<%BANNER%>

Paleoecology of Forest Environments through Time

Permanent Link: http://ufdc.ufl.edu/UFE0024701/00001

Material Information

Title: Paleoecology of Forest Environments through Time Evidence from Stable Isotopes of Mammalian Herbivores in the New World
Physical Description: 1 online resource (123 p.)
Language: english
Creator: Grawe, Larisa
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: acapulco, america, appalachians, bairdii, canopy, carbon, cenozoic, climate, craniodental, d13c, d18o, density, earth, education, elements, elevation, enamel, environment, eocene, evolution, forest, fossil, geochemistry, holocene, inquiry, isotope, mammals, mexico, microwear, miocene, morphology, neogene, north, oligocene, outreach, oxygen, palaeoecology, palaeontology, paleoecology, paleontology, pinchaque, pleistocene, pliocene, precipitation, rare, ree, refugia, serial, stable, tapiridae, tapiroidea, tapirs, tapirus, teeth, temperature, tennessee, terrestrial, terrestris, time, ungulate, vertebrates, water
Zoology -- Dissertations, Academic -- UF
Genre: Zoology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Understanding the paleoecology of forest environments is critical to comprehending the context of mammalian evolution, particularly of forest-dwelling taxa. Significant work has helped clarify the timing and consequences of the evolution of grassland biomes in North America. In contrast, the paleoecology of forest environments during the late Cenozoic in southeastern North America requires further study. Therefore, my dissertation aims to understand the ecology of forest environments using stable isotopes of mammalian herbivores. With a focus on forest-dwelling tapirs and their ancestors, in the first part I document carbon and oxygen isotope variation in extant tapirs, at the individual, population, and species level. Extant tapirs are also compared across latitudes and in areas with varying temperature and/or precipitation regimes to understand how tapirs track carbon and oxygen isotopes. These results demonstrate the conservative dietary niches of extant tapirs and the relationship between decreased drinking behavior and increased local precipitation. These data are informative to conservationists managing tapir habitats and constrain paleoecological interpretations of the southeastern fossil sites examined. In the second part, the tapir-dominated Gray Fossil Site in eastern Tennessee provides rare insight into the ecological dynamics occurring during the late Neogene, a time of dramatic global change. Isotopic evidence clarifies ecological niches, relative seasonality, and suggests that the Appalachians may have served as a forest refugium between approximately 4.5 and 7 million years ago. In the third part, extinct tapirs and tapiroids are determined to be indicators of forest environments over the past 55 million years, based on craniodental morphology and carbon isotopes, and used to understand how forest distributions in North America have changed over time. Specifically, declines in tapiroid latitudinal ranges during the Oligocene are consistent with climatic cooling of ~8degreeC while their recent retreat (within the last ~10,000 years) to only southern distributions below 21masculine ordinalN indicates human interference. Collectively, my dissertation research helps elucidate how forests and their mammalian inhabitants have changed over time during the Cenozoic in the New World.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Larisa Grawe.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Macfadden, Bruce J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024701:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024701/00001

Material Information

Title: Paleoecology of Forest Environments through Time Evidence from Stable Isotopes of Mammalian Herbivores in the New World
Physical Description: 1 online resource (123 p.)
Language: english
Creator: Grawe, Larisa
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: acapulco, america, appalachians, bairdii, canopy, carbon, cenozoic, climate, craniodental, d13c, d18o, density, earth, education, elements, elevation, enamel, environment, eocene, evolution, forest, fossil, geochemistry, holocene, inquiry, isotope, mammals, mexico, microwear, miocene, morphology, neogene, north, oligocene, outreach, oxygen, palaeoecology, palaeontology, paleoecology, paleontology, pinchaque, pleistocene, pliocene, precipitation, rare, ree, refugia, serial, stable, tapiridae, tapiroidea, tapirs, tapirus, teeth, temperature, tennessee, terrestrial, terrestris, time, ungulate, vertebrates, water
Zoology -- Dissertations, Academic -- UF
Genre: Zoology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Understanding the paleoecology of forest environments is critical to comprehending the context of mammalian evolution, particularly of forest-dwelling taxa. Significant work has helped clarify the timing and consequences of the evolution of grassland biomes in North America. In contrast, the paleoecology of forest environments during the late Cenozoic in southeastern North America requires further study. Therefore, my dissertation aims to understand the ecology of forest environments using stable isotopes of mammalian herbivores. With a focus on forest-dwelling tapirs and their ancestors, in the first part I document carbon and oxygen isotope variation in extant tapirs, at the individual, population, and species level. Extant tapirs are also compared across latitudes and in areas with varying temperature and/or precipitation regimes to understand how tapirs track carbon and oxygen isotopes. These results demonstrate the conservative dietary niches of extant tapirs and the relationship between decreased drinking behavior and increased local precipitation. These data are informative to conservationists managing tapir habitats and constrain paleoecological interpretations of the southeastern fossil sites examined. In the second part, the tapir-dominated Gray Fossil Site in eastern Tennessee provides rare insight into the ecological dynamics occurring during the late Neogene, a time of dramatic global change. Isotopic evidence clarifies ecological niches, relative seasonality, and suggests that the Appalachians may have served as a forest refugium between approximately 4.5 and 7 million years ago. In the third part, extinct tapirs and tapiroids are determined to be indicators of forest environments over the past 55 million years, based on craniodental morphology and carbon isotopes, and used to understand how forest distributions in North America have changed over time. Specifically, declines in tapiroid latitudinal ranges during the Oligocene are consistent with climatic cooling of ~8degreeC while their recent retreat (within the last ~10,000 years) to only southern distributions below 21masculine ordinalN indicates human interference. Collectively, my dissertation research helps elucidate how forests and their mammalian inhabitants have changed over time during the Cenozoic in the New World.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Larisa Grawe.
Thesis: Thesis (Ph.D.)--University of Florida, 2009.
Local: Adviser: Macfadden, Bruce J.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2011-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024701:00001


This item has the following downloads:


Full Text

PAGE 1

PALEOECOLOGY OF FOREST ENVIRONMEN TS THROUGH TIME: EVIDENCE FROM STABLE ISOTOPES OF MAMMALIAN HE RBIVORES IN THE NEW WORLD By LARISA R. G. DESANTIS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009

PAGE 2

2 2009 Larisa R. G. DeSantis

PAGE 3

3 To my family, especially my husband and daught er, for their unending support and constant source of inspiration that have enabled me to pursue my dreams

PAGE 4

4 ACKNOWLEDGMENTS I owe an enormous thanks to all who have helped support both my dissertation research and my research career. I am grateful to the continuous support of my advisor Bruce MacFadden, who has been a source of intellect ual, research, and prof essional support. From helping me establish research connections in Tennessee to en couraging my involvement in exhibit and educational outreach, he has been supportive of my res earch and educational endeavors. My Ph.D. Committ ee, including Bruce MacFadden, Karen Bjorndal, Jonathan Bloch, Robert Holt, and Ted Schuur, have helped me to integrate the fields of ecology and paleontology and provided scientific and career advice. Additionally, Jonathan Bloch improved my development as both an educator and research er while teaching verteb rate paleontology and navigating through job searches. Richard Hulb ert has helped with specimen access, tapir discussions, editorial advice, and allo wed me prime field opportunities. My dissertation research would not have been possible without access to museum collections including: the American Museum of Natural History (AMNH), East Tennessee State University and General Shale Brick Natural History Museum (ETMNH), Florida Museum of Natural History (FLMNH), and the Yale Peabody Museum (YPM). Specifically, I thank the following persons for access to, and help with, th e relevant collections under their care: John Flynn, Vertebrate Paleontology, AMNH; Eileen Westwig, Mammals, AMNH; Steven Wallace, Blaine Schubert, April Nye, and Brett W oodward, ETMNH; Richard Hulbert, Vertebrate Paleontology, FLMNH; Kristof Zyskowski and Gr egory Watkins-Colwell, Vertebrate Zoology, YPM; and Walter Joyce, Vertebrate Paleontology, YPM. The paleoecologi cal research of the Gray Fossil Site (Chapter 2) was a result of th e collaborative assistance of Steven Wallace, and the support of East Tennessee State University affiliates including: Blaine Schubert, Jeff Supplee, April Nye, Brian Compton, Shawn Ha ugrad, Brett Woodward, and Jeanne Zavada.

PAGE 5

5 Karen Carr provided the images included in Figure 3-2. Additionally, Chap ter 4 is a product of a collaborative effort with Bruce MacFadden. All is otopic analyses were run in the Department of Geological Science at the Universi ty of Florida with the assistan ce of Jason Curtis. I am very thankful for all of his help, tr oubleshooting, and use of the isotope lab for my dissertation work and other research projects. I am also thankful to John Krigbaum for isotope discussions and the use of his lab, and George Kamenov for both training and assisting me with Rare Earth Element analyses. Elizabeth Kowalski and George Case lla provided statistical advice for the analyses conducted in Chapter 2. The chapters of this dissertation have benefited from comments by anonymous reviewers in addition to Richard Hulbert, Robert Fera nec, Pennilyn Higgins, and my Ph.D. Committee. Both the Department of Biology and Florida Museum of Natural History have provided logistical and financial suppor t, including the use of mu seum specimens, geochemical preparation facilities, office spa ce, and financial support (i.e. c onsumable geochemical supplies, travel grants, end of the year funds, and four years of salary support w ith the University of Florida Alumni Fellowship). One year of sala ry support was provided by the National Science Foundation through the University of Florida Science Partners in Inquiry-based Collaborative Education program. Additional financial s upport was provided by the Florida Museum of Natural History Lucy Dickinson Scholarship. I am forever grateful to the late J.C. Dickinson for establishing and maintaining the Lucy Dickins on Scholarship, which provided me the financial freedom to conduct preliminary investigations lead ing to funded research projects. The National Science Foundation South East Alliance for Gra duate Education and the Professoriate provided financial support and valuable professional de velopment opportunities. Additional financial support was provided by the Tennessee Department of Transportation, East Tennessee State

PAGE 6

6 University, the Don Sunquist Center for Excelle nce in Paleontology, a R. Jerry Britt, Jr. Paleobiology Award, a Geological So ciety of America Grant-in-Aid of Research, a Gainesville Women's Club Scholarship, a Southwest Florida Fossil Club Scholarship, and a Sigma Xi Grantin-Aid of Research. Numerous travel awards fro m the University of Florida College of Liberal Arts and Sciences, Department of Biology, Grad uate Student Council, Southeastern Geological Society of America, and Society of Vertebrate Paleontology have enabled me to present this research at domestic and intern ational professional meetings. I am thankful for the support of the Department of Biology a nd Florida Museum of Natural History faculty, students, and staff. Brian Silliman, Jack Putz, Doug Levey, Marta Wayne, Charlie Baer, Jamie Gillooly, David Reed, Scott Robinson, Collette St. Mary, Ben Bolker, Jamie Creola, Darcie MacMahon, and Doug Jones have provided research, educa tional outreach, and/or career support. Julie Allen, Smriti Bhotika, Jason Bourque, Lisa Crummett, Dana Ehret, Donovan German, Alex Hastings, Jill Jankowsk i, Joann Labs-Hochstein, Gustavo Londoo, Ari Martinez, Julie Mathis, Paula Mejia, Sea McKeon, Bret Pasch, Kim Reich, Christine Stracey, Alfred Thomson, Judit Ungvari, and Jada-Simone White have offered diverse assistance, from job talk feedback to moral support. Karen Patterson, Vitrell Sher if, Amy Dechow, Pete Ryschkewitsch, Art Poyer, Pam Dennis, and S huronna Wilson helped with logistics. Additionally, Stephanie W ear helped me broaden my research perspective through working with the Nature Conservancy and was constantly a source of motivation. While those mentioned above played the greatest role in my dissertation research over the last few years, others including: Blaine Schubert, Robert Feran ec, Stephen Wroe, Judith Field, Gavin Prideaux, Mark Goodwin, Bill Clemens, Paul Sereno, Kevin Padian, Greg Wilson, Caroline Stromberg, Jane Mason, Laura Fawcett, Terri Stern, Kevin ORangers, Stephen Kellert,

PAGE 7

7 Oswald Schmitz, Ann Camp, Frank Beall, and Jeff Romm, have provide d research and career support over the last 10+ years. Lastly, but definitely not least, I am extremely grateful to my family. I am thankful to my parents Virginia and Clive Grawe for not trying to convince me to purse a more practical career. From enrolling me in my first paleontology class at the Los Angeles Museum of Natural History as a child to helping with the childcare of my daughter during the writi ng of this dissertation, they have encouraged my scientific pursuits throughout my life. My in-laws including Tom, Josephine, and Wes DeSantis have been a source of support and encouragement, and didnt get mad when I took their son to Florida. I owe th e greatest thanks and gratitude to my husband, Derek DeSantis and daughter, Sydney Jo DeSantis Derek moved to Gainesville sight unseen, has shaken thousands of geochemical samples at all hours of the night, helped record morphological measurements during museum visits crossed alligator infested rivers, spent numerous weekends digging up fossils, listened to the tenth through the twentieth version of grant proposals, helped count t housands of beans for educational activities, and has been a constant source of support regardless of what it entailed. Most importantly, he has stood by me during my successes and failures and always made me feel as if I could accomplish even the most challenging tasks. Derek and Sydney are my sources of inspiration and none of this would have been possible without their unending support.

PAGE 8

8 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ........10 LIST OF FIGURES.......................................................................................................................11 ABSTRACT...................................................................................................................................13 CHAPTER 1 INTRODUCTION..................................................................................................................15 2 STABL E ISOTOPE ECOLOGY OF EXTANT TAPIRS ( ) FROM THE AMERICAS................................................. 19 Introduction................................................................................................................... ..........19 Materials and Methods...........................................................................................................22 Results.....................................................................................................................................25 Individual and Popula tion Level Variation ( )...................................................25 Species Level Variation ( ) ....................................26 Climatic and Geographic Parameters ( ) ................26 Discussion...............................................................................................................................28 Implications of Ontogenetic a nd Population Level Variation ......................................... 28 Influence of Climatic and Geographic Variables............................................................ 30 Conclusions.............................................................................................................................33 3 NEOGENE FORESTS FROM THE APPA L ACHIANS OF TENNE SSEE, USA: GEOCHEMICAL EVIDENCE FROM FOSSIL MAMMAL TEETH.................................. 45 Introduction................................................................................................................... ..........45 Background.............................................................................................................................47 Stable Isotope Analysis: A Theoretical Foundation........................................................ 47 Rare Earth Element Analysis: U nderstanding Taphonomic History ............................... 49 Gray Fossil Site, Tennessee, USA................................................................................... 50 Materials and Methods ...........................................................................................................51 Morphological Measurements......................................................................................... 51 Stable Isotope Analysis................................................................................................... 51 Rare Earth Elemental Analysis........................................................................................ 52 Results and Discussion......................................................................................................... ..53 Bulk Carbon Isotope Analysis......................................................................................... 53 Bulk Oxygen Isotope Analysis........................................................................................55 Seasonal Reconstructions: Se rial Sample Analysis.........................................................56 Evidence of a Forest Refuge............................................................................................ 58 Conclusions.............................................................................................................................59

PAGE 9

9 4 IDENTIFYING FOREST ENVIRONMENTS IN DEEP TIME USING FOSSIL TAPIRS: EVIDENC E FROM EVOL UTIONARY MORPHOLOGY AND STABLE ISOTOPES....................................................................................................................... .......73 Introduction................................................................................................................... ..........73 Materials and Methods...........................................................................................................76 Morphology.....................................................................................................................76 Stable Isotopes................................................................................................................ .77 Inferred Forest Distributions...........................................................................................78 Results and Discussion......................................................................................................... ..78 Morphology.....................................................................................................................78 Stable Carbon Isotopes.................................................................................................... 81 Distributions....................................................................................................................83 Conclusions.............................................................................................................................85 5 CONCLUSION..................................................................................................................... ..95 APPENDIX A IMPORTANCE OF COMMUNICATING THE BROADER IMPACTS OF SCIENTIFIC RESEARCH TO THE PUBLIC..................................................................... 100 B STRAIGHT FROM THE MOUTHS OF HORSES AND TAPIRS: USING FOSSIL TEETH TO CLARIFY HOW ANCIENT ENVIRONMENTS HAVE CHANGED OVER TIME...................................................................................................................... ...102 LIST OF REFERENCES.............................................................................................................112 BIOGRAPHICAL SKETCH....................................................................................................... 123

PAGE 10

10 LIST OF TABLES Table page 2-1 13C and 18O values of extant tapirs ( and ) noting climate stations and asso ciated climatic and geographic variables...................................................................................................................... .......35 2-2 13C and 18O values from a population of extant tapirs ( ) in Acapulco, Mexico.............................................................................................................. 37 2-3 Stable carbon and oxygen isotope differences between various early and late erupting tooth positions, from a populat ion of extant tapirs ( ) in Acapulco, Mexico...............................................................................................................................38 2-4 Pearson correlation coefficients for stab le carbon and oxygen isotope enam el values and climatic and geographic variables............................................................................... 39 3-1 Biota from the Gray Fossil Site, Tennessee....................................................................... 62 3-2 Bulk carbon and oxygen isotopes of ma mmalian ungulate enamel, Gray Fossil Site, Tennessee...................................................................................................................... .....63 3-3 Stable carbon and oxygen values for the taxa of the Gray Fossil Site, Tennessee............ 64 3-4 Serial carbon and oxygen isotope variati on, per individual at th e Gray Fossil Site, Tennessee...................................................................................................................... .....65 3-5 Carbon and oxygen isotope values for seri al samples of ungulate enamel, Gray Fossil Site, Tennessee. ..................................................................................................................66 4-1 Com parison of craniodental featur es between the Oligocene tapiroid modern the Oligocene equid and m odern ...........................87 4-2 Summary of Paleobiology Da tabase (2006) results of tapi roid localities in North and Central America. Additionally, Holocene co llections include present loc alities taken from the IUCN Tapir Status Survey and Conservation Action Plan................................. 88

PAGE 11

11 LIST OF FIGURES Figure page 2-1 A Google Earth map showing the location of and specimens from southern Mexi co, Central America, and South America........................................................................................................................ ......402-2 Box plots of stable carbon and oxygen isot ope values from a population of extant tapirs ( ) in Acapulco, Mexico.................................................................... 412-3 Stable carbon and oxygen isotope values of extant tapirs, and ........................................................................................422-4 Relationships between precipitation frequency and oxygen isotope values in and .............................................................................................432-5 Relationship between carbon and oxyge n isotope values from individual and specimens........................................................................... 443-1 Location of the Gray Fossil Site, Tennessee, USA............................................................683-2 Stable carbon and oxygen isotope values from the ungulate taxa at the Gray Fossil Site.....................................................................................................................................693-3 Serial samples of the ungulate ta xa from the Gray Fossil Site.......................................... 703-4 Average monthly temperate (C), m onthly oxygen isotope values, and monthly precipitation (rainfall, snowfa ll) in Johnson City/Bristol Tri-City Area, Tennessee......... 713-5 Normalized REE(PAAS) concentrations of the gomphothere tusk, and teeth................................................................................................................. 724-1 Temporal distributions of the tapi roid genera analyzed in this study................................ 894-2 Specimens of the extinct tapirid extinct equid extant tapir and extant horse .........................................................................................904-3 Hypsodonty index of two contrasting clades of perissodactyls, i.e. Equidae and Tapiroidea, throughout the past 55 million years.............................................................. 914-4 North American fossil tapiroid and extant Central American and South American tapir m1 lengths through time as a proxy for body size evolution..................................... 924-5 Carbon isotope data for tooth enamel of fossil tapirs and their associated faunas for the past 10 million year s in North America....................................................................... 934-6 Maps of tapiroid localities on contin ental North America from the Eocene through the present.................................................................................................................... ......94

PAGE 12

12 B-1 A student actively measuring fossil ho rse teeth to test her null hypothesis..................... 108B-2 Activity sheet one allows students to demonstrate how ancient diets can be determined using tooth morphology................................................................................109B-3 Data sheet accompanying part one of the activity, includes actual morphological data taken from DeSantis and MacFadden (2007).................................................................. 110B-4 Activity sheet two allows students to demonstrate how ancient diets can be determined using dental microwear................................................................................. 111

PAGE 13

13 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PALEOECOLOGY OF FOREST ENVIRONMEN TS THROUGH TIME: EVIDENCE FROM STABLE ISOTOPES OF MAMMALIAN HE RBIVORES IN THE NEW WORLD By Larisa R. G. DeSantis August 2009 Chair: Bruce J. MacFadden Major: Zoology Understanding the paleoecology of forest envi ronments is critical to comprehending the context of mammalian evolution, particularly of forest-dwelling taxa. Significant work has helped clarify the timing and consequences of the evolution of grassland biomes in North America. In contrast, the pa leoecology of forest environmen ts during the late Cenozoic in southeastern North America requires further st udy. Therefore, my dissertation aims to understand the ecology of forest environments usi ng stable isotopes of mammalian herbivores. With a focus on forest-dwelling tapirs and th eir ancestors, in the first part I document carbon and oxygen isotope variation in extant tapi rs, at the individual, population, and species level. Extant tapirs are also compared across latitudes and in areas wi th varying temperature and/or precipitation regimes to understand how ta pirs track carbon and oxygen isotopes. These results demonstrate the conservati ve dietary niches of extant ta pirs and the relationship between decreased drinking behavior and increased local precipitation. These data are informative to conservationists managing tapir habitats and c onstrain paleoecological interpretations of the southeastern fossil sites examine d. In the second part, the tapi r-dominated Gray Fossil Site in eastern Tennessee provides rare insight into th e ecological dynamics o ccurring during the late Neogene, a time of dramatic global change. Isotopic evidence clarifies ecological niches,

PAGE 14

14 relative seasonality, and suggests th at the Appalachians may have served as a forest refugium between approximately 4.5 and 7 million years ago. In the third part, extinct tapirs and tapiroids are determined to be indicators of forest envi ronments over the past 55 million years, based on craniodental morphology and carbon isotopes, and used to understa nd how forest distributions in North America have changed over time. Specifica lly, declines in tapiro id latitudinal ranges during the Oligocene are consistent with climatic cooling of ~8 C while their recent retreat (within the last ~10,000 years) to only southe rn distributions belo w 21N indicates human interference. Collectively, my dissertation research helps el ucidate how forests and their mammalian inhabitants have changed over time during the Cenozoic in the New World.

