ArticlePDF Available

Abstract and Figures

Coprolites from the Rhaetian bone beds in southwest England can be assigned to crustaceans and fishes. Here, we report crustacean microcoprolites, including Canalispalliatum and Favreina, the first records from the British Rhaetian, from Hampstead Farm Quarry near Bristol, evidence for diverse lobsters and their relatives not otherwise represented by body fossils. Further, we identify five fish coprolite morphotypes that differ in shape (cylindrical, flattened) and in presence or absence of a spiral internal structure. Many coprolites show bony inclusions on the surface, often relatively large in proportion to the coprolite; these show little or no evidence for acid damage, suggesting that the predators did not have the physiological adaptations of many modern predatory fishes and reptiles to dissolve bones. CT scanning has revealed the nature, packing and identity of inclusions within the coprolites, mainly fish scales, and some coprolites can contain more than twenty. An extraordinary discovery in one coprolite comprises a single sculptured skull element of the large bony fish Severnichthys together with two caudal vertebrae of the marine reptile Pachystropheus: did the coprolite producer, likely a fish, scavenge some flesh from the head of Severnichthys and then bite off the tail of the reptile? Assigning coprolites to producers is difficult, but it seems that Gyrolepis was preyed on by nearly every predator. The faunas and trophic relations revealed by the coprolites show that this was a modern-style marine ecosystem, with abundant crustaceans and several species of durophagous fishes, evidence for an early stage in the Mesozoic Marine Revolution.
Content may be subject to copyright.
Fish
and
crab
coprolites
from
the
latest
Triassic
of
the
UK:
From
Buckland
to
the
Mesozoic
Marine
Revolution
Marie
Cueille
a
,
Emily
Green
a,b
,
Christopher
J.
Dufn
a,c,d
,
Claudia
Hildebrandt
a
,
Michael
J.
Benton
a,
*
a
School
of
Earth
Sciences,
University
of
Bristol,
Wills
Memorial
Building,
Queens
Road,
Bristol,
BS8
1RJ,
UK
b
School
of
Life
Sciences,
University
of
Lincoln,
Brayford
Pool
Campus,
Lincoln,
LN6
7TS,
UK
c
146
Church
Hill
Road,
Sutton,
Surrey,
SM3
8NF,
UK
d
Earth
Sciences
Department,
The
Natural
History
Museum,
Cromwell
Road,
London,
SW7
5BD,
UK
A
R
T
I
C
L
E
I
N
F
O
Article
history:
Received
6
May
2020
Received
in
revised
form
24
July
2020
Accepted
25
July
2020
Available
online
xxx
Keywords:
Coprolites
Decapoda
Chondrichthyes
Osteichthyes
Rhaetian
Rhaetian
bone
bed
Penarth
Group
Westbury
Formation
Mesozoic
Marine
Revolution
A
B
S
T
R
A
C
T
Coprolites
from
the
Rhaetian
bone
beds
in
south-west
England
can
be
assigned
to
crustaceans
and
shes.
Here,
we
report
crustacean
microcoprolites,
including
Canalispalliatum
and
Favreina,
the
rst
records
from
the
British
Rhaetian,
from
Hampstead
Farm
Quarry
near
Bristol,
evidence
for
diverse
lobsters
and
their
relatives
not
otherwise
represented
by
body
fossils.
Further,
we
identify
ve
sh
coprolite
morphotypes
that
differ
in
shape
(cylindrical,
attened)
and
in
presence
or
absence
of
a
spiral
internal
structure.
Many
coprolites
show
bony
inclusions
on
the
surface,
often
relatively
large
in
proportion
to
the
coprolite;
these
show
little
or
no
evidence
for
acid
damage,
suggesting
that
the
predators
did
not
have
the
physiological
adaptations
of
many
modern
predatory
shes
and
reptiles
to
dissolve
bones.
CT
scanning
has
revealed
the
nature,
packing
and
identity
of
inclusions
within
the
coprolites,
mainly
sh
scales,
and
some
coprolites
can
contain
more
than
twenty.
An
extraordinary
discovery
in
one
coprolite
comprises
a
single
sculptured
skull
element
of
the
large
bony
sh
Severnichthys
together
with
two
caudal
vertebrae
of
the
marine
reptile
Pachystropheus:
did
the
coprolite
producer,
likely
a
sh,
scavenge
some
esh
from
the
head
of
Severnichthys
and
then
bite
off
the
tail
of
the
reptile?
Assigning
coprolites
to
producers
is
difcult,
but
it
seems
that
Gyrolepis
was
preyed
on
by
nearly
every
predator.
The
faunas
and
trophic
relations
revealed
by
the
coprolites
show
that
this
was
a
modern-style
marine
ecosystem,
with
abundant
crustaceans
and
several
species
of
durophagous
shes,
evidence
for
an
early
stage
in
the
Mesozoic
Marine
Revolution.
©
2020
The
Geologists'
Association.
Published
by
Elsevier
Ltd.
All
rights
reserved.
1.
Introduction
Coprolites,
fossilized
faeces,
have
been
reported
for
over
200
years,
and
indeed
those
from
the
marine
reptiles
and
shes
of
the
British
Mesozoic
featured
in
classic
early
researches
by
luminaries
such
as
William
Buckland
(Buckland,
1829a;
Dufn,
2009,
2012b).
Although
Buckland
coined
the
term
coprolite
in
that
1829
paper,
the
rst
report
of
a
vertebrate
coprolite
had
been
written
much
earlier,
by
Edward
Lhwyd
in
1570
(Dufn,
2012a).
Coprolites
are
fossilized
faeces,
and
they
are
regarded
as
trace
fossils
(ichnofossils),
together
with
tracks,
trails
and
burrows,
because
they
record
behaviour.
Most
coprolites
have
their
original
organic
matter
replaced
by
minerals,
but
elements
of
digested
food
may
remain.
