Instructions for use

Title Restriction fragment length polymorphism of nuclear rDNA in Sorex caecutiens/shinto group (Eulipotyphla, Soricidae)

Author(s) Naitoh, Yukako; Iwasa, Masahiro A.; Ohdachi, Satoshi D.; Han, Sang-Hoon; Suzuki, Hitoshi

Citation Mammal Study, 30(2): 101-107

Issue Date 2005

Doc URL http://hdl.handle.net/2115/44410

Type article

File Information MS30-2_101-107.pdf

Hokkaido University Collection of Scholarly and Academic Papers : HUSCAP

MammalStudソ30:101-107(2005)

◎theMammalogicalSocietyofJapan

Restrictionfragmentlengthpolymorphismofnuclearrl)NAin

Sorexcaecutiens/shintogroup(Eulipotyphla,Soricidae)

YukakoNaitohi,MasahiroA.Iwasa2,SatoshiD.Ohdachii・de,Sang-HoonHan3

andHitoshiSuzuki`

11nstitute(～ プ五〇wTemperatz〃eScience,HokkaidoUniγersity,SapporoO60-0819,.ノOPan

2五ab・rat・ry・fWildlifeScience,DepartmentofAnimalScien・e・ndRes・urces,C・llege・fBi・res・urceS・iences,嶺 ・n

しiniversiリ ノ,.Ka〃leino1866,Fzグisa「 レレo,.Kanaga「wa252-8510,.ノO卯an

3∠tsiatic、Bla(]k.Bear一 ル1αnage〃zent1セa〃z,ノ 〉∂tionalParks∠{uthoriり ノ,511-1〃7hangleon-ri,ハ4asan-〃1アeon,G〃 ぞアe-gun,

泥o〃a-1>∂md()542-853,」Republic(～ プKorea

4五aboratorP・ofEcologyandGenetics,GraduateSchool()fEnvironmentalEarthScience,HokkaidoUnivez吻,Saρporo

O60-0810,.ノOPan

Abstract.WeestimatedphylogeneticrelationshipsamongshrewsoftheSorexcaecu舵 η誘 伽'o

group(Eulipotyphla,Soricidae)fヒomvariouslocationsthroughitsrange,basedonrestriction

丘agmentlengthpolymorphism(RFLP)analysisofthenuclearribosomalRNAgene(rDNA)spacer

region.SevenrDNA-RFLPrepetitivetypes(repetypes)wererecognizedamongl5shrewsexam-

ined.RestrictionpatternsofSoreac60θ6〃 舵 η5Laxmann,1788and3.sh枷oThomas,1905were

distinguishablefromeachother,buttheseparationwasnotstatisticallysupportedinthemaximum

parsimonyanalysis.TheRFLPrepetypefヒomCh(加Islandwasclosetothatof3.6α ε6〃舵 η8fヒomthe

Eurasiancontinent,indicatingthattheshrewofCh句ushouldbeclassi∬edas3.oα θ磁'θ η5.Within3.

caecutiens,thereweretwoaltemativephylogenetichypotheses.Accordingtoaparsimonioustree

andasimplenetwork,theHokkaidopopulationwasregardedtobederived仕omtheSakhalin

population,whichintumwasderivedfヒomthecontinentalpopulation.Altematively,itwasinferred

thatthecontinentandHokkaidopopulationswerefirstlyseparatedfromtheancestralpopulation,

andthenshrewsfrombothpopulationsimmigratedintoSakhalinandhybridizationoccurredthere.

Thelatterhypothesisseemstobemoreplausiblebecauseitismorecongruentwithaprevious

mitochondrialphylogeny.

Keywords:ChejOuIsland,nuclearrDNA,RFLP,Soreaccaecutiens.

TaxonomicranksofSorexcaecutiensLaxmann,1788

and3.shintoThomas,1905hadbeendebatedbyvarious

authorsbecauseinterpretationoftheirmorphologyis

difficult(seeDokuchaevetal.1999)andlittlephyloge-

neticinfbrmationwasavailablefbrtheseshrews.How-

ever,afterphylogeneticinvestigationsofsomespeciesof

Sorex(Georgel988;Ohdachietal.1997;Fumagalliet

al.1999),itisnowwidelyacceptedthatthesetwospe-

ciesaredistinct.Further,Ohdachietal.(1997,2001,

2003)revealedbasedonmitochondrialcytochromeわ

gene(mtDNAcyt一 わ)sequencesthat3.caecutiensandS.

s伽o鉤 ㎜amonophyleticgroup(3.caecu伽s/s伽o

group),andthat3.caecutiensoccursintheEurasian

continent,Sakhalin,Hokkaido,andneighboringsmall

islandswhile3.shintoisdistributedinthesouthernparts

oftheJapaneseIslands(Honshu,Shikoku,andSado).

FromCh(加Island,SouthKorea,aSorexshrewspecies

wasrecentlydiscovered(H.S.Oh,personalobserva-

tion).Ithasnotyetbeenfbrmallydescribedbutinitial

observationsuggestsitbelongstothe3.caecut'ens/

訪'η'ogroup.Ohdachietal.(2003)showedthatthe

shrewonCh〔 加Islandshouldbeincludedin3.caecu-

'蜘s,basedonmtDNAcyt一 わsequences.

AccordingtothephylogenetictreebasedonmtDNA

cyt一 わ,therearetwomainclustersin3.caecutiens:the

HokkaidoclusterandtheEurasianContinent-Sakhalin一

*7b納0〃icorrespondとnceshOU〃beadと かessed .E-〃iai1'ohd@ρ(4P.IOivte〃1.hokudai.αC.塑

Mamlllal SlIIciy 30: 10 1- 107 (2005) C the Mamma logical Society of Japan

Restriction fragment length polymorphism of nuclear rDNA in Sorex caecutienslshinto group (Eulipotyphla, Soricidae)

Yukako Naitoh', Masahiro A. Iwasa', Satoshi D. Ohdachi" ', Sang-Hoon Han.1 and Hitoshi Suzuki4

/ Instilllte of Low Temperatllre Science, Hokkaido University, Sapporo 060-0819, Japan ] Laboratory of Wildlife Science, Department of Animal Science and Resources, College of Bioresollrce Sciences, Nihon University, Kameino 1866, Fujisawa, Konagawa 252-8510, Japan J Asialic Black Bear Management Team, Nalionol Parks Aulhorify, 51 I- I Whangjeon-ri, JV!ason-myeon, Gwye-gun, Jeolla-Namdo 542-853. Republic of Korea 4 LaboralOry a/Ecology and Genetics, Graduate School oj Environmenlal Earth Science, Hokkaido University. Sapporo 060-08/0, Japan

Ahstract. We est imated phylogenetic re lationsh ips among shrews o f the SOI"ex caecl/fiens/shimo

group (Eulipotyphla, Soricidae) fro l11 va rioll s locat ions through its range, based on restriction fragment length polymorphism (RFLP) ana lys is of the nuclear ribosomal RNA gene (rONA) spacer region . Seven rDNA-RF LP repet itive types (repelypes) were recognized among 15 shrews exam­ined. Restriction patterns of Sorex caeClifiens Laxmann, 1788 and S. shinto Thomas, 1905 were disting uishab le from each other, but the separation was not stat istica ll y supported in the max imum pars imony analysis . The R FLP repetype from Cheju Island was close to that or S. caeClifiens from the Euras ian continent, ind icati ng that the shre\v ofCheju should be class itied as S. caeClifiens. Within S. caeclIfiens, there were two a lternative phylogenetic hypotheses. According to a parsimoniolls tree and a simple netwo rk , the Hokkaido populati on was regarded to be derived from the Sakhalin population, which in turn was deri ved from the continenta l populatio n. Alternati ve ly, it was in ferred ihat the conti nent and Hokkaido populat ions were firs tl y separated from the ancestra l populat ion, and then shrews from hoth populations immigrated into Sakhalin and hyhridization occurred there. The latter hypothesis seems to be more plausible because it is more congruent with a previolls mi tochondrial phylogeny.

Key words: Cheju Island, nuclear rONA, RFLP, Sorex caeclifiens.

Taxonomic ranks of Sorex caecutiens Laxmann, 1788 and S. shil1/o Thomas, 19U5 had been debated by various

authors beca use in te rpretation of their morphology is

difficult (see Dokuchaev et a l. 1999) and little phyloge­netic information was ava ilable for these shrews. How­

ever, afte r phylogenetic invest iga tions of some species of

Sorex (George 1988; Ohdachi et a l. 1997; Fumagalli et

a l. 1999). it is now widely accepted that these two spe­

c ies are di stinct. Further, Ohdac hi et al. ( 1997, 2001, 2003) revealed based on mitochondria l cytochrome b

gene (mtDNA cyt-b) sequences that S. caeclitiens and S.

shinto form a monophy leti c group (s. caecliliens/shinlo

group), and that S. caeclitiens occurs in the Euras ian

continent, Sakhali n, Hokkaido, and neighboring small islands while S. shinto is distributed in the southern parts

of the Japanese Islands (Honshu, Shikoku, and Sado). From Cheju Island, South Korea, a Sorex shrew spec ies

was recentl y discovered (H. S. Oh, personal observa­

tion). It has not yet been formally described but initia l observation sugges ts it belongs to the S. caecutiens/

shinto group. Ohdachi et al. (2003) showed that the shrew on Cheju Is land should be included in S. caecll­

liens, based on mtDNA cyt-b sequences.

