Color Polymorphism and Historical Biogeography in the Japanese Patellogastropod Limpet Cellana nigrolineata (Reeve) (Patellogastropoda: Nacellidae) Author(s): Tomoyuki Nakano, Takenori Sasaki and Tomoki Kase Source: Zoological Science, 27:811-820. 2010. Published By: Zoological Society of Japan DOI: http://dx.doi.org/10.2108/zsj.27.811 URL: http://www.bioone.org/doi/full/10.2108/zsj.27.811 BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological, and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books published by nonprofit societies, associations, museums, institutions, and presses. Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use. Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder. BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions, research libraries, and research funders in the common goal of maximizing access to critical research. ZOOLOGICAL SCIENCE 27: 811–820 (2010) ¤ 2010 Zoological Society of Japan Color Polymorphism and Historical Biogeography in the Japanese Patellogastropod Limpet Cellana nigrolineata (Reeve) (Patellogastropoda: Nacellidae) Tomoyuki Nakano1*, Takenori Sasaki2 and Tomoki Kase1 1 Department of Geology and Palaeontology, National Museum of Nature and Science, 3-23-1 Hyakunin-cho, Shinjuku-ku, Tokyo 169-0073, Japan 2 The University Museum, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-0033, Japan Cellana nigrolineata is amongst the most common and largest patellogastropod limpets in Japan, and has two color morphs. Analyses of anatomical and morphological characters, shell structure, and mitochondrial COI data (658bp) of these color morphs suggested that they represent intraspecific genetic variation. Molecular analysis demonstrated that the species can be subdivided into three genetically distinct groups: (Clade 1) Honshu, Shikoku to Eastern Kyushu, (Clade 2) Western Kyushu and (Clade 3) Southern Kyushu. Clade 1 and Clade 2 + 3 are distributed on the coastlines adjacent to two warm-water currents, the Kuroshio and Tsushima Currents, respectively. The southern population (Clade 3) is currently isolated by inhospitable sandy shores. The subdivision of these groups likely dates from the glacial period of Plio-Pleistocene time. Key words: Cellana nigrolineata, color polymorphism, Patellogastropoda, phylogeography INTRODUCTION Patellogastropod limpets are common inhabitants of littoral shores, and constitute a major component of intertidal marine ecosystems (Branch, 1985a, b). Historically, their highly variable shell morphology has led to considerable confusion in the taxonomy of patellogastropods. However, recent molecular advances have provided important clues to clarify and improve the systematics and taxonomy of the patellogastropod limpets, and revealed that intraspecific variation in shell morphology and color are much more common than previously thought (e.g. Nakano and Spencer, 2007; Nakano et al., 2009a). For example, the New Zealand lottiid Notoacmea turbatrix Nakano, Marshall, Kennedy and Spencer, 2009 displays complex variation in color and shell morphology (Nakano et al., 2009a). Individuals inhabiting rock surfaces and the shells of Lunella smaragdus (Gmelin, 1791) or other limpets have a small shell with a narrower outline, whereas those in tide pools have depressed shells that are either black or white, which resemble two other New Zealand congeners, Notoacmea badia Oliver, 1926 and Notoacmea parviconoidea (Suter, 1907), respectively (Nakano et al., 2009a). Notoacmea scapha (Suter, 1907) represents another case. Although individuals of this species that are attached to the sea grass Zostera have a long, narrow, and laterally compressed shell, those found on living cockles or dead shells are larger and have a more rounded patelliform shell (Nakano and Spencer, 2007). These * Corresponding author. Phone: +81-3-3364-7133; Fax : +81-3-3364-7104; E-mail: tomo@kahaku.go.jp doi:10.2108/zsj.27.811 intraspecific variations are thought to be ecophenotypic, due to differences in the substrata occupied. Cellana nigrolineata (Reeve, 1839) is the largest and one of the most common species of patellogastropod limpets in the Japanese zoogeographical province (Sasaki, 2000). The species has long been known to have two color morphs, the ‘radial form’ with a radially striated pattern and the ‘concentric form’ with a concentrically wavy pattern (e.g., Sasaki, 2000) (Fig. 1). These color patterns are currently treated as intraspecific variations within a single species, but this distinction has never been verified using molecular techniques. The aims of this study are to test the specific intergrity of C. nigrolineata using a combination of traditional taxonomic approaches, such as soft body anatomy, shell structure, and radular morphology, together with molecular techniques, specifically cytochrome c oxidase subunit I (COI), and to discuss the historical biogeography of this species in the light of phylogeography and distribution. MATERIALS AND METHODS Field observation and collection of samples Live samples of C. nigrolineata were collected extensively from 34 localities along the coast of the Japanese Islands between Ayukawa, Miyagi Prefecture, in the north and Banshobana, Kagoshima Prefecture, in the south (Fig. 2, Table 1). Although the type locality of this species is Camiguin Island, Philippines, Powell (1973) suggested that this is probably in error. Indeed, C. nigrolineata was never found during extensive sampling in the Philippine region (in this study; Poppe, 2008). Therefore, the distribution of the species is considered to be restricted to the Japanese Islands and the southern part of Korea (Choe, 1992), and the sampling stations covered almost all the recorded distribution of C. nigrolineata 812 T. Nakano et al. (Fig. 2, Table 1). In total, 59 individuals of the two