ArticlePDF Available

Total evidence analysis of the phylogenetic relationships of bandicoots and bilbies (Marsupialia: Peramelemorphia): Reassessment of two species and description of a new species

Authors:

Abstract and Figures

The phylogenetic relationships of bandicoots and bilbies have been somewhat problematic, with conflicting results between morphological work and molecular data. This conflict makes it difficult to assess the taxonomic status of species and subspecies within this order, and also prevents accurate evolutionary assessments. Here, we present a new total evidence analysis, combining the latest cranio-dental morphological matrix containing both modern and fossil taxa, with molecular data from GenBank. Several subspecies were scored in the morphological dataset to match the molecular data available. Both parsimony and Bayesian analyses were performed, giving similar topologies except for the position of four fossil taxa. Total evidence dating places the peramelemorphian crown origin close to the Oligocene/Miocene boundary, and the radiations of most modern genera beginning in the Late Miocene or Early Pliocene. Our results show that some species and subspecies require taxonomic reassessment, and are revised here. We also describe a new, extinct species from the Nullarbor region. This suggests that the number of recently extinct peramelemorphian species is likely to further increase.
Content may be subject to copyright.
Accepted by P. Gaubert: 13 Dec. 2017; published: ?? Month 2018
ZOOTAXA
ISSN 1175-5326 (print edition)
ISSN
1175-5334
(online edition)
Copyright © 2018 Magnolia Press
Zootaxa 0000 (0): 000
000
http://www.mapress.com/j/zt/
Article
1
https://doi.org/10.11646/zootaxa.0000.0.0
http://zoobank.org/urn:lsid:zoobank.org:pub:00000000-0000-0000-0000-00000000000
Total evidence analysis of the phylogenetic relationships of bandicoots and
bilbies (Marsupialia: Peramelemorphia): reassessment of two species and
description of a new species
KENNY J. TRAVOUILLON
1
& MATTHEW J. PHILLIPS
2
1
Western Australian Museum, Locked Bag 49, Welshpool DC, WA, 6986 Australia; E-mail: Kenny.Travouillon@museum.wa.gov.au
2
School of Earth, Environmental and Biological Sciences, Queensland University of Technology, Brisbane, Australia.
E-mail: m9.phillips@qut.edu.au
Abstract
The phylogenetic relationships of bandicoots and bilbies have been somewhat problematic, with conflicting results be-
tween morphological work and molecular data. This conflict makes it difficult to assess the taxonomic status of species
and subspecies within this order, and also prevents accurate evolutionary assessments. Here, we present a new total evi-
dence analysis, combining the latest cranio-dental morphological matrix containing both modern and fossil taxa, with mo-
lecular data from GenBank. Several subspecies were scored in the morphological dataset to match the molecular data
available. Both parsimony and Bayesian analyses were performed, giving similar topologies except for the position of four
fossil taxa. Total evidence dating places the peramelemorphian crown origin close to the Oligocene/Miocene boundary,
and the radiations of most modern genera beginning in the Late Miocene or Early Pliocene. Our results show that some
species and subspecies require taxonomic reassessment, and are revised here. We also describe a new, extinct species from
the Nullarbor region. This suggests that the number of recently extinct peramelemorphian species is likely to further in-
crease.
Key words: Bandicoot, Australia, molecular phylogeny, morphological systematics, evolution, taxonomy, new species
Introduction
The order Peramelemorphia (bilbies and bandicoots) is an order of marsupial mammals endemic to Australia, New
Guinea and surrounding islands (Van Dyck & Strahan 2008). The order has four families: Yaralidae, an extinct
family of very small insectivorous bandicoots from the Oligocene and Miocene (Muirhead 2000), Thylacomyidae,
the family of the Greater Bilby (Macrotis lagotis) and its relatives, Chaeropodidae, the family of the Pig-footed
Bandicoot (Chaeropus ecaudatus), and Peramelidae, which includes all remaining bandicoot species, though
Peroryctes, Echymipera, Microperoryctes and Rhynchomeles used to belong to their own family, Peroryctidae
(Groves & Flannery 1990). Peramelidae comprises three subfamilies, Peramelinae, which includes Isoodon and
Perameles, Peroryctinae, with only Peroryctes, and Echymiperinae, containing Echymipera, Microperoryctes and
Rhynchomeles.
There are currently 25 accepted modern species in the order, with about 20 fossil species described to date
(Warburton & Travouillon 2016). Several molecular studies have clarified the phylogenetic relationship of
peramelemorphians, grouping with Dasyuromorphia and Notoryctemorphia (Amrine-Madsen et al. 2003; Baker et
al. 2004; Nilsson et al. 2004; Phillips et al. 2006; Beck 2008; Meredith et al. 2008; Meredith et al. 2009; Mitchell
et al. 2014; Gallus et al. 2015) and also the relationships among genera within Peramelemorphia. Lower level
relationships of some taxa, especially species and subspecies of Isoodon, have proven to be quite difficult to
resolve (Baverstock et al. 1990; Westerman et al. 1999; Pacey et al. 2001; Westerman et al. 2012). Meanwhile,
morphological phylogeny of Peramelemorphia has advanced through research of the fossil record, providing an
understanding of evolution within the group (Travouillon et al. 2010; Travouillon et al. 2013a; Travouillon et al.
TRAVOUILLON & PHILLIPS
2
·
Zootaxa 0000 (0) © 2018 Magnolia Press
2013b; Gurovich et al. 2014; Travouillon et al. 2014; Chamberlain et al. 2015; Travouillon et al. 2015). There is
one major conflict between the topology of molecular trees and that of morphological trees, the position of the old
family Peroryctidae, which is sister to Isoodon and Perameles using molecular data, or sister to all
peramelemorphians (except Yaralidae) using morphological data. The latter may be a result of morphological
homoplasies such as the large auditory bullae found in both Macrotis and Isoodon, reflecting convergent evolution
in open environments (Warburton & Travouillon 2016). Kear et al. (2016) provided the first total evidence analysis
for this order, but their methodology is questionable and will be discussed in detail here.
Inferring the timescale of peramelemorphian evolution has been complicated by sparse fossil records among
modern groups, which also limit the precision of traditional node calibrations for molecular dating. Kear et al.
(2016) partly remedied this limitation by utilising fossil ages as tip dates for calibration. This promising approach
did however result in many divergences being far older than fossil records and traditional node-calibrated
molecular dates (e.g. Meredith et al. 2008; Westerman et al. 2012; Mitchell et al. 2014).
In order to improve our understanding of the evolution and phylogeny within Peramelemorphia, we combine
molecular and morphological datasets and perform a total evidence phylogenetic and dating analysis for this order.
We aim to identify which taxa are supported by both molecular and morphological data, and which require more
data in order to make informed taxonomical decisions. We also re-evaluate the taxonomic position of a couple of
taxa, and describe a new species from the Nullarbor region. Finally, the fossil taxa will be re-evaluated, to
understand how they fit in the modern phylogenetic tree.
Materials and methods
Material. Specimens examined in this study come from numerous museums across Australia and overseas
including the Western Australian Museum (prefix WAM, M; Perth, Australia), Queensland Museum (prefix QM,
JM or J or F; Brisbane, Australia), Australian Museum (prefix AM, M, P or S; Sydney, Australia), South Australian
Museum (prefix SAM, M; Adelaide, Australia), Museum Victoria (prefix NVM, C; Melbourne, Australia),
Museum and Art Gallery of Northern Territory (prefix MAGNT/NTM, U; Darwin and Alice Springs, Australia),
Natural History Museum (prefix BMNH; London, United Kingdom), Musée Nationale d'Histoire Naturelle (prefix
MNHN, ZM MO; Paris, France), Museo di Storia Naturale (prefix MSN, CE; Genoa, Italy), Zoologische
Staatssammlung Munchen (prefix ZSM; Munich, Germany) and American Natural History Museum (prefix
AMNH; New York, United States of America). The complete list of specimens is shown in the appendix.
Morphological matrix. We used the morphological matrix of Travouillon et al. (2016a) as the base of our
morphological dataset. We added morphological scores for Didelphis virginiana and Thylacinus cynocephalus as
outgroups. Instead of using Isoodon auratus, I. obesulus and Perameles bougainville, we scored some of their
subspecies, including I. a. barrowensis (from Barrow Island), I. a. auratus from the Kimberley and the desert form
of I. a. auratus (inland), I. obesulus obesulus (NSW/VIC), I. o. fusciventer (WA), Perameles b. bougainville
(Bernie and Dorre Islands, WA) and Perameles papillon sp. nov. (Ooldea, SA; previously P. b. not ina in
West erman et al. 2012). We removed Perameles allinghamensis, a fossil taxon represented by a single tooth,
because it introduced too much noise to the results (i.e. collapsing the entire tree). The extinct oldest
australidelphian Djarthia murgonensis was also removed in all analyses for the same reasons. Finally, we added
Lemdubuoryctes aruensis (Kear et al. 2016) to the morphological dataset to re-examine its relationship to other
peramelemorphians.
‘Character 14’ (Length of p3) of Travouillon et al. (2016a) was removed as it was too similar to ‘character 78’
(Relative length of p2 and p3). A new state (2, ‘enlarged anterior expansion) was added for character 54 (Shape of
narial flange of premaxilla) to reflect the unique shape of the narial flange in bilbies. Character 65 was redefined so
it is only scored in males, as canine size varies between males and females within a species as a result of sexual
dimorphism in peramelemorphians. We added a new character (155, Presence of buccal shelf on P3), a character
present in Chaeropus, some Isoodon subspecies, some Perameles subspecies and in Galadi amplus.
Molecular matrix. DNA sequences were obtained from GenBank for loci that allow the most comprehensive
taxonomic coverage for bandicoots. These include the nuclear loci, ApoB (apolipoprotein B), BRCA1 (breast and
ovarian cancer susceptibility gene 1), IRBP (interphotoreceptor retinoid binding protein gene), Rag1
(recombination activating gene-1) and vWF (von Willebrand factor gene), and the mitochondrial loci, Cytb, 12S
Zootaxa 0000 (0) © 2018 Magnolia Press
·
3
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
rRNA and 16S rRNA. Among the extinct bandicoots, partial mtDNA sequences were available for Chaeropus
ecaudatus (1,934 bp; Westerman et al. 1999, 2012) and Rhynchomeles prattorum (863 bp; Westerman et al. 2012).
A nuclear Rag1 sequence is also available for Chaeropus ecaudatus (EU369367). This sequence is most similar to
a Microperoryctes longicauda sequence (JF912124), differing only at three of the non-ambiguity sites, curiously all
as transversions (in a dataset for which transitions dominate). Chaeropus also groups with Peroryctinae and
Echymiperinae in Bayesian inference analysis with Rag1 only (Suppl. Fig. 1), when extant Peramelinae and
Peroryctinae clades are reciprocally monophyletic, in agreement with all recent molecular and morphological
studies. Morphology (e.g. Travouillon et al. 2014) and mtDNA (e.g. Westerman et al. 1999) both place Chaeropus
well outside of Peroryctinae. In caution, we await independent verification before including this sequence.
Both the nuclear and mtDNA sequences were initially aligned in ClustalW2 (Larkin et al. 2007) with penalties
of 5 for gap opening and 0.2 for gap extension. Manual adjustments were then made in Se-Al 2.0a (Rambaut 1996),
where sites with ambiguous homology were excluded. The overall length of the mtDNA and nuclear components
of the alignment are 3,266 bp and 5,967 bp respectively, totalling 9,233 bp. Inclusion of extant outgroup taxa for
the DNA data follows the morphological dataset, with Dasyurus, Antechinus and Sminthopsinae, but also including
Didelphis virginiana assists with rooting and deep calibration for the tree. The extinct outgroup, Thylacinus
cynocephalus could only be included for the mtDNA.
Among the mitochondrial data, the Cytb 3
rd
codon positions were RY-coded (A,G →R, C,T→Y). This follows
Phillips & Pratt (2008) and Mitchell et al. (2014) in acknowledging the inability of substitution models to
accurately account for high levels of saturation, and high base compositional stationarity. Cytb 1
st
and 2
nd
codon
positions, 12/16s rRNA and the nuclear protein coding sequences were maintained as standard nucleotides.
Phylogenetic inference. Bandicoot phylogeny was inferred under maximum parsimony (MP) and Bayesian
inference (BI) from the morphology and DNA datasets separately and concatenated for “total evidence”.
Substitution model categories for each data partition employed the more general of the jModelTest 0.1.1 (Posada
2008) hLRT or AIC recommendations (Table S1) or the next most general available for each phylogenetic
inference program. Substitution was modelled separately among eight DNA partitions, these being the three
nuclear and three mt protein coding codon positions, and the mt RNA stem and loop sites. For the morphological
data we employed the Mkv+Γ model for ordered and unordered characters.
MP bootstrap analyses were undertaken in PAUP 4.0b10 (Swofford 2002), and employed 1,000
pseudoreplicates. Bayesian inference was carried out with MrBayes 3.2.6 (Ronquist et al. 2012). MrBayes
phylogenetic inference analyses ran two independent sets of two MCMC chains for 10,000,000 generations, with
trees sampled every 5,000 generations and the default 25% burnin, which was sufficient to ensure that –lnL had
plateaued, clade frequencies had converged between runs, and estimated sample sizes for substitution parameters
were >200 (using Tracer v1.5, Rambaut & Drummond 2007). The substitution rates matrix, proportion of invariant
sites, gamma shape (α), and empirical base frequencies were unlinked between partitions. Morphological and
molecular branch-lengths were unlinked, and among the molecular partitions the branch-lengths were separately
scaled (ratepr=variable), allowing partitions to inform each other on relative (molecular) branch lengths across the
tree.
Total evidence geomolecular dating. We estimated the timescale of bandicoot evolution in MrBayes 3.2.6,
using the substitution and branch-length models described above for phylogenetic inference, except where
otherwise stated. First we ran “traditional” node-calibrated dating on the molecular dataset, with only the DNA
sequences included. Then we ran “total evidence” dating for the combined data. In these analyses the tip ages of the
fossil taxa provide calibration, with or without the node constraints also helping to inform rate variation across the
tree. The root calibration however, was applied to all dating analyses. To further enhance comparability between
the timetree estimates all MrBayes dating analyses employed the independent gamma rates (IGR) model, the
default prior setting igrvarpr=exp(10)prior, and a lognormal prior setting for clockratepr=lognorm(-5.0,0.6), which
was informed by initial analyses with broad, normally distributed clockrate priors. As for the MrBayes
phylogenetic inference (non-clock) analyses we employed two independent sets of two MCMC chains for
10,000,000 generations, with trees sampled every 5,000 generations, the default 25% burnin, and the same strategy
for checking MCMC stationarity and convergence.
Node-calibrated dating of the DNA-only dataset used a uniform clock prior for branch-lengths, whereas the
tip-dated analyses (with or without node calibrations) used the fossilized birth-death model, with the following
priors: speciationpr = exp(1), extinctionpr = beta(1,1), fossilizationpr = beta(1,1). Taxon sampling did not
TRAVOUILLON & PHILLIPS
4
·
Zootaxa 0000 (0) © 2018 Magnolia Press
specifically target deep-level or clustered diversity, so the default (random) sampling strategy was employed, with
sampling probability 0.3, which is roughly the geometric mean of the included bandicoot and dasyurid extant
sampling proportions. In addition, we tested the effect of linking and unlinking the molecular and morphological
relaxed clocks.
Node calibrated analyses employed seven primary calibration priors.
1. The root: Didelphis—Australidelphia (uniform bounds, 58.5–83.8 Ma). The minimum is based on the
presence of didelphoid tarsals (Szalay 1994) in the Itaborai fauna of Brazil. See Phillips et al. (2009, SI) for further
discussion. The maximum bound acknowledges the absence of crown marsupials from well-sampled Campanian
mammal faunas of Asia, North and South America, in which metatherians are present (at least in Asia and North
America), but only stem members.
2. Isoodon—Perameles (uniform bounds, 3.62–18.51 Ma). The minimum bound is based on the minimum age
of Perameles allinghamensis from the Bluff Downs Local Fauna (Mackness et al. 2000). The maximum bound
covers the absence of putative members of Isoodon/Perameles from well-sampled Riversleigh faunal zone B
faunas, such as at the radiometrically dated Neville’s Garden and Camel Sputum sites (see Woodhead et al. 2016).
There remains some uncertainty over the placement of the Middle Miocene Crash bandicoot (Travouillon et al.
2014), which if resolved outside Isoodon and Perameles may allow for a slightly younger maximum bound for this
clade.
3. Peramelidae: Perameles—Peroryctes (truncated normal prior, 3.62–23.03 Ma). The hard minimum bound is
based on the minimum age of Perameles allinghamensis from the Bluff Downs Local Fauna (Mackness et al.
