Molecular Phylogenetics and Evolution 91 (2015) 41–55

Contents lists available at ScienceDirect

Molecular Phylogenetics and Evolution

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

Revisiting the age, evolutionary history and species level diversityof the genus Hydra (Cnidaria: Hydrozoa) q

http://dx.doi.org/10.1016/j.ympev.2015.05.0131055-7903/� 2015 Elsevier Inc. All rights reserved.

q This paper has been recommended for acceptance by Bernd Schierwater.⇑ Corresponding author. Tel.: +49 0431 880 4169.

E-mail address: tbosch@zoologie.uni-kiel.de (T.C.G. Bosch).

Martin Schwentner a,b, Thomas C.G. Bosch a,⇑a Zoological Institute and Interdisciplinary Centre Kiel Life Science, Christian-Albrechts-University of Kiel, 24098 Kiel, Germanyb Museum of Comparative Zoology, Harvard University, Cambridge, MA 02138, USA

a r t i c l e i n f o a b s t r a c t

Article history:Received 12 November 2014Revised 13 May 2015Accepted 15 May 2015Available online 23 May 2015

Keywords:Divergence timesEvolutionary historyHydraMolecular clockMolecular systematicsPhylogeny

The genus Hydra has long served as a model system in comparative immunology, developmental andevolutionary biology. Despite its relevance for fundamental research, Hydra’s evolutionary origins andspecies level diversity are not well understood. Detailed previous studies using molecular techniquesidentified several clades within Hydra, but how these are related to described species remained largelyan open question. In the present study, we compiled all published sequence data for three mitochondrialand nuclear genes (COI, 16S and ITS), complemented these with some new sequence data and delimitedmain genetic lineages (=hypothetical species) objectively by employing two DNA barcoding approaches.Conclusions on the species status of these main lineages were based on inferences of reproductive isola-tion. Relevant divergence times within Hydra were estimated based on relaxed molecular clock analyseswith four genes (COI, 16S, EF1a and 28S) and four cnidarians fossil calibration points All in all, 28 mainlineages could be delimited, many more than anticipated from earlier studies. Because allopatric distri-butions were common, inferences of reproductive isolation often remained ambiguous but reproductiveisolation was rarely refuted. Our results support three major conclusions which are central for Hydraresearch: (1) species level diversity was underestimated by molecular studies; (2) species affiliationsof several crucial ‘workhorses’ of Hydra evolutionary research were wrong and (3) crown group Hydraoriginated �200 mya. Our results demonstrate that the taxonomy of Hydra requires a thorough revisionand that evolutionary studies need to take this into account when interspecific comparisons are made.Hydra originated on Pangea. Three of four extant groups evolved �70 mya ago, possibly on the northernlandmass of Laurasia. Consequently, Hydra’s cosmopolitan distribution is the result of transcontinentaland transoceanic dispersal.

� 2015 Elsevier Inc. All rights reserved.

1. Introduction

Since its first discovery in 1702 and its use in experimentalstudies in the early 18th century (van Leeuwenhoek, 1702;Trembley, 1744), Hydra has been an important model organismfor studies on regeneration, development, pattern formation, sym-biosis and more recently also for genome evolution and innateimmunity. As a result of this work, and because Hydra belongs tothe basal animal phylum Cnidaria, studies on Hydra have con-tributed significantly to our understanding of the origin and evolu-tion of developmental genes and pathways, the evolution of theimmune system and the concept of the metaorganisms or holo-biont (Bosch, 2013, 2014; David, 2012; Franzenburg et al., 2013;

Fujisawa and Hayakawa, 2012; Galliot, 2012; Grimmelikhuijzenand Hauser, 2012; Holstein, 2012; Khalturin et al., 2009; Lasiet al., 2010; Meinhardt, 2012; Nebel and Bosch, 2012; Shimizu,2012; Steele et al., 2011; Tanaka and Reddien, 2011; Technauand Steele, 2011; Watanabe et al., 2009). One of the most impor-tant tools for identifying relevant genes is the genome of Hydramagnipapillata (Chapman et al., 2010) and the ever-increasing tran-scriptome datasets of other Hydra and cnidarian species (http://www.compagen.org).

Despite Hydra’s long history as a model organism for animalevolution, key features of the evolution of Hydra itself such asthe evolutionary origins of Hydra and its species level diversityare still not well understood. As a consequence, the pace and time-frame of crucial evolutionary processes and novelties, like theemergence of numerous orphan genes specific for Hydra or someof its species (Khalturin et al., 2008, 2009), cannot be assessed.Thus, as Hydra species differ in morphology, development,

42 M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55

physiology, and ecology (Campbell, 1983; Hemmrich et al., 2007;Koizumi, 2007), the lack of a solid species level taxonomy hampersresearch on species specific differences in gene expression anddevelopment (Khalturin et al., 2008; Thomsen and Bosch, 2006).

Moreover, in every animal species divergence in host genesseems positively correlated with differentiation of the microbiome(Fraune and Bosch, 2007; McFall-Ngai et al., 2013; Brucker andBordenstein, 2013; Bosch, 2014). Parallel cladograms between thehost phylogeny and the microbiome relationships is one test of apattern termed ‘‘phylosymbiosis’’ (Brucker and Bordenstein, 2012,2013). Each Hydra species is equipped with a unique composition ofantimicrobial peptides (Franzenburg et al., 2013). Loss-of-functionexperiments have shown that species-specific AMPs sculpturespecies-specific bacterial communities by selecting for co-evolvedbacterial partners (Franzenburg et al., 2013). Current lack of infor-mation on precise phylogenetic placement of the Hydra species,however, makes it impossible to investigate the diversity of theHydra specific microbiota within a co-evolutionary framework.

Finally, unambiguous phylogenetic placement and speciesidentification of commonly used lab strains of Hydra is crucial toascertain the reproducibility of experiments (requiring the usageof identical species) and to allow intra- and interspecificcomparisons.

Resolving the evolutionary origins of Hydra has been impededby the absence of fossilized remains. Recent phylogenetic analysesunambiguously placed Hydra within the hydrozoan taxonAplanulata (Nawrocki et al., 2013), but the age of Hydra and thetiming of its diversifications are largely unknown. Different molec-ular clock approaches diverged greatly in their estimates: while theage of the viridissima group was estimated to be 156–174 millionyears based on the divergence of its symbiotic Chlorella (Kawaidaet al., 2013), the age of crown group Hydra was estimated to beonly �60 million years based on assumed substitution rates forCOI and 16S (Martínez et al., 2010).

The taxonomy of Hydra species is characterized by relativelylarge numbers of synonymizations and wrongly applied speciesnames (e.g. Hydra attenuata; see Campbell, 1989). Well acceptedand supported by molecular phylogenetic studies (Hemmrichet al., 2007; Kawaida et al., 2010; Martínez et al., 2010) is the dis-tinction of four species groups within Hydra – viridissima group(the ‘green’ Hydra featuring symbiotic Chlorella), braueri group, oli-gactis group and vulgaris group – which were outlined by Schulze(1917) and Campbell (1987). Over the last centuries �80 speciesof Hydra have been described (Jankowski et al., 2008). Many ofthese species have been subsequently synonymized and the taxo-nomic status of others is still controversial (Campbell, 1987). Forexample, Jankowski et al. (2008) suggested less than 15 valid spe-cies of Hydra whereas the World Register of Marine Species lists 40(Schuchert, 2014).

In the last years several molecular phylogenetic studies shedlight on the diversity within Hydra (Campbell et al., 2013;Hemmrich et al., 2007; Kawaida et al., 2010; Martínez et al.,2010; Reddy et al., 2011; Wang et al., 2012). The most detailedstudies regarding the number of studied individuals (=strains)were those by Kawaida et al. (2010) and Martínez et al. (2010).However, since these two studies were published nearly simulta-neously they could not take each other’s data into account andlater studies included small fractions of the previously publisheddata only. Apart from a few shared lab strains, all of these studieswere based on different sets of strains from different species andgeographic origins, limiting comparability among studies.Furthermore, despite identifying several monophyletic cladeswithin Hydra and within its four groups, few explicit conclusionsregarding the validity of Hydra species were drawn. For example,Kawaida et al. (2010) identified three monophyletic ‘‘sub-groups’’within the vulgaris group, but it remained unclear whether these

would represent three species or three clades of several specieseach. Martínez et al. (2010) recovered eight morphologicallyidentified species as monophyletic (H. viridissima, H. hymanae,H. utahensis, H. circumcincta, H. oligactis, H. oxycnida, H. canadensis,and H. vulgaris), of which H. viridissima was subdivided into severalclades, H. circumcincta into two clades and H. vulgaris into fiveclades matching geographic regions. Again, for most of these cladestheir species status was not discussed, the exception being theNorth American vulgaris group species – H. littoralis, H. carnea,and H. vulgaris AEP – which were believed to belong to a singlespecies. The latter would have far reaching consequences for evo-lutionary studies on Hydra as H. carnea and H. vulgaris AEP arecommonly used lab strains. Hydra vulgaris AEP is an artificially gen-erated strain from which all transgenic Hydra are derived (Wittliebet al., 2006). The Hydra vulgaris AEP strain originated from crossingtwo different Hydra vulgaris strains from North America (Martinet al., 1997), though Martínez et al. (2010) stated that the parentalstrains resembled H. carnea and H. littoralis morphologically.Similarly, two other important ‘workhorses’ – H. magnipapillataand the European H. vulgaris – are genetically very similar andwere assigned to the same clade (Kawaida et al., 2010; Martínezet al., 2010).

