Citation
Lacustrine deposition in the Jurassic Morrison formation, Purgatoire River region, southeastern Colorado

Material Information

Title:
Lacustrine deposition in the Jurassic Morrison formation, Purgatoire River region, southeastern Colorado
Creator:
Prince, Nancy Kathleen
Language:
English
Physical Description:
182 leaves : illustrations, maps, color photographs ; 28 cm

Subjects

Subjects / Keywords:
Geology -- Colorado -- Purgatoire River Region ( lcsh )
Geology ( fast )
Morrison Formation ( lcsh )
United States -- Morrison Formation ( fast )
Genre:
Academic theses. ( lcgft )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )
Academic theses ( lcgft )

Notes

Bibliography:
Includes bibliographical references (leaves 162-177).
General Note:
Inserts in pocket.
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Basic Science, Department of Geology.
Statement of Responsibility:
by Nancy Kathleen Prince.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
19818068 ( OCLC )
ocm19818068
Classification:
LD1190.L56 1988m .P735 ( lcc )

Downloads

This item has the following downloads:


Full Text
LACUSTRINE DEPOSITION IN THE JURASSIC MORRISON FORMATION, PURGATOIRE RIVER REGION, SOUTHEASTERN COLORADO
by
Nancy Kathleen Prince B. A. University of Colorado, 1982
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Basic Science Department of Geology
1988
\
I


This thesis for the Master of Basic Science degree by Nancy Kathleen Prince has been approved for the Department of Geology by
d*jls..2£Jnz.


Prince, Nancy Kathleen (M. B. S., Geology, Biology)
Lacustrine Deposition in the Jurassic Morrison Formation, Purgatoire River Region, Southeastern Colorado Thesis directed by Associate Professor Martin G. Lockley
The brightly red-colored rocks in the Purgatoire uplift of Las Animas, Otero and Bent counties in southeastern Colorado are Permian, Triassic and Jurassic in age. Although exact correlation of these largely terrestrial deposits is uncertain, the oldest Jurassic formation exposed in this area is the eolian Entrada Sandstone. The overlying unnamed sequence is possibly equivalent to the Bell Ranch Formation. It contains conglomeratic, sandy fluvial strata at the base and silty, sandy gypsiferous saline playa deposits above.
Cyclic lacustrine sediments of the lower Morrison Formation conformably overlie the Bell Ranch equivalent. Study of the sedimentology, paleoecology and geochemistry of outcrops within the study area allows identification of four stages of development of the lake basin in the late Jurassic Morrison sequence. From bottom to top these are:
Unit A: Laminated non-calcareous clayshale and siltstone were deposited below wavebase in a lower Morrison fresh-water lake. The lake became slightly saline and laminated to thin-bedded bioturbated micrites to rippled fossiliferous packstone formed.


«
iv
Unit B: Thinly bedded raicrite to packstone was deposited near the shoreline in a strongly fluctuating environment. Saline conditions and desiccation occurred periodically as did associated mudflat and palustrine deposition. During this stage, several smaller separate lakes may have formed in the basin.
Unit C: Sandy limestones and dolomites were formed as either primary or early diagenetic minerals while clastic deposition increased and the lake basin filled up.
Unit D: The upper Morrison in this region is dominated by thin crossbedded sandstones and colored mudstones of a fluvial floodplain.
Fluvial processes continued to dominate early in the Lower Cretaceous (Lytle Formation) followed by a marine transgression (Glencairn and Dakota formations).
Unlike other regions of "undifferentiated" Morrison Formation, in the study area there is good evidence for persistent lacustrine deposition in a tectonically stable basin over a long period of time. The climate was arid to semi-arid, with strongly seasonal moisture input. This lake had high calcium content and at times was sufficiently saline, and possibly alkaline, to precipitate unidentified salt crystals and
abundant ooids.


DEDICATION
This thesis is dedicated to my friends and family who share Edward Abbey’s view of "the lonesome Morrison hills, utterly lifeless piles of clay and shale and broken rock, a dismal scene" but who tolerated my perversions anyway. (Edward Abbey, Desert Solitaire, Ballantine Books, New York, 1968, p. 40, 303 pp.)


ACKNOWLEDGEMENTS
Thanks to Karl and Jodi Prince for their patience with their mother, "The Rock", and for their help over the years as field and lab assistants, typists and draftsmen. Thanks also to Phyllis Meyer and Pauline and Jim Buffington for their support and faith in me, and to Bob Evans, the rockhound.
Dr. M. G. Lockley suggested the thesis and provided mentorship, friendship and financial support throughout the project which is gratefully appreciated. The other members of my committee, Dr. W. L. Bilodeau, Dr. J. P. Kurtz and Dr. F.
Peterson were very patient in reading original copies of "notes" and giving excellent suggestions for cleaning up the manuscript.
To Karen Houck, Debbie Adelsperger, Kelly Conrad, and Bev Harrison go my thanks for assistance in the field and lab.
Charles Haddox, Tom Michalski, Peggy Morgan, Sheri Ransom and Myra Vaag provided much appreciated help with drafting and proofreading versions of this thesis.
Doug and Lori Nicodemus, Willard and Mary Ann Louden and Tiny Doherty graciously allowed me to roam their "backyards". -I am especially grateful for their hospitality and their help in solving logistical field problems.


Vll
The discussions with E. R. Magathan, A. S. Cohen, C. E. Turner-Peterson, B. H. Wilkinson, M. J. Kraus, R. T. Bakker, R. Forester, R. F. Dubiel and N. Mateer and the late John H. Hanley provided insight into lacustrine lithofacies, including paleoecology, geochemistry and sedimentology.
Assistance with petrology and log interpretation from M. W. Longman, M. D. Wilson, M. L. Henricks, E. B. Coalson, and J. W. Marchetti was most appreciated.
Discussions with G. R. Scott, M. McLaughlin, J. L. Ridgely, J. S. de Albuquerque, S. G. Lucas, A. P. Hunt, J. T. Kirkland, M. Parrish and R. B. O'Sullivan provided useful information about Triassic and Jurassic stratigraphy in southeastern Colorado and northern New Mexico.
I would also like to gratefully acknowledge grant assistance from the Sigma Xi Scientific Society and Research Assistantship funds through M. G. Lockley from the University of Colorado - Denver, and from the National Science Foundation,
Grant No. EAR 8618206.


CONTENTS
CHAPTER
I. INTRODUCTION....................................... 1
Previous Work.................................... 3
Purpose and Procedure............................ 7
II. GEOLOGIC SETTING................................... 12
Regional Structure.............................. 12
Paleozoic Stratigraphy and Structure............ 13
Mesozoic Stratigraphy and Structure............. 19
Triassic........................................ 20
Jurassic........................................ 23
Entrada Sandstone............................. 23
Bell Ranch Equivalent......................... 25
Pre-Morrison Jurassic Structure............... 37
Lower Morrison Contact and Chert Bed.......... 38
Morrison Formation............................ 39
Informal subdivisions within the Morrison.. 41
Age of Morrison Formation..................... 46
Cretaceous...................................... 49
Cenozoic
52


III. SEDIMENTOLOGY IX 54
Introduction 54
Grains 54
Bedding 60
Primary Structures 66
Secondary Structures 71
Pedogenic and Diagenetic Structures 75
Diagenesis 77
IV. PALEONTOLOGY 79
Introduction 79
Plants 80
Algae 80
Other Plants 85
Root Traces 86
Discussion 87
Invertebrates 89
Crustaceans 89
Molluscs .... 94
Invertebrate Trace Fossils 98
Vertebrates 100
Fish 100
Dinosaurs 101
Dinosaur Tracks 103
Other Morrison Vertebrates 104


V. GEOCHEMISTRY..................................... 105
Introduction................................... 105
Direct Evidence................................ 109
Indirect Evidence.............................. 113
Sedimentary Structures....................... 113
Absent Minerals.............................. 115
Biota........................................ 116
Basin Analysis............................... 117
Conclusions.................................... 123
VI. DEP0SITI0NAL ENVIRONMENTS........................ 125
Introduction................................... 125
Depositional Environments...................... 127
Sub-units of the Morrison Formation............ 135
Unit A....................................... 135
Unit B - 0V Lake............................. 139
Unit B - Dino Lake........................... 146
Unit C....................................... 148
Unit D....................................... 150
VII. REGIONAL PALEOGEOGRAPHY.......................... 152
Southeastern Colorado.......................... 152
Basin Development.............................. 157
Comparison to Other Lake Systems............... 160
REFERENCES............................................. 162
APPENDIX
178


XI
TABLES
Table
1. Environmental Parameters
for Biotic and Abiotic Features.............. 106
2. Water Chemistry
of Some Modern Carbonate Producing Lakes..... 110


FIGURES
1. Location Map......................................... 2
2. Purgatoire Uplift.................................... 9
3. Regional Structures...................;........... 14
4. Subsurface Correlation.............................. 15
5. Correlation Chart................................... 17
6. Pamena Gap Exposures................................ 18
7. Cross Section Locations............................. 22
8. Bell Ranch Equivalent Conglomerate.................. 26
9. Bell Ranch Equivalent Gypsum........................ 28
10. Chert Band.......................................... 30
11. Unit A.............................................. 42
12. Unit B.............................................. 43
13. Unit C.............................................. 44
14. Unit D.............................................. 45
15. Lower Cretaceous, 0V Mesa........................... 50
16. Fine-grained Sandstone.............................. 57
17. Medium-grained Sandstone............................ 59
18. Sandy Limestone, Unit C............................. 59
19. Volcanic Glass Shards............................... 61
20. Micrite and Grainstone.............................. 67
21. Debris Layer........................................ 69
22. Crystal Casts....................................... 72
23. Clastic Dikes....................................... 74
24. Circumgranular Cracking............................. 76


Xlll
25. Vadoids............................................ 76
26. Oncoids........................................... 83
27. Root Traces........................................ 86
28. Conchostrachan..................................... 93
29. Unionid Bivalve.................................... 95
30. Invertebrate Trace Fossils....................... 99
31. Dinosaur Bone..................................... 102
32. Brine Evolution................................... 108
33. Calcite Crystals.................................. 118
34. Chert Band........................................ 118
35. Legend............................................ 126
36. Depositional Environments......................... 128
37. Depositional Model................................ 128
38. Unit A, Goat Ranch................................ 137
39. Unit B, The Gap................................... 140
40. Topographic Locator Map........................... 144
41. 0V Mesa Correlation............................... 145
42. Unit B, Rock Crossing............................. 147
43. Unit C, The Gap................................... 149
44. Unit D, Colbert Canyon............................ 151
45a. Isopach of Units A and B.......................... 153
45b. Morrison Biota and Sediments...................... 153
46. Paleogeographic Recontruction..................... 158


xiv
PLATES
Plate
I. Regional Cross Sections, Southeastern Colorado
, Purgatoire Uplift, Southeastern Colorado
II. Cross Section


CHAPTER I
INTRODUCTION
Through participation in a research project on the most extensive Jurassic dinosaur trackway site found to date (Rock Crossing, Purgatoire River, Lockley et al., 1986), the author became involved in studying the regional framework of the Morrison Formation in southeastern Colorado. The area studied for this thesis (Figure 1) contains the best exposure of strata that brackets the trackbearing horizons and thereby permits a regional study of Morrison paleoenvironments and paleoecology.
The major goals for this report initially were to investigate the lacustrine deposits in the Morrison Formation and determine those factors which influenced the sedimentation and ecosystems of these lakes. It was anticipated that study of the sediments, flora and fauna should help determine the extent and nature of the lake ecosystem.
It soon became evident that the basic stratigraphy had not been sufficiently worked out, so the additional task of defining regional stratigraphy and pre-Morrison paleogeography was a necessary prerequisite to completing the study. It also became evident that other depositional environments were


Figure 1. Location map with Jurassic and older outcrops in southeastern Colorado, northeastern New Mexico and northwestern Oklahoma (after Scott, 1968; King and Beikman, 1974; Tweto, 1979). See Figure 2 for details of "Purgatoire Uplift" region.
C0L>UKAD0|


3
represented in the total Morrison sequence, and investigation of them became crucial in developing a complete paleogeographic reconstruction of the Morrison in southeastern Colorado.
The systematic approach adopted in this study involved 1) measuring sections and defining local Morrison stratigraphy, 2) determining the effects of regional structure, 3) integrating lacustrine paleoecology, sedimentology and geochemistry, and 4) developing a detailed paleoenvironmental and regional paleogeographic reconstruction of the Morrison Formation.
Previous Work
The Purgatoire River, Chaquaqua Creek and Plum Creek cut a maze of canyons in Las Animas County, southeastern Colorado. Legend says that a group of 16th century Spanish explorers lost their lives in these canyons with no one around to administer last rites, and thus the name "El Rio de las Animas Perdidas en Purgatorio", or The River of Lost Souls in Purgatory (Mosher, 1983). Local residents call it "the Picketwire". Some early geologists used the spelling "Purgatory", although on U.S.G.S. topographic maps the correct name is the Purgatoire River. "Chaquaqua" is the Indian word for elderberry and is pronounced sha-kwok* (W. Louden, pers. comm., 1984).
Gilbert (1896) was the first to report on the geology of the Purgatoire River canyons area in a study of the water potential in the Arkansas Valley and its associated drainage


4
basins. Reconnaissance along the Purgatoire River above Bent Canyon and along Two Buttes Creek in 1895 confirmed ideas developed with R. T. Hill when they mapped the foothills near Beulah the previous year, that the beds between the Cretaceous and the Paleozoic rocks in southern Colorado could be correlated with the "Juratrias" in northern Colorado.
Lee investigated exposures along the "Purgatory River" in 1901 with a side trip to the "Rio Cimarron" (Cimarron River in northeast New Mexico). Dinosaur bones in the shales between the Dakota sandstones and the gypsums of the "Red Beds" prompted Lee to announce the discovery of the Morrison Formation in the area. The next summer Lee (1902) returned to the region and explored Morrison Formation outcrops in the Cimarron and Canadian canyons of New Mexico, and the Apishapa canyon and Sangre de Cristo mountains areas of southeastern Colorado. His paper to the National Acadamy for the Sciences in 1903 reporting Comanchean fossils (similar to Lower Cretaceous of Texas) in the shales of the Morrison Formation prompted discussion between Darton, who supported these findings, and Stanton, who argued that lithologic similarity did not mean "stratic equivalency".
This discussion probably led to the productive 1905 field trip (Stanton, 1905; Darton, 1906) with Lee, Stanton, Darton and Gilmore. The reported Comanchean pelecypods were found in place in a marine shale separating "Dakota sands" above the Morrison shales along the Purgatoire River near Higbee. "Brontosaurus" bones were recovered from the Morrison shales, especially from


5
the southeast corner of the Timpas quadrangle (probably from the Beaty Canyon Quadrangle map of 1972). "Belodon (Phytosaur, Romer 1945) indicating Triassic age" (Stanton, 1905) "portion of a scapula" (Darton, 1906) was collected from the upper layers of the Red Beds below the mouth of Chaquaqua Creek.
Members of the The Colorado Geologic Survey investigated the economic potential of the geologic resources of Las Animas and Otero counties from 1916 to 1920. Results of this work, published several years later by Patton (1923), Duce (1924) and Toepleman (1924), recognized that the Morrison was probably Jurassic, not Cretaceous as was officially held by the U. S. Geological Survey at the time.
Scientific attention was again brought to the region when MacClary (1936, 1938) wrote about the "recently discovered Purgatory River dinosaur tracks". Serious students of tracks including Roland T. Bird were involved with a more accessible site in Texas and so no further study was done on the Purgatoire tracks for almost 50 years.
Ben H. Parker, who was familiar with the Cimarron Arch region, measured three sections in southeastern Colorado which Heaton (1939) included in his regional correlation of Jurassic deposits of Utah, Colorado and New Mexico. This report correlated the sandstone below the Morrison Formation (the Exeter Sandstone as named by Lee, 1902) with the Entrada Sandstone of
the Colorado Plateau.


6
In an attempt to understand the stratigraphy and structure of the Model helium gas field, Bass (1947) measured sections in Purgatoire canyons and correlated them to subsurface drill samples 20 miles to the west.
The 1956 Rocky Mountain Association of Geologists Guidebook to the Raton Basin included a road log to the "Canyon of the Purgatoire River" by E. R. Landis (1956) and a review of the problems of Triassic and Jurassic stratigraphy in southeastern Colorado by Oriel and Mudge (1956). Other papers (Oborne, 1956; Shaw, 1956) used sub-surface drill cuttings and well logs to demonstrate several late Paleozoic paleogeographic highs in southeastern Colorado.
Long's (1966) correlations of the Lower Cretaceous in southeastern Colorado included data from the mountains near Colorado Springs and Trinidad south and east to the Purgatoire canyons. Taylor (1974) measured sections in the Purgatoire and other nearby canyons focusing on early Cretaceous environments in an eastward extension of the work done by Long.
The La Junta Quadrangle map of Scott (1968) was the first geologic map of the region since the 1924 reports of the Colorado Geological Survey. This map indicated U. S. Geological Survey acceptance of the use in southeastern Colorado of Paleozoic terminology from the subsurface of Kansas.
The Rocky Mountain Association of Geologists published a series of cross sections, including two across the southeastern portion of the state assembled by J. Wilson (1977). These


7
subsurface sections utilized both geophysical logs and drill cuttings to emphasize the Paleozoic structures which are of major interest to the petroleum industry.
Investigation of the Purgatoire River tracksite by the
\
University of Colorado at Denver Geology Department began in 1982, and reports published to date include Conrad et al. (1987); Frazier et al. (1983); Lockley (1986a, b); Lockley et al. (1984; 1986); Prince (1983); Prince and Houck (1986). Some of these publications include data compiled for this thesis.
To the west and north of the study area lies the Pinon Canyon Manuever Site of the United States Army. Before this area was occupied by the Army in 1982, a survey of the geology and paleontology, supervised by Kauffman (1986), was completed by students at the University of Colorado in Boulder. The pre-Cretaceous stratigraphy was the responsibility of J. T. Kirkland and J. G. Eaton who found bone fragments of possible Late Jurassic age near the top of the massive red sandstones.
Purpose and Procedure
None of the previous studies in the region focused on the Morrison Formation. Some of the formation names applied here have been shown to be invalid, and others are inappropriate in light of new information. This research was therefore conducted at two scales: a detailed paleoenvironmental study of the Morrison Formation in the Purgatoire uplift (Figure 2) and a larger-scale regional stratigraphic and paleogeographic framework


8
(Figure 1). Because the original intent of the study was to understand the environmental setting of the beds containing the dinosaur tracks which are in the lower portion of the formation, and because the lower portion is better exposed in the study area, emphasis was placed on development of a lacustrine model for this portion of the section. Field data is the cornerstone of this study, and is supplemented by analysis of outcrop and subsurface samples and geophysical well logs.
Ten sections in the Purgatoire uplift were measured in detail and eleven nearby exposures were cursorily examined (Appendix). Using tape, the well-exposed lower shales and limestones were measured in .1 m intervals, although a few units were reported in centimeters. Covered intervals and thicker sandstones were measured in meters with a jacob staff and brunton compass. Each section is a composite, prepared by following a bedding plane laterally from one exposure*to the next (Chapter 6, Figure 46, 47). These exposures are sometimes separated by as much as .5 km of colluvium. Potential for errors exist because: 1) facies changes may occur in the covered intervals, 2) the stratum was covered or weathered beyond recognition, 3) the bed was removed by erosion just after deposition.


9
57 56 55 54
Figure 2. The Purgatoire uplift region is the area of concentration in this study. Jurassic and Triassic units are undifferentiated here because of the uncertainties of correlation of the older units (after Scott, 1968)


10
Over 360 samples were collected from the nine primary sections. Approximately 1 kg of material was taken from each unit identified in the well-exposed intervals. Representative samples of all other exposed lithologies were also collected.
Rock color was recorded using the GSA color chart. All samples were examined under a stereo microscope using the "Amstrat" technique, a simple, inexpensive way to identify components and textures in rock chips by breaking them down in dilute HC1 (American Stratigraphic Company, 1973). Sixty-eight thin sections were made and examined under the petrographic microscope to obtain greater detail. Sixty-one friable clay-rich mudstones and shales were processed to remove the clays and extract calcareous microfossils.
Some of the thin section and microfossil preparation and analysis was performed by UCD research team members K. Houck, S. Chesson and D. Adelsperger. Members of this team also examined vertebrate trace fossils in detail at the Rock Crossing site using standard mapping and replication techniques (Lockley, et al., 1986). The potential for palynologically-based biostratigraphy was not pursued. X-ray analyses were done on eight samples to obtain bulk carbonate and clay mineralogy.
Published measured section data was transcribed (by N. Prince) to the regional correlation scale of 1 in = 10 m to allow incorporation into the data base. Outcrops in each of the general areas of exposure were visited in order to develop a unified interpretation of the data.


11
Drill cuttings from five wells stored at American Stratigraphic Company were examined using the "Amstrat" technique. These were the only wells in the area for which both geophysical logs and cuttings were available through the Morrison Formation (Appendix, wells marked with *).
Gamma ray geophysical logs were obtained for 41 wells in the study area (Appendix). Formation tops were picked on these logs by incorporating data from scout cards at Petroleum Information Corporation with interpretations of cuttings from neighboring wells and with measured sections from nearby outcrops. Some of the logs, available only at 1 in = 250 ft, were xerographically reduced to 1 in = 100 ft. Outcrop data was replotted to the same scale to prepare the regional cross sections in Plate I.
Results of these investigations constitute two distinct data sets: (1) detailed descriptions of sedimentology and paleontology (Chapters III, IV and V) used to interpret microenvironments, and (2) cross sections and maps used to understand regional stratigraphy and paleogeography (Chapters VI and VII). Generalized stratigraphic columns (1 in = 10 m) which contain lithology and paleontology data for all measured sections are included on Figure 10. Details of selected intervals are included in Figures 41-47 (Chapter VI); a complete set of the ten primary measured sections (1 in = 1 m) is available from the
author.


CHAPTER II
GEOLOGIC SETTING Regional Structure
The Purgatoire River disects a broad, regional upwarp (Figure 1) called the Purgatoire uplift (Heaton, 1939) or the Red dome (Gabelman, 1956). Structural elements which define the limits of the uplift are the Black Hills, Muddy Creek and Mustang Creek monoclines (Duce, 1924; Scott, 1968). The dome is underlain by the late Paleozoic Apishapa uplift (Tweto, 1980), an uplifted basement block bounded on the north by the Apishapa fault, to the east by the Freezeout Creek fault, to the west (west of the study area) by the Wet Mountain fault and extending to the south into the Sierra Grande arch. The southern terminus of the Las Animas arch is just to the northeast.
For convenience in this report the areas of the Muddy Creek monocline and Black Hills monocline, where Lower Cretaceous sandstones rim canyons exposing relatively uniform Jurassic, Triassic, and Permian strata, will be referred to as the Purgatoire uplift (Figure 2). This includes exposures along the


13
Purgatoire River, Chaquaqua Creek and Plum Creek, within townships 26 to 31 south and ranges 54 to 57 west.
The Purgatoire uplift is at the junction of several well-studied geologic provinces: the Denver basin and Front Range, the Oklahoma panhandle, northeastern New Mexico and east-central Kansas. A hybrid stratigraphic terminology has been applied by various authors, but regional correlations across this area have been published only on the Jurassic Entrada (Heaton, 1939) and Lower Cretaceous Purgatoire Formations (Long, 1966).
Paleozoic Stratigraphy and Structure
Early Paleozoic strata are thin to absent across the crest of the Apishapa uplift (Figure 3), due to non-deposition or erosion during late Paleozoic time (Maher, 1953). Where these sediments are preserved, they are composed of limestone and dolomite with minor quartz sand, an indication that any exposed land mass was too low in relief to shed coarse clastic detritus (Gabelman, 1956).
The Texaco Cynthia True #1 (Figure 4) test well was spudded in the Lower Cretaceous on the eastern limb of the Black Hills monocline. Here a complete, although thin, section of Cretaceous through Permian strata that overlies the Precambrian granites. Other oil and gas exploration wells demonstrate that during the late Paleozoic, the Apishapa and Sierra Grande uplifts were one land mass. Karstification of Mississippian sediments along the Las Animas arch to the northeast, lack of' Pennsylvanian


Figure 3. Structural elements which influenced Jurassic deposition in the study area include the Late Paleozoic Apishapa and Las Animas and Mesozoic Sierra Grand uplifts. Present physiography is a result of Laramide and younger uplift of the Black Hills, Muddy Creek and Mustang Creek monoclines. (Area shown as in Figure 1.)


Permlan/rrlasslc Jurassic Cratacsous
TEXACO
15
O V MESA
(composite)
T29S-R56W
*1 Govt-Cynthia True 30-T28S-R56W
Figure 4. The gamma ray log from the Texaco #1 Cynthia True well, as interpreted through the use of drill cuttings, correlates with outcrops in the OV Mesa area and serves as a tool for integrating subsurface data into this study.


16
strata on the Apishapa uplift, and the presence of coarse conglomeratic Atokan age sediments on the northern and northeastern flanks of the Apishapa/Sierra Grande structure (DeVoto, 1980) are good evidence for uplift during this time period.
In the subsurface, Wolfcampian and Leonardian sediments are thin over the Apishapa uplift, Sierra Grande arch and Las Animas arch. Distribution of coarse elastics around these features indicate that although there was still some relief (Maher, 1945; Tweto, 1980) these highlands had largely been eroded by the end of Wolfcampian time.
The oldest sediments exposed in the Purgatoire uplift (Figure 5) and along Two Buttes Creek in Baca County were deposited in a marginal marine setting during Upper Permian (Guadalupian) to Lower Triassic time. Thirteen meters of orange-red, gypsiferous, very fine sand and silty shale of the Permian Whitehorse Formation are exposed in the Purgatoire River bottom (Figure 6). Above this the Permian Day Creek Dolomite contains seventeen meters of dark red silty mudstone interbedded with crinkly, stromatolitic, sometimes brecciated, limestone and dolomite. The twelve meters of dark red, thinly bedded, frequently rippled, very fine grained sandstone and siltstone which follow are assigned to the Permian Big Basin Formation.
These unique crinkly stromatolites and the sandy siltstones above the Whitehorse Formation are similar to and
correlative with beds in the Forelle and Strain members of the


Figure 5. Stratigraphic correlation chart of Paleozoic ana Mesozoic strata exposed in the Purgatoire uplift. The Morrison Formation, Purgatoire Formation and Dakota Sandstone correlations are undisputed. Earlier Jurassic, Triassic and Permian rocks lack datable fossils, contain numerous unconformities and are thus difficult to correlate with certainty to adjacent depositional basins.


19
Permian/Triassic Lykins Formation along the northern Colorado Front Range. Kansas and Oklahoma terminology (Whitehorse Formation, Day Creek Dolomite and Big Basin Formation) (Scott, 1968, Maughan, 1980) should be retained because subsurface correlation to the east is easily demonstrated, and because the Lykins is absent from outcrop in the mountains south of the Canon City Embayment (Tweto, 1979).
Mesozoic Stratigraphy and Structure
Throughout the study area 45 to 300 m of red-colored siltstone and sandstone are present between the distinctive Day Creek Formation and the gypsiferous Bell Ranch Formation. Correlation of these redbeds is complicated (Oriel and Mudge, 1956) because (1) the depositional environments were probably largely terrestrial resulting in similar but laterally variable lithologic suites within different time periods, (2) fossils are sparse, and for the most part do not yield diagnostic ages, (3) exposed unconformities are difficult to trace in the subsurface and (4) subsurface data is sparse and often of poor quality.
Figure 5 highlights the confusion regarding the correlation of these strata, and the text which follows presents new information on which the correlations of this report are
based.


20
Triassic
The deep red, thin-bedded fine-grained sandstones exposed at Two Buttes Reservoir and Purgatoire uplift (Figure 6) are probably most correctly correlated with the Late Triassic Dockum Group in northeastern New Mexico (Heaton, 1939) based largely on similar stratigraphic position and to a lesser degree on similar paleontologic and lithologic characteristics.
These sandstones, 42 m thick at Pamena Gap, overlie the maroon siltstones of the Big Basin Formation with a sharp, unconformable contact, but appear conformable with the eolian sandstones above. Although there is some similarity in lithology between the unit exposed in the Purgatoire uplift and the Travesser Formation of the Dockum Group in northeastern New Mexico, there is insufficient evidence to directly correlate the two units (K. Conrad, pers. comm., 1987).
Unidentified plant remains and bone fragments reported by Duce (1924, p. 81) from exposures near the mouth of Chaquaqua Creek suggest a terrestrial origin for these sediments. In northeastern New Mexico, terrestrial biota of the Late Triassic Dockum Group include plants (Baldy Hill Formation, Baldwin and Muehlberger, 1959), reptile body fossils (Sloan Canyon Formation, Stovall and Savage, 1939) and trace fossils (Sheep Pen and Sloan Canyon formations, Conrad et al., 1987).
Subsurface cross sections across the Sierra Grande and Apishapa arches (McLaughlin, 1954; Oriel and Mudge, 1956 and-


21
Figure 7 and Plate I of this report) support this correlation, but are inconclusive. All use incomplete geophysical well logs (Figure 4) and sporadic well cutting descriptions from widely spaced petroleum exploration wells, which commonly provide conflicting results. Outcrops in western Prowers County mapped as Triassic Sloan Canyon Formation by McLaughlin (1954) and Scott (1968) are here considered to be Morrison Formation because of their stratigraphic relationship to nearby Morrison Formation and Purgatoire Formation outcrops (See Plate I and discussion in Chapter VII, Regional Paleogeography).
A comprehensive regional study of the lower Mesozoic stratigraphy might resolve these suggested correlations of the red beds exposed in the Purgatoire canyons with those exposed at Two Buttes, Colorado and those in southwestern Kansas, northwestern Oklahoma, northeastern New Mexico and the central Colorado Front Range. Outcrop studies should include comparison of the biota and examination of possible marine sedimentary structures (J. Kirkland, pers. comm., 1987), although Late Triassic biostratigraphy is presently controversial (M. Parrish, pers. comm., 1988). This analysis should be compared with subsurface data, including drill cuttings and a full suite of geophysical logs to reach a resolution of the problems.


Figure 7. Location of cross sections combining outcrop and subsurface data. Plate I contains sections A-D and Plate II contains section E.
N)
to
COLbKAOO


23
A complete understanding of the structure during the Late Triassic awaits unravelling of the stratigraphy. There is evidence, however, for areas of positive relief to the west (Huerfano Park, Johnson, 1959) and the south (unconformities along the Cimarron arch, Baldwin and Muehlberger, 1959) and possible independent movement of the Sierra Grande and Apishapa arches during the Triassic (Baldwin and Muehlberger, 1959).
Jurassic
Entrada Sandstone
The thick pink to salmon fine-grained sandstone with large-scale cross bedding (Figure 6) exposed at Pamena Gap on the Purgatoire uplift is here correlated to the Entrada Sandstone in the Cimarron Valley of northeast New Mexico. At Pamena Gap the sandstones are 52 m thick, but thicknesses vary locally from 32 m to 65 m (Lee, 1901; Heaton, 1939; Taylor, 1974) in the canyons of the Purgatoire uplift. The top few meters of this sandstone often contain white pisolites (oolites of Lee, 1902), which are concentrations of calcium carbonate and iron oxide around sand grains. These are larger and more prominent on the upper surface of the sandstone, as are zones bleached white along fracture planes. Directly above this surface is a 3 m thick, mottled limestone. The mottled limestone, pisolites and leached texture might be the result of pedogenic processes acting on the dunes in an arid climate, and when combined with evidence for erosional


24
relief, indicate an unconformity of unknown duration at the close of red-bed deposition.
Heaton (1939) correlated the Exeter Sandstone (Lee,
1902), a white, thickly cross-bedded sandstone in northern New Mexico, with the type Entrada Sandstone of the Colorado Plateau (Gilluly and Reeside, 1926)._The thickly cross-bedded strata of the Purgatoire uplift is similar in lithology and stratigraphic position to the Entrada Sandstone of northern New Mexico. It is possible to correlate a sand body of varying thickness from Pamena Gap through the subsurface south to the New Mexico border (Plate I). This sandstone is assumed to be the equivalent of the lower portion of the Entrada Sandstone of the Cimarron Valley as designated by Lucas et al. (1985), not the Exeter member (see following discussion of Bell Ranch Equivalent). Sandstones equivalent to the Entrada may also be present (Plate I) to the east at Johnny Branch Creek (Heaton, 1939) and Two Buttes Reservoir (McLaughlin, 1954) and to the west in Cuchara Canyon (Heaton, 1939) and Huerfano Park (Johnson, 1959).
Taylor (1974) and Scott (1968) follow Oriel and Mudge (1956) in reporting a thin, white sandstone above the possible paleosol in the Purgatoire uplift as Entrada and include the thick, cross-bedded sandstone with the Triassic Dockum. This designation was based on subsurface correlation with the Dockum Group of northwestern Texas (G. R. Scott, pers. comm., 1987) but influenced by Stanton’s report (1905) of Triassic "Belodon" bone fragments (phytosaur, Romer 1945) collected from the upper layers



25
of the Red Beds below the mouth of Chaquaqua Creek. Kirkland (J. Kirkland, pers. comm., 1987; Kauffman, 1986) however, reported finding Late Jurassic bone fragments from a conglomerate just above the thick sandstone, four miles south of the junction of Chaquaqua Creek and Purgatoire River, probably in the same horizon from which the "Belodon" bone was collected. The UCD research group has also collected unidentified tooth and bone fragments from the conglomerate at several locations. Additional field and lab work to determine the presence of pre- or post-Entrada fossils here would be crucial in establishing the age of the cross-bedded sandstone.
Bell Ranch Equivalent
Basal Fluvial Unit. The mottled paleosol horizon at the top of the Entrada sandstone is truncated near OV Mesa by a coarse conglomerate about 1 m thick containing clasts of the pisolitic sandstone, clay clasts, chert pebbles and bone fragments. Dark red, ripple cross-laminated, very fine-grained sandstone overlies the basal conglomerate and is interbedded with other thin conglomerates (Figure 8). The total package is up to
4 m thick, and the top bed is a siltstone containing large (2 to
5 cm diameter, up to 20 cm long) tubular trace fossils, probably root casts. The unit appears to be fluvial, and cross sets measured at the Lost Canyon section indicated a paleocurrent direction of N 20 W.


27
The unit as described is persistent on the east side of the Purgatoire River and on the west side of Chaquaqua Creek around OV Mesa. The thickness and percentage of sandstone and conglomerate is variable throughout the Purgatoire uplift and is interpreted as interdune fluvial deposit. On the west side of the Purgatoire River near Red Rock Canyon it is up to 40 m thick (Kauffman, 1986), to the southeast in Trough Canyon of Chaquaqua Creek it is 6 m thick and it does not appear to be present 10 km east on the wall of Plum Canyon or north in Bravo Canyon near the confluence of Chaquaqua Creek and the Purgatoire River. It is also absent from other outcrop exposures of the Entrada Formation in southeastern Colorado listed above. A similar conglomerate is reported in the Canadian River Valley of northern New Mexico and is possibly correlative to the J-5 unconformity of Pipiringos and O'Sullivan (1976) (G. R. Scott, pers. comm., 1987).
Gypsiferous Unit. Conformably above the conglomerate and sandstone of the basal Bell Ranch Equivalent in the Purgatoire uplift the red siltstones grade upward into a 25 to 40 m thick unit of interbedded siltstone, mudstone, sandstone, clayshale, and gypsum. No fossils have been found in this interval. The siltstones and mudstones are predominantly reddish brown in the lower third, and gray throughout the upper portion. The white gypsiferous sandstone beds are up to 1 m thick with a cross-bedded appearance (Figure 9) and form resistant ridges throughout the generally non-resistant unit. In one measured



29
section (Lost Canyon) one meter thick gypsum beds are interbedded with thin, red and green mudstones and light green clayshales.
Both the fluvial conglomerate and the bedded gypsum are present near Canon City, Colorado. Lateral facies changes occur across paleovalleys in the underlying formations, and up to 45 m of gypsum has been deposited in the paleo-lows (Fredrickson et al., 1956).
Chert Band. In the Purgatoire uplift the top few meters of the gypsiferous unit contain a distinctive lithology, most correctly classified as a thinly bedded crystalline limestone.
The secondary calcite crystals are parallel to bedding in thin sandy clay-rich beds. Minor gypsum and barite nodules are present and granule-sized nodules of pink and white chert replace all other minerals. The chert nodules become larger upward to a horizon of solid blue and red chert about 20 cm thick (Figure 10). The calcite crystals, disseminated chert and gypsum nodules continue 50 cm above the solid chert. Overlying beds are thinly laminated, non-calcareous clayshales, with scattered selenite crystals and rare thin stringers of limonite, but with no bedded or nodular gypsum. This horizon is similar to "the agate bed" and "the welded chert bed" common in the lower Morrison (Ogden, 1954). The value of this unit as a stratigraphic marker is discussed below. Description and review of the mechanisms of formation are in Chapter V, Geochemistry.


31
Discussion. Other informal stratigraphic designations that could be used for this unit include: Ralston Creek Formation (Scott, 1968; Lockley et al, 1986), now considered an invalid name (Pipiringos and O'Sullivan, 1976), "Brown Silt Member" of the Morrison, and "Middle Unit". Because these names cannot be used without formally proposing new stratigraphic terminology and/or reinstating rejected terminology, they are discussed only briefly before focusing on the relative merits of the formally accepted terminology of time equivalent units.
Ralston Creek Formation. Pipiringos and O'Sullivan (1976), (using a chert pebble horizon) questioned the type locality of the Ralston Creek Formation (LeRoy, 1946; Van Horn, 1957) suggesting that the lower portion is equivalent to the lower Canyon Springs Sandstone Member of the Sundance Formation (Callovian) in southern Wyoming and that the upper nonconglomeratic part is equivalent to the Morrison Formation (Kimmeridgian). They did not attempt to ascertain the affinites of the gypsum present below the Morrison Formation at other localities along the Front Range however (R. O'Sullivan, pers. comm., 1986). Although the U. S. Geological Survey no longer recognizes the unit, de Albuquerque (1986) chose to retain the Ralston Creek Formation for the pre-Morrison coarse elastics on the eastern flank of the Wet Mountains as no other term seemed appropriate for the lithology. Stratigraphic position suggests


32
that these sediments are equivalent to the gypsiferous beds of the Purgatoire uplift.
Un-named "Middle Unit". The un-named "Middle Unit" of Oriel and Mudge (1956) included diverse lithologies in similar stratigraphic positions: the evaporites and varicolored mudstones of the Denver basin and northeast New Mexico, the arkosic conglomerates near Canon City, and sandstone, red mudstone and conglomerate in southeasternmost Colorado and western Oklahoma. Kauffman (1986) used this term to refer to the fluvial conglomerates and gypsiferous beds in the Purgatoire uplift.
"Brown Silt" Member of the Morrison. The "Brown silt" member of the Morrison (Bachman, 1953; Baldwin and Muehlberger, 1959) is a thin-bedded, light brown siltstone and very fine grained sandstone with minor thin limestone and gypsum about 3 m below the "agate bed", which at the base is interbedded with the Exeter Sandstone. It is present in north central New Mexico (Mankin, 1958), along the Rio Cimarron north of Kenton, Oklahoma and in Carizzo Creek in Baca County, Colorado. Dinosaur tracks, possibly Late Jurassic in age, have been documented from this horizon near Kenton (Lockley, 1986; Conrad et al., 1987). Stratigraphic position and lithologic similarity suggest that this unit is probably equivalent to the gypsiferous unit exposed in the Purgatoire uplift and possibly to the Bell Ranch Formation (Lucas et al., 1985, Conrad et al., 1987).