2000). The 97.5% soft maximum bound acknowledges the absence of crown peramelids from well-sampled Early
Miocene faunas that include stem bandicoots. The normal distribution acknowledges that Peramelidae is expected
to be substantially older than the advanced Pliocene bandicoots, and that the maximum bound is also conservative,
with Late Oligocene bandicoots being very plesiomorphic. A normal mode of 13.325 Ma allows for a higher
probability density around the Middle Miocene, close to apparently transitional forms such as Crash bandicoot.
4. Peramelemorphia—Dasyuromorphia (truncated normal prior, 25.0–72.3 Ma). The hard minimum bound is
based on the minimum age of Bulungu muirheadae (Travouillon et al. 2013b) from the Ditjimanka Local Fauna
(Etudunna faunal zone B). The 97.5% soft maximum bound acknowledges the absence of marsupials from South
American Maastrichtian mammals faunas, and the absence of crown eometatherians (the Australasian clade) from
North American marsupial faunas. The normal distribution acknowledges that the minimum bound for
Peramelemorphia/Dasyuromorphia is very conservative, given the long fossil record hiatus back to the Tingamarra
Fauna, and that the maximum bound is also conservative, with Cretaceous marsupials being substantially more
plesiomorphic. A normal mode of 54.6 Ma allows for a higher probability density around the Early Eocene, near
the basal Australian marsupial radiation found in the Tingamarra fauna, which includes dental taxa putatively close
to the divergence of bandicoots and dasyuromorphians (Godthelp et al. 1992; Black et al. 2012).
5. Phascogalini: PhascogaleAntechinus (uniform, 4.36–15.09 Ma). The minimum bound is based on the
minimum age of Antechinus sp. from the Hamilton Local Fauna (Turnbull et al. 2003). The maximum bound
covers the absence of putative phascogalines from well-sampled Riversleigh faunal zone C sites, such as at the
radiometrically dated AL90 and Ringtail sites (see Woodhead et al. 2016).
6. Dasyuridae: Dasyurus—Sminthopsinae (truncated normal prior, 4.36–23.03 Ma). The hard minimum bound
is based on the minimum age of Antechinus sp. from the Hamilton Local Fauna (Turnbull et al. 2003). The 97.5%
soft maximum bound acknowledges the absence of crown dasyurids from well-sampled Early Miocene faunas that
include stem dasyurids/dasyuromorphians. The normal distribution acknowledges that the minimum bound for
Dasyuridae is conservative, given the sparse Late Miocene dasyuromorphian record, and that the maximum bound
is also conservative, with latest Oligocene dasyuromorphians being very plesiomorphic. A normal mode of 13.695
Ma allows for a higher probability density around the Middle Miocene, near apparently transitional forms such as
Barinya wangala and Joculusium muizoni (Wroe 1999, 2001; Kealy 2013 Unpub. Honours thesis).
7. Dasyuromorphia: Thylacinus—Dasyuridae (uniform bounds, 23.03–54.65 Ma). The minimum bound is
based on the Riversleigh faunal zone A (Late Oligocene) thylacinid, Badjcinus turnbulli (see Muirhead & Wroe
1998). The maximum bound is the maximum age of the Tingamarra Fauna, which among marsupials includes only
far more “primitive” forms (Black et al. 2012). The maximum is necessarily conservative, because of a long fossil
record hiatus prior to several 25Ma faunas that include putative dasyuromorphians.
Node calibrations have typically been defined for extant clades; however, tip dating with extinct taxa also
Zootaxa 0000 (0) © 2018 Magnolia Press
·
5
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
raises the opportunity to calibrate nodes defined at least partly by fossils. We include the following example in
some analyses.
8. Crown Peramelemorphia—†Bulungu (Uniform bounds, 25.0–54.65 Ma). We sampled the Early to Middle
Miocene Bulungu palara, but the genus first appears in the Late Oligocene Etadunna faunal zone B, which
provides the minimum bound for its divergence from crown bandicoots. The maximum bound is the maximum age
of the Tingamarra Fauna, which among marsupials includes only far more “primitive” forms (Black et al. 2012).
Tip calibrations based on the fossils and localities described above in the “Morphological matrix” section, and
employed in total evidence dating were provided for the following fossil taxa: Mutpuracinus archibaldi (uniform,
12–15 Ma), Barinya wangala (uniform, 13–18 Ma), Perameles bowensis (fixed at 3.6 Ma), Yarala burchfieldi
(uniform, 13–18 Ma), Galadi speciosus (uniform, 17–19 Ma), Galadi amplus (uniform, 13–14 Ma), Bulungu
palara (uniform, 13–18 Ma), and Madju variae (uniform, 13–18 Ma). The dates approximate the temporal ranges
of the (often multiple) specimens that were required for scoring the morphological matrix.
Morphometric analysis. In order to investigate the relationship of Perameles bougainville, P. eremiana and
Perameles papillon sp. nov., a simple morphometric analysis was performed following the method of Travouillon
(2016a). All type material were included, even types of species not currently recognised (Perameles myosuros, P.
arenaria, P. notina, P. fasciata). All specimens previously referred to any of those taxa were used in the analysis
(see Table S3–4). All measurements were taken using digital callipers. External measurements including hind foot
length, tail length and ear length were taken on skins and alcohol specimens. 35 cranial and 32 dental
measurements were taken including length and width of all premolars and molars. Cranial and dental
measurements were used to perform a Principle Component Analysis (PCA) and a Canonical Variate Analysis
(CVA). External measurements were summaries in Table S2. Sexual dimorphism was investigated within the
Perameles bougainville complex, by averaging skull length for each sex in each taxon, and calculating size ratios,
following Ralls (1976).
Dental analysis of Lemdubuoryctes aruensis. In order to investigate the relatedness of Lemdubuoryctes
aruensis to species of Peroryctes, dental measurements were taken on several specimens (see Table S5), including
lengths and width of premolars and molars, following Travouillon (2016a). Dental measurements of L. aruensis
and species of Peroryctes were then plotted on an XY graph for each tooth, with convex hulls highlighting the
range of measurements for each species.
Results
Phylogenetic inference. Our non-clock maximum parsimony (MP) and Bayesian inference (BI) analyses highlight
several phylogenetic signal differences between the morphological and molecular datasets (Fig. 1 A–B).
Morphology groups Macrotis with Isoodon with moderate support (64% MP bootstrap, 0.61, MP-BP; Bayesian
posterior probability, BPP), and provides strong support for a clade including Macrotis, Isoodon, Chaeropus and
Perameles (96% MP-BP, 1.00 BPP). The molecular data made up from five nuclear and three mitochondrial genes
instead places Macrotis and Chaeropus outside all other extant and recently extinct bandicoots, which form a
monophyletic clade (98% MP-BP, 1.00 BPP).
The predominantly New Guinean bandicoots, Peroryctinae and Echymiperinae, group together in all separate
morphology and molecular analyses, although support for this grouping is only strong with these data combined
(83% MP-BP, 1.00 BPP). Within Echymiperinae, it remains unclear whether Rhynchomeles prattorum is sister to
all Echymipera (DNA only, non-clock combined DNA/morphology) or as sister to Echymipera clara (morphology
BI, relaxed-clock combined DNA/morphology). In no case was R. prattorum grouped with E. clara or excluded
from Echymipera with greater than 47% MP-BP or 0.54 BBP.
The molecular tree (Fig. 1B) fits current taxonomy, with the peroryctine-echymiperine clade grouping with the
perameline. The alternative, deep morphological placement of the plesiomorphic Peroryctinae-Echymiperinae
clade (Fig. 1A) follows Macrotis and Chaeropus grouping with the peramelines. However, Echymiperinae is
paraphyletic in both the MP and BI morphology trees, with Microperoryctes falling outside Echymipera and
Peroryctes (51% MP-BP, 0.79 BPP). In this and other cases in which morphology only weakly favoured an
alternative topology to the molecular tree, the addition of the DNA data in the combined (total evidence) tree
favoured the ‘molecular grouping’. One exception is for the placement of Isoodon (obesulus) fusciventer. This
TRAVOUILLON & PHILLIPS
6
·
Zootaxa 0000 (0) © 2018 Magnolia Press
7$%/(*HRPROHFXODUGDWHV0DIRUNH\GLYHUJHQFHVDPRQJ3HUDPHOHPRUSKLD$%DVHGRQWKH'1$GDWDDORQHDQGQRGHFDOLEUDWLRQWKHQ³WRWDOHYLGHQFH´HPSOR\LQJ
XQOLQNHG'1$DQGPRUSKRORJLFDOUHOD[HGFORFNVDQGXVLQJIRVVLOVIRUWLSGDWLQJ%ZLWKRXWQRGHFDOLEUDWLRQVRU&DOVRZLWKQRGHFDOLEUDWLRQVIRUPRGHUQFODGHVDQG'
ZLWKQRGHFDOLEUDWLRQVIRUPRGHUQFODGHVDQGDPRQJ2OLJR0LRFHQHEDQGLFRRWV(.HDUet al
&ODGH '1$RQO\ '1$DQGPRUSKRORJ\ .HDUet al
$1RGHGDWLQJ %7LSGDWLQJ &7LSQRGHGDWLQJ '7LSIRVVLOQRGHGDWLQJ (7LSQRGHGDWLQJ
3HUDPHOHPRUSKLDWRWDOJURXS  ± ± ± 
3HUDPHOHPRUSKLDFURZQJURXS ± ± ± ± ±
Chaeropus/Macrotis    ± 
Macrotis3HUDPHOLGDH ±    ±
Chaeropus3HUDPHOLGDH   ±  
Chaeropus3HUDPHOLQDH  ±   
Madju variae3HUDPHOLGDH  ±
F
 ± ± 
3HUDPHOLGDH ± ±
F
 ± ± ±
3HURU\FWLQDH(FK\PLSHULQDH ± ± ± ± ±
Peroryctes ± ± ± ± ±
(FK\PLSHULQDH ± ± ± ± ±
Microperoryctes
D
± ± ± ± ±
Echymipera/Rhynchomeles ±    ±
E. clara/Rhynchomeles  ± ± ± 
Echymipera ± ±
G
 ±
G
 ±
G
 ±
3HUDPHOLQDH ± ± ± ± ±
Isoodon ± ± ± ± ±
Perameles
E
± ± ± ± ±
D
M. longicauda WRM. papuensis
E
H[FOXGHVP. bowensis
F
DOVRLQFOXGHV Chaeropus,
G
LQFOXGHV Rhynchomeles
Zootaxa 0000 (0) © 2018 Magnolia Press
·
7
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
FIGURE 1. Peramelemorphian phylogeny inferred from (A) morphology, with the maximum parsimony (MP) tree shown and
Bayesian inference (BI) differences indicated in dashed arrows (including C. ecaudatus falling within Perameles), from (B)
DNA, with the BI tree shown and MP differences indicated as dashed arrows (including Perameles being paraphyletic relative
to Isoodon), and from combined data (C) MP, (D) BI, and (E) BI also informed by temporal signal and with posterior
probabilities indicated for nodes. Oligo-Miocene bandicoot names are shaded grey, and could not be included in the DNA-only
tree (B), which instead shows the Peroryctinae-Echymierinae expanded to genera. Outgroup (Dasyuromorphia and Didelphis)
not shown.
TRAVOUILLON & PHILLIPS
8
·
Zootaxa 0000 (0) © 2018 Magnolia Press
FIGURE 2. Total evidence Bayesian inference timescale of peramelemorphian evolution. Node and tip calibrations are
employed under a fossilized birth-death model, with molecular and morphological relaxed clock models unlinked. Blue bars
show 95% highest posterior density intervals for node ages. The outgroup includes Didelphis and eight dasyuromorphians.
taxon is placed in the molecular trees as sister to all other Isoodon, and in the morphological trees as sister to I.
macrourus and I. obesulus. The clock and non-clock combined data analyses place I. (obesulus) fusciventer as
sister to I. macrourus or I. auratus. All analyses are in agreement on I. (obesulus) fusciventer not forming a clade
with the other included I. obesulus.
The non-clock MP and BI trees (Fig. 1C–E) agree on all intergeneric relationships among the modern taxa,
except for only weak incongruence for the placement of the root between Macrotis, Chaeropus and Peramelidae.
An intriguing difference that including the DNA brings, is shifting the Miocene fossil, Madju variae from outside
crown peramelemorphians to fall as sister to the Peroryctinae-Echymiperinae clade, either with (BI) or without
(MP) the other Oligo-Miocene fossil bandicoots. This relationship appears to be facilitated by the DNA shifting
Macrotis and Chaeropus deeper, such that symplesiomorphy between the peroryctines/echymiperines and Oligo-
Miocene bandicoots (or at least Madju variae) could only be expressed by the latter being drawn within the
peramelemorphian crown. Relaxed-clock BI analyses of the combined data add a temporal signal for inferring
phylogeny, and restore the basal placement of the Oligo-Miocene bandicoots, in line with the morphology-only
trees (Fig. 1A).
Younger fossil bandicoots were also included in our analyses. The ~earliest Holocene Lemdubuoryctes
aruensis was sister to Peroryctes broadbenti with 100% MP-BP or BPP in all analyses. The Pleistocene Perameles
sobbei was clearly placed within Perameles in all analyses, although its placement within that genus varied
substantially. The Pliocene Perameles bowensis was less clearly resolved, with only the combined data MP tree
(Fig. 1C) favouring an affinity with Perameles. Morphology alone (Fig. 1A) placed Perameles bowensis outside
Peramelinae (and Macrotis/Chaeropus). Even with Chaeropus and Macrotis shifting towards the root upon the
inclusion of DNA for the combined data BI analyses, Perameles bowensis fell outside Peramelinae (Fig. 1D) or
Zootaxa 0000 (0) © 2018 Magnolia Press
·
9
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
grouped with Isoodon (Fig. 1E). The recently extinct Perameles (notina) papillon sp. nov. grouped with Perameles
bougainville with weak support in the molecular analyses (47% MP-BP, 0.49 BPP). Combined data phylogenetic
analyses instead placed Perameles (notina) papillon sp. nov. with Perameles eremiana (30% MP-BP, 0.72 BPP).
Total evidence geomolecular dating. As well as inferring the timescale of bandicoot evolution, we examined
the influence of including morphological data and fossil tips on divergence estimation. For comparison we
maintain a similar framework across analyses, using independent gamma rates (IGR) relaxed clocks within
MrBayes 3.2.6 (Ronquist et al. 2012). We begin without morphological data and fossil tips, and employ node-
calibrated molecular dating. Estimates (and 95% highest posterior density, HPD) for crown ages include 27.1
(22.1–32.6) Ma for Peramelemorphia and 11.7 (9.8–14.0) Ma for Peramelidae. Tip dating (without bandicoot node
calibrations) provides a younger estimate of 21.4 (16.5–26.5) Ma for the peramelemorphian crown, but an older
estimate of 15.3 (11.5–20.3) Ma for Peramelidae.
Including the node calibrations employed in the node-calibrated molecular dating, with the tip dating had
almost no influence on the peramelemorphian (median) divergence estimates. This tip+node dating did slightly
improve precision, compared with the tip dates (Table 1, 95% HPDs). Further including a node calibration among
the Oligo-Miocene bandicoots, based on the stem lineage of Bulungu diverging ≥ 25 Ma is temporally informative,
pushing back the divergence of Oligo-Miocene lineages from modern bandicoots, variously by 1–2.6 Ma.
Divergences within Peramelidae were unaffected.
Additional clade ages and 95% HPDs are shown in Table 1, with our favoured tip + fossil node dating
estimates (Table 1D) also presented in Fig. 2. Most notably, the crown originations for several genera (Perameles,
Peroryctes, and Echymipera) occur over a short interval close to the Miocene-Pliocene boundary. Median
estimates for these divergences range from 4.1–6.0 Ma for the DNA-only node dating to 5.0–6.9 Ma for the “total
evidence” tip+node dating (Table 1). Crown origins of Isoodon and Macrotis are younger, with median estimates in
the Late Pliocene or Early Pleistocene.
Morphometric analysis. Principal Component Analysis (PCA) and Canonical Variate Analysis (CVA)
recovered similar results for both cranial and dental measurements of the Perameles bougainville complex (Fig. 3).
In the cranial dataset (Fig. 3A–B), Component 1 in the PCA (accounting for 62% of variance) and Axis 1 in the
CVA (64.1% of variance) were separating taxa based on measurements of the skulls related to its elongation, and
overall size (onl, nl, nps, bcl and jl), while Component 2 (accounting for 9.6% of variance) and Axis 2 (19.2% of
variance) separated taxa based on measurements related to the width of the skull and the size of the bullae (ppw,
iow, BH, BL, BW). In the PCA (Fig. 3A), there is a slight overlap between P. myosuros (with P. arenar ia falling in
the middle of its distribution), P. notina and P. fasciata, and some overlap between P. papillon sp. nov. and P.
fasciata, but P. bougainville and P. eremiana do not overlap with any other taxa. In the CVA (Fig. 3B), there is no
overlap between any of the taxa, forming well defined clusters away from one another, with P. myosuros and P.
notina closest to one another. The larger specimens (larger Component 1 and Axis 1 values) of P. papillon sp. nov.
and P. myosuros (including P. arenaria) are all females, while the larger specimens for P. notina, P. fasciata and P.
eremiana are all males, and there is no particular pattern for P. bougainville. This result is also recovered by
calculating ratios of male/female skull sizes (Table 2).