color morphs were sequenced. The closely allied species Cellana grata (Gould, 1859) and Cellana mazatlandica (Sowerby, 1839) were used as outgroups (Nakano and Ozawa, 2007; Nakano et al., 2009b). Fig. 1. Shell morphology and color pattern of Cellana nigrolineata. (A) UMUT RM30718, Myojin, Kagoshima Pref., (B) UMUT RM30719, Myojin, Kagoshima Pref., (C) UMUT69663, Kinkazan, Miyagi Pref., (D) UMUT RM30683, Yura, Sumoto-shi, Hyogo Pref., (E) UMUT RM30695, Shirahama, Miyazaki-shi, Miyazaki Pref., (F) UMUT RM30673, Sakai, Minabe-cho, Wakayama Pref., (G) UMUT RM30700, Koura, Minami Satsuma-shi, Kagoshima Pref., (H) UMUT RM30684, Mitsuiwa, Manazurucho, Kanagawa Pref., (I) UMUT RM30693, Tateyama, Chiba Pref., (J) UMUT RM30671, Hikari, Yamaguchi Pref., (K) UMUT RM30705, Yakeizima, Saeki-shi, Oita Pref., (L) UMUT RM30702, Shimoda-shi, Shizuoka Pref. DNA extraction, PCR amplification and DNA sequenceing A fragment of the mantle or foot muscle tissue was dissected from each specimen. Extraction of total DNA was performed using High Pure PCR Template Preparation Kit (Roche). The universal primers LCO1490 (5’-GGTCAACAAATCATAAGAATATTGG-3’) and HCO2198 (5’-TAAACTTCAGGGTGACCAAAAAATCA-3’) (Folmer et al., 1994) were used to amplify the COI gene. PCR amplification was performed in 25 µl of reaction volume containing 10 mM Tris-HCl ph 8.3, 50 mM KCl, 1.5 mM MgCl2, 200 µM dNTPs, 0.2 µM of a forward and reverse PCR primer, 0.5 mg/ml BSA (Sigma), 2 units of Taq polymerase (Takara), and 1 µl of template DNA solution. The cycling parameters for amplification consisted of an initial denaturation for 3 min at 94°C; followed by 30 cycles each of denaturation for 45 s at 94°C, annealing for 60 s at 50°C, and extension for 90 s at 72°C; and ended with a 5 min extension at 72°C. Successfully amplified PCR products were purified using High Pure PCR Product Purification Kit (Roche). Direct double-stranded cycle sequencing of 25 to 30 ng of each PCR product was performed in both directions using the Applied Biosystems BigDye v.3 dye terminator cycle sequencing kit. Cycle sequencing was performed using an Applied Biosystems GeneAmp PCR System 9700. The cycling parameters were 25 cycles of 10 s at 96°C, 5 s at 50°C, and 4 min at 60°C. Sequencing reaction products were purified using ethanol precipitation and analysed on an ABI PRISM 3130 DNA sequencer. Sequences were verified by forward and reverse comparisons. All sequences have been deposited in GenBank under accession numbers AB548154–AB548213. Datasets Two datasets were used for constructing trees. The large dataset included all 59 individuals, whereas the small data-set included 9 OTUs after removing redundant OTUs which have no sequence variations, to unweight OTUs. Only phylogenetic Biogeography of Cellana nigrolineata 813 converged. The log-likelihood scores were not increased after 500,000 generations, suggesting likelihood values reached a plateau. Anatomy and radular morphology Pigmentation of the head, cephalic tentacles, and foot were observed in fresh and fixed specimens, including all sequenced individuals. The black pigments of these portions are not faded by the preservative (Sasaki and Okutani, 1993). After the configuration of the radular sac was traced, the radula was dissected from all sequenced individuals. Extracted radulae were placed in 20% KOH at room temperature overnight, and then rinsed in distilled water. Radulae were observed with a scanning electron microscope (SEM). Shell structure Structural trends of the first order lamellae of crossed-lamellar structure within the inner shell layer (see MacClintock, 1967, p.37– 41) were observed using a binocular microscope. RESULTS Fig. 2. study. Collection localities of the specimens used in the present trees based on the large dataset are shown in this study, as the topologies of phylogenetic trees based on each dataset are for the most part identical. Phylogenetic analyses Sequences were assembled and edited using Sequencher (version 4.1, Gene Codes Corporation). Sequences of COI were aligned using MacClade 4.03 (Maddison and Maddison, 2002), with reference to translated amino acid sequences. Models to be used in phylogenetic analyses were determined by using MrModelTest v.2.1 (Nylander, 2004), which performs hierarchical likelihood-ratio tests of the nucleotide substitution models currently implemented in MrBayes. The substitution model chosen by MrModelTest was GTR + I + G, and this model was then used to calculate pairwise molecular distances among individuals. Phylogenetic analyses were performed with PAUP* v. 4b10 (Swofford, 2002) for neighbor-joining (NJ) (Saito and Nei, 1987) and equally weighted maximum parsimony (MP), as well as their associated bootstrap values (Felsenstein, 1988). MrBayes v.3.1.2 (Ronquist and Huelsenbeck, 2003) was used to estimate Bayesian posterior probabilities. The NJ bootstrap analysis consisted of 10,000 replicates, whereas the MP bootstrap analysis comprised 1,000 replicates of a heuristic search (with 10 random addition sequence replicates and TBR branch-swapping). MrBayes was run with the following settings; the maximum-likelihood model employed six substitution types (nst = 6); rate variation across sites was modeled using a gamma distribution, with a proportion of the sites being invariant (rate = invgamma); the shape, proportion of invariable sites, state frequency, and substitution rate parameters were estimated. The Markov-chain Monte-Carlo search was run with four chains (contain 1 cold chain and 3 heated chain) for 3,000,000 generations, with trees being sampled every 100 generations and the first 5,000 trees (i.e. 500,000 generations) discarded as burn-in. The likelihood scores of Bayesian analysis were plotted to verify the MCMC was Molecular data PCR amplification of the COI gene yielded a product of