The fauna and paleoecology of the late Miocene moss acres racetrack site, Marion County, Florida

MISSING IMAGE

Material Information

Title:
The fauna and paleoecology of the late Miocene moss acres racetrack site, Marion County, Florida
Physical Description:
x, 356 leaves : ill. ; 29 cm.
Language:
English
Creator:
Lambert, W. David, 1963-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1994.
Bibliography:
Includes bibliographical references (leaves 341-354).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by W. David Lambert.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001972873
notis - AKE9653
oclc - 31736283
System ID:
AA00003216:00001

Full Text









FAUNA


AND


PALEOECOLOGY


RACETRACK


SITE,


THE


MARION


LATE


COUNTY,


MIOCENE


MOSS


ACRES


FLORIDA


DAVID


LAMBERT


DISSERTATION
UNIVERSITY OF
REQUIREMENTS


PRESENTED


FLORIDA


GRADUATE


PARTIAL


DEGREE


SCHOOL


FULFILLMENT


DOCTOR


THE


THE


PHILOSOPHY


UNIVERSITY


FLORIDA


1994









ACKNOWLEDGEMENTS


would


like


express


deep


gratitude


David


Webb,


chairman


committee,


without


whose


support


guidance


this


dissertation


would


have


been


impossible.


would


also


like


thank


other


members


advisory


committee,


Crawford


Bruce


Holling,


MacFadden,


John


Ronald


Eisenberg,


Wolff,


Douglas


Jones


their


support


assistance


during


course


this


study.


Gary


Morgan


Florida


Museum


Natural


History


provided


indispensable


help


during


this


study,


giving


invaluable


advice


assistance


different


topics


endeavors


too


numerous


list


here.


This


study


could


not


have


been


conducted


without


skilled


diligent


work


vertebrate


volunteers


fossil


like


preparator


Patrick


Russell


Hilton,


McCarty


performed


associated


numerous


miracles


converting


seemingless


hopeless


jumbles


bone


fragments


into


specimens


capable


revealing


information


about


animals


that


first


owned


them.


indebted


those


gave


their


time


effort


field


crews


from


Florida


Museum


Natural


History


collecting


Moss


Acres


Racetrack


specimens;


without


their


hard


work,


this


study


could


not


have


even


been


contemplated.


particular


note


this


regard


are


Gary


Morgan,


Art


Poyer,


Russell


McCarty,


Erica

season


Siemens,


Moss


whose

Acres


cooperation

Racetrack in


and

the


tolerance


sorincl


made


1991


field


a creat


*-









John


Claytor,


operated


backhoe


used


reopen


site


a voluntary


basis.


Great


thanks


are


John


Schimfessel,


owner


Moss


Acres


first


brought


presence


fossils


this


locality


attention


Florida


Museum


Natural


History,


then


graciously


allowed


crews


from


museum


work


site


over


course


years.


Similarly


great


thanks


are


also


Linda


David Markgraf,


later


owners


Moss


Acres


property


allowed


site


reopened


with


heavy


machinery


subsequent


field


work


occur.


Laurie


Walz


created


specimen


illustrations


presented


chapter


giant


otter


Enhvdritherium.


Barbara


Hansen


University


Minnesota


kindly


identified


pollen


Moss


United


Acres


States


Racetrack


sediment


Geological


Survey


sample.


McCartan


performed


invaluable


mineral


analysis


Moss


Acres


Racetrack


clays.


Both


University


Michigan


Smithsonian


Institution


loaned


specimens


indispensable


this


study.


McKenna


Foundation


personal


generously


salary


provided


one


funds


semester;


both


field


without


work


support


Moss


Acres


Racetrack


could


have


been


reopened


1991,


three


extraordinary


Ameba~io~don


skulls


would


remain


below


ground


undiscovered.


Jerry


Britt,


Foundation


provided


funding


field


work


via


Jerry


Britt,


Pa 1 onh i n 1 nrv


Award ira nnrl t


Pvecrcl ;rn~


1 I


Aw~ rrl


rhir ;Irrn


S iI -








provided


this


with


dissertation.


permission


Lastly,


include


wish


copyrighted


express


gures


gratitude


Florida


Museum


Natural


History


, which


provided


with


access


innumerable


invaluable


facility


personnel


during


conduction


this


study;


without


help,


study


could


have


been


accomplished.


this










TABLE


ACKNOWLEDGEMENTS...


CONTENTS


. ..... ... ... ii


LIST


COLLECTION


ABBREVIATIONS. . . . . .viii


ABSTRACT


CHAPTERS


INTRODUCTION.


THE GEOLOGY
SITE.......


Geology....
Age. . .


AND


MOSS
. .


.0 .4". 0 . .
. . . . . .


ACRES
. . 4


* a .4 4
4 0 5 4


RACETRACK
.. 4 ..


. ... 7
. ....... 24


A REDIAGNOSIS OF AMEBELODON (PROBOSCIDEA,
GOMPHOTHERIIDAE) AND A NEW SUBGENUS AND SPEC
AMEBELODON (KONOBELODON) BRITTI, FROM THE MO
ACRES RACETRACK SITE...... .... . .....


Systemat ics . . . ..
Description. . . . ..
The Rediagnosis of Amebelodon
Dentinal Rods................
Biogeography and Chronology..


........ .... 2
............ .... 3
. .. .. .. .. . 4
..... ... 0.... 5
. .. .. ... .... 5


THE
GIANT


OSTEOLOGY
T OTTER E


, ECOLOGY,


NHYDRITHERIUM


PHYLOGENY
TERRAENOVAE


Methods.....
Osteology...
Paleoecology
Phylogeny...


4* S
* S 4 *
and
* .


Func


onal


Morphology


. 61
. 62
.... 96
..... .. 107


THE


OSTEOLOGY


PLIOMETANASTES


THE


GROUND


XENARTHRA,


SLOTH
MEGALONYCHIDAE) .. .. 116


Materials and
Systematics..
Description..
The Evolution
America.....


Method


of Pli


S ..........


ometanastes


in North


.. .. .. 118
.... ....119
S. . 120


Daae











cape


THE OSTEOLOGY OF PEDIOMERYX HEMPHILLENSIS
(ARTIODACTYLA, DROMOMERYCIDAE) AND THE
PALEOECOLOGICAL SIGNIFICANCE OF DWARFING
PEDIQMERYX LINEAGE.. .. . . .. . . .


Materials and Methods..........
Systematics. . . . . .
Description. . . ....
The Paleobiological Significance
Trends in Pediomervx............


-. < '-^
S S a S ,
INTHE
.......


.* S S S S a S --
* 4 * a a a a a S S *


I
* 0 4 S -.


POPULATION OF
(PERISSODACTYLA,
ACRES RACETRACK


APHELOPS MUTILIS
RHINOCERATIDAE)
SITE............


FROM


MOSS


. ... .. .. ..1


Systematics....
Osteology.. .. .
Population Ecolo
Social Structure
Diet and Feeding


SUMMARY


gy.


Mode


. *.. S S S .
. .. .. .a
. .4 ..a S
*. a S S S
. .a . .a . S


MISCELLANEOUS


.I
* a a a S 4 5 0 S S S S
I


NON-MAMMAL


MAMMAL


TAXA


FROM


THE


MOSS


ACRES


RACETRACK


SITE..


las
las
las
las
las
las


S
S


Chondricht]
Osteichthyi


s Amphibia..
s Reptilia..
s Aves......
s Mammalia..


hyes


S SS 05 555 SS SS SSS S S.2


es........ .. .................. .2

................... .. .......... 2
...............................2 2
S- a........... ...... .. .. .. .2


THE


TAPHONOMY


MOSS


ACRES


RACETRACK


SITE..


Data...........
State of Bone Pr
Bone Association
Bone Orientation
Faunal Biases...


THE PALEOCLIMATE
ACRES RACETRACK


reservation.


."..


AND
SITE


* S S


. . . .
..a. a a a a .
. . 0 S a a. .. .a
. .a .S S . .S S S .
*.. a. S. S S. S. S S S


PALEOECOLOGY
a a. ... 4 aO0 5 S


THE


. ...2


... 242
.. .248
....253
....256


MOSS
. a a a .


Paleoclimate.....
Flora.. . .
Fauna ..........


. ..a.* a ... 5
S.. . .' 0 .a. .S.


* S S S


* S S 0
* S S S
* a a


AND


I










caoe


SUMMARY


AND


eologyC..
Taohonomy
Fauna....
Flora. ...
Climate..


ommun
OIWr 24U


* al S 4 *


remarks


............................32
....~........................32
............................332
............................32
............................32
* a a a a a 4 4 4 4 * a a * a > *
* a a . . . . . 322
* a a a * . . .32^
. a . . . . . 322
. . . a * a a .- 22
. . . . . . a 322


a
4


APPENDIX


AMEBELODON
RACETRACK
OF NATURAL


APPENDIX 2
MOSS ACRES


FLORIDA


LIST


POST


-CRANIAL ELEMEJ


(KONOBELODON) BRITTI FROM
ITE IN THE COLLECTION OF
HISTORY... .... ......


PEDIOMERYX
ACETRACK S


MUSEUM


THE


THE


NTS REFERR
MOSS ACRE
FLORIDA MU


HEMPHILLENSIS SPECIMEN
TE IN THE COLLECTION O


NATURAL


HISTORY...


THE


S EUM


THE

4 4 a a


APPENDIX 3 APHELOPS MUTILIS
ACRES RACETRACK IN THE COLLE
OF NATURAL HISTORY......


SPEC
TION


ROM TH
FLORI


EUM


..332


APPENDIX 4
MOSS ACRES


FISH, REPTILE
RACETRACK SITE


, AND BIRD SPECIMEN
IN THE COLLECTION


THE


FLORIDA


MUSEUM


NATURAL


HISTORY........


APPENDIX


ACRES
MUSEUM


MISCELLANEOUS


RACETRACK SITE IN THE
OF NATURAL HISTORY..


MAMMAL


SPECIMENS


COLLECTION


- Fl..


THE


THE


MOSS


FLORIDA


APPENDIX 6 HORSE (EQUIDAE)
ACRES RACETRACK SITE IN THE
MUSEUM OF NATURAL HISTORY..


SPECIMENS
COLLECTION


FROM
OF


THE
THE


MOSS
FLORIDA


APPENDIX 7 LIST
ACRES RACETRACK
1991, INCLUDING
DESCRIPTION OF


OF SPECIMENS RECOVERED FROM THE MOSS
SITE DURING THE SPRING FIELD SEASON
FIELD NUMBERS, DATE MEASURED, AND A
SPECIMENS . . .... .. .... .... .


APPENDIX
SELECTED
RACETRACK


POSITIONAL AND ORIENTATION
SPECIMENS COLLECTED FROM THE
SITE DURING THE FIELD SEASON


DATA
MOSS
OF S


FOR
ACRES
;PRING


1991...


LITERATURE

BIOGRAPHIC


CITED..... ... .. .. .... .. .. .- --

SKETCH .. . .. ........ --


CONCLUSIONS........................









LIST


COLLECTION


ABBREVIATIONS


(OTHER


ABBREVIATIONS


EXPLAINED


THE


TEXT)


D.M.N.H.:


Denver


Museum


Natural


History,


Denver,


Colorado


K.U.M.:


University


Kansas,


Lawrence,


Kansas


MCZ:


Museum


Comparative


Zoology,


Cambridge,


Massachusetts


T.M.M..


Texas


Florida


Memorial


Museum


Museum

Natural


, Austin,

History


Texas

Collection


Paleontology,


Gainesville,


Florida


UF(M)


: Florida


Museum


Natural


History


Collection


Mamma logy,


Gainesville,


Florida


UM-V:


University


Michigan


Vertebrate


Paleontology


Collection,


Ann


Arbor,


Michigan


UNSM:


Nebraska


State


Museum,


Lincoln,


Nebraska


USNM(M):


Smithsonian


Institution


Mammalogy


Collection,


Washington


D.C.









Abstra
The


Lct


Dissertation


University
Requirements


Presented


Florida


Degree


Partial


Doco


le Graduate Schc
Fulfillment of
r of Philosophy


THE


FAUNA


AND


PALEOECOLOGY


RACETRACK


SITE,


MARION


LATE


COUNTY,


MIOCENE


MOSS


ACRES


FLORIDA


David

April,


Lambert

1994


Chairman:


Major


Department


David


Webb


: Zoology


Moss


Hemphillian

County, Fie


Acres


age,


)rida,w:


Racetrack


clay-filled s

ith a diverse


site


inkhole

biota,


represents


western


possib


early


Marion


containing


more


evidence


local


biotic


climatic


conditions


than


other


late


Miocene


locality


Gulf


Coastal


Plain.


geology


taphonomy,


fauna,


palynoflora,


community


ecology


this


locality


were


examined


this


study.


Geological


taphonomic


evidence


suggests


that


fossil


biota


was


preserved


under


strictly


nonfluvial


conditions


isolated


pond


lake,


indicating


that


this


assemblage


represents


a real


community


rather


than


mixed-


assemblage.


faunal


description


included


detailed


study


following


mammalian


taxa:


gomphothere


proboscidean


Amebelodon


(Konobelodon)


britti


(described


a new


subgenus


species),


ground


sloth


P i~omentanastes


(represented


previously


unknown


limb


elements),


giant


otter


Enhvdritherium


(represented


partial


skeleton,


best


ol1









intact


horn


cores),


rninoceros


Aoheloos


ilia.


