University of Arizona

QUATERNARY STRATIGRAPHY, GEOCHRONOLOGY, AND CARBON ISOTOPE GEOLOGY OF ALLUVIAL DEPOSITS IN THE TEXAS PANHANDLE (RADIOCARBON).

Item Type text; Dissertation-Reproduction (electronic)

Authors STAFFORD, THOMAS WIER, JR.

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 04/10/2021 01:44:26

Link to Item http://hdl.handle.net/10150/187755 INFORMATION TO USERS

This reproduction was made from a copy of a document sent to us for microfilming. While the most advanced technology has been used to photograph and reproduce this document, the quality of the reproduction is heavily dependent upon the quality of the material submitted.

The following explanation of techniques is provided to help clarify markings or notations which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure complete continuity.

2. When an image on the film is obliterated with a round black mark, it is an indication of either blurred copy because of movement during exposure, duplicate copy, or copyrighted materials that should not have been filmed. For blurred pages, a good image of the page can be found in the adjacent frame. If copyrighted materials were deleted, a target note will appear listing the pages in the adjacent frame.

3. When a map, drawing or chart, etc., is part of the material being photographed, a definite method of "sectioning" the material has been followed. It is customary to begin filming at the upper left hand comer of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again-beginning below the first row and continuing on until complete.

4. For illustrations that cannot be satisfactorily reproduced by xerographic means, photographic prints can be purchased at additional cost and inserted into your xerographic copy. These prints are available upon request from the Dissertations Customer Services Department.

5. Some pages in any document may have indistinct print. In all cases the best available copy has been filmed.

University MicrOfilms International 300 N. Zeeb Road Ann Arbor, MI481 06

8424911

Stafford, Thomas Wier, Jr.

QUATERNARY STRATIGRAPHY, GEOCHRONOLOGY, AND CARBON ISOTOPE GEOLOGY OF ALLUVIAL DEPOSITS IN THE T~XAS PANHANDLE

The University of Arizona PH.D. 1984

University Microfilms International 300 N. Zeeb Road, Ann Arbor, MI48106

PLEASE NOTE:

In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark __-/_.

1. Glossy photographs or pages __

2. Colored illustrations, paper or print __

3. Photographs with dark background __

4. Illustrations are poor copy __

5. Pages with black marks, not original copy __

6. Print shows through as there is text on both sides of page __

7. Indistinct~ broken or small print on several pages ~

8. Print exceeds margin requirements __

9. Tightly bound copy with print lost in spine __

10. Computer printout pages with indistinct print __

11. Page(s) lacking when material received, and not available from school or author.

12. Page(s) seem to be missing in numbering only as text follows.

13. Two pages numbered . Text follows.

14. Curling and wrinkled pages __

15. ~her______

University Microfilms International

QUATERNARY STRATIGRAPHY, GEOCHRONOLOGY, AND

CARBON ISOTOPE GEOLOGY OF ALLUVIAL DEPOSITS

IN THE TEXAS PANHANDLE

Thomas Wier Stafford, Jr.

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF GEOSCIENCES

In Partial Fulfillment of the Requirements . For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

1 9 8 4 THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Final E~amination Committee, we certify that we have read

the dissertation prepared by ____T_h_o_m __ a_s __ W_i_e_r ___ S_t_a_f_f __ o_r_d~,~J_r~. ______

entitled ___Q~u_a_t_e __ r_n_a_r~y~s __ t_r_a_t_1~'g~r~a~p~h~y~,~~g~e~o~c~h~r~o~n~o~l~o~g~y~,~a~n~d~c~a~r~b~o~n~ __ __ isotope geology of alluvial deposits in the Texas Panhandle

and recommend that it be accepted as fulfilling the dissertation requirement Ph.D. for the Degree of ------

Date

CC:~c:s:~:.-.s.;;~.. ---"' .. ~;::.."..;~~··2::;:?··"... h. ~~.---.) Date . J

Pa"J. J Date I

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement •

D1s. ~~~er a 10n D1rec or Date STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for xtended quotation from or repro duction of this manuscript in whole r in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the propose use of the material is in the in- terests of scholarship. In all othe however, permissio must be obtained from the author. ACKNOWLEDGMENTS

Many people have aided me personally and financially during my graduate studies, and it is a pleasure to thank them for their generos ity. Dr. C. Vance Haynes, Jr. has supported by education from the very beginning through teaching and research positions. His generous latitude toward research gave me the opportunity to pursue my goals and it has contributed immeasurably to this dissertation. The following

National Science Foundation grants to Dr. Haynes partially or fully supported my work: EAR 77-10109 A01 and EAR 7026362. Dr. Klaus

Brendel generously gave his time, laboratory space, and equipment for the last two years while I completed the chemical aspects of this work; his guidance and friendship during that time were a major contribution to my success. Funding to Dr. Brendel was from National Institutes of

Health Grant 5-ROI-EY-1868-05. Dr. Austin Long gave me the oppor tunity to construct a laboratory for bone 14C dating based on experiments in this dissertation; the laboratory, funded by National

Science Foundation Grant BNS-8211864, was used to complete the amino acid 14C dating aspects. Dr. Paul M. Martin first helped me through the University's impersonal corridors, provided early training at

Tumamoc Hill, and at two critical times steered me toward a degree that may have otherwise eluded me. I thank Dr. Victor R. Baker for assisting with the fluvial aspects of the dissertation.

The geologic investigations were supported by the following

National Science Foundation grants to Eileen Johnson: BNS76-12006,

iii iv

BNS76-12006 AO!. Additional funding for the isotopic studies were provided by the University of Arizona Laboratory of Isotopic Geo chemistry, The University of Arizona Graduate Student Research

Development Fund and the Geological Society of America Grant 2540-79.

The accelerator 14C dates were made in cooperations with Drs. A. J.

Timothy Jull and Douglas J. Donahue from the University of Arizona

Tandem Accelerator Mass Spectrometry 14C dating facility, which is funded by the National Science Foundation Grant CHE 78-118576 to Drs.

Paul E. Damon, Department of Geosciences,. and Dr. Douglas J.

Donahue, Department of Physics, The University of Arizona.

I want to thank my parents for all their years of support that helped lay a foundation for this dissertation. I thank my brother, Bill, for his constant companionship and advice and for having to time to listen. My scientific interest was nurtured years ago by two couples, who never lost faith in me and who always shared my curiosity for what lay beyond the next hill. To Elwood and Marie Wilkins, Jr. and Jim and Marianne Akerman, thank you for all the good times. To my step daughter Melanie, I thank her for always asking "How many more pages do you have to write, Tom?" because one more page always followed.

I dedicate this dissertation to my wife, Linda, without whose love none of this would have been possible. Thank you for your laughs, your hugs, your understanding and for keeping my feathers from getting too ruffled. TABLE OF CONTENTS

Page LIST OF ILLUSTRATIONS ...... vii LIST OF TABLES ...... ix ABSTRACT •• ...... x 1. INTRODUCTION ...... 1 Geography and Regional Geology. • • • • • • 2 Methodology. • • • • • • • • • • 6 Previous Research. • • • • • • • • • • • • • 8 Summary of Late Quaternary History. • • • • 11

2. ALLUVIAL STRATIGRAPHY • 15

Geology of Lubbock Lake Site 16 Blanco Formation 16 Strata 01 and 02 16 Tahokan(?)-age Sediments •• 18 Stratum 03 • • • • • • • • • • • • • 18 Yellowhouse Draw Formation 21 Member 1 •••• •• 21 Member 2 •••••••••• 28 Member 2 Radiocarbon Dates 35 Member 3 •• 39 Member 4 ••••• 47 Member 5 ••••• 50 Geology of Upper and Lower Yellowhouse Draw and Blackwater Draw • • • • • • • • • • • • • • • • • • 54 Alluvial Strat:igraphy of Yellowhouse Draw Downstream from the Lubbock Lake Site. • • • ••• 54 Alluvial Stratigraphy of Yellowhouse Draw Upstream from the Lubbock Lake Site • 57 Trench 109 • • • • 58 Trench 114 • • • • • • • • • • • • • • • • 59 Trench 111 • • • • 60 Alluvial Stratigraphy of Blackwater Draw ••• 60 Trench 110 • 60 Trench 112 • • • • • • • • . . . 61

v vi

T ABLE OF CONTENTS--Continued

Page 3. INTERPRETATION OF STRATIGRAPHY •• ...... 63 Regional Stratigraphic Trends. • • • • • • • 63 Yellowhouse Draw Formation. • • 63 Interpretation of Sedimentary Environments and Regional Stratigraphic Histories. • 65

4. ORGANIC CHEMISTRY OF FOSSIL BONES • • • • 78

Inorganic and Organic Constituents of Bone • 78 Bone Contaminants •••••• 79 Previous Research. • • • • 80 Imino Acid Biochemistry. • • • 80 Isolation of Hydroxyproline 86 Fossil Collagen Isotopes • • • 90 Location. • • • • • • • • 94 Stable Carbon Isotopy 95 Plant Ecology • • • • • • • 98 Mammalian Isotope Compositions 99 Bison Ecology. • • • 102 Experimental Results • • • • • • 103 Discussion •••••••••• 108 Direct 14C Dating of Inorganic and Organic Phases from a Known-age Fossil Bone. 120 Results • • • • • • • ...... 128 Discussion •••• 129 Conclusions • • • • • 131 5. CONCLUSIONS ...... 132 APPENDIX: GEOLOGIC DESCRIPTION OF TRENCH 65. • 136

REFERENCES • • • • • • • • • • • • • • • • • • • • • • 146 LIST OF ILLUSTRATIONS

Figure Page

1. Location map showing the Llano Estacado and regional extent of the southern High Plains 4

2. Location ma.p showing geologic test trenches along Blackwater and Yellowhouse Draws near Lubbock, Texas • • • • • • • • • • • 7

3. Location map of backhoe trenches and archaeological excavation areas at the Lubbock Lake Site, Lubbock, Texas •••••••••••• ••• 9

4. Generalized stratigraphic column with radiocarbon dates ...... 13

5. Geologic cross section from Trench 46 to Trench 65 (type section) at the Lubbock Lake Site •••• 17 6. Geologic cross section of Trench 52 at the Lubbock Lake Site ...... 19

7. Geologic cross section of Trench 109 in Yellowhouse Draw at the Lubbock Lake Site 20

8. Geologic cross section of Trench 108 across Yellowhouse Draw at the Lubbock Lake Site 22 9. Geologic cross section of Trench 35 at the Lubbock Lake Site •••••••••• 24

10. Geologic cross section of Trench 27 showing rhythmites and shore facies of member 2 of the Yellowhouse Draw formation •••••••••••• 30

11. Geologic cross section of Trench 43 at the Lubbock Lake Site showing shore facies of all strata and marl facies of member 2 in the Yellowhouse Draw formation • • • • 32

12. Geologic cross section of Trench 18 at the Lubbock Lake Site showing shore facies of the Yellowhouse Draw formation and disconformities within section •••• • • • • • • • • • • 40

vii viii

LIST OF ILLUSTRATIONS--Continued

Figure Page

13. Longitudinal and geologic cross sections for Yellowhouse and Blackwater Draws 46

14. Time-process diagram showing geologic processes at the Lubbock Lake Site during the Pleistocene Epoch •••••••••• 74

15. Regional time process diagram showing how geologic processes aeted on a regional scale on the Llano Estacado •••••••••••••••• 75

16. Flow diagram showing steps used to isolate hydroxyproline from a fossil bone 87

17. Diagram showing the reaction of nitrous acid with primary and secondary amines •••••••• 89

18. Chromatograms of amino acids in gelatin, nitrosylated gelatin hydrolysate, and imino acid phases during isolation of hydroxyproline 92

19. Chromatograms showing elution positions of hydroxyproline and proline during preparative isolation with 1 N HCI buffer •••••••• 93

20. Flow diagram showing Ol3C composition of various fractions of a fossil bone •• • • • • • • • 106

21. Plot of collagen o13 C vs. geologic age for fossil Bison spp. bone from the Lubbock Lake Site, Lubbock Lake, Texas ••••••••••• 110 13 13 22. Plot of bone collagen 0 C vs. 0 C composition of grasslands ••••••••••••••• 117 13 23. Plot of bone collagen 0 C vs. percent C 4 grasses in the grassland ••••••••••••••••••• 118

24. Flow diagram showing radiocarbon dates on fractions of a mammoth of known age from the Domebo Site, Oklahoma ...... •...... 127 LIST OF TABLES

Table Page

1. Amino acid compositions of gelatin and fossil bone collagen and derived fractions • • • • • 91 2. ol3C isotope data for Lubbock Lake Site, Texas ••• . . 109 3. Quantitative amino acid analyses of selected fossil bison bones dating ca. 200 to ca. 12,500 yr B.P., Lubbock Lake Site, Texas • • • • • • • • • • •• 112

4. Radiocarbon dates from different phases of Domebo mammoth bone compared to dates on associated wood. 123

5. Amino acid and uranium series analyses of acid insoluble fraction from Domebo mammoth bone from the Domebo Site, Oklahoma •••••• 126

ix ABSTRACT

Sedimentology, stratigraphy, and stable-carbon isotopy were used to reconstruct geologic and climatic events on the Texas southern

High Plains from ca. 13,000 yr B.P. to the present. The alluvial sediments in Yellowhouse and Blackwater Draws were used to construct the geologic history. The oldest valley alluvium comprises the >13, OOO-yr-

B.P. fluvial sediments that were incised and buried by fluvial and lacustrine sediments dating ca. 13,000 to 4900 yr B. P. Lacustrine waters changed from oligotrophic to eutrophic and finally calcalitrophic.

Regional valley erosion at 4900 yr. B. P. developed a widespread disconformity within the Yellowhouse Draw formation, which separates lower fluvial and lacustrine sediments (ca. 13,000-4900 yr B. P.) from the overlying sediments dating 4900 yr B.P. to present. After 4900 yr

B. P., intermittent streams and eolian processes deposited several meters of sand the length of each valley. Cienegas returned to downstream reaches of both draws after 1500-2000 yr B. P.

Methods were developed to extract purified collagen residues and hydroxyproline from heavily contaminated fossil bones. Reliable 13 6 C measurements 0,P collagen require isolation of single amino acids, 14 whereas less specific purifications may yield accurate bone collagen C dates. Collagenous residues were extracted from 13,000-200-yr-B. P. fossil bison bones from the Lubbock Lake Site at Lubbock, Texas, and 13 1513 C values were determined. Collagen 15 C values changed from -8 per mil at 200 yr B.P. to -10 per mil at 4900 yr B.P. and to -17 per

x xi 13 mil at 12,500 yr B. P. The 15 C changes imply that the Lubbock area grasslands contained 30 to 40 percent C grass biomass at 12,500 yr 4 B • P. in contrast to the 95 percent C 4 grass biomass in today's grass lands.

The stratigraphic and isotopic results gave similar paleoecological histories for the Texas southern High Plains. At 12,500 yr B. P. permanent streams existed and grasslands may have resembled those in the northern central Great Plains today. The climate warmed gradually, and the water table dropped until 5000 yr B. P. when a major hydrologic shift occurred. After 4900 yr B.P., modern climatic depositional and vegetation communities were developed. Geomorphic thresholds apparently controlled the regional disconformities, depositional events, and pedogenetic episodes. Climatic change was the ultimate cause of stratigraphic changes, but individual geologic events were not coeval with any similar climatic shift. CHAPTER 1

INTRODUCTION

Geologic records from the late Quaternary are frequently so

detailed that the quantity of data obscures regional synthesis. Are

generalizations possible among a plethora of field information, especially if data may be circumstantially related? In an interdisciplinary ap

proach, how precise is each subdiscipline and are any paleoenviron

mental tools better than another? These questions are evaluated in this

dissertation by using geologic work from the Texas Panhandle. The

goal was to test isotope records against geologic data and demonstrate the complex interdependence of geologic changes. A unified geologic history was constructed by using disciplines ranging from stratigraphy

and paleontology to organic and stable isotope geochemistry.

The field area was the Llano Estacado subsection of the Texas

Panhandle where two alluviated valleys were studied. Each valley con tains a modern ca.-13,OOO-yr-B.P. sedimentary record representing numerous depositional environments and containing abundant fossil re mains. The presence of humans since at least 11,000 yr B. P. contrib uted archaeological data to the chronologies. The Llano Estacado is an ideal locality to model geology and test new methods because the strat igraphy is uncomplicated by frequent disconformities and the sedimen tary record is relatively complete for the last 12,500 years.

1 2

The research has been arranged into five chapters. Chapter 1 gives the background, scope, purpose, previous research, and general geologic history. In Chapter 2, field descriptions are assembled from sedimentary, stratigraphic, pedologic, geomorphologic, and geochrono logic data. Chapter 3 interprets the geologic data.

Chapter 4 is an analysis of l3C in fossil bison bone collagen and outlines methods for improving bone 14C dating. Included in

Chaptet: 4 are techniques for purifying collagen and isolating hydroxy proline. In Chapter 5, a chronology of geologic events is given, fol lowed by an integration of the geological and geochemical data. A un ified alluvial geologic history of the Texas southern High Plains is presented and interrelationships of geologic disciplines are discussed.

Geography and Regional Geology

The Llano Estacado is the southernmost erosional remnant of the

North American High Plains, a Cenozoic alluvial apron extending hun dreds of kilometers eastward from the Rocky Mountains for the length of the North America Great Plains. The Llano Estacado is part of the southern High Plains, which is divided by the Canadian River into the northern panhandle subsection of Oklahoma and Texas, and the south ern, Llano Estacado of eastern New Mexico and the Texas Panhandle.

The eastern, northern, and western boundaries of the Llano Estacado are erosional escarpments. The southern boun.dary is gradational be cause it merges with the Edwards Plateau.

Northward stream piracy of the Pecos River between the late

Miocene and early Pleistocene (Reeves, 1972) isolated the Llano Estacado 3

from its ancestral southern Rocky Mountains headwaters, and by early

Pleisto... ene time, the Llano Estacado had attained its isolated, localized

drainage character. The modern Llano Estacado is a virtually feature

less, 82,000 km2 plain sloping southeastward 1.5 to 1.9 m/km. Relief is

broken only by playa lakes and alluviated valleys. Elevations range

from 1,500 m in New Mexico to 750 m in the southwest. Valley reliefs

are 5-20 m in the interior to 50-70 m at the eastern Llano Estacado.

Valleys drain the region from northwest to southeast and are relics of

Rocky Mountain stream systems that in early Pleistocene time incised

valleys of significant relief into the bedrock. Today the valleys exist

as entirely or nearly sediment-filled drainages containing up to 9 m of late Quaternary alluvium.

The modern semiarid climate has an average January tempera

ture of 1°C in the northwest and 7°C in the southeast, and an average

July temperature of 23°C in the northwest and 29°C in the southeast

(Wendorf, 1975). The median rainfall is 360 mm in the northwest and

550 mm in the southeast. The original vegetation was short to mid

grasslands of the High Plains Bluestem or Mixed Prairie climaxes

(Allred, 1956), but most of the land is now cultivated.

Numerous valleys trend southeastward across the Llano Esta

cado, but only the Yellowhouse, B, .;kwater, and Tule Draws attain notable size. Because. of its excellent exposures of valley alluvium, the

Lubbock Lake Site, near Lubbock, Texas (fig. 1), in Yellowhouse Draw became the scene of concentrated geologic work.

The principal geologic units of the Llano Estacado are: (1)

Mesozoic and older bedrock, (2) Tertiary stream and eolian sediments, 1050 1040 1030 10ZO 101 0 100" 990

360 -I , CliO \\ ~I COL ",.1""" 1IwL.-.I'" t 1-36" N .I "~ ~ -~"°AMARILLO·· ~':2".... 0 u: .. 3S .-ft.- _ paID o!!!..a ~.• _ ~ 0 J ~ I ,35 OKLAHOMA

0 34 340

RiD Hondo INDEX MAP

LLANO ESTACADO 33" SUBSECTION ollhl SOUTHERN HIGH PLAINS, U.S A 33"

0 3Z "11------l.----' 3Z0 TEXAS o km. 100 STAFFORD 1978

1050 1040 1030 10ZO 101 0 100" 990 980

Figure 10 Location map showing the Llano Estacado and regional extent of the southern High Plains ,;:. 5

and (3) Quaternary-age playa, eolian, and valley alluvium irregularly

deposited throughout the southern High Plains. Regional bedrock in

the southern High Plains may be Permian, Triassic, or Cretaceous clas

tics and limestones that are exposed along escarpments around the

I;Lano Estacado. Most commonly exposed is the Ogallala, an extensive

blanket of Tertiary eC'lian and fluvial sediments deposited on an ero

sional surface of Triassic and Cretaceous rocks (Hawley, Bachman, and

Manley, 1976). The Ogallala was designated a formation by Frye and

Leon3.rd (1957), and a group by Evans (1949). The Ogallala is divis

ible into the lower Couch and upper Bridwell units, and its age em

braces the late Miocene to middle Pliocene (Reeves, 1976). By the late

Pliocene, calcification of the upper part began, and by the early

Pleistocene, a petrocalcic horizon over 5 m thick had formed. The ca

liche forms the erosionally resistant High Plains "caprock caliche. II

Early Quaternary sedimentation includes the Blanco Formation.

Its playa lake clastics, carbonates, and olive bentonitic clays and sands

were deposited in basins deflated into the caprock caliche. The age of

the Blanco Formation is imprecisely known, but it is older than 1.4. m.y., age of the Guaje ash overlying the Blanco type section at Cros byton, Texas (Izett, Wilcox, and Borchardt, 1972), and younger than

2.4 m.y (Johnson, Opdyke, and Lindsay, 1975).

The Kansan period is represented by the Tule Formation, which, like the Blanco, is hypothesized to represent modern-type playa sediments (Pierce, 1975). Overlying the Tule Formation is the Illinoian

Blackwater Draw Formation (Reeves, 1976), a northeasterly thickening surficial blanket of eolian sand originally named the "cover sands" 6

(Frye and Leonard, 1957). Soils formed in the Blackwater sands dur

ing the Sangamon include the Brownfield, Amarillo, Patricia, and

Arvana (Reeves, 1976). The youngest playa deposits are the early

Wisconsinan Double Lakes Formation and the late Wisconsinan Tahoka

Formation. The latter dates approximately 24,000 to 14,500 yr B.P.

and is divisible into the Rich Lake, Vigo Park, and Brownfield Lake

members (Reeves, 1976). The valley alluvium usually postdates

Tahokan time, but isolated buried terraces may be older than 13,000 yr

B. P. The greatest part of the valley sediments are younger than

13,000 yr B. P. and are assigned to the modern to ca. 13 , OOO-yr-B • P.

Yellowhouse Draw formation.

Methodology

Field data were collected over five summers. Work concentrated

on the Lubbock Lake Site (41LU-1) in Yellowhouse Draw just north and

outside the city limits of Lubbock, Texas (fig. 2). The site has been

excavated over many years by the Museum of Texas Tech University.

The archaeological significance of the site became apparent when a

resarvoir was dredged by the Wm::'k Project Administration in 1936 to

supply water to Lubbock. The reservoir dried as agricultural irrigation lowered the modern water table. The old reservoir walls now provide

over 1,900 linear m of exposed valley alluvium. Data from the Lubbock

Lake Site include information collected from archaeological excavations, measured sections of exposed alluvium, and descriptions of backhoe trenches surrounding the site (fig. 3). After the Lubbock Lake Site stratigraphy was completed, 17 km upstream and 47 downstream in 7

102"30' 102"15' 102'00' 101"45' '.. _ ... -1-"""'_ , 34"15' Earth 34"l5' ...... '\.. I _ LOCATION MAP ·v...... "? FOR YELLOWHOUSE AND r'" BLACKWATER DRAWS \ ... _... TEXAS SOUTHERN HIGH PLAINS

r=KILOMETERS 20 34"00' 34"00'

81 i Littlefield ~I' VBu/l ::! .' Lake ~i .-.. 'V"~'>"~"''?IIIIUSiOn , _...... 0.. Loire 1 \ Abernathy HALEr CO. LJ:,AMB co -fA '. "[HOCKLEY" 00.: Yellow" -" .,ser"· ,~.;. -" -" luBBOcK co .r. Lake ... 0...·· 'PR III .. ' "',,'. 'Ii" .,,- .J 8\ . 'oj.. YeI IO :, 33"45' Z'\ "'...... y.:_... ./ New Deal 33"45' ~ ... "\ ...../.: 5'\8 :/t' ./ / .J:,i r"

33"30' 33"30' :02"30' 102"15'

Figure 2. Location map showing geologic test trenches along Blackwater and Yellowhouse Draws near Lubbock, Texas 8

Yellowhouse Draw were investigated by using trenches along these sections (fig. 2). Trenches and gravel quarries were used in Blackwater

Draw, and auger borings and gravel quarries were examined in Run- ningwater Draw.

Laboratory research encompassed the stable carbon isotopic and amino acid chemistry of fossil bones, especially the geoch(.'· listry of bone collagen. Methods were developed to extract and purify collagen from humate-contaminated bones and to use collagen or individual amino 12/13 . acids for stable and radiocarbon studies. In the C experlment, 13 bones of known age were used and collagen t5 C was plotted versus geologic time. The stable carbon isotopes were used to evaluate dietary and grassland-climate changes during the last 12,000 yr. A technique was developed to isolate hydroxyproline, a relatively collagen-specific amino acid, which can be used to improved the accuracy of bone radiocarbon dates.

Previous Research

The Quaternary sediments of the Llano Estacado have been studied in detail since the mid-1930s; much of the impetus for research has originated from archaeological discoveries in valley alluvium.

Wendorf and Hester (1975) reviewed research on the Llano Estacado up to the mid 1970s. Some of the earliest multidisciplinary work was done at the type Clovis site in New Mexico. Early archaeological, geological, and paleontological reports for the Clovis site were published by

Howard (1936), Antevs (1935, 1941), Cotter (1937, 1938), Stock and

Bode (1937), and Patrick (1938). "9B

"93 "92 "91

LEGEND 22 Batlhoe _ lor • - QOCIoQlc """"""'" • ArchII!cIc9Icd IICIMIIbI

Figure 3. Location map of backhoe trenches and archaeological excavation areas of the Lubbock Lake Site, Lubbock, Texas ...0 10

Two comprehensive volumes on the region's Quaternary history were edited by Wendorf (1961) and Wendorf and Hester (1975). Impor tant regional summary articles are by Frye and Leonard (1957), Green

(1962), Reeves (1970, 1976), Harbour (1975), Haynes (1975), and Haw ley and others (1976).

Literature relevant to my research were that on the Clovis site or Blackwater No. 1 (Howard, 1936; Sellards, 1952; Evans, 1951; Hes ter, 1972), the Planview site (Sellards, Evans, and Meade, 1947), the

Miami site (Sellards, 1938), the San Jon locality in the northern Pan handle (Roberts, 1942; Judson, 1953), and regional playa lake strati graphy (Reeves, 1972; 1976). Regional Quaternary pollen studies were those by Green (1961), Hafsten (1961), Schoenwetter (1975), Oldfield

(1975), and Oldfield and Schoenwetter (1975). The diatom synthesis studies used were those by Patrick (1938), Hohn and Hellerman (1961), and Hohn (1975). Invertebrate treatises were those by Wendorf (1961),

Drake (1975), and Pierce (1975). A regional paleoenvironmental syn thesis was made by Wendorf (1970).

A history of research at the Lubbock Lake site was compiled by

Black (1974); Holden (1974) tabulated work done between 1939 and

1963, and Holliday (1977) compiled studies to 1978. The paleontology of the Lubbock Lake Site is cited in regional discussions by Sellards

(1952), Evans and Meade (1945), Wendorf (1961, 1970), Wendorf and

Hester (1975), and Haynes (1975). The first test trenching at the

Lubbock Lake Site was done by Wheat (1974) in 1939 and 1941; later trenching was done by Evans and Meade, who in 1948, 1950, and 1951 established the stratigraphic nomenclature still used today (Sellards, 11

1952) • Green (1962, n. d.) reported on his comprehensive geological undertakings, and Johnson (1974) presented his geologic work and re viewed previous stratigraphic schemes. Wendorf (1970) used diatom, pollen, and invertebrate records from Lubbock Lake to name the Lu.b bock subpluvial. Additional Quaternary research at the Lubbock Lake site includes diatom descriptions by Compton (1975), fossil seed inven tories by Thompson (1977), vertebrate paleontology by Johnson (1976), archaeological summaries by Kelley (1974), Holliday (1977), and Kaczor

(1978), and geology by Stafford (1977, 1978, 1981).

Summary of Late Quaternary History

Having been cut off from their Rocky Mountains headwaters in the early Pleistocene, the relic drainages crossing the southern High

Plains were left with only local sediment sources. Elimination of en vironmental variables active hundreds of kilometers upstream resulted in an unusual opportunity to analyze the effects of climatic changes pre served in locally derived sediments.

The local bedrock for Yellowhouse and Blackwater Draws is either the Ogallala group's Couch or Bridwell formations or Plio

Pleistocene Blanco Formation lacustrine sands and clays. At least two valley-floor levels of Wisconsinan age have been recognized in Yellow house Draw. The older and higher elevation valley floor underlies stratum 03. The terraces are fluvial sands and gravels perched from

8.5 to 5.2 m above the existing valley bedrock elevations and are pres ent at three locations in Yellowhouse Draw. The three outcrops are lithologically similar and date as Rancholabrean because in situ Camelops 12 and Bison cf. antiquus are present; stratum 03 dates older than 12,650 yr B ••P b y usmg. a 14C d a t e f rom a11· UVlum 0 f t hI· e over ymg mem b er 1 of the Yellowhouse Draw formation (fig. 4). After stratum 03 was deposited, stream incision formed the present bedrock valley floor.

The Yellowhouse Draw formation is the next younger unit; it is inset against and above stratum 03 and is a fluviolacustrine and eolian rock unit dating between ca. 12,650 yr B.P and the present. Five members make up the Yellowhouse Draw formation. The lower three are conformable and represent emergent water-table conditions, and the third member is disconformably overlain by the fourth. The basal unit, member 1, is channel gravel overlain by fluvial sands and upper oligotrophic lacustrine clays. The sediments were deposited between ca. 12,650 ± 250 (I-246) and 10,800 yr B.P. Conformably overlying member 1 is member 2, a series of lacustrine diatomites, diatomaceous muds, rare marls, and fossil cienega soils representing an increasing eutrophic, shallower pond system. Lake chemistry began changing around 8,000 yr B. P. when member 3 sedimentation reflected ever- increasing primary calcium carbonate deposition and a shift from eutrophic to alkalitrophic lake chemistry.

A brief but widespread disconformity at 4,900 yr B.P. separates member 3 from the overlying member. Effluent stream flow ended by 4,900 yr B.P., and thereafter only influent streams persisted. The regional drop in water table coincided with deposition of member 4.

