This Article Appeared in a Journal Published by Elsevier. the Attached

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright Author's personal copy

Molecular Phylogenetics and Evolution 57 (2010) 687–702

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution

journal homepage: www.elsevier.com/locate/ympev

Genetic variation and phylogeny of the cosmopolitan marine genus Tubiﬁcoides (Annelida: Clitellata: Naididae: Tubiﬁcinae)

a,b, c a Sebastian Kvist ⇑, Indra Neil Sarkar , Christer Erséus a Department of Zoology, University of Gothenburg, Box 463, SE-405 30 Göteborg, Sweden b Richard Gilder Graduate School, American Museum of Natural History, Central Park West at 79th Street, New York, NY 10024, USA c Center for Clinical and Translational Science, Department of Microbiology & Molecular Genetics and Department of Computer Science, University of Vermont, 89 Beaumont Avenue, Given Courtyard N309, Burlington VT 05405, USA article info a b s t r a c t

Article history: Prior attempts to resolve the phylogenetic relationships of the cosmopolitan, marine clitellate genus Received 15 February 2010 Tubificoides, using only morphology, resulted in unresolved trees. In this study, three mitochondrial Revised 16 August 2010 and three nuclear loci (5912 aligned sites) were analyzed, representing 14 morphologically separate Accepted 17 August 2010 species. Genetic distances within and between these forms on the basis of the mitochondrial genes Available online 27 August 2010 (COI, 16S and 12S) revealed that 18 distinct mitochondrial lineages were represented in the data set. After analyzing also nuclear data (28S, 18S and ITS) we conclude that 17 separately evolving lineages Keywords: (i.e., phylogenetic species) were represented, including three new, cryptic species closely related to T. Genetic variation pseudogaster, T. amplivasatus and T. insularis, respectively. Special emphasis was put on the DNA barcod- Haplotype diversity Species delimitation ing gene (COI), which was subject to haplotype diversity analysis and, for four species, diagnostic posi- Geographic distribution tion (as determined by the Characteristic Attribute Organization System [CAOS]) screening. Typically, Phylogeny the intralineage variation was 1–2 orders of magnitude smaller than the interlineage divergence, mak- Combined analysis ing COI useful for identification of species within Tubificoides. The genetic data corroborate that many of DNA barcoding the morphospecies are coherent but widely distributed metapopulations. Monophyly of the genus is Tubificoides supported and the evolutionary history of parts of the genus is revealed by phylogenetic analysis of Tubificinae the combined data set. A northern hemisphere origin of the genus is suggested, and most of the widely Naididae distributed species are members of one particular clade. Two morphological characters previously emphasized in Tubificoides taxonomy (hair chaetae and cuticular papillation) were optimized on the phylogenetic tree, revealing considerable homoplasy, belying the utility of these features as phylogenetic markers. Ó 2010 Elsevier Inc. All rights reserved.

1. Introduction smaller areas. Most species are euhaline, others are tolerant to fluc- tuations in salinity or prefer oligohaline conditions, and many of Naididae (including the former Tubificidae; see ICZN, 2007; them are favored by organic pollution and sulfide-rich sediments Erséus et al., 2008) is a large group of aquatic clitellates, compris- (e.g. Giere et al., 1988; Dubilier et al., 1995; Lerberg et al., 2000). ing about 1000 described species (Erséus, 2005). Currently divided By and large, the genus can be found in most coastal or deep-sea into eight subfamilies (Tubificinae, Rhyacodrilinae, Phallodrilinae, sediments at least in the northern hemisphere (e.g. Cook, 1969; Naidinae, Telmatodrilinae, Limnodriloidinae, Pristininae and Opist- Erséus, 1975; Brinkhurst and Baker, 1979; Brinkhurst, 1985; Kvist ocystinae [Erséus et al., 2008, 2010]), representatives of the family et al., 2008), some species being vastly abundant (e.g. Erséus and are found in marine as well as freshwater habitats. Tubificoides Las- Diaz, 1989; Harrel, 2004). However, all taxa examined in this tockin, 1937 is a species-rich genus of Tubificinae consisting of 57 study, except T. amplivasatus (Erséus, 1975), appear restricted to nominal species (unpublished compilation) including two recently shallow waters. described taxa (Kvist et al., 2008). Some species of Tubificoides The early taxonomy of current Tubificoides species was confusing. show a worldwide distribution while others appear endemic to Disagreements concerning the generic status of some species, mainly based on chaetal features, resulted in these being transferred

Corresponding author at: Richard Gilder Graduate School, American Museum of between other genera; e.g. Tubifex Lamarck, 1816, Clitellio Savigny, ⇑ Natural History, Central Park West at 79th Street, New York, NY 10024, USA. 1820 and Limnodrilus Claparède, 1862 (Dahl, 1960; Brinkhurst, E-mail address: [email protected] (S. Kvist). 1962, 1965; see Brinkhurst and Baker, 1979; Baker, 1980). These

688 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 species were later included in (and several new species were as- imen) and the Caspian Sea (one specimen) (Fig. 1). In most cases, signed to) Peloscolex Leidy, 1851 (e.g., Hrabeˇ, 1966; Cook, 1969), a the specimens were collected by Kvist and/or Erséus, otherwise genus primarily defined by its characteristic cuticular papillation. by colleagues noted at the bottom of Table 1. Intertidal and subtid- However, Holmquist (1978, 1979) regarded Peloscolex as an artificial al samples were sieved by elutriation using a mesh size of 125– assemblage and re-established Tubificoides, placing all marine spe- 300 lm, and the specimens were in toto preserved in 80% ethanol. cies of Peloscolex within it; except for P. benedii (d’Udekem, 1855), Subsequently, the worms were cut in two parts, and the posterior which was transferred to Edukemius Holmquist, 1978. These species part was transferred to 95% ethanol to be used for DNA extraction, show great resemblance in the morphology of the genitalia, which while the anterior part was stained in alcoholic paracarmine, dehy- led Brinkhurst and Baker (1979) to transfer 13 additional taxa drated in an ethanol–xylene series, and mounted whole in Canada (including E. benedii) to Tubificoides. Members of Tubificoides sensu balsam to serve as a voucher. lato are recognized by the principal morphology of their male genitalia; the vas deferens always enters the atrium subapically, more 2.3. DNA extraction, amplification and purification or less opposite to where the stalk of the prostate gland enters the same structure. The different species are then largely delimited by Total genomic DNA was extracted using DNeasy Blood and Tis- unique combinations of details in the arrangement and distribution sue KitsÓ (Qiagen Ltd.) according to the manufacturer’s protocol. of chaetae, the presence or absence of cuticular papillae and the The six loci were amplified using PuReTaq Ready-To-Go™ PCR shape and size of the penis sheaths. These features give a mosaic pat- beads (GE Healthcare) and 1 ll of each primer (Table 2). For the tern to any morphological character matrix of Tubificoides and for- PCR-reactions, the DNA extract was used in quantities of 2–4 ll mal cladistic analyses of such data sets have so far failed to and 19–21 ll of sterilized water was added to each amplification, produce any resolved phylogenetic trees (Erséus, unpublished). giving a total sample size of 25 ll. Either a PTC-100Ó (MJ Research The reported wide range of some Tubificoides species (e.g. Baker, Inc.) or an Eppendorf MastercyclerÒ was used. The thermal profile 1984; Brinkhurst, 1986; Helgason and Erséus, 1987) is suggestive was gene dependent and varied as follows: an initial step of 5 min of their significant impact on benthic coastal ecosystems in the denaturation at 95 °C (for all samples) followed by 30 (for 18S), 35 northern hemisphere. Despite this, little is known about the evolu- (for COI, 16S, 28S and ITS) or 43 (for 12S) cycles of denaturation at tionary relationships and the population genetic structure of the 95 °C (30 s for 16S, 18S and ITS; 40 s for COI, 28S and 12S), anneal- various species. In a previous study, Erséus and Kvist (2007) exam- ing at 45 °C (for COI, 16S [35 cycles] and 12S [43 cycles]), 50 °C (for ined the variation in the cytochrome c oxidase subunit I gene (COI) ITS [35 cycles]), 52 °C (for 28S [35 cycles]) or 54 °C (for 18S [30 cy- of four Scandinavian species of Tubificoides to evaluate its applica- cles]). The cycles were run for 30 s for 16S, 18S and ITS, 40 s for bility in DNA barcoding (see Hebert et al., 2003a,b, 2004; Hajiba- 28S, and 45 s for COI and 12S. Extension was performed at 72 °C baei et al., 2007). The interspecific divergence was found to be (60 s for COI, 16S, 12S and 28S; 90 s for 18S and ITS) and a final two orders of magnitude greater than the intraspecific variation extension step at 72 °C (8 min) was performed for all samples. for this gene. However, this factor (the ‘‘barcoding gap”) may have PCR products were purified using the E.Z.N.A™ Cycle-Pure kit been inflated by the limited size of the geographic area sampled (Omega Bio-Tek Inc.) following the manufacturer’s protocol. There- (Meyer and Paulay, 2005). after, sequencing was carried out in both directions and, for this The purpose of the present study was to assess the genetic varia- purpose, additional primers were used for ITS and 18S. COI se- tion in a larger sample of Tubificoides, including nuclear as well as quences of specimens CE539, 541, 551, 552, 1064, 1269, 1274 mitochondrial data, and re-examine COI using both distance and and 1724–1762 were obtained from Erséus and Kvist (2007). For character based methods. In particular, we evaluated the haplotype the new specimens, PCR products were sent to MacroGen Co. distribution among conspecific, widely separated populations. We Ltd., South Korea for sequencing. also conducted a phylogenetic analysis to estimate the general evolutionary history of the genus and, more specifically, to establish the 2.4. Sequence assembly and alignment number of separately evolving lineages within some nominal taxa with suspected cryptic speciation. Finally, we used the evolutionary Forward and reverse sequences were assembled using SeqMan tree to assess the phylogenetic signal of some of the most commonly II (DNAStarÒ Inc.). Both ends were trimmed in EditSeq (DNAStarÒ used morphological characters in the taxonomy of Tubificoides. Inc.) to delete primer sequences and manual editing was performed on obvious misreadings. Moreover, the Tubificoides COI se- 2. Material and methods quences EF675192–675228, used by Erséus and Kvist (2007), were revisited and several of these were extended in both ends; these 2.1. Genes and taxa studied were previously cut by the authors based on the rather low, yet va- lid, base-calling scores. To ensure the absence of stop codons in the Six loci were used for the combined phylogenetic analysis: par- COI sequence, these were translated using six frame translation on tial COI mtDNA, partial 16S and 12S ribosomal mtDNA, partial nu- the Baylor College of Medicine Search Launcher website (). Internal Transcribed Spacers 1 and 2 flanking the 5.8S ribosomal The consensus sequences were aligned using the web version of gene). Fourteen morphologically separated Tubificoides species MUSCLE ver. 3.7 (Edgar, 2004) on the European Bioinformatics were investigated and eight other tubificine naidids were used as Institute (EBI) server applying default settings and for ITS, which outgroup taxa. The trees were rooted with Limnodriloides anxius proved hard to correctly align, minor changes were further done (Limnodriloidinae). A complete list of specimens, sampling sites, by eye (TreeBASE study S10593; matrix M5892). and gene and voucher accession numbers can be found in Table 1. All vouchers are deposited in the Swedish Museum of Natural His- 2.5. Data analyses tory (SMNH), Stockholm. Four data sets were used: (i) 187 specimens for the COI varia- 2.2. Collection of new material tion analysis, (ii) 62 specimens for the genetic variation analysis of the 16S and ITS loci, and also for the phylogenetic analysis of Our material was assembled from 14 countries, representing the ITS locus, (iii) 34 specimens of Tubificoides and eight outgroup both sides of the Atlantic Ocean, the West Pacific Ocean (one spec- taxa for the phylogenetic analysis of each of the remaining five loci Author's personal copy

