3.“: Z.
:2 .0,

I
96...}?!
.1: :5

5

‘gr 5

L353? 2.
.22.!
I...

.
: rllra
8.31:3

.
a h.“
. :3. r..........

kn. V

cl) it
5....

; xiv. . ,
'1'}: «u. 75?: 2.1.1
.01.: (rs. : uh

. 3.1!?

u

5:,
a 113......
5‘le lbxrsa.

‘Iavatt .
.23.! I 2

7r“ .LLvr El.

.

. s f 9?
iii...

=4

5...)...

A. 21.: Y:
I... I .. 2:

EF

.. .c? .
1.. .58.;

3.1..

0.1-3.3.

\ .

1!! I

l\
(H
II

‘1
2’...

v.3.h‘
’1‘.
I .5“:
.
5.2:.
i}: y»:

.mi

ti

.2... ‘.
x. ..
f :3
\rl

0.417,:
w. Illv.’.\..
I tilts.

1...} 3; fl...

 

an! - )fi-ftt.

ICHIGAN STATE UNIVERSITY LIBRARIES

III III lllllllllllll\llll\Illlllllllllll Ill

3 1293 01020 1683

This is to certify that the

thesis entitled

CARBON AND NITROGEN ISOTOPIC EVIDENCE FOR TERTIARY
GRASSLAND DISTRIBUTIONS AND THE EVOLUTION OF HYPSODONTY
IN NORTH AMERICAN GREAT PLAINS HORSES (32 MA TO RECENT)

presented by

Shawn Gilbert Clouthier

has been accepted towards fulfillment
of the requirements for

Masters . Geological Sciences
degree 1n ‘

 

 

// W

/ 6%ressor

 

11/16/94
Date

 

0-7639 MS U is an Affirmative Action/Equal Opportunity Institution

___ i__ iA —

 

 

LIBRARY
Michigan State
University

 

 

 

PLACE N RETURN BOXtorunovoml-ehodromm ywrrocord. '
TO AVOID FINES man on or baton duo duo.

DATE DUE ‘. DATE DUE DATE DUE

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

M80 in An Affirmative Action/Equal Opponunlty Im
WM!

CARBON AND NITROGEN ISOTOPIC EVIDENCE FOR TERTIARY
GRASSLAND DISTRIBUTIONS AND THE EVOLUTION OF HYPSODONTY IN

NORTH AMERICAN GREAT PLAINS HORSES (32 MA TO RECENT)

By

Shawn Gilbert Clouthier

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE

Department of Geological Sciences

1994

ABSTRACT

CARBON AND NITROGEN ISOTOPIC EVIDENCE FOR TERTIARY
GRASSLAND DISTRIBUTIONS AND THE EVOLUTION OF HYPSODONTY IN
NORTH AMERICAN GREAT PLAINS HORSES (32 MA TO RECENT)

By

Shawn Gilbert Clouthier

Isotopic analysis of indigenous organic matter from a geographically and
temporally constrained series of fossil horse teeth is presented. Carbon and nitrogen
isotope and amino acid analyses of modern and ancient horses serve to assess natural
isotopic variation in skeletal remains and verify authenticity of ancient organic matter.
Carbon isotope values (613C) show the relative amounts of C3 and C4 plants in ancient
horse diets which is used to assess Tertiary grassland distributions and the advent of
high-crowned horse teeth. Organic 813C data reveal that C4 grasses comprised 20-30%
of the diet of Great Plains horses in the middle Miocene (15.5 Ma to 12 Ma). This is
in contrast to inorganic BBC studies of horse enamel carbonate which showed a C1
proliferation 7 Ma to 5 Ma related to decreased global CO2 levels. Since horses with
C4 dominate diets were localized to arid regions after 7 Ma and never inhabited the
Great Plains from 32 Ma to 3.3 Ma, the FT data does not support a late Miocene drop
in global C02. The origin of C4 grass in the mid-Miocene postdates the evolution of
hypsodonty in equids by at least 2 Ma. Consequently, a C4 radiation cannot be the
primary cause for the origin of high-crowned horse teeth. The expansion of C3 grasses

and abrasive grazing substrates are likely causes for the evolution of horse hypsodonty.

Dedicated to my grandmothers, Betty Clouthier and Mildred Romanack;

I will forever cherish our abridged time together.

iii

ACKNOWLEDGEMENTS

This study was partially funded by a Michigan State University Fellowship (to
Shawn G. Clouthier), by a grant from the GE Foundation (to Peggy H. Ostrom), and
by a Theodore Roosevelt Memorial Fund Grant of the American Museum of Natural
History (to Shawn G. Clouthier). I recognize Dr. Richard Tedford and J. Daniel
Bryant of the American Museum for their valued support. I owe a special debt of
gratitude to these individuals as the vast majority of fossil specimens in this thesis
came from the AMNH. Were it not for the continued assistance of Dan "SuperDan"
Bryant, long after the selection of thesis samples, this project would have taken much
longer to complete. I gratefully acknowledge Dr. Michael Voorhies at the University
of Nebraska State Museum and Dr. Everett Lindsay at the University of Arizona for
supplying additional fossil horse material. I extend my appreciation to Patrick M.
O'Connor for his help in generating my initial isotope data and his insightful
comments at the inception of my thesis.

I wish to thank my thesis committee members for their assistance in stimulating
research ideas and for editing the multiple drafts of this project. My advisor, Dr. J.
Alan Holman, along with committee members Dr. Peggy H. Ostrom and Grahame J.

Larson were helpful with their comments, constructive criticisms and encouragement.

iv

I am particularly thankful to Drs. Holman and Ostrom for their assistance and
personnel counsel throughout this Master's degree.

I would like to recognize the educators who were most instrumental in shaping
my academic and scientific growth and development: Jim Baker, my fifth grade
science teacher for farming the flame; Dr. Francis 8. Collins, the first to allow me to
express my scientific abilities in a laboratory setting and whom will always serve as an
inspiration to me; Stephen J. Gould, who has given me countless hours of intellectual
enjoyment through his writings; and all the teachers and professors, too numerous to
mention, who took a special interest in me over the years.

Though the aforementioned individuals are all deserving of my deepest
gratitude, I am especially thankful to my family as they are most responsible for the
success I have attained. Although it is not possible to individually cite each family
member deserving of recognition, I would like to extend my most heartfelt
appreciation to all the members of my immediate and extended family. Specifically, I
wish to thank: Roger, Pam and Jennifer Jennings for their interest and support
throughout the past six years; Mark Romanack, my uncle, for instilling in me a love of
nature and for playing baseball with me even while he was on crutches; Fran Roberts,
my aunt, for believing in me and for whom I have enormous admiration as an
educator. I wholeheartedly express my affection and gratitude to my cousins Bret and
Tracy Gorsline for co-authoring our shared, derived creativity and forthe multitude of
3:00 am. philosophical and scientific conversations we have enjoyed together. I

acknowledge my aunt, Francine Gorsline, for her love and friendship and for taking

care of me as one of her own all those weekends and summers throughout my
chfldhood.

I offer my most sincere appreciation to my sister, Kristy Lynn Clouthier and
her fiance and my best friend Jim Horn, for allowing me to unwind and be myself.
The same holds true for, Pebbles and Tricki, for putting up with my teasing and kitty
hugs over the years. To my parents, Gilbert and Michele Clouthier, whose love,
guidance, humor, strength, care and encouragement has made me who I am today, I
am forever indebted. Lastly and most deservedly, I thank my beautiful and loving
wife of two years, Jill Christine Clouthier, and my five year old feline daughter,
Commodore Pickelodeon Clouthier, for their unwavering support, empassioned

understanding, undying love and conviction of the heart.

With Love,

Shawn

vi

TABLE OF CONTENTS

LIST OF TABLES ............................................................................................................. xi
LIST OF FIGURES .......................................................................................................... xii
KEY TO ABBREVIATIONS AND SYMBOLS .......................................................... xiv
INTRODUCTION ............................................................................................................... 1
Stable Isotopes ......................................................................................................... 6
Carbon Isotopes- Dietary Sources ............................................................. 7

Nitrogen Isotopes- Trophic Effects ......................................................... lO

Diagenesis and Indigeneity ................................................................................... 12
Previous Studies ........................................................................................ 14

Current Research ....................................................................................... 15

Tertiary Paleoecology ............................................................................................ 17

The Miocene World .................................................................................. 18

The Pliocene Epoch .................................................................................. l9

Grasses and Hypsodonty .......................................................................... 20
MATERIALS AND METHODS ..................................................................................... 24
Sample Procurement ................................................................................. 24

Specimen Documentation ......................................................................... 24

vii

Locality and Age Descriptions .............................................................. 28

Preservational States ............................................................................... 3O
Preparatory Methods ............................................................................................. 32
Tooth and Bone Cleaning and Preparation ......................................... 32
Hydrochloric Acid Demineralization Technique ................................ 34
Dialysis and Lyophilization Procedure ................................................ 34
Homogenation Technique ..................................................................... 35

Stable Isotope Analyses ...................................................................................... 35
Quartz Tube Evacuation and Combustion ........................................... 36
Cryogenic Gas Line Separations .......................................................... 36

Dual Inlet Mass Spectrometry .............................................................. 37

Carlo Erba Method ................................................................................ 37

Amino Acid Analyses ........................................................................................ 38
Organic Matter Hydrolysis .................................................................... 39
Column Chormatographic Purifications .............................................. 39
Derivitization Procedurel ...................................................................... 40
Enantiomer Assessment ........................................................................ 41
RESULTS AND DISCUSSION ...................................................................................... 42
Yield and Recovery Experiments ..................................................................... 42
Demineralization and Lyophilization Yields ....................................... 42

High Molecular Organic Matter Yields ............................................... 42
Percent Carbon Yields ........................................................................... 46

viii

Amino Acid Recovery Experiment ........................................................ 48

Amino Acids in Fossil Horse Tooth and Bone ..................................... 51
Implications for Indigeneity Assessment of Fossils ............................. 51
Isotope Studies of Modern and Ancient Herbivores ............................................ 56
Procedural Reproducibilty ...................................................................... 57

Testing for Procedural Errors ................................................................. 57

Isotope Variation Within Modern Materials ......................................... 59
Intraspecific Variance in a Juvenile Wapiti .......................................... 59
Interspecifrc Variance Between Wapiti and Horses .............. 61

Variation Between Different Horses ....................................... 63

Variability in Horse Tooth and Bone ..................................... 66

Isotope Variation in Ancient Materials ................................................ 67
Comparisons of Modern and Ancient Isotope Values ......... 69

Evaluation of Diagenetic Effects ........................................... 71

Interpretations of Tertiary Paleoecology of the Horse Family ......................... 72
Percent C4 in Horse Diets ...................................................... 72

Temporal and Spatial Distribution of Isotope Data ............ 76

Isotope Changes Over Time .................................................. 79

Comparisons with Other Isotope Studies ............................. 81

Reinterpreting Previous Studies ............................................ 82

Paleoecological Ramifications of Equid Isotope Data ........ 86

CONCLUSIONS .............................................................................................................. 92

ix

Table 1

Table 2

Table 3

Table 4

Table 5

Table 6

Table 7

Table 8

Table 9

LIST OF TABLFS

Taxonomy and relevant information for modern horses and wapiti ............. 26
Taxonomy and relevant information for ancient horses ................................. 27
Locality, age, and preservation information of horse fossils ......................... 31
High molecular weight organic matter yields obtained from modern

horses and wapiti ................................................................................................ 43
High molecular weight organic matter yields obtained from ancient

horse tooth and bone samples ........................................................................... 44
High molecular weight organic matter yields obtained from a two hour
demineralization experiment .............................................................................. 45
Percent organic carbon yields from select modern and fossil samples ........ 47
Mean peak areas for three amino acids, before column purification

and after column purification, with varying initial concentrations ............... 49
Mean percent recoveries of aspartic acid, glutamic acid and serine

from ancient horse organic matter following column purification ............... 50

Table 10 Amino acid distributions and abundances for three fossils, 4.0 Ma,

5.5 Ma and 10.0 Ma in relation to the amino acid pattern of modern

cow bone ............................................................................................................ 52
Table 11 The 5'3C and EN values of modern wapiti and horses ............................... 58
Table 12 The 6‘3C and EN values of ancient horse teeth and bones ......................... 68

Table 13 Percent C4 vegetation in the diet of modern and ancient terrestrial

herbivores assuming four separate fractionations .......................................... 74

xi

LIST OF FIGURES

Figure l Phylogeny of the family Equidae throughout its 58 million year
evolutionary history ............................................................................................. 2

Figure 2 Horse dental evolution showing the transition from the relatively
undifferentiated brachyodont condition to the more complex enamel
foldings of the hypsodont equids ....................................................................... 4

Figure 3 Arrangement of the various cell layers in C3 and C4 plants ......................... 23

Figure 4 Modern horse skull and teeth shown in cross-section to
illustrate the pronounced hypsodont condition ............................................... 25

Figure 5 Labial view and occlusal surface of LP4-LM1 of a 15.5 Ma
merychippine horse, Mervchippus §p_. (cf. isonesus) ..................................... 29

Figure 6 Labial view and occlusal surface of RM3 of Cormohippflon
occidentale, 8.5 Ma, from Ellis County, Oklahoma ...................................... 29

Figure 7 Cross-sectional view showing the relationship of the layers
of a molar tooth from an extant hypsodont horse, Eguus, and a

low-crowned molar from an extinct horse, Parahippus ................................ 33
Figure 8 Amino acid chromatogram for a 40 Ma fossil tooth ................................... 53
Figure 9 Amino acid chromatogram for a 5.5 Ma fossil tooth ................................... 54
Figure 10 Amino acid chromatogram for a 10.0 Ma fossil tooth ................................ 55
Figure 11 Natural variation in the 5'3C and 8N values of a juvenile wapiti ........... 60

Figure 12 The relationship between the 5'3C and 5'5N values of skeletal
tissues (premolar and molar teeth and maxilla bone) from a
modern horse ................................................................................................... 64

xii

Figure 13 Ranges of 613C and 8N values of organic matter from a majority
of the published modern terrestrial herbivores including the
ancient horses analyzed in this study ............................................................ 70

Figure 14 The percent distributions of modern C4 plants in various regions
of North America ............................................................................................ 77

Figure 15 Carbon isotope values of ancient North American horses
(32.0 Ma to 3.0 Ma) in association with age and geographical

information ...................................................................................................... 78

Figure 16 The relationship between 6'3C and 8N values of ancient horse
organic matter and geologic age of the specimens ..................................... 80

Figure 17 Geographical location, age and 513C values of ancient horse
enamel carbonate ............................................................................................ 83

xii

KEY TO ABBREVIATIONS AND SYMBOLS

AMNH= American Museum of Natural History, New York.
MSU= Michigan State University, East Lansing.

SMU= Southern Methodist University, Dallas.

UA= University of Arizona, Tucson.

UCR= University of California, Riverside.

UF= University of Florida, Gainsville.

UMNH= Utah Museum of Natural History, Salt Lake City.
UNSM= University of Nebraska State Museum, Lincoln.
F:AM= Frick Collection, American Museum.

HPLC= High Performance (Pressure) Liquid Chromatography.
GC= Gas Chromatograph.

MS= Mass Spectrometer.

MSD= Mass Selective Detector.

FID= Flame Ionization Detector

HMW= High Molecular Weight.

TFA= Trifluoroacetic Acid.

RIA= Radioimmunoassay.

ELISA= Enzyme Linked Immunosorbent Assay.
PCR= Polymerase Chain Reaction.

C3= Plant photosynthetic pathway (Calvin-Benson) utilizing three carbon
intermediates.
C4= Plant photosynthetic pathway (Hatch-Slack) utilizing four carbon intermediate
molecules.
CAM= Plant photosynthetic pathway (Crassulacean Acid Metabolism) which
alternates between the C3 and C4 cycles during daylight and night conditions.

RM1= Mammalian upper molar terminology (e.g. Right First Molar).

LdP4= Mammalian upper deciduous molar terminology (e.g. Left Fourth
Deciduous Premolar).

RMax= Right Maxilla Bone.

Ma= Megannum, temporal unit meaning millions of years ago.

NALMA= North American Land Mammal Age.

xiv

INTRODUCTION

The horse family (Equidae) spans 58 million years and is perhaps the most
well-known fossil assemblage in the world. Sequential modification in skeletal
morphologies have long been cited as evidence for a gradual, stepwise evolutionary
pattern for equids (Huxley, 1870; Kowalevsky, 1874; Stirton, 1947). Among other
evolutionary trends, the gradual development of hypsodont dentition (high-crowned
teeth) became dogmatic in paleontology. Today a purely anagenetic view of horse
evolution such as this is considered simplistic. The current phylogeny of the family
Equidae, including the origin and prevalence of hypsodonty, is depicted in Figure l.

The most common hypothesis for the morphological transition from
low-crowned (brachyodont) to high-crowned (hypsodont) teeth among the Equidae is
as an adaptation in response to the widespread radiation of siliceous savanna
grasslands in the Miocene (Stirton, 1947; Simpson, 1951; MacFadden et al., 1991).
Grasses with a high silica content are more abrasive than the fleshy leaves upon which
brachyodont horses browsed. Consequently, higher tooth crowns may have evolved as
the dietary percentage of silica-bearing grasses increased. Alternatively, equid
hypsodonty may have arisen, concurrent with the transition to a grazing lifestyle, in
response to the incorporation of grit characteristic of sandy savanna substrates (Stirton,

1947; Simpson, 1951). Evolutionary change in the grinding surface of horse molars

 

 

  

 

 

       
  

      
 

              
  

  
 

    
   
  
 

 

 

 

 

 

 

 

g PHYLOGENY OF THE EOUIDAE
W“ s. AMERICA I NORTH AMERICA | ow wonu)
0 l I O . .
.5 M . t t O O I .
T Hippition. Onohipp EQUUS .
- & Pamhippan‘on
5 -_°'_ [)1.\’()HIPPUS
55
S
" i
{/2 K
10 - b S E
Proto- E § 5
8 hlpp E n. E
s u
15" .
I MER r (‘HIPPUS
O C
20 _ ........
a —l
MIOHIPPUS
x .1
MESOHIPPUS
O
35 u-
Haplo-
_ mow
PALAEOTHERE
GROUP
‘5‘ a
3 [Z] Hypsodont Equids
Brachyodont Equids
50‘ Xanico-
hippos
55-
Figure 1 Phylogeny of the family Equidae throughout its 58 million year

evolutionary history. The capitalized genera signifies those used in this
study (modified from MacFadden, 1992).

3

over their phylogeny is represented in Figure 2. Both hypotheses concerning the
origin of hypsodonty in horses have been championed and criticized with no real
evidence upon which to substantiate either position. Recent advances in the field of
isotope geochemistry have made it possible to test these and other interesting
paleobiological questions. This study, for the first time, uses stable carbon and
nitrogen isotopes of the organic remnant from a time series of fossils to determine if
the spread of savanna grasslands influenced the diet of ancient horses and
consequently contributed to the advent of hypsodonty among the Equidae.
Geochemical analyses provide information about origins and transformations of
organic matter in the natural environment. Stable isotopes have proved particularly
effective for evaluating dietary components and trophic positions of organisms owing
to the close association between the isotope composition of an organism and its diet
(DeNiro and Epstein, 1978b; 1981; Harrigan et al., 1989). This relationship exists for
modern and ancient organisms (Vogel, 1978; Tieszen et al., 1979a; Comic and
Schwarcz, 1994). Provided unique geochemical signals exist, stable isotopes can be
used to investigate the paleoecologies of extinct organisms. The transition to a
savanna-based diet would provide a unique geochemical signal resulting from the
difference in carbon isotope values (5'3C) of C3 and C, vegatation (mean 5'3C = -27%o
and -13%o, for C3 and C4 plants, respectively). Savannas exist in regions of seasonal
contrasts usually related to water stress and are among the first ecosystems to utilize
the evolutionarily more recent C4 photosynthetic pathway. Given these conditions, the

importance of savanna grasslands in ancient horse diets should be recorded in the 8'3C

 

PLEISTOCENE

 

     

 

 

"a

 
  

gunman! JANNIPPUS Pyompeus
“raucomppus /}~‘ @/
hr
§ .
'13 “cavemen: MERYCHIPPUS
g tunvcmeeusr tenoromeeusl

. w
ANCH 1111:1111»; PAR'AHIPPUS E

 

 

 

 

Low-crowned Teeth
(Brachyodont)

High-crowned teeth
(Hypsodont)

 

' i
“I W :
§ 1

1

3 mesomeeus I
I I

I

I

Q .

‘i' enmeeus '
I“ I
g 1
‘11 I
@ "

1

comreus 1

I

1

I

1

1

J

 

 

Figure 2 Horse dental evolution showing the transition from the relatively
undifferentiated brachyodont condition to the more complex enamel
foldings of the hypsodont equids (modified from Simpson, 1951).

5

values of fossil horse teeth. This information can then be used to evaluate the
traditional paleontological interpretations concerning the evolution of hypsodonty in
the family Equidae of North America.

