Ten New Genera of Oryzomyine Rodents (Cricetidae: Sigmodontinae)

2006·0.46 MB·English

by  Marcelo Weksler

PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY
CENTRAL PARK WEST AT 79TH STREET, NEW YORK, NY 10024
Number 3537, 29 pp., 1 figure, 3 tables October 19, 2006
Ten New Genera of Oryzomyine Rodents
(Cricetidae: Sigmodontinae)
MARCELO WEKSLER,1 ALEXANDRE REIS PERCEQUILLO,2
AND ROBERT S. VOSS3
ABSTRACT
In order to achieve a monophyletic classification of oryzomyine rodents, 10 new genera are
describedforspeciesorspeciesgroupspreviouslyreferredtothepolyphyleticgenusOryzomys.The
following names are proposed: Aegialomys, n.gen. (for the ‘‘xanthaeolus group’’ of authors);
Cerradomys,n.gen.(forthe‘‘subflavusgroup’’);Eremoryzomys,n.gen.(forpolius);Euryoryzomys,
n.gen. (for the ‘‘nitidus group’’); Hylaeamys, n.gen. (for the ‘‘megacephalus group’’); Mindomys,
n.gen.(forhammondi);Nephelomys,n.gen.(forthe‘‘albigularisgroup’’);Oreoryzomys,n.gen.(for
balneator); Sooretamys, n.gen. (for angouya); and Transandinomys, n.gen. (for bolivaris and
talamancae). All of the new genera thus constituted are morphologically diagnosable and have
distinct ecogeographic distributions. Pending revisionary work that is currently in progress by
other researchers, six species belonging to the ‘‘alfaroi group’’ (herein construed as including
alfaroi, chapmani, melanotis, rhabdops, rostratus, and saturatior) are provisionally referred to
Handleyomys. As a result of these changes, the genus Oryzomys is restricted to the ‘‘palustris
group’’of authors,andthe tribe Oryzomyini nowcomprises28genera.
INTRODUCTION afewgeographicallywidespreadandmorpho-
logically undiagnosable genera. Although
A striking characteristic of muroid rodent Rattus (sensu lato) is perhaps the most
classification at the middle of the last century notorious example (Musser, 1981; Musser
was the large number of species assigned to and Newcomb, 1983; Musser and Holden,
1DivisionofVertebrate Zoology(Mammalogy), AmericanMuseumofNaturalHistory.Presentaddress:Instituteof
ArcticBiologyandUniversityofAlaskaMuseum,UniversityofAlaskaFairbanks,Fairbanks,AK99775(mweksler@
amnh.org).
2DepartamentodeSistema´ticaeEcologia,CentrodeCieˆnciasExatasedaNatureza,UniversidadeFederaldaPara´ıba,
CaixaPostal5133,58051-970Joa˜oPessoa,Paraiba,Brazil([email protected]).
3DivisionofVertebrateZoology(Mammalogy),AmericanMuseumofNaturalHistory([email protected]).
CopyrightEAmericanMuseumofNaturalHistory2006 ISSN0003-0082
2 AMERICAN MUSEUMNOVITATES NO. 3537
1991), several New World genera have also ‘‘complexes’’ (e.g., by Goldman, 1918; Tate,
served as convenient receptacles for superfi- 1932; Ellerman, 1941; Gardner and Patton,
cially similar but phylogenetically heteroge- 1976; Weksler, 1996; Musser et al., 1998;
neous species. In Neotropical mammalogy, Percequillo, 1998, 2003; Sa´nchez et al., 2001).
this was the traditional role of such elastic In the accounts that follow, we designate
taxa as Akodon, Oryzomys, and Thomasomys. atypespeciesforeachnewgenus,listthevalid
Thanks to revisionary research in the last species(andsynonyms)referred toit,describe
several decades (reviewed by Musser and its geographic distribution, provide morpho-
Carleton, 1993, 2005), each of these genera is logical diagnoses and comparisons, and
now recognized in a much more restricted briefly comment on the criteria we used to
sense than formerly, but many nomenclatural determine the assignment of species not
problemsremain.Thepurposeofthisreportis represented in published analyses of character
to complete the transition from traditional data. Throughout these accounts, morpho-
usage to a phylogenetic classification of the logical characters are described using termi-
species hitherto referred to Oryzomys. nology defined and illustrated by Voss (1988,
As treated in influential mid-20th-century 1993), Carleton and Musser (1989), Voss and
checklists (e.g., Tate, 1932; Gyldenstolpe, Carleton (1993), Musser et al. (1998), Voss et
1932; Ellerman, 1941; Hall and Kelson, 1959; al. (2002), and Weksler (2006).