S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 689 ) Heterochaeta HM460278 HM460279 HM460280 HM460281 HM460282 HM460283 HM460284 HM460285 HM460273 HM460274 HM460286 HM460275 HM460276 HM460277 continued on next page ( –– –– HM459996 HM460036 –– –– HM459997 HM460037 HM459994 HM460034 –– HM459998 HM460038 HM459995 HM460035 –– –– –– HM459934 HM459935 HM459936 HM459937 HM459938 HM459939 HM459929 HM459940 HM459930 HM459941 HM459931 HM459932 HM459933 re noted as footnotes. a sets used for the genetic variation analyses. Sequences of CE196 The Netherlands; US, United States of America; ES, Spain; UK, United Kingdom; HK, COI 12S 16S 18S 28S ITS NL, Zeeland, Walcheren (8 m)NO, Bergen, near Biological Station,NO, Bergen, Espegrend near (5–10 m) Biological Station,NO, Espegrend Bergen, (5–10 near m) Biological Station, Espegrend (10–15 m) CE5430 CE5431UK, Plymouth, CE5433 SMNH108915 Millbrook Lake, PalmerUK, Point Plymouth, HM460112 SMNH108916 (intertidal) Millbrook Lake, PalmerUK, SMNH108917 Point Plymouth, – HM460113 (intertidal) Millbrook Lake, HM460114 PalmerUK, Point Wales, – (intertidal) S of HM459896 CE2691 Newport, St. Brides Wentloog CE2692 (5 m) SMNH108920 CE2693 HM460117 CE5320 SMNH108921 – – HM460118 SMNH108922 – SMNH108914 – HM460119 CE3390 HM460111 – – SMNH108923 – HM460120 – HM459897 – – – – – – – – FR, Atlantic coast, Concarneau (lowerFR, intertidal) Atlantic coast, Concarneau (lowerFR, intertidal) Atlantic coast, Roscoff, WSE, of Halland, harbour Kungsbacka, (lower Gottskär intertidal) (intertidal)SE, Halland, Kungsbacka, Gottskär (intertidal)SE, Bohuslän, CE2957 Koster area, W of CE2955 Ursholmen (38 m) SMNH108909 CE2956 SMNH108907 HM460106 HM460104 – SMNH108908 – HM460105 CE3102 – CE3109 CE3220 SMNH108910 HM460107 SMNH108911 SMNH108912 – – HM460108 HM460109 – – – – – – – – – – – – – – – – – – – – – Site (depth)SE, Bohuslän, Koster area, UrsholmenSE, (3 Bohuslän, m) Koster area, PersgrundenSE, (17 Bohuslän, m) Koster area, PersgrundenSE, (17 Bohuslän, m) Koster area, PersgrundenDK, (44–52 The m) Sound, Elsinore, EllekildeDK, have, The near Sound, shore Elsinore, (6 Ellekilde m) have (10 m) CE1744 CE539-2 CE1269-2 CE1638 SMNH85169 CE1269-3 SMNH85172 SMNH85158 SMNH85170 HM460080 SMNH108894 HM460079 HM460083 – HM460082 HM460081 – – – – CE1746 SMNH85160 Source HM460084 – – Voucher# HM459894 – – – GenBank accession # – – – – – – – – – – – – – – – SE, Bohuslän, Koster area, W of Ursholmen (38 m)CA, New Brunswick, Bocabec, PassamaquoddyCA, CE3221 Bay NB, (0.5 Blacks m) Harbour, Passamaquoddy Bay (0.5 m) SMNH108913 CE7282 HM460110 – SMNH108918 HM460115UK, CE7283 Wales, S – of Newport,UK, St. Wales, Brides S Wentloog of SMNH108919 (5 Newport,UK, m) St. Wales, Brides S HM460116 Wentloog of (5 Newport, – m) Peterstone Wentloog – (2–3 m) – CE3392 CE3397 CE3394 – SMNH108924 – SMNH108926 HM460121 SMNH108925 HM460123 – – HM460122 – – – – – – – – – – – – – – – – UK, Wales, S of Newport,UK, Peterstone Wales, Wentloog S (2–3 of m) Newport,UK, Peterstone Wales, Wentloog S (2–3 of m) Newport,UK, Peterstone Wales, Wentloog S (2 of m) Newport, Peterstone CE3398 Wentloog (2 m) CE3399 SMNH108927 HM460124 SMNH108928 CE3401 – HM460125 CE3402 – SMNH108929 HM460126 SMNH108930 – HM460127 HM459898 – – – – – – – – – – – – SE, Bohuslän, off Lysekil, BondenSE, (26–31 Bohuslän, m) off Lysekil, BondenSE, (26–31 Bohuslän, m) off Lysekil, Bonden (26–31 m) CE1778 CE1779 SMNH108895 CE1781 HM460092 SMNH108896 – HM460093 SMNH108897 – HM460094 HM459895 – – – – – – – – DK, The Sound, S ofSE, Elsinore Bohuslän, Koster harbour (16 area, m) Saltö,SE, mudflat Bohuslän, (0.5 Koster m) area, Saltö,SE, mudflat Bohuslän, (0.5 Koster m) area, Saltö,SE, mudflat Bohuslän, (0.5 Koster m) area, Saltö,SE, mudflat Bohuslän, (0.5 Koster m) area, Saltö,SE, mudflat Bohuslän, (0.5 Koster m) area, Saltö, mudflat (0.5 m) CE1754 CE1747 CE1755 SMNH85162 CE1756 SMNH85161 SMNH85163 HM460086 CE1758 HM460085 SMNH85164 – HM460087 CE1759 – SMNH85165 – HM460088 CE1760 SMNH85166 – HM460089 SMNH85167 – HM460090 – HM460091 – – – – – – – – – – – – – – – – – – – – – – – – – – – – – SE, Bohuslän, off Lysekil, BondenSE, (26–31 Bohuslän, m) Koster area, Ramsö, at Ullvillarna (20–30 m) CE2080 SMNH108899 HM460096 – CE1782 SMNH108898 HM460095 – – – – – UK, Wales, S of Newport, Peterstone Wentloog (2 m) CE3403 SMNH108931 HM460128 – – – – – SE, Bohuslän, Koster area, Ramsö, at Ullvillarna (20–30 m) CE2081 SMNH108900 HM460097 – SE, Bohuslän, Koster area, Ramsö,SE, at Västergötland, Ullvillarna Gothenburg, (20–30 Saltholmen m) (0.7SE, m) Västergötland, Gothenburg, Saltholmen (0.7SE, m) Västergötland, CE2083 Gothenburg, Saltholmen (1.5 m) SMNH108901 HM460098 CE2541 – CE2543 SMNH108902 CE2587 HM460099 SMNH108903 – HM460100 SMNH108904 – HM460101 – – – – – – – – – – – – – SE, Västergötland, Gothenburg, Saltholmen (1.5SE, m) Västergötland, Gothenburg, Saltholmen (1.5 m) CE2588 CE2589 SMNH108905 HM460102 SMNH108906 – HM460103 – – – – – – – – – (d’Udekem, 1855) SE, Bohuslän, Koster area, Ursholmen (3 m) CE539-1 SMNH85171 HM460078 – – – – – d d d a b c c c a a were derived from two different specimens from the same locality. Most specimens are collected by first and/or third author, and exceptions from this a T. benedii I T. benedii I T. benedii II T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I Species and authority T. benedii II T. benedii II T. benedii II T. benedii II T. benedii II T. benedii II T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii II T. benedii II T. benedii II T. benedii II T. benedii I T. benedii I T. benedii II T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I T. benedii I Table 1 List of included specimens, collection localitiesHong and Kong; GenBank IR, accession Iran; numbers. FR, Country France; codes ET, used Estonia; in BS, the Bahamas; site CA, column, Canada. Accession SE, Sweden; numbers DK, shown in Denmark; bold NL, face were included in the extended 16S and ITS dat costata Author's personal copy

690 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 HM460297 HM460298 HM460302 HM460303 HM460304 HM460287 HM460289 HM460293 HM460295 HM460299 HM460300 HM460288 HM460290 HM460291 HM460294 HM460292 HM460301 HM460296 HM460003 HM460043 –– –– HM460007 HM460047 –– –– HM459999 HM460039 –– HM460002 HM460042 HM460004 HM460044 HM460005 HM460045 –– HM460000 HM460040 HM460001 HM460041 –– HM460006 HM460046 –– HM459951 HM459952 HM459956 HM459957 HM459958 HM459942 HM459943 HM459947 HM459953 HM459944 HM459948 HM459949 HM459954 HM459945 HM459946 HM459955 HM459950 COI 12S 16S 18S 28S ITS 42’E (19–21 m)42’E (19–21 m)42’E (19–21 m) CE2593 CE2594 SMNH108949 CE2595 HM460160 SMNH108950 – HM460161 SMNH108951 – HM460162 – – – – – – – – – ° ° ° 34’N, 012 34’N, 012 34’N, 012 ° ° ° NO, Bergen, near Biological Station,NO, Bergen, Espegrend (90–120 near m) Biological Station,ES, Espegrend Andalucia, (90–120 Cadiz, m) Puerto CE5451 Real, S. Pedro River CE5452 (intertidal) SMNH108956 CE1247 HM460167 SMNH108957DK, The – HM460168 Sound, SMNH108958 S ofDK, Elsinore The – harbour Sound, HM460169 (16 S m) ofDK, Elsinore The harbour – Sound, (16 S m) ofDK, Elsinore The harbour Sound, (16 S m) ofSE, Elsinore Bohuslän, harbour off (16 Lysekil, m) near Bonden (26–31 m) –NL, Zuid-Holland, Dirksland, HaveloseNL, Weg Zuid-Holland, CE1734 Dirksland, HaveloseNL, Weg Zuid-Holland, CE1735 Dirksland, Havelose SMNH85189 Weg – CE1736 CE1773 SMNH85190 HM460173 CE1740 SMNH85191 SMNH108961 – HM460174 HM460177 SMNH85192 – HM460175 HM459902 – HM460176 – CE2261 – CE2262 SMNH108964 – CE2263 HM460180 SMNH108965 – – HM460181 SMNH108966 – – HM460182 – – – – – – – – – – – – – – – – – – – – – – – DK, The Sound, Elsinore, EDK, of The Ellekilde Sound, have Elsinore, (27 EDK, m) of The Ellekilde Sound, have Elsinore, (27 E m) of Ellekilde have (27 m) CE1724 CE1726 SMNH85174 CE1727 SMNH85175 HM460148 SMNH85176 – HM460149 – HM460150SE, – Kattegat, 56 SE, Kattegat, 56 SE, Kattegat, 56 – – – – – – – – NL, Zeeland, Middelburg, Walcheren, canal (8 m)FR, Atlantic coast, Pouldohan BayUK, (intertidal) Wales, Swansea, Swansea Bay, Oxwich (3 m) CE5319 SMNH108934 HM460131 – CE2964 CE3409 SMNH108937 SMNH108938 HM460134 HM460135 – HM459899 SE, Halland, Kattegat, Lilla MiddelgrundSE, (28–30 Halland, m) Kattegat, Morups Bank (30 m) – CE1056 SMNH108943 – HM460146 – CE1274 SMNH85187 – HM460147 – – – – – – – – – – ES, Andalucia, Cadiz, Puerto Real,ES, S. Andalucia, Pedro Cadiz, River Puerto (intertidal) Real, S. Pedro River (intertidal) CE1248 CE1249 SMNH108959 HM460170 SMNH108960 – HM460171 – SE, Bohuslän, off Lysekil, near Bonden (26–31 m) CE 1775 SMNH108962 HM460178 – – – – – DK, The Sound, Elsinore, EDK, of The Ellekilde Sound, have Elsinore, (27 EDK, m) of The Ellekilde Sound, have Elsinore, (27 ÅlsgårdeDK, m) (27 The m) Sound, Elsinore, ÅlsgårdeSE, (27 Skagerrak, m) Väderöarna IslandsSE, (133–137 m) Skagerrak, Väderöarna IslandsSE, (133–137 m) CE1728 Skagerrak, Väderöarna IslandsSE, (133–137 m) CE1729 Bohuslän, off Lysekil, SMNH85177 near Bonden (26–31 m) SMNH85178 HM460151 – HM460152 CE1767 CE1730 – CE1768 CE1731 SMNH108944 CE1769 SMNH85179 HM460155 SMNH108945 CE1774 SMNH85180 – HM460153 HM460156 SMNH108946 – – – HM460154 SMNH108947 HM460157 – – HM460158 – – – – – – – – – – – – – – – – – – – – – – – – CA, New Brunsw., St. Andrews,CA, Passamaquoddy New Bay Brunsw., (0.5 St. m) Andrews, Passamaquoddy Bay (0.5 CE7284 m) CE7285UK, Wales, SMNH108935 Swansea, SwanseaUK, Bay, Oxwich Wales, HM460132 SMNH108936 (3 Swansea, m) Swansea Bay, Oxwich – HM460133 (3 m) – CE3410 – CE3411 – SMNH108939 HM460136 SMNH108940 – HM460137 HM459900 – – – – – SE, Bohuslän, off Lysekil, near Bonden (26–31 m) CE1776 SMNH108948 HM460159 – – – – – UK, Wales, Swansea, SwanseaUK, Bay, Oxwich Wales, (3 Swansea, m) Swansea Bay, Mumbles Pier (3 m)SE, Bohuslän, Koster area, KosterSE, fjord Bohuslän, trench Koster (250 area, m) KosterSE, fjord Bohuslän, CE3484 trench Koster (250 area, m) KosterSE, fjord Bohuslän, CE3482 trench Koster (250 area, m) innerSE, SMNH108942 archipelago Bohuslän, (<50 Koster m) area, CE541–2 HM460139 inner SMNH108941 archipelago (<50 m) – CE541–3 HM460138 SMNH85182 – CE541–4 SMNH85183 HM460141 SMNH85184 CE551 – HM460142 CE552 – HM460143 – SMNH85185 – – SMNH85186 HM460144 – HM460145 – – – – – – – – – – – – – – – – – – – – – – – DK, The Sound, S of Elsinore harbour (16 m) CE1732 SMNH85188 HM460172 HM459901 UK, Wales, S of Newport,UK, Peterstone Wales, Wentloog S (2 of m) Newport, Peterstone Wentloog (1 m) CE3404 CE3405 SMNH108932 HM460129 SMNH108933 – HM460130 – – – – – – – – – Site (depth) Source Voucher# GenBank accession # SE, Bohuslän, Koster area, Tjärnö, at marine lab (intertidal) CE5220 SMNH108955 HM460166 – – – – – UK, Wales, S of Newport,SE, St. Bohuslän, Brides Koster Wentloog area, (5 NSE, m) of Bohuslän, Krugglö Koster (25–27 area, m) Tjärnö, at marine lab (intertidal) CE5219 CE3388 SMNH108954 CE5214 SMNH108952 HM460165 HM460163 – SMNH108953 – HM460164 – – – – – – – – – – – – – ) NL, Zuid-Holland, Dirksland, Havelose Weg CE2260 SMNH108963 HM460179 – a ) SE, Bohuslän, Koster area, Koster fjord trench (250 m) CE541–1 SMNH85181 HM460140 – – – – – Michaelsen, 1926 Erséus, 1975 ( ( f f c c f e e g g g g ) Jaroschenko, 1948 m m m m m b Brinkhurst and Baker, 1979 continued ( T. swirencowi T. heterochaetus T. heterochaetus T. amplivasatus II T. swirencowi T. swirencowi T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. diazi T. swirencowi T. swirencowi T. heterochaetus T. heterochaetus T. amplivasatus I T. amplivasatus II T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. diazi T. benedii II T. amplivasatus II T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. diazi T. benedii II T. diazi T. swirencowi T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. diazi T. diazi T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. swirencowi Species and authority T. benedii II T. benedii II T. benedii II T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I T. amplivasatus I Table 1 Author's personal copy