TABLE 2. Average skull length of males and females of Perameles bougainville complex taxon, including size ratios.
In the dental dataset (Fig. 3C–D), Component 1 in the PCA (accounting for 40.9% of variance) and Axis 1 in
the CVA (49.7% of variance) were separating taxa based on upper and lower molar length, while Component 2
Male Female
Taxa n Average n Average M/F ratio F/M ratio
Perameles eremiana 1 62.24 3 57.93 1.07 0.93
Perameles myosuros 6 61.55 6 65.72 0.94 1.07
Perameles notina 1 67.61 1 60.63 1.12 0.90
Perameles faciata 1 61 1 56.74 1.08 0.93
Perameles papillon 2 55.19 8 58.29 0.95 1.06
Perameles bougainville 5 56.54 5 57.04 0.99 1.01
TRAVOUILLON & PHILLIPS
10
·
Zootaxa 0000 (0) © 2018 Magnolia Press
(accounting for 9.8% of variance) and Axis 2 (23.56% of variance) separated taxa based on premolar length and
width. In both analyses, P. fasciata stands out, with much larger teeth and do not overlap with any taxa. All others
have some overlap. In the PCA (Fig. 3C), P. bougainville does not overlap with P. notina, but both overlap with the
remaining taxa. In the CVA (Fig. 3D), there is less overlap than in the PCA, but there is still significant overlap
between P. myosuros (plus P. arenaria) and P. eremiana.
The external measurements (ear, pes and tail lengths) are summarised in Table S2. Perameles bougainville has
the shortest ears, averaging 33mm, followed in increasing size by P. myosuros at 36mm, P. eremiana, P. notina and
P. papillon sp. nov. at 39mm (average couldn’t be calculated for P. arenaria [39mm] and P. fasciata [41mm] which
only have one specimen). P. papillon sp. nov. has the shortest pes on average, at 49mm, followed by P.
bougainville at 50mm, P. erem iana at 53mm, P. myosuros at 55mm, and P. notina at 58mm (single specimen of P.
arenaria at 57mm, and P. fasciata at 72mm, largest pes). The tail length was highly variable, as a result of
bandicoot tails being often damaged and shortened (often not preserved on specimens). P. papillon sp. nov. has the
shortest tail on average at 76mm, followed by P. bougainville at 79mm, P. notina at 85mm, P. eremiana at 98mm
and P. myosuros at 102mm (P. arenaria has a shortened tail at 60mm; P. fasciata is 84mm).
FIGURE 3. Results of the morphometric analysis of the Perameles bougainville complex. A, Principal Component Analysis of
cranial data; B, Canonical Variate Analysis of cranial data; C, Principal Component Analysis of dental data; D, Canonical
Variate Analysis of dental data. Upright rectangles, Perameles bougainville; crosses, Perameles eremiana; diamonds,
Perameles myosuros; circle, Perameles arenaria; stars, Perameles notina; triangles, Perameles fasciata; laying rectangles,
Perameles papillon sp. nov.
Zootaxa 0000 (0) © 2018 Magnolia Press
·
11
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
Dental analysis of Lemdubuoryctes aruensis. Dental measurements including premolars and molars of
Peroryctes broadbenti, P. raffrayana and Lemdubuoryctes aruensis are represented in Figure 4. In general, there is
no overlap in dental size between P. broadbenti and P. raffrayana except for the M1 (Fig. 4D), m1 (Fig. 4K), and
the P3 (Fig. 4C), where female P. broadbenti, have similar sized P3 to P. raffrayana. Teeth of L. aruensis fall within
the range sizes of P. broadbenti for the P1 (Fig. 4A), P3 (Fig. 4C), M1 (Fig. 4D), p2 (Fig. 4I), p3 (Fig. 4J), m1 (Fig.
4K), m2 (Fig. 4L) and m4 (Fig. 4N). The P2 and p1 of L. aruensis are larger than those of P. broadbenti (Fig. 4B,
4H); the M2 and M3 are smaller than those of P. broadbenti but larger than P. raffrayana (Fig. 4E–F);and the M4
and m3 are less wide than those of P. broadbenti (Fig. 4G, 4M).
FIGURE 4. X-Y graphs of the tooth length versus tooth width, for the upper (A–G) and lower (H–N) premolars and molars, for
Peroryctes raffrayana (squares), Peroryctes broadbenti (crosses), and Lemdubuoryctes aruensis (star).
Systematics
Order PERAMELEMORPHIA Kirsch 1968 (Aplin & Archer 1987)
Superfamily Perameloidea Waterhouse 1838
Family Peramelidae Gray 1825
TRAVOUILLON & PHILLIPS
12
·
Zootaxa 0000 (0) © 2018 Magnolia Press
Genus Isoodon Desmarest 1817
Isoodon fusciventer Gray 1841 new stat.
Isoodon obesulus fusciventer Gray 1841
Holotype. BMNH 41.1176, skin and skull.
Type locality. King George Sound, Western Australia, Australia
Distribution. South-west Western Australia.
Remarks. The results of our analysis show that the current subspecies of Isoodon obesulus do not form a
monophyletic clade. This was also recovered in the analysis of We sterman et al. (2012), where Isoodon obesulus
fusciventer does not form a clade with Isoodon obesulus obesulus suggesting that it is a distinct taxon. As a result,
we propose to raise Isoodon obesulus fusciventer to species status, and provide a new diagnosis.
Diagnosis. Isoodon fusciventer differs from all other Isoodon species in having a large StA on M1, with a long
crest running anteroposteriorly (this cusp and crest are small and short in all other taxa). Isoodon fusciventer differs
form I. obesulus and I. auratus in having a well-developed lingual shelf that extends to the buccal side on P3; no
stylar crest on M1; retaining a StC/D on M4. Isoodon fusciventer differs from I. obesulus and I. macrourus in
having fused StB and StC on M1 (these cusps are identifiable as distinct cusps in other taxa); posthypocristid is
oblique to the tooth row on m1and m2; preentocristid on m4 is absent. Isoodon fusciventer differs from I. auratus
in having an enlarged anterior cingulum on M1 which connects to the talon; having P1 as long as P2 (shorter in I.
auratus). Isoodon fusciventer differs from I. obesulus in having a preparacrista posterobuccally orientated to
connect to a posteriorly located StB on M1; having a ventrally angular alisphenoid tympanic process which
terminates anterior to the transverse foramen. Isoodon fusciventer differs from I. macrourus in having P2 and P3
subequal in length (P2 is shorter in I. macrourus); retaining a stylar crest on M2; cristid obliqua on m1 terminates
lingual to the protocone but buccal to the midpoint of the tooth width (buccal to the protocone in I. macrourus).
Genus Perameles Geoffroy 1803
Perameles papillon sp. nov.
Figs. 5–7
Holotype. WAM M571, skin and skull, female.
Type locality. Ooldea, South Australia (SA).
Paratypes. WAM M126, skull, sex unknown, Eucla Pass, Western Australia (WA); WAM M570, skin and
skull, female, Ooldea (SA); WAM M572, skin and skull (juvenile), male, Ooldea (SA); WAM M574, skin and skull
(Juvenile), male, Ooldea (SA); WAM M576, skin and skull, female, Ooldea (SA); WAM M577, skin and skull,
female, Ooldea (SA); AM M4352, Skin and skull, female, Rawlinna (WA); AM M31547, alcohol specimen, male,
Ooldea (SA); BMNH 1925.10.8.26, skin, male, Ooldea (SA); BMNH 1925.10.8.27, skin, female, Ooldea (SA);
SAM M3973, alcohol specimen and skull, female, collected between Ooldea and Tallaringa (SA); SAM M4640,
skin and skull, male, Ooldea (SA).
Referred specimens. AMNH 196371, 2 left subfossil dentaries, Horseshoe Cave, Nullarbor (WA); AMNH
220153 and AMNH 220154, juvenile subfossil skulls, south west Loongana, Nullarbor (WA); AM S1821, skull,
juvenile, Rawlinna (WA); AM M2986, skin and skull, female, Fisher, Nullarbor (SA); AM M4353, skin and skull,
juvenile female, Rawlinna (WA); AM M4850, skin and skull, male, Ooldea (SA); AM M4851 and M4852, skin
and skull, female, 9 miles west of Ooldea (SA); AM M3049, alcohol specimen, juvenile female, Ooldea (SA); AM
M4368–M4369, alcohol specimen, juvenile female, Rawlinna (WA); AM M4370–M4371, alcohol specimen,
juvenile male, Rawlinna (WA); AM M4939 and M4941, alcohol specimen, pouch young male, Ooldea (SA); AM
M4940, alcohol specimen, pouch young female, Ooldea (SA); AM M4942–4943, alcohol specimen, male, East
West Line (SA); AM M4944–4945, alcohol specimen, female, East West Line (SA); AM M4978–4979, alcohol
specimen, pouch young female of M4852, 9 miles west of Ooldea (SA); AM M31545, alcohol specimen, juvenile
male, Ooldea (SA); AM M31548, alcohol specimen, male, Ooldea? (SA); MAGNT U7605, skull, specimen bred in
Adelaide in 1926 (original locality unknown); NMV C7173 and C7174, subfossil skulls, N11 Cave, Nullarbor East
Zootaxa 0000 (0) © 2018 Magnolia Press
·
13
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
(SA); NMV C7175 and C7176, subfossil skulls, Weebubbie Cave, Nullarbor (WA); NMV C7177 and C7178,
subfossil skulls, Weekes Cave, Nullarbor (SA); NMV C7180, 2 subfossil dentaries, Koonalda Cave, Nullarbor
(SA); QM J5028, alcohol specimen, female, Nullarbor (SA); SAM M846, alcohol specimen and skull, juvenile
female, Ooldea (SA); SAM M1394, mount, male, Ooldea (SA); SAM M1395, mount, Ooldea (SA); SAM M1396,
mount, female, Ooldea (SA); SAM M1397, skin, male Ooldea (SA); SAM M1398, skin, female Ooldea (SA);
SAM M3926–3927, M3976, M3981–3982, M3984, M3987–3988, alcohol specimen and skull, female, no location
data; SAM M3977, M3983, M3985–3986, alcohol specimen, juvenile, no location data; SAM M3980, M3989,
alcohol specimen and skull, male, no location data; SAM M4639, skin and skull, juvenile female, Ooldea (SA);
SAM M5226, mount, no data; SAM M6302, many subfossil skulls, Weebubbie Cave, Nullarbor (WA); SAM
M6303, many subfossil skulls, Weekes Cave, Nullarbor (SA); SAM M6304, many subfossil skulls, N11 Cave,
Nullarbor East (SA); SAM M6305, many subfossil skulls, Koonalda Cave, Nullarbor (SA); WAM 76.4.21, skull,
Abrakurrie Cave, Nullarbor (WA); WAM 67.4.114–116, 67.10.305–307, skulls, Homestead Cave, Nullarbor (WA);
WAM 66.6.45–53, 66.1.54, 68.3.63–67, 69.5.17, skulls, Murra-el-elevyn Cave, Nullarbor (WA); WAM
69.7.557–559, skulls, Kjeldahl Cave, Cocklebiddy Roadhouse, Nullarbor (WA); WAM 73.1.107, 74.5.39, skulls,
Webbs Cave, Nullarbor (WA); WAM 67.5.128, 67.5.130, skulls, Warbla Cave, Nullarbor (SA); WAM
67.10.420–421, 68.3.118, skulls, Blowhole, Eucla, Nullarbor (WA); WAM 67.4.250, 72.1.8, 72.1.1088–1089,
skulls, Horseshoe Cave, Nullarbor (WA); WAM 66.6.45–53, skulls, Graham’s Cave, Nullarbor (WA); WAM
68.3.51, skull, cave 3 3/4 mile South of Cocklebiddy Cave, Nullarbor (WA); WAM 68.3.33, skull, Mullamullang
Cave, Nullarbor (WA); WAM 67.11.32, skull, Capstan Cave, Nullarbor (WA); WAM 67.4.343, skull, cave 23 miles
west of Eucla, Nullarbor (WA); WAM 68.3.20, skull, cave 6 miles south of Madura, Nullarbor (WA); WAM
67.9.134, cave near Madura, in doline, Dingo Douga, Nullarbor (WA); WAM 66.3.14–16, 67.10.177–182, skulls,
Firestick Cave, Nullarbor (WA); Nullarbor (WA).
Diagnosis. Perameles papillon sp. nov. differs from other species of Perameles in having a dark horizontal bar
on its ears, a bracelet of light brown fur around its wrists and covering the heel of the foot and the posterior edges
of the footpad, a dark butterfly pattern on its rump, shorter pes and tail (see results), enlarged lacrimal crest (except
for P. gunnii), enlarged bullae (largest of any Perameles), accessory fenestrae extending anteriorly to the canine,
maxillopalatine fenestrae extending posteriorly to the M3, well-developed squamosal epitympanic sinus.
Perameles papillon sp. nov. differs from all members of Peramelemorphia (except P. myosuros) in having females
distinctively larger than males.
Etymology. Papillon, French for butterfly, in reference to the butterfly shaped pattern on its rump.
Remarks. Westerman et al. (2012) subsampled specimen AM M31545 from the Australian Museum, which
was attributed to Perameles b. notina, from Ooldea in South Australia. After scoring specimens from this region,
including the rest of the Nullarbor Plain in Western Australia, and comparing them to the holotype of Perameles b.
notina (then named Perameles myosura notina) BMNH 43.8.12.21 from the head of St Vincent Gulf in South
Australia, it became evident that they did not match in morphology, the type being much larger and several features
on the skin and skull did not match any of the specimens recovered from the Nullarbor region. As a result, AM
M31545 and all specimens found in the Nullarbor regions (present in collections in various museums) are
considered here a new species.
The results of our analysis also show that Perameles papillon sp. nov. is distinct morphologically from
Perameles bougainville and P. eremiana, but also distinct to the currently unrecognised P. myosuros, P. notina and
P. fasciata. Our results further suggest that the latter three taxa are distinct taxa and should no longer be subsumed
in P. bougainville (see discussion below).
Distribution. Collected only in the Nullarbor Plain, on either side of the Western Australian and South
Australian border (triangles in Fig. 8).
Description. The description of the skin is a comparison of Perameles papillon sp. nov., represented by
Holotype WAM M571 (Fig. 5), to P. b. bougainville, represented by WAM M3641 (Suppl. Fig. 2), and P. b. notina,
represented by the holotype BMNH 43.8.12.21, unless specified otherwise (Suppl. Fig. 3). P. b. fasciata and P. b .
myosuros are not used in the comparison here as they were not sampled by Westerman et al. (2012), though we do
suspect they are distinct taxa and will be investigated in future research.
P. papillon sp. nov. is similar in size to P. b. bougainville but much smaller than P. b. notina, showing no
marked sexual dimorphism between males and females, though females seem to be a little larger (see Table S2–3
for measurements, Fig. 3 for morphometric analysis). The fur on the dorsal side of the head is a little darker and
TRAVOUILLON & PHILLIPS
14
·
Zootaxa 0000 (0) © 2018 Magnolia Press
FIGURE 5. Perameles papillon sp. nov., study skins. A–C, holotype WAM M571; D–F, paratype WAM M574; G–I, paratype
WAM M572. A, D, and G, dorsal view; B, E, and H, lateral view; C, F, and I, ventral view. Scale = 5cm.
