Eur. J. Phycol. (2011) 46(3): 202–228 Ribosomal DNA phylogenies and a morphological revision provide the basis for a revised taxonomy of the Prymnesiales (Haptophyta) BENTE EDVARDSEN 1 , WENCHE EIKREM 1,2 , JAHN THRONDSEN 1 , ALBERTO G. SA ´ EZ 3 , IAN PROBERT 4 AND LINDA K. MEDLIN 5 1 University of Oslo, Department of Biology, Marine Biology, NO-0316 Oslo, Norway 2 Norwegian Institute for Water Research, Gaustadalle´en 21, NO-0349 Oslo, Norway 3 Departamento de Biodiversidad y Biologı´a Evolutiva, Museo Nacional de Ciencias Naturales, CSIC, 28006 Madrid, Spain 4 CNRS/UPMC, FR2424, Station Biologique de Roscoff, BP74 29682 Roscoff cedex, France 5 University of Pierre and Marie Curie/CNRS, Laboratoire Arago, 66651 Banyuls-sur-mer, France (Received 4 January 2010; revised 30 May 2011; accepted 1 June 2011) Nucleotide sequences of the nuclear-encoded small subunit (18S rDNA) and partial large subunit (28S rDNA) ribosomal DNA were determined in 30 different species of the haptophyte genera Prymnesium, Chrysocampanula, Chrysochromulina, Imantonia and Platychrysis, all belonging to the order Prymnesiales. Phylogenies based on these and other available hapto- phyte 18S, 28S and plastid 16S rDNA sequences were reconstructed, and compared with available morphological and ultrastructural data. The rDNA phylogenies indicate that the genus Chrysochromulina is paraphyletic and is divided into two major clades. This is supported by ultrastructural and morphological data. There is a major split between Chrysochromulina species with a saddle-shaped cell form (clade B2) and the remaining species in the genus (clade B1). Clade B2 includes the type species C. parva and taxa belonging to this clade thus retain the name Chrysochromulina. The non-saddle-shaped Chrysochromulina species analysed are closely related to Hyalolithus, Prymnesium and Platychrysis species. Imantonia species are sister taxa to these species within clade B1. An amendment to the classification of the order Prymnesiales and the genera Prymnesium, Platychrysis and Chrysochromulina is proposed with one new and one emended family (Chrysochromulinaceae and Prymnesiaceae, respectively), two new genera (Haptolina and Pseudohaptolina), and one new species (Pseudohaptolina arctica). We suggest a revision of the taxonomy of the Prymnesiales that is in accordance with available molecular evidence and supported by morphological data. Key words: Chrysochromulinaceae, Haptophyta, phylogeny, phytoplankton, Prymnesiales, Pseudohaptolina arctica, ribosomal DNA, taxonomy Introduction The haptophyte order Prymnesiales Papenfuss sensu Edvardsen & Eikrem currently contains one family, the Prymnesiaceae Conrad ex O.C. Schmidt, which at present comprises the genera Chrysocampanula R.O. Fournier, Chrysochromulina Lackey, Corymbellus J.C. Green, Imantonia N. Reynolds, Hyalolithus M. Yoshida, T. Nakayama & I. Inouye, Platychrysis Geitler ex Gayral & Fresnel and Prymnesium Massart (Jordan et al., 2004). Most members of this order are unicellular, planktonic and scaly biflagellates, although colonial flagellates (e.g. Corymbellus species) occur, and some have benthic filaments or amoeboid stages in their life cycle (see reviews by Hibberd, 1980; Green et al., 1990). Most inhabit marine or brackish waters, but a few Chrysochromulina and Prymnesium spe- cies occur in fresh water (Preisig, 2003). Some spe- cies (e.g. of Chrysochromulina and Prymnesium) may form blooms, occasionally harmful to fish and other biota (reviewed by Moestrup, 1994; Edvardsen & Paasche, 1998; Edvardsen & Imai, 2006). All known species are photosynthetic, but mixotrophy appears to be common in some genera (e.g. Chrysochromulina and Prymnesium: Kawachi et al., 1991; Nygaard & Tobiesen, 1993; Jones et al., 1994; Tillmann, 1998), and uptake of dis- solved organic carbon may also occur (Pintner & Provasoli, 1968). The genus Chrysochromulina was erected by Lackey (1939) with the description of C. parva, one of only four freshwater species in the genus Correspondence to: Bente Edvardsen. E-mail: bente.edvardsen@ bio.uio.no ISSN 0967-0262 print/ISSN 1469-4433 online/11/03000202–228 ß 2011 British Phycological Society DOI: 10.1080/09670262.2011.594095 Downloaded by [Bente Edvardsen] at 09:52 27 July 2011
27
Embed
Ribosomal DNA phylogenies and a morphological revision provide
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Eur. J. Phycol. (2011) 46(3): 202–228
Ribosomal DNA phylogenies and a morphological revision
provide the basis for a revised taxonomy of the Prymnesiales
(Haptophyta)
BENTE EDVARDSEN1, WENCHE EIKREM1,2, JAHN THRONDSEN1, ALBERTO G. SAEZ3,
IAN PROBERT4 AND LINDA K. MEDLIN5
1University of Oslo, Department of Biology, Marine Biology, NO-0316 Oslo, Norway2Norwegian Institute for Water Research, Gaustadalleen 21, NO-0349 Oslo, Norway3Departamento de Biodiversidad y Biologıa Evolutiva, Museo Nacional de Ciencias Naturales, CSIC, 28006 Madrid, Spain4CNRS/UPMC, FR2424, Station Biologique de Roscoff, BP74 29682 Roscoff cedex, France5University of Pierre and Marie Curie/CNRS, Laboratoire Arago, 66651 Banyuls-sur-mer, France
(Received 4 January 2010; revised 30 May 2011; accepted 1 June 2011)
Nucleotide sequences of the nuclear-encoded small subunit (18S rDNA) and partial large subunit (28S rDNA) ribosomal
DNA were determined in 30 different species of the haptophyte genera Prymnesium, Chrysocampanula, Chrysochromulina,
Imantonia and Platychrysis, all belonging to the order Prymnesiales. Phylogenies based on these and other available hapto-
phyte 18S, 28S and plastid 16S rDNA sequences were reconstructed, and compared with available morphological and
ultrastructural data. The rDNA phylogenies indicate that the genus Chrysochromulina is paraphyletic and is divided into two
major clades. This is supported by ultrastructural and morphological data. There is a major split between Chrysochromulina
species with a saddle-shaped cell form (clade B2) and the remaining species in the genus (clade B1). Clade B2 includes the type
species C. parva and taxa belonging to this clade thus retain the name Chrysochromulina. The non-saddle-shaped
Chrysochromulina species analysed are closely related to Hyalolithus, Prymnesium and Platychrysis species. Imantonia species
are sister taxa to these species within clade B1. An amendment to the classification of the order Prymnesiales and the genera
Prymnesium, Platychrysis and Chrysochromulina is proposed with one new and one emended family (Chrysochromulinaceae
and Prymnesiaceae, respectively), two new genera (Haptolina and Pseudohaptolina), and one new species (Pseudohaptolina
arctica). We suggest a revision of the taxonomy of the Prymnesiales that is in accordance with available molecular evidence
The haptophyte order Prymnesiales Papenfusssensu Edvardsen & Eikrem currently contains onefamily, the Prymnesiaceae Conrad ex O.C.Schmidt, which at present comprises thegenera Chrysocampanula R.O. Fournier,Chrysochromulina Lackey, Corymbellus J.C.Green, Imantonia N. Reynolds, Hyalolithus M.Yoshida, T. Nakayama & I. Inouye, PlatychrysisGeitler ex Gayral & Fresnel and PrymnesiumMassart (Jordan et al., 2004). Most members ofthis order are unicellular, planktonic and scalybiflagellates, although colonial flagellates (e.g.Corymbellus species) occur, and some have benthicfilaments or amoeboid stages in their life cycle
(see reviews by Hibberd, 1980; Green et al.,1990). Most inhabit marine or brackish waters,but a few Chrysochromulina and Prymnesium spe-cies occur in fresh water (Preisig, 2003). Some spe-cies (e.g. of Chrysochromulina and Prymnesium)may form blooms, occasionally harmful to fishand other biota (reviewed by Moestrup, 1994;Edvardsen & Paasche, 1998; Edvardsen & Imai,2006). All known species are photosynthetic, butmixotrophy appears to be common in some genera(e.g. Chrysochromulina and Prymnesium: Kawachiet al., 1991; Nygaard & Tobiesen, 1993; Joneset al., 1994; Tillmann, 1998), and uptake of dis-solved organic carbon may also occur (Pintner &Provasoli, 1968).The genus Chrysochromulina was erected by
Lackey (1939) with the description of C. parva,one of only four freshwater species in the genus
Correspondence to: Bente Edvardsen. E-mail: bente.edvardsen@
bio.uio.no
ISSN 0967-0262 print/ISSN 1469-4433 online/11/03000202–228 � 2011 British Phycological Society
DOI: 10.1080/09670262.2011.594095
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
(Preisig, 2003). In the 1950s, Parke and colleagues(Parke et al., 1955, 1956, 1958, 1959) described aseries of marine Chrysochromulina species, andsince then many others have been added (see e.g.Leadbeater, 1972; Hallfors & Niemi, 1974; Estepet al., 1984; Moestrup & Thomsen, 1986; Kawachi& Inouye, 1993; Eikrem & Moestrup, 1998). Atpresent, 57 Chrysochromulina species have beenformally described. Fifty-four were listed byJordan and colleagues (Jordan et al., 2004), towhich can be added C. papillata Y. Gao,C.K. Tseng & Y. Guo (Gao et al., 1993), C. pla-nisquama X.Y. Hu, M.Y. Yin & C.K. Tseng(Hu et al., 2005), and C. palpebralis (Seoaneet al., 2009). Several others await formal descrip-tion (Jensen, 1998; LeRoi & Hallegraeff, 2004;Marchant et al., 2005), and the true number ofspecies may exceed 100 (Thomsen et al., 1994).Recent environmental clone library studiesrevealed the likely existence of hundreds of taxawithin the current confines of the genusChrysochromulina (Moon-van der Staay et al.,2000; Liu et al., 2009).Ten species of Prymnesium were listed by Jordan
