Phylogeny and evolution of body mass in didelphid marsupials (Marsupialia: Didelphimorphia: Didelphidae)

Abstract

Most extant New World marsupials belong in the Didelphidae, which comprises ca. 110 currently recognized species of opossums. Didelphids are small mammals with their mean body mass, at species level, ranging from ca. 7 g to 2.2 kg. The largest species belong in a single clade, while substantial variation remains scattered across the remaining groups. We seek out to explore the details of this mass variation in an evolutionary framework. To this end, we first reconstructed the phylogeny of didelphids based on an extensive, although fragmentary sample of sequences from ten genes. We recovered a fully resolved, highly robust phylogeny that tested and confirmed most previously reported groupings, providing a simultaneous depiction of phylogenetic relationships for 81 % of currently recognized species and all relevant supra-specific clades. As much as 69 % of total body mass variation in didelphids was explained by this phylogenetic hypothesis. Mapped on it, mass variation evolved as much as 6.8 kg of total changes, starting from a reconstructed ancestral body mass range of 22–33 g. No single, family-wide pattern was evident; in fact, the dominant pattern for mass variation was that of increases in body mass along a few successive branches, or phyletic giantism, followed by apomorphic nanism, i.e., decreases localized in single terminal branches. Phyletic trends indicated the persistence of gradual, directional changes along considerable spans of geological time and show that substantial variation of interest resides in this and perhaps most groups of small mammals.

Full text

ORIGINAL ARTICLE
Phylogeny and evolution of body mass in didelphid marsupials
(Marsupialia: Didelphimorphia: Didelphidae)
Lucila I. Amador
1
&Norberto P. Giannini
1,2,3
Received: 18 May 2015 / Accepted: 22 December 2015
#Gesellschaft für Biologische Systematik 2016
Abstract Most extant New World marsupials belong in the
Didelphidae, which comprises ca. 110 currently recognized
species of opossums. Didelphids are small mammals with
their mean body mass, at species level, ranging from ca. 7 g
to 2.2 kg. The largest species belong in a single clade, while
substantial variation remains scattered across the remaining
groups. We seek out to explore the details of this mass varia-
tion in an evolutionary framework. To this end, we first recon-
structed the phylogeny of didelphids based on an extensive,
although fragmentary sample of sequences from ten genes.
We recovered a fully resolved, highly robust phylogeny that
tested and confirmed most previously reported groupings,
providing a simultaneous depiction of phylogenetic relation-
ships for 81 % of currently recognized species and all relevant
supra-specific clades. As much as 69 % of total body mass
variation in didelphids was explained by this phylogenetic
hypothesis. Mapped on it, mass variation evolved as much
as 6.8 kg of total changes, starting from a reconstructed an-
cestral body mass range of 22–33 g. No single, family-wide
pattern was evident; in fact, the dominant pattern for mass
variation was that of increases in body mass along a few suc-
cessive branches, or phyletic giantism, followed by
apomorphic nanism, i.e., decreases localized in single terminal
branches. Phyletic trends indicated the persistence of gradual,
directional changes along considerable spans of geological
time and show that substantial variation of interest resides in
this and perhaps most groups of small mammals.
Keywords Didelphidae .Body mass .Phylogeny .Character
mapping .Phyletic giantism .Nanism
Introduction
Marsupialia (Metatheria) conforms a particular group of mam-
mals, not only with regard to anatomical and physiological
traits (e.g., osteology, dentition, reproductive system;
Dickman 2005), but also given their distinct biogeographic
and evolutionary history. With seven orders and some 330
living species, marsupials comprise ca. 7 % of the extant
mammalian diversity (Dickman 2005). These mammals have
lived in all continents, but since the Middle Miocene, they
became restricted to the Australasian region east to the
Wallace line (Groves 2005) and the New World (Gardner
2005). In the Americas, marsupials are represented by three
extant orders (Microbiotheria, Paucituberculata, and
Didelphimorphia), each one including a single extant family
(Microbiotheriidae, Caenolestidae, and Didelphidae, respec-
tively). Microbiotheriidae groups several extinct taxa and the
single extant species Dromiciops gliroides; likewise,
Caenolestidae includes many extinct but just six extantspecies
classified in three genera. Presently, Didelphidae comprises
the majority of the extant diversity of New World marsupials
(Gardner 2008), with ca. 110 currently recognized species in
18 genera (after synonymy of Micoureus with Marmosa; Voss
Electronic supplementary material The online version of this article
(doi:10.1007/s13127-015-0259-x) contains supplementary material,
which is available to authorized users.
*Lucila I. Amador
amadorlucila@gmail.com
1
Unidad Ejecutora Lillo (UEL: CONICET-FML),
Tucumán, Argentina
2
Facultad de Ciencias Naturales e Instituto Miguel Lillo,
Tucumán, Argentina
3
Department of Mammalogy, American Museum of Natural History,
New York, NY, USA
Org Divers Evol
DOI 10.1007/s13127-015-0259-x

et al. 2004;Gardner2005;Vossetal.2005; Solari 2007; Teta
et al. 2009; Voss and Jansa 2009; Gutiérrez et al. 2010; Voss
et al. 2012; Paglia et al. 2012; Solari et al. 2012;Caramaschi
et al. 2011; Pavan et al. 2012;Vossetal.2013;Giarlaand
Jansa 2014; Martínez-Lanfranco et al. 2014; Palma et al.
2014; Pavan et al. 2014;Vossetal.2014; Vilela et al. 2015;
Astúa 2015).
The evolutionary history of New World marsupials is rich-
ly documented in the fossil record (see McKenna and Bell
1997). This history has been shaped by extinction events
(e.g., Dickman and Vieira 2005) and major changes in key
characters, remarkably body mass and conspicuous functional
features associated to diet and locomotion (e.g., Argot 2003,
2004a,b). However, didelphids in particular exhibit a com-
paratively poor fossil record (see McKenna and Bell 1997)
and have experienced relatively modest changes along their
own evolutionary history. As a consequence, didelphids are
often perceived as remarkably conservative in body structure;
e.g., the didelphid molar model is only a slight modification of
an ancient one already found in Cretaceous metatherians (Fox
1987; Chemisquy et al. 2015). For this reason, didelphids, and
particularly the type genus Didelphis, have been traditionally
considered a useful functional model of generalized
metatherians and hence of primitive mammals (e.g.,
Crompton and Hiiemae 1970). However, this conservatism
is only apparent (e.g., Astúa 2009), as didelphids exhibit in-
teresting systematic variation in relevant morphological char-
acters that contribute synapomorphies to both the group and
most inner clades (e.g., Voss and Jansa 2009;Flores2009).
Here, we focus on macroevolutionary variation in body mass
in this natural group. Size is the single most important factor
affecting the biology of mammals (McNab 2007 and Citations
therein), and the particular way in which body mass varies in
different groups, or in mammals as a group, has been a con-
stant source of research in the Class (e.g., Meiri et al. 2008;
Cooper and Purvis 2010; Venditti et al. 2011;G
ianninietal.
2012;Slater2013). Gould and MacFadden (2004)describe
evolutionary body mass variation in terms of apomorphic ver-
sus phyletic change (i.e., changes traceable to a single branch
versus changes accumulated along several successive
branches of a phylogeny), as well as increases versus de-
creases. These compose, in combination, the possible out-
comes of apomorphic or phyletic giantism or nanism, in ad-
dition to stasis. Gould and MacFadden (2004) studied macro-
evolutionary size variation in large animals, specifically
equids among mammals. While the largest specimens of
Didelphis virginiana can exceptionally reach 7 kg (although
only in urban environments; McManus 1974; Cerqueira and
Tribe 2007) and some extinct forms were seemingly larger
(e.g., Thylophorops; Simpson 1972,Goinetal.2009;
Hyperdidelphis, Goin and Pardiñas 1996), the great majority
of extant didelphids are rather small mammals (Birney and
Monjeau 2003). To our knowledge, small mammals seldom
have been examined in the framework proposed by Gould and
McFadden (2004), perhaps only bats (Giannini et al. 2012).
In preliminary grounds, several aspects of body mass var-
iation in didelphids are noteworthy and invite further investi-
gation. First, the observed interspecific variation in the aver-
age of body mass in extant didelphid species ranges from 7.6 g
in Gracilinanus emiliae to 2195.5 g in Didelphis virginiana
(data from Smith et al. 2003)oranimpressive289-folddif-
ference between the average of the smallest and the largest
species. Second, cranial morphometric studies have revealed
that didelphids are among the most morphologically highly
integrated, least modular groups of mammals (Shirai and
Marroig 2010); here, body mass is important because that
magnitude of integration arises when most morphological var-
iation is highly correlated with size variation (Porto et al.
2009; see also Astúa 2009). Third, a single group of
didelphids, Didelphini sensu Steiner et al. (2005) contains
Metachirus and all the largest species in four genera (the so-
called 2n = 22 clade, composed of Chironectes,Lutreolina,
Philander,andDidelphis;Flores2009; Voss and Jansa
2009). This strongly suggests a key role of phylogenetic leg-
acy in the evolution of body mass in the group. Naturally,
discovering the specific way in which body mass evolved in
didelphids demands investigation within an explicit phyloge-
netic framework.
Recent studies made significant advances in the phylogeny
of the group. First, although the monophyly of this group was
never seriously challenged within a cladistic context, all mod-
ern phylogenies successively accrued strong support for a
didelphid clade from both molecular (e.g., Voss and Jansa
2003; Jansa and Voss 2005; Gruber et al. 2007; Voss and
Jansa 2009) and morphological data (e.g., Horovitz and
Sánchez‐Villagra 2003; Voss and Jansa 2003;Jansaand
Vo s s 2005;Flores2009). Second, and contrary to earlier stud-
ies, the new phylogenies revealed that internal relationships
among didelphids are remarkably congruent across different
sources of evidence (morphology, karyotypes, and gene
sequences; Flores 2009; Voss and Jansa 2009). These results,
defining the branching pattern of extant didelphids, have been
consistently reflected in the current systematics of the family
(Fig. 1; Voss and Jansa 2009; de la Sancha et al. 2012;Palma
et al. 2014;Vossetal.2014).
Here, we seek to investigate in detail the evolution of body
mass in this group with relatively modest, albeit interesting
variation. To this end, we contribute a comprehensive phylog-
eny that includes 81 % of currently recognized species. This
phylogeny was built upon molecular data including sequences
from ten genes from recent studies, so constructed in order to
increase the taxonomic density and, thus, the reliability of
both the phylogenetic framework used and the estimation of
evolutionary change in body mass. We show (1) that the cur-
rent phylogenetic hypothesis of didelphid relationships
(Fig. 1) is extraordinarily robust to the test of significant
L.I. Amador, N.P. Giannini

