The Phylogenetic Position of

Entoprocta, Ectoprocta, Phoronida, and

Brachiopoda1
Claus Nielsen
Integr Comp Biol (2002) 42 (3): 685-691.
DOI:
https://doi.org/10.1093/icb/42.3.685
Published:
01 July 2002

Views

PDF

Cite

Share

Tools

Abstract
Ectoprocts, phoronids and brachiopods are often dealt with under the heading
Tentaculata or Lophophorata, sometimes with entoprocts discussed in the same
chapter, for example in Ruppert and Barnes (1994). The Lophophorata is purported to
be held together by the presence of a lophophore, a mesosomal tentacle crown with
an upstream-collecting ciliary band. However, the mesosomal tentacle crown of
pterobranchs has upstream-collecting ciliary bands with monociliate cells, similar to
those of phoronids and brachiopods, although its ontogeny is not well documented.
On the contrary, the ectoproct tentacle crown carries a ciliary sieving system with
multiciliate cells and the body does not show archimery, neither during ontogeny nor
during budding, so the tentacles cannot be characterized as mesosomal. The
entoprocts have tentacles without coelomic canals and with a downstream-collecting
ciliary system like that of trochophore larvae and adult rotifers and serpulid and
sabellid annelids. Planktotrophic phoronid and brachiopod larvae develop tentacles at
an early stage, but their ciliary system resembles those of echinoderm and
enteropneust larvae. Ectoproct larvae are generally non-feeding, but the
planktotrophic cyphonautes larvae of certain gymnolaemates have a ciliary band
resembling that of the adult tentacles. The entoprocts have typical trochophore larvae
and many feed with downstream-collecting ciliary bands. Phoronids and brachiopods
are thus morphologically on the deuterostome line, probably as the sister group of the
Neorenalia or Deuterostomia sensu stricto. The entoprocts are clearly spiralians,
although their more precise position has not been determined. The position of the
ectoprocts is uncertain, but nothing in their morphology indicates deuterostome
affinities. Lophophorata is thus a polyphyletic assemblage and the word should
disappear from the zoological vocabulary, just as Vermes disappeared many years
ago.

Issue Section:
Regular Article

INTRODUCTION
The four phyla discussed here have not found their established positions in the
zoological system (see Fig. 1). They are all sessile and use ciliated tentacles in filter
feeding. However, pterobranchs also show these characteristics, and it is necessary to
discuss the four phyla more or less separately to assess their relationships (a more
extensive discussion with detailed references can be found in Nielsen, 2001). Many
recent textbooks unite Ectoprocta (Bryozoa), Phoronida and Brachiopoda under the
name Lophophorata, a name introduced by Hyman (1959), apparently because she
did not like the much older name Tentaculata. Her definition of the group was based
on their common possession of a lophophore, i.e., a crown of ciliated mesosomal
tentacles surrounding the mouth but not the anus. Her definition logically includes the
pterobranchs, but she (Hyman, 1959, p. 77) just stated that The[ir] tentaculated arms
are not to be regarded as a lophophore although no doubt phylogenetically related to
that structure. On the other hand, as pointed out earlier (Nielsen, 2000, 2001), the
tentacle crown of the ectoprocts shows only superficial similarities with that of
phoronids and brachiopods; ectoproct tentacles have lateral ciliary bands composed
of multiciliate cells, lack a longitudinal haemal vessel, and are probably not
mesosomal, whereas phoronid, brachiopod and pterobranch tentacles are monociliate,
have a longitudinal haemal vessel, and are probably mesosomal. Thus, it appears that
if the term lophophore is to be retained, it should be used for the tentacle crowns of
phoronids, brachiopods and pterobranchs.

Morphology-based studies on metazoan phylogeny do not agree on the position of the

four phyla discussed here; the phylogeny which I favour and a phylogeny which most
molecular biologists will probably favour are shown in Figure 1. The main
discrepancy between the positions of the four phyla on the two trees is the position of
phoronids and brachiopods, which are regarded as deuterostomes in the morphology-
based tree, but as protostomes in the molecular-based tree.

ENTOPROCTA
Entoprocts are undoubtedly protostomes. They show a number of characters which
are found only in spiralians, but their position within that group is difficult to assess.

