ArticlePDF Available

Devonian Dreaming Down Under - Australia's spectacular Palaeozoic fishes



No caption available
No caption available
No caption available
No caption available
No caption available
Content may be subject to copyright.
50 Australian Age of Dinosaurs
The skeletal armour of Mcnamaraspis kaprios – the official State fossil emblem of Western
Australia – is just one of many ancient marine fishes that have now been discovered at Gogo
Station and nearby Mt Pierre Station in the Kimberly district. This small predatory arthrodire
which grew to about 30cm long was found and described by Dr John Long in 1995.
Photo Prof John Long
Issue 10 51
Story by Professor John A. Long
Fishes make life very
enjoyable for most of us.
We enjoy eating many
delectable species of them, from
spicy stewed lamprey through
to delicious raw tuna. In fact
some of us enjoy eating them
so much that a 593 pound blue
fin tuna was sold at Tokyo’s
Tsukiji fish market in January
2012 for a record breaking
$736,000. That’s $2,792 per
kilo! We also enjoy keeping
fishes as pets the aquarium
industry is another big mar-
ket globally, and recreational
fishing is Australia’s largest
pastime. For me though, none
of these things comes close to
the love of working on fossil
Fishes have an incredible fos-
sil record spanning half a bil-
lion years. First appearing at
the end of the Cambrian Period
as worm-like jawless things
the size of your finger, they
became masters of evolution
as their complexity increased
to encompass all of the major
anatomical innovations that
today define the human body
plan. To fishes we owe the
origin of complex skulls, of
jaws and teeth, the advent of
paired limbs (front and rear),
of advanced sensory systems,
multi-chambered hearts, air-
breathing lungs and even com-
plex copulatory reproduction.
All of these evolutionary
events first developed within
fishes during their first 150
million years of evolution, then
later became modified and
fine-tuned as advanced lobe-
finned fishes left the water to
invade land as tetrapods, about
390 to 395 million years ago.
Australia plays a major role
in telling this story as we not
only have some of the oldest
and best preserved fish fossils
found anywhere, but also one
of the most complete records
covering nearly all major fish
52 Australian Age of Dinosaurs
A Brief History of Research
Although the discovery of fish fos-
sils in Australia goes back to Dana in
1849 working in Triassic sediments
around Sydney, the first records of
our truly ancient fishes from the
Palaeozoic Era (540-250mya) dates to
a find in the Buchan region of eastern
Victoria described by Fredrick McCoy,
the founding Director of today’s
Museum Victoria. McCoy erected the
Class Placodermi in 1848, and in
1876 described a bony plate with
tuberculated ornamentation from
the Devonian limestones of eastern
Victoria as Asterolepis cf. ornata”
a kind of antiarch placoderm well
known in the Northern Hemisphere.
Unfortunately this specimen was
later found to be indeterminate as to
its real affinities, apart from being a
kind of placoderm plate.
International attention to our
Palaeozoic fish fossils was first shown
by Dr Arthur Smith-Woodward of the
British Museum in London who in
1906 described a number of new fos-
sil fish species of acanthodians, ray-
finned and lobe-finned fishes from
the Lower Carboniferous red beds
around Mansfield, Victoria (although
his brother Henry Woodward actu-
ally preceded him in describing some
sharks teeth belonging to Helicoprion
from the Permian of Western Australia
in 1888). Arthur Smith-Woodward is
known for working solidly for 40
years straight at the British Museum,
apparently only missing half a day ’s
work in all that time due to a broken
The first serious collection and
description of Devonian fish fossils
from Australia centred on material
found in limestones of the Taemas-
Wee Jasper region around Burrinjuck
Dam of New South Wales. Geologist
Robert Etheridge Jnr described a well-
preserved lungfish skull from here in
1906 which was later fully prepared
out of the rock and described by
Prof Ken Campbell at the Australian
National University (ANU), leading
Campbell to become hooked on fos-
sil lungfish ever after. Middle-Late
Devonian fish remains were collected
from the volcanic complex sediments
around Taggerty and near Mansfield
in central Victoria by Edwin Sherbon
Hills, then at Melbourne University.
Hills was trained in vertebrate palae-
ontology for his PhD in London under
the famous D. M. S. Watson and
became the first Australian to active-
ly research Australian Palaeozoic fish
remains, using them to accurate-
Australias Early Palaeozoic Fish
ly identify the ages of major rock
strata throughout Victoria, New
South Wales and Central Australia.