PAGE 15

15 CHAPTER 1 INTRODUCTION Current global climate change has been doc umented to alter the com position and dynamics of mammalian communities and their environments (Walther et al., 2002; Parmesan and Yohe, 2003; Post and Forchhammer, 2004). Documenting past climate and environmental change is critical to understanding the context of mammalian evolution. Significant work has focused on clarifying the evolution of grassland ecosystems and their mammalian inhabitants (Wang et al., 1994, Cerling et al., 1997; Feranec and MacFad den, 2000; Koch et al., 2004; Retallack, 2001; Stromberg, 2005), however the paleoecology of fore st ecosystems are less understood. Forest environments currently receive a significant am ount of conservation attention and contain the majority of designated hotspots due to high species richness, threat ened/endangered species listings, and/or high endemism (Myers et al ., 2000; Orme et al., 2005). Their presumed evolutionary importance as incubators of biodivers ity is due in part to their potential role as refugia for taxa during periods of climate change (Haffer, 1969; Haffer and Prance, 2001). Thus, it is critical to understand the ancient ecology of forest environments and how their distributions have changed over time. Mammalian taxa hold clues to their past envi ronments in their teeth. Traditionally, the morphological characters including re lative size and shape of teeth ar e used to infer the diets of extinct mammals based on modern mammalian analogues (Solounias and Moelleken, 1993; MacFadden and Shockey, 1997; Mendoza et al., 2002). Ever since the classic work of Kowalevsky (1873), teeth that are taller then they are wide (i.e. hypsodont) are thought to indicate the consumption of grass (Janis, 1988; MacFadden, 1992). This is because grass contains phytoliths that wear down tooth enamel over time (Baker et al., 1959), thus high crowned teeth can withstand prol onged wear and enable a grazing diet. Similarly, low crowned

PAGE 16

16 teeth and other craniodental features can i ndicate a browsing diet (S olounias and Moelleken, 1993; MacFadden and Shockey, 1997; Mendoza et al., 2002). While tooth morphology can provide insight into the diets of extinct taxa, stable isotope techniques can improve dietary interpretations and identify the presence of relati vely dense forest environments (Cerling et al., 1997, 2004; van der Merwe and Medina, 1989, 1991). Stable carbon isotopes, such as 13C and 12C, naturally occur in pl ants and vary depending on how a plant photosynthesizes Taking into account the 13C enrichment from food to tooth enamel (~14) as well as the decline in 13C values (~1.5) of atmospheric CO2 due to fossil fuel burning over the past two cen turies, tooth enamel values of less than -8 indicate a diet consisting of primarily C3 vegetation whereas 13C values of greater than -2 indicate a diet of predominantly C4 vegetation (Cerling et al., 1997; Cerl ing and Harris, 1999). Additionally, stable carbon isotope data of extin ct taxa can be used to identify ancient forests due to greater 13C discrimination occurring in dense closed canopies as compared to more open C3 environments (van der Merwe and Medi na, 1989; Cerling et al., 2004). Because 13C values increase with decreasing canopy density and/or increasing distance from dense forest edges (van der Merwe and Medina, 1989; Kapos et al ., 1993; West et al., 2001), more negative 13C enamel values of mammalian herbivores reflect the cons umption of browse in forests with denser canopies (van der Merwe and Medina, 1989, 1991; Cerling et al., 2004). Therefore, 13C values of mammalian tooth enamel can indicate meaningf ul differences in habitat type within C3 floral environments and therefore assist in determin ing relative canopy density in ancient forests. Through the synergistic combination of ecologi cal ground-truthing and paleoecological investigations of late Cenozoic fossil sites, my dissertation consists of a series of research

PAGE 17

17 projects that collectively help to reconstruct anci ent forest environments. Specifically, I ask the following key questions: How do stable carbon and oxygen isotope analyses of extant tapirs help constrain paleoecological hypotheses? (Chapter 2) What was the environment and climate like in the Appalachians during the Neogene? (Chapter 3) How are resources partitioned between forest dwelling taxa? (Chapter 3) How have forest distributions changed ove r time in the New World? (Chapter 4) Using stable isotopes and vert ebrate morphology, the answers to these questions provide insight regarding the paleoecology of forest e nvironments and their mammalian inhabitants. This work improves our knowledge of fossil lo calities and provides th e necessary ecological background to further interpret ancient forest environments. Tapirs are of particular interest because th ey are potential model organisms for identifying forest environments as they are highly conservative in diet, habitat preference, and migratory behavior (Bodmer, 1990a; Salas, 1996; Henry et al., 2000; Downer 2001; Tobler, 2002; Foerster and Vaughan, 2002). With a focus on forest-dwelling ta pirs and their ancestor s, it is necessary to first understand how extant tapirs track carbon an d oxygen isotope variation, and the extent of variation at the individual, population, and species level. The second chapter clarifies stable carbon and oxygen isotope variation and provides insi ght into extant tapi r ecology. Additionally, the work discussed in Chapter 2 helps constrain paleoecological interpretations of the fossil sites discussed. The Gray Fossil Site, the most fossilif erous tapir locality in the world, provides rare insight into Neogene paleoecology in the Appa lachians region. The Gray Fossil Site was thought to have served as a forest refuge (Wallace and Wang, 2004) during a time of dramatic environmental change, incl uding the expansion of C4 grasslands globally (Cerling et al., 1997). In Chapter 3, the paleoecology and paleoclimatol ogy of the Gray Fossil Site is inferred based on

PAGE 18

18 the geochemistry of fossil tooth enamel (stable isotopes and Rare Earth Elements). Furthermore, I test the hypothesis that the Gray Fossil Site represents a forested Appa lachian refuge during the Neogene. Collectively, these chapters provide the ground work for testing the hypothesis that fossil tapirs are indicators of forest environments In order to test th is idea, I evaluated how craniodental morphology and carbon isotopes of extin ct tapirs changed over time, in Chapter 4. Using tapirs as indicator s of forests, I then document how fo rest distributions have changed in the New World throughout the last 55 million years. This body of work w ill significantly add to the field of paleoecology by improving interpre tations of late Cenozoic fossil sites through the integration of ecological and paleoecological research. The aim of this dissertation is to advance the field of paleoecol ogy and communicate some of the scientific research discussed to the public. Communi cating the broader impacts of scientific research to society is a necessary responsibility of scie ntific researchers. In Appendix A, I discuss the importance of communicating science to the pub lic. In Appendix B, I aim to improve a students understanding of ancient ecosystems by developing an educational module that allows middle school level stud ents to use actual data to test scientific hypotheses. The goal of the appended unit is to actively enga ge students in understanding how scientific interpretations and subsequent i llustrations are developed. Through inquiry-based activities that allow students to analyze actual scientific data pr oduced in this disserta tion, students develop and test hypotheses, later communi cating the products of their scientific investigation. Through the integration of modern ecology, vertebrate paleontology, and scientific outreach, I aim to improve understandings of anci ent forest environments and engage the public in the joy of scientific discovery.

PAGE 19

19 CHAPTER 2 STABLE ISOTOPE ECOLOGY OF EXTANT TAPIRS ( ) FROM THE AMERICAS Introduction The extant tapirs, Bairds tapir ( ), lowland tapir ( ), mountain tapir ( ), and Malayan tapir ( ) occupy forest environm ents in southern Mexico, Central Amer ica, South America and southeast Asia (Salas, 1996; Brooks et al., 1997; Foerster and Vaughan, 2002; Tobler, 2002; Holden et al., 2003). These habitats include lowland forests, primar y and secondary forests, Amazonian flood-plains, and montane cloud forests (Terwilliger, 1978; Eisenberg, 1989; Fragoso, 1991; Salas, 1996; Brooks et al., 1997; Downer, 2001; Foerster and Vaughan, 2002; Tobler, 2002; Holden et al., 2003; Lizcano et al., 2004). is also found almost exclus ively in areas with dense vegetation, and rarely in more open habitats (Tobler, 2002). Even the mountain tapir which is capable of inha biting treeless paramo environments of >3800m elevation, lives in dense Andean forests more frequently than any other habitat type (Downer, 2001). Tapirs can also have limited home ranges of approximately 125 ha or less (Foerster and Vaughan, 2002), making them good indicators of their local environments and thus able to provide information about local envi ronments in the fossil record. Tapirs are browsers with a di et consisting of leaves, twigs, fiber, and fruit (Terwilliger, 1978; Bodmer, 1990a, 1991; Fragoso, 1991; Salas a nd Fuller, 1996; Henry et al., 2000; Downer, 2001; Galetti et al., 2001; Tobler, 2002; Lizcano and Cavelier, 2004). They primarily browse in forest environments throughout the year (Bodmer, 1990a; Salas, 1996; Tobler, 2002; Foerster and Vaughan, 2002; Henry et al., 2000). Even with seasonal flooding, dietary differences are absent or minor, potentially varyi ng with fruiting events (Bodmer, 1990b; Henry et al., 2000). In addition, their craniodental morphology is high ly conservative over time in comparison to

PAGE 20

20 tapiroid fossils (Colbert and Schoch, 1998; DeSan tis and MacFadden, 2007) and is similar to that of other extinct and extant browsing mammals (MacFadden and Shockey, 1997; Mendoza et al., 2002). Stable carbon isotope values of tapirs ove r the past 10 million years similarly indicate that they maintained a diet of predominantly C3 plants, likely consumed in the denser canopied forests locally available (DeSantis and MacFadde n, 2007). However, it is not clear how variable extant tapir diets and subsequent 13C values are at the individua l level during ontogeny, at the population and species level, a nd how environmental and climatic variables influence their 13C and 18O values. In order to bette r understand the ecology of thes e elusive mammals both today and in the past, it is critical to und erstand their stable isotope ecology. Stable carbon isotopes are incorporated into mammalian tooth enamel, retaining isotopic signatures reflective of ones diet (DeNiro and Epstein, 1978; Krueger, 1991; Lee-Thorp and van der Merwe, 1991; Cerling et al., 1997; Ce rling and Harris, 1999). Specifically, 13C signatures of C3 and C4 plants are incorporated into the t ooth enamel of medium to large bodied herbivorous mammals with an enrichment factor of 14.1 (although non-ruminants, including tapirs, may have an enrichment factor between 12 and 13; Cerling and Harris, 1999). Based on modern analogues, 13C enamel values < -8 reflect a predominantly C3 diet (e.g. trees, shrubs, and cool season grasses) with more 13C depleted values indicating denser canopied environments (MacFadden et al., 1996; Cerling et al., 1997, 1999, 2004; van der Merwe and Medina, 1989, 1991; Cerling and Harris, 1999). Variation in 13C values between individual teeth can also indicate seasonal differences in di et, potentially reflective of seasonal changes in vegetation due to water stress (Roux et al., 2001; Ehleringer et al., 2002). As 13C signatures do not change with time during the process of fo ssilization, ratios from fossil mammals can be interpreted as similar to those of today, plus 1.5 due to modern atmospheric CO2 enrichment

PAGE 21

21 (Ehleringer and Monson, 1993; Cer ling et al., 1997; Passey et al., 2002). However, in order to properly interpret the 13C values of fossil mammals including tapirs, it is first necessary to understand 13C variation in extant indi viduals, populations, species, an d across varying climatic regimes. Stable oxygen isotopes are similarly incorpor ated into mammalian tooth enamel. Instead of a diet signal as with 13C, 18O values are a function of body wa ter that reflects the response of meteoric water to changes in temperatur e and/or precipitation/humidity (Dansgaard, 1964; Longinelli, 1984; Bryant et al., 1994, 1996a,b; Bryant and Fr oelich, 1995; Bocherens et al., 1996; Kohn et al., 1996; Sponheimer and LeeThorp, 1999; Higgins and MacFadden, 2004; MacFadden and Higgins, 2004; Hoppe 2006). In terrestrial ecosys tems, seasonal variation is recorded in tooth enamel with higher 18O values indicating high summer temperatures as compared to lower 18O values during cooler winters (Fricke and ONeil, 1996; Feranec and MacFadden, 2000; MacFadden and Higgins, 2004 ). Oxygen isotopes of mammalian tooth enamel can also vary between or among taxa oc cupying similar environments due to variations in the proportion of drinking wate r ingested, as opposed to more ev aporated plant water (Levin et al., 2006). Mammals that obtain a large proportion of their water from drinking are thought to be insensitive to large changes in 18O values with increased aridity, as opposed to taxa that obtain a larger portion of their water from plants (Levin et al., 2006). Tapi rs are thought to track meteoric water based on their consumption of water from free water sources (e.g. rivers and lakes) and prolonged periods of time spent in the water (B odmer, 1990b). However, extinct tapirs during the Pliocene and Pleistocene show a significant decline in 18O values with increased aridity associated with interglacial warming (DeSantis et al., 2009). Thus, it is critical to understand how 18O enamel values of extant tapirs are associated with climatic variables including

PAGE 22

22 precipitation and temperature. Clarifying thes e relationships will enable a better understanding of the ecology of extant tapirs and allow for informed interpre tations of the paleoecology and paleoclimatology of fossil site s containing fossil tapirs. Here, 13C variation of extant tapirs from th roughout the Americas is assessed to determine how 13C and 18O values vary over the first few years of a tapirs life, at the population and species level, and how climatic variables including precipitation and temperature affect 13C and 18O values. Individual and population variation is first examined in a Mexican population of Bairds tapir ( ) by first comparing early a nd late erupting teeth to evaluate the potential effects of weaning on 13C and 18O values. Second, 13C and 18O variation is quantified at the population level to determine dietary resource use and further evaluate if a small number of specimens can reliably estimate a populations mean. Third, species level 13C and 18O variability is quantified in the lowland tapir ( ) and the mountain tapir ( ); subsequently, quantifying the effects of climatic (i.e. temperature and precipitation) and geographic (i.e. lat itude and elevation) parameters on 13C and 18O values at the genus and species le vel. This work enables a better understanding of the ecology of extant tapirs while also constraining paleoecological interpretations of extinct tapirs and their environments through time. Materials and Methods Extant tapirs ( ) from Mexico were sampled from the Osteological Collec tion at the Yale Peabody Museum (YPM) in order to quantify dietary variation and potential ontogenetic dietary shifts, as inferred from stab le isotopes. The 11 i ndividuals sampled were initially collected with in the two-year period of 1873-1874 in Acapulco, Mexico. Because the osteological specimens of were collected from the same general area, 13C variation can be attributed to dietary vari ation. Ontogenetic diet shifts were quantified by comparing early

PAGE 23

23 erupting first molars (M1) to la te erupting fourth premolars (P4) and third molars (M3) of like individuals. The 13C and 18C values of different tooth positi ons were compared to one another using paired Students t-tests as these data were normally dist ributed (Shapiro-Wilks test). Similar to the work of Clementz and Koch (2001) I conducted a power analysis to test if small sample sizes of tapir teeth are likely to repr esent a populations mean. By comparing population means with the mean of three randomly select ed samples, I assessed if three samples can approximate a population mean (as estimated from the 13C values from one or more tooth positions). Each tooth was assigned a number from 1-33 and a random number generator was used to select three teeth, repl icated 100 times per tooth erupti on category. The resulting mean 13C and 18O values of M1s, P4s, M3s, late erupting teeth (P4s and M3s), and all tooth positions were compared to population means, and differe nces between the sample mean and population mean were calculated. In addition to the specimens from Acapulco, Mexico, extant tapirs ( n = 19; s, n = 15; and n = 3) were also samp led from Mexico, Central America, and South America based on specimens in the American Museum of Natural History (AMNH) Mammalogy Collection (Figur e 2-1, Table 2-1). Late erup ting third molars (or fourth premolars when third molars had not erupted ) were sampled from these specimens. All specimens selected are associated with geogr aphic information stating where they were collected. Based on the geographic information, da ta from the closest climate stations were associated with respective 13C and 18O values (National Climatic Data Center, 2009; Table 21). Mean data from two climate stations were averaged when specimens were located approximately equidistant from them. Data from as many years as were available were included; however, equal numbers of months was kept consistent (e.g. if only 10 January months were

PAGE 24

24 available, then only 10 February, 10 March, 10 April, etc.. months were included). The climatic data compiled and analyzed include, the following (with abbreviations and units noted): mean monthly temperature (MT, C), mean monthly precipitation (MP, mm), precipitation frequency (PF, mean number of precipitation days per month), and estimated 18O local precipitation values ( 18Op, computed using the Online Isotopes in Precipitation Calculator available at www.waterisotopes.org, based on latitude, longitu de, and elevation). Geographic information, including latitude (absolute valu e of decimal degrees north) and el evation (m) were also included in the analysis. Mean 13C and 18O values were compared at the species level using both parametric ANOVA Fisher LSD a nd non-parametric Kruskal-Wallis tests. I assessed which climatic, geographic, and isot opic variables were correlated (Pearson correlations) with 13C and 18O enamel values. Multiple regressions were used to assess the contribution of the climatic and geographic variables to 13C and 18O values. Variables that were highly correlated, Pearson correlation coefficients 0.95 were removed from individual an alyses. Linear regressions were used to determine the relationship between 18O and 13C enamel values from like specimens. Additionally, linear regressi ons were use to determine the relationship between 18O enamel values and estimated 18O local precipitation values. All analyses were done at the generic level (including all species except due to its geographic distance from the remainder of the taxa) and species level (with the added exception of due to limited sample size, n = 3). All enamel samples were acquired by drilli ng approximately 2-3 mg of sample using a low speed ForedomTM drill and carbide dental burrs. All enamel powder was pretreated with 30% hydrogen peroxide for 24 hours and 0.1N acetic acid for 12 hours to remove organics and secondary carbonates, respectively (Koch et al ., 1997). These samples (~1 mg) were run on a

PAGE 25

25 VG Prism stable isotope ratio mass spectrometer w ith an in-line ISOCARB automatic sampler in the Department of Geological Scienc es at the University of Florida. The analytical precision is +/0.1. Normalized data (to NBS-19) are reported in co nventional delta ( ) notation for carbon ( 13C) and oxygen ( 18O), where 13C (parts per mil, ) = ((Rsample/Rstandard)-1) 1000, and R = 13C/ 12C; and, 18O (parts per mil, ) = ((Rsample/Rstandard)-1) 1000, and R = 18O/ 16O; and the enamel carbonate standard is VPDB (P ee Dee Belemnite, Vienna Convention; Coplen, 1994). Results Individual and Population Level Variation ( T. bairdii ) 13C values of late erupting (P4 and M3) teet h are significantly greater (by 1.5 to 1.7) than the early erupting first molar ( < 0.0001 for both paired Student s t-tests of M1-P4, and M1-M3; Table 2-2 and 2-3). The 18O values do not show this same pattern, with M1 and P4 teeth sharing identical means (-5.8). Instead 18O values of P4 and M3 teeth are significantly different from each other (difference of -0.5, = 0.004). Additionally, dietary variation at the population level is low with total 13C variation of 2.2, 2.3, and 2.9 for M1s, P4s, and M3s, respectively (Figure 2-2). Similarly, 18O variation is low with total variation of 2.6, 1.4, and 2.2 for M1s, P4s, and M3s, respectively (Figure 2-2). Based on comparisons of the stable isotope va lues of three randomly selected samples to population means, only a few samples are required to approximate population means. This is because the mean 13C values of randomly selected sub-samples were consistently within 1.5 of the population mean, with average variation of 0.3 for P4, 0.4 for M3, 0.4 for late erupting teeth (P4s and M3s), and 0.5 for all sampled teeth (M1s, P4s, and M3s). Mean 18O values were also all within 1.1 (with an averag e difference of 0.3) using all combinations of tooth positions sampled.

PAGE 26

26 Species Level Variation ( T. bairdii, T. pinchaque, T. terrestris ) Mean 13C values (+/1 SD) of and are -14.8+/0.9, -14.3+/-1.4, and -15.6+/-1.6, respectively (Figure 2-3). The 13C values of range from -18.1 to -12.8, yielding the gr eatest total range ( 5.3) as compared to (ranging from -16.4 to -13.0, total range of 3.4) and (ranging from -15.9 to -13.1, total range of 2.8; Figure 2-3) also has individuals with the lowest 13C values of all species sampled. However, 13C values of and are not significantly differen t from one another. Mean 18O values (+/1 SD) of and are -4.5+/-1.5, -8.0+/0.4, and -3.3+/-1.4, ranging from -6.0 to -0.4, -8.4 to -7.6, and -5.2 to -0.9, respectively (Figure 2-3). In contrast, the 18O values of these tapirs are signif icantly different from one another ( = 0.001; Kruskal-Wallis), with 18O values of significantly less than both ( < 0.001; Fisher LSD) and ( < 0.0001; Fisher LSD). The 18O values of are also significantly less than ( = 0.021; Fisher LSD). Climatic and Geographic Parameters ( T. bairdii, T. pinchaque, T. terrestris ) The 18O values of all extant tapirs analyzed ( ) are negatively correlated with elevation (-0.48, = 0.003; Table 2-4). Multiple linear regressions demonstrate that precipitation fr equency (the mean number of precipitation days per month) has the greatest relative weight (standardized coefficients +/standard error, and -values noted; 2.04 +/0.29, < 0.0001), followed by mean total precipitation ( -1.35 +/0.23, < 0.001), elevation (-1.03 +/0.20, < 0.0001), latitude (0.80 +/0.17, < 0.0001), and estimated 18O local precipitation values (-0.50 +/0.20, = 0.016). The multiple linear regression model, 18O enamel values = -12.6 0.87*PF 3.45*MP + 0.26*latitude 0.78* 18Op 0.004*elevation, explains ~75% of the variation in tapir 18O enamel values (R2 = 0.75, adjusted R2 = 0.71, <

PAGE 27

27 0.0001). Mean monthly temperatures were remove d from the analysis as they are highly correlated with elevations (-0.95, < 0.0001) and elevation is more correlated with 18O values (-0.48, = 0.003) than mean monthly temperature (0.38, = 0.019; Table 2-4). The 13C values of all tapirs are instead negatively correlated with mean total mont hly precipitation (-0.50, = 0.002; Table 2-4). A multiple li near regression model was sim ilarly developed to assess the weight of the contributions of c limatic and geographic variables on 13C values, with the additional exclusion of the estimated average 18O precipitation variable. None of the variables analyzed are significant, and the model onl y accounts for ~30% of the variation (R2 = 0.30, = 0.021). Stable carbon and oxygen isotope values of and were analyzed at the species level to determine which variables were correlated with 13C and 18O variation. The 18O enamel values are positively correlated w ith both precipitation variables, precipitation frequency (0.84, < 0.0001) and mean monthly precipitation (0.63, = 0.004), and negatively correlated with estimated 18O local precipitation values (-0.76, = 0.0002) in (Table 2-4). Mean number of precip itation days explains ~70% (R2 = 0.70, < 0.0001, y = 0.56x-7.8) of the variation in 18O enamel values (Figure 2-4). Additionally, estimated 18O local precipitation values explain ~58% (R2 = 0.58, = 0.0002, y = -2.14x-13.41) of the variation in 18O enamel values. A multiple linear re gression model incorporates precipitation frequency, mean monthly temperatures, and estimated 18O local precipitation values, as all other variables were removed due to multicolinearity. 18O enamel values = -68.1 + 1.14*18Op + 2.24*MT + 1.14*PF (R2 = 0.80, adjusted R2 = 0.76, < 0.0001), with precipitation frequency yielding the greatest st andardized coefficient (+/stand error) of 1.72 +/0.48 ( = 0.003) followed by mean monthly temper ature with a value of 0.72 +/0.30 ( = 0.026). The

PAGE 28

28 13C enamel values of are not correlated with any clim atic or geographic variable, nor are they correlated with 18O enamel values of like specimens (Table 2-4). 18O enamel values are positively correlated with latitude (0.68, = 0.005) and negatively correlated with mean monthly precipitation (-0.61, = 0.016; Table 2-4). 18O enamel values = 81.5 1.04* 18Op 3.17*MT + 0.65*PF 5.62*MP 1.36*latitude 2.68*elevation (R2 = 0.89, adjusted R2 = 0.80, = 0.002), with mean monthly precipitation yielding the greatest standardized coefficient (+/stand error) of -3.12 +/0.84 ( = 0.006) followed by precipitation frequency with a value of -1.66 +/0.59 ( = 0.023). 18O enamel values are not significantly corre lated with estimated local precipitation 18O values (Table 2-4). Similarly, 13C enamel values of are negatively correlated with mean monthly precipitation (-0.56, = 0.031) and positively correl ated with latitude (0.78, = 0.001; Table 2-4) and with 18O enamel values of like specimens (0.93, < 0.0001). Although ~83% of the variation in 13C enamel values is explained using the following regression: 13C enamel values = 17.41 1.3*MT 0.02*MP + 0.28*latitude 0.014*elevation + 0.37*PF (R2 = 0.83, adjusted R2 = 0.73, = 0.003), none of the included variables significantly contribute. However, 13C enamel values explain ~85% (R2 = 0.86, adjusted R2 = 0.85, < 0.001, y = 9.95+0.85x) of the variation in 18O enamel values in (Figure 2-5). Discussion Implications of Ontogenetic and Population Level Variation Previous work has documented 18O enrichment in early erupting teeth due to weaning from the mothers milk at the time of mineraliz ation (Bryant and Froeli ch, 1995; Bryant et al., 1996a,b; Fricke and ONeil, 1996; Zazzo et al., 2002) Although tapirs are expected to exhibit the same pattern, 18O values of early and late erupting teeth are not significantly different.

