National Liirary Bibliothèque nationale I*I of Canada du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395. rue Wellington OltawaûN KlAON4 Ollawa ON K1A ON4 Canada Canada

The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bi1,liothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thése sous paper or electronic formats. la forme de microfiche/film, de reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or othexwise de ceile-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. Current tirne scales disagree on age estirnates for Jurassic stage boundaries and carry large uncertainties. The

U-Pb ~r"~~r-"~rdating of voIcaniclastic rocks of preciseiy known biochronologic age is the preferred method to improve the calibtation.

Eighteen new U-Pb zircon dates were obtained on volcaniclastic rock from Surassic volcanic arc assemblages of the North American Cordillera. The voIcanic rocks are interbedded in fossiliferous marine sedimenq rocks, which in mm were dated by ammonite biochronology at the zona1 level. In the Queen Charlotte

Islands, a tuff Iayer directly beneath the Triassic-Jurassic boundary was dated at 200I1 Ma. Lowermost Jurassic

(middle and upper Hettangian) volcaniclastic rocks ftom Alaska yielded ages of 200.8+2.7/-2.8 Ma, 1 W.8* 1 .O Ma, and 197.8+1.2/-0.4 Ma. An ash layer near the base of the Middle Toarcian Crassicosta Zone in its type section in the Queen Charlotte Islands was dated at 181.411.2 Ma. The other new dates furnish additional time scale calibration points.

The biostratigraphy of measured sections contribute to the understanding of North Arnerican Juraçsic ammonite successions, especially for the Hettangian of Alaska. In an attempt to quanti9 biochronologic correlation uncertainty, the cornputer-assisted Unitary Association mehod was used to correlate a Toarcian North American regionai ammonite zone with the northwest European standard. The correlation uncertainty between the two regions is not more than Il standard subzone.

A radiometric age database consisting of 50 U-Pb and 'OAK-~~A~ ages was compiled to construct a revised

Jurassic time scale. Apart from the newly obtained U-Pb ages, several recently reported Cordilleran dates are included with revised biochronologic ages. Additional dates were cornpiled from previous time scales and recent literaîure.

Only a few boundaries are dated directly. The chronogram method was applied for the first tirne ro estirnate ail Early and earIy Middle Jurassic chron boundaries, as well as late Middle Jurassic substage boundaries and Late

Jurassic stage boundaries. The most significant irnprovement concems the Pliensbachian and Toarcian, where six consecutive chron boundaries were detennined. The derived chron durations Vary between 0.4 and 1.6 Ma. Their disparities argue against the assumption of equal duration of chrons or subchrons, which was used as a bais for interpolation in several previous thescales. [nterpolation based on magnetochronology improved chronogram estimates for the latest Jurassic, where the isotopic database remains too sparse.

The initial boundaries of Jurassic stages are proposed as follows: Bemasian (Jurassic-Cretaceous):

144.8 +2.61-3 -7 Ma; Tithonian: 15 1.5 +1 .O14 -4 Ma; Kirnmeridgian: 154.7 + 1.014.9 Ma; Oxfordian:

156.6 +2.01-2.7 Ma; Callovian: 159.7 k1.1 Ma; Bathonian: 166.0 M.61-5.4 Ma; Bajocian: 174.0 + 1.01-7.3 Ma;

Aalenian: 178.0 +1.01-1.5 Ma; Toarcian: 183.6+1.6/-1. 1 Ma; Pliensbachian: 192.0 +3.81-5.2 Ma; Sinemurian:

197.0 + 1 -2l-4.2 Ma; Hettangian (Triassic-Jurassic): 20 1 & 1 Ma.

The revised time scale was used to assess the timing of mass extinction events and subsequent biotic recoveries. Marine and terrestrial extinctions at the end of the Triassic both occured at abou 20 1 Ma, later than suggested in most previous tirne scales. During the Pliensbachian-Toarcian rnass extinction, elevated extinction rates were sustained between 185.7 and 18 1.4 Ma and apparently followed by an immediate recovery. îhe hypothetical26 Ma extinction periodicity is plausible between the end-Permian and end-Triassic, between the

Pliensbachian-Toarcian and Callovian, and among the Tithonian, Cenomanian-Turonian, and end-Cretaceous events. However, the spacing of the end-Triassic and Pliensbachian-Toarcian events is less than 20 Ma, significantly less than predicted by the periodicity model.

iii TABLE OF CONTENTS

ABSTRACT ...... ii

TABLE OF CONTENTS ...... iv ... LIST OF TABLES...... VIII

LIST OF FIGURES...... ix

LIST OF PLATES...... xi .. ACKNOWLEDGEMENT ...... xl~ ... DEDICATION ...... xi11

CHAPTER 1. INTRODUCTION ...... I

1.1. MTRODUCTORY STATEMENT...... 1

1.2. OBJECTIVES ...... +...... - 7

1.3. METHODS...... -7

1.4. PRESENTATION ...... 3

CHAPTER 2. DEVELOPMENT OF THE JURASSIC GEOCHRONOLOGIC SCALE ...... 5

2.1. MTRODUCTION ...... 5

2.2. DATABASE AND PROCEDURES iN TIME SCALE CALIBRATION ...... 8

2.3. KEY AREAS OF PROGRESS AND CONSTRUCTIONAL DIFFERENCES BETWEEN SCALES ..... 9

2.3.1. Advances in isotopic dating ...... 9

2.3.2. Selection of dates ...... Il

2.3.3. Errors and uncertainties in data ...... 14

2.3.4. Boundary estimation and mathematical data manipulation ...... 16

2.3.5. Interpolation techniques ...... 17

2.4. DIRECTIONS IN CURRENT AND FUTURE TIME SCALE RESEARCH ...... 19

2.5. SUMMARY ...... 21 CHAPTER 3. A U-Pb ACE FROM THE TOARCLAN (LOWER JURASSIC) AND ITS USE FOR TIME

SCALE CALlBRATION THROUGH ERROR ANALYSIS OF BIOCHRONOLOGIC DATINC ...... 2;

3.1 . MTRODUCTION ...... *...*..*...... 2:

3.2. GEOLOGIC SETïING ...... 24

3.3. U-Pb GEOCHRONOMETRY ...... -3'

3 3.1. Methodology ...... -7'

3.3.2. Results ...... - 7'

3.4. BIOCHRONOLOGY ...... - 3 1

3.4.1. The ammonite fossil record...... -3 (

3.4.2. Reliability of observed ranges ...... 2! .. 3.4.3. Biochronologic correlations ...... J .

3.4.4. The effect of taxonornic noise ...... 3

3.5. DISCUSSION ...... 41

3.6. CONCLUSIONS ...... 42

CHAPTER 4. INTEGRATED AMMONITE BIOCHRONOLOGY AND U-Pb GEOCHRONOLOCY

FROM A BASAL JURASSIC SECTION IN ALASKA ...... 44

4 .I . MTRODUCTION ...... 44

4.2. GEOLOGIC SEïTWG AND PREVIOUS WORK ...... 46

4.3. HETTANGMN-SINEMURIAN BIOCHRONOLOGY ...... 4f

4.3.1. LocaI ammonite ranges ...... 49

4.3.2. Correlation and biochronologic dating ...... 54

4.3.3. Taxonomie remarks ...... 57

4.4. U-Pb GEOCHRONOLOGY ...... 63

4.4.1. Analytical methods...... 63

4.4.2. Age deteminations ...... 63

4.5. DISCUSSION ...... 67

4.6. CONCLUSIONS .~...... 71 CHAPTER 5. NEW U-Pb ZIRCON AGES INTECRATED WITH AMMONITE BIOCHRONOLOGY FROM THE JURASSIC OF THE CANADIAN CORDILLERA ...... 72

5.1. INTRODUCTION ...... 72

5.2. REGIONAL GEOLOGICAL SETTING ...... 73

5.3. DATiNG METHODS ...... 74

5.3.1. Ammonite biochronoiogy ...... 74

5.4. WTEGRATED 610- AND GEOCHRONOLOGICAL RESULTS...... 76

5.4.1. Sinemurian ...... 77

5.4.2. Pliensbachian...... 86

5.4.3. Toarcian ...... 87

5.4.4. Aalenian ...... 89

5.4.5. Bajocian ...... 90

5.4.6. Bathonian ...... 91

5.5. DISCUSSION ...... 97

CHAPTER 6 . U-Pb DATINC THE TRIASSIC-JURASSIC MASS EXTINCTION BOUNDARY ...... 99

MTRODUCTION ...... 99

A NEW U-Pb AGE NEAR THE T-J BOUNDARY IN THE QUEEN CHARLOTTE ISLANDS ...... 100

BIOCHRONOLOGY OF THE T-J BOüNDARY IN THE QUEEN CHARLOTTE ISLANDS ...... 1 O3

OTHER U-Pb AGES AROUND THE T-J BOUNDARY ...... 104

THE AGE OF THE T-J BOUNDARY ...... 106

TIMING THE END-TRIASSIC MASS EXTINCTION ...... 107

SPACMG OF MASS EXTMCTIONS...... 107

CONCLUSIONS ...... 108

CHAPTER 7. A U-Pb AND JO~r-39~rTIME SCALE FOR THE JURASSIC ...... 110

7.1. INTRODUCTION ...... 1 10

7.2. METHODS...... 1 11 7.3. CHRONOSTRATiGRAPHIC FRAMEWORK ...... 1 12

7.4. THE ISOTOPIC AGE DATABASE ...... 1 13

7.5. COMMENTS ON ITEMS USED IN THE ISOTOPIC DATABASE ...... 115

7.6. COMMENTS ON ITEMS USED IN EARLIER TIME SCALES BUT REJECTED

M THIS STUDY ...... 128

7.7. DIRECT DATMG OF STRATIGRAPHIC BOUNDAMES ...... 131

7.8. CHRONOGRAM ESTIMATION OF BOüNDARIES ...... 132

7.9. ADSUSTED STAGE BOUNDARY ESTIMATES ...... 137

7.IO . LATEST JURASSIC STAGE BOLJNDARY ESTIMATES TWROUGH

MAGNETOCHRONOLOGIC INTERPOLATION ...... 138

7.1 1. THE JURASSIC TIME SCALE...... t39

7.12. DISCUSSION ...... 140

CHAPTER 8. TIME CONSTRAINTS ON JURASSIC MASS EXTINCTIONS AND RECOVERIES ...... 143

8.1 . INTRODUCTION ...... 143

8.2. THE END-TRIASSIC MASS EXTINCTION AND EARLIEST JURASSIC RECOVERY ...... 144

8.3. THE PLIENSBACHIAN-TOARCIAN EVENT ...... 145

8.4. LATER .KI RASSIC EVENTS ...... ,., ...... 145

8.5. SPACMG OF JURASSlC AND OTHER MESOZOIC MASS EXTINCTIONS ...... 146

CHAPTER 9. SUMMARY ...... 149

REFERENCES CITED...... 153

APPENDIX 1- INPUT OF COMPUTER-ASSISTED BIOCHRONOLOGIC CORRELATION

USINC BioGraph ...... 171

A 1.1. LIST OF AMMONOID TAXA AND THEIR CODES...... 171

Al .2. AMMONOID DISTRIBUTION M REPRESENTATIVE TOARCIAN SECTIONS ...... 173

APPENDiX 2. CORRELATION TABLES USCNC UNITARY ASSOCIATIONS (BioGraph OUTPUT). 176

vii LIST OF TABLES

Table 2.1. Boundary ages and duration of the lurassic Period in the early thescales ...... 6

Table 2.2. Summary of dates obtained fiom the Guichon Creek batholith ...... 14

Table 3.1. U-Pb zircon analytical data of sample PCA-YR-1...... 78

Table 3.2. List of stratigraphie sections and sources of BioGraph input...... 35

Table 3.3. Synonymies of species of Hildocerus...... 40

Table 4.1. U-Pb zircon analytical data...... 64

Table 4.2. Comparison of ammonite identifications in Imlay (198 1) and in this report ...... 68

Table 5.1. Summary of biochronology, geochronology. and location of sarnples...... 78

Table 5.2. U-Pb anaiytical data...... 83

Table 6.1. U-Pb analytical data ...... 102

Table 7.1. Listing of selected critical isotopic ages...... II6

Table 7.2. Chronogram estimates of the initial boundary of chrons .substages and stages ...... 1%

viii Figure 2.1. Synopsis of the Jurassic part of time scales with assigned stage boundw ages...... 7

Figure 2.2. The number of critical Jurassic dates used in some key thescales ...... 9

Figure 23. Summary of geochronometnc data fiom the Guichon Creek batholith ...... 11

Figure 2.4. Proportion of dates obtained by different isotopic methods ...... 13

Figure 3.1. Location map of the Yakoun River section...... 25

Figure 3.2. Litho- and biosûatigraphy of the Toarcian of the Yakoun River section...... 26

Figure 3.3. U-Pb concordia diagram for zircons fiom sample PCA-YR- 1...... 28

Figure 3.4. Collection levets, observed ranges, and estimated maximum ranges in the Yakoun River section...... 32

Figure 3.5. Global latest Early to early Late Toarcian ammonoid taxon ranges...... 37

Figure 3.6. Reproducibility of the 40 Unitary Associations in Toarcian ammonite faunas...... 38

Figure 3.7. Simulation of the effect of 25% taxonomie noise on the unitary associations...... 40

Figure 3.8. Cornparison of boundary age estimates and numerical mid-point of the Toarcian ...... 47

Figure 4.1. Location map of the hale 8ay section ...... 45

Figure 4.2. Normal fault separating the lowest Jurassic from uppermost Triassic rocks...... 47

Figure 4.3. Measured lowest Jurassic stratigraphie section and ammonite ranges at Puale Bay .Alaska ...... 50

Figure 4.4. Biochronologic correlation chart of Henangian ammonite zonations proposed for key regions ...... 55

Figure 4.5. Ranges of Hettangian ammonite taxa known from Puale Bay as established in other regions ...... 56

Figure 4.6. One cm thick crystal tuff layer near the top of the Kamishak Formation...... 65

Figure 4.7. U-Pb concordia diagrarns for samples fiom the Hettangian section at Puale Bay ...... 66

Figure 4.8. Cornparison of Hettangian age estimates in recent time scales and the new U-Pb dates ...... 70

Figure 5.1. Index map of U-Pb sarnple tocalities ...... 73

Figure 5.2. North Arnerican composite ammonite zonal scheme for the Sinemurian through Bajocian ...... 75

Figure 5.3. U-Pb concordia diagrams of samples 1-13 ...... 79

Figure 5.4. Measured or schematic stratigraphie sections showing U-Pb samples and fossil collections...... 81

Figure 6.1. Locality index map of Triassic-Jurassic boundary sections in the Queen Charlotte Islands ...... LOO Figure 6.2. U-Pb concordia diagram of zircons From a tuff layer immediately below the T-J boundary...... 10 1

Figure 63. The U-Pb dated tuff layer in the T-J boundary section on Kunga Island...... IO4

Figure 6.4. Chronogram estimation of the T-J boundary age based on 13 U-Pb dates...... 106

Figure 7.1. Early and Middle Jurassic ammonite biochronological units ofNorth America ...... 114

Figure 7.2. Chronogram plots of chronostratigraphic boundary ages ...... 135

Figure 73. Cornparison of the proposed Jurassic time scale with other major time scales ...... 142

Figure 8.1. Periods between mas extinction events in the Mesozoic and at its boundaries...... 147 LIST OF PLATES

Plate 4.1. Middle Hettangian ammonite fauna of Puale Bay ...... 52

Plate 4.2. Late Hettangian through Early Sinemurian ammonite fauna of Puale Bay ...... 53

Plate 5.1. Late Bathonian ammonoids of the McDonell Lake and Zymoetz (Copper) River sections...... 93

Plate 5.2. Late Bathonian ammonoids of the McDonell Lake and Zymoetz (Coppet) River sections...... 94 The nature of this thesis project required an unusual degree of collaboration with numerous field geologists, geochronologists, and paleontologists to whorn I'm grateful for their advice, help, and continuing interest. The opportunity to interact with such a diverse group of knowledgeable colleagues trernendously enhanced my leaming experience. Firstly, 1 thank Paul Smith, my thesis supervisor, for suggesting a rewarding (and fnimating at times, 1 might add) research project, providing the paleontological laboratory facilities and generous Funding, accornpanying me to the field, and discussing and reviewing thesis drah and various other manuscripts. Jim Mortensen etriciently introduced a paleontologist to the world of U-Pb geochronology, allowed me to use the facilites of the Geochronology Laboratory. answered rny many simple-minded questions throughout the project, and somehow found time for corning to the field with me. Bob Anderson is thanked for providing logistical support through the Geologicai Survey of Canada for field, work during the summer seasosns, acting as a liaison to the Geochronology Laboratory of the GSC, his keen interest in the project, and insightfül and thorough reading of thesis drafts. Howard Tipper's immense experience in the Jurassic geology and biostratigraphy of western Canada was an invaluable resource in selecting field areas for dating and interpreting results. Access to his [ab at the GSC was most useful. Besides, his wisdom, encouragement and benevolence throughout the project is much appreciated. Kurt Grimm is thanked for his advice and intellectual stimulation as a rnernber of the supervisory comrnittee. Rich Friedman taught me everything I needed to know about zircon sample processing. mineral separation. zircon picking, abrasion, and data reduction, and did it patiently several times over. Rich kept my samples going through the pipe and his ever-carefùl job in chemistry and mass spectrometry was instrumental in the analytical work for this project. Randy Parrish (fonnerly at GSC) participated in the project by collecting zircon samples in the Queen Charlotte Islands and producing quality geochronological results reported in Chapter 3. Vicki McNicoll and Mike Villeneuve (GSC) produced some of the analytical results reported in Chapter 5. In various field areas throughout the Cordillera, Chris Ash (BCGS), Randy Enkin (GSC), Charlie Greig (GSC), Craig Hart (Yukon Geoscience), Julie Hunt (Yukon Geoscience), Peter Lewis (UBC), Gary Payie (BCGS), Don MacIntyre (BCGS), and Graham Nixon (BCGS)provided much useful logistical assistance and shared their knowledge of the respective areas. Fiona Childe (UBC), Larry Diakow (BCGS), Carol Evenchick (GSC), Janet Gabites (UBC), Gary Johannson (UBC), Brian Mahoney (UBC), Mitch Mihalynuk (BCGS), and JoAnne Nelson (BCGS) connibuted unpublished isotopic ages or information on the geological sening of dated or potentially dateable units. The isotopic database compilation benefited fiorn discussions with and unpublished information from Bart Kowallis (Brigham Young U.), Paul Olsen (Lamont), and Fred Peterson (USGS). Beth Carter, Fabrice Cordey, Russell Hall (U. of Calgary), Genga Nadaraju (UBC), Mike Orchard (GSC). and Tim Tozer (GSC) helped in micro- or rnacrofossil identification andlor gave advice on biochronology. Among my paleontologist colleagues. Giselle Jakobs dererves special thanks for sharing her knowledge on the biostratigraphy of the Yakoun River and Diagonal Mountain localities and her help in the identification of Toarcian ammonites and collecting in the field. Student field assistants Rocio Lopez, Sean Galway, Tanya Mauthner, and Keegan Schmidt gave help in sample collecting. Fellow graduate students at UBC, especially Mark Caplan, Ben Edwards, Bo Liang, and Hristo Stoynov are thanked for miifil discussions, their help with cornputers, and etc. Attila V6riJs and Istvrin Mabkki granted me an extended study leave frorn the Hungarian Natural History Museum to undertake PhD studies at UBC. In addition, Attila also kept me in touch with the Hungarian geological community. A number of fiiends gave me invaluable moral support in difficult times when the project was in lirnbo. Thank you Ian, Naomi, Dora, and al1 of you l can't list here. My stay at UBC was an endeavor for the whole family. To prioritize graduate school and parenting has proved to be a most delicate task in wich al1 of us had to make sacrifices. The biggest of hugs goes to my wife, Maiia Mayer-PAIS., for her tremendous support, coping with my absences, just keeping sane in crazy times, and enduring my seemingly Iifelong leaming.

xii "Sciencehath bitter routes but it bearath sweetfnrits. "

Jhos Apiczai Csere, 17th century Hungarian scholar

"...here:see this Stone, fram up [here? try as you might, you can 't; fo show all ihis in fine detail - there is no such instrument. "

Mikl6s Radnbti, 20th century Hungarian poet

[Translation by George Emery]

DEDICATION

To my children, Mirton, Maté, Lilla, and hon,

who dl took their kst steps on this very special campus.

xiii CHAPTER 1

INTRODUCTION

1.1. INTRODUCTORY STATEMENT

Time is a fundamental dimension of geology. There are numerous methods for dating rocks and geologicai

events. Primarily, tirne can be measured in million years (Ma) or expressed in terms of a standard

chronosrratigraphic Framework. Geochronologic, or time scales represent the integration of these two mcthods

whereas numeric ages are assigned to stratigraphic boundaries, The subject of the present study is the tirne scale

calibration for the Jurassic Period.

There are two fundamental observations that warrant this study (dubbed as the "Cordilleran Jurassic

Calibration Project): (1) that the available and widely used time scales for he Jurassic are conflicting and irnprecise. and (2) that the volcanosedimentary sequences of the North American Cordillera have the potential to improve the time scale through furnishing new fie points. Indeed, the Jurassic time scale is inferior to most other post-Paleozoic periods in precision and accuracy. Time scales are built from calibration ortie ponts, Le., radiometric ages of known stratigraphic position. In the North American Cordillera, volcanic arc assemblages are widespread therefore it is possible ta radiometrically date volcanic and volcaniclastic strata that are interbedded with biochronologically dated fossiliferous sedimentary rocks.