Before attempting paleoecological reconstructions, indigeneity of the fossil
organic matter must be established. When working with ancient organic material, the
possibility of postmortem and/or modern contamination must be addressed.

Numerous attempts to establish an objective indigeneity criterion have been proposed
(DeNiro, 1985; Macko and Engel, 1991; Ostrom et al., 1994). Abundances and
distributions of amino acids and isotope values of the ancient organic material isolated
from tooth and bone serve as indicators of indigenity. Diagenetic effects such as age
and depositional environment may alter elemental compositions and isotope values of
fossils (DeNiro, 1985). Therefore, this study evaluates the effects of diagenesis by
comparing the amino acid signatures and isotope values of ancient and modern
terrestrial herbivores. This thesis reports HPLC results concerning the reproducibility
and recovery of amino acids subsequent to ion exchange chromatographic purification.
In addition, gas chromatographic results detailing the distribution and relative
abundance of amino acids recovered from equid fossils are used to assess indigeneity.
Comparisons of ancient and modem isotope values designed to investigate indigeneity
require knowledge of the isotope variability of the species being studied. Very little
isotope data currently exists for modern and fossil horses and for large herbivores in
general. Thus, another aim of this research is to enhance the current data on the

natural isotopic variation in large terrestrial herbivores. This knowledge will provide a

6

framework which will facilitate the evaluation of postmortem alteration effects.
Distinguishing diagenetic alteration from natural isotopic variation is best
accomplished using a well-known spatially and temporally constrained series of fossils.
The family Equidae has a remarkably complete fossil record. For this reason, and for
the intruiging paleoecological questions concerning the origin of hypsodonty, a time

series a horse teeth and bones (32 Ma to Recent) was selected for this isotope study.

Stable Isotopes

Applications of stable isotope mass spectrometry in the field of vertebrate
paleontology are of recent origin. Nonetheless, ratios of naturally occurring isotopes,
specifically carbon (”C and 13C) and nitrogen (”N and ‘5N), enhance investigations on
prehistoric vertebrates including assessments of ancient biochemistries (Ostrom and
Fry, 1993), interpretations of paleo-community structures (Bocherens et al., 1994;
Ostrom et al, 1993) and paleoclimate analyses using vertebrate remains (Thackeray and

Lee-Thorp, 1992). By convention, isotope values are given by the following equation:

(RSampIe)
8"Y = * 103

 

(RStandard -1 )

where Y is the element of interest, x is the trace isotope and R is ”C/‘ZC, 'SN/“N, or

7

the ratio of any trace to abundant isotope of a particular element. The units for this
equation are permil (%o). Isotope values are reported relative to international
standards, Pee Dee Belemnite (Belemintella americana) for carbon and atmospheric N2

for nitrogen.

Carbon Isotopes- Dietwy Sources

Carbon isotopes have proved useful in elucidating sources of dietary carbon in
a wide variety of modern as well as fossil animals (DeNiro and Epstein, 1978b;
Tieszen et al., 1979b; Lee-Thorp and Van der Merwe, 1987; Bocherens et al., 1990).
These studies have shown that the (WC values of consumers tissues are similar to the
food consumed. Laboratory experiments demonstrate the 5'3C values of the organic
fraction of small mammalian tissues differ from the diet by about 1%o to 39611 (DeNiro
and Epstein, 1978b). Large terrestrial herbivores (e.g. elephants) commonly exhibit up
to a 6960 fractionation between the 513C values of skeletal tissues and dietary
constituents (Vogel, 1978; Vogel, 1990a). Fractionation in horses likely falls between
these two size extremes. Variation in carbon isotope values are a result of kinetic
fractionation during metabolic processes (Silfer et al., 1992). Kinetic fractionations
also account for the large differences in 5'3C values of C3 and C4 vegetation.
Knowlege of the carbon isotopic signatures of C3 and C4 plants becomes important
when attempting to reconstruct horse paleodiets.

The two priniciple plant photosynthetic pathways (C3 and C,, respectively) have

8

5‘3C values which form a nearly continuous, non-overlapping sequence. Terrestrial C3
plant 513C values range from -24%o to -34%o and average -27%o (Smith and Epstein,
1971; Galimov, 1985). Assuming a 4%o fractionation between bone and diet, horses
with pure C3 diets would have 513C values between -20%o and -30%o. Plants which
utilize the C, biochemistry are enriched in 13C relative to C3 plants. The 5'3C values
of C, plants range from -5%o to -l9%o with an average 5'3C value of about -13%o
(Smith and Epstein, 1971; Tieszen, 1978; Galimov, 1985; Keegan, 1989). Assuming
the same 4%o fractionation, horses consuming only C, vegetation would have 5'3C
values between -l%o and -15%o. A third plant biochemical pathway exists, the CAM
cycle, which in many cases is isotopically indistinguishable from C, photosynthesis.
However, CAM photosynthesis is only found in succulents and cacti which are not
likely to represent an abundant food source for horses on the Great Plains. Thus,
carbon isotope analyses allow for discrimination between C3 and C, based diets of
extinct horse species.

Modern savanna grasslands (e.g. the Serengeti plains of Africa) are
characterized by C, grasses which have adapted to avoid the problem of
photorespiration by utilizing an intermediate four-carbon molecule such as malate
during photosynthesis. This adaptation allows C, plants to sequester carbon dioxide
inside their bundle shealth cells which provides a selective advantage over their less
sophisticated C3 relatives in arid climates where conditions regularly exceed 28°C
(82°F). Other conditions which favor the C, metabolic pathway, aside from low CO2

availability and high temperatures, are high light levels, high salinity and low moisture

9

levels. In summer months these conditions are regularily attained in modern savannas.
More than half the species of known grasses utilize the C, pathway (Downton, 1975).
These are the so-called warm season (tropical) grasses as they are uncommon in
temperate regions where daily minimum summer temperatures drops below 10°C
(50°F) (Teeri and Stowe, 1976). Other C, plants include maize, sorghum, and
sugarcane. Only a single C, tree is known, the euphorbia tree of Hawaii (James A.
Teeri, personal communication). In contrast, all other deciduous and coniferous trees
and shrubs utilize the C3 pathway. Wheat, soy, oats, clover, rice, nuts, most fruits and
vegetables, and the cool season (temperate) grasses are all examples of C3 plants.
Determining the abundance of C, plants in the diet through isotope studies of
organic materials is dependent upon an understanding of the initial 5'3C values of the
individual dietary components (in this case, C3 and C, vegetation). In addition, the
turnover rate of carbon and the observed isotopic discrimination between the herbivore
and its diet must be evaluated. Isotope values of C3 and C, plants are well
characterized. The average carbon isotope values of extant C3 and C, vegetation
(-27%o and -13%o, respectively) can be used to predict the percentage of C, plants in
the diets of ancient horses provided carbon turnover rates are understood. Turnover
rates of many herbivore carbon reservoirs such as milk, fat, muscle and hair are rapid
(days to months, in most cases) (Teiszen et al., 1983; Boutton et al., 1988). Collagen,
having an extremely long turnover rate, may record yearly or longer accumulations of
carbon in large herbivores with slower metabolisms (Teiszen et al., 1983). This

knowledge of turnover rates in large herbivores is crucial since climate variaitions may

10

alter plant distributions causing seasonal differences in herbivore tissues rather than
actual dietary changes. Previous attempts have been made to assess the percent
contribution of C, plants using mass balance models (Vogel, 1978; Vogel et al., 1986;
Nordt et al., 1994; Connie and Schwarcz, 1994). This thesis reports percent C,

contributions in ancient horse diets computed as follows:

5%,, - 513cc,
% C, = * 10"-

513cc4 - 513cc3

The 8'3CC3 and 5'3CC, terms are the mean values for modern C3 and C, plants.
The 6'3CD is the isotope value of the horses diet which is determined from the 5'3C
value of the ancient horse in question taking into account the observed discrimination
between horses and their diet. To insure accuracy when determining the percent
contribution of C, plants in the diet of horses, all of these variables must be well

characterized.

Nitrogen Isotopes- Trophic Effects

Nitrogen isotope values (BUN) are useful ecological indicators. The 5'5N

values of skeletal tissues from consumers are ~3%o enriched in 5‘5N over the diet

ll

(DeNiro and Epstein, 1981; Macko et al, 1987). This observation makes it possible to
use nitrogen isotopes as trophic level indicators in modern and ancient ecosystems.
The usefulness of nitrogen isotopes in paleontological studies has recently been
demonstrated by the ability to delineate trophic level structure in Cretaceous
communities (Ostrom et al., 1993) and paleophysiologies in Pleistocene bears
(Bocherens et al., 1994).

Ancient trophic reconstructions using nitrogen isotopes rely on a knowledge of
source and cycling of nitrogen in the modern environment. Most terrestrial plants
which obtain nitrogen from the soil have 5‘5N values between 3%o and 69611, but may
be as high as 9%o (Delwiche and Steyn, 1970; DeNiro and Hastorf, 1985). Nitrogen
fixing plants such as legumes (e.g. soy, peas, clover) typically have O'SN values closer
to that of atmospheric nitrogen (0.0%»). Low 5'5N values for terrestrial herbivores
may indicate the influence of legumes or other nitrogen fixing plants in the diet.
Terrestrial herbivores with pure non-leguminous diets are expected to have 5‘5N values
between 6960 and 9%o. Terrestrial carnivores should range from 9%o to 12%;.

The nitrogen cycle in marine communities is quite different from that of the
terrestrial realm. Marine plants are up to 4%o enriched in 15N relative to terrestrial
plants (Ambrose and DeNiro, 1989). Anamolously high values from terrestrial regions
are often observed in modern and ancient communities (DeNiro and Hastorf, 1985;
Vogel et al., 1990a). These apparent discrepancies have been interpreted as climatic
effects (Heaton et al., 1986), diagenetic alteration (DeNiro and Hastorf, 1985), or

marine influences (Ambrose and DeNiro, 1989). Since it is unlikely fossil equids

12

consumed marine plants, this explanation is untenable for this research. However,
climatic effects and diagenesis are not so easily dismissed. Aridity potentially controls
S’SN values owing to differential nitrogen metabolism by animals during water stress
(Heaton et al., 1986; Vogel et al., 1990b). Soil nitrogen contamination and extensive
groundwater throughput represent possible modes of diagenetic isotope alteration.
Whether these diagenetic agents would serve to enrich or deplete isotope values of the
original organic nitrogen is unclear. The role that biomineralization reactions have in
influencing the preservation potential of organic matter in fossils is only beginning to
be studied (Masters, 1987). Preliminary research investigating fractionation associated
with diagenetic reactions exists (Silfer et al., 1992; Qian et al., 1993), but little is
currently known about this very important topic. For these reasons, a more in-depth
consideration of how diagenesis is thought to proceed and how indigenous fossil

molecules are identified is presented.

Diagenesis and Indigeneity

Underlying stable isotopic analyses of fossils from similar depositional
environments and of similar age is the notion that the original isotopic signature
persists unchanged indefinitely or is shifted by a consistent, detectable amount (Peggy
H. Ostrom, persOnal communication). However, this assumption is invalid when
dealing with fossils of different ages from different depositional settings owing to the

exceptionally labile nature of most biological molecules. Post-depositional change

l3

differentially alters the retention capabilities of organic molecules owing to newly
established affinities with the mineral phase (Masters, 1987). Therefore, it is
important to investigate diagenetic pathways of change if indigenous organic
components from fossils are to be accurately identified.

This thesis addresses the concerns of diagenesis using carbon and nitrogen
isotopes and amino acid analyses. Carbon and nitrogen isotope values of organic
matter from ancient horses that are consistent with traditional paleontological
interpretations and with trends observed in modern horses indicate a lack of diagenetic
alteration. Alternatively, the extent to which 6'3C and 8N values of the horse fossils
deviate from trends that characterize modern horses may be useful for assessing
degrees of diagenetic alteration once natural isotopic variability is taken into account.
Published isotope values of terrestrial herbivores with similar diets to horses serve as
proxies used to define the natural isotope variation within horses.

Amino acid analyses are used to verify indigeneity of ancient materials as well.
Fossil amino acid patterns that are consistent with modern amino acid signatures (e. g.
presence of hydroxyproline and high concentrations of glycine and proline) likely
retain an indigenous organic fraction. Elevated concentrations of the acidic amino
acids (aspartic acid and glutamic acid) are consistent with the observation that acidic
macromolecules with a high affinity for the mineral phase are preferentially retained
during diagenesis (Hare, 1980). Fossils which are found to have little or no glycine,
proline and/or hydroxproline and have high concentrations of aspartic acid and

glutamic acid are likely to have experienced severe diagenetic alteration. These fossils

14

are poor candidates for paleoecological investigations.

Previous Studies

Abelson (1955) detected the presence of amino acids in fossil hydrolysates.
Shortly thereafter methods were devised to determine whether or not amino acids
isolated from fossils were authentic. One of the original methods for verification of
the authenticity of organic material isolated from fossils consists of assessing amino
acid enantiomer ratios. In solution, amino acids (except for glycine) exist in one of
two stereoisomers, the L-optical isomer and the D-optical isomer (Corrigan, 1969). In
living organisms the vast majority of organic molecules are formed from L-
configuration amino acids. Upon death, a racemization reaction proceeds toward
equilibrium of the D and L enantiomers. Previously, ancient materials which were
found to contain a significant L-amino acid component were assumed to represent non-
indigenous components. However, using this technique fails to recognize indigenous
material within fossils which have undergone alteration. Polymerization and
polycondensation reactions proceed under the conditions of early diagenesis (SO-100°C)
and result in indigenous HMW material with low D/L ratios (indicative of
contamination) (Serban et al., 1987; Ostrom, 1990). Consequently, amino acid
racemization data has to be dealt with carefully. The best approach for detecting
indigenous organic components within ancient remains is one which employs several

lines of evidence.

15

Current Research

An assessment of indigeneity of organic components within fossils is presently
based on the distribution and relative abundance of amino acids and various
immunological/biochemical assays in addition to amino acid racemization studies
(Armstrong et al., 1983; Macko and Engel, 1991; Lowenstein, 1988). Amino acid
hydrolysates from ancient materials which have been desalted on an ion exchange
resin can be detected by HPLC using a fluorometer with post-column introduction of
O-phenylaldehyde (OPA). This facilitates determination of the distribution and
relative abundances of the amino acids present. Another technique used to identify
amino acids in ancient materials involves the use of a GC. In this technique a liquid
sample is combusted to a gaseous state then separated under helium flow prior to
introduction to a detector (such as a FID or a MSD). Provided the amino acids are
derivitized beforehand, a GC can be used to verify molecular identities. Derivitization
is required in order to make the amino acids more volatile during the GC combustion
step. Derivitization is complicated when dealing with fossil organic materials.
Hydrolysis of ancient organic matter results in low molecular weight (1,000-6,000mw)
peptides and salts which interfere with the derivitization step. This situation is
avoided by performing a column purification of the amino acid extract prior to
derivitization. Of considerable concern is the possible loss of amino acids during the

desalting step.

l6

Immunological approaches such as RIA and ELISA involve raising antisera
against specific fossil proteins. These methods have been very successful for verifying
fossil organic matter authenticity and elucidating relationships among extinct taxa (e. g.
the Tasmanian wolf and the Quagga)(Lowenstein et al., 1981; Lowenstein and Ryder,
1985). In addition, DNA-DNA hybridization experiments and amplification and
sequencing of intact mitochondrial or genomic DNA within ancient materials by PCR
are becoming more popular indigeneity assays (Marshall, 1988; Piabo et al., 1989).

The use of C/N ratios of extracted "collagen" for verifying sample indigeneity
(with C/N= 2.9 to 3.6 being considered indigenous) has been proposed (DeNiro and
Epstein, 1981). However, the C/N ratio of polymerized, authentic fossil "collagen" is
frequently high. Thus, the C/N criterion may well exclude some indigenous
polymerized samples and might even include altered samples which coincidentally fall
within the acceptable 2.9 to 3.6 range. Therefore, the use of C/N ratios for assessing
indigeneity of organic matter from fossils is of limited use at present.

A recent approach for paleomolecular indigeneity involves analyzing the
isotopic signature of individual amino acid enantiomers (Silfer et al., 1991; Ostrom et
al., 1993). This approach, in association with the other indicators, could provide a
more objective criterion for assessing the indigeneity of fossil organic matter.
Although this study does not report isotope values on individual enantiomers, amino
acid distribution, abundance and racemization data is utilized in conjunction with
carbon and nitrogen isotope values of the HMW material isolated from fossils to

characterize ancient organic matter for indigeneity purposes. Once indigenous organic

l7

matter has been identified, interpretations concerning the paleoecology of extinct

organisms can be formulated.

T ertiary Paleoecology

Perrisodactyl (rhinocerotids, tapiroids, titanotherids, and equids) diversity
throughout the Tertiary has recently been correlated with climatic events and
vegetational changes (Janis, 1989). Interestingly, C, grasses are first known in
abundance from the time when equid diversity sharply declines. Hindgut fermenters
(such as horses) might have been inherently less capable of processing the expanding,
nutritionally poor C, grasses compared to ruminant artiodactyls. Researchers have
hypothesized the drop in equid diversity may relate to a C, diversification (Wang et
al., 1994). This hypothesis is more parsimonious than the once dominant view that
equids were ecologically out-competed by more efficient grazing artiodactyls. The
demise of the horse family was most likely in response to Tertiary vegetational
changes which fortuitously favored ruminant artiodactyls (Janis, 1989). High-crowned
teeth confer an adaptive advantage to individuals who regularily consume materials
harder than the apatite of enamel and dentine (e.g. plant opal, soil grit particles)(Jones
and Handreck, 1967; Walker et al., 1978). Thus, the expansion of abrasive grasslands
may well explain the advent of hypsodonty in horses without invoking interspecific
competition. This view does not distinguish between siliceous components of C3 and

C, grasses or the influence that gritty substrates may have played in the evolution of

18

hypsodonty. The best way to decipher this relationship is to assess fossil grass
originations and distributions using paleontological and stable isotopic evidence.

Variables used to determine modern grass distributions can be used to
reconstruct paleotemperature and climatic conditions prevalent during the time when
C, grasses first evolved. This information provides clues to assess whether or not
there is a correlation between the radiation of abrasive grasses and the evolution of
high-crowned teeth in horses. The use of stable carbon and nitrogen isotopes provides
a more objective way to verify the existence of the postulated grassland-hypsodonty

relationship in light of all the available evidence.

The Miocene World

Paleopedological (fossil soil) evidence suggests a transition from mostly
woodlands to shrublands (with some locally widespread grasslands) occurred during
the mid to late Oligocene (Retallack, 1983a; 1983b). In the early Miocene, vegetation
of the continental interior was comprised of a forest/grassland mosaic as evidenced by
the low endemism of grasses and the presence of a large browsing fauna (Axelrod,
1985). Paleoclimatological studies suggest the middle Miocene (18 Ma to 15 Ma) was
a tropical, high-moisture environment, the warmest interval of the Neogene (Zubakov
and Borzenkova, 1990). Concommitant with the warming trend, the middle Miocene
experienced the first occurrence of high levels of seasonality in both temperature and

precipitation which profoundly affected floral distributions (Janis, 1989). The Great

l9

Plains experienced annual precipitation double present levels during the mid-Miocene
(Axelrod, 1985).

By the late Miocene, the Great Plains had become much drier (Axelrod, 1985).
A temperature-constrained herpetofauna (mainly the occurrence of the tortoise,
Geochelone. and abundant crocodiles) indicate that the late Miocene was still very
warm even in extreme northern latitudes (Holman, 1971). These hot, arid conditions
were favorable to drought-adapted floras such as the C, savanna grasses. Additionally,
the rise of a scrubland-dwelling caenophidean snake fauna at the expense of the
arboreal boid fauna supports a mid to late Miocene savanna radiation (Holman, 1976).

The middle to late Miocene in North America was a time of increasing
diversity of the hypsodont equids (Janis, 1989; MacFadden, 1992). Consequently, the
C, biochemistry which likely arose contemporaneously with the expansion of warm
season grasslands, might have influenced the development of high-crowned teeth in
horses. Alternatively, the possibility exists that the C, biochemistry is much more
ancient and merely proliferated with the expanse of grasslands during the Miocene.
Whichever is true, an assumed correlation between the diversification of siliceous

grasses and hysodonty in horses remains.

The Pliocene Epoch

By the end of the Miocene an extensive ice cap had developed in Antarctica

(Woodruff et al., 1981). Subsequent to the Miocene, continued global cooling and

20

increased aridity altered grassland distibutions once again presumably causing the
drastic decline in equid diversity throughout the Pliocene (Janis, 1989). The Miocene-
Pliocene extinction event that greatly reduced equid diversity left the world with two
horse genera, only one of which (EMS) exists today. Axelrod (1985) envisioned the
Pliocene as an epoch of gradual drying, but argued for a significant woodland habitat
on the Great Plains. An essentially modern North American snake fauna by Upper
Pliocene times indicates a progression toward recent landscape organization (Holman,
1979). Large grassland and forest-dwelling herpetofaunas from the Upper Pliocene
also indicate a gradual modernization of the landscape with tall-grass prairies pervasive
in the west (Great Plains) and short—grass plains and deciduous forest prevalent in the
Central Lowlands region of North America (Rogers, 1976). This scenario pre-
supposes the development of a moisture gradient from west to east. A continental
rain-shadow developed subsequent to the uplift of the Rocky Mountains in the middle
Miocene which was responsible, in part, for the evolution of the Pliocene landscape
structure. The Plio-Pleistocene was a time of varied climate fluctuations which most
assuredly influenced the distribution of the C3 and C, grasses (Ralph E. Taggart,

personal communication).

