• Vertebrate Origin

  • FileName: Benton_sample chapter_Vertebrate Paleontology 3e.pdf [read-online]
    • Abstract: These fossils tell us where the living vertebrates have. come from, and they show us glimpses of different ... the vertebrate are labelled, and possible homologies with amphioxus are indicated. ...

Download the ebook

Vertebrate Origin
2 Vertebrate Origin
1 What are the closest living relatives of vertebrates?
2 When did deuterostomes and chordates originate? Animals
3 What are the key characters of chordates? Fungi
4 How do extraordinary new fossil discoveries from Protists
China help us understand the ancestry of vertebrates? chloroplasts
Vertebrates are all the animals with backbones, the fish-
es, amphibians, reptiles, birds, and mammals. These
animals have attracted a great deal of study. The efforts
of generations of vertebrate palaeontologists have
been repaid by the discovery of countless spectacular
fossils: the heavily armoured fishes of the Siluro-
Devonian, seven- and eight-toed amphibians, sail-
Fig. 1.1 The ‘Universal Tree of Life’, the commonly accepted
backed mammal-like reptiles, early birds and dinosaurs
view of the relationships of all organisms. Note the location of
with feathers, giant rhinoceroses, rodents with horns, ‘Animals’, a minor twig in the tree, close to plants and Fungi.
horse-eating flightless birds, and sabre-toothed cats. (Based on various sources.)
These fossils tell us where the living vertebrates have
come from, and they show us glimpses of different
worlds that seem so bizarre that they would defy the were wrong, and that the fundamental splits in the tree
imagination of a science fiction writer.Despite all of this of life were all among Bacteria,separating the two major
information that has accumulated over the past 200 groups Bacteria and Archaea. The familiar plants, ani-
years, the origin of the group is hotly debated. mals and fungi are part of Eukaryotes, the major group
One thing is clear from examination of living ani- characterized by complex cells with a nucleus, relative
mals. The vertebrates are members of a larger group, late-comers in the broad scheme of things.
termed the Phylum Chordata, and their closest living Modern studies (e.g. Nielsen et al., 1996) confirm
relatives are marine animals such as the sea squirts and that a major clade within Metazoa, the animals, is Bila-
amphioxus (see below). These creatures do not have teria, supported by both morphological and molecular
bone,one of the characters of most vertebrates,but they evidence (Eernisse and Peterson,in press).The Bilateria
share other features, such as a notochord, a flexible includes the bilaterally symmetrical organisms, com-
tough rod that runs along the length of the body down prising three clades: Lophotrochozoa (brachiopods,
the back. The notochord in living chordates is generally phoronids, annelids, molluscs and many minor
made from an outer sheath of collagen, a tough fibrous groups), Ecdysozoa (arthropods, nematodes, pria-
connective tissue that encloses turgid fluid-filled pulids and some minor groups) and Deuterostomia
spaces. Chordates also have V-shaped muscle blocks (echinoderms, hemichordates and chordates). The
(myomeres) along the length of their body. The ques- origin of vertebrates has long been a profound mystery,
tion about the origin of vertebrates then broadens out but now some clarity is emerging.
to focus on the origin of chordates. The purpose of this chapter is to explore the various
Looked at more widely, vertebrates are a minor twig lines of evidence that can be used to reconstruct the
on the ‘Universal Tree of Life’ (Figure 1.1). Molecular story of the origin of the vertebrates: the study of
studies through the 1990s (e.g. Woese, 2000; Wolf et al., modern animals that are vertebrate-like in some fea-
2002) showed that previous conceptions of the tree tures, the study of molecular relationships, and fossils.
Sea Squirts and the Lancelet 3
Fig. 1.2 The sea squirts: (a) Ciona, external view; (b) internal anatomy and cross-section of an adult; (c) swimming larva; (d)
metamorphosing form. (Modified from Jefferies, 1986 and other sources.)