PAGE 29

29 Instead, 18O values of the late erupting P4 and M3 te eth are significantly diffe rent, likely due to differences in seasonality. As the M3 erupts af ter the P4 (based on th e examination of tapir specimens at the AMNH and YPM), thes e teeth are likely reflecting the 18O values of subsequent seasons that vary due to either temperature and/or precipitation. Additionally, because temperature is not highly variable (i.e. ~3 C variation between the lowest and highest monthly mean temperatures; National Climatic Data Center, 2009) in Acapulco, Mexico whereas the region does exhibit a distinct wet and dry season (i.e. on average 6 months out of the year experience < 2 days of precipita tion per month; National Climatic Data Center, 2009), seasonal precipitation differences are the most likely expl anation for the significa nt differences between late erupting teeth. However, 18O values of early erupting teeth may still be recording a weaning signal that is swamped out by di fferences in precipitation regimes. In contrast, 13C values of early erupting M1 teeth ar e significantly different from both of the late erupting teeth (P4 a nd M3), suggesting ontogenetic dietary differences. These differences may be due to the consum ption of lipid-rich milk that is 13C depleted (DeNiro and Epstein, 1978; Tieszen et al., 1983; Hobson and Sease, 1998) or due to the consumption of browse in a denser canopy then occupied during adulthood (van de r Merwe and Medina, 1989, 1991; Cerling et al., 2004). As th e stripped and spotted pelage of juvenile tapirs is thought to help with concealment from predators (e.g. the puma; Eisenberg, 1989), their presence in a denser canopy during juvenile years may further help protect them against predation. Alternatively, juvenile tapirs may be consuming a larger proportion of 13C depleted leaves and subsequently consuming a lower proportion of 13C enriched fruits (Cerling et al., 2004; Codron et al., 2005). Although it is difficu lt to discern the exact reason for 13C depletion in early

PAGE 30

30 erupting teeth, assessment of dietary variation at the population or species level should include like tooth positions to avoid ontogenetic dietary differences. Low 13C variation of at the population level is cons istent with their inferred conservative diet. Although 13C variation only captures an asp ect of dietary variation, limited 13C variation indicates the consump tion of food items with similar 13C values and/or from forests with similar canopy density. Low 13C and 18O variation at the population level makes tapirs ideal taxa for discerning environmental an d climatic variation. Further sub-sampling also established that 13C and 18O mean values of three randomly selected teeth, from both similar and different tooth positions, estimates the population mean. This is critical for inferring past environments as often tapir fossils are a rare component of a fauna, and the specimens available may be from teeth that erupted at different times. Thus, the stable isotope analysis of three or more extinct or extant tapir teeth is likely to estimate a populations mean. Therefore, stable carbon isotope analyses of extinct ta pirs can help elucidate past di ets and forest structure, while modern tapir isotopic data can be compared to museum sp ecimens to understand dietary variation over time. Influence of Climatic and Geographic Variables All tapirs ( ) have 13C values consistent with a predominantly C3 diet, with all valu es less than -12. also has individuals with the lowest 13C values of all species sampled (<16.5), s uggesting their consumption of browse in the densest canopied environments. This interpreta tion is consistent with their presence in dense Amazonian rainforests. Alternatively, lower 13C values may be indicati ve of a diet consisting of fewer 13C enriched fruits (Cerling et al., 2004; Codron et al., 2005). Furthermore, all 13C values of teeth are consistent with their c onsumption of browse in forest

PAGE 31

31 environments, further supporting their utilization of forest environments despite their ability to exist in treeless paramo environments (Downer, 2001). Significantly different 18O values between all tapirs likely reflects climatic differences influencing local meteoric water consumed and potential differences in tapir behavior. The 18O values of are the lowest due to their presence in high elevation environments. With increased elevation, the heavier 18O isotope is preferentia lly rained out leaving 18O depleted rainfall at higher elevations (Poage and Chambe rlain, 2001). Any water consumed via free water sources or plant water has lighter 18O values than water at lower elevations. The significant contributions of elevation and estimated 18O values of local precipitation towards explaining 18O values are consistent with the pattern of lower 18O values with increasing elevation. Once tapirs are analyzed at the speci es level and the high elevation is excluded from analyses, elevation is less of a contributing factor to either 18O values or 18O values. Although 18O values are positively correlated with elevation, this pattern is in contrast to biogeochemical processes (i.e. decreased precipitation 18O values with increased elevation, Poage and Chamberlain, 2001) and is unlikely a driving factor as all specimens are from elevations of < 60 m. Instead, precipitation variables are greater contributors to and 18O values than elevation. has the greatest 18O values, significantly greater than both and These differences in 18O values are likely due to climatic variables. Mammals living in more arid and/or warmer areas are predicted to have either similar or greater 18O values with decreased precipitation (Levin et al ., 2006); however, this pattern is in contrast to what is observed here. Instead, 18O values increase with increased precipitation (both frequency and amount; Figure 2-4). This pattern may instead demonstrate that is

PAGE 32

32 changing its behavior when present in dryer areas, dri nking a greater proportion of 18O deplete water from lakes and rivers when precipitation is low. These differences in precipitation may further be amplified by severe dry seasons that are more typical in regions such as Acapulco, Mexico (yielding < 2 days of precipitation per m onth for six months per year; National Climatic Data Center, 2009). Temperature may further influence drinking behavior, as higher temperatures are here correlated with lower 18O values, and thus indicative of increased drinking. This is in contra st to the typical pa ttern of increased 18O values with increased temperature (Fricke and ONeil, 1996; Feranec and MacFadden, 2000; MacFadden and Higgins, 2004), further supporti ng the hypothesis that tapirs are changing their drinking behavior. 18O values are instead negatively correlated with mean monthly precipitation, demonstrating that tapirs present in areas with higher prec ipitation (e.g. water is no longer a limiting factor) may instead respond as predicted. The 13C values of explain the majority of the variation in 18O values. As evaporation is low in denser canop ied forests (Klaassen, 2001), lower 18O values in plant tissues are expected to occur in denser forests with lower 13C values. Therefore, the observed correlation between 13C and 18O values is expected if is obtaining a large proportion of its water from plant leav es in the forest understory. As is present in wetter Amazonian rainforests, it is likely that is consuming a greater proportion of its water from plant leaves as compared to which may need to supplement its water intake by drinking more from free water sources (e.g. lakes and rivers). This is further evidenced by individuals present in areas with a higher precipitation frequency ( 5 days per month), which demonstrates a similar correlation between 13C and 18O values (Figure 2-5). Additionally, variation in 18O values are largely explained by estimated 18O local

PAGE 33

33 precipitation values. This pattern is expected if is obtaining a large portion of water from free water sources. Although tapirs do not occur far from permanent water sources (Eisenberg, 1989), the proportion of water they c onsume via drinking may change with climatic parameters. Thus, tapirs may present a different relationship from the evaporation sensitive (i.e. 18O values increase with increased arid ity) and evaporation insensitive (i.e. 18O values are relatively unchanged with increa sed aridity) taxa mentioned by Levin et al. (2006). Instead, tapirs exhibit decreased 18O values with increased aridity. This is due to the need to adapt and change behavior, i.e. drink more 18O depleted water, when present in dryer areas (defined according to precipitation frequency and/or amount ). Thus, the previously mentioned example demonstrating a significant decline in tapir 18O values with interglacial warming in Florida (DeSantis et al., 2009) suggests incr eased aridification based on tapir 18O values. Furthermore, the strength of the relationship between 13C and 18O values indicates that tapirs may consume a large proportion of their body water from plant leav es when precipitation is not a limiting factor. A tapir that gets a significant po rtion of its water from leaves may consume less water from rivers and open water. While it is unclear how adaptable an individual tapir is to changing its behavior in response to its local environment, further study of extant tapir species (e.g. vs. ) may elucidate their adaptability and differences in the relative proportion of time spent in or consum ing water from rivers and lakes. Conclusions In summary, low population variation in carbon and oxygen isotopes makes tapirs ideal taxa for inferring paleoenvironm ents, even when only low sample sizes are available, as is characteristically the case in many fossil a ssemblages. Tapir stable oxygen isotopes vary between species and are largely influenced by precipitation (both amount and frequency). Oxygen isotope variation at th e species level is largely a function of precipitation and 13C

PAGE 34

34 values in and respectively. This is in contra st to the expectation that tapir 18O values will most accurately reflect 18O meteoric water values. While 18O values of do show a strong rela tionship with estimated 18O precipitation values, these values are further influenced by climatic and geographic vari ables. Additionally, tapirs appear to change their drinking behavior (i.e. re lative proportion of free water cons umed) with precipitation. This demonstrates the adaptability of tapirs to compensate for low water availability, by increasing drinking behavior. A detailed understanding of modern tapir ecology also helps elucidate paleoecological and paleoc limatic interpretations.

PAGE 35

35 Table 2-1. 13C and 18O values of extant tapirs ( and ) noting climate stations and associated climatic and geographic variables. Taxon Museum ID TP 13Ce 18Oe Country Climate Station MT PF MP L E 18Op AMNH 29455 lm3 -16.4 -3.6 Nicaragua 78741 27.3 8.1 108.7 12.15 56 -5.1 AMNH 29526 lm3 -14.5 -3.3 Nicaragua 78741 27.3 8.1 108.7 12.15 56 -5.1 AMNH 35000 lm3 -15.4 -3.5 Nicaragua 78741 27.3 8.1 108.7 12.15 56 -5.1 AMNH 80075 lm3 -15.0 -1.9 Honduras 78705 26.6 11.3 206.6 15.73 26 -4.9 AMNH 80076 lp4 -14.2 -0.4 Honduras 78705 26.6 11.3 206.6 15.73 26 -4.9 AMNH 204706 LM3 -14.3 -3.5 Mexico 76833 28.4 3.9 71.3 16.17 2 -4.1 AMNH 206834 lm3 -14.7 -4.6 Mexico 76833 28.4 3.9 71.3 16.17 2 -4.1 AMNH 208259 lm3 -15.9 -5.9 Mexico 76833 28.4 3.9 71.3 16.17 2 -4.1 YPM 6712 RM3 -13.3 -4.9 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7132 RM3 -15.6 -5.6 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7133 RM3 -14.3 -5.8 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7135 RM3 -15.9 -5.2 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7136 RM3 -14.3 -5.8 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7140 RM3 -14.3 -5.5 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7141 RM3 -15.3 -5.3 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7143 RM3 -14.0 -4.6 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 7477 RM3 -14.9 -6.0 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 8626 RM3 -15.7 -3.8 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 YPM 9398 RM3 -13.0 -5.9 Mexico 76805 27.4 5 115.9 16.08 3 -3.8 AMNH 70521 lp4 -15.9 -8.4 Ecuador 84179 20.8 20.3 341.1 -1.5 960 -6.9 AMNH 149331 lm3 -13.1 -7.6 Colombia 80342 19.3 9.7 90 1.42 1826 -8.5 AMNH 149332 lp4 -13.9 -8.0 Colombia 80342 19.3 9.7 90 1.42 1826 -8.5 AMNH 14690 lm3 -15.7 -3.2 Colombia 80009 28.3 4.2 43.2 11.13 14 -4.6 AMNH 36661 lm3 -14.0 -2.4 Brazil 83361 26.1 9 125.2 -15.55 151 -5 AMNH 36662 lm3 -12.8 -0.9 Brazil 83361 26.1 9 125.2 -15.55 151 -5 AMNH 36663 lm3 -13.0 -1.2 Brazil 83361 26.1 9 125.2 -15.55 151 -5 AMNH 73596 lm3 -16.4 -4.9 Peru 84377 25.6 14.7 304.4 -3.78 126 -5.2 AMNH 73766 lm3 -17.0 -5.2 Peru 84377 25.6 14.7 304.4 -3.78 126 -5.2 AMNH 77573 lm3 -16.8 -4.7 Venezuela 80457 26.8 12.6 188 5.6 74 -4 AMNH 77576 lm3 -16.3 -3.8 Venezuela 80457 26.8 12.6 188 5.6 74 -4 AMNH 78518 lm3 -16.6 -4.0 Venezuela 80457 26.8 12.6 188 5.6 74 -4 AMNH 95133 lm3 -15.3 -2.6 Brazil 82331 27 12.7 187 -3.13 72 -5 AMNH 96130 lm3 -15.4 -2.3 Brazil 82191 26.6 18.9 261.6 -1.45 10 -2.8

PAGE 36

36 Table 2-1 Continued. Taxon Museum ID TP 13Ce 18Oe Country Climate Station MT PF MP L E 18Op AMNH 120996 lm3 -13.5 -1.4 Brazil 83611 23 8.6 122 -20.45 530 -5.5 AMNH 142280 lp4 -18.1 -5.1 Colombia 80315, 80234 24.8 13.8 191.8 -10.95 270 -5.6 AMNH 209139 LM3 -16.3 -4.3 Bolivia 85043 26.9 8.3 141.4 -11 141 -5.7 AMNH 217150 lm3 -16.6 -3.1 Bolivia 85245, 85154 25.9 11.0 166.6 -10.98 206 -5.4 Note: Table abbreviations and scales are defined as follows: Taxon, taxonomic name ; Museum, museum location of sampled specimen; ID, specimen identification number; TP, tooth positions with capital text indicating upper teeth, listing the side (l eft or right), tooth (premolar or mola r), and exact position (number); 13Ce, 13C enamel values (VPDB, ); 18Oe, 18O enamel values (VPDB, ); Country, country of origination; Climate Station, location where all c limate data were compiled and estimates of loc al 18O precipitation values were estimated; MT, mean monthly temperature (C); PF, precipitation frequency (mean number of precipitation days per month); MP, mean monthly precipitation (m m); L, latitude (decimal degr ees north); E, elevation (m); 18Op, estimated 18O values of local precipitation calculated using the Online Isotopes in Precipitation Calculator available at www.waterisotopes.org, based on latitude, l ongitude, and elevati on (VSMOW, Vienna Standard Mean Ocean Water, ).

PAGE 37

37 Table 2-2. 13C and 18O values from a population of extant tapirs ( ) in Acapulco, Mexico. 13C(VPDB, ) 18O(VPDB, ) YPM ID M1 P4 M3 M1 P4 M3 6712 -16.1 -14.3 -13.3 -5.3 -5.5 -4.9 7132 -16.4 -15.3 -15.6 -5.7 -6.2 -5.6 7133 -15.9 -14.9 -14.3 -5.5 -6.5 -5.8 7135 -16.0 -15.4 -15.9 -5.5 -5.2 -5.2 7136 -15.3 -14.5 -14.3 -5.9 -6.5 -5.8 7140 -16.0 -14.7 -14.3 -6.3 -5.9 -5.5 7141 -17.5 -14.6 -15.3 -4.8 -5.2 -5.3 7143 -15.8 -14.6 -14.0 -5.4 -5.4 -4.6 7477 -17.4 -15.5 -14.9 -7.4 -6.4 -6.0 8626 -17.1 -15.7 -15.7 -5.6 -5.1 -3.8 9398 -16.1 -13.4 -13.0 -6.3 -5.8 -5.9 mean -16.3 -14.8 -14.6 -5.8 -5.8 -5.3 SD 0.7 0.4 0.9 0.7 0.5 0.7 min -17.5 -15.7 -15.9 -7.4 -6.5 -6.0 max -15.3 -13.4 -13.0 -4.8 -5.1 -3.8 range 2.2 2.3 2.9 2.6 1.4 2.2

PAGE 38

38 Table 2-3. Stable carbon and oxygen isotope differences between various early and late erupting tooth positions, from a populat ion of extant tapirs ( ) in Acapulco, Mexico. 13C(VPDB, ) Differences 18O(VPDB, ) Differences YPM ID M1-P4 M1-M3 P4-M3 M1-P4 M1-M3 P4-M3 6712 -1.8 -2.8-10.2-0.4 -0.6 7132 -1.1 -0.80.30.5-0.1 -0.6 7133 -1 -1.6-0.610.3 -0.7 7135 -0.6 -0.10.5-0.3-0.3 0 7136 -0.8 -1-0.20.6-0.1 -0.7 7140 -1.3 -1.7-0.4-0.4-0.8 -0.4 7141 -2.9 -2.20.70.40.5 0.1 7143 -1.2 -1.8-0.60-0.8 -0.8 7477 -1.9 -2.5-0.6-1-1.4 -0.4 8626 -1.4 -1.40-0.5-1.8 -1.3 9398 -2.7 -3.1-0.4-0.5-0.4 0.1 mean -1.5 -1.7-0.20.0-0.5 -0.5 SD 0.8 0.90.60.60.7 0.4 value <0.0001* <0.0001* =0.22 =1.0 =0.04* =0.004* Indicates significant -values.

PAGE 39

39Table 2-4. Pearson correlation coefficients for stable car bon and oxygen isotope enamel ( 13Ce, 18Oe) values and climatic and geographic variables. All Tapirs 13Ce (VPDB, ) 18Oe (VPDB, ) 13Ce (VPDB, ) 18Oe (VPDB, ) 13Ce (VPDB, ) 18Oe (VPDB, ) Mean Monthly Temperature (C) -0.12 0.38* -0.08 -0.47* -0.31 -0.18 Precipitation Frequency (mean # of days per month) -0.42* 0.05 -0.08 0.84* -0.42 -0.38 Mean Monthly Precipitation (mm) -0.50* -0.03 0.13 0.63* -0.56* -0.61* Latitude (absolute value) 0.47* 0.30 0.31 -0.36 0.78* 0.68* Elevation (m) 0.19 -0.48* -0.28 0.57* 0.45 0.34 18Oprecipitation (VSMOW, ) -0.09 0.27 0.26 -0.76* -0.04 0.06 Indicates significant -values.

PAGE 40

40 Figure 2-1. A Google Earth ma p showing the location of (yellow), (purple), and (green) specimens from southern Mexico, Central America, and South America. The numbers in pare ntheses indicate the number of specimens from one location (if greater than one).

PAGE 41

41 Figure. 2-2. Box plots of stable carbon and oxyge n isotope values from a population of extant tapirs ( ) in Acapulco, Mexico. Mean values and the total ranges of variation are noted by the red plus signs and bars, respectively. Tooth positions are noted in order of eruption timing (i.e. first molar, M1; fourth premolar, P4; and third molar, M3).

PAGE 42

42 Figure. 2-3. Stable carbon (black) and oxygen (grey) isotope values of extant tapirs, (diamonds), (triangles), and (squares).

PAGE 43

43 Figure 2-4. Relationships betw een precipitation frequency (mean number of precipitation days per month) and oxygen isotope values in (diamonds; y=0.55x-7.80, R2=0.70, <0.0001) and (triangles; y=-0.15x-1.57, R2=0.15, =0.157).

PAGE 44

44 Figure 2-5. Relationship between carbon a nd oxygen isotope values from individual (triangles; y=0.85x+9.95, R2=0.86, <0.0001) and (diamonds; y=-0.08x-5.63, R2=0.002, =0.85) specimens, with grey diamonds indicating individuals present in areas with 5 precipitation days per month.

PAGE 45

45 CHAPTER 3 NEOGENE FORESTS FROM THE APPA LACHIANS OF TENNE SSEE, USA: GEOCHEMICAL EVIDENCE FROM FOSSIL MAMMAL TEETH Introduction Reconstructing the diet of ancient mammalian herbivores and their floral environment during the late Tertiary in eastern North Am eri ca is necessary to unders tanding the context of mammalian evolution in this poorly unders tood region. Global climate change and C3/C4 transitions are interpreted to have taken pl ace concurrently (Cerling et al., 1993, 1997; Wang et al., 1994). While these transitions l ead to dramatic increases in C4 grasses in North America approximately 7 mya (Cerling et al., 1993, 1997; Wang et al., 1994), it is unclear how eastern forests responded to such changes. It is possi ble that eastern North America sustained forest refugia, i.e. locations of relict populations of once widespread flora and fauna, during these transitions. The presence of a North American forest refugium has been proposed based on the Gray sites faunal and floral macrofossils including the abundance of forest-dwelling taxa (Wallace and Wang, 2004). Stable isotope sampling of the sites mammalian herbivores further clarifies our understanding of the paleoecology of this spatia lly and temporally rare site. During the late Miocene to early Pliocene, herb ivore diversity declined through a series of extinction events, with the once diverse shor t-crowned browsers experiencing proportionally greater declines than high-crowned (hypsodont ), presumed grazing he rbivores (Potts and Behrensmeyer, 1992; Janis et al ., 2000, 2002, 2004). These declines are often attributed to the increase in C4 grasslands resulting from increased aridity and/or reduced CO2 levels globally (Potts and Behrensmeyer, 1992; Cerling et al., 1993, 1997; Wang et al., 1994; Janis et al., 2000, 2002, 2004; Retallack, 2001; Stromberg, 2005) While it is clear that C4 grasses increased in abundance throughout the late Miocene to earl y Pliocene (Potts and Behrensmeyer, 1992; Cerling et al., 1993, 1997; Wang et al., 1994; Re tallack, 2001), the dr iving mechanisms

PAGE 46

46 responsible for apparent global co oling and increased seasonality are still a matter of debate. Nevertheless, it is possible that some forest envi ronments may have persisted as floral and faunal refugia within or in close proximity to C4 grasslands during this time of transition. As morphologically inferred browsers ap pear to be more numerous at th e Gray site, it is important to understand the true dietary feeding strategies of all ungulate taxa presen t. Reconstructing the diets of ungulates will also provide information on the associated flora and likely environments of the Gray Fossil Site and the broader ramificat ions for the ancient Appalachian forests. The southern Appalachians have existed fo r the past ~250 million years, potentially providing a relatively stable environment for the resident flora and fauna (Graham, 1964, 1999). Dominated by tropical flora during the Cretaceous the vegetation present dur ing the Paleocene to early Eocene suggests tropical ra in forests and a megathermal (i.e. humid and warm with mean temperatures of 20 C, de Candolle, 1874) climate (b ased on mesophyllous, entiremargined leaves in the Lower Eocene Wilcox Formation, TN). Middle Eocene tropical dry forests subsequently transitioned to modern warm-temperate deciduous vegetation at lower elevations and montane coniferous forests at higher elevations, dur ing the late Tertiary (Graham, 1964, 1999). These late Tertiary flora consisted of tropical vegetation (at the genus and family level) that were adapted to more temperate cl imates and interchanged with Asia and Europe (Graham, 1964). Additionally, mo lecular evidence of extant flor a and fauna (e.g. eastern tiger salamander red pine ) suggests that southern Appalachian refugia maintained ancestral populat ion of taxa, requiring milder climates, during periods of glaciations (Cresp i et al., 2003; Walter et al., 2005). Thus, understanding the paleoecology of the southern Appalachians may help us to better understand present floral and faunal diversity.