To obtain a comprehensive overview of the species diversityand the evolutionary origins of Hydra, we compiled all availablesequence data of Hydra from GenBank and BOLD for those geneswith the highest coverage of species and individuals (e.g. thosepublished by Campbell et al., 2013; Hemmrich et al., 2007;Kawaida et al., 2010; Martínez et al., 2010; Reddy et al., 2011and Wang et al., 2012). This extensive dataset was complementedwith some newly sequenced strains (mainly of the viridissimagroup). Species diversity was assessed by a combination of phylo-genetic and genetic distance analyses. Such analyses have becomevery popular tools for studies of species diversities and are com-monly subsumed under the term DNA barcoding. However,whether entities delimited by such approaches indeed representdistinct species is strongly dependent on the applied species con-cept (Agapow et al., 2004; Schwentner et al., 2011; Tan et al.,2008). For example, following the Phylogenetic Species Concept(Mishler and Theriot, 2000), which defines species as the ‘‘smallestmonophyletic groups worthy of formal recognition’’, all entitiesdelimited by DNA barcoding can be treated as species, if speciesdelimitation is based on phylogenetic analysis. The BiologicalSpecies Concept (Mayr, 1942), on the other hand, requires repro-ductive isolation among species. Reproductive isolation can beinferred with barcoding techniques in an integrative framework,if the respective species are consistently differentiated bymitochondrial and an independent marker system – like nuclearmarkers or morphology – and occur in sympatry (see alsoSchwentner et al., 2015).

As a first step, we identified ‘main lineages’ (=hypothetical spe-cies) by independently analyzing three molecular markers: mito-chondrial COI (cytochrome c oxidase subunit I), mitochondrial16S and nuclear ITS region (internal transcribed spacer; spanningITS1, 5.8S and ITS2). Two different analytical approaches wereemployed for each marker: Automatic Barcode Gap Discovery(ABGD; Puillandre et al., 2011), which identifies barcoding gapsin genetic distance matrixes, and the general mixed Yule coales-cent model (GMYC; Pons et al., 2006), which identifies speciesthresholds based on changes in branching rates in a phylogenetictree. The results are then evaluated to identify main lineages,which are consistently delimited across markers and analyticalmethods. These were then assessed under the different speciesconcepts. The age of crown group Hydra (monophylum comprisingall extant and ‘internal’ extinct species) and the timing of diversi-fication events within Hydra were assessed by molecular clockanalyses based on four molecular markers – COI, 16S, EF1a

M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55 43

(elongation factor 1-alpha) and 28S – and previously reported fos-silized cnidarians as calibration points. These results establish botha molecular phylogeny and a biogeographic history of the genusHydra.

2. Material and methods

2.1. DNA isolation, PCR amplification, sequencing and alignments

Total genomic DNA of several strains, which are cultured per-manently in the Bosch lab (Table 1), was isolated using theDNeasy Blood and Tissue kit (Qiagen, Germany) following manu-facturer’s instructions. Depending on their sizes, one or two indi-viduals (clones) were used per extraction. PCR reactions of allgene fragments comprised 4 ll of each primer with 10 lM each(Table 2), 0.4 ll dNTPs (100 mM each), 8 ll 5� Colorless GoTaqReaction Buffer (Promega) and 0.2 ll GoTaq polymerase(Promega) in a total volume of 40 ll. PCR amplifications had an ini-tial denaturation step at 94 �C for 2 min; followed by 35 cycles of2 min at 94 �C, 1 min annealing (COI and 16S: 47 �C, ITS: 55 �C)and 1:30 min at 72 �C (16S 1 min) and a final elongation step of10 min at 72 �C. Because of the length of the amplified productinternal primer were derived for ITS to ensure sequencing of thewhole length (Table 2). Annealing temperature was 52 �C for theinternal ITS primer. Success of PCR was assessed on 1.5%TAE/Agarose gel stained with peqGreen DNA/RNA Dye (Peqlab).PCR product purification and Sanger sequencing were performedby the Institute of Clinical Molecular Biology in Kiel. Each genefragment was sequenced bi-directionally using the same primersas for the PCR amplifications. Sequences were manually correctedand submitted to GenBank (accession numbers are provided inTable 1). Alignments for each of the gene fragments including allsequences available from GenBank and BOLD were computed withMAFFT 7.1.4 (Katoh and Standley, 2013) using standard settings.The ITS alignment featuring all available Hydra sequences con-tained a very large fraction of Indels which greatly reduced thequality of the alignment. Therefore, ITS alignments were createdseparately for each of the four main groups of Hydra, whose recip-rocal monophyly is well supported (e.g. Hemmrich et al., 2007;Kawaida et al., 2010; Martínez et al., 2010).

2.2. Identification of main genetic lineages

Main lineages were delimited based on the three markers forwhich large datasets were available: COI, 16S and ITS. Each markerwas analyzed independently; for ITS all analyses were performedindependently for each of the main groups (see Section 2.1).

Table 1Strains newly sequenced for this study. Species names refer to species identification prior tbrackets. Locality details are provided as far known, GenBank accession numbers are prov

Specimen/strain Species Locality deta

Hy1 NC64 H. viridissima (6; H. sinensis)Hy2 Würzburg H. viridissima (6; H. sinensis)Hy3 M9 H. viridissima (5) Aquarium inHy4 A250 H. viridissima (4) New South WHy5 Husum H. viridissima (3) Husum, GerHy6 M120 H. viridissima (5) MadagascarHy7 M10 H. viridissima (5) JapanHy8 A99 H. viridissima (3) Queensland,Hy10 A14c H. viridissima (6; H. sinensis)Hy11 reg16–105 H. magnipapillata (H. vulgaris 7) JapanHy16 reg-16 H. magnipapillata (H. vulgaris 7) JapanHy17 St. Petersburg H. oligactis St. PetersburHy18 L7 H. robustaHy19 Pohlsee H. oxycnida Pohlsee nearHy20 M7 H. circumcincta Japan

Strains, for which two or all three markers were available, facili-tated the identification of consistently delimited main lineagesacross markers.

Phylogenetic trees were reconstructed including all availablesequences of each marker to assess the monophyly of potentialmain lineages. Trees were reconstructed with MrBayes 3.2(Ronquist and Huelsenbeck, 2003) with four chains of for 6 ⁄ 106

generations sampling every 1200th generation. The first 10% werediscarded as burn-in. The best fitting model was determined withMEGA6 (Tamura et al., 2013). All trees were visualized with FigTreev1.4.2 (Rambaut, 2014).

Main lineages were first assessed using the general mixed Yulecoalescent model (GMYC; Pons et al., 2006), which partitions anultrametric tree by identifying changes in branching rates. Therequired ultrametric trees were generated with BEAST version2.1.3 (Bouckaert et al., 2014) employing a Yule speciation priorand a relaxed molecular clock. Analyses were run for 5 ⁄ 107 gen-erations, sampling every 5000th generation and discarding the first10% as burn-in. Trees were annotated with TreeAnnotator v2.1.2(part of the BEAST package). The single threshold GMYC modelwas then fitted onto the resulting tree using the GMYC web inter-face (http://species.h-its.org/gmyc/).

As a second method to delimit main lineages the AutomatedBarcode Gap Discovery (ABGD; Puillandre et al., 2012) wasemployed. Based on a pairwise genetic distance matrix sequencesare partitioned into main lineages, which are separated by a poten-tial barcoding gap. Across the whole range of genetic distancesmore than one putative barcoding gap (and respective partition)may be identified. Herein we focused on those derived barcodinggaps that resulted in similar partitions across markers and whichagreed with known levels of intra- and interspecific genetic dis-tances (Dawson, 2005a,b; Holland et al., 2004; Laakmann andHolst, 2014; Moura et al., 2008; Ortman et al., 2010; Pontin andCruickshank, 2012; Schroth et al., 2002). Genetic distances werecalculated as uncorrected p-distances with MEGA6 (Table 3). Theweb interface of ABGD (http://wwwabi.snv.jussieu.fr/public/abgd/) was used, setting the number of steps to 100 and the relativegap width to 0.1; otherwise standard settings were kept.

2.3. Molecular clock analyses

Molecular clock analyses were performed with the BEAST pack-age version 2.1.3 (Bouckaert et al., 2014) using two mitochondrial(COI and 16S) and two nuclear (EF1a and 28S) markers. The ITSregion was not used because of difficulties aligning sequences ofdistantly related cnidarian species. For each main lineage onesequence per gene was included, as far as possible these were from

o molecular genetic analyses, species assignment after molecular analyses provided inided for all sequenced gene fragments.

ils COI 16S ITS

KP895128 KP895104 KP895137KP895121 KP895110 KP895136

Jerusalem, Israel KP895122 KP895111 KP895135ales, Australia KP895123 KP895112 KP895138

many KP895124 KP895113 KP895142KP895125 KP895114 KP895140KP895126 KP895115 KP895141

Australia KP895127 KP895116 KP895143KP895129 KP895105 KP895139– – KP895130KP895117 KP895106 KP895131

g, Russia KP895118 KP895107 KP895133KP895119 KP895108 –

Kiel, Germany KP895120 KP895109 KP895134– – KP895132

Table 2List of all primer used in this study and its respective reference.

COICO1.Dawson.F.LCOjf GGTCAACAAATCATAAAGATATTGGAAC Martínez et al. (2010)CO1.Folmer.R.HC02198 TAAACTTCAGGGTGACCAAAAAATCA Martínez et al. (2010)

16S rRNAForward TCGACTGTTTACCAAAAACATAGC Martínez et al. (2010)Reverse ACGGAATGAACTCAAATCATGTAAG Martínez et al. (2010)

ITSITS_18S CACCGCCCGTCGCTACTACCGATTGAATGG Martínez et al. (2010)ITS_28S CCGCTTCACTCGCCGTTACTAGGGGAATCC Martínez et al. (2010)ITS_rev_intern GGACARAAGCRAGTTGAGCCTTCCG This study

44 M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55

the same individuals. Only Hydra main lineages for which atleast two of these genes were available were included in theanalyses. Several other cnidarians and the sponge Amphimedonqueenslandica were included to employ calibration points derivedfrom known fossils to date the resulting phylogenetic tree. Carewas taken to include members of all relevant cnidarians main taxa,with special emphasis of representatives of the Aplanulata. Hydrais thought to belong to Aplanulata (Nawrocki et al., 2013) andincluding as many other Aplanulata as possible is important toinfer the potential sister group of Hydra and to assess Hydra’sdivergence time. Because large sections with numerous Indelswere present in the alignments of 16S and 28S Gblocks (Talaveraand Castresana, 2007) was used to identify and removed ambigu-ously aligned stretches of the alignment. Gblocks was run withstandard settings of the online version with all ‘less stringent’parameters enabled. This led to the removal of �25% of the 16Sand �7% of the 28S alignment.

