Introduction. Studies of geographic variation within and among. closely related taxa have provided crucial information

Transcription

1 Contributions to Zoology, 68 (3) (1999) SPB Academic Publishing bv. The Hague Geographic variation taxonomy of crested newts (Triturus cristatus superspecies): morphological mitochondrial DNA data J.W. Arntzen¹ & Graham+P. Wallis ² 1 Department of Zoology Anthropology, Faculty of Sciences, University of Porto, CECA/1CETA/UP, Campus Agrdrio de Vairao, 4480 Vila do Conde, Portugal; 2 Department of Zoology Centre for Gene Research, University ofotago, P.O. Box 56, Dunedin, New Zeal Keywords : asymmetric hybridisation, biogeography, crested newt, geographic variation, morphology, mtdna, taxonomy, Triturus cristatus Abstract Introduction Within the newt genus Triturus, the largebodied species in the cristatus (crested newt) superspecies show an unusual degree ofvariation in relative trunk length as a result of amongtaxon variation in interlimb vertebral count. Here we examine the systematic value of this feature as assessed by both exterior measurement (Wolterstorff Index) direct radiographic count of ribbearing vertebrae, with particular reference to a number of confoundingfactors (sex differences,hybridisation, geographic variation, allometry, preservation effects). Using our mtdna haplotype data, which are largely concordant with geographic distributionof species, wefind that direct count ofthe ribbearing vertebrae performs more reliably (14% misclassification) than external measurement (31% misclassification) as a species identifier. We therefore recommend this feature as a taxonomic tool, although (like external measurement) it breaks down near hybrid zones. To account for the observed biogeographicalpattern phenotype genotype discrepancies, a scenario is presented that combines the movement ofthe contact zone between taxa with asymmetric hybridisation. This scenario applies to species interactions in eastern Yugoslavia western France. Studies of geographic variation within among closely related taxa have provided crucial information on the nature of species the process of speciation (Mayr, 1963). Studies of spatial morphological heterogeneity are of interest from the point of view of adaptation since differences in morphology are likely to be functional or at least have consequences on fitness ecology. A proper analysis of morphological variation requires independent assessment of the phylogenetic relationships of morphotypes (e.g. Losos, 1990). A central question to the systematist is: does geographically bounded, morphological differentiation reflect local adaptation (in which case would be convergence expected to be common), or is it a function of phylogeny (in which case morphotypes should be monophyletic) (Harvey Pagel, 1991; Bernatchez, 1995)? Adaptation phylogeny are not, of course, Contents mutually exclusive explanations of character distributions. Adaptive morphological change can occur be proliferated by cladogenesis in the Introduction 181 Material methods 182 Results 183 Discussion 193 Performance of WI RBV for species diagnosis 193 Adaptation 194 Taxonomy phylogeny 194 Distribution biogeography 195 Acknowledgements 197 References 197 Appendix 199 absence of reversal. To resolve this, the first step is to determine whether different character sets define the same taxa. Concordance of taxon boundaries resolved by different means, e.g., morphology genetics, is strong evidence for their reality (Hillis, 1987; Avise Ball, 1990). Indeed, morphological divergence often only becomes apparent after the taxa have been resolved genetically (e.g., GarciaParis, 1995; Arntzen McDowall Wallis, Here we 1996). look at

2 Geographic 182 J. W. Arntzen & G. P. Wallis variation in crested newt superspecies of the five phenotypes was associated with one or geographic variation in vertebral count in the Triturus cristatus superspecies (or Artenkreis sensu more characteristic mitochondrial genotypes (Wallis Rensch) reevaluate this character as a tool for species identification, with particular reference to the several hybrid zones in present the group our genetic analyses of these zones (Wallis Arntzen, 1989; Arntzen Wallis, 1991; JW Arntzen, in prep.). In his classic Wolterstorff paper, (1923) revised the taxonomy of the crested newt, Triturus cristatus index. using a morphological He distinguished four Arntzen, 1989). Significant differences in WI were observed between groupscarrying these genotypes, therefore WI can be used as a tool for species diagnosis. Limb length interlimb distance are sexually dimorphic characters, making it necessary to calibrate WI for males females separately. For technical mtdna reasons, data were available for (almost exclusively) femalenewts only (Wallis, 1987). However, we measured WI for males taxa at the subspecies level [carnifex (Laurenti, from the same populations because (with a few exceptions) they belong to the same species, WI 1768), cristatus (Laurenti, 1768), dobrogicus (Kiritzescu, 1903) karelinii (Strauch, 1870)] can be calibrated for males as well. The number of provided compelling phenotypic descriptions for ribbearing vertebrae (RBV) varies in parallel with each of these. He provided a crucial discriminator the inverse of WI (Arntzen Wallis, 1994; for taxon identification, which has become known CrnobrnjaIsailovic et ah, 1997). Similar patterns as the Wolterstorff Index (WI). WI is the ratio have been documented for lizards (Greer, 1987; between forelimb length (PaL) Griffith, 1990; Caputo, Lanza Palmieri, 1995), interiimb distance (LiE) is defined as WI = PaL * 100 / LiE. We have reviewed the available data on WI fish (Lindsey, 1975) other salamers (Jockush, 1997). have discussed some of the practical theoretical pitfalls associated with its use (Arntzen Here we present additional data to calibrate WI RBV document the existence of geographic Wallis, 1994). WI succinctly variation across within species. The results summarises the morphological differentiation among taxa to much the confirm the limited value of WI for diagnostic same extent as more sophisticated methods of purposes. The character RBV on the other h is multivariate morphometric analysis used in crested newts other urodeles (Kalezic et ah, 1990; shown to be diagnostic without much intraspecific geographic variation. Moreover, it appears to Arntzen Wallis, 1994; Sket Arntzen, 1994; be largely Arntzen Sket, 1997; Cvetkovic, Kalezic immune to observer bias, developmental or sexual variation, or preservation artefacts. Dzukic, 1997). From a statistical stpoint, indices may have certain undesirable properties (Atchley Applying the character at the population level produces a coherent pattern of geographic variation et ah, 1976), but the WI has become a popular tool that coincides with that of the species, although for species identification of crested newts in the (like external measurements) it breaks down near field (for example Grillitsch et ah, 1983). How hybrid zones. To account for the observed pheno ever, to avoid circular reasoning, the calibration of type genotype discrepancies biogeographical patterns in eastern Yugoslavia western WI requires independent criteria, i.e., an identification without reference to body shape. France, a scenario is presented that combines the In the crested newt superspecies its sister movement of the contact zone between taxa with taxon marmoratus (Latreille, WI 1800), increases asymmetric hybridisation. in the order: dobrogicus cristatus carnifex karelinii marmoratus, describing a morphological series from slender shortlegged to stout Material methods longlegged. The taxa also possess characteristic patterns of colouration that are described on Forelimb length interlimb distance weremeasured with plastic Plates III (for approximate species distributions callipers (0.1 mm precision) on preserved or live material, see Fig. I). We mtdna RFLP previously gathered data from female newts concluded that each representing 142 populations across the total range. Xray photographs were taken on preserved or sedated specimens (representing 116 populations) with an Elinax 90/20 an

