Norntates - COnnecting REpositories · 2017-03-10 · rodents that range from southwestern...

AMERICAN MUSEUMNorntatesPUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORYCENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024Number 2812, pp. 1-20, figs. 1-6 April 10, 1985

Skull Morphology and Relationships ofGeomyoid Rodents

JOHN H. WAHLERT1

ABSTRACT

Analysis of cranial and mandibular morpholo-gy, especially the foramina and structures relatingto the masticatory musculature, supports mo-nophyly of the living Heteromyidae. The familyincludes two major divisions, the Heteromyinaeon the one hand, and the Perognathinae and Di-podomyinae on the other. Derived features ofSchizodontomys suggest that the genus may be anearly member of the Dipodomyinae. The Heter-omyidae and Geomyidae, containing only the

Geomyinae, share common ancestry. The Ento-ptychidae are tentatively placed as the sister groupof these two families; the Florentiamyidae appearto be the earliest known branch of the geomyoidclade. The Geomyoidea, which includes thesefamilies, and the extinct Eomyoidea comprise theinfraorder Geomorpha. A classification of theGeomorpha that reflects these phylogenetic hy-potheses is presented.

INTRODUCTION

Two basic theories on the interrelation-ships of geomyoid rodents have been pro-posed. In the traditional view the geomyids(pocket gophers), and the heteromyids (kan-garoo rats and pocket mice) are two mono-phyletic groups that are descended from acommon ancestor. In the alternative view re-cent geomyids and their supposed relatives,which are extinct, are derived independentlyfrom heteromyids. Published studies havestressed morphology of the teeth and of cra-

nial structures associated with the muscles ofmastication; they have not resolved this con-troversy.

I have attempted to find cranial charactersthat can be added to the diagnoses of groupsofliving geomyoids. The new evidence agreeswith common, current subdivisions. Thefamily Geomyidae, subfamily Geomyinae,contains four genera- Geomys, Orthogeo-mys, Pappogeomys, and Thomomys. Thefamily Heteromyidae contains three subfam-

1 Research Associate, Department of Vertebrate Paleontology, American Museum of Natural History; AssistantProfessor of Biology, Department of Natural Sciences, Baruch College of the City University ofNew York.

Copyright © American Museum of Natural History 1985 ISSN 0003-0082 / Price $2. 10

AMERICAN MUSEUM NOVITATES

ilies that each include two genera: Hetero-myinae-Heteromys and Liomys; Pero-gnathinae-Chaetodipus and Perognathus;Dipodomyinae-Dipodomys and Micro-dipodops. I examined the published descnp-tions of skulls of extinct taxa to find char-acters that indicated relationships to thesegroups. Entoptychus, Pleurolicus, andSchizodontomys have heretofore been re-garded as constituting one or more subfam-ilies within the Geomyidae. Cranial evidenceof the Florentiamyidae (Wahlert, 1983) andthe Eomyoidea (Wahlert, 1978) is includedin the discussion of phylogenetic relation-ships; complete data are available in the pre-vious publications. Supposed cranial remainsofHeliscomys have been described; since theirgeneric assignments are uncertain (Wahlert,1983, 1984) I have omitted them.Rodents of the superfamily Geomyoidea

and its sister taxon, the extinct Eomyoidea,have a sciuromorphous lateral masseter. Thegeomyoids are distinguished by bilophodontcheek teeth and the presence offur-lined cheekpouches that open lateral to the mouth inliving forms. The concept of the Geomyo-idea, its subdivisions and their contents de-veloped gradually. Important early contri-butions were summarized by Coues (1877),Wood (1935), and Russell (1968).The Geomyinae are burrowers that range

from western Canada to Panama. Their fossilrecord extends back to the early Miocene andis restricted to North America with the pos-sible exception of a Miocene fossil fromShantung, China (Li, 1974). The Hetero-myidae are small, scampering and jumpingrodents that range from southwestern Can-ada to northern South America. Their fossilrecord goes back farther in time, to the earlyOligocene, and is entirely North American.The living heteromyids have certain cra-

nial and dental characters that are regardedas primitive among geomyoid rodents. Theoldest supposed heteromyid, Heliscomys, ismore ancient than other undisputed geo-myoid fossils. For these reasons systematistsexpect that the ancestors of all geomyoidswere heteromyid-like rodents. The tradition-al phylogenetic hypothesis, that geomyids andheteromyids are two monophyletic groupsdescended from a common, heteromyid-likeancestor, developed in the initial decades of

the twentieth century. Wood (1935) pre-sented a detailed phylogeny of heteromyidsthat included the known extinct and livingtaxa, and he discussed their relationship togeomyids. The first challenge to the tradi-tional view was made by Wilson (1936); hesuggested that the extinct and living geomyidsubfamilies had separate ancestry in the Het-eromyidae. Both hypotheses have been fur-ther elaborated by later authors.Wood (1935) proposed that the Dipodo-

myinae and Perognathinae are each other'sclosest relatives and that together they sharewith the Heteromyinae a common ancestryin the Oligocene genus Heliscomys. The crownmorphology of geomyid cheek teeth couldalso be derived from that of Heliscomys, butWood believed the morphologic gap betweenthis genus and the John Day geomyids to beso great and the time interval so short as topreclude the possibility of ancestor-descen-dant relationship. Recent studies of the Het-eromyidae by M. S. Hafner (1982) and Haf-ner and Hafner (1983) support Wood'shypothesis of relationship among thesubfamilies.A new subfamily ofheteromyids, the Flor-

entiamyinae, was describedby Wood (1936a).The skull of Florentiamys loomisi, on whichthe subfamily is based, is heteromyid-like buthas many primitive features. The tooth crownpattern appears more primitive than that ofHeliscomys. Wood concluded (ibid., p. 47)that this florentiamyine species could not be"a direct ancestor ofthe heteromyids, thoughit might be a structural ancestor." Wahlert(1983) described new specimens of florentia-myids and found a combination of cranialcharacters indicating that the group is neitherancestral to nor derived from heteromyids.Wood (1936b) discussed the relationships

of the Entoptychinae, an extinct geomyidsubfamily ofmid-Tertiary age, and remarkedthat "the evidence of the skulls is at least asstrong for geomyine affinities as the evidenceof the teeth is for heteromyid ones" (ibid., p.4). He described a new genus, Dikkomys, asrepresentative of an early stage in differen-tiation of the Geomyinae. This new materialdid not serve to link the Geomyinae un-equivocally to a particular branch of the En-toptychinae, but Wood found some points ofsimilarity to the genus Pleurolicus.

2 NO. 2812

WAHLERT: GEOMYOID RODENTS

Russell (1968) revised the systematics ofthe Geomyinae. He agreed with Wood on theancestral position of Dikkomys. He derivedthe geomyines and entoptychines from a hy-pothetical ancestor. Rensberger (1971, 1973a)divided the early geomyids into two subfam-ilies, the Entoptychinae and Pleurolicinae,and he added (1973b) the Florentiamyinaeto this lineage. In his phylogeny (ibid.) Hel-iscomys is shown as the common ancestor ofgeomyids and heteromyids.

In the alternative phylogenetic hypothesis,the Geomyinae are considered to be derivedfrom the Heteromyidae at a time later thanthe origin of the Entoptychinae; the hetero-myids, which include Heliscomys, are a para-phyletic stem group. Wilson (1936) describeda late Miocene geomyine, Pliosaccomys; hepointed out many heteromyid-like featuresof the dentition and suggested that the genusmay illustrate a morphological stage passedthrough by the modern Geomyinae. Since theearlier entoptychines are more specializedthan this genus in certain respects, Wilsonproposed that they might represent an earlierbranch from the central stock.A second challenge to monophyly of the

Heteromyidae was posed by Shotwell (1967).On the basis ofnew specimens, he describedthe tooth crown patterns in Pliosaccomys asa developmental step from that in Dikkomys.Morphological similarities led him to con-clude that the Geomyinae are closely relatedto the dipodomyine heteromyids; that Pleu-rolicus, too, resembles the dipodomyines andmay be a part of this lineage; and that En-toptychus appears to parallel the Geomyinaein structure but is not a close relative.

Lindsay (1972) presented a phylogenetictree of heteromyids that incorporated manyof these ideas. Heliscomys, the commonancestor, was placed in the Perognathinae.The Geomyinae and Dipodomyinae were twobranches of a distinct lineage. The Hetero-myinae comprised a fourth division; inclu-sion of Pleurolicus as the earliest represen-tative of this subfamily appears to have beenbased on the work ofReeder (1956). The En-toptychinae were shown as a fifth derivativefrom Heliscomys. The Florentiamyinae wereomitted.

Rensberger posed a similar challenge tomonophyly of the Heteromyidae. He (1971,

p. 66, fn.) stated that assignment of the En-toptychinae to the Geomyidae may be arti-ficial and that the family is then only a grade.He pointed out (ibid., p. 156) that dental sim-ilarities suggest independent origin of themodern Geomyinae from the subfamily Het-eromyinae.