Zootaxa 0000 (0) © 2018 Magnolia Press
·
15
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
FIGURE 6. Perameles papillon sp. nov., skulls. A–C and G–H, holotype WAM M571; D–F and I–J, paratype WAM M577. A,
D, G, I, dorsal view; B, E, ventral view; C, F, H, J, lateral view. Scale = 2cm.
browner than P. b. bougainville, but lighter than P. b. notina. The ventral side is cream in colour as in P. b.
bougainville whereas it is greyer in P. b. notina. The ears have a noticeable dark horizontal bar across them, absent
in both P. b. bougainville and P. b. notina or any other known peramelemorphian. Note that none of the wet
specimens of P. papillon sp. nov. show this feature (dark bar on ear) but it is present in all dry skins, and this is
suspected to be as a result of accelerated fading in alcohol specimens. The ears are long as is P. b. bougainville, but
shorter than P. b . no tina . There are long dark vibrissae present at the front of the snout, above the eye and on the
cheek below the eye. These are also present in P. b. bougainville and P. b. notina, but less obvious in P. b.
bougainville. The body is also darker and browner on the dorsal side than in P. b. bougainville but lighter than P. b.
notina. Both P. papillon sp. nov. and P. b. bougainville are cream coloured on the ventral side, while P. b. notina is
mostly grey except for patches of cream. A complex pattern of different coloured bars is present on the rump
(juveniles WAM M574 and M572 show the early stages of the development of this pattern on the rump), which is
absent in P. b. bougainville (P. b. bougainville has a single very faint darkish bar across its rump that tapers
laterally) but present in P. b. notina, though the holotype is faded, making this pattern less obvious. The pattern
resembles that of P. gunnii and P. b. notina but it is less well defined. There isn’t a central dark dorsal bar as in P.
gunnii, and most likely in P. b. notina, and the light bars are bicoloured, cream and golden brown (golden brown
only in P. gunnii and P. b. notina). The anterior most two dark bars are almost completely black or at least dark
brown (specimens are likely faded), and looks like butterfly wings in shape in dorsal view. The posterior darkest
bars are much lighter in colour and less well defined, distinguishing P. papillon sp. nov. from P. gunnii and P. b.
notina, which have a well-defined posterior bar. The tail is short as in P. b. bougainville (longer in P. b. notina), but
it is tricoloured (dark brown dorsally, beige laterally and cream ventrally) instead of bicoloured as in P. b.
bougainville (the tail is dark brown dorsally and cream ventrally) or unicoloured as in P. b. notina (the tail is dark
grey). The forelimbs are light brown in colour on the dorsal side up to the wrist, where it surrounds the wrist like a
TRAVOUILLON & PHILLIPS
16
·
Zootaxa 0000 (0) © 2018 Magnolia Press
bracelet. The manus and the ventral side of the forelimbs are cream in colour. This pattern is different to P. b.
bougainville, which lacks the bracelet of brown fur around the wrist, and is lighter in colour dorsally. The forelimbs
of P. b. notina are completely grey. The manus has the three large toes (digit II, III and IV), and the typically
reduced digits I and V as in all Peramelidae. Compared to P. b. bougainville and P. b. notina, the digits are shorter
but have more pronounced toepads as in P. b. notina (the toes of P. b. bougainville more slender). The hindlimbs, in
the region of the tibia, are cream in colour anteriorly and greyish brown on the posterior side. The darker colour
invades the heel of the cream coloured pes, and the posterior edges of the footpad, unlike in P. b. bougainville
where the pes is completely cream in colour and P. b. notina where the pes is completely grey in colour. The
footpad is hairless posterior to digit I to the heel as in P. b. notina, while in P. b. bougainville, this area is covered in
fur. As in the manus, the pes has shorter toes with thicker toepads than P. b. bougainville and P. b. notina.
The description of the skull and teeth is based on paratype WAM M577 for Perameles papillon sp. nov. (Fig.
6), and describes only features that differ from other Perameles, using WAM M6166 for P. b. bougainville, BMNH
43.8.12.21 for P. b. notina, WAM M2629 for P. eremiana, WAM M16590 for P. gunnii (Suppl. Fig. 4), QM
JM18575 for P. pallescens and WAM M44421 for P. nasu ta (see Travouillon 2016a), unless specified otherwise.
The skull of P. papillon sp. nov. is similar in size to P. b. bougainville and P. eremiana (much smaller than P. b .
notina, P. gunnii, P. pallescens and P. nasuta) but it is shorter than P. eremiana (nasals shorter in length), and much
wider than both P. b. bougainville and P. eremiana. In dorsal view, the lacrimal crest is highly developed, as seen in
P. gunnii and P. b. notina, and is visible as a distinct posteriorly orientated process in the orbit. The lacrimal crest is
less developed in all other species of Perameles. No sagittal crest is present, as in all Perameles except in males of
P. nasu ta and P. pallescens. In ventral and lateral view, the alisphenoid tympanic process, or bullae, is highly
distinctive in being the most developed (the second largest bullae) within the genus. P. b. bougainville has the
smallest bullae, followed by P. nasuta, P. gunnii, P. eremiana, P. papillon sp. nov. and P. b. notina, in increasing
order of size. On the palate, accessory fenestrae are present between the incisive fenestrae and the maxillopalatine
fenestrae as in all Perameles except P. nasuta and P. pallescens, but they extend anteriorly to the canine, unlike
other species of Perameles except P. b. notina where the anterior extension of the accessory fenestrae is posterior to
the canines. The maxillopalatine fenestrae also extend further posteriorly than in other Perameles except P. b.
notina, from the anterior of the M3 to level with the middle of M3, instead of level with M2. The dividing septa of
the maxillopalatine fenestrae are thick anteriorly and posteriorly, thinning centrally, as in Perameles nasuta (very
thin in all other Perameles). Palatine fenestrae are present and large as in P. b. bougainville (broken in P. b. notina).
In lateral view the orbitosphenoid is identified as a very reduced bone, as in P. nasuta, P. pallescens and P. gunnii
(much larger in P. b. bougainville and P. eremiana, broken in P. b. notina). The antorbital fossa is very deep as in all
Perameles except in P. nasu ta and P. pallescens. The squamosal epitympanic sinus of P. papillon sp. nov. is the
most developed of any Perameles (most visible in paratype WAM M126), being wide with a high posterior wall.
The I5 is pointed and canine-like as in P. b. bougainville, P. b. notina, P. nasuta and P. pallescens. The canines
are small with accessory cusps, in both males and females, as in P. b. bougainville, P. b. notina and P. erem i ana
(males of P. gunnii, P. nas u ta and P. pall esce ns have an enlarged canine, and males of P. nas u ta and P. pall escens
generally lack accessory cusps). The canines are entirely bordered by the maxilla as in all Perameles, except P.
eremiana (bordered by premaxilla anteriorly). The diastema between the canine and P1 is long (longer than P1) as
in all Perameles except P. nasu ta and P. pallescens. The P1 is as long as the P2 as in P. gunnii, P. nasuta and P.
pallescens (shorter in P. b. bougainville, longer in P. eremiana and P. b. notina), and the P2 is noticeably shorter
than the P3, as in P. b. notina, P. nasuta and P. pallescens (subequal in length in all other Perameles). The
development of the lingual shelf of P3 is small as in P. b. bougainville and P. eremiana. The major cusp of P3 is
similar in shape to all Perameles except P. gunnii (more conical). The upper molars of P. b. notina are completely
worn down and the holotype lacks a dentary and therefore these are not used in the comparison following. On the
M1, a short stylar crest is present anterior to the metastyle (best seen on juvenile WAM M572) with an associated
StE, while StE is a conical cusp with no associated crest in P. b. bougainville (best seen in juvenile WAM M16086).
StA is variable in size on M1–2. StB and StC are fused and indistinguishable as in P. b. bougainville and P.
eremiana. The preparacrista reconnects with the postparacrista posteriorly on the M1, as in P. eremiana. The
postprotocrista on M1 terminates posterior to the metacone as in P. b. bougainville and P. gunnii with no distinct
posterior cingulum present (terminates more buccally in all other Perameles, as a posterior cingulum). On the M2,
StB is conical as in all Perameles except P. gunnii (oval in shape). No stylar crest is present posterior to the conical
StD as seen in P. gunnii and P. b. bougainville (small crest present in all other taxa). The postprotocrista ends
Zootaxa 0000 (0) © 2018 Magnolia Press
·
17
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
posterior to the metacone as in all Perameles, except P. n asu ta. StE is present as a distinct large cusp as in P. gunnii
(it is reduced in all other species). On the M3, the postprotocrista ends posterior to the metacone as in all
Perameles, except P. nas uta . The metaconule is small with only a small shelf separating it from the lingual flank of
the metacone, as in P. b. bougainville and P. pallescens (a larger metaconule and shelf is present in all other taxa).
FIGURE 7. Perameles papillon sp. nov., upper and lower dentition of paratype WAM M577. A–B, upper premolas; C–D,
upper molars; E–F, lower premolars; G–H, lower molars. A, C, E and G, occlusal view; B, D, F and H, lateral view. Scale =
2mm.
TRAVOUILLON & PHILLIPS
18
·
Zootaxa 0000 (0) © 2018 Magnolia Press
StE is absent as in all Perameles except P. gunnii. No anterior cingulum present as in all Perameles except for some
specimens of P. pallescens and P. nasuta. On the M4, StB is variable, sometimes absent, sometimes present as a
small cusp. This feature is variable across Perameles. The postparacrista is curved, forming a small centrocrista, as
in all Perameles, except P. gunnii and P. b. bougainville, where it is straight. The postprotocrista ends anterior to
the posterior end of the postparacrista as in all Perameles except P. nasut a. The metacone is small as in all
Perameles except P. gunnii and P. nasuta, which have an enlarged metacone. StC/D is absent (sometimes present in
P. gunnii and P. n asu ta ).
The lower canine is small and has accessory cuspids as in all Perameles except in males of P. nasuta and P.
pallescens, which have enlarged unicuspid canines. The diastema between the canine and p1 is shorter than the
length of p1, as in P. pallescens (longer that p1 length in all other taxa). The p1 is as long as the p2 as in P. b.
bougainville and P. eremiana (it is shorter in all other taxa). The p2 is longer than the p3 as in P. b. bougainville and
P. eremiana (subequal in all other taxa). No cuspid is present on the hypoflexid of the lower molars as in all
Perameles except P. nasuta and P. eremiana. On the m1, the paraconid-metaconid distance is equal to the
metaconid-protoconid distance as in all Perameles except P. pallescens, which has a longer paraconid-metaconid
distance. The posthypocristid is oblique to the tooth row and connects to the hypoconulid as in all Perameles
(sometimes perpendicular in P. b. bougainville and connects to entoconid). On the m2, the posthypocristid is also
oblique and connects to a large hypoconulid (this feature varies across Perameles). On the m3, the hypoconid is
buccal to the protoconid as in all Perameles except P. b. bougainville and some specimens of P. pallescens, which
have the two cuspid level with one another. The posthypocristid connects to the entoconid as in all Perameles
except some specimens of P. nasuta and P. pallescens, and the hypoconid is either absent or reduced as in P. b .
bougainville and P. eremiana. On the m4, the talonid is reduced but not as reduced as in P. pallescens, P. b.
bougainville and P. eremiana. The posthypocristid is perpendicular to the tooth row as in all Perameles except P.
gunnii. The entoconid is large as in all Perameles, except P. b. bougainville and P. eremiana. The buccal shelf ends
at the buccal side of the hypoconid as in P. b. bougainville and P. eremiana (buccally reduced in other taxa).
Family Peroryctidae Groves & Flannery 1990
Genus Peroryctes Thomas 1906
Peroryctes aruensis Kear et al. 2016 new comb.
Lemdubuoryctes aruensis Kear et al. 2016
Holotype. WAM 14.9.6, left maxilla with P3, M1–2.
Referred material. WAM 14.9.1–14.9.5 and WAM 14.9.7–14.9.20, WAM 14.9.51–14.9.53.
Type locality. Liang Lemdubu cave, Pulau Kobroor, Aru Islands group, Eastern Indonesia.
Remarks. Kear et al. (2016) described this taxon, and assigned it to a new genus based on the result of their
phylogenetic analysis. This taxon is only represented by heavily worn specimens, and so Kear et al. (2016) scored
worn teeth in a matrix that was designed to be scored only on unworn teeth. Their principal character for placing
this taxon in a new genus is the presence of a centrocrista in M1–3. This is only seen in unworn Oligo-Miocene
bandicoot teeth, but almost every single peroryctid (Peroryctes, Echymipera, Rhnynchomeles and Microperoryctes)
do wear their teeth to the point that a centrocrista is formed on M1–3 as a result of wear. The holotype of
Peroryctes broadbenti illustrates this morphology, being an adult with worn teeth, and centrocristae are present on
M1–3. This does not happen in Peramelidae, Chaeropodidae or Thylacomyidae. As a result, the assessment of this
taxon is dubious, and was rescored here, and re-analysed. All characters that can be scored for L. aruensis were
identical to that of Peroryctes broadbenti. This does not necessarily suggest that they are the same species. Our
analysis suggests that it belongs to the genus Peroryctes, and so we reassigned it here, and provide a new diagnosis,
based on features that can be seen on those worn specimens and the results of the morphometric analysis.
New diagnosis. Peroryctes aruensis differs from Peroryctes raffrayana in having dimorphic upper and lower
third premolars. This feature only occurs in Peroryctes broadbenti, and some Echymipera (most noticeable in E.
clara). The third premolars are as long and wide as in P. broadbenti, which are narrower but longer than E. clara. It
Zootaxa 0000 (0) © 2018 Magnolia Press
·
19
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
differs from P. raffrayana in having much larger teeth. It differs from P. broadbenti in having a shorter and less
wide M4, having a reduced protocone; longer P2 and longer p1.
FIGURE 8. Distribution of species in the Perameles bougainville complex, based on museum specimens (modern and some
fossils). Circles, Perameles bougainville; squares, Perameles eremiana; crescents, Perameles myosuros; ovals, Perameles
notina; star, Perameles fasciata; triangles, Perameles papillon sp. nov.
Discussion
Our favoured total evidence phylogenetic inference (Fig. 1E) strongly supports both the predominantly Australian
Peramelinae and the predominantly New Guinean Peroryctinae-Echymiperinae. Together these form a near-
trichotomy with the two remaining Australian genera, Macrotis and Chaeropus. All of the included Oligo-Miocene
bandicoots are strongly excluded from crown Peramelemorphia, with the possible exception of Madju variae.
Bayesian inference analyses that were unrooted (non-clock) or tip-dated with linked morphological and
molecular clocks both resulted in all of the Oligo-Miocene bandicoots grouping together. Unexpected monophyly
of groups of fossil taxa appears to be common in total evidence dating, and may in part be a by-product of conflict
between molecular and fossil temporal signals, as was suggested by Ronquist et al. (2016) for apparently
associated divergence timing artefacts. But in this case, Ronquist et al.s (2016) remedy for timing (and some fossil
monophyly) concerns, of varying the fossilized birth-death models, for example to reduce the extinction
TRAVOUILLON & PHILLIPS
20
·
Zootaxa 0000 (0) © 2018 Magnolia Press
probability (100 fold in this case) had little effect. However, unlinking the molecular and morphological clock
variance parameters provided agreement with MP analysis and restored Yarala, Galadi and Bulungu to basal
paraphyly outside crown bandicoots, and Madju as close to the base of modern bandicoots. This unlinking reduces
the extent to which the molecular and morphological clocks constrain each other. Our unlinked clocks result is one
of the few examples that run counter to Donoghue & Yang (2016) suggestion that unrooted trees should generally
be favoured over clock-based trees. In the present case a strong temporal signal that enables the dated analysis to
overcome the apparent symplesiomorphy that tends to draw the Oligo-Miocene bandicoots towards the
morphologically more “primative” New Guinean bandicoots.
We were aware of the potential for morphological convergence between Macrotis and Isoodon, particularly
associated with adaptation to more open habitats. However, the stochastic influence of higher rates of
morphological evolution (resulting in long-branch attraction, LBA) might also be contributing to drawing these
two genera together. The phylogeny in Fig. 1E was inferred from both molecular and morphological data, but
shows the branch lengths from the morphology partition, and among these, Isoodon and Macrotis are by far the
longest. Likelihood models (here, within BI) can account for LBA when incorporating sufficient rate variation
among characters. In this regard it is notable that Isoodon/Macrotis is one of three clades that the non-clock
Bayesian analysis makes a substantial correction against, as shown by low BI support relative to MP support
(Supp. Fig. 5).
The results of our Total Evidence Analysis have highlighted several points of interest. The first is the clade
containing Peroryctinae and Echymiperinae. Groves & Flannery (1990) revised the families within
Peramelemorphia, and proposed the family Peroryctidae based on eight morphological characters distinguishing its
members (Peroryctes, Microperoryctes, Echymipera and Rhynchomeles) from Peramelidae (Isoodon and
Perameles). This family was later rejected as a result of a genetic study by Westerman et al. (1999), based on a
single gene (12S rRNA) which recovered members of Peroryctidae as a paraphyly. However, subsequent genetic
studies using multiple genes (e.g. Meredith et al. 2008; Westerman et al. 2012; Mitchell et al. 2014; Kear et al.
2016) and morphological studies (e.g. Travouillon et al. 2010; Travouillon et al. 2013a; Travouillon et al. 2013b;
Gurovich et al. 2014; Travouillon et al. 2014; Chamberlain et al. 2015; Travouillon et al. 2015) have recovered
Peroryctidae as a monophyletic clade sister to the clade containing Isoodon and Perameles. Our analyses also
support this topology, and as a result, we suggest that the family Peroryctidae should be re-instated to reflect the
current phylogenies. This means that there would be a total of five families within Peramelemorphia: Yaralidae
(including Yarala), Thylacomyidae (including Macrotis, Ischnodon and Liyamayi), Chaeropodidae (including
Chaeropus), Peramelidae (including Isoodon, Perameles and Crash) and Peroryctidae (including Peroryctes,
Microperoryctes, Echymipera and Rhynchomeles). We have identified additional characters, on top of the eight
characters given by Groves & Flannery (1990), which unites members of Peroryctidae (to the exclusion of
Peramelidae): presence of an anteriorly directed preentocristid; entocristid never conical; lacrimal crest absent or
poorly developed; thick septa dividing maxillopalatine fenestrae; poor development of ectotympanic, alisphenoid
tympanic process and rostral tympanic process of the petrosal; poorly defined epitympanic recess; squamosal
epitympanic sinus is absent; poor development of metaconule on M1–2; and no hypoflexid on m1.