(near 2-fold) increase in both character and taxonomic sam-
pling and (2) that while research has been centered around
trends that comprise relatively large changes in wide body
mass ranges (e.g., Cope’s rule or the island rule; Lomolino
2005), macroevolutionary variation in small mammals such
as didelphids reveals highly interesting patterns with signifi-
cant implications in the perceived evolution of body mass in
mammals.
Materials and methods
Taxo n omi c s a mpl i n g
We constructed a data matrix that includes 89 currently recog-
nized species following Gardner (2005) and other authorities
(Voss et al. 2004;Vossetal.2005; Solari 2007;Tetaetal.
2009; Voss and Jansa 2009; Gutiérrez et al. 2010;Vossetal.
2012; Paglia et al. 2012; Solari et al. 2012;Caramaschietal.
2011; Pavan et al. 2012;Vossetal.2013; Giarla and Jansa
2014; Martínez-Lanfranco et al. 2014; Palma et al. 2014;
Pavan et al. 2014;Vossetal.2014;Vilelaetal.2015). This
matrix includes all the species represented in the GenBank
database as of November 2014 (see taxonomic list and
accession numbers in Table SI.1, Supplementary
Information) for the gene sequences selected for this study
(see below). All didelphid genera and species groups are rep-
resented in this sample (Table 1), except for the monotypic
Chacodelphys for which no sequences are available. Five taxa
were added as outgroups: the caenolestids Rhyncholestes
raphanurus,Lestoros inca,andCaenolestes fuliginosus
(Paucituberculata: Caenolestidae) and Dromiciops gliroides
(Microbiotheria: Microbiotheriidae), which together with
didelphids represent the extant higher-level diversity of New
World marsupials, plus one basal representative of the large
native Australian marsupial clade, Dasyurus geoffroii
(Dasyuromorphia: Dasyuridae). Thereby, the analysis com-
prised a total of 94 terminals.
Characters and matrix compilation
We selected sequences from four mitochondrial and six nucle-
ar genes. The mitochondrial genes included the protein-
coding cytochrome b (Cyt-b) and the cytochrome c oxidase
subunit 1 (COI) and the ribosomal subunits 12S and 16S. The
nuclear genes included the interphotoreceptor retinoid binding
protein (IRBP), the breast cancer susceptibility protein 1
(BRCA1, exons 10 and 11), the dentin matrix protein 1
(DMP1), the recombination activating protein 1 (RAG1), the
von Willebrand factor (vWF, exon 28), and the transthyretin
intron 1 (TTR). These sequences are of proven systematic
utility in didelphids, and they have been generated and used,
in different combinations, by several previous authors (e.g.,
Palma and Spotorno 1999; Steiner et al.2005; Jansa et al.
2006;Gruberetal.2007; Voss and Jansa 2009). Sequences
of these genes were available for at least 23 of the study spe-
cies (Tables 2and SI.1for accession numbers).
Sequences were compiled with the program GenBank to
TNT (GB2TNT; Goloboff and Catalano 2012), a pipeline for
creating large molecular matrices that selects sequences in-
cluded in a GenBank file (by using filters defined by the user),
Tab l e 1 Taxonomic sampling, detailing number (a.f. absolute
frequency), and percentage (%) of included species of each extant
didelphid genus
Genus Number of species Number of sampled species
a.f. %
Caluromys 33100.0
Caluromysiops 11100.0
Chacodelphys 100.0
Chironectes 11100.0
Cryptonanus 5360.0
Didelphis 66100.0
Glironia 11100.0
Gracilinanus 6466.7
Hyladelphis 11100.0
Lestodelphys 11100.0
Lutreolina 22100.0
Marmosa 20 17 85.0
Marmosops 15 9 60.0
Metachirus 11100.0
Monodelphis 26 22 84.6
Philander 6466.7
Thylamys 13 12 92.3
Tlacuatzin 11100.0
Total 110 89 80.9
Fig. 1 Current hypothesis of systematic relationships in Didelphidae.
Subg. subgenus
Phylogeny and evolution of body mass in didelphid marsupials

creates a Fasta file, produces the alignment (by calling an
external program defined by the user), and generates the data
matrix in TNT (Goloboff et al. 2008b) format. From this
format, we migrated manually to PHYLIP format used by
RAxML (Stamatakis et al.2008). The completeness and ac-
curacy of the dataset so generated were checked against the
original GenBank depository to avoid missing sequences, du-
plicates, and other mistakes, which include, for instance, syn-
tactic ambiguities in the species entries. Sequences were
aligned using the program MAFFT (version 7, Katoh and
Standley 2013) with default settings. In all analyses (see be-
low), indels were treated as missing values, and the
australidelphian Dasyurus geoffroii was designated to root
the trees.
Phylogenetic analyses
We conducted two series of phylogenetic analyses with differ-
ent optimality criteria. First, a maximum likelihood (ML)
analysis was performed on the total dataset using the server
version of RAxML 7.2.8 (Stamatakis et al.2008). This pro-
gram approximates the finding of ML trees using a sample of
100 bootstrap replicates. The substitution model applied was
GTR + GAMMA + I.
Second, we performed parsimony (maximum parsimony
(MP)) analyses for tree search and branch support using the
program TNT 1.1 (Goloboff et al. 2008b). We executed
heuristic searches based on 200 random addition sequences
(RAS) each followed by Tree Bisection Reconnection (TBR)
branch swapping, saving up to 15 trees per replicate. The
search was refined with an additional TBR round applied to all
trees kept in memory. Strict consensus tree was calculated from
the set of most parsimonious trees obtained. All characters
were treated as unordered and using both equal weighting
(default option) and implied weighting (Goloboff 1993,
2014), as implemented in the current version of TNT.
Implied weighting is a traditional technique that parsimonious-
ly weights against homoplasy during tree search (Goloboff
1993), and it has been shown to outperform equally weighted
parsimony (Goloboff et al. 2008a). This technique was
applied with default settings (e.g., concavity constant k=3;
Goloboff et al. 2008a). No constraints were enforced during
searches. Branch stability and support were estimated with a
symmetric resampling (jackknife) analysis based on 1000
replicates and sequential Bremer absolute values. We
followed Giannini and Bertelli (2004) to calculate Bremer
values from ten successive samplings of up to 2000, one-
step-longer suboptimal trees. This search strategy was applied
to each gene and to concatenated datasets, specifically nuclear,
mitochondrial, total, and a nine-gene dataset excluding RAG1
sequences ran to test the impact of the homoplasy problem
caused by convergence in CG content on the third position of
this gene (Gruber et al.2007).
Evolution of body mass
We obtained body mass data, in grams, from specific sources
(Mondolfi and Pérez-Hernández 1984; Catzeflis et al.1997;
Smith et al. 2003; Voss et al.2004;Rossi2005; Voss et al.
2005; Cáceres et al.2007;Floresetal.2008;Forero-Medina
and Vieira 2009;L
einerandSilva2009; Perez-Carusi et al.
2009; Voss and Jansa 2009; Lim et al.2010; Rossi et al.2010;
Gutiérrez et al.2011; Paglia et al.2012; Pavan et al.2012;
Voss et al.2012; Voss et al.2013; Barrera-Niño and Sánchez
2014;Martínez-Lanfrancoetal.2014; and Flores, Teta, Solari
and Voss, personal communication; Table SI.2). In addition,
we gathered data from 11 didelphid species housed in the
Colección Mamíferos Lillo (CML; Tucumán, Argentina;
Tab le SI.2and see Appendix 1for specimen vouchers,
Supplementary Information).
We approached the macroevolutionary analysis of body
mass in didelphids using two contrasting methodologies.
First, we aimed at assessing the global phylogenetic effect
on body mass in the group, by means of estimating the fraction
of total mass variance explained by significant tree partitions.
This method, Canonical Phylogenetic Ordination (hereafter
CPO; Giannini 2003), considers the comparative data as de-
pendent variable(s), here the vector of body mass data for each
didelphid species, in a linear model; tree partitions represent
the independent part of the model and are equivalent to clades
of a rooted tree. Tree partitions use a presence-absence
(binary) code to represent membership of species, and each
constitutes an explanatory variable that is tested using
Tab l e 2 Character sampling, detailing mitochondrial and nuclear
markers used, number of aligned base pairs (bp), and number of
sequences (=species) for each marker
Marker Number of aligned bp Number of included sequences
Mitochondrial
Cyt-b 1149 89
COI 657 36
12S rRNA 992 26
16S rRNA 1653 23
Subtotal 4451 174
Nuclear
IRBP 1158 57
BRCA1 2139 62
DMP1 1209 38
RAG1 2790 44
vWF 972 46
TTR 1770 25
Subtotal 10,038 272
Tot al 1 4, 48 9 446
L.I. Amador, N.P. Giannini