3 Contributions to Zoology, 68 (3) Fig. curve separating species that are adjacent in mor I. Distribution of Triturus marmoratus five taxa in the cristatus superspecies (after Arntzen, 1995). The crossed hatching covering Yugoslavia adjacent regions refers to carnifex macedonicus, a taxon recognition resurrected from Karaman (1922). exposure of0.7 seconds at 38kV, 4mA on AgfaGevaert D10DW Xray film. Additional data on WI (31 populations) RBV (38 populations) were obtained from the literature(vallee, 1959; Kalezic et ah, 1990; CrnobrnjaIsailovic et ah, 1997). Analysis of variance (ANOVA) othertests followed Sokal Rohlf (1981) Siegel Castellan (1988). Logistic equations were determinedwith SPSS (SPSS, 1990),Nonparametric tests were performed with StatView (StatView, 1988), Association of intraspecific morphological variation geographical distance was tested for with the Manteltest (NTSYS 1.80, Rohlf, 1993). from populations characterised by mtdna genotype (Table I). For reference, we include published data from studies dealing with three or more taxa that were identified phenotypically (Table I). Statistically significant differences were found between all species combinations for WI (four onetailed t tests between consecutive groups for each sex: P < in all cases, except for males cristatus carnifex males carnifex karelinii with P < 0.05 females karelinii marmoratus with P < 0.01) for RBV (four Results Median Gtests between five consecutive groups with P < in all cases). Males consistently have an average WI of 810 points greater than We present estimates ofthe Wolterstorff Index (Fig. females of the same species. 2) the number of ribbearing vertebrae in newts A weighted regression was used to determine the best fitting logistic

4 Geographic 184 J. W. Arntzen & G. P. Wallis variation in crested newt superspecies

5 Contributions to Zoology, 68 (3) Table I. Number of rib bearing marbled newts, genus vertebrae observed in crested Mean ± SD; sample Herre, Lanza et al., present study size in brackets. The populations used in the present study were idententified to the species onphenotype mitochondrial DNA (Wallis & Arntzen, 1989). Triturus. dobrogicus 17(3) 17.6 ±0.51 (14) 17.5 ±0.58 (47) cristatus 16(3) 16.0 ±0.15 (44) 16.0 ± (122) carnifex 15(3) 15.1 ±0.30(148) 14.9 ± 0.33 (26) ±0,30 karelinii ± 14(1) 14.1 ±0.30 (40) 14.2 ±0.44 (41) dobrogicus from all other On species. the basis of WI, three males (4.2%) three females (2.9%) would have been classified as belonging to a species with noncorresponding ( cristatus) phenotype. Much higher levels of misclassification were obtained over all species (31%), with no significant difference in misclassification between the sexes (P > 0.05; Gtest of independence). Using RBV, is obtained between: good separation dobrogicus cristatus (6.6% misclassified), cristatus carnifex (6.8%), carnifex karelinii (16.4%) karelinii marmoratus (6.7%). Overall misclassification is marmoratus 13, ±0,34 (49) 13.7%. Application of the RBV cutoff point criteria to population means (Table II) provides inferred spe phometric space. The WI for typical each crested newt species is as follows: males: < dobrogicus 54.0, cristatus , carnifex , karelinii >67.1; females; dobrogicus < 46.2, cristatus , carnifex , karelinii > For RBV the modal values observed are: 13 in marmoratus, 14 in karelinii, 15 in carnifex, 16 in cristatus 17 or 18 in dobrogicus. Limited intraspecific variation is observed with 86% of the investigated newts showing modal values. Ranges for the identification of populations are set as follows: dobrogicus > 16.5, cristatus , carnifex , karelinii marmoratus < cies distributions that are largely internally consistent, in line with prior knowledge, hence reflect documented distributions with some exceptions as follows. 1 Populations 2 are inferred karelinii marmoratus, but are within the known geographic phenotypic ranges of marmoratus karelinii, respectively (Fig. 3). Similarly, populations 35 are obviously wrongly classified ( dobrogicus within the cristatus karelinii within the range range of carnifex carnifex). Sites 68 are known syntopic cristatus cristatus marmoratus dobrogicus populations with bimodal character state distributions (Vallee, 1959; Median Fisher s exact tests demonstrated the ab Wallis Arntzen, 1989; Arntzen Wallis, sence of significant sexual dimorphism for this 1991; Arntzen et al., 1997). Individuals with mixed character, except in dobrogicus where females dohrogicus carnifex phenotype at site 9 tendto have a higher count than males (modal RBV of 18 versus 17, P < A trend 0.05). was observed for increasing overlap of character states between species with increasing WI decreasing RBV. WI alone is usually sufficient to distinguish 10 had RBV scores typical for cristatus{ i.e., in between the values typical for dohrogicus carnifex) hence remain unclassified in Fig. 3. Populations were in the field identified as carnifex, showed RBV counts typical Fig. 2. Reported values ofthe WolterstorffIndex for seven studies that involved three or more taxa in the crested newt superspecies. W = Wolterstorff (1923): range, interrupted lines refer to extreme values; H = Herre (1932): mean ± SD range; F = Fachbach (1974): range; K = Kalezic Stevanovic (1980) Kalezic et al. (1990: tables 1 to 3 on page 3435,with populations allocated to species following Kalezic et al, (1997) but excluding the population from Tresnja which Wallis Arntzen (1989) identified as a genetically mixed dohrogicus kareliniipopulation: average weighted mean± SD range; L Lanza, Gentile Torricelli = (1991): mean range; A = Arntzen Wallis (1994): mean ± SD range; PI P2 = present paper: mean ± SD range. Populations figuring in P2 following the species distributionsas in Fig. 3. Solid bars refer to were identified aposteriori, populations classified on the basis mitochondrial DNA genotype. Sample sizes (if known) are presented at the top of the bars. Horizontal bars indicate optimal interspecies boundaries with respect to W1 (details see text). Data on marmoratus are given for comparison.

6 186 W. Arntzen & G. P. Wallis Geographic variation in crested newt superspecies

7 adjacent Contributions to Zoology, 68 (3) Fig. 3ABC. Newt populations for which mean NVR (number of rib bearing vertebrae) mtdna haplotype was determined. Solid lines connect records describing the edge of species ranges: car Triturus carnifex carnifex, cri = = cristatus, dob dobrogicus, = kar = karelinii, mac carnifex macedonicus. = Numbered populations are discussed in the text. No bigbodied newts are known for the areaapproximately coinciding with BosniaHercegovina. The area marked with a *? in northwestern Bulgaria parts of Yugoslavia is not devoid of crested newts (see Kalezic et ah, 1997), but the taxon to which they belong is undetermined. The interrupted heavy line on map C describes an area with newt populations possessing mtdna characteristic for karelinii (Wallis Amtzen, 1989); studied populations shown by triangular symbols. Arrows refer to hypothesized dispersal events discussed in the text. Note that the River Danube connects the Pannonian Dobrogean Plains through the Iron Gate. For the comprehensive distribution of dohrogicus see Arntzen et ah, for cristatus, an mtdna haplotype typical for lively low RBV scores (populations 3033, Fig. dohrogicus also remain unclassified. Indi 3c; mean RBV < 15.9) are all situated at the fringe viduals with mixed carnifex karelinii phe of the cristatus range, close to karelinii which notype (site 13) were classified as carnifex on typically has low RBV counts. Among the basis of RBV. Similarly, a newt with mixed 16 dobrogicus populations with N < 5, cristatus karelinii (site 14) phenotype was those with relatively low counts (mean RBV < 17. 5, see Table classified as karelinii. Newts from sites 1520 are phenotypically karelinii, but have RBV scores II) are, with one exception (population Alap) situated along the southern edge of the dohrogicus higher than average for that species. Sites 619 are all located at the fringe of species distributions range, whereas those with higher counts (mean RBV > 17.5) are, with one exception (population Jamena) close to another species, where one might expect situated away from the edge of the species range. interbreeding if the various taxa of crested newts Application were not genetically isolated. No dohrogicus or of the W1 criterion to the identification ofindividuals populations produces mixed cristatus samples (except for those mentioned results. For dohrogicus, the distribution pattern above) had RBV scores that did not correspond with obtained with W1 identifications neatly fits the phenotype. However, among the larger (N 2 documented distribution, whereas for the other 5) samples, four cristatus populations with rela species the results are erratic (compare Table II

