ROBERT A. ANDERSEN - American Journal of Botany

15 Pages·2004·0.27 MB·English

AmericanJournalofBotany91(10):1508–1522. 2004.
BIOLOGY AND SYSTEMATICS OF HETEROKONT AND
HAPTOPHYTE ALGAE1
ROBERT A. ANDERSEN
BigelowLaboratoryforOceanSciences,P.O.Box475,WestBoothbayHarbor,Maine04575USA
Inthispaper,Ireviewwhatiscurrentlyknownofphylogeneticrelationshipsofheterokontandhaptophytealgae.Heterokontalgae
areamonophyleticgroupthatisclassifiedinto17classesandrepresentsadiversegroupofmarine,freshwater,andterrestrialalgae.
Classesaredistinguishedbymorphology,chloroplastpigments,ultrastructuralfeatures,andgenesequencedata.Electronmicroscopy
and molecular biology have contributed significantly to our understanding of their evolutionary relationships, but even today class
relationshipsarepoorlyunderstood.Haptophytealgaeareasecondmonophyleticgroupthatconsistsoftwoclassesofpredominately
marinephytoplankton.Theclosestrelativesofthehaptophytesarecurrentlyunknown,butrecentevidenceindicatestheymaybepart
ofalargeassemblage(chromalveolates)thatincludesheterokontalgaeandotherstramenopiles,alveolates,andcryptophytes.Heter-
okontandhaptophytealgaeareimportantprimaryproducersinaquatichabitats,andtheyareprobablytheprimarycarbonsourcefor
petroleumproducts(crudeoil,naturalgas).
Keywords: chromalveolate;chromist;chromophyte;flagella;phylogeny;stramenopile;treeoflife.
Heterokontalgaeareamonophyleticgroupthatincludesall (Phaeophyceae) by Linnaeus (1753), and shortly thereafter,
photosynthetic organisms with tripartite tubular hairs on the microscopicchrysophytes(currently5Oikomonas,Anthophy-
mature flagellum (discussed later; also see Wetherbee et al., sa) were described by Mu¨ller (1773, 1786). The history of
1988, fordefinitionsofmatureandimmatureflagella),aswell heterokont algae was recently discussed in detail (Andersen,
as some nonphotosynthetic relatives and some that have sec- 2004),andfourdistinctperiodswereidentified.Thediscovery
ondarily reduced or lost tripartite hairs. Brown seaweeds, di- period (1753–1882) is that era in which brown algae were
atoms, and chrysophytes are commonly known members of described as plants, and microalgae were described as infu-
the group. Haptophyte algae are a monophyletic group that soriaandtreatedasanimals.Perhapsthemostsignificantpub-
includes all photosynthetic organisms with a haptonema, as lication of the era was the two-part publication of Ehrenberg
well as some nonphotosynthetic relatives, and some that have (1838) that contained his light microscopic observations. The
secondarily lost the haptonema. The haptonema, from which first synthesis period (1882–1914) began when brown algae
the group derives its name, is a microtubule-supported ap- and microalgae were first integrated and phylogenetic rela-
pendage that lies between two approximately equal flagella tionships were discussed (Rostafinski, 1882; Correns, 1892;
(forreview,InouyeandKawachi,1994).Thecoccolithophores Klebs, 1893a, b; Lemmermann, 1899; Blackman, 1900), but
and genera such as Pavlova and Isochrysis are commonly the period endedwhenthesetwogroupswereonceagainsep-
known members of the group. Representatives of heterokont arated (Pascher, 1914). The floristic period (1914–1950) was
algae and haptophytes are shown in Figs. 1–24. Until 1992, dominated by the description of many species. There was a
haptophytes were included or closely aligned with heterokont nearly complete absence of evolutionary discussion, for the
algae, but a nuclear small subunit ribosomal RNA (SSU primaryreasonthatthelightmicroscopewasunabletoresolve
rRNA) analysis indicated they are distantly related (Bhatta- characters for determining relationships (Fritsch, 1935). The
charya et al., 1992). Recent molecular studies, based onother second synthesis period (1950–2002) began with and was
genes, have now indicated that heterokont and haptophyte al- dominated by evolutionary and phylogenetic relationships
gae may be more closely aligned than the SSU rRNA data (e.g., Chadefaud, 1950; Bourrelly, 1957; Taylor, 1976; Leipe
indicated (Yoon et al., 2000a, b; Harper and Keeling, 2003; etal.,1996;DaugbjergandAndersen,1997a,b).Transmission
Ryall et al., 2003). electron microscopy provided a wealth of new and phyloge-
netically informative data (e.g., Dodge, 1973; Hibberd, 1976;
Historical perspective—Brown seaweeds were referred to
Taylor, 1976; Andersen, 1987), and biochemical studies were
inearlyChinese(ca.3000BC),Greek(e.g.,Theophrastos,ca.
also initiated (e.g., Strain, 1951; Quillet, 1955; Archibald et
300 BC), and Japanese(ca.500AD)writings,andknowledge
al., 1963; Ragan and Chapman, 1978; Smestad-Paulsen and
of brown seaweeds likely predated recorded history. In early
Myklestad, 1978; Bjørnland andLiaaen-Jensen,1989;Jeffrey,
human history, brown seaweeds were usedforhumanandan-
1989).Cladisticanalysisbroughtnewwaysforanalyzingevo-
imal food, medicinal purposes, and dyes. Most other hetero-
lutionary relationships (e.g., Hibberd, 1979; Lipscomb, 1989;
kont algae are microscopic, although mats of macroscopic
Andersen, 1991; Williams, 1991; Sorhannus, 2001), and mo-
Vaucheria (Xanthophyceae) may have been known but not
lecular systematics added powerful new data sets (e.g., Gun-
recorded in historical works.
dersen et al., 1987; Leipe et al., 1994, 1996; Guillou et al.,
ThefirstmodernscientificreportisthedescriptionofFucus
1999b; Moriyaet al.,2002; GoertzenandTheriot,2003).Dis-
coveries led to descriptions of many new taxa, including sev-
1Manuscriptreceived31December2003;revisionaccepted22June2004. eral classes: Eustigmatophyceae(HibberdandLeedale,1970),
IthankDavidPattersonandHiroshiKawaiforprovidingcolorphotographs
Dictyochophyceae (Silva, 1980), Synurophyceae (Andersen,
ofalgaeandStacyEdgarforassistancewithphylogeneticanalysis.Supported
1987), Coscinodiscophyceae and Fragilariophyceae(Roundet
byNSFgrantsDEB-0206590andDEB-0212138.
