FOSSIL MAMMALS OF THE COLEMAN HA
LOCAL FAUNA, SUMTER COUNTY, FLORIDA

By
ROBERT ALLEN MARTIN

A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF

THE UNIVERSITY OF FLORIDA

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE

DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA
1969

ACKNOWLEDGMENTS

Thanks go to numerous colleagues who have generously loaned me
both fossil and Recent specimens of Sigmodon and other mammals, or who
have allowed me to view collections in their care, and most are noted
in the Abbreviations section after the institution with which they are
affiliated. I thank especially Dr. and Mrs. Robert Miller and Dr. and
Mrs. Robert Hoffmann for allowing me to live in their homes while on a
recent museum tour. I remain greatly indebted to my parents, Mr. and
Mrs. Gerald M. Martin, for constant encouragement and funds.

This research was supported in part by N. I. H. and N. S. F.
summer fellowships (1966, 1967), a small grant from the Society of the
Sigma Xi, and a small allotment from the Department of Zoology, Univer-
sity of Florida.

TABLE OF CONTENTS

Acknowledgments ii

List of Tables v

List of Figures vii

Description of Measurements and Abbreviations ix

Introduction 1

Species List 8

Species Account 9

Class Mammalia

Order Marsupialia

Didelphis marsupialis 9

Order Insectivora

Cryptotis parva 9

Blarina brevicauda 9

Scalopus aquaticus 11

Order Chiroptera

Myotis cf M. austroriparius 11

Pipistrellus subflavus 11

Plecotus rafinesquii 12

Order Edentata

+Dasypus bellus 13

-fHolmesina septentrionalis 16

Order Lagomorpha

Lepus alleni 19

Sylvilagus sp. 21

Order Rodentia

Sciurus carolinensis 24

Glaucomys sp. 24

Geomys c f G. pinetis 33

Reithrodontomys humulis 33

Peromyscus floridanus 33

Ochrotomys nuttalli 34

Peromyscus sp. 34

+Sigmodon bakeri , new species 35

Key to the Extinct and Extant Species of Sigmodon 40

Account of the Extinct and Extant Species of Sigmodon 42

Genus Sigmodon 42

hispidus species group 42

Sigmodon hispidus 42

Sigmodon alleni 44

Sigmodon ochrognathus 44

TABLE OF CONTENTS - continued

Sigmodon fulviventer 44

+Sigmodon bakeri 44

leucotis species group 45

Sigmodon leucotis 45

Sigmodon peruanus 45

+Sigmodon curtisi 46

+Sigmodon hudspethensis 47

alstoni species group 48

Sigmodon alstoni 48

■fmedius species group 48

+Sigmodon medius 48

+Sigmodon roedius medius , new subspecies 50

+Sigmodon medius hibbardi, new subspecies -- 50

+Sigmodon medius, subspecies indeterminate -- 51

+Sigmodon minor 52

Evolution in the Genus Sigmodon 52

I. Dental Evolution from Pliocene to Recent Time - 54

II. Ecological Considerations 63

III. Fossil Material Examined 64

+Pitymys arata, new species 81

Neofiber alleni 88

Erethizon dorsatum 93

+Hydrochoerus 98

Order Carnivora

+Arctodus pristinus 98

Spilogale putorius 99

Conepatus sp. 99

Mephitis mephitis 99

Procyon cf. P. lotor 100

+Urocyon minicephalus , new species 101

Canis lupus 102

Felis onca - 109

Felis rufus 121

Order Artiodactyla

+Platygonus cumberlandensis 127

+My lohyus sp. 129

+Tanupolama cf. mirifica 142

Odocoileus virginianus 143

Order Perissodactyla

+Equus sp. 143

Order Proboscidea

+Mammuthus sp. 144

Age and Correlation 153

Affinities of the Coleman IIA Mammals 159

Paleoecology 16 7

Summary and Conclusions 170

Literature Cited 172

Biographical Sketch 184

LIST OF TABLES

1. Measurements (in mm) of lower dentition and mandible
of Recent and fossil Blarina brevicauda and Cryptotis

parva 10

2. Distance (in mm) from anterior border of canine alveolus
to posterior border of Mo alveolus in select fossils

and Recent specimens of Myotis 12

3. Measurenents (in mm) of some fossil and living

rabbit femora 20

4. Measurements (in mm) of mandibles and poster anial elements

of Recent and fossil Sciurus niger and S. carolinensis 26

5. Measurements (in mm) of mandibles of Recent and fossil
Glaucomys sabrinus and Glaucomys volans 32

6. Measurements (in mm) of lower dentition and mandible

in living and extinct species of Sigmodon 69

7. Measurements (in mm) of the lower dentition and
mandible of Sigmodon samples from the Vallecito

Creek-Fish Creek beds 79

8. Statistical comparison of Sigmodon samples from
the Wendell Fox Pasture (WFP) and Rexroad Loc.

3 (R3) deposits 80

9. Mandibular dimensions (in mm) of some Recent and

fossil samples of the genera Microtus and Pitymys 91

10. Measurements (in mm) of skunk mandibles 100

11. Measurements (in mm) of fossil and Recent

fox skulls 105

12. Measurements (in mm) of upper fourth premolar and

upper first molar in fossil and Recent wolves 116

13. Measurements (in mm) of fossil and Recent

wolf skulls 118

LIST OF TABLES -continued

14. Measurements (in mm) of the lower dentition of some

extinct and living cats 124

15. Measurements (in mm) of the lower dentition of fossil

and Recent jaguars 126

16. Measurements (in mm) of fossil dental samples of

Platygonus 131

17. Tanupolama ("short-limbed" species): measurements

(in mm) of metapodials 151

18. Tanupolama ("long- limbed" species); measurements

(in mm) of metapodials 152

19. Relationships of Coleman IIA mammals 165

LIST OF FIGURES

1. The major Pleistocene local faunas of Florida 5

2. Scatter diagram relating Dasypus bellus to Dasypus
novemcinctus when depth of the movable dermal plates

is plotted against width of these plates 15

3. Femora of Lepus alleni (UF 13178; right) and

Sylvilagus sp. (UF 13176; left) from Coleman HA 24

4. Scatter diagram relating Glaucomys sabrinus
(large circles) to G. volans (small circles) when

RA length is plotted against MA length 31

5. Lower teeth of some fossil and Recent cricetid rodents 38

6. A possible phylogeny for the rodent genus Sigmodon 58

7. Correlation of Kansas and Arizona deposits containing

members of the medius species group of Sigmodon 60

8. Composite ratio diagram of measurements of the
lower dentition and mandible in some members of the
Sigmodon medius species group from Kansas and

Arizona 62

9. The dentition of Pitymys arata and Equus sp. from
the Coleman HA fauna and femora of Lepus alleni

and Lepus towns endii 90

10. Upper dentition of fossil porcupine, UF 11774,

from Coleman HA 97

LIST OF FIGURES - continued

11. Recent and fossil skulls of foxes 104

12. Skulls of Canis lupus from Coleman HA 111

13. Ventral view of skull of Canis lupus from Coleman

HA (UF 11519) - 113

14. Scatter diagram relating Pleistocene and Recent
wolves when the length of the P^ (X axis) is plotted

against the width of the P^ (Y axis) 115

15. Ratio diagram comparing the lower dentitions of

various felines 123

16. Palate and partial skulls of Platygonus
cumberlandensis from Coleman HA

17. Scatter diagram relating various samples of
fossil Platygonus when length of the M (X axis) is

plotted against the width of the M, (Y axis) 141

18. Labial view of the lower dentition and mandible

of Tanupolama from Coleman HA (UF 11985) 146

19. Occlusal view of lower dentition and mandible of

Tanupolama from Coleman HA (UF 11983) 148

20. Regression analysis relating two species of

camels from the Pleistocene of Florida when proximal

width of the metapodial (X axis) is plotted against the

greatest length of the metapodial (Y axis) 150

21. Replacement of mammals in Florida during the

Pleistocene 158

MEASUREMENTS AND ABBREVIATIONS

I. Measurements

Sciurus and Glaucomys

Innominate length - greatest length

Innominate width - greatest width with the calipers held
perpendicular to the long axis of the
innominate
Femur length - greatest length
Femur width (distal) - distance between the medial and

lateral condyles
Tibia length - greatest length
Tibia width (proximal) - distance between the medial and

lateral condyles
Humerus length - greatest length
Humerus width (proximal) - distance between the greater and

lesser tuberosities
MA length - mandibular alveolar length
RA length - ramus alveolar length; distance from the greatest

angle of the posterior border of the ascending ramus
to the anterior border of the alveolus of the first
lower molar (see Martin, 1967)
Cryptotis and Blarina

Condyloid to mental foramen - distance from the most posterior

projection of the condyloid process
to the center of the mental foramen

Cryptotis and Blarina - continued

Condyloid to Mi - distance from the most posterior projection of
the condyloid process to the anterior border
of M,

Sigmodon

Length M3 - the entire visible length of the tooth in occlusal

view; not merely the occlusal surface
Width M , M„, M3- the entire visible width of the teeth in occlusal
view; not merely the occlusal surface

Canis

Width P4 and M1 - taken by first placing one edge of the calipers
flush with the labial border of the paracone and
metacone, then bringing the other ediege into
contact with the protocone

All multiple tooth measures designated by letters (e. g. , Kx - M2)
are measurements taken from the anterior border of the first tooth to
the posterior border of the second, inclusive; each is not measured
separately.

Measurements were taken with a Helios 7-inch calipers and a
Gaertner platform measuring microscope.

A cross (+) preceding the name of a taxon indicates that it is
extinct.

II Abbreviations

UT - University of Florida, Florida State Museum (Walter Auffenberg,

S. David Webb, Thomas Patton)

x

FGS - Florida Geological Survey (Stanley Olsen)

FDT - Florida Diving Tours

UT or UTMM - University of Texas, Texas Memorial Museum (Ernest Lundelius .

Jr.) .„

UMMVT - University of Michigan Museum, Vertebrate Paleontology (Claude

Hibbard)
LACM - Los Angeles County Museum (Theodore Downs, John White)
UAVP - University of Arizona, Vertebrate Paleontology (Everett Lindsay)
MUVP - Midwestern University, Vertebrate Paleontology (Walter Dalquest)
UKMVP - University of Kansas Museum, Vertebrate Paleontology (Theodore

Eaton, Orville Bonner)
UK - University of Kansas Museum, Natural History (J. Knox Jones, Robert

Hoffmann)
UCMVP - University of California at Berkeley Museum, Vertebrate

Paleontology (Donald Savage)
UCMVZ - University of California at Berkeley, Museum of Vertebrate

Zoology (William Lidicker)
AMNH - American Museum of Natural History (Richard Tedford, Malcolm

McKenna, Richard Van Gelder, Sydney Anderson, Guy Musser)
USNM - United States National Museum (Clayton Ray, Nicholas Hotton,
Charles Handley, Henry Setzer, Ronald Pine, John Paradiso)

xi

INTRODUCTION

Florida has produced, and still continues to produce, some of the
richest Pleistocene deposits in the world. Yet these deposits have,
until very recently, defied interpretation. This is particularly true
of the mammalian faunas; less so of some of the reptilian faunas.
Auffenberg's (1958) study of the genus Terrapene is the only published
attempt to correlate the numerous isolated Florida deposits both in time
and in relation to major eustatic changes in sea level. However, of
those Pleistocene faunas considered by Auffenberg (1958), only six--
Vero (Weigel, 1962), Williston (Holman, 1959), Sabertooth Cave (Simpson,
1928), Seminole Field (Simpson, 1929), Reddick IA (Gut and Ray, 1963),
and Melbourne (Ray, 1958)--had been studied in any detail with respect
to their mammalian component. In addition, Brooks (1968) presented evi-
dence that some of the ancient shorelines considered by Auffenberg (1958)
to be of Pelistocene age are probably much more archaic. Nonetheless,
my studies of the Florida Pleistocene mammalian faunas confirm most of
Auffenberg's conclusions regarding the chronological ordering of these
faunas .

The Coleman IIA local fauna represents a particularly important,
previously unsampled, time period in Florida's history, and has provided
the information necessary to develop a coherent picture of mammalian
turnover in Florida during the Ice Ages (Pleistocene). Therefore it
seems reasonable to familiarize the reader with those other Florida

-1-

Pleistocene faunas which I will constantly mention in the text. The fol-
lowing is a list of these faunas. Exact locations and geologic setting
are given only if not previously described. All localities are mapped
in Figure 1 and the numbers preceding the localities refer to their
location on the latter map:

2) Ichetucknee River, Columbia County (Simpson, 1930; Auffenberg, 1963;

Kurten, 1965; McCoy, 1963; Martin, 1969a)

3) Santa Fe River, Gilchrist County (Martin, 1969a)

Localities 1, IB, 2, 4A, 8A

6) Haile, Alachua County (Brodkorb, 1953; Auffenberg, 1963; Ligon, 1965)

Localities VIIA, VIIIA, XIA, IXB, XIIIA, XIIIB, XIIIC,
XIVA, XIVB, XVA

4) Arredondo, Alachua County (Bader, 1957; Brodkorb, 1959; Auffenberg,

1963)

Localities IA, IIA, IIB

7) Willis ton, Levy County (Holman, 1959)

Locality III

8) Devil's Den, Levy County (Arata, 1961; Kurten, 1965)

9) Reddick, Marion County (Brodkorb, 1957; Gut and Ray, 1963)

Localities IA, IB, IIC
13) Sabertooth Cave, Citrus County (Simpson, 1928; Auffenberg, 1963)
11) Withlacoochee River, Citrus County (Martin, 1968a, 1969a)

Locality 7A
20) Eichelberger Cave, Marion County (Auffenberg, 1963)

Localities A and B

18) Bradenton Field and 51st Street, Manatee County (Simpson, 1929;

Auffenberg, 1963)

14) Seminole Field, Pinellas County (Simpson, 1929; Auffenberg, 1963)
5) Payne's Prairie, Alachua County (Auffenberg, 1963; Martin, 1969a)

16) Melbourne, Brevard County (Gazin, 1950; Ray, 1958
star) Coleman, Sumter County

Locality IIA (this report)

Locality III; A, B, C, D: the west wall of Pit III is
composed of a series of superimposed fresh water
marls; the letters stand for four recognizable
spring heads .

15) Pool Branch, Polk County

Fossils from this locality were recovered from a canal

bank east of the Peace River near the Clear Spring
Mine; near the town of Fort Meade.

19) Punta Gorda, Charlotte County

Fossils were recovered from the bank of Alligator Creek
in the N edge SW-1/4 NE-1/4 Sec. 26, T 41 S, R
23 E of the Cleveland Quadrangle.
1) Aucilla River, border of Madison and Jefferson Counties

Localities from this river and their mammalian contents
have not been described.
12) Inglis, Citrus County

Locality IA; The fossil material from this local-
ity comes from a limestone sinkhole, filled with
beach sands, exposed on the north bank of the
Florida Barge Canal.

Figure 1 -- The major Pleistocene local faunas of Florida.
The numbers on the map correspond to the same
numbers preceding the names of the faunas in
the list of localities.

-5-

I NASSAU """••'"n

-6-

Lime -mining operations provide some of the most productive fossil
localities in Florida. These operations are concentrated on the Ocala
Arch, an uplifted limestone ridge in the center of the state including
formations of Eocene through Pleistocene age. The Coleman HA fauna
was removed from a filled sinkhole in the Ocala Eocene limestone at the
Dixie Lime and Stone Corporation at Coleman, Sumter County, Florida
(SE-1/4, NW-1/4, Sec. 7, T 20 S, R 23 E; 70-90 feet above present sea
level; Figure 1).

The original Coleman HA deposit no longer exists; it was com-
pletely destroyed by further mining operations. Mr. Robert Allen, one
of the first collectors at the site, provided me with the following de-
scription of the locality. The sink was approximately 30 yards long and
25 yards wide. Two facies, a wet orange-brown clay and a coarse gray-
white sand, were obvious in this perspective. Although the clay facies
appeared homogeneous, the sand facies was filled with pebble- to boulder-
sized limestone rubble. The bone removed from the clay was usually
colored white or orange, with all gradations in between; the bone from
the gray-white sand was characteristically black, occasionally white.
Bones also transgressed these facies boundaries, and are characteris-
tically colored half orange-brown and half black, testifying to the
contemporaneity of both lithologic units. Wet bone from the sand was
much harder than that from the clay upon initial removal. None of the
bone appears to have been tumbled, and semi-articulated (or at least
closely associated) skeletal materials were removed from both clay and
sand facies. This suggests a lack of transport by water or any other
agent .

An unpublished manuscript of Norm Tessman shows that the long
bones of all species of the large mammals recovered from the Coleman
IIA deposit suffered pre-fossilization breaks. Pre-fossilization breaks,
or "green" breaks, are recognized by their rough and oblique nature and
microscopic evidence of lacunae collapse along the broken surface (Tess-
man compared the fossil bones with some intentionally broken in the lab-
oratory and some known to have accumulated in a Recent sinkhole for con-
firmation of these characteristics). Bone breakage due to post-fossil-
ization phenomena usually can be recognized by the "clean" nature of the
break. Less than 10% of bones demonstrating green breaks also manifested
tooth markings indicative of large carnivore action. These data strongly
suggest that the Coleman IIA deposit was a natural-trap type of sinkhole.

That the sink was quite extensive and not filled to the top with
water during most of its depositional history is shown by the relatively
large quantity of bat remains. Those of Myotis austroriparius are
particularly common, and as this bat dwells now in limestone caves and
sinks in north Florida, we may assume that it also inhabited the more
inaccessible reaches of the Coleman IIA sinkhole. Owl remains have
also been recovered from the sink, and it is conceivable that they also
utilized the cavernous portions for roosting sites, and inadvertently
stockpiled small mammal remains. One may find fossilized owl pellets
at the Reddick IA site, which at least suggests that the phenomenon
is not particularly unusual.

Didelphis marsupialis

Cryptotis parva
Blarlna brevicauda
Scalopus aquaticus

Pipistrellus subflavus
Myotis cf austroriparius
Plecotus rafinesquii
Bat sp.

+Dasypus bellus
+Holmesina septentrlonalis

Lepus alleni
Sylvilagus sp.

Sciurus carolinensis

Glaucomys sp.

Geomys cf G. pinetis

Reithrodontomys humulis

Peromyscus floridanus

Peromyscus sp.

Ochrotomys nuttalli
+Sigmodon bakeri new sp.
+Pitymys arata new sp.

Neofiber alleni

Erethizon dorsatum
+Hydrochoerus sp.

Canis lupus
+Urocyon minicephalus new sp.
+Arctodus pristinus

Procyon lotor

Spiloqale putorius

Mephitis mephitis

Conepatus sp.
+Felis onca augusta

Felis rufus

+Mammuthus sp.

+Equus sp.

+Platygonus cumber landens is
+Mylohyus sp.

+Tanupolama cf I. mirifica
Odocoileus virginianus

SPECIES LIST

Minimum

%

Number

Total

Individuals

2.0

5

2.4

6

0.4

1

0.8

2

3.6

9

10.4

26

0.8

2

11.2

28

0.4

1

0.4

1

2.4

6

3.6

9

1.6

4

0.4

1

4.8

12

2.0

5

7.6

19

11.2

28

1.2

3

4.4

11

0.8

2

0.4

1

0.4

1

?

?

1.6

4

6.8

17

0.4

1

0.4

1

0.8

2

0.4

1

0.8

2

2.8

7

0.4

1

0.4

1

0.8

2

4.4

11

0.4

1

2.8

7

3.2

8

249

SPECIES ACCOUNT

Class Mammalia

Order Marsupialia

Family Didelphidae

Didelphis marsupialis Allen- opposum
Material: UF 11826-11829; 1 left dentary, 1 left maxilla, 5 right
humeri, 2 left femora, 4 vertebrae.

Remarks: The fossil material is identical to that of Recent D. marsupi-
alis.

Order Insectivora

Family Soricidae

Cryptotis parva (Say)- least shrew
Material: UF 11628-11641; 9 mandibles, 6 maxillae, 1 humerus.
Remarks: Table 1 shows that size differences separate mandibles of C.
parva from those of Blarina brevicauda. There is little overlap in the
measurements even though the series of Blarina measured consisted al-
most exclusively of B. brevicauda minima, the smallest subspecies of
Blarina. In those fossil specimens in which the M3 is present, the
talonid is reduced as in Recent Cryptotis (Hibbard, 1963; Guilday,
1962).

Blarina brevicauda (Say)- shorttail shrew
Material: UF 11626; 1 right mandible.

Remarks: The Mo is not preserved on the fossil, but the size of the
mandible (Table 1) and qualitative characters (after Guilday, 1962)
identify it as B. brevicauda.

■ 10-

en

cfl

0

>

u

q-i

^

(1)

01

a

TJ

u<

u

G

O Pi

u

S O *d

E C

CD cB
C --'

•H ,0 cfl

^•■-1 73

XI D

CO C Cfl

■U CO o

c e -h

0) >

Q T3 0)

(1) C M

3 _,

w C ctj

S J-1 U

B

II

IX

IX

IX

IX

T3 M-l

i- o

r^

^D

<t

vO

CN

ro

CN

in

<r

cn

^-1

3 to

CN

m

CN

* r-.

^D

<f

CN

-I CM

■H

CN

vO

0 vD

n

<)-

CN

J 1/1

LO

r^

O

■11-

Family Talpidae

Scalopus aquaticus Linnaeus- eastern mole
Material: UF 11642-11645; 1 right dentary, 2 femora, 3 humeri.
Remarks: The small sample precludes statistical treatment, but the humeri
and toothless dentary agree with the same elements of Recent Scalopus
aquaticus from Florida.

Order Chiroptera

Family Vespertilionidae

Myotis cf M. austroriparius fRhnHpg^- south-
eastern myotis
Material: UF 13244, 13245, 13251, 13252, 13256-13292: partial skull,
partial palates and mandibles.

Remarks: Considering the western affinities of a number of the Cole-
man HA mammals, the possibility that this is a small western species
of Myotis deserves careful consideration. Yet I cannot discern any qual-
itative or quantitative differences between the Coleman Myotis and Re-
cent M. austroriparius from Florida, while size differences tend to
separate it from M. keenii and M. grisescens (Table 2). On the other
hand, I cannot separate the Coleman II Myotis from M. lucifugus.

Pipistrellus subflavus (F. Cuvier)- eastern

pipistrelle
Material: UF 13246-13250, 13253-13255; dentaries and partial palates.
Remarks: The genus Pipistrellus has one less premolar than either
Plecotus or Myotis , and may be separated from other living Florida bats
on the basis of size and dental configuration. The fossil material is
identical to that of Recent P. subflavus from Florida.

■12-

Table 2

Distance (in mm) from anterior border of canine
alveolus to posterior border of M3 alveolus
in select fossils and Recent specimens of Myotis

M. austroriparius

M. grisescens

Indiana, Pennsylvania

M. keenii 10 6.2 6.0-6.6

(Miller and Allen, 1928)

Coleman HA

Devil

s Den

NX O.R.

N

X

O.R.

11 5.8 5.5-5.9

6

5.5

5.3-5.9

7

6.3

6.2-6.6

Plecotus rafinesquii Lesson- Rafinesque's

big-eared bat
Material: UF 13240-13243; a partial palate and mandibles.
Remarks: The use of Plecotus rather than Corynorhinus follows Dalquest
(1953) and Handley (1959) . The following features may distinguish
dentitions of Plecotus from those of Myotis :

1) Paraconid and metaconid less separated in Plecotus

2) Protoconid higher and more blade-like in Plecotus

3) Only two upper premolars in Plecotus

4) Hypocone on M^-M absent in P. rafinesquii (well developed in

all Myotis I have seen from Florida).
The third character is not applicable to the fossil material
available, but the others clearly permit the identification of P.
rafinesquii.

-13-

Order Edentata

Family Dasypodidae
+Dasypus bellus (Simpson)- beautiful armadillo
Material: UF 13190, 13191, 13187; dermal plates and an astragulus
Remarks: The fossil Dasypus material includes dermal plates of two size
classes (Figure 2). They may be from two individuals, but most likely
represent samples from two body areas of a single individual. As can be
seen from Figure 2 there is very little overlap in size of the dermal
plates between the extinct Dasypus bellus and the living D. novemcinctus .
The area of overlap includes, in the fossil armadillos, only the small
animal from Coleman IIA and a few plates of animals from other deposits.
Measurements of the fossil and Recent armadillo dermal plates (movable
plates only) are presented below (N = number of plates measured, X =
mean, 0. R. = observed range of measurements):

N

Thickness
X O.R.

D. novemcinctus

+D. bellus

Blanc an
(Haile XVA)

Irvingtonian
(Coleman IIA)

Rancholabrean
(Haile VIIA, Haile XIB,
Reddick IA, Reddick IB,
Bradenton 51st St.,
Arredondo II, Medford I,
Kendrick I)

235 1.6 0.9-2.4

10 3.5 2.9-4.2

21 4.0 2.7-6.1

163 3.9 1.9-6.1

Width

N X O.R.
235 5.4 3.5-8.1

10 10.5 8.2-12.4

21 8.2 6.2-10.3

163 11.4 6.6-16.5

Figure 2 -- Scatter diagram relating Dasypus bellus to
Dasypus novemcinctus when depth of the mov-
able dermal plates is plotted against width
of these plates. Large circules denote the
range of measurements in D . novemc i nc tus ,
small circules and circles with lines denote
the range of measurements of D. bellus.

Circles with lines = dermal plates from

1
Coleman HA, X = grand mean of 235 meas-

2
ured plates of D. novemcintus, X = grand

mean of 194 measured plates of D. bellus.

-15-

•

•

15.1—

#

>

•

•

•

•

•

• •

•

5

X2

1

57—

• +

:: ++ +
V*

• x

•

• •

• •

i i | " " i i i i
/./ 2.1 3.1 4.1 5.1 6.1

DEPTH

■16-

+Ho lines ina septentrionalis (Leidy)- armadillo
Material: UF 13186, 18188, 18189; dermal plates and metapodials.
Remarks: Considering that there are two names in the recent literature
for this animal, and that there are two species (one undescribed; Haile
XVA) of possibly two genera of chlamytheres in Pleistocene deposits from
Florida, a general review of this subject seems worthwhile now. Much of
this discussion is taken from Simpson (1930b).

The first chlamythere record in North America was published by
Leidy (1889 a); the material consisted of dermal plates found in Peace
Creek, Florida, and was referred to Glyptodon, new species septentrionalis.
"Later Leidy (1889 b) recognized that they did not belong to a glyptodont
but to a gigantic armadillo and referred them to Chlamytherium humboldtii
Lund, a species described from cave deposits in Brazil (Simpson, 1930 b)."
Sellards (1915) demonstrated that the Florida chlamythere was a differ-
ent species than C. humboldtii, and referred the Peace Creek material,
and material from Vero, Florida, to Chlamytherium septentrionalis.
Simpson (pers . commun.) has pointed out to me that Sellards* use of
septentrionalis was incorrect. As the species name must agree in gender
with the genus to which it is referred, the proper name for referral to
Chlamytherium is septentrionale. Following the taxonomic system of
Castellanos (1927), with new Florida chlamythere fossils from Braden-
ton Field, Simpson (1930 b) referred the Florida chlamythere to a new
genus, Holmes ina, species septentrionalis (agreeing in general with
Holmes ina, which is feminine).