8 188 J. W.Arntzen & G. P. Wallis Geographic variation in crested newt superspecies Table II. Sample site coordinates, sample size (N), mean Wolterstorffindex (WI), mean number of ribbearing vertebrae (RBV) mitochondrial DNA haplotypes observed in discussed in the text marked in Figure 3, Triturus cristatus superspecies in marmoratus. Numbers in brackets refer to sites Italicized values are outside the range typical for the taxon (no range was identified for WI in marmoratus) or atypical for the taxon (mtdna haplotypes). Identifications were made by phenotype. coordinates males females Phenotype (see Plates IIII) III1) latitude longitude N WI N WI Wl N RBV N Haplotype ft f. Triturus carnifex carnifex Acerra, Napoli, Italy $ Belovar Moravce, Croatia $ (21) Benevento, Italy (5) Bobinaco, nr. I Aquila, Italy $ Etzmorf, nr. Eggenburg, Austria $(11) ) Farma, Torniella, Italy $ Firenze, Italy $ CARS 1 CAR 6 I1 CAR 7 Fuscaldo, Italy $ II Haidlhof, nr. Baden, Austria $ (38) ,20 3 DOB II Ig (Podstrmec), Slovenia * ,95 Istarske Toplice, Croatia * Kleinmeiseldorf, Austria $ (12) DOB II 11 Kramplje, nr. Nova Vas, Slovenia $ CAR 8 Lackcnbach, Austria $ ,60 Licki Osik, Croatia (23) * Locarno, Switzerl $ CAR 9 Pisa, Italy $ , CAR 5 Plitvice, Croatia (22) Radenci (Turjanci), Slovenia * Stanjel (Goce), Slovenia * Salakovci, Croatia * Sinac, Croatia $ I Velika Vala, Slovenia * = a= Zumberak (Budinjak), Croatia * , Taxon sample size weighted mean Triturus carnifex macedonicus Ano Kaliniki, Greece $ CAR? 10 BjcloSi, BjeloSi, F.R. Yugoslavia (27) * = = Dimitrovgrad, F. R. Yugoslavia (29) (47 \ Divcibare, F.R. Yugoslavia $ II Dobrsko Selo, F.R. Yugoslavia * Donja Cadjavica, Gornja Cadjavica, BosniaHcrccgovina BosniaHercegovina $ Donja Dubrava cq. Tuzla, Bosnia Hercegovina (24) * Donji Stoj, F.R. Yugoslavia * Galicica (Djafa), Macedonia * , Gornja Cadjavica, BosniaHercegovina $ ,27 Gornji Ceklin, F.R. Yugoslavia * 42 = 21 = Grcak, F. R. Yugoslavia $ (35) KAR? 12 Karan, F.R. Yugoslavia $ Karbinci, F.R. Yugoslavia $ Lesnovo, Macedonia * Livadia, Greece Lucane, F. R. Yugoslavia $ (36) ,67 12 KAR? 12

9 Contributions to Zoology, 68 (3) 189 Table II. Continued. coordinates males females Phenotype (see Plates II1I) IIII) longitude latitude N WI N WI Wl N RBV N Haplotype 0 Novo Brdo cq. Grmija, F.R. Yugoslavia * Petrovec, Macedonia * , Prilep, Macedonia * II Probistip, Macedonia $ Radosice, F.R, Yugoslavia * Rtanj, F.R. Yugoslavia * Sarajevo, BosniaFlercegovina BosniaFIercegovina (25) Stanisinci, F.R. Yugoslavia $ Tavna Monastire, BosniaHercegovina $ ,78 Valjevo (Bukovac), F.R. Yugoslavia * Visegrad, BosniaHercegovina $ Vlasina, F.R. Yugoslavia * ,78 Vranje (Sv. Ilija), F.R. Yugoslavia * Zelengora Mountain, BosniaHercegovina (26) ) Zupa (Rataje) cq. Rataje, F.R. Yugoslavia * ,94 Zupa, F.R. Yugoslavia * Taxon sample size weighted mean Triturus crislalus cristatus Ambleteuse, France $ Biel, Switzerl $ Bogdana, Romania Bor, F. R. Yugoslavia $ (37) KAR? 12 Bumbe tijiu, Romania Canterbury, UK $ Cimpeni, Romania $ , CR1 1 Craciune ti, Romania Cra na, Romania Doetinchem,Netherls $ (4) I Drighiu, Romania Galole ti, Romania Halma d, Romania Hoensbroek, Netherls $ Houthem, Netherls $ Hurezani, Romania HutaCerteze, Romania Jabukovac, F.R. Yugoslavia $ CRI 2 Jitia, Romania Klokocevac, F.R. Yugoslavia $ (30) , KAR? 12 Lanckorona, nr. Wadowice, Pol $ Limanowa, Pol $ CRI 1 Maidstone, UK CRI 1 Marghita, Romania Mayenne, France (6) Mayenne, France $ (6) CRI 1 Milanovac, F.R. Yugoslavia $ (31) Mogielnica, Pol * Negotin, F.R. Yugoslavia * ' , Ottenstein, nr. Zwettl, Austria $ CRI 1 Oxford, UK CRI 1

10 00 01 Geographic 190 JW.Arntzen & G. P. Wallis variation in crested newt superspecies Table II. Continued. coordinates males females Phenotype (see Plates II1I) liii) latitude longitude N WI N WI N RBV N Haplotype 0(4 Peterborough, UK $ (3) CRI 1 Picaturile, Romania ebi, Romania $ (7) , I1 CRI 1 Sinaia, Romania $ II CRI 1 St L6, France I1 CRI 1 St. Michielsgestel, Netherls $ ,00 Stubik, F.R. Yugoslavia * (32) , Stubik, F.R. Yugoslavia $ (33) , , CRI 2 Ta nad, Romania Tirgoviste, Romania $ , CRI I1 Turt, Romania I1 60,3 Videle, Romania $ CRI 3 I1 CRI 4 Vipere ti, Romania Virfuri, Romania $ (8) CRI 1 Winterswijk, Netherls $ Taxon sample size weighted mean Triturus dobrogicus (Pannonian) Alap, Hungary $ , I DOB 11 Albertirsa, Hungary $ I Andrid, Romania Babe ti, Romania I I Beograd, F.R. Yugoslavia $ DOB 11 Cegenydanyad, Hungary I 50.2 Debrc, F. R. Yugoslavia $ (34) DOB 11 KAR?!2 12 Drbsing, Austria $ Dugo Selo, Croatia $ Ecka, F.R. Yugoslavia $ , DOB 11 Ghcrtenis, Gherteni, Romania Glusci, F.R. Yugoslavia $ I1 DOB 11 Ivanovo, F.R. Yugoslavia * , Jamena, F.R. Yugoslavia $ I Kormend, Hungary $ Novi Knezevac, F.R. Yugoslavia * Obedska Bara cq. Obrez, F.R. Yugoslavia * ,24 Ocsod, Hungary $ I1 DOB 11 Opovo, F.R. Yugoslavia $ Orle, Croatia * Pi colt, Romania Podgorac, Croatia $ DOB 11 ebi, Romania $ (7) I DOB 11 Senta, F.R, Yugoslavia $ DOB 11 Slavonski Brod, Croatia * 45 II Svetozar Miletic, F.R, Yugoslavia * Szolnok, Hungary $ I Tadten, Austria $ Virfuri, Romania $ (8) I1 18, DOB 11 Vrsani, BosniaHercegovina $ DOB 11 Zupanja, Croatia $ ,00 4 DOB 11 '