Coprolites
can
provide
unique
insights
into
otherwise
largely
inaccessible
aspects
of
the
palaeoecology
of
ancient
ecosystems,
and
provide
direct
evidence
of
tropic
relationships.
Their
morphologies
may
preserve
information
concerning
the
structure
of
the
gut
of
their
producers.
Further,
inclusions
within
the
faecal
groundmass
of
the
coprolite
represent
either
partially
digested
or
undigested
food
components
which,
in
turn,
reect
aspects
of
feeding
and
diet
including
prey-selection
behaviour,
the
means
of
food
ingestion
and
processing,
mechanical
and
chemical
digestion
techniques
and
their
associated
physiology.
While
inferences
can
be
made
concerning
the
assignment
of
coprolites
to
their
producers,
unequivocal
association
is
only
possible
when
the
faecal
material
is
still
within
the
body
cavity
of
its
host
(termed
a
consumulite
by
Hunt
et
al.,
2012).
Indeed,
coprolites
themselves
represent
unique
taphonomic
microenvironments
with
the
potential
to
act
as
conservation
Lagerstätten
in
their
own
right
(Qvarnström
et
al.,
2016).
Although
relatively
rare
in
the
fossil
record,
when
found,
vertebrate
coprolites
can
sometimes
be
remarkably
abundant
and
*
Corresponding
author.
E-mail
address:
mike.benton@bristol.ac.uk
(M.J.
Benton).
https://doi.org/10.1016/j.pgeola.2020.07.011
0016-7878/©
2020
The
Geologists'
Association.
Published
by
Elsevier
Ltd.
All
rights
reserved.
Proceedings
of
the
Geologists
Association
xxx
(2020)
xxxxxx
G
Model
PGEOLA-855;
No.
of
Pages
23
Please
cite
this
article
in
press
as:
M.
Cueille,
et
al.,
Fish
and
crab
coprolites
from
the
latest
Triassic
of
the
UK:
From
Buckland
to
the
Mesozoic
Marine
Revolution,
Proc.
Geol.
Assoc.
(2020),
https://doi.org/10.1016/j.pgeola.2020.07.011
Contents
lists
available
at
ScienceDirect
Proceedings
of
the
Geologists
Association
journa
l
homepage
:
www.e
lsevier.com/loca
te/pgeola
well
preserved
because
of
their
commonly
high
phosphatic
content
(Hunt
et
al.,
2007),
and
this
is
especially
true
in
marine
bone
beds.
Here
we
focus
on
coprolites
from
the
classic
Rhaetian
bone
beds
of
SW
England,
a
theme
that
combines
insights
into
the
early
work
of
Buckland,
but
also
our
understanding
of
that
time
of
turmoil,
and
the
general
use
of
coprolites
in
reconstructing
ancient
palaeoecology.
The
Rhaetian
was
the
short
(4.1
Myr;
205.5201.4
Ma;
Kent
et
al.,
2017)
nal
stratigraphic
stage
of
the
Triassic,
and
famed
in
Europe
for
evidence
of
a
major
marine
transgression
that
terminated
the
underlying
continental
red
bed
successions
and
frequently
commenced
with
a
bone
bed.
In
the
Bristol
area,
there
can
be
as
many
as
ve
or
six
bone
beds
throughout
the
Rhaetian,
classied
as
the
Penarth
Group,
and
predominantly
in
the
basal
Westbury
Formation,
but
sometimes
in
the
lower
parts
of
the
overlying
Cotham
Member
of
the
Lilstock
Formation
(Swift
and
Martill,
1999;
Allard
et
al.,
2015).
Sepkoski
(1984)
identied
a
major
changeover
in
marine
faunas
from
Permian
to
Triassic,
in
which
Paleozoic
ecosystems
comprising
brachiopods,
crinoids,
trilobites
and
graptolites
were
replaced
by
Modern
ecosystems
comprising
bivalves,
gastropods,
echinoids,
malacostracans,
and
neopterygian
shes.
Sepkoski
tied
this
changeover
to
the
devastating
effects
of
the
Permian-Triassic
mass
extinction,
and
the
Modern
fauna
emerged
during
the
recovery
of
life
in
the
Triassic
(Chen
and
Benton,
2012).
In
recent
years
(e.g.
Harper,
2003;
Hautmann,
2004;
Baumiller
et
al.,
2012),
this
restructuring
of
marine
ecosystems
in
the
Triassic
has
been
posited
as
a
rst
step
in
the
Mesozoic
marine
revolution
(Vermeij,
1977 ),
associated
with
escalations
in
predator-prey
interactions.
The
aims
of
this
paper
are
to
explore
the
diversity
of
coprolites
from
the
British
Rhaetian
and
to
use
these
to
reconstruct
a
likely
food
web
for
the
Rhaetian
seas
and
to
consider
how
this
reects
novel
trophic
relations
as
part
of
the
establishment
of
the
Modern
marine
ecosystem
during
the
Triassic.
We
develop
the
theme
of
the
historical
investigation
of
Rhaetian
coprolites
rst,
as
this
takes
us
back
nearly
200
years
to
the
origin
of
the
discipline,
and
we
reect
on
the
fact
that
Buckland
and
contemporaries
were
examining
the
same
Rhaetian
coprolite
specimens
we
study
today.
Repository
abbreviations:
BRSMG,
Bristol
City
Museum,
Geology
Collection;
BRSUG,
University
of
Bristol
Geology
Museum;
OXFUM,
Oxford
University
Museum
of
Natural
History.
Fig.
1.
William
Buckland
and
his
Rhaetian
coprolites.
(A)
The
Reverend
William
Buckland
D.D,
F.R.S.,
Canon
of
Christ
Church
and
Professor
of
Geology
and
Mineralogy
in
the
University
of
Oxford,
1833.
Painted
by
Thomas
Phillips
Esq.
R.A.
Engraved
by
Samuel
Cousins.
This
mezzotint
engraving
was
produced
by
Molteno
&
Graves
(London),
May
20th
1833.
Wellcome
Collection,
London
(CC
BY
4.0).
