ArXiv-Publication
MICHAEL GUDO & TAREQ SYED
Frankfurt am Main, 13.11.2008
100 Years of Deuterostomia (GROBBEN, 1908):
Cladogenetic and Anagenetic Relations within the Notoneuralia Domain
MICHAEL GUDO1 & TAREQ SYED2
Abstract
Results from molecular systematics and comparative developmental genetics changed the picture of metazoan and especially bilaterian radiation. According to this “new animal phylogeny” (introduced by
ADOUTTE et al. 1999/2000), GROBBEN´S (1908) widely favoured protostome-deuterostome division of the
Bilateria can be upheld, but only with major rearrangements within these superphyla. On the cladogenetic
level, the Protostomia are split into two unexpected subgroups, the Lophotrochozoa and Ecdysozoa. The
deuterostomes are split into the subgroups Chordata and Ambulacraria, which is not novel since GROBBEN
(1908) introduced the Deuterostomia in this way (together with the Chaetognatha as a third line).
However, many details of the new deuterostome phylogeny do not fit traditional, morphology-based
reconstructions. As a consequence, three relatively unexpected proposals for early deuterostome evolution
are favoured today: An ambulacraria-scenario, a xenoturbellid-scenario, and a chordate-scenario. The first
two proposals are often discussed in the literature, while the chordate-scenario is almost completely neglected. Therefore, the paper presented focuses on the chordate scenario, i.e. the hypothesis of an acrania-like
“ur-deuterostomian”. It is argued that the “acrania-hypothesis” is clearly preferable when biomechanic options of a polysegmented, hydroskeletal body plan are taken into account. The so called hydroskeleton hypothesis, rooted in the work of W. F. GUTMANN, is the most detailed anagenetic scenario which depicts an
acrania-like ur-deuterostome. Moreover, it is the only morphology-based hypothesis which is in line with
all of the unexpected molecular results of deuterostome evolution (i.e. pterobranchs, echinoderms and tunicates as highly derived forms which secondarily lost the ancestral polysegmentation).
Zusammenfassung
Seit der wiederholten Bestätigung der “new animal phylogeny” (einer molekularsystematischen, von
ADOUTTE et al. 1999/2000 ermittelten Großphylogenie des Tierreiches) ist klar geworden, dass innerhalb
der traditionellen Großeinteilung der Bilateria in Proto- und Deuterostomia (GROBBEN 1908) einige Äınderungen vorgenommen werden müssen. Kennzeichnend für die Protostomia ist eine Neueinteilung in
Ecdysozoa und Lophotrochozoa, kennzeichnend für die Deuterostomia eine Einteilung in Chordata und
Ambulacraria – letztere wurde allerdings in sehr ähnlicher Form ebenfalls von GROBBEN (1908) postuliert.
Viele Einzelresultate innerhalb dieser molekularsystematischen Großeinteilung widersprechen morphologischen Rekonstruktionen. Für die Rekonstruktion der Deuterostomier-Evolution hat dies zur Folge,
dass nur noch von drei möglichen Szenarien ausgegangen wird, welche früher allesamt als Außenseiterpositionen galten. Es sind dies ein “Ambulacraria-Szenario”, ein “Xenoturbelliden-Szenario” und ein “Chordaten-Szenario”. Das Chordaten-Szenario ist dabei in der neueren Literatur kaum detailliert worden. Im
vorliegenden Artikel wird dies mit Rückgriff auf die sogenannte Hydroskelett-Theorie getan, die auf Arbeiten von W. F. GUTMANN basiert. Das dort ermittelte Szenario eines Acranier-artigen “Ur-Deuterostomiers”
hat zwei beachtenswerte Vorteile: Erstens nimmt es sämtliche unerwarteten molekularbiologischen Einzelresultate für die Deuterostomier vorweg (Hemichordaten, Echinodermen und Tunicaten als hochabgeleitete
Formen), zweitens ist es im Laufe von über vierzig Jahren detaillierter ausgearbeitet worden als andere
Modelle zur Deuterostomier-Evolution. Aus diesen Gründen ist es auch für eine Re-Interpretation fossiler
Formen von Interesse (im Text exemplarisch für die Vetulicolia und frühe kambrische Chordaten demonstriert).
1)
Dr. MICHAEL GUDO, Morphisto Evolutionsforschung und Anwendung GmbH, Senckenberganlage 25, 60325 Frankfurt am Main, Germany, e-mail: [email protected]
2)
Dr. TAREQ SYED, Morphisto Evolutionsforschung und Anwendung GmbH, Senckenberganlage 25, 60325 Frankfurt am Main, Germany, e-mail: [email protected]
Hundred Years of Deuterostomia
1. Introduction
Since KARL GROBBEN introduced the superphylum Deuterostomia one hundred years ago (GROBBEN 1908), there have
been intense debates about validity, members and the exact
phylogenesis of this group. Today, molecular systematics and
developmental genetics offer previously unavailable clues
which, however, seem to conflict with almost all morphologybased interpretations of the Deuterostomia (for a review see
GEE 2001, and also WINCHELL et al. 2002). The vast majority
of these morphology-based reconstructions rely on comparisons of isolated characters instead of functional character complexes. To resolve some of the ambiguities, the present paper
aims at two goals. First, and most important, we will demonstrate that phylogenetic reconstructions accomplished by a
function-focused approach called constructional morphology
(German “Konstruktionsmorphologie”; SCHMIDT-KITTLER &
VOGEL 1991; GUTMANN 1993; GUDO et al. 2002) are in much
better agreement with molecular data than results from structure-based, morphological analyses. Second, we will emphasize that methods which reconstruct anagenetic pathways (i.e.
successive morphological transformation steps during evolution; RENSCH 1972) should be based on independent structuralfunctional approaches and not on mere character sorting –
however sophisticated methodologically – as in cladistic
reconstructions (pure genealogies). Such anagenetic step-bystep reconstructions are potentially helpful to interprete fossile
problematica (which will casually be demonstrated here with
regard to the cambrian Vetulicolia (Deuterostomia incerta
sedis) and Haikouella-fossils).
2. The “New Animal Phylogeny”: Results
and conflicts with traditional phylogenies
Over the past ten years, molecular phylogenetics changed textbook views on metazoan bauplan evolution. In the "New Animal Phylogeny" (NAP) introduced by ADOUTTE et al. (1999,
2000), traditional major domains of the animal kingdom were
rearranged, apparently well-established groups split, and commonly accepted evolutionary assumptions rejected (fig. 1). On
the cladogenetic level, the bilateria were devided into three
superphyla, the Deuterostomia, Ecdysozoa, and Lophotrochozoa; later, the Acoelomorpha were added as an early side
branch (COOK et al. 2004; JIMÉNEZ-GURI 2006). On the anagenetic level, the possibility of a polymer-segmented urbilaterian
(probably growing by “terminal addition” during embryogenesis; JACOBS et al., 2005) gained increasing interest, mainly on
the basis of comparative studies of the Hox gene system (e.g.
BALAVOINE et al. 2002; PRUD´HOMME et al. 2003; SEO et al.
2004; DEROSA et al. 2005). This would imply that the last common ancestor of deuterostomes, ecdysozoans and lophotrochozoans can be envisaged as a coelomate annelid-like animal,
and that in many bilaterian lines coelom as well as segments
became lost secondarily (BALAVOINE & ADOUTTE 2003). For
at least four reasons, this new view is in conflict with almost all
morphology-based phylogenies: (1) Platyhelminthes split in a
quite unexpected way (Acoelomorpha very basal, all other Platyhelminthes very derived; fig. 1); (2) Lophophorata neither
10
Fig. 1. A condensed view on the main cladogenetic and anagenetic
aspects of the new animal phylogeny (modified from PRUD´HOMME et
al. 2003; Acoelomorpha added according to COOK et al. 2004, and
JIMÉNEZ-GURI et al. 2006). On the cladogenetic level, molecular systematics confirmed the traditional group of the Deuterostomia, however,
Chaetognatha (arrow worms, not shown) and Lophophorata (comp.
position of brachiopods) are now excluded from this taxon. Lophophorata group within the new protostome-superphylum Lophotrochozoa.
The remaining protostomes are phyla of periodically moulting animals,
the Ecdysozoa. On the anagenetic level, similarities in the genetic control of body segmentation suggest a polysegmented urbilaterian and
loss of this feature in all bilaterian groups except the Acoelomorpha,
which represent an early and isolated side branch.
are closely related to deuterostomes nor do they represent basal
bilaterians (they are not even monophyletic, HALANYCH 2004,
DUNN et al. 2008); (3) Annelida and Arthropoda can no longer
be united in a taxon Articulata; (4) the bilaterian tree does not
imply an overall increase of “morphological complexity”, i.e.
gradual transformation lines such as acoelomate=>pseudocoelomate=>coelomate (anagenesis of a secondary body cavity) or
amer=>oligomer=>metamer/polymer (anagenesis of a polysegmented trunk; traditionally rated as parallel events in Chordata and Articulata) are not supported by the NAP (ADOUTTE
et al. 1999).