11 Contributions to Zoology, 68 (3) Table II. Continued. coordinates males females Phenotype (see Plates IIII) latitude longitude N WI N WI N RBV N Haplotype Flaplotype 0 (Dobrogean) Amara ti de Jos, Romania Ariciu, Romania ,5 Hogioaia, Romania # II Kladovo, FR Yugoslavia $ Obrejita, Romania , Prahova Forest, N of Bucure ti, Romania , Svistov, Bulgaria $ DOB 11 Tartal, Romania / Zimnicea, Romania $ DOB Taxon sample size weighted mean Triturus karelinii (Europe) Arjelovac, F.R. Yugoslavia $ KAR? ICAR? 12 Bansko, Macedonia $ (19) Berovo cq. Smojmirovo, Macedonia * (18) Bigla, Macedonia $ , Dafnohori, Greece $ (20) Djurinci, F.R. Yugoslavia $ (16) Gornja Sabanta, F.R. Yugoslavia $ (2) I Grivac, F.R. Yugoslavia $ Guberevac, F.R. Yugoslavia $ KAR? 12 1 KAR? 14 Istanbul, Turkey $ KAR 17 Karlovo, Bulgaria $ , KAR? 12 Kentriko, Greece $ Kozuf (Visoka Cuka), Macedonia * Levski, Bulgaria $ KAR? 12 Mitrasinci, Macedonia $ Rakovski, Bulgaria $ KAR? 12 Resaskaica Pecina (Resavica), F.R. Yugoslavia $ (17) Sevlievo, Bulgaria $ Sisevac, F.R. Yugoslavia $ , Tresnja, F.R. Yugoslavia * (15) Tresnja, F. R, Yugoslavia $ (15) DOB II 11 3 KAR? 12 3 KAR? 13 (Asia) Adapazari, Turkey $ KAR? 15 1 KAR? 16 1 KAR 17 Bartin, Turkey $ Gokgekent, Turkey # Karacabey, Turkey $ Resadiye, Turkey # , Serafiye, Turkey # f 37' Tokat, Turkey # Taxon sample size weighted mean ,

12 Split, Loja, Geographic 192 J. W. Arntzen & G. P. Wallis variation in crested newt superspecies Table II. Continued. females (Gtest of independence, P < 0.001). Clas males coordinates females Phenotype (see Plates II1I) IIII) latitude longitude N WI N WI N RBV N Flaplotype Haplotype 0 Triturus marmoratus marmoratus Barcelona, Spain $ (1) Confolcns, Confolens, France $ I MAR 18 El Berrueco, valley of Rascafria, Spain $ Mayenne, France (6) Mayenne, France $ (6) , MAR 18 1 MAR 20 Rochechouart, France $ MAR 18 1 MAR 19 Salamanca, Spain $ Santillana del Mar, Spain Valongo, Portugal $ 41 II Vilar Formoso, Portugal $ , Triturus m. pygmaeus Archidona Spain $ Cadiz, Spain , Floyo Hoyo de Manzanares Torrelodones, Spain Puerto de Galiz, nr. Ubrique, Spain $ PYG 21 Rio Alberite, Spain $ Venta del Charco, Spain $ PYG 22 Villalba, Spain $ Taxon sample size weighted mean mixed Donja Cadjavica, BosniaHercegovina $ (10) Lukovo, F. R. Yugoslavia $ (14) Mayenne, France t (6) =48 = 18 = Mayenne, France $ (6) =48= 18 = CR1 CRI 1 Vedropolje, Croatia $ (9) Vitanovac, F. R. Yugoslavia $ (13) ,00 undetermined I1 Sibenik Split, Croatia =43 30= 16 = (28) 1^ = 00 i see Wallis & Arntzen (1989) Arntzen & Wallis (1991) $ voucher material deposited at the Institute for Systematics Population Biology (Zoological Museum), University of Amsterdam II Vallee (1959) * Kalezic et al. (1990) CrnobrnjaIsailovic et al. (1997) # data courtesy of D. Cogalniceanu K. Olgun 50% in carnifex 39% in karelinii. No with Fig. 1). On the basis of the reconstructed species distributions (Fig. 3) the frequency significant differences in classification scores are observed between the sexes, except for karelinii of incorrect identification of individual crested newts applying the WI can be estimated in post hoc manner for which males are more often misclassified than as 11 % in dobrogicus, 42% in cristatus, >

13 ~ Contributions to Zoology, 68 (3) sification success is not significantly affected by cacy of WI compromise classification made sample size, except in carnifex {P < 0.05, Mann solely on this basis. The WI purports to capture Whitney U test, onetailed). information useful in taxonomy Mantel tests indicate a marginally significant association between geographic distance WI but in fact confounds the variables limb length, vertebral number, possibly vertebral length. for males females cristatus male Using mtdna haplotype (Wallis Arntzen, dobrogicus {P while the association 0.05), of geographical distance with RBV is significant (P 1989) we showed that WI makes a good approximation to species classification, but can only be used with confidence for discriminating adult < 0.05 for cristatus P < 0.01 for dobrogicus). A series of ANOVA s with sex nested under group confirmed the difference in WI between the sexes {P < 0.001) between the following a priori identified groups (cf. Fig. 1): m. dobrogicus from the other species. In contrast, the numberof ribbearing vertebrae (RBV) as assessed by radiography eliminates all but one of these problems. Because RBV is a direct discrete meristic marmoratus versus m. pygmaeus (Wolterstorff, 1905) {P < 0.01), Pannonian versus Dobrogean dobrogicus (P < 0.01) carnifex carnifex versus count, stable the through lifetime of the individual, with limited intraspecific geographic differentiation its in use conjunction with a diagnostic genetic carnifex macedonicus (Karaman, 1922) (P < character leaves only the issue of hybridisation to be addressed. Variation caused by hybridisation near 0.05), whereas no significant difference was observed for Asian versus European karelinii. A regions of parapatry is difficult to disentangle from significant difference was also found between intraspecific geographic variation on the basis of carnifex carnifex carnifex macedonicus in morphological data alone. Under both scenarios, RBV (Median (7test, P < 0.05) but not between m. mannoratus in. pygmaeus (P > 0.05). The limited sampling of Dobrogean (versus variation will be most clearly expressed when samples from remote parts of the geographic range are compared. However, the observation that char Pannonian) dobrogicus Asian karelinii acter state changes are consistently in the direction (versus European karelinii) precluded their test to that of the neighbouring species supports the ing. hypothesis of hybridisation, rather than that of intrinsic geographic variation. For example, the significantly different values in carnifex Macedonians Discussion (high WI, low RBV compared to carnifex carnifex) may well be a result of introgression from Performance ofwi RBVfor species diagnosis karelinii, whose mtdna prevails in some carnifex Macedonians populations (Wallis In our previous work concerning the capacity of Arntzen, 1989). Note, however, that the meristic the Wolterstoff Index to discriminate females of the taxa, we identified several potential problems (Arntzen Wallis, 1994): 1) statistical representation (the need for a mean value), 2) nonbiological variation (preservation measurement differences), 3) sexual variation (WI is higher formales than for females), allometric variation(wi decreases with size), 4) geographic variation (nearby animals may tend to be more similar within species), 5) hybridisation (hybrids between two species can have intermediate values typical of a third species), 6) circular reasoning (the need for an independent character set to determine the significance of WI). count in hybrids is not necessarily intermediate to that of the parental species, as documented for salmonid fishes (Leary, Allendorf Knudsen, 1985). If hybridisation betweentaxa of crested newts is a common phenomenon, it should be possible to find genetic markers for the species covarying with interspecific morphological variation. The observed breakdown of the diagnostic power of RBV in areas where taxa meet may reflect the true nature of characters in a contact zone. It is possible that RBV is influenced early on in development. Indeed, vertebral count is often highly labile in fish salamers (e.g. McDowall, 1970; All of these factors to some extent reduce the effi Jockush, 1997), with cooler conditions generally

