RSITY OF KANSAS miscellaneous

JM OF NATURAL HISTORY publication

NO. 74

Relationships of Pocket Gophers
of the Genus Geomys from the
Central and Northern Great Plains

By

Lawrence R. Heaney

and

Robert M. Timm

UNIVERSITY OF KANSAS

LAWRENCE 1983 JUNE 1,1983

hi—

737

AH 3

UNIVERSITY OF KANSAS PUBLICATIONS
MUSEUM OF NATURAL HISTORY

The University of Kansas Publications, Museum of Natural History, beginning
with volume 1 in 1946, was discontinued with volume 20 in 1971. Shorter research
papers formerly published in the above series are now published as Occasional
Papers, Museum of Natural History. The Miscellaneous Publications, Museum of
Natural History, began with number 1 in 1946. Longer research papers are pub-
lished in that series. Monographs of the Museum of Natural History were initiated
in 1970. All manuscripts are subjected to critical review by intra- and extramural
specialists; final acceptance is at the discretion of the Director.

Institutional libraries interested in exchanging publications may obtain the Occa-
sional Papers and Miscellaneous Publications by addressing the Exchange Librarian,
University of Kansas Library, Lawrence, Kansas 66045. Individuals may purchase
separate numbers of all series. Prices for all publications of the Museum may be
obtained from the Publications Secretary, Museum of Natural History, University of
Kansas, Lawrence, Kansas 66045.

HARVARD UNIVERSITY
%

Library of the

Museum of

Comparative Zoology

The University of Kansas
Museum of Natural History

Miscellaneous Publication No. 74
June 1, 1983

Relationships of Pocket Gophers of the Genus Qeomys
from the Central and Northern Great Plains

By
Lawrence R. Heaney

Museum of Natural History and Department of Systematics and Ecology

The University of Kansas
Lawrence, Kansas 66045 U.S.A.

Present address: Museum of Zoology and Division of Biological Sciences

University of Michigan
Ann Arbor, Michigan 48109 U.S.A.

AND

Robert M. Timm

Bell Museum of Natural History

University of Minnesota

Minneapolis, Minnesota 55455 U.S.A.

Present address: Division of Mammals

Field Museum of Natural History

Chicago, Illinois 60605 U.S.A.

The University of Kansas

Lawrence

1983

i/L

<f?3

University of Kansas Publications, Museum of Natural History

Editor: Robert M. Mengel
Managing Editor: Joseph T. Collins

Miscellaneous Publication No. 74

pp. 1-59; 19 figures; 8 tables

Published June 1, 19S3

83

Museum of Natural History

The University of Kansas

Lawrence, Kansas 66045

U.S.A.

Printed by

University of Kansas Printing Service

Lawrence, Kansas

CONTENTS

INTRODUCTION 1

Histoiy of Taxonomic Investigations 1

ACKNOWLEDGMENTS 3

METHODS AND MATERIALS 3

SYSTEMATIC ACCOUNTS OF SPECIES AND SUBSPECIES 10

Geomys bursarius bursarius (Shaw, 1800) 10

Geomys bursarius illinoensis Komarek and Spencer, 1931 _ 17

Geomys bursarius ivisconsinensis Jackson, 1957 . .... 17

Geomys brcviceps breviceps Baird, 1S55 .... 18

Geomys breviceps sagittalis Merriam, 1895 18

Geomys lutescens lutescens Merriam, 1890 .... 19

Geomys lutescens major Davis, 1940 22

RESULTS AND DISCUSSION 24

Size Variation 24

Cluster Analysis of OTUs 26

Discriminant Function Analyses 27

Cluster Analysis of the Taxa 38

Cranial Morphology of Gophers in the Antelope County,

Nebraska Contact Zone 38

Cladistic Analysis of Cranial Characters 42

Anatomy of the Glans Penis and Baculum 45

Kaiyotypic Evidence 47

Evidence from Parasites 49

Relationships of the Extant Species of Geomys 50

Biogeographic Interpretations 52

SUMMARY ...._ 55

LITERATURE CITED ..... ..... 55

ADDENDUM 58

APPENDIX I 58

INTRODUCTION

Pocket gophers have received much
attention recently as models of the proc-
esses of genie differentiation and specia-
tion because of their low vagility and
high degree of local morphological dif-
ferentiation (e.g., Patton and Yang, 1977;
Patton and Feder, 1978, 1981; Patton
et al, 1979; Patton and Smith, 1981;
Thaeler, 1974). Evolutionary studies,
such as those utilizing biochemical data,
are based on the current taxonomy,
which traditionally has been based on
morphological studies. Unfortunately,
the only comprehensive revision of Re-
cent geomyids (Merriam, 1895), neces-
sarily relied entirely on qualitative as-
sessments of relationships of species and
geographic variation within species, and
has been outdated by numerous publica-
tions of more limited scope. Many of
the studies on which the currently ac-
cepted taxonomy is based (see Hall,
1981) were restricted to small geographic-
areas, few taxa, and, all too often, small
sample sizes.

In the course of studies of revolu-
tion of pocket gophers of the genus
Geomys and their lice (Timm, 1979), it
became apparent that a comprehensive
review of morphological variation and
taxonomy of this genus, especially the
widespread, diverse populations referred
to the species Geomys bursarius, was
much needed. We began such an in-
vestigation, and early in the study our
attention focused on the northern and
central portions of the area in which
G. bursarius occurs, partly because of
the obvious need for work, and partly
because other studies were in progress
in New Mexico, Oklahoma, and Texas
(Baker and Genoways, 1975; Honeycutt
and Schmidly, 1979; Tucker and Schmid-
ly, 1981; Bohlin and Zimmerman, 1982).
We have given special attention to the
status of formerly recognized species
that have more recently been relegated
to subspecific status. We have not dealt
with Geomys pinetis from the southeast-
ern United States, or with the G. are-

narius/G. personatus group from Texas
and adjacent areas.

History of Taxonomic Investigations

The first species of pocket gopher to
be named in the Linnaean system was
Mtts bursarius Shaw, 1800, from "the in-
terior of Canada." Mus tuza Ord, 1815,
from the pine barrens near Augusta,
Georgia, was the next species named;
this name is now rejected as a nomen
dubium (see Harper, 1952), but is an
equivalent of Geomys pinetis. In 1817
Rafinesque described several new spe-
cies of pocket gophers, proposed new
names for the previously described spe-
cies, and erected two new genera,
Geomys and Diplostoma. Because the
two species included in Diplostoma are
now known to be junior synonyms of
Mus bursarius, which Rafinesque in-
cluded in the genus Geomys, Diplostoma
is now considered to be a junior syno-
nym of Geomys. Although some contro-
versy remains over the species to which
the name Mus bursarius was applied
(see Merriam, 1895), current usage and
relative certainty about the proper allo-
cation of the name suggest that the name
bursarius should continue to be applied
as it has been since the mid-1800's.

The next valid genus to be proposed
was Thomomys Wied-Neuwied, 1839, al-
though several genera were proposed in
the interim which are now considered to
be synonyms of Geomys. Subsequent to
1839 and prior to 1895, gophers with
grooved incisors were referred to
Geomys or one of its synonyms, and
gophers with ungrooved incisors were
referred to Thomomys. Merriam (1895)
named several new genera of gophers
based on species previously assigned to
Geomys. Although these genera have
undergone taxonomic changes (Russell,
1968), the generic status of Geomys has
been stable since that time. The follow-
ing comments refer only to those taxa in-
cluded in Geomys as currently defined.

Geomys breviceps was named by

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Baird in 1855 on the basis of specimens
from Louisiana; he referred specimens
from Louisiana, Texas, and Kansas to
breviceps, whereas specimens of this
genus from the northern portions of the
Great Plains (Illinois, Kansas, Minnesota,
Missouri, Nebraska, and Wisconsin)
were assigned to G. bursarius. No other
taxa were named until Merriam began
his studies of the genus, naming G. bur-
sarius lutescens from central Nebraska
in 1890, and G. breviceps attwateri, sagit-
talis, and texensis (all from Texas) in
1895. In Merriam's (1895) revision of the
genus, he reassigned a number of taxa,
and recognized three species groups.
One of these, the tuza {—pinetis) group,
does not concern us here. The G. bur-
sarius group was monotypic as he de-
fined it; he noted a variety of features
distinguishing it from the other groups,
including an elongate and angular skull,
high sagittal crest, and palatine bones
with ascending wings broadly articulated
with the horizontal shelf of the orbito-
sphenoids. Finally, the texensis-brevi-
ceps group included five species: are-
narius, breviceps, lutescens, personatus,
and texensis. He considered all five to be
closely related, and probably much like
the stock that gave rise to the G. tuza
(=pinetis) and G. bursarius groups.
Merriam considered the species G. lu-
tescens, occurring from Oklahoma north
to South Dakota and Wyoming, to be
the closest relative of G. breviceps. This
classification was accepted for over half
a century, although between 1895 and
1947, fourteen additional taxa were de-
scribed, all as subspecies of either G.
breviceps (llanensis, brazensis, ammo-
phihis, dutcheri, ludemani, major, pratin-
colus, and terricolus), G. bursarius (illi-
noensis and majusculus) , or G. lutescens
( hylaeus, jugossicularis, levisagittalis,
and vinaceus).

Beginning in 1947 taxonomic changes
were made which involved G. breviceps,
bursarius, lutescens, and texensis. The
following discussion deals only with
these species: i.e., it excludes G. are-
narius and personatus.

Villa-R. and Hall (1947) stated that
they had evidence of intergradation be-
tween all taxa that occurred in Kansas,
and so considered them to represent a
single species, G. bursarius. At the same
time, they described a new subspecies,
G. /;. industrius. Because subspecies of
bursarius, lutescens, and breviceps as
then defined occurred in Kansas, they
implied that all taxa previously included
in these species were conspecific, and
Baker and Glass (1951) formally sy-
nonomized G. breviceps with G. bur-
sarius on the basis of putative evidence
of intergradation between the taxa G.
breviceps dutcheri and G. bursarius ma-
jor in Oklahoma.

Subsequently, three additional forms
have been described as subspecies of
G. bursarius ( wisconsinensis, missourien-
sis, and knoxjonesi) . Bussell and Jones
(1956) considered the subspecies G. b.
vinaceus to be a synonym of G. b. lu-
tescens, and Jones (1964) also synono-
mized G. /;. hylaeus and G. b. levisagit-
talis with G. b. lutescens. Lowery (1974)
considered G. /;. pratincolus to be a
synonym of G. b. dutcheri. In the latter
three cases, the small degree of differ-
ence between surrounding populations,
and the high degree of morphological
variability (both within and between
populations) were cited as the basis for
the changes.

Iloneycutt and Schmidly (1979) re-
cently investigated the relationships of
those members of the G. bursarius com-
plex found in Texas. They synonomized
brazensis, dutcheri, ludemani, pratinco-
lus, and terricolus with G. /;. sagittalis,
and ammophilus with G. b. attwateri.
They also noted the presence of three
groups of subspecies which they sus-
pected of acting as independent species;
however, they did not elevate them to
species rank. The groups were the lu-
tescens group (including lutescens, ma-
jor, knoxjonesi, llanensis, and texensis),
the attwateri group (including only att-
wateri), and the breviceps group (in-
cluding breviceps and sagittalis). Tucker
and Schmidly (1981) have subsequently

POCKET GOPHERS OF THE GENUS GEOMYS

shown that ottwateri does not intergrade
with the G. breviceps group, and recom-
mended that it be recognized as a valid
species. Bohlin and Zimmerman (1982)
detected no electrophoretic evidence of
intergradation between breviceps and
major in Oklahoma, and elevated G.
breviceps to species status.

As a consequence of our studies,
which are described in detail below, we
became convinced that the current tax-
onomy does not reflect the actual num-
ber of species in the Geomys bursarius
complex. As is demonstrated below, the
presumed evidence for intergradation be-
tween G. bursarius and G. hitcscens was
misinterpreted or in error. We feel that
the data presented by Honeycutt and
Schmidly ( 1979) are sufficient to support
the separation of G. hitcscens and G.
breviceps, and their results are strongly
supported by recent investigations by
Bohlin and Zimmerman ( 1982 ) . There-
fore, for the reasons discussed below, we
recognize four species of pocket gophers
in the Geomys bursarius species group
in the Great Plains, G. attivateri, G.
breviceps, G. bursarius, and G. lutescens.

ACKNOWLEDGMENTS

We thank the following people for
allowing us to examine specimens under
their care: D. M. Armstrong, E. C. Bir-
ney, M. A. Bogan, R. M. Case, J. R.
Choate, J. Druecker, W. H. Elder, J. P.
Farney, R. D. Fisher, E. K. Fritzell, B. P.
Glass, J. K. Greer, Jr., H. L. Gunderson,
E. B. Hazard, R. S. Hoffmann, D. F.
Hoffmeister, G. G. Musser, P. Myers,
J. L. Patton, E. J. Spicka, D. O. Straney,
D. E. Wilson, and B. A. Wunder. We
also thank the following people for their
suggestions on various drafts of this
manuscript: D. M. Armstrong, J. Bick-
ham, R. Dowler, G. E. Glass, R. S. Hoff-
mann, P. Myers, L. H. Robbins, N. A.
Slade, D. J. Schmidly, and E. O. Wiley.
Assistance from B. L. Clauson, E. B.
Hart, P. Heideman, and G. E. Nordquist
with field work, from D. Bay and M. Van
Bolt with photography and illustrations,

and from B. L. Clauson, R. S. Hoffmann,
G. Lake, and P. Myers with various as-
pects of this project, is appreciated. We
are grateful also to D. and R. Obershaw
for permission to work on their land, and
for their hospitality. Partial support for
this work was provided by a grant to
R. S. Hoffmann from the University of
Kansas General Research Fund.

METHODS AND MATERIALS

The fundamental question addressed
in this study was, how many species of
Geomys exist in the central United States
north of Texas? We approached the
problem by grouping individuals into
operational taxonomic units (OTUs),
with each OTU consisting of gophers
from a very limited, ecologically homo-
geneous area, usually one to three coun-
ties. Means from the OTUs were used
to describe geographic variation in size.
The OTUs were then subjected to prin-
cipal components and cluster analyses
based on external and cranial measure-
ments and one qualitative cranial char-
acter in order to describe geographic
patterns of similarity. We then tested
the null hypothesis of no difference be-
tween adjacent populations by a series
of discriminant function analyses. These
tests for significant differences are most
meaningful when taken in the context
of a geographically large area and many
populations; thus, we considered approx-
imately one-fourth of the OTUs simul-
taneously in each of four analyses as
described below. We included at least
20 OTUs from three states in each analy-
sis, including some OTUs also used in
preceding or following analyses, in or-
der to simplify comparison of results.

Discriminant function analysis was
used in this study in two different con-
texts, and both deserve comment. The
first use was as a means of comparing
populations, i.e., testing the null hypothe-
sis of no difference between populations.
Discriminant function analysis is de-
signed to maximize intergroup variance
and minimize intragroup variance; it is

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

most often used as a method of assigning
individuals (cases) into predetermined
groups (usually two groups). Because
discriminant function analysis is a pow-
erful discriminator, spurious results may
be obtained if the predetermined groups
are not defined correctly. For example,
two small (n less than 50) samples
drawn from a single population may be
significantly discriminated in many cases
because of random differences due to
sampling error. However, the signifi-
cance levels drop with the addition of
more individuals and/ or groups. Ex-
perimentation has shown us that addi-
tion of an "outgroup," i.e., a population
known to be different, usually will cause
spurious discrimination to drop below
the level of statistical significance, and
for classification functions to show great
overlap between spurious or artificial
groupings. The results described below
show that, when samples of pocket go-
phers drawn from broad geographic
areas are analyzed simultaneously, many
groups are not distinguishable, in spite
of the strong discriminatory power of
the algorithm. However, some groups
are easily distinguished at high signifi-
cance levels, demonstrating that the dis-
criminatory power of the algorithm does
not break down when large numbers of
groups (up to 20) are considered simul-
taneously. Thus, one can test the hy-
pothesis of distinctness of each and all
populations of a taxonomic unit. Use of
the discriminant function analysis has
the advantage of allowing multivariate
comparison of identified groups (popu-
lations) rather than individuals; this ca-
pacity is not available in other analyses.
Definition of OTUs from small, ecologi-
cally homogeneous habitats minimizes
the likelihood of mistakenly including in-
dividuals from two different taxa, but the
possibility for such an error does remain,
and can be dealt with most effectively by
careful screening of OTUs by the in-
vestigator.

The second use of discriminant func-
tion analysis in this study was as a means
of detecting intermediacy of individuals

between two previously determined pa-
rental populations. Neff and Smith (1979)
have demonstrated that known hybrid
fish are often not correctly identifiable
as hybrids using this technique; however,
they note that the presence of at least
some hybrid individuals is always de-
tectable using this method if moder-
ately large samples are available. We
have used the analysis in a way consis-
tent with their results; i.e., we have used
the method only to detect the presence
of hybridization on a broad scale, not to
identify individuals as FiS, F2S, etc., and
have been conservative in drawing con-
clusions. However, we point out that
one may not assume that hybrid mam-
mals characteristically show a tendency
to resemble one or the other parental
type simply because fish do so, espe-
cially since some limited evidence argues
against this ( see examples of mammalian
hybrids discussed by Gray, 1972). This
could be investigated either by raising
hybrids in a laboratory setting (as Neff
and Smith did), or by determining the
correlation between morphological inter-
mediacy ( such as that graphed in Fig.
11) and an independent measure of ge-
netic intermediacy, such as an allelic in-
termediacy value derived from genetic
studies.

Approximately 1,400 adult pocket go-
phers were examined. Males and fe-
males were analyzed separately; only
adults were included in analyses. Adults
were defined as those individuals which
had the basioccipital suture fused com-
pletely, and those individuals which had
cranial crests strongly developed but
which had the basioccipital suture only
partly fused. Unless stated otherwise,
all comments below refer to adult fe-
males. Our samples of females were
larger than those of males, and females
showed less growth of the cranial crests
after suture fusion; for these reasons we
feel that female gophers are in general
more appropriate for use in multivariate
statistical analyses for taxonomic pur-
poses.

Cranial measurements were taken by

POCKET GOPHERS OF THE GENUS GEOMYS

Heaney with dial calipers graduated to
¥>o mm; these measurements were taken
as defined in DeBlase and Martin (1974),
except for the following. Length and
width measurements were taken of that
portion of the frontals which projects
between the premaxillaries on the dorsal
surface of the skull; this part of the
frontal is referred to here as the "frontal
square." Orbital length was taken from
the anteriormost point in the orbit to the
most posterior point in the "orbit" (the
orbit is confluent with, and for con-
venience is here regarded as including,
the temporal fossa). "Maxillary visibil-
ity" was taken as 0.0, 0.5, or 1.0, based
on whether the sides of the rostrum were
visible over the edge of the rostrum at
the notch anterior to the zygomatic
arches on both sides (1.0), or were ob-
scured by a horizontal projection of the
premaxillary on both sides (0.0), or one
side only (0.5). External measurements
were taken from specimen labels. Bacula
were measured by Heaney to the nearest
0.01 mm using a craniometer.

We grouped 665 adult females into
80 OTUs and 258 adult males into 40
OTUs. OTUs consisted of all adults
available for study from a given county,
group of adjacent counties, or part of
a county, as defined in the following list.
OTU numbers in tables 1 and 2 and Fig.
4 refer to these county groupings: 1.
Kansas: Morton and Stanton. 2. Kansas:
Seward. 3. Kansas: Gray. 4. Kansas:
Meade. 5. Kansas: Clark. 6. Kansas:
Comanche, Edwards, and Kiowa. 7.
Kansas: Barber. 8. Kansas: Harper. 9.
Kansas: Cowley. 10. Kansas: Hamilton.
11. Kansas: Kearny. 12. Kansas: Fin-
ney. 13. Kansas: Ford. 14. Kansas:
Cheyenne and Sherman. 15. Kansas:
Rawlins and Thomas. 16. Kansas: Deca-
tur and Norton. 17. Kansas: Graham.
18. Kansas: Rooks. 19. Kansas: Greeley,
Logan, Wallace, and Wichita. 20. Kan-
sas: Trego. 21. Kansas: Ellis. 22. Colo-
rado: Adams and Morgan. 23. Colorado:
Larimer, Logan, and Weld. 24. Colo-
rado: Boulder and Douglas. 25. Wyom-
ing: Converse, Niobrara, and Weston.

26. Wyoming: Goshen, Laramie, and
Platte. 27. Nebraska: Scotts Bluff. 28.
Nebraska: Banner, Cheyenne, and Kim-
ball. 29. Nebraska: Sioux. 30. Ne-
braska: Dawes and South Dakota: Fall
River. 31. South Dakota: Bennett, Jack-
son, Todd, and Washabaugh. 32. Ne-
braska: Boyd and Keya Paha. 33. Ne-
braska: Brown, Cherry, and Rock. 34.
Nebraska: Holt. 35. Nebraska: Lincoln.
36. Nebraska: Buffalo, Custer, Dawson,
and Valley. 37. Nebraska: Dundy and
Hitchcock. 38. Nebraska: Harlan and
Kearney. 39. Nebraska: Franklin. 40.
Nebraska: Antelope ( western edge ) . 41.
Nebraska : Antelope ( from western edge
of hybrid zone described by Heaney,
1979). 42. Nebraska: Antelope (hy-
brids). 43. Kansas: Greenwood. 44.
Kansas: Mitchell. 45. Kansas: Cloud
and Republic. 46. Kansas: Riley. 47.
Kansas: Marshall. 48. Kansas: Douglas.

49. Missouri: St. Charles and St. Louis.

50. Nebraska: Butler, Gage, and Lan-
caster. 51. Nebraska: Antelope (cen-
tral, from eastern edge of hybrid zone
described by Heaney, 1979). 52. Ne-
braska: Knox and Platte. 53. Missouri:
Atchison, Buchanan, and Clay. 54. Iowa :
Des Moines and Missouri: Clark, Marion,
and Scotland. 55. Iowa: Mahaska, Mar-
shall, Monroe, and Stoiy. 56. Iowa: Clay-
ton and Dubuque. 57. Iowa: Clay and
Emmett and Minnesota: Brown. 58.
Minnesota: Goodhue, Houston, and Wi-
nona. 59. Minnesota: Ramsey and Sher-
burne. 60. Minnesota: Becker, Cass,
Kittson, Norman, and Polk. 61. South
Dakota: Brookings, Lake, and Moody
and Minnesota: Rock. 62. South Da-
kota: Grant and Minnesota: Traverse.
63. North Dakota: Richland. 64. North
Dakota: Cass, Grand Forks, La Moure,
and Trail. 65. Wisconsin: Bayfield, Bur-
nett, and Douglas. 66. Wisconsin: Craw-
ford and Richland. 67. Illinois: DeWitt,
Logan, Mason, and McLean. 68. Illinois:
LaSalle, Marshall, Tazewell, and Wood-
ford. 69. Illinois: Cass, Madison, Mason,
Morgan, and St. Clair. 70. Illinois: Kan-
kakee and Will and Indiana: Jasper,
Newton, and Tippecanoe. 71. Oklahoma:

6

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

H3
N

D
H

O

CD

'«

o

on

CD

'ca

S

o

3

a

c

>

CD

H3

c

c8

+ 1

c
a

w

ll

2>

.tr so
C J

3 *

t< iz) pa

a!

ll .13

C CJ 6C
C 3 C

fc 00 >-)

a -a

pq

*3 •£
'5 T3

o

B -c

pq

N

o to

5 5

o
U

.5 °

o c

H
O

o d d o o d c c o o o c o d c d o o d o o o o o o o o o o d o e o o c d o o

+1 +1 +1 +1 +1 +1 +1 II +1 +1 41 +1 +1 +1 +1 +1 +1 +1 II +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 II +1 +1 +| +1 +1 +|

©O©C50C0C©c0©<COr>OClC0C~.-<O^ClrtClcC''r©C00ClDCCcr;W0CC]cDCC0C'<j"c0

^^^ddd^d-^ddd^ddddddddddd^ddddddddddddd

tiniOfeoincii'HiDOoooenoofflciHosNfHhiciiooi-iOi-inH

00 10 -f t> G5 CO

>— i ID 00 t Tj" cl

OOOOOOOOO-hOCOCOOOOOOOOOOOOOOCCOCOOCCOO

+1 +1 +1 +1 +1 +1 41 +1 +1 +1 +1 +1 +1 41 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 II +1 +1 +| +1 +1 +1

oiiisniflioiciooncifieicofflichoifiOHOciioiMOKOecoeKffiooo

NMMCCONO-l'Cl^ffiCNCrHlOlCnclclffiOOr^clOOMl^clO'-Clror-.tOt-

^^TT^Tr^Tf«lO^^^TyTf^^TfTt(Tt<-^ICO-<tM,-*Tt<TfT}ICOTjlrJlT}<lO^>Tf<T}<T}<Tl<CO

0)N,<il'JooHoociO)ffleeowiloociocciMCi"i|NNin-Hco*oico;0'HO»
c-i^^cocococicoco^^^cocorfco^rcooi^H^^oicicico^iocicocico'rioc'icooicccO'— < id

dddddddddddddddddddddddddddddddcdddddd
+1 +1 41 +1 +1 +1 +1 +1 II +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 II +1 I! +1 +! +1 +1 +1 II +1 +1

hKCrtOHloOrtcOOrHICOHHOcOOI'MCOlCOOHOHXC'rOCrHCllO

^wt-t-oiBioffiOiceefooffiOococoooioooffflffiooNffihooocffihfflffloa)

MIMiHHOCIMOi"tl(eWNO>rtCDIflini-iHOS«!!*lMMOfflOO(cm!00(Dr-tOW050l»0

Tr'^'*<coto-<fcDco'f'Occe)0'^TrTrM,"*'T,-^cO'^'-H'*io-^;OTrW'^"^cc-t,>o'oe'i-'i"

dododdodddoddodoodoodoodddocidddoddddoo
+1 +1 II +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 I! +1 +1 +1 +1 +1 +1 II +1 +1 +1 +1 +1 +1 +1 +1 +1 II II +1 +1 I! +1
BOoooMOciinooocofiioonwhooni-iinciscmoHeiooEiMMOtiMci

OKclHOOOhn^qlOOHHOt-OclrtClOOtCNCClOTflMffiOO^SDCCOOlO^

^^^^cdracdcd^^cocd^cdcdcicdcdcocdcicococdcdcd

N'*1'(MffiOn(MnNfflO(CiniiHrtCC(M10(M1,a!01tC-1,Ocnc3^0PffiO(;)0-lH
Cl^^CICOlOcOM,lDClCOClcOTrlOC01D^^lO^COCOClcOCOTt<Tt<T:t<^<clcO.--^rlDCOlDt^

ddd-ddddddddddddddddddddddddddddddddddd
II +! +1 +1 +1 +1 +1 +1 +1 +1 II II II II II II II II II II II II II II II II II II II II II II II II II II II II
o"5i'Oio;£(Minoo!£M'Tfr<cf.cotoein«c^ocrtO-tacipO'-cct'(coct
fflO-^M'irrtH(C(MOOOrtiO'!rcoi"trtioiorocociioi<(Mrt(ocoioe;(£^coTiMo

r— 1 i— I

^lOHOoiM^ffhHoofflioorticoocii-iin^ob-MtBoinfflOiC'-ioocm'Hcoo)')1*)

00Cll>l>C-lCla>-^"C]<C^Ci-it^SDcOQCCI)-*O0-Ht^CDCO0OO5CDTt,CCC)00Clt^CC'1'C'100t~c75

dddd~*-:<dd**ddr-iddddddd--:<dddddddddd~ddd'-*ddd
II +1 II II II II II II II II II II II II II II II II II II II II II II II II II II li II II l! II II II II II II

cnCOccoci^OclOcncOOOCDCOClHlOtDCDt--*1,OG!fMOOO'tOP100^'i,^O^ClCD

owmfHiniooiboiBooffiHOhbeiHi-oioioffiffleciOfincoffiiCHiMoosi
id^^^vd^^c^^^^co^^^c^^^cococo^cococo^Tpc^id^^ioc^^^^cooi

Cl OI cl cl cl Cl Cl ci Cl Cl Cl Cl Cl Cl Cl ci Cl Cl c'l ci ci Cl ci Cl Cl Cl Cl ci Cl Cl Cl Cl Cl Cl Cl Cl ci Cl

■*oxotmHtDWnrtin(Moc;oiocici(cio»a!fnociioi'CG5a'M^cioc?G)
cccofflt-rtOhiMioofffloooorHM-oooMoot-t^nt-aMnoTtafccot-cio

o o o o

^^0^1— 1 i— i O O O >-< ©

—iooo— <t-<ooo©

,-H

o o o o

i-H C O O ^h o

IH O

ii mi mi

mi ii mi ii ii ii ii ii

II

ii mi mi

till

CO CO t~- — <

CClOncChOOOclOCO

rHGMncDOOO-iftB

ID

ci -r oo re

ffOOG!O03O

o ci co oo

lO-rcicit^iocccO'-'ioo

rHlOKKTOClf-KlO

oo

GSOCI t

i— 1 Tf 00 Cl 00 CO

00 CO

IDCClolDCOCOCOCOCCcOCO^CCCOincct^CO-^COCDCOCOlDlOCClOlDr^CCt-^t^lOr^lOCCin-^
Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl

0!Cl-ft(DHrH(MNr-tl<1000C0^(0<DrHl^00(0N©a)OO-*C5OtOC0!C--i"tCCCO
(MKCrt01t-'rclbOfO!t»1,CCC10)KOC]OOiO!(Cl'5tQOcOHO(CM001tC(Mrt

nnnnnnnnnnnnnnnnnnnnnnn^

tDOClOcOO)HinOIClO,fCOW.I>^(MTl<cnOOCOOJtDOiQOa^COhOiCCCOCCtOhcl

00«lCOO<MrHrHH^tH^^C1050COOC]QOOCM0000010ClC005COCOHOCON(MlC^

h h h h ci ci ci ci ci ci ci © ci ci ■— t «— i co co © ci i— i ci i— t ^h i— i ci i— i © ci ■— » co co <— < co »—< ci i— < <j>

""^ ^ ^ ^* ^ ^ ^t1 t}* ^ "^ ^ "^ ^ ^ "^f ^ "^ "^ ^ ^J1 ^ ^f ^ "^ ^ ^T M1 "^ ^ ^ "^ "^ "^ "^ "^ ""^ *T CO

l0^lOCOOO^MOCDOOJt>C105lOCOOOOCOlOHN10lOfOOrHl010NCia5CD05005»0

i-Hi-Hi-Hi—(Oci»-HOOi—li-HOC-lOi-HC0'—l'— '•—''— t'-H.-Hi— I Cl H O O r- If— Ii— ( i— (.— «0>— if— I Cl H r- 1

II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II
•tiOiHcicoOfflot^coioOffrH^^rtittofOl-cDiOccciiOftCffii'^'HCiGsincc

dfflHHdHoiQOoiddddfliciHHdfflfflHdoiHHHdHdciddHdoidffl

COClCOCOCOCOClClCOClCOCOCOCOClCOCOCOCOClCICOCOClCOCOCOCOCOCOCOCOCOCOCOClCOCl

HHCiMt.otoom'rooioiMoioioi'cinciHnooonHoocDiocHmoiniMO)
diDdio^iDcc^'c]^deiocd^dccidc»d^iD^diDr^^dd»Dt^^i^

OOOOOOO-hOOOOOOOi-iOOOOOOOOOOOOOOOCCOOCOO

II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II II

C0C0ro-1,t~^t,C000t^lDOC0-^,lDT)<05'-;t^ OC-CltDt^ClcOlDCltDOOCDOOCO'1-OCClcOOOCD

ndoo66rt(cocrtrHooN^i>ldoo6rH6dffi^dw6ip6mrHr^cirtiriTfoS(c

t-l>t~t^t-l>t-CD«3t-t^CDvDI>COt>t^l>t^t^l>t-CD(Dt-t>t-CDt^tOI>t-CCt^lO(Cffl>D

COCC10CTlt~05^1D10t~^CDt>03CDCl'^'-;l^Clc01DlDCllDCllDOOOC»a>010COi-l->^03

d n h d io ci i< h n o d ^ d ci b ih a h io ^ d k ci h oS d -t rt ci d m" o <c ^ ci n r-i ci

^hO»— i<— iO^h^cIO^h^hO-— '^O'-^OOi-HOrHO'— 'OO*— iO^^t— i^h^h^hO^^^hCI «-h»— i

II II II II II II II II II II II II II II II II II II II II II II II II II II II II || II II II II II II II II II
Ot»00ClcHCOOOOOOOO00t-(0OtfOOOCiOMC0O0)(DlP(0POIC^QO00

d'Dcod^dr-dcccid^i-^cdood^t^cd^'^'i^t^^coiDdiDcid^ccr^cDcit^c^

COCD'^CO-^CICD'^CO'^'^OO'^CO-^'^'^Of'^-^COCOtt'^COlDCO'^tcOtOClcD"
Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl ci ci Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl

t^^CScnOOOCl-^CO^QOlDlO-^Ot^ClOt^Olr-icOOOlDCOtDlDOOlDlOlDOOlDCDOOl^CClD
Cl i— l>-liHi-lCl(Nf-lr-l CO i— I

HCicoiiw(DNceo>Ortcin^in(ot^coc50Hcin-*intDNocaio- icico-piDcct^oo

rt H H H H H rt H rt H Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl o o n o o o o o o

POCKET GOPHERS OF THE GENUS GEOMYS

T»Tt»iioTficoioowin'fio^WTiioiooonoooM»inio^oqioO'*ii!ioqiniqtqino
ooooooooooooooo'oooo'oooooooocooo'ooooooooooo

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

cocicccoxt^ocoo^t^rat^CTXxocoooOTOOocicococoxoor-oxin-rox-tfincqci

OOOOOOi^O^OOOOOOo^OrtrtO^rti^OOOOOi-ii— iO>-iOOO!—iOOC>00

HfflrMin(MOniMOn-<0000'*COHOffl(^OOOWNOiOOMni<OtOH(0 01^0«(MO)«
00000000000 0 00©000'-<0'-<00>-<OOOOOCOOOOC©000000

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 41 +! +1 +1 +1 +1 +1 +1 +1 +1 +1
0(riPtoM'1^ocococ:'^no1,'cc^cMei,(£iMOfc]oi'bceomcfflxoi'iioi'w<»o
t-fiHOTvEoiociiociHcicifflt-ciincoiHot^ofi^MccoHOOM^xnTfooinofflinc

COONnM't-IP'tOOHlCcOnWinOinOOlONCBMOt-XftSNOfflH-ticmNOHXOO

n o t ^ji a ^ o o rn co i1 o ci o c^ (D ci o o ci o o ci ci o o ci o el h en ci ei ci (M ti1 o t o co ^
o^ddooooc'ooddo'oooo^ooddoddoooodooooooooooo
+ 1 II +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 + 1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

CIMCCCOCCIO^X1,CieX»C10XOCGlHCcOC|10(OBCOClc01,COOcOOfXClXCOW

wo^ioeohxoot^ooiHOoocKiOiHMfflcoMrtOfflh;Oqcic]inf-iniEinqffl-*'i;c]
,_i ,-5 ,_5 ^ ,_; —5 -5 i-i i-i 1-5 r-i ci ci ci w ih ci ci i-h ci i-i i-5 ci ci ci i— i i-i ci ci ci i— i ci i-l ■— > i— < i-h 1— 1 i-5 •— < i— i ^^ ■ — i

(COClffitDlOlOCBTfrtHNCOt-fflcOHt-OCOOXCOfflHOfflffl'fCHOOHCSM'hXOCOOX
CO-fTfCOiCrfCOrfMt^mcOCrHClClrHflOCI-fGJ'f'^tDt-ClClM'M^I'^'t^inciroqclrtlO^

oddcdodddoodddHodddcioddooododdddooodoHdHoo

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
coMcciKc^ocicrciox^iointiocor-iONHbcoocootociciioooffixaciocoioon

TfOiEffiOCEClCbr-t-CCI'CffiClCIOOXO'l'XXTlDrHininxWCfl'OlOlOffirHOb.

cd^cdcdwdTrdcdddcdddcddT}<cdcddd^^dd^d«id

ClCOt-r-,t"*l'1a!£OXlf"tXClc»OXOXCOOClC)^lCS1<0>lO'^COc0100rtlOlPNtCr-0>M
b- CO CO CO CI i-H Cl -T- CI CO CI i? "t i-i CI W i— I i—l i-H t CO T* CO CO CO CO CO i-l CI CI CI CO CI CI Tji i-h CO lO CO 10 ci l CI

dddddddddddddddddddddddddddddddddddddddddd

+1 +1 +1 +1 +1 +1 +1 -H +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 -H 41
cocctci^^bcicooioi^coc^qco^oxcioC'r-rioxN^'Hccxr-iocobcococccocioioo
o w i-i id ci io c x r~ cc cc ci d ci d t> o i-h x ^r cc cc c o i-h o i-< iq x o r- — -h ci ■>* o i-h © ci | cc ■<* ; i-
dddffldddddcdddddffldddoaffldddQoJfflcjffiddcfflffloiffltiifflwwx

l— t l-H i-H l-H l— I I— ll-H l-H

TfrHNWC0t£MNXCCr-KHi)iMN^,O'-ic0OI/>C0Hr-Offi01WII0ffi^Cl-1iCltrHNC0HC0'^1

0!r-xt-wwc!CNiHcicit»ooffli-ii!oo>Hcoo)Hfflowi-i«)r-i»MnMO!t~^iooeoo)NNin

OOOOOCOCO-h

-h-hOOO^hOO —

O -h o O O -h

OOOOi-hOOOi-hOOCIOOOO

+1 +l+l+l+l+i

+1+1 +I+!

+1+1 +1 +1 +1+1 +1+1 +1+1

+1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1 +1+1

CICfCCclCSlPCHE'tClOOXtECOCN

co

loot^OTrocc-txoomoor^oxoTin-fio

cc ci Tf t~- ^ -*

H l- O CI

mrot^xracit^mx

CD

t> t~ I> ■* Cl CO

lnx-tit-o-Hcit-xcccoxococox

ri^co^iot^t-<ct^cciot^t^iocciooococcffl>ciot^t~cDcO'),int^t^t--t^ic-*'-i<Tt'^<Tfioc]cici

Cl Cl Cl CI C! CI CI CI CI CI CI CI CI Cl CI CI CI CI CI C] CI CI CI CI CI C] CI CI CI CI CI CI Cl CI CI CI C] CI CI ci CI CI

-ifH/lOffroinclMDtQWCCMHOOt^^OinfflcOOlOrtfflCltCCClOCOlOlOCit-tt'-'Cl
K1000«OtCffll--(EOO)C110ifN(DHiolOO)0>tCC31,K01NOOCOOC!OMhq(0,*Ciqq

o'drti-<ddddddi-5di-<^i-5i-Hdi-5i-Ii--Jddddddddcii--5i-5i-<di-5i-idde-ii-ii--<dd

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +i +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
t-o^o)HCHcii--oooi,occoccoooc3fc]icccoNcic]NC'Hh-rci«oiocniion

C£lO^Xt~i-HO-Hl-cOt~XOCO^CM^COt^^OlOC^COCCCO-^COt^ClXOClCli-HCOXO©t~lOC>

^diodt^dddxcc'cc'ddt^docddc«cco6t>ddocccdc»ddccdddddindt--''i<-*''*

Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl CO Cl Cl Cl Cl CO Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl CO Cl CO Cl Cl Cl Cl Cl Cl Cl Cl cl cl

OOXl-tDWffirtCeclinfflffiCl'*10(Offfib(DCOCOO>"tMCOCOCDWCCDM10COOC11iMClC01<

xoo«fl>ooo5co'*iHijcKOciffli,H(c^^xiDt-io-ir)rHOTi<cociht~aoot--^qi;q

pHl-il-iddl— I H O H H H H r< O C1 Ci O H ci H 1H l- id-HO-Hl-Hl-Hi-H^HCl-H-HCciddl-Hl-Hl-H-HO

+1 +1 +1 +1 -II +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 fl +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

COC10iCOOHOC100-f^mOOl~ncO!OOOfc01000NHi(c01-CO(OCOC(E(CCOCOO(00

docco-Hr^t^cst^^cciocc^ccioccffiOt^coTii-Hinco-HciaJoocicDCDt^xcqcit^i-Hcoiq
dcii-ico'iONNcc^ffiffidi^ox'iocdaiichdddNx^Nceco^aJHcici

ciffl»ffirooctccoNCNcoT(icocicD'*r-r-<oW'*ino)Ncoqcoi^ci(DccHqqb;qqqqq

ciHOdHHHiHciciHHHHHciOHHHHHHHOHHHHCOHOdHHHHlOHHClH
+ 1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

WKC!0)CJ!NXXNCCCOCBCOOKXa!i-iOXCOCOClW1«cCiHCOr-OCOiH|^OMnXCCCO'fKWq

docdd^cidcicocdi--<^^'^^cicdi--'-^cidcdididid^-Hcdcicicicid'Hdddddt~iocD
ciciclcococococococococococococococococococococococococococococococococicocicoclcici

inooi-<oocoi-iOi-HOOooioina:t~coc;ococ7)(Ccoco-*iOi-Ht^-q<iocicoocot~iot^cii^Oi-iio-^>o

O) <£ X 10 K! CO r-

r^lOC'^O'f^CX'fOlC

IOObt-tCMOh1'C(OOlNM(CXt--inilHX-flO(0

OOOOOOOOCCOOOO — OO-h

-h—OOOCOOOOOCCOCOOOO-hOOCC

+1+1+1 +1 +1 +1+1

+1+1 +1+1 +1+1 +1+1 +1+1 +1+1

M CO 00 H o t- t-

t-CCO-THOlOOCIOC

COCCOOtOCOfOClcOtOCOOXI^hXXOCOM'THCO

0)*ftti-'rocoiociinxio-ooiocoi^ccnocit>intoHiHocoi'X05HCico(Dci"HiH

Wt>l~t^l-t~X!>Xt~t~[>t^l>Xt~Xl^t>COCOt~r-t>[>t~t>XXt>»XCOCDt~CCt>CDXCDCOt~

HCDiH^N05NciincorHoowci"fciHxiomcifficocDMHij<Ht^t|cooo)tccii)'oqnHN
x ci i-5 1~ ^h' iri cc in d i-5 —5 1-5 d d ih co id d cc t^ cd t--^ d id ■— i in d co co id d ci r-i d -t ci x' ci co ci d x'

-h-h — o-hCCCi-h-h-h-h--.C-h-hOOC1C1i-hc1C;C-h^hOi-h-h^hCi-h-hC'-hOO — 1-hi-hOO

+1 +1 +1 +1 +l +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +1
incicinxcixiniooi£in-cococioxox-HMcino»t^ocicocoi,t~cocicooxqi;qci

^dxdcoo6cidcoi-HT^^^^d-t,did^t-cid^^xdddidt^dt^xdidciTi<indid-Ht^

cl-1,CO-f(E(CCDCDh(ElCt-t^COf-inXW(E(CCOTrbCCCt-'1'lD(CCDt^l^cO-1,COCOCO(Of'HrtH
Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl

CDOX'^Ot00HDCD'*t-HC0rJi^iinc0WC01,'*t-H^il0t00!HCl(»C0»C00)<0'<),lfi'1,ln'*C0H

i-H i-H i— < ,-HCl 1— 1 I— 1 rHCM I—1

ffiO^HClCO^iniONXffiOrtClCOflOfflNXOlOHClCOfWtDt-MffiO-HClCOIiWlCNMOiO

co^^H-^<'^"-q<'^<-q''^'^'Tt,io»nioi'3inioinioioiotocDcccccoco«3cococct>t^t~-t~t^t^t^t~t-t^x

O IT

i-H C5

5.

>>

P

V

nt

U

<0

a

-

-

V

a
o

<0

a

N

N

13

- 1

>

1-

-C

-c

A

o

ai

nr

-r

CJ)
13

V

E

E

is

a>

s e

o

n

fc&H

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Cimarron and Texas. 72. Oklahoma:
Beaver and Harper. 73. Oklahoma: Al-
falfa and Woods. 74. Oklahoma: Major
and Woodward. 75. Oklahoma: Custer,
Dewey, and Ellis. 76. Oklahoma: Beck-
ham, Caddo, and Washita. 77. Okla-
homa: Harmon, Jackson, and Tillman.
78. Oklahoma: Adair, Mcintosh, Musko-
gee, and Okfuskee. 79. Oklahoma: Atoka,
Choctaw, Coal, and Pittsburg. 80. Okla-
homa: Bryan and Marshall. 81. Kansas:
Atchison and Leavenworth.

Multivariate analyses were conducted
using programs in the BMDP series
(Dixon, 1975) and MIDAS, the Univer-
sity of Michigan data analysis system;
those used were the cluster ( MIDAS
CLUSTER); principal components anal-
ysis (MIDAS PRINCOM); and step-
wise discriminant function analysis
(BMDP7M). Cluster analysis was con-
ducted using data generated by the prin-
cipal components analysis. Data entered
for each OTU were the scores on the
first seven axes. Because the subsequent
axes were not statistically significant, and
accounted for less than 2% of the total
variation, they would not have added
meaningful information to the cluster
analysis. This method limits the impor-
tance of size to 1/n of the total "charac-
ters," where n = number of axes. Use of
unmodified data can allow size to play
an excessive role in clustering. The clus-
ter method used was an unweighted
pair-group sum of squares utilizing the
variance-covariance matrix. The cluster
analysis of OTLTs was used as a basis for
detecting geographic patterns of simi-
larity (Fig. 4). To construct this figure,
the cluster phenogram (which had amal-
gamation distances of 0.412 to 36.361)
was examined for identifiable geographic
groupings at amalgamation levels of 5.0,
10.0, 15.0, 20.0, 25.0, and 30.0. Below
the 10.0 level most groupings did not in-
volve geographic neighbors, but at 10.0
a geographic pattern was evident; this
is indicated on Fig. 4 by the inner ( thin-
nest) line. Level 15.0 indicated further
major groupings, and is shown as the
middle line. Level 20.0 is the outer

(heaviest) line in Fig. 4. Level 25.0
simply grouped OTLT 66 to its neighbors,
and so is not shown.

Discriminant function analyses used
a tolerance level of .01; an F-to-enter of
1.0 rather than 4.0 was used as a stopping
criterion because at 4.0 only one or two
variables entered the model, and these
primarily reflected size. An F-to-enter
of 1.0 should maximize the ability to
distinguish groups, and is thus consis-
tent with our use of this analysis to test
the null hypothesis of no difference be-
tween groups. Discriminant function
analyses were conducted on several
levels. An initial analysis was done on
all OTUs having five or more specimens.
This was done in four parts because of
program and computer limitations; the
four geographic units are defined in de-
tail below. The second level of discrimi-
nant analysis consisted of lumping all
OTLTs which could not be significantly
distinguished from one another ( as indi-
cated by non-significant F levels and
jackknife classification levels of less than
90%). OTUs with samples smaller than
five, and all specimens which had been
noted in previous studies as potential
intergrades, were entered as "unknowns."
These unknowns were then assigned to
taxa based on examination of posterior
probabilities and plots of discriminant
scores. The final level of analysis con-
sisted of grouping together all individu-
als of each taxon, in order to ascertain
the characters which are useful in identi-
fying the taxa. The means of these taxa
were used in producing a final phe-
nogram of taxa (Fig. 10) using scores
from a principal components analysis,
as discussed above.

Specimens examined were housed in
the following institutions:

American Museum of Natural History

(AMNII)
Bemidji State University (BSU)
Chadron State College (CSC)
Colorado State University (CSU)
University of Colorado Museum (CU)

POCKET GOPHERS OF THE GENUS GEOMYS

9

CO

c

CD

O

CD

CD

•—
O

u
O

s-.
3
O

CD

c

3

T3
CD
N

H

O

-s

CD

05

—

O

OS

e

3

oj

c

'■4-I
o5

>

T3

o5

4-*

in

+1

c

C«

CD

CM

W

S>

.-cr w

X c

>- Co

O J

" CO -£

[zh "3 pa

c 2 «

n -j l-

" c? n

pa

""2 •£

'3 ■"5

2pa

o

r «
O eg

Mi

>.P3
N

C be

c J
o
U

c o

4*

H

O

t^ooioooiooiooinicioioiniof'tiowi'^oioooooiooiOTftotin^o
odd dodo o'ooooodoodoodooooooooooooooooooo

+1 II +1 +1 41 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 II +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
oifflooooocioinooinicwtnooit-cifflcooooooo^oiflfflooocococoo
doHH'dbdddoddddddddodddoHdHHHHdHoddHdcidrt

Cl

in

CI O rt O U) G! H

N C5 CI IP ci in c

CO 1— 1
Cl t~-

o>cct--cDincDo:ciciT}<i>ciTt<ococj>inciinoco-*co—iooM,Tt<c'ico

tcioMOfflcCTfr-t^wocoortXccinciooorf^^nciomcc

OOOOCCO— '

+1+1 +1+1 41 +1+1 +1

<M O OD f O t^ Oi Cl
-n O (C - t- C-l CO CI

00000000000001-1— I rH ,-< © O O — OO^OOOO^nOO

+1 +1 +1 +1 +1 +1 +! +! +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1
ciioctintccifficcftDoi^onofficot — ft-fOOTfo^noinco

-Hl0MnalCOClC0fO10CPClrHC0OC10DCI)'*ClOHrt'!)iO0110C0O

CO

cc in cc m in

in co in in

mininincocomincoooocoot^t~ooo3t~tr-t-i>oooooocococoin'^,i>

CKOClffiOWCCCBCOCIPCCWlCHTfbCOONlOlnOHinifWfflClOOOOrtOclCOClCOW

■f ci ^f ci ci - icicicici^^tcnint^in^Hcoin^coinTit-^inp-icociTfTi'Tt'incicoincoci-^ci

ddddddddddddddddddddddddddddddddddddddd
+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
HrtfciomoiooccicoOHCiHoi'roioMEOoiootcfficco^oninecicohci
cotceino!rtac!cit~o)XNffiNKorooooD-!DOciiOrHowcoraoocDocDoBcD

H H H H H Cl" H H CJ H H H H H H H Cl H Cl CJ Cl H H CJ Cl Cl* Cl CJ H C] H Cl Cl H H H H H H

fOOfflOONONXOCOOtNOfClHNcTlCiCO

NtiDMOcici^ffiffi-toomincMOincjooort

t~OClCDCD^CD00c000t>C000Tt<ini>>-l
fbOCOcl^OncOMOOOhOCin

OOOOOOOOOOO— lOOOOOOOOC— lOOOOOOOOOOO— ii-HOO— I— I

+1 +1 +1 +1 +! +1 +1 +1 +1 +i +1 +1 +1 +1 +1 +! +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
^HTt-ciccinoo>cicit--t~ocit~cir-ooTi<-i"i>0'-HOOi"oot^i^c-5cooin^ooooooc-50
of^ococMncoxcoofOCciTfOBWoinTfinTfcoNosciTffflcioocirtcicoincc

^^o^cociocoon-iiooTf^'t'i,n'*Ti<Tf-ii'ii'*Tti^Trio^incD(ocoio^i'j,'iin'^

QClBHO)CDHClflOn-!fCllf!t^nc71C71CDinHCOcOClCDel»1'aOOCC'1lCDClr-NffiO

^ciciTrcir-coccinLoincoco-tininioinr-^oococcc'icirrcionrfinncjnnTiiinio
dddddddddddddddddddd^dddddddddddddddddd

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
— ocooot^oinr^mcc>i'Cot-oooocct>ini"ciccooco-HincioccO'-HcicioooooT»>
cffi«t-0!ciHci«ci-rrnoincciotci-Htccfflincfflciinc:noNO>oociOfficow
daiddidd^iddddddddddd^'dddd^id^dddddd'-^dddaiaid

en co

inoinciocici^^ro^ococo^'^cico'*05CD05|x>c7>in'i<c3>'tcDcnincnt^'i,ini--iin
«inoTi'T'roocoTrt-M,'tininM,oociO)t^cicocciccicoNxin-Hco(ioeit-niHio

o o

+1 +1

in t-

^h in

■-iOOO'H'— ii— iOi— iO>— Ii-hOi-iOi— i>— i^nOi— i(Mi— iOOi— iOi— lOOClOOi-Hi-it-Hi— li— l

+! +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
cccDOt-iioociioxcicicicciocucocicoiPWcccOrtHnrtccWTfocixcociO
conOH^Giorocih^offlcioooooooiHointcipicocoffiHincoTtihococow

oc t> t- t~

Cl Cl Cl Cl

ffitCNCSO(Ot-(CCDNhNt-(DCCOOClCBHrtpffiHCC^HOOCOCO
Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl CO CO CO Cl CO CO CO C] CO Cl Cl en CO CO Cl Cl

CO t> 10 o
ci cl cl en

r-r-ccffiinr-clclr-i-icicCcBciiCDcotfcocor-Ot-iOCBOincirHcomNO-iNclocO^i

Ttiro^^rHfflOrHincqiocoHTfMcpocinot-i^i^cerot-t^oioor-ioocccTJnocococB

HOHHHOC'WHOHdHrjHISHHHOdHrieiddHONHHHOciHNHO'd

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +1 +1
■h ^o^ciinc^omciohOot-MiHoconhccciciTcioncMOfflOciciioiioo
HcEoxinccrtcoob-in'toHfflociincociincoH'TrtOO'i't-coaooHcs^i-^co

dddfflcBGi^oi'- i^dddd—idddciTt"in'iricdinind'fdcociiriTi<di-i— iddoi't
cocDcoclclcicocicocococociocoeocoirococococoeocococococococococococococococico

C10>Clffl'<OC0t!COfflXCDc0rtOCltEClC0MClt-M0roOclCDW1icCclr-C0TtiOC0'*
CIMTfoCXCCWClrlCCHCCOhOOlOOOCBOClcONTt'HOXCllOCOfflCOCCrot-lOfflin

h h h 6 6 rt rt ci h d ci c ci -h rt ci rf d rt d ci ci t' o d ci ci d ci rt w h h rt ih rn ci h o

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
^cociocir~-Hincit^-H©oininot~cii,t~i>,i,cDcoi'co^i,inciocoinot~cimooinoo

ClObCOCCOrtifOCOclO^COffiNHCCO^ONOHClHt-clCDOMnm^O^IOincl

c»xdt^ddcr't^t^<»^t^ddoo^^dd^^incot^'M'd^t^coodindi>w

^-*Tt,-i"'^'^r^f'^,'^,'^,"*'^,'^''<i>Tj(Ti<TfT}iT}iinininininininininioininininTi,-^iM''^,TJ,io

cicocsociNcooccicoociwoooNo»cD^HcoojocinoNX«cinoc»cDionci

H H O H Cl CO O H Cl' I"! ^ H H Cl' H CO H H H CO M H H C] rH C) Cl I— I Cl" ^H ^H — H ^H Cl" © ^* Cl" ^ CO

+! +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +! +1 +1 +1 +!
cirHTfOr-iOrtciocicoocixciNwcicinipcor-coMcoHciXHinotococoNinxo

■* ci i ci co i ci ci t' ci "3* co Tji co i ■q' co co ci ci •— i co in cc d co 06 co d > in oo -^ in -ji in in co liridoito
cocococoococococoOcoococoeocoocoococococoeoco^cocococococococococoeocico

ccmcicioxoincDccoHcocicociciciHHCccocitoiinot^xciT|ixo)'*'),Ttioni^

d^iPoxw'ndiHffiffir-cid^^d^dH'i6d'ii-HNddT)i^rH^c<Jrt^ddN^x'

OOOOOOOOi-hOOOOOOO— iOO-h©— iO-h©00©0-hOO~0.-hO©©0

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

OH^xt-ooinciciBTfxi'Otccioot-t-coxciX'HrtOnxxtcinciNxcio
— <^dcdcidininccciincit-ddinino6'Heocid^Ti<— ' ^j" id -ci" Tt< d ^ co ci cd ci oo in ci d

xxocifHcsxococBincortOiniiHxxioocioin^iooHtrtOHcoxxiixcD
co d — i d ci ci d in ci oo' ci d d i-5 d d d od "*. "* o in ci t^ t~ oo d d co — < d ci d co •* d i-" °d co

i— ii— ii— irHi-Hi-iOi— iCIOi— iOOi-hO^hOOOCIi— <i— iCli— ii-hOi— iOC1C10<M(Mi-ii-iOi— lOCl

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

Hqcixi-ohocit-rtTfTfiocixcoCTfcor-ocinxinNinoxinocoqqcocoo(D
drf'r^dciTfininddd—idi^ccciix5dd^d^^dcido6c)dci^.-<indcd^o6

CCCDincDt^COCDCDr-CONCOCOCDr-COCDWr-ClOOXcliHClXOXOOOiHCDr-COCOlOM
Cl Cl CM Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl Cl CO Cl CO Cl CO CO CO Cl CO Cl Cl CO CO CO Cl Cl Cl Cl <N CO

cot^ioiOTt|Tj<coTf^cooinino^<Bcoincoi»Trt^^i-iin^cofli-iiBco^T^oocoioco-^'COin

I— I I— I I— I I— I I— I

tcDt-o^cor-oi-icoiocoxortcoinxococcxciOr-iinojHinccNOJOiH
r- i— i— ii-ii-icicicicicicococococoTr-^-*TiiTj'mininiocococococoi>l^

^ ci co in oo —l

l> t- 1> l> t~ 1> CO

05

CO

c

C3
CO

X

CJ)
CO

s

o

-

10

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Field Museum of Natural History
(FMNH)

Kearney State College (KSC)

Museum of Natural History, University
of Kansas (KU; Vertebrate Paleon-
tology, KUVP)

Museum of the High Plains, Fort Hays
State University (MHP)

Bell Museum of Natural History, Uni-
versity of Minnesota (MMNH)

Museum of Natural History, Michigan
State University (MSU)

Museum of Vertebrate Zoology, Univer-
sity of California (MVZ) '

University of Nebraska State Museum

(NSM)
Oklahoma State University (OSU)

Stovall Museum, University of Oklahoma

(SM)

Museum of Natural Histoiy, University
of Illinois (UIMNH)

University of Missouri at Columbia

(UMC)

Museum of Zoology, University of Michi-
gan (UMMZ)

United States National Museum of Nat-
ural Histoiy (USNM)

SYSTEMATIC ACCOUNTS OF SPECIES AND SUBSPECIES

Geomys bursarius (Shaw, 1800)
Plains Pocket Gopher

Geomys bursarius bursarius (Shaw, 1800)

1800. Mus bursarius Shaw, Trans. Linn. Soc.

London 5:227. Type locality Elk River,

Sherburne County, Minnesota (originally

cited as "the interior of Canada"; fixed by

Swenk, 1939).
1817. Diplostoma fusca Rafinesque, Amer.