Calibration points were chosen based on the oldest known fossilcnidarians assigned to extant cnidarian main taxa. Because thesefossils may represent the stem group of the respective main taxa,we calibrated the split between the taxon represented by the fossiland its putative sister group included in the phylogenetic analysisand used the ages of these fossils as the minimum age for therespective calibration points. This conservative coding strategyshould result in underestimations of divergence times rather thanoverestimations. The calibration points were: 570 mya for allstudied Cnidaria (cnidarian microfossils like putative embryos;see Chen et al., 2002; Xiao et al., 2000), 540 mya for the splitbetween Hydrozoa and Scyphozoa (see Waggoner and Collins(2004), Young and Hagadorn (2010) suggested an age of only505 mya, but Cartwright and Collins (2007) identified numerouscrown groups of both taxa for this time period, suggesting anearlier diversification); 500 mya for Hydrozoa (see Young andHagadorn, 2010) and 450 mya for the split between Leptomedusaand Anthomedusa (see Young and Hagadorn, 2010). All these wereassigned uniform priors with the maximum age constrained to640 mya. The maximum age is based on the earliest fossilizedmetazoan biomarkers (i.e. of sponges) in the late Cryozoan (Loveet al., 2009) and coincides with the estimated divergence timebetween sponges and cnidarians (Peterson et al., 2004). Otherstudies employing molecular clocks suggested earlier origins ofCnidaria (up to 2000 mya), which is not mirrored in the fossilrecord (Park et al., 2012; Waggoner and Collins, 2004; see alsoPeterson et al., 2004). We did not employ such extreme ages asthese may lead to overestimations of Hydra divergence times.Four nodes were constrained as monophyletic – i.e. Cnidaria,Hexacorallia, Hydrozoa and Medusozoa. To assess the impact ofthe different genes and taxa included, we re-ran the analyses withtwo different partitions: (1) including the same taxa but only thetwo mitochondrial markers (COI and 16S) and (2) including all fourgenes but reducing the number of Hydra and other Aplanulata spe-cies (including only H. viridissima 3, H. vulgaris 4, H. circumcincta 1,

Candelabrum cocksii and Ralpharia gorgoniae). The first of theseadditional analyses was performed to assess whether the mito-chondrial genes estimates deviate from estimates of the wholedata set (a separate analysis for the nuclear genes was not per-formed as the amount of missing data was higher for these genes).The number of taxa was reduced to ascertain that the overrepre-sentation of Aplanulata and especially of Hydra species in the dataset did not affect divergence time estimates.

All molecular clock analyses were performed with BEAST ver-sion 2.1.3 employing a relaxed log normal clock with a Yule prior.The best fitting substitution model was determined for each genewith MEGA6, these was the GTR + I + G for COI, 28S and EF1aand the GTR + G for 16S. The MCMC chains were run twice for100 million generations, storing every 5000th trees (20.000 treesin total). The tree files of both runs were combined withLogCombiner (part of the BEAST package). Convergence of runswas assessed with Tracer v1.6 (Rambaut et al., 2013) and treeswere generated with the burn-in removed with TreeAnnotatorv2.1.2 (part of the BEAST package). The resulting trees were visual-ized with FigTree v1.4.2 (Rambaut, 2014).

3. Results

3.1. Resolving phylogenetic relationships of the Hydra group

The phylogenetic analyses of COI and the combined data setsupport the four Hydra groups: viridissima group, braueri group,oligactis group and vulgaris group (Figs. 1 and 2; SupplementFigs. 1–4). Only the analyses of 16S alone did not fully recoverthese groups, probably due to overall lower resolution of the tree(Fig. 2, Supplement Fig. 2).

viridissima group: Within the viridissima group six main geneticlineages were delimited by COI and 16S with H. viridissima 5 and 6each being further subdivided in analyses of ITS (Figs. 1 and 2;Supplement Figs. 1, 2 and 4). Furthermore, some individuals ofthese two main lineages shared an identical ITS sequence (Tables3 and 4) and H. viridissima 5 and 6 are the main lineages withthe lowest interlineage distances in COI (3.9–4.1%) and 16S (1.5–3.0%), while all others exceed 10.5% and 3.5% in this group,respectively.

braueri group: Four main lineages were consistently differenti-ated within the braueri group: H. circumcincta 1, H. circumcincta 2,H. hymanae and H. utahensis (Figs. 1 and 2; Supplement Figs. 1, 2and 4; Tables 3 and 4). The H. circumcincta 1 lineage was furthersplit into an European and an Alaskan lineage (the latter repre-sented by only a single individual) by ABGD of 16S (thresholdbelow 1.4%) and COI (threshold below 1.3%) and GMYC of COI.However, the ITS sequence of this individual was well nestedwithin the other sequences and was not differentiated. Hydrahymanae was split into two lineages in the GMYC of 16S, but notin any other analysis (Table 4).

Table 3Genetic distances among main lineages. Uncorrected p-distances for each single marker are reported within (along diagonal) and among (above diagonal) all main lineages of (a)viridissima group, (b) oligactis group, (c) braueri group and (d) vulgaris group. From top to bottom in each cell: COI, 16S and ITS (see also bottom let of each table). Main lineagespairs that occur sympatrically (i.e. the same region) are highlighted in gray. Individuals identified as H. oligactis that were genetically closer related to H. robusta were assigned toH. robusta in this representation. H. vulgaris strain 1005 is not included in this table due to uncertainties regarding its assignment. It shares an identical COI sequence with H.vulgaris 4 (H. carnea) but its ITS sequence differs by 1.9–2.5% from other members of this group and >1.1% from members of H. vulgaris 5 and 6. H. vulgaris 7 is probably the ‘true’ H.vulgaris Pallas (1766), H. vulgaris 4 the ‘true’ H. carnea Agassiz (1851) and H. viridissima 6 the ‘true’ H. sinensis Wang et al. (2009).

(continued on next page)

M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55 45

Table 3 (continued)

46 M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55

oligactis group: Within the oligactis group H. canadensis andH. oxycnida were well differentiated from each other and all othermain lineages, but were each further subdivided in a few analyses(Figs. 1 and 2; Supplement Figs. 1, 2 and 4; Table 4). Hydra robustawas usually not delimited as a separate species, but grouped withsome or all representatives of H. oligactis. Even though they did notshare identical DNA sequences, the observed genetic distancesbetween representatives of both nominal species were lower thanamong most other main lineages (Table 3).

vulgaris group: The vulgaris group is by far the most complexand diverse group. 14 different main lineages were differentiated,but not all three markers were available for each of the main lin-eages (Figs. 1 and 2; Supplement Figs. 1–3; Tables 3 and 4). Mostindividuals of this group were originally assigned to the speciesH. vulgaris. Of the other species only H. polymorphus and H. zhujian-gensis were consistently differentiated. Individuals identified as H.carnea grouped with H. litoralis and some H. vulgaris individuals(main lineage H. vulgaris 4; this main lineage includes the H. vul-garis AEP strain); Japanese H. magnipapillata and H. japonica clus-tered with European and South African H. vulgaris and H.attenuata individuals (main lineage H. vulgaris 7).

There is only one case of incongruence among the three studiedgenetic markers in the vulgaris group and this related to main lin-eages H. vulgaris 4, 5 and 6 which all occur in North America. Oneindividual (strain 1005a from Wyoming) is well nested within H.vulgaris 4 in the analyses of COI and 16S, but in ITS this individualis in a sister group relationship to H. vulgaris 5 (SupplementFigs. 1–3). The closer relationship to H. vulgaris 5 is mirrored ingenetic distances of ITS which are 1.9–2.5% toward all other mem-ber of H. vulgaris 4 and 1.1–1.5% to members of H. vulgaris 5 and1.4–1.8% toward H. vulgaris 6 (for H. vulgaris 6 only ITS is available).Among all three of these lineages genetic distances in ITS are lowerthan among most other main lineages, however, this is not the casefor COI and 16S (Tables 3 and 4).

Several main lineages have been recorded only from geograph-ically restricted areas: e.g. H. vulgaris 1, 2 and 3 from SouthAmerica; H. vulgaris 10 from South Africa; H. vulgaris 9 fromIceland; H. vulgaris 11 and 12 from Australia and New Zealand;H. vulgaris 8 from Switzerland (Supplement Figs. 1–3). As a conse-quence, only few main lineages of the vulgaris group can be consid-ered to occur sympatrically (Table 3). Main lineages H. vulgaris 4 &7 as well as H. vulgaris 7 & 10 and H. vulgaris 1 & 2 appear to besympatrically distributed and all of these lineage pairs are consis-tently differentiated in the mitochondrial and nuclear genes(Supplement Figs. 1–3; Tables 3 and 4), thus there is no indicationof on-going reproduction among them.

3.2. Establishing a time scale for Hydra evolution

We used known Cnidarian fossils (e.g. Chen et al., 2002; Loveet al., 2009; Waggoner and Collins, 2004; Xiao et al., 2000; Youngand Hagadorn, 2010) as calibration points to infer divergence timesamong Hydra lineages and accounted for rate heterogeneity amonglineages by applying relaxed-clock models. Our divergence dateestimates indicate for the crown group Hydra an age of 193 millionyears ago (mya) (95% highest posterior density [HPD] interval 149–241 mya; Fig. 3), representing the split between the viridissimagroup and all other Hydra. The inferred sister taxon (Candelabrumcocksii) diverged 326 mya (254–404). Within Hydra the viridissimagroup has an estimated age of 112 mya (77–152 mya), the brauerigroup of 69 mya (39–110 mya), the oligactis group of 71 mya (43–102 mya) and the vulgaris group of 62 mya (35–98 mya). The lattertwo groups diverged about 101 mya (72–132 mya) and in turn sep-arated from the braueri group 147 mya (108–187 mya; Fig. 3;Table 5).