PAGE 47

47 The primary objective of this study is to reconstruct the ancient diets and paleoenvironments from the fauna of the Gr ay Fossil Site, from stable carbon and oxygen isotopes of ungulate tooth enamel. I also interpret th e paleoclimatic records of serial samples from tooth and tusk enamel, determining seasonal variability. Additionall y, I present Rare Earth Element (REE) analyses of a subset of our ungul ate taxa to determine the taphonomic context of the mammalian herbivores, determining if they represent a sympatric fauna (Trueman, 1999). The results of this study provide critical information to underst anding the unique paleoecological dynamics occurring during the late Tertia ry in the southern Appalachians. Background Stable Isotope Analysis: A Theoretical Foundation Vertebrate fossil remains can clarify pale oecological hypotheses by allowing for an independent measure of habitat type, as inferred from stable isotope ratios. Stable carbon isotopes are incorporated into th e lattice of enamel hydroxyapatite; therefore, retaining dietary isotopic signals that are reflec tive of plants consumed (DeN iro and Epstein, 1978; Krueger, 1991; Lee-Thorp and van der Merwe, 1991; Cerli ng et al., 1997; Cerli ng and Harris, 1999). Because 13C /12C ratios vary depending on a plants photosynthetic pathway and do not decay with time (Ehleringer and Monson, 1993), the ratios of the past can be interpreted as similar to those of today (Cerling et al., 1997). Additionally, 13C signatures of C3 and C4 plants are incorporated into the tooth enamel hydroxyapat ite of medium to la rge bodied herbivorous mammals with an enrichment factor of 14.1 (although non-ruminants may have an enrichment factor between 12 and 13; Cerling and Harris, 1999). However, due to modern atmospheric CO2 enrichment, floral and faunal 13C values are an additional 1.5 enriched today, as compared to the past (Cerling et al., 1997; Passey et al., 2002). Therefore, 13C values between 21 and 7 reflect a C3 diet, whereas values between 2 and 4 indicate a C4 diet

PAGE 48

48 (MacFadden et al., 1996; Cerling et al., 1997, 1999, 2004; Cerling and Harr is, 1999). Variation in 13C values within individual teeth can also indicat e seasonal differences in diet, reflective of seasonal changes in vegetation due to water stress (Roux et al., 2001; Ehleringer et al., 2002). Stable carbon isotope data of extinct and extant taxa can also be used to reconstruct rainforest distributions due to greater 13C discrimination occurring in dense closed canopies as compared to more open C3 environments (van der Merwe an d Medina, 1989; Cerling et al., 2004). Because 13C values increase with decreasing canopy density and/or in creasing distance from dense forest edges (van der Merwe and Medina, 1989; Kapos et al., 1993; West et al., 2001), more negative 13C enamel values of mammalian herbivor es reflect the consumption of browse in forests with denser canopies (van der Merwe and Medina, 1989, 1991; Cerling et al., 2004). As temperate forest floral 13C values are typically more enriched in 13C as compared to tropical forests (e.g. Cerling et al., 2004; Tu et al., 2004; BASIN Network, 2006), floral macrofossils and palynological evidence can further constrain interpretations of forest density. Therefore, 13C values of mammalian tooth enamel can indicate meaningful differences in habitat type within C3 flora and therefore assist in determining relativ e canopy density. Variation in stable oxygen isotopes of mamm alian tooth enamel is a function of body water that reflects the respons e of meteoric water to cha nges in temperature and/or precipitation/humidity (Dansgaard, 1964; Bryant et al., 1994, 1996a,b,c; Bocherens et al., 1996; Kohn et al., 1996; Sponheimer and Lee-T horp, 1999; Higgins and MacFadden, 2004; MacFadden and Higgins, 2004; Hoppe 2006). In terrestrial ecosys tems, seasonal variation is recorded in tooth enamel with more positive 18O values indicating high su mmer temperatures as compared to more negative 18O values during cooler winters (F ricke and ONeil, 1996; Feranec and MacFadden, 2000; MacFa dden and Higgins, 2004).

PAGE 49

49 Oxygen isotopes of mammalian tooth enamel can also vary between taxa occupying similar environments due to variations in the prop ortion of water ingested in the form of drinking water, as opposed to more evaporated plant water (Levin et al., 2006). By comparing the 18O values of evaporation sensitive taxa (i.e. 18Oenamel values increase with aridity) to evaporation insensitive taxa (i.e. 18Oenamel values track meteoric water) present at the same site, the 18O values of mammalian tooth enamel may be used as an index of terrestrial aridity (Levin et al., 2006). Because rhinos are evapor ation insensitive (Levin et al., 2006), comparisons of the 18O values of rhinos to the likely evaporati on sensitive camels (i.e. non-domesticated camelids typically acquire a large proportion of their water from plants), ma y elucidate relative aridity at the Gray site. Rare Earth Element Analysis: Understanding Taphonomic History REE patterns of fossilize d skeletal material within a deposit can be compared to determine whether a site has experienced significant mixi ng and/or reworking (Trueman, 1999). Because REEs are taken up in skeletal tissue in higher concentrations beginning shortly after death and continuing for approximately 10,000 to 30,000 years during diagenetic recr ystallization, REEs patterns of fossil enamel and den tin reflect the geochemistry of the local pore-water during that time (Henderson et al., 1983; Trueman, 1999; Patric k et al., 2001; MacFadden et al., 2007). As early recrystallization results in reduced porosity and flow, in itial REE patterns are typically preserved throughout latter di genesis (Trueman, 1999). Similar REE patterns indicate similarities in the geochemistry of the pore-wa ter, and comparable depositional environments (Trueman, 1999; Trueman et al., 2004). Therefore, it is possible to compare REE patterns of fossils from the Gray site to determine if they shared similar depositional environments or were reworked from a spatially and/or te mporality distinct locality.

PAGE 50

50 Gray Fossil Site, Tennessee, USA The Gray Fossil Site, located in Washington, Co., Tennessee, USA (Figure 3-1) is biostratigraphically dated between 4.5 and 7 Ma, based on the presence of the rhino and short faced bear (Wallace and Wang, 2004). The site is a sinkhole deposit consisting of finely laminated clays, silts, and fine sands with intermixed gravel lenses. The deposit resulted from a paleosinkhole lake that was approximately 2 ha in size and up to 39 m thick (Wallace and Wang, 2004; Shunk et al., 2006). The vertebrate taxa ar e of North American and, somewhat surprisingly, Eurasian ancestry (Tab le 3-1). The ungulate taxa include, the tapir rhino cf. camel cf. sp., peccary Tayassuidae, and gomphotheriid proboscidean (Table 3-1). Additionally, th e Gray site preserves a large population of the extinct tapirs ( ) represented by over 70 individuals, an order of magnitude larger than the total number of individuals of all other ungulate taxa. Given that fossil tapirs are robust indicators of ancient forests (D eSantis and MacFadden, 2007), the Gray site likely indicates the presence of forest environments. Floral macrofossils and pollen likewise suppor t a forest interpretation for the Gray site. Palynological evidence suggests a predominantly oak ( ) and hickory ( ) forest, consisting of nearly 70% of the pollen (Wallace and Wang, 1994; Table 3-1). Along with pine ( ), representing 9% of the polle n analyzed, the remaining flora present occurs in even lower abundances (Wallace and Wang, 1994). Because grass pollen occurs in such low abundance (< 2%), and anemophilous (wind dispersed) grass polle n is often over represen tative of a sites flora when present (Horrocks et al ., 2000), it is un likely that C4 grasses made up a substantial portion of the vegetation present at the Gray site. Additionally, the 13C values of all bulk organic matter sampled from the site range in value from -28 to -24, representing only C3 flora (Shunk et al., 2006).

PAGE 51

51 Materials and Methods Morphological Measurements Selected dental measurements were ta ken to quantify rela tive crown height. Measurements of crown heights and widths of th e lower third molars (m3) were taken from all specimens isotopically sampled with m3s present. A hypsodonty index (H I) was then calculated for all available m3s according to Janis (1988), by dividing the unworn crown height by the m3 width. Lower third molars with significant wear were excluded when determining mean HI values. These HI values were then used to determine if hypsodonty is predictive of diet, as inferred from stable carbon isot opes at the Gray Fossil Site. Stable Isotope Analysis A total of 32 bulk and 58 serial enamel samp les from 32 individuals were analyzed from the Gray Fossil Site in eastern T ennessee (Table 3-1, 3-5). Samples were primarily collected from late erupting teeth (i.e. third molars or f ourth premolars) in orde r to avoid sampling teeth that mineralized while nursing or weaning. Bulk samples were collected from one area parallel to the growth axis of the tooth. Serial sample s were acquired by sampling the teeth with parallel grooves at intervals of 2-3 mm perp endicular to the growth axis of the tooth. When sampling the gomphothere tusk, a total of 14 samples, each 1.5 mm wide in the grow th direction and 15 mm across the enamel band, were taken every 5 mm in order to acquire data repr esentative of at least one year of growth (Fox and Fisher, 2004). Between 2-10 mg of tooth enamel sample was collected using a low speed FOREDOMTM dental drill and carbide de ntal burrs. Samples were chemically pretreated prior to isotopic analysis with H2O2 to remove organics and weak acetic acid (0.1N, CH3CO2H) to remove secondary carbonates (Koch et al., 1997). Approximately 1 mg of treated sample was then analyzed using a VG Prism mass spectrometer in the Department of Geological Sciences at th e University of Florida.

PAGE 52

52 Stable isotope data are repor ted in conventional delta ( ) notation for carbon ( 13C) and oxygen ( 18O), where 13C (parts per mil, ) = ((Rsample/Rstandard)-1)*1000, and R = 13C/ 12C; and, 18O (parts per mil, ) = ((Rsample/Rstandard)-1)*1000, and R = 18O/ 16O. Analyzed samples were calibrated to NBS-19 and then to V-PD B (PeeDee Belemnite) following the Vienna (V-) convention (Coplen, 1994). Rare Earth Elemental Analysis Enamel and dentin from the gomphothere tus k, 1 rhino too th, and 3 tapir teeth (each from different individuals), were sampled for REE an alysis. Approximately 5-10 mg of sample was removed using a FOREDOMTM dental drill and carbide dent al burrs. These samples were cleaned in SavillexTM vials with 1 ml of 3 M HNO3, dissolved and heated overnight on a hot plate. After samples were dried until all liqui d was evaporated, the samples were weighted and dissolved in 2 ml of 5% HNO3 and left overnight on a hotplate. Approximately 3 ml of 5% HNO3 was added to each sample and the sample we ights were calculated in order to achieve a dilution factor of 2000. All samples were then run on a Thermo Finnigan ELEMENT2 inductively Coupled Plasma Mass Spectrometer in the University of Florida Department of Geological Sciences for bulk REE concentrations Internal standards and bone ash (NVS SRM 1400) were run with the samples, enabling corrections to be made due to instrument drift. All REE concentrations were normalized to PAAS (Post-Archean Australian Shale; McLennan, 1989). The REEs analyzed, range from La (Z = 57) to Lu (Z = 71). We excluded europium (Eu) from the analysis due to anomalous Eu enrichment and depletion spikes found in the Gray specimens. These anomalies are likely due to Eu partitioning under closed conditions and irrelevant for our comparisons (Trueman et al., 2004). These methods follow Trueman et al. (2004) and MacFadden et al. (2007).

PAGE 53

53 Results and Discussion Bulk Carbon Isotope Analysis The bulk carbon isotopic analyses of indicates a diet consisting entirely of C3 plants ranging in 13C values from 14.1 to 10.9 with a mean of 13.0 (1 = 0.9; Figure 3-2, Table 3-2 and 3-3). The bulk 13C values of sp., a morphologically presumed grazer (MacFadden, 1998), range from 13.6 to 13.0 with a mean of 13.3 (1 = 0.3; Figure 3-2, Table 3-2 and 3-3). Peccary (Tayassuidae) bulk 13C values range from 14.0 to 12.4, with a mean of 13.1 (1 = 0.8; Figure 3-2, Tabl e 3-2 and 3-3). These taxa are therefore interpreted to be obligate browsers, consuming only C3 vegetation. While only one tooth was available for isot opic analysis of the camel, cf. sp., the 13C value of 13.8 is consistent with a predominantly C3 diet (Figure 3-2, Table 3-2 and 3-3). The carbon isotopic niches of the peccary and the camel a ppear to overlap and are not statistically different from each other (ANOVA, = 0.81; Kruskal-Wallis, = 0.79; Figure 3-2), although it is difficult to comment on the degree of niche overlap due to the limited number of peccary, and camel teeth sampled. The bulk 13C value of the tusk from the gomphotheriid proboscidean is 0.3 (Figure 3-2, Table 3-2 and 3-3), significantly different from all other taxa sampled under parametric analyses of ANOVA ( < 0.0001) and all subsequent Fisher LSD multiple comparisons ( < 0.05). However, non-parametric KruskalWallis analysis yielded insignificant differences ( = 0.44), likely because the gomphothere tusk represents only one sample. This inconsistent 13C value prompted a comparison of enamel and dentin REEs from the proboscidean with those of the tapirs and rhinos. This analysis, discussed later, enabled me to determine if the tusk was subjected to a similar depositional environment as the more abundant forest-dwelling taxa (i.e. and ).

PAGE 54

54 As prior pollen analyses have identified the Gr ay Fossil Site to indicate a predominantly oak-hickory deciduous forest, we can constrain our interpreta tion of canopy density. Average 13C values of modern temperate deciduous flora of approximately -26 and -30 are consistent with an open forest canopy and dens e forest canopy, respectiv ely (Garten and Taylor, 1992; Tu et al., 2004; BASIN Ne twork, 2006). Accounting for bot h dietary (14.1 enrichment between large bodied ungulates and their die t; Cerling and Harris, 1999) and atmospheric enrichment (1.5; Cerling et al., 1997; Passey et al., 2002), all taxa with the exception of the proboscidean have mean 13C values consistent with a moderately dense, temperate forest ( 13C ~ -13). The bulk 13C values from the tapirs, rhinos, p eccaries, and the camel do not support the presence of C4 grasses at the Gray site. These data agree with the palynological evidence that documents low grass abundance (Wallace and Wang, 2004), of which the Gramineae pollen that is present may be from C3 grasses. However, the bulk 13C value from the gomphothere indicates the presence of C4 grasses at a distance within this species migration/home range and in large enough abundance to support a populati on of this herbi vore with a pure C4 diet, assuming that our gomphothere sample is representative of a population. Given that roughly contemporaneous sites below 37 latitude are thought to have under gone transitions to C4 grasslands during the late Mio cene/early Pliocene (Cerling et al., 1993, 1997; Wang et al., 1994), these data presented here suggest the presence of a forest environm ent at the Gray Fossil Site that may have served as a refugium to taxa requiring forest habitats among C4 grasslands. The relationship between rela tive tooth crown height ( hypsodonty index values) and 13C values are explored here. The average HI values of and peccaries are 0.7, 1.3, and 0.5, respectively. Because the sample sizes of rhinos and peccaries are small, these HI values should be viewed as preliminary. These HI values ar e not predictive of 13C values.

PAGE 55

55 Instead, the most hypsodont taxon has some of the most negative 13C values. While high-crowned teeth are no longer sy nonymous with the grazing of C4 grasses (MacFadden et al., 1999; Feranec, 2003, 2004; MacFadden, 2005), the Gray fauna likewise provides additional evidence that hypsodont t eeth do not indicate C4 grazing. Additionally, the browsing of C3 vegetation by the high-crowned a morphologically presumed grazer and isotopically classified mixed feeder/C4 grazer in Florida during the early Miocene (MacFadden, 1998), demonstrates further support of the absence of significant C4 flora at the Gray site. Bulk Oxygen Isotope Analysis Bulk 18O values of range from 5.2 to 2.3 with a mean of 4.0 (1 = 0.7; Figure 3-2, Table 3-2 and 3-3) not significantly different from peccaries, and the gomphothere (ANOVA, = 0.15; Kruskal-Wallis, = 0.21). The bulk 18O values of sp. range from 5.5 to 3.9 with a mean of 4.8 (1 = 0.7), this is the most negative mean 18O value of all ungulates sampled (Figure 3-2, Table 3-2 and 3-3). The bulk 18O values of the peccaries range from 4.9 to 4.1 with a mean of 4.4 (1 = 0.4; Figure 3-2, Table 3-2 and 3-3). Additionally, the bulk 18O value of the gomphothere tusk falls within the range of th e tapirs, rhinos, and peccaries at 4.2. The 18O of 1.7 for the camel (cf. sp.) is the most enriched in 18O (Figure 3-2, Table 3-2 and 3-3), significantly different from all other taxa sampled under parametric analyses of ANOVA ( < 0.01) and LSD multiple comparisons ( 0.01). No significant diffe rences are observed when using the non-parametric Kruskal-Wallis analysis ( = 0.13); however, this is likely a sample size issue because only one camel tooth is includ ed in the analysis. The highly enriched 18O value from the camel, a probable evaporation sensi tive taxon, may be compared to evaporation insensitive rhinos to quantify rela tive aridity (Levin et al., 2006). has the most depleted bulk 18O value of 5.5 and the camel has the most enriched 18O value of 1.7;

PAGE 56

56 therefore, the total bulk 18O range of all taxa is 3.8. Comparing the difference between the evaporation insensitive rhino and likely evap oration sensitive camel may further allow for estimates of relative aridity (Levin et al., 2006); however, additional samples of both the evaporation insensitive and sensitive taxa are first needed. Seasonal Reconstructions: Serial Sample Analysis The serial samples of and peccary teeth can reveal seasonal differences in monthly temper atures and/or precipitation. serial samp les ( = 12, 14) from two high-crowned teeth yield total 13C ranges of 0.6 ( 13.7 to 13.1) and 1.0 ( 13.7 to 12.7), with 18O ranges of 1.1 ( 6.1 to 5.0) and 1.3 ( 5.1 to 3.8; Figure 3-3B, Table 3-4, 3-5) Serial samples from two tapir individuals ( =8, 6), yield total 13C ranges of 1.1 ( 12.3 to 13.4) and 1.3 ( 10.6 to 11.9), with 18O ranges of 1.1 ( 3.9 to 2.8) and 0.8 ( 3.6 to 2.8; Figure 3-3A, Table 3-4, 3-5). One peccary tooth ( = 4) was sampled to determine if it demonstrates the same pattern of little 13C and 18O variation. The range of 13C and 18O variation was 0.6 ( 12.6 to 13.2) and 1.4 ( 5.6 to 4.2), respectively (Figure 3-3A, Table 3-4, 3-5). All serial samples of and the peccary demonstrate less than 1.5 variation in both carbon and oxygen isotopes, indicating the lack of signif icant seasonal variation. Serial samples of the gomphothere tusk ( = 14) were taken at in tervals representing over one year of growth (as per Fox and Fisher, 2004), varying 1 in both 13C and 18O values. The gomphothere 13C values ranged from 0.7 to 0.3 (Figure 3-3C, Table 3-4, 3-5), indicating the consumption of C4 grass throughout the course of a year with the absence of seasonal variations in diet. The lack of significant oxygen variati on (approximately 0.6, 4.6 to 4.0) likewise confirms the lack of significant seasonal changes in meteoric water due to temperature and/or precipitati on. The variation of tusk 13C seen here is similar to late Miocene

PAGE 57

57 gomphotheres with equable serial sample r ecords of approximately 1; however, those gomphotheres were C3 browsers and/or mixed feeders (Fox and Fisher, 2001, 2004; Figure 33C). The absence of considerable 13C and 18O variation is consistent with the isotopic patters found in the tooth enamel of th e ungulates sampled. However, th e lack of seasonal variation may be an artifact of the gomphotheres behavi or, if it actively migrat ed to areas where C4 grasses were abundant (i.e. migrating south during the winter to consume C4 grasses under similar temperature/precipitation conditions as summer grazed C4 grasses of northern latitudes). Additionally, we can not infer that the gomphothere only consumed C4 vegetation during its life including while at or near the Gray Fossil site. Instead, we ca n only state that the gomphothere sampled consumed a diet indicative of a pure C4 diet for at least one year of its life. The lack of significant variation in both the carbon and oxygen isotopes from all taxa sampled, suggests minor differences in monthly temperatures and/or pr ecipitation during the Neogene in eastern Tennessee. The Gray fauna experienced a more equable climate than today (Climate Zone, 2006; U.S. Department of Commerce and NOAA, 2006; Waterisotopes.org, 2006; Figure 3-4). Even though 18O variation is damped in mammalian tooth enamel due to the buffering of water sources and/or time av eraging (Passy and Cerling, 2002), the and peccary 18O ranges of variation are more simila r to fossil taxa from the aseasonal Gaillard Cut Local Fauna assemblage in Pana ma (MacFadden and Higgins, 2004; Figure 3-3A and 3-3B), than to taxa from highly variable climates. The serial samples of the rhinos from the 15-million-year-old site in Panama have 13C variation of 0.5 (MacFadden and Higgins, 2004), similar to 0.7 and 1.0 of from the Gray site (Figure 3-3B, Table 3-4). However, the 18O variation of 1.6 and 1.8 for from the Gaillard Cut L.F. (MacFadden and Higgins, 2004) appears greater than the range of 1.1

PAGE 58

58 and 1.2 in (Figure 3-3B, Table 3-4). Additio nally, annual variation in carbon and oxygen isotope values can yield di fferences as great as 4 in t ooth enamel of taxa that are present in seasonally variable c limates (as seen in bison, horses, and mammoths in Feranec and MacFadden, 2000); therefore, va riation of < 1.5 supports a rela tively aseasonal climate (with regard to precipitation and/or temperature). Despite the presen ce of a deciduous temperate flora that are typically present in highly seasonal envi ronments, the limited rang e of variation between serial samples likely represents a warmer and le ss seasonally variable climate than currently present in modern eastern North American temperat e forests. Therefore, the floral environment of the Gray site may resemble more equable broadleaf forests than those present in the Appalachians today. Evidence of a Forest Refuge REE analysis of the gomphothere tusk, and and teeth a llow for a comparison of patterns of REEs obtained post-mortem. Normalized REE patterns of enamel and dentin from the gomphothere tusk are nearly iden tical to, and closely para llel, those of sampled enamel and dentin from and despite differences in co ncentrations of REEs (Figure 3-5). Because similar REE patterns in dicate comparable depositional environments (Trueman, 1999), it is likely that the tusk was depos ited at the Gray site at a similar time to the rhinoceros and tapirs sampled for REEs, as opposed to being reworked. Therefore, the gomphotheres REE patterns indicate th at it dyed at or in close proximity to the Gray Fossil Site. The bulk and serial 13C values of the gomphothere tusk pr ovide conclusive evidence that C4 grasses were at least present within a distan ce no larger than the migration/home range of the individual sampled. While we are unable to determine the home range of gomphotheres or if they migrated, using modern proboscideans as analogues migration/home ranges may have been as large as 7000 km2 (Stuwe et al., 1998). Therefore, C4 grasses may have been present adjacent

PAGE 59

59 to the Gray site or at the fringes of the gomphotheres home range, potentially hundreds of kilometers away. As our gomphothere did consume a primarily C4 diet over the course of at least one year, it is likely that C4 grasses occurred in abundanc es large enough to support both itself and a population of go mphotheres with pure C4 diets, if our gomphotheres diet is representative of a population. While the gomphothere sampled has anomalous 13C values in relation to its browsing adapted dentition, gomphotheres have been described as isotopically inferred mixed C3/C4 feeders and C4 grazers, in addition to C3 browsers (MacFadden and Shockey, 1997; Cerling et al., 1999; Snchez et al., 2003, 2004; Todd et al., 2006). Because the bulk 13C values from the tapirs, rhinos, peccaries, and the camel do not support the presence of C4 grasses at the Gray site and palynological evidence indicates the low abundance of grass pollen (Wallace and Wang, 2004), it is improbable that C4 grasses were present at significant levels at the site. Instead, the Gray Fossil Site likely represents a forest environment that may have served as a refuge to taxa in the southern Appa lachians requiring C3 browse, present concurrently with C4 grasslands. The presence of forest dw elling taxa such as the tapirs and red panda ( ) at the Gray site demonstrates additional support that this forest environment may have provided uniq ue habitat to taxa that were more widespread previously. Conclusions The mammalian herbivores from the Gray Fo ssil Site provide evidence of moderately dense ancient forests, flanked by distal C4 grasslands. Both bulk and serial carbon isotopes indicate that all ungulate taxa sampled, with the exception of the gomphothere, consumed a diet of C3 vegetation. Because all of the 13C values of the peccaries, and the camel are less than 10, there is no isotopic evidence th at suggests the consumption of C4 grasses by these taxa. Therefore, the Gray site represents a forest environment large enough to support sustainable populations of its browsers (tapir rhino cf.