The unit's fluvial fine sands grade upward into eolian sands. Pedo- genesis, which prevailed during member 4 time (4,900-2,000 yr B.P.), resulted in a thick Calciustoll in the upper two-thirds of this 13

GENERALIZED STRATIGRAPHIC COLUMN FOR TERTIARY QUATERNARY ALLUVIUM, YELLOWHOUSE DRAW LUBBOCK, TEXAS

MODERN (SMU - 649).h UNIT DESCRIPTION THICKNESS 50 * SO (SMU - 650).h CIENEGA CLAY 600* 50(SMU-69S)h SC 1.7 METERS 220* 50 (SMU-555)h SI SOIL-USTOCREPT 0.15 2600 * 50 (SMU-697)h S2 SOIL-HAPLUSTOLL 0.3 z 320* 60(SMU-546)e COLLUVIAL.EOLIAN SAND < S" 5B 0.8 w {260* 50 (SMU-345)e S5 SOIL-HAPLUSTOLL 0.5 a.: 3S0*4O(SI-2700)e S4 III 505 * 55 (SI - 270l)e SOIL-OCHREPT 0.2 < 5A COLLUVIAL.EOLIAN SAND 1.1 ...I S90* 70 (SMU-65Ihh o 720 11 40 (SMU-314)h SII SOIL- CALCIUSTOLL 1.3 :t: 4960* SO (SM U-492)h Co) 4C SOIL A HORIZON 1.1 Z 61151175 (SI-3197).h 4S SOIL S HORIZON 1.0 < 4910 *50 (SMU-332-333)bc a.: 4900* 60(SMU-531)Ih 4A FLUVIAL SAND a CLAY 0.9 7100 II SO (SM U-64S).h 3C (S8) CLAY- SAPRIST 0.8 5770*SO(SMU-545)h 3B LACUSTRINE MARL 0.7 7970* SO (SMU-262)h 3A LACUSTRINE CLAY 0.5 11 7890 100 (SMU-302)h 2F COLLUVIAL SAND 0.5 28 --9S83 * 350(C-558)bb ------2E 9960 * SO (SMU-275)h 2E LACUSTRINE COLLUVIUM 1.3 {'O,060* 70 (SMU-25I)h 2C (S7) CLAY- SAPRIST 0.5 2A 7840*140 (SMU-247)r 2B DIATOMACEOUS EARTH 1.2 r030 '90 (SMU-20." 2A DIATOMITE,CLAY, MARL 0.S5 11,190 *90 (SMU-292). 10 COLLUVIAL PEBBLY CLAY 0.4 10,540 11 100 (SMU-547)p IC LACUSTRINE CLAY 0.8 9700 II 450(L-238g). 9020" 90 (SMU-317)ba IB FLUVIAL SAND a GRAVEL 1.8 11,100 * SO (SMU-263)w IA FLUVIAL GRAVEL 0.5 11,100 *100 (SMU-54S)w 03 TERRACE ALLUVIUM I.S 12,150 *100 (SMU-295). Taw EOLIAN SANDS 5+ 12,650*250(1-246). 03 02 ALLUVIAL GRAVEL 0.8 01 LACUSTRINE SAND a TAHOKA Fm. (?) DOLOMITE 23+ RANCHOLABREAN j:AUNA Te FLUVIAL a EOLtAN SAND 5+

BLANCAN FAUNA

14C LEGEND BLANCO Fm. ba - BONE APATITE bb - BURNED BONE be - BONE COLLAGEN c -CHARCOAL h -HUMATES TB P -PLANTS r -RESIDUE BRIDWELL Fm. • -SNAILS sh - SOIL HUMATES w -WOOD STAFFORD (1981)

Figure 4. Generalized stratigraphic column with radiocarbon dates 14

member. Pedogenesis ended approximately 2,000 yr B. P. when the first

of several colluvial-eolian sands of member 5 were deposited. Member 5

sands were periodically deposited and contain thin, weakly developed

soils. As modern time is approached, alluviation and colluviation be

came increasingly prevalent and strength of soil development dimin

ished. Cienegas transgressed upstream during member 5 time, and as

mentioned earlier, it was the early 20th century search for water in

Yellowhouse Draw north of Lubbock that first exposed the rich geolog

ical and archaeological record in the alluvium.

Noteworthy changes during the late Quaternary interval are the

contrast in the regional ground-water conditions before and after 4,900

yr B.P., and the abrupt and regional extinction of Pleistocene mega

fauna at 11,000 yr B. P. The contact between members 3 and 4 demar

cates the change from emergent stream flow before 4,900 yr B.P. and influent flow after 4,900 yr B.P. The boundary between strata IC and

2A separates older Rancholabrean-age Pleistocene megafauna from modern Holocene-age vertebrate faunas. These two events in the re

gion's natural history are culminations of events begun hundreds to perhaps thousands of years earlier, and their satisfactory explanations requires an understanding of every aspect of the Quaternary Epoch. CHAPTER 2

ALLUVIAL STRATIGRAPHY

The stratigraphic nomenclature at the Lubbock Lake Site was first proposed by Wheat (1974) and since then seven major stratigraphic schemes have been proposed. The stratigraphic nomenclature in this dissertation retains most of the stratigraphic divisions of Evans and

Meade as cited by Sellards (1952).

Throughout the Yellowhouse and Blackwater Draws, the local bedrock is either the Bridwell or Blanco Formation. At the Lubbock

Lake Site the Blanco Formation was informally divided into stratum 01

(Johnson, 1974) and stratum 02 (Stafford, 1977) to facilitate local archaeology work. The zero preceding the numbers 1 and 2 implied a pre-archaeological age (>12,000 yr B. P. ) • Stratum 03 underlies the

Yellowhouse Draw formation and consists of terrace gravels tentatively assigned a Tahokan age.

The name Yellowhouse Draw formation is proposed in this dis sertation for the sediments in Yellowhouse and Blackwater Draws dating from ca. 12,650 yr B.P. to the present. The Yellowhouse Draw form ation includes as members strata 1, 2, 3, 4, and 5 of previous authors

(Johnson, 1974; Stafford, 1981). Each of the five members is further divided into strata as necessary.

15 16

Geology of Lubbock Lake Site

Blanco Formation

Local bedrock at the Lubbock Lake Site is the Blanco Forma tion, a lacustrine unit underlying the Lubbock Lake Site and continuing eastward at least 1.5 km (Reeves, 1970, personal commun.) The south ern limit is Wood Ranch in Yellowhouse Draw, 40 km downstream from the Lubbock Lake Site (Reeves, 1970). The northern extent is imprecisely known, but I have drilled into the Blanco Formation 15 km upstream from the Lubbock Lake Site at Shallowwater, Texas (fig. 2).

Strata 01 and 02. Blanco Formation sediments are sands in the lower two-thirds of exposures and more dolomitic sediments in the upper section. The lower sands, which make up over 75 percent of the drilled thickness, are massive, firm, olive (5 Y 5/4), clayey, very fine to fine quartz sands, lying unconformably on red sands of the Bridwell formation. Upper Blanco Formation carbonates are very thickly bedded, massive, very to extremely hard, white dolomite or dolomite interbedded with medium- to thick-bedded olive sands and thin bentonitic clay beds.

Primary sedimentary structures are absent in the dolomites, but the prevalent structure is deep vertical jointing and slickensided curvilinear jointing. The intense fracturing facilitated incorporation of Blanco dolomites into the alluvial sediments because the granule to-boulder-size fragments are easily eroded from the valley edges.

An intraformational Blanco Formation conglomerate, stratum 02, was exposed at the Lubbock Lake Site (fig. 5) and yielded the only EAST EXCA

TRENCH 46 EXCAVATION AREA 5

02 ---diat Blancan vertebrate fauna 2A lami 1 BLA'NCO pebt LEGEND FORMATION 11111",,"1 Fossil soil 11£111£11 If I Fossil cienega root horizon S5 Paleosol oc=r

Figure 5. Geologic cross section from Trench 46

TRENCH 46 :XCAVATlON AREA 5

sediments removed durin~'~6~r~g~g=------.. :::-.~ ------••...... - I' .. .IP' . ------..... ~--- ~~:::--- .. ,/

-_9!:l"cienegCLClay~ ,.""",.".,."..",.",..~ _-= --=-=- :::::-_-= _-::. ::::-_- - - ,..,.-"'- greye 3_ _calcareous ._ --:- _ clay _ and ~rl ------~ _-. ______------p oleosol -- _olIve slit ------. . ~i~ma7e~-- -;'uds _____ ..:::~ ~~ ~~ ..:::------Pa~osol laminated diatomite ------' ------. ------pebbly,cobbley clay ------=--- BLANCO FORMATION olive sand

, i 10 meters o 2 4 6 8

,nch 46 to Trench 65 (type section of Yellowhouse Draw formation) at the Lubhock Lake Site

WEST TRENCH 65 GEOLOGIC 5 TYPE SECTION _------~:?'~ meters nts removed during \~6~r~g~g=- _ - - - -.-::,:? =:: -- 5 0 ;;;;~:~~~~o ------.. ~.... __ - F 5 • ,/ S__ 4 2 --_ _-- _ - _ _ ~ _ _,..",.,...,... _ ~ _ _ _ ,...---- ,...,- greyed sands ------_ _ - ______. _ _ _ _ _ Paleosol S6 ill1JTI1] 3C 3 __ .::~ ~ ~.: .: ~ ~ .: _ _ _ _ Paleosol S7 ~::-r-~ 3A/B

_ _ _ _ I ------Ilill'!".',"- 2C 4 ----___ ------28 ______---=--=--=--=--=-:':_=_=6=- 2A _____ .--. _ ~ _ ..- _____ .- __-----~s~:~t:>-:...;::-~O 'C 5 ;::: _ _ _ _------:o~~::~~\~;.;::;~=--).::.i.:-;~ I B oOc0g'oo I A BLANCO 6 FORMATION olive sand

I i B 10 meters

In of Yellowhouse Draw formation) at the Lubbock Lake Site

18 vertebrate fossils from the Blanco Formation in Yellowhouse Draw. The stratum disconformably overlies lacustrine sands and contains three lithologies: a basal 50- to 70-cm-thick, very hard, fine to medium sandy, bentonitic carbonate pebble conglomerate; an intermediate unit

10-45 em thick of clayey, fine to medium quartz sand, and an uppermost 20-40 cm of massive, very hard to extremely hard, coarse to very coarne, angular, blocky montmorillonitic clay. The Blancan vertebrate fossils identified by Green (n.d., p. 63), were exclusively from the basal stratum 02 conglomerate and included Geochelone, Megalonyx, Borophagus diversidens Cope, Stegomastodon -----mirificus (Lidey) , Nannippus, Plesippus simplicidens, Eguus, Platygonus, Pliauchenia(?), and Tanupolama.

Tahokan (?) -age Sediments

Stratum 03. Remnant gravel terraces predating the Yellowhouse

Draw formation occur sporadically along the inner meander bends of

Yellowhouse Draw. The gravels disconformably overlie the Blanco For- mation and are from a pre-12,650-yr-B.P. river system whose deposits were incised then buried during Yellowhouse Draw formation time.

Figures 6 and 7 illustrate the stratigraphic position of the terraces.

Common lithologies are subround to rounded, medium to very coarse carbonate pebbles with a fine to medium quartz sand matrix. Ten- centimeter-thick lenses of intermediate tabular, planar cross-stratified medium quartz sand are present within the 80-cm-thick channel gravel.

The gravels are conformably overlain by clayey fine quartz sands and 19

SOUTH NORTH elevation TRENCH 52 (meters) EXCAVATION AREA 9 -971

-970

silicified oquatlc - . - . .,; -' .. -.~ '03 :... . plant rootlets .5Y6/2 ._.~ ~~.'.o-:- ~.'. ~.~ .~·.~.~7: :'~~: ... ° .00 "0 03 . 25Y312 ;.\.0.'0.0'0"0'0'· . '0'. '0' '0 0 . ·0 •• 0 •• .. '0 0 .'0' 'o':=':u·. o· ...... -968 .

cloys and diatomite 1:1=::::I==::C:=::C=~I o 2 meters v. f. sandy cloy with scattered cobbles, few calc. roots

Figure 6. Geologic cross section of Trench 52 at the Lubbock Lake Site WEST

TRI Yellowl HIGH PLAINS CALICHE I I I I I I I Alii ···---AI3bl

g oI dilconfo . re ;OyRS/4 silt~

10YRS/2 cloye Y sill_

SMU-547: 10, 540±100

Figure 7. Geologic

RENCH 109 whouse Draw

AI>

...... _Al2cabl nformily iilly vf-f sand silt BRIDWELL FORMATION meters

o 10 20 VEx2

;ic cross section of Trench 109 in Yellowhouse Draw at the Lubock Lake Site

EAST meters. o

HIGH PLAINS CALICHE I I I I I I I I 5

-elev.=965m.-

BRIDWELL FORMATION 10

Stafford 119811

~ Lubock Lake Site

5-10-cm pebbly fine to medium quartz sands. Like all younger late

Pleistocene and Holocene clastic units, stratum 03 is bimineralic; that is, quartz and carbonate clasts co-dominate. The quartz comprises the medium sand and finer fractions, and the carbonates dominate the cobble and larger sizes; the two mineralogies overlap in abundance in the coarse sand range. Most of the carbonates are locally derived and are Blanco Formation dolomite and silicified and unsilicified caprock caliche.

Stratum 03 is sedimentologically identical with the member 1 sands and gravels of the Yellowhouse Draw formation, and similar stream characteristics are hypothesized for both. Elevation differences between stratum 03 and gravels of member 1 of the Yellowhouse Draw formation indicate that stratum 03 was incised and the valley floor cut at least 2.6 m below the base of the stratum 03 stream. Timing of the erosion is poorly known, but is older than ca. 12,650 yr B.P. and within the Rancholabrean Land Mammal Age.

Yellowhouse Draw Formation

Member 1. The basal member of the Yellowhouse Draw forma tion, is a perennial, meandering stream deposit dating between ca.

12,650 and 10,800 yr B. P. Except for a 2. 5-m-thick colluvial facies, the lower member ranges from 0.7 to 2.0 m in thickness at the Lubbock

Lake Site. Member 1 is divisible into three primary lithologies reflecting different fluvial or lacustrine depositional regimes

(Appendix). Stratum lA is a basal gravel found the entire width of

Yellowhouse Draw (fig. 8). Stratum IB is a fluvial sand, itself meters o WEST

4 01 dolomite

BLANCO FM. (011 ...... 6 ... 01 oliVi land

.... IK.r~.. Dlln

8 SMU-1I41 Mod.rn SMU-850 50!50 SMU-555 1120t50 -- 970.00 met"l .Ift SMU-5411 320tllO SMU-IIIII IIOot50 SMU-1I51 IIIOt70 • SMU-1I17 21100i50 --_ laru-144 10 aminated lOY IOYR211 clayey vi SMU-534 117ot40 R2/1 II I co SMU-531 410otllO 1 :lY5I3 vf landy lilly clay SMU-1I411 7100tllO SMU-545 577otBO SMU-544 11400!BO STAFFORD (1981) 12 01 oliv. land

Figure 8. Geologic c

Type Soil Sequence 1 TRENCH 108 YELLOWHOUSE DRAW !Ie Lowland loci ..

...... T 5'1'6/3 vi land eo lilly clay 4 5Y5/2 lilly clay 2.5'1'312 .illy Clay 5Y5II.illy clay 3C -3A pri.matlc 61\ clay .. - 2y~~:12 calc .• illy clay =---- l"U-144 IO~211 vl.and calC.lill cia 3 fR211 clayey vl-' .an1l' - Y clay :: :10_ I-m ",' ______1C '''''' """ ~_d ~2~5UY~!II2~~vl~.~a~nd~y~C~Ia~y~----~======~~==~d:ia='o=m=i=t.==::======~===:~~~~~~~ silly vl-I .and - I.Qray{2.5Y 7/2) lilly I .and e. "vl-c CoC0 3 land with vl-' quartz sand lamina. ' . ... '. ~.' .. " ... '\ ,.' .;11, ._ ,'.. 3.y\·~ . '" .•'.'!~ BLANCO FM. (01)

50 o• meters

ologic cross section of Trench 108 at the Lubbock Lake Site

TRENCH 108 EAST

5C Lowland foci .. YELLOWHOUSE DRAW

modern 5C 10YR 312 cloy ~-....!!reom channel _ _ " IOYR3I2 handy cloy \~L£~ cl0l. __ r .L .. '"T T T T T "TnT n IUIU-IIII 5Y3I2 clay 4 5Y5/2 lilly clay 38 '!ill lilly clay 3C ,-co~, ;;C~~-~~~------ c. lilly clay 3A alc.lill cia 3

------2 10 a diatomite 1C 2.5 Y6I2 clayey vf sand

1.0ray(2.5Y7/2) 01 ailly f land a Ii vf·c COCO! 5Y6I3 land wilh vf·f quartz land lamina. Clay'y vf land ~.. ...

BLANCO FM. (01) 50

3 at the Lubbock Lake Site

divisible into four secondary lithologies, and stratum lC is a lacustrine

clay and sandy clay.

1. Stratum lA: The stratum lA basal conglomerate is a 5-55-cm

thick, fine to medium sandy, subangular to subrounded, pebble and

cobble carbonate gravel. The gravel includes lenses of tabular, planar

cross-stratified, fine to medium sands. Over 95 percent of the clasts

larger than 4 mm are dolomite and caliche, with the remainder being

quartz, quartzite, chert, opal, Bridwell sandstone, or Blanco Formation

olive sandstone fragments. The maximum intermediate clast-size dimen

sion is 13 cm, with a medium value of 9.1 cm. Although the stratum IA

gravels occasionally show faint, high-angle planar cross-stratification,

the gravel is dominantly massive. Stratum IA is one of the two princi

pal gravel bodies in the lower member.

2. Stratum IB: Stratum IB makes up 50 to 75 percent of the volume of 'the member I sediments and includes 40 to 130 cm of sand

and fine pebbles carried in suspension or by traction. Although the modal lithology for the stratum is fine to medium sand, the stratum lith ology ranges from clayey sand to pebble gravel included in four principal facies reflecting changing flow conditions. The four facies are: (1) cross-stratified fine to medium sand; (2) pebbly sands; (3) convoluted muddy sands; and sandy pebble gravel. The sediments tend to be finer upward, but intraformational channeling is common within stratum IB, an aspect preventing time stratigraphic use of the IB lithologies (fig. 9). Stratum IB appears to disconformably overlie the basal gravel (stratum IA) and is both conformably and disconformably overlain by stratum IC. Stratum lB' convoluted sands, (2.5 YS/2) silty Stratum lB' fluvial sands, massive weli wf-f sand with convoluted white, f-m qtz sand sorted '-m qtz sands'rlpple X-strat'd, laminate grading to well sorted wf-f sand 1-2 em to 15 em thick sets of trough and II with convoluted clay laminate at valley axis. II tabular planar X-strat'd m-qtz sand. WEST EAST meters Stratum lB' fluvial sands, Stratum IA'basal channel gravel,m-c .9~ :.~: o ::~::.:.::: massive I.br. gr. (2.5YS/2), . 0.°.- qtz sandy,I-4cm sr-r carbonate pebble -.;: :.0 2-IOmm carbonate pebbly ...-.,; grvh4-IOem pebble/cobble grvl with ~~}:.~~ silty vf.-f.qtz.sand. 20 em clasts In lower elevation. dial earth~ 2

S o 2 4 S e 10 meters

Figure 9. Geologic cross section of Trench 35 at the Lubbock Lake Site

N ~ 25

The first facies of stratum lB is the 50-cm-thick beds of moderately sorted to very well sorted, cross-stratified fine to medium quartz sands. Sedimentary structures are 5- to lO-cm-thick, inter mediate angle, tabular planar cross-stratified sand and l-15-cm-thick cosets of small-scale festoon cross-stratification. Aquatic gastropods and freshwater clams (Pisidium and Sphaerium) abound, and large Ran cholabrean vertebrates are common.

The second fluvial facies is a light-brownish-gray (2.5 Y 6/2 m) silty, very fine to fine quartz sand with 2-l0-cm carbonate pebbles.

This facies commonly underlies the convoluted sands and conformably grades into a third facies in a fining-upwards relationship.

The third facies is convoluted silty to clayey fine sands, which cover extensive areas of the valley and are generally restricted to the upper portion of member I beneath stratum Ie. The facies ranges from white, weli-sorted to very fine to fine quartz sand with l-2-mm-thick, convoluted laminae to extensively convoluted clayey fine sands. Type B climbing ripples and recumbently folded cross-stratification are present at the unit's [facies] base. With increasing stratigraphic elevation, convoluting increases until all original sedimentary structures are destroyed.

The fourth facies of stratum IB is fine to medium sandy car bonate pebble gravel, which is a lateral facies of the stratum IB sands.

The thickness of the stratum IB gravel is often twice that of the stratum lA gravel; otherwise both gravels are lithologically and sed imentologically identical. Point bar morphologies are occasionally 26

encountered in stratum lB. Unless long trenches are available, strata

lA and IB gravels cannot be distinguished from one another.

3. Stratum Ie: Stratum Ie is a bed of clay to clayey fine sand

conformably underlying member 2 and overlying stratum lB. Stratum

Ie is from 15 to 110 cm thick and ranges from grayish brown to very

dark grayish brown (2.5 Y 5/2-3/2 m) and light gray (5 Y 6/2 m) with

either massive or faint, very thin (approximately 1 em), horizontal to

inclined and parallel bedding. A colluvial shore facies of gravelly mud

is often present along the periphery of stratum Ie (fig. 4).

Stratum Ie was originally interpreted as a flood-plain or

overback deposit (Johnson, 1974), but I now interpret the stratum Ie

clays and sandy clays as pond sediments. Along the central part of

the valley, the contact between members 1 and 2 is extremely abrupt

where diatomite (stratum 2A) overlies clay (stratum Ie). The contact

between strata lA and 2 along the basin margins is also extremely

abrupt, but only a color change exists between the units. The most

diagnostic characteristic of stratum Ie is in Trench 108 where stratum

Ie in the center of the basin is a massive, light-brownish-gray (2.5 Y

6/2 m), very fine sandy clay overlain by very dark gray (5 Y 3/1)

member 2 clay. In the east end of the trench 2.5 Y 6/2 m clayey fine

sand (stratum Ie) overlies light-olive-gray (5 Y 6/2 m) clay (stratum

2A). The contact between members 1 and 2 slopes 27 degrees shore ward, but basinward, stratum Ie is horizontal and has no appreciable

relief.

Four radiocarbon dates from member 1 establish the end of

fluvial deposition. The beginning of member 1 sedimentation is impre- 27

cisely known because the stratigraphically lowest 14C date, 12,650 ± 250

(1-246), is on Sphaerium and Pisidium from the middle of stratum IB;

Hereafter ca. 13,000 yr B. P. will be used for the initiation of member 1

deposition. The age of stratum 1A may be considerably older than

13,000 years. Initiation of stratum 1B deposition was estimated from a radiocarbon date of 12,150 ± 100 yr B.P. (SMU-295) on Pisidium and

Sphaerium shells from Area 2.

In a trench 1.5 km downstream from the Lubbock Lake site, a

12-cm-diameter tree limb from member 1 clay was 15 cm below the con tact between members 1 and 2. The radiometric age of the wood,

11,100 ± 80 yr (SMU-263), is the most reliable estimate for the end of member 1 deposition. Plant fragments from basal member 2 diatomite date 10,540 ± 100 (SMU-547). The average age of SMU-263 and SMU-

547, 10,800 yr, is the approximate age for the contact at members 1 and 2.

Several animal extinctions during member 1 time suggest dis conformities that are not sedimentologically visible. The gastropod record suggests that strata 1A and 1B have significantly different ages because stratum 1A gravds contain four terrestrial species that became extinct by stratum 1B time (Pierce, 1975). The locally extinct stratum

1A species are northern or montane taxa, or both p and include Vallonia cyclophorella, Pupilla muscorum, P. sinistra, and Vertigo gouldi basi dens (Pierce, 1975).

The most dramatic faunal extinction was in upper member 1 at approximately 11,000 yr B. P. and included the loss of Equus (horse),

Mammuthus (mammoth), Camelops (camel), Arctodus (short-faced bear), 28

and other large Rancholabrean fauna found throughout member 1. The

extinct late Pleistocene megafauna are restricted to stratum 03 and

member 1 of the Yellowhouse Draw formation.

Extinction of late Pleistocene fauna at the Lubbock Lake Site is

bracketed by four radiocarbon dates. The most reliable and uppermost

member 1 date is 11,100 ± 80 yr B.P. on wood (SMU-263). Basal stra

tum 2A dates are 11,190 ± 90 yr B.P. on shell (SMU-292), 10,530 ± 90

yr B.P. on humates (SMU-285), and 10,540 ± 100 yr B.P. on plant

fragments (SMU-547).

Member 2. Further evidence of lacustrine deposition is found

in member 2, a unit dating from 11,000 to 8,000 radiocarbon years B.P.

and representing the first of two major limnic episodes in Yellowhouse

Draw between 11,000 and 4,900 yr B.P. Member 2 is divisible into five lithologic 1:1nits that in upward stratigraphic order are: stratum 2A

diatomite, stratum 2B massive diatomaceous mud, stratum 2C peaty and prismatic mud, and stratum 2F, discontinuous olive sand lenses. Strata

2D and 2E were used between 1973 and 1977 but are now no longer valid because they are colluvial facies of strata 2A, 2B, and 2C.

Stratum 2F was proposed in 1976 (Stafford, 1977). It has not been renamed alphabetically because it could be confused with earlier 2D, 2E terminology. Member 2 is uniformly 60-90 em thick along the medial three-fourths of the valley and laterally thickens to 2 m along the valley margins.

1. Stratum 2A: The basal unit is stratum 2A, a 5-85-cm-thick bed of diatomite conformably overlying stratum Ie. Stratum 2A is divisible 29 into two lithologies, a lower 5- to 20-cm-thick bed of clean diatomite

and overlying beds of interlaminated black clay and diatomite up to 80

cm thick. When dry, the diatomite is white (10 YR 8/1) but changes upon wetting to a very dark grayish brown to black (10 YR 2/1), de pending on the amount of organic content.

The basal diatomite contains very faint, 1-2 mm-thick gray laminae (rhythmites), which can be traced shoreward until the diatomite intercalates with sand (fig. 10). The 67 individual diatomite laminae indicate the basal 10 cm of diatomite formed from at least 67 separate depositional events. It is unknown whether the periodicity was yearly or seasonal. The basal stratum 2A diatomite is rich in gastropods, and some accumulations suggest a coquina. Carbonized plant fragments less than 0.25 cm 2 are abundant along bedding and parting planes within the diatomite, whereas l-cm or less thick laminae of whole horizontal siliceous plant stems are often present between the diatomite and member 1. Diatomite laminae along the shore were used to estimate water depths during the earliest stage of the member 2 lake. If a single lamina were deposited at a Y/ater-sediment interface, relief on shore-basin laminae indicate that maximum water depths were 90 to 120 cm at 10,500 yr B.P.

Upper stratum 2A sediments are rhythmically interbedded diatomite and black mud deposited as beds a few millimeters to 1 cm thick. The contact between the interbedded diatomites and black muds and the overlying massive diatomaceous earth (stratum 2B) is abrupt to wavy and irregular or broken. The change from laminated to massive sediments was probably caused by bioturbation, because some exposures TRENCH 27 EAST WEST GEOLOGIC CROSS SECTION

IlfIrJifferf!nliolea

Slralllrn 2 5e(f; en/s rn diatomaceous mud

y. f-m. Sanely Cia

interlaminated .~am;te and dlc.gr. cloy

bosal diatomite

gry. br t -m. sand IB o 2 meters BLANCO FO~R;M;A~TrnIO~N~~~------

Figure 10. Geologic cross section of Trench 27 showing rhythmites and shore facies of member 2 of the Yellowhouse Draw formation

w o 31 of member 2 have laminated sediments in all 80 cm of an exposure, whereas adjacent trenches have 5 to 65 cm interbedded and the remain ing thickness as massive diatomaceous mud. I view the original con dition of member 2 as well stratified, the bedding being destroyed postdepositionally by crayfish, muskrats, and aquatic vegetation that produced a massive diatomaceous earth. Because the strata 2A-2B con tact is not a depositional boundary, stratum 2A is a rock stratigraphic unit whose upper surface is significantly time transgressive. Stratum

2A is much less bioturbated than stratum 2B. The stratum 2A bioturbation is limited to 2-3-cm-diameter circular disturbances and less common oval intrusions up to 12 cm in diameter. Crayfish may have caused the smaller burrows, whereas muskrats (Ondatra zibethicus) are possible causes for the large crotavinas.

Stratum 2A contains a carbonate facies only twice seen in Yel lowhouse Draw, both times at the Lubbock Lake Site. The exposures were trench 43 (fig. 11) and trench 108 (fig. 8), which show the carbonat~ unit as a wedge-shaped, basinward-thinning basal calcarenite with overlying micrite and very fine calcarenite (fig. 11). Interfin gering of the carbonates with basinal muds and diatomite of stratum 2A indicate that the carbonate is a lateral facies of much of stratum 2A.

2. Stratum 2B: Stratum 2B is generally massive diatomaceous earth, which overlies stratum 2A. The unit is very dark gray (10 YR

3/1 m) to black (10 YR 2/1 m) 20- to 95-cm-thick silty clay to clayey fine sand. As a rock unit, it is easily distinguished from other strata by its lack of both the fine interbedding seen in stratum 2A and the peaty or fibric structure of overlying stratum 2C. An occasional l-cm NW metere SE TRENCH 43 ----""""'------o • road cui _ 1936 5C~C====~4~r~-r~~~-.-~:::~ .".--- reserVOir 10m __ ..- _ .0;.,,"1--- 40 '" 'Ieooolion 2Cl.. j==7=R~I~O~Y~R~3/~I='= .~=~=~::~~::::::--~~~~:=:: %r=~ , ------olio. land 2A si t-rn 5Y1/2 rn- sandy Cloy 6\.t>.NCO 5 o )-me.ere- 5 olio. sand STAFFORD (1918)

Figure 11. Geologic cross section of Trench 43 at the Lubbock Lake Site showing shore facies of all strata and marl facies of member 2 in the Yellowhouse Draw formation. -- From Stafford (1981).

UJ N 33

or less thick lamina of plant stems is found in stratum 2B; the unit is

otherwise massive.

3. Stratum 2C: Stratum 2C, the uppermost bed of member 2, in

cludes peaty muds and columnar clays formed during cienega-pond

conditions within the valley. The most characteristic facies of stratum

2C is the fossil peaty zone, a 10- to 50-cm-thick, black (10 YR 2/1 m)

mud with to 5 to 70 percent aquatic plant stems and roots. In the

lower third of stratum 2C, the plant fossils are 0.5- to 2-cm-long frag-

ments, but in the upper two-thirds, the stratum contains whole vertical

stems and roots (fig. 4).

Although the plant remains are white, brittle, and easily

crushed, they retain minute internal and external stem morphology.

Except for carbonized remains in stratum 2A, most of the plant remains in the Lubbock Lake Site strata are white, siliceous, and postdeposi tionally oxidized. The fiber content of the muds indicates peat, and the plant remains will be considered oxidized peaty muds to distinguish them from undecomposed organics found 10-11 km downstream from the

Lubbock Lake Site. Carbonized organics occur along bedding planes within stratum 2A, but the white oxidized aquatics predominate. 7 The oxidized peaty facies of members belongs to soil S , a fibric histosol (Fibrist) 6 which is the oldest paleosol recognized at the

Lubbock Lake Site. The oxidized peaty muds are common only at the

Lubbock Lake Site and occupy the lower elevations of member 2 along the valley axis or on low-angle slopes along meander curves. In

Yellowhouse Draw, the Fibrists are absent upstream and occur for only 34

2-3 km downstream. The usefulness of these soils as time markers is

severely limited, and their regional extent is insignificant.

The second stratum 2C facies is black to dark gray (10 YR

4/1-2/1 m) clay and silty clay with a strong, very fine to fine prismatic

or columnar structure and less than 12 percent fossil plant remains. It

is unknown if the structure is from B horizon pedogenesis or from

repeated wetting and drying or a combination of both. The columnar to

prismatic structure is the dominant expression of stratum 2C throughout

most of Yellowhouse Draw. The oxidized peaty muds are a facies of the 7 columnar and prismatic clays and are included within soil S • Because

the structural muds have a sapric composition and less than 1 percent

organic carbon, they are classified as Aquepts instead of Saprists 0

4. Stratum 2F: Stratum 2F sediments are wedges of well-sorted

sand that conformably overlie member 2 and underlie member 30

Stratum 2F includes grayish-brown (2.5 Y 5/2 m) to yellowish-brown

(10 YR 5/4 m) 75-cm-thick fine to medium quartz sands that rapidly thin basinward and are intermittently distributed from the Lubbock Lake

Site to 3 km farther downstream. The sand wedges are usually 3 to 20 m or more long, are gleyed {reduced} at their lowest elevations, and become progressively more oxidized at the valley margin 0 Stratum 2F resembles small deltas or fans. They originate along the edge of

Yellowhouse Draw and are associated with faint to distinct erosion of the underlying stratum 2B muds. The texture, morphology, and basal contacts of stratum 2F are the best indicators of member 2 erosion ca 0

8,000 yr BoP. 35

Member 2 Radiocarbon Dates. Although member 2 has yielded nine 14C dates, there are still too few to calculate rates of deposition during the 3,000 years of sedimentation. The field location is precisely known for all but one date (C-558).