The two additional molecular clock analyses, either with onlymitochondrial genes or with less Aplanulata and Hydra species,resulted in relatively similar estimates of divergence times with

Fig. 1. Bayesian inference phylogenetic analyses of COI. Nodes were collapsed to depict the inferred main lineages of Hydra only (refer to Supplement Fig. 1 for fullrepresentations showing all individuals and their geographic origins). Please note that H. vulgaris 7 is probably the ‘true’ H. vulgaris Pallas, 1766 (and includes all individualsidentified as H. magnipapillata), H. vulgaris 4 the ‘true’ H. carnea Agassiz (1851) and H. viridissima 6 the ‘true’ H. sinensis Wang et al. (2009). The outgroup has been removed fora better graphic representation. ⁄⁄Posterior probability 1; ⁄Posterior probability >0.95.

M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55 47

all divergence time estimates being well within the 95% HPD rangeof the initial analysis (Table 5). When fewer taxa were included thedivergence estimates for crown group Hydra and the split betweenthe braueri and the oligactis plus vulgaris group were about 10%(�20 million years) younger. The analysis based solely on thetwo mitochondrial markers also resulted in a �20 million yearyounger divergence time for the split between the braueri andthe oligactis plus vulgaris group as well as for the split betweenthe oligactis and vulgaris groups. Conversely, the age of the viridis-sima group was estimated to be 30 million years older (Table 5).The divergence times between crown group Hydra and its inferredsister groups were increased by �30 million years in the analysiswith fewer genes and decreased by �30 million years in the anal-ysis with less taxa (see Table 6).

4. Discussion

Our study for the first time compiled all the available data andemployed objective delimitation approaches to study the specieslevel diversity of Hydra. With 28 delimited main lineages weuncovered a much greater diversity of genetic lineages, whichmay correspond to species, than suggested by previous studies thatwere based on subsets of the present dataset (Campbell et al.,2013; Hemmrich et al., 2007; Kawaida et al., 2010; Martínezet al., 2010; Reddy et al., 2011; Wang et al., 2012). This underlinesthe importance of (1) taking into account all available data and (2)to employ specific and objective criteria for delimiting lineages

which may subsequently be translated into species. In the follow-ing we will discuss the species identities of the delimited main lin-eages and we will explicitly discuss potential taxonomicimplications (e.g. synonymies).

All 28 main lineages identified herein were differentiated in atleast one of the studied markers. Consequently, all main lineagescan be assumed to fulfill the species criterion of the PhylogeneticSpecies Concept (Mishler and Theriot, 2000) and may thus be trea-ted as phylogenetic species. This is corroborated by a comparisonwith intra- and interspecific genetic distances observed in otherMedusozoa, which are of similar extend as those within and amongHydra main lineages (e.g. Dawson, 2005a,b; Holland et al., 2004;Laakmann and Holst, 2014; Moura et al., 2008; Ortman et al.,2010; Pontin and Cruickshank, 2012; Schroth et al., 2002). A moreconservative approach would be to identify reproductively isolatedmain lineages and to accept only these as species: main lineagesthat are in concordance with the Biological Species Concept(Mayr, 1942). Because of the overall congruence between the twomitochondrial markers and the nuclear ITS, the absence of recentreproduction among most main lineages can be postulated. Truereproductive isolation, however, can only be inferred for thosemain lineage pairs that occur in sympatry as only these had thechance to reproduce, for all others a delimitation based on theBiological Species Concept has to remain ambiguous for now (seeSections 4.1–4.4 for more details).

Our results highlight several instances where the delimitedmain lineages and inferred biological species (see Sections 4.1–

Fig. 2. Bayesian inference phylogenetic analyses of (A) 16S and (B)–(E) ITS. Nodes were collapsed to depict the inferred main lineages of Hydra only (refer to SupplementFigs. 2–4 for full representations showing all individuals and their geographic origins). For ITS analyses were run independently for each group: (B) vulgaris group, (C)viridissima group, (D) oligactis group and (E) braueri group. Please note that H. vulgaris 7 is probably the ‘true’ H. vulgaris Pallas (1766) (and includes all individuals identified asH. magnipapillata), H. vulgaris 4 the ‘true’ H. carnea Agassiz (1851) and H. viridissima 6 the ‘true’ H. sinensis Wang et al. (2009). All phylogenetic trees are mid-point rooted.⁄⁄Posterior probability 1; ⁄Posterior probability >0.95.

48 M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55

4.4) are at odds with current taxonomy. This seems particularly tobe the case in the viridissima and vulgaris groups in which numer-ous species were subsumed under a single species name, i.e. H.viridissima and H. vulgaris (see Sections 4.1 and 4.4 for details). Inaddition, several individuals of the vulgaris group – which wereidentified as different species – obviously belong to the same spe-cies, suggesting synonymy of the respective species. Whether theunderlying taxonomy is erroneous and in need of revision orwhether it is only wrongly applied (i.e. misidentifications of indi-viduals) cannot be answered in all cases, as we had to rely onthe species identifications provided by the respective studies.Nevertheless, at least in the viridissima group several of the delim-ited species could be assigned to previously described speciesbased on the species’ geographic distribution, suggesting inconsis-tent application of current taxonomy (see Section 4.1). The samemay be true for the vulgaris group as numerous species have beendescribed for this group but these are not represented in the cur-rent dataset (see Section 4.4). Assigning individuals (or theirclones) to putative species by molecular phylogenetic analysesprior to morphological investigation can greatly improve morpho-logical species differentiation and may result in a clearer distinc-tion between intraspecific variability and interspecific variation.Such approaches have been very successful in other taxawhere consistent morphological differentiation was shown for

supposedly cryptic species (e.g. Pepper et al., 2011; Schroth et al.,2002; Schwentner et al., 2012). The results of this and previousmolecular phylogenetic studies (e.g. Campbell et al., 2013;Hemmrich et al., 2007; Kawaida et al., 2010; Martínez et al.,2010; Wang et al., 2012) are an ideal starting point for futureanalyses combining morphological and molecular phylogeneticanalyses to revise the taxonomy and species descriptions ofHydra. Given the restricted geographic distribution of severalmain lineages delimited herein, it appears plausible that only asubset of Hydra’s total diversity has been included in molecularanalyses so far. Fine-scale sampling may reveal an even greaterdiversity.

The phylogenetic analyses support the four groups proposed bySchulze (1917) and Campbell (1987), which were also recovered inearlier phylogenetic studies (Hemmrich et al., 2007; Kawaida et al.,2010; Martínez et al., 2010). The lack of support in the analyses of16S may be an artifact cause by the overall low level of resolutionprovided by this marker in the present analyses. Their reciprocalmonophyly highlights the usefulness of morphological characteris-tic to delimit Hydra species groups. Additional genus names wereproposed for the viridissima group – Chlorohydra (Schulze, 1914)– and Pelmatohydra (Schulze, 1914) for some members of the oli-gactis group. These two genera were rejected by several otherauthors (e.g. Ewer, 1948; Hyman, 1929; Jankowski et al., 2008;

Table 4Summary of main lineages delimited by different markers and methods. The presence (X) or absence (–) of all main lineage is summarized across all markers and methods. H. vulgaris 7 is probably the ‘true’ H. vulgaris Pallas (1766), H.vulgaris 4 the ‘true’ H. carnea Agassiz (1851) and H. viridissima 6 the ‘true’ H. sinensis Wang et al. (2009).

Phylogenetic analyses

COI 16S ITS COI GMYC COI ABGD (1.1–1.29%) COI ABGD (1.3–3.7%) 16S GMYC 16S ABGD (0.5–0.97%) 16S ABGD (0.3–0.48%) ITS GMYC ITS ABGD

H. viridissima 1 X X X X X X X X X X XH. viridissima 2 X X X X X X X X X X XH. viridissima 3 X X X X X X X X X X XH. viridissima 4 X X X X X X X X X X XH. viridissima 5 X X – (CO) X X X X X X – (CO) – (FS, CO)H. viridissima 6 X X – (CO) X X X X X X – (FS, CO) – (FS, CO)H. circumcincta 1 X X X – (FS) – (FS) X – (FS) – (FS) – (FS) X XH. circumcincta 2 X X X X X X X X X XH. utahensis X X X X X X X X X X XH. hymanae X X X X X X – (FS) X X X XH. oligactis – (CO) – (CO) X – – – (CO) – – (CO) – (CO) – (CO) – (CO)H. robusta – (CO) X X – (CO) – (CO) – (CO) – (CO) – (CO) – (CO) – (CO) – (CO)H. canadensis X X X X X X – (FS) – (FS) – (FS) – (FS) XH. oxycnida X X X – (FS) X X X X X X XH. vulgaris 1 X X X X X X X – (CO) X – (CO) – (CO)H. vulgaris 2 X X X X – (FS) – (CO) – (FS) – (FS) – (FS) X XH. vulgaris 3 X X X X X X X X X – (CO) – (CO)H. vulgaris 4 X X – X X X X – (CO) X – – (FS)H. vulgaris 5 X X X X X X X – (CO) X – (CO) XH. vulgaris 6 n.a. n.a. X n.a. n.a. n.a. n.a. n.a. n.a. X XH. vulgaris 7 X X X X X – (CO) – (FS) – (CO) X X XH. vulgaris 8 X n.a. n.a. X X X n.a. n.a. n.a. n.a. n.a.H. vulgaris 9 X n.a. X X X – (CO) n.a. n.a. n.a. X XH. vulgaris 10 X X X X (FS) – (FS) – (FS) X X X X XH. vulgaris 11 n.a. n.a. X n.a. n.a. n.a. n.a. n.a. n.a. X XH. vulgaris 12 n.a. X X n.a. n.a. n.a. X X X X – (FS)H. polymorphus X X n.a. X X X X X X n.a. n.a.H. zhujiangensis X n.a. n.a. X X X n.a. n.a. n.a. n.a. n.a.

CO = clusters with other lineages (all or some individuals cluster with individuals of one or several other lineages; in case of H. oligactis and H. robusta this includes always member of the respective other lineages); FS = furthersubdivision into two to three main lineages; n.a. = not applicable (no sequence of this marker was available for the respective main lineage).