unrestricted Monte Carlo permutations (here, 4999 permuta-
tions were used). These tests of individual tree partitions are
followed by a forward stepwise selection procedure that seeks
to produce a maximally explanatory, but non-redundant model
that is a linear combination of selected partitions (clades).
These analyses were executed using the program CANOCO
version 4.0 (ter Braak and Šmilauer 1998).
Second, we chose to apply a local, node-by-node descrip-
tive approach that allowed an appropriate evaluation of evo-
lutionary changes in the framework proposed by Gould and
McFadden (2004). The body mass information was arranged
in TNT data matrix format and mapped as a continuous char-
acter (see Goloboff et al. 2006) with a single value (the species
average) per terminal, as implemented in TNT. Following
Giannini et al. (2012), we chose to interpret nodal location,
sign (increase or decrease), and magnitude of change, with the
latter being the net change or the amount of body mass in-
crease or decrease common to all reconstructions for a given
branch. We looked for macroevolutionary patterns in body
mass variation, in terms of phyletic or apomorphic changes,
giantism or nanism, and their corresponding combinations
(see Gould and McFadden 2004). To assess the persistence
of the phyletic pattern, we defined a simple metric, the order of
phyletic change, as the number of consecutive branches com-
posing the phyletic pattern, so the higher the order, the more
persistent the pattern; for instance, increases of body mass
along three consecutive branches represented a case of third-
order phyletic giantism.
Results
Phylogeny
Our molecular phylogeny includes the most comprehensive
taxonomic sampling to date, comprising 81 % of extant
didelphid diversity at the species level (i.e., 89 currently
recognized species; Table 1). The alignment produced
14,489 nucleotide characters distributed in 4451 and 10,038 bp
of mitochondrial and nuclear sequences, respectively (Table 2
and see Table SI.1for accession numbers). Clades recovered
from the various analyses in this study (see below) are summa-
rized in Table 3, and parenthetical trees are shown in Appendix 2
(Supplementary Information).
Individual genes
The nuclear genes BRCA1, IRBP, DMP1, and vWF and the
mitochondrial gen Cyt-b recovered all (or almost all) the phy-
logenetic structure of the total dataset analysis (see below
BTotal molecular evidence^section). RAG1 and TTR recov-
ered clades to approximately genus level, while 12S and 16S
so did at a higher taxonomic level (i.e., family, subfamilies).
Finally, COI exhibited a good performance chiefly within
Marmosini (sensu Steiner et al.2005).
Mitochondrial and nuclear datasets
Both the mitochondrial and nuclear analyses yielded highly
resolved, supported trees largely in agreement with the total
molecular evidence analysis (see below), but the latter recov-
ered more monophyletic groups. Except for the position of
Lestodelphys halli, which was nested within Thylamys, the
global structure of the nuclear dataset result was almost iden-
tical to that of total molecular evidence phylogeny (see be-
low), strongly suggesting that this data partition dominated the
combined analysis.
Total molecular evidence
The maximum likelihood (ML) best tree topology with boot-
strap values is shown in Fig. 2. Subfamilies Caluromyinae and
Didelphinae, as well as Marmosini, Didelphini, and
Thylamyini clades (sensu Steiner et al.2005), were recovered
as strongly supported monophyletic groups. Besides,
Didelphini and Thylamyini were reconstructed as sister
groups, and Hyladelphys kalinowskii was the basalmost taxon
of Didelphinae.
Didelphini was strongly supported, with Metachirus
nudicaudatus recovered as sister to the other genera subse-
quently branching in the order of Chironectes,Lutreolina,
and Philander +Didelphis. Within Thylamyini, Marmosops
was sister of a clade that included Gracilinanus +
Cryptonanus and Lestodelphys +Thylamys. In turn,
Thylamys consisted of five inner clades: (1) (Thylamys
velutinus,Thylamys karimii), (2) (Thylamys venustus
(Thylamys cinderella,Thylamys sponsorius)), (3) Thylamys
macrura, (4) (Thylamys tatei (Thylamys elegans,Thylamys
pallidior)), and (5) (Thylamys citellus (Thylamys pulchellus,
Thylamys pusillus)). Bootstrap values in Thylamyini were
very high (>80), with the exception of two small terminal
clades. Finally, Marmosini consisted of two main clades, the
specious genus Monodelphis and the clade formed by
Tlacuatzin +Marmosa. The latter included five clades, all
matching the subgenera recognized by Voss et al. (2014): (1)
subgenus Eomarmosa:Marmosa rubra, (2) subgenus
Exulomarmosa:(Marmosa simonsi (Marmosa robinsoni,
Marmosa xerophila)(Marmosa mexicana (Marmosa
zeledoni,Marmosa isthmica))), (3) subgenus Stegomarmosa:
Marmosa lepida, (4) subgenus Marmosa:((Marmosa
macrotarsus,Marmosa waterhousei)(Marmosa tyleriana,
Marmosa murina)), and (5) subgenus Micoureus:
((Marmosa constantiae,Marmosa regina)(Marmosa
paraguayanus (Marmosa alstoni,Marmosa demerarae))).
The majority of bootstrap values were high (>70) in these
groups. Monodelphis consisted of five inner clades: (1)
Phylogeny and evolution of body mass in didelphid marsupials

Tab l e 3 Clades recovered from different analyses following maximum likelihood (ML) and parsimony (MP) criteria
Clade Phylogenetic analyses
Individual genes (MP) Mitoc
(MP)
Nucl
(MP)
10 g
(MP)
10 g
(ML)
9g
(MP)
Cyt-b COI 12S 16S IRBP BRCA1 DMP1 RAG1 vWF TTR
Didelphidae X X X X X X X –XXXXX
Caluromyinae X X X X –X(+2) X X X X
Didelphinae X X X X X X X (−2) X X X X
Hyladelphis + (Marmosini, Didelphini, Thylamyini) –––XX X –XXXX
Marmosini X (+1) –XXX X X X X X X X
Marmosa XXX(−1)XXX XXXXX X X X X
Tlacuatzin +Marmosa –––XX X –XXX
Monodelphis XX–XXX XXX–XXXXX
Didelphini X X X X X X X X X X X X X
Didelphis X(−1) X (+1) X X (+1) X (+1) X X X X X
Philander X–XX X X(−1) X (−1) X X X X X
Didelphis +Philander XX X(+1)XX XXXXX(+3)X X X X
Thylamyini X (+4) X (−1) –XX X XX X X X X
Marmosops X–XX XXXXX(−1) X X X X
Cryptonanus X–––XX XXX–XXXXX
Gracilinanus X–––XX X(−1) X X –XXXXX
Cryptonanus +Gracilinanus ––– XXX –XXXX
Thylamys X––X
X XXXXX X(+1)X X X
Lestodelphis +Thylamys X–––X–X––XX X X X X
Cryptonanus +Gracilinanus +Lestodelphis +Thylamys –––X–X–––X(+1) X X X X
Didelphini + Thylamyini X (+3) X X X X X X X X X X
For the last, only implied weighting (IW) results are shown. Crosses correspond to monophyletic groups, with the number of taxa which have been misplaced outside of the group (−) or the number of
extraneous taxa added to the group (+) in brackets. Dashes indicate inapplicable cases
Mitoc mitochondrial, Nucl nuclear, ggenes
L.I. Amador, N.P. Giannini

Fig. 2 Best tree resulting from total dataset (ten genes) maximum
likelihood analysis. Bootstrap support values are indicated for each
node and are also reflected on degree of branch thickness. Asterisks:
Pavan et al.(2014) included Monodelphis theresa in Monodelphis
scalops,Monodelphis sorex in Monodelphis dimidiata,andMonodelphis
umbristriata in Monodelphis americana
Phylogeny and evolution of body mass in didelphid marsupials