Other


mammal


taxa


(including


horses)


were


mined


a cursory


fashion


only.


Preliminary


evidence


gests


presence


new


species


emydid


turtle


;uffemrts


a new


species


Alliaator


within


petofauna.


palynoflora


revealed


dominance


grass


I oak


suggesting


iividually


forming


savanna

minute


landscape,


portions


with

the


all

flora


other

The


taxa


(map


floral


resembles


that


a modern


warm


temperate


woodland


southeastern


United


States,


suggesting


temperatures


?ilar


those


found


region


today.


Rainfall


re
patterns


Inconclusive,


though


presence


savanna


landscape


;gests


strong


seasonal


aridity.


nature


5 largely


community


inconclusive.


interactions


Terrestrial


within


carnivores


biota nd
biota


n its


were


stually


-dence


rlcan


absent,


possibly


facilitative


savannas


was


result


grazing


searched


for,


like


with


small


that


sample


found


final


size.


until
in modern
ugh
results


ambiguous.


Amebelodon


was


shown


to be


a plausible


Logical


analog


modern


African


elephants


savannas,


hearing


woodlands


limit


their


spread


into


grasslands


thus


encouraging


maintaining


biotic


diversity ng
diversity.


* of















CHAPTER


INTRODUCTION


The Moss


Acres


Racetrack site consists of


sinkhole


within both Eocene and Miocene


limestone


in Marion County,


Florida,


near the small


town of Morriston in Levy County


(map


coordinates NW 1/4,


NE 1/4


sec.


T14S,


R19E,


Morriston


Quad.)


(Figure 1-1).


site


was


initially discovered in


1984,


when


during


construction of


a racetrack


fossils


were


discovered


Shimfessel,


Moss


the owner


Acres


South horse


the site,


brought


farm.


this


John


discovery


the attention


the Florida Museum of Natural


History


generously


gave permission for the site to be worked when


importance was revealed.


Museum crews worked Moss


Acres


Racetrack and


made


extensive


collection of


specimens


until


spring of


1987,


which point


the site was closed.


Though


small


amounts


material


was


collected in


the summer of


1989,


work at Moss


Acres


Racetrack did not


restart


significant


degree


until April


1991.


site


was


only


worked


this


until


brief


the middle of June of


period


that


the collection of Moss


year,


Acres


though during


Racetrack


material


was


enriched considerably


including


recovery


some


spectacular specimens.


The site has not


been worked


-, n n 1 1 -


". U~^














N

t


Figure
Acres


1-1.


Racetrack


of Florida showing the
locality.


location


Moss









Moss


three


Acres


reasons:


Racetrack


overall


fauna


high


remarkable


quality


primarily


spec imen


preservation,


preservational


circumstances


specimens,


unusual


ecological


biases


animals


represented.


associated


Many


even


specimens


articulated


preserved


skeletons,


are


rare


partial


occurrence


Florida,


fauna


evidently


represents


unmixed


local


community


sense


that


there


evidence


that


specimens


were


transported


water


from


other


localit


ies


mixed


with


remains


animals


that


actually


large


died


herbivores


site.


dominated


site


hooved


yielded


mammals,


rauna


most


impressive


which


shovel-tusked


gomphothere


Amebeloffon


which


this


one


richest


locai~tie


world.


Other


herbivores


found


site


lude


giant


ground


sloth


Pliometanas


tes


possible


ancestor


Meaalonvx),


wide


variety


horses


including


both


tridactyl


monodactyl


taxa,


rhinoceros


Anhe los,


unidentified


peccary,


two


camels


(one


a long-necked


giant


, and


other


comparable


a modern


lama),


hornless


deer-like


ruminant


Pseudoceras


strange


horned


ruminant


Pedomervx,


which


a single


horn


protruding


out


back


head


well


two


ordinary


frontal


horns.


However,


surprisingly


despite


abundance


herbivores


terrestrial


carnivores


are


virtually


absent


from


fauna,


only


confirmed









carnivores at


site


include an abundance of


alligators and


a single partial skeleton of the giant otter Enhvdritherium


(possibly related


the living


sea


otter),


the most


complete


specimen of a fossil


"bunodont"


otter known in the world.


The Moss


Acres


Racetrack


fauna


clearly


exceptional


terms of


faunal richness.


However,


this


faunal


richness


itself does not give this site special


paleontological


significance,


many sites


close


in age being equally rich or


richer


(e.g.,


the Coffee Ranch,


Texas


[Schultz,


1977] ;


Rhino


Hill,


Kansas


[Bennett,


1977];


and the


Love


Bone Bed,


Florida


[Webb


1981])


Rather,


the importance of


this


site


stems principally


from three


factors:


unique


specimens


known from the fauna


(including the well preserved skulls


famous genus Amebelodon,


an element unknown


from any


other


site


in the world;


a nearly complete skeleton of the


"bunodont"


otter


Enhvydritherium,


the most


complete specimen


the world for this type of otter;


the associated fore-


hindlimbs of


the ground sloth Pliometanastes,


preserving


elements previously unknown


for this genus;


and the


first


Pediomervx skull


ever found,


even preserving


intact


occipital


frontal


horns);


despite ecological biases,


the nearly


complete preservation of


one aspect


the fauna,


the ungulate

palynoflora,


"community;"

a rarity in


and 3)


Tertiary


a well preserved

sites of southeastern North


America, p


particularlyy


in association with


a good


terrestrial










taxonomic group


fauna examined


detail


horses


(Hulbert,


1987;


1988a;


1988c;


1993b),


with


other


portions


fauna remaining


undescribed.


Lacking


proper


taxonomic


description,


remained


impossible


oertorm


a proper paleoecological


comprehensive


examination


analysis


climate,


locality


flora,


:rnclucding
,,nf 1i, ri ^inl
--A _^-S L -I,A


"-ua
- tI -- ^
,d cxlld


an analysis


that


can


performed


very


locaiities


because of


such limitations


faunal


gaps and


unknown


flora.


Indeed,


such


late


Miocene


localities


nave


been


studied from


southeastern


United


States


with


result


that


little


known about


tne


ecological


conditions
r"rr~ _O^1 "r S^


in this


region during this


time.


The goal


this dissertation


a complete


taxonomic


description of


the Moss Acres Racetrack


fauna,


going


beyond


the the work done by Hulbert on


horses,


utilization of


this description


combination


with


other


data


to create a paleoecological


reconstruction


locality.


examine


first


the geology


chapter


the Moss


this


Acres


dissertation


Racetrack


will


site,


providing


insight


into the physical


habitat


which


fossil


community


existed.


Subsequent


chapters will


focus


describing


fauna,


with an emphasis placed on


osceological


description but

autecology also


with systematics


(including phylogeny)


addressed where appropriate.


Since


they


- -- m~ ~cr ?,r nrn r no


f ciivprs.. w..v


7 n f P Ym C


-1


,,,,.,:,,


fck


ri








reptiles


this


description


(birds,


amphibians,


and


fish


being

though


largely


reptiles


completely


are


addressed.


unknown


unexplained


Following


reasons),


taxonomic


descriptions


a chapter


taphonomy


Moss


Acres


Racetrack,


combining


geological


with


taxonomic


information


investigate


conditions


preservation


fauna.


that


fossil


Finally,


geological,


faunal,


taphonomic


information


from


preceding


chapters


are


combined


analysis


paleoecology


Moss


Acres


Racetrack


site,


examining


paleoclimate


possible


community


interrelationships


(including


plant-animal


animal-animal)


that


existed


originally


locality.


This


analysis


provides


new


information


about


poorly


known


ecological


conditions


that


existed


southeastern


United


States


(and


extent


North


America


a whole)


during


late


Miocene


some


factors


that


have


influenced


them.
















CHAPTER


GEOLOGY


AND


AGE


THE


MOSS


ACRES


RAC ETR AC K


SITE


Geoloav


Moss


Acres


Racetrack


site


consists


clay


unknown


having


depth


least


extent,


small,


conta


local


inning


floor


_limestone


sandy,


fragments


phosphatic


limestone.


exposure,

oriented


discerned.


stratigraphy


majority


exposures


Although


from


onl


fossils


which


bones


have


Ly known

recovered


stratigraphy


been


from

from

could


recovered


a single

horizontally


not


from


most


superficial


levels


clay


bed,


vast


majority


have


come


from


from


single


unit


approximately


one


meter


thick.


regional


stratigraphic


relationships


local


sedimentary


facies


shed


light


both


nature


preservation


fossils


site


eco


logy


preserved


paleocommunity


a whole.


Reaxo~nrl.


Geoloav


most


prominent


geological


units


northcentral


Florida


are


Ocala


Limestone


group


including


three


format ions


Iglis, W


Jilliston,


Crystal


River)


the









below),


Alachua


Formation


lithologically


diverse


unit


uncertain


stratigraphic


relationship


aforementioned


Vernon,


Moss


units)


1964) ,


Acres


Racetrac]


(Vernon,


these

k site


1951;


Scott,


formations

Other 1


are


Inlts


1988;


most


Puri


relevant


variably


present


northcentral


Florida,


including


Suwannee


(limestone)


Citronelle


Formations


(Vernon,


1951;


Scott,


1988;


Pirkle


al.,


1965) ,


have


bearing


site


are


discussed


further


here.


northcentral


Florida


Eocene


Ocala


Limestone


forms


effective


basement


which


younger


sediments


are


superposed.


Over


much


northcentral


Florida,


state


whole,


sediments


tend


overlay


Ocala


Limestone


fairly


ordinary


unconformable


fashion,


superpositioned


units


being


level


relative


each


other


(with


one


exception


being


localized


karst


infillings;


see


below)


Ocala


relatively


uniform


unit


lithologically,


composed


essentially


pure


marine


biogenic


limestone


with


clastics


nearly


completely


absent.


regions


Levy,


Marion,


Alachua,


Citrus


Counties


Ocala


Limestone


rises


locally


form


a single,


distinctly


ridge-like


structure


commonly


called


Ocala


Arch;


sediments


along


slopes


this


structure


form


skin


that


gradually


decreases


thickness


with


rising


elevation,


vanishing


near


the arch


leave


Ocala









involves


dissolution


basement


Ocala


limestone


rather


than


overlying


limestones


such


Hawthorn,


Hawthorn


karst


formation


being


more


typical


southern


haf -
^ ;s*


- orida


(Puri


Vernon,


1964;


see


below)


majority


Neogene


sediment


covering


Ocala


limestone


northcentral


Florida


consists


Hawthorn


(Scott,


:988)


Traditionally,


Hawthorn


nas


been


treated


literature


single


formation


(e.g.


Pirkle,


1956;


Purl


Vernon,


1964).


However,


Scott


(1988)


promoted


Hawthorn


status


a group,


establishing


within


number


new


formations


:nat


vary


between


northern


and


southern


halves


peninsula.


northcentral


Florida


new


formations


established


within


Hawthorn


Scott


(1988)


include


order


from


bottom


top:


Penney


Farms,


Marks


Head,


Coosawhatchee


Formations,


with


Statenville


Formation


encompassed


Coosawhatchee.


question


whether


Hawthorn


properly


formation


distinct


this


group


study,


without


specific


issue


henceforth


beyond


Hawthorn


reference


unit


scope


intent


will


referred


status.


Hawthorn


extremely


heterogeneous


unit


lithologically,


consisting


clays,


dolomitic),


sands,


with


dolomites,


phosphate


imestones


nodules


(both


(phosphorite)


calcitic


present


throughout.


concentrations


some


o c r: I


localities,


ne main


- i


extensive


bnriv


fI


phosphate


14 S rl ar hr'r


f- "R i~ c: i~s








carbonate,


(Pirkle et al.,


usually


1965) .


a mixture of these sediments


The Hawthorn represents a marine


and/or


estuarine


facies,


indicated by


mollusk


:auna


and the abundance of phosphorite


(Scott,


1988;


Vernon,


1951;


and Pirkle,


1956)


The Alachua Formation


a controversial


unit,


considered to be a true


formation by some


(e.g. ,


Sellards,


1914;


Vernon,


1951;


and Puri and Vernon,


1964) ,


and a


unit


uncertain status,


possibly not a


true


formation at all


(e.g.,


Scott,


1988).


The uncertainty about


formation


status


the Alachua Formation stems


from two factors:


discontinuous


geographical


distribution


with an associated


ambiguity as


to stratigraphic


relationship


to adjacent


units;


and 2)


a complex,


highly variable


lithology.


Alachua sediments are not continuous,


rather


form


distinct


pockets


in localized


limestone basins


scattered


throughout northcentral


Florida


found


two distinct


zones


separated by a gap,


Gilchrest,


Levy,


Alachua,


Marion Counties


to the north and Citrus,


Sumter,


Pasco


Counties to the south


[Brooks,


19811).


Thus,


unlike a proper


formation it


neither traceable nor mappable.


In addition,


Alachua sediments cannot be clearly placed into a


stratigraphic


sequence


relative


to neighboring


formations.


Alachua Formation sediments are known


in pockets of


the Ocala


Limestone


(particularly near the top of


the Ocala


Arch),









formation


mappable


have


a distinct


stratigraphic


position


(Boggs,


1987),


difficult


justify


calling


Alachua


true


formation.


difficult


characterize


litholo


Alachua


Formation


because


variability.


