Intraspecific Genetic Differentiation of the Siberian Newt (Salamandrella keyserlingii, Amphibia, Caudata) and the Cryptic Species S. schrenckii from Southeastern Russia

D. I. Berman*, M. V. Derenko*, B. A. Malyarchuk*, T. Grzybowski**, A. P. Kryukov***, and D. Miscicka-Sliwka** *

Institute of Biological Problems of the North, Far East Division, Russian Academy of Sciences, Magadan, 685000 Russia*Forensic Medicine Institute, LudwikRydygier Medical University, Bydgoszcz, 85-094 Poland ***Institute of Biology and Soil Science, Far East Division, Russian Academy of Sciences, Vladivostok, 690022 Russia 
Received March 16, 2005

Abstract—The nucleotide sequences of the mitochondrial cytochrome b gene in the Siberian newt Salaman­drella keyserlingii Dybowski 1870 from the populations of the Ural Mountains, Magadan oblast, Chukchi Pen­insula, Sakhalin Island, and Primorskii krai are analyzed. It is shown that in most populations studied (except for Primorskii krai), a low geographic variation in morphological characters corresponds to a low level of genetic variation (0.38% in the combined sample from the Magadan, Sakhalin, Chukchi, and Ural populations). Different scenarios for the origin of the genetically and morphologically homogeneous hyperpopulation are dis­cussed, taking into account the obvious lack of genetic exchange between the marginal populations of the range. They involve the rapid formation of the species range in the Holocene, which followed its gradual development in the Pleistocene; unidirectional stabilizing selection within the entire range; the maintenance of variation at a stable level by mixing of the population during the dispersal of the young and, possibly, by group fertilization. The population inhabiting the Primorskii krai, despite minor morphological differences from other populations, is characterized by a high level of mtDNA divergence (9.8-11.6%) and considerable intrapopulation variation (1.86%). In view of the data obtained, it appears feasible to restore for Salamandrella from Primorskii krai the name S. schrenckii (Strauch, 1870), which was a junior synonym for S. keyserlingii. Based on the mtDNA sequences, the times of emergence for S.
keyserlingii and S. schrenckii are dated 490 ka and 2.4 Ma, respec­tively. These should be considered two species of different ages which diverged from a common stem approx­imately 14 Ma, rather than a descendant and its ancestor.

INTRODUCTION 

The Siberian newt (Salamandrella keyserlingii Dybowski 1870) occupies an extensive range from the tundra to the steppe and from the Pacific Ocean to northwestern European Russia. Morphometric studies have revealed a low geographic variation in the mor­phological characters of this species. Although Siberian newts from the Yekaterinburg and Yakutsk regions and Primorye differ in their body proportions and plastic characteristics, the extent of the differences is too low to classify them as separate taxa. The available data allow only the conclusion that the Far East Siberian newt has some distinctive features (Ostashko, 1981). Borkin (1994) analyzed variation in eight parameters (including the lengths of the body and tail, the relative length and width of the head, the ratio of limb measure­ments, and the number of digits and grooves on the body sides) in population samples from the Yekaterin­burg region, Yakutia, Transbaikalia, Sakhalin Island, and Primorskii krai (Primorye). He also concluded that, "taking into account all the characteristics considered, it seems that Siberian newts  from the southeastern areas of the range differ to a greater extent than the others..." (Borkin, 1994, p. 79). Basarukin and Borkin (1984) proposed the probable taxonomic significance of the nonspiral shape of the spawn, the absence of nuptial behavior, and some other features distinctive of the Siberian newt population from Primorye. However, referring to unpublished data of G.P. Sapozhnikov, Borkin (1994) subsequently indi­cated that "some ecological distinctions of S. keyserlingii from Primorye are less pronounced than was previously believed" (Borkin, 1994, p. 80). At the same time, the idea of substantial differences of this population from other populations of S. keyserlingii has recently been corroborated by Vorob'eva et al. (1999) based on the development of elements of the larval locomotor system, by Litvinchuk et al. (2001) based on the genome size, and by Litvinchuk and Borkin (2003) based on the number of vertebrae and costal grooves. On this basis, Litvinchuk et al. (2004) proposed that Siberian newts from Primorye should be regarded as a separate subspecies, Salamandrella keyserlingii tridac-tyla Nikolsky 1906 .[1] Kuzmin and Maslova (2003) showed the genetic specificity of the Siberian newt from Primorye and, discussing its Latin name, pro­posed to designate it Salamandrella keyserlingii tridac-tyla Nikolsky 1905 in the case of its subspecies rank or Salamandrella tridactyla Nikolsky 1905 in the case of species rank, i.e., in the presence of reproductive isola­tion.
[1] Here and below, dates are given according to the authors cited.

The challenge to explain the surprising morpholog­ical uniformity of S. keyserlingii throughout its extremely large range, combined with the taxonomic independence of animals from southeastern Russia, required the application of molecular genetic methods. A low geographic variation in the phenotype can mask a significant genetic polymorphism and even the exist­ence of cryptic species (Kryukov and Suzuki, 2000; Borkin et al., 2001; Zink et al., 2002a; Khalturin et al., 2003). Therefore, we investigated genetic variation in S. keyserlingii over its wide range by analyzing poly­morphism in the nucleotide sequences of the cyto-chrome b gene in mitochondrial DNA (mtDNA). As is well known, mtDNA is characterized by maternal inheritance and lack of recombination; it displays a high level of variation, providing, therefore, important molecular data on the evolution of different animals, including amphibians (Avise, 1994; Garsia-Paris et al., 2000; Steinfartz et al., 2000; Riberon et al., 2000). 