(Monodelphis scalops,Monodelphis theresa), (2)
(Monodelphis sorex,Monodelphis dimidiata), (3)
(Monodelphis gardneri (Monodelphis iheringi (Monodelphis
umbristriata,Monodelphis americana))), (4) ((Monodelphis
domestica (Monodelphis glirina,Monodelphis sanctaerosae))
(Monodelphis arlindoi (Monodelphis touan (Monodelphis
brevicaudata,Monodelphis palliolata)))), and (5)
(Monodelphis kunsi ((Monodelphis adusta,Monodelphis
reigi)(Monodelphis peruviana (Monodelphis osgoodi,
Monodelphis handleyi)))). The Monodelphis groups them-
selves (and the relationships within them) were strongly sup-
ported (bootstrap values >75), but this was not the case with
their interrelationships (bootstrap values <50).
Parsimony analysis of the total dataset using implied
weighting (IW) is shown in Fig. SI.1(Supplementary
Information). The resulting topology is the strict consensus
of the three most parsimonious trees obtained, which shows
almost identical relationships as compared with the ML tree,
with just a few differences concentrated in small terminal
clades. The single politomy in the consensus involved four
species of Marmosops (Marmosops dorothea,Marmosops
invictus,Marmosops impavidus,andMarmosops noctivagus).
The parsimony total dataset analysis using equal weighting
(EW) generated 216 most parsimonious trees. The strict con-
sensus was quite similar to that of IW analysis, except for the
greater number of politomies. This was replicated in the re-
mainder of EW analysis. For that reason, we used only IW
topologies for the comparison of phylogenetical results
(Table 3).
Concatenated partition excluding RAG1
The global topology recovered with this analysis was very
similar to that of the total analysis. The main differences
consisted of the following: (1) the position of Tlacuatzin
canescens as sister to Marmosa +Monodelphis, instead of
the Marmosa clade; (2) the relative position among monophy-
letic groups within Monodelphis (although the groups
remained the same) and the position of Monodelphis emiliae
(in a more nested clade); and (3) the relative positions of
monophyletic groups within Thylamys (although the groups
remained the same) and position of Thylamys macrura (sister
to the Bvenustus^group).
Evolution of body mass
We obtained average values of body mass from the literature
and museum specimens that covered all terminal taxa
(Table SI.2). The log-transformed distribution of mass in
didelphids is shown in Fig. 3, with calculated mean = 179 g,
median= 51 g, and mode =17 g. This distribution was discon-
tinuous due to species of Didelphini and Caluromyinae ex-
cluding Glironia, which appeared as a separate group in
Fig. 3.
Phylogeny explained much of the total body mass varia-
tion, but the global, non-redundant effect of phylogeny on
body mass was quite restricted. Specifically, ten clades were
individually significant at alpha= 0.01 (defined to be interme-
diate between the true Bonferroni-corrected value and the
conventional 0.05 significance level). However, when
Fig. 2 continued.
L.I. Amador, N.P. Giannini

submitted to forward stepwise selection procedure, a single
clade was retained in the model (F= 206.17, P= 0.0002).
This clade was Didelphis, and the variation explained by this
single-variable model was as high as 69 %. This result repre-
sents the gross-scale effect of phylogeny on the body mass
variation in didelphids.
The fine-scale, detailed reconstruction of body mass evo-
lution is shown in Fig. 4. Total amount of change, i.e., the
reported length for the body mass character on the retrieved
phylogeny (steps), was 6760 g, but total net change, i.e., the
sum of changes common to all reconstructions, was 4921 g.
The ancestral node of Didelphidae was assigned states 22–
33 g. There was no net change along the backbone, so the first
changes to occur were reconstructed within each of the main
clades (i.e., Caluromyinae, Marmosini, Didelphini,
Thylamyini). The largest absolute increase in any didelphid
branch was located at the node of Didelphis virginiana (+
971 g), whereas the largest increase in relative terms was
reconstructed on the branch leading to Didelphini (967 % in-
crease in body mass with respect to the reconstructed hypo-
thetical ancestor). The largest absolute decrease was located at
the Didelphis imperfecta branch (−430 g), while in relative
terms, the largest decrease was reconstructed at the
Marmosa lepida branch (−264 % with respect to the
descendant).
A summary with frequency percentagesof stasis, increases,
and decreases is shown in Table 4. A prevalence of stasis was
observed both at global level (62.7 %) as well as within par-
ticular groups. With regard to the frequency of net changes,
although increases predominated over decreases globally,
both types of change were relatively balanced within groups,
with the exception of Caluromyinae where only increases
were reconstructed. However, the magnitude of increases
was twice as much (or more) as the magnitude of decreases,
both at global level as in particular groups (Table 5).
With respect to the observed patterns of body mass evolu-
tion, no clear global pattern (e.g., Cope’s rule or the like) was
detected. However, multiple cases of either phyletic and
apomorphic changes were reconstructed in particular clades
or branches, being the first type nearly twice as common as the
second type (Table 6). Considering these specific patterns, the
more frequent combinations were phyletic giantism and
apomorphic nanism, both at global level and within principal
clades (Table 6). The most notable cases of phyletic giantism
were, as expected, concentrated in the Didelphis clade
(Fig. 4), specifically Didelphis virginiana,Didelphis
marsupialis,andDidelphis pernigra, representing a fourth-
order case of phyletic giantism. In the other hand, the most
notable cases of apomorphic nanism were located in
Marmosini (Fig. 4), with Marmosa lepida (−37 g or 264 %
decrease with respect to its descendant) and Monodelphis
sanctaerosae (−43 g or 187 % decrease).
Discussion
In this research, we aimed at understanding body mass evolu-
tion in a group of small mammals ofour interest, the didelphid
marsupials. This is particularly important because didelphid
marsupials are morphologically highly integrated taxa (see
Porto et al.2009), meaning that they are evolutionarily less
flexible than most other mammals (Marroig et al.2009)sothat
their evolutionary responses to selection are basically aligned
with size variation (Marroig et al. 2009; Shirai and Marroig
2010). Macroevolutionary patterns, such as this one of body
mass evolution, can only be discerned in an explicit phyloge-
netic framework (Gould and MacFadden 2004;Honeetal.
2005; Butler and Goswami 2008). Thus, we contributed a
comprehensive, strong phylogeny based on unconstrained tree
searches that included 89 currently recognized didelphid spe-
cies. Previously, the most comprehensive phylogenetic hy-
pothesis of the group was the marsupial supertree provided
by Cardillo et al. (2004), containing 62 didelphid species;
however, this supertree lacks resolution particularly for New
World marsupials. Here, we carried out our phylogeny of
didelphids from a supermatrix approach, with no constraints
on recovered clades. This analysis contains at least twice as
many species included in recent supermatrix phylogenies
(e.g., Jansa et al.2006;Gruberetal.2007; Voss-Jansa 2009;
Flores et al.2009;Astúa2009; May-Collado et al.2015). We
were able to provide a fully resolved, highly supported phy-
logenetic hypothesis for the Didelphidae. This result is
Fig. 3 Log-transformed distribution of body mass in Didelphidae
Phylogeny and evolution of body mass in didelphid marsupials

somewhat unexpected given the level of character conflict
apparent in earlier studies of didelphid systematics (e.g.,
Kirsch 1977; Kirsch and Archer 1982; Kirsch and Palma
1995;Reigetal.1987); by contrast, our phylogeny and its
systematic implications were generally consistent with all re-
cent analyses, both at a suprageneric and infrageneric level.
Moreover, the two distinct analyses performed (ML and par-
simony) produced almost identical results. The tree reflected,
for each major group, the resolution found in specific previous
studies (see below). Although this may seem trivial given that
sequences came from those studies, results need not to be
congruent globally when data are gathered together from dif-
ferent partial studies attacking specific subgroups, but we ob-
tained a remarkable result of high congruence when all taxa
and data were analyzed together, as we discussed next. We
recovered Hyladelphys as sister to all didelphids as in Jansa
and Voss (2005), and the internal structure of the remaining
clades, particularly Marmosa, and the recently proposed sub-
division in five subgenera (see Gutiérrez et al.2010;dela
Sancha et al.2012;Vossetal.2014), Monodelphis
Fig. 4 Evolution of body mass in Didelphidae. Reconstructed values are shown below branches. Black arrows denote increases and gray ones,
decreases. Body mass values in grams for taxa are indicated in brackets
L.I. Amador, N.P. Giannini

(Caramaschi et al.2011;Carvalhoetal.2011; Pavan et al.
2012,2014), Philander (Chemisquy and Flores 2012), and
Thylamys (Giarla et al.2010; Giarla and Jansa 2014;Palma
et al. 2014). Our tree differs slightly from others only in poorly
supported clades (e.g., inside Monodelphis; see Vilela et al.
2015). This means that the data accrued to date by many
independent researchers proved to be of high quality and
congruence. It should be noted that the global topology was
not significantly modified when RAG1 sequences were
excluded from the supermatrix, thus contrasting with Gruber
et al. (2007) whose results substantially improved with the
exclusion of the RAG1 third position. This might be due to
the fact that homoplasy by convergence in CG content on this
position, which Gruber et al. (2007) claimed to cause the
Fig. 4 continued.
Phylogeny and evolution of body mass in didelphid marsupials