In this paper I use aspects of cranial mor-phology to investigate the interrelationshipsof heteromyid and geomyid subfamilies. Ihave grounded the study in recent taxa, be-cause skulls are abundant, and the contentsofany foramen can be checked by dissection.This information is sufficient to test the mo-nophyly of the living heteromyids. The ex-tinct taxa are examined in the context of thecranial evidence that defines living taxa andthe concepts ofcharacter polarity in the Geo-morpha that were developed in previouspublications (Wahlert, 1978, 1983).

ACKNOWLEDGMENTS

This research was funded by National Sci-ence Foundation grant no. DEB79-03286.The departments of Vertebrate Paleontologyand Mammalogy at the American Museumof Natural History gave me free access tocollections; together with the Department ofBiology at Franklin and Marshall College theyprovided the facilities for the research. TheNikon dissecting microscope with camera lu-cida that I used for drawing small skulls waspurchased with funds granted by the Profes-sional Staff Congress of the City Universityof New York, grant no. 6-63171. A cameralucida given to me by Miss Elinore Blaisdellof Lancaster, Pennsylvania, has been an in-valuable tool for drawing large specimens.

I appreciate the help ofDrs. Karl F. Koop-man, Malcolm C. McKenna, Guy G. Musser,Richard H. Tedford, and Ms. Marie A. Law-rence. Drs. Albert E. Wood, John E. Storer,and Mark S. Hafner criticized the manuscriptand helped me to straighten muddled thoughtsand clarify muddy verbiage. Mr. Chester Tar-ka and Ms. Lorraine Meeker guided me inillustrating the recent specimens. The hos-pitality of Mr. and Mrs. Byron Bell maderesearch trips to New York City an affordablepleasure.

I am particularly indebted to Mr. TimothyScott Hall who carefully dissected the heads

1985 3


and described the anatomy of Geomys andDipodomys. He carried out this independentstudy project under my direction, when hewas a biology major at Franklin and MarshallCollege.

SPECIMENS EXAMINED

All recent specimens are from the collec-tion of the Department of Mammology,American Museum of Natural History, andbear the acronym AMNH. The generic andspecific names are in accord with Hall (1981)as it has been updated by Dr. Karl Koopmanfrom the papers of Anderson (1966), Best(1978), Hafner and Hafner (1983), Nader(1978), and Williams and Genoways (1978,1980, 1981).Geomys arenareus. 132080, Otero Co., N.

Mex. Geomys bursarius. 3467, Hennepin Co.,Minn.; 6399, Aransas Co., Tex.; 121661,Adams Co., Nebr.; 130275, Brazos Co., Tex.Geomyspersonatus. 2727, 2977, Nueces Co.,Tex. Geomys pinetis. 23023-23024, CamdenCo., Ga.; 23036, Glynn Co., Ga.; 166457,Clarke Co., Ala.; 166538, Pinellas Co., Fla.Orthogeomys cavator. 18846, Panama. Or-

thogeomys cherriei. 140385, Costa Rica. Or-thogeomys cuniculus. 206830, Oaxaca, Mex-ico. Orthogeomys dariensis. 37982, Panama.Orthogeomys grandis. 123388, Honduras;167393, 171622, 185055, Oaxaca, Mexico.Orthogeomys heterodus. 139272, 139762,Costa Rica. Orthogeomys hispidus. 79104,Guatemala; 148005, Tamaulipas, Mexico;165989, 178729, Oaxaca, Mexico; 177335,Veracruz, Mexico. Orthogeomys under-woodi. 123389, Costa Rica.Pappogeomys bulleri. 4303, Jalisco, Mex-

ico. Pappogeomys castanops. 63812, San LuisPotosi, Mexico; 136722, Brewster Co., Tex.;188613, 188617, Chihuahua, Mexico. Pap-pogeomys gymnurus. 26168, 26184, 26220,Jalisco, Mexico. Pappogeomys merriami.10881, Veracruz, Mexico; 143208, DistritoFederal, Mexico. Pappogeomys tylorhinus.143555, Distrito Federal, Mexico.Thomomysbottae. 13837, 124268, Merced

Co., Calif.; 124230, San Bernardino Co., Cal-if.; 132034, Otero Co., N. Mex.; 132077, Ber-nalillo Co., N. Mex. Thomomys bulbivorous.40581, Multonomah Co., Oreg. Thomomysmonticola. 121122, Tuolumne Co., Calif.

Thomomys talpoides. 123145, Rio BlancoCo., Colo. Thomomys umbrinus. 15784,17513, Chihuahua, Mexico.Heteromys anomalus. 14755, Venezuela;

34593, Tolima, Colombia; 186715, Trini-dad. Heteromys australis. 32959, Cauca, Co-lombia. Heteromys desmarestianus. 79248,Guatemala. Heteromys gaumeri. 91197, Yu-catan, Mexico. Heteromys lepturus. 172467,172482, Veracruz, Mexico.Liomys adspersus. 147781, Panama. Lio-

mys irroratus. 189221, 189223, Oaxaca,Mexico. Liomyspictus. 143452,190271, Oa-xaca, Mexico; 24258, Sinaloa, Mexico. Lio-mys salvini. 79271, Guatemala; 126291,128159, Honduras; 177012, Chiapas, Mex-ico.

Chaetodipus baileyi. 136973, 136975, PimaCo., Ariz. Chaetodipus californicus. 4559,Orange Co., Calif. Chaetodipusfallax. 3250,Orange Co., Calif. Chaetodipus formosus.122256, Washoe Co., Nev. Chaetodipus his-pidus. 3192, Cameron Co., Tex. Chaetodipusintermedius. 132458, Lincoln Co., N. Mex.Chaetodipus nelsoni. 21009, Durango, Mex-ico. Chaetodipus penicillatus. 139586, SanDiego Co., Calif.; 171029, Sonora, Mexico.Chaetodipuspernix. 25902, Sinaloa, Mexico.Chaetodipus spinatus. 31838, Baja Calif.,Mexico.Perognathus amplus. 136987, Pima Co.,

Ariz. Perognathus fasciatus. 6708, ShannonCo., S. Dak. Perognathus flavescens. 6692,Navajo Co., Ariz. Perognathusflavus. 8755,Bexar Co., Tex.; 132441, Colfax Co., N. Mex.Perognathus longimembris. 138595, KernCo., Calif. Perognathus parvus. 33505, Mal-heur Co., Oreg.Dipodomys agilis. 94303, San Diego Co.,

Calif. Dipodomys deserti. 139605, Yuma Co.,Ariz.; 145389, Pinal Co., Ariz. Dipodomysheermanni. 124158, San Luis Obispo Co.,Calif. Dipodomys merriami. 184199, PimaCo., Ariz.; 188637, Chihuahua, Mexico. Di-podomys ordii. 39815, Malheur Co., Oreg.;124717, Beaver Co., Utah. Dipodomys pan-amintinus. 138663, Tulare Co., Calif. Dipod-omysspectabilis. 145390, Pinal Co., Ariz. Di-podomys venustus. 145108, Monterey Co.,Calif.Microdipodops megacephalus. 135602,

135610,135612, Washoe Co., Nev.; 140689,Elko Co., Nev.

4 NO. 2812


Mr. Timothy Hall dissected a specimen ofGeomys bursarius from Douglas Co., Kansas,and of Dipodomys ordii from Thomas Co.,Nebraska. I dissected a specimen of Tho-momys bottae.

LIVING GEOMYOIDS

CRANIAL FORAMINA

A guide to the foramina in each ofthe fourliving subfamilies is presented in figures 1through 4. A small, median interpremaxillaryforamen is present immediately behind theincisors in geomyines; the structure is lackingin heteromyids. The incisive foramina ofgeo-myines, heteromyines, and perognathines aresituated in the anterodorsal part of a pair ofdepressions in the diastema; the length ofthese deep openings is 15 percent or less ofthe diastemal length. A spur of the maxillareaches the back of each foramen in mostspecimens; in others the median and lateralparts of the premaxilla meet behind the fo-ramina, and the maxilla is excluded from theirmargins. The posterolateral extensions ofthepremaxillaries are usually longer than themedian spine of that bone between the twoforamina. In dipodomyines the incisive fo-ramina are at the surface in the diastema andoccupy from 22 to 38 percent of its length;the premaxillary-maxillary suture intersectsthe lateral margins of the incisive foraminanear the middle.The posterior palatine foramina are in the

maxillary-palatine suture in all taxa exceptMicrodipodops, where they are within thepalatine bones. The maxillary-palatine su-ture is fused in many of the geomyine spec-imens. The major pair of these foraminaranges in position through a zone that extendsfrom the middle ofthe first molar to thejunc-tion of the second and third molars. In geo-myines a deeply incised furrow runs forwardfrom each foramen; in some specimens ex-cavation of the palatine also extends poste-riorly from each foramen. The palate in het-eromyids is flat or has shallow furrows. Avariable number of small foramina com-monly occur in the palate of geomyoids.The back ofthe palate is peculiar. The pal-

atine bones extend posteriorly and form apair ofdepressions, the parapterygoid fossae,which are separated by a median ridge. The