Grasses and Hypsodonty

If C, grasses had an influence on the evolution of hypsodonty, they must have

existed in the early-middle Miocene (circa 18 Ma) when horse tooth crown heights

21

began to increase. The C, photosynthetic pathway may have existed as far back as the
Cretaceous (Brown and Smith, 1972). The fossil record, however, has yielded only one
definitive C, grass of Miocene age (from Nebraska, approximately 7 Ma to 5 Ma)
(Thomasson et al, 1986). A C, grass from California originally assigned to the
Pliocene (Nambudiri et al., 1978) is likely late Miocene (7 Ma to 5 Ma) based upon
mammalian fauna] correlations (MacFadden, 1984). This paucity of fossil evidence
might suggest that no relationship exists between the evolution of C, grasses and the
evolution of hypsodonty in horses. However, many fossil plants have a notoriously
poor fossil record and the identification of C, photosynthesis in fossil plants is difficult
even under the best circumstances. Carbon isotope analyses do not require
morphological preservation to determine photosynthetic pathways of extinct plants. In
fact, 5‘3C analyses do not even require the plant itself. All that is required is tissue
from the herbivore. Therefore, 6‘3C analyses of equid remains should prove useful for
determining when C, grasses were first incorporated into horse diets.

One reason C, grasses are thought to be responsible for equid hypsodonty
despite the lack of ample fossil evidence is the greater abrasiveness of most C, grasses
compared to C3 grasses. Silica inclusions may have evolved as an anti-predation
strategy of the diversifying C, grasses in response to herbivore grazing. Silica bodies
are often found along the ribs of C, grasses, but rarely are they observed in C3 grasses
(Kaufman et al., 1985). Silica bodies are located over the vascular bundle cells of C,
grasses where they alternate with cork cells (Kaufman et al., 1985). Plants which

exhibit "Kranz anatomy" (a specialized C, leaf structure) have low mesophyllzbundle-

22

shealth area (Thomasson et al., 1986)(Figure 3). Consequently, many C, grasses
possess more hypoderrn, a lignified, fibrous material surrounding the vascular bundle-
shealth cells within the stem (Elias, 1942). This larger amount of lignin, together with
cellulose and siliceous components make C, grasses less digestable, less nutritious and
necessitate herbivores to process much more grass. This extra processing is inefficient
for an herbivore with access to a more nutritous food source (e.g. C, vegetation).
However, some C, grasses contain opal phytoliths within the leaf epidermis which
may deter herbivory in much the same way that silica inclusions function in C, grasses
(Jones, 1964). The identification of opal phytoliths and silica inclusions within
anatomical features of plants represents the first empirical evidence suggesting the
evolution of a more abrasive grass structure might have led to the evolution of high-
crowned teeth in horses. The use of stable isotopes promise a fuller insight into the
relative importance of C, and C, plants in horse diets and into the influence these

vegetations had on the development of hypsodont teeth in horses.

23

 

Stoma
#914. Ahead“
effing-95:11,; N».

 
  
  

‘79"

 
 
  

 

? ‘

~‘

\‘J‘mu 31V» ‘ 1 ,1‘1‘9 A .

45,73)

 

\

 

Cuticle Mesophytl Cell
C , Plant 5““

 

 

 

 

 

 

 

 

 

Bundle-Sheath Cell Mesophyll Cells

C, Plant

Figure 3 Arrangement of the various cell layers in C, and C, plants. The
vascular bundle-shealth cells of C, plants, absent in C, plants, are
surrounded by silica bodies and lignin fibers making them less

digestible (modified from Wessells and Hopson, 1988).

24

MATERIALS AND METHODS

Sample Procurement

Modern horse skeletal and dental specimens were obtained from MSU

Veterinary Medical Center (courtesy, Peter Ocello) and the AMNH (courtesy, J. Daniel

 

Bryant) (Figure 4; Table 1). Modern wapiti (Cervus canadensis) material was also
obtained through MSU (courtesy, Dr. William Cooper) (Table 1). Fossil horse
material was supplied by the AMNH (courtesy, Dr. Richard Tedford), the UNSM
(courtesy, Dr. Michael Voorhies) and the UA (courtesy, Dr. Everett Lindsay) (Table
2). Specimen selection was confined to individuals 32 Ma and younger owing to the
size constraints of geologically older specimens. At present, horse teeth from the
Oligocene and earlier are too small to obtain enough HMW organic matter for isotopic

analysis without combining samples.

Specimen Documentation

Isolation of organic matter for isotopic analysis requires that a portion of the

specimen be destroyed. For modem specimens this does not usually present a

problem. When dealing with an exhaustible, exceptionally valuable entity such as a

25

 

 

Figure 4 Modern horse skull and teeth shown in cross-section to illustrate the
pronounced hypsodont condition. Premolar and molar tooth position
nomenclature is given (modified from Simpson, 1951).

5.25)) 1C... Rowena; FCOUCPC CC.— COZ=ZCCkp= .2:>0_9-_ DC: \AFCCCCXQLI ~ 07:45

26

 

8.3502 .60 5..
case: .8 8.56
59:22 .8 5.56
£822 .8 828
59:22 .8 5.56
59:22 .8 .528
53:22 .8 8.56
525:2 .8 82:0
E8222 :00 5.5.0
58222 .8 5.26
59:22 .8 8.56
59:22 .8 8.56
525:2 .8 8.56
5322 .8 5.56
c8222 .60 53:0
5322 .8 828
case: .8 5.56
595.2 .8 8.56
59:22 .8 5.56
59:22 .8 8.56
595:2 .8 8E6

c8222 :00 9:822
9:833 .60 «can:
mega»? :00 «Sam
mic—0&5 :00 «:5:
9:50»? :00 ~58:
95:33 :00 «:5:
mega»? :00 «can:
wage»? :00 «can:
minced azamweoc

.533 can 830; E82: .8 cones—BE. Ego—2 v5 fleeces...

no: 23 a: a
so: 29 :5.
so: 28 Ex
ocom «:38; M—

:2

Ex

2m

22¢

22m
$85.: :2m
22%.: :2m
2285.: :8
$58 Em
€38 Em
case: .32
22%,: :5
€38 :5
22%.: Ex
€38 Ea
€38 :2m

2288 :2m
.95 3832

©8333: 28:2
035. 9:80:80

532 2:8 .5:

.3:
25m @8ch a
3:8
:5.

EquE Biz—am

 

a
“leg
.88
33%;
9138
88
88138
888
a;
888
3.38
84%
8408
8.188
88.8.88
88
8.81.38
8831.8
888
88
3:38 3.6”.—

N.

288qu 3:00

8%

22888 $580

 

 

 

 

 

£203.50 Shoo

 

amen—Ugo mPCoO

mace—cacao mstou
flap—338 £5.80

3.89. 35¢

22 2.2a
a .8 5m:

m .8 5m:

5 .8 522

e .8 5m:

n .8 5m:

4 .8 5m:

m .8 5m:

N .8 5m:
:8 5m:
V: .8 5m:

.. _ .8 5m:
:8 5m:
£8 5m:
28 5m:
:8 5m:
:8 5w:
3.8 5w:
2.8 522
£ .8 5m:
2 .8 5m:

_ .8 5m:
2 .8 32
2 .8 5m:

_ _ .8 5m:
2 .8 5m:
3 .8 5m:
3 .8 5m:
2 .8 Mom:

BEE-.2 9.958

_ 033.

27

 

 

 

 

 

 

 

 

  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

8.888 N. 8.. 88.8 mm 88.8..8 88.8.2 so... 2< Hu.
8.888 4 8.. n8... 2... 188488.884 282 32 H..
25. .85 m 883 m8:8... 3. 4.388.388. 3.2.2 32 N...
NE; .88 82a n: 8.8. .8888... :82 :2 H..

.2: v.86 88m 2 85.8 88.88:. m . 98 $2.)...

85-..: .83. .. .88 8....m t 82.8 88.88.)... 88.. $2 Hh.
.28.: :85 2. 888m... .8 88.882 3.2. 3... Un.

$3- r: 8.2 A 28.0 m. 8.8. 88.882 o . RN. 2... Un.
Em 8..8_.> N. 858 88.8.... 80.... .2... Mn.

25. 8.82.3 N. .8... 882 32 H...

ms... 8.8.a> N. 858 88.8.... .338. 2... H..

ms... 8..8_.> N. 8.8.. 88.8.... 2.82 32 Hn.

N238: .83. 1. .88 88886. 3. 21888.80 8.88.888 .382 32 H..
N... 32.2. fi< o. 8.8208 8.88.888 88.... 2< a.

9: 32.0.. 8.. m... 8.8280 8.88.888 882 32 H.

w: 32.3. 8.. no 0.88888 8.8.8.888 88m. 2... H...

9... .88 :88? m... 28.8208 88888.58 282 .>.< Mb.

82.. 88. 828:8 .. .88. .0 8.88.8 8-3 . m 352:

ms: 88. 825:8 5 .2. .8 88:8 m2-.. .8 352:

N... .88 3.25 . 88.8.2 .8885? <88. :23...

2):. .88 :88... 3 0.88.8 8.89847. 888 $2 Hh.

.288. s... 88. 898...: 8 28.8.8. 88.882 8-8 . m 552..
.233. ms... 88... 828...: o 8.888. .88....85. new? 3575
ms... .88 :88... mm 388.2 .8885? 88m. 5... H..

.25. 8 :5. E.3. v 88.23:... 8:8 v. .8... $2 Hn.
8.5. 8 .2”. 88.. .828... m 888 0.8 5.5
3.5805 3.38% cots—FE...— QE. amt. 8.92... 8:60 .3537. cos—Beam

 

«8.2. .565 .8 co..m:..8=. Ego—o. v.3 ancoxah

N 2an

28

fossil it is of utmost concern. Consequently, all fossil material utilized in this thesis
represents either highly abundant fossils (e.g. horse teeth) or remains which are
unlikely to be useful in morphological/taphonomical studies. In the future digital,
three-dimensional computer imaging will make possible the ability to preserve the
morphometric dimensions of fossils in virtual reality. Unfortunately, this technology is
not currently available. Therefore, all specimens were measured, sketched and
photographed (from two separate angles) to preserve the morphologies. A portion of
every fossil was saved in its original state for comparative purposes to future studies.
Illustrations of two representative fossil horse specimens obtained from the AMNH are

included (Figure 5, Figure 6).

Locality and A ge Description

Documentation concerning stratigraphy and chronometry, depositonal
environment and paleontological information was obtained from Cenozoic Mammals of
North America (Woodbume, 1987) or from a series of bulletins of the AMNH
(Galusha, 1975; Skinner et al., 1977; Skinner and Johnson, 1984) (Table 3). Voorhies
(1990) provided additional biostratigraphic and radiometric information relevant to the
horses analyzed in this study. Without exception, the geologic age assigned to each
fossil was determined by the loaning institution (Table 3). In most instances there is
good agreement between the age calibrations of the separate institutions. For one

locality (Burge Quarry) there is considerable dispute concerning its proper NALMA

29

 
     
 

  

 

 

 

 

 

 

 

4 M1
P P4 Ml
Matrix
Bone -r. Enamel
Dentine
Thickness: I. "x
, 25““ Q 331 46368
17mm \ l - 5. 5mm a. Slow!
‘- \
\ .l- Cementum
t—-——-1 4
20mm r 20mm 4
if I
40mm

Figure 5 Labial view (left illustration) and occlusal surface (right) of LP4-LM1
of a 15.5 Ma merychippine horse, Megchippus 3;. (cf. isonesus)
(F:AM 129453).

 

Complex Enamel Folds

32mm Dentine

   

 

 

 

J.

Figure 6 Labial view (left illustration) and occlusal surface (right) of RM3 of
Cormohipparion occidentale, 8.5 Ma, from Ellis County, Oklahoma
(F:AM 129442).

30

placement. The loaning institutions (AMNH) designation of Latest Barstovian (12
Ma) has been preserved in this case in lieu of Early Clarendonian (11 Ma) as some
paleontologists advocate. Aside from Burge quarry, all other age estimates can be

considered accurate to within :t 0.5 Ma (J. D. Bryant, personal communication).

Preservational States

The state of preservation was defined for each fossil and used to determine
which samples would be analyzed (Table 3). Preservation was defined herein to
include the gross morphological condition of the fossil (e.g. exposed tooth roots,
extensively worn crown enamel, root casts from plants present, etc.), the amount of
preserving agent (e.g. consolidants) and the amount of encasing matrix around the
fossil. Fossils which were found to be mostly or wholly intact, predominantly free
from perserving agents and easy to remove from their matrix were selected for isotopic
analysis. With the exception of one specimen (F:AM 129441), all horses analyzed in
this study had attained adulthood. Maturity was evaluated by surveying the wear

patterns on the tooth enamel.

3l

 

wccaou gustomoi Emfi

 

 

865 :63 coo—(.4 205840 40:34 .485 canoe—430 .4 mm So: 24 m
.35 oz .cc:§__sm.coom 5:50 348.4 EaE coo; 22220 40:34 sue—tam e£35 .m Won $32 3?. ”m
9::on PICNCDmam $01015. m. .0 100.40 89:55:00 GEUHOHMEEQE .m n.w_ VbeN— E ”L
mccmou 959/335 Ewfi €830 €425; 02.0.82 50 wco4 caBSEEEoE .4 M: Swan— 3 ”“4
52; 2:52 5:22 224 44880 42:56 930 assofi 54234520: .4 t m _ 8N mzz<
wEEoU oEmKomoi Ewfi .6030 3:520 E90 :8895 5653550: .4 E :8: 32 i
92:30 92:5.Comoum .425. :E 524054058440 Easo ficaw 4mm”; Curioumumm— .m mfi— mmvom— E ”m
Scam ECCF E @0105,— 74522 50 anbm E540 ogom Cagoamuwm .m m— 0_ MoN— 5 “h
9:6.cumoi EH4 .Cu> 8204 >20 Exam 4533* E45 0945 :mSBmSm .4 .4 N4 woo:— 2< um
5:22 4256262805 234 3894 >20 7.28m 333* >530 093m 53235 .4 .4 N. 332 2< um
5:22 424.4..0.§.c§i 4%: $254 .36 Exam 8332 EEO omsm 5:23am .4 .4 2 $32 $2 a
€4.42 425.2688: .42.: 8254 Eu 35m 333x E30 ems: 53222. .4 .4 2 $32 :2 u...
:mmc; 330cm anE 4. E450 «Ewe/‘82 54:35.20 .m no. .302 E< U..—
wEEoU P/CSK—Qmoi Emma mama—U gun—am 53a vcam >50 E30 Cogs—U camcovcougo .4 O_ vaom _ 24‘ ”m
45480 23385 224 2% 5:30 828-433 43 .594 E5 «sax 54822.46 .4 ma 8:2 :2 a
£522 224 .uéahafii 2E4 Beam 53:0 3:84—03 can .58 .530 88x 5:84.530 .4 no 24va— 2< ”"4
33225 44:32 diatomem oz 3:520 535m «E him we tom 5:42an: .m .m Wm $.32 32 v.4
as: 5&3 .6 84:94 .243 9. a 3:804 32: E5 82305. 54.44350: 42 a 2.3: 352:
455443 502 424454 ESE: 9. a 3:804 362: E30 82322 Séfieom .2 a moses 362:
42a .5444 .325 634 as: :4... 42455 _ .0 33o 323% 54.4295: .m h 43%: $232
.35 .98: .825. 48m .55 2.85 5.3 .2820 885m >550 ,4. xom can—Eng: .m .4 We 2334 Z< ”“4
522 5%: .0 SEE .243 S4 5 3:804 .2sz E26 26.4%: 552,445: .4 4 one: 52:
52:42.4 34o: 84 5 3:804 :52: E30 0.644%: 5529544 .4 4 mean 32:
5:22 food 52:. 4480qu 348400 3:52.40 8.42% E50 8250 54:54:43,. .4 Ow 33.3— 3 E
cocoom 83.65.; V: 4. 55m 532 sauce—m .4 v 3 3S 3 E
522 3%: banEmEm 98m .4420 65882900 «:0 4.859284% cacao—m .4 m 055 @4445
32m 3:23.438..— .55:e.:.Em 3.828.424 .0350 442442 :4): um< 34:52 55325

.2438 9.30: we 484888.45 4.4232305 v5 .03 $5304 m «Bah

32

Preparatory Methods
Tooth and Bone Cleaning and Preparation

All samples were thoroughly cleaned with dental instruments to remove any
exogenous contamination (e.g. matrix) within tooth or bone cavities. On occassion
small arthropod exoskeletons were removed from the pulp cavity of a few of the
modem horse teeth. The outer enamel surface from every sample was removed using
a dremel tool or a chisel to remove external contamination and any preservative when
present. If this procedure did not visibly remove the preserving agent the fossil was
not analyzed further. Throughout the cleaning procedure gloves were worn and only
ashed glassware and dental tools were used to prevent contamination. The complex
inner enamel folds of the hypsodont horse teeth made dentine-specific isolation
virtually impossible (Figure 7). Consequently, all interior tooth layers were extracted
together. The inner enamel, dentine and cementum complex, free from preservative,
was then cleaned rigorously with dental probes to remove any remaining sediments or
impurities from within the pulp cavity. The cleaned inner complex material from
either the tooth or bone was then sectioned with a dremel saw. These fragments were
pulverized in a steel mill and subsequently blended to a fine grain size in a Waring
blendor or, alternatively, ground in a glass mortar. Tooth and bone powders were

stored at room temperature prior to demineralization.

Figure 7

33

'T'O-z-unn ,

Iv . .
O
~u,
. .
. . .

‘l
11“

3

‘b

1.
opfi-m-n-n.
. . .

 

 

 

Cross-sectional view showing the relationship of the layers of a molar
tooth from an extant hypsodont horse, Eguus (right), and a low-crowned
molar from an extinct horse, Parahippus (left). Note the complex inner
enamel foldings of the hypsodont horse (modified from MacFadden,
1992).

34

Hydrochloric A cid Dem ineralization Technique

Granulized samples were demineralized in cold (4°C) hydrochloric acid (either
12.5%, 25%, or 50% HCl for modern samples; 50% HCl for all fossil material).
Hydrochloric acid dilutions (1N, 3N, and 6N) were accomplished using 12N HCl
(Optima, Sigma) in varying volumes of distilled, deionized water. Modem samples
were treated with different acid normalities to determine the relationship between acid
strength and demineralization efficiency. Samples were allowed to demineralize for
twenty-four hours under these conditions. Some modem samples did not acheive a
full demineralization after twenty-four hours. In these cases the supernatant was
decanted from the undemineralized pellet and fresh, chilled, dilute HCl was added to
the beaker. If the sample still was not fully demineralized after another hour at 4°C it

was ground in a glass mortar over ice to facilitate demineralization.

Dialysis and Lyophilization Procedure

Immediately following demineralization, the HCl-soluble fraction from each
sample was placed into dialysis. The supernatant (containing the HCl-soluble organic
fraction) was pipetted from any residual undemineralized pellet. Typically there was a
small percentage (>5% dry weight) of non-calcitic material which did not
demineralize. The supernatant was then placed into a newly hydrated dialysis bag

(Spectra/por 1, 6,000-8,000 molecular weight cutoff) secured with labeled clips. The

35
dialysis bags (usually 3-4 together) were then placed into a 4.0L beaker of distilled

water at 4°C for 14 days. Dialysis water was changed three times on the first day of
dialysis, twice per day for the first week and once per day for the final week in

dialysis. This procedure purified the HMW organic extract by removing anions,

cations and other low molecular weight material. The HMW organic isolates were
removed from dialysis and placed into individual 250ml ball flasks. The inner surface
of the dialysis tubing was rinsed with ~5ml of quartz distilled water to minimize sample
loss. This fraction was added to the ball flask. Samples were lyophilized for twenty-

four to forty-eight hours depending upon the initial volume of the sample.

Horn ogenation Technique

The lyophilized material from each sample was homogenized with a
Wig-L-Bug amalgamator (Henry Schein Dental Supply Company). Compared to other
techniques (e. g. mortar and pestel), the Wig-L-Bug provided the best homogenation
efficiency and the least sample loss. The HMW homogenates were stored in a freezer

at -5°C prior to isotopic analysis.

Stable Isotope Analyses

Isotope analyses were performed using a modification of the Dumas

combustion followed by cryogenic gas line separation (Macko et al, 1987).

36

Alternatively, an automated technique (Carlo Erba analyzer) involving separation on a
GC column and subsequent transfer to the MS under helium flow was utilized. The
next several sections deal with the methodology involved in the preparation of samples

for the manual mass spectrometry method.