2.1 Consequences of the new protostome phylogeny
The origin of the NAP can be traced back to the introduction of
the new protostome superphyla Lophotrochozoa and Ecdysozoa in 1995 and 1997, respectively (comp. reviews in
ADOUTTE et al. 1999, 2000; and HALANYCH 2004). After the
exclusion from the Deuterostomia of the Chaetognatha,
Pogonophora and the three lophophorate phyla (Phoronida,
Brachiopoda, and Bryozoa), there now are far more protostome
than deuterostome lines. GERHART (2006, p.678) recognizes 25
protostome phyla, whereas deuterostomes consist of only 6
phyla (fig. 2b and next section). Here we refrain from exploring protostome radiation; molecular analyses of protostome
relationships will be mentioned only if they bear on our subsequent discussion of deuterostome phylogeny.
Obviously, if the anagenetic scenario depicted in fig. 1 is
correct, then an annelid-like body plan (gradually achieved
from a common, unsegmented ancestor of acoelomorphs and
MICHAEL GUDO & TAREQ SYED, 13. November 2008
early annelids, see also fig. 3) will be the key to early deuterostome evolution. Some supporters of the new phylogeny,
though, doubt the significance of the annelid body plan (e.g.
HALANYCH 2004, p.245). However, they neglect the striking
similarities in the structure and function of Hox-genes and
other “segmentation genes” that function as developmental
regulators on a higher hierarchical level (see for example
PRUD´HOMME et al. 2003 and DEROSA et al. 2005) without
offering an evolutionary explanation. Another reason for rejecting the hypothesis of a polysegmented urbilaterian is that this
scenario implies a multiple, “non-parsimonious” loss of segmentation in several bilaterian lines (e.g. JENNER, 1999; comp.
also fig. 1). On the other hand, BALAVOINE & ADOUTTE (2003)
pointed out that many of these non-segmented phyla possess
characters that can be interpreted as traces of a former segmentation, and that comparative developmental genetics should be
applied for further clarification (as in the case of the
widespread “terminal addition” mode of growing, JACOBS et
al. 2005).
A significant feature of the “polysegmented urprotostomian” concept is that only in this case the classical anagenetic
model of arthropods evolving from annelid-like precursors can
be upheld, despite the abolition of the taxon Articulata. The
“polysegmented urprotostomian” concept also implies the derivation of the Acrania and Craniota from annelid-like precursors, an old idea which has often been critized by pointing at
the ventrally localized nervous system of annelids (gastroneural condition), which contrasts with the dorsal nerve tube of
the Acrania and Craniota (notoneural condition). However, this
argument is valid only if two currently existing body plans are
forced into a phylogenetic sequence (for example, by turning
annelids upside down to reach the chordate-condition), a procedure which is anti-evolutionistic and forms the root of countless misconceptions in evolutionary biology. In our reconstruction, we prefer to follow RUPPERT (2005, p.12) and explore the
implications and consequences of the assumption that a common ancestor of Gastroneuralia and Notoneuralia was a polysegmented animal without a highly specialized nervous
system. Also LOWE (2008: 1575) points out “that a hypothetical ancestor was not necessarily characterized by a central nervous system” (which of course does not exclude the possibility
of a common ancestry of bilaterian nervous system centralization, see DENES et al. 2007). Quite obviously, distinct architectures of the neural systems are possible at the end-points of different anagenetic pathways that originate from this common
ancestor (thus, by using HATSCHEK´S (1891) terms Gastroneuralia and Notoneuralia we refer to existing groups rather
than their ancestors). In this case, the question of deuterostome
origins becomes linked to the question of the gradual evolution
of the notoneural condition. It can best be answered by discussing probable evolutionary advantages of a dorsal muscle concentration in the line that leads to the earliest deuterostomes,
which in turn requires sound biomechanical reconstructions of
hypothetical ancestors (the constructional approach; see section 3).
Fig. 2. Two distinct but partly similar views on deuterostome cladogenesis and anagenesis. 2A. Karl GROBBEN introduced the term “Deuterostomia” for the subgroups Chordonia, Ambulacralia and Chaetognatha. Deuterostomia and Protostomia (not shown) originate from a common coelomate precursor (detail of a phylogenetic tree from GROBBEN 1908). 2B. Molecular systematics reanimated the often rejected idea of a ChordataAmbulacraria dichotomy in the Deuterostomia. In contrast to 2.A., Chaetognatha are excluded from the Deuterostomia; instead, the monotypic
phylum Xenoturbellida is now regarded as a basal branch of the Ambulacraria. Following fig. 1A, a segmented, probably coelomate ur-deuterostomian is suggested. Gill slits are present in ur-deuterostomes, and are secondarily lost in the Xenoturbellida and recent echinoderms (carpoids are
extinct echinoderms with gill slits). The double arrow indicates that the position of Tunicata (urochordates) and Acrania (cephalochordates) has to
be changed according to newer results. Independent of these modifications, the scheme implies a secondary loss of segmentation in tunicates, and
also in xenoturbellids and echinoderms (modified from GEE 2001; Xenoturbellida added following BOURLAT et al. 2006).
11
Hundred Years of Deuterostomia
2.2 Consequences of the new deuterostome
phylogeny
Molecular systematics of the Deuterostomia unanimously support the division of this superphylum into Chordata and Ambulacraria (fig. 2b; see HALANYCH 2004 for review). Together
with the exclusion of the lophophorate phyla from the deuterostomes, this “new” picture somehow resembles
GROBBEN´S (1908) introduction of the Deuterostomia (fig. 2a).
However, GROBBEN´S inclusion of the Chaetognatha (arrow
worms) into the Deuterostomia is now refuted by molecular
results (HALANYCH 2004; MARLÉTAZ et al. 2006, DUNN et al.
2008). Instead, the flatworm-like Xenoturbella bocki became a
new member of the deuterostomes, probably as the basic
ambulacrarian line; this rearrangement contradicts former
views of echinoderms as basal ambulacraria (comp. fig. 3b and
STACH et al. 2005; BOURLAT et al. 2006; DUNN et al. 2008, but
also PERSEKE et al. 2007). Another surprise came with the
molecular phylogeny of the hemichordates, because here the
Pterobranchia appear as a highly derived line (CAMERON et al.,
2000) that probably originated from enteropneust-like precursors; this reverses the relationship favored by most morphology-based phylogenies. Similarly, the division pattern along
the Chordata-branch was rearranged (comp. fig. 2b). Instead of
the previously almost universally accepted Acrania/Craniota
taxon, there now is support for a sister-group relationship
between Tunicata and Craniota (PHILIPPE et al. 2005; BOURLAT et al. 2006; DELSUC et al. 2006; WADA et al. 2006, DUNN
et al. 2008; for a contradicting interpretation, see MALLATT &
WINCHELL 2007). In any case, tunicates now are seen as a
highly specialized side branch, and no longer as representatives of the ancestral chordates, due to their reduced genome
size and dispersed Hox cluster (e.g. IKUTA et al. 2004; SEO et
al. 2004; HUGHES & FRIEDMAN 2005). The derived status of
the Pterobranchia and Tunicata leaves Xenoturbellida,
Enteropneusta, and Acrania as the only possible representatives of the earliest deuterostome line. In the context of the idea
of a polysegmented ancestor, the Acrania would appear the
Fig. 3. Functional design of the ur-deuterostomes
and ur-protostomes. Here, the last common ancestor of deuterostomes and protostomes is reconstructed as a worm-like animal with a metameric
hydraulic skeleton functioning as force absorber
and force transmitter (evolving from a compact,
gelatinous bilaterian with inner canal systems).
Left side: Through specific coordinations of the
activity of circular and longitudinal muscles, lateral
undulations could be generated which resulted in
swimming movements. The cranial region of this
ancestral animal was designed as a filtering device,
giving rise to the branchial apparatus which is a
common structure in deuterostomes (“filter coelomates”). Right side: The alternative design is presented by creeping, benthic forms which developed
a sucking pharynx for taking up particles from the
substratum (“worm coelomates”). The pelagic
forms developed a dorsally concentrated nervous
system (innervation of muscles around the chorda
dorsalis, comp. fig. 4), while in the benthic forms a
ventral concentration of nerves took place (probably in connection with parapodial movement).
12
best candidates for relatively basal deuterostomes due to their
polysegmented body (chorda-myomere system). However,
many developmental biologists favour an enteropneust-like
condition at the basis of the clade, arguing that the dorsoventral
expression pattern of bmp and chordin genes is protostomelike in enteropneusts (e.g. GERHART 2006, LOWE 2008). If this
assumption is correct, the polysegmentation of the urbilaterian
became reduced during evolution of an enteropneust-like urdeuterostomian, followed by a “re-polymerization”-process in
the line leading to the chordates. In the alternative scenario, an
acrania-like ur-deuterostomian evolved from the polysegmented urbilaterian, and polysegmentation vanished in the
lines leading to the ambulacraria and tunicata (fig. 2b). This
would imply that the protostome-like bmp-chordin-axis of the
enteropneusts is a case of evolutionary parallelism that probably occurred during a process of strong simplification and
modification. Both of these scenarios have their merits, at least
from a developmental biologist´s point of view, and further
studies must be awaited (see also section 5.3). For example, if
a bmp-chordin-axis could be demonstrated in the xenoturbellids, with the mouth located at the chordin side (as in craniotes), this could weaken the view that the enteropneust condition is ancestral in deuterostomes (obviously, this argument
also depends on the systematic position of Xenoturbella, which
is relatively uncertain at the moment – comp. 5.1. for details).