According to the phylogenetic tree based on on tDNA

cyt-b, there are two mai n c lusters in S. caeculiens: the Hokkaido c luster and the Eurasian Cont inent-Sakhalin-

· To whol1l corrcspolldclIcesholilci be addressed. E-mail: ohd@pop.lolI.fem.hokudai.ac.jp

Mamlllal SlIIciy 30: 10 1- 107 (2005) C the Mamma logical Society of Japan

Restriction fragment length polymorphism of nuclear rDNA in Sorex caecutienslshinto group (Eulipotyphla, Soricidae)

Yukako Naitoh', Masahiro A. Iwasa', Satoshi D. Ohdachi" ', Sang-Hoon Han.1 and Hitoshi Suzuki4

/ Instilllte of Low Temperatllre Science, Hokkaido University, Sapporo 060-0819, Japan ] Laboratory of Wildlife Science, Department of Animal Science and Resources, College of Bioresollrce Sciences, Nihon University, Kameino 1866, Fujisawa, Konagawa 252-8510, Japan J Asialic Black Bear Management Team, Nalionol Parks Aulhorify, 51 I- I Whangjeon-ri, JV!ason-myeon, Gwye-gun, Jeolla-Namdo 542-853. Republic of Korea 4 LaboralOry a/Ecology and Genetics, Graduate School oj Environmenlal Earth Science, Hokkaido University. Sapporo 060-08/0, Japan

Ahstract. We est imated phylogenetic re lationsh ips among shrews o f the SOI"ex caecl/fiens/shimo

group (Eulipotyphla, Soricidae) fro l11 va rioll s locat ions through its range, based on restriction fragment length polymorphism (RFLP) ana lys is of the nuclear ribosomal RNA gene (rONA) spacer region . Seven rDNA-RF LP repet itive types (repelypes) were recognized among 15 shrews exam­ined. Restriction patterns of Sorex caeClifiens Laxmann, 1788 and S. shinto Thomas, 1905 were disting uishab le from each other, but the separation was not stat istica ll y supported in the max imum pars imony analysis . The R FLP repetype from Cheju Island was close to that or S. caeClifiens from the Euras ian continent, ind icati ng that the shre\v ofCheju should be class itied as S. caeClifiens. Within S. caeclIfiens, there were two a lternative phylogenetic hypotheses. According to a parsimoniolls tree and a simple netwo rk , the Hokkaido populati on was regarded to be derived from the Sakhalin population, which in turn was deri ved from the continenta l populatio n. Alternati ve ly, it was in ferred ihat the conti nent and Hokkaido populat ions were firs tl y separated from the ancestra l populat ion, and then shrews from hoth populations immigrated into Sakhalin and hyhridization occurred there. The latter hypothesis seems to be more plausible because it is more congruent with a previolls mi tochondrial phylogeny.

Key words: Cheju Island, nuclear rONA, RFLP, Sorex caeclifiens.

Taxonomic ranks of Sorex caecutiens Laxmann, 1788 and S. shil1/o Thomas, 19U5 had been debated by various

authors beca use in te rpretation of their morphology is

difficult (see Dokuchaev et a l. 1999) and little phyloge­netic information was ava ilable for these shrews. How­

ever, afte r phylogenetic invest iga tions of some species of

Sorex (George 1988; Ohdachi et a l. 1997; Fumagalli et

a l. 1999). it is now widely accepted that these two spe­

c ies are di stinct. Further, Ohdac hi et al. ( 1997, 2001, 2003) revealed based on mitochondria l cytochrome b

gene (mtDNA cyt-b) sequences that S. caeclitiens and S.

shinto form a monophy leti c group (s. caecliliens/shinlo

group), and that S. caeclitiens occurs in the Euras ian

continent, Sakhali n, Hokkaido, and neighboring small islands while S. shinto is distributed in the southern parts

of the Japanese Islands (Honshu, Shikoku, and Sado). From Cheju Island, South Korea, a Sorex shrew spec ies

was recentl y discovered (H. S. Oh, personal observa­

tion). It has not yet been formally described but initia l observation sugges ts it belongs to the S. caecutiens/

shinto group. Ohdachi et al. (2003) showed that the shrew on Cheju Is land should be included in S. caecll­

liens, based on mtDNA cyt-b sequences.

According to the phylogenetic tree based on on tDNA

cyt-b, there are two mai n c lusters in S. caeculiens: the Hokkaido c luster and the Eurasian Cont inent-Sakhalin-

· To whol1l corrcspolldclIcesholilci be addressed. E-mail: ohd@pop.lolI.fem.hokudai.ac.jp

102 ル勿〃1〃ialStu4ソ30(2005)

Ch()jucluster(Ohdachietal.2003;Ohdachi2005).Fur-

thermore,itwasdemonstratedthatgeneticvariationwas

lowandtherewasalmostnolocaldivergenceofhaplo-

typeamongindividualsof3.caec〃 舵 η8fヒomtheEura-

siancontinentandSakhalin(Ohdachietal.2001,2003).

Basedontheseresults,Ohdachietal.(2001,2003)sug-

gestedthattheancestralpopulationof3.eoeeutie〃8were

firstdividedintotwosubpopulations,oneinHokkaido

andtheotherintheeastempartoftheEurasiancon-

tinent.Then,thelattersubpopulationspreadrapidly

throughoutthecontinentandSakhalinafterthelast

glacialperiod.

PhylogenetictreesinferredfrommtDNA,however,

sometimescontradictthoseinferredfromnuclearDNA

(e.g.Mukai2001;SotaandVogler2001;Iwasaand

Suzuki2002).Insuchcases,taxonomyandevolutionary

eventsinferredfrommtDNAalonewillleadtoerroneous

conclusions.Inthe3.cα θα4〃(～ns/sh'ntogroup,phylo-

geneticinvestigationsbasedonnuclearDNAhavenot

beenconducted.Thus,itisneededtoin色rthephylog-

enyfヒomtheinfbrmationofnuclearDNA.

Inthepresentpaper,weinvestigaterestrictionfrag-

mentlengthpolymorphism(RFLP)ofnuclearriboso-

malRNAgenes(rDNA)andestimatethephylogenetic

relationshipsbetweenspeciesorpopulationsinthe3.

caecutノ θη3/タ痂 η'ogroup.Inaddition,phylogeneticstatus

oftheSorexshrewf士omChq.uIslandisinvestigated.

Herein,weshortlysummarizethenatureofnuclear

rDNA.TherDNAgeneunitmultipliesseveralhundred

timesinanucleargenomeandformsamultigenefamily.

Thus,itisusefUltoestimateanoverallgeneticvariation

ofanucleargenomeofanindividualinapopulation.

TherDNAgeneunitiscomposedofthecodingand

spacerregions.Theformerregions,whichencodehighly

conservedrDNA(18S,5.8S,and28S),aresuitableas

probe-anchoringregions(Arheimetal.1980).Thespacerregions,incontrast,areknowntoevolverapidlyandare

usefultodetectvariationbetweenandwithinspeciesin

phylogeneticinferences(e.g.Arheimetal.1980;Suzukietal.1990;Suzukil994).

Materialsandmethods

Specimens

EightindividualsofS.caecutiensandfiveS.shinto

weresampledfroml210calitiesofnorthemEurasia

(TablelandFig.1).TwoindividualsofSorexshrewspecies(Sorexsp.oftheS.caecutiens/shintogroup)from

ChejuIsland,SouthKorea,werealsoexamined.

S伽hernhybridizationand〃 即 卿9(ゾrestrictionsites

TotalgenomicDNAwasextractedfromlivertissue

preservedin90%ethanolbythephenol/proteinaseK/

SDSmethod(Sambrooketal.1989).ThegenomicDNA

wasdigestedwithsevenrestrictionenzymes,Aatl(A),

Bglll(B),Dral(D),-EcoRI(E),五 伽dIII(H),-Pstl(P)

Table1.Sorexshrewsamplesandtheircollectionlocalitiesexaminedinthisstudy.

Species Repetype LocalityNo, Locality SpecimenCode

S.shinto

Shi-A

Shi.B

Shi-B

Shi.B

Shi-C

1

(∠

弓」

3

4.

Mt,Kamegamori,Shikoku,Japan

NaganoPref.,Honshu,Japan

Ryotsu,SadoIs.,Japan

Ryotsu,SadoIs。,Japan

Mt。Hayachine,Honshu,Japan

SO97/5/25-l

SO96misc-13

HA5961

HA5962

SO97/8/5-1

s.caecutiens

Cae-A

Cae-A

Cae.B

Cae-C

Cae.C

Cae-C

Cae-C

Cae-C

5

6

7

8

9

0

2

3

1

1

1

Kaminokuni,Hokkaido,Japan

Sarobetsumoor,Hokkaido,Japan

Trudovoe,Sakhalin,Russia

Kedrovaya,Primorye,Russia

Mt.Gaya,SouthKorea

Mt.Odae,SouthKorea

Milkovo,Kamchatka,Russia

Pallasj註rvi,Lapland,Finland

SO97/6/7-l

SO94scl

SO94sak5

SO96/7/16-2

SO99misc.9

SO99/10/17-l

SO97misc-82

VH365

Sorexsp.

Sp-Cheju

Sp-Cheju

l

l

l

l

Mt。Halla,ChejuIs.,SouthKorea

Mt.Halla,Ch()juIs.,SouthKorea

SO99/10/12-l

SO99/10/13-1

102

Cheju cluster (Ohdachi et a J. 2003; Ohdachi 2005). Fur­thermore, it was demonstrated that genetic variation was low and there was almost no local divergence of haplo­type among ind ividuals of S. eoeeliliens from the Eura­sian continent and Sakhalin (Ohdachi et a J. 200 I, 2003). Based on these results, Ohdachi et a J. (2001 , 2003) sug­gested that the ancestral populat ion of S. eoeell/iens were first div ided into two subpopu lations, one in Hokkaido and the other in the eastern part of the Eurasian con­tinent. Then, the latter subpopu lat ion spread rapid ly

throughout the continent and Sakha lin after the last glacial period.