et al. (2004). Of these, six have been described indetail and can be delineated by scale morphologyand DNA sequences (P. annuliferum, P. calathi-ferum, P. faveolatum, P. nemamethecum,P. parvum and P. zebrinum), to which can beadded P. lepailleurii Fresnel (Probert & Fresnel,2007). Five additional species have been reportedin the literature (P. czosnowskii Starmach, P. gla-diociliatum (Buttner) R.W. Jordan & J.C. Green,P. minutumN. Carter, P. papillatum Jiao-Fen Chen& C.K. Tseng and the type species P. saltansMassart), but these have not been well character-ized by electron microscopy (EM) or moleculargenetics. Type material of P. saltans no longerexists and was never studied using EM. Reportsof this species (e.g. Wang & Wang, 1992) havebeen scarce since its description by Massart(1920) and it has been suggested that P. parvumis identical to P. saltans (e.g. Moestrup, 1994;Guo et al., 1996).Although the number of species in both of these
genera is likely to increase as new taxa aredescribed, it may also be reduced as organisms for-merly believed to be different species are found tobe life-cycle stages. The two forms of P. parvum,f. parvum and f. patelliferum, were previously rec-ognized as separate species, but are now consideredto be stages in the life cycle of the same species(Larsen & Medlin, 1997; Larsen & Edvardsen,1998; Larsen, 1999). Chrysochromulina polylepisalso has an alternate stage with distinct scales(Paasche et al., 1990; Edvardsen & Paasche,1992), which is similar in overall cell morphologyand ultrastructure as well as having identical gene
sequences (Edvardsen et al., 1996; Edvardsen &Medlin, 1998) to the authentic cell type describedby Manton & Parke (1962). Flow cytometric anal-ysis has documented the presence of two ploidylevels in C. polylepis (Edvardsen & Vaulot, 1996)and in four other Chrysochromulina species(Edvardsen, 2006), suggesting that heteromorphiclife cycles may be common in the Prymnesiales.Members of the order Prymnesiales possess two
smooth flagella that are equal or unequal in length,but similar in form (Green & Hori, 1994). Anorganelle unique to haptophytes, termed the hap-tonema (Parke et al., 1955; Inouye & Kawachi,1994), is located between the flagella. The hapto-nema differs structurally from the flagella, usuallycontaining six or seven microtubules in thePrymnesiales, but sometimes eight as inChrysochromulina kappa (Manton & Leedale,1961a). The length of the haptonema variesgreatly, from being many times the cell diameterin some species of Chrysochromulina (e.g. 160 mmlong in C. camella: Leadbeater & Manton, 1969a)and coiling, to short and non-coiling inPrymnesium species (Green et al., 1982), and evenabsent in Imantonia (Green & Pienaar, 1977). Theflagellar apparatus typically consists of two basalbodies, the base of the haptonema, microtubular(simple or compound) roots (R1–R4), fibrousroots (‘cytoplasmic tongue’), and accessory andconnecting fibres (Green & Hori, 1994; Jordanet al., 1995; Inouye, 1997; Billard & Inouye,2004). The microtubular root R1, which is con-nected to the basal body of the mature or left fla-gellum (BB1), is present in all prymnesialeanspecies hitherto examined. R1 may be eithersimple (a sheet of microtubules; Fig. 1), or com-pound (a sheet and a crystalline bundle of micro-tubules; Fig. 2), and its structure varies within
Fig. 1. Drawing of the simple flagellar apparatus inChrysochromulina scutellum showing the positions of the
microtubular roots R1-R4, haptonematal base (H) andthe basal bodies 1 and 2 (BB1 and BB2). Modified fromEikrem & Moestrup (1998).
Taxonomic revision of the Prymnesiales 203
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
genera as they are presently defined. The microtu-bular root R2, with its origin between the twobasal bodies, is simple and comprises a few micro-tubules in all species examined so far (Billard &Inouye, 2004). The microtubular roots R3 andR4, associated with the right or immature flagellarbase (BB2), contain only a few microtubules, whichalso appears to be the rule in other haptophytes(Green & Hori, 1994).The two yellow-brown chloroplasts are envel-
oped by four chloroplast membranes and havelamellae composed of three thylakoids, as in mostmembers of the Haptophyta. The prymnesialeanchloroplasts have immersed or bulging pyrenoidstraversed by thylakoids or tubes. The nucleus isposterior or central and the outer nuclear mem-brane is continuous with the outer chloroplastmembrane. The large Golgi body, in which scalesare produced, lies anterior to the nucleus and justbeneath the basal bodies, and is composed of asingle fan-shaped dictyosome with many cisternae.The cisternae can be dilated, as in Chryso-chromulina chiton (Manton, 1967a; Hibberd,1976, 1980; Pienaar & Birkhead, 1994).There are no clear-cut morphological or ultra-
structural features that distinguish existing generawithin the Prymnesiales. The main character usedfor species identification within genera is the mor-phology and ornamentation of body scales, fea-tures that can usually be seen only by using theelectron microscope. The cell surface of all prym-nesialean species is covered by one to several layersof these scales. The basic scale is a round or ovalplate composed of microfibrils. In many species,the proximal side consists of microfibrils arrangedin a radial pattern, whereas the pattern of the distalside is more variable. The scales consist of organic
substances, mainly proteins and carbohydrates(Leadbeater, 1994 and references therein). InHyalolithus neolepis, one scale type of the non-motile form undergoes intracellular silicification(Yoshida et al., 2006). Chrysochromulina parkeaeJ.C. Green & B. Leadbeater has scales that maybe slightly calcified (R. Andersen, pers. comm., inSaez et al., 2004), and may be allied with coccolith-bearing species (Saez et al., 2004, Edvardsen &Medlin, 2007). The length of the haptonema rela-tive to either the flagellar length or cell body diam-eter, the cell size and form, and the swimmingbehaviour are additional features used in speciesidentification that can be seen under the lightmicroscope (LM). In addition, some species havelarge conspicuous scales that are visible by LM.Early on, Parke and co-workers recognized thegreat morphological variation within the genusChrysochromulina, but were reluctant to erectnew genera on characters that could only be seenin the electron microscope (Parke et al., 1955).Suggestions that the genus Chrysochromulina isan artificial grouping have also been based onultrastructural studies (Birkhead & Pienaar,1995a), 18S rDNA (Medlin et al., 1997;Edvardsen et al., 2000; Edvardsen & Medlin,2007) and rbcL DNA sequence analysis (Inouye,1997; Fujiwara et al., 2001).The aim of this study was to reconstruct the phy-
logeny of the Prymnesiales based on nuclear andplastid-encoded ribosomal DNA (18S, partial 28S,16S and concatenated) sequences and availablemorphological and ultrastructural data, and torevise the taxonomy in accordance with thesedata, thereby attempting to identify synapo-morphic and systematically useful morphologicalcharacter states within this ecologically and eco-nomically important group of photosyntheticprotists.
Materials and methods
Cultures
Twenty-one strains representing 14 prymnesialean spe-cies, presently maintained in the University of OsloCulture Collection of Algae (UIO), were isolated fromNorwegian coastal waters by the serial dilution methodor by single cell capillary isolation (Table 1). Of these 11were used in this study. Twenty-six additional strains ofPrymnesiales were obtained from the Algobank CaenCollection de Cultures de Microalgues – Universite deCaen Basse-Normandie (ALGOBANK-CAEN, previ-ously ALGO), Provasoli–Guillard National Center forCulture of Marine Phytoplankton (CCMP) and thePlymouth Culture Collection of Marine Algae (PCC)(Table 1). These originated mainly from NorthAtlantic coastal waters, but also from the Arctic, thePacific, New Zealand and from freshwater lakes.
Fig. 2. Drawing of the flagellar apparatus in Prymnesiumpalpebrale, comb. nov. (syn. Chrysochromulina palpebralis)
including a compound R1 flagellar root with the crystallinebundle of microtubules. R2 not shown. Abbreviations as inFig. 1. Modified from Birkhead & Pienaar (1995a).
Cultures of these strains were generally grown in IMR1/2 medium (Eppley et al., 1967), supplemented with10 nM selenium, in a 12:12 h light:dark cycle at aphoton flux rate of 50–100 mmolm–2 s–1. Cultures fromthe ALGOBANK culture collection were grown as inProbert & Houdan (2004). Cultures were harvested byfiltration or centrifugation (20 min, 7000 rpm, 4�C,RC-5B Sorvall Centrifuge).