M.Schw

entner,T.C.G.Bosch

/Molecular

Phylogeneticsand

Evolution91

(2015)41–

5549

Fig. 3. Molecular clock dated phylogenetic tree. Tree reconstruction is based on four genes, COI, 16S, EF1a and 28S, and the application of four cnidarians fossil calibrationpoints. Bars represent 95% highest posterior densities (HPD) intervals. All main lineages of Hydra which may correspond to distinct species and for which at least two of thefour genes were available were included in the analysis. Numbers indicate the four implemented fossil calibration points, (1) all Cnidaria (570–640 mya), (2) split betweenHydrozoa and Scyphozoa (540–640 mya), (3) Hydrozoa (500–640 mya) and (4) split between Anthomedusa and Leptomedusa (450–640 mya). ⁄⁄Posterior probability 1;⁄Posterior probability >0.95.

Table 5Estimated divergence times. For a set of nodes relevant in the evolution of Hydra therespective divergence times are reported for all three molecular clock analyses. Allestimates are in million years ago (mya) and include the 95% HPD interval; see alsoFig. 3.

All taxa,all genes

All taxa,COI+16S

Less taxa,all genes

Split between Hydraand its sister taxon

326 (254–404)a 358 (270–443)a 294 (193–422)b

Crown group Hydra 193 (149–241) 202 (134–276) 175 (119–236)viridissima group 112 (77–152) 150 (94–210) –braueri + oligactis +

vulgaris group147 (108–187) 123 (74–181) 121 (75–174)

oligactis + vulgarisgroup

101 (72–132) 84 (49–127) –

braueri group 69 (39–110) 80 (45–121) –oligactis group 71 (43–102) 63 (45–121) –vulgaris group 62 (35–98) 63 (33–97) –

a Split between crown group Hydra and Candelabrum cocksii.b Split between crown group Hydra and Candelabrum cocksii plus Ralpharia

gorgoniae.

Table 6Summary of proposed species affiliation for several common ‘lab strains’ of Hydra.

Common lab strains Species affiliation

H. carnea H. carneaH. circumcincta M7 H. circumcincta 1H. magnipapillata (all strains) H. vulgarisa

H. oligactis (all strains) H. oligactisH. viridissima (or its synonym H.

viridis)Several different species

H. viridissima A14c H. sinensisH. viridissima A250 H. viridissima 5H. viridissima A99 H. viridissima 3H. viridissima M9 H. viridissima 5H. viridissima M10 H. viridissima 5H. viridissima M120 H. viridissima 5H. viridissima NC64 H. sinensisH. vulgaris ‘AEP’ H. carneaH. vulgaris ‘Basel’ H. vulgarisa

Pelmatohydra robusta H. robusta (possibly synonymous to H.oligactis)

a Herein identified as H. vulgaris 7 but probably represent the ‘true’ H. vulgarisPallas (1766).

50 M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55

Schuchert, 2010) and were adopted only rarely in the literature.Although both genera would be monophyletic (if Pelmatohydrawas applied to all species of the oligactis group), we follow theother authors and suggest retaining the genus Hydra for all speciesof all four groups.

4.1. Translating main lineages into species: viridissima group

Although four species of ‘green’ Hydra have been described – H.viridissima Pallas, 1766 from Europe; H. hadleyi (Forrest, 1959)from North America; H. plagiodesmica Dioni, 1968 from South

M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55 51

America and H. sinensis Wang et al., 2012 from China – it hasbecome common practice to refer to all ‘green’ Hydra as H. viridis-sima (e.g. Habetha et al., 2003; Habetha and Bosch, 2005;Hemmrich et al., 2007; Jankowski et al., 2008; Kawaida et al.,2010, 2013; Martínez et al., 2010). That this may be an oversimpli-fication became apparent by high levels of genetic differentiationobserved among various strains of this group (Kawaida et al.,2010; Martínez et al., 2010). But also earlier studies based on mor-phology alone suggested the presence of additional species of‘green’ Hydra (Hyman, 1929; McAuley, 1984; Schulze, 1917). Byanalyzing all the data available we could identify six main lineageswithin the viridissima group, which translate to five or even six bio-logical species. Most of them occur sympatrically with no indica-tion of gene flow and reproduction. The sole exceptions arelineages H. viridissima 5 and 6. Their symbiotic Chlorella are genet-ically nearly identical (Kawaida et al., 2013) and individualsassigned to either H. viridissima 5 or 6 mitochondrially featuredITS sequences of the other lineage. Whether this is an artifact(i.e. an ancestral polymorphism) or an indication of ongoing repro-duction has to remain open for now.

Linking the herein delimited species to existing species namesis not unambiguously possible. With the data currently availabletwo cases appear straight forward giving the geographic distribu-tion of the species. The South American H. plagiodesmica Dioni,1968 most likely corresponds to H. viridissima 4 and H. sinensisWang et al., 2009 to H. viridissima 6. Hydra viridissima Pallas,1766 was first described from Europe without a precise type local-ity (see Schuchert, 2010). Thus, H. viridissima 2, 3 and 4 all are pos-sible candidates to represent H. viridissima Pallas, 1766. Similarly,one of the three species identified from North America – H. viridis-sima 1, 2, and 5 (the latter potentially including H. viridissima 6) –may represent H. hadleyi (Forrest, 1959). If H. viridissima 5 and 6represent a single species and if this species is the ‘true’ H. hadleyi,H. sinensis would be a junior synonym of H. hadleyi.

Given the wide distribution of ‘green’ Hydra, only few strainswere incorporated into this and all previous analyses. Most likely,more species will be discovered if sampling is performed moredensely, as exemplified by the three species recorded fromSwitzerland. This may further complicate the correct assignmentof existing species names. If one keeps in mind that the divergencewithin the extant member of the viridissima group is nearly as oldas among all other Hydra species, a much greater number of spe-cies can be anticipated for this particular group.

4.2. Translating main lineages into species: oligactis group

The main lineage pairs H. oxycnida & H. oligactis and H. canaden-sis & H. oligactis can be considered to have sympatric distributions,the former in Europe and the latter in northern North America. Dueto their consistent differentiation in mitochondrial and nuclearmarkers they qualify as biological species. The additional subdivi-sion within H. oxycnida and H. canadensis proposed by some anal-yses are viewed as a result of intraspecific genetic variation, as thissubdivision is not supported by all markers.

The species status of H. robusta is questionable. Previous studiesidentified H. robusta and H. oligactis as closely related sister species(Hemmrich et al., 2007; Kawaida et al., 2010), which were notreciprocally monophyletic in the phylogenetic analyses of each sin-gle marker. Of course, reciprocal monophyly is not a prerequisitefor species differentiation and shared identical sequences couldbe the result of ancestral polymorphisms (Funk and Omland,2003). However, the additional COI sequences of European individ-uals morphologically identified as H. oligactis obtained from theBOLD database (submitted by P. Schuchert) clustered within theH. robusta clade. This questions the distinctiveness of H. robusta.Regrettably, no ITS sequences were available for these important

individuals. If their ITS sequences correspond to the ITS sequencesfound so far in H. robusta (which are distinct from those of H. oli-gactis), this would strongly support reproductive isolation amongthese two species and would expand the distribution of H. robustafrom Japan to central Europe. Conversely, if their ITS sequencescorrespond to those of other H. oligactis from central Europe, thespecies status of H. robusta would be doubtful.

4.3. Translating main lineages into species: braueri group

All four main lineages identified in the braueri group can eachbe ascribed to a distinct biological species: H. circumcincta 1, H. cir-cumcincta 2, H. hymanae and H. utahensis. Because all of them occurin northern North America (Alaska) and H. circumcincta 1 and H.circumcincta 2 also in northern Europe, their distributions can beconsidered as sympatric. All four are consistently differentiatedin the studied mitochondrial and nuclear markers, which stronglyimply reproductive isolation among these species. A further subdi-vision of H. circumcincta 1, as indicate by a few analyses, is doubtfuland may rather be caused by intraspecific genetic diversity. Whichof the two H. circumcincta species is the true H. circumcinctaSchulze, 1914 cannot be determined unambiguously here. Theavailable data points to H. circumcincta 1. It was recorded also incentral Europe including northern Germany and the type localityfor H. circumcincta is in Berlin (Germany). The respective other spe-cies may correspond to one of the species previously synonymizedwith H. circumcincta (see Schuchert, 2010).

4.4. Translating main lineages into species: vulgaris group

The vulgaris group seems to be the most diverse group of Hydra.Across all markers 14 main lineages were delimited; however, ofonly seven main lineages all three markers were available. Thisand the fact that the majority of main lineages appear to beallopatrically distributed impedes their delimitation as biologicalspecies and impedes reliable estimated of biological species num-bers, even though for most main lineages no evidence pointsagainst them representing separate biological species.Nevertheless, several important conclusions can be drawn.

First of all, H. magnipapillata and several H. vulgaris strains fromEurasia and South Africa belong to a single species (H. vulgaris 7).This includes H. vulgaris strain Basel, which is an important labstrain and which has been used in comparative studies alongsideH. magnipapillata, and strains identified as H. japonica, H. attenuataand H. shenzhensis. The synonymy of H. magnipapillata and some H.vulgaris is not surprising as previous studies had revealed only lit-tle genetic differentiation among strains (Kawaida et al., 2010;Martínez et al., 2010); albeit Martínez et al. (2010) summarizedall member of the vulgaris group as H. vulgaris. Hydra shenzhensiswas only recently described (Wang et al., 2012) which may beattributed to the erroneous assumption that H. japonica and H.attenuata referred to valid species. H. vulgaris 7 is most likely thetrue H. vulgaris Pallas, 1766. Following Pallas (1766) H. vulgaris issupposed to be common throughout Europe and the otherEuropean H. vulgaris main lineage identified herein (H. vulgaris 8)seems to be rarer and possibly restricted to Switzerland.

Secondly, the majority of North American H. vulgaris strainsconstitute a single species (H. vulgaris 4) including strains fromJapan and South America and those identified as H. carnea and H.littoralis. It also includes H. vulgaris AEP, the strain used for gener-ating transgenic Hydra (e.g. Wittlieb et al., 2006), as it is a crossbetween strains identified as H. carnea and H. littoralis (Martínezet al., 2010). The valid name for this species is most likely H. carneaAgassiz, 1851. Previous studies suggested a close relationshipbetween H. carnea and H. vulgaris AEP (Hemmrich et al., 2007;Kawaida et al., 2010), but not necessarily them being conspecific.