Phylogenetic trees inferred from mtDNA, however, sometimes contradict those inferred from nuclear DNA (e.g. Mukai 2001; Sota and Vogler 2001 ; Iwasa and Suzuki 2002). In such cases, taxonomy and evolutionary events inferred from mtDNA alone wil l lead to erroneous conclusions. In the S. caecliliens/shinlo group, phylo­genetic investigations based on nuclear DNA have not

been conducted. Thus, it is needed to infer the phy log­eny from the information of nuclear DNA.

In the present paper, we investigate restriction frag­ment length po lymorphism (RFLP) of nuclear riboso­mal RNA genes (rDNA) and est imate the phylogenetic relationships behveen species or popu lations in the S.

caeclil iens/shinta group. fn addition, phylogenetic status of the Sorex shrew from Cheju Island is investigated. Herein, we shortly summarize the nature of nuclear

Mammol Study 30 (2005)

rDNA. The rDNA gene unit mUltip lies several hundred times in a nuclear genome and form s a ITI ultigene fam ily. Thus, it is useful to estimate an overall genetic variation of a nuclear genome of an individual in a popu lation. The rDNA gene unit is composed of the coding and spacer regions. The fanner regions, which encode highly conserved rDNA ( 18S, 5.8S, and 28S), are suitable as probe-anchoring regions (Arheim et aJ. 1980). The spacer regions, in contrast, are known to evo lve rapidly and are usefu l to detect variat ion between and withi n species in phylogenetic inferences (e.g. Arhe im et aJ. 1980; Suzuk i et aJ. 1990; Suzuki 1994).

Materials and methods

Specimens

Eight individuals of S. caecllliens and five S. shinJO

were samp led from 12 locali ties of northern Eurasia (Table I and Fig. I). Two individuals of Sorex shrew

species (Sorex sp. of the S. eoeell/iens/shin/o group) from Cheju Island, South Korea, were also examined.

SOUlhern hybridization and mapping afrestriction siles Tota l genomic DNA was extracted from liver tissue

preserved in 90% ethanol by the phenol/proteinase K/ SDS method (Sambrook et a J. 1989). The genomic DNA was digested with seven restriction enzymes, Aa/I (A),

8gnJ (B), Dral (D), EeoRl (E), HindllJ (H), Psll (P)

Ta ble 1. Sorer shrew samples and thei r co llection loca liti es examined in this study.

Species Repcrype Loca lity No. Loca lity Specimen Code

S. shima

Shi·A Mt. Kamegamori, Shikoku, Japan 5097/5/25-1

Shi-B 2 Nagano Pref.. Honshu. Japan S096misc-13

Shi-O J Ryotsu, Sado \s. , Japan l lA5961

Shi-B 3 Ryotsu, Sado Is. , Japan HA5962

Shi -C 4 ML Hayachin e, Honshu, Japan 5097/8/5- 1

S. caecllfiens

Cae-A 5 Kmninokuni, Il okkaido, Japan S097/6/7-1

Cae-A 6 Sarobetsu moor, Hokka ido. Japan S094sc l

Cae- B 7 Trudovoe, Sakha lin, Russia S094sak5

Cae-C 8 Kedrovaya, Primorye, Russ ia S096/7/ 16-2

Cae-C 9 Mt. Gaya, South Korea S099misc-9

Cae-C 10 Me Odae, South Korea S099/ IO/l7- 1

Cae-C 12 Milkovo, Kamchatka, Russia S097misc-82

Cae-C 13 Pa llasjarvi, Lapland, Fin land VH365

Sorex sp.

Sp-C hcj u II Mt. I lall a, Chcj u Is., Soulh Korea S099/ I 0112-1

Sp-Cheju II Mt. Halla, Chej u Is., Soulh Korea S099/ I 0/13- 1

102

Cheju cluster (Ohdachi et a J. 2003; Ohdachi 2005). Fur­thermore, it was demonstrated that genetic variation was low and there was almost no local divergence of haplo­type among ind ividuals of S. eoeeliliens from the Eura­sian continent and Sakhalin (Ohdachi et a J. 200 I, 2003). Based on these results, Ohdachi et a J. (2001 , 2003) sug­gested that the ancestral populat ion of S. eoeell/iens were first div ided into two subpopu lations, one in Hokkaido and the other in the eastern part of the Eurasian con­tinent. Then, the latter subpopu lat ion spread rapid ly

throughout the continent and Sakha lin after the last glacial period.

Phylogenetic trees inferred from mtDNA, however, sometimes contradict those inferred from nuclear DNA (e.g. Mukai 2001; Sota and Vogler 2001 ; Iwasa and Suzuki 2002). In such cases, taxonomy and evolutionary events inferred from mtDNA alone wil l lead to erroneous conclusions. In the S. caecliliens/shinlo group, phylo­genetic investigations based on nuclear DNA have not

been conducted. Thus, it is needed to infer the phy log­eny from the information of nuclear DNA.

In the present paper, we investigate restriction frag­ment length po lymorphism (RFLP) of nuclear riboso­mal RNA genes (rDNA) and est imate the phylogenetic relationships behveen species or popu lations in the S.

caeclil iens/shinta group. fn addition, phylogenetic status of the Sorex shrew from Cheju Island is investigated. Herein, we shortly summarize the nature of nuclear

Mammol Study 30 (2005)

rDNA. The rDNA gene unit mUltip lies several hundred times in a nuclear genome and form s a ITI ultigene fam ily. Thus, it is useful to estimate an overall genetic variation of a nuclear genome of an individual in a popu lation. The rDNA gene unit is composed of the coding and spacer regions. The fanner regions, which encode highly conserved rDNA ( 18S, 5.8S, and 28S), are suitable as probe-anchoring regions (Arheim et aJ. 1980). The spacer regions, in contrast, are known to evo lve rapidly and are usefu l to detect variat ion between and withi n species in phylogenetic inferences (e.g. Arhe im et aJ. 1980; Suzuk i et aJ. 1990; Suzuki 1994).

Materials and methods

Specimens

Eight individuals of S. caecllliens and five S. shinJO

were samp led from 12 locali ties of northern Eurasia (Table I and Fig. I). Two individuals of Sorex shrew

species (Sorex sp. of the S. eoeell/iens/shin/o group) from Cheju Island, South Korea, were also examined.

SOUlhern hybridization and mapping afrestriction siles Tota l genomic DNA was extracted from liver tissue

preserved in 90% ethanol by the phenol/proteinase K/ SDS method (Sambrook et a J. 1989). The genomic DNA was digested with seven restriction enzymes, Aa/I (A),

8gnJ (B), Dral (D), EeoRl (E), HindllJ (H), Psll (P)

Ta ble 1. Sorer shrew samples and thei r co llection loca liti es examined in this study.

Species Repcrype Loca lity No. Loca lity Specimen Code

S. shima

Shi·A Mt. Kamegamori, Shikoku, Japan 5097/5/25-1

Shi-B 2 Nagano Pref.. Honshu. Japan S096misc-13

Shi-O J Ryotsu, Sado \s. , Japan l lA5961

Shi-B 3 Ryotsu, Sado Is. , Japan HA5962

Shi -C 4 ML Hayachin e, Honshu, Japan 5097/8/5- 1

S. caecllfiens

Cae-A 5 Kmninokuni, Il okkaido, Japan S097/6/7-1

Cae-A 6 Sarobetsu moor, Hokka ido. Japan S094sc l

Cae- B 7 Trudovoe, Sakha lin, Russia S094sak5

Cae-C 8 Kedrovaya, Primorye, Russ ia S096/7/ 16-2

Cae-C 9 Mt. Gaya, South Korea S099misc-9

Cae-C 10 Me Odae, South Korea S099/ IO/l7- 1

Cae-C 12 Milkovo, Kamchatka, Russia S097misc-82

Cae-C 13 Pa llasjarvi, Lapland, Fin land VH365

Sorex sp.

Sp-C hcj u II Mt. I lall a, Chcj u Is., Soulh Korea S099/ I 0112-1

Sp-Cheju II Mt. Halla, Chej u Is., Soulh Korea S099/ I 0/13- 1

Naitohetal.,RF乙PqプnuclearD八 「Ainshrei・vs 103

andPvull(V).AfterelectrophoresiswithO.7%agarose

gelfbrl7hoursat30V,thedigestedDNAwasimmo-

bilizedonnylonmembrane(Hybond-N,Amersham)and

hybridizedwithdigoxigenin-labeled(DIGDNALabel-

ingKitandDetectionKit,RocheDiagnostics)28Sprobe

(0.7μg/ml)frommouse(Suzukietal.1994).Detection

oftheprobeonthemembranefollowedthesupplier's

protocol.Fragmentlengthsweredeterminedbycom-

parisonwithasizemarker(DNAMolecularWeight

MarkerIIdigoxigenin-labeledlambdaHindlll).Restric-

tionmapsofthe28Sdownstreamspacerregionwere

constructedbyrefeningtothestandardrestrictionmap

ofthel8S,5.8S,and28SrRNAgenesthatareconser-

vativeamongmammalianspecies(Suzukietal.1994;

Iwasaetal.2001).

Phylogeneticona!yses

Allrestrictionsiteswerenumberedfromltol3(Fig.

60。N 80。N

.ノ

、

ぷ避

ぐ

詳

ノ

ギ

/

/

≠

ゲ

縁

溝

Cheju7

話11

Kyushu

や

締 、

Hokkaido

XtTsug・ ・uSt・ait

⊂=⊃S.caecutiens

-s.shinto

9〈'12

Fig.1.Collectionlocalitiesofshrewsamples.LocalitynumberscorrespondwiththoseofTablel.