Monthly Mag. 2:44. Type from Missouri

River region.
1817. Diplostoma alba Rafinesque, Amer.

Monthly Mag. 2:44. Type from Missouri

River region.
1821. Mus saccatus Mitchill, Med. Repos. (n.s.)

[New York], 6(21):249. Type from "area

bordering on Lake Superior."
1825. Ascomys canadensis Lichtenstein, Abh.

K. Akad. Wiss., Berlin, for 1822, p. 20.

Type from "Canada."
1829. Geomys bursarius Richardson, Fauna

Boreali-Americana 1:203. First use of name

combination.
1939. Geomys bursarius majusculus Swenk,

Missouri Valley Fauna 1:6. Type from

Lincoln, Lancaster County, Nebraska.
1958. Geomys bursarius missouriensis Mc-
Laughlin, Los Angeles County Mus. Cont.

Sci. 19:1. Type from 2 mi. N Manchester,

St. Louis County, Missouri.

Type specimen. — Uncertain; perhaps
the specimen from the Bullock collec-
tion, now in the Rijksmuseum van Nat-
uurlijke Historie, Leiden (see Merriam,
1895:123-127).

Distribution. — Occurs principally in
tall-grass prairie in the northeastern por-

tion of the Great Plains west of the Mis-
sissippi River (Fig. 1).

Description. — For a complete descrip-
tion of Geomys, see Merriam (1895).
This description and those following re-
fer to adult females. Size large for ge-
nus; adult females averaging head and
body length and condylobasal length
from 170 mm and 44.4 mm (in north-
western Wisconsin) to 210 mm and 50.5
mm (in northeastern Iowa), respectively
(Fig. 3). Skull large, robust (Fig. 2A);
zygomata broad, widely divergent an-
teriorly; sagittal crest narrow and usually
high; rostrum long (both relatively and
absolutely); frontal-premaxillary suture
on dorsum falling anterior to or reaching
(not exceeding) a line drawn between
the anterior-most points in the orbits;
mastoid processes long, projecting at an
angle approximately 20° from horizontal;
braincase oval; naso-frontal region usu-
ally domed slightly; anterior root of zy-
gomatic arches sloping back strongly.
Face of upper incisors strongly bisulcate;
minor sulcus fine and close to inner edge
of tooth; principal sulcus much deeper
and wider, and lying on or slightly ex-
ternal to median line; enamel face
rounded externally and between sulci.
Incisors procumbent. Upper premolar
bilobate, curved, sloping forward ba-
sally, and concave anteriorly; last molar
curving backwards basally, and concave

POCKET GOPHERS OF THE GENUS GEOMYS

11

posteriorly. M1 and M2 curved back-
ward, M1 slightly and M2 more strongly.
P, bilobate, large, curving forward ba-
sally, concave anteriorly; Mi through M3
decreasing progressively in length, and
increasing in basal curvature towards
posterior; M:; concave posteriorly.

Color a rich chocolate or chestnut

brown dorsally, hairs without dark tips;
slightly to considerably lighter ventrally;
ventrum often with silver cast. Dorsum
sometimes with a medial darkened re-
gion from snout to base of tail, 10 to 15
mm wide; this dorsal "stripe" is most
conspicuous in populations in Kansas
and Missouri, and only barely discerna-

102

90

i_

KILOMETERS

102

Fig. 1. — Map showing distribution of Geomys in the central and northern Great Plains. Solid
circles = G. bursarius bursarius; open circles = G. b. illinoensis; open squares = G. b. wiscon-
sinensis; solid squares = G. lutescens lutescens; solid triangles = G. /. major; open triangles =

G. breviceps sagittalis.

12

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

ble in Nebraska, Iowa, and most of
Minnesota. Pelage near month often
white. Dorsal surface of fore- and hind-
feet with white fur (hindfeet sparsely
covered ) ; white fur often on ventral base
of forelimbs, sometimes extending to
base of humerus. Mystacial vibrissae
stiff, relatively fine, 5-30 mm in length;
supraorbital and preauricular vibrissae
very fine, few in number, up to 10 mm
in length. Color of vibrissae ranges from
light to dark brown. Tail moderately
long (36-42% of head and body) and
thick, with blunt tip. Basal quarter of
tail usually well furred, with density of
hair decreasing toward tip. Distal por-
tion ( half to four- fifths ) often with white
hairs, sometimes nearly naked. Melanism
rare, at frequencies less than 1% in known
populations.

Forefoot large with five stout claws;
digits III, IV, II, V, and I with progres-
sively shorter claws. Hindfoot not un-
usually large, claws normal.

Diagnosis and comparisons. — Differs
from G. breviceps in having much
greater overall size; dorsal color without
"peppering" due to dark distal bands on
hairs; sagittal crest prominent and high,
rather than a poorly-defined ridge 4-5
mm wide between the temporal crests;
zygomatic arches wider at anterior than
posterior angles, rather than approxi-
mately equal in width; auditory bullae
proportionately longer, less dome-
shaped. Differs from G. lutescens in
having rich brown cast to dorsal pelage,
rather than having yellowish cast; tem-
poral ridges fused into a sagittal crest
in adult females, rather than being sepa-
rated by a ridge 2-3 mm wide defined
by the temporal ridges; rostrum abso-
lutely and relatively longer; frontal-pre-
maxillary suture on dorsum usually fall-
ing anterior to or reaching (not exceed-
ing) a line drawn between anteriormost
points in the orbits; mastoid processes
longer, less closely appressed to skull;
braincase oval, not rectangular; naso-
frontal junction domed, not flat. Differs
from G. bursarius illinoensis in being
smaller, having a proportionately shorter

rostrum and shorter tail (mean 39.1%;
33.9 to 44.3% length of head and body),
and having brown rather than slate-gray
("melanistic") fur in most individuals.
Differs from G. bursarius loisconsinensis
in having a proportionately shorter tail
and in having the anterior portion of the
frontals forming a rectangle rather than
a square ( Jackson, 1961 ) . Baculum long
(mean = 11.3 mm) with proportions
typical for the species-group (Fig. 15;
Table 8).

Specimens examined. — Iowa: Clay
Co.: no specific locality (2 UMMZ);
Clayton Co.: 4 mi. NE Garbin (1 SM);
4 mi. E Monona (2 SM); Monona (7
SM); Dubuque Co.: 4 mi. NE Bankston
(1 SM); Emmett Co.: 3Vz mi. S, 234 mi.
E Wallingford (1 KU); Grundy Co.:
m mi. S, 2V. mi. W Wellsburg (1 KU);
Mahaska Co.: 2% mi. E New Sharon
(1 KU); 2 mi. N, 3 mi. E Oskaloosa (1
KU); Marion Co.: Knoxville (11 USNM);
Marshall Co.: 3 mi. W Green Mountain
(1 KU); SE ]4 sec. 2, T82N, R17W (1
KU); Monroe Co.: Vh mi. N Melrose
(1 KU); Plymouth Co.: 3 mi. N Le
Mars; Story Co.: 1 mi. N Ames (1
MSU); SW % sec. 10, T83N, R24W (1
KU); Winneshiek Co.: Decorah (1
UMMZ).

Kansas: Atchison Co.: 3 mi. N Cum-
mings (2 KU); Vk mi. S Muscotah (1
KU); Butler Co.: 8 mi. W Rosalia (2
KU); Cloud Co.: 2% mi. N, 2 mi. E
Jamestown (6 KU); 4 mi. E Jamestown
(1 KU); Douglas Co.: 1% mi. N, Vh mi.
E Lawrence (4 MHP); Vi mi. N, Vh mi.
E Lawrence (1 KU); Lawrence (8 KU);
Vs mi. W Lawrence (1 KU); 2% mi. W
Lawrence ( 1 KU ) ; 1 mi. S, 4 mi. W
Lawrence (2 KU); no specific locality
(2 KU); Greenwood Co.: V\ mi. E Ham-
ilton (1 KU); Vm mi. E Hamilton (1
KU); Hamilton (6 KU); ¥a mi. S Hamil-
ton (1 KU); V> mi. S Hamilton (2 KU);
1 mi. S Hamilton (1 KU); 8V2 mi. S
Toronto (2 KU); no specific locality (2
KU); Jackson Co.: Birmingham (1 KU);
Jefferson Co.: 1 mi. NW Midland (1
KU); Jewell Co.: NE Vi sec. 12, T1S,
R6W (2 MHP); Leavenworth Co.: Fort

POCKET GOPHERS OF THE GENUS GEOMYS

13

co

-S

a

Co

'--
3

►a
U

3

eg

-a

Co

3

-a

co

S
o

U

3
03

«5

a

a!
H
bO
O

4-»
O

JS

PL,

d

14

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

8

<4i

c

^
u

c

u

CO

Ss

o
U
U

eg

<D

3

a
a

O

ft

Si

o

■*->
O

-C

t3

3

C

T

POCKET GOPHERS OF THE GENUS GEOMYS

15

Leavenworth ( 1 KU ) ; no specific locality
(1 KU); Marion Co.: Vh mi. NE Lin-
colnville (1 KU); Marshall Co.: Vi mi.
N, 1 mi. W Blue Rapids (2 KU); Water-
ville (2 KU); McPherson Co.: 1 mi. S,
y2 mi. W Lindsborg (1 KU); Mitchell
Co.: 1 mi. N, 2 mi. E Beloit (1 KU)
:i mi. S, 31/. mi. W Beloit (8 KU)
Nemaha Co.: 2 mi. N Sabetha (1 MSU)
Osborne Co.: % mi. S, 8 mi. E Osborne
(2 KU); Pottawatomie Co.: 1 mi. E
Olsburg (1 KU); 1% mi. N, 1 mi. W
Olsburg (1 KU); Onaga (3 USNM)
Republic Co.: Scandia (4 KU); 2 mi. N
V2 mi. W Scandia (1 KU); Riley Co.
5V2 mi. N, 2% mi. E Randolph (2 KU)
5Vo mi. N, m mi. E Randolph (5 KU)
V2 mi. S, Yi mi. W Randolph (1 KU)
Washington Co.: 7v4 mi. N, 3% mi. W
Washington (1 KU); Wyandotte Co.:
Wyandotte (1 MHP).

Minnesota: Anoka Co.: 5!4 mi. N,
1 mi. W Lino Lakes (1 KU); Carlos
Avery Game Mgmt Area (1 KU);

Becker Co.: SW Yt sec. 24, T139N, R42W
(1 MMNH); Beltrami Co.: 11 mi. N
Bemidji (1 BSU); SEM sec. 5, T148N,
R35W (2 MMNH); Brown Co.: SW Y*
sec. 7, T110N, R21W (1 MMNH); Cass
Co.: Cass Lake (1 USNM); Goodhue
Co.: Goodhue (1 MMNH; 1 UMMZ);
Hennepin Co.: Minneapolis (1 MMNH);
Houston Co.: La Crescent (1 USNM);
Kittson Co.: Karlstad (1 MMNH); St.
Vincent (1 USNM); Marslmll Co.: 1 mi.
N Alvarado (1 MMNH); Morrison Co.:
1 mi. S Pillager (1 MMNH); Norman
Co.: NW M sec. 13, T145N, R44W (1
MMNH); Pennington Co.: no specific
locality (1 BSU); Polk Co.: 4% mi. S
Fisher ( 1 MMNH ) ; NW U sec. 6, T149N,
R39W (1 MMNH); Vk mi. S Alvarado
(1 MMNH); Ramsey Co.: St. Paul (3
KU; 22 MMNH); Rock Co.: 5 mi. WSW
Luverne (1 MSU); Slierburne Co.: Elk
River (5 KU, 6 USNM); 6 mi. SE St.
Cloud ( 1 UMMZ); Swift Co.: NE V± sec.
2, T120N, R40W (2 MMNH); Traverse

Fig. 2. — (continued). Photographs of crania of adult female of: E. Geomys breviceps sagittalis.

16

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Co.: near Brown's Valley (1 USNM);
Wheaton (1 MMNH); Winona Co.: 8
mi. S, 5 mi. W Winona (1 MMNH).

Missouri: Andrew Co.: 1 mi. E Flag
Springs (1 UMC); Atchison Co.: 5 mi.
S. 21. mi. W Rock Port (1 UMC); Bu-
chanan Co.: St. Joseph (1 UMC); 5 mi.
SW St. Joseph (3 UMC); Clark Co.:
Kahoka (2 UMC); no specific locality
(4 UMC); Clay Co.: Smithville (1
UMC); Crawford Co.: Steelville (1
UMC); Franklin Co.: 2Vj mi. E Sullivan
(1 SM); Sullivan (1 UMC); Holt Co.:
Mound City (1 UMC); no specific lo-
cality (1 UMC); Jackson Co.: Buckner
(1 UMC); no specific locality (1 UMC);
Lewis Co.: Wakonda State Park (1 KU);
no specific locality (2 UMC); Marion
Co.: no specific locality (2 UMC); Perry
Co.: Perryville (1 UMC); Scotland Co.:
no specific locality ( 1 UMC); St. Charles
Co.: 3% mi. S Orchard Farm (1 KU);
no specific locality (3 UMC); St. Louis
Co.: Afton (2 UMC); Baden (1 UMC);
H mi. N Black Jack (1 UMC); 2 mi. E
Creve Coeur Lake (1 UMC); 1V_- mi. S
Creve Coeur Lake (1 UMC); 2!72 mi.
NE Cross Keys (2 KU); VA mi. NW
Cross Keys (1 KU); Florissant (4 KU);
Ladue (1 UMC); 1 mi. W Lindberg
(1 UMC); St. Louis (2 USNM); Clayton
Rd. and Hwy. 340 (1 UMC); no specific
locality (4 UMC).

Nebraska: Adams Co.: Hastings (4
AMNH); Antelope Co.: V* mi. N, 2% mi.
E Oakdale (5 UMMZ); 2% mi. E Oak-
dale (5 FMNH); W edge Oakdale (S
SM); V> mi. W Oakdale (3 SM); %o mi.
S, %o mi. W Oakdale (5 KU); %o mi. S,
%o mi. W Oakdale (6 KU); V-i mi. S, V/i
mi. E Oakdale (1 UMMZ); 2!io mi. S
Oakdale (1 SM); 1 mi. W Tilden (2
SM); 5 mi. W Tilden (1 KU); Butler
Co.: 4 mi. E Rising City ( 1 KU); Dodge
Co.: Ames (1 USNM); Gage Co.: 2 mi.
S, % mi. E Barnston (1 KU); Knox Co.:
!■• mi. N, 3 mi. W Center (1 KU); V-i mi.
S, 1 mi. W Niobrara (3 KU); 2V> mi. S,
9 mi. W Niobrara (2 KU); mouth of
Niobrara River (1 USNM); Verdigre
(1 USNM); Lancaster Co.: V-> mi. S
College View (1 KU); 5 mi. N Lincoln

(1 KU); 5% mi. E Lincoln (1 KU);
Lincoln (2 KU, 10 NSM); 2 mi. S, 1 mi.
E Malcolm (1 NSM); 3 mi. S Malcolm
(4 NSM); Madison Co.: Norfolk (2
USNM); Pierce Co.: 7>io mi. N Meadow
Grove (1 KU); Platte Co.: Columbus
(1USNM).

North Dakota: Barnes Co.: Valley
Citv (1 USNM); Cass Co.: Casselton
(1 USNM); Grand Forks Co.: Grand
Forks (1 USNM); Manvel (1 USNM);
La Moure Co.: La Moure (1 USNM);
Ransom Co.: Lisbon (1 USNM); Rich-
land Co.: Blackmir (1 USNM); 5 mi. E
Fairmount (2 USNM); Lidgerwood (6
USNM); Trail Co.: Portland (7 USNM).
South Dakota: Bon Homme Co.:
\v-< mi. E Scotland (1 SM); Brookings
Co.: 5!-j mi. N, 1 mi. E Volga (1
MMNH); 5 mi. N, 3 mi. W Volga (2
MMNH); 4 mi. N, 2 mi. W Volga (2
MMNH); 2'j mi. S, 3 mi. W Volga (1
MMNH); 3% mi. S, 2% mi. W Volga
1 MMNH); 3% mi. S, 3 mi. W Volga
1 MMNH); 3!4 mi. S, 4% mi. W Volga
1 MMNH); AYi mi. S, VA mi. W Volga

1 MMNH); 5 mi. S, 1 mi. W Volga

2 MMNH); 6!4 mi. S, V'z mi. E Volga

1 MMNH); 7!1> mi. S, 2% mi. E Volga

2 MMNH); Tf-i mi. S, 2% mi. E Volga
1 MMNH); IVi mi. S, 2 mi. E Volga
1 MMNH); Grant Co.: 6 mi. N Mil-

)ank (4 KU); Lake Co.: 1 mi. S, 1 mi.
E Madison (1 MMNH); IVi mi. S, 1 mi.
E Madison (1 MMNH); Marshall Co.:
Roy Lake State Park (3 KU); Moody
Co.: 14 mi. S, 1% mi. E Brooking (1
MMNH); 12 mi. S, V> mi. E Brooking
(1 MMNH); 3¥i mi. S, 1% mi. W Volga
(1 MMNH).

Wisconsin: Bayfield Co.: 4 mi. E
Iron River (1 UIMNH); 5 mi. W Iron
River (1 UIMNH); 7 mi. W Iron River
(1 UIMNH); Burnett Co.: Danbury (1
USNM); Chippewa Co.: Anson Town-
ship (3 USNM); Douglas Co.: 3 mi. S,
3 mi. W Brule (8 UIMNH); 3 mi. N, 3
mi. E Solon Springs (1 UIMNH); Solon
Springs (1 USNM); Polk Co.: V-i mi. N
Dresser (1 UIMNH); Vi mi. S, 1 mi. W
Dresser (2 UIMNH); 3 mi. S St. Croix

POCKET GOPHERS OF THE GENUS GEOMYS

17

Falls (1 UIMNH); Trempealeau Co.:

4 mi. N Arcadia (2 UIMNH).

Geomys bursarius illinoensis Komarek
and Spencer, 1931

1931. Geomys bursarius illinoensis Komarek
and Spencer, J. Mammal. 12:405.

1936. Geomys illinoensis Lyon, Amer. Midi.
Nat. 17:216. First use of name combination.

1941. Geomys bursarius illinoensis Necker and
Hatfield, Bull. Chicago Acad. Sci. 6:51.

Type specimen. — Chicago Academy
of Sciences 713, from 1 mi. S Momence,
Kankakee County, Illinois. Type now
housed in die Field Museum of Natural
History.

Distribution. — Occurs in tall-grass
prairie — oak-hickory savannah commu-
nity ( Kiichler, 1964 ) in eastern and cen-
tral Illinois and west-central Indiana
(Fig. 1).

Diagnosis and comparisons. — May be
distinguished from G. breviceps, G. bur-
sarius bursarius, and G. Jutescens, as
noted in the G. b. bursarius description
and diagnosis. Differs from G. bursarius
wisconsinensis in being larger with a
proportionately longer rostrum ( Fig. 2B )
and tail (40 to 44% length of head and
body), and having slate-gray ("melanis-
tic") pelage in nearly all individuals.

Specimens examined. — Illinois: Cass
Co.: 2%o mi. ESE Bluff Springs (1 KU);
DeWitt Co.: VA mi. ENE Clinton (1
UIMNH); 2 mi. E Clinton (1 KU); VA
mi. E Clinton ( 1 UIMNH); 1 mi. E Clin-
ton (1 UIMNH); Clinton (1 UIMNH,

5 USNM); 1 mi. S, 1 mi. E Clinton (2
UIMNH); 1 mi. SSE Clinton (1
UIMNH); % mi. NE Junction US Rt. 51
and 54 (1 KU); Kankakee Co.: VA mi.
S, VA mi. W Kankakee (1 UIMNH);
LaSalle Co.: §io mi. S Lostant (1
UIMNH); Logan Co.: 5 mi. E Lincoln
(1 UIMNH); 2 mi. E Lincoln (1
UIMNH); Madison Co.: Collinsville (3
KU); Marshall Co.: 2 mi. E Lacon; Ma-
son Co.: 4V> mi. S Bath (1 KU); 3 mi.
E Havanna (1 KU); Havanna (1 USNM);
McLean Co.: 10 mi. N Bloomington
(1 UIMNH); 2 mi. SE Bloomington (1
KU); 2 mi. W Hudson (2 UIMNH); 4

mi. E McLean (2 UIMNH); 1 mi. S
McLean (1 UIMNH); 9 mi. N Normal
(1 UIMNH); 5% mi. N Normal (1
UIMNH ) ; 3 mi. N Normal ( 1 UIMNH ) ;
Normal (5 KU); Morgan Co.: 5 mi. NE
Jacksonville (1 UIMNH); Tazewell Co.:
Lilly (1 UIMNH); Will Co.: 2 mi. W
Wilmington (1 UMMZ); Woodford Co.:
1 mi. W Benson (1 UIMNH); 4 mi. N
El Paso (1 UIMNH); 2% mi. N El Paso
(1 UIMNH); 2 mi. N El Paso (1
UIMNH); 6 mi. N Eureka (1 UIMNH);
1 mi. NW Kappa (1 UIMNH); 1 mi. S,
S mi. W Minonk (1 UIMNH).

Indiana: Jasper Co.: 3 mi. E Rose-
lawn (1 USNM); 2 mi. E Roselawn
(1 UIMNH); Newton Co.: Lake Village
(1 UMMZ); 1 mi. E Roselawn (1
UIMNH, 1 USNM); Junction US Rt. 10
and 55 (1 USNM); no specific locality
(1 UMMZ); Tippecanoe Co.: Battle
Ground ( 1 UMMZ, 6 USNM); Lafayette
(2 USNM); Warren Co.: VA mi. NE
Pine Village (1USNM).

Geomys bursarius wisconsinensis Jack-
son, 1957

1957. Geomys bursarius wisconsinensis Jackson,
Proc. Biol. Soc. Washington 70:33.

Type specimen. — USNM 243055,
taken on 13 September 1922 at Lone
Rock, Richland County, Wisconsin.

Distribution. — Occurs in southwest-
ern Wisconsin, in the vicinity of the
north bank of the Wisconsin River, in
oak savannah intermixed with maple-
basswood forest (Fig. 1).

Diagnosis and comparisons. — Differs
from other G. bursarius in having that
portion of the frontals which projects
between the premaxillaries contact the
nasals in the shape of a square rather
than an elongated rectangle; tail moder-
ately long (43.2% length of head and
body). Comparisons with other taxa as
noted above.

Specimens examined. — Wisconsin:
Crawford Co.: 1 mi. W Wauzeka (5
UIMNH); Richland Co.: tt mi. E
Gotham (3 SM); Gotham (3 SM); Lone
Rock (7 MMNH, 2 USNM).

18

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Geomys breviceps Baird, 1S55
Southern Plains Pocket Gopher

Geomys breviceps breviceps Baird, 1855

1855. Geomys breviceps Baird, Proc. Acad.
Nat. Sci. Philadelphia 7:335.

Type specimen. — USNM 156 (skin)/
1138 (skull), obtained in 1852 at Prairie
Mer Rouge, Morehouse Parish, Louisi-
ana.

Distribution. — Extralimital to study
area; vicinity of Mer Rouge, Louisiana
only (see Lowery, 1974).

Diagnosis and comparisons. — See di-
agnosis of G. /;. sagittalis.

Specimens examined. — Louisiana:
Morehouse Parish: Prairie Mer Rouge
(1 USNM, holotype); plus approxi-
mately 45 others in USNM from the
vicinity of the type locality.

Geomys breviceps sagittalis Merriam,
1895

1895. Geomys breviceps sagittalis Merriam, N.
Amer. Fauna 8:134.

1938. Geomys breviceps brazensis Davis, J.
Mammal. 19:489. Type from 5 mi. E
Kurten, Grimes County, Texas.

1940. Geomys breviceps dutcheri Davis, Texas
Agr. Exp. Sta. Bull. 590:12. Type from
Fort Gibson, Muskogee County, Oklahoma.

1940. Geomys breviceps terricolus Davis, Texas
Agr. Exp. Sta. Bull. 590:17. Type from 1
mi. N Texas City, Galveston County, Texas.

1940. Geomys breviceps pratincolus Davis,
Texas Agr. Exp. Sta. Bull. 590:18. Type
from 2 mi. E Liberty, Liberty County,
Texas.

1940. Geomys breviceps ludemani Davis, Texas
Agr. Exp. Sta. Bull. 590:19. Type from 7
mi. SW Fannett, Jefferson County, Texas.

Type specimen.— USNM 32936 (skin)/
44957 (skull), obtained 28 March 1892
at Clear Creek, Galveston Bay, Galveston
County, Texas.

Distribution. — Ranges from southern
Texas near Galveston Bay (extralimital;
see Honeycutt and Schmidly, 1979) to
northeastern Oklahoma and eastern Ar-
kansas ( Fig. 1 ) .

Description. — Size small for genus;
adult females (within study area) aver-
aging head and body length from 145
mm to 155 mm, condylobasal length
39.3 mm to 40.2 mm (Fig. 3). Skull

(Fig. 2E) small, sturdy; zygomatic arches
only slightly broader anteriorly than at
posterior angle, or equal in width; inter-
orbital region flat or concave; rostrum
absolutely and relatively short, moder-
ately narrow; temporal crests separated
medially by poorly defined ridge 4-5 mm
wide; mastoid processes short, nearly as
broad (at base) as long; foramen ro-
tundum relatively low on alisphenoid,
not obscured in lateral view by zygo-
matic arch; braincase smoothly rounded.
Incisors as in G. bursarius, but less pro-
cumbent. P4 and P4 usually straight in
lateral profile, less often slightly con-
cave. Molars as in G. bursarius. Fore-
feet as in G. bursarius, although propor-
tionately less enlarged. Tail shape and
pelage density as in G. bursarius; tail
length moderate, 41-49% of head and
body length.

Color of dorsal pelage variable, show-
ing a general cline from medium brown
with only a trace of yellow in gophers
from the eastern quarter of Oklahoma,
to a lighter brown suffused with a yellow
(in the north) or orange (in the south)
tone suffused through both dorsal and
ventral pelage along the western margin
of the distribution of the species (from
Payne through Love and Marshall coun-
ties). "Peppering" due to black tips on
dorsal hairs prominent in all popula-
tions. Mid-dorsal region sometimes show-
ing darkening; development of this is
variable within populations, but seems
never to be so prominent as in some
populations of G. bursarius and G. lutes-
cens. Ventral pelage often with a silver
sheen, and often with white spotting.

Diagnosis and comparisons. — Differs
from G. bursarius as noted above. Dif-
fers from G. lutescens in being generally
smaller; temporal crests separated by
poorly defined ridge 4-5 mm wide rather
than prominent ridge 2-3 mm wide;
breadth across anterior and posterior an-
gles of zygomatic arches approximately
equal, rather than anterior angle width
distinctly greater; rostrum narrow, rela-
tive to length, rather than relatively short
and broad; foramen rotundum relatively

POCKET GOPHERS OF THE GENUS GEOMYS

19

low on alisphenoid, not obscured from
view laterally by zygomatic arch, rather
than relatively high on alisphenoid and
obscured from lateral view by zygomatic
arch. Baculum smallest for members of
species group, proportions typical (Fig.
15; Table 8).

Specimens examined. — Arkansas:
Crawford Co.: Fort Smith (3 USNM);
Ouachita Co.: Camden (1 USNM); Pu-
laski Co.: 3 mi. E Wrightsville (1 SM);
2 mi. E Wrightsville (2 SM, 5 USNM);
Saline Co.: Benton (2 USNM).

Oklahoma: Adair Co.: 2Vz mi. E
Stilwell (3 USNM); Atoka Co.: 5 mi. E
Atoka (1 UMMZ); Bryan Co.: 4 mi. E
Durant (2 OSU); Carter Co.: 2 mi. W
Ratliff City (1 UMMZ); Choctaw Co.:
7 mi. SSE Fort Towson (4 SM); Cleve-
land Co.: Norman, Reeve's Park ( 1 SM);
1.6 mi. E Norman (1 SM); Coal Co.:
1 mi. W Coalgate (1 OSU, 1 UMMZ);
Creek Co.: 15 mi. W Sapulpa (1
UIMNH); Garvin Co.: Washita River
Bottom (1 OSU); Grady Co.: 1 mi. W
Chickasha (1 OSU, 1 SM); Haskell Co.:
Whitefield (1 OSU); Hughes Co.: Ca-
nadian River, S of Holdenville (1 OSU);
Lincoln Co.: Chandler (2 UMMZ);
Love Co.: 7 mi. S Marietta (1 UMMZ);
Marshall Co.: 2 mi. E Willis (5 SM);
1 mi. W Willis (5 SM); S side Washita
River on Hwy. 12 (1 UMMZ); Lake
Texoma (1 UMMZ); McClain Co.: 2 mi.
W Byars (1 OSU); Rosedale (1 OSU, 1
SM); Mcintosh Co.: Eufaula (1 SM);
Muskogee Co.: Wildlife Cons. Sta.,
Braggs (1 UMMZ); Ft. Gibson (1
UMMZ, 2 USNM); Okfuskee Co.: 3 mi.
N, 1 mi. E Mason (1 OSU); Oklahoma
Co.: 2 mi. E Tinker Field (1 OSU, 1
SM, 1 UMMZ); Pay tie Co.: Stillwater,
Boomer Lake ( 1 OSU ) ; 2 mi. S Stillwater
(2 OSU); 8i<> mi. E Stillwater Cemetery
(1 OSU); Pittsburg Co.: McAlester (2
OSU); Pottawatomie Co.: Asher (1
OSU, 1 UMMZ); Pushmataha Co.: Ant-
lers (1 UMMZ); Tulsa Co.: 5 mi. W
Sand Springs Bridge, Arkansas River ( 1
OSU); Mohawk Park (3 UMMZ); Wag-
oner Co.: Cowetta, S of Arkansas River
(1 UMMZ).