MATERIALS AND METHODS 

We analyzed samples of Salamandrella (tissues of adults, larvae, and embryos frozen or preserved in 70% alcohol) from the following regions (Fig. 1): 

(1) southern Sakhalin Island, Lake Tunaicha,
9 specimens; 

(2) vicinity of Vladivostok (Fig. 2a), 28 specimens,
including 2 specimens from the vicinity of the Bogatinskoe Reservoir, 7 from the Malaya Sedanka River valley, 8 from the Botanical Garden, 5 from the Bol'shaya Sedanka River valley, and 6 from the vicinity of Lazurnaya Bay;

(3) northeastern Asia (Fig. 2b), 41 specimens:
Chukchi Peninsula (2 specimens from the village of
Markovo and 1 from the town of Anadyr); Kolyma
River Basin (5 specimens from the vicinity of the village of El'gen); Ola, Oira, and Yana rivers (Pacific Basin, west and east of Magadan), lakes nearby Magadan (at the top of the Kamennyi Venets Hill, Staritskogo Peninsula), 33 specimens; and (4) vicinity of Yekaterinburg, 8 specimens. 
DNA was extracted by the standard method involv­ing tissue lysis in a solution containing 100 mM Tris-HCl (pH 8.0), 10 mM EDTA, 100 mM NaCl, 1% sodium dodecylsulfate, and 0.2 mg/ml proteinase K (Sigma, USA) at 56°C for 12-16 h, with subsequent deproteination in a phenol-chloroform mixture. The quality and quantity of DNA were estimated by electro-phoresis in 1% agarose gel followed by gel staining with ethidium bromide and examination under UV 
light. The cytochrome b gene was amplified as two over­lapping DNA fragments by polymerase chain reaction (PCR) with the pairs of primers MVZ15L and MVZ18H, and MVZ25L and ControlWH proposed by Goebel et al. (1999). Amplification included 40 cycles at the following temperatures: 94°C for 60 s, 46°C for 90 s, and 72°C for 90 s. The PCR products were purified by ultrafiltration in Microcon 100 columns (Am-con, United States). Fragments of the cytochrome b gene were sequenced in an ABI PrismTM 377 automatic sequencer using a BigDye Terminator v. 3.1 Cycle Sequencing Kit (Applied Biosystems, United States) and primers MVZ15L and MVZ25L. The sequences obtained were aligned and analyzed using the Sequence Navigator program (Applied Biosystems), and phylo-genetic analysis was performed using the MEGA 2.1 program package (Kumar et al., 2001). Pairwise genetic distances (p distances) between individual DNA sequences were calculated based on the number of nucleotide substitutions per base pair. The time of divergence and evolutionary age of the mtDNA lin­eages were estimated on the assumption that the rate of substitutions in the cytochrome b gene was 0.77% of divergence (for transitions and transversions) per mil­lion years (Caccone et al., 1997). This value was obtained by calibrating the "molecular clock" based on paleontological data on the time of divergence between species of the genus Euproctus (Caccone et al., 1997). Furthermore, this value agrees with the rate of diver­gence for this gene (ranging from 0.47 to 1% per mil­lion years) in a number of species of the  family Sala-mandridae (for reviews, see Tan and Wake, 1995; Cac-cone et al., 1997; Riberon et al., 2000). 

The neighbor-joining (NJ) method (Saitou and Nei, 1987), unweighted pair group method with arithmetic averages (UPGMA), and maximum parsimony (MP) method (Nei, 1987) were used to construct phyloge-netic trees. The data from GenBank (http://www.ncbi.nlm.nih.gov/entrez) on the nucleotide sequences of the cytochrome b gene (a 743-bp frag­ment) in members of different genera of the family Hynobiidae were used for comparative analysis. Three species of the genus Euproctus from the family Sala-mandridae were used as outgroups (Table 1).

The sequences of the cytochrome b gene obtained in this study for the 86 Siberian newts were deposited in the GenBank (accession nos. AY701904-AY701989).

RESULTS 

The nucleotide sequences of a 825-bp fragment of the mitochondrial cytochrome b gene between posi­tions 14214 and 15 308 (according to the numbering in the complete mitochondrial genome of Ranodon sibiricus, see Zhang et al., 2003) were determined in 86 Siberian newts from five regions. The total sample included 15 variants (haplotypes or mitotypes) of the gene differing in 112 polymorphic positions, of which 102 were phylogenetically informative (i.e., occurred in more than one mtDNA haplotype). The majority of mutations (99) are transitions, while transversions are far less numerous (13 positions). There are 19 and 1 mutations in the first and second codon positions, respectively. There are 9 positions where mutations result in amino acid substitutions. Thus, most mutations are synonymous (Fig. 3). The most frequent haplotypes in the samples exam­ined were designated Magadan-1 (recorded in 23 Sibe­rian newts from the Magadan sample) and Ural-1 (the same haplotype revealed in all specimens from the Ural sample). Another haplotype observed in 12 out of 38 specimens from the Magadan sample (Magadan-4) was also detected in all specimens from the Chukchi Peninsula (haplotype Chukotka-1). The most frequent haplotype of the Primorye sample (Vladivostok-1) was revealed in 14 out of 28 specimens (Fig. 3).

The distribution of haplotypes of the cytochrome b gene manifests obvious geographic differentiation, as is evident from the NJ phylogenetic tree constructed on the basis of both transitions and transversions (Fig. 4). 

Two large monophyletic clusters are distinct—the first contains only mtDNA variants from Primorye, while the second includes mtDNA variants found in other local samples, i.e., those from the Ural, Sakhalin, Magadan, and Chukchi populations. The Sakhalin mtDNA variants in the second cluster are segregated and form an individual second-order cluster. The mito-chondrial variants from Primorye are also clearly heter­ogeneous; the Vladivostok-2 and Vladivostok-3 vari­ants are distinctly segregated from the others. It should be emphasized that the phylogenetic trees constructed by different methods (i.e., UPGMA or MP) have the same topology.


The pairwise genetic distances between the nucle-otide sequences of the cytochrome b gene variants range from 0.1 to 11.6%, that is, from 1 to 96 nucleotide substitutions (Table 2).

 Siberian newts from Primorye make the main contribution to these values; thus, their mtDNA variants differ from those in other samples by 81-96 substitutions (9.8-11.6%). The intrapopulation variation in Siberian newts from Primorye (1.86%) is ten times that in other local populations, i.e., 0.16% in the Sakhalin sample and 0.18% in the combined Magadan, Chukotka, and Ural sample (Table 3).

 More­over, in the first-order clusters, the degrees of intra-group differentiation of the mtDNA variants recorded in Primorye are 0.84 and 0.61% (Fig. 4); i.e., this dif­ferentiation is also substantially higher than in other local populations of S. keyserlingii.