However,


can


crudely


described


terrestrial,


lacustrine,


and/or


fluviatile


mixture


interbedded,


irregular


deposits


clays,


sand,


sandy


clay


diverse


characteristics.


Though


considerable


var


nation


exists,


Alachua


Formation


typically


following


stratigraphy.


upper


portion


characterized


faintly


stratif


led,


cross-bedded,


poor ly


indurated,


fine-grained


gray


quartz


sand


contained


matrix


unidentified


phosphatic


clay


mineral.


Interbedded


with


commonly


underlying


these


sands


are


occasional


pebbles


waterworn


flint,


erratic


boulders


limestone,


silicified


limestone


flint,


clay


lenses,


pebbles,


boulders


phosphorite,


phosphatic


clay


conglomerate,


colloidal


phosphate,


occasional


concentration


vertebrate


fossils.


abovementioned


clay


lenses


are


formed


montmorillonite


light


pastel


shades


green


blue,


occasionally


containing


phosphorite


best


a minor


constituent)


silicified


clay.


This


clay


waxy,


blocky,


irregular


distribution


though


commonly


separates


concentrations


phosphorite


from


underlying


,: rr nne-4 'r


41~~C nr~ T rIL Alrin7


fnrrnei


811r hll~


flnnr


RT TnP


I" ^'r~ -


; -









phosphatic


clay


greenish-gray


waxy


clays


(Vernon,


1951;


Vernon


Puri,


1964).


Terrestrial


vertebrate


faunas


collected


from


Alachua


sediments


range


from


early


Miocene


(e.g.,


Hemingfordian


Thomas


faunas


Farm


Local


from


Fauna)


Haile


Pleistocene


complex)


age


(e.g. ,


(Vernon,


many


1951;


local


Puri


Vernon,


1964) .


Thus,


these


sediments


have


chronological


range


least


million


years,


which


could


intuitively


reduce


credibility


Alachua


a formation.


should


noted,


however,


that


neither


chronological


range


nor


continuity


are


crucial


considerations


determining


formational


status


a unit


according


North


American


Stratigraphic


Code


(Boggs,


1987).


Stratiaranhv


Moss


Acres


Racetrack


Site


Virtually


stratigraphic


information


Moss


Acres


Racetrack


site


based


a single


vertical


exposure


excavated


Moss


when


Acres


site


Racetrack


was


worked


site


can


1991.


divided


clay


into


bed


four


distinct

excluding


units

the


numbered c

superficial


:onsecutively


covering


from

grass


bottom,


topsoil.


Approximations


relative


thicknesses


units


along


with


brief


descriptions


are


presented


Figure


2-1.


Unit


a white,


unconsolidated


clay


layer


roughly


centimeters


or less


thickness,


differina


trom all o


















unmeasured



unmeasured




100cm








> 300cm


sandy white clay



loose orange-brown clay



green, brown, and orange clay,
peat, and silicified clay







brown and orange clay, silicified
clay



exposure floor


Figure


2-i.


stratigraphy


Schemati
of one


c diagram
exposure


illustrating the
t the Moss Acres


sedimentological
Racetrack site.









deteriorated


caliche


nodules.


fossils,


vertebrate


invertebrate,


have


been


found


this


unit.


Unit


clay


darker


than


unit


being


orange-brown


color


(possibly


oxidation)


Unlike


unit


pure


clay,


with


signs


sand


other


impurit


beyond


isolated


chunks


suspended


limestone.


Unit


resembles


unit


being


poorly


consolidated,


possibly


weathering.


Vertebrate


fossils


have


been


recovered


from


unit


these


fossils


are


poorly


preserved,


condit


consistent


with


hypothesis


that


this


unit


been


subjected


considerable


weathering


(see


chapter


Unit


most


fossiliferous


unit


site.


Like


unit


unit


clay


appears


quite


pure.


impurities


except


some


large


chunks


limestone


phosphate


pebbles,


presumably


reworked


from


adjacent


walls.


Small


peat


amounts


also


fine


occur


silica


Unit


sand


clay


isolated


varies


pockets


considerably


color,


ranging


from


orange


brown,


blue,


green;


much


most


heavily


streaked


numerous


dark


lines


with


preferred


orientation,


perhaps


some


elemental


contamination


(similar


streaks


are


abundant


fossil


bones


taken


from


site)


Isolated


pockets


unit


clay


contain


degraded


peat,


identified


Newsome


(pers.


comm.),


which


gives


clay


dark


gray


black


color


greasy


texture.


When


freshly


exposed,


he clay of


this


unit









"sludge,


drying


becoming


extremely


crumbly


s W1 C~r


expanding


when


exposed


water


form


foam.


addition


blocky,


waxy


clay,


there


are


pocKets


1i
C, ~a


* -'
/4 ~ K. ,


texture


reminiscent


soft


limestone.


zSucfl


ockets


presumably


represent


silicif


clay,


which


s known


both


Hawthorn


Alachua


Formations


(Pirkle


a. ..


1965;


Vernon,


1951) .


Pockets


this


hardened


clay


tend


particularly


abundant


this


unit


near


accumulations


fossil


bone,


suggesting


two


possibilities


local


diagenetic


processes


involving


biological


material


have


influenced


chemistry


surrounding


matrix;


organisms


not


helped


normally


introduce


deposited


were


site.


accompanied


While


material


significantly


thicker


than


units


unit


not


particularly


thick


whole,


measuring


approximately


meter


measured


section


vertical


exposure


probably


not


exceeding


meters


units


anywhere


fairly


exposure.


distinct;


contact


however,


between


contrast


contact

because


between


units

strong


rather


similarity


colo


subtle,

r and


part


superficial


texture


between


clays


(see


below).


Unit


not


dramatically


different


appearance


from


unit


Unit


typically


brown


orange,


unlike


unit


which


displays


blue


green


well


brown


orange


1-ninaNP1oPrt~r~ hPczS


ITni t


ran ne


nii s ri nrauished he


from


hrlac









unit


clay


blocky


waxy


when


unsilicified.


Vertebrate


fossils


are


essentially


absent


from


this


unit,


although


one


heavily


worn,


isolated


AoheloDs


vertebra


was


encountered.


Unit


thickest


four


units,


having


minimal


thickness


three


meters


(the


depth


floor


deepest


part


exposure).


full


thickness


this


unit


could


determined,


nor


could


possibility


lateral


variation


investigated


small


size


exposure.


clay


Moss


Acres


Racetrack


must


contact


underlying


limestone


unit,


such


unit


was


exposed


measured


section.


exposure


yellowish


sandy


limestone


was


observed


infield


region


track


when


site


was


worked


between


1985


1987


(Webb


Morgan,


pers.


comm)


This


limestone


exposure


just


north


productive


clay


exposures


protruded


upward


into


clay,


either


Unit


This


limestone


contained


considerable


amounts


sand


water


worn


phosphatic


gravel


(Morgan,


pers.


comm.),


almost


certainly


represents


Hawthorn limestone.

phosphatized shark


This

teeth,


limestone


e.g. ,


produced


Neaaorion


brevirostris,


Galeocerdo


sp. ,


Carcharodon


auriculatus,


which


help


corroborate


early


Miocene


age.


Reworked


invertebrate


fossils


as well,


diagnostic


indicating


Eocene


that t


have


clay


been


or was


found


one


clay


time in










positional


relationship


Ocala


Si me s tone


clay


Hawthorn


unknown.


Mineraloav


sample


clay


from


unit


was


sent


McCartan


United


analysis


States


X-ray


Geological


Survey


differaction.


mineralog


result


ical


this


analysis


follows.


attributed


some


clay


not


residues


montmorillonite,


Hawthorn


common


well


Alachua


Formation


(Vernon


1 OR6.


Puri


Vernon,


1964),


rather


illite


kaolinite.


clay


remarkably


pure,


showing


signs


heavy


minerals


such


garnet,


zirconium,


tourmaline,


common


Hawthorn


sediments


including


clays


(Pirkle


al. ,


1956).


Phosphates


are


also


absent


except


rare,


isolated


nodules


probably


reworked


from


Hawthorn.


only


significant


mineralogical


impurity


observed


silica,


present


only


minute

silicic


grains

acid


perhaps


resulting


froth


formed

n the


precipitation


hydrolyt


breakdown


silicate


minerals


(see


below).


This


mineralogy


strong


implications


conditions


under


which


these


sediments


were


formed.


Both


illite


kaolinite


are


formed


subaerial


weathering.


Kaolinite


tends


form


under


severe


chemical


leaching


*1 - 1 - - -


I.- A


..


* n, rr r* t*


r I I


- A


I i









such


when potassium


feldspar-rich


rocks


(e.g. ,


granites)


are broken down


(Boggs,


1987) .


The presence of


the silica


grains


in the


clay


suggests


the possibility that


localized


mineral breakdown occurred at


the site.


The hydrolysis


silicates


that


leads


formation of clays


such


kaolinite and illite also


forms silicic acid


(H4SiO4)


by-product,


which can precipitate


into small


grains of


silica


(SiO2)


well as combine with cations


form clay


(Boggs,


1987) .


Though not


examined by


X-ray


analysis,


the hardened


clays


in units


and unit


are


in all


likelihood


silicified,


conceivably


through the


process


percolation and


precipitation of


silicic acid.


Formational


Status of


the Moss


Acres


Racetrack Site


The Moss


Acres


Racetrack site has


two major geological


features,


the clay bed and the underlying


limestone formation(s)


have only


limestone.


a minor bearing on


The

the


paleoecology and taphonomy of


the mammal


fauna of Moss


Acres


Racetrack,


are discussed above,


and will


further


considered here.


Thus,


following


discussion


concentrates


the clays.


The Moss


Acres


Racetrack clay bed can be plausibly


referred to only the Hawthorn or the Alachua


Formation,


since


the major geological


units


found in northcentral


Florida


only these extend into


the Miocene and


are


known


to contain









likely


source


clay


basis


a number


factors,


including:


virtual


absence


phosphorite


from


least


units


(with


condit ion


unif
' 1 -| t


currently


unknown),


complete


absence


such


ne avY


minerals


garnet,


tourmal


mne,


zircon,


a probable


origin


clay


rom


fresh


water


rather


tnan


a3 m~ann C


facies,


indicated


presence


peat


unit


absence


a distinctive


marine


fauna.


clay


bears


distinct


resemblance


clay


lenses


someti mes


found

(e.g.,


the Ale

blocky,


ichua


waxy


Formation,

appearance


having

with


a similar


some


texture


slic i icat ion),


color


(shades


blue


green


Alachua


clay


compared


with


blue,


green,


orange,


brown


Moss


Acres


Racetrack


clay


unit


blue


green


unit


facies


(both


fresh


water/terrestrial).


mineralogy


appears


differ


considerably


between


two,


Alachua


clay


being


composed


montmor


illonite


rather


than


kaolinite


illite


mixture


found


Moss


Acres


Racetrack.


However,


possibility


exists


that


presence


montmorillonite


Alachua


clay


cited


Vernon


was


result


a misidentification,


since


illite/kaolinite


mixture


Moss


Acres


Racetrack


superficially


analysis


resembles


Alachua


montmorillonite


Formation


formal


montmorillonite


X-ray


was


mpnrt i onnPd


Tha.i -


Moss


Acres


Racetrack


clay


can


srtu


I


?


. ,-









Puri


(1964).


However,


questions


validity


Alachua


Formation


whole


aside,


lithostratigraphic


unity


these


clay


beds


doubtful.


Considering


that


"Alachua


clays"


can


overlie


Ocala,


Hawthorn,


even


Suwannee


limestone


wide


range


faunas


found


these


clays,


appears


likely


that


Alachua


Formation


sediments,


particularly


clay


lenses,


are


result


number


chronologically


geographically


isolated


local


events


rather


than


single


chronologically


continuous,


geographically


wide-spread


deposition


event.


Origins


Moss


Acres


Racetrack


aii~te


This


section


addresses


question


sediments


(specifically


clays)


were


laid


down


site,


since


their


lithogenesis


relevance


preservation


vertebrate


bones.


geologist


would


normally


study


geomorphology


surrounding


sedimentary


units


determine


whether


else


sediments


filled


been


a depression


laid


such


down


a graben,


a floodplain,


karst,


channel


such


river


bed.


Unfortunately


sediments


area


Moss


Acres


Racetrack


are


poorly


exposed


not


mapped


geologically


detail,


that


there


virtually


direct


information


regarding


geomorphology


adjoining


units.


Thus,


inferences


regarding


1ithg.ei









clay


mineralogy


nature


condition


preserved


bones


themselves.


mineralogy


Moss


Acres


Racetrack


clays


strong


implications


their


origins.


discussed


above


these


clays


are


nearly


pure,


being


virtually


free


contaminants


such


sand


heavy


minerals


(excluding


minutE

clay


silica


itself


grains,


which


therefore


were


not


probably


formed


contaminant)


with


clays


are


composed


minerals


kaol


nite,


both


which


are


products


chemical


breakdown


other


parent


minerals


suC


feldspar


during


subaeri


weathering.


However,


immediately


unclear


in sirn


whether


or ex sltu


this process

with subseque:


occurred

nt transport


sediments


into


site


(see


below).