Naifoh el al.. RFLP oillllclear DNA ill shrews

and PvurJ (V). After electrophoresis with 0.7% agarose gel for 17 hours at 30 Y, the digested DNA was immo­bilized on nylon membrane (Hybond-N, Amersham) and hybridized with digoxigenin-Iabeled (DIG DNA Label­ing Kit and Detection Kit, Roche Diagnostics) 28S probe

(0.7 ,"g/ml) from mouse (Suzuki et al. 1994). Detection of the probe on the membrane followed the supplier' s protocol. Fragment lengths were determined by com­parison with a size marker (DNA Molecular Weight

SO ON BO ' N

0 1000 km

1:'>0 0

q,~.§'

t1

103

Marker J] digoxigenin-labeled lambda HindlJl). Restric­

tion maps of the 28S downstream spacer region were constructed by referring to the standard restriction map of the 18S, S.8S, and 28S rRNA genes that are conser­vative among mammalian species (Suzuki et al. 1994;

Iwasa et al. 200 I).

Phylogenetic analyses All restriction sites were numbered from I to 13 (Fig.

Baika lake

7

'* ·r

<'f , §

~O ; 0 0'" Cj t~

Pacific Ocean

2

c=J S. caecutiens _ So shinto

Fig. I. Collect ion loca lities o f shrew sampl es. Locality numbers correspond with those of Table I.

Naifoh el al.. RFLP oillllclear DNA ill shrews

and PvurJ (V). After electrophoresis with 0.7% agarose gel for 17 hours at 30 Y, the digested DNA was immo­bilized on nylon membrane (Hybond-N, Amersham) and hybridized with digoxigenin-Iabeled (DIG DNA Label­ing Kit and Detection Kit, Roche Diagnostics) 28S probe

(0.7 ,"g/ml) from mouse (Suzuki et al. 1994). Detection of the probe on the membrane followed the supplier' s protocol. Fragment lengths were determined by com­parison with a size marker (DNA Molecular Weight

SO ON BO ' N

0 1000 km

1:'>0 0

q,~.§'

t1

103

Marker J] digoxigenin-labeled lambda HindlJl). Restric­

tion maps of the 28S downstream spacer region were constructed by referring to the standard restriction map of the 18S, S.8S, and 28S rRNA genes that are conser­vative among mammalian species (Suzuki et al. 1994;

Iwasa et al. 200 I).

Phylogenetic analyses All restriction sites were numbered from I to 13 (Fig.

Baika lake

7

'* ·r

<'f , §

~O ; 0 0'" Cj t~

Pacific Ocean

2

c=J S. caecutiens _ So shinto

Fig. I. Collect ion loca lities o f shrew sampl es. Locality numbers correspond with those of Table I.

104 ル勿〃1〃ialStu4ソ30(2005)

　-spacerregionatdownstreamof28Sgene

k--28S・RNAg・ ・e-一 明Il PHEGDA V

1kb一一

PH EGD AG V E

.畠=

PH D AG V E

PH D AGG V E一V PH ADG E

▲

V PH D DADG V E一V PH DAG V E一1 23 456 789101112 13

restrictionsitenumber

Fig.2.RestrictionmapsofthemajorrepeatingunittypesoftheribosomalRNAgenedownstreamofthe28ScodingregionintheSorex

caecutiens/shintogroupfromvariouslocalitiesofEurasia.Abbreviationsofrestrictionenzymes:A,Aatl;B,Bglll;D,Dral;E,EcoRI;H,Hindlll;

P,Pstl;V,1)vull.Amongl5individualsexamined,somehadidenticalrepetitivetypes(repetypes).SeeTablelf()rindividualsexaminedandtheir

RFLPrepetyes.Opentrianglesindicatepolymorphicsitesassizevariationwithinagenomeofanindividual。

2)andconvertedintoaO/ldatamatrix,whereaOindi-

catesabsenceofthebandandalindicatespresence.

Phylogeneticanalyseswereconductedusingthemaxi-

mumparsimonymethod,appliedusingPAUPver.

4.OblO(Swofford2000)withtheexhaustivesearch

methodandDollo'scriterion.Thecostfbrgainofa

restrictionsitewassixtimeshigherthanfbrlossofasite

becausetherestrictionenzymesusedallhadasixbase

recognitionsequence(SwoffordandOlsenl990).Nodal

supportwasevaluatedbylO,000bootstrapreplicateswith

50%majorityruleconsensus.Aparsimoniousnetwork

wasalsoconstructedaccordingtoBandelt(1994).This

methodindicateseverychangeofstateinanetwork.

Thus,whenthenumberofinformativesitesissmall,this

methodismorestraightfbrwardthanothermethodsof

phylogenyreconstruction.

Results

Restrictionmap

SevenrDNA-RFLPrepetitivetypes(repetypes)were

recognizedamongl5shrewsexamined(Fig.2).The

restrictionpattemsofS.shintoandS.caecutienswere

distinguishedbysitel(Fig.2).FourindividualsofS.

shintoshowedthreerDNA-RFLPrepetitivetype:Shi-A,

Shi-BandShi-C(Tablel,Fig.2).Incontrast,fivein-

dividualsofS.caecutiensfromseverallocalitiesonthe

EurasiancontinentincludingtheKoreaPeninsulashared

onlyonerepetype(Cae-C;Tablel,Fig.2).TwoS.

caecutiensindividualsfromHokkaidohadasingle

repetype(Cae-A;Tablel,Fig.2),andsites4and5were

notobservedinCae-Cwhilesitesllandl3shownin

Cae-CwereabsentinCae-A(Fig.2).Therepetypefrom

Sakhalin(Cae-B)hadsites4,5,ll,andl3(Fig.2),

demonstratingitcontainedcharacteristicsofrepetypes

bothfromthecontinent(Cae-C)andHokkaido(Cae-A).

TwoindividualsfromChej'uIslandhadasinglerepe-

type(Sp-Cheju),whichshowedauniquerestrictionsite

(sitelO)butothersiteswereidenticaltotherepetype

(Cae-C)fromthecontinent(Fig.2).

C1α ∂bgra〃2α 掘 雁wo鴻

Oneshortesttree(treelength=16)wasobtainedby

themaximumparsimonymethod(Fig.3).Sorexsh'η'o

and3.6α θ6〃'1θη8weredistinguishedinthecladogram

(Fig.3)buttheconfidenceofthisdivisionwasstatisti-

callylow(bootstrapvalues,<50%).Therepetypefヒom

Sp-Ch〔 加(Chq.uIsland)wasincludedinthe3.oα θo〃一

伽scladeandformedasubcladewithCae-C(Eurasian

continent).TherepetypesfbrCae-A(Hokkaido)and

Cae-B(Sakhalin)formedasecondsubcladewithinthe

3.caecutiensclade(Fig.3).

Inthephylogeneticnetwork(Fig.4A),3.caecutiens

and3.3痂 η'oweredistinguishedbyrestrictionsitel,and

104 Mammal Study 30 (2005)

spacer region at downstream of 28S gene 1kb

28S rR NA gene • PH E G O A V ---- I 1 I \1 1 1 PH EGO A G V E ---- I I I \1 1 1 I I

PH 0 A G V E ---- I I I 1 !. 1 I

PH 0 AG G V E ____ smHeI§i'_ I I I 11 !. I I

V PH AOG E

~ I I I 1(!. I

V PH 0 OAOG V E

~ I I I I 'if !. 1 I

V PH OA G V E

~ I I I ';j, !. 1 I

2 3 4 5 6 7 8 9 10 11 12 13 restriction site number I I I ~I '-'l l ~ ~ I I

Fig. 2. Restr ict ion maps of the major repeating unit types of the ri bosomal RNA gene dow nstream of the 28S codi ng region in the Sorex caeclItiells/shinfo group From various localit ies of Ellrasia. Abbrevi~Hions o f restriction enzymes: A, Aal l ; B, BgflI: 0 , Dral ; E. EcoRl ; H. Hilldll[ ~

P, Pstl ; V, PVII II . Among 15 ind ividua ls examined. some had identical repetitive types (repetypes) . See Table I for individuals exam ined and thei r RFLP repetyes. Open triang les indicate polymorphic s ites as size var iat ion within a genome of an ind ividual.

2) and converted in to a 0/ 1 data matrix, where a 0 indi­

cates absence of the band and a 1 indicates presence. Phylogenetic ana lyses were conducted us ing the maxi­

mum pars imony method, appl ied using PAUP ver. 4 .0b I 0 (Swofford 2000) with the exhausti ve search

method and Da lla ' s criterion. The cost for ga in of a

restriction site was six times higher than for loss of a site

because the restriction enzymes used a ll had a six base

recogniti on sequence (Swofford and O lsen 1990). Nodal

support was evaluated by 10,000 bootstrap replicates with 50% majority ru le consensus. A parsimoniolls nehvork

was a lso constructed according to Bandelt ( 1994). This

method indicates every change of state in a network. Thus, when the number of into nnative sites is small , th is

method is more stra ightforward than other methods of

phylogeny reconstruction.

R es u lts

Restriction map

Seven rDNA-RFLP repetiti ve types (repetypes) were recognized among IS shrews examined (Fig. 2). The

restriction patte rns of S. shinto and S. caeclIfiens were distinguished by site I (Fig. 2). Four individua ls of S.

shil/to showed three rDNA-RFLP repetitive type: Shi-A,

Shi-B and Shi-C (Table I , Fig. 2). In contrast, fi ve in­dividua ls of S. caecutiens from severa l localities on the

Euras ian continent including the Korea Pen insula shared onl y one repetype (Cae-C; Table I, Fig. 2). Two S.

caecllfiens individuals from Hokkaido had a single repetype (Cae-A; Table I, Fig. 2), and sites 4 and 5 were

not observed in Cae-C whil e sites I I and 13 shown in

Cae-C were absent in Cae-A (Fig. 2). The repetype from

Sakhalin (Cae-B) had sites 4 , 5, I I, and 13 (Fig. 2), demonstrating it conta ined characterist ics of repetypes

both from the continent (Cae-C) and Hokkaido (Cae-A). Two individua ls from C heju Island had a sing le repe­

type (Sp-Chej u), which showed a unique restriction si te

(s ite 10) but other sites were identical to the repetype (Cae-C) from the continent (F ig. 2).