DISCUSSION 

The above analysis has shown that all the popula­tions studied (from the Urals, Magadan oblast, Chukchi Peninsula, and Sakhalin Island), except for the popula­tion from Primorye, combine insignificant geographic variation in morphological characters with a low level of genetic variation in the mitochondrial cytochrome b gene. Thus, S. keyserlingii is a genetically uniform spe­cies throughout its vast range (except for Primorye). Only the Sakhalin variants of mtDNA form a separate subcluster. At present, it is premature to discuss the tax-onomic significance of this subcluster because of the limited sample size and the relatively small differences from the other samples (0.54%, Table 3). Note that a separate species, Salamandrella cristata Anderson, 1917, was described from southern Sakhalin; however, it was subsequently synonymized with S. keyserlingii (Borkin, 1994). Therefore, it is desirable to use a larger sample from this area, because the nine spawn samples examined come from the same water body, but, none­theless, belong to three different variants of the cyto-chrome b gene. It may well be that the Siberian newt population from southern Sakhalin differs from the continental populations to an even greater extent than is presently thought. The insignificant morphological specificity of Sala-mandrella from Primorye concealed the fact that this population is characterized by considerable genetic dif­ferences (10.8%) from all the other populations stud­ied. To reveal the probable taxonomic significance of these differences, we performed a comparative phylogenetic analysis of the nucleotide sequences of the cytochrome b gene in 13 species belonging to different genera of the family Hynobiidae (Table 4). 

The inter­specific distances within the genus Batrachuperus (with six species included in the comparison) are 4.4­9.3%; those in the genera Pseudohynobius (two spe­cies) and Hynobius (three species) are 15.5 and 10.7­12.7%, respectively. The interspecific distances in the genus Euproctus (three species) from the family Sala-mandridae (used as an outgroup in our analysis) are even greater, 10.7-21%. Thus, the distance of 10.8% between two clusters of the Salamandrella samples examined is in the range of values characteristic of interspecific differences in several genera of the family Hynobiidae and in the genus Euproctus. 

These results support the conclusion of our prede­cessors concerning the specificity of Salamandrella from Primorye, which was based on differences in mor­phology, biology, and genome size (Ostashko, 1981; Basarukin and Borkin, 1984; Borkin, 1994; Vorob'eva et al., 1999; Litvinchuk et al., 2001, 2004; Litvinchuk and Borkin, 2003; Kuzmin and Maslova, 2003). All these data in the aggregate strongly suggest that Sala-mandrella from Primorye is a separate species that should be named S. schrenckii (Strauch, 1870) (Berman et al., 2005). This name was previously regarded as a senior synonym of S. keyserlingii (see review, Borkin, 1994), which was proposed by A. Strauch 35 years earlier than the name Salamandrella keyserlingii var. tridactyla introduced by Nikol'skii (1905). In the fol­lowing text, this author indicates that the establishment of this form is tentative: "The Zoological Museum of the Imperial Academy of Sciences has sent me for examination one specimen of this species from Vladivostok; the forelimbs of this animal have only three digits, which are equally devel­oped; the third of the four digits present on normal limbs is absent. Based on this sole specimen, it is impossible to conclude whether it represents a particu­lar variety or an anomaly; if this feature proves suffi­cient for distinguishing a variety, I will name it var. tridactyla" (Nikol'skii, 1905, p. 491). 
Subsequent studies of the Siberian newt have shown that the proportion of animals with tridactylous fore-limbs averages 17%, increasing in Primorye to 37%; nevertheless, these individuals were justly considered to be abnormal (Borkin, 1994) and were not segregated in a separate taxon.

The name S. schrenckii was introduced earlier to designate normal animals; therefore, it should be pref-ered to S. tridactyla, which could have been used if the Siberian newt from Primorye was advanced in taxo-nomic rank (relative to the subspecies Salamandrella keyserlingii tridactyla) (Litvinchuk et al., 2004). 

It should also be emphasized that the data obtained in this study corroborate the generic rank of Salaman-drella, which was separated from Hynobius. Table 4 shows that the differences between either cluster of Salamandrella and 13 species of other genera of the family Hynobiidae (used for comparison) range from 15.5 to 18.8%, while the corresponding values for spe­cies of the genera Pseudohynobius, Hynobius, Batrach-uperus, Ranodon, and Liua are 14.4-18.1, 14.4-20, 14.3-18.9, 20, and 19.4%, respectively. Naturally, spe­cies of the genus Euproctus, which are members of a different family, differ from the same 13 species to a greater extent (18.9-25.7%).

The detached position of Salamandrella as a sepa­rate genus is clearly seen in the maximum parsimony tree with local bootstrap probabilities (Fig. 5). 

The genus Salamandrella forms a monophyletic cluster in the tree, with a high bootstrap value (92%). The two groups of mtDNA variants which correspond to the species Salamandrella keyserlingii and S. schrenckii, are differentiated within this cluster with an even higher bootstrap value (99%).

Supplement to the Description of Salamandrella shrenckii 

Salamandrella schrenckii (Strauch, 1870), sp. dist. Schrenck's Siberian newt. 
Isodactylium schrenckii Strauch, 1870: p. 56. Salamandrella keyserlingii var. tridactyla Nikolsky, 1905: p. 491. 
Salamandrella keyserlingii var. kalinowskiana 
Dybowski, 1928: p. 1080. 
Salamandrella keyserlingii tridactyla: Litvinchuk 
et al., 2004, p. 282. 
Salamandrella schrenckii: Berman et al., 2005. 

Lectotype. Specimen labeled "no. 115. Isodactylium schrenckii Str. Agdeki ad Ussuri. Dr. L. v. Schrenck. 1855," "no. 115. Trdet. O. Gumilevskii. Hynobius key-serlingii (Dyb). Ogdeki Ussuri 1855. Leg.: Schrenck." The lectotype is housed at the Zoological Institute of the Russian Academy of Sciences (ZIN); it comes from the vicinity of the village of Agdeki [2] (= Kukolevo, Kha­barovsk krai), situated on the Ussuri River south of the Podkhorenok River mouth (less than 20 km from the railroad station Dormidontovka); designated by Berman et al. (2005).