(B)
Rhaetian
coprolites
as
gured
by
Buckland
(1835,
pl.
30,
gs.
1329).
(C,
D)
OXFUM
J23743,
holotype
of
Strabelocopros
pollardi
Hunt,
Lucas
and
Spielmann,
2012
from
the
Rhaetian
of
Watchet,
Somerset.
2
M.
Cueille
et
al.
/
Proceedings
of
the
Geologists
Association
xxx
(2020)
xxxxxx
G
Model
PGEOLA-855;
No.
of
Pages
23
Please
cite
this
article
in
press
as:
M.
Cueille,
et
al.,
Fish
and
crab
coprolites
from
the
latest
Triassic
of
the
UK:
From
Buckland
to
the
Mesozoic
Marine
Revolution,
Proc.
Geol.
Assoc.
(2020),
https://doi.org/10.1016/j.pgeola.2020.07.011
2.
Rhaetian
coprolites:
history
of
research
William
Buckland
(17841856;
Fig.
1A)
read
a
paper
to
the
Fellows
of
the
Geological
Society
of
London
on
6th
February
1829
that
spanned
several
topics
the
rst
pterosaur,
fossil
sepia
in
cephalopods
and
fossil
faeces
in
ichthyosaurs,
all
from
the
Liassic
rocks
outcropping
at
Lyme
Regis
(Buckland,
1829a).
Four
months
later,
at
the
meeting
of
the
Society
on
May
1st,
he
revisited
the
idea
of
the
existence
of
fossil
faeces
and
recorded
that
:
He
has
also
ascertained,
by
the
assistance
of
Mr.
Miller
and
Dr.
Prout,
that
the
small
black
rounded
bodies
of
various
shapes,
and
having
a
polished
surface,
which
occur
mixt
with
bones
in
the
lowest
strata
of
the
lias
on
the
banks
of
the
Severn,
near
Bristol,
are
also
of
faecal
origin:-
they
appear
to
be
co-extensive
with
this
bone
bed,
and
occur
at
many
and
distant
localities.
(Buckland,
1829b,
p.
142 )
Later
in
the
same
paper,
concerning
these
and
other
specimens,
he
proposed
to
include
them
all
under
the
generic
name
of
Coprolite
(Buckland,
1829b,
p.
143 ).
The
Rhaetian
Stage
was
not
recognised
in
Britain
until
the
1850s
(Dufn,
2019),
so
Bucklands
remarks
concerning
the
Lias
bone
bed
actually
refer
to
the
Rhaetian
Bone
Bed.
The
fuller
account
of
his
paper
was
eventually
published
in
the
Transactions
of
the
Geological
Society,
dated
1829
but
not
actually
published
until
1835.
The
rst
word
in
this
account
was
his
newly
coined
term
coprolites
and
he
added
further
detail
to
the
remarks
given
in
the
preliminary
accounts
of
his
lectures
which
had
appeared
in
the
Proceedings.
Firstly,
he
records
the
presence
of
these
structures,
which
he
termed
faecal
balls
of
digested
bone,
in
the
Rhaetian
bone
beds
at
Westbury
Garden
Cliff,
Aust
and
Watchet,
remarking
that:
Mr.
Conybeare
and
myself
have
described
these
Coprolites
as
irregular
bodies
of
various
form,
usually
cylindrical,
with
rounded
ends,
some
having
a
black
and
glossy
surface
and
fracture,
others
being
of
a
dull
brown
colour;
and
have
conjectured
them
to
be
rolled
palates,
or
rolled
fragments
of
very
solid
bone:
at
that
time
no
one
suspected
that
they
were
bone
reduced
to
the
state
of
faeces.
(Buckland,
1835,
p.
227)
Buckland
also
noted
that
Mr.
Dillwyn
has
applied
to
them
the
name
of
nigrum
graecum,
from
their
resemblance
in
form
to
the
album
graecum
of
the
cave
of
Kirkdale
(Buckland
1835,
p.
227).
The
Mr
Dillwyn
here
is
Lewis
Weston
Dillwyn
(17781855),
a
porcelain
manufacturer
who,
at
various
times,
served
as
High
Sherriff
of
Glamorgan,
MP
for
Glamorganshire
and
Mayor
of
Swansea.
He
was
also
a
renowned
and
published
naturalist,
and
one
of
his
sons
married
Henry
de
la
Beches
daughter
in
1838.
Buckland
described
Rhaetian
coprolite
specimens
from
the
collection
of
J.S.
Miller
in
typically
economical
and
clear
style,
stating
that,
in
comparison
to
the
Lyme
Regis
material,
they
were:
much
smaller,
and
differ
in
the
absence
of
spiral
structure,
and
the
rare
occurrence
of
scales
or
bones
in
them.
Externally
they
are
of
a
bright
glossy
black,
internally
of
a
dark
brown
colour;
their
substance
is
compact,
their
fracture
splintery,
and
sometimes
conchoidal;
their
surface
often
smooth
as
if
they
had
been
polished.
They
vary
in
size
from
that
of
a
small
potatoe
to
a
hemp
seed:
in
shape,
many
of
them
resemble
the
subangular
concretions
found
in
the
human
gall-bladder,
and
in
the
cavities
of
a
diseased
kidney;
others
are
spherical,
like
sheep's
dung,
or
cylindrical,
like
that
of
rats
and
mice,
with
various
intermediate
varieties
of
size
and
form;
some
are
at
like
a
bean,
others
polygonal.
(Buckland,
1835,
p.
228)
He
even
speculated
about
the
producers
of
these
Rhaetian
coprolites,
although
his
account
pre-dates
the
work
of
Louis
Agassiz
who
later
named
many
of
the
Rhaetian
sh
species
(see
Cross
et
al.,
2018):
There
is
no
direct
evidence
to
show
from
what
animals
the
smaller
varieties
of
these
Coprolites
have
been
derived.