Ordinary


fluvial


deposition


Moss


Acres


Racetrack


sediments


contradicted


presence


pockets


peat


unit


Peat


humic


organic


matter


that


accumulates


fresh


water


brackish


water


swamps


bogs


where


stagnant,


anaerobic


conditions


(Boggs,


prevent


1987;


complete


Behrensmeyer


oxidation


bacterial


1992) .


signif


decay

icant


current


flow


into


site


would


have


stirred


water


column an

they were


introduced

considerably


oxygen

deeper


than


bottom

the cur


sediments


rent


unless


level,


allowing


plant


material


completely


decompose.


Further


evidence for


absence


a current


_ the site


crov


ided


f=r; al,


-


\ -


i- -









both


strong


tendency


towards


association


lack


clear


orientation


pattern


(see


chapter


noted


above,


origins


fossiliferous


clay


found


Moss


Acres


Racetrack


are


unclear.


one


hand,


absence


evidence


significant


fluvial


deposition


lack


sedimentary


contaminants


(e.g.,


sand),


presence


small,


pure


silica


grains


suggest


that


clay


formed


n siLu


from


subaerial


weathering


silicate-rich


(e.g. ,


granitic)


rocks.


However,


pointed


Jones


(pers.


comm.)


oxidizing


conditions


necessary


aforementioned


weathering


process


occur


would


surely


have


destroyed


plant


material


(i.e. ,


peat).


Thus,


despite


evidence


against


significant


fluvial


deposition


into


site,


clays


must


have


somehow


been


transported.


Jones


pers.


comm.)


proposed


that


clays


were


not


transported


ordinary


fluvial


fashion


surficial


water,


rather


were


carried


deposited


flowing


water


moving


through


aquifer


, a phenomenon


that


not


particularly


uncommon.


However,


this


any


other


hypothesis


explaining


origins


clays


Moss


acid


Acres


that


Racetrack


quite


must


likely


explain


percolated


source


precipitated


silicic


within


clays


anomaly


form


silica.


consistent


possible


with


explanation


aforementioned


this


hypothesis


that


r~~~~ or rt~r~- nr c 4 mon 1


out


cl r ht C3 r r ~ n ~ ~ n


; tc~lf


ll~t ~r


e ~ rl ; m n n t


1


' r -L-


r









In summary,


origin


Moss


Acres


Racetrack


sediments can


plausibly


explained


following


scenario.


Earlier


Miocene


a sinkhole


formedd


location of


the Moss


Acres


Racetrack


site


(possibly


response to a drop


sea


level),


this


sinkhole


was


filled by water and


sediments


ultimate


source


these


sediments


unclear.


Scott


(1988)


proposed


that


Alachua


Formation sediments


are


residue


weathered


Hawthorn


limestone.


However,


near


complete


absence


here


phosphorite and


heavy


minerals


abundant


Hawthorn,


well


absence


potassium-rich


feldspars


trom


Hawthorn required


form


kaolinite


found


Moss


Acres


Racetrack


(Pirkle et al. ,


1965) ,


renders


this


hypothesis


implausible.


One possible source


sediments


out


which


illite and kaolite could have


formed


granite-rich


Appalachians


to the north,


though presumably


transformation of


these


parent


minerals


into


clay


occurred


prior to their deposition


in Moss


Acres


Racetrack.


During


after this clay deposition a shallow bog or swamp came


exist

lake


water


in the site,


completely


lake succession.


a state that


reached just before


filled during the normal


This


pattern


infilling continued


the

fresh


until


sinkhole had completely disappeared,


lake succession merging with the adjacent


final


stage


topsoil.


role


,~ I -


_ r


* *


rt .,..,


r 'i 1 1'








clay


suggests


that


its origins were different


from those that


first


filled


sinkhole.


Gulf Coastal


Plain fossil


localities other than


those of


very


late


Pleistocene age tend to be very difficult


to date


directly due to the absence or near absence of useful


radioisotopes


in useful


states


(e.g. ,


40K sealed in volcanic


crystal


in sediments of


region.


Strontium,


specifically


the 87Sr/86Sr ratio,


has been used to help


date


some Florida


localit


ies


such


Leisey


site


(Webb


al.,


1989);


however,


strontium isotope dating requires


a site


have an


invertebrate


fauna


with diagenetically unaltered


tests and represent either a marine or else at a


marine/terrestrial


interface


facies


(Elderfield,


1986),


prerequisite that Moss Acres Racetrack clearly


fails.


Thus,


the only viable method for dating the Moss Acres


Racetrack


site

fauna


faunal


biostratigraphic correlation,


dated relative


radioisotopically


in which


dated


faunas


other regions,


primarily the Great


Plains.


facilitate


this


faunal analysis,


list


of all mammalian taxa


in the


Moss


Acres


Racetrack


fauna


identified to genus or below and


their


chronological


ranges are presented in


Table 2-1.


survey of Table 2-1 shows that almost all


the taxa


present


are


restricted to


the Clarendonian and Hemphillian


land mammal


ages,


a chronological


range of


approximately


Acre


In situ











Table 2-1. A list of mammal taxa in the Moss Acres Racetrack
fauna identified to genus or below and their chronological ranges
in terms of land mammal ages.


Taxon


Chronological


Range


Pliometanastes


~RXOfl stu s1


Osteoboru s


Enhvdri therium


early
early


terraenovae


Hemphil
Hemphil


early


early


Amebelodon britti


ear


Aoheloos


-late


early-late


DinohiDous


CaliDDaus
Hioarion


mccartviv


ear
ear


tefionense


hian2


Hemphil
Blancan3
HemDhi 1
Hemphii


lian-


ian4
ian5


e Clarendonian-latest
Hemphillian6
ly Hemphillian7
ly Clarendonian-late
Hemphillian-


Nanni ous


NanniDDus


aztecus


late


Hemph
early


moraani


Clarendonian-
illian9
Hemphilliani


tate


CormohiDDarion


emslei


early


late


CormohinDDaron


Neohiooarion


olicatile


late


tramoasense


late


Hemiauchenia


late


Hemp-1l


Blancanri
Clarendonian-early
Hemphi 1lian11
Clarendonian-early
Hemphi 1lian
Clarendonian-late


Rancholabrea


new


genus


species


near


Pseudoceras


early
early


Pediomervx hemohillensis
Hemphillian13


Hemphillian
-latest


Reference


1969;
1977;


Hulbert,
Hulbert,
13: Webb,


abbreviations-


: Berta


MacFadden,
1988b; 9:
1988c; 12:


* Is


Morgan, 1I
1992, Hu]
MacFadden,


Webb


Webb ar
985; 4
ibert,
1984;


al.


Hirschfeld,


see
1992;
10:


1981,


chapter


see c
Kurten


1983a.


:ulbert,
chapter
and AJ


Webb,


: Bennett,


1988a;

nderson,


8:

1980;


lian-


0.5~


mytiIis


i









radioisotopic dating and stratigraphic correlation


Great


Plains


faunas)


Pliometanastes


considered


diagnostic


the beginning of


the Hemphillian


(Tedford


al.,


1987),


while AmebeloQdon,


ADheloos mutilis,


Neohionarion


trampasense,


and Pediomervx hemohillensis are all wide


ranging taxa


that are restricted to


the Hemphillian


(Cormohiaoarion olicatile


is also restricted to the


Hemphillian,


[Hulbert,


1988b] ) .


is geographically restricted


Florida


Thus the Moss Acres Racetrack can be


clearly assigned a Hemphillian age,


with a range


from 9-4.5


No diagnostic


latest Hemphillian taxa


are


present


Moss Acres Racetrack


fauna


(e.g. ,


Plesioculo,


Fem.


Meaantereon,


Aariotherium),


while neither Cormohionarion


olicatile nor Neohinoarion tramoasense are known


to extend


into


late and/or


latest


Hemphillian


(Schultz,


1977;


Tedford et al. ,


1987) .


Thus the Moss Acres


Racetrack


fauna


must be considered to be early Hemphillian in age,


roughly


1/2 Ma.


Hulbert


(1988a)


concluded that


the Moss


Acres


Racetrack


fauna


the basis of


is older than the very


following:


early Hemphillian on


the morphologically


advanced


state of members of two equid species,


Nannipous aztecus and


Cormohipoarion olicatile;


and 2)


the presence of Anheloos


mutilis,


a species more advanced than the


late


Clarendonian/early


Hemphillian


species


malacorhinus


(see


ml -


1 ~~~ ~ ~ ~ n an r'nr~ or Ini rrto VI\C rcr 1 Fr =, ,n n


D~rn ir3~L


.r-- -


fi -


31Cn rh~n~nr


F~rln~


r^ ^ n









Hemphillian


Florida


sites


Mixon's


Bone


McGeehee,


circa


older


than


such


classic


Great


Plains


late


Hemphillian


faunas


Rhinoceros


Hill


Kansas)


and


Coffee


Ranch


Texas),


circa


approximately


(Ted


ford


ai.,


1987


; Bennett,


1977;


see


chapter


Moss


Acres


Racetrack


fauna


strongly


resembles


another


Marion


County


late


ocene


fauna,


Withlacoochee


River


, sharing


following


taxa:


Pliometanastes


.rotistus


(see


chapter


Os t. .bo rus


orc


(see


chapter


Enhvdrither


lum


terraenovae


(see


chapter


Acheloos


mu ti l


(Becker,


1985),


Cali.ous


m.e cartrv:


(see


chapter


HiDDarion


* ceho nens..e


(Hulbert,


1988b),


Nanni.Dus


aztecus


(Becker,


1985),


Nann iDDU S


mo raan .


(see


chapter


new


genus


species


near


Pseudoceras


(see


chapter


8) *Of


these


taxa,


Osteaoborus


Q-% I


CalioDus


mccartvi,


NanniDous


morcaani


possibly


Pseudoceras-like


animal


are


endemic


these


two


faunas


(see


chapter


This


resemblance


associated


endemi


cism


strongly


suggests


that


these


two


unas


are


o.ose


age


well


location.


presence


ursid


Indarctos


Withlacoochee


River


fauna,


taxon


diagnostic


late


early


Hemphillian


(Tedford


al.,


1987),


supports


a late


early


Hemphillian


age


both


these


faunas.


Withlacoochee


River


fauna


will


discussed


further


future


chapters.















CHAPTER


REDIAGNOSIS


THE


GENUS


AMEBELODON


(PROBOSCIDEA,


GOMPHOTHERIIDAE)
(KONOBELODON)


WITH


BRITTI,


A NEW


FROM


SUBGENUS


THE


MOSS


AND


SPECIES,


ACRES


AMEBELODON


RACETRACK


SITE


shovel-tusked


gomphotheres


represent


one


most


distinctive


proboscidean


groups.


indicated


their


monicker,


shovel-tuskers


are


characterized


broad


flat


lower


tusks


that


resemble


heads


shovels.


shovel-tuskers


include


such


well


known


genera


Amebelodon,


Platvbelodon,


Torvnobelodon.


genus


Amebelodon


been


subject


considerable


Barbour


systematic


(1927)


dispute


been


since


considered


was


both


established


monophyleti


(e.g.


, Gregory,


1945


polyphylet ic


(e.g


Bennett


, 1977)


different


arrive


authors.


a clear


controversy


diagnosis


stems


genus;


from


some


a failure


specimens


referred


Amebelodon


possess


characters


considered


diagnostic


other


shovel-tusker


genera,


Platvbelodon


Torvnobelodon,


have


been


linked


these


genera


some


authors.


Amebelodon


fossil


s described


Gaziry


(1987),


well


numerous


exceptionally


preserved


Amebelodon


fossils


North


from


Florida,


Moss


help


Acres


resolve


Racetrack


this


site


controversy


(Hemphillian)


provide












On the basis


this


rediagnosis


new


subgenus,


Amebelodon


(Konobelodon)


a new


species,


(K.)


are


established.


Measurements


were


made with


following


devices:


calipers,


tree


terminology


fol


calipers,

lows that


and an anthropometer.


Tobien


Dental


{1973)


Svstematics


Order


Proboscidea


Family


Gomphotheriidae


Cabrera,


1929


Genus Amrebelodon


Barbour,


Subgenus Amebelodon


(Amebelodon)


Barbour,


1927


Type species:


Amebelodon


(Amebelodon)


Barbour,


1927.


Included species:


Amebelodon


(A.)


floridanus


Leidy,


1886;


(A.)


hicksi


Cook,


1922;


oaladentatus


Cook,


1922;


(A.)


sinclairi


Barbour,


1930.


Diagnosis:


Trilophodont


intermediate


cheek


teeth


(defined


this paper


DP4,


and M2);


lower


tusks


with


a simple,


laminated


internal structure;


and M3's with


five


fewer


lophs(ids)


Remarks:


The subgenus Amebelodon


(Amebelodon)


encompasses


the species


traditionally referred to


genus


Amebelodon.


shared with typical


Its diagnostic characters are plesiomorphies


gomphotheres such as Gomohotherium.


1927


britti,


ciai









Included


species:


Amebelodon


cvrenaicus


Gaziry,


1987.


Etymology:


3-Qn-Q I


Greek


refers


dentinal


rods,


tapered


form


cones,


found


lower


tusks


this


subgenus.