The cleavage is spiral, and mesoderm is reported to develop from the 4d-cell; the two
first cleavages divide the embryo into four quadrants, called A, B, C, and D (Marcus,
1939; Malakhov, 1990; Fig. 2). Gastrulation is described as embolic, but the funnel-
shaped oesophagus (stomodaeum) is formed at about the same time, and more
detailed studies are needed to make the distinction between these two structures quite
clear. The larva is a typical trochophore with apical organ, prototroch and metatroch
of compound cilia, adoral ciliary zone and gastrotroch of separate cilia, and a pair of
protonephridia; only the telotroch is missing (Nielsen, 1971, 1979). The apical organ
resembles that of other spiralians, with paired nerves and muscles to the
oesophagus/prototroch area and a pair of serotonergic cells (Nielsen, 1971; Hay-
Schmidt, 2000; Fig. 3). There is a primary body cavity which functions as a
hydrostatic skeleton (Nielsen, 1971). Some species of Loxosoma and Loxosomella
have larvae with a long planktotrophic stage, but the development of these larvae
through the planktonic stage to metamorphosis has not been followed. Particle
collecting has not been described in detail, but Jgersten (1964) observed particles
being caught by cilia of the prototroch and carried along the food groove to the
mouth. So both structure and function appear to conform to the general picture of the
downstream-collecting mechanism characteristic of spiralian larvae (Riisgrd et al.,
2000). Most other entoproct species have a short, free-swimming larval stage and the
larvae are apparently competent to settle shortly after liberation (Nielsen, 1971).

Settling and metamorphosis show enormous variation between species (Nielsen,

1971). Settling with the frontal organ appears to be the ancestral type, because it
involves the least morphogenetic change, whereas the types with budding from a
larval body which disintegrates appear highly derived. Larvae of the colonial forms
settle with the ring of cells just apical to the retracted prototroch, glued to the
substratum by secretions from three sets of glands in the foot (gastrotroch). The
apical organ is lost after metamorphosis and a new brain develops from the
epithelium of the ventral side of the atrium. Also the ciliary bands disintegrate, and
degenerating prototroch cells each with a bundle of slowly beating cilia can often be
seen in the narrow primary body cavity of the newly settled juveniles (Nielsen, 1971).

The adult loxosomatids are solitary, whereas the other families comprise only
colonial forms. The adult feeding structures, i.e., tentacles with a downstream-
collecting ciliary system, develop anew after metamorphosis, but could possibly have
been derived as loops on the prototroch in the ancestor. Structure and function of the
ciliary bands are very similar to those of the larvae, except that the two opposed
bands of compound cilia are of equal size (Riisgrd et al., 2000). There is a narrow
primary body cavity with protonephridia and few mesodermal elements (Nielsen and
Jespersen, 1997).

The only character which is deuterostome-like is the loss of the apical organ at
metamorphosis; this is typically seen in all non-chordate deuterostomes. However, the
structure of the apical organ is of the protostome type and sessile organisms often
show highly modified nervous systems, so this character probably represents a
convergence.

No entoproct character indicates a close relationship with any specific spiralian

phylum, so until further knowledge has been gathered it seems practical to place them
at the base of the spiralian tree in a polytomy with Schizocoelia, Parenchymia and
Gnathifera (Nielsen, 2001; Fig. 1).

ECTOPROCTA
The ectoproct bryozoans are perhaps the most puzzling phylum in phylogenetic
studies of the Bilateria. The strong influence from the treatise of Hyman (1959),
where they are classified as lophophorates, seems more or less to have stifled
discussions of alternative phylogenetic positions. Even some quite recent authors
have regarded the lophophorates as placed between protostomes and
deuterostomes, which is not in agreement with modern phylogenetic systematics.