Two dates are from Evans and Meade's excavations; the remainder are from the 1974-1978 excavations. The date of 9,883 ± 350 yr B.P. (C-558) on charred bison bone collected by Evans and Meade was a solid-carbon whole-bone date. It was the first radiocarbon date on what Evans and Meade believed to be a Folsom-age Bison antiguus bone bed. Although the original bone bed is easily relocated and is now feature F AS-5, its vertical position within member 2 remains uncertain. The bone bed is within a black silty clay 60 cm below the contact between members 2 and 3. Immediately north of the feature,

Trench 64 shows at least 90 cm of member 2 underlying the bone bed.

Consequently, I believe that the C-558 bone date is from the upper third of stratum 2B and to date ca. 8,500 to 9,500 yr B.P. Although the C-558 date is reasonably accurate, it was originally assigned to the wrong stratum.

Another 1950 radiocarbon date wal:!l from the base of stratum 2A where Helisoma and Lymnae (probably Lymnae reflexa) shells were dated as 9,700 ± 450 yr B.P. (L-238g; Broecker and Kulp, 1957). Green

(n.d., p. 35) noted that Evans and Meade had given the provenance of the shells as ". • • from the Folsom horizon and slightly higher than the charred bone sample." Green discovered during his excavation at

Blocks X, Y, and Z (adjacent to Evans and Meade's Station C) that the gastropods were actually from ". • • the pure diatomite of Zone a," 36

that is, the basal diatomite of the present stratum 2A. The L-238g

date is therefore at least 1000 years too young because: (1) It under

lies in situ Folsom deposits, (2) overlies by only a few centimeters the

11,100 ± 80 yr B. P. wood date from member 1, and (3) disagrees with

several dates in the 11,000- to 10,000-yr-B.P. range.

Three radiocarbon samples from stratum 2A were processed

after 1974. One date was on Lymnae reflex a shells collected from the

same diatomite used for L-238g. The shell date is 11,190 ± 90

(SMU-292) and included shells from the entire 10-cm thickness of the

pure diatomite at the base of stratum 2A. The SMU-292 sample was

collected in a trench a few meters east of Area 6 and from a bed strat igraphically identical with stratum 2A, local bed 1 in Area 6.

The second date was on organics from a 9-cm-diameter horizontal core from the pure diatomite bed at the base of stratum 2A.

Area 1 (fig. 10). The base of the sample was 7 cm above the contact

of members 1 and 2; the humates dated 10,530 ± 90 (SMU-285).

The third date, from stratum 2A basal diatomite in Area 3, was significantly younger than the two previous stratum 2A dates and had values of 7,840 ± 140 (SMU-247) on residue and 10,060 ± 70 (SMU-251) on humates. I eliminate the 7,840 yr B.P. date because: (1) It is incompatible with the Folsom archaeology of stratum 2A and (2) it con flicts with the overall 10,000-8,000 yr B. P. chronology of member 2.

The humate date, although substantially younger than the other stratum

2A dates, may not be significantly erroneous if the sampled 18-cm thickness of the basal diatomite is considered. Such a thickness would yield an average of old and young humate fractions. 37

A carbon-14 date from the middle of member 2 in Trench 27 gave 9,960 ± 80 (SMU-275) on humates. The sample was from very fine sandy clay 60 cm above the member 1-2 contact, and extrapolation showed that SMU-275 was overlain by approximately 80 cm of member 2 mud and underlain by 60 cm of sediment.

The end of member 2 deposition was approximated by using two radiocarbon dates, both of which were from stratum 2C cienega.

Oxidized peaty muds in Area 3 are 21 cm thick. Upper and lower sections were used; humates from the upper 9 cm dated 7,970 ± 80 yr

B. P. (SMU-262) and humates from the lower 12 cm of the fossil marsh dated 7,890 ± 100 yr B.P. (SMU-302).

A wealth of paleontological data supports a lacustrine and cienega origin for member 2. Twelve species of aquatic gastropods and abundant Sphaeriid clams and Cypridid ostracodes were present in member 2 (Pierce, 1975). Plant remains from member 2 are limited to seeds, large amounts of roots and stems, and numerous but usually unidentifiable stem molds and impressions. Thompson (l977) identified several aquatic seed genera, including bullrush (Scirpus): 1,720 seeds; spikerush (Eleocharis): 2 seeds; and water lily (Nymphaea): 1 seed.

Other plants identified from seeds were goose foot (Chenopodium), nightshade (Solanum rostrum), Gaura, hackberry (Celtis reticulata), and devills claw (Proboscidea). Macrofossils of Scripus are most common in stratum 2C, and 1-3-cm-Iong stems are commonly preserved.

Additional 'macrofloral remains have been tentatively identified as

Equisetum. Plant impressions are most abundant along horizontal bedding planes within diatomite and clay in stratum 2A and occur as 38

either impressions (external molds) or molds containing traces of

carbonized plant material. Large (1-cm) fragments of lignitized plants

are disseminated throughout the diatomite; one concentration from the

base of 2A in Trench 109 was sampled for radiocarbon dating. The

only plant impression identifiable to genus was the alga Chara cf.

vulgaris found in a fine clay at the base of stratum 2A.

Johnson (1976) identified several aquatic or water-favoring

genera including catfish (cf. Ictalurus), leopard frog (Rana pipiens),

bull frog (Rana catesbeiana), salamander (Ambystoma), green water

sm:.ke (Natrix cyclopion), snapping turtle (Chelydra serpentina), yellow

mud turtle (Kinosternum flavescens), Sora (Porzana), rail (Ra1lus),

marsh hawk (Circus cyaneus), and green-winged teal (Anas cf.

carolinesis) ,

At ca. 11,000 yr B.P., the water depth in member 2 ponds is

estimated to have been 90 to 120 cm; by 8,000 yr B.P., the depth had

decreased to 10-30 cm. The lakes were probably filled with water from

springs within or on the valley edges. Paleontologic, geologic, and

historic data suggest this origin. Examples of spring activity are the

diatom Pinnularia micro&tauron (Ehrenberg) Cleve, a diatom found today

only in springs (Compton, 1975), fossil cienega soils, and calcium

carbonate wedges in stratum 2A. The carbonate resembles modern car bonate spring mounds described by Reeves (1968). Historic records

mention springs near the Lubbock Lake Site as recently as the early

1900s (Holden, 1974), and marshy areas appear in 1939 aerial photo

graphs of the site (Wheat, 1974, p. 18). 39

Unanswered is how the valley was dammed to form ponds over

90 cm deep. Member 2 lacustrine sediments occur along a 60-km-Iong segment of Yellowhouse Draw and over a relief of 240 m. The most probable origin for the ponding is that natural depressions and irregu larities remaining at the top of stratum 1B filled with water and the ponds eventually became self-enhancing by aquatic vegetative growth.

Member 3. The third member of the Yellowhouse Draw forma tion is member 3, a series of 4,900-8,000-yr-B.P. lacustrine clays, marls, and cienega soils deposited by lakes having a water chemistry different from that in previous ponds. Five lithologies are mappable in member 3. They are grouped into three strata (3A, 3B, and 3C) and have a total thickness of 100 to 180 cm. The thickest accumulations are along the western valley edges where eolian and colluvial deposition was active. A~though the top of member 3 is an erosional surface, radio carbon dates suggest that erosion was brief and decreased in magnitude downstream.

Member 3 conformably overlies member 2 in mid-valley and dis conformably overlies member 2 at valley margins (fig. 12). The boundary between members 2 and 3 is abrupt if fossil peats and peaty muds (stratum 2C) are overlain by member 3; within 2-cm vertical distance, the plant content of the muds decreased from 70 percent or more in stratum 2C to 5-7 percent in stratum 3A. Plant content in member 3 diminishes upward until less than 1 percent plant fragments exist in upper member 3. Where stratum 2C is columnar and prismatic and peats are absent, the contact between members 2 and 3 grades over WEST

TRENCH 18 GEOLOGIC CROSS SECTION ...r:

BLANCO FORMATION olive m. sandy clay

Figure 12. Geologic cross section of Trench 18 at the Lubbock Lake Site sl disconformities within section

EAST MODERN GROUND SURFACE dk.gr: (5Y4/1) f. sand with 0.4 to I cm carbonate clasts gr: br.(2.5Y5/2) m. sandy silty clay with carbonate clasts to 5 cm 5 C v. dk.gr: (5Y3/lHsandy clay with carbonate clasts to o.5cm . dk. Qr.C5Y4/D ON m. sandy clay with 0.4 to 2 em carbonate clasts; bed of 2-4cm dia. carbonate pebbles,one clast 0 0 0 0 o 0000 0 -'-r-J- .L..rl- <0 C --'- ,.. ,- C) -'-r 0

--'-- -.- c::::, ~

o 2 meters

BLANCO FORMATION olive m. sandy clay

the Lubbock Lake Site showing shore facies of the Yellowhouse Draw formation and

2 to 10 cm vertically, the color of member 3 lightens, and the columnar structure becomes finer and weaker upward. The contact between members 2 and 3 may be indeterminate where both members are struc tureless and the color of the unstratified muds differ only from 2.5 Y

3/2 m to 2.5 Y 5/2 m over a vertical distance of 20 to 30 cm.

1. Stratum 3A: Stratum 3A at the Lubbock Lake Site is an 80-

150-cm-thick bed of grayish-brown (10 YR 5/3-5/3 m) noncalcareous to moderately calcareous silt-clay. Plant fossils are less than 7 percent by volume, and fine vertical silicified roots are common only in the basal 10 cm. An unexplained but characteristic structure of stratum 3A is fine continuous random tubular pores that occur throughout the sediment.

At the Lubbock Lake Site, the carbonate content of stratum 3A gradually increases upward and eventually forms the thick dense cal cium carbonate unit, stratum 3B (Appendix). Although exposures with in the reservoir area show this vertical lithologic succession, a more accurate stratigraphy exists in Trench 108 where calcium carbonate con tent in member 3 increases both upward and laterally away from the axis of the valley.

2. Stratum 3B: Stratum 3B contains white calcareous clay grading into clayey micrite (marl) that attain composite thicknesses of 85 cm. A moderate very fine sub angular blocky structure grades upward into strong fine to medium platy as carbonate content increases (fig. 7).

Where the carbonate percentage is greatest, faint to distinct antiforms and synforms develop with wavelengths and amplitudes of 10 cm.

3. Stratum 3C: The uppermost unit, stratum 3C, contains two apparently contemporaneous lithologies. The more widespread is a 42

25-50-cm-thick, very dark gray (10 YR 3/1 m) soft silty clay interpreted to be a cienega soil conformably overlying stratum 3B. The second lithology is a hard, strong, medium to coarse angular blocky, dark-grayish-brown (2.5 Y 4/2 m) clay disconformably overlying strata

3A and 3B. The blocky clay is also found in shallow « 10-cm) depres sions within cienega soil S6 and contains 5 percent 1-5-cm-diameter concentrations of autochthonously precipitated calcium carbonate similar to the carbonate muds in stratum 5C.

Member 3 sediments coarsen along the valley margins, particularly the western windward borders where the colluvial facies is more sandy and gravelly than member 2 colluvium. Member 3 colluvium extends further toward the valley axis than member 2 colluvial sediments. The member 3 colluvium is slightly more brown (10 YR 4/6 m) than its cC?eval basinward sediments (2.5 Y 5/2 m) and includes carbon ate rock clasts up to 1 em in diameter.

A radiocarbon age for member 3 was determined from dates on bison bone from stratum 3C blocky clay in Trench 65. The clay was the uppermost unit of member 3 and disconformably overlaid stratum 3B marl. The bone date was 4910 ± 50 yr B.P., which was an average of

4680 ± 50: apatite (SMU-309), 4880 ± 80: collagen (SMU-332), and 4950

± 60: collagen residue (SMU-333).

The depositional environment of member 3 is more difficult to determine than that of any alluvial unit because diagnostic sedimentary structures are rare, stratification is poor, and paleontological evidence is sparse. Depositional environments hypothesized for member 3 include 43 pedogenic (Johnson, 1974), brackish bog (Hohn and Hellerman, 1961), and damp ground (Wendorf (1961).

Large vertebrates are extremely rare in member 3. Only two concentrations of bison bone have been found at the Lubbock Lake Site, and together they total less than 50 skeletal elements. Antelope

(Antilocapra americana) is the only other large mammal found in this member. An additional 14 vertebrate species of small carnivores, rodents, and assorted herpetofauna give member 3 a significantly lower species diversity than member 2, which contained 54 vertebrate species.

The most diagnostic species from member 3 were yellow mud turtles

(Kinosternum flavescens) and red-bellied water snakes (Natrix erythrogaster), two marsh and pond inhabitants (Johnson, 1976).

Fossil seeds from member 3 are scarce and include Scirpus and water lily (Nymphaea) from stratum 3B (Thompson, 1977), and external molds of cocklebur (Xanthium) seed pods from the stratum 3C blocky clay.

Although the meager vertebrate and floral evidence suggests at least partial ponding, snails and diatoms provide the most conclusive evidence for pond-marsh conditions between 8,000 and 4,900 yr B. P.

Member 3 contained four &quatic gastropods: Gyralus circumstriatus

Tyron (15%), Physa gyrina Say {31~), Aplexa hypnorum (1%), and

Lymnaea humilis (53%). In comparison, member 2 had 10 species of aquatic snails (Pierce, 1975). Pierce also found that Sphaeriid clams had become rare in member 3, whereas Cypridid ostracodes remained common. Diatoms decreased from 48 species in member 2 to 27 taxa in 44

stratum 3A. Hohn and Hellerman (1961) interpreted the absence of

planktonic diatoms to indicate boglike areas with shallow water.

A pedogenic origin for strata 3A and 3B carbonate is rejected

because (1) the carbonate occurs in more than one distinct bed, (2) the

carbonate varies laterally in percentage and laterally interdigitates with

eolian sands, and (3) the invertebrate fauna indicates moist conditions

during deposition of member 3. A pedogenic origin is hypothesized for

the uppermost dark-gray friable silty clays (soil 8 6) of member 3, be-

cause of their similarity to modern cienega soils in Yellowhouse Draw.

Playa deposition is not a tenable explanation for member 3 be-

cause (1) gypsum and halide crystals or their molds are absent, (2)

multiple unconformities and intraformational conglomerates are absent,

(3) cienega soils and marls are facies of one another, and (4) the

strata 3A and 3B lithologies extend the studied length of Yellowhouse

Draw (>60 'km) and are not limited to small basins.

The depositional environment most compatible with the data from

member 3 is a hard-water marl-depositing lake. The following facts

support this interpretation: presence of aquatic vertebrate and invertebrate fossils, the lateral and vertical changes in carbonates, the horizontal gradation of marl into cienega muds, the interfingering of marls with fringing eolian sands, and the vertical gradation of member 2 into member 3.

In summary, the data indicate a period of accelerating drying beginning ca. 8,000. Vegetative changes in the uplands may have

caused increased runoff and coarsening of colluvium and accelerated 45 eolian deflation. Increases in surface runoff and eolian deposition would have significantly increased the amount of calcium carbonate washing into the valley. As the water chemistry of the ponds and marshes changed, biological productivity would decrease and marl dep osition predominate. The cienega soil, S6, may represent a local decrease in calcium carbonate influx into the valley and a return to more organic sedimentation. Despite this late shift to less calcareous conditions, the top of member 3 marks the end of at least 8,000 years of continuous high water table conditions within Yellowhouse Draw. Ciene- gas disappeared by 4,900 yr B. P. Similar palustrine environments would not reoccur in Yellowhouse Draw until at least 2,600 yr B.P.

The Yellowhouse Draw formation contains a regional disconformity recognizable in Yellowhouse and Blackwater Draws (figs. 8 and 13).

The disconformity divides the late Quaternary alluvium into two genetically distinct rock stratigraphic units. Member 3 of the Yellowhouse

Draw formation contains fluvial and lacustrine sediments, and member 4 contains fluvial, eolian, and colluvial sediments separated by several paleosols.

Although the two members record a marked sedimentologic shift, the disconformity between the members represents a short time interval.

The post-member 3 erosion in Yellowhouse Draw at the Lubbock Lake site is dated at 4,900 yr. B. P., based on two radiocarbon dates from the top of member 3 (4,900 ± 60 (SMU-531), 4,910 ± 50 (SMU-332, 333), and a hearth date near the base of member 4 (4,960 ± 80 (SMU-492).

At the Lubbock Lake Site, the contact between members 3 and 4 is extremely abrupt and wavy and the surface of member 3 is often LONGITUDINAL AND GEOLOGICAL CROSS SECTIONS FOR YELLOWHOUSE AND BLACKWATER DRAWS LLANO ESTACADO t TEXAS TRENCH III VALLEY PROFILES X200 VE /JE STRATIGRAPHIC SECTIONS X50 VE W ----fill

en W ...J ii: o TRENCH 110 a:: 11. >- 10 W ...J ...J« 1021 >

in II: w ~6 :E ~ SCALE FOR ~ STRATIGRAPHIC ei CROSS SECTIONS >o~====::::;.! o 400 HORIZONTAL (KILOMETERS)

,--,-- 1-' I , I I I I I I , o 10 20 30 40 50 60 10 80 90 100 110 120 SCALE FOR VALLEY PROFILES (KILOMETERS)

Figure 13. Longitudinal and geologic cross sections for Yellowhouse and Blackwater Draws a-~ 47 channeled and deeply eroded. The contact is distinct farther south in

Yellowhouse Draw but is clear to diffuse in the remainder of Yellow house and Blackwater Draws. In the latter examples, the change from lacustrine to eolian deposition is unequivocal; only the contact has become obscured by later bioturbation and pedogenesis.

Member 4. Member 4 includes two lithologic associations. One dominates Yellowhouse Draw from the Lubbock Lake Site to Dam 6 and includes basal olive sands and clays overlying reddish-brown sands and an uppermost soil A horizon. The second lithofacies is massive oxidized brown sands with a thick cumulic soil A horizon. The latter facies is far more extensive and is found along the western edges of the valley downstream from the Lubbock Lake Site and in the remainder of Yellow house Draw and all of Blackwater Draw.

In. contrast to earlier usage where member 4 (stratum 4) was separated into two or three subunits (Green, 1962; Johnson, 1974), member 4 is not subdivided in my study. Previous subdivisions 4A,

4B, 4C (Johnson, 1974) were based on postdepositional rather than depositional criteria and are not useful stratigraphic markers. In earlier terminology, stratum 4A was reduced olive sand designated a soil

C horizon, stratum 4B was oxidized red sand with calcic and cambic horizons, and stratum 4C was the soil A horizon.

The thickness of member 4 at the Lubbock Lake Site ranges from 170 cm at the type section (Appendix) to 220 cm along the western edges of the canyon (fig. 7). Its sedimentary origin is often indeterminate because most bedding and sedimentary structures were 48

destroyed by pedogenesis and bioturbation. A t least half of the

thickness of member 4 at the Lubbock Lake Site is fluvial, where the

basal part of member 4 is light-olive-gray (5 Y 6/2 d) fine to medium

sand and clay. Thin horizontal interbeds of 0.5- to 2-cm-thick clay

and coarse silt are the common type of stratification, whereas recum-

bently folded cross-stratified and small-scale trough cross-stratified fine

sands are less abundant. The olive sediments are either in extremely

abrupt, smooth to wavy contact" with member 3 of the Yellowhouse

formation or are intermixed by soft sediment deformation with the

stratum 3C muds. The lower 5 to 25 em of the olive bed is commonly

brown (7.5 YR 5/4 d). The oxidation-reduction front is postdeposition-

al because the redox alteration cuts across sedimentary structures,

particularly where basal member 4 is sandy.

The olive sands and" muds of member 4 are the C soil horizon of 5 paleosol S. The gleyed unit contains up to 20 percent fine to medium,

distinct to prominent carbonate mottling coincident with the bedding

planes of the clay and silt. This Cca horizon is unrelated to the over lying S5 soil and represents either the calcic horizon of a truncated buried soil not presently recognized or is a primary deposit of calcium

carbonate formed by local ponding or elevated ground-water conditions.

The olive color of member 4 persists along the axial three-quarters to four-fifths of the valley. but along either valley wall the sediments are oxidized and the color changes from basinal 5Y hues to strong browns

(7.5 YR 5/6 d) over a horizontal distance of 30 to 40 m.

The greater part of member 4 within the Lubbock Lake Site is included in the A and B horizons of soil S5, the most strongly 49

developed of any buried soil at the site and one containing a mollic

epipedon and cambic and calcic horizons. The soil is classified a

Calciustoll.

In all exposures examined, soil S5 is polygenetic, composite, or

both because its uppermost calcic horizon (Aca) was inherited from

younger (S4) soil pedogenesis. Soil S5 includes the Al horizon of 54 in exposures of S5 to the west of the site. No monogenetic examples of

soil S5 have been found.

At the type section (Appendix), the base of the S5 A horizon

contains an upper calcic unit (Aca), which together with the All

horizon is separated from the principal calcic horizon (B2ca) by a reddish-brown cambic horizon having 110 secondary carbonate. The

B2ca for S5 is 50 cm thick with stage III carbonate morphology and is

the only calcic unit in member 4 formed during S5 pedogenesis.

Beneath the B2ca is a brown oxidized B3 with no secondary carbonate.

The B2ca, B3, B24ca, and B23ca horizons have faint light-olive-gray mottling. With increasing depth, olive mottling increases, the amount of brown matrix diminishes until the sediment becomes light olive gray at the base. The olive mottling within the B horizon suggests that 50 percent of member 4's thickness may have undergone postdepositional oxidation-reduction.

The lithology and pedogenetic horizonation of soil S5 change

considerably from the type section along the western 100 m of the valley. The most significant changes are: formation of a composite soil profile by merging of the Al horizons of soils S4 and S5, thickening or

cumulization of the A horizon of 55, an increase in sand content, a so diminution and attenuation of the Aca horizon, and a 1. S-times increase in the overall stratigraphic thickness.

Thick cumulic profiles and sandy lithologies are characteristic of member 4 north of the Lubbock Lake Site and in Blackwater Draw.

Trench 109 (fig. 6) illustrates member 4 and its sandy, cumulic multiple calcic horizon and oxidized character.

Member S. Between 2600 ± so (SMU-697) and 4960 ± 80

(SMU-492) yr B.P. stream erosion along the axis of Yellowhouse Draw at the Lubbock Lake Site removed all or most of member 4 and an unknown thickness of member 3 to form a 80-1S0-m-wide, 1.6-m-deep channel. Following the erosion, cienegas developed at the Lubbock

Lake Site for the first time since 4900 yr B. P., and soon afterward colluvial and eolian sands were deposited as slope facies of the paludal muds. The muds and sands constitute member S, the uppermost member of the Yellowhouse Draw formation.

Member S is divisible into three units, strata SA, SB, and SC.

Strata SA and SB are eolian and colluvial pale-brown sands and clayey sands intermittently deposited during several sedimentary cycles. Most were coeval with the deposition of cienega clays and marls in the valley lowlands. Strata SA and SB temporally succeed one another, and both interfinger laterally with the stratum SC marsh muds.

Strata SA and SB represent two major periods of eolian and col luvial deposition, each of which ended in the formation of a moderately well developed soil. Weaker soils within the sands record shorter 51 1 periods of stability, and four soils were mapped in member 5: Sand

S2 in stratum SB and S3 and S4 in stratum SA.

Stratum SA consists of brown to light-yellowish-brown clayey to silty, very fine to medium quartz sands with discontinuous one-clast- thick laminae of O.S-cm carbonate pebble gravel. Within the stratum's

90-cm thickness, a~ least four pebble horizons occur. Each pebble bed is interpreted to be a minor unconformity or depositional surface because concentrations of lithic artifacts commonly coincide with the pebble horizons.

The uppermost soil of stratum SA is S3, a moderately well developed Haplustoll with a lS-cm-thick A horizon, a cambic B2 and

5-20 percent filamentous carbonate in the basal C horizon. A second soil, S4, occurs toward the base of SA and is an Ochrept with a 7-cm thick A horizon and a cambic B2 overlying member 4. Although S4 appears to' be a very weak soil, it may be more developed than S3 if the calcic horizon (Alca) of sS belongs to S4.

Stratum SB is finer grained and thinner than SA. The pebbly sands in SB are restricted to its base, whereas the remaining sediment lacks gravel. The soils of stratum SB (Sl and S2) are at the top of the unit and include the modern surficial soil, Sl. Soil Sl is an

Orthent with a 1- to 3-cm-thick A horizon and a thin cambic B2. Soil

S2 is a Haplustoll underlying Sl with no intervening C horizon. It is 1 1 better developed than S , because the S A horizon is twice as thick and B2 and B3 horizons exist in S2. There is no field evidence of pedogenetic carbonate accumulation in either Sl and S2. 52

Where traced upslope to the west, S 2 through S 5 show succes-

sive truncation by younger depositional units and development of sur-

faces of increasingly lower angle to horizontal. At the extreme western

edge of the valley, soils 52, S3, and S4 have been eroded and S5 lies

immediately below S2 and within 30-40 cm of the modern surface (fig.

7) •

Stratum 5C was deposited along the valley axis within a 1.7-

m-deep channel eroded into members 3 and 4 of the Yellowhouse Draw

formation. The stratu.m 5C marsh sediments include 30 cm of very dark

gray clay and lesser amounts of marl and calcareous clay (fig. 8) that

partially predate 5A and 5B deposition and pedogenesis but are mostly

coeval with them. The depositional chronology of 5C was: (1) basal

black (5 Y 5/2) clay, (2) intermediate mud, and (3) uppermost, very

dark gray clay to sandy mud. The basal black clay disconformably

overlies either members 2, 3, and 4 of the Yellowhouse Draw formation

and changes vertically to marl or weakly to massively bedded dark-gray

muds. The upper surface of 5C shows the beginning of pedogenesis,

i.e., a slight brown coloration possibly dating to the post-1940s des-

iccation of the valley. The slope sands and marsh muds interdigitate as

illustrated by figure 7, which shows the contemporaneity of marsh and

colluvial environments.

Hearths and bone beds in member 5 haye been radiocarbon

dated. The first radiocarbon determinations for member 5 were two

dates on charred bison bone collected by Green (1962) in the lower third of his stratum 6, presently stratum SA in this dissertation. The

dates were 70 ± 70 yr B. P. (I-140) and 160 ± 90 yr B.P (1-208) 53

(Trautman and Walton, 1962). Two superimposed hearths were exca vated in stratum SA as feature F A8-6 lying between soils 53 and 54 in the center half of stratum SA. 1 believe that FA8-6 is stratigraphically very close to Green's (1962) two dated bones. The upper hearth dated

505 ± 55 yr B.P.: charcoal (51-2700 ) and the lower hearth dated 505 ±

55 yr B.P. (51-2701) (Kaczor, 1978). The close agreement between the

FA8-6 dates suggests that the bone dates are in error. Another date,

160 ± 60:humates (5MU-343), available for member 5 is a soil date pre- sumably on 5 3 from Trench 65. The exact field position of the sample is known; however, the soil is the only A horizon between fill and member 4. The moderate soil development of the dated horizon re- 3 sembles 5. Because the date is from a soil whose stratigraphy is imprecisely known, considerable caution is advised in using the date.

An additional date from stratum 5A has no additional provenance; the date is on' hearth charcoal from feature FA15-1 that dated 285 ± 60 (51-

2703). Another hearth, known only to be from member 5, dated 315 ±

50:charcoal (51 2704). Both the 51-2703 and 51-2704 samples were associated with Garza projectile points. Finally, a hearth from either stratum 5A or 5B (stratigraphy is unknown) yielded an apparently con taminated date of 100.03% modern 14C (51-2702).

The contrasting relationships of lithologies and soil stratigraphies in member 5 give paleoclimatic data for the late Holocene in

Yellowhouse Draw. The long period of pedogenesis that formed soil 55 ended when stream downcutting occurred ca. 2600 yr B.P. Afterwards cienegas returned to Yellowhouse Draw for the first time since 4900 yr

B • P. From at least 600 yr B. P. and probably 1000 yr B. P. to the 54 present, eolian and colluvial deposition were contemporaneous with the deposition of stratum 5C marsh muds. The contemporaneity of eolian and marsh sedimentation appears contradictory, but the processes have independent controls. The marsh environments were controlled by fluctuations in the valley's perched water table, whereas slope and upland stability were controlled by the regional water table of the uplands.

Geology of Upper and Lower Yellowhouse Draw and Blackwater Draw

Alluvial geology for the remainder of Yellowhouse and Black- water Draws is explained by using the Lubbock Lake Site as a refer- ence. Although the sedimentology throughout the two valleys varies little from that at the Lubbock Lake Site, the stratigraphic complexity is different. The stratigraphy at the Lubbock Lake Site is deceptively more complex than regional trends indicate.

Alluvial Stratigraphy of Yellowhouse Draw Downstream from the Lubbock Lake Site

Trenching and boring data are available downstream from the

Lubbock Lake Site as far south as the Fort Worth and Denver Railroad bridge (fig. 2). Little is known of the depth to bedrock or the detailed stratigraphy of member 1 in this interval because the valley water table is high and prevents excavation to bedrock. Only the upper 5-10 cm of member 2 are directly observable at Dam 6. Where member 1 could be examined, only its upper sands and their contact with basal conglomerates wero seen. The uppermost member 1 sediments were lithologically equivalent to strata IB and Ie at the Lubbock Lake Site and were light-brownish-gray (2.5 Y 6/2 m) clayey very fine to fine 55

sand or light-gray (2.5 Y 7/2 m) silty clays whose maximum observed

thickness was 130 cm. The member 1 sands and clays were in abrupt

contact with carbonate pebble gravels that were 20 cm thick and con

sidered equivalent to basal stratum lA gravels.

A log found 1.5 m downstream from the Lubbock Lake Site and

in the upper member 1 sands dated 11,100 ± 100 yr B.P (SMU-548).

The' tree limb, 7 x 12 cm in cross section, laid horizontally with its

upper surface 15 cm below the contact of members 1 and 2 and was

enclosed in clayey very fine to fine sands.

Member 2 varies little from Lubbock Lake Site exposures. The

principal difference is the lack of observable fine internal stratification

due to water saturation. Thicknesses of member 2 were from 150 to

180+ cm, although 250 ± 10 cm was most common along the valley axis.

The lithology was very dark gray (10 YR 3/2) to olive (5 Y 4/3) clays,

silty clays' and very fine sandy muds in the basin and coarser sandy to pebbly muds toward the valley edge. The contact between members 1 and 2 was abrupt and conformable with a predominance of diatomite at the base. Although an upper cienega (stratum 2C) was visually absent and members 2 and 3 graded into each other, 20-25-cm-thick prismatic clays were found with 10-15 percent vertical silicified rootlets.

Excepting the most basinward locations, the black muds of member 2 were often separated from member 3 by the very fine sands to gravels of stratum 2F. Stratum 2F sediments were at least 5 times more common downstream than at the Lubbock Lake Site. The stratum 2F sediments were usually olive (5 Y 5/3) to olive gray (5 Y 5/2), The valley margin deposits were very pale brown (10 YR 7/3) to brown (7.5 Y 56

5/4) and were thoroughly oxidized. Sandy and gravelly stratum 2F deposits were up to 65 em thick. Where the sediments were fine sand and coarser, they disconformably overlaid member 2 into which channels

50 cm wide and 40 cm deep had been cut. Some channels trended 60 to

90 degrees to the valley axis and gradually diminished in relief toward the basin. Where stratum 2F thinned and fined to very fine sands and clays, its basal contact with black member 2 muds was abrupt but not obviously erosional. Only by tracing the sands shoreward was the dis conformity evident.

Member 3 downstream from the Lubbock Lake Site is identical with exposures at the Lubbock Lake Site. The member ranges from 110 to 190 cm in thickness and is divisible into the upper cienega soil, stratum 3C, and underlying marls and calcareous clays of strata 3A and

3B. Strata 3A and 3B were combined because vertical and horizontal gradation precludes differentiation.

The base of member 3 is conformable with member 2 in the center of the basin, and the two members are often inseparable over a vertical distance of 40 em. Member 3 is dominantly clay to clayey silt, with medial marls and calcareous clays. Colors range from light gray

(10 YR 7/1) to light brownish gray (2.5 Y 6/2) for clays, white (10 YR

8/1) for dense marls, and very dark gray (10 YR 3/1), gray brown (10

YR 5/1), or dark gray (5 Y 4/1) for stratum 3C. Stratum 3C changes little in thickness, usually being 20-50cm. The thickness of strata

3A-3B sediments varies the most and ranges from 60 to 155 cm. The range exist~ because there is often difficulty establishing the contact between members 2 and 3. 57

Members 4 and 5 most closely duplicate Lubbock Lake Site descriptions. Only minor differences in thickness and stratigraphic complexity were noted. Member 4 retains the same disconformable base and strongly developed calcic profiles described for the type section at

Lubbock Lake Site. The only measurable change is an increase in the thickness of member 4 to 210 to 230 cm.