E-mail:[email protected]. al., 1990), Chrysomerophyceae (Cavalier-Smith et al., 1995),
1508
October 2004] ANDERSEN—HETEROKONT AND HAPTOPHYTE ALGAE 1509
Bolidophyceae (Guillouetal.,1999a),Pelagophyceae(Ander- Ecology—Heterokontalgaearefoundinalmostallenviron-
sen et al., 1993), Phaeothamniophyceae (Bailey et al., 1998), ments where life exists, but the occurrence varies widely
Pinguiophyceae (Kawachi et al., 2002b), and Schizocladi- among the classes. The ecological literature is extensive and
ophyceae (Kawai et al., 2003). The sequencing of the Thal- impossible to summarize here; the references listed later are
assiosira pseudonana genome, initiated in 2002, was thought goodsourcesforadditionalinformation.Bolidophyceae,Chry-
to be the start of a new period, but it is too early to define somophyceae,Pelagophyceae,Pinguiophyceae,andSchizocla-
this period. dophyceaeareonlyknownfrommarineenvironments(Billard,
The first record of haptophyte algae might begin with Eh- 1984; Guillouetal., 1999a;AndersenandPreisig,2002b;Ka-
renberg (1836), who discovered that chalk was composed of wachi et al., 2002b; Kawai et al., 2003). Phaeophyceae are
tiny crystallites that he considered to be formed by precipita- almost exclusively marine organisms,butfivefreshwatergen-
tion rather than biological activity (see Green and Jordan, era are known (Bold and Wynne, 1985). Synurophyceae are
1994;Siesser,1994).Inthemid1800s,aseriesofarticleswere probablyrestrictedtofreshwater,althoughacoupleofdubious
concerned withthebiologicaloriginofcoccolithsandcoccos- marineoccurrenceshavebeenreported(AndersenandPreisig,
pheres(Huxley,1858;Wallich,1860,1861;Sorby,1861;Cart- 2002a). Chrysophyceae, Phaeothamniophyceae, and Xantho-
er, 1871; Wyville-Thomson, 1874), and the matter was re- phyceae are predominately freshwater organisms, although a
solved in 1898 when Murray and Blackman described and substantial number of xanthophytes are terrestrial (Ettl, 1978;
Reith, 1980; Starmach, 1985; Kristiansen and Preisig, 2001;
illustrated a dividing cell inside the coccosphere (see Green
Hibberd, 1990b; Ettl and Gaertner, 1995; Bailey et al., 1998;
and Jordan, 1994; Siesser, 1994). The first description of the
Preisig and Andersen, 2002). Dictyochophyceaeoccurinboth
haptonemawasbyScherffel(1901)whenhedescribedPhaeo-
marineandfreshwaterhabitats(Moestrup,1995;Moestrupand
cystis,butheconsideredthehaptonematobeathirdflagellum.
O’Kelly, 2002), and Eustigmatophyceae occur in freshwater,
Pascher (1910, 1913, 1914) placed golden microalgae with
marine,andterrestrialhabitats(Hibberd,1990a).Raphidophy-
two equal flagella into order Isochrysidales, classChrysophy-
ceae are sharply divided into two groups, marine genera with
ceae,andthisincludednotonlyorganismswerecognizetoday
fucoxanthin–violaxanthin type pigments, and freshwater gen-
as haptophytes but also some of Synurophyceae and Chryso-
era with heteroxanthin–diatoxanthin type pigments (Table 2;
phyceae.Additionaltaxaweredescribedintheyearsfollowing
Heywood, 1990; Potter et al., 1997; Heywood and Leedale,
Pascher’s classification (e.g., Prymnesium,Chrysochromulina;
2002). Finally, diatoms are found in allcommonhabitatssup-
Carter, 1937; Lacky, 1939), and with the advent of electron
porting life (Round et al., 1990). Regarding haptophytes,they
microscopy,manyadditionalspeciesweredescribed(e.g.,Par-
are predominately marine, but several freshwater species are
keetal.,1955;MantonandLeedale,1969;MantonandLead-
well known (Green and Leadbeater, 1994).
beater, 1974). Electron microscopy also demonstrated the
Diatoms are widely used as indicator species in paleoeco-
uniquestructureofthehaptonema(Parkeetal.,1955),unusual
logicalstudies(forreview,seeStoermerandSmol,1999).Sil-
featuresoftheGolgiapparatus(Manton,1967),andultrastruc-
ica-scaled algae are also good indicator species (e.g., Siver,
tural differences between haploid and diploid phases of the
1991;Smol,1995).Heterokontandhaptophyteclassescontain
life cycle (e.g., Parke and Adams, 1960). These differences
toxic or harmful species. A number of diatoms areharmfulto
led Christensen (1962) to propose a separateclass,Haptophy-
marine life, and domoic acid from Pseudo-nitzschia, concen-
ceae, which he made approximately equal to Chrysophyceae,
trated in shellfish, has killed humans (see Fryxell and Hasle,
Xanthophyceae, Phaeophyceae, etc. Hibberd (1976) provided
2003 for review). Aureococcus and Aureoumbra (Pelagophy-
additional support for the separation of Haptophyceae, Cava-
ceae) form coastal blooms that are harmful to marine inver-
lier-Smith (1986, 1989) divided Haptophyta into two classes, tebrates (Cosper et al., 1989; Buskey et al., 1997; Bricelj et
and most recently, Edvardsen et al. (2000) summarized the al.,2001).ChattonellaandHeterosigma(Raphidophyceae)are
classification of Division Haptophyta, including several no- well-known fish killers (Okaichi, 1989; Hallegraeff and Hara,
menclaturalproposalstobringclassificationinaccordwiththe 2003).Also,Chrysochromulina,Prymnesium,andPhaeocystis
IBCN. (Prymnesiophyceae) are known to kill fish or be harmful to
marine life (Moestrup and Thomsen, 2003).
A number of flagellate heterokont and haptophytealgaeare
Currently recognized classes—The taxonomic class is the
mixotrophic,usuallybyphagocytosis,andmanyutilizeorgan-
primary currency for classifying heterokont algae. In large
ic molecules. The ‘‘biflagellate’’ Chrysophyceae, for which
part, this stems from an inadequate understanding of phylo-
Epipyxis is the model system, may all be phagocytotic, and
geneticrelationships.Thus,someworkerslumpallclassesinto
they have a sophisticated capturing mechanism that involves
a single division, Heterokontophyta (e.g., Hoek, 1978; Hoek
microtubules of the flagellar apparatus. Bacteria captured by
et al., 1995), whereas others raise classes to division level
flagella are pressed into a feeding basket near the flagellar
(e.g.,Corliss,1984).Therearecurrently17recognizedclasses, basesat theanteriorend ofthecell(AndersenandWetherbee,
and, except for the three diatom classes, all classes are listed 1992; Wetherbee and Andersen, 1992). Phagocytosis also oc-
in Tables 1–3. Diatoms are currently classified in Coscinod- curs in the haptophytes, in which Chrysochromulina is the
iscophyceae(centricdiatoms),Fragilariophyceae(araphidpen- model organism (Kawachi et al., 1991; Inouye and Kawachi,
nates), and Bacillariophyceae (raphid pennates; Round et al., 1994). The haptonema captures food particles, wraps around
1990). However, diatom classification will change soon be- the cell, and then particlesareengulfedattheposteriorendof
cause the two pennate classes form a monophyletic group, the cell.
whereas centric diatoms form two clades (e.g., Medlin et al.,
1996). Haptophyta are recognized as a division divided into Cell biology—Chloroplasts and their pigments—The chlo-
two classes, Pavlovophyceae and Prymnesiophyceae (Cava- roplaststructuresofallheterokontalgaeandhaptophytesshare
lier-Smith, 1998; Edvardsen et al., 2000). some features (Dodge, 1973). The chloroplast is surrounded
1510 AMERICAN JOURNAL OF BOTANY [Vol. 91
October 2004] ANDERSEN—HETEROKONT AND HAPTOPHYTE ALGAE 1511
bythechloroplastendoplasmicreticulum,andthusfourmem- Many heterokont swimming cells as well as some Pavlo-
branes separate the stroma from the cytosol. Each chloroplast vophyceae have an eyespot that is located within the chloro-
lamella consists of three adpressed thylakoids. Finally, al- plast or associated with it (e.g., Dodge, 1973; Green, 1980).
though not strictly a chloroplast feature, the photosynthetic Eyespots are part of the photoreceptor apparatus (also called
carbohydratestorageproductisab-1,3-linkedglucanofsmall the eyespot apparatus), shielding light so that the other ele-
molecular size (20–50 glucose residues), which for osmotic mentscanmorepreciselydeterminethedirectionoflight(Fos-
reasons is stored in a vacuole outside the chloroplast. ter and Smyth, 1980). In a wide variety of heterokont and
Distinguishing features include the presence of a girdle la- haptophyte algae, one flagellum possesses an autofluorescent
mella,whichisasaclikethree-thylakoidstructurethatencloses substance (flavin and pterin-like in brown algae) that plays a
all other (sheet type) lamellae. Most heterokont classes (Eus- role in phototaxis (Mu¨ller et al., 1987; Kawai and Inouye,
tigmatophyceae excepted) have a girdle lamella, but it is ab- 1989;Kawaietal.,1996).Inthetypicalcase(mostheterokont
sent in Haptophyta (Table 1). In most heterokont classes as algae, Pavlovophyceae), the eyespot lies just inside the chlo-
well as haptophytes, the outer membrane of the chloroplast roplast in the area immediately adjacent to the mature flagel-
endoplasmicreticulumiscontinuouswiththeoutermembrane lum. Eustigmatophytes have a large eyespot located outside
ofthenucleus.Theinnerchloroplastendoplasmicreticulumis the chloroplast but adjacent to the mature flagellum; this un-
consideredtobeeithertheremnantplasmalemmaofanancient usual eyespot is the basis of the class name. For a recent re-
endosymbiotic event or derived from the outer nuclear enve- view, see Kawai and Kreimer (2000).