An improved time scale will find its application in a diverse suite of lurassic research. From a Cordilleran perspective, it is particularly important for better correlation of magmatic, sedirnentary and metamorphic events thereby refining the geologic history of Cordilleran termes. As Jurassic rocks are hosts to numerous minerai deposits, mineral exploration will also benefit from establishing more realistic Iinkages of known ages of mineralization events, sedimentary and magmatic host rocks, etc. The primary objectives of this study are as follows:

(1) To review the available Jurassic time scales, assessing their methodology, strengths, and deficiencies.

(2) To identiQ the mon suitable geo- and biochronologicat methods for modem time scale research and to obtain reliable calibration points.

(3) To develop methods to estimate the uncertainty of biochronologic dating and correlation between North

American and northwest European zonations.

(4) To compile an updated global database of acceptable and stratigraphically well-constrained Jurassic isotopic ages.

(5) To consmct a Jurassic time scale based on the above, attempting to refine resolution to the zonal Ievel.

(6) To assess the implications of timing constraints provided by the above scale on major Jurassic bio-events.

1.3. METHODS

In the course of the present study, the following methods were employed:

Literature review - An extensive literature review was carricd out to surnmarize the state of affairs in

Jurassic tirne scale research and to select Cordillemn sampling sites for integrated radiometric and ammonite biochronologic dating. Preference was given to sections with a potential or proven record of zona1 resolution ammonite biochronology.

Ammonite biochronology - The standard chronostratigraphy of the Jurassic is based on ammonoid biochronology, therefore its use to establish biochronologic ages was preferred wherever practical. Ammonite and other faunas were collected from measured sections wherever feasible. After identification, we followed the practice of using regional or local North American zonations that are correlated with the northwest European standard.

Error estimation of biochronologic dating - To make a rigorous assessment of correlation uncertainty, we employed a model calculation of range extension and the computer-assisted Unitary Association rnethod to detennine a realistic error range. U-Pb geochronology - For U-Pb zircon geochronology, the most felsic, crystal-tich volcaniclastic units were selected for sarnpling in the field. Depending on the Iithology and ease of collecting, 5 to 30 kg sarnples were taken, preferably fiom measured sections. Procedures for minera1 separation. selection of grains, chemisny, mass spectometry, data reductions and interpretation follow those outlined by Monensen et al. (1 995) with additional remarks given in later chapten.

Time scale construction - Steps of building the time scale include initial screening and compilation of a database, selection of suitable dates as direct boundary estimates, calculation of chronograms for stage and, if possible, tonal boundaties, adjustments of calculated boundary ages, and magnetochronologic interpolation for the

Late Jurassic stage boundaries.

1.4. PRESENTATION

The main part of this dissertation consists of seven chapters (Chapters 2-8) that are self-contained research articles. Chapters 2 and 3 have been published whereas Chapters 4 to 8 are to be submitted for publication in various earth science journals. mis format facilitates rapid communication of the main results to the scientific comrnunity. However, it also results in some inevitable repetition, especially in the introduction and, to a lesser degree, in the discussion sections. Care was taken to minimize the redundancy, although some overlap was unavoidable.

Chapter 2 provides a review of existing Jurassic time scales with emphasis on their different methodologies and concomitant discrepancies. By demonstrating the deficiencies of available scales, we establish the need for the present study. The analysis of previous work helps define the goals and methods CO be followed here.

Chapter 3 is centered around a single U-Pb age obtained fiom the type section of a North American regional standard ammonite zone. It is presented separately as an example for a high-quality calibration point, on the strength of a precise isotopic age with zona1 biochronologic constraints. A case study in analysis of biochrono1o;ic dating error is also given here.

Chapters 4 and 5 contain the bulk of the new U-Pb ages obtained in the course of this study frorn Alaska and

British Columbia (three and thirteen, respectively). All isotopic dates are integrated with ammonite biochronology. Chapter 6 reports another new U-Pb age that has the singular importance of providing a direct date for the

Triassic-Jurassic boundary. It is considered together with twelve other U-Pb ages fiom the uppennost Triassic and lowemost Jurassic to define the age of this boundary that is abo marked by a severe mass extinction event.

Chapter 7 represents the culmination of this study wheceby a cornprehensive Jurassic isotopic database is presented and is used to constnict a revised time scale, arguably the mast important outcorne of this study.

Chapter 8 is a preiiminary assessment of the timing of major Jurassic and other Mesozoic biotic events (mm extinction and subsequent recoveries) in the light of the new tirne scale.

It is only one of the several exarnples of how the revised time scale might shed new light on problems of

Jurassic earth history. Exploring more of such avenues of research will be Fniittül but are beyond the scope of this study. CHAPTER 2

DEVELOPMENT OF TAE JURASSIC GEOC~ONOLOGICSCALE'

2.1. JNTRODUCTION

Time is a fundamental dimension of geologic processes and phenomena. Stratigraphers and geoscientists in general interested in the Jurassic (or any other geologic period) need a chronometrically calibrated chronostratigraphicscale which facilitates correlation of rocks and events dated by various methods. it is also necessary for measuring rates of geologic processes. Conclusions in a wide range of problems are affected by the accuracy and precision of the tirne scale used. For example, in an active margin sening the correlation of isotopically dated magmatic and metamorphic events with biochronologically dated sedimentary events depends on the reliabiliry of time scale. Also, calculation of sedimentation rates, sedimentary cycle duration, purported periodicity in extinction and other phenomena, subsidence or uplift rates, plate velocity vectors. etc. needs accurate measurements of tirne, again dependent on the tirne scale.

The nomencIanire of different kinds of time scales is extensively discussed by Harland ( 1978, see also

Harland et al., 1990). The focus of this review is the calibration of the chronostratigraphic scale made up of standard subdivisions with the Iinear chronometric scales which uses Ma (million years) units. The resuIting calibrated scale is termed the geochronologic scale, sometimes also referred to as the numeric or geoIogic time scale. Over the past 40 years the Jurassic chronostratigraphic scale has remained relatively stable except for the status of the Aalenian and the subdivisions of the latest Jurassic. Even at the zona1 level of standard ammonite biochronology, refinements (e-g. Cope et al., 1980) since Arkell's synthesis (1956) have been minor compared with developments in the field of isotopic dating and the emergence of new concepts and methods of time scale calibration.

t Pubiished in 1995 under the same title in Hantkeniana, v. 1, pp. 13-25. Barre11 1917 Holmes Holmes HoImes Kulp Minimum Maximum 1937 1947 1959 1961 lurassic/Cretaceous 120 150 108 127 13515 135 boundary Triassic/Jurassic 155 195 145 152 18W5 181 boundary

Table 2.1. Boundary ages md duration of the Jurassic Period in the early rime scales.

Many geologic time scales have been published since BarreIl's ( 19 17) fint anempt at systematically estimating the ages of geologic system boundaries (including those of the Jurassic). The early scales (Barrell. 19 17,

Holmes, 1937, 1947, 1959, Kulp, 1961) were based on very limited numben of isotopic dates and provided estimates for only the upper and lower boundaries of the period (Table 2.1).

Beginning with Harland et al. (19641, the next generation of researchen attempted ro assign numeric age estimates to stage boundaries based on a growing database of primarily K-Ar ages (e.g Van Hinte. 1976. Amsaong. 1978). although some workers (e.g. Lambert, 1971 and Fitch et al., 1974) rejected this approach as inappropriate given the available dataset, Modem time scaIes published between 1982 and 1994 (see Fig. 2.1 ) are based on a much larger. although still insufficieot, isotopic age database which includes a slowly increasing proportion of and U-

Pb ages. Many of these time scales introduce some kind of statistical treatment of accepted isotopic data and express the uncertainties of boundary age estimates. Some of these time scales are extensively referred to in the recent geologic Iiterature (especially Harland et al., 1982 superseded by Harland et al., 1990, Palmer, 1983. and to a lesser extent Kennedy and Odin, 1982, Haq et al., 1987, 1988). Time scales which oversirnplified the problem

(Saivador, 1985) or concenbated on the mathematical aspects only (Carr et al., 1984, Bayer, 1987) did not gain widespread acceptance, nor did those based on reinterpretations of pre-existing datasets (e-g. Westermann, 1984,

1988, Halla.cn et al., 1985, Menning, 1989). The Soviet time scales were not widely available and thus had relatively lirtle impact in the western geoscience Iiterature (e.g. Afanasyev and Zykov, 1975, Afanasyev, 1987).

Omitted from my treatment are compilations published only in a chart format which lack adequate explanatory notes or reproduce an earlier compilation (e.g. Van Eysinga, 1975, Cowie and Bassett, 1989). 205 SIN

Figure 2.1. Synopsis of the Jurassic pan of tinie scales witli assigned stage boundary ages.

a approxirnate readings from graph, no nurneric value given in text; no error limits given for interpolated ages; * errors given here are one half of quoted chronogram error range; uncertainties not given in Kent and Gradstein (1985, 1986). Palnier (1983) quotes the unceriainties given by Harland et al. (1 982); boundary ages and 0.50 errors from Table 1 of Bayer ( 1987) - noie ihat his conlparative lime scale (cg. 6A) proposes 20% IO for Triassic/Jurassic and l38Iî for Jiirassic/Cretaceous boundary; estiniated average limits of uncertainiy quoted in text 13.5 Ma. As illustrated in Fig. 2.1, the wide range of available time scales disagree significantly in their boundary

estimates. Where uncertainties are quoted, they ofien exceed the duration of adjacent stages. The aims of this paper

are to (1) provide some background regarding the construction of time scales; (2) review the historic development

of Jurassic time scales; (3) present the variety of modem time scales and analyse their differences and limitations:

and (4) discuss current and future avenues of research which rnay lead to improved accuracy. It is hoped that

Jurassic stratigraphers will benefit from an increased awareness of the differences in the available time scales and

will be more critical in their use.

2.2. DATABASE AND PROCEDURES IN TIME SCALE CALIBRATION

A calibrated time scale depends on biostratigraphically constrained isocopic dates. The following types of dates are commonly used, listed here in order of decreasing frequency in most scales: (1) ages of intrusive rocks whose stratigraphic constnints are established from crosssuning and superposirional relationships with fossiliferous sedimentary rocks; (2) ages of authigenic glauconite found in fossiliferous sedimentary rocks; (3) ages of volcanic flow or pyroclastic layers intercaiated in fossiliferous sedimmtary sequences.

Isotopic age determinations applicable for time scale construction carry three kinds of uncertainties (e.g.

Odin, 1982): analytical uncenainty (rneasurernent error of the isotopic age detemination), geochemical uncertain-

(the possibility that the rneasured apparent age differs from the true age due to disturbance of the isotopic system). and suatigraphic uncertainty (the width of the stratigraphic interna1 to which the sample is confined. coupled with potential correlation error).

Every isotopic age determination must be evaluated for analytical precision, geochronologic accuracy. and quality of biochronologic constraints. The construction of a time scale which best represents this screened database cm be achieved by statistical or graphical methods, or intuitively, i.e. by judging and weighting each date individually. If direct isotopic dating of stratigraphic boundaries is not available, as has so far been the case for the

Jurassic, then interpolation between calibration points becornes necessary. 2.3. KEY AREAS OF PROGRESS AND CONSTRUCTIONAL DIFFERENCES

BETMEN SCALES

Successive arnendments of the geologic time scale reflect progress in certain key areas of building the scale and the differing methodologies of research groups. An analysis of these factors is attempted in the following.

There has been a slow but steady increase in the number of useful Jurassic isotopic dates (commonly called

"items" following Harland et al., 1964) (Fig. 2.2). Armstrong (1978) pioneered a cornputer database for elficient management of large number of isotopic dates, a practice followed by subsequent worken who relied on cornprehensive and original databases with the intention of mathematical treatment of dates (Harland et al.. 1982.

IWO, Gradstein et al,, 1994).

2.3.1. Advances in isotopicdating

A succinct overview of progress in geochronornetry can be found in Armstrong (1 99 1 ). Developrnents in this field were reflected in the time scales. The earliest scdes (BarreIl, 19 17, Holrnes, 1937, 1947) relied mostly on age deteminations of uranium minerals. In the 1950's. the K-Ar method became available and for decades it was the rnost frequently used dating method. The databases used in the rnost recent attempts at rime scale calibration

Holmes Kulp Hariand et al. Odin i-bhd GradStein 1959 1961 1964 1982 1982 et aL1990 et at.1~94~

Figure 2.2. The number of critical Jurassic dates used in some key time scales buiIt on comprehensive isotopic databases. a seven additional glauconite dates listed but no

of ages obtained during the 1960's and 1970's. (The historically oldest item (PTS76) still used by Harland et al..

1990 and Gradstein et al., 1994 is based on an age determination reported in 1958.) Most of these early data were

obtained using the K-Ar rnethod along with a lesser number of Rb-Sr analyses.

Complicating the early time scale compilations was the need to recalculate ages because of inconsistent decay

constants. Following the adoption of standard decay constants by the IUGS Subcornmission on Geochronology

(Steiger and Jager, 1977), this problem was resolved.

Over the last ten years the U-Pb and 40Ar-J9Ar methods emerged as superior in precision to the K-Ar. Rb-Sr. and other rnethods. They also offer the advantage of interpretation of the geological meaning of the ages obtained.

The detection of radiogenic dau&ter isotope loss greatly reduces the geochemical uncertainty of these ages.

The U-Pb dating of accessory minerals. most notably zircon. is one of the rnost accurate isotopic dating methods presently available. The closing temperature of zircon is estimated to be >800 Co (Heaman and Parrish.

1991), thus it retains the radiogenic Pb isotopes even in upper amphibolite grade rnetamorphism. Recent advances conmbuting to the increased accuracy of the method include new analytical procedures Iowering the anaIytica1 Pb blank (Heaman and Parrish. 199 1 ): the introduction of air abrasion to reduce or eliminate discordance due to post- crystallization Pb loss (Krogh, 1982); and the use of an ultra-clean ?05pb tracer (Parrish and Krogh. 1987).

Minerals other than zircon cm also be used for U-Pb dating, e.g. badelleyite, an accessory mineral in mafic rocks

(for a Jurassic example see Dunning and Hodych. 1990), and the relatively rare accessory minera1 monazite

(Parrish, 1990). Monazite has been successfully used (for a Paleozoic example see Roden er al.. 1990) in cases where inheritance of older xenocrystic zircon cornplicates the interpretation of U-Pb zircon data. as inherited monazite appears to be extremely rare.

The40Ar-39~rdating technique was introduced in 1966 by Merrihue and Turner. and its application became more widespread by the early 1970's (McDougall and Harrison, 1988). Nevenheless the HarIand et al. (1990) time scale lists only three J0Ar-34Ardates (as opposed to 30 conventional K-Ar aga) for the Jurassic. It is now well established that the age spectra generated by step-heating technique can be used to interpret the thermal history of rocks- Ar loss or gain can therefore be detected and a true cooling age can be inferred. Among the datable, common rock-forming minerals, hombIende has the highest closure temperature (>500 OC). The precision of-'OArj9~r technique approaches that of the U-Pb chronometer, attaining *l-2 Ma for Jurassic samples in favourable circumstances. K-Ar dates

Figure 2.3. Sumrnary of geochronomebic data from the Guichon Creek batholith (British Columbia, Canada) (modified afier Mortimer et al., 1990). Selected dates from this dataset were used in different time scales for constraining the TriassidJurassic boundary (see Table 2.2). K-Ar dates (published benveen 1960 and 1979, and unpublished data) are summarized on a histogram with a 5 Ma class interval. reponed analytical erron are typically rt8 Ma. Two Rb-Sr dates (published in 1969 and 1979) and one U-Pb date are plotted with their respective error bar. Sources of data are listed in ,Mortimer et al. ( IWO).

As an example, dates obtained from the Guichon Creek batholith (British Columbia, Canada) are illustrated here. This intrusion is perhaps the most geochronometrically investigated igneous body in the Canadian Cordillera

(Mortimer et al., 1990). The histogram of Fig. 2.3 sumrnarizes the distribution of apparent K-Ar ages reponed between 1960 and 1979 which are cornpared with two Rb-Sr isochron ages and a recent U-Pb zircon age. It is evident thar (1) the nearly concordant U-Pb age is analytically precise, (2) the rnajority of the K-Ar ages are too young, and (3) at least one (the more precise) of the nvo Rb-Sr isochron ages is also too young, casting doubt on the reliability of those geochronometers.

2.3.2. Selection of dates

It is clear hmthe criteria listed above that not every isotopic date is suitable for use in time scale calibration.

Lambert ( 1971) pointed out that each Jurassic item quoted by Howarth (1964) and listed in Harland et al. (1964) can be criticized on one or more grounds. The greatest risk posed by minerals with low closure temperature is their susceptibility to loss of radiogenic daughter isotopes in post-crystallization geological processes. The result is a young apparent age which is hard to detect especially using the K-Ar method. It became apparent that glauconite ages often tend to be younger than their high-temperature counterparts (e.g. Obradovich, 1988). However, Odin advocated the use of glauconite arguing that its geochernical behaviour can be independently evatuated (Odin,

1982). Glauconitic sediments are abundant in the European Jurassic and most glauconite dates are

biochronologicalIy tightly constrained (e-g. Fischer and Gygi, 1989)- therefore their use in tirne scale studies is

tempting. In the time scales of Kennedy and Odin (1982), Afanasyev and Zykov (1975). Haq et al. (1988). Odin

and Odin (I990), and Odin (1994) glauconite ages dominate the Late Jurassic dataset. Other workers, however,

gradually abandoned the use of glauconite ages. Harland et al. (1982) qualified some glauconite ages as too young

and later treated them as minimum ages only (Harland et al., 1990). Gradstein et al. (1994) Iist glauconite ages but

reject all of them for their final calculations. The 10 to 15 Ma discrepancy in the Jurassic/Cretaceous boundary

estimates between the hvo types of scales (see Fig. 2.1) is clearly the result of the inclusion or rejection of glauconite dates.

Apart from the glauconite question, the curent scales are not selective with respect to the dating method or dated mineral used. There is a significant time lag between major advances in geochronornetry and their utilization

for time scale studies. U-Pb and 40Ar-39Ar dates still represent only a minor fraction of items used in most recent time scales (Fig. 2.4). Moreover. the all-inclusive nature of databases. advocated by Harland et al. ( 1982. 1990) and also adopted by Gradstein et al. (1994) for sake of relative stability in numeric boundary estimates. dampens the effect of new dates with superior precision and accuracy. It is revealing that the most recent time scales contain items gleaned fiorn late 19501s-early 1960's publications in their data suite. A more radical approach, such as

Obradovich's (1993) exclusive use ~f~~Ar-~'Arages with high analytical and stratigraphic precision for the

Creiaceous the scale, has not yet been taken for the Jurassic.

The example of the Guichon Creek batholith can be used again to illustrate ihis point. Dates from rhis pluton have been used in time scales since the work of Holmes (1959), because the emplacement of the batholith is thought to be near the TriassicIJurassic boundary. Successive anempts at dating the batholith. their representation in various time scales, and their effect on the placement of the Triassic/lurassic boundary are lisred in TabIe 2.2. It is notable that (1) the more recently obtained ages tend to be geologically older then previous ones: (2) some time scales (e.g. Gradstein et al., 1994 using sources published more than 20 years earlier) rely on outdated information:

(3) the U-Pb age should be preferred over K-Ar or Rb-Sr ages. The Iast point has not been considered in any existing Jurassic time scale but it may indicate that K-Ar or Rb-Sr ages are best treated as minimum ages only. rGadstein et al.

Odin 1982

K-ArW RM . . 1 I Hadand et al. 1982 1

; K-Ar(g1) 1 I I I Hariand et al. 1964 1 1 1 K-Ar(gl) 1 1 Kulp 1961 1 1 K-Ar(@) I

Figure 2.4. Proportion of dates obtained by different isotopic rnethods used in sorne widely cited time scales. gl - glauconite age; ig - age determined on igneous mineral.

The elimination of biochronologically well constrained glauconite dates increases the weight of ages frorn intrusive rocks whose stratigraphic position is often poorly bracketed although their isotopic age can be precise.

Multi-phase intmsions, whose emplacement can span long intervals of tirne, are cornmon in the geologic record.

The large percentage of plutonic ages in the database of every the scale is accountable for the large uncertainties attached to stage boundary estimates. The listing of inconsistent interpretations of the stratignphic constraints on the Guichon Creek batholith (Table 2.2) reveals the problems associated with intmsive bodies and their serious effect on boundary age estimates. TIME SCALE ITEM QUOTED AGE QUOTED CHRONOS'TRATIC PUBLICATION METHOD' TRIASSICI CODE' POSITION YEAR OF QUOTED JURASSIC SOURCE OF AGE' BOUNDARY ESTlMATE

- - Upper Ttiassic or Lower Jurassic post-lamian. pn-Bajocian Hariand et al.. 1964 1 O (Tozer & Nonan to Bajocian Hacker) Harland & Francis 366 post-early Upper Triassic. 'OO? 1971 (Lambert) ptc-Midd lc Jurassic Armsuong, 1978 366 Camian-Ptiensbachian not given -21 1; 475 brackets Armsuong, 1982 209 (oldest) 203 (mean) Odin. 1982 NDS 177 K- Ar Kennedy & Odin. 1982 (Armstrong) Rb-Sr Harland et aI.. 1982 PTSS 366

HarIand et ai.. 1990 NDS 177 Norian-Hettmgian brackets 'OS (A3661 Gradstein et al.. 1994 284 Norian-Henmgian bmckets (NDS 177)

2 le3 post-Carnian (possibly post- Mortimer U-Pb Norian). pre-Pl icnsbachian et al.. 1990~ (possibly pte-Sinemurian)

Table 2.2. Surnmary of dates obtained from the Guichon Creek batholith (British Columbia, Canada) and their role in successive time scales with reference to the Triassic/lurassic boundary age estimate. hame of abstractor or reference to item used in previous rime scales given in parentheses; 2parentheses indicate that no direct source given but reference made to an item used in previous time scales; 3approximate reading from graph, no numeric value given in rext; 4recaiculated in Harland et al. (1982) using standard decay constants in Steiger and Jtiger (1977); 5age not referred to in Gradstein et al.. 1994 or any other time scale.