PCR amplification and DNA sequencing
Total nucleic acids were extracted using a modifiedCTAB extraction (Doyle & Doyle, 1990) or by theQiagen DNeasy Plant Mini Kit (Qiagen, Hilden,Germany), and these served as the template for amplifi-cation of the 18S rDNA following Medlin et al. (1988),Chesnick et al. (1997), and Edvardsen & Medlin (1998).28S rDNA was amplified as in Edvardsen et al. (2003).PCR products were directly sequenced using a solidphase sequencing method with radioisotopes (Chesnicket al., 1997) or cycle-sequenced (Sequi-Therm,BIOZYM) using infra-red-labelled primers and analysedwith a LICOR automatic sequencer (MWG, Everbat,Germany). Alternatively, purified PCR products weresequenced directly using the DYEnamic ET terminatorCycle sequencing kit (Amersham Biosciences, USA)according to the manufacturer’s recommendations.The PCR fragments were then bidirectionally sequencedusing the primers as described in Edvardsen et al. (2003)on a MEGABACE (Amersham Biosciences, Germany)automatic sequencing device at the Department ofBiology, University of Oslo, Norway. Some templateswere cloned (LigAtor, R&D Systems, the Netherlands)prior to automatic sequencing. Amplification andsequencing of the partial 16S rDNA region of some ofour strains were undertaken by S. McDonald(McDonald et al., 2007).
Phylogenetic analyses
The rDNA sequences generated in this study and allpublicly available nearly full-length 18S, and partial28S and 16S rDNA sequences of cultured prymnesialeantaxa were included, and then duplicates removed(Table 1). A set of 18S rDNA environmental sequencesof oceanic picoplankton (Moon-Van der Staay et al.,2000, Table 1) were also included to increase thenumber of taxa. Sequences of haptophyte taxa basal toPrymnesiales in previous phylogenetic studies (e.g.Edvardsen et al., 2000; Medlin et al., 2008) were selectedas outgroups to root the trees. Four datasets were gen-erated: 18S (63 taxa, 1838 characters), 28S (37 taxa, 1086characters), 16S (22 taxa, 701 characters) and a conca-tenated dataset of the three genes, including taxa whereboth 18S and 28S rDNA sequences were available(30 taxa, 3625 characters), with the 16S rDNA for 13taxa (Table 1). The single gene datasets were alignedusing MAFFT v6 Q-INS-I model (Katoh & Toh,2008), considering secondary RNA structure (defaultparameters) followed by editing by eye in BioEditv7.0.9 (Hall, 1999). The concatenated dataset was gen-erated by combining the three alignments. The 28S
rDNA and concatenated alignments included hypervar-iable regions and were treated with Gblocks v0.91b(Castresana, 2000), under default parameters, to excludepoorly aligned positions from the phylogenetic infer-ence, reducing the number of characters to 2758 (1669in 18S, 389 in 28S, 700 in 16S) in the concatenated and384 in the 28S rDNA datasets. All four datasets werethen analysed with MODELTEST (Posada & Crandall,1998) to establish the optimal model of nucleotide evo-lution; for all alignments the general time reversiblemodel (GTR) was preferred for both the Akaike andBayesian information Criterion (AiC and BiC).Maximum-likelihood (ML) analyses were performedusing RAxML v.7.0.3 (Stamatakis, 2006). The generaltime reversible model with parameters accounting forinvariable sites (I) and gamma-distributed (G) rate var-iation across sites with four discrete rate categories wasused for all four rDNA datasets. The bootstrap analyseshad 100 replicates for the ML analysis. Bayesian infer-ence (BI) under the same evolutionary model was per-formed with MrBayes v. 3.1.2 (Ronquist &Huelsenbeck, 2003). Two Markov Chain Monte Carlo(MCMC) runs each with four chains were performed for5 000 000 generations, where the average standard devi-ation of split frequencies were 50.01. Trees were sam-pled every 100 generations. Bayesian posteriorprobabilities (PP) were calculated from the majority-rule consensus of the tree sampled after the initialburn-in phase. The MAFFT alignments were performedonline at CBRC (www.mafft.cbrc.jp), the Gblock atPhylogeny.fr (www.phylogeny.lirmm.fr), and theModeltest, ML and BI at BioPortal (www.bioporta-l.uio.no). The alignments are available as supplementaryfiles (available via the Supplementary Content tab onthe article’s online page at http://dx.doi.org/10.1080/09670262.2011.594095).
Results and discussion
Molecular phylogenetics
This study produced 53 new sequences from 37different strains representing 30 species in the hap-tophyte order Prymnesiales: 29 nearly complete18S and 24 partial 28S rDNA sequences(Table 1). Bayesian and ML analyses recoveredtrees of almost identical topology. The four con-sensus Bayesian rDNA trees generated are shownin Fig. 3 (18S), Fig. 4 (28S), Fig. 5 (16S) and Fig. 6(combined 18S, 28S and 16S ¼ concatenated) usingthe new names proposed in this study. These fourphylogenies are largely congruent with each otherand subsequently we refer mainly to the 18S rDNAtree because it contains the most characters andtaxa. The single gene partial 28S rDNA dataset,including only 384 characters, generated a weaklyresolved phylogeny.In our 18S rDNA tree (Fig. 3), clades are num-
bered according to our previous work (Edvardsenet al., 2000) and newly recognized clades B1-1 toB1-6. The first divergence within the class
B. Edvardsen et al. 210
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
Prymnesiophyceae (the ingroup) corresponds tothe order Phaeocystales, clade A (see Langeet al., 2002 for full description) plus clade D, con-sisting of some environmental sequences knownonly from a picoplanktonic clone library
originating from oligotrophic waters in thePacific Ocean, here termed OLI sequences(Moon-Van der Staay et al., 2000). Clade D is suf-ficiently genetically divergent to be recognized as aseparate order of haptophyte algae once the
0.1
Prymnesium sp. HAPptPrymnesium pigrum
0.90/58
Prymnesium pienaarii + P. simplex0.57/-
Prymnesium zebrinum
0.89/-
Prymnesium nemamethecumPrymnesium parvum
1.00/850.88/-
Prymnesium faveolatumPrymnesium calathiferum + P. annuliferum
1.00/85
0.87/-
Prymnesium polylepisPrymnesium neolepis
0.89/71
Prymnesium aff. polylepis1.00/56
0.99/-
Prymnesium chitonPrymnesium minus0.72/-
1.00/76
Prymnesium kappaPrymnesium palpebrale1.00/86
1.00/77
Haptolina hirta + H. ericinaHaptolina cf. herdlensis
Fig. 3. Consensus Bayesian tree based on nuclear 18S ribosomal encoding DNA sequences of members of thePrymnesiophyceae. Posterior probability (left) and maximum-likelihood bootstrap values (right)40.5/50 are shown above
or below the branches. Pavlova gyrans and Rebecca salina (Pavlovophyceae) were used as outgroups. Scale bar representsnumber of substitutions/site.
Taxonomic revision of the Prymnesiales 211
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
morphology of these cells is known. The nextmajor divergence is that of the orderPrymnesiales, corresponding to clades B1 and B2,from the cluster of taxa in clades C and E. Clade Cembraces the orders Coccolithales andIsochrysidales. Clade E here contains two pico-planktonic clone library OLI sequences found togroup with a sequence of a newly described non-mineralized species Chrysoculter rhomboideusT. Nakayama, M. Yoshida, M.-H. Noel,M. Kawachi & I. Inouye in the phylogeny pre-sented by Nakayama et al. (2005).
Prymnesiales
The divergences within the order Prymnesiales arethe primary focus of this study. In Table 2, avail-able morphological and ultrastructural data for the
prymnesialean species analysed here are summa-rized. In the 18S rDNA Bayesian tree (Fig. 3) wefind two distinct clades with high posterior proba-bilities (PP 0.91 and 1.00, respectively), termed B1(consisting of several subclades) and B2. These twomajor clades are also recovered and strongly sup-ported in previous SSU rDNA phylogenies(Medlin et al., 1997; Simon et al., 1997;Edvardsen et al., 2000; Saez et al., 2004;Edvardsen & Medlin, 2007). This major divergencebetween clades B1 and B2 is also recovered in the16S and concatenated rDNA phylogenies (Figs 5,6, support values PP 0.7–1.0), but with an unre-solved placement of Chrysocampanula spinifera(Fig. 6, see below). Species of the genusChrysochromulina fall into both of these cladesand our findings support the views of Birkhead &Pienaar (1995a), based on ultrastructural data, and
Fig. 4. Consensus Bayesian tree based on partial nuclear 28S ribosomal encoding DNA sequences of members of thePrymnesiales. Posterior probability (left) and maximum-likelihood bootstrap values (right)40.5/50 are shown above orbelow the branches. Cruciplacolithus neohelis, Emiliania huxleyi and Pleurochrysis carterae were used as outgroups. Scale
bar represents number of substitutions/site.
B. Edvardsen et al. 212
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
of Inouye (1997), based on rbcL DNA sequencedata, that the genus Chrysochromulina is notmonophyletic.
