RESEA
RCH
ARTICL
E
Late
middle
Miocene
caviomorph
rodents
from
Tarapoto,
Peruvian
Amazonia
Myriam
Boivin
ID
1
*
, Laurent
Marivaux
2
, Walter
Aguirre-Diaz
3
, Aldo
Benites-Palomino
ID
3,4
,
Guillaume
Billet
5
, Franc
̧ ois
Pujos
ID
6
, Rodolfo
Salas-Gismondi
3,7
, Narla
S.
Stutz
ID
2,8
, Julia
V.
Tejada-Lara
2,3
, Rafael
M.
Varas-Malca
ID
3
, Anne
H.
Walton
9
, Pierre-Olivier
Antoine
2
1
Laboratori
o de
Paleontolog
ı
́
a de
Vertebrad
os,
Instituto
de
Ecorregion
es
Andinas
(INECOA
), Universidad
Nacional
de
Jujuy,
CONICE
T, San
Salvado
r de
Jujuy,
Jujuy,
Argentina,
2
Laboratoir
e de
Pale
́
ontolog
ie,
Institut
des
Sciences
de
l’Evolution
de
Montpelli
er (ISEM,
UMR
5554,
CNRS/UM
/IRD/EPHE
), Universite
́
de
Montpelli
er,
Montpellier,
France,
3
Departamen
to de
Paleontolog
ı
́
a de
Vertebrad
os,
Museo
de
Historia
Natural—
Universidad
Nacional
Mayor
San
Marcos
(UNMSM,
DPV-MUSM
), Lima,
Peru,
4
Pala
̈
ontologis
ches
Institut
und
Museum
, Universita
̈
t Zu ̈ rich,
Zu ̈ rich,
Switzerlan
d,
5
De
́
partemen
t Origine
et Evolution,
Muse
́
um
nationa
l d’Histoire
naturelle
(MNHN,
CR2P
-UMR
7207
CNRS/M
NHN/Sorbonn
e Universite
́
), Paris,
France,
6
Laboratori
o de
Paleontolog
ı
́
a, Instituto
Argentino
de
Nivologı
́
a, Glaciologı
́
a y Ciencias
Ambien
tales
(IANIGL
A),
CCT–CONIC
ET–Mend
oza,
Mendoza,
Argentina,
7
BioGeoCiencia
s Lab,
Facultad
de
Ciencias
y
Filosofı
́
a/CIDIS,
Laboratori
os
de
Investigac
ion
y Desarroll
o (LID),
Centro
de
Investigacio
́
n para
el Desarroll
o
Integra
l y Sostenible
(CIDIS),
Univers
idad
Peruana
Cayetano
Heredia,
Lima,
Peru,
8
Programa
de
Po
́
s-
Graduac
̧
ã
o
em
Geoci
ê
ncia
s, Univers
idade
Federal
do
Rio
Grande
do
Sul,
Porto
Alegre,
Rio
Grande
do
Sul,
Brazil,
9
Springfield
Technical
Commun
ity
College,
Springfield
, Massach
usset,
United
States
of America
*
mboivin@
idgym.unju
.edu.ar
Abstract
Miocene
deposits
of South
America
have
yielded
several
species-rich
assemblages
of cavio-
morph
rodents.
They
are
mostly
situated
at high
and
mid-
latitudes
of the
continent,
except
for
the
exceptional
Honda
Group
of La
Venta,
Colombia,
the
faunal
composition
of which
allowed
to describe
the
late
middle
Miocene
Laventan
South
American
Land
Mammal
Age
(SALMA).
In this
paper,
we
describe
a new
caviomorph
assemblage
from
TAR-31
locality,
recently
dis-
covered
near
Tarapoto
in Peruvian
Amazonia
(San
Martı
́
n Department).
Based
on
mammalia
n
biostratigraphy,
this
single-phased
locality
is unambiguou
sly
considered
to fall
within
the
Laven-
tan
SALMA.
TAR-31
yielded
rodent
species
found
in La
Venta,
such
as
the
octodontoid
Ricard-
omys
longidens
Walton,
1990
(
nom
.
nud
.),
the
chinchilloids
Microscleromys
paradoxalis
Walton,
1990
(
nom
.
nud
.)
and
M
.
cribriphilus
Walton,
1990
(
nom
.
nud
.),
or closely-related
taxa.
Given
these
strong
taxonomic
affinities,
we
further
seize
the
opportunity
to review
the
rodent
dental
material
from
La
Venta
described
in the
Ph.D.
volume
of Walton
in 1990
but
referred
to
as
nomina
nuda
.
Here
we
validate
the
recognition
of these
former
taxa
and
provide
their
formal
description.
TAR-31
documents
nine
distinct
rodent
species
documenting
the
four
extant
superfamilies
of Caviomorpha,
including
a new
erethizontoid:
Nuyuyomys
chinqaska
gen.
et
sp.
nov.
These
fossils
document
the
most
diverse
caviomorph
fauna
for
the
middle
Miocene
interval
of Peruvian
Amazonia
to date.
This
rodent
discovery
from
Peru
extends
the
geographi
-
cal
ranges
of
Ricardomys
longidens
,
Microscleromys
paradoxalis
,
and
M
.
cribriphilus
,
1,100
km
to the
south.
Only
one
postcranial
element
of rodent
was
unearthed
in TAR-31
(astragalus).
This
tiny
tarsal
bone
most
likely
documents
one
of the
two
species
of
Microscleromys
and
its
morphology
indicates
terrestrial
generalist
adaptations
for
this
minute
chinchilloid.
