Euglenozoa Cavalier Smith

Available online at www.sciencedirect.
com
ScienceDirect
European Journal of Protistology 56 (2016) 250–276
Higher classification and phylogeny of Euglenozoa

Thomas Cavalier-Smith
Department of Zoology, University of Oxford, South Parks Road, Oxford OX1 3PS, UK
Received 2 March 2016; received in revised form 5 September 2016; accepted 8 September 2016
Available online 15 September 2016
Abstract
Discoveries of numerous new taxa and advances in ultrastructure and sequence phylogeny (including here the first site-
heterogeneous 18S rDNA trees) require major improvements to euglenozoan higher-level taxonomy. I therefore divide
Euglenozoa into three subphyla of substantially different body plans: Euglenoida with pellicular strips; anaerobic Postgaardia
(class Postgaardea) dependent on surface bacteria and with uniquely modified feeding apparatuses; and new subphylum Gly-
comonada characterised by glycosomes (Kinetoplastea, Diplonemea). Euglenoida comprise two new infraphyla: Entosiphona
with three feeding rods and Dipilida ancestrally with two. Dipilida comprise basal superclass Rigimonada with longitudinal rigid
strips [i.e. new classes Stavomonadea (Petalomonadida, Decastavida and new order Heterostavida) and Ploeotarea (Ploeotiida)
with contrasting oral cytoskeletons] and derived superclass Spirocuta with more numerous spirally arranged, often slideable,
strips (clade Peranemea/Euglenophyceae) and a different, highly conserved microtubule pattern at strip joints. Peranemea com-
prise four orders: Peranemida (anterior gliding, protrusible rods), and three new, Anisonemida (posterior gliders), Natomonadida
(swimmers including phagotrophic new suborder Metanemina and osmotrophic suborder Rhabdomonadina), and Acroglissida
(anterior gliders with cytoproct). I establish orders Entosiphonida, Rapazida, Bihospitida; and seven new euglenoid families
(Entosiphonidae, peranemean Neometanemidae, Rapazidae, two stavomonad, two ploeotiid) and three new postgaardian, and
three kinetoplastid families (Neobodonidae, Rhynchomonadidae, Parabodonidae), plus new diplonemid family Hemistasiidae
for Hemistasia.
© 2016 The Author. Published by Elsevier GmbH. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).
Keywords: Diplonemea; Euglenoida; Glycomonada; Kinetoplastea; Spirocuta; Postgaardia
Introduction trees (mitochondrial cytochrome c: Schwartz and Dayhoff

1978). Euglenozoa are ancestrally aerobic, non-pseudopodial
When establishing phylum Euglenozoa to embrace zooflagellates with a microtubule-rich pellicle, a unique com-
euglenoids and kinetoplastids I argued that they differ plex feeding apparatus (FA), tubular extrusomes, parallel
in many respects from all other Protozoa, and seriously centrioles attached within a deep ciliary pocket by three dis-
considered putting them in a separate kingdom (Cavalier- tinctive microtubular roots to the pellicle and FA, and cilia
Smith 1981) as suggested earlier (Cavalier-Smith 1978). ancestrally with unique dissimilar latticed paraxonemal rods
They were ultrastructurally unique and the deepest eukary- (Cavalier-Smith 1981; Simpson 1997). More recent work
ote branch on the first prokaryote-rooted protein sequence amplifies their distinctiveness and confirms their evolutionary
unity (Cavalier-Smith 2010a, 2013a). Of the six classes previ-
ously recognised, only Euglenophyceae secondarily acquired
E-mail address: [email protected]
http://dx.doi.org/10.1016/j.ejop.2016.09.003
0932-4739/© 2016 The Author. Published by Elsevier GmbH. This is an open access article under the CC BY license (http://creativecommons.org/licenses/
by/4.0/).
T. Cavalier-Smith / European Journal of Protistology 56 (2016) 250–276 251
a green algal chloroplast by symbiogenesis (Cavalier-Smith and kinetoplastids treated as separate single-class subphyla
2013b; Gibbs 1981) and became algae that are well classi- (Cavalier-Smith 1993). That revision was largely based on the
fied (Bicudo and Menezes, 2016; Kim et al. 2010; Marin structure of the euglenoid and diplonemid FA (Triemer and
et al. 2003) except for there being no family or order for Farmer 1991a,b) and its phylogenetic implications (Cavalier-
the highly distinctive phagotroph Rapaza (Yamaguchi et al. Smith 1995), but did not adequately evaluate ultrastructural
2012). All other Euglenozoa are heterotrophic protozoa, aspects of euglenoid pellicle morphogenesis (Belhadri and
phagotrophic or more rarely osmotrophic, symbiotrophic or Brugerolle 1992; Mignot et al. 1987; Triemer and Fritz 1988).
parasitic, whose classification is currently confused and very It also preceded discovery of Postgaardi (Fenchel et al. 1995)
incomplete. with novel ultrastructure placing it in Euglenozoa (Simpson
Mignot (1964) first noted paraxonemal rod similarities et al. 1996/97), where it was formally accommodated
between a euglenoid and kinetoplastid, but they were first within new class Postgaardea grouped with kinetoplastids
suggested as related because of similar mitosis (Leedale, as subphylum Saccostoma (Cavalier-Smith 1998). Currently,
1970), and recognized as a clade by Taylor (1976) and Euglenozoa are classified in subkingdom Eozoa of kingdom
Cavalier-Smith (1978, both separated from other eukaryotes Protozoa and subdivided into six classes (three in subphylum
as ‘kingdom Euglenoida’) for these and other reasons. They Euglenoida: Peranemea, Euglenophyceae, and one unnamed)
have a unique cytochrome c with haem attached via one and 12 orders, subphylum Saccostoma being abandoned
cysteine (Pettigrew et al. 1975), not two as in all other orga- because Postgaardea proved to be much more distinctive than
nisms, which also requires different biosynthetic machinery was originally recognised (Ruggiero et al. 2015).
from the radically contrasting machineries used by bacteria As recently revised (Cavalier-Smith et al. 2015a), Eozoa
and excavates on the one hand and higher eukaryotes on the include only three phyla, phenotypically radically different
other; a novel single-protein non-bacterial machinery evolved and genetically deeply divergent but internally relatively uni-
in the last common protozoan ancestor of animals, fungi, form: (1) Euglenozoa with ciliary paraxonemal rods; (2) the
plants, and chromists (Allen 2011; Cavalier-Smith 2010a). biciliate grooved Eolouka (i.e. jakobids with a vaned posterior
Euglenozoa have dozens of molecular properties that are cilium and the most primitive mitochondrial genomes, plus
unique among eukaryotes. Many of these are clearly sec- Tsukubamonas without ciliary vanes); and (3) the quadri-
ondary acquisitions that would not have been present in the ciliate Percolozoa without paraxonemal rods or vanes but
ancestral eukaryote, but about a dozen have been interpreted with unstacked Golgi membranes (unlike Euglenozoa and
as primitive bacteria-related characters that collectively sug- most eukaryotes other than higher fungi and diplomonads).
gest that the root of the eukaryote evolutionary tree may Percolozoa share discoid mitochondrial cristae with Eugleno-
lie between Euglenozoa and all other eukaryotes (Cavalier- zoa so these discicristates were once made an infrakingdom
Smith 2010a,b). Some multiprotein trees are consistent with (Cavalier-Smith 1998). The presence of a few respiratory
that root position (Lasek-Nesselquist and Gogarten 2013) chain proteins shared by Euglenozoa and Percolozoa (Perez
though others are not, suggesting instead that it may be et al. 2014), some of which might conceivably be involved
between discicristates (Euglenozoa plus Percolozoa) and in causing their discoid cristal form, but which are absent
other eukaryotes (Raymann et al. 2015) or that discicristates from higher eukaryotes and bacteria, makes it possible
plus jakobids (Eozoa sensu Cavalier-Smith et al. 2015a) may that Discicristata (or Eozoa if also present in Eolouka, not
be a clade (e.g. Derelle et al. 2015) not the ancestral eukaryote currently known) are a clade, implying that the eukary-
group as Cavalier-Smith (2010a,b) argued. ote root is not between Euglenozoa and Percolozoa unless
The uniqueness of Euglenozoa is further emphasised by the these are ancestral proteins lost by other eukaryotes. Pre-
remarkable discovery that euglenoid and kinetoplastid mito- viously Eolouka were lumped with Malawimonas that has
chondrial respiratory chains share 34 proteins absent from a cytoskeletally related feeding groove (but a positionally
all other eukaryotes and bacteria (Perez et al. 2014). This non-homologous posterior ciliary vane) as phylum Louko-
respiratory chain uniqueness and their unique cytochrome c zoa (Cavalier-Smith 1999), which was later grouped with
biogenesis (Allen 2011) would both have arisen immediately Percolozoa and the anaerobic, ancestrally quadriciliate Meta-
after the origin of mitochondria if the divergence between monada as infrakingdom Excavata (Cavalier-Smith 2010a;
Euglenozoa and all other eukaryotes was the primary one Ruggiero et al. 2015). However multigene trees repeat-
in eukaryote evolution as Cavalier-Smith (2010a,b, 2013a, edly showed that Malawimonas and Metamonada are more
2014b) argued is most likely, or somewhat later if the root closely related to the podiate clade that includes opisthokonts,
were elsewhere. Amoebozoa, and Sulcozoa (Cavalier-Smith 2013a) than to
Euglenozoa now comprise four ultrastructurally very dis- Eolouka or Euglenozoa, and that excavates are a para-
tinct types of phylogenetically related protozoan flagellates: phyletic organisational grade (Brown et al. 2013; Burki et al.
euglenoids, postgaardids, diplonemids, and kinetoplastids 2016; Cavalier-Smith et al. 2014, 2015a,b, 2016b). Therefore
(Cavalier-Smith 1998; Simpson 1997). Higher classifica- Malawimonas and Metamonada were formally transferred
tion of Euglenozoa and euglenoids was last seriously from subkingdom Eozoa to protozoan subkingdom Neozoa
revised two decades ago when three euglenoid classes were by Cavalier-Smith et al. (2015a), Excavata (originally made
established within subphylum Euglenoida and diplonemids an infrakingdom by Cavalier-Smith 2002) being abandoned
252 T. Cavalier-Smith / European Journal of Protistology 56 (2016) 250–276
as a taxon, a substantial rearrangement of kingdom Protozoa level the major innovation is the new subphylum Gly-
that better fits eukaryote deep phylogeny and the distribution comonada to embrace kinetoplastids and diplonemids, which
of the two types of ciliary vanes on the eukaryote tree. 187-gene trees show are undoubtedly sister classes (Cavalier-
Of particular importance for overall euglenozoan tax- Smith et al. 2014, 2015a,b, 2016b). The name emphasises
onomy is the demonstration that Calkinsia, originally that both flagellate classes have glycosomes that contain the
considered a euglenoid (Lackey 1960), is ultrastructurally glycolytic enzymes (Makiuchi et al. 2011; Gualdrón-López
more similar to Postgaardi (Yubuki et al. 2009, 2013), thus et al. 2012), in contrast to the homologous more typical per-
confirming the placement of Calkinsia in class Postgaardea oxisomes of all other eukaryotes whose glycolysis takes place
by Cavalier-Smith (2003a,b), which was not accepted by Adl in the cytosol. Euglenoida remain a subphylum as in Cavalier-
et al. (2005) or by Yubuki et al. (2009) who proposed a new, Smith (1993). Postgaardea are placed in the new subphylum
in my view unnecessary, clade name Symbiontida for Calkin- Postgaardia, as they are radically different ultrastructurally
sia plus unspecified relatives, then excluding Postgaardi. from both euglenoids and glycomonads and are a third deep-
The recently discovered postgaardean Bihospites (Breglia branching euglenozoan clade that may be sister to Euglenoida
et al. 2010) revealed novel ultrastructural features confirming but do not branch within them or Glycomonada on the evo-
the distinctiveness of Postgaardea/Symbiontida. Though Adl lutionary most realistic sequence trees presented in the next
et al. (2012) followed Cavalier-Smith (1998, 2003a,b) in three sections, contrary to some poorly resolved earlier trees.
treating them as a group distinct from Euglenoida, Lax and Lower level innovations are predominantly amongst
Simpson (2013), Lee and Simpson (2014a,b), and Chan et al. phagotrophic euglenoids as numerous ultrastructurally and
(2015) reverted to including all Postgaardea in euglenoids as a phylogenetically distinctive new genera have been recently
discrete subgroup. For euglenoids sensu stricto (i.e. excluding discovered or better characterised (Breglia et al. 2013;
Postgaardea), a recent protist system (Adl et al. 2005, 2012) Cavalier-Smith et al. 2016a; Chan et al. 2013, 2015; Lax
reverted to Klebs’ (1892) oversimplified three-group system, and Simpson 2013; Lee and Simpson 2014a,b; Yamaguchi
ignoring numerous advances over the past century; there is a et al. 2012). This classification is the first to use families
pressing need to use them to make a far better classification, comprehensively since Euglenozoa was established; 14 of
especially for the phagotrophs. the 31 recognised here are new, as are seven of the 18 orders
Similarly important was the demonstration that kinetoplas- and three out of eight classes. Families were entirely omitted
tid ultrastructural diversity is much greater than is even now by Leedale (1967, 2002) for euglenoids. Indeed, all works
generally appreciated (Frolov and Karpov 1995; Frolov et al. on heterotrophic euglenoids known to me since Vasileva
2001), and certainly at the time of the previous major revi- (1987) ignored them; that deficiency reflects the difficulty
sion (Cavalier-Smith 1993) and when kinetoplastids were of delimiting euglenoid families noted by Pringsheim (1948)
first compared in detail with euglenoids (Brugerolle 1985; who rightly stressed that the three dating from Klebs (1892)
Kivic and Walne 1984). Moreira et al. (2004) established – Euglenidae (phototrophs), Astasiidae (osmotrophs), and
an improved phylogenetic classification of kinetoplastids, Peranemidae (phagotrophs) – ‘do not, however, seem to coin-
which can readily accommodate the additional genera dis- cide with true taxonomic groups’. Fortunately, comparative
covered since (e.g. Flegontov et al. 2013; Hirose et al. ultrastructure and better taxon-rich sequence trees (next sec-
2012; Stoeck et al. 2005). Diplonemid evolution and diplone- tion) now allow definitions of euglenoid families with good
mid/kinetoplastid relationships have been recently greatly prospects of future stability, as already well done for non-
clarified by the demonstration that Hemistasia previously phagotrophic phototrophs (Marin et al. 2003; Kim et al.
considered a kinetoplastid (Adl et al. 2012; Elbrächter 2010) and essayed here for the rest. I discuss classification of
et al. 1996) is actually a diplonemid (Yabuki and Tame each subphylum in sequence, starting with euglenoids, after
2015). Many other advances over the past two decades, first explaining my rather comprehensive site-heterogeneous
improved and taxonomically more comprehensive sequence sequence trees.
trees (including those presented in this paper), and critical
reevaluation of euglenozoan comparative anatomy (which
I shall publish in more detail elsewhere), make a radically
improved euglenozoan taxonomy now essential especially Euglenozoan Site-heterogeneous 18S rDNA
for the relatively neglected phagotrophic euglenoids and Phylogeny: Rationale and Methods
Postgaardea and at higher ranks.
The purpose of this paper is to present such a revised Using macgde v. 2.4 (http://macgde.bio.cmich.edu/) I
classification (Table 1) comprehensive to the family level manually aligned 18S rDNAs of 217 Euglenozoa plus 481
(placing almost all genera); to explain the reasons for the outgroup taxa representing all major eukaryote groups and
most important innovations and simplifications; to provide selected by eye 1577 or 1541 reasonably well aligned
enough historical background to put them into perspective; nucleotide positions for preliminary phylogenetic analysis
and to provide the first site-heterogeneous 18S rDNA tree for (>50% more than in some euglenozoan studies), depending
Euglenozoa (Fig. 1) to enable comparison of sequence phy- on whether the highly divergent Percolozoa were excluded
logeny with the new classification. At the highest taxonomic or included. 18S rDNA of Percolozoa, sometimes included
as outgroups for euglenozoan phylogeny (Lax and Simpson tic in allowing different patterns of nucleotide substitution
2013; von der Heyden et al. 2004), evolves much faster than across sites in the molecule (a site-heterogeneous model, not
in most eukaryotes making its alignment especially difficult site-homogeneous as in ML (Lartillot and Philippe 2004);
and potentially introducing long-branch artefacts (Cavalier- ML can be misleading if unrealistic models are used) and
Smith 2015). Ribosomal DNA trees for Euglenozoa have should give more accurate trees (never previously done for
typically been basally poorly resolved and contradictory for Euglenozoa), as it appears to have done for other difficult
deep relationships (Chan et al. 2013; Lax and Simpson 2013; to resolve protist groups with rDNA long-branches such as
von der Heyden et al. 2004; Yubuki et al. 2009). To increase gregarines and Percolozoa (Cavalier-Smith 2014a, 2015);
resolution I did four things: (1) used not only maximum like- (2) included many more sequences than any previous study,
lihood (ML) as in most other studies, but also the CAT-GTR which should increase accuracy of ancestral state recon-
model of PhyloBayes, which is evolutionarily more realis- struction; (3) included a much higher proportion of the 18S
Table 1. Classification of phylum Euglenozoa and its 8 classes, 18 orders, and 31 families.
Subphylum 1. Euglenoida Bütschli, 1884 (as Euglenoidina) stat. n. Cavalier-Smith, 1993 (aerobes; pellicular strips)
Infraphylum 1. Entosiphona infraphyl. n. (three feeding apparatus support rods)
Class Entosiphonea cl. n. (posterior ciliary gliders)
Order Entosiphonida ord. n.
Family Entosiphonidae fam. n. (Entosiphon)
Infraphylum 2. Dipilida infraphyl. n. (two feeding apparatus support rods)
Superclass 1. Rigimonada* supercl. n. (rigid longitudinal pellicle strips)
Class 1. Stavomonadea cl. n. (simple strip joints, not below prominent ridges)
Subclass 1. Homostavia subcl. n. (uniform pellicle strips)
Order 1. Decastavida Cavalier-Smith in Cavalier-Smith et al., 2016a (posterior gliders)
Family 1. Decastavidae Cavalier-Smith in Cavalier-Smith et al., 2016a (Decastava)
Family 2. Keelungiidae fam. n. (Keelungia)
Order 2. Petalomonadida Cavalier-Smith, 1993 (anterior gliders)
Family 1. Scytomonadidae Ritter von Stein, 1878 (syn. Petalomonadidae Bütschli, 1884)
(Atraktomonas, Biundula, Calycimonas, Dolium, Dylakosoma, Notosolenus, Pentamonas,
Petalomonas, Scytomonas, Tropidoscyphus)
Family 2. Sphenomonadidae Saville Kent, 1880 (Sphenomonas)
Subclass 2. Heterostavia subcl. n. (heteromorphic strips)
Order Heterostavida ord. n. (posterior gliders)
Family Serpenomonadidae fam. n. (Serpenomonas)
Class 2. Ploeotarea cl. n. (prominent dorsal pellicle ridges and 10 microgrooves)
Order Ploeotiida Cavalier-Smith, 1993 (posterior gliders)
Family 1. Ploeotiidae fam. n. (Ploeotia)
Family 2. Lentomonadidae fam. n. (Lentomonas)
Superclass 2. Spirocuta supercl. n. (spiral pellicle strips; often squirm)
Class 1. Peranemea Cavalier-Smith, 1993 em. (heterotrophs)
Subclass 1. Peranemia subcl. n. (protrusible rod apparatus)
Order Peranemida Cavalier-Smith, 1993 (anterior gliders)
Family Peranemidae Bütschli, 1884 (Peranema, Peranemopsis, Jenningsia, Urceolus)
Subclass 2. Anisonemia subcl. n. (non-protrusible rod apparatus)
Order 1. Anisonemida ord. n. (posterior gliders)
Family Anisonemidae Saville Kent, 1880 em. (Anisonema, Dinema*, Heteronema)
Order 2. Natomonadida ord. n. (swimmers)
Suborder 1. Metanemina subord. n. (phagotrophs)
Family Neometanemidae fam. n. (Neometanema)
Table 1 (Continued)
Suborder 2. Rhabdomonadina Leedale, 1967 em. Cavalier-Smith, 1993 (osmotrophs: originally
excluded Astasia and Distigma )
Family 1. Astasiidae Saville Kent, 1884 (as Astasiadae; incl. Menoidiidae Hollande, 1942)
(Astasia , Rhabdomonas, Gyropaigne, Menoidium, Parmidium)
Family 2. Distigmidae Hollande, 1942 (some violently plastic others rigid) ( Distigma )
Subclass 3. Acroglissia subcl. n. (anterior gliders)
Order 1. Acroglissida ord. n.
Family Teloproctidae Cavalier-Smith in Cavalier -Smith et al., 2016 a (Teloprocta )