inconsistent topology problem in the group, would be com-
pensated by the congruent signal of many other independent
characters. In summary, we simultaneously tested, with more
taxa and characters, the monophyly of all didelphid groups
proposed to date, found strong support for most previously
recognized groups in a single, robust phylogenetic hypothesis,
and thus offered a strong confirmation to the current classifi-
cation of the Didelphidae. This provides a firm basis for future
taxonomic, evolutionary, and biogeographic studies in the
group.
A first application of this remarkable phylogenetic result is
precisely our analysis of body mass evolution. In spite of
being small and therefore limited to the lower range of the
whole mammalian variation, didelphids revealed considerable
and highly interesting variation. This is both in terms of com-
parisons between observed extremes, with a ca. 289-fold dif-
ference between the average mass of smallest and largest spe-
cies, and in evolutionary terms (see below). We detected one
major tree partition with large, significant differences between
the set of its respective members, specifically Didelphis,
which explained as much as 69 % of total interspecific average
mass variation in the group. We also reconstructed, by means
of character mapping, many changes that allowed a detailed
analysis of body mass evolution in the group. The ancestral
interval, estimated at 22–33 g, is quite narrow, so the confi-
dence on its reconstructed range increases. From this small
root value, dated at a point estimate of 31.4 mya in the
Oligocene (or range estimated in 23–38 mya, between late
Eocene and Oligocene; Meredith et al.2011), didelphids
evolved as much as 6760 g of total mass changes. This sce-
nario differs from a larger previous estimate of ancestral body
mass based on morphometric data, corresponding to the mass
of Marmosa (subgenus Micoureus)orGlironia (Astúa 2009),
that is between 75 and 130 g (Table SI.2). In all likelihood, this
difference with our ancestral estimate (at 22–33 g, see above)
may be due to the fact that the taxonomic sample in Astúa
(2009) lacked small outgroups (caenolestids not included) and
small relevant basal taxa (particularly Hyladelphis;16g,
Tab le SI.2). The amount of evolutionary change that we re-
constructed did not unfold in a specific pattern for didelphids
as a group. We see, as discussed next, that evolution of body
mass was group-specific, and this reflects, within a particular
clade of mammals, a pattern found at the level of the Class
(e.g., Venditti et al.2011).
The total magnitude of changes in absolute terms was
greater for increases than for decreases, as expected. This bias
Tabl e 4 Results from body mass
optimization Clade Stasis
frequency
Increases
frequency
Decreases
frequency
Tot al n um be r of nodes
a.f. % a.f. % a.f. %
Family Didelphidae 111 62.7 38 21.5 28 15.8 177
Subfamily Caluromyinae 5 55.6 4 44.4 0 0.0 9
Subfamily Didelphinae 105 62.9 34 20.4 28 16.8 167
Marmosini 49 62.0 18 22.8 12 15.2 79
Didelphini 15 55.6 7 25.9 5 18.5 27
Thylamyini 38 66.7 9 15.8 10 17.5 57
Frequencies of stasis, increases, and decreases are indicated for each main Didelphidae clade, both in absolute
terms (a.f. absolute frequency) and in relative terms (%)
Tabl e 5 Results from body mass
optimization (cont.) Clade Increases magnitude Decreases magnitude Total
a.v. (g) % a.v. (g) %
Family Didelphidae 3599 73.14 1322 26.86 4921
Subfamily Caluromyinae 512 100.00 0 0.00 512
Subfamily Didelphinae 3087 70.02 1322 29.98 4409
Marmosini 312 66.24 159 33.76 471
Didelphini 2660 70.84 1095 29.16 3755
Thylamyini 115 64.97 62 35.03 177
Magnitudes of increases and decreases are indicated for each main Didelphidae clade, both in absolute terms (a.v.
absolute value, in g) and in relative terms (%)
L.I. Amador, N.P. Giannini

toward a greater magnitude of increases is not an artifact of the
reconstruction technique but a property of the character scale
(restricted to the set of positive rational numbers). Large in-
creases are more likely to occur than large decreases; e.g.,
from any small value, like the root at 27.5 g (middle between
22 and 33 g), it is possible to have a large evolutionary in-
crease of, say 200 g, thus producing a descendant at 227.5 g,
but it is not possible to have the same absolute amount of
decrease. By contrast, the frequencies of increases and de-
creases were relatively balanced within groups (except in
Caluromyinae). A major difference was though that increases
(giantism in Gould and McFadden’s(2004)terminology)
were arranged more commonly in a phyletic pattern, while
decreases (nanism) were more often single, or apomorphic,
changes. In this way, the largest values of body mass observed
in extant Didelphidae were reconstructed as the cumulative
evolution along successive branches through time, while the
smallest values of body mass were often the product of single
evolutionary events. Given this pattern, the most persistent
case of phyletic giantism was that of Didelphis.Particularly
in Didelphis virginiana, four changes along six branches
added a total of 2032 g, that is, a ca. 62-fold increase from
the ancestral Didelphini and a ca. 80-fold increase between the
reconstructed ancestral body mass at 22–33 g (mean 27.5 g)
and the largest terminal (Didelphis virginiana at average
2195 g). Still, 82 % of that change occurred at the ancestral
Didelphis and its descendant species. If we assume a basal
position to the oldest known Didelphis specimen, which is
also the smallest member of the genus, of late Miocene
(Huayquerian) age from the Solimões formation in the
Brazilian Amazon (Cozzuol et al.2006), it shows that this
trend of increasing body mass has persisted for at least 9–
6.8 my. The rate of change, estimated between 2032 g/
9 my = 225.7 g my
−1
and 2032 g/6.8 my = 298.8 g my
−1
,
was very high (4-to-6-fold the median value for the family at
51 g). This trend is consistent with fast but gradual (phyletic)
change and suggests directional selection for increased body
mass sustained over substantial geological time. This also
holds if extinct forms (e.g., the largest known didelphid, the
late Pliocene Thylophorops lorenzinii,Goinetal.2009)are
considered.
This evolution must have affected the biology of Didelphis
at a critical geologic age. Modern Didelphis appeared shortly
before the mammalian peak of geodispersal during the Great
American Biotic Interchange (GABI; Cozzuol et al.2006;also
see Bacon et al.2015) and must have speciated during the
latest GABI phases. At this time, placental carnivores gradu-
ally entered South America, and no other marsupials as large
or larger than Didelphis escaped extinction; Didelphis not
0
An anonymous reviewer suggested that large body mass may also be
physiologically advantageous for colonization of cold areas, as is the case
of Didelphis virginiana in its northern distributional limit.
Tab l e 6 Patterns of body mass evolution in terms of phyletic giantism (PG), phyletic nanism (PN), apomorphic giantism (AG), and apomorphic nanism (AN)
Clade Total changes Phyletic changes Apomorphic changes
PG PN Total AG AN Total
a.f.% a.f.% a.f.% a.f.%
Family Didelphidae 66 32 48.5 8 12.1 40 6 9.1 20 30.3 26
Subfamily Caluromyinae 4 4 100.0 0 0.0 4 0 0.0 0 0.0 0
Subfamily Didelphinae 62 28 45.2 8 12.9 36 6 9.7 20 32.3 26
Marmosini 30 16 53.3 2 6.7 18 2 6.7 10 33.3 12
Didelphini 12 7 58.3 0 0.0 7 0 0.0 5 41.7 5
Thylamyini 19 5 26.3 6 31.6 11 4 21.1 4 21.1 8
Numbers indicate the absolute frequency (a.f.) and the percentage (%) of changes under the different patterns for each main didelphid clade
Phylogeny and evolution of body mass in didelphid marsupials