PLOS ONE
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
1 / 80
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN
ACCESS
Citation:
Boivin
M, Marivaux
L, Aguirre-D
iaz W,
Benites-Palom
ino A, Billet G, Pujos
F, et al. (2021)
Late middle
Miocene
caviomorph
rodents
from
Tarapoto,
Peruvian
Amazonia.
PLoS
ONE 16(11):
e0258455.
https://d
oi.org/10.1371/j
ournal.
pone.025845
5
Editor:
Ju ̈rgen Kriwet,
University
of Vienna,
AUSTRIA
Received:
July 19, 2021
Accepted:
September
24, 2021
Published:
November
3, 2021
Peer Review
History:
PLOS
recognize
s the
benefits
of transpar
ency
in the peer review
process;
therefore,
we enable
the publication
of
all of the content
of peer review
and author
response
s alongside
final,
published
articles.
The
editorial
history
of this article
is available
here:
https://doi.o
rg/10.1371/jo
urnal.pone.0
258455
Copyright:
©
2021
Boivin
et al. This is an open
access
article
distributed
under
the terms
of the
Creative
Commons
Attribution
License,
which
permits
unrestricte
d use, distribu
tion, and
reproduction
in any medium,
provided
the original
author
and source
are credited.
Data
Availabilit
y Statement:
All relevant
data are
within
the manuscript
and its Support
ing
Information
files (S1 Table
and S2 File). The fossil
material
is permanently
stored
in the Vertebrate
Introduction
Caviomorphs
(Caviomorpha
Wood,
1955)
are
hystricognathous
rodents
originating
from
South
America,
such
as
guinea
pig,
New
World
porcupines,
chinchillas,
and
spiny
rats.
Nowa-
days,
they
show
a great
taxonomical
diversity
and
a large
array
of
lifestyles
(arboreal,
fossorial,
semi-aquatic,
and
terrestrial)
[1].
Among
caviomorphs,
two
main
clades,
each
divided
into
two
superfamilies,
are
recognised
[2–4]:
Erethicavioi
Boivin,
2019
(comprising
Cavioidea
Fischer,
1817
+ Erethizontoidea
Bonaparte,
1845
+ stem-groups)
and
Octochinchilloi
Boivin,
2019
(comprising
Chinchilloidea
Bennett,
1833
+ Octodontoidea
Waterhouse,
1839
+ stem-
groups).
Their
evolutionary
history
extends
back
to
the
late
middle
Eocene,
with
the
earliest
representatives
of
the
group
so
far
recorded
in
Peruvian
Amazonia
[5–8].
The
Miocene
record
of
caviomorphs
is particularly
well
documented.
Numerous
fossil-
bearing
localities
yielded
abundant
specimens
and
species-rich
assemblages
of
rodents
[9].
Most
of
them
are
situated
at
high
and
middle
latitudes,
notably
in
Argentina
(e.g.,
Pinturas,
Sarmiento,
Santa
Cruz,
Collo
́
n Cura,
Cerro
Azul
and
Ituzaingo
formations)
[9–15]
and
Chile
(e.g.,
Laguna
del
Laja
and
Pampa
Castillo)
[16,
17].
At
low
latitudes
of
South
America,
two
remarkable
and
very
species-rich
caviomorph
faunas
are
known,
La
Venta
in
Colombia
(Honda
Group;
late
middle
Miocene)
[18]
and
Acre
in
Brazil
(Solim
õ
es
Fm.
[Formation];
late
Miocene;
see
[19]).
Since
the
early
discoveries
of
fossil-bearing
localities
in
the
Honda
Group
of
the
upper
Magdalena
valley
in
1923,
considerable
field
efforts
led
by
petroleum
companies
and
several
scientific
teams
were
performed
in
this
region
of
Colombia,
and
notably
in
the
La
Venta
area,
until
1992
[20,
21].
As
a result,
140
localities
over
52
different
stratigraphic
levels,
yielding
thousands
of
fossils
of
many
vertebrate
groups
were
discovered
in
the
Honda
Group
dating
from
the
late
middle
Miocene
[22].
In
particular,
the
recognition
in
1945
of
the
first
fos-
sil
remains
of
primates
in
the
Magdalena
valley
[23,
24]
has
been
one
of
the
most
significant
discoveries
that
motivated
numerous
subsequent
field
expeditions
[20].
The
mammal
fauna
from
La
Venta
encompasses
at
least
77
species
and
80
genera,
representing
29
families
[25,
26].
Concerning
rodents,
around
20
taxa
documenting
the
four
main
caviomorph
superfamilies
have
so
far
been
reported
(excluding
the
indeterminate
genera
and
species)
[18,
27–30].
In
addition
to
La
Venta,
the
Peruvian
faunas
from
MD-67
(Madre
de
Dios)
[31],
the
Fitzcarrald
local
fauna
(near
Atalaya,
Ucayali)
[32,
33],
and
CTA-45
(near
Contamana,
Loreto)
[34],
also
middle
Miocene
in
age,
document
representatives
of
the
four
caviomorph
superfamilies
(4,
8,
and
2 taxa,
respectively).
Finally,
a fourth
late
middle
Miocene
caviomorph-bearing
locality
also
occurs
in
low
latitudes,
from
the
Socorro
Formation
in
Venezuela,
but
so
far
it has
only
yielded
a fragmented
distal
femur
of
a giant
caviomorph
[35].
Over
the
last
decade,
our
yearly
paleontological
surveys
in
the
Tarapoto
area
(San
Martı
́
n
Department,
Peruvian
Amazonia),
have
allowed
the
discovery
of
a new
fossil-bearing
locality,
TAR-31,
in
the
vicinity
of
the
Juan
Guerra
village.
This
locality
is considered
to
be
late
middle
Miocene
in
age
(Laventan
SALMA)
based
on
mammalian
biostratigraphy
(see
[36];
see
Mate-
rial
and
Methods
section
of
this
paper).