Class 2. Euglenophyceae Schoenichen in Eyfurth and Shoenichen, 1925 orth. em. Smith, 1933** (ancestrally and largely phototrophs)
Subclass 1. Rapazia subcl. n. (phagophototrophs)
Order Rapazida ord. n. (16 pellicular strips)
Family Rapazidae fam. n. (Rapaza)
Subclass 2. Euglenophycidae Busse and Preisfeld, 2003 (non-phagotrophs)
Order 1. Eutreptiida* Leedale, 1967 (=Eutreptiales; 48 pellicle strips)
Family Eutreptiidae* Hollande, 1942 (Eutreptia*, Eutreptiella, Tetreutreptia)
Order 2. Euglenida Ritter von Stein, 1878 stat. n. Calkins, 1926 (=Euglenales Leedale, 1967; 40 pellicular strips, two posterior whorls) ***
Family 1. Euglenidae Dujardin, 1841 em. Kim et al., 2010

Subfamily 1. Eugleninae Hollande, 1952 em. (Euglena, Euglenaria, Khawkinea, Monomorphina, Cryptoglena)
Subfamily 2. Colaciinae Smith, 1933 (as family) (Colacium, Strombomonas, Trachelomonas)
Family 2. Phacidae Kim et al., 2010 (=Phacaceae) (Phacus, Lepocinclis, Discoplastis)
Family 3. Euglenamorphidae Hollande, 1952 (as subfamily) stat. n. (Euglenamorpha, Hegneria)
Subphylum 2. Postgaardia subphyl. n. (anaerobes; epicellular bacteria; strips, glycosomes absent)
Class Postgaardea Cavalier-Smith, 1998 (synonym Symbiontida Yubuki et al., 2009; no rank)
Order 1. Postgaardida Cavalier-Smith, 2003
Family 1. Postgaardidae fam. n. (Postgaardi)
Family 2. Calkinsiidae fam. n. (Calkinsia)
Order 2. Bihospitida ord. n.
Family Bihospitidae fam. n. (Bihospites)
Subphylum 3. Glycomonada subphyl. n. (heterotrophs with glycosomes, ancestrally phagotrophs)
Class 1. Diplonemea Cavalier-Smith, 1993
Order Diplonemida Cavalier-Smith, 1993
Family 1. Diplonemidae Cavalier-Smith, 1993 (Diplonema, Rhizopus)
Family 2. Hemistasiidae fam. n. (Hemistasia)
Class 2. Kinetoplastea Honigberg em. Vickerman, 1976 stat. n. Margulis, 1974
Subclass 1. Prokinetoplastina Vickerman in Moreira et al., 2004
Order 1. Prokinetoplastida Vickerman in Moreira et al., 2004
Family Ichthyobodonidae Isaksen et al., 2007 (Ichthyobodo, Perkinsela)
Table 1 (Continued)
Subclass 2. Metakinetoplastina Vickerman in Moreira et al., 2004

Order 1. Bodonida* Hollande, 1952 em. Vickerman, 1976, Krylov et al. 1980
Suborder 1. Neobodonina Vickerman in Moreira et al., 2004 stat. n.
Family 1. Neobodonidae* fam. n. (Neobodo*, Rhynchobodo, Actuariola, Azumiobodo, Cruzella, Cryptaulaxoides, Klosteria)
Family 2. Rhynchomonadidae fam. n. (Rhynchomonas, Dimastigella)
Suborder 2. Parabodonina Vickerman in Moreira et al., 2004 stat. n.
Family 1. Parabodonidae fam. n. (Parabodo)
Family 2. Cryptobiidae* Vickerman, 1976 (Cryptobia*, Procryptobia, Trypanoplasma, Cephalothamnion)
Suborder 3. Eubodonina Vickerman in Moreira et al., 2004 stat. n.
Family Bodonidae Bütschli, 1883 em. (Bodo)
Order 2. Trypanosomatida Saville Kent, 1880 stat. n. Hollande, 1952
Family Trypanosomatidae Doflein, 1901 (e.g. Blastocrithidia, Crithidia, Leishmania,
Leptomonas, Paratrypanosoma, Phytomonas, Sergeia, Trypanosoma, Wallaceina)
* Probably paraphyletic.
** I avoided the much older potential name Euglenea Bütschli, 1884 for Euglenophyceae, used contradictorily as a synonym for all Euglenophycidae by Busse
and Preisfeld (2003) and by Adl et al. (2012) for Euglenidae only, as it was also published as a beetle genus in 1896. There is no good reason to regard any
Euglenozoa except class Euglenophyceae as algae, which are best defined as ‘oxygenic photosynthesisers other than embryophyte land plants’ (Cavalier-Smith
2007); it is simplest to treat all Euglenozoa including Euglenophyceae uniformly under the Zoological Code of Nomenclature, under which the names Dinema,
Entosiphon, and Peranema are valid; thus younger replacement names Dinematomonas, Entosiphonomonas and Pseudoperanema, are entirely unnecessary,
contrary to some statements, and their use is nomenclaturally destabilizing. If botanists wish for historical reasons to treat photosynthetic euglenoids under
the International Code of Nomenclature for algae, fungi and plants, that is not unreasonable provided that this policy is strictly limited to Euglenophyceae as
circumscribed here (as it includes Rapazida, broader than Euglenophycidae of Busse and Preisfeld (2003) and compositionally equivalent Euglenophyceae
sensu Marin et al. 2003). The botanical division/phylum name Euglenophyta Pascher, 1931 should not be applied to Euglenozoa or Euglenoida as a whole as it
is profoundly misleading for either of these ancestrally phagotrophic, non-photosynthetic, non-algal protozoan taxa. I considered using Euglenidea, already in
use decades ago (e.g. Fernández-Galiano 1990), as a ‘zoological’ tradition class name instead of Euglenophyceae, but settled on the latter as it should be more
acceptable to both phycologists and protozoologists and thus promote future stability.
*** Euglenida of Ritter von Stein (1878) was ranked as a family and referred only to Euglenidae, not all euglenoids, whose unity was first recognised by
Bütschli (1884/85) under the name Euglenoidina accepted as an order by many from Blochmann (1895) through Walton (1915) to Kudo (1966) and class or
higher by Hollande (1952), Leedale (1967), and numerous more recent authors, e.g. Cavalier-Smith (1993, 1998); Hausmann and Hülsmann (1996), and Karpov
(2000). The traditional vernacular term euglenoids (Walton 1915) applies to all suphylum Euglenoida. Now disused early spellings (Eugleninae Lemmermann,
1913; Euglenineae) predate general use of rank-informative suffixes and are only of historical interest. A recent fashion (e.g. Adl et al., 2012) for spelling all
euglenoids Euglenida (possibly related to Calkins’ (1926) influence on American protozoologists and their slowness to accept a rank above order (Honigberg
et al. 1964; Levine et al. 1980), as always favoured by phycologists and most other protozoologists since 1952) or keeping that ordinal suffix even at phylum rank
(Walne and Kivic 1990) causes confusion with the much narrower order Euglenida (and still more restricted family Euglenidae). The vernacular ‘euglenids’ is
similarly ambiguous, but might usefully be restricted in scope to refer just to order Euglenida.
rDNA molecule than before to provide more phylogenetically 2013; Lee and Simpson 2014a,b; Yamaguchi et al. 2012).
useful characters (up to 1577 nucleotides, over 50% more Restricted outgroups (especially just two species of one sub-
than in some studies) made possible by including so many order) risk biasing the position of the euglenozoan root by
taxa, which enables accurate alignment in regions previously being unrepresentative; a problem with including Percolo-
excluded because alignment is harder when taxa are sparser; zoa is that they have systematically the longest branches of
and (4) included many more outgroup taxa than before, which any whole eukaryotic phylum (Cavalier-Smith 2014a, 2015)
should enable the position of the euglenozoan tree’s root to and risk biasing the root position by convergent evolution
be more accurately calculated, reducing potential errors from with long ingroup branches. A broadly representative out-
its misplacement. group without excessively long branches is probably best;
Preliminary analysis investigated effects of outgroup as many non-excavates have shorter branches and are evo-
choice on euglenozoan phylogeny for the first time by analyz- lutionary diverse, excluding them in previous studies was
ing by both methods three different taxon and gene segment unwise. I initially analysed a 698-taxon sample including 382
samples. Since von der Heyden et al. (2004) it has been short-branch representatives of all non-excavate outgroups,
customary to use only excavates as outgroups in eugleno- 43 short-branch excavates, and a thorough sampling of Per-
zoan rDNA trees; Lax and Simpson (2013) used Jakobea, colozoa, but with only nucleotides accurately alignable with
Tsukubamonas and Percolozoa, whereas others used even them included (1541 nucleotides) to allow comparison with
more restricted outgroups: just two Jakobina (Chan et al. previously published trees using only jakobid outgroups; and
2013) or two Andalucina (Yubuki et al. 2009); most used a 609 taxon sample with 382 short-branch representatives
no outgroups so were arbitrarily rooted (e.g. Breglia et al. of all neokaryote outgroups other than Percolozoa. To allow
faster convergence of PhyloBayes trees in later analyses I accurately aligned or seem too divergent, but excluded sev-
pruned outgroup diversity of close relatives without sac- eral substantial segments that I included and which are not
rificing phylogenetic breadth. The two PhyloBayes chains especially difficult to align, quite well conserved, and phylo-
in the resulting 240 taxon sample (Supplementary Fig. S1) genetically informative. These serious alignment errors and
converged well onto the same topology (maxdiff 0.16). Con- non-ideal choice of segments for analysis probably explain
vergence was less good for the 609 and 698 taxa samples; why Chan et al. (2015) did not even show Euglenoida as
nearly all their branch patterns were the same, but there were monophyletic, as it is with strong support on all but one of
a few differences, especially near the base of Euglenozoa. my trees (an ML tree including the extremely long-branch
There were significantly more differences with ML. Entosiphon; Euglenoida was invariably a clade with Phy-
To reduce discrepancies I added ∼35 more Euglenozoa loBayes). I did not make similar comparisons with other
to the alignment, removed the longest branch Percolozoa papers using widely varying nucleotide numbers, down to
(mainly Percolatea) and further improved parts of the align- 734 (Yubuki et al. 2009), as their alignments were not avail-
ment, mainly to certain partially misaligned Entosiphon able; however this comparison cautions against assuming that
segments, but in small ways also to some relatives of added alignments including fewer nucleotides are necessarily more
Euglenozoa. To lessen computing time and ensure conver- accurate. As my alignment took much effort and care, and is
gence in this improved alignment I restricted sampling of taxonomically and in nucleotide positions much more com-
non-eozoans to a representative subset, giving 323 taxa alto- prehensive than any other to date (and I suspect more accurate
gether; as Entosiphon has the longest branch in the tree, I also than most), it is in the supplementary material for general use
analysed 318 taxa excluding Entosiphon to see if its presence and criticism (together with the three masks used for selecting
distorted topology or reduced bootstrap support for biparti- positions for analysis).
tions of interest. I also ran trees excluding all Percolozoa to
check whether their divergent sequences change the apparent
position of the euglenozoan root (282 taxa without and 287 Euglenozoan Molecular Phylogeny
taxa with Entosiphon: Fig. 1). For accurate topology among
closer relatives one should include as many nucleotide pos- Fig. 1 is a site-heterogeneous tree based on 1577 nucleotide
itions as possible, whereas for accuracy for deep branches it positions of 18S rDNA with 224 Euglenozoa plus outgroups
may be better to exclude some of the fastest evolving sites restricted to short-branch excavate lineages as in most previ-
(unfortunately there is probably no overall optimal number ously published euglenoid rDNA trees (also for ease of fitting
and no objective way of deciding a best compromise). For onto one page). It excluded long-branch Entosiphon whose
our first Euglenozoa-wide trees (von der Heyden et al. 2004) position is plotted on it from a separately run tree with iden-
we included only 1233 positions, but much improved taxon tical alignment. On the latter tree tree Entosiphon appears as
sampling now allows me to align 1577 nucleotides with rea- sister to Stavomonadea (Fig. 1), whereas by Hsp90 it is sister
sonable confidence. To check whether excluding the fastest to all Dipilida (Cavalier-Smith et al. 2016a). Supplementary
evolving (but generally reasonably aligned) sites gives differ- Fig. S1 is a tree run before Neometanema parovale (Lee and
ent results, I also ran trees including only 1425 nucleotides. Simpson 2014a,b) and Serpenomonas costata sequences of
Phylogenetic analysis was by RAxMLHPC-PTHREADS- Chan et al. (2015) were published, and before I improved
SSE3 v. 7.3.0 (Stamatakis 2006) using the GTRGAMMA Entosiphon’s alignment to my present satisfaction; CAT mis-
model with four rate categories and 400 or 1000 fast leadingly put Entosiphon within Stavomonadea as in Lax and
bootstraps (4 processors) and by the site-heterogeneous Simpson (2013) but as sister to Keelungia not petalomon-
CAT-GTR-GAMMA (4 rates) model of PhyloBayes v. 3.3 ads as on their tree (and by ML for Fig. S1 alignment). This
(Lartillot and Philippe 2004) with two chains for thousands shows unsurprisingly that some misalignment can change the
of generations after log likelihood values plateaued, early pre- apparent positions of Entosiphon, and also that the relatively
plateau trees being removed as burnin before summation of small degree of Fig. S1 misalignment (much less than in Chan
all other trees (for brevity called CAT only in the text). Using et al. (2015)) predominantly of Entosiphon does not change
the final improved rDNA alignment I ran 16 trees with differ- euglenozoan topology generally, with the single exception of
ent taxon and sequence samples and algorithm in all the above the rather close basal branching order of the three stavomonad
combinations. All Bayesian trees converged with maxdiff orders (different in Figs 1, S1 CAT though S1 ML had the
<0.3, mostly 0.1 or less. I prepared trees for publication using same stavomonad topology as Fig. 1) and of metakinetoplas-
FigTree v. 1.2.2 (http://tree.bio.ed.ac.uk/software/figtree/) tids (Fig. S1 with a slightly less refined alignment wrongly
and Eazydraw. put Trypanosomatida as sister to rather than within bodonids
I compared my 18S rDNA alignment to the 1374 nucleotide as in Figs 1, S2 and previous multiprotein trees). Extremely
alignment of Chan et al. (2015); that is grossly misaligned close bush-like branching of several lineages is generally the
in several tracts, especially for Serpenomonas (=Ploeotia most important factor (more so than long-branches alone) that
costata) the subject of their paper. Moreover it included makes parts of sequence trees extremely hard to resolve, even
many nucleotide positions here excluded (from both my with trees based on hundreds of proteins (Cavalier-Smith et al.
1425 and 1577 nucleotide selections) because they cannot be 2015b).
0.64/- Entosiphon oblongum type strain

X2 0.65/75 activated
Entosiphon oblongum TCS-2003 AY425008
ENTOSIPHONA 1/100
sludge DNA Q4C041QESU GU920505
Entosiphon sulcatum CCAP 1220/1B AY061999
0.73/59 Entosiphon sulcatum CCAP 1220/1A AF220826
E 0.73/68 Decastava edaphica Decastavida
Keelungia pulex Scytomonas saepesedens
Arctic oxygen-depleted marine tidal sediment DNA D2P04B10 EF100248 .94/80
0.99/98 Petalomonas cantuscygni CCAP 1259/1 AF386635 Petalomonas cantuscygni
U 0.47/34 0.98/66 0.94
/-
Petalomonas cantuscygni U84731
deep sea microbial mat cold seep DNA RM2-SGM31 AB505539
Arctic oxygen-depleted marine tidal sediment DNA D3P06F06 EF100316
freshwater sediment DNA CH1_S2_16 AY821957
G 0.99/100 0.6/51
peat bog water DNA PR3_3E_63 GQ330643
Biundula (=Petalomonas) sphagnophila 3
Petalomonadida Stavomonadea
0.99/70
L isolate U1 KC990930
Notosolenus urceolatus
Notosolenus ostium
E
AF403149
Serpenomonas aff. costata KF586333 DIPILIDA
E 0.98/80 Serpenomonas aff. costata KF586332
Serpenomonas costata CCAP 1265/1
Ploeotiida Ploeotarea
Heterostavida
Ploeotia cf. vitrea KC990993 Montegut-Felkner and
N Triemer strain
CCAP 1260/1B AF386636 Peranema trichophorum Peranemida Peranemea
Dinema sulcatum AY061998
Dinema platysomum U3 Anisonema/Dinema sp. KC990932 Anisonemia
0.99/100 0.99 0.99/100
O 0.79/49
0.99/100
/63
BIPA 2001 misidentified as Peranema sp. AY048919
G1 Anisonema sp. KC990936
W1 Anisonema sp. KC990935
B1 Anisonema acinus KC990937
Anisonemida
0.93/63 Anisonema acinus AF403160
I Spirocuta 0.48 /- Neometanema parovale KJ690254 Metanemina
Neometanema (=Heteronema) cf. exaratum Distigma proteus var. longicauda
0.99/42 Distigma gracilis SAG 216.80 AF386637
D 0.91/75
Rhabdomonadina 0.99/99
0.99/100
Distigma gracilis CCAP 1216/2 AY061997
Astasia curvata 4
Astasia torta
Menoidium gibbum 3 Natomonadida
Rhabdomonas intermedia
A .99
Parmidium circulare
Parmidium scutulum 0.99/89
0.99/82 /92 Rhabdomonas incurva AF403151
Rhabdomonas intermedia AF247601
Gyropaigne costata AF110418
Rhabdomonas costata AF295021
E
Teloprocta (=Heteronema) scaphurum Acroglissida
0.61/53
Rapaza viridis
Eutreptiella gymnastica AJ532400
Eutreptiella eupharyngea
Rapazia
Eutreptiella pomquetensis CCMP 1491 AJ532398
U
0.99/95 0.96/100 AJ532397 Eutreptiella braarudii Eutreptiida
0.99/100
Eutreptiella braarudii
Eutreptiella CCMP389
Discoplastis sp. FJ719606
Discoplastis spathyrhyncha AJ532454
G
0.53/- Phacus limnophila
0.8/61 Euglenida
0.99/100 0.99
Phacus platalea
Phacus acuminatus 0.98/34
Phacus tortus
L
Phacus ranula
/73 .5/- Phacus warszewiczii
0.99/63
Phacus pleuronectes var. minimus
Phacus skujae
1/98 Phacus
Phacus oscillans 1/84
E
0.98/25 .94 orbicularis
0.79/-
/25 .43/-
. 74/-
Phacus swirenkoi
Phacus acuminatus var. triquetra
Phacus segretii
Phacus hamelii
Euglenophyceae N
0.99/58 .99/70 Phacus applanatus Euglenophycidae
Lepocinclis salina
Lepocinclis 4 1/95
Euglenaformis proxima
Monomorphina aenigmatica
O
.99 Monomorphina pseudopyrum
0.99/56
/46
0.99/81
Monomorphina rudicola
Monomorphina inconspicua
Monomorphina ovata
Z
Monomorphina sp. MI 73
0.95/60 0.45/- Cryptoglena