Cladogram and network One shortest tree (tree length = 16) was obta ined by

the maximum parsimony method (Fig. 3). Sorex shinto and S. caeclftiens were distinguished in the c ladogram

(Fig. 3) but the confide nce of thi s d ivision was stat isti­

ca lly low (bootstrap va lues, <50%). The repetype from Sp-C heju (Chej u Island) was included in the S. caec,,­liens clade and formed a sube lade with Cae-C (Euras ian

continent). The repetypes for Cae-A (Hokkaido) and Cae-B (Sakhalin) formed a second subc lade within the

S. caeculiens clade (F ig. 3).

In the phylogenetic network (Fig. 4A), S. caeculiens and S. shinto were d istinguished by restr iction site I , and

104 Mammal Study 30 (2005)

spacer region at downstream of 28S gene 1kb

28S rR NA gene • PH E G O A V ---- I 1 I \1 1 1 PH EGO A G V E ---- I I I \1 1 1 I I

PH 0 A G V E ---- I I I 1 !. 1 I

PH 0 AG G V E ____ smHeI§i'_ I I I 11 !. I I

V PH AOG E

~ I I I 1(!. I

V PH 0 OAOG V E

~ I I I I 'if !. 1 I

V PH OA G V E

~ I I I ';j, !. 1 I

2 3 4 5 6 7 8 9 10 11 12 13 restriction site number I I I ~I '-'l l ~ ~ I I

Fig. 2. Restr ict ion maps of the major repeating unit types of the ri bosomal RNA gene dow nstream of the 28S codi ng region in the Sorex caeclItiells/shinfo group From various localit ies of Ellrasia. Abbrevi~Hions o f restriction enzymes: A, Aal l ; B, BgflI: 0 , Dral ; E. EcoRl ; H. Hilldll[ ~

P, Pstl ; V, PVII II . Among 15 ind ividua ls examined. some had identical repetitive types (repetypes) . See Table I for individuals exam ined and thei r RFLP repetyes. Open triang les indicate polymorphic s ites as size var iat ion within a genome of an ind ividual.

2) and converted in to a 0/ 1 data matrix, where a 0 indi­

cates absence of the band and a 1 indicates presence. Phylogenetic ana lyses were conducted us ing the maxi­

mum pars imony method, appl ied using PAUP ver. 4 .0b I 0 (Swofford 2000) with the exhausti ve search

method and Da lla ' s criterion. The cost for ga in of a

restriction site was six times higher than for loss of a site

because the restriction enzymes used a ll had a six base

recogniti on sequence (Swofford and O lsen 1990). Nodal

support was evaluated by 10,000 bootstrap replicates with 50% majority ru le consensus. A parsimoniolls nehvork

was a lso constructed according to Bandelt ( 1994). This

method indicates every change of state in a network. Thus, when the number of into nnative sites is small , th is

method is more stra ightforward than other methods of

phylogeny reconstruction.

R es u lts

Restriction map

Seven rDNA-RFLP repetiti ve types (repetypes) were recognized among IS shrews examined (Fig. 2). The

restriction patte rns of S. shinto and S. caeclIfiens were distinguished by site I (Fig. 2). Four individua ls of S.

shil/to showed three rDNA-RFLP repetitive type: Shi-A,

Shi-B and Shi-C (Table I , Fig. 2). In contrast, fi ve in­dividua ls of S. caecutiens from severa l localities on the

Euras ian continent including the Korea Pen insula shared onl y one repetype (Cae-C; Table I, Fig. 2). Two S.

caecllfiens individuals from Hokkaido had a single repetype (Cae-A; Table I, Fig. 2), and sites 4 and 5 were

not observed in Cae-C whil e sites I I and 13 shown in

Cae-C were absent in Cae-A (Fig. 2). The repetype from

Sakhalin (Cae-B) had sites 4 , 5, I I, and 13 (Fig. 2), demonstrating it conta ined characterist ics of repetypes

both from the continent (Cae-C) and Hokkaido (Cae-A). Two individua ls from C heju Island had a sing le repe­

type (Sp-Chej u), which showed a unique restriction si te

(s ite 10) but other sites were identical to the repetype (Cae-C) from the continent (F ig. 2).

Cladogram and network One shortest tree (tree length = 16) was obta ined by

the maximum parsimony method (Fig. 3). Sorex shinto and S. caeclftiens were distinguished in the c ladogram

(Fig. 3) but the confide nce of thi s d ivision was stat isti­

ca lly low (bootstrap va lues, <50%). The repetype from Sp-C heju (Chej u Island) was included in the S. caec,,­liens clade and formed a sube lade with Cae-C (Euras ian

continent). The repetypes for Cae-A (Hokkaido) and Cae-B (Sakhalin) formed a second subc lade within the

S. caeculiens clade (F ig. 3).

In the phylogenetic network (Fig. 4A), S. caeculiens and S. shinto were d istinguished by restr iction site I , and

ハlaitohetal.,」RF乙 、PofηuclearDハIAinshrews 105

Sp・Cheju

Fig.3.UnrootedmaximumparsimonytreebasedonnuclearrDNA-

RFLPrepetypesintheSorexcaecutiens/shintogroup.Percentvalues

arethebootstrapvaluesafterlO,000replications.SeeTablelfbr

acronymsoftherDNA-RFLPrepetypes.

theshrewsfromCh(加Island(Sp-Ch(加)weremost

closelyconnectedwithS.eoeeutiensfromthecontinent,

asinthemaximumparsimonytree.Thephylogenetic

networksuggeststhattherepetypeofHokkaido(Cae-A)

wasderivedfromthatofSakhalin(Cae-B),whichwas

inturnderivedfromthatoftheEurasiancontinent

(Cae-C).ThedivergencepattemwithinS.shintowas

unclear(Fig.4A).

Inaddition,itwasobservedthatthemappingpattem

oftherestrictionsitesofSakhalin(Cae-B)wasthesum

ofthecontinental(Cae-C)andHokkaido(Cae-A)types

(Fig.2);i.e.,Cae-BwastheintermingledtypeofCae-C

andCae-A.Thus,allowinghybridizationoftherDNA-

RFLPpattem,analternativenetworkhasbeenrecon-

structed(Fig.4B).

Discussion

PreviousphylogeneticstudiesbasedonmtDNAcyt-b

sequences(Ohdachietal.1997,2001,2003;Fumagalli

etal.1999)andallozymes(Georgel988)demonstrated

A.parsimoniousnetwork

B.networkwithhybridization

S.caecu亡1eηs

11N)x

13/Q

Hybridization

Fig.4.ParsimoniousnetworkbasedonrDNA-RFLPrepetypes(A)andaphylogeneticnetworkallowinghybridization(B),basedonrDNA-

RFLPrepetypesinthe5「orexcaecutiens/shintogroup.Numbersnearshortbarsdenotethoseofthechangeofrestrictionsites。Sitenumbers

correspondwiththoseofFig。2.Anopencircleisahypotheticaltaxonomicunit.IntraspecificrelationshipsofS.shintowereomittedinthe

networks.SeeTablelfbracronymsoftherDNA-RFLPrepetypes.

Naifoh el al.. RFLP oillllclear DNA ill shrews

Cae-B 57% 52% Cae-C

Fig. 3. Unrooted maximum parsimony tree based on nuc lear rDNA ­RFLP repetypes in the Sorex caeclifiells/shillfO group. Perce nt va lues are the bootstrap va lu es after 10,000 repl ications. See Table I for acronyms oflhe rDNA-RFLP repetypes .

A. parsimonious network

S. caecutiens Sp-Cheju

Cae-A Cae-8

11 13

B. network with hybridization

S. caecutiens

Sp-Cheju

105

the shrews fro m Cheju Island (Sp-Cheju) were most closely connected with S. caecutiens from the continent,

as in the maximum parsimony tree. The phylogenetic network suggests that the repetype of Hokkaido (Cae-A) was derived from that of Sakhalin (Cae-S), which was in turn derived from that of the Eurasian continent

(Cae-C). The divergence pattern within S. shinlo was unclear (Fig. 4A).

In addition, it was observed that the mapping pattern of the restriction sites of Sakhal in (Cae-S) was the sum of the continenta l (Cae-C) and Hokkaido (Cae-A) types (Fig. 2); i.e. , Cae-S was the interm ingled type of Cae-C and Cae-A. Thus, allowing hybridi zation of the rDNA­RFLP pattern, an alternative network has been recon­structed (Fig. 4S).

Discussion

Previous phylogenetic studies based on mtDNA cyt-b

sequences (Ohdach i et al. 1997, 200 1, 2003; Fumagalli et al. 1999) and allozymes (George 1988) demonstrated

S. shinto

Cae -8 O ... -------------I!:f+t-+o - S. shinto

~

\ .. ------- Hybridization

Cae-A

Fig. 4. Parsimonious net\vork based on rDNA -RFLP repetypes (A) and a phylogenetic network allow ing hybridization (8 ), based on rDNA ­RFLP repety pes in the Sorex caeclifiens/sh infO group. Numbers near ShOlt bars denote those of the change of restriction si tes. Site num bers correspond wi lh those of Fig. 2. An open c irc le is a hypolhet ical taxonomic unit. Inlraspecific relationships of S. shimo were omitted in the networks. See Table I fo r ac ronyms of the rDNA-RFLP repetypes.

Naifoh el al.. RFLP oillllclear DNA ill shrews

Cae-B 57% 52% Cae-C

Fig. 3. Unrooted maximum parsimony tree based on nuc lear rDNA ­RFLP repetypes in the Sorex caeclifiells/shillfO group. Perce nt va lues are the bootstrap va lu es after 10,000 repl ications. See Table I for acronyms oflhe rDNA-RFLP repetypes .