Member 5 is divisible into three stratigraphic members: the two sandy alluvial facies, strata SA and 5B, and a basinal contemporaneous cienega clay, stratum 5e. The principal difference from the Lubbock

Lake Site is the absence in strata SA and 5B of four soil horizons.

Instead, three pedogenetic horizons are recognizable, one atop stratum

SA and two in upper stratum 5B. These soils appear to correlate with soils SI, S2, and S3 at the Lubbock Lake Site; the missing soil is assumed to be S4, itself a very weakly developed and often absent soil at the Lub"bock Lake Site. Another similarity to the Lubbock Lake Site is the abundance of Bison bison and occasional Equus caballos remains within the upper stratum ~C cienega. An example of this association is at the Nash Site, a kill site 2.2 km downstream from the Lubbock Lake

Site. At the Nash Site butchered bison and horse were recovered over a O. 6-km distance along the vaney axis.

Alluvial Stratigraphy of Yellowhouse Draw Upstream from the Lubbock Lake Site

Upstream from the Lubbock Lake Site, members 1-3 of the Yel- lowhouse Draw formation are easily recognized; however, members 4 and 5 are undifferentiable upstream from Trench 109. The alluvial sedimentary environments as far northwestward as Trench 111 are identical 58 to those in Trench 108. The most significant changes along the 47 km in Yellowhouse Draw are the valley bedrock cross-sectional geometry, the progressive erosion of large volumes of member 4 and some of member 5 and the disappearance of all but one recognizable paleosol in the modern to 8000 yr B. P interval.

There are several similarities between the Lubbock Lake Site and upstream localities: the presence of all Yellowhouse Draw formation members, persistence of the disconformity between members 3 and 4 and ability to assign Yellowhouse Draw and Lubbock formation nomenclature as far upstream as Trench 111.

Trench 109. In Trench 109, valley width decreased to 78 m and the strata are concomitantly compressed laterally and to some degree vertically (fig. 7). The narrow width enabled sandy colluvium from the ~plands to spread across the valley and mask the lithology of original sedimentary environments.

The oldest alluvium found in Trench 109 was a remnant terrace, stratum 03, resting high on the bedrock (fig. 7). The buried terrace contained indurated quartz and carbonate sand conformably overlain by very fine carbonate pebble gravels. The elevation of the terrace implies that a minimum of 4 m and as much as 8 m of erosion occurred prior to ca. 12,500 yr B.P.

Members 1 and 2 are lithologically identical to those at the

Lubbock Lake Site type section (Appendix), even to the inclusion of a basal diatomite in member 2. Plant fragments at the base of member 2 dated 10,540 ± 100 (SMU-547). Member 3 is discernible despite 59 replacement of its upper calcareous clays and marls by sands. At the extreme western edge of the valley member 2 has been eroded. In the center of the valley, the lower bed of member 3 lies conformably upon member 2.

The disconformity between members 3 and 4 is distinct, and the erosional surface is overlain by members 4 and 5 sands. Stratum 5C cienega clays are absent. Only two soils were differentiable in members

4 and 5. One soil is a 5-cm-thick modern soil, and the other is a 180-

240-cm-thick buried soil differentiable into horizons only in the center of the valley. A massive cumulic A horizon is the only recognizable horizon along the valley walls.

Trench 114. At Trench 114, 14.8 km upstream from Trench

109, members 1,.2, and 3 and stratum 5C were exposed. Thicknesses of member 1 equaled downstream thicknesses, and there was a basal stratum 1A conglomerate 50 cm thick overlain by 85 cm of carbonate gravels interfingered with stratum 1B sands.

Member 2 was lithologically similar to that of Trench 108, except that internal bedding was not pronounced and the unit graded westward into a sand. Most of member 3 had been removed by erosion, but the remaining 60 cm were identical to downstream member 3 exposures.

The greatest difference in Trench 114 stratigraphy from that at the Lubbock Lake Site is the major erosion that removed most of member

3, part of member 2, and all of any deposited member 4. In the re sulting channel, 2.1 m of stratum 5C clay were deposited. This 60 thickness is greater than at any previous exposure of stratum 5C seen in Yellowhouse Draw.

Trench 111. The most upstream trench in Yellowhouse Draw is

Trench 111 at Anton, 47.1 km from the Lubbock Lake Site. At this locality, only members 1, 2, and 3 of the Yellowhouse Draw formation were observed (fig. 13). The 12 em of member 1 include lateral interfingering carbonate pebble gravels and sands, but no basal conglomerate exists. Member 2 is 15 em thick, which is thinner than any exposures except trench 112 in Blackwater Draw. Member 3 most closely resembles the Lubbock Lake Site type section and in trench 111 is a gray-brown (10 YR 5/2 m) clayey silt.

Members 4 and 5 sediments were not found, but road-fill operations may have removed them. The fill shown in cross section

(fig. 13) ~ates from the 20th century because bricks and metal were found throughout the thickness of upper sediments.

Alluvial Stratigraphy of Blackwater Draw

Two backhoe trenches were cut in Blackwater Draw. Trench

110 showed a valley profile and stratigraphy unlike any exposure in

Yellowhouse Draw, whereas Trench 112 revealed a stratigraphic section virtually identical to the Lubbock Lake Site section.

Trench 110. Trench 110 is 24 km upstream from the confluence of Blackwater and Yellowhouse Draws and where the valley is 900 m wide. The trench, in the eastern 600 m of the valley, exposed

70-cm-thick fine carbonate pebble basal conglomerates and overlying 61 pink (7.5 YR 8/2 m) fine to medium quartz sands sedimentologically identical to Lubbock Lake Site member 1 deposits. The gravels and sands were traced for 550 m east-west, but only the eastern contact with bedrock was established. Overlying member 1 were 45 cm of con formable, light-brownish-green (2.5 Y 6/2 m) sandy clays. The sandy clays were largely eroded in the center of exposure and were strati

graphically uncorrelative with any known unit in Yellowhouse Draw.

Filling most of Trench 110 were 260 cm of dark-yellow-brown (10 YR

4/4 m) silty very fine to fine sand above which were 30 cm of sandy clay. No 14C dates exist for this member. The sediments are very friable and otherwise lithologically similar to late member 5 alluvium at the southern extreme of Yellowhouse Draw.

Trench 112. Trench 112 near Amherst is 109 km upstream from

Trench 11~ and 133 km upstream from the confluence of Blackwater and

Yellowhouse Draws (fig. 2). The trench's stratigraphy is similar to the

Lubbock Lake Site sequence and very unlike that of Trench 110. All five members of the Yellowhouse Draw formation are present. Basal sediments resemble those of member 1 (Appendix) and are 65-cm-thick

1-2-cm carbonate pebble conglomerates overlain by 130 cm of clean well-sorted fine to medium quartz sand. This member 1 thickness is equivalent to that at Trench III and 114 exposures. Capping the clean sands is 10 cm of gray-brown (2.5 Y 5/2 m) very fine sandy clay

(stratum IC) conformably overlain by 15 to 50 cm of black silt containing a I-cm-thick bed of clean diatomite. The black silts, assigned to member 2, grade upward into 50-65 cm of gray-brown (2.5 62

Y 5/2 m) silty clays (stratum 3A), which, in turn, grade vertically to

40 to 65 em of marl and calcareous clay (stratum 3B).

A valley-wide disconformity separates member 3 from overlying brown to dark-brown fine to medium sands denoted member 4. The sands contain a massive I55-cm-thick cumulic soil A horizon, which was eroded along with an unknown thickness of member 3. Very dark gray clays and silty clays of stratum 5C were deposited in the resulting channel. CHAPTER 3

INTERPRETATION OF STRATIGRAPHY

Regional Stratigraphic Trends

One hundred thirty-three kilometers of Blackwater Draw from its confluence with the Yellowhouse Draw to Trench 112 were tested as were 64 km of Yellowhouse Draw from Trench 111 to Dam 6 at the southern edge of Lubbock, Texas. Regional stratigraphic trends were recognized throughout these distances by observing bedrock geometry, sediment volume and composition, unconformities, formation continuity, and pedologic horizonation.

In the two valleys, widths of cross sections changed from 78 m at Trench ,109 to 900 m at Trench 110 and depth of alluvial fill ranged from 4 m at Trench 110 to 8.5 m at Trench 109. Although thickness, color, and sand content of individual units differed between localities, the Yellowhouse Draw formation remained mappable; individual members were not always persistent and only the lower three subdivisions were regionally mappable. Members 4 and 5 were frequently inseparable up stream from the Lubbock Lake Site, whereas both members were map pable downstream from the Lubbock Lake Site.

Yellowhouse Draw Formation

Member 1 has a uniform thickness of 180 em in both valleys, and the member includes a basal conglomerate and overlying sands with contemporaneous gravel facies. Member 1 is frequently capped with

63 64

30-50 cm of clay that grades upward into overlying member 2 diatomites and black clays.

Member 2 varies greatly in thickness, is lithologically uniform, and thickens downstream. Member 2 is 15 cm thick in Trench 111 and

50 cm thick in 112. The thickest member 2 sediments were encountered at Dam 6 where 280+ cm are exposed. Basal diatomites and higher diatomaceous muds are predominant lithologies, whereas in central sections of both draws, member 2 clays gradually grade vertically into member 3.

Member 3 is uniformly thick in both valleys and is lithologically homogeneous; the basal half is a grayish-brown (2.5 Y 5/2) silty clay to clay that grades upward into calcareous clays and marls.

Disconformably overlying member 3 is member 4, a thick series of sands gleyed at their base from the Lubbock Lake Site south and oxidized in all other exposures. Internal stratification becomes complex downstream because sedimentary structures are not masked by eolian sand in these localities.

Some unconformities extend the lengths of a valley, whereas others are localized. The dis conformity separating members 3 and 4 is the single most important post-13,OOO yr B.P. erosional event. The disconformity was identifiable in all trenches in Yellowhouse and

Blackwater Draws. The next most extensive erosion occurred ca. 2500 yr B.P., when several meters of sediment were removed along the cen tral portion of both valleys. A third and local disconformity, dating ca. 8000 yr B. P., was caused by erosion that removed highest elevation member 2 pond sediments. 65

Unlike the lithostratigraphic units, no pedologic horizons are regionally nappable, and none is correlative with any certainty between 7 valleys. The oldest soil at the Lubbock Lake Site is S upper member

2 muds of the Yellowhouse Draw formation. It is a very localized paleosol; pedogenetically altered member 2 sediments are found only at the Lubbock Lake Site and for a few kilometers farther downstream.

Soil S6 in upper stratum 3C is equally limited in extent and is absent north of the Lubbock Lake Site. The remaining soils, S5 in member 4 and Sl-4 in member 5 are traceable within the Lubbock Lake Site; in upstream exposures un differentiable cumulic horizons reach 2.4 m in thickness and condense at least 4000 yr of pedogenesis into a single soil profile.

Interpretation of Sedimentary Environments and Regional Stratigraphic Trends

In' the study area, no unconsolidated valley alluvium older than

Rancholabrean was observed. The oldest alluvium, stratum 03 included moderately indurated fluvial gravels and sands remaining as buried terraces preserved along the valley bedrock walls (figs. 6 and 7). The maximum exposed thickness of stratum 03 was 3.8 m. This thickness represents the minimum amount of alluvium in Yellowhouse Draw before stratum 03 was incised and the Yellowhouse Draw formation deposited.

The post-stratum 03 erosion removed an additional 1 to 2.5 m of bedrock to form the present valley bedrock configuration. The suggested valley development is: (1) during the early to middle Pleis- tocene the Yellowhouse Draw formed by erosion into the Blanco For- mation bedrock and a valley floor was formed 2.5 m above its present 66 elevation, (2) the valley aggraded and filled with a minimum of 6.3 m of fluvial gravel and sand between ca. 100,000 and ca. 13,000 yr B.P.

(Rancholabrean time), and (3) the Yellowhouse Draw was scoured of all but small volumes of stratum 03 sediments sometime before ca. 13,000 yr

B. P. The age for stratum 03 deposition is unknown. Also unknown is how much time elapsed between the erosion of stratum 03 and the deposition of Yellowhouse Draw formation basal gravels.

Due to erosion into and below the base of stratum 03, the next younger sediments (member 1 of the Yellowhouse Draw formation) occupy the lowest elevations in both draws. It is valid to divide member 1 into a basal conglomerate (stratum lA) and overlying sands

(stratum IB); however, only long transverse trenches will expose the stratigraphic relationship between the true basal conglomerate and coarse gravels that are facies of the fine to medium stratum IB sands.

Me'mber 1 was deposited by a moderately sinuous permanent stream. Large volumes of clay and silt overlying coarse clastics are absent, as are vertically alternating gravels, sands, and intraforma tional conglomerates. Multiple cut-and-fill episodes are evident in the upper sands, but the lenses are invariably sands or clayey sands with minor gravel lenses. Where gravel does constitute a large volume of upper member 1, it is massive and resembles bar gravels interfingering laterally with fine to medium sands. Therefore, no evidence exists for overbank-dominated meandering stream deposition or flashy, ephemeral, to braided stream flow. Pierce's (1975) gastropod data suggest a depositional hiatus between strata lA and IB. The lowest stratigraphic date for stratum IB is on Pisidium and Sphaerium shells from gravels 67

(fig. 13). The age was 12,650 ± 250 yr B.P. (1-246). Stratum lA may have been deposited several thousand years prior to 13,000 yr B. P.

Fluvial deposition and reworking of sands persisted until the final sand deposition that dates between 12,400 and 11,000 yr B.P.

The transition from fluvial member 1 sedimentation to lacustrine member 2 sedimentation is represented by stratum Ie olive low-organic clays and sandy· clays found underlying member 2. The bedding in stratum Ie is parallel to overlying lacustrine units. The Ie clays represent a brief oligotrophic episode between the transition from fluvial to eutrophic lacustrine deposition. The stratum Ie sediments represent initial ponding before major aquatic vegetation became established.

The excellent stratigraphic control at the Lubbock Lake Site enables testing of the contemporaneity of the Rancholabrean faunal extinction and depositional environment changes around 11,000 yr B. P.

The extinCtion is not temporally coincident with a fluvial to lacustrine shift; instead, species extinctions occurred du:dng evolution of the pond system. Disparate timing of extinction and depositional environ ment change does not preclude a cause-and-effect relationship. Drop ping of the water-table may not have stressed animal popUlations until many years latter. The only definite conclusion is that there is no obvious coincidence between megafaunal losses and stratigraphic changes.

The cause of the late Pleistocene megafaunal extinction is subordinate to its potential as a regional time-stratigraphic event. In all exposures examined in both the Yellowhouse and Blackwater Draws, no extinct vertebrate fauna were found above member 1 of the 68

Yellowhouse Draw formation. This abrupt, regionally identifiable event portends equally large-scale use for the extinction. I have used the late Pleistocene megafauna extinction in the Llano Estacado to define the

Pleistocene-Holocene boundary in that region. Furthermore, the post-ll,OOO yr B.P. interval may warrant its own North American Land

Mammal Age; however, formally naliling this age is withheld pending more regional investigation, but the use of a nonclimatic continentwide

Pleistocene-Holocene boundary seems justified.

The lacustrine origin of member 2 is unquestionable; only the detailed chemical and biological evolution of the fossil ponds needs further elucidation. Starting at 11,000 yr B.P., oligotrophic ponds at least 240 cm deep fostered oligophilic, epiphytic diatom populations that formed the finely stratified basal diatomites. Detailed cross sections

(figs. 5, 7, 8, 12, 13) reveal a progressive shallowing of the lake basin with time, . an increase with time in organic detritus, and finally shallow marshes at 8000 yr B. P. that were associated with drought-resistant gastropods and eutrophylic diatoms. No stratigraphic or sedimentologic evidence was seen to support Wendorf's (1970) conclusion that a sub pluvial, wetter, fresh-water interval existed near the base of member 2.

Geologic data imply a gradual lowering of the water table until 9000-8000 yr. B. P. Geologic evidence does not support the Scharbal:.er Interval

(Wendorf, 1970) during stratum Ie time. Temporary drying is indi cated by peaty muds and the erosion of peripheral member 2 muds.

This implies that by late member 2 time ponds had shrunk well below their older water levels. At 8,000 yr B.P., ponds may have existed only along the center of the valleys. Member 2 was not eroded at the 69 lowest elevation of those muds along the valley axis; however, marginal

erosion and subaerial pedogenesis predominate the highest elevation member 2 sediments.

The stratigraphic change from member 2 to member 3 has less

genetic importance than the lithology changes would indicate. At the type section, the extremely abrupt contact between members 2 and 3 is

due to differences in sediment structure and weathering characteristics.

Carbonate content is the principal cause for the abrupt contact.

Regionally, the two members grade vertically into each other over several tens of centimeters. The type section's abrupt boundary be tween members 2 and 3 is therefore not significant. The significance of member 3 is that it is the uppermost unit of a conformable, genetically coherent sedimentological sequence. The calcareous clays and marls that terminate member 3 are the final phase of a lacustrine cycle that began around 11,000 to 11,500 yr B.P. Member 3 is interpreted as alkalitrophic lake deposits similar to those from calcareous terranes described by Wetzel (l975). A marl-depositing lake origin is favored by upward decreases in organic content, diatoms, and invertebrates and the upward increase in calcium carbonate content. Eolian and colluvial sand became important components of shore facies, beginning around

9500 yr B.P. and this sedimentation continued past 4,900 yr B.P. The increase in upland-derived clastics would have added carbonates, which would have absorbed nutrients and decreased biologic activity in the ponds. This self-enhancing cycle would ev~ntually CUlminate in ponds where chemically induced oligotrophy would favor low organic matter content and marl deposition. Increased eolian deposition during member 70

3 time undoubtedly enhanced alkalitrophic conditions; however, alkali trophic ponds might have developed irregardless of upland erosion.

The evolution of lakes in carbonate terranes often follows an ontogeny from oligotrophic to eutrophic to alkalitrophic (Wetzel, 1975). Climatic changes between 11,000 and 5000 yr B.P. could have accelerated this evolution, but development of carbonate-depositing lakes may have been inevitable once ponds began forming around 11,000 to 11,500 yr B. P.

The dis conformity between members 3 and 4 is recognizable in both valleys. The disconformity is a regional stra.tigraphic marker, and it demarcates two ground-water regimes. The earlier is the late

Pleistocene through early Holocene emergent water table and the later is the later Holocene influent conditions during which ephemeral streams and eolian and colluvial deposition predominated. The erosion around

4900 yr B" P. was the consequence of the drying during the late Pleis tocene that caused valley water tables to drop to a level where erosion could begin. The 4900-yr-B.P. erosion was not coincident with an iso lated dramatic climatic change. The erosion resulted from processes begun 7000 years earlier. The degradation was due to the intersection of a constant stream energy with decreasing alluvial resistance to erosion. The post-member 3 erosion was brief; its duration is esti mated by a date of 4960 ± 80 yr B.P (SMU-492) above the disconformity and dates of 4910 ± 50 yr B.P. (SMU-331. 333) and 4900 ± 60 yr B.P.

(SMU-531) below the disconformity. During that time a thick bed of fluvial sand was deposited, after which eolian and colluvial deposition added to the thickness of member 4. Duration of the 4900-yr-B • P • erosion and deposition is considered instantaneous: a threshold was 71

crossed, erosion followed, and the valley rapidly reeq uilibrated by the

deposition of the fluvial sands.

Although the 4900-yr-B.P. geomorphic perturbation and ensuing

eolian deposition might imply extreme aridity, regional stratigraphy does

not support this interpretation. The valley water table dropped, but

only enough to initiate erosion around 4900 yr B. P. Member 4 dates

between 2600 and 4900 yr B. P. At least 2,000 years was needed to

develop the thick Calciustoll within member 4. The soil's presence and

its development imply considerable landscape stability within the valley.

The existence of a high water table during early member 4 time is indi

cated by gleyed basinal sands. Similar postdepositional reduction of

surface alluvium is occurring in Yellowhouse Draw today. The high

water table could not have persisted more than a few hundred years

because the developing B horizon of soil S5 reoxidized the upper parts

of the gleyed member 4 sediments. A high water level would have pre

vented S5 calcic horizon formation.

The strong development of 55 at the Lubbock Lake Site could

indicate a regional landscape stability far greater than that seen from

stratigraphic evidence. In the region between the Lubbock Lake Site

and Dam 6, S5 is distinct and readily separable from overlying soils;

however, in the remainder of Yellowhouse and Blackwater Draws, only a

single massive cumulic buried soil is seen. I conclude that some or all 5 of the post-S Boils were incorporated into the cumulic profiles seen

upstream from the Lubbock Lake Site. The cumulic development pre vents formal soil nomenclature being used on a regional scale. If the

Sl through S5 chronology were valid, repetitive erosional-pedogenic 72 episodes would be datable; however, if the Lubbock Lake Site paleosol

series is only a local phenomenon, regional climatic data would be un- available. A third explanation of the different soil stratigraphies could be that the Llano Estacado has large geomorphic subsections, each of which has a varying susceptibility to eolian deflation. As aridity in creases northwestward across the Llano Estacado, a threshold may exist where wetter and drier regions are separated. I conclude that formal nomenclature for alluvial soils is inapplicable for the Llano Estacada.

Although repetitive paleosols do record small climatic fluctuations, the soils can be used only for climatic changes in the vicinity of the soil profile.

Member S cienega clays in Yellowhouse Draw interfinger with eolian and colluvial sands. The coincidence of arid-initiated deflation and high water table pond development seems contradictory. It is not because different ground-water regimes control the upland and valley geology. The uplands rely on regional ground wat'i!r and rainfall to establish soils and vegetation communities. The water tables in the valleys are perched and are frequently asynchronous with the local climate. An example of the latter is that modern cienegas exist in

Yellowhouse Draw, which is within a semiarid climate.

If strata SA and SB deposits are related to upland processes, do their four soils, 51_54 have paleoclimatic significance? Repetitive eolian deposition followed by pedogenesis implies rapid climatic shifts.

An alternative explanation is that the cycles of deflation and pedo genesis are geologic processes active near a threshold. The annual rainfall and temperatures of the Lubbock area coincide with the 73

maximum sediment yield in Schumm's (1973) plot of sediment yield vs.

climate. If the rainfall of the Lubbock region were to vary even 5

percent above or below today's 18 inches per year, either stable vege

tated uplands or erosion would ensue. Very small, not large changes

in yearly rainfall could easily shift geomorphic stability from pedo

genesis to deflation in the uplands. The Lubbock area may today

straddle such a threshold. Historic climatic events indicate that the

region is "stable" compared to the instability experienced during the

Great Dust Bowl era. A similar threshold relationship is hypothesized

for at least the last 900 years and possibly the last 2,000 years.

Explaining complex soil stratigraphies by using thresholds eliminates the need for major climatic changes to cause the complex strata 5A and 5B

stratigraphy.

The geologic history of the Llano Estacado is summarized in

10ca1- and' regional-scale time-geologic process diagrams (figs. 14 and

15). On each diagram erosion, deposition, and stability were plotted against geologic time. Erosion or degradation is the net loss of sed iment (mass) from the valleys, whereas deposition (aggradation) is the reverse. Stability is either the absence of erosion and deposition or an equilibrium where both processes are active but equal in magnitude. A fourth division is the pedogenetic window, an energy regime wherein soils can form and be preserved. Soils such as Calciustolls form at equilibrium, or zero, where the landscape surface is neither losing or gaininK mass. As deposition increases, pedogenesis continues but with

-the formation of cumulic soil profiles. Above a certain deposition rate, pedogenesis ceases and soil formation will not begin until deposition I04 104 T l--STRATUM 03-t I-+STRATUM 2+STRATUM 3-1 en en 3 ~STRATUM en 0::: 10 : .., ...... ! .. w U u Z STRATUM 4 o 2 0::: Z 10 n. o U 1 r-- -1H (5 .... 10 en rSTRATUM 5- 9 o o 100 w fh (!) oP~ ,~, ,~O~ PM1N W f.en 100 SOIL s5 0::: solLs4 @ U SOIL S'3 ~~~ Z ~ 101 . «(!) z SOlLS2 ~ o 102 W en SOIL SI/ o I > 3 ~ 10 ~ ...J ffi W 0::: 4 ~104 ••.? ... ! ..... ""10

, , 1- , - , I I I I I I I L I I I -. E-M -'r 17 16 15 14 13 12 II 10 9 8 7 6 5 4 3 2 0 PLEISTOCENE 14C YEARS B.P.(x 103)

Figure 14. Time-process diagram showing geologic processes of the Lubbock Lake Site during the Pleistocene Epoch

-.J ~ 4 4 ,10 10 en en rSTRATUM 03-1l-STRATUM I-f-STRATUM 2+STRATUM 3-1 en .0= 103 : ..? ...... ,... 1 w o u Z 2 ~ 10 0.. Z r---STRATUM 4/5 o U (5 !::: 101 en o 9 0 @ f:j 10 (!) J.SOIL BURIED OR tL. PEDOGENIC - (GUfJ(JLlC'£PIP'EDONS 0 WINDOW lCALCIUSTOLLS 0 ] [ERODED EPIPEDONS "1S0ILERODED -- ~ t~0 => en 10 t- 0::: Z (J Z 101 ~ z CUMULIC W 2 Qen 10 PEDOGENESIS > o ~ 3 ..J ffi 10 W 0::: ~104 ...! .. , ... 104

• -- I --.- ---. - I - I - I I I I I I I I I I I E-M ..Jor 17 16 15 14 13 12 II 10 9 8 7 6 5 4 3 2 0 PLEISTOCENE 14C YEARS B.P.(x 103 )

Figure 15. Regional time process diagram showing how geologic processes acted on a regional scale on the Llano Estacado

-J U1 76 abates. Conversely, a finite amount of erosion may coexist with pedogenesis, but the results are thin A horizons or soils stripped to their lower B or C horizons.

The Y axis of figures 13 and 14 has qualitatively assigned, non dimensional logarithmic v21ues reflecting sediment consistency, grain size, volume of sediment and its thickness, horizonation, and type of soil formed. The Y-axis values reflect the energy stored in the landscape or the work needed to initiate an erosional, depositional, or pedogenetic event.

The earliest event plotted on figures 14 and 15 is the erosion of the Llano Estacado valleys. Dotted lines imply uncertainty of the duration and timing of that erosional event. Stratum 03 is also dotted because its age is undetermined. Following deposition of stratum 03, incision occurred and the basal member 1 gravels were deposited.

Aggradation dominated until 4900 yr B. P. when erosion and subsequent 5 deposition of member .; occurred. Extended S pedogenesis was terminated ca. 2600 yr B.P. when erosion cut a channel along the valley axis. Geomorphic events during the last 1,000 to 2,000 years were a series of increasing1y more frequent deposition and weaker soil formation.

Figures 14 and 15 compare the kinetic energy of eolian, fluvial, gravitational, and colluvial processes to landscape resistance. Erosion, deposition, or pedogenesis results when specific intrinsic thresholds are crossed by external energy changes (geologic processes) or internal changes such as the landscape's susceptibility to erosion. Landscape form is controlled by its activation energy, and this energy must be 77 exceeded before erosion or deposition will begin. The activation energy is a summation of resistance (lithology, induration, and vegetative

cover) and stored gravitation energy (elevation, sediment volume, and slope of the alluvium). Activation energy is increased by lithification or plant growth and decreased by vegetation loss and a lowered water table. As a landscape's activation energy decreases. less energy is needed to initiate erosion and deposition and the probability of geologic change increases. Conversely, if activation energy remains constant, erosion can still occur by changes in external factars such as amount and timing of rainfall.

Superimposed upon the time-process concept is the relationship between the magnitude of a geologic event (erosion) and the following response (deposition). The different magnitudes of the event- response in figures 14 and 15 suggest that several activation energies (thres holds) may have existed for the Llano Estacado and, depending upon which level is crossed, the scale of the geologic event will similarly vary.

The series of soils and colluvium dating from 1000 yr B.P. to the present may be associated with a lower order threshold than the threshold for the 4900-yr-B.P. erosion. An even -greater threshold

(activation energy level) had to be crossed for post-stratum 03 erosion to take place. CHAPTER 4

ORGANIC CHEMISTRY OF FOSSIL BONES

Inorganic and Organic Constituents of Bone

The alluvial geology is so well known from the Llano Estacado that the geology should be used to test hypothetical paleoenvironmental techniques. A concept that can be tested is the determination of late

Pleistocene-Holocene vegetational changes by using stable carbon isotopes. The abundance of bison remains at the Lubbock Lake Site presented an ideal opportunity to evaluate stable carbon isotopes against a known geologic history.

Fossil bone collagen isotopy has been widely used to unravel human agt1cultural innovations (van der Merwe, 1982); however, these techniques determined only the presence or absence of certain foods and in popUlations temporally separated by hundreds to thousands of years. The availability of a continuous isotopic record for a grazing animal was an ideal opportunity to test whether stable-carbon isotopes in collagen were a valid paleoenvironmental tool.

The purpose of the isotopic work was to determine if the 613C of bison collagen changed over the last 12,000 to 13,000 years. The fossils were extremely contaminated and considerable chemical pretreat ment was necessary to obtain the pure organic constituents. Once separated, these organic residues were usable not only for stable carbon isotopy but also for radiocarbon dating. The search for a 78 79

collagen-specific bone compound culminated in the isolation of the amino

acid hydroxyproline. The purification and chromatographic isolation of bone amino acids is introduced at the beginning of this chapter. The

013C study is presented next, and the chapter is concluded by a report on radiocarbon dating of several inorganic and organic fractions from a mammoth bone of known age. 14 The contamination of fossil hones has hampered C workers for years. Questions about bone 14C accuracy have been raised by

Berger, Horney, and Libby (1964), Tamers and Pearson (1965); Sells tedt, Engstrand, and Gejvall (1966), and Berglund, Hakansson, and

Largerlund (1976). The use of bones for o13C studies necessitates that carbon-bearing phases be even more highly purified.

Bone Contaminants

Bone has an intrinsic propensity for its own contamination, both physically by particulate debris and chemically through adsorption or exchange with carbon-bearing inorganic and organic molecules. Partic- ulate contaminants include root hairs, partially decayed plant fragments, and carbonaceous muds lodged in cancellous tissue and Haversian canals. Calcium carbonate contamination may be macroscopic, e.g., coatings of soil or ground-water caliche. Less obvious is molecular- -2 - scale contamination whereby exogenous C0 is exchanged for OH or 3 Co;2 in the hydroxyapatite crystal lattice. Organic contamination is predominantly by tannic, humic, and fulvic acids; free amino acids, noncollagen pep tides , and more rarely petroleum residues may also be 80 adsorbed by hydroxyapatite, a compound with highly adsorptive proper ties (Brooks, 1981).

Previous Research

Historically, pretreatment of bone for 14C dating has used either the insoluble collagen fraction remaining after HCl or EDTA decalcification (Berger, Horney, and Libby, 1964; Tamers and Pearson,

1965; Sellstedt and others, 1966; Haynes, 1967; Olsson and others,

1974) or gelatinization of the acid-insoluble collagen (Longin, 1971).

The method most commonly used now is HCl decalcification followed by

0.5 M NaOH leaching of the residue (Haynes, 1967). The insoluble collagen is finally dissolved by using pH 3 water at 90 0 C for 24 hr

(Weber and others, 1981; Berglund and others, 1976). The purity of each residue has been assessed by micro-Kjeldahl analyses (Sellstedt and others, 1966) and quantitative amino acid and C/H/N determinations

(Hassan and Hare, 1978; Dennisen, 1980). Dating petroleum-impreg nated bones requires acid hydrolysis of the collagen and purification of the hydrolysate by binding the amino acids to cation-exchange resins and washing the chromatographic column with organic solvents (Ho,

Marcus, and Berger, 1969). Despite the specificity of certain reagents for collagen dissolution 14C researchers have not used hot trichloracetic acid, sodium dodecylsul£ate, and the enzyme collagenase because these organic reagents could contaminate the sample.