TAR-31
yielded
a diverse
assemblage
of
aquatic
and
terrestrial
vertebrates,
including
numerous
caviomorph
specimens
that
are
the
subject
of
this
paper.
The
primary
purpose
of
the
present
work
is to
describe
the
new
caviomorph
materials
found
in
TAR-31.
This
locality
shares
taxa
or
closely
related
taxa
with
La
Venta
fauna
described
by
one
of
us
(AHW)
in
the
framework
of
her
PhD
[37]:
Microsteiromys
,
Ricardomys
,
and
Microscleromys
.
Because
Microsteiromys
,
Ricardomys
,
Microscleromys
,
and
the
species
included
within
these
genera
were
not
formally
described
(
nomina
nuda
)
[18,
37,
38],
here
we
also
revise
the
material
from
La
Venta
attributed
to
these
taxa,
in
order
to
clarify
their
systematics.
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
2 / 80
Paleontologi
cal collection
of the “Museo
de Historia
Natural,
Universida
d Nacional
Mayor
de San
Marcos”
(MUSM)
, Lima,
Peru.
Funding:
LM and POA received
funding
from
The
Leakey
Foundation
and the LabEx
CEBA
(ANR-10-
LABX-0025-
01). POA received
funding
from
the
National
Geograph
ic Society
and from
the French
‘Agence
Nationale
de la Recherche
’ (ANR)
in the
framework
of the GAARAnti
program
(ANR-17-
CE31-000
9). LM received
funding
from
the
CoopIntEER
CNRS-CO
NICET
(n ̊ 252540).
POA and
FP received
funding
from
the ECOS-SU
D/FONCyT
(n ̊ A-14U01
) internationa
l collabora
tion programs.
MB received
funding
from
the ‘Laborato
ire de
Plane
́
tologie
et de Ge
́
odynamiqu
e de Nantes‘.
The
funders
had no role in study
design,
data collection
and analysis,
decision
to publish,
or preparation
of
the manuscript.
Competing
interests
:
The authors
have declared
that no competing
interests
exist.
Material
and
methods
TAR-31,
Tarapoto,
Peru
The
rodent
fossil
material
from
Peruvian
Amazonia
described
in
the
present
work
originates
from
the
TAR-31
locality,
situated
in
the
vicinity
of
the
small
town
Juan
Guerra,
along
the
Mayo
River
(Tarapoto
area,
San
Martı
́
n Department,
Peru;
Fig
1A
and
1B).
TAR-31
consists
of
a 10–15
cm-thick
yellow
microconglomerat
e interbedded
within
a grey
cross-stratified
and
sandstone-dominated
fluvial
unit
(Fig
2A
[36]).
The
latter
is intercalated
between
thick
violin-
grey
variegated
paleosols
pointing
to
the
existence
of
a meandering
river
with
sustainable
floodplain
[36].
These
levels
were
originally
mapped
as
belonging
to
the
lower
member
of
Ipururo
Formation
(middle
Miocene
in
age
in
the
Huallaga
basin)
[39–42],
an
assignation
in
agreement
with
the
nature
of
their
facies
and
depositional
environment
sequence.
The
fossil
content
of
the
TAR-31
encompasses
plants,
amber
clasts,
crabs,
chondrichthyans,
osteichthyans,
anurans,
chelonians,
crocodylomorphs,
birds,
and
mammals
(including
metatherians,
xenarthrans,
liptoterns,
notungulates,
sirenians,
chiropterans,
primates,
and
caviomorph
rodents).
The
TAR-31
mammal
assemblage
includes
a didelphid
marsupial
(cur-
rently
under
study
by
one
of
us,
NSS),
the
interatheriine
notoungulate
Miocochilius
sp.,
the
didolodontid
Megadolodus
sp.,
the
platyrrhine
primate
Neosaimiri
aff.
fieldsi
[36],
and
the
caviomorph
fauna
described
here
(see
‘Age
of
TAR-31’
section
of
this
present
work).
This
assemblage
recalls
some
of
the
Laventan
SALMA
localities:
Quebrada
Honda
in
Bolivia
(13.1–
12.2
Ma)
[43–46],
the
Fitzcarrald
local
fauna
in
SE
Peru
[33],
and
especially
the
lower
part
of
the
Villavieja
Formation
in
the
La
Venta
area
in
Colombia
(13.8–11.6
Ma;
[47]
and
see
above).
Accordingly,
the
TAR-31
locality
most
likely
documents
the
late
middle
Miocene
Laventan
SALMA
[36].
The
material
from
TAR-31
was
collected
by
excavating
and
wet
screening
(2
and
1 mm
meshes)
about
550
kg
of
sediment
during
our
yearly
field
expeditions
(August
2015,
2016,
and
Fig
1.
Geographi
c location
of
TAR-31
(San
Martı
́
n Departm
ent,
Peru;
late
middle
Miocene,
Laventa
n SALMA)
.
(A)
Location
map
of
the
Tarapoto
area
(star
symbol)
in
Peru.
(B)
Location
map
of
the
TAR-31
(star
symbol)
in
the
Tarapoto
area.
SALMA,
South
American
Land
Mammal
Age.
https://d
oi.org/10.1371/j
ournal.pon
e.0258455.g0
01
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
3 / 80
2018).
The
caviomorph
material
corresponds
to
about
400
isolated
teeth
(complete
or
frag-
mentary),
one
fragment
of
mandible
(MUSM
4643),
one
fragment
of
maxilla
(MUSM
4375),
and
one
astragalus
(MUSM
4658).
No
permits
were
required
for
the
described
study,
which
complied
with
all
relevant
regulations.