Many
may
probably
be
referred
to
the
small
reptiles,
and
others
to
the
shes,
whose
broken
and
scattered
bones,
teeth,
palates,
and
spines,
are
so
frequent
in
the
same
breccia
with
themselves:
others
may
possibly
be
derived
from
the
inhabitants
of
the
Nautili,
Ammonites,
Belemnites,
and
other
Cephalopodes
which
abounded
at
the
period
of
the
lias
formation.
(Buckland
1835,
p.
228)
Briey
referring
to
records
of
coprolites
from
numerous
other
Rhaetian
localities,
and
guring
17
specimens
(reproduced
here
in
Fig.
1B),
Buckland
marvels
at
the
geographical
extent
of
the
Rhaetian
bone
bed,
sometimes
comprising
25
%
by
volume
of
coprolites,
and
accounting
for
the
richness
of
the
deposit
by
long
periods
of
slow
or
non-deposition
in
an
area
which
must
for
a
long
time
have
been
the
bottom
of
an
ancient
sea,
and
the
receptacle
of
the
faeces
and
bones
of
its
inhabitants,
the
cloaca
maxima,
as
it
were,
of
primaeval
Gloucestershire
(Buckland,
1835,
p.
229).
The
comment
regarding
the
cloaca
maxima
is
an
allusion
to
the
main
sewer
in
ancient
Rome,
constructed
around
600
BCE,
and
used
to
carry
the
citizens
efuent
to
the
River
Tiber.
Coprolites
were
mentioned
in
numerous
papers
listing
the
faunal
components
of
the
Rhaetian
bone
bed
and
Penarth
Group
sediments
in
general
from
a
wide
range
of
localities
in
subsequent
literature.
No
serious
attempt
was
made
to
characterise
the
Rhaetian
coprofauna
in
any
way
until
Dufn
(1979),
who
focused
on
specimens
from
the
conglomeratic
bone
bed
at
Aust
Cliff.
Dufn
recognised
four
broad
morphotypes
and
discussed
the
identities
of
potential
producers.
In
brief,
he
identied:
Type
1:
Large
(up
to
80
mm
long),
brown,
often
tapered,
amphipolar
spiral
coprolites
with
well-dened
internal
struc-
ture
and
visible
inclusions
of
sh
remains.
Producers
probably
included
hybodontiform
and
neoselachian
sharks,
and
perhaps
palaeoniscid
chondrosteans.
Type
2:
Light
brown
to
black,
elongate
amphipolar
and
occasional
heteropolar
spiral
coprolites
with
relatively
poorly
dened
internal
structure,
measuring
up
to
30
mm
in
length
and
lacking
inclusions.
Producers
may
have
included
the
dipnoan
Ceratodus.
Type
3:
Capsule-shaped,
non-spiral
coprolites
lacking
inclu-
sions
and
measuring
up
to
30
mm
long.
Type
4:
Flattened
non-spiral
coprolites
lacking
inclusions
and
measuring
up
to
30
mm
across.
The
only
British
Rhaetian
coprolite
to
be
designated
by
a
formal
taxonomic
ichnological
binomen
is
Strabelocopros
pollardi
Hunt,
Lucas
and
Spielmann,
2012.
The
holotype
(OXFUM
J23743;
Hunt
et
al.,
2012,
Fig.
4AD;
Fig.
1C,
D)
comes
from
an
unrecorded
level
in
the
Watchet
succession,
which
also
preserves
Lower
Jurassic
rocks.
These
authors
also
gure
an
isolated,
un-named
coprolite
from
the
Buckland
Collection
at
the
OXFUM
(Hunt
et
al.,
2012,
Fig.1D).
Further
information
on
Bucklands
studies
of
Rhaetian-age
coprolites,
and
illustrations
of
more
OXFUM
specimens
are
provided
by
Hunt
et
al.
(2013,
Fig.
12).
Recent
studies
by
members
of
the
Palaeobiology
Research
Group
at
Bristol
University
have
concentrated
on
the
Rhaetian
microvertebrate
faunas
from
a
wide
range
of
localities
in
the
West
of
England,
and
the
coprolite
components
of
the
various
faunas
have
been
noted
(e.g.
Korneisel
et
al.,
2015;
Nordén
et
al.,
2015;
Allard
et
al.,
2015;
Lakin
et
al.,
2016;
Slater
et
al.,
2016;
Mears
et
al.,
2016;
Landon
et
al.,
2017;
Cavicchini
et
al.,
2018;
Cross
et
al.,
2018;
Ronan
et
al.,
2020).
One
locality,
Hampstead
Farm
Quarry,
near
Chipping
Sodbury
in
Gloucestershire,
has
yielded
a
numerically
abundant
coprofauna.
The
specimens
have
the
advantage
over
coprolites
from
the
Aust
Cliff
bone
bed
in
that
they
have
all
been
M.
Cueille
et
al.
/
Proceedings
of
the
Geologists
Association
xxx
(2020)
xxxxxx
3
G
Model
PGEOLA-855;
No.
of
Pages
23
Please
cite
this
article
in
press
as:
M.
Cueille,
et
al.,
Fish
and
crab
coprolites
from
the
latest
Triassic
of
the
UK:
From
Buckland
to
the
Mesozoic
Marine
Revolution,
Proc.
Geol.
Assoc.
(2020),
https://doi.org/10.1016/j.pgeola.2020.07.011
Fig.
2.
Summary
scheme
of
the
12
coprolite
morphotypes
from
the
Rhaetian
bone
beds
at
Hampstead
Farm
Quarry,
Gloucestershire.
There
are
ve
sh
coprolite
morphotypes
(AE),
some
of
these
subdivided
into
sub-morphotypes
(A1A4,
B1,
B2),
and
a
broad
category
of
crustacean
coprolites.
Summary
of
morphological
characters,
size,
inclusions,
numbers,
and
sketch
outlines
in
transverse
and
lateral
views.
4
M.
Cueille
et
al.