When the famous Canowindra fish
site in NSW was first discovered in
1956, Hills was contacted to study
the finds. Later, when snowed under
with other duties that lead him away
from his love of palaeontology, he
contacted British expert Keith Stewart
Thompson to take over the descrip-
tion of the first of these well-pre-
served fossils; the lobe-finned fish
named by Thompson as the genus
Canowindra. It would be another 25
years before the true significance of
this site was realised when the road
was widened and new excavations in
the 1990s under the leadership of Dr
Alex Ritchie revealed a great wealth
of highly significant new specimens.
In the late 1950s-early1960s Dr
Errol White of the Natural History
Museum, London, made detailed
descriptions of Australian Devonian
placoderms surreptitiously collected
from the Taemas region and taken
back to London by his preparator
Mr Harry Toombs. Having received
in the late 1940s from an expatri-
ate William Bedford, a well-preserved
Devonian fish fossil (which he named
Williamsaspis bedfordi) White, who
was keen to obtain further speci-
mens, sent Harry Toombs out to
Australia in 1963 on a secretive mis-
sion to collect fish fossils from the
sites. Toombs, walking and camping
around the region like a swagman,
collected a large number of superb
fossil fish specimens but told no-
one at the nearby ANU what he was
doing. Later this became a real issue
with Australian palaeontologists and
following a second collecting trip to
Australia in which Toombs visited
Perth and spied a Gogo fish inside
Founding Director of the
National Museum of Victoria,
Sir Frederick McCoy (left) was
the first person to work on
Australia’s Palaeozoic fishes,
erecting the class Placodermi
or ‘plated skin’ in 1848 when
describing fossils discovered
in the Buchan region of east-
ern Victoria. However it was
not until after the discovery
of Devonian fish fossils near
Burrinjuck Dam in New South
Wales that significant research
of Australia’s Palaeozoic fishes
commenced. Early research
on specimens from this site
was carried out by Robert
Etheridge Jnr of the Australian
Museum (below) in 1906.
Issue 10 53
a nodule, the first steps were taken
towards protecting Australia’s rich
fossil heritage from these events.
Toombs’ Gogo discovery, combined
with an acetic acid etching method
he had recently developed to get fossil
bones out of the limestone, resulted in
two joint expeditions being launched
by the Western Australian (WA)
Museum and the Natural History
Museum of London to collect Gogo
Following the discovery of the Late
Devonian Gogo site in the Kimberly
region of Western Australia, joint
expeditions were undertaken by
the Western Australian Museum
and the Natural History Museum
of London in 1963 and again
in 1967. The 1967 expedition
(above) shows George Kendrick of
the Western Australian Museum
pointing across Paddy’s Valley site
on Gogo Station. Collection and
research of Australian Devonian
fishes continues today and includes
the work of Dr Carole Burrows, Dr
Gavin Young and Prof John Long,
shown at the Craven Peak fish beds
in southwest Queensland (right).
Photo courtesy Ian Rolfe
Photo Prof John Long
54 Australian Age of Dinosaurs
fish fossils from the north of Western
Australia in 1963 and 1967. The Gogo
sites proved superb for recovering 3D
complete fossil fishes of many varie-
ties and were considered treasures of
the Devonian fish world. Because of
this and what had happened previ-
ously at Taemas, Dr David Ride, then
Director of the W.A. Museum, made a
firm written contract with the Natural
History Museum of London that all
Holotypes from the expeditions, and a
fair share of all material found, must
be returned to Australia after the
British scientists had completed their
scientific work on the fossils. This was
the first step taken to prevent inter-
national field parties from entering
Australia and taking whatever they
wished back to their museums with-
out consulting Australian authorities.
In years to come the formalisation of
our Protection of Moveable Cultural
Heritage Laws would place Australia
in the forefront of protecting its fossil
riches a law still lacking in many
major countries today, most nota-
bly the USA. Gogo fish fossils were
further collected in the 1970s by
Gavin Young, Alex Ritchie and Ken
Campbell. In1986 I revisited the site
and made a number of significant
finds there and have been collecting at
the sites continuously since.