fossae are dorsal to the level of the anteriorpart of the palate. They are shallow in Mi-crodipodops and shallow and anteroposte-riorly short in Perognathus. Merriam (1895)described these structures in the Geomyinaebut did not mention their function. Our dis-sections of Geomys and Thomomys revealedthat the fossae house salivary glands. In Di-podomys an anterior part ofthe internal pter-ygoid muscle originates here (Howell, 1932,p. 412). Foramina occur in the region. In mostgeomyoid specimens the posterior maxillaryforamen is present laterally in the maxillary-palatine suture just posterior or medial to thethird molar. The foramen is usually small andis underthrust by a flange of the maxilla; theforamen is absent in some specimens. A fo-ramen is always present in the anterior endof each fossa. In geomyines it plunges dor-sally or anterodorsally and appears to jointhe canal of the descending palatine artery.Hill (1935, p. 124) stated that in geomyinesit transmits the palatine vein. In heteromyidsthe foramen opens dorsally into the anteriorpart of the sphenopterygoid canal. Hartman(1980) stated that this parapterygoid foramentransmits a vein in Dipodomys. A medial fo-ramen in the wall separating the fossa andnasal passage is common in Liomys and Het-eromys and occurs in some geomyine speci-mens. It appears to have been covered bymembrane; in a few specimens it is presenton only one side. In geomyines the roof ofthe fossa ranges from solid to lacy. In mostgeomyine specimens, but not in heteromyids,a furrow passes from the fossa onto the pal-ate; in some specimens it is covered by boneand forms a short canal through the back ofthe palate. Dissection of Geomys and Tho-momys revealed that the palatine vein runsposteriorly through the furrow to join the in-ternal maxillary vein. This route is an alter-native to the primitive, dorsal course in whichthe vein ascends through the posterior max-illary foramen, and it may account for thesmall size and variability of the aperture.The infraorbital foramen is not visible in

front view, because its lateral margin is flushwith the wall of the snout. The infraorbitalcanal in geomyines is ventral or medial to thelong incisor alveolus, whereas in heteromy-ids, which have a shorter incisor, it is ventralto the alveolus. The infraorbital canal emerges

1985 5


op umlpgl paf eth nl

fty msc,sy bu' t

fo 5 mm

sty paf ac

ju mbf vf ppl inhy pom vt

1 Pappogeomys 2 Heteromys

FIG. 1. Pappogeomys merriami, AMNH 143208.FIG. 2. Heteromys lepturus, AMNH 172467. Skulls. Cross hatching indicates imagined cut through

bone.Abbreviations: For foramina and other structures (as named): ac, posterior aperture of alisphenoid

canal; asq, anterior squamosal; bu, buccinator; cHu, canal of Huguier; eth, ethmoid; fo, foramen ovale;foa, accessory foramen ovale; hy, hypoglossal; ifo, infraorbital; in, incisive; ipm, interpremaxillary; ito,interorbital; ju, jugular; mbf, fissure medial to bulla; ms, mastoid; msc, masticatory; mt, mental; nl,nasolacrimal; op, optic; paf, post-alar fissure; pgl, post glenoid; pom, posterior maxillary; ppl, posteriorpalatine; rp, rostral perforation; spl, sphenopalatine; spt, sphenopterygoid canal; spv, sphenopalatinevacuity; st, stapedial; sty, stylomastoid; t, temporal; uaf, unossified area between alisphenoid and frontalbones; uml, unossified area between maxillary and lacrimal bones; vf, venous foramen in parapterygoidfossa.

at a depression in the side of the snout. Ingeomyines the depression has a medial wall,but in heteromyids the bone always is per-forated; the prominent hole in the maxillawas covered by membrane in life. The pre-maxillary-maxillary suture is anterior to thedepression in most geomyine specimens. Inheteromyines, Microdipodops, and somespecimens of perognathines the suture runsinto the perforation, whereas in Dipodomys

and other perognathines a thin lamina of themaxilla surrounds the anterior end ofthe per-foration. The scar of the superficial masseteris spread ventral and posteroventral to theforamen in geomyines. It is more restrictedin most heteromyid specimens, and the scaris confined ventral to the foramen. The lengthof the infraorbital canal in most geomyinesis between 15 and 20 percent of the condy-lobasilar length; in Thomomys the range is

6 NO. 2812


mt

3 Perognathus 4 Microdipodops

FIG. 3. Perognathus amplus, AMNH 136987.FIG. 4. Microdipodops megacephalus, AMNH

from 1 1 to 15 percent, and in Liomys from19 to 24 percent. The anterior alveolar fo-ramen is hidden deep within the canal.The lacrimal foramen is anterodorsal to the

orbital aperture ofthe infraorbital canal. Thelacrimal spine overhangs it to a varying de-gree. The descending channel and lateralmargin ofthe foramen are formed by the lac-rimal bone posteriorly and by the maxillaanteriorly. In geomyines the initial course ofthe lacrimal canal is ventral or posteroven-tral. The canal passes ventrolateral to the endofthe incisor alveolus, which extends low andterminates at or in the root ofthe zygoma; atthis point the lacrimal canal is lateral to theinfraorbital canal. The lacrimal canal thenascends anteriorly until it is dorsal to the in-fraorbital foramen, where it may form a bulge

135612. Skulls. Abbreviations as in figure 2.

in the rostral depression. In heteromyids theincisor alveolus ends high on the snout anddoes not obstruct passage of the lacrimal ca-nal. The canal has an anterior orientation; itbegins dorsolateral to the posterior entranceto the infraorbital canal. The infraorbital ca-nal is visible through the thin wall of therostrum and can be seen to pass dorsomedialto the perforation. A nonossification is com-mon in geomyines and heteromyids at or nearthe intersection of the frontal-maxillary su-ture and the edge of the Lacrimal bone. It isusually absent in Dipodomys.The sphenopalatine foramen is medial to

the alveolar capsules that protrude into theorbit in geomyines. Its position in living formsranges from just anterior to the capsule ofthefirst molar to the space between the capsules

71985


ofthe first and second molars. In heteromyidsthe sphenopalatine is dorsal to a portion ofthe range from the junction ofthe fourth pre-molar and first molar to the posterior part ofthe first molar. The maxilla always forms theanterior and ventrolateral boundaries of theaperture and never surrounds the entireopening. The orbital lamina of the palatinereaches the back ofthe foramen in most spec-imens; the frontal and orbitosphenoid bonesparticipate in the margin of the foramen insome specimens. Variation in the number ofbones bounding the sphenopalatine foramenis great, and no taxonomic significance canbe assigned to any particular combination ofthem.

Ossification of the dorsal part of the or-bitosphenoid and posteroventral part of theorbital lamina of the frontal is commonlyincomplete in geomyines. This aperture isconfluent ventrally with the optic foramenand dorsally with the ethmoid foramen inmany specimens. Commonly, it extends ven-trally between the orbitosphenoid and ali-sphenoid bones and is continuous with theanteriorly enlarged orbital fissure. Amongheteromyids complete orbital walls are foundin Heteromys and Liomys. The nonossifiedarea is present in Microdipodops and somespecimens of Dipodomys, and it is large inall perognathines. The orbital fissure is largein all heteromyids, and a thin bar of boneseparates it from the posterior part of theoptic foramen.The ethmoid foramen is dorsal to a zone

that extends from the middle of the first mo-lar to the middle of the second. It is unitedwith the nonossified area in many geomyinespecimens, in perognathines, Microdipodops,and many Dipodomys. In other geomyine andDipodomys specimens it is within the frontalbone. In heteromyines it is in or near thefrontal-orbitosphenoid suture.The optic foramen in Heteromys and most

geomyine specimens is less than 1 mm indiameter, about 1 mm in Liomys, and largerthan 1 mm in perognathines and dipodo-myines. It is completely surrounded by theorbitosphenoid in heteromyids except someperognathine specimens; the degree to whichit is enclosed varies in geomyines. The po-sition ofthe optic foramen ranges from abovethe junction of the first and second molars to

the middle ofthe second molar in geomyines.It is farther posterior in most heteromyidsand ranges from the junction of the secondand third molars to a position posterior tothe third. An interorbital foramen is presentin the orbitosphenoid bone anteroventral orventral to the optic foramen in geomyinesand heteromyines. The interorbital foramenis present in a few specimens of perognath-ines; this suggests that it may be united withthe large optic foramen in the rest of the pe-rognathine specimens and in the dipodo-myines. Hartman (1980) called this aperturethe presphenoid foramen; he observed it andrecorded its variability in Dipodomys. Thelateral sides of the orbitosphenoid are flat-tened where the internal pterygoid muscleoriginates. In heteromyines the area is pos-teroventral to the optic foramen; in some di-podomyines it extends farther anteriorly. Ingeomyines, too, it extends anteriorly and isventral to the optic foramen. The region isnot sculpted in this fashion in perognathines.The dorsal palatine foramen is hidden

within the anterior-alar fissure (formerlytermed sphenoidal fissure) in geomyines andis medial or anteromedial to the anterior edgeof the fissure in heteromyids. In most spec-imens the foramen is between the alveoli ofthe second and third molars.The sphenofrontal foramen is lacking in all

but Dipodomys where it may be the largeopening in the dorsal curve of the anterior-alar fissure. The content of the opening hasnot been verified by dissection. The foramenis not seen, but may be included in the largeorbital nonossification in Perognathus. Bothof these heteromyid genera have a stapedialartery that retains all three branches (Bugge,1971, p. 347), and the foramen is to be ex-pected. Nutritive foramina are present in themaxillary floor of the orbit.The ventral root ofthe anterior-alar fissure,

which is formed by the maxilla and ali-sphenoid bones, arises above a zone thatranges from the second molar alveolus to thejunction ofthe second and third molar alveoliin geomyines. It arises dorsal or slightly an-terior to the third molar in heteromyines,usually just posterior to the third molar inperognathines and Microdipodops, and pos-terior to that tooth in Dipodomys.The internal maxillary artery runs within