/
Proceedings
of
the
Geologists
Association
xxx
(2020)
xxxxxx
G
Model
PGEOLA-855;
No.
of
Pages
23
Please
cite
this
article
in
press
as:
M.
Cueille,
et
al.,
Fish
and
crab
coprolites
from
the
latest
Triassic
of
the
UK:
From
Buckland
to
the
Mesozoic
Marine
Revolution,
Proc.
Geol.
Assoc.
(2020),
https://doi.org/10.1016/j.pgeola.2020.07.011
isolated
from
the
surrounding
sediment.
This
coprofauna
can
be
used
to
update
the
classication
and
discussion
given
by
Dufn
(1979)
and
Swift
and
Dufn
(1999),
be
subject
to
modern
non-
destructive
imaging
techniques
and
form
the
basis
for
simple
statistical
analysis
with
a
view
to
elucidating
more
of
the
palaeoecological
factors
at
play
during
the
Late
Triassic.
3.
Materials
and
methods
The
studied
coprolites
come
from
two
Rhaetian
bone
bed
localities
near
Bristol,
Hampstead
Farm
Quarry
(HFQ)
and
Aust
Cliff.
The
specimens
from
Hampstead
Farm
(National
Grid
Reference:
ST
726840)
were
collected
and
donated
by
Mike
Curtis
(19502008),
who
was
in
charge
of
the
Chipping
Sodbury
quarries
for
around
20
years
(Mears
et
al.,
2016).
Coprolites
were
prepared
out
of
the
sediment
by
Curtis
himself,
using
acid
digestion
with
dilute
acetic
acid,
and
sieving
the
residue
to
425
mm
(Cross
et
al.,
2018).
That
huge
collection
is
now
maintained
between
the
BRSUG
and
BRSMG.
A
few
specimens
from
Aust
Cliff
(ST
566
898)
were
added
to
complete
and
compare
the
two
localities
and
to
help
identify
some
of
the
inclusions
noted
in
the
Hampstead
Farm
quarry
coprolites.
This
second
collection
is
also
stored
in
BRSUG
and
was
collected
in
the
1950s
and
1960s
(Cross
et
al.,
2018).
In
total,
1062
coprolites
were
studied
from
HFQ.
The
Aust
Cliff
specimens
were
not
included
in
the
quantitative
data.
All
measurements
were
taken
with
a
ruler
to
the
nearest
millimetre
by
eye:
maximum
length
and
width
were
recorded
for
all
specimens.
The
length
is
the
distance
between
the
two
ends
of
the
coprolite,
and
the
width
is
the
maximum
diameter.
Other
characteristics
were
recorded:
colour,
surface
sculpture,
morphol-
ogy,
breakage
and
general
condition
for
each
coprolite
from
this
locality.
Special
attention
was
paid
to
inclusions
within
the
coprolites,
whether
identiable
or
not.
An
example
of
each
morphotype
was
photographed,
and
the
digital
images
were
then
processed
using
GIMP
software
to
remove
backgrounds
and
adjust
colour
balance
to
be
as
realistic
as
possible.
We
observed
inclusions
in
many
of
the
coprolites,
and
these
were
identied
by
external
examination
of
the
specimens.
In
addition,
we
CT-scanned
three
macrocoprolites
(BRSMG
Cf15467,
BRSMG
Cf10155
and
BRSMG
Cf10162)
and
a
few
microcoprolites
(BRSUG
MF606i-38)
to
enable
production
of
a
3D
model,
showing
hidden
parts
of
the
inclusions
using
a
non-destructive
method.
We
scanned
the
coprolite
specimens
on
the
Nikon
XT
H,
225
ST
CT
scanner
in
the
Palaeobiology
Research
Group
laboratories
at
the
University
of
Bristol.
Scan
parameters
were
set
at
170
kV,
76
uA,
reection
225,
detector
pixel
size
0.2
mm,
and
3141
projections
with
one
frame
per
projection.
The
data
les
were
then
imported
into
Avizo
(version
8.0,
Visualization
Science
Group),
running
on
HP
Workstations
with
72
GB
of
RAM,
using
Windows
10.
The
study
of
vertebrate
coprolites
received
an
enormous
stimulus
with
the
publication
of
a
dedicated
volume
in
2012,
which
we
follow
(Hunt
and
Lucas,
2012a,
2012b).
Hunt
and
Lucas
(2012b)
developed
a
terminology
for
vertebrate
coprolites
and
related
trace
fossils,
termed
collectively
as
bromalites,
food
items
that
have
entered
the
oral
cavity
or
gastrointestinal
tract
of
an
animal
and
have
been
expelled
(either
orally
or
rectally
and
either
pre-
or
post-mortem)
from
or
retained
within
them
(Hunt
and
Lucas
2012b,
p.
140 ).
Several
sub-types
were
recognised
by
Hunt
and
Lucas
(2012b)
and
are
relevant
to
discussion
here.
Con-
sumulites
embrace
all
trace
fossils
comprising
ingested
food
and
found
within
the
body
cavity.
Classied
according
to
their
position
in
the
body,
these
include
gastrolites
(preserved
in
the
stomach),
cololites
(parts
of
the
gut
posterior
to
the
stomach,
enterospirae
being
found
in
a
spiral
valvular
intestine),
intestinelites
(preserved
in
the
body
cavity)
and
evisceralites
(in-lled
intestinal
material
found
separate
from
and
outside
the
body
cavity).
In
its
strictest
sense,
the
term
coprolite
is
reserved
for
faecal
material
that
has
been
ejected
from
the
posterior
end
of
the
gastro-intestinal
tract,
although
the
word
is
generally
used
informally
for
any
faecal
mass
found
in
the
fossil
record.
4.
Fish
coprolites
4.1.
Classication
Coprolites
can
be
classied
by
the
use
of
informal
terms
(e.g.,
cylindrical
coprolite
A)
or
by
the
application
of
a
Linnaean
parataxonomy,
because
they
are
trace
fossils.