The second point of interest is the position of Perameles bowensis which oscillates between forming a clade
with Perameles or Isoodon. In earlier morphological studies, its position remained elusive, as it was often
recovered as a polytomy within Perameloidea (e.g. Travouillon et al. 2015) or as a basal member of Peramelinae
(e.g. Travouillon et al. 2013b; Travouillon et al. 2014; Chamberlain et al. 2015). Travouillon et al. (2017)
recovered P. bowensis as sister to Isoodon, and suggested that it is possible that this taxon and others (e.g. P.
allinghamensis) may need to be reassigned to a new genus, as it is likely to be ancestral to either or both Perameles
and Isoodon. Here, our results would support this conclusion that P. bowe nsis probably does not belong to the
genus Perameles. We therefore caution against the common practice of using P. bowensis as a minimum bound for
calibration the divergence between Isoodon and Perameles, although the affinities of this extinct bandicoot do
appear to be close to the most recent common ancestor of Isoodon and Perameles.
The Quenda, or Southern Brown Bandicoot from Western Australia (Isoodon obesulus fusciventer) does not
seem to form a monophyletic clade with its eastern counterpart I. o. obesulus. This paraphyly of the species I.
obesulus has also been recovered by We sterman et al. (2012), but with a slightly different topology. Westerman et
al. (2012) recovered I. o. fusciventer as sister to all Isoodon while in our analyses it tends to be sister to I. auratus.
This result is similar to that of Pope et al. (2001) and Zenger et al. (2005) who recovered I. obesulus fusciventer
Zootaxa 0000 (0) © 2018 Magnolia Press
·
21
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
and Isoodon auratus as a clade sister to I. o. obesulus. Contra Pope et al. (2001) and Zenger et al. (2005), we
propose here to raise I. o. fusciventer to species level as it does not form a monophyletic group with other I.
obesulus, rather than sinking I. auratus within I. obesulus. Morphologically, I. o. fusciventer has a unique set of
characters (see diagnosis for list of characters) that separate it from other species of Isoodon. While it seems to
share more characters with I. auratus spp., its large size and pelage colour are closer to I. obesulus. It also shares
the enlarged metacone on the M4 with I. macrourus, which is a key feature separating I. macrourus from I.
obesulus. I. o. fusciventer seems to have diverged from I. auratus around 1.5 (0.8–2.5) Ma, not long after the
general diversification of Isoodon taxa 3.1 (1.8–4.8) Ma. As a result of the elevation of I. o. fusciventer to full
species, its vernacular name needs to be reconsidered. It can no longer be referred to as a Southern Brown
Bandicoot, and therefore the term ‘Quenda’ is more appropriate, and is already widely used (Woinarsky et al.
2014).
Kear et al. (2016) described Lemdubuoryctes aruensis, a fossil species from the Holocene of Aru Island. In
their total evidence analysis, they recovered L. aruensis as a stem taxon, amongst Miocene taxa such as Yarala,
Bulungu and Galadi, outside of the crown Peramelemorphia. This is largely the result of the presence of a
’complete’ centrocrista with continuous postparcirstae-premetacristae on all upper molars on L. aruensis as seen in
Yar al a and Bulungu muirheadae. However, there is one major flaw in this assessment of the morphology of these
taxa and the way that Kear et al. (2016) scored these characters. The morphological matrix, first designed by
Travouillon et al. (2010), requires that characters are scored on unworn dentition. This has been followed in all
other subsequent publications of the dataset (Travouillon et al. 2013a; Travouillon et al. 2013b; Gurovich et al.
2014; Travouillon et al. 2014; Chamberlain et al. 2015; Travouillon et al. 2015), where only unworn teeth were
scored, as many characters are erased or reconfigured as a result of wear. The complete centrocrista is indeed seen
in unworn specimens only of Yarala, Bulungu and Galadi, but all specimens presented by Kear et al. (2016) are
heavily worn (as noted in their description), which makes the comparison ambiguous. While unworn upper molars
of Peroryctes and Echymipera do not form a complete centrocrista, when worn, they do form a centrocrista as a
result of wear (see Supp. Fig. 6). This suggests that scoring worn specimens is misleading if one is to understand
the evolution of this feature through time. This has significant ramification for the results obtained by Kear et al.
(2016) which rely on their scoring of the centrocrista and is the focus of their diagnosis. In our study, we have re-
scored L. aruensis taking into account dental wear, and therefore scored worn features as a question mark. This has
resulted in a completely different relationship for both L. aruensis and the stem taxa, with L. aruensis found to be
the sister taxon to P. broadbenti, separating only 0.6 (0.0–1.8) Ma. This is a much more likely scenario, considering
the amount of overlap in morphology between these two taxa (see diagnosis), especially the sexually dimorphic P3/
p3 (enlarged in males, not enlarged in females; see results), which is unique to P. broadbenti (Echymipera clara has
also a sexually dimorphic P3/p3 but these teeth are shorter and wider than those of P. broadbenti and L. aruensis).
This feature has never been recorded in any stem peramelemorphian or any ‘Australian’ peramelemorphian.
Considering the biogeographical history of Aru Island, it was connected to mainland New Guinea (and by
extension to Australia) during the last glacial maximum (and several times before that), with several ‘Australian
taxa’ (e.g. Isoodon macrourus) present in the same fossil locality as L. aruensis (O’Connor et al. 2005a, 2005b). A
close relationship between P. broadbenti and L. aruensis is therefore not surprising, and could have formed a single
uniform population before the isolation of Aru Island since the last glacial maximum. As a result, we have
reassigned L. aruensis to the genus Peroryctes. Note that Kear et al.s (2016) other conclusions on early origins of
xeric-adapted peramelemorphians rely on their scoring of L. aruensis, and therefore these conclusions are
questionable. Our results here show the reverse, with mesic-adapted stem taxa giving rise to xeric-adapted taxa.
This repeats a similar temporal/ecological pattern inferred for habitat transitions among kangaroos (Dodt et al. In
Press), and is further supported by the recent discovery of a Late Pliocene Chaeropus, which showed a late but
rapid evolution of herbivory in this genus (Travouillon 2016b).
With the exclusion of the most incomplete fossil taxa and placement of L. aruensis within Peroryctes, our total
evidence “tip-dates” are far younger than Kear et al.’s (2016) tip dates, and instead closely agree with our node-
dating analysis (Table 1) and that of Westerman et al. (2012). Adding node calibrations for modern clades had very
little effect on date estimates. However, the further inclusion of a node calibration for the divergence of the Oligo-
Miocene Bulungu from crown bandicoots did have an influence, increasing the estimate for the age of origin of all
included stem and crown bandicoots from 29.8 (23.9–37.1) Ma to 32.4 (27.0–39.1) Ma. Date estimates within the
crown group were barely affected, although precision was improved in some cases (Table 1). As far as we are
TRAVOUILLON & PHILLIPS
22
·
Zootaxa 0000 (0) © 2018 Magnolia Press
aware this is the first example of total evidence dating taking advantage of fossil inclusion to calibrate internal
nodes that do not appear on the molecular tree.
The Western Barred-Bandicoot, Perameles bougainville (Quoy & Gaymard 1824), has been puzzling
taxonomists for many decades. With the mainland population completely extinct, it has been difficult to establish
whether it represents a complex of species or not (Friend 2008; Warburton & Travouillon 2016). There are
currently several taxa synonymised within this species, including P. myosuros (Wagner 1841), P. arenaria (Gould
1844), P. notina (Thomas 1922) and P. fasciata (Gray 1841). Tate (1948) was the first to re-examine this complex
of species, but he only examined the types of P. notina and P. fasciata and a specimen of Perameles b. bougainville
(total sample size = 3 specimens). He did not see any other specimens, yet he came to opposite conclusions in his
study; one conclusion being that they all belonged to the same species with regional variation, and the other that
Perameles b. bougainville was distinct to P. notina and P. fasciata. Later on, Freedman (1967) and Freedman &
Joffe (1967) re-examined this complex of species, yet they did not examine any type material. Instead, they
compared specimens of Perameles b. bougainville with a mix of specimens from the Nullarbor (which they called
P. notina) and from central Australia (which they seem to be aware belonged to P. eremiana). No specimens from
the south west of Western Australia were included (P. myosuros/P. arenaria) and all specimens east of the
Nullarbor were dismissed as being P. gunnii (despite the types of P. no tina and P. fasciata occurring there). Their
analysis showed that Perameles b. bougainville was distinct from the Nullarbor/Central Australia population. The
results of this study gave rise to the current understanding of the species P. bougainville, having two subspecies,
Perameles b. bougainville and Perameles b. fasciata.
West erman et al. (2012) provided the first molecular study to include molecular data for a mainland P.
bougainville. They sampled one of the specimens from the Nullarbor, which they named P. b. notina, probably a
name associated with the studies of Freedman (1967) and Freedman & Joffe (1967), though there has never been
any study showing that the Nullarbor population was indeed belonging to P. n ot ina . In their study, Westerman et al.
(2012) recovered P. b bougainville as sister to P. b. notina, but with a separation date of over 3 Ma. This already
suggests that these two taxa might be separate species. Our results presented here show a slightly different
relationship, with P. b. notina more closely related to the Desert Bandicoot, P. e re mia na, which diverged around 3.5
(1.8–5.2) Ma. As a result, we consider the Nullarbor population as a distinct species, but also, upon comparison
with all the types within the P. bougainville complex, we found that it was morphologically distinct from P. n ot ina ,
and therefore represented a new unnamed species (Perameles papillon sp. nov.). The results of our morphometric
analysis also suggests that each member of the P. bougainville complex may represent a distinct morphology,
except for P. arenaria, which we consider a synonym of P. myosuros (morphologically identical and geographically
close).
While we are not revising P. myosuros, P. notina or P. fasciata in this study, as we are awaiting to recover some
DNA sequences first, our results do show that they are sufficiently distinct (little to no overlap in size and
morphology) to warrant being re-elevated to full species. The distribution of each taxon based on museum
specimens, shown in Figure 8, also tells an important story. Perameles papillon sp. nov. has been well-sampled
through the Nullarbor, through both modern and subfossil specimens seem to be restricted to those plains. P.
myosuros occurred in the Wheat belt of WA (mostly woodland), and has some records from the southern edge of
the Nullarbor caves, all associated with a thin strip of woodland. Perameles papillon sp. nov. and P. myos uros are
the only two taxa in the P. bougainville complex to be found sympatrically. There are huge gaps in the distribution
of the P. bougainville complex taxa (Fig. 8). Further examination of subfossils will be required to better understand
the past distribution of these taxa (Warburton & Travouillon 2016). Despite this current lack of knowledge, this
distribution as an important impact for the conservation of P. bougainville. Current conservation effort has been in
trying to reintroduce this species within its past distribution. With the discovery of a new species, P. papillon sp.
nov., and better understanding of taxonomy, it is becoming apparent that the living P. bougainville, from Bernie
and Dorre Island, has never had any records outside of Western Australia. In fact other than on these islands, only
two specimens on the mainland can be confidently attributed to P. bougainville, the holotype on Peron Peninsula,
and a specimen recovered from Onslow.
So is it really wise to translocate individuals of P. bougainville outside of Western Australia, given the current
knowledge of its distribution, and not knowing how well the ecology P. bougainville matches the ecology of now
extinct species that occurred outside of Western Australia.
The biology of the members of the P. bougainville complex is quite interesting. Both P. papillon sp. nov. and P.
Zootaxa 0000 (0) © 2018 Magnolia Press
·
23
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
myosuros have females much larger than males, while P. fasciata and P. notina have larger males, and P.
bougainville seems to have males and females similar in size (Table 2). Larger females than males are rare in
mammals (found in some bats, primates, rodents, etc.), but it is not unique to these two taxa, though it is the first
occurrence within Peramelemorphia (Ralls 1976; some dasyurids, phalangerids and vombatids are listed in that
paper as having larger females, but this is not supported by later research, summarised in Van Dyck & Strahan
2008, which only lists Lagorchestes hirutus and Lagostrophus fasciatus as having larger females). In humans, the
males are on average 1.07 times larger (height) than females (Ralls 1976). This is the ratio found in P. eremiana, P.
fasciata and P. notina (though the sample size was small), but we found the opposite for P. papillon sp. nov. and P.
myosuros. While many hypotheses have been proposed for this phenomenon, Ralls (1976) suggested that more
intense competition among females for resource than among males might be the cause. Though, more than one
selection could be involved in affecting the size of the females (including decreasing pressure in large male
selection). In arid and semi-arid environments, where P. myosuros and P. papillon sp. nov. occur, competition for
resources is likely to play an important part of the biology of these animals, and perhaps explain why females
became larger. The size and shape of the bullae, which was also noted by Freedman & Joffe (1967) as being an
important difference, also separate all taxa within the P. bougainville complex, with P. bougainville having the
smallest and roundest, and P. papillon sp. nov. having the largest relative to its size. While the function of the
bullae is not completely understood, most studies that have looked into its function agree that they do play an
important role as part of the auditory system. Keen & Grobbellar (1940) suggested that enlarged bullae are an
adaptation for the better transmission of air-conducted sounds. Webster’s (1962) experiments on kangaroo rats
concluded that their function was to dampen the tympano-ossicular system to allow resonance phenomena to occur,
which increases the ear’s sensitivity to resonant frequencies and as a result enable to better detect predators. Lay
(1972) noted that the hypertrophied bullae was more common in arid adapted taxa and that this is probably because
sound reception is more difficult in deserts. Many carnivores also have hypertrophied bullae, which are likely to
help with prey detection (Hunt 1974). There are very likely to be differences in hearing ability between members of
the P. bougainville species complex, which would need to be taken into account in conservation programs,
especially when P. bougainville from Dorre and Bernie Island, which has the smallest bullae, is introduced into the
distribution of other Perameles, which had very large bullae.
In regards to the vernacular name of the P. bougainville complex, as mentioned by Travouillon (2016a), the
name ‘Western Barred Bandicoot’ is less than adequate for P. bougainville, considering it lacks clear barring on its
rump. Friend (2008) provides a list of other possible common names, including ‘Barred Bandicoot’, ‘Striped
Bandicoot’, ‘Zebra Rat’, ‘Marl’ or ‘Nyemmel’, the last two derived from the indigenous names (‘Mal’ and ‘Nymal’
in Woinarski et al. 2014). Iredale & Troughton (1934) called P. bougainville the ‘Little Marl’, and P. myosuros the
‘Marl’, which seem more suitable since neither taxon is barred. They named P. fasciata the ‘Eastern Barred
Bandicoot’ but this name is currently used to refer to P. gunnii (which they named ‘Tasmanian Barred Bandicoot’).
Friend (2008)’s ‘New South Wales Striped Bandicoot’ most likely refers to P. fasciata and ‘South Australian
Striped Bandicoot’ to P. notina, both of which would be suitable and unambiguous. For P. papillon sp. nov., we
propose either the ‘Nullarbor Barred Bandicoot’ or ‘Butterfly Bandicoot’ (direct translation of its species name).
Acknowledgments
Many thanks to the Winston Churchill Memorial Trust for providing K. Travouillon the opportunity to undertake
this project and travel around the world to achieve this goal. Thanks to the sponsor of my Churchill Fellowship, the
Australian Biological Resources Study (ABRS) for providing the ideal avenue to undertake taxonomic projects.
Many thanks to the WA Museum for letting K. Travouillon undertake this project and support throughout the
journey. Thanks also to the Australian Research Council for support (DP150104659 to M. Phillips). Thanks to
Sandy Ingleby and Anja Divljan at the AM in Sydney, David Stemmer at the SAM in Adelaide, Kevin Rowe and
Karen Roberts at Museum Victoria in Melbourne, Gavin Dally at the MAGNT in Darwin, Roberto Portela Miguez
and Pip Brewer at the NHM in London, Géraldine Véron, Cécile Callou and Aurélie Verguin at the MNHN in
Paris, Giuliano Doria at the MSN in Genoa, Anneke H. van Heteren at ZSM in Munich, Eileen Westwig and
Eleanor Hoeger at AMNH in New York, Darrin Lunde at the Smithsonian in Washington DC, for collection access
and allowing this research. Many thanks to Alex Baynes for valuable discussions that have helped improved this
research.