1.1 SEA SQUIRTS AND THE LANCELET siphons, at the top. The body is firmly fixed to a hard
There are two basal groups of living chordates, the sea The internal structure is fairly complex (Figure
squirts and the cephalochordates (amphioxus). Am- 1.2(b)).A large pharynx fills most of the internal space,
phioxus certainly looks superficially fish-like, but the and its walls are perforated by hundreds of gill slits,each
adult sea squirts could hardly look like less likely rela- of which bears a fringe of cilia, fine hair-like vibratile
tives of the vertebrates! structures. Seawater is pumped through the inhalant
siphon into the pharynx by beating movements of the
cilia, and the water is then passed through a surround-
1.1.1 Urochordata: sea squirts ing cavity, the atrium, and ejected through the exhalant
siphon. The pharynx serves mainly to capture food par-
A typical sea squirt,or tunicate,is Ciona (Figure 1.2(a)), ticles from the stream of seawater that flows through it.
which lives attached to rocks in seas around the world. The seawater is drawn into a filter bag of mucus, which
It is a 100–150 mm tall bag-shaped organism with a is produced inside the pharynx by a gland known as the
translucent outer skin (the tunic) and two openings, or endostyle. During feeding, this gland continuously se-
4 Vertebrate Origin
Fig. 1.3 Amphioxus, a cephalochordate:
(a) modes of life, including swimming and
burrowing into sand for protection; (b)
internal anatomy. (Modified from Pough
et al., 2002 and other sources.)
cretes mucus into the oesophagus, together with the Branchiostoma, a representative of the Cephalochor-
food particles that it has filtered from the seawater, and data (or Acraniata). The adult amphioxus is convinc-
the food is passed to the stomach for digestion. ingly chordate-like, being a 50 mm long cigar-shaped
Why is Ciona identified as a chordate? The pharynx animal which looks like a young lamprey or eel,yet lack-
and other structures are in fact very like those of the ing a head. Amphioxus swims freely by undulating its
cephalochordates and lamprey larvae, but further evi- whole body from side to side, and it burrows in the sed-
dence is to be found in the larval stage, when the sea iment on the sea-floor (Figure 1.3(a)).
squirt is a tiny free-swimming tadpole-shaped animal Amphioxus feeds by filtering food particles out of
with a head and a tail. The larval sea squirt (Figure the seawater. Water is pumped into the mouth and
1.2(c)) has a notochord that runs along the tail, and this through the pharynx by cilia or the gill slits, and food
identifies it as a chordate. There are muscles on either particles are caught up in a bag of mucus produced by
side of the notochord that contract alternately, causing the endostyle, the feeding system seen also in tunicates
the tail to beat from side to side, and this drives the ani- and in the larvae of the lamprey. The mucus with its
mal forward in the water. The larva has a dorsal nerve contained food particles is pulled into the gut for diges-
cord, running along the tail just above the notochord, tion, whereas the seawater passes through the gill slits
and this expands at the front into a very simple brain into the atrium. Oxygen is also extracted, and the waste
which includes a light sensor (an ‘eye’) and a tilt water then exits through the atriopore.
detector. The anatomy of amphioxus, with its pharynx,
The larva then settles on a suitable surface. It up- notochord, dorsal nerve cord, myotomes, and
ends on to the tip of its ‘snout’ and attaches itself by endostyle (Figure 1.3(b)), is typically chordate.
means of adhesive suckers (Figure 1.2(d)). The noto- Swimming and burrowing are by means of lateral
chord and tail portion wither away, and the pharynx contractions of the myomeres acting against the stiff
and gut expand to fill up the body cavity.This extraordi- rod-like notochord.