There appears to be agreement regarding the ancestral
deuterostome organization in one crucial point: nearly all
developmental genes so far studied in comparative analysis of
enteropneusts and acrania/craniota are not expressed in larval
or juvenile stages; their activation starts when reaching the
adult condition. Thus, as GERHART (2006, p. 681) states, the
ancestral larva is a poor candidate for ur-deuterostomian organization, compared with the ancestral adult. In consequence,
the various “neotene larva”-scenarios for deuterostome evolution (e.g. SALVINI-PLAWEN 1998) are rejected (see also
LACALLI 2005), and anagenetic reconstructions have to focus
on stepwise biomechanical modifications of adult forms.
ancestral bilaterian
with internal canals
filter-coelomates
(Notoneuralia)
worm coelomates
(Gastroneuralia)
MICHAEL GUDO & TAREQ SYED, 13. November 2008
3. Anagenetic reconstructions by
Konstruktionsmorphologie
Anagenetic reconstructions should present evolutionary scenarios that show how the functional design of organismic body
structures can be transformed without gaps in functionality.
Plausible scenarios can be established by ruling out dysfunctional intermediates, and by explaining how the proposed
pathways of transformation are canalized and constrained. In
the quasi-engineering approach of “Konstruktionsmorphologie“, the structures of living organisms are treated as biotechnical units. This conceptualization of plausible reconstructions is
not comparable with anagenetic hypotheses established by the
mapping of characters on genealogies; it does not even require
a cladogenetic working hypothesis. Instead, a detailed analysis
of the functional design of recent organisms is the starting
point for a “retro-engineering” evolutionary interpretation.
Here, an initial analytical dissection phase identifies and describes the structural elements and analyses their mechanical
roles and mutual interdependence. A subsequent integrative
phase uses these data to create a mechanically coherent model
of the organismic apparatus. It takes into account the biomechanical and material properties of the various tissues and fluid
fillings which constitute every organism. Thus, organisms are
interpreted as “hydraulic machines” in this approach. Konstruktionsmorphologie originated in a series of papers which
aimed to reconstruct vertebrate anagenesis from invertebrate
precursors (GUTMANN 1967a et mult.). The evolutionary modifications of hydroskeletal body plans are a central aspect in
invertebrate-vertebrate transformation models, and these anagenetic models automatically affect the understanding of deuterostome radiation.
4. Evolutionary Scenario for Deuterostomes
This section presents the basic reconstructions for the Notoneuralia domain, as elaborated by the approach of Konstruktionsmorphologie over the past fourty years. The hypothesis of
a polysegmented hydroskeletal Urdeuterostomian, which gives
rise to a chordate branch on the one hand and an ambulacrarian
branch on the other, has been revised and detailed many times
(e.g. GUDO & GRASSHOFF 2002). Central arguments of the
latest version are summarized in the following, offering a conclusive morphological interpretation of deuterostome molecular phylogeny. Moreover, we think that this is one of the most
detailed scenarios of deuterostome evolution. As such it is of
special interest for palaeontologists, because fossil findings
can be interpreted with regard to the hypothetical intermediates
described in the step-by-step-reconstructions.
4.1 The Ur-Deuterostomian
The first aspect we have to deal with is the body structure of
the supposed deuterostome ancestor, i.e. the Ur-Deuterostome
and the Ur-Bilaterian. The polysegmentation of these forms
was proposed by GUTMANN (1966, 1967a, 1972a,b), now gaining support from developmental genetics (fig. 1). By the retroengineering approach, the body structure of the deuterostome
ancestor can be outlined in more detail, i.e., the interaction of
anatomical and histological structures is described: It was an
elongated worm-like animal with serially arranged coelomic
cavities that served as a hydraulic skeleton (comp. fig. 3). Lateral undulations, one of several types of body movements possible in such a construction, were performed by muscular contraction and antagonistic interaction transmitted by the fluid
filling of the coelom. The coelom is one of several differentiations of former fluid-filled canals lined with a ciliated epithelium. One of these canals functioned as intestinal tract, others
functioned in the distribution of nutrients and the excretion of
metabolic products. Additional canals might have been present,
but speculations about their number and function lies beyond
the possibilities of our reconstruction. One specific longitudinal canal, located dorsally of the intestinal tract, will be of special importance in the subsequent argumentation (shown in
fig 4).
Biomechanicly, the lateral, serial arrangement of hydraulic cavities minimizes internal friction during body movements. When the body bends laterally, the fluid in these proposed coelomic cavities is deformed by the muscles of one body
side and thereby stretches the muscles of the opposite side. The
tissues that separate the chambers of this segmented hydraulic
skeleton constrain the deformation of the fluid and permit
smooth movement. We will assume a state in which, similar to
living annelids, there were two types of internal walls that bordered coelom chambers. First, there were the transverse septa
oriented perpendicular to the long body axis, which are called
dissepiments. Second, septa along a plane parallel with the
long body axis divided the serial body cavities into two lateral
portions, thus establishing a truely bilateral biomechanic architecture without which it would be problematic to define lateral
undulation as a specific mode of motility (HERKNER 1991).
Such a longitudinally oriented septum is called mesenterium. It
should be noted that while this hypothetical annelid-like urbilaterian resembles modern annelids biomechanicly, there is no
requirement for body tissues to be as highly differentiated as in
recent annelids. Assuming that our urbilaterian originated from
gelatinous early metazoa (fig. 3), the body can be assumed to
have consisted of relatively undifferentiated gelatinous tissues
with contractile cells and a grid of connective fibres (the latter
being components of the extracellular matrix). The potential of
such “gelatinous fibrous tissues“ or, in German, “GallertfaserGewebe“, for evolutionary differentiation has been discussed
elsewhere (GRASSHOFF & GUDO 2001, 2002).
The fluid-filled body works as a hydrostatic skeleton
(CLARK 1964), which means that forces exerted by contracting
muscles are transmitted to antagonistic muscle groups by the
volume-invariant deformation of fluid-filled cavities. In such a
hydraulic softbody system performing lateral undulations, the
hydraulic system transmits the muscular forces between the
left and right body side, which contract alternatingly. Under
contraction of the longitudinal muscles alone, the body would
be shortened stepwise and increase in width, but circular muscles as well as contractile elements that might be present in the
dissepiments and mesenteria work against these forces
(HERKNER 1991) and so the body bends, and ultimately moves
foreward. For physiological reasons it is necessary that each of
the hydraulic units (=segments) has its own excretory system
such as the metanephria in existing annelids. These metanephridia have their internal opening in one coelomic cavity
and their terminal end in the adjacent (caudal) segment (fig. 4).
13
Hundred Years of Deuterostomia
This organism moved by lateral undulations; a pelagial
life made the active acquisition of food possible. Any morphological transformation that improved the efficiency of locomotion or reduced the amount of muscular activity required for
the dynamic control of body shape during locomotion was
advantegous at this comparatively unspecialized stage. In an
engineering sense, the most efficient modification of this body
structure would be a reduction of the mass of the circular
muscles, which only function to stabilize body length during
undulating movements and enforce a more or less constant
body diameter. Controling body shape by muscular action
alone is energy-consuming and metabolicly costly, and it limits
the overall body size (GUTMANN 1972a; GUDO & GRASSHOFF
2002).
coelome
proto-gills
internal canals
transverse septum
neural tube
metanephridium
4.2 The Chordates
chorda dorsalis
differentiated coelomic cavities
(dorsal parts become sclerocoels)
early acrania stage
myoseptum
myomere
early craniote stage
Fig. 4. Evolutionary scenario for the origin of chordates from polymersegmented ancestors possessing a hydraulic skeleton. The sequence
starts with the hypothetical proto-deuterostome, an actively swimming,
filtrating organism. The evolution of the neural tube and the chorda
dorsalis are optimizations of the lateral undulating mode of swimming.
Thus, deuterostome radiation reflects different options of the notoneuralian and chordate organization. Due to the dorsal concentration of
muscles, the coelom and excretory system (metanephridia) are rearranged in the lineages leading to the acrania and craniota (modified from
GUDO & GRASSHOFF 2002; see text for details).
This makes sense hydromechanicly as no hydraulic fluid is lost
in an uncontrolled way when muscular activity sets the fluid
filling under pressure (CHAPMAN 1950; CHAPMAN 1958; QUILLIN 1998).
The front end of this polysegmented animal carries a filtration apparatus; canals run from the foremost part of the gut
to the lateral body wall so that water can flow in through the
mouth and out through lateral openings. Nutrition particles are
collected and held by mucus, and transported to the pharynx
via a ciliated ridge (i.e. the putative endostyle, GUDO & GRASSHOFF 2002).
14
The evolutionary transformations leading to a reduction of the
circular muscles have to be interpreted in the context of
interactions between existing structures and of causal histogenetics. If the body length is stabilized by circular muscle activity, the tissues in the body can differentiate in specific ways.
During swimming movements – performed by the lateral bending of the body via longitudinal and circular muscles – tissues
are set under compressive and tensile stresses. Especially the
tissues in the sagittal plane of the body have to absorb lateral
and longitudinal forces (HERKNER 1991). The tissues react by
histological reorganization: the connective tissue fibres and
contractile cells orientate perpendicularly to the working
forces. In interaction with the already existing fluid-filled canal
dorsal of the intestinal tract, force-absorbing structures
developed in the sagittal plane of the body.