Remarks. — Geomys attwateri, an ad-
ditional species related to G. breviceps,
occurs in south-central Texas, west of
the Brazos River. See Honeycutt and
Schmidly (1979) and Tucker and
Schmidly (1981) for discussion.

Geomys lutescens Merriam, 1890
Yellow Pocket Gopher

Geomys lutescens lutescens Merriam,
1890

1890. Geomys bin sarins lutescens Merriam, N.
Amer. Fauna 4:51.

1895. Geomys lutescens Merriam, N. Amer.
Fauna 8:127. First use of name combina-
tion.

1938. Geomys lutescens hyiaeus Blossom, Oc-
eas. Papers Mus. Zool., Univ. Michigan
368:1. Type from 10 mi. S Chadron, Dawes
County, Nebraska.

1940. Geomys lutescens jugossicularis Hooper,
Occas. Papers Mus. Zool., Univ. Michigan
420:1. Type from Lamar, Prowers County,
Colorado.

1940. Geomys lutescens levisagittalis Swenk,
Missouri Valley Fauna 2:4. Type from
Spencer, Boyd County, Nebraska.

1940. Geomys lutescens vinaccus Swenk, Mis-
souri Valley Fauna 2:7. Type from Scotts-
bluff, Scotts Bluff County, Nebraska.

Type specimen. — USNM 23595, taken
on 27 May 1889 from sandhills on Bird-
wood Creek, Lincoln County, Nebraska.

Distribution. — Widespread in west-
ern Nebraska and Kansas and adjacent
parts of Colorado, South Dakota, and
Wyoming in short-grass and mixed grass
prairie (Fig. 1).

Description. — Size medium for ge-
nus; adult females averaging head and
body length 162 mm to 182 mm, condy-
lobasal length 39.4 mm to 43.3 mm
(Fig. 3). Skull (Fig. 2C) broad and
robust, dorsoventrally flattened; zygo-
mata broad and roughly square in out-
line, broader anteriorly than posteriorly;
temporal ridges separated by 2-3 mm;
braincase short and broad; mastoid proc-
esses of moderate length, projecting at
angle averaging 40J-50° from horizontal;
posterior edge of prem axillary-frontal
suture usually exceeding anterior edge
of orbit; naso-frontal junction approxi-
mately flat in lateral view; foramen ro-

20

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

tundum relatively high on alisphenoid,
usually obscured in lateral view by zy-
gomatic arch. Incisors as in G. bursarius,
but less procumbent. Anterior edge of
P4 usually straight in lateral profile; edge
of P4 usually convex. Molars as in G.
bursarius.

Pelage color variable. Dorsal pelage
usually with yellowish cast, especially
on sides. A mid-dorsal stripe of dark fur,
ranging from two to five cm wide pres-
ent on most individuals. Dorsal color
darkest and mid-dorsal stripe widest in
north and east, especially north of Nio-
brara River, becoming more pallid to
west; gophers from Wyoming and Colo-
rado often sandy yellow. Localized re-
gions with unusually dark or light soils
often with similarly colored gophers ( see
Hendrickson, 1972, and "Nebraska re-
gion," below). "Peppering" due to black
tips on dorsal hairs often present, though
usually not conspicuous.

Diagnosis and comparisons. — Differs
from all G. bursarius as noted in the dis-
cussion of G. h. bursarius. Differs from
G. /. major in being slightly larger over-
all, maxillary usually not visible on sides
of rostrum when viewed dorsally. Bacu-
lum moderate in length (mean = 10.8),
unusually broad for species-group (Fig.
15; Table 8).

Specimens examined. — Colorado:
Adams Co.: Barr (1 CSU); M mi. W, 1
mi. E Bennett ( 11 KU); 5 mi. E Brighton
(1 KU); 2 mi. E Brighton (6 KU); m
mi. N, 3 mi. W Strasburg (2 KU); 2 mi.
N, 3% mi. W Strasburg (2 KU); Bent
Co.: Las Animas (3 USNM); 4 mi. W
Las Animas (3 MVZ); Boulder Co.: 8L'io
mi. E Boulder (1 KU); 1% mi. S, 1 mi.
E Lafayette (1 KU); Douglas Co.: 2 mi.
N Parker (1 KU); El Paso Co.: Vk mi.
N, 9 mi. E Colorado Springs (1 CSU);
Colorado Springs (1 KU); Fremont Co.:
4 mi. SSE Canon City (4 KU); Larimer
Co.: 7 mi. NW Ft. Collins ( 1 CSU); 3(/io
mi. N Ft. Collins (1 CSU); 3 mi. N, 2
mi. W Ft. Collins (1 CU); 2V2 mi. NE
Ft. Collins (1 CSU, 1 CU); 2 mi. N Ft.
Collins (1 CU); 2 mi. NW Ft. Collins
(1 CSU); Vi mi. N Ft. Collins (1 CU);

3 mi. E Ft. Collins (2 CU); 2%o mi. E
Ft. Collins (1 CU); Ft. Collins (1 CSU,

4 CU); 1 mi. S, 3% mi. E Ft. Collins (1
CSU); 1 mi. N Laporte (1 CU); 1 mi.
NE Masonville (1 CU); Vi mi. WSW
Masonville (4 KU); Logan Co.: Vk mi.
S, 13 mi. W Peetz (1 KU); Morgan Co.:
2 mi. N, T'i mi. W Ft. Morgan (5 KU);
Prowers Co.: Lamar (1 UMMZ); 1 mi.

5 Lamar (1 CU, 1 UIMNH); Washing-
ton Co.: 9 mi. N, % mi. E Cope (1 KU);
Cope (2 KU); v-i mi. S, 3 mi. W Cope
(2 KU); % mi. S, 1 mi. W Cope (2 KU);
Weld Co.: V2 mi. N, 1 mi. W Briggsdale
(1 KU); 5% mi. E Ft. Lupton (1 KU);
:;i mi. N, 2U mi. E Hudson (1 KU);
Yuma Co.: Wray (2 CU); 7 mi. S Wray
(2 CU).

Kansas: Barton Co.: 8 mi. N Ellin-
wood (1 KU); Cheyenne Co.: 10 mi. N,
2 mi. W Bird City' (2 MHP); 1 mi. N
St. Francis (1 KU); 1 mi. W St. Francis
(2 KU); 8 mi. S, 11% mi. W St. Francis
(1 KU); Decatur Co.: 1 mi. N, 1 mi. W
Dresden (1 MIIP); 1 mi. W Dresden
(2 MHP); Oberlin (1 KU); Ellis Co.:
V* mi. N, 2 mi. E Ellis (3 MHP); 16 mi.
N, 1 mi. W Hays (1 MHP); V'i mi. N,
2 mi. W Havs (3 MHP); 1 mi. N, l\'-i
mi. W Hays (1 MHP); v-i mi. N, 3 mi.
W Havs (1 MHP); :L> mi. N, 6 mi. W
Hays (1 MIIP); Hays (1 KU); 2 mi. W
Hays ( 1 MHP); 6 mi. W Hays 1 (MHP);
:;i mi. W Yocemento (6 MHP); NE U
sec. 11, T15S, R20W (2 MHP); SW V*
sec. 16, T15S, R19W (2 MHP); NE M
sec. 25, T13S, R19W (1 MHP); NW \\
sec. 30, T13S, R18W (3 MHP); EUs-
icorth Co.: 2 mi. S Ellsworth (1 KU);
Finneij Co.: 37i« mi. N, V-i mi. W Garden
City (1 KU); Graham Co.: 3% mi. N,

6 mi. E Hill City (3 MHP); 2 mi. N,
2% mi. E Hill City (2 MHP); % mi. N,
5% mi. E Hill City (1 MHP); Greeley
Co.: 4U mi. E Tribune (1 KU); Kearny
Co.: 10 mi. N Lakin (1 KU); 31^ mi. N,
4 mi. E Lakin (1 KU); 2M mi. W Lakin
(2 KU); 15 mi. S, 4% mi. E Leoti (1
KU); Lane Co.: Pendennis (2 KU);
Logan Co.: 17%o mi. N Leoti (1 KU);
Ness Co.: 1 mi. S, 16 mi. W Ness City
(2 KU); Norton Co.: % mi. N, 4 mi.

POCKET GOPHERS OF THE GENUS GEOMYS

21

E Lenora (3 MHP); Norton (1 MHP);
3 mi. S Norton (2 KU); 4 mi. S, 4 mi.
W Norton (1 MHP); Osborne Co.: V>
mi. E Alton (2 KU); V-i mi. S, VA mi. E
Alton (1 KU); Vi mi. S, 1 mi. E Alton
(1 KU);%mi. S, % mi. E Alton (2KU);
Phillips Co.: 1 mi. S, 2 mi. W Glade
(1 MHP); V2 mi. N, 7 mi. W Kirwin
(1 MHP); 6 mi. W Kirwin (1 MHP);
M mi. S, 6 mi. W Kirwin (1 MHP);
Rawlins Co.: 2 mi. E Atwood (1 KU);
S bank Lake Atwood (1 MHP); Vi mi.
N, 3 mi. E Herndon (5 MHP); Rooks
Co.: 4 mi. N, 27io mi. E Damar (2
MHP); M mi. S, 6 mi. E Stockton (5
KU); 1 mi. S, 8 mi. W Stockton (3 KU);
Russell Co.: 7 mi. S, 1 mi. E Lucas (3
KU); W % sec. 9, T13S, R11W (1 MHP);
E Vi sec. 10, T13S, R11W (1 MHP);
Sherman Co.: 15 mi. N, 3 mi. E Edson
(15 KU); Thomas Co.: 7 mi. N, 2% mi.
E Colby (7 KU); Trego Co.: 15 mi. N,
2 mi. E Brownell (6 MHP); 8% mi. N
Ogallah (1 MHP); 11 mi. S, 2 mi. W
Ogallah (1 MHP); 4 mi. N, 1 mi. E
WaKeeney (1 MHP); Wallace Co.: IOVj
mi. N, 4% mi. W Weskan (1 MHP); 5%
mi. N, m mi. W Weskan (3 MHP); 5%
mi. N Weskan (1 MHP); I1.' mi. N Wes-
kan (1 MHP); Wichita Co.: Ww mi. S
Leoti (1 KU); 15 mi. W Scott City
(1 KU); 17 mi. W Scott City (1 KU). '
Nebraska: Antelope Co.: 7 mi. W
Clearwater (5 NSM); 214 mi. N, 7 mi.
W Elgin (1 FMNH, 2 UMMZ); &A mi.
N, 1 mi. E Neligh (1 KU); 5V> mi. N
Neligh (1 SM); 4 mi. N Neligh (1 SM);
Neligh (1 NSM, 3 SM, 3 USNM); Vi mi.
S, % mi. W Neligh (1 KU); 1 mi. SW
Neligh (1KU); Vk mi. S Neligh (2SM);
2 mi. S, 1 mi. W Neligh (4 KU); 4 mi. S
Neligh (2SM);4M>mi. S Neligh (1 SM);
5 mi. S, 1 mi. E Neligh (1 SM); %o mi.
N, 1 mi. W Oakdale (5 KU); 1io mi. N,
1 mi. W Oakdale (2 FMNH); Mo mi. N,
IMo mi. W Oakdale (1 KU); Ho mi. N,
1-io mi. W Oakdale (2 KU, 1 UMMZ);
%o mi. W Oakdale (1 UMMZ); IMo mi.
W Oakdale (1 KU); l%o mi. W Oakdale
(2 UMMZ); V/w mi. W Oakdale (9 KU);
VA mi. W Oakdale (1 SM); 2 mi. W
Oakdale (1 SxM); Mo mi. S, 1 mi. W

Oakdale (2 FMNH, 2 UMMZ); Mo mi.
S, l-.o mi. W Oakdale (3 KU, 1 UMMZ);
Tio mi. S, 1 mi. W Oakdale (1 FMNH,
1 KU); L'io mi. S, l'-'io mi. W Oakdale
(1 UMMZ); % mi. S, 3 mi. W Oakdale
(2 FMNH, 1 UMMZ); V-i mi. S, 3% mi.
W Oakdale (2 FMNH); % mi. S, 3~i mi.
W Oakdale (1 FMNH, 1 UMMZ); (vi-
cinity of) Oakdale (1 USNM). Runner
Co.:' 10 mi. S, 2' 2 mi. E Gering (4 KSC);
no specific locality (1 NSM); Boyd Co.:

1 mi. W Bristow (2 KU); 1% mi. W
Bristow (1 KU); 2 mi. S, Vi mi. W Butte
(1 KU); 5 mi. WNW Spencer (1 KU);

2 mi. N Spencer (2 KU); 1 mi. WNW
Spencer (4 KU); M mi. N Spencer (2
KU); Brown Co.: 11% mi. N, 6 mi. E
Ainsworth (1 KU); 12 mi. N Johnstown
(1 KU); 7 mi. N Johnstown (1 KU);
61- mi. N Johnstown (2 KU); 2% mi. N,
& mi. E Long Pine (2 KSC); l1^ mi. S
Long Pine (6 KU); 23 mi. S Long Pine
(1 KSC); Buffalo Co.: 3 mi. N Kearney
(1 KSC); 2 mi. S, 2 mi. E Kearney (1
KSC); 3 mi. S, 5 mi. W Kearney (1
KSC); Cherry Co.: Hackberry Lake (3
KU); 10 mi. S Nenzel (1 KSC); 4 mi.
E Valentine (1 KU); Valentine (1 NSM);
15 mi. S, 3 mi. W Valentine (2 UIMNH);
15 mi. S, 4 mi. W Valentine (1 UIMNH);
23 mi. S, 6 mi. W Valentine (3 UIMNH);
23 mi. S, 9 mi. W Valentine (3 UIMNH);
Niobrara Wildlife Refuge (3 NSM);
Chei/cnnc Co.: 2 mi. N Dalton (1 KU);
Dalton (1 NSM); 15 mi. S Dalton (5
KU); Lodgepole (1 NSM); Sidney (2
NSM); Custer Co.: 18 mi. S Ansley (1
KSC); IS mi. S, 3 mi. W Ansley (1
KSC); 1 mi. S, 2 mi. W Broken Bow
(3 KSC); 23 mi. S, 5 mi. E Broken Bow
(1 KSC); 4 mi. N, 3 mi. E Oconto (1
KSC); Dawes Co.: Chadron (1 NSM);
10 mi. S Chadron (2 KU, 4 UMMZ);
13 mi. S, 3 mi. E Chadron (3 KU, 44
NSM); 5& mi. S Crawford (3 KSC);
Dawson Co.: Lexington ( 1 NSM); Deuel
Co.: Chappell ( 1 NSM); 4 mi. W Chap-
pell (3 KSC); Dundy Co.: 2 mi. SW
Benkelman (1 KU); 5 mi. N, 2 mi. W
Parks (6 KU); 4 mi. N, FL- mi. W Parks
(1 MHP); Franklin Co.: 1 mi. SW
Franklin (8 KU); Greeley Co.: 10 mi.

22

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

N Greeley (1 KSC); Harlan Co.: 1 mi.
S, 2 mi. W Alma (3 KSC); Hitchcock
Co.: 10 mi. S, 2 mi. E Stratton ( 1 NSM ) ;
Trenton (1 KU); Holt Co.: 6 mi. S, %
mi. W Butte (2 KU); 6 mi. N Midway
(4 KU); 24 mi. N O'Neill (1 KU); 23%
mi. N O'Neill (2 KU); Kearney Co.: 1
mi. S, 8 mi. E Kearney (1 KSC); 3% mi.
S Kearney (2 KU); 4 mi. S, 4 mi. E
Kearney (3 KSC); 4 mi. S, 14 mi. E
Kearney (2 KSC); Keith Co.: 4 mi.
WNW Keystone (1 KU); no specific lo-
cality (1 NSM); Keija Paha Co.: V> mi.
N, V-i mi. W Norden (2 KU); Norden
(1 KU); 12 mi. N Springview (2 KU);
10 mi. N Springview (1 KU); 4 mi. S
Springview (3 KU); Kimball Co.: Kim-
ball (3 NSM); Lincoln Co.: Brady (1
NSM); 1 mi. W Brady (11 KSC); 8 mi.
N, 4 mi. W Hershey (8 KU); 7 mi. N,
4 mi. W Hershey (2 KU); 6 mi. N, 2
mi. E Sutherland (1 KU); 2 mi. N, 1
mi. E Sutherland (2 KU); 4 mi. S, 2
mi. W Sutherland (2 KU); 7 mi. S, 1
mi. W Sutherland (1 KU); McPherson
Co.: 25 mi. N North Platte (2 NSM);
Morrill Co.: Bridgeport (1 NSM); Rock
Co.: 8!1> mi. N, 2% mi. W Bassett (1
KSC); 4V> mi. N Bassett (1 KU); 3% mi.
N Bassett (3 KU); H£> mi. W Bassett
(1 KSC); ZVi mi. S, 4 mi. W Bassett
(2 KU); Scotts Bluff Co.: Scottsbluff
(2 NSM); 1 mi. E Scottsbluff (3 KU);
2 mi. S, 1 mi. W Scottsbluff (2 KU); 12
mi. S Scottsbluff (2 KU); Sheridan Co.:
NW side Snow Valley, N Lakeside ( 1
NSM); Sioux Co.: 3 mi. E Agate (1
KU); 6% mi. W Crawford (4 NSM); 1
mi. S, 4 mi. W Crawford (2 NSM);
Glen (1 NSM); 8 mi. N Harrison (1
UMMZ); 5 mi. N, 2!i> mi. W Harrison
(1 KSC); Harrison (1 NSM); Thomas
Co.: 3 mi. S, 6 mi. W Halsey (1 KSC);
Valley Co.: 4 mi. N, 2 mi. W Arcadia
(2 KSC); 2V> mi. N, Y> mi. W Elyria
(3 KSC); Webster Co.: Y> mi. S, 3 mi.
W Red Cloud (1 KSC); Wheeler Co.:
1 mi. S Ericson (1 KSC).

South Dakota: Bennett Co.: La-
Creek Nat. Wildlife Refuge (1 MHP);
4 mi. S, 8 mi. E Martin ( 1 KU); 7 mi. S,
4 mi. E Martin (1 MHP); 8 mi. S Martin

(3 KU); 10 mi. S Martin (1 KU); Fall
River Co.: 1 mi. E Edgemont (2 KU);
Jackson Co.: 2 mi. S, 2 mi. E Interior
(1 KU); Todd Co.: 2 mi. N, 3 mi. W
Rosebud (2 KU); Washabaugh Co.: 10
mi. N, 4 mi. E Potato Creek (2 KU).

Wyoming: Converse Co.: 3 mi. N,
5 mi. E Orin (3 KU); Goshen Co.: 8 mi.
SSE Torrington (1 KU); Laramie Co.:
1 mi. W Gallio ( 1 KU); 6 mi. W Meriden
(2 KU); &/i mi. W Meriden (13 KU);
2% mi. SW Meriden (1 KU); Niobrara
Co.: 10 mi. N Hat Creek ( 16 KU ) ; 2 mi.
S, V-i mi. E Lusk (2 KU); Platte Co.:
3 mi. W Guernsey (1 KU); Weston Co.:
23 mi. SW Newcastle (4 KU).

Geomys lutescens major Davis, 1940

1940. Geomys lutescens major Davis, Texas

Agr. Exp. Sta. Bull. 590:32.
1947. Geomys bursarius industrius Villa-R. and

Hall, Univ. Kansas Publ., Mus. Nat. Hist.

1:226. Type from IJ2 mi. N Fowler, Meade

County, Kansas.

Type specimen. — Texas Cooperative
Wildlife Collection (Texas A&M Uni-
versity) 819, obtained 29 December 1938
at 8 mi. W Clarendon, Donley County,
Texas.

Distribution. — Occurs in southeastern
Colorado, southwestern Kansas, western
Oklahoma, east-central New Mexico, and
northwestern Texas in short-grass prairie
and mesquite prairie and in southeastern
Missouri in tall-grass prairie/ deciduous
forest mosaic (Fig. 1; see also Honeycutt
and Schmidly, 1979).

Diagnosis and comparisons. — Differs
from G. bursarius as discussed above.
Differs from G. 1. lutescens in being
smaller on average, and having propor-
tionately shorter skulls and narrower zy-
gomatic arches, but greater mastoid
breadth; frontal-premaxillary suture on
dorsum of skull falling posterior to an-
terior edge of orbits; maxillary usually
visible on sides of rostrum when viewed
dorsally ( Fig. 2D ) . Color extremely vari-
able, generally dark brown in east and
pale yellow-brown in west, but irregular
in geographic pattern. Dorsal stripe of
darkened fur sometimes present, but
rarely conspicuous to the degree seen

POCKET GOPHERS OF THE GENUS GEOMYS

23

in G. I. lutescens. Baculum long (mean
= 11.2 mm) but narrow distally (Fig. 15;
Table 8). '

Specimens examined. — Kansas: Bar-
Iyer Co.: 2 mi. E Aetna (1 MHP); Aetna
(1 MHP); 1 mi. W Aetna (1 KU); 1 mi.
S Aetna (1 KU); 1 mi. N, 19% mi. E
Coldwater (1 KU); 8 mi. N, 2 mi. E
Hardtner (3 MHP); 5 mi. N, 4% mi. E
Hardtner (1 MHP); 7 mi. N, 7 mi. W
Kiowa (1 KU); 6 mi. NW Medicine
Lodge ( 1 KU ) ; 2 mi. W Medicine Lodge
(1 KU); Vi mi. S, 1 mi. W Medicine
Lodge (1 MHP); % mi. S, 1 mi. W
Medicine Lodge (1 MHP); 1 mi. S, 1
mi. W Medicine Lodge (2 MHP); 1%
mi. S, 1 mi. W Medicine Lodge ( 1
MHP); 2 mi. S, 1 mi. W Medicine Lodge
(2 MHP); 3U mi. S Medicine Lodge
(1 MHP); 3 mi. N, VA mi. E Sharon
(2 MHP); 2% mi. N, 1% mi. E Sharon
(1 KU); Sharon (1 KU); 2V> mi. S Sun
City (1 KU); 3 mi. S Sun City (1 KU);
4% mi. S Sun City (1 KU); Clark Co.:
371.) mi. N, 12 mi. W Ashland (1 KU);
1 mi. N, 12 mi. W Ashland (1 KU); 12
mi. E Ashland ( 1 KU); % mi. E Ashland
(1 KU); l%o mi. W Ashland (1 KU);
5%o mi. W Ashland (1 KU); 6 mi. S
Kingsdown (1 KU); 7 mi. SW Kings-
down (1 KU); Comanclie Co.: 1 mi.
N, 15% mi. E Coldwater (2 KU); 1 mi.
N, 14% mi. E Coldwater (1 KU); 5 mi.
S, 11 mi. W Coldwater (1 KU); lO!^
mi. S Protection (1 MHP); Cowley Co.:
3 mi. SE Arkansas City (2 KU); Ed-
wards Co.: 1 mi. N Kinsley (1 MHP);
Kinsley (1 SM); 3 mi. E Offerle (1
MHP); Finney Co.: 2 mi. S Garden City
(1 KU); 4Yv, mi. S Garden City (2 KU);
6% mi. S Garden City (1 KU); 1 mi. S
Pierceville (2 KU); Ford Co.: 2 mi.
SW Dodge City (6 KU); 3/io mi. SW
Dodge City (1 KU); 10 mi. N, 4% mi.
W Mullinville (1 MHP); Gray Co.: 1
mi. S Cimarron (1 KU); 2 mi. S Cimar-
ron ( 1 KU); 2% mi. S Cimarron (2 KU);
4% mi. S Cimarron (1 KU); 5*4 o mi. S
Cimarron (1 KU); 6 mi. S Cimarron (2
KU); 64/io mi. S Cimarron (1 KU); 7/io
mi. S Cimarron (1 KU); 7%o mi. S
Cimarron (1 KU); Hamilton Co.: 1 mi.

E Coolidge (4 KU); 1 mi. S Coolidge
(3 MHP); 2% mi. N, % mi. W Syracuse
(3 KU); m mi. N, % mi. W Syracuse
(1 KU); Harper Co.: 1 mi. N, % mi. E
Corwin (1 MHP); VA mi. S, 1% mi. E
Corwin (2 MHP); 4% mi. N Danville
(2 KU); 1 mi. N Harper (2 KU, 1
UIMNH); Harvey Co.: Halstead (1
KU); VA mi. N, 13% mi. W Newton
(2 KU); Kearny Co.: 2/io mi. E Lakin
(1 KU); 27.<> mi. E Lakin (1 KU); ls/io
mi. E Lakin (1 KU); 4 mi. S Lakin
(1 KU); Kiowa Co.: 5 mi. N Belvedere
(1 KU); m mi. S, 3V2 mi. E Belvedere
(2 MHP); Greensburg (1 MHP); 7% mi.
S, 4 mi. E Haviland (1 MHP); Meade
Co.: 3r2 mi. NE Fowler (1 KU); 2 mi.
N Fowler (1 KU); 1% mi. N Fowler
(1 UMMZ); 7 mi. N Meade (1 KU);
8 mi. S, 2 mi. W Meade (1 MHP);
13 mi. SW Meade (6 KU); Meade
County State Park (4 KU, 1 UMMZ);
Morton Co.: 12 mi. N Elkhart (1 KU);
7% mi. N, V'-i mi. W Elkhart ( 1 KU, 2
MHP); 7 mi. N Elkhart (1 MHP); l%o
mi. N Elkhart (2 KU); 7% mi. S Rich-
field ( 1 KU ) ; no precise locality ( 1 KU ) ;
Pawnee Co.: 3%o mi. NE Larned (2
KU); Larned (1 KU); 1 mi. S, 1 mi. E
Larned (2 KU); Pratt Co.: Pratt (3
KU); Reno Co.: V-i mi. E Hutchinson
(1 MHP); Rice Co.: % mi. N, 12 mi. E
Sterling (3 MHP); Seward Co.: 10%
mi. N Liberal (1 SM); 6 mi. N, 8 mi. E
Liberal ( 1 KU); 3 mi. N, 4 mi. E Liberal
(2 KU); 1 mi. S, 1 mi. W Liberal (1
KU); Stanton Co.: 1 mi. N, 6 mi. W
Manter (1 KU); Stafford Co.: Little Salt
Marsh (1 KU).

Missouri: Carter Co.: Hunter (4
USNM); Wayne Co.: Willamsville (8
USNM).

Oklahoma: Alfalfa Co.: 3 mi. N,
6% mi. E Cherokee (1 SM); Cherokee
(1 OSU, 1 UMMZ); Great Salt Plains
Wildlife Ref. (2 OSU); Beaver Co.: 2
mi. W Forgan (1 UMMZ); Beckham
Co.: % mi. S Sayre (1 OSU); Blaine Co.:
Canton Shooting Grounds (1 OSU);
Roman Nose State Park (1 OSU); Wa-
tonga (1 OSU); Caddo Co.: 5 mi. W
Cogar (1 USNM); 1 mi. S Hinton (4

24

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

SM); 2 mi. N Hinton (1 OSU); Cana-
dian Co.: 2 mi. S, 4 mi. E Union City
(1 OSU); Cimarron Co.: 8 mi. W Boise
City (2 OSU, 3 UMMZ); Cimarron
River, N of Boise City (1 OSU); Cleve-
land Co.: 3 mi. N Lexington (2 SM);
L%o mi. N Norman (1 SM); Vh mi. E
Norman (1 SM); Norman (1 OSU);
Norman, Univ. Oklahoma campus ( 1
SM); Norman, Reeve's Park (1 SM);
Norman, intersection Timberdell and
Asp (2 SM); XA mi. S Norman (2
USNM); 1 mi. S Norman (1 SM); 2%
mi. S Norman (1 SM); 3 mi. SW Nor-
man (2 USNM); 2 mi. S Slaughterville
(2 SM); Comanche Co.: Lawton (2
USNM); Cotton Co.: 5 mi. SE Taylor
(1 SM); 8 mi. E Walters (1 UMMZ);
Custer Co.: Weatherford (2 KU); Dewey
Co.: 5 mi. W Canton (1 KU); 5 mi. SE
Vici (1 OSU); Ellis Co.: Shattuck (3
UMMZ); Garfield Co.: 8 mi. W Enid
(1 OSU); Grady Co.: ]/i<> mi. S, 3 mi.
W Blanchard (1 SM); Grant Co.: 1 mi.
N, l4/io mi. W Hawley (2 SM); Harmon
Co.: 1 mi. SW Hollis (1 UMMZ); Har-
per Co.: USDA Southern Plains Exp.
Range Sta. (4 OSU, 4 UMMZ); Buffalo
Creek on Hwy. 64 (1 OSU); 10 mi. W
Buffalo (2 OSU, 1 UMMZ); 5 mi. N
Fort Supply (1 USNM); 3 mi. N Fort
Supply (1 USNM); Jackson Co.: 5 mi.
SW Eldorado (1 UIMNH); Jefferson
Co.: 1 mi. E Ringling (2 UMMZ); Kay
Co.: Ponca Agency (3 MSU); 2% mi. S,
10 mi. W Tonkawa (1 OSU); SE M sec.
18, T26N, R2E (1 OSU); Major Co.: 5%
mi. S Waynoka (1 SM); McClain Co.:
1 mi. S, 1 mi. W Norman (ISM); Wayne
(2 OSU, 1 SM); Oklahoma Co.: Okla-
homa City (3 KU); Pawnee Co.: V-i mi.
W Cleveland (1 OSU); Pay tic Co.: V-i

mi. S, 5 mi. W Stillwater (1 OSU); 4^2
mi. W Stillwater (1 OSU); V/i mi. SW
Stillwater (1 OSU); 2 mi. S, 1 mi. W
Stillwater (1 OSU); Pottawatomie Co.:
1 mi. W Tecumseh (1 SM); Roger Mills
Co.: l7io mi. N Cheyenne (1 SM); 1 mi.
N Cheyenne (1 OSU); Stephens Co.:
M mi. E Claude (1 SM); 1 mi. S, Yi mi.
E Claude (1 SM); 1 mi. E Marlow (2
OSU, 1 SM, 1 UMMZ); Texas Co.: Guy-
mon (1 OSU, 1 UMMZ); Tillman Co.:
Frederick Cemetery (1 UMMZ); Woods
Co.: Waynoka (1 UMMZ); 5 mi. S
Waynoka (1 SM); 12 mi. N Alva (1
OSU); Alva (1 USNM); 3 mi. E Camp
Houston (1 OSU); E side Cimarron
River on Hwy. 64 (1 UMMZ); Washita
Co.: 4 mi. E Cordell ( 1 UMMZ); Wood-
ward Co.: 2 mi. NNE Woodward (1
SM); 2 mi. NNW Woodward (1 SM);
3^2 mi. WNW Woodward (1 SM);
Woodward (1 USNM).