Numerous
sh
coprolite
ichnotaxa
have
been
noted
in
the
Rhaetian-aged
sedi-
ments
of
Europe
and
North
America
(Hunt
and
Lucas,
2012a),
and
yet
it
was
difcult
to
assign
coprolites
in
our
collections
to
these
named
morphotypes.
Therefore,
we
establish
our
own
scheme,
but
do
not
apply
formal
ichnotaxon
names.
We
identify
four
main
categories
of
coprolite
forms
involving
at
least
seven
morphotypes
(some
of
them
with
different
varieties)
in
the
HFQ
collection
(Fig.
2),
whereas
in
the
Hunt
and
Lucas
(2012a)
scheme,
there
are
27
morphotypes
of
coprolites
in
11
main
categories.
We
divide
some
morphotypes
into
varieties
or
sub-
morphotypes
because
of
the
great
diversity
of
shapes:
morphotype
A
comprises
four
sub-morphotypes
and
morphotype
B
comprises
two
sub-morphotypes.
The
average
size
of
HFQ
coprolites
is
6.8
mm
wide
and
11.4
mm
long
(complete
and
incomplete
specimens).
Most
vertebrate
coprolites
are
sub-cylindrical
(Hunt
and
Lucas,
2012a),
and
we
nd
this
at
HFQ,
where
around
70
%
of
coprolites
are
cylindrical
or
ellipsoidal.
The
most
represented
morphotype
at
HFQ
is
A,
with
almost
41
%
of
the
sample.
In
spiral
coprolites,
morphotype
B,
they
may
be
slightly
cone-shaped
or
lens-shaped.
The
rest
are
irregular
in
shape
or
too
incomplete
to
be
sure,
and
they
are
listed
as
unknown
morphotype.
44
%
of
the
coprolites
are
incomplete,
and
20
%
of
the
complete
specimens
are
irregular
in
shape.
Because
of
their
nature
and
preservation,
it
is
difcult
to
link
coprolite
ichnotaxa
with
their
producers
(Luo
et
al.,
2017;
see
Discussion).
There
are
many
larger
coprolites
from
Aust,
but
they
are
almost
all
incomplete
and
any
inclusions
are
very
rare,
making
it
impossible
to
use
them
to
help
reconstruct
a
food
web.
4.1.1.
Morphotype
A
Cylindrical
to
sub-cylindrical
in
shape
and
straight
to
slightly
curved,
and
complete
specimens
are
252
mm
long
(average
812
mm)
and
226
mm
wide
(average
5
mm).
The
external
surface
of
the
coprolite
possesses
occasional
longitudinal
and
transverse
striae,
but
no
spiral
structure
is
visible.
Some
specimens
have
a
phosphatised
skin,
presumed
to
be
a
consequence
of
diagenesis
(Dufn,1979),
and
only
visible
in
transverse
sections
of
incomplete
coprolites.
Relatively
homogenous
overall,
the
coprolitic
matrix
varies
from
coarse-grained,
resulting
in
an
irregular
external
surface,
to
ne
grained,
giving
rise
to
a
smooth
surface.
The
majority
of
morphotype
A
specimens
(89
%)
are
cylindrical
(Fig.
3A)
while
the
remaining
11
%
are
more
rounded
and
sub-
cylindrical
in
shape.
A
wide
range
of
forms
is
present,
but
all
are
cylindrical
in
cross-section,
which
is
the
basis
for
assignment
to
this
category.
60
%
of
the
cylindrical
coprolites
have
fewer
than
ve
inclusions
visible
at
the
surface,
almost
all
of
which
are
scales
(unidentiable
inclusions
are
very
few
in
number).
Only
16
%
have
more
than
ten
inclusions
breaking
the
surface
of
the
coprolite.
The
nature
of
the
surface
is
not
indicative
for
this
morphotype,
with
51
%
of
the
surfaces
being
irregular
and
49
%
smooth.
Most
morphotype
A
coprolites
are
light
in
colour;
many
are
white,
but
light
brown
and
light
grey
specimens
are
also
present.
Only
7%
are
dark
coloured.
There
is
apparently
no
correlation
between
the
M.
Cueille
et
al.
/
Proceedings
of
the
Geologists
Association
xxx
(2020)
xxxxxx
5
G
Model
PGEOLA-855;
No.
of
Pages
23
Please
cite
this
article
in
press
as:
M.
Cueille,
et
al.,
Fish
and
crab
coprolites
from
the
latest
Triassic
of
the
UK:
From
Buckland
to
the
Mesozoic
Marine
Revolution,
Proc.
Geol.
Assoc.
(2020),
https://doi.org/10.1016/j.pgeola.2020.07.011
nature
of
the
surface
and
the
colour
of
the
coprolite.
The
rst
cylindrical
form,
here
designated
sub-morphotype
A1
(Fig.
3B),
is
a
classic
cylinder
with
both
terminations
rounded
and
convex
(isopolar),
and
a
constant
diameter
along
the
full
length
of
the
specimen.
These
specimens
are
quite
rare
in
the
collection
because
only
fairly
complete,
well
preserved
specimens
can
be
assigned
with
condence
to
this
category.
Many
specimens
are
incomplete
being
preserved
as
short
segments
of
coprolite
with
an
ovoid
cross-
section;
that
shape
allows
them
to
be
classied
as
sub-morphotype
A,
but
there
is
the
danger
that
some
might
be
misidentied.
The
second
type,
A2
(Fig.
3C),
is
more
ellipsoidal
in
lateral
view,
but
still
thick,
with
ends
which
are
less
rounded
than
in
A1,
or
slightly
tapered.
Most
of
the
specimens
are
almost
cone-shaped
with
one
tapered
end
(sometimes
strongly
tapered),
while
the
other
end
is
always
rounded
and
convex
(anisopolar).
The
third
type,
A3
(Fig.
3DE),
embraces
attened
coprolites,
but
specimens
which
are
not
at
enough
to
qualify
as
Morphotype
C.
Both
ends
are
usually
rounded
with
an
ellipsoidal
cross-section.