In the past 40 years there has
been a great surge of research on
Australia’s early fossil fishes, plac-
ing Australia as one of the leading
countries worldwide in the number
of publications and high-impact
papers. Important contributions in
this field have been made by Dr Alex
Ritchie on agnathans, placoderms and
osteichthyans; Dr Gavin Young on
agnathans, chondrichthyans, placo-
derms and osteichthyans; Prof Ken
Campbell, Dr Richard Barwick and
Dr Peter Pritchard on fossil lungfish-
es and other osteichthyans; and Dr
Susan Turner and Dr Carole Burrows
on agnathans, acanthodians and
shark and placoderm microfossils in
particular. More recently, work has
been undertaken by Dr Kate Trinajstic
on placoderms and sharks; Dr Brian
Choo on Devonian palaeoniscoids; Dr
Timothy Holland on actinopterygians
and tetrapodomorphs, and Dr Alice
Clement on fossil lungfishes.
Discovery of the famous Canowindra fish site in 1956 led to the naming of
Canowindra grossi, (above) by Dr Keith Thompson in 1973. Completely new
to science, this exceptionally well-preserved sarcopterygian was the first
complete lobe-finned fish found in the Southern Hemisphere – the first of
several new species of sarcopterygians discovered following extensive exca-
vations at Canowindra by Dr Alex Ritchie (left) in 1993.
Photo courtesy Dr Alex Ritchie
Photo Prof John Long
Issue 10 55
Australia’s First Fishes
Fish today are easy to define as
‘aquatic back-boned animals that
lack developed limbs with digits’.
However, when fossils are taken into
account this definition becomes nebu-
lous as the first fishes emerge from a
mixture of worm-like animals. Early
Cambrian fossils from Chengjiang,
Yunnan, China, dated at around 520
mya show early jawless forms with
possible vertebral elements preserved
and these are now generally accept-
ed as the oldest true ‘fishes.’ Yet
the oldest bony tissues of vertebrates
come from south central Queensland
found in geological samples of late
Cambrian limestone, around 490-500
million years old, and described by
Dr Gavin Young and his colleagues
back in 1996 in Nature magazine.
Although some other colleagues cast
doubt on these as being vertebrate,
the presence of dentine-like tissues
and external sculptured bone orna-
mentation makes them hard to refute.
Unfortunately we don’t have larger
plates, only scraps, so the shape of
these mysterious early boned fishes
still remains a mystery.
The oldest partially articu-
lated fishes in the world are also
from Australia. This is the jawless
Arandaspis (named for the Aranda
tribe of central Australian indigenous
peoples) and its kin from the 460 to
470 mya Middle Ordovician Stairway
Sandstone sites near Mt Watt in the
Amadeus Basin, Northern Territory.
First described by Dr Alex Ritchie
and Joyce Gilbert-Tomlinson in 1977,
they represent simple elongated fishes
about 25cm long with a complete
external cover of elaborately sculpted
dermal bone, ornamented with a dis-
tinct pattern of large plates covering
the head and trunk area, and long
comb-like scales surrounding the rest
of the body. No fins are determined
from the material. The fish has head-
light-like eyes facing frontwards and
an open mouth, and most likely fed
as either a filter feeder or organic
sediment ingester. A second form,
named Porophoraspis, was identified
by its different type of sculptured
plates. There is additional material of
other new species found at this site
by Dr Ritchie, who is now retired but
continues to work on the specimens
with colleagues. Articulated fishes
found in Bolivia, in slightly younger
Ordovician rocks than Arandaspis,
were identified by European work-
ers and named as Sacabambaspis.
Recently, dermal bones of this fish
have also been reported from central
Australia by Dr Gavin Young. It thus
seems that this Ordovician jawless
fish fauna containing Arandaspis-
Sacabambaspis was widespread across
Gondwana around 450 to 470 million
years ago.
Other additions to our know-
ledge of the Ordovician fishes from
Australia shows the fish fauna at this
time was more diverse than previ-
ously thought. In his 1997 study
of Ordovician microfossils from the
Amadeus Basin, Gavin Young identi-
fied a few other kinds of fossil ver-
tebrate scales which he attributed to
possible early sharks (Areyongalepis’),
but as we only have the scales, the
true nature of these early fishes still
remains obscure. Another strange
shark-like scale was recently identi-
fied from the Stairway Sandstone by
UK researcher Ivan Sansom and col-
leagues (including Alex Ritchie from
Australia), who, in 2012 named it
Tantalepis in reference to the enig-
matic nature of the scales. The prob-
lem here is that they look shark-like
in their structure but until such scales
are found fully associated with more
complete remains they could alterna-
tively belong to an as yet unknown
group of early jawed fishes or ‘stem
Australia has a sparse but inter-
esting assemblage of Silurian fish
remains, mostly known from iso-
lated scales, teeth and bits of der-
mal bone extracted from residues
of limestones. They represent early
osteichthyans (bony fishes), placo-
derms, sharks, acanthodians and jaw-
less thelodonts, based on work done
largely by Carole Burrows and Susan
Turner of Queensland. We do pos-
sess one almost complete fish of this
age and that is Yealepis, an enigmatic
acanthodian-like gnathostome (jawed
fish) about 20cm long with tiny
scales. This specimen was described
by Carole Burrows and Gavin Young
in 1999 and came from the world
famous Baragwanathia plant site in
central Victoria which is of upper-
most Silurian age, around 420 mya.