Remarks. — In addition to the two
subspecies of G. lutescens found in our
study area, there are three other taxa
which must be considered. The form
named knoxjonesi by Baker and Geno-
ways (1975) is similar to G. I. major in
cranial moiphology, karyology (Honey-
cutt and Schmidly, 1979), and ectopara-
sites (Timm and Price, 1980) and we
consider it to be a subspecies of G.
lutescens. The status of llanensis and
texensis, which are isolated populations
on the Edwards Plateau of central Texas,
is less certain. They are karyotypically
very similar to G. /. knoxjonesi, but are
distinct cranially, and have ectoparasites
more allied with G. breviceps. Until ad-
ditional studies are carried out, we sug-
gest that they be tentatively listed as
subspecies of Geomys lutescens.

RESULTS AND DISCUSSION
SIZE VARIATION

Geographic variation in size is shown
in Fig. 3; mean condylobasal length of
adult females was used in each OTU as
our estimator of size. The largest
Geomys occur in northeastern Iowa, with

size decreasing radially in all directions;
the smallest pocket gophers in our study
area occur in south-central Nebraska and
in southeastern Oklahoma. The figure
shows that, although most size variation
is clinal, there is a geographically con-

POCKET GOPHERS OF THE GENUS GEOMYS

25

tinuous discontinuity in the cline in east-
ern Nebraska and Kansas. In this region
there is an abrupt change from less than
44 mm to greater than 46 mm; only a
single population in southeastern Kansas
is intermediate. Single specimens that
were not included in the OTU means
indicate that the clinal discontinuity is
clearly no more than 30 km wide, and is
narrower in many places. The detailed
nature of the discontinuity has been
studied in two areas. In Antelope

County, Nebraska, the change in size is
extremely abrupt, with a change from
42 mm to 47 mm taking place over a
distance of about 2 km (Heaney, 1979,
and data reported below). In north-
eastern Kansas, the width is less certain,
but is less than 30 km and perhaps as
little as 10 km ( Hendrickson, 1972). In
both of these areas, and apparently
throughout the region of the size discon-
tinuity, there is an abrupt change in
dominant vegetation from tall-grass

Fig. 3. — Map of the central and northern Great Plains showing geographic variation in mean
condylobasal length ( in mm ) of adult female Geomys; scale as in Fig. 1.

26

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

prairie to the east, and short- or mid-
grass prairie to the west, with a parallel
change from dark silt-loam soils to light
sandy loams, as discussed below.

CLUSTER ANALYSIS OF OTUS

An initial assessment of multivariate
geographic patterns in similarity of the
gophers was conducted by cluster analy-
sis of the OTUs that contained five or

more individuals (see Methods). The
resulting phenogram was then used to
group taxa into three levels of similarity;
the "similarity isoclines" are shown in
Fig. 4. Two major groups are apparent.
In the eastern group, which corresponds
to the species G. bursarius, the most dis-
tinct subgroups are the subspecies G. b.
illinoensis and G. b. luisconsinensis; a
population in eastern Iowa is less dis-

Fig. 4.— Map of the central and northern Great Plains showing levels of similarity of OTUs of
adult female Geomys based on cluster analysis; scale as in Fig. 1. For locations of numbered localities,

see "Methods and Materials."

POCKET GOPHERS OF THE GENUS GEOMYS

27

tinct. There is no tendency for the re-
maining populations to cluster into
northern and southern groups as would
be expected if the gophers in this region
should be divided into two subspecies
(G. /;. bursarius and G. b. majusculus) ,
as has often been done (e.g., Villa-R.
and Hall, 1947; Jones, 1964; Bowles,
1975; Hall, 1981).

The western group corresponds to
the species G. lutescens and G. breviceps
as defined in this study. The "similarity
isoclines" show the populations of G.
lutescens to be fairly homogeneous; the
gophers from southwestern Kansas and
adjacent Oklahoma (G. 1. major) did not
form a distinct group in the analysis.
Geomys breviceps was clearly most simi-
lar to G. lutescens, but was distinct at all
but the highest levels of clustering.

DISCRIMINANT FUNCTION
ANALYSES

These analyses were first carried out
to test the null hypothesis of no differ-
ence between a large number of OTUs,
all of which were considered simultane-
ously. In the initial analyses, all OTUs
having five or more specimens were con-
sidered as distinct groups and differences
were sought. For convenience, analyses
were done on four geographic units;
these were 1) Colorado, Kansas, and
Missouri; 2) Nebraska, southern South
Dakota, and Wyoming; 3) Illinois, Indi-
ana, Iowa, Minnesota, Missouri, North
Dakota, eastern South Dakota, eastern
Nebraska, and eastern Kansas; and 4)
Arkansas, Missouri, and Oklahoma.

Kansas region. — Analysis of the first
region indicated that all OTUs from
southwestern Kansas overlapped greatly
and were not distinguishable from one
another (Fig. 5). These populations had
previously been divided into three sub-
species (industrius, jugossicuhris, and
major) (Villa-R. and Hall, 1947), but
they all may be referred to G. I. major
(see below). They were distinct from
most, but not all, northwestern Kansas
OTUs, and were highly distinct from all

groups from northeastern Kansas and
Missouri. The northwestern OTUs, which
overlapped greatly among themselves
(Fig. 5), are not distinguishable from
gophers from western Nebraska (see be-
low), and are considered to be G. I.
lutescens. The OTU from east-central
Missouri was not distinguishable from
eastern Kansas OTUs, but was distin-
guishable from all others. The eastern
OTUs in Kansas were highly distinct
from the two western groups; the east-
ern group, and also the Missouri OTU,
are part of G. bursarius as defined here.
In the second Kansas region analysis,
small samples and specimens from the
areas of contact between the taxa were
entered into the analysis as unknowns,
and compared with the aggregated sam-
ples of G. /;. bursarius, G. 1. lutescens,
and G. I. major. The specimens from
Colorado identified as G. I. lutescens also
were used as a group of "knowns," as
was the sample of gophers from eastern
Missouri. The Missouri sample was not
significantly different from the Kansas
G. bursarius sample (F = 1.49, p > .05),
but all other groups entered as known
could be significantly distinguished.
However, 21 of 100 G. /. lutescens from
Kansas were misclassified with the Colo-
rado group, and eight of 50 from Colo-
rado were misclassified as being from
Kansas, thus indicating their high level
of similarity.

Villa-R. and Hall (1947:231) stated
that it was the ". . . intermediate nature
of . . . specimens from Butler County
and . . . McPherson County, Kansas, that
have caused us to treat G. b[ursarius]
majusculus ... as only subspecifically
distinct from the more western subspe-
cies, [G. lutescens] major . . . ." One
Butler County and all but one McPher-
son County specimens were juveniles,
and so not usable by us. The remaining
two specimens were entered into our
analysis as unknowns, and were assigned
to G. bursarius with posterior proba-
bilities in excess of .95; they also had
all qualitative cranial characters asso-
ciated with G. bursarius, and we con-

28

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

sider them to be typical representatives
of that species.

Hendrickson (1972) suggested that
specimens from the vicinity of Osborne
County, Kansas, represent intergrades
between lutescens and bursarius. We
entered three females from western Os-
borne County, two from eastern Osborne
County, one from Jewell County, and
two from eastern Russell County as un-
knowns in the second analysis. Speci-
mens from western Osborne and Russell
counties were identified as G. lutescens
with posterior probabilities in excess of
.95. Those from eastern Osborne County
and Jewell County were identified as G.
bursarius with posterior probabilities of
.92, .95, and 1.00, respectively. Males
from eastern Jewell and western Osborne
counties were included in an analysis
conducted in the same fashion using

adult males. The one male from Jewell
County was assigned to G. bursarius
with a .99 probability, and the three
from western Osborne County were as-
signed to G. lutescens with probabilities
in excess of .95.

Additionally, we entered females
from Barton ( 1 ) , Ellsworth ( 1 ) , Rice
(1), Pratt (2), and Cowley (3) counties
as unknowns; all but the one from Rice
County were assigned to G. lutescens
with probabilities in excess of .95; the
one from Rice County had a probability
of .81. Males from Stafford (1), Barton
(1), Harvey (2), Rice (2), Reno (1),
Harper ( 3 ) , and Cowley ( 1 ) , were anal-
yzed similarly and were all identified as
G. lutescens with posterior probabilities
in excess of .95.

The cluster analysis and discriminant
function analyses discussed above indi-

2.5

II

0.0

■2.5

44-

46-
48-

47-
45

-6.0

I

Fig. 5. — Graph of the first two axes of the discriminant function analysis of the Kansas region OTUs.

Polygons on the left are G. bursarius, in the upper right are G. /. lutescens, and in the lower right

are G. /. major. Numbers correspond to OTU numbers; see text.

POCKET GOPHERS OF THE GENUS GEOMYS

29

cated that gophers in the species G.
lutescens fall into two groups for which
the names lutescens and major are avail-
able. In order to test for evidence of
intergradation between the two taxa,
we entered all gophers from the area of
contact (as defined by Villa-R. and Hall,
1947, and the results of our cluster anal-
ysis shown in Fig. 4) as unknowns in the
second-level discriminant function anal-
ysis described above. The results indi-
cated that gophers from the floodplain
of the Arkansas River in Kearny and
Finney counties in west-central Kansas
and in Barton, Cowley, Harper, Harvey,
and Rice counties in south-central Kan-
sas (a zone 40-50 km wide) showed in-
dications of intergradation between lu-
tescens and major, and it is on this basis
that we consider them to be conspecific.
The assignment of a small number of
these individuals is arbitrary in that
the probability of belonging to either
taxon is roughly equal (e.g., specimens
from Barton and Harvey counties), and
in some cases individuals from the same
locality (e.g., specimens from Hamilton
and Harper counties) could be assigned
to different subspecies. We interpret this
as evidence of a fusion type of hybrid
zone (Heaney, 1979) in which selection
for the two genomes has occurred at dif-
ferent intensities in different localities.
It would appear that part of the lutes-
cens genome extends into southern
Kansas from Barton through Harper
counties, but because the majority of in-
dividuals are most similar to major, we
have assigned all specimens to that taxon.
The subspecies G. I. jugossicularis
was named on the basis of two speci-
mens from Lamar, Prowers County,
Colorado (Hooper, 1940); nine speci-
mens from Morton County, Kansas, were
also referred to that subspecies by
Hooper. The Morton County specimens
were entered into the first discriminant
function analysis as knowns in a single
OTU; in that analysis, all were nearly
identical to those gophers to the east
which had been included in the subspe-
cies major and industrius. The paratype,

an adult female, and a young adult fe-
male from 1 mi. S Lamar were entered
into the second-level discriminant func-
tion analysis as unknowns. They were
both assigned to G. I. lutescens with pos-
terior probabilities of .767 and .749, re-
spectively. Two females from 4 mi. W
Las Animas, Bent County, Colorado,
were also assigned to G. /. lutescens with
probabilities in excess of .85. Although
few specimens are available from south-
eastern Colorado, it appears that G. I.
lutescens extends south at least as far as
Lamar and Las Animas in Colorado. In
spite of evidence that some intergrada-
tion does occur in the area (e.g., the
specimen from El Paso County, which
had a posterior probability of .691 of
belonging with G. /. major, and the two
from 1 mi. S Lamar), and in spite of the
fact that most of the specimens referred
to this taxon in the past are consub-
specific with those in southwestern Kan-
sas, the name jugossicularis must be re-
garded as a junior synonym of G. I.
lutescens because specimens from the
type locality are most similar to that
taxon.

As noted by McLaughlin (1958), the
populations he named as G. b. missouri-
ensis represent a zoogeographic enigma.
One of the two known main popula-
tions occurs primarily on the floodplain
and river bluffs of the Missouri River
near its confluence with the Mississippi
River. The other population occurred
along a single railroad right-of-way in
the Ozark uplands of Carter and Wayne
counties in southeastern Missouri; Mc-
Laughlin searched for them there 60
years after their discoveiy, and con-
cluded that they were extinct. Although
G. /;. illinoensis occurs across the Mis-
sissippi River immediately to the east
of the St. Louis area, we suspect that
the river forms a very rarely crossed
barrier because we found no evidence
of interbreeding between the two taxa.
The St. Louis area population is about
150 km from the nearest known popu-
lation of G. b. bursarius. The southern
population was about 120 km from the

30

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

St. Louis gophers (although new speci-
mens from Crawford and Perry counties
narrow this apparent gap), 175 km from
the nearest known population of G.
breviceps sagittalis in Arkansas, and 200
km from the nearest populations of G.
lutescens major in northeastern Okla-
homa. Our analyses indicated that seven
females from the St. Louis area popula-
tion were not significantly distinguish-
able from G. b. bur sarins in Kansas, and
we can see no reason to recognize this
population as subspecifically distinct (see
further discussion below). One adult
male from St. Louis County, two from
Carter County, and one from Wayne
County were available for study. These
were entered as unknowns in the dis-
criminant function analysis of males. The
results indicate that the St. Louis speci-
men may be assigned to G. bursarius
(posterior probability = 1.00), whereas
those from Carter and Wayne counties
are most likely G. lutescens ( probability
= 1.00, 1.00, and .898, respectively), and
are closest to G. I. major. In the analysis
of Oklahoma gophers (discussed below),
these males were all assigned to G. I.
major, with probabilities of .99, .741, and
.581, rather than to G. breviceps from
Oklahoma or G. bursarius from St. Louis.
On this basis we have assigned the Car-
ter and Wayne County specimens to
G. lutescens, and have tentatively listed
them as G. I. major. If this population
is extinct, as suggested by McLaughlin,
it may not be possible to assign these
gophers to any taxon with any greater
degree of assurance.

Nebraska region. — Pocket gophers
from Nebraska, South Dakota, and Wy-
oming show considerable variation in
body size and pelage coloration, and this
has led to the suggestion that various
populations should be considered sub-
specifically distinct (Blossom, 1938;
Swenk, 1939, 1940). Each nominal sub-
species was considered to be quantita-
tively distinct from the "typical" form
of the Sandhills region in north-central
Nebraska, and was thought to be con-
fined to a certain soil and vegetation

type. To test for patterns in variation of
these pocket gophers, we entered all
OTLls having five or more specimens
(Table 1) as knowns in a discriminant
function analysis. The results (Fig. 6)
showed two groups; one of these, re-
ferable to G. bursarius, was composed of
individuals found in eastern Nebraska,
among which no subgroups could be
distinguished on the basis of F-statistics
and jackknife classification matrices.
Specimens from a hybrid zone in central
Nebraska (described in detail by Hea-
ney, 1979) are discussed below. The
absence of overlap between the two
groups is taken as evidence of a lack of
significant introgression into either spe-
cies.

We found no distinct groups of OTUs
within G. lutescens in the Nebraska re-
gion. There is a general trend for de-
creasing size from north to south and
west to east (Table 1, Fig. 3), and al-
though OTUs from opposite ends of this
region are distinguishable, none is distin-
guishable from its neighbors, and all
variation seems to be in the form of
gradual clines.

Although we did not quantify color-
ation, some qualitative assessments and
judgments are possible. As noted by
Jones ( 1964), pocket gophers from north
of the Niobrara River in Boyd and Key a
Paha counties are darker (especially in
having a large, dark mid-dorsal stripe)
than those from the Sandhills region;
they are not paler, as was stated by
Swenk (1940). Those from immediately
south of the Niobrara River in Brown
and western Rock counties are dark also,
as are those from south of Chadron in
Dawes County, Nebraska (Swenk, 1940).
In both regions the color change is rapid,
defined by the Niobrara River in Boyd
and Holt counties, and by Bone Creek
in Brown County. The color change is
correlated with change in habitat from
mixed grass prairie on sandy loam soil
to sandhills prairie on sandy soil in Boyd,
Brown, Holt, and Keya Paha counties,
and from mixed ponderosa forest-short-
grass prairie on dark sandy loam to short-

POCKET GOPHERS OF THE GENUS GEOMYS

31

grass prairie on sandy soil in Dawes
County (Kaul, 1975; Swenk, 1940). A
correlation between coat color and soil
color has been previously documented
in Geomys ( Hendrickson, 1972), and we
feel that this variation, and other, less
dramatic color variation in the species
elsewhere, is probably the result of local
adaptation and carries little or no in-
formation about gene flow. Because we
did not find concordant differentiation
in cranial morphology, and in fact have
evidence of broad clines in morphology,
we have chosen to follow Jones ( 1964 )
in recognizing a single subspecies of
pocket gopher (lutescens) in the western
portion of Nebraska and adjacent areas.
Jones ( 1964 ) suggested that a broad
zone of intergradation between bursarius
and lutescens exists in Adams, Antelope,
Boyd, Buffalo, Harlan, Holt, Kearney,
and Webster counties on the basis of
east-west clines in color and size; cranial
morphology was not considered. He be-

lieved that the zone of intergradation
widened towards the south, especially
south of the Platte River. Lack of inter-
gradation in Kansas, as was demon-
strated above, puts his observation in
doubt, and Fig. 3 shows that the size
cline is discontinuous. We tested for
intergradation in Nebraska by running
a second discriminant function analysis
in which all specimens from the above-
named counties, plus those from Frank-
lin and Wheeler counties, were entered
as unknowns for comparison with a
group of G. bursarius and of G. lutes-
cens, each composed of all specimens of
the respective taxa from outside the
putative zone of intergradation. All
specimens from Boyd, Buffalo, Custer,
Dawson, Franklin, Harlan, Kearney,
Kaya Paha, Valley, Webster, and
Wheeler counties, and those from west-
ern Antelope County, were identified as
G. lutescens with posterior probabilities
in excess of .99. Those from Adams (1),

2.5 -

II

pIGi 6. — Graph of the first two axes of the discriminant function analysis of the Nebraska region
OTUs. Polygons on the left are G. bursarius bursarius, and on the right are G. lutescens lutescens.

Numbers correspond to OTU numbers; see text.

32

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Knox (3), and Platte (1) counties were
identified as G. bursarius with posterior
probabilities of 1.00, .74, .99, .98, and .63.
None of these specimens had qualitative
features which characterized the hybrids
in Antelope County (Heaney, 1979).
Males from the contact area were anal-
yzed similarly, and all specimens (Boyd,
Franklin, Greeley, Harlan, Holt, Kear-
ney, Keya Paha, and Valley counties)
except one were assigned to G. lutescens
with posterior probabilities in excess of
.95. One specimen from Boyd County
was assigned to G. lutescens with a
probability of .51, but another from one
mile away was assigned to G. lutescens
with a probability of 1.00; the one "inter-
mediate" specimen was a large, old male
with qualitative features characteristic
of G. lutescens. Also included in the
analysis were males from the hybrid zone
described by Heaney ( 1979 ) . Specimens
previously identified as G. lutescens were
assigned to G. lutescens with probabil-
ities over .95; five specimens previously
identified as G. bursarius were identified
as G. bursarius with probabilities of .59,
.89, .95, .99, and 1.00. One specimen
from Antelope County identified by
Heaney (1979) as a hybrid was assigned
to G. bursarius with a posterior proba-
bility of .81 (see also below).

We conclude that in Nebraska intro-
gression between the two species of
Geonujs is so low that it cannot be
measured by the morphometric tech-
niques employed here, or is absent alto-
gether. However, there are few speci-
mens from many parts of the potential
zone of contact, and further studies are
warranted.

Northeastern region. — In the third
series of discriminant function analyses,
all OTUs from Illinois, Indiana, Iowa,
Minnesota, North Dakota, eastern South
Dakota, and Wisconsin were entered as
knowns, and those from Kansas, Mis-
souri, and Nebraska previously identi-
fied as G. bursarius were entered as
knowns also. The results (Fig. 7) indi-
cated that the Indiana and Illinois sam-
ples, although not distinguishable from

each other, were distinct from all other
groups. These gophers were named as
a subspecies of G. bursarius (G. b. illi-
noensis) by Komarek and Spencer (1931).
Because they are isolated from para-
patric populations by the Mississippi
River (Fig. 1), there is little to no op-
portunity for gene flow; only one G. b.
illinosnsis was misclassified as belonging
in the St. Louis area OTU, and the con-
verse is true, also. G. /;. illinoensis was
also found to be distinct from Wisconsin
pocket gophers, which also occur east
of the Mississippi River; the two are
separated by 300 km and several large
rivers, and show no evidence of close
relationship.

The Wisconsin OTU, referable to
G. /;. wisconsinensis Jackson, 1957, was
nearly as distinct as G. /;. illinoensis. The
F-matrix indicated significant differences
from all other OTUs, and one of the
eleven specimens was misclassified as
belonging with the northwestern Mis-
souri OTU. This level of morphological
distinctness contrasts with the absence
of apparent karyological differences ( dis-
cussed below). Females from Chippewa
(1) and Polk (2) counties were entered
as unknowns in a discriminant function
analysis in which all G. /;. bursarius,
G. b. illinoensis, and G. b. wisconsinensis
were grouped as three units. The speci-
mens from Polk County were assigned
to G. b. bursarius with a probability of
.81 and .99, and the one from Chippewa
County was assigned to G. b. wiscon-
sinensis with a probability of .54. Two
males from Trempealeau County were
both assigned to G. b. bursarius with a
probability of .98. The intermediate po-
sition of the specimen from Chippewa
County may indicate some introgression.
Nevertheless, we retain G. b. wiscon-
sinensis as a valid subspecies because
the qualitative and quantitative differ-
ences suggest that some selective forces
have acted on it separately from other
members of the species, perhaps during
a period of allopatry during the late
Pleistocene (see below). Studies in
northwestern Wisconsin, where G. b.

POCKET GOPHERS OF THE GENUS GEOMYS

33

bursarius and G. b. wisconsinensis meet,
may clarify the relationship between the
two.

All other OTUs included in this anal-
ysis overlapped with each other exten-
sively, and variation appears to be clinal
(Fig. 7); for example, OTUs from Min-
nesota and Kansas are distinguishable,
but they both overlap with OTUs from
Iowa and eastern Nebraska. Division of
this large population into subspecies
would be arbitrary and would not impart
any information about relationships, so
we have chosen to recognize all OTUs
as belonging to a single subspecies, G. /;.
bursarius. This is contrary to Swenk
( 1940), Villa-R. and Hall ( 1947), Bowles
(1975), and Hall (1981) who divided
the gophers into two subspecies, G. h.
bursarius and G. b. majusculus. We also
include G. b. missouriensis in G. b. bur-
sarius. As discussed above, as originally
defined, this subspecies was composed

of populations of two species. The north-
ern populations near St. Louis are not
distinguishable from pocket gophers
from eastern Kansas. Because the holo-
type is from St. Louis, G. b. missouriensis
should be regarded as a junior synonym
of G. J), bursarius. The southern popu-
lations, from Wayne and Carter counties,
are here referred to G. 1. major, as dis-
cussed above.

Oklahoma region. — Discriminant
function analysis of gophers from Okla-
homa, Arkansas, and Missouri indicated
the presence of three principal groups
(Fig. S). Gophers from eastern Okla-
homa (OTUs 79 and 80) are clearly
distinct from those elsewhere in the re-
gion, and the specimens from adjacent
counties (OTU 78) were indistinguish-
able from these. These specimens are
referable to G. breviceps sagittalis
(Honeycutt and Schmidly, 1979). Go-
phers from central and western Okla-

5.0

67

61 45

2.5-

70-

68-

II o.o

-2.5

50

48
66

-5.0

49 43

44.46,47, 51,54, 56,59,
60, 63, 64, 65

-6.0

-3.0

0.0

3.0

6.0

I

Fig. 7. — Graph of the first two axes of the discriminant function analysis of the Northeastern re-
gion OTUs. Polygons 67, 68, and 70 represent G. b. illinoensis; 66 represents G. b. wisconsinen-
sis; and all others are G. b. bursarius. Numbers correspond to OTU numbers; see text.

34

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

honia formed a second cluster of widely
overlapping groups (OTUs 72, 73, 75,
and 76); these are referable to G. lu-
tescens major. A final cluster is formed
by gophers from northern and eastern
Missouri (OTUs 49, 53, and 54). Al-
though this last cluster falls close to
G. 1. major on the graph, no specimens
from either group were misidentified as
belonging to the other.

Baker and Glass (1951) studied a
series of 31 Geomys from Pottawatomie
and Cleveland counties, Oklahoma, and
concluded that the two characters used
previously to distinguish between major
and "dutcheri" ( = sagittalis) (namely,
length of hind foot and relative length
of dorso-lateral exposure of the jugal)
showed evidence that the two taxa inter-
graded. They also noted a line of con-
tact between the taxa running from Paw-
nee through Lincoln, Cleveland, Potta-
watomie, and McClain counties. To test

their hypothesis of widespread intergra-
dation, we conducted a discriminant
function analysis in which gophers from
the OTUs mentioned above were entered
as three groups of knowns, and all speci-
mens from the following counties in
central Oklahoma were entered as un-
knowns: Canadian, Cotton, Creek, Gar-
vin, Grady, Jefferson, Kay, Lincoln,
Love, McClain, Okfuskee, Oklahoma,
Payne, Pottawatomie, Stephens, and
Tulsa. We found evidence for "intergra-
dation" or close contact only in Cleve-
land, Grady, McClain, Oklahoma, Payne,
and Pottawatomie counties. In Cana-
dian, Cotton, Jefferson, Kay, and Ste-
phens counties, all gophers were assigned
to G. 1. major with probabilities in ex-
cess of .90, and in Creek, Garvin, Love,
Okfuskee, and Tulsa counties all were
assigned to G. breviceps with proba-
bilities in excess of .85. The two speci-
mens from Lincoln County were identi-

5.0

2.5

II

0.0

Fig. 8. — Graph of the first two axes of the discriminant function analysis of the Oklahoma region

OTUs. Polygons on the right are G. hreviccps sagittalis, those in the upper left are G. lutescens,

and those in the lower left are G. bursarius. Numbers correspond to OTU numbers; see text.

POCKET GOPHERS OF THE GENUS GEOMYS

35

fied as G. I. major, but both are large,
old adults (one male and one female)
and have qualitative features typical of
G. breviceps. We tentatively refer these
to G. breviceps, but recognize the need
for additional information.

Our analysis included all adults from
Cleveland and Pottawatomie counties
used by Baker and Glass ( 1951 ) . Un-
fortunately, many they included were
not full adults, and so could not be
utilized here, but some additional speci-
mens were available to us. We found
that, contrary to statements by Baker
and Glass (op. cit. ), three adult speci-
mens from Pottawatomie County are
clearly identifiable as G. lutescens (fe-
male from 1 mi. S Tecumseh, prob. =
1.00) or G. breviceps (two males from
Asher, prob. = .98, .99). These are from
two distinctly different soil and vegeta-
tional types. The G. lutescens are from
an extensive area of oak-hickory savan-
nah on light-colored sandy soil, whereas
G. breviceps are from riverine forest and
saturated floodplain along the Canadian
River. In Cleveland County, our exami-
nation of 24 females and 15 males from
the vicinity of Norman revealed that
there is an abrupt contact between G.
lutescens and G. breviceps. Specimens
assigned to G. breviceps and G. lutescens
meet in an apparent zone of intergrada-
tion extending from central Norman to
about three miles east of Norman; pos-
sible hybrids are from VA mi. E Norman
(SM 10835), 2.85 mi. E Norman (SM
10842), and V> mi. S Norman (USNM
263479). Unfortunately, we have too
few specimens to document the north-
south extent of the zone. Hart (1978)
found gophers with karyotypes typical
of G. breviceps at 4 mi. E Norman, and
typical G. I. major in Norman, further
identifying this as a contact zone. The
vicinity of Norman was originally tall-
grass prairie on dark, clayey soils, with
an oak-hickory savannah on sandy soil
a few miles to the west (Gray, 1959), but
agriculture and urbanization have un-
doubtedly had a major impact. At pres-
ent, we feel that such a narrow zone in-

dicates either a recent contact between
the taxa, which seems unlikely, or past
selection against introgression. Detailed
studies currently underway by Zimmer-
man (pers. comra.) should clarify the
dynamics and significance of the hybrid
zone.