Cabrera,1961),thegenusOryzomyscontained
anywhere from 60 to 120 nominal taxa in five COMMON ATTRIBUTES
to seven subgenera that collectively ranged
from Patagonia to New Jersey. The artifici- The taxa named below share many attri-
ality of this usage was subsequently empha- butes that it would be pointless to repeat in
sized by karyotypic and morphological re- eachdiagnosis.Forexample,insofarasknown
searchers (e.g., Gardner and Patton, 1976; (postcranialskeletonshavenotbeenexamined
Carleton and Musser, 1989) who raised all of fromalltaxa),theyresembleothermembersof
the subgenera recognized by midcentury the cricetid subfamily Sigmodontinae by
authors to generic rank. The species that are having a double articulation of the first rib,
still referred to Oryzomys, however, do not lacking an entepicondylar foramen of the
compriseamonophyleticgroup(fig. 1),andit humerus, and lacking an entoglossal process
isintolerablethatthissituationshouldpersist. of the basihyal. Similarly, the material we
In order to achieve a monophyletic classi- examined indicates that they consistently re-
fication, we now name 10 new genera for sembleothermembersofthetribeOryzomyini
species currently classified as Oryzomys. in lacking a posterior suspensory process of
Pending the description of other new genera the squamosal, having 12 ribs, having uniloc-
(byM.D.CarletonandG.G.Musser,personal ular-hemiglandular stomachs, and lacking
commun.), we provisionally transfer members a gall bladder.
of the ‘‘alfaroi group’’ (herein understood to Some characters that vary within Oryzo-
includealfaroi,chapmani,melanotis,rhabdops, myini are likewise uninformative in the con-
rostratus, and saturatior) to Handleyomys, text of these comparisons and need not be
a suboptimal but phylogenetically defensible repeated below. Like most oryzomyines (with
nomenclatural option previously discussed by exceptions as noted), all of the taxa described
Weksler (2006). Table 1 lists all of the herein have soft fur (Neacomys and Scolomys
oryzomyine genera that we recognize as valid, have spiny fur); the manual claws are small
including those described as new herein. and unkeeled (Lundomys has long, ventrally
keeled manual claws); the hind feet lack well-
developed natatory fringes and interdigital
NEW GENERA
webs (well-developed natatory fringes and/or
Most of the clades for which we provide webbing are present in Amphinectomys,
new generic names have been recognized in Holochilus, Lundomys, Nectomys, Oryzomys
one form or another for many years, usually palustris, and Pseudoryzomys); the mammary
as informally designated species ‘‘groups’’ or complementconsistsofeightteatsininguinal,
2006 WEKSLER ET AL.:TEN NEW ORYZOMYINE GENERA 3
Fig. 1. Phylogenetic relationships of oryzomyines based on a heuristic parsimony analysis of sequence
data from the Interphotoreceptor Retinoid Binding Protein (IRBP, 1266bp from exon 1) and 99
morphological characters (after Weksler, 2006: fig. 37). Numbers above and below branches represent
jackknifesupport (.50%) and decay indices (.1), respectively. Vertical bars on the right-hand side of the
figureindicate taxonmembership in cladesA–D. SeeWeksler (2006: 14–17)for methodological details.