2.3.3. Errors and uncertainties in data

The three main sources of uncertainties related to items used in time scale studies were mentioned above.

Quantification oiuncertainties in form of error limits is required if errors on mathematically derived boundary age estirnates are to be rigorously calculated. Errors need to be taken into account even if an intuitive rnethod is preferred when each date is individually evaluated and weighted.

Any isotopic age determination is subject to analytical ermr. It is generally reponed at la (68% confidence) or 20 (95% confidence) level. AIthough this is common practice now, many early dates used in even modem scales do not have published error lirnits. The geochemical uncertainty, Le. the possibility that the apparent age may deviate fiom the me age of the

dated rock, cannot be quantified. It is dealt with at the selection of acceptable isotopic dates when for exarnple,

glauconite ages are treated as minimum ages only (HarIand et al.. 1990) or rejected altogether as suspect (Gradstein

et al.. 1994: see discussion above). As noted earlier, none of the Surassic scales acknowledges that U-Pb and Jo~r-

39Aranalyses can prove the closed system behaviour of the isotopic system, thereby minimizing the geochemical

uncertainty.

The stratigraphie uncertainty also poses serious problems. Chronostratigraphic assignment of intrusive rocks

is problematic and cornmonly even the tightest constraints Ieave an interval of several stages. as demonstrated

above by the exarnple of the Guichon Creek batholith (Table 2.1). Here the overlying sediments were dated as

Early Jurassic based on the bivalve Wqfa(Frebold and Tipper, 1969). The upper bracket of the batholith's

emplacement was arbitrarily interpreted as Hettangian. Sinemurian. or Pliensbachian in different time scales.

without due consideration to the Hettangian to Toarcian age range of Weyfu.Although Late Pliensbachian

ammonites were subsequently found as the oldest fossils in overlying sediments (Tipper in Monger and McMillan.

1984). chis is not reflected in the time scales of Harland et al. (1990) and Gradstein et al. ( 1994) which cite a

Hettangian upper bracket.

Most authors are satisfied to accept the reported stage-level chronostratigraphic brackets of isotopic ages and

they take them at face value implying that biochronologic correlation is free of error. Most biostratigraphers.

however, would agree that chronologie correlation between stratigraphies of different fossil groups and separate

biogeographic provinces is not straightfonvard. Only Harland et al. (1990) attempted to quanti@ this kind of

chronostratigraphic uncertainty as "fossil error" by arbitrarily assigning an additional 2.5 Ma error to al1 Jurassic

ages. As acknowledged by Smith (1993), it is an incorrect overestimation of the uncertainty which was chosen only

for simplicity and in lack of other available error estimation method. Another example from near the

Triassic/Surassic boundary fûnher illustrates the diEcuIties faced. Hodych and Dunning (1992) obtained a U-Pb age from a basalt flow within the non-marine. Upper Triassic to Lower Jurassic Newark Supergroup in Nova

Scotia. The TriassidJurassic boundary within the Newark Supergroup is esrablished not far below the dated basalt. based on palynology and vertebrate biostratigraphy. However, correlation with standard chronostratigraphy is debated (Hallam, 1990). Clearly, the usefulness of this analytically precise date depends on the estimation of the uncertainty involved in correlating non-marine faunas and floras with the standard ammonite zonation. 2.3.4. Boundary estimation and mathematical data manipulation

Afier building the dataset of acceptable isotopic dates, the stage and period boundary ages need to be

estimated. With the "geochronologic approach" (Odin, 1994), the boundary ages are derived from intuitive judgement and weighting of each individuai data point. As there are ofien conflicts between isotopic ages and their

chronostratigraphic position. an element of subjectivity is introduced in resolving the disagreements. Pre-1982 time

scales were al1 constnicted in this manner. and Odin's research group still favours this method (Kennedy and Odin.

1982, Odin and Odin, 1990, Odin, 1994).

With an increase in the nurnber of dates used, the intuitive method becomes more diffîcult and the need arises

for mathematical treatment of data. On the other hand, a large dataset is needed to statistically validate such

procedures. Armstrong (1978) graphicaIIy plotted al1 of the relevant dates accepted by him. thus provided a visual rneans of evaluating consistency and choosing the best fit for boundary ages. This was a step towards removing potential bias by using the dataset coIlectively.

There have been three mathematical rnethods applied to the boundary age estimation problem. The formulae and detailed descriptions are not reproduced herein, only the key features are discussed. Harland et al. (1982. 1990) used the chronogram rnethod which was originally introduced by Cox and Dalrymple (1967) to calculate the age of magnetic polarity revenais ftom the availabie isotopic dates. The boundary is approximated by an Eft) error function which is calculated at regulariy pIaced trial ages (for details see Harland et al., 1990, p. 106). In the calculation only the inconsistent dates are considered, Le. those where the isotopic age and the trial age for stage boundary are in conflict given the known chronostratigraphic age of the sample. The standard deviation of the date is taken into account, therefore analytically precise ages are more heavily weighted regardless of their geochernical uncertainties. A chronogram is graphically produced by plotting E, the error function against r, the trial age. The ideal chninogram curve is pambdic and assumes its minimal value at the boundary age. The error range is established statistically non-rigorously as the time interval where Eqi,+I.

Agterberg (1988, 1994) noted that although the chronogram method does provide unbiased boundary age estimates, it does not use the dataset efficiently inasmuch as it excludes consistent dates which carry important information. He proposed using the maximum likelihood method which was shown to be superior to the chronogram rnerhod by obtaining boundary age estimares with lower error limits from the same data set. As with the chronogram method, the maximum Iikelihood method is also based on the assumptions that the crue ages of samples are randomly distributed and their error distribution is normal. Gradstein et al. (1994) used the maximum

Iikelihood method and noted that it yielded significantly better results (Le. smaller standard deviations) than the chronogram when the available data was sparse, such as for the Jurassic. [t is, however, not fiilly documented whether in the case of srnaIl sarnple size the assumption of nonnal distribution of samples remains warranteci-

Furthemore. Gradstein et al. (1994) ignore the chronostratigraphic error in their time scale construction, which is partly responsible for their smaller boundary age errors as cornpared with the Harland et al. ( 1990) scale.

Carr et al. (1984) also noted that a weakness of the chronogram method was that the consistent dates were ignored in the calculation. They proposed a linear regression model instead, based on the assumption that the relative stratigraphic position of isotopically dated samples can be expressed numerically and the plot of isotopic ages versus stratigraphic position is expected to be a straight line. A serious practical limitation to this method is that only the ages of stratigraphicalIy precisely constrained sampIes can be used. not many ofwhich are available in the Jurassic. In fact, Carr et al. (1984) were only able to calculate the boundary ages for the Jurassic period but not for its stages. Theoretically, the quantificationof stratigraphic position requires the assumption of constant rate of the geological process used in the Iinearization. As it will be discussed below, most such assumptions are controversial.

2.3.5. Interpolation techniques

Whether using a statistical approach or intuitive evaluation of data for best fit, establishing a satisfactory boundary age requires a minimum number of good quality. consistent dates frorn both side of the boundary.

Lacking adequate control for boundaries to be calibrated. estimates must be made by interpolation benveen some anchored calibration points using a Iinearizing operator based on a geologic process with assumed constant rate.

BarreII (191 7) had only one Tertiary and one Late Paleozoic isotopic date to calibrate the intervening

Mesozoic. His approach. foliowed in the subsequent the scales of Holmes including his last anempt (1 959). \vas to use cumulative sediment thicknesses as a proxy for the. Although the underlying assumption of constant sedimentation rate cannot be defended, this method allows a crude estimate of relative duration of geologic periods.

An attempt to revive the concept by Smith (1993) suggests that a new look at this method taking into account recent research on measuring stratigraphic completeness (e.g. Sadler and Strauss, 1990) couId yield more sophisticated resulrs.

For interpolation within a period, Harland et al. (1964, 1982) assumed equal duration of constituent stages.

The influential Harland et al. (1982) scaIe used no Jurassic tie-points (Le. al1 Jurassic stage boundary chrono,~rarns were rejected as having error ranges in excess of 5 Ma). Interpolation benveen the AnisianfLadinian and

AptiadAlbian tie points was used to determine ail Jurassic stage boundaries. Chronostratigraphic subdivisions are based on faunal evolution but the usage of stages records historical developments and tradition in stratigraphy. ft is

therefore less unrealistic to assume an equaI duration for biochronozones (ammonite zones for the Jurassic) in

which case a constant rate of evolution is implied. This method was introduced by Van Hinte (1976) and later

followed by Kent and Gradstein (1985, 1986, for Hettangian to Oxfordian only) and Harland et al. ( 1990). The

latter scale used five tie-points for interpolation of Jurassic stages: the chronogram ages to define NoriadRhaetian.

HettangianJSinemurian. and Oxfordian/Kimmeridgian boundaries which are supplemented by the mid-Bajocian

and TithoniadBemasian pseudo-tie-points. Westermann (1 984. 1988) carried the argument further by deveIoping

the scaled equal subzone interpolation method which considers the ammonite subzone as a basic unit and scales the

duration of a subzone at 0.75 of that of an undivided zone. His time scale is based on the re-apportion of Jurassic

according to this method, using the period boundary ages proposed by Palmer (1983). (However. after scaling and

rounding, the sum of Westetmann's (1984, fig. 1) stage durations exceeds the total duration of Jurassic in Palmer

(1983) by 1.5 Ma.) Hallam et al. (1985) also used a sirnilar, subzone based interpolation scheme for their lunssic

time scale.

A fundamental criticism leveled against these methods is that the assumption of constant rate of evolurion is unwamnted (e.g. House, 1985). Direct isotopic dating of ammonite zones is not yet available for the Jurassic but

Cretaceous data of Obradovich (1993) points to widely disparate zonal durations. A possible independent merhod of estimating zonal durations utilizes orbitally forced sedimentary cycles. House (1985) suggested that

Milankovitch cycles could be demonstrated in the rhythmically-bedded lowermost Jurassic Blue Lias and the

Upper Jurassic Kimmeridge Clay of England and the recognition of cycles could have chronostratigraphic application. Schwarzacher (1 993) pointed out that House's (1985) calculations supponing cycle durations within the Milankovitch range first used ammonite zona1 durations derived from the equal zone assumption, before the oversimplified counting of bed couplets was in turn used to provide arguments against the equai zone durations.

Subsequent studies confimed the presence of cycles in the Blue Lias utilizing sophisticated filtering and spectral analysis (Weedon, 1985, Smith, 1990). Smith's (1990) data, derived from the cyclostratigraphic correlation of three biostratigraphically well-controlled sections, are strongly suggestive of disparate ammonite zonal durations in the

Hettangian and earliest Sinemurian. Based on the number of cycles represented, he found an order of magnitude difference between some zones, independent of the absolute cycle length. Similar disparity in ammonite zonal length was also suggested for the Cenomanian (Cretaceous) based on Milankovitch cycles studied by Gale ( IWO).

Laçtly, another possible linearizing operator for interpolation is the width of sea-floor magnetic anomalies.

The oldest known preserved oceanic cmst is Callovian in age, therefore the post-Middle Jurassic geochronologic time scale cm be calibrated using the magnetic polarity time scale. A thorough review of the topic can be found in Hailwood (1989). ffa constant spreading rate is assumed, then the width of sea floor magnetic anomalies is

proportional to the duration of the respective magnetochrons. The spacing of the Hawaiian sequence of oceanic

magnecic lineations (Larson and Hilde, 1975) was used for Late Jurassic interpolation in the DNAG tirne scale

(Palmer, 1983, Kent and Gradstein, 1985, 1986) and by Gradstein et al. (1994). It is interesting, although perhaps

coincidental, that the proportions of Late Jurassic stages in these magnetochronologically interpolated time scales

are remarkably similar to the proportions obtained by Westermann (1984. 1988) using the equal scaled subzone

method.

Cubic spline srnoothing was proposed as an additional mathematical technique to improve interpolation

(Agterberg, 1988. 1994). The time scale of Gradstein et al. (1994) uses this method which offers several

advantages over the simple linear interpolation: it takes into account the standard error of tie-points. reduces the

asymmetry of unit length on the two sides of tie-points, and reduces the uncertainty of intemiediate boundaries.

especially for stages with large error.

Despite the availability of these techniques, Odin (1994 and references therein) argued against interpolation

and used it only to complement his "geochronologic approach" when the lack of isotopic dates othenvise precludrd

the assignment of a boundary age (e.g. for his Sinemurian/Pliensbachian. Pliensbachian/Toarcian.

ToarcianIAalenian, and Oxfordian/Kimmeridgian boundaries).

2.4. DIRECTIONS IN CURRENT AND FUTURE TIME SCALE RESEARCH

The relatively poor quality of calibration in the Jurassic has been repeatedly pointed out by various time scale authors (e.g. Harland et al., 1990, Gradstein et al., 1994, Odin, 1994). Its primary cause is the scarcity of adequate critical dates from this period. Therefore significant improvement of the Jurassic time scale cannot be achieved without new isotopic dates. Statistical data analysis and interpolation will be relegated to secondary importance as the pool of acceptable ages increases.

Recently the gap between the resolution of biochronology and isotopic dating has been closing with the advent of accurate and precise U-Pb and J0~r-39~rdates with uncertainties near *1Ma, thus comparable with the average ammonite zona1 duration (e.g. Westermann, 1984). it is expected that zonally constrained high-precision dates will become available, as it has been the case for the Cretaceous (Obradovich, 1993). Dating of volcanic

layers intercalated with fossiliferous sediments should be given priority, since glauconite is genenlly rejected as an unreliable geochronometer yielding anomalously low ages (e.g. Obradovich, 1988) and plutonic rocks cannot

generally have tight enough stratigraphie brackets required for zona1 resolution. The U-Pb and .'O~r-~%rdating

methods yield superior analytical precision with the benefit of offering an intemal checks for closed system

behaviour. Their use reduces or eliminates the geochemical error. Le. the apparent venus true age dilemma

plaguing other dating methods. In this light the traditional time scale database needs to be re-evaluated and the bulk

of old K-Ar and Rb-Sr ages replaced (rather than supplemented) by newly obtained U-Pb and 40Ar-39Ar ages.

Volcanosedimentary successions with the potential for stratigraphically well-constrained dateable horizons

are scarce in Europe but known to exist abundantly in areas of Jurassic active margins, e.g. in the North American

Cordillera and the Andes. Dedicated srudies towards improving the Jurassic time scale are now undenvay in

Canada (Pal@ et al., 1995). A prerequisite for such projects is the developrnent of regional ammonite zona1

schemes outside Europe (e.g. Hillebrandt et al., 1992, Smith et al.. 1994). The increase in precision of

intercontinental correlation will render isotopic age determinations from different parts of the world rnuch more

useful for calibration.

If chronostratigraphic age assignments are made at the zona1 Ievel, the need for an accurate representation of correlation uncertainty increases. Prornising recent research in quantitative biostratigraphy (e.g. tavon range confidence intervals: Strauss and Sadler, 1989. Marshall, 1994) may be adapted and further developed to assess uncertainties of zonal correlation. Alternatively, a semi-quantitative uncertainty assessment may prove satisfactory, where expert judgernent is used to numerically express the reliability of biochronologic correlation to the standard scherne.

Refinements in the Sr isotope reference curve for the Jurassic enable correlations comparable to the subzonal level of resolution at the steepest parts of the curve (Jones et al., I994a, 6). As an independent method. it can corroborate biochronologic correlation or substitute it where zona1 fossils are not avaiIable to deterrnine the chronostratigraphic age of radioisotopically dated sarnples.

Developments in cyclostratigraphic analysis appear prornising for interpolation below the stage level when it rernains necessary. Where the completeness of record and the orbitai forcing of sedirnentary cycIes can be documented, relative zonal durations can be independently assessed. The building blocks of geochronologic scales are isotopic dates of known chronosmtigraphic age. Plutonic rocks are frequently dated isotopically but their chronostratigraphic position is diffïcult to establish and cornrnonly poorly constrained. Glauconite, an authigenic mineral occurring in sedimentary rocks, can also be isotopically dated, but the apparent age is commonIy anomalously low. The biochronologic age of the sampled horizon. however, can be very precisely known. Volcanosedimentary sequences provide opportunities for isotopic dating of samples fiom intercalated volcanic rocks. In many cases, fossils from adjacent sediments tightly bracket the chronosmtigraphic age of the sarnple.

The darabase of critical age determinations has been growing gradually. allowing the estimation of Jurassic stage boundaries for the first time in 1964 (Harland et al., 1964). In the past 30 years numerous amrndrnenrs were published, resulting in a perplexing array of time scales available for the Jurassic. The sigificant discrepancies of the proposed stage boundary ages and their quoted uncenainties (which often exceed the stage durations) may confuse thescale users, who cannot assume that the most recently published scale is necessarily the most reliable.

An obvious reason for repeated revisions of time scales is that new isotopic dates become available and technologic advances increase their analytical precision. Yet the differences between tirne scales cannot be explained by these factors alone. Diversent approaches regarding the selection of data. handling of uncertainties. procedures for boundary age estimation and interpolation al1 contribute to the disagreement among tirne scales.

Among the major modem works (published in 1982 or later), Odin's successive time scales (Kennedy and

Odin, 1982, Odin and Odin, 1990, Odin, 1994) represent the "geochronologic approach". whereby each date is individually evaiuated, the boundary ages are intuitively selected to provide a subjective best fit for the dataset. uncertainties are assigned manually, and interpolation is used, if at all, only as a fast resort where critical data are lacking. Glauconite ages, especially common for the Late Jurassic, are weighted heavily. resulting in younger boundary estimates than in most other scales.

Harland et aI. (1982, 1990) treat glauconite dates as minimum ages but their database is conservative in preserving most items from previous compilations. AH accepted dates are handled equally by using the chronogram rnethod for calculating boundary ages and estimating their associated error range. Only the best constrained chronogram ages are accepted as tie-points. the intervening stage boundaries are interpolated assuming equal duration of stages or zones.

The DNAG time scale (Palmer, 1983, Kent and Gradstein, 1985, 1986) uses "hand-picked" tie-points and interpolation based on sea-floor magnetic anomalies for the Kimmeridgian-Tithonian and the equal-zone-duration 2 1 method for the older part of the Jurassic. Gradstein et al. (1994) append the isotopic database of Harland et al. (1990) but reject the use of glauconite ages. The maximum likelihood method is used to obtain first estirnates for . stage boundaries, then cubic spline smoothing is employed to refine the interpolation results of equal-subzone- duration (for the EarIy and Middle Jurassic) and sea-floor rnagnetochronology (for the Late Jurassic) methods.

The frarnework of Jurassic isotopic ages is still too sparse for reliable and precise calibrarion. Existing scales do not reflect recent directions in geochronometry which increasingly employ the U-Pb and jOAr-j9Ar methods for precise and accurate dating. Perhaps the most prornising approach to improve the Jurassic geochronoIogic scale is the construction of a scale From U-Pb and J0ArA9Arages which are directly constrained at the zona1 Ievel. CHAPTER 3

A U-Pb AGE FROM THE TOARCIAN (LOWER JUFUSSIC)

AND TTS USE FOR TIME SCALE CALIBRATION

THROUGH ERROR ANALYSIS OF BIOCHRONOLOGIC DATING*

3.1. INTRODUCTION

The paradox of Jurassic geochronology is that the biochronologic framework was laid out in the mid- 19th century (Oppel, 1856- 1858) and has been continuously refined (see CalIomon. 1993 for one of several recent reviews) but its calibration with isotopic ages is still poor (Palfy, 1995). Recently published tirne scaIes (Gradstein et al., 1994; Gradstein et al., 1995; Haq et al., 1987; Haq et al.. 1988; Harland et al., 1990; Kent and Gradstein.

1985; Kent and Gradstein, 1986; Odin. 1994; Palmer. 1983) suggest conflicting boundary age estirnates and stase durations. Where uncertainties are quoted (Gradstein et al., 1994; Gradstein et al., 1995; Harland et al., 1990;

Palmer, 1 983), the boundaries of the Toarcian appear the most poorly constrained of al1 the Jurassic stages.

Underlying this problem. there is a scarcity of isotopic dates (eight items listed in Gradstein et al., 1994; Gradstein et al., 1995; Harland et al., 1990), of which al1 except one are K-Ar or Rb-Sr ages with inferior precision and accuracy. Also, rnost dated samples are stratigraphically poorly constrained and none of them is unambiguously confined to the Toarcian stage.

Published in 1997 under the same title by Pal@, J., Parrish, R.R., and Smith, P.L., Earth and PIanetary

Science Letters, v. 146, pp. 659-675. The U-Pb analysis was done by R.R. Parrish who is also the prirnary author of section 3.3. ALI other parts of this chapter were written by the present author. To hprove the calibration ofthe Jurassic time scale, concerred effort is undenvay in the Canadian Cordillera

(Pa@ et al., 1995). U-Pb dating of volcanic units is preferred where fossiliferous volcanosedimentary sequences

allow independent ammonite biochronologic dating at the zonal level.

We report a new, precise (I error is less than 1% at 20 level) U-Pb zircon age from the Queen Charlotte

Islands. The dated sample is biochronologically constrained at the zonal level. Moreover, it was obtained from the

type section of three consecutive Middle and Upper Toarcian North American standard ammonite zones (Jakobs et

al., 1994). To demonstrate the usefulness of this date as a time scale calibration point. we rigorousIy assess the ammonite fossil record of the section using classical confidence intervals for taon ranges (Marshall. 1990;

Springer and Lilje, 1988), the global biochronological correlations using the unitary association (UA) method

(Guex, 199 l), and the effect of Fossil identification errors using mode1 calculations. These methods serve to evaluate the biochronoIogic dating error that is not easily quantifiable. In previous time scales it was either not entered into calculations (Gradstein et al., 1994; Gradstein et al., 1995) or overestimated (Harland et al.. 1990) .