Imino Acid Biochemistry

A reliable 14C date from bone protein demands isolation of one or more collagen-specific constituents. One such molecule is 81 hydroxyproline, a secondary amino acid or imino acid that is rare in nature except in collagen where it makes up 10-14 mole percent of the protein's 19 amino acids. Ho anJ others (1969) suggested radiocarbon 14 dating hydroxyproline to enhance the accuracy of a collagen e date.

Hydroxyproline has value for radiocarbon dating work because its biochemical synthesis is a test of its purity. Hydroxyproline is formed in vivo by hydroxylation of proline (Stetten and Shoenheimer, 1944;

Udenfriend, 1966; Adams and Frank, 1980); because only OH is added to proline, hydroxyproline and proline will have identical 12e, l3e , and

14e compositions. Discordant stable or radiocarbon values for proline and hydroxyproline would indicate contamination with exogenous proline or hydroxyproline.

The carbon composition of both proline and hydroxyproline must be compared because proline is common through nature and hydroxy proline occurs not only in vertebrate collagen but in nonskeletal vertebrate and invertebrate collagens and non collagenous proteins. The most common occurrences of hydroxyproline are in collagen and collagenlike proteins such as elastin and extensin and in bacterial and ver" tebrate metabolic products.

Hydroxyproline includes two position isomers, 3- and 4-hydroxyproline, together having six diastereomers: D- and L-3-hydroxyproline and the cis- and trans-isomers of D- and L-4-hydroxyproline. The

3-hydroxy-D- and trans-4-hydroxy-D-proline isomers are unreported in living systems (Kuttan and Radhakrishnan, 1973), whereas the others exist bound or as the free imino acid. 3-hydroxy-L-proline constitutes approximately 1 residue per 1000 in vertebrate collagen and occurs in 82 spongin A and B, Mediterranean sponges, the legume Delonix regia, and the antibiotic bacterial peptide telomycin (Kuttan and Radha krishnan, 1973). Cis-4-hydroxy-D-proline is bound in the antibiotics etamycin and actinomycin Xor3 (Kuttan and Radhakrishnan, 1973). Cis

(allo)-4-hydroxy-L-proline is bound in the toxic Amanita mushroom pep tide phalloidin but is most abundant in a soluble protein in sandalwood

(Santalum album L. (Radhakrishnan and Giri, 1954), other species of

Santalum, and Osyrus arborea in the Santalaceae family.

The principal, but not sole repository of 4-hydroxyproline is collagen, a connective tissue found in all metazoan life forms; that is, all higher invertebrates and vertebrates (Adams, 1978; Adams and

Frank, 1980). Collagen has a characteristic wide-angle X-ray diffrac tion pattern, approximately one-third glycine residues and a triple helix structure. Usually, but not always, collagen contains a high percent age of pyrl'olidine amino acids and has repeating units of gly-X-Y

(Adams, 1978). Only silk of the sawfly (Nematis ribesii has been re ported 4-hydroxyproline free (Adams, 1978).

Vertebrate collagen constitutes approximately 20 percent of an animal's weight and is the single largest source for 4-hydroxyproline.

Proline and hydroxyproline each average approximately 10 mole percent of a collagen's residues. A third of the collagen's 19 amino acids are glycine and another third a.re glutamic acid plus imino acids; tryptophan, cystein, and cysteic acid are usually absent (Kuttan and

Radhakrishnan, 1973).

Invertebrate collagen, which is found in every phylun from

Porifera through Hemichordata, differs from vertebrate collagen by 83 having a more variable amino acid spectrum, sometimes cystein (Hunt,

1970), different ratios of proline:hydroxyproline and lysine:hydroxy lysine and the occasional presence of hydroxyproline=x in (gly-X-Y)n instead of hydl'oxyproline=y in vertebrate collagen (Adams, 1978).

Noncollagenous vertebrate protein contains 4-hydroxyproline in elastin, a connective-tissue protein that contains 0.8-3 percent hydroxyproline

(Kuttan and Radhakrishnan, 1973). Other vertebrate proteins con taining 4-hydroxyproline are reticulins, collastromin, elastroidin, dentinal proteins (Kuttan and Radhakrishnan, 1973), the C1q component of complement, a small protein in bone (acetyl cholinesterase), lung lavage glycoproteins, a collagenlike protein absorbed by human blood cells, and a serum component not identical with the Clq protein (Adams and Frank, 1980).

Nonccllagenous invertebrate protein yields 4-hydroxyproline predominantly in polysaccaride-protein complexes in at least eight invertebrate phyla from Porifera upward (Hunt, 1970). In insects, the protein sclerotin contains 4-hydroxyproline, whereas the sc1eroproteins and resilin, an analog of elastin, do not contain hydroxyproline.

In plants, 3- and 4-hydroxyproline is bound within cell walls

(Lamport and Northcote, 1960; Lee, Bada, and Peterson, 1976), which are constructed of cellulose microfibrils embedded in a noncellulose matrix containing high amounts of extensin, a 4-hydroxyproline-rich glycoprotein containing up to 33 percent hydroxyproline residues

(Kuttan and Radhakrishnan, 1973; Sadava, Walker, and Chrispeels,

1973; Rosenthal, 1982). In marine diatoms, both 3- and 4-hydroxypro line are present in a protein associated with siliceous skeletal formation 84

(White and others, 1978). All plant groups from algae through angio

sperms have genera with significant amounts of hydroxyproline-bearing

arabinisides in which hydroxyproline is glycosodically linked to OH to

form insoluble arabinose oligosaccharides (Adams and Frank, 1980).

Extensin is very insoluble and yields hydroxyproline only after harsh

hydrolysis (Kuttan and Radhakrishnan, 1973; Lee and others, 1976).

Unbound or free trans-4-hydroxyproline is limited to urine

(Klein, Faulkner and de la Pena, 1970), some fruits (prunes and young

apples), pollen, and other exotic biological occurrences (Kuttan and

Radhakrishnan, 1973). Both free and peptide-bound hydroxyproline

occurs in urine; human urine contains 11-55 mg hydroxyproline in a

24-h collection period (Klein and others, 1970).

The 4-hydroxyproline in sediments is ultimately derived from

animal or plant tissues, and contamination of bone would first occur in

this matrix. Soils contain hydroxyproline (Bremner, 1950); however,

free hydroxyproline in soils appears very unstable and is rapidly de

aminated and oxidized. Bound hydroxyproline is more stable and can

be liberated by using hot 6 N Hel hydrolysis of the soil to yield up to

4 mg hydroxyproline/100 g soil.

Hydroxyproline was once considered specific to collagen, but it

is known to be widespread. The abundance of nonvertebrate collagen hydroxyproline does not vitiate 'Using hydroxyproline for isotopic

studies because many of its isomers are those other than trans-4- hydroxy-L-proline. Most of the nonskeletal 4-hydroxyproline sources are from rare exotic biological circumstances not encountered in a

geological context. The most probable source of 4-hydroxyproline 8,5

contamination would be extensin in plant cell walls and free or small

peptide-bound hydroxyproline formed during diagenesis of buried animal

tissues. Extensin is insoluble in hot acidic water, which dissolves

collagen, whereas free hydroxyproline is dissolved and removed during

nCI decalcification of bones. Small peptic1e-bound hydroxyproline re

mains a possible contaminant, and its mobility and preservation in

sediments is unknown.

Purification of proline and hydroxyproline for carbon isotopy imposes demanding strictures: The imino acids must be totally free of

foreign carbon thus precipitation of imino acids as picrate, reinekate,

copper, and rhodanilate salts (Greenstein and Winitz, 1961) is pro hibited. Semi-preparative ion exchange chromatography isolates amino

acids on a milligram scale, but the eluting buffers contain' foreign carbon from sodium citrate, detergents, and preservatives (Moore and

Stein, 1951). Volatile ammonium acetate and ammonium formate buffers have been used to separate 50-300-mg amounts of amino acids (Hirs,

Moore, and Stein, 1952), and they considered the removal of these buffers.

A volatile inorganic elutant such as HCl would be ideal for amino acid chromatography. An HCI system was tested originally by

Stein and Moore (1950) and Hirs, Moore, and Stein (l954) on microgram quantities of amino acids and both step and gradient HCI solutions were used by Lynch, Hughes, and Rhodes (1959). Up to a few hundred milligrams of imino acids are separable on short HCI-eluted DOWEX-50 columns (Brownlee and Spiro, 1979); however, all primary amino acids 86 must first be removed, otherwise the imino acid peaks coelute with aspartic acid, alanine, and glycine.

Proline and hydroxyproline are separable by nitrosylating collagen hydrolysates and removing the primary amine degradation products with ethyl ether to leave water-soluble imino acids. Hiller and

Van Slyke (I919) found that only primary (I0) amines react with HN0 2 to yield hydroxy acids (ROHGOOH). Hamilton and Ortiz (1950a, 1950b) formed hydroxy acids from 1° amino acids and generated N-nitroso derivates from secondary (2°) amines using NaNO with either acetic Z acid or 6 N HGI 'acid to liberated nitrous acid. Volatile N-nitroso 2° derivates formed at pH Z.5, whereas none formed in 6 N HGI. The change from weak to strong HGI solutions eliminated hydrolyzing the nitrosylate to convert N-nitroso amines to the original imino acid.

Nitrosylation yields were 96-98 percent for the imino acids. The 1° hydroxy-acid derivatives were removed with ethyl ether (Hamilton and

Ortiz, 1950b; Dziewiatkowski, Hascall, and Riolo, 1972). In my pro cedure (Stafford and others, 1982) aqua regia generates the nitrosylating species NOGI. The method eliminates formation of NaGI, which would otherwise interfere with the chromatographic separation of the amino acids.

Isolation of Hydroxyproline

Bone is mechanically cleaned by scraping and sonicating, then finely powdered. The powder is sized, and the fraction smaller than

250 lJm is demineralized for 5 hr in cold (6°G) 0.6 N HGI. Approxi mately 50 ml of acid is required for each gram (fig. 16). Micro 87

BONE POWDER ID.5 N HCt, 4'C, % h~A I CelltJt.i61lg~1 CRUDE ORGArHC PELLET I IH 20, 90'C, 24 h~A Cellt.\.i~Ilge I CRUDE GELATIN EXTRACT

IL!loph.ii..(.:~ 6 N He! i J1 5' C! 2" Itu I CRUDE HYDROLYSATE

IKAP-% Co tillll II I I PURIFIED HYDROLYSATE

IEuapoJta~~, AQila Reg.i4 EuapoJt4.t~1 I NiTROSYLATED HYDROLYSATE I Etlte.\ ElttJt4c:.t I 'S&p4Jtat~ In4otllb!~1 EXTRACTED HYDROLYSATE

• PRO + HYPRO

lAG SDW-X', IN Hctl EU4pc~ate PROLINE HYDROXYPROLINE

Figure 16. Flow diagram showing steps used to isolate hydroxy proline from a fossil bone. -- From Stafford and others (1982). 88

Kjeldahl or a fluorescamine assay (Bohlen and others, 1973) helped to estimate the total amount of bone powder needed to isolate a given amount of hydroxyproline. After demineralization, the insoluble col lagen is isolated by centrifugation. The collagen pellet is extracted with water at 105°C for 24 hr under nitrogen, and the soluble gelatin separated from insolubles. The yields of gelatin in this step depend on the degree of collagen cross-linking, which may be a function of age and diagenesis. The slightly acidic gelatin solution often contains sig nificant contamination by fulvic acid, which is removed by passing the solution through a 0.5 x 10-cm column of 20-25 mesh XAD-2 resin (Junk and others, 1974; Aiken and others, 1979; Thurman and Malcolm, 1979;

Curtis and others, 1981). This resin not only completely absorbs the fu1vic acid but also removes from solution some of the gelatin. Collagen absorption is overcome by hydrolyzing the gelatin extract for 24 hr at

1050 C in 6' N HCl under N2 gas. Fulvic acid is removed from the amino acids by passing the hydrolysate over the XAD-2 column. Negligible amounts of amino acids are held by the column, but all of the fulvic acids are absorbed. Yellow coloration of the hydrolysate is due to Fe+3 ions. The purified hydrolysate is rotary evaporated to a syrup, then redissolved in 2 ml concentrated HCl per each 100 mg of amino acids.

Five hundred microliters of concentrated nitric acid are added to yield a solution of nitrosylhydrochloric acid (aqua regia). The solution is heated 5 min in a boiling water bath and then immediately rotary evap orated. The reactions occurring in this step are similar to those oc curring in the presence of nitrous acid and are based on the evolution of nitrosyl chloride (NOCl) in aqua regia. Figure 17 illustrates the 8-9

R, /\ HC - NH2 NC - NH I I COOH AQUA REGIA COOH +NOCL

ret l-HCL R R I H . /\ HC - N - N • a , He - N - N• a COOH I COOH

R I r + He - N • N OH I eOOH

R ~f"' I \ ·He - OH He - rlH t I COOH COOH

Q-hljdlLcx!J 41!.i.d ~tI!C"d4ILd 4.ino 41!id Uh!! ':. ~oLubl.~ I!thl!·'I, in4oLubl.t

Figure 17. Diagram showing the reaction of nitrous acid with primary and secondary amines. -- From Stafford and others (1982). 90 difference in reaction of primary amino acids and secondary imino acids with nitrosyl chloride. The residue from the nitrosylation reaction contains unreacted imino acids and a mixture of hydroxy acids and their oxidation products, which elute before hydroxyproline. This mixture is separable in one of two ways. The residue is either re-evaporated after addition of 10 ml of water or dissolved in 1 ml 1 N HCl and directly applied to the column described below. The re-evaporated residue is dissolved in 2 ml acetone at 50 o-60oC, then chilled to OOC and 100 ml diethyl ether added to this solution. The resultant milky suspension settles as an oily residue, and the organic layer is de- canted. A second extraction with 50 ml ether removes remaining traces of the hydroxy acids. The oily residue, when analyzed, consists of a mixture of proline and hydroxyproline (table 1, fig. 18) and can be separated into individual proline and hydroxyproline fractions on an

DOWEX 50W-X8 column eluted with 1 N HCL (fig. 19). Combined fractions containing hydroxyproline or prQline are rotary evaporated, and the residue dissolved in water and transferred to storage vials and lyophilized. Overall yields are between 80 and 90 percent, and the purity of hydroxyproline and proline determined by amino acid analysis is given in table 1.

Fossil Collagen Isotopes

The stable carbon isotope composition of an animal is repre- . 12 13 13 sented as the rabo of C to C and is expressed as 15 C, a value proportional to the 12e-13C ratio in the food or CO metabolized by the 2 animal or plant. Laboratory and field experiments show a biochemically Table 1. Amino acid compositions of gelatin and fossil bone collagen and derived fractions

Residues per 1000

a b Gelatin Total Bone Collagen

Hydro- Nitro- Dowex Dowex Hydro- Nitros- Dowex Dowex Amino Acid lyzed sylated 1 2 lyzed XAD sylated 1 2

4-hydroxyproline 86 470 1000 104 109 426 1000 Aspartic acid 43 44 28 Threonine 16 15 18 Serine 33 25 22 Glutamic acid 76 64 76 Proline 130 530 1000 112 124 574 1000 Glycine 356 346 366 Alanine 103 125 111 ! Cystine Valine 21 25 23 Methionine 5 2 Isoleucine 9 12 10 Leucine 22 25 23 Tyrosine 4 1 Phenylalanine 10 12 5 Hydroxylysine 6 3 3 Histidine 6 6 5 Lysine 27 29 26 Arginine 46 52 48

a. Commercial gelatin (Knox Gelatin, Inc.). b. Demineralized fossil bison bone dated 4910 ± 50 yr B.P. (SMU-332,333). (12) were carried through the isolation procedure of figure 16. '"...... 92

gly

glu pro G.'atin Hydrolyzat. ; 1:5000 Dilution

U Nltroaylated Hydrolyzat. C-1005 i 1:5000 Dilution ...:c c c:J= Water Solubl. Ph •• e 0 .! ':5000 Dilution ~6 ,.J =w u z c CD a:: Eth.r Pha •• 0 ., 1:20 dilution C-1007 II) CD :\ c

a.a.lln.: 0.1N HCI".) ~a

--~i~--~i----~i~--~i-----ri----~i-----~i----~i~---Ti-----~--~,----~,-- 20 40 10 10 100 120 ' .. 0 '80 '10 200 220 240

ELUTION TIME (MINUTES)

absorption at 440 nm characteristic for imino acids

absorption at 570 nm characteristic for amino acids and ninhydrin reactive products

Figure 18, Chromatograms of amino acids in gelatin, nitrosyl ated gelatin hydrolysate, and imino acid phases during isolation of hydroxyproline. -- From Stafford and others (1982). 93

1"C PRO I' I \ 600 I \ 3000 I \ I \ :I: I \ :I: 0. "00 2000 C I \ ~ :c J \ (J to) , \ ~ J \ 200 I \ 1000 I " I " o -_./ ------o 250 300 400 500 600 700

VOLUME (ml)

Figure 19. Chromatograms showing elution positions of hydroxyproline and proline during preparative isolation with 1 N Hel buffer. -- From Stafford and others (1982). 94 predetermined isotopic shift (fractionation) from source to product carbon. Diet is most commonly inferred by measuring living or fossil animal tissues, a technique that has successfully established the agri cultural introduction of maize (Zea) or shifts from marine to terrestrial human food subsistences (van der Merwe, 1982).

An unstudied paleoenvironmental aspect is how accurate and precise are millennia-long isotopic records. Fluctuations between \513 C extremes are easily recognized in fossil data, but are climatic shifts equally discernible and are they valid when compared to more conven tional paleoecological techniques? To evaluate carbon isotopic histories, 13 a 13,OOO-year-Iong fossil bison collagen \5 C record was established.

The isotopically inferred climatic changes were compared to paleoen vironmental conclusions made geologically from the fossil-bearing sedi ments. The test is strengthened by the availability of historic climatic records an'd subrecent bison, both of which establish a climate-isotope relationship that can be extrapolated back to the late Pleistocene.

Location

The vertebrates analyzed were fossil bison collected at the

Lubbock Lake Site whose geology has been discussed in the previous chapters. The alluvial site contains unconsolidated sediments dating from modern to ca. 13,000 yr B. P., and the locality is one of only a few valley exposures that reveals the fluvial, lacustrine, and eolian sediments found within other relict Pleistocene drainages crossing the

Texas southern High Plains (Stafford, 1981). 95

Several methods were used to limit laboratory variation and

enable others to evaluate the conclusions. A single mammalian genus,

Bison, was used to eliminate differences between ecologically dissimilar

genera. Both the modern bison, Bison bison bison, and its predeces-

Bor, Bison antiquus, were considered to have had the same grazing

habits and ecological niches, and both were assumed not to have been

browsers. Modern grasslands of the Texas Panhandle are predominantly

C grass communities with a high l3C content. Sedimentologic and 4 paleontological data indicate that C grasslands might have been the 3 dominant grass type 11,000-12,000 yr ago, and if so, an isotopic tran-

sition would exist between modern and late Pleistocene time. Finally,

collagen in large (>100 kg) mammals has such a long turnover time (>10 yr) that a bison's diet would have been averaged over its lifetime, and l3 the bone collagen 0 C could therefore be taken as an average of the

grassland composition.

Stable Carbon Isotopy

In the four decades following the discovery of differing 12C and

l3C abundances in nature by Nier and Gulbransen (1939), researchers have tabulated 12C-l3C variations throughout the geosphere and bio- sphere (Craig, 1953), discovered photosynthetic pathways for CO iso 2 topic fractionation and inferred human-environment interactions from fossil remains. Reviews of this literature by van der Merwe (1982),

Lerman and Troughton (1975), and Smith (1972) stressed the importance of carbon fractionation by plants: the differential utilization of 12C02 96 13 and CO2 that yields compounds whose isotopic signatures are carried throughout the food chain.

Depending on their photosynthetic pathway, plants can be divided into three groups, C , C , and Crassulacean acid metabolism 3 4 . 12 13 (CAM), and each has a correspondmg C- C composition depending how CO is metabolized (O'Leary, 1981). The C plants utilize the 2 3 Calvin-Benson photosynthetic pathway (Calvin and Benson, 1948) in which ribulose 1, 5-diphosphate is carboxylated to two molecules of the three carbon product, phosphoglyceric acid. As currently known, all trees and most shrubs are C plants and many other families are also 3 C ; one exception is the Gramineae, which are mixed C and C • Suc 3 3 4 culents are unknown as C and are either C 4 or CAM. As a group, C 3 3 plants have an average 013 C value of -26.5 per mil and a variation from

-20 to -35 per mil (Vogel, Fuls, and Ellis, 1978). In contrast, C 4 plants discriminate an average of 1. 4 percent less against 13C02 than do C taxa, thus C plants range in Ol3C value from -9 to -15 per mil 3 4 and average -12.5 per mil.

The C 4 plants fix CO via the Hatch-Slack photosynthetic path 2 way (Kortschak, Hartt, and Burr, 1965; Hatch and Slack, 1966, 1970) whereby phosphoenolpyruvate (PEP) is carboxylated to form the four- carbon dicarboxylic acid oxaloacetate, which is hydrolyzed to malate or aminated to aspartate, depending on which C subgroup the plant be 4 longs. The malic or aspartic acid. synthesized in the mesophyll is sub sequently transported inward to the bundle sheath cells where decarboxylation occurs by one of three different enzymes. If malic acid is formed, the NADP-malic enzyme would decarboxylate malate, whereas 97

in C plants synthesizing aspartic acid, either PEP carboxykinase or 4 NAP-malic enzyme is used (Ellis, Vogel, and Fuls, 1980). After

decarboxylation of malate-aspartate, the CO is converted to 3-phos 2 phoglyceric acid via the normal Calvin-Benson pathway. The two-step

fixation of CO in C 4 plants demands spatial isolation of the bio 2 chemistry in different cell types, a biochemical-anatomical linking

accomplished by the Kranz anatomy, also termed the "Kranz syndrome"

(Laetsch, 1968). Independent of a distinct o13C signature, the unusual

Kranz anatomy morphologically characterizes those belonging to the C 4

group. The Hatch-Slack, or dicarboxylic acid, pathway is especially

abundant in the Gramineae (grasses) but is also found in the Acan- thaceae, Aizoaceae, Amaranthaceae, Asclepiadaceae, Boraginaceae,

Capparidaceae, Carylophyllaceae, Chenopodiaceae, Compositae, Cyper- aceae, Euphorbiaceae, Liliaceae, Nyctaginaceae, Polygalaceae, Portu- lacaceae, Scrophulariaceae, and Zygophyllaceae (Laetsch, 1968; Brown,

1977; Stowe and Teeri, 1978; Teeri, 1979).

Crassulacean acid metabolism (CAM) plants are isotopically intermediate between C and C taxa and range from -37 to -9 per mil 3 4 013 C (Lerman and Troughton, 1975). The group is named for the metabolism of the plant acid malate, a pathway common in the Crassulaceae family in which CO is first fixed as malic acid during daylight. De 2 carboxylation and CO incorporation into hexoses also occur during 2 daylight periods (Beevers, Stitter, and Butt, 1966). The C plants 4 physically separate the two photosynthetic CO fixation steps, whereas 2 CAM plants temporally alternate the synthesis of the same compounds. 98

Plant Ecology

The ecologic position of C and C 4 plants was initially equated 3 with cool-moist and hot-dry environments, respectively (Downtown,

1971), but this generalization has been disproved by regional vegeta tional analyses along climatic gradients (Tieszen, Senyimba, and others,

1979; Ellis and others, 1980; Livingstone and Clayton, 1980; Hattersley,

1983). Temperature is believed to control C abundance; more 4 examples are daily minimum July (summer) temperature (Teeri and

Stowe, 1976); temperature of the warmest growing month (Vogel and others, 1978; Ehleringer, 1978), mean annual temperature (Boutton,

Harrison, and Smith, 1980; Tieszen, Hein, and others, 1979), and local temperature as controlled by elevation (Livingstone and Clayton, 1980).

Temperature is not the only control for C 4 distribution because absolute rainfall does modify C -C abundance in the optimal thermal 3 4 growing regime of each group. The C plants can dominate where 3 precipitation extremes are from 100 to over 1,250 mm per year and equally high C 4 percentages are found where the rainfall is 100 mm to

1,000 mm rainfall per year (Vogel and others, 1978). In Australia, C 4 composition depends positively on the average minimum January

(Australian summer) temperature, whereas species numbers followed different criteria (Hattersley, 1983). The C species numbers increased 3 with decreasing average January (summer) temperature and also increased as spring rainfall increased. The C species increased both 4 as October (spring) average minimum temperature and average February precipitation increased. 99

The success of C and C plants appears primarily temperature 3 4 dependent, although less exclusively than Livingstone and Clayton

(1980) posited. Relative C to C abundance is modified by amount of 3 4 rainfall and coincidence of precipitation with the optimal growing season. The C plants would be most abundant where summers are 4 both hot and wet, whereas C taxa would flourish wherever the spring 3 is cool and wet (Hattersley, 1983).

Mammalian Isotope Compositions

The C and C 4 plants are valid indicators of paleoenvironments 3 because modern plant communities change in percentage of C to C 4 3 along climatic gradients, and the taxonomic gradient is duplicated by

Ol3C values for bulk vegetation along these transects. The isotopic composition of an animal's tissues parallels the isotopic composition of its food source; therefore animal remains will reflect the diet and climate during the animal's lifetime.

Controlled laboratory feeding experiments (DeNiro and Epstein,

1978b; Tieszen, Hein, and others, 1983) were used to demonstrate the parallelism between diet and tissue ol3e content. Field experiments have shown the direct relationship between in an animal's habitat e3-c4 and ol3e of animal tissues (Minson, Ludlow, and Troughton, 1975; van der Merwe and Vogel, 1978; DeNiro and Epstein, 1978a; Vogel, 1978;

Tieszen, Hein, and others, 1979; Land, Lundelius, and Valastro, 1980;

Sullivan and Krueger, 1981; Chisholm, Nelson, and Schwarcz, 1982).

The total animal is slightly enriched in l3e (approximately 1-2 per mil relative to its diet) (DeNiro and Epstein, 1978b). Individual 100 tissues, organs, chemical fractions, and excreta may range from 13 per mil heavier to 2 per mil lighter than the diet (Minson and others,

1975; DeNiro and Epstein, 1978a, 1978b; Lyon and Baxter, 1978; Vogel,

1978; Jones and others, 1979; Tieszen, Hein, and others, 1979). Bone is usually the only tissue available for paleoecological studies, but no one agrees how much fractionation occurs between diet and bone collagen. Values for bone collagen enrichment over diet include the fol lowing 013 C values: +6.1 per mil (Vogel and Waterbolk, 1967), +5.1 per mil (van der Merwe and Vogel, 1978), +3.7 to +2.8 per mil (DeNiro and Epstein, 1978b). The controlled laboratory experimentation on mice may be the most accurate (DeNiro and Epstein, 1978b).

Carbonate from hydroxycarbonate apatite (bone mineral) is more enriched in 13C than collagen, and values for bone carbonate enrich ment vs. collagen are +6.5 to 7 per mil (DeNiro and Epstein, 1978a),

+10 per mil (DeNiro and Epstein, 1978b), +8.8 to +7.0 per mil (Sullivan and Krueger, 1981), and +7 to +10 per mil (Land and others, 1980).

No mechanism is presently known for collagen to exchange carbon with exogenous carbonaceous molecules. Bone apatite carbonate does ex 2 change and is highly susceptible to exogenous C03 contamination (Hassan, Termine, and Haynes, 1977). Although the apatite carbonate controversy has been rekindled by Sullivan and Krueger (1981, 1983) and Schoeninger and DeNiro (1982, 1983), fossil bone apatite is con sidered unsatisfactory for paleoecological work and I have used only collagen in the Bison research.

If fossil animal popUlations are to be reliable indicators of temporal vegetation changes, the collagen isotopes must reflect not only 101 the isotopic value of the food but also the percentage of C to C 3 4 plants in the herbivore's food. Modern grazing animals reflect the l3C content of their grassland habitats or, if the animal is a mixed grazer- browser. the proportion of woody (C ) to grass (C ) vegetation will be 3 4 indicated. South African hyrax (Procavia johnstoni) grazing on C 4 grasses were isotopically distinguishable from a sympatric species

(Heterohyrax brucei) predominantly. browsing woody (C plants (DeNiro 3 and Epstein, 1978a). The 013C values for hair, milk, and hide of cattle grazing either temperate or tropical grassland paralleled the values measured for respective C and C 4 grasslands (Minson and 3 others, 1975). Eleven species of African browsers, including elephant, 13 giraffe, and kudu, had an average 0 C of -21 per mil, whereas grazers such as zebra and wildebeast from C 4 grasslands had an average ol3C of -9 to -9.3 per mil (Vogel, 1978). Stomach contents of African species parallel grassland composition because Ol3C analyses of rumen contents matched 013 C estimates made from macroscopic identification of rumen contents (Tieszen, Hein, and others, 1979). In North America, bone collagen o13 C on eastern white-tailed deer (Odocoileus virginianus) and western mule deer (0. hemionus) showed the dichotomy between humid C food sources and arid to semiarid C foods (Land and others, 3 4 1980).

The proportion of C to C 4 vt:getatiotl is climatically controlled 3 13 and 0 C values change as the C -C ratios change. Animals feeding 3 4 in a specific ecosystem unquestionably attain isotopic compositions reflecting their diet. A final question is whether Bison spp. precisely 102 reflect the grasslands they graze or can grazing behavior vary enough to alter the isotopic results.

Bison Ecology

The living North American Great Plains bison is a grazer with no major preference for anyone or group of grasses {Peden and others, 1974}. Other bison species or populations are not as re stricted. The European wisent {Bison bonasus} is a marked browser;

Athabascan and Utah populations of Bison bison eat a high percentage of sedges, whereas Arizona herds favor saltbush {Atriplex} {Peden and others, 1974}. Plains buffalo will eat shrubs on heavily grazed pas tures but not on lightly grazed prairies where shrubs are <4 percent of the total vegetation {Peden and others, 1974}. Compared to other grazers like antelope, cattle and sheep, bison are the least selective

{Schwartz ,and Ellis, 1981}. Bison in northeastern Colorado eat grasses in nearly the abundance found in the prairies, and only during May will ingested plants differ from the habitat's grass composition. Peden

{1976} determined from esophogeally fistulated bison that their diets were 76 to 90 percent C grass species and 0-13 percent C g~~ass 4 3 species during the year except for May. In May, C 4 grass intake was

46 percent, whereas the grasslands contained 76 percent C 4 grass species.

Until extensive grassland-rumen-collagen isotope analyses are available for bison on high and low C 4 percent grasslands, the possi bility exists that a bison's eating habits could make the animal's isotopic values appear a few per mil lighter than that of the prairie biomass. 103

Experimental Results

The fossil bison bones were extensively humate contaminated, and extensive purification was necessary before the 13C values were reliable. The effectiveness of the pretreatment was tested by isolating hydroxyproline and comparing its t513C value with those for other purified fractions.

A complete Bison bison left tibia dating 4910 ± 90 yr B. P.

(SMU-332, 333) was used to evaluate how t513C changes as purification of collagen proceeds. The specimen was collected from stratum 3C alkalitrophic lacustrine sediments at the Lubbock Lake Site. Epiphyses and cancellous tissue were removed, and the diaphysis was broken into l-cm fragments. After 2i hr· ultrasonification and drying at 50°C, the bone was ground to less than 0.25 mm. The l1-mm-thick cortical bone had its outer 3 mm and inner 2 mm discolored ·a light yellowish brown

(10 YR 6/4) by adsorbed humic and fulvic acids. Micro-Kjeldahl analysis of the bone gave 3329 ppm nitrogen (approximately 2 percent protein). The powder was divided into two fractions; fraction A was used for direct hydrolysis and fraction B for gelatinization. Aliquots 13 13 from each step were removed for t5 C. To evaluate the t5 C of inorganic carbonate, 5.9 g of powdered bone were stirred 14 hr at 21°C under line vacuum in 500 ml 1 M acetic acid to remove secondary carbonate. The insoluble residue was washed three times with H 0 and 2 lyophylized 12 hr.