From Payne County, to the north of
Cleveland County, we examined nine
adult females and six adult males, all
from near Stillwater. There appears to
be an abrupt contact between G. lutes-
cens and G. breviceps just to the west
of Stillwater, with possible hybrids from
4 mi. N (OSU 497) and 2 mi. S (OSU
8407) of Stillwater. Another apparent
hybrid (prob. of belonging to G. brevi-
ceps of 58%) is from 1 mi. W Perkins
Corner (OSU 2594), which is 10 miles
south of Stillwater. All of those west of
Stillwater are typical G. lutescens, and
those to the east are G. breviceps, but
our sample sizes are too small to allow
definitive conclusions. Like Norman,
Stillwater originally was tall-grass prairie
on dark, clayey soil, but it has been
greatly disturbed by human activity.
Also like Norman, a pocket of light sandy
soil that once supported oak-hickory sa-
vannah occurred nearby (to the south
and west).

A third contact area between G.
breviceps and G. /. major exists in Okla-
homa County. Two specimens from 2
mi. E Tinker Field (adjacent to Okla-
homa City) both have .99 probability of
belonging to G. breviceps, whereas three
from Oklahoma City were assigned to
G. /. major (prob. = .99, .99, and .92).
No specimens from intervening areas are
available.

In McClain County, two specimens
from 2 mi. W Byars and Rosedale were
assigned to G. breviceps (prob. = .94
and 1.00, respectively), whereas a speci-
men from about 10 miles to the west at
Wayne was assigned to G. I. major
( prob. = .99 ) ; all were in dark, clayey
soils under tall-grass prairie, but several
other soil types occur very close by.
Finally, in Grady County, a specimen
from riverine floodplain at 1 mi. W

36

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Chickasha was assigned to G. breviceps
(prob. = .98), whereas a specimen from
about 15 miles away in a sandy soil /oak-
hickory savannah association (Gray,
1959) at Vi mi. S, 3 mi. W Blanchard was
assigned to G. /. major (prob. = 1.00).

The above soil and vegetation desig-
nations were taken from a map prepared
by Gray (1959). When the various lo-
calities from which Geomys have been
taken in Oklahoma are plotted, it is ap-
parent that G. I. major is associated with
light-colored, moderately sandy, well-
drained soils underlying short-, mid-, or
tall-grass prairie. G. breviceps sagittal is,
on the other hand, is associated with
dark-colored clayey or moist riverine
soils under oak-hickory savannah or tall
grass. G. breviceps apparently is en-
tirely absent from the Ozark and Oua-
chita Highlands where oak-hickory-pine
savannah and forest predominate. The
two taxa do show some ability to move

4.0 -

into "atypical" habitat away from the
general area of contact. In the Payne to
Grady counties contact zone, interdigi-
tation of soil and vegetation types cre-
ates an extremely complex pattern. In
spite of this, affinity of the two taxa to
their usual soil type is moderately strong,
with G. /. major nearly always found in
or near extensive sandy soils and G.
breviceps in or near darker soil that
supports some trees, as well as tall-grass
prairie. The zone of contact meanders
due to interdigitation of soils, but where
specimens are available, the potential
zone of introgression appears to be no
more than five miles wide, and may be
much less. These data all indicate to us
that G. breviceps and G. /. major are eco-
logically and evolutionarily independent
of one another, and should be recognized
as distinct species. Studies of chromo-
somal variation support this hypothesis
(Honeycutt and Schmidly, 1979; Tucker

II

-4.0

FIG- 9. — Graph of the first two axes of the discriminant function analysis of all taxa. 1 = Geomys
bursarius bursarius, 2 = G. bursarius wisconsinensis, 3 = G. bursarius illinoensis, 4 = G. breviceps
sagittalis, 5 = G. lutescens lutescens, 6 = G. lutescens major. Elipses enclose approximate 95% con-
fidence limits. For classification functions, see Table 4.

POCKET GOPHERS OF THE GENUS GEOMYS

37

and Schmidly, 1981), and current testing
by Zimmerman (pers. eomm.) should
clarify the matter still further.

As mentioned above, gophers from
Arkansas were also included in this anal-
ysis. We have only three adult females
and nine males available from a total of
four counties, however, so our results
must be considered to be tentative.

One male and one female from Craw-
ford County, which is adjacent to Okla-
homa, were assigned to G. breviceps
with probabilities of 1.00 for both. Two
males from Saline County, in central
Arkansas, also were assigned to G. brevi-
ceps (prob. = 1.00, .83). However, in
Pulaski County, which is immediately to
the east of Saline County, one male was
assigned to G. breviceps (prob. = .78),
four males (prob. = .83, .99, .95, .99)
were assigned to G. I. major, and one
female and one male were essentially
intermediate (prob. of belonging with
G. breviceps — .55 and .47, respectively).
Our one specimen from Ouachita County
was a young male, but seemed typical
of G. breviceps. We consider it possible
that the population of G. 1. major from
southern Missouri once extended to Pu-
laski County, but we feel that much more
evidence is needed before this can be
stated with confidence. For now, we
list all Arkansas gophers as G. breviceps
sagittalis as a matter of convenience, but
emphasize that this is a tentative assign-
ment in need of further investigation.

As indicated in the Kansas region
analysis, gophers from Missouri fall into
two species. All gophers from northern
Missouri and the St. Louis area appear
to be typical G. b. bursarius, and three
specimens from south of St. Louis in
Crawford, Franklin, and Perry counties
were assigned to G. bursarius (prob. all
in excess of .95). The populations from
Carter and Wayne counties were as-
signed to G. I. major, as discussed above.
In this case there is little evidence
against assigning them to G. I. major,
and we have done so, but further studies
are warranted.

Discriminant functions analysis of

CD
to

x

71

c

CD

-D.

o
a

•4-J

a
o
o

a

a

o

a
en

o

U §

93 3

03 *0

£ 0
CD

H3 CD

<a c

c

CD

c

OS
CD

U-i
O

a

c

>

CD

03

nr}

C

cc!

+1

a

a

CD

co
w

<

CO

C

c»

c_>

V.

CO

.©

co

c»

5

CO

s

—

i-H

r-*A

R

u

CO

r*

CO

co

CJ

CO

O

CO

i— 1

co

CM

"^

Sj

CO

1^.

"1

u

CO

rs

Cj"

&s

_CO

Q

'S

■+*»

00

co

■t^

cm

-S

Cud

43

u

03

J3

CO

5^

Co

Q

e

co

co

cm

S

a

•*

-ja

js

O

Co

o?

CO

_*3

fj

"C

co

Q

g

S*

**»»

i— i

te

c-

1— 1

-a

o

CO

U

,co

CO

^

&5

'w

3

a

£

i-H

2

a

oo

3

£

1— 1

-£>

hC

d

z

cqwcqcor-joicoocowco
CO t^i-Hr-Hi— ioooooo

Ti +i ii +i ii ii ii ii ii ii ii

t-ococoffioooH
ffiin^ocNCrHooioinoo

COr-idcNicDrtcicOHTfd
CO l> CO XT CM CM i— I

cm

(DCOCiHrf^oocO
CO CD b C? O Q in 1/5 ^f CD -t

(Nt-^i— iHHdcidddd

II II II II II II II II II II II

TfcicDcoonooo

COqCDOCOOlOWMMT}"

d d d oi d ■* oi o i-l rf d
co i^ co -^ cm cm i—i

CM

Ot-<MNbCDW(M

TFinOcOOCCDCOCOCOCDTf

Gibpinddddddd

II II II II II II II II II II II
t- t> ■* i> in ■>+ cm co

©l/5b;l/5b;b;CDCDnCDM

'^cdcdcJTt'cModcMi— i co ©

■—I CO (M CO CM CM i—l
CM

CO^COOON'I'N'f
<M "* © ■-! P 00 <M in CO rt] CO

in i-~ ih i-i t-i o o © © © o

II II II II II II II II II II II

COCOCOCO©^^-^t,
O ■*+< Tfi CO 00 Tf_ © TF © O 00

cS— n'cMcdodt^cJincMCD©

CO 00 CO -^ CM CM i-H

CM

COOOCBO'fOO

-* in co in i> i> i-H cm co p -*

ndHHOo'ddddd

i^ II II II II II II II II II II II

i— ICM^i-HCOCO-^O

cococoi— icooomi— ;i>coin
cJ^'c^b-'odincJcdcMino

CD oo CO Tt1 CM CM --I

CM

COO)CMincOl>Cj5tM
OOTOOcOTT^ppt^"^

t> »' CM cm' rl rH O © © O O

II II II II II II II II II II +1

CMNOhOCDCOCO
MinCOTt-t-OOQOSCOh;

co^cdt-'odcDCJcdi-HincJ

CD I— CO -tf CM CM i-l

CM

c« —

J 03 ■£

«^ s odd.

CO?'*-1
CD [£, ^i 03 T3

l~l rrt S "8

CO M _0

CD
CD

CD

■—

CD

rSS

$£ -

CD CT _

J- oo cy: *- ' >•
CQ _^ _ 5

M

C i«

CD r>

CO

§3

hhKU n'S ZfcfcOS

38

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

taxa. — A final discriminant function anal-
ysis was run on all members of the six
taxa; this included 711 specimens (means
in Table 3), and excluded those from
contact areas (92 specimens). Scatter
along the first two axes is shown in Fig.
9, and classification functions for the
first axis are given in Table 4. The jack-
knifed classification matrix indicated that
only 11 out of 454 (2.4%) G. lutescens
were misclassified as G. bursarius, and
13 (2.9%) were misclassified as G. brevi-
ceps. Of 234 G. bursarius, 16 (6.8%)
were misclassified as G. lutescens, and
none as G. breviceps. However, 10.7%
and 19.6% of the G. bursarius and G.
lutescens, respectively, were misclassi-
fied as to subspecies within the proper
species. These data reflect the higher
level of morphological distinctness be-
tween than within species.

CLUSTER ANALYSIS OF THE TAXA

A final quantitative analysis of mor-
phological similarity was conducted us-
ing a cluster analysis of all taxa ( Fig.
10), based on the results of a principal
components analysis (see Methods).
The analysis indicates that two major
groups are present. The first group con-
sists of the three subspecies of G. bur-
sarius, with G. /;. wisconsinensis being
the most distinct of the three. The sec-
ond group contains both G. breviceps
and G. lutescens. The analysis indicates
a roughly equivalent level of distinctness
in the pairs G. I. lutescens-G. I. major
and G. b. bursarius-G. b. illinoensis.

Table 4. — Pairwise classification functions for
all adult females of taxa that are parapatric.
Variable names are: 1= total length; 2= tail
length; 3= hind foot length; 4 = condylobasal
length; 5= zygomatic breadth; 6= mastoid
breadth; 7= nasal breadth; 8= frontal square
length; 9= frontal square width; 10= orbital
length; 11= maxillary visibility.

1. G. h. bursarius vs. G. h. wisconsinensis
X- - 0.096(V,) + 0.535(V-) + 0.209(V3)

- 0.827(V.) - 0.675(V,) + 1.686(V.)
+ 3.324(V7) + 1.448(V8) - 6.507(V0)
+ 1.523(V10) + 0.223(Vix) - 21.995.

2. G. b. bursarius vs. G. b. illinoensis

X= 0.380(V,) - 0.184(V2) + O.G39(V3)
+ 0.032(V*) - 0.861 (V.-,) + 0.315(V8)

- 0.151 (V7) -- 4.583(V8) - 1.729(V,)
+ 0.299(Vio) - 0.758(V„) + 20.427.

3. G. b. bursarius vs. G. I. lutescens

X = 0.0156(V,) - 0.094(V2) + 0.232(V3)
+ 2.968(V4) - 0.736(V,) + 0.223(V6)

- 3.859 (V7) + 0.059 (Vs) + 0.350(V.)
+ 0.493(V10) + 0.107(Vn) - 93.689.

4. G. b. bursarius vs. G. I. major

X = 0.039(V.) - 0.156(V2) + 0.316(V3)
+ 3.361 (V4) -- 0.123(Vn) - 1.440(V«)

- 1.780(V7) - 0.663 (V„) + 0.549(V0)

- 0.446(V,„) - 1.087(V„) - 92.250.

5. G. I. major vs. G. I. lutescens

X = - 0.024(V,) + 0.065(V2) - 0.084(V3)

- 0.393(V4) - 0.613(V,) + 1.663(V„)

- 2.079(V7) + 0.722(V8) - 0.199(V„)
+ 0.939 (V,„) + 1.939(Vu) - 1.439.

6. G. breviceps sagittalis vs. G. 1. major

X = - 0.071(V,) + 0.047(V2) - 0.790(V0
+ 0.752 (V.) + 0.565 (V,) - 0.946(V6)

- 1.81 1(V7) - 3.350(Vh) - 1.995(V.)

- 1.233(V1„) - 2.784(V„) + 61.932.

7. G. breviceps sagittalis vs. G. I. lutescens

X = - 0.095(V,) + 0.112(V2) - 0.358 (V3)
+ 0.359(V,) + 0.134(Vr,) + 0.716(V6)

- 3.890(V7) - 2.628(V„) - 2.195(V»)

- 0.293(Vio) - 1.590(Vii) + 60.493.

G. b. wisconsinensis

G. b. illinoensis

G. 1. lutescens i

i

0.0

1
.0

1 1
2.0

1

3.

Fig. 10. — Phenogram of cranial similarity of six
taxa of Geomys from the study area based on
cluster analysis of transformed taxon means.
Units on the axis are amalgamation distances.

CRANIAL MORPHOLOGY OF

GOPHERS IN THE ANTELOPE

COUNTY, NERRASKA

CONTACT ZONE

As noted in the discussion of gophers
from Nebraska, specimens that appear
to be hybrids between G. bursarius and
G. lutescens are available from a single,
very limited area near Oakdale, Antelope
County. The town of Oakdale lies about
1.5 km east of the extreme eastern edge
of the Nebraska Sandhills. To the east

POCKET GOPHERS OF THE GENUS GEOMYS

39

of Oakdale, tall-grass prairie predomi-
nates, and the soil is generally silty
loam; to the west of Oakdale is short-
grass prairie that lies on the sandy soil
of the Sandhills. Results described above
show that two readily distinguishable go-
phers, G. bursarius and G. lutescens,
occur in the tall-grass/ silty loam asso-
ciation and short-grass/ sandy soil asso-
ciation, respectively, and no intergrada-
tion could be detected on a broad scale.
We examined 30 adult females and 14
adult males from the vicinity of Oakdale
(Table 5; Appendix 1), with a special
interest in those from 0.5 to 1.5 km west
of Oakdale, where the two soil and vege-
tation types come into contact and inter-
grade. Details of soil and plant distribu-
tions, and gopher distribution and
reproduction are provided by Heaney
( 1979 ) . We provide data here on cra-
nial morphology that is critical to deter-
mining the presence of hybridization and
the width of the hybrid zone.

In order to quantify the rapidity of
change in cranial morphology of go-
phers in the contact area, a series of
stepwise discriminant function analyses
(BMDP7M) was conducted. A refer-
ence sample of G. bursarius from Butler,
Gage, Knox, Lancaster, Madison, and
Platte counties, Nebraska, and a sample
of G. lutescens from western Antelope,
Custer, Dawson, Greeley, and Valley
counties, Nebraska, were used as typical
representatives of those taxa; males and
females were analyzed separately, and
only adults were used. These samples
were chosen because they are geograph-
ically adjacent to the contact zone, so
that the direct and indirect effects of
climate on variation should be minimal.
In these analyses, all specimens from the
study area were entered as "unknowns"
to be compared with the two reference
samples.

In the analysis of females, a combi-
nation of three variables, considered si-
multaneously, was found to be the best
method of distinguishing between ref-
erence samples of G. bursarius and
G. lutescens; these were condylobasal

length, hind foot length, and frontal
square length. An F-to-enter of 4.0 was
used as a stopping criterion for deter-
mining the number of variables to be in-
cluded. All specimens of the two refer-
ence samples were correctly identified
to species by the classification function
that was generated, and the two samples
were significantly different ( p < 0.01 ) .
Females from the study area were plot-
ted in Fig. 11 using the same classifica-
tion function; thus, their positions along
the discriminant axis indicate their rela-
tive similarity to G. bursarius or G.
lutescens. There is a significant correla-
tion between discriminant scores and
distance west of Oakdale ( r = 0.733;
p < 0.01). Reference samples of G.
bursarius and G. lutescens were not in-
cluded in this regression.

Males from this area (Fig. 12) were
subjected to the same type of analysis,
again using an F-to-remove of 4.0 as stop-
ping criterion. The variables which en-
tered as significant were total length, zy-
gomatic breadth, mastoid breadth, and
appearance of the posterior portion of the
maxillary. The classification matrix indi-
cated that reference samples were sig-
nificantly different from one another
(p<0.05), but also that one member
of each reference sample (i.e., 6%) was
mistakenly identified by this classifica-
tion function as belonging to the wrong
species. As with the females, the corre-
lation between discriminant score and
distance west of Oakdale is significant
(r = 0.654, p<0.05). For both males
and females, visual inspection of the
graphs indicates that most of the change
occurs at about 1.5 km west of Oakdale,
at the extreme eastern edge of the Sand-
hills. The fact that specimens of inter-
mediate moi-phology exist, and that the
variation is nearly continuous, suggests
that introgression probably takes place.
However, introgression must be slight,
since the change from typical G. bur-
sarius cranial morphology to typical G.
lutescens morphology takes place over
a distance of no more than two km, with
most of the change occurring over a dis-

40

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

cd

s

3
O

O

V

a

o

c
<

a?

TJ

o

u
eS

CD

c

CD
C

o

N

4-J

o

s
o
o

CD

g CO

- %

S as

£ g

=o 3
:=> o

S *J
n §

O <-lH

° |

CO 3

ft s
p -C

H o
W) co

«> *3
> c

CCS -3
T3

0)

CD

/ — - CO
3 g

.2 £

>

cd

cS
3

+1

c

5

<d

B

a

<D

E

cd
>-.

3

CO

cd

in

w

<

-2

X

cd

CO

e

<

CO

>»

■4—

T3

3

— _

co u

CD
O

3
CD

CD

<-w

CD

PS

co
g

CO

U

CD .£
I- £

< «
Jo

CD

l>3

"C

a

S

3

«o

o+

*o

o+

<0

o+

«o

ot

*o

0+

Tt<

00

CO

00

00

in

CO

in
oq

in oq o m; q

rf h 6 05 O Tf

i— I cq

00 00 CO in CO cq

cq

H CO

cq CD

^_ 00 CD CD © l> ■■* "* TF CO in

n in d d h d d ci 6 6 d

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 II
ln^ci^ooqcDdcocico

05-^OCOCDTt<C>cd^^O
f l^ O 1< O] (M i— I

CN

CO CO CD CO CO

O lrf O H H

+1 +1 +1 +1 +1

cd oo tf oo ■*

CD

cq

CO

co
co

00

1—1
co

cq

o

i— i

t}<

i— i

co

02

02

CO

t»

rt<

in

i—5
i— i

+1

CO

co"

+1

i— 5

+ 1
CO

O

+ 1

i—l

©

+1

io

o
+1

CM

d
+1

CO

©

+1
o

CO

CO

cq

C5
CO

d

CXI

oq

•>*

CD
Cq

cq

03

"tf

— i in t- in oo in

--5 © © do ©

II +1 +1 +1 +1 +1

b II II t> Tf Tf

i-i CD O

^ in in

d d d

oo in rf
i-i tH o

cqcqo3cocqincqinint>in
cd i> cq cq' cq" 1-5 d d d d d

Ti +i +i +i +i +i +i +i +i +i +i

t- K3 O) b

co 1>
(M

i-h i-H in in oq co co

HNHbO^Htod

CO "* CO CNl i— I i—l

lnco'i-idddddddd
+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

CO 05 CO

in oq co

^0505«dTf05xf

co co oq ^ cq oq
oq

» oo t
■^ d

in

t-'

i—i

+1

oo

oq

i—i

CO

o
d

a>coinpi>cococDoqin

i>i-5di-5ddddr-5d

+1 +1 +1 +1 +1 +1 +1 +1 II +1

oq co cq -* in

i— 5 d ■* in i-5
en co in co co

o cd oq i-i cq
Tf oq oo d

iqwHcoNTtCNinw
t^^-5i— 5r-5dddddo

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1

i— l05COOq01>05CBr- j O l>

ooTri>'di>dcdoq'dd

T(l

oq

o co -tf co cq

oo ^f co

d d oq"

i—l i-H

+ 1 +1 II

WHO)

T|i 3 b

i-H 00 CO
CO

t~; cq cq

i— 5 d i— 5

+1 +1 +1

t H t»

in" in ■*

t- » CO
cq

CqOi— ICDCDTfCOCO

-<cti cd cq d d d i-5 d

II +1 +1 +1 +1 +1 +1 II

oiTfcJOiOHinq

in ■<* d --5 "*' cq i> d

in co co i—i i—i

oo co in

+1 +1 +1

co in cq

co as i>

tji oq cq

to

TJH Tf CO l~- CO

dodo d

+1 +1 +1 tl +1

oq cq o o en

d cd cq d d

3

J3

4-t

XI

It

0

rD.

T3
CD

X.

-t-»

T3

3

-4-J

o

4-J

as
CD

4-»

o

T3

O
T3

c8

s

o

T2
'o

-D

13

•t-H

3

3

W)

CO

cS
i«ri

CO

O

H

as
H

S

O
U

aS

111 >
CD ?

<D ^

cd 2

CJi C71

CD

ti) -S
3 "

O i — i +j 4-» xj *~^
+_» rt c! fi -^ 3

§ s

cd

fa fa o

POCKET GOPHERS OF THE GENUS GEOMYS

41

tance of less than one km. Data on
bacular and chromosomal variation given
below also indicates a hybrid zone less
than two km wide. Given the post-
Pleistocene history of the area ( summar-
ized below, and in greater detail by
Heaney, 1979), it is likely that the two
species have been in contact for up to
9,000 years. Because introgression is

slight, as indicated by the data given
here, it appears that a "fusion model"
type of hybrid zone does not exist at the
locality. The most likely alternatives are
the "isolation" and "hybrid superiority"
models; neither of these is indicative of
the two taxa of gophers belonging to a
single species ( see Heaney, 1979; Moore,
1977).

7.0-

5.0-

UJ

cc
O
O

C/)

\-
z
<

3.0-

1.0-

o -1.0-1

- 3.0-

5.0-

c

<D
O

CD

O

CO

3

hm

co

CO

k-
3
.£3

6

0.0

1.0

2.0

3.0

DISTANCE (KM)

Fig. 11. — Results of discriminant function analysis of adult female Geomys from the vicinity of
Oakdale, Antelope Co., Nebraska. Reference samples of G. bursarius and G. lutescens were en-
tered as knowns, and all Oakdale area specimens were entered as unknowns. Ranges for the ref-
erence samples are given as vertical bars on the left and right for G. bursarius and G. lutescens,
respectively. Distances given are west of the Oakdale post office. The equation for the regression
line shown is: discriminant score = 5.06 (distance) —2.50. (r = 0.733, p < 0.01).

42

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

CLADISTIC ANALYSIS OF
CRANIAL CHARACTERS

As noted in the descriptions of taxa
given above, there are a number of cra-
nial characters useful in distinguishing
between taxa; seven of the most consis-
tent are summarized in Tables 6 and 7.
We have used these in a cladistic analy-
sis of relationships, utilizing the Wagner
Tree method ( see Kluge, 1976; Carleton,
1980). We have included data on
Geomys personatus, G. pinetis, and Zy-
gogeomys trichops as well. Geomys per-
sonatus was considered by Merriam
( 1895) to be a close relative of G. brevi-
ceps, whereas G. pinetis is generally

thought to be either the sister-group to
all other Geomys (Russell, 1968), or to
form a clade with G. bursarius and G.
lutescens, with G. breviceps as the sister-
group to these three (Merriam, 1895).
Zygogeomys is thought to represent the
most primitive surviving stock in the
tribe to which Geomys belongs, the
Geomyini, and is also considered the
sister-group to Geomys (Merriam, 1895;
Russell, 1968). The analysis we con-
ducted did not include the following
characters unique to Geomys pinetis:
rostrum long and slender; incisors rela-
tively narrow; incisors procumbent. Our
objective was to evaluate relationships
of the species we studied, and the re-

5.0-

3.0-

LU
DC
O

0 1.0H

c/>

z
<

1 -1.0H

K
O
C/)

Q

-3.0-

5.0-

(A

3
i_
CO
(/)

3

d

CO

c

<D
O
CO
0)

O

0.0

1.0

2.0

3.0

DISTANCE IKMI

Fig. 12.— Results of discriminant function analysis of adult male Geomys from the vicinity of
Oakdale, Antelope Co., Nebraska (see Fig. 11). The equation for the regression line is: discrimi-
nant score = 3.70 (distance) - 4.17. (r = 0.654, p < 0.05).

POCKET GOPHERS OF THE GENUS GEOMYS 43

Table 6. — Summary of qualitative cranial characters that differ among taxa of Geomys.

Sagittal

Zygomatic

Mastoid

crest

arch width

process

Frontal

Foramen

Sub-

develop-

(anterior/

Rostral

develop-

square

rotundum

lacrimal

ment

posterior)

length

ment

shape

position

fossa

OTU

(1)

(2)

(3)

(4)

(5)

(6)

(7)

1.

G. b. bursarius

great

>1.0

long

great

long
rectangle

high

deep

2.

G. b. illinoensis

great

>1.0

very long

great

long
rectangle

high

deep

3.

G. b. wisconsinensis

great

>1.0

long

great

square

high

deep

4.

G. 1. lutescens

moderate

>1.0

short

moderate

long
rectangle

high

deep

5.

G. 1. major

moderate

>L0

short

moderate

long
rectangle

high

deep

6.

G. breviceps

slight

^1.0

short

slight

long
rectangle

low

deep

7.

G. personatus

slight

>1.0

moderate

slight

long
rectangle

high

deep

8.

G. pinetis

slight

ML0

long

slight

square

high

shallow

9.

Zygogeomys trichops

moderate

>1.0

long

moderate

short
rectangle

high

shallow

suits should be viewed in that context.
Polarity of characters was based on
out-group comparison with the presumed
primitive sister-group, Zygogeomys tri-
chops. Although this leaves open the
possibility of error through character
convergence of Z. trichops with one or
more species of Geomys, it provides a
non-arbitrary reference point that is
likely to contain many primitive charac-
ters. The characters used (see Table 6)
were the following.

1. Sagittal crest development.

a. moderate

b. slight

c. great

In Z. trichops, the sagittal crest aver-
ages 2 mm wide, and varies in height
from about 0.5 mm anteriorly to 1.0 mm
posteriorly. Geomys bursarius has a
higher, narrower crest, whereas the other
species of Geomys have a broad, low
ridge between the temporal crests. Be-
cause the latter group of Geomys vary
greatly in cranial size, this does not ap-
pear to be a strictly allometric trait.

2. Zygomatic arch width.

a. width at anterior angle greater
than at posterior angle

b. width at anterior and posterior
angles approximately equal

Zygogeomys and most Geomys have
zygomatic arches that flare laterally, then
deflect posteriorly at an angle over 90°.
At the level of the condylar fossa the
zygoma abruptly turn and merge with
the braincase. In G. breviceps and G.
pinetis, the zygoma are not proportion-
ately as broad anteriorly, so that the
arches lie roughly parallel to the skull,
and the width at anterior and posterior
angles is approximately equal.

3. Rostral length.

a. long

b. moderate

c. short

d. very long

The rostrum and diastema of Zygo-
geomys are long, and appear to be
equivalent in Geomys b. bursarius and
G. b. ivisconsinensis; the rostrum of G. b.
illinoensis is even more elongate. The
other species of Geomys have propor-
tionately shorter rostra, with G. pinetis
intermediate between Zygogeomys and
the very short rostrum typified by G.
breviceps and G. lutescens.

44

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Table 7. — Data matrix of cranial characters coded for Wagner Tree analysis. OTUs, character
numbers, and character states are from Table 6; Zygogeomys trichops is assumed to represent the
primitive condition. Multistats characters were coded using the method described by Carleton (1980).

G.b.

G. b.

G.b.

G. 1.

G. 1.

G.

G.