2. In the map with Schrenck's route given by Strauch (1870), the settlement is designated Agdiki; in the original label of the type specimen, no. 115 (and in the old registration book, see below) it is named Agdeki, while in the label with the identification of Gumilevskii, it is named Ogdeki. The village of Kukalevo is named in some maps as Kukelevo.

Strauch described the species based on several spec­imens. The number of specimens or information from their labels were absent, only the following data were provided: "Habitat. Ost-Sibirien, am Ussuri, an der Schilka und am Baikal-See" (Strauch, 1870, p. 57). The old registration book of the Department of Herpetology of ZIN contains the following data on 12 specimens listed under the name "Isodactylium schrenckii Str.:" 
nos. 110 and 111: Kunstkamer; 
nos. 112 (2 specimens), 113, and 114: fl. Schilka, 
Popoff, 1854; 
no. 115: Agdeki ad Ussuri, N 187, Dr. L. v. 
Schrenck, 1855; 
no. 116 (2 specimens): Sibiria orient., Radde; 
no. 117: Fl. Schilka, Maack, 1855; 
no. 118 (marked "var."): Lac Baikal, Maack, 1855; 
no. 119 (marked "var."): Des. Kirgis.??, Motschulsky. 

All these specimens were collected before 1870, and their labels are not in conflict with the points of capture indicated in the original description (see below); therefore, it is reasonable to suppose that all of them were included in the type series. The specimens stored at the Cabinet of Curiosities (Kunstkamer) were collected before 1830 and, probably, beyond the area of Primorye or Khabarovsk krai, because these regions were investigated later. G.I. Radde collected materials near Lake Baikal in 1855, in eastern Transbaikalia in 1856, and in the Amur River basin, in the area of the present-day Jewish Autonomous Region, in 1856 to 1858 (Ruzskii, 1937). Motchoulski (1844) traveled to Lake Baikal and western Transbaikalia in 1839 and 1840, and his specimen was most likely collected there, rather than in the steppes of Kazakhstan (Desertis Kirgisorum), as is specified in the registration book (with two question marks). Out of these specimens, nos. 111­113, 115 (designated above as the lectotype), and 116 are currently kept in the collection of ZIN. It is hardly probable that specimens from Lake Baikal and Shilka belong to S. schrenckii, because S. keyserlingii was originally described from the Baikal Region.

As indicated above, all the specimens analyzed by us from Primorye (28 individuals) are assigned on the basis of mtDNA to S. schrenckii. Therefore, we refer the d3ata on Salamandrella from this region to this species.³

3.During the preparation of this paper, N.E. Dokuchaev collected five additional Salamandrella specimens in the vicinity of the vil­lage of Georgievka (southern Khabarovsk krai, the middle reaches of the Kiya River, a right tributary of the Ussuri River), which is situated less than 50 km from the village of Kukalevo (about 25 km to the north by latitude), i.e., approximately within the terra typica of Salamandrella schrenckii. All of them were referred to this species based on their mtDNA sequences.

Salamandrella schrenckii differs from S. keyser-lingii in the following features:

(1) in a different (nonoverlapping) set of haplotypes (11.6% of the mtDNA distance); 

(2) in genome size; Salamandrella from the vicinity of Vladivostok significantly differs in the amount of nuclear DNA (by 3.1%) from those collected in Yaku­tia, Nizhni Novgorod, Tomsk, Kamchatka, and Sakha­lin oblasts (Litvinchuk et al., 2001);

(3) in the average number of vertebrae and costal grooves; in Salamandrella from Primorye, these values are less than 16.5 and less than 11.5, respectively, but they are greater in Salamandrella from other areas of the range (Ostashko, 1981; Litvinchuk and Borkin, 2003); 

(4) in certain developmental characteristics of limbs and other elements of the larval locomotor system, which are basically reduced to heterochronies (Vorob'eva et al., 1999). However, the authors of the paper cited believe that all differences are of a cenoge-netic nature and are mostly associated with different conditions of larval development in the Ural region (relatively warm 4water bodies) and Primorye (cold mountain creeks);[4]

(5) in the earlier completion of metamorphosis and emergence on land (Sapozhnikov, 1990). This point requires additional study, because Salamandrella lar­vae have not been examined in detail in other southern regions;

(6) in the shape and dimensions of spawns; they are not only spiral (as in all other parts of the range), but also straight, bagshaped (Korotkov, 1977; Chernichko, 1982; Regel', personal communication, cited from Basarukin and Borkin, 1984; Sapozhnikov, personal communication, cited from Borkin, 1994). Bag-shaped spawns of the thoroughly investigated Sala-mandrella from the Ural population form 1-4 coils (Sytina et al., 1987), while those of S. schrenckii form only 0.5-1.5 coils; in Primorye, bags are on average half as long (86.0 ± 1.0 mm versus 180 ± 5.1) and 1.5 times narrower in cross-section area than in the Ural population (G.P. Sapozhnikov, personal commu­nication). Sapozhnikov believes that the number of coils in spawn depends basically on the bag length; the lengths of bags with 0.5, 1, 1.5 coils average 75.0 ± 2.4, 87.9 ± 2.6, and 106.4 ± 5.5 mm, respectively; in Primorye, the degree of coiling is independent of the num­ber of embryos. As regards the fecundity of Salaman-drella from Primorye, contradictory data have been reported: according to Bromlei et al. (1977), the aver­age number of embryos in a spawn is 140-160; accord­ing to Korotkov (1977), it is up to 132; and Sapozhni-kov (personal communication) believes that it is 106.2 ± 1.4. 

---

^[4] All other published and original data on Salamandrella from Pri-morye, which are discussed in the present paper, concern amphib­ians and their spawn collected in small stagnant water bodies only.

These differences apparently result from 
making measurements of spawns in water bodies dif­fering in the age and size structure of the population, rather than from artifacts. 

