Sea Urchins of the Genus Gracilechinus Fell & Pawson, 1966

Home , Echinidae, Echinoida, Echinus melo, Gracilechinus

This article was downloaded by: [Kirill Minin] On: 02 October 2014, At: 07:19 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Marine Biology Research Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/smar20 Sea urchins of the genus Gracilechinus Fell & Pawson, 1966 from the Pacific Ocean: Morphology and evolutionary history Kirill V. Minina, Nikolay B. Petrovb & Irina P. Vladychenskayab a P. P. Shirshov Institute of Oceanology, Russian Academy of Sciences, Moscow, Russia b A. N. Belozersky Research Institute of Physico-Chemical Biology, Moscow State University, Moscow, Russia Published online: 29 Sep 2014.

Click for updates

To cite this article: Kirill V. Minin, Nikolay B. Petrov & Irina P. Vladychenskaya (2014): Sea urchins of the genus Gracilechinus Fell & Pawson, 1966 from the Pacific Ocean: Morphology and evolutionary history, Marine Biology Research, DOI: 10.1080/17451000.2014.928413 To link to this article: http://dx.doi.org/10.1080/17451000.2014.928413

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions Marine Biology Research, 2014 http://dx.doi.org/10.1080/17451000.2014.928413

ORIGINAL ARTICLE

Sea urchins of the genus Gracilechinus Fell & Pawson, 1966 from the Pacific Ocean: Morphology and evolutionary history

KIRILL V. MININ1*, NIKOLAY B. PETROV2 & IRINA P. VLADYCHENSKAYA2

1P. P. Shirshov Institute of Oceanology, Russian Academy of Sciences, Moscow, Russia, and 2A. N. Belozersky Research Institute of Physico-Chemical Biology, Moscow State University, Moscow, Russia

Abstract Gracilechinus is a widely distributed genus of regular echinoids with a centre of species diversity in the North Atlantic (five of 10 species). Only three species of this genus are known from the Pacific Ocean: Gracilechinus multidentatus, G. euryporus and G. lucidus. The first two species were previously assigned to the genus Echinus, which differs from Gracilechinus in ambulacral tuberculation. According to our data, all three Pacific species possess ambulacral tuberculation characteristic of the genus Gracilechinus and differ from each other in the structure of the valves of tridentate pedicellariae and primary spines. Assignment of these three species to the genus Gracilechinus is also confirmed by phylogenetic analysis of the 657 bp fragment of the COI gene. The COI tree provides evidence for the validity of the genera Gracilechinus and Echinus and agrees in general with morphological data. Among species of Gracilechinus, two lineages are distinguished on the COI tree, differing in the shape of the tridentate pedicellariae valves. One lineage comprises only the North Atlantic species Gracilechinus alexandri, the other one includes the North Atlantic, South Atlantic and Pacific species. The genus Gracilechinus apparently originated in the North Atlantic and only representatives of the second lineage have penetrated the South Atlantic and the Pacific.

Key words: Biogeographic history, COI, Echinus, Gracilechinus, Pacific, phylogeny

Introduction other species of this genus, E. elegans Düben & Koren, 1844, E. alexandri Danielssen & Koren, Gracilechinus and Echinus are very closely related 1883, E. affinis Mortensen, 1903, E. atlanticus genera of sea urchins belonging to the family Echinidae. According to Mortensen (1943), the Mortensen, 1903, E. stenoporus Mortensen, 1942, genus Echinus included 17 species. Fell & Pawson and E. lucidus Döderlein, 1885, were transferred (1966) split Echinus and founded the new genus, to Gracilechinus (Smith & Kroh 2011; Kroh & Gracilechinus, to accommodate the species Echinus Hansson 2013).

Downloaded by [Kirill Minin] at 07:19 02 October 2014 acutus Lamarck, 1816 and E. gracilis A. Agassiz, The assignment of E. multidentatus and E. eur- 1869. The new genus comprised species with a yporus H. L. Clark, 1912 to Echinus still remains primary tubercle on each ambulacral plate. Other questionable. Both for E. multidentatus (McKnight species were retained in the genus Echinus, charac- 1967) and E. euryporus (Clark 1925; Mortensen terized by having a primary tubercle on every second 1943) it was shown that they have a primary tubercle or third, or second and third ambulacral plate. on every ambulacral plate, which is the main McKnight (1967) transferred another Echinus spe- diagnostic feature of Gracilechinus. cies, Echinus multidentatus H. L. Clark, 1925, to the Gracilechinus has its centre of species diversity in genus Gracilechinus on the basis of the ambulacral the North Atlantic (five of eight species listed in tuberculation pattern of the type specimen and Smith & Kroh (2011) and Kroh (2012a)) with specimens from the Chatham Rise. Recently species occurring in a wide range of habitats and G. multidentatus was moved back to Echinus, but six depths from intertidal to the deep sea. According to

*Corresponding author: Kirill V. Minin, P. P. Shirshov Institute of Oceanology, Russian Academy of Sciences, Nakhimovsky Prospekt, 36, Moscow 117997, Russia. E-mail: [email protected]

Published in collaboration with the Institute of Marine Research, Norway

Smith and Kroh (2011) and Kroh (2012a), Graci- to the Pacific Ocean from 1958 to 1987. Echinoids lechinus is represented in the Pacific Ocean by the were collected from the Southeast Pacific by the RV single species, G. lucidus (Döderlein, 1885). Two Akademik Kurchatov and RV Professor Shtokman, Echinus species, E. multidentatus and E. euryporus, from the South Pacific by RV Dmitry Mendeleev, possibly belonging to Gracilechinus, also occur in the and from the Northwest Pacific by RV Vityaz and Pacific Ocean (Mortensen 1943; Clark 1912; RV Odissey (Table I). In total 103 specimens of McKnight 1967, 1968; Kroh 2012b, 2012d). The Gracilechinus were obtained from six stations. All morphology and distribution of these three Pacific specimens were preserved in 96% ethanol on board species remain poorly known. the ships and then stored in the collections at the P. A study of the biogeographic history of Gracilechi- P. Shirshov Institute of Oceanology. Spine muscles nus was performed by Tyler et al. (1995). These and tube feet of one specimen of Echinus euryporus authors considered species of Gracilechinus in the H. L. Clark, 1912 (RV Professor Shtokman, St. 1964; genus Echinus. It was assumed that Echinus has a EspStock1964) and two specimens of Gracilechinus North Atlantic origin and that colonization of the lucidus (Döderlein, 1885) (RV Vityaz, St. 3768; different regions of the Pacific occurred separately. ElucV3768-25 and ElucV3768-27) were used for One possible pathway passed through the South the DNA extraction. Specimens of E. multidentatus Atlantic, turning eastward towards the Southwest H. L. Clark, 1925 (RV Dmitry Mendeleev, St. 1268) Pacific. This route coincides with the flow of the were unsuitable for DNA extraction because of poor North Atlantic Deep Water (NADW) capable of preservation condition. transporting larvae from the North Atlantic to the South Atlantic and then to the Southwest and Northwest Pacific. Because the NADW does not Morphological examination enter the Southeast Pacific, two possible pathways of Spines and pedicellariae for examination were colonization of this region were proposed: from the removed from the test and placed in sodium hypo- West Atlantic through the Straits of Panama or chlorite solution (household bleach), then washed through the Drake Passage (Tyler et al. 1995). with water and studied under a binocular dissecting In the present paper, redescriptions of the three microscope. The ultrastructure of cleaned spines and Pacific species of Gracilechinus are given. Phylogen- pedicellariae was studied using scanning electron etic relationships between the Pacific and the microscopy (SEM). Spines, pedicellariae and soft Atlantic Gracilechinus species are discussed and a tissue were removed from the test with a hard brush biogeographic history of this genus is proposed to make the plating and tuberculation patterns visible. based on the molecular phylogenetic tree generated Character terminology of the test follows Morten- using a fragment of the mitochondrial COI gene. sen (1943). Character terminology of pedicellariae is shown in Figures 3 and 4. In the present paper the Material and methods following abbreviations and symbols are used: TD, test diameter; TH, test height; Ap, diameter of the Material apical system; Per, diameter of the peristome; A, Material for morphological examination was col- number of ambulacral plates; I, number of inter- Downloaded by [Kirill Minin] at 07:19 02 October 2014 lected by Russian research vessels during five cruises ambulacral plates.