nary metamorphosis occurs rapidly to allow the adult
to start feeding in its new way as soon as possible. 1.2 PHYLUM HEMICHORDATA:
1.1.2 Cephalochordata: amphioxus
Another unusual group of living marine deuterostomes
Another chordate generally reckoned to be related may offer further clues about the origin of the chor-
closely to the vertebrates is the amphioxus or lancelet, dates. These are the hemichordates, a phylum that in-
Phylum Hemichordata: Pterobranchs and Acorn Worms 5
cludes two superficially very different kinds of marine current, and food particles are captured by mucus on
animals. The first, the pterobranchs such as Cephalodis- the arms,while water passes out of the pharynx through
cus (Figure 1.4(a, b)), are small animals that live in loose the gill slits. The animal lives in or around a group of
colonies on the sea-bed in the southern hemisphere and horny tubes that the colony has constructed, and it at-
in equatorial waters. Cephalodiscus has a plate-like head taches itself inside these tubes by means of a sucker on
shield, a collar with five to nine pairs of feeding arms, the end of the stalk.
and a sac-like trunk perforated by a pair of gill slits and The second hemichordate group, the acorn worms,
containing the gut and gonads, and the body ends in a or enteropneusts, such as Saccoglossus, are worm-like
contractile stalk. Cilia on the arms produce a feeding animals varying in length from 20 mm to 1.8 m. They
Fig. 1.4 Typical hemichordates: (a) the pterobranch Cephalodiscus, internal anatomy and (b) mode of life; (c) the enteropneust
Saccoglossus, mode of life and external anatomy. (Modified from Jefferies, 1986.)
6 Vertebrate Origin
live in burrows low on the shore in Europe and else- Eventually a hollow ball of cells is produced, called the
where. Saccoglossus (Figure 1.4(c)) has a long muscular blastula stage (Figure 1.5(d)). A pocket of cells then
proboscis that fits into a fleshy ring or collar behind.The moves inwards, forming the precursor of the gut and
mouth is placed beneath this collar, and seawater and other internal structures. The opening of this deep
sand are pumped through the gut and expelled through pocket is called the blastopore. You can imagine push-
an anus at the posterior end of the body. The long body ing in the walls of a hollow rubber squash ball with your
is pierced by small holes at the front end, probably thumb to produce a model of this embryonic pattern,
equivalent to the gill slits of Cephalodiscus, sea squirts, known as the gastrula stage (Figure 1.5(e–g)).
and amphioxus. Embryologists noticed some time ago that animals
It was suggested that the Pterobranchia and En- fall into two large groups depending on the relative ori-
teropneusta should be regarded as two separate, entation of the mouth and anus. The classic story is that
but closely-related, groups (Peterson, 1995), although in most invertebrates (the protostomes), the blasto-
more recent molecular work (Winchell et al., 2002) pore becomes the mouth (Figure 1.5(h)), whereas in
concurs with morphological data (Smith et al., in press) others (the deuterostomes), including the chordates,
that Hemichordata is indeed a valid phylum, and this opening becomes the anus (Figure 1.5(i)), and the
more closely related to echinoderms than to chordates.
Hemichordates do not have a notochord at any stage,
but they possess gill slits, as in chordates, and giant
nerve cells in the nerve cord of the collar region that are
probably equivalent to similar nerve cells in amphioxus
and primitive vertebrates. Both pterobranchs and en-
teropneusts share morphological characters indicating
monophyly of the Hemichordata, such as the stomo-
chord (an anterior buccal tube on the dorsal part of the
pharynx) and mesocoelomic ducts.
The relationships of chordates used to be rather prob-
lematic, but intensive analyses of morphological and
molecular data have shown a clearer picture (Eernisse
and Peterson, in press; Smith et al., in press). The Phy-
lum Chordata is part of a larger clade, the Deuterosto-
mia, which in turn is part of a yet larger clade of all the
bilaterally symmetrical animals, the Bilateria (see p. 2).
But what exactly diagnoses the Deuterostomia? The
clue comes from embryology, the study of the early
phases of development in, and just out of, the egg.
Fig. 1.5 Embryonic development: (a–g) sequence of cell
1.3.1 Embryology and the position of the anus
division in amphioxus, from the single-cell stage (a), through
the blastula stage (d), to the gastrula stage (g). (h) Fate of the
In early development each animal starts as a single cell. blastophore in protostomes, and (i) in deuterostomes. [Figures
Soon this cell begins to divide, first into two cells, then (a–g), after Hildebrand and Goslow, 2001, copyright © 2001 John
four, then eight, sixteen, and so on (Figure 1.5(a–c)). Wiley & Sons, New York; (h, i), after Jefferies, 1986.]