33
Wanakah Formation. The Wanakah Formation was originally named as a member of the Morrison Formation (Burbank, 1930) for shale, limestone and sandstone exposures in mines in the Ouray district of southwestern Colorado. In the San Juan basin, Condon and Peterson (1986) have elevated the Wanakah Formation from member to formation status, with the Todilto and Beclabito members. Along the Front Range, west and south of Trinidad (Wood et al., 1957; Johnson, 1959) and in the subsurface of southeastern Colorado the marls of the lower Morrison have been called the Wanakah since they are similar lithologically to the Wanakah Formation at the type locality. A fish-bearing unit north of Canon City below the Morrison Formation was assigned to the Ralston Creek Formation and later to the Wanakah Formation, based on middle to lower Callovian age of the fish (Schaeffer and Patterson, 1984).
Todilto Formation. The Todilto Formation, whether used to denote a member of the Wanakah Formation, of the Morrison Formation, of the Entrada Sandstone or a distinct formation (Gregory, 1916) defines a closed basin of deposition in northwestern New Mexico (Anderson and Kirkland, 1960) with limestone at the outer margin and gypsum above the limestone in the center of the basin. The Todilto basin was extended into east-central New Mexico, and south-central and south-eastern Colorado by Lucas et al. (1985).


34
Based on the age of abundant fish fossils, the Todilto Formation and time-equivalent facies, the Bell Ranch and Exeter formations, are thought to be early Callovian (late Middle Jurassic) by Lucas. There is currently debate about the environment of the Todilto Formation sediments: it is considered to be either an inland lake (Anderson and Kirkland, 1960), a marine embayment (Ridgley and Goldhaber, 1983) or a land locked salina with subsurface recharge from a nearby sea (Lucas et al., 1985).
Bell Ranch Formation. The type section of the Bell Ranch Formation (Griggs and Read, 1959) was proposed for the sequence between the Entrada Sandstone and Morrison Formation previously called the Wanakah Formation at Carpenter's Point, east-central New Mexico. It was further defined by Lucas et al. (1985) as a reddish, ribbed sandy/silty formation with disseminated or nodular gypsum and minor limestone. It is in sharp, possibly conformable, contact with the Entrada Sandstone and appears to interfinger to the south and east with the Exeter Member of the Entrada Sandstone. Locally it overlies the Todilto Formation suggesting it may be a landward shoreline facies of the Todilto water body. As suggested in the following discussion, the weight of evidence favors designation of the strata in the Purgatoire uplift area as being equivalent to the Bell Ranch
Formation.


35
Discussion. The J-5 unconformity is represented by an erosion surface, possibly correlative to a slight eustatic lowering of sea level (Vail et al., 1977). If the lower fluvial unit at the Purgatoire uplift marks the J-5 unconformity, then it might be possible that these gypsiferous units should be included in the Morrison, as is the Tidwell member of the Morrison on the Colorado plateau (O'Sullivan, 1984). However, this surface has not been clearly identified in the San Juan basin (Peterson and Turner-Peterson, 1987), and it would require careful analysis to confirm the J-5 correlation in southeastern Colorado. This, then, does not seem to provide a convincing solution to the stratigraphic problem.
Current feeling (J. Ridgely, pers. comm., 1985), based on preliminary geochemical observations, is that the gypsum in southeastern Colorado belongs to a different (non-marine) depositional system and possibly different time from that of the Wanakah and/or Todilto Formations and therefore should not carry the San Juan basin terminology.
At most outcrops in southeastern Colorado, northwestern Oklahoma and northeastern New Mexico examined by the author there is a brownish silty, usually gypsiferous unit between the Entrada/Exeter cross-bedded sandstone and the Morrison Formation drab shales. This unit can be traced on subsurface well logs, varying from 10 m in the south and east to 45 m in the Purgatoire uplift (Plate I).


36
The lithologies of this unit vary, but are explainable in terms of a playa-type flood plain with elastics coming off the Front Range and Cimarron arch area grading into an evaporite basin in the Purgatoire uplift area. The thinly interbedded siltstones and sandstones and evaporitic beds (especially gypsum and limestone) are common to both the unit in question in the study area and the Bell Ranch Formation of northeastern New Mexico.
If Lucas et al. (1985) are correct in that the Todilto Formation, Exeter Formation and Bell Ranch Formation are all late Callovian in age and if the gypsiferous unit in the Purgatoire uplift is equivalent to them, then the bone-bearing conglomerates below the gypsum are also Callovian (Middle Jurassic) or older. Three independent, unconfirmed lines of evidence, however, suggest that the gypsiferous unit is Late Jurassic in age: the large size of the "ornithopod" dinosaur footprints in the "Brown Silt Member" near Kenton (Conrad et al., 1987), Kirkland's suggested date for the bone fragments in the lower fluvial unit, and the possibility that the conglomerate in the fluvial unit may mark the J-5 unconformity, which is taken as the base of the Kimmeridgian (Berman et al., 1980).
The term Bell Ranch Equivalent therefore is used for the purposes of this report with the reservation that, although it can be mapped as a lithologic unit with definite upper and lower boundaries, the problem of time equivalence has yet to be completely resolved.


37
Pre-Morrison Jurassic Structure
Several lines of evidence suggest that north-south and east-west trending low relief remnant highlands may have separated southeastern Colorado from areas to the west and southwest.
Baldwin and Meuhlberger (1959) infered that the Exeter/Entrada formations were deposited on a relatively flat surface along the Colorado-New Mexico border but low relief along the ancestral Front Range (Long, 1966) might have contributed to linear isopach trends (Lucas, et al., 1985) within this eolian formation.
A coarse arkosic fluvial outwash plain in the "Ralston Creek" Formation (possibly a Bell Ranch equivalent, see Chapter 3) of the Wet Mountains (de Albequerque, 1986) further supports the idea that Precambrian rocks were still exposed to the west (Fredrickson et al., 1956). Other thickness, lithofacies and geochemical changes in Todilto and Bell Ranch equivalent formations indicate separation of depositional basins in north-central New Mexico from those in southeastern Colorado (J. Ridgely, pers. comm., 1986). In the study area, it appears that Bell Ranch equivalents are predominatly reddish-brown thinly-bedded fluvial floodplain siltstones and sandstones around the eastern and southern margins. Although these beds contain minor gypsum, the thickest units of bedded gypsum and thin, white, gypsiferous sandstones appear in the center of the region, north of the Sierra Grande uplift.


38
Isopach maps of the Entrada Sandstone, the basal fluvial unit and the gypsiferous unit of the Bell Ranch Equivalent in the Purgatoire uplift might help reconstruct local areas of relief which remained prior to the deposition of the Morrison lakes.
The thin, even-bedded nature of the uppermost Bell Ranch Equivalent and lowermost Morrison Formation sediments in the uplift suggest however, that there was little relief within the immediate area.
Lower Morrison Contact and Chert Bed
Most authors working in the Purgatoire uplift have included the bed containing red and blue chert nodules described above with the underlying gypsiferous units. Here the sandy, cherty crystalline limestone is followed immediately by laminated, gray non-calcareous clayshales. In adjacent exposures at Peacock Canyon, northeastern New Mexico, and north of Kenton in western Oklahoma, the chert is in the clayshales of the lower Morrison as reported by Baldwin and Muehlberger (1959) and Lucas et al. (1985).
At other outcrop localities in the study area, there appears to be a similar change from sand-sized grains below to clay-size grains above, as well as loss of nodular and bedded gypsum above the chert. As described above, however, the chert occurs at several horizons within the same unit (Mankin, 1958). The contact then should be taken at the change in grain size and


39
loss of gypsum although the presence of the red and blue chert may have utility in areas of poor exposure (Baldwin and Muehlberger, 1959; Taylor, 1974). If the presence of chert does represent a "time-surface marker" as suggested by Ogden (1954) this "time surface" crosses the formation boundary.
Morrison Formation
Eldridge (in Emmons, 1896) named the Morrison Formation for fresh-water marls averaging 200 feet thick between the Dakota Sandstone and the brown and pink "Trias" sandstone exposed just north of the town of Morrison, Colorado. Waldschmidt and LeRoy (1944) re-defined the Morrison Formation and described a type locality in Jefferson County, Colorado, where they subdivided the formation into six units. The upper sandstone and shale unit is equivalent to the Purgatoire Formation (Long, 1966); the five other Morrison units are: basal sandstone; gray and red shale; gray clay and limestone; gray shale and sandstone; and red shale. The re-correlation of the Ralston Creek Formation at its type locality by Pipiringos and O'Sullivan (1978) suggests that portions of the shale unit below the "basal sandstone" at the Morrison Formation type locality should also be included in the Morrison Formation.
On the Colorado Plateau the Morrison Formation was divided into four members by Craig et al., (1955): the Salt Wash, Westwater Canyon, Recapture and Brushy Basin Members. Later revisions of the stratigraphy of the San Juan Basin include other


40
local members: Tidwell, Bluff Sandstone, Junction Creek, Fiftymile, and Jackpile Sandstone Members (Condon and Peterson 1986; Peterson and Turner-Peterson, 1987). The lower contact of the Morrison Formation on the Colorado Platueau is now defined as being a widespread erosion surface (J-5) above which is the Tidwell Member, consisting of redbeds with occasional basal gypsum or gray and green mudstone. None of these members have been identified outside of the Colorado Plateau, and the Morrison Formation in these regions is therefore called "undifferentiated" (Craig et al., 1955).
In Red Rocks Canyon and Plum Canyon of the Purgatoire uplift, Lee (1901) correlated the shales between the gypsum and the overlying resistant sandstones of the Dakota with the Morrison Formation of central Colorado based on dinosaur bone fragments (Morosaurus and Diplodocus vertebrae were identified by Barnum Brown, Lee, 1902) and lithologic characteristics. Brontosaur bones were recovered later from the southeast corner of the Timpas Quadrangle (probably from Clark Hill on the 1972 Beaty Canyon Quadrangle map) (Stanton, 1905).
The gypsiferous beds were included with the Morrison Formation by Heaton (1939) and Bass (1947), but Scott (1968) and Taylor (1974) agreed that the lower contact should be the top of the resistant "chert bed". Most workers are in agreement that the contact of the Morrison Formation with the overlying Cretaceous rocks is the base of the massive white Lytle sandstone in the Purgatoire uplift area.


41
Informal Subdivisions within the Morrison
Within the Purgatoire uplift the author has informally subdivided the Morrison Formation into four units from bottom to top: Unit A, Unit B, Unit C and Unit D (Plate II). Away from the uplift, Units A and B thin and cannot be separated (Plate I), appearing as one single unit of calcareous clayshale interbedded with argillaceous limestones. Unit C thins and is absent in the Cimarron Valley but is thicker to the east at Johnny Branch (Plate I). Unit D is the thickest and almost always completely covered except for the thin sandstones and limestones. These four units are described in detail and their environmental significance discussed in Chapter VI, Depositional Environments.
These units have not been correlated with published sections from other regions but the general sequence of calcareous clayshales followed by sandstones and mudstones is interpreted as a general shallowing of a lacustrine system and has been recognized as typical at different locations within the "undifferentiated" Morrison region (Dodson et al., 1980).
Although Craig and his co-workers (Craig et al., 1955) felt that the "undifferentiated" Morrison Formation lithologies were similar to those of the Brushy Basin Member, recent work (Peterson and Turner-Peterson, 1987) suggests that the lacustrine limestone units such as in Unit A and B of this report are more similar, and perhaps more equivalent, to the Tidwell Member.


46
Age of Morrison Formation
Imlay (1980) using dates established on marine molluscs in northwestern Colorado, Utah and Wyoming, suggested that Morrison deposition probably began during middle to late Oxfordian and ended during lower Tithonian in the southern part of the western interior.
Although it is difficult to determine ages in a largely terrestrial system like the Morrison, U. S. Geological Survey personnel, especially Pipiringos, O'Sullivan and Peterson (Pipiringos and O'Sullivan, 1976; O'Sullivan, 1984; Condon and Peterson, 1986) have devised a system of unconformities (J-l through J-5) based largely on chert pebble horizons which they can identify throughout the Colorado Plateau region. The J-2 unconformity corresponds to the Bajocian/Bathonian boundary (lower Middle/middle Middle Jurassic), the J-4 to the Callovian/ Oxfordian boundary (Middle/Late Jurassic) and the J-5 to the Oxfordian/Kimmeridgian boundary (lower Late/middle Late Jurassic).
These regional unconformities have been correlated beyond the Colorado Plateau, and Pipiringos & O'Sullivan (1978) conclude that the J-3 and J-4 are not present in southeastern Colorado. This is based on the correlation drawn from Wyoming to Ralston Creek Reservoir which shows the Canyon Springs member bounded by J-5 and J-2, directly on the Permian/Triassic Lykins Formation in Jefferson County. From southeast Utah across Four Corners to northwest New Mexico the Morrison Formation is above the J-5


47
surface, and the Middle Jurassic Entrada (and Todilto member) is above the J-2 and the Triassic Chinle. The Lower Jurassic Glen Canyon Group is not present in the southern San Juan Basin (Condon and Peterson, 1986).
Ages established for the Morrison Formation independent of this system are often inconclusive and sometimes conflicting. Pollen from two localites in Wyoming was similar to that of Purbeckian beds, therefore considered uppermost Jurassic (Dodson et al, 1980). However, pollen recovered from just below the Lytle sandstone in southern Colorado was diagnosed as "Jurassic" (Long, 1966) but other pollen samples from a variety of upper Morrison sites in Colorado, New Mexico and Wyoming yielded a possible pre-Kimmeridgian age (Hotton, 1986).
Molluscs were suggested by Yen (1952) to be upper Jurassic, but more recent studies by Hanley et al. (1986) indicate that upper Morrison molluscs are probably facies controlled, and therefore not good biostratigraphic indicators.
Ostracods and charophytes studied by Peck (1956) in Wyoming were considered to be Upper Jurassic. A collection of ostracods and charophytes from the Morrison of Oklahoma, Colorado, and New Mexico was identified and curated by Dr. Wedel and Betty Kellett Nadeau during the. 1940's (Stovall, 1943; Peck and Reker, 1948). This data was never published, however, and the disposition of this collection is unknown. It is probable that this biota will, as the molluscs, be highly facies controlled and inconclusive for biostratigraphic correlations.


48
In northern China there is some interfingering of marine and non-marine Jurassic strata, allowing a more refined biostratigraphy. Chen Pei-ji (Nanjing Institute of Geology and Palaeontology) has tentatively identified a conchostrachan collected from the Goat Ranch section of the Purgatoire uplift as Nestoria. This genus is also found in the Da Hinggan Mountains north of Beijing, near the Mongolian border. Other specimens collected in the Purgatoire uplift area and from the Oklahoma panhandle have been referred to the above worker for identification in the hope that this international cooperation will shed light on these biostratigraphic problems.
Six biochron zones based on dinosaur fossils collected from reference sections near Canon City, Colorado and Como Bluff, Wyoming have been proposed (Bakker, 1986). Zones 2 and 3 seem to be missing from Canon City and Como Bluff as well as any of the Front Range and Denver Basin sites. The youngest, Zone 6, is typical of Brushy Basin faunas, and is well represented in the Stovall quarries in northern Oklahoma, which occur in the upper half of the Morrison Formation there.
Results obtained through radiometric dating are relatively consistent although they seem to indicate a younger age for the Morrison than had been established based on biostratigraphy. This apparent discrepancy might be due to "re-setting" of the radiometric clocks. Fission-track dates of Neocomian (early Cretaceous) for Dry Mesa, Colorado, and Cleveland-Lloyd, Utah quarries have been reported from zircons


49
and apatites in bentonites (Kowallis, 1986). Potassiura/argon dates from biotite crystals from a bentonite bed at the Cleveland-Lloyd quarry suggest that it is Tithonian, but those from Dinosaur National Monument are Neocomian (Bowman et al., 1986).
Cretaceous
In the Purgatoire uplift a thick cross-bedded white sandstone, usually with basal conglomerate of chert, clay pebbles and occasional petrified wood, caps the Morrison Formation mudstones and sandstones (Figure 15). This is the Lytle Sandstone member of the Purgatoire Formation. The upper portion of this sandstone is often stained tints of purple and yellow. Deep maroon mudstone is somtimes interbedded and some of the brown thin, discontinuous sandstone beds about 10 m below the white sandstone possibly should be included in the Lytle Sandstone also. Long (1966), outside the Purgatoire uplift, was able to identify the Morrison/Lytle contact in resistent beds below the sandstone in a few sections where the contact was not covered. In most sections this contact is covered, however, and Long (1966) and Taylor (1974) placed the contact at the base of the white sandstone, the major identifiable lithologic change. That convention is followed in this report.
Yellowish brown bioturbated and fossiliferous sandstone and siltstone above the massive sandstone represents the first major marine transgression in the region since the


51
Late Permian/Early Triassic (Taylor, 1974). This Glencairn Shale member of the Purgatoire Formation sometimes includes a dark gray shale at its top just below the massive brown weathering sandstones of the Dakota Sandstone. The Dakota Sandstone forms the caprock of the canyons and is the youngest formation exposed in the Purgatoire uplift.
Stanton (1905) first identified the separation of "Dakota sands" by a marine unit with Comanche fauna in the Purgatoire uplift area. Stose (1913) attached the formation name Purgatoire to these sandstones and "fireclays" in the Apishapa Quadrangle. The type section is in Purgatoire Canyon on the 1890 Mesa de Maya Quadrangle, probably one of the exposures south of 0V Mesa (approximately 37 degrees, 30 minutes north latitude). Finlay (1916) named the Lytle Sandstone Member and Glencairn Shale Member for exposures in El Paso County, Colorado.
Long (1966) concluded that the Morrison/Lytle contact is conformable with zones of unconformity present where Lytle channels have incised into Morrison shales and mudstones. He recommends returning to Waage's (1953) useage of Purgatoire Formation and Dakota Sandstone in south central and southeast Colorado instead of the Kansas terminology (Kiowa and Cheyenne members) of Sanders (1934), Stovall (1938) and McLaughlin (1954).
Upper Cretacesous Graneros Shale, Greenhorn Limestone, Carlile Shale and Niobrara Formation are exposed in the cliffs south and west of the Purgatoire uplift and constitute most of the high plains surface from the Purgatoire uplift west to the


52
Rocky Mountains (Scott, 1968). The thick late Cretaceous strata of the Pierre, Laramie and Fox Hills formations so evident in the Denver basin was removed by erosion from the Purgatoire uplift largely prior to the Pliocene.
Cenozoic
In the southern portion of the study area the Niobrara Formation is capped by the Pliocene Ogalalla Formation caliches and Miocene and Pliocene Mesa de Maya basalts (Tweto, 1979). Surface sediments dip slightly, and become younger away from the central portion of the Purgatoire uplift, where the Ogalalla is missing, possibly due to erosion (McLaughlin, 1954).
Within the canyons themselves, entrenched meanders are seen along the Purgatoire in the area of the Black Hills monocline. Above the stream terraces, the shale slopes show major slumping which probably occurred during early Pleistocene to early Holocene time (Colton, et al., 1975).
In the Black Hills monocline, strata as old as Permian are exposed, but in the Muddy Creek monocline the oldest exposed rocks are Jurassic in age. Terraces with boulder size gravels, caliche and other associated soils appear in the Black Hills monocline. These differences support the possibility that at least two basement blocks, such as documented in the Northern Denver Basin (Sonnenberg and Weimer, 1981) with different tectonic histories, have been active. If the caliche in the Black Hills is Ogallala in age, much of the movement may have


53
occurred along that structure prior to the Pliocene, possibly coincident with movement of the Wet Mountain and Apishapa uplifts during the Eocene documented by Tweto (1980) and Gabelman (1956). Some of the present physiography however, is probably due to post-Ogallala movement of both the Black Hills and Muddy Creek monoclines.
Evidence for the latter structural movement is present in the Muddy Creek area along the Purgatoire River where numerous small anticlines with relief of less than one meter are present. These structures disrupt Jurassic formations, but not the surrounding Quaternary alluvium. Earthquakes reported by residents of Branson, along the base of Mesa de Maya, within the past 20 years (W. Louden, pers. comm., 1984) suggest that this area is still tectonically active.


CHAPTER III
SEDIMENTOLOGY
Introduction
Siliciclastic and non-skeletal carbonate grains, bedforms and non-biogenic sedimentary structures in the Morrison Formation are described in this chapter as are structures of probable pedogenic or diagentic origin. Grains and sedimentary structures clearly of biologic origin will be discussed in Chapter IV, Paleontology and chemically precipitated sediments are discussed in Chapter V, Geochemistry.
Grains
Coated Grains
Coated grains are abundant in the study area and may be described by Peyrt's (1983) categories: ooids (concentrically laminated, chemically precipitated in a phreatic environment), vadoids (thin, irregular to concentric laminated, chemically precipitated in a vadose environment) and oncoids (irregularly laminated, encrusted by green and blue-green algae and bacteria). Ooids and possible vadoids will be discussed in this chapter;


55
oncoids will be covered in Chapter IV, Paleontology. This division assumes that biologic influence in the formation of ooids is less important to mineralogy and internal structure than is water chemistry and physical setting, and that the reverse is true in the development of those coated grains classified here as oncoids.
Ooids
Ooids are found at several horizons within Units A and B. Although some have nuclei of quartz grains, most appear to have micrite pellets in the center. SEM examination is required to determine cortical structure and mineralogy, but the presence of a psuedo-uniaxial cross when viewed with crossed polars is taken to suggest tangential structure and aragonitic mineralogy (Medwedeff and Wilkinson, 1983; Halley, 1985). Radial structure may produce a faint cross, and may represent either calcite or aragonite mineralogy. Ooids from the Goat Ranch and The Gap measured sections (Figure 2, Plate II) appear to be radial in nature, but ooids from the Rock Crossing section display pseudo-uniaxial crosses (Lockley et al., 1986, Figure 6) and are therefore tangential in nature.
None of the Morrison Formation ooids were examined under SEM, but X-ray analysis confirms that those from Rock Crossing are now calcite although these may have been precipitated as aragonite. The probable radial structure and lack of extensive


56
recrystallization of most of the ooids however, suggest original calcite mineralogy, magnesium component unknown.
Pellets
Rounded micrite pellets commonly occur in some of the wackestones. Areas of distinct pellets grade into regions with micrite texture, the result of compaction prior to lithification and/or recrystallization. These pellets are possibly of fecal origin. Other oblong grains common in the algal grainstones may also be of biologic origin.
Pelloids and Lithoclasts
Large (up to several cm in long dimension) current oriented flat micrite chips are found in many of the thin conglomerates of Unit B, often associated with oncoid fragments and a variety of bioclastic debris. Rounded clay pebbles are also characteristic of these conglomerates. Both the chips and the pebbles probably represent eroded and transported fragments of mud-cracked polygons similar to those of the Green River Formation (Eugster and Hardie, 1975) although wind deflation of dry flats might have helped concentrate these grains and transport them as at Amboseli (Behrensmeyer and Boaz, 1981)
Siliciclastic Grains
Most sandy horizons in Units A and B range from a few millimeters to a few centimeters thick (Figure 16). These calcareous, well-sorted, very fine-grained, angular sub-arkosic


58
sandstones often contain ostracode and chara debris and occasional larger, rounded frosted quartz grains. These laminae are interbedded with micrite or wackestone. Both potassium feldspars and plagioclase are present, as is biotite and a variety of unidentified opaque and heavy minerals. The silt- to very fine-grained quartz and feldspar fraction is ubiqitous; very few micrites contain less than 10% randomly scattered siliciclastic grains. A few thin lenses of very well sorted, rounded, medium-grained sandstone are present (Figure 17), as is local poorly sorted, subround to subangular, fine- to coarse-grained conglomerate.
The grain composition suggests that the source for most of the fine-grained sediment was acid igneous rocks, although exposed siliciclastic and carbonate sedimentary rocks appear to have sourced some of the coarser conglomerate (Lewis, 1984).
Unit C is typified by sandy limestone (Figure 13, 18) and dolomite, units with sparry carbonate matrix and moderately sorted, angular, fine- to medium-grained quartz sand. These appear in places to grade laterally into sandy mudstone, also containing abundant quartz silt and sand.
Although Unit D is volumetrically dominated by colored mudstone and shale, the thin sandstone beds are the most commonly exposed. They are moderately sorted, subround to subangular medium-grained clay cemented sub-arkose. Chert grains and rock fragments are common, and the beds are often stained brown by the weathering of iron minerals (mica and hornblende seem prevalent).


60
Clay-rich Sediments
Clay mineralogy is discussed in Chapter V, Geochemistry, because of the interaction between clays and ambient waters. The clay-sized sediments containing less than 50% calcium carbonate and displaying laminations are called clayshales in this report (Potter et al., 1980). The term claystone is applied to those massive clay-rich sediments lacking fissility or bedding.
Volcanic Glass Shards
In Unit C, a few small grains with the three-cornered appearance of volcanic glass shards (Figure 19) have been identified. These are silica, and their close association with smectitic clays lend credence to their volcanic origin. Lower in the section similar shaped grains are composed of sparry calcite, and are probably bioclastic fragments rather than calcitized shards.
Bedding
Bedding Style
Laminated to thin-bedded. Units A and B are composed of interbedded thinly laminated to thickly bedded, fine- to coarse-grained beds, most with ripple and flaser morphologies (see below). Although some units show disruption of bedding by crystal casts, burrowing or other animal activity, the bedding is


62
usually preserved. Possible explanations for this phenomenon include:
1) Deposition below wave base in a stratified (chemically and/or thermally) lake where unsuitable bottom conditions retard reworking by biota (Van Houten, 1964; Hakanson and Jansson, 1983).
2) Deposition in a shallow, relatively quiet environment where wave action is baffled by surrounding landforms or plants (Cohen and Thouin, 1987).
3) Deposition below wave base in an unstratified lake where alternations are due to fluctuations in sediment input from rivers such as the Omo River into Lake Turkana in Africa (Yuretich, 1979; Cohen et al., 1986) This mechanism may be active where graded laminations similar to those forming in deep water in Lake Turkana are found.
4) A high sedimentation rate which does not allow enough time for plants or animals to rework the sediment (Cohen, 1984).
5) Dominance of epifaunal organisms (Cohen, 1984).
6) Growth of stromatolic blue-green algae in a low energy environment (Monty, 1976).
7) Bioturbation during one season when the lake was higher or lower, followed by reworking by wave action before preservation (Laporte and Behrensmeyer, 1980).
8) Deposition in ponded (standing) water on mudflats where sediment-charged sheetwash rapidly decelerates and quickly deposits load. Graded thin beds or thick lamina result from


63
deposition of the waning turbid underflow or settle out from mixed inflow into the lake (Hardie et al., 1978).
9) Suspended load fluvial deposition produces this type of bedding in the final stages.
10) Highly productive, well-oxygenated lakes may be depleted in oxygen on the lake bottom (Olsen, 1980a; Cohen, 1984) therefore restricting infauna.
These laminations do not contain highly organic or reduced layers, and are not the thin, regular laminations typical of stratified lakes; therefore, the first model is one of the least likely processes. The others were probably active at different times or places during deposition of the lacustrine limestones.
Cross-bedded. A few thin (less than .5 m), laterally very discontinuous (visible only for a few meters) tabular cross-bedded sand bodies are interbedded with the lacustrine marls of Unit B. Some of these have a quartz and clay pebble lag and fine upward into medium-grained sandstone; others are uniform poorly sorted medium- to fine-grained sandstone. These sands contain abundant oncoid and pelloid debris and are often capped with a clay drape containing abundant fish scale and plant fragment debris. Small channels on the surface of the dry lake bed probably were the sites of these deposits.
The sandy limestone beds of Unit C show relict cross-bedding partly obscured by bioturbation and diagenesis.
*4


64
The tabular planar sets do not display the flat bottomed, convex upward shape typical of lacustrine bars (Peterson and Turner-Peterson, 1987). Lacustrine beach or bar origin is suggested, however, by the abundant well-washed and well-sorted siliciclastic grains not quite in grain to grain contact. Rare ostracode shells and abundant bioturbation also help support the lacustrine interpretation over the other possible environments of alluvial fan or floodplain limestone or fluvial tufa. A change in the groundwater levels and chemistries, with some eolian volcanic input possibly account for dolomitization and silicification of the original calcite matrix.
Several sandstone bodies in Unit B and Unit D which have low-angle cross bedding are laterally continuous for at least 20 m. One exposed at Lost Canyon (Plate II) has the appearance of a small delta (Jackson, 1979; Reineck and Singh, 1980), Figure 173. Another, noted at The Gap measured section (Plate II), appears to grade laterally into a tabular cross-bedded sandstone and may represent a shallow ephemeral stream (Reineck and Singh, 1980). This particular type of sand body is typified by very fine grained quartz with extremely thin laminae of dark colored heavy minerals and weak current lineations on the upper surface. Scoyenia and Planolites trails are also associated with these sheet sands.
Unit D sandstones are thinner (1 m) at the base, and become thicker (3-5 m) near the contact with the Lytle Sandstone. They are largely trough cross-bedded, sometimes tabular planar


65
cross-bedded (Figure 14) and often have a basal lag of chert and clay pebbles. These sandstones can be traced laterally for long distances, and probably represent times when braided stream deposition (low discharge to sediment load ratio) was the rule.
In the few good exposures of Unit D bright red and green mudstone containing isolated discontinuous tabular cross-bedded sandstone beds is present. During most of the later part of Morrison deposition, meandering, ephemeral streams on an oxidizing fluvial floodplain was the most common depositional environment, although suspended load streams could also have produced some of these sediments. It is this facies which could be examined in greater detail for vertebrate bones.
Because of the evidence for the influence of wind generated waves on carbonate precipitation (Lockley et al.,
1986), it is possible that there was also some eolian influence on distribution of carbonate and siliciclastic grains (Glennie, 1970). The frosted quartz grains common in some Unit B packstone lenses may have an eolian origin, but it is also possible that they were derived from older eolian sediments. In this report therefore, no eolian origin has been proposed for any of the sandstones, but this possiblity should not be totally excluded.


66
Primary Structures
Lenticular Ripple Bedding and Flaser Bedding
Lenticular bedding, slightly connected or isolated lenses (Figure 20) of grainy material distributed in a muddy matrix, is common in Unit B. Flaser bedding, muddy lenses in a grainy matrix is observed less frequently. These bedding styles, although common on marine tidal flats, may be produced in any environment with alternating current energy and sediment supply. For example, lake bottom sediments in front of small deltas (Coleman, 1966) often are characterized by lenticular laminated sediments. On playa flats wind waves rework flood deposits to produce sand lenses and muddy drapes or simple surface rippling (Hardie et al. 1978, p. 21). It should be noted that in these Morrison structures the muddy sediments are micrites and calcareous shales and the sand lenses contain abundant bioclasts.
Ripple Marks
Several bedding planes, especially subaerially exposed carbonate packstones, contain large wave ripples. Measurements of amplitude, wavelengths and grain sizes of 41 fields of ripples were taken at the Rock Crossing site (Lockley et al., 1986). Wave-fetch distance estimates of over 100 km were derived, using the method of Tanner (1971, 1974). Although the apparent storm-generated nature of these beds make it difficult to substantiate that fetch distance, the fetch was probably greater than 40 km, a size predicted for most Morrison lakes


68
(Dodson et al., 1980). This study was not attempted for Black Hills sections because bedding plane exposures of ripple marks were limited. Visual observation of a few at The Gap (Figure 21) and Round Corral indicate similar ripple dimensions and therefore similar wave fetch distances for this area.
Debris Layers
Pebble conglomerates with abundant oncoids, flat micrite chips, quartz grains and fish debris cap many subaerial exposure surfaces in Unit B (Figure 21). Some are grainstones, some packstones; some are cross-bedded, others rippled; some fine upward, some are imbricate, some have clasts randomly distributed throughout the matrix. A scenario of "catastrophic expansion, gradual contraction" (Hardie et al., 1978) where a storm lifts mud clasts from the mudflat during a storm and deposits them with thin silt and clay drapes as water level and energy lowers, may explain some of these layers. If water level stayed constant for some time after the storm then wave activity could wash and sort the clasts. Wind deflation could also concentrate clasts on the dry lake or mudflat, and this debris could then be deposited in shallow ephemeral channels prior to renewed influx of water (Behrensmeyer and Boaz, 1981; Glennie, 1970).


70
Admixed Sediment
Near the top of Unit B bedding is not often preserved. Amorphous "pockets" of grains are found in fine-grained sediment, or vice versa, "pockets" of micrite or claystone in wackestone and packstone textures. Possible explanations include the following: 1) Bioturbation by large vertebrates (Laporte and Behrensmeyer, 1980), in this case dinosaurs, can be demonstrated at Rock Crossing (Lockley et al., 1986) where a random admixture of sediment results. Most admixtures in the Black Hills area have this random pattern, and although three dimensional exposures are not available to identify tracks and trampling, "dinoturbation" (Lockley and Conrad, 1987) of exposed or shallow water deposits are probably responsible for much of this texture. 2) Flooding may dump heavy coarse sand on top of soupy mud, producing a variety of convoluted bed forms and load structures (Reineck and Singh, 1980). This type of load structure is visible at Rock Crossing in one unit, but it appears to result, for the most part, in discrete packages of one type of sediment encased in the other, and is probably not the dominant operative process. 3) Some of the large clay pockets have distinct boundaries along one surface, and are gradational into micrite along another. These might have been semi-lithified mud, dislodged and transported as a clast only a short distance. 4) Loading and differential sinking of sediment into soft muddy lake bottom by earthquake shock may produce admixed sediment (Collinson, 1977b). This is difficult to demonstrate in the


71
ancient record, and is only listed as a possible alternative or additional explanation for some of these structures.
Secondary Structures
Salt Casts and Molds
Crystal casts, both large rhombohedral- and small tabular-shaped, are disruptive features in a few beds in Unit B throughout the study area. Some are along bedding planes between sandstone and shale (Lockley et al., 1986 Figure 6), and others are randomly distributed within micritic limestone (Figure 22). Although crystal growth appears to have disrupted bedding somewhat, primary despositional structures are still very apparent.
Playa lakes and associated mudflats are often characterized by bedded salts or highly disrupted bedding due to formation of authigenic crystals (Eugster et al., 1978).
Although salts frequently return to solution during the next rainy season, the right conditions may result in their preservation. These crystals appear similar to those reported from the lacustrine Lockatong Formation of the Triassic Newark rift basin of the eastern U. S. (Van Houten, 1965), but crystals are less abundant and sediment less disrupted in the Morrison, possibly because the Morrison waters were not as saline (see Chapter IV, Geochemistry), and desiccation not as complete as in the Triassic example.


73
Mudcracks
Small (several centimeters on a side) polygonal structures occur throughout the sequence in the Purgatoire uplift. Most often the cracks disrupt interbedded micrite and packstone, and are filled with wackestone. Thin mud drapes above rippled oncoidal conglomerates (Figure 21) and algal mats are often mudcracked. These probably represent a period of desiccation (lowstand) following a period of higher water level (Hardie et al., 1978).
The very tiny, millimeter sized mudcracks in the clay-rich calcareous shales of Unit A may have been caused by subaqueous synaresis (Collinson, 1978b) and are not necessarily evidence of subaerial exposure.
Clastic Dikes
Larger polygonal dikes have been observed in plan view at Rock Crossing (Prince and Lockley, in press) which are wackestone/packstone filled, 2-3 cm across, 5-10 cm in relief and form polygons up to 2 m on a side. Similar features have been observed in cross section at several horizons in sections measured at The Gap and 0V Mesa (Figure 23). These polygons are reminiscent of those formed in temporary desert lake basins in Libya (Oomkens, 1966) which form when thoroughly saturated sediments undergo extended desiccation (R. Dubiel, pers. comm., 1987). Glennie (1970) notes that often only the upper layers dry and crack, and that when the natives remove


75
this salt crust, the sediment beneath it is wet and salt free, suitable for growing crops.
Pedogenic and Diagenetic Structures
Mottling and Clotted Textures
Mottling, clotted textures and circumgranular cracking (Figure 24) in some of the micrites suggest a palustrine environment (Freytet, 1973; Brown and Wilkinson, 1981; Esteban and Klappa, 1983). These features are formed by the action of roots and other soil organisms on the wet, marshy soil.
Nodules, Geodes
Nodular textures are observed in some of the micrites, as are geodes. These are possible rhizoliths and as such, suggest soil-forming processes or a palustrine environment (Klappa, 1980; Cohen, 1982). This nodular texture is especially well-exposed in Unit D at the Louden Ranch where the generally vertically oriented nodules sometimes coalesce to form a continuous horizontal limestone bed.
Vadoids
In a few beds coated grains with irregular, laminated coatings are found in micrites with clotted textures, and are therefore interpreted as vadoids. Other horizons contain grains with clay halos (pseudo ooids) (Figure 25) developed during soil formation (Brewer, 1964).


77
Colored Mudstone
The maroon and green-colored mudstone of Units C and D may represent diagenetic alteration of hematite grains (Walker, 1967), but because of associated pedogenic nodules, probably represent paleosols (Kraus et al., 1985). This mudstone is poorly exposed, and no attempt was made to identify horizons, to name the soils or to attempt a climatic interpretation.
Diagenesis
The most common diagenetic minerals are sparry calcite, sparry quartz and barite. Minor dolomite, pyrite, limonite and hematite are present in some nodules and as alteration of pre-existing minerals. Some of the present clay mineralogy is probably the result of diagenetic processes, but the determination of the extent of clay diagenesis is beyond the scope of this project.
The paragenetic sequence seems to be 1) sparry calcite,
2) chert, 3) clear silica, 4) barite. Where pyrite is present, it accompanies the precipitation of barite. Dolomite is found as the final stage mineral in small geodes at the upstream Rock Crossing section (Plate II), where calcite and clear euhedral silica are first and second. Pyrite accompanies the formation of
the dolomite rhombs.


78
A unique mineral, but one characteristic of many early Morrison deposits in the western U. S., is flesh-colored acicular barite. A possible intraformational source for the barium would be barium substituted for potassium in feldspars and illitic clays. Barium could also have been introduced from barite deposits in the upper Cretaceous marine sediments. Barite is the last mineral in the paragenetic sequence, and is often seen filling both horizontal and vertical fractures, some of which are truncated by a Tertiary erosion surface, suggesting movement during the late stages of tectonism. The strontium carbonate, celestite, was reported by Duce (1924). These two minerals have very similar optical properties and so are easily misidentified (Shelley, 1985). The one sample examined with X-ray techniques for this project did not indicate the presence of any strontium.


