Arch. Pol. Fish. (2011) 19: 161-166 DOI 10.2478/v10086-011-0020-9 RESEARCH ARTICLE

On the variation and distribution of the lake , Eupallasella percnurus (Pall.)

Received – 15 July 2011/Accepted – 18 August 2011. Published online: 30 September 2011; ©Inland Fisheries Institute in Olsztyn,

Jan Kusznierz, £ukasz Paœko, Daniyar Tagayev

Abstract. The , Eupallasella percnurus (Pall.), extends from the Oder River system in the west to the has a vast distribution range which extends from Poland to the Anadyr River system in the Chukchi Peninsula, Chukchi Peninsula, Sakhalin, Hokkaido, and Korea. This area Sakhalin, Hokkaido, Korea, northern China, Mongo- and its adjacent regions are inhabited by an array of other lia, and Kazakhstan. Throughout this area it inhabits minnow species which are morphologically close to the lake minnow and are currently included in the genera Phoxinus almost exclusively small, usually isolated, distrophic and . However, their systematic relationships water bodies. It is very rarely found in rivers. remain largely obscure. The vast distribution range of the The systematic position of the lake minnow re- species, including numerous river systems divided by many mains very controversial. According to the most mountain ranges that provide it with a very diverse and rich up-to-date morphological analysis (Howes 1985), it hydrographic and glaciological history, favors the evolution of locally differentiated forms. As a result, five subspecies have is a member of the genus Eupallasella, to which it been distinguished within the lake minnow: E. p. percnurus; was transferred from the genus Phoxinus. The hy- E. p. stagnalis; E. p. ignatowi; E. p. mantschuricus; pothesis was supported by the results of allozyme E. p. sachalinensis. Because of great morphological variation, studies by Ito et al. (2002). Based on a limited-range their status is still the subject of debate. The most recent genetic study Sakai et al. (2006) placed it in the ge- morphological studies suggest a separate specific status of the nus Rhynchocypris. Volga minnow E. p. stagnalis. The extremely large distribution range and the Keywords: lake minnow, variation, distribution, fact that the fish inhabits small and most often iso- morphology, systematics lated water bodies favor great morphological varia- tion, which has been emphasized by all the authors who have dealt with the infraspecific systematics of Introduction the lake minnow (Berg 1949, Thienemann 1950, Kaj 1954, G¹sowska and Rembiszewski 1967, Howes The lake minnow, Eupallasella percnurus (Pall.), is 1985, Mitrofanov 1987). In the past, this variation a Eurasian representative of cyprinid fishes led to establishing numerous subspecies which were () of a vast distribution range which then subject to a few limited systematic revisions (Kaj 1954, G¹sowska and Rembiszewski 1967). Based on the literature (Berg 1949, Kaj 1954, G¹sowska and + J. Kusznierz [ ], £. Paœko, D. Tagayev Rembiszewski 1967, Mitrofanov 1987, Nikitin and Zoological Institute, University of Wroc³aw Sienkiewicza 21, 50-335 Wroc³aw, Poland Safronov 2009), the following subspecies are distin- e-mail: [email protected] guished at present: E. percnurus percnurus inhabit- ing most of and the territories of Poland, 162 Jan Kusznierz et al.

2 5

1 a 4 b 3

e c d

Ranges of subspecies: a–E. percnurus percnurus b–E. percnurus stagnalis Origin of samples: c–E. percnurus ignatowi d–E. percnurus mantschuricus 1 – Volga riv. basin, 2 – Pechora riv. basin, 3 – Ob riv. basin, e–E percnurus sachalinensis 4 – Lena riv. basin, 5 – Sea of Okhotsk coast Figure 1. Distribution range of Eupallasella percnurus (Pall.) subspecies and sampling site locations.

Table 1 Sampling site location and number of specimens of Eupallasella percnurus (Pall.)

Subspecies River system Locality N

Eupallasella percnurus percnurus Pechora Ust-Tsilma 15 Ob Sarafanovka village 14 Lena Yakutsk 25 Sea of Okhotsk (Tauy R.) Talon village 15 Eupallasella percnurus stagnalis Volga Nizhniy Novgorod 31 western , and northwestern Russia; and Rembiszewski 1967) represent the nominate E. percnurus stagnalis which occurs in the Volga, subspecies E. percnurus percnurus, most of whose Kama, and Oka river systems; E. percnurus ignatowi range is located east of the Urals. Considering the from undrained areas of northern Kazakhstan; role of these mountains as a barrier, it seems more E. percnurus mantschuricus which is found in the likely that Europe was populated by the minnow Amur River system, Sakhalin, and Korea; originating from the Volga River system. E. percnurus sachalinensis distributed in Sakhalin The aim of this paper was to analyze morphologi- and Hokkaido. The relationships among these sub- cal variation of the lake minnow using geometric species remain obscure. morphometry based on the distribution of the sub- According to earlier studies, the European popu- species, and to attempt to specify the origin of the Eu- lations of the lake minnow (Berg 1949, G¹sowska ropean populations of the species. On the variation and distribution of the lake minnow, Eupallasella percnurus (Pall.) 163