ConcIusions derived here can be extended to the use of other dates from the North American Cordillera as calibration points for the Jurassic time scale.

3.2. GEOLOGIC SETTING

The studied section is exposed on the left bank of Yakoun River on Graham Island, Queen Charlotte Islands,

British Columbia (Fig. 3.1). Jurassic rocks of the archipelago are assigned to Wrangellia, one of the major tectonostratigraphic temanes comprising the North American Cordillera (Monger et al.. 199 1). The Jurassic stratigraphy of WrangelIia is characterized by arc-related volcanosedimentary sequences, although the Lower

Jurassic Kunga and Maude groups are predominantly sedimentary on the Queen Charlotte Islands. The Yakoun

River section is the designated type section of the Whiteaves Fonnation within the Maude Group (Section 15 in

(Cameron and Tipper, i985)). The stratigraphy of the section is shown on Fig. 3.2. The Whiteaves Formation consists of dark, poorly bedded concretionary mudstone with minor sandstone interbeds and rare, thin bentonitic ash fayen. [t records deposition in a marine basin oFmoderate water depth (Cameron and Tipper, 1985). There is no apparent gap within the exposed section. The Whiteaves Formation is comformably overlain by thin-bedded to massive sandstone of the Phantom Creek Formation although the basal part of that unit is largely concealed. Yukon

Albert,

Columbia i..--.- Pacific Ocean9- ' U.S.A.

Figure 3.1. Location map of the Yakoun River section on Graham Island, Queen Charlotte Islands. Grid reference for the U-Pb sarnple locality is UTM Zone 9,68 1500E 592 1830N. Legend . . ,massive tuff layer 1.. , -.sandstone YAKOUN RIVER

: bedded SECTION

Figure 3.2. Litho- and biostratigraphy of the Toarcian of the Yakoun River section (modified after Jakobs et al., 1994). "2" denotes the level of U-Pb zircon sample. Ammonite ranges (vertical lines) with collection levels (horizontal bars) are shown for the Planulata, Crassicosta, and Hillebrandti zones only (after Jakobs, in press (1997)). Dotted lines denote imprecisely located early collections. 3.3. U-PB GEOCHRONOMETRY

3.3.1. Methodology

Sample PCA-YR-1 was collected from a IO-cm thick, clay-rich, white gritty ash bed interbedded within black

mudstone of the Whiteaves Formation in the Yakoun River section. The sarnple was excavated with a knife and

spoon and stored in a rnoist state. ln the laboratory, the sarnple was blended into a sluny, allowed to senle, and the

fines decanted. The washed concentrate was further ulmonically agitated and cleaned. passed through heavy

liquids, and separated into aIiquots with different magnetic susceptibilities. Euhedral needles of zircon were

identified in the least magnetic material, hand picked, photographed, and abraded (Krogh. 1983) for U-Pb analysis.

Crystals selected were reasonably homogeneous, but allowing for the possibility that xenocrysts or transponed zircons may have been present, the first set of analyses were of single grains weighing only a few micrograms. The concordance and reproducibility of these analyses established a high probability of a single population of zircon. and additional analyses consisted ofup to 10 grains to achieve somewhat smailer U-Pb analyticai erron. The methods and decay constants used in U-Pb analyses are described eisewhere (Parrish et al.. i987; Roddick et al..

1987; Steiger and Mger, 1977). Table 3.1 presents these data and errors (Roddick, 1987), and they are shown graphically on Fig. 3.3. U and Pb blanks were approxirnately 0.2 and 3 picograrns, respectively.

3.3.2. Results

Six analyses of single or multipIe grains of abraded zircons al1 produced consistently concordant and overlapping error ellipses on the concordia diagram (Fig. 3.3). In Jurassic zircons the abundance ratio of '"~b/'~'Pb is about 20, therefore the '06~b/'3E~age fiom single grain analyses is inherentIy more precise than the '07~b/sS~ age and it reliably estimates the crystallization age when concordance of multiple analyses is demonstrated and the

206 P~/~~uages form a tight cluster(Mundil et al., 1996). The strong abrasion of the grains (Krogh. 1982), their mutual concordance, and agreement in their "PbtZ8~ agas definitively argue against any subsequent Pb loss in the analyzed crystals. The weighted rnean of '"Pb??J ages based on al1 six fractions is 18 1.k0.6 Ma (2a). We take a more conservative approach by selecting the three most precise analyses (A- 1, B-3. C-2). and noting their concordance, we use the mutual overlap of their '06~b/"E~ages as the best estimate of the true age and uncertainty and assign 18 l.4*1.2 Ma (2a) as the crystallization age of zircons. a s. single grain; *, weight estimatcd. b Weighing error 0.00 1 mg. c Radiogenic Pb. d Mcafured ratio, comcted for spik and Pb fra~tionatinnof 0.09Y&O.O30/dAMU. e Total common Pb in analysis corrected for hctionation and spike. / Corrccted for blank Pb and U. and model common Pb composition (Shccy and Kramers. 1975) for 181 Ma Pb: cmrs lrrr I standard cmr of the mean in percent. g Corrected for blank and common Pb: enors arc 2 standard crron of the mean in Ma

Table 3.1. U-Pb zircon analytical data of sample PCA-YR-1

0.032 -

PCA-YR-t tuff layer in Whiteaves Fm. 0.030 - Yakoun River

3 m (? N \ 0.028 -

Ns

lnterpreted age: 181.4 î1.2 Ma

0.024

Figure 3.3. U-Pb concordia diagram for zircons from sample PCA-YR-1, a volcanic ash layer in the Whiteaves Formation at Yakoun River. Errors are shown at 20 (95%) level. Error ellipses with thick outline denote single grain analyses, grey shading indicates fractions not used in final age interpretation (see text for details and Table 3.1 for anaiytical data). 28 3.4. BIOCKRONOLOGY

Biochronologic dating is based on sequences of associations of taxa that have been shown to maintain their stratigraphie relationships over a wide geographic area. This approach is well-suited to ammonoids and has been successfilly used in developing zonations for the North American Lower Jurassic (Jakobs et al., 1994; Smith et al..

1988). The use of the new U-Pb date as a tirne scale calibration point requires correlation between the secondary standard North American and the primary standard northwest European zonations (see (Callomon, 1984) for discussion). Here we develop such a correlation and apply additional rnethods to test for biochronologic error.

3.4.1. The ammonite fossil record

The Whiteaves Formation in the Yakoun River section is richly fossiliferous. The rnost common rnacrofossils are ammonites with a total of 1 19 collections obtained through initial reconnaissance work (Cameron and Tipper,

1985) followed by three field seasons of carefil systernatic sarnpling (Jakobs, in press (1997)). This collection history and the abundance of fauna distinguishes the Yakoun River section as one of the best studied Early Jurassic ammonite localities in North America. Well-preserved ammonites occur rnost commonly in non-septarian calcareous concretions, the majority of which are distributed randomly in the section while some form distinct layers. Although rare, ammonites were also found crushed within the shale [G. Jakobs. pers. cornm. 19961. The uniform Iithology throughout the Whiteaves Formation suggests that there is no significant preservationat bias affecting the vertical distribution of fossils which was carefdly rneasured and docurnented (Jakobs, in press (1997):

Jakobs et al., 1994).

Five regional standard ammonite zones were proposed for the Toarcian of North America (Jakobs et al..

1994). All of them are recognizable in the Yakoun River section which serves as the stratotype for the Planulata.

Crassicosta, and Hillebrandti zones (Jakobs et al., 1994). The ash layer yielding the analyzed U-Pb sample is by definition assigned to the lower part of Crassicosta Zone. Therefore the following discussion is concerned with the

Crassicosta and the adjacent Planulata and Hillebrandti zones only.

3.4.2. Reliability of observed ranges

The regional standard zona[ scherne for the Toarcian (Jakobs et al., 1994) is based on a Iiteral reading of observed vertical ranges ofammonites. It is widely accepted, however, that the fossil record is incornpiete (Paul. 1982) and it underestimates the meranges (Signor and Lipps, 1982). Our aim in assessing the robusmess of

ammonite zones as defined in the Yakoun River section is twofold: to use rigorous statistical methods to test the

validity of zones and to compare the empirical data derived fiom successive collecting campaigns with the

predictions from statistical methods. The fotlowing caIcuIations are based on published fossil occurrence data

(Jakobs, in press (1997); Jakobs et al., 1994).

The simplest way of providing a maximum estirnate of the completeness of the ammonite record is to calculate a "hitlmiss" ratio, where the hits are confirmed occurrences and the misses are collection levels with no record of a given taon within its known vertical range (aIso termed virtual occurrences (Guex, 199 1)). For the 54 levels reported within the three zones, there are 74 hits and 1 10 misses suggesting that the maximum average probability of finding a species in a fossiliferous horizon within its vertical range is no bener than 40%. It can be argued that the average probability is biased towards lower values by rare taxa which are given little weight in devising biozonations. Of the 22 taxa considered, onIy six (Rarenodia plantilara, Phymatoceras crassicosfa, P. hillebrundti. P. cf. pseudoerbaense, Leukadiella ionica, and Denckmannia tumefacru) were found at more than two levels. Indeed, these six taxa score a hithiss ratio of 51/54 giving a nearly 50% chance of being found. Not surprisingly, the three zonal index species among them score the highest (hit/miss=36/30+chance of being found=64%).

It is evident that misses would occur beyond the lowest and highest hits, Le. the observed ranges are shorter than the me local ranges. To estimate the meranges, gaps separating occurrence levels rnust also be considered

(Paul, 1982; Springer and Lilje, 1988). Fig. 3.2 is a re-plotting of observed ranges (Jakobs et al., 1994) considering al1 available information on vertical distribution (Jakobs, in press (1997)). The distribution of gap lengths in the fossil record is show to have Dirichlet distribution and classical confidence intervals can be calculated for estimating the nue stratigtaphic ranges (Marshall, i 990; Springer and Lilje, 1988). These statistics are only valid if the fossil distribution is assurned to be randorn. As pointed out above, there is no appreciable prese~ationalbias or change in sedimentation rate in the rather uniform lithologic sequence of the Whiteaves Formation. Likewise, collection bias can also be discounted as coilection intensity was constant during the first field season and collecting strategy remained systematic (although slightly biased towards parts of the Planulata and Hillebrandti zones) in the following two campaigns [Ci.Jakobs, pers. comrn. 19961.

Extensions of the observed ranges for the six common taxa were calculated at the 95% confidence level

(Marshall, 1990) and plotted on Fig. 3.4. Note that the Iength of the range extension is a function of the length of the observed range and the number of occurrences within it. The statistical meaning of the range extensions is that there is less than 5% probability of finding a given species beyond the limits of the extended range. For the three 3 0 zonal index species, occurrences are separated according to the year of their first finding and changes in cumulative observed ranges are also plotted. The following observations cm be made: (1) For commonly occurring ammonite taxa collected through systematic collecting effort, range extensions at the 95% confidence are not justifiable. A lesser degree of confidence would better agree with range extensions realized through further collecting but its approximate IeveI needs to be established empirically through tests on a larger database. (2) The Planulata Crassicosta, and Hillebrandti zones are defined as assemblage zones (Jakobs et al., 1994). in practice the detinition of their boundaries at the stratotype depends, to a large degree. on the range of their zonal index species, The

Planulata and Hillebrandti zones are based on robust evidence ffom the distribution of their respective zonal indices. Their range extension even at the 95% confidence level would not modify the zonation significantly. The

Crassicosta Zone, on the other hand, is Iess well-defined by the range of its zonal index but is well-constrained by the sub- and superjacent zones and the cohort of CO-occurringtaxa. (3) Of the six range end-points of other relatively common species, only one would require modification of zonal assiçnments if the range extension at the

95% confidence level is considered. (4) The isotopically dated tuff layer (within the observed range of

Phymatoceras crussicosta) falls very near to the top of the extended range (at the 95% confidence level) of

Rarenodia planulata, therefore its assignment to the Crassicosta Zone is undoubted. Figure 3.1. Collection levels (arrowheads), obsewed ranges (unfilled boxes), and estimated maximum ranges at 95% confidence level (shaded boxes) of common ammonite species (i.e., occurring at more than two levels) fiom the Planulata, Crassicosta, and Hillebrandti zones in the Yakoun River section. Changes in observed and estimated ranges afler successive collection years are show for the three zonal index species (Rarenodia planulata, Phymatoceras crassicosta, and P. hillebrandt~).Occurrence data fiom (Jakobs, in press (1 997)). "Zdenotes the level of U-Pb zircon sample. See text for discussion. 3.4.3. Biochronologic correlations

Traditional ammonite biochronology was founded and proved extremely powerful within single bioprovinces

(e-g., Boreal northwest Europe, Tethys) where it is possible ro establish reproducible hi& resolution uni& (faunal

horizons). Limitations and dificulties are imposed by facies dependence and limited paleobiogeog-aphic

distribution of taxa resulting in complex timelspace distribution patterns overprinted by the vagaries of fossil record

(Le., preservation and collection biases). Correlation across biogeogmphical boundaries is often controversial. as

exernplified by long-standing debates on TethydBoreal correlation in the Toarcian of Europe (Elmi et al.. 1994).

The traditional approach is based on expert judgement in emphasizing the correlation value of certain ammonoid

taxa and their associations at the expense of others when discrepancies in the first and last appearance datums

(FADILAD) are detected.

A traditional solution for the correlation of North American Toarcian ammonite zones with northwest

European standard zones (sensu (Dean et al., 196 1)) equates the PlanuIata Zone with the Bifrons and basal

Variabilis zones, the Crassicosta Zone with the rest of the Variabilis Zone, and the Hillebrandti Zone with the

Thouanense Zone (Jakobs et al., 1994). It is acknowledged that (1) correlation is hampered by die absence of key

European taxa in North Arnerican faunas (e-g., Midocerus and Haugia, to which index species of the Bifrons and

Variabilis zones belong); (2) North American faunas have greater similarity to Tethyan and South American ones than those of Boreal northwest Europe; (3) in North America, the FAD and/or LAD of several important genera

(e-g.. Phymatoceras. Podagrosires, Peronoceras. Mercaticeras) are anomalous with respect to those found elsewhere (Jakobs et al., 1994). Consequently, a cautious approach to correlation needs to integrate global stratigraphie distribution from al1 faunal provinces, should seek maximum ranges of taxa, and is expected to result in a correlation scheme with apparent resolution sacrificed for increased reproducibility and confidence, The amount of available data calls for quantitative treatment which also provides the advantage of eIiminating potential bias introduced by subjective judgements.

There is a variety of quantitative biosûatigraphic techniques available and the choice depends on the nature of data and the expected outcorne (Edwards, 1982). In our case, the data are obtained from measured sections in different sedimentary basins and bioprovinces, hence the total number of taxa is high p-100) while the number of comrnon taxa is often low. Also, fossil distribution among sections rnay be non-random due to faunal migration. As the timing of migration events is not known independently, maximum ranges are sought to use for correlation.

Among the most widely used and tested quantitative rnethods, graphic correlation seeks maximum ranges but is not practical beyond a single sedimentary basin (Mann and Lane, 1995). The probabilistic ranking and scaling method

3 3 also works best within a single basin/bioprovince where it produces average ranges with the maximum liketihood of FADLAD sequence while assuming random distribution of fossils (Agterberg, 1990). The unitary association

(UA) method (Guex, 1991) appears to be best suited to our problem. It furnishes maximum ranges based on a deteministic approach and it does not require random fossil distribution or homogeneity of source data. It applies the rigor of graph theory to the familiar concept of Oppelian assemblage zones thus it retains the philosophy underlying traditional ammonite biochronology. The pitfalls of intuitive correlation schemes are minirnized. The algorithmic formulation of the UA method resulted in an efficient computer program, the BioGraph (Savary and

Guex, 1991). The UA method was found to eficiently constmct biochronologically meaningful zonations from complex data (Baumgartner, 1984). Notably, it was successfully used for correlation between different bioprovinces (Baumgartner, 1993) and was demonstrated to closely reproduce the ammonite zonation developed by traditional methods in northwest Europe (Dommergues and Meister, 1987).

We used the following procedure for computer-assisted biochronologic correlation. Conventional correlation of the Planulata, Crassicosta, and Hillebrandti zones (Jakobs et al., 1994) was accepted as a guide. The scanning range was set to latest EarIy to early Late Toarcian. All established ammonite provinces were considered and the available literature was culled for representative sections from each province. The selection was based on evidence that collections were made fiom measured sections spanning several zones with no indication of condensed horizons or reworking, that the ammonite fauna is abundant and diverse, and that local ranges are well- documented. Sources wiih sound taxonornic documentation (preferably with illustration) were chosen if possible.

No adequate sections meeting the above criteria were located f'rom the western Pacific and Arctic provinces. Beside the Yakoun River section, data from one other North American, four South American. five western Tethyan and five northwest European sections were compiled (see Table 3.2 for details of sources and remarks). Amans the taxa occurring, only species of Ammonitina reported from two or more localities were used. Poiyplecrtcs and

Pseudoliocerus, two long-ranging genera with Iittle stratigraphic value, were omitted. Limited anempt was made to homogenize taxonomy (see discussion below). mainly to consider synonymies of taxa reported from the Yakoun

River (Jakobs, in press (1997)). Composite genus ranges were added for each section to mitigate the effect of paleobiogeographic differences at the species level. A conservative generic classification scheme (Donovan et aI..

1981) was adopted with the exception of recognition of Rarenodia (Jakobs, in press (1997)) and the combining of

Porpoceras with Peronocerar to avoid confusion stemming from different opinions as to their species content. The

103 taxa and the stratigraphic ranges in each section are listed in Appendix 1. LOCAUrY REFERENCE

NORTH AMERICA 1. Yakoun River, British Columbia Jakobs (in press. 1997); Jakobs et ai. (1 994) 2. Joan Lake, British Columbia Jakobs (in press. 1997); Jakobs et al. (1994)

WESTERN TEfHYS 3. Valdorbia. Urnbria, ltaly Cresta et al. (1989); Gallitelli Wendt (1970) (As bed-by-bed correlation between the two sources are ambiguous at some levels, data fmm the two sources was first processed by BioGraph and the resuiting taxa ranges wilh UA treated as levels were entered into the main BioGraph input file.) 4. Djebel-es-Saffeh (Section 3B), Djebel Nador EImi et al. (1974) area, Algena 5. Djebel-es-Saffeh (Secüon 2), Djebel Nador Elmi et al. (1974) area. Algeria (Supplementary section to Djebel-es-SaRh 3B, mainly to better document the ammonite fauna fmthe lowerpart of the succession.) 6. Paghania, Greece Kottek (1966) 7. Monte di Civitella. Umbna, ltaly Venturi (1975)

NORTHWEST EUROPE 8. St-Paul des Fonts. Aveyron, France Guex (1972) (Ammonite faunas of horizons, said to represent one to thebeds. are given and entered here as ievels.) 9. Camplong. Aveyron, France Guex (1975) IO. Rida and La Almunia, lberian Range, Spain Goy and Martinez (1990) 11. Anse Saint-Nicolas, Vendée, France Gabilly (1 976) (Fmm the vicinity of Thouars, the classical type area for the Toarcian. this section contains the best documented and most diverse ammonoid fauna.) 12. Ravenscarand Whitby, Yorkshire, England Dean (1954); Howarth (1962); Hawarth (1992) (Data compiled followhg the bed-by-bed correhfion belween the two sections (Howarfh, 1992).)

SOUTH AMERICA 13. Quebrada El Bolito. northem Chile Hillebrandt and Schmidt-Effing (1981) 14. Quebrada Yerbas Buenas, northem Chile Hillebrandt and Schmidt-Effing (1981) 15. Quebrada Larga. northem Chile Hillebrandt and Schmidt-Effing (1 981) 16. Rio del Toro. northern Chile Hillebrandt and Schmidt-EfEng (1981) (For al1 South American section, emended taxonomy was used (Hilebrandt, Ig87).)

Table 3.2. List ofstratigraphic sections and sources of BioGraph input.

ïhese data were entered into and processed by the BioGraph program (Savary and Guex, 199 1). Sections within each province were ordered with priority given to the most species-rich and complete one. As the resotution of conmdictory stratigraphic relationships may be affected by the order in which the data are processed (Guex,

199I), al1 six permutations of North American, Tethyan and northwest European sections were tried with the least informative South American sections consistently entered 1st.The output of 40 successive UA (Fig. 3.5), however. was not sensitive to the permutation.

The UA assigned to specific beds or collection levels in the source sections were determined from the correlation tables produced by BioGraph (Appendix 2). Thus at the Yakoun River, the base of the Crassicosta Zone corresponds to UA 22-23, the base of the Hillebrandti Zone to UA 33, and the isotopically dated tuff layer to UA

22-25. Standard subzonal and zonal boundaries reported from the analyzed northwest European sections were used to determine the maximum permissible extent of these uni& in terms of the UA scheme (Fig. 3.6). (Lacking a universally accepted Toarcian standard zonation for northwest Europe, we adopt a recently updated scheme (Elmi et al., 1994) which differs only slightly from a more traditional zonation (Dean et al., 196 1). Notably. the subzones of the Bifrons Zones are based on the species sequcnce of Hildoceras. the Variabilis Zone is subdivided into three subzones. and the subzonal scheme within the Thoumense Zone is revised, increasing the correlation potential

(Elmi et al., 1994)) îhe most Iikely range. as expressed by the overlap or minimum required range of UA frorn the anaIyzed sections, was also inferred for the standard units. It is evident from Fig. 3.6 that individual UA are seldom reproducible in more than one or two sections and across bioprovinces. However, their groupings which correspond to naditional zones or subzones, are commonly present in several sections from different provinces.