abdominal, postaxial, and pectoral pairs less conspicuous epidermal scales and lacks
(Handleyomys and Scolomys have six mam- a terminal tuft of long hairs (the well-haired
mae because they lack pectoral teats); the tailofNesoryzomysdoesnotappearscaly,and
sparsely haired tail is covered with more or some species of Oecomys have prominently
4 AMERICAN MUSEUMNOVITATES NO. 3537
TABLE1
Contentsofthe TribeOryzomyini
tufted tails); the zygomatic plate lacks an a well-developed postorbital ridge); the bony
anterodorsal spinous process (a spinous pro- palate between the molar rows is smooth or
cess is present in Pseudoryzomys, Lundomys, weaklysculpted(thepalatesofHolochilusand
andHolochilus);thenasalboneshaverounded Lundomys have a well-developed median keel
or squared posterior margins (Nectomys, flanked by deep lateral gutters); the alisphe-
Scolomys, and Sigmodontomys have acutely noid canal has a large anterior opening (the
angled posterior nasal margins); the posterior anterior opening of the alisphenoid canal is
wall of the orbit is smooth (Holochilus has absent or very small in Scolomys); the upper
2006 WEKSLER ET AL.:TEN NEW ORYZOMYINE GENERA 5
incisors have smoothly rounded enamel bands eye when laid forward. Mystacial and super-
(theupperincisorenamelisdistinctlyfacetedin ciliary vibrissae not extending posteriorly
Holochilus); the molars are low-crowned and beyond pinnae when laid back. Hind foot
bunodont or terraced (Holochilus has high- withconspicuoustuftsofungualhairsatbases
crowned, planar molars); the labial flexi are of claws on dI–dV; plantar surface densely
enclosed by a cingulum (the labial flexi are coveredwithdistinctsquamaedistaltothenar
unenclosed in Holochilus and Lundomys); the pad; hypothenar pad present and large; claw
maxillary toothrows are parallel (Holochilus of dI extending beyond middle of phalange 1
and Lundomys have anteriorly convergent (almost to first interphalangeal joint) of dII;
toothrows);mesolophsarepresentonallupper claw of dV extending just beyond first in-
molars(Holochilus,Lundomys,Pseudoryzomys, terphalangeal joint of dIV. Tail about as long
Scolomys, and Zygodontomys lack mesolophs as head and body in A. galapagoensis but
ononeormoreupperteeth);themedianmure distinctly longer than head and body in A.
is connected to the protocone on M1 (the xanthaeolus; weakly to distinctly bicolored
median mure is connected to the paracone in (dark above, pale below).
Holochilus); the paracone of M2 lacks an
Skull with stout rostrum flanked by deep
accessory loph (an accessory loph is present
zygomatic notches; interorbital region anteri-
in Oecomys); and a posteroflexid is presenton
orly convergent with strongly beaded supra-
m3 (posteroflexids are absent on m3 in
orbital margins; braincase oblong, usually
Holochilus, Lundomys, Pseudoryzomys, and
with well-developed temporal crests; lambdoi-
Zygodontomys). Likewise, all dissected oryzo-
dal and nuchal crests often well developed in
myines(exceptNesoryzomys)havemaleacces-
older adults. Posterior margin of zygomatic
sory reproductive gland complements that
plate dorsal toM1 alveolusinsome examined
include one pair each of bulbourethral, dorsal
specimens, anterior to M1 alveolus in others.
prostate, anterior prostate, vesicular, and
Jugal present but small (the maxillary and
ampullary glands, and two pairs of ventral
squamosal zygomatic processes broadly over-
prostateglands.
lapping in lateral view but not in contact).
In effect, the species that still remain in
Nasals extending posteriorly behind lacrimals
Oryzomys are those that lack the conspicu-
in A. galapagoensis but shorter (extending to
ously divergent morphological traits of ory-
but usually not behind lacrimals) in A.
zomyines hitherto referred to other genera.
xanthaeolus; lacrimals usually with longer
However,thetaxanamedbelowdifferinother
maxillary than frontal sutures. Fronto-
characters that provide an unambiguous basis
squamosal suture usually colinear with fron-
for the following diagnoses and comparisons.
toparietal suture. Parietals with broad lateral
expansions. Incisive foramina long, typically
Aegialomys, new genus extending posteriorly to or between M1
TYPE SPECIES: Oryzomys xanthaeolus Tho- alveoli; almost parallel-sided (in A. galapa-
mas, 1894. goensis) or widest at midlength and tapering
CONTENTS: galapagoensis Waterhouse, symmetricallyanteriorlyandposteriorly(inA.