CHAPTER IV
PALEONTOLOGY
Introduction
The fossils found in the Morrison Formation in the Purgatoire uplift are useful environmental indicators even though they are neither abundant or diverse. Based on functional morphology, taxonomic relationships and sediment relationships, inferences may be made about the chemistry, temperature, paleotopography and climate cyclicity. These inferences must be weighed carefully however: 1) complete uniformitarianism is not possible due to lack of relationship to extant genera; 2) organisms may be absent if they have not been introduced into that system (Beadle, 1981); 3) taphonomic effects must be
evaluated.
Presence/absence and visual estimates of relative abundance were reported on representative body fossil samples because the scope of the project and outcrop characteristics prevented detailed quantitative analysis. Thorough analysis of absolute abundance, density and dispersion especially on the


80
ostracodes and charophytes from selected outcrops could be useful in understanding the microenvironments identified in the system.
Evidence for species interaction in predator/prey ratios, nutrient flow and accidental death (the clam killer story) for the Rock Crossing site are discussed in detail elsewhere (Lockley et al., 1986; Lockley, 1986a). Some discussion of primary production is included here.
This chapter includes a despcription of the flora and fuana observed and a discussion of the paleoecology (geochemistry, energy, depth, faunal associations etc.) of each species as determined from literature to use as a basis for developing an environmental model.
Plants
Algae
Charophytes. Chara oogonia (Figure 24) are abundant components of many of the wackestones and grainstones from Unit A and Unit B in the Purgatoire uplift. Four or five species of chara appear to be represented. Alistochara was identified, but identification of the other forms is difficult, pending completion of the Treatise on plants (R. Forester, pers. com., 1988). Since the external ornament and whorls are well preserved on the oogonia, and stem fragments have been found in a few samples, these plants were probably not transported far prior to deposition.


81
Usually about 3 cm tall, these bushy plants live totally submerged in the photic zone (Ott, 1958), most commonly in water 1 to 8 m in depth (Tappan, 1980). They tolerate a wide range of pH (5.2 to 9.8 although 7 to 8 is characteristic), and alkaline, well oxygenated waters are favored. Although the stems are not always calcified (Burne et al., 1980), chara thrives, and precipitates more calcite, in lakes supersaturated with calcium carbonate. Salinity is only partially a control; Lamprothamnium chara are present today in highly saline lakes (up to 70%o) in Australia, but are usually less diverse in higher salinities (Burne et al., 1980). These chara are suited to the seasonal changes in water level; they regenerate after desiccation (Burne et al., 1980). Jurassic Alistochara and modern Lamprothamnium are in the same family, and therefore may have some similar environmental requirements or tolerances.
Chara are abundant in the Morrison Formation and other species have been catalogued from numerous other localities by Peck (1937, 1956, 1957), Waldschmidt and Leroy (1944), Johnson (1954) and others (Ott, 1958).
Oncolites. Micro-oncoids (less than 2 mm in diameter, Peryt, 1983) make up much of the grainstone and packstone of Unit B. In thin section irregular bands can be seen, often coating mud clasts or quartz grain centers (Figure 20); this irregular surface resembles algal and bacterial growth, not abiotic growth through accretion. Some of the grains described as micro-oncoids


82
may include "reworked carbonate" nodules (cf Dodson et al., 1980a), but probably are fragments or crumbs of larger oncoids such as are found in modern Lake Constance, Germany (Schottle and Muller, 1968).
Larger oncoids up to 3 cm across (Figure 26) are locally abundant in distinctive conglomerates and grainstones that are useful marker horizons. These large oncoids are disk-shaped with no apparent centers and appear to have formed on a mud substrate with little rounding or disturbance during growth. Lack of fossil centers, as are common in the Green River Formation (Weiss, 1969), suggests that little shell material was available to be incorporated into the oncoids, further evidence of the lack of a diverse invertebrate fauna (see below).
Since they are formed by blue green algae, oncoids grow within the photic zone, generally at or near river mouths. Here algal growth is favored and the periodic turbulence of the soft mud bottom disrupts growth, creating the characteristic disk shapes. In Lake Constance, a temperate marl lake, oncoids grow embedded in loose rippled sediment below mean water level (Shafer and Stapf, 1977) and are most common and largest near the Rhine River inflow. Smooth hard ones pile up like pebbles on shore, forming pebble conglomerate. Oncoids are generated by many of the same organisms that generate stromatolites, and therefore have many of the same requirements, including calcium carbonate saturation of the water, but form where there is sufficient wave action to overturn and round them (Eggleston and Dean, 1976).


84
Stromatolites. Thin, .1 to .3m thick, slightly undulating, cryptalgal laminated horizons are common in Unit B in the Purgatoire uplift. Stromatolitic structures form in many environments, but are common in intertidal and supertidal zones. Mucilagenous blue-green algal filaments calcify as they grow, forming micritic layers which alternate with layers of pellet packstone/wackestone deposited during severe storms. A change in micro-environmental geochemical conditions may also cause calcite precipitation. The filaments of algae bind available clasts such as ooids and bioclasts with this calcite (Monty, 1976) in micro-laminated sediments. The resulting mats may be preserved if drying occurs as a result of lowered water level in a tidal flat environment (Brock, 1976) or if the sedimentation rate increases rapidly in an ephemeral lake (DeDeckker, 1983).
The range of pH tolerance for blue-green stromatolitic algae is 4 to 14, with optimum values of 7 to 8.5. These algae can tolerate low oxygen levels (Tappan, 1980) and high salinities (DeDeckker, 1983) and can survive periods of desiccation.
Study of growth forms in Shark Bay, on the coast of Australia indicate that low lying stromatolites develop in lower energy environments than do strongly undulating, columnar ones (Hoffman, 1976). In the hardwater lakes of New York, however, form appears also to be related to lake size, depth, slope, and
calcium carbonate saturation of the water.


85
Thrombolitic fabrics (microscopic clots, Monty, 1976) are characteristic of certain modern lacustrine stromatolites, including those from Green Lake, New York; Mono Lake, California; and Lake Tanganyika, Africa but have not been recognized in Morrison Formation sediments.
Other Plants
Equiseteum. Impressions of thin, unbranched, ribbed stems with nodes several centimeters apart are found only at the Rock Crossing site, in a single bed interpreted as a marsh (Lockley et al., 1986). Related to modern horsetails Equiseteum, these plants had a horizontal rhizome as the main support, and thin roots branching off of it (Bold, 1967) indicating a source of groundwater near the surface.
Petrified wood. Silicified conifer logs with "typical Morrison preservation" (Tidwell, 1987) are common in upper Morrison sandstones at Two Buttes, Colorado. Thin sections were made from this material, but the cell structure was not preserved well enough to allow identification of the plant species or their age. Petrified wood has also been reported in the Purgatoire uplift (Taylor, 1974), but none was recovered by this author.
Carbonaceous fragments. Small (less than 1 mm long), black carbonaceous fragments present on bedding planes of some mudstones are interpreted as plant debris. A few conglomerates have yielded larger (1 cm diameter) pieces of possible plant


86
material but no identifications have been attempted due to the fragmented condition of the specimens. This lithology could be an excellent source of pollen for biostratigraphic studies.
Root Traces
Two types of lithified root casts have been identified in the study area (Figure 27). At Rock Crossing a horizontal root network spreads out through a clay drape just above footprint bed C, indicating a high water table, possibly marshy conditions at that time.
Figure 27. Vertical root traces indicate a lower water table, horizontal root traces indicate a near surface water table, and coalescing nodules were probably casts of larger roots.


87
More common are thin, 1 mm diameter and up to 10 cm long, downwardly branching tubules filled with sparry calcite and barite. In the micrite of footprint bed B and packstone of footprint bed C at Rock Crossing (Lockley t al., 1986) 145 holes were counted in a square 20 cm by 20 cm (3625 per square meter). 8900 per square meter were found in cryptalgal laminites at The Gap. The size, density and downward branching suggest that all were made by a small plant with a simple root system reaching for a somewhat deeper water table.
Silicified carbonate nodules at several horizons within Unit B and D are possible rhizocretions, casts of the roots of larger plants. Limited exposures and samples of these nodules prevent positive identification however.
Discussion
Plant fossils are traditionally considered scarce in the Morrison Formation either because of original low productivity, oxidation of plant material (Dodson, et al., 1980a) or some combination of both factors. Factors which limit plant growth include the availability of sunlight, water and nutrients. Productivity might also have been controlled by groundwater chemistry when Units A and B were being deposited. Plants are generally less tolerant of alkaline than saline (Beadle, 1981) water chemistries in many African lakes today. Excess salinity is also a limiting factor at modern Lake Amboceli where the rising water table in the dry lake basin brings salt with it, killing the trees (Western and Van Praet, 1973).


88
It is more likely, given the evidence for large numbers of herbiverous dinosaurs in the area of the Purgatoire uplift during the time of deposition of Unit A and Unit B that there was abundant plant life in the vicinity. Eolian deposits are also lacking in this part of the section, a less reliable indicator that plants were available to stabilize exposed loose topsoil.
Oxidizing conditions which prevent preservation of plant material might have been caused by either a hot, arid climate (Dodson et al., 1980b), the lack of thick stands of high-growing plants around the marshes (Beadle, 1981) or by well drained swamps (Coleman, 1966). It has also been suggested that the dinosaurs ate all of the plants before they could be preserved. This explanation could have credence on a local level, as for example, at Lake Amboceli, where the elephants consume all available foliage in an area during a drought (Western and Van Praet, 1973) when hunger kills more animals than thirst.
Significant Morrison flora from several sites in the western U.S. include conifer trees and shrubs, shrubby cycads, shrubby and herbaceous ferns, herbaceous lycopsids and sphenopsids (Tidwell, 1975; McAlester, 1977; Seward, 1933; Dunbar and Waage, 1969). Although there is some speculation that primitive angiosperms, probably trees or shrubs, had begun to develop (McAlester, 1977) there is no fossil evidence to support this. It is likely that shrubby ferns and cycadedoids provided the major groundcover (Dunbar and Waage, 1969), although one might speculate that some hardy, drought resistant plants,


89
possibly similar to gnetophytes might have also been available to stabilize loose sediment.
Well developed domal stromatolites form a prominant ledge in Cimarron River Valley of Oklahoma and New Mexico (West, 1978; Neuhauser et al, 1987) but are not present in the Purgatoire uplift. Leaf and stem impressions have been recovered from siltstones just above the stromatolites. The presence of crocodilian remains and tracks (Conrad et al., 1987) in associated clastic sediments suggests that this environment was very different from that exposed in the Purgatoire uplift, possibly with a strong fluvial influence.
Invertebrates
Crustaceans
Ostracodes. Ostracodes are the most abundant and diverse fauna (Figure 21, 27) in the sections measured in the Purgatoire uplift. Modern non-marine ostracodes are generally detritus filter feeders that swim several centimeters above a relatively firm substrate in low to moderately turbulent water, although some may live interstially within the sediment and groundwater aquifers. Fresh to brackish water forms generally have smooth shells which can be classified by life mode. "Swimmers" are usually high relative to length, and are able to tolerate seasonal temperature and salinity variations. "Groundwater"


90
forms live interstitially or in small ponds and littoral zones of lakes and are more likely to be small and robust (Benson, 1961).
Different genera within a species may have different chemical tolerances, but most cyprids and darwinulids are freshwater forms (< 3 ppm salinity) and most cythrids tolerate more brackish water (3 to 30 ppm salinity) (Van Morkhoven, 1962). Many ostracodes can live in relatively stagnant water with low oxygen levels but prefer to lay their eggs in oxygenated water. Darwinula and Theriosynoecum brood their eggs, but other species produce eggs which can survive desiccation up to 20 years (Brasier, 1980).
Selected ostracodes from this project were examined by R. M. Forester, U.S.G.S., a specialist in Cenozoic non-marine ostracodes. He described the specimens in terms of possible environments based on comparison of morphology and taxonomy with modern forms. In samples from Goat Ranch, The Gap and Rock Crossing sections (Plate II) Forester identified fresh to slightly brackish forms Darwinula (a small groundwater form), a small groundwater form (Candona type), a large swimming cyprid (Midlocypris type), an unknown small swimming form, and a brackish water form Theriosynoecum (a larger cythrid, groundwater form),
A modern assemblage like this might be found in slightly saline waters with more Ca+2, and less HC03-, dominated by Na+ or Mg+2; S04-2 or Cl- ions (Forester, 1986), such as modern Australian lakes. Here ostracode diversity is greater in the


91
fresher lakes and less in the large closed basins with extensive internal drainage area (DeDeckker, 1983). This assemblage could also tolerate water which was at times slightly alkaline (enriched in HC03-), which might be more similar to African lakes, where oxygen content also seems to be a limiting factor (Cohen, 1986).
Most of the ostracodes have well-preserved external ornament and are frequently preserved as articulated, spar-filled individuals. These whole carapaces are possible indications of high sedimentation rate which covered the animals before the valves could be opened by weakening of muscles or by predators. Beds with disarticulated valves may have been deposited in times of low sedimentation rate (Brasier, 1980).
Thin ostracode coquinites are present in Round Corral and The Gap sections (Plate II). It is probable that they are winnowed concentrations of shells, such as are common just below wave base in large African lakes today (Cohen, in press). These could also represent "blooms", or times of high productivity in the lake, or less likely, anoxic conditions could help preserve the abundant ostracodes (Murphy and Wilkinson, 1980).
A large collection of Morrison ostracodes collected by Wedel and Hallet-Nadeau during 1930's and 1940's from northern Oklahoma, New Mexico and southern Colorado was reported by Peck (Peck and Reker, 1948). The author has been unable to locate the disposition of this collection, which reportedly included some 70 species from the western U.S. and would be a valuable


92
contribution to our knowledge of the Morrison formation in the area.
Conchostrachans. The genera Nestoria was identified by Chen Pei-ji of Nanjing, China, from photographs (Figure 28) of specimens collected at Goat Ranch. Original material has been forwarded to him from Rock Crossing and The Gap, and the State Line section in northwestern Oklahoma.
These "clam shrimp" have chitinous shells and occupy a variety of environments including large playa lakes and marshes, and more rarely, littoral regions of lakes and small permanent ponds. They prosper in waters with 7 to 9.7 pH, with a thermal tolerance of 4 to 30 degrees Centigrade. Their eggs must pass through a dry state of torpidity and the juveniles come out of diapause when moisture, photoperiod, oxygen, alkalinity and salt concentration are optimal. Some specimens have been known to survive this desiccatation up to 15 years. Conchostrachans have a total life span between 2 weeks and 4 months. These slow-moving invertebrates have no defense against predators (insect larvae and fish) so favor habitats where they will be sheltered, or the predators absent (Tasch and Zimmerman, 1961; Webb, 1979). Some modern conchostrachans prefer clear ponds, while others prefer alkaline, muddy water. Because they are often found in ephemeral environments, conchostrachans seem to be adapted for living in conditions of fluctuating salinity and alkalinity (Pennack, 1978).


94
Preservation of the chitinous carapaces of the conchostrachans indicates that these delicate shells were not abraded. Although lack of abrasion does not necessarily mean lack of transport, these microfossils are usually found in relative abundance within a given bed, and it is assumed that the majority of these animals lived near where their shells were buried.
Molluscs
Bivalves and gastropods apparently have very limited distribution within the study area, having only been collected from a few horizons at Rock Crossing. It is possible, but not likely, that the lack of molluscs in other measured sections in the study area is an artifact of the exposures or oversight by the researcher. Over 500 samples have been taken from a dozen sections during this study and each has been examined in detail. Also, none of the abundant oncoids or ooids have shell fragment centers common where shell material is available, suggesting that there truly was some limiting factor(s) responsible for the limited distribution of the molluscs.
Bivalves. Bivalves have been found in only two beds exposed at Rock Crossing. A few silicified disarticulated unionid valves have been collected from the basal portion of footprint bed 1 (Lockley et al., 1986). Articulated, spar-filled unionids (Figure 29) are found in place, associated with dinosaur tracks and plant


96
stem impressions (Footprint bed 1, Lockley et al., 1986), apparently trampled to death in a marshy mudflat by the large vertebrates.
All modern species of the genus Unio are found in fresh water. Each species has different requirements, but most prefer clean, oxygenated, shallow (2 to 7 m), calcium-rich (20 ppm Ca) (Burky, 1983) water with pH greater than 7 (5.6 to 8.3 total range). They live infaunally in rivers and lakes with stable substrate, good food source and seasonally warm temperatures (Hanley, 1976), and require permanent water habitats with current for dispersal of larvae, and fish to host the glochidial larvae stage. Although permanent habitat is required for reproduction, a few North American unionid species are reported to have remained alive in moist mud for several months (Pennack, 1978).
Bivalves are reported (Stose, 1913) from limestones along the Huerfano River. These outcrops are 90 km from the Purgatoire uplift, and appear to be in a different depositional basin based on available subsurface data (see Chapter VII, Regional Paleogeography).
Gastropods. Several specimens of small gastropods of the genus Lymnaea, Gyraulus and Amplovalvata were recovered from the calcareous shales between footprint beds 2 and 3 at Rock Crossing (Lockley et al., 1986; Plate II). Spar filled voids with shapes reminiscent of larger gastropods were noted in the top bed exposed at Rock Crossing a few meters from the dinosaur bone


97
discussed below. Other small spar-filled voids were noted in thin sections from the southern sections, but none were distinctive enough to be called gastropods with confidence.
Modern pulmonates such as Lymnaea and Gyraulus live in water less than 3 m deep on rooted, emergent vegetation, (Pennak, 1978). Lymnaeids can live in waters with reduced oxygen concentration by breathing at the air-water interface, or breathing oxygen directly from the water through their skins (Hanley, 1976).
Valvata is the closest living relative of the gill bearing prosobranch Amplovalvata found in the study area.
Although different species have vaying requirements, many Valvata prefer alkaline waters with a pH of 7.3 to 8.4 and carbonate concentrations of 80 to 250 ppm (Harman, 1974). Quiet water with aquatic vegetation (Hanley, 1976) provides the most suitable environment. Their amphibious strategy may have evolved to compensate for periodically lower oxygen (Aldridge, 1983).
Gastropods are sensitive to the total ionic composition of the ambient water, and salinity tolerances change as other ions are added. Many can live in higher salinities if the water is calcium rich, so it is difficult to predict chemistry based on the presence of a certain species. They are restricted somewhat in highly alkaline water, but four species live today in Lake Tanganyika (Cohen and Thouin, 1987) and eight species in Lake Turkana (Cohen, 1984). The Turkana species are often thin-shelled, however, because the water is deficient in calcium.


98
Valvata scabrita was reported (Stose, 1913) from limestones 30 m below the Cretaceous sandstones along the Huerfano River near the boundary between Pueblo and Huerfano Counties. These outcrops are 90 km from the Purgatoire uplift, and appear to be in a different depositional basin based on available subsurface data (see Chapter VII, Regional Paleogeography).
Invertebrate Trace Fossils
Thin sandstones of Units C and D locally contain abundant invertebrate burrows. Several of the more distinct traces were identified by Dr. Kent Chamberlain (consulting geologist) as Skolithos, Scoyena and Planolites (Figure 30). These wandering .1 to .5 cm diameter trails occur both horizontally in thin mud drapes and vertically from the top of fine-grained, well-sorted planar cross-bedded sandstones in Unit D.
Numerous random burrows .5 to 1 cm in diameter riddle the tops of the sandy limestones of Unit C. Lacking the distinct structure for taxonomic identification, they appear to have been made by invertebrates mining thin clay pockets and clay drapes in the limestone.
Random burrows 1 mm in diameter and several centimeters long appear infrequently in the more friable, laminated calcareous clayshales of Units A and B. These burrows are often filled with calcite spar and medium-sized quartz and bioclastic


99
grains. They are most common in beds with abundant ostracodes suggesting small crustaceans as possible trace makers.
Figure 30. Invertebrate trace fossils identified in the thin sandstones of Unit C and Unit D include Skolithus, Scoyenia, and Planolites. Tiny traces and large, unidentifiable burrows are common in the sandy limestones of Unit C.


100
Vertebrates
Fish
Fish fragments from the study area have been tentatively identified by Gerald R. Smith and Ralph Stearley of the University of Michigan Museum of Paleontology. Material collected from Rock Crossing includes scales from a new species of semionotoid (related to to modern gar, Lepidosteus) and an unidentifiable amioid (related to the modern bowfin, Amia).
Only scales, external spines and isolated vertebrae and cephalic bones have been found from the fish. Lack of other material appears to be preservational bias. The frequent association of abundant scales with storm surfaces indicates the fish probably lived adjacent to, but not along with, the micro-organisms.
Semiontids probably were slow swimmers, feeding on plants and molluscs on the bottom sediments (Norman, 1963). Modern gar can tolerate brackish water and both groups are potential air breathers with divided swim bladders; they often surface to gulp air (Norman, 1963). The male Bow-fin of North America makes a nest in the swampy ends of lakes near abundant plants then guards the eggs and newborn young (Norman, 1963).
It should be noted that although Callovian fish localities are relatively abundant in the Western U. S. in the Todilto and Wanakah Formations, fish remains are rare in the Morrison Formation. Fish are reported from Como Ridge, Wyoming,


101
in possible lacustrine and fluvial facies and from Fruita in southwestern Colorado with other aquatic vertebrates (Dodson et al., 1980; Schaeffer and Patterson, 1984). Lack of proper preservational environment is probably more a factor than lack of fish, at least in the Purgatoire uplift, since scales are very abundant at some horizons.
Dinosaurs
Sauropod hind limb and pes bones were found in a micritic wackestone at Rock Crossing (Figure 31). Although there may have been some post-mortem transport/scattering of the animal remains, these large bones were fractured in-place after burial, suggesting trampling (Beherensmeyer and Boaz, 1980; Western, 1980). The presence of abundant dinosaur tracks in the bed (Lockley, et al., 1986) supports this interpretation.
Abundant small, unidentifiable fragments are found in micrite and sandstone near the top of Unit B and in Unit D at various localities near The Gap. The micrites on Bresden Trail also contain rounded quartz pebbles and cobbles with clay coatings and other suggestions of modification by pedogenic processes. Bone scrap-bearing conglomerate and sandstone are usually thin, cross-bedded and laterally discontinuous; they appear to be small channels on the dry lake bed.


103
Miscellaneous bone fragments have been reported in the area by ranchers (Taylor, 1974). Morosaurus, Diplodocus and brontosaur bones were collected at the turn of the century (Stanton, 1905) from the Purgatoire uplift. Bone quarries have been active at Canon City and Cimarron County, Oklahoma, and have produced a wide variety of dinosaur fossils. Dinosaurs and their paleoecology from these and other Morrison quarries throughout the western U. S. are discussed in detail by Dodson et al., (1980b).
Dinosaur Tracks
Over 1300 tracks forming 100 trackways have been mapped in four footprint-bearing beds at the Rock Crossing site (Lockley et al., 1986, Figure 3, 7, 8, and 9). These probably represent sauropods, ornithopods (camptosaur and/or iguanodontid) and large and small theropods. These tracks and trackways and paleoecologic interpretation based on them are discussed in detail in several publications of the UCD dinosaur research team (Lockley et al., 1986; Lockley, 1986a, 1986b).
Relatively fresh waters in tropical Africa have more large vertebrate trails than do alkaline waters (Laporte and Behrensmeyer, 1980). The abundance of tracks in several layers suggest that the dinosaurs returned repeatedly to the lake at the Rock Crossing site, which was fresh enough then to attract the large vertebrates.


104
Other Morrison Vertebrates
It should be noted that other Morrison bone localities have yielded more diverse fauna in limited numbers including crocodiles, turtles, and small mammal bones. These are most often recovered from the fluvial floodplain lithofacies A and B of Dodson et al. (1980b). These lithofacies correspond somewhat to facies of Unit D. This unit is very poorly exposed in the study area and was not thoroughly investigated in this research.
The oxidized mudstones (cf. Dodson et al., 1980b) appear to be well exposed in the Johnny Branch drainage near the Bent -Las Animas county line. Thorough investigation of this outcrop and others in the Smith Canyon and Johnny Branch drainages might increase knowledge of Morrison biota in southeastern Colorado.


CHAPTER V
GEOCHEMISTRY
Introduction
As indicated in the preceeding chapter, there is good evidence for the lacustrine nature of the lower Morrison Formation exposed in the Purgatoire Uplift. Determining the geochemistry of the system is not a simple process however. The predominance of limestone and presence of ooids and salt casts suggest deposition in other than a freshwater lake, but diagenesis has altered the calcite and other minerals, obscuring the original mineralogy and evidence concerning the ionic composition of the lake waters. Indirectly, the possible requirements of the biota and characteristics of the basin help suggest what the lake waters might have been like, but are in no way conclusive (Table 1).
Modern lakes are classified geochemically as fresh, hardwater and saline and/or alkaline. Salinity is defined, based on biological tolerance, as greater than 5 ppm total dissolved solids (Beadle, 1981). Alkalinity (or carbonate alkalinity) is defined as the sum of all titratable weak acids, which is assumed to be [HC03-] + [C03—] in meq/1 in freshwaters with a pH of


pH Eh Oxygen
Feature
(sedimentary structure/
Salinity Alkalinity Brine Type
Permanent/ Depth Ephemeral
Wave Action
Ooids slight slight high Mg/Ca permanent shallow strong
Glauberite crystal casts high low Na(Ca)S04Cl ephemeral?
Gypsum crystal casts moderate low MgNaCaS04Cl ephemeral?
Chara high tol. slight "high" Ca ephemeral rr 0 *-* 1 3 8light-moderate 7-8
Oncoids high tol. Ca cement shallow moderate 7-8
Stromatolites high tol. Ca cement ephemeral rr O I-* slight 7-8
Ostracodes moderate not likely CaNaMgS04Cl
Conchostrachans ephemeral 7-9 low
Bivalves Ca 20 ppm permanent 2-7 m slight-moderate 6-8
Gastropods slight tol. ‘slight Ca 80-250ppm 3 m 7-8 low tol.
Fish slight tol. permanent shallow low tol.
Missing pyrite oxidizing
Missing zeolites low
TABLE 1. Environmental Parameters for Biotic and Abiotic Features.
(Tol. indicates a tolerance of given paramenter. Discussion and references in text)


107
6.5 to 8.5. Several measures of carbonate hardness are used in the literature, including the European designation as the weight of CaC03 precipitate derived from carbonates in solution, the same value used by these workers to express alkalinity. Other hardness values are determined by measuring the amount of CaO or CaC03 per mg or per mole per liter (Kelts and Hsu, 1978).
Critical range of calcium concentration for most fauna appears to be .25 to .50 meq/1 or 5 to 10 mg/1 (Beadle, 1981), although many groups may thrive at higher concentrations, for example, 47 mg/1 in hardwater Lake Constance, Germany.
Saline/alkaline waters are further classified according to the dominant anions and cations (Eugster and Hardie, 1978).
The major cations are Na+, Ca+2, Mg+2, and K+, the major anions Cl—, S04-2 and HC03-. Alkaline lakes are those in which HC03- is the dominant anion, and are usually lower in Ca+2 and Mg+2 concentrations, and may be high in Na+, Cl- and S04-2. As evaporation occurs and concentrations increase, precipitation occurs along three major pathways (Figure 32), depending on the original composition of the water. As ions are removed from the system by precipitation, the remaining ions are effectively concentrated, moving along the pathway until the next precipitation branchpoint is reached (Eugster and Jones, 1979).


Full Text

PAGE 1

LACUSTRINE DEPOSITION IN THE JURASSIC MORRISON FORMATION, PURGATOIRE RIVER REGION, SOUTHEASTERN COLORADO by Nancy Kathleen Prince B. A. University of Colorado, 1982 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Basic Science Department of Geology 1988

PAGE 2

This thesis for the Master of Basic Science degree by Nancy Kathleen Prince has been approved for the Department of Geology by

PAGE 3

Prince, Nancy Kathleen (M. B. S., Geology, Biology) Lacustrine Deposition in the Jurassic Morrison Formation, Purgatoire River Region, Southeastern Colorado Thesis directed by Associate Professor Martin G. Lockley The brightly red-colored rocks in the Purgatoire uplift of Las Animas, Otero and Bent counties in southeastern Colorado are Permian, Triassic and Jurassic in age. Although exact correlation of these largely terrestrial deposits is uncertain, the oldest Jurassic formation exposed in this area is the eolian Entrada Sandstone. The overlying unnamed sequence is possibly equivalent to the Bell Ranch Formation. It contains conglomeratic, sandy fluvial strata at the base and silty, sandy gypsiferous saline playa deposits above. Cyclic sediments of the lower Morrison Formation conformably overlie the Bell Ranch equivalent. Study of the sedimentology, paleoecology and geochemistry of outcrops within the study area allows identification of four stages of development of the lake basin in the late Jurassic Morrison sequence. From bottom to top these are: Unit A: Laminated non-calcareous clayshale and siltstone were deposited below wavebase in a lower Morrison fresh-water lake. The lake became slightly saline and laminated to thin-bedded bioturbated micrites to rippled fossiliferous packstone formed.

PAGE 4

iv Unit B: Thinly bedded micrite to packstone was deposited near the shoreline in a strongly fluctuating environment. Saline conditions and desiccation occurred periodically as did associated mudflat and palustrine deposition. During this stage, several smaller separate lakes may have formed in the basin. Unit C: Sandy limestones and dolomites were formed as either primary or early diagenetic minerals while clastic deposition increased and the lake basin filled up. Unit D: The upper Morrison in this region is dominated by thin crossbedded sandstones and colored mudstones of a fluvial floodplain. Fluvial processes continued to dominate early in the Lower Cretaceous (Lytle Formation) followed by a marine transgression (Glencairn and Dakota formations). Unlike other regions of "undifferentiated" Morrison Formation, in the study area there is good evidence for persistent lacustrine deposition in a tectonically stable basin over a long period of time. The climate was arid to semi-arid, with strongly seasonal moisture input. This lake had high calcium content and at times was sufficiently saline, and possibly alkaline, to precipitate unidentified salt crystals and abundant ooids.

PAGE 5

DEDICATION This thesis is dedicated to my friends and family who share Edward Abbey's view of "the lonesome Morrison hills, utterly lifeless piles of clay and shale and broken rock, a dismal scene" but who tolerated my perversions anyway. (Edward Abbey, Desert Solitaire, Ballantine Books, New York, 1968, p. 40, 303 pp.)

PAGE 6

ACKNOWLEDGEMENTS Thanks to Karl and Jodi Prince for their patience with their mother, "The Rock", and for their help over the years as field and lab assistants, typists and draftsmen. Thanks also to Phyllis Meyer and Pauline and Jim Buffington for their support and faith in me, and to Bob Evans, the rockhound. Dr. M. G. Lockley suggested the thesis and provided mentorship, friendship and financial support throughout the project which is gratefully appreciated. The other members of my committee, Dr. W. L. Bilodeau, Dr. J. P. Kurtz and Dr. F. Peterson were very patient in reading original copies of "notes" and giving excellent suggestions for cleaning up the manuscript. To Karen Houck, Debbie Adelsperger, Kelly Conrad, and Bev Harrison go my thanks for assistance in the field and lab. Charles Haddox, Tom Michalski, Peggy Morgan, Sheri Ransom and Myra Vaag provided much appreciated help with drafting and proofreading versions of this thesis. Doug and Lori Nicodemus, Willard and Mary Ann Louden and Tiny Doherty graciously allowed me to roam their "backyards". I am especially grateful for their hospitality and their help in solving logistical field problems.

PAGE 7

vii The discussions with E. R. Magathan, A. S. Cohen, C. E. Turner-Peterson, B. H. Wilkinson, M. J. Kraus, R. T. Bakker, R. Forester, R. F. Dubiel and N. Mateer and the late John H. Hanley provided insight into lacustrine lithofacies, including paleoecology, geochemistry and sedimentology. Assistance with petrology and log interpretation from M. W. Longman, M. D. Wilson, M. L. Henricks, E. B. Coalson, and J. W. Marchetti was most appreciated. Discussions with G. R. Scott, M. McLaughlin, J. L. Ridgely, J. S. de Albuquerque, S. G. Lucas, A. P. Hunt, J. T. Kirkland, M. Parrish and R. B. O'Sullivan provided useful information about Triassic and Jurassic stratigraphy in southeastern Colorado and northern New Mexico. I would also like to gratefully acknowledge grant assistance from the Sigma Xi Scientific Society and Research Assistantship funds through M. G. Lockley from the University of Colorado Denver, and from the National Science Foundation, Grant No. EAR 8618206.

PAGE 8

CHAPTER I. II. CONTENTS INTRODUCTION ••••••••••••• ...................... 1 Previous Work................................ 3 Purpose and Procedure........................ 7 GEOLOGIC SETTING ••••••••••••••.••••.••••••••••. 12 Regional Structure........................... 12 Paleozoic Stratigraphy and Structure......... 13 Mesozoic Stratigraphy and Structure.......... 19 Triassic..................................... 20 Jurassic..................................... 23 Entrada Sandstone.......................... 23 Bell Ranch Equivalent...................... 25 Pre-Morrison Jurassic Structure............ 37 Lower Morrison Contact and Chert Bed....... 38 Morrison Formation......................... 39 Informal subdivisions within the Morrison.. 41 Age of Morrison Formation.................. 46 Cretaceous................................... 49 cenozoic. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

PAGE 9

ix III. SEDIMENTOLOGY •••••••••••••••••••••••••••••••••• 54 Introduction ••• . .......................... . 54 Grains •••• . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 Bedding ••••••• . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Primary.Structures ••• . . . . . . . . . . . . . . . . . . . . . . . . 66 Secondary Structures ••••••••••••••••••••.•••• 71 Pedogenic and Diagenetic Structures •••• 75 Diagenesis ••• . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 IV. PALEONTOWGY ••••••••••••••••••••••••••••••••••• 79 Introduction ••••••••••••.•••••••••••••••••••• 79 Plants ...................................... . 80 Algae ••••••••• . ................... . 80 Other Plants ••••••••••••••••••••••••••••••• 85 Root Traces •••••••••••••••••••••••••••••••• 86 Discussion ••••••••••••••••••••••••••••••••• 87 Invertebrates ••••••••••.••••••••.•••••••••••• 89 Crustaceans •••••••••••••••••••••••••••••••• 89 Molluscs .............•..................... 94 Invertebrate Trace Fossils ••••••••••••••••• 98 Vertebrates •................................. 100 Fish ...................................... . 100 Dinosaurs •••••••••••••••••••••••••••••••••• 101 Dinosaur Tracks .....•.•••••.••••••••.•.••.• 103 Other Morrison Vertebrates ••••••••••••••••••• 104

PAGE 10

v. VI. VII. GEOCIIEMISTRY ..•..•..•.•.•.....••.•••...•..•••.• In trod uc tion ................................ . Direct Evidence ••.•••.••.•.•.••••••••....••.• Indirect Evidence ........................... . Sedimentary Structures. Absent Minerals •••.•••••..•• B iota ....................... . Basin Analysis. Conclusions ................................. . DEPOSITIONAL ENVIRONMENTS ••.•.•••••.••••••••••• Introduction ......•••.••••.•••••••.••••••.•.. Depositional Environments ••.•.•••••...••.•... Sub-units of the Morrison Formation •.•••••••• Unit A ..•.................................. Unit B OV Lake .......•.••••.•.•.....•.... Unit B -Dina Lake ........................ . Unit C .................................... . Unit D .................................... . REGIONAL PALEOGEOGRAPHY •••••••••••••••.•••••••. Southeastern Colorado .•.••••.•••.•.•.•••.•..• Basin Development .•.....••••..•.•• Comparison to Other Lake Systems •••••••••.•.• REFERENCES •••••••.•••..•••••.••••••••••••••••••••••••• 105 105 109 113 113 115 116 117 123 125 125 127 135 135 139 146 148 150 152 152 157 160 162 APPENDIX ................................................ 178 X

PAGE 11

Table 1. TABLES Environmental Parameters for Biotic and Abiotic Features ••.••••••••••• 2 . Water Chemistry xi 106 of Some Modern Carbonate Producing Lakes ••.•• 110

PAGE 12

1. 2 . 3 . 4 . 5 . 6 . 7. 8 . 9 . 10. 11. 1 2 . 13. 14. 15 . 16 . 17. 18 . 19. 20. 21. 22. 23. 24. FIGURES Location Map ...••• ..................... . .......... Purgatoire Uplift ..• Regional Structures •••••• Subsurface Correlation •..••.•.•......•••••.••••..• Correlation Chart ................................ . Pamena Gap Exposures .•••• Cross Section Loca tions .......................... . B e l l R a nch Equivalent C o nglomerate .• Bell Ranch Equivalent Gypsum ••••.•••.••••••.•••••• Chert Band . ................................... . . . Unit A .......................................... . Unit B .......................................... . Unit C .......... ................................ . Unit D .......................................... . Lower Cretaceous, OV Mesa ••••...••••••••••.•••.•• Fine-grained Sandstone ... M e d i u m -grain ed Sa ndston e . Sandy Limestone, Unit C ..•..•.•••..••••.•••••.••. Volcanic Glass Shards .•.•••.•......•.•.•••••.•••. Micrite an d Grainstone .......................... . Debris Layer •. Crystal Casts ................................... . Clastic Dikes ................................... . Circumgranular Cracking •.••••.••••.•...•.•.•••.•• 2 9 14 15 17 18 22 26 28 30 4 2 43 44 45 50 57 59 59 61 67 6 9 72 7 4 76

PAGE 13

25. 26. 27. 28. 29. 30 . 31. 32 . 33. 34 . 35 . 36 . 37 . 38 . 39. 40 . 41. 42 . 43 . 44 . V adoid s ......................................... . Oncoids .......... ............... . ............... . Root Traces ............................... ...... . Conchostrachan ...•......... . •••..•••••••.••••...• Unionid Bivalve ................................. . Invertebrate Trace Fossils •.••...• Dinosau r B one ••.••.••..•.•••.•••..••.•• Brine Evolution •........•...•...••..•••...••••••• Calcite Crystals. Chert Band ...................................... . Legend ........•.......... . Depositional Environments ..••..•••••.••..••.••••• Depositional Model .•..•••••••...•••.•••.•••••.• •• Unit A , Goat Ranch .............................. . Unit B , T h e Gap ••.•••••••. T opog raphic Locator Map ••• OV Mesa Correlatio n ••.••••••••.••.••.••.••••••••• Unit B , Unit C , Unit D , Rock Crossing. The Gap •••••••. Colbert Canyon • . . • 45a . Isopach of Units A and B ••••••••.•••••••••••••••. 45b. Morrison Biota and Sediments .•..••••••••••••••••• 46 . Paleogeographic Recontruction •••.•••••••.••••.••• 76 83 86 93 95 99 102 108 118 118 126 128 128 137 1 40 144 145 147 149 151 153 153 1 58 xiii

PAGE 14

xiv PLATES Plate I. Regional Cross Sections, Southeastern Colorado II. Cross Section, Purgatoire Uplift, Southeastern Colorado

PAGE 15

CHAPTER I INTRODUCTION Through participation in a research project on the most extensive Jurassic dinosaur trackway site found to date (Rock Crossing, Purgatoire River, Lockley et al., 1986), the author became involved in studying the regional framework of the Morrison Formation in southeastern Colorado. The area studied for this thesis (Figure 1) contains the best exposure of strata that brackets the trackbearing horizons and thereby permits a regional study of Morrison paleoenvironments and paleoecology. The major goals for this report initially were to investigate the lacustrine deposits in the Morrison Formation and determine those factors which influenced the sedimentation and ecosystems of these lakes. It was anticipated that study of the sediments, flora and fauna should help determine the extent and nature of the lake ecosystem. It soon became evident that the basic stratigraphy had not sufficiently worked out, so the additional task of defining regional stratigraphy and pre-Morrison paleogeography was a necessary prerequisite to completing the study. It also became evident that other depositional environments were

PAGE 16

T " "' • FJ JURASSIC & h.] TRIASSIC []PERMIAN G"-J JURASSIC TRACKWAYS r .. T .. •• . I ' " . .!-----,' ' ' •. ,. ' :... ,. ' : • • • r Figure 1 . Location map with Jurassic and older outcrops in southeastern Colorado, northeastern New Mexi c o and northwestern Oklahoma (after Scott, 1968 ; King and Bei kman, 1974; Tweto, 1979). See Figure 2 for details of "Purgatoire Uplift" region. • T ' . N

PAGE 17

3 represented in the total Morrison sequence, and investigation of them became crucial in developing a complete paleogeographic reconstruction of the Morrison in southeastern Colorado. -The systematic approach adopted in this study involved 1) measuring sections and defining local Morrison stratigraphy, 2) determining the effects of regional structure, 3) integrating lacustrine paleoecology, sedimentology and geochemistry, and 4) developing a detailed paleoenvironmental and regional paleogeographic reconstruction of the Morrison Formation. Previous Work The Purgatoire River, Chaquaqua Creek and Plum Creek cut a maze of canyons in Las Animas County, southeastern Colorado. Legend says that a group of 16th century Spanish explorers lost their lives in these canyons with no one around to administer last rites, and thus the name "El Rio de las Animas Perdidas en Purgatorio", or The River of Lost Souls in Purgatory (Mosher, 1983). Local residents call it "the Picketwire". Some early geologists used the spelling "Purgatory", although on U.S.G.S. topographic maps the correct name is the Purgatoire River. "Chaquaqua" is the Indian word for elderberry and is pronounced sha-kwok' (W. Louden, pers. comm., 1984). Gilbert (1896) was the first to report on the geology of the Purgatoire River canyons area in a study of the water potential in the Arkansas Valley and its associated drainage

PAGE 18

4 basins. Reconnaissance along the Purgatoire River above Bent Canyon and along Two Buttes Creek in 1895 confirmed ideas developed with R. T. Hill when they mapped the foothills near Beulah the previous year, that the beds between the Cretaceous and the Paleozoic rocks in southern Colorado could be correlated with the "Juratrias" in northern Colorado. Lee investigated exposures along the "Purgatory River" in 1901 with a side trip to the "Rio Cimarron" (Cimarron River in northeast New Mexico). Dinosaur bones in the shales between the Dakota sandstones and the gypsums of the "Red Beds" prompted Lee to announce the discovery of the Morrison Formation in the area. The next summer Lee (1902) returned to the region and explored Morrison Formation outcrops in the Cimarron and Canadian canyons of New Mexico, and the Apishapa canyon and Sangre de Cristo mountains areas of southeastern Colorado. His paper to the National Acadamy for the Sciences in 1903 reporting Comanchean fossils (similar to Lower Cretaceous of Texas) in the shales of the Morrison Formation prompted discussion between Darton, who supported these findings, and Stanton, who argued that lithologic similarity did not mean "stratic equivalency". This discussion probably led to the productive 1905 field trip (Stanton, 1905; Darton, 1906) with Lee, Stanton, Darton and Gilmore. The reported Comanchean pelecypods were found in place in a marine shale separating "Dakota sands" above the Morrison shales along the Purgatoire River near Higbee. "Brontosaurus" bones were recovered from the Morrison shales, especially from

PAGE 19

5 the southeast corner of the Timpas quadrangle (probably from the Beaty Canyon Quadrangle map of 1972). "Belodon (Phytosaur, Romer 1945) indicating Triassic age" (Stanton, 1905) "portion of a scapula" (Darton, 1906) was collected from the upper layers of the Red Beds below the mouth of Chaquaqua Creek. Members of the The Colorado Geologic Survey investigated the economic potential of the geologic resources of Las Animas and Otero counties from 1916 to 1920. Results of this work, published several years later by Patton (1923), Duce (1924) and Toepleman (1924), recognized that the Morrison was probably Jurassic, not Cretaceous as was officially held by the U. S. Geological Survey at the time. Scientific attention was again brought to the region when MacClary (1936, 1938) wrote about the "recently discovered Purgatory River dinosaur tracks". Serious students of tracks including Roland T. Bird were involved with a more accessible site in Texas and so no further study was done on the Purgatoire tracks for almost 50 years. Ben H. Parker, who was familiar with the Cimarron Arch region, measured three sections in southeastern Colorado which Heaton (1939) included in his regional correlation of Jurassic deposits of Utah, Colorado and New Mexico. This report correlated the sandstone below the Morrison Formation (the Exeter Sandstone as named by Lee, 1902) with the Entrada Sandstone of the Colorado Plateau.