Some workers since that time (Bader, 1957; Weigel, 1962) have fol-
lowed Simpson's conclusions, but others (James, 1957; Ray, 1958; Gut

•17-

and Ray, 1963; Hibbard and Dalquest, 1966) have reinstated Chlamytherium.

There are presently six named genera of Plio-Pleistocene chlamytheres
Chlamytherium (=Pampatherium) , Plaina, Hoffstetteria, Kraglievichia,
Vassalia, and Holmesina. The first five were originally described from
South American deposits (reviewed by Castellanos, 1927, 1937, 1957), the
last solely from the United States (Simpson, 1930b). Kraglievichia,
Plaina, and Vassalia are Pliocene forms (from Entrerian, Monte Hermosan,
and Araucanian time), and South American Chlamytherium and Hoffstetteria
are restricted by Castellanos to the Pleistocene (Pampean of Argentina
and Brazilian and Ecuadorian deposits presumably of Pleistocene age).
As Simpson (1930b) states, "If . . . [Chlamytherium, Kraglievichia, and
Vassalia] ... be retained in a single genus, then the Florida form
belongs in that genus. If Castellanos is followed, which is probably
preferable . . . , then the Florida species cannot be referred to any
of his three South American genera, for it differs from them as much
as they differ among themselves."

Considering that there is no recent taxonomic revision of these
forms available, Simpson's statements seem to outline the most logical
present course of action regarding the taxonomic status of the Florida
late Pleistocene (Rancholabrean) chlamythere, recorded by Ray (1958)
from Melbourne, Weigel (1962) from Vero, Bader (1957) from Hornsby
Springs, Gut and Ray (1963) from Reddick IA, and by Simpson (1928,
1929, 1930b) from Sabertooth Cave, Bradenton Field, Peace Creek, With-
lacoochee River, Venice, Sarasota, and Seminole Field. The same can-
not necessarily be said of the chlamytheres noted by Hibbard and Dal-
quest (1966) from the middle Pleistocene of Texas, and by Cahn (1922)

• 18-

and Hay (1926) from deposits of uncertain age in Texas, as there is
evidence that anthoe genus of chl-amythere was also present in North
America during part of the Pleistocene.

Castellanos (1927, 1957) and Simpson (1930 b) assumed that a maxi-
mum of three genera of chlamytheres existed during the entire Pleistocene
in both North and South America. Further, only one species of each genus
was recognized during that time (South America, Chlamytherium humboldtii
and Hoffstetteria occidentale; North America, Holmesina septentrionalis) .
Recent collecting in Florida directed by S. David Webb has produced a
new, small chlamythere of early Pleistocene (late Blancan) age. This
species is now known from the Haile XIIA, Haile XVA (Robertson, Ph. D.
disser., U. Fla.), and the Santa Fe River 1 (intrusive) and IB faunas.
The generic status of the Blancan form is not now known; in size and
some features of the dentition the beast probably best approximates
Kraglievichia.

In the Coleman IIA chlamythere the dermal plates and metapodials
are comparable in size to those of Holmesina septentrionalis. This is
the earliest record of this genus in Florida. The following is a list
of localities in Florida, other than those noted previously, that con-
tain remains of Holmesina septentrionalis:

Alachua County:- Arredondo, Pit I (site unknown), Pit II (just S of B);
Haile, Loc. VIIA, VIIIA; and Payne's Prairie, Loc . III.
Calhoun County:- Chipola River, Loc. 2.
Charlotte County:- Charlotte Harbor.
Citrus County:- Withlacoochee River, Loc. 7A.
De Soto County:- Joshua Creek; and Prairie Creek.

•19-

Gilchrist County:- Ichetucknee River; and Santa Fe River, Locs . 2, 3,

7B, 11C, 17.
Levy County : - Waccasassa River.

Marion County:- Medford, Cave 1; Orange Springs, Loc. 1; and Reddick IIC,
Orange County:- Rock Springs.
Palm Beach County:- Jupiter Inlet.
Polk County : - Pool Branch.
Sarasota County:- Apollo Beach.
Sumter County : - Coleman III, Locs. B, C.
Suwanee County:- Branford IA.

Order Lagomorpha

Family Leporidae

Lepus alleni Mearns- antelope jackrabbit
Material: UF 13178, 13179; femora and innominates .

Remarks: The Coleman material referrable to the antelope jackrabbit
consists of ten femora and two innominates. All lagomorph dental com-
ponents in the fauna were assignable to Sylvilagus. The femora are
rather large, and most closely approximate the antelope jackrabbit,
Lepus alleni (Table 3; Figure 3). There is some overlap in measure-
ments between L. alleni and L. townsendii, but configuration of the
lesser trochanter and accompanying muscle scars (of the obturators and
quadratus femoralis especially) readily separate these species (Figure
10). Lepus alleni is also a member of the Inglis IA Irvingtonian de-
posit, and its occurrence in these deposits marks the first record of
this genus from Florida.

Table 3
Measurements (in mm) of fossil and living rabbit femora

•20-

Gr. w shaft

Gr. length

Gr. w distal

Gr. w head

Gr. w shaft
Gr. length
Gr. w distal

Gr. w shaft
Gr. length
Gr. w distal
Gr. w head

Gr. w shaft
Gr. length
Gr. w distal
Gr. w head

NX 0. R.

Lepus californicus

9 8.6 7.2-9.5

9 108.2 100.5-114.8

9 16.1 15.2-16.8

9 8.2 7.6-8.9

Lepus towns end ii

10 10.1 9.3-10.8

10 123.1 117.7-129.0

10 19.0 18.3-20.5
Lepus alleni

6 9.3 8.4-9.8

6 118.9 114.4-122.6

6 18.6 18.3-19.3

6 9.6 8.9-10.1

Lepus americanus

5 6.6 6.0-7.5

5 89.9 84.7-96.7

5 12.7 11.7-13.9

5 6.5 5.9-6.8

-21-

Table 3 (continued)

Gr. w. shaft
Gr. length
Gr. w distal
Gr. w head

Gr. w shaft
Gr. length
Gr. w distal
Gr. w head

NX 0. R.

Coleman II Lepus alleni
7 9.0 8.0-10.3

1 118.3

2 17.8 17.6-18.0

3 9.5 9.2-9.7
Coleman II Sylvilagus

3 6.2 6.0-6.5
1 71.4

3 11.2 11.0-11.6
1 5.5

Sylvilagus sp.- marsh and /or cottontail

rabbit
Material: UF 13158-13161, 13165, 13166, 13174-13177: 3 femora, 6
humeri, 2 scapulae, 4 calcanea, 11 mandibles, 2 tibia, 1 innominate,
5 ulnae, isolated teeth.

Remarks: The material is too poor to be identifiable to species, as
characteristic portions of the mandibles have been shorn away. Both
Sylvilagus floridanus and S. palustris are presently sympatric in Florida,
found in almost every situation providing dense underbrush. Some meas-
urements of the Coleman IIA Sylvilagus are presented in Table 3.

Figure 3 -- Femora of Lepus alleni (UF 13178; right) and
Sylvilagus sp. (UF 13176; left) from Coleman
HA.

-23-

• 24-

Order Rodentia

Family Sciuridae

Sciurus carolinensis Gmelin- Gray squirrel
Material: UF 11650-11658; 3 innominates, 2 right mandibles, 1 maxilla,
5 femora, 2 humeri, 2 tibia, 2 ulna, 2 incisors.

Remarks: Table 4 shows that in almost all measurements the fossils con-
form to Recent Sciurus carolinensis. Sciurus niger is consistently
larger. The MA length of S. niger shown in Table 4 agrees with the
dentary tooth row measurements presented by Moore (1956) for S. niger
shermani from central Florida.

Glaucomys sp.- flying squirrel
Material: UF 11646-11649; 2 dentaries, 1 femur.

Remarks: Table 5 shows that the Glaucomys fossils are not identifiable
to species when either the MA or RA lengths are considered separately.
This is true also when the MA length is plotted against the RA length
(Figure 4). The one fossil from which both measurements were obtain-
able lies intermediate between the two species.

Guilday, et al. (1964) identified both G. volans and G. sabrinus
from the New Paris No. 4 Sinkhole in Pennsylvania. Although the fossil
G. volans averaged 8% larger than modern samples from Pennsylvania,
overlapping with modern G. sabrinus macrotis in MA length, they felt
that G. volans was larger in the Pleistocene. This interpretation gains
support from the fact that in the same fauna G. sabrinus averaged 13%
larger than Recent Appalachian races of that species.

The Coleman IIA Glaucomys are 8% larger than the average Recent

-25-

G. volans from the southeastern U. S. However, because the MA length
is so variable in G. sabrinus (V = 8.97) and because the theoretical
range of measurements for the MA length of G. sabrinus at -3s includes
the entire theoretical range at 13s for G. volans, the Coleman IIA

Glaucomys must remain unidentified at present.

-26-

»

m

e

C

0

0

c

•H

«H

IT)

4J

4J

01

m

m

E

U

•H

>

<U

a

cu

J3

>

<o

CO

4-1

4-1

LH

r)

c

en

M-l

O

u

CU

•r-l

o

ttj

R

m

4J

TJ

hi

C

CO

£

c

i-H

0.)

4->

"J

crt

0)

C

•i-l

■H

4-1

•H

a

CJ

en

.-1

^1

•■-i

•M

OS

o

i-i

M-4

m

>H

u

4-1

d

crt

CU

CI)

en

RJ

o

o

0

3

U

fl

C)

C

a

cu

•H

-u

in

T>

II

b

DQ

•r^

n

T1

4-i

>

>-4

o>

C

C

0

g

m

crt

CO

cm:

•r-l

CD

c

^D

rn

r-l

3

T)

U

c

3

en

•r-l

y

0

CO

a

•H X>

^ CJ
ctf
co

■U JJ

B a

CD CD

S o

cu cu

u <&

3 ,

CO M-t

crt O
&

II > *

cu

X X> CO

4-J

B crt
O T3

m co cu

o> (U

it a

CU • •>

a cu cu

t-I 60 60

cu cb c8

CO

T3 i-H

u-i cu crt
O > o

U CU 4J
CU CO cu
,0 ,0 H

Boo

3 cu

C II -C

4-1
II •

O oo

+ 1

in CI .
o> 0|X
U

+ icn -fin + 1 m -ficN

0J

0)

60

4J

M

4J

4-1

C

XI

C

nj

CO

0)

■r-l

CU

c

C

r-4

3

r-l

r-l

■r-l

H

S

h

M

c«

o

o

3

3

•r4

c

c

s

s

J3

c

s

QJ

CU

•r-l

■27-

>

+ '

"-> a \x

<y\ o
O

+ l ro

ON ON

O IN

4J

x:

T3

■u

•H

M

9

(3

a)

01

^H

3

M

OJ

<

H

3

ffi

2

^^

cfl

.-H

H

CS

■H

ja

4J

X

jj

Si

cn

0

M

4-1

■H

S-4

C

13

T3

a

a)

•H

•^

^H

3

J3

X

4J

,G

4-)

J3

OJ

a)

M

4J

M

4-1

4-1

4-1

C

T3

C

X)

cd

tfl

0)

■H

0)

■rl

a

C

^H

3

1—1

3

•.-1

«H

id

B

H

l-l

cfl

OJ

o

o

3

3

■i-(

■H

c

c

£

S

X

Xi

c

c

01

QJ

•H

•H

•28-

>

in C |X

a <u

3 S T)

+ 1 on

+ 1 NO

3 X!

i /->

U 4J

X rH

0) T3

O cfl

1

i

i

1

a -h

n a

3 3

0.-M

SC

CO
3 X!

u u

a c

3 <U
5S i-i

i

i

1

co x

X r-l

•i-l 4J

O CO

XI T3

n H

I

i

i

1

•H T^

a-H

H &

XI
CO 4J

•r-l 50
XI (3
•H CU

H r-i
r-l

i

'

'

rJ

Xi CO

3

4-1 4J

00

00

1

X) CD

I

i

0)

•rJ T-l

ON

ON

Pn

x:

M 4J

3 cxt

O

o

a c

I

i

Q) CU

o

r-l

Pn rH

NO

nD

1

<U Xi

ON

r^-

4J 4J

1

3

CO T3

«*

LO

3

3 -H

i-l

I— 1

H

1

•H &

X

g

CD 4-1

NO

ON

0

4J M

■

1

a

cfl 3

ON

o

a

3 CU

<r

m

m m

ON o

vj- <t in m
m m m m

NO vO NO NO

co co co
m m in

VO NO NO

^^ggg

m

m

CD

ID

CD

CD

CD

H

•r-l

•H

•H

■r-l

1-1

•r-l

cd

CD

CD

rn

CD

CD

CO

C

3

3

3

C

a

3

(1)

0)

(1)

CU

CU

CU

a)

3

3

C

3

3

3

3

r-l

•r-l

•r-l

•H

•r^

•H

•■-i

i— 1

rH

r-l

r-l

H

H

r-l

o

o

o

O

0

0

o

r-i

u

U

(-1

u

M

u

«

cfl

rt

n)

cfl

cfl

cfl

a

O

y

u

u

U

CJ

.29-

• XI

cu

TJ ID

H

•H -i-l

3

£ X)

ffi

s-^

en

3

,.c

1 /-n

M

4-1

X rH

0)

TJ

O CO

H

•H

u 0

n

IS

Ci-H

a

X

4J

M

C
<D

03

X

X .-i

•H

4-1

O cfl

/J

X)

n a

■H

•H

a -,-1

H ?

Jj X cfl

3 u u

a t) co

CO t4 t^

Pn 3 TJ

u

4J

3

eXI

b

c

aj

a)

Ph

i—i

1

0

0)

X

0

4J

4J

c

cfl

TJ

c

C

•.-I

M
1

•r-4

13

X

H

OJ

j-i

<)

4-1

U)

a

CO

c

a

C

0)

till

TJ

00

^o

vO

o

i—i

4J

in

m

in

in

m

c

vd

vO

v£3

^o

vO

g^gp

CO

m

0)

en

co

•H

•H

T-l

•■-I

•rH

CO

en

en

rn

CO

a

C

C

a

G

cu

a)

ci)

cu

CI)

c

C

C

c

(3

•H

•.-1

■r-J

■H

•i-t

r-l

i—i

l—l

■—I

i—i

o

n

O

O

o

u

M

u

J-I

J-l

cfl

crt

ca

cfl

cfl

O

a

o

O

o

i

Figure 4 — Scatter diagram relating Glaucorays sabrinus
(large circles) to G. volans (small circles)
when RA length is plotted against MA length.
Star = fossil Glaucomys from Coleman IIA.

-31-

$£-

1

9.0-

8.5-

•
•

8.0-

•

5

s

-i 7.5-

•

^

• ••

•

^

•
• ® —
• •

7.0-

••
• t
• • •
• « 1 t •
• t •• •

6.5-

• : t • •

• i» i • •

• • • • •
t

• •

6.0-

1

• •
•
• • •

10.0 11.0 12.0 13.0

r

14.0

HA. LENGTH

-32-

•rl fl fi

■n

c

G

cti

cfi

LO

ti

3

(!)

m

C

■—I

o

•H

£>

u

cfl

X>

H

cfl

I

B O

a m-i

m d |x

o m m on

<y\ <n in r--.

I

I

O CO

r* o-

vO OS

CO O OS CM

os cm m r»»

+ • <r

+ 1 oo
m o

s s

■33-

Family Geomyidae

Geomys cf G. pinetis Rafinesque- southeast-
ern pocket gopher
Material: UF 11659-11683; 1 partial cranium, 10 right mandibles, 8 left
mandibles, 5 left femora, 2 left humeri, 3 right humeri, 4 right innomin-
ates, 4 left innominates , 2 right tibia-fibulae, 12 left upper incisors,
11 right upper incisors, isolated teeth.

Remarks: The fossil specimens are identical to Recent G. pinetis from
Florida. However, they are not certainly unseparable from G. bursarius
and G. personatus either. These three species are obviously closely
related and probably represent remnants of a common Pleistocene Gulf
coast population.

Family Cricetidae

Reithrodontomys humulis (Audubon and Bachman)-
eastern harvest mouse
Material: UF 11686-11695; 8 mandibles, isolated teeth.
Remarks: Reithrodontomys humulis is easily identifiable by the pres-
ence of a distinct labial cingulum, usually with associated cusplets,
on the ML and M2 (Hall and Kelson, 1959; Hooper, 1952). The grooved
upper incisors are also diagnostic.

Peromyscus floridanus (Chapman) - gopher or
Florida mouse
Material: UF 11731, 11733-11736, 11738-11743, 11746-11747, 11749-11761,
11763-11768; 31 mandibles.
Remarks: Identification of P. floridanus follows Martin (1967). This

•34-

mouse is one of the largest of North American Peromyscus. It is identi-
fiable from the MA length alone. The dental pattern of the fossils agrees
with the description given by Bader (1959), in that the ectolophid and
ectostylid are rare or absent, while on the M-. and M2 the mesostylid and
mesolophid are poorly developed or absent.

Ochrotomys nuttalli (Harlan)- golden mouse
Material: UF 11730; 3 humeri (2 left, 1 right), 1 right mandible, 1
right maxilla, 2 isolated molars.

Remarks: 0. nuttalli mandibles without teeth can be separated from other
Recent Florida Peromyscus species when RA length is plotted against MA
length (Martin, 1967). The position of the mental foramen will further
separate 0. nuttalli from some additional Peromyscus species (Martin 1968b) ,
but identification of toothless mandibles is not possible if either P.
maniculatus or P. leucopus is presumed present in the fossil fauna under
study. Molars of the golden mouse are easily identified. The molar
pattern is one of the most complex of the Peromyscines (Osgood, 1909;
Hooper, 1957; Bader, 1959; Martin, 1968b).

Humeri like those of Peromyscus , lacking an entepicondylar fora-
men, may also be referred to 0. nuttalli. The entepicondylar foramen
is absent in 0. nuttalli and is present in specimens of Peromyscus
polionotus , P. gossypinus , and P. leucopus.

Peromyscus sp.
Material: UF 11720-11729, 11732, 11737, 11744, 11745, 11748, 11762,
11769-11784; 17 left mandibles, 7 right mandibles, 5 right humeri, 8
left humeri, 5 right femora, 10 left femora, 3 right innominates, 3

-35-

right tibia-fibulae, 3 left tibia- fibulae, numerous isolated teeth.
Remarks: This material probably represents a mixture of three or four
Peromyscus and Ochrotomys species. It undoubtedly contains some small
P. floridanus, a few specimens of 0. nuttalli (toothless mandibles),
and either one or two of the species P. gossypinus, P. leucopus, or P^
maniculatus. The incidence of accessory cusps increases on the molars
of these smaller mandibles, as they do in P. gossypinus, P. leucopus, and
P. maniculatus. Three mandibles (UF 11720, 11723, 11724) have an MA
length which falls in the range of P. polionotus (fossils: 3. 6-3. 7mm;
Recent P. polionotus: 3. 1-3. 7mm), but this length also falls within
the measured range of all southeastern Peromyscus species except P^
floridanus (Martin, 1967). Since no mandibles or teeth were identi-
fied as P. polionotus it seems likely that the oldfield mouse is ab-
sent from the Coleman IIA fauna.

Western Peromyscus species, such as P. pectoralis, P. boylii, and
P. eremicus now found in Texas are definitely not represented in Cole-
man IIA (see Martin, 1968b).

+Sigmodon bakeri, new species

= Sigmodon hispidus , Holman, 1959;

Bull, Florida State Museum, 5(1): 1-24
Holotype: UF 11700, a left mandible with all teeth

Horizon and Locality: late Irvingtonian; Coleman IIA, Sumter County,
Florida.

Referred specimens:

-36-

Coleman IIA, Sumter County, Florida

UF 11696-11719; 11 left mandibles, 8 right mand-
ibles, 2 left maxillae, 3 left femora, 4 right
femora, isolated teeth.
Bradenton Field, Manatee County, Florida; early Sangamon:

UF 2002; left mandible.
Williston III, Levy County, Florida; early Sangamon:
FGS 5862; 2 left mandibles, 1 right mandible,
1 lower right incisor.
Haile VIIA, Alachua County, Florida; early Sangamon:
UF 9844; right mandible

UF 15153; 2 right mandibles, 1 left mandible,
3 isolated molars.
Diagnosis: The first lower molar of Sigmodon bakeri possesses 4 well
developed roots, indicating alliance with the hispidus species group as
defined on page 42 of this treatment. Sigmodon bakeri differs from all
of these species in characteristically lacking an anterior cingulum on
the M2 and M3 (Figure 5). The configuration of the anteroconid of the
Mi will also separate S . bakeri from other hispidus group species. In
S . bakeri the anteroconid is grossly asymmetrical and is extended lab-
ially and posteriorly. Other hispidus group species will evidence this
character occasionally, but normally the anteroconid is symmetrical or
only slightly asymmetrical. Sigmodon bakeri is smaller than any living
Sigmodon species except S. ochrognathus (Table 6) .

Etymology: This species is named after Dr. Rollin Baker of Michigan
State University for his work on living cotton rats.

i

Figure 5 — Lower teeth of some fossil and Recent cricetid

rodents. A = Sigmodon bakeri, holotype UF 11700,
Coleman IIA, B = Sigmodon hispidus, UF 549, Re-
cent specimen from Brevard Co., Fla., C = Pitymys
pinetorum, UF 14387, Reddick IA: small numbers
correspond to triangles, D = Pitymys pinetorum
nemoralis, UK 54262, Recent specimen from Green-
wood Co., Kansas, E = Pitymys ochrogaster haydeni,
UK 20772, Recent specimen from Scott's Bluff,
Nebraska, F = P. 6. haydeni, UK 20760, Scott's
Bluff, Neb., G = Pitymys pinetorum, UF 4860,
Haile XIB. The arrows point out the salient fea-
ture that separates Pitymys pinetorum from Pj^
ochrogaster (see discussion under Pitymys arata) .

■38-

left

5S
^

■39-

Discussion: Discovery of other fossil Sigmodon species in Florida, from
the Blanc an Haile XV deposit in Alachua County and the Inglis IA site
in Levy County, prompted a study of other North American fossil samples
of this genus. An annotated list of the living and extinct species and
extinct subspecies is presented below. Some of these taxa are new, and
are described in the species account which follows the list. A cross in
front of a taxon indicates that it is extinct.
hispidus species group
hispidus
alleni

ochrognathus
fulviventer
+bakeri new species
leucotis species group
leucotis
perunanus
+curtisi
+hudspethensis
alstoni species group
als ton!
+ medius species group
+medius

4-medius medius new subspecies
-hnedius hibbardi new subspecies
-hninor

-40-

Specimens referred to new taxa, except for those of S. bakeri, as well
as those of all other fossil Sigmodon taxa studied, are listed under
"Fossil Material Examined" at the end of the section of this report
dealing with Sigmodon.

Key to the Extinct and Extant Species of Sigmodon

l.a. First lower molar with four well developed roots 2

b. First lower molar with three well developed roots 7

2. a. Relatively low crowned molars; reentrant folds broad and shal-
low; lingual root reduced +S. hudspethensis

b. Hypsodont molars; reentrant folds characteristically narrow

and deep; lingual root well developed 3

3. a. Mental foramen located at anterior base of M]^ and lingually

directed (not visible from labial view) S. alleni

b. Mental foramen anterior to the base of M-^ and partly labially

directed (visible from labial view) 4

4. a. Anterior cingulum on M2-M3 usually well developed; anteroconid
of M;l anteriorly -posteriorly flattened in specimens with

little tooth wear 5

b. Anterior cingulum on M2-M3 moderately developed to absent;
anteroconid of M-. not anteriorly-posteriorly flattened in

specimens with little tooth wear 6

5. a. Size small; lower third molar relatively short S . ochrognathus

b. Size large; lower third molar relatively long S. fulviventer

-41-

6. a. Anterior cingulum on M2-M3 grading from moderately developed

to absent, characteristically present; Mi usually without

posteriorly extended anteroconid; size large S.hispidus

b. Anterior cingulum on M2-M3 characteristically absent; M-^

with posteriorly extended anteroconid; size small +S. bakeri

7 . a. Upper incisors grooved S. alstoni

b. Upper incisors, when known, not grooved 8

8. a. Anterior cingulum on Mo-Mo absent S. peruanus

b. Anterior cingulum on M2-M3 usually well developed 9

9. a. Size and dental pattern approximating that of

S . hispidus +S. curtisi and S. leucotis*

b. Size small; dental pattern not as in S. hispidus 10

10. a. Size relatively large; reentrant folds relatively

broad +S . medius

b. Size relatively small; reentrant folds relatively

narrow +S . minor

* I cannot separate the dentitions of these two species,
though perhaps the anterior cingulum on M2-M3 is a bit
less developed in S. leucotis. As noted in the Species
Account, the interparietal bone is absent or at most
vestigial in S. leucotis, but as no skull material of
S. curtisi has been discovered I retain the integrity
of this species.

-42-

Account of the Extinct and Extant Species of Sigmodon

Genus Sigmodon Say and Ord 1825
Characteristics of the genus are adequately delimited by Hershkovitz
(1955). Taxonomy of the Recent North American species follows Baker
(pers. commun.), although the species groups are in part my develop-
ment. Much work needs to be done with living South American Sigmodon,
and for now I have merely lumped all beasts in that continent with three
roots on the M1 into either Sigmodon peruanus or S. alstoni, depending
upon the condition evidenced by the upper incisors.

hispidus species group
Diagnosis: This group is characterized by possessing four well de-
veloped roots on the first lower molar. All species except S. bakeri
are extant and distributed almost entirely in North America.
Geologic Range: Latest Irvingtonian to Recent.

Sigmodon hispidus Say and Ord 1825
Diagnosis: The largest North American cotton rat (Table 6), Sigmodon
hispidus is the most advanced cotton rat with respect to dental evolu-
tion. The process of lamination as described by Hershkovitz (1962)
has proceeded in this species to the point of separation of the ante-
roconid from the paraconid and entoconid in some specimens. The cheek
teeth are highly prismatic and the reentrant folds are deep and nar-
row. The enamel surface at the termination of each reentrant fold is
thinned relative to the enamel occlusal surface along the remainder of
the border of each fold. The anterior cingulum on the M2 and M3 tends
to be reduced.