14 to J.W. Arntzen & G. P. Wallis Geographic variation in crested newt superspecies slowing development increasing annually approximately three months in the water, carnifex four months, cristatus five several meristic counts (Barlow, months (Bouton, 1986; Griffiths Mylotte, 1987; 1961 references in Jockush, 1997). In Triturus vulgaris the average RBV count increases with the at temperature which the embryos are raised (Orska Iraiolek, 1962), while for other salamer species more complicated environmental effects were found (Lindsey, 1966; Peabody Brodie, 1975). To address the question to what extent the variation in RBV is genetically determined to gain insight Andreone Giacoma, 1989), while the aquatic phase of lasts dobrogicus usually six months (Karaman, 1948; Jehle et ah, 1997). No data are available on the phenology of karelinii. This species is predicted to have a short aquatic phase, of intermediate length to that of carnifex marmoratus. In terms of the performance, sinusoidal swimming ability shouldbe best in the lowest into the relationship between embryonic development adult WI / highest RBV count taxon, he., dobrogicus morphology requires experimental work. Triturus relatively poor in marmoratus. Indeed, the dobrogicus might be the best species to work with ecological niche of dobrogicus is different from because it is naturally polymorphic for RBV. that of the other bigbodied species. It may coexist with fish in oxbows, river margins other nontemporary water bodies (Arntzen et ah, 1997). Adaptation The observed ventral aposematic coloration pattern in the cristatus superspecies versus the dorsal Elongation of the body a reduction in the length of limbs in vertebrates generally indicates a more piscine locomotion by sinusoidal body undulation. Although this can be associated with some terrestrial habitats (as in sswimming skinks), it is aposematic colouration of marmoratus (particularly evident in the entirely terrestrial juveniles) provides further to our support interpretations. Predictions about performance, such as in the gathering of food or predator avoidance behavior in more usually an adaptation to a more aquatic mode the aquatic versus terrestrial habitat could be tested experimentally. of life. Reduction of trunk size the development of robust legs is more unequivocally associated with a terrestrial mode of life where the body requires more support (Young, 1950;Le, 1978). Taxonomy phytogeny The wide variation in body shape in the legless Gymnophiona, ranging from stout to threadlike, The depth of the differences among taxa, the may also be associated with locomotory behavior relative sharpness of the contact zones led us to ecological adaptation (Renous Case, 1989). follow earlier suggestions to raise the taxa to full In the salamer genus Batrachoseps, extensive species status (Buccilnnocenti, Ragghianti geographic variation in RBV, possibly related to Mancino, 1983) as have others (Frost, 1985). The fossoriality, has been observed, most ofwhich was ' available data, unfortunately, do not support a single shown to be genetically determined (Jockush, 1997). for the phylogenetic hypothesis four taxa comprising Selection on female fecundity (correlated the cristatus superspecies. RBV is primitively with interlimb length) male sexual performance (stature at display correlated with leg length) may also play a role (Arntzen Wallis, 1994). 14 in the genus Triturus (B. Lanza et ah, in prep.), rendering RBV of 13 an autapomorphic character state for marmoratus RBV of 15 a Morphology predicts the aquatic period of synapomorphic character state series for carnifex marmoratus to be short, that of carnifex to be cristatus dobrogicus. This character alone intermediate that of cristatus dobro would suggest that karelinii represents the old gicus in particular be long. While of course est extant crested newt lineage, followed by carni the phenology of breeding may be different from fex, cristatus dobrogicus. This evolutionary year to year, this prediction appears to be corroborated by field data. Triturus marmoratus spends classification is supported somewhat ambiguously by the phenetic analysis of protein electrophoretic data (Crnobrnja, Kalezic Dzukic, 1989) but

15 Contributions to Zoology, (3) 195 contradicted by another such study (Litvinchuk et of the eastern group were described as Molge karelinii var. macedonica Karaman, Considering ah, 1994). The phylogenetic analysis of molecu the morphological genetic differentiation be lar data (mtdna RFLP s) suggests a different tween the forms, we propose raising this taxon to phylogeny. Looking at the mostparsimonious the subspecies mtdna tree (Wallis Arntzen, 1989: Fig. 4), a level, supported by phylogenetic arguments, classifying it as belonging to tree that optimizes RBV characterstate change [tree: character structure (MAR,PYG: 13)/ancestor: 14/ (KAR?,KAR: 14(CAR?,CAR: 15(CRI: 16,DOB: 17,18)))] carnifex (not karelinii as suggested by Karaman, 1922). Therewith, Triturus carnifex var. albanicus Dely, 1959 is a junior synonym of carnifex macedonicus (Karaman, 1922). Following involves moving only the DOB branch (with terminal taxon number 11). [CAR? KAR? refer to deeply differentiated haplotype lineages within our correspondence with coauthor J. CrnobrnjaIsailovic (J. W. Arntzen, in letter, 1996) this taxonomic solu carnifex karelinii. We now recognize the tion is accepted by Kalezic et al. (1997), although first of these as belonging to carnifex mace the taxon is incorrectly referred to in feminine donicus (see below) while the other will be subject to taxonomic description at the subspecific level (S. Litvinchuk et al., in. prep.)]. If DOB were placed with the cristatus (CRI) haplotypes, increased RBV becomes a derived character interior to the tree, with the more massive built newts basal. Although this haplotype tree has 70 steps as opposed to 67 in the published maximum parsimony tree (Wallis Arntzen, 1989: Fig. 4), there is no bootstrap support above 50% for any of the crested newt specieslevel structure. That is to say, the relationship DOB(KAR(CAR,CRI)) is only gender. The two more massive newt species, karelinii carnifex, show much greater restriction site variation than the other two species (Wallis Arntzen, 1989). They also have slightly larger mitochondrial genomes a greater tendency for insertions in the control region (Wallis, 1987). These factors that the two suggest northern species may have been subjected to longterm small population size during glaciations (Wallis Arntzen, 1989), it is conceivable that the evolutionary change in vertebral count is related to this popula defined by three synapomorphies in total the tion genetic feature. tree could more conservatively be depicted as a fourway polychotomy at this level. This incomplete resolution is appreciated by Wallis Arntzen Distribution biogeography (1989: 99) emphasised by further analysis (Faith Cranston, 1991; Faith, 1992). However, strong support is obtained from the CAR CAR? Crested newts appear to be absent from the largest part of BosniaHercegovina (see for example mtdna haplotypes for the sister taxon status of Schmidtler Schmidtler, 1983; Kalezic, Dzukic Italian central Balkan crested newts. These groups of populations are also united by the synapomorphic character state RBV = 15. We therefore Tvrtkovic, 1990; Kalezic et al., 1997). The southeasternmost localities of carnifex to the consider the crested newts from the central Balkan northeast of the perceived gap in the species distribution are sites 2123 [Belovar Moravce (Table to belong to carnifex. The range of this species is disjunct (see below). Newts from both parts of the range are phenotypically distinct: carnifex from the western (Italian Slovenian) of part the range typically have few, large, illdefined black dots on the bellies, whereas carnifex from the II); Plitvice (FejervaryLangh, 1943) Licki Osik (Kalezic et al, 1990)]. Further to the southeast carnifex is found at sites 2427 [Donja Dubrava (Kalezic et al., 1990), Sarajevo (Bolkay, 1929; communicated by G. Dzukic), the Zelengora Mountain (Bolkay, 1928) Dobrsko Selo (Kalezic eastern part of the range (F. R. Yugoslavia Dzukic, 1990)]. The easternmost recorded local Greece) have ventral coloration patterns with many sharpedged as spots, Freytag (1988) observed, not unlike thatof cristatus (Plates III). Crested newts is site 29 at ity Dimitrovgrad (Radovanovic, 1964). Crested newts of unknown taxonomic affinity were recorded at the Dalmatian coast [site 28, situated in between Sebenico (= Sibenik) Spalato (=