In the Ussuriiskii Nature Reserve, spawn in the shape of straight bag prevails: such bags accounted for three fourths of the 212 spawn samples examined by G.P. Sapozhnikov, ranging from 50 to 97% in particular water bodies (personal communication). In places, only straight bags are recorded in the reserve (S.M. Lyapkov, personal communication, cited from Ishchenko et al., 1995; Kuzmin and Maslova, 2003). Straight spawn also occurs southeast of Khabarovsk, in the region named after Lazo near the village of Sidim (N.E. Dokuchaev, personal communication) approximately 100 km north­east of the type region in the upper reaches of the Nemta River, a right tributary of the Amur River. In the vicinity of Komsomolsk-na-Amure, only usual spiral spawn of S. keyserlingii is observed, while straight spawn has not been recorded (V.A. Mutin, personal communication). In contrast, in central Yakutia "...spawn of Siberian newts in the shape of relatively small conical bags was repeatedly observed" (Belimov and Sedalishchev, 1983, p. 42) in the Zarechnaya group of regions (V.T. Sedalishchev, personal communica­tion). Unfortunately, at present, it is impossible to com­pare in detail the shape of straight spawn from Primorye and Yakutia. Isolated findings of straight spawn were reported from the Ural Mountains (Ishchenko et al., 1995). Thus, the ranges of variation in the majority of spawn characteristics of the two spe­cies also overlap.[5] At present, the sole character distin­guishing the spawn of S. schrenckii is its considerably smaller size, namely, the length and thickness of bags. This character should also be verified, however. Another feature that was usually thought to distin­guish Salamandrella from Primorye is the absence of nuptial behavior of the kind observed in S. keyserlingii (Korotkov, 1977; Basarukin and Borkin, 1984; Borkin, 1994). However, G.P. Sapozhnikov observed well-man­ifested courtship rituals performed by males, which were identical to those described by us in the upper reaches of the Kolyma River. This researcher believes that the nuptial behavior of males in Primorye is rarely observed in the daytime, but their activity probably increases at night (Sapozhnikov, personal communica­tion). In the water bodies containing only straight spawn (undoubtedly belonging to S. schrenckii), we observed only weak manifestations of male nuptial behavior, which did not differ from that in S. keyser-lingii. The presence of unique nuptial behavior resem­bling a "round dance" in Salamandrella from Primorye, which was described by Yu.B. Pukinskii (cited from Grigor'ev, 1981), should also be checked. Note that, in Primorye, constant and temporary creeks are used by Salamandrella for reproduction more fre­quently than in other parts of the range (Kuzmin, 1999).

---

^[5] During the preparation of this paper, we analyzed mtDNA of 22 spawn samples collected in the Ussuriiskii Nature Reserve. All of them were shown to belong to S. schrenckii, irrespective of the degree of coiling. No correlation between the spawn shape and the mtDNA haplotype was revealed.


According to the molecular data obtained by us, S. schrenckii is divided into two subclusters with a fairly large genetic distance between them: on average, 3.2%, with a maximum of 3.5% (Table 2). Note that the two haplotypes occur together. Such a great difference formally corresponds to the subspecies rank; however, comparative data on the intraspecific structure in mem­bers of various genera of the family Hynobiidae are as yet absent.

On Factors Accounting for Insignificant Geographic Variation in the Siberian Newt 

The low level of genetic variation in S. keyserlingii from different parts of the vast species range with dif­ferent natural conditions requires explanation. The geo­graphic populations examined display such a great genetic uniformity that they appear to form a single hyperpopulation (except for the above-mentioned Sala-mandrella from Sakhalin). This phenomenon is surpris­ing, as it is impossible to propose a mechanism for effi­cient gene exchange between marginal populations which could have provided for this uniformity.

The simplest way to explain the insignificant geo­graphic variation (both morphological and genetic) is to attribute it to late formation of the present-day range, possibly, during one of the most humid periods of the Holocene. The genetic uniformity of the hyperpopula-tion suggests the rapid distribution of one haplotype. A similar phenomenon was described using molecular genetic data in the three-toed (Zink et al., 2002b) and great spotted woodpeckers (Zink et al., 2002a), which colonized Eurasia in the postglacial period along with the restoration of forests. It is evident that Salaman-drella has a much more restricted migration ability (compared to birds), since it is confined to water bodies during reproductive period and to moist habitats at other stages of the life cycle. The outstanding (for amphibians) ability to live at low summer temperatures (which is evident from the successful existence of this species in the southern tundra subzone) suggests that abundant moistening or, more precisely, the presence of many small water bodies of different types, rather than heat supply, could be the factor promoting the expan­sion of the Siberian newt. Note that in northeastern Asia, interfluves with elevations of up to 1000 m above sea level, covered with forests or mountain tundra, do not prevent its expansion (Berman and Sapozhnikov, 1994). 

In principle, not only rapid, but also gradual forma­tion of the range was possible even during the cold phases of the Pleistocene. This was probably facilitated by the outstanding cold hardiness of the Siberian newt: during wintering, it survives temperatures as low as - 40°C (Berman et al., 1984). This is not in conflict with the age of this species, which was estimated by us as 490 ka based on mtDNA variation. However, the conclusion concerning fast and recent expansion of the species is supported by the undoubted absence of gene exchange between marginal populations during recent times (see above). 

Berman (2003) examined the biological and ecolog­ical mechanisms providing for the maintenance of vari­ation at the same level in different local populations in order to find another explanation for the genetic unifor­mity combined with low morphological variation in S. keyserlingii throughout the continental part of its range (Berman, 2003). The extremely low level of geo­graphic variation within such an extensive range seems extraordinary, since environmental conditions in its extreme zones differ greatly and, hence, the abiotic component of natural selection must be well-mani­fested. In spring, when ambient temperatures increase above zero, the conditions in different natural zones become more or less similar and the impact of the cli­mate is alleviated. In summer and autumn, however, this impact must be significant because of differences in the sum of temperatures and their biocenotic conse­quences. The young of the year are especially sensitive to climatic conditions: their growth rate, body size, and other parameters strongly depend on the period of time between their emergence on land and the onset of the cold season, which ranges from a few days in cold years in the northern taiga (in the upper reaches of the Kolyma River) to 1.5-2 months in the southern taiga (Novosibirsk) and forest-steppes (the Baraba region).