S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 691 ) HM460314 HM460319 HM460321 HM460305 HM460307 HM460315 HM460322 HM460306 HM460308 HM460320 HM460316 – HM460309 HM460317 – HM460318 HM460310 HM460311 HM460312 HM460313 continued on next page ( HM460016 HM460056 HM460020 HM460060 HM460021 HM460061 HM460008 HM460048 HM460010 HM460050 HM460017 HM460057 HM460022 HM460062 HM460009 HM460049 HM460011 HM460051 –– –– –– –– HM460018 HM460058 –– HM460019 HM460059 HM460012 HM460052 HM460013 HM460053 HM460014 HM460054 HM460015 HM460055 HM459970 HM459975 HM459977 HM459959 HM459962 HM459971 HM459978 HM459960 HM459963 HM459976 HM459972 HM459961 HM459964 HM459973 HM459965 HM459974 HM459966 HM459967 HM459968 HM459969 NL, Zeeland, Middelburg, Walcheren, canalNL, (8 Zeeland, m) Middelburg, Walcheren, canalNL, (8 Zeeland, m) Middelburg, Walcheren, canal (8 m)US, Virginia, Gloucester Co., YorkUS, River, Virginia, Gloucester Gloucester Pt Co., (2 YorkUS, CE5321 m) River, Virginia, Clay Gloucester Bank Co., (8 YorkUS, CE5322 CE2377 m) River, Virginia, Clay Gloucester Bank Co., SMNH109006 (8 York CE5323 m) River, Clay HM460223 Bank SMNH109007 SMNH109010 (8 m) CE2428 – HM460224 HM460227 SMNH109008 CE2429 – – HM460225 SMNH109011 CE2430 HM459915 HM460228 SMNH109012 – HM460229 SMNH109013 – HM460230 – HM459916 – – – – – – – – – – – – – – – – – – – NO, Bergen, at Biological Station,NO, Espegrend Bergen, at (intertidal) Biological Station,US, Espegrend Virginia, (intertidal) York Co., York Bridge Harbour (2 CE5442 m) CE5443 SMNH108997 HM460214 SMNH108998 – HM460215 – CE2408 SMNH108999 HM460216 – HM459911 – – – – – – – SE, Halland, Fladen (15–20 m)DK, The Sound, Elsinore, EllekildeSE, have Bohuslän, Koster (10 m) area, W of Nord-Koster (16–18 m) CE2062 SMNH108976 CE1272 HM460193 HM459905 SMNH108975 HM460192 – CE1064 SMNH95193 HM460191 – – – – – – – – – NL, Zuid-Holland, Dirksland, Havelose WegNL, Zuid-Holland, Dirksland, Wittebrug Goededreede CE2266 SMNH108968 CE2264 HM460184 HM459903 SMNH108967 HM460183 – – – – – US, Virginia, Gloucester Co., YorkUS, River, Virginia, Clay Gloucester Bank Co., (8 YorkUS, m) River, Delaware, Clay Sussex Bank Co., (8 Lewes,US, m) Marine Delaware, Lab Sussex (4 Co., m) Lewes,US, Marine Delaware, CE2438 Lab Sussex (4 Co., m) Lewes,US, Marine Delaware, CE2439 Lab Sussex (4 Co., m) Lewes,US, SMNH109014 Marine Delaware, Lab Sussex (4 Co., m) HM460231 Lewes,US, SMNH109015 Marine Delaware, Lab Sussex (4 Co., – m) HM460232 Lewes,US, CE2474 Marine Delaware, Lab Sussex (4 Co., – m) Lewes,UK, CE2482 Wales, Marine Lab S SMNH109016 (4 of m) Newport, CE2484 St. Brides HM460233 SMNH109017 Wentloog (5 CE2485 m) – HM460234 SMNH109018 CE2487 – – HM460235 SMNH109019 CE2488 – – HM460236 SMNH109020 CE2475 – HM460237 SMNH109021 CE3387 – HM460238 SMNH109022 – – HM460239 SMNH109023 – – – HM460240 – – HM459917 – – – – – – – – – – – – – – – – – – – – – – – – – – – – – US, Delaware, Sussex Co., Lewes, Marine Lab (4 m) CE2491 SMNH109000 HM460217 HM459912 SE, Bohuslän, Koster area, W of Nord-Koster (16–18 m) CE2134 SMNH108977 HM460194 HM459906 US, Virginia, Middlesex Co., LaUS, Grange Virginia, Creek Middlesex (1–2 Co., m) La Grange Creek (1–2 m) CE2446 CE2447 SMNH108969 HM460185 SMNH108970 – HM460186 HM459904 – – – – US, Virginia, Gloucester Co., York River, Gloucester Pt (2 m) CE2376 SMNH109009 HM460226 – UK, Wales, S of Newport, St. Brides Wentloog (5 m) CE3389 SMNH109024 HM460241 – – – – – US, Delaware, Sussex Co., Lewes, Marine Lab (4 m) CE2473 SMNH109001 HM460218 – US, Virginia, Middlesex Co., GroomsUS, Landing Virginia, (0.5 Middlesex m) Co., GroomsUS, Landing Virginia, (0.5 Middlesex m) Co., GroomsUS, Landing Virginia, (0.5 Middlesex m) Co., La Grange Creek (1–2 m) CE2468 CE2469 SMNH108971 CE2470 HM460187 SMNH108972SE, CE2496 Bohuslän, Koster – area, HM460188 SMNH108973 WSE, of Bohuslän, Nord-Koster Koster – (16–18 area, HM460189 SMNH108974 m) WSE, of Bohuslän, Nord-Koster Koster – (16–18 area, HM460190 m) WSE, of Bohuslän, Ursholmen Koster – (38 area, m) WSE, of Bohuslän, CE2135 Ursholmen Koster (38 area, m) W – of CE2136 Ursholmen (38 m) – SMNH108978 HM460195 SMNH108979 – – HM460196 CE3210 – CE3211 – SMNH108980 CE3212 – HM460197 SMNH108981 – – HM460198 SMNH108982 – – HM460199 – – – – – – – – – – – – – – – – – – – – – – US, Delaware, Sussex Co., Lewes,US, Marine Delaware, Lab Sussex (4 Co., m) Lewes,US, Marine Delaware, Lab Sussex (4 Co., m) Lewes, Marine Lab (4 m) CE2476 CE2486 SMNH109002 CE2489 HM460219 SMNH109003 – HM460220 SMNH109004 – HM460221 HM459913 – – – – – – – – SE, Bohuslän, Koster area, WSE, of Bohuslän, Ursholmen Koster (38 area, m) S of Sydkoster, Skåreskär (15 m) CE3250 SMNH108984 CE3213 HM460201 SMNH108983 – HM460200 – – – – – – – – – US, Delaware, Sussex Co., Lewes, Marine Lab (4 m) CE2490 SMNH109005 HM460222 HM459914 SE, Bohuslän, Koster area, Tjärnö, at marine lab (intertidal) CE2077 SMNH108986 HM460203 HM459907 SE, Bohuslän, Koster area, between Saltö and Tjärnö (1 m) CE3205 SMNH108988 HM460205 HM459908 SE, Halland, Kungsbacka, Hanhalsholme (1 m) CE3107 SMNH108987 HM460204 – – – – – SE, Bohuslän, Koster area, Tjärnö, at marine lab (0.5 m) CE199–2 SMNH108989 HM460206 HM459909 SE, Bohuslän, Koster area, between Saltö and Tjärnö (1 m) CE1753 SMNH108990 HM460207 HM459910 SE, Västergötland, Gothenburg, Saltholmen (1.5SE, m) Västergötland, Gothenburg, Saltholmen (1.5SE, m) Västergötland, Gothenburg, Saltholmen (1.5SE, m) Bohuslän, Koster area, betweenUK, Saltö Wales, and N Tjärnö of (1 Caerphilly,UK, m) Rhymney Wales, CE2578 River S (3–4 of m) Newport, Peterstone CE2580 Wentloog CE3206 (1 SMNH108991 m) CE2581 HM460208 SMNH108992 SMNH108994 – HM460209 SMNH108993 HM460211 – CE3355 HM460210 – CE3406 – SMNH108995 SMNH108996 HM460212 HM460213 – – – – – – – – – – – – – – – – – – – – – – – – – – ) SE, Bohuslän, Koster area, Tjärnö, at marine lab (0.5 m) CE199–3 SMNH108985 HM460202 – e Dahl, 1960 Brinkhurst, 1985 ( c c b b b g h Brinkhurst and Baker, 1979 i i Brinkhurst and Baker, 1979 Baker, 1983 T. brownae T. parapectinatus T. parapectinatus T. brownae T. brownae T. pseudogaster II T. wasselli T. kozloffi T. kozloffi T. heterochaetus T. brownae T. brownae T. wasselli T. kozloffi T. heterochaetus T. brownae T. brownae T. brownae T. brownae T. brownae T. brownae T. brownae T. brownae T. brownae T. heterochaetus T. heterochaetus T. brownae T. parapectinatus T. kozloffi T. heterochaetus T. kozloffi T. kozloffi T. heterochaetus T. heterochaetus T. kozloffi T. kozloffi T. heterochaetus T. brownae T. parapectinatus T. kozloffi T. kozloffi T. parapectinatus T. parapectinatus T. kozloffi T. pseudogaster I T. parapectinatus T. pseudogaster I T. parapectinatus T. pseudogaster I T. pseudogaster I T. pseudogaster II T. pseudogaster II T. pseudogaster II T. pseudogaster II T. pseudogaster II T. pseudogaster II T. pseudogaster II T. pseudogaster II T. pseudogaster II Author's personal copy