TRAVOUILLON & PHILLIPS
24
·
Zootaxa 0000 (0) © 2018 Magnolia Press
References
Amrine-Madsen, H.A., Scally, M., Westerman, M., Stanhope, M.J., Krajewski, C. & Springer, M.S. (2003) Nuclear gene
sequences provide evidence for the monophyly of australidelphian marsupials. Molecular Phylogenetics and Evolution,
28, 186–196. https://doi.org/10.1016/S1055-7903(03)00122-2
Aplin, K.P. & Archer, M. (1987) Recent advances in marsupial systematics with a new syncretic classification. In: Archer, M.
(Ed.), Possums and Opossums, Studies in Evolution. Surrey Beatty and Sons, Sydney, pp. xv–lxxii.
Baker, M.L., Wares, J.P., Harrison, G.A. & Miller, R.D. (2004) Relationships among the families and orders of marsupials and
the major mammalian lineages based on recombination activating gene-1. Journal of Mammalian Evolution, 11, 1–16.
https://doi.org/10.1023/B:JOMM.0000029143.39776.ec
Baverstock, P.R., Flannery, T., Aplin, K., Birrell, J. & Krieg, M. (1990) Albumin immunologic relationships of the bandicoots
(Perameloidea: Marsupialia)—A preliminary report. In: Seebeck, J.H., Brown, P.R., Wallis, R.L. & Kemper, C.M. (Eds.),
Bandicoots and Bilbies. Surrey Beatty and Sons, Chipping Norton, N.S.W, pp. 13–18.
Beck, R.M.D. (2008) A dated phylogeny of marsupials using a molecular supermatrix and multiple fossil constraints. Journal
of Mammalogy, 89, 175–189.
https://doi.org/10.1644/06-MAMM-A-437.1
Black, K.H., Archer, M., Hand, S.J. & Godthelp, H. (2012) The rise of Australian marsupials: a synopsis of biostratigraphic,
phylogenetic, palaeoecologic and palaeobiogeographic understanding. In: Talent, J.A. (Ed.), Earth and Life. Springer,
Dordrecht, pp. 983–1078.
Chamberlain, P.M., Travouillon, K.J., Archer, M. & Hand, S.J. (2015) Kutjamarcoot brevirostrum gen. et sp. nov., a new short-
snouted, early Miocene bandicoot (Marsupialia: Peramelemorphia) from the Kutjamarpu Local Fauna (Wipajiri
Formation) in South Australia. Alcheringa: An Australasian Journal of Palaeontology, 40, 1–10.
https://doi.org/10. 1080/03115518.2016.1103525
Desmarest, A.G. (1817) Nouveau Dictionnaire d’Histoire Naturelle, appliquée aux arts, a l’agriculture, a l’économie rurale et
domestique, a la médecine, etc. Par une société de naturalistes et d’agriculteurs. Nouvelle édition presqu’entièrement
refondue et considérablement augmentée. Chez Deterville, Paris, 595pp.
Donoghue, P.C.J. & Yang, Z. (2016) The evolution of methods for establishing evolutionary timescales. Philosophical
Transactions of The Royal Society B Biological Sciences, 371, 20160020.
https://doi.org/10.1098/rstb.2016.0020
Dodt, W.G., Gallus, S., Phillips, M.J. & Nilsson, M.A. (2018) Resolving kangaroo phylogeny and overcoming retrotransposon
ascertainment bias. Scientific Reports. [in press]
Freedman, L. (1967) Skull and tooth variation in the genus Perameles. Part I: anatomical features. Records of the Australian
Museum, 27, 147–165.
Freedman, L. & Joffe, A.D. (1967) Skull and tooth variation in the genus Perameles. Part 3: Metrical features of P. gu nn ii and
P. bougainville. Records of the Australian Museum, 27, 197–212.
Friend, J.A. (2008) Western barred bandicoot, Perameles bougainville (Quoy and Gaimard, 1824). In: Van Dyck, S. & Strahan,
R. (Eds.), The Mammals of Australia. Reed Holland Publishers, Sydney, pp. 182–184.
Gallus, S., Janke, A., Kumar, V. & Nilsson, M.A. (2015) Disentangling the relationship of the Australian marsupial orders
using retrotransposon and evolutionary network analyses. Genome Biology and Evolution, 7, 985–992.
Geoffroy, E. (1803) Note sur les genres Phascolomis et Perameles, nouveaux genres d’animaux a bourse. Bulletin des Sciences,
par la Société Philomathique de Paris, 80, 149–150.
Godthelp, H., Archer, M., Cifelli, R., Hand, S.J. & Gilkeson, C.F. (1992) Earliest known Australian Tertiary mammal fauna.
Nature, 356, 514–516.
Gould, J. (1844) Exhibition and character of a number of animals, &c. transmitted from Australia by Mr. Gilbert. Proceedings
of the Zoological Society of London, 12, 103–107.
Gray, J.E. (1825) Outline of an attempt at the disposition of the Mammalia into tribes and families with a list of the genera
apparently appertaining to each tribe. Annals of Philosophy (New Series), 10, 337–344.
Gray, J.E. (1841) Contributions towards the geographical distribution of the Mammalia of Australia, with notes on some
recently discovered species, in a letter addressed to the Author. Appendix C. In: Grey, G. (Ed.), Journals of Two
Expeditions of Discovery in North-west and Western Australia, During the Years 1837, 38 and 39, Under the Authority of
Her Majesty’s Government. Describing many newly discovered, important, and fertile districts, with observations on the
moral and physical condition of the aboriginal inhabitants, &c., &c. T. & W. Boone, London, pp. 397–414.
Groves, C.P. & Flannery, T.F. (1990) Revision of the families and genera of bandicoots. In: Seebeck, J.H., Wallis, R.L., Brown,
P.R. & Kemper, C.M. (Eds.), Bandicoots and Bilbies. Surrey Beatty, Sydney, pp. 1–11.
Gurovich, Y., Travouillon, K.J., Beck, R.M.D., Muirhead, J. & Archer, M. (2014) Biogeographical implications of a new
mouse-sized fossil bandicoot (Marsupialia: Peramelemorphia) occupying a dasyurid-like ecological niche across
Australia. Journal of Systematic Palaeontology, 12, 265–290.
https://doi.org/10.1080/14772019.2013.776646
Hunt, R.M. (1974) The auditory bulla in carnivora: An anatomical basis for reappraisal of carnivore evolution. Journal of
Morphology, 143, 21–75. https://doi.org/10.1002/jmor.1051430103
Iredale, T. & Troughton, E.G. (1934) A check-list of the mammals recorded from Australia. Memoirs of the Australian
Zootaxa 0000 (0) © 2018 Magnolia Press
·
25
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
Museum, 6, i–xii 1–122.
Kealy, S. (2013) The Phylogeny of Australia’s Carnivorous Marsupials (Dasyuromorphia): implications for their biogeography
and diversity through time. Unpublished Honours Thesis, University of New South Wales, Sydney, 103 pp.
Kear, B.P., Aplin, K.P. & Westerman, M. (2016) Bandicoot fossils and DNA elucidate lineage antiquity amongst xeric-adapted.
Scientific Reports, 6, 37537. https://doi.org/10.1038/srep37537
Keen, J.A. & Grobbelaar, C.S. (1940) The comparative anatomy of the tympanic bulla and auditory ossicles, with a note
suggesting their function. Transactions of the Royal Society of South Africa, 28, 307–329.
Kirsch, J.A.W. (1968) Prodromus of the comparative serology of Marsupialia. Nature, 217, 418–420.
Larkin M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam, H., Valentin, F., Wallace, I.M., Wilm,
A., Lopez, R., Thompson, J.D., Gibson, T.J. & Higgins, D.G. (2007) Clustal W and Clustal X version 2.0. Bioinformatics,
23, 2947–2948.
Lay, D.M. (1972) The anatomy, physiology, functional significance and evolution of specialized hearing organs of gerbilline
rodents. Journal of Morphology, 138, 41–120. https://doi.org/10.1002/jmor.1051380103
Mackness, B.S., Wroe, S., Muirhead, J., Wilkinson, C. & Wilkinson, D. (2000) First fossil bandicoot from the Pliocene
Chinchilla local fauna. Australian Mammalogy, 22, 133–136.
Meredith, R.W., Westerman, M. & Springer, M.S. (2008) A timescale and phylogeny for “Bandicoots” (Peramelemorphia:
Marsupialia) based on sequences for five nuclear genes. Molecular Phylogenetics and Evolution, 47, 1–29. https://doi.org/
10.1016/j.ympev.2008.01.002
Meredith, R.W., Krajewski, C., Westerman, M. & Springer, M.S. (2009) Relationships and divergence times among the orders
and families of Marsupialia. Museum of Northern Arizona Bulletin, 65, 383–406.
Mitchell, K.J., Pratt, R.C., Watson, L.N., Gibb, G.C., Llamas, B., Kasper, M., Edson, J., Hopwood, B., Male, D. & Armstrong,
K.N. (2014) Molecular phylogeny, biogeography, and habitat preference evolution of marsupials. Molecular Biology and
Evolution, 31, 2322–2330. https://doi.org/10.1093/molbev/msu176
Muirhead, J. (2000) Yaraloidea (Marsupialia, Peramelemorphia), a new superfamily of marsupial and a description and analysis
of the cranium of the Miocene Yarala burchfieldi. Journal of Paleontology, 71, 521–523.
Muirhead, J. & Wroe, S. (1998) A new genus and species, Badjcinus turnbulli (Thylacinidae: Marsupialia), from the late
Oligocene of Riversleigh, northern Australia, and an investigation of thylacinid phylogeny. Journal of Vertebrate
Paleontology, 18, 612–626.
Nilsson, M.A., Arnason, U., Spencer, P.B.S. & Janke, A. (2004) Marsupial relationships and a timeline for marsupial radiation
in South Gondwana. Gene, 340, 189–196. https://doi.org/10.1016/j.gene.2004.07.040
O’Connor, S., Aplin, K., Pasveer, J. & Hope, G. (2005a) Liang Nabulei Lisa, a late Pleistocene and Holocene sequence from the
Am Islands. In: O’Connor, S., Spriggs, M. & Veth, P. (Eds), Terra Australis. Vol. 23. The Archaeology of the Aru Islands,
Eastern Indonesia. Pandanus Books, Canberra, pp. 125–162.
O’Connor, S., Aplin, K., Szabó, K., Pasveer, J.M., Veth, P. & Spriggs, M. (2005b) Liang Lemdubu, a Pleistocene cave site in
the Am Islands. In: OConnor, S., Spriggs, M. & Veth, P. (Eds.), Terra Australis. Vol. 23. The Archaeology of the Aru
Islands, Eastern Indonesia. Pandanus Books, Canberra, pp. 171–204.
Pacey, T.L., Baverstock, P.R. & Jerry, D.R. (2001) The phylogenetic relationships of the bilby, Macrotis lagotis
(Peramelimorphia: Thylacomyidae), to the bandicoots –DNA sequence evidence. Molecular Phylogenetics and Evolution,
21, 26–31. https://doi.org/10.1006/mpev.2001.0987
Phillips, M.J., Mclenachan, P.A., Down, C., Gibb, G.C. & Penny, D. (2006) Combined mitochondrial and nuclear DNA
sequences resolve the interrelations of the major Australian marsupial radiation. Systematic Biology, 55, 122–137. https://
doi.org/10.1080/10635150500481614
Phillips, M.J. & Pratt, R.C. (2008) Family-level relationships among the Australasian marsupial "herbivores" (Diprotodontia:
Koala, wombats, kangaroos and possums). Molecular Phylogenetics and Evolution, 46, 594–605.
Phillips, M.J., Bennett, T.H. & Lee, M.S.Y. (2009) Molecules, morphology, and ecology indicate a recent, amphibious ancestry
for echidnas. Proceedings of the National Academy of Science of USA, 106, 17089–17094.
Pope, L., Storch, D., Adams, M., Moritz, C. & Gordon, G. (2001) A phylogeny for the genus Isoodon and a range extension for
I. obesulus peninsulae based on mtDNA control region and morphology. Australian Journal of Zoology, 49, 411–434.
Posada, D. (2008) jModelTest: phylogenetic model averaging. Molecular Biology and Evolution, 25, 1253–1256.
Ralls, K. (1976) Mammals in Which Females are Larger Than Males. The Quarterly Review of Biology, 51, 245–276.
https://doi.org/10.1086/409310
Rambaut, A. (1996) Sequence alignment editor. Available from: http://tree.bio.ed.ac.uk/software/seal (accessed 5 May 1997)
Rambaut, A. & Drummond, A. J. (2007) Tracer v1.5. Available from: http://tree.bio.ed.ac.uk/software/tracer (accessed 23
November 2008)
Ronquist, F., Kloppstein, S., Vilhelmsen, L., Schulmeister, S., Murray, D.L. & Rasnitsyn, A.P. (2012) A total-evidence
approach to dating with fossils, applied to the early radiation of the Hymenoptera. Systematic Biology, 61, 973–999.
Ronquist, F., Lartillot, N. & Phillips, M.J. (2016) Closing the gap between rocks and clocks using total-evidence dating.
Philosophical Transactions of The Royal Society B Biological Sciences, 371, 20150136.
https://doi.org/10.1098/rstb.2015.0136
Swoddord, D. (2002) PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4 (Updated to 10 Beta).
Sinauer Associates, Sunderland, Massachussets.
TRAVOUILLON & PHILLIPS
26
·
Zootaxa 0000 (0) © 2018 Magnolia Press
Tate, G.H.H. (1948) Results of the Archbold Expeditions, no 60. Studies in the Peramelidae (Marsupialia). Bulletin of the
American Museum of Natural History, 92, 317–346.
Thomas, O. (1906) On mammals collected in south-western Australia for Mr. W. E. Balston. Proceedings of the Zoological
Society of London, 1906, 468–478.
Thomas, O. (1922) On bandicoots allied to Perameles bougainvillei. Annals & Magazine of Natural History, 9, 143–145.
Travouillon, K.J., Gurovich, Y., Beck, R.M.D. & Muirhead, J. (2010) An exceptionally well-preserved short-snouted bandicoot
(Marsupialia; Peramelemorphia) from Riversleigh's Oligo-Miocene deposits, Northwestern Queensland, Australia.
Journal of Vertebrate Paleontology, 30, 1528–1546.
https://doi.org/10.1080/02724634.2010.501463
Travouillon, K.J., Gurovich, Y., Archer, M., Hand, S.J. & Muirhead, J. (2013a) The genus Galadi: three new bandicoots
(Marsupialia, Peramelemorphia) from Riversleigh’s Miocene deposits, northwestern Queensland, Australia. Journal of
Vertebrate Paleontology, 33, 153–168.
https://doi.org/10.1080/02724634.2012.713416
Travouillon, K.J., Beck, R.M.D., Hand, S.J. & Archer, M. (2013b) The oldest fossil record of bandicoots (Marsupialia;
Peramelemorphia) from the late Oligocene of Australia. Palaeontologia Electronica, 16.2.13A, 52. Available from: http://
palaeo-electronica.org/content/2013/400-late-oligocene-bandicoot (accessed 18 May 2016)
Travouillon, K.J., Hand, S.J., Archer, M. & Black, K.H. (2014) Earliest modern bandicoot and bilby (Marsupialia, Peramelidae
and Thylacomyidae) from the Miocene of the Riversleigh World Heritage Area, northwestern Queensland, Australia.
Journal of Vertebrate Paleontology, 34, 375–382.
https://doi.org/10.1080/02724634.2013.799071
Travouillon, K.J., Archer, M., Hand, S.J. & Muirhead, J. (2015) Sexually dimorphic bandicoots (Marsupialia,
Peramelemorphia) from the Oligo-Miocene of Australia, first cranial ontogeny for fossil bandicoots and new species
descriptions. Journal of Mammalian Evolution, 22, 141–167.
https://doi.org/10.1007/s10914-014-9271-8
Travouillon, K.J. (2016a) Investigating dental variation in Perameles nasuta Geoffroy, 1804, with morphological evidence to
raise P. nasuta pallescens Thomas, 1923 to species rank. Zootaxa, 4114, 351–392.
Travouillon, K.J. (2016b) Oldest fossil remains of the enigmatic pig-footed bandicoot show rapid herbivorous evolution. Royal
Society Open Science, 3, 160089.
Travouillon, K.J., Louys, J., Price, G.J., Archer, M., Hand, S.J. & Muirhead, J. (2017) A review of the Pliocene bandicoots of
Australia, and descriptions of new species and genus. Journal of Vertebrate Paleontology, 37 (5), e1360894.
https://doi.org/10.1080/02724634.2017.1360894
Turnbull, W.D., Lundelius, E.L.J. & Archer, M. (2003) Dasyurids, perameloids, phalangeroids, and vombatoids from the Early
Pliocene Hamilton Fauna, Victoria, Australia. Bulletin of the American Museum of Natural History, 279, 513–540.