Clade B2 – Chrysochromulina
Clade B2 contains the type species of the genusChrysochromulina, C. parva, and therefore thesespecies retain the name Chrysochromulina. All ofthe taxa in clade B2, except for C. leadbeateri, areallied by their basic cell shape and morphology(Fig. 7). This clade includes all of the saddle-shaped Chrysochromulina species examined here(C. acantha, C. campanulifera, C. cymbium, C. lead-beateri, C. rotalis, C. simplex, C. scutellum, C. stro-bilus and C. throndsenii). Available data on theultrastructure of the saddle-shaped species havebeen reviewed by Eikrem & Moestrup (1998).Along with a common shape, these species allhave a very long, coiling haptonema with six or
seven microtubules in the free part. Their two fla-gella have both proximal and distal transitionalplates, which is the case in all species ofChrysochromulina and Prymnesium where thischaracter has been examined. All taxa in cladeB2, except for C. leadbeateri, have flagella thatare inserted subapically and ventrally. The flagellarroot R1 is generally simple (only a sheet of micro-tubules with no crystalline bundle of microtubules)and contains relatively few microtubules (usually510, Fig. 1). One exception to this is C. acantha,which may have up to 20 microtubules in the R1 ina sheet (Table 2). The cells contain two chloro-plasts with immersed pyrenoids traversed bytubes or thylakoids and are covered by manysmall, delicate scales that may be cup-shaped orhave a relatively short spine.Chrysochromulina leadbeateri is grouped
together with the saddle-shaped species in the 18SrDNA tree (Fig. 3), albeit in a distinct lineage with
0.1
Prymnesium polylepis UIOA
Prymnesium polylepis UIOB152j1.00/99
Prymnesium aff. polylepis PCC200
0.67/62
Prymnesium sp. MBIC10516
0.99/68
Haptolina cf. herdlensis
Haptolina hirta0.96/96
0.91/52
Imantonia rotunda UIO101
Imantonia rotunda RCC406
Imantonia rotunda RCC305
Imantonia sp. MBIC10497
1.00/100
Prymnesiaceae sp. MBIC10518
0.93/61
Chrysochromulina leadbeateri
Chrysochromulina cymbium0.99/100
Chrysochromulina acantha UIO024
Chrysochromulina acantha RCC3390.94/99
0.79/-
Chrysochromulina simplex
Chrysochromulina sp. MBIC10513
Chrysochromulina throndsenii
1.00/100
Chrysochromulina campanulifera
0.87/55
0.70/-
Chrysochromulina sp. clone124
Emiliania huxleyi
Dicrateria inornata1.00/100
B1-5
B2
B1-2
B1-4
Outgroup
B1-6
Fig. 5. Consensus Bayesian tree based on partial plastid 16S ribosomal encoding DNA sequences of members of the
Prymnesiales. Posterior probability (left) and maximum-likelihood bootstrap values (right)40.5/50 are shown above orbelow the branches. Dicrateria inornata and Emiliania huxleyi were used as outgroups. Scale bar represents number of sub-stitutions/site.
Taxonomic revision of the Prymnesiales 213
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
four OLI sequences corresponding to taxa withunknown morphology. Chrysochromulina simplexis sister to this lineage in the 18S, 28S and conca-tenated rDNA trees (Figs 3, 4, 6). Although cells ofC. leadbeateri may be round and have a coiledhaptonema approximately equal in length to theflagella, and flagella that are inserted apically(Eikrem & Throndsen, 1998), it shares other fea-tures with the saddle-shaped species, such asimmersed pyrenoids, six microtubules in the emer-gent part of its haptonema and a simple R1 flagel-lar root. Chrysochromulina leadbeateri is believedto constitute a species complex that exhibits somevariations in scale and cell morphology, and somecells belonging to this complex possess the saddle-shape characteristic of clade B-2 and a long hap-tonema (Eikrem & Throndsen, 1998). We thereforesuggest that the saddle-shaped cell form is a syn-apomorphy for clade B2 Chrysochromulina species,
but that it has been lost in one or more formswithin the C. leadbeateri complex. The saddleshape has been reported in the calcified, non-photosynthetic genus Ericiolus H.A. Thomsen,which is only known from TEM whole-mounts ofnatural samples (Thomsen et al., 1995). In ouropinion, from the TEM images presented byThomsen et al. (1995), the cells in the two speciesof Ericiolus correspond in shape more to the’strawberry-shaped’ cells of certain coccolitho-phores (e.g. Syracosphaera) than to the typicalsaddle shapes of clade B2 Chrysochromulina spe-cies. In addition, unlike the saddle-shapedChrysochromulina, the two species of Ericiolushave a haptonema that is shorter than the flagella.Chrysochromulina campanulifera, C. strobilus
and C. cymbium are grouped together in our 18SrDNA tree (Fig. 3). Cells of these species are cov-ered by small cup-shaped scales, in addition to an
0.01
Prymnesium zebrinum
Prymnesium pigrum
0.61/-
Prymnesium nemamethecum
Prymnesium parvum
1.00/100
0.81/58
Prymnesium pienaarii
0.96/59
Prymnesium faveolatum
Prymnesium calathiferum1.00/84
1.00/71
Prymnesium aff. polylepis
Prymnesium polylepis
0.99/67
Prymnesium neolepis1.00/68
1.00/65
Prymnesium palpebrale
0.59/-
Prymnesium kappa
1.00/88
Haptolina fragaria
Haptolina brevifila
1.00/99
Haptolina hirta + H. ericina
Haptolina cf. herdlensis1.00/100
1.00/96
1.00/77
Imantonia rotunda
Imantonia sp. CCMP14041.00/100
Pseudohaptolina arctica CCMP1204
1.00/89
1.00/72
Chrysochromulina acantha
Chrysochromulina rotalis
0.71/57
Chrysochromulina throndsenii
1.00/98
Chrysochromulina campanulifera
1.00/74
Chrysochromulina cymbium
1.00/90
Chrysochromulina simplex
Chrysochromulina leadbeateri1.00/83
1.00/98
Chrysocampanula spinifera
Cruciplacolithus neohelis
Pleurochrysis carterae
0.99/70
Emiliania huxleyi
1.00/100
B1-5
B2
B1-2
B1-4
Outgroup
B1-3
B1-1
Fig. 6. Consensus Bayesian tree based on concatenated nuclear 18S and partial 28S, and plastid 16S ribosomal encoding DNAsequences of members of the Prymnesiales. Posterior probability (left) and maximum-likelihood bootstrap values (right)40.5/50 are shown above or below the branches. Cruciplacolithus neohelis, Emiliania huxleyi and Pleurochrysis carterae were used as
outgroups. Scale bar represents number of substitutions/site.
B. Edvardsen et al. 214
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
Table
2.Morphologicalandultrastructuralinform
ationonselected
speciesfrom
theorder
Prymnesiales.ND¼nodata;MT¼microtubules.
Species
Cellshape
Flagella
Haptonem
a(H
)
length
relative
toflagella
(F)
Scales
Scale
patterning
No.of
MT
in
emergent
part
of
haptonem
a
Flagellar
microtubular
roots
Flagella
withproxim
al
and/or
distalplates
Flagella
with
helical
band
or
other
structures
Cytoplasm
ic
tongue
Pyrenoids
‘Peculiar’
golgi
body
References
CladeB2
Chrysochromulina
acantha
saddle
equal
H4F
plate,spine
loosely
woven
radial
ribsoverlaying
concentric.ribs
7simple
r1
(manyMT)
proxim
aland
distalplates
absent
absent
immersed
ND
Leadbeater&
Manton,
1971;Gregson
etal.,1993
C.campanulifera
saddle
equal
H44
Fplate,cup
radialpatternboth
faces
6simple
r1ND
absent
absent
immersed,
traversedby
thylakoids
present
Unpublished
results
C.campanulifera
saddle
equal
H44
Fplate,cup
radialpatternboth
faces
6ND
ND
ND
ND
immersed,tra-
versedby
thylakoids
ND
Manton&
Leadbeater,
1974
C.cymbium
saddle
equal
H44
Fplate,cup
radialpatternboth
faces
6ND
ND
ND
ND
immersed
present
Parkeet
al.,1959;
Leadbeater&
Manton1969a
C.leadbeateri
spherical
equal–slightly
unequal
H¼
Fplate
(2types)
radialribsandconcen-
tric
rings,central
open
ringwith
cross
6simple
r1proxim
aland
distalplates
ND
ND
immersed,
traversedby
tubules
present
Eikrem
&Throndsen,
1998
C.parva
saddle
equal
H44
Fplate
loosely
woven
radial
ribsoverlaying
concentric
ribs
ND
simple
r1ND
ND
ND
ND
ND
Lackey,1939;Parke
etal.,1962;
Moestrup&
Thomsen,1986
C.rotalis
saddle
equal
H44
Fplate,spine
radialribsproxim
al
face,concentric
fibrilsdistalface
ND
ND
ND
ND
ND
immersed
ND
Eikrem
&Throndsen,
1999
C.scutellum
saddle
equal
H44
Fplate,spine(2
types)
radialribsproxim
al
face,concentric
fibrilsdistalface
7simple
r1proxim
aland
distalplates
absent
absent
immersed,
traversedby
thylakoids
absent?
Eikrem
&Moestrup,
1998
C.simplex
saddle
equal
H44
Fplate
radialribsproxim
al
face,concentric
fibrilsdistalface
7simple
r1proxim
aland
distalplates
absent
absent
immersed,
traversedby
tubules
present
Estep
etal.,1984;
Birkhead&
Pienaar,1995b
C.strobilus
saddle
equal
H44
Fplate,cup
radialpatternboth
faces
6ND
ND
ND
ND
immersed
present
Parkeet
al.,1959;
Leadbeater&
Manton
1969a,1969b
C.throndsenii
saddle
equal
H44
Fplate
(2types)
radialribsoverlayinga
spirallingrib,cen-
tralopen
ringwith
cross
6simple
r1proxim
aland
distalplates
absent
absent
immersed,
traversedby
tubules
absent
Eikrem,1996;Eikrem,
unpublished
CladeB1
Chrysocampanula
spinifera
bell-shaped
unequal
H5F
plate,spine
radialribsoneface,
fibrilsother
face
ND
ND
ND
ND
ND
ND
ND
Fournier,1971;
Pienaar&
Norris,
1979
Haptolinabrevifila
spherical
equal
H5F
plate
(2types),spine
radialribsoverlaying
concentric
fibrils
both
faces
7simple
r1
(manyMT)
proxim
aland
distalplates
tubularrings
present
immersed,
traversedby
thylokoids
ND
Birkhead&
Pienaar,
1994a
H.ericina
oval–oblong
equal
H4F
plate,spine
radialribsoneface,
concentric
fibers
other
face
7ND
ND
ND
ND
immersed
present
Parkeet
al.,1956;
Manton&
Leedale,
1961b
H.herdlensis
conical
equal
H5F
plate
(3types)
radialribs
ND
ND
ND
ND
ND
ND
ND
Leadbeater,1972
(continued
)
Taxonomic revision of the Prymnesiales 215
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
Table
2.Continued.