Material and methods last (interior) ray of the pectoral fin; V – posterior margin of the supraoccipital bone; VI – base of the first ray of the dorsal fin; VII – base of the first ray of The material included 100 specimens of the lake the ventral fin; VIII – base of the first ray of the anal minnow representing 5 populations from the fin; IX – base of the last ray of the dorsal fin; X – base Pechora, Volga, Ob, and Lena river systems and the of the last ray of the anal fin; XI – origin of the dorsal Sea of Okhotsk coast (Fig. 1, Table 1). The fish were lobe of the caudal fin (anteriormost point of the first caught with nets and preserved in 4% formaldehyde ray base); XII – origin of the ventral lobe of the caudal in horizontally positioned plastic containers, thus the fin (anteriormost point of the last ray base). specimens were not deformed and mostly well Distances between the landmarks were mea- stretched. In the laboratory, the specimens were sured with the SigmaScan Pro 5.0 program (SPSS rinsed in running water for two hours and then pre- Inc.) pared to be photographed. Each photographed fish was placed on its right side and pinned to a cork board using thick entomological needles in such Statistical analyses a way that the sagittal plane of the specimen was par- allel to the board surface. Thinner entomological Statistical calculations were done with Microsoft Ex- needles were used to mark the landmarks. This facil- cel 2003 (Microsoft Corporation), NTSYS 1.8 for itated the localization of the landmarks in the photo- DOS (Rohlf), NTSYS 2.21l for Windows (Rohlf), and graphs. A millimeter scale was placed at the level of Statistica 9.1 (StatSoft Inc.). The data were the sagittal plane of each specimen for the purpose of log-transformed, and the effect of varied body size of later calibration of the measurements taken from the the specimens was removed by Allometric Burnaby’s photographs. Thus prepared, the specimens were Method (Bookstein 1991, Reyment 1991). In order to photographed with a Canon EOS D350 digital cam- detect patterns of inter-population variation, the era. Only non-deformed straight specimens were number of variables (characters) was subject to re- used for the photographs and, consequently, mea- duction in data matrices freed from the effect of the surements. Twelve landmarks were designated on first eigenvector using Canonical Variate Analysis each specimen according to the Truss Network (CVA) since the data met the assumption of normal- scheme (Strauss and Bookstein 1982) as modified by ity of distribution of characters in the populations Paœko and Maœlak (1997) (Fig. 2). The landmarks and of homogeneity of their variance which were were: I – anteriormost point of the premaxillary bone; tested with Shapiro-Wilk’s test of normality of the II – posterior margin of the posterior naris; III – distribution of characters and with Levene’s test of posteriormost point of the mandible; IV – base of the the homogeneity of variances.

Figure 2. Landmark location and the Truss Network scheme used for morphometric measurements. 164 Jan Kusznierz et al.

6 E. p. stagnalis (Volga riv. basin) E. p. percnurus (Sea of Okhotsk coast) 5 E. p. percnurus (Lena riv. basin) E. p. percnurus (Ob riv. basin) E. p. percnurus (Pechora riv. basin) 4

3

2

1

0

-1

Canonical vector 2 -2

-3

-4

-5 -8 -6 -4 -2 0 2 4 6 8 Canonical vector 1 Figure 3. Results of the Canonical Variate Analysis (CVA) of the Truss Network morphometric features of Eupallasella percnurus (Pall.).

