaf1 ad 527 55 13 1 4,5 18 1 1 1 2 25 149 21 F12 v.vognaf2 ad 548 59 13 1 4,5 21 1,5 2 1 4 28 150 21 F13 a.lepianef1 (MSNG33637A) ad 588 59 14 1 2 19 1,5 2 1 2 29 147 21

olSs vlRs(2016) Res Evol Syst Zool J F14 s.giovannidandornof1 (MSNG34485) ad 570 55 16 1 5,5 21 2 2 1 5 31 147 21 F15 a.finestref1 (MSNG30818B) subad. 263 27,5 17 1 5 20 2 2 1 4 29 148 21 F16 m.rossodelcrosof1 (MSNG30286) juv 232 24,5 17 1 4,5 22 1,5 2 1 6 27 147 21 F17 a.finestref2 (MSNG33637C) juv 213 21 20 1 4,5 19 1,5 2 1 2 28 154 19 F18 a.finestref3 (MSNG33637D) juv 191 17,5 16 1 2,5 20 1 1 1 6 23 149 21 © F19 v.sesiaf1 (MSNG2171A) ad 593 55 7 17 1 1 1 3 30 156 21

06BakelVra GmbH Verlag Blackwell 2016 F20 r.valdobbiaf1 (MSNG2171B) juv 219 22 20 1,5 2 1 29 151 23 F21 a.ﬁnestref4 (MSNG30818A) ad 19 1 5 18 1,5 3 1 6 F22 oropaf1 ad 17 1 5,5 19 2 3 1 3 F23 oropaf2 juv 19 1 5 23 2 3 1 6 F24 sesseraf1 ad 15 1 3 20 1 2 1 3 F25 v.stronaf5 ad 535 50 17 1 4 18 1,5 3 1 9 30 148 21

54 F26 v.stronaf6 ad 520 20 1 5,5 21 2 3 1 3 27 145

3,161--173 (3), F27 v.stronaf7 ad 460 22 1 4 20 1,5 3 1 7 24 141 F28 v.stronaf8 ad 16 1 5,5 20 1,5 3 1 7 27 151 F29 v.stronaf9 ad 580 21 1 4,5 22 1,5 3 2 4 26 145 21 F30 sesseraf2 ad 640 54 14 1 4,5 19 1,5 2 1 2 24 146 F31 piedicavallof1 (MSNG41663) ad 16 1 4,5 19 1,5 2 1 10 30 151 169 170 GHIELMI,MENEGON,MARSDEN,LADDAGA and URSENBACHER presence of a new distinct lineage, more related to the Caucasian ‘critically endangered’ (Tuniyev et al. 2009), whereas V. kaz- vipers, strongly suggests an additional, more recent, colonization nakovi (related to V. darevskii and thus to V. walser) is consid- of western Europe (from the V. kazankovi complex or during the ered ‘endangered’, meaning that the entire clade is highly split between the V. kaznakovi complex and V. ursinii complex) threatened with extinction. than the one involving the V. berus group, and possibly one that Within its restricted range, V. walser appears to be quite com- was concurrent with that of V. ursinii (Early Pliocene; Ferchaud mon in suitable habitat. However, to date, no systematic survey et al. 2012). has been undertaken, either to estimate its population density or Given that the European viper species tend to exclude each identify its habitat requirements. Such surveys are clearly a prior- other geographically, resulting in limited portions of overlapped ity for the future research. Estimates of current abundance, using ranges (Saint Girons 1978), we can assume that V. walser found mark–recapture or distance sampling (e.g. Mazerolle et al. 2007), refugial areas different from those of V. berus during the numer- would be useful to determine total population size and trends, ous glaciations of the Pleistocene. Currently, both V. berus and and to more precisely assign the species to a Red List category. V. walser seem to occupy very similar habitats, suggesting a Occupancy modelling (Larson 2014) might also be suitable to possible competition (or ecological differentiation as that determine areas of occupancy at appropriate scales. between V. aspis and V. berus; Guillon et al. 2014). It is, how- Perhaps more important would be detailed studies of the spe- ever, possible that, like V. kaznakovi, V. walser can tolerate war- cies’ precise habitat requirements, to determine how past and mer temperatures than can V. berus so long as sufficient current land use changes have affected the species, and how they humidity is present. Yet, this possibility needs to be investigated might be altered to benefit the species in the near future. Based as it could have important implications for future conservation on our preliminary observations, this species inhabits open areas, programmes. often with rocky outcrops (Fig. 9), and may not tolerate wood- land unless it is very sparse. European