Early ectoproct development is poorly known. Cyclostomes and phylactolaemates

have very peculiar embryologies, which must be considered highly derived.
Gymnolaemates show more usual developmental types, with a few species having
planktotrophic cyphonautes larvae (in genera such as Membranipora, Electra and
Alcyonidium) whereas most other species have more or less modified lecithotrophic
larvae (Hayward and Ryland, 1998; Cadman and Ryland, 1996). The early cleavage
shows an unusual, biradial pattern, which could perhaps be considered as modified
spiralian, but the planes of the cleavage have not been related to the planes of the
larva. The origin of the mesoderm is uncertain, but there is no sign of mesothelia or
coelomic sacs. The fully formed cyphonautes is laterally compressed with a pair of
shells; it swims with the corona, a ring of multiciliate cells with separate cilia situated
along the lower edge of the shells, surrounding a concave ventral side or vestibule.
The corona could perhaps be interpreted as a modified prototroch; its situation and
normal direction of beat corresponds to that of a trochophore, but the cilia are not
compound and the direction of the beat can be reversed, a feature which is apparently
not known from bands of compound cilia in spiralians (Nielsen, 1987). The vestibule
is divided into a larger, anterior inhalant chamber and a smaller, posterior exhalant
chamber by a U-shaped ciliated ridge, which carries three rows of cilia: Lateral cilia
at the lateroposterior side of the ridge create a water current across the ridge from the
anterior to the posterior chamber; laterofrontal cilia apparently function as a
mechanical filter which intercepts the particles carried by the water flow and in some
unknown way transfers them to frontal cilia at the anterior side of the ridge; the
particles are then in some way transported to the mouth (Strathmann and McEdward,
1986). This filter-feeding structure resembles one side of the tentacle of the adult (see
below). The apical organ is of the protostome type with nerves and muscles leading to
the prototroch and a pair of serotonergic cells (Stricker et al., 1988a; Hay-Schmidt,
2000; Fig. 3). A pair of ciliated tubes have been observed in each side of the larval
body (Stricker et al., 1988b); they could possibly be derived protonephridia, but more
detailed studies are needed.

Settling and metamorphosis show much variation. The cyphonautes larva settles with
an everted adhesive sac which develops into the epithelium of the underside of the
ancestrula, whereas the upper epithelium derives from the ectoderm carrying the
shells. Other types, such as Bowerbankia and Tanganella (Reed, 1991) have
metamorphoses resembling that of the colonial entoprocts, with a glandular adhesive
sac secreting the adhesive which glues the ancestrula to the substratum after
retraction of the hyposphere. Larval organs such as apical organ and gut degenerate,
the first polypide is formed from a blastema of varying origin (Zimmer, 1997), and
further zooids in the colony develop through budding. This makes it impossible to
infer a dorsoventral orientation of any zooid in a colony. In the zooids, the mesoderm
becomes organized as a coelomic epithelium surrounding a large body compartment
and a smaller compartment surrounding the mouth and sending canals into the
tentacles, but the two compartments are in open connection in an area behind the
mouth (Nielsen, 2000). This origin and adult organization of the body cavity bears no
resemblance to the archimery of the deuterostomes, including phoronids and
brachiopods. The zooids have no nephridia.

The ciliary bands show some variation between the main groups, but particle capture
by the ciliary sieving method has been observed in cyclostomes and
gymnolaemates. Cyclostomes have the most simple tentacles with only lateral and
laterofrontal cilia. The lateral cilia create a water current towards the mouth and out
of the tentacular funnel between the tentacles, and particles are strained from the
water by the filter formed by the stiff laterofrontal cilia. The drag on the captured
particle elicits a tentacle flick, which pushes the particle to the central current
carrying the particle to the mouth (Nielsen and Riisgrd, 1998). Gymnolaemates have
tentacles with lateral, laterofrontal and frontal ciliary bands; they use the same ciliary
sieving method as the cyclostomes, but have a more varied behavioural repertoire
(Riisgrd and Manrquez, 1997). The phylactolaemates are in need of further study.
The ciliary sieving method of the ectoprocts (gymnolaemates and stenolaemates)
seems to be unique in the animal kingdom.

Neither ontogeny nor adult structure of ectoprocts shows any specific similarities
(synapomorphies) with phoronids and brachiopods, so the traditional classification of
these three phyla as lophophorates finds no support in morphology, and it is not
supported by molecular analyses either.

PHORONIDA AND BRACHIOPODA

Phoronids and brachiopods are undoubtedly closely related. As mentioned above, the
two phyla are regarded as deuterostomes by some morphologists, whereas most
molecular biologists regard them as protostomes.