Diagnosis:


Tetralophodont


second


molar;


permanent


lower


tusks


with


internal


dentinal


rods;


M3 's


with


well


developed


lophs(ids);


lower


tusk


dentinal


rods


with


distinctive


distinct


morphology,


near


relatively


proximal


narrow


tusk,


structurally


expanding


tending


merge


towards


distal


tip;


mandibular


morphology


and


size


resembles


Amebelodon


fricki,


though


generally


somewhat


larger.


Remarks:


subgenus


Amebelodon


(Konobelodon)


includes


those


Amebelodon


specimens


convergent


with


shovel-tusker


genera


Platvheto~don


Torvnobelodon


having


tetralophodont


M2's


(absent


from


holotype


subgenus


type


species,


K.U.M.


present


3852


[Barbour


referred


Hibbard,


material


1941


T.M.M.


; Gregory,


31075-1


1945]


having


dentinal


rods


permanent


lower


tusks


(the


condition


deciduous


lower


tusks


unknown).


Amebelodon


cvrenaicus


included


subgenus


despite


having


plesiomorphic


features:


tusks


with


internal


lamination


trilophodont


(DP4


unknown


this


spec


ies)


(Gaziry,


1987) .


Though


resembling


subgenus


lodon)


these


plesiomorDhic


features,


ne presence


cone,


( ,4m e b e









tetralophodont


second


molar


clearly


place


vr..lena .cus


subgenus


(Konobelodon)


Amebelodon


(Konobelodon)


New


Species


Synonyms:


Amebelodon


(Amebelodon)


fricki


Osborn


Granger,


1931;


Barbour


Etymology:


Hibbard,


Named


1941;
f /" fl


Jerry


Gregory,


Britt,


Jr. ,


1945.


a deceased


graduate


student


University


Florida,


honor


contributions


Florida


paleontology.


Diagnosis:


Internal


lamination


absent


from


powerr


C, -*- 0 Jv S


with


both


cortex


medulla


solidly


,-,~e


with


zent Inal


rods;


tetralophodont.


Holotype:


70000


, a


partial


mandible


lacking


most


symphysis


associated


axis


lower


fragments


vertebrae,


vertebra,


tusks


hyoid


centrum


collection


possessing


apparatus,


only


atlas


indeterminate


Florida


Museum


cervical


Natural


History.


Paratypes:


69994,


intact


upper


tusk


, and


97269,


completely


intact


lower


tusk


with


root


preserved,


collection


Florida


Museum


Natural


History.


Referred


material:


Amebelodon


from


Rhino


Hill


site,


Kansas


(K.U.M.


3852,


reconstructed


palate


with


K.U.M.


3477,


juvenile


palate


with


K.U. M


~nA V i" M rc trTM 'AV ti


Sritti


3Rc;~


IRe;A


Fraci


K TT M


K ii M


\/3 V ti


r71 -^^


anri


< J








History,


University


Kansas)


mandible


from


Parcell


Ranch


Texas


(T.M.M.


31075-1,


Texas


Memorial


Museum


Austin);


isolated


lower


tusk


fragment


from


Texas


(D.M.N.H.


1319,


Denver


Museum


Natural


History);


material


from


Moss


Acres


Racetrack


site


135286,


mandible


with


one


partial


lower


97261,


lower


tusks;


partial


tusk;


69995,


69998,


upper


distal


tusk;


97268,


half


69993,


124373,


upper


isolated


135802,


tusk;


mandibular


symphysis;


three


unprepared


uncatalogued


skulls;


and


assorted


post-cranial


elements


including


limb


bones,


Florida


Museum


Natural


History


Gainesville,


Florida-


see


Appendix


a lower


from


Bone


Valley


region


Central


Florida


95778,


Florida


Museum


Natural


History


Gainesville,


Florida)


DescriDtion


Holcrvoeve


70000


most


diagnostic


holotype


specimens


partial


mandible


(Figure


3-1)


Two


M3's


are


present,


without


even


alveolus


originally


recovered,


indicate

this s


nature


pecimen


was


badly


M2 's.

crushed


When

medio-


laterally;


present


three


dimensional


construction


result

Natural


extensive


History.


restoration


work


Measurements


: the Florida

specimen are


Museum o

presented


Table


3-i.


he symphysis


poorly


preserved.


Most












































Sb~


*
P "4-
48dlbka* ;p


Figure 3-1. Mandible UF 70000,


(Konobelodon)


bratt 1.


(ic osos


ho lot ype
1 r t e r a I


specimen of


view;


Amebe lodon


close- up


left M3,


occlusal


Vi Ct W


i he1










Table


3-1.


Measurements


(millimeters)


partial


mandible


70000,


Amebelodon


(Konobelodon)


britti,


with


comparative


measurements


(Amebelodon)


type


fricki,


specimen


UNSM


Amebelodon


1226.


UF 700000 UNSM 1226
total length: 1384* 2133
length from condyle to the
edge of the symphyseal
ridge: 985
length from coronoid to
condyle: 272
intercondylar width: 555 508
condylar width: 157
height of symphyseal ridge
at proximal edge: 301
maximal width of symphyseal
ridges: 139
maximal height of symphyseal
ridges (from symphyseal
floor): 70
depth from coronoid to
angular process: 375 356
length of third molar: 286 229
width of third molar 90 89
* specimen broken and value not indicative


I










posterior


sections


lower


tusk


alveoli


are


exposed.


medial


alveolar


wall


missing


except


deep


symphys


near


posterior


walls


alveol


tu1SKS


are


largely


missing;


however,


some


tusk


materna


remains


wirhir
W^ ^- ^


alveoli


distinctly


showing


dentinal


rods


internal


structure.


suprasymphy


seal


ridges


differ


from


those


type


mandible


fricki


(UNSM


also


rom


referred


mandible


e (T..M.


31075-1)


high


above


symphyseal


margin


having


distinct


lateral


lare.


right


lower


badly


damaged.


cuss


molars


will


based


large


which


almost


perfectly


preserved.


most


striking


aspects


these


molars


their


extreme


length.


left


molar


long,


considerably


longer


than


lower


type


specimen


(229


, from


Barbour,


1929


upper


numbered


value


from


Barbour


Hibbard,


1941) .


M3's


have


lophids,


sinai


cingula,


small


talonids.


lophids


are


highly


worn,


that


most


detail


morphology


conids


conulids


been


lost;


however,


remaining


enamel


folds


compri


sing


lophids


remain


intact.


ectolophids


have


highly


developed


trefoils.


conid-derived


fold


trefoil


- ,


t~~~~~n~~ .A r. I- l .t r % % V r e- ,44-9,r ,-"


N ---


. / *


*^





rr- /->


-^ V- <-^


b^-


-


t r( n


T V -


*


^-


*3: ne









entolophids


fold


are


broad,


incipiently


medio-laterally


foiled.


conid-derived


compressed,


elongated


antero-posteriorly.


conulid-derived


folds


are


sinai


resulting


poorly


developed


trefoil


pattern.


rest


mandible


this


fricki


partial


size


mandible


resembles


morphology


. This


type


specimen


distinctly


larger


than


type


differences


are


great


shown


Table


-1).


There


strong


resemblance


between


two


specimens


such


features


angular


mandibular


process,


condyle,


horizontal


coronoid


vertical


process,


mandibular


rami.


Paratvne.


229AP


This


specimen


right


upper


tusk


(Figure


-2).


overall


preservation


excellent;


however,


sustained


three


ma or


breaks.


Measurements


specimen


are


presented


Table


3-2.


large


enamel


band


extends


along


lateral


surface


tusk


(orientation


references


will


assume


that


tusk


life


position,


oriented


perpendicular


ground).


enamel


band


present


throughout


length


tusk,


though


preservation


uniform.


largely


intact


over


distal


one-third


tusk,


shows


varying


degrees


post-mortem


stripping


over


proximal


two-thirds


















Lt.


-"* 'S -
-
St


9. .-
- -2~


Figure


3-2.


UF 69994,


right upper tusk,


paratype of Amebelodon


(Konobelodon) britti. A)


lateral


view;


close-up


of the medio-


proximal


wear


f'


facet.










Table 3-2. Measurements (millimeters) of UF 69994, a right
upper tusk of Amebelodon (Konobelodon) britti.

total length: 1570
maximum anteroposterior diameter: 284
width of enamel band: 66
length of wear facet: 566










present


medial


surface


distal


tusk,


extending


medial


thin


edge


from


surface,


tip.


rather


latero-poster


facet


heavily


ior


does


tlat


slanted


surface


termin~g


tusk.


proximal


portion


latero-posterior


facet


surface


directly


enamel


bounded


band


, forming


apparent


cutting


edge.


Where


bounds


facet


enamel


band


highly


facet,


crenulated


with


name


possibly


bordered


result


cutting


abrasion.


edge,


result


wear


caused


stripping


perhaps


vegetation


aspects


cutting


this


behavi


facet


funct


similar


ional


facets


ecological


other


gomphotheres


are


discussed


detail


Lambert,


989;


1992


ParatvDoe


97269


This

preserved


specimen


Figure


s a right


-3).


lower


However,


tusk,

has


exce


lently


sustained


break,


repaired


with


plaster,


from


proximal


tip.


Measurements


specimen


are


presented


Table


3-3.


tusk


extremely


flattened


dorso-ventrally,


with


very


shallow


dorsal


concavity.


lateral


edge


thicker


than


medial


edge,


giving


tusk


a cross-


sectional


profile


similar


type


specimen


cki.


Medio-laterally


tusk


gently


curved


sigmoidally,


*1


1CS


1


I


__,


nr^


~r L Ck C


_


n ir -r ur


C~. v *r


f-d -.


-fc^ -
























































Figure


3-3.


UF 97269,


right lower tusk, paratype of Amebelodon


(Konobelodon) britti.










Table 3-3. Measurements (millimeters) of UF 97269, a right
lower tusk of Amebelodon (Konobelodon) britti.
total length: 1236
maximum width: 175
thickness at inner border: 30
thickness at outer border: 55









well


preserved


root,


which


extremely


flattened


relative


rest


tusk.


dentinal


rods


are


directly


exposed


this


specimen.


However


corrugations,


which


are


more


obvious


proximal


than


distal


regions,


reveal


their


presence.


Skull


Three


Amebelodon britti


skulls


were


recovered


from


Moss


Acres


Racetrack


during


1991


field


season,


first


reasonably


well


preserved


skulls


ever


found


genus


Amebelodon.


Preparation


these


specimens


progress,


that


detailed


preliminary


description


preparation


one


currently


skull


impossible.


reveals


However,


heavily


retracted


nasal


bones,


considerably


more


retracted


than


those


modern


elephants.


This


retraction


indicates


that


Amebelodon


britti,


possibly


Amebelodon


whole,


large,


well


developed


proboscis,


possibly


larger


than


that


modern


elephants


(see


Lambert,


1992


further


discussion)


Post-cranial


Skeleton


s Acres


Racetrack


Amebelodon britti


material


exceptional


containing


a variety


post-cranial


elements,


especially


limb


bones,


from


a single


population.


The


manus


and


are


especially


well


represented,


with


nearly


fully


complete


associated


specimens


known.


Table


provides


9t *


.- r~ -% 4- .w' .a I-mk -'%, -'n- r-


A 4 ~


rnn r nrrr hmnnC


-*


i


I


1


*


~nnhn


h


C


Inn










Table 3-4. Measurements (millimeters) of Amebelodon
(Konobelodon) britti post-cranial elements.
I-- -n -r -n -


69997
left


left


right


radius
total
ulna
total


length


length


: 800


including


olceranon):


scapula


total


length


along


spine:


right


tibia


total


minimum
width


length


m shaft
across


width:
distal


: 744


articulating


width


surface
across
surface


: 265


proximal


articulating


right


fibula


right


total
femur
total
length


length


length
from


-: 679


ball


greater


trochanter:


width
minimum


across


condyles:


mediolateral


shaft


width:


minimum


anteroposterior


width:


shaft


69944
right


pelvis
maximum


length


from


ischium


ileum


ramus


edge:


..._.._._









Limb


height


bone


weight


measurements


this


allow


animal.


estimate


forelimb


original


incomplete


partial


skeleton


69997,


lacking


a humerus;


thus,


shoulder


height


cannot


directly


determined


this


individual.


However,


hindlimb


complete,


allowing


estimate


pelvic


height


approximately


2.3-2


meters.


height


shoulder


was


probably


not


greatly


different.


This


figure


suggests


that


this


animal


was


roughly


similar


height


a modern


Asian


elephant


(EleDhas


maximus)


(Nowak


femoral


Paradise,


regress


(see


1983) .


chapter


However


10) ,


mass,


been


based


estimated


approximately


Columbian


10,000


Mammoth


kilograms,


(Mammuthus


figure


columbi)


size


(Shipman,


range


1992).


Thus,


Azweba~fon britn


have


been


one


largest


mammals


ever


walk


in North


America.