Euglenaria sp. JQ691737
0.97/95 Euglena longa
Euglena
hiemalis
Euglena gracilis
O
skujae
0.84/21
1/97 Cryptoglena pigra
Trachelomonas zorensis
Trachelomonas sp. T603
Strombomonas triquetra
A
0.99/100 Strombomonas costata.98/97
Strombomonas ovalis
0.54/- 0.46/- Strombomonas borysthensis
Calkinsia aureus Colacium 7 0.99/100
POSTGAARDIA 0.99/100 environmental DNA 6
0.91/100 marine DNA from super-sulfidic anoxic fiord water DQ310255 marine DNA clade Postgaardida Postgaardea
marine DNA from sulfidic suboxic Black Sea water HM749952
0.99/100 HM004354
Bihospites baccati (synonym Symbiontida)
0.99/63 Hemistasia phaeocysticola
HM004353 Bihospitida
G 0.99/100 marine DNA 5 0.99/100
Diplonema sp. ATCC 50224
Hemistasidae
deep ocean DNA DQ504350
L 0.8/81 Rhynchopus sp. ATCC 50229 AY425014
Rhynchopus euleeides ATCC 50226 0.52/83 0.99/63 Diplonemea
Y 0.42/- Diplonema sp. ATCC 50232 Diplonemidae
Diplonema sp. ATCC 50225
0.94/54 Diplonema ambulator ATCC 50223
C 0.88/52 0.98/98 AF30522 Mid Atlantic Ridge sediment DNA
Trypanosomatida
unidentified marine clade AF30521 0.82/93 Paratrypanosoma confusum
O other Trypanosomatina 20 0.99/97
0.50/69
0.99/100
Parabodonina 8 0.99/86
Eubodonina 12 .86/71+ Neobodo saliens 0.21/- Bodonida
M 0.99/100 .38/-
AY998650 'Neobodo designis' SCCAP BD54
AY998651 'Neobodo designis' SCCAP BD55
Metakinetoplastina AY490220Neobodo Auckland 0.99/100
O DQ207578 Neobodo designis HFCC8 brackish Baltic
0.51/- .99 AY753625 Neobodo Lake sediment Panama2
.99 /94
Neobodo AY753610 Lake sediment San Francisco .68/85 Kinetoplastea
N 0.99/100 0.51/- /99 AY753618 Neobodo marine Raglan1
0.68/55
.99/1 DQ207589 Neobodo saliens HFCC11
A 0.52/-
.99/71 AY753600 Neobodo Devon 2 marine rockpool .93/73
AY425021 Cryptaulaxoides-like Costa Rica
Rhynchobodo 4 .98/70
D 0.37/- Neobodo designis 20 .42/-
Dimastigella 4 Neobodonina
0.27/- 1
Rhynchomonas nasuta 4
A 0.96/100 Perkinsiella-like AFSM3
Ichthyobodo 7
AY163355
Prokinetoplastina
0.99/90 Tsukubamonas globosa
.54/30
Jakobina 5
Breviata anathema Tsukubamonadea
.98/84
Anaeromonadea 3 0.97/-
Neozoa 430.41/-
Malawimonas jakobiformis
neokaryotes
Andalucina 4 Malawimonadea
0.97/90
0.3
Fig. 1. PhyloBayes CAT-GTR-GAMMA tree for 18S rDNA of 282 eukaryotes (224 Euglenozoa) and 1577 nucleotide positions. Both chains
(each run for nearly half a million genrations) converged to the same topology (maxdiff 0.14; 163,998 trees summed, excluding first 1604 as
burnin). Newly sequenced taxa are in bold. Support values for bipartitions are posterior probabilities (left) and bootstrap percentages for 1000
fast bootstraps for a RAxML GTR gamma analysis for the same alignment (right). Black blobs mean maximal (1.0, 100%) support by both
methods, for example that between Euglenozoa and all other eukaryotes. Taxa for some outgroups whose internal phylogeny is irrelevant to
this paper are collapsed to enable fitting onto one page; the number of species for each is beside its name; their names and internal topology
are on an uncollapsed 323-taxon tree that included Entosiphon and Percolozoa (Fig. S2). The Entosiphon branch is from the equivalent
CAT-GTR-GAMMA tree with 287 taxa; its ultra-long stem has been halved in length, the arrow marked E and dashed line showing where it
joined that tree, the rest of whose topology is identical to this except for the position of Dinema sulcatum that grouped with other anisonemids
(see text); when Percolozoa are included in the outgroup Entosiphon moves down one node to be sister to Dipilida and Teloprocta becomes
sister of Anisonemia (Fig. S2). Tree-rooting follows Cavalier-Smith (2010a, 2013a, 2014b). Clade names follow the taxonomy of Table 1.
Table 2. Sensitivity of Postgaardea placement to sequence sampling and substitution model.
Nucleotide positions and algorithm No Percolozoa in outgroup Percolozoa in outgroup

Entosiphon Entosiphon
−(282 taxa) +(287 taxa) −(318 taxa) +(323 taxa)
1425 positions CAT E (0.40) E (0.43) E (0.52) E (0.40)

ML G (30%) G (26%) E (45%) E (40%)
1577 positions CAT E (0.53) E (0.52) E (0.67) E (0.48)
ML G (25%) E (46%) E (70%) Ento (40%)
Position of Postgaardea on 16 18S rDNA trees: E = sister to Euglenoida; G = sister to Glycomonada; Ento = sister to Entosiphon. The differences between E
and G do not really imply a difference in topology within Euglenozoa; they simply reflect where the chosen outgroup joins the tree and therefore are a rooting
rather than internal branching uncertainty; ML is clearly more sensitive to outgroup choice. Trees were run including (+) or excluding (−) Entosiphon. Support
values (PP or BS %) are in brackets.
Table 3. Sensitivity of Teloprocta and Entosiphon placement to sequence sampling and model.
Nucleotide positions and algorithm No Percolozoa in outgroup Percolozoa in outgroup
Entosiphon Entosiphon
−(282 taxa) +(287 taxa) −(318 taxa) +(323 taxa)
1425 positions CAT T: EP (0.66) EP (0.65) A (0.54) EP (0.39)

E: PS (0.35) Sp (0.44)
ML T: EP (53%) EP (53%) EP (49%) EP (48%)
E: K (27%) K (32%)
1577 positions CAT T: EP (0.61) EP (0.63) A (0.66) A (0.75)
E: St (0.56) D (0.35)
ML T: EP (53%) R (21%) EP (49%) EP (50%)
E: K (22%) P (30%)
Position of Teloprocta (T: shown in bold) on 16 18S rDNA trees: EP = sister to Euglenophyceae; A = sister to Anisonemia; R = sister to Rhabdomonadina.
Position of Entosiphon (E:) on 8 18S rDNA trees: PS = sister to Ploeotaria plus Spirocuta; Sp = sister to Spirocuta; D = sister to all other euglenoids (Dipilida)
St = sister to Stavomonadea; K = sister to Keelungia; P = sister to Postgaardea.
Trees were run including (+) or excluding (−) Entosiphon. Support values (PP or BS %) are in brackets.
Supplementary Fig. S2 using 1577 nucleotide positions sampling including broadly sampled outgroups give a bet-
from 323 taxa including Entosiphon and a much broader and ter resolved deep euglenozoan branching order than previous
more representative outgroup selection including Percolo- less well sampled site-homogeneous trees.
zoa exemplifies the 16 trees obtained after the best alignment On all 18S rDNA trees Diplonemea and Kinetoplastea
(also used with fewer taxa for Fig. 1) was established. It shows group together as a clade (here called Glycomonada because
the same branching order for Stavomonadea as Fig. 1, differ- of their shared glycosomes; Fig. 1), but support varies greatly
ing within euglenoids only in the positions of Entosiphon with different taxon samples, outgroups, and number of
and Teloprocta, each moving by one node, and also in the included nucleotides (maximally with posterior probability
deep branching within Metakinetoplastina (though like Fig. 1 (PP) 0.99 and ML bootstrap support (BS) 88%). A con-
correctly placing trypanosomatids within bodonids). Fig. S2 currently published Hsp90 tree, though much more sparsely
shows Entosiphon as sister to Dipilida (therefore the deep- sampled (Cavalier-Smith et al. 2016a, Fig. 9), is also congru-
est branching euglenoid) in complete harmony with Hsp90 ent for both methods: Diplonemea are sisters to kinetoplastids
(Cavalier-Smith et al. 2016a) but much lower posterior proba- with maximal support by CAT and near maximal (99%) by
blity support. However, even with this superior alignment the ML. This agreement between Hsp90 and rDNA trees con-
position of Postgaardea (Table 2), Entosiphon and Teloprocta trasts with many previous 18S rDNA studies noted above
(Table 3) are sensitive to whether or not Percolozoa is present that incorrectly grouped Diplonemea with Euglenoida not
in the outgroup, to algorithm (site-heterogeneous or not), and with Kinetoplastea as Hsp90 showed. Within Kinetoplas-
to whether more or fewer faster evolving parts of the molecule tea there is maximal support for the deep divergence of
are included. In all other respects euglenozoan tree topol- prokinetoplastids and metakinetoplastids (bodonids plus try-
ogy except deep metakinetoplastid topology was insensitive panosomatids) for both molecules, but basal branching within
to these technical variations with this improved alignment metakinetoplastids is slightly contradictory between them in
with almost all nodes strongly or maximally supported (Fig. two respects: Hsp90 puts Bodo as sister to trypanosomatids,
S2). Thus site-heterogeneous 18S trees with massive taxon like 192-gene trees (Cavalier-Smith et al. 2014), whereas
rDNA groups Bodo and parabodonids with insignificant sup- Postgaardea Are Not Euglenoids, but
port (most CAT trees, e.g. Figs 1, S1) or groups parabodonids Possibly Their Sisters
and trypanosomatids with insignificant support (e.g. Fig. S2
CAT and Fig. 1 and other ML trees); Hsp90 shows neo- Unlike some published ones, my 18S rDNA trees
bodonids as paraphyletic whereas rDNA either agrees (ML; collectively argue strongly against the bacteria-covered
CAT with 1425 positions) or shows them as a weakly sup- postgaardians branching within Kinetoplastea or classical
ported clade (CAT with 1577 positions, e.g. Figs 1, S2). euglenoids. Contrary to previous conflicting studies, all
The internal phylogeny of euglenoids shown by my except one rDNA tree firmly excluded Postgaardea (=Sym-
present rDNA trees was discussed in detail by Cavalier- biontida) from euglenoids, which are a robust clade with
Smith et al. (2016a). In brief, they strongly confirm holophyly support for exclusion of Postgaardea varying with sampling
of Spirocuta with numerous spiral strips and that they are (maximally 0.99, 80%). However the position of post-
evolutionarily derived from (ancestral or paraphyletic) Rigi- gaardeans is sensitive to outgroup choice and algorithm.
monada with fewer longitudinal strips and ancestrally a 2-rod Amongst 16 trees for the most refined alignments, CAT
FA. Entosiphon, also with few longitudinal strips but a rad- always showed Postgaardea as sister of euglenoids with typi-
ically distinct 3-rod FA is clearly the most deeply divergent cally insignificant support (PP 0.44–67; details Table 2); four
euglenoid on Hsp90 trees; though some rDNA trees support ML trees also gave that topology (BS 40–70%); three contra-
this others do not (Table 2). All contradict earlier inclusion dictorily put them as sister to Glycomonada (25–32%) and
of Entosiphon in Peranemida (Cavalier-Smith 1993), as does one as sister of Entosiphon only (30%). All preliminary trees
Hsp90 strongly. In most cases, inclusion of Entosiphon or also strongly excluded them from Euglenoida. The unique
not did not alter tree topology, just support values for deep- grouping with Entosiphon is clearly artefactual. A relation-
est euglenoid branches. However its exceptionally divergent ship with euglenoids is likely (and consistent with most
rDNA prevents its conclusive placement by rDNA sequence previously published ML trees), but protein trees are needed
trees, making its position unstable and highly sensitive to to decide between the two alternatives and the third possibil-
taxon and site sampling, so Hsp90 is probably more reliable. ity that Postgaardea are sisters of all other Euglenozoa, which
The only rDNA tree that correctly placed it is the CAT tree rDNA trees do not exclude.
with the most complete sequence and broadest outgroup rep- Critical pellicle comparisons lead me to conclude that the
resentation (1577 positions, 323 taxa: Fig. S2); this shows that Bihospites pellicle does not have euglenoid-like S-shaped
(provided one takes exceptional care with the alignment in a pellicular strips as Breglia et al. (2010) claimed. Postgaar-
very large database) using the largest practical number of sites dian pellicle local ultrastructure is essentially the same as
can be more accurate than trimming alignments drastically in diplonemids and gives no support to a relationship with
to include only the most easily aligned parts that is too often euglenoids; Simpson et al. (1996/97) stressed that the even
done (and never positioned Entosiphon concordantly with microtubule (mt) array of Postgaardi pellicles is ‘indistin-
Hsp90 trees). The position of Teloprocta is also unstable as guishable from the pattern seen in large kinetoplastids and
previously found (Table 2). Organisms called Ploeotia in the diplonemids’ and ‘do not follow the “euglenid pattern”’.
past are so deeply divergent that they clearly belong in sev- Bihospites and Calkinsia pellicles have essentially identical
eral genera, e.g. Serpenomonas costata does not group with mt arrangement and cross-bridging. In my view FA homolo-
Ploeotia cf. vitrea, showing that phagotrophic euglenoids are gies of Postgaardea have also been misinterpreted, as briefly
much more diverse than often supposed. As noted below, explained in a later section. Fig. 2 summarises the findings
Serpenomonas pellicle is unique but previously greatly misin- of rDNA and Hsp90 trees.
terpreted, so Serpenomonas is here placed in a separate order
distinct from order Decamonadida recently established for
Decamonas and Keelungia (Cavalier-Smith et al. 2016a), and
grouped with it and Petalomonadida as new class Stavomon- Euglenoid High-level Taxonomy Needs
adea that excludes Ploeotiida sensu stricto (Table 1). Radical Revision
Stavomonadea are a clade on rDNA trees where
Entosiphon is excluded or if included does not (arguably arte- Bütschli (1884) first established euglenoids as an almost
factually) intrude amongst them. Petalomonadida are weakly unified group under the name Euglenoidina (not Euglenida
sister to Decamonadida (Fig. 1, S1 ML, S2). Though multi- as used by Adl et al. (2012) who claim to select the old-
gene trees are needed to establish the branching order of est names). Lankester (1885) included six families – only
the three stavomonad orders with confidence, their collec- five actually euglenoids and only one (Euglenina) photo-
tive unity and consistent strong divergence from Ploeotia cf. synthetic – in order Euglenoidea Bütschli, still the most
vitrea on Fig. 1, S1, S2 and all my other trees make it clearer suitable spelling if euglenoids are treated as a class (e.g.
than before that petalomonad FAs are secondarily simpli- Cavalier-Smith 2003a,b; Hausmann and Hülsmann 1996).
fied by rod loss, not primitively simple as often assumed, Lankester also had a fourth phagotrophic euglenoid fam-
and suggest that their pellicles are also secondarily simplified ily, Anisonemina comprising Anisonema and Entosiphon
(Cavalier-Smith et al. 2016a). (narrower than Anisonemidae Saville Kent, 1880 that also
18 orders 8 classes 3 subphyla