mountains experienced a long period of agricultural/agropastoral expansion from the Late Near-future threats and conservation Middle Ages to the 19th century, with large areas of the Alps Vipera walser appears to occur only in a very limited area in the converted to upland grasslands and heathlands (e.g. Vives et al. Alps north of Biella (Fig. 8). It is very likely that all native pop- 2014). These open landscapes were presumably beneficial for ulations of adder south of the Alps and west of the river Ticino V. walser. However, the decline in agropastoral activities in the belong to the species herein described. Based on the Italian Atlas last 100 years and associated afforestation (Carlson et al. 2014; of Amphibians and Reptiles (Sindaco et al. 2006), the current Garbarino et al. 2014) is probably the greatest threat to the spe- distribution area (‘extent of occurrence’) is almost certainly cies, and it is an urgent priority to assess such changes within <1000 km2. Consequently, V. walser should be classified as ‘en- the range of V. walser. More immediate and major threats in the dangered’ according to IUCN (2014) Red List criteria B1a/B2a. short term are culling and collection. Indeed, the description of If we consider that the population is strongly fragmented, or that several new vipers species (e.g. V. kaznakovi and Montivipera the actual area of occupancy is probably <500 km2 and frag- wagneri (Nilson & Andren, 1984)), as well as the attraction of mented (IUCN Red List Categories and Criteria: Version 3.1. being peculiar and rare (e.g. Macrovipera schweizeri (Werner, Second edition), then V. walser appears to be among the most 1935), has led to the illegal collection of numerous individuals threatened vipers in the world. The new taxon’s sister species for the international pet trade (Nilson et al. 1999, http://www.iuc- V. darevskii, with area of occupancy <10 km2, is now listed as nredlist.org), causing local extinctions. Because this species

Fig. 9. Habitat of Vipera walser sp. nov. Valle Strona at about 1650 m (upper left), Valle Strona at about 1800 m (upper right), Valle Mastallone at 2070 m (bottom left) and Valle della Vecchia at 1830 m (bottom right) J Zool Syst Evol Res (2016) 54(3), 161--173 © 2016 Blackwell Verlag GmbH Description of Vipera walser 171 occurs only in Italy, we strongly suggest that a specific legal pro- Conclusion tection for the species should be implemented very quickly. The present study described and named a new viper species, V. walser, which shows strong genetic divergence and clear mor- Longer term prospects and climatic change phological differentiation from all other known European viper species. The new taxon occurs in a restricted area of the south- Of course, it can be argued that V. walser, as a restricted-range fi relic species, is likely heading down an evolutionary dead-end western Italian Alps and shows close af nities with the Cau- path (Allendorf and Luikart 2007), in the sense of Darwin’s casian species V. dinniki, V. darevskii and V. kaznakovi, opening ‘wreck of ancient life’ (Darwin 1859) or Jeannel’s ‘fossiles unexpected and interesting biogeographic scenarios. The very vivants’ (Jeannel 1943). Its eventual natural extinction may take small extent of occurrence of the new species implies a particu- many millennia, but its ability to survive the next 100 years may larly high threat level, and thus conservation managements hang on two important aspects of its biology. First, there is a real should be developed. The protection of its habitat, the limitation lack of genetic variability within the population as compared to of the forest regrowth, but also the evaluation of its likely future that in other vipers (e.g. Ursenbacher et al. 2006a,b; Ferchaud distribution given