Species
Cellshape
Flagella
Haptonem
a(H
)
length
relative
toflagella
(F)
Scales
Scale
patterning
No.of
MT
in
emergent
part
of
haptonem
a
Flagellar
microtubular
roots
Flagella
withproxim
al
and/or
distalplates
Flagella
with
helical
band
or
other
structures
Cytoplasm
ic
tongue
Pyrenoids
‘Peculiar’
golgi
body
References
H.hirta
spherical–oblong
equal
H4F
plate,spine(2
types)
radialribs.oneface,
fibrilsother
face
ND
ND
ND
ND
ND
ND
ND
Manton,1978
H.fragaria
spherical
equal
H5F
plate
(2types)
radialribsoverlaying
concentric
fibrils
both
faces
7ND
ND
tubularrings
ND
immersed
ND
Eikrem
&Edvardsen,
1999
Imantonia
rotunda
spherical
equal
Absent
plate
(2types)
radialribs(few
)some
concentric
ribs
nohapt.
simple
r1(possible
vestigeof
crystal)
proxim
aland
distalplates
ND
ND
immersed,
traversedby
thylakoids
ND
Reynolds,1974;
Eikrem,unpub-
lished;Green
&
Pienaar,1977;
Green
&Hori,
1986
P.annuliferum
oblong
subequal
H55
Fplate
(2types)
radialribsoneface,
concentric
fibrils
other
face
ND
ND
ND
ND
ND
ND
ND
Billard,1983
P.calathiferum
oblong–spherical
equal–subequal
H55
Fplate
(2types)
radialribs.both
faces
ND
ND
ND
ND
ND
immersed
ND
Chang&
Ryan,1985
P.chiton(PCC
146)
spherical–oblong
subequal–equal
H4F
plate
(2types)
radialribsdistalface,
fibrilsproxim
al
face
7ND
ND
ND
ND
bulging,
traversedby
thylakoids
present
Parkeet
al.,1958;
Manton,1967a
P.chiton
(Leeds
isolate)
probably
spherical–
oblong
probably
subeq-
ual–equal
Probably
H4F
plate
(2types)
radialpatternboth
faces
7ND
ND
ND
ND
bulging,
traversedby
thylakoids
present
Manton,1966,1967a,
1967b
P.faveolatum
oblong–veryelongated
subequal
H55
Fplate
(2types)
radialribs.both
faces
7simple
r1(m
any
MT)
ND
tubularrings
present
immersed,
traversedby
thylakoids
ND
Fresnel
etal.,2001
P.kappa
spherical
equal
H5F
plate,spine
radialribsproxim
al
face,fibrilsdistal
face
7-8
ND
ND
ND
ND
bulging
present
Parkeet
al.,1955;
Manton&
Leedale,
1961a
P.kappa
(Norw
egianisolate)
spherical
equal
H5F
plate
(2types),spine
radialribsproxim
al
face,fibrilsdistal
face
7compoundr1
proxim
aland
distalplates
ND
ND
bulging,
traversedby
thylakoids
present
Eikrem,unpublished
P.minus
spherical–oval
equal
H5F
plate
(2types)
radialpatternboth
faces
7ND
ND
ND
ND
bulging,
traversedby
thylakoids
Parkeet
al.,1955;
Manton&
Leedale,
1961a;Eikrem
etal.,1998
P.nem
amethecum
oblong–spherical
equal–subequal
H55
Fplate
(2types
onbody,
1typeon
haptonem
a)
radialribsboth
faces
7compoundr1
proxim
aland
distalplates
tubularrings
present
(reduced)
bulging,
traversedby
thylakoids
present
Birkhead&
Pienaar,
1994b;Pienaar&
Birkhead,1994
P.neolepis
(motile
form
)
subspherical,pear-
shaped
equal
H�F
spine(1
type)
perforatedporesin
radiatingpattern
ND
ND
ND
ND
ND
ND
ND
Yoshidaet
al.,2006
P.neolepis
(non-m
otile
form
)
spherical–subspherical
equal
H4F
plate
(1type)
siliceous
radialandconcentric
ribs
7compoundr1
proxim
aland
distalplates
ND
ND
immersed,
traversedby
thylakoid
present
Yoshidaet
al.,2006
P.neustophilum
rounded
orangular
shape
subequal
H55
FND
ND
ND
ND
ND
ND
ND
ND
ND
Norris,1967
P.palpebrale
oblong–asymmetrical
equal–subequal
H5F
plate
(2types),spine(1
type)
radialribsboth
faces
7compoundr1
proxim
aland
distalplates
tubularrings
absent
immersed,
traversedby
thylokoids
ND
Birkhead&
Pienaar,
1995a;Seoane
etal.,2009
P.parvum
f.parvum
oblong–spherical
equal–subequal
H55
Fplate
(2types)
radialribsdistalface,
fibrilsproxim
al
face
7simple
r1(m
any
MT)
proxim
aland
distalplates
ND
ND
immersed
ND
Manton&
Leedale,
1963;Manton,
1964
B. Edvardsen et al. 216
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
P.parvum
f.
patelliferum
oblong–spherical
equal–subequal
H55
Fplate
(2types)
radialribsboth
faces
ND
simple
r1(m
any
MT)
proxim
aland
distalplates
ND
ND
immersed,
traversedby
thylakoids
ND
Green
etal.,1982;
Green
&Hori,
1990
P.pienaarii
rounded
orangular
subequal
H55
Fplate
(2types)
radialribsandconcen-
tric
pattern
ND
compoundr1
ND
ND
ND
immersed,
traversedby
thylakoids
present
Gayral&
Fresnel,1983
P.pigrum
motile
cells
sphericalwith
apicaldepression
equal–subequal
H55
Fplate
(2types)
radialribsandconcen-
tric
pattern
7broadroot430
MT,com-
poundr1
ND
ND
ND
immersed,
traversedby
thylakoids,
withstigma
present
Chretiennot,1973
P.polylepis
(alternate
type)
oblong–spherical
equal
H5F
plate
(2types),spine
radialribsboth
faces
overlayingconcen-
tric
fibrils
7compoundr1
proxim
aland
distalplates
other structure
present
immersed
traversedby
thylakoids
present
Manton&
Parke,1962;
Eikrem
unpublished
P.polylepis
(authentictype)
oblong–spherical
equal
H5F
plate
(4types)
radialpatternboth
faces,perforations
7compoundr1
proxim
aland
distalplates
other structure
present
immersed
traversedby
thylakoids
present
Paascheet
al.,1990;
Edvardsenet
al.,
1996;Eikrem
unpublished
P.aff.polylepis
(PCC
200)
oblong–spherical
equal
H5F
plate
(5types)
radialribsboth
faces
ND
ND
ND
ND
ND
ND
ND
Edvardsen&
Medlin,
1998;Edvardsen&
Eikrem,
unpublished
P.simplex
rounded
orangular
H55
Fplate
(2types)
radialribsboth
faces
ND
compoundr1
ND
ND
ND
immersed,
traversedby
thylakoids
ND
Gayral&
Fresnel,1983
P.zebrinum
oblong
subequal
H55
Fplate
(2types)
radialribsboth
faces
ND
ND
ND
ND
ND
ND
ND
Billard,1983
Pseudohaptolina
arctica
(CCMP1204)
spherical
equal
H�
Fplate
(2types)
radialribsboth
faces
ND
ND
ND
ND
ND
ND
ND
Eikrem
&Edvardsen,
1999
Taxonomic revision of the Prymnesiales 217
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
inner layer of plate scales. Chrysochromulinacamella, whose 18S rDNA sequence has not beenanalysed, clusters with C. cymbium in the 28SrDNA tree (Fig. 4), as expected from similaritiesto this group in body shape and scale morphology(Leadbeater & Manton, 1969a, 1969b).Ultrastructural data are missing for C. cymbium.Chrysochromulina ahrengotii M.Ø. Jensen &
Moestrup, C. alifera Parke & Manton, C. aphelesMoestrup & H.A. Thomsen, C. ephippium Parke &Manton and C. pontica Rouchijajnen also have asaddle-shaped cell form, ventrally inserted flagellaand a long, coiling haptonema and are expected tobelong to clade B2.