approximately 700 bp, and subsequent sequencing of this product routinely yielded a sequence of 658 bp. The aligned dataset of 658 bp characters, including the outgroup taxa (C. grata and C. mazatlandica), had 94 variable and 90 parsimony-informative characters. Molecular phylogeny All phylogenetic trees, whether constructed using NJ, equally weighted MP or Bayesian analyses, yielded three well-supported clades (bootstraps for NJ > 96% and MP > 86%, and Bayesian posterior probabilitiy > 0.99), corresponding to the following geographical distributions: Honshu, Shikoku to Eastern Kyushu (Clade 1), Western Kyushu (Clade 2) and Southern Kyushu (Clade 3), irrespective of color morphs (Fig. 3). Genetic distances between clades ranged from 0.3– 2.0% (Table 2). Clade 1 includes members with up to 0.3% genetic distance between them, whereas there were no variations within Clades 2 and 3. Anatomy and radular morphology The side of the foot (lf: Fig. 4C) is entirely gray except for the ventral margin. The cephalic tentacles (ct: Fig. 4C, F) are pigmented entirely in black (ct: Fig. 4), while the snout (sn: Fig. 4C, F) lacks dark pigmentation. The mantle margin (Fig. 4D) is bordered by a black line. The pallial tentacles (pt: Fig. 4D) are uniformly gray. There are four intestinal loops visible on the dorsal surface (i1–i4: Fig. 4E). The radular sac (rs: Fig. 4F, G) is very long and coiled in four loops on the right anterior side of the visceral mass. The docoglossate radula is typical of Cellana, with the formula 3–2–1–2–3 (Fig. 5). The central tooth is very narrow, represented by a vestigial small medio-central plate between the first pair of laterals. The first lateral teeth each bear a long, slender, erect, simple-pointed cusp. The second lateral teeth are long, with moderately curved cusps: each cusp is sharply pointed at the tip and weakly denticulate on each side at the base. The marginal teeth are in 814 T. Nakano et al. Table 1. Specimens and localities examined in the present study. *UMUT: The University Museum, The University of Tokyo, NUGB: Nagoya University, Geobiology. Species Locality Color pattern ID in Fig. 1 GenBank Voucher Cellana nigrolineata (Reeve, 1839) Kaino, Mie Pref., Japan Hikari, Yamaguchi Pref., Japan Hikari, Yamaguchi Pref., Japan Shishizima, Kagoshima Pref., Japan Sakai, Minabe-cho, Wakayama Pref., Japan Sakai, Minabe-cho, Wakayama Pref., Japan Tomioka, Reihoku-cho, Kumamoto Pref., Japan Tomioka, Reihoku-cho, Kumamoto Pref., Japan Koganezaki, Nishi Izu-cho, Shizuoka Pref., Japan Iwagaya, Shodoshima-cho, Kagawa Pref., Japan Ukusu, Nishi Izu-cho, Shizuoka Pref., Japan Seto, Shirahama-cho, Wakayama Pref., Japan Hazu, Okagaki-cho, Fukuoka Pref., Japan Yura, Sumoto-shi, Hyogo Pref., Japan Yura, Sumoto-shi, Hyogo Pref., Japan Mitsuiwa, Manazuru-cho, Kanagawa Pref., Japan Chouzyagasaki, Hayama-cho, Kanagawa Pref., Japan Karakizaki, Gosyourazima, Kumamoto Pref., Japan Karakizaki, Gosyourazima, Kumamoto Pref., Japan Ukishima, Nishi Izu-cho, Shizuoka Pref., Japan Ukishima, Nishi Izu-cho, Shizuoka Pref., Japan Ayukawa, Ishinomaki-shi, Miyagi Pref., Japan Ayukawa, Ishinomaki-shi, Miyagi Pref., Japan Tateyama, Chiba Pref., Japan Tateyama, Chiba Pref., Japan Hane-misaki, Muroto-shi, Kochi Pref., Japan Shirahama, Miyazaki-shi, Miyazaki Pref., Japan Shirahama, Miyazaki-shi, Miyazaki Pref., Japan Hiraura, Uwazima-shi, Ehime Pref., Japan Hiraura, Uwazima-shi, Ehime Pref., Japan Koura, Minami Satsuma-shi, Kagoshima Pref., Japan Koura, Minami Satsuma-shi, Kagoshima Pref., Japan Banshobana, Minami Kyushu-shi, Kagoshima Pref., Japan Shimoda-shi, Shizuoka Pref., Japan Motosaru, Saeki-shi, Oita Pref., Japan Yakeizima, Saeki-shi, Oita Pref., Japan Yakeizima, Saeki-shi, Oita Pref., Japan Tozakibana, Ichikikushikino-shi, Kagoshima Pref., Japan Tozakibana, Ichikikushikino-shi, Kagoshima Pref., Japan Nagasakibana, Ichikikushikino-shi, Kagoshima Pref., Japan Nagasakibana, Ichikikushikino-shi, Kagoshima Pref., Japan Okawa, Akune-shi, Kagoshima Pref., Japan Okawa, Akune-shi, Kagoshima Pref., Japan Banshonohana, Akune-shi, Kagoshima Pref., Japan Banshonohana, Akune-shi, Kagoshima Pref., Japan Seto, Nagashima-cho, Kagoshima Pref., Japan Banbabana, Nagashima-cho, Kagoshima Pref., Japan Banbabana, Nagashima-cho, Kagoshima Pref., Japan Myojin, Nagashima-cho, Kagosima Pref., Japan Myojin, Nagashima-cho, Kagosima Pref., Japan Myojin, Nagashima-cho, Kagosima Pref., Japan Myojin, Nagashima-cho, Kagosima Pref., Japan Myojin, Nagashima-cho, Kagosima Pref., Japan Fukudomari, Nagashima-cho, Kagoshima Pref., Japan Fukudomari, Nagashima-cho, Kagoshima Pref., Japan Fukudomari, Nagashima-cho, Kagoshima Pref., Japan Fukudomari, Nagashima-cho, Kagoshima Pref., Japan Fukudomari, Nagashima-cho, Kagoshima Pref., Japan Mie, Nagasaki-shi, Nagasaki Pref., Japan Mie, Nagasaki-shi, Nagasaki Pref., Japan Radial Ray Radial Ray Radial Ray Radial Ray Radial Ray Concentric Radial Ray Concentric Radial Ray Radial Ray Concentric Radial Ray Radial Ray Radial Ray Concentric Radial Ray Radial Ray Concentric Radial Ray Concentric Radial Ray Radial Ray Concentric Concentric Radial Ray Radial Ray Concentric Both Radial Ray Concentric Concentric Radial Ray Concentric Both Both Radial Ray Radial Ray Radial Ray Concentric Radial Ray Concentric Radial Ray Concentric Radial Ray Concentric Radial Ray Radial Ray Concentric Radial Ray Concentric Concentric Concentric Concentric Radial Ray Radial Ray Radial Ray Concentric Concentric Radial Ray Concentric L30 L37 L38 L39 L888 L889 L890 L891 L892 L893 L894 L895 L896 L897 L898 L899 L900 L901 L902 L903 L904 L905 L906 L934 L935 L946 L1010 L1011 L1012 L1013 L1014 L1015 L1016 L1101 L1117 L1118 L1119 L1120 L1121 L1122 L1123 L1124 L1125 L1126 L1127 L1128 L1129 L1130 L1131 L1132 L1133 L1134 L1135 L1136 L1137 L1138 L1139 L1140 L1145 L1146 I P P X J J V V G M