lopeaswell(byanout-foldingmodel).Someheterokontalgae Brown algae produce two types ofswimmingcells,asexual
lack a chloroplast endoplasmic reticulum–nuclear envelope zoospores and male (and sometimes female) gametes. Kawai
continuity,andtheseincludethosediatomswithmultiplechlo- et al. (1990, 1991) showed that swimming cells have photo-
roplasts, raphidophytes and synurophytes. A relationshipwith tacticresponsestophotosyntheticallyactivewavelengths.Iken
symbiotic bacteria occurs in the lumen of the chloroplast en- et al. (2001) described five different swimming patterns for
doplasmicreticulumofthediatomPinnularia(Schmid,2003a, Hincksia by employing computer-assisted motion analysis.
b). The bacteria are blocked from passing down the lumen of The patterns were associated with finding suitable attachment
the endoplasmic reticulum to the nucleus. The bacteria also sites for settlement or with positive or negative reactions to
cause, or at least occupy, invaginations in the plastid, giving certain environmental stimuli.
it an irregular margin. Chloroplastsfunctionprimarilyforphotosynthesis,andhet-
Notallspecieshavechloroplasts.Leucoplasts(unpigmented erokont and haptophyte algae have a wide variety of light-
plastids) are present in some chrysophytes, e.g., Paraphyso- harvestingpigments,manyofwhicharephotosyntheticallyac-
monas and Spumella (Mignot, 1977; Preisig and Hibberd, tive. Characterization of pigments has advanced dramatically
1982a, b, 1983). Recently, Sekiguchi et al. (2002) described in the past 50 years, and new techniques as well asmorecrit-
the presence of leucoplasts in two colorless pedinellids, Pter- ical characterization of molecules have been significant.Nev-
idomonas and Ciliophrys (Dictyochophyceae), and they also ertheless, pigment scientists have not always kept abreast of
amplified and sequenced the rbcL gene fromtheseorganisms. taxonomicchanges,andrelativelyfeworganismsineachclass
This provided clear evidence that the colorless taxa were de- have been critically studied (e.g., Jeffrey and Vesk, 1997). A
rived from photosynthetic ancestors, falsifying an earlier hy- summary of chloroplast pigments, by taxonomic class, is
pothesis that the pigmented forms arose from colorlessances- shown in Table 2, but the reader should keep in mind the
tors via an endosymbiotic event (Cavalier-Smith et al., 1995). limited taxon sampling. All heterokont and haptophyte algae,
A taxonomic reevaluation of pedinellids was subsequently except Eustigmatophyceae, have one or more types of chlo-
published(Sekiguchietal.,2003).Leucoplastsmaybeentirely rophyll c, but variability and diversity probably exceeds that
absentinsomeheterokonts,e.g.,Picophagus(Chrysophyceae; shown. These algae are rich in carotenoids, giving them a
Guillou et al., 1999b), but recent cautions indicate that rem- golden or brown color (Eustigmatophyceae, Xanthophyceae,
nantsofplastidsmayremain(HarperandKeeling,2003).Col- some Raphidophyceae excepted). In addition to other roles
orless diatoms, especially Nitzschia, are known (Lewin and (e.g., ultraviolet light protection, photosynthetic quenching),
Lewin, 1967), but whether or not they have leucoplastsisun- one or more photosynthetically active carotenoids are usually
clear. Finally, Sphaeropsis pascheri Schiller (Chrysophyceae) present (e.g., Alberte and Andersen, 1986; Porra etal.,1997).
was described as having cyanelles (Schiller, 1954); however,
this light microscopic work has not been verified using elec- Cell coverings—Heterokont algae have awiderangeofcell
tron microscopy or molecular techniques. This is apparently coverings. Bolidophytes are naked flagellates (Guillou et al.,
theonlyreportofacyanelle-bearingheterokontalga,andthere 1999a); diatoms have siliceous frustules (Round et al., 1990);
are no reports of cyanelles in haptophytes. chrysomerophytes have cell walls (Billard, 1984); chryso-
‹
Figs.1–24. Haptophyteandheterokontalgae.Figs.1–4.Haptophytealgae.1.Prymnesium(Prymnesiophyceae).Scalebar510mm.2.Chrysochromulina
(Prymnesiophyceae).Scalebar510mm.3.Emiliania(Prymnesiophyceae).Scalebar510mm.4.Phaeocystis(Prymnesiophyceae).Scalebar510mm.5.
Pavlova (Pavlovophyceae). Scalebar 5 5 mm.Figs.6–24.Heterokontalgae. 6. Nannochloropsis(Eustigmatophyceae).Scalebar51mm.7.Pinguiococcus
(Pinguiophyceae).Scalebar55mm.8.Polypodochrysis(Pinguiophyceae).Scalebar55mm.9.Pylaiella(Phaeophyceae).Scalebar510mm.10.Dictyopteris
(Phaeophyceae).Scalebar51cm.11.Tribonema(Xanthophyceae).Scalebar510mm.12.Odontella(Bacillariophyta).Scalebar510mm.13.Phaeoplaca
(Chrysophyceae).Scalebar510mm.14.Chrysamoeba(Chrysophyceae).Arrows5flagellum.Scalebar55mm.15.Lagynion(Chrysophyceae).Scalebar
55mm.16.Epipyxis(Chrysophyceae).Scalebar55mm.17.Phaeothamnion(Phaeothamniophyceae).Scalebar510mm.18.Pelagomonas(Pelagophyceae).
Scale bar 5 1 mm. 19. Aureococcus (Pelagophyceae). Scale bar 5 1 mm. 20. Chattonella (Raphidophyceae). Scale bar 5 10 mm. 21. Pseudopedinella
(Dictyochophyceae). Scale bar 5 5 mm. 22. Halodiscus (Bacillariophyta). Scale bar 5 10 mm. 23. Mallomonas (Synurophyceae). Scale bar 5 10 mm. 24.
Rhizochromulina(Dictyochophyceae).Scalebar510mm.
1512 AMERICAN JOURNAL OF BOTANY [Vol. 91
TABLE 1. Chloroplast features of the heterokont and haptophyte algae. The eyespot of eustigmatophytes is located outside the chloroplast.Gen-
ophoretype referstothe arrangementofplastidDNA(ringofDNA,scatteredgranulesofDNA).