Discovered near Mt Watt in the Northern Territory, the bony shields of
Arandaspis (above) represent the oldest partially articulated fishes in the
world. This 25cm long jawless species was probably a filter
feeder or organic sediment ingester and is
characterised by its forward
facing eyes and open
mouth (right).
Image Nobu Tamura
56 Australian Age of Dinosaurs
Early Devonian Fishes
Australia excels with one of the finest
records of fish fossils in the world for
the Devonian period (359-419mya),
not only through its high diversity
of species but also because it has the
most superb preservation of any fish-
es of this age known. Although we
have many sites scattered throughout
Australia, the most significant assem-
blages are the Taemas-Wee Jasper
fauna of New South Wales and the
similar aged Buchan Fishes in east-
ern Victoria. At this time fish also
began to grow very large. The larg-
est fishes before the Devonian period
are all under 50cm or so, most being
about 10-20cm long. By the end of
the Early Devonian, around 400mya,
some placoderms grew to almost 4m
in length and the biggest bony fishes
– enormous lungfishes – were around
1.5 metres long. This massive increase
in size of fishes is now thought to be
related to the development of myelin
sheaths around the nerves, enabling
them to grow longer.
Imagine southern New South Wales
and eastern Victoria around 400 mil-
lion years ago. You would be swim-
ming in warm, shallow seaways with
colourful patches of reef (bioherms)
here and there, with well-developed
reef complexes clustered around
where the Burrinjuck Dam exists
today. Today these murky grey-blue
limestones are packed full of inver-
tebrate fossils of many kinds includ-
ing snails, clams, corals, crinoids,
bryozoans and brachiopods as well as
yielding bones of the first great fish
radiation documented in Australia.
The fauna is exceptionally diverse
with close to 70 different species now
known from the sites, of which some
45 species belong to the placoderms.
Their remarkable preservation as 3D
skulls, isolated bones and occasional
partially-articulated skeletons means
that these fossils have had very high
scientific value in debates about early
vertebrate evolution. Their great clar-
ity of anatomical detail makes them
a definitive reference to turn to when
deciphering the origins of the major
vertebrate body plan at this time,
such as determining when and how
paired limbs or jaws, or a certain pat-
tern of skull bones, first developed.
The most abundant fishes of the
Taemas fauna are the armour-plated
placoderms (Greek for ‘plated skin’).
These were the first jawed fishes
and sit at the very base of the ver-
tebrate evolutionary tree. The placo-
derms of the Taemas-Buchan district
were first described by Errol White
of the Natural History Museum in
London, but the most recent series of
publications have been by Dr Gavin
Young of ANU, Canberra. Young
has described exquisitely preserved
braincases showing pathways of
cranial nerves, arteries and veins,
muscle attachments and so on, ena-
bling accurate estimation of the soft
anatomy of these early jawed fishes.
His 1980 monograph describing the
cranial anatomy of Brindabellaspis,
an acanthothoracid placoderm with
beautifully sculptured dermal orna-
ment, is still widely used by research-
ers around the world. Similarly his
1979 work on Buchanosteus, actu-
The Spectacular Devonian Fishes of Australia
The Taemas Wee Jasper region of New South Wales has produced a wealth of
Early Devonian fishes, the most abundant being armour-plated placoderms such
as Buchanosteus (above left). As the first jawed fishes, placoderms fit at the very
base of the vertebrate family tree and some of the Taemas faunas had remarkable
features – such as the heavily armoured eyeball of Murrindalaspis (above right).
Photos Dr Gavin Young
Issue 10 57
ally the first of the new placoderm
species shows the structure of the
early arthrodire skull. This fish was
named from work published earlier
by Edwin Hills and Erik Stensio of
Sweden in 1945. A recent paper by
Young also describes the jaws and the
snout of a Buchanosteus-like placo-
derm, with armoured eyes forming
the front of the fish and enclosing
the nasal openings. Arthrodires were
the most diverse group of all placo-
derms, and the Taemas-Buchan fos-
sils are intrinsic to understanding the
evolutionary radiation of the group.