PAGE 20

6 In an attempt to understand the stratigraphy and structure of the Model helium gas field, Bass (1947) measured sections in Purgatoire canyons and correlated them to subsurface drill samples 20 miles to the west. The 1956 Rocky Mountain Association of Geologists Guidebook to the Raton Basin included a road log to the "Canyon of the Purgatoire River" by E. R . Landis (1956) and a review of the problems of Triassic and Jurassic stratigraphy in southeastern Colorado by Oriel and Mudge (1956). Other papers (Oborne, 1956; Shaw, 1956) used sub-surface drill cuttings and well logs to demonstrate several late Paleozoic paleogeographic highs in southeastern Colorado. Long's (1966) correlations of the Lower Cretaceous in southeastern Colorado included data from the mountains near Colorado Springs and Trinidad south and east to the Purgatoire canyons. Taylor (1974) measured sections in the Purgatoire and other nearby canyons focusing on early Cretaceous environments in an e astward extension of the work done by Long. The La Junta Quadrangle map of Scott (1968) was the first geoloeic map of the region since the 1924 reports of the Colorado Geolog ical Survey. This map indicated U . S. Geological Survey acceptance of the use in southeastern Colorado of Paleozoic terminology from the subsurface of Kansas . The Rocky Mountain Association of Geologists published a series of cross sections, including two across the southeastern portion of the state assembled by J . Wilson (1977). These

PAGE 21

7 subsurface sections utilized both geophysical logs and drill cuttings to emphasize the Paleozoic structures which are of major interest to the petroleum industry. Investigation of the Purgatoire River tracksite by the \ University of Colorado at Denver Geology Department began in 1982, and reports published to date include Conrad et al. (1987); Frazier et al. (1983); Lockley (1986a, b); Lockley et al. (1984; 1986); Prince (1983); Prince and Houck (1986). Some of these publications include data compiled for this thesis. To the west and north of the study area lies the Pinon Canyon Manuever Site of the United States Army. Before this area was occupied by the Army in 1982, a survey of the geology and paleontology, supervised by Kauffman (1986), was completed by students at the University of Colorado in Boulder. The pre-Cretaceous stratigraphy was the responsibility of J. T. Kirkland and J. G. Eaton who found bone fragments of possible Late Jurassic age near the top of the massive red sandstones. Purpose and Procedure None of the previous studies in the region focused on the Morrison Formation. Some of the formation names applied here have been shown to be invalid, and others are inappropriate in light of new information. This research was therefore conducted at two scales: a detailed paleoenvironmental study of the Morrison Formation in the Purgatoire uplift (Figure 2) and a larger-scale regional stratigraphic and paleogeographic framework

PAGE 22

8 (Figure 1). Because the original intent of the study was to understand the environmental setting of the beds containing the dinosaur tracks which are in the lower portion of the formation, and because the lower portion is better exposed in the study area, emphasis was placed on development of a lacustrine model for this portion of the section. Field data is the cornerstone of this study, and is supplemented by analysis of outcrop and subsurface samples and geophysical well logs. Ten sections in the Purgatoire uplift were measured in detail and eleven nearby exposures were cursorily examined (Appendix). Using tape, the well-exposed lower shales and limestones were measured in .1 m intervals, although a few units were reported in centimeters. Covered intervals and thicker sandstones were measured in meters with a jacob staff and brunton compass. Each section is a composite, prepared by following a bedding plane laterally from one exposure'to the next (Chapter 6, Figure 46, 47). These exposures are sometimes separated by as much as .5 km of colluvium. Potential for errors exist because: 1) facies changes may occur in the covered intervals, 2) the stratum was covered or weathered beyond recognition, 3) the bed was removed by erosion just after deposition.

PAGE 23

57 56 55 54 ,------r-------r------T------1 D CRETACEOUS II I AND YOUNGER I 1::::1 J U A A s s I c & I 2 61 L:J TRIASSIC I [;01 PERMIAN I r I I 271 I I I J-1 I I 2 8 1 I I I r I I I 2 9 1 I I .. I I 3 o I I I . I r I I 3 I I ROCK I : I : I 9 Figure 2. The Purgatoire uplift region is the area of concentration in this study. Jurassic and Triassic units are undifferentiated here because of the uncertainties of correlation of the older units (after Scott, 1968)

PAGE 24

10 Over 360 samples were collected from the nine primary sections. Approximately 1 kg of material was taken from each unit identified in the well-exposed intervals. Representative samples of all other exposed lithologies were also collected. Rock color was recorded using the GSA color chart. All samples were examined under a stereo microscope using the "Amstrat" technique, a simple, inexpensive way to identify components and textures in rock chips by breaking them down in dilute HCl (American Stratigraphic Company, 1973). Sixty-eight thin sections were made and examined under the petrographic microscope to obtain greater detail. Sixty-one friable clay-rich mudstones and shales were processed to remove the clays and extract calcareous microfossils. Some of the thin section and microfossil preparation and analysis was performed by UCD research team members K. Houck, S. Chesson and D. Adelsperger. Members of this team also examined vertebrate trace fossils in detail at the Rock Crossing site using standard mapping and replication techniques (Lockley, et al., 1986). The potential for palynologically-based biostratigraphy was not pursued. X-ray analyses were done on eight samples to obtain bulk carbonate and clay mineralogy. Published measured section data was transcribed (by N. Prince) to the regional correlation scale of 1 in = 10 m to allow incorporation into the data base. Outcrops in each of the general areas of exposure were visited in order to develop a unified interpretation of the data.

PAGE 25

11 Drill cuttings from five wells stored at American Stratigraphic Company were examined using the "Amstrat" technique. These were the only wells in the area for which both geophysical logs and cuttings were available through the Morrison Formation (Appendix, wells marked with *). Gamma ray geophysical logs were obtained for 41 wells in the study area (Appendix). Formation tops were picked on these logs by incorporating data from scout cards at Petroleum Information Corporation with interpretations of cuttings from neighboring wells and with measured sections from nearby outcrops. Some of the logs, available only at 1 in = 250 ft, were xerographically reduced to 1 in = 100 ft. Outcrop data was replotted to the same scale to prepare the regional cross sections in Plate I. Results of these investigations constitute two distinct data sets: (1) detailed descriptions of sedimentology and paleontology (Chapters III, IV and V) used to interpret microenvironments, and (2) cross sections and maps used to understand regional stratigraphy and paleogeography (Chapters VI and VII) • . Generalized stratigraphic columns (1 in= 10m) which contain lithology and paleontology data for all measured sections are included on Figure 10. Details of selected intervals are included in Figures 41-47 (Chapter VI); a complete set of the ten primary measured sections (1 in = 1 m) is available from the author.

PAGE 26

CHAPTER II GEOLOGIC SETTING Regionai Structure The Purgatoire River disects a broad, regional upwarp (Figure 1) called the Purgatoire uplift (Heaton, 1939) or the Red dome (Gabelman, 1956). Structural elements which define the limits of the uplift are the Black Hills, Muddy Creek and Mustang Creek monoclines (Duce, 1924; Scott, 1968). The dome is underlain by the late Paleozoic Apishapa uplift (Tweto, 1980), an uplifted basement block bounded on the north by the Apishapa fault, to the east by the Freezeout Creek fault, to the west (west of the study area) by the Wet Mountain fault and e xtending to the south into the Sierra Grande arch. The southern terminus of the Las Animas arch is just to the northeast. For convenience in this report the areas of the Muddy Creek monocline and Black Hills monocline, where Lower Cretaceous sandstones rim canyons exposing relatively uniform Jurassic, Triassic, Permian strata, will be referred to as the Purgatoire.uplift (Figure 2). This includes exposures along the

PAGE 27

Purgatoire River, Chaquaqua Creek and Plum Creek, within townships 26 to 31 south and ranges 54 to 57 west. 13 The Purgatoire uplift is at the junction of several well-studied geologic provinces: the Denver basin and Front Range, the Oklahoma panhandle, northeastern New Mexico and east-central Kansas. A hybrid stratigraphic terminology has been applied by various authors, but regional correlations across this area have been published only on the Jurassic Entrada (Heaton, 1939) and Lower Cretaceous Purgatoire Formations (Long, 1966). Paleozoic Stratigraphy and Structure Early Paleozoic strata are thin to absent across the crest of the Apishapa uplift (Figure 3), due to non-deposition or erosion during late Paleozoic time (Maher, 1953). Where these sediments are preserved, they are composed of limestone and dolomite with minor quartz sand, an indication that any exposed land mass was too low in relief to shed coarse clastic detritus (Gabelman, 1956). The Texaco Cynthia True #1 (Figure 4) test well was spudded in the Lower Cretaceous on the eastern limb of the Black Hills monocline. Here a complete, although thin, section of Cretaceous through Permian strata that overlies the Precambrian granites. Other oil and gas exploration wells demonstrate that during the late Paleozoic, the Apishapa and Sierra Grande uplifts were one land mass. Karstification of Mississippian sediments along the Las Animas arch to the northeast, lack of Pennsylvanian

PAGE 28

Muddy . / . , "' Block Hlllt_J_; lf'\ M on o c II n 1 .,-, .... / Creek Monocline ) _ _ _ .) _, ...... FtllliOUf F 1 ul t / / Eaprettlon of Aplahlpa Hq)hlancs Animal Arch W u t tano Creek Wonocllne Laccotlthtc Dome OTwo 8uttol C lmerrof'l Arch Figure 3 . Structural elements which influenced Jurassic deposition in the study area include the Late Paleozoic Apishapa and Las Animas and Mesozoic Sierra Grand uplifts. Present physiography is a result of Laramide and younger uplift of the Black Hills, Muddy Creek and Mustang Creek monoclines. (Area shown as in Figure 1.)

PAGE 29

r--.. DAKOTA 0 • u II GLENCAIRN .. • ... (,.) LYTLE Unit D MORRISON Unit C Unit B Unit A ENTRADAiEXETER u DOCKLt.c. • • • ... 0 V MESA (composite) T29S-R56W I: .. : . : .. , .. ,. :. J : <:: . . . . . . : ) .. . . ...:.;:-..:.. . : ... : : . . . . . --,.. . . -15 TEXACO + 1 Govt-Cynthia True 30-T28S-R56W -__ I I '-? I . --_---. . U . . I :... =t.. . .. -4.81G 8AIIN DAY CREEK 1 2 I . c: c \. ... .. 5 • -E WHITEHORSE .. • I gamma ray neutron Figure 4. The gamma ray log from the Texaco #1 Cynthia True well, as interpreted through the use of drill cuttings, correlates with outcrops in the OV Mesa area and serves as a tool for integrating subsurface data into this study.

PAGE 30

strata on the Apishapa uplift, and the presence of coarse conglomeratic Atokan age sediments on the northern and northeastern flanks of the Apishapa/Sierra Grande structure (DeVoto, 1980) are good evidence for uplift during this time period. 16 In the subsurface, Wolfcampian and Leonardian sediments are thin over the Apishapa uplift, Sierra Grande arch and Las Animas arch. Distribution of coarse clastics around these features indicate that although there was still some relief (Maher, 1945; Tweto, 1980) these highlands had largely been eroded by the end of Wolfcampian time. The oldest sediments exposed in the Purgatoire uplift (Figure 5) and along Two Buttes Creek in Baca County were deposited in a marginal marine setting during Upper Permian (Guadalupian) to Lower Triassic time. Thirteen meters of orange-red, gypsiferous, very fine sand and silty shale of the Permian Whitehorse Formation are exposed in the Purgatoire River bottom (Figure 6). Above this the Permian Day Creek Dolomite contains seventeen meters of dark red silty mudstone interbedded with crinkly, stromatolitic, sometimes brecciated, limestone and dolomite. The twelve meters of dark red, thinly bedded, frequently rippled, very fine grained sandstone and siltstone which follow are assigned to the Permian Big Basin Formation. These unique crinkly stromatolites and the sandy siltstones above the Whitehorse Formation are similar to and correlative with beds in the Forelle and Strain members of the

PAGE 31

CAITACIOUI JUIUIIIC T"IAIIIC S I I nt OIL Owe•, " • 11 on, 8 I' I , O r lal l "'"'''' Sco 1 t, A IOOrl La • • t t 0 I It 0 5 It 2 4 ,, 31 t •• ., Wwdo•, It 51 11511 It II I DAKOTA DAIICOTA DAM: OT A DAKOTA DAKOfA DAKOTA w c GLENCAIAN 0 DAKOTA UNNAioiD KIOWA GLENCAIRN 0 0 < ,_UAGATOIAE 0 COMANCHEAN PUAOATOIAE < .., 0 c LYTLE ChEYENNE LYTLE L WOAAISON NORRIS ON MORRISON WOARISON WOAAISON WORAISON WOARISON WORAISON MORRISON CYPSIFEA •-.... ous UN IT TRANSITION TRANSITION TRANSITION Zw . MIDDLE UNif RALSTON CA NIDDL UNIT > ---FLU V l A L ENTRADA ENTRADA wo UN IT UN•NAWE D ENTRADA 0[0 ENTRADA CANYON ENTRADA > ILO AN , CANYON IOUIYALfNT . u IUD lfDI 0 0 UN•NAN I!NTAADA DOCKUM JURA/TRIAS z DOCKUW DOCK UN DOCIICUW UN•NAMEO 0[0 IHALIE TALOOA HAOUAOU TALOGA L Y K I N I 110 lAtiN 110 IAIIN DAY CAI!EK DAY CRI!!K DAY CAEI!K DAY CREEK WHITE HORSE Sl 4 IN E HOASEC>< WHITE LYONS WHITE HORSE Figure 5 . Stratigraphic correlation chart of ana Mesozoic strata exposed in the Purgatoire uplift. The Morrison Formation, Purgatoire Formation and Dakota Sandstone correlations are undisputed. Earlier Jurassic, Triassic and Permian rocks lack datable fossils, contain numerous unconformities and are thus difficult to correlate with certainty to adjacent depositional basins. T
PAGE 32

19 Permian/Triassic Lykins Formation along the northern Colorado Front Range. Kansas and Oklahoma terminology (Whitehorse Formation, Day Creek Dolomite and Big Basin Formation) (Scott, 1968, Maughan, 1980) should be retained because subsurface correlation to the east is easily demonstrated, and because the Lykins is absent from outcrop in the mountains south of the Canon City Embayment (Tweto, 1979). Mesozoic Stratigraphy and Structure Throughout the study area 45 to 300 m of red-colored siltstone and sandstone present between the distinctive Day Creek Formation and the gypsiferous Bell Ranch Formation. Correlation of these redbeds is complicated (Oriel and Mudge, 1956) because (1) the depositional environments were probably largely terrestrial resulting in similar but laterally variable lithologic suites within different time periods, (2) fossils are sparse, and for the most part do not yield diagnostic ages, (3) exposed unconformities are difficult to trace in the subsurface and (4) subsurface data is sparse and often of poor quality. Figure 5 highlights the confusion regarding the correlation of these strata, and the text which follows presents new information on which the correlations of this report are based. -

PAGE 33

20 Triassic The deep red, thin-bedded fine-grained sandstones exposed at Two Buttes Reservoir and Purgatoire uplift (Figure 6) are probably most correctly correlated with the Late Triassic Dockum Group in northeastern New Mexico (Heaton, 1939) based largely on similar stratigraphic position and to a lesser degree on similar paleontologic and lithologic characteristics. These sandstones, 42 m thick at Pamena Gap, overlie the maroon siltstones of the Big Basin Formation with a sharp, unconformable contact, but appear conformable with the eolian sandstones above. Although there is some similarity in lithology between the unit exposed in the Purgatoire uplift and the Travesser Formation of the Dockum Group in northeastern New Mexico, there is insufficient evidence to directly correlate the two units (K. Conrad, pers. comm., 1987). Unidentified plant remains and bone fragments reported by Duce (1924, p. 81) from exposures near the mouth of Chaquaqua Creek suggest a terrestrial origin for these sediments. In northeastern New Mexico, terrestrial biota of the Late Triassic Dockum Group include plants (Baldy Hill Formation, Baldwin and Muehlberger, 1959), reptile body fossils (Sloan Canyon Formation, Stovall and Savage, 1939) and trace fossils (Sheep Pen and Sloan Canyon formations, Conrad et al., 1987). Subsurface cross sections across the Sierra Grande and Apishapa arches (McLaughlin, 1954; Oriel and Mudge, 1956 and-

PAGE 34

21 Figure 7 and Plate I of this report) support this correlation, but are inconclusive. All use incomplete geophysical well logs (Figure 4) and sporadic well cutting descriptions from widely spaced petroleum exploration wells, which commonly provide conflicting results. Outcrops in western Prowers County mapped as Triassic Sloan Canyon Formation by McLaughlin (1954) and Scott (1968) are here considered to be Morrison Formation because of their stratigraphic relationship to nearby Morrison Formation and Purgatoire Formation outcrops (See Plate I and discussion in Chapter VII, Regional Paleogeography). A comprehensive regional study of the lower Mesozoic stratigraphy might resolve these suggested correlations of the red beds exposed in the Purgatoire canyons with those exposed at Two Buttes, Colorado and those in southwestern Kansas, northwestern Oklahoma, northeastern New Mexico and the centra l Colorado Front Range. Outcrop studies should include compar ison of the biota and examination of possible marine sedimentary structures (J. Kirkland, pers. comm., 1987), although Late Triassic biostratigraphy is presently controversial (M. Parrish, pers. comm., 1988). This analysis should be compared with subsurface data, including drill cuttings and a full suite o f geophysical logs to reach a resolution of the problems.

PAGE 35

I B l I Pui r g a Reg' o n I I I . I II i .. I I E' ; I • I i l 23 . 1 i I -:-1 . ---r-, t T lC " • T .. • T " 1 I I 1--+---+----1 lO , zci-,---' I ' I .. J ! • • -,,., .... Figure 7 . Location of cross sections combining outcrop and subsurface data. Plate I contains sections A D and Plate II contains section E . N N

PAGE 36

23 A complete understanding of the structure during the Late Triassic awaits unravelling of the stratigraphy. There is evidence, however, for areas of positive relief to the west (Huerfano Park, Johnson, 1959) and the south (unconformities along the Cimarron arch, Baldwin and Muehlberger, 1959) and possible independent movement of the Sierra Grande and Apishapa arches during the Triassic (Baldwin and Muehlberger, 1959). Jurassic Entrada Sandstone The thick pink to salmon fine-grained sandstone with large-scale cross bedding (Figure 6) exposed at Pamena Gap on the Purgatoire uplift is here correlated to .the Entrada Sandstone in the Cimarron Valley of northeast New Mexico. At Pamena Gap the sandstones are 52 m thick, but thicknesses vary locally from 32 m to 65 m (Lee, 1901; Heaton, 1939; Taylor, 1974) in the canyons of the Purgatoire uplift. The top few meters of this sandstone often contain white pisolites (oolites of Lee, 1902), which are concentrations of calcium carbonate and iron oxide around sand grains. These are larger and more prominent on the upper surface of the sandstone, as are zones bleached white along fracture planes. Directly above this surface is a 3 m thick, mottled limestone. The mottled limestone, pisolites and leached texture might be the result of pedogenic processes acting on the dunes in an arid climate, and when combined with evidence for erosional

PAGE 37

24 relief, indicate an unconformity of unknown duration at the close of red-bed deposition. Heaton (1939) correlated the Exeter Sandstone (Lee, 1902), a white, thickly cross-bedded sandstone in northern New Mexico, with the type Entrada Sandstone of the Colorado Plateau (Gilluly and Reeside, 1926). _1be thickly cross-bedded strata of the Purgatoire uplift is similar in lithology and stratigraphic position to the Entrada Sandstone of northern New Mexico. It is possible to correlate a sand body of varying thickness from Pamena Gap through the subsurface south to the New Mexico border (Plate I). This sandstone is assumed to be the equivalent of the lower portion of the Entrada Sandstone of the Cimarron Valley as designated by Lucas et al. (1985), not the Exeter member (see following discussion of Bell Ranch Equivalent). Sandstones equivalent to the Entrada may also be present (Plate I) to the east at Johnny Branch Creek (Heaton, 1939) and Two Buttes Reservoir (McLaughlin, 1954) and to the west in Cuchara Canyon (Heaton, 1939) and Huerfano Park (Johnson, 1959). Taylor (1974) and Scott (1968) follow Oriel and Mudge (1956) in reporting a thin, white sandstone above the possible paleosol in the Purgatoire uplift as Entrada and include the thick, cross-bedded sandstone with the Triassic Dockum. This designation was based on subsurface correlation with the Dockum Group of northwestern Texas (G. R . Scott, pers. comm., 1987) but influenced by Stanton's report (1905) of Triassic "Belodon" bone fragments (phytosaur, Romer 1945) collected from the upper layers

PAGE 38

25 of the Red Beds below the mouth of Chaquaqua Creek. Kirkland (J. Kirkland, pers. comm., 1987; Kauffman, 1986) however, reported finding Late Jurassic bone fragments from a conglomerate just above the thick sandstone, four miles south of the junction of Chaquaqua Creek and Purgatoire River, probably in the same horizon from which the "Belodon" bone was collected. The UCD research group has also collected unidentified tooth and bone fragments from the conglomerate at several locations. Additional field and lab work to determine the presence of pre-or post-Entrada fossils here would be crucial in establishing the age of the cross-bedded sandstone. Bell Ranch Equivalent Basal Fluvial Unit. The mottled paleosol horizon at the top of the Entrada sandstone is truncated near OV Mesa by a coarse conglomerate about 1 m thick containing clasts of the pisolitic sandstone, clay clasts, chert pebbles and bone fragments. Dark red, ripple cross-laminated, very fine-grained sandstone overlies the basal conglomerate and is interbedded with other thin conglomerates (Figure 8). The total package is up to 4 m thick, and the top bed is a siltstone containing large (2 to 5 em diameter, up to 20 em long) tubular trace fossils, probably root casts. The unit appears to be fluvial, and cross sets measured at the Lost Canyon section indicated a paleocurrent direction of N 20 W.

PAGE 39

27 The unit as described is persistent on the east side of the Purgatoire River and on the west side of Chaquaqua Creek around OV Mesa. The thickness and percentage of sandstone and conglomerate is variable throughout the Purgatoire uplift and is interpreted as interdune fluvial deposit. On the west side of the Purgatoire River near Red Rock Canyon it is up to 40 m thick (Kauffman, 1986), to the southeast in Trough Canyon of Chaquaqua Creek it is 6 m thick and it does not appear to be present 10 km east on the wall of Plum Canyon or north in Bravo Canyon near the confluence of Chaquaqua Creek and the Purgatoire River. It is also absent from other outcrop exposures of the Entrada Formation in southeastern Colorado listed above. A similar conglomerate is reported in the Canadian River Valley of northern New Mexico and is possibly correlative to the J-5 unconformity of Pipiringos and O'Sullivan (1976) (G. R. Scott, pers. comm., 1987). Gypsiferous Unit. Conformably above the conglomerate and sandstone of the basal Bell Ranch Equivalent in the Purgatoire uplift the red siltstones grade upward into a 25 to 40 m thick unit of interbedded siltstone, mudstone, sandstone, clayshale, and gypsum. No fossils have been found in this interval. The siltstones and mudstones are predominantly reddish brown in the lower third, and gray throughout the upper portion. The white gypsiferous sandstone beds are up to 1 m thick with a cross-bedded appearance (Figure 9) and form resistant ridges throughout the generally non-resistant unit. In one measured

PAGE 40

29 section (Lost Canyon) one meter thick gypsum beds are interbedded with thin, red and green mudstones and light green clayshales. Both the fluvial conglomerate and the bedded gypsum are present near Canon City, Colorado. Lateral facies changes occur across paleovalleys in the underlying formations, and up to 45 m of gypsum has been deposited in the paleo-lows (Fredrickson et al., 1956). Chert Band. In the Purgatoire uplift the top few meters of the gypsiferous unit contain a distinctive lithology, most correctly classified as a thinly' bedded crystalline limestone. The secondary calcite crystals are parallel to bedding in thin sandy clay-rich beds. Minor gypsum and barite nodules are present and granule-sized nodules of pink and white chert replace all other minerals. The chert nodules become larger upward to a horizon of solid blue and red chert about 20 em thick (Figure 10). The calcite crystals, disseminated chert and gypsum nodules continue 50 em above the solid chert. Overlying beds are thinly laminated, non-calcareous clayshales, with scattered selenite crystals and rare thin stringers of limonite, but with no bedded or nodular gypsum. This horizon is similar to "the agate bed" and "the welded chert bed" common in the lower Morrison (Ogden, 1954). The value of this unit as a stratigraphic marker is discussed below. Description and review of the mechanisms of formation are in Chapter V, Geochemistry.

PAGE 41

31 Discussion. Other informal stratigraphic designations that could be used for this unit include: Ralston Creek Formation (Scott, 1968; Lockley et al, 1986), now considered an invalid name (Pipiringos and O'Sullivan, 1976), "Brown Silt Member" of the Morrison, and "Middle Unit". Because these names cannot be used without formally proposing new stratigraphic terminology and/or reinstating rejected terminology, they are discussed only briefly before focusing on the relative merits of the formally accepted terminology of time equivalent units. Ralston Creek Formation. Pipiringos and O'Sullivan (1976), (using a chert pebble horizon) questioned the type locality of the Ralston Creek Formation (LeRoy, 1946; Van Horn, 1957) suggesting that the lower portion is equivalent to the lower Canyon Springs Sandstone Member of the Sundance Formation (Callovian) in southern Wyoming and that the upper nonconglomeratic part is equivalent to the Morrison Formation (Kimmeridgian). They did not attempt to ascertain the affinites of the gypsum present below the Morrison Formation at other localities along the Front Range however (R. O'Sullivan, pers. comm., 1986). Although the U.S. Geological Survey no longer recognizes the unit, de Albuquerque (1986) chose to retain the Ralston Creek Formatton for the pre-Morrison coarse clastics on the eastern flank of the Wet Mountains as no other term seemed appropriate for the lithology. Stratigraphic position suggests

PAGE 42

32 that these sediments are equivalent to the gypsiferous beds of the Purgatoire uplift. Un-named "Middle Unit". The un-named "Middle Unit" of Oriel and Mudge (1956) included diverse lithologies in similar stratigraphic positions: the evaporites and varicolored mudstones of the Denver basin and northeast New Mexico, the arkosic conglomerates near Canon City, and sandstone, red mudstone and conglomerate in southeasternmost Colorado and western Oklahoma. Kauffman (1986) used this term refer to the fluvial conglomerates and gypsiferous beds in the Purgatoire uplift. "Brown Silt" Member of the Morrison. The "Brown silt" member of the Morrison (Bachman, 1953; Baldwin and Muehlberger, 1959) is a thin-bedded, light brown siltstone and very fine grained sandstone with minor thin limestone and gypsum about 3 m below the "agate bed", which at the base is interbedded with the Exeter Sandstone. It is present in north central New Mexico (Mankin, 1958), along the Rio Cimarron north of Kenton, Oklahoma and in Carizzo Creek in Baca County, Colorado. Dinosaur tracks, possibly Late Jurassic in age, have been documented from this horizon near Kenton (Lockley, 1986; Conrad et al., 1987). Stratigraphic position and lithologic similarity suggest that this unit is probably equivalent to the gypsiferous unit exposed in the Purgatoire uplift and possibly to the Bell Ranch Formation (Lucas et al., 1985, Conrad et al., 1987).

PAGE 43

33 Wanakah Formation. The Wanakah Formation was originally named as a member of the Morrison Formation (Burbank, 1930) for shale, limestone and sandstone exposures in mines in the Ouray district of southwestern Colorado. In the San Juan basin, Condon and Peterson (1986) have elevated the Wanakah Formation from member to formation status, with the Todilto and Beclabito members. Along the Front Range, west and south of Trinidad (Wood et al., 1957; Johnson, 1959) and in the subsurface of southeastern Colorado the marls of the lower Morrison have been called the Wanakah since they are similar lithologically to the Wanakah Formation at the type locality. A fish-bearing unit north of Canon City below the Morrison Formation was assigned to the Ralston Creek Formation and later to the Wanakah Formation, based on middle to lower Callovian age of the fish (Schaeffer and Patterson, 1984). Todilto Formation. The Todilto Formation, whether used to denote a member of the Wanakah Formation, of the Morrison Formation, of the Entrada Sandstone or a distinct formation (Gregory, 1916) defines a closed basin of deposition in northwestern New Mexico (Anderson and Kirkland, 1960) with limestone at the outer margin and gypsum above the limestone in the center of the basin. The Todilto basin was extended into east-central New Mexico, and south-central and south-eastern Colorado by Lucas et al. (1985).

PAGE 44

34 Based on the age of abundant fish fossils, the Todilto Formation and time-equivalent facies, the Bell Ranch and Exeter formations, are thought to be early Callovian (late Middle Jurassic) by Lucas. There is currently debate about the environment of the Todilto Formation sediments: it is considered to be either an inland lake (Anderson and Kirkland, 1960), a marine embayment (Ridgley and Goldhaber, 1983) or a land locked salina with subsurface recharge from a nearby sea (Lucas et al., 1985). Bell Ranch Formation. The type section of the Bell Ranch Formation (Griggs and Read, 1959) was proposed for the sequence between the Entrada Sandstone and Morrison Formation previously called the Wanakah Formation at Carpenter's Point, east-central New Mexico. It was further defined by Lucas et al. (1985) as a reddish, ribbed sandy/silty formation with disseminated or nodular gypsum and minor limestone. It is in sharp, possibly conformable, contact with the Entrada Sandstone and appears to interfinger to the south and east with the Exeter Member of the Entrada Sandstone. Locally it overlies the Todilto Formation suggesting it may be a landward shoreline facies of the Todilto water body. As suggested in the following discussion, the weight of evidence favors designation of the strata in the Purgatoire uplift area as being equivalent to the Bell Ranch Formation.

PAGE 45

35 Discussion. The J-5 unconformity is represented by an erosion surface, possibly correlative to a slight eustatic lowering of sea level (Vail et al., 1977). If the lower fluvial unit at the Purgatoire uplift marks the J-5 unconformity, then it might be possible that these gypsiferous units should_ be included in the Morrison, as is the Tidwell member of the Morrison on the Colorado plateau (O'Sullivan, 1984). However, this surface has not been clearly identified in the San Juan basin (Peterson and Turner-Peterson, 1987), and it would require careful' analysis to confirm the J-5 correlation in southeastern Colorado. This, then, does not seem to provide a convincing solution to the stratigraphic problem. Current feeling (J. Ridgely, pers. comm., 1985), based on preliminary geochemical observations, is that the gypsum in southeastern Colorado belongs to a different (non-marine) depositional system and possibly different time from that of the Wanakah and/or Todilto Formations and therefore should not carry the San Juan basin terminology. At most outcrops in southeastern Colorado, northwestern Oklahoma and northeastern New Mexico examined by the author there is a brownish silty, usually gypsiferous unit between the Entrada/Exeter cross-bedded sandstone and the Morrison Formation drab shales. This unit can be traced on subsurface well logs, varying from 10 m in the south and east to 45 m in the Purgatoire uplift (Plate I).

PAGE 46

36 The lithologies of this unit vary, but are explainable in terms of a playa-type flood plain with clastics coming off the Front Range and Cimarron arch area grading into an evaporite basin in the Purgatoire uplift area. The thinly interbedded siltstones and sandstones and evaporitic beds (especially gypsum and limestone) are common to both the unit in question in the study area and the Bell Ranch Formation of northeastern New Mexico. If Lucas et al. (1985) are correct in that the Todilto Formation, Exeter Formation and Bell Ranch Formation are all late Callovian in age and if the gypsiferous unit in the Purgatoire uplift is equivalent to them, then the bone-bearing conglomerates below the gypsum are also Callovian (Middle Jurassic) or older. Three independent, unconfirmed lines of evidence, however, suggest that the gypsiferous unit is Late Jurassic in age: the large size of the "ornithopod" dinosaur footprints in the "Brown Silt Member" near Kenton (Conrad et al., 1987), Kirkland's suggested date for the bone fragments in the lower fluvial unit, and the possibility that the conglomerate in the fluvial unit may mark the J-5 unconformity, which is taken as the base of the Kimmeridgian (Berman et al., 1980). The term Bell Ranch Equivalent therefore is used for the purposes of this report with the reservation that, although it can be mapped as a lithologic unit with definite upper and lower boundaries, the problem of time equivalence has yet to be completely resolved.

PAGE 47

37 Pre-Morrison Jurassic Structure Several lines of evidence suggest that north-south and east-west trending low relief remnant highlands may have separated southeastern Colorado from areas to the west and southwest. Baldwin and Meuhlberger (1959) infered that the Exeter/Entrada formations were deposited on a relatively flat surface along the Colorado-New Mexico border but low relief along the ancestral Front Range (Long, 1966) might have contributed to linear isopach trends (Lucas, et al., 1985) within this eolian formation. A coarse arkosic fluvial outwash plain in the "Ralston Creek" Formation (possibly a Bell Ranch equivalent, see Chapter 3) of the Wet Mountains (de Albequerque, 1986) further supports the idea that Precambrian rocks were still exposedto the west (Fredrickson et al., 1956). Other thickness, lithofacies and geochemical changes in Todilto and Bell Ranch equivalent formations indicate separation of depositional basins in north-central New Mexico from those in southeastern Colorado (J. Ridgely, pers. comm., 1986). In the study area, it appears that Bell Ranch equivalents are predominatly reddish-brown thinly-bedded fluvial floodplain siltstones and sandstones around the eastern and southern margins. Although these beds contain minor gypsum, the thickest units of bedded gypsum and thin, white, gypsiferous sandstones appear in the center of the region, north of the Sierra Grande uplift.