.43-

Geologic Range: Rancholabrean to Recent
Florida:

Devil's Den: Martin, 1968a; Wisconsin

Haile VIIIA: unpubl.; Sangamon

Haile, Locs . XIB, XIIIA, C: unpubl.; ?Sangamon

Withlacoochee River 7A: unpubl.; Wisconsin

Ichetucknee River; unpubl.; Wisconsin

Vero: Weigel, 1962; Wisconsin

Seminole Field: Simpson, 1929; Wisconsin

Melbourne: Ray, 1958; Wisconsin

Reddick IIC: unpubl.; Sangamon

Reddick IA: Gut and Ray, 1963; Sangamon

Sabertooth Cave: Simpson, 1928; Sangamon

Kendrick I: unpubl.; Sangamon

Maximo Moorings: unpubl.; Wisconsin

Arredondo IA: unpubl.; Wisconsin

Arredondo IIB: unpubl.; Sangamon

Arredondo IIC: unpubl.; Sangamon

Louisiana:

Georgia:

Texas

New Mexico:

Aruba:

Little Bayou Sara: Martin, 1968a; Wisconsin

Ladds : Ray, 1967; Sangamon?

Moore Pit: Slaughter, 1966; Sangamon?

Sims Bayou: Slaughter and McClure, 1965; Wisconsin?

Howard Ranch: Dalquest, 1965; Wisconsin

Clear Creek: Slaughter and Ritchei, 1963; Wisconsin

Friesenhahn Cave: Lundelius, 1960; Wisconsin

Longhorn Cavern: Semken, 1961; Wisconsin

Ben Franklin: Slaughter and Hoover, 1963; Wisconsin

Brown Sand Wedge: Slaughter, 1962; Wisconsin

Isla: Hooijer, 1967; Wisconsin?

-44-

Sigmodon alleni Bailey 1902
Diagnosis: Sigmodon alleni is best characterized by position of the
mental foramen of the mandible. This foramen cannot be seen when the
mandible is viewed from the labial side, as it is located close to the
base of the first lower molar and more lingual ly directed than in other
Sigmodon species. The dental pattern is most similar to that of Sigmodon
hispidus .
Geologic Range: No fossil record.

Sigmodon ochrognathus Bailey 1902
Diagnosis: Sigmodon ochrognathus is the smallest of the living Sigmodon
species (Table 6), and like S. fulviventer, demonstrates a peculiar
anteriorly-posteriorly compressed anteroconid on the M-^ in specimens
with little tooth wear.
Geologic Range: No fossil record.

Sigmodon fulviventer Allen 1889
Diagnosis: This diagnosis applies to S . f . minimus , the only sub-
species I have viewed. In size S. fulviventer approaches S. hispidus
(Table 6), and with the exception only of the anteriorly-posteriorly
compressed anteroconid of the M-^, the dental pattern approximates that
of S. hispidus.
Geologic Range: No fossil record.

+ Sigmodon bakeri Martin 1969
Diagnosis: As discussed previously. Sigmodon bakeri represents the
earliest record in North America of the hispidus species group, but the
specialized condition of absence of the anterior cingulum on the M2-M3

-45-

disallows it as an ancestor for any other hispidus group species. The

dental pattern, with the exception of the character mentioned above, is

somewhat intermediate between the primitive 3 -rooted Mi species and the

hispidus species group in that the enamel borders of the reentrant folds

are not as thinned as they are in the other hispidus group species. The

reentrant folds also tend to be wider.

Geologic Range: Latest Irvingtonian through early Rancholabrean of

Florida :

Coleman HA: this report; late Illinoian?
Haile VIIA: this report; Sangamon
Williston III: this report; Sangamon
Bradenton 51st St.: this report; Sangamon

leucotis species group
Diagnosis: Characteristically with three, occasionally with four,
roots on the first lower molar.

Sigmodon leucotis Bailey 1902
Diagnosis: The dental pattern of S. leucotis is essentially the same as
that of S. hispidus. Examination of 11 specimens demonstrated 7 with
three roots on the M^ (lingual root missing) and 4 with four roots on
the Mi; of the latter 4, the lingual root was developed as well as in
S. hispidus in 2, and as a small peg in the other 2.
Geologic Range: No fossil record.

Sigmodon peruanus Allen 1897
Diagnosis: Sigmodon peruanus is the largest living Sigmodon species
(Table 6). It is further characterized through loss of the anterior
cingulum on the M2-M3 and by the highly prismatic, flatly worn enamel

-46-

reentrant fold borders. Of 10 specimens examined, 9 evidenced only three
roots on the Mj_, and one specimen, USNM 303005 from Tumbes, Peru,
possessed four well developed roots on this tooth.
Geologic Range: No fossil record.

+Sigmodon curtisi Gidley 1922

= Sigmodon cf S . hispidus , Hibbard, 1952;
Vertebrata, Art. 2: 1-14.

Diagnosis: Sigmodon curtisi is a large cotton rat (Table 6) which demon-
strates a dental pattern as in S. hispidus and S. leucotis. In fact, I
am unable to separate the fossil material of S. curtisi from Recent speci-
mens of S. leucotis. The interparietal bone is absent or vestigial in
S. leucotis, and is well developed in all other Sigmodon species, but as
no skull material of S. curtisi has yet been recovered I retain the in-
tegrity of both species. Two out of six first lower molars of S. curtisi
from the Inglis IA deposit of north Florida evidence 4 roots on the Mj,
placed as in the hispidus species group. Non-Florida curtisi that I have
examined apparently evidence only 3 roots on the M^, but the preserva-
tion of this material is not very good and the teeth are imbedded in the
mandibles, which makes examination quite difficult.
Geologic Range: Irvingtonian:

Arizona: Curtis Ranch; Gidley, 1922

Kansas: Kentuck Assemblage; Hibbard, 1952

California: Vallecito Creek; unpubl.

Florida: Inglis IA; unpubl.

-47-

+Sigmodon hudspethensis Strain 1966
Diagnosis: Unfortunately the holotype, a lower right first molar, UT
40857-10, was lost before it could be properly studied. It had been
measured, however, and these data are included in Table 6. It, there-
fore, seems prudent to designate a neotype. The paratype, another lower
right Mi, is available, but a more recently collected specimen, found
after Strain's (1966) paper was published, makes a better neotype. The
new specimen, collected by William Akersten, is described below.

Neotype: UT 40857-10, a left mandible with all cheek teeth. Both
anterior and posterior portions of the ramus are broken off, but the in-
cisor is present. Collected by William Akersten from the Hudspeth
Fauna (Red Light Local Fauna), Love Formation, Hudspeth County, Texas;
late Blancan?

Amended Diagnosis: Contrary to Strain (1966), the dental pattern of
S. hudspethensis does not vary significantly from that of S. medius .
It does, however, differ from the medius species group in possessing a
well developed labial root on the first lower molar, obvious in the
neotype. The first lower molar of S. hudspethensis also has a fourth
peg-like lingual root, developed as in the two speciments of S. leucotis
mentioned previously that also evidenced peg-like fourth lingual roots.
This lingual root is not as well developed as it is in the hispidus
species group.

Strain (1966) hinted at the fact that Sigmodon hudspethensis was
larger than either S. medius or S. minor, but provided no comparative
measurements. These data are presented in Table 6, and support Strain's
suggestion that S. hudspethensis was "A cotton rat about the size of

-48-

Sigmodon hispidus ..."

Geologic Range: Late Blancan? of Texas:

Hudspeth Fauna: Strain, 1966; Akersten, 1968.

alstoni species group

Sigmodon alstoni Thomas 1880
Diagnosis: Sigmodon alstoni is relegated to its own species group by
virtue of the fact that it is the only Sigmodon species to have developed
grooved upper incisors. The primitive nature of this species is evi-
denced by the fact that it possesses only three roots (lingual root
absent) on the first lower molar.
Geologic Range: No fossil record.

+medius species group
Diagnosis: The members of this group, S . medius and S. minor, are the
the most primitive of known Sigmodon species. They are small cotton
rats (Table 6) with little to no development of labial and lingual
roots on the first lower molar. The incisor capsular process of the
mandible is least developed in these species; a primitive trait for
cricetid rodents in general.

+Sigmodon medius Gidley 1922

= Sigmodon intermedius ; Hibbard, 1938,

Trans. Kansas Acad. Sci. 40: 239-265,

Diagnosis: Sigmodon medius is characterized by having only 2 or 3

roots on the first lower molar. The labial root is always more greatly

developed than is the lingual root, but both are usually tiny pegs and

-49-

may be centrally, rather than lingually and labially, located. The lin-
gual root is about as well developed as is that root in the neotype of S^
hudspethensis, but it is much less developed than is the lingual root in
the hispidus species group. In two out of fifty-four specimens viewed,
solely from the Rexroad Local Fauna, Locality No. 3 (Figure 7), there were
four roots on the first lower molar, but the two accessory roots were
centrally located and very minute.

There are no morphological differences between Sigmodon medius
Gidley and S. intermedius Hibbard, and Figure 8 and Table 6 demonstrate
that they are also equal in size.

Compared to all hispidus group species and S. leucotis and S^
curtisi of the leucotis species group, the dental pattern of S. medius
is much less prismatic. The reentrant folds are not particularly deep
as they are in the other species, nor are they as narrow. In the species
mentioned above the enamel borders of the reentrant folds are greatly
thinned at the termination (apex) of each fold; a development I believe
which is correlated with structural relationships allowing the posterior
enamel border of a particular reentrant fold to double back (Figure 5)
closely adhered to the anterior enamel border of the same fold. An-
other way of describing this pehnomenon in the hispidus species group
and in S. leucotis and S. curtisi is, for example, that the posterior
enamel border of the protoconid is closely allied with the anterior
enamel border of the hypoconid (dental terminology after Hooper, 1957).

The anteroconid of the M-^ in S. medius is generally small and
usually symmetrical, with a wide fourth reentrant fold.

-50-

+Sigmodon medius medius , new subspecies
Holotype: USNM 10519, the holotype of Sigmodon medius described by
Gidley (1922).

Diagnosis: Species characteristics as described for S. medius. This
subspecies differs from S. m. hibbardi by virtue only of smaller size
(Figure 8, Tables 6 and 8). The distinction between the subspecies is
further discussed under S. m. hibbardi.
Geologic Range: Early Blancan through latest Blancan:
Kansas :

Rexroad Local Fauna, Loc . No. 3; Hibbard, 1938
Sanders Local Fauna; unpubl.

Arizona:

Benson Ranch Local Fauna: Gidley, 1922
Tusker Local Fauna: unpubl.

California:

Arroyo Seco Fauna: unpubl
Layer Cake Fauna: unpubl.

+Sigmodon medius hibbardi, new subspecies
Holotype: UMMVP 35093, a right mandible with all cheek teeth, from the
Wendell Fox Pasture locality, Meade County, Kansas; Blancan.
Etymology: This subspecies is named in honor of Dr. Claude Hibbard of
the University of Michigan for his manifold contributions to the field
of vertebrate paleontology, and especially for providing a Pleistocene
stratigraphic sequence in Kansas, complete with fossil mammals, which is
unparalleled in any area of the New World.

Diagnosis: The Wendell Fox Pasture sample of S. m. hibbardi is plotted
in Figure 8, and the log difference plot for this subspecies is seen to

-51-

be different from that of S. m. medius. The mean values of four of the
seven measurements taken (Table 8, Figure 8) in S. m. hibbardi are
statistically significantly different from those taken in S. m. medius
at the 90% level of probability (p < .1). The dental pattern is other-
wise identical to that of S. m. medius.
Geologic Range: Early or late Blancan:

Kansas :

Wendell Fox Pasture Locality: unpubl.

California:

Transition zone between the Arroyo Seco Fauna
and the Vallecito Creek Fauna: unpubl.

+Sigmodon medius, subsp. indeterminate
Sigmodon medius has been recovered from the following localities,
but the small samples preclude statistical treatment. Measurements of
the samples are presented in Table 6.
Kans as :

Benders Local Fauna: unpubl., early Blancan
Texas :

Blanco Local Fauna: unpubl., Blancan
Florida:

Haile XVA Local Fauna: unpubl., late Blancan?
Nebraska:

Sand Draw Local Fauna: unpubl., Blancan.

■52-

+Sigmodon minor Gidley 1922

= Sigmodon hilli; Hibbard, 1941. State Geol.
Surv. Kansas, Bull. 38: 197-220.

Diagnosis: Sigmodon minor is the smallest known species of Sigmodon
(Table 6), and appears to have been derived directly from Sigmodon
medius . As seen in Figure 8 there are no dental proportional differ-
ences between S. minor and S. medius. The major difference between
S. minor and S. medius is size. The reentrant folds are deeper and
narrower in S . minor than they are in S . medius , especially obvious in
teeth with little wear, but well worn teeth of S. minor approximate
very closely those of S. medius.

There are no qualitative differences between Sigmodon minor Gidley
and S. hilli Hibbard, and Figure 8 and Table 6 demonstrate that these
species are equal in size as well.

The Borchers Sigmodon minor include five first lower molars with
4 roots; out of 27 specimens studied. A random selection of 27 S.
medius included two with 4 roots. This difference is not statistically
significant (jrwith Yates Continuity Correction= .657; one degree of
freedom) .

Geologic Range: Irvingtonian:
Kansas :

Borchers Local Fauna: Hibbard, 1941
Arizona:

Curtis Ranch Local Fauna: Gidley, 1922.

Evolution in the Genus Sigmodon
Living sigmodont rodents include the genera Reithrodon, Neotomys ,

-53-

Sigmodon, and Holochilus (Hershkovitz, 1955). All but Sigmodon are re-
stricted to South America. The genus Sigmodon exists now in northern
South America and extends northward through Central America into the
southern United States. All sigmodonts possess a comlex penis and a
long-wide palate. According to Hershkovitz (1955), "Diagnostic charac-
ters of sigmodonts include reduced outer hind toes, webbed middle hind
toes, spinous process of zygomatic palate, specialized posterior palatal
region, simplified molars characterized by absence or obsolescence of a
functional mesoloph(id) in, at least, My^-, and the S-shaped enamel pat-
tern of M3." Hershkovitz (1955) further suggests that Sigmodon is the
most generalized form of the living sigmodonts.

"Evolution of American cricetines has paralleled that of the New
World Perissodactyla. The primitive brachydont, buno-mesolophodont
cricetines have survived, like the tapir, in forested parts of the
range. Like the horse, the progressive branch of cricetines, with
mesoloph absent or vestigial, has become increasingly specialized for
life in open country and a diet of grasses. The young have lost, or are
in the process of losing, the distinctive juvenal phase of pelage compar-
able to the spotted coat of young tapirs and other woodland ungulates.
The molars have become higher, their crown surfaces flatter and tending
toward lamination. The third molar of grazing cricetines is nearly
always precociously functional. This condition provides the needed
maximum grinding surface which, in young ungulates, is supplied by
deciduous premolars.

Sigmodont rodents are members of the relatively recently evolved,
essentially grazing, New World mammalian fauna of open country. Morpho-
logically, they merge into the phyllotine group (Phyllotis, Hesperomys,

-54-

and others) by a combination of characters. In both groups the denti-
tion is fundamentally similar (Hershkovitz , 1955)." Hershkovitz (1962)
further states, "Phyllotines probably evolved from the same line that
gave rise to akodont rodents. The sigmodonts ... are more special-
ized but appear to be progressive offshoots from the main phyllotine
line."

I. Dental Evolution from Pliocene to Recent Time
The following information is now available:

1) The most generalized South American phyllotine rodents,
from which Sigmodon was supposedly derived, possess 4
roots on the first lower molar (Hershkovitz, 1962).

2) The earliest fossil Sigmodon species known is S. medius.
This species is found in late Pliocene deposits, has
the most primitive dental pattern of all recognized
Sigmodon species (see the discussion under S. medius

in the Species Account), and characteristically evi-
dences 2 or 3, rarely 4, roots on the first lower
molar. The lingual and labial roots (accessory roots)
are typically minute and most often located more to-
wards the center of the tooth than they are in the hispi-
dus species group.

3) Sigmodon hudspethensis , a larger animal than S . medius ,
approximating the size of S. curtisi and S. hispidus,
is the only Sigmodon species other than S. medius known
from early Pleistocene (late Blancan) deposits. This
species may possess 4 roots on the first lower molar,

-55-

the accessory roots developed more fully than they are
in S. medius, but less fully than in the hispidus species
group.

4) Sigmodon curtisi, with a dental pattern as in the his-
pidus species group and as in S. leucotis, first appears
during the early Irvingtonian (early middle Pleistocene)
in North America, and may have 4 well developed roots

on the first lower molar.

5) Three living species of Sigmodon in the New World char-
acteristically evidence 3 roots on the first lower molar,
and all may evidence 4 well developed roots occasionally.
One of these species, S. leucotis, is restricted to iso-
lated populations at higher elevations in Mexico (Baker,
pers . comrnun.) surrounded by hispidus group species.
Teeth of S. leucotis are similar to those of S. curtisi.
The remaining two species, S. alstoni and S . peruana? ,
are restricted to South America, and appear to have
evolved in isolation from their North American relatives.

6) The hispidus species group (M^ with 4 well developed
roots and complex molars) is first seen in North America
during the latest middle Pleistocene (latest Irvington-
ian) .

The fossil record for cotton rats is one of the most complete of
any genus of small mammal that existed and evolved in the New World
through Pleistocene epoch. A perfectly adequate evolutionary scheme

-56-

can be developed for the genus based solely on the fossil material
available from deposits in the United States. However, this scheme will
undoubtedly be distasteful to many people, as it requires an evolution-
ary reversal in root count. An alternative model may also be developed,
but it requires hypothesizing extinct creatures which are not repre-
sented by fossil material.

If Hershkovitz (1955, 1966) is correct, the first Sigmodon species
evolved from a generalized phyllotine rodent in South America and then
dispersed northward. The living phyllotine forms which according to
Hershkovitz (1962) most closely approximate a generalized beast, such
as Phyllotis darwini, Galenomys garleppi, and Calomys spp. , possess 4
roots on the first lower molar. As demonstrated by the information
summarized on page 54 , there, is an increase in numbers of roots on the
first lower molar and increased dental complexity in fossil Sigmodon
from the late Pliocene S. medius (2 or 3 roots) to the late Pleistocene
hispidus species group (4 roots). Therefore, if the evolutionary trend
disclosed by the available fossil material is valid, an evolutionary
reserval in root count is indicated (4-3-2-3-4). This sytem is outlined
in Figure 6 .

The alternative model would require that: 1) S. medius does not
give rise to S. hudspethensis , 2) S. curtisi is not ancestral to the
hispidus species group, and 3) evolution to the curtisi stage in
North America is paralleled in Central or South America by another
line leading to a curtisi-like animal in which, however, all grades
retain 4 well developed roots on the first lower molar. The resultant
end product of this line then evolves into the hispidus species
group .

-58-

i\

\

1^1

\

Ml AWV3

6

V 1/

!^

J!

iivi awj iin Aim
mwomiAm #rarr«

,Q

TI

,fi

b

0

.£>

01

H

•W

ti

•H

EC

00

w

6

-60-

PLIOCENE

PLEISTOCENE

£

s*

^

is,

a*

S?

1

3

53

03
53
!/>
O)

5'

=3

00

ST

=3

53

Sj'

53

53

£5

5-

Late Blancan

Irving ton/an

c
a
j

c

s
C

Rexroad-S
Fox Canyon

CO
CO

53

1
Co

*'

o

53

Sanders- S
Deer Park

5j-

00

>
f"3

5j-
Cb

on

1

ft

=3

3

to

p.

ft;

3

ft.

CO

o'
3

■5:

Co

a

1

CD

ft^

ft'

On

§-

Co

ft'

53

on

ft

'~I

Do

«6

53
to

S1

On

C~5

ft

30

§

■>

on

N
5>

£

l<0

=0
ft

8

53

<1

3

~i

ft-
1

re

5>-
1

•to

ft

On

ICo

5*

5

C6

ft.

ft'

C5

5'

On

Figure 8 — Composite ratio diagram of measurements of the
lower dentition and mandible in some members
of the Sigmodon medius species group from
Kansas and Arizona. A = Sigmodon medius
medius ; Benson Ranch, B - Sigmodon medius
medius ; Tusker, C = Sigmodon minor; Curtis
Ranch, D - Sigmodon medius hibbardi; Wendell
Fox Pasture, E = Sigmodon medius medius; Rex-
road Loc. 3, F = Sigmodon medius medius;
Sanders, G = Sigmodon minor; Borchers.
MA = mandibular alveolar length, 1 = length,
w = width. The standard is the Coleman IIA
Sigmodon bakeri.

-62-

-63-

II. Ecological Considerations

Early and middle Pliocene deposits in the United States contain
cricetine species referrable to Copemys , Peromyscus , Tregomys , or
Gnomomys . Sigmodon has not been recorded from deposits of the above
ages, and yet is the most abundant cricetine in deposits of early and
late Blancan age in the United States. One might call the appearance
of Sigmodon in these deposits "sudden." Sigmodon entrance into the United
States also corresponds in general to the influx of rooted-cheek- toothed
microtines such as Pliopotamys, Ophiomys , Pliophenacomys , Pliolemmus,
and Ogmodontomys . The beginning of this microtine immigration is re-
corded, for example, in the late Hemphillian McKay Reservoir Local
Fauna of Oregon (Shotwell, 1956), a fauna which contains Teleoceras and
the extinct microtine Prosomys mimus . Teleoceras is unknown in early
Blancan faunas such as Rexroad, and is generally believed to have be-
come extinct by the end of the Hemphillian (Hirschfeld and Webb, 1968).

It is interesting to speculate how, in the earlier Pleistocene,
the vole-like Sigmodon could be sympatric with such an array of rooted-
cheek- toothed microtines, and the answer, I believe, is readily appar-
ent. The dental evolutionary trends which characterize Sigmodon in
general, and especially the medius -minor line, are those which point to
convergence with the true (rootless-cheek-toothed) voles; notably hypsodonty
and involution. Disappearance of the medius species group coincides ap-
proximately with the appearance of the rootless-cheek-toothed microtines
(Pitymys , Pedomys , Neodon) which first appear in the Cudahy Fauna of
Kansas, of Kansan age (Paulson, 1961; Hibbard, 1967).

These considerations suggest that the earliest Sigmodon species
were not in direct competition with the early, rooted-cheek- toothed

-64-

microtines, and were filling the grassland niche later to be filled
more adequately by the rootless-cheek-toothed microtines. I further sug-
gest that most of the early rooted-cheek-toothed microtines were sylvan
and aquatic rather than pastoral, as is the case in the living North
American rooted-cheek-toothed forms Clethrionomys , Phenacomys , and
Ondatra. This hypothesis is partly substantiated by Hibbard's and
Zakrzewski's statement (1967) that Pliopotamys was ancestral to the
aquatic Ondatra.

Sigmodon is found today primarily in the Lower Austral and Tropical
life zones in North America. Microtines are generally Holarctic mammals,
and only Microtus mexicanus , Pitymys pinetorum, P. quasiater, and tiny
populations of three species, Pitymys guatemalensis , Microtus umbrosus,
Pitymys oaxacensis are able to exist in life zones lower and warmer than
the Upper Austral zone. Both P. pinetorum and P. quasiater are sylvan
fossorial species, and thus in general Microtus and Pitymys are rare in
grasslands in the southern United States, Mexico, and Central America.
This niche is today filled primarily by Sigmodon, although other genera
such as Reithrodontomys and Oryzomys also contribute an appreciable
amount of biomass.

III. Fossil Material Examined
Sigmodon medius

S. m. hibbardi:

1. Wendell Fox Pasture Locality, Meade County, Kansas:

UMMVP 35093, 57050-57054

2. Transition between Arroyo Seco Fauna and Vallecito
Creek Fauna, 3200 to 3250 feet from top of sequence,

-65-

San Diego County, California:

LACM 1588/4442, 1451/4447

S. m. medius :

1. Arroyo Seco Fauna, 4900 to 5350 feet from top of
sequence, San Diego County, California:

LACM 6554/13754, 6552, 6550, 6552/12505

2. Layer Cake Fauna, 6750 feet from top of sequence,
San Diego County, California:

LACM 1711/7005

3. Rexroad Local Fauna, Locality 3, Meade County, Kansas:

UMMVP 29162, 29669, 31085, 31086, 41193,
44589, 56249

4. Sanders Local Fauna, Localities 1, 2, 4, Meade
County, Kansas:

UMMVP 32003-32005, 31997, 31998, 56247,
56248, 50263, 50264

5. Tusker Local Fauna, Locality 15-24, Graham County,
Arizona:

UAVP 899, 905, 910, 913, 914, 922, 924, 925,
927-936, 938-941, 945, 949, 966-970, 972-1003,
1007,1020, 1023-1025, 1030, 1036-1042, 1054,
1056, 1057, 1059-1069, 1075, 1077-1079, 1081,
1087, 1089, 1090, 1094, 1100, 1104, 1105,
1111, 1113, 1114, 1118, 2494-2510, 2513, 2514,
2519, 2700-3053

-66-

Sigmodon medius

S. m. medius

6. Benson Ranch Local Fauna, Cochise County, Arizona:
USNM 10520-10523
S. medius, subsp. indeterminate

1. Benders Local Fauna, Meade County, Kansas:

UMMVP 45820

2. Haile XVA Local Fauna, Alachua County, Florida:

UF 12334, 12336, 12338, 12342

3. Sand Draw Local Fauna, Brown County, Nebraska:

UMMVP 57056

4. Blanco Local Fauna, Crosby County, Texas:

MUVP 7146

Sigmodon minor

1. Borchers Local Fauna, Meade County, Kansas:

UMMVP 35766, 56244, 56245, 51302, 51312,

51313, 51307, 51305, 51303, 51309, 51311,

51314, 51308, 51306, 51304,

2. Curtis Ranch Local Fauna, Cochise County, Arizona:

USNM 10512-10518, 16608-16611

Sigmodon hudspethensis

1. Hudspeth Fauna, Madden Arroyo, Hudspeth County,
Texas :

UTMM 40240-1, 40240-2

-67-

2. Hudspeth Fauna;

Red Light Local Fauna, Hudspeth County, Texas:

UTMM 40857-10, 40857-11

Sigmodon curtisi

1. Curtis Ranch Local Fauna, Cochise County, Arizona:

USNM 10511, 16606, 16607, 16605

2. Kentuck Assemblage, McPherson County, Kansas:

UKMVP 7361

3. Vallecito Creek Fauna, 900 to 1520 feet from top
of sequence, San Diego County, California:

LACM 1615/4389, 1114/3394, 1297/6941, 1114/3395,
1461/4445, 1615/4396, 1615/4398

4. Inglis IA Local Fauna, Citrus County, Florida:

UF 15155

Sigmodon hispidus

1. Reddick IA, rodent beds, Marion County, Florida:

UF 14347-14360

2. Reddick IIC, Marion County, Florida:

UF 15204

3. Devil's Den, Levy County, Florida:

UF 13440, 13453, 13444, 13593-13600

4. Kendrick I, Marion County, Florida:

UF 2658

5. Maximo Moorings, Pinellas County, Florida:

UF 3062

-68-

6. Haile VIIIA, Alachua County, Florida:

UF 9844, 15153, 12680-12684

7. Haile XIB, Alachua County, Florida:

UF 13471-13592

8. Haile XIIIA, Alachua County, Florida:

UF 13096-13097

9. Haile XIIIC, Alachua County, Florida:

UF 13049

10. Ichetucknee River, Gilchrist County, Florida:

UF 15205

11. Withlacoochee River, Locality 7A, Citrus County,
Florida:

UF 15206

12. Arredondo IA, Alachua County, Florida:

UF 15207

13. Arredondo IIB, Alachua County, Florida:

UF 12589

14. Arredondo IIC, Alachua County, Florida:

UF 12297-12303

-69-

Table 6

Measurements (in mm) of the lower dentition
and mandible in living and extinct species of Sigmodon

MA = mandibular alveolar length, 1 = length, w = width

MA

O.R.