Table I. Samples examined of the Pacific species of Echinus and Gracilechinus. Date format: dd.mm.yyyy.

Depth Number of Species Vessel Station Date Latitude Longitude (m) specimens

E. multidentatus RV Dmitry 1268 11.01.1976 43°00′S 172°21′E 850 14 Mendeleev E. euryporus RV Akademik 3615 13.02.1982 46°04′S 83°50′W 220–240 24 Kurchatov RV Professor 1964 30.04.1987 24°56.3′–24° 88°32.6′–88° 580–564 2 Shtokman 58.2′S 29.8′ W G. lucidus RV Vityaz 3768 08.10.1958 30°30.8′N 129°19.4′E 818–824 57 RV Odissey Trawl 16.08.1984 37°55′N 170°22′E 1400 1 no. 33 RV Odissey Trawl 25.08.1984 34°48′N 171°47′E 820 5 no. 40 Pacific Gracilechinus: Morphology and evolutionary history 3

Table II. Sequences of COI gene obtained for different species of genera Echinus, Gracilechinus and outgroup species, Sterechinus neumayeri.

Source of COI Length of COI GenBank Species Code sequences Primers used fragment (bp) no.

Echinus esculentus EeDZ05 Authors AsF1, urR2 849 KF500005 EeDZ07 Authors AsF1, urR2 849 KF500006 Echinus euryporus EspStock1964 CCDB LCOech1aF1, HCO2198 658 KF642986 Echinus melo EmV75 CCDB LCOech1aF1, HCO2198 658 KF642990 Echinus multidentatus Ech179 GenBank See Ward et al. (2008) 836 EU869930 Ech181 GenBank See Ward et al. (2008) 841 EU869928 Ech182 GenBank See Ward et al. (2008) 841 EU869927 Gracilechinus affinis EafJC4816-162 Authors AsF1, urR2 849 KF500007 Gracilechinus alexandri EalJC4816-162 Authors AsF1, urR2 849 KF500014 EalJC4816-173 Authors AsF1, urR2 849 KF500008 Gracilechinus atlanticus EatlAI2146-4-2 Authors AsF1, urR2 849 KF500009 Gracilechinus elegans EelMa1 Authors AsF1, urR2 851 KJ729124 EelSka1 Authors AsF1, urR2 846 KJ729122 EelMo2 Authors AsF1, urR2 830 KJ729123 Gracilechinus lucidus ElucV3768-25 Authors AsF1, urR2, EcoinF, EcoinR 849 KF500011 ElucV3768-27 Authors AsF1, urR2, EcoinF, EcoinR 849 KF500012 Gracilechinus stenoporus EstAI2187-1 Authors AsF1, urR2 849 KF500013 Sterechinus neumayeri AY275548 GenBank See Lee & Chung (2003) 1077 AY275548

DNA extraction, amplification and sequencing PCRs were performed in the Biorad PTC-02206 Total genomic DNA was extracted from the tube Dyad thermocycler using the Evrogen Encyclo PCR feet and spine muscles of two specimens of Graci- kit (including Encyclo polymerase mix) according to lechinus lucidus (RV Vityaz, St. 3768), two specimens the manufacturer’s protocol in 25 µl reaction volumes. of G. alexandri (Danielssen & Koren, 1883), one After 10 min of initial heating at 94°C, amplifica- specimen of G. affinis (Mortensen, 1903), three tion was performed in 30 repetitions of the three- specimens of G. elegans (Düben & Koren, 1844), step cycle (denaturation, 94°C for 30 s; annealing, one specimen of G. atlanticus (Mortensen, 1903), 50°C for the primer pair AsF1-urR2 and 48°C for one specimen of G. stenoporus (Mortensen, 1942) primer pairs AsF1-EcoinR and EcoinF-urR2, for and two specimens of Echinus esculentus Linnaeus, 30 s; elongation, 72°C for 1 min) with the final 1758 (Table II). DNA was extracted using the extension at 72°C for 7 min. PCR products were Macherey–Nagel NucleoSpin® Tissue kit according purified with the Cytokine DNA purification kit to the manufacturer’s protocol. according to the manufacturer’s protocol. The A 830–849 bp fragment of COI gene corresponding sequencing was accomplished by the Center of – to the positions from 5886 to 6715 6734 of the Collective Use ‘Genome’, in an Applied Biosystems complete mitochondrion of Paracentrotus lividus 3730 DNA Analyzer automatic sequencer using the Downloaded by [Kirill Minin] at 07:19 02 October 2014 (Lamarck, 1816) (Cantatore et al. 1989; Genbank ABI PRISM® BigDye™ Terminator v. 3.1 kit. accession number NC_001572) was amplified from COI sequences of Echinus melo Lamarck, 1816 and the total genomic DNA using the polymerase chain E. euryporus (RV Professor Shtokman, St. 1964) were reaction (PCR). PCR primers were designed from the obtained at the Canadian Center for DNA Barcoding conserved regions with reference to the complete (CCDB, Table II) as a part of the international mitochondrial DNA of P. lividus (Cantatore et al. Barcode of Life programme (project code KMBSU). 1989). The forward primer was taken from the 5′-part of the COI gene (AsF1, 5′-TTT-CTA-CTA- Total genomic DNA was extracted from tube-feet AAC-ATA-AGG-ACA-TTG-G-3′) and the reverse tissue and a 658 bp fragment of COI gene correspond- – primer from the 3′-part of the COI gene (urR2, ing to positions 5885 6542 of the P. lividus mitochon- 5′-TCG-TGT-GTC-TAC-GTC-CAT-TCC-TAC- drion (Cantatore et al. 1989) was amplified and TGT-RAA-CAT-RTG-3′). Owing to the poor qual- sequenced in accordance with standard CCDB meth- ity of the samples of G. lucidus, it appeared neces- ods and protocols (Ivanova et al. 2005, 2006). Amp- sary to design intermediate primers from sequences lification and sequencing at CCDB were carried out obtained from other Gracilechinus species (EcoinF, using the forward primer LCOech1aF1 (5′-TTT- 5′-GAT-GAA-CTA-TTT-ACC-CCC-CTC-TAT-3′; TTT-CTA-CTA-AAC-ACA-AGG-ATA-TTG-G-3′) EcoinR, 5′-ATA-GAG-GGG-GGT-AAA-TAG-TTC- designed by Doug Eernisse (unpublished data) and ATC-3′). the reverse primer HCO2198 (5′-TAA-ACT-TCA- 4 K. V. Minin et al.