Rigimonada
Entosphonida Entosiphonea
Decastavida
Petalomonadida Stavomonadea
Heterostavida
Ploeotiida Ploeotarea
Peranemida
Anisonemida
Euglenoida
Dipilida Peranemea
Natomonadida
Spirocuta Acroglissida
Rapazida
chloroplast Eutreptiida Euglenophyceae
Euglenophycidae Euglenida
Postgaardida
Bihospitida Postgaardea Postgaardia
Diplonemida Diplonemea
Prokinetoplastida
Glycomonada
Bodonida Kinetoplastea
Metakinetoplastina
Trypanosomatida
Fig. 2. Schematic relationships amongst the major euglenozoan taxa as shown by 18S rDNA (Figs 1, S1, S2) and Hsp90 trees (Cavalier-Smith
et al. 2016a). Eutreptiida, Bodonida, Rigimonada, and Peranemea are known to be paraphyletic and ancestral to the groups indicated. All
other named taxa are probably clades, though a minority of my trees (e.g. Fig. 1) raise the possibility that Anisonemida might be paraphyletic
ancestors of Natomonadida because of deeper branching of Dinema scaphurum (most, as Cavalier-Smith et al. (2016a) explained, show
Anisonemida as a clade, e.g. Fig. S1), and only one sequence each is known for Ploeotiida and Acroglissida (Teloprocta); more are needed
to test this. There is also uncertainty whether Acroglissida are sister to Euglenophyceae as shown (e.g. Fig. 1) or to Anisonemia (dashed line
and Fig. S2).
had Heteronema), but did not consider them euglenoids, Eugleninae (Lemmermann 1913; later a class, Schoenichen
placing it instead in Bütschli’s now defunct order Hetero- 1925) or division Euglenophyta with two classes (Pascher
mastigoda together with a much more heterogeneous family 1931). Eventually phycologists settled on spelling the class
Bodonina of heterotrophic flagellates including Bodo, Het- Euglenophyceae (Huber-Pestalozzi 1955; Smith 1933) and
eromita, and four non-euglenozoan genera. For a century consistently using euglenoid in the vernacular (Leedale
classification was dominated by a simpler, less accurate 1967; Triemer and Farmer 2007). Protozoologists in con-
division into just three families (Klebs 1892) that lumped trast long treated all flagellates as just one class, slavishly
Lankester’s four phagotrophic families into one: phototrophs following Bütschli, so ranked euglenoids only as an order
(Euglenida, i.e. not all euglenoids), osmotrophs (Astasi- – spelt Euglenoidina (e.g. Walton 1915 with families
ida) and phagotrophs (Peranemida), as exemplified by the Euglenidae, Astasiidae, Peranemidae) or Euglenida, e.g.
first comprehensive treatise on euglenoids in the mod- Calkins (1926) also with three families, photosynthetic
ern sense (Walton 1915), though some authors had two Euglenidae, phagotrophic Heteronemidae (contrary to Adl
or three photosynthetic families (Hollande 1942; Smith et al. 2005, 2012, Heteronemina Leedale, 1987 was not
1933). Hollande (1942) split euglenoids primarily into three, its first use as a suprageneric taxon), and Astasiidae with
but unlike Klebs grouped osmotrophs (2 families) with a mix of osmotrophs and phagotrophs including both Per-
three families of phototrophs (=Aphagea of Cavalier-Smith anema and Petalomonas – a more heterogeneous mixture
1993) and split phagotrophs into two: Péranémöidinées and than Lankester’s Astasiidae, which excluded Peranemidae
Pétalomonadinées, the latter including Ploeotia, Entosiphon, and Petalomonadidae as separate families. A protozoological
and Anisonema with petalomonads sensu stricto. His later committee (Honigberg et al. 1964) divided order Euglenida
superior system (Hollande 1952) like Lankester’s sepa- into suborders Euglenina, Peranematina and Petalomonad-
rated Petalomonadidae and Anisonemidae (i.e. Anisonema, ina, clearly based on Hollande (1942 not 1952). Given so
Entosiphon) and put Ploeotia (omitted by Lankester and Sav- many past conflicting spellings, the most sensible protistolog-
ille Kent) incertae sedis. ical compromise is to harmonise vernacular and formal terms
In the twentieth century some phycologists initially by uniformly using the oldest vernacular term euglenoid for
ignored euglenoids (West 1904), following Klebs (1883) the whole group and Euglenoida for their subphylum (Table 1
who considered them protozoan flagellate Euglenoidina not and Cavalier-Smith 1978, 1993), as in Ruggiero et al. (2015).
algae. But others embraced them as Euglenineae (Senn Electron microscopy provided germs of a more funda-
1900; later a class: West and Fritsch 1932; Fritsch 1948) or mental subdivision in the six orders or suborders of Leedale
(1967) still extant in some publications, e.g. Leedale (2002) all plants, a criticism also made by Ruggiero et al. (2015).
where it remained deficient in having no families. Leedale Such idiosyncratic ranking gives as much weight to igno-
(1967) retrogressed somewhat by lumping petalomonads and rance as to knowledge. Cladifications can be useful, but have
Anisonema as Sphenomonadina, a new name composition- different aims from and do not replace the function of tra-
ally equivalent to Pétalomonadinées of Hollande (1942), and ditional Linnean classification (Mayr and Bock 2002), i.e.
adopting Heteronematina in an only slightly narrower sense to provide a user-friendly, hierarchical classification of all
than Calkins (1926) extremely heterogeneous Heterone- life with taxa ranked according to phenotypic distinctive-
matidae (by segregating Sphenomonadina). Levine et al. ness and arranged in an evolutionarily sound manner without
(1980, with Leedale coauthor) accepted his six suborders. polyphyletic groups, comprehensively at each rank (Cavalier-
Cavalier-Smith (1993) abandoned Sphenomonadales/-ina Smith 1998; Ruggiero et al. 2015).
as too heterogeneous, using the older name when estab- To this end Table 1 offers the product of a single mind inte-
lishing order Petalomonadida that like Petalomonadidae grating over a century of progress since pioneering Bütschli
sensu Hollande (1952) excluded Ploeotia, Entosiphon, and (1884) and Klebs (1883, 1892); I hope it provides a good
Anisonema, and segregated a new order Ploeotiida from basis for further improvement and debate. I need not defend
Peranemida; for the first time Euglenoida was ranked as a pro- establishing the ancestral superclass Rigimonada or retaining
tozoan subphylum with three classes with contrasting feeding paraphyletic orders Eutreptiida and Bodonida (all flagged in
machinery: Aphagea, and the phagotrophic Petalomonadea Table 1 as paraphyletic to prevent readers misinterpreting
and Peranemea. them as clades), as the standard arguments against ances-
Adl et al. (2005) reverted to Klebs’ (1892) oversimpli- tral (paraphyletic) taxa are unsound and refuted in detail
fied three-group system with no subordinate phagotrophic elsewhere (Cavalier-Smith 2010b). However, those averse
euglenoid suprageneric taxa, but changed two of his names. to such groups have the consolation that my changes make
Instead of adopting his Peranemida for phagotrophs they used Petalomonas, Ploeotia, Heteronema, Ploeotiida, and Perane-
Heteronematina of Leedale (1967) by radically changing its mida no longer paraphyletic on rDNA trees. To provide a
meaning to including all Sphenomonadina (in this broadened hierarchical, ranked reference classification suitable for gen-
sense it ought to have been called Heteronematidae Calkins, eral use (Ruggiero et al. 2015) it is comprehensive for the
1926 even though his group was narrower than theirs and standard categories of subphylum, class, order, and fam-
different from Leedale’s), and followed Busse and Preisfeld ily, and introduces additional intermediate categories only
(2003) in restricting Aphagea to osmotrophs, i.e. a differ- for those subgroups where necessary to subdivide them
ent meaning from the original (Cavalier-Smith 1993); both appropriately. In places this necessarily includes intermediate
changes contravened their assertion that they ‘used the older categories that can be omitted by endusers whose purposes
name that describes each taxon unless its composition was are better served by listing only classes (as in Fig. 2 cen-
substantially modified’. Thus three generations of Society of tre column), only orders (Fig. 2 middle column) or only
Protozoologists/International Society of Protistology com- families within a particular group. Arteficially restricting the
mittee taxonomy (Adl et al. 2005, 2012; Honigberg et al. number of categories used for higher taxa like Entosiphona
1964; Levine et al. 1980) simply returned us to 1892 for having only one genus by avoiding standard intermediate cat-
euglenoids; the Adl et al. (2005) system was not ‘the new egories that some might imagine to be superfluous, would
higher level classification’ it proclaimed. Despite its first sen- have deprived end users of that flexibility and also left some
tence wrongly saying Adl et al. (2005) ‘established name genera unassigned to families, orders or classes, which in
stability’, Adl et al. (2012) without explanation changed cir- itself could be a potential source of ambiguity or confusion
cumscription of their phototroph name (Euglenea in 2005) by that is best avoided. The notion expressed by one referee that
excluding Eutreptiales, and grouped Eutreptiales, Euglenea, one should never place a single genus alone in its own higher
and the photophagotroph Rapaza (Yamaguchi et al. 2012) taxon is irrational and contrary to 300 years of taxonomic
with equal rank within Euglenophyceae. practice.
Unlike earlier systems (Cavalier-Smith 1993, 1998;
Hollande 1952; Honigberg et al. 1964; Leedale 1967, 2002;
Walton 1915) that of Adl et al. (2005, 2012) was not a
balanced comprehensive classification with taxa employing Five Classes and Seven New Orders of
Linnean categories like phylum, class and order; nor was Euglenoids
it a strict cladifaction (Mayr and Bock 2002) as like tra-
ditional taxonomy it accepted some paraphyletic groups, Though one might treat all euglenoids as one class
though seeming to regard their elimination as intrinsically (Cavalier-Smith 2003a,b; Hollande 1952; Smith 1933), gen-
desirable (Adl et al. 2012, p. 430). Using only four ranks eral use of Euglenophyceae as a class for phototrophs only
severely limited its usefulness – it unwisely gave the same (Bicudo and Menezes, 2016; Marin et al. 2003) accepts
rank to single genera like Pelomyxa, Ancyromonas, Gromia that dispensing with the classical distinction between green
or Stephanopogon (previously all placed sensibly by oth- and colourless forms (Leedale 1967) was a mistake, and
ers in higher taxa) as to Euglenozoa, Fungi, all animals or makes thorough reevaluation of phagotrophic euglenoid class
demarcation overdue. Previously euglenoids were divided stricto (here restricted to Ploeotia and Lentomonas) that I
into three classes solely using feeding apparatus (FA) dif- place in a new class Ploeotarea because of their distinc-
ferences, before rDNA was sequenced for phagotrophs, so tive pellicle structure and Ploeotia’s deep separation from
Ploeotiida and Entosiphonida were wrongly included in Per- stavomonads on rDNA trees.
anemea (Cavalier-Smith 1993). The five euglenoid classes Leedale placed Entosiphon in Heteronematina whereas
here, and grouping two as superclass Spirocuta, better express Cavalier-Smith included them in Ploeotiida (1993, 1995
euglenoid megadiversity and phylogeny than the former three implicitly; explicitly in von der Heyden et al. 2004); Adl
(Cavalier-Smith 1993). One, long accepted Euglenophyceae, et al. (2012) left them in Heteronematina but their footnote
is identical to one of Klebs (1892) three groups; the other implied a preference for Ploeotiida. Because of their inordi-
four reflect much improved understanding of phagotrophic nately long-branch for Entosiphon rDNA trees were unable
euglenoids and their relationships in the past 20 years, includ- to establish its correct position, but Hsp90 trees lack that
ing evidence that rhabdomonads are more closely related to long-branch problem and strongly (and site-heterogeneous
some of them than to Euglenophyceae, and a fundamen- rDNA trees weakly) place Entosiphon as sister to all other
tal reevaluation of pellicle and FA comparative anatomy euglenoids (Cavalier-Smith et al., 2016a). Entosiphon does
(Cavalier-Smith unpublished), so are evolutionarily more not group with either Peranema or Serpenomonas as it did
realistic than previous heterotrophic classes. inconclusively on distance rDNA trees (von der Heyden et al.
Adl et al. (2012) did not subclassify euglenoid 2004). This deeply branching position is consistent with its
phagotrophs, but a footnote accepts Petalomonadida as prob- FA having three supportive rods of very distinct structure
ably a clade, but suggests expanding Ploeotiida to include from the two present in Peranemea and Ploeotiida (Triemer
Entosiphon, Lentomonas, and Keelungia with Ploeotia. How- and Farmer 1991a,b) and with major differences in pellicle
ever, putting Entosiphon in Ploeotiida (von der Heyden et al. structure. Therefore as well as removing Entosiphon from
2004) made it too heterogeneous in FA structure for an order Ploeotiida as a separate order I place it in a new infraphy-
and was little improvement on grouping it with Anisonema lum and class (Table 1). That entails removing Entosiphon
(Saville Kent 1880–Saville Kent 18801882; Lankester 1885) from Peranemea. New class Entosiphonea with protrusible
with which it was originally congeneric (Dujardin 1841), so I siphon and three microtubule-bundle support rods empha-
establish a separate order Entosiphonida. Even the remaining sises that its rods differ greatly from those of other euglenoids
Ploeotia-like taxa are not ultrastructurally similar enough for here grouped as new infraphylum Dipilida; I consider that
one order, which would be radically broader than the original this structural dichotomy represents the primary phylogenetic
Ploeotiida. Such a group would be deeply paraphyletic and split within euglenoids.
lack a clear unifying morphology. All recent sequence trees Ribosomal DNA trees also show the original ultrastruc-
show Petalomonadida as a robust clade (Figs 1, S2 and for turally diverse Peranemida to be deeply paraphyletic (Figs
Hsp90 Cavalier-Smith et al. 2016a), confirm the polyphyly of 1, S2 where the former peranemids Peranema, Anisonema,
Sphenomonadina, and show that ‘Ploeotia cf. vitrea’ does not Dinema, Neometanema, Teloprocta are not collectively a
group with Peranemida or Petalomonadida, and thus support clade, but rhabdomonads and Euglenophyceae are indepen-
separateness of the three phagotrophic orders established by dently derived from them) confirming early distance-tree
Cavalier-Smith (1993). Fig. 2 summarises conclusions from evidence (von der Heyden et al. 2004), so I subdivide former
euglenozoan sequence phylogeny at the ordinal level based Peranemida by establishing three new holophyletic orders of
on both Hsp90 (Cavalier-Smith et al. 2016a) and the present contrasting motility within Peranemea: Anisonemida glide
very comprehensive site-heterogeneous 18S rDNA trees. on the posterior cilium not the anterior one as in Perane-
Serpenomonas pellicle strip ultrastructure and morpho- mida sensu stricto; Natomonadida swim instead of gliding
genesis differ so radically from Ploeotia and all other and include ultrastructurally and phylogenetically related
euglenoids, as I shall explain in detail elsewhere, that they phagotrophs (Neometanema) as well as the osmotrophic
should never have been put in the same genus (Farmer and Rhabdomonadina; Acroglissida at present includes only the
Triemer 1988). Serpenomonas (Triemer 1986) deserves its anterior gliding Teloprocta scaphurum (Cavalier-Smith et al.
own order Heterostavida to emphasise its radically different 2016a) that uniquely for Euglenozoa has a cytoproct and was
pellicle morphogenesis (heteromorphic strips, not homomor- formerly wrongly treated as a Heteronema (Breglia et al.
phic ones as in Ploeotiida and most other euglenoids) and 2013). Thus there are now nine orders of non-photosynthetic
phyletic distinctiveness. All trees confirm its extreme diver- euglenoids, each with distinctive ultrastructure and uniform
gence from Ploeotia cf. vitrea (Cavalier-Smith et al. 2016a). motility mode. Each order is rather homogeneous in ultra-
I group Heterostavida with Petalomonadida and Decastavida structure and lifestyle, whereas previously Ploeotiida and
(recently established for Decastava and Keelungia: Cavalier- Peranemida were deeply paraphyletic and ultrastructurally
Smith et al. 2016a) as a new phagoheterotrophic euglenoid extremely diverse. Eight orders are clearly clades on rDNA
class Stavomonadea, ancestrally with a pellicle of 10 rigid trees; Anisonemida is either a clade (Figs 1, S1, S2) or ances-
simple strips and non-protrusible mouthparts, with feed- tral to the closely related Natomonadida.
ing comb and two hollow support rods. The simplicity of Sequence trees show that the difference between anterior
stavomonad strips distinguishes them from Ploeotiida sensu ciliary gliding (Petalomonadida, Peranemida, Acroglissida,
Postgaardida) and posterior ciliary gliding (Entosiphonida, means (see Cavalier-Smith et al. 2016a) make this ordinal
Decastavida, Heterostavida, Ploeotiida, Anisonemida) is evo- name best abandoned. Most described Heteronema species
lutionarily profoundly important. From their distribution on need to be assigned to a new genus or genera when char-
the tree we can conclude that gliding on the posterior cil- acterized by ultrastructure and/or sequencing. Their anterior
ium is the ancestral state for euglenoids. These contrasting gliding and second shorter cilium suggest that most belong
locomotory patterns, which must use different molecular in Acroglissida. Only H. marina, the type, is surely a Het-
motors (kinesin for posterior gliding and dynein for ante- eronema! On Fig. 1 Peranemea and Euglenophyceae together
rior gliding: Cavalier-Smith 2013a, 2014b), appear to have form a very strong (0.99, 100%) clade, here made new
been stable for hundreds of millions of years with rare euglenoid superclass Spirocuta on account of their spirally
switches between them, as no euglenoid orders recognised arranged, often contractile pellicle. The spirocute pellicle
here have a mixture of anterior and posterior ciliary gliders with strips interlocking laterally by complementary hooks
and their divergences are all ancient. Swimming is phyloge- at their heel [the name used by Mignot et al. (1987) for
netically much more restricted in euglenoids, characterising the thickened edge of spirocute strips clearly visble when
only two independently derived clades: Euglenophyceae Euglena is squashed and strips appear as separate laminae
and Natomonadida. Neometanema normally swim like their (Leedale 1966)] and opposite or ‘toe’ edges. This interlocking
rhabdomonad relatives, and should never have been called contrasts sharply with the primitive arrangement of laterally
Heteronema, conforming neither with Dujardin’s original flush-butted, barrel-stave-like strips that form the rigid pelli-
concept (posterior ciliary gliding) nor with Ritter von Stein’s cle of the new ancestral superclass Rigimonada (Ploeotarea,
totally contradictory one (anterior ciliary gliding). The skid- Stavomonadea) as well as in Entosiphonea.
ding mode of swimming close to surfaces of the predatory Here I designate the unthickened edge of spirocute strips
Neometanema can be regarded as the ancestral state for the toe. This new term is desirable because although Mignot
natomonads, evolutionarily and behaviourally intermediate et al. (1987) called the recurved edge of the spirocute toe
between gliding and the fully planktonic swimming of rhab- the hook, Leander and Farmer (2001) confusingly used ‘the
domonads that presumably evolved when these osmotrophs hook’ instead for the raised edge of the heel and adopted
first abandoned phagotrophy. the little-used term ‘overhang’, which originally referred
Rhabdomonadina and Euglenophycidae were once to the overall shape of each Euglena ‘pellicle complex’
grouped together as class Aphagea on the hypothesis of (Sommer and Blum 1965) not specifically to part of a
a common loss of phagotrophy and switch from gliding strip, changing its meaning to refer instead to the hook
to swimming (Cavalier-Smith 1993). However, rDNA trees of Mignot et al. (1987). Retaining Mignot’s older name
(Fig. 1) show that the ancestor of Natomonadida switched ‘heel’ and adopting an unambiguous complementary new
from gliding to swimming independently of Euglenophyceae name ‘toe’ resolves the confusion caused by the now con-
and before Rhabdomonadina lost phagotrophy. As the discov- tradictory double meaning for ‘hook’ and usefully allows
ery of phagophototrophic Rapaza showed (Yamaguchi et al. the same terminology for the contrasting strip edges of
2012), even Euglenophycidae did not switch to swimming all euglenoids, even those of Entosiphon and rigimonads
and lose phagotrophy simultaneously. Swimming evolved where the toe is straight, not hooked. I recommend that the
first in the ancestor of all Euglenophyceae at the time of now ambiguous term ‘hook’ be no longer used to desig-
green algal chloroplast enslavement (Cavalier-Smith 2013b), nate specifically just one strip edge in spirocutes, both of
whereas phagotrophy was lost later in only one subclade which are hooked. Other recent rDNA trees (e.g. Lax and
(Euglenophycidae), being retained by its sister Rapaza. Simpson 2013) also show a sharp bipartition between ances-
Therefore I abandon polyphyletic Aphagea. Unlike Adl et al. tral rigid euglenoids with 12 or fewer flush-butted pellicle
(2005, 2012) and Lee and Simpson (2014a,b), I do not accept strips mostly with straight toes (Entosiphon plus Rigimon-
restricting Aphagea to Rhabdomonadina alone (Busse and ada) and derived classes with 16 or more spiral interlocking
Preisfeld 2003), as sequence trees favour making Anisonemia strips with hooked toes and widespread euglenoid motility
a subclass with two orders of contrasting ciliary/locomotory (Spirocuta).
organisation (Table 1). Following Cavalier-Smith (1993) and Within Euglenophyceae adding a new phagophototrophic
contrary to Adl et al. (2005, 2012), I retain Leedale’s suborder order for Rapaza, yields 12 euglenoid orders overall, not six
Rhabdomonadina (as expanded previously to include Astasia as in Leedale (1967). This increase stems partly from discov-
and Distigma). On its own, loss of phagotrophy is insufficient ery of Rapaza, Serpenomonas, and decastavids, and partly
to merit even ordinal separation; it would be unwarranted rank from ultrastructural and sequencing revelations of radically
inflation to treat both natomonad suborders as a class, which greater phagotrophic euglenoid diversity than appreciated in
accepting class Aphagea for Rhabdomonadina would entail. the 1960s. According to the latest sequence trees (e.g. Fig. 1
Unlike Adl et al. (2005, 2012), I follow Cavalier-Smith and Cavalier-Smith et al. 2016a) only one euglenoid order is
(1993) in rejecting Leedale’s excessively heterogeneous Het- undoubtedly paraphyletic (Eutreptiida); in accepting it Adl
eronematida that embraced genera now included in two et al. (2012) did not realize its probable paraphyly, presum-
separate classes (Peranemea, Entosiphonea). This hetero- ably because other recent sequence trees of Euglenophyceae
geneity and Ritter von Stein’s muddle over what Heteronema included too few outgroups to show that (Marin et al. 2003)
or none at all (Linton et al. 2010), and so may have been laterally and do not overlap. Strip shape varies from gently
incorrectly rooted. undulating to markedly grooved. Feeding apparatus anteri-
Contrary to a mistaken assertion (Adl et al. 2005), cre- orly with a C-shaped cement skeleton with a primary row
ating a higher-rank taxon to include a single lower-ranked of closely spaced microtubules embedded at its inner sur-
taxon is not taxonomically ‘superfluous’. For example, new face; posteriorly to this cement arc three bundles of closely
subclass Rapazia (with at present a single order, family packed microtubule angled rows are added to every other pri-
and genus) allows us (a) to emphasise by its rank that the mary microtubule outside the now U-shaped primary row and
dichotomy between the photophagotroph Rapaza and the its associated cement to make three cytopharyngeal support
non-phagotrophic phototrophs is the most important one rods. Curved anterior supplementary plaque of cement with a
within Euglenophyceae; and (b) by giving equal rank to its thick hinged apical cap, lacking microtubules on its inner face
sister subclass Euglenophycidae to indicate that Eutrepti- lies inside the main U-shaped cement, thickened as a meshlike
ida and Euglenida together are a clade and mutually more scaffold beneath the cap; these inner structures differ greatly
closely related than either is to Rapaza. The oversimplified from the broader, more massive left/dorsal jaw supports of
Adl et al. (2012) system achieves neither benefit; its categori- dipilid euglenoids. Four slightly curved microtubule-attached
cally confusing equal ranking of a single genus Rapaza, order vanes, associated with a fifth reinforced microtubule; all five