PAGE 48

38 Isopach maps of the Entrada Sandstone, the basal fluvial unit and the gypsiferous unit of the Bell Ranch Equivalent in the Purgatoire uplift might help reconstruct local areas of relief which remained prior to the deposition of the Morrison lakes. The thin, even-bedded nature of the uppermost Bell Ranch Equivalent and lowermost Morrison Formation sediments in the uplift suggest however, that there was little relief within the immediate area. Lower Morrison Contact and Chert Bed Most authors working in the Purgatoire uplift have included the bed containing red and blue chert nodules described above with the underlying gypsiferous units. Here the sandy, cherty crystalline limestone is followed immediately by laminated, gray non-calcareous clayshales. In adjacent exposures at Peacock Canyon, northeastern New Mexico, and north of Kenton in western Oklahoma, the chert is in the clayshales of the lower Morrison as reported by Baldwin and Muehlberger (1959) and Lucas et al. (1985). At other outcrop localities in the study area, there appears to be a similar change from sand-sized grains below to clay-size grains above, as well as loss of nodular and bedded gypsum above the chert. As described above, however, the chert occurs at several horizons within the same unit (Mankin, 1958). The contact then should be taken at the change in grain s ize and

PAGE 49

39 loss of gypsum although the presence of the red and blue chert may have utility in areas of poor exposure (Baldwin and Muehlberger, 1959; Taylor, 1974). If the presence of chert does represent a "time-surface marker" as suggested by Ogden (1954) this "time surface" crosses the formation boundary. Morrison Formation Eldridge (in Emmons, 1896) named the Morrison Formation for fresh-water marls averaging 200 feet thick between the Dakota Sandstone and the brown and pink "Trias" sandstone exposed just north of the town of Morrison, Colorado. Waldschmidt and LeRoy (1944) re-defined the Morrison Formation and described a type locality in Jefferson County, Colorado, where they subdivided the formation into six units. The upper sandstone and shale unit is equivalent to the Purgatoire Formation (Long, 1966); the five other Morrison units are: basal sandstone; gray and red shale; gray clay and gray shale and sandstone; and red shale. The re-correlation of the Ralston Creek Formation at its type locality by Pipiringos and O'Sullivan (1978) suggests that portions of the shale unit below the "basal sandstone" at the Morrison Formation type locality should also be included in the Morrison Formation. On the Colorado Plateau the Morrison Formation was divided into four members by Craig et al., (1955): the Salt Wash, Westwater Canyon, Recapture and Brushy Basin Members. Later revisions of the stratigraphy of the San Juan Basin include other

PAGE 50

40 local members: Tidwell, Bluff Sandstone, Junction Creek, Fiftymile, and Jackpile Sandstone Members (Condon and Peterson 1986; Peterson and Turner-Peterson, 1987). The lower contact of the Morrison Formation on the Colorado Platueau is now defined as being a widespread erosion surface (J-5) above which is the Tidwell Member, consisting of redbeds with occasional basal gypsum or gray and green mudstone. None of these members have been identified outside of the Colorado Plateau, and the Morrison Formation in these regions is therefore called "undifferentiated" (Craig et al., 1955). In Red Rocks Canyon and Plum Canyon of the Purgatoire uplift, Lee (1901) correlated the shales between the gypsum and the overlying resistant sandstones of the Dakota with the Morrison Formation of central Colorado based on dinosaur bone fragments (Morosaurus and Diplodocus vertebrae were identified by Barnum Brown, Lee, 1902) and lithologic characteristics. Brontosaur bones were recovered later from the southeast corner of the Timpas Quadrangle (probably from Clark Hill on the 1972 Beaty Canyon Quadrangle map) (Stanton, 1905). The gypsiferous beds were included with the Morrison Formation by Heaton (1939) and Bass (1947), but Scott (1968) and Taylor (1974) agreed that the lower contact should be the top of the resistant "chert bed". Most workers are in agreement that the contact of the Morrison Formation with the overlying Cretaceous rocks is the base of the massive white Lytle sandstone in the Purgatoire uplift area.

PAGE 51

41 Informal Subdivisions within the Morrison Within the Purgatoire uplift the author has informally subdivided the Morrison Formation into four units from bottom to top: Unit A, Unit B, Unit C and Unit D (Plate II). Away from the uplift, Units A and B thin and cannot be separated (Plate I), appearing as one single unit of calcareous clayshale interbedded with argillaceous limestones. Unit C thins and is absent in the Cimarron Valley but is thicker to the east at Johnny Branch (Plate I). Unit D is the thickest and almost always completely covered except for the thin sandstones and limestones. These four units are described in detail and their environmental significance discussed in Chapter VI, Depositional Environments. These units have not been correlated with published sections from other regions but the general sequence of calcareous clayshales followed by sandstones and mudstones is interpreted as a general shallowing of a lacustrine system and has been recognized as typical at different locations within the "undifferentiated" Morrison region (Dodson et al., 1980). Although Craig and his co-workers (Craig et al., 1955) felt that the "undifferentiated" Morrison Formation lithologies were similar to those of the Brushy Basin Member, recent work (Peterson and Turner-Peterson, 1987) suggests that the lacustrine limestone units such as in Unit A and B of this report are more similar, and perhaps more equivalent, to the Tidwell Member.

PAGE 52

46 Age of Morrison Formation Imlay (1980) using dates established on marine molluscs in northwestern Colorado, Utah and Wyoming, suggested that Morrison deposition probably began during middle to late Oxfordian and ended during lower Tithonian in the southern part of the western interior. Although it is difficult to determine ages in a largely terrestrial system like the Morrison, U. S. Geological Survey personnel, especially Pipiringos, O'Sullivan and Peterson (Pipiringos and O'Sullivan, 1976; O'Sullivan, 1984; Condon and Peterson, 1986) have devised a system of unconformities (J-1 through J-5) based largely on chert pebble horizons which they can identify throughout the Colorado Plateau region. The J-2 unconformity corresponds to the Bajocian/Bathonian boundary (lower Middle/middle Middle Jurassic), the J-4 to the Callovian/ Oxfordian boundary (Middle/Late Jurassic) and the J-5 to the Oxfordian/Kimmeridgian boundary (lower Late/middle Late Jurassic). These regional unconformities have been correlated beyond the Colorado Plateau, and Pipiringos & O'Sullivan (1978) conclude that the J-3 and J-4 are not present in southeastern Colorado. This is based on the correlation drawn from Wyoming to Ralston Creek Reservoir which shows the Canyon Springs member bounded by J-5 and J-2, directly on the Permian/Triassic Lykins Formation in Jefferson County. From southeast Utah across Four Corners to northwest New Mexico the Morrison Formation is above the J-5

PAGE 53

47 surface, and the Middle Jurassic Entrada (and Todilto member) is above the J-2 and the Triassic Chinle. The Lower Jurassic Glen Canyon Group is not present in the southern San Juan Basin (Condon and Peterson, 1986). Ages established for the Morrison Formation independent of this system are often inconclusive and sometimes conflicting. Pollen from two localites in Wyoming was similar to that of Purbeckian beds, therefore considered uppermost Jurassic (Dodson et al, 1980). However, pollen recovered from just below the Lytle sandstone in southern Colorado was diagnosed as "Jurassic" (Long, 1966) but other pollen samples from a variety of upper Morrison sites in Colorado, New Mexico and Wyoming yielded a possible pre-Kimmeridgian age (Hotton, 1986). Molluscs were suggested by Yen (1952) to be upper Jurassic, but more recent studies by Hanley et al. (1986) indicate that upper Morrison molluscs are probably facies controlled, and therefore not good biostratigraphic indicators. Ostracods and charophytes studied by Peck (1956) in Wyoming were considered to be Upper Jurassic. A collection of ostracods and charophytes from the Morrison of Oklahoma, Colorado, and New Mexico was identified and curated by Dr. Wedel and Betty Kellett Nadeau during the. 1940's (Stovall, 1943; Peck and Reker, 1948). This data was never published, however, and the disposition of this collection is unknown. It is probable that this biota will, as the molluscs, be highly facies controlled and inconclusive for biostratigraphic correlations.

PAGE 54

48 In northern China there is some interfingering of marine and non-marine Jurassic strata, allowing a more refined biostratigraphy. Chen Pei-ji (Nanjing Institute of Geology and Palaeontology) has tentatively identified a conchostrachan collected from the Goat Ranch section of the Purgatoire uplift as Nestoria. This genus is also found in the Da Hinggan Mountains north of Beijing, near the Mongolian border. Other specimens collected in the Purgatoire uplift area and from the Oklahoma panhandle have been referred to the above worker for identification in the hope that this international cooperation will shed light on these biostratigraphic problems. Six biochron zones based on dinosaur fossils collected from reference sections near Canon City, Colorado and Como Bluff, Wyoming have been proposed (Bakker, 1986). Zones 2 and 3 seem to be missing from Canon City and Como Bluff as well as any of the Front Range and Denver Basin sites. The youngest, Zone 6, is typical of Brushy Basin faunas, and is well represented in the Stovall quarries in northern Oklahoma, which occur in the upper half of the Morrison Formation there. Results obtained through radiometric dating are relatively consistent although they seem to indicate a younger age for the Morrison than had been established based on biostratigraphy. This apparent discrepancy might be due to "re-setting" of the radiometric clocks. Fission-track dates of Neocomian (early Cretaceous) for Dry Mesa, Colorado, and Cleveland-Lloyd, Utah quarries have been reported from zircons

PAGE 55

49 and apatites in bentonites (Kowallis, 1986). Potassium/argon dates from biotite crystals from a bentonite bed at the Cleveland-Lloyd quarry suggest that it is Tithonian, but those from Dinosaur National Monument are Neocomian (Bowman et al., 1986) . Cretaceous In the Purgatoire uplift a thick cross-bedded white sandstone, usually with basal conglomerate of chert, clay pebbles and occasional petrified wood, caps the Morrison Formation mudstones and sandstones (Figure 15). This is the Lytle Sandstone member of the Purgatoire Formation. The upper portion of this sandstone is often stained tints of purple and yellow. Deep maroon mudstone is somtimes interbedded and some of the brown thin, discontinuous sandstone beds about 10 m below the white sandstone possibly should be included in the Lytle Sandstone also. Long (1966), outside the Purgatoire uplift, was able to identify the Morrison/Lytle contact in resistent beds below the sandstone in a few sections where the contact was not covered. most sections this contact is covered, however, and Long (1966) and Taylor (1974) placed the contact at the base of the white sandstone, the major identifiable lithologic change. That convention is followed in this report. Yellowish brown bioturbated and fossiliferous sandstone and siltstone above the massive sandstone represents the first major marine transgression in the region since the

PAGE 56

51 Late Permian/Early Triassic (Taylor, 1974). This Glencairn Shale member of the Purgatoire Formation sometimes includes a dark gray shale at its top just below the massive brown weathering sandstones of the Dakota Sandstone. The Dakota Sandstone forms the caprock of the canyons and is the youngest formation exposed in the Purgatoire uplift. Stanton (1905) first identified the separation of "Dakota sands" by a marine unit with Comanche fauna in the Purgatoire uplift area. Stose (1913) attached the formation name Purgatoire to these sandstones and "fireclays" in the Apishapa Quadrangle. The type section is in Purgatoire Canyon on the 1890 Mesa de Maya Quadrangle, probably one of the exposures south of OV Mesa (approximately 37 degrees, 30 minutes north latitude). Finlay (1916) named the Lytle Sandstone Member and Glencairn Shale Member for exposures in El Paso County, Colorado. Long (1966) concluded that the Morrison/Lytle contact is conformable with zones of unconformity present where Lytle channels have incised into Morrison shales and mudstones. He recommends returning to Waage's (1953) useage of Purgatoire Formation and Dakota Sandstone in south central and southeast Colorado instead of the Kansas terminology (Kiowa and Cheyenne members) of Sanders (1934), Stovall (1938) and McLaughlin (1954). Upper Cretacesous Graneros Shale, Greenhorn Limestone, Carlile Shale and Niobrara Formation are exposed in the cliffs south and west of the Purgatoire uplift and constitute most of the high plains surface from the Purgatoire uplift west to the

PAGE 57

52 Rocky Mountains (Scott, 1968) . The thick late Cretaceous strata of the Pierre, Laramie and Fox Hills formations so evident in the Denver basin was removed b y erosion from the Purgatoire uplift largely prior to the Pliocene. Cenozoi c In the southern portion of the study area the Niobrara Formation is capped by the Pliocene Ogalalla Formation caliches and Miocene and Pliocene Mesa de May a basalts (Tweto, 1979). Surface sediments dip slightly, and become younger away from the central portion of the Purgatoire uplift, where the Ogalalla is missing, possibly due to erosion (McLaughlin , 1954). Within the canyons themselves, entrenched meanders are seen along the Purgatoire in the area of the Black Hills monocline. Above the stream terraces, the shale slopes show major slumping which probably occurred during early Pleistocene to early Holocene time (Colton, et al., 1975). In the Black Hills monocline, strata as old as Permian are exposed, but in the Muddy Creek monocline the oldest exposed rocks are Jurassic in age. Terraces with boulder size gravels, caliche and other associated soils appear in the Black Hills monocline. These differences support the possibility that at least two basement blocks, such as documented in the Northern Denver Basin (Sonnenberg and Weimer , 1981) with different tectonic histories, have been active. If the caliche in the Black Hills is Ogallala in age, much of the movement may have

PAGE 58

53 occurred along that structure prior to the Pliocene, possibly coincident with movement of the Wet Mountain and Apishapa uplifts during the Eocene documented by Tweto (1980) and Gabelman (1956). Some of the present physiography however, is probably due to post-Ogallala movement of both the Black Hills and Muddy Creek monoclines. Evidence for the latter structural movement is present in the Muddy Creek area along the Purgatoire River where numerous small anticlines with relief of less than one meter are present. These structures disrupt Jurassic formations, but not the surrounding Quaternary alluvium. Earthquakes reported by residents of Branson, along the base of Mesa de Maya, within the past 20 years (W. Louden, pers. comm., 1984) suggest that this area is still tectonically active.

PAGE 59

CHAPTER III SEDIMENTOLOGY Introduction Siliciclastic and non-skeletal carbonate grains, bedforms and non-biogenic sedimentary structures in the Morrison Formation are described in this chapter as are structures of probable pedogenic or diagentic origin. Grains and sedimentary structures clearly of biologic origin will be discussed in Chapter IV, and chemically precipitated sediments are discussed in Chapter V, Geochemistry. Grains Coated Grains Coated grains are abundant in the study area and may be described by Peyrt's (1983) categories: ooids (concentrically laminated, chemically precipitated in a phreatic environment), vadoids (thin, irregular to concentric laminated, chemically precipitated in a vadose environment) and oncoids (irregularly laminated, encrusted by green and blue-green algae and bacteria). Ooids and possible vadoids will be discussed in this chapter;

PAGE 60

55 oncoids will be covered in Chapter IV, Paleontology. This division assumes that biologic influence in the formation of ooids is less important to mineralogy and internal structure than is water chemistry and physical setting, and that the reverse is true in the development of those coated grains classified here as oncoids. Ooids Ooids are found at several horizons within Units A and B. Although some have nuclei of quartz grains, most appear to have micrite pellets in the center. SEM examination is required to determine cortical structure and mineralogy, but the presence of a psuedo-uniaxial cross when viewed with crossed polars is taken to suggest tangential structure and aragonitic mineralogy (Medwedeff and Wilkinson, 1983; Halley, 1985). Radial structure may produce a faint cross, and may represent either calcite or aragonite mineralogy. Ooids from the Goat Ranch and The Gap measured sections (Figure 2, Plate II) appear to be radial in nature, but ooids from the Rock Crossing section display pseudo-uniaxial crosses (Lockley et al., 1986, F igure 6) and are therefore tangential in nature. None of the Morrison Formation ooids were examined under SEM, but X-ray analysis confirms that those from Rock Crossing are now calcite although these may have been precipitated as aragonite. The probable radial structure and lack of extensive

PAGE 61

56 recrystallization of most of the ooids .however, suggest original calcite mineralogy, magnesium component unknown. Pellets Rounded micrite pellets commonly occur in some of the wackestones. Areas of distinct pellets grade into regions with micrite texture, the result of compaction prior to lithification and/or recrystallization. These pellets are possibly of fecal origin. Other oblong grains common in the algal grainstones may also be of biologic origin. Pelloids and Lithoclasts Large (up to several em in long dimension) current oriented flat micrite chips are found in many of the thin conglomerates of Unit B, often associated with oncoid fragments and a variety of bioclastic debris. Rounded clay pebbles are also characteristic of these conglomerates. Both the chips and the pebbles probably represent eroded and transported fragments of mud-cracked polygons similar to those of the Green River Formation (Eugster and Hardie, 1975) although wind deflation of dry flats might have helped concentrate these grains and transport them as at Amboseli (Behrensmeyer and Boaz, 1981) Siliciclastic Grains Most sandy horizons in Units A and B range from a few millimeters to a few centimeters thick (Figure 16). These calcareous, well-sorted, very fine-grained, angular sub-arkosic

PAGE 62

58 sandstones often contain ostracode and chara debris and occasional larger, rounded frosted quartz grains. These laminae are interbedded with micrite or wackestone. Both potassium feldspars and plagioclase are present, as is biotite and a variety of unidentified opaque and heavy minerals. The silt-to very fine-grained quartz and feldspar fraction is ubiqitous; very few micrites contain less than 10% randomly scattered siliciclastic grains. A few thin lenses of very well sorted, rounded, medium-grained sandstone are present (Figure 17), as is local poorly sorted, subround to subangular, fine-to coarse-grained conglomerate. The grain composition suggests that the source for most of the fine-grained sediment was acid igneous rocks, although exposed siliciclastic and carbonate sedimentary rocks appear to have sourced some of the coarser conglomerate 1984). Unit C is typified by sandy limestone (Figure 13, 18) and dolomite, units with sparry carbonate matrix and moderately sorted, angular, fine-to medium-grained quartz sand. These appear in places to grade laterally into sandy mudstone, also containing abundant quartz silt and sand. Although Unit D is volumetrically dominated by colored mudstone and shale, the thin sandstone beds are the most commonly exposed. They are moderately sorted, subround to subangular medium-grained clay sub-arkose. Chert grains and rock fragments are common, and the beds are often stained brown by the weathering of iron minerals (mica and hornblende seem prevalent).

PAGE 63

60 Clay-rich Sediments Clay mineralogy is discussed in Chapter V , Geochemistry, because of the interaction between clays and ambient waters. The clay-sized sediments containing less than 50% calcium carbonate and displaying laminations are called clayshales in this report (Potter et al., 1 980) . The term claystone is applied to those massive clay-rich sediments lacking fissility or bedding. Volcanic Glass Shards In Unit C, a few small grains with the three-cornered appearance of volcanic glass shards (Figure 19) have been identified. These are silica, and their close association with smectitic clays lend credence to their volcanic origin. Lower in the section similar shaped grains are composed of sparry calcite, and are probably bioclastic fragments r ather than calcitized shards. Bedding Bedding Style Laminated to thin-bedded . Units A and B are composed of interbedded thinly laminated to thickly bedd ed , fine-to coarse-grained beds, most with ripple and flaser morphologies (see below). Although some units show disruption of bedding by crystal casts, burrowing or other animal activity, the bedding is

PAGE 64

62 usually preserved. Possible explanations for this phenomenon include: 1) Deposition below wave base in a stratified (chemically and/or thermally) lake where unsuitable bottom conditions retard reworking by biota (Van Houten, 1964; Hakanson and Jansson, 1983). 2) Deposition in a shallow, relatively quiet environment where wave action is baffled by surrounding landforms or plants (Cohen and Thouin, 1987). 3) Deposition below wave base in an unstratified lake where alternations are due to fluctuations in sediment input from r ivers such as the Omo River into Lake Turkana i n Africa (Yuretich, 1979; Cohen et al., 1986) This mechanism may be active where graded laminations similar to those forming in deep water in Lake Turkana are found. 4) A high sedimentation rate which does not allow enough time for plants or animals to rework the sediment (Cohen, 1984). 5) Dominance of epifaunal organisms (Cohen, 1984). 6) Growth of stromatolic blue-green algae in a low energy environment (Monty, 1976). 7) Bioturbation during one season when the lake was higher or lower, followed by reworking by wave action before preservation (Laporte and Behrensmeyer, 1980). 8) Deposition in ponded (standing) water on mudflats where sediment-charged sheetwash rapidly decelerates and quickly deposits load. Graded thin beds or thick lamina result from

PAGE 65

deposition of the waning turbid underflow or settle out from mixed inflow into the lake (Hardie et al., 1978). 63 9) Suspended load fluvial deposition produces this type of bedding in the final stages. 10) Highly productive, well-oxygenated lakes may be depleted in oxygen on the lake bottom (Olsen, 1980a; Cohen, 1984) therefore restricting infauna. These laminations do not contain highly organic or reduced layers, and are not the thin, regular laminations typical of stratified lakes; therefore, the first model is one of the least likely processes. The others were probably active at different times or places during deposition of the lacustrine limestones. Cross-bedded. A few thin (less than .5 m), laterally very discontinuous (visible only for a few meters) tabular cross-bedded sand bodies are interbedded with the lacustrine marls of Unit B. Some of these have a quartz and clay pebble lag and fine upward into medium-grained sandstone; others are uniform poorly sorted medium-to fine-grained sandstone. These sands contain abundant oncoid and pelloid debris and are often capped with a clay drape containing abundant fish scale and plant fragment debris. Small channels on the surface of the dry lake bed probably were the sites of these deposits. The sandy limestone beds of Unit C show relict cross-bedding partly obscured by bioturbation and diagenesis.

PAGE 66

64 The tabular planar sets do not display the flat bottomed, convex upward shape typical of lacustrine bars (Peterson and TurnerPeterson, 1987). Lacustrine beach or bar origin is suggested, however, by the abundant well-washed and well-sorted siliciclastic grains not quite in grain to grain contact. Rare ostracode shells and abundant bioturbation also help support the lacustrine interpretation over the other possible environments of alluvial fan or floodplain limestone or fluvial tufa. A change in the groundwater levels and chemistries, with some eolian volcanic input possibly account for dolomitization and silicification of the original calcite matrix. Several sandstone bodies in Unit B and Unit D which have low-angle cross bedding are laterally continuous for at least 20 m. One exposed at Lost Canyon (Plate II) has the appearance of a small delta (Jackson, 1979; Reineck and Singh, 1980), Figure 173. Another, noted at The Gap measured section (Plate II), appears to grade laterally into a tabular cross-bedded sandstone and may represent a shallow ephemeral stream (Reineck and Singh, 1980). This particular type of sand body is typified by very fine grained quartz with extremely thin laminae of dark colored heavy minerals and weak current lineations on the upper surface. Scoyenia and Planolites trails are also associated with these sheet sands. Unit D sandstones are thinner (1 m) at the base, and become thicker (3-5 m) near the contact with the Lytle Sandstone. They are largely trough cross-bedded, sometimes tabular planar

PAGE 67

65 cross-bedded (Figure 14) and often have a basal lag of chert and clay pebbles. These sandstones can be traced laterally for long distances, and probably represent times when braided stream deposition (low discharge to sediment load ratio) was the rule. In the few good exposures of Unit D bright red and green mudstone containing isolated discontinuous tabular cross-bedded sandstone beds is present. During most of the later part of Morrison deposition, meandering, ephemeral streams on an oxidizing fluvial floodplain was the most common depositional environment, although suspended load streams could also have produced some of these sediments. It is this facies which could be examined in greater detail for vertebrate bones. Because of the evidence for the influence of wind generated waves on carbonate precipitation (Lockley et al., 1986), it is possible that there was also some eolian influence on distribution of carbonate and siliciclastic grains (Glennie, 1970). The frosted quartz grains common in some Unit B packstone lenses may have an eolian origin, but it is also possible that they were derived from older eolian sediments. In this report therefore, no eolian origin has been proposed for any of the sandstones, but this possiblity should not be totally excluded.

PAGE 68

66 Primary Structures Lenticular Ripple Bedding and Flaser Bedding Lenticular bedding, slightly connected or isolated lenses (Figure 20) of grainy material distributed in a muddy matrix, is common in Unit B. Flaser bedding, muddy lenses in a grainy matrix is observed less frequently. These bedding styles, although common on marine tidal flats, may be produced in any environment with alternating current energy and sediment supply. For example, lake bottom sediments in front of small deltas (Coleman, 1966) often are characterized by lenticular laminated sediments. On playa flats wind waves rework flood deposits to produce sand lenses and muddy drapes or simple surface rippling (Hardie et al. 1978, p. 21). It should be noted that in these Morrison structures the muddy sediments are micrites and calcareous shales and the sand lenses contain abundant bioclasts. Ripple Marks Several bedding planes, especially subaerially exposed carbonate packstones, contain large wave ripples. Measurements of amplitude, wavelengths and grain sizes of 41 fields of ripples were taken at the Rock Crossing site (Lockley et al., 1986). Wave-fetch distance estimates of over 100 km were derived, using the method of Tanner (1971, 1974). Although the apparent storm-generated nature of these beds make it difficult to substantiate that fetch distance, the fetch was probably greater than 40 km, a size predicted for most Morrison lakes

PAGE 69

68 (Dodson et al., 1980). This study was not attempted for Black Hills sections because bedding plane exposures of ripple marks were limited. Visual observation of a few at The Gap (Figure 21) and Round Corral indicate similar ripple dimensions and therefore similar wave fetch distances for this area. Debris Layers Pebble conglomerates with abundant oncoids, flat micrite chips, quartz grains and fish debris cap many subaerial exposure surfaces in Unit B (Figure 21) . Some are grainstones, some packstones; some are cross-bedded, others rippled; some fine upward, some are imbricate, some have clasts randomly distributed throughout the matrix. A scenario of "catastrophic expansion, gradual contraction" (Hardie et al., 1978) where a storm lifts mud clasts from the mudflat during a storm and deposits them with thin silt and clay drapes as water level and energy lowers, may explain some of these layers. If water level stayed constant for some time after the storm then wave activity could wash and sort the clasts. Wind deflation could also concentrate clasts on the dry lake or mudflat, and this debris co uld then b e deposited in shallow ephemeral channels prior t o renewed influx of water (Behrensmeyer and Boaz , 1981 ; Gle nnie, 1970) .

PAGE 70

70 Admixed Sediment Near the top of Unit B bedding is not often preserved. Amorphous "pockets" of grains are found in fine-grained sediment, or vice versa, "pockets" of micrite or claystone in wackestone and packstone textures. Possible explanations include the following: 1) Bioturbation by large vertebrates (Laporte and Behrensmeyer, 1980) , in this case dinosaurs, can be demonstrated at Rock Crossing (Lockley et al. , 1986) where a random admixture of sediment results. Most admixtures in the Black Hills area have this random pattern, and although three dimensional exposures are not available to identify tracks and trampling, "dinoturbation" (Lockley and Conrad, 1987) of exposed or shallow water deposits are probably responsible for much of thi s texture. 2) Flooding may dump heavy coarse sand on top of soupy mud, producing a variety of convoluted bed forms and load structures (Reineck and Singh, 1980) . This type of load structure is visible at Rock Crossing in one unit, but it appears to result, for the most part, in discrete packages of one type of sediment encased in the other, and is probably not the dominant operative process. 3) Some of the large clay pockets have distinct boundaries along one surface, and are gradational into micrite along another. These might have been semi-lithified mud, dislodged and transported as a clast only a short distance. 4) Loading and differential sinking of sediment into soft muddy lake bottom by earthquake shock may produce admixed sediment (Collinson, 1977b) . This is difficult to demonstrate in the

PAGE 71

71 ancient record, and is only listed as a possible alternative or additional explanation for some of these structures. Secondary Structures Salt Casts and Molds Crystal casts, both large rhombohedral-and small tabular-shaped, are disruptive features in a few beds in Unit B throughout the study area. Some are along bedding planes between sandstone and shale (Lockley et al., 1986 Figure 6), and others are randomly distributed within micritic limestone (Figure 22). Although crystal growth appears to have disrupted bedding somewhat, primary despositional structures are still very apparent. Playa lakes and associated mudflats are often characterized by bedded salts or highly disrupted bedding due to formation of authigenic crystals (Eugster et al., 1978). Although salts frequently return to solution during the next rainy season, the right conditions may result in their preservation. These crystals appear similar to those reported from the lacustrine Lockatong Formation of the Triassic Newark rift basin of the eastern U. S. (VanHouten, 1965), but crystals are less abundant and sediment less disrupted in the Morrison, possibly because the Morrison waters were not as saline (see Chapter IV, Geochemistry), and desiccation not as complete as in the Triassic example.

PAGE 72

73 Mudcracks Small (several centimeters on a side) polygonal structures occur throughout the sequence in the Purgatoire uplift. Most often the cracks disrupt interbedded micrite and packstone, and are filled with wackestone. Thin mud drapes above rippled oncoidal conglomerates (Figure 21) and algal mats are often mudcracked. These probably represent a period of desiccation (lowstand) following a period of higher water level (Hardie et al., 1978). The very tiny, millimeter sized mudcracks in the clay-rich calcareous shales of Unit A may have been caused by subaqueous synaresis (Collinson, 1978b) and are not necessarily evidence of subaerial exposure. Clastic Dikes Larger polygonal dikes have been observed in plan view at Rock Crossing (Prince and Lockley, in press) which are wackestone/packstone filled, 2-3 em across, 5-10 em in relief and form polygons up to 2 m on a side. Similar features have been observed in cross section at several horizons in sections measured at The Gap and OV Mesa (Figure 23). These polygons are reminiscent of those formed in temporary desert lake basins in Libya (Oomkens, 1966) which form when thoroughly saturated sediments undergo extended desiccation (R. Dubiel, pers. comm., 1987). Glennie (1970) notes that often only the upper layers dry and crack, and that when the natives remove

PAGE 73

75 this salt crust, the sediment beneath it is wet and salt free, suitable for growing crops. Pedogenic and Diagenetic Structures Mottling and Clotted Textures Mottling, clotted textures and circumgranular cracking (Figure 24) in some of the micrites suggest a palustrine environment (Freytet, 1973; Brown and Wilkinson, 1981; Esteban and Klappa, 1983). These features are formed by the action of roots and other soil organisms on the wet, marshy soil. Nodules, Geodes Nodular textures are observed in some of the micrites, as are geodes. These are possible rhizoliths and as such, suggest soil-forming processes or a palustrine environment (Klappa, 1980; Cohen, 1982). This nodular texture is especially well-exposed in Unit D at the Louden Ranch where the generally vertically oriented nodules sometimes coalesce to form a continuous horizontal limestone bed. Vadoids In a few beds coated grains with irregular, laminated coatings are found in micrites with clotted textures, and are therefore interpreted as vadoids. Other horizons contain grains with clay halos (pseudo ooids) (Figure 25) developed during soil formation (Brewer, 1964).

PAGE 74

77 Colored Mudstone The maroon and green-colored mudstone of Units C and D may represent diagenetic alteration of hematite grains (Walker, 1967), but because of associated pedogenic nodules, probably represent paleosols (Kraus et al., 1985). This mudstone is poorly exposed, and no attempt was made to identify horizons, to name the soils or to attempt a climatic interpretation. Diagenesis The most common diagenetic minerals are sparry calcite, sparry quartz and barite. Minor dolomite, pyrite, limonite and hematite are present in some nodules and as alteration of pre-existing minerals. Some of the present clay mineralogy is probably the result of diagenetic processes, but the determination of the extent of clay diagenesis is beyond the scope of this project. The paragenetic sequence seems to be 1) sparry calcite, 2) chert, 3) clear silica, 4) barite. Where pyrite is present, it accompanies the precipitation of barite. Dolomite is found as the final stage mineral in small geodes at the upstream Rock Crossing section (Plate II), where calcite and clear euhedral silica are first and second. Pyrite accompanies the formation of the dolomite rhombs.

PAGE 75

78 A unique mineral, but one characteristic of many early Morrison deposits in the western U.S., is flesh-colored acicular barite. A possible intraformational source for the barium would be barium substituted for potassium in feldspars and illitic clays. Barium could also have been introduced from barite deposits in the upper Cretaceous marine sediments. Barite is the last mineral in the paragenetic sequence, and is often seen filling both horizontal and vertical fractures, some of which are truncated by a Tertiary erosion surface, suggesting movement during the late stages of tectonism. The strontium carbonate, celestite, was reported by Duce (1924). These two minerals have very similar optical properties and so are easily misidentified (Shelley, 1985). The one sample examined with X-ray techniques for this project did not indicate the presence of any strontium.

PAGE 76

CHAPTER IV PALEONTOLOGY Introduction The fossils found in the Morrison Formation in the Purgatoire uplift are useful environmental indicators even though they are neither abundant or diverse. Based on functional morphology, taxonomic relationships and sediment relationships, inferences may be made about the chemistry, temperature, paleotopography and climate cyclicity. These inferences must be weighed carefully however: 1) complete uniformitarianism is not possible due to lack of relationship to extant genera; 2) organisms may be absent if they have not been introduced into that system (Beadle, 1981); 3) taphonomic effects must be evaluated. Presence/absence and visual estimates of relative abundance were reported on representative body fossil samples because the scope of the project and outcrop characteristics prevented detailed quantitative analysis. Thorough analysis of absolute abundance, density and dispersion especially on the

PAGE 77

80 ostracodes and charophytes from selected outcrops could be useful in understanding the microenvironments identified in the system. Evidence for species interaction in predator/prey ratios, nutrient flow and accidental death (the clam killer story) for the Rock Crossing site are discussed in detail elsewhere (Lockley et al. , 1986; Lockley , 1986a ) . Some discussion of primary production is included here. This chapter includes a despcription of the flora and fuana observed and a discussion of the paleoecology (geochemistry, energy , depth, faunal associations etc. ) of each species as determined from literature to use as a basis for developing an environmental model . Plants Algae Charophytes. Chara oogonia (Figure 24) are abundant components of many of the wackestones and grainstones from Unit A and Unit B in the Purgatoire uplift. Four or five species of chara appear to be represented. Alistochara was identified, but identification of the other forms is difficult, pending completion of the Treatise on plants (R. Forester, pers. com. , 1988). Since the external ornament and whorls are well preserved on the oogonia, and stem fragments have been found in a few samples, these plants were probably not transported far prior to deposition.

PAGE 78

81 Usually about 3 em tall, these bushy plants live totally submerged in the photic zone (Ott, 1958), most commonly in water 1 to 8 m in depth (Tappan, 1980). They tolerate a wide range of pH (5.2 to 9.8 although 7 to 8 is characteristic), and alkaline, well oxygenated waters are favored. Although the stems are not always calcified (Burne et al., 1980), chara thrives, and precipitates more calcite, in lakes supersaturated with calcium carbonate. Salinity is only partially a control; Lamprothamnium chara are present today in saline lakes (up to 70%o) in Australia, but are usually less diverse in higher salinities (Burne et al., 1980). These chara are suited to the seasonal changes in water level; they regenerate after desiccation (Burne et al., 1980). Jurassic Alistochara and modern Lamprothamnium are in the same family, and therefore may have some similar environmental requirements or tolerances. Chara are abundant in the Morrison Formation and other species have been catalogued from numerous other localities by Peck (1937, 1956, 1957), Waldschmidt and Leroy (1944), Johnson (1954) and others (Ott, 1958). Oncolites. Micro-oncoids (less than 2 mm in diameter, Peryt, 1983) make up much of the grainstone and packstone of Unit B. In thin section irregular bands can be seen, often coating mud clasts or quartz grain centers (Figure 20); this irregular surface resembles algal and bacterial growth, not abiotic growth through accretion. Some of the grains described as micro-oncoids

PAGE 79

82 may include "reworked carbonate" nodules (cf Dodson et al., 1980a), but probably are fragments or crumbs of larger oncoids such as are found in modern Lake Constance, Germany (Schottle and Muller, 1968) • Larger oncoids up to 3 em across (Figure 26) are locally abundant in distinctive conglomerates and grainstones that are useful marker horizons. These large oncoids are disk-shaped with no apparent centers and appear to have formed on a mud substrate with little rounding or disturbance during growth. Lack of fossil centers, as are common in the Green River Formation (Weiss, 1969), suggests that little shell material was available to be incorporated into the oncoids, further evidence of the lack of a diverse invertebrate fauna (see below). Since they are formed by blue green algae, oncoids grow within the photic zone, generally at or near river mouths. Here algal growth is favored and the periodic turbulence of the soft mud bottom disrupts growth, creating the characteristic disk shapes. In Lake Constance, a temperate marl lake, oncoids grow embedded in loose rippled sediment below mean water level (Shafer and Stapf, 1977) and are most common and largest near the Rhine River inflow. Smooth hard ones pile up like pebbles on shore, forming pebble conglomerate. Oncoids are generated by many of the same organisms that generate stromatolites, and therefore have many of the same requirements, including calcium carbonate saturation of the water, but form where there is sufficient wave action to overturn and round them (Eggleston and Dean, 1976).

PAGE 80

84 Stromatolites. Thin, .1 to .3 m thick, slightly undulating, cryptalgal laminated horizons are common in Unit B in the Purgatoire uplift. Stromatolitic structures form in many environments, but are common in intertidal and supertidal zones. Mucilagenous blue-green algal filaments calcify as they grow, forming micritic layers which alternate with layers of pellet packstone/wackestone deposited during severe storms. A change in micro-environmental geochemical conditions may also cause calcite precipitation. The filaments of algae bind available clasts such as ooids and bioclasts with this calcite (Monty, 1976) in micro-laminated sediments. The resulting mats may be preserved if drying occurs as a result of lowered water level in a tidal flat environment (Brock, 1976) or if the sedimentation rate increases rapidly in an lake (DeDeckker, 1983). The range of pH tolerance for blue-green stromatolitic algae is 4 to 14, with optimum values of 7 to 8.5. These algae can tolerate low oxygen levels (Tappan, 1980) and high salinities (DeDeckker, 1983) and can survive periods of desiccation. Study of growth forms in Shark Bay, on the coast of Australia indicate that low lying stromatolites develop in lower energy environments than do strongly undulating, columnar ones (Hoffman, 1976). In the hardwater lakes of New York, however, form appears also to be related to lake size, depth, slope, and calcium carbonate saturation of the water.

PAGE 81

85 Thrombolitic fabrics (microscopic clots, Mont y , 1976) are characteristic of certain modern lacustrine stromatolites, including those from Green Lake, New York; Mono Lake, California; and Lake Tanganyika, Africa but have not been recognized in Morrison Formation sediments. Other Plants Eguiseteum. Impressio ns of thin, unbranched, ribbed stems with nodes several centimeters apart are found only at the Rock Crossing site, in a single bed interpreted as a marsh (Lockley et al., 1986). Related to modern horsetails Eguiseteum, these plants had a horizontal rhizome as the main support, and thin roots branching off of it (Bold, 1967) indicating a source of groundwater near the surface. Petrified wood. Silicified conifer logs with "typical Morrison preservation" (Tidwell, 1987) are common in upper Morrison sandstones at Two Buttes, Colorado. Thin sections were made from this material, but the cell structure was not preserved well enough to allow identification of the plant species or their age. Petrified wood has also been reported in the Purgatoire uplift (Taylor, 1974), but none was recovered b y this author. Carbonaceous fragments. Small (less than 1 mm long), black carbonaceous fragments present on bedding planes of some mudstones are interpreted as plant debris. A few conglomerates have yielded larger (1 em diameter) pieces of possible plant

PAGE 82

86 material but no identifications have been attempted due to the fragmented condition of the specimens. This l ithology could be an excellent source of pollen for biostratigraphic studies. Root Traces Two types of lithified root casts have been identified in the study area (Figure 27). At Rock Crossing a horizontal root network spreads out through a clay drape just above footprint bed C, indicating a high water table, possibly marshy conditions at that time. Nodules -o-Q-o Oo-1 • ! .' , , r I ' Vertical Root Traces Figure 27. Vertical root traces indicate a lower water table, horizontal root traces indicate a near surface water table, and coalescing nodules were probably casts of larger roots.