H K T L L D C W W F F A A B B N S S O O g g h E Q R R f f e e d d c c b a a Z Z Z Z Z Y Y Y Y Y U U AB238548 AB548154 AB548155 AB548156 AB548157 AB548158 AB548159 AB548160 AB548161 AB548162 AB548163 AB548164 AB548165 AB548166 AB548167 AB548168 AB548169 AB548170 AB548171 AB548172 AB548173 AB548174 AB548175 AB548176 AB548177 AB548178 AB548179 AB548180 AB548181 AB548182 AB548183 AB548184 AB548185 AB548186 AB548187 AB548188 AB548189 AB548190 AB548191 AB548192 AB548193 AB548194 AB548195 AB548196 AB548197 AB548198 AB548199 AB548200 AB548201 AB548202 AB548203 AB548204 AB548205 AB548206 AB548207 AB548208 AB548209 AB548210 AB548211 AB548212 UMUT RM30669 UMUT RM30670 UMUT RM30671 UMUT RM30672 UMUT RM30673 UMUT RM30674 UMUT RM30675 UMUT RM30676 UMUT RM30677 UMUT RM30678 UMUT RM30679 UMUT RM30680 UMUT RM30681 UMUT RM30682 UMUT RM30683 UMUT RM30684 UMUT RM30685 UMUT RM30686 UMUT RM30687 UMUT RM30688 UMUT RM30689 UMUT RM30690 UMUT RM30691 UMUT RM30692 UMUT RM30693 UMUT RM30694 UMUT RM30695 UMUT RM30696 UMUT RM30697 UMUT RM30698 UMUT RM30699 UMUT RM30700 UMUT RM30701 UMUT RM30702 UMUT RM30703 UMUT RM30704 UMUT RM30705 UMUT RM30706 UMUT RM30707 UMUT RM30708 UMUT RM30709 UMUT RM30710 UMUT RM30711 UMUT RM30712 UMUT RM30713 UMUT RM30714 UMUT RM30715 UMUT RM30716 UMUT RM30717 UMUT RM30718 UMUT RM30719 UMUT RM30720 UMUT RM30721 UMUT RM30722 UMUT RM30723 UMUT RM30724 UMUT RM30725 UMUT RM30726 UMUT RM30727 UMUT RM30728 Cellana grata (Gould, 1859) Akagurizaki, Fukui Pref., Japan Kaino, Mie Pref., Japan L51 L54 AB548213 AB238546 NUGB-L51 NUGB-L54 L717 L718 L719 L720 AB433635 AB433636 AB433637 AB433638 NUGB-L717 NUGB-L718 NUGB-L719 NUGB-L720 Cellana mazatlandica Anezima, Ogasawara Is., Tokyo, Japan (Sowerby, 1839) Hirazima, Ogasawara Is., Tokyo, Japan North port, Hahazima, Ogasawara Is., Japan Sankakuiwa, Hahazima, Ogasawara Is., Japan Biogeography of Cellana nigrolineata 815 three pairs and are fused basally. Despite wide geographic distribution of the samples and irrespective of color morphs, no difference was detected in the internal anatomy, radular sac, or radular morphologies among the specimens examined. Shell structure The first-order lamellae exhibited a bilaterally symmetric pattern with lamellae extending laterally perpendicular to the median line in the anterior part, but progressively becoming curved posteriorly and radiating towards a point near the posterior margin (Fig. 6). Although the point of radiation in the posterior part varies in position to some extent, the general pattern of first-order lamellae is consistent in all shells of C. nigrolineata. Fig. 3. NJ phylogram generated from 658 bp COI data. Number above or under the branches are NJ bootstrap/MP bootstrap/Bayesian posterior probabilities. R and C indicate the radial and concentric forms of color pattern, respectively. B indicates both color patterns are observed in the specimen. Dimorphic color markings of C. nigrolineata Cellana nigrolineata is known to have dimorphic color markings. The abundant color form has reddishbrown radial rays over the whole shell surface (Fig. 1F–K), whereas the form with concentric markings (Fig. 1A–E) is less common. In addition, there are other rare color variations. For example, large dark spots may be arranged radially. If the radial rays are obscure, the color pattern appears to be intermediate between the two morphs (Fig. 1L). These uncommon patterns were found in less than 10% of individuals around the Kanto area, central Japan, and western Kyushu, in the southern part of Japan. The background color is blue-gray in all color morphs, which is unusual among patellogastropods. This bluegray pigment is only present in a thin surface layer and is easily removed by erosion (Fig. 1J, K), so that it is often absent in beach-drifted dead shells. DISCUSSION Table 2. Genetic distances (%): within and among clades. The genetic distances were calculated using a model selected by MrModeltest (GTR + I + G) Clade1 Clade2 Clade3 Clade1 Clade2 Clade3 0.00–0.3 1.5–1.7 1.8–2.0 0.00 0.3 0.00 Species identification The two color morphs of shells of C. nigrolineata are indistinguishable by anatomical characters. Among soft-part characters, the color pattern of the head, cephalic tentacles, and the side of the foot are recognized as useful characters for species identification of intertidal gastropods (e.g. Sasaki and Okutani, 1993; Reid, 1996; Meyer et al., 2005). The configuration of the radula sac has also been used for specieslevel classification of the lottiid limpet Nipponacmea (Sasaki and Okutani, 1993) and nacellid Nacella (Valdovinos and 816 T. Nakano et al. Rüth, 2005). None of these characters differed between the two color forms of C. nigrolineata. Shell microstructure also did not reveal any difference between the radial and concentric color morphs of C. nigrolineata. Recently, Kase et al. (submitted) suggested the orientation of the first order lamellae of crossedlamellar structures within the inner layer is useful as a supplementary character in species-level classification of fossil limpets. This character did not differ between the two color morphs of C. nigrolineata. The number and arrangement of shell layers may differ among species, but these are identical in most Japanese Cellana species, including C. nigrolineata (Fuchigami and Sasaki, 2005). Although three geographical clades are recognized in the phylogenetic trees, the two color morphs of the species can be found in sympatry within each clade, and there is no genetic differentiation between the two color morphs. The two color morphs of C. nigrolineata are, therefore, not distinguishable by either anatomical, shell morphological, or genetic data. Our results suggest that these color patterns represent intraspecific variation. Fig. 4. Anatomy of Cellana nigrolineata. (A) Dorsal view of animal. (B) Left