only survived, but also speciated and even expanded its range
to North America (Giannini et al.2011, and Citations therein).
This marsupial clade accumulated defensive morphologies
and skills (e.g., presence of hypertrophied spinous processes
in cervical vertebrae that function as neck armor, catatonia or
death feigning, secretion of anal glands, etc.; see Giannini
et al.2011), shifted to more generalized feeding and locomo-
tor habits, and achieved a relatively large body mass as com-
pared with most of their ancestors and extinct close relatives
(cf. Goin et al.2009). Possibly, this body mass permitted to be
large enough for accessing vertebrate prey and for promoting
the development of efficient defensive traits
1
but, at the same
time, to be still small enough for avoiding potential placental
omnivore competitors (e.g., Procyonidae), and for maintain-
ing those generalized habits (Astúa 2009). Thus, we speculate
that phyletic giantism in didelphids is associated with long-
term, persistent changes in diet (e.g., from insectivory to in-
creased frugivory or omnivory, as in caluromines and
didelphines) and/or locomotion (see Astúa 2009). By contrast,
evolution of body mass showed that the most prominent de-
creases were more recent (i.e., near the terminals), specifically
within Marmosini (see above). Apomorphic nanism is possi-
bly related to specific limitations (e.g., physiological or eco-
logical) for reducing mass in a mammal whose ancestor was
already small.
Previous studies on marsupials relate body mass with risk
of extinction (e.g., Cardillo and Bromhan 2001;Fisheretal.
2003; Johnson and Isaac 2009). These investigations
attempted to test the controversial hypothesis of Bcritical
weight range^(CWR) proposed by Burbidge and McKenzie
(1989), who stated that species of non-volant terrestrial mam-
mals of intermediate body mass (between 35 and 5500 g) have
declined mostseverely. However, these studies were restricted
to Australasian species, and the picture seems to be different
for American marsupial species. On preliminary grounds, the
surviving species have a small to intermediate body mass
(being the lastsuccessfully widespread), while the large clades
(order Sparassodonta) became all extinct (see above). The
traditional explanation for this pattern has been the competi-
tion with placental carnivorans, which entered from North
America during the GABI and caused the decline and extinc-
tion of large carnivorous marsupials. However, this idea has
been questioned, criticized, or rejected in recent years, as there
is no evidence of temporal overlap between carnivorans and
their sparassodont ecological counterparts (Prevosti et al.
2013), so apparently, carnivorans occupied a niche left vacant
by extinct sparassodonts (a passive replacement pattern), as
already advanced by Marshall (1978)andcommentedby
Vieira and Astúa de Moraes (2003). Whatever the cause, the
association between risk of extinction and body mass in
American marsupial species appears to follow a different pat-
tern from that of Australian species and deserves further
investigation.
In conclusion, here, we provided a robust phylogenetic
basis for evolutionary comparisons of relevant life history
traits in a group of particular biological interest, the largest
New World radiation of marsupials, the didelphids. On this
basis, we revealed interesting patterns of body mass evolution
in a group of typical small mammals, including no clade-wide
dominating trend, predominance of phyletic increases and
apomorphic decreases, and persistent clade-specific trends
consistent with directional selection over extended periods
of geological time. This is particularly relevant given the con-
straint imposed by mass on morphological evolutionary re-
sponses to directional selection in highly integrated mammals,
such as didelphids (Marroig et al.2009). We conclude that
evolutionary body mass variation of substantial interest may
reside in groups of small mammals; this should be explored
further in more groups in the light of robust phylogenetic
reconstructions, as demonstrated here with didelphids.
Acknowledgments We thank Diego Astúa, David Flores, Sergio
Solari, Pablo Teta, and Robert Voss for providing us with unpublished
body mass data and Rubén Barquez and Mónica Díaz (CML) for granting
access to specimens under their care. Special thanks to Santiago Catalano
for his help with using his program GB-to-TNT. We acknowledge support
from Consejo Nacional de Investigaciones Científicas y Técnicas
(CONICET), Argentina, and grant PICT 2008–1798 to NPG.
Data archiving Data are archived in TreeBASE (http://purl.org/phylo/
treebase/phylows/study/TB2:S17654)
References
Argot, C. (2003). Functional adaptations of the postcranial skeleton of
two Miocene borhyaenoids (Mammalia, Metatheria), Borhyaena
and Prothylacinus,fromSouthAmerica.Palaeontology, 46, 1213–
1267.
Argot, C. (2004a). Functional-adaptive features and paleobiologic impli-
cations of the postcranial skeleton of the late Miocene sabretooth
borhyaenoid Thylacosmilus atrox (Metatheria). Alcheringa, 28,
229–266.
Argot, C. (2004b). Functional-adaptive analysis of the postcranial skele-
ton of a Laventan Borhyaenoid, Lycopsis longirostris (Marsupialia,
Mammalia). Journal of Vertebrate Paleontology, 24,689–708.
Astúa, D. (2009). Evolution of scapula size and shape in didelphid mar-
supials (Didelphimorphia: Didelphidae). Evolution, 63(9), 2438–
2456.
Astúa, D. (2015). Family Didelphidae (Opossums). In: D.E. Wilson and
R.A. Mittermeier (Orgs.), Handbook of the mammals of the world,
1st ed., Vol. 5 - Monotremes and Marsupials, pp: 70–18. Barcelona,
Lynx Editions.
Bacon, C. D., Silvestro, D., Jaramillo, C., Smith, B. T., Chakrabarty, P., &
Antonelli, A. (2015). Biological evidence supports an early and
complex emergence of the Isthmus of Panama. Proceedings of the
National Academy of Sciences, 112(19), 6110–6115.
Barrera-Niño, V., & Sánchez, F. (2014). Forrajeo de Didelphis pernigra
(Mammalia: Didelphidae) en un área suburbana de la Sabana de
Bogotá, Colombia. Therya, 5(1), 289–302.
Birney, E.C., & Monjeau, J.A. (2003). Latitudinal patterns in South
American marsupial biology. Carnivorous with pouches: biology
of carnivorous marsupials,293–313.
L.I. Amador, N.P. Giannini

Burbidge, A. A., & McKenzie, N. L. (1989). Patterns in the modern
decline of Western Australia’s vertebrate fauna: causes and conser-
vation implications. Biological Conservation, 50(1), 143–198.
Butler, R. J., & Goswami, A. (2008). Body size evolution in Mesozoic
birds: little evidence for Cope’srule.Journal of Evolutionary
Biology, 21(6), 1673–1682.
Cáceres, N. C., Napoli, R. P., Lopes, W. H., Casella, J., & Gazeta, G. S.
(2007). Natural history of the marsupial Thylamys macrurus
(Mammalia, Didelphidae) in fragments of savannah in southwestern
Brazil. Journal of Natural History, 41(29–32), 1979–1988.
Caramaschi, F. P., Nascimento, F. F., Cerqueira, R., & Bonvicino, C. R.
(2011). Genetic diversity of wild populations of the grey short-tailed
opossum, Monodelphis domestica (Didelphimorphia: Didelphidae),
in Brazilian landscapes. Biological Journal of the Linnean Society,
104(2), 251–263.
Cardillo, M., & Bromham, L. (2001). Body size and risk of extinction in
Australian mammals. Conservation Biology, 15(5), 1435–1440.
Cardillo, M., Bininda‐Emonds, R. P., Boakes, E., & Purvis, A. (2004). A
species-level phylogenetic supertree of marsupials. Journal of
Zoology, 264(1), 11–31.
Carvalho, B.D. A., Oliveira, L. F., Langguth, A., Freygang, C. C., Ferraz,
R. S., & Mattevi, M. S. (2011). Phylogenetic relationships and phy-
logeographic patterns in Monodelphis (Didelphimorphia:
Didelphidae). Journal of Mammalogy, 92(1), 121–133.
Catzeflis, F., Richard-Hansen, C., Fournier-Chambrillon, C., Lavergne,
A., & Vié, J. (1997). Biométrie, reproduction et sympatrie chez
Didelphis marsupialis et D. albiventris en Guyane française
(Didelphidae: Marsupialia). Mammalia, 61(2), 231–244.
Cerqueira, R., & Tribe, C. J. (2007). Genus Didelphis Linnaeus, 1758.
Mammals of South America, 1,17–25.
Chemisquy, M. A., & Flores, D. A. (2012). Taxonomy of the southern-
most populations of Philander (Didelphimorphia, Didelphidae),
with implications for the systematics of the genus. Zootaxa, 3481,
60–72.
Chemisquy, M. A., Prevosti, F. J., Martin, G., & Flores, D. A. (2015).
Evolution of molar shape in didelphid marsupials (Marsupialia:
Didelphidae): analysis of the influence of ecological factors and
phylogenetic legacy. Zoological Journal of the Linnean Society,
173(1), 217–235.
Cooper, N., & Purvis, A. (2010). Body size evolution in mammals: com-
plexity in tempo and mode. The American Naturalist, 175(6), 727–
738.
Cozzuol, M. A., Goin, F., Reyes, M. D. L., & Ranzi, A. (2006). The oldest
species of Didelphis (Mammalia, Marsupialia, Didelphidae), from
the late Miocene of Amazonia. JournalofMammalogy,87(4), 663–
667.
Crompton, A. W., & Hiiemae, K. (1970). Molar occlusion and mandib-
ular movements during occlusion in the American opossum,
Didelphis marsupialis. Zoological Journal of the Linnean Society,
49,21–47.
de la Sancha, N. U., D’Elía, G., & Teta, P. (2012). Systematics of the
subgenus of mouse opossums Marmosa
(Micoureus)(Didelphimorphia, Didelphidae) with noteworthy re-
cords from Paraguay. Mammalian Biology-Zeitschrift für
Säugetierkunde, 77(4), 229–236.
Dickman, C.R. (2005). Introduction. In R.M. Nowack (ed.), Walker’s
marsupials of the world. Johns Hopkins University Press, 226 pp.
Dickman, C.R., & Vieira, E. M. (2005). Ecology and life histories. In P. J.
Armati,C.R.Dickman,&I.D.Hume(Eds.),Marsupials.
Cambridge: Cambridge University Press.
Fisher, D. O., Blomberg, S. P., & Owens, I. P. (2003). Extrinsic versus
intrinsic factors in the decline and extinction of Australian marsu-
pials. Proceedings of the Royal Society of London B: Biological
Sciences, 270(1526), 1801–1808.
Flores, D. A. (2009). Phylogenetic analyses of postcranial skeletal mor-
phology in didelphid marsupials. Bulletin of the American Museum
of Natural History, 320,1–81.
Flores, D. A., Barquez, R. M., & Díaz, M. M. (2008). A new species of
Philander Brisson, 1762 (Didelphimorphia, Didelphidae).
Mammalian Biology-Zeitschrift für Säugetierkunde, 73(1), 14–24.
Forero-Medina, G., & Vieira, M. V. (2009). Perception of a fragmented
landscape by neotropical marsupials: effects of body mass and en-
vironmental variables. Journal of Tropical Ecology, 25(01), 53–62.
Fox, R. C. (1987). Paleontology and the early evolution of marsupials.
Possums and Opossums: Studies in Evolution, 1,161–169.
Gardner, A.L. (2005). Order Didelphimorphia, Paucituberculata and
Microbiotheria. In D.E. Wilson and D.M. Reeder (eds.), Mammals
species of the world: a taxonomic and geographic reference,3rded.,
Vol. 1, 743 pp. Johns Hopkins University Press.
Gardner, A.L. (2008). Cohort Marsupialia. In A.L. Gardner (ed.),
Mammals of South America, Volume 1, marsupials, xenarthrans,
shrews, and bats, 669 pp. The University of Chicago Press.
Giannini, N. P. (2003). Canonical phylogenetic ordination. Systematic
Biology, 52(5), 684–695.
Giannini,N. P., & Bertelli, S. (2004). Phylogeny of extant penguins based
on integumentary and breeding characters. The Auk, 121(2), 422–
434.
Giannini, N. P., Gaudioso, P., Flores, D. A., & Gaudin, T. J. (2011). A
possible function for an enigmatic synapomorphy of Didelphis.
Mammalian Biology-Zeitschrift für Säugetierkunde, 76(4), 512–
514.
Giannini, N.G., Gunnell, G.F., Habersetzer, J. & Simmons, N.B. (2012).
Early evolution of body size in bats. In G.F. Gunnell and N.B.
Simmons (eds.), Evolutionary history of bats: fossils, molecules,
and morphology (530–555 pp). Cambridge University Press.
Giarla, T. C., & Jansa, S. A. (2014). The role of physical geography and
habitat type in shaping the biogeographical history of a recent radi-
ation of Neotropical marsupials (Thylamys: Didelphidae). Journal of
Biogeography, 41(8), 1547–1558.
Giarla, T.C., Voss, R.S., & Jansa, S.A. (2010). Species limits and phylo-
genetic relationships in the didelphid marsupial genus Thylamys
based on mitochondrial DNA sequences and morphology. Bulletin
of the American Museum of Natural History,1–67.
Goin, F. J., & Pardiñas, U. F. J. (1996). Revisión de las especies del
género Hyperdidelphis Ameghino 1904 (Mammalia, Marsupialia,
Didelphidae). Su significación filogenética, estratográfica, y
adaptativa en el Neógeno del Cono Sur sudamericano. Estudios
Geológicos, 52,327–359.
Goin, F. J., Zimicz, N., de los Reyes, M., & Soibelzon, L. (2009). A new
large didelphid of the genus Thylophorops (Mammalia:
Didelphimorphia: Didelphidae) from the Late Tertiary of the
Pampean Region (Argentina). Zootaxa, 2005,35–46.
Goloboff, P. A. (1993). Estimating character weights during tree search.
Cladistics, 9(1), 83–91.
Goloboff, P. A. (2014). Extended implied weighting. Cladistics, 30(3),
260–272.
Goloboff, P. A., & Catalano, S. A. (2012). GB‐to‐TNT: facilitating crea-
tion of matrices from GenBank and diagnosis of results in TNT.
Cladistics, 28(5), 503–513.
Goloboff, P. A., Mattoni, C. I., & Quinteros, A. S. (2006). Continuous
characters analyzed as such. Cladistics, 22(6), 589–601.
Goloboff, P. A., Carpenter, J. M., Arias, J. S., & Esquivel, D. R. M.
(2008a). Weighting against homoplasy improves phylogenetic anal-
ysis of morphological data sets. Cladistics, 24(5), 758–773.
Goloboff, P. A., Farris, J. S., & Nixon, K. C. (2008b). TNT, a free pro-
gram for phylogenetic analysis. Cladistics, 24(5), 774–786.
Gould, G. C., & MacFadden, B. J. (2004). Gigantism, dwarfism, and
Cope’srule:Bnothing in evolution makes sense without a
phylogeny^.Bulletin of the American Museum of Natural History,
285,219–237.
Phylogeny and evolution of body mass in didelphid marsupials