PAGE 83

B7 More common are thin, 1 mm diameter and up to 10 em long, downwardly branching tubules filled with sparry calcite and barite. In the micrite of footprint bed B and packstone of footprint be d Cat Rock Crossing (Lockley tal., 1986) 145 holes were counted in a square 20 em by 2 0 em (3625 per square meter). 8900 per square meter were found in cryptalgal laminites at The Gap. The size, density and d o wnward branching suggest that all were made by a small plant with a simple root system reaching for a somewhat deeper water table. Silicified carbonate nodules at several horizons within Unit B and D are possible rhizocretio ns , casts of the roots of larger plants. Limited exposures and samples of these nodules prevent pos i t i ve identification how e v e r . Discussion Plant fossils are traditionally considered scarce in the Morrison Formation either because of original low productivity, oxidation of plant material (Dodson, et al., 1980a) or some combination of both factors. Factors which limit plant growth include the availability of sunlight, water and nutrients. Productivity might also have been controlled by groundwater chemistry when Units A and B were being deposited. Plants are generally less tolerant of alkaline than saline (Beadle, 1981) water chemistries in many African lakes today. Excess salinity is also a limiting factor at modern Lake Amboceli where the rising water table in the dry lake basin brings salt with it, killin g the trees (Western and Van Praet, 1973).

PAGE 84

88 It is more l'ikely, given the evidence for large numbers of herbiverous dinosaurs in the area of the Purgatoire uplift during the time of deposition of Unit A and Unit B that there was abundant plant life in the vicinity. Eolian deposits are also lacking in thi s part of t h e section, a less reliable indicator that plants were available to stabilize exposed loose topsoil. Oxidizing conditions which prevent preservation of plant material might have been caused by either a hot, arid climate (Dodson et al., 1980b), the lack of thick stands of high-growing plants around the marshes (Beadle, 1981) or by well swamps (Coleman, 1966). It has also been suggested that the dinosaurs ate all of the plants before they could be preserved. This explanation could have credence on a local level, as for example, at Lake Amboceli, where the elephants consume all available foliage in an area during a drought (Western and Van Praet, 1973) when hunger kills more animals than thirst. Significant Morrison flora from several sites i n the western U.S. include con ifer trees and shrubs, shrubby cycads, shrubby and herbaceous ferns, herbaceous lycopsids and sphenopsids (Tidwell, 1975; McAlester, 1977; Seward, 1933; Dunbar and Waage, 1969). Although there i s some speculation that primitive angiosperms, probably trees or shrubs, had begun to develop (McAlester, 1977) there is no fossil evidence to support this. It is likely that shrubby ferns and cycadedoids provided the major groundcover (Dunbar and Waage, 1969), although one might speculate that some hardy, drought resistant plants,

PAGE 85

89 possibly similar to gnetophytes might have also been available to stabilize loose sediment. Well developed domal stromatolites form a prominant ledge in Cimarron R iver Valley of Oklaho m a and New Mexico (West, 1978; Neuhauser et al, 1987) but are not present in the Purgatoire uplift. Leaf and stem impressions have been recovered from siltstones just above the stromatolites. The presence of crocodilian remains and tracks (Conrad et al. , 1987) in associated clastic sediments suggests that this environment was very different from that exposed in the Purgatoire uplift, possibly with a strong fluvial influence. Invertebrates Crustaceans Ostracodes. Ostracodes are the most abundant and diverse fauna (Figure 21, 27) in the sections measured in the Purgatoire uplift. Modern non-marine ostracodes are generally detritus filter feeders that swim several centimeters above a relatively firm substrate in low to moderately turbulent water, although some may live interstially within the sediment and groundwater aquifers. Fresh to brackish water forms generally have smooth shells which can be classified by life mode. "S wimmers " are usually high relative to length, and are able to tolerate seasonal temperature and salinity variations. "Groundwater"

PAGE 86

90 forms live interstitially or in small ponds and littoral zones of lakes and are more likely to be small and robust (Benson, 1961). Different genera within a species may have different chemical tolerances, but most cyprids and darwinulids are freshwater forms (< 3 ppm salinity) and most cythrids tolerate more brackish water (3 to 30 ppm salinity) (Van Morkhoven, 1962). Many ostracodes can live in relatively stagnant water with low oxygen levels but prefer to lay their eggs in oxygenated water. Darwinula and Theriosynoecum brood their eggs, but other species produce eggs which can survive desiccation up to 20 years (Brasier, 1980). Selected ostracodes from this project were examined by R. M. Forester, U.S.G.S., a specialist in Cenozoic non-marine ostracodes. He described the specimens in terms of possible environments based on comparison of morphology and taxonomy with modern forms. In samples from Goat Ranch, The Gap and Rock Crossing sections (Plate II) Forester identified fresh to slightly brackish forms Darwinula (a small groundwater form), a small groundwater form (Candona type), a large swimming cyprid (Midlocypris type), an unknown small swimming form, and a brackish water form Theriosynoecum (a larger cythrid, groundwater form), A modern assemblage like this might be found in slightly saline waters with more Ca+2, and less HC03-, dominated by Na+ or Mg+2; S04-2 or Cl-ions (Forester, 1986), such as modern Australian lakes. Here ostracode diversity is greater in the

PAGE 87

91 fresher lakes and less in the large closed basins with extensive internal drainage area (DeDeckker, 1983). This assemblage could also tolerate water which was at times slightly alkaline (enriched in HC03-), which might be more similar to African lakes, where oxygen content also seems to be a limiting factor (Cohen, 1986). Most of the ostracodes have well-preserved external ornament and are frequently preserved as articulated, spar-filled individuals. These whole carapaces are possible indications of high sedimentation rate which covered the animals before the valves could be opened by weakening of muscles or by predators. Beds with disarticulated valves may have been deposited in times of low sedimentation rate (Brasier, 1980). Thin ostracode coquinites are present in Round Corral and The Gap sections (Plate II). It is probable that they are winnowed concentrations of shells, such as are common just below wave base in large African lakes today (Cohen, in press). These could also represent "blooms", or times of high productivity in the lake, or less likely, anoxic conditions could help preserve the abundant ostracodes (Murphy and Wilkinson, 1980). A large collection.of Morrison ostracodes collected by Wedel and Hallet-Nadeau during 1930's and 1940's from northern Oklahoma, New Mexico and southern Colorado was reported by Peck (Peck and Reker, 1948). The author has been unable to locate the disposition of this collection, which reportedly included some 70 species from the western u.s. and would be a valuable

PAGE 88

92 contribution to our knowledge of the Morrison formation in the area. Conchostrachans. The genera Nestoria was identified by Chen Pei-ji of Nanjing, China, from photographs (Figure 28) of specimens collected at Goat Ranch. Original material has been forwarded to him from Rock Crossing and The Gap, and the State Line section in northwestern Oklahoma. These "clam shrimp" have chitinous shells and occupy a variety of environments including large playa lakes and marshes, and more rarely, littoral regions of lakes and small permanent ponds. They prosper in waters with 7 to 9.7 pH, with a thermal tolerance of 4 to 30 degrees Centigrade. Their eggs must pass through a dry state of torpidity and the juveniles come out of diapause when moisture, photoperiod, oxygen, alkalinity and salt concentration are optimal. Some specimens have been known to survive this desiccatation up to 15 years. Conchostrachans have a total life span between 2 weeks and 4 months. These slow-moving invertebrates have no defense against predators (insect larvae and fish) so favor habitats where they will be sheltered, or the predators absent (Tasch and Zimmerman, 1961; Webb, 1979). Some modern conchostrachans prefer clear ponds, while others prefer alkaline, muddy water. Because they are often found in ephemeral environments, conchostrachans seem to be adapted for living in conditions of fluctuating salinity and alkalinity (Pennack, 1978).

PAGE 89

94 Preservation of the chitinous carapaces of the conchostrachans indicates that these delicate shells were not abraded. Although lack of abrasion does not necessarily mean lack of transport, these microfossils are usually found in relative abundance within a given bed, and it is assumed that the majority of these animals lived near where their shells were buried. Molluscs Bivalves and gastropods apparently have very limited distribution within the study area, having only been collected from a few horizons at Rock Crossing. It is possible, but not likely, that the lack of molluscs in other measured sections in the study area is an artifact of the exposures or oversight by the researcher. Over 500 samples have been taken from a dozen sections during this study and each has been examined in detail. Also, none of the abundant oncoids or ooids have shell fragment centers common where shell material is available, suggesting that there truly was some limiting factor(s) responsible for the limited distribution of the molluscs. Bivalves. Bivalves have been found in only two beds exposed at Rock Crossing. A few silicified disarticulated unionid valves have been collected from the basal portion of footprint bed 1 (Lockley et al., 1986). Articulated, spar-filled unionids (Figure 29) are found in place, associated with dinosaur tracks and plant

PAGE 90

96 stem impressions (Footprint bed 1, Lockley et al., 1986), apparently trampled to death in a marshy mudflat by the large vertebrates. All modern species of the genus Unio are found in fresh water. Each species has different requirements, but most prefer clean, oxygenated, shallow (2 to 7 m), calcium-rich (20 ppm Ca) (Burky, 1983) water with pH greater than 7 (5.6 to 8.3 total range). They live infaunally in rivers and lakes with stable substrate, good food source and warm temperatures (Hanley, 1976), and require permanent water habitats with current for dispersal of larvae, and fish to host the glochidial larvae stage. Although permanent habitat is required for reproduction, a few North American unionid species are reported to have remained alive in moist mud for several months (Pennack, 1978). Bivalves are reported (Stose, 1913) from limestones along the Huerfano River. These outcrops are 90 km from the Purgatoire uplift, and appear to be in a different depositional basin based on available subsurface data (see Chapter VII, Regional Paleogeography). Gastropods. Several speci mens of small gastropods of the genus Lymnaea, Gyraulus and Amplovalvata were recovered from the calcareous shales between footprint beds 2 and 3 at Rock Crossing (Lockley et al., 1986; Plate II). Spar filled voids with shapes reminiscent of larger gastropods were noted in the top bed exposed at Rock Crossing a few meters from the dinosaur bone

PAGE 91

97 discussed below. Other small spar-filled voids were noted in thin sections from the southern sections, but none were distinctive enough to be called gastropods with confidence. Modern pulmonates such as Lymnaea and Gyraulus live in water less than 3 m deep on rooted, emergent vegetation, (Pennak, 1978). Lymnaeids can live in waters with reduced oxygen concentration by breathing at the air-water interface, or breathing oxygen directly from the water through their skins (Hanley, 1976). Valvata is the closest living relative of the gill bearing prosobranch Amplovalvata found in the study area. Although different species have vaying requirements, many Valvata prefer alkaline waters with a pH of 7.3 to 8.4 and carbonate concentrations of 80 to 250 ppm (Harman, 1974) . Quiet water with aquatic vegetation (Hanley, 1976) provides the most suitable environment. Their amphibious strategy may have evolved to compensate for periodically lower oxygen (Aldridge, 1983). Gastropods are sensitive to the total ionic composition of the ambient water, and salinity tolerances change as other ions are added. Many can live in higher salinities if the water is calcium rich, so it is difficult to predict chemistry based on the presence of a certain species. They are restricted somewhat in highly alkaline water, but four species live today in Lake Tanganyika (Cohen and Thouin, 1987) and eight species in Lake Turkana (Cohen, 1984). The Turkana species are often thin-shelled, however, because the water is deficient in calcium.

PAGE 92

98 Valvata scabrita was reported (Stose, 1913) from limestones 30 m below the Cretaceous sandstones along the Huerfano River near the boundary between Pueblo and Huerfano Counties. These outcrops are 90 km from the Purgatoire uplift, and appear to be in a different depositional basin based on available subsurface data (see Chapter VII, Regional Paleogeography). Invertebrate Trace Fossils Thin sandstones of Units C and D locally contain abundant invertebrate burrows. Several of the more distinct traces were identified by Dr. Kent Chamberlain (consulting geologist) as Skolithos, Scoyena and Planolites (Figure 30). These wandering .1 to .5 em diameter trails occur both horizontally in thin mud drapes and vertically from the top of well-sorted planar cross-bedded sandstones in Unit D. Numerous random burrows .5 to 1 em in diameter riddle the tops of the sandy limestones of Unit C. Lacking the distinct structure for taxonomic identification, they appear to have been made by invertebrates mining thin clay pockets and clay drapes in the limestone. Random burrows 1 mm in diameter and several centimeters long appear infrequently in the more friable, laminated calcareous clayshales of Units A and B. These burrows are often filled with calcite spar and medium-sized quartz and bioclastic

PAGE 93

99 grains. They are most common in beds with abundant ostracodes suggesting small crustaceans a s possible trace makers. Scoyenia Skollthu• Large Planolltee Small burrow• Figure 30 . Invertebrate trace fossils identified in the thin sandstones of Unit C and Unit D include Skolithus, Scoyenia, and Planolites. Tiny traces and large, unidentifiable burrows are common in the sandy limestones of Unit C .

PAGE 94

100 Vertebrates Fish Fish fragments from the study area have been tentatively identified by Gerald R. Smith and Ralph Stearley of the University of Michigan Museum of Paleontology. Material collected from Rock Crossing includes scales from a new species of semionotoid (related to to modern gar, Lepidosteus) and an unidentifiable amioid (related to the modern bowfin, Amia) . Only scales, external spines and isolated vertebrae and -cephalic bones have been found from the fish. Lack of other material appears to be preservational bias. The frequent association of abundant scales with storm surfaces indicates the fish probably lived adjacent to, but not along with, the micro-organisms. Semiontids probably were slow swimmers, feeding on plants and molluscs on the bottom sediments (Norman, 1963) . Modern gar can tolerate brackish water and both groups are potential air breathers with divided swim bladders; they often surface to gulp air (Norman, 1963). The male Bow-fin of North America makes a nest in the swampy ends of lakes near abundant plants then guards the eggs and newborn young (Norman, 1963). It should be noted that although Callovian fish localities are relatively abundant in the Western U. S. in the Todilto and Wanakah Formations, fish remains are rare in the Morrison Formation. Fish are reported from Como Ridge, Wyoming,

PAGE 95

101 in possible lacustrine and fluvial facies and from Fruita in southwestern Colorado with other aquatic vertebrates (Dodson et al., 1980; Schaeffer and Patterson, 1984). Lack of proper preservational environment is probably more a factor than lack of fish, at least in the Purgatoire uplift, since scales are very abundant at some horizons. Dinosaurs Sauropod hind limb and pes bones were found in a micritic wackestone at Rock Crossing (Figure 31). Although there may have been some post-mortem transport/scattering of the animal remains, these large bones were fractured in-place after burial, suggesting trampling (Beherensmeyer and Boaz, 1980; Western, 1980). The presence of abundant dinosaur tracks in the bed (Lockley, et al., 1986) supports this interpretation. Abundant small, unidentifiable fragments are found in micrite and sandstone near the top of Unit B and in Unit D at various localities near The Gap. The micrites on Bresden Trail also contain rounded quartz pebbles and cobbles with clay coatings and other suggestions of modification by pedogenic processes. Bone scrap-bearing conglomerate and sandstone are usually thin, cross-bedded and laterally discontinuous; they appear to be small channels on the dry lake bed.

PAGE 96

103 Miscellaneous bone fragments have been reported i n the area by ranchers (Taylor, 1974). Morosaurus, Diplodocus and brontosaur bones were collected at the turn of the century (Stanton, 1905) from the Purgatoire uplift. Bone quarries have been active at Canon City and Cimarron County, Oklahoma, and have produced a wide variety of dinosaur fossils. Dinosaurs and their paleoecology from these and other Morrison quarries throughout the western U. S. are discussed in detail by Dodson et al., (1980b). Dinosaur Tracks Over 1300 tracks forming 100 trackways have been mapped in four footprint-bearing beds at the Rock Crossing site (Lockley et al., 1986, Figure 3, 7, 8, and 9). These probably represent sauropods, ornithopods (camptosaur and/or iguanodontid) and large and small theropods. These tracks and trackways and paleoecologic interpretation based on them are discussed in detail in several publications of the UCD dinosaur research team (Lockley et al., 1986; Lockley, 1986a, 1986b). Relatively fresh waters in tropical Africa have more large vertebrate trails than do alkaline waters (Laporte and Behrensmeyer, 1980). The abundance of tracks in several layers suggest that the dinosaurs returned repeatedly to the lake at the Rock Crossing site, which was fresh enough then to attract the large vertebrates.

PAGE 97

104 Other Morrison Vertebrates It should be noted that other Morrison bone localities have yielded more diverse fauna in limited numbers including crocodiles, turtles, and small mammal bones. These are most often recovered from the fluvial floodplain lithofacies A and B of Dodson et al. (1980b). These lithofacies correspond somewhat to facies of Unit D. This unit is very poorly exposed in the study area and was not thoroughly investigated in this research. The oxidized mudstones (cf. Dodson et al., 1980b) appear to be well exposed in the Johnny Branch drainage near the Bent Las Animas county line. Thorough investigation of this outcrop and others in the Smith Canyon and Johnny Branch drainages might increase knowledge of Morrison biota in southeastern Colorado.

PAGE 98

CHAPTER V GEOCHEMISTRY Introduction As indicated in the preceeding chapter, there is good evidence for the lacustrine nature of the lower Morrison Formation exposed in the Purgatoire Uplift. Determining the geochemistry of the system is not a simple process however. The predominance of limestone and presence of ooids and salt casts suggest deposition in other than a freshwater lake, but diagenesis has altered the calcite and other minerals, obscuring the original mineralogy and evidence concerning the ionic composition of the lake waters. Indirectly, the possible requirements of the biota and characteristics of the basin help suggest what the lake waters might have been like, but are in no way conclusive (Table 1). Modern lakes are classified geochemically as fresh, hardwater and saline and/or alkaline. Salinity is defined, based on biological tolerance, as greater than 5 ppm total dissolved solids (Beadle, 1981). Alkalinity (or carbonate alkalinity) is defined as the sum of all titratable weak acids, which is assumed to be [HC03-] + [C03--] in meq/1 in freshwaters with a pH of

PAGE 99

Feature Salinity Alkalinity Brine Type Permanent/ Depth Wave Action pH Eh Oxygen (sedimentary structure/ Ephemeral biota/mineralo Ooids alight alight high Mg/Ca permanent shallow strong Glauberite crystal casts high low Na(Ca)S04Cl ephemeral? Gypsum crystal casts moderate low MgNaCaS04Cl ephemeral? Char a high tol. alight "high" Cn ephemeral tol. 1-8m alight-moderate 7-8 Oncoids high tol. Ca cement shallow moderate 7-8 Stromatolites high tol. Ca cement ephemeral tol. slight 7-8 Ostracodes moderate not likely CaNaMgS04Cl Conchostrachan s ephemeral 7-9 low Bivalves Ca 20 ppm permanent 2-7 m slight-moderate 6-8 Gastropods slight tol. 'slight Ca 80-250ppm 3 m 7-8 low tol. Fish slight tol. permanent shallow low tol. Missing pyrite oxidizing Missing zeolites low TABLE 1. Environmental Parameters for Biotic and Abiotic Features. (Tol. indicates a tolerance of given paramenter. Discussion and references in text)

PAGE 100

107 6.5 to 8.5. Several measures of carbonate hardness are used in the literature, including the European designation as the weight of CaC03 precipitate derived from carbonates in solution, the same value used by these workers to express alkalinity. Other hardness values are determined by measuring the amount of CaO or CaC03 per mg or per mole per liter (Kelts and Hsu, 1978). Critical range of calcium concentration for most fauna appears to be .25 to .50.meq/l or 5 to 10 mg/1 (Beadle, 1981), although many groups may thrive at higher concentrations, for example, 47 mg/1 in hardwater Lake Constance, Germany. Saline/alkaline waters are further classified according to the dominant anions and cations (Eugster and Hardie, 1978). The major cations are Na+, Ca+2, Mg+2, and K+, the major anions Cl-, S04-2 and HC03-. Alkaline lakes are those in which HC03-is the dominant anion, and are usually lower in Ca+2 and Mg+2 concentrations, and may be high in Na+, Cl-and S04-2. As evaporation occurs and concentrations increase, precipitation occurs along three major pathways (Figure 32), depending on the original composition of the water. As ions are removed from the system by precipitation, the remaining ions are effectively concentrated, moving along the pathway until the next precipitation branchpoint is reached (Eugster and Jones, 1979).

PAGE 101

108 In all cases low magnesium calcite (less than 5 mole % magnesium in the calcite lattice) is the first precipitate, followed by gypsum or magnesium calcite, protodolomite or magnesite. The most concentrated brines precipitate halite and a variety of other minerals, depending on the pathway and on associated trace elements. In this chapter parameters which can be measured in the lab will be discussed first, followed by less direct means of estimating the chemistry of this system, and the confidence level of these methods. Flow diagram for brine evolution. Solid rectangles represent critical precipitates ; rectangles with dashed borders are typical water co mpositions . Final brine types together with examples of salt lakes are surrounded by dash-dot rectangles. All paths are numbered and referred to in the text. Mg-silicate precipitation and 504 reduction are possible for most paths . Saline MgS(\CI ' BRINE i BasQue L . . Ho t L I ' HCC1>> C o•Mq I : HCO:l C o + M u I ' HCCl;] • Co+Mg NaC<1 so4c1 BRINE . Alkal 1 Valley M ono L lilA Alkaline Figure 32. The evolution of chemical brines through evaporative concentration and precipitation in saline and alkaline lakes. (Taken from Figure 5, Eugster, H. P. and Hardie, L. A. 1978 , Saline Lakes in Lakes: Chemistry, Geology, Physics: Springer-Verlag, p. 237-294)

PAGE 102

109 Direct Evidence Mineralogy/Chemically Precipitated Sediments Calcium carbonates. Most beds i n Units A and B contain at least some calcium carbonate either as sparry calcite cement, biogenic or abiogenic grains. Major mechanisms for calcium carbonate precipitation in lacustrine s ystems include: 1) evaporative concentration, where evaporation e xceeds inflow, usually in arid climates and/or closed basins; 2) loss of gases, especially C02, caused by agitation of the water (beachroc k formation by waves, etc.) or biochemical respiration; 3) mixing of waters with different chemistries from streams or springs; 4) supersaturation due to change in water temperature; 5) biochemical precipitation of shells of invertebrates and support structures in plants. In hardwater systems the carbonate phase precipitated usually is low magnes ium calcite (Wilkinson et al. , 1980). However, in lakes with high concentrations of dissolved ions, high magnesium calcite, aragonite or proto-dolomite may form . The p hase which precipitates depends on a variety of factors including temperature, (Schottle and Muller, 1968), the pressure of C 02 at the site of crystal nucleation (Given and Wilkinson, 1985), the magnesium/calcium ratio (Goldsmith et al., 1961) and presence of other ions in the water (Kelts and Hsu, 1968). See Table 2 for comparison of the chemistries of some modern lakes and their precipitates.

PAGE 103

Lake Reference Na K Ca Mg C03+HC03 Cl Tanganyika Beadle 2 meq• .9 . 5 4 Africa Cohen 55 mg 29 12 45 Great Salt Hutchinson 33 mg 2 .2 3 .1 55 Lake, Utah Eugster and 63,600 Hardie ppm Turkana, Beadle 35 meq .5 .3 .3 25 14 Africa Lake Higgins Wilkins on Minnesota Jones/Bowen .3 meq .5 15 .6 17 Pyramid Lake Eugster and 1630 134 10 113 1390 Nevada Hardie ppm Lake Chad Beadle .5 meq . 2 .6 .3 2 0 Africa Lake Constance Muller Germany August 3 mg? 47 8 146 3 March 3 47 7 165 4 TABLE 2 . Water Chemistry of Some Modern Carbonate Producing Lakes (* Units in Na column apply to all analyses in that row) 504 Mg/Ca Brine Type Water Type Ooids Comments .2 high Na-Mg C03 mildly Mg-calcite Deep, windy, perennial alkaline "marine" biota 7 high Na(Mg)Cl saline aragonite (III B) 4-1 low NaC03 alkaline Moderate to high waves hardwater oncoids-low CaC03 calcite Sediments 50% CaC03 high Na(Mg)ClC03 alkaline aragonite Mixing of Ca-rich spring (lake) (III) 4-6 & calcite water (low Mg/Ca) .l NaCaCl slightly iron saline 37 low CaC03 hardwater oncoids-41 low calcite

PAGE 104

111 The present mineralogy of six different samples from Rock Crossing and the Purgatoire uplift was determined by X-ray diffraction (analysis by S. Sutley, U.S.G.S.) following the method of Goldsmith and Graf (1961) for determination of magnesium substitution in the calcite lattice. All samples now are composed of low magnesium calcite as high magnesium calcite is an unstable phase, stabilizing to low magnesium calcite within 50,000 years in incongurent dissolution when exposed to fresh water (Halley, 1985). Clays. The second most abundant mineral group present in the Morrison lacustrine system is the clay minerals. One sample from 3m below the chert bed at Goat Ranch (Plate II), and one sample from Unit A, 3 m above the chert bed at Round Corral, were analyzed by x-ray diffraction by K. Houck, U.C.D, and one sample from Unit B at The Gap, about 20 m above the chert bed was analyzed by C. Turner-Peterson, U.S.G.S. The Bell Ranch equivalent yielded 70% illite and discrete illite and 30% smectite, as did the Unit A sample. The Unit B sample was all smectite. Although they were not analyzed, thin clay layers in both Units A and B in the Black Hills monocline and in Unit B at Rock Crossing had the physical swelling and slick paper-shale qualities indicative of smectite (C. Turner-Peterson, 1987). The montmorillonitic mudstones of the Morrison Formation may be alteration products of volcanic ash (Dawson, 1970); ash falls are reported from the San Juan Basin (Turner-Peterson,

PAGE 105

112 1985), Canon City, Colorado, (Brady, 1969) and Clayton, New Mexico (Mankin, 1958). In the Brushy Basin Member in the San Juan basin, ash was altered in a zoned alkaline/saline playa lake setting with smectite on the mud flat and zeolites and potassium feldspar in the inner zones (Turner-Peterson, 1985). The Unit B sample was a few meters below a sandy limestone bed which contains possible volcanic glass shards. Although no other indication of volcanic ash was identified, it is likely that several ash falls are preserved in Units A and B in the study area. Other clays could be the result of flocculation of suspended material from surrounding basin, in quiet water or enhanced by saline waters (Sly, 1978). Clay deposition increases and carbonate deposition decreases during times of high lake level in Lake Tanganyika (Cohen and Thouin, 1987). Calcite production and organic carbon production increase when nutrient supply is higher due to volcanic activity in the surrounding area. Possible sources for these clays are the weathering of silicate minerals, diagenetic alteration of silicate and carbonate minerals and soil processes such as produced the kaolinitic clays in northeastern New Mexico (Mankin, 1958).

PAGE 106

113 Indirect evidence Sedimentary Structures Salt casts. Individual crystal casts are preserved in Unit B in the Purgatoire uplift. None of the original minerals remain. The largest molds occur along permiability barriers in thin sandstones at the Rock Crossing Section (Lockley et al., 1986, Figure 6). The majority of occurrences, however , are randomly distributed in a micritic matrix and sometimes disrupt bedding. Although possibly originally rhombohedral, these are now distorted and filled variously with barite, calcite and silica. Glauberite (CaS04 : NaS04) is one mineral which forms crystals with similar interfacial angles as those found at Rock Crossing (Lockley et al., 1 986) . Smaller molds are blade or tabular shaped, reminiscent of gypsum blades (Figure 25). Gypsum forms early along several pathways from calcium-rich saline water. Glauberite forms today from sodium enriched saline brines, and can form as a reaction of precursor gypsum with sodium rich water (Eugster and Hardie, 1975). Without further evidence it is possible to suggest only that these crystals precipitated during the evolution of a brine enriched in Na-(Ca) -S04-Cl or Mg-Na(Ca) -S04-Cl (Eugster and Hardie, 1978). Tentative identification of the mineral glauberite implies a sodium enrichment difficult to support from available evidence, and may therefore be misleading.

PAGE 107

114 Ooids. Ooids similar to the ones found in Units A and B (Chapter III, Sedimentology) are presently forming in lakes which are saline and/or alkaline, some with high magnesium/calcium .ratios and all with strong wind-generated wave action. Cortical structure determined by the presence or absence of a pseudo-uniaxial cross suggests that most of the Morrison Formation ooids were originally calcite, although those from Rock Crossing might have been aragonite. Modern lacustrine ooids are reported from only four localities: Lake Tanganyika, Africa, magnesium calcite (Cohen and Thouin, 1987); Great Salt Lake, Utah, aragonite (Halley, 1977); Pyramid Lake , Nevada, calcite and aragonite (Popp and W ilkinson, 1983); and Lake Malinou, U.S.S.R. (Vital, 1948) . "Ool ites" reported from Higg ins Lake (Wilkinson et al., 1980) are calcite, do not show uniform cortical structures and do show strong biotic influence in formation (B. Wilkinson, pers. comm., 1987) and so are discussed as oncoids in Chapter IV, Paleontology. Table 2 gives chemical and physical data for these lakes. The chemistry is slightly different in each lake, but all have elevated ionic concentrations: high magnesium to calcium ratio, mildly alkaline/saline water and strong surf action in Tanganyika; high magnesium to c&lcium ratio, hypersaline waters and strong surf action in Great Salt Lake; and mixing of bicarbonate enriched lake waters with calcium enriched spring waters in a zone of turbulence in Pyramid Lake . It should be noted (Kelts and Hsu, 1978) that often a rise in

PAGE 108

115 magnesium/calcium ratio follows a rise in salinity, therefore ooid production may indicate more saline water. Although ooids have been formed in artificial enviro nments in less than a month, modern lacustrine ooids have been dated at around 1000 years (Popp and Wilkinson, 1983). The ooids from Lake Tanganyika are in the process of being dated now (A. Cohen pers. comm. , 1987 ) and it is likely that these ooids will also indicate some stability in the lake basin. Desiccation features. Desiccation features (mud cracks) are not chemical indicators in themselves, but when combined with other evidence such as disruption of bedding by crystal growth, suggest that evaporative concentration might have influenced carbonate deposition during some phases of lake development (Eugster and Hardie, 1978) . Absent Minerals Sometimes negative evidence is the strongest. Units A and B in the study area lack early authigenic pyrite, zeolites, and dolomite, clues to original geochemistry. The lack of early authigenic pyrite and organic plant material suggests that the depositional and early diagenetic environments were oxidizing, not reducing. The Morrison environment in the Purgatoire uplift was therefore oxygenated, probably without stratification and anoxic highly saline mudflats.

PAGE 109

Zeolites are a common alteration of volcanic ash in strongly alkaline/saline lakes. Their absence , and the preservation of unaltered volcanic shards and ash, are an indication that the system was not strongly alkaline. 116 None of the carbonate in Units A or B appears to be dolomite. Protodolomite often precipitates from saline waters, especially if the magnesium/calcium ratio is high. Since most dolomitization occurs as a replacement process of aragonite or high-magnesium calcite (Zenger and Dunham, 1980), the lack of dolomite may only hint at original low magnesium calcite, in relatively fresh waters. The sandy limestones of Unit C have a crystalline calcite matrix that is sometimes partially dolomitized. This recrystallization could have happened in late-stage co ncentration of brines as the lake system was drying u p , with possible groundwater or airborne ash contributing ions. It is more likely, however, that replacement occurred much later. A careful petrographic study of a number of these beds might help resolve the uncertainty associated with timing of dolomitization. Biota Limiting factors. A relatively "high" calciu m content is suggested b y the presence of calcite shell b uilders, although it is difficult to quantify because of the c omplex interactions of the biotic and chemical components . Chara probably produced thicker calcareou s stems and egg cases because of "high" calcium

PAGE 110

117 content, which in turn contribut ed significantly to the p roduction of the micrites. Nonbiogenic carbonate precipitation was also enhanced b y respiration of abundant oncoid-forming algae. Some of t he biot a present also have high salinity tolerances. General distribution. Alkali ne lakes often restrict nlants more tha n an imals (Beadle, 1981) . Although lack o f plant fossils co u l d be due t o poor preservation i n oxidizing enviro nment, (see C. Turner-Peterson, 1985 for review) slightly alkali ne conditions might also have p l a yed a part. The predominance of biotic elements that require high calcium c oncentrations suggest that the lake system was at least "hard", and that it possibly was also slightly alkaline. Basin Analysis Pre-cursor basin. The Bell Ranch equivalent gypsum was deposited in a playa complex after partial erosion of t he underlyin g Entrada Formation. Thinly bed d ed crystalline limestone near the top is composed of secondary calcite crystal s which occur a s hexagonal prisms with doubly terminated crystals l n a clay matrix (Figure 33). They are oriented parallel to bedding and many h a v e a shadow in t he cert:er. Granule-sized nodules oi pink and white chert appear as rep lacemen t m inerals i n clays, and some gypsum nodules are present. The density and size of t h e chert increase u pward to a horiz o n

PAGE 111

119 red chalcedony and chert about 20 em thick intergrown with pyrite, barite, calcite, and clear quartz • The calcite crystals, disseminated chert and gypsum nodules continue 50 em above the solid chert horizon in a non-calcareous clayshale matrix. The mechanism for formation of the crystalline limestone and chert bed is unclear. The thin bedding style interrupted by authigenic crystals suggests that the clay and fine sand grain matrix was deposited in a saline lake. Unknown crystals nucleated around some clastic component or authigenic mineral which grew from brines in the clay matrix (M. Longman, pers. comm., 1987). The calcite crystals which are clearly untransported were formed as an early diagenetic replacement of this precursor, and later replaced by silica, barite and pyrite. Crystals with a strong resemblance to these are forming by unknown processes in alkaline saline lakes in Africa's rift (Hay, 1976). This similarity in morphology and depositional style is not an indicator of original brine chemistry, but does lead to the conclusion that the water was not "fresh". The "welded chert band" near Canon City, Colorado, was mentioned by Fredrickson et al. (1956) who described replacement of aragonite and suggested that groundwater replacement of CaC03 concretions was the mechanism. This same horizon and locality were examined in thin section and described in detail by King and Merriam (1969). Although their descriptions are very similar to those of the Purgatoire uplift, the doubly terminated hexagonal

PAGE 112

120 crystals are quartz, and aragonite inclusions have been identified in some of the large calcite crystals. The pink chert is identified as the mineral "beekite". At this locality, the chert nodules begin 1 m above the last gypsum, and continue for 3 m to the top of the massive chert bed. They propose five stages of diagenesis: 1) calcareous cement, 2) growth of the authigenic quartz around quartz grains, 3) beekite, 4) chert, 5) calcite and barite. Devitrification of tuffaceous materials in the shale is suggested as source of the silica. Although they did not describe the "agate beds" in detail, Lucas et al. (1985) report finding root casts and concretions associatd with the siliceous beds at Bull Canyon in east-central New Mexico, suggesting pedogenic processes for their formation. Farther north along the Rio Cimarron, however, root casts were not obvious, and samples collected by this author appear very similar (in regard to relationships of gypsum, chalcedony, sandy clay and small calcite crystals) to those of the Purgatoire uplift which do not show fabrics typical of silcretes (Goudie, 1973). Muehlberger et al. (1961) propose soil development in volcanic ash in northeastern New Mexico. Calichified tuffs at Lake Amboceli in Africa also contain zeolites (Hay and Reeder, 1978), although these carbonatite ashes are probably not good analogies for Morrison ash falls. Similar chert nodules from the Morrison Formation in southern Wyoming near Como have been described as "magadi-type cherts" (Surdam et al., 1972) formed in an alkaline/saline lake.

PAGE 113

121 This analysis is based upon surface textures of the nodules, and the paper does not report on the associated lithologies and mineralogies, so it is difficult to make comparisons w ith material from the study area. The "agate bed" or "chert band" problem is an interesting one. The presence of siliceous bands near the base of the Morrison has been suggested as a significant stratigraphic marker since the time of Lee (1901) (See K i ng and Merriam, 1969 for review of reported occurrences). If these bands do represent a similar period of time, then some mechanism has to be called upon to explain the widespread distribution of silica and chemistry for the precipitation of the distinctive cherts. If volcanic ash in the sediments is the source of silica as suggested (King and Merriam, 1969), then groundwater has to become alkaline regionally to mobilize the silica. At least in the Purgatoire uplift and immediately adjacent areas it requires a big change to go from a gypsum precipitating water system to one which would mobilize silica. Whatever the mechanism, it might be one that would tie in with more global changes, and it is a problem that deserves attenti on beyond the scope of this study. The beds immediately above, however, are non-calcareous clayshales, an obvious change in chemistry and depositional environment. These first Morrison beds have millimeter-size laminations of very fine-grained quartz and feldspar, abundant mica, some plant debris and thin reddish lenses. A total of 2 m

PAGE 114

122 thick and devoid of fauna except for a single ostracode, these beds probably are fresh-water, nearshore deposits (VanHouten, 1964). Reminiscent of the Olduvai Gorge sediments described by Hay (1976), they could, however, also have resulted from precipitation in a chemically stratified lake. The first abundant biota occur with ooids, evidence that carbonate precipitating lacustrine deposition was established, and are present about 4 m above the last bedded gypsum. Source of ions. It is common in modern lakes for ion-bearing waters to enter the system as sub-aqueous spring inflow as well as surface streams, and for the waters to be of different compositions. Surface water entering the early Morrison lake system in southern Colorado might have come from the east and southeast (see Chapter VII, Paleogeography) where it was probably draining exposed sandstones and shales, sources of HC03and variable cations. Some surface drainage might also have entered from the southwest where some Precambrian igneous and metamorphic rocks would also supply HC03-, Si02, Na+ and Ca+2, as well as limited amounts of Mg+2 and Fe+2. Because the evidence points to fluctuating water levels, possibly related to a seasonally varying climate, this surface input probably was not constant. Because the outline of the basin has not been determined, it is difficult to suggest what lithologies would have been host to springs and groundwater. The underlying sandy , shaly,

PAGE 115

123 gypsiferous Bell Ranch equivalent, although not showing signs of wholesale dissolution and brecciation, almost certainly contributed some Ca+2 and S04ions as well as additional HC03ions to the lake system, probably through baseflow. This was possibly the most constant inflow into the lake system. Airborne ionic input in the form of rainwater and rhyolitic ash falls probably contributed some Na+, Ca+2, Mg+2, K+, HC03-, Cland S04-2 (Eugster and Hardie, 1978). Both sources probably were sporadic, and had varying degrees of effect on the system. Effects of climate, drainage. Inferences of a seasonal climate point to variation in chemical composition in the lake system, especially if the basin was hydrologically closed at times. Biotic distribution and sedimentologic structures do support the idea that there was variation in the chemical composition. The ramifications of this are discussed more fully in Chapter VII, Paleogeography. Conclusions Mixing was proposed. by Lockley et al. (1986) for precipitation of the micrites of Unit B at the Rock Crossing dinosaur track site because of the ooids in the micrites and salt casts along permeability barriers. There is also evidence for evaporative concentration, especially during times of maximum desiccation. Biochemically induced precipitation probably was

PAGE 116

124 most important, especially since calcium carbonate-rich units are also the ones with maximum preservation of blue-green and green algae. Most evidence indicates that the waters were generally calcium-rich (but probably not hardwater), and slightly saline, especially during times of maximum desiccation. Any alkaline influence, sodium enrichment or high magnesium/calcium ratios were temporary, and probably the result of spring water or airborne ash input. Although the environment was generally oxidizing, the waters were probably sluggish and not well-oxygenated except when agitated during stormy, windier intervals.

