The origin and genetic relationships of the Baikal seal, Phoca sibirica, were studied by restriction fragment length polymorphism analysis of mitochondrial DNA (mtDNA). Using 17 different six-base recognition restriction endonucleases, we examined 98 Baikal seals, and two other related species, the ringed seal, P. hispida, (n=87), and the Caspian seal, P. caspica, (n=94). Analysis revealed the existence of 87 mtDNA haplotypes in the total of 279 specimens. The haplotypes of each species were divided into different clusters on a dendrogram obtained by UPGMA based on haplotype frequency and mtDNA base substitution. No common haplotypes were found among the species examined. The Baikal seal is much more closely related to the ringed seal than the Caspian seal. The amount of divergence suggested that an ancestor of the Baikal seal came down to the lake approximately 0.4 million years ago as was previously indicated by paleontological studies. The seals examined here showed lower variabilities.
INTRODUCTION
The Baikal seal, Phoca sibirica, is the only species of phocids (true seals) distributed in a freshwater region. This species is endemic to Lake Baikal, located in the midst of Siberia far from the habitat of their marine relatives. The origin of this species has therefore received substantial attention. This species is considered a close relative of the Caspian seal, Phoca caspica, and the ringed seal, Phoca hispida, as classified as the subgenus Pusa (Bonner, 1989) or into a separate genus Pusa (Rice, 1998). The Baikal seal differs externally from the latter two species by its unspotted or rarely spotted appearance (Reeves et al., 1992). The Caspian seal, endemic to the Caspian Sea, is very similar externally to the ringed seal. However, its rings are sparser than on ringed seals, although a few specimens lack rings completely as the Baikal seal. The ringed seal is widely distributed in the circumpolar Arctic and subarctic coasts, and is classified into five subspecies – P. h. hispida in the Arctic basin, P. h. ochotensis in the Seas of Okhotsk and Japan, P. h. saimensis in Lake Saimaa, P. h. ladogensis in Lake Ladoga, and P. h. botnica in the Baltic Sea (Jefferson et al., 1993).
Several reports based on fossil records have been published to date on the origins of the Baikal seal (Kozhov, 1963; Ray, 1976; Repenning et al., 1979; Davies, 1958; MacLaren, 1960). The results of these studies present two primary hypotheses on the origins of the Baikal seal. Ray (1976) and Repenning et al. (1979) suggested that about 2.5 to 3 million years ago, when the Arctic Ocean extended below 61° north latitude, an ancestor of Pusa gained access to the Arctic Ocean from the Paratethys of southeastern Europe, and then about 300,000 years ago migrated to Lake Baikal via the Yenisey River. Alternatively, the Pusa group may have evolved along with the Baikal, Caspian and ringed seals from a common ancestor in the trace of the Paratethyan Basin, and then moved eastward through canals and lakes formed during glaciation. Some of these groups may have eventually settled in Lake Baikal, while the ancestor of the ringed seal continued northward and distributed over a wide range in the Arctic Ocean (Davies, 1958; MacLaren, 1960). The latter hypothesis suggests that the ancestor of the Baikal seal reached Lake Baikal much earlier than suggested by the former hypothesis.
Studies on the biochemical or molecular evolution of the Baikal seal have been rather scarce to date. There is no report on the differences between the closely related species of the subgenus Pusa, but for Malikov et al. (1997) examining nucleotide sequence divergence of the Baikal seal with more distantly related harbor and spotted seals.
In the present report, the origin and differentiation time of the Baikal seal are examined by analyzing restriction fragment length polymorphisms (RFLP) of mtDNA in 279 specimens representing three species of the subgenus Pusa; the Baikal seal, Caspian seal, and the ringed seal. Intra-and interspecific genetic variability are compared with among the Pusa species. We present and discuss data on the differentiation time of the Baikal seal ancestor from the other two species ancestor.