Plastid-nucleus
Girdle membrane Genophore
lamella connection Eyespot type
Heterokontalgae
Bacillariophyta 1 6 2 ring
Bolidophyceae 1 1 2 ring
Chrysomerophyceae 1 1 1 ring
Chrysophyceae 1 1 1 ring
Dictyochophyceae 1 1 2 scattered
Eustigmatophyceae 2 1 1(outside) ring
Pelagophyceae 1 1 2 scattered
Phaeophyceae 1 1 6 ring
Phaeothamniophyceae 1 1 1 ring
Pinguiophyceae 6 1 2 scattered
Raphidophyceae 6 2 2 scattered
Schizocladophyceae 1 1 1 ring
Synurophyceae 1 2 2 ring
Xanthophyceae 1 1 6 ring
Haptophyta
Pavlovophyceae 2 1 1 scattered
Prymnesiophyceae 2 1 2 scattered
Note:15present;25absent;65presentorabsent.
phytes have cell walls, organic loricas, organic or silica scale raphidophytesarenakedcells(Heywood,1990;Heywoodand
cases, gelatinous coverings, and completely naked cells (Star- Leedale, 2002); Schizocladia has cell walls without cellulose
mach, 1985; Kristiansen and Preisig, 2001; Preisig and An- but impregnated with alginates (Kawai et al., 2003); synuro-
dersen, 2002); dictyochophytes have silica skeletons, organic phyteshavebilaterallysymmetricalsilicascalesgluedtogether
scales,ornakedcells(Moestrup,1995;MoestrupandO’Kelly, to form a highly organized scale case (Ludwig et al., 1996);
2002);eustigmatophyteshavecellwalls(Hibberd,1990a);pe- xanthophytes have predominately cell walls, some with H-
lagophytes have cell walls, thecae, gelatinous coverings, and shaped overlapping sections, as well as plasmodial and naked
naked cells (Andersen and Preisig, 2002b); phaeophytes have forms (Hibberd, 1990b).
cellulosic cell walls impregnated with alginates and often in- Although silica frustules of diatoms have long beenstudied
terconnected via plasmodesmata (Bisalputra, 1966; Pueschel for taxonomic purposes (e.g., Hustedt, 1928), newtechnology
and Stein, 1983); phaeothamniophytes have cell walls(Bailey has allowed scientists to investigate the nonsiliceous compo-
et al., 1998); pinguiophytes have mineralized loricas, gelati- nents of the cell wall. Higgens etal. (2003) used atomicforce
nous coverings, or naked cells (Kawachi et al., 2002a, b, c); microscopy to study the topology and properties of the mu-
TABLE2. Chloroplastpigmentationofheterokontand haptophytealgae.
199- 199- Diato-
Chlorophylls Fuc hex but dino Viola Hetero Vauch
Heterokontalgae
Bacillariophyta a, c (c) 1 1 2 1 2 2 2
1,2 3
Bolidophyceae a, c 1 1 2 1 2 2 2
1–3
Chrysomerophyceae a, c 1 2 2 2 1 2 2
1,2
Chrysophyceae a, c 1 1 2 2 1 2 2
1,2
Dictyochophyceae a, c 1 2 2 1 2 2 2
1,2
Eustigmatophyceae a 2 2 2 2 1 2 1
Pelagophyceae a, c 1 (1) 1 1 2 2 2
1,2
Phaeophyceae a, c 1 2 2 2 1 2 2
1,2
Phaeothamniophyceae a, c 1 2 2 1 2 1 2
1,2
Pinguiophyceae a, c 1 2 2 2 1 2 2
1,2
Raphidophyceae-FW a, c 6 2 2 1 2 1 1
1,2
Raphidophyceae-Mar a, c 6 2 2 2 1 2 2
1,2
Schizocladophyceae a, c(type?) 1 ? ? ? ? ? ?
Synurophyceae a, c 1 2 2 2 1 2 2
1
Xanthophyceae a, c 2 2 2 2 1 1 1
1,2
Haptophyta
Pavlovophyceae a, c 1 1 1 1 2 2 2
1–3
Prymnesiophyceae a, c 1 1 1 1 2 2 2
1–3
Notes:Fuc5fucoxanthin;199-hex5199-hexanoyloxyfucoxanthin;199-but5199-butanoyloxyfucoxanthin;diato-dino5pigmentsofthediatox-
anthin; diadinoxanthin cycle; viola5pigmentsoftheviolathanin;antheraxanthin;zeaxanthincycle;hetero5heteroxanthin;vauch5vaucheriox-
anthin;+5present;25absent;65presentorabsent;? 5unknown.
October 2004] ANDERSEN—HETEROKONT AND HAPTOPHYTE ALGAE 1513
TABLE3. Featuresoftheflagellarapparatusofthe heterokontandhaptophytealgae.
# Flagellar Lateral mt Green Parafag- Striated
Tri-H flagella TH beat hairs roots flagellum ellarrod root
Heterokontalgae
Bacillariophyta 1 1,0 0 pulls 2 0 2 ? 2
Bolidophyceae 1 2 0 pulls 2 0 ? 2 2
Chrysomerophyceae 1 2 6› pulls 2 4 ? 2 ?
Chrysophyceae 1 2 4–6 › pulls 1 2–4 1 2 1
Dictyochophyceae 1 1 0–2 fl pulls 2 0 2 1 2
Eustigmatophyceae 1 2,0 6› pulls 2 4 2 2 1
Pelagophyceae 6 2,1,0 0–2 fl pulls 2 0 2 1 2
Phaeophyceae 1 2 0 pulls 2 4 1 2 2
Phaeothamniophyceae 1 2 6› pulls 2 4 1 2 1
Pinguiophyceae 6 2,1,0 0,2fl pulls,none? 2 3,4 6 2 1
Raphidophyceae 1 2 0 pulls 2 ? 2 2 ?
Schizocladophyceae 1 2 6› pulls 2 ? 1 2 ?
Synurophyceae 1 2,1 6–9› pulls 1 2 1 2 1
Xanthophyceae 1 2,0 236› pulls 2 4 1 2 1
Haptophyta
Pavlovophyceae 2 2 0 pulls 2 2 2 2 2
Prymnesiophyceae 2 2 0 pushes,breastpull 2 4 1 2 2
Notes: Tri-H 5 tripartite tubular hairs; TH 5 transitional helix; mt roots 5 microtubular roots. Arrows (› ; fl ) indicate the position (distal,
proximal)ofthe transitionalhelixwithrespecttothe majortransitionalplate.+5present;25absent;65presentorabsent;?5unknown.
cilage layer that coats diatom frustules. They found two dif- Pasini and Alberte, 2001). Important cell wall features that
ferenttypesofmucilagenanostructureontwobenthicspecies, distinguish Phaeophyceae and Schizocladophyceae are the
andonathirdspeciestheydemonstratedthecompleteabsence presenceofcelluloseandplasmodesmatainthewallsofbrown
of a mucilage layer. They also measuredtheadhesive-binding algae but the absence of both in Schizocladia (Kawai et al.,
properties and elasticity properties of the polymer chains that 2003). Like brown algae, however, Schizocladia contains al-
makeupthemucilage.Silicificationindiatomsoccursinsilica ginates that impregnate its (unknown) cell wall fibers.
deposition vesicles that are shaped into the form of the final Haptophytes also have a variety of cell coverings. Benthic
valve or girdle band (Simpson and Volcani, 1981; Schmid, stages of some have cell walls, coccolithophorids have calci-
2003a, b). Silica scales and siliceous cysts of synurophytes fied scales (usually mineralized onto organic scales) that are
and chrysophytes as well as the siliceous skeleton of Dictyo- termed coccoliths, some have only organic scales, a silica-
cha (Dictyochophyceae) are also formed in silica deposition scaled prymnesiophyte was recently reported, some are sur-
vesicles (Schnepf and Deichgra¨ber, 1969; Mignot and Brug- rounded by gelatinous material, and others are naked (see
erolle, 1982; Beech et al., 1990; Moestrup and Thomsen, Green and Leadbeater, 1994; Winter and Siesser, 1994).