lateral view of animal. (C) Same view with mantle removed. (D) Ventral view of mantle. (E) Dorsal view of animal after epithelium of visceral mass and part of digestive glands are removed. (F) Dorsal view of animal after digestive tract and digestive glands are removed. (G) Dorsal view of buccal mass and radular sac. (H) Dorsal view of odontophoral cartilages. (I) Ventral view of odontophoral cartilages. Abbreviations: a, auricle; ac, anterior cartilage; ct, cephalic tentacle; dg, digestive gland; epl, left efferent pallial sinus; epr, right efferent pallial sinus; ev, epithelium covering visceral mass; esv, esophageal valve; g, gonad; i1–i4, first to fourth intestinal loops on dorsal surface; lc, lateral cartilage; lf, lateral wall of foot; lk, left kidney; lp, lateral protractor muscle of odontophore; m, mantle; me, mid-esophagus; mm, mantle margin; mo, mouth; p, pericardium; plc, pallial cavity above head; pc, posterior cartilage; pdc, posterodorsal cartilages; pe, posterior esophagus; pm, pallial muscle; ps, pedal sole; pt, pallial tentacles; rds, radular sac; rdt, radular teeth; rk, right kidney; rsm, retractor muscle of subradular membrane; s, stomach; sm, shell muscle; srm, subradular membrane; sn, snout; v, ventricle; va, ventral approximator muscle; vlc, ventrolateral cartilage. Color polymorphism in Patellogastropoda It has often been reported that patellogastropods exhibit shell color polymorphism in connection with inhabiting different substrates. Giesel (1970) showed that Lottia digitalis (Rathke, 1833), on the Pacific coast of North America, has two color morphs: individuals living among colonies of barnacles display light color, while those on rocks display darker color. Nakano et al. (2009a) showed a similar phenomenon in the New Zealand taxa Notoacmea parviconoidea. The cause of shell color polymorphism in patellogastropods is believed to be ecophenotypic. By means of transplant experiments, Lindberg and Pearse (1990) showed that differences in the shell color of Lottia asmi (Middendorf, 1847) and L. digitalis on the Pacific coast of North America are caused by variation in food availability (especially that of algae) in different habitats. While not adaptive in an evolutionary sense, such ecophenotypic variation may still functional. Some Biogeography of Cellana nigrolineata 817 central Japan (Clade1) and western Kyushu, southern part of Japan (Clade2). The radial form is usually abundant, whereas the concentric form is found in less than 10% of individuals in each region. Therefore, a physical effect is unlikely for C. nigrolineata, at least at the present time. Fig. 5. Radular morphology of Cellana nigrolineata. Fig. 6. Distribution of shell microstructure and structure trends in internal shells of Cellana nigrolineata. authors have suggested that the color polymorphism of patellogastropods is a form of camouflage against predators (Giesel, 1970; Hocky et al., 1987; Byers, 1989). Sorensen and Lindberg (1991) suggested that a change from one habitat to the other exposes mismatched individuals to fatal attacks from predators. Such selective visual predation can maintain the strict spatial distribution of the color polymorphism (Giesel, 1970; Hocky et al., 1987; Byers, 1989). The two color forms of C. nigrolineata are micro-sympatric at all collection sites, suggesting that these color variations are unlikely to be controlled by differences in diet or habitat. The nature of selection on shell color forms is a matter of speculation. In addition to visual predation, physical agents, such as desiccation and temperature, can exert selection on color forms of intertidal molluscs (Sokolova and Berger, 2000). Dark-colored shells warm under the sun faster than do light-colored shells, and the former may be advantageous in cold climates, whereas light-colored shells may prevent fatal heating in warmer regions (Jones, 1973; Stine, 1989). Miura et al. (2007) found such examples in the mud snails Batillaria attramentaria (Sowerby, 1855) and B. multiformis (Lischke, 1869) in Japan. The dark-colored forms are dominant in colder regions, whereas the light colored forms increase in number towards warmer regions. In the case of C. nigrolineata, the two color morphs are appeared at the same frequencies in at least Kanto area, Phylogeography of C. nigrolineata Intraspecific genetic structure in marine organisms is strongly affected by historical changes of the environment, such as climate, ocean currents, and sea-level changes (e.g. Kojima et al., 1997, 2003, 2004). Patellogastropod limpets produce short-lived planktonic larvae that survive for up to ten days (Amio, 1963). Due to this limited dispersal ability, populations of patellogastropod limpets are believed to be readily isolated from each other. Therefore, the intraspecific genetic variation of patellogastropod limpets may reveal details of their historical biogeography. In the northwestern Pacific, phylogeographic studies have been conducted for marine intertidal gastropods, including Turbo cornutus Lightfoot, 1786, Batillaria attramentaria, and B. multiformis in Japan (e.g. Kojima et al., 1997, 2003, 2004). Two geographical clades have been reported, corresponding to the Pacific coast and Japan Seaside coast in the first two of these species: these have respectively been named the ‘Kuroshio’ and ‘Tsushima’ types, after the dominant warm-water currents in each area. In the present study, C. nigrolineata has been shown to be subdivided into three geographical populations: Honshu, Shikoku to Eastern Kyushu (Clade 1), Western Kyushu (Clade 2) and Southern Kyushu (Clade 3). This subdivision is similar to those previously found in T. cornutus and B. attramentaria (Kojima et al., 1997, 2004), although to our knowledge this is the first report that the population from southern Kyushu