The
field
work
on
TAR-31
was
carried
out
in
the
frame-
work
of
the
ongoing
cooperation
agreement
between
the
‘Museo
de
Historia
Natural
de
la
Universidad
Nacional
Mayor
San
Marcos’
(Lima,
Peru)
and
the
‘Institut
des
Sciences
de
l’Evo-
lution
de
Montpellier-Universite
́
de
Montpellier’
(France).
The
TAR-31
fossil
specimens
are
permanently
stored
in
the
palaeontological
collection
of
the
‘Museo
de
Historia
Natural,
Uni-
versidad
Nacional
Mayor
de
San
Marcos’,
Lima,
Peru
(MUSM).
La
Venta
area,
Colombia
The
fossil
rodent
material
from
La
Venta
re-analysed
here
(originally
studied
in
the
framework
of
the
Ph.D.
of
Anne
H.
Walton)
[18,
37]
originates
from
the
badland
exposures
located
in
the
vicinities
of
the
Villavieja
and
La
Victoria
villages,
in
the
upper
valley
of
the
Magdalena
River
(Huila
Department,
Colombia).
Numerous
fossil-bearing
localities
are
known
in
the
La
Venta
area
(and
the
Magdalena
upper
valley
in
general)
[26].
These
are
assigned
to
two
successive
formations
of
the
Honda
Group:
La
Victoria
and
Villavieja
formations
[48].
The
studied
mate-
rial
from
this
area
was
found
in:
(i)
one
locality
from
La
Victoria
Formation
(IGM-DU
loc.
075,
between
the
Chunchullo
Beds
and
the
Tatcoa
Sandstone
Beds),
and
(ii)
distinct
localities
from
the
Villavieja
Formation,
including
several
localities
in
the
Fish
Bed
(CVP
5,
8,
9,
10,
10A,
and
13B),
two
localities
in
the
Monkey
Bed
(IGM-DU
loc.
006–1
and
IGM-DU
loc.
022
or
UCMP
loc.
V4536),
and
one
in
the
El
Cardo
́
n Red
Beds
(IGM-DU
loc.
032
and
its
screen-
washed
locality
032–5).
The
Unit
between
the
Chunchullo
and
Tatacoa
Sandstone
Beds
is
149.3
m-thick,
and
is composed
of
very
fine
sandstones
and
mudstones
alternating
with
metric
medium-grained
sandstone
lenses
[48].
The
Villavieja
Formation
is divided
into
the
Baraya
Member,
including
the
Fish
and
Monkey
beds,
and
the
Cerro
Colorado
Member,
including
the
El
Cardo
́
n Red
Beds
[48].
The
Baraya
Member
is very
fossiliferous
and
is composed
mainly
of
gray
mudstones
and
sandstones
with
minor
layers
of
red
mudstones,
while
the
Cerro
Colo-
rado
Member
is less
fossiliferous
and
consists
of
thick
horizons
of
red
mudstones
with
a very
small
amount
of
volcanic
litharenites
and
chert
litharenites
[48].
The
La
Venta
fauna
is late
middle
Miocene
in
age
(13.8–12.05
Ma)
based
on
40
Ar/
39
Ar
datings
realised
on
volcanic
units
from
different
localities
of
the
Honda
Group
(on
biotites,
hornblendes,
and
plagioclases;
13.8–
12.2
Ma)
and
magnetostratigraphic
correlations
of
the
Honda
Group
(13.61
[base
of
C5ABn
chron]
–12.05
Ma
[top
of
C5An.1n])
[26,
47,
49].
The
La
Venta
area
yielded
a very
rich
verte-
brate
fauna
including
chondrichthyans,
osteichthyans,
amphibians,
squamates,
chelonians,
crocodylomorphs,
birds
and
mammals
(metatherians,
xenarthrans,
astrapotheres,
litopterns,
notoungulates,
cetaceans,
chiropterans,
sirenians,
primates
and
caviomorph
rodents)
[21,
22].
The
caviomorph
remains
from
La
Venta
included
in
this
work
were
formerly
studied
by
one
of
us
[18,
37].
In
a doctoral
thesis
manuscript,
Walton
[37]
described
several
new
genera
and
species
such
as
Microsteiromys
jacobsi
,
Ricardomys
longidens
,
Microscleromys
paradoxalis
,
and
Microscleromys
cribriphilus
.
As
the
latter
taxa
were
not
formally
described,
they
should
be
considered
as
not
available
following
the
articles
8,
9,
11,
13
of
the
International
Code
of
Zoo-
logical
Nomenclature
(i.e.,
nomen
nudum
;
p.
132–143,
246
[38]).
As
formally
recognised
for
nomina
nuda
(“A
nomen
nudum
is not
an
available
name,
and
therefore
the
same
name
may
be
made
available
later
for
the
same
or
a different
concept;
in
such
a case
it would
take
author-
ship
and
date
[Arts.
21,
50]
from
that
act
of
establishment,
not
from
any
earlier
publication
as
a nomen
nudum”
[38],
we
chose
to
keep
the
taxonomic
names
proposed
by
Walton
[37].
For
reasons
related
to
the
covid
pandemic
and
confinement,
the
original
material
from
La
Venta
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
4 / 80
could
not
be
directly
observed.
The
description
and
comparison
of
this
material
in
the
present
work
is based
on
published
photos
and
drawings
[18,
37].
Most
of
these
specimens
were
col-
lected
by
surface
collecting
and
wet
screening
(2
and
0.6
or
0.25
mm
meshes)
in
the
years
1985
through
1989.