The dorsal canal above the intestinal tract became the
key structure for the evolving notoneuralian organization
(GUDO & GRASSHOFF 2002). It provided a place for nervous
fibres stimulating the longitudinal muscles on the left and right
body flank. As a consequence, control and stimulation of locomotive muscles is not disturbed by the generated mechanical
forces. In a fluid-filled canal, bending forces are low and compressive forces are compensated by the internal liquor and the
enveloping connective tissue. This effect is further enhanced
when the canal is closed at both ends. It represents a hydraulic
system in the midline of the body, stabilizing body length at
least to a certain extent. As a consequence, radial muscles lose
some of their importance as control elements of body width,
and can be gradually reduced. Moreover, the tissues between
the neural canal and the intestinal tract could develop additional histological differentiations that enhanced their functionality as a length-stabilising yet flexible structure. As mentioned
before, these tissues formed a force-absorbing structure, possessing fibres orientated perpendicular to the working forces.
Finally, the notochord resulted: a connective tissue enclosing a
contractile system that provided length stability (HERKNER
1991). The notochord can perform the function of length stabilization much better than the ancestral interaction of lateral
muscles and the fluid-filled neural canal. Obviously, the
functional design which evolved by these transformations is
the body structure of an early chordate (fig. 4, for more details
see GUTMANN 1972a; GUTMANN 1997; GUDO & GRASSHOFF
2002).
MICHAEL GUDO & TAREQ SYED, 13. November 2008
The histological configuration of such an early chordate
is represented by Branchiostoma lanceolatum and other extant
acranians. Histological investigations show the notochord as a
contractile hydraulic organ surrounded by the hydraulic sclerocoels, which allow free gliding of the notochord in the muscle
packages during lateral undulation (GUTMANN 1971; BONIK &
GUTMANN 1978). With respect to the body cross-section, the
notochord is positioned somewhat dorsal of the midpoint. This
configuration leads to particular biomechanical problems for
lateral undulations. Engineering considerations show: if (1)
muscle packages would be oriented in a way that the generated
forces work exactly parallel to the body axis, and (2) the length
stabilizing, force absorbing structure would not be in the
middle of the body diameter, then the horizontal undulation of
the body would include a slight dorso-ventral bending, causing
inefficient locomotion. But in existing animals, muscle fibres
are arranged in a specific manner, viz. in a long elongated spiral that extends from the front to the hind end and the
notochord in its center. This orientation of muscle fibres is
attained by a reorganisation of the dissepiments (which later
will evolve into myosepta) which connect the several portions
of longitudinal muscles coherently with each other into the
well known zig-zag pattern of recent chordates (BONIK & GUTMANN 1978). However, an alternative solution is also possible:
an enlargement of the dorsal muscle packages, which results in
a midline position of the chorda over the entire body length and
allows the perpendicular orientation of the myosepta. This evolutionary solution is shown by some fossil early chordates,
such as Haikouella lanceolatum or Pikaia, found in early cambrian sediments (BRIGGS & KEAR 1994; CHEN et al. 1995;
CHEN & LI 2000; MALLATT & CHEN 2003; SHU et al. 1999,
2003). It can be interpreted as an ancestral functional design in
the evolutionary field of chordates; however, these organisms
represent side branches.
The transverse septa can be reduced in the region of the
coelomic cavities when length stabilisation is achieved by the
notochord, so that the coelomic fluid does no longer function
as part of a hydraulic skeleton for lateral movements. Consequently, the tissue mass – which always comes at material and
energetic costs for production and maintenance – is reduced.
The result is one large body coelom, containing the intestinal
tract and the internal organs. This fluid-filled cavity still functions as a hydraulic skeleton for the dorsoventral bending
driven by the epaxial and hypaxial musculature. From this
chordate animal, further improvements of lateral undulation
open up the pathway to the craniotes (GUTMANN 1967b; GUTMANN 1972a; GUTMANN 1972b). Another anagenetic pathway
originating from the simple chordate construction defined
above leads to the recent acranians which preserved features of
this ancestral structure, due to some physiological constraints.
As discussed by SCHMITZ et al. (2000), respiration through the
branchial gut contributes only up to 10 percent to the overall
respiratory activity of small acranians, while the bulk of the
gas exchange takes place in the tissues of the peribranchial
chamber, over the entire body surface, and at the surfaces of
the coelomic cavities. This explains why acranians have to preserve their locomotory apparatus although they are hemisessile
organisms dwelling in sediments. The tunicates, on the other
hand, represent an alternative: they enlarged their peribranchial
cavity and perform respiration across its surface alone. As a
consequence, they were able to reduce their muscular tail, pre-
served in the famous tadpole larvae. Among adults, only in the
tunicate subgroup of the copelates which have no peribranchial
cavity, the tail persists; the copelates are an early tunicate side
branch (comp. GUDO 2004). The enigmatic Cambrian Vetulicolia fossils (SHU et al. 2001) probably represent intermediates
between swimming chordates with enlarged peribranchial cavity and ascidia-like tunicate forms (comp. the predicted prototunicate in fig. 5 and fig. 8).
As an interim summary, we conclude that our postulated
ancestral deuterostome gave rise to three major evolutionary
lineages: the acranians, the craniotes, and the tunicates. The
ancestral deuterostome was a swimming, filter-feeding chordate construction. This mode of life led to the craniotes on one
hand and, as an alternative specialization, to the peribranchialchordates which divided early into the acranians and the tunicates, on the other (fig. 5).
4.3 Ambulacrarians
The ambulacrarians (hemichordates plus echinoderms;
METSCHNIKOFF 1881) are commonly seen as primitive organisms and therefore assumed to be basal in evolutionary
history. However, not alone the NAP implicates a different
interpretation, but also biomechanical considerations make it
plausible that ambulacrarians are highly derived animals.
Starting again with the early chordate defined above, the
following scenario can be reconstructed. In the ancestral chordate, the coelomic cavities of the metameric ancestor had been
preserved as sclerocoels around the notochord, facilitating
practically friction-free gliding of the notochord past the lateral
muscle groups when the body undulated. This organism had a
peribranchial chamber, enabling a hemisessile life-style burrowed in the sediment, as in recent acranians. Modern lancelets
burrow by wriggling; they ‘hammer’ themselves into the sediment by notochord-contractions, supported by intense lateral
undulations of the trunk. However, this type of motility is energetically costly, and is only applied for short, fast movements
into sandy sediments. For a more persistent burrowing in softer
sediments, peristaltic body deformation would effectively support the wriggeling movements. However, such quasi-peristaltic movements can be perfomed only by the front end of the
ancestral chordate, when the longitudinal muscle fibres in the
oral region attain a plywood-like crisscross orientation, and
when the sclerocoels around the notochord enlarge. During
these transformations, the cranial end of the chordate was
transformed into an organ which allowed burrowing in fine
sediment by peristaltic movements. Finally, when the peristaltic organ differentiated into a proboscis and a collar, the wriggeling movements of the hindpart of the body became redundant. As a consequence, the notochord and the massive longitudinal muscles could be reduced in caudo-cranial direction. The
original peribranchial chamber behind the peristaltic organ
(that is, caudal of the collar), was dorsally opened, enforcing
nerve fibres to rearrange in the way known from recent enteropneusts. The gonads maintained their place in the “wings“ of
the peribranchial chamber. The result of this transformation
was an ancestral enteropneust-like organism (GUTMANN &
BONIK 1979).
This organism possessed an organisation of its coelomic
cavities which became important not only for the functioning
of the enteropneust construction, but also for subsequent evo-
15
Hundred Years of Deuterostomia
ancestral chordate
metameric ancestor
Tunicata (vetulicolia stage)
Craniota
Acrania
Fig. 5. Anagenetic relations of the chordate branch of the deuterostomes. According to biomechanical considerations, chordates appear basal,
while tunicates (= urochordates) are derived from acrania-like intermediates. Note that the branching order of the Tunicata, Acrania and Craniota
still remains unresolved in molecular phylogenies (comp. fig. 2b).
lutionary transformations. A pair of trunk coeloms extended
along the mesenterium into the collar region. The burrowing
enteropneust moves by peristaltic creeping, anchors its collar
in the sediment and contracts the longitudinal muscles of the
trunk pulling it behind. In this way, the pressure in the trunk
coelom including its protrusions into the collar rises. The resulting stiffening of the collar supports its anchoring in the walls
of the burrowed tube (fig. 6).