MATERIALS AND METHODS
Samples and Restriction Analysis
A total of 279 specimens, consisting of three related species (98 Baikal seals, 94 Caspian seals, and 85 ringed seals), were sampled from a wide range of localities (Fig. 1). Crude total DNAs isolated by SDS / phenol extraction from muscle or liver tissues of each animal were digested with 17 different six-base sequence recognition endonucleases: Acc I, Apa I, Ava I, Bam H I, Bgl I, Bgl II, Cla I, Dra I, EcoR I, EcoR V, Hind III, Hpa I, Kpn I, Nco I, Pst I, Pvu II, and Ssp I. The resulting mtDNA fragments were electrophoresed in a horizontal 0.8% agarose gel, transferred to nylon membranes. MtDNA fragments were visualized with digoxigenin labeled closed circular mtDNA probes from chum salmon, Oncorhynchus keta, or Caspian seals. The probes were purified from liver tissue by CsClethidium bromide density gradient ultracentrifugation (Numachi et al., 1990). The salmon mtDNA probe exhibited the same fragmental patterns as the seal probe (Sasaki et al., 2001).
Data Analysis
Specimens were classified into the composite genotypes (haplotypes) assigned by the arrangement of fragmental patterns produced by each restriction enzyme. Sequence diversity D, nucleotide diversity π, nucleon diversity H and net nucleotide difference between populations were calculated (Nei and Li, 1979; Nei and Tajima, 1981; Nei, 1987). Quantification of the genetic differentiation of mtDNA was completed by analysis of molecular variance (AMOVA) (Excoffier et al., 1992). Phylogenetic relationships of haplotypes and populations were assessed using the unweighted paired group method with arithmetic average (UPGMA) (Nei, 1987) and neighbor-joining method (Saitou and Nei, 1987) included in PHYLIP (version 3.572, Felsenstein, 1993). Phylogenetic relationship among haplotypes was caluculated on the basis of a matrix of the number of base substitutions per nucleotide. When comparing among populations, the net difference among them was estimated by subtracing the average level of variation within populations from the level of variation among them (Nei, 1987). Haplotypes of the harbor seal and gray seal deduced from the complete mtDNA sequences were used as an outgroup (Arnason and Johnsson, 1992; Arnason and Gullberg, 1993).
RESULTS AND DISCUSSION
Haplotypes Demonstrated in Pusa
A total of 98 restriction sites was identified in the 279 samples examined. Of these sites, 67 (68.37%) were polymorphic. Specimens were classified into the haplotypes assigned by the arrangement of fragmental patterns produced with each of the enzymes. In total, 87 different haplotypes were discriminated from the 279 specimens. Fragment patterns data for these haplotypes are available from one of the authors (HS) upon request.
Relationships Among Haplotypes and Species
Haplotypes of each species were clearly divided into their own distinct clusters, with no common haplotype among the species (Fig. 2). The results clearly show that each species is polymorphic, represented by 12 to 47 haplotypes, and the haplotypes found within a species are genetically similar to each other and distinct from those of other species.
Fig. 2
Genetic relationships of the 87 haplotypes of mtDNA demonstrated in all the total of 279 specimens examined of the three species of subgenus Pusa. The harbor seal, Phoca vitulina and gray seal, Halichoeus grypus were added as the outgroup. The UPGMA method based on a matrix of the number of base substitutions per nucleotide was used (Nei, 1987). Numeral on right side of the dendrogram represents haplotype ID number.
Population Structure of the Examined Species
Heterogeneity of population samples from different localities and different sampling years was examined in the Baikal and Caspian seals by AMOVA, using the haplotypic frequencies and differences between haplotypes. Neither the Baikal or Caspian seal exhibited a heterogeneous population structure (P > 0.05). Results of the Baikal seal agreed with a previous report which revealed that the Baikal seal moves throughout the entire basin of Lake Baikal (Stewart et al., 1996). Three sample sets of the Caspian seals examined, n=18 from a frozen area in the northeastern quadrant in February 1993, and n=30 and n=46 from a more southern area of the quadrant in November 1992 and November 1993 respectively showed no genetic heterogeneity (P > 0.05). Jefferson et al. (1993) reported that Caspian seals are prompted to move to the northeastern quadrant by ice formation and south into deeper and cooler regions after the pupping season. Average value of haplo-type frequency presented here might represent the genetic composition of this species.
Genetic Differentiation of the Ringed Seal Subspecies
The haplotypes discriminated in the ringed seals collected from various locations were all assigned to one group (Fig. 2). The ringed seals examined in the present study were previously classified into three subspecies according to their distribution range (Nishiwaki, 1965; Jefferson et al., 1993): the arctic ringed seal, which is distributed in the Arctic Ocean including the Kara Sea and White Sea, the Baltic ringed seal, distributed in the Baltic Sea, and the Okhotsk ringed seal, distributed in the Okhotsk Sea. However, the haplotype distribution in four population samples belonging to these subspecies from different localities was complex, as shown in Fig. 3. AMOVA indicated that the common ringed seal from the White Sea differed from the other populations (P < 0.05). In addition, the common ringed seal from the Kara Sea differed from the Okhotsk ringed seal (P < 0.05).