PAGE 117

CHAPTER VI DEPOSITIONAL ENVIRONMENTS Introduction The Morrison Formation in the study area is characterized by a low diversity biota, most abundant in Units A and B, which lived in a carbonate producing lacustrine system. Thickly laminated and thinly bedded sediments predominate; there are few thinly sediments which indicates the lake was probably not often chemically or thermally stratified. The cyclic chemical precipitates and biotic assemblages which might have either tolerated or required saline and infrequent slightly alkaline environments (Table 1) suggest that fluctuating water chemistry was an important factor in the ecosystem. The sedimentary structures in these units are evidence of shallow water and periodic exposure in an arid environment with seasonally high moisture input during most of lacustrine deposition. Figure 35 is a legend for the drawtngs in this chapter. Sluggish ephemeral rivers have been suggested as a likely depositional environment for similar sediments at other places in the Morrison Formation, but the abundance of ooids and other

PAGE 118

!ill Llmeatone Q Gypeum D .olomlte Chert [J 8andetone U Slltatone g Mudetone j: -1 Clayehale Caloareoua Clerehale Ripple• "••er Ripple• M11doracke Claetlc Dlkae Conwolutlone Nodule a Crretal Caete Cia eta 126 (} D lnoeau r Traoke SS BloturbatloR Root Caate .:J' Dlnoeaur lone Fhh Debrh e ..... e. Oetracodee Conchoetracflane Plant Fragment• Char a Algal Mat Oncolde Pellet a . 0 Oolde Volcanic Shard• Figure 35 . Legend for cross sections and paleo-environmental interpretations. All biota and most sedimentary structures identified in the Purgatoire uplift measured sections are depicted on this diagram.

PAGE 119

equant coated grains and lack of fluvial sandstones in this interval indicate a lacustrine setting for Units A and B. 127 Clastic infill of the lake in a possibly more restricted basin is represented by the sandy limestones and dolomites of Unit C, which contains only limited trace fossils. Rare dinosaur bone and petrified wood fragments from Unit D fluvial floodplain sandstones and mudstones indicate complete filling of the lake basin. Contacts between these four units are gradational. The facies represented probably are in part lateral equivalents, and although some lithologies appear to vary across an outcrop, poor exposures and subsurface data make it difficult to establish lateral relationships over any distance. Depositional Environments The generalized environments shown in Figures 36 and 37 and listed below were developed by combining lithofacies and biofacies. Some facies overprint others, such as mudcracks and salt casts disrupting a laminated nearshore sequence. More specific interpretations of the environments in three examples from Units A and B are developed by integrating information about the biota and sedimentary structures found in the previous chapters.

PAGE 121

IILL ftANCH IOUIYALINT eJ L • . • F W 1 t 1 i • I • lftdiiOftl I , colore d mwd• t•n•• PLAYA L AKE COWPLf:X z .. . c w .. . 0 w c " w .. • z 0 c ,; z z 0 w . . 0 . .. .. w 0 .. 0 c .. c z w w .. • . . UftTOPONLKJJDI A • .. Q c H I 0 I ••• lit • lt41 • •+'"" •• ,., c •• ,.,, r"• ••• c.t>! -------roo 0 ... , ... 128 '

PAGE 122

129 Although these facies represent some of the environments included by Dodson et al. (1980b) in their four Morrison lithofacies (channel sands, oxidized floodplain, reduced marshy flood plain and lacustrine), direct comparison is not attempted here. There are some obvious differences from the environments they studied (presence of evaporites, for example). This study also differs in scope as it attempts a more detailed environmental reconstruction of one area instead of the regional synthesis of Dodson et al. (1980b). Perennial Lake Facies A. Chara, ostracodes, fish debris, ooids, oncoids and bivalve fragments occur in grainy lenses. The ooids suggest some turbulence as well as permanence and the bivalves require permanent water habitats. When ooids were being produced the lake probably also had elevated ion concentrations. Bedding was preserved because of high sedimentation rate, or reworking of the sediments by wave activity prior to preservation (Units A and B) although some of these units are thoroughly bioturbated. Nearshore -Alternating Energy Conditions, Occasional Exposure: Facies B. Ripple-interlaminated to thinly interbedded micrite and packstone or grainstone is the most abundant lithology in Units A and B. The micrite is usually clay-rich, dense and homogenous, although it sometimes contains carbonized plant fragments, mica and wispy laminations. The grainy laminations contain combinations of very fine to fine

PAGE 123

130 siliciclastic grains, skeletal components, micro-oncoids, micrite coated, well-rounded medium-sized quartz grains, and micrite clasts. The skeletal components are most commonly . ostracodes, both disarticulated carapaces and complete spar filled shells, with some layers restricted to adult forms, and other layers containing both juveniles and adults. Chara oogonia are best preserved in layers where ostracodes are present. Micro-oncoids are medium to coarse sand size; occasionally oncoids up to granule size are present in this facies. Bedding is sometimes slightly disturbed by bioturbation or rooting (Facies B-1), or disrupted by mudcracks, crystal casts or clastic dikes (Facies B-2) (Units A and B). Facies C. Dense thick micrite with spar and packstone/grainstone patches randomly distributed throughout is probably a variation of Facies B which has been bioturbated or dinoturbated (Unit B). Facies D. Thin (1 to 2 em thick), rippled quartzose sandstones have rare salt casts on the upper surface. Loading structures into the soft sediment below suggest that these siliciclastic grains represent times of increased sediment input into the lake (Units A and B). Nearshore -Alternating Energy Conditions, Fresh Water Facies E. This facies is similar in appearance to Facies A and B. Flaser ripples of very fine-grained feldspathic quartz sand grains with minor micrite suggest an increase in

PAGE 124

131 availability of siliciclastic grains accompanied by a decrease in ionic content during increased freshwater input from rivers. Although there are no bioclasts, tiny distinct burrows indicate the waters were hospitable to some infaunal organisms (Units A and B). Nearshore -Low Energy Conditions Facies F. Clay rich micrite or wackestone with abundant ostracodes, mica and plant debris floating in the matrix occur in very thin beds between grainier layers. Homogenous texture of this facies is probably caused by extensive bioturbation (Units A and B). Facies G. Calcareous clayshale with thin, discontinuous calcareous lenses contains gastropods, chara, and ostracodes. This facies has been identified only from one horizon at the Rock Crossing site (Unit B). Shoreline -Low Energy: Facies H. Thinly laminated micrite and grainstone that have a "crinkly" appearance and minor "birdseye" fabric are probably cryptalgal laminites created by blue-green algae trapping and binding sediments during times of greater sedimentation. The low relief of these structures indicates that they developed in low energy and were preserved by sediment cover as the water level rose. This facies often has numerous small root traces which produce a distinctive texture on the weathered surface. (Unit B)

PAGE 125

132 Shoreline -Periodic High Energy, Storms Facies I. Lenses of carbonate packstone with abundant oncoid and fish debris sometimes have large oscillation ripples and are often capped with mudcracked clay drapes. The grains range from sand to cobble size and are randomly distributed in the poorly sorted packstone which was rapidly deposited in a waning current (Unit B). Facies J. Grainstone composed of imbricated flat algal pebbles, pellets and coated grains, usually fine upward. Although poorly sorted, ranging to cobble-sized grains, there is no fine sediment between the grains which indicates slight washing in a high energy environment (Unit B). Beaches, Shoals Facies K. Sandy limestone and dolomite have minor relict cross bedding and rare volcanic shards. Top surfaces are frequently riddled with centimeter diameter sub-horizontal burrows. The matrix supported nature of these suggests lacustrine origin for these facies. They were apparently deposited in a high energy environment, followed by a time of quiescence and/or exposure which allowed burrowing of the upper surface (Unit C). Ephemeral Pond Facies L. Dark calcareous shales with abundant conchostrachans on bedding planes are present locally in the sequence. The conchostrachans probably were deposited in shallow

PAGE 126

133 ponds at a time of low water level in the lake because their fragile chitinous shells would not be preserved in a high energy environment. This facies could also represent deposition in a quiet, shallow portion of the main lake, especially near abundant plant cover (Unit A and B). (Note: The report in 1986 by Lockley et al. of conchostrachans in the high energy shoreline facies at Rock Crossing was an error.) Marsh, Paludal Facies M. Micrite with sub-conchoidal fracture and nodular texture are common and suggest lower overall energy, possibly bioturbation by roots in an oxygenated paludal environment. A unique marsh environment is represented by the in-situ unioni d bivalves and Eguisitium impressions which occur in wackestones at Rock Crossing (Lockley et. al, 1986). Reducing swamp conditions are largely absent, although there is limited preservation of carbonized plant debris in the upper portion of Unit B (Unit A, B). Mudflat Facies N. The thin lenticular laminations of silty and very fine sand-sized grains of quartz or reworked carbonate, mudcracked and disrupted by salt crystal casts of this facies may have formed in ponded water on a mudflat. This facies may include some beds assigned to Facies B (Unit B).

PAGE 127

134 Delta and Delta Front Facies 0. Some of the thin sheet sandstones have the geometries of foreset and topset beds of thin Gilbert-type deltas. They lack coarse sediments, however, and may also be stream sands (see Facies Q below). (Unit B C) Floodplain Complex-overbank ponds, soils, fluvial channels Facies P. The lenses of cross-bedded conglomerate less than .3 m thick, fine upward from a basal lag of bone fragments and rounded siliciclastic pebbles and were deposited in small ephemeral streams on exposed mudflat/lake bottoms. These are similar in composition to Facies I, but lack the upper rippled surface. Some of the material in the lag may have been concentrated on the dry surface by deflation, but the typical sand lens does not appear to be eolian in nature. (Unit B) Facies Q. Thin sheet sandstones of fine-to medium-grained quartzose sand were formed as bars of ephemeral streams. These beds display a variety of sedimentary structures including current laminations of very fine-grained opaque minerals, horizontal trails along the upper surface and occasionally climbing ripples (Unit B, C). Facies R. Although not often exposed, sand lenses less than . 3 m thick and 2 m wide encased in dull red and green mudstones probably are a large percentage of the deposits of Unit D. They resulted from deposition by suspended load streams (Unit D).

PAGE 128

Facies S. Fine-to medium-grained brown weathering sandstones with shallow cross beds and minor clay and chert pebble lag were deposited in braided streams (Unit D). 135 Facies T. Red and green mudstones formed by oxidation of floodplain sediments (Unit C, D). Facies U. Nodular micrite, geodes, micrite with clotted texture, abundant spar and quartz filled fractures are ubiquitous. In Units A and B they are associated with drab colored sediments and represent marsh or soil horizons. In C and D they are most often assoicated with deep red and green mudstones and resulted from pedogenic processes on the exposed floodplain (Units A, B, C and D). Ash Fall Facies V. Papery-parting clayshales "sticky" to the tongue, waxy and micaceous are smectitic, possibly bentonitic. In beds less than .1 m thick, these probably represent ash falls in relatively fresh water. An alternate explanation for these clays could be that they settled out of suspension from streams as the water moved into a quiet environment (Units A, B, C and D) Sub-units of the Morrison Formation Unit A Non-calcareous clayshales lacking biotic diversity and subaerial exposure features predominate in the lowest Morrison beds, Un. i t A of this report. The similarity of these sediments

PAGE 129

136 to those of the "detrital cycles" of the Triassic Lockatong Formation of the eastern U. S. (Van Houten, 1964) suggest they were deposited in a fresh-water lacustrine system. Less than 30% of this unit consists of flaser-rippled and laminated packstone and micrite of a slightly saline lake. The grainy lenses contain ooids, bioclasts and abundant fine-grained quartz. The facies packages in this unit are relatively thick: 1 m or more, indicating a stable environment. Evidence of subaerial exposure such as salt casts, mud cracks, or large ripple marks is not present. Often poorly exposed, Unit A appears at the base of the Morrison Formation throughout much of the study area. Although varying in thickness, it provides evidence for one large lake of similar dimensions to the precursor gypsum basin persisting for a relatively long period of time. At Goat Ranch, the exposure is continuous from the chert bed and crystalline limestone and clayshale (the contact with the Bell Ranch Equivalent) to the lowermost Morrison Formation. Figure 38 is an interpretive log of Unit A, which is 10 m thick here. The lack of biota in the lower noncalcareous clayshale beds might be explained merely by the fact that it takes time for flora and fauna to be distributed to a newly formed lake. Although ostracodes and micro-oncoids become abundant, stromatolites and large oncoids are absent, possibly because high

PAGE 130

137 :z: ... .. ... ... a:: z ... ... 0 ... ... ::> z a:: 2 0 ... ... .... :z: 0 : ... .. z 0 .. a:: a:: z a:: 0 ... ... 0 ... :z: ... a:: .... 2 .. z "' •
PAGE 131

139 resistant, micritic limestone, which is highly weathered at this locality. Unit B -Black Hills Sections, "OV Lake" Thickly laminated to thinly bedded micrite and packstone with abundant, relatively diverse biota are the most common lithology in this unit. Interbedded with the resistant micrite is thin, usually calcareous clayshale which also contains bioclasts. A complete section of Unit B is exposed at The Gap (Figure 39) which is easy to correlate with other sections near OV Mesa (Round Corral, OV Mesa) using a few key beds (see below). It is about 6 m thick at these exposures on the southern wall of OV Mesa. Most of the interbedded sequence probably resulted from variation in water flowing into the lake, especially from surface streams, because the grainy units contain a large percentage of siliciclastic grains. It is not possible to suggest that these • fluctuations in water level represent climatic changes of a given duration (annual, etc.), but they do appear similar to the alternations described by Hackenson and Jansson (1983). During the dry "season" there is no deposition. The rains bring mineral grains, nutrients and ions into the lake (the grainy layers). After the rains, plants and animals take advantage of the additional nutrient supply to raise productivity which also increases calcium carbonate precipitation (the micritic layers).

PAGE 132

140 .. l: .. < .. "' ... ... 0 .. < ::1 .. "' . :I 0 .. < :z: 0 i .. .. z 0 .. "' "' z ! 0 "' 0 < .. .. < :z: .. "' .. :I • z .. u @) .. ., .. I _ _ , . -liil II IIII I _ o old • c A TromDIIno T--,1-..j: I I p -8 M ..... Lenticular ..... ... . -ll l I I I I 8 I tamlnatlone F .... ..1. I I ... -I I" o I I I I I I N 5 Lenttcu 1i'am '"-• l"-" 8 ..... .... ..... e 4 ...... 1 .. Unlquo ottracodet " e l . l j_: ll I E I I M l J Lenticular 8 1 lamination• . -: ._I 8 ..... . ...... I I I I 3 t. 0 m I I e _l. J . . J II I I I I ThIn bedded . -I . :j--" ll I I II H ..:=!.. Clotted t • • t u r • 1 M 81 2 t-=-r-.... -,=-. Wf!J ThI n bedded e 2 1 Figure 39. Unit B as measured at The Gap contains 6 shallowing upward cycles, each capped with an exposure surface. Four of these surfaces can be traced to other outcrops in the OV Mesa area. The section was measured from the lowest exposed bed of Unit B to the contact with Unit C.

PAGE 133

141 Groundwater input in the form of baseflow is difficult to estimate, but the localized presence of ooids and salt casts suggest ionized spring flow in some areas. These frequent, small scale fluctuations are part of an overall larger scale pattern as seen in the six shallowing upward cycles exposed at The Gap. Although each is different, in general, shallow, sometimes non-calcareous fresh-water clayshale grade up into micrite as the water becomes more saline. These cycles are usually capped with an exposure surface (ripples, mudcracks, crystal casts) followed by a conglomeratic storm bed. The most diverse biota is found in the storm beds, probably due to mixing of elements from several environments during the storm. Some penecontemporaneous deformation structures suggest that the lake did not always reach desiccation stage before renewe d influx of water and sediment began. Only the upper portion of Cycle 1 is exposed. Here Darwinula and limited chara suggest a shallow water bod y of constant salinity. However, as the lake shrank, small gypsum crystals grew in the upper few inches of the drying mud. A thin ash bed caps this cycle. Cycle 2 contains Darwinula and a small cyprid but no chara. The clotted texture of the micrites and the rooted algal mats are indications that this thin cycle represents a marsh deposit. A chara meadow in the shallows of a well-established lake helped shelter a variety of ostracodes including the small

PAGE 134

142 unidentified "swimmer", Darwinula, and Theriosynoecum during Cycle 3. Mudcracks capping an exposed rooted algal mat indicate desiccation prior to the storm event which deposited a thick, rippled bed of fish debris, chara and pebble-sized oncoids, but no ostracodes. The presence of a unique ostracode fauna including a large, unidentified "open water" cyprid and Darwinula signals the opening of Cycle 4, but this assemblage did not persist. Ostracode and chara communities similar to those of Cycle 3 followed, and conchostrachans also enjoyed the protection of the chara meadows at times. A very thin packstone of chara, clay clasts and micrite-coated rounded quartz grains caps this cycle. Cycle 5 was probably deposited in a mudflat environment because there are no fauna and only few siliciclastic flaser ripples in the laminated mudstone. The upper layers of this bed are disrupted by crystal casts and a thin grainstone of clay chips and abundant micrite-coated quartz grains follows. The grainstone has thin clay drapes with a few fragmented conchostrachan carapaces along the bedding plane. Following this mudflat, Cycle 6 contains a very different biotic assemblage. Ostracodes are much less abundant and there are no chara, but fish scales and coalified plant fragments appear on numerous bedding planes. Micrite-coated quartz grains and micro-oncoids are the principal components of the laser ripples within the mudstone. Cycle 6 still appears to represent a shallow lacustrine environment, but possibly nearer the river

PAGE 135

143 inlet into the lake. The capping oncoid conglomerate of this cycle is interbedded with a centimeter thick, rippled sandstone which has abundant plant debris on the upper bedding plane. Some of these clastic beds could have been deposited by streams on the dried up lake bed. The upper .5 m of this unit is friable and not well exposed. Trenching reveals an admixture of wackestone and packstone containing plant debris, micrite-coated grains, oncoids, well-developed ooids and traces of chara and ostracodes. The texture and composition of this bed is very similar to well exposed "dinoturbated" lacustrine beds at the top of the exposure at Rock Crossing, and may represent a similar environment. Fluctuating water level, along with a variety of peripheral environments, produced a mosaic of lithologies, many of which seem to be extremely laterally variable. There are exceptions however. The rooted, algally laminated bed capping Cycle 2, the rippled, mud-cracked storm bed above Cycle 3; the salt-cast disrupted micrite of Cycle 5; and the highly disturbed micrite at the top of the sequence all seem to be present at similar elevations at sections around OV Mesa (Figure 40, 41). These beds are all exposure surfaces, and they appear contemporaneous. It is possible therefore that OV Mesa outcrops define a recurring minimum perimeter of the lake. To the north of OV Mesa at the Purgatoire Canyon section, Unit B is several meters thicker and a larger number of the lithologies represent exposure surfaces. A disturbed micrite at

PAGE 136

144 Figure 40 . Topographic map showing location of measured sections in the Purgatoire uplift. Because these hillsides are badly slumped, all sections are composites of several traverses.

PAGE 137

SOUTH ROUND CORRAL THE GAP OV MESA 4 3 6 l' .. ! . I : ! , . 3 I , 1064m 243m 61ITI 912m 304m Figure 4 1 . Although m ost individual b e d s are laterally discontinu o u s , fou r expo s ure surf aces appear t o correlate over 2 km from t h e R o und Corral section t o t h e OV Mesa section. NORTH 4 , 5

PAGE 138

146 the top of the sequence appears very similar to that at OV Mesa, and they may be the same. No other surfaces seem to correlate because of lateral variation in the depositional environment. Unit B Rock Crossing Section, "Dino Lake" North of Purgatoire Canyon, at Rock Crossing t n e exposures (Figure 42) appear similar to Unit B near OV Mesa, and have approximately the same stratigraphic relationship to Units C, D and the overlying Cretaceous sandstones. The section at Rock Crossing also contains thinly bedded micrite and packstone cyclic sediments with exposure surfaces; the difference between the two localities is in the paleontology. Additional fauna at Rock Crossing includes gastropods, bivalves, more abundant and diverse fish debris, dinosaur tracks and Eguisetium plant impressions. This association indicates a possibly fresher lake than can be documented at OV Mesa. Extensive (greater than 1 km) bedding planes exposed here, have allowed examination of a paleo drainage, ripple mark trends, and dinosaur trackway trends. A compilation of this information (Lockley et al., 1986) indicates a large body of water to the north of this outcrop, away from the deposits to the south at OV Mesa. Figure 42 is modified from Figure 5, Lockley et al (1986). Please note additional fossil data which has become

PAGE 139

147 X ... "' • .... ... a: ... ... Q .. ... • :I z a: -' ... 2 0 < ... ... X ... Q :z: .. "' z .. 0 • a: a: z ... ' 0 '"' 0 • .. u ... • :z: ... .. .. a..,. fJ<>WJ!C: ... 2 • z ... u ... I I )0 I I l l ll Dlno1Yrbo1od • I 1 Q I l L l 82 I . . I . ) 0 I : I I I II Dlnorurbatec .. I I !"11<1 _lii._I.LI. • 5 . l I I 8 1 I I ;-w-_l J 4 .. -. . \ E / 'Q I I I I I ll Trackwar• J A .I.. I fl} 3 ..L a ..1-I I ) Q l I 11 Olnolurbaltd w ..L I I ...I.. Laminated L 2 . . . . : --..-.L -"--L. _I I F •• I I, , ........ 1 I I 8 Figure 42. Section exposed at Rock Crossing at the main trackway site appears correlative to Unit Bat ovMesa. Just upstream of this section, an additional 3 m of section are exposed below this which were not sampled as part of this research.

PAGE 140

148 available since publication of that diagram. Environmental interpretations vary slightly, based in part on the new data, and in part on the wider perspective of this report. Four of the six shallowing upward cycles at Rock Crossing are capped with surfaces exhibiting desiccation features and dinosaur tracks. One cycle is capped with a thin sandstone with crystal casts, and one with a micritic layer with mudcracks. Unit C Unfossiliferous sandy limestone with a few volcanic shards, relict cross-bedding and bioturbated tops, recrystallized sparry calcite and minor dolomite are the characteristic beds of this unit (Figure 43). Thin micrite with mottled and brecciated texture, and cracks filled with calcite and silica are often associated with colored (maroon and green) structureless mottled mudstone. As exposed at The Gap, the beds above the last bioclastic limestone of Unit B are green mudstone and orange and green calcareous sandstone lenses. A possible ash fall is suggested by a thin, non-calcareous smectitic "paper shale" just above the colored sandstone lenses. The limestone above is matrix-supported, about 40% quartz sand, and was deposited in a lacustrine environment. The intimate association with adjacent floodplain deposits suggests nearshore deposition, such as on a beach or bar. They apppear to

PAGE 141

I .... "' < ... z ..... ... a:: ... ... ... < 0 z a:: ..... :> 0 ..... ... 2 ..... :r < 0 ... "' 0 a:: a:: z J: z 0 "' 0 < ... ..... a:: J: ... a:: ... < "' z ... 2 ... )-. II K . ) -'Pur p lo T Oreen Rod ond T green II l' . -'-. _._ II -'-K II m I " "'-._ . . / Rod T Green ::::!"' Orange 0 ( Green T Figure 43 . Unit C sandy limestones at The Gap contain few fossils and appear to represent shoreline and floodplain environments. 149

PAGE 142

150 be discontinuous lenses, but their friable, easily eroded nature makes it difficult to determine the true geometry. Slightly saline waters either at the time of deposition, or introduced, possibly as groundwater, might have induced recrystallization and dolomitization. Magnesium from volcanic ash could have influenced dolomite formation also. The thin brecciated and crystalized micrites formed either from pedogenic processes or in shallow paludal areas. Calichification requires a net moisture loss during the year, suggesting that overall aridity continued even into the upper Morrison deposits. Unit D Colored (red and green) mudstone, and covered intervals which are probably drab to lightly colored mudstone make up the majority of this unit (Figure 44). Brown weathered, porous, medium-grained cross-bedded sandstone with lag of chert and clay pebbles often form a resistant ledge below the massive white Lytle sandstone. Occasional dinosaur bone and wood fragments weather out of the mudstone or thin, discontinuous sandstone lenses. These sediments are interpreted as fluvial flood plain and overbank deposits, with some soil formation and a few thin beds. Because some of the thick, upper sandstones are quite widespread, they probably were deposited as part of a braided stream system.

PAGE 143

Green mudstone R Figure 44. A rare section of Unit D was exposed in Colbert Canyon and appears to contain floodplain deposits. 151

PAGE 144

CHAPTER VII REGIONAL PALEOGEOGRAPHY Southeastern Colorado The lacustrine sequence exposed in the Purgatoire uplift is part of a more extensive system which occupied much of the study area during late Jurassic time. The predominance of clays in the lacustrine facies may be equated to the high gamma count on gamma ray logs (Potter et al., 1980 and Figure 4). The isopach map of the gamma peak at the base of the Morrison Formation correlates to Unit A plus Unit B, and serves as a basis for defining the limits of the lake basin (Figure 45a). Outcrop and subsurface paleontologic and sedimentologic data constrain these limits somewhat (Figure 45b), but control is too sparse to draw a paleogeographic reconstruction with great accuracy. As noted in the previous chapter, Unit B beds have a different biotic composition at Rock Crossing from equivalent beds in the Black Hills monocline sections. It is assumed that these beds are generally time equivalent because of their position in the section relative to the thin medial brown sandstone and the overlying Lytle Sandstone. If they are time equivalent, and if at Rock Crossing the water body was to the

PAGE 145

&c A 0"'"'' . f 111'' C i t , e"' ' 11l!M I I 81 Jfot PIIII C I IlliCit Ill !Ill \ Sv•t..,III C I D11tr ''"''' r '1 153 Figure 45. a. Isopach of Unit A plus Unit B from the high gamma signal indicating lacustrine clays. b. Distribution of sedimentologic features and biota from Morrison deposits (Units A, B, C and D) plotted against important Unit A plus Unit B contours. (Area shown is as in Figure 1)

PAGE 146

154 north, one can then assume that either Unit B represents one large lake with an extensive headland, or two lakes separated by a narrow strip of land. The second is favored because Unit B is thicker at the intermediate Purgatoire Canyon section (Plate II) and the outcrops to the northeast of the study area contain molluscs which are lacking in sections to the south and east. Two exposures to the west and north of the Purgatoire uplift: one at the junction of the Cuchara and Huerfano rivers, the other along the Apishapa River, lack gypsiferous beds and contain abundant gastropods and minor bivalves (Stose, 1913) . Well logs show thinning (and therefore separation of deposition) between Rock Crossing and these distant exposures (Plate I). In general however, the sequence seems to freshen to the west, an observation supported by the abundant bivalves reported in association with the dinosaur quarries at Garden Park near Canon City (J. Hanley, pers. comm., 1986). Going east from sections measured at OV Mesa , along the east wall of the Purgatoire uplift canyons (Briggs Canyon), the gypsiferous beds are slightly thicker and the clay-rich beds still contain lacustrine features similar in all respects to that of The Gap and OV Mesa. Slightly to the north the exposures appear slightly sandier, and the resistant micrites of Unit B are not obvious, possibly indicating a facies change. Throughout Smith Canyon (at the eastern boundary of the Purgatoire uplift), exposures are largely in the upper half of the section. Lacustrine limestone beds are not exposed (Taylor,

PAGE 147

155 1974) and no well logs are available to ascertain the continuity of the lacustrine sequence. Johnny Branch Creek (east of Smith Canyon) has good exposures of thin gypsiferous and cherty beds in contact with the overlying Morrison Formation clayshales. Here, Unit A is very thin, and Unit B, although of similar thickness, is dominated by oolite grainstone interbedded with light purple mudstone. Unit C is thicker here, but Unit D appears very similar to outcrops in the Purgatoire uplift. The same lake probably covered much of the region initially, but as the lake became smaller, the eastern edge was near the present Johnny Branch Creek, even while the upper part of Unit A and the lower part of Unit C were being deposited at the Purgatoire uplift. The few meters of sediment exposed less than 8 m below the Cretaceous strata along Two Butte Creek near the Baca Las Animas county line have been correlated with the Triassic of Oklahoma and New Mexico because of their purplish color. Plate I however shows that the lacustrine units of the Morrison Formation thin eastward from Johnny Branch, and correlate with the County Line section sediments. These light purple sandstones and ostracode-bearing mudstones appear to have been deposited in isolated floodplain ponds at a time possibly equivalent to the period of deposition of Unit B at the Purgatoire uplift. To the north around the Two Buttes Reservoir and the Tertiary Two Buttes uplift no single outcrop exposes more than a small portion of the section. By comparing Sanders' (1934)

PAGE 148

composite section with well logs and isolated outcrops, it appears that the lower Morrison Formation may be truncated entirely. Only a thin sequence resembling the "Brown Silt member" of the Morrison of northeast New Mexico is present. 156 None of the lacustrine beds of Units A and B of the uplift are present here. Abundant conifer logs and the predominance of fluvial sandstones also lead one to the conclusion that this area had some relief relative to the lake basins of the Purgatoire uplift. Northern perimeters of the lacustrine system are difficult to delineate because there are no outcrops in the southern Denver basin. One sample set from an oil well near the northeastern edge of the study area contained a trace of molluscan debris in the lacustrine interval. The only other available clues about the system in this direction are t he general isopach trends reported by Merriam (1955) in eastern Kansas . These isopachs include everything between the redbeds and the Cretaceous and reflect the clastic component of the entire sequence. The western edge of the thickest clastic deposits in Kansas corresponds quite well with a High'' line inferred througheastern Colorado (Figure 45b). Both in Kansas and Colorado the sediments become more clay rich west of this line. To the southeast, well logs show thinning of the clay-rich interval. Outcrops of the lower Morrison Formation in northwestern Oklahoma and northeastern New Mexico are

PAGE 149

157 distinguished by well-developed domal stromatolites which indicate high energy. Leaf and stem impressions occur in thin, silty sandstones capping these stromatolites. The presence of crocodile remains and tracks (Conrad et al., 1987) corraborate the evidence that these deposits represent a different basin than OV Lake, possibly a fluvial system. Figure 46 is one possible paleogeographic reconstruction which incorporates the isopach data with paleoecologic and sedimentologic evidence obtained from studying outcrop and subsurface samples. Basin Development By middle to late Jurassic time the eolian dune field of the Entrada Sandstone in the Purgatoire uplift region had been largely eroded and the interdunal areas filled by fluvial conglomerates and sandstones. A gypsum precipitating playa formed which covered much of the central portion of the study area. Gypsum precipitation ceased, and the next deposits of this basin were clay-rich sandstones which were recrystalized later to calcite, chert and chalcedony. The first deposits of the Morrison Formation were largely unfossiliferous, thinly laminated non-calcareous clayshales which were deposited in a fresh water lake with increased fine-grained clastic input.

PAGE 150

N DINO LAKE OV LAKE Z / Figure 46 . Paleogeographic reconstruction of southeastern Colorado during t he time of deposition of dinosaur trackway bearing horizons at Rock Crossing.

PAGE 151

159 Cycles of calcium carbonate precipitation occurred however, with evidence that chemically-dominated lacustrine conditions prevailed: ooids, micro-oncoids, chara and ostracodes. Although this lake was shallow and well-oxygented, the lake was large and relatively stable. There is some evidence for fluctuating water levels, but few shoreline or exposure features are present. During the midlife of the lake, fluctuations in size and water level created a cyclic sedimentary pattern. The basin may have separated into two or more lakes with different chemical and biotic characteristics. Shorelines moved back and forth frequently across the region, and at times when the water was fresh in Dino Lake to the north, the lake was host to a relatively diverse community of plants, small invertebrates and large vertebrates. Chemical processes were still important during the last stages of the lake, but as clastic input increased, base level rose. Crystalline limestone and dolomite with fine-grained quartz sand deposits document the filling of the lake basin. Fluvial sands and muds accompanied by paleosol development and scattered thin overbank ponds predominated throughout the region during the last-portion of Morrison deposition.

PAGE 152

160 Comparison to Other Lake Systems During late Jurassic the estimated paleolatitude of the Purgatoire uplift was about 20 to 30 degrees north (Hallam, 1973, Steiner, 1975). Late Jurassic global temperatures were generally warmer than today (Hallam, 1973), and at these subtropical latitudes there was probably no large seasonal change in temperature. Dune fields, followed by thick bedded gypsum immediately below the Morrison were deposited in an arid environment, which continued intermittently in the Late Jurassic. The carbonate minerals of the lower Morrison Formation precipitated in a slightly saline lake in a sometimes hydrologically closed basin where evaporation only slightly exceeded inflow (Eugster and Hardie, 1978). This balance could easily swing toward fresher water during wet times and slightly more alkaline during dry spells (Beadle, 1981; Van Houten, 1964). Variation in moisture regime probably played a larger role in producing the cycles than tectonics, although slight changes in the rate of subsidence could have also been a factor. Even as base level rose to subsidence during deposition of the upper Morrison Formation fluvial floodplain, the presence of soil caliches suggest net arid conditions at times (Dodson et al, 1980b). The best modern analogies for these sedimentary processes and ecosystems would be those of the sub tropical and tropical regions, especialy where seasonal climate controls the moisture

PAGE 153

161 regime: the belt including parts of Africa, China, South America, North America. and Australia. Most published studies are available for African, U. S. and Australian lakes, and these are therefore easiest to compare with the Morrison Formation. Suggested models for the Morrison Formation include the wet and dry extremes of Gran Chaco and the high runoff and suspended load of Hwang Ho and Yangtsee River basins (Mook, 1916; Dodson et al., 1980b). This comparison was made based on data collected at the large dinosaur bone quarries, and is prejudiced toward the more fluvial sequences. Analogies for the lacustrine facies have been made to the desert lakes of Kenya and the Okevango delta in Africa because of moisture extremes and the predominance of large terrestrial vertebrates (R. Bakker, pers. comm., 1987; A. Cohen, pers. comm., 1987). Although some Morrison lakes probably had chemistries similar to the African soda lakes and the U. S. basin and range playas (Turner-Peterson, 1985), OV and Dino lakes were probably more like the fluctuating perennial saline lakes (Reeves, 1968) of arid Australia (DeDeckker, 1983) and southwestern United States (Anderson and Kirkland, 1969) both in chemistry and in tectonic stability. Evidence in the Purgatoire uplift region is for alternating wet and dry climate extremes and for lakes that at times were large, lasted for many years, and appear to have occupied the same basin for much of Morrison lacustrine deposition.

PAGE 154

REFERENCES American Stratigraphic Company, 1973, Sample examination and logging manual: 49 p. Aldridge, D. W., 1983, Physiological ecology of freshwater prosobranchs, Russell-Hunter, W. D., ed., The Mollusca, Volume 6, Ecology: Orlando, Florida, Academic Press, Inc., p. 330-359. Anderson, R. Y., and Kirkland, D. W., 1960, Origin, varves, and cycles of Jurassic Todilto Formation, New Mexico: American Association of Petroleum Geologists Bulletin, v. 44, p. 37-52. Anderson, R. Y., and Kirkland, D. W., 1969, Paleoecology of an early Pleistocene lake on the High Plains of Texas, Anderson, R. Y. and Kirkland, D. W., eds., Geological Society of America Memoir 113: 215 p. Bachman, G. 0., 1953, Geology of a part of northwestern Mora County, New Mexico: U. S. Geological Survey Oil and Gas Investigation Map 137, Bakker, R. T., 1986, The Dinosaur Heresies: New York, William Morrow and Co. Inc., 481 p. Baldwin, B., and Muehlberger, W. R., 1959, Geologic Studies of Union County, New Mexico: New Mexico Bureau of Mines Bulletin 63, 171 p. Bass, N. W., Straub, C. E., and Woodbury, H. 0., 1947, Structure contour map of the surface rocks of the Model Anticline, Las Animas County, Colorado: U. S. Geological Survey Oil and Gas Investigations Preliminary Map 68, scale 1:42,240. Beadle, L. C., 1981, The inland waters of tropical Africa: New York, Longman, 475 p.

PAGE 155

163 Behrensmeyer, A. K., and Boaz, D. E . D., 1980, The recent bones of Amboseli Park, Kenya, in relation to East African paleoecology, Behrensmeyer, A. K., and Hill, A. P., eds., Fossils in the making: Chicago, Illinois, University of Chicago Press, p. 72-92. Behrensmeyer, A. K., and Boaz, D. D., 1981, Late Pleistocene geology and paleontology of Amboseli National Park, Kenya: Paleoecology of Africa, v. 13, p. 175-188. Benson, R. H., 1961, Ostracoda, Teichert, C., ed., Treatise on invertebrate paleontology, Part Q, Arthropods 3. Berman, A. E., Poleschook, D., and Dimelow, T. E., 1980, Jurassic and Cretaceous systems of Colorado: Colorado Geology, Rocky Mountain Association of Geologists, Guidebook, p. 111-128. Bold, H. C., 1967, Morphology of plants: New York, Harper & Row, Publishers, Incorporated, 541 p. Bowman, S. A. B., Bowman, J. T., and Drake, R. E., 1986, Interpretation of the Morrison Formation as a time-transgressive unit: North American Paleontological Convention Abstracts with Programs, v. 4, p. AS. Brady, L. L., 1969, Stratigraphy and petrology of the Morrison Formation (Jurassic) of the Canon City, Colorado, area: Journal of Sedimentary Petrology, v. 39, p. 632-648. Brasier, M.D., 1980, Microfossils: Boston, George Allen & Unwin, 193 p. Brewer, R., 1964, Fabric and mineral analysis of soils: New York, John Wiley and Sons, 470 p. Brock, T. D., 1976, Environmental microbiology of living stromatolites, Walter, M. R., ed., Stromatolites: New York, Elsevier Scientific Publishing Company, p. 141-148. Brown, R. E., and Wilkinson, B. H., 1981 , The Draney Limestone: Early Cretaceous lacustrine carbonate deposition in western Wyoming and southeastern Idaho: University of Wyoming Contributions to Geology, v. 20, p. 23-31. Burbank, W. S., 1930, Revision of geologic structure and stratigraphy in the Ouray District of Colorado, and its bearing on ore deposition: Proceedings of the Colorado Scientific Society, v. 12, p. 175-180.

PAGE 156

164 Burky, A. J., 1983, Physiological ecology of freshwater bivalves, Russell-Hunter, W. D., ed., The Mollusca, Volume 6, Ecology: Orlando, Florida, Academic Press, Inc., p. 281-329. Burne, R. V., Bauld, J., and DeDeckker, P., 1980, Saline lake charophytes and their geological significance: Journal of Sedimentary Petrology, v. 50, p. 281-293. Cohen, A. S., 1982, Paleoenvironments of root casts from the Koobi Fora Formation, Kenya: Journal of Sedimentary Petrology, v. 52, p. 401-414. Cohen, A. S., 1984, Effect of zoobenthic standing crop on laminae preservation in tropical lake sediment, Lake Turkana, East Africa: Journal of Paleontology, v. 58, p. 499-510. Cohen, A. S., 1986, Distribution and faunal associations of benthic invertebrates at Lake Turkana, Kenya: Hydrobiologia, v. 141, p. 179-197. Cohen, A. S., Ferguson, D. S., Gram, P.M., Hubler, S. L., and Sims, K. W., 1986, The distribution of coarse-grained sediments in modern Lake Turkana, Kenya: implications for clastic sedimentation models of rift lakes, Frostick, L. E. et al., eds., Sedimentation in the African rifts: Geological Society Special Publication No. 25, p. 127-139. Cohen, A. S., and Thouin, C., 1987, Nearshore carbonate deposits in Lake Tanganyika: Geology, v. 15, p. 414-418. Cohen, A. S., in press, Fossil ostracodes from Lake Mobutu (L. Albert): paleoecologic and taphonomic implications: Coleman, J. M., 1966, Ecological changes in a massive fresh-water clay sequence: Transactions-Gulf Coast Association of Geological Societies, v. 16, p. 159-174. Collinson, J. D., 1978a, Deserts, H. G. Reading, ed., Sedimentary Environments and Facies: New York, Elsevier, p. 80-96. Collinson, J.D., 1978b, Lakes, H. G. Reading, ed., Sedimentary environments and facies: New York, Elsevier, p. 61-79. Colton, R. B., Anderson, L. W., and Holligan, J. A., 1975, Preliminary map of Landslide deposits, La Junta 1 X 2 Quadrangle, Colorado and Kansas: U. S. Geological Survey Miscellaneous Field Studies Map 708, scale 1:250,000.