PAGE 60

60 camel cf. sp., peccary Tayassuidae). The majority (63%) of specimens sampled have depleted 13C values of 13, further suggesting that this dominant oak-hickory forest had a relatively dense canopy for a temperate forest. The presence of organisms that currently live in humid mesothermal (i.e. moderate moisture and heat with mean temperatures between 15-20 C, Wolf 1975 modified from de Candolle 1874) and/or megathermal areas such as tapirs and alligators provides evidence for a warmer and more equable climate than today. Se rial samples of tooth and tusk enamel document 13C and 18O variation of < 1.5, demonstrating neglig ible seasonal changes in temperature and/or precipitation. Assuming that the climate wa s warmer (as inferred from the taxa present) and that the relatively constant annual precipitation pa tterns seen today (Sankovski and Pridnia, 1995; Climate Zone, 2006; U.S. Department of Commerce and NOAA, 2006; Figure 3-4) occurred in the past, we would expect to see minor differences in 18O values. Additionally, relatively constant precipitati on and warmer mean annual temp eratures could explain why C3 and C4 floras do not experience season al water stress (as inferred from the lack of seasonally enriched 13C values), due to greater evaporation duri ng periods of increased temperature and/or aridity. Currently, the southern Appalachians ar e relatively warm and humid at low elevations, while precipitation shows little seasonality. Ther efore, the Gray site may have served as a refugium to taxa requiring C3 vegetation and more equable/warmer environments than may have been available at other geographical localities during the Miocene/Pliocene C4 grassland transition. Due to limited numbers of peccary, and camel specimens sampled, it is too early to speculate on the true degree of niche overlap present at this site. Continued excavation of the Gray Fossil Site will provide larger sample s from which future analyses of isotopic niche

PAGE 61

61 overlap can be clarified. Add itionally, microwear anal ysis of the ungulates sampled isotopically will provide further resolution to whether C3 consumers were feeding on C3 grasses and/or C3 browse. The continued sampling of the vertebrate fauna present at the Gray site will further clarify ecological niches and s easonal variation. In particular additional gomphothere samples will shed light on population level dietary va riation including estimates of percent C4 grass consumption. These mammalian herbivores from the Neogene of eastern North America provide unique opportunities to understand paleoecological phenomenon o ccurring during a period of dramatic global change.

PAGE 62

62 Table 3-1. Biota from the Gray Fossil Site Tennessee. Compiled from Wallace and Wang (2004) and Schubert and Wallace (2006). Fauna Osteichthyes Amphibia Anura Plethodontidae sp. Reptilia sp. sp. sp. Chelydridae sp. Viperidae Colubridae Aves Passeriformes Mammalia Soricidae Talpidae Lagomorpha Rodentia Xenarthra Gomphotheriidae cf. Tayassuidae cf. sp. Canidae Mustelidae cf. sp. sp. Flora Conifers Deciduous Shrubs Herbs -type Cyperaceae Gramineae Umbelliferae Caryophyllaceae

PAGE 63

63 Table 3-2. Bulk carbon and oxygen isotopes of mammalian ungulate enamel, Gray Fossil Site, Tennessee. Taxon Specimen No.* Tooth Position 13C () 18O () 291 LM2 -13.0 -3.7 586 partial RP4,M2,or M3 -13.9 -3.8 587 partial LP4,M2,or M4 -12.7 -3.5 588 Rm3 -13.7 -4.2 595 Lm3 -13.1 -2.8 602 Lp4 -12.1 -4.5 606 LM3 -13.4 -4.2 607 RM3 -13.3 -4.1 608 RM3 -11.2 -4.3 623 RM3 -14.1 -4.6 639 LM2 -12.7 -4.2 652 RM2 -13.1 -4.5 653 LM3 -14.0 -4.6 661 LM2 -13.9 -2.3 664 RM3 -11.5 -3.9 666 LM3 -14.0 -5.2 683 LM3 -13.7 -4.1 3424 RM3 -12.9 -3.8 3425 RM3 -13.4 -4.5 3426 LM2 -13.7 -3.0 3427 LM2 -13.6 -4.1 2/20/04-027 Lm3 -10.9 -3.4 2002-5-119 LM3 -12.4 -4.7 cf. 566 Rp3 -13.4 -3.9 cf. 609 Lm3 -13.2 -5.5 cf. 780 RM1 -13.0 -4.6 cf. 781 LM1 -13.6 -5.3 Tayassuidae 593 LM3 -14.0 -4.2 Tayassuidae 778 LM3 -12.4 -4.1 Tayassuidae 779 Rm3 -12.9 -4.9 cf. sp. 738 deciduous premolar -13.8 -1.7 Gomphotheriidae 305 tusk -0.3 -4.2 All specimen numbers are East Tennessee Museum of Natural History catalogue numbers with the exception of two uncatalogued specimens (2002-5-119, 2/20/04-27).

PAGE 64

64 Table 3-3. Stable carbon and oxygen values for th e taxa of the Gray Fossil Site, Tennessee. 13C () 18O () Taxon a Mean SD Range Mean SD Range 23 -13.0 0.9 -14.1 to -10.9 -4.0 0.7 -5.2 to -2.3 cf. 4 -13.3 0.3 -13.6 to -13.0 -4.8 0.7 -5.5 to -3.9 Tayassuidae 3 -13.1 0.8 -14.0 to -12.4 -4.4 0.4 -4.9 to -4.1 cf. sp. 1 -13.8 -1.7 Gomphotheriidae 1b -0.3 -4.2 a = the number of different individuals sampled, the descriptive statistics do not include more than one tooth from an individual (no teeth we re included that could have been missing teeth from an included individual). b Gomphothere tusk enamel was on ly available for isotope analysis.

PAGE 65

65 Table 3-4. Serial carbon and oxyge n isotope variation, per indivi dual at the Gray Fossil Site, Tennessee. 13C () 18O () Taxon ETMNH # Tooth N Min. Max. Range Min. Max. Range cf. 609 Lm3 12 -13.7 -13.0 0.7 -6.1 -5.0 1.1 cf. 781 LM1 14 -13.7 -12.7 1.0 -5.1 -3.8 1.2 595 Lm3 6 -11.9 -10.6 1.2 -4.0 -2.8 1.2 3424 RM3 8 -13.4 -12.3 1.2 -3.9 -2.8 1.1 Tayassuidae 779 Rm3 4 -13.2 -12.6 0.6 -5.6 -4.2 1.4 Gomphotheriidae 305 tusk 14 -0.7 0.2 1.0 -4.6 -4.0 0.6

PAGE 66

66 Table 3-5. Carbon and oxygen isotope values for serial samples of ungulate enamel, Gray Fossil Site, Tennessee. Sample 13C () 18O () cf. ETMNH 781 LM1 LGD-ETR1-S1 -13.3 -4.9 LGD-ETR1-S2 -13.4 -5.0 LGD-ETR1-S3 -13.7 -5.1 LGD-ETR1-S4 -13.4 -4.7 LGD-ETR1-S5 -13.2 -4.9 LGD-ETR1-S6 -12.8 -4.5 LGD-ETR1-S7 -13.1 -4.7 LGD-ETR1-S8 -13.2 -4.8 LGD-ETR1-S9 -13.3 -4.6 LGD-ETR1-S10 -13.2 -4.7 LGD-ETR1-S11 -13.3 -4.4 LGD-ETR1-S12 -13.4 -4.3 LGD-ETR1-S13 -13.2 -4.5 LGD-ETR1-S14 -12.7 -3.8 cf. ETMNH 609 Lm3 LGD-ETR2-S1 -13.2 -5.7 LGD-ETR2-S2 -13.3 -5.1 LGD-ETR2-S3 -13.0 -5.0 LGD-ETR2-S4 -13.2 -5.2 LGD-ETR2-S5 -13.1 -5.0 LGD-ETR2-S6 -13.1 -5.3 LGD-ETR2-S7 -13.2 -5.3 LGD-ETR2-S8 -13.5 -5.7 LGD-ETR2-S9 -13.7 -6.1 LGD-ETR2-S10 -13.3 -5.9 LGD-ETR2-S11 -13.5 -5.9 LGD-ETR2-S12 -13.5 -5.8 ETMNH 3424 Lm3 LGD-ETT5-S1 -10.9 -4.0 LGD-ETT5-S2 -10.6 -3.6 LGD-ETT5-S3 -11.0 -2.8 LGD-ETT5-S4 -11.8 -3.1 LGD-ETT5-S5 -11.9 -3.1 LGD-ETT5-S6 -11.7 -2.8

PAGE 67

67 Table 3-5 Continued. ETMNH 608 RM3 LGD-ETT15-S1 -12.6 -3.0 LGD-ETT15-S2 -12.9 -3.0 LGD-ETT15-S3 -13.4 -3.7 LGD-ETT15-S4 -13.4 -3.9 LGD-ETT15-S5 -13.3 -3.3 LGD-ETT15-S6 -12.9 -2.9 LGD-ETT15-S7 -12.7 -2.9 LGD-ETT15-S8 -12.3 -2.8 Tayassuidae ETMNH 779 Rm3 LGD-ETP1-S1 -12.6 -4.2 LGD-ETP1-S2 -13.2 -4.7 LGD-ETP1-S3 -12.8 -5.0 LGD-ETP1-S4 -12.8 -5.6 Gomphotheriidae ETMNH 305 tusk LGD-ETG1-S1 0.2 -4.2 LGD-ETG1-S2 -0.1 -4.2 LGD-ETG1-S3 -0.1 -4.1 LGD-ETG1-S4 -0.1 -4.2 LGD-ETG1-S5 -0.2 -4.0 LGD-ETG1-S6 -0.5 -4.1 LGD-ETG1-S7 -0.4 -4.2 LGD-ETG1-S8 -0.5 -4.1 LGD-ETG1-S9 -0.4 -4.2 LGD-ETG1-S10 -0.4 -4.4 LGD-ETG1-S11 -0.6 -4.4 LGD-ETG1-S12 -0.3 -4.5 LGD-ETG1-S13 -0.7 -4.4 LGD-ETG1-S14 -0.6 -4.6

PAGE 68

68 Figure 3-1. Location of the Gray Fossil Site, Tennessee, USA.

PAGE 69

69 Figure 3-2. Stable carbon and oxygen isotope values from the ungulate taxa at the Gray Fossil Site. Symbols represent the mean value, wh ile the error bars co rrespond to 1 standard deviation.

PAGE 70

70 Figure 3-3. Serial samples of the ungulate taxa from the Gray Fossil Site, including: A) ( ,), peccary (*), B) ( ), and C) the gomphothere ( ). ( ) from the 15 Ma Gaillard Cut Local Fa una in Panama (MacFadden and Higgins, 2004) and late Miocene Gomphotheres ( ) from the Port of Entry Pit in Oklahoma, USA (Fox and Fisher, 2001), are included for comparisons in part B and C, respectively. Dashed lines represent oxygen values and solid lines indicate carbon values, while like colors correspond to the same individual per taxon.

PAGE 71

71 Figure 3-4. Average monthly temperate (C), monthly oxygen isotope values, and monthly precipitation (rainfall, snowfall) in Johnson City/Bristol Tri-City Area, Tennessee. Oxygen isotope data are from Waterisotope s.org, temperature data are a 10-yr mean taken from station 401094/13877 and provided by the U.S. Department of Commerce and NOAA, www.ncdc.noaa.gov/oa/climate/climatedata.html. Remaining precipitation data (rainfall, snowfall) are from Climat e Zone www.climate-zone.com.

PAGE 72

72 Figure 3-5. Normalized REE(PAAS) concentrations of the (asteris k) gomphothere tusk, (triangles) and (square) teeth. Dentine and enamel REE concentrations are noted with dotted lines and solid black lines respectively. Because gomphothere tusk enamel and dentin have higher absolute concentrations of REEs, they are plotted on the secondary y-axis (right).

PAGE 73

73 CHAPTER 4 IDENTIFYING FOREST ENVIRONMENTS IN DEEP TIME USING FOSSIL TAPIRS: EVIDENCE FROM EVOLUTIONARY MORPHOL OGY AND STABLE ISOTOPES Introduction Modern tapirs occupy densely-canopied fo rests throughout southern Mexico, Central America, South America, and southeast Asia (Salas, 1996; Brooks et al., 1997; Foerster and Vaughan, 2002; Tobler, 2002; Holden, 2003). Li ving tapirs are browsers and possess morphological features that are present in diverse clades of browsing mammals (e.g. lowcrowned teeth, short m andibular diastema; MacFadden and Shockey, 1997; Mendoza et al., 2002). The masticatory morphology (morphological characters associated with the oral processing of food) of tapirs is interpreted to be highly conservative, retaining plesiomorphic characters through time (Colbert and Schoch, 1998). In addition to the conservative nature of morphological characters of browsers, tapiroids al so appear to have browsed through time, as inferred from the stable carbon isotopes of th eir tooth enamel (MacFadden and Cerling, 1996; MacFadden and Shockey, 1997; Koch et al., 19 98; Kohn et al., 2005; Feranec and MacFadden, 2006). Thus, if tapiroids are morphologically conservative and maintain stable carbon isotope values consistent with browsing, then their distributions can be used to identify forest environments in Deep Time. The four extant tapirs, Bairds tapir ( ), the lowland tapir ( ), the mountain tapir ( ), and the Malayan tapir ( ) occupy forest environments, including: lowla nd forests, primary and secondary forests, Amazonian floodplains, and montane cloud forests (B rooks et al., 1997). Tobler (2002) also noted that tracks of were almost exclusively found in areas with dense vegetation and were rare in more open habitats. While and inhabit forest environments (Salas, 1996; Brooks et al., 1997; Foerster and Va ughan, 2002; Tobler, 2002;

PAGE 74

74 Holden, 2003), even the most eco logically divergent species, (capable of occupying treeless paramo environments of greater than 3800 m elevation) lives in dense Andean forests more frequently than any other habitat type (Downer, 2001). Additionally, the Andean forests are necessary habitats for as their canopies offer prot ection from predators and icy storms (Downer, 1996, 2001). Since living tapirs are consistently found either within, or in close proximity to, dense closed-canopy forests, th ey are model organisms for inferring forest environments of the past. The diets of living tapirs ge nerally consist of leaves, twig s, fiber, and fruit (Bodmer, 1991; Henry et al., 2000; Downer, 2001; Galetti et al., 2001; Tobler, 2002; Lizcano and Cavelier, 2004). Despite seasonal flooding, there is little variation in the diet of in Perus Amazon floodplain (Bodmer, 1990). In contrast, in French Guiana consume fewer fruits and greater fiber seasonally, after peak frui ting (Henry et al., 2000). Even with potential seasonal variability in diet corresponding with fruiting events, living tapirs predominantly browse and forage for food throughout the year in forest environments (Bodmer, 1990; Salas, 1996; Tobler, 2002; Henry et al., 200 0; Foerster and Vaughan, 2002). Stable carbon isotope signatures of mammalian tooth enamel can be used to reconstruct the diet of extinct herbivores, in cluding tapirs (DeNiro and Ep stein, 1978; Quade et al., 1992; Cerling et al., 1997; MacFadden et al., 1999). Since the stable carbon isotope signatures of mammalian tooth enamel reflects the diet co nsumed (with a dietar y enrichment rate, *, of approximately 14.1 for medium to large bodied mammalian herbivores; Cerling and Harris, 1999), the diet of fossil tapirs can likewise be reconstructed through time. All previous studies of extinct mammals have demonstrated that with in a given herbivore fauna, tapirs consistently have among the most negative 13C values (along with camelids; MacFadden and Cerling, 1996;

PAGE 75

75 Koch et al., 1998; Kohn et al., 2005; Feran ec and MacFadden, 2006). Based on modern analogues (e.g. Cerling et al., 2004), these values are interpreted to represent ancient forest habitats. In addition to their known ecological preference for forest s, tapirs are model organisms for comparisons through Deep Time due to low intra-population variation of enamel stable carbon isotopic values (Grawe DeSantis, 2005). Given that a recen t adult population of from Acapulco, Mexico was isotopically homoge neous in diet, with 13C variation of only 3, the carbon isotope values of adult fossil ta pir specimens are likely to reflect those of a fossil population (Grawe DeSantis, 2005). T hus, comparisons of carbon isotope values of fossil tapirs through time will elucidate their diet ary niche and proximity to forest environments. Throughout the Cenozoic, North Americ a underwent dramatic environmental transformations, as evident from anomalously wa rm tropical forests in high latitude North America during the Paleocene-Eocene Thermal Maximum (PETM) (Wing et al., 2005), dramatic declines in browsing taxa throughout the middle to late Miocene (Janis et al., 2000), and the expansion of C4 grasslands during the late Miocene/earl y Pliocene (Wang et al., 1994; Cerling et al., 1997; Koch et al., 2004; Retallack, 2001; St romberg, 2005). Because the expansion and contraction of forest environments are likely to affect the distributions of resident taxa, the presence of obligate forest dwellers can help reconstruct the distributi ons of their corresponding forest habitats in the past. Colbert and Schoch (1998) suggested that tapiroids have always resided in humid, mesothermal areas and that de clining clade diversity during the Oligocene and Miocene reflects the contraction of these ar eas in comparison to Eocene distributions. The primary objective of this study is to recons truct the distribution of forest environments through time using fossil tapiroids as indicator taxa. While plant macrofossils and pollen are usually analyzed to understand ancient forest distributions, mammalian herbivores potentially

PAGE 76

76 provide another line of evidence t ypically not available to paleobotan ists. In order to use tapirs this way, I will first document their conservative morphology through time and compare them to the closely related horses (Equida e), a family with a considerab ly different evolutionary and adaptive history. Secondly, I will compare stable isotope values of tapirid tooth enamel through time to confirm their occupation as forest dwel ling browsers. Lastly, I will use the Paleobiology Database (2006) to map tapiroid distributions through time, ther efore, reconstructing forest distributions in Deep Time. Materials and Methods Morphology Selected dental measurements were taken to quantify the evolution of crown height and body size proxies in relevant tapir specim ens, ranging in age from early Eocene (Wasatchian, North American Land Mammal Age) to Recent. These measurements include: greatest anteriorposterior length of M1, m1, and m3; greatest transverse width of M1, m1, and m3; and greatest enamel crown height of M1, m1, and m3 (Mx, upper molar position; mx, lower molar position; Px, upper premolar position; px, lower premolar position). Relative ontogeny (based on tooth wear) was coded for each specimen as unworn, little wear, intermediate w ear, and heavily worn. Individuals in the latter categor y were not used to calculate hypsodonty index (ratio of molar length to crown height). Other measured characters that were taken, as available on individua l specimens, include: greatest length of i1; greatest length of i3; lengt h of mandibular diastema (i.e. from posterior to the canine to the anterior point of the p2); lo wer premolar row length (PRL); and lower molar row length (MRL). Following Solounias and Moelleken (1993) MacFadden and Shockey (1997), and Mendoza et al. (2002), several cran ial characters were coded to de scribe morphological evolution

PAGE 77

77 related to browsing and grazing adaptations: gle noid fossa height above occlusal plane (low, high), paracondylar process length (short, long ), anterior zygomatic arch (poorly or well developed), position of anterior most part of the orbit (dorsal to P3, P4, M1, M2, M3, or posterior to M3); shape of in cisor arcade (curved, straight), and masseteric process above M1 (absent, present). Specimens measured were from: the Amer ican Museum of Natural History (AMNH) Vertebrate Paleontology and Mammalogy collec tions in New York, New York, USA; the University of Florida/Florida Museum of Natu ral History (UF) Verteb rate Paleontology and Mammalogy collections in Gainesville, Florid a, USA; and, the Yale Peabody Museum (YPM) Vertebrate Paleontology and Vertebrate Zoology collections in New Haven, Connecticut, USA. Data from the Paleobiology Database (2006) were also included, when available. Stable Isotopes Stable isotopic evidence was com piled from all available publications dealing with fossil tapirs from North America (M acFadden and Cerling, 1996; Koch et al., 1998; Kohn et al., 2005; Feranec and MacFadden, 2006). To this I have added previously unpublished data produced in our laboratory for specimens from the late Mioc ene McGehee and early Pliocene Palmetto Fauna localities in Florida and extant tapirs from Acapulco, Mexico. All carbon isotope data are reporte d in the standard convention: 13C (per mil, ) = [Rsample-Rstandard] ] x 1,000, where R is the ratio of 13C/12C and the standard is VPDB (Pee Dee Belemnite, Vienna Convention; Coplen, 1994). A ll newly analyzed data were prepared using standard pre-treatment techniques (e.g. Koch et al., 1997; MacFadden and Higgins, 2004) and then analyzed on a VG Prism stable isotope ra tio mass spectrometer with an in-line ISOCARB

PAGE 78

78 automatic sampler in the Department of Geological Sciences at the University of Florida. The analytical precision based on replicate analyses is + 0.1 Inferred Forest Distributions The locations and therefore known geographic di stributions were plotted for all tapiroid taxa (Tapiroidea; classifica tions based on McKenna and Bell, 1997 and Colbert, 2005) compiled in the Paleobiology Database (2006). These data were used to produce the range maps for the Eocene through Recent. Present tapir distribution s as determined from the International Union for Conservation of Nature and Natural Re sources (IUCN) and the Species Survival Commission-Tapir Specialist Groups (SSC), Tap ir Status Survey and Conservation Action Plan (Brooks et al. 1997), were added to the Present map. Results and Discussion Morphology Radinsky The superfamily Tapiroidea has a fossil record extending back into the early Eocene, ~55 million years ago (McKenna and Bell, 1997; Colbert, 2005; Paleobiology Database, 2006; Figure 4-1). Here we present both qualitative and quantitative morphological re sults demonstrating the conservative nature of fossil tapi rs as compared to a more rapidly evolving clade within the Perissodactyla, i.e. the classic example of fo ssil horses (family Equidae; e,g, Simpson, 1953; MacFadden, 1992). As I assert above, the bradytelic (i.e. very slow, Simpson, 1953) evolution in Cenozoic tapirs allows ecol ogical interpretations relative to modern far back into the fossil record. 1 Paraphrased comment, L. Radinsky to B. MacFadden, mid 1970s