Rediaanosis


Amebelodon


confusion


over


correct


diagnosis


Amebelodon


basis


a failure


many


authors


properly


evaluate


phylogenetic


nature


two


features


considered


diagnostic


genus


many


authors:


internal


structure


lower


tusks,


number


lophs(ids


intermediate


cheek


teeth.


genus


Am elie 1 n 4 n


traditionally


diagnosed


following


characters:


narrow,


extremely


elongated


mandibular


symphysis;









composed


simple


concentric


amina


(see


Figure


3-4A);


intermediate


cheek


teeth


with


three


lophs(


ids)


condition


known


trilophodonty


(Barbour,


lo29;


Gregory,


1945;


Osborn,


1936


; Tobien,


1972


However,


shovel-cusker


lower


material


oeen


referred


Amebe1 odon


genocypic


species,


fric.ki,


basis


overall


similar


despite th

tusks with


dentinal


rods


presence of

n internal


such


following


structure


found


characters:


composed


Platvbelodon


tightly

see Fig


lower

packed


tre


, and


tetralophodont


intermediate


cheek


teeth


SOsborn


Granger,


1931;


Barbour


Hibbard,


1941;


Gregory,


1945) .


This


character


combination


considered


diagnostic


Platvbelodon


Torvnobelodon


rather


than


Amebeiodon


(Gregory,


1945)


Thus,


there


a clear


conflict


proper


diagnosis


Amebelodon.


This


conflict


leads


two


possible


conclusions.


traditional


diagnosis


Amebelodon


involving


lower


tusk


internal


structure


loph


number


intermediate


cheek


teeth


erroneous


Only


mandibular


symphyseal


morphology


superficial


lower


tusk


morphology


can


used


structure


diagnose


loph


number


genus,

divide


while

the


lower

genus


tusk

into


internal

two


distinct


morphotypes,


one


which


convergent


with


Torvnobelodon


and Platvbelodon


(here


placed


subaenus










Figure 3-
different
internal


-4.


Cross-sectional


Amebelodon
structure.


species


views


lower


demonstrating


Amebelodon


(Amebelodon)


tusks


from


differences


sp. ,


showing


simple


lamination


plesiomorphic


gomphotheres;


Amebelodon
lamination
(medulla);


(Konobelodon)


(cortex)


cvrena.icus,


densely


Amebelodon


packed


showing


both


dentinal


(Konobelodon)


internal


rods


britti,


showing


uniform,


densely-packed


dentinal


rods.






























































i "s l_ -.









traditional


diagnosis


Amebelodon


correct.


anomalous


diagnosis


specimens


that


Amebelodon


not


conflict


actually


with


traditional


belong


genus


rather


are


more


closely


related


Torvnobelodon


Platvbelodon,


are


only


convergent


with


Amebelodon


certain


features.


Both


conclusions


have


been


drawn


different


authors,


former


Gregory


(1945)


latter


Bennett


(1977).


Evidence


intermediate


molar


favoring


loph


first


numbers


conclusion


taxa


exists


question.


Platvbelodon


(and


presumably


Torvnobelodon)


are


tetralophodont


their


intermediate


molars,


though


Ml's


some


individuals


have


only


three


lophs(ids)(Osborn


, 1936) .


However,


least


derived


anomalous


Amebelodon


species,


cvrenaicus,


only


tetralophodont


definitely


known


trilophodont


(Gaziry,


1987).


more


derived


anomalous


Amebelodon


species,


britti,


both


tetralophodont


Moreover,


trilophodont


contrast


tetralophodont


DP4's


Platvbelodon/Torvnobelodon


(Mebrate,


1987;


Osborn,


1936) .


This


situation


consistent


with


scenario


convergence


between


two


groups.


Additional


evidence


convergence


between


these


groups


found


internal


structure


lower


tusks


rvrsna i riin


P1 aiirP


AIn Ar~~n~h]r


1nxApn TA


tiir 1ck


i-dA


~h nwc:


rri th


I i I i









Amebelodon


lower


tusk


that


solidly


filled


with


dent inal


rods,


derived


condit


ion


shared


with


Platvbelodon/Torvnobelodon


see


Figure


4---,


However


cvrenaicus,


shown


Figure


3-4B,


revea


a lower


condition


clearly


intermediate


between


those


shown


- gures


3-4A


3-AC.


cortex


simple


lamination


plesiomorphic


Amebelodon


gomphotheres


a whole,


while


medulla


idly


filled


with


dentinal


rods


identical


those


found


derived


lower


tusks


anomal


ous


Ameoelodon


specimens.


Thus,


c'zrenai cus


plausibly


represents


intermediate


stage


between


:ne


two


Amebelodon


morphotypes


This


indicates


that


derived


character


state


shared


between


anomalous


Amebelodon


morphotype


that


Torvnobelodon/Pla


tvbelodon


analogous,


therefore


that


cradiri


ional


diagnosis


Amebelodon


based


this


character


state


erroneous.


Additional


exists


evidence


detailed


analogy


this


morphology


character


dentinal


rods


state


found


two


groups


see


below).


Since


have


shown


traditional


diagnosis


Amebelodon


formulated,


untenable,


requiring


new


a drastic


diagnosis


must


revision


inal


diagnosis


rather


simplification.


genus


Amebel


odon


can


diagnosed


narrow,


extremely


elongated


/ -


,,7t,,nlfllrc,~~ ~ ~ ~ ~ I7 C'n Ir r n r1 ch C, fl rr -nA ,n


.4;t


A~O*le


mhrn Fn rn3n


~1 imn k~ i ~ ; II


- 1_


*3 17\


?rr.


J











Dentinal


Rods.


preceding


section


was


shown


through


evidence


gross


morphology


intermediate


molars


lower


tusk


internal


structure


that


genus


Amebe


lodon


contains


two


morphotypes,


one


convergent


with


Platvbelodon


Torvnobelodon.


This


conclusion


further


supported


differences

individual


between

dentinal


these

rods


groups

within


morphology


their


lower


tusks.


study,


four


Amebelodon


(Konobelodon)


britti


lower


tusk


specimens


from


Moss


Acres


Racetrack


site


and


eleven


Torvnobelodon


lower


tusk


fragments


from


Withlacoochee


site


(North


Florida,


early


late


Hemphillian)


were


examined


(all


specimens


are


from


collection


Florida


Museum


Natural


History)


This


examination


revealed


following


aspects


Torvnobelodon


dentinal


morphology.


rods


are


densely


packed


within


external


shell


tusk.


individual


rods


are


structurally


indistinct,


tend


merge


into


themselves


cross-section


(Figure


-5A)


From


side


view,


dentinal


rods


are


extremely


difficult


resolve,


resembling


solid


dentine.


dentinal


rods


this


same


appearance


examined


specimens,


indicating


uniform


that


dentinal


throughout


lower


tusks


morphology


essentially


Torvnobelodon.


similar


morphology


can De


seen


lower


tusk o










Figure 3-5
Ameebelodon
region, cr
view near
sectional


. Views o
britti l
oss-secti
the tusk
view near


f dentinal
.ower tusks.
onal view;
root; and C
the distal


2t

B)


ods in Torvnobelodon
A) Torvnobelodon, un
) AmebelQdon britti,
Amebelodon britti,
though not distalmos


and
known tusk
lateral
cross-
t end.







52
















A
















B











,%




*I 'it
:f '~ki^^^ ~aL^ ^- ^W^L_
*- | ^' "^^^ ^ / r^^K "' ^
"^Ri~e HL *^- a K,!'^F H-. L ^^ P ^
r' JK^ln .'4 W-t O-









dentinal


rods


Amebelodon


(Konobelodon)


bri ti


differ


from


those


Platvbel


odon/Torvnobelodon


being


morphologically


heterogeneous


throughout


tusks.


proximal


region


tusks,


dentinal


rods


are


extremely


thin


have


clear


boundaries,


being


distinct


cross-


sect


ional


view.


Similarly,


proximal


dentinal


rods


are


visible


lateral


view,


appearing


distinctly


layered


strands

reveals


(Figure

an orde


3-5:


nr y


B) A

change


distal


progression


nature


along


dentinal


tusk

rods;


rods


become


wider


begin


merge,


approaching


condition


found


Platvbelodon


Torvnobeiodon


(Figure


3-5C).


very


distal


tusk,


dentinal


rods


completely


disappear


are


replaced


undif


ferentiated


solid


dentine.


pattern


just


described


britti


lower


tusk


material


from


Moss


Acres


Racetrack


site


also


been


documented


other


(Konobelodon)


lower


tusk


specimens,


notably


those


studied


Gregory


1945)


Gaziry


(1987)


Thus,


morphological


differences


between


individual


dentinal


rods


(Konobelo 1don


Platvbelodon/Torvnobelodon


offer


strong


evidence


their


analogous


nature


addition


that


supplied


lower


tusk


gross


morphology.


However,


premise


dentinal


analogy


these





1


-------


"


--


* tr









Platvbelodon/Torvnobelodon


are


characterized


lower


tusks


that


are


broad


dorso-ventrally


thin,


therefore


presumably


more


vulnerable


breakage


than


deeper,


more


rounded


tusks


typical


gomphotheres.


commonly


proposed


functional


explanation


evo


lution


these


dentinal


rods


involves


internal


bracing


resist


breakage


(e.g.,


Osborn


Granger,


1931) .


this


plausible


untested

surprising,


proposal

a that


adaptations


correct,


shovel-tusked


face


this


then


not


gomphotheres


common


terribly


evolved


selective


parallel


pressure.


However,


question


arises


these


gomphothere


unusual


groups


independently


dentinal


rods,


evolved

adaptation:


a bracing


mechanism

unknown in


any


other


gomphothere


groups.


purely


speculative


answer


this


question


may


nature


mammalian


tooth


development.


tusks


(canines)


walruses


(Odobenidae)


contain


internal


cavity


that


packed


with


distinct


though


irregularly


arranged


pieces


dentine,


known


globular


osteodentine,


that


superficially


similar


appearance


dentinal


rods


Torvnobelodon


Platvbelodon


lower


tusks


(Ray,


1960) .


Walrus


tusks


are


used


grasping


devices


during


locomotion


ice,


must


endure


enormous


stress


(Nowak


Paradise,


1983) .


Presumably


globular


resist


osteodentine


breaking.


adaptation


presence


help


apparently


these


sinitlar


tusks


cent: ina


i *r~


- u









that


potential


structures


above


form


dentinal


distinct


tubule


internal


leveI


aentina


common


mammalian


teeth


genera


Thus,


evo


iution


cent i nal


rods


in A.


( Konobelodon),


Platvbelodon/


To rynobelo don,


similar


separate


dentinal


responses


structures


common


walruses


developmental


represent


systems


common


selective


pressure,


need


prevent


elongated


tooth


breakage


shearing


forces.


evolution


dentinai


rods


tne


Platvbel


o~dmn/T


orvnobelodon


- neage


can


explained


simple


structural


adaptation


linked


evo


lution


extremely


flattened


lower


ta's


(their


tusks


are


relatively


much


thinner


evolution


than


within


those


genus


Amebelodon).

Amebelodon


However


more


their

ficult


explain


, since


species


within


subgenus


Amebelodon


(Amebel


apparently


managed


well


without


dentinal


rods.


A oriori


most


plausible


explanation


evolution


dentinal


rods


(Konobelodon)


that


chan


feeding


habits


involving


lower


tusks


individuals


this


dlade


resulted


greater


lower


tusk


stress.


However,


a study


Lambert


(1989)


Amebelodon


lower


tusk


wear


patterns


their


feeding


habit


implications


found


significant


different


trends


between


(Amebelodon)


Konobel


odon


individuals

behavioral


Within


variationn


both


was


subspecies


area


some


within


individuals


between


(the


species)


J *


odon


L-* h









shaving bark


from


trees


with


tusk.


Thus,


evidence


casts


strong doubt


this


hypothesis.


Another possible explanation


could


involve


simple


structural


engineering such


was


proposed


Platvbelodon and


Torvnobelodon above.


change in lower tusk


size


and/or


shape in members of the


resulted in a need for


(Konobelodon)


clade could have


increased internal bracing.


However,


the lower tusks of some


(Konobelodon)


individuals


are


similar


even


smaller


size


than


those


(A.)


fricki


which lacks dentinal


rods


(Gregory,


1945) .


Moreover,


in cross-sectional profile the


lower tusks of


(Konobelodon)


individuals strongly resemble those of


(A.)


fricki.


Thus


the evidence fails to support


this


hypothesis


well.


I have no other plausible explanation


for now the evolution of dentinal


rods


in the


(Konobelodon)


lineage remains an enigma.


Bioaeoaraohv


and Chronoloav


Known records of


the subgenus Amebelodon


(Konobelodon)


are quite circumscribed both biogeographically


chronologically.


The earliest


record of


the subgenus


(Konobelodon)


species


(K.)


britti


from Moss


Acres


Racetrack,


which


late early Hemphillian


age,


about


seven


'Jfz rczj


-- n


t -- i -


?Ffi ii r


Am~}m; I- c-; nn


offer,


-'1 1n


ry
A


>, -


i


rlY:









Texas.


Tedford


{1987)


gave


matter


fauna


early


late


Hemphillian


age,


about


million


years


old.


britti


also


known


from


tne


Bone


Valley


region


central


Florida.


Texas


specimens


occur


both.


from


unknown


(Osborn


Granger,


1931),


Panhandle


(Grecory,


1945)


None


these


latter


specimens


suff ic:


biostratigraphic


record


provide


accurate


:aunal


correlation,


though


each


instance


a Hemphillian


most


likely.


North


America,


Amebe odon


became


ext Ic Z


S-,rE


-ate


Hemphill


latest


VMi OCene)i


along


with


.mu


Clarendonian


ungulate


chronofauna


(terminology


Webb,


1983).