Quartz Tube Evacuation and Combustion

Samples for manual isotope analysis were prepared by placing between 17mg
and 24mg of HMW fossil powder in a 6mm quartz tube. Approximately two inches of
copper oxide and one inch of copper metal shavings (Cu) were added to the sample
prior to sealing the quartz tube under vaccuum. The sealed tubes were shaken to
thoroughly mix the combustion reagents with the sample powder. The tubes were then
combusted in a furnace at 850°C, then 650°C, and finally 550°C for one hour each.
Following the combustion, the tubes were allowed to return to room temperature over
night so as to quantitatively convert the nitrous oxides to N2 gas. This protocol
converts the solid tissue sample to C02, N2 and H20 vapor which can be cryogenically

purified then isotopically analyzed.

Cryogenic Gas Line Separations

Sealed, combusted quartz tubes were introduced to a cracking apparatus

connected to an evacuated gas line. A liquid nitrogen dewer (-195°C) in contact with

37

the gas line served to trap the carbon dioxide and water vapor once the sample gases
were introduced. The sample N2 was allowed to equilibrate for five minutes after
cracking. The nitrogen gas was collected on a molecular sieve containing zeolite. In
a similar manner, carbon dioxide gas was collected using an isopropyl-liquid nitrogen
slush (~65°C to -70°C) into pyrex tubes which were sealed before being removed from
the gas line. Prior to collection of the CO2 gas, percent carbon determinations were
made using a calibrated manometer. The purified CO2 and N2 gases were then

analyzed on a VG PRISM dual inlet mass spectrometer.

Dual Inlet Mass Spectrometry

Samples of N2 gas were introduced into the dual inlet of the mass spectrometer
by heating the molecular sieve to approximately 300°C with a heating cord. Carbon
dioxide samples were cracked into the spectrometer. All samples were analyzed in
comparison to laboratory gas standards which were previously calibrated against NBS
standards (e.g. PDB and N2). Precision for (WC and EN replicate analyses on the MS

was :1: 0.1%o.

Carlo Erba Method

In addition to isotope analyses approximately ten 5'3C values were determined

using a Carlo Erba elemental analyzer interfaced to the VG PRISM stable isotope ratio

38

mass spectrometer. This required no more than 1mg of sample. The Carlo Erba
combusts the sample carbon and nitrogen to CO2 and N2 using a modification of the
Dumas method. Water resulting from this combustion is trapped using magnesium
perchlorate. The purified sample gases are separated via a GC then sent on to the MS

for analysis. Instrument precision for the Carlo Erba is typically 02%;.

A mino A cid A nalyses

An amino acid recovery experiment designed to quantify amino acid loss
during column purification was performed. Solutions of known concentration of ten
amino acids were placed over a cation exchange resin column. Standard solutions of
identical amino acid concentration were made for each sample. After column
purification, sample eluents were run by HPLC to assess amino acid recoveries.
Amino acid analyses were subsequently performed on fossil organic matter
hydrolysates by gas chromatography. Three representative fossil samples (4.0 Ma, 5.5
Ma and 10.0 Ma) were chosen for amino acid analysis. The 4.0 Ma sample was
chosen because its appearance was consistent with being diagenetically altered
(exposed inner surfaces and abberant coloration indicating possible diagenesis). The
5.5 Ma sample was selected because it appeared to be the least altered (intact
structure, no visible signs of alteration) of the entire time series. The 10.0 Ma sample
was neither visibly well preserved nor significantly altered, but was intact.

Calculations of the mole percent of each amino acid were made. The HMW material

39

extracted from modern bovine bone (Ostrom, 1990) was used as the standard for

comparisons to ancient equid amino acid patterns.

Organic Matter Hydrolysis

Five to ten milligram aliquots of HMW organic matter isolated from individual
fossil horse teeth were each hydolyzed in lml of 6N HCl at 100°C for a period of
twenty-four hours. Hydolyzates were dried under filtered air, then rehydrated with
500pls double distilled water to concentrate the amino acids before proceeding with
the desalting procedure. Modern amino acid standards did not require concentration in

this manner and were placed directly over the resin column.

Column Chromatographic Purifications

To assess percent recoveries and reproducibilities, amino acid standards of
known composition and concentration were placed over an affinity ion exchange
chromatography column. Fossil samples were desalted by the same procedure used for
the amino acid standards. Glass burets (25ml) were packed with approximately 10ml
of ion exchange resin (BioRad, Amberlite 50W-X8 100 mesh size) freshly hydrated in
pH 2 quartz distilled water. The column was then equilibrated to a slightly acidic pH
by washing with 3 to 4 bed volumes of quartz distilled water. Protonation of the

column in this manner facilitates the ion exchange reaction between the resin and the

40

sample amino acids. Next, the sample amino acid solution was added to the top of the
column. Two bed volumes of quartz distilled water were placed over the column
immediately at flow rates between 0.2ml/min. and 0.7ml/min. Slower flow rates
increase the ion exchange capabilities of the resin. Multiple bed volumes of quartz
distilled water were added to the column to remove all the salt contaminants. Amino
acids were eluted from the column by the addition of 2 to 3 bed volumes of 3N
ammonium hydroxide (NH40H). The eluent was rotoevaporated at 40-45°C to a
volume of approximately lml, dried down under filtered N2 gas and rehydrated to
approximately 500p] in pH 2 water. A 20011] aliquot of the resulting solution was

dried and used for derivitization.

Derivitization Procedure

Desalted hydrolyzates were derivitized to their TFA esters (Engel and Hare,
1985). The derivitization involved esterification with an acidified isopropyl alcohol
followed by acylation using TFA (Engel and Hare, 1985). To each vial 500p] of acid
alcohol (1:4 acetyl chloridezisopropanol) was added. The samples were allowed to
reflux for one hour at 100-1 10°C in teflon sealed vials, quenched in an ice bath and
dried under filtered N2 gas at room temperature. To each sample 100“] of methylene
chloride (CHZClz) was added and the drying step was repeated. Samples were
esterified with 500p] of TFA and 500 pl CH2Cl2 for 10 minutes at 100-1 10°C. This

reaction was quenched in an ice bath then dried under N2 gas. The derivitized amino

41

acids were resuspended in a known volume (typically lul) of CHzCl2 and injected into

the GC for analysis.

Enantiom er A ssessm ent

The relative abundance of amino acids in a sample were estimated using a GC
(I-IP5790A) interfaced with a MSD (HP5971). Individual derivitized amino acid
solutions were injected into the GC inlet where they were combusted (at 280°C) to the
vapor phase. Separation of amino acids was achieved by maintaining a constant
column temperature of 90°C for seven minutes then changing to a final temperature of
200°C at a rate of 1.6°C per minute. The mobile phase, helium gas, transported the
sample gas down the capillary column (internal diameter 0.25mm). The coating on the
50m long Chirasil-Val (Applied Science) column acts as the stationary phase and was
composed of N-propionyl-L-valine terbutylamide coupled to a co-polymer of
carboxyalkylmethylsioxene and dimethylsiloxene. Amino acids were identified both
by retention time and by comparison of the sample ion fragmentation patterns to an
amino acid standard mass spectral library. Ion fragmentation patterns were analyzed

over the range from 50-500 mass to charge (m/z) units at 1.7 cycles per second.

42

RESULTS AND DISCUSSION

Yield and Recovery Experiments

Dem ineralization and Lyophilization

Fossil material ground to a uniform powder in a shorter time period and
demineralized more efficiently than did modern materials. Lyophilized material
consisted of a light brown powder and occassionally a fluffy white component. These
may represent organic fractions in different conformations. Some of the powderized

material probably was silicified crystals which would not demineralize.

High Molecular Weight Organic Matter Yields

Percent yields of HMW organic matter obtained from all modern and fossil
samples were determined. The results from four separate demineralization techniques
used for modern and ancient materials (1N, 3N, 6N HCl and 6N-2Hr.) are reported
(Table 4, Table 5, Table 6). Duplicate sample numbers denote separate elements from
the same individual. Modern horse and wapiti tissues treated in IN HCl yield from

0.3% to 10.5% HMW material and have an average value of 3.0 i 1.9% (Table 4).

43

 

 

 

 

Table 4 High molecular weight organic matter yields obtained from modern
horses and wapiti.

Sample/Element Genugpecies Demin. Mass (mg) Organic Mass (mg) Percent Yield (°/o)
MSU: Ccr 1 RMl Cervus canadensis 1540 19 1.2
MSU: Cer l RMl Cervus canadensis 1560 12 0.8
MSU: Cer 1 RM] Cervus canadensis 1550 26 1.7
MSU: Cer 2 dP4 Cervus canadensis 1070 29 2.7
MSU: Cer 2 dP4 Cervus canadensis 1080 32 3.0
MSU: Cer 2 dP4 Cervus canadensis 1080 23 2.1
MSU: Cer 3 Dent. Cervus canadensis 1220 128 10.5
MSU: Cer 3 Dent. Cervus canadensis 1380 41 3.0
MSU: Ccr 3 Dent. Cervus canadensis 1090 32 2.9
MSU: Eq la RM! Eguus caballus 1960 71 3.6
MSU: Eq 1d RMI Eguus caballus 1150 48 4.2
MSU: Eq 1fRM1 Eguus caballus 870 66 7.6
MSU: Eq lj RMl Eguus caballus 2260 48 2.1
MSU: Eq 11: RMl Eguus caballus 2410 44 1.8
MSU: Eq 11 RMl Eguus caballus 2180 28 1.3
MSU: Eq 2 RMZ Eguus caballus 2200 78 3.5
MSU: Eq 3 RM3 Eguus caballus 5040 16.8 0.3
MSU: Eq 4 RP2 Eguus caballus 2520 83 3.3
MSU: Eq 5 RP3 Egyus caballus 2500 87 3.5
MSU: Eq 6 RP4 Eguus caballus 2560 41 1.6
MSU: Eq 7 R Max. Eguus caballus 2000 64.5 3.2
MSU: Eq 8 RM] Elms caballus 2000 185.2 9.3
MU: Eq 9 RM! Eguus caballus 2500 I59. 9 6.4
F.M.M. 1219 R Max. guns 3g. 2480 305.3 12.3

Mean (1N HCI) 3.0 1.9
Mean (3N HCl) 6. 4 NA
Mean (6N HCI) 10.8 NA

44

High molecular weight organic matter yields obtained from ancient

horse tooth and bone.

Table 5

 

 

 

 

 

 

 

 
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

mm 3.. o2. N... . . . .82. $9... :2 m
3 2b 88 new 33332 .8... 232 .2... m
3. 2m 28 2: 33 3.2.2 .2... m
f 3.. 88 w. 33 233.332.. 5.2. $2 a
2 n3 coon : 3...... .3359? 2 SN :23.
no 2.: 82 2 23... .3365? .32 2.8: :2 H...
N. 3. 9.2 S 3...... ..m.........oEE.. 3.53: 2... m
2 3mm 82 n2 33.5.. 3.33353: 3.3 :2 H...
S. 3. 82 2 m. .3 v... .5 22 .32 2 RN. :2 a
3 2.: 82 2 .33 382052 2.. 2 82 :2 H...
3.. S: 82 N. .33 9.33: $2.: 2... H...
3. m: can N. .3”... 2.33: 352 2... m
I. am 8% N. .32. 333... $32 2... m
2 3.2 9.2 2: 22.308 8.333.638 2,: _ 3.3 $2 a.
on 36 9.2 2: 232.68 333.635 33.33 :2 H...
2. .2 82 2: 3.338.. 338.358 :2.— .33 :2 n...
o. _ 2mm 28 2: 3.2.68 3.382038 .32 :32 52 m
3 a. can 2 338.. 333.638 :32 32 N...
am on 82 3 3338.. 333.638 323 2.... a
2 SN 8: 3 33.68 8.382958 $.52 32 m
on 8. Sam 3 23358 333.638 ~33. 52 m
2 3.2 82 5 13.133 9.5-22 :52:
I. 22 8...... e 32...... ............_o...o.. <3“: :22...
3 3.: 32 e :83 .3 .8 333: 2-3: :52:
i 22 82 3 8.33... 3.35282 2.3. :2 H...
E 33 82 0 32m $52:
3 ”a 3: o . $3 32:
N... $2 82 3 3.9.22 ..m........_o..5.. $32 $2 a.
E. 22 o3. .. 2.0.3.3.... 3m 3 32 $2 a.
w... 2.: 8? m .9 3mm on: .35

33 22> 33.8.— qfiv ".32 g 13.5 222 .5599 A2): om<i cozacwga ago—Snub 538.5

 

45

 

 

  

 

 

 

 

«on... .\+ 3 .33 3.5: an N 5: 23 532
3. NE 82 ~_ 350m mammEEE $3va 3?. E
0m 3% 82 S . . Womvmfi 3?. m
m.m 9mm 82 S . 52E Tomvmfi S? ”m
3.; 23> 2.33m #35 was: umamMmO 3.5 «an: .5509 A35 ow< 8.9.2.» 2:30 5:5on

 

.EoEtoqxo 533—95580...

So: 95 a :8.“ 3:330 mEoS house 2590 £303 530.08 :3: o 2an

46

These same tissues yield 6.4% and 10.8% HMW material when treated with 3N HCl
and 6N HCl, repectively. Modern organic matter yields of this magnitude are
consistent with the small fraction of organic matter present in mammalian dentine. In
most cases, bones yield a larger percentage of HMW organic matter compared to teeth
(e.g. F:AM 129441 and F:AM 110575, Table 5). The HMW organic material
retrieved from horse fossils demineralized twenty-four hours in 6N HCl range from
1.1% to 17.4% dry weight yield and average 6.1 i 3.7% (Table 5). This yield is
similar to the dry weight percent yields obtained from prehistoric remains by other
researchers (DeNiro and Epstein, 1981; Nelson et al., 1986). Fossil HMW material
obtained from the two hour demineralization experiment gave a mean value of 3.8 :L-
0.3% (Table 6). Variations in the yield among fossils could relate to either differential
diagenetic alteration or to variable contributions from aluminosilicate minerals that
were not removed during demineralization. The mean yield of I-MW organics from
modern materials isolated after demineralization in IN HCl (3.0 i 1.9%) is similar to
that obtained from fossils treated with 6N HCl (6.1 i 3.7%). Consequently, modern

and fossil yields are comparable despite the difference in demineralization technique.

Percent Carbon Yields

The mean percent organic carbon yields for the fossils analyzed in this study

was 2.2 :1: 1.0% (Table 7). Despite being small, this value is similar to the published

47

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 7 Percent organic carbon yields from select modem and fossil samples.
Specimen Genus species Age (Ma) Percent Carbon (%)
F.M.M. 1219 Eguus Sp. 0 28.5
UALP 7810 Eguus sp. 3 3.7
F: AM 129114 Equus simplicidens 4 5.3
UNSM 35-568 "Dinohippus" leidvafis 6 3.5
UNSM 316-56 Nannippus lenticularis 6 1.8
UNSM 316-56 Nannippus lenticularis 6 2.3
UNSM 5157-73 Hipparion sp. (cf. forcei) 7 1.9
AMNH 17599A "Dinohippus" leidvgnus 7 2.3
UNSM 2010-708 Calippus sp. nov. 7 0.7
F: AM 129445 Cormohipparion occidenglg 9.5 1.5
F: AM 129446 Connohippgrion occidenta—le 9.5 1.3
F: AM 129454 Cormohipparion occiden_ta_1§ 10 1.4
F: AM 129454 Cormohipparion occiden_t_alg 10 3.5
F: AM 129441 Max. Cormohipparion occidemLe 10.5 2.5
F: AM 129441 Max. Cormohipparion occidenjglg 10.5 3.0
F: AM 129441 LMl Cormohippamn occidentale 10.5 2.0
F: AM 129441LdP4 Cormohipparion occidentale 10.5 2.7
F: AM 129441LM2 Connohipparion occidegalg 10.5 3.4
F: AM 129447 P_1iohiDDus pernix 12 1.2
F: AM 129448 Pliohippus pgmix 12 1.8
F: AM 129450 Pliohippus pemix 12 1.7
F: AM 129450 Pliohippus pernix 12 1.7
F: AM 129450-A Pliohippus pernix 12 1.6
F: AM 129450-B Pliohinpus pernix 12 1.7
F: AM 129450-C Pliohippus pernix 12 1.2
F: AM 129316 LP3 Mervchippus insigms 15 1.4
F: AM 129316 Max. Mervchippus ins_i,gm§ 15 1.1
F: AM 129453 Megchippus sp.(cf. isonesus) 15.5 2.8
F: AM 110575 Max. "Mervchippus" primug 17 1.7
AMNH 20513 "Mervchippus" prim_u_s 17 1.8
F: AM 129451 "Me_rychipp_us" tertius 18 1.7
F: AM 128784 RM2 "Parahippus" sp. 18.5 3.7
F: AM 128784 Max. "Parahippus" 52. 18.5 1.8
F: AM 129455 Acct. Mesohippus sp. 30.5 1.9
F: AM 74067 Acet. Miohippus obliguidens 32 3.7

 

 

Mean

2.2% +/- 1.0 °/o

48

percent carbon yields from ancient materials of DeNiro and Werner (1988) (0.3% to

10.7%) and Ostrom et a1. (1990) (2.0% to 9.0%).

A mino A cid Recovery Experiment

The results from three representative amino acids with varying initial
concentrations (aspartic acid, glutamic acid and serine) are reported (Table 8). Percent
recoveries are reported as mean percent recovered of each amino acid (Table 9). The
data in Table 8 and Table 9 demonstrate that column purification does not result in
appreciable amino acid loss. Aspartic acid and serine experienced no appreciable loss
irrespective of initial sample concentration. As much as 12.6% of the original
glutamic acid may be lost during the purification process when the initial concentration
is 0.1nmol/ul or higher. A loss of this magnitude is not problematic for interpreting
overall amino acid abundances and distributions. For lower concentrations, glutamic
acid loss is of less concern (e.g. 98.9% mean recovery when C, = 0.01nmol/u1). The
increase in loss of glutamic acid at higher concentrations relates to a saturation of the
theoretical plates within the column and might be avoided by using a larger resin bed.
As such, amino acid loss during purification is minimal. Consequently, all fossil
samples analyzed for their amino acid content were first purified by ion exchange
chromatography to remove salts, minerals and HMW material remaining after

hydrolysis which might inhibit the derivitization step.

Table 8

Mean peak areas for three amino acids, before column purification and

49

after column purification, with varying initial concentrations.

 

Amino Acid Standard [1

0.2nlnoI/20ml

 

 

 

 

 

 

 

 

 

 

 

 

(Pm-purification)
Amino Acid Mean Peak Area (mV—sec) 0
ASP 8.80E+06 1. 1313-1-06
GLU 9.09E+06 3.49E+05
SER 2.06E+07 1.26E+06
Amino Acid Standard 11 0.2nm01120ml
(Post-purification)
Amino Acid Mean Peak Area (mV-sec) 0
ASP 1.0SE+07 2.42E+05
GLU 1.0ZE+07 1 .96E+05
SER 2.79E+07 2. 16E+05
Amino Acid Standard 1 2.0nmol/20ml
(Pm-purification)
Amino Acid Mean Peak Area (mV-sec) 0
ASP 1.5 lE+07 2.7OE+05
GLU 1.91E+07 2.3 113-106
SER 3.5815407 6.7913405
Amino Acid Standard I 2.0nm01120ml
(Post-purification)
Amino Acid Mean Peak Area (mV-sec) 0
ASP 1.54E+O7 5.601905
GLU 1 .67E+O7 3.92E+06
SER 3 JOE-+07 5. 12E+06

 

 

 

50

Table 9 Mean percent recoveries of aspartic acid, glutamic acid and serine
from ancient horse organic matter following column purification.

 

 

 

Amino Acid Concentration Minimum % Recovery Mean °/o Recovery
ASP 2.0nmol/20ml 96. 1% 100.0%
ASP 0.2nmoll20ml 97.5% 100.0%
GLU 2.0nmol/20ml 59.8% 87.4%
GLU 0.2nmoll20ml 87.8% 98.8%
SER 2.0mnol/20rnl 87.4% 100.0%

SER 0. 2nmol/20ml 100.0% 100.0%

 

 

 

51

A mino A cids in Fossil Horse Tooth and Bone

The relative abundance and distribution of amino acids in three fossils of
different ages were determined (Table 10). The presence of hydroxyproline and high
concentrations of glycine and proline (verified both by retention time and mass
fragmentation patterns) in each sample suggest an indigenous component is retained in
these fossils (Figure 8, Figure 9, Figure 10). Modern tooth and bone collagen is
predominantly composed of glycine, proline and hydroxyproline (Hare, 1980). High
concentrations of aspartic acid and glutamic acid are consistent with the suggestion
that acidic macromolecules are preferentially preserved during diagenesis owing to
their high selectivity for the mineral phase (Masters, 1987; Robbins and Brew, 1990).
The lack of large concentrations of serine and threonine (common contaminating
amino acids) also suggest indigenous organic matter was recovered from each of these
three fossils. The presence of several D/L pairs in the chromatogram of the 4.0 Ma
fossil could be construed as evidence for contamination of this specimen (Figure 8).
However, this is not the case given the resemblance of the hydroxyproline, glycine and

proline amino acids to the modern bovine bone pattern.
Implications for Indigeneity Assessment of Fossils

The amino acid patterns of bones and teeth change as a function of time.