1839 (including bauri J.A. Allen, 1892) and xanthaeolus). Posterolateral palatal pits large,
xanthaeolus Thomas, 1894 (including baroni complex, and recessed in deep fossae; mesop-
J.A. Allen, 1897, and ica Osgood, 1944). terygoid fossa penetrating anteriorly between
DISTRIBUTION: In the lowland dry forests maxillae in A. galapagoensis but often not in
of western Ecuador (including the Galapagos A. xanthaeolus; bony roof of mesopterygoid
Islands) and western Peru, but also at higher fossa perforated by very large sphenopalatine
elevations (to about 2500 m) in the upper vacuities. Alisphenoid strut absent (buccina-
Maran˜o´n valley of northern Peru. tor-masticatory foramen and accessory fora-
MORPHOLOGICALDIAGNOSIS: Dorsalpelage men ovale confluent). Stapedial foramen and
coarsely grizzled yellowish- or grayish-brown; posterior opening of alisphenoid canal small;
ventral pelage abruptly paler (superficially squamosal-alisphenoid groove and spheno-
whitish or pale yellow), but ventral hairs frontal foramen absent; secondary anastomo-
alwaysgray-based.Pinnaesmall,notreaching sisofinternalcarotidcrossesdorsalsurfaceof
6 AMERICAN MUSEUMNOVITATES NO. 3537
pterygoid plate (5 carotid circulatory pattern brae with posterior spinous process. Sup-
3 of Voss, 1988). Postglenoid foramen large ratrochlear foramen of humerus present.
and rounded; subsquamosal fenestra small Stomach without extension of glandular
but distinct in most forms, but vestigial or epithelium into corpus. One pair of preputial
absent in an unnamed species from coastal glands present. Distal bacular cartilage of
Ecuador. Periotic exposed posteromedially glans penis small and trifid (with a short and
between ectotympanic and basioccipital, but slendercentraldigit);nonspinoustissueonrim
usually not extending anteriorly to carotid of terminal crater does not conceal bacular
canal; mastoid unfenestrated or with a small mounds; dorsal papilla spineless; urethral
but distinct posterodorsal fenestra (in speci- processes without subapical lobules.
mensfromcoastalEcuador).Capsularprocess COMPARISONS: Aegialomyswasrepresented
of lower incisor alveolus well developed in by ‘‘Oryzomys’’ xanthaeolus4 in the phyloge-
most fully adult specimens; superior and netic analyses of Weksler (2003, 2006), who
inferior masseteric ridges conjoined anteriorly consistently recovered it asa member ofclade
as single crest below m1. D. Within clade D, ‘‘O.’’ xanthaeolus usually
Labial and lingual flexi of M1 and M2 not appeared as the sister taxon of a group
interpenetrating. First upper molar (M1) composed of Amphinectomys, Melanomys,
anterocone divided into anterolabial and Nectomys, and Sigmodontomys (as in fig. 1).
anterolingual conules by distinct anterome- Phenetically,however,Aegialomysmoreclose-
dian flexus in some forms (e.g., A. galapa- ly resembles Oryzomys sensu stricto (the
goensis and an undescribed species from ‘‘palustris group’’ of authors), Cerradomys
coastal Ecuador), undivided in others (e.g., (the ‘‘subflavus group’’), and Eremoryzomys
A. xanthaeolus, which, however, has a small (the ‘‘polius group’’). Comparisons with
Oryzomys sensu stricto and Cerradomys are
internal fossette that seems to represent
provided here, and comparisons with
a vestigial anteromedian flexus); anteroloph
Eremoryzomys are included in the account
well developed and fused with anterostyle on
for that genus (below). Table 2 summarizes
labialcingulum,separatedfromanteroconeby
key morphological comparisons among all of
persistentanteroflexusinsomespecies(e.g.,A.
the new taxa belonging to clade D.
xanthaeolus) but fused with anterocone (ante-
Aegialomys differs from Oryzomys in nu-
roflexus reduced or absent) in others; proto-
merous traits, among which the most note-
style absent; paracone usually connected by
worthy are its large, distinct hypothenar pad
enamel bridge to posterior moiety of proto-
onthehindfoot(thehypothenarpadisabsent
cone. Second upper molar (M2) protoflexus
orvestigialinOryzomys);conspicuoustuftsof
present; mesoflexus present as single internal
long ungual hairs at the bases of the claws on
fossette. Third upper molar (M3) with poster-
pedal digits II–V (the ungual hairs are sparse
oloph and diminutive hypoflexus (the latter
and short in Oryzomys); M1 anteromedian
tending to disappear with moderate to heavy
flexus present or vestigial (the anteromedian
wear). Accessory labial root of M1 often
flexus is unambiguously absent in Oryzomys);
present.