Van Dyck, S. & Strahan, R. (2008) The mammals of Australia. Reed New Holland, Sydney, 887 pp.
Wagner, A. (1841) Beschreibung einer neuen Art von Bandikuts, Perameles myosurus, nebts Bemerkungen iiber Perameles
obesula. Archiv für Naturgeschichte, 7, 289–297.
Warburton, N.M. & Travouillon, K.J. (2016) The biology and palaeontology of the Peramelemorphia: a review of current
knowledge and future research directions. Australian Journal of Zoology, 64, 151–181.
https://doi.org/10.1071/ZO16003
Waterhouse, G.R. (1838). Catalogue of the Mammalia Preserved in the Museum of the Zoological Society of London. R. Taylor
& J.E. Taylor, London, 42 pp.
Webster, D.B. (1962) A Function of the Enlarged Middle-Ear Cavities of the Kangaroo Rat, Dipodomys. Physiological
Zoology, 35, 248–255.
https://doi.org/10.1086/physzool.35.3.30152809
Westerman, M., Springer, M.S., Dixon, J. & Krajewski, C. (1999) Molecular relationships of the extinct pig-footed bandicoot
Chaeropus ecaudatus (Marsupialia: Perameloidea) using 12S rRNA sequences. Journal of Mammalian Evolution, 6,
271–288.
Westerman, M., Kear, B.P., Aplin, K., Meredith, R.W., Emerling, C. & Springer, M.S. (2012) Phylogenetic relationships of
living and recently extinct bandicoots based on nuclear and mitochondrial DNA sequences. Molecular Phylogenetics and
Evolution, 62, 97–108.
https://doi.org/10.1016/j.ympev.2011.09.009
Woinarski, J., Burbidge, A. & Harrison, P. (2014) Action Plan for Australian Mammals 2012. CSIRO Publishing, Melbourne,
1038 pp.
Woodhead, J., Hand, S.J., Archer, M., Graham, I., Sniderman, K., Arena, D.A., Black, K.H., Godthelp, H., Creaser, P. & Price,
E. (2016) Developing a radiometrically-dated chronologic sequence for Neogene biotic change in Australia, from the
Riversleigh World Heritage Area of Queensland. Gondwana Research, 29, 153–167.
Wroe, S. (1999) The geologically oldest dasyurid, from the Miocene of Riversleigh, north-west Queensland. Paleontology, 42,
501–527.
Wroe, S. (2001) A new genus and species of dasyuromorphian from the Miocene of Rivers1eigh, northern Australia. Memoirs
of the Australian Association of Palaeontologists, 25, 53–59.
Zenger, K.R., Eldridge, M.D.B. & Johnston, P.G. (2005) Phylogenetics, population structure and genetic diversity of the
endangered southern brown bandicoot (Isoodon obesulus) in south-eastern Australia. Conservation Genetics, 6, 193–204.
Zootaxa 0000 (0) © 2018 Magnolia Press
·
27
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
SUPPLEMENTARY FIGURE 1. Bayesian inference phylogeny of peramelemorphians from RAG1 DNA sequences, carried
out in MrBayes 3.2.6 with a GTR+I+G4 model. Extant members of Peramelinae and Echymiperinae+Peroryctinae were
constrained to be reciprocally monophyletic. Clade support values are Bayesian posterior probabilities.
TRAVOUILLON & PHILLIPS
28
·
Zootaxa 0000 (0) © 2018 Magnolia Press
SUPPLEMENTARY FIGURE 2. Perameles bougainville, study skins. A–C, WAM M3640, adult; D–F, WAM M16086,
juvenile. A and D, dorsal view; B and E, lateral view; C and F, ventral view. Scale = 5cm.
Zootaxa 0000 (0) © 2018 Magnolia Press
·
29
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
SUPPLEMENTARY FIGURE 3. Perameles myosura notina, holotype, BMNH 43.8.12.21, study skin. A, dorsal view; B,
lateral view; C, ventral view. Grid scale = 1cm.
TRAVOUILLON & PHILLIPS
30
·
Zootaxa 0000 (0) © 2018 Magnolia Press
SUPPLEMENTARY FIGURE 4. Skulls and jaws of Perameles bougainville bougainville, WAM M6166, (A–E), Perameles
myosura notina, BMNH 43.8.12.21, (F–H), Perameles gunnii, WAM M16590, (I–M), and Perameles eremiana, WAM M2629,
(N–R). A, D, F, I, L, N, Q, dorsal view; B, G, J, O, ventral view; C, E, H, K, M, P, R, lateral view. Scale = 2cm.
Zootaxa 0000 (0) © 2018 Magnolia Press
·
31
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
SUPPLEMENTARY FIGURE 5. Relationship between Bayesian posterior probability (BPP) and maximum parsimony
bootstrap (MP-BP) on the morphological data for clades favoured in both of these unrooted (non-clock) analyses. Logarithmic
curve fitted as BPP=0.3912*ln(MP-BP)—0.8015 (R2 = 0.7575). The two clades that fall furthest below the trend are Macrotis-
Isoodon (filled square), and a clade that includes all Isoodon except Isoodon auratus barrowensis (filled triangle).
TRAVOUILLON & PHILLIPS
32
·
Zootaxa 0000 (0) © 2018 Magnolia Press
SUPPLEMENTARY FIGURE 6. Specimens of Peroryctes broadbenti (A) and Echymipera kalubu (B) showing centrocrista
formation as a result of dental wear.
Zootaxa 0000 (0) © 2018 Magnolia Press
·
33
PERAMELEMORPHIAN PHYLOGENY & NEW SPECIES DESCRIPTION
TABLE S1. Substitution model categories for the molecular data partitions. Each is the most general available within
MrBayes that is consistent with the more general of the jModelTest hLRT or AIC recommendations.
TABLE S2. Univariate statistics for the ear, pes and tail length of the Perameles bougainville complex.
Partition Substitution model
Mt Cytb codon positions 1&2 GTR+I+G
4
Mt Cytb codon position 3 (RY coded) F81+I+G
4
Mt RNA stems GTR+I+G
4
Mt RNA loops GTR+I+G
4
Nuclear codon position 1 GTR+I+G
4
Nuclear codon position 1 GTR+I+G
4
Nuclear codon position 1 GTR+G
4
Ear length N Min Max Mean SE Var SD CV
P. bougainville 17 26 39 33.2059 0.771366 10.7101 3.27263 9.8556
P. eremiana 5 35.62 43 38.754 1.19887 10.0611 3.17192 8.1848
P. myosuros 8 26.58 43.79 35.5262 1.69013 34.2786 5.85479 16.48
P. notina 4 29.13 47.6 38.61 3.45774 59.7797 7.73173 20.025
P. papillon 29 31.68 49 38.9845 0.813916 19.2113 4.38307 11.243
Pes length N Min Max Mean SE Var SD CV
P. bougainville 18 46.06 54.18 50.4244 0.53269 5.10766 2.26001 4.482
P. eremiana 7 48 59 53.0371 1.37647 13.2627 3.6418 6.8665
P. myosuros 12 50 62 55.4267 1.07941 13.9815 3.73918 6.7462
P. notina 5 54.56 59.69 57.784 0.999263 4.99263 2.23442 3.8668
P. papillon 29 41 56 49.2 0.502639 7.32673 2.70679 5.5016
Tail length N Min Max Mean SE Var SD CV
P. bougainville 12 52.32 100.4 79.3267 4.09769 302.239 17.385 21.916
P. eremiana 5 60 124 98.614 8.9519 560.955 23.6845 24.017
P. myosuros 8 95.91 110.073 102.407 1.42797 24.4693 4.94664 4.8304
P. notina 4 75.27 89.77 85.34 3.04897 46.4811 6.81771 7.9889
P. papillon 24 54 96.57 75.7775 2.1033 128.292 11.3266 14.947
... May-Collado et al. [15] utilized all sequences available at the time to construct their marsupial supermatrix tree, in which Chaeropus was sister to the peroryctine bandicoots. Subsequently, Travouillon and Phillips [16] combined the mtDNA sequences with morphological data (including for fossil taxa) and placed Chaeropus as sister to Macrotis or outside all extant peramelemorphians (as did Kear et al. [17] and Beck et al. [18]). Travouillon and Phillips [16] cautioned the use of the Chaeropus RAG1 sequence, because its phylogenetic signal is largely confined to ambiguous sites and, unusually, the inferred substitutions along this lineage are dominated by transversions over transitions. ...
... Subsequently, Travouillon and Phillips [16] combined the mtDNA sequences with morphological data (including for fossil taxa) and placed Chaeropus as sister to Macrotis or outside all extant peramelemorphians (as did Kear et al. [17] and Beck et al. [18]). Travouillon and Phillips [16] cautioned the use of the Chaeropus RAG1 sequence, because its phylogenetic signal is largely confined to ambiguous sites and, unusually, the inferred substitutions along this lineage are dominated by transversions over transitions. Cytb provided the most remarkable result, with Upham et al. [19] nesting the Chaeropus yirratji sequence within a different marsupial order (Dasyuromorphia), among the dunnarts (Sminthopsis). ...
... Peramelemorphia phylogeny (in green) for all extant genera, with dasyuomorphian and Notoryctes outgroups. The potentially extinct Rhynchomeles may fall within Echymipera [16]. Placements of Chaeropus in molecular and combined molecular-morphological studies; 1. Westerman et al. [12,14], Kear et al. [17], Travouillon and Phillips [16], Beck et al. [18], 2. Meredith et al. [13], Travouillon et al. [20], 3. Travouillon and Phillips [16], 4. May-Collado et al. [15], 5,6. ...
Article
Full-text available
Our understanding of the biology of the extinct pig-footed bandicoots (Chaeropus) has been substantially revised over the past two decades by both molecular and morphological research. Resolving the systematic and temporal contexts of Chaeropus evolution has relied heavily on sequencing DNA from century-old specimens. We have used sliding window BLASTs and phylogeny reconstruction, as well as cumulative likelihood and apomorphy distributions, to identify contamination in sequences from both species of pig-footed bandicoot. The sources of non-target DNA that were identified range from other bandicoot species to a bird—emphasizing the importance of sequence authentication for historical museum specimens, as has become standard for ancient DNA studies. Upon excluding the putatively contaminated fragments, Chaeropus was resolved as the sister to all other bandicoots (Peramelidae), to the exclusion of bilbies (Macrotis). The estimated divergence time between the two Chaeropus species also decreases in better agreement with the fossil record. This study provides evolutionary context for testing hypotheses on the ecological transition of pig-footed bandicoots from semi-fossorial omnivores towards cursorial grazers, which in turn may represent the only breach of deeply conserved ecospace partitioning between modern Australo-Papuan marsupial orders.
... For example, hearing characteristics could have implications for their ability to detect and therefore escape from terrestrial predators, or their ability to locate insect prey. Bandicoots and bilbies have marked species variation in pinna and tympanic bulla sizes (Hall et al. 2016;Travouillon and Phillips 2018) but it is not understood how this variation may be related to their ecology. In this study, we tested the relationships between relative bulla and pinna sizes in 29 peramelemorphian species and subspecies and examined to what extent these correlated with environmental characteristics. ...
... Species' distribution maps (Supplementary Figure 3) were drawn in QGIS (QGIS Development Team 2019) using data obtained from the IUCN (2017) terrestrial mammal database, Warburton and Travouillon (2016), modified from Travouillon and Phillips (2018), and the Mammals of New Guinea (Flannery 1995). These maps were used to obtain maximum and average values for temperature (as a measure of the level of thermal challenge for each species) and precipitation (as a proxy for vegetation productivity and therefore habitat complexity; Sala et al. 1988 6 as a proxy of the type of ground and overhead cover, was also considered in modelling. ...
... To control for the potential non-independence of species' data due to phylogenetic relatedness between species (Harvey and Pagel 1991;Chamberlain et al. 2012;Nakagawa and Santos 2012), we implemented phylogenetic information as a variance-covariance matrix in analyses. A phylogenetic tree was constructed by combining two trees constructed using genetic information (Westerman et al. 2012;Travouillon and Phillips 2018) and creating a tree with branch segments representing approximate million year intervals that was built using Mesquite ( Figure 4). Four species were not found in either of the existing trees (Perameles myosuros, Isoodon m. moresbyensis, I. o. nauticus, and Microperoryctes murina) and were therefore placed in polytomies with the expected closest relative (Cruz-Neto et al. 2001). ...
Article
Full-text available
Bandicoots and bilbies (Order Peramelemorphia) occupy a broad range of habitats across Australia and New Guinea, from open, arid deserts to dense forests. This once diverse group has been particularly vulnerable to habitat loss and introduced eutherian predators, and numerous species extinctions and range retractions have occurred. Understanding reasons for this loss requires greater understanding of their biology. Morphology of the pinnae and tympanic bullae varies markedly amongst species. As hearing is important for both predator avoidance and prey location, the variability in ear morphology could reflect specialisation and adaptation to specific environments, and therefore be of conservation relevance. We measured 798 museum specimens representing 29 species of Peramelemorphia. Controlling for phylogenetic relatedness and head length, pinna surface area was weakly negatively correlated with average precipitation (rainfall being our surrogate measure of vegetation productivity/complexity), and there were no environmental correlates with effective diameter (pinna width). Controlling for phylogenetic relatedness and skull length, tympanic bulla volume was negatively correlated with precipitation. Species that inhabited drier habitats, which would be open and allow sound to carry further with less obstruction, had relatively larger pinnae and tympanic bullae. By contrast, species from higher rainfall habitats, where sounds would be attenuated and diffused by dense vegetation, had the smallest pinnae and bullae, suggesting that low-frequency hearing is not as important in these habitats. Associations with temperature did not reach statistical significance. These findings highlight linkages between hearing traits and habitat that can inform conservation and management strategies for threatened species.
... More species are likely to be described in north Australia, and the taxonomy of a number of species need to be resolved [23]. Since the Ziembicki et al. (2015) [23] review, there have been revisions of two major taxa, one within the order Peramelemorphia [68] and one in the genus Dasyurus [69] (see below). Disturbingly, the revision of the Peramelemorphia (bandicoots and bilbies) prompted Travouillon and Phillips (2018) [68] to suggest that the known number of recently extinct bandicoots is likely to increase. ...
... Since the Ziembicki et al. (2015) [23] review, there have been revisions of two major taxa, one within the order Peramelemorphia [68] and one in the genus Dasyurus [69] (see below). Disturbingly, the revision of the Peramelemorphia (bandicoots and bilbies) prompted Travouillon and Phillips (2018) [68] to suggest that the known number of recently extinct bandicoots is likely to increase. ...
Article
Full-text available
Northern Australian biomes hold high biodiversity values within largely intact vegetation complexes, yet many species of mammals, and some other taxa, are endangered. Recently, six mammal species were added to the 20 or so already listed in the Australian endangered category. Current predictions suggest that nine species of mammal in northern Australia are in imminent danger of extinction within 20 years. We examine the robustness of the assumptions of status and trends in light of the low levels of monitoring of species and ecosystems across northern Australia, including monitoring the effects of management actions. The causes of the declines include a warming climate, pest species, changed fire regimes, grazing by introduced herbivores, and diseases, and work to help species and ecosystems recover is being conducted across the region. Indigenous custodians who work on the land have the potential and capacity to provide a significant human resource to tackle the challenge of species recovery. By working with non-Indigenous researchers and conservation managers, and with adequate support and incentives, many improvements in species' downward trajectories could be made. We propose a strategy to establish a network of monitoring sites based on a pragmatic approach by prioritizing particular bioregions. The policies that determine research and monitoring investment need to be re-set and new and modified approaches need to be implemented urgently. The funding needs to be returned to levels that are adequate for the task. At present resourcing levels, species are likely to become extinct through an avoidable attrition process .
... Faunal records for Tungawa Rockshelters 2 and 6, Tartanga and Ngaut Ngaut were compiled from published records (Hale and Tindale 1930;Mulvaney 1960;Mulvaney et al. 1964;Smith 1982), and updated to reflect current taxonomy (see notes supplied with Table S2). Taxonomy for mammals follows Jackson and Groves (2015), excluding Peramelids (bandicoots), for which we referred to Travouillon and Phillips (2018) in recognising Perameles notina (south-eastern striped bandicoot). Bird and reptile taxonomy follows Dickinson and Remsen (2014) and Wilson and Swan (2013) respectively. ...
... Arid zone marsupials are icons of Australia and have an inferred evolutionary history that extends back over some ~15 Ma 1 . Nevertheless, the precise divergence timings of the major extant clades are ambiguous, as are the possible drivers behind their adaptive radiations [2][3][4][5][6][7][8][9][10][11][12][13] . Macropodoids (Macropodiformes: Macropodoidea) ¾ the group encompassing living kangaroos, wallaroos, wallabies, pademelons and tree-kangaroos (Macropodidae), bettongs and potoroos (Potoroidae), the Musky rat-kangaroo (Hypsiprymnodon moschatus: Hypsyprymnodontidae), and their stem antecedents 14 ¾ incorporate some of the most distinctive Australian arid zone marsupials, as epitomised by the famous Red kangaroo, Osphranter rufus 15 . ...