Clades B1 and B1-1
Clade B1 is morphologically more heterogeneousthan clade B2 and is also less robust in all trees(Figs 3–6). It contains members of the currentgenera Chrysocampanula (in the 18S rDNA tree),Chrysochromulina, Hyalolithus, Imantonia,Platychrysis and Prymnesium. We propose severaltaxonomic revisions within this clade and the pro-posed new names (for which see ‘Taxonomic rec-ommendations’) are henceforth used in figures andtables and introduced in the text below. In the 18SrDNA tree, clade B1 is a polytomy of threelineages: the first (B1-1) is composed ofChrysocampanula spinifera, the second (B1-6) ofChrysochromulina parkeae and the cultured strainMBIC10518 and the third consists of three well-supported subclades (B1-2 þ B1-3, B1-4, B1-5, allPP¼ 1.0) and a clade of picoplanktonic OLIsequences. The placement of Chrysocampanula spi-nifera (B1-1) within the B1 clade is well supportedby the 18S Bayesian analysis (PP40.9), but the B1clade is weakly supported in ML analysis of thisgene. In the 28S rDNA (Fig. 4) and the concate-nated (Fig. 6) trees the position of C. spinifera isunresolved. Chrysocampanula spinifera was trans-ferred to Chrysochromulina by Pienaar & Norris(1979), but was moved back to Chrysocampanulaby Jordan et al. (2004) based on the molecular datapresented in Saez et al. (2004). Chrysocampanulaspinifera has bell-shaped cells covered by platescales and conspicuous spine scales (Fig. 8). Thehaptonema is rigid and does not coil and is shorterthan the heterodynamic and unequal flagella(Pienaar & Norris, 1979). The flagellar root R1 isprobably compound (Kawachi, pers. comm., inBirkhead & Pienaar, 1995a). Morphological char-acters thus suggest that C. spinifera is affiliated toclade B1. The clear divergence of its 18S and 28SrDNA sequences from the rest of the Prymnesialesand its unique combination of characters justifythe reinstatement of the genus Chrysocampanula.Its unresolved placement within the Prymnesiales
in our rDNA phylogenies and its aberrant cellform could suggest that it represents a separatefamily but more molecular and morphologicaldata are needed for C. spinifera and closely relatedtaxa to clarify their systematic position.In this study Chrysochromulina parkeae was
placed within Prymnesiales, in a well-supportedclade together with the cultured strainMBIC10518 (B1-6). Previous phylogenetic analy-ses of haptophytes based on 18S rDNA placedC. parkeae within the coccolithophorid clade(Saez et al., 2004; Edvardsen & Medlin, 2007). Aswith C. spinifera, more molecular and morpholog-ical data of C. parkeae are needed to clarify itssystematic position.
Clades B1-2 and B1-3
Clade B1-2 þ B1-3, recovered in the 18S and con-catenated rDNA phylogenies, is composed ofImantonia and strain CCMP 1204. Imantoniarotunda and Imantonia sp. (strain CCMP 1404)form a monophyletic lineage (clade B1-2), andtogether with strain CCMP 1204 (B1-3) form asister group to the remaining clade B1 species(clades B1-4 and B1-5) in the concatenatedrDNA tree (Fig. 6) and a polytomy with theseclades in the 18S rDNA phylogeny (Fig. 3).Imantonia rotunda has immersed pyrenoids and fla-gella with distal and proximal transitional plates,which supports a closer relationship to species ofChrysochromulina and Prymnesium than toIsochrysis, with which it was previously grouped(Jordan & Green, 1994). Imantonia was alsoplaced near the non-saddle-shaped Chryso-chromulina species in the rbcL gene tree presentedby Inouye (1997). Members of Imantonia are pico-plankters with spherical cells and with no traces ofan emergent haptonema except for a proboscis(Fig. 9). The cells possess a deviant flagellar appa-ratus with many flagellar roots, each composed ofonly a few microtubules. There is no bundle ofmicrotubules associated with R1 or R2, but theroot termed R5 by Green & Hori (1986) may beinterpreted as a vestige of a crystalline array (abundle of many closely packed microtubulesbranching off the sheet) associated with R1 (R1c,Inouye, 1997; Eikrem & Moestrup, 1998). We sug-gest that the haptonema was lost in the ancestorsof Imantonia and its absence is a derived characterfor this genus. Imantonia rotunda is the type speciesof the genus and is included in all our rDNA phy-logenies. Our molecular data and morphologicalrevision support the interpretation that taxawithin the Imantonia group should all retain theirgenus name.In contrast, the sister taxon to Imantonia, repre-
sented by strain CCMP 1204 isolated from Arctic
B. Edvardsen et al. 218
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
waters, has many features in common with theremaining non-saddle-shaped Chrysochromulinaspecies in clade B1-4 (rounded cell shape, non-coil-ing haptonema shorter than flagella, and plate-likebody scales; Eikrem & Edvardsen, 1999), suggest-ing these are shared primitive characters (symple-siomorphies) of clade B1. Based on a combinationof morphological and molecular rDNA data, wepropose the erection of a new genus for this spe-cies, Pseudohaptolina arctica, gen. et sp. nov.
Clades B1-4 and B1-5
The remaining taxa in clade B1 fall into twowell-supported clades, B1-4 and B1-5, in our 18S,16S and concatenated rDNA Bayesian inferences(Figs 3, 5, 6, PP� 0.96). Clade B1-4 contains thecurrent species Chrysochromulina brevifilum, C. eri-cina, C. fragaria, C. cf. herdlensis and C. hirta, andwe propose the creation of a new genus, Haptolina,to embrace species in this clade. Members of thegenera Hyalolithus (H. neolepis), Platychrysis(Pl. pienaarii, Pl. pigra, Pl. simplex and Pl. sp.strain RCC1385 in this study) and Prymnesium,together with some non-saddle-shaped Chryso-chromulina species (C. chiton, C. kappa, C. minor,C. palpebralis, C. polylepis and C. sp. strain MBIC10516 in this study) cluster together in clade B1-5in the 18S, 16S and concatenated trees, and arehere transferred to the genus Prymnesium (seebelow). Members of Haptolina (clade B1-4) forma sister group to Prymnesium (clade B1-5) withhigh support (PP40.96) in the concatenatedrDNA tree.Haptolina species of clade B1-4 have a symmet-
rical spherical to oval cell shape, apically insertedappendages and a homodynamic swimming pat-tern. The haptonema is shorter, equal to or some-what longer than the flagella and is non-coiling orcoiling (see below; Fig. 10). The cells may havelarge spiny scales (Haptolina (Chrysochromulina)ericina and Haptolina (Chrysochromulina) hirta,Fig. 10). The haptonema of Haptolina (Chryso-chromulina) fragaria is shorter than the flagellaand has seven microtubules in the emergent part.Like Haptolina brevifila (Chrysochromulina brevifi-lum), the proximal part of the flagella ofH. fragaria has tubular rings (Eikrem, unpublishedobservations), and the pyrenoids may be bulging(Eikrem & Edvardsen, 1999). The flagellar appara-tus in H. brevifila includes a R2 root similar to thatin Prymnesium species, but which was reduced orlacking in two saddle-shaped Chrysochromulinaspecies (Birkhead & Pienaar, 1994a). The R1 fla-gellar root in H. brevifila, however, is simple andlacks closely packed bundles of microtubulesforming a compound root, as in Prymnesiumspecies including Prymnesium polylepis
(Chrysochromulina polylepis) within clade B1-5(Birkhead & Pienaar, 1994a, see below). The hap-tonema in H. brevifila, although shorter than theflagella, has the ability to coil and in fact it appearsthat this is associated with the presence of a simpleR1 flagellar root throughout the order Prymne-siales. The cells of H. ericina and H. hirta have ahaptonema that is longer than the flagella and ableto coil. Ultrastructural data are limited for thesetwo latter species, but their body shape, scale types(including large spiny scales) and 18S rDNAsequences ally them with clade B1 species.The species belonging to clade B1-5 are here
transferred to Prymnesium. They include the spe-cies known as Chrysochromulina polylepis,C. chiton, C. minor, C. kappa and all members ofPlatychrysis, as well as species already classified inPrymnesium. These species have a non-coiling hap-tonema slightly or much shorter than their flagella[Fig. 11, Table 2; the haptonema in Prymnesium(Chrysochromulina) chiton is however longer thanthe flagella], oblong and asymmetrical cell shape,and more or less heterodynamic flagella. They mayalso have features commonly associated with coc-colithophorids, for example a compound R1 root(a sheet plus a bundle of microtubules) and bulgingpyrenoids, or they possess features that are usuallyabsent in the Chrysochromulina species of clade B2(see above). Both P. nemamethecum and P. palpeb-rale (Chrysochromulina palpebralis), which was firstdescribed by Birkhead & Pienaar (1994b, 1995a)under the name ‘eyelash Chrysochromulina sp.’and formally described by Seoane et al. (2009),have a compound root associated with their rightflagellum (R1) and a fibrous root resembling whatis termed a cytoplasmic tongue in the coccolitho-phorids (Beech & Wetherbee, 1988; Fresnel, 1989;Seoane et al., 2009). In addition to the distal andproximal bands in the flagella, P. palpebrale has ahelical band in its flagella (Birkhead & Pienaar,1994b, 1995a; Seoane et al., 2009), as do some coc-colithophorids. Of the species we have sequenced,both P. polylepis and P. kappa (C. kappa) havecompound R1 flagellar roots (Edvardsen et al.,1996; Eikrem & Moestrup, unpublished observa-tions). Both P. parvum f. patelliferum andP. parvum f. parvum have simple R1 flagellarroots, but the sheet contains many microtubules(420) as in P. nemamethecum. Clade B1 specieshave immersed pyrenoids, except for P. kappaand P. nemamethecum, which have bulging pyre-noids. Because a compound root associated withR1 is found both in coccolithophorids and in sev-eral branches of clade B1, it seems that this may bea primitive feature that has been lost many times indifferent lineages.Members of the genus Platychrysis fall within
the clade of Prymnesium species in our 18S, 28S
Taxonomic revision of the Prymnesiales 219
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
and concatenated rDNA trees. Platychrysis isknown from morphological and ultrastructuralstudies to have a close affinity to Prymnesium,and the two genera have been shown to have clo-sely related rbcL DNA sequences (Inouye, 1997).There are at present four described species ofPlatychrysis, Pl. neustophila R.E. Norris,Pl. pienaarii, Pl. pigra and Pl. simplex. Their dom-inant life-cycle phase is non-motile, but they pro-duce flagellate cells that resemble Prymnesiumparvum in cell microanatomy and the sculpturingof scales. Our rDNA phylogenies support the pro-posal that all Platychrysis species should be trans-ferred to the genus Prymnesium. Non-motile cellsof Hyalolithus neolepis show several of the featurescharacterizing members of clade B1-5: spherical tosubspherical cell shape, haptonema shorter thanthe flagella, a compound flagellar root and cellscovered by organic plate scales. The non-motilecells also have siliceous scales. Non-motile cystsof Prymnesium parvum similarly have a siliceouscell covering, here composed of layers of organicscales with siliceous material deposited on the
outermost scales (Pienaar, 1980; Green et al.,1982). Electron microscopical X-ray microanalysisof scales of P. polylepis indicates that they containsilicate (Jahn Throndsen and Wenche Eikrem,unpublished data), suggesting that silicified scalesare not unique to H. neolepis within thePrymnesiales. Motile cells covered only by organicscales may also occasionally occur in cultures of H.neolepis. This species falls within clade B1-5 in our18S and 28S rDNA trees and a tree using rbcL(Yoshida et al., 2006). In order to avoid having anumber of paraphyletic genera in clade B1-5, wepropose the transfer of all species within this cladeto the genus Prymnesium.Among other Chrysochromulina species that
have not yet been sequenced, we expect allnon-saddle-shaped species (e.g. C. bergenensisB. Leadbeater, C. birgeri G. Hallfors & Niemi,C. brachycylindra G. Hallfors & H.A. Thomsen,C. fragilis B. Leadbeater, C. mantoniaeB. Leadbeater, C. mactra Manton and C. pring-sheimii Parke & Manton) to belong to clade B1.The distinction within this clade between
Figs 7–11. Drawings of selected species of Prymnesiales. 7. Chrysochromulina strobilus. 8. Chrysocampanula spinifera. 9.