However, a direct relationship between age of the fossil and amino acid loss

52

U_o< oc_E<

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

min 30 >46 ._<> (.2.
. . . an.” a. . a.” O
Cm—H / : Ea
”
7 e
a»
H a L
/ m.“ 1 cm
/ .2
/ .. ...
/
/
U .. 8
/
b
m i 3
y
ms. 0.? EmEoEooo :otmofiEoEcoo D as. oé 9:330 333 1 cm
as. mam $298.29 _.m:QQ.EoEQ.. @ ocom oc_>om Soto—2 I 8

.83— .Eozmo an @8232 E 8% Box 05.5 EovoEV econ
Boo 525.: Mo Eaten Boa 05:8 2: 9 cog—2 5 9.03833.

«2&2 32¢ 2a 3.va 2,: .332 25: a: 3: Ba 22 2
.32 9v 5:33 025 com mousse—Spa Ea muonsnmbflu Eon oEE<

2 035.

wede a|ow

53

 

 

 

 

 

 

 

 

Abundance TIC: 3282.0

: GLU; UNRESOLVED HMW COMPLEX

7000003 I
2 l

eooooof : i
; l I

500000 : : i
_' ' !

400999? D/L ASP i
. l I

200000 2 i i a; H i
7 PRO PHE .
3A” i i i w")
J LEO I I I ‘1

1000007 §VALH I HYP . a .
i i‘ \ii . ii I I a

‘MVMAILA i- '1 m LVW.. -
0 1 g T r V T j I 7 fl 1 j I l ‘l' f T I I 7 7
wine -> 10.00 20.00 30.00 40.00 50.00
Figure 8 Amino acid chromatogram for a 4.0 Ma fossil tooth (F :AM 129114).

Note the presence of hydroxyproline and D/L aspartic acid peaks.

 

54

 

Thundance TIC: 3252.0

5000004 GLU UNRESOLVED HMW COMPLEX

-
I
.

4000004

. WWW

zooooog

 

.J

 

 

 

 

I I I U ' V I I
40.00 50.00

 

 

 

 

fiEundance Scan 3225 (28.752 min):
50000 134

.A.-h-h—h_—

40000

.1

43

I - .d
——-————g-——

aooooé

LA..- .._.-. .

20000.

u—u—J-a-A-oul—
—

10000'

—-al=r——-———— U

fiddg—L—J—L
H
H
.b

A: 59 H
0 .l.!!l!:. ...L....llL.. 9.5.11 1. . H..- .199".

 

 

      

 

 

I .
LIZ -> 40 60 80 100 120 140 160 180 200 220 240 260

 

 

Figure 9 Amino acid chromatogram for a 5.5 Ma fossil tooth (F:AM 129444).
The mass fragmentation pattern for hydroxyproline is given below the
chromatogram.

55

 

e TIC: 3244.0

7ooooo§

soooooi

. UNRESOLVED HMW COMPLEX
500000;

0
t"
_——————_——C~§C_

l

- I
4000005 ASP l H

- I

|

i
.IW,
I .LH'

PHE l Jur
l ' M
l

ILJ*'
U.. l HLJV'
.-W

aoooooé

200000:

 

1000003: ALA CITY LEU P130 nyp
~ . VAL“ ‘ l ‘

1

fl V fi‘!
Time -> 10.00

h-——_—

T Tfi

j I
50.00

 

U I I I V ‘l I I V I f l I
20.00 30.00 40.00

 

 

Figure 10 Amino acid chromatogram for a 10.0 Ma fossil tooth (F:AM 129454).
Note the large abundance of the glycine and proline peaks and the
presence of a hydroxyproline peak.

56

does not exist. Complexification of acidic amino acids with organic molecules may
stabilize organic matter in fossils over time. Further research directed toward
assessing how organic matter survives diagenesis is needed in order to better
understand indigeneity of ancient materials.

For this study, amino acid analyses supply several indications that an
indigenous organic component has been isolated from these fossils including: 1) The
distribution and abundance of amino acids is similar to the amino acid pattern of a
modem terrestrial herbivore (a cow); 2) The presence of a distinct hydroxyproline peak
which is unique to tooth and bone; 3) The lack of large concentrations of serine and
threonine (common soil contaminants). These indications suggest an indigenous
component was obtained for fossils which were intact and which appeared to be
moderately to well preserved. Even the poorly preserved 4.0 Ma fossil had an
indigenous organic signal. Unquestionably, the organic fraction in these fossils has
undergone alteration, but most of them should retain an indigenous fraction judging by
this amino acid data, particularly if the fossil is intact. In order to more conclusively
demonstrate the indigeneity of organic material isolated from this time series of fossil

horses, an evaluation of the carbon and nitrogen isotope values was undertaken.

Isotope Studies of Modern and Ancient Herbivores

Isotope values of modern equids were determined in order to better understand

the natural isotopic variation in horse hard parts. This understanding provides

57

comparative data which may facilitate an indigeneity assessment of fossil organic
matter. Carbon isotope values of the modern terrestrial herbivore tissues measured
range from -20.l%o to -24.2%o (Table 11). The nitrogen isotope range for these same

specimens is 2.3%o to 6.8%o (Table 11).

Procedural Reproducibility

Procedural reproducibility on modern samples was evaluated through replicate
demineralizations and dialyses of five aliquots of the labial half of a single horse
molar (RMl). Of the five replicate demineralizations, three 5'3C and five 5'5N
determinations were made. The procedural reproducibility for modern horse 513C and
5‘5N measurements was 0.2960. Reproducibility for the fossil 5'3C analyses was

typically 0.1%o to 0.3%o. Reproducibility for fossil 8‘5N values was 0.4%.

Testing For Procedural Errors

The three samples which compose the two hour demineralization experiment
(F:AM 129450-1, F:AM 129450-2, and F:AM 129450-3) have mean 5‘3C and 8N
values of -23.6 :h 0.1% and 4.2 i 0.5%, respectively (Table 6). Although the FT
values are not significantly affected by a reduction in demineralization acid exposure
time, the 5'5N values are on average 3.9%» depleted in ”N. This depletion is

accounted for by a larger incorporation of MN-rich low molecular weight material

58

 

 

 

 

 

Table 11 The 5'3C and EN values of modern wapiti and horses.

Sample Number Genus species Element Sampled 613C 815N
MSU: Cer. 1 Cervus canadensis RM] ND 5.0
MSU: Cer. 2 Cervus canadensis RM] -22.9 5.7
MSU: Cer. 3 Cervus canadensis RM] —22.9 5.4
MSU: Cer. 4 Cervus canadensis RdP4 -23.2 5.8
MSU: Cer. 5 Cervus canadensis RdP4 -22.8 6.1
MSU: Cer. 6 Cervus canadensis RdP4 -23.0 5.]
MSU: Cer. 7 Cervus canadensis R Dentary Bone ND 5.1
MSU: Cer. 8 Cervus canadensis R Dentary Bone -22.6 4.9
MSU: Cer. 9 Cervus canadensis R Dentary Bone -23.0 4.0
MSU: Cer. 10 Cervus canadensis Hair -24.0 ND
MSU: Cer. ]] Cervus canadensis Plant Debris inTeeth ND 6.3
MSU: Cer. 12 Cervus canadensis Connective Tissue -24.2 5.3
MSU: Cer. 13 Cervus canadensis Muscle (Masseter) ND 5.0
MSU: Bot. l ? MI Meadow Hay -28.] ND
MSU: Eq. la Eguus caballus RM]. 1A (labial) -21.1 6.]
MSU: Eq. lb Eguus caballus RM]. 13 (labial) -21 .1 5.9
MSU: Eq. 1c Eguus caballus RM1.1C (labial) -20.9 6.0
MSU: Eq. 1d Eguus caballus RM1.2 (labial) -20.8 5.8
MSU: Eq. 1e Eguus caballus RM1.3 (labial) -20.3 5.9
MSU: Eq. 1f Eguus caballus RM] .4A (labial) -20.6 6.]
MSU: Eq. 1g Eguus caballus RM1.4B (labial) ND 5.7
MSU: Eq. 1h Eguus caballus RM1.4C (labial) ND 5.9
MSU: Eq. j Eguus caballus RM 1 .5 (lingual) -21.4 5.7
MSU: Eq. 2 Eguus caballus RM2 -20.5 6.3
MSU: Eq. 3 Eguus caballus RM3 -20.8 6.0
MSU: Eq. 4 Eguus caballus RP2 -20.1 5.8
MSU: Eq. 5 Eguus caballus RP3 -20.2 5.9
MSU: Eq. 6 Eguus caballus RP4 -20.3 5.6
MSU: Eq. 7 Eguus caballus R Maxilla Bone -20.4 5.6
MSU: Eq. 8 Eguus caballus RM] (6N HCl) ND 6.0
MSU: Eq. 9 Eguus caballus RM] (3N HCl) -20.7 6.8
F.M.M. 1219 Eguus sp. R Maxilla Bone -21.8 2.3

 

 

59

remaining in response to the poor demineralization efficiency. Hence, samples which
are not fully demineralized will not provide reliable nitrogen isotope data.
Correspondingly, these three samples are not used when interpreting horse

paleocology.
Isotope Variation Within Modern Materials

Separate tissues from a single organism are often found to be isotopically
distinct from one another (Teiszen et al., 1983). For example, DeNiro and Epstein
(1978b) found the 8'3C values of soft anatomies of mice to be approximately -21.0%o
while bones from the same mouse strain average approximately -18.0%o. Differences
such as these results from fractionation during metabolism. Natural isotopic variation
within a species must be understood before accurate paleoecological assessments can

be made.
Intraspecific Variance in a Juvenile Wapiti

In an effort to understand natural isotope variation among separate tissues from
an individual, various skeletal and soft tissues from a single juvenile wapiti were
measured. The various wapiti tissues range from -22.8%o to -24.2%o in 6‘3C and from
4.0960 to 6.1% in 8‘5N (Figure 1]). Comparisons of 5'3C and 8N values among

tissues from an individual wapiti shows that variation is typically larger for nitrogen

6.5

0’)

Nitrogen Isotope Value
01
01 m

4.5

Figure 1 1

6O

 

 

 

 

 

T

RM1
4? I
__ RdP4

' Connective Tissue
l Dentary Bone Muscle Tissue

_+_ I I
22 22.5 23 23.5 24 24.5 25

Carbon Isotope Value

Natural variation in the 5'3C and 5‘5N values of a juvenile wapiti.

61

isotopes than for carbon isotopes. For example, the mean 8‘3C value for modern
wapiti bone is -22.8 i 0.1%o whereas the mean 5’5N value for modern wapiti bone is
4.7 i 0.6%. Natural isotope variation is small in comparison to isotope effects from
alternate sources such as diagenesis or dietary changes (e.g. transition from a C3 to a
C4-based diet) (DeNiro and Hastorf, 1985; Vogel, 1978). The mean 5'3C value for the
modern juvenile wapiti skeletal materials (teeth and bones) is -22.9 :t 0.2% while the
corresponding mean 5‘5N value is 5.2 i 0.6%o (Figure 11). The large variation in
nitrogen isotope values could be due to changes in the juvenile's diet that occurred
during development. The transition from breast milk to the herbivorous adult diet can
result in natural isotope differences between tissues emplaced at different times.
Enrichment in the 5'5N of milk teeth of 2960 over bone collagen occurs in juveniles
and in species whose teeth stop growing early in life (Bocherens et al., 1994). A lack
of similar variation in 813C and other lines of evidence (e.g. grass found within teeth)
indicate this juvenile was weaned from its milk diet. Nonetheless, the potential
confounding variable that juvenile specimens introduce may make them poor indicators
of natural isotopic variation. Consequently, the natural variation in the carbon and

nitrogen isotope values of modern horse skeletal materials was also investigated.

lnterspecific Variance Between Wapiti and Horses

The carbon isotope values of modern horse teeth and bones range from

-20.1%o to -21.8%o (Table 11). The mean 5'3C value for all modern horse skeletal

62

material is -20.7 d: 0.4% versus -22.9 i 0.2%» for wapiti materials. Natural variation
in 5’3C values of skeletal remains of horses and wapiti are similar whereas the 5'3C
values themselves are different (t-test, t = —12.2, df = 16.2, a = 0.05). A 5'3C
difference of 2.2%o exists between modern wapiti skeletal material and modern horse
skeletal material. This likely reflects the different feeding strategies of wapiti and
horses, wapiti consuming a considerably larger amount of browse (C3 plants). A
"canopy" effect might also explain the 2.2%o depletion given preliminary results which
indicate a 5960 to 6% depleted carbon pool on dense forest floors (Van der Merwe and
Medina, 1989). Alternatively, this isotope difference may be attributed to the age of
the wapiti since a milk diet would serve to deplete the its 5’3C value (Boutton et al.,
1988)

Currently, little information is available concerning the magnitude of depletion
associated with milk-based diets in undomesticated mammals. Studies have shown
that lipids are regularly 7-8%o more negative than other biological fractions (DeNiro
and Epstein, 1977). Whole wapiti milk is assuredly not as depleted as a pure lipid
fraction. Tyrrell et a1. (1984) found dairy cow milk to be 1%» depleted in 5'3C and
Boutton et a1. (1988) found 1.8%o to 2.2%o depletions in bovine 5'3C milk values
compared to the diet. These depletions are consistent with the observation that milk
fat is typically less than 10% in most mammals (Iverson, 1992). The presence of a
substantial amount of plant debris within the selenodont lophs of this juvenile's cheek
teeth indicates a weaned individual. If the individual was recently weaned a milk

signature might persist for some time. This does not appear to be the case, however,

63

since the mean 513C for this individual is similar to the 5'3C values published for
other mature artiodactyls (deer, mainly) (DeNiro and Weiner, 1988; Cormie and
Schwarcz, 1994). The wapiti had not reached full maturity since it had not yet lost its
deciduous premolars. Therefore, the juvenile wapiti appears to be in transition to
adulthood and it's adult diet.

The 5'5N values of modern horse teeth and bones range between 5.6% and
6.3% (Table 1]). The mean 5‘5N values for the modern horse skeletal material
analyzed is 5.9 :t 0.3%o compared to 5.2 i 0.6%» for the wapiti. A statistical
difference in the mean horse and wapiti 5‘5N values exists (t-test, t = -4.0, df = 11.7,
a = 0.05). This result may indicate fundamental differences in the diets of wapiti and

horses in terms of the amount of legumes consumed.

Variation Between Different Horses

Carbon isotope values for modern horses are similar between separate
individuals and within the same individual (Figure 12). A difference in BBC of 1.4%
was observed for bone between a horse from Michigan and another from Nebraska.
The difference in 5‘3C for these two horses is similar to the 1.5%o variance Vogel
(1978) found between individual ungulates with like diets. The 5'3C value for bone
from the modern Michigan horses is -20.4%o while the Nebraska horse bone is

-21.8%o. These values are also in good agreement with the organic 6‘3C values of

64

 

 

 

 

7 , ~
I
, szRP3 ‘ RM]
3 6T ‘ A RP4 RM2 A
- , ‘ ‘ . RM3
g 5 ,_ Maxulla
a 9
3 4 1
3 i
2 3
g A
g 2 ‘* Maxilla
.1:
Z 1 .1
O 4 . t t t + 4 e 4
20 20.2 20.4 20.6 20.8 21 21.2 21.4 21.6 21.8 22
Carbon Isotope Value
Figure 12 The relationship between the 5'3C and 6'5N values of skeletal

tissues (premolar and molar teeth and maxilla bone) from a modem
horse.

65

modern horse bone (-21.2%o) published elsewhere (Schoeninger and DeNiro, 1984).
No such comparison was possible for recent horse teeth in that this study is the first to
characterize modern equid dentine isotopically.

Modern horse nitrogen isotope values were also found to be similar within an
individual (Figure 12). Nitrogen isotope variation between individuals can be large
even for individuals with similar diets (Bada et al., 1990). The nitrogen isotope value
of the modern horse from Nebraska (F.M.M. 1219) differed substantially from the
Michigan horse. The abberant 2.3%o nitrogen isotope value of the Nebraska horse may
reflect consumption of plants with altered nitrogen values due to the influence of
fertilizers which have 5'5N values of about 0.0%: (Schoeninger and DeNiro, 1984).
Alternatively, this horse may have eaten a large amount of legumes such as clover
which have 6‘5N values near 0.0%» (Delwiche et al., 1979; Schoeninger and DeNiro,
1984). A 60% incorporation of total nitrogen with 3 EN signature of 0.0%o will shift
a natural bone value of 5.6%» to 2.3%.

The mean 5‘5N value for modern horse skeletal materials is 5.9 :t 0.3%o
excluding the 23% value which may record fertilizer influences. The mean nitrogen
isotope value of modern horse is similar to the modem mean 5‘5N value for terrestrial
herbivores (5.3 :1: 1.9%) determined by Schoeninger and DeNiro ( 1984). However, the
mean 5‘5N value of the modern horse is slightly more depleted than the expected value
for a terrestrial herbivore consuming a non-leguminous diet. Consequently, the
modern horses analyzed in this thesis likely ate a percentage of clover or some other

leguminous, herbaceous ground cover. Noteworthy is the fact that the 6‘5N values

66

reported herein are consistent with expectations based on modern horse trophic level
position as well as nitrogen isotope values of horses reported in the literature (DeNiro

and Epstein, 1981; Schoeninger and DeNiro, 1984).

V ariability in Horse Tooth and Bone

The mean 513C value for all modern equid teeth is -20.6 :1: 0.4%o. Replicate
analyses of RM] yielded a mean 513C value of ~20.9 :t 0.4%o (n=7). All other adult
molars and premolars had 5'3C values slightly enriched over the RM] mean (Table
10). There is no statistical difference in 5'3C values within molars (t-test, t = -l.2,
df = 3.0, a = 0.05), within premolars (t-test, t = 3.0, df = 1.0, a = 0.05) or between
tooth and bone (t-test, t = 0.7, df = 2.3, or = 0.05). The mean 6'3C values for modern
horse bone and teeth are -21.0 i 0.6%o and -20.6 i 0.4%», respectively. Undoubtedly,
natural isotopic differences exist within tooth types and between tooth and bone, but
these differences are smaller than the reproducibility of the technique. No differences
were observed in the 5‘5N values within molars (t-test, t = -1.4, df = 1.2 a = 0.05),
within premolars (t-test, t = 5.0, df = 1, or = 0.05) and between molars and premolars
(t-test, t = -l.4, df = 4.1, or = 0.05). In addition, there was no difference between
tooth and bone of the same individual (t-test, t = -0.6, df = 2.], on = 0.05).
Determination of the 513C and EN values of horse dentine are significant as a means
to begin distinguishing natural isotopic variation from the potential effects of

diagenesis or dietary shifts recorded in fossil teeth.

67

Isotope Variation in A ncient Horses

Carbon isotope values of the fossil Equidae ranges from -]8.4%o to -26.7%o
(Table 12). Nitrogen isotope values fall between 0.5960 and 14.0%o (Table 12).
Consistent with the modern modern data, no significant carbon or nitrogen isotope
differences were observed between fossil horse premolars and molars (carbon t-test, t =
-0.6, df = 2.9, a = 0.05; nitrogen t-test, t = -] .2, df = 29.9, or = 0.05) or between fossil
horse bone and teeth (carbon t-test, t = -0.8, df = 3.2, or = 0.05; nitrogen t-test, t =
-0.7, df = 4.3, a = 0.05). Within an individual, 6‘3C and 8N values of bone
frequently deviate from tooth isotope values (e.g. F:AM 129316, F:AM 110575)(Table
12). These differences cannot be ascribed to tooth formation prior to weaning as with
the juvenile wapiti resulting in a milk signature for two reasons. First, all but one of
the ancient horses were fully mature (as evidenced by their lack of deciduous teeth and
the wear patterns present on their adult teeth). Secondly, the hypsodont teeth of horses
continue to grow throughout life (Bocherens et al., 1994). Therefore, any milk
signature would soon be replaced by the adult diet signature. The intraspecific isotope
differences between tooth and bone might relate to differential susceptabilities to
diagenesis of these two skeletal elements. This can best be evaluated through a
comparison of the isotope values of modern and ancient skeletal elements from various

horses.