The reconstruction of the next evolutionary steps starting from the enteropneust organization is straight-forward
(comp. fig. 7). Extensions of the collar became tentacle-like
structures as found in recent pterobranchs (GUTMANN 1972a,
BONIK et al. 1978, GUTMANN & BONIK 1979). The transition
to echinoderms occurred in a separate lineage in which not the
gut and its terminal openings were shifted in the mesentery – as
it happened in the pterobranchs – but in which the gut together
with the mesentery became bent into a U-shaped structure. In
this way the trunk coelom was enlarged, forming a hydraulic
capsule which could be pressurized continuously by muscular
activity. Additional loops of gut and mesentery developed, providing internal tethering structures. These tethering structures
controlled the patterned bulging of the body wall when the
fluid filling was set under pressure. Due to the bending of the
gut and to basic physical principles acting when hydraulic
systems come into close contact, the trunk formed five bulges
forcing the body into a overall pentaradial organisation. These
bulges enlarged more and more and provided mechanical support for the tentacles that developed from the collar. Various
complex transformations of the geometry of the coelom and
the nervous system as well as histological transformations
occurred which we will not discuss in detail; for more com-
16
plete descriptions of the relatively straight-forward derivation
of echinoderm body-plans from enteropneust-like constructions, see GUDO (2005) and GUDO & DETTMANN (2005).
With this biomechanics-based evolutionary scenario, the
evolutionary history of the deuterostomes is complete. The
ancestral deuterostome-construction offered the biomechanical
option to develop a peristaltic organ (consisting of proboscis
and collar) by simple morphological transformations. When
these structures were fully developed, the typical chordate features could be reduced and the body structure of an enteropneust evolved. Thus, pterobranchs and echinoderms are highly
derived hemichordates. However, it remains unclear at this
stage whether pterobranchs and echinoderms are widely separated (i.e. independently from enteropneust-like predecessors)
or whether they evolved from an ancestor with a body structure
quite similar to pterobranchs.
5. Discussion
If the molecular reconstruction of deuterostome phylogeny as
shown in fig. 2b is accepted, then the anagenetic scenario of
the hydroskeleton hypothesis, published in a comprehensive
form as early as 1966 by W. F. GUTMANN, will be of high
interest as it sheds light on the causes for the contradictions
between phylogenies based on structural characters as opposed
to molecular ones. Provided that the cladogenetic and anagenetic implications of the NAP are correct, it appears that approaches based on morphological structures suffer from an inability
to identify secondary simplification events (JENNER, 2004). In
our constructional morphology approach, this is not the case.
MICHAEL GUDO & TAREQ SYED, 13. November 2008
zing that on the basis of recent molecular findings, the lophophorate phyla do not seem closely related to the Deuterostomia. The taxon “Epineuralia” consisting of Lophophorata and
Deuterostomia, which formed the basis of the – now fully refuted – cladogenetic and anagenetic proposals of V. SALVINIPLAWEN (1998) and NIELSEN (2001), appears today to be artificial. Lophophorates are highly derived protostomia, according to the hydroskeleton hypothesis; this probably is one of
the most specific similarities between the hypothesis and the
NAP, as noted by BALAVOINE & ADOUTTE (2003, p.142) with
regard to the brachiopod derivation model of GUTMANN et al.
(1978). Furthermore, the hydroskeleton hypothesis interpretes
the pelagic Copelata as an early branch within the Tunicata
(mentioned in section 4.2), which is also supported by molecular data (MALLATT & WINCHELL 2007). Thus, it now appears
that a very first “total evidence model” is possible, i.e. a complete integration of molecular results and independently elaborated morphological reconstructions (SYED 2003; SYED et al.
2007). Concerning early deuterostome radiation, it has been
stated that three competing models have to be discussed: an
ambulacrarian model, a chordate model, and a xenoturbellidmodel (SWALLA & SMITH 2008: 1565). It is quite surprising
that the chordate model is completely ignored in the newer literature, and that the scenario of the hydroskeleton-hypothesis is
still the only detailed proposal here. Nevertheless, several problematic aspects of the chordate model remain to be clarified,
as discussed below.
5.1 The unresolved position of Xenoturbella bocki
Fig. 6. Evolutionary transformation from chordates to ambulacrarians
(modified from GUTMANN 1989; see text for details).
Decades ago, “Konstruktionsmorphologie” has interpreted
tunicates, hemichordates, and echinoderms as highly derived
forms featuring secondary losses of typical chordate characters, which nicely fits the molecular results shown in fig. 1 and
fig. 2b. Moreover, the exclusion of the lophophorate phyla, the
Chaetognatha, and the Pogonophora from the Deuterostomia is
in full agreement with the hydroskeleton hypothesis (GUTMANN 1966 included the Pogonophora into the Deuterostomia,
but corrected this in 1972 by describing them as derived annelids. The latter is now confirmed by molecular data, see
HALANYCH 2004, p.239 f.). In particular, it is worth emphasi-
The possible deuterostome relationship of xenoturbellids is a
feature of current molecular phylogenies which has not been
subjected so far to a rigorous retro-engineering analysis in the
context of the hydroskeleton hypothesis. If Xenoturbellida are
an early side branch of Ambulacraria as suggested by fig. 2b,
then an enteropneust-like precursor would be the most obvious
starting point for an anagenetic interpretation. We do not
attempt to detail such a model here, because the notion of
Xenoturbella as a basic ambulacrarian line (BOURLAT et al.
2006) has recently been challenged by PERSEKE et al. (2007),
according to whom xenoturbellids are closest relatives of all
other deuterostomes, or in some analyses even a sister group of
the Protostomia + Deuterostomia-clade. However, the latest
results from molecular systematics, including an analysis of
their Hox-repertoire, support the inclusion of Xenoturbellida
into Deuterostomia, and their branching as basal Ambulacraria
(FRITZSCH et al. 2007, DUNN et al. 2008). These results imply
that Xenoturbella has to be interpreted as a secondarily simplified animal, which lost coelomic cavities and trimeric segmentation from a hemichordate-like stage. In support of this idea, it
has to be stressed that according to the NAP the vast majority
of flatworm-like body plans has to be interpreted as “coelomates without a coelom” (BALAVOINE, 1998), i.e. as highly
derived side branches, independent of their exact branching
position (comp. position of the platyhelminthes in fig. 1).
5.2 Echinoderms: Descendants of pterobranchs or
of enteropneusts?
The hydroskeleton hypothesis has been challenged on the
grounds that the NAP does not support a Pterobranchia + Echi-
17
Hundred Years of Deuterostomia
ancestral enteropneust
pterobranch-like
intermediate
Echinodermata
Enteropneusta
Pterobranchia
Fig. 7. Anagenetic relations of the ambulacrarians. According to the presented scenario (fig. 5), the functional design of enteropneusts can be
derived from an ancestral chordate. From such a body structure, the recent enteropneusts, pterobranchs, and echinoderms developed as independent lineages. This does not necessarily imply that Echinodermata are the sister-group of the Pterobranchia, but rather that echinoderms have to be
derived from ancestors that had a functional design similar to pterobranchs.
nodermata-clade (WINCHELL et al. 2002, p.774). This argument is not sound since the anagenetic model of echinoderm
evolution as given by GUTMANN (1972a) and, in more detail,
by GUDO (2005), does not necessarily imply a sister group
relationship between Pterobranchia and Echinodermata. As
GUTMANN (1981, p.74) explained:
“Because the adult morphology of echinoderms is so profoundly modified from that of typical acrania-like ancestors, it
is difficult to suggest a reasonable transformation series and to
link possible intermediate stages with known groups. Most
evidence suggests a very close link between echinoderms and
hemichordates so that I will assume that the pterobranchs or at
least the enteropneusts represent the chordate construction
from which echinoderms evolved.”
In other words, an independent evolution of pterobranchs and echinoderms from an enteropneust-like ancestor is
not excluded (in fact, this possibility was explicitly demonstrated by BONIK et al. 1978). In this scenario, which fits the
results of the NAP, the tentacles of pterobranchs and echinoderms evolved independently. Our figs. 7 and 8 show both
possibilities: A close relationship of echinoderms and pterobranchs in fig. 7fig. 1, and a probably independent evolution of
these lines in fig. 8. What both reconstructions have in com-
18
mon is that the tentacles of pterobranchs and echinoderms arise
from extensions of the enteropneust collar (see also fig. 6). In
general, the collar region of recent enteropneusts shows considerable variability as demonstrated by the bizarre, enlarged
collars of the former “Lophenteropneusts”, deep sea enteropneusts described by HOLLAND et al. (2005). Taken together, the
reconstruction of the ambulacraria-domain does not imply one
single cladogenetic tree, but rather an anagenetic scenario in
which echinoderms and pterobranchs are more derived than
enteropneusts. It cannot be overstressed that this long-time
minority opinion is now fully supported by molecular phylogenetics.
5.3 The basic deuterostome: pelagic or benthic
mode of life?
The swimming capacity of acranians now is widely accepted as
a plesiomorphic character of the Chordata (since Tunicata are
no longer regarded as ancestral forms), and this raises the question whether early deuterostomes were capable of active
swimming, too. The idea usually is rejected by authors who
favour an enteropneust-like predecessor, and thus a fully
benthic mode of life for the ur-deuterostomia (e.g. LACALLI
2005). As mentioned in section 2.2, developmental biologists
tend to support this notion because the localization of the ente-
MICHAEL GUDO & TAREQ SYED, 13. November 2008
Fishes
metameric hydraulic system
head, jaws,
scale-belts
Conodonts
rostral filter system
pharyngealelements
Agnatha
head
capsule
notochord/somite system
Acranians
mouth
tentacles
Tunicates
branchial
basket
proboscis:
burrowing
Enteropneusts
Fig. 8. Combined evolutionary scenario (= anagenetic relations) of the
deuterostomes. Biomechanical considerations strongly suggest that within
the deuterostomes, two major lineages diverged very early: the lineage of
the ambulacraria (secondarily benthic
forms) and the lineage of the chordata
(primarily pelagic; the tunicates
secondarily became benthic with the
possible exception of the pelagic tunicate subgroup of the Copelata [syn.