M1 paracone usually attached by an enamel
First lower molar (m1) anteroconid usually
bridge to the posterior moiety of the proto-
without an anteromedian flexid; anterolabial
cone(theattachmentisusuallytotheanterior
cingulum present on all lower molars; ante-
moiety in Oryzomys); M2 mesoflexus form-
rolophidpresentonm1butabsentonm2and
ing a single internal fossette (the M2 meso-
m3; ectolophid absent on m1 and m2;
flexus usually forms two internal fossettes in
mesolophid distinct on unworn m1 but re-
Oryzomys); lack of distinct anterolophids on
duced on m2; m2 hypoflexid short. Accessory
lingualandlabialrootsofm1present;m2and 4Subsequentstudyindicatesthatthisterminaltaxonwas
m3eachwithtwosmallanteriorrootsandone a composite based on material of the unnamed
large posterior root. Ecuadoreanspeciesmentionedintheprecedingdiagnosis
Fifth lumbar (17th thoracicolumbar) verte- together with Aegialomys xanthaeolus sensu stricto.
Taxonomicdifferencesthereforeaccountforsomeofthe
bra with well-developed anapophysis. Hemal
character variation scored as polymorphisms of A.
arch between second and third caudal verte- xanthaeolusinWeksler’s(2006)analyses.
2006 WEKSLER ET AL.:TEN NEW ORYZOMYINE GENERA 7
TABLE 2
SelectedMorphological Comparisons AmongNew Genera fromCladeD
m2and m3 (anterolophidsare usually distinct each genus: for example, as documented in
on unworn m2 and m3 in Oryzomys); and Langguth and Bonvicino’s recent (2002) de-
mesolophidsthat tend to disappearas distinct scriptions of new species of Cerradomys, and
structures with only moderate wear (mesolo- by our remarks about character variation in
phids are persistent as distinct structures in Aegialomys (above). In fact, Aegialomys and
Oryzomys). In addition, A. xanthaeolus has Cerradomys do not appear to differ consis-
a well-developed anapophysis on the fifth tently in any integumental or cranial feature
lumbar vertebra that is absent in O. couesi that we have been able to identify. Several
and O. palustris; atridigitate bacular cartilage dental and genitalic characters, however,
with a short and slender central digit (the suggest that these are distinct taxa that merit
central digit is robust in O. couesi and O. formal recognition. Because they are so few,
palustris);aspinelessdorsalpapilla(thedorsal each character merits particular attention.
papillaisprovidedwithspinesinO.couesiand In Aegialomys galapagoensis, the unworn
O. palustris); and urethral processes that lack anterocone of M1 is divided into subequal
subapical lobules (subapical lobules are pres- anterolabial and anterolingual conules by an
ent on the urethral processes of O. couesi and anteromedian flexus, but in A. xanthaeolus
O. palustris). (from coastal Peru) the anterocone is un-
Although Aegialomys xanthaeolus and Cer- dividedandtheanteromedianflexusispresent
radomys subflavus differ in numerous mor- only as an internal fossette whose faint
phological characters and were never recov- connection to a shallow median sulcus in the
ered as sister taxa in Weksler’s (2003, 2006) anteriorfaceoftheanteroconeistransientand
phylogeneticanalyses, only a few traits distin- can only be seen on minimally worn teeth
guish their respective genera as recognized (e.g., AMNH 10111, 42398). By contrast, the
herein. This difficulty arises from substan- anterocone of M1 in Cerradomys is never
tial character variation among species within divided into labial and lingual conules by an
8 AMERICAN MUSEUMNOVITATES NO. 3537
anteromedian flexus, and the internal fossette Cerradomys, new genus
of the procingulum that is visible in some
TYPE SPECIES: Hesperomys subflavus Wag-
unworn dentitions (e.g., AMNH 134566,
ner, 1842.
illustrated by Musser et al., 1998: fig. 144) is
CONTENTS: maracajuensis Langguth and
clearly derived from the anteroflexus, a labial
Bonvincino, 2002; marinhus Bonvincino,
enamel infolding.