The bone powders for hydrolysis and gelatin pathways were decalcified with 3,900 ml 0.6 N HCl at 7°C for 22 hr, the last 14 hr under line vacuum. The insoluble residue was concentrated by 104 centrifugation, and the clear pale-yellow supernatant saved. The very dark brown pellet was washed three times with 150 ml 0.6 N HCL at

7°C. The residue for fraction A was hydrolyzed 24 hr at 105°C with 6

N HCl bubbled with N2 gas for 3 min. Fraction B was heated 24 hr at

90°C in pH 3 water to gelatinbe collagen. Gelatinization yielded a slightly viscous solution, which was centrifuged and the supernatant rotary evaporated. The resulting syrup was dissolved in 20 ml 6 N HCl for hydrolysis. Both fractions followed the final purification methods.

Fulvic acids were removed from the 6 N HCl hydrolysate by passing the solution through a 2. 4-cm by 20-cm column of 20-50 mesh

XAD-2 equilibrated with 20 ml 6 N HCl (Stafford and others, 1982).

The resin was washed with pH 1 water to remove excess HCl before the fluvic acids were eluted with 1 M NH 0H. 4 Primary and secondary amino acids were separated by nitrosy lating the 'hydrolysate with 4 ml: 1 ml 12 M HCl: 18 M HN0 per 200 mg 3 total amino acids. The syrup was first dissolved in 12 N HCl, and the

18 N HN0 added afterward. The solution was heated in a boiling 3 water bath for 5 min after efferescencr..: began. The solution was rotary evaporated, water was added, and the solution was re-evaporated to a yellow syrup. The syrup was dissolved in 1.5 ml acetone, and the flask was cooled in an ice bath before 50 ml ethyl ether was added. A clear colorless supernatant (the hydroxy acid fraction) and an immis cible oil (imino acid fraction) resulted. After decanting the acetone ether and re-extracting the oil with ether, the nonpolar phase was evaporated and the oil dissolved in 1 ml water for chromatography. 105

The imino acids were separated by injecting the aqueous imino acid mixture onto a 77 x 1.3-cm, 200-400 mesh DOWEX SOW X-S column equilibrated with 1 N Hel. The column was eluted with 1 N HCl over

30 hr at a flow rate of 19 mllhr. Hydroxyproline was collected in the

150- to 360-ml fractions and proline in the 425- to sOO-ml fractions.

Each imino-bearing solution was rotary evaporated and lyophylized.

Sequential purification of the fossil bison bone is shown in figure 20. Two different pathways, gelatinization and direct hydrolysis, are used to remove humic and fulvic acids. In both schemes, the o13C values of humates were 3-4 per mil lighter than those of the col lag en , but collagen became progressively heavier following each purifi cation. In the direct hydrolysis method, collagen hydrolysate 013 C values prior to XAD were -11. 9 per mil, but -11.1 per mil following

XAD; the 013C value for combined imino acids was -10.5 per mil, where as values for hydroxyproline and proline were -10.2 and -9.S per mil, respectively. Comparative o13 C values after gelatinization were per sistently heavier than the respective hydrolysis step and were -10.6 per mil for "gelatin" before XAD and -9.9 per mil afterwards; value for combined imino acids was -9.4 per mil. In addition to humic and fulvic acids, values for the hydroxy acids from the nitrosylation reaction were substantially lighter than those of the collagen; values of hydroxy acids were 4.9 per mil lighter by direct hydrolysis and 2 per mil lighter by gelatinization.

The significance of the results shown in figure 20 are: (1)

XAD-2 effectively removes humate contamination, (2) o13 C values for the unseparated imino acids are close to those in individual pro and 106

SUPERNATANT INSOLUBLE RESIDUE XAD-2 RESIN

PASSED THROUGH AMINO ACIDS -9.9%0 N ITROSYLATION -10.5%0 4:1 12m HCI: IBm HN02 ACETONE/ETHER CHROMATO GRAPHIC EXTRACTION SEPERATION

INSOLUBLE HXDROXYPROLI NE PHASE (IMINO ACIDS) -10.2%0 -9.8%0 -9.4%0

Figure 20. Flow diagram showing 613 C composition of various fractions of a fossil bone 107 hyp, (3) collagen values after XAD purification may be as much as 1 per mil lighter than either imino acid, and (4) hydroxy acid fractions are contaminated by 12C from organic solvents and are not useable for stable carbon isotopy.

The 013 C vs. time curve for fossil collagen (fig. 21) was established by collecting in situ fossil bison remains from modern to ca.

13, OOO-yr-B • P. alluvium and chronologically positioning the specimens by associated 14C dates or stratigraphic correlation. Bison metapodials were analyzed whenever possible, but the rarity of vertebrate remains in some units (stratum lA, lower member 4) necessitated using other skeletal elements.

Laboratory purification of the severely humate-contaminated bones followed the preceding method, except that individual amino acids were not isolated. After all cancellous tissue was removed and the diaphyseal' cortical bone was powered, 50 to 150 g of bone were dis solved in 0.6 N HCI at SOC to obtain an insoluble collagenous residue.

The acid-insoluble protein was hydrolyzed with 6 N HCI under N2 gas at 105°C for 24 hr and the filtered solution then passed through a 1 x

20-cm column of 20-50 mesh Amberlite XAD-2 resin to remove fulvic acids. The purified hydrolysate was rotary evaporated and finally dried at 50°C in a vacuum drying oven. The collagen hydrolysate was burned to CO for mass spectrometric analysis by heating 10-20 mg of 2 sample, 1 g CuO wire, and 5-10 mg Ag foil to 800°C for 3 hr in evac- uated, sealed Vycor tubes (Buchanan and Corcoran, 1959; DesMarais and Hayes, 1976). 108

The results from the fossil collagen stratigraphic study are tabulated in table 2 and figure 21. The most important findings are:

1. There is minimal change in the 0 13C value of CO from acetic 2 acid-treated bone apatite over 13,000 years. The average value of -1. 7

± 2.5 per mil is markedly different from the collagen values over ,time.

2. The o13C value of collagen becomes lighter with increasing geologic age of the fossil specimen. The average o13C values changed from -8 per mil at 200-700 yr B.P. to -11.5 per mil at 5000 yr B.P. and to -16 to -17 per mil at 8000 to 12,500 yr B.P.

3. Populations of samples from the same member varied 2-3 per mil, whereas reproducibility from the same bone was approximately ±0.5 per mil.

Discussion

The isotopic studies are valid only if several qualifications are true. Failure of anyone aspect is sufficient to jeopardize the tech nique's . usefulness. It must be unequivocally established that:

1. Diagenesis of collagen during burial does not alter the original protein 13C value.

2. Laboratory methods are capable of purifying highly contaminated bone.

3. Collagen has a sufficiently long in vivo half-life that seasonal dietary changes do not alter lifelong 13C averages.

4. Feeding behavior of fossil bison populations did not vary from that of modern populations. Table 2. aBC isotopic data for Lubbock Lake Site, Texas Values for Nos. 12 throu

13 - 15 C (per Collagen pDB Lab/Field Date Skeletal Isolation No. Number Stratum Matrix (yr B.P.) Element Method Organics Ap

a 1 77LLP-122a 5C clay ca.500 metatarsal Gelatin - 9.12 a 2 77LLP-122b 5C clay ca.500 metatarsal Gelatin - 8.61 a 3 77LLP-5-1a 3C clay 4910 tibia Gelatin -11. 00 a 4 77LLP-5-1b 3C clay 4910 tibia Gelatin - 9.98 a 5 77LLP-5-1c 3C clay 4910 tibia Gelatin -11.33 a 6 77LLP-5-1d 3C clay 4910 tibia Gelatin -12.24 a 7 77LLP-5-2a 3C clay 4910 metatarsal Gelatin -12.82 a 8 77LLP-5-2b 3C clay 4910 metatarsal Gelatin -10.71 a 9 77LLP-17 1A gravel ca.12,000 metacarpal Gelatin -17.95 a 10 77LLP-17' 1A. gravel ca.12,000 metatarsal Gelatin -13.09

11 Bos bos A.D.1982 femur 6 N HCL/XA~b -12.68 12 78LLP-6 5C clay ca. 200 metacarpal 6 N HCl/XAD -6.09; -7.1 b 13 78LLP-6' 5C clay ca. 200 metatarsal 6 N HClIXAD -7.19; -7.1 14 78LLP-2 5C clay ca. 400 metatarsal 6 N HClIXAD~ - I 15 78LLP-32 5C ca. 700 tibia 6 N HCl/XAD - 1 b 16 78LLP-22 4A 4400 ramus 6 N HClIXAD -1: b 17 78LLP-14a femur 6 N HClIXAD 18 78LLP-14b femur hydrol~zed -14.47 gelatin b 19 78LLP-27 4A 4900 humerus 6 N HCL/XAD-2 -9.52 20 78LLP-17 4A b clayey sand 4900 medial 6 N HClIXAD-2 -1- phalanx 21 78LLP-28 3C clay 5050 humerus 22 78LLP-8 3C clay 5050 tibia 6 N HCl/XAD-2~ °-1 1 23 77LLP-5 3C clay 5050 tibia 6 N HClIXAD-2 -1 b 24 78LLP-23 2C clay 8000 metacarpal 6 N HCl/XAD-2 -1 b 25 78LLP-23 2C 8100 6 N HCl/XAD-2 -1 clay metatarsal b 26 78LLP-29 mid 2 clay 9700 metacarpal 6 N HCl/XAD-2 -1 b 27 78LLP-4 2A diatomite 10,700 metatarsal 6 N HCl/XAD-2 -1 28 78LLP-10 1C sandy clay 11,100 metacarpal 6 N HCl/XAD-2~ -1 29 77LLP-17 1B sandy clay 12,300 metacarpal 6 N HCl/XAD-2 -1 b 30 77LLP-14 1A gravel ca.12,500 thoracic 6 N HCl/XAD-2 -1 a. Gelatin isolation procedure of Longin (1971). b. 0.6 N HCL decalcificati c. Gelatin extraction, 6 N HCL hydrolysis, XAD-2 purification.

109

Values for Nos. 12 through 30 are plotted on figure 21.

13 - IS C (per mil) Hagen pDB )lation thod Organics Apatite Comments

a latin - 9.12 Distal end of 77LLP-122 a latin - 8.61 Proximal end of 77LLP-122 a latin -11.00 Distal end of 77LLP-5-1 a latin - 9.98 Distal mids~ction of 77LLP-5-1 a latin -11.33 Proximal midsection of 77LLP-5-1 a latin -12.24 Proximal end 77LLP-5-1 a latin -12.82 Proximal end 77LLP-5-2 a latin -10.71 Distal end 77LLP-S-2 a latin -17.95 Distal end of 77LLP-17 a latin -13.09 Proximal end of 77LLP-17

~ HCL/XA~b -12.68 -1.24 Univ. Ariz. Meat Sci. Lab. ~ HClIXAD -6.09; -7.85 b ~ HClIXAD -7.19; -7.08 - Same bone powder extracted two different times b - 9.71 -0.83 ~ HCl/XADb ~ HClIXAD - 8.12 -0.19 b -11.40 -2.16 ~ HClIXADb ~ HClIXAD 6 N HCL and gelatin extraction from same bone powder drol~zed -14.47 Latin b ~ HCL/XAD-2 -9.52 -2.96 b ~ HCl/XAD-2 -14.81 -4.00 ~ HClIXAD-2~ "-10.19 -2.59 ~ HCl/XAD-2 -11. 90 - Compare with HYP /PRO fractions below b ~ HClIXAD-2 -15.77 b ~ HCl/XAD-2 -17.19 -3.09 b ~ HClIXAD-2 -14.25 -4.29 b ~ HCl/XAD-2 -13.63 -3.51 ~ HCl/XAD-2~ -16.2 -1.99 ~ HClIXAD-2 -17.04 -2.47 b ~ HCl/XAD-2 -17.15 -5.40

0.6 N HCL decalcification, 6 N HCl hydrolysis, XAD-2 purification.

~n.

110

I I

I I I I I .1 APATITE 13 I 8 C I

-6

II) 0 a.. -8 0 !, 0~ u ~ co- -10 I Z IJ.J ,II (!) c::[ -12 -l -l 0 t) -14 I !I I -16 I I COLLAGEN I II 8 13C -18

-20

~----~-----~------~----~------I~----~I-----~I~------o 2 4 6 8 10 12 14 xl03 yrs8.P.

Figure 21. Plot of collagen 15 13 C vs. geologic age for fossil Bison spp. bone from the Lubbock Lake Site, Lubbock Lake, Texas 111

Diagenesis of bone protein during burial could alter the aBe

value of the collagen hy changing the amino acid by selective degrada-

tion. Amino acids in modern collagen should have different isotopic

compositions because they have different dietary origins. The ol3e of

inta.ct total collage~ replicates its dietary sources, but if the molecule is

differentially degraded, amino acid ratios respective to one another will

change and the resulting ol3e value will vary from that originally in

bone collagen.

Diagenesis was monitored by using quantitative amino acid

analyses of bone protein and by the aBe value of specific amino acids

isolated from purified hydrolysates. Table 3 tabulates the amino anal

yses of nine bones used for Q13 e analysis; the six youngest specimens

have collagenlike residues per 1000, that is, a value close to glycine

(300/1000), proline (110/1000), hydroxyproline (90/1000), aspartic acid

(50/1000), . alanine (130/1000), and glutamic acid (70/1000). The oldest

three had noncollagen compositions. Noncollagenous compositions would

indicate either diagenetic loss of some collagen amino acids or con-

tamination by bacteria or other exogenous amino acids (Armstrong and

. others, 1983). 13 The XAD-2-purified collagen hydrolysate was used for the 0 e

study. Figure 20 summarizes the purification of a 4900-yr-old bone. 13 The XAD-purified collagen hydrolysate has a 15 e value of -11.1 per

mil, whereas the hydroxyproline and proline values are -10.2 and -9.8

per mil, respectively. The absence of significant fractionation between

hydroxyproline and the XAD-2-purified hydrolysate implies that these hydrolysates are suitable for Q13e work. Table 3. Quantitative amino acid analyses of selected fossil bison bones dating ca. 200 to ca. 12,500 yr B.P., Lubbock Lake Site, Texas. -- Values expressed as residues per thousand.

1978- 1978- 1977- 1978- 1978- 1978- 1978- 1977- 1977- LLP-6 LLP-2 LLP-32 LLP-27 LLP-28 LLP-8 LLP-4 LLP-17 LLP-14

Geologic age, yr B.P. 200 400 700 4900 5050 5050 10,700 12,300 12,500 Geologic stratum 5C 5C 5 4A 3C 3C 2A 1B 1A Nitrogen, % 2.32 0.83 3.00 0.35 0.95 0.023 0.09 0.07 0.018 Amino acid Cysteic acid 0 0 0 0 0 0 0 0 0 3-hydroxyproline 1 1 1 0 0 0 0 0 0 4-hydroxyproline 94 101 97 66 95 87 30 52 43 Aspartic acid 50 57 53 79 55 60 152 104 112 Threonine 16 14 16 17 16 13 11 16 19 Serine 31 29 32 28 28 28 18 23 39 Glutamic acid 80 73 77 94 82 77 172 133 133 Proline 132 143 133 113 134 113 86 78 110 Glycine 296 269 293 287 291 312 267 320 277 Alanine 130 136 124 159 132 162 114 138 11~ i cystine 0 0 0 0 0 0 0 0 0 Valine 22 26 25 29 26 21 48 38 65 Methionine 6 5 6 3 5 6 10 6 4 Isoleucine 12 13 12 13 12 9 8 12 11 Leucine 28 29 28 28 27 23 27 33 28 Tyrosine 4 2 3 1 1 2 8 0 0 Phenylalanine 13 14 15 13 13 13 16 16 11 Hydroxylysine 6 6 5 3 7 9 11 4 0 Lysine 27 25 26 26 26 21 23 22 24 Histidine 6 3 4 2 8 2 0 5 0 Arginine 46 52 49 40 43 43 0 0 13

1000 998 1001 1001 1001 1001 1000 1001 I-' Total (R/lOOO) 999 I-' N 113

Figure 20 also demonstrates contaminant removal and to what

degree bones must be purified to achieve reliable 013 C values. The major contaminants in the bones are fulvic and humic acids absorbed

from enclosing sediments. The collagen hydrolysate must be passed through the XAD-2 resin to remove fulvic acid. Humates are only partially removed by Longin's (1971) gelatinization method, and inclusion of humates in the hot water-soluble phase may explain why 013C values obtained in earlier analyses (table 2) were not reproducible.

The l3C content of an animal's diet could be substantially altered by seasonal vegetative changes or migration of bison to different

food sources. Unless the carbon in the animal's tissues is metabolically inert, seasonal dietary variations could dramatically affect the isotopic composition of the collagen.

Collagen is an extremely inert mammalian constituent, and its metabolic half-life is so long that seasonal and even decadal variations

In. carb· on lnta k e h ave neg1· 19l ·bl e e ff ect on I·t s 13C composl·t· Ion. The metabolic half-life of rat collagen has been estimated by Thompson and

Ballou (1956) to be 500 to 1000 days. Libby and others (1964) used

14C f rom atomIC . b om b testIng . an d f oun d no 14C In· corporat·· Ion In t 0 human cartilage collagen during a 10-year period. Stenhouse and

Baxter (1979) calculated a human bone collagen mean residency time of more than 30 yr by using detailed bomb 14C analyses. The slow metabolic turnover of bone collagen implies that the 013C values will reflect an average of the carbon ingested during at least the last 5-10 years of the animal's life. Seasonal fluctuations in grassland 114 composition or an animal's migration into different grasslands should not alter the long-term carbon composition in bone collagen.

The Longin (1971) procedure was used for samples 77-5-1,

77-5-2, 77-122, and 77-17 (table 2) from which gelatin was extracted from proximal and distal halves of the same specimen. The halves from bone 77-122, the youngest and least humate contaminated fossil, had the closest 1513 C agreement (-9.12 and -8.61 per mil), whereas the greatest 13 discordancy was in old and humate-impregnated specimens where c5 C ranged from -9.98 to -12.24 on the same bone. Fulvic acid is considered to be the primary contaminant. Hydrolyzing the gelatin in 6 N

HCl and passing the solution through XAD-2 resin removes fulvic acid with insignificant loss or fractionation of collagenous amino acids. 13 Apatite c5 C data were not used because Hassan (976), Hassan and others (1977), Schoeninger and DeNiro (1982, 1983), and Land and others (1980), found postmortem carbonate exchange in bone hydroxy apatite. The apatite c513C was determined to test the premise of Sull- ivan and Krueger (1981) that bone leached with 1 M acetic acid to 13 remove secondary inorganic carbonate would give c5 C date parallel to 13 15 C data on collagen.

The clustering of apatite c513C values at -2.7 per mil is interpreted as postmortem exchange of hydroxyapatite carbonate with exogenous carbonate from bedrock and alluvium. Carbonates are present in the field area as at least two ages of caliche: one is early Pleistocene and the other is Holocene. Sedimentary carbonates reservoirs include lacustrine dolomite and limestone in the Blanco Formation and early to middle Holocene lacustrine marls in member 3 of the Yellowhouse Draw 115

formation. Exchange of carbonate apatite with foreign carbonate was

accelerated by conditions of high ground water that saturated fossils

older than 5000 yr. Sediments younger than 5000 yr are sandy and

facilitated rainwater percolation and caliche formation. Incorporation of -2 C0 into 200-yr-old bones could not be determined because the apa- 3 tite ~13C would fortuitously have been the same as inorganic carbonate.

Values of -1 to 2 per mil for apatite CO would be expected from col 2 lagen ~13C of approximately -10 per mil. If a +8 to +10 per mil value

enrichment relative to collagen is assumed, apatite would have a natural

~13C of 0.0 to -2 per mil. 13 The collagen ~ C data can be used to infer vegetation changes

for the southern High Plains. The isotopic and behavioral fractionation

of carbon during metabolism and ingestion is still imprecisely known,

but uncertainties are lessened if modern quantitative grassland descrip- 13 tions are combined with subfossil bison and their C analyses.

The Pantex site in the northwest Texas Panhandle (Sims,

Singh, and Lauenroth, 1978) was used as a modern example of a

southern High Plains grasslanci. The vegetation at Pantex is a short

grass prairie containing 23 plant species whose biomass is 80 percent

grasses. Ninety-five percent of the grasses are C 4 varieties, of which

Bouteloua gracilis is dominant. The C grasses and C forbs make up 4 4 89 percent of the ungrazed pastures. The most recent fossil bison at 13 the Lubbock Lake Site date ca. 200 yr B. P. and have ~ C values of

-8.1 per mil. These subfossil bison are used to approximate "modern" bison herds on the southern High Plains grazing 90 to 95 percent C 4 grasslands. It' i~ possible to estimate the isotopic fractionation between 116 vegetation and fossil collagen by comparing sub fossil bison collagen 13 . 13 t5 C to the percent C 4 bIomass amd ots t5 C in modern grasslands.

Figure 22 was constructed from modern vegetational and subfos sil t513C data. The composition of older grasslands is extrapolated from 13 the modern data. If the relationship between bone t5 C and grassland percentage C4 is linear, figure 22 results. Inferred grassland changes would be from modern grasslands (95 percent C biomass) changing to 4 85 percent C 4 at 4900 yr B. P., then to 65 percent C 4 at 8000 yr B. P. , and finally to 45 percent C 4 at ca. 12,500 yr B. P. If bison preferen tially select C plants over C 4 plants and if they ingest more C plants 3 3 as these grasses increase in abundance, a nonlinear relationship would result (fig. 23). To evaluate this potential fractionation by behavior, modern bison bone collagen would have to be analyzed from animals living on grasslands of varying C composition. 4 Substantial vegetation changes are hypothesized for the southern High Plains over the last 15,000 years. If the relationship between vegetation and bone collagen is linear (fig. 22), C grasses may have 3 made up 50 to 55 percent of the prairie biomass at 12,000 yr B.P. In historic times, the grasslands contained <5 percent C biomass. Cor 3 relating the fossil grasslands to modern areas of the North America

Great Plains superficially indicates that late Pleistocene climates in the

Texas Panhandle were similar to those in today's northern central Great

Plains province. It is premature to take C percentages from a 5000- 4 or 12, OOO-yr-old isotopic record, assign them to a modern location in the Great Plains, and then use that longitude's temperatures and rainfalls for those of the late Pleistocene to Holocene. Extreme caution % A Grassland 13 = Collagen 13C -12.5 0 100% " C ~ C4 " B Parallel line to A based an Z " modern vegetation record~ and " (J) " (J) (J) "" A a " lL " -19.5°/00 0 ~ 50% " " 0 LLo " " 0 LEGEND " 0- c1..... " " It)U It) .1 200 yr. B.P. " U 2 w 0 4900 yr. B.P. "" C> :.!! " 0 12,500 yr. B.P. " > ~ "

Figure 22. Plot of bone collagen 013 C vs. 013 C composition of grasslands

I--' I--' -J 100% C4

(f) W (f) (f) .r u "'" B ___ _ S" 50% ------0~ W > i= A Linear collagen to diet

100%JL----.-~==:;~~~~~--~~--~~----~~--~~--~,~i2----~~~-C 3 -8 -10 -12 -14 -16 -18 -20 -22 -24%0 BONE COLLAGEN 8 13C PDB

Figure 23. Plot of bone collagen vs. percent C grasses in the grassland ol3c 4

I-' I-' 00 119 is warranted because C 4-C3 compositions for the Great Plains vary state wide (Teeri and Stowe, 1976). Finer resolution would require at least county-scale plant surveys before fossil and modern C -C biomass 3 4 values could be compared. More importantly, the magnitude and periodicity of temperature and rainfall are too interdependent to use C per 4 centages to infer absolute temperature or rainfall. The geologic data indicate higher rainfalls and lower temperatures around 10,000 to 13,000 yr B. P. Although the isotopic data agree with the sedimentology, a shift in precipitation from summer to winter would yield a similar C to 3 C 4 change. The effect of climatic periodicity versus magnitude was illustrated by Hattersley (1983), and his conclusions are apropos to my research. Unless either temperature or rainfall is accurately known, neither can be absolutely inferred from a C percentage. In the Texas 4 example, where paleoclimate is roughly known geologically, the C 4 data help refine the sedimentological climatic curve. Sedimentology may reveal the correct climatic trend, but isotopic data will provide better resolution.

Much remains to be done before fossil collagen isotope techniques become a truly precise paleoclimatic tool. Although broad isotopic changes were discernible in the Texas study, any fine fluctuations were mas k eye d b th 1 t 0 3 per ml'I u.r13 C varIa't' IOn f or b ones f rom the same member. Organic content, degree of collagen and amino acid diagenesis, and inorganic contamination are highly variable and not pres- 13 ently compensated for; even sample pretreatment changes 15 C results.

All factors must be assessed and understood before isotopic data are refined. One major conclusion is that the percentage of C 4 plants in 120 the diet cannot be determined more accurately than ±10 percent for many fossil bones. Even the compilation of published data becomes sus- pect because numerous laboratory methods were used to extract collagen. Only the youngest and least contaminated and diagenetically a Itere d b ones currentl y gIve" h"Ig h precIsIon"" u",13 e vaI ues Irregar" dl ess 0 f pretreatment method.

Precision could be increased by separating one or several specific amino acids; this method would eliminate many diagenetic effects and would standardize future isotopic work. Although hydroxyproline is the most specific constituent of collagen, this imino acid is often decomposed during diagenesis. Other amino acids such as glycine, aspartic or glutamic acid, and alanine might be better substitutes because their molar percentages are often higher in fossil bones. An objection to using noncollagen-specific amino acids is that they may include contamination . from noncollagen sources; only the analysis of numerous fossil bones will establish how easily exogenous amino acids pass through the otherwise rigorous collagen extraction schemes developed in this research'J

Direct 14C Dating of Inorganic and Organic Phases from a Known-age Fossil Bone

Radiocarbon dating of fossil bones has been only partially suc- cessful in providing accurate age determinations, and despite repeated attempts to purify and isolate diagnostic fractions, most bone 14e dates remain questionable. I have attempted to resolve these uncertainties by radiocarbon dating 25 inorganic and organic carbon fractions from bone and evaluating which phases give the most accurate 14e dates for a 121 fossil of known age. I have isolated not only those fractions most commonly cited in the literature but also hydroxyproline, a constituent of bone collagen that should be one of the most specific and reliable fractions for fossil bone dating. The method for isolating hydroxyproline has been published (Stafford and others, 1982) p and a similar technique has been independently developed by Wand (1981).

In the long history of bone radiocarbon dating, researchers have concentrated on the purification of either the collagenous organic fraction or the inorganic phase, carbonate-hydroxyapatite:

(Neuman and Neuman, 1958). Inorganic or apatite bone 14C dating has vacillated between favor and disfavor among geochronologists and has been considered unusable if bone carbonate has exchanged with ground water carbonate (Berger and others, 1964), reliable if first pretreated with acetic acid (Haynes, 1968), and irreversibly contaminated by exog- -2 enous C0 exchange (Hassan, 1976; Hassan and others, 1977). More 3 accurate 14C dates were obtained if CO was derived by sequential 2 heating of the bone (Haas and Banewicz, 1980).

Although several methods have been used to extract organic carbon from bone (Olsson and others, 1974; Taylor, 1982), these techniques adapt solubility properties of modern collagen to fossil bone, a tissue whose collagenous residues often behave differently from modern proteins and yield extracts of varying molecular size and composition. 122

A fossil mammoth (Mammuthus imperator) excavated from the

Domebo 5ite in Caddo County, southwestern Oklahoma, was chosen for

analysis. The bone was selected for dating because the elephantid was

datable archaeologically by its association with Clovis Culture projectile points and radiometrically by wood in the stratum from which the bone was obtained. The bone retained 10 percent of its original organic content, a common value for fossilc that makes the results applicable to a wide range of similarly preserved vertebrates.

The Domebo mammoth site yielded a single immature, possibly female mammoth (Mehl, 1966), whose bones were enclosed by reduced paludal, olive clayey silts, denoted unit 2 of the Domebo Formation's lower member (Albritton, 1966; Leonhardy and Anderson, 1966). The mammoth's geologic age was first established by six 14C determinations, four of which are considered accurate (table 4). One date is 11,045 ±

647 yr B.P. (5M-695) on an in situ tree stump, cf. Ulmus alata (elm) found in the upper part of the clayey silts enclosing the elephant. A second date is 10,123 ± 280 yr B.P. (5M-610) on lignitic wood 2 feet above the bone bed and in clayey silts of the lower Domebo formation.

The third is 11,220 ± 500 yr B.P. (51-172) on mammoth bone organics, and the fourth is 11,200 ± 600 yr B.P. (51-175) on humates from the mammoth bone (Leonhardy and Anderson, 1966). Two additional 14C dates, which are considered spurious (Leonhardy and Anderson, 1966), are 4,952 ± 304 yr B.P. (TBN-311) on untreated mammoth tusk and

9,400 ± 300 yr B.P. (GX-56) on organic earth near the mammoth skull.

Archaeologically, the mammoth dates between 11,000 and 11,500 yr B. P. by association with Clovis Period artifacts (Haynes, 1982). 123

Table 4. Radiocarbon dates from different phases of Domebo mammoth bone compared to dates on associated wood

Material 14C Date

Published wood and soil dates from a Domebo Paleo-Indian Site, Oklahoma

Clovis-level wood from bone stratum (SM-695) 10,145 ± 647 Lignitic wood 2 ft above bone bed (SM-610) 10,123 ± 280 Organic earth near elephant skull (GX-56) 9400 ± 300

Published bone dates from Domebo Paleo-Indian Site, Oklahomaa

Mammoth bone organics (51-172) 11,220 ± 500 Humic acid from mammoth bone (51-175) 11,200 ± 600 Untreated mammoth tusk (TBN-311) 4952 ± 304

University of Arizona wood dates (1983)

Clovis-level wood from bone stratum (C-987) 11,140 ± 280 (redate of M-695) by direct 14C)

C-14 dates on the mammoth bone from Domebo Site, Oklahoma (this dissertation) . 14 ranked according to Cage

Target Carbon Fraction Number

Hydroxy acids-1 from 6N HCl C555 11,840 ± 420 hydrolysis Hydroxy acids-II from gelatin C551 11,630 ± 450 phase XAD-purified gelatin CI038 11,360 ± 310 XAD-purified 6 N Hel C544B 11,270 ± 700 hydrolysate Imino acids from HCL C559 10,850 ± 300 phase 0.6 N HCL acid-insoluble C606 10,760 ± 560 phase Raw bone after acetic C542 10,700 ± 300 acid Acid-insoluble organizes II CI002 10,670 ± 280 124

Table 4. Radiocarbon dates from different phases of Domebo mammoth bone compared to dates on associated wood--Continued

Material 14C Date

C-14 dates on the mammoth bone from Domebo Site, Oklahoma (this dissertation) ranked according to 14C age--Continued

Target Carbon Fraction Number

XAD-purified gelatin C543 10,580 ± 290 Imino acids from gelatin C556A 10,310 ± 620 phase C556B 10,580 ± 360 Gelatin C480 10,250 ± 310 0.6 N HCL acid-soluble C477 9810 ± 460 organic Inorganic CO from acetic C615 9250 ± 260 2 acid residue Raw bone C474B 7940 ± 470 Bone carbonate C662 7180 ± 400 Fulvic acids, gelatin method CFFI 5210 ± 280 Fulvic adds from 6 N HCL C561 5070 ± 310 Lipid fraction - C783 1640 ± 490 Fulvic . acids, conc. C573 recent

a. Leonhardy (1966). 125

Analysis of the mammoth bone included CHN, Kjeldahl, and quantitative amino acid analyses to evaluate the degree of collagen pre servation in addition to 230Th/234U dating and direct 14C determina tions (tables 4 and 5).