Larvacea/Appendicularia]).
For
details, see text.
tentacles
Pterobranchia
skeletal elements
in the skin
Echinoderms
ropneust mouth at the chordin-side is protostome-like and thus
indicates that enteropneust are links between proto- and deuterostomes (NÜBLER-JUNG & ARENDT 1999; GERHART 2006).
However, newer evo/devo-studies realize that the enteropneusts may represent a highly modificated deuterostome condition. DENES et al. (2007, p. 285) offer an interpretation which
perfectly fits the scenario of the hydroskeleton hypothesis
(compare fig. 6 & 7):
“One possible explanation is that the enteropneust trunk has
lost part of its neuroarchitecture due to an evolutionary
change in locomotion. While annelids and vertebrates propel
themselves through trunk musculature (and associated trunk
CNS), the enteropneust body is mainly drawn forward by
means of the contraction of the longitudinal muscles in their
anterior proboscis and collar (…). Possibly, enteropneusts
have partially reduced their locomotor trunk musculature concomitant with motor parts of the CNS (while the peripheral
sensory neurons prevailed in “diffuse” arrangement).”
This justifies the assumption of swimming organisms as
protostome-deuterostome intermediates, as more as this seems
plausible within the phylogenetic context of the NAP, especially with respect to the systematic position of the Chaetognatha
(VARGAS & ABOITIZ 2005). As mentioned above, Chaetognatha appear unrelated to the deuterostomes in molecular phylogenies, and group as an isolated protostome branch ancestral to
the Ecdysozoa + Lophotrochozoa-clade. Therefore, VARGAS &
ABOITIZ (2005) conclude that ancestral protostomes might well
have been pelagic organisms (as chaetognaths typically are),
and that this also holds true for the ancestral deuterostomes
(even more so as Chaetognatha are “protostomes with deuterostome-like development”, MARLÉTAZ et al. 2006). Moreover,
BOURLAT et al. (2006: 88) emphasize that the nearest protostome-like relatives of the deuterostomes possess a centralized
nervous system rather than diffuse nerve nets, so that the former is more likely to represent the condition of basal deuterostomes (comp. also DENES et al. 2007). Although both the
swimming capacity and a centralized nervous system are in
agreement with the anagenetic scenario shown in fig. 4, we feel
that this argumentation is problematic. For example, VARGAS
& ABOITIZ (2005) do not discuss evidence provided by COOK
et al. (2004) that Acoelomorpha represent the most basal bila-
19
Hundred Years of Deuterostomia
terian line. Defenders of the “benthic deuterostome ancestor”hypothesis could interpret the epibenthic lifestyle and the
basoepithelial nerve net of acoelomorphs as an ancestral condition, and this could also be claimed for the same features in
xenoturbellids (given that they group outside the Deuterostomia, as supposed by PERSEKE et al., 2007). This case exemplifies why instead of parsimony arguments, we prefer detailed
anagenetic models to reconstruct the morphological organization and habitat of the urdeuterostomia. According to our argumentation in section 4.2, flatworm-like creeping forms or neotenic larva of such benthic organisms can be ruled out as direct
chordate ancestors. For example, it is entirely unclear how a
well-regulated muscle/myosept-system could evolve gradually
in such forms. GUTMANN (1988, p.265) demonstrated this for
the tail of the tunicate larva: These larvae have been claimed to
represent chordate ancestors, although no explanation or model
has ever been offered for a gradual evolution of myosepta in
those organisms. This is also the case in the larva-hypothesis
given by SALVINI-PLAWEN (1998, p.138): In his fig. 21, the
longitudinal muscles suddenly develop into a serial pattern, but
no indication of the causes and constructional consequences of
this evolutionary step is provided. The hydroskeleton model
does not suffer from such shortcomings, as here the myosepta
can be derived from the transverse septa of an annelid-like
adult form without unexplained sudden ocurrences of novel
structures in the course of the anagenetic development.
5.4 Final Conclusions
Anagenetic hypotheses are always present in phylogenetic
trees but usually not revealed in full detail, especially in purely
cladistic approaches. Following the integration of molecular
data and results from the “Konstruktionsmorphologie”approach as given above for the deuterostome phyla, one can
outline a simple, two-step method to evaluate “naked” cladograms (i.e. pure genealogical schemes). The first step is to
derive a scenario from the tree considered; the second step is to
compare this scenario with anagenetic models that are based on
quasi-engineering investigations and reconstructions of the
functional designs of the evolutionary stages postulated. In
turn, molecular investigations are extremely helpful to determine starting points of anagenetic models, especially for
highly modified or miniaturized forms. Above, we have mentioned xenoturbellids, echinoderms and the tunicate subgroup
of the copelates as examples how molecular genealogies support the determination of phylogenetic precursor stages. However, secondary reductions and modifications also occur on the
genetic level and are difficult to handle by molecular systematics (which actually is a problem in the Tunicata, compare
MALLATT & WINCHELL 2007, p.1012). Waiting for robust
molecular phylogenies may require considerable patience in
some cases, but it is wiser than accepting molecular trees too
early, as totally artificial results are always possible. A warning
example is the previous molecular-based classification of
Xenoturbella as a mollusk, more specifically a nuculid clam
(literature in PERSEKE et al. 2007). This finding immediately
led some authors to suggest that body plan plasticity is nearly
incalculable (e.g. LEROI 2000). This view ignores biomechanical constraints and physiological limitations that canalize morphological transformations. Contrarily, in the context of the
“Konstruktionsmorphologie”-approach, the lack of any plausi-
20
ble derivation of the Xenoturbella body-plan from a clam-like
organization forms a strong argument against the validity of
the molecular analyses that motivated the inclusion of Xenoturbella in the molluscs. Evolution does not stand for a “morphing” of character patterns. Therefore we are skeptical when
for example GERHART (2006) presents a character-based body
inversion scenario leading from enteropneusts (turned upside
down) to chordates. GERHART himself admits on p. 682 that
“this is, of course, a speculative >morphing< exercise”. In contrast, anagenetic reconstructions that are based on analyses of
the biomechanics and physiological functioning of the organisms in question, are open to criticism of their physical plausibility, and – speculative as they may be – can never be rated as
mere “morphing exercises” (the latter being a purely statistical
approach). We agree with BOCK (2000) who rated process-oriented approaches as “nomological deductive explanations”;
such explanations can always be critically evaluated in terms
of causations and effects, because they deal with transformations of functional organisms, not with patterns of descriptive
characters.
In the case of the deuterostomes, detailed anagenetic
reconstructions suggest that all phylogenetic trees are highly
questionable which, for example, shift echinoderms to an ancestral position, or which combine echinoderms or hemichordates with tunicates (known as the pharyngopneust-hypothesis).
Similarly, assumptions of a xenoturbellid-like or an oligomeric
(trimeric) ur-deuterostomian have to be rejected: The mechanical conditions for the development of a notochord and an
efficient locomotory apparatus as known from recent acranians
and craniotes, are not given in any of these body structures. On
the other hand, trees are plausible that place echinoderms and
hemichordates together as Ambulacraria, as the sister group of
the Chordata. Highly convincing are those phylograms which
suggest a polysegmented ancestor for the deuterostomes (as for
example fig. 1), because only in a polysegmented system the
biomechanical conditions for the development of a notochord
are present.
What does this imply regarding the Deuterostomia-phylogenies (as shown in fig. 2b) of the NAP? Even though there
still is some ambiguity in the cladogenetic model, the overall
anagenetic implications appear quite plausible from a constructional morphologist´s point of view, especially when results
from molecular systematics and developmental genetics are
combined, as shown in fig. 1. In other words, the Deuterostomia-phylogenies elaborated by newer molecular approaches
can be rated as being in the realm of possibility. If so,
GROBBEN’S (1908) structure-based phylogeny, deriving deuterostomes from coelomates and deviding them into Ambulacraria and Chordata implies one of the most favourable anagenetic models – however, a lot of details of this scenario still
await plausible reconstructions.
6. Acknowledgements
We are indebted to our colleagues and students at the Senckenberg Institute for discussions of the topic. Special thanks to Dr.
MANFRED GRASSHOFF, who read the manuscript several times
and helped considerably to develop the argumentation presented, and to Dr. WINFRIED S. PETERS for improving the manuscript linguisticly.
MICHAEL GUDO & TAREQ SYED, 13. November 2008
7. Literature
ADOUTTE, A., BALAVOINE, G., LARTILLOT, N. & DE ROSA, R. (1999):
The end of intermediate taxa? Trends in Genetics 15: 104-108.
ADOUTTE, A., BALAVOINE, G., LARTILLOT, N., LESPINET, O.,
PRUD´HOMME, B. & DE ROSA, R. (2000): The new animal phylogeny: Reliability and implications. Proceedings of the National
Academy of Science, USA 97: 4453-4456.
BALAVOINE, G. (1998): Are Platyhelminthes coelomates without a
coelom? An argument based on the evolution of Hox genes. Amer.