692 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 HM460327 HM460328 HM460330 HM460323 HM460324 HM460325 HM460326 HM460329 HM460331 HM460332 HM460333 HM460334 HM460026 HM460066 HM460027 HM460067 –– HM460023 HM460063 HM460024 HM460064 HM460025 HM460065 –– HM460028 HM460068 HM460029 HM460069 –– –– HM460030 HM460070 HM459983 HM459984 HM459979 HM459980 HM459981 HM459982 HM459986 HM459985 HM459987 HM459988 HM459989 HM459990 COI 12S 16S 18S 28S ITS CE3424 SMNH109031 HM460248 – CE3417 SMNH109029CE3418 HM460246 HM459919 SMNH109030 HM460247 HM459920 CE5460 SMNH109025 HM460242 – – – – – CE2422 SMNH109040CE2448 HM460257 – SMNH109041 HM460258 – – – – – – – – – CE211 – HM460267 DQ459921 AY885610 AF411879 HM460072 – . Erséus and Kvist (2007) UK, Scotland, Cambpeltown, Campbeltown LochUK, (18 Scotland, m) Campbeltown, Campbeltown Loch (18 m) CE4438 CE4439 SMNH109032 HM460249 SMNH109033 – HM460250 – HK, Conic Island, submarine cave (19 m) CE536 SMNH109026 HM460243 HM459918 IR, Caspian Sea, Noor coastUS, (15–30 Virginia, m; Gloucester salinity Co., 13 Glouc. ppt) Pt., Sarah’s Cr. (3 m) CE1273 CE2396 SMNH109035 SMNH109036 HM460252 HM460253 – – – – – – Site (depth) Source Voucher# GenBank accession # BS, Great Exuma, Lee Stocking Island, Lobster Pond (0.5 m) CE71 SMNH109034 HM460251 HM459921 US, Virginia, Gloucester Co., Glouc. Pt., Heywood Cr. (2 m) CE2412 SMNH109037 HM460254 HM459922 UK, Wales, Swansea, Swansea Bay,UK, Mumbles Wales, Pier Swansea, Swansea (3 m) Bay, Mumbles Pier (3 m) UK, Wales, Swansea, Swansea Bay, Mumbles Pier CE3416 (3 m) UK, Wales, Swansea, Swansea Bay, Mumbles SMNH109028 Pier (3 m) HM460245 – – – – – US, New York, New York,Rossville Staten (0.5 m) Island, US, Virginia, Gloucester Co., Glouc.US, Pt., Virginia, Gloucester Heywood Cr. Co., (2 Glouc.US, m) Pt., Virginia, Gloucester Heywood Cr. Co., (2 Glouc.Heywood m) Cr. Pt., (2 m) CE2413US, Virginia, Middlesex CE2414 Co., La Grange SMNH109038 Creek (1–2 m) US, HM460255 Virginia, SMNH109039 Middlesex Co., LaUS, – HM460256 Grange Virginia, Middlesex Creek (1–2 Co., m) UrbannaUS, – Creek Virginia, harbour Middlesex (2 Co., m) Urbanna Creek harbour (2 m) CE2458 CE2450 – CE2459 SMNH109043 – SMNH109042 HM460260 SMNH109044 HM460259 – HM460261 – – – – – – – – – – – – – – – – US, Virginia, Middlesex Co., Urbanna Creek harbour (2 m) CE2460 SMNH109045 HM460262 – US, Virginia, Middlesex Co., UrbannaUK, Creek Wales, harbour N (2 of m) Caerphilly, Rhymney River (3–4 m) CE2495 SMNH109046 HM460263 CE3354 – SMNH109047 HM460264 HM459923 – – – – Lången Lake (littoral) by ) EE, lab culture kept by Tarmo Timm CE282 – HM460269 HM459927 HM459992 HM460032 HM460074 – T. heterochaetus (Grube, 1861) EE, lab culture kept by Tarmo Timm CE289 – HM460270 HM459928 HM459993 HM460033 HM460075 – Claparède, 1862 EE, lab culture kept by Tarmo Timm CE290 – HM460271 DQ459923 AY885613 AF469007 HM460076 – Erséus, 1990 BS, Great Exuma, Lee Stocking Island (0.5 m) CE131 – HM460272 DQ459919 AY885621 AF411866 HM460077 – Southern, 1909 (Claparède, 1863) SE, Bohuslän, Koster area, near Tjärnö (>10 m) CE196 – HM460266 HM459925 AY340460 AY340432 AY340397 – ( (Piguet, 1906) EE, lab culture kept by Tarmo Timm CE281 – HM460268 HM459926 HM459991 HM460031 HM460073 – (Müller, 1776) SE, Bohuslän, Koster area, Tjärnö Marine lab (0.3 m) CE112 – HM460265 HM459924 AY885615 AF411863 HM460071 – ) (Stolc, 1886) SE, Västergötland, Vårgårda, Fly, Helgason and (Stephenson, 1922) UK, Wales, Swansea, Swansea Bay, Oxwich (3 m) CE3415 SMNH109027 HM460244 – – – – – k k Brinkhurst, 1985 j Brinkhurst, 1986 l continued ( Erséus, 1987 Andrew Mackie (2008). Mershad Taheri (2005). Individuals identiﬁed as Ton van Haaren (2008). Pierre De Wit (2008). Colin Kilvington (2007). Pierre De Wit (2005). Judith Fuchs (2007). Rienk Geene (2006). Tjeerd du Bois (2006). Alice Brylawski (2007). Hong Zhou (2002). Pierre De Wit (2007). T. fraseri authority Species and T. insularis II T. parviductus T. fraseri T. insularis I T. insularis I T. insularis I T. insularis II T. insularis I T. insularis I T. brownae T. imajimai T. fraseri T. fraseri T. fraseri T. fraseri T. fraseri T. fraseri T. fraseri T. fraseri T. fraseri T. fraseri T. fraseri Limnodrilus hoffmeisteri Outgroup taxa Clitellio arenarius Heterochaeta costata Aulodrilus pluriseta Ilyodrilus templetoni Psammoryctides barbatus Limnodriloides anxius Tubifex ignotus i j l f c e a g k b d h m Table 1 Individuals collected by: Author's personal copy

S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 693

Fig. 1. Map of the collection localities for the specimens of Tubificoides. Fourteen countries and three continents are represented among the 187 samples used in the present study. and the mitochondrial and the nuclear loci separately, and (iv) 18 (Kimura, 1980) and with the following settings: uniform rates specimens of Tubificoides and eight outgroup taxa for the phyloge- among sites and complete deletion of gaps. Haplotype netic analysis of the six loci combined. Failure to recover 12S for T. diversities ± S.E. were calculated in DnaSP ver. 4.90.1 (Rozas insularis II, T. amplivasatus I and II, and ITS for the outgroup taxa re- et al., 2003) using the Jukes and Cantor model of nucleotide sulted in these being represented by the five remaining loci, substitution (Jukes and Cantor, 1969) and the sliding window respectively. option (window length 100, step size 25). DnaSP does not Nucleotide diversities ± standard errors (S.E.) of COI, 16S and handle ambiguous nucleotides (e.g. N, M or Y) and, therefore, ITS were calculated in MEGA ver. 4 (Tamura et al., 2007) using 33 such positions were changed to gaps in the extended ITS data the Kimura 2-parameter (K2P) model of base substitution set.

Table 2 List of primers used in the ampliﬁcation and sequencing PCR-reactions.

Gene Primer Sequence Reference

ITS ITS4 50-TCCTCCGCTTATTGATATGC-30 White et al. (1990) ITS5 50-GGAAGTAAAAGTCGTAACAAGG-30 White et al. (1990) 5.8MussR 50-GATGTCGATGTTCAATGTGTCCTGC-30 Källersjö et al. (2005) 5.8MussF 50-CGCAGCCAGCTGCGTGAATTAATGT-30 Källersjö et al. (2005)

18S TimA 50-AMCTGGTTGATCCTGCCAG-30 Tim Littlewood (pers. comm. in Norén and Jondelius, 1999) TimB 50-TGATCCATCTGCAGGTTCACCT-30 Tim Littlewood (pers. comm. in Norén and Jondelius, 1999) 1100R 50-GATCGTCTTCGAACCTCTG-30 Tim Littlewood (pers. comm. in Norén and Jondelius, 1999) 660F 50-GATCTCGGGTCCAGGCT-30 Erséus et al. (2002) 1806R 50-CCTTGTTACGACTTTTACTTCCTC-3 Michael Norén (pers comm. in Hovmöller et al., 2002) 18S4FB 50-CCAGCAGCCGCGGTAATTCCAG-30 Norén and Jondelius (1999) 18S4FBK 50-CTGGAATTACCGCGGCTGCTGG-30 Norén and Jondelius (1999) 18S5F 50-GCGAAAGCATTTGCCAAGAA-30 Marta Riutort (pers comm. in Norén and Jondelius, 1999) 18S7FK 50-GCATCACAGACCTGTTATTGC-30 Marta Riutort (pers. comm. in Norén and Jondelius, 1999)

28S 28SC1´ 50-ACCCGCTGAATTTAAGCAT-30 Dayrat et al. (2001) 28SC2 50-TGAACTCTCTCTTCAAAGTTCTTTTC-30 Dayrat et al. (2001)

12S 12SE1 50-AAAACATGGATTAGATACCCRYCTAT-30 Jamieson et al. (2002) 12SH 50-ACCTACTTTGTTACGACTTATCT-3 Jamieson et al. (2002)

COI LCO1490 50-GGTCAACAAATCATAAAGATATTGG-30 Folmer et al. (1994) HCO2198 50-TAAACTTCAGGGTGACCAAAAAATCA-30 Folmer et al. (1994) COI-E 50-TACTTCTGGGTGTCCGAAGAATCA-3 Bely and Wray (2004)

16S 16SAR-L 50-CGCCTGTTTATCAAAAACAT-30 Palumbi (1996) 16SBRH 50-CCGGTCTGAACTCAGATCACGT-30 Palumbi (1996) Author's personal copy

694 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702

Table 3 Gene and input data for the combined analysis.

Locus and codon position Total number of characters Number of informative characters Model Priors COI 1st position 219 83 GTR + G Lset: nst = 6, rates = gamma Prset: statefreqpr = dirichlet (1, 1, 1, 1) COI 2nd position 219 17 SYM + I + G Lset: nst = 6, rates = invgamma Prset: statefreqpr = ﬁxed (equal) COI 3rd position 220 185 GTR + I Lset: nst = 6, rates = propinv Prset: statefreqpr = dirichlet (1, 1, 1, 1) 16S 507 200 GTR + I + G Lset: nst = 6, rates = invgamma Prset: statefreqpr = dirichlet (1, 1, 1, 1) 12S 415 186 GTR + I + G Lset: nst = 6, rates = invgamma Prset: statefreqpr = dirichlet (1, 1, 1, 1) 18S 1777 56 K80 + I + G Lset: nst = 2, rates = invgamma Prset: statefreqpr = ﬁxed (equal) 28S 331 25 GTR + I + G Lset: nst = 6, rates = invgamma Prset: statefreqpr = dirichlet (1, 1, 1, 1) ITS 2224 579 GTR + I + G Lset: nst = 6, rates = invgamma Prset: statefreqpr = dirichlet (1, 1, 1, 1) Total 5912 1331

Because morphological discrepancies exist for most of our sam- et al., 2002a,b) was used to ﬁnd diagnostic positions (molecular pled species but not for the groups with suspected cryptic forms, synapomorphies or Characteristic Attributes [CA’s]) in a phyloge- the Characteristic Attribute Organization System (CAOS; Sarkar netic-free context (sensu Lowenstein et al., 2009) for these groups.

Fig. 2. Midpoint-rooted Neighbor-Joining (NJ) trees of the COI locus. (a) Tree derived from the NJ analysis of the complete 187 taxa data set with each of T. fraseri, T. brownae, T. heterochaetus and T. swirencowi (discussed in detail in Section 4.3) highlighted. The tree shows a distinct separation of 18 different COI lineages within Tubiﬁcoides. (b–e) show NJ trees of the isolated alignments of T. heterochaetus, T. swirencowi, T. brownae and T. fraseri, respectively. Author's personal copy