GGG-TGA-CCA-AAA-AAT-CA-3′) designed by to construct the majority rule consensus tree and to Folmer et al. (1994). calculate posterior probability for the nodes (average All sequences are deposited in GenBank under standard deviation of split frequencies = 0.005149). accession numbers KF500005–KF500009, In our analysis, another convergence diagnostic, the KF500011–KF500014, KF642986, KF642990 and Potential Scale Reduction Factor, PSRF (Gelman & KJ729122–KJ729124 (Table II). Rubin 1992), summarizing the samples of the parameter values, had values between 1.00 and 1.02 (average PSRF for parameter values (excluding NA Phylogenetic analyses and > 10.0) = 1.002). In addition to the above sequences, 836–841 bp Sequences were grouped based on the results of fragments of the COI gene corresponding to positions the phylogenetic analyses. Average evolutionary from 5885 to 6720–6725 of the Paracentrotus divergence over sequence pairs between and within lividus mitochondrion (Cantatore et al. 1989) were groups were estimated using MEGA software (ver- obtained for Gracilechinus multidentatus from Gen- sion 5; Tamura et al. 2011). Analyses were con- Bank (Ward et al. 2008; voucher Ech179, GenBank ducted using the Kimura 2-parameter model of access EU869930; voucher Ech181, GenBank access sequence evolution (Kimura 1980). The rate vari- EU869928; voucher Ech182, GenBank access ation among sites was modelled with a gamma EU869927). The species Sterechinus neumayeri distribution (shape parameter = 0.7). (Meissner, 1900) was used as an outgroup, with the sequence obtained from GenBank (Lee & Chung Results 2003; GenBank access AY275548; Table II). All sequences were aligned by eye. Owing to the shorter Taxonomy length of the COI fragment of Echinus melo (Emv75) Family Echinidae and G. euryporus (EspStock1964), only the 657 bp Genus Gracilechinus Fell & Pawson, 1966 fragment of sequences common to all species was Gracilechinus Fell & Pawson, 1966: U431. used for analyses. Phylogenetic trees were generated using the maximum parsimony (MP) method and the Type species: Gracilechinus gracilis (A. Agassiz, 1869) PAUP* program (version 4.0b10; Swofford 1999), by original designation. the maximum likelihood (ML) method using the RAxML software (Stamatakis 2006) in the raxmlGUI Species graphic user interface (Silvestro & Michalak 2012) and by the method of Bayesian inference (BI) using Assigned species: G. acutus (Lamarck, 1816), MrBayes version 3.1.2 (Huelsenbeck & Ronquist G. affinis (Mortensen, 1903), G. alexandri (Danielssen 2001). For the MP all sites were treated as unordered & Koren, 1883), G. atlanticus (Mortensen, 1903), characters with equal weight using the heuristic search G. elegans (Düben & Koren, 1844), G. euryporus method with random addition of sequences and (Clark, 1912), G. gracilis (Agassiz, 1869), G. lucidus branch swapping (TBR branch swapping algorithm). (Döderlein, 1885), G. multidentatus (Clark, 1925), G. ML and BI analyses were carried out using the GTR stenoporus (Mortensen, 1943).

Downloaded by [Kirill Minin] at 07:19 02 October 2014 +G substitution model of sequence evolution that was selected by the Akaike information criterion (Akaike Gracilechinus euryporus (H. L. Clark, 1912) 1974) of 40 models in jModeltest 2 (Darriba et al. (Figures 1, 3 and 5) 2012). This model was applied to a data set parti- tioned by codon positions for both ML and BI Echinus norvegicus. A. Agassiz 1881: 117 pars (non E. analyses. The robustness of each branch was evalu- norvegicus Düben & Koren, 1844). ated by the bootstrapping method with 500 replica- Pseudechinus euryporus H. L. Clark, 1912: 266–267, tions for both MP and ML analyses. pl. 93, figures 2–3, pl. 109, figures 4–6; Lambert & For Bayesian inference, two Markov Chain Monte Thiéry 1914: 244. Carlo (MCMC) runs of four chains each were Echinus euryporus. H. L. Clark 1925: 112; Mortensen performed starting from different random trees for a 1943:90–91, pl. LV, figures 23–25; Smith & Kroh total of 500,000 generations with sampling frequency 2011; Kroh 2012b. every 100 generations and diagnostic frequency every 50,000 generations. The first 3000 trees with convergence diagnostic values (average standard Material examined deviation of split frequencies) > 0.01 were discarded RV Akademik Kurchatov, St. 3615, 46°04′S, 83°50′W, (burnin = 3000), and the remaining sampled trees unnamed seamount in the Southeast Pacific, with convergence diagnostic values < 0.01 were used 220–240 m, six partly broken and 18 complete Pacific Gracilechinus: Morphology and evolutionary history 5

Figure 1. Ambulacra of Gracilechinus euryporus,RVAkademik Kurchatov, St. 3615, TD = 91 mm (A,B) and Gracilechinus multidentatus,RV Dmitry Mendeleev, St. 1268, TD = 78 mm (C,D). A,C, aboral part; B,D, oral part. Scale for A and B at the left, for C and D at the right. Abbreviations used to mark morphological structures: pt, primary tubercle; st, secondary tubercle.

specimens; RV Professor Shtokman, St. 1964, 24° occur on every second to sixth ambulacral plate. The 56.3′–24°58.2′S, 88°32.6′–88°29.8′W, Sala y Gomez rest of the primary tubercles are similar in size Ridge, 580–564 m, two complete specimens. to secondaries. Larger primary tubercles are often more numerous in one of two adjacent ambulacral rows. In other specimens the difference in size Description among primary ambulacral tubercles is less pro- The material studied comprises six partly broken nounced (Figure 1A,B). and 20 complete specimens with TD ranging from Each interambulacral plate has a primary tubercle,

Downloaded by [Kirill Minin] at 07:19 02 October 2014 60.5 to 95.5 mm. Test high, hemispherical in all these tubercles are arranged in regular longitudinal specimens. TH ranges from 0.56 to 0.74 TD, Ap series. In most specimens these tubercles aborally from 0.23 to 0.32 TD. Peristome is more or less show some irregularity in size. sunken, Per varies from 0.25 to 0.3 TD. The Secondary tubercles on the aboral side of the test smallest specimen has 25 ambulacral and 16 inter- are well developed, but do not cover most of the test ambulacral plates, the largest 30 ambulacral and 20 surface. interambulacral plates. I/TD ratio varies from 0.20 In 12 of 26 specimens small, club-shaped spines to 0.27, A/TD from 0.31 to 0.46 (Table III). present on buccal plates. No correlation between All specimens with primary tubercle on almost presence of spines and TD, the smallest and one of every ambulacral plate, only a few plates lack a the largest (RV Akademik Kurchatov, St. 3615, TD = primary tubercle. Number of ambulacral plates 91.5 mm) specimens possessing buccal spines. without primary tubercle is up to seven per ambu- Spines on periproctal plates only in five specimens. lacral row in some specimens, but the proportion of On one specimen (RV Professor Shtokman, St. 1964, such plates never exceeds 25%. Ambulacral primary TD = 77 mm) periproctal spines numerous, other tubercles in irregular longitudinal series. They vary specimens have only one or two spines. Periproctal in size, especially on the aboral side of the test. In spines present mainly in large specimens, the smal- some specimens large aboral primary tubercles, lest of them with TD = 77 mm. Four specimens gradually decreasing in size towards apical system, possess both buccal and periproctal spines. 6 K. V. Minin et al.

Figure 2. Ambulacra of Gracilechinus lucidus,RVVityaz, St. 3768, TD = 32.5 mm (A), RV Odissey, Trawl 33, TD = 52 mm (B,C). B, oral part; A,C, aboral part. Scale for A at the left, for B and C in the centre. Abbreviations used to mark morphological structures: pt, primary tubercle; st, secondary tubercle.