OnIy four UA are unambiguously identifiable in North America and they are not directly correlatable with other provinces.

Fig. 3.6 is used to determine the most likely and the maximum permissible correlations of the Crassicosta

Zone (and the isotopically dated level within it) with the northwest European standard zonation. The beginning of the zone appean equivalent or marginally older (but cannot be younger) than that of the standard Variabilis

Subzone (Variabilis Zone). It could also be correlative with (but not older than) the later part of Semipolitum

Subzone (Bifrons Zone). The beginning of Hillebrandti Zone lies most Iikely within the latest Variabilis Zone

(Vitiosa Subzone) but it could be as old as the Illustris Subzone (middle part of Variabilis Zone) or as young as the earliest Bingmanni Subzone of the Thouarsense Zone. The isotopically dated tuff layer is constrained to the

Semipolitum through Illustris subzones (inclusive), with the earlier part of the Variabilis Zone being its most likely age.

SECTIONS 1111111 1234567890123456

NAm. fethys NW Europe S Am.

Figure 3.6. Reproducibility of the 40 Unitary Associations recognized in latest Early to early Late Toarcian ammonite faunas analyzed using the BioGraph program. The left side of diagram shows the occurrences of UA (solid bars) in the 16 selected sections (listed in Table 3.2). Ploned on the right side are minimal (solid bars) and maximal (hatched bars) groupings of UA corresponding to the northwest European standard zones and subzones (Elmi et al., 1994) and their correlation with the North American Crassicosta Zone (box) and the isotopically dated level (gray shaded) within it. As expected, the quality of UA based correlation decreases towards either end of the scanning range (Le.,

earliest Bifrons Zone and latest Thouarsense Zone) where there is insufficient superpositional control deduced fiom the raw data. With this caveat, the correlative of the base of Planulata Zone (UA 7-14) appears to lie somewhere between the late Sublevisoni and early Semipolitum subzones (Bifrons Zone) The top of Hillebrandti Zone (UA

38) shows robust correlation with the Thouarsense Subzone but correlation with any of the subzones (Bingmanni through Fallaciosum) of the Thouarsense Zone is permitted.

3.4.4. The effect of taxonomic noise

In addition to differences between observed and true fossiI ranges and correlation uncenainty .a third potential source of error in biochronologic dating is the uncertainty of fossil identification. Given the variety of authors and their differing opinions on species, taxonomic noise (Le., random inconsistencies in identifications) is expected in the source data for biochronologic correlation. The following exampIe is given to illustrate the effect of hi& Ievels of taxonomic noise.

The smtigraphically important genus Hildocerm was recently revised (Howarth, 1992). Four widely distributed species were accepted with many orher nominal species synonyrnized (Table 3.3). Tabulated are the revisions made to figured specimens from sources used in our correlation database (Elmi et al.. 1974; Gabilly.

1976; Gallitelli Wendt, 1970; Guex. 1972; Kottek, 1966). in approximately 25% of the cases, the original author's identification is revised in addition to reassignments of synonyrnized nomina1 species. In three cases. specimens originally figured under one name are assigned to two different species by the reviser. It follows that even with amply illustrated material, the stratigraphic distribution data cannot be unambiguously revised without a comprehensive review of al1 considered specimens, a daunting task not atternpted here.

If we adopt this high figure of 25% erroneous identifications at the species levej. we can then examine the effect of this taxonomic noise on the UA correlation method. The simplest mode1 calculation is to construct UA from two species (A and B), both represented by one specimen at two leveis of a hypothetical stratigraphic section

(Fig. 3.7). The stratigraphic ranges of A and B cm be concurrent, overlapping, or exclusive. There is a single UA in case of concurrent or overlapping ranges whereas two successive UA are identified for exclusive ranges. Let one specimen Le., one quarter of the total, be misidentified (A erroneously called B or B called A). In each case, there are four permutations to place the misidentified specimens. Recording the changes in UA and the proportion of number of levels with erroneously assigned UA provides a measure of sensitivity of the UA method to taxonornic Fiifdoceras luficosia BELLHl H. /usifunicum MEiSiEIt H. b$kms (BRUGUIERE) H. semipoliium BUW

(=ilsuble*oni FUCMI; H. (=H. graecwn &W (=H.aperhun GABILLY) (=H angusrisiphonarum PRINZ; carerinii MERU) H cmmm Mrnomuros: H. H. semicosfa BUCWAN) iethysi Gt(2YI

H graenim stiblevisoni (Konck, ? H. sunipolilum (Kottek 1966) H. semigoliisrm (iioitek, 1966) H. xemipoli~um(Guex. 1972) L 966) <-> ? Hildoiier Ievhoni (Gallitelli ff. gruemm graecum (Kottek. H. sedcosta (Guex. 1972) H. semicosru (GUCL 1972) Wendt. 1970) 1966) <-> if. sublevisoni (Elmi et al.. 1974) ff. groecum lusitaninun (Kottck, bi/roru (Guex. 1972) H angusrlriphunutu (Guex. 1966) 1972) H mblewisoni (Gabilly, 1976) H. sublevisoni (Gallhelli Wendt. H. bijrom (Etmi ct al.. 1974) H. bifrm (Elmi cc al.. 1974) 1970) c-> H. carerinii (Gabilly. 1976) ff. sublcvisoni (Guec 1972) H. apertum (Elmi et al.. 1974) H bi/ons var. mguriisiphoncira ? H. graecurn (Gucx. 1972) bi/rom bfionr (Koitek. 1966) (Elmi et al.. 1974) ? H. fusitanicum (Gu=. 1972) H. bi/ronr ivalcofi (Kottek. 1966) iL wnipolitwn (Elmi et ai.. 1974) H. Iusiranicum (EIrni et al., 1974) H. bijrom bijrom (Gallitelli H. ~emipoltium(Gabilly. 1976) Wend!. 1970) H rc~hysi(Gabilly, 1976) H. burons angustisiphonarum H. msum(Gabilly. 1976) (Gallitelli Wendt. 1970) H. Iusitunicum (Gabilly. 1976) H. semipoli~um(Gallitelli Wrndt 1970) H. aperhun (Gabilly, 1976) fi. bifrom (Gabilly. 1976)

Table 3.3. Synonymies of species of Hildoceras figured in sources used in our biochronoIogic correlation database (Elmi et al., 1974; Gabilly, 1976; Gallitelli Wendt, 1970; Guex, 1972; Kottek, 1966), as recognized in the revision of Howarth (Howarth, 1992), Subjective junior synonyrns are given in parentheses in the heading (Howarth, 1992). Bold tàce indicates disagreement between the reviser and the original author afier reassignment of synonyrnized nominal species. Arrows mark specimens originally figured under one name that are now assigned to wo different species (Howarth, 1992).

CORRECTID 1 25% ID ERROR CONCURRENT RANGES UA UA UA UA Tl1AB 1 FI1 BB 1 m; OVERLAPPING RANGES FI; i EXCLUSIVE RANGES

Figure 3.7. Simulation of the effect of 25% taxonomie noise on the unitary associations (UA). Left column: stratigraphic relationships of species A and B known from two Ievels in each section and derived UA. Other four coiumns: permutations of one misidentified specimen (in boId italic type). Heavy border denotes modification in UA pattern compared with Iefi column, circled UA number denotes levels of critical error. See text for discussion. noise. We shall only consider the critical emrs where the misidentification leads to the addition of artificial UA or to a shift of UA boundary. If the result is the merger of two UA, the net effect is a decrease in resolution and loss of correlation power without introduction of correlation error.

As shown on Fig. 3.7, there are two permutations for the overlapping ranges and two for the exciusive ranges where a total of five levels suffer critical error in UA assignment. A cornparison with the total number of levels considered in al1 scenarios (36) reveals that at the 25% taxonomic noise IeveI, the critical stratigraphic error is less than 14%. A remarkable property of the UA method is that increase in the density or complexity of source data greatly reduces the sensitivity to taxonomic noise. In the mode1 calculation. it can be demonstrated that a twofold increase in the number of occurrence levek (from hvo to four) leads to less than 3% critical stratigraphic error at the sarne 25% taxonornic noise level. Furthemore, we note that misidentification rnost ofien occurs among rnorphologically sirnilar forrns that are often phylogeneticalIy closely related and have simiiar stratigraphic distributions. Taxonomie mistakes invoIving taxa with concurrent ranges wiI1 not introduce stratigraphic error. The amount and complexity of data used in the world-wide correlation together with the foregoing considerations lead us to regard the contribution of taxonomic noise to biochronologic dating error as negiigible.

3.5. DISCUSSION

The results of the correlation using the UA rnethod can be compared with those obtained traditionally (Jakobs et al., 1994). The boundary of Planulata and Crassicosta zones, previously correlated with the early part of

Variabilis Zone, is now allowed to fail within the interval of Semipoliturn to Variabilis subzones. with the greatest likelihood indeed near the base of the Variabilis Zone. The boundary of Crassicosta and Hillebrandti zones. previously equated with the base of the Thouarsense Zone, is now ailowed to faIl within the interval of IIlustris

Subzone to the base of Thouarsense Zone, most Iikely within the Vitiosa Subzone.

As demonstrated above, the maximum uncertainty in biochronologic age assignment of the isotopically dated sample is one standard subzone. It is superior to al1 previously available dates which are constrained at the stage level at best. The precision of the U-Pb age also surpasses that of any other isotopic age fmm, or near, the Toarcian. nius the U-Pb age from near the middle of Middle Toarcian at Yakoun River serves as an important new calibration point for the Jurassic time scale. When plotted against the major recently published and widely used time scales, the new U-Pb date of 18 1.4* 1.2 Ma falls within the Toarcian in al1 except the DNAG (Kent and

Gradstein, 1985; Kent and Gradstein, 1986; Palmer, 1983) seale (Fig. 3.8). The mid-point of Toarcian is within the error of our date in the EXX88 (Haq et al., 1987; Haq et al., 1988) and GTS89 (Harland et al.. 1990) scales only.

41 The new date is significant in that it, together with other newly obtained Early Jurassic isotopic ages, will help

greatly reduce the large uncertainty associated with previous stage boundary estimates. Rigorous error estimates are

only given in GTS89 (Harland et al., 1990) and MTS (Gradstein et al., 1994; Gradstein et al.. 1995). Considering

the error ranges, in GTS89 (Harland et al.. 1990) the base of the Toarcian can be as young as 172 Ma and the top as

old as 188.5 Ma, both outside the error lirnits of the new U-Pb date. in MTS (Gradstein et al.. 1994; Gradstein et

al.. 1995), the top of the Toarcian can be as old as 184.1 Ma which is in conflict with the U-Pb date reported here.

- 175 NEW U4b DATE - MID-TOARCIAN ODIN ------'lm ,O - ~,,,.,,'= - i EXX88 cc GTS89 'iw 186 f3.5 4 * - -- 187.0 t15 MTS 187 188.6 r4.0

Figure 3.8. Comparison of boundary age estimates and numerical mid-point (marked with arrowheads) of the Toarcian in some recent time scales and the new U-Pb date from the mid- Toarcian of Queen Charlotte Islands. Sources of time scales quoted: DNAG (Kent and Gradstein. 1985; Kent and Gradstein, 1986; Palmer, 1983); EXX88 (Haq et al., 1987; Haq et al., 1988); GTS89 (Harland et al., 1990); Odin (Odin, 1994); MTS (Gradstein et al., 1994; Gradstein et al., 1995).

3.6. CONCLUSIONS

We report a newly obtained U-Pb zircon age of 18 1.41 1.? Ma (20) from the Queen Charlotte Islands. The

interpreted age is based on concordant and overlapping analyses of single-crystal as well as muiti-grain fractions.

The dated sample was collected from a bentonitic ash layer within the Toarcian Whiteaves Fonnation in the

Yakoun River section. The ash layer lies within the North American srandard Crassicosta Zone as defined by ammonite biostratigraphy (Jakobs et al., 1994).

The Yakoun River is the designated type section of the Crassicosta Zone as well as the subjacent Planulata and superjacent Hillebrandti zones (Jakobs et al., 1994). We statistically anaIyzed the quality of the ammonite fossil record in this section. All three zones are assemblage zones by definition but in practice are delimited at the stratotype by their respective index species. The collection density of Rarenodia planulata and Phyinatoceras hillebrandti is adequate in that their range extension using 95% confidence intervals would not modify the pIacement of the zona1 boundaries significantly. Collection data fmm three successive field seasons (Jakobs, in press) suggest that generally rnuch less than 95% confidence level is required when proposing range extensions for comrnonly occurring ammonite taxa. The assignrnent of the dated tuff Iayer to the lower part of Crassicosta Zone wouId not change even when range extensions at the 90% confidence level are considered.

Biochronologic conelation was done using the cornputer-assisted unitary association (UA) method (Guex.

1991). Ammonite local range data from representative North American, western Tethyan, northwest European, and South Amencan sections were processed to provide unbiased estimates of global maximum taxon ranges and a sequence of UA. The UA framework was then used to establish the maximum extent of permissible correlation between the secondary standard North Arnerican and the primary standard northwest European zonations.

Correlation of the Crassicosta Zone is bracketed by the Semipolitum Subzone (late Bifrons Zone) and Bingmanni

Subzone (early Thouanense Zone). In particular, the dated tuff layer cannot be older than the Semipolitum

Subzone or younger than the Illustris Subzone (Variabilis Zone).

Random misidentifications are known to occur in the world-wide dataset used for correlation. However, the

UA rnethod is shown to respond to taxonomic noise by loss of resolution rather than erroneous correlation. Thus we regard the proposed correlation as conservative best estimates.

With less than *l% (20) error in the isotopic age and a maximum ofione standard subrone uncertainty in biochronologic correlation, the U-Pb age from the Yakoun River section serves as an important calibration point for the Early Jurassic time scale. A cornparison with recently proposed tirne scales reveals that only rninor adjustments are required for the Toarcian stage boundary age estimates but the associated uncertainties can be greatly reduced. U-Pb dating of volcanic horizons within fossiliferous sequences in the North American Cordillera hoIds promise for providing more useful calibration points for the Jurassic time scale. INTEGRATED AMMONITE BIOCEIRONOLOGY

AND U-Pb GEOCHRONOLOGY FROM A

BASAL JURASSIC SECTION IN ALASKA

4.1. INTRODUCTION

Lowennost Jurassic rocks are rare world-wide and their dating and correlation is often difficult. Fossiliferous marine sediments are scarce and ammonites are known From only a few localities. The Triassic-Surassic boundary is marked by one of the five major Phanerozoic mas extinction events but its age and the timing ofpost-extinction recovery is poorIy constrained. Numerical estimates of the age of the Triassic-Jurassic boundary vary widely between 2 13 and 300 Ma for tirne scales published in the last 15 years (Pal@, 1995). The main problern. the srnall number of relevant isotopic ages is amibutable to the scarcity of databIe and stratigraphically well-constrained volcanic rocks near the boundary. The Cordilleran Jurassic CaIibration Project is an initiative to irnprove the

Jurassic time scale through U-Pb dating of volcanic units whose age is well-constrained by ammonite biochronology From the North American Cordillera (Piilfi et al., 1995).

To refine the time scale near the Triassic-lurassic boundary, we studied the Puale Bay section on the Alaska

Peninsula (Fig. 4.1) because it was long known to contain fossififerous uppermost Triassic and lowenost lurassic sedimentary strata and volcanics (Capps, 1923; Smith and Baker, 1934; Martin, 1926; Kellum et al., 1945). The rnonographic tteatment of its Hettangian ammonoids by Imlay (198 1 ) indicated the presence of earliest Jurassic faunas. Moreover, Newton (1989) suggested a continuous Triassic-Jurassic transition. a situation that is only known at a hancifiil of localities world-wide (Hallam, 1990). PUALE

BAY

-- -

Figure -1.1. Location map of the Puale Bay section. The studied section is Iocated on the southeastem shore of Puale Bay, on the rugged coastline of the Alaska

PeninsuIa, across tbe Shetikof Strait flom Kodiak Island (Fig. 4.1). A wave-cut intertidal piatform and adjacent

coastat bIuffs provide excellent exposures. Beds dip moderately to the northwest and their continuous erosion by

waves aid macrofossil collecting.

in 1995, we examined the Triassic-Jurassic transition. coIlected ammonoids and other macrofauna horn 26

levels in the Hettangian-Sinemurian strata, measured the basa1 Jurassic stratigraphic section, and sarnpled

potentially zircon-bearing volcanic units for U-Pb dating,

in this report we: (1) document the Early Jurassic ammonoid succession in the PuaIe Bay section based on the

new colIections; (2) establish a biochronologic fiamework to constrain the dated volcanic uni& at the zonai level:

(3) report three new zircon U-Pb ages; (4) discuss the significance of new data in comparison with recent time

scales; and (5) reconsider the status of the Puale Bay section as a Triassic-Surassic boundary section.

4.2. GEOLOGIC SETTING AND PREVIOUS WORK

South-central Alaska is a collage of acronostratigraphic terranes. The three largest. the Alexander,

Peninsular. and Wrangellia terranes are thought to have arnalgamated into the Wrangellia composite terrane prior to their accretion to North Amenca in the Crecaceous (Nokleberg et al., 1994 and references therein). The Puale

Bay area foms part of the Peninsular terrane, also known as the Alaska Peninsula terrane (Wilson et aI., 1985). Its

Triassic and Jurassic stratigraphy records the geological evolution of a volcanic island arc (Wang et al., 1988) that is cIosely Iinked to Wrangellia on the basis of shared shailow marine carbonate, clastic, and volcanic (mainly volcaniclastic) sequences.

The Puale Bay section was recognized as one of the most complete and relativeiy undeformed Triassic and lurassic successions in south-central Alaska. Eady work is summarized by Imlay and Detterman ( 1977). Based on a comprehensive evaluation of al[ Early Jurassic ammonite collections made in the area (including neighbouring

Alinchak Bay), Imlay (1981) demonstrated the presence of i-iettangian rocks. The underlying Upper Triassic sequence was studied in detail by Wang et al. (1988). A complete (Upper Triassic to Upper Jurassic), measured stratigraphic section is found in Detterman et al. (1985). The most recent stratigraphic summary is given by

Nokleberg et a[. (1994). îhe Upper Triassic Kamishak Formation is the oldest Mesozoic unit at Puale Bay. It comprises some 700 m of shallow marine, biogenic carbonate that is, in its upper pa* interbedded with basaltic volcanic rocks (Wang et al., 1988). Wiison and Shew (1992) obtained an imprecise whole rock K-Ar age of 197* 12 Ma (1o) for the basalt.

Dark grey, organic-rich, thin-bedded siliceous limestone in the upper part of the formation yielded ammonites

(Merasibirites. Rhabdoceras), bivalves (Monofis spp-), and the hydrozoan Heterastridium that indicate a Late

Norian age (Silberling in Detterman et al., t985, Wang et al., 1988, Newton, 1989, and this study). Above the fossiliferous Norian beds, there is some 50 m of suata apparently devoid of macrofossils that contain only trace fossils (Newton, 1989). We found the lowest lurassic ammonoid-bearing beds immediately above a fault of unknown (but probably not large) offset (Fig. 4.2). There is only a slight and gradua1 change upsection manifested in the decrease of carbonate content. Rocks on two sides of the fault are nearly identical. The Hettangian-

Sinemurian section above the fault was measured in detail and is shown on Fig. 4.3. It contains an abrupt transition fiom fine-grained, calcareous sediments to andesitic volcaniclastic rocks. The latter is referred to as the Talkeetna

Formation, a widespread, predorninantly volcanigenic unit of the Peninsular Terrane (Nokleberg et al., 1994). In the Puale Bay section, we sugzest drawing the base of the Talkeetna Formation at the lowermost massive green

Figure 4.2. Normal fault separating the lowest Jurassic (Middle Hettangian) ammonite- bearing beds to the left from unfossiliferous, Iithologically very sirnilar. presumably uppermost Triassic rocks to the right. tuff. This differs fiom the traditional approach of appraxirnately equating the TriayhHurassic system boundary

with the Kamishak-Talkeetna formation boundary by arbitrarily dividing the non-volcanigenic sedimentary

sequence (Nokleberg et ai., 1994 and references therein). The lower 120 m of the Tafkeetna Formation consists of

mainly green tuff, volcanic breccia and agglomemte with minor volcanigenic sandstone. The rocks are interpreted

as subaqueous deposits of a major volcanic episode in moderate proximity to the eruptive centres. Higher up, there

is some 200 m ofpredominantly green-grey. corne tuffaceous sandstone that is commonly thick bedded to massive

and locally cross-stratified. Shale interbeds occur only rarely. The abundant volcanic detritus was likely derived from the active parts of the TaIkeema island arc.

The HettangianSinernurian section is truncated by a fault that juxtaposes the Talkeetna Formation and the

several hundred metres thick, dark shale sequence of the Kialagvik Formation. The oldest ammonites from the

laner unit occur merely 6 m above the fault. From this level lrnlay (198 1) identified Haugia cf. compressa of

Middle Toarcian age that was later revised as Pleydellia maudemis, a guide ammonite of the latest Toarcian

Yakounensis Zone (Jakobs et al., 1994; Jakobs, in press (1997)). Our collection of Pleydelfiu sp. confirms the

presence of uppennost Toarcian. Beds a few metres upsection from this uppermost Toarcian level yielded

Tmetocerm, a characteristic Aalenian genus. Stratigraphically younger parts of the Puale Bay section are beyond

the scope of this study.