Eutreptiales, and quasi-class ‘Euglenea’ (changed in sense microtubules loop over from a shallow side pouch of the
from 2005!) (Adl et al. 2012) is better expressed by ranking ciliary pocket. Unlike Rigimonada, siphon protrusible and
them all as orders. The standard ordinal suffix – ida automat- feeding comb absent. Glide on posterior cilium or swim with
ically travels with the name, telling readers their equal rank anterior cilium using an oar-like beat. Etymology. Named
whenever mentioned, which is not so with their cumbrous after sole included family:
multiple blob ranking, which lacks advantages over standard, New family Entosiphonidae. Diagnosis: as for
broadly understood Linnean categories like order, class, and Entosiphona and Entosiphonida. Type genus Entosiphon
phylum. Using blobs to denote rank, motivated primarily to Ritter von Stein, 1878. Comment: it was essential to
avoid minor suffix changes that some rank changes entail (a remove Entosiphon from Anisonemidae (Lankester 1885),
non-problem) (Adl et al. 2005), did more harm than good. The Peranemea (Cavalier-Smith 1993) or Ploeotiida (von der
square bracket system citing up to four names for each taxon Heyden et al. 2004) into a new higher taxon, as its U-shaped
(Adl et al. 2005) proposed to circumvent the harm done by FA has a more primitive character more easily related to
unwisely abandoning rank-denoting suffixes is cumbersome that of diplonemids than those of dipilid euglenoids and its
in the extreme, a decisive argument against it. protrusion machinery is unrelated to that of Peranemida, as
Taxonomy of photosynthetic non-phagotrophic euglenoids I shall explain in detail elsewhere.
is in a relatively good state with morphological and sequence New infraphylum Dipilida. Diagnosis: Photosynthetic,
evidence well integrated and a proper Linnaean hierarchy osmotrophic or phagotrophic; if phagotrophic ancestrally
of categories (Marin et al. 2003). That for phagotrophic with two, typically hollow cemented feeding apparatus sup-
and osmotrophic euglenoids has lagged considerably. That is port rods (never three), four vanes, and distinct dorsal and
partly because some who have been most active in studying ventral cemented arc-like cytostomal skeletons that face each
them recently have been hostile to classical Linnean taxon- other across a ciliary pocket extension, dorsal arc wider than
omy altogether (as exemplified by Adl et al. 2005, 2012) and ventral. Etymology: di- L two-fold, double; pila L. pillar.
partly because the influential, pioneering system of Leedale New superclass Rigimonada. Diagnosis: Rigid, non-
(1967, 2002), despite being Linnean in character, for simplic- squirming heterotrophic euglenoids with 10 or fewer largely
ity did not attempt to revise or even use families. The present longitudinal pellicular epiplasmic strips; epiplasmic strips
paper is the first to include families for non-photosynthetic abut laterally and do not overlap and interlock laterally as
and phagotrophic euglenoids since Vasileva (1987) who had in Spirocuta; minimally two closely interlinked microtubules
three phagotrophic families (Peranemataceae, Petalomon- attached to strip heel. Ancestrally with non-protrusible feed-
adaceae and Scytomonadaceae as in Mignot 1967), and two ing apparatus rods, without central microtubules or with one
osmotrophic families (Astasiaceae and Menoidiaceae) for the or a few short microtubule rows not in hexagonal array
other osmotrophs. Cavalier-Smith et al. (2016a) provided the (unlike Peranemea). Rod-vane complex extends almost the
first sequence data for a Scytomonas, showing that a separate whole cell length, unlike Peranemea, but cement and rods lost
family Petalomonadidae is unjustified. in petalomonads. Mostly aerobic phagotrophs. Etymology:
Diagnoses of the new euglenoid taxa follow in Table 1 rigeo L. I am stiff; monas Gk unit.
order. FA and pellicle structure terminology reflects the New class Stavomonadea. Diagnosis: species with
revised interpretations of homologies that I shall explain, cemented mouthparts have a double incurved C-shaped dense
illustrate, and reference in more detail in a separate paper. cement layer supporting its dorsal jaw, whose continuous
New infraphylum Entosiphona, class Entosiphonea and outer lip has an associated arc of two rows of closely linked
Entosiphonida ord. n. Cavalier-Smith. Diagnosis: Biciliate microtubules on its inner face (the outer arc with numer-
bacterivorous euglenoids with fluted cell surface; pellicle has ous projections facing the inner arc) and a third layer of
4, 8, 10 or 12 longitudinal strips, whose epiplasmic layers abut widely spaced microtubules on the side facing the cytophar-
ynx, which consists of five reinforced microtubules plus one used to group them are arguably either plesiomorphies for
to many pairs of unreinforced microtubules, each pair having all rigimonads or more likely convergent. rDNA trees show
a connecting dense lamina; the fifth reinforced microtubule is that Serpenomonas and P. cf. vitrea do not group together
connected to the first pair by a longer lamina wrapped round and are as far apart on the tree as any two dipilid euglenoid
a cement spur from the inner left lip. The reinforced micro- genera can possibly be, whereas Serpenomonas is sister to
tubules loop over from the left/dorsal jaw support to the other Decastavida (Fig. 1). More importantly, its strip heteromor-
side of the cytostome, where four bear vanes except in most phism is unique and not present in Ploeotia or as markedly in
petalomonads; the adjacent unreinforced microtubule pairs any other euglenoids; and it lacks Ploeotia’s ventral groove.
similarly loop over in parallel. FA rods hollow, with few or no I do not accept that Serpenomonas has five vanes (Linton
internal microtubules; rods and cement lost in petalomonads. and Triemer 1999); as I shall explain more fully elsewhere,
Usually glide on posterior or anterior cilium rather than swim. it has six, shown most clearly by Belhadri et al. (1992): four
Strip junctions not located just below the crest of narrow standard curved vanes with a microtubule along one edge,
prominent ridges, unlike Ploeotarea. Unlike Spirocuta and plus two additional strongly folded (one surrounding a pro-
Entosiphon, cytostome separate from and ventral to reservoir jecting ridge from each rod) that lack edge microtubules and
canal, except in a few smaller petalomonads (Petalomonas might arise by distal splitting of two standard microtubule-
mediocanellata, Scytomonas). Etymology: stave E. from the borne vanes. Linton and Triemer’s (1999) description of the
resemblance of the strips to barrel staves; monas Gk unit. mouthparts is not fully self-consistent: the structure labeled
New subclass Homostavia. Diagnosis: 10 or fewer mor- DL in their Fig. 5 is consistent with their text ‘the anterior-
phologically similar, subequal stave-like pellicular strips, most portion of the comb . . . formed what was seen externally
lacking deep troughs, never with strong alternating hetero- as the dorsal lip’, but DL in Figs 7, 17 instead label a dorsal
morphism as in Serpenomonas. Etymol: homos Gk same; extension of the ventral lip cement, not the comb at all; this
stave E. because all strips have the same barrel stave, non- dorsal extension may correspond to the ventral part of the
trough-like morphology. accessory lip structure present but much less prominently in
New family Keelungiidae. Diagnosis: With only one pair Keelungia (Chan et al. 2013).
of unreinforced microtubules looping from dorsal jaw sup- New class Ploeotarea. Diagnosis: Phagotrophic bicili-
port to cytostome; outer rod unflanged. Type genus Keelungia ate heterotrophic euglenoids; two cemented pharyngeal rods
Chan et al. 2013. with homogeneous dense matrix, unlike Entosiphonida and
New subclass Heterostavia. Diagnosis as for sole order Spirocuta never with hexagonal-array microtubules. 10 mor-
Heterostavida: phologically similar pellicular strips; joints with rounded
New order Heterostavida. Diagnosis as for sole family overhangs and strongly inrolled heel region, making 10
Serpenomonadidae: microgrooves below the crest of 10 longitudinal ridges. Glide
New family Serpenomonadidae. Diagnosis: 10 pellicu- on posterior cilium. Etymol: ploion Gk boat, from boat-like
lar strips symmetrically arranged but dimorphic in width and cell shape.
form, unlike Decastavida and Ploeotiida: five broad alternat- New family Ploeotiidae. Diagnosis: Biciliate bacteri-
ing with five very narrow and concave strips form five deep vorous rigimonads with 10 longitudinal pellicle strips. Cell
grooves underlying five, barely projecting, ridges; outermost in cross section rounded, with 10 narrow unequally termi-
part of ridges regularly crenate. Strip joints not on strongly nally bifurcate, non-crenate ridges, each with a strip joint
projecting narrow ridges. Feeding apparatus with two dense just below its crest. One ventral strip narrower than others
hollow lateral rods with no internal microtubules and only forms base of a central ventral posterior ciliary groove, unlike
a surface row facing the vanes; in posterior regions support Lentomonas. Glide on posterior cilium. Feeding apparatus
rods have six prominent ribs (three each) associated with four with two oval hollow cement rods without obvious inter-
unfolded vanes with one edge attached to a microtubule and nal microtubules; posteriorly vanes associate with four rod
two folded vanes with no edge microtubules; cytostome dor- cement ridges. Type genus Ploeotia Dujardin, 1841. Com-
sal to anterior cilium, with prominent dorsal jaw-supporting, ment. Most nominal ‘Ploeotia’ are probably wrongly in the
cemented doubly crescentic comb, its outer lip with two genus as they lack 10 prominent narrow ridges. Many are
close-packed layers of microtubules; 12 pairs of unreinforced more similar to Serpenomonas costata, as Lax and Simpson
microtubules loop over in parallel with five reinforced micro- (2013) noted; others may not belong to either genus. From its
tubules from comb to cytopharynx; narrower but thicker morphology, sequenced Ploeotia cf. vitrea (Lax and Simpson
domed ventral jaw-support connects cement rods apically. 2013) probably belongs in Ploeotiidae, but is much too
Unlike Ploeotia, there is no ventral ciliary groove delimited large to be P. vitrea; it might not even be a Ploeotia as the
by two raised ridges. Type genus Serpenomonas Triemer, dorsoventral arrangement of ridges differs from P. vitrea of
1986. Etymol: hetero Gk different; stave E. because the strips Farmer and Triemer (1988; not necessarily from Dujardin’s),
are heteromorphic, only alternate ones resembling barrel being more like Lentomonas. The assertion that P. vitrea
staves, the others being deeply trough-like. Comment. Trans- and Serpenomonas costata FA are ‘almost identical’ was
ferring Serpenomonas to Ploeotia (Farmer and Triemer 1988) ill-documented (Farmer and Triemer 1988); no micrographs
was not justified; the asymmetric bifurcate pellicle ridges were shown of P. vitrea jaw supports, so their properties must
be omitted from the diagnosis; their Fig. 9 is so fuzzy that I here emend this class by excluding Entosiphon, Ser-
one cannot even count vanes or decide which are folded or penomonas, and Ploeotiida and by adding Rhabdomonadina,
have edge mts, and they cannot be clearly distinguished from which makes it more homogeneous in pellicle and FA struc-
mt rows; there appear to be four posterior rod cement ridges ture if one allows for obviously secondary FA simplifications
not six as in Serpenomonas, but this region also differs from in the osmotrophs that as losses do not deserve class-level
Lentomonas. separation. The four orders are grouped into three subclasses,
New family Lentomonadidae. Diagnosis: Biciliate bac- each a clade on rDNA trees and each homogeneous with
terivorous rigimonads; 10 subequal pellicle strips; strip heels respect to locomotory mode and FA features.
strongly recurved, making 10 microgrooves; dorsoventrally Subclass Anisonemia, the most diverse heterotrophic
differentiated, three ventral strips, making flat surface, 7 spirocute clade, comprises two new orders with distinct
dorso-lateral strips strongly curved, making 7 broad ridges, locomotory modes, each a well-supported clade. Order
with strip joints just below prominent longitudinal ridge Anisonemida comprise biciliate phagotrophic spirocute
crests; no obvious ventral groove or ridges. Left jaw sup- euglenoids that invariably glide on their posterior cilium,
port with one arc of close-packed microtubules bearing in marked contrast to the non-sister subclasses Peranemia
slanted pairs of unreinforced microtubules; several widely and Acroglissia that have a similar pellicle but glide on
spaced reinforced microtubules associated with ciliary pocket their anterior cilium. Anisonemida do not group with Per-
extension between left and right jaw supports. Cytostome anema, being usually sister (Figs S1, S2, 1 ML; but with
separate from ciliary reservoir, no common vestibulum. Type outgroups restricted as in Fig. 1 CAT without support put
genus Lentomonas Farmer and Triemer, 1994. Comment. L. them ancestral) to a consistently well supported clade of
applanatum had one narrow lateral ventral strip and three swimming heterotrophic spirocutes (new order Natomona-
broad ones, but no ventral groove or ridges (Farmer and dida) and have nonprotrusible FA. Natomonadida comprise
Triemer 1994), whereas Entosiphon applanatum (Preisig the purely osmotrophic suborder Rhabdomonadina and new
1979) had a central narrow ventral groove bounded by two phagotrophic suborder Metanemina (Neometanema).
ridges. This implies a very different ventral pellicle struc- Subclass Acroglissia (order Acroglissida) contains only
ture, so they are not the same species. I agree with Ekebom the anterior ciliary glider Teloprocta scaphurum (formerly
et al. (1996) and Patterson and Simpson (1996) that ultra- lumped in Heteronema: Skuja 1932; Breglia et al. 2013;
structurally studied L. applanatum and Ploeotia corrugata Schroeckh et al. 2003), which unlike Peranemida sensu
Larsen and Patterson (1990) are the same species, and with stricto has a non-protrusible rod apparatus and a cytoproct.
Linton and Triemer (2001) that they are not the same genus Though Neometanema and Teloprocta are both phagotrophic
as S. costata. I think they are also not congeneric with spirocutes with simplified cytostomal supports, they never
P. vitrea, so transfer P. corrugata and similar P. azurina group together (despite formerly being lumped in one exces-
to Lentomonas. New combinations: Lentomonas corru- sively wide genus Heteronema); as no special morphological
gata comb. nov. Basionym Ploeotia corrugata Larsen and features unite them, Cavalier-Smith et al. (2016a) made Telo-
Patterson, 1990 p. 867; synonym Lentomonas applanatum procta a separate genus. Most trees put it weakly sister to
Farmer and Triemer (1994), but not Entosiphon applana- Euglenophyceae (Lax and Simpson 2013; Lee and Simpson
tum Preisig, 1979. Lentomonas azurina comb. nov. Basionym 2014a,b; my Fig. 1); however three of my 16 best aligned and
Ploeotia azurina Patterson and Simpson, 1996 p. 432. I based most comprehensively sampled trees group it instead with
the family diagnosis entirely on Farmer and Triemer (1994), Anisonemia (Table 3), including Fig. S2 which appeared to
who had no light micrographs to support synonymy with E. be the most reliable for 18S rDNA judging by its congruence
applanatum, nor even said if FA is visible in the light micro- with Hsp90, making it possible that this minority position
scope as in P. corrugata and E. applanatum (as Patterson and is correct and Acroglissida plus Anisonemida are really a
Simpson (1996) stressed, it had a protrusible siphon unlike clade that is sister to Euglenophyceae. Highest support for an
P. corrugata). Acroglissida/Anisonemida clade (0.75) was for the largest
New superclass Spirocuta. Diagnosis: Pellicle of 16–56 taxon sample (323 including Entosiphon) and largest num-
ancestrally spirally arranged narrow strips; ancestrally ber of included nucleotides (1577), but it was still supported
squirm by lateral sliding of strips; ancestrally phagotrophic (0.66) for 1577 nucleotides when Entosiphon was removed.
feeders on eukaryotes with four unfolded vanes and two oval Suppport for the competing grouping with Euglenophyceae
cytopharyngeal supporting rods with core of close-packed was weak. Its position was unresolved in Breglia et al. (2013)
microtubules, typically in hexagonal array. Cytostome opens and should become more stable when further species are
into upper part of reservoir (in the common vestibulum) sequenced. As explained by Cavalier-Smith et al. (2016a), the
unlike most rigimonads where it is separate from the reservoir original Heteronema marina of Dujardin (1841) had a thicker
canal as in Glycomonada. Includes secondary osmotrophs trailing cilium and thinner undulating anterior-pointing one,
and phototrophs with reduced or absent mouthparts. Etymol: thus was quite similar to his Anisonema though he put them
Spira L. coil, spire; cutis L. skin. in separate families. Ritter von Stein (1878) caused much
Class Peranemea Cavalier-Smith, 1993 confusion by ignoring Dujardin’s species and applying the
same generic name to two radically different euglenoids that
glide instead on their anterior cilium: Peranema globulosa New family Neometanemidae. Diagnosis: flattened
Dujardin 1841 (one cilium only seen) and Astasia acus Ehren- broadly ellipsoidal biciliate phagotrophs that swim, skidding
berg, 1838. close to substrates, but do not normally glide on either cilium;
Subclass Peranemia comprises classical Peranemidae, the cilia of equal thickness and length, anterior lashing, posterior
only family now remaining in order Peranemida (Table 1): trailing. Feeding apparatus visible or not by light microscopy.
New subclass Peranemia. Diagnosis: Squirming Weak to moderate squirming. Type genus Neometanema Lee
phagotrophic Spirocuta that glide on anterior cilium that and Simpson, 2014a,b.
is rigid except for tip flickering; protrusible FA of two New subclass Acroglissia and new order Acroglissida.
pharyngeal rods, with hexagonally or irregularly arranged Diagnosis: biciliate phagoheterotrophic spirocutes; glide on
microtubule core surrounded by cortex of microtubule-free proximally rigid anterior-directed cilium; weak to moderate
dense cement, anteriorly closely linked by cemented con- squirming; two non-protrusible FA rods visible in light micro-
nector supporting dorsal jaw; additional microtubules face scope. Etymol: acro Gk topmost, glisser Fr. slide, as the
vanes and sometimes (Urceolus) also surround cortex as a gliding cilium projects forwards from the cell tip. Sole genus
thin 1-mt thick peripheral ‘skin’; rods each anteriorly with a Teloprocta Cavalier-Smith in Cavalier-Smith et al. 2016a.
long oblique, robust, lateral cemented strut (inner and outer Historical Note: Separation here of heterotrophic
laminas) linked by long striated fibres to each end of outer euglenoids into four ultrastructurally and phyletically dis-
part of strongly cemented double dorsal/left jaw support tinct classes shows how much wiser Lankester (1885) was in
(LJS) and in this region only an outer groove. LJS with separating them into four separate families than was Klebs
lateral dense bodies; outer one linked by cemented rod-like (1892) who lumped them into just one, an oversimplifi-
anchor to reservoir canal peripheral cement support. Jaw cation that so many since, up to and including Adl et al.
supports separated at outer edge and centrally by a ciliary (2005, 2012), have essentially followed. Allowing for genera
pocket extension across which reinforced microtubules then unknown, Peranemidae is compositionally identical in
and a row of unreinforced microtubule pairs loop from Lankester’s and Table 1; his Petalomonadidae, a junior syn-
LJS to cytopharynx. Four slightly curved (Peranema) or onym by a year of Scytomonadidae, is also compositionally
folded (Urceolus) vanes edged by reinforced microtubules. homogeneous, but unlike Scytomonadidae now excluded the
Posterior cilium not visible by LM: absent (Urceolus, biciliate Tropidoscyphus. His other two families are more
Jenningsia) or attached laterally to body within specialized heterogeneous, but before electron microscopy the FA con-
groove, so invisible in LM (Peranema). trasts of Entosiphon and Anisomema could hardly have been
New subclass Anisonemia. Diagnosis: Ancestrally forseen; nor could those amongst genera in his Astasiidae;
biciliate squirming non-photosynthetic spirocutes; FA of his Menoidina is equivalent to Rhabdomonadina except for
phagotrophs with two anteriorly linked rods of hexago- excluding Astasia.
nally close packed microtubules, peripheral ones embedded Class Euglenophyceae – the only algal class of Eugleno-
in variably developed dense cement, without a thick zoa: chloroplast with triple envelope
microtubule-free cortex (unlike Peranemia); non-protrusible, New subclass Rapazia and new order Rapazida.
inner and outer laminas and linked striated fibres absent or Diagnosis: phagotrophic photosynthetic eukaryovorous
vestigial. LJS often less strongly cemented than in Peranemia, euglenoids with no rod/vane feeding apparatus; simple
without cement anchor to canal cement; rods and vanes lost reservoir-associated feeding pocket supported by a bundle
by non-phagotrophic Rhabdomonadina, but reservoir canal of four rows of close-spaced microtubules, not just one
peripheral fibrous supports well developed. Posterior cilium row as in Euglenophycidae; non-gliding swimmers in the
lost by Astasiidae. water column. Etymol: rapax L. seizing, grasping, as the
New order Anisonemida. Diagnosis: Typically non- only predatory Euglenophyceae. Comment: I consider that
swimming phagotrophic biciliates with spiral pellicular strips homologies of the Rapaza FA and the MTR pockets of
and lashing anterior cilium that glide on prominently thick- Euglenophycidae (Shin et al. 2002) compared with other
ened posterior cilium. Etymol: aniso Gk unequal; nema Euglenozoa were previously partially misinterpreted and will
thread, because of unequal ciliary morphology and function present a new synthesis elsewhere.
and because it includes Anisonema. New family Rapazidae. Diagnosis: as for Rapazida, plus
New order Natomonadida. Diagnosis: Heterotrophic pellicle of 16 strips and one posterior whorl, and feed on
swimming, non-gliding, biciliates with spiral pellicular eukaryote algae. Type genus Rapaza Yamaguchi et al. (2012).
strips. Apical end of vestibulum surrounded by a scroll-like
structure consisting of microtubules and dense fibrous ele-
ments. Etymol: nato L. I swim, monas Gk unit. New subphylum Postgaardia
New suborder Metanemina. Diagnosis: phago-
heterotrophic spirocutes with two swimming cilia of equal Reconstructions of FA ultrastructure in Postgaardi
thickness. Non-gliding skidding motility close to surfaces (Simpson et al. 1996/97; Yubuki et al. 2013) and Calkinsia
driven by curved anterior-directed cilium. Etymol: meta Gk (Yubuki et al. 2009) confirmed that they are fundamentally
after, nema Gk thread. Non-typified name. similar and deserve to be classified together as a distinct
order Postgaardida (Cavalier-Smith 2003a) and class Post- dom artifacts of including too few nucleotide positions. Two
gaardea (Cavalier-Smith 1998), as both genera uniquely share other even less well resolved trees were more grossly mislead-
six finger-like projections inside the cytopharynx mouth ing: in Breglia et al. (2010) diplonemids, kinetoplastids, and
with identical underlying fibrous and microtubular skele- postgaardians all wrongly nested within euglenoids, group-
ton. Cavalier-Smith (2003a,b) had formally placed Calkinsia ing with spirocutes as a collapsed tetrafurcation, whereas in
within Postgaardea and Postgaardida, but Yubuki et al. (2009) Chan et al. (2013) (whose alignment was very inaccurate)
did not accept that these two genera were related and proposed both diplonemids and postgaardians were wrongly within
a new clade name Symbiontida for Calkinsia plus unspeci- euglenoids. Trees of Lee and Simpson (2014a,b) without
fied related anaerobic Euglenozoa with epibiotic bacteria and glycomonads or non-euglenozoan outgroups were arbitrarily
some environmental rDNA sequences; oddly they did not rooted to make it wrongly appear that postgaardians are
include Postgaardi in Symbiontida, but confusingly incor- nested within euglenoids. Adl et al. (2012) more reasonably
rectly regarded the whole class Postgaardea as a synonym treated Symbiontida as a fourth euglenozoan group of equal
for the species Postgaardi mariagerensis alone. Breglia et al. rank with Euglenoida (unwisely called Euglenida: see above),
(2010) discovered a substantially different third postgaardean Diplonemea, and Kinetoplastea.
genus Bihospites whose FA can be more readily homologised Given my much more robust and consistent rDNA trees
with that of diplonemids than with that of euglenoids, con- and reinterpretation of their ultrastructural homologies, I
trary to their interpretations (Cavalier-Smith unpublished). still firmly exclude Postgaardea from Euglenoida, and rank