Some of the Taemas placoderms had
remarkable features, like the heavily
armoured eyeball of Murrindalaspis,
covered in thick dermal bones and
with the internal surface completely
ossified to show the muscle attach-
ments and soft tissue features. Using
the remarkable micro CT scanner
at the ANU’s department of Applied
Mathematics, Prof Tim Senden has
revealed details inside the eye to fol-
low the pathways of its nerves, arter-
ies and internal muscle attachments
for the retractor bulbi muscle.
Taemas is also well known for its
superb fossil bony fishes. The ancient
lungfishes were fully marine crea-
tures with thick shiny bones, quite
unlike the living Queensland lung-
fish that inhabits the freshwater riv-
ers of central south Queensland. We
have around half a dozen species
known from the Taemas deposits,
some like Dipnorhynchus cathlesae
were enormous, with massive crush-
ing lower jaws that were 22cm wide.
This fish must have been around
1.5-2 metres in length, making it
the largest osteichthyan at this time
anywhere in the world. The detailed
descriptions of these fishes by Prof
Ken Campbell and Dr Richard Barwick
of ANU revealed remarkable details
of their reconstructed jaw muscula-
ture, soft tissue anatomy and brain
structure obtained from endocranial
casts and CT scanning. These fishes
show that dental tissues evolved into
many experimental types. Some had
palates covered with a massive thick
sheet of dentine, others with dentine
and patches of denticles, and others
with rows of conical to blunt rounded
teeth as in Speonesydrion. Almost all
of the Taemas lungfishes possessed
powerful crushing jaws to grind up
hard-shelled invertebrate prey.
The modern ray-finned fishes are
the most abundant species on Earth
with nearly 29,000 species known.
The oldest members of this group are
known largely from isolated scales
and bits of bone dated as late Silurian
age such as Naxilepis from China. A
tiny skull named Ligulalepis found at
Taemas is now considered as the old-
est well-preserved ‘stem osteichthyan’
which means it is placed in the evolu-
tion of bony fishes as a node on the
tree just before true actinopterygians
appeared. Its discovery was so signifi-
cant that it was published in Nature
in 2000 by Alison Basden, Gavin
Young, Alex Ritchie and Mike Coates.
It retains certain primitive features in
its braincase shared with other basal
osteichthyans like Psarolepis from the
Early Devonian of China.
A layer of nodular concretions col-
lected in the 1960s and 1970s from
sediments immediately below the
limestones at Buchan (called ‘The
Fairy Beds’) by a team from Monash
and Macquarie Universities have
produced a fauna of tiny isolated
small fish bones and teeth from early
osteichthyans. From this material,
Zerina Johanson and colleagues have
described the oldest known member
of the coelacanth group (Eoactinistia)
based on a lower jaw as well as the
oldest member of the predatory ony-
chodontid fishes (Bukkanodus).
The ray-finned fishes are today the most
abundant species on Earth with over 29,000
species known. However this group was
only just beginning to develop in the Early
Devonian. At around 400 million years
old, the tiny skull of Ligulalepis (below)
from Taemas, is considered the oldest well
preserved stem osteichthyan in the world.
Photo Prof John Long
Unlike freshwater lungfish inhabiting the rivers of
central south Queensland today, Early Devonian
lungfish from the Taemas Wee Jasper region were
fully marine with some species growing to over 1.5
metres long. About six different species possessing a
variety of dentition types are now known from this
site. The blunt rounded teeth and powerful crushing
jaws of Speonesydrion (above) were used for crushing
up hard-shelled prey that would have formed part
of the animal’s diet.
Photo Prof John Long
58 Australian Age of Dinosaurs
Middle Devonian Fishes
A diverse assemblage of mostly
placoderms with some bony fish-
es and strange jawless fishes comes
from the sandstone strata straddling
the Early-Middle Devonian boundary
from outcrops in two main regions,
one near Cobar in western NSW,
and the other from the Cravens Peak
beds near the Toko Range in south-
western Queensland. In 1973, Alex
Ritchie described a peculiar placo-
derm from here which he named
Wuttagoonaspis – considered to be a
bizarre kind of arthrodire. The faunas
have been named the Wuttagoonaspis
fauna because of this. Some species
of Wuttagoonaspis were quite large,
with W. milligani reaching about a
metre long. Today these faunas are
significant for the very large number
of endemic placoderms that inhabited
these ancient drainage basins, and in
particular for the very odd jawless
armoured fishes called ‘pituriaspids’.