G.

z.

bursarins

illinoensis

wisconsmcnsis

hitcscens

ma/or

breviceps

personatus

pinetis

trichops

(1)

0

0

0

0

0

1

1

1

0

1

1

1

0

0

0

0

0

0

(2)

0

0

0

0

0

1

0

1

0

(3)

0

0

0

1

1

1

1

0

0

0

0

0

1

1

1

0

0

0

0

1

0

0

0

0

0

0

0

(4)

1

1

1

0

0

0

0

0

0

0

0

0

0

0

1

1

1

0

(5)

1

1

1

1

1

1

1

1

0

1

1

0

1

1

1

1

0

0

(6)

1

1

1

1

1

0

0

0

0

(7)

1

1

1

1

1

1

1

0

0

4. Mastoid process development.

a. moderate

b. great

c. slight

The mastoid process of Zygogeomys
is prominent, robust, and triangular in
dorsal view; its width (from the medial
margin above the auditory canal to the
margin point directly posterior) is
slightly less than the length (measured
lateral to the line defined by width to
the lateral tip), and it is deflected at
about 30° from the horizontal. In
Geomys bursarius, the mastoid is longer
than wide, and deflected at about 25°.
In G. breviceps, G. personatus, and G.
pinetis, the mastoid is of about equal
length and width, and is deflected at 40°
to 45°.

5. Frontal square shape.

a. short rectangle

b. square

c. long rectangle

The anterior portion of the frontals
lies between the posterior projections of
the premaxillae, forming a small rec-
tangle that is open on the posterior end.
In Zygogeomys this "frontal square" is
wider than it is long; in G. pinetis and
G. /;. ivisconsinensis it is of equal length
and width; and in the other species, it
is longer than wide.

6. Foramen rotundum position.

a. high on alisphenoid

b. low on alisphenoid

The foramen rotundum of Zygo-
geomys lies high on the alisphenoid, at
or near the top of the horizontal portion
of the alisphenoid, and adjacent to the
condylar fossa. It is usually obscured
in lateral view by the zygomatic arch.
The same condition is found in all
Geomys examined except G. breviceps,
in which the foramen rotundum lies
about one-third of the way down the
horizontal portion of the alisphenoid and
is easily visible in lateral view.

7. Sub-lacrimal fossa.

a. shallow

1). deep

At the anterior edge of the orbit is a
fossa that lies beneath the horizontal
portion ( = dorsal process ) of the lacri-
mal. This fossa is bounded posteriorly
by a thin ridge on the horizontal wall of
the orbit, and ventrally by a continuation
of the ridge. The fossa has a foramen
that pierces its postero-ventral edge, de-
scending ventrally into the maxillary. In
Zygogeomys the fossa extends anteriorly
only a short distance into the zygomatic
root of the maxilla, the ventral foramen
is small, and the ridge at the posterior
margin of the fossa is low and rounded.
This condition appears to be identical
to that in G. pinetis. In all other species
of Geomys examined, the fossa extends
substantially deeper into the zygomatic

POCKET GOPHERS OF THE CENUS GEOMYS

45

root of the maxilla, the ventral foramen
is large, and the ridge at the posterior
margin is high and sharp-edged.

Results of the Wagner Tree analysis
are depicted in Fig. 13. Geomys pinetis
lies closest to Zygogeomys tricliops, i.e.,
possesses few derived character states
that are shared with other Geomys. How-
ever, it should be borne in mind that
the three characters related to the ros-
trum (cited above) that are unique to
G. pinetis were not included in this
analysis. For the characters considered,
G. pinetis is intermediate between Zy-
gogeomys and the node (#1) that links
G. breviceps and G. personatus. These
two species differ little from each other,
and although the Wagner Tree shown
illustrates G. breviceps as the root for
G. lutescens, both G. breviceps and G.
personatus have an equal probability of
occupying this position. Geomys lutes-
cens and G. bursarius appear to be sister-
species that are derived from a G. brevi-
ceps or G. personatus-\ike ancestor.

illinoensis

major,
lutescens

personatus

wisconsmensis
bursarius

breviceps

it pinetis

Zygogeomys

Fig. 13. — Results of a Wagner Tree analysis of
seven cranial characters, using data in Table 6.
Index of consistency = 0.667. Nodes 1, 2, and
3 represent hypothetical ancestors for the vari-
ous clades.

The analysis suggests that the break
of G. pinetis from the ancestral stock
was an early one, and that G. personatus
and G. breviceps are species that have
retained traits that are similar to the
population that gave rise to G. lutescens,
and subsequently to G. bursarius.

ANATOMY OF THE GLANS PENIS
AND BACULUM

The anatomy of the male reproduc-
tive tract has been shown to provide
useful taxonomic characters in many
groups of mammals (e.g., Burt, 1960;
Carleton, 19S0). However, the anatomy
of the glans penis has not previously
been described in Geomys, and few ob-
servations have been made on bacula
of Geomys. Sherman (1940) illustrated
a baculum of Geomys fontanelus ( =G.
pinetis fontanelus; see Williams and
Genoways, 1980), and Burt (1960) illus-
trated a baculum from Geomys bursarius
bursarius. Kennerly (1958) pointed out
that some variation exists among species
of Geomys.

We examined the glandes penes of
three adult Geomys bursarius bursarius
and three G. lutescens lutescens, and
found no substantial variation from the
morphology shown in Fig. 14, although
specimens of G. /. lutescens were not
well preserved. The glans is cylindrical
and is entirely covered with evenly-
spaced, extremely small tubercles. Each
tubercle has two rows of spines that are
directed posteriorly; each row had three
to four spines (Fig. 14). The glans
measures about 5 mm, with a tip ( bacu-
lar mound) extending some 2.5 mm be-
yond. The rim of the terminal crater is
slightly flared, and in lateral or dorsal
view obscures the urethral pore, urethral
lappets, and other features within the
terminal crater. The raphe (mid-ventral
ridge) is distinct but low and narrow,
extending the entire length of the glans.
The baculum is visible under strong
light within the semi-transparent bacular
mound. The mound is nearly as broad
as it is long; there is no cartilaginous tip
on the baculum. The structure described

46

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Fig. 14. — Glans penis of an adult Geomys bur-

sarius hursarius from Anoka Co., Minnesota.

On right is an enlargement of a tubercle from

the lower right on the glans.

and illustrated here is nearly identical
to that exhibited by Liomys and Hetero-
mys, except that both heteromyid genera
lack tubercles on the glans (Genoways,
1973).

Our small samples of glandes from
the two species of Geomys failed to show

differences between species, but this was
not the case for the bacula we examined.
As described by Burt (1960), all bacula
were simple rod-like bones, with the
rounded shaft curving gently upwards.
The distal tip is laterally expanded and
dorsoventrally flattened ( Fig. 15 ) . Table
8 lists measurements for G. hursarius,
G. 1. lutescens, and G. I. major, and gives
values for a single example of G. brevi-
ceps sagittalis. We found that basal
width varied greatly, and that it ap-
peared to be correlated with age of the
gopher. Midshaft width showed less
variation, and no obvious correlation
with age, once adult size ( and presuma-
bly breeding condition) was reached.

The bacula of G. hursarius were the
largest we examined on average, al-
though those of G. /. major were only
slightly smaller. The bacula of G. hur-
sarius were generally more robust than
those of G. 1. major (Fig. 15), although
this was not reflected in the midshaft
diameter/ length ratio (Table 8).

The bacula of G. I. lutescens were
shorter than those of G. hursarius or G. Z.
major, but averaged appreciably wider
at midshaft, and all gave the appearance
of being more robust overall.

Our one specimen of G. hrcviceps
sagittalis was much smaller than any

B

10 mm

Fig. 15. — Bacula of adult Geomys. A = G. hursarius
81848). B = hybrid G. h. hursarius X G. 1. lutescens
braska (KU). C = G. lutescens lutescens from Brown
major from Cimarron Co., Oklahoma (UMMZ 99812).

hursarius from Jasper Co., Iowa (UMMZ
from 1 mi. W Oakdale, Antelope Co., Ne-
Co., Nebraska (KU). D = G. lutescens
E = G. hreviceps sagittalis from Love Co.,

Oklahoma (UMMZ 99722).

POCKET GOPHERS OF THE GENUS GEOMYS 47

Table 8. — Measurements and proportions (mean ± standard deviation) of bacula of Geomys.

Width of

Midshaft

Midshaft width

Taxon

N

Length

Base

Width

Length

G.

h. bursarius

5

11.32±0.46

1.82±0.19

0.95±0.08

0.084 ±0.007

G.

brcviceps sagittalis

1

9.56

1.69

0.79

0.0826

G.

attwateri*

9

9.86

1.70

—

—

G.

I. lutescens

8

10.78±0.41

1.98±0.42

1.12±0.14

0.105 ±0.014

G.

I. major

4

11.21±0.11

2.09±0.33

0.94±0.15

0.083 ±0.014

G.

h. bursarius X

G. 1. lutescens

1

11.68

1.94

1.21

0.104

* = from Kennedy, 1958

other. It was similar in size and appear-
ance to nine bacula of G. attwateri de-
scribed by Kennerly ( 1958 ) , and was
proportioned much the same as G. bur-
sarius and G. 1. major.

Although the morphology of the
bacula generally indicates the distinct-
ness of these four taxa, we find the de-
gree of difference between G. 1. lutescens
and G. 1. major to be surprising. It would
be worthwhile to examine bacula from
gophers from the putative zone of inter-
gradation between the two taxa, but no
specimens are currently available to us.

We also examined bacula from ten
adult gophers from an area of contact
and apparent hybridization between G.
b. bursarius and G. I. lutescens in Ante-
lope County, Nebraska (see Heaney,
1979, and above). Fig. 16 is a graph of
the midshaft diameter/ length ratio for
all specimens from this contact zone,
with the ranges and means for G. b.
bursarius and G. 1. lutescens from out-
side the zone for scale. Although the
data are scanty, it appears that at least
a few gophers from the area about one
mile west of Oakdale (the eastern edge
of the hybrid zone) have bacula inter-
mediate between typical representatives
of the two taxa, and that substantially
more variation exists in the population
in the contact zone than is usual for
populations elsewhere.

In summary, the taxa of gophers ex-
amined in this study all appear to have
bacula differing recognizably from each
other. Most differences are in size, and
appear to be correlated with the overall

size of the animal. Geomys lutescens
lutescens has the most distinctive bacu-
lum, in that it differs from the others in
relative mid-shaft width. A contact zone
between G. b. bursarius and G. 1. lutes-
cens yields gophers with unusually vari-
able bacula, with no consistent indica-
tion of intermediacy.

KARYOTYPIC EVIDENCE

Few pocket gophers from the north-
ern portion of our study area (i.e., north

.13

.12

K

a

.11-

.10

z

< .09
a

.08"

— r-
1.0

2.0
DISTANCE |KM|

3.0

Fig. 16. — Morphology of bacula of adult Geo-
mys from the vicinity of Oakdale, Antelope Co.,
Nebraska. Ranges for reference samples of
G. bursarius and G. lutescens are from Table 8.
Distances given are distance west of the Oak-
dale post office.

48

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

of Oklahoma) have been karyotyped;
all previously available information was
summarized by Hart (1978). He found
that Geomijs from Wisconsin, Iowa, and
east-central Missouri had the same karyo-
type ( 2N = 72, FN = 72, 2 biarmed auto-
somes), and these differed only slightly
from gophers from Illinois and eastern
Kansas ( 2N = 72, FN = 70, no biarmed
autosomes), and eastern Nebraska and
eastern South Dakota ( 2N = 70, FN =
68, no biarmed autosomes). This group
is what we consider to be Geomys bur-
sarius sensu stricto. Thus, the karyotype
for this species is 2N = 70-72, FN = 68-
72, and the X chromosome is a large
metacentric in southeastern South Da-
kota and eastern Nebraska, whereas in
all other parts of the range the X is a
large acrocentric.

More detailed chromosomal data are
available for Geomys from Oklahoma,
Texas, and New Mexico ( Baker et al.,
1973; Baker and Genoways, 1975; Hart,
1978). Geomys lutescens knoxjonesi
(2N = 70, FN = 68-70, 0-2 biarmed auto-
somes) has a karyotype much like that
of G. I major (2N = 70, FN = 70, 0-2
biarmed autosomes) (Baker and Geno-
ways, 1975; Hart, 1978). Hart (1978)
reported that the kaiyotype of popula-
tions previously recognized as industrius
had autosomes identical to those of ma-
jor, the only difference being that in the
industrius populations, the X was a large
metacentric whereas in major it was a
large acrocentric. Other members of G.
lutescens ( = the "lutescens group" of
Honeycutt and Schmidly, 1979) are G.
lutescens texensis and G. I. llanensis,
both with 2N = 70, FN = 68-69, and no
biarmed autosomes. In contrast, G. lu-
tescens lutescens from central and west-
ern Nebraska differs greatly in the num-
ber of biarmed autosomes (2N = 72,
FN = 86-98, 16-28 biarmed autosomes;
Hart, 1978), with some populations
showing polymorphism. This includes
populations previously placed in two
other subspecies that are no longer con-
sidered valid, G. /. hylaeus and G. I.
vinaceus, in addition to lutescens.

The contact zone between G. attwa-
teri and G. breviceps has been described
in detail by Honeycutt and Schmidly
(1979) and Tucker and Schmidly (1981).
They found that G. breviceps sagittalis
has a diploid number of 74 and funda-
mental number of 72-74, and G. attwateri
has a diploid number of 70, and funda-
mental number of 72-74.

This brief summary shows that dip-
loid number varies from 70 to 74, and
with one exception, fundamental number
varies from 68 to 74. The one aberrant
karyotype within the group is that of G.
lutescens lutescens, with a diploid num-
ber of 72, but a fundamental number of
86 to 98.

Timm et al. (1982) reported karyo-
types from 24 Geomys from the vicinity
of the contact zone between G. bursarius
and G. lutescens near Oakdale, Antelope
Co., Nebraska. Two G. b. bursarius from
2 '4 mi. E Oakdale, Antelope Co. (which
is four miles east of the center of the
contact zone), had karyotypes identical
to those reported by Hart (1978) for a
single individual from 1 mi. W Tilden,
Antelope Co. ( 2N = 70, FN = 68, with
no biarmed autosomes, and a large meta-
centric X). Eight gophers karyotyped
from l?lo mi. W Oakdale and areas to
the west of that point had karyotypes
identical to those described by Hart
(1978) for G. /. lutescens from Chadron,
Dawes Co., and 4 mi. S Neligh, Ante-
lope Co. (2N = 72, FN = 86-98, 16-28
biarmed autosomes ) .

Twelve gophers from the area be-
tween the parental types (from %o mi.
W Oakdale to Bio mi. W Oakdale) had
fundamental numbers ranging from 75
to 95. Six individuals had fundamental
numbers intermediate between the pa-
rental types (FN = 75 to 82), and one
of these six had 2N = 71. Because of
the high variability in FN in the parental
population of G. lutescens ( FN = 84-
92), it is not possible to say whether the
karyotypically intermediate individuals
are F, hybrids or backcrosses. It is ap-
parent that the change from the G.
bursarius to G. lutescens cytotype takes

POCKET GOPHERS OF THE GENUS GEOMYS

49

place over a distance of about one-half
mile (Timm et al, 1982).

EVIDENCE FROM PARASITES

As part of our study of relationships
of Geomys in the Great Plains, lice ( Mal-
lophaga: Trichodectidae) of the genus
Geomydoecus were obtained from all
species and subspecies of Geomys that
were recognized previously, for use as
an additional taxonomic character. A re-
vision of the Geomydoecus on the
Geomys bursarius complex resulted in
redescription of the four previously
recognized species of lice, description of
four additional species, and refinement
of our knowledge of the distribution of
lice on pocket gopher taxa (Timm, 1979;
Timm and Price, 1980; Timm, 1983).
Taxa of Geomydoecus are distinguished
primarily on the morphology of the geni-
talia of both males and females, but also
on differences in chaetotaxy, size, and
the distinctive antennal scape of the
males. These characters exhibited little
intrapopulational variation, but varied
considerably geographically. No indi-
vidual or population of Geomys was
found to be parasitized by more than
one species of Geomydoecus, and in gen-
eral, a single species of louse was found
throughout the range of a taxon or
group of taxa of pocket gophers.

The Geomydoecus on the Geomys

bursarius complex cluster into two main
groupings that Timm and Price (1980)
termed the "northern group" and the
"southern group." The "northern group"
is composed of two complexes of species,
the "geomydis" complex and the "okla-
homensis" complex; the "southern group"
is composed of three species (see Fig.
17).

Northern Group. — Geomydoecus geo-
mydis is found only on two subspecies
of pocket gophers, Geomys bursarius
bursarius and G. b. ivisconsinensis. It
was found on all individuals examined
from throughout the range of the two
subspecies, except for those pocket go-
phers in the vicinity of St. Louis, Mis-
souri, previously referred to as G. b.
missouriensis. The second member of
the "geomydis" complex, Geomydoecus
Ulinoensis, is restricted to one geograph-
ically isolated subspecies of pocket go-
pher, G. b. Ulinoensis. Although defi-
nitely a member of the "geomydis"
complex, this louse is quite distinctive
in several metric features and so clusters
separately from G. geomydis (Timm and
Price, 1980).

The second complex in the northern
group includes three species, Geomydoe-
cus oklahomensis, G. nebrathkensis, and
G. spickai, and was termed the "oklaho-
mensis" complex. Geomydoecus oklaho-
mensis was found on the following

G spick a /
G. oklahomensis
G nebrathkensis
G. geomydis
G. Ulinoensis
G subgeomydis
G. ewingi
G. heaneyi

— i —
4.0

— i —
8.0

— I —
7.0

6.0

5.0

3.0

2 0

.0

Amalgamation Distance

Fig. 17. — Phenogram of morphological similarity of eight species of lice (Geomydoecus) that para-
sitize Geomys based on cluster analysis of species means. ( From Timm and Price, 1980. Re-
printed with permission of the Journal of Medical Entomology).

50

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

pocket gopher taxa: G. I. knoxjonesi,
G. I. lutescens, and G. 1. major. The
pocket gophers from southwestern Ne-
braska, western Kansas, Oklahoma, and
Texas, and eastern Colorado and New
Mexico all are parasitized by this spe-
cies of Geomydoecus. Although this
louse was the most variable and the
most widely distributed of the eight spe-
cies found on the G. bursarius complex,
there was no evidence that any of the
populations warranted classification as
a distinct taxon. The lice on two previ-
ously recognized subspecies of pocket
gophers, industrius and jugossiadaris,
appeared to be one uniform population.
Geomydoecus nebrathkensis is found
only on populations of G. 1. lutescens
found north of the Platte River in north-
ern Nebraska, northeastern Colorado,
eastern Wyoming, and southern South
Dakota and not on other populations of
this gopher subspecies. The presence of
G. nebrathkensis north of the Platte
River may represent a speciation event
for the lice which is not reflected in mor-
phological differentiation of the hosts.

Geomydoecus spickai is found only
on the pocket gophers in the vicinity of
St. Louis, Missouri, the population that
previously was considered a distinct sub-
species, G. b. missouricnsis. The pres-
ence of this species of louse on pocket
gophers that are now considered Geo-
mys b. bursarius on morphological and
karyotypic grounds represents a discor-
dance between relationships suggested
by the parasite data and that suggested
by other kinds of data. At present we
know nothing about what lice may have
occurred on pocket gophers in southern
and central Missouri, and we must ten-
tatively conclude that the presence of
G. spickai on this one population of Ge-
omys bursarius represents a dispersal
event not parallel to host speciation.

Southern Group. — Geomydoecus hea-
neyi is found on only two subspecies of
pocket gophers, llanensis and texensis.
G. heaneyi is a very distinctive louse
and its presence on both llanensis and

texensis suggests that they are most
closely related to each other.

Geomydoecus subgeomydis is found
on two species of pocket gophers, Geo-
mys attwateri (in part) and G. breviceps
sagittalis. Geomydoecus ewingi is found
on two species of pocket gophers, G.
attwateri (in part), and G. breviceps.
The presence of G. ewingi on G. /;.
breviceps and G. /;. sagittalis supports
the conclusion of Honeycutt and
Schmidly (1979) that breviceps is most
closely related to sagittalis.

The boundary that Honeycutt and
Schmidly (1979) drew between Geomys
attwateri and G. breviceps sagittalis is
in close agreement with the boundary
between the two species of lice, Geomy-
doecus subgeomydis and G. ewingi;
however, the match is not perfect ( Timm
and Price, 1980). In general, G. ewingi
is found on the eastern species, G. brevi-
ceps, and G. subgeomydis is found on
the western species, G. attwateri. How-
ever, along the Brazos River some pop-
ulations of Geomys that Honeycutt and
Schmidly (1979) referred to the eastern
species, G. breviceps, were parasitized
by the western louse, G. subgeomydis.
Also, there is a population of the eastern
louse, G. ewingi, in Atascosa, Bexar,
Goliad, and Wilson counties, Texas, that
is separated by a population of G. sub-
geomydis from the main body of G.
ewingi. An analysis of the lice showed
that there was no justification for split-
ting the two separated populations of
G. ewingi into two or more taxa (Timm
and Price, 19S0). These discrepancies
suggest that our understanding of the
systematica and host relationships of the
Geomys and Geomydoecus of southeast-
ern Texas warrants further investigation.

RELATIONSHIPS OF THE EXTANT
SPECIES OF GEOMYS

Relationships of the living species of
Geomys have been considered in detail
only by Merriam (1S95) and Russell
(1968); the phylogenetic trees they pro-
posed are shown in Fig. 18. The pre-

POCKET GOPHERS OF THE GENUS GEOMYS

51

ceding pages contain a body of data
that may be used to test these previously
proposed phytogenies, and may also be
used if necessary to construct a new
phylogenetic tree that is more in con-
cordance with available data. Because
we have detailed data from only three
of the extant species, some of our con-
clusions must be tentative. However,
some definitive statements can be made.

In his revision of the Recent Geomyi-
dae, Merriam (1895) considered Zygo-
geomys trichops to represent a species
similar to the ancestor of Geomys.
Within Geomys, he considered the G.
breviceps group ( including G. arenarius,
G. breviceps, G. personatus, and G. trop-
icalis) to be the most primitive, with
G. pinetis, G. lutescens, and G. bursarius
representing progressively more derived
species (Fig. 18A). Russell's (1968)
study of relationships of genera was
based on fossil as well as Recent ma-
terial. Like Merriam, he considered Z{/-
gogeomys trichops to be close to the
ancestry of Geomys (Fig. 18B). How-
ever, he considered G. pinetis to be the
most primitive living species, and placed
G. breviceps (in synonymy with G. bur-
sarius) among the most derived forms.
He placed G. personatus and G. are-
narius, respectively, as progressively
more derived species in a lineage lead-
ing to G. bursarius (including G. brevi-
ceps and G. lutescens) .

The first conclusion that we may
draw is that Russell (1968) was incor-
rect in considering G. breviceps and G.
lutescens to be synonyms of G. bursarius.
This might have had little effect on his
concept of relationships if the three spe-
cies form a monophyletic clade, but it
should be recognized that he lumped
them because he followed Villa-R. and
Hall (1947), rather than because of cri-
tical examination of evidence. Merriam
(1895) was correct in separating the
three species.

Our phenogram of cranial similarity
(Fig. 10) suggests a closer relationship
between G. lutescens and G. breviceps
than between either of these and G.

„*»

«\S6 re?* c.

0^'>>;>,e;<'.<^

**<**"*

^\<v»s ,^6

««*>

Fig. 18. — Hypothesized phylogenetic trees of

geomyids suggested by: A. Merriam (1895:

24), andB. Russell (1968).

bursarius. This is not relevant to Rus-
sell's phytogeny (Fig. 18B), because he
considered all three to be conspecific. It
does not support Merriam's phytogeny
(Fig. ISA), since he indicated a closer
relationship between G. lutescens and G.
bursarius than between either and G.
breviceps.

Morphology of the bacula appears
to be of little help in discerning rela-
tionships, since with one exception the
difference between taxa appears to be
well correlated with size, which cannot
necessarily be assumed to indicate rela-
tionship. The one exception is G. lutes-
cens lutescens, whose relatively stout
baculum may be considered a derived
character.

Because the homologies of individual
chromosomes are unknown in most of

52

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

the species of Geomys (i.e., little band-
ing has been carried out), there is little
definitive information content in the data
available. The most aberrant karyotype
among the taxa considered in this study
is that of G. lutescens lutescens, which
has a large number of biarmed chromo-
somes. It is most probable that this is
a derived condition, with the possession
of all or nearly all acrocentric chromo-
somes representing the primitive condi-
tion in this group.

The phylogeny of the ectoparasitic
lice (Geomydoecus) found on Geomys
(Timm and Price, 1980; Fig. 17) sug-
gests a close relationship between G.
bursarius and G. lutescens, with G. brevi-
ceps being the sister-group, and so is in
accordance with Merriam's tree (Fig.
ISA). Relationship of lice on other spe-
cies of Geomys are not explicitly ex-
pressed (Price and Emerson, 1971).

Our cladistic analysis of seven cranial
features using the Wagner Tree method
( Fig. 13 ) is in better concordance with
the phylogeny proposed by Merriam
(1895) than Russell's (1968), but is not
identical to either. Our analysis supports
Merriam's suggestion that G. bursarius
and G. lutescens are highly derived sis-
ter-groups. The analysis supports Rus-
sell's suggestion that G. pinetis is an
early off-shoot of the Geomys stock, and
that G. personatus is close to the an-
cestry of G. lutescens and G. bursarius.

We conclude from this discussion
that the weight of evidence does not
support either Merriam's (1895) or Rus-
sell's ( 1968 ) hypothesized phylogeny of
Geomys. A tree that fits the data should
have the following features. First, Geo-
mys pinetis should be shown as an early
offshoot of the main stock, as indicated
by its generally primitive set of charac-
ters. Second, G. personatus should be
shown as similar to G. breviceps, and
both as potential sister-species to the
G. lutescens-G. bursarius clade. Third,
G. lutescens should be shown as being
intermediate between G. breviceps and
G. bursarius. Fig. 19 is our estimate of

the phylogeny that best represents these
features.

One significant feature that emerges
from this study and is emphasized in
our consideration of phylogenetic rela-
tionships is that parapatric and allopatric
populations of pocket gophers often de-
velop discrete, unique cranial characters
before they become genetically inde-
pendent of one another, that is, before
they attain species level (i.e., the sub-
species of G. bursarius and G. lutescens) .
This suggests that analyses of taxonomic
relationships near the species level of
differentiation must give substantial at-
tention to evidence of intergradation be-
tween taxa. Any analysis that deals only
with qualitative characters, and does not
investigate geographic variation, is likely
to err in overestimating the number of
species present in any given group. A
second important feature that is evident
is that limited hybridization, such as ap-
pears to exist between G. bursarius and
G. lutescens, and between G. lutescens
and G. breviceps, is not a useful indi-
cator of relationship, since such hybrid-
ization is not confined to sister-species.

BIOGEOGRAPHIC
INTERPRETATIONS

Although our knowledge of the Pleis-
tocene flora and fauna of the Great
Plains has advanced greatly in recent
years ( see, for example, Dort and Jones,
1970), current paleobotanical data are
insufficient to infer the extent of prairie
vegetation patches that might have been

««*<^

of

Fig. 19. — Hypothesized phylogenetic tree
selected extant species of Geomys based on
results of this study.

POCKET GOPHERS OF THE GENUS GEOMYS

53

critical to differentiation of pocket go-
phers. For this reason, the following dis-
cussion is general in scope, and deals
with major glacial advances and floristic
changes. We present this as a plausible
hypothesis which may be tested with
more detailed knowledge of the Pleisto-
cene history of the Great Plains and the
gophers themselves.

The Geomys bursarius and G. pinetis
complexes originated no later than the
late Irvingtonian, as offshoots from either
G. garbanii or G. tobinensis (Kurten and
Anderson, 1980). The earliest records
of the G. bursarius complex are from the
Yarmouthian of Ellsworth County, Kan-
sas (Hibbard et al, 1978). We suggest
that the ancestors of the two major
clades in the G. bursarius complex, the
breviceps group (including att water i,
arenarius, personatus, and tropicalis),
and the bursarius group (bursarius and
lutescens), split during the Kansan gla-
ciation, with the breviceps group speciat-
ing during some uncertain later time.
The split between G. bursarius and G.
lutescens probably occurred during the
Illinoian glaciation when glacial ad-
vances, and probably forest expansion,
were at a maximum. The subsequent
interglacial, the Sangamonian, was most
likely a period of some expansion for the
prairie (or savannah) flora and fauna,
and Geomys were recorded from Kansas,
Nebraska, Oklahoma, and Texas during
this period (Hibbard, 1970; Russell,
1968). We suggest that during the San-
gamonian, approximately 200,000 years
BP, Geomys spread to most regions it
currently occupies. During the subse-
quent stage of glaciation, the Wiscon-
sinan, many of the populations of pocket
gophers were isolated as forest and gla-
ciers advanced; most must have become
extinct as their habitat was reduced.
Continental glaciers gradually moved to
cover much of the area now occupied
by Geomys in the northern portions of
its range. Probably all of Minnesota
(Wright and Ruhe, 1965) and eastern
North and South Dakota (Lemke et ah,
1965) were covered by ice. Only a small

portion of southwestern Wisconsin was
ice-free (Frye et ah, 1965), but the pres-
ence of Geomys cf. bursarius (Hay, 1923:
343), Thomomys talpoides, and other
small mammals shows that a region of
forest parkland remained throughout the
Wisconsinan ( Rasmussen, 1971 ) .