16 an Geographic the 196 J. W. Arntzen & G. P. Wallis variation in crested newt superspecies Split) (Werner, 1897, also mentioned by Buresh On a gross geographic scale, phenotype distri Zonkov, 1941)], but with no clearly indepen butions, mtdna haplotype distributions are dent confirmation for over a century we doubt the concordant. However, in northern Yugoslavia the validity of this record. Dzukic (1993) considers KAR? mtdna haplotype is more widespread than the distribution of carnifex not to be interrupted but continuous, following a strip of l to the south the karelinii phenotype distribution would suggest (Fig. 3) (Wallis Arntzen, 1989). The of the Sava river, without, however, presenting data KAR? mtdna haplotype is locally found in supporting this view. The area where crested newts are absent coincides with the core area ofthe karst (Sket, 1994), where most natural water bodies are dobrogicus, cristatus, carnifex macedonicus populations. The reverse situation, with a foreign haplotype in karelinii, has been observed ephemeral do often not support once (the DOB haplotype in population 15). the larva! development of species with a prolonged larval phase, Populations with foreign haplotypes possess either such as crested newts. The smallbodied newts such two haplotypes original plus an alien, such as at site in karelinii dobrogicus), as alpestris vulgaris in contrast are widespread locally abundant. They may reproduce or just the alien haplotype [ KAR? in cristatus successfully in shallow temporary ponds such (site 30 37, N = a wheel ruts (Winkler Brauns, dispersal rate for the small newt 1990) the 5) in carnifex macedonicus (site 35 36, N= 16; Wallis Arntzen, vulgaris is estimated to be higher than that for the big newt 1989)]. To account for these observations we suggest the following scenario. In former times cristatus (Stensjo, karelinii was more widespread than at present, with 1998). While most contemporary newt ponds are manmade rarely desiccate (i.e., watering holes for cattle), the puddles a range approximately coinciding to the present formed by fallen trees springs may originally day distribution of the KAR? haplotype. By dispersing southwards northwards, respectively have been the typical breeding habitat for the small bodied species. cristatus carnifex macedonicus superseded karelinii, in which process the range of The distribution of the four crested newt karelinii south ofbelgrade became isolated from the main stock (see the arrows in Fig. 3c). The species in F. R. Yugoslavia is complex (Fig. 3c). Tritiums dobrogicus is found all over the Pannonian genetic interactions between karelinii at one side Dobrogean Plains. Both parts of the range are probably connected the by Danube where flowing through the Iron Gate (Arntzen et al., 1997). Triturus cristatus carnifex macedonicus at the other where such that the formation of FI hybrids was asymmetric, with hybrid offspring cristatus has a wide European range, is widespread the subsequent backcrosses possessing over Romania reaches southwards over the Iron the (maternally inherited) karelinii mtdna. This scenario is surprisingly similar to the one we described for Gate into Yugoslavia. Triturus carnifex macedonicus is widespread over most of Yugoslavia, the cristatus Former Yugoslavian Republic of Macedonia, Albania, northern Greece. Triturus karelinii is found immediately south southeast of Belgrade. The available evidence that the local dis suggests marmoratus interactions in western France (Arntzen Wallis, 1991). In France, cristatus supersedes marmoratus, forming marmoratus enclaves in the process. Hybridisation between the species is strongly asymmetric, with tribution is in a small pocket enclave, geo FI adults derived from matings of cristatus graphically isolated from the main karelinii mothers marmoratus fathers significantly distribution in Bulgaria, Thrace, Turkey (Fig. outnumbering the reverse combination. The facts 3c). However, a link between the parts, along a narrow strip in northeastern Yugoslavia (as in Arntzen, 1995 in Kalezic et al, 1997: fig. 6), responsible for this phenomenon are largely unknown but may involve the genetic incompatibility of the nuclear mtdna genomes (J. W. cannot be excluded. The further surveying ofeastern Arntzen et ah, in prep.). By comparing past Yugoslavia northwestern Bulgaria is required present distributions, the rate at which cristatus to settle this issue. takes over from marmoratus has been estimated

17 Contributions to Zoology, (3) 197 as averaging one km a year. The process may be triggered, or accelerated, by the removal of hedgerows, modifying a with terrestrial lscape Arntzen JW, Bugter RJF, Cogalniccanu D, Wallis GP The distribution conservation status of the Danube crested newt, Triturus 18: 133 dobrogicus.amphibiareptilia 142. features favourable to marmoratus, the most Arntzen JW, GardaPan's M Morphological terrestrial among the bigbodied newt species, into allozyme studies of midwife toads (genus Alytes), including one favourable to the more aquatic cristatus. the description of two new taxa from Spain. Contrib. Zool. The habitat preferences of the various crested newt species in eastern Europe, with the of exception dobrogicus, are poorly understood it is unclear which ecological parameters affect their distribution or change in distribution. Another area of complexity is that around Vienna, where 65: 534. Arntzen JW, Sket B Morphometric analysis of black white European cave salamers, Proteus anguinus. J. Zool. 241: Arntzen JW,Tennis SFM A six year study on the population dynamics ofthe crested news (Triturus cristatus) following the colonization of a newly created pond. Herpetol. carnifex, cristatus dobrogicus meet (Fig. 3b). At sites newts were found with carnifex phenotype the mtdna haplotype typical for dobrogicus, matching a similar observation at Tasovice in the Czech Republic (48 49 N, E; J. Pialek, V. Zavadil J. W. Arntzen, unpubl.). As noted by CrnobrnjaIsailovic et al. (1997), 1 3: Arntzen JW, Wallis GP Restricted flow in gene a moving hybrid zone of the newts Triturus cristatus marmoratus in western France. Evolution 45: Arntzen JW, Wallis GP The Wolterstorff Index its value to the taxonomyof the Crested Newt superspecies. Abh. Ber. Naturk. Vorgesch., Magdeburg 17: Atchley WR, Gaskins CT, Andersen D Statistical properties of ratios. I. Empirical results. Syst. Zool. 25: Avise JC, Ball RM Principles of genealogical concor the remarkable variability in Balkan crested newts dance in species concepts biological taxonomy. Oxford should provide valuable insights into the evolu Surv. Evol. Biol. 7: tion of the group. Palaeontological various molecular methods have some provided clues towards the timing of the radiation of the cristatus superspecies (reviewed in Oosterbroek Arntzen, 1992). This period, which can be placed at 25 Ma, was one of great geographical complexity in southeastern Europe (CrnobrnjaIsailovic geological et ah, 1997 references therein). However, our ability to associate the historical patterns of fragmentation, speciation dispersal with palaeogeography is, as yet, hampered by the absence of a wellsupported phylogeny. Barlow GW Causes significance of morphological variation in fishes. Zool. Syst. 10: , Bernatchez L A role for molecular systematics in defining evolutionarily significant units in fishes. In: Nielsen JL, ed. Evolution the Aquatic Ecosystem. Defining Unique Units in Population Conservation. American Fisheries Society Symposium; Bethesda, Maryl, 17: SJ Die Schadel der Salamrinen mit Bolkay besonderen Rucksicht auf ihre systematische Bedeutung. Zeitschr. Anat. Entwickelungsgesch. 86: Bolkay SJ Die Amphibien und Reptilien von Sarajevo und Umgebung. Glasnik Zem. muzeum BosnieHerzegovina 41: Bouton N Donnees sur la migration de Triturus cristatus et marmoratus (Urodola, Salamridae) dans le Departement de la Mayenne (France!. Bui. Soc. Herpetol. France Acknowledgements We thank friends close colleagues for constructivecomments all those who helped in collecting. 40: Buccilnnoccnti S, Ragghianti M, Mancino G Investigations of karyology hybrids in Triturus boscai villatus, with a reinterpretation of the species groups within Triturus (Caudata: Salamridae). Copeia 1983: Bu resell I, Zonkov J Untersuchungen iiber die References Verbreitung der Reptilien und Amphibien in Bulgarien und auf der Balkanhalbinsel. Mitt. ksnigl. naturwiss. Inst. Sofia Bulgarien 14: (abstract). Andreone F, Giacoma C Breeding dynamics oftriturus Caputo V, Lanza B, Palmier! R Body elongation carnifex at a pond in northwestern Italy (Amphibia, Urodela, Salamridae). Holarct. Ecol. 12: limb reduction in the genus Chalcides Laurenti 1768 (Squamata Scincidae): a comparative study. Trap. Zool. 8: Arntzen JW European newts: a model system for evolutionary studies. In: Llorente GA, Montori A, Santos X, Carretero MA, eds. Sci. Herpetol AsociacionHerpetologica Espanola: Barcelona, Crnobrnja J, Kalezic ML, Dzukic G Genetic divergence in the crested newt (Triturus cristatus complex) from Yugoslavia. Biosistematika 15: 8192.