Gavin Young gave the fishes this
name after the hallucigenic Pituri
plant as the fish were so strange
it was like he was hallucinating.
Pituriaspis belongs in its own Class of
Vertebrates. The ‘Pituriaspida’ was
the first one named in over 70
years when erected by Young
in 1991. Today we only have
casts of the armours of these
strange fishes (belonging
in two genera), so more
information on them
is vital to determine
their relationships to
the first jawed verte-
A Devonian fossil
fish site in the beau-
tiful central highlands
of eastern Victoria near
Mt Howitt was discovered by
geologist Mark Marsden in the late
1960s. It became of great interest as
it produced whole fishes preserved
between layers of finely laminated
black shale. The site was then collected
over several field seasons by Monash
University teams lead by Prof James
Warren. I was very fortunate to
work on these fishes for my honours
and PhD theses whilst at Monash
University between 1980 and 1983.
In the late 1960s a late Middle Devonian fish fossil site was discovered near Mt Howitt in the central highlands
of eastern Victoria. This site was extensively collected over several years by teams from Monash University (above)
led by Prof Jim Warren. The Mt Howitt fishes are particularly significant as they represent periodic mass kills
thus preserving fishes in all stages of their growth from tiny larvae to fully mature adults.
Photo courtesy Jim Warren
The skull of Beelarongia, an unusual
tetrapodomorph (lobe-finned fish)
from the Mt Howitt fish site in
eastern Victoria (above).
Photo Prof John Long
Issue 10 59
The Mt Howitt fishes are now
known to be of late Middle Devonian
age (c385 mya), and are especially
significant because they represent
periodic mass kills that have been
preserved complete showing all stages
of their growth – from tiny larvae
through to full grown adults. The
most important finds coming out of
the site are the well-preserved early
phyllolepid placoderms. These were
flattened placoderms (like flounders)
with distinct patterns of concentric
bony ornament on the external sur-
faces of their plates. Austrophyllolepis
from Mt Howitt was one of the first
endemic members of this group to
be described from Australia along
with Placolepis from NSW (both pub-
lished in 1984). Recently the pelvic
girdle of Austrophyllolepis showed an
elongated bone (a ‘basipterygium’)
that may have borne a clasper in the
males. This was figured in a 2009
Nature paper on the origins of inter-
nal fertilisation in vertebrates. An
early ray-finned fish from the site,
which I named Howqualepis, has been
of great interest to researchers as a
well-documented example of a basal
ray-finned fish. Although looking
vaguely trout-like, it has inflexible
cheeks and long lower jaws – quite
unlike the modern ray fins with their
protruding mouths.
385 million years ago Mt Howitt
was home to a diverse group of
freshwater fishes including the
phyllolepids – an extinct taxon of
flattened placoderms, lobe-finned
and ray-finned fishes and several
species of lungfishes. The trout-like
Howqualepis (right) was a basal
ray-finned fish that had inflexible
cheeks and long lower jaws – quite
unlike the protruding mouths of
its modern equivalent. The skull of
Howqualepis (above) is a latex peel
of the original that has been whit-
ened with ammonium chloride for
Image Dr Brian Choo
60 Australian Age of Dinosaurs
Late Devonian Fishes
The Late Devonian Gogo sites cover
a wide area of land on Gogo and Mt
Pierre Stations in the Kimberley dis-
trict of Western Australia. Here fossils
are collected from weathered nodules
that litter the ground although fish
are a rare occurrence in them; most
nodules having either nothing or the
occasional remains of a crustacean
inside them. The fossils can be pre-
pared out of the nodules with weak
acetic or formic acid. The preservation
of these fishes is truly superb, with
all the bones extracted in perfect 3D
form, and often the entire fish is pre-
served in the nodule. Sometimes even
muscles, nerves and other soft tissues
are preserved through a process of
bacterial replacement.
Early research on the fauna from
the mid 1970s through to the late
1990s was published by British sci-
entists Dr Roger Miles, Brian Gardiner
and Sheila Mahala Andrews, who
described the detailed anatomy of
the fishes to test phylogenetic rela-
tionships of placoderms and early
osteichthyans. Work on the Gogo
fauna to date includes research by Ken
Campbell, Richard Barwick and Peter
Pridmore on Gogo lungfishes (1980-
2012); by Ken Campbell, Richard
Barwick and me on the Gogo onycho-
dontid (2006) and tetraopodomorph
Gogonasus (1997); by Gavin Young,
Kate Trinajstic, Tim Senden and me
on Gogo placoderms (1990s-2012)
and Gogonasus (2006); on the origins
of Gogo placoderm teeth by Zerina
Johanson and Moya Smith (2003);
and most recently by Brian Choo on
revision of Gogo ray-finned fishes
(2009-2012), and Alice Clement on
new Gogo lungfishes (2010-2012).