4.3. HETTANGIANSINEMUIUAN BIOCHRONOLOGY

Ammonoids are the predominant megafossi1.s in the Lower Jurassic section at Puale Bay. More than one hundred specimens were collected from 26 levels (numbered From the base up in Fig. 4.3) in a measured stratigraphic section. Bivalves, plant fossils, and nautiloids also occur but they do not rival the abundance and stratigraphic importance of ammonoids and are not discussed here. The ammonoids are rnoderately to poorly preserved interna1 molds that suffered postdepositional compression to varying degrees. In most cases, much important morphologie information such as whorl cross section, ventral features, and suture line are lost. The set of characteristics availabie for identification is commonly restricted to volution, whorl expansion rate, and ornamentation. The less than ideal preservation necessitates the frequent use of open nomenclature (Bengtson, l988), in several cases only a group of taxa cm be given which share the observed characters but their differences cannot be resolved in the absence of preserved distinguishing features. In some cases comparable species have been attributed to more than one genus. We indicate this by a question mark appended to the genus considered the most likely. The low

Figure 4.3. Measured lowest Jurassic stratigraphie section and ammonite ranges at Puale Bay, Alaska. Taxa identified at the species Ievel or assigned to a well-defined group of species are show in normal print. Bold typewriter font denotes confidence in genus level identification only or cases where closely comparable species are allocated to different genera. Normal typewriter font marks the most tentative assignments when a genus is not or only provisionally suggested. Plate 4.1. (On p. 52.) Middle Hettangian ammonite fauna of Puale Bay. All specirnens are 95% of natural sîze. Collection Ievels are shown on Fig. 4.3. Specimens are deposited in the type collection of the Department of Earth and Ocean Sciences, University of British Columbia under the type numbers with the prefut UBC.

A, D: Discamphiceras cf. silberlingi Guex

A: Levet 7, UBC 0 I8; D: Level 12, UBC 02 1.

B: Level 5, UBC 019; E: Level 10, UBC 022; F: Level8, UBC 023; H: Level 7. UBC 025.

Discornphiceras aff. reissi (Tilmann)

Level 13, UBC 020.

Srnoceras? sp.,

G: Levet 8, UBC 024.; J: Level4, UBC 027.

Smcoceras? ex gr. portlocki (Wright)

Level6, UBC 026.

Lytoceratid indet.

Level 14. UBC 028.

Pleuroacanrhifesex gr. mulleri Guex

Level9, UBC 029 (note parabolic nodes).

Franzicerm? sp.

Level8, UBC 030.

Mullerires cf. pleuroacanthitoides Guex

Level 1 1, UBC 03 1 and UBC 032.

Psiloceratid indet.

Level9 UBC 033 (counterparts of the same specimen, note nodes on inner whorls).

Kammerhrites ex gr. megastoma (GUmbel)

Level3 UBC 034. Plate 4.1 Plate 4.2

Plate 4.2, Late Hettangian through Early Sinemurian ammonite fauna of Puale Bay. All specimens natural size. Collection levels are shown on Fig. 4.3. Specimens are deposited in the type collection of the Department of Earth and Ocean Sciences, University of British Columbia, under the type numbers with the prefut UBC.

Sunrisites? sp., Level 16, UBC 035, (counterparts of the same specimen); Paracaioceras cf. rursicosfatum Frebold, C: Level 18, UBC 036; J: float specimen, UBC 043; Eolytoceras cf. tasekoi Frebold D: Level 19, UBC 037; H: Level 19, UBC 041 (nucleus only); Badouxia? sp. E: Level22, UBC 038; F: Badouxia? sp., Level2 1. UBC 039; Badouria columbiae (Frebold), Level23, UBC 040; Badouxia canadensis (Frebold), Level 18, UBC 042; Arnioceras sp., Level9, UBC 044 (inner whorls). The third ammonite-bearing intewal is found within the top 80 rn below the fault contact with Toarcian strata where sparsely fossiliferous sandstone and rninor shale interbeds yield a monogenenc Arnioceras fauna (PI. 4.2. fig. J, K).

4.3.2. Correlation and biochronologic dating

The standard chronostratigraphy of the Jurassic is based on the northwest European ammonite succession. North Arnerican Early Jurassic ammonite faunas are different enough from the northwest European ones to wamt independent regional standard zonations, as demonstrated for the Pliensbachian and Toarcian (Smith et al., 1988;

Jakobs et al., 1994). To date, no regional North Arnerican zonal scheme has been proposed for the Hettangian but local biostratigraphy has recently been elaborated for two areas with the most cornplete faunal succession: Nevada

(Guex, 1995) and the Queen Charlotte Islands (Tipper and Guex, 1994). We first compare the Alaskan faunas to these areas to which they exhibit great faunal sirnilanty. Next we consider the South Arnerican and Alpine regional zonal schemes that are usefùl for correlation on the basis of a host of cornmon taxa. Finally, we attempt a correlation with the standard chronozones (Le., northwest European zones) through a web of interregional correlation and a few direct links. Fig. 4.4 shows a global compilation of Hettangian ammonite zonations and their approximate correlation.

The stratigraphic distribution of the Alaskan Hettangian taxa known frorn other regions is sumrnarized in Fig.

4.5. Vertical ranges are shown at zonal resolution except for Nevada and the Queen Charlotte Islands. where data are available to further confine ranges to certain parts of zones. The Alpine zonation (Wahner, 1886) is based on condensed sequences that rnay not necessarily represent al1 the time of other units shown as their correlatives. It is evident that the lower fauna frorn Puale Bay correlates with Middle Hettangian units world-wide. It rnay be further constrained to its lower part (Le., Portlocki Subzone equivalents). Direct correlation with northwest Europe using

Srnoceras ex gr. portlocki and Kammerkarites ex gr. megarlama corroborates this conclusion. The only species that apparently contradicts this correlation is Mulierites pleuroacanthitoides, wh ich is only known ftorn a single bed in Nevada representing a slightly higher stratigraphic position within the Middle Hettangian. The second fauna contains East Pacific elements to such extent that there is a conspicuous lack of direct links to European faunas. Nevertheless, a placement within the Late Hettangian through Henangian-Sinemurian boundary interval is undoubted. Sunrisites, occurring in the lowest bed of this fauna, suggests the lower pari of the Upper Hettangian as the genus appean at this level in North and South Arnerica. Sornewhat higher, the association of Badolaia, Eolyraceras, and Paracaloceras is characteristic of the Canadensis Zone (Frebold, 1967; Phlfy er al..

1994) but the fint representatives of these genera may appear earlier (Tipper and Guex, 1994; Guex. 1995).

Correlation of the Canadensis Zone is dificult. Several East Pacific workers favor its placement straddling the Hettangian-Sinemurian stage boundary (Taylor, 1990; Riccardi et al.. 1991; Palfy et al., 1994). The correlation of this zone with the Alpine Marmorea Zone is more conclusive but the position of the Marmorea Zone itself is debated (Guex and Taylor, 1976; Bloos, 1983; Taylor, 1986). Recent studies by Bloos (1994; 1996) provide evidence supporting the placement of Marmorea Zone in the Hettangian. In AIaska as well as in the Queen Charlotte

Pafacaloceras Canadensis S. rnarmorea 37 44 56 P Croefzkirchneri S. montana

36 43 55 A. proaries F ranziceras -g Megastoma D. reissi Portlocki

Calliphyllum E Planorbis

NW EUROPE ALPS NEVADA QUEEN S AMERICA CHARLOTE IS. Figure 4.4. Biochronologic correlation chart of Hettangian ammonite zonations proposed for key regions. Only approlrirnate correlation is implied and the Hettangian-Sinemurian boundary is not precisely located outside northwest Europe. Zone reference numbers are the same as in Fig. 4.5. Compiled fmm the following sources: Northwest Europe - Dean et al. (1961), Alps - Wahner (1886), Nevada - Guex (1995), Queen Charlotte Islands - Tipper and Guex (1994). South America - Hillebrandt (1994). NW EUROPE ALPS NEVADA (G95) QUEEN CHARLOTTE IS. S AMERICA

Figure 4.5. Ranges of Hettangian ammonite taxa known from Puale Bay as established in other regions. Numbered biostratigraphic uni& are the same as in Fig. 4.4. Ail taxa listed in the taxonornic remarks as included in the identitied species groups are considered. Lighter shade denotes taxa of uncertain generic assignment. Solid bars indicate taxa occurring in the lower faunal level, open bars indicate taxa fiom the rniddle faunal levei at Puale Bay. Genus abbreviations: B: Badouxia; D: Discamphicerus; E: Eolytoceras; F: Franziceras; K: Kammerknrites; M: Mullerires; P: P~racaioceras;P: Pleuroncanthites; Pc: Paracaloceras; S: Smoceras. Sources: (395: Guex (1995); GF90: Guérin-Franiatte (1990); H90: Hillebrandt (1990); H94: Hillebrandt (1994); L52: Lange (1952); P94: PAlfjl et al. (1994); R91: Riccardi et al. (1991); R93: Rakus (1 993a); TG94: Tipper and Guex (1 994); W: Wühner (1 882- 1898). Islands, the füst appearance of Badolaia columbiae postdates that of most other elements of the Canadensis Zone and thus it can be pragmatically used to approximate the base of the Sinemurian.

The third fauna contains Arnioceras only, A-cf arnouldi being the only species identified. lt is a guide fossil of the Arnouldi Assemblage recognized in the Lower Sinemurian of the Queen Charlotte Islands (Palfi et al.,

1994). Although Arnioceras is known to range up to the lower part of the Upper Sinemurian. monogeneric faunas typically occur in the upper Lower Sinemurian. The global record ofthis cosrnopolitan genus also suggests that the standard Semicostatum Zone is the most likely correlative of this fauna.

4.3.3. Taxonomic remarks

As full h~onornictreatment is not intended here, taxa are listed alphabetically. The generic and even higher

Ievel classification of Hettangian ammonites is far from settled (for two different approaches. see Donovan et al.,

198 1; Guex, 1987). For practical reasons we mainly follow the scheme used by Guex (1 980; 1987: 1995) as it is convenientiy applicable to the Nevadan and Alpine faunas to which our Alaskan material shows remarkable affinity. Uncenainty in genenc assignrnent of early schlotheimiids reflects the fact that critical ventral features are rarely preserved in the Alaskan material.

Arnioceras cf. arnouldi (Dumortier)

A large specimen exceeding 25 cm in diameter was found in situ (level24) on a bedding surface in the intertidal platform where its removal was not feasible. It is identical in coiling and curved rursiradiate ribbing to -4. arnouldi known from the Queen Charlotte Islands (Pal@ et al., 1994). A srnaller, fragmentary specimen (PI. 4.2, fig. K), is also referred to Arnioceras.

Bodouxia? sp. (PI. 4.2, fig. E, F)

Several specimens from levels 19-22 are comparable to B. columbiae but differ in the occasional presence of bifurcating nbs on the inner whorls (see PI. 4.2, fig. F), a feature common in Schlotheimia or other schlotheimiids.

The venter of the slightly crushed specimens is not well preserved but it doesn't appear to show a schlotheimiid- type, well-defined sulcus or abnipt termination of ribs such as seen on a specimen of Schlotheimia sp. figured by

Imlay (198 1, pl. 2, fig. 16-17) from Puale Bay. Discamphiceras aff. missi (Tilmann) (Pl. 4.1, fig. C)

Two crushed, weakly omamented specimens fiorn levels 13-14 are more involute than D. cf. silberlingi and show weaker and Iess regular ribbing. They closely resernble D. aff. reissi from Nevada as figured by Guex ( 1995. pl. 2, fig. 13-19). The typical D. reM itself is distinguished by its regular, more prominent ribs which persist to the upper flank as shown by the holotype (Tilmann, 19 17. pl. 2 1, fig. 4) and other South American specimens

(Hillebrandt, 1994, pl. 1, fig. 12; Quinzio Sinn. 1987, pl. 1, fig. 12. pl. 2, fig. 1). We note that there are transitional forms behveen D. aff. reissi and D. silberlingi in both the Nevadan (Guex, 1995, pl. 15, fig. 13- 16) and Alaskan material.

Discamphiceras cf. silberlingi Guex (PI. 4.1. fig. A, D)

Four crushed specimens from levels 2.7, and 12 show midvolute coiling and slightly prorsiradiate ribs that are strongen on mid-flank and fade towards the venter. This form appears to be the most common species of D. in the section and similar specimens were referred to D. cf. toxophonm by Irnlay (198 1. pl. 1. fig. 3.4.8-10) who also discussed their affinities to other species of Discamphiceras from the Eastern Alps. As Guex (1 995) noted, at least some of the specimens figured by lrnlay can be identified with D. silberlingi, a species originally described from Nevada. We favor the assignment of the Alaskan form to D. silberlingi while emphasizing its similarity to the

Alpine species. The South American D. reissi (see below) is also closely related (see Hillebrandt. 1994. pl. 1. fis. 13).

Eoiyoceras cf. rasehi Frebold (PI. 4.2, fig. D, H)

This species is common in the middIe part of the section (levels 18-2 1) where it is represented by srnall. presumably juvenile specimens. Widely spaced constrictions, faint and irregular ribbing, and an evolute coiling with moderate expansion rate suggest assignment to E. tasekoi.

Euphyllifes?sp.

Several cnished specimens of srnaII size from levels 13-13 resemble EuphyUites in their evolute coiling, moderate expansion of whorls, and lack of ornamentation except for nodes on the nucleus. Positive indentification is precluded by the poor preservation and lack of larger specimens. The lest involute and most weakly ornamented

Discamphiceras species are aIso morphologically similar but differ from Euphyiiileci by their lack of nodes on the

58

Guex and Rakiis (199 1) who showed that Primapsilocerm primtrlum can be regarded as a juvenile or microconch

Kammerkarites, consistent with a reassignment of Psiloceras mberugafum,an index ammonite frorn higher levels in northeast Siberia, to Pleuroacanrhires.

Kammerkarires ex gr. megrnoma (GUmbeI) (PI. 4.1, fig. R)

A single large body chamber fragment was recovered from level3. The entire specimen would have a diameter of about 30 cm and it shows straight, slightly prorsiradiate ribs that fade above mid-flank.

Several species of Kammerkarites have similar morphology al a comparable diameter. These include K. megasfoma (e.g., Guérin-Franiatte, 1990, pl. 12, fig. 4), K. longiponrinum (Oppel) (e.g., Guérin-Franiatte. 1990. pl.

15. fig. 2, Riccardi et al., 1991, fig. 4/74), and K. artnonense (Repin) (Repin, 1988, pl. 1. fig. 7). K. huplop~chirs

(Wtihner) is aiso closely related but it differs in its ribs penisting higher on the upper flank (e-g., Wiihner, 1882. pl.

17, fig. 1-5, Guex, 1995, pl. 15, fig. 1-2).

Lytoceratid indet. (PI. -1.1, fig. K)

Only one quarter of a body chamber of a large specimen (D > 13 cm) is available from level 14. Only faint, irregular ribbing is seen together with widely spaced flares, one of which terminates in an indistinct node.

MuIIerires cf pleuroacanfhitoides Guex (PI. 4.1, fig. N, O)

This species is described in detail by Guex (1980: 1995). The Iarger of the two Alaskan specimens €rom level

1 1 clearly shows a change in ribbing style from straight and rectiradiate to aborally curved and possesses several fine, broadly parabolic striae superirnposed on ribs of the sarne trajectory. There is an apparent increase in the expansion rate of the last whorl of the Alcskan specimen, a change not seen in the Nevadan type material. lt may be the result of post-depositional distortion and differential preservation of the body chamber.

Paracaloceras cf. rursicosfarumFrebold (Pl. 4.2, fig. C, J)

A few poorly preserved Fragments from level 18 and a float specimen show evolute coiling, carinate-bisulcate venter, and coarse, dense, mrsiradiate ribbing that suggest cornparison with P. rursicostatum, a common species at nurnerous other localities thoughout the East Pacific realm. Pleuroacanihifes ex gr. mulleri Guex (Pl. 4.1, fig. L)

A single specimen of small size (D = 20 mm) fiom level9 shows highly evolute coiling, well-developed nodes on the innermost whorls and irregular ribbing interspersed with prominent parabolic nodes between shelI diameters of 10 and 20 mm, consistent with its assignrnent to Pleuroacanthites.

Similar sized specirnens of P. rnulferi Guex, known only from the middle Henangian of Nevada bear strong resemblance to the Alaskan specimen (Guex, 1995, p. 43, especially pl. 24, figs. 5-6 and 1 1-12). The Alaskan form is also close to mal1 specimens of P. biformh (Sowerby) fi-om the Megastoma Zone of Austrian Alps (Wahner.

1894, especially pl. 5, figs. 3-6; Lange, 1952, p. 93, pl. 1 1. fig. 4). and the Hettangian of West Carpathians (Rakus.

1993b. p. 13, pl. 6, fig. 7). Distinction between P. muileri and P. biformis is based on features of the adult body chamber (Guex, 1980; Guex, 1995) which is not preserved in our specimen.

Psiloceratid indet. (PI. 4.1, fig. P. Q)

One specimen fiom levei 9 appears very similar to Imlay's (198 1, pl. 1, fig. 2) Psiloceras cf. planorbis. It shows snongly nodose inner whorls, a feature cornmon to Psiloceras, Pleuroucunthite.s, and other psiloceratids

(Guex, 1995).

Srnoceras? ex gr. portlocki (Wright) (PI. 4.1. fig. I)

Three incornplete, crushed intemal molds from levels 6 and 7 differ from K. cf.frigga in their higher whorl and nearly straight, rectiradiate ribbing. Such foms from Puale Bay were previously referred to Waehneroceras porrlocki by Imlay (198 1, pl. 2, tig. 7, 10-1 1, 14-1 5). There are several closely related species known fiom the

Eastern Alps such as S. errracostatum (Wahner) (a possible synonym ofS. porrlocki, according to Donovan, 1952. and Imlay, 198 1) and S. pa~neri(Wahner).We note that the latter species (e.g.. Wahner, 1882, pl. 15, fig. 1-2).

Lange (1952, pl. 16, fig. 12-14)) is closerto our form in its high whorls, while there may be a continuum of whorl height distribution and species distinction on this basis would rernain arbin-ary. A recent treatment of the portlocki species group fiom France by Guérin-Franiatte (1990) includes " Waehneroceras" prometheus (Reynès) (lectotype refigured in Guérin-Franiatte (1990, pl. 12. fig. 2)) which our fonn also closely resembles. Assignment of these species to Saxoceras follows Guex (1995, p. 35). Srnoceras? sp. (Pl. 4. t, fig. G,J)

This form, occurring in levels 4,6. and 8, is distinguished from other representatives of Saxoceras andor

Kammerkurites by its marked coarsening of ribbing on the body chamber. Two specimens from Puale Bay figured by Imlay ( 198 1, pl. 2, fig, 12. 13) as Waehneroceras cf. portlocki closely match Our form but clearly differ From similar-sized specimens which do not show an abrupt change in rib density. A specimen from the Eastern Alps figured by Wtihner (1882, pl. 16, fig. 1) as Aegoceras n. f. indet.. cf. atracostatum shows abrupt coarsening of ribs although at somewhat Iarger diameter. Guérin-Fmniatte (1990) mentions some change in ribbing of

"Waehneroceras"port1ocki but it appears to be les pronounced than in the Alaskan form. This form rnay be a new species but betîer preserved material is needed for a definitive treatment.

Schlotheimiid indet.

One very large specimen (D 3.40 cm) from level22 that resembles bath "Charmasseiceras" and the

"marmorea" pup(sensu Bloos. 1988) in its inner whorls. nie last whorl becomes almost completely smooth.

Sunrisites? sp. (PI. 4.1, fig. A, B)

A single specimen from IeveI 16 shows evolute coiling and coarse, blunt, rectiradiate ribbing. It is morphoIogically similar to Franziceras? sp. discussed above, although they are clearly separated stratigraphically and occur within different ammonite assemblages. The lack of tuberculation of the inner whorls serves as a distinguishing feature. The ribs are Iess flexuous and more widely spaced than in S. sumisense Guex or S. hadroptychrm (Wahner). Reasonable comparison can be made with S. peruvianum (Lange) that was originally assigned to Caloceras but is now placed within Sunrisites based on its stratigraphic position within the upper

Hettangian (Hillebrandt, 1994). Two specimens figured as Psiloceras (Franziceras) sp. €rom northern Alaska by

Imlay (198 1, pl. 1, fig. 1 1. 15- 16) were assigned to Sunrisites by Guex (1 995) and are close to the Puale Bay specimen differïng only by their rectiradiate ribbing trend. Comparable coarse ribbing also occurs in Nevadan specimens referred to Alsatires nigrornontanus (GClmbel) (Guex, 1995). The latter species is said to show transitional forms towards Sunrisires (W&ner, 1886, Guex, 1995). 4.4.1. Analytical methods

U-Pb analytical work was done at the Geochronology Laboratoty of the University of British Columbia.

Zircon was obtained hm8 to 25 kg samples using cnishing, grinding, wet shaking, and heavy Iiquid separation.

Fractions were hand-picked based on differences in magnetic susceptibility, size, and crystal rnorphology.

Whenever possible, the least magnetic and best quality grains (i.e., free of cracks, inclusions, and visible cores or

zoning) were used for the analyses. tn al1 but one sample with the srnallest zircon yieId (95JPJ), the fiactions were

strongly air-abraded using the technique of Krogh (1982) to minimire or eliminate the effect of surface-correlated

Pb-loss. Details of techniques employed in chemistry and mass spectrometry are given by Mortensen et al. ( 1995).

During the coune of this study, U and Pb procedural blanks were 0.6-2.5 and 5-9 pg, respectively. Blank isotopic

compositions, crucial in calculation of "'pb/'j5~ and '07~bl'"~bages, were carefully monitored.