This
material
is stored
in
the
‘Museo
Geolo
́
gico
Jose
́
Royo
y Go
́
mez,
Servio
Geo-
lo
́
gico
Colombiano’
(IGM;
formerly,
‘Instituto
Nacional
de
Investigaciones
en
Geociencias,
Mineria
y Quı
́
mica’)
and
the
‘Departamento
de
Geociencias,
Universidad
Nacional
de
Colom-
bia’
(UNC),
Bogota
́
, Colombia,
as
well
as
in
the
Field
Museum
of
Natural
History,
Chicago
(FMNH),
USA.
Some
specimens
do
not
have
collection
number
(unnumbered
specimens).
We
used
their
field
number
to
refer
to
them
(S1
Table).
Imaging
For
the
largest
specimens
(mandibular
and
maxillary
fragments)
and
high-crowned
teeth
found
at
TAR-31,
we
used
X-ray
microtomography
(
μ
CT-scan)
for
obtaining
three-dimen-
sional
digital
models
(3D
surface
renderings).
The
dental/jaw
specimens
and
the
astragalus
were
scanned
with
a resolution
of
6
μ
m
and
5
μ
m,
respectively,
using
a
μ
CT-scanning
station
EasyTom
150/Rx
Solutions
(Montpellier
RIO
Imaging
[MRI],
ISEM,
Montpellier,
France).
They
have
been
virtually
delimited
by
manual
segmentation
under
AVIZO
7.1
software
(Visu-
alization
Sciences
Group).
The
specimens
were
prepared
within
a ‘Label
Field’
module
of
AVIZO,
using
the
segmentation
threshold
selection
tool.
The
other
teeth
figured
in
this
paper
were
photographed
with
a Scanning
Electron
Microscope
(SEM)
HITACHI
S 4000
(Institut
des
Neurosciences
de
Montpellier
[INM],
France).
Some
of
the
pictures
of
the
astragalus
are
photographs
of
the
original
specimen,
which
are
the
result
of
the
fusion
of
multifocus
images
obtained
with
an
optical
stereomicroscope
(Leica
M
205C)
connected
to
a camera
(Leica
DFC
420C;
ISEM).
Nomenclature
and
comparisons
The
main
terminology
for
the
rodent
dentition
follows
Boivin
&
Marivaux
[50],
and
the
litera-
ture
cited
therein.
Some
extant
caviomorphs
and
related
extinct
forms
show
a highly
derived/
specialised
dental
pattern
with
respect
to
that
of
the
oldest
representatives
of
this
group
(for
which
the
aforementioned
nomenclature
was
based
on).
This
is the
case
of
two
taxa
described
below:
(i)
Caviidae
indet.
gen.
et
sp.,
for
which
nomenclatures
proposed
by
several
authors
[51–55]
were
followed
(Fig
2A);
and
(ii)
the
adelphomyine
Ricardomys
longidens
,
for
which
we
partly
followed
the
nomenclature
proposed
by
Patterson
&
Pascual
[56]
(Fig
2B
and
2C).
Lower
case
letters
are
used
for
lower
dentition
(dp,
for
decidious
premolar;
p,
for
premolar;
m,
for
molars)
and
upper
case
letters
for
upper
dentition
(dP,
for
decidious
premolar;
P,
for
pre-
molar;
M,
for
molars).
The
astragalar
nomenclature
is based
on
Ginot
et
al.
[57],
Rose
&
Chin-
nery
[58],
and
Wible
&
Hughes
[59].
The
caviomorph
taxa
cited
in
the
text
and
used
for
dental
comparisons
are
listed
in
S2
Table.
Three
isolated
teeth
from
MD-67
locality
in
Peru
(MUSM
1974,
1975,
and
4298;
early
middle
Miocene)
were
originally
attributed
to
cf.
Microsteiromys
sp
[31].
They
are
reassigned
here
in
light
of
the
new
material
from
Tarapoto.
For
the
MUSM
4658
astragalus,
we
used
the
same
comparative
sampling
as
Boivin
et
al.
[60].
In
addition,
we
consulted
supplementary
astragalar
material
attributed
to
the
erethizontid
Steiromys
duplicatus
(MACN
A 10055–78,
10081
and
10082;
see
[61]),
the
caviids
Dolicavia
minuscula
(MMP
10055–78)
and
Galea
leucoblephara
(INBIAL-CV
00290),
the
neoepiblemid
Neoepiblema
sp.
(probably
N
.
acreensis
;
UFAC
5249
and
61840),
the
octodontids
Abalosia
castellanosi
(MMP
1439-M)
and
Pithanotomys
columnaris
(MACN-Pv
7429–7431),
the
echimyid
Eumysops
cha-
palmalensis
(MACN-Pv
17868),
and
the
ctenomyid
Actenomys
priscus
(MMP
367-S
and
395-S).
When
the
fossils
described
in
this
paper
are
compared
with
several
taxa,
they
are
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
5 / 80
primarily
listed
according
to
their
chronostratigraphic
order
(from
the
oldest
to
the
most
recent)
and
then
alphabetically.
Nomenclatural
acts
The
electronic
edition
of
this
article
conforms
to
the
requirements
of
the
amended
Interna-
tional
Code
of
Zoological
Nomenclature,
and
hence
the
new
names
contained
herein
are
avail-
able
under
that
Code
from
the
electronic
edition
of
this
article.
This
published
work
and
the
nomenclatural
acts
it contains
have
been
registered
in
ZooBank,
the
online
registration
system
for
the
ICZN.
The
ZooBank
LSIDs
(Life
Science
Identifiers)
can
be
resolved
and
the
associated
information
viewed
through
any
standard
web
browser
by
appending
the
LSID
to
the
prefix
“http://zoobank.org/”.
The
LSID
for
this
publication
is:
urn:lsid:zoobank.org:pub
:ECE79D67-
9203-4929-9CCF-7C8D51C83
26E.