2003; scotti Langguth and Bonvincino, 2002
The mesoflexus of M2 is present as a single
(including andersoni Brooks et al., 2004); and
internal fossette in Aegialomys. Although
subflavus Wagner, 1842 (including vulpinus
occasional rare variants are to be expected in
Lund, 1840; vulpinoides Schinz, 1845; and
such traits, this morphology appears to be
laticeps Winge, 1888 [not Lund, 1840]).
exhibited consistently by examined specimens
DISTRIBUTION: In dry tropical and sub-
of A. galapagoensis, A. xanthaeolus (including
tropical forests of the Caatinga, Cerrado,
baroni), and the unnamed form from coastal
andChacofromnortheasternBraziltoeastern
Ecuador. The mesoflexus of Cerradomys,
Bolivia.
however, is usually represented by two in-
MORPHOLOGICALDIAGNOSIS: Dorsalpelage
ternal fossettes, of which one is labial and
coarsely grizzled, usually some shade of
other is nearer the midline of the tooth (as
reddish- or yellowish-brown; ventral pelage
illustrated for ‘‘Oryzomys’’ subflavus by
abruptly paler in some species (superficially
Musser et al., 1998: fig. 144).
whitish or yellowish) or not (the ventral
The male genitalia of Aegialomys galapa-
coloration merging gradually with that of the
goensis (‘‘Oryzomys bauri’’) and A. xanthaeo-
dorsum), but ventral hairs always gray-based.
lus (‘‘O. xantheolus’’) were described and
Pinnae small, not extending to eye when laid
illustrated by Patton and Hafner (1983). In
forward. Mystacial and superciliary vibrissae
both species, the distal bacular cartilage is
notextendingposteriorlybeyondpinnaewhen
unambiguously trifid, with a slender but
laidback.Hindfootwithconspicuoustuftsof
distinct central digit. By contrast, the glans
long ungual hairs at bases of claws on dI–dV;
penis of Cerradomys scotti, C. subflavus, and plantar surface densely covered with distinct
an undescribed congener from northeastern squamaedistaltothenarpad;hypothenarpad
Brazil have a bifid distal bacular cartilage smallbutdistinct;clawofdIextendingbeyond
because the middle digit is vestigial or absent. middle of phalange 1 but not quite to first
REMARKS: Although ‘‘Oryzomys’’ galapa- interphalangeal joint of dII; claw of dV
goensis and ‘‘O.’’ xanthaeolus have long been extendingtobutnotbeyondfirst interphalan-
recognized as closely related species (e.g., by geal joint of dIV. Tail longer than combined
Thomas, 1894; Gardner and Patton, 1976; length of head and body, weakly bicolored in
Patton and Hafner, 1983), no published most species but distinctly bicolored in others
phylogenetic analysis of biochemical or mor- (e.g., C. scotti).
phological data is currently available to Skull with long, tapering rostrum flanked
support the monophyly of Aegialomys as bydeepzygomaticnotches;interorbitalregion
constituted herein. The presence of at least anteriorly convergent, with strongly beaded
one undescribed species among the material supraorbital margins; braincase oblong, with
we examined, together with questions that well-developed temporal crests; lambdoidal
have been raised elsewhere concerning the and nuchal crests well developed in older
taxonomic status of ica (by Musser and adults. Posterior margin of zygomatic plate
Carleton, 2005: 1156) and our own reserva- usually anterior to M1 alveolus. Jugal present
tions about baroni, suggest that a revision of butsmall(maxillaryandsquamosalzygomatic
this group is needed to identify the terminal processes overlapping in lateral view but not
taxa that should be represented in future in contact). Nasals not extending posteriorly
phylogenetic analyses. beyond lacrimals; lacrimals usually sutured
ETYMOLOGY: Fromaegialos(Greekforthe equallytomaxillaryandfrontalbones(except
seashore), in reference to the predominantly in C. maracajuensis, which has longer maxil-
coastaldistributionofthesespeciesinwestern lary than frontal sutures). Frontosquamosal
South America. suture usually colinear with frontoparietal
2006 WEKSLER ET AL.:TEN NEW ORYZOMYINE GENERA 9
suture. Parietals with broad lateral expan- present on all lower molars; anterolophid
sions. Incisive foramina long, usually extend- present on m1 but absent on m2 and m3;
ing posteriorly to or between M1 alveoli; ectolophid usually absent on m1 and m2;
usually widest at midlength and tapering mesolophid well developed on m1 and m2 in
symmetrically anteriorly and posteriorly. mostspecies,butmesolophid(andmesostylid)
Posterolateral pits large, complex, and re- reduced or absent in C. scotti; m2 hypoflexid
cessed in deep fossae; mesopterygoid fossa short. Accessory lingual and labial roots
penetrating anteriorly between maxillae but present on m1; m2 and m3 each with one
usually not between molar rows; bony roof of large anterior root and one large posterior
mesopterygoid completely ossified in some root.