Chordate Origins 7
mouth is a secondary perforation. Such a dramatic 1.4 CHORDATE ORIGINS
turnaround, a switch from mouth to anus, seems in-
credible. Note, however, that many protostomes show There are many putative early fossil chordates, and their
deuterostomy, and this condition may be primitive and numbers have grown hugely since 1995, with the
shared by all Bilateria (Eernisse and Peterson, in press). announcement of remarkable new finds from the
Nevertheless, this peculiarity of embryological devel- Chengjiang Formation of China, an Early Cambrian
opment appears to solve the question of the broader re- deposit (see Box 1.2). These new specimens, combined
lationships of chordates. with studies of modern forms, give clues about the early
evolution of chordates, but there are many disputes.
1.3.2 Relationships of the Deuterostomia
1.4.1 Diverse early chordates
The deuterostomes are the phyla Chordata, Hemichor-
data and Echinodermata. Another minor phylum, the There are four main categories of possible early
Chaetognatha, or arrow worms, was formerly included chordates: possible urochordates, possible cephalo-
here, but they show more protostome than deuteros- chordates, vetulicolians, and carpoids. At one time,
tome characters. The closest major group of living rela- conodonts, represented in the fossil record generally
tives of the chordates and hemichordates are thus the only by their tooth elements, were treated as dubious
echinoderms — sea urchins, star fish, sea lilies, and sea chordates. Conodonts are now placed firmly within the
cucumbers. Vertebrata, as jawless fishes, as are some of the taxa from
Can the status of the Deuterostomia be confirmed? Chengjiang, such as Haikouichthys and Myllokunmin-
The assumption is that Deuterostomia is a mono- gia (see Chapter 3).
phyletic group, or a clade, in other words, a group that Urochordates have a patchy fossil record. Isolated
had a single common ancestor,and which includes all of impressions of sac-like bodies, and trace fossils, mark-
the descendants of that ancestor (see p. 31). The mono- ings made in or on the sediment by the activities of ani-
phyly of the Deuterostomia is confirmed by the fact that mals, have been ascribed to tunicates. The best fossils
they possess unique characters that are not seen in other are small sac-like specimens from Chengjiang, Shank-
animals (Smith et al., in press): a posterior blastopore ouclava, that shows a large perforated branchial basket,
that generally becomes the anus, gill slits (present only branchial slits, and an elongate endostyle (Chen et al.,
in precursors of the echinoderms) and other characters. 2003). There is also a possible degenerating tail, sug-
There has been some dispute over the relationships of gesting this might be a larva that had just settled (cf.
the taxa within Deuterostomia (see Box 1.1), although Figure 1.2(d)).
this is now largely resolved. The fossil record of cephalochordates is not much
The chordates all share several unique features better. The Chengjiang locality has also yielded a
such as a notochord, a dorsal hollow nerve cord with a a superficially amphioxus-like cephalochordate,
shared developmental pattern, an endostyle organ Cathaymyrus, as well as the yunnanozoons, which have
(equivalent to the thyroid gland of vertebrates), and a also been identified as cephalochordates, although
tail used for swimming. It is generally reckoned that others assign them to the Vetulicolia (see below). In
only chordates have true tails. A tail technically may be the absence of hard tissues such as bone, these non-
defined as a distinct region extending behind the viscer- vertebrate chordates are not often preserved.
al cavity, and in particular located entirely behind the
anus. Non-chordates, such as insects, worms, molluscs,
jellyfish, and sea urchins, do not have tails. What of the 1.4.2 Vetulicolians and yunnanozoons
fossil evidence?