These
coprolites
might
have
been
attened
after
extrusion,
by
compaction
during
the
burial
processes.
Most
of
their
diameters
are
around
half
the
size
of
specimens
belonging
to
morphotypes
A1
and
A2,
and
they
can
be
isopolar
or
anisopolar.
The
remainder
(11
%)
of
Morphotype
A
coprolites
(type
A4)
are
round
to
ovoid
forms
(Fig.
3F).
Complete
specimens
have
lengths
from
3
to
18
mm
and
widths
from
3
to
18
mm.
The
general
shape
is
ovoid,
and
they
have
no
ends.
Some
of
the
specimens
are
slightly
attened.
Many
of
them
are
pellet-shaped
to
spherical
in
form.
Discussion.
The
absence
of
spiral
structures
in
Morphotype
A
coprolites
could
reect
their
true
structure,
or
such
structures
might
have
been
lost
(see
Discussion
of
Morphotype
B).
Inclusions
are
quite
rare
in
these
coprolites;
where
absent,
the
coprolites
could
have
been
produced
by
durophagous
shes
such
as
Lepidotes
or
more
probably
Sargodon
tomicus.
This
is
because
durophagous
shes
generally
ate
only
invertebrates.
Many
small
shes
could
also
be
potential
producers,
such
as
the
prey
of
Gyrolepis
albertii
(Cross
et
al.,
2018).
Morphotype
A
coprolites
with
included
scales
usually
have
just
one
(95
%),
and
62
%
of
those
with
scales
show
fewer
than
ve.
This
contrasts
with
the
spiral,
morphotype
B,
coprolites,
where
inclusions
are
much
more
abundant,
suggesting
probably
differences
in
diets
rather
than
taphonomic
losses
(all
coprolites
are
preserved
together
in
the
same
conditions).
4.1.2.
Morphotype
B
These
coprolites
all
show
spiral
structure,
attesting
to
passage
through
a
spiral
valve
in
the
intestine.
Most
of
the
spiral
coprolites
in
this
study
have
fewer
than
ten
longitudinal
striations
(and
many
have
fewer
than
ve)
on
the
outer
surface
of
the
coprolite.
Some
25
%
of
all
specimens
possess
more
than
ten
scales
breaking
the
coprolite
surface,
the
highest
percentage
for
all
coprolite
morphotypes.
There
are
two
predominant
colours,
grey
and
brown,
with
a
few
white
specimens
and
one
black
(BRSMG
Cf9993).
There
are
two
sub-morphotypes
of
Morphotype
B.
The
most
common
is
variety
B1
(Fig.
4
A,
E),
the
scroll
type
which
has
a
paper-roll
shape,
and
makes
up
28
%
of
Morphotype
B
specimens.
These
coprolites
have
lengths
ranging
from
5
to
22
mm
and
widths
from
2
to
9
mm
(based
on
complete
specimens
only).
Most
of
these
coprolites
are
anisopolar,
with
one
end
very
tapered
and
the
other
slightly
rounded.
Two-thirds
of
the
specimens
have
an
irregular
surface
with
a
rough
texture,
and
the
rest
have
a
smooth
surface.
These
scroll
specimens
are
distinctive,
with
a
visible
last
whorl
which
is
a
large
fold
of
the
coprolite
along
the
length
of
the
entire
specimen,
even
if
no
characteristic
spiral
Fig.
4.
The
spiral
coprolite
morphs.
(A)
Coprolite
morph
B1,
BRSMG
Cf15557,
lateral
view,
(BD)
Coprolite
morph
B2,
BRSMG
Cf9823
(B);
BRSMG
Cf15621
(C);
BRSMG
Cf9848
(D),
all
lateral
views,
(E)
Coprolite
morph
B1,
BRSMG
Cf9833,
transverse
view
of
a
cross
section.
Fig.
3.
The
smooth-surfaced
coprolite
morphs.
(A)
Coprolite
morph
A,
BRSMG
Cf9630,
lateral
view,
(B)
Coprolite
morph
A1,
BRSMG
Cf10055,
lateral
view,
(C)
Coprolite
morph
A2,
BRSMG
Cf10158,
lateral
view,
(D
and
E)
Coprolite
morph
A3,
BRSMG
Cf9819,
in
lateral
(D)
and
longitudinal
(E)
views,
(F)
Coprolite
morph
A4,
BRSMG
Cf15516,
transverse
view.
6
M.
Cueille
et
al.
/
Proceedings
of
the
Geologists
Association
xxx
(2020)
xxxxxx
G
Model
PGEOLA-855;
No.
of
Pages
23
Please
cite
this
article
in
press
as:
M.
Cueille,
et
al.,
Fish
and
crab
coprolites
from
the
latest
Triassic
of
the
UK:
From
Buckland
to
the
Mesozoic
Marine
Revolution,
Proc.
Geol.
Assoc.
(2020),
https://doi.org/10.1016/j.pgeola.2020.07.011
striations
are
visible.
Two
specimens
have
a
visible
spiral
structure
in
cross-section
(BRSMG
Cf9833
and
BRSMG
Cf10326;
Fig.
4E).
The
overall
shape
is
cylindrical,
and
slightly
ovoid
for
the
smaller
ones.
Some
have
both
ends
tapered,
giving
a
lensoid
shape
overall.
Inclusions
consist
of
scales
only
and
30
%
lack
visible
inclusions
altogether.
61
%
of
B1
coprolites
have
more
than
ve
scales
and
70
%
have
at
least
one
inclusion,
the
highest
proportions
in
all
the
coprolite
morphotypes
in
the
collection.
Variety
B2
is
the
second
morphotype
in
order
of
importance,
making
up
10
%
of
Morphotype
B
specimens
(Fig.
4BD).
These
coprolites
are
amphipolar,
with
both
ends
of
similar
shape
(isopolar)
and
spiral
striations
are
visible
from
one
end
to
the
other
of
the
specimen.