S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 695

For the present study, single positions (‘‘simple” CA’s) were identi- 3.3. Genetic variation, haplotype diversity and CA’s in COI fied, which were further categorized as ‘‘pure” (occurring across all members of a given group) or ‘‘private” (occurring exclusively In preliminary diagnoses, specimens of T. benedii, T. amplivasa- within a given group but not across all members). tus, T. pseudogaster and T. insularis all showed morphological con- PAUP* 4.0b10 (Swofford, 2002) was used to construct a Neigh- formity with the original descriptions of these taxa, but COI soon bor-Joining (NJ) tree of the COI locus using the K2P correction mod- suggested that each of them is separated into two different lin- el with gaps treated as missing data and using midpoint-rooting. eages. The suffices I and II are hereafter used to distinguish be- Using PAUP*, stability of the branches (in all but the NJ tree) was tween the different types within these morphospecies and the estimated by bootstrap re-sampling under the maximum parsi- terms ‘‘intralineage” and ‘‘interlineage” variation are used to avoid mony (MP) criterion with 1000 bootstrap replicates each consist- confusion between the terms ‘‘lineages” and ‘‘species”. Thus, a total ing of 10 random addition sequence replicates and employing of 18 different mitochondrial lineages of Tubificoides were investi- TBR branch swapping. gated (Table 1). MrModeltest ver. 2.2 (Nylander, 2004) was used to find the The mean intralineage COI variation in the total data set was most likely model of evolution under a Bayesian framework (see 0.33% ± 0.11 (T. fraseri showing a notably high value Johnson and Omland, 2004), with the data partitioned for each [0.920% ± 0.210]; see Supplementary Table 1), and the mean value gene and also for each codon position in COI. The partitions were of the interlineage divergence was 24.31% ± 2.19 (T. benedii I vs. II tested separately according to the Akaike Information Criterion showing a lower value than other comparisons [4.19% ± 0.74]; (Akaike, 1974; see Table 3) and the posterior probabilities were see Supplementary Table 2). Low values were also noted for T. ins- estimated through Bayesian analysis (BA) using MrBayes3 (Ron- ularis I vs. II (9.04% ± 1.20), T. amplivasatus I vs. II (12.74% ± 1.50) quist and Huelsenbeck, 2003). Two simultaneous runs, each with and T. imajimai vs. T. parapectinatus (13.60% ± 1.55). The mean dis- one cold and three hot Markov chains, were performed for 20 mil- tance between lineages I and II of the morphospecies T. pseudogas- lion generations (except for the extended nuclear analysis, which ter was 22.07% ± 2.01. These five pairings are further discussed in ended prematurely at 16 million generations due to temporal Section 4.1. restrictions of the server) on the CBSU computer cluster at the Cor- Fig. 2 shows the midpoint-rooted NJ tree of the COI locus, in nell Theory Center, New York. Trees were sampled every 1000th which 18 distinct lineages can be identified; the four highlighted generation and the first 2000 trees were discarded as ‘‘burn-in”. and enlarged portions of the tree are important for the discussion This yielded 14,001–18,001 trees to be used for the estimation of on geographical distribution (Section 4.3). In the COI haplotype fre- posterior probabilities. Using the cumulative function, the web quency analysis, T. fraseri showed the largest number of different version of AWTY (Nylander et al., 2008) was used to verify that haplotypes (10) despite being represented by only 13 specimens post burn-in generations had reached stationarity in all data sets. (Supplementary Table 1). The highest diversity (1.00 ± 0.50) was found in T. wasselli but this is only based on two specimens; the closest value to this was 0.95% ± 0.02, present in T. fraseri. At the COI locus, CA’s were recovered for the four taxa with sus- 3. Results pected cryptic forms. The CA’s are listed in their entirety in Table 4. When separating T. benedii I and II, 24 pure and 15 private attri- 3.1. The correct identity of previously deposited GenBank sequences butes were found; for T. amplivasatus I and II, 71 pure and 7 private attributes were found; for T. pseudogaster I and II, 124 pure and 3 Specimens CE1732–1740 (see Table 1) were previously identi- private attributes were found; and for T. insularis I and II, 55 pure fied as T. heterochaetus by Erséus and Kvist (2007), but re-examina- and no private attributes were found. tion of the morphological vouchers showed that they are T. swirencowi. Thus, five previously published COI sequences (Gen- Bank Nos. EF675222–675226) should be attributed to the latter 3.4. Genetic variation in 16S species. Moreover, a worm erroneously identified as T. pseudogaster was The ratio between interspecific and intraspecific variation in previously used for sequencing of 18S (AF411873; Erséus et al., 16S largely reflects that of COI, but 16S showed consistently lower 2002), 16S (AY885609; Sjölin et al., 2005), and 12S (DQ459922; values than COI. The mean intralineage variation was 0.15% ± 0.12 Envall et al., 2006; Erséus et al., 2010). During the present study, (notably, the highest variation occurring in T. fraseri [0.58% ± 0.30]) we found that this worm instead is identical to Heterochaeta costa- and the mean interlineage divergence was 18.11% ± 2.57, with the ta Claparède, 1863. This is further treated in Section 4.4. comparisons T. benedii I vs. II, T. insularis I vs. II, T. amplivasatus I vs. II and T. imajimai vs. T. parapectinatus showing low values (Supple- mentary Tables 3 and 4).

3.2. New geographical records 3.5. Genetic variation in ITS Tubificoides swirencowi, previously known from the Black and Mediterranean Seas (Jaroschenko, 1948; Casellato, 1999; Casellato The ITS region failed to separate T. benedii I and II (see Fig. 3). and Salmaso, 2000) is here formally reported for the first time from Therefore, these were lumped into one lineage in both the genetic Denmark and Sweden. As noted above, this replaces the erroneous variation and haplotype diversity analyses of this region (Supple- record of T. heterochaetus (from Denmark) by Erséus and Kvist mentary Table 5). The mean intralineage variation of the ITS region (2007). Furthermore, T. fraseri, previously known from Australia, was 0.28% ± 0.16 and the mean interlineage divergence was New Zealand and several sites on the Atlantic and Pacific coasts 9.36% ± 1.72, with low values noted for T. imajimai vs. T. parapectin- of the US (Brinkhurst, 1986; Erséus, 1989), is reported for the first atus (0.94% ± 0.51), T. pseudogaster I vs. II (1.26% ± 0.63) and T. bene- time from Wales and the Caspian Sea. Finally, T. parapectinatus, dii vs. T. kozloffi (1.28% ± 0.61). T. imajimai and T. parviductus previously known from the Pacific coast of North America (Brink- (represented by single specimens) were separated from other spe- hurst, 1985) is reported for the first time from the Netherlands cies by a mean of 8.33% ± 1.70 and 8.70% ± 1.58, respectively (Sup- and from the US Atlantic coast. plementary Table 6). Author's personal copy

696 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 re pure. de position at the locus the difference is situated. Nucleotides in parentheses indicate the CTCGCAATCCATATCGCTCTATAACACG TTCGACGATCTACCTTCCCTCTAGTACCTCTTTGCTATC ATCTATCTGGGCTAGGATAATTTCTTTA TCTCGATCAAAACTAAGCGGCCATCCCG CCGCCCATATACCATCTTCCTCCCTAGTTATTTGAGCGTTGCTCCATATTTATAC GC A(T)T A T G C C A G G(A) T C T A T C C C T T G A GCCAGCGAACCTCGTGCTCAGCCTACGTCACCCGAGGCAGACCCCTTACCTAATT T(C)C ATAATTATAATTACCCACGACACGC T C(T)TCCCCAAGGGAATCCGACCCCAATACA TC G A T CCTGTGACGATTGCGGCGCATACTCGCCTCACCGATTCCCAAATATTCTAATAT C T CTCCACGTAGCCTTCTAATATACTCTCA ACTACAACCATGTCTACCACCGACTGTTT T G A A C T C A(T) G G A(C) C T C C C G T A(C) A T T T T T C T T A T C T A C T A G A G A G T T A C C A T G(A) T G(A) A T C T A C C A C T T CA T T(C)A(G)G T C G A GA G G(C) G A(G) T A T T(C) A(T) T C T T T A(G) A G C T T(A) G(A) C T T C A C A T T(C) A T T G C G C(T) C(G) T(C) A G C T C A A T G A A(G) T T G A A C C A T G G A C T C G(A) A C A A T C G C T C T T(A) T T G C A C A T A T T C G T A T. insularis II PositionT. insularis I T. insularis II Position 005T. insularis I T. insularis 010 II Position 016 403T. insularis I 031 409 034 428 653 049 439 064 451 076 456 100 474 109 481 112 484 122 493 127 496 148 502 190 508 214 520 217 538 229 562 241 571 250 598 265 601 274 607 283 613 328 616 334 619 340 622 373 628 631 652 T. pseudogaster I T. pseudogaster II PositionT. pseudogaster I T. pseudogaster II Position 436T. pseudogaster I T. pseudogaster 439 II 457 568 460 574 469 577 474 580 478 581 484 583 496 586 502 589 505 595 508 596 512 598 513 604 515 613 517 631 523 637 526 640 532 643 538 646 539 655 541 547 550 556 562 565 T. amplivasatus II PositionT. pseudogaster I T. pseudogaster II Position 005T. pseudogaster I T. pseudogaster 010 II Position 011 142 014 143 016 145 277 022 158 284 023 160 286 030 163 295 031 172 298 037 181 307 043 184 046 319 187 325 049 197 343 055 202 059 346 205 061 347 208 076 350 211 079 358 212 088 359 214 094 361 220 118 379 223 121 373 226 122 376 232 124 379 238 130 385 241 388 133 250 139 391 256 397 263 400 271 415 421 424 PositionT. amplivasatus I T. amplivasatus II Position 001T. amplivasatus I T. amplivasatus 010 II Position 019 271T. amplivasatus I 021 274 049 292 547 067 293 553 085 295 565 109 298 571 112 307 574 133 313 583 142 315 157 586 328 158 592 331 160 593 350 178 595 373 181 598 407 187 599 418 197 610 428 206 616 436 214 619 454 220 622 460 235 625 478 247 628 484 251 631 487 253 637 505 643 259 508 268 646 514 649 526 658 538 PositionT. benedii I T. benedii II Position 003T. benedii I T. benedii 009 II 019 415 047 477 082 508 088 521 100 542 127 556 142 576 181 586 184 618 187 625 200 630 202 636 205 208 214 220 223 242 265 278 343 370 379 382 405 Table 4 Characteristic Attributes (CA’s) in theminority COI private locus of attributes, the i.e., four attributes different present morphospecies in as retrieved some by representatives CAOS. of ‘‘Position” one indicates lineage in which and nucleoti not in the other; all other attributes in the table a Author's personal copy

S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 697

Fig. 3. Bayesian tree of the ITS region (average ln L of 14437.61 and an average harmonic mean of the two runs of 14501.56). The tree shows 17 distinct lineages of À À Tubiﬁcoides and a failure to resolve T. benedii I from II. However, for T. amplivasatus, T. pseudogaster and T. insularis the tree shows a distinct separation of the different types (I and II). Posterior probabilities (>0.5) are shown above, and bootstrap support values (>50%) below the branches.

3.6. Phylogenetic analyses Tubificoides shows a basal dichotomy, with a clade of T. kozloffi and T. benedii (supported by pp 0.96) being sister to all remaining Parsimony and Bayesian analyses were accomplished for the taxa (with pp 1.00). The latter group consists of T. pseudogaster combined data set, the mitochondrial loci, the nuclear loci and (I + II, supported by pp 1.00, bs 96%), and a large assemblage (sup- each of the six loci, separately. The tree from the combined data ported by pp 1.00) with four clades that are unresolved from each set is shown in Fig. 4 and the mitochondrial and nuclear trees other. The first two of these clades are T. parviductus and T. swirenc- are shown in Supplementary Figs. 1 and 2, respectively (some sup- owi, respectively. The third (supported by pp 1.00, bs 99%) contains port values present in the single gene trees are presented in Sup- T. amplivasatus I + II and T. insularis I + II, both morphospecies with plementary Table 7). To test if the resulting topology of the maximum support. The fourth clade (pp 1.00) is a group of seven combined tree was compromised by the absent ITS data for the species forming two primary subclades, one (with only pp 0.93) outgroups, this region was also excluded from the ingroup taxa comprising T. heterochaetus and T. fraseri, the other (pp 1.00) con- and the analysis was re-run. The resulting topology showed a taining T. parapectinatus + T. imajimai (pp 1.00, bs 100%), and T. monophyletic Tubificoides, and the locus was therefore included diazi + T. brownae + T. wasselli (pp 0.94); among the last three spe- in the combined analysis. cies, T. brownae and T. wasselli are sister taxa (pp 1.00, bs 96%). The combined data set included 5912 characters; 1331 of which The extended mitochondrial and nuclear data sets, with 34 were parsimony informative (Table 3). The BA tree (Fig. 4) is lar- Tubificoides taxa each, consisted of 1577 and 4549 characters, gely congruent with, but more resolved than, the MP tree (not respectively. The mitochondrial data set, under both optimality cri- shown). Thirteen of its nodes, relevant to the phylogeny of Tubifi- teria, supports the monophyly of Tubificoides with pp 1.00 and bs coides, show posterior probabilities (pp) > 0.95, but only nine of 82% (Supplementary Fig. 1), but for the remaining interrelation- these received bootstrap support (bs) > 90% in the MP analysis ships, the MP tree was somewhat unresolved. In total, 13 nodes (see Fig. 4). Monophyly of Tubificoides is strongly supported (pp relevant to Tubificoides (excluding the terminal nodes showing 1.00, bs 94%), and Ilyodrilus templetoni is suggested as this genus’ the alliance of the two individuals of the same species, which all sister taxon (pp 1.00, bs < 50%). showed maximum support) received pp P 0.95 and 12 of these Author's personal copy