Yubuki et al. (2013) conjectured that the environmental DNA postgaardeans (=Symbiontida), Euglenoida, and Glycomon-
clade that is sister to Calkinsia in Fig. 1 is Postgaardi, thereby ada, which collectively embrace all euglenozoan diversity,
(like Adl et al. 2012) accepting that Postgaardi belongs in the as subphyla in my revised higher classification of Eugleno-
same clade as Calkinsia and Bihospites, for which the old- zoa (Table 1). This ranks equally the three deepest branching
est name is Postgaardea, but overlooked that Cavalier-Smith euglenozoan clades, whose exact branching order is still
(2003a,b) had already placed Calkinsia in Postgaardea and uncertain (Figs 1, S1, S2); each has radically distinct variants
order Postgaardida; their Fig. 24 effectively made the clade of the basic euglenozoan body plan.
name Symbiontida a junior synonym of both Postgaardea and New subphylum Postgaardia. Diagnosis: Biciliate
Postgaardida. Initially I intended to adopt Symbiontida as the free-living anaerobes covered with epibiotic bacteria in
subphylum name, to validate it as a ranked taxon and thereby longitudinal rows. Highly contractile pellicle underlain by
facilitate its continued use even though when first published numerous equally spaced microtubules without specially
I considered it superfluous, but a referee requested that I use differentiated morphogenetic pairs. Simplified cytopharynx
a new name instead. without cytostomal or reservoir encircling fibres, cemented
Breglia et al. (2010) mistakenly considered the radi- jaw supports or rigid longitudinal straight cemented rods.
cally novel ‘rod apparatus’ of Bihospites as homologues Etymology: based on on sole included class Postgaardea.
of euglenoid rods and overlooked their greater ultra- Class Postgaardea Cavalier-Smith, 1998. Revised diag-
structural similarities with diplonemid FA structures; I nosis: Biciliate heterotrophic anaerobes with well developed
explain elsewhere my view that they represent hypertrophied heteromorphic latticed paraxonemal rods and rows of
diplonemid-like ultrastructural features (Cavalier-Smith epibiotic bacteria, and simplified FA. Centrioles parallel, con-
unpublished). The postgaardean FA is not specifically nected by two dissimilar striated roots, thinner on cell’s right
euglenoid in character and I consider that the S-shaped associated with the intermediate centriolar root; subapical.
pellicle units of Bihospites were misinterpreted as euglenoid- Cytostome dorsal, not ventral as in euglenoids. Corset of
like. Here I establish a new postgaardean order Bihospitida longitudinal pellicle microtubules, evenly and closely spaced
for these highly distinctive flagellates, which makes Sym- with frequent cross linkers and direct links to plasma mem-
biontida a junior synonym only of Postgaardea. Presumably brane, apart from interruption by tubular extrusome docking
partly because of this ultrastructural misinterpretation and sites; extending over the whole cell surface a short way into
partly because of poorly sampled site-homogeneous 18S the neck (canal) of the reservoir as a dorsal row of micro-
rDNA trees that sometimes put postgaardeans weakly within tubules linked to the dorsal centriolar root by a dense dorsal
euglenoids (e.g. Yamaguchi et al. 2012; Yubuki et al. 2009), amorphous fibre. Unlike euglenoids without discrete longitu-
Lax and Simpson (2013), Lee and Simpson (2014a,b), and dinal epiplasmic strips. Mitochondria without obvious cristae
Chan et al. (2015) all controversially treat Postgaardea as or kinetoplasts.
euglenoids and do not accept a separate class. However, Order Postgaardida Cavalier-Smith, 2003a,b. Revised
my comprehensive site-heterogeneous trees all exclude Post- diagnosis: Microtubule band of four or about sixteen micro-
gaardea/Symbiontida from Euglenoida (robustly in Figs 1, S1 tubules loops over from ciliary pocket to cytopharynx dorsal
more weakly in Fig. S2), as do all but one of my ML trees, and margin in a plane orthogonal to five reinforced microtubules
weakly suggest that Postgaardea are sisters of Euglenoida, not that loop over apically from ciliary pocket to cytopharynx via
Glycomonada. Thus their insignificantly supported intrusion six finger-like projections inside mouth of cytopharynx; no
into euglenoids in Yubuki et al. (2009), Yamaguchi et al. supporting rods or dorsal/left jaw support homologues. Rigid
(2012), and Lax and Simpson (2013) are most likely ran-
cells with homogeneous pellicle with a dense thin (∼25 nm) proposals now appear to be erroneous, as I argue else-
lamina underlying microtubules; cytostome ventral groove. where that the four euglenoid vanes were not ancestrally
New family Calkinsiidae. Diagnosis: Phagotrotrophs plicate and that diplonemids do not have five vanes, three
with MTR pocket cytopharynx that eat diatoms and bacte- plicate, as originally assumed, but eight vanes of which
ria; cytostome opens on right of dorsal cilium into a shared six are paired, these pairs having been misinterpreted as
apical depression (vestibulum) close to ciliary pocket; the folded vanes. Furthermore, the previously misinterpreted
cytopharyngeal loop has about 16 microtubules with dense ‘accessory rod’ of the postgaardean Bihospites is probably
flanges on the inner side of the loop at the cytopharyngeal an 8-fold multiplication of diplonemid-like paired vanes,
end. Glide on largely rigid anterior cilium. Type and sole implying that such paired (not-plicate) vanes were ances-
genus Calkinsia Lackey, 1960. tral to all Euglenozoa (Cavalier-Smith unpublished). Though
New family Postgaardidae. Diagnosis: Heterotrophs Plicostoma (euglenoids plus Diplonemea) was a weakly
with cytopharynx opening separately from ciliary pocket; supported clade on many 18S rDNA trees (e.g. von der
5 reinforced microtubules (MTR) loop from ciliary pocket, Heyden et al. 2004), phylogenetic trees for several pro-
along right side of U-shaped gutter oriented posteriorly from teins showed this group to be paraphyletic; instead classes
the cytostome, turn anteriorly along its left side halfway back Kinetoplastea (trypanosomatids and their free-living bodonid
towards cilia, then loop inwards and posteriorly alongside and prokinetoplastid relatives) and Diplonemea were sister
cytopharynx that opens into left gutter; an orthogonal loop- groups (Simpson et al. 2006; Simpson and Roger 2004).
ing band of four non-flanged microtubules passes from ciliary Recent multigene trees strongly support holophyly of the
pocket end of MTR along the anterior half of the left gutter; Diplonemea/Kinetoplastea clade (Cavalier-Smith et al. 2014,
gutter covered by two overlapping longitudinal flaps, inner 2015a,b, 2016b), which is also now consistently strongly
reinforced ridge on left and anterior lip on right. Swim, not supported by the most comprehensive site-heterogeneous
glide. Type and sole genus Postgaardi Fenchel et al., 1995. 18S rDNA trees published here as well as by Hsp90
New order Bihospitida. Diagnosis: as for new family trees (Cavalier-Smith et al. 2016a). I call the Diplone-
Bihospitidae. Diagnosis: Metabolic anaerobic bacterivorous mea/Kinetoplastea clade Glycomonada because uniquely
phagotrophs with pellicle longitudinally subdivided into among eukaryotes peroxisomes were modified in their last
extrusome-delimited S-shaped regions formed by grooves common ancestor to form glycosomes containing glycolytic
containing epibiotic bacteria and ridges with underlying enzymes that are present only in the cytosol in other lineages
mitochondria; cytostome opens on right of dorsal cilium (Makiuchi et al. 2011; Gualdrón-López et al. 2012).
into shared apical depression (vestibulum) close to ciliary The seemingly simple vaneless mouthparts of Postgaardi
pocket. Strongly curved C-shaped rod apparatus originates and most kinetoplastids are almost certainly independent
at vestibulum, loops round nucleus within a nuclear envelope simplifications of a vaned ancestor, analogous to those
groove; nucleus-attached ‘main rod’ comprises a dense lam- that occurred independently in Euglenophycidae and most
ina (immediately adjacent to adhering accessory rod) bearing petalomonads (Cavalier-Smith et al. 2016a). Figs 1, S1, S2
a stack of ∼75 broad, parallel, laminas; ‘accessory rod’ is robustly confirm that Plicostoma are paraphyletic and provide
a membrane-associated row of ∼40 reinforced microtubules the first rooted euglenozoan trees to unambiguously show
with proximal dense flanges and distal paired vanes like those that Saccostoma is polyphyletic; as both were also based on
of diplonemids, initiated at a cytostomal funnel at the ciliary ultrastructural/evolutionary misinterpretations, they are dis-
pocket and looping over to cytopharynx; an adjacent bundle continued as taxa [in agreement with Ruggiero et al. (2015)].
of non-reinforced microtubule pairs at the ciliary pocket is As diplonemids and kinetoplastids clearly have the same fun-
not part of the ‘accessory rod’ Type genus Bihospites Breglia damental body plan, I make Glycomonada a new subphylum
et al., 2010. of Euglenozoa:
New subphylum Glycomonada. Diagnosis: Het-
erotrophic Euglenozoa without pellicular strips; microbodies
are glycosomes containing glycolytic enzymes, not per-
Subphyla Plicostoma and Saccostoma oxisomes. Mitochondria ancestrally polykinetoplastid;
Abandoned mitochondrial genomes of multiple heterogeneous cir-
cles; transcripts undergo RNA editing including uridine
Larsen and Patterson (1990) grouped diplonemids and insertion. Ciliary pocket without dorsal row microtubules,
euglenoids as ‘plicostome Euglenozoa’ because they consid- unlike Euglenoida and Postgaardia; pellicle microtubules
ered the ‘plicate’ (folded) vanes of Diplonema, Ploeotia and ancestrally in a continuous cross-linked corset, nucleated
Serpenomonas mouthparts homologous and a synapomorphy posteriorly; cytostome ancestrally present at tip of pro-
absent in kinetoplastids. Cavalier-Smith (1998) formalised nounced apical rostrum separated from ciliary pocket by
this group as subphylum Plicostoma and erected a com- narrow preoral crest; feeding apparatus ancestrally with
plementary subphylum Saccostoma for Kinetoplastea and left and right ‘jaw-bone’-like cement lip supports, each
Postgaardea, assuming that their simpler mouthparts with- associated with short lateral dense rods that extend only
out vanes were primitive. The assumptions behind these a short way from cytostome; ancestrally with hairs on
preoral crest, and circumferential encircling microtubules rods; differ from Diplonemidae in having mitochondria with
surrounding cytostome; three microtubule sets (nucle- giant flat cristae and dispersed kinetoplast nodules and two
ated in ciliary pocket) loop over to support the left side long cilia in trophic cells, ciliary pocket extension very
of cytopharnx: 4–8 central widely spaced microtubules shallow, and having smooth cortical alveoli. Paraxonemal
reinforced by characteristic dense material (MTR) flanked rods and tubular extrusomes prominent. Peripheral lacunae
by close-set less reinforced microtubules on dorsal (parallel in cytoplasm. Type genus Hemistasia Griessmann (1913).
microtubule loop (PML) microtubules, mostly paired with Comment: Transferring Phyllomitus amylophagus to Hemis-
intrapair thin laminas) and ventral external microtubule band tasia (Lee 2002) was wrong if Mylnikov et al. (1998)
(EMB) sides; ancestrally MTR microtubules had flanges correctly identified their strain that did not have cortical
on inner face of loop and vanes on cytostome-associated alveoli or a well developed set of encircling microtubules;
outer part, secondarily lost in most kinetoplastids. Etym: this species is probably a neobodonid deserving a new genus;
glycys Gk sweet, monas Gk unit, because of glycosomes. I agree with Lee (2002) that it is not Phyllomitus and that
Comment: 187-8-gene trees maximally support (by CAT genus must be restricted to marine P. undulans Stein. As Phyl-
and ML) Diplonemea plus Kinetoplastea being a clade lomitus is unassigned to a suprageneric taxon I now formally
(Cavalier-Smith et al. 2014, 2015a,b, 2016b), and they classify it within phylum Cercozoa, Class Imbricatea, order
share a unique mitochondrial genome structure of multiple Marimonadida that contains the only other known genera
heterogeneous circles (Marande et al. 2005; Roy et al. with mutually adhering posterior-pointing cilia: Auranti-
2007). Though mitochondrial genomes of euglenoids are cordis with four; Rhabdamoeba with two as in Phyllomitus
also peculiar and poorly characterised compared with (Howe et al. 2011). Abollifer with parallel centrioles and two
non-Euglenozoa, in Petalomonas cantuscygni at least they non-adhering cilia shown to be a marimonad by sequencing
appear predominantly linear (Roy et al. 2007). Euglena (Shiratori et al. 2014) has a similar apical ventral pit or short
gracilis mitochondrial genomes comprise a pool of het- groove to Phyllomitus and likewise is highly flexible. Pseudo-
erogeneous DNA molecules encoding fewer proteins than phyllomitus with four species, perhaps not a coherent genus
the 12 in kinetoplastids – only seven, whose transcripts do (Lee 2002), might in part belong to diplonemids or (more
not undergo RNA editing (Dobáková et al. 2015). RNA likely) Cercozoa or elsewhere, but must be left incertae sedis
editing by uridine insertional and/or deletion is therefore a in no phylum. ‘Phyllomitus apiculatus’ ultrastructurally stud-
derived shared character of glycomonads, not ancestral for ied by Mylnikov (1986) was misidentified and probably an
Euglenozoa whose last common ancestor clearly reduced undescribed relative of Rhynchobodo.
mitochondrially encoded protein numbers compared with
excavate protozoa, but retained 43 proteins from their
bacterial ancestors that were lost by excavates and their
higher eukaryote descendants. Cytopharynx and associated Class Kinetoplastea: Three New Bodonid
mts lost by some trypanosomatids. Families
Multigene trees strongly confirm that Prokinetoplastina
Class Diplonemea, Order Diplonemida (e.g. Ichthyobodo: Joyon and Lom 1969) and Metakineto-
plastina are deeply divergent sister clades (Cavalier-Smith
Montegut-Felkner and Triemer (1994) made the first et al. 2016b). Bodonida, though paraphyletic (Cavalier-Smith
three dimensional reconstruction of a diplonemid FA. Until et al. 2014, 2015a,b, 2016b; Deschamps et al. 2011), are
recently the taxonomic position of Hemistasia was unknown. kept as an order, neobodonids, parabodonids, and eubodonids
Initially considered a dinoflagellate (Scherfell 1900) or being reduced to suborders as their phenotypic differences
euglenoid relative (Griessmann 1913), ultrastructure showed are too slight to merit the ordinal rank assigned in Moreira
many similarities to kinetoplastids (Elbrächter et al. 1996), et al. (2004). My site-heterogeneous trees (e.g. Figs 1,
as Senn (1911) supposed. Simpson (1997) suspected it was S1, S2) strongly confirm the holophyly of Parabodonina
a diplonemid; Yabuki and Tame (2015) confirmed this by and Eubodonina shown originally by site-homogeneous
rDNA sequencing and suggested it merits a new family. I now trees (Moreira et al. 2004; von der Heyden et al. 2004),
establish new family Hemistasiidae in order Diplonemida. but are inconclusive over whether Neobodonina are para-
Hemistasia FA resembles that of Diplonemidae more than phyletic [ancestral metakinetoplastids; also seen (statistically
previously realised; elesewhere I use comparisons between insignificantly) on a recent neighbour joining tree (Hirose
them to reconstruct the ancestral diplonemid FA. Similarities et al. 2012)] or a sister clade to all other metakinetoplastids.
between the cytoskeleton of the now broadened Diplonemea As no families exist for neobodonids and the nine gen-
and Kinetoplastea are much greater than between diplone- era fall naturally into two groups with substantially distinct
mids and euglenoids. ultrastructure, I establish two neobodonid families, Rhyn-
New family Hemistasiidae. Diagnosis: phagotrophic chomonadidae being a maximally supported clade on all my
biciliates with obvious rostrum and apical cytopharynx trees. Neobodonidae is paraphyletic like Neobodo itself and
with supporting microtubules and two hollow dense cement Neobodo designis that is so deeply diverse it needs thorough
study to make numerous separate species (von der Heyden all swimming plastidless euglenoids (Neometanema and the
and Cavalier-Smith 2005). Parabodo can no longer remain osmotrophic Rhabdomonadina) as Natomonadida. Anterior-
in Bodonidae; I could have simply transferred it to Crypto- gliding Teloprocta is separated from Peranemida as order
biidae, but it is sufficiently different to merit its own family; Acroglissida. Thus Peranemida sensu stricto now has only
Cryptobia is paraphyletic and needs revision. These three Peranemidae (four genera). All these orders of Peranemea are
new families correct the irrational situation prevailing since ultrastructurally/behaviourally uniform but distinct from the
Moreira et al. (2004) of these 10 genera remaining in fam- other three, as well as phylogenetically so deeply divergent
ily Bodonidae with this family spread across three separate that their distinctive ultrastructure and ciliary locomotory
orders! Formerly a catch-all, Bodonidae is now homogeneous patterns must be very ancient, dating back at least to the
and restricted to Bodo for the first time in over 130 years since Palaeozoic.
it was established in 1883. Ciliary gliding may have played a key role in the origin
New family Neobodonidae. Diagnosis: biciliate kineto- of cilia in the ancestral eukaryote as an intermediate stage
plastids with apical cytostome and gently curved longitudinal in evolution prior to the mechanistically more complicated
cytopharynx occupying half to two thirds cell length, sub- ciliary swimming (Cavalier-Smith 2014b). According to that
stantially occupying a rigid rostrum on the same side as scenario the last common ancestor of all eukaryotes moved
the anterior undulating cilium; mouthparts only slightly by kinesin-driven posterior ciliary gliding and fed by fish-
retractile. Posterior cilium trails behind during swimming. ing for bacteria with an undulating anterior cilium that used
Intermediate root of several microtubules. Cross-linked pel- surface gliding motility to move trapped bacteria to the cell
licle microtubules evenly spread under surface. Cytostome surface [as is done today in the amoebozoan Phalansterium
supported on cilium side by five or more widely spaced micro- filosum (Smirnov et al. 2011) and the euglenoid Scytomonas
tubules, with dense reinforcing matrix and V- or Y-shaped saepesedens, and possibly Teloprocta scaphurum (Breglia
dense linkers to plasma membrane, that loop over from ciliary et al. 2013)], but had no discrete cytopharnyx nor specialized
pocket in parallel with a 2–3 mt band of closely interlinked ventral feeding groove. It was further argued that following a
microtubules, and by a prism or trapezoidal rod organ of putatively basal eukaryote bifurcation between neokaryotes
2–6 microtubule rows, apically fixed to MTR/cytopharynx and Euglenozoa (Cavalier-Smith 2010a,b), excavates (the
complex by dense amorphous cement, C-shaped in at least ancestral neokaryotes) evolved planktonic swimming and
some species; outer cytostome lip supported by curved micro- a ventral feeding groove and lost gliding motility, whereas
tubule band. Phagotrophic; bacterivores or eukaryovorous. Euglenozoa evolved a cytopharynx and retained a benthic
Type genus Neobodo Vickerman in Moreira et al. (2004). or surface-associated life style retaining ciliary gliding. The
New family Rhynchomonadidae. Diagnosis: the non- ancestral euglenozoan might have inherited posterior ciliary
swimming anterior cilium adheres to the mobile proboscis, gliding directly from the ancestral eukaryote, it being lost
swinging it from side to side as it beats; this, plus gliding by ancestral neokaryotes when feeding grooves evolved. If
on posterior cilium, and left jaw support being a sin- the basal euglenozoan bifurcation is between glycomonads,
gle microtubule row, without rod organ, distinguish them in which anterior gliding is unknown but posterior gliding
from Neobodonidae. Five MTRs and two or three external present in some neobodonids, and euglenoids/postgaardians,
microtubule band microtubules border cytopharynx; separate where Calkinsia and three euglenoid orders have anterior cil-
arc-like microtubule bands at proboscis base and tip. Type iary gliding and five euglenoid orders have posterior ciliary
genus Rhynchomonas Klebs, 1892. gliding, anterior gliding probably evolved four times inde-
New family Parabodonidae. Diagnosis: biciliates with pendently in Euglenozoa.
anterior swimming cilium and posterior trailing one, with- In other protists also posterior gliding is much more
out proboscis or rostrum. Trailing cilium attached basally widespread, being the predominant motility mode in Sul-
at least to cell body. Cytopharynx straight, shorter than in cozoa and Cercozoa and present in one heterokont genus
neobodonids, obliquely positioned with cytostome behind (Caecitellus). In Sulcozoa their ancestral posterior gliding
trailing cilium; supported by MTR and 2-mt parallel micro- was replaced by anterior gliding when breviates lost the
tubule loop; microtubular rod organ absent or reduced to posterior cilium (Cavalier-Smith 2013a,b), and in Cercozoa
5-mt vestige or single row. Intermediate centriolar root of one anterior gliding evolved in skiomonads, which alone among
microtubule. Outer cytostome lip’s curved microtubular sup- eukaryotes glide on both cilia (Cavalier-Smith and Karpov
porting band with numerous or as few as two microtubules. 2012; Howe et al. 2011). In Viridiplantae Chlamydomonas
Non-gliding, bacterivorous phagotrophs or osmotrophs. Type has bidirectional intraciliary motility in both cilia mediated
and sole genus Parabodo Vickerman in Moreira et al. (2004). by intraciliary transport particle trains driven one way by
dynein and the other by kinesin motors (Shih et al. 2013),
their dynein 1b causing cell gliding with anterior-directed
Euglenozoan Ciliary Functional Divergence cilium. All ciliated eukaryotes use kinesin-based anterograde
transport for ciliary growth (Scholey 2013a,b) and I sus-
My new classification segregated the posterior-gliding pect dynein-based retrograde transport for retraction (Dentler
Anisonemidae as a new order Anisonemida and grouped 2005). If, as is likely, all ciliated eukaryotes have intraciliary
transport particle trains driven by both dynein and kinesin, are roughly half as old as Euglenozoa, and Euglenophyceae
the direction of ciliary gliding on substrata would depend only about three quarters as old as Spirocuta, so Spirocuta
on which of these motors is activated when coupled to sur- probably evolved only after eukaryotic algae, whose origin
face glycoproteins (Shih et al. 2013) rather than on their would have given a selective advantage to greater cell size
presence or absence. The more even spread of anterior and via strip multiplication. Perhaps, the classical view should be
posterior gliding in Euglenozoa might be attributable solely turned on its head; eukaryovory may be the ancestral condi-
to differential loss from a common ancestor that had both, tion for Euglenozoa and may have stimulated the origin of
but the fact that no modern euglenoids have retained both the FA. Some postgaardeans, diplonemids and kinetoplastids
and that only two eukaryote genera (the skiomonads Trem- are eukaryovores, so why should we suppose that all early
ula and Glissandra (Howe et al. 2011)) glide on both cilia euglenoids ate only bacteria?
suggests that biciliary gliders occupy a very narrow zone and
that biciliary gliding was probably an evolutionarily unstable
state for euglenoids, though it might have existed tempo- Concluding Comment
rarily during the switches from one mode to another when
orders diverged. If all euglenoid anterior gliding evolved In my view this revised classification better partitions the
secondarily, as in Sulcozoa and Cercozoa, intermediates immense cellular diversity of phagotrophic euglenoids that
where switching occurred could have been dual gliders like is now increasingly apparent as well as the fundamental
skiomonads, but perhaps more likely near-surface skidding ultrastructural differences amongst the three subphyla than
swimmers like Neometanema, an evolutionary stable adap- do any previous classifications of Euglenozoa. However, as
tive zone for euglenoids; loss of posterior gliding coupled Euglenozoa of interestingly distinct ultrastructure will prob-
with retention of surface-hugging swimming behaviour could ably continue to be discovered for at least another decade, it
have preadapted them for secondary evolution of anterior is likely to need some further revision. Nonetheless I hope
gliding, simply explaining why it is commoner in euglenoids its main features will stand the test of time and contribute to
than other phyla. greater taxonomic stability in the future.
Derived Nature of Petalomonad Feeding Acknowledgements