Pore zones usually do not widen near the peri- close to the end of the spine; thorns and protuber- stome (Figure 1B); in some specimens they are ances disappear, streaks become less prominent and slightly widened. also often disappear from the central part of the Tridentate pedicellariae in two forms, small rostrate- wedge. In the present paper this type of spine surface like and large. However, some transitions between microsculpture is identified as type ‘a’. the two general forms are also found. The large form Test pale brown or purple, creamy in some speci- is characterized by slender, straight valves meeting at mens. Spines usually same colour as test, sometimes

Downloaded by [Kirill Minin] at 07:19 02 October 2014 their distal parts over 1/3 or less of their length. Base with a faint tinge of green at base. In one specimen part of the valve usually has distinct cutoffs at each (RV Professor Shtokman, St. 1964), tips of primary side. Blade is widened proximally and slightly con- spines violet. stricted in the middle (Figure 3A–D). It usually has a more or less developed inner calcareous meshwork. In two specimens (RV Professor Shtokman, St. Distribution 1964), valves of large tridentate pedicellariae have This species is known from three records: the type blades broader than those of the other specimens and locality off the Chilean coast (Clark 1912), a lack an inner calcareous meshwork (Figure 3C,D). seamount close to the type locality and the Sala-y- Valves of globiferous pedicellariae with tubular Gomez Ridge, the top of Yantarnaya Seamount blade and more or less rounded base without prom- (Mironov & Detinova 1990; Parin et al. 1997). The inent corners. Lateral teeth are usually abundant, 1–6 depth range is from 220 to 580 m. at each side of the valve (Figure 4A,B). Triphyllous and ophicephalous pedicellariae typical of this genus. Large primary spines (more than 1.2 mm in Remarks diameter) have 15–31 wedges. Their surface is longit- According to The World Echinoidea Database udinally streaked with small thorns and protuberances (Kroh 2012b), this species is assigned to the genus (Figure 5A). This sculpturing becomes less developed Echinus. However, the ambulacral tuberculation Pacific Gracilechinus: Morphology and evolutionary history 7 Downloaded by [Kirill Minin] at 07:19 02 October 2014

Figure 3. Valves of tridentate pedicellariae in Paciﬁc species of Gracilechinus. A,B, Gracilechinus euryporus,RVAkademik Kurchatov, St. 3615, TD = 85 mm; C,D, G. euryporus,RVProfessor Shtokman, St. 1964, TD = 77 mm; E,F, Gracilechinus multidentatus, RV Dmitry Mendeleev, St. 1268, TD = 72 mm; G,H, Gracilechinus lucidus,RVVityaz, St. 3768, TD = 34.5 mm; I, G. lucidus,RVOdissey, Trawl 33, TD = 52 mm. Scales 300 µm. Abbreviations used to mark morphological structures: ba, valve base; bl, blade; bw, widening of the proximal part of the blade; co, cutoff in the base of the valve; im, inner calcareous meshwork.

pattern in this species is typical of Gracilechinus: in all diagnostic feature of this species, showed considerable examined specimens almost every ambulacral plate intraspecific variation. Pore zones widen near the has a primary tubercle. This tuberculation pattern peristome, as they do in the type specimen, in only a was previously recorded for the type specimen few of the specimens examined. Periproctal and buccal (Clark 1912). spines are recorded for the first time in this species. Some new details of morphology were revealed in The tridentate pedicellariae were never carefully the present study. Width of the pore zones, the main described before (Clark 1912;Mortensen1943). The 8 K. V. Minin et al. Downloaded by [Kirill Minin] at 07:19 02 October 2014

Figure 4. Valves of globiferous pedicellariae in Paciﬁc species of Gracilechinus.A,Gracilechinus euryporus,RVAkademik Kurchatov, St. 3615, TD = 87.5 mm; B, G. euryporus,RVProfessor Shtokman, St. 1964, TD = 77 mm; C, G. multidentatus,RVDmitry Mendeleev, St. 1268, TD = 72 mm; D, G. lucidus,RVVityaz, St. 3768, TD = 32 mm; E, G. lucidus,RVVityaz, St. 3768, TD = 34.5 mm; F, G. lucidus,RVOdissey, Trawl 33, TD = 52 mm. Scales 100 µm. Abbreviations used to mark morphological structures: a, angle of the valve base; ba, valve base; bl, blade; lt, lateral teeth; th, thorn on the valve base; tt, terminal tooth.

valves of the large tridentate pedicellariae are of a very Gracilechinus lucidus (Döderlein, 1885) peculiar shape. Cutoffs at the base of the valve, com- (Figures 2–5) bined with widening of the proximal part of the blade, make the tridentate pedicellariae of G. euryporus dif- Echinus norvegicus. A. Agassiz 1881: 117 pars (non E. ferent from those of other Gracilechinus species. norvegicus Düben & Koren, 1844). Pacific Gracilechinus: Morphology and evolutionary history 9

Figure 5. Surface microsculpture of primary spines at the base. A, types a and c; B, type b. A, Gracilechinus multidentatus (RV Dmitry Mendeleev, St. 1268); B, G. lucidus (Northwest Paciﬁc Ridge, RV Odissey, Trawl 33). Scales 100 µm.

Echinus lucidus Döderlein, 1885:97–98; Mortensen specimen has seven interambulacral and 10 ambu- 1903: 100, 161, 177, pl. XIX, figure 18; A. Agassiz lacral plates, the largest one 19 and 26, respectively. & H. L. Clark 1907: 127; H. L. Clark 1912: 167– I/TD ratio from 0.30 to 0.89; A/TD from 0.41 to 268, pl. 107, figures 1–3; H. L. Clark 1925: 113; 1.28 (Table III). Mortensen 1943:74–79, pl. III, figures 5–8, pl. IV, Specimens from different regions differ signifi- figure 2, pl. LIV, figures 14–17, 20–24, 28, 30. cantly in test size and shape. Specimens from the Echinus norwegicus. Yoshiwara 1906: pl. XII, figures East China Sea (ECS, RV Vityaz, St. 3768) are 5–6. characterized by smaller size, a low subconical test Pseudechinus lucidus. Lambert & Thiéry 1914: 244. and a sunken peristome. Specimens from the North- Gracilechinus lucidus. Smith & Kroh 2011; Kroh west Pacific Ridge (NPR, RV Odissey, Trawl 33 and 2012c. Trawl 40) are larger, with a higher, hemispherical test and the peristome not sunken. Specimens also differ in some other parameters of the test shape Material examined (Table III). RV Vityaz, St. 3768, 30°30.8′N, 129°19.4′E, Tokara Almost all ambulacral plates have a primary Strait, 818–824 m, 57 complete specimens; RV tubercle. In some specimens it can be absent on Odissey, Trawl 33, 37°55′N, 170°22′E, Odjik Sea- one to three plates per ambulacral row. Ambulacral mount, 1400 m, one complete specimen, Trawl 40, primary tubercles form an almost regular longitud- 34°48′N, 171°47′E, Kinmei Seamount, 820 m, four inal series in most of specimens. Only in the largest complete and one partly broken specimen. specimens (TD > 30 mm) are they arranged in a somewhat irregular series. The primary ambulacral Downloaded by [Kirill Minin] at 07:19 02 October 2014 tubercles of large specimens show conspicuous Description irregularity in size. This is most pronounced on the One partly broken and 62 complete specimens with aboral side of the test, where neighbouring primary TD varying from 8 to 64 mm. Test shape varies from tubercles usually differ in size (Figure 2A–C). In almost flattened to more or less high, with TH some specimens the average size of primary tubercles ranging from 0.37 to 0.69 TD. Ap from 0.21 to is larger in one of two adjacent ambulacral rows. 0.55 TD, Per from 0.25 to 0.47 TD. The smallest Each interambulacral plate has a primary tubercle.

Table III. Test shape parameters in Gracilechinus lucidus populations from the Northwest Pacific Ridge (NPR) and East China Sea (ECS) compared to other Pacific Gracilechinus species.

Species TD (mm) TH (%TD) Ap (%TD) Per (%TD) I/TD A/TD

G. lucidus (ECS) 8–34.5 37–54 26–55 31–47 0.37–0.89 0.56–1.28 G. lucidus (NPR) 27.5–64 56–69 21–27 25–38 0.30–0.48 0.41–0.73 G. euryporus 60.5–95.5 56–74 23–32 25–30 0.20–0.27 0.31–0.46 G. multidentatus 30–78 33–57 21–28 23–33 0.21–0.43 0.39–0.68 10 K. V. Minin et al.