698 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702

Fig. 4. Bayesian tree derived from the combined analysis of the six loci. The two BA runs returned trees with an average ln L of 34869.98 and an average harmonic mean of À 34945.10. Posterior probabilities (>0.5) are shown above, and bootstrap support values (>50%) below the branches. Following each taxon is an abbreviation of the collection À country and, when present, a ‘‘leaf” indicates cuticular papillation and a ‘‘needle” indicates hair chaetae. The image of the worm is taken from Erséus and Bonomi (1987). also exist with high support (save for one node showing pp 0.59, All the studied mitochondrial loci separate T. benedii into two bs < 50%) in the combined tree. However, only six nodes show different lineages (Fig. 2a; Supplementary Fig. 1). These were both bs P 90% and these are all present with high support in the com- represented in the same localities in the Netherlands and Canada bined data set tree. Using the nuclear data set, both the MP and but only one of the lineages was found in other localities (T. benedii BA trees nested Ilyodrilus templetoni unresolved within Tubificoides I: Sweden, Norway, Denmark and France; T. benedii II: Wales and (Supplementary Fig. 2). Nine nodes relevant to the phylogeny of England; see Fig. 1). Thus, both lineages are distributed across Tubificoides (excluding the terminal nodes as noted above, which the Atlantic Ocean. There is no evidence for ceased gene flow be- again received maximum support) received pp > 0.95 and all of tween or within populations and we therefore conclude that the these exist with high support in the combined tree. Only six nodes mitochondrial divergence is due to a former separation of mito- show bs > 90% and these also overlap with highly supported nodes chondrial lineages and we suggest that T. benedii be regarded as in the combined tree. a single species. The sister group relationship of the two lineages of T. pseudogaster is surprising considering their large COI distance (Supplemen- 4. Discussion tary Table 2). In fact, the comparisons between T. pseudogaster II and, respectively, T. benedii I, T. benedii II and T. kozloffi all show 4.1. Species delimitation and cryptic speciation smaller distance than that between and T. pseudogaster I and II. The nuclear genes (especially ITS) also resolve T. pseudogaster line- De Queiroz (2007), when reviewing the concepts of species and age I from lineage II. Specimens of the two lineages are almost species delimitation, suggested a unifying concept of species being identical morphologically but a putative difference may be the defined as ‘‘separately evolving metapopulation lineages”. Further, noted rich mucus formation on the body surface of T. pseudogaster he argued that the actual delimitation of such lineages as species is II, not seen in lineage I (personal observation). As each lineage brought through the practical process of establishing lines of evi- exhibits (reciprocal) monophyly in the combined tree it thus seems dence for independent evolution (e.g. morphological differences, appropriate to regard them as separate species. genetic divergence, evidence of ceased gene flow, reproductive iso- Tubificoides amplivasatus I and II are morphologically indistin- lation). That is, the more evidence supporting that two populations guishable, and yet genetically distinct in both their mitochondrial have separate histories, the stronger the case for them being differ- and nuclear genomes (Supplementary Tables 2 and 6; Supplemen- ent species. Each of the morphospecies T. benedii, T. pseudogaster, T. tary Figs. 1 and 2). Specimens of T. amplivasatus I were all collected amplivasatus and T. insularis were found to contain two distinct in the northeastern Atlantic Ocean (the area of the type locality for mitochondrial lineages and the question is whether or not these the species; Erséus, 1975) and the three specimens of T. amplivas- lineages are separate species. atus II were collected on the Atlantic coast of southern Spain. We Author's personal copy

S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 699 thus conclude that speciation has occurred and a formal descrip- However, the frequently employed distance-based method has tion of ‘‘T. amplivasatus II” as a new (cryptic) taxon is currently in recently met opposition in a more character-based approach to preparation (Kvist and Erséus). Interestingly, the two species pairs DNA barcoding (e.g. DeSalle et al., 2005; Rach et al., 2008). The T. amplivasatus I and II, and T. imajimai and T. parapectinatus show rationale behind this approach is that a reproducible standard similar values of genetic divergence in both the mitochondrial and threshold of genetic distance for species discrimination cannot be nuclear loci suggesting that the species split occurred at approxi- determined. Instead, barcoding may rely on discrete nucleotide mately the same time for both pairs. substitutions to identify families, genera and species (Rach et al., The COI divergence between T. insularis I and II is intermediate 2008). As with traditional distance-based barcoding efforts, this between that of T. benedii I and II, and T. amplivasatus I and II, and in character-based approach is not intended to be used as a tool for the few specimens available (five of T. insularis I, two of T. insularis species discovery but rather a tool for easy and unequivocal species II) there was only one COI haplotype in each lineage. The ITS dis- identification. In the present study, this character-based approach tance between T. insularis I and II, however, is comparable to the was adopted to provide a possible means for subsequent robust intralineage variation within T. wasselli as well as the interlineage species identification for the cryptic species present in the T. ampli- divergence between T. pseudogaster I and II. The T. insularis speci- vasatus, T. pseudogaster and T. insularis complexes. The different mens were collected over a narrow range (T. insularis I: Wales, T. species within these three complexes are each clearly distinguish- insularis II: Scotland) and are morphologically inseparable yet the able on the basis of the CA’s recovered by CAOS (Table 4). trees based on nuclear data suggest that they are evolving sepa- The 16S data of Tubificoides showed lower distance values than rately. For the time being, T. insularis I and II may be regarded as those of COI in both the intralineage and interlineage comparisons. two putative cryptic species, but a formal taxonomic revision is The higher genetic rate in COI is likely due to its hypervariable not warranted until the population genetics of this complex has third positions (Halanych and Janosik, 2006). The ITS region, how- been studied in greater detail, using additional specimens not only ever, showed an intralineage variation comparable to that of COI, from the Northeast Atlantic; the morphospecies T. insularis has also but an interlineage divergence only about one-third the size of that been recorded from New Brunswick and Maine in the Northwest of COI. Atlantic (Brinkhurst, 1985). The need for additional data also applies to T. wasselli, the two 4.3. Geographical distribution specimens of which differed significantly in both COI and ITS. Its intraspecific COI variation was about twice the size of the second Considering that so many species of Tubificoides have a wide highest value (present in T. fraseri), and there would be evidence geographic range, often covering different continents and/or for cryptic speciation also in T. wasselli should the rather high ITS oceans, sustaining and reproducing in non-indigenous areas ap- divergence show a consistent pattern of separation in a larger sam- pear to be common characteristics within this genus. However, ple of this taxon. at the same time, most of the widely distributed species treated Most of the other nominal species examined in this study are by us have surprisingly high nucleotide and haplotype diversities distinct on the basis of both mitochondrial and nuclear genetic dis- (in COI and partly in 16S; see Supplementary Tables 1 and 3). tances (Fig. 2a; Supplementary Tables 1–6), but also in terms of Tubificoides brownae, T. fraseri, T. diazi, T. heterochaetus, T. benedii phylogeny (Figs. 3 and 4; Supplementary Figs. 1–2). In addition and T. parapectinatus are, in this study, reported from both sides of to this, there are clear morphological discrepancies between these the North Atlantic Ocean, and in one case (T. fraseri) an Asian in- other lineages. In summary, we thus conclude that our material of land water (the Caspian Sea). Beyond this, these particular species, Tubificoides represents 17 separately evolving metapopulations and also T. kozloffi, T. pseudogaster, T. swirencowi, T. insularis and T. (i.e., species), three of which are cryptic forms within T. pseudogas- wasselli have previously been reported from other widespread ter, T. amplivasatus, and T. insularis, respectively. locations, such as the Indian Ocean and the Gulf of Mexico (T. brownae: Erséus, 1985; Milligan, 1991), the Southwest Pacific Ocean (T. fraseri: Erséus, 1989), the Arctic region (T. kozloffi: Helgason and 4.2. DNA barcoding and genetic variation Erséus, 1987), the Northwest Atlantic Ocean (T. kozloffi: Brinkhurst and Baker, 1979; T. insularis: Brinkhurst, 1985), the Black and Med- In line with the results presented by Erséus and Kvist (2007), iterranean Seas (T. swirencowi: Jaroschenko, 1948; Casellato, 1999; our new data show that the interlineage divergence of COI is 1–2 T. heterochaetus: Popescu-Marinescu et al., 1966) and the Northeast orders of magnitude larger than the intralineage variation of the Pacific Ocean (T. brownae, T. fraseri, T. pseudogaster, T. diazi, T. bene- same gene (Supplementary Tables 1–2) also in this taxonomically dii, T. wasselli: Brinkhurst, 1986; T. heterochaetus: Thompson et al., and geographically wider representation of Tubificoides. Only a 2000). Moreover, T. motei Brinkhurst, 1986 was described from few interlineage comparisons show mean divergence values below both sides of the North American continent. The high intraspecific 20% and, conversely, only the mean distance within T. wasselli (two nucleotide and haplotype diversities shown by many species stud- specimens only) is above 1%. In the two putative species within T. ied by us suggest that multiple introduction events may have insularis, as well as in T. pseudogaster I, COI has shown no variation shaped the genetic compositions of their populations (Frankham, thus far. These results indicate that, regardless of the increased 2005). While the high diversity could potentially also be explained geographical area sampled and the inclusion of more taxa, the bar- by putative large effective population sizes, mitochondrial genetic coding gap is still large enough to effectively separate species from diversity has been shown to be unaffected by population size (Ba- congeners (see Hebert et al., 2003a) in Tubificoides. These COI val- zin et al., 2006). Thus, as the high genetic diversity in our samples ues parallel prior studies of oligochaetous clitellates. For example, is largely confined to the mitochondrial loci (with the exception of Huang et al. (2007) reported an interspecific divergence of P15% ITS in a few cases), population size alone would not sufficiently ex- as opposed to an intraspecific variation of <1% (in most cases) in plain the high diversity found in our samples. closely related species of Amynthas (Megascolecidae), Voua Otomo Because of its abundance in certain areas as well as the lack of et al. (2009) noted a variation of >16% between, and <1% within, evidence that the specimens actually mate, Brinkhurst (1986) pro- species of Eisenia (Lumbricidae) and Gustafsson et al. (2009) posed that T. fraseri is capable of reproducing either by self-fertil- showed that the commonly used lab worm Lumbriculus variegatus ization or parthenogenesis. If this species reproduces by (Lumbriculidae) is subdivided into at least two distinct species, parthenogenesis only, the ten COI haplotypes found in our material separated by an average COI distance of 17.7%. (Fig. 2e; Supplementary Table 1) can each be regarded as repre- Author's personal copy