A. parsimonious network

S. caecutiens Sp-Cheju

Cae-A Cae-8

11 13

B. network with hybridization

S. caecutiens

Sp-Cheju

105

the shrews fro m Cheju Island (Sp-Cheju) were most closely connected with S. caecutiens from the continent,

as in the maximum parsimony tree. The phylogenetic network suggests that the repetype of Hokkaido (Cae-A) was derived from that of Sakhalin (Cae-S), which was in turn derived from that of the Eurasian continent

(Cae-C). The divergence pattern within S. shinlo was unclear (Fig. 4A).

In addition, it was observed that the mapping pattern of the restriction sites of Sakhal in (Cae-S) was the sum of the continenta l (Cae-C) and Hokkaido (Cae-A) types (Fig. 2); i.e. , Cae-S was the interm ingled type of Cae-C and Cae-A. Thus, allowing hybridi zation of the rDNA­RFLP pattern, an alternative network has been recon­structed (Fig. 4S).

Discussion

Previous phylogenetic studies based on mtDNA cyt-b

sequences (Ohdach i et al. 1997, 200 1, 2003; Fumagalli et al. 1999) and allozymes (George 1988) demonstrated

S. shinto

Cae -8 O ... -------------I!:f+t-+o - S. shinto

~

\ .. ------- Hybridization

Cae-A

Fig. 4. Parsimonious net\vork based on rDNA -RFLP repetypes (A) and a phylogenetic network allow ing hybridization (8 ), based on rDNA ­RFLP repety pes in the Sorex caeclifiens/sh infO group. Numbers near ShOlt bars denote those of the change of restriction si tes. Site num bers correspond wi lh those of Fig. 2. An open c irc le is a hypolhet ical taxonomic unit. Inlraspecific relationships of S. shimo were omitted in the networks. See Table I fo r ac ronyms of the rDNA-RFLP repetypes.

106 ル勿〃1〃ialStu4ソ30(2005)

A.hypotheticaIphylogeny(Tree1)

Eurasian

Continent

Sakhalin

S.caeα ノオleηs

B.hypotheticalphylogeny(Tree2)

S.caecutiens

SakhalinAr

o

Eurasian

Continent

Fig.5.TwohypotheticalphylogeniesofnuclearDNAintheSorexcaecutiens/shintogroup,basedontherDNA-RFLPrepetypes,Treel(A)

wasinferredfromtheparsimoniousnetwork(Fig.4A)andtree2(B)fromthenetworkallowinghybridization(Fig。4B)。Arrowswithbrokenlines

indicatehybridization.

thatS.caecutiensandS.shintoshouldbetreatedastwo

separatespecies.Inthepresentstudy,itwasnotstatis-

ticallysupported,althoughtopologyshowedthesetwo

wereseparated(Fig.3).However,theshrewfromCh〔 加

IslandwasmostclosetoS.caecutiensintheEurasian

continentanditshouldberecognizedasS.caecutiens

(Fig.3).Inaddition,althoughsamplinglocationsfrom

theEurasiancontinentweregreatlyscattered,therewas

novariationintherDNA-RFLPrepetypewithinS.cae-

cutiensoverthisrange(Tablel,Fig.2).Thesameten-

dencywasobservedinthegeneticvariationofmtDNA

cyt一わ(Ohdachietal.2001,2003;Ohdachi2005).These

findingsfピombothnuclearDNAandmtDNAsuggest

thatS.caecutienshasgonethrougharecentandrapid

rangeexpansionintheEurasiancontinent.Pattemsof

geneticvariationwerelargelydifferentbetweenHokkaido

andthecontinentbothinrDNA-RFLPrepetypes(Fig.4)

andmtDNAhaplotypes(Ohdachietal.2001,2003;

Ohdachi2005).Therefbre,therangeexpansionofS.

caecutiensinthecontinentmusthaveoccurredafter

Hokkaidoandthecontinentweregeologicallyseparated

afterthelastglacialperiod.ItisestimatedthatHokkaido

andSakhalin+Eurasiancontinentwereseparatedca.

12,000yearsago(Ohshimal990).Thus,thegreat

rangeexpansioninthecontinentpresumablybeganafter

Variouslocatities

intheEurasian

Continentand

Sakhalin

Fig.6.SchematictreeofthemitochondrialDNAphylogenyinthe

Sorexcaecutiens/shintogroup,basedonmtDNAcytochromebgenesequences(Ohdachietal.2003;Ohdachiinpress),Anopencircleisa

hypotheticaltaxonomicunit.

12,000yearsago.

Themaximumparsimonytree(Fig.3)andthephylo-

geneticnetwork(Fig.4A)wereusedtoconstructahypotheticalphylogenyofthenuclearrDNA(Treel,

Fig.5A).Ontheotherhand,allowinghybridization

betweentherDNA-RFLPs,analtemativenetworkwas

reconstructed(Fig.4B),andfromthisphylogeneticnet-

workwithhybridization,anotherhypotheticalphylogeny

wasinferred(Tree2,Fig.5B).

Wecannotconcludewhichphylogeny(Fig.5:treesl

and2)ismoreappropriateasanevolutionaryprocessof

106 Mammal Study 30 (2005)

A. hypothetical phylogeny (Tree 1)

S. shinto Cheju

Hokkaido Shikoku

Sakhalin Honshu

S. caecutiens

B. hypothetical phylogeny (Tree 2)

S. caecutiens S. shinto Cheju

Shikoku

Sakhalin ...... ••••• Honshu

o .... ~~.

Hokkaido

Fig. 5. Two hypothetical phylogenies of nuclear DNA in the Sorel' caeculiells/sitilllo group, based on the rDNA- RFLP rcpetypcs. Tree I (A) was inferred from the parsimonious network (Fig. 4A ) and tree 2 (B) from the network allowing hybridization (Fig. 4B). Arrows \vith broken lines indicate hybridi zation .

that S. caecutiens and S. shinto should be treated as t\vo separate species. In the present study, it was not stat is­

tically supported, although topology showed these two were separated (Fig. 3). However, the shrew trom Cheju

Island was most close to S. caeculiens in the Eurasian

cont inent and it should be recognized as S. caecliliens

(Fig. 3). In addi tion, a lthough sampling locations from

the Eurasian continent were great ly scattered, there was

no variation in the rDNA-RFLP repetype within S. cae­

cliliens over this range (Table I , Fig. 2). The same ten­dency was observed in the genetic variation of IlltDNA

cyt-b (Ohdachi et a l. 2001, 2003 ; Ohdachi 2005). These findings from both nuclear DNA and mtDNA suggest

that S. caecliliens has gone through a recent and rapid range expansion in the Eurasian cont inent. Patterns of

genetic variation were largely different between Hokkaido

and the continent both in rDNA-R FLP repetypes (Fig. 4)

and mtDNA haplotypes (Ohdachi et al. 200 1, 2003;

Ohdachi 2005). Therefore, the range expansion of S.

caeclfliens in the continent Illust have occurred after

Hokkaido and the continent were geologica lly separated

after the last glac ial period. It is estimated that Hokkaido and Sakha lin+ Eurasian continent were separated ca.

12 ,000 years ago (Ohshima 1990). Thus, the great

range expansion in the continent presumably began after

S. caecutiens

Various local ities in the Eurasian Continent and Sakhalin

S. shinto

Fig. 6. Schematic tree of the mitochondrial DNA phylogeny in the So/'ex caecutiens/shillto group, based on mtDNA cytochrom e b gene sequences (Ohdachi et al. 2003; Ohdachi in press). An open ci rcle is a hypothetical taxonomic unit.

12,000 years ago.

The maximum parsimony tree (Fig. 3) and the phylo­genetic network (Fig. 4A) were used to construct a

hypothetical phylogeny of the nuclear rDNA (Tree I, Fig. SA). On the other hand, a llowing hybridization

between the rDNA-RFLPs, an a lternative network was

reconstructed (Fig. 4B), and from this phylogenetic net­

work with hybridi zation, another hypothetical phylogeny

was inferred (Tree 2, Fig. 5B). We can not conclude which phylogeny (F ig. 5: trees I

and 2) is more appropriate as an evolutionary process of

106 Mammal Study 30 (2005)

A. hypothetical phylogeny (Tree 1)

S. shinto Cheju

Hokkaido Shikoku

Sakhalin Honshu

S. caecutiens

B. hypothetical phylogeny (Tree 2)

S. caecutiens S. shinto Cheju

Shikoku

Sakhalin ...... ••••• Honshu

o .... ~~.

Hokkaido

Fig. 5. Two hypothetical phylogenies of nuclear DNA in the Sorel' caeculiells/sitilllo group, based on the rDNA- RFLP rcpetypcs. Tree I (A) was inferred from the parsimonious network (Fig. 4A ) and tree 2 (B) from the network allowing hybridization (Fig. 4B). Arrows \vith broken lines indicate hybridi zation .

that S. caecutiens and S. shinto should be treated as t\vo separate species. In the present study, it was not stat is­

tically supported, although topology showed these two were separated (Fig. 3). However, the shrew trom Cheju

Island was most close to S. caeculiens in the Eurasian

cont inent and it should be recognized as S. caecliliens

(Fig. 3). In addi tion, a lthough sampling locations from

the Eurasian continent were great ly scattered, there was

no variation in the rDNA-RFLP repetype within S. cae­

cliliens over this range (Table I , Fig. 2). The same ten­dency was observed in the genetic variation of IlltDNA

cyt-b (Ohdachi et a l. 2001, 2003 ; Ohdachi 2005). These findings from both nuclear DNA and mtDNA suggest

that S. caecliliens has gone through a recent and rapid range expansion in the Eurasian cont inent. Patterns of

genetic variation were largely different between Hokkaido

and the continent both in rDNA-R FLP repetypes (Fig. 4)

and mtDNA haplotypes (Ohdachi et al. 200 1, 2003;

Ohdachi 2005). Therefore, the range expansion of S.

caeclfliens in the continent Illust have occurred after

Hokkaido and the continent were geologica lly separated

after the last glac ial period. It is estimated that Hokkaido and Sakha lin+ Eurasian continent were separated ca.