Amehe.lo on


unknown


from


faunas


younger


than


that


Coffee


Ranch


in North


America.


World,


(Konobelodon)


known


only


species


cvrenaicus


from


Sahabi


fauna


North


Africa,


believed


earliest


Pliocene


(Ruscinian)


(Gaziry,


1987;


Tobien,


1986)


This


record


cvrena1cus


represents


youngest


record


Amebelodon


known.


Overall,


genus


Ameoeb odon


rare


World,


known


only


other


isolated


lower


tusk


specimens,


one


from


Turkey


other


from


China.


Both


specimens


belong


subgenus


( Ameoe lodo r i


with


one


referred


fricki


(Tobien,


1986).


-Ite


ent

















CHAPTER


THE


OSTEOLOGY,


PALEOECOLOGY,


AND


PHYLOGENY


THE


GIANT


OTTER


ENHYDRITHERIUM TERRAENOVAE


otters,


both


living


extinct,


can


placed


into


two


distinct


groups


based


relative


robustness


their


carnassial


cusps.


One


these


groups


includes


most


living


Lutra-like


genera


well


extinct


genus


Satherium,


characterized


carnassials


with


relatively


slender


cusps


that


tend


form


blades,


especially


the paracone

These otters


and

were


metacone

loosely


upper


named


fourth


"fish-eatir


premolar.

ng" otters


Repenning


(1976)


other


group


characterized


carnassials


with


relatively


thick


enamel


bulbous


cusps


that


show


tendency


merge


into


blades.


This


group,


including


living


genus


Enhvdra


well


extinct


genera


Enhvdriodon


Enhvdritherium,


was


loosely


named


"crab-eating"


otters


Repenning


(197


reference


presumed


durophagous


diet


members


(henceforth


this


paper


when


referring


this


group,


morphologically


descriptive


term


bunodont


will


replace


term


"crab-


eating"


extinct


bunodont


otters


can


divided


into


distinct


-4 tv,- I -a ILi-. 1 4- -- rrL. -l

I -.rr~~i


-, -.
-, I -


-*


i


^- *-- -- *- T-


T J


1,,1 ~j


**










Sivaonvx


relationship


Enhvdriodon


unclear,


will


not


further


considered


here;


see


Colbert,


1935


and


Repenning,


1976) .


one


time


New


World


bunodont.


otters


were


referred


Enhvdriodon


well,


single


species


Iluecai.


However,


Berta


Morgan


removed


them


from


Enhvdriaod n


placed


them


into


a new


monospecific


genus,


Enhvdritherium


terraenovae.


Enhvdriodon


represented


a reasonably


good


assortment


cranial


postcrani


mater als,


though


one


species


particularly


well


known.


-I, h nrrnr'rnr,


s ivalensis


known


a well


preserved


sKUll,


no iot


ype,


though


postcranial


specimens


have


been


assi


gned


this


species


(Colbert,


1935) .


contrast,


limbs


are


represented


dubious


species


latiDes


described


Pilgrim


been


(1931),


referred


though


this


dental


species.


cranial


Most


elements


other


have


species


Enhydriodo n


are


known


basis


mandibular


material


largely


consisting


first


molar,


though


some


postcranial


elements


have


been


assigned


africanus


(Repenning,


1976;


Hendey,


1978).


contrast


Enhvdriodon,


genus


Enhvdritherium


known


literature


basis


extremely


limited


material.


Berta


Morgan


(1985)


based


their


diagnosis


Enhvdritherium


terranovae


small


number


mandibular


A-


I -*


-










insufficient


to clarify details of


its osteology,


autecology,


or phylogeny.


Over the years since the publication of


Berta and Morgan,


the paper by


the Florida Museum of Natural History has


accumulated a


fairly


extensive collection


additional


Enhvdritherium material.


The majority of this material


terms


individual animals


represented)


consists


mandibular elements


(though a partial


ulna has also been


found)


from the Bone


Valley region of Central


Florida


(latest


Hemphillian)


far greater


importance,


however,


partial


skeleton of Enhvdritherium terraenovae that was


found


the Moss Acres Racetrack site.


This specimen


, bearing the


number UF


100000,


surpasses all


other documented bunodont


otter


specimens


in the


literature,


including


a reasonably


intact


skull,


an intact mandible,


a significant


portion of the axial


column,


virtually


complete


fore-


hindlimbs.


This


Enhvdritherium terraenovae partial


skeleton


provides


impetus


this paper.


The purposes of


this


study


are


investigate


the osteology,


autecology,


and


phylogeny of


this species


in a level


of detail previously


impossible.


Specifically,


will


undertake


following


three


tasks:


on the basis of


this new specimen,


describe the


- S ~~~~- ,- L


-I


f-


* "


j


n 1 ? I i i. I .. L ...~~









analyze


depositional


environments


Enhvdritherium


occurrences


therefore


determine


their


likely


Lha.UiLQ


preferences;


interpreted


osteological


features


Enhydri other i um


with respect


their


adaptive


significance


comparison


with


other


otters;


reevaluate


phylogeny


this


taxon.


Methods


examining


osteology,


hyloce ny


auctecology


Enhvdritherium


- ~ r ra e m ra~


toliowing


taxa


specimens


were used for


comparison:


Amblonvx cinerea:


UF(M)


14791


Enhdra lutris:


UF(M)


24196,


UF(M)


22397,


UF(M)


24197


UF(M)


24922


Lutra canadensis:


UF(M)


22598


Pteronura brasiliensis:


USNM(M)


304663


Satherium Discinaria:


UM-V 6187


Enhvdriodon sivalensis:


based on


the verbal


diagnosis


aiven


for Enhydriodon in Pilgrim


(1931)


(skull


only)


Enhvdri odon latiDes:


based on verbal


descriptions


drawings


from Pilgrim


(1931)


(limb elements


only)


Mustela vision:


UF(M)


Za lohus californianus:


UF(M)


24910


C -ir.,n n 'r 1 ,rc cAL'r1~ A 1z14r~ c


~n 3 dr rnrj


I.r; ih


13; 31


'*31; nRrC


r.TtT V-,-









Osteoloav


Skull


skull


this


partial


skeleton


was


badly


crushed


when


discovered


from


field,


required


extensive


reconstruction


preparation


excluding


basicraniu;


study.


good


maxillary

m were ci


dermal


condition.


region


:ushed


roof


bones


Unfortunately


toothrow


beyond


repair,


survived


both


, and


limiting


palate,


most

the


detail

dorsal


with

and


which

lateral


cranial


dermal


anatomy


bones


can


discerned.


cranium


rostrum


are


essentially


intact


(Figures


through


Cranial


measurements


are


presented


Table


4-1.


lateral


profile


skull


generally


resembles


that


Lutra.


cranium


very


flat,


with


shallow


sagittal


crest.


facial


region


nearly


level


with


cranium,


similar


while


that


rostrum


This


gentle


differs


from


slope


condition


observed


Enhvdriodon


sivalensis,


which


a highly


swollen


cranium


similar


that


Aonvx


(Pilgrim,


1931),


Enhvdra,


which


a steeply


sloped


rostrum.


enlarged


triangular


jugal


process


(suborbital


present


process)


Similar


dorsal surface

processes were


observed


same


location


other


examined


otters,


these


were


little


more


than


small,


rounded


bumps.


dorsal


view


cranium


shows


a distinctly


rounded


J-


Lutra.


, J


.%.


*-- 4











Table 4-1. Cranio-mandibular measurements millimetersi of
the Enhydritherium terraenovae specimen UF I00000.

Skull


length
width


from


across


occipital
Mandible


occipital
mastoid


height


from


condyle
process
condyle


premaxilla


nuchal


crest


length


width


from


across


condyle


corono


sympnys is


processes


height


coronoid


ramus


process


from


ventral


border


Dentition


upper
upper

upper


length,
length,

width,


abial
ingua


anterior


border
border

border


upper


upper
upper

upper


upper


width,


length,
length,

width,


width,


posterior


labial
lingua


anterior


posterior


border


border
border

border


border


lower
lower


length
width,


down


midl


trigonid


me


15.9


(metaconid-protoconid)


lower


width,


talonid


~~___



























































Figure


4-1.


Enhydritherium skull,


lateral


view.















C-




































t if





I* if if I ~~










;. I I
)*
i


if I, i r



It-.t


~ i *~ i f~





if if


49w'
*~ if.

*: S; *






111




*






U 4 i f I

I S


'4 14if

'S. if i

:- IC,
if


*6 ,.~w~Z..* *j.. -
if .i f f *,4-- -i


itt






'4lr"













(the


frontals


are


particularly


constricted


Pteronura)


Lambdoidal


crests


are


completely


absent,


supraorbital


processes


are


represented


only


inconspicuous


bumps;


both


these


mustelids


features


(Neal,


are


1977)


known


The


vary


length


ontogenetically


height


sagittal


crest


similar


that


Lutra.


front


view,


skull


this


specimen


resembles


that


Lutra.


not


within


external


greatly


nasal


nareal


different


region,


opening


from


such


somewhat


that


turbinals,


broken,


bones


are


compi


etely


absent.


infraorbital


foramina


are


large


proportionate


those


Lutra


Enhvdra.


premaxil


only


moderately


preserved.


ventral


region


skull


poorly


preserved,


with


only


portion


maxilla


containing


toothrow


good


condition.


medial


portion


maxilla,


pterygoids,


virtually


bones


ventral


floor


braincase


are


absent


except


floor


right


auditory


bulla,


uninflated,


which


compares


largely intact.

favorably with


Relatively


that


Enhvdra


auditory


differs


bulla


greatly


Lutra.


from


opening


thin-walled,


into


inflated


antero-medial


corner


auditory


bulla


presumably


remnant


Lutra


t r


r F


1









anterior


observed


wall


auditory


in Enhvdra,


Lutra.


bulla,


condition


and Amnblonvx.


In both


also


Pteronura


Satherium


external


meatus


even


with


middle


auditory


bulla,


while


Enhvdr i odon


sivalensis


external


meatus


positioned


even


with


posterior


wall


auditory


bulla,


condition


opposite


that


observed


Enhvdritherium


specimen


(Pilgrim,


1931) .


glenoid


post-glenoid


walls,


ossae

are


squamosals,


essentially


intact


including


though


there


some


damage.


comparison


possible,


glenoid


region


resembles


that


typical


otters.


mastoid


process


robust


distinctly


elongated,


contrasting


relatively


short,


poorly


developed


mastoid


processes


typical


mustelids.


Most


basicranial


foramina


this


specimen


have


been


destroyed.


only


surviving


basicranial


foramen


this


skull,


foramen


postglenoideum


located


immediately


behind


post-glenoid


wall


glenoid


fossa,


resembles


that


Lutra


in both


shape


location.


occipital


region


skull


moderately


well


preserved,


though


a large


present.


occipital


condyles


are


perfectly


preserved,


show


special


features.


nuchal


crest


highly


developed,


both


more


robust


proportionately


taller


than


that


other


otter,


addition,


there


a deep


fossa


nuchal


region


that


contrasts


near ly


fTiat


nuirhal


rz 6-,nn


.


.. -









that


Enhvdritherium


t erraenovae


possessed


neck


muscles


more powerful


ability


than


chose


forcefully


turn


any

the


otter,


.& in


head


and/or


implying


resist


turning.(see


below).


Mandible


The mandible


excellently


preserved,


with


ofiy


significantly


region


breakage


having


occurred


relative


symphyseal


length


mandible


similar


that


Lutra,


much


greater


than


that


Enhvdra.


mandible


moderately


rocust,


more


tnan


Lutra and the other


fish-eating


otters


much


less


tnan


in Enhvdra.


From


dorsal


view,


mandibular


arcade


typical


mustelid profile.


coronoid


process


broad


untaperea,


with


a distal


edge


that


rounded


into


symmetrical,


distinctly


semicircular


profile.


This


contrasts


the condition observed


fish-eating


otters,


which


the coronoid process

Like Enhvdritherium,


tapered


Enhvdra


with


a has


a narrow


broad


distal


coronoid


edge.


process,


though


is distinctly


irregular


profile.


masseteric


fossa of Enhvdritherium


relatively deep,


with


long


axis oriented parallel


to the toothrow and extending


point below the anterior end of


the m2,


while ventrally


mandibular ramus


bordered by


large,


robust


sheif


tnvt i,.n 4n ri i


1 nrv 1 1


T T 3 rr r h t


* r 1


r~r j2C C t NV.tC -


(Figure 4-4)


C,,,-
































e: ./. .
*- -
-^ f -- C

--' '- \



j *t'


es
*



"if
-'
. e* *




--






*t '
.
*

.


4 i f.






- if *


.- -
S.
I--


if


i
if -
* .* -i


4 4


* -a


,,


r' *
(r


'.J ^









shallow,


bordered


steeply


slanted


ventrally


relative


very


small


toothrow,


shelf


and


mandibular


ramus,


a condition


only


shared


with


Ambi


onvx.


(Berta


Morgan


1985]


erroneously


attributed


deep


masseteric


fossa


with


distinct


ventral


shelf


Enhvdra


, considering


synapomorphy


linking


Enhvdra


Enhvdri


therium.


mandibular


condyles


are


well


preserved,


not


sign


ficantly


different


from


those


typical


otters.


Measurements


mandible


are


listed


Table


4-1.


Dentition


Incisors.