8 NO. 2812


the alisphenoid canal up to the level of themasticatory foramen where its medial wallends, and the passage joins the sphenopter-ygoid canal. This anterior opening can be seenwithin the masticatory foramen in geo-myines, Heteromys, and some Liomys. In di-podomyines and perognathines the enclosedpart of the alisphenoid canal is extremelyshort. Masticatory and buccinator foraminaare separated in Microdipodops, most speci-mens of geomyines and many perognathinespecimens. The foramina are joined as a sin-gle aperture in heteromyines and Dipodomys.Hartman (1980) incorrectly called this theforamen ovale; he noted that it is occasionallydivided in Dipodomys. The canals to theseforamina open laterally from the enclosed partof the alisphenoid canal.The lateral pterygoid flange forms the lat-

eral wall ofthe sphenopterygoid canal. In liv-ing geomyoids it usually fails to reach theauditory bulla, and the accessory foramenovale is rarely enclosed posteriorly by bone.The alisphenoid canal enters the bone ante-rior to the foramen ovale and just medial tothe anterior margin of the accessory foramenovale. In all specimens the foramen ovale isopen posteriorly, and thus is bounded by theauditory bulla. Usually the foramen ovale iscontinuous laterally with an opening that in-cludes the post-alar fissure and postglenoidforamen. In dissected specimens and in someof the skulls examined, a membrane coversthe fissure. In some specimens ofGeomys andin one of Liomys a spur of the alisphenoiddorsolateral to the foramen ovale reaches thebulla and separates the foramen ovale fromthe post-alar fissure. In most geomyine spec-imens this spur is present in varying degreebut does not reach the bulla. Occasionally thepostglenoid foramen is separate and sur-rounded by the squamosal and the bulla; itis not entirely enclosed by the squamosal.The sphenopterygoid canal is a prominent

feature of the pterygoid region. The internalpterygoid muscle arises from its walls andfrom a flattened area on the medial wall ofthe orbit posteroventral to the optic foramen.In Dipodomys a large vein emerges from thecanal (Howell, 1932, p. 480). The position ofthe entrance to the transverse canal is vari-able. In some geomyines it is at the base ofthe hamular process in the posteromedial wall

of the sphenopterygoid canal; in others it be-gins laterally, in the medial wall of the ali-sphenoid canal. It is lacking in some geo-myine and most heteromyid specimens.The carotid foramen is situated medial to

the auditory bulla near the junction of thebasioccipital and basisphenoid bones. Inperognathines it is usually behind the suture.A slight gap between the anterior part of thebulla and the medial basicranial elements en-larges the foramen in heteromyids. The gapis very long in dipodomyines and includesthe jugular foramen.The internal structure of the bulla is vesic-

ular in all geomyoids except dipodomyines.The vesicles are open into the middle earcavity. The vesicular texture is most exten-sive in perognathines. The bulla in dipodo-myines is made of a thin lamella ofbone, butthis is achieved through remodeling of a tra-becular precursor (Webster, 1975), which isprobably the beginning of vesicular texturein other taxa. The auditory bullae are strik-ingly enlarged in perognathines and are enor-mous in dipodomyines. In perognathines theanteromedial parts ofthe bullae, which housethe eustachian tubes, approach one anotherand are appressed to the ventral side of thebasisphenoid; in a few specimens these ex-tensions meet in the midline. In dipodo-myines the extensions meet in a broad junc-tion and are similarly applied to the surfaceof the basisphenoid; in these rodents this isthe only contact of the anterior and medialsides of the bulla with the rest of the skull.The stapedial foramen is absent in hetero-

myines and absent or marked only by a pitor tiny perforation in geomyines. The fora-men is ofmoderate size in perognathines anddipodomyines and enters the medial wall ofthe bulla anteroventral to the jugular fora-men. In dipodomyines it is quite far anterior,and the course of the stapedial canal is pos-terolateral. Bugge (1971) observed that inPerognathus and Dipodomys the stapedial ar-tery supplies blood to the dura mater, orbit,and upper and lower jaws; this is the prim-itive pattern in rodents. In Geomys the ex-ternal carotid has, via anastomosis, taken overthe entire area of stapedial supply. Bugge didnot examine the carotid circulation in het-eromyines.The jugular foramen is a lenticular slot be-

91985


tween the basioccipital and the postero-medial part ofthe auditory bulla and periotic.It is open anteriorly into a gap between theanterior parts of these elements in dipodo-myines. The hypoglossal foramen is single ormultiple and faces anterolaterally toward thejugular foramen; in some specimens it is hid-den by an anterolaterally projecting flangefrom the ventral surface of the basioccipital.Temporal foramina are absent in most spec-imens. Dipodomyines have one or two fo-ramina in the squamosal bone dorsal to thezygomatic root; they are just in front of theanterior bulge of the inflated auditory bulla.I have named the aperture the anterior squa-mosal foramen; Hartman (1980) called it thesquamosal foramen.The stylomastoid foramen is between the

bony auditory tube and the mastoid. Thesetwo elements are often fused. In many spec-imens a furrow leads ventromedially fromthe foramen; it runs along the anterior edgeof the paroccipital process and approachesthe jugular foramen. The feature is not seenin dipodomyines. The mastoid foramen ingeomyines is a thin slit dorsal to the medialpart ofthe mastoid on the face ofthe occiput,and it is overhung by the edge ofthe occipitalbone in many specimens. The foramen is mi-nute or absent in heteromyids. Tiny condy-loid foramina are variable in their presenceor absence.

OTHER CRANIAL FEATURES

The adductor musculature of the jaw hasshifted anteriorly in geomyoids relative tothat in primitive rodents. The sciuromor-phous condition in which the origin of thelateral masseter extends anteriorly onto therostrum illustrates this point. The shift is ap-parent also in the temporal and internal pter-ygoid muscles and in the insertion of the lat-eral masseter on the mandible. The origin ofthe internal pterygoid muscle is describedabove in conjunction with the orbit and thesphenopterygoid canal.The origin of the temporal muscle differs

among geomyoids. The muscle is extensivein geomyines; a median sagittal crest is com-mon but not universal in the family. A pairof temporal crests occurs in some specimensof each genus and is most common and far-

thest lateral in Thomomys. The temporalmuscle is smaller in heteromyids. Temporalcrests are far lateral in heteromyines, and thenarrow origin of the muscle reaches the backof the squamosal. The inflated bulla in pe-rognathines and dipodomyines restricts themuscle origin still more, and it does not ex-tend posteriorly beyond the zygomatic root.The temporal muscle is modified uniquely

in geomyines. Dissection of Geomys andThomomys revealed that a tough, smoothcapsule of connective tissue covers the pro-tuberance of the squamosal anterior to theglenoid fossa. A part ofthe posterior divisionofthe temporal muscle rides over this pulley-like structure, and the direction of its forceis made vertical to the skull axis. The deepanterior part of the temporal muscle arisesin a unique fossa on the posterior wall of theorbit. The alisphenoid bone forms most ofthe fossa and ascends almost to the skull ta-ble; it is separated from the parietal by thelong anterodorsal extension of the squamo-sal. The fossa is a single channel in Thomo-mys. In the other living geomyines a verticalridge divides the dorsal part of the fossa intwo.