Anal9ical data are reported in Table 4.1. The age calculations are based on the decay constants

recommended by Steiger and Jtiger (1977); the correction for initial common Pb folIows the mode1 of Stacey and

Kramers (1975). The analytical errors are propagated through the age calculations using the method of Roddick

(1987). In some analyzed zircon fractions, U-Pb systematics are affected by Pb-loss and inheritance. The rationale

for assignment of the crysrallization age and its associated error at the 2 sigma Ievel is described separately for each sample.

4.4.2. Age determinations

The lowermost sample (95JPI) was collected fiom a conspicuous, 1 cm thick, pink, coarse-gained crystal niff layer interbedded with the siliceous marl sequence (Fig. 4.6). A small, less than 1 O kg sample yielded abundant and excellent quality zircons. The apparently homogenous population consisted of pale brown, mainly euhedral. multifaceted, doubly temninated prisms with varying aspect ratio, ranging from stubby to needle-like grains.

Seven of the analyzed nine fractions are concordant to nearly concordant (Fig. 4.7A). The 'Obpb/% ages range from 194.7 to 198.3 Ma and many of them overlap. Slight Pb loss not entirely removed by air-abrasion is indicated for these fractions, perhaps with the exception of fraction A that is concordant and yields the oldest

'"Pb?% age (t98.3I0.7 Ma). Notably, fractions E and F, abraded to a lesser degree, yielded the youngest

-336 P~/~~uages. Since the concordance and age of Fraction A has not been duplicated, we calculated a seven- fraction discordia line through the ongin (MSWD.1.98) that yields an upper intercept age of 2OO.4*2.7 Ma and

only requires a 0.1 Ma extension to include the age of A and its range of error. Thus a conservative ben

estimate for the age of the rock is 200.4+2.7/-2.8 Ma. Fraction J was not considered as it is revenely discordant and

it clearly represents an inferior analysis. Fraction D diffen from al1 other fractions by its older 'M~b/?J age that is

interpreted as a result of small amount of inherited older Pb cornponent, likely as undetected, cryptic cores.

Unequivocal evidence for inherited or xenocrystic zircon will be presented below for samples stratigraphically

higher in the section. On this basis, fraction D was also omitted From the final age calculation. U hpic/ traction1 Wt Pb.2 =PD- E'b' ?IfPb) Lrotopic ratios (rl a.?/6) Calculatcd ages (SuMa) mg ppm ppm mPb pg % rtnpa21-ru 3>7&3nPb 95JPI ' A [>134. NI. etprs. s-al 8-4 0.03I244.17% B [>134. NI. ecs+t. s-al 8.9 0.03 f 05 IO. IO% D [SI 34. pce, s-a] 8.9 0.03206 MX% E [(14. N 1. n+c. m-a] 10.1 0.03089 t0.10% F [74- 134. N1. WC. m-a] 9.4 0.03066 IO. 1 1% G [74-134, NI, p. s-a] 9.1 0.03097 &16% H [74-134. N2, pe. s-al 95 0.03104 d.lI% 1 [> 134, M, @p. s-a] 8.7 0.03091 4.10% i 174-134, NZ, pcs-a] 8.8 0.03044 4.26%

95JP3 ' A [>IO4 NS, prq. s-al B [74-104. NS. pr,s-al C [(14, N5. ah+sh s-a] D [> 104, N5, ps-a] E [404, N5. prshqa]

95JP4 A [W-l34. N2. q, n-a1 B [q1.N2. q. sh. n-al 95P5 '" A [>13J. NI. clip, s-al B [>lX. NI. p. s-al C [>134, NI. syp. s-al F[>IW.Nt.p+eLs-a] G [104-134. N1. p. s-al H[>I34. NI. ptels-a] i [104-134. NI.p+s, s-a]

[AII zircon fi-actions. Listed in brackets: Grain sire range in microns; Side slope of Franr rnagnrtic separator (in degrea) at which grains are non-magnetic (N) or magnetic (M), using 20" front slope and 1.8A field strength: Grain character: b=broken pieces, edongate, eq=equant, n=needles, p-prismatic, s-stubby, t=tabular ah=anhedral, sh=subhedral; Degrer of air abrasion: n-a-on-abraded, 1-a=lightly abraded, m-a=moderateIy abraded, s-a=strongIy abraded. 2~adiogaiicPb 3~casuredratio corrected for spikc and Pb fractionation of O.OO43/amu12O% (DaIy collector) and 0.00 12mu + 7% (Faraday collector). '%otal comrnon Pb in anaiysis based on blank isotopiç composition S~adiogenicPb 6~orrectedfor blank Pb. U, and initial cornmon Pb based on the Stacey and Kramers (1975) model at the age of the rock or the 20TbPPbage of the Fraction. '~ocation: SE shore of Pulae Bay, Map Karluk C-415, UTM 6400950N. 3576308 *~ocation:SE shore of Pulae Bay, Map Karluk C-415. üTM 6400950N. 3576308 9~ocation:SE shore of Pulae Bay, Map Karluk C-415, üTM 6400950N, 3576308 10~ocation:SE shore of Pulac Bay. Map Karluk C-415, UTM 6400950N. 3576308

Table 4.1. U-Pb zircon analytical data Figure 4.6. One cm thick crystal tuff layer (SampIe 95JP1) interbedded with the siliceous marl near the top of the Kamishak Formation. Shotgun for scale (120 cm in length) at left center.

Three samples were collected fiom the main voIcaniclastic unit. The lowermost of these, 95JP5, is from a nearIy 6 m thick bed ofgreen to red, unaltered, crystal-rkh tuff. Common phenocrysts include plagioclase, biotite. and hornblende. Abundant gem-quality zircon was recovered from this sample. Colorless to pale brown, euhedral. doubly-tenninating prismatic grains form a single population ranging continuously from needles with simple. square cross-section to more multifaceted, stubby crystals. Six of the analysed seven fractions intercept or touch the concordia line and overlap one another (Fig. 4.7I3).

Fraction H that yielded the youngest '"pb/UgIJ age is slightly discordant, likely due to Pb loss not completely removed by abrasion. The other six fiactions yield a weighted mean '06~b/738~age of 197.8*0.4 Ma. Allowing for the possibility of slight Pb Ioss in some of these fiaction, we extend the error to include the oldest '06~b/'js~age of the concordant fraction B (198.610-4). Thus the best age estimate of this sample is 197.8+1.2/-0.4 Ma.

SampIe 95JP4 was collected fiom a green, resistant, !O cm thick. likely somewhat reworked tuff layer that contains rare plagioclase and biotite phenocrysts, devitrified glas and lithic fragments. A 20 kg sarnple yielded les than 0.1 mg zircon that was divided into two fractions. Both contained subhedral grains of good quality with no visible cores or zoning. Abrasion was prohibited by the small quantity of grains. Both fractions show a strong inherited Pb component (Fig. 4.7C). A discordia line yields a Iower intercept of 1 13h29 Ma. The deviation from the 9.- 95JPl (Middle Hettangian)

3 omis f a na

0-

lnterpreted age: 200.4 +2.7/-2.8 Ma

ODZDJ

95JP3 and 95JP4 combineci (Middle ta Lptc Hallripin)

O 15 OU 0 55

@ 1 95JPS (Middle to Late iieüangian) ~4

I lnterpreted age: 197.8 +t .2i-û.4 Ma

Figure 4.7. U-Pb concordia diagrams for samples from the Hettangian section at Puale Bay. A: Sample 95JP1, Middle Hettangian, crystal tuff from near the top of Kamishak Formation (?); B: Sample 95JP5, Middle to Late Hettangian, tuff from the Talkeetna Formation; C: Samples 95JP3 and 951P4, tuff fiom the Talkeetna Formation. expected crystailization age could result fiom Pb Ioss andlor mixing of inherited Pb of different aga. ResuIts fiom

Our other samples suggest that Pb loss is expected for unabraded zircons. The upper intercept of 1270I68 Ma points to a Proterozoic zircon component that was not detected as cryptic cores or xenocrysts.

nie stratigraphically highest sample, 95JP3, was collected From massive, sofl, poorly lithified, plagioclase-

phyric water-laid tuff with green, somewhat altered groundmass. Zircon in this sample was not abundant but the

recovered material was sufficient to analyze five fractions. Colorless to pale brown grains of good to excellent

clarity were separated into fractions of euhedral or dominantly subhedral to anhedral crystals. Fractions A and D.

both containing euhedral prisms, are overlapping and slightly discordant to concordant whereas the other three

fractions reveal variable amount of inherited Pb. It is possible that some resorbed grains are xenocrysts or cryptic

zircon cores remained visually undetected. When plotted on a composite concordia diagram (Fig, 4.7C), these three

fractions and the two discordant fractions of sample 95JP4 fail to defiiie a single chord. It is therefore likely that

rnixing of inherited Pb of varying ages occured in both samples. A bounding chord through fractions A, D, and C

defines an older upper intercept age of XWk3O Ma The other bounding chord through A, D, and E yields an

upper intercept age of 1094k30 Ma. The crystallization age of the mff is estimated from the m~b/'js~age of

Fractions A and D that are vimally fiee of inheritance and Pb loss. An estimate of 197.811.0 Ma is derived From

the weighted mean '06~b?% age with its error extended to encompass the error range of both Fractions.

4.5. DISCUSSCON

In this study we identified 20 ammonite taxa From new collections as compared to the nine taxa reported by

Irnlay (1 98 1) whose work is based on all previous collections from the same area. Table 4.2 summarizes the revised taxonomy in cornparison with that of Irnlay (198 1). Only two previously known species have not been found in the new collection. Loqueoceras cf. sublaqueiis is a serpenticone alsatitid characteristic to the upper part of the Liasicus Zone or its equivaIents. Schloiheimia sp., with the genus ranging from the Middle through the Late

Hettangian, can also be easily accomrnodated within the stratigraphie Framework outlined above.

The biochronologically most important revision concems PsiIoceras cf. planorbis in ImIay (198 1). Similar. midvolute unormanented forms occumng in the lower fauna are now tentatively interpteted as EuphyIIi~es?sp.

Also similar is a large, smooth whorl fragment at the top ofthe lower fauna that is probably a lytoceratid. Both of these identifications are in accordance with the Middle Hettangian age assignment. There appears to be no fim evidence for the presence of Early Hettangian faunas. The youngest proven Triassic is the Norian Cordilleranus

67 Zone and the oldest proven Jurassic is equivalent to the Liasicus Zone. Between the two at Puale Bay, there are several tens of rnetres of mata containing only trace fossils (Newton, 1989) and a fault. The presence of Middle

Hettangian ammonites in the lowest beds of the hanging wall and the lack of significant lithological change across the critical interval suggest that deposition may have been continuous throughout the Triassic-Jurassic boundary but the basal lurassic is faulted out. The Puale Bay section is therefore niled out as a continuous Triassic-lurassic boundary section. However, outcrops at neighbouring Alinchak Bay should be scrutinized as they may contain an uninterrupted record spanning the boundary.

THIS REPORT IMLAY 1981

~oiytocerasîf.tasekoi Pleuroacanthites ex gr. mulleri - lytoceratid indet ? Psiloceras cf. P. planorbis Discamphiceras cf. silberlingi D. cf. D. toxophorum Discamphiceras aff. reissi - Euphyllites? sp. ? Psiloceras cf. P. planorbis psiloceratid indet - Kammerkantes? cf. fngga Waehneroceras cf. W. tenerum Kammerkantes ex gr. megastoma - Saxoceras? ex gr. portlocki Waehneroceras cf. W. portlocki Saxoceras? sp. Waehnemceras cf. W. portlocki Franziceras? sp. - Schlotheimia sp. schlotheimiid indet.. - Mullerites cf. pleuroacanthitoides - Laqueoceras cf. L. subiaqueus Sunrisites? sp. - Badouxia canadensis Badouxia canadensis B. columbiae - Badouxia? sp. - Paracaloceras cf. rursicostatum Paracaloceras cf, rursimstatum Amioceras arnouldi Amioceras cf. A. densicosta - --

Table 4.2. Cornparison of ammonite identifications in ImIay (1981) and in this report. The detailed documentation of ammonoid ranges in the Puale Bay section is an important step toward the development a regional standard zonation for the North American Hettangian. The main stratigraphie relationships known from Nevada and the Queen Charlotte Islands were upheid. Our new observations underscore the need for the study of several sections before the biochronologic significance of certain taxa can be fully understood. The lower fauna is not easiiy amenable to subdivision yet it contains elernents of four Iocal zones fiom Nevada (Guex.

1995) and two fiom the Queen Charlotte islands (Tipper and Guex, 1994). The occurrence of Discamphihceras cf. silberlingi below D. aff. reissi is the reverse of the succession in Nevada. There is a more substantial overIap in the ranges of Kammerkarites, Saroceras and Franziceras than found elsewhere. Euphyiiites, if correctly identified, occurs above MulIerires. The succession of Sunrisites? and Badolrxia in the second fauna supports the idea that this lineage is the mon usefil for subdividing the Late Hettangian and the Hettangian-Sinemurian transition in North

Arnerica.

The isotopically dated levels are firmly constrained by the ammonite biochronology. SampIe 95JP1 cornes from within the lower fauna, ix., it belongs in the Middle Hettangian (Liasicus Zone equivalent). This thin tuff layer predates the onset of a major volcanic phase that is dated by the other three samples (JP955.lP954. and

JP953). They are not older than Middle Hettangian (Liasicus Zone equivalent) and not younger than Late

Hettangian (Angulata Zone equivalent). The isotopic ages of the Middle Hettangian and the Middle or Upper

Hettangian samples overlap within error therefore the duration of these Hettangian ammonite zones cannot be directly estirnated but were likely short.

The new U-Pb ages do not correspond to the Hettangian in most of the recently publisbed tirne scales, although they fa11 within the range of error proposed for the terminal Hettangian boundary (Fig. 4.8). Based on the new results, the Hettangian-Sinemurian boundary cannot be older than 199 Ma. The age of the Triassic-Jurassic boundary cannot be directly derived from our data. Many tirne scales (e.g., Harland et al.. 1990. Gradstein et al..

1994) use interpolation based on the equal duration of biochronologic units. This method is ill-suited for the period of biotic recovery fo[Iowing the end-Triassic mass extinction events. Disparate zonal durations in the Hettangian.

Le., the brevity of the Planorbis Zone compared to the Liasicus Zone, were demonstrated using Milankovitch cyclicity (Smith, 1990). Accepting that the Pfanorbis Zone is not longer than average, the Triassic-lurassic boundary is likely to fall in the 200-205 Ma interval. This estimate is younger than those in most curent time scales with the exception of Odin's (1 994) (Fig. 4.8). It is also compatible with the 20 1-702 Ma boundary age suggested fiom the Newark Basin (Olsen et al., 1996b) through the integration of cyclostratigraphy (Olsen et al.. Middle to Late

NEW HETTANGIAN U-Pb DATES

Figure 4.8. Comparison of Hettangian stage boundary age estimates in recent time scales and the new Hettangian U-Pb dates reported here. Sources of time scales quoted: DNAG (Palmer, 1983; Kent and Gradstein, 1985; Kent and Gradstein. 1986); EXX88 (Haq et al., 1987; Haq et al., 1988); GTS89 (Harland et al., 1990); Odin (Odin, 1994); MTS (Gradstein et al., 1994: Gradstein et al., 1995).

1996a), U-Pb dating (Dunning and Hodych, 1990; Hodych and Dunning, 1992) and palynosnatigraphy (Fowell et al., 1994).

The presence of inherited Proterozoic zircon in the Talkeetna Formation is documented here for the first time.

Five fractions in two of the dated samples (95JP3 and 95JP4) contained a strong inherited zircon component. The lack of CO-linearityin the discordant Fractions is probably atrributed to the mixing of inherited zircons of different ages. The two bounding chords (Fig. 4.7C) suggest a range fkom at least late Middle Proterozoic (1094iJO Ma) ta

Late Archean (2775*30 Ma) upper intercept ages. The admixing of such oId zircon requires the proximity of evolved Precambrian crut during the magmatic processes, a scenario that is not compatible with some curent tectonic models. On the basis of elemental abundances, Barker et al. (1994) found the Talkeema volcanics transitional between tholeiitic and calc-alkaline type and proposed an intra-oceanic volcanic arc as their likely tectonic setting. In another geochemical study, DeBari and Sleep (1991) documented high-Mg and low-Al "bulk arc" composition for the Talkeetna arc and suggested that the magma was sourced fiom a mantle wedge rather than a subducting plate. A new tectonic model for the Taikeetna arc should account for the observed zircon inheritance pattern while remaining compatible with the geochemical afinities. 4.6. CONCLUSIONS

The lowest Jurassic ammonite volcano-sedimentary sequence of the Puale Bay section was successfully dated

using ammonite biochronology and U-Pb geochronomeûy. The oldest Jurassic ammonites in the Puale Bay section

are of Middle Hettangian age (Liasicus Zone, probably Portlocki Subzone equivalent). No record of basal

Hettangian was found and the Triassic-lurassic transition is probabty missing localty due to a srnaII fault. The

andesitic volcanism recorded in the Talkeetna Formation staned during or immediately after the Middle Hettangian

and was terminated, at lem locally, before the end of Hettangian. Hettangian and Early Sinemurian ammonite

faunas are cIoseIy comparable with those of the other major North Amencan localities in Nevada and the Queen

Charlotte Islands pemitting global correlation at approximately the zonal Ievel.

Three new U-Pb zircon dates from the stratigraphically tightly constrained Hettangian volcanic units fumish

calibration points for the Jurassic time scale. A thin tuff layer from within the correlatives ofthe Middle Hettangian

Liasicus Zone is dated at 200.8+2.7/-2.8 Ma. Tuffs of the Talkeetna Formation bracketed by Middle and Late

Hettangian ammonites yield crystallization ages of 197.8+ 1 .U-0.4 Ma and 197.81: 1 .O Ma. Zircon inheritance panems reveal the proximity of the Talkeetna arc to evolved, Proterozoic to Late Archean crusta1 components that is not compatible with current tectonic models.

The new dates suggest that the Hettangian-Sinemurian boundary is younger than 199 Ma and its age is overestimated in nearly aIL currently used time scales. The age of the Triassic-Jurassic boundary is likely between

200 and 205 Ma, younger than most current estimates. CHAPTER 5

NEW U-Pb ZIRCON AGES INTEGRATED WITH

AMMONITE BIOCHRONOLOGY FROM THE JURASSIC

OF THE CANADIAN CORDILLERA

5.1. INTRODUCTION

More than a dozen time scales published in the last 15 years provide different age estimates for Jurassic stage boundaries. Even the most widely used time scales (Palmer, 1983; Haq et al., 1988; Harland et al., 1990: Gradstein et al., 1994), however, commonly disagree considerably and contain large uncertainties. The flaws are atvibuted to the relatively smalI number and uneven temporal distribution of isotopic ages used, the predominance of K-Ar ages, the fiequent use of minerals with low isotopic closure temperature in dating, and the large proportion of ages obtained from stratigraphically poorly-constrained plutonic rocks (PAlQ, 1995).

The Cordilleran Jurassic Calibratîon Project was initiated to improve the time scale through generation of new calibration points (Pal@ et al., 1995). lntegrated ammonite biochronology and U-Pb geochronology was adopted as a combination of one of the best fossil and isotopic dating techniques presently available that is well-suited to the abundant volcanosedimentary arc sequences in the Canadian Cordillera. The usefiilness of this approach is demonstrated by PdQ et al. (1997). Results of a study using a similar approach in Alaska are presented in Chapter

4. Here we report 13 new U-Pb ages obtained from volcaniclastic units that are well constrained by ammonite biochronology. This concened dating effort is unprecedented for the Jurassic in that it attempts to systematicaIly generate time scale calibration points for much of the Early and Middle Jurassic and uses zona1 resolution in biochronologic dating. 5.2. REGIONAL GEOLOGICAL SETTING

Target sections for field work and sample collecting were identified in Stikinia and Wrangellia, two of the largest volcanic arc-type Cordilleran tectonostratigraphic ternes (Fig. 5. I). A voluminous Lower to Middle

Jurassic calc-alkaline volcanic suite and associated marine sedimentary mcks make up the Hazelton Group (Tipper and Richards, 1976), a key stratigraphie unit of Stikinia. Local Iithostratigraphy mies along the length and width of the terrane as summarized by Marsden and Thorkelson (1992) who interpret the wide volcanic belt as the result of paired subduction-related arc volcanism and sedimentation in a common back-arc basin. This basin. the

Hazelton Trough, contains an intermittent and scattered ammonite record that nevertheless represents most zones of the Late Sinemurian through Late Bajocian interval. The waning of widespread Hazelton volcanism corresponded to the inception of the Bowser basin, a successor basin of the Hazelton Trough. This depocenter accumuIated thick

Middle Jurassic to mid-Cretaceous molasse shed fiom the emerging CordiIIera (Ricketts et al., 1992). The Ashman

Formation, the lowermost unit of Bowser Group, still contains a minor volcaniclastic component. The cessation of volcanism in the Late Jurassic resmcts our study to EarIy and MiddIe Jurassic time.

C Wrangellia L Stikinia L Cache Creek

ZOO km N

Figure 5.1. Index map of U-Pb sample localities with relation to the major accreted ternes of the Canadian Cordillera. A - Ashman Ridge; B - Mount Brock range; C - Zymoetz (Copper) River; D - Diagonal Mountain; M - McDonell Lake; R - Rupert Inlet; Te - Telkwa Range; To - Todagin Mountain; Tr - Troitsa Ridge. In Wrangellia, Jurassic volcanic arcs are recorded in the Lower to Middle Jurassic Bonanza Group volcanic rocks on Vancouver island (Jeletzky, 1976) and Middle Jurassic Yakoun Group of the Queen Charlotte Islands

(Cameron and Tipper, 1985). The only Wrangellian sample reported in this paper is from the Harbledown

Formation, a thin-bedded Sinemurian shale unit that contains thin niff layers and is interpreted a distal marine facies equivalent of parts of the Bonanza Group volcanic rocks (Muller et al., 1974).