There are also other, ordinary factors responsible for the heterogeneity of populations, for example, variation in conditions of animal growth, development, repro­duction, etc., in different aquatic and terrestrial habi­tats. The initial stages of the divergence of local popu­lations (Ishchenko, 1989; Siberian Newt..., 1995) are short-term, probably accomplished within the life span of one generation. Initial differentiation remains insig­nificant, however, as the individuality of populations is smoothed because of group fertilization of females,[1] and wide, fan-shaped migrations of the young, which result in constant intermixing. Both factors apparently provide for assimilation of deviant varieties, thus main­taining the morphological standard of the population and, eventually, restricting geographic variation (Ber-

man, 2003). It is intuitively clear that the absence of geographic variation is a consequence of unidirectional stabilizing selection throughout the species range. Selection may be directed toward the maintenance of adaptation to winter climatic conditions, which are extremely severe in almost all parts of the vast range of the Siberian newt. Due to the specific features of the Siberian High (with the values of January isotherms decreasing from west to east rather than from north to south), differences in winter temperatures are leveled off from the tundra to the steppe, irrespective of natural zonality (Alfimov, 2005). The most severe conditions are observed at the southern boundary of the range in Siberia, where low ambient temperatures are

combined with the absence of snow cover (Berman, 2003). Unfortunately, we cannot yet propose a concrete mechanism of natural selection.

---

^[6] The possibility of group fertilization in S. keyserlingii was put into doubt by Savel'ev et al. (1993).

CONCLUSIONS^[7]

Thus, the results of this study confirm that wide Palearctic species ranges have a specific phylogeo-graphic structure in which populations from the south­east clearly stand out against a background of general genetic uniformity of virtually all other populations. A similar situation was previously described for the car­rion crow Corvus corone, in which southeastern popu­lations (from Primorye and southern Sakhalin) were distinguished by a special mtDNA haplotype (Kryukov and Suzuki, 2000). Moreover, a similar division of the range was revealed in the great spotted woodpecker Dendrocopos major: one mtDNA cluster comprised variants characteristic of the southeastern area of the range (including Sakhalin, Primorye, and Hokkaido), while the other included those of other areas of the range in northern Eurasia (Zink et al., 2002a). Based on the sequences of two mitochondrial genes of the mag­pie Pica pica, its range was also divided into two parts, with the tentative boundary between them lying along the Amur River (Kryukov et al., 2004). Increased vari­ability of mtDNA in the population from Primorye was also revealed in the course of wide geographic studies on variation in the mouse Apodemus peninsulae (Serizawa et al., 2002). The situation with Salaman-drella and the other examples discussed in this paper is obviously accounted for by the same factor. Primorye is a refugium well-known for its highly stable natural conditions, at least, in the Late Quaternary; unlike other areas of the boreal Palearctic, it did not suffer from cat­astrophic breaks in development such as those pro­duced by glaciations (Kolesnikov, 1969; Nazarenko, 1990). Hence, it is not surprising that the age of S. schrenckii is much greater than that of S. keyserlingii calculated by the same method. It is estimated as 2.4 Ma, which suggests that S. schrenckii is a native inhabitant of Primorye. We believe that the existence of two haplotypes in S. schrenckii, the greatest genetic distance between which is 3.5% (see above), is also associated with the early emergence of this species. ^[7] During the preparation of this paper, the sample of mtDNA sequences of Salamandrella schrenckii and S. keyserlingii increased approximately twofold (from 86 to 160 specimens) due to analysis of animals from new areas (southern Primorye, south­ern Khabarovsk krai, the lower reaches of the Kolyma River, the upper reaches of the Indigirka River, and the vicinity of Tomsk) and additional experiments with mtDNA variants described above. The data obtained confirm the conclusions made in this paper.

Although the age estimates for the two Salaman-drella species are tentative and approximate, it is possi­ble to conclude, taking into account only the magnitude of values and their ratios, that S. keyserlingii and S. schrenckii are two relatively young species of differ­ent ages, the ancestral lineages of which diverged from a common stem approximately 14 Ma (estimated according to the calibration accepted in this study), rather than a descendant and its ancestor. The scheme of relationships and the time of the divergence of these species are shown in Fig. 6. 

These estimates appear to be quite credible, since the available data on the diver­gence of caudate amphibians within a genus (in partic­ular, in the genus Euproctus used here as an outgroup) are similar to ours (Tan and Wake, 1995; Caccone et al., 1997; Riberon et al., 2000).


The high level of genetic differences between Sala-mandrella from Primorye and from all other regions strongly suggest the existence of two species. However, they are difficult to distinguish morphologically. At present, it is impossible to identify each individual without employing molecular genetic methods, which provides a basis for applying the term "cryptic species" to Salamandrella schrenckii. In this connection, it is important to outline the range of S. schrenckii and, in particular, its northern boundary, which is the zone of contact between the two species. To date, the spread of S. schrenckii to the north has been delimited by two points in the vicinities of the villages of Georgievka and Sidim in Khabarovsk krai, which are located several dozens of kilo­meters south and north of the 48th parallel.

ACKNOWLEDGMENTS 

We are sincerely grateful to N.B. Anan'eva and V.F. Orlova for the opportunity to examine the herpe-tological collections of the Zoological Institute of the Russian Academy of Sciences and the Zoological Museum of Moscow State University; to P.A. Gudkov, N.E. Dokuchaev, V.G. Ishchenko, A.N. Leirikh, and G.P. Sapozhnikov for placing specimens at our dis­posal; to I.M. Kerzhner for valuable discussions on the systematics; and to Ya. Charnyi, M. Voznyak, E. Levandovskaya, and E. Meshcheryakova for their help in laboratory studies. 

This study was supported by the Russian Founda­tion for Basic Research, project no. 04-04-48187-a; by the Far East Division of the Russian Academy of Sci­ences, project no. 04-3-A-06-009; and by Ludwik Rydygier Medical Academy, project no. BW 59/05.

REFERENCES

1.      Alfimov, A.V., "Distribution of Minimal Temperatures in the Surface Soil Layer under Snow in Northern Eurasia,"

Pochvovedenie, No. 4, 438-445 (2005).

2.      Avise, J.C., Molecular Markers, Natural History, and Evolution (Chapman and Hall, London, 1994), pp. 1­482.

3.      Basarukin, A.M. and Borkin, L.A., "Distribution, Ecol­ogy, and Morphological Variation of the Siberian Newt Hynobius keuserlingii Dub. in Sakhalin Island," in Ecol­ogy and Faunistics of Amphibians and Reptiles of the Soviet Union and Neighboring Countries (Zool. Inst. Akad. Nauk. SSSR, Leningrad, 1984), pp. 12-54 [in Russian].