Sigmodon medius

Benson Ranch

Tusker

Rexroad , Loc . 3

Sanders

Benders

Sand Draw

Blanco

Haile XV

Wendell Fox Pasture

3

25

8

5

5.75
5.70
6.17
6.07

5.59-6.01
5.30-6.13
5.97-6.42
5.76-6.44

1

5.58

-

3

6.27

6.18-6.38

Sigmodon minor

Curtis Ranch
Borchers

Sigmodon curtisi

Curtis Ranch
Inglis IA
Kentuck

Sigmodon hudspethensis

Hudspeth and Red Light

Sigmodon bakeri

Coleman IIA
Willis ton III

25

11

2

5.30
5.63

6.90

6.28
6.69

4.97-5.80
5.27-6.00

1

7.03

-

4

6.68

6.27-6.91

1

6.78

-

5.93-6.71
6.40-6.98

Table 6 - continued

-70-

MA (continued)

Sigmodon hispidus

Reddick IA

Florida (Recent)

Texas (S. h. berlandieri)

Sigmodon ochrognathus

Texas
Sigmodon fulviventer

Mexico; Durango
Sigmodon alleni

Mexico; Michoacan and Oaxaca

Sigmodon leucotis

Mexico; Durango, Morelos,

Guererro, Guanajuato

Sigmodon peruanus

Ecuador and Peru

N X

O.R.

18 7.27 6.83-7.65

30 7.26 6.56-7.80

8 6.68 6.34-6.93

4 6.12 5.81-6.29

8 7.17 6.95-7.34

5 6.64 6.38-7.08

11 6.76 6.44-7.05

10 7.72 7.17-8.23

LM,

Sigmodon medius

Benson Ranch

Tusker

Rexroad, Loc. 3

Sanders

Benders

Sand Draw

Blanco

Haile XV

Wendell Fox Pasture

4 2.09 1.91-2.29
25 2.03 1.86-2.26
39 1.98 1.80-2.26
10 2.05 1.94-2.14

1 1.98

5 1.99 1.98-2.00

2 2.26 2.18-2.58
5 2.16 2.10-2.22

-71-

Table 6 - continued

LM (continued)

Sigmodon minor

O.R.

3

2.41

2.28-2.49

5

2.35

2.22-2.50

1

2.42

-

Curtis Ranch 9 1.92 1.73-2.19

Borchers 38 1.88 1.72-2.07

Sigmodon curtisi

Curtis Ranch
Inglis IA
Ken tuck

Sigmodon hudspethensis

Hudspeth and Red Light 4 2.34 2.19-2.58

Sigmodon bakeri

Coleman IIA 20 2.21 2.04^2.46

Williston III 3 2.46 2.28-2.61

Sigmodon hispidus

Reddick IA

Florida (Recent)

Texas (S. h. berlandieri)

Sigmodon ochrognathus

Texas 4 2.15 2.10-2.20

Sigmodon fulviventer

Mexico; Durango 8 2.27 2.20-2.39

Sigmodon alleni

Mexico; Michoacan and Oaxaca 5 2.41 2.27-2.57

Sigmodon Leucotis

Mexico; Durango, Morelos,

Guererro, Guanajuato 9 2.54 2.44-2.75

Sigmodon peruanus

Ecuador and Peru 9 2.68 2.31-3.03

18

2.49

2.24-2.72

30

2.47

2.11-2.73

8

2.35

2.18-2.49

■72-

Table 6 - continued

N

2
X

O.R.

4

1.51

1.43-1.57

26

1.52

1.35-1.71

23

1.56

1.33-1.81

9

1.54

1.40-1.69

1

1.67

-

1

1.53

-

1

1.41

-

1

1.49

-

5

1.60

1.53-1.65

Sigmoid on medius

Benson Ranch

Tusker

Rexroad , Loc . 3

Sanders

Benders

Sand Draw

Blanco

Haile XV

Wendell Fox Pasture

Sigmodon minor

Curtis Ranch
Borchers

Sigmodon curtisi

Curtis Ranch
Inglis IA
Kentuck

Sigmodon hudspethensis

Hudspeth and Red Light

Sigmodon baker i

Coleman IIA
Williston III

Sigmodon hispidus

Reddick IA

Florida (Recent)

Texas (S. h. berlandieri)

Sigmodon ochrognathus

Texas
Sigmodon fulviventer

Mexico; Durango

9
40

13
2

1.42
1.38

1.75

1.61
1.76

1.30-1.62
1.22-1.54

3

1.87

1.85-1.89

6

1.80

1.75-1.85

1

1.91

-

1.73-1.76

1.49-1.75
1.66-1.85

18

1.82

1.67-1.95

30

1.77

1.59-1.96

8

1.72

1.57-1.79

1.82

1.50-1.82

1.73-1.91

■73-

Table 6 - continued

LM2 (continued)

Sigmodon medius

Sigmodon minor

O.R.

Sigmodon alleni

Mexico; Michoacan and Oaxaca 5 1.76 1.65-1.87

Sigmodon leucotis

Mexico; Durango, Morelos ,

Guererro, Guanajuato 9 1.78 1.63-1.85

Sigmodon peruanus

Ecuador and Peru 9 2.13 1.83-2.29

LM~

Benson Ranch 4 1.94 1.88-2.04

Tusker 26 1.97 1.61-2.28

Rexroad, Loc. 3 15 1.93 1.75-2.14

Sanders 7 1.95 1.88-2.14

Benders 1 2.15

Sand Draw -

Blanco 2 2.02 2.01-2.02

Haile XV 1 1.93 -

Wendell Fox Pasture 2 2.02 1.88-2.16

Curtis Ranch 6 1.84 1.64-2.12

Borchers 40 1.75 1.40-2.06

Sigmodon curtisi

Curtis Ranch 1 2.55

Inglis IA 2 2.25 2.14-2.35

Kentuck 1 2.49

Sigmodon hudspethensis

Hudspeth and Red Light 2 2.21 2.07-2.35

■ 74-

Table 6 - continued

LM3 (continued)

Sigmodon bakeri

Sigmodon medius

O.R.

Coleman IIA 13 1.98 1.72-2.37

Williston III 1 2.02

Sigmodon hispidus

Reddick IA 18 2.59 2.24-3.00

Florida (Recent) 30 2.56 2.17-2.89

Texas (S. h. berlandieri) 8 2.40 2.02-2.74

Sigmodon ochrognathus

Texas 1 1-75

Sigmodon fulviventer

Mexico; Durango 8 2.73 2.59-2.92

Sigmodon alleni

Mexico; Michoacan and Oaxaca 5 2.33 2.25-2.40

Sigmodon leucotis

Mexico; Durango, Morelos,

Guererro, Guanajuato 9 2.20 2.05-2.41

Sigmodon peruanus

Ecuador and Peru 9 2.65 2.25-2.87

WM,

Benson Ranch

Tusker

Rexroad , Loc . 3

Sanders

Benders

Sand Draw 1 1.37

4

1.44

1.43-1.71

26

1.40

1.29-1.51

41

1.43

1.27-1.58

11

1.40

1.21-1.50

■75-

Table 6 - continued

O.R.

5

1.39

1.20-1.68

2

1.27

1.25-1.29

6

1.53

1.51-1.57

3

1.62

1.55-1.71

5

1.67

1.51-1.74

1

1.56

-

WMi (continued)

Sigmodon raedius (continued)

Blanco

Haile XV

Wendell Fox Pasture

Sigmodon minor

Curtis Ranch 9 1.31 1.21-1.44

Borchers 39 1.31 1.17-1.48

Sigmodon curtisi

Curtis Ranch
Inglis IA
Ken tuck

Sigmodon hudspethensis

Hudspeth and Red Light 4 1.48 1.40-1.57

Sigmodon bakeri

Coleman HA 20 1.51 1.35-1.86

Williston III 3 1.61 1.56-1.76

Sigmodon hispidus

Reddick IA 18 1.73 1.63-1.86

Florida (Recent) 30 1.65 1.55-1.88

Texas (S. h. berlandieri) 8 1.63 1.56-1.77

Sigmodon ochrognathus

Texas 4 1.61 1.57-1.64

Sigmodon fulviventer

Mexico; Durango 8 1.71 1.65-1.81

Sigmodon alien!

Mexico: Michoacan and Oaxaca 5 1.63 1.53-1.72

Sigmodon leucotis

Mexico; Durango, Morelos ,

Guererro, Guanajuato 9 1.67 1.58-1.75

Table 6 - continued

-76-

KM-, (continued)
Sigmodon peruanus

Ecuador and Peru

1.91

O.R.

1.76-2.07

WM„

Sigmodon medius

Benson Ranch

Tusker

Rexroad , Loc . 3

Sanders

Benders

Sand Draw

Blanco

Haile XV

Wendell Fox Pasture

Sigmodon minor

Curtis Ranch
Borchers

Sigmodon curtisi

Curtis Ranch
Inglis IA
Kentuck

Sigmodon hudspethensis

Hudspeth and Red Light

Sigmodon bakeri

Coleman IIA
Williston III

Sigmodon hispidus

Reddick IA

Florida (Recent)

Texas (S. h. berlandieri)

4

1.54

1

43-1.71

26

1.55

1

39-1.76

23

1.56

1

39-1.68

9

1.59

1

53-1.68

1

1.72

-

1

1.51

-

1

1.49

-

1

1.63

-

5

1.67

1

62-1.71

9 1.43
40 1.42

1.27-1.55
1.20-1.55

3 1.89 1.85-1.92
6 1.87 1.73-1.97
1 1 . 85

2 1.52 1.39-1.65

13 1.72 1.55-1.86
2 1.91 1.74-2.08

18 2.02 1.88-2.14
30 1.97 1.79-2.13

■77-

Table 6 - continued

Sigmodon minor

O.R.

1.81 1.77-1.84

WM9 (continued)

1 N

Sigmodon ochrognathus

Texas 4

Sigmodon fulviventer

Mexico; Durango 8 1.93 1.86-2.05

Sigmodon alleni

Mexico; Michoacan and Oaxaca 5 1.87 1.71-2.00

Sigmodon leucotis

Mexico; Durango, Morelos,

Guererro, Guanajuato 9 1.90 1.81-2.06

Sigmodon peruanus

Ecuador and Peru 9 2.23 2.10-2.33

Sigmodon medius

Benson Ranch

Tusker

Rexroad, Loc . 3

Sanders

Benders

Sand Draw

Blanco

Haile XV

Wendell Fox Pasture

WM
3

4

1.48

1.33-1.62

26

1.45

1.31-1.64

15

1.48

1.34-1.64

7

1.53

1.44-1.64

1

1.72

-

2

1.47

1.45-1.52

1

1.53

-

2

1.59

1.52-1.65

Curtis Ranch 6 1.35 1.30-1.40

Borchers 40 1.36 1.20-1.50

-78-

Table 6 - continued

O.R.

1

1.81

-

2

1.79

1.76-1.82

1

1.83

-

WMo (continued)

Sigmodon curtisi

Curtis Ranch
Inglis IA
Kentuck

Sigmodon hudspethensis

Hudspeth and Red Light 2 1.73 1.70-1.76

Sigmodon bakeri

Coleman IIA 12 1.60 1.43-1.69

Williston III 1 1.61

18

1.88

1.77-1.96

30

1.91

1.77-2.10

8

1.79

1.65-1.89

Sigmodon hispidus

Reddick IA

Florida (Recent)

Texas (S . h. berlandieri)

Sigmodon ochrognathus

Texas 2 1.73 1.70-1.76

Sigmodon fulviventer

Mexico; Durango 8 1.98 1.93-2.09

Sigmodon alleni

Mexico; Michoacan and Oaxaca 5 1.82 1.80-1.88

Sigmodon K-ucotis

Mexico; Durango, Morelos,

Guererro, Guanajuato 9 1.75 1.66-1.90

Sigmodon peruanus

Ecuador and Peru 9 2.13 2.00-2.35

•79-

Table 7

Measurements (in mm) of the lower dentition and mandible of
Sigmodon samples from the Vallecito Creek-Fish Creek beds

LCF = Layer Cake Fauna; ASF = Arroyo Seco Fauna;

VCF-ASF = transition zone between Arroyo Seco Fauna

and Vallecito Creek Fauna; VCF = Vallecito Creek Fauna;

MA = mandibular alveolar length; 1 = length; w = width

MA

L^

LM2

LM3

WML

WM2

WM3

Sigmodon curtisi

VCF 900

-

-

1.84

-

-

2.02

-

VCF 1100

7.64

-

1.94

2.66

1.98

2.29

2.05

VCF 1450

-

-

1.85

-

-

1.99

-

VCF 1520

7.64

2.79

2.06

2.58

1.85

2.00

1.97

6.99

2.47

2.00

2.82

2.07

2.24

2.04

-

2.41

1.96

2.47

2.00

2.27

1.83

-

-

1.82
1.92

-

1.75
1.93

1.86
2.10

-

X

7.42

2.56

2.63

1.97

Sigmodon medius hibbardi

VCF-ASF 3200

6.10

2.14

1.73

2.25

1.47

1.70

1.60

-

-

1.60

2.08

-

1.50

1.48

6.20

2.24

1.77

2.04

1.48

1.65

1.63

VCF-ASF 3250

6.53
6.28

-

1.69
1.70

2.04
2.10

-

1.73
1.65

1.62

X

2.19

1.48

1.58

Sigmodon medius medius

ASF 4900

5.68

-

1.54

2.06

1.45

1.50

1.36

ASF 5100

-

-

1.70

2.21

-

1.64

1.59

5.97

-

1.67

1.96

1.66

1.60

1.51

ASF 5350

5.92

-

1.59

1.80

-

1.56

1.39

6.15
5.93

2.14
2.14

1.62
1.62

1.76
1.96

-

1.55
1.57

1.49

X

1.56

1.47

Sigmodon medius medius

LCF 6750

5.79

2.03

1.72

1.89

1.44

1.51

1.45

-

-

1.51
1.62

2.02
1.96

-

1.44
1.48

1.47

X

5.79

2.03

1.44

1.46

-80-

Table 8

Statistical comparison of Sigmodon samples from the

Wendell Fox Pasture (WFP) and Rexroad , Loc . 3 (R3)

deposits

— o

N = number of specimens; X = mean; S = variance;

t = student's t value; p = probability value;

* = statistically significant difference

WFP

R3

N

X

s2

N

X

s2

t

P

MA length

3

6.27

.007

8

6.17

.028

.01

.90C

Length M,

5

2.16

.002

39

1.98

.015

*3.09

.010

Length M2

5

1.60

.002

23

1.56

.009

.90

•4> P>

.3

Length Mo

2

2.02

.040

15

1.93

.016

.90

•4>P>

.3

Width ML

6

1.53

.001

41

1.43

.004

*10.46

.001

Width M2

5

1.67

.003

23

1.56

.005

*3.19

.010

Width M3

2

1.59

.003

15

1.48

.006

*1.89

• 1>P7.

05

-81-

+Pitymys arata new species
Holotype: UF 11685 (Figure 9) , a right dentary with all lower molars.
Both the coronoid process and the angular process are broken, but the
condyloid process is complete. The lower incisor is broken at the
anterior edge of the ramus.

Referred specimens: UF 11684, a partial right dentary with incisor but
no cheek teeth.

Horizon and locality: Coleman HA Local Fauna, Sumter County, Florida;
late Irvingtonian, ?Illinoian.

Diagnosis: Pitymys arata is a large vole, the size of Microtus richard-
soni (Table 9), and is thus larger than any known Pitymys , extinct or
extant. Configuration of the first lower molar clearly identifies P.
arata as a member of the latter genus. The deep sixth reentrant angle
and posterior sloping anterior enamel border of the fourth triangle sep-
arates P. arata from P. ochrogaster (Figure 5 and Martin, in prep.).

The pattern of the first lower molar combines characteristics of
both P. pinetorum and P. quasiater. The anterior loop, with incipient
sixth and seventh triangles, but with shallow eighth and ninth reentrant
angles, best approximates P. pinetorum. These reentrant angles are
typically deep in P. quasiater. The closed anterior loop in P. arata
(caused by deep penetrance of reentrant angles 6 and 7) is characteris-
tic of P. quasiater. Yet I have seen a few specimens of P. pinetorum
nemoralis (the largest P. pinetorum subspecies) from Kansas with this
feature. The anterior loop on the M, is widely open in Florida fossil
and extant Pitymys pinetorum. Pitymys arata further differs from P.
pinetorum, but agrees with P. quasiater, in possessing a reduced capsular

-82-

process, especially bulbous in the largest, oldest individuals. The un-
pronounced capsular process of P. arata, related to short length of the
lower incisor, is a primitive trait for all voles.

Etymology: This species is named for Dr. Andrew A. Arata of the World
Health Organization (Geneva) for his guidance early in my graduate
career.

Remarks: The following is a tentative classification of those fossil
and living voles that have cementum in the reentrant angles and rootless
cheek teeth. Emphasis is placed on New World species, and the species
list for neither the Old World nor New World is meant to be exhaustive.
Recent species of the New World are listed only if I have transferred
them from the genus in which they were located by Hall and Kelson (1959)
to some other genus. The only Old World species I have transferred to
another genus are those of the taxon Eothenomys , which for my conven-
ience and for reasons which will be explained, appear to be closely
related to those of the genus Anteliomys . Old World species of other
genera are listed below only for illustrative purposes and do not, as in
the case of the New World species, indicate a change in taxonomic status.
This classification is based solely upon the dentition of these animals,
and particularly upon evolution of the Mi as seen in the fossil record
(OW = Old World, NW = New World):

Grade I

Arvicola Lacepede 1799- OW
+A. grccnii Hinton
+A. praeceptor Hinton
A. amphibius (Linnaeus)

Phaiomys Blyth 1863- OW, NW

-83-

Phaiomys (cont'd) =+Allophaiomys Kormos 1933, Neue Jahrb. Paleont.

& Miner., Beil-Bd. 69B: 323-346. Munchen.

+P. pliocaenicus Kormos- OW, NW

= +Microtus (Pedomys) llanensis, Hibbard,
1952. Vertebrata, Art. 2: 1-14.

+P. laguroides Kormos- OW
+P. ruffoi- Pass a- OW
P. leucurus Blyth - OW

Grade II

Neodon Hodgson 1849- OW, NW

+N. paroperarius (Hibbard)- NW

= +Microtus paroperarius Hibbard 1944.
Bull. Geol. Soc. Amer., Vol. 55:
707-754.

N. irene Thomas- OW

N. sikimensis Hodgson- OW

N. carruthersi Thomas- OW

Grade III

Pitymys McMurtie 1831- OW, NW
= Pedomys Baird 1857, in Rept. Expl. Surv., 8(1):
517.
+P. involutus (Cope)- NW

=+Arvicola involuta Cope 1871. Proc . Amer.
Philos. Soc. : 89.

=+Microtus (Pitymys or Pedomys) involutus ,
Hibbard, 1955. Proc. Acad. Nat. Sci,
Phila., Vol. CVII: 87-97.

+P. dideltus (Cope)- NW

=+Arvicola didelta Cope 1871. Proc. Amer.
Philos. Soc. ; 89.

=+Microtus (Pitymys or Pedomys) dideltus ,

Hibbard, 1955. Proc. Acad. Nat. Sci,
Phila., Vol. CVII: 87-97.

+P. hibbardi Holman- NW

+P. arata Martin- NW

+P. meadensis Hibbard- NW

+P. arvaloides Hinton- OW

+P. gregaloides Hinton- OW

P. guatemalensis (Merriam)- NW

-84-

Pitymys McMurtie (cont'd)

=Microtus guatemalensis Merriam 1898. Proc.
Biol. Soc. Wash., 12: 108.

P. oaxacensis (Goodwin)- NW

=Microtus oaxacensis Goodwin 1966. Amer.
Museum Novitates, No. 2243: 1-4.
+P. llanensis (Hibbard)

=Microtus (Pedomys) llanensis Hibbard 1944.

Bull. Geol. Soc. Amer., Vol. 55: 707-754.

P. ochrogaster (Wagner)- NW

= Hypudaeus ochrogaster Wagner 1842. in
Schreber, Die Saugthiere . . . ,
suppl., 3: 592.

=Microtus (Pedomys) ochrogaster, Hall and
1959. The Mammals of North America,
Ronald Press, New York, Vol. II.

+P. mcknowni Hibbard 1937. Jour. Mammal.,
18(2): 235.

Microtus Schrank 1798- OW, NW
Alticola Blanford 1881- OW
Neofiber True 1884- NW
Anteliomys Miller 1896- OW

= Eothenomys Miller 1896. North Amer. Fauna,
No. 12: 45.

My studies of fossil and living voles demonstrate three major
grades of dental evolution, each with various side branches. Although
characteristics of the other molars are useful taxonomically , only
the first lower molar appears to be a reliable indicator of evolution-
ary grade. The first grade, including Arvicola and Phaiomys , may be
characterized by the presence of only three closed triangles on the
M^ . This grade, first seen in either the upper Pliocene or lower
Pleistocene of the Old World (Hinton, 1926; Kowalski, 1960; Kurteli, 1968)
is clearly derivable (probably polyphyletically) from the Old World
rooted-cheek-toothed Mimomys . Arvicola may be separated from Phaiomys
by characteristics of the Mo ; this tooth in Arvicola retains a well

■85-

developed fourth triangle, making the pattern of M3 identical to that of
M£. The fourth triangle of M„ in Phaiomys is reduced or is absent. The
anterior loop of the M, tends to be simpler in Arvicola, but in reality
there probably is continuing complexity in this feature from Arvicola
through Phaiomys to Neodon.

The genus Neodon (which may eventually be shown to include
Orthriomys , Proedromys , and Blanfordimys) represents the second evolu-
tionary grade, and is intermediate in complexity of the M-^ between
grades I and III. The M, in grade II may have from three to five well
developed triangles, of which only three or four are usually closed.
Triangles four and five are usually confluent and open widely into the
anterior loop, but these triangles are less confluent than in the
genus Pitymys, and the anterior loop is, in Neodon, usually more elongate,
complex, and asymmetrical than it is in Pitymys . Phaiomys ruffoi
(Pasa, 1947) of middle or early Pleistocene sediments from Italy is
possibly the ancestor for living Neodon species and for the extinct
N. paroperarius from the Cudahy fauna in Kansas (Hibbard, 1944: Paul-
son, 1961).

Grade III includes all those voles with an M, containing at least
five well developed triangles, of which all may be closed. This grade
includes the genera Pitymys (including Pedomys and Herpetomys) , Microtus
(including Aulacomys , Chionomys , Stenocranius , Chilotus, and Lasiopodomys) ,
Alticola, Anteliomys (including Eothenomys) , and Neofiber. This grade
is first seen in lower Pleistocene deposits of Europe (Kurten, 1968), but
does not appear in the New World until the early middle Pleistocene
(the Cudahy fauna of Kansas; Paulson, 1961). These early North

-86-

American records include only the genus Pitymys (sympatric with Neodon
paropcrarius) ; Neofiber appears in North America in Port Kennedy Cave
time (Irvingtonian; probably somewhat later than the Cudahy fauna) and
Micro tus is not found in North America prior to the Rancholabrean
(Microtus speothen of the Port Kennedy Cave site is either a synonym of
Neodon paroperarius or of the living Microtus oeconomus, the latter
species which is at the Neodon grade of M, evolution, and perhaps more
properly belongs in the latter genus). The genera Alticola and Anteliomys
are now restricted to the Old World, and I am not aware of any extinct
species allocated to these genera. I have included the genus Eothenomys
within the genus Anteliomys because I do not believe that confluency of
all triangles evolved more than once. The following is a key to the
genera of this grade:

1. Triangles 1 & 2 of Mi confluent Anteliomys

Triangles 1 & 2 of M-^ not confluent 2

2. Triangles 4 & 5 of M1 confluent Pitymys

Triangles 4 & 5 of M-^ not confluent 3

3. Triangle 6 of M^ directed posteriorly; triangle 7 not

usually developed on M-, Alticola

Triangle 6 of M^ directed mesially; triangle 7 of M^

at least incipiently developed 4

4. -^.Capsular process of mandible undeveloped in adults;
enamel borders of reentrant angles extremely

thick; size large Neofiber

Capsular process of mandible usually well developed
in adults; enamel borders of reentrant angles
relatively thin; size small Microtus

■87-

Th e following discussion and species treatment ace limited to the
genus Pitymys.

Pitymys fossils are known from early Pleistocene deposits of
Europe, but are recorded in North America for the first time in the
early middle Pleistocene Cudahy fauna (Paulson, 1961). Pitymys meadensis
may be the ancestor for most later North American species of the same
genus, and based upon published information it is inseparable from the
Old World P. gregaloides of early Pleistocene age (Hinton, 1926, re-
ferred this species to the Upper Pliocene, but the deposits from which
this species is known, the upper freshwater beds at West Runton, have
since been referred to the earliest Pleistocene; Zeuner, 1959). Both
of these species are characterized by the closed anterior loop of M^^
and closed third and fourth triangles on the M2. The living species
P. oaxacensis and P. guatemalensis also demonstrate these features,
and thus are relictual species of this earliest Pitymys radiation in
North America.

Pitymys llanensis is also recorded from the Cudahy fauna of Kansas,
and is separable from P. meadensis in that the anterior loop of ^ is
open and triangles three and four of the M2 are confluent. The shallow
sixth reentrant angle and mesially directed anterior enamel border of
the fourth triangle of the ^ indicate affinities with the living
P. ochrogaster, and I believe that P. llanensis is the ancestor of the
former species.

Pitymys involutus and P. dideltus of the Port Kennedy and
Cumberland Cave faunas are inadequately described and illustrated in
published accounts, and I have not studied these forms in detail. Both

of these faunas are most likely of Irbingtonian age and somewhat younger
than the Cudahy fauna (Hibbard, 1958).

The extinct Pitymys arata and the living P. quasiater are related
to P. meadensis in possessing a closed anterior loop on the M. , but are
more advanced in that the third and fourth triangles of the M2 are con-
fluent. The sixth reentrant angle of the M-^ in both these species is
deep and may curve up into the anterior loop. The anterior border of
the fourth triangle is usually sloped posteriorly as well. Both char-
acteristics indicate alliance with the line leading to Pitymys pinetorum.