1990; Preisig, 1994). Despite the unusual nature of siliceous
wall coverings as well as the similar silicification processes Flagellar apparatus—The typical swimming cell of heter-
found among diatoms, chrysophytes, Dictyocha, and synuro- okont algae has two flagella, a long immatureflagellumanda
phytes, only Chrysophyceae and Synurophyceae appear to be short mature flagellum (Table 3). It is the marked and nearly
closely related (see phylogeny section). consistent nature of these two flagella that defines the term
Parmales, a poorly known group of heterokont algae not heterokont. The control of flagellar length in heterokonts is
discussed elsewhere in this paper,aretinymarinephytoplank- unknown, but it may be similar to that for green algae (see
ters that are characterized by relatively large silica platessur- Beech, 2003, for review). An immature flagellumisproduced
rounding the protoplasm (Booth and Marchant, 1987, 1988; de novo during cell division, and the previous immature fla-
Kosman et al., 1993; Bravo-Sierra and Herna´ndez-Becerril, gellum is transformed into a mature flagellum by a process
2003). The silicification process is not known for Parmales, termed flagellar transformation (e.g., Wetherbee et al., 1988).
butpresumablyitinvolvessilicadepositionvesicles.Parmales Thus, each typical cell has a longer immature flagellum bear-
are known only from field samples, and their classification ingtripartitehairsandashortermatureflagellum(seelaterfor
remains an enigma. They have been nominally classified in exceptions).
Chrysophyceae, but the lack of distinctive ultrastructural fea- In heterokont algae, orientation of flagella on biflagellate
tures, apparent absence of flagellate stages, no knowledge of cells varies greatly, from cells with two forward-directed fla-
photosyntheticpigments,andabsenceofgenesequencesmake gella to those with one forward-directed flagellum and one
an informed classification impossible. trailing flagellum. Sometimes, but not always, orientation of
Brown seaweeds (Phaeophyceae) include kelps, the largest basal bodies matches that of flagella. Mismatched direction
and most structurally complex of heterokont algae. Of heter- occurs,forexample,inzoosporesofbrownalgae(basalbodies
okont algae, they most resemble plants with regard to cell at 908, flagella at1808)andflagellatecellsofRaphidophyceae
walls. Adjacent cells are often interconnected via plasmodes- and some Synurophyceae (basal bodies nearly parallel or 08,
mata (Bisalputra, 1966; Pueschel and Stein, 1983), a feature flagella at 1808).
not found in other heterokont algae. Biochemical studiespro- Flagellated vegetative cells of Bolidophyceae, Chrysophy-
vided evidence of intercellular transport, such as movement ceae, and Raphidophyceae as well as most vegetative cellsof
fromtheleafyfrondstothemeristematicregion(e.g.,Cabello- Synurophyceae and Phaeomonas (Pinguiophyceae) have two
1514 AMERICAN JOURNAL OF BOTANY [Vol. 91
typical flagella (e.g., Hibberd, 1976; Andersen, 1989; Hey- menopilesincludesheterokontalgae,o¨omycetes,labyrithulids,
wood, 1990; Guillou et al., 1999b; Honda and Inouye, 2002). thraustochytridsandcertainbiflagellateprotozoa.Thebipartite
Similarly, flagellated zoospores or sperm of Chrysomerophy- hairsofPelagomonasandthehairlessflagellaofGlossomastix
ceae, Eustigmatophyceae, Phaeophyceae, Phaeothamniophy- and Polypodochrysis are presumed to be derived conditions.
ceae, Schizocladophyceae, and Xanthophyceae as well as Flagella(orflagellum)areputatively‘‘anchored’’inthecell
some Pelagophyceae have two typical flagella (e.g., Billard, with various structures that are generally referred to as the
1984; O’Kelly, 1989; Hibberd, 1990a, b; Lobban et al., 1995; flagellar root apparatus. In broad terms, the flagellar root ap-
Andersen et al., 1998b; Kawai et al., 2003). Conversely, the paratus consists of microtubular roots, striated roots, and a
flagellate sperm of the diatoms as well as armored vegetative complex transitional region. Because of considerablevariabil-
cells of Dictyochophyceae and some Mallomonas species ityamongheterokontalgae,itisdifficulttodesignateatypical
(Synurophyceae) have only a single, immature flagellum,i.e., organization (Andersen, 1991).
they lack a mature flagellum although they possess a mature Microtubular roots are found in swimming cells of allclas-
basal body (e.g., Manton and von Stosch, 1966; Beech and ses,exceptdiatoms,Dictyochophyceae,andPelagomonas(Pe-
Wetherbee,1990a,b;MoestrupandThomsen,1990).Ofthese, lagophyceae; Andersen, 1991; Andersen et al., 1993; Moes-
the diatom sperm are noteworthy in that the flagellum axo- trup, 1995; Sekiguchi et al., 2003). These are designated R–
neme has a 9 1 0 microtubular arrangement; in all other het- R (Andersen, 1987). The R root typically consists of two1to
erokonts,theflagellumhasatypical912arrangement(Man- fo4ur microtubules and assoc1iated dense materials. It attaches
ton and von Stosch, 1966; Heath and Darley, 1972). In Pe- tothebasalbodyoftheimmatureflagellum,andwhenviewed
lagomonas (Pelagophyceae), only the immature flagellum is from the cell anterior, forms a clockwise arc around the an-
present, and no remnant of the mature flagellumbasalbodyis terior of the cell. In most groups, the arc consists of approx-
present (Andersen et al., 1993). A paraxonemal rod lies be- imately 180 degrees (Andersen, 1991), but in Synurophyceae,
tweentheaxonemeandimmatureflagellarmembraneofsome R forms a complete loop of 360 degrees (Andersen, 1985,
1
Dictyochophyceae, Pelagomonas (Pelagophyceae), and possi- 1989). In most organisms (Eustigmatophyceae excepted, see
bly diatom sperm (Heath and Darley, 1972; Zimmermann et SantosandLeedale,1991),R nucleatesnumerouscytoskeletal
1
al., 1984; Moestrup and Thomsen, 1990; Andersen et al., microtubules that extend out and putatively form structural
1993; Sekiguchi et al., 2003). Paraxonemal rods are absentin support for the cell (see Andersen, 1991). In some organisms
other heterokont algae, but a similar rod is present in some (e.g., brown algae or phaeothamniophytes), a special set of
dinoflagellates. In Glossomastix (Pinguiophyceae), the single cytoskeletalmicrotubulestermedthebypassingrootlet,extend
flagellum was designated the mature flagellum, with the ac- from the R root past the basal bodies and into the central
1
companying basal body identified as immmature (O’Kelly, regionofthecell(O’Kelly,1989;Andersenetal.,1998b).The
2002). In Polypodochrysis (Pinguiophyceae), a similar situa- R roottypicallyconsistsofonetotwomicrotubulesthatorig-
2
tion was found, but mature and immature structures were not inate on the side opposite the immature basal body (with re-
identified (Kawachi et al., 2002c). In somemembersofChry- spect to the R root) and probably terminates at or near the
1
sophyceae, diatoms, Eustigmatophyceae, Pelagophyceae, arc of the R root (Andersen, 1991). This root is not always
1
Phaeothamniophyceae, and Xanthophyceae, flagellate stages present. The R root consists of approximately five to seven
3
are unknown. microtubules arranged in a trough or flat arrangement, and a
The typical heterokont swimming cell has tripartite tubular layeredstructureistypicallyassociatedwithmicrotubules.The
hairs (5 mastigonemes) arranged in two rows along the im- R rootextendsfromthematurebasalbodyand,whenviewed
3
mature flagellum. The flagellum beat is sinusoidal, the hairs from the cell anterior, curves in a counterclockwise arc (see
reverse the thrust of the flagellum, and therefore the beating Andersen, 1991). The length, curvature, and path for R vary
3
flagellum pulls the cell forward (Sleigh, 1974, 1989). Mem- widely. For example, in the brown algal zoospores of Lami-
bers of Chrysophyceae and Synurophyceae have lateralfibers naria, the R is short (O’Kelly, 1989), whereas in the phago-
3
on the central shaft of the tripartite hair (e.g., Bouck, 1972; trophic chrysophyte Epipyxis, the R forms a long, complex
3
Andersen, 1989), but such lateral hairs are absent in all other looping structure that is involved in the engulfing of bacteria
heterokont algae. It may be worth noting that Hemiselmis (Andersen and Wetherbee, 1992). The R microtubular root
4
(Cryptophyceae) also has short and long lateral filaments on arises along the mature basal body opposite the R root. The
3
itsbipartitehairs(Bouck,1972).InPelagomonas(Pelagophy- R root is short, extending slightly away from but parallel to
4
ceae), hairs are bipartite, lacking the basal portion, but nev- the mature basal body before terminating. Like the R root,
2
ertheless, the hairs reverse thrust and swimming direction is the R root isapparentlyabsentinmanyheterokontflagellates
4
unchanged(Andersenetal.,1993).Therearenotripartitehairs that possess microtubular roots.