The moderately hard, light-brown (10 YR 5/4) mammoth bone had its cancellous tissues impregnated with light-olive-gray clayey silt, and its outer surface sporadically coated with "Gelva," an acetone- soluble preservative. Following removal of all cancellous tissue and hand scraping of the surfaces, the bone was broken into ~O. 5 cm fragments and washed three times at room temperature with acetone to remove preservative before drying and grinding to <125 lim with a mortar and pestle. Aliquots of this powder were left untreated or reacted with

1 M acetic acid, and the various fractions either combusted with CuO powder for total carbon or reacted with 95% H P0 to evolve inorganic 3 4 carbonate CO (table 4; fig. 24). The remaining bone powder (100 g) 2 was dissolved in 0.6 N HCl at 6°C and 23 different organic fractions were isolated. Hydroxyproline and proline were separated from the water-soluble imino acid phase by HPLC methods that used 1 N HCl as an eluting buffer. The sample was dissolved in approximately 500 III

0.1 N HCl and injected onto a 250 x 10-mm glass column of Benson

BC-X8, 20-25 micron cation exchange resin. The 1 N HCl was pumped with a Milton-Roy dual piston pump with titanium liquid ends, and the 4 imino acid peaks were detected by an ISCO V absorbance monitor set at 215 nm. The CO for TAMS 14C dating was generated from organic 2 residues by the methods of Jull and others (1983) in which approximately 10 mg of sample are combusted with 500 mg CuO powder 126

Table 5. Amino acid and uranium series analyses of acid-insoluble fraction from Domebo mammoth bone from the Domebo Site, Oklahoma

Micrograms Amino Acid Residues/l000 per gram bone

4-hydroxyproline 71 60.9 Aspartic acid 59 50.4 Threonine 24 20.4 Serine 34 29.3 Glutamic acid 85 72.9 Proline 111 95.1 Glycine 296 253.5 Alanine 130 111.5 Valine 36 31.0 Methionine 6 5.2 Isoleucine 15 12.7 Leucine 36 30.9 Tyrosine 7 6.1 Phenylalanine 16 14.1 Histidine 4 3.8 Hydroxylysine o 0.0 Lysine 26 22.4 Tryptophan 2 1.8 Arginine 41 35.2 Tot?lls 999 857.2

Kjeldahl nitrogen: 434 ppm (cortical bone)

CHN analysis: Cortical bone 5.19% C, 0.64% H, 0.69% N Cancellous bone 4.37% C, 0.59% H, 0.24% N

a Uranium Series Analyses

234U/238U = 1.14 ± 0.02 238U = 4.57 ± 0.09 ppm 2 230Th/232Th = 455,551 x 10 230Th/234U = 0.09 ± 0.00 Corrected age = 9640 yr +525 Age (yr) = 9512 -400

Sample B-53 11,500 ± 2000 yr B.P. b (uranium series) a. Bischoff (1983, personal commun.) b. Szabo (1980). RAW BONE I INITIAL CLEANING (hand scraping. break Into 1/2cm fragments. ultrasonify. dH20. acetone extraction) I GRIND TO < 125,wn POWDER

A ce tic A cid ;..T~r~e~a~tm~e::n~t____ -.:S~o~/~v:..::e~n~t~E~x~tr~a~c~t::i~o~n __-r~-;::-;~~;::7;:::-1- ___"::G~e~/a~t~i~n~M~1 CLEAN BONE Soxhlet 0., extraction. POWDER de Acetone

Combustion: H~ O. EXTRACTED SOLUBLE INSOLUBLE 1M Acetic Acid TOTAL Hydrolysis: BONE POWDER ORGANICS RESIDUE 24h. Soluble J CARBON TOTAL INORGANIC Phase Discarded; ~ __~J~ __~ CARBON H20 Wash 1'640 ± 490 I \ 7940±470 ! 17180±4001 ACETIC ACtO WASHED 0.6NHC/ BONE POWDER decalcification

INSOLUBLE SOLUBLE RESIDUE (discard) ! 1'0250 + 310 r---:'UPERNATANT INSOLUBL • - "GELATIN" RESIDUE Itt.,00 ± 1000 I INSOLUBLE SOLUBLE INORGANIC, • ORGANICS (discard) CARBONATE I I ! 6N HCI hydrolysis 1,0.700 ± 310 I 19254 ± 260 I INSO~ATANT RESIDUE I

XAD-2 COLUMN ~ ADSORBED FULVIC EFrLUEN~ Inorganic and Organic ACID ELUTED WITH ,NH'OH tlitrc Carbon Extraction Scheme' 4:1 15210±2801 For Domebo Mammoth ACETONE/ET 14 C Dates

INSOLUBLE~ l Associated Wood Dates 10.580 ± 360 POLAR PHASE NON- lO.310! 620 "IMINO (CIDS" ~ 11,140±280 CHROMA TOGRAPHIC 11,045±647 SEPARATION

PROLINE~ HYDROXYPRO

Figure 24. Flow diagram showing radiocarbon dates on fractions isolated from a mammoth of known age from the Domebo Site, Oklahoma -.-, -" 127

RAW BONE I INITIAL CLEANING (hand scraping. break Into 1/2cm fragments. ultrasonily. dH20. acetone extraction) I GRIND TO < 125,um POWDER traction "Gelatin Methodn 11Hydrolysis Method" CLEAN BONE rhlet O.BNHCI O.BN HCI :ractlon, POWDER decalcification decalcification etone

H3 O. SOLOBLE INSOLUBLE SOLUBLE INSOLUBLE SOLUBLE Hydrolysis: ORGANICS RESIDUE ORGANICS RESIDUE ORGANICS , TOT AL INORGANIC I CARBON 1640 ± 490 1 7940 ± 470 I 17180±4001

:/on "HYDROLYSIS METHOD"

BN HCI, 24hrs, 105C LE under N2 gas 'd) 1'0250 + 310 ~UPERNATANT INSOLUBLE ,- "GELA TIN" RESIDUE INSOLUBLE RESIDUE SUPERNAT ANT I "HUMIC ACID" I 6N HCI hydrolysis LVOpJLlIZ. ~ INSO~ATANT ,...----..., ADSORBED FULVIC EFFLUENT RESIDUE I 15070 ± 3 10 ~ ACID ELUTED "HYDROLYSED WITH NH'OH AMINO ACIDS"

XAD-2 COLUMN

ADSORBED~ FULVIC EFFLUENT-1 10,5S0 ± 290 1 ACID ELUTED Nitrosylation WITH NH.OH Nitrosylation 4:112hfHC/: , 4:112MHCI:1SMHN03 IBM HN03 15210 ±2S01 ~ ACETONE/ETHER ACETONE/ETHER EXTRACTION

INSOLUBLE~ SOLUBLE INSOLUBLE.~ SOLUBLE PHASE 10.580:!: 360 POLAR PHASE NON-POLAR PHASE POLAR PHASE "HYDROXY ACIDS' 10,310! 620 "IMINO ACIDS' .--_..1.' __--... "IMINO ACIDS' r---'-(__ --, I 1,1 ,630 ! 450 1 1"·840!4201 CHROMA TOGRAPHIC

CHROMA TOGR APHIC SEPARATION PROLINE~ HYDROXYPROLINE PROLINE~ HYDROXYPROLINE showing radiocarbon dates on known age from the Domebo Site,

128 and 10 mg Ag powder at 1000oC, and the CO subsequently reduced by 2 O using Mg turnings and 1-2 mg Cd metal as a catalyst. The carbon O was mixed with 17 times its weight in Fe powder and heated to approximately 15000 C to form an iron carbide (Fe C) bead that was 3 pressed into an aluminum sample holder (Jull and others, 1983).

The 14C_13C ratios were measured by tandem accelerator mass spectrometry at the National Science Foundation facility at The

University of Arizona (Donahue and others, 1983; Zabel and others,

1983) and are presented in figure 24 and table 4 as l4C yr B. P.

Results

The Domebo mammoth bone is diagenetically similar to many late

Pleistocene-Holocene fossils because a considerable amount (90%) of its organics was removed during burial; the percent nitrogen values of

0.43, 0.69, and 0.24 indicate only a tenth of the normal collagen content. The residual protein has the amino acid composition for modern collagen; therefore milligrams of imino acids can be isolated from the

Domebo bone.

The age of the Domebo mammoth was independently established by using wood that dated 11,045 ± 647 yr (SM-695). Bark from the same log was l4C-dated by accelerator mass spectrometry and had an age of 11,140 ± 280 yr (C-987). The number of tree rings in the wood was 94 ± 2 (M. A. Thompson, 1984, personal commun.).

The data in table 4 and figure 24 indicate that 6 of the 19 bone dates, fractions 3-8 inclusive, are statistically identical to the wood date. Two fractions, C-555 and C-551, are significantly older and are 129 hydroxy acids, which are the nitrosylation derivatives of primary amino acids. Six dates, fractions 9-14, are younger than the established age.

In addition to the radiocarbon age determinations, a 230Th/234U + 525 age of 9512 _ 400 yr was calculated for the same Domebo mammoth cortical bone by Bischoff (1983, personal commun.) (table 5). Szabo

(1980) previously published a uranium series age of 11,500 ± 2000 yr.

Discussion

The l4C dates from Domebo mammoth bone fractions included dates concordant and discordant with the known age. Total bone carbon (collagen + apatite) (C-474B: 7940 ± 470 yr) and inorganic apatite carbon (C-662: 7180 ± 400 yr) from Domebo were younger than the known age. For bone leached 24 hr with 1 M acetic acid, the inorganic

CO was still too young (C-615: 9250 260 yr); however, the age of 2 ± combusted total bone carbon after 1 M acetic acid was 10,700 ± 300 yr

(C-542). The weak acid (0.6 N HCn-soluble phase is commonly dated because severely degraded bones have more organics that are soluble than insoluble in HC}' The 14C date of 9810 ± 460 yr B.P. of the

HCI-soluble organics was too young, a finding that signified that soluble organic(s) had been adsorbed or absorbed by the fossil. Gel- atin or hot water-soluble collagen was first dated by Longin (1971), and variations of this method are widely used today. The age (10,250 ± 310 yr) obtained from the gelatin (fraction 9) was younger than expected, and I believe that the contamination resulted from absorbed humates because when they were removed by using XAD-2 resin, the gelatin 14 extraction yielded an older C age of 11,270 ± 700 yr. 130

Hydroxy acids from the nitrosylation reaction dated 11,840 ± 420 and 11,630 ± 450 yr B. P • Although these compounds are primary amino acids whose amine groups have been replaced by OH and would not be expected to differ in age from the secondary collagenous amino acids, the hydroxy acids were apparently contaminated during their extraction into the acetone and ether used to separate the primary amine derivatives from the more polar imino acids. The evaporated hydroxy acids had an aromatic, fruity odor reminiscent of natural esters, and the hydroxy acids possibly combined with one of the organic solvents during evaporation in the presence of traces of HCI. This reaction would permanently bind petroleum-derived hydrocarbons with no 14C activity to the primary amino acids and would result in a 14C age that was too old. Independent evidence for foreign carbon addition is the 013 C value of -15.4 per mil for the hydroxy acids isolated from Bison bone

(fig. 20); this value is considerably lighter than the average value of

-10.0 per mil for the hydroxyproline or proline and the -11.1 per mil for the hydrolysate preceding the nitrosylation reaction. Acetone and 13 ethyl ether are petroleum byproducts and would have 0 C values more negative than -25 to -30 per mil.

The next six fractions were statistically identical with the known age and illustrate that although imino acids are the most specific and diagnostic fractions of fossil bone, several other phases are equally usable. The advantage of using hydroxyproline is that little question remains regarding elimination of contamination. The only prominent source of nonbone hydroxyproline is extensin, a protein in all plant cell walls that is abundant but also highly resistant to hydrolysis (Kuttan 131 and Radhakrishnan, 1973; Sadava and others, 1973). Utilization of the gelatin pathway would eliminate extensin, which is not hot water soluble. Despite its specificity to collagen, the extraction of hydroxy proline requires 30-35 work hours over 6-7 days and relatively expen sive and sophisticated liquid chromatography equipment. Fractions pre ceding the isolation of imino acids such as XAD-purified hydrolysates or even combined imino acids require only 10-15 hours of work over 3-H days and yield as accurate dates as pure hydroxyproline. Unusual geochemical settings such as lakes, marshes, or marine environments with high amounts of organic contaminants might encourage dating of bone by the hydroxyproline method. If the bone is poorly preserved and hydroxyproline sparse or absent, other fractions would provide reliable if not the only attainable 14C dates.

Conclusions

Radiocarbon dating has shown that at least 6 fractions from a fossil bone yielded accurate 14C dates and of those 6, hydrolysates purified with XAD resin, the individual imino acids hydroxyproline and proline or a mixture of those imino acids would be good choices.

Whole-bone and apatite CO dates were found contaminated, whereas 2 soluble organics should be used as a last preference. How extensive a bone should be 14C dated must depend on its diagenetic history, the lithology of enclosing sediments, and funding available for pretreatment. CHAPTER 5

CONCLUSIONS

Combining stratigraphic and geochemical data throughout the

Llano Estacado eliminates many local geologic aberrations and yields a reliable approximation of regional paleoecological conditions for the last ca. 13,000 years. The alluvium in the Yellowhouse and Blackwater

Draws is geologically differentiable into three mappable rock strati graphic units. The first are the unnamed, pre-ca.-13,000-yr. B.P.

Rancholabrean fluvial sediments present as remnant buried terraces; next younger are the fluviolacustrine members 1, 2, and 3, of ca.

12,600-4900-yr B. P. fluviolacustrine Yellowhouse Draw formation, and the young~st are the 4900-yr B. P. to modern fluvial-eolian-paludal sedi ments of members 4 and 5 in the upper Yellowhouse Draw formation.

The preserved Rancholabrean fluvial sediments are the remnants of a thick Wisconsinan stream deposit, which was subsequently incised before ca. 13,000 yr B. P. The present valley bedrock configuration developed during this terminal Pleistocene erosion when additional bed rock was removed. No later erosion would reach to the bedrock base of the valleys. An effluent but constantly dropping water table existed from the late Pleistocene to 4900 yr B. P., a period during which per manent, moderate sinuosity streams at ca. 12,600-11,500 yr B.P. evolved into oligotrophic lakes from 11,500-11,000 yr B.P. Eutrophi cation increased until 8000 yr B. P., after which alkaline lacustrine and

132 133

pond conditions existed until 4900 yr B. P. The water table gradually

dropped between 13,000 and 4900 yr B. P., but ca. 8000 yr B. P. there

was a sudden drop then rise in a ground-water level. High-elevation

pond sediments of member 2 were exposed and eroded; however, ponds

persisted in the center of the valleys. When deposition of member 3

began a few hundred years later, the ponds and marshes transgressed

the eroded member 2 surface and eventually member 3 sediments filled

the valley to higher elevations than those inundated during member 2

time.

A major regional sedimentary change occurred ca. 4900 yr B.P.

when fluvial events scoured both valleys, member 4 deposition began,

and the region attained a near-modern climate. The stratigraphy of

members 4 and 5 of the Yellowhouse Draw formation is an alternating

series of progressively weaker soils and eolian-colluvial sands remi

niscent of' modern depositional cycles. The low water table that had

enabled calcic soil horizons to form in the valley center raised,

transgressed upstream ca. 2600 yr B. P., and formed cienegas still per

sisting in downstream stretches of Yellowhouse Draw.

The o13C values from fossil bone collagen becomes more nega

tive (lighter) as age of the fossil increases. I conclude that the late

Pleistocene grasslands had a C 4 biomass of 30 to 40 percent and that

this percentage changed to >95 percent C by modern times. The iso 4 topic data parallel the stratigraphic conclusions, and both suggest that

climatic change was gradual until 8000 yr B. P. By 5000 yr B. P. the

region had attained essentially modern climatic conditions. The rate of isotopic change between 8000 and 5000 yr B. P. was not calculatable, 134

because fossils have not been found in sediments of that time interval;

however, sedimentologic data indicate a rapid change to modern condi

tions. During late member 3 time (6500-5000 yr B. P. ), eolian influx

into both valleys markedly increased, and by 4500 yr B.P. recent-type

pedogenesis of calcic Ustolls had begun.

The geologic and isotopic information contributes to an under

standing of the late Pleistocene megafaunal extinction by showing the

relative timing of sedimentologic and vegetative events. The extinction

at the Lubbock Lake Site was dated at 11,050 yr B.P., and at that time

the valleys were undergoing a transition from oligotrophic to eutrophic

lacustrine conditions and the grasslands may have been dominated by

C -type grass species. The only major sedimentologic change, a shift 3 from fluvial to lacustrine deposition, preceded the extinction by at least

500 years; no other major sedimentologic or hydrologic change was

identified until 8000 yr B. P. Clearly, a catastrophic climate event is

not coincident with the extinction; however, environmental change can

not be entirely excluded because a significant lag time may have existed

between the water-table drop and its direct effect on the animals.

Another regional implication of the isotopic-geologic data is that

episodes of erosion and deposition were not associated with climatic

changes but were apparently the culmination of long-duration, naturally

evolving sedimentologic cycles. The changes in lacustrine geology of

members 1, 2, and 3 are not coincident with any known external event.

Likewise, the mUltiple pedogenesis and deposition in members 4 and 5 had no obvious regional climatic origin. The initiation of the various

geologic events may have been caused by the crossing of discrete 135 thresholds. The 4900-yr B. P. regional erosion and later cyclic pedo genesis were threshold-related events. I imply that a modern threshold exists and that on either side of it erosion or deposition occurs, de pending on how rainfall varies from year to year. The amount of pre cipitation need change only a few centimeters from equilibrium to initiate a marked geomorphic change.

Investigation of the stable and radiocarbon isotopes in the fossil bones revealed that accurate 14C dates are attainable on skeletal remains and that isolation of specific amino acids is not always necessary for accurate radiocarbon dating of bones. The most reliable and easily obtained 14C bone dates would be on hydrolyzed collagen passed through XAD-2 resins.

Reliable stable carbon isotopic results from bones requires the isolation of individual amino acids. Total collagen hydrolysates will approximat'e \S13 C changes over time, but as collagen degrades, its amino acid composition changes and the residual II collagen II becomes isotopically fractionated. The amount of isotopic change in collagen depends on the degree of protein loss during burial: this factor cannot be reliably evaluated or corrected for at the present. The second com ponent of bone, the mineral or carbonate-hydroxyapatite phase, was found too susceptible to exogenous carbonate exchange; even for spec imens treated to remove foreign CO;2, the stable carbon isotope data were spurious. APPENDIX

GEOLOGIC DESCRIPTION OF TRENCH 65

136 137

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description

0-40 fill Dredging debris from 1936 Dis conformity • • • • • • • • • • • • Yellowhouse Draw formation, ca. 12,500 yr B.P. to present 40-63 Al 5 Sand, gray-brown (10 YR 5/2d), dark-brown (10 YR 313m) fine, clayey; moderately hard; weak fine, subangularblocky and moderate very fine granular; granular texture caused by worm burrows; effer vescent (5) ; < 1% fine carbonate pebbles and discontinuous 2-cm lenses of fine carbonate pebble beds; bone common throughout; clear, wavy base; unit distin guishable from Alb by a more granular, softer consistency. 63-75 5 Sand, gray-brown (10 YR 5/2d), daJ:'k'-brown (10 YR 313m), silty, very fine; hard; weak, coarse, subangular blocky, breaking to weak, very fine granular; effer vescent (5); very weak stage I carbonate; 1%-2% of 5 mm long x 0.5-mm-diameter root molds have CaC0 films coating interiors. 3 Scattered, 2%-5% very fine to fine carbonate pebbles; base is grad ual by color but abrupt, wavy using filamentous carbonates. 75-85 5 Lithology identical to that for 63-75 cm, except filamentous carbonate absent; clear, smooth base. 85-90 5 Sand, pale-brown (10 YR 6/3d), dark yellow-brown (10 YR 4/4m), silty, very fine to fine; moderately hard to hard; weak, very fine sUbangu lar blocky; 10%-15% 2-mm-diameter yellow-brown mottles within gray brown matrix; scattered granule lenses; effervescent (5); visible carbonate absent; abrupt, wavy base. 138

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description

90-97 5 Clay, light-brown 10 YR 6/4d), yel low-brown (10 YR 5/4m), silty, fine sandy; hard; weak to moderate, coarse subangular blocky; effer vescent (5); abrupt, smooth base.

97-103 5 Clay, light-gray (10 YR 6/1d), dark brown (10YR 313m), silty, very fine-fine sandy; slightly hard; weak, very fine subangular blocky; effervescent (5); very weak A hori zon; very dark gray organic matter disseminated throughout a yellow brown matrix; discontinuous, disap pears 4 m to the south; abrupt, smooth base.

103-113 5 Sand, very pale brown (10 YR 7/4d), dark-yellow-brown (10 YR 4/4), clayey, very fine to fine sandy; hard; weak, coarse blocky; effer vescent (5); «1% faint filamentous CaC03 within root hair molds; abrupt, smooth base.

113-118 5 Sand, light-gray (10 YR 7/2d), brown (10 YR 4/3m), clayey very fine to fine; hard; weak, fine sub ang'ular blocky; effervescent (5); discontinuous, very weak A hori zon; abrupt, smooth base.

118-123 Cl 5 Sand, pink (7.5 YR 7/4d), brown (10 YR 4/3m), silty very fine to fine; hard; massive; effervescent (5); abrupt, smooth base.

123-126 5 Sand, gray-brown (10 YR 5/2d), dark-brown (10 YR 313m), silty, fine; hard; massive; moderate A horizon development; abrupt, smooth base. 139

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description

126-129 C2 5 Clay, light-gray (10 YR 7/1d) to gray (10 YR 6/1d), black (10 YR 211m), silty coarse to very coarse carbonate sandy; soft to slightly hard; massive; abrupt, smooth base.

129-132 C3 5 Sand, light-yellow-brown (10 YR 6/4d), yellow-brown (10 YR 5/4m), silty, fine; hard; massive; effer vescent (5); scattered carbonate granules; abrupt to extremely abrupt, smooth and wavy base.

132-142 4 Sand, dark-gray-brown (10 YR 4/2d), very dark gray brown (10 YR 312m), clayey, fine, hard; weak medium to coarse subangular blocky, breaking to moderate very fine sub angular blocky; efferves cent (4); 2% I-em-diameter burrows and worm burrows, both filled with white silt; abrupt to clear, wavy base.

142-150 4 Sand, dark-brown (10 YR 4/3d), dark-brown (10 YR 3/3;n), silty, fine; slightly hard, weak coarse subangular blocky; effervescent (4); abrupt, wavy to irregular base.

150-156 A3 b 4 Sand, white (10 YR SlId), pale 5 ca brown (10 YR 6/3m), silty, fine, with mottles of dark-brown (10 YR 4/3d), dark-brown (10 YR 3/3), constituting 40%-50% of matrix in upper portion and 5% in lower sec tion; very hard, moderate, medium subangular blocky, breaking to weak, very fine sub angular blocky; effervescent (5); carbonate in creases toward base; clear, wavy base. 140

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description lS6-165 B21 b 4 Clay, white (10 YR SlId), brown ca 5 (10 YR 5/3m); silty, fine sandy; very hard; medium sub an gular blocky; effervescent (S); Stage III carbonate development (unit with strongest carbonate development); clear, wavy base.

16S-17S B22 b 4 Sand, pink-white (7.5 YR S/2d), ca S brown (7.S YR S/4m); silty, fine; carbonate masks faint background color of pink to light brown (7.5 YR 7.4, 6/4d); very hard; modt!r ate to strong, fine to medium sub angular blocky; effervescent (5); Stage II to III carbonate develop ment; S5%-60% distinct, medium to coarse carbonate mottles; 30%-50% less CaC0 than 156-165 cm; diffuse 3 base.

17S-190 . B23 b 4 Sand, yellow-brown (10 YR S/4d), ca 5 dark-yellow-brown (10 YR 4/4m); very hard; weak medium subangular blocky; faint to distinct, 10%-15% fine carbonate mottles, including I-mm-diameter filamentous carbonate (7%-10%); effervescent (S); Stage I - II; transition zone; clear, wavy base.

190-203 B24 b 4 Sand, white (10 YR SlId) to faint ca S pink (10 YR 7/4d), brown (7.S YR 5/4m); clayey, fine; very hard; very weak, fine subangular blocky to massive; effervescent (5); Stage III carbonate; strong calcic horizon; carbonate disseminated throughout matrix; clear, wavy base. 141

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description

203-212 B2S b 4 Sand, 40% light-brown-gray (2.S Y S ca 6/2d,m) and 60% brown (7.S YR S/4m), light-brown (7.S YR 6/4d) clayey fine; very hard; weak, fine, sub angular blocky: brown coloratioll occurs as 1-2-mm wide, 10-1S-cm long, vertical, wavy streaks run ning through gleyed matrix; very faint, 3%-S% fine carbonate mottles; highest elevation occurrence of gleyed sediments of member 4; effervescent (S); gradual base.

212-234 C 4 Sand, light-gray (S Y 7/2d), pale 40x olive (S Y 6/3m), fine, silty-clayey, with 30%-40%, o. S-2-cm mottles of light-brown (7.S YR 6/4d), brown (7. S YR S/4m); hard, massive; ef fervescen t ( 4-S); no visible car bonate mottling; brown oxidized sands vertical as in 203-212 cm, ex cept 1-3 cm wide; gradual, irregu lar base.

234-273 CSgca 4 Sand, light-gray (S Y 7/2d), pale olive (S Y 6/3m), fine to medium quartz; slightly hard, massive; thick interbedded fine to medium quartz sand, silty sand, and very fine to fine sandy clay, all massive; sands contain no carbonate mot tling, sandy clay has >20% carbon ate mottles, often concentrated along sand-clay bedding plane; sands have rare small-scale trough cross-stratification; abrupt, wavy to irregular base. 142

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description

273-302 C6g 4 Sand, brown (7.5 YR 5/4d), dark brown (7.5 YR 4/4m), silty, very fine; slightly hard; massive to weak fine angular blocky; effervescent (3); 1%-2% scattered I-mm-diameter carbonate· filaments; soft sediment deformation has severely deformed the large- to small-scale trough cross-stratification; tongues of stratum 3C extend upward into C6g; C6g sands are incorporated into stratum 3C; extremely abrupt, wavy, irregular base...... • Disconformity 302-342 IIC7 3C Clay, light-brown-gray (2.5 Y 6/2d), dark-gray-brown (2.5 Y 4/2m), sil ty; very hard; medium to coarse prismatic, breaking to strong, medi um to coarse angular blocky; effer vescent (4); contains 7%-15%, 0.5- 2-mm-diameter CaCa particles (autochthonous) and r-cm-long exter nal molds of Xanthium seed pods; extremely abrupt, wavy base.

342-373 IlIC8 3B Marl, white (10 YR 8/1d), white (10 YR 812m), clayey, with 30% to 50% fine to medium mottles of light-gray (10 YR 7/1d), gray-brown (10 YR 5/2m); soft to slightly hard; very fine subangular blocky (carbonate) and weak very fine to fine, soft subangular blocky (gray mottling); laterally a facies of moderate, very fine to fine platy marl; effervescent (5); clear, wavy base. 143

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description

373-390 IIIC9 3A Clay, white (10 YR 8/1d), brown (10 YR S/3m), silty; slightly hard; weak medium subangular blocky to weak, very fine columnar, breaking to moderate, very fine subangular blocky; effervescent (S); columnar structure increases in strength to ward base; gradual, wavy base.

390-401 IIICI0 3A Clay, light-gray (10 YR 7/1d), gray brown (10 YR S/2m), silty; slightly hard; moderate to strong, very fine columnar; effervescent (S); S%-7% vertical, white, very fine to fine silicified rootlets; abrupt to ex tremely abrupt, wavy base.

401-411 2C Silt, gray-brown (10 YR S/2d), very dark gray brown (10 YR 312m); soft; moderate, very fine subangu lar blocky; effervescent (1); con tains 7%-10% silicified rootlets in upper S cm, increasing abruptly to 30%-40% in lower Scm; rootlets are white, very hard, horizontal, 0.S-2 mm in diameter; extremely abrupt, wavy base.

411-418 IVCll 2B Clay, gray (10 YR SlId), very dark gray (10 YR 311m), silty; very hard; strong, very fine to fine prismatic and columnar; ped sur faces stained dark yellow brown (10 YR 6/3d) with limonite; clear, wavy base. 144

Depth Below Ground Surface Pedogenic Geologic (em) Horizon Horizon Description

418-430 VC12 2 Silt, light-brown-gray (2.S Y 6/2d), olive (S Y 5/3m), very fine sandy; hard to moderately hard; weak coarse angular blocky; effervescent (1); 10%-15% horizontal and oblique fine, I-em-long silicified rootlets; base exhibits soft sediment deforma tion; abrupt to clear, wavy to ir regular and broken base.

430-4S1 VIC 13 2 Silt, dark-gray (10 YR 4/1d), black (10 YR 211m); very hard; massive, effervescent (1); with discontinuous 2-3-cm-thick bed of vertical and oblique silicified plants stems; extremely abrupt to abrupt, wavy base.

4S1-4f30 VIIC14 2 Diatomite, white (10 YR 8/1d), dark gray (10 YR 4/1m) interlaminated with dark-gray (10 YR 4/1d), black (10 YR 211m) carbonaceous diatom ite; hard; massive pure diatomite 80% of unit thickness; with < l-cm thick laminae of horizontal silicified rootlets; internally, diatomite beds contain 1-3-mm-thick, gray laminae; horizontal, carbonized plant frag ments on bedding planes; extremely abrupt to clear, wavy and irregular base.

480-S10 CIS lC Sand, light-brown-gray (2.5 YR 6/2d), gray-brown (2.S Y 5/2m), clayey fine; very hard; weak coa!"se angular blocky; ped surfaces stained dark yellow brown (10 YR 4/6d); effervescent (2); with S%- 10%, 0.5-2-cm subround to subangu lar carbonate pebbles increasing to 2S% at base; numerous extinct Ran cholabrea species; clear, wavy base. 145

Depth Below Ground Surface Pedogenic Geologic (cm) Horizon Horizon Description

510-555 C16 IB Gravel, white, fine to medium sandy; 0.5-3-cm carbonate gravel with clasts to 4 cm; massive, abrupt, irregular base.

555-610 C17 lA Gravel, white, medium sandy; loose quartz sand and 3-5-cm subangular to subround carbonate gravel; mas sive; extremely abrupt, wavy base.

Disconformity

Blanco Formation (late Pliocene-early Pleistocene)

610-650+ CIS 01 Sand, light-gray (2.5 Y 7/2d), clayey; hard, massive; base not exposed. REFERENCES

Adams, Elijah, 1978, Invertebrate collagens: Science, v. 202, p. 591-598.

Adams, Elijah, and Frank, Leonard, 1980, Metabolism of proline and the hydroxyprolines: Annual Review of Biochemistry, v. 49, p. 1005-1061.

Aiken, G. R., Thurman. E. M., Malcolm, R. L., and Walton, H. F., 1979, Comparison of XAD macroporous resins for the concentration of fulvic acid from aqueous solution: Analytical Chemistry, v. 51, p. 1799-1803.

Allred, B. W., 1956, Mixed prairie in Texas, in Weaver, J. E., and Albertson, F. W., eds., Grasslands of the Great Plains: Lincoln, Nebraska, Johnsen Publishing Co., p.267-283.

Albritton, C. G., Jr., 1966, Stratigraphy of the Domebo site, in Leonhardy, F. C., ed., Domebo: A Paleo-Indian mammoth kill in the prairie-plains: Lawton, Oklahoma, The Great Plains Histor ical Association, Contributions of the Museum of the Great Plains No.1, p. 11-13. ' ..

Antevs, Ernst, 1935, The" occurrence of flints and extinct animals in pluvial deposits near Clovis, New Mexico. Pt.II. Age of the Clovis deposits, in Proceedings, Philadelphia Academy of Natural Sciences: -p. 304-311. Antevs, Ernst, 1949, Geology of the Clovis sites, in Wormington, H., Ancient man in North America: Denver, Colorado p Denver Museum of Natural History, Popular Series 4, p. 185-192.

Armstrong, W. G., Halstead, L. B., Reed, F. B., and Wood, Liliana, (1983), Fossil proteins in vertebrate calcified tissues: Philosophical Transactions of the Royal Society [London], v. -B301, p. 301-343.

Beevers, R., Stitter, M. L., and Butt, V. S., 1966, Metabolism of the organic acids, in Steward, F. C., ed., Plant Physiology, Vol. IV B: New York, Academic Press, p. 119-261.

Berger, Rainer, Horney. A. G., and Libby, W. F., 1964, Radiocarbon dating of bone and shell from their organic components: Science, v. 144, p. 999-1001.

146 147

Berglund, B. J., H~kansson, S6ren, and Lagerlund, Erik, 1976, Radiocarbon-dated mammoth (Mammuthus primigenius Blumenbach) finds in south Sweden: Boreas, v. 5, p. 177-191.

Bischoff, J. L., Personal communication: Staff geologist, U.S. Geo logical Survey, Menlo Park, California.