4.      Belimov, G.T. and Sedalishchev, V.T., "On the Ecology of the Siberian Newt Hynobius keuserlingii Dub. in Cen­tral Yakutia," Biol. Nauki, No. 10, 37-13 (1983).

5.      Berman, D.I., "An Ideal Timeserver, or an Adaptive Strategy of Hynobius keuserligii Dub.," in Russian Sci­ence: Nature Destined Us to..., Ed. by Skulachev, V.P. (Octopus, Moscow, 2003), pp. 184-200 [in Russian].

6.      Berman, D.I. and Sapozhnikov, G.P., "Spatial Population Structure of the Siberian Newt Hynobius keuserlingii Dub. in the Upper Kolyma Basin," Ekologiya, No. 2, 79­88 (1994).

7.      Berman, D.I., Leirikh,A.N., and Mikhailova, E.I., "On Wintering of the Siberian Newt (Hynobius keuserlingii Dub.) in the Upper Reaches of the Kolyma," Zh. Evol. Biokhim. Fiziol. 20 (3), 323-327 (1984).

8.      Berman, D.I., Derenko, M.V., Malyarchuk, B.A., et al., "Genetic Polymorphism of Siberian Newt (Salaman-drella keuserlingii, Caudata, Amphibia) in Its Range and the Cryptic Species S. schrenckii from Primorie," Dokl. Akad. Nauk 403 (3) (2005) [Dokl. Biol. Sci 403, 275­278 (2005)].

9.      Borkin, L.Ya., "Systematics," in Siberian Newt: Zoogeog­raphy, Systematics, and Morphology, Ed. by Vorob'eva, E.I. (Nauka, Moscow, 1994), pp. 54-80 [in Russian].

10.   Borkin, L.Ya., Litvinchuk, S.N., Mil'to, K.D., et al., "Cryptic Speciation in Pelobates fuscus (Amphibia, Pelobatidae): Cytometric and Biochemical Evidence," Dokl. Ross. Akad. Nauk 376 (5), 707-709 (2001) [Dokl.

Biol. Sci. 376, 86-88 (2001)].

11.   Bromlei, G.F., Vasil'ev, N.G., Kharkevich, S.S., and Nechaev, V.A., Flora and Fauna of the Ussuri Reserve (Nauka, Moscow, 1977) [in Russian].

12.   Caccone, A., Milinkovitch, M.C., Sbordoni, V., and Powell, J.R., "Mitochondrial DNA Rates and Biogeogra-phy in European Newts (Genus Euproctus)," Syst. Biol. 46, 126-144 (1997).

11.    Chernichko, I.A., "Amphibians and Reptiles," in Flora and Fauna of the Sikhote Alin Reserve (Nauka, Moscow, 1982), pp. 191-195 [in Russian].

12.    Dybowski, B., "Uber die Urodelen Ostsibiriens," Bul. Acad. Polonaise Sci. Lettres Classe Sci. Math. Nat., Ser. B: Sci. Nat., Cracovie 8/10, 1073-1081 (1928).

13.    Fu, J., Wang, Y., Zeng, X., et al., "Genetic Diversity of Eastern Batrachuperus (Caudata: Hynobius)," Copeia 4, 1100-1107 (2001).

14.    Garsia-Paris, M., Good, D.A., Parra-Olea, G., and Wake, D.B., "Biodiversity of Costa Rican Salamanders: Implications of High Levels of Genetic Differentiation and Phylogeographic Structure for Species Formation," Proc. Natl. Acad. Sci. USA 97, 1640-1647 (2000).

15.    Goebel, A.M., Donnelly, J.M., and Atz, M.E., "PCR

Primers and Amplification Methods for 12S Ribosomal DNA, the Control Region, Cytochrome Oxidase I, and Cytochrome B in Bufonids and Other Frogs, and an Overview of PCR Primers Which Have Amplified DNA in Amphibians Successfully," Mol. Phylogenet. Evol. 11 , 163-199 (1999).

18.    Grigor'ev, O.V., "Nuptial Dances of the Siberian Newt Hynobius keuserlingii Dub.," Priroda, No. 3, 104-105

(1981).

19.    Ishchenko, V.G., "Ecological Mechanisms for Support of Population Stability of Amphibians," Ekologiya, No. 2, 12-19 (1989).

20.    Ishchenko, V.G., Ledentsov, A.V., Godina, L.B., and Kuz'min, S.P., "Development and Growth," in Siberian Newt: Ecology, Behavior, and Conservation, Ed. by Vorob'eva, E.I. (Nauka, Moscow, 1995) [in Russian].

21.    Khalturin, M.D., Litvinchuk, S.N., Borkin, L.Ya., et al., "Genetic Variation in Two Forms of Pelobates fuscus (Pelobatidae, Anura, Amphibia) Differing in the Genome Size," Tsitologiya 45 (3), 308-323 (2003).

22.    Kolesnikov, B.P., "Vegetation," in The Southern Far East (Nauka, Moscow, 1969), pp. 206-250 [in Russian].

23.    Korotkov, Yu.M., "On the Ecology of Triton (Onycho-dactylus fischeri) and the Siberian Newt (Hynobius keu­serlingii) in Primorskii Krai," Zool. Zh. 56 (8), 1258­1260 (1977).

24.    Kryukov, A.P. and Suzuki, Kh., "Phylogeography of Carrion, Hooded, and Jungle Crows (Aves, Corvidae) Inferred from Partial Sequencing of the Mitochondrial DNA Cytochrome b Gene," Genetika 36 (8), 1111-1118

(2000) [Russ. J. Genet. 36 (8), 922-929 (2000)].

25.    Kryukov, A., Iwasa, M.A., Kakizawa, R., Suzuki, H., Pinsker, W., and Haring, E., "Synchronous East-West Divergence in Azure-Winged Magpies (Cyanopica cya-nus) and Magpies (Pica pica)," J. Zool. Syst. Evol. Res.

42, 342-351 (2004).