68

 

 

Table 12 The 5'3C and 5'5N values of ancient horse teeth and bones. Italicized
isotope values are presummed to be diagenetically altered or had ion
beams to low to achieve an accurate isotope value.

firecimen Genus species Age (Ma) 6 13C 815N

UALP 7810 Eguus Sp. 3 -23.3 5.9

F: AM 129114-1 Eiuus simplicidens 4 -25.2 1.3

F: AM 1291 14-2 Equus simplicidens 4 -24.9 0.5

F: AM 129444-1 "Dinohjppus” leidvanus 5.5 -23.4 5.7

F: AM 129444-2 "Dinohippus" leidyanus 5.5 -22.9 ND

UNSM 35-568 "Dinohiypus" leidvanus 6 -22.4 4.2

UNSM 316-56-1 Nannippus lenticularis 6 -22.() 5.4

UNSM 316-56-2 Nannippus lenticularis 6 -23.3 ND

F: AM 129443 Neohippan'on elm/styles 6.5 -22.9 5.3

UNSM 5157-73 Hi anon s . cf. forcei 7 -24.5 6.8

AMNH l7599A-l ”Dinohippus" leidvanus 7 -24.5 5.6

AMNH l7599A-2 "Dinohippus" leidym 7 -25.5 ND

UNSM 2010-708 Calippus sp. nov. 7 -24.2 5.6

F: AM 129442 Cormohipparion occidentale 8.5 -22.7 7.2

F: AM 129445 Cormohipparion occidentale 9.5 -23.6 6.0

F: AM 129446 Cormohipparion occidentale 9.5 -24.3 6.1

F: AM 129454-1 Connohipparion occidentale 10 -23.6 6.2

F: AM 129454-2 Cormohipparion occidentale 10 -235 ND

F: AM 129441Max. Connohymm occidentale 10.5 -22.6 4.3

F: AM 129441Max. Cormohipgarion occidentale 10.5 ND 4.2

F: AM 129441LdP4 Cormohipparion occidentale 10.5 -23.1 5.8

F: AM 129441LM1 Cormohipgarion occidentale 10.5 -21.() ND

F: AM 129441LM2 Connohipparion occidentale 10.5 -2 l .3 ND

F: AM 129447 Pliohippus gnu); 12 -19.3 7.4

F: AM 129448 Pliohippus pemix 12 -23.() ND

F: AM 129450-A Pliohippus pemix 12 -22.9 8.5

F: AM 129450-B Pliohippus p_emix 12 -22.8 6.6

F: AM 129450-C Pliohippus pemix 12 -22.7 8.6

F: AM 129450-D Pliohippus pernix 12 -23.3 7.1

F: AM 129450-1 Pliohipgus mix 12 -23.4 4.1

F: AM 129450-2 Pliohippus pcmix 12 -23.2 4.5

F: AM 129450-3 Pliohippus pernix 12 -24.2 3.9

F: AM 1 14068 Pliohippus pernix 12 -25.7 ND

F: AM 129316 LP3 Merychippus insigm 15 -21.1 1.6

F: AM 129316 Max. Mgchippus insignis 15 -.?5.0 14.0

F: AM 129453 Mervchippus 5p. (cf. isonesus) 15.5 -21.1 6.9

F: AM 110575 LP4 "Megchippus" primus 17 -25.5 8.5

F: AM 110575 Max. ”Merychippus” primus 17 -18.4 10.0

AMNH 20513 ”Mmchippus" primus 17 -22.6 ND

F: AM 129451 "Mervchfiirpus" tertius 17.5 -26.2 8.3

F: AM 128784 RM2 "Parahippus" .512. 18.5 -27.5 9.8

F: AM 128784 Max. "Parahippus" 3;. 18.5 -26.7 ND

F: AM 129455-1 Mesohippus s12. 30.5 -25.4 6.2

F: AM 129455-2 Mesohippus s12. 30.5 -25.3 ND

F: AM 74067 Miohippus obliguidens 32 -26.6 ND

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

69

Comparisons of Modern and Ancient Isotope Values

With an understanding of natural isotopic variability in modern and ancient
horses, inferences about the indigeneity of organic matter extracted from fossil bones
and teeth are possible. Isotope fluctuations not attributable to natural variation may be
due to diagenetic alteration of the isotope-signal or to actual changes in the paleodiet
of the horse in question. Evidence for indigeneity of ancient organic matter includes:
1) Similarity in the distribution and relative abundance of amino acids of the fossil
compared to the modern pattern 2) Percent yields of organic matter obtained from the
fossil which are consistent with the range of yields from other studies 3) Similarity in
the BBC and 5'5N values of the fossil compared to a modern analog. Fossils which
conform to these criteria likely retain an indigenous organic matter fraction. Fossils
which do not conform to these criteria have undergone some amount of diagenetic
alteration and consequently are not good candidates for reconstructing equid
paleoecologies. Amino acid and percent yield data used to assess indigeneity of fossil
horse organic matter has previously been discussed. Stable isotope data is particularly
important as an independent method for verifying indigeneity.

Modern and fossil isotope values of the organic matter from terrestrial
herbivores were compiled from the literature for comparative purposes (Figure 13).
Values of herbivores from arid and non-arid regions are reported separately since 6'5N
values may change as a function of aridity (Heaton et al., 1986; Ambrose and DeNiro,

1989). In general, the isotope values of the fossil horses fall within the range of

70

.m_0>0_ 0:56,: 0:9: :23 000.0 8 023:5... 00 20:82 25

E E20006 coco 0.3 0.033 22m 00:33 mowqfl 8000006 02E 0:230
28-8: 05 EB 50¢ 00829.0: .030 0:: E 005205

0020: E065 05 wig—0E 008390: 3:000:00 E0008 00:23.5 05

 

 

 

 

    

 

 

 

.00 €00.an 0 820 0:2: 2:090 .«o 0.02: 220 0:0 Dem 00 0095; 2 05m.»
020
m. o w. m 7 ON. mm. on. mm.
2 z z _ z _ o
008E=0 DIE- coz iEdt. 00.0290:
6 w020> 002053”. 00:0,”.
1 m
9L
1 9
or N
as. 0.0 . as. 0.3 I
a: 0.: - as. 0.8 E 020820 0_.._<. E0: 0029280:
£00» 85: 20.05. B 0020> 0030.530. 35». 1 mp
mafiivmvwwwmmmmmoaom. 50.1004 fiwmmi no 00 0mfiMwmlouolmi_ c0903

 

 

 

 

ON

71

modern values of herbivores from arid and non-arid regions. Depleted 5'3C values of
four ancient horses (F:AM110575, F:AM 128784, F:AM 129451, F:AM 129455) likely
reflect a higher dietary contribution of C3 plants. These horses are 17.0 Ma and older
and likely existed before the widespread evolution of C4 plants. Two horse bones
(F:AM 129316, F:AM110575) have 513C and S'SN values which are not consistent
with those of modern herbivores (Figure 13). These specimens may exhibit diagenetic
alteration. Supporting this contention, bones and teeth from the same individual may
have different isotope values. Three ancient bones (for which 6‘5N values of tooth and
bone from the same horse were determined) have 5‘5N values which differ by 15960 or
more from tooth values. Bone appears less likely to retain an indigenous organic
fraction owing to its greater porosity compared to teeth. The evaluation of these bones
as diagenetically altered precludes their use as reliable paleoecological indicators. The
possibility exists that low 8'5N bone values record a dietary contribution of legumes.
Similarly, high 5‘5N values could document a sick organism in nitrogen stress or an
individual from an especially arid locality (Heaton et al., 1986; Ambrose and DeNiro,
1989). Owing to the extensive isotopic alteration present in these bone samples, it is

unlikely that any causal factor other than diagenesis is indicated.

Evaluation of Diagenetic Effects

An initial aim of this research was to attempt to quantify diagenesis in the

fossils which were deemed to be altered in some regard. A time series of fossils from

72

similar depositional environments should have provided insight into diagenetic
processes. However, quantification of diagenesis was not possible because some
fossils from the same deposits exhibited very different degrees of alteration. For
example, a maxilla bone (F:AM 110575 LMax.) from Thompson Quarry was

7.1960 enriched in 5‘3C and 1.5%o enriched in 5‘5N relative to a tooth (F:AM 110575
LP4) from the same location. Additional effort must be placed on obtaining samples
with precise stratigraphic control within a quarry to best investigate diagenetic effects.
Geochemical characterization of the associated sediment would also aid in
understanding diagenesis. Provided precise depositional and stratigraphic information
is attainable, isotope studies may be useful for correlating the extent of postmortem
processing with one or more environmental parameters. This would enhance our
understanding or organic matter preservation in fossils. Conceivably, altered fossils
may be shown to be useful for paleoecological reconstructions if diagenesis can be
accurately modeled and predicted. As this was not possible for this study, the

diagenetically altered samples were not used for evaluating equid paleoecology.

Interpretations of Tertiary Paleoecology of the Horse Family

Percent C 4 in Horse Diets

The first step towards an understanding of Tertiary paleoecology and the

evolution of hypsodonty in equids is to assess the abundance of siliceous grasses in

73

the diets of various horses. This assessment relies on two assumptions: 1) Knowledge
of the BBC values of the dietary constituents; 2) The degree to which these
endmember values are fractionated by the horse. Mean 5‘3C values for modern C3 and
C, plants are well characterized (-27%o and -l3%o, respectively). A modern C3 plant
eaten by horses was measured to determine if grasses common to horse diets were
similar to the 5‘3C mean of C3 plants. Meadow hay from New Baltimore, Michigan
was found to have a 513C value of -28.1%o. Therefore, mean BBC values should be
useful when modeling the amount of C, vegetation in horse diets. Metabolic
fractionation is not well constrained in horses, so several representative values for
other mammals were used to model equid fractionation (DeNiro and Epstein, 1978b;
Vogel, 1978; Schoeninger and DeNiro, 1984). The component of C, plants in the diet
of all fossil and modern horses assuming four separate theoretical equid 513C
fractionations equal to 1%o, 3%o, 4%o and 5960 were calculated using mixing equations
(Table 13). Carbon isotope fractionations of about 1%: to 2960 are characteristic of
small mammals (DeNiro and Epstein, 1981) and are probably too small for herbivores
as large as horses. The largest terrestrial herbivores may have 5'3C fractionations of
5.1960 to 5.3%» (Ambrose and DeNiro, 1989) and possibly as high as 6.1%: (Vogel,
1978). Thus, carbon isotope fractionation in horses is likely on the order of 3% to
4%o. The relative abundance and distribution of C, plants in modern environments in
relation to the percent C, values determined for modern horses can be used to model

the appropriate fractionation associated with horses. The percentage of C, plants

Table 13

74

Percent C, vegetation in the diet of modem and ancient terrestrial

herbivores assuming four separate fractionations. Based upon modern C,
distributions in North America, horse metabolic fractionation is on the
order of 3%o to 4%o.

 

 

 

Age °/o C4 °/o C4 % C4 % C4
613C Sample Number Genus species (Ma) 6‘ = 1 G = 3 G = 4 G = 5
-22.93 MSU: Cer l Cervus canadensis 0 21 9 7 6 0 5 (l 0
-22.94 MSU: Cer 1 Cervus canadensis () 21.9 7.6 0.4 0.0
-23.18 MSU: Cer 2 Cervus canadensis (1 20.1 5.9 0.0 0.0
-22.79 MSU: Cer 2 Cervus canadensis 0 22.9 8.6 1.5 0.0
-22.99 MSU: Cer 2 Cervus canadensis 0 21.5 7.2 0.1 0.0
-22.59 MSU: Cer 3 Cervus canadensis 0 24.4 10.1 2.9 0.0
-22.98 MSU: Cer 3 Cervus canadensis 0 21.6 7.3 0.1 0.0
-23.98 MSU: Cer 4 Cervus canadensis 0 14.4 0.1 0.0 0.0
~24.20 MSU: Cer 6 Cervus canadensis 0 12.9 0.0 0.0 0.0
-24.50 MSU: Cer 7 Cervus canadensis 0 10.7 0.0 0.0 0.0
-20.89 MSU: Eq lc Eguus caballus 0 36.5 22.2 15.1 7.9
-20.77 MSU: Eq 1d Eguus caballus 0 37.4 23.1 15.9 8.8
-20.31 MSU: Eq 1e Eguus caballus 0 40.6 26.4 19.2 12.1
-20.60 MSU: Eq 1f Eguus caballus 0 38.6 24.3 17.1 10.0
-21.39 MSU: Eq lj Euus caballus 0 32.9 18.6 11.5 4.4
-21.07 MSU: Eq 11; Eguus caballus 0 35.2 20.9 13.8 6.6
-21.06 MSU: Eq 11 firms caballus 0 35.3 21.0 13.9 6.7
-20.52 MSU: Eq 2 Eguus caballus 0 39.1 24.9 17.7 10.6
-20.83 MSU: Eq 3 anus caballus 0 36.9 22.6 15.5 8.4
~20.06 MSU: Eq 4 guns caballus 0 42.4 28.1 21.0 13.9
-20.20 MSU: Eq 5 Eguus caballus 0 41.4 27.1 20.0 12.9
-20.31 MSU: Eq 6 Eguus caballus 0 40.6 26.4 19.2 12.1
-20.37 MSU: Eq 7 @uus caballus 0 40.2 25.9 18.8 11.6
~20.70 MSU: Eq 9 E11115 caballus 0 37.9 23.6 16.4 9.3
-21.80 F.M.M. 1219 Max. Eguus sp. 0 30.0 15.7 8.6 1.4
-23.30 UALP 7810 Eguus Sp. 3 19.3 5.0 0.0 0.0
-25.20 F: AM 129114-1 @uus simplicidens 4 5.7 0.0 0.0 0.0
-24.90 F: AM 129114-2 Eguus simplicidens 4 7.9 0.0 0.0 0.0
-23.40 F: AM 129444 "Dinohippus" leidyanus 5.5 18.6 4.3 0.0 0.0
-22.40 35-5613 "Dinohippus' leidyanus 6 25.7 1 1.4 4.3 0.0
-21.98 316-56-1 Nannigpus lenticularis 6 28.7 14.4 7.3 0.1
-23.30 316-56-2 Nannippus lenticularis 6 19.3 5 .0 0.0 0.0
-22.87 F: AM 129443 Neohippa_n'on eggtyles 6.5 22.4 8.1 0.9 0.0
-24.50 5157-73 Hipm'on sp. (cf. forcei) 7 10.7 0.0 0.0 0.0
-24.45 AMNH 17599A-1 ”Dinohippus” leidyanus 7 1 1.1 0.0 0.0 0.0
-25.51 AMNH 17599A-2 "Dinohippus" leidyanus 7 3.5 0.0 0.0 0.0
-24. 16 2010-708 Calippus §p_. nov. 7 13.1 0.0 0.0 0.0

75

Table 13 (cont'd)

 

 

 

 

Age °/o C4 o/o C4 °/o C4 o/o C4
613C Sample Number Genus species (Ma) Q = 1 E = 3 G = 4 G = 5
-22.67 F: AM 129442 C. occidentale 8.5 23.8 9.5 2.4 0.0
-23.60 F: AM 129445 C. occidentale 9.5 17.1 2.9 0.0 0.0
-24.26 F: AM 129446 C. occidentale 9.5 12.4 0.0 0.0 0.0
-23.58 F: AM 129454-l C. occidentale 10 17.3 3.0 0.0 0.0
-23.50 F: AM 129454-2 C. occidentale 10 17.9 3.6 0.0 0.0
-22.60 F: AM 129441Max C. occidentale 10.5 24.3 10.0 2.9 0.0
-23.10 F: AM 129441LdP4 C. occidentale 10.5 20.7 6.4 0.0 0.0
-20.96 F: AM 129441LM1 C. occidentale 10.5 36.0 21.7 14.6 7.4
-21.30 F: AM 129441LM2 C. occidentale 10.5 33.6 19.3 12.1 5.0
-19.30 F: AM 129447 Pliohippus mix 12 47.9 33.6 26.4 19.3
-23.00 F: AM 129448 Pliohippus mix 12 21.4 7.1 0.0 0.0
-22.75 F: AM 129450-B Pliohippus mix 12 23.2 8.9 1.8 0.0
-25.70 F: AM 114068 Pliohippuspemix 12 2.1 0.0 0.0 0.0
-21.10 F: AM 129316 LP3 Mgchippus insignis 15 35.0 20.7 13.6 6.4
-25. 00 F: AM 129316 Max. Megchippus insignis 15 7.1 0.0 0.0 0.0
~21.07 F: AM 129453 Machippus s2. 15.5 35.2 20.9 13.8 6.6
-25.50 F: AM 110575 LP4 "ngchippus" primus 17 3.6 0.0 0.0 0.0
-18.40 F: AM 110575 Max. "ngchippus" Qn_t_n' us 17 54.3 40.0 32.9 25.7
-22. 60 AMNH 20513 ”Mmehippus" primus 17 24.3 10.0 2.9 0.0
-26.18 F: AM 129451 "Me_rychippus" tertius 17.5 0.0 0.0 0.0 0.0
-27.45 F: AM 128784 ”Parahippus" g2. 18.5 0.0 0.0 0.0 0.0
~26.70 F: AM 128784 "Parabippus" 52. 18.5 0.0 0.0 0.0 0.0
-25.40 F: AM 129455 Mesohippus sp. 30.5 4.3 0.0 0.0 0.0
-25.33 F: AM 129455 Mesohippus sp. 30.5 4.8 0.0 0.0 0.0
-26.60 F: AM 74067 Miohippus obliguidens 32 0.0 0.0 0.0 0.0

76

in various regions throughout North America is known (Figure 14). The modern day
percent distributions of C, plants in the Midwest and Nebraska most closely parallels
the 3960 and 4%o fractionation models (Table 13, Figure 14). Having established a way
to assess percent C, vegetation in horse diets, an evaluation of the paleoecological

significance of the data can then be made.

Temporal and Spatial Distibution of Isotope Data

Despite being constrained to the Great Plains, some of the horses which
compose this time series are from distant geographical regions (e. g. Nebraska and
Texas). An investigation of 5'3C values in association with age and locality
information aids in interpreting the stable isotOpe data (Figure 15). A striking feature
of viewing data in pictoral format is the perspective it provides concerning the
distribution of the data points. The relative proximities of the horses (both
geographically and temporally) is another salient feature of viewing data pictorally.
Without this map, the fact that 85% of this time series is constrained to western
Nebraska might easily go unrecognized. The distribution of data points in space and
time are fundamental concerns of the interpretive aspect of science. Approaching the
isotope data with a visual perspective facilitates interpretations and enhances

paleoecological assessments.

77

 

 

 

Figure 14 The percent distributions of modem C, plants in various regions of
North America (modified from Teeri and Stowe, 1976).

78

 

 

,_.__— _ -—

.:oc0E.o-«E 00:30.30» 0:0 0m0 at? 020060000 E 02 o.m

 

 

 

 

 

 

 

 

 

     

8 02 onV 0008: 5000.02 552 2.06:0 .«o 002g 380$ 03:5 2 05wE
0.0 0.00- /
0.0. 0. P0- 0.0 0.00- / .
0.2 o 0 0.2 o 0 \ . 0 0 0 00-
MF 0P \\ \1 \\\.As.. /.
- 0 /1-\../. 0.2 00. 0
v
M0 “.00- ,3. 05.00 \

2 0P 0 /,/
1 - .-/ 80.0.2 .502
0.00 0.00- 0.0. 0.00- (20050.0
0.00 v.00- 3: 0.00- 11 5209....

. . 9525.
0.9 0.00- m m m .00- / / 005.98
0.: 0.00- 0.0 0.00- /

. . (xwémwz 1
0.: 0.00- 0 0 0 00-
0.9 P. E- as New- \ 02.20:.) v
0.00 P. .0. 0.0 0.00- \ «5020 :50- -11 Z
0.0. 0.00- 0.0 0.00- .\ 111-. 11
0.0. 0.00- 0.0 0.00- \..
0.0. 0.0.- 0.0 0.00- .\
m5. 0 m P m0 as. O m.- m0

 

 

79

Isotope Changes Over Time

If horse diets changed from C3 ~based to C, -based any time in the last 32.0 Ma
this signature should be recorded in the 513C values of this time series. Linear
regression of the 5'3C values (r2 = 0.01) versus time reveals no correlation between the
isotopic signature and age of the fossils (Figure 16). Similarly, no trend exists
between 5‘5N values and time (r2 = 0.15; Figure 16). The lack of an obvious trend in
BBC and EN versus time is not at all surprising given the wide array of species and
Great Plains depositional localities.

Rather than looking for linear temporal trends, a better way to interpret the data
is to divide it into separate populations based on traditional paleobotanical
assessments. Before about 17 Ma savanna grasslands composed only a small fraction
of the Tertiary landscape (Axelrod, 1985). The mean 513C value of horses from this
study that are 17 Ma or older is -26.2 i 0.7%- which is consistent with a pure C3 diet
given the range and mean BBC values for modern C3 plants. The oldest C, grass
known is from the latest Miocene (7 to 5 Ma) (Thomasson et al., 1986). Therefore,

7 Ma is another convenient place to subdivide the time series. Prior to 7 Ma, but
subsequent to 17 Ma, the mean 813C value of horses is -22.8 1 1.8%». Although not a
large component (~25%) of horse diets, the percent C, mixing equations used in this
thesis suggests that C, plants were present by somewhere between 15.5 Ma and 12 Ma
on the North American Great Plains (Table 13). After 7 Ma, when C, grasses are first

documented in the fossil record, the mean 6‘3C value of horses is -23.7 at 10960. If

 

21 11

Carbon lootopo Vat-Io

 

80

 

 

 

 

 

 

 

 

 

14

12‘

 

Nltogon lootopo Vain

 

 

 

 

M0 (Ma)

 

Figure 16

The relationship between 5'3C and 8N values of ancient horse organic
matter and geologic age of the specimens.