Pitymys pinetorum individuals I have studied usually demonstrate
an open anterior loop on the M, (it is closed in a few P. p. nemoralis,
but I have not seen it closed in any other living P. pinetorum subspecies)
and triangles three and four of the M2 are open in all specimens I have
seen. These characteristics are first seen in a member of the P. pinetorum
line in the early Rancholabrean species from the Willis ton III deposit
of Florida, P. hibbardi (Holman, 1959). This species differs from P_.
pinetorum in its larger size and reduced capsular process. The cap-
sular process is well developed in adult P. pinetorum. Pitymys hibbardi
may have been derived from the Coleman IIA P. arata.

Neofiber alleni
Material: UT 11785-11789; 1 femur, 1 humerus, isolated teeth.
Remarks: The fossil material is inseparable from Recent material with
which it was compared.

Figure 9 . — The dentition of Pitymys arata and Equus sp,
from the Coleman IIA fauna, and femora of
Lepus alleni and Lepus townsendii.

-90-

.91-

O tH

s

co cu

6 J2

0 JJ

Tl

I-l

a

H

h

Cfl

CTi

.— 1

m

3

4J

^H

•H

•r-l

Tl

CO

c

CO

s "h

i

Cfl

QJ

a . .

33

o

fei

■u

cfl

>>

<U

T)

TJ

C

rH

3

•H

■r-l

jj

a

3

CO

o

18 I

-92-

a

00 06
C •
d) O

<

B

n

«

sO

c^

T3

r-(

M

cd

,o

-a

.Q

M

a

03
,0

,0

h

t^

01

1-1

^-i

•H

C/l

B

B

•u

(1)

a)

En

G

^

cu

I

S 3

> IH

0 w

b._m_

-a

3

O EC

15 $

15
> ec

-93-

Family Erethizontidae

Erethizon dorsatum (Linnaeus)- porcupine
Material: UF 11774, 11776: partial skull with all cheek teeth and a
fragmentary mandible. (Figure 10)

Remarks: John White (1968) has recently described a new species of
Coendu from middle Pleistocene deposits in California. In the same
paper he referred two other fossil porcupines originally described as
Erethizon to Coendu. In doing so, White has implied that the earliest
records of Erethizon in North America are late Pleistocene, and that
perhaps the evolutionary divergence of the two genera is that of age,
or at the very earliest middle Pleistocene. The Coleman IIA porcupine,
relegated to the late middle Pleistocene, therefore takes on added im-
portance, and a study of the skull from this deposit was undertaken.
The Coleman specimen lacks the nasals and the entire braincase.

Palates, upper cheek teeth, and mandibles of Coendu adults may
be separated from those of Erethizon adults by the following criteria:

Coendu

1. P^ equal or slightly
wider than M .

2. Upper tooth rows subparallel.

3. Posterior border of palate
at midline located opposite
center or posterior border

of M3.

Erethizon

1. P^ markedly wider than M .

2. Upper tooth rows widely
divergent.

3. Posterior border of palate
at midline located opposite
center or posterior border
of M2.

■94-

Coendu (cont'd) Erethizon (cont'd)

4. Angular process of mandible 4. Angular process of mandible
not sharply inflected; not sharply inflected; flattened
particularly flattened on on ventral surface,
ventral surface.

5. Upper cheek tooth row less 5. Upper cheek tooth row greater
than 1/4 of skull length. than 1/4 of skull length.

Criteria 1), 2), and 4) are from White (1968). Criterion 3) is the re-
sult of my observations on a small sample of Erethizon (6 individuals)
and Coendu (3 individuals). Criterion 5) is taken from the following
data made available to me by S. David Webb. Measurements were taken
on specimens at the American Museum of Natural History:
Coendu : 19/88, 21/92, 20/87, 19/82, 20/89
Erethizon: 23/83, 25/90, 25/92, 25/88, 25/98

The measurement to the left of the line is the length of the cheek
tooth row; to the right of the line is the greatest length of the
skull.

Criteria 4) and 5) are not applicable to the Coleman II porcupine
skull, but in all other criteria this skull best approximates Erethizon

dorsatum. Webb excavated the P from beneath the functional premolar,

4 1 4

the DP , and the former is much wider than the M (width P = 7.2 mm,

width M1 = 6.3 mm). In divergence of the upper tooth rows the Coleman

porcupine measurements fall within the variation expressed by Erethizon

dorsatum as presented by White (1968): width of palate between P4

alveoli =2.8 mm; width of palate between M3 alveoli =8.6 mm.

-95-

3
A mandible and partial palate (that part housing the M s is miss-
ing) of a porcupine from the Kansan Inglis IA deposit of Florida is re-
ferrable also to Erethizon, as the angular process of the mandible is
clearly inflected and ventrally flattened and the P is wider than the
M1 (P4 width - 7.0 mm, M width = 6.3).

Although White (1968) may be correct in his assignment of the
Grand View (Wilson, 1935), Aguascalientes (Hibbard and Mooser, 1963),
and Vallecito (White, 1968) porcupines to Coendu, the data presented by
him are far from convincing. The approximation of the M width by the
P^ width in the Vallecito porcupine is certainly suggestive of Coendu,
but as White (1968) demonstrates, the California fossils are otherwise
inseparable from Erethizon, cranially and postcranially . I have not
found either the scratches on the occlusal surface of the teeth, the con-
figuration of the masseter scars, or the projection of the longitudinal
axis of the lower tooth row onto the incisor as described by White
(1968) reliable features with small samples. However, if the longitud-
inal axis of the lower tooth row projects far laterad to the incisor,
the individual may be referred to Erethizon.

The presence of Erethizon in Florida during the Irvingtonian
further indicates my reasons for doubting White's (1968) taxonomic
references, and suggests that the evolutionary dichotomy between
Erethizon and Coendu occurred prior to Irvingtonian time.

Figure 10 -- Upper dentition of fossil porcupine, UF 11774,
from Coleman IIA.

-97-

■98-

Family Hydrochoeridae
+Hydrochoerus sp. Brunnich- capybara
Remarks: The capybara material from Coleman IIA is being studied by
John Lance at the U. S. National Museum. Correspondence with him, and
personal observation of some of the material suggests that the capybara
represented is Hydrochoerus (rather than either Neochoerus or Hydro-
choeropsis) ; but it may be a new species.

Capybara remains are common in river deposits in Florida, includ-
ing the Ichetucknee River (Simpson, 1930a), Santa Fe River (unpubl.),
Waccasassa River (unpubl.), Withlacoochee River (unpubl.), and Oklawaha
River (unpubl.). These materials usually include either Hydrochoerus
or Neochoerus or both genera. Capybaras have also been reported from
terrestrial deposits at Vero (Weigel, 1962), Seminole Field (Simpson,
1929, 1930a), Sabertooth Cave (Simpson, 1928), and Bradenton Field
(Simpson, 1930a), and West Palm Beach (unpubl.).

Order Carnivora

Family Ursidae
+Arctodus pristinus Leidy- eastern short-faced

bear
Material: UF 12363; right M1.

Remarks: The only specimen, an upper right first molar, is the first
record of Arctodus pristinus from Florida. Size of the molar (length
23.9 mm, width 20.4 mm) clearly indicates A. pristinus rather than
A. simus (Kurten, 1967). Although there is some overlap in length of
the M between the two species Kurten found no overlap in measurements
of the width of this tooth (A. simus : 22.3-27.3 mm; A. pristinus: 20.2-

■99-

22.2 mm). Previously this species had been reported from deposits of
Irvingtonian age in eastern North America (Port Kennedy Cave, Cumberland
Cave) and from deposits of heterogeneous origin in the Ashley River,
North Carolina (Kurten, 1967).

Family Mustelidae

Spilogale putorius (Linnaeus)- spotted skunk
Material: UF 13168; isolated teeth and mandibular fragments.
Remarks: The fossil material is identical to that of Recent Spilogale
putorius from Florida.

Conepatus sp. Gray- hog-nosed skunk
Material: UF 13169, 13170; 2 right mandibles, one with P., the other

with P2-p4-

Remarks: The crowded condition of the premolars, plus the lingually
directed, enlarged posterior heel of the P^ identify the fossils as
Conepatus . Conepatus leuconotus remains have been reported from the
Haile VILA and Williston III lime pits of Florida (Ray, et_al . , 1963;
Churcher and Van Zyll de Jong, 1962), but the measurements of the one
complete mandible from Coleman IIA do not allow an identification to
species (Table 10). Conepatus leuconotus is also known from the
Pleistocene Ladds deposit in Bartow, Georgia (Ray, 1967).

Mephitis mephitis (Schreber)- striped skunk
Material: UF 13167; right and left mandibular fragments.
Remarks : Mephitis and Conepatus are sympatric now only in the arid to
semiarid southwest and northern Mexico. An unusual feature of the
Florida Rancholabrean faunas is the abundance of skunks (in numbers of

-100-

Table 10

Measurements (in mm) of skunk mandibles
N = number of specimens,
mean above and observed range (in parentheses) below,
1 = alveolar length ?2~^-\> 2 = depth of ramus at middle of M-^,
3 = breadth of ramus at middle of Mi

C. mesoleucas 17.4 7.1 4.7

(15.9-19.9) (6.0-8.8) (3.9-5.3)

N=8 N=7 N=7

C. leuconotus 20.1 8.1 5.4

(19.2-20.9) (6.7-10.3) (5.0-5.8)

N=9 N=9 N=9

Coleman IIA 19.0 8.0 5.0

individuals fossilized) and almost total absence of weasels. This may
be somewhat related to weasel habits; they are taken only infrequently
by raptorial birds and possibly were able to escape natural-trap sink-
holes. A fossil weasel (Mustela frenata) has been reported from only
one deposit in Florida; the Wisconsin Seminole Field locality (Simpson,
1929) . Although weasel habits may play a part in the scarcity of their
remains, it is further conceivable that weasels were Wisconsin immi-
grants into Florida.

Family Procyonidae

Procyon cf P. lotor (Linnaeus)- Raccoon
Material: UF 13163, 13164, 13171-13173; mandible, partial palate,
humerus , RP, .

■101-

Remarks: The Coleman IIA raccoon is tentatively identified as P. lotor
following Arata and Hutchison (1964) and comparison with Recent Florida
material. Study of the Recent material convinces me that Procyon lotor
is one of the most variable carnivores, and perhaps mammals in general,
in North America. One well preserved fossil mandible, with all teeth
save the canine and carnassial, was duplicated by only very young
Recent individuals, in both size and dental morphology. A partial upper
palate, with the M to M^ is within the size range of Recent specimens
of young individuals, but the fossil is clearly from an old individual,
as the teeth are heavily worn and cusp patterns have all but been oblit-
erated. This suggests that the Coleman IIA raccoon may have been
smaller than that subspecies which presently inhabits the Coleman
area. Yet the fossil sample is so small, and the Recent species so
variable, that the material must simply be referred to Procyon lotor.

Family Canidae
■fUrocyon minicephalus , new species
Holotype: UF 13146; a skull, complete except for broken zygoma, nasals,
and premaxillaries . The dentition lacks the canines and incisors.
Referred specimens: UF 13137-13145, 13147-13151; all cranial and
postcranial elements.

Horizon and locality: Coleman IIA Local Fauna, Sumter County, Fla. ;
late Irvingtonian; ?Illinoian.

Diagnosis: Urocyon minicephalus is a fox similar to the living gray
fox, differing in the relatively closely allied sagittal crests and
narrow occiput (Table 11; Figure 11). The greatest width between
these crests (measured from the outside of each) in three fossil

• 102-

skulls does not approximate that in living or other Pleistocene gray
foxes of comparable developmental age (as judged by tooth wear and skull
suture patterns). In this respect the Coleman Urocyon are similar to
the living Vulpes fulva, the red fox. The sagittal crests are widely
spaced in the California island gray fox, U. littoralis, and in the
Blancan species from the Rexroad fauna, U. progressus (Stevens, 1965
and personal observation).

Remarks: The relationships and origin of this species are unknown, and
it is conceivable that U. minicephalus was a Florida endemic.

Canis lupus Linnapus- gray wolf
Material: UF 11518-11520, 12113-12126; 4 mandibles, 2 partial skulls,
3 radii, 1 scapula, 6 astraguli, 2 innominates , 6 femora, 4 tibia, 5
humeri, 6 ulnae, 3 phalanges, 8 metapodials.

Remarks: North American Pleistocene canids, especially the wolves, are
badly in need of revision. At present five nominal species of wolves
are recognized in the Pleistocene: Canis dirus Leidy, Canis ayersi
Sellars, Canis milleri Merriam, Canis armbrusteri Gidley, and the liv-
ing Canis lupus Linnaeus (= C. occidentalis of Merriam, 1912).

The most abundant collections of Pleistocene wolves from a
single locality are those from the Rancho La Brea tar pits. Merriam
(1912) recognized three species from these pits: Canis dirus, Canis
milleri , and Canis lupus (= C. occidentalis). This is significant, as
it is one of three fossil assemblages in which both living and extinct
wolf species have been found together (the others include Fossil Lake;
Elftman, 1931, and McKittrick; Schultz, 1938) . As these pits are still
trapping animals (Stock, 1930) heterochrony may be suspected.

ft

>H

p

c

cfl

03

3

H

y

OJ

m

CO

CO

3

3

3

i— (

■-I

,— 1

m

cfl

C8

XI

42

42

a

D.

&

CO

(1)

CU

y

O

O

•r-l

•W

•H

c

C

c

■H

1-1

•l-l

0

H

0

p

£>

£>"

-104-

•105-

Table 11

Measurements (in mm) of fossil and Recent fox skulls
N = number of specimens, mean above and observed range (in
parentheses) below

Great w across
sag. crest

Great w brain-
case at zygema

Great 1 from inion
to horiz. line
betw. po processes

Coleman II

13.5
(11.2-15.1)
N = 3

42.0
(41.6-42.5)

N - 3

56.1
(55.0-57.5)

N = 3

Recent U. cinereo-

argenteus- Florida

22.4
(17.4-28.1)
N = 19

44.3

(42.8-47.0)

N = 7

58.0
(55.2-60.4)

N = 7

Great w across

Great w brain-
case at zygoma

Great 1 from inion
to horiz. line
betw. po processes

Arredondo IIA

26.4
N = 1
44.9
N = 1
62.3
N = 1

Recent U. cinereo-

argenteus- Fla. , Ga.,Mass,
111., Ala., Ariz., Ark.,

24.0
(15.8-36.8)

N = 58

Great w across
sag. crest

Vulpes fulva- N.Y. , Penn.,
Md., Del., Va., Alaska

11.4
(3.0-16.9)
N = 39

•106-

Nevertheless, the fact that Merriam could recognize three wolf species
will be important to later considerations.

As shown in Tables 12 and 13, and Figure 14, Canis dirus includes
some individuals larger than the largest living Canis lupus . These data
also show that, even in the southeastern states, the extinct dire wolf
averaged larger than lupus .

After studying over 1,000 Recent wolf skulls at the U. S. National
Museum, I am convinced that Canis dirus is distinct from C . lupus . Two
qualitative characters clearly separate these species. First, as noted
by Merriam (1912), the well developed inion in C. dirus projects back-
wards and downwards to a greater extent than in lupus . Although there
is some variability within each species, and between males and females,
even the largest male Yukon lupus do not demonstrate the pronounced
overhang that is characteristic of all dire wolves. The Coleman IIA
Canis agrees with lupus in this respect , (Figure 12).

Second, the configuration of the anterior cingulum of the first
upper molar appears to be diagnostic. In Canis dirus the anterolingual
cingulum does not reach the hypocone, but ends at the protocone. If
samples of living wolves that approximate the average size of Canis
dirus are selected (Canis lupus pambas ileus and C. 1. occidentalis) , in
only 4 out of 75 the cingulum fails to reach the hypocone. The vast
preponderance exhibit a cingulum that joins the hypocone. If Canis
lupus subspecies are studied at random, without regard to size, 20%
demonstrate an incomplete anterior cingulum. This indicates that,
although there is a significant increase in the dirus trait in lupus
from areas other than the Yukon and northern Canada, this increase is

■107-

present only in wolves smaller than the extinct dire wolf.

The oldest true wolves in North America are not as large as
dims, and appear to be good lupus (those from Cumberland Cave, Inglis
IA, and this fauna). In Florida Canis dirus appears during the middle
or late Sangamon and lasts until the latest Wisconsin. Although Savage
(1951) recorded Canis dirus from the Irvington fauna, the materials on
which he based his identification (a mandibular fragment with a broken
carnassial and a heavily worn P^) do not permit a positive species
reference. Indeed, Savage (1951) noted that a Canis femoral shaft
from the Irvington, also assigned to dirus, appeared shorter and stockier
than those of Rancho La Brea dirus to which it was compared. This sug-
gests that the Irvington wolf is lupus , as dirus is characterized
partly by "light" limbs (Merriam, 1912).

According to Sellards (1916) Canis ayersi from Vero, Florida, has
a narrower snout than C. dirus, but is otherwise identical to the
Rancho La Brea dire wolf. Another dire wolf skull, UF 2923, has since
been collected from the Reddick IC Rancholabrean deposit in Marion
County, Florida, and measurements of this animal are compared to those
of the Vero and Rancho La Brea wolves in Table 13. These data show
that the Reddick IC Canis is clearly dirus, and as the Vero skull is
well within the size range of Rancho La Brea dirus it seems reason-
able to consider it and other dire wolf material from Florida Rancho-
labrean deposits as Canis dirus.

Canis armbrusteri from the Cumberland Cave fauna presents an
enigmatic taxonomic situation. First, the skulls referred to
armbrusteri range in size from that typical of lupus through that

■108-

typical of dirus , the average approximating that of the large northern
lupus subspecies (Table 13). One skull, USNM 7994, is essentially
identical to the Coleman HA skull UF 11519. In general, the Cumber-
land Cave wolves approximate large northern lupus for all qualitative
characters (all traits mentioned as unique by Gidley and Gazin, 1938,
can be reproduced in these lupus ) . Yet there is one skull from Cumber-
land Cave (USNM 1186) in which the inion is as pronounced and hooked
as in dirus . All but one of the first upper molars in the Cumberland
Cave wolves have a complete anterior cingulum as in lupus .

The large size of the Cumberland Cave wolves can be explained by
deposition during a glacial period (or periods); either Illinoian or
Kansan, or both. It is evident from published studies (Dalquest, 1965;
Guilday, et_al . , 1964, 1966; Martin, 1968a) that mammals now confined
to northerly regions existed in more southerly regions during times of
glacial advances. The largest Canis lupus now reside in north Canadian
and Alaskan woods. During the Kansan and/or Illinoian glacial maxima
it is conceivable that wolves of this size would have existed farther
south, quite probably as far south as the Cumberland Cave area in
Maryland. Therefore, with the exception of skull USNM 1186, it seems
reasonable to refer the wolf skulls from the Cumberland Cave deposit
to Canis lupus . The unusual skull (USNM 1186) with the pronounced
inion may represent Canis dirus, but it seems more probable to me that
it too is lupus , perhaps evidencing the first trace of dirus features.
I suggest that dirus is derived from lupus ; perhaps from the Cumber-
land Cave lupus .

The only wolf species remaining is Canis milleri (Merriam, 1912).

• 109-

This species is represented only by the type skull, UCVP 11257. Sup-
posedly this species differs from C . lupus only in greater width of the
skull and more massive dentition. My measurements (Tables 12, 13) show
no significant difference between this type and Canis lupus . As shown
by Merriam (1912; figure 31), the anterior cingulum of the M connects
with the hypocone.

In summary, I conclude that there are only two wolves common to
middle and late Pleistocene deposits of North America, Canis dirus
Leidy (=Canis ayersi Sellards) and Canis lupus Linnaeus (=Canis dirus
of Savage, 1951, =Canis armbrusteri Gidley, =Canis milleri Merriam,
=Canis occidnetalis of Merriam, 1912). The typical Irvingtonian wolf
is Canis lupus . The common Rancholabrean wolf is Canis dirus. After
the latest Wisconsin extinctions only lupus remains.

Family Felidae

Felis onca Linnaeus- jaguar
+Felis one a august a Leidy
Material: UF 12128-12165; 9 partial dentaries , 12 partial maxillae,
1 brain case, numerous skull pieces, 10 calcanea, 7 astraguli, numer-
ous metapodials, 5 humeri, 63 phalanges, 6 femora, 1 sacrum, 6 ulnae,
4 innominates , 7 tibia, 7 radii, 4 scapulae, 1 fibula.
Remarks: Numerous large felines have been reported from Pleistocene
deposits in North America (Simpson, 1941; Kurten, 1965) and to present
a complete summary of these forms is now quite unfeasible. The large
cat from Coleman IIA is clearly a jaguar, and is referrable to the
extinct subspecies Felis onca augusta following Simpson (1941) and

3 g

-Ill-

Figure 13 -- Ventral view of skull of Canis lupus
from Coleman IIA (UF 11519).

-113-

c

4-1

3

a)

a

u

4-1

3

aj

co

i-l

pel

0

•r-l

Tt

cfl

o

C

bO

CB

«

<

(JO

CD

CO

CO

CO

c

•r-l

ai

3

a)

•r-l

X

j-t

&

i— 1

4-1

ctj

CTJ

3

W>

Cfl

3

i-i

C

rH

X

cr

a)

(1)

^^

co

•H

h

U

u

<t

T3

4-1

H

CM

•r-l

cu

cfl

^

>

c

4-1

X

CU

hi)

CO

rU

3

b

CD

(H

3

rH

^

u

CO

•i-l

3

a

&

T— 1

o

•H

•r-l

M

r/1

>!

a;

en

(U

!-i

rd

i—l

M

cl

•H
Cfl

U

a

rQ

•i-l

rH

1-1

c

u

O

o

Cfl

3

LM

03

3

3

h

&

3

i—l

3

MH

o

o

LO

<t

r-l

X

MM

O

CO

3

ctj

cfl

3

cu

■v

H

•r-l

>

•1-1

CO

T3

3

a

<

CN

O

H

X

>,

M

>

n

co

<-<

(11

rH

rH

cfl

•r-f

3

U

Cfl

T3

IS

•H
>

ii

II

•H

en

T3

X

M

3

ai

•H

4J

CO

3

r^

rH

a>

rH

Cfl

C)

3

m

Tl

c

o

•H

•H

>

d

•H

4J

cfl

TJ

O

CU

3

-115-

CV3

4

1>

o

0 #

rC3)

X D O

o • o »
o o o

o o *

+

o

+

X

i i— — i — i 1 r"

I0** C£3 lf3 «c3* CO C-^J

n r-

CO

€73
CM

€33

CO ^

CM

CO
CM

CM

CM

HiaiM

■116-

M

«

<u

ti

TJ

a

C/l

CU

Cu

h

3

II

3
01

T)

cd

c

C

(U

cd

•H

B

b

H

C

cd

«

cd

H

.e

b

O

4-1

•H

H

OT

X)

CJ

CD

CU

•H

CD

M

>

s

ft

a. i-i

en

0

J3

IS

4J

u

i*

in

n

4-1

01

3

a

•*

w

CM

O

CD

x:

M

H

M-4

t)

4J

m

(1)

on

■— i

01

w

{*:

C

.— 1

a;

QJ

II

X)

an3

rH

CD

a

c

H

3

cs

II

X

«H J g

m

ca

T1

0

e

aj

CM

CD

w

b

3

C

•r4

en

i-l

0

cd

(U

a)

u

a

b

a

1-4

CO

C

o

<+4

CD

0

o

b

•H

M

O

■".)

Cl)

r"

p,

H

en

3

c

ii

z

vo i—i

co m

1-4 00

vO CM

r-«

cr> m

i—l o\

CM vo
CM CM

i—i m

CM CM

o -3-
CM CM

r- O

■-I CM

o

CM

i-l CO
CM CM

o> o>
1-1 1—1

r-» cm

ro cm

CM CO

O CM

■tf

•tf o

m r*

vD On

vO Ov

vo r^

co m

\a

r-» o\

vO >o

•h cd .h cd
B B SB

B B

B B

a cs

CO l"»

O co

O CM

co oo

CM <f

53 r-l

X)
3

M

•i-l

•r4

u

i-l

CU

cu

4-1

i—i

CO

rH

3

•r4

t-i

y

,0

B

en

H

•H

cd

C

cd

O

u

C X

•r-l Cfl

S B

•tf

k)

l/~l Ol
CM CM

CM CO

CM CM

O CM
CM CM

00

vo O
CM CO

u-l vo
CM CM

•r-i cd
B B

a x

•m cd

B B

a x

•i-l cd
B B

C X

•h cd
B B

C X*

•t-4 cd
B B

c x'

•m cd
B B

i-» m

o o

oo vt

CN CN

0> t— 1

t^. O

00 CN
r-l CM

00 o
.-1 CN

s e

3

in o>

•H CO

a s

..,

ca

-X

i-i

ca

XI

o

•H

X.

m

u

o

s

r-t

cd

fe

I*

■117-

■ 118-

co

QJ

3

<4-l

4-J

XI

- CO

O

3
0

1-1

CO

cu co
jj a)

-o

4-J

CO B

co

a

4-1

3

6

iH

QJ

a

o

•r-l C

i-l

0

X QJ

3vf P>J

u

c

4-1

o E

^

pn

P-,

o

CO

3

rH -H

co

•H

O

a. o

M-l

MH

)-l

4-1

C

CX 0)

m

O

O

QJ

o

CO

co a

i-H

4-1

CO

0

CO

co

CO

x

M-l

II

2

3

3

O

4J

O

4-1

i-H

i—i

0-xl

Cu CO

■u

o

o

•H

X

tO QJ

C

QJ

QJ

E

1*

4-1

00

cu

>

>

o

X)

- U

u

rH

1"H

i-l

II

■H

C CO

QJ

tfl

CO

M-l

3

cO i-l

«

m

X

rJ

u

QJ

II

W II

X)

o

o

u

#1

C

•r-l

•iH

c

CO

[~-

M •

CO

U

Sh

CO

QJ

QJ X

QJ

QJ

4J

i-l

4J CO

Cn i-H

u

4-1

CO

^

rH

co B

i-l -i-l

CO

co

•H

-a

i—l

CU

CO

O

o

X)

c

3

U i

QJ co

a

Cl

o

^

00 X>

i-H O

II

u

CO

CU

X M— i

4-J

4-1

II u

cO

cO

CO

<f

1— 1

4-1

3

H MH

CO

0

CO

O

X

x

•*

4-1

co

4J

4-1 CN

■H

X

QJ

r— >

X!

XI

£

D.

4J

• B

CO

CO

•,H

60X

QJ

OJc-0

0

(3

w C

Sh

Sh

Oh

o

cu

*0 <U

c

X

X

o

i-H

co S

•r-l

X

cu -i-i

4J

4J

4J

IH

4-1

rH O

CO

CO

60

o

CO

X (U

CO

QJ

QJ

c

QJ

a.