The Vetulicolia is a newly-named group, one of the
most extraordinary findings from the Chengjiang
8 Vertebrate Origin
Three substantially different schemes for deuterostome relationships have been proposed. The ‘traditional’ view (e.g. Maisey,
1986; Peterson, 1995; Donoghue et al., 1998; illustration (a)) was to place the hemichordates as basal to chordates since they
both share ciliated gill slits and giant nerve cells, as well as other features, which are not seen in echinoderms. Enteropneusts
were sometimes said to be closer relatives of chordates since their gill slits are similar, they have a very short dorsal hollow
nerve cord, and a number of other features of the gut not seen in pterobranchs (Peterson, 1995). Most authors regard
amphioxus as the closest relative of the Vertebrata on the basis of 10–15 features that are not seen in tunicates.
The ‘calcichordate’ model (Jefferies, 1986, 1997; illustration (b)) places hemichordates basal to echinoderms and uro-
chordates as sister group to vertebrates, based on evidence from embryology and fossils.
The third view (illustration (c)) is supported by morphological and molecular data and is now widely accepted (Smith et al.,
in press). The first molecular studies in which the 18 S rRNA genes of echinoderms, hemichordates, and chordates were com-
pared were inconclusive, but newer work (e.g. Bromham and Degnan, 1999; Cameron et al., 2000; Peterson and Eernisse,
2001; Furlong and Holland, 2002; Winchell et al., 2002) definitively pairs hemichordates with echinoderms, as the clade Am-
bulacraria, and places cephalochordates closer to chordates than urochordates. See Box 3.1 for phylogeny of Vertebrata.
rte at a
Ve ord dat
En ra ta
U aloc ta
U ho ta
C ord ta
C ode ta
br rd
C ord a
br rd
ro hia
Ec opn ia
ob ma
ic ma
ch us
ch hor
n es
ob es
n es
ha ta
br a
ha ta
te nch
ch da
rte ho
rte ho
hi eu
ro ne
Ec tom
Ec stom
em r
Ve loc
Ve loc
Pt ode
H ode
En ra
(a) (b) (c)
Cladograms showing the relationships of the main deuterostome groups: (a) the ‘traditional’ model, (b) the ‘calcichordate’ model, and (c) the
molecular model. Synapomorphies: A DEUTEROSTOMIA, blastopore becomes anus during development, bipartite mesocoel, mesocoelomic
ducts; B, stomochord, paired gill slits; C, multiple pairs of gill slits, pharyngeal slits U-shaped, dorsal hollow nerve cord, preoral ciliary organ,
mouth anterior and ventral and anus posterior and ventral or dorsal, multiciliated cells; D CHORDATA, notochord present and not attached to
gut, dorsal hollow nerve cord with neural-plate stage in development, endostyle organ, a true tail used in swimming; E, digestive caecum, open
capillary junctions, somites present, lateral-plate mesoderm, neural tube differentiated into grey and white matter, cerebral vesicle in brain;
F, ciliated extensions of the mesocoel either absent or present as water vascular system (but not as lophophore), anus not anterior and dorsal;
G DEXIOTHETICA, dexiothetism (rotation and partial loss of right side of precursor form), stone canal, calcite skeleton internalization of
protostome; H, specialized olfactory areas in buccal cavity, hind-tail tripartite, dorsal longitudinal canal connected with notochord; I
AMBULACRARIA, trimeric arrangement of the adult coelom, axial complex with hydropore, dipleureula larva with neotroch.
Formation, and still highly controversial. The group sections in front of, and behind, a flexible connection.
was named by Shu et al. (2001) on the basis of three gen- There is a large mouth with a strengthened rim, and
era, Vetulicola, Xidazoon and Didazoon (Figure 1.6(a, preserved internal structures include the guts and a
b)). These animals look like sausage balloons, knotted possible endostyle. Both parts of the body appear to
in the middle: the body is in two parts, with bulbous be crossed by transverse bands of tissue. On the
Chordate Origins 9
The Chengjiang fossil site in Yunnan Province, south-west China, is exciting because it is one of the oldest sources of excep-
tionally preserved organisms, falling early in the great Cambrian radiation of animals in the sea. The fossils come from differ-
ent levels through several hundred metres of mainly fine-grained sediments. When the site was discovered, in 1984, it was
thought to correspond to the already well-known Burgess Shale, a Middle Cambrian locality in Canada that has yielded numer-
ous exceptionally preserved arthropods and the putative chordate Pikaia. Chengjiang, however, is older, dating from the
middle of the Early Cambrian, some 525–520 Myr ago.