700 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 senting a specific clone. It is then noteworthy that the single worm ulations. No sample of any reasonable size from one area bears any from the Caspian Sea (CE1273) belongs to the same clone as two signs of bottlenecked diversity. worms from Virginia, USA (CE2422, CE2460), whereas the remain- There may be a few records of recent introductions in our data ing nine individuals from Virginia represent seven other clones and set; one is the finding of T. heterochaetus in The Netherlands (see the single specimen from the UK (Cardiff) represents yet another above), another the finding of an otherwise ‘‘American” haplotype clone. This high diversity suggests that the Virginia population of of T. fraseri in the Caspian Sea. However, as both these species are T. fraseri either is the result of multiple introductions, or belongs known also from several other parts of the world it is still difficult to the source population of this taxon. to establish their native areas. Moreover, as will be discussed be- Great COI diversity was also found in T. brownae (17 specimens, low, there is not much help from their phylogeny either. With eight haplotypes; Fig. 2d) and T. heterochaetus (12 specimens, se- the emerging patterns of so many different species being discov- ven haplotypes; Fig. 2b). The former is often a dominant contribu- ered across the expected boundaries of marine biogeography, there tor to biomass of benthic invertebrates (Diaz, 1984), and it has seems to be overwhelming evidence that these taxa are ubiquitous, been shown to inhabit degraded (Lerberg et al., 2000) as well as opportunistic worms that have become common in several coastal unpolluted sediments (Nichols and Thompson, 1985). Thus, its sediments. In some sense of the word they are thus ‘‘invasive”, but geographic expansion is probably not confined to any particular as they appear to be so well established, it is still difficult to under- oxygen level or nutritional preference and it is probably capable stand if they adversely have affected, or will affect, the habitats of rapid reproduction, regardless of mode. Our sample of T. brow- they invade economically, environmentally or ecologically. nae contains two European specimens (CE3387, CE3389), and a total of 15 worms from Virginia, Delaware and New York, all off the 4.4. Phylogeny East coast of USA (Fig 2d; Table 1). CE3387 is close to CE2475, an aberrant US haplotype from Delaware, while CE3389 is nested Based on a comparative study of the arrangement of chaetae, among the other American haplotypes, and the sample as a whole and the morphology and orientation of the male genital organs, may be part of one large, genetically diverse, amphi-Atlantic Erséus (1984) discussed the possibility of Tubificoides being re- population. cently diverged from the otherwise mainly freshwater-based gen- Tubificoides heterochaetus has also been found dominant in era comprising the subfamily Tubificinae. Moreover, using DNA many types of environments, particularly in oligohaline tidal data, this subfamily has been shown to be closely related to Limno- creeks and estuaries (Diaz, 1989). For example, de Vos (1936) driloidinae, an exclusively marine taxon (Sjölin et al., 2005; Envall had not found this species while collecting in Zuiderzee, The et al., 2006; Erséus et al., 2010). Therefore, the phylogenetic analy- Netherlands in 1920–21, but when sampling the same brack- ses of the combined data set of the present study included seven ish-water area in 1927–32, he found that T. heterochaetus had tubificine and one limnodriloidine species as outgroups. The anal- become the most dominant oligochaete at many sampling sta- ysis (Fig. 4) supports Tubificoides as a monophyletic group in Tub- tions. He therefore concluded that it was introduced to Zuider- ificinae (pp 1.00, bs 94%), and that Ilyodrilus templetoni is its sister zee after 1921 (as cited by Wolff, 2005). Interestingly, this taxon (pp 1.00, bs 93%). In some regards, I. templetoni is morphospecies was first described and named, from the Southern Baltic logically similar to Tubificoides, e.g., both taxa have cuticularized Sea, by Michaelsen (1926) at about the same time. The rapid in- penis sheaths and the placement of the prostate gland is subapical crease in abundance of T. heterochaetus in Zuidersee suggests on the atrium (for I. templetoni, see Southern, 1909, Plate VIII, that this species is not native to the Dutch coast. Our six speci- Fig. 6E–F). However, in I. templetoni the vas deferens enters the at- mens from a single site in Suid-Holland (CE2260-2264, CE2266) rium apically whereas in Tubificoides this structure enters the at- represent four different haplotypes (Fig. 2b), suggesting multiple rium subapically in a position more or less opposite to the introductions. At the other side of the Ocean, at two sites in Vir- prostate gland; this latter vas-atrium arrangement is likely to be ginia, the six other specimens of T. heterochaetus represent five an autapomorphy for Tubificoides. separate haplotypes, two of which are identical to those of Dutch Further, Ilyodrilus Eisen, 1879 is a freshwater genus as is Psam- specimens. Thus, this species exhibits a rather high diversity in moryctides Hrabeˇ, 1964, which in our Bayesian analysis (Fig. 4) is both areas investigated. sister to Ilyodrilus + Tubificoides (with pp 0.97), whereas all mem- Tubificoides swirencowi, first described from the Black Sea (Jar- bers of Tubificoides live in brackish or fully marine waters. This oschenko, 1948), and later reported from the Mediterranean Sea phylogeny thus supports a secondary invasion of the sea by Tubif- (Casellato, 1999), now appears to be established also in the Katte- icoides (as suggested by Erséus, 1984) unrelated to the origins of gat–Skagerrak region of Denmark and Sweden (Erséus and Kvist, other marine groups within the Naididae. It should be noted, how- 2007; present paper). This taxon shows rather high haplotype ever, that this placement of Psammoryctides is not well-supported diversity (Supplementary Table 1), but this diversity only corre- in the bootstrap analysis and because Bayesian approaches have sponds to five nucleotide substitutions, evenly distributed among been shown to be more prone to incorrectly supporting false phy- the individuals, indicating that the Scandinavian population is a re- logenetic relationships than a bootstrap approach (Douady et al., cent introduction, possibly from the Mediterranean area. Further 2003), this node should be viewed with appropriate caution. supporting this notion is the fact that the ITS region shows no var- Some of the previous molecular assessments, including mem- iation in our (limited) sample. In the COI NJ tree presented by Ersé- bers of Tubificoides, failed to recover the monophyly of the genus us and Kvist (2007), the five specimens of T. swirencowi (then (Erséus et al., 2002, 2010; Envall et al., 2006) but these studies labeled T. heterochaetus; see Section 3.1.) also display five different were based on fewer loci and included a misidentified specimen haplotypes. However, the haplotype difference between CE1732 (‘‘T. pseudogaster”), now known to be H. costata (see Section 3.1.). and CE1736 is solely a length inequality; no actual nucleotide sub- The basal topology of the Tubficoides phylogeny involves a para- stitutions separated the two specimens. These sequences were ex- phyletic assemblage of taxa (T. benedii, T. kozloffi, T. pseudogaster I tended to the same length in the present study and no further and T. pseudogaster II), which share a more northern distribution differences were found between them (Fig. 2c). range than the remaining species (Fig. 4). This suggests that the The overall tendency in these examples of the widely distrib- common ancestor of Tubificoides (i.e., of the species sampled here) uted species of Tubificoides is that there is great mitochondrial var- evolved in the colder part of the northern hemisphere. The remain- iation throughout their ranges, but that there has been, and may ing part of the tree contains mostly temperate and boreal species, still be, considerable gene flow between some of the various pop- but also a few known to extend into tropical waters (T. parviductus Author's personal copy

S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702 701 in the Caribbean, T. imajimai in Hong Kong and Florida, T. wasselli in We thank M. Lindström, P. De Wit, L. Matamoros, J. Fuchs, K. Stur- Florida). This large group contains a particular clade (T. heterocha- divant, K. Mortimer, C. Kilvington, M. Taheri, H. Zhou, A. Brylawski, etus through T. wasselli in Fig. 4), in which all species show a distri- R. Geene, T. du Bois, T. van Haaren, S. Montanari, H. Yi, M. Siddall, F. bution involving more than one continent or ocean. Among the Fontanella, C. Calmerfalk and T. Timm for help with additional col- other species in this group, T. swirencowi, T. amplivasatus I and II lecting; A. Ansebo, E. Sjölin and M. Lindström for assistance in the have yet to be found outside Europe and T. swirencowi has probably molecular lab; P. De Wit for software assistance. We also acknowl- been introduced to Scandinavia in recent times (see above). The edge the Italian Journal of Zoology for kindly allowing us to use the morphospecies T. insularis, however, has also been recorded from image of the worm in Fig. 4. All of the molecular work was done at the North American coast of the Atlantic Ocean. At present, it is dif- the Department of Zoology, Göteborg, Sweden. An early draft of ficult to see any biogeographical pattern reflected in the topology this paper was much improved by thoughtful considerations from of this large group, but it can be noted that all its species, except A. Oceguera-Figueroa, A. Phillips, M. Siddall and three anonymous T. amplivasatus I and II, and T. swirencowi, have at some point been reviewers. This study has been financially supported by the Swed- recorded from the Northwest Atlantic. Moreover, considering that ish Research Council (grant #621-2007-5313 to C.E.). in addition to the taxa included in the present study, about 25 other species of Tubificodes have been described from this area (from Canada in the north to Central America in the south [our Appendix A. Supplementary data compilation]), the Northwest Atlantic seems to be a hotspot for species diversity of Tubificoides. Therefore, it is not unreasonable Supplementary data associated with this article can be found, in to suggest that at least the group containing the seven species, T. the online version, at doi:10.1016/j.ympev.2010.08.018. heterochaetus through T. wasselli (in Fig. 4), may have originated and diversified in this region. However, additional taxa from the References North American continental shelves and the Caribbean are needed to test this hypothesis. Akaike, H., 1974. A new look at the statistical model identification. IEEE Trans. Automat. Control 19, 716–723. 4.5. Evolution of morphological characters Baker, H.R., 1980. A redescription of Tubificoides pseudogaster (Dahl) (Oligochaeta: Tubificidae). Trans. Am. Microsc. Soc. 99, 337–342. Baker, H.R., 1984. Diversity and zoogeography of marine Tubificidae (Annelida, As mentioned above, penis sheath morphology, presence/ab- Oligochaeta) with notes on variation in widespread species. Hydrobiologia 115, sence of hair chaetae and cuticular papillation, as well as the 191–196. arrangement of the bifid chaetae, general body shape and size Bazin, E., Glémin, S., Galtier, N., 2006. Population size does not influence mitochondrial genetic diversity in animals. Science 312, 570–572. are characters that are commonly used in diagnoses of Tubificoides Bely, A.E., Wray, G.A., 2004. Molecular phylogeny of naidid worms (Annelida; species. The molecular phylogeny presented here allows for evalu- Clitellata) based on cytochrome c oxidase I. Mol. Phylogenet. Evol. 30, 50–63. ation of the stableness of these characters as phylogenetic markers, Brinkhurst, R.O., 1962. A check-list of British Oligochaeta. Proc. Zool. Soc. Lond. 138, 317–330. as the morphological traits can be optimized on the tree. Here we Brinkhurst, R.O., 1965. Studies on the North American aquatic Oligochaeta II: focus on two conspicuous characters: hair chaetae and cuticular Tubificidae. Proc. Acad. Nat. Sci. Phila. 118, 1–33. papillation and these were optimized under a parsimony Brinkhurst, R.O., 1985. A further contribution to the taxonomy of the genus Tubificoides Lastockin (Oligochaeta: Tubificidae). Can. J. Zool. 63, 400–410. framework. Brinkhurst, R.O., 1986. Taxonomy of the genus Tubificoides Lastockin (Oligochaeta, Judging from the topology of the tree (Fig. 4), there seems to be Tubificidae): species with bifid setae. Can. J. Zool. 64, 1270–1279. a significant amount of homoplasy in the occurrence of hair chae- Brinkhurst, R.O., Baker, H.R., 1979. A review of the marine Tubificidae (Oligochaeta) T. kozloffi T. amplivasatus I II, T. of North America. Can. J. Zool. 57, 1553–1569. tae, as the species with hairs ( , and Casellato, S., 1999. Oligochaetes of karavasta lagoon (Albania) preliminary results. swirencowi, T. imajimai, T. insularis and T. parapectinatus) are partly Hydrobiologia 406, 175–182. scattered across the tree. In light of the phylogeny, the hair chaetae Casellato, S., Salmaso, C., 2000. Strategies of heavy metal detoxification in feature seems to have been lost or gained on at least four indepen- Tubificoides swirencowi (Jarosh, 1948): an estuarine species widespread in northern Adriatic area. Biol. Mar. Mediterr. 7, 662–665. dent occasions in Tubificoides (see Fig. 4). Cook, D.G., 1969. Peloscolex dukei n.sp. and P. aculeatus n.sp. (Oligochaeta, Cuticular papillation is not unique for Tubificoides but is present Tubificidae) from the North-west Atlantic, the latter being from abyssal in several other tubificine naidids (e.g. Spirosperma, Quistadrilus), as depths. Trans. Am. Microsc. Soc. 88, 492–497. Dahl, I.O., 1960. The oligochaete fauna of 3 Danish brackish water areas (Taxonomic well as representatives of other subfamilies including Phallodrili- and biological observations). Meddelelser fra Danmarks Fiskeri-og nae (Duridrilus) and Limnodriloidinae (Tectidrilus). The species that Havundersøgelser (New Series) 26, 1–20. show cuticular papillation (T. wasselli, T. benedii, T. heterochaetus Dayrat, B., Tillier, S., Lecointre, G., Tillier, S., 2001. New clades of euthyneuran gastropods (Mollusca) from 28S rRNA sequences. Mol. Phylogenet. Evol. 19, and T. insularis) are disunited in our tree (Fig. 4), suggesting that 225–235. this character is also homoplasious in Tubificoides; cuticular papil- DeSalle, R., Egan, M.G., Siddall, M., 2005. The unholy trinity: taxonomy, species lation appears to have been gained independently on at least four delimitation and DNA barcoding. Phil. Trans. R. Soc. B. 360, 1905–1916. De Queiroz, K., 2007. Species concepts and species delimitation. Syst. Biol. 56, 879– different occasions in the phylogeny of Tubificoides. 886. However, the supporting bootstrap values for many of the De Vos, A.P.C., 1936. Chaetopoda. In: Redeke, H.C. (Ed.), Flora en fauna der nodes discussed above are very low (<50%) and this may be due Zuiderzee. Monografie van een brakwatergebied. De Boer, Den Helder, pp. 85– 96. to a recent radiation within the group. It should also be noted that Diaz, R.J., 1984. Short term dynamics of the dominant annelids in a polyhaline most of the deep-water taxa (which constitute the majority of temperate estuary. Hydrobiologia 115, 153–158. known Tubificoides species) were not included in this study and Diaz, R.J., 1989. Pollution and tidal benthic communities of the James River estuary, inclusion of these may still prove pivotal for increasing the support Virginia. Hydrobiologia 180, 195–211. Douady, C.J., Delsuc, F., Boucher, Y., Ford Doolittle, W., Douzery, J.P., 2003. of these problematic nodes. Comparison of bayesian and maximum likelihood bootstrap measures of phylogenetic reliability. Mol. Biol. Evol. 20, 248–254. Acknowledgments Dubilier, N., Giere, O., Grieshaber, M.K., 1995. Morphological and ecophysiological adaptations of the marine oligochaete Tubificoides benedii to sulphidic sediments. Am. Zool. 35, 163–173. We are greatly indebted to Robert J. Diaz (Virginia Institute of Edgar, R.C., 2004. MUSCLE: multiple sequence alignment with high accuracy and Marine Science) and Andrew S.Y. Mackie (National Museum Wales) high throughput. Nucleic Acids Res. 32, 1792–1797. Envall, I., Källersjö, M., Erséus, C., 2006. Molecular evidence for the non- for all their aid in collecting material; without their generous sup- monophyletic status of Naidinae (Annelida, Clitellata, Tubificidae). Mol. port, this project would have been immensely harder to conduct. Phylogenet. Evol. 40, 570–584. Author's personal copy