The
electronic
edition
of
this
work
was
published
in
a jour-
nal
with
an
ISSN,
and
has
been
archived
and
is available
from
the
following
digital
repositories:
LOCKSS.
Dental
associations
Working
on
isolated
teeth
requires
the
consideration
of
distinct
levels
of
morphological
varia-
tion:
(i)
variation
depending
on
the
dental
locus;
(ii)
variation
owing
to
the
occlusal
wear;
(iii)
intra-specific
or
intra-taxonomic
variation
or
polymorphism;
and
(iv)
inter-specific
or
inter-
taxonomic
variation.
In
this
study,
several
taxa
from
TAR-31
are
described.
The
association
of
the
material
attributed
to
these
taxa
is based
on
several
complementary
criteria:
(i)
a similar
occlusal
size;
(ii)
a similar
crown
height
(for
a given
occlusal
size
and
stage
of
wear);
(iii)
a sim-
ilar
proportion
of
the
maximum
mesiodistal
length
and
the
maximum
linguolabial
width;
(iv)
a similar
occlusal
outline;
(v)
a similar
occlusal
pattern
(bunolophodonty/lophodo
nty;
taenio-
donty/non-taeniodonty;
presence/absence,
development,
position
of
cusp[-id]s/styl[e/id]s;
Fig
2.
Dental
nomenclatu
re
for
cavioid
(A)
and
adelphomyine
(B,
C)
rodent
teeth
in
occlusal
view.
(A)
p4;
1,
anterior
projection
of
the
anterior
lobe;
2,
interprism
atic
furrow;
3, apexes;
4, hypoflexid
; 5, enamel;
6, dentine;
7,
posteri
or
lobe;
8, longitudinal
furrow
opposite
to
the
hypoflexid
(=
primary
internal
flexid);
9, anterior
lobe.
(B)
Lower
molar;
1,
metacon
id
(+
its
posterior
arm?);
2, first
lamella;
3, metalophuli
d I; 4, protoconi
d;
5, ectolophid
; 6, hypoflexid
; 7, anterior
outgrowth
of
the
hypoconid
; 8, hypoconid
; 9, anterior
arm
of
the
hypoconid;
10,
posterolophi
d;
11,
third
lamella;
12,
metaflexid;
13,
second
lamella;
14,
entoconid;
15,
hypolophid
; 16,
confluen
ce
between
the
anteroflexid
and
the
mesoflexid.
(C)
Upper
molar;
1, first
lamella;
2, anterolo
ph;
3, posterior
arm
of
the
protoc
one
(=
lingual
protolop
h);
4, protocone
; 5, posterior
outgrowth
of
the
protocone
; 6 hypoflexus;
7,
mure;
8, anterior
arm
of
the
hypocone;
9, hypocone;
10,
posteroloph;
11,
confluence
of
the
distal
mesoflexus
with
the
metaflexu
s; 12,
mesostyle;
13,
third
lamella;
14,
third
transverse
crest
(=
mesoloph
and/or
mesolophu
le);
15,
mesial
mesoflexus
; 16,
second
lamella;
17,
paracone;
18,
labial
protoloph;
19,
paraflexus
; ant.,
anterior;
ling.,
lingual.
(B,
C)
Note
that
the
positio
n of
the
fused
structures
is
speculative.
The
dental
terminolo
gy
is modified
after
several
authors
[50–56].
Not
to
scale.
https://doi.
org/10.1371/j
ournal.pone
.0258455.g002
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
6 / 80
presence/absence,
continuity/discontinuity,
length,
height
and
obliquity
of
loph[-id]s;
and
presence/absence
and
types
of
fusion
of
dental
structures);
and
(vi)
a compatible
occlusal
pat-
tern
between
upper
and
lower
teeth.
For
teeth
at
advanced
stages
of
wear,
we
took
into
account
the
potential
impact
of
occlusal
wear
on
characters
aforementioned.
Based
on
a wide
array
of
closely-
or
distantly-related
rodent
taxa,
we
prefered
to
use
a size
criterion
for
the
recognition
of
m1s
vs
m2s
and
M1s
vs
M2s
(i.e.,
m1/M1s
to
be
smaller
than
the
m2/M2s
in
a given
species).
The
recognition
of
other
loci
is based
on
the
occlusal
shape
of
teeth
(for
more
details,
see
descriptions
in
the
Systematic
Palaeontology
section)
and
on
the
presence
and/or
absence
of
mesial/distal
vertical
contact
facets
due
to
contiguous
teeth.
The
occlusal
shape
of
M3s
is fur-
ther
characterised
by
a smaller
and
more
labial
hypocone
with
respect
to
the
protocone
than
on
M1–2s.
However,
the
recognition
of
molar
loci
is quite
ambiguous
for
several
taxa,
espe-
cially
for
the
first
two
upper
molars.
In
that
case,
we
determined
these
loci
with
uncertainty
(e.g.,
lower
molar,
m1?,
m2?,
m3?,
upper
molar,
M1
or
M2,
M1
or
M2?,
M3?).
Measurements
and
statistics
For
each
dental
specimen
of
TAR-31,
we
measured
both
the
maximum
mesiodistal
length
(len)
and
linguolabial
width
(wid),
as
well
as
the
maximum
lingual
crown
height
(Hg)
and
labial
crown
height
(Hb;
S1
Table).
These
measurements
were
made
by
M.
Boivin
with
a mea-
suroscope
Nikon
10
and
a Keyence
Digital
Microscope
VHX-2000F.
For
Microsteiromys
jacobsi
and
Ricardomys
longidens
from
La
Venta,
we
used
the
measurements
available
in
Wal-
ton
[37]
(S1
Table).