Typical Chengjiang fossils, the vetulicolian Xidazoon (a), and the basal vertebrate Myllokunmingia (b), both facing right. Scale bars in
millimetres. Compare with interpretive drawings in Figures 1.6 (b) and 3.1(a). (Courtesy of Shu Degan.)
10 Vertebrate Origin
The Chengjiang site is rich, having produced more than 10,000 specimens, and the fauna consists of 90 or more species,
mainly of arthropods (trilobites and trilobite-like forms), sponges, brachiopods, worms, and other groups, including possible
basal deuterostomes, such as the vetulicolians and yunnanozoons (Figure 1.7), as well as the first fishes (Shu, 2003). Some of
the arthropods are like Burgess Shale animals, but others, such as the basal deuterostomes, seem to be unique. Most of the
animals lived on the bottom of the sea-bed, filtering organic matter from the sediment. There were a few floaters and swim-
mers, and some of the larger arthropods were clearly predators, feeding on the smaller bottom-dwellers.
The Chengjiang beds are grey marine mudstones that preserve soft tissues of many animals in exquisite detail, some re-
placed by phosphate and others by pyrite. Some soft tissues survive as thin organic films. The grey sediment weathers on con-
tact with the air to a light grey or yellow colour, and the fossils may also be grey, or sometimes reddish, and with internal
anatomical details picked out in shades of grey, brown, and black.
Read more at http://palaeo.gly.bris.ac.uk/Palaeofiles/Lagerstatten/chngjang/index.html and http://www.palaeos.com/
gills 1–5 mouth-bearing segment, presumably the front part of
mouth the body, are five circular structures in a row that have
been interpreted as pharyngeal gill slits.
The vetulicolians have been accorded three posi-
gut tions in the phylogenetic tree (Figure 1.7): as basal
deuterostomes, as urochordates or as basal chordates.
Evidence that vetulicolians are deuterostomes are the
10 mm
(a) gill slits and the possible endostyle. They have been in-
terpreted as basal deuterostomes by Shu et al. (2001)
gills 1–5
because they apparently lack an atrium, the internal
chamber in tunicates into which the gill slits and
anus open. In vetulicolians, the intestine terminates
at the end of the body, and the gill slits presumably
opened directly to the outside through openings in the
external body wall. There are, however, some general
resemblances to swimming tunicates in the bulbous
streamlined body shape, the thin external tunic, and
(b) ?endostyle
the regularly spaced transverse bands, which might be
dorsal nerve cord segments
muscles that ran round the body in rings (Lacalli,2002).
The absence of a notochord in vetulicolians is not
critical, since most adult tunicates also have lost this
Additional invertebrate chordates from Chengjiang,
the yunnanozoons Yunnanozoon and Haikouella (Fig-
(c) mouth external gills gut ure 1.6(c)), have been interpreted by rival researchers
both upwards and downwards in the scheme of things
Fig. 1.6 Basal deuterostomes: (a, b) the vetulicolians Didazoon
(a) and Xidazoon (b), showing how the body is divided into two (Figure 1.7). One team identified these animals first as
sections that are joined by a flexible connection; (c) Haikouella. possible cephalochordates (Chen et al., 1995), and then
(Courtesy of Shu Degan.) upwards as vertebrates (Chen et al., 1999; Holland and
Chen, 2001; Mallatt and Chen, 2003). The other team
preferred to regard the yunnanozoons first as hemi-
Chordate Origins 11
hu 200
(C et

Use: 0.0256