Preprint
Full-text available
The evolution of Australia’s distinctive marsupial fauna has long been linked to the onset of continent-wide aridity. However, how this profound climate change event affected the diversification of extant lineages is still hotly debated. Here, we assemble a DNA sequence dataset of Macropodoidea — the clade comprising kangaroos and their relatives — that incorporates a complete mitogenome for the Desert ‘rat-kangaroo’, Caloprymnus campestris. This enigmatic species went extinct nearly 90 years ago and is known from a handful of museum specimens. Caloprymnus is significant because it was the only macropodoid restricted to extreme desert environments, and therefore calibrates the group’s specialisation for increasingly xeric conditions. Our robustly supported phylogenies nest Caloprymnus amongst the bettongs Aepyprymnus and Bettongia. Dated ancestral area optimisations further reveal that the Caloprymnus-Bettongia lineage originated in nascent arid zone settings from the later-middle to early-late Miocene, ~12 million years ago (Ma), but subsequently dispersed into mesic habitats during the Pliocene and Pleistocene. This coincides with ancestral divergences amongst kangaroos in disparate woodland-forest and shrubland settings, but predates their adaptive radiation into proliferating grasslands during the late Miocene to Pliocene, after ~7 Ma. We thus demonstrate that protracted changes in both climate and vegetation likely staged the emergence of modern arid zone macropodoids.
... Of the five extant species of Isoodon, only I. macrourus is found in New Guinea (Pope et al. 2001;Travouillon and Phillips 2018;Cooper et al. 2020). Fungi, plants and invertebrates are commonly eaten by the Australian species I. fusciventer and I. peninsulae (Opie 1980;Lobert 1985;Quin 1988;Keiper and Johnson 2004;Maclagan et al. 2021). ...
Article
Little is known about the diets and ecology of New Guinea's 14 bandicoot species. In order to better understand the diet and digestive morphology of these marsupials, we reviewed the literature, studied the dental morphology, conducted analysis of gastrointestinal contents, and measured the digestive tracts of: Echymipera clara, E. davidi, E. kalubu, E. rufescens, Isoodon macrourus, Microperoryctes ornata, M. papuensis and Peroryctes raffrayana. These species consume a mix of fungi, insects and plant material that is broadly consistent with the omnivorous diet characteristic of most Australian bandicoots; however, morphological observations reveal variation between species that likely reflect finer-scale differences in diet. Dental morphology suggests a wider variety of diets (insectivore, omnivore, frugivore) than on the Australian mainland (mostly omnivore). Dissections and measurements of the digestive tract of seven New Guinean species indicate variation linked to diet. The relatively short caecum in all New Guinean species, but especially in E. clara and E. kalubu, is particularly suggestive of limited consumption of fibrous plant material; the relative length of the large intestine suggests variable capacity for water reabsorption. Our dietary data also suggest that some of these species also play an important role in the dispersal of hypogeous fungi.
Article
Full-text available
The current literature on marsupial phylogenetics includes numerous studies based on analyses of morphological data with limited sampling of Recent and fossil taxa, and many studies based on analyses of molecular data with dense sampling of Recent taxa, but few studies have combined both data types. Another dichotomy in the marsupial phylogenetic literature is between studies focused on New World taxa and those focused on Sahulian taxa. To date, there has been no attempt to assess the phylogenetic relationships of the global marsupial fauna based on combined analyses of morphology and molecular sequences for a dense sampling of Recent and fossil taxa. For this report, we compiled morphological and molecular data from an unprecedented number of Recent and fossil marsupials. Our morphological data consist of 180 craniodental characters that we scored for 97 terminals representing every currently recognized Recent genus, 42 additional ingroup (crown-clade marsupial) terminals represented by well-preserved fossils, and 5 outgroups (nonmarsupial metatherians). Our molecular data comprise 24.5 kb of DNA sequences from whole-mitochondrial genomes and six nuclear loci (APOB, BRCA1, GHR, RAG1, RBP3 and VWF) for 97 marsupial terminals (the same Recent taxa scored for craniodental morphology) and several placental and monotreme outgroups. The results of separate and combined analyses of these data using a wide range of phylogenetic methods support many currently accepted hypotheses of ingroup (marsupial) relationships, but they also underscore the difficulty of placing fossils with key missing data (e.g., Evolestes), and the unique difficulty of placing others that exhibit mosaics of plesiomorphic and autapomorphic traits (e.g., Yalkaparidon). Unique contributions of our study are (1) critical discussions and illustrations of marsupial craniodental morphology including features never previously coded for phylogenetic analysis; (2) critical assessments of relative support for many suprageneric clades; (3) estimates of divergence times derived from tip-and-node dating based on uniquely taxon-dense analyses; and (4) a revised, higher-order classification of marsupials accompanied by lists of supporting craniodental synapomorphies. Far from the last word on these topics, this report lays the foundation for future research that may be enabled by the discovery of new fossil taxa, better-preserved material of previously described taxa, novel morphological characters (e.g., from the postcranium), and improved methods of phylogenetic analysis.
Article
Residential gardens can provide essential opportunities for native wildlife and represent a valuable way of creating new habitats. Bandicoots (marsupial family Peramelidae) are medium-sized digging mammals that play a valuable role in maintaining ecosystem health; retaining these important ecosystem engineers across urban landscapes, including in private gardens, can have enormous conservation benefits. Urbanisation is a significant threat for some bandicoot species, and therefore understanding the factors associated with their activity can help guide urban landscape and garden design. To identify key features associated with the activity of a local endemic bandicoot species, the quenda (Isoodon fusciventer), we carried out a camera trap survey of front and back yards for 65 residential properties in the City of Mandurah, Western Australia. We compared quenda activity with biotic and abiotic factors that could indicate potential predation risk (activity of domestic dogs Canis familiaris and cats Felis catus, and the presence of artificial or natural protective cover), food availability (including deliberate or inadvertent supplementary feeding, provision of water, and diggable surfaces) and garden accessibility (distance to bushland, permeability of boundary fencing, and garden position). Supplementary feeding was strongly associated with quenda activity. Quenda were also more active in back yards, and in gardens where there was greater vegetation cover. Of concern, quenda activity was positively associated with cat activity, which could reflect that straying pet cats are attracted to gardens that harbour wildlife populations, including quenda. Furthermore, almost half of the gardens showed cat activity despite only a small sample of the surveyed residents owning a pet cat. Results of this study can help guide the design of residential gardens to increase useful habitat for these important digging mammals. Vegetation, wood mulch and semi-permeable fencing can provide valuable resources needed to support the persistence of quendas across the rapidly changing urban landscape mosaic, where natural and managed (e.g., gardens and parks) green spaces are becoming less common and more isolated.
Article
Full-text available
The evolution of Australia’s distinctive marsupial fauna has long been linked to the onset of continent-wide aridity. However, how this profound climate change event affected the diversification of extant lineages is still hotly debated. Here, we assemble a DNA sequence dataset of Macropodoidea—the clade comprising kangaroos and their relatives—that incorporates a complete mitogenome for the Desert ‘rat-kangaroo’, Caloprymnus campestris . This enigmatic species went extinct nearly 90 years ago and is known from a handful of museum specimens. Caloprymnus is significant because it was the only macropodoid restricted to extreme desert environments, and therefore calibrates the group’s specialisation for increasingly arid conditions. Our robustly supported phylogenies nest Caloprymnus amongst the bettongs Aepyprymnus and Bettongia . Dated ancestral range estimations further reveal that the Caloprymnus - Bettongia lineage originated in nascent xeric settings during the middle to late Miocene, ~ 12 million years ago (Ma), but subsequently radiated into fragmenting mesic habitats after the Pliocene to mid-Pleistocene. This timeframe parallels the ancestral divergences of kangaroos in woodlands and forests, but predates their adaptive dispersal into proliferating dry shrublands and grasslands from the late Miocene to mid-Pleistocene, after ~ 7 Ma. We thus demonstrate that protracted changes in both climate and vegetation likely staged the emergence of modern arid zone macropodoids.
Article
Within-species morphological variation is often observed across spatial and climatic gradients. Understanding this variation is important to conservation planning, as specialised adaptations may influence a population’s persistence following translocation. However, knowing whether local adaptations are prevalent within a species can be challenging when the species has undergone range contractions. Here, we used museum specimens to study size and shape variation of the greater stick-nest rat (Leporillus conditor). We aimed to determine whether intraspecific size and shape variation previously existed within the species across its historical range, and inform on possible implications for translocations of the remaining extant population. We found significantly larger skull size in the Franklin Islands and arid populations, possibly indicating a historically continuous population experiencing similar selection pressures such as high predation pressure, competition with other large arid zone rodents or climatic extremes. Conversely, skull shape variation within the species adheres to an allometric trajectory, indicating no specific local adaptations of skull shape. This absence of local skull shape adaptation suggests that the Franklin Islands population is likely suitable for mainland translocations. However, further research into the historical phylogeography of the species is recommended to identify whether large size resulted from shared ancestry or convergent evolution.
Article
Full-text available
Reconstructing phylogeny from retrotransposon insertions is often limited by access to only a single reference genome, whereby support for clades that do not include the reference taxon cannot be directly observed. Here we have developed a new statistical framework that accounts for this ascertainment bias, allowing us to employ phylogenetically powerful retrotransposon markers to explore the radiation of the largest living marsupials, the kangaroos and wallabies of the genera Macropus and Wallabia. An exhaustive in silico screening of the tammar wallaby (Macropus eugenii) reference genome followed by experimental screening revealed 29 phylogenetically informative retrotransposon markers belonging to a family of endogenous retroviruses. We identified robust support for the enigmatic swamp wallaby (Wallabia bicolor) falling within a paraphyletic genus, Macropus. Our statistical approach provides a means to test for incomplete lineage sorting and introgression/hybridization in the presence of the ascertainment bias. Using retrotransposons as “molecular fossils”, we reveal one of the most complex patterns of hemiplasy yet identified, during the rapid diversification of kangaroos and wallabies. Ancestral state reconstruction incorporating the new retrotransposon phylogenetic information reveals multiple independent ecological shifts among kangaroos into more open habitats, coinciding with the Pliocene onset of increased aridification in Australia from ~3.6 million years ago.
Article
Full-text available
The Pliocene fossil record of Australia has revealed the oldest known occurrences of modern peramelemorphian (bandicoot and bilbies) genera such as Perameles, cf. Peroryctes, and Chaeropus. Recent phylogenetic analyses based on morphology have questioned the previously accepted understanding about generic relationships of some of these Pliocene taxa. These doubts limit our ability to develop independent divergence models based on molecular data because they depend on fossil records in order to calibrate minimum rates of change. Hence, there is a need to critically review the Pliocene fossil record. To this end, we have examined Pliocene specimens of peramelemorphians in museum collections across Australia and performed a comprehensive phylogenetic analysis based on dental and cranial morphology to review and accordingly revise subfamilial taxonomy. As part of this revision, we describe here two new species, one from the Chinchilla Local Fauna (LF) in Queensland (Perameles wilkinsonorum, sp. nov.) and one from both the Big Sink LF in the Wellington Caves area and the Bow LF in New South Wales (Silvicultor karae, gen. et sp. nov.). We reassign the two ‘Peroryctes’ species from Hamilton LF, Victoria, to a new genus (Silvicultor). We summarize the distribution of peramelemorphians during the Pliocene and show how climate change appears to have shaped their subsequent Quaternary distribution. SUPPLEMENTAL DATA—Supplemental materials are available for this article for free at www.tandfonline.com/UJVP Citation for this article: Travouillon, K. J., J. Louys, G. J. Price, M. Archer, S. J. Hand, and J. Muirhead. 2017. A review of the Pliocene bandicoots of Australia, and descriptions of new genus and species. Journal of Vertebrate Paleontology. DOI: 10.1080/02724634.2017.1360894.
Article
Full-text available
Bandicoots (Peramelemorphia) are a unique order of Australasian marsupials whose sparse fossil record has been used as prima facie evidence for climate change coincident faunal turnover. In particular, the hypothesized replacement of ancient rainforest-dwelling extinct lineages by antecedents of xeric-tolerant extant taxa during the late Miocene (~10 Ma) has been advocated as a broader pattern evident amongst other marsupial clades. Problematically, however, this is in persistent conflict with DNA phylogenies. We therefore determine the pattern and timing of bandicoot evolution using the first combined morphological + DNA sequence dataset of Peramelemorphia. In addition, we document a remarkably archaic new fossil peramelemorphian taxon that inhabited a latest Quaternary mosaic savannah-riparian forest ecosystem on the Aru Islands of Eastern Indonesia. Our phylogenetic analyses reveal that unsuspected dental homoplasy and the detrimental effects of missing data collectively obscure stem bandicoot relationships. Nevertheless, recalibrated molecular clocks and multiple ancestral area optimizations unanimously infer an early diversification of modern xeric-adapted forms. These probably originated during the late Palaeogene (30–40 Ma) alongside progenitors of other desert marsupials, and thus occupied seasonally dry heterogenous habitats long before the onset of late Neogene aridity.
Article
Full-text available
The pig-footed bandicoot, Chaeropus ecaudatus, is one of the most enigmatic Australian marsupials, which went extinct in the late 1950s probably as a result of European colonization. It is unusual in being the only marsupial to have evolved reduction of digits on both fore and hind feet, with the forefeet being pig-like (two toes) and the hind feet being horselike (one toe). According to molecular phylogenetic analyses, Chaeropus diverged from other bandicoots (Peramelidae), and the bilbies (Thylacomyidae) by the mid-Late Oligocene. This is considerably earlier than suggested by the fossil record, with the current oldest specimens being Late Pleistocene in age. Here, I report the oldest fossils of Chaeropus, representing a new species, Chaeropus baynesi from the Late Pliocene to Early Pleistocene (2.47–2.92 Ma) Fisherman’s Cliff Local Fauna, Moorna Formation, New South Wales, Australia, and extending the fossil record of the genus and family by at least 2 million years. Chaeropus baynesi is less high crowned than C. ecaudatus and lacks lateral blade development on lower molars, suggesting that it was unlikely to be grazing. This suggests that Chaeropus must have adapted rapidly to the drying conditions and changes in environments, and would have become a grazer in a very short period of time.
Book
The Action Plan for Australian Mammals 2012 is the first review to assess the conservation status of all Australian mammals. It complements The Action Plan for Australian Birds 2010 (Garnett et al. 2011, CSIRO Publishing), and although the number of Australian mammal taxa is marginally fewer than for birds, the proportion of endemic, extinct and threatened mammal taxa is far greater. These authoritative reviews represent an important foundation for understanding the current status, fate and future of the nature of Australia. This book considers all species and subspecies of Australian mammals, including those of external territories and territorial seas. For all the mammal taxa (about 300 species and subspecies) considered Extinct, Threatened, Near Threatened or Data Deficient, the size and trend of their population is presented along with information on geographic range and trend, and relevant biological and ecological data. The book also presents the current conservation status of each taxon under Australian legislation, what additional information is needed for managers, and the required management actions. Recovery plans, where they exist, are evaluated. The voluntary participation of more than 200 mammal experts has ensured that the conservation status and information are as accurate as possible, and allowed considerable unpublished data to be included. All accounts include maps based on the latest data from Australian state and territory agencies, from published scientific literature and other sources. The Action Plan concludes that 29 Australian mammal species have become extinct and 63 species are threatened and require urgent conservation action. However, it also shows that, where guided by sound knowledge, management capability and resourcing, and longer-term commitment, there have been some notable conservation success stories, and the conservation status of some species has greatly improved over the past few decades. The Action Plan for Australian Mammals 2012 makes a major contribution to the conservation of a wonderful legacy that is a significant part of Australia’s heritage. For such a legacy to endure, our society must be more aware of and empathetic with our distinctively Australian environment, and particularly its marvellous mammal fauna; relevant information must be readily accessible; environmental policy and law must be based on sound evidence; those with responsibility for environmental management must be aware of what priority actions they should take; the urgency for action (and consequences of inaction) must be clear; and the opportunity for hope and success must be recognised. It is in this spirit that this account is offered. Winner of a 2015 Whitley Awards Certificate of Commendation for Zoological Resource.
Book
This volume focuses on the broad pattern of increasing biodiversity through time, and recurrent events of minor and major ecosphere reorganization. Intense scrutiny is devoted to the pattern of physical (including isotopic), sedimentary and biotic circumstances through the time intervals during which life crises occurred. These events affected terrestrial, lacustrine and estuarine ecosystems, locally and globally, but have affected continental shelf ecosystems and even deep ocean ecosystems. The pattern of these events is the backdrop against which modelling the pattern of future environmental change needs to be evaluated.