climatic changes (for the long term) or struggle et al. 2011). The population is already fragmented into two main against culling (short term) are key elements to investigate. subpopulations, and, presumably, the complex topography of Involvement of local authorities, foundations and other stake- ridges and valleys may work to further isolate populations, as in holders will be crucial in realizing effective protection of this V. berus (Ursenbacher et al. 2009). A high priority for future species. study is an examination of habitat suitability at the landscape scale coupled with research on dispersal mechanisms and ability Acknowledgements in the species. Second, and related to the above, its ability to withstand or We are grateful to Paolo Eusebio Bergo, Gianluca Danini and Tiziano adapt to climatic change expected to take place within its range Pascutto for the useful information they have provided. Thanks to Ana Rodriguez Prieto for her invaluable help in the laboratory. Special thanks will be crucial. The current habitat of V. walser is restricted to 2 go to Roberta Salmaso of the Natural History Museum of Verona, Italy, an area of around 800 km within a few valleys, which experi- and Giuliano Doria, Director of the Natural History Museum ‘Giacomo ence some of the highest rainfall in the Alps (Mercalli et al. Doria’ of Genova, Italy, for facilitating the examination of specimens in 2008). Point estimates of annual rainfall from presence locations his care and for the loan of the type series. Thanks to Sebastiano Salvidio within its area of occupancy range from 1018 to 1604 mm for his critical reading of the manuscript and contribution to the statistical (mean = 1348 mm 111 SD) and mean minimum temperature analyses. Finally, the manuscript was greatly improved following com- between May andÆ October from 3.1 to 10.0°C ment by three anonymous reviewers. (mean = 6.1 1.2°C SD). Climate models (CMIP5 IPPC Fifth Æ Assessment; www.worldclim.org) indicate that in the next References 20 years, these valleys will become far wetter (mean = 1536 mm 129 SD) and warmer (mean = 8.5 1.2°C SD; Fig. 10). Allendorf FW, Luikart G (2007) Conservation and the Genetics of Consequently,Æ species distribution modelling,Æ and how this distri- Populations. Blackwell Publishing, Malden, Massachusetts, USA, pp bution might change under realistic climate change scenarios, 642. especially including the influence of habitat and habitat change Anderson MJ (2001) A new method for non-parametric multivariate analysis of variance. Austral Ecol 26:32–46. and dispersal ability (e.g. Pearson and Dawson 2003), is clearly Arevalo E, Davis SK, Sites JW (1994) Mitochondrial DNA sequence a priority. divergence and phylogenetic relationships among eight Chromosome races of the Sceloporus grammicus complex (Phrynosomatidae) in Central Mexico. Syst Biol 43:387–418. Avcy A, Cßetin Ilgaz C, Baþkaya Þ Baran Y, Kumlutaþ Y (2010) Contribution to the distribution and morphology of Pelias darevskii (Vedmederja, Orlow et Tuniyev 1986) (Reptilia: Squamata: Viperidae) in Northeastern Anatolia. Russ J Herp 17:1–7. Balmford A, Jayasuriya AHM, Green MJB (1996) Using higher-taxon richness as a surrogate for species richness: II. Local applications. Proc Roy Soc Lond B 263:1571–1575. Beebee TJC (2008) Buccal swabbing as a source of DNA from squamate reptiles. Conserv Gen 9:1087–1088. Bickford D, Lohman DJ, Sodhi NS, Ng PKL, Meier R, Winker K, Ingram KK, Das I (2007) Cryptic species as a window on diversity and conservation. Trends Ecol Evol 22:148–155. Capra F (1954) La Vipera berus L. in Piemonte. Annali del Museo civico di storia naturale di Genova 66:301–312. Carlson BZ, Renaud J, Biron PE, Choler P (2014) Long-term modelling of the forest-grassland ecotone in the French Alps: implications