Only


are


present.


their


overall


appearance


they


resemble


13's


such


extant


otters


Lu tr~a.


Pteronura,


Enhvdra


, except


being


relatively


longer


more


robust


(see


Figures


-3).


Pilgrim


(1931


described


robust


Enhvdri


odon


sivalensis.


Canines.


four


canines


are


present


this


specimen,


upper


canines


being


well


preserved


lowers


being


poor


condition.


upper


canines


are


extremely


robust


(see


Figures


and


4-3) ,


much


more


than


in any


examined


otters


feature


shared


with


Enhvdriodon


(Pilgrim,


1931).


Aside


from


their


robustness,


however,


they


resemble


upper


canines


fish-eating


otters


differ


from


shorter


, more


slender


canines


Enhvdra.


lower


C =i fln ir a


(r


r5 nV -


nnr'r~rlW


V


c...


- -


; iknr;l~m


*IYnnn~. r~r~


Cn k~ tri r









are much shorter than their


opposing uppers,


though


they are


similarly


robust.


Premolars


excludinga


p4).


Both the upper and


lower


premolars are reasonably well


preserved


this


specimen.


Both the upper and lower premolar

(P4 will be discussed separately


series

below),


range from P/p2-4

with P/pl absent


typical


approximate


otters.


size


and similar in


badly


still


size


to P3


fractured,


discernible.


(contrasting Enhydra,


though its


is double-rooted,


in which P2


vestigial)


The P3


reasonably intact;


like P2


double-rooted,


possess


a distinct


posterior


cingulum.


The lower premolars are similar


size,


and all


possess


well


developed posterior cingula


(like P2,


is vestigial


Enhvdra).


The p4 of the Moss


Acres


Ractrack specimen differs


from that of all


of the examined fish-eating otters and


Enhvdra


in lacking an


accessory


cusp


(in Satherium two


accessory


cusps are present)


However this absence


evidently the


result


intraspecific variation;


accessory


cusps are present on an Enhvdritherium


terraenovae


mandible,


UF 125000,


recovered from the Whidden Creek Local


Fauna of


the Bone


Valley region.


Carnassials.


The carnassials


(P4 and ml)


this


specimen comply with the diagnosis


Enhvdritherium


terraenovae given by Berta and Morgan 1985),


who along with


with Reoennin(


(1Q7 g;


carnassial


moroholoqv


Av u- -


I


ri ~


,


-


d \ ( .









following


features:


transversely


broad


with


a roughly


quadrangular


profile;


paracone


metacone


separate;


deeo


valley


than


between


hypoconal 1


paracone


crest


ana


present;


protocone;


hypocone


nypocone


rather


relatively


close


protocone.


characterized


short


length


with


rectangular


occlusal


outline;


large


metaconid


posit


ioned


opposite


protoconid;


metastylid


beyond


metaconid;


talonid


broad,


with


shallow


basin


posterior


margin


squared


off;


entoconid


a low


ridge


that


meets


hypoconid


talonid


r 'ge


-c~usOs


are


-nfla e,


relative


those


Lutra,


though


less


than


those


Enhvdra


(see


Figures


4-3,


4-14).


carnass ia s


this


specimen


are


reasonably


intact,


though


they


show


evidence


great


masticatory


wear,


especial


Size


measurements


tinoer


both


first


and ml


molar.


are


presented


superficially


Table


resembles


that


Lutra


(see


Figure


-3).


roughly


trapezoidal


lusal

the


profile,

lingual


with


edge


broader


long

-thar


axis

ithe


oriented

labial


labio-lingually


edge.


L~utxr.


a protoloph,


composed


both


protocone


paraconule,


extends


from


middle


lingual


edge


through


interior


middle


anterior


edge.


protoloph


occludes


with


posterior


surfaces


- ,, -.-,A, -n-,4


.-% 4 A, t~ AA,. t 9 A- .- -.I- .


--


---


," ,,,


ChnnM:










possessing


distinct


hypocone.


paracone


metacone


are


roughly


equal


size,


a small


metaconule


present.


Lower


second


molar.


this


specimen


conforms


important


respects


that


described


Enhvdritherium


terraenovae


Berta


Morgan


1985) .


Axial


Skeleton


Except


skull,


axial


skeleton


this


specimen


poorly


represented.


following


elements


are


associated


with


skeleton:


atlas


vertebra,


axis


vertebra,


two


cervical


vertebrae,


four


thoracic


vertebrae,


three


lumbar


vertebrae,


sacrum,


five


caudal


vertebrae,


nine


indeterminate


comparison


significantly


isolated


actually

different


vertebral


possible,


from


their


central.


none


extent


these


equivalents


that


specimens


extant


otters.


Limbs


Both


represented


forelimb


this


hindlimb


specimen,


are


with


excellently


hindlimb


elements


represented


(except


possibly


fifth


metatarsal)


forelimb


elements


except


trapezoid.


thorough


descriptions


and


drawings


limb


elements


Enhvdriodon


species


latioes


Pilgrim


(1931)


allow


many


-,~~ ,


'I *


*^


rr 1


I1











Table 4-2. Select limb element measurements (in millimeters)
of Enhvdritherium terraenovae specimen UP 100000

Humerus


length f
distance


rom


head


across


trochlea


epicondy


Radius


length


from


head


style


process


Ulna


length


from


olecranon


styloid


process


Femur


length


distance
distance


from


ball


across
from b


to condy
condyles


'all


greater


trochanter


Tibia


length


distance


from


proximal


across


distal


medial


articulating


malleolus


surface


Fibula


length


rom


proximal


medial


malleoius









Scanula.


scapula


virtually


absent


from


this


specimen,


represented


only


intact


left


glenoid


region


with


a small


part


scapular


spine.


comparison


possible,


this


specimen


not


significantly


different


from


scapulae


any


observed


extant


otters.


Humerus.


left


humerus


well


preserved,


while


right


humerus


only


represented


intact


distal


(Figure


thickness


4-5).


humeral


that


shaft


femur,


contrasting


Enhvdra


e Tu a


which


femoral


shaft


thicker


than


that


its


humerus.


greater


tubercle


well


developed,


deltoid


pectoral


ridges


are


both


hypertrophied


merge


form


single


platelike


ridge;


these


features


features


are


much


more


poorly


developed


L12-La


virtually


absent


Enhvdra.


deltoid


tuberosity


prominent,


deflected


laterally.


trochlea


small


(the


facet


separates


presence/absence


entepicondyle


this


facet


from


varies


considerably


among


otters)


RadIiLs.


Only


left


radius


fully


preserved


right


radius


represented


only


proximal


(Figure


4-5)


does


differ


significantly


from


radii


other


observed


otters.


lzna.


4-5).


Both


left


ulna


right


this


ulnae


specimen


are


shows


well


some


preserved


distinct


a mi rn ; 1 r 1


tVm~h


Pnbtrclrs


j ikn


WnhnvArA


nri rnnnnn


approximately


(Figure


r i


; S


r )










Figure 4-5.


Forelimb


elements


Enhvdritherium.


left


humerus,


anterior view;


left


radius,


anterior


view;


right


ulna,


lateral


view.








78



















C

















i






~' I










unlike


Enhvdra


coronoid


process


Enhvdrit her ium


replaced


a relatively


large,


well


developed


facet.


ulna


relatively


deep


antero-posteriorly,


and


medial


groove


quite


shallow.


styloid


process


well


developed.


Scaoholunar.


Both


left


right


scapholunars


are


represented


(Figure


4-6).


scapholunar


disproportionately


large


compared


Enhydra,


animal


roughly


deep


similar

compared


size.


that


unc form


Enhvdra


facet :

Lutra.


especially

However, i


overall


appearance


scapholunars


these


otherwise


otters.


quite


similar


scapholunar


Enhvdriodon


unknown.


Unciform.


Both


left


right


unciforms


are


present


(Figure


4-6).


uniform


Enhvdri theriumr


not


significantly


different


from


that


Enhv


dra.


uniform


Enhvdriodon


differs


greatly


from


that


Enhvdritherium


other


extant


otters


however,


being


tapered


antero-posteriorly


rather


than


basically


oval


profile


having


posterior


that


broader


than


anterior


end,


with


distinct


medial


constriction.


Maanum.


Only


left


magnum


represented


(Figure


lateral


portion


missing,


detailed


comparison


4 -~~~ L'A .


.__


Lutra


i


.i


L L


r Z.











Figure 4-6.
scapholunar,


Carpal elements of Enhvdritherium.


anterior view;


left magnum,


left


posterior view;


right


cuneiform,


posterior view;


right


cuneiform,


anterior view;


right


trapezius,


dorsal


view;


left


uniform,


posterior view.
















I .






*t. ~ a *









A; B











C~I 1 '; '~ ~ ~i
$ E D








I ,
: '~**


I* U t?.2


1 ~*, i -* 4









Cuneiform.


Both


left


right


cuneiforms


are


represented


(Figure


4-6)


This


cuneiform


very


similar


that


both


Lutra


and Enhvdra.


cuneiform


Enhvdriodon


unknown.


Tranezius.


Only


right


trapezius


present


(Figure


4-6).


does


not


differ


significantly


from


trapezius


Enhydra,


while


trapezius


Enhvdriodon


adequately


pictured


Pilgrim


(1931)


allow


a comparison.


Metacarnals.


metacarpals


are


preserved,


including


left


right


left


III,


right


left


and


right


metacarpal


pictured


Figure


4-7.


Their


morphology


similar


that


Enhvdriodon,


their


proportions


similar


those


Lutra.


contrast,


metacarpals


Enhvdra


are


extremely


short,


almost


stumpy


comparison.


Pelvis.


pelvis


present,


poorly


preserved


allow


any


comparisons


with


pelves


other


otters.


Femur.


Only


left


femur


represented.


femur


this


specimen


significantly


different


from


femora


observed


fish-eating


otters


beyond


greater


size


(Figure


4-8)


However,


femur


Enhvdra


differs


greatly


from


femora


other


otters


examined,


including


Enhydritherium,


in being


highly


robust


short


relative


humerus,


having


greatly


enlarged


trochanters,


h ,rrinar


h vn f r rtnnh i pd


mP ri a 1


ronndvi


r-


,-i ilr ]








83









.A







:4






Ik
$









"II



i'I

4 id
L" ":
II;j t


i' :













described


in Berta and Morgan


(1985)


considerably


from


smaller


tibiae


than


other


described


observed


specimen.


OuTt erlS


differs


naving


greatly reduced medial


malleolus


a poorly


developed


tibial


crest.


Fibula.


Only


left


fibula


represented


(Figure


This specimen


differs


from


fibulae


other


examined otters


in being unflattened and relatively


robust.


In addition,


muscle attachment,


poorly


and an extra


lateral malleolus absent


Astraaalus.


from the


Both the left a


developed


process

other

Lnd righ


crests


adjacent t

observed ot

t astragali


ters.

are


preserved


(Figure 4-9)


The astragalus


this


specimen


disproportionately

The neck is distill


small


nctly


relative


abbreviated


those


(this


other


feature


otters.


varies


considerably among the otters,


Enhvdriodon


example


having


a relatively


astragalus of


elongated neck)


other


Enhvdritherium resembles


that


respects


the other


observed


otters.


Calcaneum.


Both the right and left


calcanei are


preserved


(Figure 4-10)


However,


the right


calcaneum shows


coarse extra growth

the right ankle of


h indicating a pathological


the animal.


condition


This condition appears


many of


the right


tarsals of this specimen and thus,


when


t ~ough





























































Figure


4-10.


dorsal


Left


calcaneum of Enhydritherium.


ventral


view.


view;










tuber,


condition


contrasting


other


otters


observed


including


Enhydriodon.


trait


shared


with


Enhvdriodon,


absent


from


other


otters,


absence


latero -


anterior


astragalar


articulating


facet


dorsal


surface.


Cuboid.


Both


right


left


cuboids


are


present


(Figure


4-11).


cuboid


very


similar


that


Enhvdr iodon


except


that


possesses


extra


process


ventral


surface


articulating


with


calcaneum.


overall


resemblance


very


close.


Navicular.


Both


left


right


naviculars


are


present


(Figure


4-11)


Compared


Enhvdriodon,


navicular

prominent


this


process


specimen


more


slender,


antero-medial


edge


there


absent


from


Enhvdriodon.


Ectocuneiform.


Both


left


right


ectocuneiforms


are


preserved


(Figure


4-11).


ectocuneiform


this


specimen


almost


identical


that


Enhvdriodon,


with


one


difference


being


small


process


dorsomedial


surface.


ectocuneiforms


both


taxa


broadly


resemble


those


Lutra


Enhvdra.


Mesocuneiform.


Both


left


right


mesocuneiforms


are


present


(Figure


4-11).


mesocunei forms


Enhvdritherium


and Enhydriodon


are


virtually


identical,


which


turn


are


almost


identical


that


Lutra.


Enhvdra





Lutra










Figure 4-11.


Tarsal


elements


Enhvdritherium.


left


cuboid,


anterior view;


left


ectocuneiform,


anterior


view;


left mesocuneiform,


anterior


view;


right


entocuneiform,


posterior view;


left


navicular,


posterior


view.







90



A
A.


.: '. .






4e.













D*
*- A j it ,.
','




"" '.





'.': ..
Ii a
D '


..r E ~;;r