Results Discussion

Wilks’ Lambda for the general model obtained in the The results justify the separation of E. percnurus analysis was 0.00 at P £ 0.05, indicating the very stagnalis as a distinct taxon which should most prob- strong, statistically significant discrimination power ably have species status. The population of this form of the model. The correctness of the classification differs decidedly from the populations of nominate (based on the generalized distance for particular form, which is slightly surprising, considering the specimens from cluster centroids) for the data matrix spatio-temporally extensive palaeohydrographic used to calculate the model was very high and ranged contact of the present Volga system with the areas of from 72 to 100% for individual populations. On aver- Siberia and northeastern Europe inhabited by age, nearly 92% of each population was correctly E. percnurus percnurus. classified within the model. The first three canonical According to most authors, and especially Berg vectors were significant at P £ 0.05, but already the (1949), most of the range of the lake minnow (except first two explained 78% of the total variation. the Volga and Amur river systems, Sakhalin, The discrimination localization of the studied Hokkaido and an undrained region of northern populations within the space of the first two canoni- Kazakhstan) is inhabited by lake minnow popula- cal vectors is presented in Figure 3. The population of tions representing the nominate subspecies. In the E. percnurus stagnalis from the Volga River system is case of Siberia, the situation can be accounted for by distinctly separated along vector I which explains its glaciological past which is much less complex as 63% of the total variation observed. The remaining compared to that of Europe. However, the fact that populations form essentially one cluster, although the subspecies also inhabits Europe at present is sur- the population of E. percnurus percnurus from the prising. According to Kaj (1954), the lake minnow is Pechora River system occupies an intermediate posi- a post-glacial immigrant from the northeast, or a gla- tion between the Volga population and the remaining cial relic of the Central European mixed fauna. The populations which originate from the areas east of results of previous genetic studies (Kusznierz et al. the Urals. Vector II does not discriminate distinctly 2006), which revealed small genetic variation among any of the populations studied. Polish populations, suggest that Europe was On the variation and distribution of the lake minnow, Eupallasella percnurus (Pall.) 165 recolonized by the lake minnow returning from the comments on its relationships and distribution – Bull. Br. so-called northern refugia located in present-day Mus. Nat. Hist. (Zool.), 48(1): 57-74. Ito Y., Sakai H., Shedko S., Jeon S-R. 2002 – Genetic differen- Hungary and Slovakia (Willis et al. 2000, Stewart tiation of the northern Far East cyprinids Phoxinus and and Lister 2001). This partly confirms Kaj’s (1954) Rhynchocypris – Fish. Sci. 68 (suppl. I): 75-78. view, the more so that the presence of lake minnow Kaj J. 1954 – Contribution to the knowledge of the species populations in western Ukraine (Movtchan and Phoxinus percnurus Pall. in Poland – Pol. Arch. Smirnov 1981) and their absence between the Dnie- Hydrobiol. 1: 49-78. Kusznierz J., Borsuk P., Kamiñski R., Koper M., Myszkowski per and Volga rivers indicate that recolonization did L., Popoviæ D., Wolnicki J., Stankoviæ, A. 2006 – Low not proceed from the Volga system, though at first genetic variability among the lake minnow Eupallasella glance such a migration route would seem most percnurus (, Cyprinidae) populations in likely. The extensive contact between the Volga River Poland – Fish. Manag. Ecol. 13: 131-134. system and the region of central and northeastern Mangerud J., Jakobsson M., Alexanderson H., Astakhov V., Clarke G.K.C., Henriksen M., Hjort Ch., Krinner G., Europe (Mangerud et al. 2004) in the not very remote Lunkka J.P., Möller P., Murray A., Nikolskaya O., past suggests that the populations inhabiting the Saarnistou M., Svendsen J.I. 2004 – Ice-dammed lakes Oder, Vistula, Dnieper, Dnester, North Dvina, and and rerouting of the drainage of northern Eurasia during Pechora river systems should be closer to the Last Glaciation – Quaternary Sci. Rev. 23: 1313-1332. Mitrofanov V.P. 1987 – Genus Phoxinus agassiz, 1835 – In: E. percnurus stagnalis from the Volga system than Fishes of Kazakhstan, vol. 2, (Eds) V.P. Mitrofanov, G.M. they are to the Siberian populations of the nominate Dukraviec, A.F. Sidorova, L.N. Soloninova, Alma-Ata: subspecies, which are separated from Europe by the 123-147 (in Russian). Urals. The very distinct morphological differences Movtchan Y.V., Smirnov A.I. 1981 – Fauna of Ukraine, vol. 8 – between E. percnurus stagnalis and E. percnurus Naukova Dumka Publishing House, Kiev, 367 p. (in Russian). Nikitin V.D., Safronov S.N. 2009 – History of studies, species percnurus seem to confirm the results of the most re- composition, morphology and distribution of cent palaeohydrographic studies (Mangerud et al. from the genus Rhynchocypris (Cyprinidae) on Sakhalin 2004), which suggest the possible existence of a con- Island – Proc. Irkutsk State University. Biology, Ecology. nection between the Pleistocene periglacial Lake 2(2): 41-44. Komi with the Ob River system; the connection, lo- Paœko £., Maœlak R. 1997 – Morphometric schemes and mea- surement error in ichthyological studies – Przegl. Zool. cated on the northern fringe of the Urals, could ex- 41: 221-231 (in Polish). plain the presence of E. percnurus percnurus in the Reyment R. A. 1991 – Multidimensional Palaeobiology – Pechora River system. The species could have Pergamon Press, 426 p. reached European waters along this route. Sakai H., Ito Y., Shedko S.V., Safronov S.N., Frolov S.V., Chereshnev I.A., Jeon S-R., Goto A. 2006 – Phylogenetic and Taxonomic Relationships of Northern Far Eastern Phoxinin Minnows, Phoxinus and Rhynchocypris (Pisces, References Cyprinidae), as Inferred from Allozyme and Mitochon- drial 16S rRNA Sequence Analyses – Zool. Sci. 23: 323-331. Berg L.S. 1949 – Freshwater fishes of the USSR and adjacent Stewart J.R., Lister A.M. 2001 – Cryptic northern refugia and countries (4th edition), vol. 2 – USSR Academy of Sci- the origins of the modern biota – Trends Ecol. Evol. 16: ences, Moscow-Leningrad, USSR, 456 p. (in Russian). 608-613. Bookstein F.L. 1991 – Morphometric Tools for Landmark Strauss R. E., Bookstein F. L. 1982 – The Truss: Body Form Data – Geometry and Biology. Cambridge University Reconstruction in Morfometric – Syst. Zool. 31: 113-135. Press. 455 p. Thienemann, A. 1950 – Verbreitungsgeschichte der G¹sowska M., Rembiszewski J.M. 1967 – The revision of the Süßwassertierwelt Europas – Die Binnengewässer 18: subspecies of the swamp-minnow Phoxinus percnurus 1-809. (Pallas) in Poland – Ann. Zool. 24: 305-341. Willis K.J., Rudner E., Sümegi, P. 2000 – The full-glacial for- Howes G.J. 1985 – A revised synonymy of the minnow genus ests of central and southeastern Europe – Quaternary Phoxinus Rafinesque, 1820 (Teleostei: Cyprinidae) with Res. 53: 203-213. 166 Jan Kusznierz et al.