However,
we
noted
some
inconsistencies
regarding
the
m3
of
the
type
of
Ricardomys
,
IGM
183847,
between
the
measurements
reported
in
Table
2,
p.
33
[37]
and
the
scaled
photo
of
the
specimen
in
Fig
9A,
p.
31
[37].
Similarly,
the
measurements
of
the
type
of
M
.
jacobsi
(FMNH
PM
54672
in
Table
6,
p.
58
[37])
do
not
match
with
its
scaled
representa-
tion
(in
Fig
16A,
p.
56
[37]).
For
these
two
specimens,
we
chose
to
use
the
measurements
made
from
their
figuration
in
the
Figs
9A
and
16A
[37].
For
the
Laventan
material
attributed
to
Microscleromys
,
the
measurements
made
from
the
figures
of
the
specimens
in
Walton
(Figs
11–13
[37])
and
those
from
Walton
(Fig
24.2
[18])
are
different.
We
extracted
the
measure-
ments
from
the
first
study
(Ph.D.)
because
the
specimens
were
photographed
(instead
of
drawn)
with
a larger
scale
and
more
material
is represented.
The
hypsodonty
index
of
a tooth
(HI)
[62]
equals
its
crown
height
divided
by
its
anteroposterior
length
(H/ML);
teeth
with
a
HI
<
1 are
considered
as
brachydont,
those
with
a HI
= 1 are
considered
as
mesodont,
and
those
with
a HI
>
1 are
considered
as
hypsodont.
The
latter
are
either
protohypsodont
when
they
still
have
roots,
or
euhypsodont
if they
lack
roots
[51].
Two
hypsodonty
indexes
were
cal-
culated
in
function
to
the
side
of
the
crown
considered:
from
Hb
(HIb)
and
from
Hg
(HIg).
The
HI
values
mentioned
in
the
descriptions
were
measured
on
specimens
at
the
early
stages
of
wear.
The
MUSM
4658
astragalus
from
TAR-31
was
measured
following
the
protocol
of
Ginot
et
al.
[57]
(Table
1).
Table
1.
Measuremen
ts
(mm)
of
the
astragalus
from
TAR-31
, MUSM
4658.
ABW
AmTL
ATL
ATW
EL
EW
HH
HW
LBH
2.265
~1.220
3.088
2.649
1.447
1.231
0.954
1.316
1.069
LTL
MBH
MTL
NL
SL
SW
TW
mTAH
1.860
1.682
2.015
1.382
1.211
1.035
1.832
0.725
ABW,
Astragalus
Body
Width;
AmTL,
Astragalar-
medial
Tarsal
facet
Length;
ATL,
Astragalus
Total
Length;
ATW,
Astragalus
Total
Width;
EL,
Ectal
facet
Length;
EW,
Ectal
facet
Width;
HH,
Head
Height;
HW,
Head
Width;
LBH,
Lateral
Body
Height;
LTL,
Lateral
Trochlear
Length;
MBH,
Medial
Body
Height;
MTL,
Medial
Trochlear
Length;
NL,
Neck
Length;
SL,
Sustenta
cular
Facet
Length;
SW,
Sustenta
cular
Facet
Width;
TW,
Trochlea
r Width;
mTAH,
medial
Trochlear
Arc
Height.
https://do
i.org/10.1371/j
ournal.pone
.0258455.t001
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
7 / 80
In
order
to
help
identifying
the
number
of
species
and
discriminating
which
species
of
Microscleromys
are
present
in
TAR-31
and
La
Venta,
we
performed
three
types
of
statistical
analyses
on
the
maximum
mesiodistal
length
and
linguolabial
width
by
locus
(i.e.,
dp4,
p4,
m1–
2(?),
m3(?),
dP4,
P4,
M1–2(?),
M3(?))
with
R v.4.0.3.
[63]
(S1
and
S2
Files).
First,
we
tested
sta-
tistical
differences
between
potential
taxonomic
groups
mentioned
by
Walton
(
M
.
paradoxalis
nom
.
nud
.,
M
.
cribriphilus
nom
.
nud
.,
M
.
?
paradoxalis
,
and
Microscleromys
sp.)
[37]
and
Micro-
scleromys
from
TAR-31
using
Kruskal-Wallis,
Permutation
Analysis
of
Variance,
Student
and
Wilcoxon
tests,
depending
to
what
preconditions
were
satisfied,
with
the
‘dunn.test’,
‘RVAide-
Memoire’,
‘stats’
and
‘trend’
packages
[63–66].
Second,
partitioning
analyses
were
conducted
until
five
(dp4),
nine
(m3)
or
fifteen
(other
loci)
groups
depending
the
number
specimens
for
each
loci
with
the
‘cluster’
package
[67].
Finally,
we
performed
several
Bayesian
models
on
the
most
numerous
loci
including
several
specimens
previously
attributed
to
M
.
cribriphilus
and
M
.
paradoxalis
(i.e.,
p4,
m1–2(?),
and
M1–2(?))
in
order
to
obtain
the
most
probable
partioning
with
the
‘glm’
function
of
‘MASS’
and
‘stats’
packages
[63,
68].
We
tested
39
models
for
the
M1–
2(?)
and
51
models
for
the
p4
and
m1–2(?)
including
the
partionings
following
the
previous
sys-
tematic
hypotheses
and
those
obtained
with
the
partitioning
analyses
(see
S1
File).
Abbreviations
Institutional
abbreviations.
FMNH,
Field
Museum
of
Natural
History,
Chicago,
USA;
IGM,
Servicio
Geolo
́
gico
Colombiano
(before
Instituto
Nacional
de
Investigaciones
en
Geo-
ciencias,
Mineria
y Quı
́
mica
[INGEOMINAS]),
Museo
Geolo
́
gico
Jose
́
Royo
y Go
́
mez,
Bogota
́
,
Colombia;
IGM-DU,
Field
numbers
from
expeditions
by
the
INGEOMINAS
in
cooperation
with
Duke
University.