12 ,000 years ago (Ohshima 1990). Thus, the great

range expansion in the continent presumably began after

S. caecutiens

Various local ities in the Eurasian Continent and Sakhalin

S. shinto

Fig. 6. Schematic tree of the mitochondrial DNA phylogeny in the So/'ex caecutiens/shillto group, based on mtDNA cytochrom e b gene sequences (Ohdachi et al. 2003; Ohdachi in press). An open ci rcle is a hypothetical taxonomic unit.

12,000 years ago.

The maximum parsimony tree (Fig. 3) and the phylo­genetic network (Fig. 4A) were used to construct a

hypothetical phylogeny of the nuclear rDNA (Tree I, Fig. SA). On the other hand, a llowing hybridization

between the rDNA-RFLPs, an a lternative network was

reconstructed (Fig. 4B), and from this phylogenetic net­

work with hybridi zation, another hypothetical phylogeny

was inferred (Tree 2, Fig. 5B). We can not conclude which phylogeny (F ig. 5: trees I

and 2) is more appropriate as an evolutionary process of

Naitohetal.,RF乙PqプnuclearD八 「Ainshrei・vs

thenuclearrDNA.AmtDNAphylogenyoftheS.

eaecutiens/shintogroup,wasinferred(Fig.6)basedon

Ohdachietal.(2003)andOhdachi(2005),whereS.

eoeeutienswasdividedintoHokkaidoandContinent-

Sakhalin-Chej●ugenealogiesafterS.(3aecutiensandS.

shintodiverged.Hence,tree2ismorecongruentwith

themtDNAphylogenythantreelandfbrthisreason

seemsthemoreappropriatephylogenetictree(network)

ofthenuclearrDNA.

Acknowledgments:WewishtothankH.Abe,N.E.

Dokuchaev,V.Haukisalmi,M.Yamagishiforoffering

shrewsamples,G.Hintenforreviewingthemanuscript,

andK.Yoshizawafbrsupportinganalysisofthemaxi-

mumparsimonymethod.Wealsowouldliketoexpress

ourgratitudetoanonymousreviewersfbrcriticalcom-

ments.

References

Arheim,N.,Krystal,M,,Schmickel,R,,Wilson,G,,Ryder,0.and

Zimmer,E.1980.Molecularevidencefbrgeneticexchanges

amongribosomalgenesonnonhomologouschromosomesinman

andapes.ProceedingsoftheNationalAcademyofSciencesof

theUnitedStatesofAmerica73:7323-7327.

Bandelt,H.J.1994.Phylogeneticnetworks.Verhandlungendes

NaturwissenschaftlichenVereinsinHamburgN.F.34:51-71.

Dokuchaev,N,E,Ohdachi,S,andAbe,H,1999.Mo叩hometric

statusofshrewsoftheSorexcaecutiens/shintogroupinJapan.

MammalStudy24:67-78.

Fumagalli,L,Taberlet,P。,Stewart,D.T.,Gielly,LHausser,J.and

Vogel,P.1999.MolecularphylogenyandevolutionofSorex

shrews(Soricidae:Insectivora)inferredfrommitochondrialDNA

sequencedata,MolecularPhylogeneticsandEvolutionll:222-

235.

George,S。1988.Systematics,historicalbiogeography,andevolution

ofthegenusSorex。JournalofMammalogy69:443-461。

Iwasa,M。A。,Ohdachi,S.,Han,S.-H.,Oh,H.-S.,Abe,H。andSuzuki,

H.2001.KaryotypeandRFLPofthenuclearrDNAofthe

Croeidurasp,onChejuIsland,SouthKorea.Mammalia65:

451-459,

Iwasa,M.A.andSuzuki,H.2002.Evolutionarynetworksofmaternal

andpaternalgenelineagesinEothenomysvolesendemictoJapan.

JoumalofMammalogy83:852-865.

107

Mukai,T.2001.Hybridizationandintrogressioninthespeciation

processoffishes.JapaneseJoumalofIchthyology48:1-18(in

JapanesewithEnglishsummary).

Ohdachi,S.2005.Historyofcommunityorganizationofshrewsin

Hokkaido,inferredfromDNA.In(R.MasudaandH.Abe,eds。)

NaturalHistoryofAnimalBiogeography.HokkaidoUniversity

P「ess,SaPporo(inJapanese)(inpress).

Ohdachi,S.,Masuda,R.,Abe,H.,Adachi,J.,Dokuchaev,N.E.,

Haukisalmi,V,andYoshida,M,C,1997,PhylogenyofEurasian

soricineshrews(lnsectivora,Mammalia)inferredfromthemito-

chondrialcytochromebgenesequences.ZoologicalSciencel4:

527-532。

Ohdachi,S.,Dokuchaev,NE.,Hasegawa,M.andMasuda,R.2001.

Intraspecificphylogenyandgeographicvariationofsixspeciesof

northeasternAsiaticSorexshrewsbasedonthemitochondrial

cytochromebsequences.MolecularEcologylO:2199-2213.

Ohdachi,S。D.,Abe,H.andHan,S.-H。2003.Phylogeneticalposi-

tionsofSorexsp。(lnsectivora,Mammalia)fromCh(加Islandand

S.caecutiensfromtheKoreanPeninsula,inferredfrommitochon-

drialcytochromebgenesequences.ZoologicalScience20:91-

95.

Ohshima,K.1990.ThehistoryofstraitsaroundtheJapaneseIs-

landsinthelate-Quaternary.DaiyonkiKenkyu[TheQuatemary

Research]29:193-208(inJapanesewithEnglishabstract).

SambrookJ.,Fritsch,E.F.andManiatis,T.1989.MolecularCloning:

ALaboratoryManual,2nded.ColdSpringHarborLaboratory,

NewYork.

Sota,T.andVogler,A.P.2001.Incongruenceofmitochondrialand

nucleargenetreesinthecarabidbeetlesOhomoρterus.System-

aticBiology50:39-59。

Suzuki,H。1994.GeneticdiversityofribosomalDNA:phylogenetic

analysisofsmallmammals.HonyuruiKagaku[Mammalian

Science]34:67-79(inJapanesewithEnglishabstract),

Suzuki,H.,Tsuchiya,K.,Sakaizumi,M.,Wakana,S.,Gotoh,0.,

Saitou,N,Moriwaki,K.andSakurai,S.1990.Differentiation

ofrestrictionsitesinribosomalDNAinthegenusApodemus.

BiochemicalGenetics28:137-149.

Suzuki,H.,Kawamoto,Y.,Takenaka,0.,Munechika,1.,Hori,H。and

Sakurai,S.1994,PhylogeneticrelationshipsamongHomo

sapiensandrelatedspeciesbasedonrestrictionsitevariationin

rDNAspacers.BiochemicalGenetics32:257-269.

Swofford,D.L2000.PAUP*:phylogeneticanalysisusingparsi-

mony(*andothermethods).Version4.Sinauer,Sunderland,

Massachusetts.

Swofford,D,L.andOlsen,G,Ll990.Phylogenyreconstruction.In

(D.M.HillisandC.Moritz,eds.)MolecularSystematicsPp.

411-501.Sinauer,Sunderland,Massachusetts.

Received18January2005.Accepted4June2005.

Nai foh el al.. RFLP oillllclear DNA ill shrews

the nuclear rDNA. A mtDNA phylogeny of the S.

eaeel/liens/shinlo group, was inferred (Fig. 6) based on Ohdachi et a l. (2003) and Ohdachi (2005), where S.

caeclffiens was divided in to Hokkaido and Continent­Sakhalin-Cheju genealogies after S. eaeel/liens and S.

shinto diverged. Hence, tree 2 is more congruent with the mtDNA phylogeny than tree I and for this reason seems the more appropriate phylogenetic tree (network) of the nuclear rDNA.

Acknowledgments : We wish to thank H. Abe, N. E. Dokuchaev, V. Haukisalmi , M. Yamagi shi for offering shrew samples, G. Hinten for reviewing the manuscript, and K. Yoshizawa for supporting ana lysis of the maxi­mum parsimony method. We also would like to express our gratitude to anonymous reviewers for critical com­

ments.

References

Arhcim, N., KrystaL M., Schmickcl , R .. Wil son, G., Ryder, O. and Zi mmer, E. 1980. Molecular ev idence for genetic exchanges among ribosomal genes on nonhomo logous chromosomes in man and apes. Proceedin gs of the National Academy of Sc iences of the United States o f America 73: 7323-7327.

Bandelt, H. J. 1994. Phy logenetic networks. Verhand lungen des Naturwissensehaftl ichen Vereins in Hamburg N. F. 34: 5 1- 7 1.

Dokudmcv, N. E., Ohdachi, S. amI Abe, H. 1999. Morphometric status of shrews of the Sorex caecllf;enslsiJinto group in Japan . Mammal SHldy 24: 67- 78.

FUl11agalli, L.. Taberlet. P .. Stewart. D. T .. Gidly. L. Hausser. J. and Voge l, P. 1999. Molecular phylogeny and evol ution of Sorex shrews (Soricidae: Insectivora) inferred from mitochondrial DNA sequence data . Molecular Phylogenetics and Evol ut ion II : 222-235 .

George, S. 1988. Systematics, hi storica l biogeography, and evo lution of tile genus Sorex. Journal of Mammalogy 69: 443-461.

Iwasa , M. A ., Ohdachi , S., Han, S.-H. , Oh , H.-S., Abe, H. and Suzu ki, H. 200 1. Karyotype and RFLP of the nu clear rONA of the Crocidum sp. on Cheju Is land. South Korea. Mammalia 65: 45 1-459.