Tubercles are arranged in regular longitudinal series Distribution and gradually increase in size towards the ambitus. Gracilechinus lucidus was known from depths of 180– Secondary tubercles on the aboral side of the test 1750 m along the east coast of Japan, from Hakodate are well developed, in one specimen from NPR (RV to Kagoshima (Döderlein 1885; Mortensen 1943; Odissey, Trawl 33) they cover almost the entire Clark 1912; Shigei 1986). Material from the Tokara surface of the test (Figure 2B,C). Strait (East China Sea, RV Vityaz, St. 3768) falls Buccal and periproctal spines are absent, pore zones within the range of previously known findings. Two do not widen near the peristome (Figure 2A,B). other stations on the Northwest Pacific Ridge (RV Tridentate pedicellariae are of two forms: small Odissey, Trawls 33 and 40) are northeast of that rostrate-like and large, with some transitions between region. The depths of both stations (1400 and them. The rostrate-like form is usual for this genus. 820 m, respectively) are within the known vertical The large form has slender, curved valves meeting at distribution range of the species. their distal parts over 1/5 of the length. Edges of the base of the valve are straight, and the proximal part of the blade is not conspicuously widened. The blade is Remarks slightly constricted in the middle (Figure 3G–I). As predicted by Mortensen (1943), this species Valves of the globiferous pedicellariae are very shows conspicuous morphological variability. Speci- variable in shape. Edges of the valve base are usually mens from the Northwest Pacific Ridge differ in test smooth and rounded (Figure 4D), but sometimes and spine morphology from the East China Sea angular. The angle is often placed in the middle of specimens. The former are larger in size, with fewer the lateral side of the base (Figure 4E), but in some coronal plates, higher test and smaller Per and Ap pedicellariae it is placed at the outer corner of the than the latter. The most marked differences base. The angle can be very strongly developed, in between the Northwest Pacific Ridge and East China which case it ends with a thorn. ECS specimens have Sea specimens are the shape of the peristome (not globiferous pedicellariae with the base shape of both sunken in the former and conspicuously sunken in types. In a single specimen from the NPR (RV the latter) and the primary spine surface microsculp- Odissey, Trawl 33) only one globiferous pedicellaria ture. The single specimen from Trawl 33, RV was found with thorns at the outer corners of the Odissey, differs from other specimens in having base (Figure 4F). The blades of the valves in all many more secondary tubercles. specimens are tubular, with 1–4 lateral teeth on each side. The valves of triphyllous and ophicephalous pedicellariae are typical of this genus. Gracilechinus multidentatus (H. L. Large primary spines (more than 1.2 mm in Clark, 1925) diameter) with 19–23 wedges. Microsculpture of (Figures 1, 3–5) their surface shows considerable variation. Besides type ‘a’ there are two other types of microsculpture. Echinus norvegicus. A. Agassiz 1881: 114 pars (non E. Some spines have wedges covered by thorns and norvegicus Düben & Koren, 1844). protuberances, with only the margins of each wedge Echinus multidentatus H. L. Clark, 1925: 115–116,

Downloaded by [Kirill Minin] at 07:19 02 October 2014 ‘ ’ longitudinally streaked (type b )(Figure 5B). pl. VI, figures 1–2; Mortensen 1943:88–90, pl. LV, Thorns and protuberances gradually decrease in figures 1–5, 8; Smith & Kroh 2011; Kroh 2012d. number and size towards the end of the spine. The Gracilechinus multidentatus. McKnight 1967: 295; ‘ ’ sculpturing pattern of type c is the same as in type McKnight 1968:94–99, figures 4–6; McKnight & ‘ ’ a , but streaks and protuberances do not decrease in Probert 1997: 505, 509. size and number towards the spine end. Sometimes sculpturing at the end of the spine is even more developed. Specimens of Graciechinus lucidus from Material examined different populations differ in microsculpture of the RV Dmitry Mendeleev, St. 1268, 43°00′S, 172°21′E, spine surface. Specimens from ECS (RV Vityaz,St. 850 m, one broken and 13 complete specimens. 3768) are characterized by type ‘a’. The NPR specimens (RV Odissey, Trawls 33 and 40) have sculpturing of type ‘b’. Spines with microsculpture of type ‘c’ were Description found in some specimens from both populations. The material comprises one broken and 13 complete The test and spines of ECS specimens (RV Vityaz, specimens with TD ranging from 30 to 78 mm. St. 3768) are white. Specimens from NPR (RV Test shape is very variable, from low, flattened to Odissey, Trawls 33 and 40) have pale cream test more or less high, hemispherical. TH ranges and spines. from 0.33 to 0.57 TD, Ap from 0.21 to 0.28 TD, Pacific Gracilechinus: Morphology and evolutionary history 11

Per from 0.23 to 0.33 TD. Peristome sunken, Distribution deeper in the largest specimens. The smallest speci- This species is known from off Tasmania and the men with 20 ambulacral and 13 interambulacral southern coast of Australia and from the waters plates, the largest with 32 and 19, respectively. A/TD surrounding New Zealand, from off the Kermadec varies from 0.39 to 0.68, I/TD from 0.21 to 0.43 Islands to the north Campbell Plateau and the (Table III). Chatham Rise in the east (McKnight 1967, 1968; Almost every ambulacral plate with a primary McKnight & Probert 1997; Mortensen 1943;Aus- tubercle. The tubercle was absent on several plates, tralian Museum Marine Invertebrate Collection especially close to the apical system, but the propor- 2013). Depth from 531 to 1326 m. Our record lies tion of these plates never exceeds 25%. Primary within this distribution range. ambulacral tubercles are arranged in an irregular series and differ considerably in size (Figure 1C,D). Remarks This irregularity is more pronounced aborally. As in This species was recently reassigned to the genus some Gracilechinus euryporus, primary tubercles are Echinus (Smith & Kroh 2011; Kroh & Hansson of two size classes, large and small, the small ones 2013). However, the presence of a primary tubercle almost similar in size to secondary tubercles. Small on almost every ambulacral plate indicates that this primary tubercles alternate with the large ones, the is a true Gracilechinus species. This agrees with the latter gradually increase in size towards the ambitus opinion of McKnight (1968). Similar to specimens (Figure 1C). Each interambulacral plate has a of McKnight from the Chatham Rise and the north primary tubercle, these tubercles are arranged in a Campbell Plateau, our specimens have ambulacral regular longitudinal series and gradually increasing primary tubercles of large and small size, almost in size towards the ambitus. equal to the size of secondary tubercles. McKnight Secondary tubercles are well developed on the (1968) noted that the tuberculation of the type aboral side but do not cover most of the test surface. specimen also shows this pattern. The small primary Buccal and periproctal spines are absent in all tubercles can be easily confused with the secondaries, specimens. Pore zones do not widen close to the which may have resulted in the erroneous opinion peristome (Figure 1D). that the primary tubercle in this species develops on Tridentate pedicellariae occur in two usual forms, every third or second ambulacral plate (Clark 1925; small rostrate-like and large, with some transitions Mortensen 1943; Smith & Kroh 2011). Other details between them. Valves of the rostrate-like form are of the test morphology in our specimens agree well typical of this genus. Large tridentate pedicellariae of with existing descriptions (Clark 1925; McKnight this species are similar to those of G. lucidus. They 1968). The pedicellariae are very similar to those have slender, curved valves meeting at their distal described and illustrated by Mortensen (1943: parts over 1/4 to 1/6 of their length. Edges of the plate LV). base of the valve straight, proximal part of the blade not conspicuously widened. Blade slightly con- Phylogenetic relationships among species of the stricted in the middle (Figure 3E,F). In one speci- genera Echinus and Gracilechinus Downloaded by [Kirill Minin] at 07:19 02 October 2014 men (TD = 75.5 mm) tridentate pedicellariae differ from those in other specimens. Their valves are The length of the COI gene fragment used for the broader and meet at their widened distal parts over analyses was 657 bp. Of these, 124 sites were 1/3 of their length. variable and 63 were parsimony-informative. The Valves of the globiferous pedicellariae usually have Maximum Parsimony (MP) analysis revealed 7 equally parsimonious trees with a consistency index smooth and rounded base. In few pedicellariae outer (CI) of 0.8654, a homoplasy index (HI) of 0.1346, a corners of the base are angular. Blade of the valve is retention index (RI) of 0.8871 and a rescaled tubular, with 1–6 lateral teeth on each side (Figure consistency index (RC) of 0.7677. The tree topology 4C). Valves of triphyllous and ophicephalous pedi- shows two major clades, one consisting of two cellariae are of a usual shape for this genus. Echinus species (Echinus esculentus and E. melo) and Large primary spines (more than 1.2 mm in the other gathering species of the genus Gracilechi- – diameter) with 25 46 wedges. Microsculpture of nus. Each clade shows high bootstrap support (100 ‘ ’ ‘ ’ their surface of type a or c (Figure 5A) described and 98 respectively). Sterechinus neumayeri emerges above. as the most distantly related species to either of the Tests and spines in all specimens straw-coloured. two clades. Inside the Gracilechinus clade, individuals In some specimens aboral primary spines with a faint of Gracilechinus alexandri cluster separately from tinge of green at the base. the group that comprises other species of the 12 K. V. Minin et al.