PAGE 79

79 Several previous studies have demonstr ated that there are suites of qualitative morphological characters of the cr anium and mandible in mammalian herbivores that represent adaptations for either browsing or grazing (S olounias and Moelleken, 1993; MacFadden and Shockey, 1997; Mendoza et al., 2002; Table 4-1). Based on the distribution of these characters in the fossil record, it is known that browsing adaptations are pr imitive (plesiomorphic), whereas grazing characters are derived (apom orphic). These morphologies are illustrated (Figure 4-2) in four representative examples, the Oligocene tapiroid and modern as compared to the Oligocene equid and modern As can be seen in both table 4-1 and Figure 4-2, primitive morphologies of the cranium and mandible are demonstrated in and whereas derived morphologi es are demonstrated in This indicates that the adaptive morphology of has evolved little since the early Cenozoic when both the Tapiroidea and Equidae had browsing adaptations (Table 4-1, Figure 42). While the Equidae subsequently underwent explosive evolution in cranial and mandibular morphology, particularly since the Miocen e (Simpson, 1953; Radinsky, 1984; MacFadden, 1992), tapirs demonstrate bradytel ic (slow) evolution. The comp arative evolution and analysis of multiple characters that are associated w ith dietary strategies strengthens our dietary interpretations through time. Most of the evolution described above results from majo r morphological changes to the masticatory complex, including the great expansi on of cheek tooth crown heights observed in advanced horses (Simpson, 1953; Radinsky, 1984; MacFadden, 1992). Given the relatively common occurrence of fossil horse teeth, as opposed to the rarely preserved cranial and mandibular morphology, equid dentitions are fr equently cited as prime examples of macroevolution. One such dental character, hyps odonty index (HI), is an informative means of

PAGE 80

80 comparing the evolutionary morphology in clades of mammalian herbivores. Previous studies have shown that extinct mammal species with HIs < 1 are primarily browsers and those with HIs > 1, although classically interp reted as primarily grazers (e .g. Simpson, 1953; MacFadden, 1992), actually have the evolutiona ry capacity to be either brow sers or grazers (MacFadden et al., 1999; Feranec, 2003; MacFadden, 2005). Hors es were primitively short-crowned, with HIs all < 1 until about 20 million years ago. Ther eafter, several clades of horses underwent explosive, rapid evolutio n of crown heights, resulting in HI s ranging from > 1 to 3 (Figure 4-3; although one equid clade, the browsing anchithe res, retained the primitive morphology). The explanation for this rapid increase in HIs duri ng the Miocene is initially to exploit a new food resource, grasses, which were spreading over many continental landscapes. This Great Transformation (Simpson, 1953; Stromberg, 2005) fundamentally affected both the morphological and ecological evolution of the Equidae. In cont rast to the Equidae, a very different pattern is seen in the evolution of HIs in Tapiroidea (F igure 4-3). All tapiroid taxa measured for this study ranging in age from ~53 Ma to the present have HIs < 0.7. Thus, while horses were rapidly evolving in response to the changing environments, the bradytelic (slowly evolving) tapirs are characterized by stasis in crown heights. Body size is a fundamentally important char acter in understanding ecological adaptations of individual species (e.g. Eise nberg, 1981). Although body size is diffi cult to estimate in extinct species, molar dimensions can serve as a pr oxy for relative body size (e.g. MacFadden, 1986; Damuth and MacFadden, 1990). In this paper, I use m1 length as a relative indicator of tapir body size. With the exception of the three modern species and and the fossil taxon that have declined in m1 length since the late Miocene, the m1 length within ot her species within the Tapiroid ea has increased in size by about

PAGE 81

81 2.5 times in approximately 50 million years (Figure 4-4). This increase in m1 length, and corresponding inferred body size increas e, appears to be linear and re latively gradual, interpreted to represent relative stas is, both in morphology and diet. This is in contrast to the explosive pulse of evolution see in fossil horses duri ng the Neogene after about 20 million years ago (MacFadden, 1986). As will be seen below, the conservative morphology and browsing diet demonstrated here for fossil tapirs are also corroborated by evidence from stable isotopes. Stable Carbon Isotopes As opposed to the traditional method of inte rpreting extinct mammalian herbivores as grazers or browsers using morphol ogical characters, stable carbon is otope ratios can be used to interpret ancient diets (Cerling et al., 1993, 1997; Wang et al., 1994; MacFadden and Shockey, 1997; MacF adden et al., 1999; Zazzo et al., 2000). Stable carbon isotope analysis of fossil tooth enamel provides dietary information about the re spective taxon, because carbon is incorporated into the lattice of enamel hydroxylap atite retaining an isotopic signal that is reflective of plants consumed (Krueger, 1991; Lee-Thorp and van de r Merwe, 1991; Cerling et al., 1997; Cerling and Harris, 1999). Because 13C /12C ratios in plants vary depending on plant photosynthetic pathways and stable carbon isotopes do not decay with time (Ehleringer and Monson, 1993) in the absence of diagenesis (postmortem chemical alteration), we can look at 13C and 12C ratios of the past similarly to ratios of 13C and 12C today (Cerling et al., 1997). The 13C values of medium to large-bodied ungulates are en riched by approximately 14.1 (although nonruminants, such as tapirs, may be enriched by 12-13), as compared to plants consumed (Cerling and Harris, 1999). Therefore, enamel 13C values -9 (possibly as enriched as -7 due to modern atmospheric CO2 enrichment; Cerling et al., 1 997; Passey et al., 2002) reflect a pure C3 diet and 13C values -2 indicate a predominantly C4 diet (MacFadden et al., 1996; Cerling et al., 1997, 1999, 2004; Cerling and Harris, 1999).

PAGE 82

82 Stable carbon isotopes of extinct/extant taxa can also be used to r econstruct forest canopy density due to greater 13C discrimination occurring in dense closed canopies as compared to less dense/open C3 environments (van der Merwe and Medi na, 1989; Cerling et al., 2004). Since 13C values increase with decreasing canopy density and/or increasing distance from dense forest edges (van der Merwe and Medina, 1989; Kapos et al., 1993; West et al ., 2001), more depleted 13C values of mammalian herbivores reflect the consumption of browse in denser canopied forests (van der Merwe and Medina, 1991; Cerlin g et al., 2004). By examining the stable isotopes of fossil tapirs and other extinct mammalian herbivores, we can document dietary changes through time. Stable carbon isotopes of fossil tapir enamel have been analyzed from several North American sites spanning the past ~10 million years (MacFadden and Cerling, 1996; Koch et al., 1998; Kohn et al., 2005; Feranec a nd MacFadden, 2006; Figure 4-5). The localities included in this analysis are, from oldest to youngest (Ma): Love Bone Bed (~9.5); McGehee and Mixons (~7.5); Withlacoochee (~7); Palmetto (~4.5); Ha ile 15A, Macasphalt, Port Charlotte, Punta Gorda and Santa Fe 1 (~2.5); Leisey 1A (~1.5); Harleyville (~0.4); Cutler, Hornsby Springs, Ichetucknee, Page-Ladson, Rock Springs, a nd Santa Fe (~0.1-0.01) (MacFadden and Cerling, 1996; Koch et al., 1998; Kohn et al., 2005; Fera nec and MacFadden, 2006; Figure 4-5). Fossil tapirs consistently demonstrate di ets composed of predominantly C3 vegetation, with enamel 13C values ranging from -10.1 to -14.3 (Figure 4-5). Additionally, fossil tapirs likely maintain their presence in denser canopied forests because their 13C values consistently are among the most 13C depleted isotopic values as comp ared to other co-occurring mammalian herbivores (Figure 4-5). For example, fossil horses from these sites demonstrate a pattern different from tapirs, instead consuming more is otopically enriched vege tation (Figure 4-5).

PAGE 83

83 Since their first appearance approximately 2.5-1.5 Ma during the Great American Biotic Interchange (Stehli and Webb, 1985), fossil and ex tant South American tapirs inhabit dense canopied environments, as inferred from stab le isotopes (MacFadden and Shockey, 1997; Grawe DeSantis, 2005). MacFadden and Shockey (1997) interpreted as a browser (based on morphological charac ters), along with peccaries ( sp.), deer ( sp.), and llamas ( ) at the Tarija Pleistocene (0.5 to 1.0 Ma) site in Bolivia. As is also seen in ancient North American ecosystems, the isotopic values of of 13.4 to -10.5 are among the most 13C depleted of all herbivores sampled within the Tarija fauna (total whole faunal range of -13.4 to -3.4; MacFadden and Shockey, 1997). Similarly, the modern tapir (from specimens collected be tween 1873-74 in Acapulco, Mexico) demonstrates mean 13C values of -14.6 (Grawe DeSa ntis, 2005), consistent with their classification as forest occupying browse rs. Because tapirs consume fruits that have isotopic values more enriched than corres ponding sub-canopy foliage (due to vertical stratification of 13C values; van der Merwe and Medina, 198 9; Cerling et al., 2004) their dietary 13C values are likely more enriched than the foliage in these environments. Therefore, tapirs likely reside in even more densely forested environments than indicated isotopically. In conjunction with their conserva tive morphology, the depleted 13C values justify tapirs as robust indicators of forest environments in Deep Time. Distributions From the Paleocene-Eocene Thermal Maxim u m to the Miocene-Pliocene expansion of C4 grasslands, North America underw ent great environmental change throughout the Cenozoic. Paleobotanical studies document dramatic vegetation shifts in mesothermal (i.e. moderate moisture and heat with mean temperatures between 15-20 C, Wolf, 1975 modified from de Candolle, 1874) broad-leaved evergreen forests, contracting from Eocen e distributions of 60

PAGE 84

84 latitude from the equator, to only 35 during the sharp cooling of the early Oligocene (Potts and Behrensmeyer, 1992; Wing, 1998). Megathermal (i.e humid and warm with mean temperatures 20 C, de Candolle, 1874) vegetation also became restricted to within 15 from the equator during the Oligocene, compared to early Eocene latitudinal distributions of 60-65 and modern ranges of 20-25 (Potts and Behrensmeyer, 1992). Corresponding to these shifts in vegetation, cooling occurred from the late Eocene to early Miocene, with periodic warming events that consistently declined in temper ature from preceding warming events (Wolfe, 1994). While these changes in forest distributions ar e inferred from paleobotan ical evidence including pollen, plant macrofossils, and subsequent C limate-Leaf Analysis Multivariate Program (CLAMP) analyses, tapirs can pr ovide an independent line of evidence for the presence of ancient forests in Deep Time. As Colbert and Schoch (1998) s uggested, the lack of significant tapiroid remains in the Oligocene and Miocene ma y be a result of declines in the mesothermal forests they likely inhabited. The Paleobiology Database (2006) is a valuab le tool for analyzing the distribution of ancient faunas both temporally and spatially. Using the Paleobiology Database (2006), I generated maps of tapiroid dist ributions in North and Central America from the Eocene to the present (Table 4-2, Figure 4-6). During the Eocene, tapiroid dist ributions are most wide-ranging, extending from central Mexico to Arctic Canada (Table 4-2, Figure 4-6). The widespread distribution of Tapiroidea dur ing the Eocene correlates with paleobotanical evidence that suggests extensive mesothermal and megatherma l vegetation ranges with boreotropical flora extending to at least 65 N in North America (Wolf, 1975; Wing and Sues, 1992; Wing, 1998). However, based on the presence of tapiroids at latitudes up to 78.5 N during the Eocene, boreotropical flora may have extended farther north then previously antic ipated. This is in

PAGE 85

85 agreement with the presence of semi-tropical fo ssil forests found at Axil Heiberg Island of ~80 N latitude (Christie and McMill an, 1991). Additionally, the dr amatic decline in tapiroid latitudinal ranges during the Oligocene (Table 4-2, Figure 4-6) is likely a result of sharp global cooling that decreased their mesothermal habitats (Colbert and Schoch, 1998; Wing, 1998). After the Oligocene, tapirid ranges expand th rough the Pleistocene (Table 4-2, Figure 4-6) concurrently with fluctuating warming and c ooling events (includi ng the Miocene Thermal Maximum at ~16 Ma; Wing, 1998). These tapirid range expansions occur from northern to southern regions with relatively consistent upper latitudinal range limits of between 40-46 N, from the Miocene to the Pleist ocene (Table 4-2, Figure 4-6). Subsequently, the upper range limits of contracted from 40 to 21 N during the last ~2 million years (Table 4-2, Figure 4-6). This recent contraction likel y resulted from the inability of to live in the seasonally cool and/or glacially inundated hi gher latitudes. Overall, tapiroid distributions correlate well with paleobotanically derived climatic inferenc es. Along with tapiroid morphology and isotopic data, these distribution maps provi de reconstructions of ancient forest environments with both temporal and spatial resolution. Conclusions Cenozoic tapirs have been under appreciated, mo stly because as Len Radinsky noted in the 1970s, the most interesting phases of their evol ution occurred during th e Eocene. Thereafter, despite significant global change in terrestrial ecosystems, tapirs are prime examples of Simpsons (1953) concept of bradytely, i.e. a gro up demonstrating slow, or arrested evolution. This mostly resulted from Tapiroids being well ad apted to their respective adaptive zone (niche complex in modern parlance) in ancient forests. Tapirs may not be a good group to investigate evolution in the fast lane, but they are mode l taxa for paleoecological reconstructions. The morphologic, isotopic, and biogeographic analyses presented above indicate that tapiroids are

PAGE 86

86 excellent indicators of ancient fo rest environments, and adds to our knowledge of these ancient habitat types based on other fossil evidence, e .g. as derived from paleobotany. As additional localities are discovered and analyzed in the future, we are bound to find more evidence of herbivore-plant interactions such as those exemplified by extinct forest-dwelling tapirs.

PAGE 87

87 Table 4-1. Comparison of craniodental features between the Oligocene tapiroid modern the Oligocene equid and modern Craniodental feature (Oligocene) (Modern) (Oligocene) (Modern) Hypsodonty Index < 1 < 1 < 1 > 1 Relative size of incisors i1 > i3 i1 > i3 i1 > i3 i1 i3 Shape of Incisor arcade curved curved curved straight Relative length of premolar tooth row PRL < MRL PRL < MRL PRL < MRL PRL > MRL Length of mandibular diastema short <2 x m1 length short <2 x m1 length short <2 x m1 length long >2 x m1 length Masseteric prominence above M1 absent absent absent present Position of the orbit above or anterior to M2 above or anterior to M2 above or anterior to M2 posterior to M3 Anterior extension of the zygomatic arch poorly developed poorly developed poorly developed well developed Height of the glenoid fossa above the occlusal plane low low low high The craniodental features were modified from those compiled by Mendoza et al. (2002). PRL=premolar row length, and MRL=molar row length.

PAGE 88

88 Table 4-2. Summary of Paleobiology Database ( 2006) results of tapiroid localities in North and Central America. Additionally, Holocene collections include presen t localities taken from the IUCN Tapir Status Survey and Conservation Action Plan (Brooks et al., 1997). Geological Epoch Age Range (million years ago) Latitudinal RangeTotal Latitudinal Range Eocene 54.8-33.7 21 N to 79 N 58 Oligocene 33.7-23.8 42 N to 50 N 8 Miocene 23.8-5.3 28 N to 46 N 18 Pliocene 5.3-1.8 20 N to 42 N 22 Pleistocene 1.8-0.01 9 N to 40 N 31 Holocene 0.01-present 9 N to 21 N 12

PAGE 89

89 Figure 4-1. Temporal distributions of the tapiroid genera analy zed in this study. The ranges are taken from the Paleobiology Database ( 2006). Following McKenna and Bell (1997) and Colbert (2005), Heptodon is included in the Tapiroidea.

PAGE 90

90 Figure 4-2. Fossil specimens of the (A) tapirid (left: YPM 11165; right: AMNH 661) from the middle Oligocene (Protoceras beds), South Dakota, and (B) equid (cranium, UF 201941, and UF191532, mandible) from the late Eocene/early Oligocene White River Group, western Nebraska. Modern (C) tapir ( ; UF 24112) and (D) horse ( ; UF225366) specimens are from the UF/FLMNH collection. Abbreviations ar e, as follows: dl=diastema length; gf=glenoid fossa; ia=incisor arcade; mrl=molar row length; mp=masseteric prominence; po=position of orbit; and, prl=premolar row length. All scale bars equal 5 cm.

PAGE 91

91 Figure 4-3. Hypsodonty index of two contrasting clades of perissodactyls, i.e. Equidae and Tapiroidea, throughout the past 55 million ye ars. Data for Equidae are taken from MacFadden (1992; Figure 11.6). Tapiroid values were measured from collections at the AMNH, YPM, and FLMNH. Mean values of the following taxa are included: (including ), (including ), (including ), and (including , ).

PAGE 92

92 Figure 4-4. North American fossil tapiroid and extant Central American and South American tapir m1 lengths through time as a proxy for body size evolution. The linear regression of all data analyzed, R2=0.73. If you exclude the three extant tapirs ( and ) and the fossil which demonstrate recent declines in body size since the late Miocene, R2=0.94. Data were measured from specimens at the AMNH, YPM, FLMNH, and compiled from the Paleobiology Database ( http://paleodb.org ). The following taxa were included: (including ), (including ), (including ), (including ), (including ), and (including , ).

PAGE 93

93 Figure 4-5. Carbon isotope data for tooth enamel of fossil tapirs and their associated fauna s for the past 10 million years in North America (compiled from MacFadden and Cerling, 1996; Koch et al., 1998; Kohn et al., 2005; Feranec and MacFadden, 2006). Additional new isotopic data sampled from the McGeh ee and the Palmetto Fauna loca lities in Florida are also included. Family abbreviations are, as follows (number of samples noted in parentheses): Ame=Amebelodontidae (N=12), Bov=Bovidae (N=3), Cam=Camelidae (N=51), Cer=Cervidae (N=9), Dro=Dromom erycidae (N=11), Ele=Elephantidae (N=11), Equ=Equidae (N=141), Gom=Go mphotheriidae (N=5), Mam=Mammutidae (N=14), Rhi=Rhinocerotidae (N=12), Tap=Tapiridae (N=33), and Tay=Tayassuidae (N=4). Each data point represents a single specimen. Additionally, all data points between 9-10 Ma are from the Love Bone Bed Local Fauna ~9.5 Ma; however, they are slightly offset to improve visual clarity.

PAGE 94

94 Figure 4-6. Maps of tapiroid localities on continental North America from the Eocene through the present. The maps were created from the Paleobiology Database ( http://paleodb.org). For the Eocene, gray circles ( ) indicate early Eocene (Wasatchian, Bridgerian; ~55-46 Ma) localities and black circles ( ) represent late Eoc ene (Uintan, Chadronian; ~46-34 Ma) sites. The Miocene is also sub-divided, with gray circles ( ) signifying early Miocene (late Arik areean, Hemingfordian, Barstovian; ~24-11 Ma) ages, black circles ( ) representing the late Miocene (Clarendoni an, Hemphillian; ~11-5 Ma), and black and white circles () indicating undifferentiated Miocene localities. Additionally, pres ent tapir distributions as summarized from the IUCN Tapir Status Survey and Conservation Action Plan (Brooks et al., 1997) were added to the present maps.

PAGE 95

95 CHAPTER 5 CONCLUSION Understanding the ecology of modern and ancient forests is critical to comprehending the context of mammalian evolution. Collectively, the ch apters presented in th is dissertation clarify the paleoecological dynam ics of forest environments a nd their mammalian inhabitants throughout the Cenozoic. Through a detailed ex amination of the stab le isotope ecology of modern tapirs, I clarified the pa leoecology and paleoclimatology of ancient tapir environments in the Appalachians and throughout the New World. Specifically, I document dietary variation in extant tapirs and further cl arify relationships between st able carbon and oxygen isotopes and climatic and geographic variables (Chapter 2). Next (Chapter 3), I elucidate the paleoecology and paleoclimatology of a potential Neogene Appal achian forest refuge. Subsequently (Chapter 4), I document the conservative dietary niches of ex tant tapirs and their an cestors and uses tapirs as model organisms to document forest distributions throughout the Cenozoic in the New World. In Chapter 2, stable carbon and oxygen isotopes of extant tapirs are quantified to test hypotheses regarding ontogenetic di etary shifts, stable isotope va riation at the population level, and relationships between stable isotopes and clim atic variables. A population of extant tapirs ( ) from Acapulco, Mexico demonstrates that 13C values of late erupting teeth are significantly greater than the earl y erupting first molar, indicati ng that juveniles are consuming 13C deplete milk and/or browsi ng in the denser canopy. The 18O values of late erupting teeth (P4, M3) are significantly different from each other, likely reflecting seasonal differences due to their chronological eruption. Dietary vari ation at the population level is low with 13C variation of ~2-3 for individual tooth positions. Additionally, extant tapir ( ) 18O values are constrained by climatic and geographic variables. Most notably, 18O values of decrease with decreasing precipitation frequency (mean number of

PAGE 96

96 precipitation days per month). is typically present in areas with greater precipitation than and 18O values are instead significantly correlated with 13C values. These data indicate that tapirs in wette r areas derive a larger proporti on of their water from leaves experiencing less evaporation in denser canopies, while is interpreted to increase water consumption by drinking when present in drier areas. An understanding of extant tapir stable isotope ecology improves and validates ecological interpretations of thes e elusive mammals both today and in the past. Additionally, these da ta help constrain the paleoecological and paleoclimatic hypotheses tested in Chapters 3 and 4. In Chapter 3, I clarified the paleoecology of mammalian herbivores and their floral environments during the Neogene in the poorly un derstood Appalachians, which is characterized by a paucity of relevant fossil lo calities during this time period. Global climate change and the expansion of C4 grasslands were interpreted to have ta ken place concurrently (Cerling et al., 1993, 1997; Wang et al., 1994); however, the Gray s ites flora and fauna suggest a forest refugium (Wallace and Wang, 2004). Therefore, stable isotope analyses of bulk and serial samples of fossil tooth enamel from all ungulates present at the Gray site were used to elucidate the ancient ecology of resident fauna. All of the ungulate taxa (tapir s, rhinos, camels, and peccaries), with the exception of the gomphothere, yield mean stable carbon isotope tooth enamel values suggestive of forest-dwelling browsers in a moderately dense forest. The lack of significant C4 plant consumption suggests the presence of forests large enough to independently support the continued browsing of su stainable populations of browse rs from the Gray site. In contrast, bulk and serial carbon isot opes from the gomphothere tusk support a diet consisting of C4 grasses, suggesti ng the presence of C4 grasslands within the individuals home range. The rare earth element (REE) analyses of the gomphothere tusk and the teeth of and