is genetically different from those in other areas. It is difficult to estimate divergence times of limpets, due to their limited fossil record (Koufopanou et al., 1999). If the divergence rate of COI is taken as 0.85–1.15% per million years, as calculated from cowries, and used for estimating divergence times of the limpet genus Eoacmaea (Kirkendale and Meyer, 2004), the divergence between Clade 1 and Clades 2 + 3 is estimated at about 1.28–4.0 Mya. This date is consistent with those of Turbo estimated by Kojima et al. (1997). Clade 2 is estimated to have diverged from Clade 3 about 0.26–0.35 Mya. The Kuroshio Current flows along the southern coast of Honshu from the Philippines via the west side of the Ryukyu Islands, and the Tsushima Current branches off from the Kuroshio Current at the southern end of Kyushu Island (Fig. 7). The routes of these two warm water currents appear to effectively prevent gene flow between the populations on the Pacific and the Japan Sea coasts, as shown by the distinct clades in the present and previous studies (Kojima et al., 1997, 2004). Chinzei (1991) suggested that this system of warm-water currents along the Japanese coasts had already been established at 15 to 17 Mya. However, most of the East China Sea was dry land during low sea-level phases in the early Pleistocene. In this period, there were no suitable habitats in Western Kyushu, and the Kuroshio Currents flowed only along the Pacific coast. If the three clades of C. nigrolineata had diverged before the early Pleistocene, they 818 T. Nakano et al. Fig. 7. Summary of the distribution of Cellana nigrolineata and warm-water currents around Japan. Square = Clade 1; black circle = Clade 2; white circle = Clade 3; triangle = records from literature (see text for details). would be expected to have had secondary contact, and lost genetic differentiation, in this period. Following sea-level rise, suitable habitats may have become available on the western side of Kyushu. It is likely that divergence between Clade 1 and Clade 2 + 3 took place during this period. Western Kyushu is thought to have been connected several times to the Asian continent during the glacial intervals, based on the similarity of the benthic fauna (Sato, 2000), the occurrence of warm-water molluscs, planktonic foraminiferans and diatoms (Kitamura and Kimoto, 2006) and the migration routes of mammals (Kawamura, 1998). At this time, the Japan Sea is thought to have been isolated and during the Last Glacial Maximum most of its marine benthic fauna may have become restricted or extinct due to reduced salinity and anoxic conditions (Oba et al., 1991). Indeed, the genetic diversity of Clades 2 and 3 is lower than that of Clade 1 (Pacific coast). According to Kojima et al. (1997), the Tsushima group may have been restricted to the western coast of Kyushu as a refuge during the glacial periods. Our results are consistent with this hypothesis. Kojima et al. (1997, 2004) suggested that the Tsushima group of T. cornutus and B. attramentaria are likely to have reinvaded northward along the coast of the Japan Sea with the inflow of the Tsushima Current after 10,000–8,000 years ago (Oba, 1983). Although there are several records of C. nigrolineata from the Noto Peninsula, Toyama Prefecture (Kawabata, 2004) and Sado Island, Niigata Prefecture (Kuroda, 1957; Honma and Kitami, 1978), the species has not been reported from the Oga Peninsula, Akita Prefecture (Sasaki, 1995) or southern Hokkaido (Conchological Club of Northern Regions, Hokkaido University, 2009). Judging from these records, the northern limit of C. nigrolineata is Sado Island, Niigata Prefecture, on the Japan Sea coast. During our field collections, we did not find any specimens of the species from the middle to northern part of the Japan Sea coast and the species may be rare or absent there. The oceanic conditions of southern Kyushu changed dramatically during the Plio-Pleistocene epoch. For example, based on a compilation of fossil and recent distribution records of Ostracoda, it is believed that Neomonoceratina delicata (Ishizaki and Kato, 1976) migrated south of the Tokara Strait, while nearly all populations of this species disappeared to the north of the Tokara Strait during the Last Glacial Maximum due to the decrease of water temperature (Irizuki et al, 2009). In contrast, the Kuroshio Current may have flowed strongly into Kagoshima Bay 7,000–5,000 years ago, since fossil Lunulicardia retusa (Linnaeus, 1758) have been reported, whereas this species is currently distributed in the Amami Islands and southwards (Oki, 2002). Recent molecular work has also suggested the connection between southern Kyushu and Okinawa; for example, Stomatella planulata (Lamarck, 1816) are genetically identical in these areas (Williams et al., 2010). The low genetic diversity of the southern Kyushu population of C. nigrolineata may be also explained by a bottleneck due to the dramatic changes of water temperature during the PlioPleistocene. The southern part of Kyushu seems at present to be isolated by the sandy shores of southwestern and southeastern Kyushu. These sandy shores may prevent the extension of the southern population even if the ocean current flows northwards along the coastline of eastern Kyushu, since the haplotypes of southern populations are not found intermixed with those of the other populations. Dispersal is determined not only by larval period and geographic distance, but also by habitat specificity (Kirkendale and Meyer, 2004; Ayre et al., 2009). Members of the limpet genus of Eoacmaea all inhabit calcareous substrata (Lindberg and Vermeij, 1995; Nakano et al., 2005; Kirkendale and Meyer, 2004). So, they have actually limited geographical distribution for their limited habitats and show fine-scale endemisms (Kirkendale and