on the emergent flagellum (whether designated mature or im- A special striated flagellar root, also termed a rhizoplast,is
mature) of flagellate eggs of Laminaria angustata Kjellman found in swimming cells of Chrysophyceae, Eustigmatophy-
(Phaeophyceae; Motomura and Sakai, 1988) or the zoospores ceae, Phaeothamniophyceae, Pinguiophyceae, Raphidophy-
of Glossomastix and Polypodochrysis (Pinguiophyceae); pin- ceae, Synurophyceae, and Xanthophyceae (e.g., Hibberd,
guiophyte zoospores glide along the substrate in amoeboid 1976, 1990a, b; Heywood, 1990; Andersen, 1991; Andersen
fashion (O’Kelly, 2002; Kawachi et al., 2002c). However, et al., 1998b; Kawachi et al., 2002b). One end of this striated
Phaeomonas (Pinguiophyceae) has typical tripartite tubular root lies along the nuclear envelope, and the other end is typ-
hairs on its immature flagellum (Honda and Inouye, 2002). ically attached to proximal end of the immature basal body.
The terms stramenopiles and stramenochromes have been ap- However, in Synurophyceae, it attaches to both basal bodies
plied to heterokont algae and their relatives (Patterson, 1989; (Andersen, 1985, 1989; Beech and Wetherbee, 1990b). The
Leipe et al., 1996), with both terms referring (strameno 5 nucleusispositionedsomedistancefromthebasalbodies,and
straw)totripartiteflagellarhairsasasynapomorphiccharacter. the striated root is probably contractile.
Stramenochromes is equal to heterokont algae, whereas stra- There has been no report of a rhizoplast-type striated root
October 2004] ANDERSEN—HETEROKONT AND HAPTOPHYTE ALGAE 1515
in Bolidophyceae, diatoms, Dictyochophyceae, Pelagophy- openings at the poles (Vesk et al., 1984). Pelagococcus (Pe-
ceae, Phaeophyceae, or Schizocladophyceae. Some Dictyoch- lagophyceae; Vesk and Jeffrey, 1987), Synura (Synurophy-
ophyceae have astriated bandthatextendsfromtheimmature ceae; Andersen, 1989), and most Phaeophyceae (see Green,
basal body to the nucleus, but because the nucleus is posi- 1989,forreferences)behavesimilarlytoHydrurus.Vaucheria
tioned against the basal bodies, it is unclear if this is a ho- (Xanthophyceae) has an intactnuclear envelopeatmetaphase,
mologous structure (e.g., Koutoulis et al., 1988; Sekiguchi, and spindle microtubules form completely within the nuclear
2003). envelope (Ott and Brown, 1972). Vacuolaria (Raphidophy-
The transitional region of the flagellum,thatareawherethe ceae) is perhaps the most unusual situation, in which the nu-
basal body connects to the flagellum, is also variable among clear envelope of daughter cells forms inside the dispersing
heterokont algae (Preisig, 1989). A major transitional plate is old mother nuclear envelope (Heywood, 1990; Heywood and
found in all heterokont flagella, and in a few instances, asec- Leedale, 2002). Mitosis has not been reported for Bolidophy-
ond transitionalplate occurs. The majorplateislocatedinside ceae, Chrysomerophyceae, Dictyochophyceae, Eustigmato-
the nine pairs of microtubules so that it is distal to the third phyceae, Phaeothamniophyceae, Pinguiophyceae, and Schi-
microtubuleofthebasalbodytripletsandproximaltothecen- zocladophyceae. Among haptophytes, mitosis has been de-
traltwomicrotubulesoftheflagellaraxoneme.Thereisatran- scribedforPavlova(Pavlovophyceae)aswellasforEmiliania,
sitional helix above the major transitional plate in Chryso- Chrysochromulina, Imantonia, Isochrysis, Pleurochrysis, and
merophyceae, Chrysophyceae, Eustigmatophyceae, Phae- Prymnesium (Prymnesiophyceae; Manton, 1964; Stacey and
othamniophyceae, Pinguiophyceae, Schizocladophyceae, and Pienaar,1980;HoriandInouye,1981;HoriandGreen,1985a,
Synurophyceae; a double transitional helix occurs above the b, c; Green and Hori, 1988; Green et al., 1989). The spindle
plate in Xanthophyceae. There is a transitional helix between is U- or V-shaped in PavlovabutisstraightinPrymnesiophy-
major and minor plates in Dictyochophyceae, Pelagophyceae, ceae. In general, the nuclear envelope disperses during pro-
and Pinguiophyceae. There is no report of a transitionalhelix phase but is often replaced with rough ER during metaphase;
of any kind in Bolidophyceae, diatoms, Phaeophyceae, and see Green (1989) and Hori and Green (1994) for further de-
Raphidophyceae. tails.
Haptophyte algae are biflagellate, but they completely lack
tripartite tubular hairs. Pavlovophyceae sometimes have knob Other ultrastructural features—All heterokonts and hapto-
scales on the immature flagellum; these scales appear to re- phytes have mitochondria with tubular cristae (Taylor, 1976;
verse the thrust of the flagellum, thereby causing the cells to StewartandMattox,1980).HeterokontalgaehavetypicalGol-
swim forward. Prymnesiophycae lack even knob scales, and gi bodies, and in most classes (Dictyochophyceae excepted),
when their flagella beat with a sinusoidal wave, the cells are Golgi bodies areanteriortothenucleus,withcis-cisternaead-
pushed backward. However, these organisms can also beat jacentthenuclearenvelope(e.g.,Hibberd,1976).Haptophytes
their flagella using the ‘‘breast stroke’’ action, similar to the typically have Golgi bodies that are anteriorly adjacent the
green alga Chlamydomonas, and with this flagellar beat pat- nucleus, but they are oriented at 908 so that the cis-trans axis
tern, the cell swims forward. lies parallel tothenuclearenvelope(e.g.,Manton,1967).Fur-
The microtubular flagellar roots of haptophytes resemble thermore, cisternae are unusually inflated. Brown algae often
those of heterokont algae. Typically, Prymnesiophyceae have contain numerous vesicles of phenolic-type compounds, and
four microtubular roots that correspond to heterokonts with these structures are referred to as physodes. Mucocysts are
regard to origin and general path through the cell. Pavlovo- common in Raphidophyceae (Heywood, 1990; Heywood and
phyceae differ in that the immature flagellum lacks microtu- Leedale, 2002), and various mucosal vesicles occur in some
bular roots. The unique structure of haptophytes is the hap- members of Chrysomerophyceae (Billard, 1984) and Chryso-
tonema, a microtubule-supported appendage that extends for- phyceae(e.g.,Hibberd,1970;Mignot,1977;Andersen,1982).
ward betweenthetwo flagella.Thefunctionofthehaptonema Haptophytesarecharacterizedbyaperipheralendoplasmicre-
includes the capture of prey particles in mixotrophic and het- ticulum, which lies just beneath the plasmalemma in most ar-
erotrophic species (Kawachi et al., 1991), attachment to sur- eas of the cell (flagellar region excluded; e.g., Hibberd,1976;
faces, and various other poorly documented roles(Inouyeand Beech and Wetherbee, 1988). It has been suggested that the
Kawachi, 1994). A fibrous root extends from the immature peripheral endoplasmic reticulum of haptophytes is homolo-
basal body in Pavlova, but fibrous roots are apparentlyabsent gous to alveoli of ciliates, amphiesmal vesicles of dinoflagel-
in Prymnesiophyceae. The transitional region of haptophytes lates,theinnermembranecomplexofapicomplexans,theperi-
contains one or more transitional plates, but typical hetero- plast of cryptophytes, and possibly mucosal structures of het-
kont-like transitional helices are absent. Pleurochrysis(Prym- erokont algae (Daugbjerg and Andersen, 1997b; Cavalier-
nesiophyceae) has a helix, but its structure is different(Beech Smith, 2002; Andersen, 2004). If these are truly homologous
and Wetherbee, 1988). structures, they would be a synapomorphic character for
chromalveolates.