Groves, C.P. (2005). Order Notoryctemorphia, Dasyuromorphia,
Peramelemorphia and Diprotodontia. In D.E. Wilson and D.M.
Reeder (eds.), Mammals species of the world: a taxonomic and
geographic reference, 3rd ed., Vol. 1, 743 pp. Wilson Johns
Hopkins University Press.
Gruber, K. F., Voss, R. S., & Jansa, S. A. (2007). Base-compositional
heterogeneity in the RAG1 locus among didelphid marsupials: im-
plications for phylogenetic inference and the evolution of GC con-
tent. Systematic Biology, 56(1), 83–96.
Gutiérrez, E.E., Jansa, S.A., & Voss, R.S. (2010). Molecular systematics
of mouse opossums (Didelphidae: Marmosa): assessing species
limits using mitochondrial DNA sequences, with comments on phy-
logenetic relationships and biogeography. American Museum
Novitates,1–22.
Gutiérrez, E. E., Soriano, P. J., Rossi, R. V., Murillo, J. J., Ochoa-G, J., &
Aguilera, M. (2011). Occurrence of Marmosa waterhousei in the
Venezuelan Andes, with comments on its biogeographic signifi-
cance. Mammalia, 75(4), 381–386.
Hone, D. W. E., Keesey, T. M., Pisani, D., & Purvis, A. (2005).
Macroevolutionary trends in the Dinosauria: Cope’s rule. Journal
of Evolutionary Biology, 18(3), 587–595.
Horovitz, I., & Sánchez‐Villagra, M. R. (2003). A morphological analysis
of marsupial mammal higher-level phylogenetic relationships.
Cladistics, 19(3), 181–212.
Jansa, S. A., & Voss, R. S. (2005). Phylogenetic relationships of the
marsupial genus Hyladelphys based on nuclear gene sequences
and morphology. Journal of Mammalogy, 86(5), 853–865.
Jansa, S. A., Forsman, J. F., & Voss, R. S. (2006). Different patterns of
selection on the nuclear genes IRBP and DMP-1 affect the efficiency
but not the outcome of phylogeny estimation for didelphid marsu-
pials. Molecular Phylogenetics and Evolution, 38(2), 363–380.
Johnson, C. N., & Isaac, J. L. (2009). Body mass and extinction risk in
Australian marsupials: the ‘critical weight range’revisited. Austral
Ecology, 34(1), 35–40.
Katoh, K., & Standley, D. M. (2013). MAFFT multiple sequence align-
ment software version 7: improvements in performance and usabil-
ity. Molecular Biology and Evolution, 30(4), 772–780.
Kirsch, J. A. W. (1977). The comparative serology of Marsupialia, and a
classification of marsupials. Australian Journal of Zoology, 52,1–
152.
Kirsch, J.A.W., & Archer, M. (1982). Polythetic cladistics, or, when
parsimony’s not enough: the relationships of carnivorous marsu-
pials. In M. Archer (ed.), Carnivorous marsupials, Vol. 2, 595–
620 pp. Mosman, NSW: Royal Society of New South Wales.
Kirsch, J. A. W., & Palma, R. E. (1995). DNA/DNA hybridization studies
of carnivorous marsupials. V. A further estimate of relationships
among opossums (Marsupialia: Didelphidae). Mammalia, 59,
403–425.
Leiner, N. O., & Silva, W. R. (2009). Short communication: territoriality
in females of the slender opossum (Marmosops paulensis)inthe
Atlantic forest of Brazil. JournalofTropicalEcology,25,671–675.
Lim, B. K., Engstrom, M. D., Patton, J. C., & Bickham, J. W. (2010).
Molecular phylogenetics of Reig’s short-tailed opossum
(Monodelphis reigi) and its distributional range extension into
Guyana. Mammalian Biology-Zeitschrift für Säugetierkunde,
75(4), 287–293.
Lomolino, M. V. (2005). Body size evolution in insular vertebrates: gen-
erality of the island rule. Journal of Biogeography, 32(10), 1683–
1699.
Marroig, G., Shirai, L. T., Porto, A., de Oliveira, F. B., & De Conto, V.
(2009). The evolution of modularity in the mammalian skull II:
evolutionary consequences. Evolutionary Biology, 36(1), 136–148.
Marshall, L. G. (1978). Evolution of the Borhyaenidae, extinct South
American predaceous marsupials. University of California
Publications in Geological Sciences, 117,1–87.
Martínez-Lanfranco, J. A., Flores, D., Jayat, J. P., & D’Elía, G. (2014). A
new species of lutrine opossum, genus Lutreolina Thomas
(Didelphidae), from the South American Yungas. Journal of
Mammalogy, 95(2), 225–240.
May-Collado, L., Kilpatrick, C. W., & Agnarsson, I. (2015). Mammals
from Bdown under^: a multi-gene species-level phylogeny of mar-
supial mammals. PeerJ, 3, e805. doi:10.7717/peerj.805.
McKenna, M.C., & Bell, S.K. (1997). Classification of mammals above
the species level. M.C. McKenna and S.K. Bell (eds.), 631 pp.
Columbia University Press.
McManus, J. J. (1974). Didelphis virginiana. Mammalian Species, 40,1–
6.
McNab, B.K. (2007). The evolution of energetics in birds and mammals.
In D.A. Kelt, E.P. Lessa, J. Salazar-Bravo, and J.L. Patton (eds.), The
quintessential naturalist: honoring the life and legacy of Oliver P.
Pearson,67–110 pp. University of California Publications.
Meiri, S., Meijaard, E., Wich, S. A., Groves, C. P., & Helgen, K. M.
(2008). Mammals of Borneo—smallsizeonalargeisland.
Journal of Biogeography, 35(6), 1087–1094.
Meredith, R. W., Janečka, J. E., Gatesy, J., Ryder, O. A., Fisher, C. A.,
Teeling, E. C., Goodbla, A., Eizirik, E., Simão, T. L. L., Stadler, T.,
Rabosky, D. L., Honeycutt, R. L., Flynn, J. J., Ingram, C. M.,
Steiner, C., Williams, T. L., Robinson, T. J., Burk-Herrick, A.,
Westerman, M., Ayoub, N. A., Springer, M. S., & Murphy, W. J.
(2011). Impacts of the Cretaceous Terrestrial Revolution and KPg
extinction on mammal diversification. Science, 334(6055), 521–
524.
Mondolfi, E., & Pérez-Hernández, R. (1984). Una nueva subespecie de
zarigüeya del grupo Didelphis albiventris (Mammalia: Marsupialia).
Acta Científica Venezolana, 35,407–413.
Paglia, A.P., Fonseca, G.A.B., Rylands, A.B., Herrmann, G., Aguiar,
L.M.S., Chiarello, A.G., Leite, Y.L.R., Costa L.P., Siciliano, S.,
Kierulff, M.C.M., Mendes, S.L., Tavares, V. da C., Mittermeier,
R.A. y Patton, J.L. (2012). Lista Anotada dos Mamíferos do Brasil
/ Annotated Checklist of Brazilian Mammals. 2ª Edição / 2nd
Edition. Occasional Papers in Conservation Biology,6,76pp.
Palma, R. E., & Spotorno, A. E. (1999). Molecular systematics of mar-
supials based on the rRNA 12S mitochondrial gene: the phylogeny
of Didelphimorphia and of the living fossil microbiotheriid
Dromiciops gliroides Thomas. Molecular Phylogenetics and
Evolution, 13(3), 525–535.
Palma, R. E., Boric-Bargetto, D., Jayat, J. P., Flores, D. A., Zeballos, H.,
Pacheco, V., Cancino, R. A., Alfaro, F. D., Rodríguez-Serrano, E. &
Pardiñas, U. F. J. (2014). Molecular phylogenetics of mouse opos-
sums: new findings on the phylogeny of Thylamys
(Didelphimorphia, Didelphidae). Zoologica Scripta, 43,217–234.
Pavan, S. E., Rossi, R. V., & Schneider, H. (2012). Species diversity in the
Monodelphis brevicaudata complex (Didelphimorphia:
Didelphidae) inferred from molecular and morphological data, with
the description of a new species. Zoological Journal of the Linnean
Society, 165(1), 190–223.
Pavan, S. E., Jansa, S. A., & Voss, R. S. (2014). Molecular phylogeny of
short-tailed opossums (Didelphidae: Monodelphis): taxonomic im-
plications and tests of evolutionary hypotheses. Molecular
Phylogenetics and Evolution, 79,199–214.
Perez Carusi, L. C., Farace, M. I., Ribicich, M. M., & Gómez Villafañe, I.
E. (2009). Reproduction and parasitology of Didelphis albiventris
(Didelphimorphia) in an agroecosystem landscape in central
Argentina. Mammalia, 73(2), 89–97.
Porto, A., de Oliveira, F. B., Shirai, L. T., De Conto, V., & Marroig, G.
(2009). The evolution of modularity in the mammalian skull I: mor-
phological integration patterns and magnitudes. Evolutionary
Biology, 36(1), 118–135.
Prevosti, F. J., Forasiepi, A., & Zimicz, N. (2013). The evolution of the
Cenozoic terrestrial mammalian predator guild in South America:
L.I. Amador, N.P. Giannini