Specimens
deposited
at
the
Museo
Geolo
́
gico
Jose
́
Royo
y Go
́
mez,
Bogota
́
, Colombia;
INBIAL-CV,
Instituto
de
Biologı
́
a de
la Altura,
San
Salvador
de
Jujuy,
Argentina;
ING-KU,
Field
numbers
from
expeditions
by
the
INGEOMINAS
in
cooperation
with
Kyoto
University.
Specimens
deposited
at
the
Museo
Geolo
́
gico
Jose
́
Royo
y Go
́
mez,
Bogota
́
, Colombia;
LACM,
Los
Angeles
County
Museum,
Los
Angeles,
USA;
MCZ,
Museum
of
Comparative
Zoology,
Cambridge,
USA;
MACN,
Museo
Bernardino
Rivadavia,
Buenos
Aires,
Argentina;
MLP,
Museo
de
La
Plata,
La
Plata,
Argentina;
MMP,
Museo
Municipal
De
Ciencias
Naturales
Lorenzo
Scaglia,
Mar
del
Plata,
Argentina;
MNHN,
Museum
National
d’Histoire
Naturelle,
Paris,
France;
MPEF-PV,
Museo
Paleontolo
́
gico
Egidio
Feruglio,
Trelew,
Argentina;
MUSM,
Museo
de
Historia
Natural
de
la Universidad
Nacional
Mayor
San
Marcos,
Lima,
Peru;
UCMP,
University
of
California
Museum
of
Paleontology,
Berkeley,
USA;
UFAC,
Colec
̧
ã
o
de
Paleovertebrados
do
Laborato
́
rio
de
Pesquisas
Paleontolo
́
gicas,
Universidade
Fed-
eral
do
Acre,
Rio
Branco-AC,
Brazil;
UNC,
Departamento
de
Geociencias,
Universidad
Nacio-
nal
de
Colombia,
Bogota
́
, Colombia.
Other
abbreviations.
CVP,
locality
belongs
to
the
Fish
Bed
in
La
Venta
area
collected
by
screenwashing
during
the
expeditions
by
the
INGEOMINAS
in
cooperation
with
Duke
Uni-
versity;
IGM-DU
loc.,
fossiliferous
locality
found
during
the
expeditions
by
the
INGEOMI-
NAS
in
cooperation
with
Duke
University
in
La
Venta
area;
UCMP
loc.,
fossiliferous
locality
found
during
the
expeditions
by
the
University
of
California
in
La
Venta
area.
Systematic
paleontology
Nomenclatural
Remark:
The
new
species
and
genera
described
below
must
be
referred
to
Boi-
vin
&
Walton,
2021,
following
the
article
50.1
and
the
“recommendation
50A
concerning
mul-
tiple
authors”
of
the
International
Code
of
Zoological
Nomenclature
(p.
182
[38]).
This
is of
particular
interest
for
nomina
nuda
that
were
previously
used
by
Walton
[18,
37]
and
made
available
here,
following
the
International
Code
of
Zoological
Nomenclature
[38].
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
8 / 80
Rodentia
Bowdich,
1821
Ctenohystrica
Huchon,
Catzeflis
&
Douzery,
2000
Hystricognathi
Tullberg,
1899
Caviomorpha
Wood,
1955
Erethicavioi
Boivin,
2019
in
Boivin,
Marivaux
&
Antoine,
2019
Erethizontoidea
Bonaparte,
1845
Microsteiromys
Boivin
&
Walton
gen.
nov.
urn:lsid:zoobank.org:act
:70F9B9F7-51EB-4FFF-
913C-2E89EBF5F9D
B
Type
species
Microsteiromys
jacobsi
gen.
et
sp.
nov.
Species
content
Only
the
type
species.
Derivation
of
name
Based
on
the
original
description
of
the
species
[37]:
‘In
reference
to
its
small
size
and
general
resemblance
to
the
erethizontoid
Steiromys
Ameghino,
1887’.
Geographical
and
stratigraphical
distribution
La
Venta,
Baraya
and
Cerro
Colorado
members,
Villavieja
Fm.
(Laventan
SALMA,
late
middle
Miocene),
Huila
Department,
Colombia.
Generic
diagnosis
As
for
the
type
and
only
known
species.
Microsteiromys
jacobsi
Boivin
&
Walton
sp.
nov.
urn:lsid:zoobank.org:act
:1DF51272-C949-
4105-BC5C-299F8
8252206
Fig
3 and
S1
Table
Nom
.
nud
.
Microsteiromys
jacobsi
Walton,
1990,
Fig
16A–16C,
p.
56.
Nom
.
nud
.
Microsteiromys
jacobsi
Walton,
1997,
Figs
24.1B
and
24.2B,
p.
394,
395.
Holotype
FMNH
PM
54672,
right
mandibular
fragment
bearing
dp4–m3
(Walton,
1990:
Fig
16A;
Wal-
ton,
1997:
Figs
24.1B
and
24.2B;
Fig
3A).
Referred
material
IGM-DU
89–249,
left
mandibular
fragment
bearing
p4–m1
(Walton,
1990:
Fig
16C;
Fig
3B);
IGM-DU
88–034,
right
posterior
fragment
of
lower
tooth
(Walton,
1990:
Fig
16B;
Fig
3C).
PLOS ONE
Miocene
caviom
orphs
from
Tarapo
to,
Peru
PLOS
ONE
| https://doi.or
g/10.137
1/journal.po
ne.02584
55
November
3, 2021
9 / 80