Mitosis—Mitosis is known only for a few heterokont and
haptophyte algae, and these few examples vary considerably. Phylogenetic relationships—Phylogenetic relationships of
In diatoms (see Green, 1989, for references) and most Chry- heterokontalgaearestilllargelyunresolved.Lightmicroscopy
sophyceae (e.g., Ochromonas, Poterioochromonas, Uroglen- provided few characters that could be used, and the onedom-
opsis; Slankis and Gibbs, 1972; Bouck and Brown, 1973; inating relationship, Pascher’s (1914) division Chrysophyta
Schnepf et al., 1977; Tippit et al., 1980; Andersen, 1989),the (classes Bacillariophyceae sensu lato, Chrysophyceae and
nuclear envelope disperses during prophase. Spindle microtu- Xanthophyceae) was quickly demolished when electron mi-
bules attach to either basal bodies (diatoms) or the striated croscopy reached widespread use. Cladistic analyses were at-
flagellarroots(Chrysophyceae).However,inHydrurus(Chry- tempted (e.g., Hibberd, 1979; Lipscomb, 1989; Andersen,
sophyceae), the nuclear envelope remains largely intact, with 1991; Williams, 1991), but these suffered from a lack of
1516 AMERICAN JOURNAL OF BOTANY [Vol. 91
knowledge of homologous structures. Molecularphylogenetic
analyses have made some progress. An early study showed
that a heterokont alga was related to an o¨omycete fungus
(Gundersenetal.,1987),bringingfurthersupporttoagrowing
consensus that photosynthetic and nonphotosynthetic hetero-
konts formed a clade (e.g., Cavalier-Smith, 1986). Another
early study showed that Xanthophyceae and Phaeophyceae
were closely related, as were Chrysophyceae and Synurophy-
ceae; however, the two clades were unrelated (Ariztia et al.,
1991). These molecular data provided perhaps the final evi-
dence that Pascher’s Chrysophyta was not a natural group. A
total evidence approach, using ultrastructural, biochemical,
and molecular data, showed that Dictyochophyceae and Pela-
gophyceae were closely related to each other but distantly re-
lated to Chrysophyceae in which species of the former two
classes were once classified (Saunders et al., 1995). Further-
more, this study indicated that these classes may berelatedto
diatoms, forming a clade of organisms with reduced flagellar
apparatuses. One subsequent total evidence analysis also pro-
vided support for this idea (Sorhannus, 2001). To date, mo-
lecular phylogenetic analyses including mostorallheterokont
algal classes have been based on either the 18S rRNA or the
rbcL gene. Other genes have been examined, e.g., the fuco-
xanthin/chlorophyll photosystem-I- and -II-binding proteins
(Caron et al., 1996; Green and Durnford, 1996), the alpha-
tubulin gene (Keeling and Doolittle, 1996), large subunit
(LSU) rRNA gene (Van der Auwera and De Wachter, 1996;
Ben Ali et al., 2001), the GAPDH gene (Fast et al., 2001;
Harper and Keeling, 2003), plastid psaA, psbA, 16S rRNA,
rbcL and tufA genes (Medlin et al., 1997; Yoon et al., 2002a,
b), and the type II fatty acid synthetase gene (Ryall et al.,
2003). However, in all cases, taxon sampling was limited,
omitting most heterokont algal classes and often including
only one to three taxa for classes that were studied.
Two recent studies have combined these more extensively
sampled genes (SSU rRNA, rbcL; Sorhannus,2001;Goertzen
and Theriot, 2003), and the Sorhannus study also included
partialLSUrRNA,ultrastructural,andbiochemicaldata.From
thesetwostudies,aswellasmanyotherstudiesthatseparately
examined SSU rRNA and rbcL sequences, a few consensus
relationships can be identified. Three two-class clades, Chry-
sophyceae/Synurophyceae, Dictyochophyceae/Pelagophyceae,
Bolidophyceae/diatoms, are always recovered.However,there
isweaksupport(e.g.,,50%bootstrapvalues)andnoconsen-
sus regarding relationships among these pairs of classes.
Phaeophyceae and Xanthophyceae are closely related, but
whentaxaofChrysomerophyceae,Phaeothamniophyceae,and
Schizocladophyceae are added, thePhaeophyceae/Xanthophy-
ceae relationship is weakened or disrupted (e.g., Bailey et al.,
1998; Kawai et al., 2003). Eustigmatophyceae, Pinguiophy-
ceae, and Raphidophyceae have no clear relationship among Fig.25. Singlemostparsimonioustree(oneofthree)fromamixed(nu-
cleotideandaminoacid)TNT(TreeAnalysisusingNewTechnology,version
themselves or with other heterokont classes(e.g.,Potteretal.,
1.0, by Goloboff, Farris and Nixon, website: http://www.cladistics.org/
1997; Andersen et al., 1998a; Kawachi et al., 2002b). Figure
downloads/webtnt.html) analysis of the concatenated SSU rRNA and rbcL
25illustratesaphylogenetictreeconstructedfromacombined genes. Most SSU rRNA sequences were obtained in an aligned form from
analysisofSSU rRNAandrbcLgenesfromheterokonts,hap- the European Ribosomal RNA Database (website: http://www.psb.ugent.be/
tophytes, alveolates, cryptophytes, and rhodophytes. This tree rRNA/index.html); a few additional taxa (e.g., Pinguiophyceae and Phae-
is poorly resolved when compared to trees from a rbcL gene othamniophyceae)wereaddedandalignedbyeye.TherbcLgeneswerepri-
marily obtained from GenBank; a few Chrysophyceae were from our labo-
onlyanalysis(notshown),butthenonphotosynthetictaxacan-
ratory.TherbcLgenewasconvertedfromnucleotidestoaminoacids.1000
not be included in the rbcL analysis.
bootstrapreplicateswereconductedandthepercentagesupportisshownfor
The two classes of haptophyte algae have a sister relation- allnodeswith.50%support.A5alveolatetaxa,B 5 haptophytetaxa, C
ship in phylogenetic analyses (e.g., Edvardsen et al., 2000), 5allheterokonttaxa,andD5heterokontalgae.Triangleheightis propor-
but the relationship of Haptophyta to other protists is unre- tionaltonumberoftaxa.Solidtrianglesrepresentgroupswithtaxaknownto
solved in SSU rRNA and rbcL phylogenetic analyses. How- possessplastids.
October 2004] ANDERSEN—HETEROKONT AND HAPTOPHYTE ALGAE 1517
ever, recent studies using other genes, albeit with limitedtaxa phytes is necessary for a better understanding of their evolu-
and few classes, are beginning to support a chromalveolate tionary relationships. This task will require substantial work
assemblage. That is, Cryptophyceae, Haptophyta, alveolates because there are many classes of heterokont algae, and the
(dinoflagellates,ciliates,apicomplexans),andheterokontalgae nonalgalheterokontsareequallychallenging.Phylogeneticre-
are perhaps related, but the branching order is still unclear lationships of heterokont and haptophyte algae are fertile
(e.g.,Fastetal.,2001;Yoonetal.,2002a,b;HarperandKeel- groundthathasbeenbarelyscratched,andmuchexcitingwork
ing, 2003; Ryall et al., 2003). The geological time for the remains in this diverse group.
originofthechromalveolateswasplacedat1300millionyears
ago (Yoon et al., 2004). LITERATURE CITED
Nonpigmented heterokonts are close relatives ofheterokont
algae, but no details are provided here. Blackwell and Powell ALBERTE, R. S., AND R. A. ANDERSEN. 1986. Antheraxanthin, a light-har-
(2000) provided an excellent review. Some nonpigmentedfla- vesting carotenoid found in a chromophyte alga. Plant Physiology 80:
583–587.
gellates are described by Moestrup (2002) and Patterson
ANDERSEN,R.A.1982. Alightandelectronmicroscopicalinvestigationof
(2002).