competition or replacement? Journal of Mammalian Evolution,
20(1), 3–21.
Reig, O.A, Kirsch, J.A.W., & Marshall, L.G. (1987). Systematic relation-
ships of the living and neocenozoic American Bopossum-like^mar-
supials (Suborder Didelphimorphia), with comments on the classi-
fication of these and of the Cretaceous and Paleogene New World
and European metatherians. In M. Archer (ed.), Possums and opos-
sums: studies in evolution,Vol.1,1–89 pp. Sydney: Surrey Beatty.
Rossi, R. V. (2005). Revisão taxonômica de Marmosa Gray, 1821
(Didelphimorphia: Didelphidae). Tese de Doutorado:
Universidade de São Paulo, São Paulo.
Rossi, R.V., Voss, R.S., & Lunde, D.P. (2010). A revision of the
Didelphid Marsupial Genus Marmosa. Part 1. The species in
Tat e’s‘mexicana’and ‘mitis’sections and other closely related
forms. Bulletin of the American Museum of Natural History,1–83.
Shirai, L. T., & Marroig, G. (2010). Skull modularity in neotropical mar-
supials and monkeys: size variation and evolutionary constraint and
flexibility. Journal of Experimental Zoology Part B: Molecular and
Developmental Evolution, 4(8), 663–683.
Simpson, G. G. (1972). Didelphidae from the Chapadmalal Formation in
the Museo Municipal de Ciencias Naturales of Mar del Plata.
Revista del Museo Municipal de Ciencias Naturales de Mar del
Plata, 2,1–40.
Slater, G. J. (2013). Phylogenetic evidence for a shift in the mode of
mammalian body size evolution at the Cretaceous-Palaeogene
boundary. Methods in Ecology and Evolution, 4(8), 734–744.
Smith, F. A., Lyons, S. K., Ernest, S. M., Jones, K. E., Kaufman, D. M.,
Dayan, T., Marquet, P. A., Brown, J. H., & Haskell, J. P. (2003).
Body mass of Late Quaternary mammals: ecological archives E084-
094. Ecology, 84(12), 3403.
Solari, S. (2007). New species of Monodelphis (Didelphimorphia:
Didelphidae) from Peru, with notes on M. adusta (Thomas, 1897).
JournalofMammalogy,88(2), 319–329.
Solari, S., Pacheco, V., Vivar, E., & Emmons, L. H. (2012). A new species
of Monodelphis (Mammalia: Didelphimorphia: Didelphidae) from
the montane forests of central Perú. Proceedings of the Biological
Society of Washington, 125(3), 295–307.
Stamatakis, A., Hoover, P., & Rougemont, J. (2008). A rapid bootstrap
algorithm for the RAxML web servers. Systematic Biology, 57(5),
758–771.
Steiner, C., Tilak, M. K., Douzery, E. J., & Catzeflis, F. M. (2005). New
DNA data from a transthyretin nuclear intron suggest an Oligocene
to Miocene diversification of living South America opossums
(Marsupialia: Didelphidae). Molecular Phylogenetics and
Evolution, 35(2), 363–379.
Ter Braak, C.J.F., & Šmilauer, P. (1998). CANOCO 4. CANOCO refer-
ence manual and user’s guide to Canoco for Windows. Centre of
Biometry, Wageningen.
Teta, P., D’Elía, G., Flores, D., & de La Sancha, N. (2009). Diversity and
distribution of the mouse opossums of the genus Thylamys
(Didelphimorphia, Didelphidae) in northeastern and central
Argentina. Gayana, 73(2), 180–199.
Venditti, C., Meade, A., & Pagel, M. (2011). Multiple routes to mamma-
lian diversity. Nature, 479(7373), 393–396.
Vieira, E. M., & Astúa de Moraes, D. (2003). Carnivory and insectivory
in Neotropical marsupials. In M. E. Jones, C. R. Dickman, & M.
Archer (Eds.), Predators with pouches: the biology of carnivorous
marsupials (pp. 271–284). Collingwood: CSIRO Publishing.
Vilela, J. F., Oliveira, J. A., & Russo,C. A. M. (2015). The diversification
of the genus Monodelphis and the chronology of Didelphidae
(Didelphimorphia). Zoological Journal of the Linnean Society,
174(2), 414–427.
Voss, R.S., & Jansa, S.A. (2003). Phylogenetic studies on didelphid mar-
supials II. Nonmolecular data and new IRBP sequences: separate
and combined analyses of didelphine relationships with denser tax-
on sampling. Bulletin of the American Museum of Natural History,
1–82.
Voss, R.S., & Jansa, S.A. (2009). Phylogenetic relationships and classi-
fication of didelphid marsupials, an extant radiation of New World
metatherian mammals. Bulletin of the American Museum of Natural
History,1–177.
Voss, R.S., Tarifa, T., & Yensen, E. (2004). An introduction to
Marmosops (Marsupialia: Didelphidae), with the description of a
new species from Bolivia and notes on the taxonomy and distribu-
tion of other Bolivian forms. American Museum Novitates,1–40.
Voss,R.S.,Lunde,D.P.,&Jansa,S.A.(2005).Onthecontentsof
Gracilinanus Gardner and Creighton 1989, with the description of
a previously unrecognized clade of small didelphid marsupials.
American Museum Novitates,1–36.
Voss, R. S., Pine, R. H., & Solari, S. (2012). A new species of the
didelphid marsupial genus Monodelphis from eastern Bolivia.
American Museum Novitates, 3740,1–14.
Voss, R. S., Lim, B. K., Díaz-Nieto, J. F., & Jansa, S. A. (2013). A new
species of Marmosops (Marsupialia: Didelphidae) from the
Pakaraima Highlands of Guyana, with remarks on the origin of the
endemic Pantepui mammal fauna. American Museum Novitates,
3778,1–27.
Voss, R. S., Gutiérrez, E. E., Solari, S., Rossi, R. V., & Jansa, S. A. (2014).
Phylogenetic relationships of mouse opossums (Didelphidae,
Marmosa) with a revised subgeneric classification and notes on
sympatric diversity. American Museum Novitates, 3817,1–27.
Phylogeny and evolution of body mass in didelphid marsupials