Meyer, 2004). Ayre et al. (2009) analyzed ten rocky intertidal invertebrates, including a limpet Cellana tramoserica (Holton, 1802), to test the hypothesis that larval type and habitat specificity are predictors of gene flow. They showed that there is a phylogeographic break at the southeast corner of Australia for rock specialists with planktonic larvae, whereas species that have relaxed habitat specificity show no genetic differentiation. They suggested that a sheltered-shore without suitable habitat in the southeast corner of Australia may present a barrier to gene flow for rocky specialists, despite the larval type. We also did not find any patellogastropod limpets on the small rocks on sandy beaches at western and eastern Kyushu, except for a small number of juvenile Cellana toreuma (Reeve, 1854) and Patelloida saccharina (Reeve, 1855). These individuals likely represent a temporal colonization that will soon disappear. It therefore seems plausible that inhospitable sandy shores can prevent the gene flow in these areas. According to Sasaki (2000), the previous record of the northern limit of C. nigrolineata on the Pacific coast was the Boso Peninsula, central Japan. However, C. nigrolineata Biogeography of Cellana nigrolineata was found in Ayukawa, Oshika Peninsula, northern Japan, in the present study. In this area, the cold-water species Lottia emydia (Dall, 1914), warm-temperate species Nipponacmea gloriosa (Habe, 1944), and C. nigrolineata are found sympatrically (Nakano, personal observation). The occurrence of these species suggests that this is a mixing zone of warm- and cold-water currents. In the case of B. attramentaria, the Tsushima type of the species reaches Iwate Prefecture on the Pacific coast via the Tsugaru Strait between the northernmost Honshu and southern Hokkaido. However, Clade 2 of C. nigrolineata does not reach the Pacific coast, since the haplotypes of Clade 2 was not found on the Pacific coast. Therefore, the distribution of the population of Oshika Prefecture appears likely to extend from Boso Peninsula, central Japan along the Pacific coast. There may be broader suitable habitats in the Pacific coast during the glacial period, since the population of Clade 1 has higher genetic diversity than those of Clade 2 and 3. ACKNOWLEDGMENTS We thank Dr D. G. Reid, Dr G. A. Williams, Dr B. K. K. Chan and anonymous reviewers for making many useful comments to improve the manuscript. We thank Dr Y. Kano, Mr T. Haga, Dr Y. Ise, Mr. M. Okanishi, Mr T. Hamada, Mr H. Fukumori and Ms M. Fuji for providing specimens. Dr K. Hasegawa provided literature information. This study was funded by a Grant-in-Aid for Scientific Research project no. 207024 to TN from the Japan Society for the Promotion of Science and the National Museum of Nature and Science, Tokyo under the 130th anniversary research project entitled ‘Studies on the Geography and Evolution of Biodiversity in Japan’. REFERENCES Amio M (1963) A comparative embryology of marine gastropods, with ecological considerations. J Shimonoseki Univ Fish 12: 229–358. (in Japanese with English abstract) Ayre DJ, Minchinton TE, Perrin C (2009) Does life history predict past and current connectivity for rocky intertidal invertebrates across a marine biogeographic barrier? Mol Ecol 18: 1887– 1903 Branch GM (1985a) Limpets: evolution and adaptation. In “The Mollusca. Vol 10” Ed by ER Trueman, MR Clarke, Academic Press, New York, pp 187–220 Branch GM (1985b) Limpets: their role in littoral and sublittoral community dynamics. In “The ecology of rocky coasts” Ed by PG Moore, R Seed, Hodder & Stoughton, London, pp 97–116 Byers BA (1989) Habitat-choice polymorphism associated with cryptic shell-color polymorphism in the limpet Lottia digitalis. Veliger 32: 394–402 Chinzei K (1991) Late Cenozoic zoogeography of the Sea of Japan area. Episodes 14: 231–235 Choe BL (1992) Illustrated Encyclopedia of Fauna and Flora of Korea. Vol 33. Mollusca (II). Ministry of Education, South Korea, Seoul, pp 860 (in Korean) Conchological Club of Northern Regions, Hokkaido University (2009) Molluscan Fauna of Usujiri, Hokkaido. Conchological Club of Northern Regions, Hokkaido University, pp 576 Felsenstein J (1988) Phylogenies from molecular sequences: inference and reliability. Ann Rev Genet 22: 521–565 Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R (1994) DNA primers for amplification of mitochondrial cytochrome c oxidase subunit I from diverse metazoan invertebrates. Mol Mar Biol Biotech 3: 294–299 Fuchigami T, Sasaki T (2005) The shell structure of the Recent 819 Patellogastropoda (Mollusca: Gastropoda). Paleo Res 9: 143– 168 Giesel JT (1970) On the maintenance of a shell pattern and behavior polymorphism in Acmaea digitalis, a limpet. Evolution 24: 98– 119 Hocky PAR, Bosman AL, Ryan PG (1987) The maintenance of polymorphism and cryptic mimesis in the limpet Scurria variabilis by two species of Cinclodes (Aves: Furnariinae) in central Chile. Veliger 30: 5–10 Honma Y, Kitami T (1978) Fauna and flora in the waters adjacent to the Sado Marine Biological Station, Niigata University. Ann Rep Sado Mar Biol Station, Niigata Univ 8: 7–81 Irizuki T, Taru H, Taguchi K, Matsushima Y (2009) Paleobiogeographical implications of inner bay Ostracoda during the Late Pleistocene Shimosueyoshi transgression, central Japan, with significance of its migration and disappearance in eastern Asia. Palaeogeo Palaeoclim Palaeoeco 271: 316–328 Jones JS (1973) Ecological genetics and natural selection in mollusks. Science 182: 439–453 Kawabata Y (2004) Catalogue of the specimens preserved in Nanao Children Science Museum: Catalogue of the Mollusks from Noto Peninsula, central Japan. Rep Nanao Children Sci Mus 8: 43–82 Kawamura Y (1998) Immigration of Mammals into the Japanese Islands during the Quaternary. Quat Res 37: 251–257 