Mode I thank NERC for past grant and Professorial Fellowship
support and funding open access and Ya-Fan Chan for kindly
The common idea that the first euglenoids were strict bac-
providing his sequence alignment.
terivores (Leander 2004; Triemer and Farmer 1991a,b) is
ecologically and evolutionarily unrealistic, and the associated
idea that petalomonad simple FA is primitive is contradicted Appendix A. Supplementary data
by my improved rDNA trees showing petalomonads as nested
within lineages with highly complex FA and the discov- Supplementary data associated with this arti-
ery of more complex FA in Scytomonas than other studied cle can be found, in the online version, at
petalomonads (Cavalier-Smith et al. 2016a). More likely http://dx.doi.org/10.1016/j.ejop.2016.09.003.
ancestral euglenoids opportunistically fed on both bacteria
and protists, and were of moderate, not tiny, size. Abil-
ity of the 54 ␮m ‘Ploeotia aff. vitrea’ to eat algae (Lax References
and Simpson 2013) and the phylogenetically distant Ser-
penomonas to eat yeast (Linton and Triemer 1999) show that Adl, S.M., Simpson, A.G., Farmer, M.A., Andersen, R.A., Ander-
stavomonad rigid pellicles are not barriers to eukaryovory, son, O.R., Barta, J.R., Bowser, S.S., Brugerolle, G., Fensome,
as often assumed (Triemer and Farmer 1991b); small size, R.A., Fredericq, S., James, T.Y., Karpov, S., Kugrens, P., Krug,
mouthpart simplification, or phylogenetic specialization in J., Lane, C.E., Lewis, L.A., Lodge, J., Lynn, D.H., Mann,
food preference better explain why stavomonads are mostly D.G., McCourt, R.M., Mendoza, L., Moestrup, O., Mozley-
bacterivorous. Most are so little studied that our view of Standridge, S.E., Nerad, T.A., Shearer, C.A., Smirnov, A.V.,
their diet may be biased; eukaryovores are harder to culti- Spiegel, F.W., Taylor, M.F., 2005. The new higher level classifi-
vate than bacterivores. Even Scytomonas ingests quite large cation of eukaryotes with emphasis on the taxonomy of protists.
J. Eukaryot. Microbiol. 52, 399–451.
cells (Cavalier-Smith et al. 2016a) and could in principle have
Adl, S.M., Simpson, A.G., Lane, C.E., Lukeš, J., Bass, D.,
enslaved a cyanobacterium, as did Paulinella and the ances- Bowser, S.S., Brown, M.W., Burki, F., Dunthorn, M., Hampl,
tor of plants (Cavalier-Smith 2013b). However, if euglenoids V., Heiss, A., Hoppenrath, M., Lara, E., Le Gall, L., Lynn, D.H.,
are ∼1.2 Gy old but green algae only ∼750–800 My old McManus, H., Mitchell, E.A., Mozley-Stanridge, S.E., Parfrey,
(see Cavalier-Smith 2013a), for much of their early his- L.W., Pawlowski, J., Rueckert, S., Shadwick, R.S., Schoch, C.L.,
tory stavomonads existed in a world with no green or other Smirnov, A., Spiegel, F.W., 2012. The revised classification of
eukaryotic algae to eat. According to Fig. 1 Euglenophyceae eukaryotes. J. Eukaryot. Microbiol. 59, 429–493.
Allen, J.W.A., 2011. Cytochrome c biogenesis in Malawimonas): their evolutionary affinities and new higher taxa.
mitochondria—systems III and V. FEBS J. 278, 4198–4216. Int. J. Syst. Evol. Microbiol. 53, 1741–1758.
Belhadri, A., Brugerolle, G., 1992. Morphogenesis and ultra- Cavalier-Smith, T., 2003b. Protist phylogeny and the high-level clas-
structure of the feeding apparatus of Entosiphon sulcatum: an sification of Protozoa. Eur. J. Protistol. 39, 338–348.
immunofluorescence and ultrastructural study. Protoplasma 168, Cavalier-Smith, T., 2007. Evolution and relationships of algae:
125–135. major branches of the tree of life. In: Brodie, J., Lewis, J. (Eds.),
Belhadri, A., Bayle, D., Brugerolle, G., 1992. Biochemical and Unravelling the Algae—The Past, Present, and Future of Algal
immunological characterization of intermicrotubular cement in Systematics. CRC, Boca Raton, pp. 21–55.
the feeding apparatus of phagotrophic euglenoids: Entosiphon, Cavalier-Smith, T., 2010a. Kingdoms Protozoa and Chromista and
Peranema, and Ploeotia. Protoplasma 168, 113–124. the eozoan root of eukaryotes. Biol. Lett. 6, 342–345.
Bicudo, C.E.de M., Menezes, M., 2016. Phylogeny and classifica- Cavalier-Smith, T., 2010b. Deep phylogeny, ancestral groups, and
tion of Euglenophyceae: a brief review. Front. Ecol. Evol. 4, the four ages of life. Philos. Trans. R. Soc. B 365, 111–132.
http://dx.doi.org/10.3389/fevo.2016.00017, article 17. Cavalier-Smith, T., 2013a. Early evolution of eukaryote feeding
Blochmann, F., 1895. Die microscopische Pflanzen- und Thierwelt modes, cell structural diversity, and classification of the proto-
des Susswassers. Theil II. Abteilung I: Protozoa. Lucas Gräfe & zoan phyla Loukozoa Sulcozoa, and Choanozoa. Eur. J. Protistol.
Sillem, Hamburg. 49, 115–178.
Breglia, S.A., Yubuki, N., Hoppenrath, M., Leander, B.S., 2010. Cavalier-Smith, T., 2013b. Symbiogenesis: mechanisms, evolution-
Ultrastructure and molecular phylogenetic position of a novel ary consequences, and systematic implications. Ann. Rev. Ecol.
euglenozoan with extrusive episymbiotic bacteria Bihospites Evol. Syst. 44, 145–172.
bacati n. gen et sp. (Symbiontida). BMC Microbiol. 10, 145. Cavalier-Smith, T., 2014a. Gregarine site-heterogeneous 18S rDNA
Breglia, S.A., Yubuki, N., Leander, B.S., 2013. Ultrastructure and trees, revision of gregarine higher classification, and the evo-
molecular phylogenetic position of Heteronema scaphurum: a lutionary diversification of Sporozoa. Eur. J. Protistol. 50,
eukaryovorous euglenid with a cytoproct. J. Eukaryot. Micro- 472–495.
biol. 60, 107–120. Cavalier-Smith, T., 2014b. The neomuran revolution and
Brown, M.W., Sharpe, S.C., Silberman, J.D., Heiss, A.A., Lang, phagotrophic origin of eukaryotes and cilia in the light of
B.F., Simpson, A.G., Roger, A.J., 2013. Phylogenomics demon- intracellular coevolution and a revised tree of life. In: Keel-
strates that breviate flagellates are related to opisthokonts and ing, P.J., Koonin, E.V. (Eds.), The Origin and Evolution
apusomonads. Proc. R. Soc. Biol. Sci. 280, 20131755. of Eukaryotes. Cold Spring Harbor Perspectives in Biology,
Brugerolle, G., 1985. Des trichocystes chez les bodonides, un car- http://dx.doi.org/10.1101/cshperspect.a016006.
actère phylogénétique supplémentaire entre Kinetoplastida et Cavalier-Smith, T., 2015. Mixed heterolobosean and novel gregarine
Euglenida. Protistologica 21, 339–348. lineage genes from culture ATCC 50646: long-branch artefacts,
Burki, F., Kaplan, M., Tikhonenkov, D.V., Zlatogursky, V., Minh, not lateral gene transfer, distort ␣-tubulin phylogeny. Eur. J.
B.Q., Radaykina, L.V., Smirnov, A., Mylnikov, A.P., Keeling, Protistol. 51, 121–137.
P.J., 2016. Untangling the early diversification of eukaryotes: a Cavalier-Smith, T., Karpov, S.A., 2012. Paracercomonas kinetid
phylogenomic study of the evolutionary origins of Centrohelida, ultrastructure, origins of the body plan of Cercomonadida, and
Haptophyta and Cryptista. Proc. R. Soc. Biol. Sci. 283. cytoskeleton evolution in Cercozoa. Protist 163, 47–75.
Busse, I., Preisfeld, A., 2003. Systematics of primary osmotrophic Cavalier-Smith, T., Chao, E.E., Snell, E.A., Berney, C., Fiore-
euglenids: a molecular approach to the phylogeny of Distigma Donno, A.M., Lewis, R., 2014. Multigene eukaryote phylogeny
and Astasia (Euglenozoa). Int. J. Syst. Evol. Microbiol. 53, reveals the likely protozoan ancestors of opisthokonts (animals,
617–624. fungi, choanozoans) and Amoebozoa. Mol. Phylogenet. Evol.
Bütschli, O., 1884. Dr. H.G. Bronn’s Klassen und Ordnungen des 81, 71–85.
Thier-Reichs. Vol. 1 Abt. II Mastigophora. C. F. Winter, Heidel- Cavalier-Smith, T., Chao, E.E., Lewis, R., 2015a. Multiple origins
berg. of Heliozoa from flagellate ancestors: new cryptist subphylum
Calkins, G.N., 1926. The Protozoa. Lee and Fibiger, New York. Corbihelia, superclass Corbistoma, and monophyly of Haptista
Cavalier-Smith, T., 1978. The evolutionary origin and phylogeny of Cryptista, Hacrobia and Chromista. Mol. Phylogenet. Evol. 93,
microtubules, mitotic spindles and eukaryote flagella. BioSys- 331–362.
tems 10, 93–114. Cavalier-Smith, T., Fiore-Donno, A.M., Chao, E., Kudryavtsev, A.,
Cavalier-Smith, T., 1981. Eukaryote kingdoms: seven or nine? Berney, C., Snell, E.A., Lewis, R., 2015b. Multigene phylogeny
BioSystems 14, 461–481. resolves deep branching of Amoebozoa. Mol. Phylogenet. Evol.
Cavalier-Smith, T., 1993. Kingdom Protozoa and its 18 phyla. 83, 293–304.
Microbiol. Rev. 57, 953–994. Cavalier-Smith, T., Chao, E.E., Vickerman, K., 2016a. New
Cavalier-Smith, T., 1995. Zooflagellate phylogeny and classifica- phagotrophic euglenoid species, (new genus Decastava; Scy-
tion. Tsitologiia 37, 1010–1029. tomonas saepesedens; Entosiphon oblongum), Hsp90 introns,
Cavalier-Smith, T., 1998. A revised six-kingdom system of life. and putative euglenoid Hsp90 pre-mRNA insertional editing.
Biol. Rev. Camb. Philos. Soc. 73, 203–266. Eur. J. Protistol. 56, 147–170.
Cavalier-Smith, T., 2002. The phagotrophic origin of eukaryotes Cavalier-Smith, T., Chao, E.E., Lewis, R., 2016b. 187-gene phy-
and phylogenetic classification of Protozoa. Int. J. Syst. Evol. logeny of protozoan phylum Amoebozoa reveals a new class
Microbiol. 52, 297–354. (Cutosea) of deep-branching, ultrastructurally unique, enveloped
Cavalier-Smith, T., 2003a. The excavate protozoan phyla marine Lobosa, and clarifies amoeba evolution. Mol. Phylogenet.
Metamonada Grassé emend. (Anaeromonadea, Parabasalia, Evol. 99, 275–296.
Carpediemonas, Eopharyngia) and Loukozoa emend. (Jakobea,
Chan, Y.F., Chiang, K.P., Chang, J., Moestrup, O., Chung, C.C., 2012. When, how and why glycolysis became compartmen-
2015. Strains of the morphospecies Ploeotia costata (Eugleno- talised in the Kinetoplastea. A new look at an ancient organelle.
zoa) isolated from the Western North Pacific (Taiwan) reveal Int. J. Parasitol. 42, 1–20.
substantial genetic differences. J. Eukaryot. Microbiol. 62, Hausmann, K., Hülsmann, N., 1996. Protozoology. Georg Thieme,
318–326. Stuttgart.
Chan, Y.F., Moestrup, O., Chang, J., 2013. On Keelungia pulex nov. Hirose, E., Nozawa, A., Kumagai, A., Kitamura, S., 2012. Azumio-
gen. et nov. sp., a heterotrophic euglenoid flagellate that lacks bodo hoyamushi gen. nov. et sp. nov. (Euglenozoa Kinetoplastea,
pellicular plates (Euglenophyceae Euglenida). Eur. J. Protistol. Neobodonida): a pathogenic kinetoplastid causing the soft tunic
49, 15–31. syndrome in ascidian aquaculture. Dis. Aquat. Organ. 97,
Dentler, W., 2005. Intraflagellar transport (IFT) during assembly 227–235.
and disassembly of Chlamydomonas flagella. J. Cell Biol. 170, Hollande, A., 1942. Étude cytologique et biologique de quelques
649–659. flagellés libres. Arch. Zool. Exp. Gén. 83, 1–268.
Derelle, R., Torruella, G., Klimes, V., Brinkmann, H., Kim, E., Hollande, A., 1952. Classe des Eugléniens (Euglenoidina Bütschli
Vlcek, C., Lang, B.F., Elias, M., 2015. Bacterial proteins pin- 1884). In: Grassé, P.-P. (Ed.), Traité de Zoologie. Masson, Paris,
point a single eukaryotic root. Proc. Natl. Acad. Sci. U. S. A pp. 238–284.
112, E693–E699. Honigberg, B.M., Balamuth, W., Bovee, E.C., Corliss, J.O., Gojdics,
Deschamps, P., Lara, E., Marande, W., López-García, P., Ekelund, M., Hall, R.P., Kudo, N.D., Levine, N.D., Loeblich Jr., A.R.,
F., Moreira, D., 2011. Phylogenomic analysis of kinetoplastids Weiser, J., Wenrich, D.H., 1964. A revised classification of phy-
supports that trypanosomatids arose from within bodonids. Mol. lum Protozoa. J. Protozool. 11, 7–20.
Biol. Evol. 28, 53–58. Howe, A.T., Bass, D., Scoble, J.M., Lewis, R., Vickerman, K.,
Dobáková, E., Flegontov, P., Skalický, T., Lukeš, J., 2015. Unex- Arndt, H., Cavalier-Smith, T., 2011. Novel cultured protists iden-
pectedly streamlined mitochondrial genome of the euglenozoan tify deep-branching environmental DNA clades of Cercozoa:
Euglena gracilis. Genome Biol. Evol. 7, 3358–3367. new genera Tremula, Micrometopion, Minimassisteria, Nudifila,
Dujardin, F., 1841. Histoire naturelle des zoophytes infusoires. Peregrinia. Protist 162, 332–372.
Roret, Paris. Huber-Pestalozzi, G., 1955. Das Phytoplankton des Süsswassers;
Ekebom, J., Patterson, D.J., Vørs, N., 1995/6. Heterotrophic flagel- Sytematic und Biologie Euglenophyceen. Teil 4. E. Schweitzer-
lates from coral reef sediments (Great Barrier Reef, Australia). bart’sche Verlagsbuchhandlung, Stuttgart.
Arch. Protistenk. 146, 251–272. Joyon, L., Lom, J., 1969. Etude cytologique, systématique et
Elbrächter, M., Schnepf, E., Balzer, I., 1996. Hemistasia phaeo- pathologique d’Ichtyobodo necator (Henneguy, 1883) Pinto,
cysticola (Scherffel) comb. nov. redescription of a free-living, 1928 (Zooflagelle). J. Protozool. 16, 703–719.
marine, phagotrophic kinetoplastid flagellate. Arch. Protistenk. Karpov, S.A. (Ed.), 2000. Protista: Handbook on Zoology. Nauka,
147, 125–136. St Petersburg.
Farmer, M.A., Triemer, R.E., 1988. A redescription of the genus Kim, J.I., Shin, W., Triemer, R.E., 2010. Multigene analyses of pho-
Ploeotia Duj. (Euglenophyceae). Taxon 37, 319–325. tosynthetic euglenoids and new family, Phacaceae (Euglenales).
Farmer, M.A., Triemer, R.E., 1994. An ultrastructural study of J. Phycol. 46, 1278–1287.
Lentomonas applanatum (Preisig) n. g. (Euglenida). J. Eukaryot. Kivic, P.A., Walne, P.L., 1984. An evaluation of a possible phyloge-
Microbiol. 41, 112–119. netic relationship between the Euglenophyta and Kinetoplastida.
Fenchel, T., Bernard, C., Esteban, G., Finlay, B.J., Hansen, P.J., Origins Life 13, 269–288.
Iversen, N., 1995. Microbial diversity and activity in a Danish Klebs, G., 1883. Über die Organisation einiger Flagellaten-Gruppen
fjord with anoxic deep water. Ophelia 43, 45–100. und ihre Beziehungen zu Algen und Infusorien. Unters. Bot. Inst.
Fernández-Galiano, D., 1990. Las nuevas clasificaciones de los Tübingen 1, 233–262.
organismos eucarióticos unicelulares protistología versus proto- Klebs, G., 1892. Flagellatenstudien I and II. Zeit. Wiss. Zool. 55,
zoología. Rev. R. Soc. Esp. Hist. Nat. (Sec. Biol.) 85, 107–125. 265–445.
Flegontov, P., Votýpka, J., Skalický, T., Logacheva, M.D., Penin, Kudo, R.R., 1966. Protozoology, 5th ed. Thomas, Springfield.
A.A., Tanifuji, G., Onodera, N.T., Kondrashov, A.S., Volf, P., Lackey, J.B., 1960. Calkinsia aureus gen. et sp. nov.: a new marine
Archibald, J.M., Lukeš, J., 2013. Paratrypanosoma is a novel euglenid. Trans. Am. Microsc. Soc. 79, 105–107.
early-branching trypanosomatid. Curr. Biol. 23, 1787–1793. Lankester, E.R., 1885. Protozoa. In: Baynes, T.S., Smith, W.R.
Fritsch, F.E., 1948. The Structure and Reproduction of the Algae, (Eds.), The Encyclopaedia Britannica, Vol. 19, 9th ed. A. and
vol. 1. Cambridge University Press, Cambridge. C. Black, Edinburgh, pp. 830–866.
Frolov, A.O., Karpov, S.A., 1995. Comparative morphology of kine- Larsen, J., Patterson, D.J., 1990. Some flagellates (Protista) from
toplastids. Tsitologiia 37, 1072–1096. tropical marine sediments. J. Nat. Hist. 24, 801–937.
Frolov, A.O., Karpov, S.A., Mylnikov, A.P., 2001. The ultrastruc- Lasek-Nesselquist, E., Gogarten, J.P., 2013. The effects of model
ture of Procryptobia sorokini (Zhukov) comb. nov. and rootlet choice and mitigating bias on the ribosomal tree of life. Mol.
homology in kinetoplastids. Protistology 2, 85–95. Phylogenet. Evol. 69, 17–38.
Gibbs, S.P., 1981. The chloroplasts of some algal groups may have Lax, G., Simpson, A.G.B., 2013. Combining molecular data with
evolved from symbiotic green algae. Ann. N. Y. Acad. Sci. 361, classical morphology for uncultured phagotrophic euglenids
193–208. (Excavata): a single-cell approach. J. Eukaryot. Microbiol. 60,
Griessmann, K., 1913. Über marine Flagellaten. Arch. Protistenk. 615–625.
32, 1–78. Leander, B.S., 2004. Did trypanosomatid parasites have photosyn-
Gualdrón-López, M., Brennand, A., Hannaert, V., Quiñones, W., thetic ancestors? Trends Microbiol. 12, 251–258.
Cáceres, A.J., Bringaud, F., Concepción, J.L., Michels, P.A.,
Lee, W.J., 2002. Redescription of the rare heterotrophic flagellate Mignot, J.-P., Brugerolle, G., Bricheux, G., 1987. Intercalary strip
(Protista) – Phyllomitus undulans Stein, 1878, and erection of development and dividing cell morphogenesis in the euglenoid
a new genus – Pseudophyllomitus gen. n. Acta Protozool. 41, Cyclidiopsis acus. Protoplasma 139, 51–65.
375–381. Montegut-Felkner, A.E., Triemer, R.E., 1994. Phylogeny of
Lee, W.J., Simpson, A.G.B., 2014a. Morphological and molecular Diplonema ambulator (Larsen and Patterson): 1. Homologies
characterisation of Notosolenus urceolatus Larsen and Patterson of the flagellar apparatus. Eur. J. Protistol. 30, 227–237.
1990, a member of an understudied deep-branching euglenid Moreira, D., López-García, P., Vickerman, K., 2004. An updated
group (petalomonads). J. Eukaryot. Microbiol. 61, 463–479. view of kinetoplastid phylogeny using environmental sequences
Lee, W.J., Simpson, A.G.B., 2014b. Ultrastructure and molecu- and a closer outgroup: proposal for a new classification of the
lar phylogenetic position of Neometanema parovale sp. nov. class Kinetoplastea. Int. J. Syst. Evol. Microbiol. 54, 1861–1875.
(Neometanema gen. nov.), a marine phagotrophic euglenid with Mylnikov, A.P., 1986. Ultrastructure of a colourless flagellate,
skidding motility. Protist 165, 452–472. Phyllomitus apiculatus Skuja 1948 (Kinetoplastida). Arch. Pro-
Leedale, G., 1966. Euglena: a new look with the electron micro- tistenk. 182, 1–10.
scope. Adv. Sci. 23, 22–37. Mylnikov, A.P., Mylnikova, Z.M., Tsvetkov, A.H., Elizarova, V.A.,
Leedale, G., 1967. Euglenoid Flagellates. Prentice-Hall, Englewood 1998. Fine structure of the predatory flagellate Phyllomitus amy-
Cliffs, N.J. lophagus (in Russian). Biol. Vnutr. Vod 2, 21–27.
Leedale, G., 1970. Phylogenetic aspects of nuclear cytology in algae. Pascher, A., 1931. Systematische Übersicht über die mit Flagellaten
Ann. N. Y. Acad. Sci. 175, 429–453. in Zusammenhang stehenden Algenreihen und versuch einen
Leedale, G., 2002 dated 2000. Class Euglenoidea Bütschli, 1884, in: Einreihung dieser Algenstämme in die Stämme des Pflanzen-
Lee, J.J., Leedale, G., Bradbury, P. (Eds.), The Illustrated Guide reiches. Biol. Zentralbl. 48, 317–332.
to the Protozoa. Society of Protozoologists, Lawrence, Kansas, Patterson, D.J., Simpson, A.G.B., 1996. Heterotrophic flagellates
pp. 1136–1157. from coastal marine and hypersaline sediments in Western Aus-
Lemmermann, E., 1913. Flagellatae 2. Gustav Fischer, Jena. tralia. Eur. J. Protistol. 32, 423–448.
Levine, N.D., Corliss, J.O., Cox, F.E., Deroux, G., Grain, J., Honig- Perez, E., Lapaille, M., Degand, H., Cilibrasi, L., Villavicencio-
berg, B.M., Leedale, G.F., Loeblich 3rd, A.R., Lom, J., Lynn, D., Queijeiro, A., Morsomme, P., Gonzalez-Halphen, D., Field,
Merinfeld, E.G., Page, F.C., Poljansky, G., Sprague, V., Vávra, M.C., Remacle, C., Baurain, D., Cardol, P., 2014. The mito-
J., Wallace, F.G., 1980. A newly revised classification of the chondrial respiratory chain of the secondary green alga Euglena
protozoa. J. Protozool. 27, 37–58. gracilis shares many additional subunits with parasitic Try-
Linton, E.W., Karnkowska-Ishikawa, A., Kim, J.I., Shin, W., Ben- panosomatidae. Mitochondrion 19 (Pt B), 338–349.
nett, M.S., Kwiatowski, J., Zakrys, B., Triemer, R.E., 2010. Pettigrew, G.W., Leaver, J.L., Meyer, T.E., Ryle, A.P., 1975.
Reconstructing euglenoid evolutionary relationships using three Purification, properties, and amino acid sequences of atypical
genes: nuclear SSU and LSU, and chloroplast SSU rDNA cytochrome c from two Protozoa, Euglena gracilis and Crithidia
sequences and the description of Euglenaria gen. nov. (Eugleno- oncopelti. Biochem. J. 147, 291–302.
phyta). Protist 161, 603–619. Preisig, H., 1979. Zwei neue Vertreter der farblosen Euglenophyta.
Linton, E.W., Triemer, R.E., 1999. Reconstruction of the feeding Schweiz. Z. Hydrol. 41, 155–160.
apparatus in Ploeotia costata (Euglenophyta) and its relationsip Pringsheim, E.G., 1948. Taxonomic problems in the Euglenineae.
to other euglenoid feeding apparatuses. J. Phycol. 35, 313–324. Biol. Rev. 23, 46–61.
Linton, E.W., Triemer, R.E., 2001. Reconstruction of the flagellar Raymann, K., Brochier-Armanet, C., Gribaldo, S., 2015. The
apparatus in Ploeotia costata (Euglenozoa) and its relationship two-domain tree of life is linked to a new root for the
to other euglenoid flagellar apparatuses. J. Eukaryot. Microbiol. Archaea. Proc. Natl. Acad. Sci. U. S. A. 112, 6670–6675,
48, 88–94. http://dx.doi.org/10.1073/pnas.1420858112.
Makiuchi, T., Annoura, T., Hashimoto, M., Hashimoto, T., Aoki, T., Ritter von Stein, F., 1878. Der Organismus der Infusionsthiere. III.
Nara, T., 2011. Compartmentalization of a glycolytic enzyme Der Organismus der Flagellaten I. Engelmann, Leipzig.
in Diplonema, a non-kinetoplastid euglenozoan. Protist 162, Roy, J., Faktorová, D., Lukeš, J., Burger, G., 2007. Unusual mito-
482–489. chondrial genome structures throughout the Euglenozoa. Protist
Marande, W., Lukeš, J., Burger, G., 2005. Unique mitochondrial 158, 385–396.
genome structure in diplonemids, the sister group of kinetoplas- Ruggiero, M.A., Gordon, D.P., Orrell, T.M., Bailly, N.B., Bourgoin,
tids. Eukaryot. Cell 4, 1137–1146. T., Brusca, T., Cavalier-Smith, T., Guiry, M., Kirk, P.M., 2015.
Marin, B., Palm, A., Klingberg, M., Melkonian, M., 2003. A higher level classification of all living organisms. PLoS One
Phylogeny and taxonomic revision of plastid-containing 10, e0119248, http://dx.doi.org/10.1371/journal.pone.0119248.
euglenophytes based on SSU rDNA sequence comparisons and Saville Kent, W., 1880–1882. A manual of the Infusoria. Bogue,
synapomorphic signatures in the SSU rRNA secondary structure. London.
Protist 154, 99–145. Scherfell, A., 1900. Phaeocystis globosa nov. spec. nebst einigen
Mayr, E., Bock, W.J., 2002. Classifications and other ordering sys- Betrachtungen über die Phylogenie niederer inbesondere brauner
tems. J. Syst. Evol. Res., 169–194. Organismen. Wiss. Meeresunters. Abt. Helgoland 4, 1–29.
Mignot, J.-P., 1964. Observations complémentaires sur la struc- Schoenichen, W., 1925. Einfachste Lebensformen des Tier- und
ture des flagelles d’Entosiphon sulcatum (Duj.) Stein, flagellé Pflanzenreiches. 5. Aufl. Band I. Spaltpflanzen, Geissellinge,
Euglénien. C. R. Séances Acad. Sci. 258, 3360–3363. Algen, Pilze. In: Eyferth, B., Schoenichen, W. (Eds.). Lichter-
Mignot, J.-P., 1967. Affinités des euglénamonadines et des felde, Berlin, pp. vii + 519.
chloromonadines: remarques sur la sytématique des Euglénida. Scholey, J.M., 2013a. Cilium assembly: delivery of tubulin by
Protistologica 3, 25–60. kinesin-2-powered trains. Curr. Biol. 23, R956–R959.
Scholey, J.M., 2013b. Kinesin-2: a family of heterotrimeric and Triemer, R.E., Farmer, M.A., 1991a. An ultrastructural comparison
homodimeric motors with diverse intracellular transport func- of the mitotic apparatus, feeding apparatus, flagellar apparatus
tions. Ann. Rev. Cell Dev. Biol. 29, 443–469. and cytoskeleton in euglenoids and kinetoplastids. Protoplasma
Schroeckh, S., Lee, W.J., Patterson, D.J., 2003. Free-living het- 164, 91–104.
erotrophic euglenids from freshwater sites in mainland Australia. Triemer, R.E., Farmer, M.A., 1991b. The ultrastructural organiza-
Hydrobiologia 493, 131–166. tion of heterotrohic euglenids and its evolutionary imlications.
Schwartz, R.M., Dayhoff, M.O., 1978. Origins of prokary- In: Patterson, D.J., Larsen, J. (Eds.), The Biology of Free-
otes, eukaryotes, mitochondria, and chloroplasts. Science 199, Living Heterotrophic Flagellates. Clarendon Press, Oxford, pp.
395–403. 185–204.
Senn, G., 1900. Eugleninae. In: Engler, A., Prantl, K. (Eds.), Triemer, R.E., Farmer, M.A., 2007. A decade of euglenoid molec-
Die natürlichen Pflanzenfamilien. 1. Engelmann, Leipzig, pp. ular systematics. In: Brodie, J., Lewis, J. (Eds.), Unravelling
173–185. the Algae—The Past, Present, and Future of Algal Systematics.
Senn, G., 1911. Oxyrrhis, Nephroselmis und einige Euflagellaten, CRC, Boca Raton, pp. 315–330.
nebst Bemerkungen über deren System. Zschr. Wiss. Zool. 97, Triemer, R.E., Fritz, L., 1988. Ultrastructural features of mitosis
605–672. in Ploeotia costata (Heteronematales, Euglenophyta). J. Phycol.
Shih, S.M., Engel, B.D., Kocabas, F., Bilyard, T., Gennerich, A., 24, 514–519.
Marshall, W.F., Yildiz, A., 2013. Intraflagellar transport drives Vasileva, I.I., 1987. Yakut Euglenoid and Xanthophyte Algae.
flagellar surface motility. eLife 2, e00744. Nauka, Leningrad (in Russian).
Shin, W., Brosnan, S., Triemer, R.E., 2002. Are cytoplasmic pockets von der Heyden, S., Chao, E.E., Vickerman, K., Cavalier-Smith,
(MTR/pocket) present in all photosynthetic euglenoid genera? T., 2004. Ribosomal RNA phylogeny of bodonid and diplone-
J. Phycol. 38, 790–799. mid flagellates and the evolution of Euglenozoa. J. Eukaryot.
Shiratori, T., Yokoyama, A., Ishida, K., 2014. Phylogeny, ultra- Microbiol. 51, 402–416.
structure, and flagellar apparatus of a new marimonad flagellate von der Heyden, S., Cavalier-Smith, T., 2005. Culturing and
Abollifer globosa sp. nov. (Imbricatea, Cercozoa). Protist 165, environmental DNA sequencing uncover hidden kinetoplas-
808–824. tid biodiversity and a major marine clade within ancestrally
Simpson, A.G.B., 1997. The identity and composition of the freshwater Neobodo designis. Int. J. Syst. Evol. Microbiol. 55,
Euglenozoa. Arch. Protistenk. 148, 318–328. 2605–2612.
Simpson, A.G.B., Inagaki, Y., Roger, A.J., 2006. Comprehensive Walne, P.L., Kivic, P.A., 1990. Phylum Euglenida. In: Margulis, L.,
multigene phylogenies of excavate protists reveal the evolution- Corliss, J.O., Melkonian, M., Chapman, D.J. (Eds.), Handbook
ary positions of “primitive” eukaryotes. Mol. Biol. Evol. 23, of Protoctista. Jones & Bartlett, Boston, pp. 270–287.
615–625. Walton, L.B., 1915. A review of described species of the order
Simpson, A.G.B., Roger, A.J., 2004. Protein phylogenies robustly Euglenoidina Bloch, Class Flagellata (Protozoa) with particu-
resolve the deep-level relationships within Euglenozoa. Mol. lar reference to those found in the city water supplies and in
Phylogenet. Evol. 30, 201–212. other localities in Ohio. Ohio State Univ. Bull. 19, 341–449.
Simpson, A.G.B., van den Hoff, J., Bernard, C., Burton, H.R., Patter- West, G.S., 1904. A Treatise on the British Freshwater Algae. The
son, D.J., 1996/97. The ultrastructure and systematic position of University Press, Cambridge.
the euglenozoon Postgaardi mariagerensis, Fenchel et al. Arch. West, G.S., Fritsch, F.E., 1932. A Treatise on the British Freshwater
Protistenk. 147, 213–225. Algae, Revised Ed. The University Press, Cambridge.
Skuja, H., 1934 dated 1932. Beitrag zur Algenflora Lettlands I. Acta Yabuki, A., Tame, A., 2015. Phylogeny and reclassification of
Hort. Bot. Univ. Latt. 7, 25–85. Hemistasia phaeocysticola (Scherffel) Elbrächter & Schnepf,
Smirnov, A.V., Chao, E., Nassonova, E.S., Cavalier-Smith, T., 2011. 1996. J. Eukaryot. Microbiol. 62, 426–429.
A revised classification of naked lobose amoebae (Amoebozoa: Yamaguchi, A., Yubuki, N., Leander, B.S., 2012. Morphostasis in
Lobosa). Protist 162, 545–570. a novel eukaryote illuminates the evolutionary transition from
Smith, G.M., 1933. The Freshwater Algae of the United States. phagotrophy to phototrophy: description of Rapaza viridis n.
McGraw-Hill, New York and London. gen et sp. (Euglenozoa, Euglenida). BMC Evol. Biol. 12, 29.
Stoeck, T., Schwarz, M.V., Boenigk, J., Schweikert, M., von der Yubuki, N., Edgcomb, V.P., Bernhard, J.M., Leander, B.S., 2009.
Heyden, S., Behnke, A., 2005. Cellular identity of an 18S rRNA Ultrastructure and molecular phylogeny of Calkinsia aureus:
gene sequence clade within the class Kinetoplastea: the novel cellular identity of a novel clade of deep-sea euglenozoans with
genus Actuariola gen nov. (Neobodonida) with description of epibiotic bacteria. BMC Microbiol. 9, 16.
the type species Actuariola framvarensis sp. nov. Int. J. Syst. Yubuki, N., Simpson, A.G., Leander, B.S., 2013. Reconstruction
Evol. Microbiol 55, 2623–2635. of the feeding apparatus in Postgaardi mariagerensis pro-
Taylor, F.J.R., 1976. Flagellate phylogeny: a study in conflicts. J. vides evidence for character evolution within the Symbiontida
Protozool. 23, 28–40. (Euglenozoa). Eur. J. Protistol. 49, 32–39.
Triemer, R.E., 1986. Light and electron microscopic description of
a colourless euglenoid, Serpenomomas costata n. g., n. sp. J.
Protozool. 33, 412–415.