species (e.g., C. marinhus and C. maracajuen- Fifth lumbar (17th thoracicolumbar) verte-
sis) but perforated by large sphenopalatine bra with well-developed anapophysis. Hemal
vacuitiesinothers(e.g.,C.scotti).Alisphenoid arch between second and third caudal verte-
strut absent (buccinator-masticatory foramen brae with posterior spinous process. Supra-
and accessory foramen ovale confluent) in trochlear foramen of humerus present.
most species, but present (foramina separate) Stomach without extension of glandular
in C. scotti. Stapedial foramen and posterior epithelium into corpus. Distal bacular carti-
opening of alisphenoid canal vestigial or lage of glans penis bifid (the central digit is
absent; squamosal-alisphenoid groove and vestigial or absent); nonspinous tissue on
sphenofrontal foramen absent; secondary crater rim does not conceal bacular mounds;
anastomosis of internal carotid crosses dorsal dorsal papilla spineless; urethral processes
surface of pterygoid plate (5 carotid circula- without subapical lobules.
tory pattern 3 of Voss, 1988). Postglenoid COMPARISONS: Cerradomys was represent-
foramen large and rounded; small subsqua- ed by ‘‘Oryzomys’’ subflavus in the phyloge-
mosal fenestra distinct and patent in most netic analyses of Weksler (2003, 2006), who
species but absent or vestigial (not patent) in consistently recovered it as a member of clade
C. scotti. Periotic exposed posteromedially D (as in fig. 1). Although its phylogenetic
between ectotympanic and basioccipital but position within clade D was never strongly
usually not extending anteriorly to carotid supported in any analytic permutation based
canal; mastoid completely ossified or fenes- on morphological and/or IRBP sequence
trated (variation observed within and among characters, Cerradomys never appeared as
species). Capsular process of lower incisor the sister taxon of any other species of
alveolus strongly developed below base of Oryzomys sensu lato. However, it is pheneti-
coronoid process; superior and inferior mas- cally most similar to Aegialomys (the
seteric ridges converge anteriorly as open ‘‘xanthaeolus group’’ of authors), Oryzomys
chevron below m1. sensu stricto (the ‘‘palustris group’’), and
Labial and lingual flexi of M1 and M2 not Sooretamys (‘‘Oryzomys’’ angouya). Compar-
interpenetrating.Firstuppermolar(M1)ante- isons with Oryzomys are provided here,
rocone not divided into labial and lingual whereas comparisons with Aegiolomys and
conules (anteromedian flexus absent); antero- Sooretamys are provided in the accounts for
lophwelldevelopedandfusedwithanterostyle those taxa (above and below, respectively).
on labial cingulum, usually separated from Cerradomys differs from Oryzomys sensu
anterocone by persistent anteroflexus; proto- stricto by its distinct hypothenar pad on the
style usually absent; paracone connected by hindfoot(thehypothenarisabsentorvestigial
enamelbridgetomiddleortoposteriormoiety in Oryzomys); conspicuous tufts of long un-
of protocone. Second upper molar (M2) pro- gual hairs at the bases of the claws on pedal
toflexus present; mesoflexususually present as digits II–V (the short ungual hairs of
twointernalfossettes.Thirduppermolar(M3) Oryzomys do not form distinct tufts); lacri-
withposteroloph;hypoflexuspresentorabsent. mals that are usually sutured equally to the
AccessorylabialrootofM1present. maxillary and frontal bones (the lacrimals are
Firstlowermolar(m1)anteroconidwithout primarily sutured with the maxillaries in
ananteromedianflexid;anterolabialcingulum Oryzomys); shorter palate (the mesopterygoid
10 AMERICAN MUSEUMNOVITATES NO. 3537
fossa does not extend anteriorly between the forward. Mystacial and superciliary vibrissae
maxillary bones in Oryzomys); absence of notextendingposteriorlybeyondpinnaewhen
an anterolophid on m2 and m3 (the ante- laidback.Hindfootwithconspicuoustuftsof
rolophid is distinct on unworn m2 and m3 in long ungual hairs at bases of claws on dI–dV;
Oryzomys);presenceofananapophysisonthe plantar surface densely covered with distinct
fifth lumbar vertebra (absent in Oryzomys); squamaedistaltothenarpad;hypothenarpad
bifiddistalbacularcartilage(thedistalbacular largeanddistinct;clawofdIextendingalmost
cartilage is trifid in Oryzomys); dorsal papilla tofirstinterphalangealjointofdII;clawofdV
ofglanspenisspineless(spinousinOryzomys); extending just beyond first interphalangeal
and urethral processes without subapical jointofdIV.Taillongerthancombinedlength
lobules (present in Oryzomys). of head and body; distinctly bicolored (dark
REMARKS: Compelling evidence for the above, pale below).