PAGE 97

97 indicate that these individuals shared similar depositional environments; thus, demonstrating the conc urrent presence of C3 forests and C4 grasslands. These data therefore support the interpretation of a North American fo rest refugium in the southern Appalachians during a time typified by more open environm ents. Additionally, stable carbon and oxygen serial sample variation of less than 1.5, sugge sts minor differences in seasonal temperature and/or precipitation. The high-crowned serial samples also had lower variation and overlapped the values of from a 15-million year old aseasonal forest in Panama (MacFadden and Higgins, 2004). As the southern Appalachians today are relatively warm and humid at lower elevations with aseasonal precipitation (Sankovski and Pridnia, 1995; Climate Zone, 2006; U.S. Department of Commerce and NOAA, 2006), th e Neogene Appalachians may have also served as a thermal refugium to ta xa requiring more equable/warmer environments. In Chapter 4, I first documented the conserva tive evolution of fo ssil tapirs and their subsequent diets. Tapir mas ticatory morphology is consistent with browsing dietary niches through time, despite dramatic global environm ental changes during the Cenozoic (Potts and Behrensmeyer, 1992). This is in contrast to closely related, fast-evolving equids (Simpson, 1953; Radinsky, 1984; MacFadden, 1992). Similarl y, stable carbon isotope analyses indicate that tapirs have consis tently had diets of C3 vegetation in denser canopied environments than most other sympatric mammalian herbivores. Thus, tapirs are robust indicators of ancient forest habitats and can be used to document how forest distributions have changed over time. After a relatively widespread distribution during the Eoce ne, declines in tapior oid latitudinal ranges during the Oligocene are consistent with paleobotanical evidence for contracting mesothermal and megathermal vegetation (Potts and Behrensmeyer, 1992; Colbert and Schoch, 1998; Wing, 1998). Tapirs subsequently increase their latitudinal ranges from the Miocene through the

PAGE 98

98 Pleistocene, and recently retreated to occupy only southern distributions below 21 N by ~10,000 years ago, with the most recent contraction likely a result of human activity. Maps of tapiroid (i.e. tapirs and their ancestors) fossil distributions in continental North America provide spatial and temporal proxy evidence for the pres ence of forest environments. The work discussed in this dissertation lays the groundwork for interpreting stable isotope data of extinct tapirs through time and understa nding how mammalian evolution may have been influenced by an Appalachian forest refugium du ring the Neogene and forest environments in the New World during the Cenozoic. Fu ture work can instead investigate the paleoecology of tapir dominated environments. Typicall y, when tapirs are found in the fossil record they are a minor component of the fauna. However, the Gray Fo ssil Site is a clear ex ception to this trend (Wallace and Wang, 2004). Since the discovery of the Gray Fossil Site, a second Neogene site with a high proportion of tapir fossils was discover ed in north central Fl orida (Hulbert et al., 2006). This highly fossiliferous Florida fossil si te, Haile 7G, is approximately 2 million years old and contains abundant tapirs and xenarthrans, whereas hors es and carnivores are rare (Hulbert et al., 2006). Thus, fu ture studies can examine how di etary niches are partitioned within a tapir dominated fauna at Haile 7G, enab ling comparisons to the Gray Fossil Site. Such comparisons may provide additiona l insight into the paleoecology of extinct tapirs and their environments; specifically, clarif ying what ecological conditions contribute to the dominance of fossil tapirs and the rarity of typically dominant horses. Lastly, all of the interpretations of forest environments discussed are based on the work of van der Merwe and Medina (1989, 1991) and others (e.g. Ka pos et al., 1993, West et al., 2001) who have documented the relationship be tween increased canopy density and decreased 13C values in tropical forests. However, little work has been done on documenting how 13C

PAGE 99

99 values vary with increased canopy density in te mperate forests. Understanding the strength of this relationship and the degree of isotopic variation in temperate forests at various latitudes is critical to proper interpretations of canopy dens ity in the fossil record. Thus, the continued integration of ecological data w ith paleontological data is required to properly interpret the paleoecology of past environments. The fossil record can provide valuable insigh t regarding how environments have changed over time. Modern ecological st udies are often limited by time, rarely exceeding one or more decades. Additionally, studies that are of a longer duration are often very costly and logistically difficult. Instead, the fossil reco rd allows long-term ecological que stions to be asked that cannot be asked by neontologists. Here, the fossil record helps clarify how forest environments have changed over time. However, our understanding of thes e forest environments is contingent on understanding extant tapir stable isotope ecology. Thus, the inte gration of the disparate fields of ecology and paleontology can syne rgistically improve understandings of past environments and ecological processes through time.

PAGE 100

100 APPENDIX A IMPORTANCE OF COMMUNICATING THE BROADER IMPACTS OF SCIENTIFIC RESEARCH TO THE PUBLIC ~Baba Dioum Dioums words, originally addressed to the In ternational Union for C onservation of Nature and Natural Resources, are a personal reminder of the synergistic need fo r innovative scientific research, quality education, and c onservation. As a scientist, e ducator, and conservationists, I aim to conduct innovative interdis ciplinary research, engage in inquiry-based science education, and communicate the conservation implications of my research to the public. Through the education of undergraduate and graduate students, I hope to not only inspire students to pursue careers in the sciences but to also become involved in educati on and outreach activities that will inspire future generations of scientists. However, as a graduate student in paleoecology, I have often felt as if I have been in the middle of a tug-of-war with educa tion on one side and science on the other. If I chose to engage in educational activities during my graduate caree r, my research would consequently receive less attention. Was this okay? Can scientists do both, conduct innovative res earch and communicate their science to society? I soon realized that not only scientists do both, we do both if we aim to improve scientific l iteracy. However, it is a consta nt challenge to figure out the meaning behind broader impact activities and determine how to achieve a necessary balance between research and public outreach. Currently, the National Science Foundation is attempting to ch ange the culture of ivory tower science, instead requiring scientists to de velop a broader impacts component to research grants. One way to excite student s about science while extending the reach of scientific grants is to develop educational lessons th at communicate actual hot off the press research (similar to

PAGE 101

101 that presented in Appendix B). Effective inquiry-based lessons should hook a students attention and engage them in the process of scientific di scovery. In addition, educational models are best disseminated through teacher read educational journals (e.g. the National Association of Biology Teachers and/or the National Associ ation of Science Teacher journals ), internet resources, and/or at local or national science teacher meetings. Idea lly, scientists should help present an aspect of educational modules in K-12 classrooms. This improves a scientists ability to communicate complex concepts and concisely explain the importa nce of their research. Furthermore, it allows students the opportunity to engage in open dialogue with a profe ssional scientist. By infusing young graduate students in the science, tec hnology, engineering, and mathematics (STEM) disciplines into K-12 classrooms, students begin to learn that scientists are represented by individuals from diverse ethnicities and genders. Participating in outreach that directly extends from current research has been a valued opportunity that I hope all gra duate students have the chance to experience, and hopefully something we will all continue with as young pr ofessionals. Currently, I aim to continue education and outreach activities associated with ongoing research. In addition, I hope to help future generations of scientific professionals va lue broader impact activitie s and gain experience communicating science to the public. By equipping future academics with the skills needed to communicate the broader impacts of scientific re search to the public, we will hopefully improve scientific literacy in future generations. Many scientists today have childhood stories that discuss their early passion for discov ery. Fostering this innate desire in students of all ages is the responsibility of scientists. For as Diom stat es, We will love only what we understand. We will understand only what we are taught.

PAGE 102

102 APPENDIX B STRAIGHT FROM THE MOUTHS OF HORSES AND TAPIRS: USING FOSSIL TEETH TO CLARIFY HOW ANCIENT ENVIRONMENTS HAVE CHANGED OVER TIME Introduction Do you or your students ever look out the window and imagine prehistoric animals wandering around millio ns of years ago? Or, when watching movies or shows about the ancient past, do you wonder how scientists know what the environment was like? Clarifying ancient environments millions of years ago is necessary to better understanding how ecosystems change over time, providing insight as to the potential impacts of curre nt global warming. However, understanding how scientists recons truct past environments is not always straight forward. The activity described here allows st udents to carry out the same res earch as professional scientists, develop hypotheses, collect and analyze data, and infer how Nort h American environments have changed over the last 55 million years. Using toot h measurements and dental microwear (i.e. the microscopic wear features that result from the processing of food) methods, students will develop science process skill s through the captivating discipline of paleontology. This module engages middle school students in th e scientific process, asking them to test the null hypothesis that horse and tapir diets have not ch anged over time using tooth measurements (Figure B-1). Based on their t ooth study students are then asked to make a new hypothesis regarding the diets of these animals, testing their second hypothesis with dental microwear data. Students utilize multiple learning styles during their paleontology research projects, ultimately making scientific illustrations based on their analysis of quantitative data. Determining Ancient Diets Using Tooth Morphology The size and shape of teeth allow paleontologists to interpret the diet of extinct animals based on modern analogues. For example, sharp s licing teeth in lions are used to interpret their dietary strategy as being carnivorous In contrast, flat blunt teeth as seen in domestic cows are

PAGE 103

103 used for grinding vegetation such as grasses. The proportions of teeth can further elucidate dietary categories by looking at the height of teeth as compared to their anterior to posterior length (i.e. hypsodonty index; Fi gure B-2; MacFadden, 2000). Modern grazers, such as horses and cows, have high crowned teeth (height > length) that are able to withstand abrasive vegetation including the glass-like silica in grasses. Unlike graze rs, living browsers (animals that consume leafy vegetation from trees and shrubs ), such as deer and tapirs, typically have low crowned teeth (height < length). With an unders tanding of how living animal diets relate to tooth shape, we can infer the diet of extinct animals. Cutting Their Teeth To begin the activity, provide students with a variety of specimens (e.g. skulls of cows, dogs, cats, horses, etc. which can be purchased from a biological supply company or borrowed from local museums or universities) or images of skulls (useful images can be found at www.d91.k12.id.us/skyline/teachers/robertsd/mammal 1.htm ). Ask your students to work in small groups and take ~10 minutes to examine the specimens and figure out what the animals ate. Once students have articulated how they made such inferences, discuss these dietary assignments collectively as a class. Through this opening whole-class discussion students collaboratively inquire as to the diets of the specimens provided, using the shape of teeth to determine if an animal is an herbivore (flat a nd blunt teeth) or a carni vore (sharp and/or pointy teeth). This discussion pr ovides a necessary lead in to the stud ents determination of the diets of the fossil animals listed in the data table (Figur e B-2) and testing the following hypothesis, The diets of horses and tapirs ha ve not changed over time. Collectively as a class, you can intro duce and discuss the hypsodonty index (HI) and have them figure out how they might use it to de termine if extinct animals ate grasses or leaves. Once students have determined their experimental design together as a class, they can begin

PAGE 104

104 measuring the heights and lengths of the fossil teeth using the Data Sheet provided (Figure B-3). Using the hypsodonty index (HI) ratio, students calculate HI values by simply measuring and then dividing the tooth height by the tooth length for the specimen s on their data sheet (Figure B2, B-3). Once all HI values have been determin ed, students graph these data points with the age of the fossil as the independent variable (millions of years, x-axis) and HI value as the dependant variable (unit-less ratio, y-axis; Figure B-2). Please note that wh ile Activity Sheet 1 (Figure B-2) gives explicit directions regarding data collect ion methods, these directions should only serve to refresh students on the experiment al design they previously came up with as a class. Next, students can (in small groups of 2) evaluate their null hypotheses based on data they collected and graphed. Specifically, students should note that HI values for horses increased approximately 15 million years ago while tapir HI values remained the same over time. This can easily be seen when comparing tapir HI data which forms a st raight horizontal line while the horse HI values increase dramatically at ~15 milli on years ago. Therefore, students can infer that HI values of horses changed over time, likely reflecti ng a shift in diet from leafy browse to grass. The students inferences should be discussed co llectively as a class, ensuring that everyone thoroughly understands the scientif ic process and how these data support the conclusions. Testing Dietary Interpretatio ns with Dental Microwear Now having a better understanding of the scie n tific process, data collection, and the graphing of data, students are prepared to embark on their next scientific study in small groups. Based on the results of their tooth morphology st udy, students develop thei r second investigation evaluating a new hypothesis with hy pothetical microwear data. For example, students can test the hypothesis that the diet of horses changed from that of a browser to a grazer using microscopic marks that result from the processing of food during the animals lifetime. Because these microscopic features have been studied in living animals w ith known diets, scientists have

PAGE 105

105 determined that animals with ~1.5 times more s cratches than pits in dicate a diet of grass while the reverse is indicative of a diet of leafy browse (MacFadden et al., 1999). By simply counting the pits (dots) and scratches (lin es) on the cartoon microwear slides provided (Figure B-4), student can compare the number of pits versus the number of scratches to determine how the diets of horses have changed over time. To do this, students should think about their prior study that utilized a ratio of tooth height to length. In groups, students collaborat ively inquire and determine the steps for their study and begin collecting data. In the end, students s hould plot the relative num ber of scratches vs. pits in the form of a ratio referred to as the microwear index (i.e. total number of scratches/total number of pits). If the animal has a microwear index value of greater than 1.5 (1.5 times more scratches than pits) they are infe rred to be grazing, while a value of less than 1.5 indicates a browser (MacFadden et al., 1999). Once all data are plotted, students can then test their hypothesis with data demonstrating that horses change from a mixed diet to one of primarily grasses. This evidence subsequent ly supports the idea that grasslands expanded approximately 20 million years ago in North Am erica. Microwear studies provide another line of evidence for looking at how diets of fossil animals have changed over time and can test earlier dietary interpretations as inferred from tooth morphology. Subsequently, as a class, ask students to interpret how environments have changed over time based on the diets of horses and tapirs. A dditionally, ask students to think about some of the problems with their data and what they could do to improve their study. For example, students may mention that by onl y analyzing the two microwear images provided they may not be capturing what the average horse is eating. Th is is because each slide is from a different time period, representing only one horse. If this point is not mentione d, students can be asked if they

PAGE 106

106 see any problem with inferring average human di ets based on only one person. Students should think about how they could improve their study by increasing the number of microwear images examined for each time period sampled; thus, comp aring average values over time (like they did with the tooth morphology study). Lastly, the last supper effect of microwear data (i.e. the idea that microwear data captures the last few meals of the animal s, as compared to a long-term dietary average) can bias these data. For exampl e, an older horse with worn down teeth may be eating a different diet th en it did when it was younger. This i ssue can be discussed as part of a general wrap-up discussion that focuses on iden tifying the strengths and weaknesses of the microwear study. Thus, students can discuss ways to improve scientific studies and recognize the benefits of using multiple tools to look at similar questions. Illustrating and Communicati ng Scientific Information After com pleting the research project described, your students should have an understanding of the scientific process, includ ing how to falsify hypotheses, how sample size plays a role in testing hypothese s, and techniques used to acqui re data (i.e. tooth morphology and microwear data). The class discussion mentioned above should be used to assess student understandings thus far. Using their knowledge of the scientific process, their acquired data, and the results of their data analysis students are now asked to devel op an artistic reconstruction of typical horse and tapir environments in North America at ~50 million years ago and also ~5 million years ago. Because both tapirs and horses were browsers ~50 million years ago (based on the first graph of HI values over time), the artistic reconstruction at ~50 million years ago should be of a forest environment. In contra st, the change in horse diets from browsing to grazing ~15 million years ago suggests a ~5 million year old environment as one that has both grasslands and forests. Upon completion of thes e illustrations, a brief wrap-up discussion should

PAGE 107

107 involve students in recapping how scientific hypo theses were developed, tested, and scientific information interpreted artistically. Resources Fossil Horses in Cyberspace, a web exhibit by the Florida Museum of Natural Hist ory where you can learn about paleontology and evolution through fossil horses Megalodon Educators Guide by the Florida Museum of Natural History contains numerous lessons centered around the larg est shark that ever lived, Megalodon Explorations Through Time by the University of California Museum of Paleontology Understanding Evolution for Teachers PBS Evolution Resources

PAGE 108

108 Figure B-1. A student actively measuring fossil horse teeth to test her null hypothesis.

PAGE 109

109 Figure B-2. Activity sheet one allows students to demonstrate how ancient diets can be determined using tooth morphology. All tooth images are from tapir and horse specimens from the Leisey Shell Pit 1A site in Florida, USA.

PAGE 110

110 Figure B-3. Data sheet accompanying part one of the activity, includes actual morphological data taken from DeSantis and MacFadden ( 2007). All tooth images are from tapir and horse specimens from the Leisey Shell Pit 1A site in Florida, USA.

PAGE 111

111 Figure B-4. Activity sheet two allows student s to demonstrate how ancient diets can be determined using dental microwear. Microwear slides are cartoon reconstructions of SEM microwear images from a fossil and modern horse analyzed by Hayek and others (1991) and cited by MacFadden (2000).

PAGE 112

112 LIST OF REFERENCES Baker, G., Jones, L.H.P., Wardrop, I.D., 1959. Cause of wear in sheeps teeth. Nature 4698, 1583-1584. BASIN Network, 2006. BASIN Databa se. Biosphere-Atmosphere Stable Isotope Network. 1 Septem ber 2006. < http://basinisotopes.org/basin/sites/list.html > Bocherens, H., Koch, P.L., Mariotti, A., Geraad s, D., Jaeger, J.J., 1996. Isotopic b iogeochemistry (13C, 18O) of mammalian enamel from African Pleistocene hominid sites. Palaios 11, 306318. Bodmer, R.E., 1990a. Fruit patch size and frugivory in the lowland tapir ( ). Journal of Zoology 222, 121-128. Bodmer, R.E., 1990b. Responses of ungulates to seasonal inundations in the Amazon floodplain. Journal of Tropical Ecology 1990, 191-201. Bodmer, R.E., 1991. Influence of digestive morphology on resource partitioning in Amazonian ungulates. Oecologia 85, 361-365. Brooks, D.M., Bodmer, R.E., Matola, S., 1997. Tapirs : Status Survey and Conservation Action Plan. IUCN/SSC Tapir Specialist Group, Gland. Bryant, J.D., Froelich, P.N., 1995. A model of oxygen isotope fractionation in body water of large mammals. Geochimica et Cosmochimica Acta 59, 4523-4537. Bryant, J.D., Luz, B., Froelich, P.N., 1994. Oxyge n isotopic composition of fossil horse tooth phosphate as a record of continental pale oclimate. Palaeogeography, Palaeoclimatology, Palaeoecology 107, 303-316. Bryant, J.D., Froelich, P.N., Showers, W.J., Ge nna, B.J., 1996a. A tale of two quarries: biologic and taphonomic signatures in the oxygen isotope composition of tooth enamel phosphate from modern and Miocene equids. Palaios 11, 397-408. Bryant, J.D., Froelich, P.N., Showers, W.J., Ge nna, B.J., 1996b. Biologic a nd climatic signals in the oxygen isotopic composition of Eocen e-Oligocene equid enamel phosphate. Palaeogeography, Palaeoclimat ology, Palaeoecology 176, 75-89. Bryant, J.D., Koch, P.L., Froelich, P.N., Show ers, W.J., Genna, B.J., 1996c. Oxygen isotope partitioning between phosphate and carbonate in mammalian apatite. Geochimica et Cosmochimica Acta 60, 5145-5148. Cerling, T.E., Harris, J.M., 1999. Carbon isotope fr actionation between diet and bioapatite in ungulate mammals and implications for ecologi cal and paleoecological studies. Oecologia 120, 347-363.

PAGE 113

113 Cerling, T.E., Wang, Y., Quade, J., 1993. Expansion of C4 ecosystems as an indicator of global ecological change in the late Miocene. Nature 361, 344-345. Cerling, T.E., Harris, J.M., MacFadden, B.J., Leakey, M.G., Quade, J., Eisenmann, V., Ehleringer, J.R., 1997. Global vegetation ch ange through the Miocene/Pliocene boundary. Nature 389, 153-158. Cerling, T.E., Harris, J.M., Leakey, M.G., 1999. Br owsing and grazing in el ephants: the isotope record of modern and fossil proboscideans. Oecologia 120, 364-374. Cerling, T.E., Hart, J.A., Hart, T.B. 2004. Stable isotope ecology in the It uri forest. Oecologia 138, 5-12. Christie, R.L., McMillan, N.J., 1991. Tertiary Foss il Forests of the geodetic hills, Axel Heiberg Island, Arctic Archipelago. Geological Survey of Canada, Bulletin 403, 1-227. Clementz, M.T., Koch, P.L., 2001. Differentiati ng aquatic mammal habita t and foraging ecology with stable isotopes in toot h enamel. Oecologia 129, 461-472. Climate Zone, 2006. Bristol Johnson City. Climate Zone (www.climate-zone.com). 28 August 2006. < http://www.climate-zone.com/climate/united-states/tennessee/bristol-johnsoncity/ > Codron, J., Codron, D., Lee-Thorp, J.A., Sponheimer M., Bond, W.J., Ruiter, D.D., Grant, R., 2005. Taxonomi c, anatomical, and spatio-tempor al variations in the stable carbon and nitrogen isotopic compositions of plants from an African savanna. Journal of Archaeological Science 32, 1757. Colbert, M.W., 2005. The facial skeleton of the Earl y Oligocene Colodon (Perissodactyla, Tapiroidea). Palaeontologia Electronica 8, 12A, 1-27. Colbert, M.W., Schoch, R.M., 1998. Tapiroidea and other moropomorphs. In: Janis, C.M., Scott, K.M., Jacobs, L.L. (Eds.), Evolution of Tertiary Mammals of North America 1. Cambridge Univ. Press, Cambridge, UK, pp. 569-582. Coplen, T.B., 1994. Reporting of stable hydroge n, carbon, and oxygen isotopic abundances. Pure and Applied Chemistry 66, 273-276. Crespi, E.J., Rissler, L.J., Browne, R.A., 2003. Testing Pl eistocene refugia theory: phylogeographical analysis of a high-elevation salamander in the southern Appalachians. Mo lecular Ecology 12, 969-984. Damuth, J., MacFadden, B.J., 1990. Body Size in Mammalian Paleobiology: Estimation and Biological Implications. Cambri dge Univ. Press, Cambridge. Dansgaard, W., 1964. Stable isotopes in precipitation. Tellus 16, 273-276.

PAGE 114

114 de Candolle, A.P.A., 1874. Constitution dans le rgne vgtal de groupes physiologiques applicables la geographie botanique ancienne et moderne. Archives des Science Physiques et Naturelles 50, 5-42. DeNiro, M.J., Epstein, S., 1978. Influence of diet on the distribution of carbon isotopes in animals. Geochimica et Cosmochimica Acta 42, 495-506. DeSantis, L.G., MacFadden, B., 2007. Identifying fo rested environments in Deep Time using fossil tapirs: evidence from evolutionary morphology and stable isotopes. Courier Forschungsinstitut Senckenberg 258, 147-157. DeSantis, L.R.G., Feranec, R.S., MacFadden, B. J., 2009. Effects of global warming on ancient mammalian communities and their environments. PLoS ONE 4, e5750. Downer, C.C., 1996. The mountain tapir, endanger ed flagship of the high Andes. Oryx 30, 4458. Downer, C.C., 2001. Observations on the diet and habitat of the mountain tapir ( ). Journal of Zoology 254, 279-291. Eberle, J.J., 2005. A new tapir from Ellesmer e Island, Arctic Canada Implications for northern high latitude paleobiogeography and tapir palaeobiology. Palaeogeography, Palaeoclimatology, Palaeoecology 227, 311-322. Ehleringer, J.R., Monson, R.K., 1993. Evolutiona ry and ecological aspe cts of photosynthetic pathway variation. Annual Review of Ecology and Systematics 24, 411-439. Ehleringer, J.R., Bowling, D.R., Flanagan, L.B., Fessenden, J., Helliker, B., Martinelli, L.A., Ometto, J.P., 2002. Stable isotopes and carbon cy cle processes in forests and grasslands. Plant Biology 4, 181-189. Eisenberg, J.F., 1981. The Mammalian Radiations: An Analysis of Trends in Evolution, Adaptation, and Behavior. Univ. of Chicago Press, Chicago. Eisenberg, J.F., 1989. Mammals of the Neotropi cs: The Northern Neotropics 1. Univ. of Chicago Press, Chicago. Feranec, R.S., 2003. Stable isotopes, hypsodonty, and the paleodiet of (Mammalia: Camelidae): a morphological speciali zation creating ecologi cal generalization. Paleobiology 29, 230-242. Feranec, R.S., 2004. Geographic variation in the diet of hypsodont he rbivores from the Rancholabrean of Florida. Palaeogeog raphy, Palaeoclimatology, Palaeoecology 207, 359369. Feranec, R.S., MacFadden, B.J., 2000. Evolution of the grazing niche in Pleistocene mammals from Florida: evidence from stable isot opes. Palaeogeography, Palaeoclimatology, Palaeoecology 162, 155-169.