Prymnesium, Haptolina and Pseudohaptolina is,however, less clear. Thus, we will not transfer addi-tional Chrysochromulina species until further ultra-structural information and/or molecular data areavailable. Similarly, Corymbellus aureus has not yetbeen characterized genetically, but is expected tofall in clade B1 based on its morphology, cellsbeing oblong and irregular in shape and the hap-tonema being shorter than the flagella. Again, it isunclear whether Corymbellus should be transferredinto one of Prymnesium, Haptolina or Pseudo-haptolina, or whether this genus forms a distinctlineage within clade B1.
Conclusion
Reconstructed phylogenies based on nuclear ribo-somal SSU (18S), the plastid ribosomal SSU (16S),and a combination of the 18S, 28S and 16S rDNAwere congruent and resulted in very similar majorclades. Our data are in close concordance with thetrees for the plastid-encoded rbcL gene includingeight or nine Prymnesiales taxa presented byFujiwara et al. (2001) and Yoshida et al. (2006),as well as with previous studies based on 18SrDNA datasets (e.g. Edvardsen et al., 2000; Saezet al., 2004). Chrysochromulina species are clearlydistributed in two clades. In clade B1, someChrysochromulina species cluster with membersof Imantonia, Hyalolithus, Platychrysis and Prymn-esium, whereas clade B2 contains only Chrysochro-mulina species, including the type species C. parva.The ultrastructural relationships between and
within genera in the Prymnesiales tend to berather complicated, as can be seen in Table 2where we summarize available morphological andultrastructural data for the species included in thisstudy. Characters showing shared derived stateswithin taxonomic units and thus apparently valu-able systematic indicators within the Prymnesiales
are: cell shape, nature of the haptonema, flagellarapparatus including the flagellar roots, and tosome extent scale type. Scale morphology appearsto be useful for species identification and in somecases may indicate phylogenetic relatedness.However, scale morphology may vary betweenlife-cycle stages within a species, and some fea-tures, such as spiny scales, may have evolvedmore than once (e.g. in C. parkeae, Haptolina eri-cina and Chrysocampanula spinifera). There is over-lap in some of these features between variousgenetic sub-groups, making it difficult to dividethis order based solely on morphology. Anumber of features appear to have independentlybeen reduced or lost, or have arisen by convergentevolution, obscuring phylogenetic relationships.Molecular data for different genes are consistent
and strengthen inferences from certain morpholog-ical and ultrastructural features, making it possibleto revise the systematics of this widely distributedgroup of haptophyte algae. Table 3 lists nucleotidesites with consistent sequence differences in the 18SrDNA among the genera of Prymnesiales, wherethis information is available, and between cladesB1 and B2. Clade B1 (excluding B1-1 and B1-6)is recognized here as Prymnesiaceae, and B2 asChrysochromulinaceae. Chrysocampanula spinifera(B1-1) and Chrysochromulina parkeae (B1-6) areconsidered as incertae sedis for the time being. Inaddition to the specific sites in the 18S rDNA(Table 3), we were able to find two probe regionsin this gene that are specific for these families,which can be used as synapomorphic characters.At positions 229–247 (related to sequenceAJ004866 of Prymnesium polylepis) in 18S rDNA,the sequence TGCCGGTTGCGTGCTGAGT isspecific for all members of Prymnesiaceae. At posi-tions 168–185 (reference sequence as above), thesequence TACATGCAGGAAGACCCG is speci-fic for all the taxa sequenced so far in theChrysochromulinaceae.
Table 3. Sites with sequence differences in the 18S ribosomal DNA among the genera of Prymnesiales, where data are
available. The site position is related to sequence AJ004866 of Prymnesium (Chrysochromulina) polylepis.
Nucleotide site
position
741
761
1813
1983
4912
6181
6442
6501
6782
7042
10902
13391
13401
Genus
Chrysochromulina – A A T C G G/A A C/G T A G C
Chrysocampanula T G T A T A G G G T A A T
Imantonia T G G C C A G G G G A A T
Pseudohaptolina T G G C C A G G C T T A T
Haptolina T G T A C A C G A T A A T
Prymnesium T G T/C A/G T A G G G/C G T A T
1Nucleotide site with sequence difference between Prymnesiaceae (clade B1) and Chrysochromulinaceae (clade B2).2Nucleotide site with sequence difference between Haptolina (clade B1-4) and Prymnesium (clade B1-5).3Nucleotide site with sequence difference between ImantoniaþPseudohaptolina (clades B1-2 and B1-3) and the rest.
Taxonomic revision of the Prymnesiales 221
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
The ultrastructure of a number of previously cul-tured species within the Prymnesiales has not beenthoroughly investigated and many more describedspecies have not yet been brought into culture. Inaddition, prymnesialean diversity newly discoveredin environmental DNA samples (Moon-Van derStaay et al., 2000; Liu et al., 2009) awaits morpho-logical study and description. Future ultrastruc-tural and genetic studies may well lead to furtherrevisions of the taxonomy proposed here.Nevertheless, in light of the clear deficiencies inprevious taxonomic schemes, there is an immediateneed for a revised taxonomy that more accuratelyreflects phylogenetic relationships and provides amore informative framework for interpreting theever-increasing amount of metagenomic data (e.g.Cuvelier et al., 2010). Here we propose a conserva-tive revision of the taxonomy of the Prymnesialesthat is in accordance with available molecular evi-dence and supported by morphological data.Figure 12 shows typical morphological characterstates for the genera of Prymnesiales included inthis study (all genera described to present exceptfor Corymbellus).
Taxonomic recommendations
We propose to make the following taxonomicchanges to the taxa falling in clades B1 and
B2: The order Prymnesiales is to be divided intotwo families. The family Prymnesiaceae is emendedand the Chrysochromulinaceae fam. nov. is erected.The latter family should be distinguished from afamily Chrysochromulinidae proposed by Lackeyin 1939 to accommodate his new speciesChrysochromulina parva and other species withthree appendages, for which no description ordiagnosis was provided. The previously describedfamily Prymnesiaceae Conrad (1926) with the samedistinguishing feature made Lackey’s new familysuperfluous. The genera Haptolina andPseudohaptolina are erected, Pseudohaptolina arc-tica is described, and selected members ofChrysochromulina, and all members ofPlatychrysis and Hyalolithus are transferred toPrymnesium. The previous reinstatement of thegenus Chrysocampanula is confirmed. Taxonomicreassignment of Chrysochromulina parkeae will bepublished by others elsewhere.
Family Prymnesiaceae W. Conrad ex O.C. Schmidt
emend. Edvardsen, Eikrem & Medlin
DESCRIPTION: Cells oblong and asymmetrical tospherical and symmetrical, not saddle-shaped,with two smooth, equal to unequal, homodynamicor heterodynamic flagella. Haptonema predomi-nantly shorter than or equal in length to flagella,
Fig. 12. Typical morphological features in the genera of Prymnesiales included in this study. Illustrations of cells are modified
from Throndsen et al. (2007).
B. Edvardsen et al. 222
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
and predominantly non-coiling. Flagella and hap-tonema inserted apically or slightly subapically.Cells covered by simple to elaborate scales.Flagellar apparatus with a compound R1 flagellarroot with a bundle of microtubules, or simple witha sheet of410 microtubules. Cells with two yellow-brown chloroplasts with immersed or bulging pyre-noids. Nucleotide sequences for the SSU and LSUrDNA distinct.TYPE GENUS: Prymnesium Massart, 1920. Generaincluded in the family: Chrysocampanula,Corymbellus, Haptolina, Imantonia, Prymnesium,Pseudohaptolina.