In heteromyines and perognathines theprotuberance on the squamosal is present butweaker than in most geomyids; its functionhas not been verified by dissection. The struc-ture is not present in dipodomyines. In het-eromyids there is no furrow in the orbit thatwould indicate the presence of a special an-terior part of the temporal muscle; the ali-sphenoid bone does not ascend to the skulltable, and it usually meets the squamosal nearthe middle of the vertical dimension of theorbit.The mandibles of heteromyids and geo-

myines are markedly different. In hetero-myids the angle of the jaw is a distinct flangethat curves ventrolaterally from the body ofthe ramus. It terminates posterodorsally in aflaring superior angular process. Howell(1932, p. 411) observed that in Dipodomysthe posterior part ofthe deep lateral masseter,which he termed masseter posterior, insertson the lateral and medial surfaces ofthis pro-cess, which he distinguished from the angle.The region is similar in other heteromyids.Nikolai and Bramble (1983) proposed thatthe eversion of the angular process reduces

10 NO. 2812


crowding of the mandible by the enlargedbullae in perognathines and dipodomyines.Since a somewhat everted angle occurs in het-eromyines, which lack great bullar inflation,the design may be associated primarily witha certain mode of chewing and gnawing. Ingeomyines the angle is reduced and does notform anything more than a ridge on the bodyof the ramus; the superficial masseter andinternal pterygoid insert here on the lateraland medial sides ofthe jaw, respectively. Thesuperior angular process is robust and pro-jects laterally from the back ofthe ramus; theinsertion of the posterior part of the deeplateral masseter surrounds it. Hill (1937, p.104) called the structure the angular process,and Merriam (1895) labeled it the angle inhis figures. Such usage is incorrect, since thetrue angle is associated with insertion of thesuperficial masseter and internal pterygoidmuscles.The coronoid process, on which the tem-

poral muscle inserts, is strong in geomyinesand extends farther dorsally than the con-dyloid process. It is much weaker and lessextensive in heteromyids. In geomyines theinsertion ofthe temporal muscle is primarilyon the medial side of the process and in aprominent depression (the basitemporal fos-sa of Russell, 1968) that extends anteriorlybetween the process and the last two molars.The depression is shallowest in Thomomys,and it is lacking in heteromyids. In somespecimens ofDipodomys a depression is pres-ent, but it is associated with the mandibularforamen. In geomyines this foramen is pos-terior to the depression.The alveolus of the lower incisor is long

and produces a lateral bulge on the mandible.In geomyines the bulge is between the su-perior angular process and the condyloid pro-cess; in heteromyids it is farther anterior. Theposterior part of the medial masseter insertsbetween the alveolar bulge and the condyloidand coronoid processes in geomyines; the in-sertion is usually marked by a pit in geo-myines, but it is not a distinct structure inheteromyids.The insertions of the lateral and medial

divisions of the masseter create a shallowmasseteric fossa on the lateral side of themandible. The fossa is roughly triangular, andan apex marks the anteriormost extent ofthe

insertion. The single mental foramen is an-terior to this point. In all geomyoids the apexis farther forward than in primitive rodents.In geomyines it is ventral to the anterior partof the fourth premolar. In heteromyids theapex is usually extended by a ridge or is itselfanteroventral to that tooth.

EXTINCT TAXAIn addition to the florentiamyids, de-

scribed elsewhere (Wahlert, 1 983, 1 984), cra-nial remains of three other extinct geomyoidtaxa are known in detail sufficient for inclu-sion in this study. Entoptychus occurs in theearly Miocene John Day Formation of Ore-gon, and Pleurolicus in slightly older levelsof the same formation. Rensberger (1971,1973a) and Wahlert (1972) described theseskulls. Schizodontomys of early middle Mio-cene age occurs in both the upper part of theJohn Day Formation and strata ofsimilar agein South Dakota; Rensberger (1973a) de-scribed a partial skull, and Munthe (1981)expanded this description with a more com-plete cranium and partial skeleton. All ofthesetaxa bear the hallmarks of the Geomyoidea,but they have a variety of specializations.Entoptychus (fig. 5) retains certain primi-

tive characters that do not occur in the livinggeomyines and heteromyids. The accessoryforamen ovale (identified by Rensberger asthe foramen ovale) is enclosed within the ali-sphenoid; the foramen ovale appears to bebetween the pterygoid region and the bulla.The alisphenoid reaches the bulla and sepa-rates the foramen ovale from the post-alarfissure. Rensberger (1971) said that thesphenopterygoid canal is lacking. I have dis-agreed (1972), but the region is damaged inall specimens that I have seen. This detail isan important point in placement ofthe groupand should be reexamined. The postglenoidforamen is within the squamosal bone, anda temporal foramen is present above the pos-terior part of the zygomatic root.Entoptychus shares certain derived fea-

tures with the heteromyids and geomyines.A temporal tuberosity is present on the squa-mosal bone anterior to the glenoid fossa. Thesuperior angular process projects laterallyfrom the mandible.Entoptychus has derived characters of its

1985 I1I


1 cm

I j* St I ?sPt v I ihy ac ?sp ¾f pp in

FIG. 5. Entoptychus sp., University of California Museum of Paleontology no. 65251. Skull. Abbre-viations as in figures 1 and 2. Modified from Wahlert (1972, fig. 19).

own. The posterior part of the squamosal isnarrow above the auditory tube. The audi-tory bullae, which are not trabecular, areswollen ventrally; anteromedial processesfrom opposite sides touch in the midline justanterior to a peculiarly swollen region of thebasioccipital and basisphenoid. Rensbergerreported that a narrow slit separates the bullaand basioccipital; it is confluent posteriorlywith the jugular foramen. A fossa for inser-tion of the temporalis muscle occurs on themandible lateral to M3; it is not as pro-nounced as that of geomyines.The skull of Pleurolicus is not as fully

known. Rensberger stated that the pterygoidfossa is shallow. The postglenoid foramen issituated between the squamosal and the au-ditory tube. There is a single temporal fora-men. A large temporal tuberosity can be seenon the squamosal anterior to the glenoid fossain Rensberger's figure and plates of skulls.

The superior angular process projects later-ally. Specialization of the auditory and basi-cranial regions is like that of Entoptychus;the bullae are somewhat more inflated. Rens-berger also described skeletal remains andfound many features indicating fossorial spe-cialization in Entoptychus (1971, p. 140) andPleurolicus (1973a, p. 81).Schizodontomys is superficially similar to

these two genera, but Munthe (1981) notedimportant differences. The infraorbital fora-men appears to be enlarged by an unossifiedarea anterior to the canal. The sphenopter-ygoid canal is present posterior to the shallowparapterygoid fossa. The foramen ovale is en-larged by a gap between the pterygoid regionand the bulla. The accessory foramen ovaleand temporal foramen are lacking. The post-glenoid foramen, between the squamosal boneand the auditory bulla, is confluent with thepost-alar fissure; the alisphenoid bone sepa-

12 NO. 2812


co

0

0

oOC-4f

0)Q -

FIG. 6. Proposed relationships of the Geomyoidea. Numbers at nodes correspond to numbers indiscussion.

rates this aperture from the foramen ovale.Masticatory and buccinator foramina areconfluent. As in Entoptychus and Pleurolicusthere is a stapedial foramen; the bony tubefor the artery is preserved crossing the prom-ontorium. The auditory bullae are inflatedand meet in a broad junction in the ventralmidline, but the basioccipital is not swollenbehind the junction. Inflation ofthe auditoryregion exceeds that ofEntoptychus and Pleu-rolicus, since the mastoid is swollen poste-riorly and dorsally. The posterior part of thesquamosal is very narrow above the auditorytube. There is no area ofthe orbit specializedfor origin of a part of the temporal muscle.The angle of the jaw exhibits gentle lateralbending and the superior angular process isnot set apart. The extent of the incisors is asin heteromyids. Munthe described the skel-eton ofSchizodontomys and found many fea-

tures indicating that it was at least quadru-pedally saltatorial; she pointed out that thesespecializations make it morphologically in-termediate between Heteromys and Dipod-omys.

DISCUSSION

The data of cranial foramina and othercharacters that were observed support thetraditional hypothesis ofrelationships amongliving geomyoids. The cladogram (fig. 6) andfollowing summary of shared, derived char-acters add to the basis of this phylogeny andinclude data from extinct taxa. Numbers inthe list correspond to numbered nodes in thecladogram. Additional characteristics of theliving taxa may be found in the recent anal-yses by Hall (1981); Rensberger and Woodbest summarized other characters of the

00

O ..0

q.)0

131985


extinct taxa in the papers cited. The spectrumof primitive and derived conditions of char-acters is based on comparison with Paramys,Sciuravus, and a sample of muroid rodents(Wahlert, 1974, 1978).

1. Myomorpha: Carotid canal short, nearbasisphenoid-basioccipital junction; en-trance to transverse canal separated from ali-sphenoid canal; zygomatic arch slender; in-cisor enamel with uniserial Hunter-Schregerbands in the portio interna.

2. Geomorpha: Fossa on rostrum for an-teriorly extended lateral masseter (sciuro-morphy); long, low infraorbital canal sunkinto side of rostrum at anterior end; spheno-palatine foramen far anterior, dorsal to M1;interorbital foramen present (apparentlyunited with optic foramen in many perogna-thine and dipodomyine specimens); wear oncheek teeth forming transverse lophs.

3. Eomyoidea: Complex Schmelzmuster inenamel of lower incisors with uniserial Hun-ter-Schreger bands longitudinal; sphenopa-latine foramen completely surrounded bymaxilla; masticatory and buccinator foram-ina united with accessory foramen ovale;temporal foramen absent.

4. Geomyoidea: Incisive foramina short,30 percent or less ofdiastemal length (excep-tion, Dipodomyinae); parapterygoid fossae,usually with foramen, present between palateand pterygoid fossae; auditory bulla formingback of foramen ovale; cheek teeth bilopho-dont; transverse lophs in upper and lowercheek teeth widened by addition of styles.