monophyly of Cerradomys is provided by Skull with long, stout rostrum flanked by
parsimony and maximum likelihood anal- moderately deep zygomatic notches; interor-
yses of cytochrome b mtDNA sequences bital region anteriorly convergent, with
(Bonvicino and Moreira, 2001; Bonvicino,
beaded supraorbital margins; braincase
2003). The highly distinctive anatomy of the
rounded, with more or less distinct temporal
glans penis, which lacks a central bacular
crests; lambdoidal and nuchal crests devel-
digit, likewise supports this conclusion.
oped in older adults. Posterior margin of
Hopefully, the flurry of recently published
zygomatic plate dorsal to M1 alveolus; jugal
descriptions of new species of Cerradomys
present and large (the maxillary and squamo-
(e.g., by Langguth and Bonvicino, 2002;
sal zygomatic processes widely separated, not
Bonvicino, 2003; Brooks et al., 2004) will
overlapping in lateral view). Nasals short, not
soon be followed by more synthetic and
extending posteriorly beyond lacrimals; lacri-
comprehensive studies to convincingly docu-
mals equally sutured to maxillary and frontal
ment species identifications and geographic
bones. Frontosquamosal suture usually co-
distributions across the entire range of this
linear with frontoparietal suture. Parietals
obviously diverse clade. At least some of the
with broad lateral expansions. Incisive foram-
many unstudied specimens representing this
ina very long, usually extending posteriorly
genusfromBoliviaandParaguayarelikelyto
between M1 anterocones or protocones; with
belong to taxa recently described from
subparallel lateral margins. Posterolateral
Brazilian material, but others may represent
palatal pits large, complex, and recessed in
new forms. Unfortunately, no taxonomic
deep fossae; mesopterygoid fossa penetrating
study published to date has effectively trans-
anteriorly to or slightly between molar rows;
cended national boundaries.
bony roof of mesopterygoid fossa perforated
ETYMOLOGY: For the Cerrado, a vast mo-
bylargesphenopalatinevacuities.Alisphenoid
saic of savannas and dry forests, where many
strut usually present (buccinator-masticatory
species of this clade are found.
foramen and accessory foramen ovale sepa-
rate), but strut unilaterally absent on some
Eremoryzomys, new genus
skulls. Stapedial foramen and posterior open-
TYPE SPECIES: Oryzomys polius Osgood, ing of alisphenoid canal small; squamosal–
1913. alisphenoid groove and sphenofrontal fora-
CONTENTS: polius Osgood, 1913. menabsent;secondaryanastomosisofinternal
DISTRIBUTION: Known only from a few carotid crosses dorsal surface of pterygoid
localities in the upper R´ıo Maran˜o´n valley of plate(5carotidcirculatorypattern3ofVoss,
northern Peru. 1988). Postglenoid foramen large and round-
MORPHOLOGICALDIAGNOSIS: Dorsalpelage ed; subsquamosal fenestra large and patent.
coarsely grizzled-grayish (but brownish- or Periotic exposed posteromedially between
yellowish-grayinsomeoldandpossiblysoiled ectotympanic and basioccipital but not ex-
specimens); ventral pelage paler (superficially tending anteriorly to carotid canal; mastoid
whitish), but ventral hairs always gray-based. perforated by small or large posterodorsal
Pinnae small, not reaching eye when laid fenestra. Capsular process of lower incisor