18 Geographic 198 J. W. Arntzen & G. P Wallis variation in crested newt superspecies Crnobrnjalsailovic J, Dzukic G, Krstic N, Kalezic ML Kalezic ML, Dzukic G Evidence for paedomorphosis Evolutionary paleogeographical effects on the distribution of the Triturus cristatus superspecies in the central in the Crested newt (Triturus cristatus complex) from Yugoslavia. Bull. Mus. Hist. Nat. Belgrade Serie Biol. B 45: 127 Balkans. AmphibiaReptilia 18: Cvctkovic D, Kalczic ML, Dzukic G Sexual size Kalezic ML, Stevanovic AM Morphological variation shape difference in the crested newt (Triturus carnifex): ontogenetic growthaspects. Alytes 15: Dely OG Contribution nouvelles la connaissance des a Tritons a Crete (Triturus cristatus Laurenti). Ann. Hist.Nat. in the crested newts, Triturus cristatus (Urodela, Amphibia) from Kru'sevacka Z upa (central Serbia). Biosistematika 6: Kalezic ML, Dzukic G, Tvrtkovic N Newts (Triturus, Mus. Nat. Hungarici 51: Salamridae, Urodela) of the Kukovica Ravni Kotari Dzukic GV Fauna, zoogeografija zas tita repatih regions. Spixiana 13: vodozemaca (Caudata) Srbije. [Tailed amphibians (Caudata) of Serbia: a faunistical zoogeographical study]. PhD Kalezic ML, Dzukic G, Stamenkovic AM, Crnobrnja J Thesis, Faculty of Biology, University of Beograd, Fachbach G Das SerumEiweissbild der Triturus Morphometries ofthe crested newt (Triturus cristatus complex) from Yugoslavia: Relevance Nauka, Beograd 42: for taxonomy. Arh. Biol. cmto/kiunterarten, Zeitschr. Zool. Syst. Evol. 12: Kalezic ML, Dzukic G, Mesaros G, Crnobrnjalsailovic, J The crested newt (Triturus cristatus superspecies) in Faith DP Conservation evaluation phylogenetic diversity. Biol. Conserv. 61: 110. exyugoslavia: morphological structuring distribution Faith DP, Cranston PS Could a cladogram this short have arisen by chance alone? On permutation tests for ciadistic structure, Cladistics 7: 128, patterns, The University Thought Publication in Natural Sciences 4: Karanian S Beitrage zur Herpetologie von Mazedonien. FejervaryLangh AM Beitrage und Berichtigungen zum AmphibienTeil des ungarischen Faunenkatalogus. Fragmenta Faunistica Hungarici 6: Freytag GE Erinnerungen an Willy Wolterstorff. Wie kompliziert ist die Rassengliederung des Kammolches (Triturus cristatus [Laurenti, 1768])? (Amphibia, Caudata, Salamridae).Bin Blick in die Geschichte der Salamer Glasnik Kroat. Naturwiss. Ges. Zagreb 34: Karaman SL Beitrag zur Herpetologie Nordserbiens. Jugoslavenska Akad. Znanosti Umjetnosti 1948: 627 (abstract). Le R Evolutionary mechanisms oflimb loss in tetrapods. Evolution 32: Lanza B, Gentile E, Torricelli Preliminary data on the kunde. Zool. Ahh. Staatl. Mus. Tierk. Dresden 44: 110. Frost DR (ed.) Amphibian Species of the World. A Taxonomic Geographical Reference. Allen Press the Association of Systematics Collections: Lawrence, Kansas. Greer AE Limb reduction in the lizard Lerista. 1. genus number of presacral vertebrae in Triturus cratatoartenkreis (Amphibia: Caudata: Salamridae). In: Ghiari G. ed., Symposium on the evolution of terrestrial vertebrates. Selected Symposia MonographsU. Z. I. Mucchi: Modena. 4: Variation in the number of phalanges presacral vertebrae. Herpetologica 21: Leary RF, Allendorf FW, Knudscn KL, Developmental instability high meristic counts in interspecific hybrids Griffith II Miniaturization elongation in Eumeces of salmonid fishes. Evolution 39: (Sauria: Scincidae). Copeia 1990: , Lindsey CC Temperaturecontrolled meristic variation Griffiths RA, Mylottc VJ Microhabitat selection in the salamer Ambystoma gracile. Nature 209: Lindsey CC Pleomerism, the widespread tendency feeding relations ofsmooth warty newts, Triturus vulgaris cristatus, at an upl pond in midwales. Holarct. Ecol. 10: 17. Grillitsch B, Grillitsch H, Iliiopl M, Tiedemann F. 1983]. Lurche und Kriechtiere Niederdsterreichs. Facultas Verlag: among related fish species for vertebral number to be correlated with maximum body length. J. Fish. Res. Board Canada 32: Wien. Litvinchuk S, Sokolova TM, Borkin LJ Biochemical Harvey PH, Pagcl MD The Comparative Method in Bi differentiation ofthe crested newt (Triturus cristatus group) ology. Oxford University Press: Oxford. in the territory of the former USSR. Abh. Ber. Naturk. Hcrrc W Vergleichende Untersuchungen an den Vorgesch., Magdeburg 17: Unterarten des Triturus cristatus Laur. Zeitschr. Anat. Entwicklungsgesch. IAbt. 99: 162. Hillis DM Molecular versus morphological approaches to systematics. Annual Rev. Ecol. Syst. 18: Jehle R, PauliThonke A,TamnigJ, Hodl W Phanologie Losos JB A phylogenetic analysis of character displacement in Caribbean Anolis lizards. Evolution 44: Mayr E Animal Species Evolution. Belknap Press Cambridge, Massachusetts. McDowall RM The galaxiid fishes of New Zeal. Bull. und Weraktivitat des Donaukammolches (Triturus Mus. Comp. Zool. 139: dobrogicus) an einem Gewasser auf der Wiener Donauinsel. Stapfia 51: McDowall RM, Wallis GP Description redescription of Galaxias species (Teleostei: Galaxidae) from Otago Jockuseh EL Geographic variation phenotypic plasticity of number of trunk vertebrae in slender salamers, Batrachoseps(Caudata: Plethodontida). Evolution 51: 1966 Southl. J. Roy. Soc. New Zeal 26: Oosterbroek P, Arntzen JW Areacladograms of Circum Mediterranean taxa in relation to Mediterranean palaeogeography. J. Biogeogr. 19: 320,