Zool. 38: 843-858.
BALAVOINE, G., DE ROSA, R. & ADOUTTE, A. (2002): Hox clusters and
bilaterian phylogeny. Molecular Phylogenetics and Evolution 24:
366-373.
BALAVOINE, G. & ADOUTTE, A. (2003): The segmented Urbilateria: A
testable scenario. Integrative and Comparative Biology 43: 137147.
BOCK, W. J. (2000): Explanations in a historical science. In: D. S.
PETERS & M. WEINGARTEN (eds.): Organisms, Genes and Evolution - Evolutionary Theory at the Crosscroads. Proceedings of the
7th International Senckenberg Conference. Steiner, Stuttgart: 3342.
BONIK, K. & GUTMANN, W. F. (1978): Die Biotechnik der DoppelHydraulik (Chorda-Sklerocoelen-Myomeren-System) bei den
Acraniern. Senckenbergiana biologica 58: 275-286.
BONIK, K., GUTMANN, W. F. & HAUDE, R. (1978): Stachelhäuter mit
Kiemen-Apparat: Der Beleg für die Ableitung der Echinodermen
von Chordatieren. Natur und Museum 108: 211-214.
BOURLAT, S. J., JULIUSDOTTIR, T., LOWE, C. J., FREEMAN, R., ARONOWICZ, J., KIRSCHNER, M., LANDER, E. S., THORNDYKE, M.,
NAKANO, H., KOHN, A. B., HEYLAND, A., MOROZ, L. L., COPLEY,
R. R. & TELFORD, M. J. (2006): Deuterostome phylogeny reveals
monophyletic chordates and the new phylum Xenoturbellida.
Nature 444: 85-88.
BRIGGS, D. E. G. & KEAR, A. J. (1994): Decay of Branchiostoma:
implications for soft-tissue preservation in conodonts and other
primitive chordates. Lethaia 26: 275-287.
CAMERON, C. B., GAREY, J. R. & SWALLA, B. J. (2000): Evolution of
the chordate body plan: new insights from phylogenetic analyses of
deuterostome phyla. Proceedings of the National Academy of Science, USA 97: 4469-4474.
CHAPMAN, G. (1950): Of the movements of worms. Journal of Experimental Biology 27: 29-39.
CHAPMAN, G. (1958): The hydrostatic skeleton in the invertebrates.
Biological Reviews of the Cambridge Philosophical Society 33:
338-360.
CHEN, J. Y., DZIK, J., EDGECOMBE, G. D., RAMSKÖLD, L. & ZHOU, G.
Q. (1995): A possible Early Cambrian chordate. Nature 377: 720721.
CHEN, J. Y. & LI, C. W. (2000): Distant ancestor of mankind
unearthed: 520 million year-old fish-like fossils reveal early history
of vertebrates. Scientific Progress 83: 123-133.
CLARK, R. B. (1964): Dynamics in the metazoan evolution. The origin
of the coelom and segments. Clarendon, Oxford.
COOK C. E., JIMENEZ E., AKAM M. & E. SALO (2004): The Hox gene
complement of acoel flatworms, a basal bilaterian clade. Evolution
and Development 6: 154-163.
DELSUC, F., BRINKMANN, H., CHOURROUT, D. & PHILIPPE, H. (2006):
Tunicates and not cephalochordates are the closest living relatives
of vertebrates. Nature 439 (7079): 965-968.
DENES, A.S., JÉKELY, G., STEINMETZ, P.R.H., RAIBLE, F., SNYMAN,
H., PRUD´HOMME, B., FERRIER, D.E.K., BALAVOINE, G. &
ARENDT, D. (2007): Molecular architecture of annelid nerve cord
supports common origin of nervous system centralization in Bilateria. Cell 129: 277-288
DEROSA, R., PRUD´HOMME, B. & BALAVOINE, G. (2005): caudal and
even-skipped in the annelid Platynereis dumerilii and the ancestry
of posterior growth. Evolution and Development 7: 574-587.
DUNN, C. W., HEJNOL, A., MATUS, D. Q., PANG, K., BROWNE, W. E.,
SMITH, S. A., SEAVER, E., ROUSE, G. W., OBST, M., EDGECOMBE,
G. D., SORENSEN, M. V., HADDOCK, S. H. D., SCHMIDT-RHAESA,
A., OKUSU, A., KRISTENSEN, R. M., WHEELER, W. C., MARTINDALE, M. Q. & GIRIBET, G. (2008): Broad phylogenomic sampling
improves resolution of the animal tree of life. Nature 452: 745-749.
FRITZSCH, G., BÖHME, M.U., THORNDYKE, M., NAKANO, H., ISRAELSSON, O., STACH, T., SCHLEGEL, M., HANKELN, T. & STADLER, P.
F. (2007): PCR survey of Xenoturbella bocki Hox genes. J. Exp.
Zool. (Mol. Dev. Evol.) 310B: 278-284.
GERHART, J. (2006): The deuterostome ancestor. Journal of Cellular
Physiology 209: 677-685.
GEE, H. (2001): Deuterostome phylogeny: the context for the origin
and evolution of the vertebrates. In: AHLBERG P.E. (ed.): Major
events in early vertebrate evolution. Taylor & Francis, London,
New York: 1-14.
GRASSHOFF, M. & GUDO, M. (2001): The evolution of animals –
poster with explanations. Frankfurt am Main.
GRASSHOFF, M. & GUDO, M. (2002): The Origin of Metazoa and the
Main Evolutionary Lineages of the Animal Kingdom – The Gallertoid Hypothesis in the Light of Modern Research . Senckenbergiana lethaea 82: 295-314.
GROBBEN, K. (1908): Die systematische Einteilung des Tierreiches.
Verhandlungen der kaiserlich-königlichen zoologisch-botanischen
Gesellschaft in Wien 58: 491-511.
GUDO, M. (2004): Der Generationswechsel der Tunikaten: Ein evolutionsbiologisches Erklärungsmodell. Senckenbergiana biologica
84: 97-117.
GUDO, M. (2005): An Evolutionary Scenario for the Origin of Pentameric Echinoderms – Implications from the Hydraulic Principles of
form Determination. Acta Biotheoretica 53: 191-216.
GUDO, M. & DETTMANN, F. (2005): Evolutionsmodelle für die Entstehung der Echinodermen. Paläontologische Zeitschrift 79: 305-338.
GUDO, M. & GRASSHOFF, M. (2002): The Origin and Early Evolution
of Chordates: The ‘Hydroskelett-Theorie’ and New Insights
Towards a Metameric Ancestor. Senckenbergiana lethaea 82: 325346.
GUDO, M., GUTMANN, M. & SCHOLZ, J. (2002): Concepts of Functional, Engineering and Constructional Morphology: Introductory
Remarks. Senckenbergiana lethaea 82: 7-10.
GUTMANN, W. F. (1966): Coelomgliederung, Myomerie und die Frage
der Vertebraten-Antezedenten. Zeitschrift für zoologische Systematik und Evolutionsforschung 4: 13-57.
GUTMANN, W. F. (1967a): Die Entstehung des Coeloms und seine phylogenetische Abwandlung im Deuterostomier-Stamm. Zoologischer Anzeiger 179: 109-131.
GUTMANN, W. F. (1967b): Das Dermalskelett der fossilen "Panzerfische" funktionell und phylogenetisch interpretiert. Senckenbergiana lethaea 48: 277-283.
GUTMANN, W. F. (1971): Zu Bau und Leistung von Tierkonstruktionen
14. Was ist urtümlich an Branchiostoma? Natur und Museum 101:
340-356.
GUTMANN, W. F. (1972a): Die Hydroskelett-Theorie. Aufsätze und
Reden der Senckenbergischen Naturforschenden Gesellschaft 21:
1-91.
GUTMANN, W. F. (1972b): Vom Hydroskelett zum Skelettmuskelsystem. Eine biotechnisch begründete Evolutions-Studie. Mitteilungen
des Institutes für leichte Flächentragwerke der Universität Stuttgart
(IL) 4: 16-38.
GUTMANN, W. F. (1981): Relationships between invertebrate phyla
based on functional-mechanical analysis of the hydrostatic skeleton. American Zoologist 21: 63-81.
GUTMANN, W. F. (1988): The hydraulic principle. American Zoologist
28: 257-266.
GUTMANN, W. F. (1989): Die Evolution hydraulischer Konstruktionen
– organismische Wandlung statt altdarwinistischer Anpassung.
Kramer, Frankfurt am Main.
GUTMANN, W. F. (1993): Organismic machines – The hydraulic principle and the evolution of living constructions. In: K. KULL & T.
TIIVEL (eds.): Lectures in theoretical biology – The Second Stage,
Estonian Academy of Sciences, Tallinn: 171-188
GUTMANN, W. F. (1997): Chordaten-Konstruktionen und ihre Evolution. Von der Kiemenreuse zur zahnbesetzten Kieferapparatur: 363389. In: W. K. ALT & J. C. TÜRP (eds.): Die Evolution der Zähne –
Phylogenie, Ontogenie, Variation, Quintessenzverlag, Berlin.
GUTMANN, W. F. & BONIK, K. (1979): Detaillierung des Acranier- und
Enteropneusten-Modells. Senckenbergiana biologica 59: 325-363.