Figure 6. The Bayesian inference (BI) tree of COI sequences in Echinus and Gracilechinus species; outgroup species Sterechinus neumayeri. BI performed using Mr. Bayes program version 3.1.2 (Huelsenbeck & Ronquist 2001). All codon positions were treated separately and the GTR+G model of nucleotide substitution was used. Groups of sequences used for estimation of average evolutionary divergence are marked (1 to 6).

genus – G. elegans, G. lucidus, G. multidentatus, Discussion G. euryporus, G. affinis, G. stenoporus and G. atlanti- Morphological comparison of the Pacific species of the cus (bootstrap support values 100 and 96, respect- genus Gracilechinus ively). In some of the MP trees the latter three All three Pacific species of the genus Gracilechinus species cluster together, although with low bootstrap have similar morphological features. Gracilechinus supports. The phylogenetic relationships among euryporus, G. multidentatus and G. lucidus have well other species of this group are not resolved. developed secondary tuberculation on the aboral Maximum likelihood (ML) and Bayesian analyses side of the test, ambulacral primary tubercles of (BI, Figure 6) resulted in trees with a similar topology, irregular size, tridentate pedicellariae with slender distinguishing all the above principal clades. The valves and globiferous pedicellariae with usually bootstrap values for these clades revealed by ML and Downloaded by [Kirill Minin] at 07:19 02 October 2014 more than one lateral tooth on each side of the MP are similar. At the same time, Bayesian posterior valve. All three species are characterized by the pale probabilities are in general higher (Figure 7). The colour of the test and spines. Gracilechinus euryporus, support value for the clade of G. stenoporus, G. multidentatus and G. lucidus share the last four G. atlanticus and G. affinis in the Bayesian inference features with their South Atlantic congeners tree is 72 (compared to 48 and 45 in MP and ML, (G. stenoporus and G. atlanticus) and the Northeast respectively). Atlantic species Gracilechinus affinis (for descriptions All sequences were arranged into six groups based see Mortensen 1943). This makes the Pacific Graci- on the topology of the phylogenetic trees (Figure 6). lechinus morphologically closer to the latter three The Kimura 2-parameter intergroup distances (0.007– species than to other congeners. 0.177) exceed the intragroup values (0.002–0.008), Some diagnostic features, such as the width of confirming the accuracy of the phylogenetic grouping pore zones near the peristome and the number of (Table IV). Average distance between sequences of lateral teeth on the valves of globiferous pedicellar- groups 5 and 6 (0.007) is the closest to the intragroup iae, showed significant intraspecific variability and distances. It agrees with the phylogram: these groups therefore are unsuitable for recognition of the cluster together and are weakly supported (Figure 7). species. At the same time, new features were found All other intergroup distances are distinctively higher distinguishing the Pacific species of Gracilechinus (0.020–0.177). (Table V). The main difference between G. euryporus Pacific Gracilechinus: Morphology and evolutionary history 13

Figure 7. The maximum likelihood (ML) tree of COI sequences in species of Echinus and Gracilechinus compared to their geographic distribution and morphological characters. Sterechinus neumayeri used as outgroup. Branch supports for principal nodes are from MP bootstrap (500 repetitions)/ML bootstrap (500 repetitions)/BI analyses. Abbreviations: NA, North Atlantic; SA, South Atlantic; P, Paciﬁc; E, primary tubercles on alternating ambulacral plates; G, primary tubercles on all ambulacral plates; B, tridentate pedicellariae with broad valves; S, tridentate pedicellariae with slender valves.

and other Pacific species is the peculiar shape of the two features: the shape of the valves of tridentate valves of large tridentate pedicellariae. In G. eur- pedicellariae and the presence of buccal and peri- yporus they are straight, with a cutoff in the basal part proctal spines in some specimens. (Figure 3). The other difference is a presence of Populations of G. lucidus from the East China periproctal and buccal spines in some specimens. Sea (ECS) and the Northwest Pacific Ridge (NPR) Gracilechinus multidentatus differs from its Pacific differ significantly and possibly represent different congeners in having more wedges in large primary species. Differences in the test shape parameters spines. Although the number of wedges shows some (TH/TD, Ap/TD, Per/TD, A/TD and I/TD, the intraspecific variation, these numbers do not overlap shape of the peristome) between the ECS and NPR among examined specimens of G. multidentatus and specimens are almost the same as between Pacific G. lucidus.InG. multidentatus and G. euryporus the Gracilechinus species (Table III). However, the test numbers overlap only slightly (Table V). Gracilechi- shape parameters alone are not enough for distin- nus multidentatus and G. lucidus appear to be the guishing species. Such morphological differences

Downloaded by [Kirill Minin] at 07:19 02 October 2014 closest pair of species, differing only in one feature, can indicate a high level of intraspecific variability the number of wedges on large primary spines. in G. lucidus, as in another Gracilechinus species, Gracilechinus euryporus differs from each of them in G. acutus (Mortensen 1943). The only discrete

Table IV. Estimates of evolutionary divergence over sequence pairs within and between groups.

Group 1 Group 2 Group 3 Group 4 Group 5 Group 6

Group 1 – 0.022 0.023 0.022 0.021 0.021 Group 2 0.170 0.008 ±0.003 0.012 0.013 0.012 0.012 Group 3 0.177 0.071 0.002 ±0.001 0.009 0.008 0.009 Group 4 0.175 0.085 0.050 0.006 ±0.003 0.005 0.005 Group 5 0.171 0.081 0.043 0.020 0.004 ±0.002 0.002 Group 6 0.170 0.078 0.044 0.020 0.007 0.004 ±0.002

Groups 1–6 correspond to those marked on Figure 6. Numbers of base substitutions per site from averaging over all sequence pairs between groups are shown below the diagonal. Standard errors are shown above the diagonal. Numbers of base substitutions per site from averaging over all sequence pairs within each group (± standard error) are shown on the diagonal. Analyses were conducted using the Kimura 2-parameter model. The rate variation among sites was modelled with a gamma distribution (shape parameter = 0.7). Codon positions 1st+2nd+3rd+Noncoding. All positions containing gaps and missing data were eliminated. For Group 1, comprising a single sequence of Sterechinus neumayeri, estimation of intragroup evolutionary distances was not possible. 14 K. V. Minin et al.

Table V. Morphological differences in Pacific species of the genus Gracilechinus.