702 S. Kvist et al. / Molecular Phylogenetics and Evolution 57 (2010) 687–702

Erséus, C., 1975. Peloscolex amplivasatus sp.n. and Macroseta rarisetis gen. et sp.n. Jukes, T.H., Cantor, C.R., 1969. Evolution of protein molecules. In: Munro, H.N. (Ed.), (Oligochaeta, Tubificidae) from the West coast of Norway. Sarsia 58, 1–8. Mammalian Protein Metabolism. Academic Press, New York, pp. 21–123. Erséus, C., 1984. Cladistic analysis of the subfamilies within the Tubificidae Källersjö, M., von Proschwitz, T., Lundberg, S., Eldenäs, P., Erséus, C., 2005. (Oligochaeta). Zool. Scr. 19, 57–63. Evaluation of ITS rDNA as a complement to mitochondrial gene sequences for Erséus, C., 1985. Annelida of Saudi Arabia. Marine Tubificidae (Oligochaeta) of the phylogenetic studies in freshwater mussels: an example using Unionidae from Arabian Gulf coast of Saudi Arabia. Fauna of Saudi Arabia 6, 130–154. north-western Europe. Zool. Scr. 34, 415–424. Erséus, C., 1989. Two new species of the marine genus Limnodriloides and a record of Kimura, M., 1980. A simple method for estimating evolutionary rates of base Tubificoides fraseri Brinkhurst (Oligochaeta: Tubificidae) from New Zealand. NZ substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. J. Mar. Freshwat. Res. 23, 557–561. 16, 111–120. Erséus, C., 2005. Phylogeny of oligochaetous Clitellata. Hydrobiologia 535 (536), Kvist, S., Dreyer, J., Erséus, C., 2008. Two new species of Tubificoides (Annelida: 357–372. Clitellata: Naididae) from the Blake Ridge methane seep in the North-west Erséus, C., Bonomi, G., 1987. A new species of Tubificoides (Oligochaeta, Tubificidae) Atlantic Ocean. Proc. Biol. Soc. Wash. 121, 531–540. from the Adriatic Sea. Ital. J. Zool. 54, 165–168. Lerberg, S.B., Holland, F.A., Sanger, D.M., 2000. Responses of tidal creek Erséus, C., Diaz, R.J., 1989. Population dynamics of Tubificoides amplivasatus macrobenthic communities to the effects of watershed development. (Oligochaeta, Tubificidae) in the Oresund, Denmark. Hydrobiologia 180, 161– Estuaries 23, 838–853. 176. Lowenstein, J.H., Amato, G., Kolokotronis, S.-O., 2009. The real maccoyii: identifying Erséus, C., Envall, I., Marchese, M., Gustavsson, L., 2010. The systematic position of tuna sushi with DNA barcodes–contrasting characteristic attributes and genetic Opistocystidae (Annelida, Clitellata) revealed by DNA data. Mol. Phylogenet. distances. PLoS ONE 4, e7866. Evol. 54, 309–313. Meyer, C.P., Paulay, G., 2005. DNA barcoding: error rates based on comprehensive Erséus, C., Källersjö, M., Ekman, M., Hovmöller, R., 2002. 18S rDNA phylogeny of the sampling. PLoS Biol. 3, 2229–2238. Tubificidae (Clitellata) and its constituent taxa: dismissal of the Naididae. Mol. Michaelsen, W., 1926. Oligochäten aus dem Ryck bei Greifswald und von Phylogenet. Evol. 22, 414–422. benachbarten Meeresgebieten. Mitt. Hamb. Zool. Mus. Inst. 42, 21–29. Erséus, C., Kvist, S., 2007. COI variation in Scandinavian marine species of Milligan, M.R., 1991. Two new species of Tubificoides (Oligochaeta, Tubificidae) and Tubificoides (Annelida: Clitellata: Tubificidae). J. Mar. Biol. Assoc. UK 87, new records of T. Brownae and T. imajimai from the Gulf of Mexico and 1121–1126. Caribbean, with a redescription of T. bakeri. Zool. Scr. 20, 339–345. Erséus, C., Wetzel, M., Gustavsson, L., 2008. ICZN rules – a farewell to Tubificidae Nichols, F.H., Thompson, J.K., 1985. Persistance of an introduced mudflat (Annelida, Clitellata). Zootaxa 1744, 66–68. community in South San Fransisco Bay, California. Mar. Ecol. Prog. Ser. 24, Folmer, O., Black, M., Hoeh, W., Lutz, R., Vrijenhoek, R., 1994. DNA primers for 83–97. amplification of mitochondrial cytochrome c oxidase subunit I from diverse Norén, M., Jondelius, U., 1999. Phylogeny of the Prolecithophora (Platyhelminthes) metazoan invertebrates. Mol. Mar. Biol. Biotechnol. 3, 294–299. inferred from 18S rDNA sequences. Cladistics 15, 103–112. Frankham, R., 2005. Resolving the genetic paradox in invasive species. Heredity 94, Nylander, J.A.A., 2004. MrModeltest V2. Program distributed by the author, 385. Evolutionary Biology Centre. Uppsala University. Giere, O., Rhode, B., Dubilier, N., 1988. Structural peculiarities of the body wall of Nylander, J.A.A., Wilgenbusch, J.C., Warren, D.L., Swofford, D.L., 2008. AWTY (are we Tubificoides benedii (Oligochaeta) and possible relations to its life in sulphidic there yet?): a system for graphical exploration of MCMC convergence in sediments. Zoomorphology 108, 29–39. Bayesian phylogenetics. Bioinformatics 24, 581–583. Gustafsson, D.R., Price, D.A., Erséus, C., 2009. Genetic variation in the popular lab Palumbi, S.R., 1996. Nucleic acids II: the polymerase chain reaction. In: Hillis, D.M., worm Lumbriculus variegatus (Annelida: Clitellata: Lumbriculidae) reveals Mortiz, C., Mable, B.K. (Eds.), Molecular systematics, 2nd ed. Sinauer Associates cryptic speciation. Mol. Phylogenet. Evol. 51, 182–189. Sunderland, MA, pp. 205–247. Hajibabaei, M., Singer, G.A.C., Hebert, P.D.N., Hickey, D.A., 2007. DNA barcoding: Popescu-Marinescu, V., Botea, F., Brezeanu, G., 1966. Untersuchungen über die how it complements taxonomy, molecular phylogenetics and population Oligochaeten im rümanischen Sektor des Donaubassins. Arch. Hydrobiol. Suppl. genetics. Trends Genet. 23, 167–172. 30, 161–179. Halanych, K.M., Janosik, A.M., 2006. A review of molecular markers used for Annelid Rach, J., DeSalle, R., Sarkar, I.N., Schierwater, B., Hadrys, H., 2008. Character-based phylogenetics. Integr. Comp. Biol. 46, 533–543. DNA barcoding allows discrimination of genera, species and populations in Harrel, R.C., 2004. Systematic and ecological notes on Tubificoides heterochaetus Odonata. Proc. R. Soc. B. 275, 237–247. (Oligochaeta: Tubificidae) from the Neches River estuary, Texas. Texas J. Sci. 56, Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference 263–267. under mixed models. Bioinformatics 19, 1572–1574. Hebert, P.D.N., Cywinska, A., Ball, S.L., deWaard, J.R., 2003a. Biological identifications Rozas, J., Sánchez-DelBarrio, J.C., Messeguer, X., Rozas, R., 2003. DnaSP, DNA through DNA barcodes. Proc. R. Soc. Lond. B. 270, 313–321. polymorphism analyses by the coalescent and other methods. Bioinformatics Hebert, P.D.N., Ratnasingham, S., deWaard, J.R., 2003b. Barcoding animal life: 19, 2496–2497. cytochrome c oxidase subunit 1 divergences among closely related species. Sarkar, I.N., Planet, P.J., Bael, T.E., Stanley, S.E., Siddall, M., DeSalle, R., Figurski, D.H., Proc. R. Soc. Lond. B 270, 96–99. 2002a. Characteristic attributes in cancer microarrays. J. Biomed. Inform. 35, Hebert, P.D.N., Stoeckle, M.Y., Zemlak, S.T., Francis, C.M., 2004. Identification of birds 111–122. through DNA barcodes. PLoS Biol. 2, 1657–1663. Sarkar, I.N., Thornton, J.W., Planet, P.J., Figurski, D.H., Schierwater, B., DeSalle, R., Helgason, G.V., Erséus, C., 1987. Three new species of Tubificoides (Oligochaeta, 2002b. An automated phylogenetic key for classifying homeoboxes. Mol. Tubificidae) from the North-west Atlantic and notes on geographic variation in Phylogenet. Evol. 24, 388–399. the circumpolar T. Kozloffi. Sarsia 72, 159–169. Sjölin, E., Erséus, C., Källersjö, M., 2005. Phylogeny of Tubificidae (Annelida, Holmquist, C., 1978. Revision of the genus Peloscolex (Oligochaeta, Tubificidae). 1. Clitellata) based on mitochondrial and nuclear sequence data. Mol. Morphological and anatomical scrutiny; with discussion on the generic level. Phylogenet. Evol. 35, 431–441. Zool. Scr. 7, 187–209. Southern, R., 1909. Contributions towards a monograph of the British and Irish Holmquist, C., 1979. Revision of the genus Peloscolex (Oligochaeta, Tubificidae). 2. Oligochaeta. Proc. Roy. Irish Acad. 27, 119–182. Scrutiny of the species. Zool. Scr 8, 37–60. Swofford, D., 2002. PAUP*: phylogenetic analysis using parsimony (*and other Hovmöller, R., Pape, T., Källersjö, M., 2002. The Palaeoptera problem: basal methods). Version 4.0b. Computer software and manual: sinauer associates, pterygote phylogeny inferred from 18S and 28S rDNA sequences. Cladistics Sunderland, MA. 18, 313–323. Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: molecular evolutionary Hrabeˇ, S., 1966. New or insufficiently known species of the family Tubificidae. Spisy genetics analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599. Prirodoved Fak. Univ. J.E. Purkyne Brne. 470, 57–76. Thompson, B., Lowe, S., Kellogg, M. 2000. Macrobenthic assemblages of the San Huang, J., Xu, Q., Sun, Z.J., Tang, G.L., Su, Z.Y., 2007. Identifying earthworms through Francisco bay-delta, and their responses to abiotic factors. Results of the Benthic DNA barcodes. Pedobiologia 51, 301–309. Pilot Study 1994–1997 Part 1, pp. 1–41. International Commission of Zoological Nomenclature [ICZN], 2007. Opinion 2167 Voua Otomo, P., Jansen van Vuuren, B., Reinecke, S.A., 2009. Usefulness of DNA (Case 3305). Naididae Ehrenberg, 1828 (Annelida, Clitellata): precedence over barcoding in ecotoxicological investigations: resolving taxonomic uncertainties Tubificidae Vejdovsky, 1876 maintained. Bull. Zool. Nomen. 64, 71–72. using Eisenia Malm 1877 as an example. Bull. Environ. Contam. Toxicol. 82, Jamieson, B.G.M., Tillier, S., Tillier, A., Justine, J.-L., Ling, E., James, S., McDonald, K., 261–264. Hugall, A.F., 2002. Phylogeny of the Megascolecidae and Crassiclitellata White, T.M., Bruns, T., Lee, S., Taylor, J., 1990. Amplification and direct sequencing of (Annelida, Oligochaeta): combined versus partitioned analysis using nuclear fungal ribosomal RNA for phylogenetics. In: Innis, M.A., Gelfrand, D.H., Sninsky, (28S) and mitochondrial (12S, 16S) rDNA. Zoosystema 24, 707–734. J.J., White, T.J. (Eds.), PCR Protocols. Academic Press, San Diego, CA, pp. 315–322. Jaroschenko, M.F., 1948. Oligochaetes of the Dnieper-Bug Liman. Nauc Zap. Modlav. Wolff, W.J., 2005. Non-indigenous marine and estuarine species in the Netherlands. Nauc. Issl. Bazy Acad. Nauk. SSSR 1, 57–68. Zool. Meded. (Leiden) 79, 1–116. Johnson, J.B., Omland, K.S., 2004. Model selection in ecology and evolution. Trends Ecol. Evol. 19, 101–108.