19 Contributions to Zoology, (3) 199 Orska J, Imiolek Z Preliminary studies on the effect of SPSS Statistical packagefor the social sciences, version temperatureon the development ofmeristic characters in the 4.0. SPSS Inc.: Chicago. Urodela. Acta Universitatis Wratislaviensis 3: StatView. Peabody RB, Brodie El) Effect of temperature,salinity photoperiod The solution for data analysis presentation graphics. Abacus Concepts: Berkeley. Stensjo JO Population genetics of the common newt on the number of trunk vertebrae in Ambystoma maculatum. Copeia 1975: (Triturus vulgaris,) the crested newt ( cristatusj, with Radovanovic M Die Verbreitung der Amphibien und implications for conservation. Thesis, Uppsala Universitet: Reptilien in Jugoslawien. Senckenbergiana Biol. 45: Renous S, Case JP Body vertebral proportions in Uppsala. Vallee L Recherches sur Triturus blasii de ITsIe, hybride naturel de Triturus cristatus Laur. x Triturus marmoratus Latr. Mem. Soc. Zool. France 39: 195. Wallis GP Mitochondrial DNA insertionpolymorphism Gymnophiona(Amphibia): Diversity ofmorphological types. Copeia 1989: Rolilf FJ NTSYSpc. Numerical multivari taxonomy ate analysis system. Exeter Publishing; New York. germ line heteroplasmy in the Triturus cristatus complex. Heredity 58: Schmidtlcr JJ, Schmidtlcr JF Verbreitung, Okologie Wallis GP, Arntzen JW Mitochondrial DNA variation und innerartliche Gliederung von Triturus vulgaris adriatischen Kustengebieten. Spixiana 6: in den in the crested newt superspecies: Limited cytoplasmic gene flow among species. Evolution 43; Siegel S, Castellan NJ Nonparametric statistics for the Werner F Die Reptilien und Amphibien Osterreich behavioural sciences (second edition). McGrawHill: New Ungarns und der Occupationsler. [PUBLISHER] Wien. York. Winkler C, Brauns C Zur Okologie von Molchen in Sket B Yugoslavia (BosniaHerzegovina, Croatia, wassergefiillten Wagenspuren einer Mischwaldflache im Macedonia, Montenegro, Serbia, Slovenia). In: Juberthie C, Sudniedersachsischen Bergl. Salamra 26: , Decu J, eds. Encyclopaedia Biospeologica. Tome I. Moulins, Wolterstorff W Ubersicht der Unterarten und Formen Bucarest: Societe de Biospeologie, des Triton cristatus Laur. Blatter Terrarienkunde, Stuttgart 34: Aquarien und Sket B, Arntzen JW A black, nontroglomorphic amphibian from the karst ofslovenia: Proteus anguinusparkelj n. ssp. (Urodela: Proteidae,). Contrib. Zool. 64: Young JZ The Life of Vertebrates. Clarendon Press: Ox ford. Sokal RR, Rohll'FJ Biometry, second edition. Freeman: San Francisco. Received: 30 October 1998 Appendix PlateI. Sexual display by male Triturus dobrogicus. Symptomatic ofthe species are the bulging black white throat high head crest, giving the appearance of increased size. Photo by M. Sparreboom.

20 J. W. Arntzen & G. 200 Plates II. III. Plate II: b, Triturus c, e. carnifex carnifex; k, Geographical origin Tadten, Austria; k, Italy; surface Ventral of crested newts in of five taxa dobrogicus; carnifex macedonicus; of the is: individuals i, Plate III: 1, dobrogicus individuals from Yugoslavia; i, Karacabey, colour the cloaca coordinates is size black large, as one limbs of the Triturus spots, Throat spots surface are a ) but build, Lean orange short than in males (a, e, g, the female it is h, j, n Karan, s, Farma, Italy; the t, n, v, Tavna r, q, Federal Republic of m, Visegrad, BosniaHercegovina, in (the others) limb by digit flattened. the tail females females v) tail distance is shorter in males than from r, u, of the small, yellow in is entirely black; in legs, underside In males is females. Note the Istanbul, toes some L.A. be most side of In males the tail of its all or not are van distinguished by body shape. ventral yellow along that can length marking by toeclipping. Photography by many fine white narrow sharp, medium age white tend darker throats with black spots. function sized legs, (Arntzen more & of the narrow Teunis, 1993) build, large legs, Medium larger with white black blase, heavily whitestippled sides, is lenght. original size (such der Laan. tail few base Notes: conspicuous with irregular individual (a) medium wide, large, roundish, stipples Ventral spots throat throat with tend to coloration large angular lineup is shown little black is than females. likely or illdefined spots. no Notes: The black to be younger white The newts T o as Plate white in (b). III. muddied mix spots than the of grow others. on sides, throat black spots. Notes: Males stippling muddygray to fuse by tail base, heavily whitestippled sides, throat a stippling, ventral surface yelloworange stipples. Ventral surface yellow variable with tail roundish females. A possible Leanmediumbuild, with carnifex carnifex. have it in males of the underside regenerating with spot pattern becomes denser with Triturus cloaca or newt deep larger Triturus cristatus. to carnifex phenotypes black yellow the in dobrogicus. ventral the colour twolobed the that 2. Note Table longer interlimb inner toes of Description are see cloaca, by of the yellow only directly behind Digits u, o, Yugoslavia; of Federal Republic Monastire, BosniaHercegovina; Federal qw, Turkey. geographical shape, Benevento, Italy; p, n, karelinii. Geographical origin of m, 1 Divcibare, w, Turkey. For is: carnifex carnifex; Triturus o, p, macedonicus, karelinii. superspecies newt superspecies. the Triturus cristatus cristatus; j, Djurinci, g, Yugoslavia; h, Istanbul, Turkey, of g variation in crested Geographic Mayenne, France; j, Farma, cristatus from Visegrad, BosniaHercegovina; Republic d, f, a, P. Wallis colour represent opposite sides of the range. Triturus carnifex muddied mix macedonicus. of black pattern of small, irregular spots. dobrogicus (n), Triturus small cristatus KAR?, Notes: see (s), to many, Notes: no Wallis black obvious medium spots, extending coloration on to the characters base medium sized white stipples. Ventral The coloration characteristics carnifex carnifex (t) Arntzen, 1989). tail heavy build, large legs, or tail particularly sides yellow variable whitestippled sides, (especially distinguish are wide, surface densely whitestippled, to orangeyellow individuals may with throat a a dense resemble karelinii (v). karelinii. Heavy build, large legs, wide tail base, heavily to mediumsized angular. Medium yellow with between in females) ventral surface yelloworange continuous karelinii possessing with throat where different mtdna spots with many tend to be haplotypes (KAR

21 1999 Contributions to Zoology, 68 (3) 201 Plate II

22 1999 Contributions to Zoology, 68 (3) 203 Plate III