This
is
quite
similar
to
specimens
of
Hyronocopros
amphipolar
described
by
Hunt
et
al.
(2007).
The
spiral
form
is
well
dened,
but
never
perfectly
concentric
(Dufn,
1979 ).
Some
coprolites
are
incomplete,
but
broken
surfaces
reveal
the
entire
spiral,
which
is
useful
when
the
exterior
spiral
striations
are
not
clearly
visible.
Some
spiral
coprolites
cannot
be
classied
as
either
B1
or
B2
because
of
poor
preservation.
They
may
have
only
one
to
three
visible
spiral
turns,
and
the
spiral
conguration
cannot
be
dened
precisely.
These
are
the
most
common
spiral
specimens
in
the
HFQ
collection,
comprising
61
%
of
morphotype
B
specimens.
Discussion.
The
HFQ
specimens
of
morphotype
B1
are
similar
to
those
described
as
Morphotype
F1
by
Hunt
and
Lucas
(2012a),
and
HFQ
Morphotype
B2
specimens
are
similar
to
their
Morphotype
F2.
Our
examples
show
the
highest
numbers
of
inclusions
of
all
the
HFQ
coprolite
morphotypes,
and
especially
the
Scroll
coprolites
show
morethan
ve
or
ten
scales
as
inclusions,
whichsuggeststhey should
be
assigned
to
a
larger
predator
such
as
large
bony
shes
or
sharks.
Fossilised
faecal
masses
showing
spiral
form
are
cololites
(enterospirae)
when
found
in
the
body
cavity.
These
are
fairly
rare
in
the
fossil
record
and
tend
to
be
found
in
faunas
where
special
conditions
of
fossilisation
occur,
such
as
various
Conservation
Lagerstätten,
as
in
specimens
of
Cladoselache
clarki
from
the
Devonian
Cleveland
Shale
of
Ohio
and
a
number
of
xenacanth
sharks
from
the
Permian
of
the
Czech
Republic
(Woodward,
1917;
Williams,
1972 ).
Isolated
specimens,
such
as
those
here,
could
be
either
coprolites
sensu
stricto
(ejected
from
the
posterior
part
of
the
gut),
or
evisceralites,
intestinal
spiral
valve
lls
never
ejected
as
coprolites
but
separated
and
removed
from
their
original
site
of
residence
in
the
body
cavity,
perhaps
following
scavenging
or
decomposition
of
the
host
carcass
(Dufn
and
Ward,
2020).
The
holotype
of
Strabelocopros
pollardi
from
the
Rhaetian
of
Watchet,
Somerset
(Fig.
1C,
D)
is
most
likely
an
example
of
an
evisceralite.
Spiral
coprolites
were
rst
designated
by
Buckland
and
their
origins
(coprolite
versus
enterospirae,
and
the
nature
of
the
source
animal)
have
been
actively
debated
(e.g.
Buckland,
1835,
1836;
Ammon,
1889;
Neumayer,
1904;
Woodward,
1917;
Williams,
1972;
Dufn,
1979;
McAllister,
1985,
1988,
1996,
2003;
Hunt
et
al.,
2015).
Much
of
what
we
know
about
the
morphology
and
variation
in
intestinal
spiral
valves
relies
on
the
original
work
of
Thomas
Jeffrey
Parker
(18501897)
who
recognised
four
types
of
spiral
valve
or
valvula
spiralis
in
which
coiling
takes
place
around
a
longitudinal
axis,
and
the
internal
structure
comprises
a
series
of
stacked,
spiralling
cones
whose
apices
may
be
directed
either
anteriorly
or
posteriorly
(Parker,
1885;
Jain,
1983;
McAllister,
1985,
1996;
Hunt
and
Lucas,
2012b).
In
addition,
the
valvula
voluta
or
scroll
valve
is
wound
a
little
like
a
roll
of
paper.
Valvular
intestines
are
indicated
for
some
or
all
agnathans,
placoderms,
possibly
some
acanthodians
and
all
chondrichthyans
(McAllister,
1996;
Hunt
and
Lucas,
2012b;
Bajdek
et
al.,
2019).
Among
bony
shes,
spiral
valves
are
known
in
sarcopterygians
(lobe-nned
shes)
and
have
been
described
for
extant
coela-
canths
and
lungsh
(Millot
et
al.,1978;
Hassanpour
and
Joss,
2009).
They
are
also
reported
in
non-teleostean
primitive
actinopterygians,
in
particular
polypterids
(bichirs
and
reedsh),
acipenseriforms
(sturgeons
and
paddleshes),
holosteans
(includ-
ing
gars,
bowns
and
caturids)
and
pachycormids
(Arratia
and
Schultze,
2013;
Cataldi
et
al.,
2002).
There
is
circumstantial
evidence
also
for
a
spiral
valve
in
palaeoniscid
chondrosteans
(Price,
1927;
Dufn,
1979);
direct
evidence
of
a
spiral
valve
is
known
from
certain
fossil
sturgeons
(Peipiaosteus
pani
from
the
Early
Cretaceous
of
Liaoning,
China;
Capasso,
2019)
and
pachy-
cormids
(Arratia
and
Schultze,
2013).
However,
spiral
coprolites
are
not
seen
in
teleost
shes
or
tetrapods
(Bajdek
et
al.,
2019).
The
spiral
valve
is
explained
as
an
adaptation
to
life
in
variable
and
hostile
environments
(Capasso,
2019,
p.
23).
It
is
a
means
to
increase
the
surface
area
of
the
intestinal
wall
without
increasing
its
length,
thereby
optimising
nutrient
absorption
and
conserving
total
gut
volume.
It
is
also
associated
with
slow
transit
of
food
through
the
gut,
another
adaptation
to
maximise
nutrient
uptake.
The
number
of
turns
of
the
spiral
valve
may
relate
to
diet.
From
the
above
discussion,
it
is
clear
that
many
groups
of
marine
vertebrates
could
have
contributed
spiral
coprolites
in
the
Rh