A recent review of work published
on the Gogo fishes up until 2010 is
provided in Long & Trinajstic (2010).
The most spectacular finds from
Gogo include the ‘mother fish’, a
ptyctodontid placoderm collected on
my 2005 expedition which we named
Materpiscis attenboroughi (after Sir
David Attenborough). It showed the
first ever in situ placoderm embryo,
still attached by a mineralised umbili-
cal cord. Although previous finds of
this strange little placoderm showed
them to be sexually dimorphic with
males having bony claspers to insert
inside the females, no ‘hard evidence’
for their mode of sexual reproduction
was known until the discovery of this
fossil. It confirmed that ptyctodon-
tids mated by copulation and pushed
back the fossil evidence for viviparity
(live birth) in vertebrates by about
180 million years. The discovery
Looking for the tell-tale sign of a treasure within, Dr John Long studies yet another nodule on Gogo Station in the
central Kimberly region of Western Australia (above). Commonly referred to as the Gogo Fish Beds, this region is
part of the Gogo Reef Formation, a Late Devonian shale and limestone deposit that has preserved a diverse range
of ancient marine life including some of Australia’s most spectacular and significant fossil fishes.
Photo Prof John Long
Issue 10 61
The discovery in 2005 of Materpiscis attenboroughi, or “Mother
Fish” (above) showed the first ever in-situ placoderm embryo,
still attached by a permineralised umbilical cord. Due to prior
discoveries of males with bony claspers that would have enabled
them to mate with females (right), this species was believed to
be sexually dimorphic but there was no exact evidence of their
mode of reproduction. Discovery of Materpiscis confirmed that
this ptyctodontid placoderm mated by copulation and pushed
back the fossil evidence of live birth by 180 million years (inset
above). Of the 45 or so species of fish now discovered from the
Gogo site, the majority are placoderms. The well preserved skull
of Bothriolepis, an antiarch placoderm (below), is one of the rarer
species found at Gogo and is believed to be a mainly freshwater
species that migrated between river systems via the sea.
Photos Dr John Long
62 Australian Age of Dinosaurs
was published in Nature (2008) and
even made it into the Guinness World
Records 2010. Other finds from Gogo
that have made scientific headlines
include the oldest male sexual organ,
or clasper identified in an arthrodiran
placoderm (Nature 2009), and the first
evidence of a large spiracular opening
on the top of the head of the lobe-
finned fish Gogonasus (Nature 2006).
The latter feature is now thought to
be related to these fishes being able to
take in air through the spiracles to the
lung via a chamber inside the mouth.
Studies of the superb lungfish-
es from Gogo have revealed many
startling features, not only in the
strange ways they grow and remodel
their dental structures through life,
but also how the great variety of
lungfishes present at Gogo adapted
to a wide variety of feeding strate-
gies on the ancient reef. Some like
Holodipterus had very strong crush-
ing jaws, whilst others like the long-
snouted Griphognathus, (nicknamed
‘the duck-billed lungfish’) possessed
a delicate shagreen of fine denticles
like sandpaper on its palate and lower
jaws for rasping down food such
as branching bryozoans and corals.
Another recent discovery was a lung-
fish from Gogo with skeletal adapta-
tions for air breathing, described by
Alice Clement.
The Canowindra Site in central
NSW represents one of the most
famous mass kill sites for Australian
Devonian fishes. Many thousands of
complete fishes, represented by natu-
ral casts of the external and internal
bones in red sandstone and fine mud-
stone, have been collected from here
by Alex Ritchie and his colleagues
since the road widening took place
in the 1990s. The Canowindra fauna
Issue 10 63
The Late Devonian equivalent of today’s Great
Barrier Reef, the Gogo Reef was a 1,400-kilometre
barrier reef – positioned off the northern coastline
of Gondwana 380 million years ago. Once home
to a menagerie of bony fishes, placoderms,
crustaceans and conodont animals (above) the
limestone sediments of this ancient structure
have now revealed a diverse fish fauna including
lobe-finned fishes such as Gogonasus (top left)
and this rare, complete skeleton of a coelacanth
(middle left). Several species of ray-finned fishes
are also present and include this exceptionally
well preserved specimen (right). The long-snouted
lungfish Griphognathus (left) is one of many superb
lungfish specimens also recovered from the site.