Illinois and Indiana were nearly com-
pletely covered by glaciers during the
Illinoian glaciation, but were ice-free
during the Sangamonian, and were only
partly glaciated during the Wisconsinan
(Frey et al., 1965). G. bursarius per-
sisted in this region, as indicated by the
discovery of G. bursarius of Wisconsinan
age in central Indiana, southern Illinois,
Kentucky, and central Tennessee (Par-
malee and Klippel, 1981). The presence
of Geomys and Microtus ochrogaster
show that forest parkland existed south
of the glacier front in Indiana and Illi-
nois (Pannalee et al., 1978). The Wis-
consinan faunas reported by Parmalee
(op. cit. ) and Guilday et al. (1971) from
this area are much like that occurring
today in the vicinity of Minneapolis,
Minnesota, in mixed deciduous and co-
niferous forest with patches of tall-grass
prairie where G. bursarius is abundant.
The analyses of relationships discussed
above do not support the suggestion by
Pannalee and Klippel (1981) that the
now-extinct Kentucky and Tennessee
populations of G. bursarius gave rise to
G. pinetis, since these are only distantly
related.

Although the Des Moines Lobe cov-
ered much of the central part of the
current range of G. bursarius, many areas
in southeastern Iowa, southeastern Ne-
braska, and eastern Kansas were not
glaciated. The presence of grass, sage-
brush, and ragweed pollen in late Wis-
consinan lake sediments in northeastern
Kansas indicates the persistence of prai-
rie islands in conferous forest in this
region (Wright, 1970). A specimen of
late Pleistocene age from Pottawatomie
County, Kansas (KUVP 25299), that ap-
pears to us to be G. bursarius indicates
that this species occurred in the area.
It is possible that G. bursarius occurred

54

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

farther south in the central plains dur-
ing this period (i.e., outside of its pres-
ent range), but we have no evidence
of this.

Geomys lutescens may have been ab-
sent from northern Nebraska and ad-
jacent areas during the Wisconsinan
stage because of the coniferous forest
widespread there; certainly they were
absent at the close of the Wisconsinan
because the shifting sands of the Ne-
braska Sandhills (Wright, 1970) could
not have supported them. Geomys lu-
tescens probably persisted throughout
the Wisconsinan in western Kansas
where coniferous parkland was wide-
spread (Stewart, 1979); fossil Geomys
(initially referred to lutescens, but in
recent years referred to bursarius) are
known from several faunas in Meade
County, Kansas (Russell, 1968; Davis,
1975), and from Rooks County ( Mc-
Mullen, 1978). The current level of di-
vergence between G. 1. lutescens and
G. 1. major suggests that theirs is sec-
ondary intergradation, but we know of
no certain isolating factor. Perhaps a
band of continuous forest separated pop-
ulations of G. 1. lutescens in Kansas from
populations of G. I. major in Texas; de-
tailed studies of Wisconsinan gophers
from these areas and of their associated
faunas may yield sufficient information
to evaluate this hypothesis.

While Wisconsinan climatic changes
in the central and northern Great Plains
probably resulted in restrictions in the
distribution of Geomys, the cooler, wet-
ter conditions may have allowed expan-
sion in Texas, Oklahoma, New Mexico,
and northern Mexico. The southern spe-
cies of Geomys (G. arenarius, G. per-
sonatus, and G. tropicalis) may have
originated via allopatric speciation prior
to the Wisconsinan, perhaps during the
hotter, drier conditions of the Sanga-
monian. The hypothesis that they origi-
nated via parapatric ( = "stasipatric" of
some authors) speciation during the
Wisconsinan is equally possible, although
equally untested.

At the close of the Pleistocene, mas-

sive changes took place in the flora of
central North America, marked by the
extinction of many mammals, and great
changes in the distributions of most
others (Guilday, 1967; Martin and Neu-
ner, 1978). Pocket gophers of the genus
Geomys probably expanded their ranges
in most areas because the climatic
changes taking place favored the spread
of their habitat, the prairies. Geomys
lutescens expanded into relatively dry
portions of the central Great Plains in
western Nebraska and adjacent states,
and Geomys bursarius expanded into
moist portions of the plains in eastern
Nebraska, Iowa, and Minnesota. At least
two of the isolated populations survived
the Wisconsinan; these are now repre-
sented by G. /;. illinoensis and G. b. ivis-
consinensis. In the southern Great Plains,
gopher distributions undoubtedly be-
came restricted as aridity increased,
leaving relict populations of G. arenarius,
G. attwateri, G. breviceps, G. personatus,
and G. tropicalis scattered across the
southern plains. Subsequent climatic
events undoubtedly modified the distri-
butions of the gophers, but the major
features of Geomys distribution we see
today were probably determined by
about 8,000 RP.

The location of the current zone of
contact between G. bursarius and G.
lutescens in Kansas was investigated by
plotting all known localities of occur-
rence on the map of potential natural
vegetation prepared by Kuchler (1974).
The distribution of the two species is
defined with remarkable accuracy by
the line marking the western boundary
of prairie with significant forest islands.
In the southern part of the state the line
lies well to the east of the border of
tall-grass prairie, and in the northern
part of the state, especially near the
Nebraska border, the line lies to the
west of the limit of tall-grass prairie, but
in all cases the line defines the distribu-
tion of the gophers. We plotted all
known localities of Geomys in Nebraska
on the map of natural vegetation pre-
pared by Kaul (1975). All records of

POCKET GOPHERS OF THE GENUS GEOMYS

55

G. lutescens lie in mixed, sandhills, or
short-grass prairie. Most records of G.
bursarius are from areas having tall-
grass prairie, but a few records from
Adams and Knox counties are from the

transition zone between mixed and tall-
grass prairie. We have drawn our esti-
mate of the boundary line between the
two groups ( Fig. 1 ) based on these
observations.

SUMMARY

Three species of pocket gophers of
the genus Geomys exist in the Great
Plains north of Texas. The northernmost
of these, G. bursarius, consists of the
widespread, variable nominate subspe-
cies and two more restricted, less vari-
able subspecies (G. /;. illinoensis and
G. /;. wisconsinensis) . The second spe-
cies, from the central plains, G. lutescens,
has two subspecies within the study area
(G. /. lutescens and G. /. major), and
one or more extralimital subspecies. The
third species, G. breviceps, is repre-
sented in the southern plains (in Okla-
homa) by G. b. sagittalis, and by several
subspecies outside of our study area.
All species are distinguishable on the
basis of cranial morphology, karyology,
and ectoparasite fauna, and show minor
differences in bacula. All specimens pre-
viously thought to be intergrades be-
tween the three species were reexamined
and found to be clearly assignable to

one species or the other, with no evi-
dence of intergradation. Hybridization
between G. bursarius and G. lutescens
occurs only at a single locality in Ne-
braska, and introgression there is in-
consequential. Hybridization between
G. lutescens and G. breviceps probably
occurs in several places in east-central
Oklahoma, but the hybrid zones are nar-
row, and gene flow appears to be re-
stricted. Variation within all species is
mostly clinal; non-clinal variation is
thought to be due to the late-Pleistocene
history of the gophers. Analysis of cra-
nial and other characters indicates that
G. bursarius and G. lutescens are re-
cently derived from a stock similar to
G. breviceps and G. personatus. Geomys
pinctis had the most primitive features
of the species considered, and was prob-
ably separated from the others at an
early date.

LITERATURE CITED

Baker, R. H., and B. P. Glass. 1951. The
taxonomic status of the pocket gophers,
Geomys bursarius and Geomys breviceps.
Proc. Biol. Soc. Washington, 64:55-58.

Baker, R. J., and H. H. Genoways. 1975. A
new subspecies of Geomys bursarius
(Mammalia: Geomyidae) from Texas
and New Mexico. Texas Tech Univ.,
Occas. Pap. Mus. No. 29. 18 pp.

Baker, R. J., S. L. Williams, and J. C.
Patto.w 1973. Chromosomal variation
in the plains pocket gopher, Geomys
bursarius major. J. Mammal., 54:765-
769.

Blossom, P. M. 1938. Description of a new
race of pocket gopher (Geomys lutescens
hylaeus) from northwestern Nebraska.
Univ. Michigan, Mus. Zool., Occas. Pap.
No. 368. 2 pp.

Bohlix, R. G., and E. G. Zimmerman. 1982.
Genie differentiation of two chromosome
races of the Geomys bursarius complex.
J. Mammal., 63:218-228.

Bowles, J. B. 1975. Distribution and biogeog-
raphy of mammals of Iowa. Texas Tech
Univ., Spec. Pub. Mus. No. 9. 184 pp.

Burt, W. H. 1960. Bacula of North American
mammals. Univ. Michigan, Mus. Zool.,
Misc. Publ. No. 113. 76 pp., 25 pi.

Carleton, M. D. 1980. Phylogenetic relation-
ships in Neotomine-Peromyscine rodents
( Muroidea ) and a reappraisal of the
dichotomy within New World Cricetinae.
Univ. Michigan, Mus. Zool., Misc. Publ.
No. 157. 146 pp.

Davis, L. C. 1975. Late Pleistocene geology
and paleoecology of the Spring Valley
Basin, Meade County, Kansas. Unpubl.
Ph.D. Dissertation, Univ. Iowa, Iowa
City. 170 pp.

DeBlase, A. F., and R. E. Martin. 1974. A
manual of mammalogy with keys to
families of the world. Wm. C. Brown
Company Publishers, Dubuque, Iowa,
xv + 329 pp.

Dixon, W. J. [ed.]. 1975. BMDP Biomedical

56

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

computer programs. University of Cali-
fornia Press, Berkeley. 792 pp.

Dort, W., Jr., and J. K. Jones, Jr. (eds.).
1970. Pleistocene and Recent environ-
ments of the Central Great Plains. Univ.
Kansas, Dept. Geol., Spec. Publ. No. 3.
433 pp.

Frey, J. C, H. B. Willmax, and R. F. Black.
1965. Outline of glacial geology of Illi-
nois and Wisconsin. Pp. 43-61, in The
Quaternary of the United States, H. E.
Wright, Jr., and D. G. Frey, eds. Prince-
ton Univ. Press, Princeton, x + 922 pp.

Genoways, H. H. 1973. Systematica and evo-
lutionary relationships of spiny pocket
mice, genus Liomys. Texas Tech Univ.,
Spec. Publ. Mus. No. 5. 368 pp.

Gray, A. P. 1972. Mammalian hybrids. A
check-list with bibliography. Common-
wealth Agric. Bur., Farnham Royal
(England), x + 262 pp.

Gray, F. 1959. Soil map (generalized) of
Oklahoma. Oklahoma Agric. Exp. Sta.,
Stillwater, map.

Guilday, J. E. 1967. Differential extinction
during late-Pleistocene and Recent times.
Pp. 121-140, in Pleistocene extinctions:
the search for a cause, P. S. Martin and
H. E. Wright, Jr. eds. Yale Univ. Press,
New Haven, x + 453 pp.

Guilday, J. E., H. W. Hamilton, and A. D.
McGrady. 1971. The Welsh Cave pec-
caries (Platygonus) and associated fauna,
Kentucky Pleistocene. Ann. Carnegie
Mus., 43:249-319.

Hall, E. R. 1981. The mammals of North
America, vol. 1. John Wiley & Sons,
New York, xv + 600 + 90 pp.

Harper, F. 1952. History and nomenclature
of the pocket gophers (Geomys) in
Georgia. Proc. Biol. Soc. Washington,
65:35-38.

Hart, E. B. 1978. Karyology and evolution of
the plains pocket gopher, Gcomijs bur-
sarius. Univ. Kansas, Mus. Nat. Hist.,
Occas. Pap. No. 71. 20 pp.

Hay, O. P. 1923. The Pleistocene of North
America and its vertebrated animals
from the states east of the Mississippi
River and from the Canadian provinces
east of longitude 95°. Carnegie Inst.
Washington, No. 322. viii + 499 pp.

Heaney, L. R. 1979. Taxonomy and hybrid-
ization of Great Plains pocket gophers:
a study of mammalian speciation. Un-
publ. Ph.D. Dissertation, Univ. Kansas,
Lawrence. 101 pp.

Hendricksen, R. L. [1972]. 1973. Variation
in the plains pocket gopher ( Geomys
bursarius) along a transect across Kansas
and eastern Colorado. Trans. Kansas
Acad. Sci., 75:322-368.

Hibbard, C. W. 1970. Pleistocene mammalian
local faunas from the Great Plains and

central lowland provinces of the United
States. Pp. 395-433, tn Pleistocene and
Recent environments of the Central
Great Plains, W. Dort, Jr., and J. K.
Jones, Jr., eds. Univ. Kansas, Dept. Geol.
Spec. Publ. No. 3. 433 pp.

Hibbard, C. W., R. J. Zakrzewski, R. E.
Eshelman, G. Edmund, C. D. Griggs,
and C. Griggs. 1978. Mammals from
the Kanopolis local fauna, Pleistocene
( Yarmouth ) of Ellsworth County, Kan-
sas. Univ. Michigan Mus. Paleontol.,
Contrib., 25:11-44.

Honeycutt, R. L., and D. J. Schmidly. 1979.
Chromosomal and morphological varia-
tion in the plains pocket gopher, Geomys
bursarius, in Texas and adjacent states.
Texas Tech Univ., Occas. Pap. Mus. No.
58. 54 pp.

Hooper, E. T. 1940. A new race of pocket
gopher of the species Geomys lutescens
from Colorado. Univ. Michigan, Mus.
Zool., Occas. Pap. No. 420. 3 pp.

Jackson, H. II. T. 1961. Mammals of Wis-
consin. Univ. Wisconsin Press, Madison.
xii + 504 pp.

Jones, J. K., Jr. 1964. Distribution and tax-
onomy of mammals of Nebraska. Univ.
Kansas Publ., Mus. Nat. Hist. No. 16.
356 pp.

Kaul, R. B. 1975. Vegetation of Nebraska
(circa 1850). Univ. Nebraska Conserv.
Survey Div., Inst. Agric. Nat. Res. Map.

Kennerly, T. E., Jr. 1958. The baculum in
the pocket gopher. J. Mammal., 39:
445-446.

Kluge, A. O. 1976. Phylogenetic relationships
in the lizard family Pygopodidae: an
evaluation of theory, methods and data.
Univ. Michigan, Mus. Zool., Misc. Publ.
No. 152. 72 pp.

KOMAREK, E. V., AND D. A. SPENCER. 1931.

A new pocket gopher from Illinois and
Indiana. J. Mammal., 12:404-408, 1 pi.

Kuchler, A. W. 1964. Potential natural vege-
tation of the conterminous United States.
American Ceogr. Soc, New York, Spec.
Publ. No. 36. v + 1-39 + map + 116
pis.

Kuchler, A. W. 1974. A new vegetation map
of Kansas. Ecology, 55:586-601, suppl.

Kurten, B., and E. Anderson. 1980. Pleisto-
cene mammals of North America. Co-
lumbia Univ. Press, New York, xvii +
442 pp.

Lemke, R. W., W. M. Laird, M. J. Tipton,
and R. M. LiNirvALL. 1965. Quarter-
nary geology of northern Great Plains.
Pp. 15-27, in The Quaternary of the
United States. H. E. Wright, Jr., and
D. G. Frey, eds. Princeton Univ. Press,
Princeton, x + 922 pp.

Lowery, G. H., Jr. 1974. The mammals of
Louisiana and its adjacent waters. Lou-

POCKET GOPHERS OF THE GENUS GEOMYS

57

isiana State Univ. Press, xxiii -f- 565 pp.

Martin, L. D., and A. M. Neuner. 1978. The
end of the Pleistocene in North America.
Trans. Nebraska Acad. Sci., 6:117-126.

McLaughlin, C. A. 1958. A new race of the
pocket gopher Geomys bursarius from
Missouri. Los Angeles Co. Mus., Contr.
Sci. No. 19. 4 pp.

McMullen, T. L. 1978. Mammals of the
Duck Creek local fauna, late Pleisto-
cene of Kansas. J. Mammal., 59:374-
386.

Merhiam, C. H. 1895. Revision of the pocket
gophers, family Geomyidae, exclusive of
the species of Thomomys. N. American
Fauna, No. 8. 258 pp., 4 pi.

Moore, W. S. 1977. An evaluation of narrow
hybrid zones in vertebrates. Quart. Rev.
Biol., 52:263-277.

Neff, N. A., and G. R. Smith. 1979. Multi-
variate analysis of hybrid fishes. Syst.
Zool., 28:176-196.

Parmalee, P. W., and W. E. Klippel. 1981.
A late Pleistocene population of the
pocket gopher, Geomys cf. bursarius, in
the Nashville Basin, Tennessee. J. Mam-
mal., 62:831-835.

Parmalee, P. W., P. J. Munson, and J. E.
Guilday. 1978. The Pleistocene mam-
malian fauna of Harrodsburg Crevice,
Monroe County, Indiana. Bull. Nat.
Speleolog. Soc., 40:64-75.

Patton, J. L., and J. H. Feder. 1978. Genetic
divergence between populations of the
pocket gopher, Thomomys umbrinus
(Richardson). Z. Siiugetierk., 43:17-30.

Patton, J. L., and J. H. Feder. 1981. Micro-
spatial genetic heterogeneity in pocket
gophers: non- random breeding and
drift. Evolution, 35:912-920.

Pattox, J. L., J. C. Hafner, M. S. Hafner,
axd M. F. Smith. 1979. Hybrid zones
in Thomomys bottac pocket gophers:
genetic, phenetic, and ecologic con-
cordance patterns. Evolution, 33:860-
876.

Pattox, J. L., axd M. F. Smith. 1981.
Molecular evolution in Thomomys: phy-
letic systematics, paraphyly, and rates
of evolution. J. Mammal., 62:493-500.

Pattox, J. L., axd S. Y. Yaxg. 1977. Genetic
variation in Thomomys bottae pocket
gophers: macrogeographic patterns.
Evolution, 31:697-720.

Price, R. D., axd K. C. Emerson. 1971. A
revision of the genus Geomydoccus
(Mallophaga: Trichodectidae ) of the
New World pocket gophers (Rodentia:
Geomyidae). J. Med. Entomol., 8:228-
257.

Rasmussen, D. L. 1971. Microvertebrates from
a fissure deposit in the "driftless area"
of southwestern Wisconsin (abstr. ).

Proc. North-Central Sec, Geol. Soc.
Amer., 3:275-276.

Russell, R. J. 1968. Evolution and classifica-
tion of the pocket gophers of the sub-
family Geomyinae. Univ. Kansas Publ.,
Mus. Nat. Hist., 16:473-579.

Russell, R. J., axd J. K. Jones, Jr. 1956. The
taxonomic status of Geomys bursarius
vinaceus Swenk. Trans. Kansas Acad.
Sci., 58:512-513.

Sherman, H. B. 1940. A new species of pocket
gopher (Geomys) from eastern Georgia.
J. Mammal., 21:341-343.

Stewart, J. D. 1979. Paleontology and paleo-
ecology of the Trapshoot Local Fauna,
Rooks County, Kansas. Unpubl. M.A.
Thesis, Univ. Kansas, Lawrence. 146 pp.

Swexk, M. H. 1939. A study of local size
variation in the prairie pocket gopher
(Geomijs bursarius), with description of
a new subspecies from Nebraska. Mis-
souri Valley Fauna, 1:1-8.

Swexk, M. H. 1940. A study of subspecific
variation in the yellow pocket gopher
(Geomys lutescens) in Nebraska, and of
the geographical and ecological distribu-
tion of the variants. Missouri Valley
Fauna, 2:1-12.

Thaeler, C. S., Jr. 1974. Four contacts be-
tween ranges of different chromosome
forms of the Thomomys talpoides com-
plex (Rodentia: Geomyidae). Syst.
Zool., 23:343-354.

Timm, R. M. 1979. The Geomydoecus (Mal-
lophaga: Trichodectidae) parasitizing
pocket gophers of the Geomys complex
(Rodentia: Geomyidae). Unpubl. Ph.D.
Dissertation, Univ. Minnesota, St. Paul.
124 pp.

Timm, R. M. 1983. Farenholz's Rule and Re-
source Tracking: a study of host-para-
site coevolution. Pp. 225-266, in Co-
evolution. M. H. Nitecki, ed. Univ.
Chicago Press, Chicago. 362 pp.

Timm, R. M., E. B. Hart, axd L. R. Heaney.
1982. Karyotypic variation in pocket go-
phers (Geomyidae: Geomys) from a
narrow contact zone in Nebraska. Mam-
mal. Chromosomes Newsletter, 23:108-
117.

Timm, R. M., and R. D. Price. 1980. The
taxonomy of Geomydoecus (Mallophaga:
Trichodectidae ) from the Geomys bur-
sarius complex (Rodentia: Geomyidae).
J. Med. Entomol, 17:126-145.

Tucker, P. K., and D. J. Schmidly. 1981.
Studies of a contact zone among three
chromosomal races of Geomys bursarius
in east Texas. J. Mammal., 62:258-272.

Villa-R., B., and E. R. Hall. 1947. Sub-
speciation in pocket gophers of Kansas.
Univ. Kansas Publ., Mus. Nat. Hist.,
1:217-236.

Williams, S. L., and H. H. Genoways. 1980.

58

MISCELLANEOUS PUBLICATION MUSEUM OF NATURAL HISTORY

Morphological variation in the south-
eastern pocket gopher, Gcomys pinetis
(Mammalia: Rodentia). Ann. Carnegie
Mus., 49:405-453.
Wright, H. E., Jr. 1970. Vegetational history
of the Central Plains. Pp. 157-172, in
Pleistocene and Recent environments of
the Central Great Plains, W. Dort, Jr.,

and J. K. Jones, Jr., eds. Univ. Kansas,
Dept. Geol, Spec. Publ. No. 3. 433 pp.
Wright, H. E. Jr., and R. V. Ruhe. 1965.
Glaciation of Minnesota and Iowa. Pp.
29-41, in The Quaternary of the United
States. H. E. Wright, Jr., and D. G.
Frey, eds. Princeton Univ. Press, Prince-
ton, x + 922 pp.

ADDENDUM

Several important papers have ap-
peared since this manuscript was ac-
cepted for publication. In the first,
Honey cutt and Williams (1982) used
starch-gel electrophoresis to examine in-
tergeneric relationships within the sub-
family Geomyinae. Both a cladistic anal-
ysis and a cluster analysis using similarity
coefficients indicated that Geomys is one
of the earliest lateral branches in the
subfamily, but is generally most similar
to Zy go geomys, also an early lateral
branch. They suggested that Orthogeo-
mys is the sister-group of Zy go geomys.

Williams (1982) described the phalli
of geomyids in detail, including those of
the taxa referred to here as Geomys
arenarius, G. attwateri, G. breviceps
sagittalis, G. lutescens lutescens, G. lu-
tescens major, G. personatus, and G.
pinetis. A cluster analysis of a set of
phallus measurements suggested that G.
breviceps is the sister-group of G. lu-
tescens; that G. arenarius is the sister-
group to that lineage; that G. attwateri
is the sister-group to G. "lutescens"
Uanensis and texensis; and that G. pinetis
and G. personatus lie outside of this
group. Because no G. bursarius as de-
fined here were included, no direct com-
parison to our phylogeny of the G. bur-
sarius group is possible. We suggest
that a phenetic analysis that restricts the

effect of size and a cladistic analysis of
these data, with a sample of G. bursarius
added, would comprise a suitable test
of our proposed phylogeny.

Finally, Hafner (1982) used electro-
phoretic and immunological data to ex-
amine relationships of the Geomyoidea.
His data support the monophyly of the
Geomyidae and Geomyinae. Unlike the
results of Honeycutt and Williams
( 1982 ) , he found evidence for consider-
ing Pappogeomys to be the sister-group
of Geomys, but in general found his data
inadequate for consideration of generic
or subgeneric relationships.

In summary, although these studies
add substantially to the perspective of
this paper, they do not affect our con-
clusions regarding the relationships of
species in the Geomys bursarius species
group.

Hafner, M. S. 1982. A biochemical investiga-
tion of geomyoid systematics (Mammalia:
Rodentia). Z. Zool. Syst. Evolutionforsh.,
20:118-130.

Honeycutt, R. L., and S. L. Williams. 1982.
Genie differentiation in pocket gophers
of the genus Pappogeomys, with com-
ments on intergeneric relationships in
the subfamily Geomyinae. J. Mammal.,
63:208-217.

Williams, S. L. 1982. Phalli of Recent genera
and species of the family Geomyidae
(Mammalia: Rodentia). Bull. Carnegie
Mus. Nat. Hist., 20:1-62.

APPENDIX I

All specimens from the zone of con-
tact between Geomys bursarius and G.
lutescens near Oakdale, Antelope Coun-
ty, Nebraska, utilized in the preceding
analyses are listed here.

Geomys bursarius

Specimens examined. — Nebraska:
Antelope Co.: 2Vw mi. S Oakdale (1 SM);
W edge Oakdale (8 SM); Y2 mi. W Oak-

POCKET GOPHERS OF THE GENUS GEOMYS

59

dale (3 SM); %o mi. S, 7/io mi. W Oak-
dale (6KU); %o mi. S, 9io mi. W Oak-
dale (5 KU).

Geomys bursarius X Geomijs Jutescens

Specimens examined. — Nebraska:
Antelope Co.: %o mi. S, 7/io mi. W Oak-
dale ( 1 KU ) ; %o mi. S, 6/io mi. W Oakdale
(1 KU); Mo mi. N, f/io mi. W Oakdale
(3 KU); %o mi. N, %o mi. W Oakdale
(2 KU); 1 mi. W Oakdale (16 KU, 6
SM); (vicinity of) Oakdale (1 USNM).

Geomys lutescens

Specimens examined. — Nebraska:
Antelope Co.: %o mi. S, 1 mi. W Oakdale
(1 KU); 740 mi. N, 1 mi. W Oakdale
(5 KU); Mo mi. N, lMo mi. W Oakdale
(1 KU); IMo mi. W Oakdale (1 KU)
Mo mi. N, l%o mi. W Oakdale (2 KU)
-io mi. S, l%o mi. W Oakdale (3 KU)
ltto mi. W Oakdale (9 KU); 1% mi. W
Oakdale (1 SM); 2 mi. W Oakdale (1
SM); (vicinity of) Oakdale (1 USNM).

QL737.R654 H43 1983

Relationships of pockcl gophers ol

Harvard MCZ Library VHC078

3 2044 062 388 848

RECENT MISCELLANEOUS PUBLICATIONS
UNIVERSITY OF KANSAS MUSEUM OF NATURAL HISTORY

52. Reproductive cycles in lizards and snakes. By Henry S. Fitch. Pp. 1-247, 16 fig-
ures in text. June 19, 1970. Paper bound.

53. Evolutionary relationships, osteology, and zoogeography of leptodactyloid frogs.
By John D. Lynch. Pp. 1-238, 131 figures in text. June 30, 1971. Paper bound.

55. Middle American lizards of the genus Ameiva (Teiidae) with emphasis on geo-
graphic variation. By Arthur C. Echternacht. Pp. 1-86, 28 figures in text. De-
cember 14, 1971. Paper bound.

57. A systematic review of the Teiid lizards, genus Bachia, with remarks on Heter-
odactylus and Anotosaura. By James R. Dixon. Pp. 1-47, 15 figures in text.
February 2, 1973. Paper bound.

59. Systematics and evolution of the Andean lizard genus Tholidobolus (Sauria:
Teiidae). By Richard R. Montanucci. Pp. 1-52, 8 figures in text. May 14, 1973.
Paper bound.

61. Reproductive strategies in a tropical anuran community. By Martha L. Crump.
Pp. 1-68, 13 figures in text. November 15, 1974. Paper bound.

62. A demographic study of the ringneck snake (Diadophis punctatus) in Kansas. By
Henry S. Fitch. Pp. 1-53, 19 figures in text. April 3, 1975. Paper bound.

65. The biology of an equatorial herpetofauna in Amazonian Ecuador. By William
E. Duellman. Pp. 1-352, 198 figures in text. August 30, 1978. Paper bound.

66. Leptodactylid frogs of the genus Eleutherodactylus from the Andes of southern
Ecuador. By John D. Lynch. Pp. 1-62, 23 figures in text. February 28, 1979.
Paper bound.

67. An ecogeographic analysis of the herpetofauna of the Yucatan Peninsula. By
Julian C. Lee. Pp. 1-75, 27 plates, 22 figures in text. February 29, 1980. Paper
bound.

68. Internal oral features of larvae from eight anuran families: Functional, systematic,
evolutionary and ecological considerations. By Richard Wassersug. Pp. 1-146,
37 figures in text. June 24, 1980. Paper bound.

69. The Eleutherodactylus of the Amazonian slopes of the Ecuadorian Andes ( Anura:
Leptodactylidae ) . By John D. Lynch and William E. Duellman. Pp. 1-86, 8 fig-
ures in text. August 29, 1980. Paper bound.

70. Sexual size differences in reptiles. By Henry S. Fitch. Pp. 1-72, 9 figures in text.
February 27, 1981. Paper bound.

71. Late Pleistocene herpetofaunas from Puerto Rico. By Gregory Pregill. Pp. 1-72,
26 figures in text. May 8, 1981. Paper bound.

72. Leptodactylid frogs of the genus Eleutherodactylus in the Andes of northern
Ecuador and adjacent Colombia. By John D. Lynch. Pp. 1-46, 22 figures in text.
July 8, 1981. Paper bound.

73. Type and figured specimens of fossil vertebrates in the collection of the Uni-
versity of Kansas Museum of Natural History. Part I. Fossil fishes. By H.-P.
Schultze, J. D. Stewart, A. M. Neuner and R. W. Coldiron. Pp. 1-53. October 6,
1982. Paper bound.