4-J

4J

4->

QJ

Sh

4J

CJ CO

a

cd

CO

i—l

cu

CO

•H

QJ

CU

cu

4J

QJ

4-J 4-1

E

u

u

U

c

u

CO CO

01

60

oo

CO

QJ

60

B <U

u

i-l

u

O 1-4

3

II

ii

0

II

60 rH

CO

QJ

o

>> CO

cfl

r-l

CN

>

4-1

^D

N B

CU

i-4

CO

£

CO

CU
4-1

QJ

II
II

II

cd
1—1

c

•r-l

00

m

CO
Cu

CO

o

•H

E

i-l co

c

o

c

1—1

1-1

c

CO

^

M-l CO

n

^_^

r-i cO

60

X X

C 00

E

[V,

S-i 00

a

XI X)

cO C

CO

X3

o c

^^

3

3 3

QJ 3

«H

3

M-l 3

>^

O

4-1 4-1

B O

u

4-1

O

X)

X

CO CO

>>-^ >h

u

QJ

CO

§*

2

E

CO CO

CO

§

s

CO

QJ b

CO

O

•H •<->

3

•H

B O

u

Xl X!

QJ

M

g

X

»—* in

CO

M-l ^-s

4-J 4J

i—l

M-l

4-1

MH

•iH

<t-

N— ' <W

•r-l

s*

M

S— r

•H

<f

X

• •> v^5

co

\0

M-l

>

vO

4-1

CO CTi

CN O

CO

CO

CTi

••—^

^D

QJ

CO

C5>

i— I i—l

00 00

X

i—l

■—1

Is.

i-H

i-H

rH

CO

-* <t

E

CO

«tf

LO

•W

CO

m

3 •>

<r <r

cO

3

00

m

CO

3

CN

•o a

i-i i-i

a

X)

C

C3>

V0

X

X)

C

ro

•i-l cO

CN CN

•H

CO

CN

•r4

CO

in

> B

>

E

N

>

E

T3 i-l

1! H

i-i

•r-l
X)

i—l

%

g

rH

•H

xs

XJ

i— i

§

C O

CO CO

C

o

U

CO

C

0

CO

•H O

1-3 &

a

•H

o

a

»

u

•H

a

r3

o oo

a b

a

cd
>,

3

-119-

cm m

ON o

•l-l

•H

u

S-l

<u

01

4J

i-H

CO

i-l

3

•r-l

!-!

H

XI

s

t/i

Pi

•r-4

cs

C

cd

u

o

01

to

3

3

u

K

3

•rH

i-l

T3

CJ

CJ

d)

vO

^

SO

>SVO

0

O i-4

ca i— i

•H

XI ON

r»»

T)

O iH

•

T-S

c

O w

(1)

cd «

II O

p4

PJ CO

<f vO

N-'fH

v— '

WO

en in

H

co oo

en

CO

3

3

M

^

H

•i-l

T3

TJ

U

CJ

• 120-

Kurten (1965). Some dental measurements of the Coleman IIA jaguar are
presented with those of other fossil and living jaguar samples in
Table 15.

Large cats such as Felis atrox and Felis onca augusta were common

predators in the Rancholabrean fauna of North America, yet their taxo-

s

nomic status remains somewhat uncertain. Both Simpson (1941) and Kurten

(1965) related Felix atrox to the living jaguar, but Kurten (1965)
states, "Although Felis atrox may be related to the jaguar . . . , it
must have looked very different in the flesh. With its slim build,
long legs, and relatively small head it was obviously highly cursorial."
The largest living jaguars, F. onca milleri and F. onca palustris are
found now only in South America, and yet are apparently the only
races of the living jaguar which approximate F. o. augusta in size
(Simpson, 1941).

These considerations occasioned a study of morphometric variation
in the lower dentition of most of the New World fossil and living Felis
(excluding Lynx) and the lion, tiger, and leopard of the Old World.
One would perhaps expect that the proportions of the teeth would vary
significantly in such a widespread and externally variable group of
mammals. The measurements of these forms are presented in Table 14.
As the most useful graphical method for noting proportional differ-
ences is the ratio diagram, the measurements were portrayed by this
method in Figure 15.

Surprisingly, there is virtually no difference in the form of
these diagrams between the species, although some size differences
are obvious. The only possible exception to the above statement is

• 121-

seen in the plot for the puma, Felis concolor. I am not certain of the
significance of this dichotomy; if it were to be a fundamental differ-
ence between the Panthera and Felis groups, it might be expected that
the ratio diagram plots for the margay and ocelot would approximate
that of the puma as well, but they do not. Meade (1945) and Savage
(1955) also compared members of the Felis and Panthera groups by the
ratio diagram method and came to the conclusion, with regard to dental
dimensions, that the resultant plots simply do not demonstrate any
major differences between the two groups. Simpson (1941) did not pre-
sent any ratio diagrams comparing only dental dimensions, but rather
combined dental with postcranial dimensions or dealt solely with limb
material. His plots may demonstrate quite succinct differences be-
tween the puma and Recent jaguars, but these differences depend upon
the bones compared (e. g., plotted dimensions of metatarsal II will
separate pumas from jaguars, but the diagram for dimensions of metacarpal
IV will not separate these forms except on size).

These data suggest that although the pelage and limbs of the cats
studied have undergone selective modification, the dentition has re-
mained conservative.

Felis rufus Schreber- bobcat
Material: UF 14996; 1 right humerus and 1 right fibula.
Remarks: The fossil material recovered, although conceivably the re-
mains of a small cat other than the bobcat, was inseparable from the
same elements of Recent F. rufus with which it was compared.

Figure 15 — Ratio diagram comparing the lower dentitions
of various felines. A = Felis wiedii, B =
F. pardalis, C = F. concolor, D = F. pardus ,
E = F. onca milleri, F = F. tigris , G =
Coleman HA F. onca augusta, H = F. leo,
I = Ichetucknee River F. atrox. Numbers of
specimens and sources of measurement data
are given in Table 15.

-123-

0.5 0.4 0.3 0.2 0.1 -0+ 0.1 0.2 0.3

I I I.I J_J£ ■ I

r \ J \ ft '

\ \ v W

\

>

< <

I l
I I

1

() O (I

/
/

/I

\ \ [\

\\ \
»

u i

,7 I /

it / f

;» / I

) I I

*\ \

w

w

w

CD

\
\ \ \
V\ \ \

'3

wP3

IP,

wP4

IM,

wM,

■124-

4-1 T3

CJ CD

a >

T>N

0) CO

r-l B

C B

►J IX

4-1

12 IX

►J IX

•A

rH

CTi

u

<t

4-1

CI)

o>

CO

I— 1

r-l

a

r-l

bl

■H

3

H

a

o

cfl

ca

CO

cd

u

H,

CJ

c

b

C

0

•H
CO

o

CO

CO

•H

h

T-I

rH

n

rH

ci)

n

CO

It.

M-l

+

a) m
o cy.

cd

cd

x

id

•H

C

a)

4-1

13

0

S

CO

6

N

0

M

■H

X3

CJ

H

e

X)

<

Cd

X3

c

a

cd

r^

3

0

■r-l

cd

X! .-s

N

0

0

BJ >,

cd

o

•H

XI X)

t-i

•r-l

h

M 3

PQ

X ^

m

N-/ JJ

v-^

<U ;V>

<

CO

a xi

^^

M

CO

O 3

o

CO

•H

4J

to

r-l

■r-l

r-l

•r-l

CO

3

^— s

O

X!

cd

S— s

•r-l

T)

>1

u

4-1

X)

>1

XI

CO

rJ

X)

c

r-l

XI

QJ

•r-l

cd

n

0

crt

3

•r-l

rC

u.

■u
c/1

o

cd

c

p.

4-1
CO

3

4J

CO

CO

m

CO

CO

•H

to

•r-l

4-1

t4

CO

■H

cd

r-l

•H

r-l

c

r-l

■r-l

r-l

C)

a)

.C

CI)

o

01

rC

OJ

•r-l

it.

4-)

u.

S

1*4

4-1

rH

0£

■ 125-

P6

o
.-I

£

■u

•l-l
3 IX

*r°

iJ IX CN

d

-3
i-t
3 IX

cu m

M r-1

a

rt

X

03

i-i

c

a

4-J

X)

0

H

CD

(2

N

O

H

l-l

X)

a

u

-o

«U

nj

«

p

c

o3

o

i-l

03

03

,13

>~s

N

0

ej

03

>1

o3

U

•H

T3

X)

U

•H

W

M

3

pq

X

<",

CO

N— ^

3

•O

0

S-i

<u

ctf

r-<

a

r/1

CO

■H

•i-l

7-H

i-i

8)

CU

Pn

fll

>>

(1)

T3

hi

3

03

4-1

U

co

0)

>

CO

<

•i-l

*

■ 126-

a

3

M

en

cO

CO

•r-1

m

H

4J

a

^H

OJ

cfl

t)

3

cu

T)

St

■H

>

T>

■H

C

TO

td

C

•r-4

•h

4J

t/i

co

co

<U

O

.

■r-l

X

LO

■u

cfl

iH

•r-l
4-1

H

a)

c

r-l

(1)

,0

•n

-n

cfl

01

H

M

h

5 w

C cfl

0) g

£ S

CO 4J

QJ X)

•H

•H

M

h

ai

cu

r-4

i-i

<-t

.-<

>H

•H

H

a

0

o

f"J

fxn

oo co o o

i-l I— I en r-l

CM <N CN <N

CD

*

>

C

•H

cu

Pi

4-1

M

cu

3

Ph

W

CU

S

4-1

0

C

H

CO

14-1

X-s

,—,

VO

c

ITl

LO

ON

Vu

vO

O

r-l

4-)

C^

o>

(-4

.— 1

t— 1

I

C

fed

soj

C

c

4-1

H

'OJ

NOJ

M

n

■u

4-1

3

M

(-J

(-4

w

4-1

3

3

^ '

w

w

E

O

O

g

I

M

st

o

0

M-l

m

B ■-» '->

<t o

cn en
<n cm

CN CM

&

•a

3

c

4J

o

CO

w

ex

CO

S

CO

CU

a

i

t-i

4J

01

0)

>

JD

CO

C3

C

cfl

T)

a)

nj

£

OJ

J3

-:<

60

-127-

Order Artiodactyla

Family Tayassuidae

Platygonus cumberlandensis Gidley- peccary
Material: UF 12070-12112; 28 ulnae, 8 scapulae, 29 humeri, 14 femora,
3 innominates, 22 tibia, 13 astraguli, 14 calcanea, 114 phalanges , 143
metapodials, 2 sacra, 17 dentaries, 8 palates, 1 partial skull, numerous
isolated teeth and vertebrae.

Remarks: There are presently five species of Platygonus known from the
middle to the late Pleistocene of North America. According to the latest
review of these forms (Slaughter, 1966) there are two species groups.
The first is represented by P. compressus, P. leptorhinus, and P. alemani,
These species are characterized by non-expanded zygoma and are consid-
ered synonymous by Slaughter (1966). The second group is represented
by P. vetus and P. cumberlandensis, and has characteristically expanded
zygoma.

The Florida samples I have measured form a fairly coherent pat-
tern, both in time and in anatomical and morphometric features.
Platygonus remains are present in Blanc an through latest Rancholabrean
deposits in Florida. Measurements of the Florida material and locali-
ties are listed in Table 16. When anterior-posterior length of the
Mi is plotted against transverse length of the same tooth, the Florida
Platygonus samples form what I consider to be three distinct groups
representing three species (Figure 17). The first species, represented
in the Santa Fe IB (late Blancan) and Inglis IA (Irvingtonian; ?Kansan)
deposits, is a large form characterized also by the presence of three
pairs of lower incisors. Zygomatic structure is unknown in this form,

■128-

but as this species includes very large animals, presumably it was ex-
panded. The second species in Florida, somewhat smaller than the Santa
Fe lB-Inglis IA form (Figure 17, Table 16), is characterized by the
presence of only two pairs of lower incisors and expanded zygoma. This
species, found in the Coleman HA and Haile VIIA faunas, is referrable to
Platygonus cumber landensis , and is clearly derived from the Santa Fe-
Inglis species. One mandible (UF 12085) from the Coleman HA fauna
demonstrates two shallow alveoli in the position of the third postero-
lateral pair of incisors of the earlier species. The alveoli in UF 12085
are so narrow and tiny that I rather doubt if this pair of incisors was
present for more than a short period in the individual's development.
Two other mandibular symphyses from Coleman HA do not demonstrate these
vestigial alveoli. As can be seen from Figure 17 the Cumberland Cave
P. cumberlandensis conform to the variation expressed by the Coleman
Platygonus . Although zygomatic morphology is not known for the Haile
VIIA Platygonus , first lower molars from this fauna plotted in Figure
17 coincides more with P. cumberlandensis than with the later, smaller
form, and on this basis I refer the Haile VIIA material to P. cumber-
landensis .

The third species, which will here be referred to as P. compressus,
is characterized by small size (Figure 17, Table 16), non-expanded zygoma,
and two pairs of lower incisors. Large samples of this species are
found in the Illinois Cherokee Cave (Simpson, 1949) and Texas Laubach
Cave (Slaughter, 1966) faunas. Although the zygomatic structure of
the Florida Rancholabrean Platygonus is not known, the variation in
size of the M, and the other teeth (Figure 17, Table 16) from the
Reddick IA, Reddick IIC, Reddick IB, and Devil's Den deposits is

■129-

contained completely within that expressed by the Laubach Cave and
Cherokee Cave samples, and these teeth can therefore be referred with
some confidence to P. compressus.

Platygonus vetus supposedly differs from P. cumberlandensis in
lacking accessory structures (lophs and cuspules) on the molars (Gidley,
1920). Such structures are moderately well developed in the Coleman HA
Platygonus (Figure 16). Although Slaughter (1966) suggests that dental
complexity in these forms may have little taxonomic utility, it is not
necessarily valid to cite variation in a Blancan or Hemphillian species
(P. bicalcaratus) as proof. Besides, the taxonomy of that species too
needs review. As I have not completely analyzed dental variation in
P. vetus or in P. cumberlandensis I will temporarily accept their
integrity, and refer the Coleman HA material to P. cumberlandensis.

+ Mylohyus sp. Cope- peccary
Material: UF 14243; upper first or second molar.

Remarks: Mylohyus is obviously in need of revision. There are now
about eight nominal species of this genus known from both middle and
late Pleistocene deposits, which makes allocation of any Florida mate-
rial somewhat difficult. There is now much more Mylohyus material
available from Florida than that discussed by Lundelius (1960), some
of which has been reported by Semken and Griggs (1965).

The older Pleistocene faunas in Florida, those of Blancan and
Irvingtonian age (pre-Coleman HA) contain only Platygonus . Mylohyus
is first seen in Florida in the Coleman HA fauna, and was possibly
much less abundant during that time than was Platygonus . In all of
the Sangamon and Wisconsin deposits in Florida Mylohyus remains are

■130-

more common than are those of Platygonus , and some do not contain any
Platygonus material. Examples of some Rancholabrean deposits with only
Mylohyus are; Arredondo II (Bader, 1957), Vero (Weigel, 1962), Williston
(Holman, 1959), and Seminole Field (Simpson, 1929). Although it must
be considered as pure speculation, these considerations suggest that
Mylohyus was the common peccary in Florida subsequent to the Illinoian
glaciation. However, Platygonus was sympatric with Mylohyus during the
Sangamon at Reddick IA, Reddick IIC, and Haile VIIIA. The latest occur-
rence of peccaries in Florida is during Devil's Den time, from which
deposit there is Platygonus , but no Mylohyus . Again, in all the above
localities, excepting Devil's Den, Mylohyus remains are much more com-
mon than are those of Platygonus.

According to Hershkovitz (1962) and Hooper (1957) populations
of cricetid rodents that inhabit arid situations have less compli-
cated teeth than do those living in more humid situations. Teeth of
Platygonus are much less complicated and more tubercularly hypsodont
(terminology after Hershkovitz, 1962) than are those of Mylohyus.
Thus, if dental adaptive trends of peccaries correspond at all to those
of rodents, and particularly to the cricetid rodents, then the peccar-
ies suggest a change from open, drier country in Blancan and Irving-
tonian time to denser, more mesic situations in Rancholabrean time in
Florida. Possible corroborative evidence was presented by Lundelius
(1960), who considered My lohyus the ecological equivalent of the Euro-
pean Sus , and associated Mylohyus with mesic forest. The idea that
Mylohyus was a woodland peccary and Platygonus a plains form is not
a new one according to Slaughter (1961), and was suggested early in
this century by Barnum Brown (1908) .

• 131-

Table 16

Measurements (in mm) of fossil dental samples of Platygonus
N = number of specimens, X = mean, O.R. = observed range

Santa Fe River, Loc. 1

Santa Fe River, Loc. IB

Inglis IA

Coleman HA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

2
Laubach Cave

3

Cherokee Cave

Length P2

Width P2

O.R.

O.R.

P. leptorhinus'

4

11.6

11.2-11.9

4

8.6

7.9-9.0

1

10.3

-

1

7.8

-

2

9.8

9.5-10.0

2

7.8

7.5-8.0

2

10.5

8.7-12.3

1

6.3

-

9

10.0

8.8-11.8

9

7.6

6.3-8.5

7

8.9

8.0-10.3

7

6.7

5.2-8.6

4

8.8

7-10

4

6.3

6-7

Length P3

Width P,

Santa Fe River, Loc. 1

Santa Fe River, Loc. IB

Inglis IA

Coleman HA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

6 13.9 12.6-14-6 6 10.4 9.5-11.6

113.3 - 19. 2

2 12.2 11.6-12.8 2 9.5 9-10

4 10.5 9.6-11.1 4 9.1 8.1-9.8

13 11.4 10.1-12.5 13 9.0 8.5-10.0

•132-

Table 16 - continued
(continued)

Cherokee Cave
P. Leptorhinus

Santa Fe River, Loc . 1

Santa Fe River, Loc. IB

Inglis IA

Coleman IIA

Haile VILA.

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. Leptorhinus

Length P3 Width P3

N X O.R. N X O.R.

11 10.6 9.7-11.4 24 8.6 7.8-10.2

4 10.0 10 4 7.8 7-8

Length P4

Width P.

3 13.2 12.6-13.8 3 12.1 11.5-12.7

14.3 13.3-15.1

12.2 11.6-12.8

11.9 11.3-13.2

11.6 11.5-11.7

3 12.5 12.3-12.6 3 10.7 10.5-10.9
13 12.0 10.3-13.1 13 10.4 9.1-11.3
10 12.1 11.0-13.8 25 9.9 8.9-12.2

4 10.8 10-12 4 9.5 9-10

Santa Fe River, Loc. 1
Santa Fe River, Loc. IB
Inglis IA
Coleman IIA
Haile VIIA
Cumberland Cave

Length M,
1 12.6*

Width M,

1 12.0

1 18.3 - 1 14.1

15 17.0 15.9-18.0 15 12.2 11.4-13.2

4 16.9 16.3-17.8 4 11.6 11.4-11.7

4 16.4 15.5-17.1 4 11.9 11.5-12.7

Table 16 - continued

■133-

(continued)

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Santa Fe River, Loc. 1

Santa Fe River, Loc. IB

Inglis IA

Coleman IIA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Santa Fe River, Loc. 1
Santa Fe River, Loc. IB
Inglis IA
Coleman IIA

Length M-^
NX O.R.

Width Mi

N

O.R.

3 14.6 14.3-15.2 3 10.5 10.2-10.8

14 14.7 13.1-16.3 14 11.3 10.5-11.9

15 14.2 13.3-15.0 19 10.3 9.7-11.0
6 12.3 10-15 6 9.8 9-11

Length M,.

Width nr

'2 '" 2

2 16.4 15.7-17.1 2 14.5 14.0-14.9

1 18.1 - 1 15.0

1 18.5 - 1 15.4

12 19.7 18.9-21.1 12 14.6 13.5-15.2

2 20.5 20.2-20.8 2 14.9 14.1-15.6
5 20.6 19.1-22.6 5 14.8 14.5-15.0

5 17.4 16.8-18.4 5 13.0 12.2-14.1

19 18.0 16.9-19.6 19 13.1 12.2-14.5

13 17.0 15.0-19.5 42 12.5 10.4-14.4
5 15.4 15-16 5 11.4 10-13

Length M3 Width M3

2 26.8 25.6-27.9 1 15.7

1 25 . 3

7 26.5 25. 8-27. J

1 15.2

7 15.3 14.7-15.9

-134-

Table 16 - continued

(continued)

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Santa Fe River, Loc . 1

Pool Branch

Inglis IA

Coleman IIA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Santa Fe River, Loc. 1
Pool Branch
Inglis IA

Length M„
N X O.R.

2 24.2 22.8-25.5

Width M3
NX O.R.

15.5

15-16

4 23.2 22.8-23.6 4 13.5 12.8-14.2
14 24.1 21.6-27.3 14 13.5 12.8-14.6
24 23.7 21.2-26.5 24 13.4 12.0-14.6

5 21.2 20-23 5 11.6 11-12

Length P'

Width P"

2 12.1 11.5-12.7 2 11.4 10.1-12.7

1 11.3 - 1 10.1

2 11.4 11.0-11.7 2 10.9 10.5-11.3

8 10.7 9.3-11.5 7 9.9 8.2-11.1

7 9.8 9.1-10.7 7 9.8 9.3-10.1

3 10.0 9-11 3 9.7 9-10

Length P~
1 12.5

Width P"
1 12.5

■135-

Table 16 - continued

(continued)

Coleman IIA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Santa Fe River, Loc. 1

Pool Branch

Inglis IA

Coleman IIA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Length P-3 Width PJ

N X O.R. N X O.R.

2 12.9 12.8-13.0 2 13.3 12.7-13.8

1 12.6 - 1 12.0

2 11.4 11.0-11.7 2 13.4 13.3-13.5

14 11.3 10.6-13.3 13 11.3 10.8-12.8

12 10.4 9.5-11.5 12 11.5 10.4-12.1

4 10.3 10-11 4 11.3 10-12

Length P

Width PL

3 12.6 11.9-13.0 3 14.7 14.4-15.4

1 12.6 - 1 14.5

2 11.7 11.6-11.7 2 14.2 14.0-14.3

1 12.7 - 1 13.1

14 11.5 10.7-12.8 14 13.7 12.9-14.6

21 10.8 9.3-12.7 21 13.3 12.3-14.8

4 9.3 9-10 4 12.0 11-13

■ 136-

Table 16 - continued

Santa Fe River, Loc . 1

Pool Branch

Inglis IA

Coleman IIA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Santa Fe River, Loc. 1

Pool Branch

Inglis IA

Coleman IIA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,
Haile VIIIA, Devil's Den

Laubach Cave

Cherokee Cave

P. leptorhinus

Length M1 Width M1

NX O.R. NX O.R.

1 17.4

1 15.7

6 16.9 15.7-17.7 6 14.7 13.3-15.8

3 16.9 16.2-17.4 3 14.5 13.7-15.0

5 16.2 15.6-16.5 5 14.6 14.1-15.5

1 15.3 - 1 13.4

19 15.1 11.8-16.1 19 14.2 13.5-15.8

19 14.2 13.3-15.4 29 13.6 11.9-16.0

5 12.6 11-14 5 12.2 12-13

Length M^

Width *T

1 22.2 - 1 17.4

8 19.5 18.9-20.2 8 17.0 16.4-17.7

1 19.9 - 1 17.0

2 18.8 18-19 2 17.0 16-18

2 18.4 18.1-18.6 2 16.6 16.2-16.9

19 17.6 16.0-19.7 19 16.3 15.1-17.4

17 17.5 15.1-19.3 34 15.6 13.8-17.4

4 16.0 15-17 4 14.3 14-15

■137-

Table 16 - continued

1

24.6

-

1

19.0

-

2

24.3

23.6-24.9

2

18.8

17.4-20.2

2

21.4

20.7-22.1

-

-

-

2

22.6

20.6-24.5

2

17.6

17.0-18.2

Length M3 Width M3

N X O.R. N X O.R.
Santa Fe River, Loc. 1 - - - -

Pool Branch - - - - -

Inglis IA

Coleman IIA

Haile VIIA

Cumberland Cave

Reddick IA, Reddick IB,

Haile VIIIA, Devil's Den 2 21.4 20.7-22.1 2 16.9 16.2-17.5

Laubach Cave 13 20.5 18.9-23.5 13 16.2 15.0-17.7

Cherokee Cave 12 21.0 19.0-23.6 26 16.2 14.8-17.9

P. leptorhinus 2 19.5 19-20 2 14.0 14

P. cumberlandensis ; this study and from Gidley and Gazin, 1938.

2
from Slaughter, 1966.

3

from Simpson, 1949.
4

recalculated from Simpson, 1949.

worn down to the roots.

Figure 16 -- Palate and partial skulls of Platygonus
cumber landens is from Coleman IIA (A, UF
12077 adult; B, UF 12084 young; C, UF 12084
young; D, UF 12076 adult).

-139-

Bl

C

H

T-1

a

O

•H

c

ctf

14-1

nJ

en

03

•i-i

t>0

■i-l

i— 1

^

4-1

as

>— I

M

4J

n

bC

a)

T3

C

rO

T)

01

03

H

b

•r-l

m

■u

"!

1—1

^-i

4-1

II

O

o

a

o

CO

h

r^

a>

03

a

i—i

a.

"

c

T)

01

C

H

03

q)

■—I

i-H

M

0

cu

;j

.a

ti

ii

=)

X

C3

4=

0]

T3

4J

i-4

-r)

on

[H

r-i

c

c

^-^

O

n3

OJ

CO

■—I

i—i

i— 1

H

»

1—1

cu

C

^H

rU

CI)

<D

0$

H

X3

jd

y

d

IS

4-1

CO

o

a

M

03

o

■—I

IS

M

CU

CO

.a

•r-l

b

3

c

0

»

o

a

M

cfl

03

CU

4J

n

J3

S

U

03

CO

• o.

-141-

e«j

_C9

O

• «!

_4fi>

c'.fj

M.J

< o

-4*

+

.fc»

CO

«>j

l

'J HJ.QIM

■142-

Family Camelidae
+ Tanupolama cf mirifica Simpson- camel
Material: UF 11967-12034, UF 16371; 10 dentaries , 3 palates, 2 fibulae,
60 phalanges, 8 calcanea, 6 scapulae, 10 ulnae, 15 humeri, 13 tibia, 8
femora, 10 astraguli, 17 metapodials , isolated teeth and foot bones.
Remarks: The camelid material from this deposit is from a single
species, and includes a good comparative sample. Preliminary investiga-
tion by S. David Webb and myself indicates that there are at least two
species of Pleistocene camels represented in Florida. Although they may
be labeled "larger and longer-limbed" versus "smaller and shorter-
limbed" it is clear from Figure 20 and Tables 17 and 18 that the main
difference between the metapodials of the two species is the ratio of
length to width, not the absolute size of either dimension. The Coleman
HA sample consists entirely of the latter species, and is the earliest
record of this form in Florida. Earlier faunas (Punta Gorda, Inglis IA,
and Payne's Prairie III) contain the other form. We have not as yet
worked out the nomenclatorial problems dealing with these beasts, but
it appears possible that the "shorter-limbed" variety will eventually
bear the name mirifica, a species named by Simpson (1929) from the
Florida Seminole Field deposit. Unfortunately the taxonomic characters
by which we are now able to adequately separate the two forms do not
include the lower molars, and the type specimen of T. mirifica includes
only the lower molars. Based upon Simpson's published figures of the
camel premolars, both species are represented in the Seminole Field
collections. The metapodial figured and labeled by Simpson (1929) as
Camelops is from the "shorter- limbed" Tanupolama. The lower dentition

•143-

and mandible of the Coleman HA camel may be viewed in Figures 18 and 19.