81

anything, there appears to be a decrease in the amount of C, vegetation in the diet of
these Great Plains horses after 7 Ma. These findings are not in agreement with some
lines of evidence concerning the distribution and prevalence of C, grasses during

the Tertiary (Cerling, 1992; Cerling et al., 1993; Wang et al., 1994). However, the
appearance of C, vegetation by 15.5 Ma to 12 Ma on the Great Plains closely agrees
with age estimates determined for the first occurrence of C, vegetation in South

America (MacFadden et al, 1994) and Kenya (Morgan et al., 1994).

Comparisons with Other Isotope Studies

Savanna-type grasslands of limited extent first appear in East Africa about
9 Ma to 8 Ma based on paleosol isotope data (Cerling, 1992). Cerling et al. (1993),
analyzing 513C values of carbonates and palaeosols and mammalian tooth enamel,
provided evidence for a C, expansion presumably at the expense of C3 vegetation at
approximately 7 Ma to 5 Ma in North America and Pakistan. A C, proliferation did
not occur between 7 Ma and 5 Ma on the Great Plains based upon carbon isotope data
from the organic material preserved in fossil horse teeth (Table 13). Even if horse
diets were to record a C3/C, shift there is no reason to assume all equids would change
uniformly. A shift in 6'3C values of all horses between 7 Ma and 5 Ma would only be
expected if C, grasses rapidly radiated blanketing the Great Plains in a short period of
time providing a new food resource for horses, a situation which has little evidence

currently.

82

Researchers have suggested that a global drop in atmospheric CO2 levels may
account for the apparently rapid appearance of C, dominated habitats in North America
and Pakistan (Cerling et al., 1993). A subsequent study of the enamel carbonate of
prehistoric herbivores from Pakistan and Kenya was not consistent with a decrease in
global CO2 levels (Morgan et al., 1994). In that study, Morgan et a1. (1994) provided
evidence for the existence of C, grasses by 9.4 Ma in Pakistan and 15.3 Ma in Kenya.
These ages are too old to be ascribed to a late Miocene global drop in atmospheric
carbon dioxide levels. Another recent study of the inorganic tooth enamel carbonate
from a subset of North American horses reports a change in carbon isotope values at
approximately 7 Ma to 5 Ma (Wang et al., 1994) (Figure 17). This evidence in
association with enamel carbonate isotope data from South American ungulates led
MacFadden et al. (1994) to postulate the appearance of C, grasses 10 million years
earlier in the Southern Hemisphere. Obviously, the late Miocene global CO2 issue has
raised considerable interest. Consequently, a re-examination of some of the 813C data
upon which the CO2 hypothesis rests might facilitate a better understanding of this

controversial topic.
Reinterpreting Previous Studies
A spatial evaluation of prehistoric horses studied by Wang et a1. (1994) reveals

an extensive geographical distribution which is not as apparent when the data is

viewed in tabular form. Specimens are located in the southeast, along the western

83

. 0.0..
mKP
0K?
m.m.-
m6?
0.).

900
0.00
0.00
0.00
0.2

 

.360. ...0 00 0:03 82-. 00.0695 20:09.00

 

 

 

 

 

 

 

 

 

 

 

 

 

.0805 00.3: 2.0.05 00 0020., 0:0 0:0 000 620000. .00.:0000000 0. 050....
0.0 1.9 0.: 0.0. 0.0 0.0
0.0 0. F- 0.: 0.0. 0.0 0.0
10.0-11-1.11. 1 0.2 0.0 0.? 0.. 0.0
_ Xv /. 2 J 00.0
0.2 0 m0 .. 0.. . 4...-.111111111 0.0 F.0- Né F...
9 .. ham // .//
1 1 II/ /. \x . m m:
1 1 1 1.111111... /\.v\: /1\\.//./// 2 00—.0 OWPQ
0.0. 0.0- / fl ./ ...,.,,.-.--.11 -Hu.
050.0 / ,,
o.N—. m.m| / //
0.0- . . ,
. - . . 0000.2 2.02
0 or m m h or- $201506 (200.5
0.0- 0.0 0.07
. - . 5209.3
0 0 0.0 0 0. 0025.
0.0- 0.0 0.0- 09098
/ +25
0 n 0.2 0 m. \2/ 5.03.002 <0<>mz
- .my 11 111W. ;\ r11
1 - 1 1111 -11 02.295. 11--
Q00 $2.- 7 \ 5020 :5
Ti- 0.00 0. P T \ /
0 3- 0 00 0 :- .\ 0.0. 0.0- 0.: 0.0.
- 07 o 00 0 9- \ 0.9 0.0. 0.0 0.0-
0.0 T o 00 0.0 T \ 0.9 od- o.m F. :-
0 m. 0.2 0 m 0 0
00 9 .2 02.0 .2 00%

 

 

84

coast and on the Great Plains of North America. Today these regions are highly
disparate in terms of climate with extensive desert and scrubland in the southwest and
a more variable climate regime in the central Great Plains region. These localities
differ not only in minimum, maximum and mean temperature, but also precipitation,
soil moisture content (owing to differences in soil water retentions), light intensity,
degree of shading, seasonality, latitude, altitude and a whole host of other climate
variables. These conditions are the same envionmental parameters which influence the
distribution of C3 and C4 plants in modern ecosystems (Teeri and Stowe, 1976).
Variable climatic conditions similar to today inevitably existed during the latest
Miocene and Pliocene. The presence of fossil reptiles and amphibians (Egelhoff and
Norden Bridge Faunas) in regions well outside modern ranges and a dramatic reptile
turnover indicate variable climates throughout the late Tertiary (Holman, 1973;
Holman 1982). Indeed, paleoclimates may have been as variable as in modern times.
Consequently, it seems very unlikely a single parameter such as an atmospheric CO2
decrease would control Tertiary vegetational changes. In addition, a global drop in
carbon dioxide levels would serve to cool global temperatures offsetting any inherint
advantage C4 plants would enjoy under low pCO2 conditions. Moreover, the isotope
data of Wang et a]. (1994) does not support the global CO2 hypothesis upon closer
inspection.

Enamel carbonate 513C values below -12%o are indicative of C3 dominated
diets. Carbonate BBC values between -1%o to +196» indicate pure C4 diets.

Consequently, equal portions of C3 and C4 plants in a horses diet will result in

85

carbonate 5‘3C values near -5%o to -6%o. The only horses of Wang et a1. (1994)
which show a clear diet shift between 7 Ma and 5 Ma are from the arid soutnwestem
region of North America (Ocate, New Mexico not Ocote, Mexico as published; SMU-
uncat). If all specimens younger than 7 Ma are considered, only five more individuals
from Arizona (UA 12/526, UA 52/1539, UA 7352, UA 7354/6552, UA 25-0/718A)
show the diet shift. Noteworthy, the 7 Ma to 5 Ma shift does not begin from a pure
C3 diet immediately prior to 7 Ma.

Both inorganic and organic 513C values reveal a small percentage (20-30%) of
C4 vegetation in horse diets as early as the middle Miocene (~15.5 Ma to 12.0 Ma).
Due to limitations imposed by the mass balance model assumptions (e.g. mean 5'3CC3
5'3CC4, etc.), only percent contributions of C4 vegetation larger than 7% are considered
significant. Grasslands containing C4 plants may well have existed prior to 15.5 Ma,
but would have constituted less than 1/10 of the vegetation in horse diets.
Contributions of less than 10% C4 grasses in the diet of ancient horses are not likely to
have had a major influence on tooth morphology.

None of the Great Plains horses in the 7 Ma to 5 Ma time range possess C4
dominanted diets. The Nebraska horses (AMNH 17599A and F :AM 128963) from
7.5 Ma to 3.3 Ma clearly have C3-based diets. Furthermore, only one Texas horse
(SMU 70533) and one Florida horse (UF uncat. 3A) exhibit transitional diets.
Consequently, isotope data of Wang et a1. (1994) does not demonstrate a diet shift
between 7 Ma and 5 Ma and thus does not support a decrease in global CO2 levels in

the late Miocene as proposed. Rather, the isotope data only indicates an abundance of

86

C4 vegetation in the southwest region of North America after 7 Ma, a condition likely
attributable to elevated temperature and/or aridity of this localized region. The
absence of an abrupt C3/C4 transition in the Great Plains horses of Wang et al. (1994)
is consistent with the findings of this thesis. This suggests that distributions of C4
plants were far more provincial than previously thought.

Three horses measured by Wang et al. (1994) (AMNH l7599A, F:AM 128784
and F :AM 114068) were also analyzed in this thesis. The 5'3C values of both the
enamel carbonate and organic matter from these horses predict similar C3-based diets.
This result supports the view that both isotope methods can provide reliable

paleoecological data.

Paleoecological Ram ifications of Equid Isotope Data

All the available evidence regarding the rise of the C4 grassland biome during
the Tertiary points to a provincial history. Plants utilizing the C4 biochemistry have
evolved independently in thirteen families (James A. Teeri, personal communication).
A phylogeny such as this is best explained by adaptation to local envionmental
conditions, not global changes. The C4 pathway may have evolved first in the
Southern Hemisphere if tropical climate conditions began earlier in the southern
latitudes. However, the first appearance of C4 vegetation seems to be nearly
contemporaneous for both North America (this study), South America (MacFadden et

al, 1994) and regions within Africa (Morgan et al., 1994). This suggests that

87

conditions favorable to the C4 biochemistry (e.g. aridity, warmer temperatures, etc.)
were expanding during the middle Miocene. A multitude of herpetological,
paleopedological and paleobotanical evidence agrees with this contention (Holman,
1971; Retallack, 1983a; 1983b). Despite the ongoing large scale climate trends, local
environmental conditions have been far more important in determining the abundance
and distribution of C4 grasses throughout their evolutionary history. Were CO2 the
only controlling variable, C4 vegetation would be far more pervasive between 7 Ma
and 5 Ma than the fossil record indicates. In fact, sound evidence in support of a
widespread C4 vegetation radiation at the end of the Miocene has yet to be
demonstrated. Recent research indicates that C4 grasses were minor constituents in
Greece throughout the last 11 Ma (Quade et al., 1994). This is consistent with a
provincial distribution for C4 grasses throughout their history.

Savanna grasses began with patchy seasonal distributions in local favorable
environments. As these environments expanded so did the territory of the C4 grasses.
Only in restricted high temperature and aridity regions did the C4 grasses proliferate at
the expense of the C3 grasses (Tieszen, 1979b). Whereas the radiation of the first
grasses (C3) can rightfully be called an explosion, C4 grasses diversified rather
unimpressively. This interpretation is consistent both with the meager fossil record of
C, grasses as well as the stable isotope record contained within fossil teeth and soils
(Thomasson et al., 1986; Morgan et al., 1994; Quade et al., 1994).

Siliceous C4 vegetation was present during the middle Miocene as traditionally

supposed (Simpson, 1951). But, the first occurrence of C4 plants does not precisely

88

coincide with the advent of hypsodont dentition in North American horses.
Merychippine high-crowned horses are known from the early Miocene (~18 Ma) at least
2 Ma to 3 Ma before C4 vegetation is first documented. Hence, C4 grasses cannot be
the main cause for the evolution of hypsodonty in horses. The cause might then be
the transition from a browsing to a grazing diet consisting of C3 grasses. The
incorporation of grit particles most likely influenced the evolution of horse hypsodonty
as well.

The most primitive hypsodont horse, "Merychippus" gunteri, arose between
17.7 Ma and 16.2 Ma (MacFadden et al., 1991). "Merychippus" gunteri is thought to
be ancestral to "Merychippus" primus (MacFadden and Hulbert, 1988). These horses
are often assumed to be the first equid grazers. If so, they must have been consuming
abrasive C3 grasses off gritty substrates. Grazing ungulates were not uncommon in the
middle Miocene. For example, Teleoceras (an extinct hypsodont rhinoceros) has been
discovered with siliceous grass anthoecia contained within the teeth presumably
indicative of a grazing diet. Some authors justifiably caution against the assumption
that all hypsodont ungulates were grazers (Janis, 1988; Hulbert, 1993). Digestive
strategy (e.g. hindgut versus foregut fermentation), substrate quality and plant
abrasiveness all influence whether or not high-crowned teeth evolve. Possibly the
most important prerequisite for hypsodonty is preadaptation of enamel microstructure
(Pfretzschner, 1993). The parahippine ancestor which gave rise to the merychippine
linneage had an enamel microstructure capable of undergoing the transition to the

hypsodont condition (Pfretzschner, 1993). Not all ungulates possess this fortuitous

89

enamel microstructure. Presumably, hypsodonty will evolve in response to an abrasive
grazing diet provided the precursor enamel structure exists. All members of the
Equinae possess this prerequisite enamel microstructure (Pfretzschner, 1993).
Therefore, it is not unreasonable to assume a grazing diet led to equid hypsodonty in
light of the isotope data generated by this thesis as well as recent research on tooth
wear patterns in horses (Hulbert, 1982) and other mammals (Walker et al., 1978). A
diet of predominantly siliceous C3 grasses (with associated grit) thus appears to have
been the major causal factor in the evolution of equid hypsodonty.

The presence of C4 vegetation on the Great Plains after about 15.5 Ma to
12.0 Ma served only to perpetuate the runaway selection toward higher crown heights
as equids continued to graze on the diversifying savanna grasses. In some local
favorable environments C4 grasses may have dominated year round. However, C4
grasses did not achieve modem-day savanna distributions until the Pleistocene
(Cerling, 1992). Organic and inorganic carbon isotope data suggest that C4 plants
never comprised more than 40% of the vegetational biomass consumed by Great Plains
horses throughout the Tertiary (32 Ma to 3.3 Ma). This realization is surprising given
the dominance of C4 grasses in some modern environments.

Several reasons might explain a maximum of 40% C4 vegetation in horse
paleodiets. First, evidence exists that demonstrates some herbivores have an aversion
to C4 plants (Caswell et al., 1973; Vogel et al., 1986). Some biochemical
intermediates (e.g. malate, aspartate, etc.) are present in C4 plants that do not exist in

C3 plants. However, these are generally short-lived intermediates which do not

90

accummulate to any extent in plant tissues (Ralph E. Taggart, personal
communication). There are few if any well documented cases where C, plants
sequester secondary metabolites to deter herbivory. Horses prefer to eat clover over
crab grass given a choice (personal observation), but this is not selective herbivory so
much as dietary preference of the horse. To an extent, herbivore preference will
account for the amount and kinds of grasses consumed (Tieszen et al., 1979a).
However, resource partitioning among megaherbivores and grass abundances also
determine which food type is consumed.

Selective herbivory in horses might evolve if C, vegetation was of poorer
nutrional quality than C3 grasses. Quite a bit of evidence suggests that C, plants are
indeed nutritionally a less valuable food resources to a wide range of taxa (Bryant et
al., 1989; Maschinski and Whitham, 1989). Relative to C3 plants, C, have a lower
nitrogen content indicating the different nutritional values of these two plant types
(James A. Teeri, personal communication). Grasses utilizing the C, pathway produce
four to five times less photosynthetic products and quickly transport their synthesized
starches and sugars to their stem and roots thus reducing the availability of high-
energy metabolites (Steve Stephenson, personal communication). As compared to C3
grass, a larger percentage of C, grass high in silica, lignin and cellulose remains
unprocessed following digestion of hindgut fermenters like horses (Janis, 1989). These
observations suggest that C, grasses are an energetically and nutritionally poor food
source. Yet, African zebras live quite well on C,-based diets today even in regions

where C, grasses are available (Vogel, 1978). Thus, there does not seem to be much

91

evidence in favor of C, avoidance among the Equidae.

Another possibility why most fossil horses do not exhibit pure C, diets relates
to seasonal influences of available vegetation. Modern Nebraskan grasslands contain
about 40% C, species (Teeri and Stowe, 1976). The relative abundance of C3 and C,
grasses is seasonally controlled. Vogel et al. (1986) found that C, species dominate all
of southern Africa except in the winter rainfall areas. If horses grazed on essentially
pure C3 diets during the wet season and ate 100% C, grasses during the drier summer
months, 5‘3C values would be intermediate depending on how long the wet and dry
seasons lasted. In other words, C, vegetation may never have been available more
than 40% of the year on the Great Plains through the Mic-Pliocene. Migration in
horses would bring them into contact with new types of vegetation. Unless migrating
horses traveled through the harshest climates, an increase in the available abundance of
C, vegetation would not occur. In all likelihood, the contribution of C, in the diet of
fossil horses roughly records the percentage available in the horses home range.
Consequently, it is fair to assume that C, grasses were seasonally controlled, locally
important, but far less widespread on the Great Plains during the Mio-Pliocene than

has recently been hypothesized.

92

CONCLUSIONS

Based upon stable isotopic evidence from the organic remnant of horse tooth
and bone the rise of the savanna grassland biome on the Great Plains of North
America began between 15.5 Ma and 12.0 Ma. These ages are consistent with
traditional interpretations, but disparate with recent isotope studies of inorganic equid
enamel carbonate which purportedly provided evidence for largescale savanna
grassland proliferation between 7 Ma and 5 Ma attributed to a global CO2 drop at the
end of the Miocene. Mass balance calculations demonstrate that C, vegetation
comprised fewer than 40% of the biomass consumed by Great Plains horses between
32 Ma and 3.3 Ma. In contrast, horses from the arid southwest region of North
America had diets predominantly (SO-90%) composed of C, vegetation in the late
Miocene and Pliocene. A provincial grassland distribution pattern such as this
indicates that global atmospheric CO2 fluctuations in the late Miocene are insufficient
for explaining the expansion of C, grasslands. Other factors which control the
distribution of modern C, vegetation such as regional temperature, moisture availability
and latitudinal effects controlled the distribution and abundance of savanna grasses
throughout their phylogeny.

Savanna vegetation did not become an important dietary constituent of Great

Plains horses until the middle Miocene (15.5 Ma to 12.0 Ma), approximately 2 Ma

93

after the evolution of horse hypsodonty. Therefore, C, vegetation cannot be the
primary causal agent of high-crowned teeth in horses. Although not the original cause,
siliceous C, grasses likely contributed to the evolution of hypsodonty among the
Equidae. The incorporation of C3 grasses containing opal phytoliths and the expansion
of abrasive grazing substrates are the most likely causes for the increase in equid tooth
crown heights. The precise role that abrasive C3 grasses and gritty prairie soils played
in the evolution of horse hypsodonty requires further investigation. Undeniably, the
combination of C3 grasses, C, grasses and grit particles worked in concert to produce
the remarkable hypsodont teeth which characterize members of the family Equidae for

the last eighteen million years.

APPENDIX

94

APPENDIX 1

Vertebrate taxonomy to the specific level of the ancient and modern ungulates
analyzed in this study including the original references when known (modified from
Carroll, 1988).

Class Mammalia (Linnaeus, 1758)

Superorder Ungulata (Novacek, 1986)

Order Artiodactyla

Family Cervidae

Cervus canadensis

 

Order Perissodactyla (Owen, 1848)
Family Equidae (Gray, 1821)

Calippus §Q.&
Cormohipparion occidentale
"Dinohippus" leidyms
Eguus caballus

Eguus simplicidens

Hipparion sp.(cf.forcei)
Nannippus lenticularis

Neohipparion eurvstvle

Mesohippus §p_.
Merychippus insignis

'Merychippus" primus
"Merychippus" tertius
"Merychippus" §p_.(cf.isonesus)
Miohippus obliguidens
"Parahippus" s9.

Pliohippus m

 

LIST OF REFERENCES

95

LIST OF REFERENCES

Abelson P. H., 1955. Organic constituents of fossils. Carnegie Institute Washington
Yearbook 54: 107-109.

Ambrose, S. H. and DeNiro, M. J., 1989. Climate and habitat reconstruction using
stable carbon and nitrogen isotope ratios of collagen in prehistoric herbivore
teeth from Kenya. Quaternary Research 31: 407-422.

Armstrong, W. G., Halstead, L. B., Reed, F. B. and Wood, L., 1983. Fossil proteins
in vertebrate calcified tissues. Philosophical Transactions of the Royal Society
of London 301: 301-343.

Axelrod, D. I., 1985. Rise of the grassland biome. Botanical Review 51: 163-201.

Bada, J. L., Peterson, R. O., Schimmelmann, A., and Hedges, R. E. M., 1990. Moose
teeth as monitors of environmental isotopic parameters. Oecologia 82: 102-106

Bocherens, H., Fizet, M. and Mariotti, A., 1990. Evidence for vegetarian diet of cave
bear (Ursus spelaeus) from isotOpic biogeochemistry (”C, 15N) of fossil
vertebrate collagen. C. R. Academie des Sciences 311 (Serie II): 1279-1284.

Bocherens, H., Fizet, M., and Mariotti, A., 1994. Diet, physiology, ecology of fossil
mammals as inferred from stable carbon and nitrogen isotope biogeochemistry:
Implications for Pleistocene bears. Palaeogeography, Palaeoclimatology,
Palaeoecology 107: 213-225.