Density of Lateral teeth on Wedges of large Buccal and secondary Valves of Valves of the valves of Primary (> 1.2 mm periproctal tubercles on the tridentate globiferous globiferous sipine diameter) spines aboral side of test pedicellariae pedicellariae pedicellariae surface primary spines

Character no. 1 2 3 4 5 6 7 G. lucidus a a,b a a,b 1–4a,b,c19–23 G. lucidus (ECS) a a a a, b 1–4a,c G. lucidus (NPR) a a, b a b 2–3b,c19–23 G. euryporus a, b a b a 1–6a15–31 G. multidentatus aa aa1–6a,c25–46

1. (a). Buccal and periproctal spines absent. (b). Buccal and periproctal spines present. 2. (a). Secondary tubercles well developed on the aboral side of the test, but do not densely cover the whole surface of the test. (b). Secondary tubercles very densely cover the whole aboral side of the test. 3. (a). Valves of tridentate pedicellariae slightly curved, without conspicuous cutoff in the basal part. (b). Valves of tridentate pedicellariae straight, with conspicuous cutoff in the basal part. 4. (a). Basal part of the valve of globiferous pedicellariae with smooth edges. (b). Basal part of the valve of globiferous pedicellariae angular, sometimes bears small thorns. 6. (a). Surface of the wedges of primary spines longitudinally streaked with small thorns and protuberances. Streaks and protuberances disappear close to the end of the spine. (b). Surface of the wedges of primary spines covered with protuberances, only edges longitudinally striated. Protuberances disappear close to the end of the spine. (c). Surface of the wedges of primary spines longitudinally streaked with small thorns and protuberances all the way to the end of the spine.

feature that distinguishes the ECS and NPR speci- by high bootstrap values and Bayesian posterior mens is primary spine surface microsculpture. The probabilities (Figure 7, ‘G+B’ and ‘G+S’). microsculpture of the type ‘b’ characteristic of the The assignment of Gracilechinus lucidus, G. multi- NPR specimens was never recorded in G. euryporus, dentatus and G. euryporus to the genus Gracilechinus is G. multidentatus or the ECS specimens of G. lucidus. confirmed by both morphological and molecular The intraspecific variability and the taxonomic sig- data. On the COI tree all the three species appear nificance of this feature in Gracilechinus is unknown. within one clade comprising slender-valved Gracile- Examination of other species of this genus and chinus species (Figure 7, ‘G+S’). Inside this clade, additional NPR specimens are required to under- phylogenetic relationships among the species are not stand if the NPR and ECS G. lucidus are separate well resolved. Only individuals of G. multidentatus species. and one other species, G. elegans, form well-differ- entiated clades. At the same time, both G. lucidus and G. euryporus have peculiar morphological fea- Phylogenetic relationships among Pacific species of the tures and can be distinguished from other Gracile- genus Gracilechinus and their evolutionary history chinus species. Phylogenetic relationships among The genus Gracilechinus differs from Echinus in the the three other species of this clade (G. affinis, pattern of ambulacral tuberculation. Species of Ech- G. stenoporus and G. atlanticus), clearly distinguished inus have a primary tubercle on every second or third morphologically (Mortensen 1903; Mortensen

Downloaded by [Kirill Minin] at 07:19 02 October 2014 (or second and third) ambulacral plate. In the genus 1943), also remain unresolved by the COI tree Gracilechinus almost all ambulacral plates bear a (Figure 7). The absence of significant molecular primary tubercle. Species of this genus can be differences in the examined fragment of the COI divided according to the shape of their tridentate gene (Figure 6; Table IV) in species of this clade pedicellariae. The tridentate pedicellariae of Gracile- might be explained by their relatively young age. The chinus alexandri have broad, leaf-shaped valves. Other relatively small lengths of the COI fragment used for species of the genus are characterized by slender the analysis (657 bp) can be another explanation. valves of the tridentate pedicellariae (Mortensen The distribution of species of the two major clades 1943). of Gracilechinus is different. The only species of the The COI tree including eight species of Gracile- broad-valved Gracilechinus clade, G. alexandri, chinus and two of Echinus agrees in general with occurs in the North Atlantic. The clade of the morphological data and provides evidence for the slender-valved Gracilechinus includes both Atlantic validity of these genera. Species of each genus form and Pacific species (Figure 7, ‘G+S’). Gracilechinus distinct, well-supported clades. Inside the Gracilechi- elegans and G. affinis occur in the North Atlantic. nus clade, slender-valved species form a distinct Two more species, G. stenoporus and G. atlanticus, group, separated from the species with broad valves are distributed in the southeastern and central (Gracilechinus alexandri). The clades of slender- Atlantic. Other species from this clade occur in the valved and broad-valved species are corroborated Pacific Ocean. Pacific Gracilechinus: Morphology and evolutionary history 15