Black, Craig, ed., 1974. History and prehistory of the Lubbock Lake Site: Lubbock, West Texas Museum Association, The Museum Journal XV.

B6hlen, Peter, Stein, Stanley, Dairman, Wallace, and Udenfriend, Sidney, 1973, Fluormetric assay of proteins in the nanogram range: Archives of Biochemistry and Biophysics, v. ISS, p. 213-220.

Boutton, T. W., Harrison, A. T., and Smith, B. N., 1980, Distri bution of biomass species differing in photosynthetic pathway along an altitudinal transect in southeastern Wyoming grassland: Oecologia, v. 45, p. 287-298.

Brand, J. P., 1974, Guidebook to the Mesozoic and Cenozoic geology of the southern Llano Estacado: Lubbock, Texas Tech University, Department of Geosciences, Lubbock Geological Society.

Bremner, J. M., 1950, The amino-acid compostion of the protein ma,terial in soil: The Biochemical Journal, v. 47, p. 538-542.

Broecker, W. S., and Kulp, J. L., 1957, Lamont natural radiocarbon measurements IV: Science, v. 126, p. 1324-1334.

Brooks, T. M., 1981, Hydroxylapatite. Fast flow and high resolution. Document Number 8158-282: San Diego, California, Calbio chern-Behring.

Brown, W. V., 1977, The Kranz syndrome and its subtypes in grass systematics: Memoirs of the Torrey Botanical Club, v. 23, p. 1-97.

Brownlee, Michael, and Spiro, R. G., 1979, Glomerular basement membrane metabolism in the diabetic rat: Diabetes, v. 28, p. 121-125.

Buchanan, D. L., and Corcoran, B. J., 1959, Sealed tube combustions for the determination of 14C and total carbon: Analytical Chemistry, v. 31, p. 1635-1638.

Calvin, M., and Benson, A. A., 1948, The path of carbon in photo synthesis: Science,. v. 107, p. 476-480. 148

Chisholm, B. S., Nelson, D. E., and Schwarcz, H. P., 1982, Stable carbon isotope ratios as a measure of marine versus terrestrial protein in ancient diets: Science, v. 216, p. 1131-1132.

Compton, J. L., 1975, Diatoms of the Lubbock Lake Site, Lubbock County, Texas: unpublished M. S. thesis, Texas Tech Univer- sity, Lubbock.

Cotter, J. L., 1937, The occurrence of flints and extinct animals in pluvial deposits near Clovis, New Mexico. Part IV, Report on excavations at the gravel pit, 1936: Philadelphia Academy of Natural Sciences, Proceedings, v. 89, p. 2-16.

Cotter, J. L., 1938, The occurrence of flints and extinct animals in pluvial deposits near Clovis, New Mexico. Part VI, Report of the field season of 1937: Philadelphia Academy of Natural Sciences, Proceedings, v. 90, p. 113-117.

Craig, Harmon, 1953, The geochemistry of the stable carbon isotopes: Geochimica et Cosmochimica Acta, v. 3, p. 53-92.

Curtis, M. A., Witt, A. F., Schram, S. B., and Rogers, L. B., 1981, Humic acid fractionation using a nearly linear pH gradient: Analytical Chemistry, v. 53, p. 1195-1199.

DeNiro, M. J., and Epstein, Samuel, 1978a, Carbon isotopic evidence for different feeding patterns in two Hyrax species occupying the same habitat: Science, v. 201, p. 906-908.

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

Dennison, K. J., 1980, Amino acids in archaeological bone: Journal of Archaeological Science, v. 7, p. 81-86.

DesMarais, D. J., and Hayes, J. M., 1976, Tube cracker for optming glass-sealed ampoules under vacuum: Analytical Chemistry, v. 48, p. 1651-1652.

Donahue, D. J., Zabel, T. H., Jull, A. J. T., and Damon, P. E., 1983, Results of tests and measurements from the NSF Regional Accelerator Dating Facility for Radioisotope Dating: Radio carbon, v. 25, p. 719-728.

Downton, W. J. S., 1971, Adaptive and evolutionary aspects of C 4 photosynthesis, in Hatch, M. D., Osmond, C. B., and Slayter, R. 0., eds., Photosynthesis andphotorespiraton: New York, Wiley-Interscience, p. 3-17. 149

Drake, R. J., 1975, Fossil nonmarine molluscs of the 1961-63 Llano Estacado paleoecology study, in Wendorf, F., and Hester, J. J., eds., Late Pleistocene environments of the southern High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 201-245.

Dziewiatkowski, D., Hascall, V. C., and Riolo, R. L., 1972, Epimerization of trans-4-hydroxy-L-proline to cis-4-hydroxy-D-proline during acid hydrolysis of collagen: Analytical Biochemistry, v. 49, p. 550-558.

Ehleringer, J. R., 1978, Implications of quantum yield differences on the distribution of C and C grasses: Oecologia, v. 31, p. 255-267. 3 4

Ellis, R. P., Vogel, J. C., and Fuls, A., 1980, Photosynthetic path way!; and the geographical distribution of grasses in South West Africa/Namibia: South African Journal of Science, v. 76, p. 307-314.

Evans, G. L., 1949, Upper Cenozoic of the High Plains, in Hills, J. M., chairman, Cenozoic geology of the Llano Estacado and Rio Grande Valley: Guidebook field trip No.2: Lubbock, Texas. West Texas Geological Society and New Mexico Geological Society, p. 1-22.

Evans, G. L., 1951, Prehistoric wells in eastern New Mexico: Amer ic~n Antiquity, v. 17, p. 1-8.

Evans, G. L., and Meade, G. E., 1945, Quaternary of the Texas High Plains, in Contributions to Geology, 1944: Austin, University of TexasPublication 4401: p. 485-507.

Frye, J. C., and Leonard, A. B., 1957, Studies of Cenozoic geology along eastern margin of Texas High Plains, Armstrong to Howard Counties: Austin, University of Texas, Bureau of Economic Geology Report of Investigations No. 32, p. 1-62.

Green, F. E., n.d., The Lubbock Reservoir site: Lubbock, Texas Tech University, The Museum, unpublished manuscript.

Green, F. E., 1961, Discussion of the pollen and stratigraphic data, in: Wendrof, F., assembler, Paleoecology of the Llano Estacado: Sante Fe, The Museum of New Mexico Press, Fort Burgwin Research Center Publication No.1, p. 94-97.

Green, F. E., 1962, The Lubbock Reservoir site, 12,000 years of pre human history: Journal of the West Texas Museum Assodation, v. 6, p. 85-123. 150

Greenstein, J. P., and Winitz, Milton, 1961, Chemistry of the amino acids, Vol 3: New York, John Wiley and Sons.

Haas, Herbert, and Banewicz, John, 1980, Radiocarbon dating of bone apatite using thermal release of CO • Radiocarbon, v. 22, p. 2 537-544.

Hafsten, Ulf, 1961, Pleistocene development of vegetation and climate in the southern High Plains as evidenced by pollen analysis, in Wendorf, F., assembler, Paleoecology of the Llano Estacado: Santa Fe, The Museum of New Mexico Press, Fort Burgwin Research Center Publication No.1, p. 59-91.

Hamilton, P. B., and Ortiz, P. J., 1950a, Proline and hydroxyproline: determination of the sum of their a-nitrogen: Journal of Biological Chemistry, v. 187, p. 733-742.

Hamilton, P. B., and Ortiz, P. J., 1950b, Proline and hydroxyproline: purification reaction with ninhydrin, and some properties of their N-nitroso derivatives: Journal of Biological Chemistry, v. 184, p. 607-615.

Harbour, Jerry, 1975, General stratigraphy, in Wendorf, F., and Hester, J., eds., Late Pleistocene Environments of the oouthern High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 33-56.

Hassan, A,. A., 1976, Geochemical and mineralogical studies on bone material and their implications for radiocarbon dating. Un published Ph.D. dissertation, Southern Methodist University, Dallas, Texas.

Hassan, A. A., and Hare, P. E., 1978, Amino acid analysis in radio carbon dating of bone collagen, in Carter, G. F., ed., Ad vances in chemistry series, No. 171, Archaeological chemistry II: Washington, D. C., American Chemical Society, p. 109-116.

Hassan, A. A., Termine, J. D., and Haynes, C. V., Jr., 1977, Mineralogical studies on bone apatite and their implications for radiocarbon dating: Radiocarbon, v. ,19, p. 364-374.

Hatch, M. D., and Slack, C. R., 1966, Photosynthesis by sugar-cane leaves: Biochemical Journal, v. 101, p. 103-111.

Hatch, M. D., and Slack, C. R., 1970, Photosynthetic CO -fixation 2 pathways: Annual Review of Plant Physiology, v. 21, p. 141-162.

Hattersley, P. W., 1983, The distribution of C 3 and C 4 grasses in Australia in relation to climate: Oecologia, v. 57, p. 113-128. 151

Hawley, J. W., Bachman, G. 0., and Manley, K., 1976, Quaternary stratigraphy in the Basin and Range and Great Plains provinces, New Mexico and western Texas, in Mahaney, w. C., ed., Quaternary stratigraphy of North America: Stroudsburg, Pennsylvania, Dowden, Hutchinson and Ross, p. 235-274.

Haynes, C. V. [Jr.], 1967, Bone organic matter and radiocarbon dat ing, in Radioactive dating and methods of low level counting: Vienna, Austria, International Atomic Energy Agency, p. 163-168.

Haynes, C. V., Jr., 1968, Radiocarbon analysis of inorganic carbon of fossil bone and enamel: Science, v. 161, p. 687-688.

Haynes, C. V., Jr., 1970, Geochronology of man-mammoth sites and their bearing on the origin of the Llano Complex, in: Dort, W., Jr., and Jones, J. K., Jr., eds., Pleistocene and Recent environments of the Central Great Plains: Lawrence, Kansas, University of Kansas Press, p. 77-92.

Haynes, C. V., [Jr.], 1975, Pleistocene and Refent Stratigraphy, in Wendorf, F., and Hester, J. J., eds., Late Pleistocene Environments of the southern High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 57-96.

Haynes, C.' V., Jr., 1982, Were Clovis progenitors in Beringia?, in Hopkins, D., Matthews, J., Jr., Schweger, C., and Young; S., eds., Paleoecology of Beringia: New York, Academic Press, p. ~~3-398.

Hester, J. J., 1972, Blackwater Locality No. 1. A stratified, early man site in eastern New Mexico: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.8.

Hiller, Alma, and Van Slyke, D. D., 1919, Direct determination of non-amino nitrogen in the products of protein hydrolysis: Journal of Biological Chemistry, v. 39, p. 479-488.

Hirs, C. H. W., Moore. Sta.:lford, and Stein, W. H., 1952, Isolation of amino acids by chromatography on ion exchange columns; use of volatile buffers: Journal of Biological Chemistry, v. 195, p. 669-683.

Hirs, C. H. W., Moore, Stanley, and Stein, W. H., 1954, The chromatography of amino acids on ion exchange resins. Use of vOiatile acids for elution: Journal of the American Chemical Society, v. 76, p. 6063-6065. 152

Ho., T. Y., Marcus, L. F., and Berger. Ranier. 1969. Radiocarbon dating of petroleum-impregnated bone from tarpits at Rancho La Brea, California: Science, v. 164. p. 1051-1052.

Hohn, M. H., 1975, The diatoms, in Wendor:i: F., and Hester, J., eds., Late Pleistocene environment of the southern High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 197-200.

Hohn, M. H., and Hellerman, Joan, 1961, The diatoms, in Wendorf, F., assembler, Paleoecology of the Llano Estacado: Santa Fe, The Museum of New Mexico Press. Fort Burgwin Research Center Publication No.1, p. 98-104.

Holden, W. C., 1974, Historical background of the Lubbock Lake Site, in History and prehistory of the Lubbock Lake Site: Lubbock, West Texas Museum Association, The Museum Journal XV, p. 11-14.

Holliday, Vance, 1977, Cultural chronology of the Lubbock Lake Site, unpublished M.S. thesis, Texas Tech University, Lubbock.

Howard, E. B., 1936, The occurrence of flints and extinct animals in pluvial deposits near Clovis, New Mexico: Proceedings, Philadelphia Academy of Natural Sciences. v. 87. p. 299-303.

Hunt, Stephen, 1970, Polysaccharide-protein complexes in inverte br~tes: London. Academic Press.

Izett, G. A., Wilcox, R. E., and Borchardt, G. A., 1972, Correlation of a volcanic ash bed in Pleistocene deposits near Mount Blanco, Texas, with the Guaje pumice bed near the Jemez Mountains, New Mexico: Quaternary Research. v. 2. p. 554-578.

Johnson, Charles II, 1974, Geologic investigatibns at the Lubbock Lake Site, in Black, C., ed., History and prehistory of the Lubbock Lake Site: Lubbock, West Texas Museum Association, The Museum Journal XV, p. 79-105.

Johnson, Eileen, 1976, Investigations into the zooarchaeology of the Lubbock Lake Site. Unpublished Ph.D. dissertation, Texas Tech University, Lubbock.

Johnson, N., Opdyke, N. D., and Lindsay, Everett, 1975, Magnetic polarity stratigraphy of Pliocene-Pleistocene terrestrial deposits and vertebrate faunas, San Pedro Valley, Arizona: Geological Society of America Bulletin, v. 86. p. 5-12. 153

Jones, R. J., Ludlow, M. M., Troughton, J. H., and Blunt, C. G., 1979, Estimation of the proportion of C and C 4 plant species 3 in diet of animals from the ratio of natural 12C and l3C isotopes in the faeces: Journal of Agricultural Science, v. 92, p. 91-100.

Judson, Sheldon, 1953, Geology of the San Jon Site, eastern New Mexico: Smithsonian Miscellaneous Collection 121: 1.

Jull, A. J. T., Donahue, D. J., and Zabel, T. R., 1983, Target preparation for radiocarbon dating by tandem accelerator mass spectrometry. Paper submitted to the Sixth International Conference on Ion Beam Analysis, Tempe, Arizona.

Junk, G. A., Richard, J. J., Grieser, M. D., Witiak, D., Witiak, J. L., Arguello, M. D., Vick, R., Svec, H.,J. Fritz, J. S., and Calder, G. V., 1974, Use of macroreticular resins in the analysis of water for trace organic contaminants: Journal of Chromatography, v. 99, p. 745-762.

Kaczor, M. J., 1978, A correlative study of the West Texas Museum excavations at the Lubbock Lake Site, 1959-61: an example of applied museum collection management techniques within a research analysis design: unpublished M. S. thesis, Texas Tech University, Lubbock.

Kelley, J. R., 1974, A brief r~sum~ of artifacts collected at the Lubbock Lake Site prior to 1961, in Black, C., ed., History and prehistory of the Lubbock Lake Site: Lubbock, West Texas Museum Association, The Museum Journal XV, p. 43-78.

Klein, LeRoy, Faulkner, W. R., and de la Pena, Armanda, 1970, Hydroxyproline in urine and tissues: Standard Methods in Clinical Chemistry, v. 6, p. 41-56.

Kortschak, H. P., Hartt, C. E., and Burr, G. 0., 1965, Ca:t'"i:;on dioxide fixation in sugar can leaves: Plant Physiology, v. 40, p. 209-213.

K uttan, Ramadasan, and Radhakrishnan, A. N., 1973, Biochemistry of the hydroxyprolines: Advances in Enzymology, v. 37, p. 273-347.

Laetsch, W. M., 1968, Chloroplast specialization in dicotyledons pos sessing the C 4-dicarboxylic acid pathway of photosynthetic CO2 fixation: American Journal of Botany, v. 55, p. 875-883.

Lamport, D. T. A. , and Northcote, D. H., 1960, Hydroxyproline in primary cell walls of higher plants: Nature, v. 188, p. 665-666. 154

Land, L. S., Lundelius, E. L., Jr., and Valastro, Salvatore, 1980, Isotopic ecology of deer bones: Palaeogeography, Palaeoclimatology, Palaeocology, v. 32, p. 143-151.

Lee, Cindy, Bada, Jeffrey, and Peterson, Etta, 1976, Amino acids in modern and fOSRil woods: Nature, v. 259, p. 183-186.

Leonhardy, F. C., ed., 1966, Domebo: A Paleo-Indian mammoth kill in the prairie-plains: Lawton, Oklahoma, The Great Plains His torical Association, Contributions of the Museum of the Great Plains, No.1.

Leonhardy, F. C., and Anderson, A. A., 1966, Archaeology of the Domebo Site, in Leonhardy, F., ed., Domebo: A Paleo-Indian mammoth kill in the prairie-plains: Lawton, Oklahoma, The Great Plains Historical Association, Contributions of the Museum of the Great Plains, No.1, p. 14-26.

Lerman, J. C., and Troughton, J. H., 1975, Carbon isotope discrimi nation by photosynthesis: implications for the bio- and geosciences, in Klein, E. R., and Klein, P. D., eds., Pro ceedings of the Second International Conference on Stable Isotopes: Oak Brook, Illinois, p. 630-644.

Libby, W. F., Berger, Rainer, Mead, J. F., Alexander, G. V., and Ross, J. F., 1964, Replacement rates for human tissue from atmospheric radiocarbon: Science, v. 146, p. 1170-1172.

Livingstone, D. A., and Clayton, W. D., 1980, An altitudinal cline in tropical African grass floras and its paleoecological significance: Quaternary Research, v. 13, p. 392-402.

Longin, R., 1971, New method of collagen extraction for radiocarbon dating: Nature, v. 230, p. 241-242.

Lynch, D. L., Hughes, D. H., and Rhodes, Y. E., Jr., 1959, Pressure and gradient elusion in ion exchange chromatography of the amino acids in soils: Soil Science, v. 87, p. 339-344.

Lyon, T. D. B., and Baxter, M. S., 1978, Stable carbon isotopes in human tissues: Nature, v. 273, p. 750-751.

Mehl, M. G., 1966, The Domebo mammoth: vertebrate paleomortology, in Leonhardy, F. C., ed., Domebo: A Paleo-Indian mammoth kill in the prairie plains: Lawton, Oklahoma, Great Plains Historical Association, Contributions of the Museum of the Great Plains, No.1, p. 27-30. 155

Minson, D. J., Ludlow, M. M., and Troughton, J. R., 1975, Dif ferences in natural carbon isotope ratios of milk and hair from cattle grazing tropical and temperate pastures: Nature, v. 256, p. 602.

Moore, Stanford, and Stein, W. H., 1951, Chromatography of amino acids on sulfonated polystyrene resins: Journal of Biological Chemistry, v. 192, p. 663-681.

Neuman, W. F., and Neuman, M. W., 1958), The chemical dynamics of bone mineral: Chicago, University of Chicago Press.

Nier, A. 0., and Gulbransen, E. A., 1939, Variations in the relative abundance of the carbon isotopes: Journal of the Americal Chemical Society. v. 61, p. 697-698.

Oldfield, Frank, 1975, Pollen-analytical results, Part II, in Wendorf, F., and Hester, J., eds., Late Pleistocene environments of the southern High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 121-147.

Oldfield, Frank, and Schoenwetter, James, 1975, Discussion of the pollen-analytical ('vidence, in Wendorf, F., and Hester, J., eds., Late Pleistocene environments of the southern High Plains, Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 149-177.

O'Leary, ~. H., 1981, Carbon isotope fractionation in plants: Phyto chemistry, v. 20, p. 553-567.

Olsson,!. U., EI-Daoushy, M. F. A. F., A.bdel-Mageed, A. I, and Klasson, Martin, 1974, A comparison of different methods for pretrp.atment of bones. 1. Geologiska Foreningens i Stockholm Forhandlinar, v. 96, p. 171-181.

P"'J;sons, T. D., and Kavanagh, Maureen, 1982, Bone composition and the reconstruction of diet: Examples from the midwestern United States: Midcontinental Journal of Archaeology, v. 7, p. 61-79.

Patrick, R., 1938, The occurrence of flints and extinct animals in pluvial deposits near Clovis, New Mexico. Part V. Diatom evidence from the mammoth pit: Proceedings of the Philadelphia Academy Natural Sciences, v. 90, p. 15-24.

Peden, D. G., 1976, Botanical composition of bison diets on shortgrass plains: The American Midland Naturalist, v. 96, p. 225-229.

Peden, D. G., Van Dyne, G. M., Rice, R. W., and Hansen, R. M., 1974, The trophic ecology of Bison bison L. on shortgrass plains: Journal of Applied Ecology, v. 11, p. 489-497. 156

Pierce, H. G., 1975, Diversity of late Cenozoic gastropods on the southern High Plains, unpublished Ph. D. dissertation, Texas Tech University, Lubbock.

Radhakrishnan, A. N., and Giri, K. V., 1954, The isolation of allohydroxy-L-proline from sandal (Santalum album L.). Biochemical Journal, v. 58, p. 57-61.

Reeves, C. C., Jr., 1968, Introduction to paleolimnology: New York, Elsevier Publishing Co.

Reeves, C. C., Jr., 1970, Some geomorphological, structural, and stratigraphic aspects of the Pliocene and Pleistocene sediments of the southern High Plains. Ph.D. dissertation, Texas Tech University, Lubbock.

Reeves, C. C., Jr., 1972, Tertiary-Quaternary stratigraphy and geo morphology of West Texas and southeastern New Mexico, in Kelley, V. C., and Trauger, F. D., eds., East-center New Mexico, New Mexico Geological Society guidebook, 23rd Field Conference: Socorro, p. 108-117.

Reeves, C. C., Jr., 1976, Quaternary stratigraphy and geologic his tory of southern High Plains, Texas and New Mexico, in Mahaney, W. C., ed., Quaternary stratigraphy of North America: Stroudsburg, Pennsylvania, Dowden, Hutchinson and Ross., p. 213-234.

Reeves, C~ C., Jr., 1980, Personal communication: Professor of Geol ogy, Texas Tech University, Lubbock, Texas.

Roberts, F. H. H., Jr., 1942, Archaeological and geological investi gations in the San Jon district, eastern New Mexico: Smith sonian Miscellaneous Collection, v. 103, no. 4.

Rosenthal, G. A., 1982, Plant nonprotein animo and imino acids: bio logical, biochemical and toxicological properties: New York, Academic Press.

Sadava, D., Walker, F., and Chrispeels, M. J., 1973, Hydroxyproline rich cell wall protein (extensin): biosynthesis and accumulation in growing pea epicotyls: Developmental Biology, v. 30, p. 42-48.

Schoeninger, Margaret, and DeNiro, M. J., 1982, Carbon isotope ratios of apatite from fossil bone cannot be used to reconstruct diets '" of animals: Nature, v. 297, p. 577-578.

Schoeninger, M. J., and DeNiro, M. J., 1983,· Carbon isotope ratios of bone apatite and animal diet reconstruction: Nature, v. 301, p. 177-178. 157

Schoenwetter, James, 1975, Pollen-analytical results, Part I in Wendorf, F., and Hester, J., eds., Late Pleistocene environments of the southern High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 103-120.

Schumm, Stanley, 1973, Geomorphic thresholds and complex response of drainage systems, in Morisawa, Marie, ed., Fluvial geomorphology: Binghamton, New York State University Publications in Geomorphology, p. 299-310.

Schwartz, C. C., and Ellis, J. E., 1981, Feeding ecology and niche separation in some native and domestic ungulates on the short grass prairie: Journal of Applied Ecology, v. 18, p. 343-353.

Sellards, E., 1938, Artifacts associated with a fossil elephant: Geological Society of America Bulletin, v. 49, p. 999-1009.

Sellards, E., 1952, Early man in America: Austin, University of Texas Press.

Sellards, E., Evans, G. L., and Meade, G. E., 1947, Fossil bison and associated artifacts from Plainview, Texas, with description of artifacts by Alex D. Krieger: Geological Society of America Bulletin, v. 58, p. 927-954.

Sellstedt, H., Engstrand, L., and Gejvall, N.-G., 1966, New application of radiocarbon dating to collagen residue in bones: Nature, v. 212, p. 572-574.

Sims, P. 0., Singh, J. S., and Lauenroth, W. K., 1978, The struc ture and function of ten western North American grasslands: Journal of Ecology, v. 66, p. 251-285.

Smith, B. N., 1972, Natural abundance of the stable isotopes of car bon in biological systems: BioScience, v. 22, p. 226-231.

Stafford, T. W., Jr., 1977, Late Quaternary alluvial stratigraphy of Yellowhouse Draw, Lubbock, Texas, in King, M., ed., Cultural adaption to ecological change on the Llano Estacado. Progress report of the 1976 field season of the Lubbock Lake Site by E. Johnson and T. Stafford, Jr. for National Science Foundation Grant No. BNS 7612006, Lubbock, Texas Tech Univeristy" The Museum. 15

Stafford, T. W., Jr., 1978, Late Quaternary alluvial stratigraphy of Yellowhouse and Blackwater Draws, Llano Estacado, Te:x:as, in King, M., ed., Cultural adaption to ecological change on. the Llano Estacado, Progress report of the 1977 field season.. of thE Lubbock Lake Site by E. Johnson and T. Stafford ~ Jr _ for Grant BNS 7612006 AOl, Lubbock, Texas Tech University, ThE Museum.

Stafford, T. [W.], Jr., 1981, Alluvial geology and archaeological potential of the Texas southern High Plains: Atnerican Antiquity, v. 46, p. 548-565.

Stafford, T. W., Jr., Duhamel, R. C., Haynes, C. V., Jr.. and Brendel, K., 1982, Isolation of proline and hydro:x:yproIine from fossil bone: Life Sciences, v. 31, p. 931-938.

Stein, W. H., and Moore, Stanford, 1950, Chromatographic determination of the amino acid composition of protei:n..s. Cold Sp:r.ing Harbor Symposia on Quantitative Biology, v _ 14 • p. 179-190.

Stenhouse, M. D., and Baxter, M. S., 1979, The uptake of bomb 14( in humans, in Berger, R., and Suess, H., eds., Radiocarbon dating: Berkeley, University of California Press. p. 324-341.

Stetten, M. R., and Schoenheimer, Rudolf, 1944, The metabolism of L(-)-proline studied with the aid of deuteriurn and isotopic ni~rogen: Journal of Biological Chemistry, v. 153, p _ 113-132.

Stock, Chester, and Bode, F. D., 1937, The occurrence of f1i:n.. ts anc extinct animals in pluvial deposits near Clovis. NeVil' Mexico, Part III--Geology and vertebrate paleontology of the late Quaternary near Clovis, New Mexico: Proceedings of the Philadelphia Academy of Natural Sciences, v. 88 ~ p. 219-241.

Stone, B. L., and Gray, W. R., 1981, Occurence of hydro:x:yproline i a toxin from the marine snail Conus geographus: Archiv-es of Biochemistry and Biophysics, v. 216, p. 765-767.

Stowe, L. G., and Teeri, J. A., 1978, The geographic distribu.tion 0 C 4 sp~cies of the Dicotyledonae in relation to cIiInate : The American Naturalist, v. 112, p. 609-623.

Sullivan, C. H.• , and Krueger, H. W., 1981, Carbon isotope a:n..al ysis of separate chemical phases in modern and fossil bone: NaturE v. 292, p. 333-335.

Sullivan, C. H., and Krueger, H. W., 1983, Carbon isotope ratios 0; bone apatite and animal diet reconstruction: N a. ture • "V. 301, p. 177. 159

Szabo, B. J., 1980, Results and assessment of uranium-series dating of vertebrate fossils from Quaternary alluvium in Colorado. Arctic and Alpine Research, v. 12, p. 95-100.

Tamers, M. A., and Pearson, F. J., 1965, Validity of radiocarbon dates on bone: Nature, v. 208, p. 1053-1055.

Taylor, R. E., 1982, Problems in the radiocarbon dating of bone, in Currie, L. A., ed., Nuclear and chemical dating techniques: Washington, D. C., American Chemical Society Symposium Series No 176, p. 453-473.

Teeri, J. A., 1979, The climatology of the C 4 photosynthetic pathway, in Solbrig, ed., Topics in plant population biology: New York, Columbia University Press, p. 356-374.

Teeri, J. A., and Stowe, L. G., 1976, Climatic patterns and the dis tribution of C grasses in North America: Oecologia, v. 23, p. 4 1-12.

Thompson, J. L., 1977, Investigations into the modern and late Pleis tocene flora at the Lubbock Lake Site. Unpublished M. S. thesis, Texas Tech University, Lubbock.

Thompson, M. A., 1984, Personal communication: Research assistant, Tree Ring Laboratory, University of Arizona, Tucson.

Thompson,.R. C., and Ballou, J. E., 1956, Studies of metabolic turn over with tritium as a tracer : Journal of Biological Chemistry, v. 223, p. 795-809.

Thurman, E. M., and Malcolm, R. L., 1979, Concentration and frac tionation of hydrophobic orga.':lic acid constituents from natural waters by liquid chromatography: U. S. Geological Survey Water-Supply Paper No. 1817-G.

Tieszen, L. L., Boutton, T. W., Tesdahl, K. G., and Slade, N. A., 1983, Fractionation and turnover ?I3 stable carbon isotopes in animal tissues: implication for c5 C analysis of diet: Oecologia, v. 57. p. 32-37.

Tieszen, L. L., Hein, D., Qvortrup, S.1~.' Troughton, J. H., and Imbamba, S. K., 1979, Use of c5 C values to determine vege tation selectivity in East African herbivores: Oecologia, v. 37, p. 351-359.

Tieszen, L. L., Senyimba, M. M., Imbamba, S. K., and Troughton, J. H., 1979, The distribution of C and C 4 grasses and carbon isotope discrimination along an al?itudinal and moisture gradient in Kenya: Oecologia, v. 37, p. 337-350. 160

Trautman, Milton, and Walton, Alan, 1962, Isotopes, Inc. radiocarbon measurements II: Radiocarbon, v. 4, p. 35-42.

U denfriend, Sidney, 1966, Formation of hydroxyproline in collagen: Science, v. 152, p. 1335-1340. van der Merwe, N. J., 1982, Carbon isotopes, photosynthesis and archaeology: American Scientist, v. 70, p. 596-606. 13 van der Merwe, N. J., and Vogel, J. C., 1978, C content of human collagen as a measure of prehistoric diet in woodland North America: Nature, v. 276, p. 815-816.

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

Vogel, J. C., Fuls, A., and Ellis, R. P., 1978, The geographical dis tribution of Kranz grasses in South Africa: South African Journal of Science, v. 74, p. 209-215.

Vogel, J. C., and Waterbolk, H. T., 1967, Groningen radiocarbon dates VII: Radiocarbon, v. 9, p. 107-155.

Wand, J. 0, 1981, Microsample preparation for radiocarbon dating. Unpublished Ph.D. thesis, Oxford University, England.

Weber, F., Hamilton, T., Hopkins, D., Repenning, C., and Haas, H., 19tH, Canyon Creek. A late Pleistocene vertebrate locality in interior Alaska: Quaternary Research, v. 16, p. 167-180.

Wendorf, Fred, assembler, 1961, Paleoecology of the Llano Estacado: Santa Fe, New Mexico, The Museum of New Mexico Press, Fort Burgwin Research Center Publication No. 1.

Wendorf, Fred, 1970, The Lubbock subpluvial, in Dort, W., and Jones, J., eds., Pleistocene and Recent environments of the central Great Plains: Lawrence, Kansas, University of Kansas Press, p. 23-35.

Wendorf, Fred, 1975, The modern environment, in Wendorf, F., and Hester, J., eds., Late Pleistocene environments of the southern t, High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9, p. 1-12.

Wendorf, Fred, and Hester, J., eds., 1975, Late Pleistocene environ ments of the southern High Plains: Ranchos de Taos, New Mexico, Fort Burgwin Research Center Publication No.9.

Wetzel, R. G., 1975, Limnology: Philadelphia, Pennsylvania, W. B. Saunders Co. 161

Wheat, J. B., 1974, First excavations at the Lubbock Lake Site, in Black, Craig, ed., History and prehistory of the Lubbock Lake Site: Lubbock, West Texas Muserum Association, The Museum Journal XV, p. 15-42.

White, Abraham, Handler, Philip, Smith, E. L., Hill, Robert, and Lehman,!. R., 1978, Principles of biochemistry: New York, McGraw-Hill.

Zabel, T. H., Jull, A. J. T., Donahue, D. J., and Damon, P. E., 1983, Quantitative radioisotope measurement with the NSF Arizona Regional Accelerator Facility: IEEE Transactions on Nuclear Science, v. NS-30, p. 1371-1373.