Fig. 4. Scheme of inferred evolutionary history of Hydra groups. It highlightsimportant times of divergence among and within the various Hydra groups.Divergence within each Hydra group is limited to the first divergence, for viridissimagroup the first two divergence events are shown due the group’s greater age. Thebranch leading to Hydra is depicted in red and the one leading to viridissima groupin green, as these indicate the time intervals in which the Hydra and viridissimagroup specific traits (e.g. taxonomically restricted genes/orphan genes) evolved,respectively. The scheme is based on the fossil dated molecular clock analysisdepicted in Fig. 3. (For interpretation of the references to colour in this figurelegend, the reader is referred to the web version of this article.)

52 M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55

Whether H. carnea is the only representative of the vulgaris groupin North America or whether the two other main lineages (H. vul-garis 5 and 6) present in North America constitute separate biolog-ical species has to remain an open question for now.

All other main lineages are genetically well separated in allavailable markers and it seems likely that they represent distinctspecies each. Apart from the erroneous application of the speciesname H. vulgaris there is no further apparent conflict with taxo-nomic nomenclature. Given their high levels of intralineage geneticdiversity H. vulgaris 2 and H. vulgaris 10 may each represent twodistinct species. However, given the low number of sequencesavailable from each of these lineages we hesitated to split themfurther. Hydra polymorphus Chen and Wang (2008) and H. zhujian-gensis Liu and Wang (2010) are both genetically differentiated fromother Asian species and thus well supported as valid species. Hydravulgaris 1 and 2 both occur sympatrically in South America (Chileand Argentina) and are genetically well differentiated in mitochon-drial and nuclear markers and are thus most likely biological spe-cies. One of these could be H. vulgaris pedunculata Deserti et al.(2011), which was described from eastern Argentina. If this werethe case, H. vulgaris pedunculata would need to be raised fromsub-species to species level. Similarly, other main lineages identi-fied herein may correspond to previously described species of thevulgaris group; however, elucidating this issue requires detailedmorphological and taxonomic studies.

4.5. Evolutionary origins and biogeographic implications

Current hypotheses regarding the origin and biogeographic his-tory of Hydra are widely diverging. While the age of the viridissimagroup was estimated to be 156–174 million years (Kawaida et al.,2013), the age of crown group Hydra was estimated to be only�60 million years (Martínez et al., 2010). These are highly contra-dictory estimates. Accordingly, in the latter study Hydra wasassumed to have originated on the prehistoric northern hemi-sphere continent Laurasia. Members of the vulgaris and viridissimagroup would have colonized the southern continents by intercon-tinental dispersal, whereas the oligactis and braueri group did notdisperse to southern continents. Conversely, based purely on thedistribution of the four groups is has been assumed that the

viridissima and vulgaris groups evolved prior to the separation ofcontinents (possibly on Pangea) and the braueri and oligactisgroups evolved after the separation of northern and southern con-tinents (Jankowski et al., 2008). This scenario does not require anyintercontinental dispersal.

The divergence time estimates obtained in the present studydate the evolutionary origins of crown group Hydra to the lateTriassic or early Jurassic (Fig. 4). This finding is somehow inter-mediate to the divergence time estimates of previous studies(Kawaida et al., 2013; Martínez et al., 2010). In the lateTriassic/early Jurassic all continental plates were still part ofPangaea, though Pangea started breaking apart into northernLaurasia and southern Gondwana (Bortolotti and Principi,2005). Given its Pangean origin, today’s global distribution ofHydra could be a result of vicariance with little or nolarge-scale (i.e. without trans-oceanic) dispersal as suggestedby Jankowski et al. (2008). But the divergence between theoligactis, braueri and vulgaris groups and the ages of each ofthese groups are younger than the separation betweenGondwana and Laurasia. Even Gondwana most likely brokeapart before the four extant groups of Hydra evolved(Upchurch, 2008). Consequently, the cosmopolitan distributionsof the vulgaris group and most likely also of the viridissimagroup are due to transcontinental and transoceanic dispersalas suggested by Martínez et al. (2010), although the underlyingtimeframe differs greatly. This is in direct contrast to theprevailing assumption that Hydra is not capable of transoceanicdispersal (Campbell, 1999; Jankowski et al., 2008).

Of course, Hydra species may have spread across all conti-nents through the global trade of aquatic organisms. This mayexplain why genetically identical individuals (or strains) wererecovered from different continents (Campbell et al., 2013;Martínez et al., 2010), for example identical genotypes ofH. vulgaris 4 in America and New Zealand, of H. viridissima 3in Australia and Germany and of H. viridissima 5 in Europe,North America and Africa (a potential source of error could bethe long-term maintenance of some strains in different laborato-ries). However, for several species with wide geographic distri-butions – like H. vulgaris 7 across Eurasia and South Africa orH. oligactis and H. circumcincta 1 & 2 across northern Europeand North America – the recovered genotypes were closelyrelated but not identical as would be expected for recenthuman-mediated dispersal. In addition, the presence of endemicand genetically divergent species of the vulgaris group in SouthAfrica, Australia/New Zealand and South America corroboratestransoceanic dispersal in pre-human times. These are strongindications that Hydra species are well capable of long-distance and even trans-oceanic dispersal. Also members ofthe oligactis and braueri group seem to be present on southerncontinents as well (e.g. Campbell, 1999; Deserti et al., 2012),questioning cross-equatorial dispersal limitations, which havebeen suggested for these groups (Martínez et al., 2010), giventhe time-frame of diversification proposed herein.

It is noteworthy that the diversification of three Hydra groups –oligactis, braueri and vulgaris groups – as well as the diversificationwithin the two largest monophyletic clades within the viridissimagroup occurred between �60 and �70 mya. This coincides roughlywith the Cretaceous-Paleogene boundary, a time of severe climaticchange and of global mass extinction. Possibly former Hydra diver-sity was reduced to few species, which subsequently gave rise totoday’s groups. Because the majority of extant species occur inNorth America and/or Eurasia, the ancestors of today’s groups mostlikely survived on Laurasia and spread further south from there.Transoceanic dispersal after the separation of all main land masseswould explain why Hydra species from Africa, South America andAustralia do not show the close relationship expected for taxa with

M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55 53

Gondwana distributions, however, this may also be due to theoverall low resolution of interspecific relationships.

Our results provide a robust framework to assess the timing andrate of key evolutionary processes. Of course, all molecular clockestimates should be treated with some caution, but our analysisallows setting certain limits for the timing of evolutionary pro-cesses. For example, Hydra specific genes that are present in allextant Hydra but missing in all other taxa must have evolved afterHydra split from its proposed sister taxon and before crown groupHydra diversified; thus in a �130 million year window between�325 and �195 mya. Another important question regards theacquisition of symbiotic Chlorella by all members of the viridissimagroup. We can assume that the ancestor of all extant Hydra lackedChlorella because this is the plesiomorphic condition observed inall other Aplanulata and all non-green Hydra species. Therefore,the acquisition of Chlorella symbionts and related adaptationsand evolutionary novelties (like taxon specific genes) would haveoccurred between �110 and �195 mya over a period of �80 mil-lion years. Of course, comparable estimates are possible amongall groups of Hydra, offering new possibilities to assess the ratesof genomic changes.

5. Conclusion

One may think that the taxonomy of Hydra and the preciseassignment of single individuals (or clonal strains) to certainspecies is a mere academic exercise, which is of little importancefor others than taxonomists or possibly ecologist. However,Hydra is an important model system in both evolutionary devel-opmental biology and research on the metaorganism or holo-biont. Such studies rely on a precise and unambiguousassignment of the studied individuals to species as the prerequi-site for intraspecific or interspecific comparisons and for theidentification of common or diverging processes and functionsamong species.

With respect to Hydra research three findings are crucial: (1)species level diversity was underestimated by previous molecu-lar studies; (2) species affiliations of several crucial ‘workhorses’of Hydra evolutionary research were wrong and (3) the evolu-tionary origins of the genus Hydra date back nearly 200 millionyears to the late Triassic or early Jurassic. Our re-analysis ofexisting data highlights the need for a careful and extensiverevision of the taxonomy of Hydra species. For evolutionaryresearch the fact that H. vulgaris strain Basel and H. magnipapil-lata as well as H. carnea and H. vulgaris AEP are conspecific andthat H. viridissima features several species have far reachingconsequences as inter- and infraspecific differences may havebeen wrongly assessed. Informative examples are the associatedbacterial communities of H. carnea and H. vulgaris AEP, whichshow great similarity (Franzenburg et al., 2013). In light of thepresent study, this is not surprising given the fact that bothbelong to the same species. Previous studies have indicatedtheir close relationships, but not necessarily them being conspe-cific. Our divergence time estimates set the time frame to datethe origins of evolutionary novelties and to assess the rates ofgenomic changes for Hydra and its groups and species.

Acknowledgments

The work was supported in part by grants from the DeutscheForschungsgemeinschaft (DFG) and by grants from the DFGClusters of Excellence programs ‘‘The Future Ocean’’ and‘‘Inflammation at Interfaces’’. We thank four anonymous reviewersfor their helpful comments on an earlier version of the manuscript.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.ympev.2015.05.013.

References

Agapow, P.-M., Bininda-Emonds, O.R.P., Crandall, K.A., Gittleman, J.L., Mace, G.M.,Marshall, J.C., Purvis, A., 2004. The impact of species concepts on biodiversitystudies. Quat. Rev. Biol. 79, 161–179.

Agassiz, L., 1851. On the little bodies seen on Hydra. Proc. Bost. Soc. Nat. Hist. 3,354–355.

Bortolotti, V., Principi, G., 2005. Tethyan ophiolites and Pangea break-up. Isl. Arc 14,442–470.

Bosch, T.C.G., 2013. Cnidarian-microbe interactions and the origin of innateimmunity in metazoans. Annu. Rev. Microbiol. 67, 499–518.

Bosch, T.C.G., 2014. Rethinking the role of immunity: lessons from Hydra. TrendsImmunol. 35, 495–502.

Bouckaert, R., Heled, J., Kühnert, D., Vaughan, T., Wu, C.H., Xie, D., Suchard, M.A.,Rambaut, A., Drummond, A.J., 2014. BEAST 2: A software platform for Bayesianevolutionary analysis. PLoS Comput. Biol. 10, e1003537. 3537.