The two phyla are usually regarded as sister groups when morphological data are
analysed (e.g.,Carlson, 1995; Giribet et al., 2000; Nielsen, 2001), but many molecular
analyses resolve the phylogeny with phoronids as an ingroup of the brachiopods,
grouping variously with Craniida, Articulata and Inarticulata (Cohen and
Gawthrop, 1997; Zrzav et al., 1998; Giribet et al., 2000); this finds no support in
morphology.

Brachiopods have a number of characters which are not shared with the phoronids
and which must be regarded as autapomorphies: Two shells, a brachial and a pedicle
valve, secreted by characteristic mantle folds, which are extensions of the metasome
and contain metacoelomic mantle canals; rows of setae, each secreted by one cell,
along the mantle edges; a short ventral side, as shown by the ontogeny (Nielsen,
1991); and two coelomic systems in the lophophore, a large brachial canal which is
restricted to the base of the lophophore and a small brachial canal which sends a canal
into each tentacle and which represents the mesocoel (at least in Neocrania, as
demonstrated by its ontogeny, see Nielsen, 1991). On the contrary, phoronids lack
mantle folds with setae and valves, have a short dorsal side (as shown by their
ontogeny), and have only one (mesosomal) cavity in the lophophore.

On the other hand, two characters indicate a sister-group relationship of the two
phyla. Their planktotrophic larvae have long, ciliated tentacles similar in structure
and function to those of the adults and forming a postoral horseshoe which is almost
closed near the apical organ in the area where new tentacles develop; these tentacles
directly become the juvenile tentacles at metamorphosis in most species. The adults
have one (or two) pair(s) of metanephridia with a large ciliated funnel; these also
serve as gonoducts. These characters are not seen in the Neorenalia (=Deuterostomia
sensu stricto), although this can of course not be stated categorically for the larval
tentacles, because all pterobranchs studied so far have direct development.

A whole series of characters distinguish the Deuterostomia (sensu lato,i.e., including

phoronids and brachiopods) from the Protostomia (Nielsen, 2001).

The first two cleavages are generally median and transverse, whereas they form a
cross dividing the embryo into right, anteroventral, left, and posterodorsal cells in
most spiralians (Fig. 2). Among neorenalians, only few exceptions are known, the
best example being the sea urchin Strongylocentrotus (Davidson et al., 1998).
Phoronids have a somewhat labile pattern, but the first cleavage is transverse and the
second one median in most of the embryos (Freeman, 1991). Brachiopods show
considerable variation: Terebratalia and Neocrania show no fixed relationship
between early cleavage planes and adult body symmetry (Freeman, 1993, 2000),
whereas in Glottidia and Discinisca, the first cleavage is median in almost all
embryos (Freeman, 1995, 1999). The third cleavage separates apical and blastoporal
regions in all species.

Regulative powers are generally considerable in deuterostome embryos and very

limited in protostomes, which are generally incapable of regulating, with a few being
able to regulate at the 2-cell stage. Phoronid embryos are generally able to regulate
when bisected along the primary axis between the 2-cell stage and the gastrula stage;
apical halves of 8-cell to late blastula stages develop into ciliated spheres, whereas
the blastoporal halves develop almost normally (Freeman, 1991). Studies of the
brachiopods Terebratalia, Neocrania, Glottidia, and Discinisca (Freeman, 1993,
1995, 1999, 2000) have demonstrated some variations, but the general pattern is that
isolated blastomeres of the 2-cell stages and halves of the 4-cell stage show complete
regulation. Apical halves of blastulae develop into ciliated vesicles, whereas the
blastoporal halves developed into miniature larvae with small apical lobes.
Experiments with halving blastulae along the primary axis showed that the bilaterality
becomes established during this stage and that the power of regeneration becomes
lost. The regulative powers of phoronids and brachiopods resemble those of the sea
urchins (Hrstadius, 1939) in very many details.