for land management and conservation. Ecol Appl 24:1213–1225. Clement M, Posada D, Crandall KA (2000) TCS: a computer program to estimate gene genealogies. Mol Ecol 9:1657–1660. Cornetti L, Belluardo F, Ghielmi S, Giovine G, Ficetola GF, Bertorelle G, Vernesi C, Hauffe HC (2015) Reproductive isolation between oviparous and viviparous lineages of the Eurasian common lizard Zootoca vivipara in a contact zone. Biol J Linn Soc 114:566–573. Darriba D, Taboada GL, Doallo R, Posada D (2012) jModelTest 2: more Fig. 10. Current and projected (2035; CMIP5 [IPPC Fifth Assessment]) models, new heuristics and parallel computing. Nat Methods 9:772– mean annual rainfall and mean minimum temperatures (°C) for the 772. months May–October within the current range of Vipera walser sp. nov. Darwin C (1859) On the Origin of Species. John Murray, London, UK. J Zool Syst Evol Res (2016) 54(3), 161--173 © 2016 Blackwell Verlag GmbH 172 GHIELMI,MENEGON,MARSDEN,LADDAGA and URSENBACHER

Dayrat B (2005) Towards integrative taxonomy. Biol J Linn Soc 85:407– Miralles A, Vences M (2013) New metrics for comparison of taxonomies 415. reveal striking discrepancies among species delimitation methods in Dowling HG (1951) A proposed standard system of counting ventrals in madascincus lizards. PLoS ONE 8:e68242. snakes. British J Herp 1:97–99. Nilson G, Andren C (1984) Systematics of the Vipera xanthina complex Ferchaud AL, Lyet A, Cheylan M, Arnal V, Baron JP, Montgelard C, (Reptilia: Viperidae). 2. An overlooked viper within the xanthina Ursenbacher S (2011) High genetic differentiation among French species-group in Iran. Bonner Zoologische Beitr€age 35:175–184. populations of the Orsini’s viper (Vipera ursinii ursinii) based on Nilson G, Tuniyev BS, Orlov NL, Hoggren€ M, Andren C (1995) mitochondrial and microsatellite data: implications for conservation Systematics of the vipers of the Caucasus: polymorphism or sibling management. J Hered 102:67–78. species? Asiatic Herpetol Res Berkeley 6:1–26. Ferchaud AL, Ursenbacher S, Cheylan M, Luiselli L, Jelic D, Halpern B, Nilson G, Andren C, Ioannidis Y, Dimaki M (1999) Ecology and Major A, Kotenko T, Keyans N, Behrooz R, Crnobrnja-Isailovic J, conservation of the Milos viper, Macrovipera schweizeri (Werner, Tomovic L, Ghira I, Ioannidis Y, Arnal V, Montgelard C (2012) 1935). Amphibia-Reptilia 20:355–375. Phylogeography of the Vipera ursinii complex (Viperidae): Orlov NL, Tuniyev BS (1990) Three species in the Vipera kaznakovi mitochondrial markers reveal an east–west disjunction in the complex (Eurosiberian Group) in the Caucasus: their present Palaearctic region. J Biogeogr 39:1836–1847. distribution, possible genesis and phylogeny. Asiatic Herpetol Res Garbarino M, Sibona E, Lingua E, Motta R (2014) Decline of traditional Berkeley 3:1–36. landscape in a protected area of the southwestern Alps: the fate of Padial JM, Miralles A, De la Riva I, Vences M (2010) The integrative enclosed pasture patches in the land mosaic shift. J. Mt Sci-Engl future of taxonomy. Front Zool 7:1–14. 11:544–554. Pearson RG, Dawson TP (2003) Predicting the impacts of climate change Geniez F, Teynie A (2005) Discovery of a population of the critically on the distribution of species: are bioclimate envelope models use. endangered Vipera darevskii Vedmerdeja, Orlov & Tuniyev, 1986 in Glob Ecol Biogeogr 12:361–371. Turkey, with new elements on its identification (Reptilia: Squamata: Polasky S, Camm JD, Solow AR, Csuti B, White D, Ding R (2000) Viperidae). Herpetozoa 18:25–33. Choosing reserve networks with incomplete species information. Biol Ghielmi S, Eusebio Bergo P, Andreone F (2006) Nuove segnalazioni di Conserv 94:1–10. Zootoca vivipara Jaquin e di Vipera berus Linnaeus, in Piemonte, Pons J, Barraclough TG, Gomez-Zurita J, Cardoso A, Duran