PAGE 115

115 Feranec, R.S., MacFadden, B.J., 2006. Isotopic di scrimination of resource partitioning among ungulates in C3-dominated communities from the Miocene of California and Florida. Paleobiology 32, 191-205. Foerster, C.R., Vaughan, C., 2002. Home range, habitat use, and activity of Bairds tapir in Costa Rica. Biotropica 34, 423-437. Fox, D.L., Fisher, D.C., 2001. Stable isotope ecology of a late Mi ocene population of (Mammalia, Proboscidea) from Po rt of Entry Pit, Oklahoma, USA. Palaios 16, 279-293. Fox, D.L., Fisher, D.C., 2004. Dietary reconstruction of Miocene (Mammalia, Proboscidea) from the Great Plains region, USA, based on the carbon isotope composition of tusk and molar enamel. Palaeogeogr aphy, Palaeoclimatology, Palaeoecology 206, 311335. Fragoso, J.M., 1991. The effect of selective loggi ng on Bairds tapir. In: Mares, M.A., Schmidly, D.J., (Eds.), Latin American Mammalogy: Hi story, Biodiversity, and Conservation. Univ. of Oklahoma Press, Norman, OK, pp. 295-304. Fricke, H.C., ONeil, J.R., 1996. Interand in tra-tooth variation in the oxygen isotope composition of mammalian tooth enamel phospha te: Implication for palaeoclimatological and palaeobiological resear ch. Palaeogeography, Palaeoc limatology, Palaeoecology 126, 91-99. Galetti, M., Keuroghlian, A., Hanada, L., Morato, M.I., 2001. Frugivory and seed dispersal by the lowland tapir ( ) in southeast Brazil. Biotropica 33, 723-726. Garten, C.T., Taylor, G.E., 1992. Foliar 13C within a temperate deciduous forest: spatial, temporal, and species sources of variation. Oecologia 90, 1-7. Graham, A., 1964. Origin and evolution of the bi ota of southeastern No rth America: evidence from the fossil plant record. Evolution 18, 571-585. Graham, A., 1999. Late Cretaceous and Cenozoi c History of North American Vegetation. Oxford Univ. Press, Oxford. Grawe DeSantis, L.R. 2005. Straight from the tapi rs mouth: applying stable isotope analyses of extant tapirs to constrain pa leoeclogical hypotheses. In: Kellner, A.W.A., Henriques, D.R. H., Rodrigues, T., (Eds.), Congresso LatinoAmericano de Paleontol ogia de Vertebrados: Boletim de Resumos 2. Museu Nacional, Rio de Janeiro, Brazil, pp. 94-95. Haffer, J., 1969. Speciation in Amazoni an forest birds. Science 165, 131-137. Haffer, J., Prance, G.T., 2001. Climate forcing of evolution in Amazonia during the Cenozoic: on the refuge theory of biotic differentiation. Amazoniana 47, 579-607.

PAGE 116

116 Hayek, L.A., Bernor, R.L., Solounias, N., St eigerwald, P., 1991. Preliminary studies of hipparionine horse diet as measured by tooth microwear. Annales Zoologici Fennici 28, 187-200. Henderson, P., Marlow, C.A., Molleson, T.I., Williams, C.T., 1983. Patterns of Chemical Change During Bone Fossi lization. Nature 306, 358-360. Henry, O., Feer, F., Sabatier, D., 2000. Diet of the lowland tapir ( L.) in French Guiana. Biotropica 32, 364-368. Higgins, P., MacFadden, B.J., 2004. Amount Ef fect recorded in oxygen isotopes of Late Glacial horse ( ) and bison () teeth from the Sonoran and Chihuahuan deserts, southwestern United States. Palaeogeogr aphy, Palaeoclimatology, Palaeoecology 206, 337-353. Hobson, K.A., Sease, J.L., 1998. Stable isotope anal yses of tooth annuli re veal temporal dietary records: an example using Stellar s ea lions. Marine Mammal Science 14, 116-129. Holden, J., Yanuar, A., Marty, D.J., 2003. The as ian tapir in Kerinci Seblat National Park, Sumatra: evidence collected through photo-trapping. Oryx 37, 34-40. Hoppe, K.A., 2006. Correlation between the oxygen isotope ration of North American bison teeth and local waters: implication for paleoc limate reconstructions. Earth and Planetary Science Letters 244, 408-417. Horrocks, M., Deng, Y., Ogden, J., Sutton, D.G., 2000. A reconstruction of the history of a Holocene sand dune on Great Barrier Island, northern New Zealand, using pollen and phytolith analyses. Journa l of Biogeography 27, 1269-1277. Hulbert, R., Bloch, J., Poyer, A., 2006. Exception pr eservation of vertebrates from Haile 7G, a new Late Pliocene site from Florida. J ournal of Vertebrate Paleontology 26, 78A. Janis, C.M., 1988. An estimation of tooth vol ume and hypsodonty indices in ungulate mammals, and the correlations of these factors with dietary preference In: Russell, D.E., Santoro, JP., Sigogneau-Russell, D., (Eds.), Teeth Revisite d: Proceedings of the VIIth International Symposium on Dental Morphol ogy, Paris, Mmoires du Musum National d'Histoire Naturelle, series C, Paris, pp. 367. Janis, C.M., Damuth, J., Theodor, J.M., 2000. Miocene ungulates and terrestrial primary productivity: Where have all of the browsers gone? Proceedings of the National Academy of Sciences 97, 7899-7904. Janis, C.M., Damuth, J., Theodor, J.M., 2002. The origins and evolution of the North American grassland biome: the story from the hoofed mammals. Palaeogeography, Palaeoclimatology, Palaeoecology 177, 183-198.

PAGE 117

117 Janis, C.M., Damuth, J., Theodor, J.M., 2004. Th e species richness of Miocene browsers, and implication for habitat type and primary pr oductivity in the North American grassland biome. Palaeogeography, Palaeo climatology, Palaeoecology 207, 371-398. Kapos, V., Ganade, G., Matsui, E., Victoria, R.L., 1993. 13C as an indicator of edge effects in tropical rainforest reserves Journal of Ecology 81, 425-432. Klaassen, W., 2001. Evaporation from rain-wetted forest in relation to canopy wetness, canopy cover, and net radiation. Water Resource Research 37, 3227. Koch, P.L., Tuross, N., Fogel, M.L., 1997. The effects of sample treatmen t and diagenesis on the isotopic integrity of carbonate in biogenic hydroxylapat ite. Journal of Archaeological Science 24, 417-429. Koch, P.L., Hoppe, K.A., Webb, S.D., 1998. The is otopic ecology of late Pleistocene mammals in North America: Part 1. Florida. Chemical Geology 152, 119-138. Koch, P.L., Diffenbaugh, N.S., Hopp e, K.A., 2004. The effects of th e late quaternary climate and CO2 change on C4 plant abundance in the south-central United States. Palaeogeography, Palaeoclimatology, Palaeoecology 207, 331-357. Kohn, M.J., 1996. Predicting animal 18O: accounting for diet and physiological adaptation. Geochimica et Cosmochimica Acta 60, 4811-4829. Kowalevsky, V., 1873. Sur Cuv. Et sur LHistorie palontologique des chevaux. Mm. L Acad. Impriale Sci. St. Ptersbourg 7th Sr. (20) 5, 1-73.+ Krueger, H.W., 1991. Exchange of carbon with biological apatite. Journal of Archaeological Science 18, 355-361. Lee-Thorp, J.A., van der Merwe, N. J., 1991. Aspects of the chemistry of modern and fossil biological apatites. Journal of Archaeological Science 18, 343-354. Levin, N.E., Cerling, T.E., Passey, B.H., Harris, J.M., Ehleringer, J.R., 2006. A stable isotope aridity index for terrestrial environments Proceedings of the National Academy of Sciences 103, 11201-11205. Lizcano, D. J., Cavelier, J., 2004. Ca ractersticas qumicas de salados y hbitos alimenticios de la danta de montaa (Tapirus pinchaque Roulin, 1829) en los Andes Centrales de Colombia, Mastozoologia Neotropical 11, 193-201. Longinelli, A., 1984. Oxygen isotopes in mammal bone phosphate: a tool for paleohydrological and paleoclimatological research? Ge ochimica et Cosmochimica Acta 48, 385-390. MacFadden, B.J., 1986. Fossil horses from Eohippus ( ) to : scaling, Copes law, and the evolution of body size. Paleobiology 12, 355-369.

PAGE 118

118 MacFadden, B. J., 1992. Fossil Horses: Systematics, Paleobiology, and Evolution of the Family Equidae. Cambridge Univ. Press, Cambridge. MacFadden, B.J., 1998. Tale of two rhinos: isotopic ecology, paleodiet, and niche differentiation of and from the Florida Neogene. Paleobiology 24, 274-286. MacFadden, B.J., 2000. Cenozoic mammalian herb ivores from the Amer icas: reconstructing ancient diets and terrestrial communities. Annual Review of Ecology and Systematics 31, 33-59. MacFadden, B.J., 2005. Diet and habitat of t oxodont megaherbivores (M ammalia, Notoungulata) from the late Quaternary of South and Cent ral America. Quaternary Research 64, 113-124. MacFadden, B.J., Cerling, T.E., 1996. Mammalian herbivore communities, ancient feeding ecology, and carbon isotopes: A 10 million-year sequence from the Neogene of Florida. Journal of Vertebrate Paleontology 16, 103-115. MacFadden, B.J., Shockey, B.J., 1997. Ancient feeding ecology and nich e differentiation of Pleistocene mammalian herbivores from Ta rija, Bolivia: morphological and isotopic evidence. Paleobiology 23, 77-100. MacFadden, B.J., Higgins, P., 2004. Ancient ecol ogy of 15-million-year-o ld browsing mammals within C3 plant communities from Panama. Oecologia 140, 169-182. MacFadden, B.J., Cerling, T.E., Prado, J., 1996. Cenozoic terrestrial ecosystem evolution in Argentina: evidence from carbon isotopes of fossil mammal teeth. Palaios 11, 319-327. MacFadden, B.J., Solounias, N., Cerling, T.E., 1999. Ancient diets, ecology, and extinction of 5million-year-old horses from Florida. Science 283, 824-827. MacFadden, B.J., Labs-Hochstein, J., Hulbert, R.C., Baskin, J., 2007 Revised age of the late Neogene terror bird ( ) in North America during the Great American Interchange. Geology 35, 123-126. McKenna, M.C., Bell, S.K., 1997. Classificatio n of Mammals Above the Species Level. Columbia Univ. Press, New York. McLennan, S.M., 1989. Rare Earth Elements in sedimentary rocks; influence of provenance and sedimentary processes. Reviews in Mineralogy and Geochemistry 21, 169-200. Mendoza, M., Janis, C.M., Palmqvist, P., 2002. Ch aracterizing complex craniodental patterns related to feeding behavior in ungulates: a multivariate approach. Journal of Zoology 258, 223-246. Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000. Biodiversity hotspots for conserva tion priorities. Nature 403, 853-858.

PAGE 119

119 National Climatic Data Center, 2009. Monthly global surface data. 1 May 2009. < http://www.ncdc.noaa.gov/oa/ncdc.html > Orme, C.D.L., Davies, R.G., Burgess, M., Eige nbrod, F., Pickup, N., Olson, V.A., Webster, A.J., Ding, T-S., Ras mussen, P.C., Ridgely, R.S., St attersfield, A.J., Bennett, P.M., Blackburn, T.M., Gaston, K.J., Owens, I.P.F., 2005. Globa l hotspots of species richness are not congruent with endemism or threat. Nature 436, 1016-1019. Paleobiology Database, 2006. Tapiroidea, Pa leobiology Database Online Archives. < http:// paleodb.org > Parmesan, C., Yohe, G., 2003. A globally coherent fingerprint of climate change impacts across natural systems. Nature 421, 37-42. Passey, B.H., Cerling, T.E., 2002. Tooth enam el mineralization in ungulates: Implications for recovering a primary isotopic time-series. Geochimica et Cosmochimica Acta 66, 32253234. Passey, B.H., Cerling, T.E., Perkins, M.E., V oorhies, M.R., Harris, J.M., Tucker, S.T., 2002. Environmental change in the Gr eat Plains: an isotopic record from fossil horses. Journal of Geology 110, 123-140. Patrick, D., Terry Jr., D.O., Grandstaff, D.E., 200 1. Rare earth element (REE) variation in fossil and modern bones: the influence of osteological materials an d time. Geological Society of America Abstracts with Programs 33, 27. Poage, M.A., Chamberlain, C.P., 2001. Empirical re lationships between elevation and the stable isotope composition of precipitation and surf ace waters: considerations for studies of paleoelevational change. American Journal of Science 301, 1-15. Post, E., Forchhammer M.C., 2004. Spatial sync hrony of local populations has increased in association with the recent Northern Hemi sphere climate trend. Proceedings of the National Academy of Sciences 101, 9286-9290. Potts, R., Behrensmeyer, A.K., 1992. Late Cenozoi c terrestrial ecosystems. In: Behrensmeyer, A.K., Damuth, J.D., DiMichele, W. A., Potts, R., Sues, H-D., Wing, S.L. (Eds.), Terrestrial Ecosystems Through Time: Evolutionary Paleo ecology of Terrestrial Plants and Animals. Univ. of Chicago Press, Chicago, pp. 419-541. Quade, J., Cerling, T.E., Barry, J.C., Morgan, M.E., Pilbeam, D.R., Chivas, A.R., Lee-Thorp, J.A., van der Merwe, M.J. 1992. A 16-Ma reco rd of paleodiet using carbon and oxygen isotopes in fossil teeth from Pa kistan. Chemical Geology 94, 183-192. Radinsky, L.R. 1984. Ontogeny and phylogeny in hor se skull evolution. Evolution 38, 1-15. Retallack, G.J., 2001. Cenozoic expansion of gra sslands and climatic cooling. Journal of Geology 109, 407-426.

PAGE 120

120 Roux, X.L., Bariac, T., Sinoquet, H., Genty, B., Piel C., Mariotti, A., Girardin, C., Richard, P., 2001. Spatial distribution of leaf water-use efficiency and carbon isotope discrimination within an isolated tree crow n. Plant, Cell and Environment 24, 1021-1032. Salas, L.A., 1996. Habitat use by lowland tapirs ( L.) in the Tabaro River Valley, southern Venezuela. Canadian Journal of Zoology 74, 1452-1458. Salas, L.A., Fuller, T.K., 1996. Diet of the lowland tapir ( L.) in the Tabaro River Valley, southern Venezuela. Canadian Journal of Zoology 74, 1444-1451. Snchez, B., Prado, J.L., Alberdi, M.T., 2003. Pa leodiet, ecology, and extinction of Pleistocene gomphotheres (Proboscidea) from the Pampean Region (Argentina). Coloquios de Paleontologa 1, 617-625. Snchez, B., Prado, J.L., Alberdi, M.T., 2004. Feeding ecology, dispersal, and extinction of South American Pleistocene gomphotheres (Gomphotheriidae, Proboscidea). Paleobiology 30, 146-161. Sankovski, A., Pridnia, M., 1995. A comparison of the Southern Appalachians (U.S.A.) and Southwestern Caucasus (Russia) forests: in fluences of historical events and present environments. Journal of Biogeography 22, 1073-1081. Schubert, B., Wallace, S., 2006. Amphibians and rept iles of the Mio-Pliocene Gray Fossil Site and their paleoecological implications. J ournal of Vertebrate Paleontology 26, 122A. Shunk, A.J., Driese, S.G., Clark, G.M., 2006. Latest Miocene to earliest Pliocene sedimentation and climate record derived from paleosinkhole fill deposits, Gray Fossil Site, northeastern Tennessee, U.S.A. Palaeogeography, Pa laeoclimatology, Palaeoecology 231, 265-278. Simpson, G.G., 1953. The Major Features of E volution. Columbia Univ. Press, New York. Solounias, N., Moelleken, S.M.C., 1993. Dietar y adaptation of some extinct ruminants determined by premaxillary shape. Journal of Mammalogy 74, 1059-1071. Sponheimer, M., Lee-Thorp, J.A., 1999. Oxygen isotopes in enamel carbonate and their ecological significance. Journal of Archaeological Science 26, 723-728. Stehli, F.G., Webb, S.D., 1985. The Great American Biotic Interchange. Plenum Press, New York. Stromberg, C.A.E., 2005. Decoupled taxonomic radiation and ecological expansion of openhabitat grasses in the Cenozoic of North America. Proceedings of the National Academy of Sciences 102, 11980-11984. Stuwe, M., Abdul, J.B., Nor, B.M., Wemm er, C.M., 1998. Tracking the movements of translocated elephants in Malaysia using satellite telemetry. Oryx 32, 68-74.

PAGE 121

121 Terwilliger, V.J., 1978. Natural history of Bair ds tapir on Barro Coolorado Island, Panama Canal Zone. Biotropica 10, 211-220. Tieszen, L.L., Boutton, T.W., Tesdahl, K.G., Sl ade, N.A., 1983. Fractionation and turnover of stable carbon isotopes in anim al tissues: implications for 13C analysis of diet. Oecologia 57, 32-37. Tobler, M.W., 2002. Habitat use an d diet of Bairds tapirs ( ) in a montane cloud forest of the Cordillera de Talamanca, Costa Rica. Biotropica 34, 468-474. Todd, N., 2006. Trends in Proboscidean divers ity in the African Cenozoic. Journal of Mammalian Evolution 13, 1-10. Trueman, C.N., 1999. Rare Earth Element geochemi stry and taphonomy of terrestrial vertebrate assemblages. Palaios 14, 555-568. Trueman, C.N.G., Behrensmeyer, A.K., Tuross, N., Weiner, S., 2004. Mineralogical and compositional changes in bone exposed on soil surfaces in Amboseli National Park, Kenya: diagenetic mechanisms and the role of sediment pore fluids. Journal of Archaeological Sciences 31, 721-739. Tu, T.T.N., Krschner, W.M., Schouten, S ., Van Bergen, P.F., 2004. Leaf carbon isotope composition of fossil and extant oaks grown under differing atmospheric CO2 levels. Palaeogeography, Palaeoclimat ology, Palaeoecology 212, 199-213. U.S. Department of Commerce and National Oceanic and Atmospheric Administration, 2006. Get/View on-line climate data. NOAA Satellite and In formation Service and National Climatic Data Center. 28 August 2006 < http://www.ncdc.noaa.gov/oa/climate/ climatedata.html> van der Merwe, N.J., Medina, E., 1989. Photosynthesis and 13C/12C ratios in Amazonian rain forests. Geochimica et Cosmochimica Acta 53, 1091-1094. van der Merwe, N.J., Medina, E., 1991. The canop y effect, carbon isotope ratios and foodwebs in Amazonia. Journal of Archaeological Science 18, 249-259. Wallace, S.C., Wang, X., 2004. Two new carnivores from an unusual late Tertiary forest biota in eastern North America. Nature 431, 556-559 + Supplement Walter, R., Epperson, B.K., 2005. Geographi c pattern of genetic diversity in : contact zone between descendants of glacial refugia. American Journal of Botany 92, 92100. Walther, G-R., Post, E., Convey, P., Menzel A., Parmesan, C., Beebee, T.J.C., Fromentin, J-M., Hoegh-Guldberg, O., Bairlein, F., 2002. Ecologi cal responses to recent climate change. Nature 416, 389-395.

PAGE 122

122 Wang, Y., Cerling, T.E., MacFadden, B.J., 1994. Foss il horses and carbon isotopes: new evidence for Cenozoic dietary, habitat, and ecosystem changes in North America. Palaeogeography, Palaeoclimatology, Palaeoecology 107, 269-279. Waterisotopes.org, 2006. OIPC, The online isotopes in precipitation calculator. Waterisotopes.org. 28 August 2006. < http://wateriso.eas.purdue.edu/waterisotopes/pages/ data_access/oipc.html> West, A.G., Midgley, J.J., B ond, W.J., 2001. The evaluation of 13C isotopes of trees to determine past regeneration environments. Forest Ecology and Management 147, 139-149. Wing, S.L., 1998. Tertiary vegetatio n of North America as a context for mammalian evolution. In: Janis, C.M., Scott, K.M., Jacobs, L.L., (E ds.), Evolution of Tertiary Mammals of North America 1. Cambridge Univ. Pr ess, Cambridge, UK, pp. 37-65. Wing, S.L., Harrington, G.J., Smith, F.A., Bloc h, J.I., Boyer, D.M., Freeman, K.H., 2005. Transient floral change a nd rapid global warming at th e Paleocene-Eocene boundary. Science 310, 993-996. Wing, S.L., Sues, H.D., 1992. Mesozoic and ea rly Cenozoic terrestrial ecosystems. In: Behrensmeyer, A.K., Damuth, J.D., DiMichel e, W.A., Potts, R., Sues, H-D., Wing, S.L., (Eds.), Terrestrial Ecosystems Through Time: Evolutionary Paleoeco logy of Terrestrial Plants and Animals. Univ. of Chicago Press, Chicago, pp. 327-416. Wolfe, J.A., 1975. Some aspects of plant geography of the Northern Hemisphere during the Late Cretaceous and Tertiary. Annals of the Missouri Botanical Garden 62, 264-279. Wolfe, J.A., 1994. Tertiary climatic changes at middle latitudes of we stern North America. Palaeogeography, Palaeoclimat ology, Palaeoecology 108, 195-205. Zazzo, A., Mariotti, A., Lcuyer, C., Heintz, E ., 2002. Intra-tooth isotope variations in late Miocene bovid enamel from Afghanistan: paleobiological taphonomic, and climatic implications. Palaeogeography, Pal aeoclimatology, Palaeoecology 186, 145-161.

PAGE 123

123 BIOGRAPHICAL SKETCH Larisa R. G. DeSantis, a Los Angeles native born in 1979, has had a passion for discovery and the outdoors science childhood. Always as king, why and how, she became fascinated with earths current and past inhabitants and en rolled in her first dino saur class at the Los Angeles Natural History Museum at the age of 6. She began studying paleobiology during her freshman year at the University of Chicago in 1997. In 2000, she earned a Bachelor of Science degree in Resource Management with double honors from the University of California, Berkeley. She later went on to earn a Mast er in Environmental Management degree at Yale University in 2003. While at Yale, she was active in muse um education and outreach through the Yale Peabody Museum. Subsequently, she accepted a pos ition with the American Museum of Natural History in New York City, driv ing a 38-ft Paleontology of Dinos aurs Moveable Museum where she educated students and the public about pale ontology. Determined to integrate her passions for paleontology and ecology, she moved to Gaines ville, Florida in 2004 where she began her doctorate studying the paleoecology of ancient forests and their mammalian inhabitants. While at the University of Florida she maintained involvement in numerous educational outreach programs and completed her doctoral research in 2009. Currently, she is an Assistant Professor in the Department of Earth and Environmen tal Sciences at Vanderbilt University.