Euglenozoa Cavalier Smith

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Euglenozoa Cavalier Smith

Uploaded by

Copyright:

Available Formats

Available online at www.sciencedirect.

Higher classiﬁcation and phylogeny of Euglenozoa

Keywords: Diplonemea; Euglenoida; Glycomonada; Kinetoplastea; Spirocuta; Postgaardia

Introduction trees (mitochondrial cytochrome c: Schwartz and Dayhoff

Suborder 2. Rhabdomonadina Leedale, 1967 em. Cavalier-Smith, 1993 (osmotrophs: originally

excluded Astasia and Distigma )

(Astasia , Rhabdomonas, Gyropaigne, Menoidium, Parmidium)

Order 1. Acroglissida ord. n.

Family Teloproctidae Cavalier-Smith in Cavalier -Smith et al., 2016 a (Teloprocta )

Family 1. Euglenidae Dujardin, 1841 em. Kim et al., 2010

Subclass 2. Metakinetoplastina Vickerman in Moreira et al., 2004

0.64/- Entosiphon oblongum type strain

0.95/60 0.45/- Cryptoglena

Table 2. Sensitivity of Postgaardea placement to sequence sampling and substitution model.

Nucleotide positions and algorithm No Percolozoa in outgroup Percolozoa in outgroup

1425 positions CAT E (0.40) E (0.43) E (0.52) E (0.40)

Nucleotide positions and algorithm No Percolozoa in outgroup Percolozoa in outgroup

1425 positions CAT T: EP (0.66) EP (0.65) A (0.54) EP (0.39)

18 orders 8 classes 3 subphyla

Derived Nature of Petalomonad Feeding Acknowledgements

You might also like