5. Florentiamyidae: Special process frompalatine, together with alisphenoid, formingedge of anterior-alar fissure; canal for de-scending palatine vein lateral to anterior-alarfissure; optic foramen larger than 1.0 mm;masticatory and buccinator foramina unitedwith accessory foramen ovale; entostyle ofupper molars elongated anteroposteriorly andblocking lingual end of transverse valley.

6. Unnamed node: Pterygoid fossa deep;superior angular process of mandible flaredlaterally.

7. Entoptychidae: Auditory bullae inflatedventrally with anteromedial processes meet-ing in midline; gap between bulla and basi-occipital; contiguous parts ofbasioccipital andbasisphenoid swollen ventrally.

8. Unnamed node: Fur lined cheek pouch-

es present outside the mouth (this charactermay have arisen at a lower node); incisiveforamina 15 percent or less ofdiastemal length(exception, Dipodomyinae); pterygoid fossacontaining entrance to sphenopterygoid ca-nal; origin of internal pterygoid muscle ex-tending through sphenopterygoid canal to-ward orbit; posterior margin of accessoryforamen ovale not ossified, and lateral pter-ygoid flange often failing to reach auditorybulla; postglenoid foramen between squa-mosal bone and auditory bulla, often con-fluent with post-alar fissure; temporal fora-men absent; auditory bullae with highlyvesicular texture (exception, Dipodomyi-nae).

9. Heteromyidae: Large perforation ofros-trum at and anterior to infraorbital foramen;interpremaxillary foramen absent; posteriorend ofoptic foramen narrowly separated fromorbital fissure; mastoid foramen minute orlacking; origin of temporal muscle restrictedto lateral part of skull roof; coronoid processon mandible small and low; insertion ofmas-seter on mandible usually marked by ridgeextending anteroventral to P4.

10. Heteromyinae: Ventral root of ante-rior-alar fissure rising above M3; masticatoryand buccinator foramina united; stapedial andsphenofrontal foramina absent.

11. Heteromys: Optic foramen smaller than1.0 mm.

12. Unnamed node: Optic foramen largerthan 1.0 mm; nonossification common dor-sal to orbitosphenoid; auditory bullae highlyinflated with anteroventral processes ap-proaching each other and touching basisphe-noid; squamosal bone extending as thin pro-cess posteriorly above bony auditory meatus;origin of temporal muscle restricted chieflyto orbit.

13. Perognathinae: Small skull; large or-bital nonossification usually including eth-moid foramen.

14. Dipodomyinae: Auditory bullae fur-ther inflated, especially dorsal part; antero-ventral processes of bullae meet in broadjunction in midline.

15. Unnamed node: Incisive foraminaelongated posteriorly with premaxillary-maxillary suture intersecting lateral marginsnear middle; posterior part of squamosallacking; postglenoid foramen absent; anterior

14 NO. 2812


squamosal foramen present; further inflationof auditory bullae; gap present between bullaand medial basicranial elements and includ-ing carotid foramen at anterior end and jug-ular foramen at posterior; stapedial canal en-tering auditory region anterior to its usualposition.

16. Microdipodops: Posterior palatine fo-ramina in palatine bones (may be primitivecondition); bullae slightly more inflated thanin Dipodomys (Webster and Webster, 1975).

17. Dipodomys: Masticatory and buccina-tor foramina usually united; cheek teeth hyp-sodont.

18. Geomyidae: Diastema strongly arched;palate deeply furrowed anterior to posteriorpalatine foramina; furrow or canal connect-ing parapterygoid fossa and palate; initialcourse of lacrimal canal ventral; optic fora-men smaller than 1.0 mm and far anterior,dorsal to a zone from the junction of M'M2to M2; nonossification common dorsal to or-bitosphenoid; ventral root of anterior-alarfissure above a zone from alveoli of M2 toM2M3 junction; dorsal palatine foramen hid-den within anterior-alar fissure; anteromedialopening of alisphenoid canal visible withinmasticatory foramen; sphenofrontal foramenlacking; stapedial foramen and functional ar-tery lacking; origin ofinternal pterygoid mus-cle extending anteriorly through sphenopter-ygoid canal as far as or beyond optic foramen;posterior part of maxilla meeting anteroven-tral part of alisphenoid in long suture; bosson squamosal redirecting pull ofposterior partof temporal muscle; deep anterior part oftemporal muscle arising in broad vertical fos-sa in orbit; alisphenoid bone extended dor-sally almost to skull table and sculpted byfossa for temporal muscle; insertion of tem-poral muscle in broad pit between coronoidprocess ofmandible and posterior cheek teeth;angle reduced to ridge on mandible; robustsuperior angular process projecting laterallybelow condyloid process; alveoli ofupper andlower incisors with great posterior extent.

19. Thomomys: Infraorbital canal relative-ly short.

20. Geomys, Orthogeomys, Pappogeomys:Fossa on alisphenoid for origin of deep an-terior part oftemporal muscle divided in twoby vertical ridge in dorsal part.

21. Other Myomorpha: Enlarged infraor-

bital foramen; lacrimal foramen low in max-illa, medial or anteromedial to infraorbitalforamen.

It is tempting to treat sciuromorphy as aprimary character and unite the eomyoids,geomyoids, sciuroids, and castoroids in a sin-gle taxonomic group as did early mammal-ogists. However, sciuromorphy, the exten-sion ofthe origin ofthe lateral masseter ontothe rostrum, has arisen more than once amongrodents. The extinct ischyromyoid Titano-theriomys is sciuromorphous (Wood, 1976),but it is not related to any ofthe groups men-tioned above. Anterior extension of the ori-gin ofthe lateral masseter is also a componentof the myomorphous condition seen in mu-roids. The living sciurids exhibit a range indevelopment of the infraorbital canal thatmakes them more primitive than the knownextinct and living Geomorpha. The canal isno more than a hole in the maxillary root ofthe zygoma in Protoxerus, Eutamias, andTamias; it is a long, enclosed passage in mostother sciurids (Wahlert, 1972). Emry andThorington (1982) have described the Oli-gocene genus Protosciurus which has manymorphological features that are special to theSciuridae, but it retains the primitive, pro-trogomorphous condition in which the originof the lateral masseter is restricted to the zy-gomatic arch. Therefore, sciuromorphy doesnot even apply to all the members of thetaxon on which the name is based. In sciuridsand castorids the anterior margin of the in-fraorbital canal projects laterally from therostrum. The fact that in the Geomorpha theanterior end of the infraorbital canal is de-pressed into the rostrum and the anteriormargin of the canal is flush with the rostralwall may suggest independent origin of sciu-romorphy in the group. Anterior extensionof the origin of the lateral masseter is a com-mon specialization in rodents. By itself sciu-romorphy is not a sufficient criterion on whichto base a hypothesis of monophyly. Its in-discriminate use in classification masks thecomplexity of rodent evolution.The evidence ofcranial foramina that places

the Geomorpha in the Myomorpha is slim.This placement is based on the work of Hill(1937) and Wilson (1949). Hill found thatmany of the characters that geomyoids sharewith sciurids are interdependent, because they

1985 15


are related to sciuromorphous origin of thelateral masseter. Some other similarities areretentions of primitive characters. Hill listedmany characters of the skull, skeleton, andsoft anatomy that are chiefly specializationsshared by geomyoids and muroids. He sug-gested that the Geomyoidea be placed in theMyomorpha. Wilson (1949) compared theskull ofParadjidaumo, an Oligocene eomyid,with other sciuromorphous and myomor-phous rodents. He, too, distinguished prim-itive from specialized characters and con-cluded that "the Eomyidae should perhapsbe viewed as a relatively primitive group ofrodents which departed from the same branchof the Aplodontoidea that gave rise to theheteromyids and muroids -all three beingabout equally remote from one another, butthe heteromyids and eomyids less so; all ofcloser affinity to one another than to the sci-urids and castorids" (ibid., p. 48). Wilsonclearly proposed a monophyletic clade whosebranches include the Eomyidae, Heteromyi-dae, and Muroidea.The evidence for relationship of eomyids

to heteromyids was set forth by Wilson (1949).Wahlert (1978) included the geomyids inanalysis of cranial foramina and reached thesame conclusion. Unique, derived charactersof the eomyids indicate that they constitutea monophyletic clade that shares its ancestrywith the Geomyoidea.The Geomyoidea share a suite of derived

characters that make them distinctly differentfrom the eomyids. Correct identification ofthe parapterygoid fossae is important. Fossilskulls are often broken immediately behindthese fossae, especially if a sphenopterygoidcanal is present, and the parapterygoid fossaeare easily mistaken for the pterygoid fossaeproper.The skulls of florentiamyids were de-

scribed by Wood (1936a), Rensberger(1973b), and Wahlert (1983). The group isunusual for its combination of a primitivegeomyoid tooth crown pattern with a derivedmorphology of the anterior-alar fissure andcanal for the descending palatine vein that isunique among rodents. Damage anterior tothe infraorbital canal makes Wahlert's sup-position that the rostrum was imperforateuncertain. On the basis ofcurrent knowledge,