26.    Kumar, S., Tamura, K., Jakobsen, I.B., and Nei, M., MEGA2: Molecular Evolutionary Genetics Analysis Software (Arizona State University, Tempe, Arizona,

USA, 2001).

27.    Kuzmin, S.L., Amphibians of the Former USSR (Tova-rishchestvo Nauchnykh Izdanii KMK, Moscow, 1999) [in Russian].

28.    Kuzmin, S.L. and Maslova, I.V., The Amphibians of the Russian Far East (Pensoft, Sofia, 2003).

29.    Litvinchuk, S.N. and Borkin, L.J., "Variation in Number of Trunk Vertebrae and in Count of Costal Grooves in

Salamanders of the Family Hynobiidae," Contributions to Zoology. SPB Academic Publishing by the Hague 72 (4), 195-209 (2003).

30.    Litvinchuk, S.N., Borkin, L.J., and Rosanov, J.M., "Intra- and Interspecific Genome Size Variation in Hynobiid Salamanders of Russia and Kazakhstan: Determination by Flow Cytometry," Asiat. Herpetol. Res., No. 10, 282-294 (2004).

31.    Litvinchuk, S.N., Rozanov, Yu.M., Borkin, L.Ya., et al., "The Size of Genome and the Problems of Systematics of Urodele Amphibians (with Reference to Salaman-dridae and Hynobiidae)," in Problems of Herpetology (Pushchino-Moscow, 2001), pp. 168-170.

32.    Motchoulski, V., "Insectes de la Siberie rapports d'un voyage fait en 1839 et 1840," Mem. Imp. Acad. Sci. St. Petersburg, Mem. Savants Etrangers 5, 274 (1844).

33.    Nei, M., Molecular Evolutionary Genetics (Columbia University Press, New York, 1987).

34.    Nazarenko, A.A., "Ornithofaunistic Exchange between the Southern and Northern Eurasia on the Eastern Periphery of the Continent: The Last Glacial-Intergla-cial Cycle," Zh. Obshch. Biol. 51 (1), 89-106 (1990).

35.    Nikol'skii, A.M., "Reptiles and Amphibians of the Rus­sian Empire," Zap. Imp. Akad. Nauk, Ser. Fiz.-Mat. 17 (1), 1-518 (1905 (1906)).

36.    Ostashko, N.G., "On Geographic Variation in the Sibe­rian Newt Hynobius keuserlingii," in Proceedings of V All-Union Herpetological Conference on Issues of Herpetology (Nauka, Leningrad, 1981), p. 98.

37.    Riberon, A., Miaud, C., Grossenbacher, K., and Taber-let, P., "Phylogeography of the Alpine Salamander, Sala-mandra atra (Salamandridae), and the Influence of the Pleistocene Climatic Oscillations on Population Diver­gence," Mol. Ecol. 10, 2555-2560 (2000).

38.    Ruzskii, M.D., "Gustav Radde as a Pioneer Explorer of Siberia," Tr. Biol. Nauchno-Issled. Inst. Tomsk. Gos. Univ. 4, 310-311 (1937).

39.    Saitou, N. and Nei, M., "The Neighbour-Joining Method: A New Method for Reconstructing Phyloge-

netic Trees," Mol. Biol. Evol. 4, 406-425 (1987).

40.    Sapozhnikov, G.P., "Some Specific Features of Develop­ment and Growth of Larval Siberian Newt Salaman-drella keuserlingii," Tr. Zool. Inst. Akad. Nauk SSSR 213, 75-90 (1990).

41.    Savel'ev, S.V., Kuranova, V.N., and Besova, N.B., "Reproduction of the Siberian Newt Salamandrella keu-

serlingii," Zool. Zh. 72 (8), 59-69 (1993).

42.    Serizawa, K., Suzuki, H., Iwasa, M.A., et al. , "A Spatial Aspect of Mitichondrial DNA Genealogy in Apodemus peninsulae from East Asia," Biochem. Genet. 40 (5/6),

149-161 (2002).

43.    Siberian Newt: Ecology, Behavior, and Conservation, Ed. by Vorob'eva, E.I. (Nauka, Moscow, 1995) [in Rus­sian].

44.    Steinfartz, S., Veith, M., and Tautz, D., "Mitochondrial Sequence Analysis of Salamandra Taxa Suggests Old Splits of Major Lineages and Postglacial Recoloniza-tions of Central Europe from Distinct Source Popula­tions of Salamandra salamandra," Mol. Ecol. 9, 397­410 (2000).

45.    Strauch, A., "Revision der Salamandriden-Gattungen nebst Beschreibing einiger oder weniger bekannten

Arten dieser Familie," Mem. Acad. Imp. Sci. St. Peters­burg 16 (4), 1-110 (1870).

46.    Sytina, L.A., Medvedeva, I.M., and Godina, L.B., Devel­opment of the Siberian Newt (Nauka, Moscow, 1987) [in Russian].

47.    Tan, A.M. and Wake, D.B., "mtDNA Phylogeography of the California Newt, Taricha torosa (Caudata, Salaman-dridae)," Mol. Phylogenet. Evol. 4, 384-394 (1995).

48.    Vorob'eva, E.I., Antipenkova, T.P., and Hinchliff, J.R., "Characteristics of Limb Development in the Far Eastern Population of the Siberian Newt (Salamandrella keuser-lingii, Hynobiidae, Caudata)," Dokl. Akad. Nauk 364 (1), 130-133 (1999) [Dokl. Biol. Sci. 364, 34-37

(1999)].

49.    Zhang, P., Chen, Y.Q., Zhou, H., et al., "The Complete Mitochondrial Genome of a Relic Salamander, Ranodon sibiricus (Amphibia: Caudata), and Implications for Amphibian Phylogeny," Mol. Phylogenet. Evol. 28,

620-626 (2003).

50.    Zink, R.M., Drovetski, S.V., and Rohwer, S., "Phylogeo-graphic Patterns in the Great Spotted Woodpecker Den-drocopos ajor Across Eurasia," J. Avian Biol. 33, 175­178 (2002a).

51.    Zink, R.M., Rohwer, S., Drovetsky, S., et al., "Holarctic Phylogeography and Species Limits of Three-Toed Woodpeckers," Condor 104, 167-170 (2002b).