Fig. 3
Comparison of haplotype frequency among the four populations of the ringed seal and dendrogram of the mtDNA haplotypes of the ringed seal obtained by the UPGMA. Numeral on right side of dendrogram represents haplotype ID number. PHB; Phoca hispida botnica in Baltic Sea, PHW; P. h. hispida in White Sea, PHK; P. h. hispida in Kara Sea, and PHO; P. h. ochotensis in Okhotsk Sea.
Intraspecies Genetic Variability
Both H and π values were lowest in the Baikal seal, and highest in the ringed seal inclusively (Table 1). The Caspian seal showed intermediate values. All of the ringed seal sub-species also showed higher values, although samples from the White Sea showed monomorphism. These lower values in the Baikal and Caspian seals may be caused by the smaller population sizes in the lake. The seal population in Lake Baikal was estimated to be 58,000 in 1986 (Reeves et al., 1992). In the Caspian Sea, the seal population rewrite according to Ridgeway and Harrison (1981) estimated at approximately 450,000 – 600,000. Monomorphism in the ringed seal from the White Sea remains to the further studies.
Table 1
Genetic variability within the Pusa species by nucleon diversity (H) and nucleotide diversity (π)
Genetic Relationships Among Pusa Species and Origin of the Baikal Seal
The UPGMA dendrogram constructed from a matrix of net nucleotide divergence revealed that the Baikal seal is much more closely related to the ringed seal, and the Caspian seal was positioned far from the other two (Fig. 4a). The dendrogram obtained by the neighbor-joining method also showed a similar result (Fig. 4b). Skull morphometry based on the same specimens as the present study supports our results (Koyama et al., 1997). The Baikal seal presumably came from the north and not directly from the area west of the old Paratethyan basin, which was located in the present Caspian Sea. Our results support the paleontological and geographical hypotheses presented by Ray (1976) and Repenning et al. (1979).
Fig. 4
Dendrogram (A) and (B) of the four species of subgenus Pusa including harbor seal, Phoca vitulina and gray seal, Halichoeus grypus, as the outgroup, respectively obtained by the UPGMA method and the Neighbor-joining method which was based on the net difference among them was estimated by subtracing the average level of variation within populations from the level of variation among them (Nei, 1987).
Differentiation Time
The period of seal immigration into Lake Baikal via the Yenisey River appears to have occurred approximately 0.4 million years ago from estimates of genetic distance between these seal species (Table 2), assuming that the rate of nucleotide substitution in mtDNA is 2% per one million years, as is suggested for animal species (Hoelzel and Dover, 1991). This strongly supports the immigration course and time suggested by Ray (1976) and Repenning et al. (1979). On the other hand, the common ancestor of the Baikal and ringed seals was estimated to have diverged from the Caspian seal approximately 0.7 million years ago. This figure is much more recent than the 2.5 to 3 million years estimated by the above paleontological studies. The close relationship between the Pusa species and harbor seals shown in the present study suggests that differentiation time of the Pusa species and harbor seals is more recent than previously suggested.
Table 2
Estimates of percent of net nucleotide substitution between the seal species examined.
Acknowledgments
The authors thank Dr. E. Petrov of the Limnological Institute, Russian Academy of Science, Irukutsk, Russia, Dr. K. Sergeevich of the Caspian Scientific Research Institute, Astrahan, Russia, Dr. O. Minnibaeva, Sev PINRO, Russia, and Dr. V. A. Vladimirov, Marine Mammal Laboratory, VINIRO, Russia, for collecting samples and their technical support in the field. We are also grateful to Dr. M. Amano and Mr. Y. Koyama, Ocean Research Institute of University of Tokyo, Japan, Dr. M. Olsson, Swedish Museum of Natural History, Sweden, the staff of Otaru Aquarium in Japan and Dr. T. Oike D.V.M. and the staff of Minami Chita Beachland Aquarium in Japan for their kind donations of the Baikal, Caspian and ringed seal samples.