placement of the florentiamyids in any othergeomyoid group would make the definitionof that group unworkable.The entoptychids share derived characters

with geomyids and heteromyids. The supe-rior angular process of the mandible is flaredlaterally, and the pterygoid fossa is deep ifnot open dorsally into a sphenopterygoid ca-nal. Retention of primitive characters indi-cates that the Entoptychidae branched offbe-fore the splitting of the Heteromyidae andGeomyidae. These characters are the sup-posed lack ofa sphenopterygoid canal, reten-tion ofthe temporal foramen, enclosure with-in a single bone of the accessory foramenovale and of the postglenoid foramen. En-toptychus and Pleurolicus lack the rostral per-foration seen in heteromyids.The Heteromyidae and Geomyidae share

many derived characters. Chiefamong themis the presence of a sphenopterygoid canalwhich permits anterior extension of the in-ternal pterygoid muscle. Hill (1937, p. 159)stated that in heteromyids the internal pter-ygoid muscle does not invade the spheno-pterygoid canal; he said (1935) that the canalis reduced to a foramen for a large vein inPerognathus, Dipodomys, and Microdipo-dops. This claim appears to have been basedon Howell (1932) whom he cites, and on ex-amination of cleaned skulls of the taxa men-tioned. New dissection of Dipodomys re-vealed that a part of the internal pterygoidmuscle passes through the sphenopterygoidcanal anteriorly to the level ofthe base oftheanterior-alar fissure; passage of a large veinthrough the canal was verified. Other char-acters, such as the reduction of the posteriorparts of the alisphenoid and squamosal thatpermits confluence of several foramina mayreflect a common, derived arrangement ofthemasticatory system and supporting bony ele-ments. The fur-lined cheek pouches that areexternal to the mouth in geomyids and het-eromyids may have been present in extincttaxa of the other related groups.The Heteromyidae are united by the pres-

ence of a large perforation in the wall of therostrum anterior to the infraorbital canal. Theimportance ofthis character has been stressedby systematists for more than a century. Thereduced temporal muscle origin and small

16 NO. 2812


coronoid process are probably interdepen-dent characters. The Heteromyinae have themost primitive skulls among living geo-myoids. It is the shared primitive charactersand parallelism in loss ofthe stapedial arteryand forward position of the anterior-alar fis-sure that make heteromyines and geomyinesappear similar.The perognathine-dipodomyine branch is

characterized by enlargement of the auditorybullae and reduction of the squamosal andrestriction ofthe origin ofthe temporal mus-cle. The sense of vision appears to be espe-cially important. Anteroventral enlargementof the auditory bullae is accompanied in di-podornyines, as in Entoptychus, by separa-tion ofthe bulla on anterior and medial sidesfrom other bones of the cranium. The sepa-ration may reduce transmission through boneof sound that is generated in chewing. Theunion of bullar processes and contact be-tween this specialized region and the basi-sphenoid may give the animal a means ofdistinguishing and ignoring the sound pro-duced in its mouth. The incisive foramina ofdipodomyines are very long; their extensionfar posterior to the premaxillary-maxillarysuture indicates that this is a secondarily de-rived condition. The thin bone of the audi-tory bulla is remodeled from a trabecular pre-cursor and also appears to be secondarilyderived. The remarkable cranial specializa-tions of the Dipodomyinae stand in strongcontrast to those of the Geomyinae; close re-lationship of the two groups is unlikely, anddental similarities are interpreted as paral-lelism.The family Geomyidae, which includes

only the Geomyinae, is characterized by acomplex set of characters that in part reflectextreme specialization. Wilkins and Woods(1983) found that chewing is propalinal in allforms, and they proposed that both masti-catory and fossorial adaptations are impor-tant determinants of skull form. Such a longlist ofderived characters that set a group apartfrom its closest relatives indicates remotecommon ancestry, or rapid evolution, or acombination ofthe two hypotheses. The pro-posed phylogeny supports remote commonancestry of geomyids and heteromyids. Thelow level ofcranial differentiation among liv-

ing geomyine genera suggests either recentdiversification or long conservation of a suc-cessful design after a period of rapid evolu-tion.

CONCLUSIONSThe following classification is based on the

proposed phylogeny of the infraorder Geo-morpha. The genera described in this paperare given in parentheses.Infraorder Geomorpha Thaler, 1966

Superfamily Eomyoidea Deperet andDouxami, 1902

Family Eomyidae Deperet and Doux-ami, 1902

Superfamily Geomyoidea Weber, 1904Family Florentiamyidae Wood, 1936Family Entoptychidae Miller and Gid-

ley, 1918(Pleurolicus and Entoptychus)

Family Heteromyidae Allen and Chap-man, 1893

Subfamily Heteromyinae Alston, 1876(Liomys and Heteromys)

Subfamily Perognathinae Coues, 1875(Perognathus and Chaetodipus)

Subfamily Dipodomyinae Coues, 1875(?Schizodontomys, Microdipodops,

and Dipodomys)Family Geomyidae Gill, 1872

Subfamily Geomyinae Baird, 1857(Geomys, Orthogeomys, Pappogeo-

mys, and Thomomys)

This classification is conservative and fol-lows both mammalogists and paleontologistswith regard to the taxonomic rank of the liv-ing groups. Elevation of the Entoptychidaeand Florentiamyidae to families is based onthe early branching of these taxa from thegeomyoid lineage.The Geomyidae, which includes only the

Geomyinae, has the longest defining list ofderived characters among the groups studied;monophyly of this group has not been ques-tioned. I retain the subfamily Geomyinae fortwo reasons-it is commonly used by mam-malogists, and it distinguishes these generafrom the entoptychids, which are consideredto be a subfamily ofthe Geomyidae by manypaleontologists. Many of the geomyid spe-

1985 17


cializations are directly associated withchanges in the arrangement of the mastica-tory muscles and in dental morphology. Smallsize of the optic foramen may be associatedwith reduction of eyes and life in burrows.The reasons for loss of the stapedial arteryand new connections of its branches are ob-scure.Monophyly of the Heteromyidae is indi-

cated by reduced temporal musculature anda unique perforation in the rostrum anteriorto the infraorbital foramen. Among hetero-myids the Heteromyinae are primitive, andthey share many primitive cranial featureswith the Geomyidae. Rensberger (1971) not-ed this similarity. Enlargement of the bullaeand modification of the posterior part of theskull together with large size of the optic fo-ramen characterize the perognathine-dipod-omyine branch of the Heteromyidae. Thepresence of a sphenopterygoid canal and ros-tral perforation and dipodomyine-like spe-cializations reported by Munthe (1981) per-mit incorporation of Schizodontomys in thislineage. Shotwell (1967) and Lindsay (1972)grouped the living geomyids with the dipod-omyines; there is dental similarity, but lackof shared, derived cranial characters beliesthis proposed relationship.The Heteromyidae and Geomyidae are

each other's closest relatives. The presenceof a large sphenopterygoid canal and reduc-tion of the posterior edges of the alisphenoidand squamosal bones characterize thesegroups.The data do not support close relationship

of living geomyids and extinct entoptychids.The Entoptychidae have a deep pterygoidfossa, but the sphenopterygoid canal is lack-ing. The alisphenoid bone is extensive pos-teriorly and encloses the accessory foramenovale and postglenoid foramen. Despite re-tention of some primitive characters, thereare specializations; the enlarged bullae andanterior separation of bullae from the basi-occipital are derived features seen also in di-podomyines. Entoptychids lack the rostralperforation that is a characteristic ofthe Het-eromyidae. Other aspects of skull morphol-ogy resemble geomyid specializations; theseare interpreted as examples ofparallelism thatare associated with a burrowing mode of life.

I have described the skulls of the Floren-tiamyidae (1983) and Eomyoidea (1978) pre-viously. The new data do not contradict theconclusions reached in those studies. The in-fraorder Geomorpha is used here as a gath-ering taxon for all ofthese rodents that sharederived features; these include sciuromor-phous lateral masseter, long infraorbital ca-nal that is depressed into the rostrum, ante-rior position of the sphenopalatine foramen,and tooth crown pattern with transverselophs. Although these rodents are sciuro-morphous, other derived characters suggestthat they are members of the suborder Myo-morpha.

LITERATURE CITED

Allen, J. A., and F. M. Chapman1893. On a collection of mammals from the

island of Trinidad, with descriptions ofnew species. Bull. Amer. Mus. Nat. Hist.,vol. 5, pp. 203-234.

Alston, E. R.1876. On the classification ofthe order Glires.

Proc. Zool. Soc. London, pp. 61-98, pl.4.

Anderson, S.1966. Taxonomy of gophers, especially Tho-

momys in Chihuahua, Mexico. System-atic Zool., vol. 15, pp. 189-198.

Baird, S. F.1857. Mammals: general report upon the zo-

Best, T. ]1978.

Bugge, J.1971.