Ochromonas sphaerocystis Matvienko (Chrysophyceae):the statospore,
vegetativecellanditsperipheralvesicles.Phycologica21:390–398.
Unity and diversity—Heterokont and haptophyte algae ANDERSEN, R. A. 1985. The flagellar apparatus of the golden alga Synura
sharethefollowingfeatures:mitochondriawithtubularcristae; uvella:fourabsoluteorientations.Protoplasma128:94–106.
an extraplastidal carbohydrate storage product consisting of ANDERSEN, R. A. 1987. Synurophyceae classis nov., a new class of algae.
short b-1,3-linked glucan chains; a plastid with three adpres- AmericanJournalofBotany74:337–353.
ANDERSEN, R. A. 1989. The Synurophyceae and their relationship to other
sedthylakoidsinternallyandtwoendoplasmicreticulummem-
golden algae. In J. Kristiansen, G. Cronberg, and U. Geissler [eds.],
branes externally; photosynthesis predominating; most organ- Chrysophytes developments and perspectives. Nova Hedwigia Beiheft
isms with chlorophyllsa and c. Thesefeaturesarealsoshared 95:1–26.
by a number of other protist groups and therefore cannot be ANDERSEN,R.A.1991. Thecytoskeletonofchromophytealgae.Protoplasma
considered synapomorphic characters. Heterokont algae are 164:143–159.
united only by the presence of tripartite tubular hairs on the ANDERSEN, R. A. 2004. A historical review of heterokont phylogeny. The
JapaneseJournalofPhycology52:153–162.
immature flagellum. This feature is shared with nonphotosyn-
ANDERSEN,R.A.,R.W.BRETT,D.POTTER,ANDJ.P.SEXTON.1998a. Phy-
thetic heterokonts and perhaps the bipartite hairs of crypto- logenyoftheEustigmatophyceaebaseduponthe18SrRNAgene,with
phytes. Unifying morphological characters define heterokont emphasisonNannochloropsis.Protist149:61–74.
algalclasses,butestablishinghomologouscharactershasbeen ANDERSEN,R.A.,D.POTTER,R.R.BIDIGARE,M.LATASA,K.ROWAN,AND
difficult, restraining efforts to establish phylogenetic relation- C.J.O’KELLY.1998b. Characterizationandphylogeneticpositionofthe
enigmatic golden alga Phaeothamnion confervicola: ultrastructure, pig-
ships among classes. Molecularanalyses,basedupononetoa
mentcompositionandpartialSSUrDNAsequence.JournalofPhycology
fewgenes,haveindicatedsomephylogeneticrelationships,but
34:286–298.
considerablymoremolecularandmorphologicaladvanceswill ANDERSEN,R.A.,ANDH.PREISIG.2002a. Synurophyceae.InJ.J.Lee,G.
be required before consensus is reached on their broad phy- F.Leedale,andP.C.Bradbury[eds.],Anillustratedguidetotheprotozoa,
logenetic relationships. Similarly, the pendulum continues to 2nded.,vol.2,759–775.SocietyofProtozoologists,Lawrence,Kansas,
swing regarding opinions about the relationship betweenhap- USA.
tophyte and heterokont algae. The uncertain phylogenetic re- ANDERSEN, R. A., AND H. PREISIG. 2002b. Pelagophyceae. In J. J. Lee, G.
F.Leedale,andP.C.Bradbury[eds.],Anillustratedguidetotheprotozoa,
lationships for other related protistan groups (e.g., alveolates,
2nded.,vol.2,733–743.SocietyofProtozoologists,Lawrence,Kansas,
cryptophytes, cercozoans) confound the problem. USA.
Despite our limited knowledge about their phylogeneticre- ANDERSEN,R.A.,G.W.SAUNDERS,M.P.PASKIND,ANDJ.P.SEXTON.1993.
lationships, the heterokont algae are certainly a large and di- Ultrastructureand18SrRNAgenesequenceforPelagomonascalceolata
verse group of living organisms. There are many species of gen.etsp.nov.andthedescriptionofanewalgalclass,thePelagophy-
ceaeclassisnov.JournalofPhycology29:701–715.
diatoms, with estimates of up to a million ormorespeciesyet
ANDERSEN,R.A.,Y.VANDEPEER,D.POTTER,J.P.SEXTON,M.KAWACHI,
to be described (Round et al., 1990). Heterokont algae range
ANDT.LAJEUNESSE.1999. PhylogeneticanalysisoftheSSUrRNAfrom
in size from eustigmatophyte and pelagophyte picoplankters membersoftheChrysophyceae.Protist150:71–84.
(;1mm)tobrownalgalkelp(100minlength).Cellcoverings ANDERSEN,R.A.,ANDR.WETHERBEE.1992. Microtubulesoftheflagellar
include cellulosic walls, glass walls, organic and mineralized apparatusareactiveduringpreycaptureinthechrysophyceanalgaEpi-
scales, organic and mineralized loricas, and gelatinous sub- pyxispulchra.Protoplasma166:8–20.
stances. Theflagellarapparatusis highlyvariable,tothepoint ARCHIBALD,A.R.,W.L.CUNNINGHAM,D.J.MANNERS,ANDJ.R.STARK.
1963. StudiesonthemetabolismofProtozoa.10.Themolecularstruc-
thathomologousstructuresare difficulttoestablish.Similarly,
tureofreservepolysaccharidesfromOchromonasmalhamensisandPer-
haptophyte algae are diverse, although more fossilspeciesare anematrichophorum.BiochemicalJournal88:444–451.
known than living species. Conversely, Schizocladophyceae ARIZTIA,E.V.,R.A.ANDERSEN,ANDM.L.SOGIN.1991. Anewphylogeny
contains a single species, and Bolidophyceae, Chrysomero- forchromophytealgaeusing16S-likerRNAsequencesfromMallomonas
phyceae, Eustigmatophyceae, Pinguiophyceae, and Raphido- papillosa (Synurophyceae) and Tribonema aequale (Xanthophyceae).
JournalofPhycology27:428–436.
phyceae have fewer than 25 described species. At present, it
BAILEY, J. C., R. R. BIDIGARE, S. J. CHRISTENSEN, AND R. A. ANDERSEN.
is unclear whether these classes are ancient and consist of a
1998. Phaeothamniophyceaeclassisnova:anewlineageofchromophy-
few remnant species or if they are newly evolved groups that tes based upon photsynthetic pigments, rbcLsequence analysis and ul-
have not yet radiated. trastructure.Protist149:245–263.
Although studies in nuclear genes have been initiated(e.g., BEECH,P.L.2003. Thelongandtheshortofflagellarlengthcontrol.Journal
Fast et al., 2001; Yoon et al., 2002a, b; Harper and Keeling, ofPhycology39:837–839.
2003; Ryall et al., 2003; Yoon et al., 2004), a greater use of BEECH, P. L., AND R. WETHERBEE. 1988. Observations on the flagellarap-
paratus and peripherial endoplasmic reticulum of the coccolithophorid
multiple nuclear genes in a wide range and large number of
Pleurochrysiscarterae(Prymnesiophyceae).Phycologia27:142–158.
photosynthetic and nonphotosynthetic heterokonts and hapto- BEECH, P. L., AND R. WETHERBEE. 1990a. Direct observations on flagellar