Photos Dr John Long
Image Brian Choo
64 Australian Age of Dinosaurs
contains about seven or eight spe-
cies. Today the best examples are dis-
played in Canowindra’s ‘Age of Fishes’
museum. The fauna is dominated by
many individuals of the antiarch pla-
coderms Bothriolepis and Remigolepis,
and rarer examples of the arthrodire
Groenlandaspis, but is best known
internationally for its very inter-
esting large lobe-finned fishes. The
most primitive known rhizodontid
Goologongia was described from here
by Zerina Johanson and Per Ahlberg
in Nature (1999), as were large preda-
tory lobe-finned tristicopterid fishes
like Mandageria and Cabonnichthys. A
long-snouted lungfish, Soderberghia,
also known from North America and
Greenland, has also been described
from the site.
Early Steps on Land
The biggest evolutionary event at
the end of the Devonian period was the
transition from advanced lobe-finned
fishes called ‘tetrapodomorphs’, to
primitive amphibians (‘tetrapods’)
whose fish-like bodies and skulls
are only distinguished by the digits
on the end of their arms and legs.
Early aquatic forms like Acanthostega
from East Greenland bore up to eight
fingers and toes, whilst others like
Ichthystega bore seven digits on its
limbs, and Tulerpeton had six digits.
The pattern of having just five fingers
and toes in tetrapods only stabilised
after terrestrialisation was complete
in the early Carboniferous period.
In 1972 zoologist Norman
Wakefield discovered the world’s old-
est trackways of early tetrapods on
the banks of the Genoa River in far
western Victoria. Accompanied by
Jim Warren from Monash University,
Wakefield collected the specimens by
helicopter. Recent geological studies
by Gavin Young has confirmed cor-
relation of the trackway site with
nearby NSW east coast Devonian
fish sites near Eden, indicating that
the Genoa trackways are most like-
ly lower Frasnian in age (around
380 mya). Potentially much older
tetrapod trackways were described
from sandstone coming from Glenilsa
in the Grampians by Anne Warren
and Barry Bolton in 1986 but were
doubted by Northern Hemisphere
tetrapod workers as too old to be
valid. However since the recent find of
395-million-year-old tetrapod tracks
Canowindra in central New South Wales is regarded as one of Australia’s most famous mass kill sites for
Australian Devonian fishes – and this beautifully preserved specimen of Cabonnichthys burnsi (above) is but one of
several well documented lobe-finned fishes from the site. Canowindra’s advanced sarcopterygians, which have
primitive limb bones, would have been prime candidates for evolutionary adaptation into early amphibians or
‘tetrapods’ at the end of the Devonian period.
Photo Prof John Long
Issue 10 65
from the Middle Devonian of Poland
(which featured on the cover of Nature
in January 2010), the Grampians
tracks are now taken a lot more seri-
ously and are currently being fur-
ther investigated by a team involving
Australian, US, UK and Spanish work-
ers. Australia might well have been the
place for the most historic step taken
by any vertebrate – not
Neil Armstrong’s first
step on the moon –
but the first step
on land by a
tetrapod animal.
The Author
Professor John Long graduated from Monash
University with a PhD in Palaeozoic fish evolution
in 1984. Following postdoctoral positions in several
state universities, he took on a position as Curator of
Vertebrate Palaeontology at the Western Australian
Museum in 1989 where he remained for 15 years. He
was Head of Sciences at Museum Victoria from 2004
until 2009 before relocating overseas as Vice President of
Research and Collections at the Natural History Museum
of Los Angeles County in USA. In November 2012 John
returned to Australia and is now based at the School of
Biological Sciences at Flinders University. His main area
of research is on the Late Devonian Gogo fish of northern Western Australia.
Professor John Long is an accomplished author who has published numerous popular articles and books. One of his most
famous works The Rise of Fishes was first published in 1995 with an updated edition following in 2010. Other works on prehis-
toric fishes by Professor Long include: Swimming in Stone (2006); Hung Like an Argentine Duck (2011) and Frozen in Time (2011).
Like an immature tadpole undertaking
its first tentative steps on land, the early
tetrapods were simply a fish with stumpy
legs that enabled them to make increas-
ingly longer forays out of the safety of
water (right). Tetrapod trackways found on
the banks of the Genoa River in far western
Victoria (above) have been dated at about
380 million years old.
Photo Prof John Long Image © John Sibbick
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.