GUTMANN, W. F., VOGEL, K. P. & ZORN, H. (1978): Brachiopods:
Biomechanical Interdependences Governing Their Origin and Phylogeny. Science 199: 890-893.
21
HALANYCH, K. M. (2004): The new view of animal phylogeny. Annual
review of ecology, evolution and systematics 35: 229-256.
HATSCHEK, B. (1891): Lehrbuch der Zoologie. Third edition, Gustav
Fischer Verlag Jena.
HERKNER, B. (1991): Neue Betrachtungen zur Chordatenevolution.
Natur und Museum 121: 193-203.
HOLLAND, N. D., CLAGUE, D. A., GORDON, D. P., GEBRUK, A., PAWSON, D. L. & VECCHIONE, M. (2005): “Lophenteropneust” hypothesis refuted by collection and photos of new deep sea
hemichordates. Nature 434: 374-376.
HUGHES, A.L. & FRIEDMAN, R. (2005): Loss of ancestral genes in the
genomic evolution of Ciona intestinalis. Evolution and Devolopment 7: 196-200.
IKUTA T., YOSHIDA N., SATOH N. & SAIGA, H. (2004): Ciona intestinalis Hox gene cluster: Its dispersed structure and residual colinear
expression in development. Proceedings of the National Academy
of Science 101: 15118-15123.
JACOBS D. K., HUGHES N. C., FITZ-GIBBON S. T. & WINCHELL, C. J.
(2005): Terminal addition, the cambrian radiation and the phanerozoic evolution of bilaterian form. Evolution and Development 7:
498-514.
JENNER, R. M. (1999): Metazoan phylogeny as a tool in evolutionary
biology: Current problems and discrepancies in application. Belgian Journal of Zoology 129: 245-262.
JENNER, R. M. (2004): When molecules and morphology clash: reconciling conflicting phylogenies of the Metazoa by considering secondary character loss. Evolution and Development 6: 372-378.
JIMÉNEZ-GURI, E., PAPS, J., GARCÍA-FERNÀNDEZ & SALÓ, E. (2006):
Hox and ParaHox genes in Nemertodermatida, a basal bilaterian
clade. International Journal of Developmental Biology 50: 675679.
LACALLI, T.C. (2005): Protochordate body plan and the evolutionary
role of larvae: old controversies resolved? Canadian Journal of
Zoology 83: 216-224.
LEROI, A.M. (2000): The scale independence of evolution. Evolution
and development 2: 67-77.
LOWE, C. J. (2008): Molecular genetic insights into deuterostome evolution from the direct developing hemichordate Saccoglossus kowalevskii. Phil. Trans. R. Soc. B 363: 1569-1578.
MALLATT, J. & CHEN, J. Y. (2003): Fossil sister group of craniates:
predicted and found. Journal of Morphology 258: 1-31.
MALLATT, J. & WINCHELL, C. J. (2007): Ribosomal RNA genes and
deuterostome phylogeny revisited: More cyclostomes, elasmobranches, reptiles, and a brittle star. Molecular Phylogenetics and
Evolution 43: 1005-1022.
MARLÉTAZ, F., MARTIN, E., PEREZ, Y., PAPILLON, D., CAUBIT, X.,
LOWE, C.J., FREEMAN, B., FASANO, L., DOSSAT, C., WINCKER, P.,
WEISSENBACH, J. & LE PARCO, Y. (2006): Chaetognath phylogenomics: a protostome with deuterostome-like development. Current
Biology 16: R577-R578.
METSCHNIKOFF, V. E. (1881): Über die systematische Stellung von
Balanoglossus. Zoologischer Anzeiger 4: 139-157.
NIELSEN, C. (2001): Animal Evolution. Interrelationships of the living
phyla. Oxford University press.
NÜBLER-JUNG, K. & ARENDT, D. (1999): Dorsoventral axis inversion:
Enteropneust anatomy links invertebrates to chordates turned
upside down. Journal of Zoological Systematics and Evolutionary
Research 37: 93-100.
PERSEKE, M., HANKELN, T., WEICH, B., FRITZSCH, G., STADLER, P. F.,
ISRAELSSON, O., BERNHARD, D. & SCHLEGEL, M. (2007): The
mitochondrial DNA of Xenoturbella bocki: genomic architecture
and phylogenetic analysis. Theory in Biosciences 126: 35-42.
PHILIPPE, H., LARTILLOT, N. & BRINKMANN, H. (2005): Multigene
analyses of bilaterian animals corroborate the monophyly of Ecdysozoa, Lophotrochozoa and Protostomia. Molecular Biology and
Evolution 22: 1246-1253.
PRUD´HOMME B., DE ROSA R., ARENDT D., JULIEN J. F., PAJAZITI R.,
DORRESTEIJN A.W.C., ADOUTTE A., WITTBRODT J. & BALAVOINE,
G. (2003): Arthropod-like expression patterns of engrailed and
wingless in the annelid Platynereis dumerilii suggest a role in segment formation. Current Biology 13: 1876-1881.
QUILLIN, K. J. (1998): Ontogenetic scaling of hydrostatic skeletons:
geometric, static stress and dynamic stress scaling of the earthworm Lumbricus terrestris. The Journal of Experimental Biology
201: 1871-1883.
RENSCH, B. (1972): Neuere Probleme der Abstammungslehre. Verlag
F. Enke, Stuttgart.
RUPPERT, E.E. (2005): Key characters uniting hemichordates and chordates: homologies or homoplasies? Canadian Journal Zoology 83:
8-23.
SALVINI-PLAWEN, L. V. (1998): The urochordate larva and archichordate organization: chordate origin and anagenesis revisited. Journal
of Zoological Systematics and Evolutionary Research 36: 129-145.
SCHMIDT-KITTLER, N. & VOGEL, K. P. (1991): Constructional Morphology and Evolution. Springer. Berlin, Heidelberg, New York,
Tokyo.
SCHMITZ, A., GEMMEL, M. & PERRY, S. F. (2000): Morphometric partitioning of respiratory surfaces in Amphioxus (Branchiostoma lanceolatum Pallas). The Journal of Experimental Biology 203: 33813390.
SEO, H.-C., EDVARDSEN, R. B., MAELAND, A.D., BJORDAL, M.,
JENSEN, M. F., HANSEN, A., FLAAT, M., WEISSENBACH, J.,
LEHRACH, H., WINCKER, P., REINHARDT, R. & CHOURROUT, D.
(2004): Hox cluster disintegration with persistent anteroposterior
order of expression in Oikopleura dioica. Nature 431: 67-71.
SHU, D. G., LUO, H. L., MORRIS, S. C., ZHANG, X. L., HU, S. X.,
CHEN, L., HAN, J., ZHU, M., LI, Y. & CHEN, L. Z. (1999): Lower
Cambrian vertebrates from south China. Nature 402: 42-46.
SHU, D. G., CONWAY MORRIS, S., HAN, J., CHEN, L., ZHANG, X. L.,
LIU, H. Q., LI, Y. & LIU, Y. N. (2001): Primitive deuterostomes
from the Chengjiang Lagerstätte (lower cambrian, China). Nature
414: 419-424.
SHU, D., MORRIS, S. C., ZHANG, Z. F., LIU, J. N., HAN, J., CHEN, L.,
ZHANG, X. L., YASUI, K. & LI, Y. (2003): A new species of yunnanozoan with implications for deuterostome evolution. Science 299:
1380-1384.
STACH, T., DUPONT, S., ISRAELSSON, O., FAUVILLE, G., NAKANO, H.,
KÄNNEBY, T. & THORNDYKE, M. (2005): Nerve cells of Xenoturbella bocki (phylum uncertain) and Harrimania kupfferi (Enteropneusta) are positively immunoreactive to antibodies raised against
echinoderm neuropeptides. Journal of the Marine Biology Association of the United Kingdom 85: 1519-1524.
SWALLA, B. J. & SMITH, A. B. (2008): Decipherinmg deuterostome
phylogeny: molecular, morphological and palaeontological perspectives. Phil. Trans. R. Soc. B 363: 1557-1568.
SYED, T. (2003): Wie neu ist die "New Animal Phylogeny"? Eine
mögliche Synthese morphologischer und molekularer Befunde zur
Bauplan-Evolution. Jahrbuch für Geschichte und Theorie der Biologie 9: 33-76.
SYED, T., GUDO, M. & GUTMANN, M. (2007): Die neue Großphylogenie des Tierreiches: Dilemma oder Fortschritt? Denisia 20: 295312.
VARGAS, A.O. & ABOITIZ, F. (2005): How ancient is the adult swimming capacity in the lineage leading to euchordates? Evolution and
Development 7: 171-174.
WADA, H., OKUYAMA, M. SATOH, N. & ZHANG, S. (2006): Molecular
evolution of fibrillar collagen in chordates, with implications for
the evolution of vertebrate skeletons and chordate phylogeny. Evolution and Development 8: 370-377.
WINCHELL, C. J., SULLIVAN, J., CAMERON, C. B., SWALLA, B. J. &
MALLATT, J. (2002): Evaluating hypotheses of deuterostome phylogeny and chordate evolution with new LSU and SSU ribosomal
DNA data. Molecular Biology and Evolution 19: 762-776.