Coues, E

ology of the several Pacific railroadroutes. Repts., Explorations and Sur-veys for railroad route from MississippiRiver to Pacific Ocean, Washington,D.C., vol. 8, pt. 1, pp. xix-xlviii, 1-757,pls. 17-60.

L.Variation in kangaroo rats (genus Di-podomys) ofthe heermanni group in BajaCalifornia, Mexico. Jour. Mammal., vol.59, pp. 160-175.

The cephalic arterial system in sciuro-morphs with special reference to thesystematic classification ofrodents. ActaAnat., vol. 80, pp. 336-361.

1875. A critical review ofthe North AmericanSaccomyidae. Proc. Acad. Nat. Sci.Philadelphia, vol. 27, pp. 272-327.

1877. No. 8-Saccomyidae, pp. 481-542; andNo. 10-Geomyidae, pp. 601-629. InCoues, E., and J. A. Allen, Monographs

18 NO. 2812


ofNorth American Rodentia. Rept. U.S.Geol. Surv. Terr., vol. 11.

Deperet, C., and H. Douxami1902. Les vertebres oligocenes de Pyrimont-

Challonges (Savoie). Mem. Soc. Pale-ont. Suisse, vol. 29, pp. 1-90, pls. 1-6.

Emry, R. J., and R. W. Thorington, Jr.1982. Descriptive and comparative osteology

ofthe oldest fossil squirrel, Protosciurus(Rodentia: Sciuridae). SmithsonianContrib. Paleobiol., no. 47, pp. i-iv, 1-35.

Gill, T.1872. Arrangement of the families of mam-

mals with analytical tables. Smithson-ian Misc. Coll., vol. 11, pp. i-vi, 1-98.

Hafner, J. C., and M. S. Hafner1983. Evolutionary relationships of hetero-

myid rodents. Great Basin NaturalistMem., no. 7, pp. 3-29.

Hafner, M. S.1982. A biochemical investigation of geo-

myoid systematics (Mammalia: Roden-tia). Z. Zool. Syst. Evolut.-forsch., vol.20,pp. 118-130.

Hall, E. R.1981. Mammals ofNorth America. N.Y., John

Wiley and Sons, vol. 1, 600 pp. plusindexes 90 pp.

Hartman, S. E.1980. Geographic variation analysis ofDipod-

omys ordii using nonmetric cranial traits.Jour. Mammal., vol. 61, pp. 436-448.

Hill, J. E.1935. The cranial foramina in rodents. Jour.

Mammal., vol. 16, pp. 121-129.1937. Morphology ofthe pocket gopher mam-

malian genus Thomomys. Univ. Calif.Publ. Zool., vol. 42, pp. 81-172.

Howell, A. B.1932. The saltatorial rodent Dipodomys: the

functional and comparative anatomy ofits muscular and osseous systems. Proc.Amer. Acad. Arts Sci., vol. 67, pp. 375-536.

Li, C.-k.1974. A probable geomyoid rodent from mid-

dle Miocene of Linchu, Shantung. Ver-tebrata PalAsiatica, vol. 12, pp. 43-53,pls. 1-2. (Chinese with English sum-mary.)

Lindsay, E. H.1972. Small mammal fossils from the Barstow

Formation, California. Univ. Calif.Publ. Geol. Sci., vol. 93, pp. 1-104.

Merriam, C. H.1895. Monographic revision ofthe pocket go-

phers, family Geomyidae (exclusive ofthe species of Thomomys). U.S. Dept.Agr., North American Fauna, no.8, pp.1-258, pls. 1-19, maps 1-4.

Miller, G. S., Jr., and J. W. Gidley1918. Synopsis of the supergeneric groups of

rodents. Jour. Washington Acad. Sci.,vol. 8, pp. 431-448.

Munthe, K.1981. Skeletal morphology and function ofthe

Miocene rodent Schizodontomys hark-seni. PaleoBios, no. 35, pp. 1-33.

Nader, I. A.1978. Kangaroo rats: intraspecific variation in

Dipodomys spectabilis Merriam and Di-podomys deserti Stephens. Illinois Biol.Monogr., vol. 49, pp. 1-116.

Nikolai, J. C., and D. M. Bramble1983. Morphological structure and function in

desert heteromyid rodents. Great BasinNaturalist Mem., no. 7, pp. 44-64.

Reeder, W. G.1956. A review ofTertiary rodents ofthe fam-

ily Heteromyidae. University Micro-films Internat., Ann Arbor (facsimileproduced in 1979), 618 pp.

Rensberger, J. M.1971. Entoptychine pocket gophers (Mam-

malia, Geomyoidea) of the early Mio-cene John Day Formation, Oregon.Univ. Calif. Publ. Geol. Sci., vol. 90,pp. i-vi, 1-209.

1973a. Pleurolicine rodents (Geomyoidea) ofthe John Day Formation, Oregon. Ibid.,vol. 102, pp. i-vi, 1-95, pls. 1-17.

1973b. Sanctimus (Mammalia, Rodentia) andthe phyletic relationships of the largeArikareean geomyoids. Jour. Paleont.,vol. 47, pp. 835-853.

Russell, R. J.1968. Evolution and classification ofthe pock-

et gophers ofthe subfamily Geomyinae.Univ. Kansas Publ. Mus. Nat. Hist., vol.16, pp. 473-579.

Shotwell, J. A.1967. Late Tertiary geomyoid rodents of Or-

egon. Bull. Mus. Nat. Hist. Univ. Ore-gon, no. 9, pp. 1-51.

Thaler, L.1966. Les rongeurs fossiles du Bas-Languedoc

dans leurs rapports avec l'histoire desfaunes et la stratigraphie du Tertiaired'Europe. Mem. Mus. Natl. d'Hist.Natur., Ser. C, Sci. Terre, vol. 17, pp.1-295, pls. 1-27.

Wahlert, J. H.1972. The cranial foramina of protrogomor-

1985 19


phous and sciuromorphous rodents; ananatomical and phylogenetic study. Un-pubi. Ph.D. thesis, Harvard Univ.,Cambridge, 230 pp.

1974. The cranial foramina of protrogomor-phous rodents; an anatomical and phy-logenetic study. Bull. Mus. Comp. Zool.,vol. 146, pp. 363-410.

1978. Cranial foramina and relationships ofthe Eomyoidea (Rodentia, Geomor-pha). Skull and upper teeth of Kan-sasimys. Amer. Mus. Novitates, no.2645, pp. 1-16.

1983. Relationships of the Florentiamyidae(Rodentia, Geomyoidea) based on cra-nial and dental morphology. Ibid., no.2769, pp. 1-23.

1984. Kirkomys, a new florentiamyid (Roden-tia, Geomyoidea) from the Whitneyanof Sioux County, Nebraska. Ibid., no.2793, pp. 1-8.

Weber, M.1904. Die Saiugetiere. Gustav Fischer, Jena,

866 pp.Webster, D. B.

1975. Auditory systems of Heteromyidae:postnatal development of the ear of Di-podomys merriami. Jour. Morphol., vol.146, pp. 377-394.

Webster, D. B., and M. Webster1975. Auditory systems of Heteromyidae:

functional morphology and evolution ofthe middle ear. Jour. Morphol., vol. 146,pp. 343-376.

Wilkins, K. T., and C. A. Woods1983. Modes of mastication in pocket go-

phers. Jour. Mammal., vol. 64, pp. 636-641.

Williams, S. L., and H. H. Genoways1978. Review of the desert pocket gopher,

Geomys arenarius (Mammalia: Roden-tia). Ann. Carnegie Mus., vol. 47, pp.541-570.

1980. Morphological variation in the south-eastern pocket gopher, Geomys pinetis(Mammalia: Rodentia). Ibid., vol. 49,pp. 405-453.

1981. Systematic review of the Texas pocketgopher, Geomys personatus (Mamma-lia: Rodentia). Ibid., vol. 50, pp. 435-473.

Wilson, R. W.1936. A Pliocene rodent fauna from Smiths

Valley, Nevada. Contrib. Paleontol.,Carnegie Inst. Washington Publ. no. 473,pp. 15-34, pls. 1-2.

1949. On some White River fossil rodents.Ibid., no. 584, pp. 27-50, pls. 1-2.

Wood, A. E.1935. Evolution and relationships of the het-

eromyid rodents. Ann. Carnegie Mus.,vol. 24, pp. 73-262.

1936a. A new subfamily ofheteromyid rodentsfrom the Miocene of western UnitedStates. Amer. Jour. Sci., ser. 5, vol. 31,pp. 41-49.

1936b. Geomyid rodents from the middle Ter-tiary. Amer. Mus. Novitates, no. 866,pp. 1-3 1.

1976. The Oligocene rodents Ischyromys andTitanotheriomys and the content of thefamily Ischyromyidae. Athlon: Essayson palaeontology in honour of LorisShano Russell, C. S. Churcher (ed.).Royal Ontario Mus., Life Sci. Misc.Publ., pp. 244-277.

NO. 281220

Norntates - COnnecting REpositories · 2017-03-10 · rodents that range from southwestern...

Documents

Transcript of Norntates - COnnecting REpositories · 2017-03-10 · rodents that range from southwestern...