Streszczenie

O zró¿nicowaniu i rozmieszczeniu strzebli b³otnej Eupallasella percnurus (Pall.)

Strzebla b³otna Eupallasella percnurus (Pall.) odznacza siê W prezentowanej pracy przedstawiono wyniki badañ morfo- bardzo szerokim zasiêgiem wystêpowania siêgaj¹cym od Pol- metrycznych przeprowadzonych z wykorzystaniem metody ski po Pó³wysep Czukocki, Sachalin, Hokkaido i Koreê. Na Truss Network oraz analizy kanonicznej, do których u¿yto tym rozleg³ym obszarze oraz terenach bezpoœrednio do niego 100 osobników strzebli b³otnej reprezentuj¹cych 5 populacji przyleg³ych wystêpuje ponadto szereg innych gatunków strze- pochodz¹cych z dorzeczy Peczory, Wo³gi, Obu i Leny oraz bli morfologicznie zbli¿onych do strzebli b³otnej, zaliczanych z wybrze¿a Morza Ochockiego. Uzyskane rezultaty wskazuj¹ aktualnie do rodzajów Phoxinus i Rhynchocypris. Ich relacje zasadnoœæ nadania strzebli wo³¿añskiej E. p. stagnalis statusu systematyczne w znacznej mierze pozostaj¹ ci¹gle niejasne. samodzielnego taksonu, najprawdopodobniej na poziomie ga- Ogromny zasiêg wystêpowania gatunku obejmuj¹cy szereg tunkowym. Uzyskany obraz zró¿nicowania morfometryczne- dorzeczy wielkich rzek, podzielony wieloma pasmami górski- go potwierdza wiêksze podobieñstwo populacji z dorzecza mi i w zwi¹zku z tym odznaczaj¹cy siê bardzo zró¿nicowan¹, Peczory do populacji wystêpuj¹cych na wschód od Uralu ni¿ bogat¹ histori¹ hydrograficzn¹ i glacjologiczn¹, z za³o¿enia do znacznie bli¿szej, w sensie geograficznym, populacji z do- sprzyja powstaniu zró¿nicowania lokalnych form. W efekcie rzecza Wo³gi. Wskazuje to na mo¿liwoœæ zasiedlenia dorzecza u strzebli b³otnej wyró¿niono piêæ podgatunków E. p. percnu- Peczory, a prawdopodobnie tak¿e Pó³nocnej DŸwiny, Dnie- rus, E. p. stagnalis, E. p. ignatowi, E. p. mantschuricus, pru, Dniestru, Wis³y i Odry przez strzeble pochodz¹ce z do- E. p. sachalinensis. Ze wzglêdu na du¿¹ zmiennoœæ morfolo- rzecza Obu, które do dorzecza Peczory mog³y dostaæ siê przez giczn¹ status tych form w dalszym ci¹gu budzi wiele w¹tpli- po³¹czenie tych dorzeczy usytuowane na pó³nocnych stokach woœci. Uralu.