Brucker, R.M., Bordenstein, S.R., 2012. Speciation by symbiosis. Trends Ecol. Evol.27, 443–451.

Brucker, R.M., Bordenstein, S.R., 2013. The hologenomic basis of speciation: gutbacteria cause hybrid lethality in the genus Nasonia. Science 341, 667–669.

Campbell, R.D., 1983. Identifying Hydra species. In: Lenhoff, H.M. (Ed.), HydraResearch Methods. Plenum Publishing, New York.

Campbell, R., 1987. A new species of Hydra (Cnidaria Hydrozoa) from North Americawith comments on species clusters within the genus. Zool. J. Linn. Soc. 91, 253–263.

Campbell, R., 1989. Taxonomy of the European Hydra (Cnidaria: Hydrozoa): are-examination of its history with emphasis on the species H. vulgaris Pallas,H. attenuata Pallas and H. circumcincta Schulze. Zool. J. Linn. Soc. 95,219–244.

Campbell, R., 1999. The Hydra of Madagascar (Cnidaria: Hydrozoa). Ann. Limnol. 35,95–104.

Campbell, R.D., Iñiguez, A.R., Iñiguez, A.J., Martínez, D.E., 2013. Hydra of Hawaii:phylogenetic relationships with continental species. Hydrobiologia 713, 199–205.

Cartwright, P., Collins, A., 2007. Fossils and phylogenies: integrating multiple linesof evidence to investigate the origin of early major metazoan lineages. Integr.Comp. Biol. 47, 744–751.

Chapman, J.A., Kirkness, E.F., Simakov, O., Hampson, S.E., Mitros, T., Weinmaier, T.,Rattei, T., Balasubramanian, P.G., Borman, J., Busam, D., Disbennett, K.,Pfannkoch, C., Sumin, N., Sutton, G.G., Viswanathan, L.D., Walenz, B.,Goodstein, D.M., Hellsten, U., Kawashima, T., Prochnik, S.E., Putnam, N.H., Shu,S., Blumberg, B., Dana, C.E., Gee, L., Kibler, D.F., Law, L., Lindgens, D., Martinez,D.E., Peng, J., Wigge, P.A., Bertulat, B., Guder, C., Nakamura, Y., Ozbek, S.,Watanabe, H., Khalturin, K., Hemmrich, G., Franke, A., Augustin, R., Fraune, S.,Hayakawa, E., Hayakawa, S., Hirose, M., Hwang, J.S., Ikeo, K., Nishimiya-Fujisawa, C., Ogura, A., Takahashi, T., Steinmetz, P.R.H., Zhang, X., Aufschnaiter,R., Eder, M.-K., Gorny, A.-K., Salvenmoser, W., Heimberg, A.M., Wheeler, B.M.,Peterson, K.J., Böttger, A., Tischler, P., Wolf, A., Gojobori, T., Remington, K.A.,Strausberg, R.L., Venter, J.C., Technau, U., Hobmayer, B., Bosch, T.C.G., Holstein,T.W., Fujisawa, T., Bode, H.R., David, C.N., Rokhsar, D.S., Steele, R.E., 2010. Thedynamic genome of Hydra. Nature 464, 592–596.

Chen, Z., Wang, A., 2008. A new species of the genus Hydra from China (Hydrozoa,Hydraridae). Acta Zootaxonomica Sin. 33, 737–741.

Chen, J.-Y., Oliveri, P., Gao, F., Dornbos, S.Q., Li, C.-W., Bottjer, D.J., Davidson, E.H.,2002. Precambrian animal life: probable developmental and adult cnidarianforms from southwest China. Dev. Biol. 248, 182–196.

David, C.N., 2012. Interstitial stem cells in Hydra: multipotency and decision-making. Int. J. Dev. Biol. 56, 489–497.

Dawson, M.N., 2005a. Five new subspecies of Mastigias (Scyphozoa: Rhizostomeae:Mastigiidae) from marine lakes, Palau. Micronesia. J. Mar. Biol. Assoc. UK 85,679–694.

Dawson, M.N., 2005b. Incipient speciation of Catostylus mosaicus (Scyphozoa,Rhizostomeae, Catostylidae), comparative phylogeography and biogeography insouth-east Australia. J. Biogeogr. 32, 515–533.

Deserti, M., Zamponi, M., Escalante, A., 2011. The genus Hydra from Argentina. I.Hydra vulgaris pedunculata subsp. nov. (Cnidaria, Hydrozoa). Rev. Real Acad.Galega Ciencias 30, 5–14.

Deserti, M.I., Zamponi, M.O., Escalante, A.H., 2012. The genus Hydra from Argentina.II. Hydra pseudoligactis Hyman, 1931 (Cnidaria; Hydrozoa), a new record. Rev.Real Acad. Galega Ciencias 31, 5–14.

Dioni, W., 1968. Hydra (Chlorohydra) plagiodesmica sp. nov. una hidra verde del RioSalado, Republica Argentina (Cnidaria, Hydrozoa). Phys., B. Aires 28, 203–210.

Ewer, R., 1948. A review of the Hydridae and two new species of Hydra from Natal.Proc. Zool. Soc. Lond. 118, 226–244.

Forrest, H., 1959. Taxonomic studies on the Hydras of North America. VII.Description of Chlorohydra hadleyi, new species, with a key to the NorthAmerican species of Hydras. Am. Midl. NAt. 62, 441.

54 M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55

Franzenburg, S., Walter, J., Künzel, S., Wang, J., Baines, J.F., Bosch, T.C.G., 2013.Distinct antimicrobial peptide expression determines host species-specificbacterial associations. PNAS E 3730–E 3738.

Fraune, S., Bosch, T.C.G., 2007. Long-term maintenance of species-specific bacterialmicrobiota in the basal metazoan Hydra. PNAS 104, 13146–13151.

Fujisawa, T., Hayakawa, E., 2012. Peptide signaling in Hydra. Int. J. Dev. Biol. 56,543–550.

Funk, D.J., Omland, K.E., 2003. Species-level paraphyly and polyphyly: frequency,causes, and consequences, with insights from animal mitochondrial DNA. Annu.Rev. Ecol. Evol. Syst. 34, 397–423.

Galliot, B., 2012. Hydra, a fruitful model system for 270 years. Int. J. Dev. Biol. 56,411–423.

Grimmelikhuijzen, C.J., Hauser, F., 2012. Mini-review: the evolution of neuropeptidesignaling. Regul. Pept. 177, 6–9.

Habetha, M., Bosch, T.C.G., 2005. Symbiotic Hydra express a plant-like peroxidasegene during oogenesis. J. Exp. Biol. 208, 2157–2164.

Habetha, M., Anton-Erxleben, F., Neumann, K., Bosch, T.C.G., 2003. The Hydra viridis/Chlorella symbiosis. Growth and sexual differentiation in polyps withoutsymbionts. Zoology 106, 101–108.

Hemmrich, G., Anokhin, B., Zacharias, H., Bosch, T.C.G., 2007. Molecularphylogenetics in Hydra, a classical model in evolutionary developmentalbiology. Mol. Phylogenet. Evol. 44, 281–290.

Holland, B.S., Dawson, M.N., Crow, G.L., Hofmann, D.K., 2004. Globalphylogeography of Cassiopea (Scyphozoa: Rhizostomeae): molecular evidencefor cryptic species and multiple invasions of the Hawaiian Islands. Mar. Biol.145, 1119–1128.

Holstein, T.W., 2012. A view to kill. BMC Biol. 10, 18.Hyman, L.H., 1929. Taxonomic studies on the Hydras of North America. I. General

remarks and descriptions of Hydra americana, new species. Trans. Am.Microcop. Soc. 48, 242–255.

Jankowski, T., Collins, A.G., Campbell, R., 2008. Global diversity of inland watercnidarians. Hydrobiologia 595, 35–40.

Katoh, K., Standley, D.M., 2013. MAFFT multiple sequence alignment softwareversion 7: improvements in performance and usability. Mol. Biol. Evol. 30, 772–780.

Kawaida, H., Shimizu, H., Fujisawa, T., Tachida, H., Kobayakawa, Y., 2010. Molecularphylogenetic study in genus Hydra. Gene 468, 30–40.

Kawaida, H., Ohba, K., Koutake, Y., Shimizu, H., Tachida, H., Kobayakawa, Y., 2013.Symbiosis between Hydra and Chlorella: molecular phylogenetic analysis andexperimental study provide insight into its origin and evolution. Mol.Phylogenet. Evol. 66, 906–914.

Khalturin, K., Anton-Erxleben, F., Sassmann, S., Wittlieb, J., Hemmrich, G., Bosch,T.C.G., 2008. A novel gene family controls species-specific morphological traitsin Hydra. PLoS Biol. 6, e278.

Khalturin, K., Hemmrich, G., Fraune, S., Augustin, R., Bosch, T.C.G., 2009. More thanjust orphans: are taxonomically-restricted genes important in evolution?Trends Genet. 25, 404–413.

Koizumi, O., 2007. Nerve ring of the hypostome in Hydra: is it an origin ofthe central nervous system of bilaterian animals. Brain Behav. Evol. 69,151–159.

Laakmann, S., Holst, S., 2014. Emphasizing the diversity of North Sea hydromedusaeby combined morphological and molecular methods. J. Plankton Res. 36, 64–76.

Lasi, M., David, C.N., Böttger, A., 2010. Apoptosis in pre-Bilaterians: Hydra as amodel. Apoptosis 15, 269–278.

Liu, H., Wang, A., 2010. A new species of the genus Hydra from Guangdong,China (Hydrozoa, Anthoathecatae, Hydraridae). Acta Zootaxonomica Sin. 35,857–862.

Love, G.D., Grosjean, E., Stalvies, C., Fike, D.A., Grotzinger, J.P., Bradley, A.S., Kelly,A.E., Bhatia, M., Meredith, W., Snape, C.E., Bowring, S.A., Condon, D.J., Summons,R.E., 2009. Fossil steroids record the appearance of Demospongiae during theCryogenian period. Nature 457, 718–721.

Martin, V., Littlefield, C., Archer, W., Bode, H., 1997. Embryogenesis in Hydra. Biol.Bull. 192, 345–363.