Iwasa, M. A. and Suzuki , H. 2002. Evo lutionary networks of maternal and paternal gene lineages in Eorhenomys voles endemic to Japan. Journa l of Mamma logy 83: 852- 865.

107

Mukll i. T. 2001. Hybrid ization lind in trogression in the speciation process of fi shes . Japanese Journal of Ic hthyology 48 : \- 18 ( in Japanese with Eng lish summary).

Ohdachi , S. 2005 . Hi story of co mmunity organ ization o f shrews in Hokkaido, inferred from DNA. In (R. Masuda and H. Abe , eds.) Natural History of Anim al Biogeography. Hokkaido Uni versity Press, Sapporo (in JlIpanese) (in press).

Ohdachi, S., Masuda, R., Abe, 1-1 ., Adachi, J., Dokuchaev, N. E., Hauki sa lmi, V. and Yos hida, M. C. 1997. Phylogeny of Eurasian soric ine shrews ( Insecti vora, Mam malia) in fe rred from the mito­chondrial cytochrome b gene sequences. Zoo log ical Sc ience 14: 527-532.

Ohdachi, S .. Dokuchaev, N . E. , Hasegawa, M. and Masuda, R. 2001. Intraspecific phy loge ny and geographi c variation of s ix spec ies o f northeastern Asiatic Sorex shrews based on the mitochondrial cytoc hrome h sequences . Molecular Ecology 10: 2 199-22 13.

Ohdachi. S. D., Abe, H. and Han, SAl. 2003. Phylogenetical posi­tions of Sorex sp. (Insecti vora, Mammalia) from Chej u Is land and S. caeclIl;ellS from Ihe Korean Peni nsula. in ferred fro m mitochon­dria l cytoc hrome b gene seq uences. Zoo logical Sc ience 20: 9 1-95.

Ohsilima, K. 1990. The history of st rait s aro und th e Japanese Is­lands in the late·Quatemary. Daiyonki Kenkyu lThe Quaternary Research1 29: 193-208 (in Japanese with Eng lish abstract).

Sambrook J ., Fritsc h, E. F. and Man iati s, T. 1989. Molecul ar Clonin g: A Laboratory Manua l, 2nd ed . Co ld Spring Harbor Laboratory, New York.

Sota, T. and Vog ler, A. P. 2001. In congruence of mitochondri al and nuclear gene trees in th e carabid beetles Oholl/opferus. System­at ic Bio logy 50: 39- 59,

Suzuki, H. 1994. Genetic diversity of ribosomal DNA: phylogenetic ana lysis of small mammals. Honyurui Kagaku [Mammali an Sc ience] 34: 67~79 (in Japanese with English abstract).

Suzuki, H., Tsuchiya, K., Sakai zumi , M., Wakana, S., Gotoh, 0., Saitou, N., Mori waki, K. and Sakurai , S. 1990. Differentiation of restri ction sites in r ibosoma l DNA in the gen us Apodellllls. Biochemica l Genetics 28: 137- 149.

Suzuki , H., Kawamoto, Y., Takenaka, 0 ., Mun echi ka, I., Hori, H. and Saku rai , S, 1994. Phyloge net ic relationships among Homo sapiens and reialeu spec ies based un rest ri ction site variation in rONA spacers. Biochemical Genetics 32: 257-269.

Swofford, D. L. 2000. PAUP*: phy logenetic analysis using parsi­mon y (* and olher met hods). Version 4, S inauer. S underl and. Massacllusetts.

Swofford, D. L. and Olsen , G. L. 1990. Phylogeny reconstruction. In (D. M. Hill is and C. Mori tz. eds.) Molecular Systematics Pp. 4 1\- 50 I. S inauer, Sunderland, Massachusetts .

Received 18 Jallt/lllY 2005. Accepted 4 JUlie 2005.

Nai foh el al.. RFLP oillllclear DNA ill shrews

the nuclear rDNA. A mtDNA phylogeny of the S.

eaeel/liens/shinlo group, was inferred (Fig. 6) based on Ohdachi et a l. (2003) and Ohdachi (2005), where S.

caeclffiens was divided in to Hokkaido and Continent­Sakhalin-Cheju genealogies after S. eaeel/liens and S.

shinto diverged. Hence, tree 2 is more congruent with the mtDNA phylogeny than tree I and for this reason seems the more appropriate phylogenetic tree (network) of the nuclear rDNA.

Acknowledgments : We wish to thank H. Abe, N. E. Dokuchaev, V. Haukisalmi , M. Yamagi shi for offering shrew samples, G. Hinten for reviewing the manuscript, and K. Yoshizawa for supporting ana lysis of the maxi­mum parsimony method. We also would like to express our gratitude to anonymous reviewers for critical com­

ments.

References

Arhcim, N., KrystaL M., Schmickcl , R .. Wil son, G., Ryder, O. and Zi mmer, E. 1980. Molecular ev idence for genetic exchanges among ribosomal genes on nonhomo logous chromosomes in man and apes. Proceedin gs of the National Academy of Sc iences of the United States o f America 73: 7323-7327.

Bandelt, H. J. 1994. Phy logenetic networks. Verhand lungen des Naturwissensehaftl ichen Vereins in Hamburg N. F. 34: 5 1- 7 1.

Dokudmcv, N. E., Ohdachi, S. amI Abe, H. 1999. Morphometric status of shrews of the Sorex caecllf;enslsiJinto group in Japan . Mammal SHldy 24: 67- 78.

FUl11agalli, L.. Taberlet. P .. Stewart. D. T .. Gidly. L. Hausser. J. and Voge l, P. 1999. Molecular phylogeny and evol ution of Sorex shrews (Soricidae: Insectivora) inferred from mitochondrial DNA sequence data . Molecular Phylogenetics and Evol ut ion II : 222-235 .

George, S. 1988. Systematics, hi storica l biogeography, and evo lution of tile genus Sorex. Journal of Mammalogy 69: 443-461.

Iwasa , M. A ., Ohdachi , S., Han, S.-H. , Oh , H.-S., Abe, H. and Suzu ki, H. 200 1. Karyotype and RFLP of the nu clear rONA of the Crocidum sp. on Cheju Is land. South Korea. Mammalia 65: 45 1-459.

Iwasa, M. A. and Suzuki , H. 2002. Evo lutionary networks of maternal and paternal gene lineages in Eorhenomys voles endemic to Japan. Journa l of Mamma logy 83: 852- 865.

107

Mukll i. T. 2001. Hybrid ization lind in trogression in the speciation process of fi shes . Japanese Journal of Ic hthyology 48 : \- 18 ( in Japanese with Eng lish summary).

Ohdachi , S. 2005 . Hi story of co mmunity organ ization o f shrews in Hokkaido, inferred from DNA. In (R. Masuda and H. Abe , eds.) Natural History of Anim al Biogeography. Hokkaido Uni versity Press, Sapporo (in JlIpanese) (in press).

Ohdachi, S., Masuda, R., Abe, 1-1 ., Adachi, J., Dokuchaev, N. E., Hauki sa lmi, V. and Yos hida, M. C. 1997. Phylogeny of Eurasian soric ine shrews ( Insecti vora, Mam malia) in fe rred from the mito­chondrial cytochrome b gene sequences. Zoo log ical Sc ience 14: 527-532.

Ohdachi, S .. Dokuchaev, N . E. , Hasegawa, M. and Masuda, R. 2001. Intraspecific phy loge ny and geographi c variation of s ix spec ies o f northeastern Asiatic Sorex shrews based on the mitochondrial cytoc hrome h sequences . Molecular Ecology 10: 2 199-22 13.

Ohdachi. S. D., Abe, H. and Han, SAl. 2003. Phylogenetical posi­tions of Sorex sp. (Insecti vora, Mammalia) from Chej u Is land and S. caeclIl;ellS from Ihe Korean Peni nsula. in ferred fro m mitochon­dria l cytoc hrome b gene seq uences. Zoo logical Sc ience 20: 9 1-95.

Ohsilima, K. 1990. The history of st rait s aro und th e Japanese Is­lands in the late·Quatemary. Daiyonki Kenkyu lThe Quaternary Research1 29: 193-208 (in Japanese with Eng lish abstract).

Sambrook J ., Fritsc h, E. F. and Man iati s, T. 1989. Molecul ar Clonin g: A Laboratory Manua l, 2nd ed . Co ld Spring Harbor Laboratory, New York.

Sota, T. and Vog ler, A. P. 2001. In congruence of mitochondri al and nuclear gene trees in th e carabid beetles Oholl/opferus. System­at ic Bio logy 50: 39- 59,

Suzuki, H. 1994. Genetic diversity of ribosomal DNA: phylogenetic ana lysis of small mammals. Honyurui Kagaku [Mammali an Sc ience] 34: 67~79 (in Japanese with English abstract).

Suzuki, H., Tsuchiya, K., Sakai zumi , M., Wakana, S., Gotoh, 0., Saitou, N., Mori waki, K. and Sakurai , S. 1990. Differentiation of restri ction sites in r ibosoma l DNA in the gen us Apodellllls. Biochemica l Genetics 28: 137- 149.

Suzuki , H., Kawamoto, Y., Takenaka, 0 ., Mun echi ka, I., Hori, H. and Saku rai , S, 1994. Phyloge net ic relationships among Homo sapiens and reialeu spec ies based un rest ri ction site variation in rONA spacers. Biochemical Genetics 32: 257-269.

Swofford, D. L. 2000. PAUP*: phy logenetic analysis using parsi­mon y (* and olher met hods). Version 4, S inauer. S underl and. Massacllusetts.

Swofford, D. L. and Olsen , G. L. 1990. Phylogeny reconstruction. In (D. M. Hill is and C. Mori tz. eds.) Molecular Systematics Pp. 4 1\- 50 I. S inauer, Sunderland, Massachusetts .

Received 18 Jallt/lllY 2005. Accepted 4 JUlie 2005.