According to our data, there are at least two References lineages of Gracilechinus species in the North Atlantic Agassiz A. 1881. Report on the Echinoidea dredged by H.M.S. characterized by different types of tridentate pedi- Challenger, during the years 1873–1876. In: Thomson cellariae valves. Only one lineage penetrated the CW, editor. Report on the Scientific Results of the Voyage of South Atlantic and the Pacific. All South Atlantic H.M.S. Challenger. Zoology. Volume 3. Edinburgh: Neill, – and Pacific species occur at bathyal depths and they p1 321. Agassiz A, Clark HL. 1907. Hawaiian and other Pacific Echini. apparently descend from a common bathyal The Cidaridæ. Memoirs of the Museum of Comparative ancestor. Zoology 34(1):1–42. Tyler et al. (1995) suggested separate colonization Akaike H. 1974. A new look at the statistical model identification. of the western and eastern Pacific by Atlantic IEEE Transactions on Automatic Control 19(6):716–23. Gracilechinus (considered by the authors as Echinus) Australian Museum Marine Invertebrate Collection. 2013. http:// collections.ala.org.au/public/show/co113/ (accessed 3 April via three possible routes: one through the South 2013). Atlantic eastward, second through the South Cantatore P, Roberti M, Rainaldi G, Gadaleta MN, Saccone C. Atlantic westward, or alternatively through the 1989. The complete nucleotide sequence, gene organization, West Atlantic and the Straits of Panama (more and genetic code of the mitochondrial genome of Paracentrotus – details in Introduction). However, both the morpho- lividus. Journal of Biological Chemistry 264:10965 75. Clark HL. 1912. Hawaiian and other Pacific Echini. Memoirs of logical and molecular similarity of the three Pacific the Museum of Comparative Zoology 34(4):205–383. Gracilechinus species suggest that colonization Clark HL. 1925. A Catalogue of the Recent Sea Urchins occurred once via a single route. This route appar- (Echinoidea) in the British Museum (Natural History). Lon- ently passed through the South Atlantic because the don: Trustees of the British Museum, Oxford University Press. South Atlantic G. atlanticus and G. stenoporus and the 250 pages. Darriba D, Taboada GL, Doallo R, Posada D. 2012. jModelTest Northeast Atlantic G. affinis are morphologically 2: More models, new heuristics and parallel computing. Nature closest to the Pacific Gracilechinus species. Further Methods 9:772. dispersal from the South Atlantic most likely Döderlein L. 1885. Seeigel von Japan und den Liu-Kiu-Inseln. occurred in the eastward direction. Most of the Archiv für Naturgeschichte 51(1):73–112. deep-water currents, capable of transporting the Fell HB, Pawson DL. 1966. Echinoida. In: Moore RC, editor. larvae of South Atlantic bathyal Gracilechinus to the Treatise on Invertebrate Paleontology. Part U, Echinodermata 3. Lawrence, Kansas: Geological Society of America and Pacific Ocean, flow eastwards (Stramma & England University of Kansas Press, p 431–36. 1999). Colonization of the Pacific was apparently Folmer O, Black M, Hoeh W, Lutz R, Vrijenhoek R. 1994. DNA followed by dispersal across the Pacific Ocean. primers for amplification of mitochondrial cytochrome C oxidase subunit I from diverse metazoan invertebrates. – Acknowledgements Molecular Marine Biology and Biotechnology 3:294 99. Gelman A, Rubin DB. 1992. Inference from iterative simulation Molecular analysis was partially carried out within using multiple sequences. Statistical Science 7:457–72. ‘ Huelsenbeck JP, Ronquist F. 2001. MRBAYES: Bayesian infer- the project KMBSU ( Bathyal Fauna of Sea Urch- – ’ ence of phylogenetic trees. Bioinformatics 17:754 55. ins ) of the BOLD. We thank Erin Corstorphine, Ivanova NV, DeWaard JR, Hajibabaei M, Hebert PDN. 2005. Nikolay Poyarkov, Alex Borisenko, Natalia Ivanova Protocols for high volume DNA barcoding. Draft submission and Paul D. N. Hebert for collaboration. The to: DNA working group Consortium for the Barcode of Life. research was partially supported by the NSERC, http://barcoding.si.edu/PDF/Protocols_for_High_Volume_DNA_ Downloaded by [Kirill Minin] at 07:19 02 October 2014 Genome Canada, Gordon and Betty Moore Founda- Barcode_Analysis.pdf (accessed 23 February 2014). Ivanova NV, DeWaard JR, Hebert PDN. 2006. An inexpensive, tion, Canada Foundation for Innovation and Ontario automation-friendly protocol for recovering high-quality DNA. Innovation Trust, Canadian Centre for DNA Bar- Molecular Ecology Notes 6:998–1002. coding (BIO, ON, Canada) and funded through Kimura M. 1980. A simple method for estimating evolutionary rate Genome Canada and the Ontario Genomics Insti- of base substitutions through comparative studies of nucleotide – tute (2008-OGI-ICI-03). The authors thank Endre sequences. Journal of Molecular Evolution 16:111 20. Kroh A. 2012a. Gracilechinus Fell & Pawson, in Moore, 1966. In: Willassen and Jon A. Kongsrud of the Zoological Kroh A, Mooi R, editors. World Echinoidea Database. http:// Museum, University of Bergen, for the donation of www.marinespecies.org/Echinoidea/aphia.php?p=taxdetails&id= tissue samples of some Gracilechinus species. The 411391/ (accessed 1 April 2013). authors thank Alexander Mironov, Andrey Gebruk Kroh A. 2012b. Echinus euryporus H.L. Clark, 1912. In: Kroh A, and David Pawson for invaluable comments on the Mooi R, editors. World Echinoidea Database. http://www. marinespecies.org/echinoidea../aphia.php?p=taxdetails&id=513 manuscript and editing the style of the English. 248/ (accessed 1 April 2013). Kroh A. 2012c. Gracilechinus lucidus (Döderlein, 1885). In: Kroh Funding A, Mooi R, editors. World Echinoidea Database. http://www. marinespecies.org/echinoidea../aphia.php?p=taxdetails&id=532 The present study was supported by the Russian Founda- 036/ (accessed 1 April 2013). tion for Basic Research (grant numbers 12-05-31525, Kroh A. 2012d. Echinus multidentatus H.L. Clark, 1925. In: Kroh 12-05-33049 and 13-04-01332). A, Mooi R, editors. World Echinoidea Database. http://www. 16 K. V. Minin et al.

marinespecies.org/echinoidea../aphia.php?p=taxdetails&id=513 Shigei M. 1986. The Sea Urchins of Sagami Bay. Tokyo: 250/ (accessed 1 April 2013). Maruzen. 204 pages. Kroh A, Hansson H. 2013. Echinus Linnaeus, 1758. In: Kroh A, Silvestro D, Michalak I. 2012. RaxmlGUI: A graphical front-end Mooi R, editors. World Echinoidea Database. http://www. for RAxML. Organisms Diversity & Evolution 12:335–37. marinespecies.org/echinoidea/aphia.php?p=taxdetails&id=1233 Smith AB, Kroh A. 2011. The Echinoid Directory. WorldWide 86/ (accessed 1 April 2013). Web electronic publication. http://www.nhm.ac.uk/research- Lambert J, Thiéry P. 1914. Essai de Nomenclature Raisonnée des curation/projects/echinoid-directory/ (accessed 4 March 2013). Echinides. Chaumont, France: L. Ferrière. 294 pages. Stamatakis A. 2006. RAxML-VI-HPC: Maximum likelihood- Lee YH, Chung WS. 2003. Sterechinus neumayeri COI sequence. based phylogenetic analyses with thousands of taxa and mixed http://www.ncbi.nlm.nih.gov/nuccore/33439469?report=genbank/ models. Bioinformatics 22(21):2688–90. (accessed 1 April 2012). Stramma L, England M. 1999. On the water masses and mean McKnight DG. 1967. Additions to the echinoderm fauna of the circulation of the South Atlantic Ocean. Journal of Geophysical Chatham Rise. New Zealand Journal of Marine and Freshwater Research 104(C9):20863–83. Research 1:291–313. Swofford DL. 1999. PAUP*. Phylogenetic Analysis Using Parsi- McKnight DG. 1968. Additions to the echinoid fauna of New mony (*and Other Methods). Version 4.0b3a. Sunderland, Zealand. New Zealand Journal of Marine and Freshwater Massachusetts: Sinauer Associates. Computer program. Research 2:90–110. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S. McKnight DG, Probert PK. 1997. Epibenthic communities on 2011. MEGA5: Molecular evolutionary genetics analysis using the Chatham Rise, New Zealand. New Zealand Journal of maximum likelihood, evolutionary distance, and maximum Marine and Freshwater Research 31:505–13. parsimony methods. Molecular Biology and Evolution Mironov AN, Detinova NN. 1990. Bottom fauna of the Nazca and 28:2731–39. Sala-y-Gomez ridges. In: Mironov AN, Rudjakov JuA, editors. Tyler PA, Young CM, Serafy DK. 1995. Distribution, diet and Plankton and Benthos from the Nazca and Sala-y-Gomez reproduction in the genus Echinus: Evidence for recent diver- Submarine Ridges. Moscow: Nauka, p 269–278. (in Russian) sification? In: Emson RH, Smith A, Campbell AC, editors. Mortensen T. 1903. Echinoidea. The Danish Ingolf-Expedition. Echinoderm Research 1995. Proceedings of the Fourth Euro- Volume 4. No 1. Copenhagen: Bianco Luno. 193 pages. pean Echinoderms Colloquium, London, UK, 10–13 April Mortensen T. 1942. New Echinoidea (Camarodonta). Preliminary 1995. Rotterdam: Balkema, p 325–34. notice. Videnskabelige Meddelelser fra Dansk Naturhistorisk Ward RD, Holmes BH, O’Hara TD. 2008. DNA barcoding Forening 106:225–32. discriminates echinoderm species. Molecular Ecology Mortensen T. 1943. A Monograph of the Echinoidea. Volume III. 3. Resources 8:1202–11. Camarodonta. II. Echinidae, Strongylocentrotidae, Parasalenii- Yoshiwara S. 1906. Japanese Echini. Zoological Magazine, Tokyo dae, Echinometridae. Copenhagen: C. A. Reitzel. 446 pages. 18: plate XII:5–6. Parin NV, Mironov AN, Nesis KN. 1997. Biology of the Nazca and Sala y Gomez submarine ridges, an outpost of the Indo- Editorial responsibility: Torsten Struck West Pacific fauna in the eastern Pacific Ocean. Composition and distribution of the fauna, its communities, and history. Advances in Marine Biology 32:145–242. Downloaded by [Kirill Minin] at 07:19 02 October 2014