Martínez, D.E., Iñiguez, A.R., Percell, K.M., Willner, J.B., Signorovitch, J., Campbell,R.D., 2010. Phylogeny and biogeography of Hydra (Cnidaria: Hydridae)using mitochondrial and nuclear DNA sequences. Mol. Phylogenet. Evol. 57,403–410.

Mayr, E., 1942. Systematics and The Origin of Species from the Viewpoint of aZoologist. Columnia University Press, New York.

McAuley, P., 1984. Variation in green Hydra. A description of three cloned strains ofHydra viridissima Pallas 1766 (Cnidaria: Hydrozoa) isolated from a single site.Biol. J. Linn. Soc. 1766, 1–13.

McFall-Ngai, M., Hadfield, M.G., Bosch, T.C., Carey, H.V., Domazet-Loso, T., Douglas,A.E., Dubilier, N., Eberl, G., Fukami, T., Gilbert, S.F., et al., 2013. Animals in abacterial world, a new imperative for the life sciences. Proc. Natl. Acad. Sci. USA110, 3229–3236.

Meinhardt, H., 2012. Modeling pattern formation in hydra: a route to understandingessential steps in development. Int. J. Dev. Biol. 56, 447–462.

Mishler, B.D., Theriot, E.C., 2000. The phylogenetic species concept (sensu Mishler &Theriot): monophyly, apomorphy and phylogenetic species concepts. In:Wheeler, Q., Meier, R. (Eds.), Species and Phylogenetic Theory. ColumbiaUniversity Press, New York, pp. 44–54.

Moura, C.J., Harris, D.J., Cunha, M.R., Rogers, A.D., 2008. DNA barcoding revealscryptic diversity in marine hydroids (Cnidaria, Hydrozoa) from coastal anddeep-sea environments. Zool. Scr. 37, 93–108.

Nawrocki, A.M., Collins, A.G., Hirano, Y.M., Schuchert, P., Cartwright, P., 2013.Phylogenetic placement of Hydra and relationships within Aplanulata(Cnidaria: Hydrozoa). Mol. Phylogenet. Evol. 67, 60–71.

Nebel, A., Bosch, T.C.G., 2012. Evolution of human longevity: lessons from Hydra.Aging 4, 730–731.

Ortman, B.D., Bucklin, A., Pagès, F., Youngbluth, M., 2010. DNA Barcoding theMedusozoa using mtCOI. Deep Sea Res. Part II Top. Stud. Oceanogr. 57, 2148–2156.

Pallas, P.A., 1766. Elenchus zoophytorum sistens generum adumbrationesgeneraliores et specierum cognitarium succinctas descriptiones cum selectisauctorum synonymis. Fransiscum Varrentrapp, Hagae.

Park, E., Hwang, D.-S., Lee, J.-S., Song, J.-I., Seo, T.-K., Won, Y.-J., 2012. Estimation ofdivergence times in cnidarian evolution based on mitochondrial protein-codinggenes and the fossil record. Mol. Phylogenet. Evol. 62, 329–345.

Pepper, M., Doughty, P., Hutchinson, M.N., Scott Keogh, J., 2011. Ancient drainagesdivide cryptic species in Australia’s arid zone: morphological and multi-geneevidence for four new species of Beaked Geckos (Rhynchoedura).Mol. Phylogenet. Evol. 61, 810–822. http://dx.doi.org/10.1016/j.ympev.2011.08.012.

Peterson, K.J., Lyons, J.B., Nowak, K.S., Takacs, C.M., Wargo, M.J., McPeek, M., 2004.Estimating metazoan divergence times with a molecular clock. Proc. Natl. Acad.Sci. USA 101, 6536–6541.

Pons, J., Barraclough, T., Gomez-Zurita, J., Cardoso, A., Duran, D., Hazell, S., Kamoun,S., Sumlin, W., Vogler, A., 2006. Sequence-based species delimitation for theDNA taxonomy of undescribed insects. Syst. Biol. 55, 595–609.

Pontin, D.R., Cruickshank, R.H., 2012. Molecular phylogenetics of the genus Physalia(Cnidaria: Siphonophora) in New Zealand coastal waters reveals crypticdiversity. Hydrobiologia 686, 91–105.

Puillandre, N., Lambert, A., Brouillet, S., Achaz, G., 2012. ABGD, Automatic BarcodeGap Discovery for primary species delimitation. Mol. Ecol. 21, 1864–1877.

Rambaut, A., 2014. FigTree 1.4.2, Available from <http://tree.bio.ed.ac.uk/>.Rambaut, A., Suchard, M.A., Xie, D., Drummond, A.J., 2013. Tracer v1.6, Available

from <http://beast.bio.ed.ac.uk/Tracer>.Reddy, P.C., Barve, A., Ghaskadbi, S., 2011. Description and phylogenetic

characterization of common Hydra from India. Curr. Sci. 101, 736–738.Ronquist, F., Huelsenbeck, J.P., 2003. MrBayes 3: Bayesian phylogenetic inference

under mixed models. Bioinformatics 19, 1572–1574.Schroth, W., Jarms, G., Streit, B., Schierwater, B., 2002. Speciation and

phylogeography in the cosmopolitan marine moon jelly. Aurelia sp. BMC Evol.Biol. 10, 2.

Schuchert, P., 2010. The European athecate hydroids and their medusae (Hydrozoa,Cnidaria): Capitata Part 2. Rev. Suisse Zool. 117, 337–555.

Schuchert, P., 2014. World Hydrozoa database. Accessed through: World Register ofMarine Species at <http://www.marinespecies.org/aphia.php?p=taxdetails&id=290703> (06.11.14).

Schulze, P., 1914. Bestimmungstabelle der deutschen Hydraarten. Sitzungsberichteder Gesellschaft naturforschender Freunde zu Berlin 9, 395–398.

Schulze, P., 1917. Neue Beiträge zu einer Monographie der Gattung Hydra. Arch.Biontol. 4, 39–119.

Schwentner, M., Timms, B.V., Richter, S., 2011. An integrative approach to speciesdelineation incorporating different species concepts: a case study ofLimnadopsis (Branchiopoda: Spinicaudata). Biol. J. Linn. Soc. 104, 575–599.

Schwentner, M., Timms, B.V., Richter, S., 2012. Description of four new species ofLimnadopsis from Australia (Crustacea: Branchiopoda: Spinicaudata). Zootaxa3315, 42–64.

Schwentner, M., Just, F., Richter, S., 2015. Evolutionary systematics of the AustralianCyzicidae (Crustacea, Branchiopoda, Spinicaudata) with the description of anew genus. Zool. J. Linn. Soc. 173, 271–295.

Shimizu, H., 2012. Transplantation analysis of developmental mechanisms in Hydra.Int. J. Dev. Biol. 56, 463–472.

Steele, R.E., David, C.N., Technau, U., 2011. A genomic view of 500 million years ofcnidarian evolution. Trends Genet. 27, 7–13.

Talavera, G., Castresana, J., 2007. Improvement of phylogenies after removingdivergent and ambiguously aligned blocks from protein sequence alignments.Syst. Biol. 56, 564–577.

Tamura, K., Stecher, G., Peterson, D., Filipski, A., Kumar, S., 2013. MEGA6: molecularevolutionary genetics analysis version 6.0. Mol. Biol. Evol. 30, 2725–2729.

Tan, D., Ali, F., Kutty, S., Meier, R., 2008. The need for specifying species concepts:how many species of silvered langurs (Trachypithecus cristatus group) should berecognized? Mol. Phylogenet. Evol. 49, 688–689.

Tanaka, E.M., Reddien, P.W., 2011. The cellular basis for animal regeneration. Dev.Cell. 21, 172–185.

Technau, U., Steele, R.E., 2011. Evolutionary crossroads in developmental biology:Cnidaria. Development 138, 1447–1458.

Thomsen, S., Bosch, T.C.G., 2006. Foot differentiation and genomic plasticity inHydra: lessons from the PPOD gene family. Dev. Genes Evol. 216, 57–68.

Trembley, A., 1744. Mémoires, pour servir à l’historie d’un genre de polypes d’eaudouce, à bras en forme de cornes (Jean and Herman Verbeek).

Upchurch, P., 2008. Gondwanan break-up: legacies of a lost world? Trends Ecol.Evol. 23, 229–236. http://dx.doi.org/10.1016/j.tree.2007.11.006.

van Leeuwenhoek, A., 1702. In: Palm, L.C. (Ed.). The Collected Letters of Antoni vanLeeuwenhoek, vol. XIV, pp. 169–173 (Swets and Zeitlinger, 1996).

Waggoner, B., Collins, A.G., 2004. Reducto ad absurdum: testing the evolutionaryrelationships of ediacaran and paleaozoic problematic fossils using moleculardivergence dates. J. Paleont. 78, 51–61.

M. Schwentner, T.C.G. Bosch / Molecular Phylogenetics and Evolution 91 (2015) 41–55 55

Wang, A.-T., Deng, L., Lai, J.-Q., Li, J., 2009. A new species of green Hydra (Hydrozoa:Hydrida) from China. Zool. Sci. 26, 664–668.

Wang, A.-T., Deng, L., Liu, H.-T., 2012. A new species of Hydra (Cnidaria: Hydrozoa:Hydridae) and molecular phylogenetic analysis of six congeners from China.Zool. Sci. 29, 856–862.

Watanabe, H., Hoang, V.T., Mättner, R., Holstein, T.W., 2009. Immortality and thebase of multicellular life: lessons from cnidarian stem cells. Semin Cell Dev.Biol. 20, 1114–1125.

Wittlieb, J., Khalturin, K., Lohmann, J.U., Anton-Erxleben, F., Bosch, T.C.G., 2006.Transgenic Hydra allow in vivo tracking of individual stem cells duringmorphogenesis. Proc. Natl. Acad. Sci. USA 103, 6208–6211.

Xiao, S., Yuan, X., Knoll, A.H., 2000. Eumetazoan fossils in terminal Proterozoicphosphorites? PNAS 97, 13684–13689.

Young, G.A., Hagadorn, J.W., 2010. The fossil record of cnidarian medusae.Palaeoworld 19, 212–221.