Family Cervidae

Odocoileus virginianus (Zimmermann) - white-
tailed deer
Material: UF 11912-11956; 15 calcanea, 7 astraguli, 12 femora, 15 tibia,
3 scapulae, 4 ulnae, 8 humeri, 19 phalanges, 5 dentaries , isolated teeth.
Remarks: The fossil material appears inseparable from comparable material
of the living Florida white-tailed deer.

Order Perissodactyla

Family Equidae

+ Equus sp.- horse

Material: UF 12035-12050; 1 scapula, 4 ulnae, 2 humeri, 6 metapodials,

2 femora, 2 calcanea, 2 astraguli, 6 phalanges, vertebrae and isolated

teeth.

Remarks: To refer any fossil Equus (subgenus or subgroup Equus ) from
Florida to species at this time appears to me impossible. Although it
would be possible to study the fossil Equus of Florida, voluminously
represented in Blancan through latest Wisconsin deposits, a study of
this nature is beyond the scope of this treatment.

The only conclusion I am now able to reach is that the Coleman
HA horse material is not identical to that of a large horse found in
almost all later Pleistocene faunas from Florida. Preliminary analysis
by myself and others indicates that horse teeth similar to those of
the Coleman IIA species also occur in some of these faunas, but I
cannot say whether this represents dental evolution or partial

■ 144-

replacement of one species by the other. The premolars of the Equus
(Plesippus) from the Blancan Haile XVA Florida deposit show a remark-
able similarity to both premolars and molars of the Coleman II horse,
teeth of the latter which are pictured in Figure 9 . Molars of the
Haile XVA horse are clearly representative of Plesippus .

Order Proboscidea

Family Elephantidae
+ Mammuthus sp.
Material: UF 14389; right lunar.

Remarks: The single lunar is the only element from the Coleman IIA
deposit definitely referrable to the Proboscidea. It is almost iden-
tical to the same element of a young Mammuthus from the late Wisconsin
Aucilla River deposits.

c

C

tfl

crt

H

y

CU

T)

r- 1

C

0

ctf

u

-146-

Figure 19 — Occlusal view of lower dentition
and mandible of Tanupolama from
Coleman HA (UF 11983).

-148-

a

CN

cfl

•H

<r

s

i-J

M

ii

ra

j3

0)

■H

o

>

u-l

X

en

P3

•H

cd

QJ

nj

Pd

M

M

CM

x

ni

PQ

0)

^-^

3

QJ

rH

cr

.— 1

a

cd

rH

en

O

x

d

crt

o

o

CB

i-J.

4-1

Cfl

PM

3

4-1

u

0

cfl

II

01

d

<x

01

.d

cd

TI

d

d

H

x:

01

3

•H

;s

4J

X.

s

►J

<

M

u

H

.-<

c

II

ctf

0)

>,

. ~

o

h

rH

X

CN

vO

T3

H

I

d

X

4J

T3

.-<

0

0

CO

0)

H

T)

S-i

OJ

4-)

H

M

0)

0,

4J

4-)

H

0)

l-l

a

•r-1

>

S-l

d

0)

01
S-i

CW

0)
•H

•H

PS

<C

H

d

d

cx)

o

>■,

(JU

•H

cfl

-150-

W *% mm

H19H31

■151-

Table 17

Tanupolama ("short-limbed" species); measurements

(in mm) of metapodials

I/II = greatest length divided by greatest proximal width

Regression line B in Figure 21 corresponds to

these measurements

Coleman IIA

UF 11977
UF 11976
UF 11976
UF 11976
UF 11991
UF 11994
UF 11994

II

I

Greatest

Greatest

Proximal

Length

Wid th

I/II

263

51.3

5.13

304

51.5

5.90

295

54.2

5.44

296

52.9

5.60

294

52.5

5.60

264

52.8

5.00

296

55.2

5.36

Ichetucknee River
UF 11320

310

60.7

5.11

Bradenton 51st Street
UF 3546

285

56.7

5.03

Arredondo IA

UF 828

UF 828

295
290

54.7
54.5

5.39
5.32

Sebastian Canal 2
UF 14247

289

51.2

5.64

Manatee Springs
UF 15159

Reddick IIC
UF 14173

275

242

56.5

48.8

4.87

4.96

Reddick ID
UF 10930

345

63.0

5.47

■152-

Table 18

Tanupolama ("long- limbed" species); measurements

(in mm) of metapodials

I/II = greatest length divided by greatest proximal width

Regression line A in Figure 21 corresponds to

these measurements

_

I Greatest

Greatest Proximal

Length Width I/II

Payne's Prairie III
UF 12521

394

51.0

7.73

Punta Gorda

UF 11187
UF 9717
UF 9718

314 40.3 7.79
321 43.4 7.40
319 40.2 7.94

Pool Branch
UF 11432

351

45.8

7.80

Haile VIIIA
UF 10928

374

50.0

7.48

Santa Fe River, Loc . 1-6

UF 11560
UF 10855

374
367

55.5 6.74

49.4 7.43

-153-

AGE AND CORRELATION

The Coleman IIA Local Fauna apparently stands at the Irvington-
ian-Rancholabrean boundary (Figure 21). We may tentatively label the
fauna as latest Irvingtonian rather than earliest Rancholabrean because
the fauna lacks Bison and includes mammals unknown in the majority of
later Pleistocene faunas of Florida (some of these forms include
Pitymys arata, Urocyon minicephalus , Sigmodon bakeri , Lepus alleni, and
Platygonus cumber land ens is ) , some of which may well be ancestral to
later Pleistocene species. In Florida the Irvingtonian-Rancholabrean
boundary is no longer clear. The Coleman IIA, Williston III, Haile VIIA,
and Bradenton localities bridge the gap between "typical" Irvingtonian
faunas such as Inglis IA and "typical" Rancholabrean faunas such as
Reddick IA. Of course, even if the boundary is no longer clear, the names
serve to denote periods of time in Florida (and elsewhere) which in
general evidenced quite distinct faunas.

Auffenberg's (1958, 1967) studies of the fossil and living box
turtles of Florida suggest that correlation may be made between cer-
tain Terrapene Carolina subspecies and sea level in the past. The
large, extinct subspecies T. c. putnami is considered a coastal,
savanna form, whereas the smaller T. c. Carolina is considered a
forest race which occurred at higher elevations. The Coleman IIA
Terrapene are small, and although they may represent T. c. putnami and
T. c. Carolina intergrades , they are not good putnami. According to
Auffenberg (1958) this suggests either that sea level was not much
above its present level or that sea level may have been lower than
at present. However, if H. K. Brooks (pers . coram.) is correct in

■154-

his view that the 25 foot stand of sea level corresponds to the latest
Yarmouthian, then it is conceivable that Terrapene Carolina putnami
might not be expected as far inland as Coleman during this time. If
the Coleman IIA site represents deposition during a glacial period, it
is fairly certain that this period would be of Illinoian, rather than
of either Wisconsin or Kansan time. Faunas of known Wisconsin age,
such as Vero (Weigel, 1962), Devil's Den (Martin and Webb, in prep.;
H. K. Brooks, pers. comm.), Melbourne (Ray, 1958), Seminole Field
(Simpson, 1929; Auffenberg, 1958), contain the extant species Pitymys
pinetorum and Sigmodon hispidus as well as the extinct species Platygonus
compressus , whereas Coleman IIA contains the extinct species Pitymys
arata, Sigmodon bakeri, and Platygonus cumber landens is. Further, a
fauna in Florida assigned to the Kansan glaciation, Inglis IA, contains
the extinct Sigmodon curtisi (a more primitive grade than S. bakeri),
the antelope Capromeryx, and an undescribed species of Platygonus that
was clearly ancestral to the Coleman IIA P. cumber landens is .

My present ideas on mammalian faunal changes in Florida during
the Pleistocene are noted in Figure 21. Florida is the only area in
the New World in which four species of Sigmodon, representing three
evolutionary grades, have been recovered. These species, along with
others noted in the same illustration (Figure 21), facilitate correla-
tion of most major Pleistocene deposits of Florida. The following
deposits correspond to the sequence noted in Figure 21:
Hemphillian

McGehee Farm

Withlacoochee River, Loc. 4A

Manatee Dam

• 155-

Hemphillian (continued)

Bone Valley
Mixson's Bone Bed
Emathla

Late Blanc an

Santa Fe River, Locs . IB, 4A, 8A
Haile XVA

Irvingtonian

Punta Gorda (includes Mammuthus (Archidiskodon)
haroldcooki)

Inglis IA
?Payne's Prairie III
?Pool Branch

Coleman IIA

Sangamon

Early

Williston III

Bradenton 51st St. and ?Bradenton Field

Haile VIIA

Late

Reddick IA, IB, IIC

Haile VIIIA
?Haile XIIIA, B, C
?Arredondo IIA, B

Sabertooth Cave
?Payne's Prairie B

Wisconsin

Early ( >30,000 years B.P.)

Vero, Bed 2 ( > 30,000 years B. P., cl4; Weigel,

1962)
Melbourne, Golf Course Loc.
Seminole Field
?Arredondo IA

Middle (c . a. < 30,000, > 11,000 years B.P.)

Ichetucknee River (part; H. K. Brooks, pers.comm.)
Withlacoochee River, Loc. 7A
Aucilla River (part)

-156-

Wisconsin (continued)

Late ( < 11,000 years B.P.)

Devil's Den (c. a. 8,000 years B. P., cl4; H. K.
Brooks, pers. comra.)

Coleman IIA cannot clearly be correlated with any deposit outside

of Florida. The closest approximation appears to be the Cumberland

Cave fauna of Maryland (Gidley and Gazin, 1938), which contains Platygonus

cumberlandensis , Canis lupus (see my discussion in the Species Account)

and at least one extinct species of Pitymys (personal observation).

Although the Cumberland Cave fauna appears to be of Irvingtonian age,

the mammalian remains have not been reviewed in 30 years, and until this

is done it would be too hazardous to speculate further.

Figure 21 — Replacement of mammals in Florida
during the Pleistocene.

-158-

ttCUl

■Jr

1 pennsylvat

S hispidus

Bison latifrons /

i < 1 1 —

Ctmepatus \ Pitymys aiaia Mybhyus

Symplomys tislrtlis'.

C«le»ai II»

P cumberltndeitsis

iniNTMMN

LUTE ILMCM

Sigmodon medius Ctpromeryx

Iqmis

HILT MIKM

-159-

AFFINITIES

OF

THE

COLEMAN

HA MAMMALS

The

Coleman HA fauna

contains four

living

taxa,

Erethizon

dorsatum,

Conepatus sp. , Lcpus

all<

jni , and

Felis

one a,

that

are

no

longer found in Florida. Erethizon could have entered Florida either
from the north or from the west, but as the fauna contains other obvi-
ous western immigrants (Lepus , Conepatus) , and lacks any mammals now
confined north of Florida, it seems more likely that the porcupine also
wandered in along the Gulf Coast corridor during the Illinoian. In
fact, it is quite probable that almost the entire body of Coleman HA
mammals, and most of the living mammals of Florida as well, entered
Florida from the west along the Gulf Coast. Table 19 shows the Cole-
man HA mammal species (exclusive of the bats) of which any dispersal
and evolutionary centers are known or can be postulated with any degree
of confidence. Only two out of eighteen (11%) possibly entered Florida
from the north, or Appalachian route, while sixteen (89%) presumably
entered Florida from the west along the Gulf of Mexico. The balance of
living Florida terrestrial mammals may be added to this list (Table 19):
of the six species for which immigration routes can be determined, four
entered via the Gulf Coast route. Thus, approximately 83% of the liv-
ing terrestrial mammalian fauna of Florida has ultimately western
affinities.

Table 19, however, requires some discussion. Including the most
closely related living species to the Coleman HA mammalian species
in many cases was useful in defining the evolutionary and dispersal
centers for the latter species, but in other cases actually clouds
the issue. For instance, it was assumed by the author that those areas

■160-

which now attain the highest density of living species related to the
Florida fossil and living species represents the general areas from
which the Florida forms originated. In general this may be valid, but
in certain instances this is clearly not the case.

The Coleman IIA Glaucomys cannot be identified to species. It may
be either G. sabrinus, G. volans, or an intermediate between the two.
At any rate, the geographic (and evolutionary) connection between the
two appears most probably to be the Appalachian chain. Glaucomys
sabrinus exhibits a positive Bergmann's response, with populations of
small individuals, such as G. s. coloratus, located on isolated moun-
tain tops in the southern Appalachians. I believe that one of these
small subspecies ultimately gave rise to G. volans during and subse-
quent to a major glacial period; probably Illinoian.

The fossil Geomys are referred to G. pinetis purely on a geo-
graphical basis; they are morphologically referrable to G. personatus
as well. Pocket gophers probably existed in continuous populations
across the Gulf Coast in either or both Illinoian and Wisconsin time.
High sea levels, plus the great Mississippi alluvial fan, probably
isolated the eastern Geomys pinetis from its western relatives.

Reithrodontomys humulis is a grassland or field rodent, most
closely related to the Latin American R. burti (Hooper, 1952) of the
subgenus Reithrodontomys . There is little doubt that the ultimate
roots of R. humulis, and most' of the other species of the genus, are
in southwestern North America.

Sigmodon bakeri may be most closely related to S. hispidus, an
extant Florida species. However, both may trace their ancestry to the

-161-

hodgepodge of Mexican Sigmodon.

Dasypus bellus is most closely related to the living D. novemcinctus
which, until rather recently, was confined south of the Rio Grande. In
recent time D. novemcinctus has moved northward and eastward (movement of
this species in Florida is not relevant to this discussion, as D.
novemcinctus was recently introduced into Florida; Neill, 1952). This
same Gulf Coast pathway was undoubtedly utilized also by D. bellus,
which reached Florida in the late Blancan (Haile XVA time) and remained
there until the late Wisconsin.

Lepus alleni now exists primarily in northwest Mexico. Perhaps
its closest living relative is L. californicus , but recording this fact
has no application to the origin of L. alleni, the latter species which
simply moved into Florida from the west.

Peromyscus floridanus has no close living relative. It lives al-
most exclusively in open, sandy, scrubby areas. According to Layne
(1963), its origin was in the west.

The western mule deer, Odocoileus hemionus , is the most closely
related living species to the white-tailed deer, 0. virginianus, but
recording this fact, as in the case of Lepus californicus, simply ful-
fills the requirements of the table. The evolutionary and dispersal
patterns of these species are unknown.

Peromyscus gossypinus was probably derived from its leucopus-
group relative, P. leucopus, as the latter species' Austroriparian
equivalent, much in the same fashion as Synaptomys australis was de-
rived from S. cooperi. I suggested (1967) that speciation of the
leucopus group may have been a Wisconsin or post -Wis cons in phenomenon,

-162-

but new evidence proves that P. gossypinus was present in Florida during
the early Sangamon. Peromyscus gossypinus probably evolved from a P.
leucopus population pushed south into the Gulf Coast region during the
latest Illinoian. Although P. leucopus can now be found in Texas and
Louisiana, and is sympatric in east Texas and parts of Louisiana with
P. gossypinus, in these areas P. leucopus is always smaller than else-
where (the smallest leucopus subspecies is the east Texan texanus ;
Martin, 1968b), while P. gossypinus is larger in these areas of sympatry
than it is elsewhere (e. g. , P. g. megacephalus) . This perhaps repre-
sents character displacement to a certain extent, and I look for another
area to the east, where the morphological traits overlap more (Dismal
Swamp, Virginia; Dice, 1940; Martin, 1967) for the species connection.
Speciation of the two forms perhaps culminated in the early or middle
Sangamon as sea level rose and warming trends separated populations of
the two forms .

I do not know which living species of squirrels are most closely
related to Sciurus carolinensis and S. niger, unless they are each
other, which avails nothing in determining past dispersal patterns.
The origin of S. carolinensis is obscure, but S. niger did not appear
in Florida until about 8- to 10,000 years ago, and is recorded as a
fossil only in the Devil's Den fauna (Martin and Webb, in prep.), the
latest, coldest time period in the Pleistocene history of Florida.
Mammals from this fauna indicate movement solely from the north.

According to Bowen (1968) the ancestor of Peromyscus polionotus
was not, as Osgood (1909) suggested, the Texas P. maniculatus pallescens.
but rather another maniculatus subspecies such as bairdii, the latter

■163-

ranging now as far south as southern Arkansas, and typically an animal
of the Upper Austral and Transition Life Zones (Hall and Kelson, 1959).
Bowen suggests this to explain the absence of P. polionotus in appar-
ently suitable habitat west of the Alabama River. He states (1968),
"Had the species evolved from an eastward extension of maniculatus stock
in Texas, the suitable habitats in western Alabama would surely be
occupied today." This is not convincing, as the same could be said of
Geomys pinetis, which was obviously derived from some ancestral Gulf
Coast stock (possible G. personatus of eastern Texas), and which also
does not cross the Alabama River into "suitable habitats." The Mobile
Bay beaches west of the Alabama River are isolated and not extensive.
Woodland comes down very close from the narrow beaches there from the
north, and the Mississippi delta encroaches from the west as the Ala-
bama River alluvial plain does from the east. Minor glacial retreats
coincident with raised sea levels during the Wisconsin are recognized ,
and a good example is the Two Creeks interstade at ca. 11,800 years
B. P. Minor interstades would possibly be capable of flooding out the
entire Mobile Bay region and connecting the Mississippi and Alabama
River flood plains, or at the very least producing riparian forests in
this region. Such minor sea level fluctuations would effectively
decimate any grassland or beach mammals west of the Alabama River,

1
The position of Bowen that some subspecies of P. polionotus
had evolved by the Yarmouthian is very unlikely, especially consid-
ering that the 150 foot Okefenokee shoreline which he correlated with
this interglacial is in actuality of Pliocene age (Brooks, 1968).

•164-

but would not be capable of removing all such populations to the east,
as any part of Florida would make an adequate refugium. In actuality
it does not appear to me to be particularly significant that neither
G. pinetis nor P. polionotus occurs west of the Alabama River. Al-
though I believe that my explanation is more reasonable than that of
Bowen (1968), there are so many examples of areas that appear hospitable
to a particular species that remain unoccupied that I must agree with
Deevey (1967) who suggested that "... a niche not occupied, or occupi-
able, by a known organism is an unhelpful construct, like a hoop snake."

•165-

>

CO

rQ

fe

X)

•r-l

0)

o

u

rfl

4J

o

4J

r-l

M

fe

o

IW

o

r-4

0

4J

QJ

c

S*

>.

•H

03

4-1

i-l

(U

0J

E

CJ

•H

<

i-l

c

O

M

X

cd

QJ

H

o

r4

ft

u

4-1

CO

C

ft

Cfl

s

QJ

QJ

cd
0

«

r-l

,n

QJ

QJ

cfl

14-1

4-1

r-l

CO

s

O

O

o

u

r-4

cd

0)

<

CJ

B

4-1

H

E

3

M

0

II

<

o

!-4

H

H

3

M-l

H

M

cd

M

QJ

B

X)

3

J3

0)

QJ

cd

4J

r-l

r4

CO

E

O

QJ

3

5

0

o

>

3

& r-1

4-)

O

QJ

r-l O

c

U

60

o

(-1

«H

QJ

CD

QJ

U

QJ

r-l 0)

■4-1

TJ

E

rQ J3

QJ

C

4-J

cd

cd u

r4

3

O

CO

H

O

s

>4-l

3

<H

M-l

O

cd

cu

o

•r4

QJ

H

co

Si

rJ

QJ

CO

ft

0

cd

|S

QJ

i-i

cd

•r-l

rC

r-l

QJ

4-1

CJ

CO

cd

s

3

QJ

a

a,

■H

^3

ft

o

ft t-I

CO

•H

<

4-J

4J

a)

a

60

cd

T3

r-4

cd

3

i— i

s

>fl

4J

•H

0)

cd

g

X

>

pi

0

QJ

•H

4-1

xi

H

CO

0)

cd

c

M

T3

o

QJ

cd

QJ

U

-M

4-1

0

QJ

cd

M-l

^4

3

H

r-4

rQ

•H

QJ

3

r-4

M

Q

QJ

.0

CJ

0) cfl

si

4-)

H CO

CO

<u

0

•r4

B

o

QJ

II

I
cd

r-l C

CO cd

a. -h
< o

cd

QJ

3

3

•i-l

3

H

QJ

o

H

o u

o

o

o

cd

O

o cd

o

■H

u

CJ

3

CJ

•H U

•r-l

ft

CO

•t-4

•H

cd

•r-l

X

X

ft

X

X

•H

X

0) J3

QJ

•r-l

CO

CO

CU

CJJ

CO

0)

E *->

a

co

cd

cd

i cd

cd

a «

i cd

a i

1

•i-i cd i

i a

r4

CO

X

T3

T3

X

TJ

cd

3 T3

• O

•H

QJ

■i-l

•r4

QJ

•H

E

O -H

w s

w

co

CO

•H

H

o
1— 1

U

o

r-l

H

i

(-1
O
r-l

w

cd
3
cd

rJ r4

O

. r-l

3

S5 W*

23

a

W

pen

fe

W

S

fe

CO

ft

[2 Pf

CO

3

CO

i—l

3

cd

3

0-

•r4
U

S*

CO

cd

CO

u

CO

co

o

CO

3

3

a

QJ

4-J

CJ

co

co

4J

C)

■1-1

3

M

3

cd

CO

CO

3

3

3

O

4J

3

CIJ

3

•H

U

cd

Cfl

cd

I-l

4J

C)

0

-o

CO

c

CJ

cd

cd

ta

M-l

•rl C^-

3

O -r4

•H

&

r-l

cu

•H

r4

cfl

CO 4J

X

H

M-l

>

r-l

O

r-4

r4 U

QJ

QJ

3

CJ

cd

H

0

QJ 3

b

4J

1-1

3

CJ

M-l

>

ft ,n

u

PQ

CO

P

P

CO

o

O OS

CO

QJ

3

4J

Td

Cfl

•rl

•H

(X

CO

CO

cfl

w a

C— C^- O- CN-

XX X X X X e~ ">■ X X X! «"• X! X <>• X X! p- <>■ »• <>• X

cd

cd

CO

u

CO

3

cd

r-l

B

rH

CO

CO

•H

-a

3

4J

3

CO

-3

TJ

r-l

3

cd

3

3

cx

3

O

3

QJ

•t-4

•H

cd

4-J

J3

3

•rl

3

cd

cd

•i-l

QJ

3

4-1

co

)-l

4J

■H

•rl

cd

ft

ft

•rl

!-4

CO

>

CJ

4J

CO

4J

• 1-4

QJ

1-

o

4J

!-4

cd

3

CO

CO

a)

4J

O

to

u

•rl

cd

3

ft

ft r-l

3

r-l

3

QJ

4J

QJ

u

CJ

CO

U

4J

cd

cd

>

3

i-i

QJ

i-i

cfl O

ft

•r4

o

M-l

3

Jd

cd

r-l

O

CO

•rl

•r4

o

3

E

ft

QJ

cr

i—i

co

3

U

CO

ft

■u

cfl

u

1-1

TD

3

CO

3

^4

4J

ft

l-i

cfl

QJ

QJ

co cd

3

CO

CO

-a

cd

cd

^4

3

•rl

ft

o

CO

CO

43

rQ

cd

r-4

3 O

CO

M-l

O

3

IT

3

QJ

ft

B

H

QJ

•rl

•H

CO

3

r-l

M

^

0

X)

CJ

3

co

^4

O

O

3

co

r4

J3

4J

cd

3

CO

•r4

cd

cd CO

0

CO

0

0

3

QJ

N

j3

r-4

c

3

c

cd

a

O

3

ft

3

CO

r^ 3

O

CO

u

r

4J

-a

4J

rQ

•rl

CJ

o

T3

o

bO

r-l

4J

•rl

O

Q-

QJ

CO

•rl U

a

r

Si

o

o

o

•rl

J3

0

CO

>,

O

>>

O

QJ

ft

^

i—l

>-,

E

3

> 3

3

4J

0

u

gj

!-4

M-l

4-)

u

•r-l

CJ

4J

u

r-4

T3

r?

cd

cfl

CO

r-l

ft r-l 1-1

cd

0

•rl

!-4

-3

0

o

QJ

T3

3

O

o

o

•H

•rl

r-l

CJ

cd

O

CU

>> O

r-4

QJ

QJ

QJ

CJ

•H

•rl

QJ

U

>,

cd

u

!-i

M

ft

P

u

PQ

CO

P

+

KB

+

r4

00 CO

O

o

«

Pm

O

CO

+

a
+

S3

w

ffl

+

CJ

+

<
+

ft

CO

-166-

cd

3
cfl cfl

<J o

X X

01 0>

S H

o w >
Sot ja

cu

td

.

O

h

o

0}

CD

3

h

4-1

3

c

0

O P.

H

0

0

3

nj

0)

S u

^* X X p" ^ ^' ^" ^* ^"

B M

S U

an

B

M

3

3 4-4

•H

cfl

CO

O • O

>

cfl

3

CO t/> tfl

CO

0

cw

2 S

3

c

3

(X

R eg id

OJ

0

H

10

O 3 rH

60 >> O

1-1
•H

n

CO

en

>>J3 ft

0

T-l

•r-l

3

■U O 3

a

H

H

3

cfl i-i c

o

01

(!)

rr

e: ^f2

T)

ta

h

W

rt

+

+

+ + +

1 co O co

X X

X

1 -H co O «

o> 0)

CO CD

1 co cfl v-i co 1

i H H

a a

1 co X X! C

•■-1

1 -H CU 0) cfl

cfl •

1 S H S M

1

W W

a u

O O 3: O*

CO CO

co w

CO H fl H CO C 3

3 r-l 0) -i-l 3 O CU

a 3 > 45 T-l ft X) 4J

O O -H «H e- P- c^- CO O c- O O

S «

a J

{x; ><; c— o- e- X X i>- e- **■•

3 3

a w

•r-l o

CO

CO

ft c

co

•H

•r-l Cfl

>> o

3

rn

M

U 3

CO

CO iH

T)

"5

4J

4-1 Cfl

cfl

•H

CO i-H

•H

3

M

CO

CO T)

■u

CO

o o

ft

cfl

O

O

u

3 -r-l

cfl

3

3

60 ft

co

o

■—I

r-4

CD

i-H U

C

o

CU

•T-l

•r-l

0

•r-l

hi)

cfl O

a)

en

t>

co co

,3

u

o

61)

•r-l

(X i-H

r-l

•r-l

cfl

3 3

a

3

3

3

4-1

m

>

3

o o

c

H

0

O

CO

cfl

CO CO

0

co

a

r-H

en

o o

Cfl

cfl

0

I"!"