Boutton, T. W., Tyrell, H. F ., Patterson, B. W., Verga, G. A., and Klein, P. D., 1988.
Carbon kinetics of milk formation in Holstein cows in late lactation. Journal of
Animal Science 66: 2636-2645.

Brown, W. V. and Smith, B. N., 1972. Grass evolution, the Kranz syndrome, 13C/‘zC
ratios and continental drift. Nature 239: 345.

Bryant, J. P., Tahvanainen, J., Sulkinoja, M., Julkunen-Tiitto, R., Reichardt, P., and
Green, T., 1989. Biogeographic evidence for the evolution of chemical
defense by boreal birch and willow against mammalian browsing. The
American Naturalist 134 (1): 20-34.

96

Carroll, R. L., 1988. Vertebrate paleontology and evolution. W. H. Freeman and
Company, New York, NY, 629-645.

Caswell, H., Reed, F., Stephenson, S. N., and Werner, P. A., 1973. Photosynthetic
pathways and selective herbivory: A hypothesis. The American Naturalist 107
(956): 465-480.

Cerling, T. E., 1992. Development of grasslands and savannas in East Africa during
the Neogene. Palaeogeography, Palaeoclimatology, Palaeoecology 97: 241-
247.

Cerling, T. B, Wang, Y. and Quade, J., 1993. Expansion of C, ecosystems as an
indicator of global ecological change in the late Miocene. Nature 361: 344-
345.

Cormie, A. B., and Schwarcz, H. P., 1994. Stable isotopes of nitrogen and carbon of
North American white-tailed deer and implications for paleodietary and other
food web studies. Palaeogeography, Palaeoclimatology, Palaeoecology 107:
227-241.

Corrigan J. J., 1969. D-Amino acids in animals. Science 164: 142-149.

Delwiche, C. C., and Steyn, P. L., 1970. Nitrogen isotope fractionation in soils and
microbial reactions. Environmental Science and Technology 4: 929-935.

Delwiche, C. C., Zinke, P. J., Johnson, C. M. and Virginia, V. A., 1979. Nitrogen
isotope distribution as a presumptive indicator of nitrogen fixation. Botanical
Gazette Supplement 140: 65-69.

DeNiro, M. J., 1985. Postmortem preservation and alteration of in vivo bone collagen
isotope ratios in relation to paleodietary reconstruction. Nature 317: 806-809.

DeNiro, M. J., and Epstein, S., 1977. Mechanism of carbon isotope fractionation
associated with lipid synthesis. Science 197: 261.

DeNiro, M. J., and Epstein, S., 1978a. Carbon isotopic evidence for different feeding
patterns in two hyrax species occupying the same habitat. Science 201: 906-
907.

DeNiro, M. J., and Epstein, S., 1978b. Influence of diet on the distribution of carbon
isotopes in animals. Geochimica et Cosmochimica Acta 42: 495-506.

DeNiro, M. J., and Epstein, S., 1981. Influence of diet on the distribution of nitrogen
isotopes in animals. Geochimica et Cosmochimica Acta 45: 341-351.

97

DeNiro, M. J., and Hastorf, C. A., 1985. Alteration of the 15N/“N and l3C/‘zC ratios of
plant matter during the initial stages of plant diagenesis: Studies utilizing
archaeological specimens from Peru. Geochimica et Cosmochimica Acta 49:
97-115.

DeNiro, M. J., and Weiner, S., 1988. Chemical, enzymatic and spectroscopic
characterization of collagen and other organic fractions in prehistoric bone.
Geochimica et Cosmochimica Acta 52: 2197-2206.

Downton, W. J. S., 1975. The occurrence of C, photosynthesis among plants.
Photosynthetica 9: 96-105.

Elias, M. K., 1942. Tertiary praire grasses and other herbs from the High Plains.
Geological Society of America Special Paper 41: 1-176.

Engel, M. H., and Hare, P. E., 1985. Gas-liquid chromatographic separation of amino
acids and their derivitives. In: Barrett, G. C. (ed), Chemistry and
biochemistry of the amino acids, Wiley & Sons, New York, NY, 462-478.

Galimov, E. M., 1985. The biological fractionation of isotopes. Academic Press, Inc.,
Orlando, FL, 21-23.

Galusha, T., 1975. Stratigraphy of the Box Butte Formation, Nebraska. Bulletin of
the American Museum of Natural History 156 (1): 1-68.

Gray, J. E., 1821. On the natural arrangement of vertebrose animals. London Medical
Repository Review 15: 296-310. [unseen].

Hare, PE, 1980. Organic geochemistry of bone and its relation to the survival of
bone in the natural environment. In: Behrensmeyer, AK. and AP. Hill
(eds), Fossils in the making: Vertebrate taphonomy and paleoecology, Univ.
of Chicago Press, Chicago, IL, 208-219.

Harrigan, P., Zeamen, J. C., and Macko, S. A., 1989. The base of nutritional support
for the gray snapper (Lutjanus griseus): An evaluation based on combined
stomach content and stable isotope analysis. Bulletin of Marine Science 44:
65-77.

Heaton, T. H. E., Vogel, J. G, von la Chevallerie, G. and Collettt, G., 1986. Climatic
influence on the isotopic composition of bone nitrogen. Nature 322: 822-823.

Holman, J. A., 1971. Climatic significance of giant tortoises from the Wood Mountain
Formation (Upper Miocene) of Saskatchewan. Canadian Journal of Earth
Sciences 8: 1148-1151.

98

Holman, J. A., 1973. Reptiles of the Egelhoff fauna (Upper Miocene) of Nebraska.
Contributions of the Museum of Paleontology, University of Michigan 24 (12):
125-134.

Holman, J. A., 1976. Snakes and Stratigraphy. Michigan Academician 8 (4): 387-396.

Holman, J. A., 1979. A review of North American Tertiary snakes. Publication of the
Museum, Michigan State University. Paleontological Series 1 (6): 203-260.

Holman, J. A., 1982. New herpetological species and records from the Norden Bridge
Fauna (Miocene: Late Barstovian) of Nebraska. Transactions of the Nebraska
Academy of Sciences, 10: 31-36.

Hulbert, R. C., 1982. Population dynamics of the three-toed horse Neohipparion from
the late Miocene of Florida. Paleobiology 8: 159-167.

Hulbert, R. C. Jr., 1993. Taxonomic evolution in North American Neogene horses
(subfamily Equinae): The rise and fall of an adaptive radiation. Paleobiology
19 (2): 216-234.

Huxley, T. H., 1870. Anniversary address of the president. Quarterly Journal of the
Geological Society of London 26: 29-64 [unseen].

Iverson, S. J., 1992. Milk secretion in marine mammals in relation to foraging: Can
milk fatty acids predict diet? Symposia of the Zoological Society of London
66 (1): 263-291.

Janis, C. M., 1988. An estimation of tooth volume and hypsodonty indices in ungulate
mammals, and the correlation of these factors with dietary preferences.

Memoires Museum National d'Historie Naturelle, Paris, Serie C 53: 367-387.

Janis, C. M., 1989. A climatic explanation for patterns of evolutionary diversity in
ungulate mammals. Palaeontology 32: 463-481.

Jones, R. L., 1964. Note on occurrence of opal phytoliths in some Cenozoic
sedimentary rocks. Journal of Paleontology 38 (4): 773-775.

Jones, L. H. P., and Handreck, K. A., 1967. Advances in Agronomy 19:107.

Kaufman, P. B., Dayanandan, P., Franklin, C. 1., and Takeoka, Y., 1985. Annual
Botanica 55: 487.

Keegan, W. F., 1989. Reconstructions of life from the skeleton. Alan R. Liss, Inc,
New York, NY, 223-236.

99

Kowalevsky, W., 1874. Monographie der gattung Anthmcotherium cuv.
Paleontographica 22: 131-346 [unseen].

Lee-Thorp, J. A., and Van der Merwe, N. J., 1987. Carbon isotope analysis of fossil
bone apatite. South African Journal of Science 83: 712-715.

Linnaeus, C., 1758. Systema Natume per Regna m'a Naturae, secundum Classes,
Ordines, Genera, Species, cum Characteribus, Differetiis Synonymis, Locis,
Stockholm [unseen].

Lowenstein, J. M., 1988. Immunological methods for determining phylogenetic
relationships. In: Broadhead, T.W. (ed.), Molecular evolution and the fossil
record, 11th annual short course of the Paleontological Society, The
Paleontological Society, Knoxville, TN, 12-19.

Lowenstein, J. M., Sarich, V. M., and Richardson, B. J., 1981. Albumin systematics
of the extinct mammoth and Tasmanian wolf. Nature 291: 409-411.

Lowenstein, J. M., and Ryder O. A., 1985. Immunological systematics of the extinct
quagga (Equidae). Experientia 41: 1192-1193.

MacFadden, B. J., 1984. Systematics and phylogeny of Hippan'on, Neohipparion,
Nannippus, and Cormohipparion (Mammalia, Equidae) from the Miocene and
Pliocene of the New World. Bulletin of the American Museum of Natural
History 179 (1): 4-195.

MacFadden, B. J., and Hulbert, R. C., 1988. Explosive speciation at the base of the
adaptive radiation of Miocene grazing horses. Nature 336: 466-468.

MacFadden, B., Bryant, D. J., and Mueller, PA, 1991. Sr-isotopic, paleomagnetic, and
biostratigraphic calibration of horse evolution: Evidence from the Miocene of
Florida. Geology 19: 242-245.

MacFadden, B. J., 1992. Fossil horses: Systematics, paleobiology, and evolution of
the family Equidae. Cambridge University Press, New York, NY 1-336.

MacFadden, B. J., Wang, Y., Cerling, T. E., and Anaya, F., 1994. South American
fossil mammals and carbon isotopes: A twenty-five million year sequence
from the Bolivian Andes. Palaeogeography, Palaeoclimatology, Palaeoecology
107: 257-268.

Macko, S. A., Estep, M. F., Hare, P. E., and Hoering, T. C., 1987. Isotopic
fractionation of nitrogen and carbon in the synthesis of amino acids by
microorganisms. Isotope Geoscience 65: 79-92.

100

Macko, S. A., and Engel, M. H., 1991. Assessment of indigeneity in fossil organic
matter: amino acids and stable isotopes. Philosophical Transactions of the
Royal Society of London B 333: 367-374.

Marshall, C. R., 1988. DNA-DNA hybridization, phylogenetic reconstruction and the
fossil record. In: Broadhead, T.W. (ed.), Molecular evolution and the fossil
record, 11th annual short course of the Paleontological Society, The
Paleontological Society, Knoxville, TN, 75-88.

Maschinski, J. and Whitham, T. G., 1989. The continum of plant responses to
herbivory: The influence of plant association, nutrient availability and timing.
The American Naturalist 134 (1): 1-19.

Masters, P. M., 1987. Preferential preservation of noncollagenous protein during bone
diagenesis: Implication for chronometric and stable isotope measurements.
Geochimica et Geochimica et Cosmochimica Acta 51: 3209-3214.

Morgan M. E., Kingston, J. D., and Marino, B. D., 1994. Carbon isotopic evidence
for the emergence of C, plants in the Neogene from Pakistan and Kenya.
Nature 367: 162-165

Nambudiri, E. M. V., Tidwell, W. D., Smith, B. N., and Hebbert, N. P., 1978. A C,
plant from the Pliocene. Nature 276: 816.

Nelson, B. K., DeNiro, M. J., Schoeninger, M. J., De Paolo, D. J., and Hare, P. E,
1986. Effects of diagenesis on strontium, carbon, nitrogen, and oxygen
concentration and isotopic composition of bone. Geochimica Cosmochimica
Acta 50: 1941-1949.

Nordt, L. C., Boutton, T. W., Hallmark, C. T., and Waters M. R., 1994. Late
Quaternary vegetation and climate changes in central Texas based on the
isotopic composition of organic carbon. Quaternary Research 41: 109-120.

Novacek, M. J., 1986. The skull of leptictid insectivorans and the higher-level
classification of eutherian mammals. Bulletin of the American Museum of
Natural History 183: 1-112.

Ostrom, P. H., 1990. Geochemical characterization of high molecular weight material
isolated from Late Cretaceous Fossils. PhD Dissertation, Memorial University
of Newfoundland, 1-202.

Ostrom, P. H., Macko, S.A., Engel, M.H., Silfer, IA, and Russell, D., 1990.
Geochemical characterization of high molecular weight material isolated from
Late Cretaceous Fossils. Organic Geochemistry 16: 1139-1144.

101

Ostrom, P. H., and Fry, B., 1993. Sources and cycling of organic matter within
modern and prehistoric food webs. In: Engel and Macko (eds), Organic
Geochemistry: Principles and applications. Plenum Press, New York, NY,
785-798.

Ostrom, P. H., Macko, S. A., Engel, M. H., and Russell, D. A., 1993. Assessment of
trophic structure of Cretaceous communities based on stable nitrogen isotope
analyses. Geology 21: 491-494.

Ostrom, P. H., Zonneveld, J. P., and Robbins, L. L., 1994. Organic geochemistry of
hard parts: Assessment of isotopic variability and indigeneity.
Palaeogeography, Palaeoclimatology, Palaeoecology 107: 201-212.

Paéibo, 8., Russell, G. H., and Wilson, A. C., 1989. Ancient DNA and the polymerase
chain reaction. The Journal Of Biological Chemistry 264 (17): 9709-9712.

Pfretzschner, H. U., 1993. Enamel microstructure in the phylogeny of the Equidae.
Journal of Vertebrate Paleontology 13 (3) 342-349.

Qian, Y., Engel M. H., Macko, S. A., Carpenter, S., and Deming, J. W., 1993.
Kinetics of peptide hydrolysis and amino acid decomposition at high
temperature. Geochimica et Cosmochimica Acta 57: 3281-3293.

Quade, J., Solounias, N., and Cerling, T. E., 1994. Stable isotope evidence from
paleosol carbonates and fossil teeth in Greece for forest or woodlands over the
past 11 Ma. Palaeogeography, Palaeoclimatology, Palaeoecology 108: 41-53.

Retallack, G. J., 1983a. Late Eocene and Oligocene paleosols from Badlands National
Park, South Dakota. Geological Society of America Special Paper 193: 1-61.

Retallack, G. J., 1983b. A paleopedological approach to the interpretation of terrestrial
sedimentary rocks: The mid-Tertiary fossil soils of Badlands National Park,
South Dakota. Geological Society of America Bulletin 94: 823-840.

Robbins, LL, and Brew, K., 1990. Proteins from the organic matrix of core-top
fossil planktonic foraminifera. Geochimica et Cosmochimica Acta 54:
2285-2292.

Rogers, K. L., 1976. Herpetofauna of the Beck Ranch Local Fauna (Upper Pliocene:
Blancan) of Texas. Publication of the Museum, Michigan State University.
Paleontological Series 1 (5): 167-200.

102

Schoeninger M. J. and DeNiro, M. J., 1984. Nitrogen and carbon isotopic composition
of bone collagen from marine and terrestrial animals. Geochimica et
Cosmochimica Acta 48: 625-639.

Serban, A., Engel, M. H., Macko, S. A., 1987. The distribution, stereochemistry and
stable isotopic constituents of fossil and modern mollusk shells. In: Matavelli,
L. and Norvelli L. (eds), Advances in Organic Geochemistry, Pergamon Press,
Oxford. 13: 1123-1129.

Silfer, J.A., Engel, M.H., Macko, S. A., and Jumeau, El, 1991. Stable carbon isotope
analysis of amino acid enatiomers by conventional ratio mass spectrometry and
combined gas chromatography/isotope ratio mass spectrometry. Analytical
Chemistry 63: 370-374.

Silfer J. A., Engel M. H., and Macko S. A., 1992. Kinetic fractionation of stable
carbon and nitrogen isotopes during peptide bond hydolysis: Experimental
evidence and geochemical implications. Chemical Geology 101: 211-221.

Simpson, G. G., 1951. Horses: The story of the horse family in the modern world and
through sixty million years history. Oxford University Press, New York, NY,
1-247.

Skinner, M. F., Skinner, S. M., and Gooris, R. J., 1977. Stratigraphy and
biostratigraphy of Late Cenozoic deposits in Central Sioux County, Western
Nebraska. Bulletin of the American Museum of Natural History 158 (5): 1-371.

Skinner, M. F., and Johnson, F. W., 1984. Tertiary stratigraphy and the Frick
collection of fossil vertebrates from North-Central Nebraska. Bulletin of the
American Museum of Natural History 178 (3): 1-368.

Smith, B. N., and Epstein, S., 1971. Two categories of l3C/‘ZC ratios for higher plants.
Plant Physiology 47: 380-384.

Stirton, R. A., 1947. Observations on evolutionary rates in hypsodonty. Evolution 1:
32-41.

Teeri, J. A. and Stowe, L. G., 1976. Climatic patterns and the distribution of C,
grasses in North America. Oecologia 23: 1-12.

Teeri, J. A. and Schoeller, D. A., 1978. 5‘3C values of an herbivore and the relation
of C3 to C, plant carbon in its diet. Oecologia 39: 197-200.

103

Thackeray, J. F., Lee-Thorp, J. A., 1992. Isotopic analysis of equid teeth from
Wonderwerk Cave, Northern Cape Province, South Africa. Palaeogeography,
Palaeoclimatology, Palaeoecology 99: 144-150.

Thomasson, J. R., 1978. Tertiary fossil plants found in Nebraska. National
Geographic Society Research Reports, 1978 Projects.

Thomasson, J. R., Nelson, M. E., and Zakrzewski, R. J., 1986. A fossil grass
(Grarnineae: Chloridoideae) from the Miocene with Kranz anatomy. Science
233: 876-878.

Tieszen, L. L., 1978. Carbon isotope fractionation in biological materials. Nature
276: 97-98.

Tieszen, L. L., Hein, D., Qvortrup S., Troughton, J., and Imbarnba, S., 1979a. Use of
5‘3C values to determine vegetation selectivity in East African herbivores.
Oecologia 37: 351-359.

Tieszen, L. L., Senyimba, M. M., Imbamba, S. K., and Troughton, J. H., 1979b. The
distribution of C3 and C, grasses and carbon isotope discrimination along an
altitudinal and moisture gradient in Kenya. Oecologia 37: 337-350.

Tieszen, L. L., Boutton, T.W., Tesdahl, K.G., Skadem BA, 1983. Fractionation and
turnover of stable carbon isotopes in animal tissues: Implications for 13C
analysis of diet. Oecologia 57: 32-37.

Tyrrell, H. F., Pelletier, G., Chevalier, R., Hilliare-Marcel, C., and Gagnon, M., 1984.
Use of carbon-13 as a tracer in metabolic studies. Canadian Journal of Animal
Science 64 (suppl): 127.

Van der Merwe, N. J., and Medina, E., 1989. Photosynthesis and 13C/12C ratios in
Amazonian rain forests. Geochimica Cosmochimica Acta 53: 1091-1094.

Vogel, J. C., 1978. Isotopic assessment of the dietary habits of ungulates. South
African Journal of Science 74: 298-301.

Vogel, J. C., Fuls, A., and Danin, A., 1986. Geographical and environmental
distribution of C3 and C, grasses in the Sinai, Negev and Judean deserts.
Oecologia 70: 258-265.

Vogel, J. C., Talma, A. S., Hall-Martin, A. L., and Viljoen, P.J., 1990a. Carbon and
nitrogen isotopes in elephants. South African Journal of Science 86: 147-150.

104

Vogel, J. C., B. Eglinton, and Auret, J. M., 1990b. Isotope fingerprints in elephant
bone and ivory. Nature 346: 747-749.

Voorhies, M. R. and Thomasson, JR, 1979. Fossil grass anthoecia within Miocene
rhinoceros skeletons: Direct evidence of diet in an extinct species. Science
206: 331-333.

Voorhies, M. R., 1990. Vertebrate biostratigraphy of the Ogalla Group in Nebraska.
In: Gustavson, T.C. (ed.), Geological framework and regional hydrogeology:
Upper Cenozoic Blackwater Draw and Ogalla Formations, Great Plains, Austin,
TX, 115-151.

Walker, A., Hendrick, N. H., and Perez, L., 1978. Microwear of mammalian teeth as
an indicator of diet. Science 201: 908-910.

Wang, Y., Cerling, T. E. and MacFadden, B. J., 1994. Fossil horses and carbon
isotopes: New evidence for Cenozoic dietary, habitat, and ecosystem changes
in North America. Palaeogeography, Palaeoclimatology, Palaeoecology 107:
269-279.

Wessells N. K., and Hopson, J. L., 1988. Biology. Random House, Inc., New York,
NY, 188-190.

Woodbume, M. O., 1987. Cenozoic mammals of North America. University of
California Press, London, UK, 118-210.

Woodruff, F ., Savin, S. M., and Douglas, R. G., 1981. Miocene stable isotope record:
A detailed deep Pacific ocean study and its paleoclimatic interpretation.
Science 212: 665-668.

Zubakov, V. A., and Borzenkova, I. I., 1990. Global Paleoclimate of the Late
Cenozoic, Elsevier, Amsterdam.