Kirkendale LA, Meyer CP (2004) Phylogeography of the Patelloida profunda group (Gastropoda: Lottiidae): diversification in a dispersal-driven marine system. Mol Ecol 13: 2749–2762 Kitamura A, Kimoto K (2006) History of the inflow of the warm water Tsushima Current into the Sea of Japan between 3.5 and 0.8 Ma. Palaeogeo Palaeoclim Palaeoeco 236: 355–366 Kojima S, Segawa R, Hayashi I (1997) Genetic differentiation among populations of the Japanese turban shell Turbo (Batillus) cornutus corresponding to warm currents. Mar Ecol Prog Ser 150: 149–155 Kojima S, Kamimura S, Kimura T, Hayashi I, Iijima A, Furota T (2003) Phylogenetic relationships between the tideland snails Batillaria flectosiphonata in the Ryukyu Islands and B. multiformis in the Japanese Islands. Zool Sci 20: 1423–1433 Kojima S, Hayashi I, Kim D, Iijima A, Furota T (2004) Phylogeography of an intertidal direct-developing gastropod, Batillaria cumingi, around the Japanese Islands. Mar Ecol Prog Ser 276: 161–172 Koufopanou V, Reid DG, Ridgway SA, Thomas RH (1999) A molecular phylogeny of the patellid limpets (Gastropoda: Patellidae) and its implications for the origins of their antitropical distribution. Mol Pylogenet Evol 11: 138–156 Kuroda T (1957) Catalogue of Mollusks in Sado. Trans Sado Nat Hist Soc 1: 13–32 Lindberg DR, Pearse JS (1990) Experimental manipulation of shell color and morphology of the limpets Lottia asmi (Middendorff) and Lottia digitalis (Rathke) (Mollusca, Patellogastropoda). J Exp Mar Biol Ecol 140: 173–185 Lindberg DR, Vermeij GJ (1995) Patelloida chamorrorum spec. nov.: A new member of the Tethyan Patelloida profunda group (Gastropoda: Acmaeidae). Veliger 27: 411–417 MacClintock C (1967) Shell structure of patelloid and bellerophontid gastropods (Mollusca). Bull Peabody Mus Nat Hist 22: 1–140 Maddison DR, Maddison WP (2002) MacClade4: Analysis of phylogeny and character evolution, version 4.03. Sinauer Associates, Sunderland, Massachusetts Meyer CP, Geller JB, Paulay G (2005) Fine scale endemism on coral reefs: archipelagic differentiation in turbinid gastropods. Evolution 59: 113–125 Miura O, Nishi S, Chiba S (2007) Temperature-related diversity of shell colour in the intertidal gastropod Batillaria. J Mollus Stud 73: 235–240 820 T. Nakano et al. Nakano T, Ozawa T (2007) Worldwide phylogeography of limpets of the order Patellogastropoda: Molecular, morphological and palaeontological evidence. J Mollus Stud 73: 79–99 Nakano T, Spencer HG (2007) Simultaneous polyphenism and cryptic species in an intertidal limpet from New Zealand. Mol Phylogenet Evol 45: 470–479 Nakano T, Aswan, Ozawa T (2005) A new limpet (Gastropoda: Lottiidae) of the Patelloida profunda group from Java, Indonesia, with notes on co-occuring species. Venus 64: 31–38 Nakano T, Marshall BA, Kennedy M, Spencer HG (2009a) The phylogeny and taxonomy of New Zealand Notoacmea and Patelloida species (Mollusca: Patellogastropoda: Lottiidae), inferred from DNA sequences. Mollus Res 29: 33–59 Nakano T, Yazaki I, Kurokawa M, Yamaguchi K, Kuwasawa K (2009b) The origin of the endemic patellogastropod limpets of the Ogasawara Islands in the northwestern Pacific. J Mollus Stud 75: 87–90 Nylander, J (2004) MrModeltest v. 2.1. Department of Systematic Zoology, Uppsala University, Sweden Oba T (1983) Paleoenvironment in Japan Sea after the glacial maximum period. Chikyu 5: 37–46 (in Japanese) Oba T, Kato M, Kitazato H, Koizumi I, Omura A, Sakai T, Takayama T (1991) Paleoenvironmental changes in the Japan Sea during the last 85,000 years. Paleoceanography 6: 499–518 Oki K (2002) Changes in depositional environments during the postglacial stage in Kagoshima Bay and seas around the northern part of the Ryukyu Islands. Quat Res 41: 237–251 Poppe GT (2008) Philippine Marine Mollusks. In “Gastropoda - Part 1”, ConchBooks, pp 1–759 Powell AWB (1973) The patellid limpets of the world (Patellidae). Indo-Pacific Mollusca 3: 75–206 Reid DG (1996) Systematics and evolution of Littorina. Ray Society, London, pp 363 Ronquist F, Huelsenbeck JP (2003) MrBayes3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572– 1574 Saito N, Nei M (1987) The neighbor-joining method: a new method for reconstructed phylogenetic trees. Mol Biol Evol 4: 406–425 Sasaki T (1995) Limpet fauna in Oga Peninsula, northern part of Sea of Japan. Chiribotan 26: 1–6 (in Japanese) Sasaki T (2000) Patellogastropoda. In “Marine mollusks in Japan” Ed by T Okutani, Tokai University Press, Japan, pp 24–33 Sasaki T, Okutani T (1993) New genus Nipponacmea (Gastropoda: Lottiidae): a revision of Japanese limpets hitherto allocated in Notoacmea. Venus 52: 1–40 Sato M (2000) Life in the Ariake Sea: biodiversity in tidal flats and estuaries. Kaiyusha, Tokyo (in Japanese) Sokolova IM, Berger VY (2000) Physiological variation related to shell color polymorphism in White Sea Littorina saxatilis. J Exp Mar Biol Ecol 245: 1–23 Sorensen FE, Lindberg DR (1991) Preferential predation by American black oystercatchers on transitional ecophenotypes of the limpet Lottia pelta (Rathke). J Exp Mar Biol Ecol 154: 123–136 Stine OC (1989) Cepaea nemoralis from Lexington, Virginia: the isolation and characterization of their mitochondrial DNA, the implications for their origin and climatic selection. Malacologia 30: 305–315 Swofford DL (2002) PAUP*: Phylogenetic Analysis Using Parsimony (* and other Methods), version 4. Sinauer Associations, Sunderland, Massachusetts Valdovinos C, Rüth M (2005) Necellidae limpets of the southern end of South America: taxonomy and distribution. Revista Chilena de Historia Natural 78: 497–517 Williams ST, Donald KM, Spencer HG, Nakano T (2010) Molecular systematics of the marine gastropod families Trochidae and Calliostomatidae (Mollusca: Superfamily Trochoidea). Mol Phylogenet Evol 54: 783–809 (Received March 5, 2010 / Accepted May 16, 2010)