Genus Prymnesium Massart emend. Edvardsen,
Eikrem & Probert
DESCRIPTION: Cells predominantly elongated andirregular, often with an apical shoulder with twoequal or subequal, homodynamic or heterody-namic flagella. Haptonema predominantlyshorter or equal in length to flagella, and usuallyunable to coil. Flagella and haptonema insertedapically or slightly subapically. The flagellarapparatus has a compound R1 flagellar rootwith a bundle of microtubules or a sheet ofmany (420) microtubules. The organic bodyscales are simple to elaborate plate scales andmay have spines or other protuberances. Scalesand cysts may be silicified. Life cycles with alter-nating haploid and diploid phases occur. Severalspecies are toxic.SYNONYMS: Chrysochromulina pro parte excl. typus,Platychrysis, Hyalolithus.TYPE SPECIES: Prymnesium saltans Massart, 1920.
NEW COMBINATIONS:Prymnesium neolepis (M. Yoshida, M.-H. Noel, T.Nakayama, T. Naganuma & I. Inouye) Edvardsen,Eikrem & Probert, comb. nov.BASIONYM: Hyalolithus neolepis M. Yoshida, M.-H.Noel, T. Nakayama, T. Naganuma & I. Inouye.Yoshida et al. (2006). Protist, 157: 214, fig. 2A.
BASIONYM: Chrysochromulina kappa Parke &Manton. Parke et al. (1955). J. Mar. Biol. Assoc.UK, 34: 583, figs 1–19, ‘Type’ culture PCC K.
Prymnesium chiton (Parke & Manton) Edvardsen,Eikrem & Probert, comb. nov.BASIONYM: Chrysochromulina chiton Parke &Manton. Parke et al. (1958). J. Mar. Biol. Assoc.UK, 37: 225, figs 1–37, ‘Type’ culture PCC 146.
Prymnesium minus (Parke & Manton) Edvardsen,Eikrem & Probert, comb. nov.BASIONYM: Chrysochromulina minor Parke &Manton. Parke et al. (1955). J. Mar. Biol. Assoc.UK, 34: 594, figs 36–64, ‘Type’ culture PCC 52.
Haptolina Edvardsen & Eikrem, gen. nov.DESCRIPTIO: Cellulae symmetricae, sphaericae seuovales vel conicae. Haptonema paululo brevius,aequale aut longius quam flagella, tortum autnon tortum. Haptonema cum flagellis duobusapicaliter inserta. Cellulae squamis spineismagnis interdum ornatae. Cellulae chloroplastispraeditae. Apparatus flagellaris radice flagellarisimplici RI interdum instructus. Sequentiaenucleotidis genorum SSU et LSU rRNAdistinctae.DESCRIPTION: Cells symmetrical, spherical to ovalor conical, with two equal, homodynamic flagella.Haptonema slightly shorter, equal or longer thanthe flagella, and can be coiling or non-coiling.Haptonema and flagella inserted apically. Cellsmay have large organic spine scales. The flagellarapparatus may have a simple R1 flagellar root.Nucleotide sequences for the SSU and LSUrDNA distinct.
Taxonomic revision of the Prymnesiales 223
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
SYNONYM: Chrysochromulina pro parte excl.typus.TYPE SPECIES:Haptolina brevifila (Parke & Manton)Edvardsen & Eikrem, comb. nov.
NEW COMBINATIONS:Haptolina brevifila (Parke &Manton) Edvardsen &Eikrem, comb. nov.BASIONYM: Chrysochromulina brevifilum Parke &Manton. Parke et al. (1955). J. Mar. Biol. Assoc.UK, 34: 601, figs 65–81.
Haptolina ericina (Parke & Manton) Edvardsen &Eikrem, comb. nov.BASIONYM: Chrysochromulina ericina Parke &Manton. Parke et al. (1956). J. Mar. Biol. Assoc.UK, 35: 389, figs 1–19.
Pseudohaptolina Edvardsen & Eikrem, gen. nov.DESCRIPTIO: Cellulae rotundae vel oblongae flagel-lis duobus haptonemateque instructae. Haptonemacum flagellis duobus apicaliter insertum.Haptonema similis quam flagella et haptonemanon tortum. Periplastus squamis organicis planiscostis radiantibus in faciebus et distali et proximalitectus. Cellulae chloroplastis praeditae. Sequentiaenucleotidis genorum SSU et LSU rRNA distinctae.DESCRIPTION: Cells round to oblong with twoequal, homodynamic flagella. Haptonema similarin length to flagella and non-coiling. Haptonemaand flagella inserted apically. Periplast covered byorganic plate scales with radiating ribs on bothdistal and proximal faces. Cells with two chloro-plasts. Nucleotide sequences for the SSU and LSUrDNA distinct.TYPE SPECIES: Pseudohaptolina arctica Eikrem &Edvardsen
Pseudohaptolina arctica Edvardsen & Eikrem,
sp. nov.DESCRIPTIO: Cellulae rotundae vel oblongae(8–14� 6–10 mm), saepe dorsoventraliter compres-sae chloroplastis aureo-brunneis duobus praeditae.
Haptonema (12–20mm) cum flagellis duobus(14–24 mm longis) apicaliter inserta. Tegumentumsquameum generibus squamarum rotundarum velovalium duobus compositum. Stratum squamaruminterius (0.7–0.9� 0.8–1.0 mm) margine imbricato,exterius (0.5–0.8� 0.6–0.9 mm) margine erecto.Squamae omnes costis radiantibus (c. 75–85) infaciebus et distali et proximali.DESCRIPTION: Cells round to oblong (8–14�6–10 mm), often dorso-ventrally compressed withtwo golden brown chloroplasts. Haptonema(12–20 mm) and two flagella (14–24mm long)inserted apically. Scaly covering composed of tworound to oval scale types. Inner layer scales(0.7–0.9�0.8–1.0 mm) with imbricated rim. Outerlayer scales (0.5–0.8� 0.6–0.9 mm) with uprightrim. All scales with radiating ribs (c. 75–85) onboth distal and proximal faces.HOLOTYPE: Eikrem & Edvardsen (1999).Phycologia 38: 153, figs 17, 18 (asChrysochromulina sp. 4), CCMP 1204.TYPE LOCALITY: Arctic waters, 76�250N 82 �550WHABITAT: MarineETYMOLOGY: It was isolated by R. Selvin in 1989from Arctic waters.
Family Chrysochromulinaceae Edvardsen, Eikrem
& Medlin, fam. nov.DESCRIPTIO: Cellulae maxime ephippoideae flagellisduobus levibus, aequalibus, homodynamicisinstructae. Haptonema longum ad longissimumatque tortum. Flagella et haptonema ventraliteret subapicaliter inserta. Cellulae squamis organicisparvis tectae, structuris costarum vel fibrillarumradiantium et concentricarum. Apparatus flagel-laris radice flagellari R1 simplici microtubulos(510) saepissime paucos continenti fasciculomicrotubulorum carenti. Cellulae chloroplastisflavo-brunneis duobus utroque pyrenoide immersapraeditae. Sequentiae nucleotidis genorum SSU etLSU rDNA distinctae.DESCRIPTION: Cells predominantly saddle-shaped,with two smooth, equal, homodynamicflagella. Haptonema long to very long and coil-ing. Flagella and haptonema inserted ventrallyand subapically. Cells covered by small organicscales with patterns of radiating and concentricribs or fibrils. Flagellar apparatus with a simpleR1 flagellar root that predominantly containsfew microtubules (510) and lacks a bundle ofmicrotubules. Cells with two yellow-brown chlo-roplasts each with an immersed pyrenoid.Nucleotide sequences for the SSU and LSUrDNA distinct.TYPE GENUS: Chrysochromulina Lackey 1939.The Chrysochromulinaceae contains one genus,Chrysochromulina.
B. Edvardsen et al. 224
Dow
nloa
ded
by [B
ente
Edv
ards
en] a
t 09:
52 2
7 Ju
ly 2
011
Chrysochromulina Lackey emend. Eikrem &
Edvardsen
DESCRIPTION: Cells predominantly saddle-shaped,with two equal, homodynamic flagella.Haptonema long to very long and coiling.Flagella and haptonema inserted ventrally andsubapically. Cells covered by small organic scaleswith patterns of radiating and concentric ribs orfibrils. Some of the scale types bear short spinesor are cup-shaped. Cells with two chloroplastswith immersed pyrenoids traversed by tubes or thy-lakoids. Flagellar apparatus with a simple R1 fla-gellar root that predominantly contains fewmicrotubules. Nucleotide sequences for the SSUand LSU rDNA distinct.TYPE SPECIES: Chrysochromulina parva Lackey,1939. Lackey (1939), Lloydia 2: 128–143.
Acknowledgements
The authors wish to thank Dr W. Kooistra,S. Wrieden, C. Harms, Ms U. Wellbrock, S. Eikvarand S. Brubak for technical assistance with thesequences and S. Brubak and L. Broch for technicalassistance with the cultures. Drs C. Billard, J. Fresnel,J. Green and R. A. Andersen kindly provided cul-tures. We thank B. Tosterud for the Latin transla-tions, Dr S. Ota and R. Orr for expert advice onphylogenetic analyses and Drs W. Kooistra,Ø. Moestrup, E. Paasche and M. Kawachi for valu-able comments on a previous version of the manu-script. This research was funded in part by theNorwegian Research Council to BE and WE(HAPTODIV, 190307/S40), by the BMBF(03F0161) and the EU CODENET to LKM, andby the EU I3 project ASSEMBLE (227799) to IP.
Supplementary material
The following supplementary material is availablefor this article, accessible via the Supplementarycontent tab on the article’s online page at http://dx.doi.org/10.1080/09670262.2011.594095.
SSU FIN.txtLSUFIN.txt16SFIN.txt3genesFIN.txt
References
BEECH, P.L. & WETHERBEE, R. (1988). Observations on the flagellar
apparatus and peripheral endoplasmic reticulum of the cocco-