X)

o

HI

CO

X

3

i—l

.— 1

cu

cfl

o o

h

3

•1-1

a;

3

N 4-1

4-1

4-1

n

r-l n

on

rn

i-l

u

•r-l

u cu

CO

CO

4-1

0) 0)

■H

(-J

cu

()

0

1

3

3

ft ft

CO

3

ta

CO

CO

O 55

X

s

r-l

-167-

PALEOECOLOGY

There is actually very little that can be said of the paleoecol-
ogy of an isolated aggregation of fossil mammals which accumulated over
an interval of unknown duration. The limited area and depth (less than
3 feet) from which the Coleman fossils were collected, the good pre-
servation of these fossils, and the associated elements of single in-
dividuals which transgress the facies boundaries, suggest rapid deposi-
tion (perhaps less than 1,000 years), but there is no way that this may
be substantiated. The term isochronous may be applied to the Coleman
IIA fauna as a unit relative to other faunas of, for example, early or
late Sangamon time, but to consider the mammalian remains from the Cole-
man sink as a homogeneous, natural unit, analagous to what one could
slaughter in a year or so on an African savanna, would be making an
assumption which, although conceivable, is highly unlikely. Further,
there is not now an adequate body of information concerning the habitat
preferences of living Florida mammals on which to draw for comparison
to the fossil fauna. Some data may be gleaned from the works of
Layne (1963), Pournelle (1947), Ivey (1947), Barrington (1949), and
Pearson (1951), but there is a good deal of disagreement among these
authors, and a more thorough, Florida-wide study is required.

Peromyscus floridanus is one of the few Florida mammals for which
there is enough data to conclusively determine its habitat preference
(Layne, 1963). This species is confined almost entirely to sand pine
scrub and longleaf pine/turkey oak sandhill associations. According
to Layne (1963) these two habitats are relatively xeric and attain the
highest overall evaporation rates of Florida plant communities.

•168-

Peromyscus floridanus remains were the most numerous of all peromyscine
rodents recovered from the Coleman IIA deposit. The predominant vege-
tation surrounding the Coleman IIA deposit now is mesic hardwood forest,
which does not support any populations of this species.

Although the number of fossil specimens is small, the six to one
ratio of least shrew (Cryptotis parva) to shorttail shrew (Blarina
brevicauda) remains are further suggestive of a drier situation than
dense mesic hammock. Both Pournelle (1947) and Barrington (1949)
testify that the shorttail shrew is most common in mesic forest and
forest border areas; found especially in deep humous and forest bottom
litter. The least shrew was most commonly found by these authors in
pine flatwoods and burned marginal thicket, and I have trapped this
species mostly in grass or sedge fields in Sigmodon runways.

Remains of the antelope jackrabbit Lepus alleni also support a
suggestion of xeric habitat during the Coleman IIA time. This species,
now confined to western Mexico and southern Arizona, is found primarily
in the open, arid to semi-arid Pacific coastal lowlands (Burt, 1938;
Hall and Kelson, 1959).

The presence of gray squirrel (Sciurus carolinensis) and golden
mouse (Ochrotomys nuttalli) remains suggest that some denser, perhaps
more mesic vegetation was also being sampled, but these species have
been recorded in xeric habitats (Ivey, 1947; Pournelle, 1947), and the
small number of individuals fossilized does not allow anything other
than this suggestion.

The above information, scanty though it is, leads me to be-
lieve that the Coleman area during Coleman IIA time was somewhat more

■169-

open and more xeric than it is today. The predominance of savanna species
in the earlier Inglis IA fauna, and the predominance of mesic forest and
mesic forest border species in faunas later than Coleman IIA suggest that
Coleman represents the transition between a purely savanna mammalian
fauna to one of subtropical to tropical mesic forest nature. Concordant-
ly, the species diversity increases (Martin, 1969b) from 35 in Coleman
IIA time to 39 during the early Sangamon, and to 48 in the middle to
late Sangamon time.

SUMMARY AND CONCLUSIONS
Fossil mammals from the Coleman IIA local fauna were apparently
trapped and died during a relatively xeric period between classical Irv-
ingtonian and classical Rancholabrean time as recognized by most mammalo,
gists. The Coleman IIA local fauna clearly demonstrates affinities to
the earlier Irvingtonian Inglis IA deposit, both faunas containing
Lepus alleni, Erethizon, and Canis lupus. The fossil Sigmodon from the
Inglis deposit, S. curtisi, is of a more primitive grade than the Cole-
man S. bakeri. Sigmodon bakeri is found also in the Haile VIIA, Bra-
denton Field, and Williston III deposits, but Coleman IIA is clearly
older than these three deposits because they contain Pitymys hibbardi,
Synaptomys australis, or Bison latifrons. The Coleman IIA Pitymys
arata is of more primitive nature than is P. hibbardi. The new species
described from Coleman IIA, Sigmodon bakeri, Pitymys arata, and Urocyon
minicephalus do not appear readily derivable from any known extinct
species .

The Coleman IIA wolf is Canis lupus. Canis lupus was replaced by
C. dirus throughout most of North America during the Rancholabrean.

The Inglis IA Platygonus was ancestral to the Coleman IIA P.
cumber landens is, and as a clear trend in size diminution can be seen
in Platygonus throughout the Pleistocene, it seems reasonable to con-
clude also that P. cumber land ens is gave rise to the smaller P.
compressus .

Coleman IIA time is the earliest period from which Tanupolama
mirifica, or if that name does not prove to be valid, "the shorter-
limbed" Tanupolama with complex fourth upper and lower premolars,"

-170-

■171-

has been recorded. This species is apparently an immigrant as it was
probably not directly derived from the large, "longer-limbed" Tanupolama
which has been recorded from earlier Florida deposits such as Inglis IA
and Punta Gorda.

The Coleman IIA Sigmodon bakeri is a member of the hispidus
species group of Sigmodon, and in that respect heralds the Rancholabrean
period in North America, dominated by Sigmodon hispidus. Sigmodon
hispidus replaces S. bakeri during the Sangamon in Florida, and is the
living Florida representative of the genus.

The Coleman IIA local fauna has aided immensely in solving some of
the major problems of mammalian immigration, evolution, and extinction
during the Pleistocene, in Florida especially, but also throughout
North America. It is my hope that soon we will be able to sequence all
the North American Pleistocene deposits as now appears possible for
those from Florida.

LITERATURE CITED

Akersten, W. 1968. Master of Science thesis, U. Texas, Austin.
Arata, A. A. 1961. Meadow vole (Microtus pennsylvanicus) from the

Quaternary of Florida. Quart. J. Florida Acad. Sci.,

24: 117-121.
and J. H. Hutchison. 1964. The raccoon (Procyon) in the

Pleistocene of North America. Tulane Stud. Zool., 2(2):

21-27.
Auffenberg, W. 1958. Fossil turtles of the genus Terrapene in Florida.

Bull. Florida State Museum, 3(2): 53-92.
. 1963. The fossil snakes of Florida. Tulane Stud. Zool.,

10(3): 131-216.
Bader, R. S. 1957. Two Pleistocene mammalian faunas from Alachua

County, Florida. Bull. Florida State Museum, 2(5):

57-75.
. 1959. Dental patterns in Peromyscus from Florida. J.

Mammal., 40(4): 600-602.
Barrington, B. A., Jr. 1949. Mammals of a north Florida flatwoods.

Ph. D. dissertation, University of Florida, Gaimesville.
Bowen, W. W. 1968. Variation and evolution of Gulf Coast populations

of beach mice, Peromyscus polionotus . Bull. Florida

State Museum, 12(1): 1-91.
Brodkorb, P. 1953. Pleistocene birds from Haile, Florida. The

Wilson Bull., 67(1): 49-50.

-172-

-173-

I

. 1957. New passerine birds from the Pleistocene of

Reddick, Florida. J. Paleont., 31: 129-138.
. 1959. The Pleistocene avifauna of Arredondo, Florida.

Bull. Florida State Museum, 4(9): 269-291.
Brooks, H. K. 1968. The Plio-Pleistocene of Florida in Late Cenozoic

Statigraphy of Southern Florida- A Reappraisal; compiled

by R. D. Perkins; Second Annual Field Trip Miami Geol.

Soc. : 3-42.
Brown, B. 1908. The Conard Fissure, a Pleistocene bone deposit in

northern Arkansas: With description of two new genera

and twenty new species of mammals. Amer. Mus . Natur.

Hist. , Mem. , 9: 157-208.
Burt, W. H. 1938. Faunal relationships and geographic distribution

of mammals in Sonora, Mexico, Misc. Pubs. No. 39, Mus.

Zool., Univ. Michigan: 7-77.

Cahn, A. R. 1922. Chlamytherium septentrionalis , a fossil edentate
new to the fauna of Texas. Jour. Mammal., 3: 22-23.

Castellanos, A. 1937. Anotaciones sobre la linea filogenetica de los
clamiterios. Publ. Fac. Gien. , Univ. Nac . Litoral,
Santa Fe, Argentina; Ser. Tecnico-Cientifics , Publ.
No. 8: 1-35.

. 1927. Breves notas sobre los clamidoterios . Centro.

Estud. de Ingeneria de Rosario: 1-8.

. 1957. Nuevos generos de clamiterio y de Gervo y una

nueva especie de Pliohophorus . Ameghiniana, 1(3): 5-14.

-174-

Churcher, C. S. and Van Zyll de Jong, C. G. 1965. Conepatus talarae

n. sp. from the Talara Tar-Seeps, Peru. U. Toronto,

Royal Ontario Museum, Life Sci. Contribs. 62: 1-15.
Dalquest, W. W. 1953. Mammals of the Mexican state of San Luis Potosi.

La. State Univ. Stud., Biol. Sci. Ser. No. 1: 112 pp.
. 1965. New Pleistocene formation and local fauna from

Hardeman County, Texas. J. Paleont. 39: 63-79.
Deevey, E. S., Jr. 1967. Introduction; in Pleistocene Extinctions, P.

S. Martin and H. E. Wright, Jr., eds . , Yale Univ. Press,

New Haven and London, 453 pp.
Dice, L. R. 1940. Relations between the wood-mouse and the cotton-
mouse in eastern Virginia. J. Mammal. ,21(1) : 14-23.
Elftman, H. 0. 1931. Pleistocene mammals of Fossil Lake, Oregon.

Amer. Museum Novitates , No. 481: 1-21.
Gazin, C. L. 1942. The Late Cenozoic vertebrate fauna from the San

Pedro Valley, Arizona. Proc. U. S. National Museum,

92(3155): 475-518.
. 1950. Annotated list of fossil Mammalia associated

with human remains at Melbourne, Florida. Wash. Acad.

Sci. J., 40: 397-404.
Gidley, J. W. 1913. Preliminary report on a recently discovered cave

deposit near Cumberland, Maryland. Proc. U. S. National

Museum, 46: 93-102.
. 1920. Pleistocene peccaries from the Cumberland Cave

deposit. Proc. U. S. National Museum, 57: 651-678.
. 1922. Preliminary report on fossil vertebrates of the

San Pedro Valley, Arizona. U. S. Geol. Surv. Prof.

Paper 131-E: 119-131.

I

•175-

and Gazin, C. L. 1938. The Pleistocene vertebrate fauna

from Cumberland Cave, Maryland. U. S. National Museum
Bull. 171: 1-99.
Guilday, J. E. 1962. The Pleistocene local fauna of the Natural
Chimneys, Augusta County, Virbinia. Annals Carnegie
Museum, 36: 87-122.

and Bender, M. S. 1959. Late Pleistocene records of the

yellow-cheeked vole, Microtus xanthognathus (Leach).
Annals Carnegie Museum, 35: 315-330.

, Hamilton, H. W. , and McCrady, A. D. 1966. The bone

breccia of Bootlegger Sink, York County, Pa. Annals
Carnegie Museum, 38: 145-163.

, Martin, P. S. , and McCrady, A. D. 1964. New Paris No. 4:

A Pleistocene cave deposit in Bedford County, Pennsyl-
vania. Bull. Natl. Speleol. Soc, 26(4): 121-194.

Gut, H. J. and Ray, C. E. 1963. The Pleistocene vertebrate fauna

of Reddick, Florida. J. Fla. Acad. Sci., 26(4): 315-328.

Hall, E. R. and Kelson, K. R. 1959. The Mammals of North America.
The Ronald Press, New York: 1079 pp.

Handley, C. 0., Jr. 1959. A revision of American bats of the genera

Euderma and Plecotus. Proc. U. S. Natl. Museum, 110(3417)
95-246.

Hay, 0. P. 1926. A collection of Pleistocene vertebrates from south-
western Texas. Proc. U. S. Natl. Museum, LXVIII,
Art. 24: 1-18.

■176-

Hershkovitz, P. 1962. Evolution of neotropical cricetine rodents

(Muridae) with special reference to the Phyllotine Group.

Fieldiana, Zoology, 46: 1-524.
. 1955. South American marsh rats, genus Holochilus ,

with a summary of sigmodont rodents. Fieldiana, Zoology,

37: 639-687.
. 1966. Mice, land bridges and Latin American faunal

interchange, in Ectoparasites of Panama; Wenzel, R. L.

and Tipton, V. J., eds . ; Field Museum Nat. Hist., Chicago;

725-751.
Hibbard, C. W. 1937. A new Pi tymys from the Pleistocene of Kansas.

J. Mammal. , 18(2): 235.
. 1938. An upper Pliocene fauna from Meade County, Kansas.

Kansas Acad. Sci. Trans., 40: 239-265.
. 1941. The Borchers fauna, a new Pleistocene interglacial

fauna from Meade County, Kansas. U. Kansas Pubs. State

Geol. Surv. Kansas, Bull. 38: 197-220.
. 1949. Pleistocene stratigraphy and paleontology of

Meade County, Kansas. Contribs. Museum Paleont. U. Mich.,

VII (4): 63-90.
. 1952. Vertebrate fossils from late Cenozoic deposits

of central Kansas. Vertebrata, Art. 2: 1-14.
. 1955. a. Notes on the microtine rodents from the Port

Kennedy Cave deposit. Proc . Nat. Sci. Phil. CVII: 87-97.
. 1955. b. The Jinglebob interglacial (Sangamon?) fauna

from Kansas and its climatic significance. Contribs.

Museum Paleont. U.. Mich. , XII(IO): 179-228.

■177-

. 1963. A late Illinoian fauna from Kansas and its climatic

significance. Papers Mich. Acad. Sci., Arts and Letters,
XLVIII: 187-221.

. 1967. New rodents from the late Cenozoic of Kansas.

Papers Mich. Acad. Sci., Arts and Letters, LII: 115-131.

, and Dalquest, W. W. 1966. Fossils from the Seymour Forma-
tion of Knox and Baylor Counties, Texas, and their bearing
on the late Kansan climate of that region. Contribs. Mus.
Paleont., U. Mich., XXI (1): 1-66.

. 1944. Stratigraphy and vertebrate paleontology of Pleis-
tocene deposits of Southwestern Kansas. Geol. Soc. Amer.
Bull., 55: 767-754.

and Mooser, 0. 1963. A porcupine from the Pleistocene of

Aguascalientes , Mexico. Contribs. Mus. Paleont., U. Mich.
XVIII (16): 245-250.

and Rinker, G. C. 1943. A new mouse (Microtus ochrogaster

taylori) from Meade County, Kansas. U. Kansas Sci. Bull.,
XXIX, Part 11(4): 255-268.

and Zakrzewski, R. J. 1967. Phyletic trends in the late

Cenozoic microtine Ophiomys gen. nov., from Idao. Contribs.
Museum Paleontol. U. Mich., XXI(12): 255-271.

Hinton, M. A. C. 1926. Monograph of the Voles and lemmings (Microtinus) ,
Br. Mus. Nat. Hist. I: 1-488.

Hirschfeld, S. E. and Webb., S. D. 1968. Plio-Pleistocene megalonychid
sloths of North America. Bull. Florida State Museum,
12(5): 213-296.

=178-

Hooijer, D. A. 1967. Pleistocene vertebrates of the Netherlands Antilles,
in Pleistocene Extinctions, P. S. Martin and H. E. Wright,
Jr., eds., Yale Univ. Press, New Haven: 399-406.
Holman, J. A. 1959. Birds and mammals from the Pleistocene of Williston,

Florida. Bull. Florida State Museum, 5(1): 1-25.
Hooper, E. T. 1952. A systematic review of the harvest mice (genus

Reithrodontomys) of Latin America. Misc. Pubs. Museum

Zool. U. Mich., 77: 1-255.
. 1957. Dental patterns in mice of the genus Peromyscus .

Misc. Pubs. Museum Zool. U. Mich., 99: 1-54.
Ivey, R. D. 1947. The mammals, excluding bats, of Palm Valley, Florida.

M. Sci. thesis, Univ. Florida, Gainesville.
James, G. T. 1957. An edentate from the Pleistocene of Texas. J.

Paleont., 49(1): 134-319.
Kowalskl, K. 1960. An early Pleistocene fauna of small mammals from

Kamyk (Poland). Folia Quaternaria, 1 Krakow: 1-24.
Kurten, B. 1965. The Pleistocene Felidae of Florida. Bull. Florida

State Museum, 9(6): 215-273.
. 1967. Pleistocene bears of North America. 2. Genus

Arctodus, short-faced bears. Acta Zool. Fennica, 117: 1-60.
. 1968. Pleistocene Mammals of Europe. Aldine Publ. Co.,

Chicago, 317 pp.
Layne, J. N. 1963. A study of the parasites of the Florida mouse,

Peromyscus f loridanus , in relation to host and environ-
mental factors. Tulane Stud. Zool., 11(1): 1-27.

-179-

Leidy, J. 1889. a. Fossil vertebrates from Florida. Proc. Acad.

Nat. Sci. Phil.: 96-97.
. 1889. b. Description of vertebrate remains from Peace

Creek, Florida. Trans. Wagner Free Institute Sci., 2: 19-31.
Ligon, J. D. 1965. A Pleistocene avifauna from Haile, Florida. Bull.

Florida State Museum, 10(4): 127-158.
Lundelius, E. L. , Jr. 1960. Mylohyus nasutus: long-nosed peccary of

the Texas Pleistocene. Bull. Texas Memor. Museum, 1: 1-40.
Martin, R. A. 1967. A comparison of two mandibular dimensions in

Peromyscus , with regard to identification of Pleistocene

Peromyscus from Florida. Tulane Stud. Zool., 14(2): 75-79.
. 1968. a. Late Pleistocene distribution of Microtus

pennsylvanicus . J. Mammal., 49(2): 265-271.
. 1968. b. Further study of the Friesenhahn Cave Pero-
myscus. Southwest. Natur. , 13(3): 253-266.
. 1969. a. Taxonomy of the giant Pleistocene beaver

Castoroides from Florida. J. Paleont. , in press.
. 1969. b. Extinction and immigration of mammals during

the late Pleistocene in north Florida. Paper presented

at the June, 1969 meetings of the American Society of

Mammalogists , New York, N. Y.
. Notes on dental variation in fossil and living Microtus ;

in prep.
and Webb, S. D. Pleistocene mammals from Devil's Den,

Levy County, Florida; in prep.
McCoy, J. J. 1963. The fossil avifauna of Ichetucknee River. Auk,

80: 335-351.

■180-

Merriam, J. C. 1912. The fauna of Rancho La Brea. Part II. Canidae.

Memoirs U. Calif. 1(2): 217-262.
Miller, G. M. , Jr. and Allen, G. M. 1928. The American bats of the

genera Myotis and Pi z onyx. U. S. Natl. Museum Bull.

144: 1-218.
Moore, J. C. 1956. Variation in the fox squirrel in Florida. Amer.

Midland Nat., 55(1): 41:65.
Neill, W. T. 1952. The spread of the armadillo in Florida. Ecology,

33(2): 282-284.
Osgood, W. H. 1909. Revision of the mice of the American genus Pero-

myscus . North Amer. Fauna, No. 28: 1-285.
Paulson, G. R. 1961. The mammals of the Cudahy Fauna. Papers Mich.

Acad. Sci., Arts, and Letters, XLVI: 127-153.
Pasa, A. 1947. I Mammiferi di alcune antiche brecce veronesi. Mem.

Mus. Stor. Nat. Verona, 1:1-111.
Pearson, P. G. 1951. Mammals of Gulf Hammock, Levy County, Florida.

M. Sci. thesis, Univ. Florida, Gainesville.
Pournelle, G. H. 1947. Mammals of a swamp region near Gainesville,

Florida. M. Sci. thesis, Univ. Florida, Gainesville.
Ray, C. E. 1958. Additions to the Pleistocene mammalian fauna from

Melbourne, Florida. Bull. Mus. Comp. Zool., 119(7): 421-449.
. 1967. Pleistocene mammals from Ladds , Bartow County,

Georgia. Bull. Ga. Acad. Sci., XXV (3): 120-150.
, Olsen, S. J., and Gut, H. J. 1963. Three mammals new to

the Pleistocene fauna of Florida, and a reconsideration of

five earlier records. J. Mammal., 44(3): 373-395.

•181-

Savage, D. E. 1951. Late Cenozoic vertebrates of the San Francisco Bay

region. U. Calif. Pubs. Bull. Dept. Geol.Sci., 28(10): 215-314.
Schultz, J. R. 1938. A late Quaternary mammal fauna from the tar seeps

of McKittrick, Calif. Cam. Inst. Wash. Publ., No. 487: 111-215.
Sellards, E. H. 1915. Chlamytherium septentrionalis , an endentate from

the Pleistocene of Florida. Amer . J.Sci.XL(4) : 139-145.
. 1916. Human remains and associated fossils from the

Pleistocene of Florida. Eighth Ann. Report Fla. State Geol.

Surv.: 121-160.
Semken, H. A., Jr. 1961. Fossil vertebrates from Longhorn Cavern, Burnet

County, Texas. Texas J. Sci., 12: 290-310.
. 1966. Stratigraphy and paleontology of the McPherson

Equus beds (Sandahl local fauna), McPherson County, Kansas.

Contribs. Museum Paleont. U. Mich., XX(6): 121-178.
j and Griggs, C. D. 1965. The long-nosed peccary, Mylohyus

nasutus , from McPherson County, Kansas. Papers Mich. Acad.

Sci., Arts, and Letters. L: 267-274.
Shotwell, J. A. 1956. Hemphillian mammalian assemblage from north-
eastern Oregon. Bull. Geol. Soc. Amer., 67: 717-738.
Simpson, G. G. 1928. Pleistocene mammals from a cave in Citrus County,

Florida. Amer. Museum Novitates, 328: 1-16.
. 1929. Pleistocene mammalian fauna of the Seminole Field,

Pinellas County, Florida. Bull .Amer .Mus .Nat. Hist. , 56(8):

561-599.
. 1930. a. Additions to the Pleistocene of Florida. Amer.

Museum Novitates., 406: 1-14.

»

-182-

. 1930. b. Holmesina septentrionalis , extinct giant arma-
dillo of Florida. Amer. Museum Novitates, 442: 1-10.
. 1941. Large Pleistocene felines of North America. Amer.

Museum Novitates, 1136: 1-27.
. 1949. A fossil deposit in a cave in St. Louis. Amer.

Museum Novitates, 1408: 1-46.
. 1961. Principles of Animal Taxonomy. Columbia U.

Press, New York; 247 pp.
Slaughter, B. H. 1961. The significance of Dasypus bellus (Simpson)

in Pleistocene local faunas. Texas J. Sci., XIII(3):

311-315.
. 1962. An ecological interpretation of the Brown Sand

Wedge local fauna, Blackwater Draw, New Mexico: and a

hypothesis concerning late Pleistocene extinction, in

Paleoecology of the Llano Estacade; F. Wendorf and J. J.

Hester, eds., 2: 1-37.
. 1966. Platygonus compressus and associated fauna from the

Laubach Cave of Texas. Amer. Midland Nat., 75(2): 475-494.
and Hoover, B. R. 1963. Sulphur River Formation and the

Pleistocene mammals of the Ben Franklin Local Fauna.

J. Grad Research Center, 31(3): 132-148.
and McClure, W. L. 1965. The Sims Bayou Local Fauna:

Pleistocene of Houston, Texas. Texas J. Sci., XVII (4):

404-417.
and Ritchie, R. 1963. The Pleistocene mammals of the

Clear Creek local fauna, Denton County, Texas. J.

Grad. Research Center, 31(3): 117-131.

■183-

Stevens, M. S. 1965. A new species of Urocyon from the upper Pliocene

of Kansas. J. Mammal., 46(2): 265-269.
Stock, C. 1930. (Sixth printing, 1968). Rancho La Brea a record of

Pleistocene life in California. Los Angeles County Museum

Nat. Hist., Science Series No. 20, Paleont.No.il: 1-83.
Strain, W. S. 1966. Blancan mammalian fauna and Pleistocene forma-
tions, Hudspeth County, Texas. Bull. Texas Memor. Museum,

No. 10: 1-55.
Webb, S. D. and Tessman, N. 1968. A Pliocene vertebrate fauna from low

elevation in Manatee County, Florida. Amer. J. Sci.,

266: 777-811.
Weigel, R. D. 1962. Fossil vertebrates of Vero, Florida. Fla. Geol.

Surv., Special Pub. No. 10: 1-59.
White, J. A. 1968. A new porcupine from the middle Pleistocene of the

Anza-Borrego Desert of California. Contribs. in Sci.,

Los Angeles County Museum, No. 136: 1-15.
Whitmore, F. C. , Jr. and Foster, H. L. 1967. Panthera atrox (Mammalia:

Felidae) from central Alaska. J. Paleontol. , 41(1):

247-251.
Young, S. P. and Goldman, E. A. 1964. The Wolves of North America.

Dover Pubic, Inc., New York, 632 pp.
Zeuner, F. E. 1959. The Pleistocene Period. Hutchison & Co., London,

447 pp.

BIOGRAPHICAL SKETCH
Robert A. Martin was born in New York City on February 19, 1944.
He lived on Long Island in the town of Westbury until 1965, graduating
from W. Tresper Clarke High School and obtaining a Bachelor of Arts
degree in Biology from Hofstra University. From there he journeyed to
New Orleans, and earned a Master of Science degree in Biology from
Tulane University in 1967. Since that time he has been working towards
a Doctor of Philosophy degree in Zoology at the University of Florida.

•184-

This dissertation was prepared under the direction of the chair-
man of the candidate's supervisory committee and has been approved by
all members of that committee. It was submitted to the Dean of the
College of Arts and Sciences and to the Graduate Council, and was ap-
proved as partial fulfillment of the requirements for the degree of
Doctor of Philosophy.

August, 1969.

Dean, Collegf'of Art/s and Sciences

Dean, Graduate School

Supervisory Committee:

4&h :

■ fartUm**

UNIVERSITY OF FLORIDA

3 1262 08666 444 7