DP, Hazell Italia nord-occidentale (Novitates Herpetologicae Pedemontanae II). S, Kamoun S, Sumlin WD, Vogler AP (2006) Sequence-based species Acta Herpetol 1:29–36. delimitation for the DNA taxonomy of undescribed insects. Syst Biol Goc€ßmen B, Konrad M, I_gci N, Akman Z, Yıldız MZ, Oguz MA, AltınC 55:595–609. (2014) New locality records for four rare species of vipers (Reptilia: Rambaut A, Suchard MA, Xie D, Drummond AJ (2014) Tracer v1.6, Viperidae) in Turkey. Zool Middle East 60:306–313. Available from http://beast.bio.ed.ac.uk/Tracer. Guillon M, Guiller G, DeNardo DF, Lourdais O (2014) Microclimate R Core Team (2014) A language environment for statistical computing, preferences correlate with contrasted evaporative water loss in R Foundation for Statistical Computing, Vienna, Austria, 2013, ISBN parapatric vipers at their contact zone. Can J Zool 92:81–86. 3-900051-07-0. [Online]. Available: http://www.R-project.org. Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to Saint Girons H (1978) Morphologie externe comparee et systematique estimate large phylogenies by maximum likelihood. Syst Biol 52:696– des Viperes d’Europe (Reptilia, Viperidae). Revue Suisse de Zoologie, 704. Geneve 85/3:565–595. Hammer Ø, Harper DAT, Ryan PD (2001) PAST: paleontological Scali S, Gentilli A (1998) Morphometric analysis, sexual dimorphism and statistics software package for education and data analysis. Palaeontol distribution of extinct adders (Vipera berus) of the Po Plane (Northern Electron 4:1–9. Italy). In: Miaud C, Guyetant R (eds), Current Studies in Herpetology, Howard PC, Viskanic P, Davenport TRB, Kigenyi FW, Baltzer M, Le Bourget du Lac, France, S.H.E., pp 391–396. Dickinson CJ, Lwanga JS, Matthews RA, Balmford A (1998) Sindaco R, Doria G, Razzetti E, Bernini F (eds) (2006) Atlante degli Complementarity and the use of indicator groups for reserve selection Anfibi e dei Rettili d’Italia/Atlas of Italians Amphibian and Reptiles. in Uganda. Nature 394:472–475. Societas Herpetologica Italica, Edizioni Polistampa, Firenze, Italy. Huelsenbeck JP, Ronquist F (2001) MrBayes: Bayesian inference of Stumpel€ N (2012) Phylogenie und Phylogeographie eurasischer Viperinae phylogeny. Bioinformatics 17:754–755. unter besonderer Berucksichtigung€ der orientalischen Vipern der Huelsenbeck JP, Imennov NS (2002) Geographic origin of human Gattungen Montivipera und Macrovipera. Unpublished thesis, mitochondrial DNA: Accommodating phylogenetic uncertainty and Technischen Universit€at Carolo-Wilhelmina, Braunschweig, model comparison. Syst Biol 51:155–165. Deutschland. IUCN (2014) The IUCN Red List of Threatened Species. Version 2014.3. Surget-Groba Y, Heulin B, Ghielmi S, Guillaume CP, Vogrin N (2002) http://www.iucnredlist.org. Downloaded on 17 November 2014. Phylogeography and conservation of the populations of Zootoca Jeannel R (1943) Les Fossils Vivants des Caverns. Gallimard, Paris, vivipara carniolica. Biol Conserv 106:365–372. France. Surget-Groba Y, Heulin B, Guillaume CP, Puky M, Semenov D, Orlova Joger U, Stumpel€ N (eds) (2005) Handbuch der Reptilien und V, Kupriyanova L, Ghira I, Smajda B (2006) Multiple origins of Amphibiens Europas. Band 3/IIB, Schlangen (Serpentes) III viviparity, or reversal from viviparity to oviparity? The European (Viperidae). Aula-Verlag, Wiesbaden (Germany), 420 pp. common lizard (Zootoca vivipara, Lacertidae) and the evolution of Kalyabina-Hauf S, Schweiger S, Joger U, Mayer W, Orlov N, Wink M parity. Biol J Linn Soc 87:1–11. (2004) Phylogeny and systematics of adders (Vipera berus complex). Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular Mertensiella 52:439–459. Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol Larson DM (2014) Grassland fire and cattle grazing regulate reptile and Biol Evol 24:1596–1599. amphibian assembly among patches. Environ Manage 54:1434–1444. Townsend TM, Larson A, Louis E, Macey RJ (2004) Molecular Lenk P, Kalyabina S, Wink M, Joger U (2001) Evolutionary phylogenetics of Squamata: the position of snakes, amphisbaenians, and relationships among the true vipers (Reptilia: Viperidae) inferred from dibamids, and the root of the squamate tree. Syst Biol 53:735–757. mitochondrial DNA sequences. Mol Phylogenet Evol 19:94–104. Townsend TM, Alegre RE, Kelley ST, Wiens JJ, Reeder TW (2008) Lessona M (1879) Intorno al Pelias berus in Piemonte. Atti della Reale Rapid development of multiple nuclear loci for phylogenetic analysis Accademia delle Scienze di Torino 14:748–749. using genomic resources: an example from squamate reptiles. Mol Mazerolle MJ, Bailey LL, Kendall WL, Royle JA, Converse SJ, Nichols Phylogenet Evol 47:129–142. JD (2007) Making great leaps forward: accounting for detectability in Tuniyev B, Nilson G, Agasyan A, Orlov N, Tuniyev S (2009) Vipera herpetological field studies. J Herpetol 41:672–689. darevskii. The IUCN Red List of Threatened Species. Version 2014.3. Mercalli L, Cat Berro D, Acordon V, Di Napoli G (2008) Cambiamenti www.iucnredlist.org. Downloaded on 28 April 2015. climatici sulla montagna piemontese. Rapporto tecnico realizzato da Ursenbacher S, Sasu I, Rossi M, Monney JC (2005) Are there Societa meteorologica Subalpina per conto di Regione Piemonte. Societa morphological differences between two genetically differentiated Meteorologica Subalpina Castello Borello, Bussoleno (TO), Italy. clades in the adder Vipera berus berus?. In: Ananjeva N, Tsinenko O

(eds), Herpetologica Petropolitana. Proc. of the 12th Ord. Gen.Meeting Werner F (1935) Reptilien der Ag€ €aischen Inseln. Sitzungsber Akad Wiss Soc. Eur. Herpetol., August 12 – 16, 2003, St. Petersburg, Russ, Wien 144:81–117. J. Herpetol., 12(Suppl.): 96–98. Zhang J, Kapli P, Pavlidis P, Stamatakis A (2013) A general species Ursenbacher S, Carlsson M, Helfer V, Tegelstrom€ H, Fumagalli L delimitation method with applications to phylogenetic placements. (2006a) Phylogeography and Pleistocene refugia of the adder (Vipera Bioinformatics 29:2869–2876. berus) as inferred from mitochondrial DNA sequence data. Mol Ecol Zinenko O, Turanu V, Stugariu A (2010) Distribution and morphological 15:3425–3437. variation of Vipera berus nikolskii Vedmederja, Grubant et Rudaeva, Ursenbacher S, Conelli A, Golay P, Monney J-C, Zufﬁ MAL, Thiery G, 1986 in Western Ukraine, The Republic of Moldova and Romania. Durand T, Fumagalli L (2006b) Phylogeography of the asp viper Amphibia-Reptilia 31:51–67. (Vipera aspis) inferred from mitochondrial DNA sequence data: evidence for multiple Mediterranean refugial areas. Mol Phylogenet Evol 38:546–552. Supporting Information Ursenbacher S, Monney JC, Fumagalli L (2009) Limited genetic Additional Supporting Information may be found in the online diversity and high differentiation among the remnant adder (Vipera version of this article: berus) populations in the Swiss and French Jura Mountains. Conserv Genet 10:303–315. Figure S1 Phylogenetic reconstruction for each gene sepa- Vacher J-P, Geniez M (coords), (2010) Les Reptiles de France, Belgique, rately Luxenbourg et Suisse. Biotope, Meze (Collection Parthenope); Table S1 PERMANOVA (Permutational Multivariate Analysis of Museum national D’Histoire naturelle, Paris, 544p. Variance. Vives GS, Miras Y, Riera S, Julia R, Allee P, Orengo H, Paradis- Appendix S1 List of morphologically analysed samples of Grenouillet S, Palet JM (2014) Tracing the land use history and Vipera berus and Vipera walser. vegetation dynamics in the Mont Lozere (Massif Central, France) during the last 2000 years: the interdisciplinary study case of Countrasts peat bog. Quat Int 353:123–139.