PAGE 157

165 Condon, S.M., and Peterson, F., 1986, Stratigraphy of Middle and Upper Jurassic rocks of the San Juan basin: historical perspective, current ideas, and remaining problems, Turner-Peterson, C. E., Santos, E. S. and Fishman, N. S., eds., A basin analysis case study: the Morrison Formation Grants Uranium region New Mexico, American Association of Petroleum Geologists Studies in Geology #22: p. 7-26. Conrad, K., Lockley, M.G., and Prince, N. K., 1987, Triassic and Jurassic vertebrate-dominated trace fossil assemblages of the Cimarron Valley region: implications for paleoecology and biostratigraphy: Northeastern New Mexico, New Mexico Geological Society, 38th Field Conference, Guidebook, p. 127-138. Craig, L. C., Holmes, C. N., Cadigan, R. A., Freeman, V. 1., Mullens, T. E., and Weir, G. W., 1955, Stratigraphy of Morrison and related formations, Colorado Plateau region, a preliminary report: U. S. Geological Survey Bulletin 1009-E, p. 125-168. Darton, N.H., 1905, Discovery of the Comanche Formation in southeastern colorado: Science, v. 22, p. 120. Darton, N.H., 1906, Geology and underground waters of the Arkansas Valley in eastern Colorado: U. S. Geological Survey Professional Paper 52, 90 p. Dawson, J. C., 1970, The sedimentology and stratigraphy of the Morrison Formation (Upper Jurassic) in northwestern Colorado and northeastern Utah [Ph.D. thesis]: Madison, Wisconson, University of Wisconsin, 125 p. deAlbuquerque, J. S., 1986, Stratigraphy of the Jurassic (Callovian/Oxfordian?) Ralston Creek Formation, south-central Colorado: Geological Society of America Abstracts with Programs, Rocky Mountain Section, v. 18, p. 350. DeDeckker, P., 1983, Australian salt lakes: their history, chemistry, and biota -a review: Hydrobologia, v. 105, p. 231-244. DeVoto, R. H., 1980, Pennsylvanian stratigraphy and history of Colorado: Colorado Geology, Rocky Mountain Association of Geologists, Guidebook, p. 71-101. Dodson, P., Bakker, R. T., Behrensmeyer, A. K., and Mcintosh, J. S., 1980a, Paleoecology of the dinosaur-bearing Morrison Formation: National Geographic Society Research Reports, 1974, p. 145-151.

PAGE 158

166 Dodson, P., Behrensmeyer, A. K., Bakker, R. T., and Mcintosh, J. S., 1980b, Taphonomy and paleoecology of the dinosaur beds of the Jurassic Morrison Formation: Paleobiology, v. 6, p. 208-232. Duce, J. T., 1924, Geology of parts of Las Animas, Otero, and Bent Counties, Colorado: Colorado Geological Survey Bulletin 27 Part III, p. 73-102. Dunbar, C. 0., and Waage, K. M., 1969, Historical geology: New York, John Wiley & Sons, Inc., 556 p. Eggleston, J. R., and Dean, W. E., 1976, Freshwater stromatolitic bioherms in Green Lake, New York, Walter, M. R., ed., Stromatolites: New York, Elsevier Scientific Publishing Company, p. 479-488. Emmons, S. F., Cross, W., and Eldridge, G. H., 1896, Geology of the Denver basin in Colorado, with a section on the Morrison Formation by G. W. Eldridge: U. S. Geological Survey Monograph 27, 556 p. Esteban, M., and Klappa, C. F., 1983, Subaerial exposure, Scholle, P. A., Bebout, D. G. and C. H., eds., Carbonate Depositional Environments: American Association of Petroleum Geologists Memoir 33, p. 1-54. Eugster, H. P., and Hardie, L. A., 1975, Sedimentation in an ancient playa-lake complex: the Wilkins Peak member of the Green River Formation of Wyoming: Geological Society of America Bulletin, v. 86, p. 319-334. Eugster, H. P., and Hardie, L.A., 1978, Saline Lakes, Lerman, A., ed., Lakes: Chemistry, Geology, Physics: New York, Springer-Verlag, p. 237-294. Finlay, G. I., 1916, Geologic Atlas of the U. S., Colorado Springs: U. S. Geological Survey Folio 203, p. 7-8. Forester, R. M., 1986, Determination of the dissolved anion composition of ancient lakes from fossil ostracodes: Geology, v. 14, p. 796-798. Frazier, F., Houck, K., Lockley, M.G., Prince, N., Vest, W., and Corringrato, V., 1983, Interpretations of some depositional environments and paleoecology of a part of the Morrison Formation in SE Colorado: Geological Society of America Abstracts with Programs, v. 15, p. 333.

PAGE 159

167 Frederickson, E. A., DeLay, J. M., and Saylor, W. W., 1956, Ralston Formation of Canon City embayment, Colorado: Bulletin American Association of Petroleum Geologists, v. 40, p. 2120-2148. Freytet, P., 1973, Petrography and paleo-environment of continental carbonate deposits with particular reference to the upper Cretaceous and lower Eocene of Languedoc (southern France): Sedimentary Geology, v. 10, p. 25-60. Gabelman, J. W., 1956, Tectonic history of the Raton basin region: Raton Basin, Rocky Mountain Association of Geologists, Guidebook, p. 35-39. Gilbert, G. K., 1896, Underground water of the Arkansas Valley in eastern Colorado: U. S. Geological Survey Seventeenth Annual Report, Part II, 51 p. Gilluly, J., and Reeside, J. B., 1928, Sedimentary rocks of the San Rafael swell and some adjacent areas in eastern Utah: U. S. Geological Survey Professional Paper 150-D, p. 61-110. Given, R. K., and Wilkinson, B. H., 1985, Kinetic control of morphology, composition and mineralogy of abiotic sedimentary carbonates: Journal of Sedimentary Petrology, v. 55, p. 109-119. Glennie, K. W., 1970, Desert sedimentary environments: New York, Elsevier Publishing Company, 222 p. Goldsmith, J. R., Graf, D. L., and Heard, H. C., 1961, Lattice constants of the calcium-magnesium carbonates: American Mineralogist, v. 46, p. 453-457. Goudie, A., 1973, Duricrusts in tropical and subtropical landscapes: Oxford, U.K., Clarendon Press, Gregory, H. E., 1916, The Navajo Countrya geographic and hydrographi c reconnaissance of parts of Arizona, New Mexico, and Utah: U. S. Geolog ical Survey Water-Supply Paper 380, 161 p. Griggs, R. L., and Read, C. B., 1959, Revisions in stratigraphic nomenclature in Tucumcari-Sabinoso area, northeastern New Mexico: American Association of Petroleum Geologists Bulletin, v. 43, p. 2003-2007. Hakanson, L., and Jansson, M., 1983, Principles of Lake Sedimentology: New York, Springer-Verlag, 316 p. Hallam, A., 1973, Atlas of Paleobiogeography: New York, Elsevier, 531 p.

PAGE 160

168 Halley, R. B., 1977, Ooid fabric and fracture in the Great Salt Lake and the geologic record: Journal of Sedimentary Petrology, v. 47, p. 1099-1120. Halley , R. B., 1985, Carbonate sedimentation and diagenesis: Short Course sponsored by the Rocky Mountain Section of Society of Economic Paleontologists and Mineralogists, 124 p. Hanle y , J. R., 1976, Paleosynecology of nonmarine mollusca from the Green R iver and Wasatch Formations (Eocene), southwestern Wyoming and northwestern Colorado, Scott, R. W., and West, R. R., eds., Structure and classification of paleocommunities: Dowden, Hutchinson and Ross, Inc., p. 235-261. Haniey, J. H., Evanoff, E., and Good, S.C., 1986, Paleobiology of nonmarine mollusks from the Morrison Formation, Western Interior, U. S. an overview: North American Paleontological Convention Abstracts with Programs, v. 4, p. A19. Hardie, L.A., Smoot, J.P., and Eugster, H. P., 1978, Saline lakes and their deposits: a sedimentological approach, Matter, A. and Tucker, M. E., eds., Modern and ancient lake sediments: International Association of Sedimentologists, Special Paper No. 2, p. 7-42. Harman, W. N., 1974, Snails (Mollusca: Gastropoda), Hart, C. W. et al., eds., Pollution ecology of fresh water invertebrates: New York, Academic Press, 389 p. Hay, R. L., 1976, Geology of the Olduvai Gorge: Berkeley, University of California Press, 203 p. Hay, R. L. and Reeder, R. J., 1978,Calcretes of Olduva i Gorge and the Ndolanya beds of northern Tanzania: Sedimentology, v. 25, p. 649-673. Heaton, R. L., 1939, Contribution to Jurassic stratigraphy of Rocky Mountain Region: American Association of Petroleum Geologists Bulletin, v. 23, p. 1153-1177. Hoffman, P., 1976, Stromatolite morphogenesis in Shark Bay, Western Australia, Walter, M. R., ed., Stromatolites: New York, Elsevier Scientific Publishing Company, p. 261-272. Hotton, C. L., 1986, Palynology of the Morrison Formation: North American Pa leontological Convent ion Abstracts with Programs, v. 4, p. A20. Imlay, R. W., 1980, Jurassic paleobiogeography of the conterminous United States in its continental setting: U. S. Geological Survey Professional Paper 1062, 134 p.

PAGE 161

Jackson, T. J., 1979, Lacustrine deltaic deposition of the Jurassic Morrison Formation of north-central Colorado: 169 Northern Front Range and Northwest Denver basin, Colorado, Rocky Mountain Section Geological Society of America Guidebook, p. Johnson, R. B., 1959, Geology of the Huerfano Park area, Huerfano and Custer Counties, Colorado: U. S. Geological Survey Bulletin 1040-0, p. 557-583. Jones, B. F., and Bowser, C. J., 1978, The mineralogy and related chemistry of lake sedi'ments, Lerman, A., ed., Lakes: Chemistry, Geology, Physics: New York, Springer-Verlag, p. 179-236. Kauffman, E. G., 1986, Geological and paleontological site analysis of the Pinon Canyon Maneuver Site, Las Animas County, Colorado, report to the U. S. Army, with a section on Mesozoic Stratigraphy by J. I. Kirkland and J. G. Eaton: Boulder, Colorado, University of Colorado Department of Geological Sciences, 132 p. Kelts, K., and Hsu, K. J., 1978, Freshwater Carbonate Sedimentation, Lerman, A., ed., Lakes: Chemistry, Geology, Physics: New York, Springer-Verlag, p. 295-324. King, P. B., and Beikman, H. M., 1974, Geologic map of the United States exclusive of Alaska and Hawaii: U. S. Geological Survey Map no.G82198, scale: 1:2,500:000. King, R. J., and Merriam, D. F., 1969, Origin of the "Welded Chert", Morrison Formation (Jurassic), Colorado: Geological Society of America Bulletin, v. 80, p. 1141-1148. Klappa, C. F., 1980, Rhizoliths in terrestrial carbonates: cassification, recognition, genesis and significance: Sedimentology, v. 27, p. 613-629. Kowalis, B. J., 1986, Fission track dating of bentonites and bentonitic mudstones from the Morrison Formation, Utah and Colorado: North American Paleontological Convention Abstracts with Programs, v. 4, p. A26. Kraus, M. J., Bown, T. M., Kvale, E. P., and Vondra, C. F., 1985, Upper Jurassic/Lower Cretaceous and Paleogene alluvial sediments of the Bighorn basin, northwest Wyoming: Modern and ancient fluvial systems in the United States, Third International Fluvial Sedimentology Conference, Guidebook, p. 19-44. Landis, E. R., 1956, Road Log-Third Day: Raton Basin, Rocky Mountain Association of Geologists, Guidebook, p. 118-125.

PAGE 162

170 Laporte, L. F., and Behrensmeyer, A . K., 1980, Tracks and substrate reworking by terrestrial vertebrates in Quaternary sedimentsof Kenya: Journal of Sedimentary Petrology, v . 50, p. 1337-1346. Lee, W . T., 1901, The Morrison Formation of southeastern Colorado: Journal of Geology, v. 9, p. 343-352. Lee, W . T., 1902, The Morrison shales of southern Colorado and northern New Mexico: Journal of Geology , v . 10, p. 36-58. Lee, W. T., 1903, Age of the Atlantosaurus Beds: Science, v. 17, p . 292-293. LeRoy, L . W., 1946, Stratigraphy of the Golden-Morrison Area, Jefferson County , Colorado: Colorado School of Mines Quarterly, v. 41, 115 p. Lewis, D. W., 1984, Practical Sedimentology: Stroudsberg, Pennsylvania, Hutchinson Ross Publishing Company, 229 p. Lockley, M. G., 1986a, The paleobiological and paleoenvironmental importance of dinosaur footprints: Palaios, v. 1, p. 37-47. Lockley, M.G., 1986b, Dinosaur tracksites, Colorado University at Denver Geology Department Magazine, special issue no. 1 . A field guide published in conjunction with the First International Symposium on Dinosaur Tracks and Traces: Albuquerque, New Mexico, 56 p. Lockley, M. G., and Conrad, K., 1987, Mesozoic tetrapod tracksites and application in paleoecological census studies, Mesozoic Terrestrial Ecosystems, 4th Symposium: p. 144-149. Lockley, M.G., Houck, K., and Prince, N. K., 1986, North America's largest dinosaur trackway site: Implications for Morrison Formation paleoecology: Geological Society of America Bulletin, v. 97, p. 1163-1176 . Lockley, M.G., Prince, N . K., Houck, K., and Carpenter, K., 1984, Reconstruction of a Late Jurassic lacustrine ecosystem: Geological Society of America Abstracts with Programs, v. 15, p. 228. Long, C. S. ,J., 1966, Basal Cretaceous strata, southeastern Colorado (Ph.D. thesis]: Boulder, Colorado, University of Colorado, 479 p. Lucas, S. G., Keitze, K. K., and Hunt, A. P., 1985, The Jurassic System in east-central New Mexico: Santa Rosa, New Mexico Geological Society, 36th Field Conference, Guidebook, p. 213-242.

PAGE 163

171 MacClary, J. S., 1936, Footprints in Colorado: Life Magazine, v. 1, p. 5. MacClary, J. S., 1938, Dinosaur trails of Purgatory, Scientific American, v. 158, p. 72. Maher, J. C., 1953, Paleozoic history of southeastern Colorado: American Association of Petroleum Geologists Bulletin, v. 37, p. 2475-2489. Mankin, C. J., 1958, Stratigraphy and sedimentary petrology of Jurassic and pre-Graneros Cretaceous rocks, northeastern New Mexico [Ph.D. thesis]: Austin, Texas, Texas University, 231 p. Maughan, E. K., 1980, Permian and lower Triassic geology of Colorado: Colorado Geology, Rocky Mountain Association of Geologists, Guidebook, p. 103-110. McAlester, A. L., 1977, The history of life: Englewood Cliffs, New Jersey, Prentice-Hall, 168 p. McLaughlin, T. G., 1954, Geology and Ground-water resources of Baca County, Colorado: U. S. Geological Survey Water Supply Paper 1256, 232 p. Medwedeff, D. A., and Wilkinson, B. H., 1983, Cortical fabrics in calcite and aragonite ooids, Peryt, T. M., ed., Coated Grains: Berlin, Springer-Verlag, p. 109-114. Merriam, D. F., 1955, The geologic history of Kansas: State Geological Survey of Kansas, Bulletin 162, p. 1-317. Monty, C. L. V., 1976, The origin and development of cryptalgal fabrics, Walter, M. R., ed., Stromatolites: New York, Elsevier Scientific Publishing Company, p. 141-148. Mook, C. C., 1915, Origin and distribution of the Morrison Formation: Geological Society of America Bulletin, v. 26, p. 315-322. Mosher, L. F., 1983, In southwestern canyons, Crump, D. J., ed., America's wild and scenic rivers: National Geographic Society, p. Muehlberger, W. R., Baldwin, B., and Foster, R. W., 1961, High plains, northeastern New Mexico, Raton, Capulin Mountain, Clayton: State Bureau of Mines and Mineral Resources Guidebook to Scenic Trips to the Geologic Past, 101 p.

PAGE 164

172 Murphy, D. H., and Wilkinson, B. H., 1980, Carbonate deposition and facies distribution in a central Michigan marl lake: Sedimentology, v. 27, p. 123-135. Neuhauser, K. R., Lucas, S. G., deAlbuquerque, S., Louden, R. J., Hayden, S. N., Kietzke, K. K., Oakes, W., and DesMarais, D., 1987, Stromatolites of the Morrison Formation (Upper Jurassic), Union County, New Mexico: a preliminary report: Northeastern New Mexico, New Mexico Geological Society 38th Annual Field Conference Guidebook, p. 153-160. Norman, J. R., 1963, A history of fishes: New York, Hill and Wang, 398 p. O'Sullivan, R. B., 1984, The base of the upper Jurassic Morrison Formation in east-central Utah: U. S. Geological Survey Bulletin 1561, 17 p. Oborne, H. W., 1956, Wet Mountains and Apishapa uplift: Raton Basin, Rocky Mountain Association of Geologists, Guidebook, p. 58-64. Ogden, L., 1954, Rocky Mountain Jurassic time surface: American Association of Petroleum Geologists Bulletin, v. 38, p. 914-916. Olsen, P. E., 1980, Triassic and Jurassic formations of the Newark basin: Field Studies of New Jersey Geology, 52nd Annual Meeting, New York State Geological Association, Guidebook, p. 2-41. Oomkens, E., 1966, Environmental significance of sand dikes: Sedimentology, v. 7, p. 145-148. Oriel, S. S., and Mudge, M. R., 1956, Problems of lower Mesozoic stratigraphy in southeastern Colorado: Raton Basin, Rocky Mountain Association of Geologists, Guidebook, p. 19-24. Ott, H. L., 1958, Stratigraphic distribution of Charophyta i n the Morrison Formation of Colorado and Utah [M.S. thesis]: Columbia, Missouri, University of Missouri , Columbia, Patton, H. B., 1924, Underground water possibilities for stock and domestic purposes in La Junta area, Colorado: Colorado Geological Survey Bulletin 27, Part 1, p. 1-58. Peck, R. E., 1937, Morrison charophyta from Wyoming: Journal of Paleontology, v. 11, p. 83-90. Peck, R. E., 1956, Rocky Mountain Mesozoic and Cenozoic nonmarine microfossils: Jackson Hole, Wyoming Geological Association Guidebook, p. 95-98.

PAGE 165

173 \ Peck, R. E., 1957, North American Mesozoic Charophyta: U. S. Geological Survey Professional Paper 294-A, 44 p. Peck, R. E., and Reker, C. C., 1948, The Morrison and Cloverly Formations: Wind River basin, Wyoming Geological Association Guidebook, p. 125-139. Pennack, R. W., 1978, Fresh water invertebrates of the United States (2nd edition): New York, Wiley, 803 p. Peryt, T. M., 1983, Classification of coated grains, T. M. Peryt, ed., Coated grains: New York, Springer-Verlag, p. 3-7. Peterson, F., and Turner-Peterson, C. E., 1987, The Morrison Formation of the Colorado Plateau; recent advances in sedimentology, stratigraphy, and paleotectonics: Hunteria, v. 2, p. 1-18. Pipiringos, G. N., and O'Sullivan, R. B., 1976, Stratigraphic sections of some Triassic and Jurassic rocks from Douglas, Wyoming, to Boulder, Colorado: U. S. Geological Survey Oil and Gas Investigations Chart 69, 1 plate. Popp, B. N., and Wilkinson, B. H., 1983, Holocene lacustrine ooids from Pyramid Lake, Nevada, Peryt, T. M., ed., Coated grains: Berlin, Springer-Verlag, p. 142-153. Potter, P. E., Maynard, J. B., and Pryor, W. A., 1980, Sedimentology of Shale: New York, Springer-Verlag, 306 p. Prince, N. K., 1983, Late Jurassic dinosaur trackways from SE Colorado: Colorado University at Denver Geology Department Magazine, v. 2, p. 15-19. Prince, N. K., and Houck, K., 1986, Sedimentology of the Purgatory River Tracksite, Jurassic Morrison Formation, Southeast Colorado: First International Symposium on Dinosaur Tracks and Traces Abstracts with Programs, p. 23. Reeves, C. C., 1968, Introduction to paleolimnology: New York, Elsevier Publishing Company, 228 p. Reineck, H. E., and Singh, I. B., 1980, Depositional sedimentary environments: New York, Springer-Verlag, 549 p. Ridgley, J. L., and Goldhaber, M., 1983, Isotopic evidence for a marine origin of the Todilto Limestone, north-central New Mexico: Geological Society of America, Abstracts with Programs, v. 15, p. 414.

PAGE 166

174 Romer, A. S., 1945, Vertebrate Paleontology: Chicago, University of Chicago Press , 687 p. Sanders; C. W., 1934, Geology of Two Buttes dome in southeastern Colorado: Bulletin American Association of Petroleum Geologists, v. 18, p. 860-870. Schaeffer, B., and Patterson, C., 1984, Jurassic fishes from the western United States, with comments on Jurassic fish distribution: American Museum Novitates, no. 2796, p. 1-86. Schafer, A., and Stapf, K. R. G., 1978, Permian Saar-Nahe Basin and Recent Lake Constance (Germany): two environments of lacustrine algal carbonates, Matter, E. and Tucker, M., eds., Modern and ancient lake sediments: p •• Schottle, M., and Muller, G., 1968, Recent carbonate sedimentation in the Gnadensee (Lake Constance), Germany, Muller, G. and Friedman, G. M., eds., Carbonate sedimentology in Central Europe: New York, Springer-Verlag Inc., p. 148-156. Scott, G. R., 1968, Geologic and Structure Contour Map of the LaJunta Quadrangle, Colorado and Kansas: U.S. Geological Survey Miscellaneous Geologic Investigations Map I-560, scale 1:250,000. Seward, A. C., 1966, Plant life through the ages: A geological and botanical retrospect: New York, Hafner Publishing Company, 603 p. Shaw, G. L., 1956, Subsurface stratigraphy of the Permian-Pennsylvanian Beds, Raton basin Colorado: Raton Basin, Rocky Mountain Association of Geologists, Guidebook, p. 14-18. Shelley, D., 1980, Manual of Optical Mineralogy: New York, Elsevier, 239 p. Sly, P. G., 1978, Sedimentary Processes in Lakes, Lerman, A., ed., Lakes: Chemistry, Geology, Physics: New York, Springer-Verlag, p. 65-90. Sonnenberg, S. A., and Weimer, R. J., 1981, Tectonics, sedimentation, and petroleum potential, northern Denver Basin, Colorado, Wyoming and Nebraska: Colorado School of Mines Quarterly, v. 76, 45 p. Stanton, T. W., 1905, The Morrison Formation and its relations with the Comanche Series and the Dakota Formation: Journal of Geology, v. 23, p. 657-669.

PAGE 167

175 Steiner, M. B., 1975, Reversal pattern and apparent polar wander for the Late Jurassic: Geological Society of America Bulletin, v. 86, p. 1537-1543. Stose, G. W., 1913, Description of the Apishapa Quadrangle, Colorado: U. S. Geological Suurvey Atlas 186, 12 p. Stovall, J. W., 1938, The Morrison of Oklahoma and its dinosaurs: Journal of Geology, v. 46, p. 583-600. Stovall, J. W., 1943, Stratigraphy of the Cimarron Valley (Mesozoic Rocks), Schoff, S. L. and. Stanton. J. W., Geology and groundwater resources of Cimarron County, Oklahoma: Oklahoma Geological Survey Bulletin 64: p. 43-71. Stovall, J. W., and Savage, D. E., 1939, A phytosaur in Union County, New Mexico, with notes on the stratigraphy: Journal of Geology, v. 47, p. 759-766. Surdam, R. C., Eugster, H. P., and Mariner, R. H., 1972, Magadi-type chert in Jurassic and Eocene to Pleistocene rocks, Wyoming: Geological Society of America Bulletin, v. 83, p. 2261-2266. Tanner, W. F., 1971, Numerical estimates of ancient waves, water depth and fetch: Sedimentology, v. 16, p. 71-88. Tanner, W. F., 1974, History of Mesozoic lakes of northern New Mexico: Ghost Ranch, New Mexico Geological Society 25th Field Conference, Guidebook, p. 219-223. Tappan, H., 1980, Paleobiology of Plant Protists: San Francisco, W. H. Freeman and Company, 1028 p. Tasch, P., and Zimmerman, J. R., 1961, Comparative ecology of living and fossil conchostracans in a seven county area of Kansas and Oklahoma: University of Wichita Bulletin, v. 47, p. 1-14. Taylor, A.M., 1974, Stratigraphy and depositional environments of Upper Jurassic and Cretaceous rocks in Bent, Las Animas and Otero Counties, Colorado [Ph.D. Thesis]: Golden, Colorado, Colorado School of Mines, 211 p. Tidwell, W. D., 1975, Common Fossil Plants of Western North America: Provo, Utah, Brigham Young University Press, 197 p. Toepelman, W. C., 1924, Preliminary notes on the revision of the geological map of eastern Colorado: Colorado Geological Survey Bulletin 20, 21 p.

PAGE 168

176 Turner-Peterson, C. E., 1985, Lacustrine-humate moodel for primary uranium ore deposits, Grants uranium region, New Mexico: American Asociation of Petroleum Geologists Bulletin, v. 69, p. 999-2020. Tweto, 0., 1979, Geologic Map of Colorado: U. S. Geological Survey General Map 77115, scale 1:500,000. Tweto, 0., 1980, Tectonic history of Colorado: Colorado Geology, Rocky Mountain Association of Geologists, Guid ebook, p. 5-9. Vail, P.R., Mitchum, R. M. ,J., Todd, R. G., Widmier, J . M., Thompson, S., Sangree, J. B., Bubb, J. N., and Hatelid, W. G., 1977, Seismic stratigraphy and global changes of sea level, Payton, C. E., ed, Seismic stratigraphy-applications to hydrocarbon exploration: American Association of Petroleum Geologists Memoir 26, p. 49-212. Van Horn, R., 1957, Ralston Creek Formation, new name for Ralston Formation of LeRoy (1946): American Association of Petroleum Geologists Bulletin, v. 41, p. 755-756. VanHouten, F. B., 1964, Cyclic lacustrine sedimentation, Upper Triassic Lockatong Formation, central New Jersey and adjacent Pennsylvania: State Geological Survey of Kansas Bulletin 169, v. 2, p. 497-532. VanHouten, F. B., 1965, Crystal casts in upper Triassic Formations: Sedimentology, v. 4, p. 301-313. Van Morkhoven , F. P., 1962, Post-Paleozoic Ostracoda: Their morphology, taxonomy and economic uses: New York, Elsevier Publishing Company, 300 p. Vital, D. A., 1948, Modern lime-magnesium concretions and oolites in the lakes of the Kulindin Steppe, Altai Region, U.S.S.R.: Moscow Society of Naturalists, v. 23, p. 83-100. Waage, K. M., 1953, Refactory clay deposits of south-central Colorado: U . S. Geological Survey Bulletin, 993, 104 p . Waldschmidt, W. A., and LeRoy, L. W., 1944, Reconsideration of the Morrison Formation in the type area, Jefferson County, Colorado: Bulletin of the Geological Society of America, v. 55, p. 1097-1114. Walker, T. R., 1967, Formation of red beds in modern and ancient deserts: Geological Society of America Bulletin, v. 78, p. 353-368.

PAGE 169

Webb, J. A., 1979, A reappraisal of the paleoecology of conchostracans (Crustacea:Branchiopoda): N. Jb. Geol. Palaont., Abh, v. 158, p. 259-275. 177 Weiss, M.P., 1969, Oncolites, paleoecology, and Laramide tectonics, central Utah: Bulletin American Association of Petroleum Geologists, v. 53, p. 1105-1120. West, E. S., 1978, Biostratigraphy and paleoecology of the lower Morrison Formation of Cimarron County, Oklahoma [M.S. thesis]: Wichita State University, 61 p. Western, D., 1980, Linking the ecology of past and present mammal communities, Behrensmeyer, A. K. and Hill, A. P., eds., Fossils in the Making: Chicago, Illinois, University of Chicago Press, p. 41-54. Western. D., and Van Praet, C., 1973, Cyclical changes in the habitat and climate of an East African ecosystem: Nature, v. 241, p. 104-106. Wilkinson, B. H., Pope, B. N., and Owen, R. M., 1980, Nearshore ooid formation in a modern temperate region marl lake: Journal of Geology, v. 88, p. 697-704. Wilson, J. M., 1977, Southeast Colorado basin, Plate 22 and Plate 23, Irwin, D., ed., Subsurface cross-sections of Colorado, Rocky Mountain Association of Geologists Special Publication 2: p. 33-36. Wood, G. H. ,J., Johnson, R. B., and Dixon, G. H., 1957, Geology and Coal Resources of the Starkville-Weston Area, Las Animas County, Colorado: U. S. Geological Survey Bulletin 1051, 68 p. Yen, T. C., 1952, Molluscan fauna of the Morrison Formation: U. S. Geological Survey Professional Paper 233-B, p. 21-51. Yuretich, R. F., 1979, Modern sediments and sedimentary processes in Lake Rudolf (Lake Turkana) eastern Rift Valley, Kenya: Sedimentology, v. 26, p. 313-331. Zenger, D. H., and Dunham, J. B., 1980, Concepts and models of dolomitization-an introduction, Zenger, D. H., Dunham, J. B. and Ethington, R. L., eds, Concepts and models of dolomitization: Society of Economic Paleontologists and Mineralogists Special Publication No. 28, p. 1-10.

PAGE 170

Appendix I Measured Sections and Wells Figures 7-10,49 MEASURED SECTIONS (1-13) 1. Lost Canyon: 6-30S-56W, Las Animas County The Lost Canyon section was measured just north of the point on the Black Hills Monocline where the red beds dip below the surface. The Morrison is poorly exposed, and along the main transect the Lytle sandstones have slumped, creating the impression of several individual sand bodies. The Morrison section, where partially exposed, appears to be different from the closest other measured outcrop, Goat Ranch, 3 miles away. Unit A appears to be thinner and Unit B equivalents are sandier here than a few miles to the north, although this may be an illusion caused by slumping of Morrison sediments. 2. Goat Ranch: 4-30S-56W, Las Animas County The Entrada, Bell Ranch equivalent, lower Morrison contact and Unit A are completely exposed at this section. The lower part of Unit B can be pieced together, but the upper portion, as well as Unit C and much of Unit D is covered by large Cretaceous sandstone blocks. 3. Bresden Trail: 27, 28-29S-56W, Las Animas County The limestones were too weathered to process for paleontology, and the lower portion of the exposure is a large slide block. This section is important however, because it is the only section where a fluvial sandstone, capped by a bone-bearing conglomerate in Unit D is well exposed. 4. Round Corral: 26,34-29S-56W, Las Animas County A good exposure of the upper portion of the gypsiferous beds and the chert bed. The lower portion of Unit A and upper portion of Unit B were covered. One of the Pleistocene (?) gravels with large blocks of Lytle sandstone was observed on a terrace just above the highest exposed Unit B limestones. Units C and D were completely covered. 5. The Gap: 26,27-29S-56W, Las Animas County The most complete section, with the chert bed just below the contact well exposed, Unit A mostly covered, and Unit B and C well exposed. Unit D was largely covered by massive blocks of Cretaceous sandstone. OV Mesa and The Gap sections are both on the south face of OV Mesa, less than a mile apart.

PAGE 171

179 6. OV Mesa: 22-29S-56W, Las Animas County The basal contact and Unit A are obscure, and Unit B only partially exposed in a few gullies. Unit C is well exposed however, and several good samples of trace fossils were collected here. The sandstones of Unit D were well exposed, and apparently undisturbed by slump, although none of the upper mudstone was exposed. 7. Pamena Gap: 14-29S-56W, (unsurveyed) Las Animas County This section was chosen for Permian through Entrada exposures. It is on the north wall of OV Mesa, a mile directly north of the OV Mesa and The Gap sections. 8. Purgatoire Canyon: 35-28S-56W, Las Animas County The northernmost exposures of red beds, at the junction of Chaquaqua Creek and the Purgatoire River, was chosen as a tie between Rock Crossing and The Gap. Although it is possible to identify the various units, none of them were well exposed, and the upper portion of the section was covered entirely. 9. Rock Crossing: 8-28S-55W, Las Animas County The dinosaur trackways are well-exposed in the riverbed here in beds probably belonging to Unit B. Units C and D can be identified here, but it was difficult to get a good section because it appears that a small fault runs through the valley, slightly offsetting the strata. 10. Higbee: 35-26S-54W, Las Animas County A small dinosaur trackway was exposed here in a micritic bed similar to those of Unit B. Well developed fluvial sandstones, possibly transitional to Unit C are a few meters above the track-bearing bed. Unit C and D are covered, and the Lytle sandstone is exceptionally thin here, with a thick section of Glencairn sandstone above that. 11. Colbert Canyon: 30-34S-55W, Las Animas County An isolated exposure of the mudstones and sandstones of Unit D was identified 50 km south of the Purgatoire Uplift sections. It was examined in detail as no exposure of these mudstones had been located in the Purgatoire Uplift. 12. Peacock Canyon: Union County, New Mexico This locality is near the type locality of the Exeter Sandstone. It is also is adjacent to a Triassic vertebrate trackway site • • 13. State Line: 13-35S-50W, Baca County, Colorado Located on the side of a small butte on the Colorado/Oklahoma State Line, this section is near one Jurassic dinosaur trackway site, and several Triassic vertebrate trackway sites.

PAGE 172

180 PUBLISHED SECTIONS, FIELD CHECKED (14-20) 14. Two Buttes, Sanders 1934 Composite: 28S-46W; 27S-45W, Baca County 15. County Line: 19-29S-50W, Baca County, Prince (1987) 16. Johnny Branch; Duce (1924) p. 86, Heaton (1939) p. 116, Prince (1987): 22 and 27-28S-52W, Las Animas County 17. Briggs Canyon: 2-30S-55W, Las Animas County, Prince (1987) 18. Jones Lake Fork, Stose (1913) p. 21,29: 31-27S-61W, Las Animas County 19. Cuchara Canyon, Heaton (1939) p. 1166: 25S-64W, Huerfano County 20. Huerfano Canyon, Stose (1913) p. 21,26: 1 2-24S-64 W , Pueblo County 21. Baldwin and Muehlberger (1959) 10a,b: 11-31,32N-32E, Union City, New Mexico PETROLEUM EXPLORATION WELLS ON CROSS SECTION (21-44) * Wells with sample sets examined at American Stratigraphic 22. J. Champlin #1 State (1958): SESE-16-28S-62W, Las Animas County 23. Texaco # 1 Hall (1963): NENE-6-28S-59W, Las Animas County 24. Texaco #1 Schneider (1963): NWSW-26,27S-58W, Otero County * 25. Texaco #1 Cynthia True (1963): NENW-30-28S-56W, Las Animas County * 26. Phillips #1-B Denton (1970): NWNW-1-30S-55W, Las Animas County Phillips # 1 -B Colorado (1970): NENW-35-29S-54W, Las Animas County 28. Trend #1-13, Watkins (1977): NESW-13-29S-52W, Las Animas County 29. Michigan-Wisconsin Pipeline #1-28, Freezeout (1974): SESW-28-29S-50W, Baca County

PAGE 173

181 30. Forest Oil #1-3, Forest, Grynberg, McEndree (1963): NWNE-2-29S-49W, Baca County 31. Kerr-McGee #1, McEndree (1968): NESE-20-28S-48W, Baca County 32. Mobil #1, Niles Estate (1962): NESW-2-29S-44W, Baca County 33. Addington 4-State (1974): NWSW-36-30S-53W, Las Animas Count y 34. Addington 12-State (1974): NWSW-14-32S-53W, Las Animas County 35. Addington 6-Nicholson Autry (1975): SENW-26-32S-52W, Las Animas County 36. Addington 5-Mustang Creek (1975): SWNW-15-33S-52W, Las Animas County 37. Addington 9-Wren (1975): SWNW-9-34S-52W, Las Animas County 38. Phillips Petroleum Company, Cedarwood #1: SESE-17-24S-64W, Pueblo County 39. Colorado Oil #32-11, Sniff Ranch (1972): SWNE-11-24S-49W, Bent County * 40. Tom Brown #1-5, Herrin (1972): NESW-5-25S-47W, Prower County 41. Marathon #1, Plagge (1970): NWSE-19-31S-49W, Baca County * 42. Marathon #1, Cook (1970): SWSW-5-32S-49W, Baca County 43. Champlin #1 Gut (1968): SESW-11-25S-59W, Otero County 44. Model Energy #1 State (1982): SENW-36-29S-60W, Las Animas County PETROLEUM EXPLORATION WELLS, ISOPACH DATA POINTS (45-62) 45. W.C. McBride #1-19, Frank: NWNW-19-24S-44W, Prowers County 46. Michigan-Wisconsin Pipeline #1-26, Bailey: 26-25S-45W, Prowers County 47. Phillips Petroleum #1, Sample Nose (1956): NWNW-9-25S-64W, Pueblo County 48. Ferguson Oil Company #1, Walker (1967): NESW-17-25S-46W, Prowers County

PAGE 174

182 49. Texas Oil and Gas Corp. #1-4, Bailey's Inc. (!977): SWNE-4-26S-45W, Prowers County SO. Pan American Petroleum #1, Bennett (!960): SWSW-1-26S-43W, Prowers County 51. R. H . Fulton, Inc. #1, McNally (1970): SWSW-15-26S-56W, Otero County *52. Mobil Oil Company #1, Rober Unit (1963): NWNW-30-27S-42W, Prowers County 53. Sage Oil Company #1, Braly (1962): SESE-19-27S-51W, Bent County 54. W.C. McBride, Inc. #1-6, Hardin (1968): NESE-6-28S-42W, Baca County 55. Texaco, Inc. #1, Gov't. Davis (1960): SWNE-12-28S-52W, Las Animas County 56. Chevron Oil Company #1, Colorado State (1967): 16-28S-48W, Baca County 57. Trend Resources Limited #1-35, Palmer (1977): 35-29S-52W, Las Animas County 58. Forest Oil Corp. #1-3, Forest, Grynberg, and McEndree (1963): NWNE-3-29S-49W, Baca County 59. J. Champlin #1, Gov't. (1958): 10-29S-62W, Las Animas County 60. J . Champlin #1, Marques (1959): NWSW-9-29S-63W, Las Animas County 61. T. Brown #1, Tucker (1975): SENW-26-31S-46W, Baca County 62. C&K Petroleum #1, Jenkins-Cooper (1975): NWNE-8-35S-45W, Baca County