The mesoderm of protostomes and deuterostomes typically originates from different

sources, with protostomes having mesoderm developing from the blastopore lip/4d-
cell (and ectomesoderm) whereas deuterostomes have the mesoderm originating from
the archenteron (in some cases in the shape of mesodermal pouches, i.e., by
enterocoely, but this is not essential). Mesoderm formation is not well known in
phoronids, but Zimmer (1980, 1991), Emig (1977) and Herrmann (1986) agree that
mesoderm proliferates from the dorsal and lateral sides of the anterior archenteron
and forms the protocoel; meso- and metacoel become organized at later stages from
cells of uncertain origin. This is very similar to the formation of the mesodermal sacs
in the enteropneust Ptychodera flava (Peterson et al., 1999). In brachiopods, the
mesoderm develops from the archenteron, through differentiation of particular
regions of the archenteron in Crania, Discinisca, Glottidia, and articulates (Nielsen,
1991; Freeman, 1995, 1999; Long and Stricker, 1991), although Freeman (2000)
described a quite deviant type of differentiation in Crania. Thus, the origin of the
mesoderm is clearly of the deuterostome type in phoronids and brachiopods.

The larval apical organ contains a kidney-shaped group of many serotonergic cells in
Phoronis and an elongate group on the median tentacle of the brachiopods Glottidia
and Lingula, whereas the protostome larvae have only 23 serotonergic cells (Hay-
Schmidt, 1990, 1992; Fig. 3).

Archimery, the organization of the body in three regions each with characteristic
coelomic compartments, is clearly recognized during ontogeny and in the full-grown
larvae of Crania (Nielsen, 1991), but the protocoel is not well developed in phoronids
and has not been identified with certainty in adult brachiopods. However, both meso-
and metacoels have been followed from the larval stages through metamorphosis to
the adults, so their identity cannot be doubted.

The ciliary bands on the phoronid and brachiopod tentacles are upstream-collecting
bands with separate cilia on monociliate cells just like those of larvae or adults of
pterobranchs, echinoderms and enteropneusts (Nielsen, 1987). The particle-collecting
method is generally believed to involve a reversal of the ciliary beat (Strathmann,
1973), but further investigations are needed.

All these characters indicate that phoronids and brachiopods belong to the
deuterostome line. I cannot identify any unequivocal protostome character, but
blastopore fate and protonephridia of phoronids are among the characters which could
indicate protostomian affinities (or be ancestral).

DISCUSSION
The above summaries have concentrated on morphological characters, and it appears
that entoprocts and ectoprocts are protostomes (and possibly closely related), whereas
phoronids and brachiopods share a number of key characters with the other
deuterostomes (Figs. 23; Table 1).

Molecular studies support the monophyly of Phoronida + Brachiopoda but almost

invariably place them within the Lophotrochozoa, which in most studies
comprising many species appear as a complete hotchpotch of ciliated protostomes
(Winnepenninckx et al., 1998; Halanych, 1998; Giribet et al., 2000). However, it
should be stressed that whereas the Deuterostomia sensu stricto is recognized as a
monophyletic group with a number of well defined phyla in a rather constant
phylogenetic pattern, the protostomes do not show a clear or consistent pattern.
Analyses including a variety of annelids and molluscs do not show Annelida and
Mollusca as monophyletic (e.g.,Giribet et al., 2000; Peterson and Eernisse, 2001),
and the Ectoprocta are usually polyphyletic too. To me this indicates that the 18S
rDNA molecule does not contain enough information to resolve the old speciation
events at the base of the Bilateria, and that only the branch leading to Deuterostomia
sensu stricto has been isolated (i.e., given off no surviving groups) for so long that it
can be recognized in the analyses.

Morphologically it is rather easy to define both the Deuterostomia sensu lato and the
Deuterostomia sensu stricto (=Neorenalia), whereas I find it impossible to identify
any synapomorphies of a Protostomia sensu lato,i.e., including phoronids and
brachiopods; all characters that can be listed are symplesiomorphies, viz.
deuterostome synapomorphies which are absent, and this supports the impression of a
paraphyletic Protostomia sensu lato. The only logical position of the Protostomia-
Deuterostomia split therefore places Phoronida + Brachiopoda in the Deuterostomia
(sensu lato).

Although no firm conclusion can be drawn about the phylogenetic position of the
ectoprocts, it can be stated that there is no indication from morphology or molecules
in favour of uniting them with phoronids and brachiopods. In consequence, the word
Lophophorata must be removed from the phylogenetic vocabulary, just as the word
Vermes was abandoned several decades ago.

https://academic.oup.com/icb/article/42/3/685/724040/The-Phylogenetic-Position-of-
Entoprocta-Ectoprocta