53. DATING METHODS

5.3.1. Ammonite biochronology

Ammonite biochronology is preferred as the method to date sedimentary rocks intercalated with the isotopically-dated volcanic rocks because: (1) ammonites, along with biostratigraphically less usefid bivalves, are the most cornmon macrofossils in Cordilleran Jurassic strata; (2) standard stratigraphic units in the Jurassic are based on ammonite biochronology: (3) ammonite faunal successions of the Cordillera have been intensively studied and regionally applicable standard or preliminary zonal schemes are now available for most parts of the

Jurassic.

The Sinemurian through Bajocian North American ammonite biochronologic scherne used in this study (Fig.

5.2) comprises formal regional standard zones and informal assemblages based on local studies or less well- understood faunas awaiting Further studies. Sources of compilation include Pzilfy et al. (1994) for the Sinemurian.

Smith et al. (1988) for the Pliensbachian, Jakobs et al. (1994) for the Toarcian, Poulton and Tipper (199 1) for the

AaIenian, Hall and Westermann (1980) and Hillebrandt et al. (1992) for the Bajocian. The Bathonian and Callovian ammonite successions are less well understood and no regionally applicable and widely accepted zonal scheme is available. An excellent sumrnary of faunal associations of this age is given by Callomon (1984) and recent updates are found in Arthur et al. (1993) and Poulton et al. (1994). The following broader units are distinguished: the Early

Bathonian based on Inkkinites, Parareineckeia and perisphinctids; the MiddleiLate Bathonian based on early representatives of Kepplerites along with [niskinites, the first Liffoettia. and perisphinctids, the Early Callovian with Cadocerus comma,coarsely ribbed Kepplerites spp., and Liffoettiaspp. The Middle Callovian with its

Cordilleran index ammonite Cadoceras stenoloboide is beyond the stratigraphic range covered in this report NORTH AMERICAN AWMONiïE ZONES OR ASSEMBLAGES Epidgzagiœras

Rotundum Zone

Kirschneri Zone rassic cos ta tus zone 1 Widebayense Zone 1

1 WSSU~Zone 1

Yakounensis Zone I Hillebrandti Zone Cassicosta Zone PlanuiataZone E Kanense Zone Carloîtense Zone L I - KunaeZone Freboldi Zone E WhiteavesiZone lmiayi Zone - Tetraspidoceras Assemblage PleSBdlioœras? L harbledownense Assemblage

. --- ~- - 1 Assemblage 1 Amioc8ms amouldi

Canadensis Zone

Figure 5.2. North American composite ammonite zonal scherne for the Sinemurian through Bajocian. See text for explmation and sources of compilation.

Correlation with the northwest European, primary standard zonal scheme is discussed in the sources quoted and not dealt with here, except where warranted by new data. Zona1 assignments are based on the published ranges of ammonites found in the studied sections. Where fossils occur only either below or above a U-Pb dated volcanic unit, the temporal proximity of fossils to the dated stratum is inferred from the stratigraphic relationships. U-Pb zircon dating was carried out in the Geochronology Laboratory of the Department of Earth and Ocean

Sciences, University of British Columbia (al1 sarnples except for those listed below) and in the Geochronology

Laboratory of the Geological Survey of Canada (samples 2 and 5).

Zircon was obtained fiom 15 to 40 kg samples using conventional crushing, grinding, wet shaking. and heavy

liquid separation. Fractions were hand-picked based on differences in magnetic susceptibility, size, and crystal

morphology. Whenever possible, the least magnetic and best quality grains (Le., free of cracks, inclusions, and

visible cores or zoning) were used for the analyses. Most hctions were air-abraded using the technique of Krogh

(1982) to minimize or elirninate the effect of surface-correlated Pb loss. Non-abraded hctions were sometimes

used as a necessity dictated by the small amount of zircon available (e-g.. samples land 9)or to assess the amount

Pb loss and to help constrain the Pb loss chord (samples 7 and 1 1). Details of techniques employed in chemistry and mas spectrometry at UBC are given by Mortensen et al. (1995) and those used at the GSC are discussed by

Parrish et al. (1 987). During the course of this study, U and Pb procedural blanks were 0.6-2.5 and 5-9 pg, respectively. Blank isotopic compositions, crucial in calculation of '"~bl?J and 'O'P~/'~P~ages, were careh lly monitored. The age calculations are based on the decay constants recommended by Steiger and Jager (1977); the correction for initial common Pb follows the mode1 of Stacey and Kramers (1975). The analytical errors are propagated through the age calculations using the method of Roddick (1987). In the analyzed samples, zircon U-

Pb systematics are commonly affected by Pb loss andlor inheritance. The rationale for assignment of the crystallization age and its associated error at the 2 sigma IeveI is desctibed individually in each case.

5.4. INTEGRATED BIO- AND GEOCHRONOLOGICAL RESULTS

In the following section we present the results in order of oldest to youngest stages. Biochronology, geochronoIogy, and location of samples are surnmarized in Table 5.1. For each sample, a brief description of the local geology and the studied section, the ammonite biochronology, the U-Pb systematics, and a discussion of the age assignment are given. The U-Pb analytical data are Iisted in Table 5.2. In many cases ammonite data fiom the sections under discussion have already been published. However, it is necessary to illustrate newly discovered

Bathonian faunas. Sample 2. Tuflayer in Harbledown Formation, near Rupert Met, northern Vancouver Island

A predominantly dark shale sequence is exposed in a quarry east of the head of Rupert inlet on northern

Vancouver Island (see Table 5.1 for locality information) and was assigned to the Harbledown Formation and dated as Sinemurian based on the presence of Arniocem (Haggart and Tipper, 1994). Stratignphic retationships with the Triassic (and possibly basal Jurassic) Parson Bay Formation and Lower to Middle Jurassic Bonanza Group are discussed in Nixon et al. (1994). Sample 1 was collected from one of several, few centimetre thick, light grey. reworked air-fa11 mff layers interbedded with thin-bedded, dark shale and siltstone. Our new ammonite collections yielded Arniocerrl~cf. arnouldi 2.5 m below the dated tuff layer and Arnioceras cf. oppeli another 5 m fanher downsection. Both species range from the late Early Sinemurian Arnouldi Assemblage to the early Late Sinemurian

Varians Asemblage (Pale et al., 1994). In the Queen Charlotte Islands where the ammonite succession is better known, the absence of associated other ammonites (e.g., asteroceratids)characterizes the Amouldi Assemblage.

Although no ammonites were recovered from above the tuff layer, no Iithologic break or other indication of a gap was obçerved between the tuff and the fossiliferous beds. Ammonite zones in the Harbledown Formation in its type area on Harbledown Island (Crickmay, 1928), as well as in the temporaily and lithologically equivalent SandiIands

Formation on the Queen Charlotte Islands (Cameron and Tipper, I985), are typicalIy represented by several tens of metres of strata. The isotopically dated tuff is therefore assigned to the Amouldi or possibly to the Varians

Assemblage.

Approximately 45 kg of processed rock yielded less than 0.5 mg of zircon. An apparently homogenous population consisted of mostly subhedral, prismatic grains of excellent clarity. One of the two analyzed fractions. B is slightly reversely discordant (Fig. 5.3A), likely due to analytical problems in measuting '07pb in the small amount of available material and the sensitivity of resulrs to blank Pb composition (see Table 5.2 for analytical data). Significant Pb loss is evident in the non-abraded Fraction A. It is likely that the moderate abrasion of fraction

B did not remove al1 the surface-correlated Pb loss, therefore we interpret its '06~b/"8~age of 19 I.3a.4 Ma as the minimum age of the tuff. Sample Field No. 150 000 UTM UTM Locality Lower age bracketl Upper age bracket' lnterprebd &Pb age No, NTS easting northing E of Rupert lnlet Arnouldi Ass. (E. Sinemurian) Varians Ass. (L. Sinemurian) 191 Ma minlmum

Ashman Ridge (not directly known) Varians Ass. (L. Sinemurlan) 189 Ma mlnimum

Ashman Ridge (no! diredly known) Varlans Ass. (L. Sinemurian) 192 Ma mlnlmum

S Telkwa Range Harbledownense Ass. (L. Slnemurian) Harbledownense Ass. (L. Slnemurian) 194.0+9,11-1.8 Ma

S Telkwa Range Harbledownense Ass. (L. Sinemurian) Whiteavesi 2. (E. Pliensbachlan) 191.5î0.8 Ma

Todagin Mtn. Freboldi Z. (E. Pliensbachian) Freboldi 2. (E. Pliensbachlan) 185.6+6.11-0.6 Ma

W of Mt. Brock Kanense 2. (E. Toarcian) Planulata 2. (M. Toarclan) 180.4+6.0/4.4 Ma

Diagonal Mtn. Yakounensis Z. (L. Toarclan) Early Bajocian 179.8I6.3 Ma

W of Trotsa Peak Howelli 2. (L. Aalenian) Howelli 2. (L. Aalenian) 175+161-12 Ma

Diagonal Mtn. Rotundum 2. (L. Bajocian) Late Bajocian 167.2+10.5/-0.4 Ma

McDonell Lake Bathonian Late Bathonian 158.2+1.9/-0.4 Ma

Diagonal Mtn. Bathonian L. BathonianlE. Callovian 158 Ma minlmum

Zymoetz Rlver Late Bathonian L. BathonianlE. Callovlan 162.6+2.9/-7.0 Ma

Z.-Zone, Ass. -Assemblage, E. - Early, M. - Middle, L. - Late

Tablc 5.1. Suiiiniary of biochronology, geoclironology. and location of saniples. @ SlNEYURlAN rVarians Assemblage 1,6 - Sample 2

ID r: 0.028 - B n m 5?

A Minimum age 191 Ma - tly Minimum age 189 Ma 0.024 0.165 0.180 O 185 O 210

SINEMURIAN SINEMURIAN Harbledownense Assemblage Sarnple 4

180

Minimum age 189 Ma

sSlNEMURlAN PLIENIBACHIAN sHarbledownense @ 1 lgY" Assemblage - Sample 5 194.

Yakounensls Zone to Early BAJOCIAN Sarnple 8

Figure 5.3. Continued on following page, 79 BAJOCIAN Howelll Zone Rotundurn Zone Sample 9 Sample IO

,'- Middle to Late BATHONIAN

Late BATHONIAN to Early CALLOVIAN Sample 13

Figure 5.3 (this and preceding page). U-Pb concordia diagrams of samples 1-13. (B) Troitsa ridge

0~r3 @TL! @ Trl

(CI McDonell Lake

Figure 5.4. Continued on following page. 8 t LEGEND

limestone

shale

siltstone

sandstone sandstone coquina

conglomerate

tuffaceous (D) West of Mount Brock sandstone volcanic tuff (non-welded)

welded tuff

densely welded tuff

volcanic flow

@ Dacîylioceras spp volcanic breccia

dyke

Figure 5.4. Measured or schematic smtigraphic sections showing U-Pb sarnples and fossil collections. A - Ashrnan Ridge; B - Troitsa ridge; C - McDonelI Lake; D - West of Mount Brock; E - Telkwa Range; F - Zyrnoetz (Coppet-) River 82

'AII zircon fractions. Listed in brackets: Grain size range in microns: Side slope of Franz magnetic separator (in degrees) at . which grains are non-magnetic 0 or magnetic (M), using 20° front dope and 1.8A field strength: Grain character: b=broken pieces, =Iongate, eq=equmt. meedles, p=prismatic. ~stubby.mbular, Degree of air abrasion: n-a=non- abraded, a=ab raded, Ga=Iigh tIy abraded. m-a=moderateIy abraded, s-a=strong I y abraded. *Etadiogenic Pb 3~eaairedratio correctcd for spike and Pb fractionation oFO.O043/amu+ 20% (Daly collector) and 0.00 1 Yamu k 7% (Faraday collector). 4~otalcommon Pb in analysis based on blank isotopic composition

6~orrectedfor blank Pb. U. and initial cornmon Pb based on the Stacey and Kramcrs ( 1975) mode1 at the agr of the rock or the =07PbPPbage of the tiaction.

Table 5.2. (Continued from preceding page) U-Pb analytical data.

Sample 2. Rhyolite tuf/ in Telkwa Formation, Ashman Ridge, Srnithers map area

The section on Ashman Ridge is one of the mon complete and best known Jurassic successions within

Stikinia in north-central British Columbia (Tipper and Richards, 1976). Its lower part comprises a thick pile of volcanic and volcaniclastic strata of the Telkwa Formation and it contains the oldest known fossiIiferous Jurassic sedimentary rocks in the area (Fig. 5.4A). A detailed description is given in Pal@ and Schmidt (1994). A densely welded, rhyolite to dacite crystal-lithic tuff unit near the base of the exposed section was sarnpled for U-Pb dating

(Sarnple 2). It lies more than 200 rn below a locally very fossiliferous, calcareous tutfaceous sandstone (Level.4 1 on Fig. 5.4A) which in tum overlies 2 thick-bedded, micritic limestone both of which appears ro record the only short-lived episode of marine sedirnentation in the Telkwa Formation in the Ashman Ridge area. The ammonite fauna is composed exclusively of asteroceratids (Asreroceras cf.. A. aff. margarira, and Aegasterocerar? aff. jeleekyi, al1 figured in PAIS and Schmidt (1994)), indicating the early Late Sinemurian Varians Assemblage. The

Telkwa Formation appears to represent geologically rapid accumulation of predominantly subaerially deposited volcanic and volcaniclastic rocks (Tipper and Richards, 1976), hence the age of the isotopically dated tuff unit is considered to be early Late Sinemurian or slightly older.

No precise age assignment is possible because al1 three zircon fractions analyzed are discordant to varying degrees and do not form a linear array (Fig. 5-38). The older '"P~/Z~~P~age of fraction B suggests the presence of mhrinherited cores or xenocrysts undetected by visual inspection of the grains. The '07~b/206~bages of fractions

A and C agree well and a regression line fit through them is consistent with recent Pb loss. A weighted mean

'07pdMpb age of these two fractions. 202.W7.5 Ma, is interpreted as the maximum age of the tuff. A minimum age of 189.3kO.4 Ma is assigned based on the fM~b/U8~age of the least discordant fraction A. Sample 3. Andesite tuf in Telkwa Formation, Ashman Ridge, Srnithers ntap area

Sample 3 was also collected ti-om the Ashman Ridge section described above. The sampled level is 2 IO m

snatigraphically above Sample 2, near the top of a 75 ni thick massive, purple dacite co andesite flow unit that

immediately underlies the asteroceratid-bearing sedirnentary rocks described above (Level A 1 on Fig, 5.4A). Thus the biochronoIogical upper age consmint of the two sarnples are identical: the earIy Late Sinemurian Varians

Assemblage.

A moderate arnount of good qualiiy zircon with prismatic habit was recovered from the sample. The five analyzed fractions do not permit a precise age assignment. There is a good agreement in the 'W~b/"8~ages of

Fractions A and B as well as the '07~b/'06~bages of B and D that converge around 192 Ma. This is regarded as a minimum age. Pb loss effects are clearly indicated despite the strong abrasion but the Iack of colinearity of discordant Fractions also suggests minor inheritance of older Pb in fractions A, C, and D.

Sample 4 Dacite tuf in Telkwa Formation, sourhern Telkwa Range. Smithers map areu

A more than 150 m thick section comprising voIcanic, volcaniclastic, and carbonate rocks of the Telkwa

Formation is exposed in the southern Telkwa Range (Fig. 5.4E). The geology of the area is discussed by Desjardins et al. (1990) and the section is described in detail by Ph19 and Schmidt (1994). Sample 4 was obtained From a maroon, crystal-lithic dacite tuff unit in the Iower part of the section. Paitechioceras cf. boehmi occurs in sparsely fossiliferous, micritic to bioclastic limestones both below and above the tuff. P. CEroihpieei and P. ex gr. apianarum was collected some 15 rn above the tuff. The ammonite fauna unequivocally indicates that the isotopically dated level is within the Upper Sinemurian Harbledownense Assemblage.

Zircons from Sample 4 were colourless to pale brown, rnostly stubby prismatic or tabular pins. The six analyzed fractions clearly define a discordia Line indicating recent Pb loss (Fig. 5.3D). Analytical problems encountered during mass spectrometry of fractions E and F render them less reliable, as manifest in the slight reverse discordance of E. Fractions A ttirough D fonn a linear array and yield a weighted mean '07~b/'0b~bage of

194.0I9.1 Ma. Considering the 206~b/?Jage of 192.Si0.6 Ma from the most concordant fraction C as a minimum age, we interpret 194.0+9.l/-1.8 Ma as the age of the tuff. Sumple j. Rhyolite in Telkwa Formation, southern Telkwu Range, Smithers map area

Sample 5 was collected fiom near the top of the sarne section that yielded Sample 4 (Fig. 5.4E). More than

100 m of epichstic and pyroclastic strata separate the sampled rhyolite flow unit fiom the underlying fossiliferous

limestone of Late Sinemurian (Harbledownense Assemblage) age described above. The stratigraphie distance does not necessarily represent significant difference in age as sudden facies changes and rapid deposition of thick volcanigenic strata are typical in the proximal volcanic facies of the Hazelton Group (Tipper and Richards. 1976).

Faunas younger than Late Sinemurian are not known from the Telkwa Formation. Regionally, the oldest faunas reported from the overlying Nilkitkwa Formation are of Early Pliensbachian (Whiteavesi Zone) age (Tipper and

Richards, 1976). Therefore the age of the rhyolite is regarded as Late Sinemurian (Harbledownense Assemblage) or slightly younger but no younger than the Early Pliensbachian Whiteavesi Zene.

The three zircon fractions analyzed form a tight cluster with overlapping '06~b/?Jages (Fig. SE).

Fractions B and C are concordant while fraction A is only slightly discordant but overlaps with B. We interpret the age of the rhyolite as 19 1 SIO.8 Ma. based on the '06~b/38~ages of B and C.

5.4.2. Pliens bachian

Sample 6. Dacitic lapilli tuf in Hazelton Group, Todagin Mountrrin, Spuwizi mup urea

A well-exposed section of Lower Jurassic volcanic, volcaniclastic and sedimentary rocks occurs southwest of

Todagin Mm, east of Kinaskan Lake (Ash et al., 1996). Locally, the upper part of the HazeIton Group consists of a bimodal, basaltic to rhyolitic volcanic suite with common, predominantly fine-grained, sedimentary intercalations

(Ash et al., 1997). Ammonites fiom the sequence suggest a Pliensbachian age (Tipper in Ash et al., 1996) whereas a spatially closely related felsic volcanic unit yielded a preliminary U-Pb age of 18 1 .O +5.9/-0.4 Ma (Ash et al.,

1997). Subsequent detailed field work for this study aimed at sampling and dating felsic volcanic rocks unambiguously bracketed by ammonite-bearing strata. Sample 6 was collected from a waterlain, quartz-phyric lapilli tuff that locally contained rip-up clasts fiom the underlying rnudstone. The section is structurally intact and the tuff is underlain and overlain by shale, siltstone, and sandstone of undoubted Freboldi Zone age. proven by the zonal index Dubaricerrzsfreboldi. Additional taxa, Metaderoceras sp. and Oisfoceras?sp. found immediately above the tuff, confm the late Early Pliensbachian age assignment. The fauna is closely similar to that known

hmelsewhere in the Spatsizi area (ïhomson and Smith, 1992).

Sample 6 yielded a more than average quantity of zircon for volcanic rocks. Quality was variable and visually the best grains were included in Fractions C and D, of which D had high-U content and contained slightly magnetic zircons. As expected, fraction C is the most concordant fraction with a slight reverse discordance while fraction D is only slightIy discordant and the two error ellipses nearly overlap (Fig. 5.3F). The '06~bl?Jages of these fractions are in good agreement at 185.610.6 Ma that it can be argued to represent the mecrystallization age. The obvious Pb loss in fractions A and B and the somewhat older 207~b/2"~bages of A, 6,and D suggest caution as the weighted mean '"~b/'~~bage of al1 fractions is 19 I-7k4.9 Ma (Le.. outside the error of the oldest 'm~b~"8~ages).

ïhe corresponding discordia line is poorly fitted (MSWD=3.0). We derive a conservative age estimate of

185.6+6.11-0.6 Ma by placing greater confidence in the '"Pbl?J ages but extending the error range to include the weighted rnean '07~b/r"~bage.

5-43. Toarcian

Sample 7. Crystal tufin Hazelton Group, west of Mount Brock Spafsizi map area

One of the few places in the Hazelton Trough where a substantial pile of Toarcian voIcanic rocks is exposed is at and near Mount Brock in the northwest corner of the Spatsizi map area. This volcanic unit within the Hazelton

Group is informally known as the Mount Brock volcanics (Thorkelson et al., 1995). On an unnarned rnountain 10 km northwest of Mt. Brock, the volcanic rocks interfinger with fossiliferous sedimentary rocks (Smith et al.. 1984:

Thomson et al., 1986; Read and Psutka, 1990). Part of the section measured on the west-facing rnountainside is show in Fig. SAD; further details wananted on the merit of its ammonite record that spans the

Pliensbachiafloarcian boundw will be published elsewhere. Out of three sarnpled volcanic units here, only the stratigraphically highest yielded interpretable U-Pb results. Sarnple 7 was taken from a t .5 rn thick, felsic, non- welded, waterlain, feldspar-phyric crystal tuff. The sarnpled interval lies about 8 m above a 3-4 m thick, rernarkably fossiliferous bioclastic limestone bed (Fig. 5.4D). Earlier collections by Read and fsutka (1990, GSC

Loc. C-90698) yielded Dactyliocerus, identifred either as D. cf. commune (Jakobs et a[., 1994) or D. cf. pseudocommune, D. cf. simpla, and D. cf. knnense (Jakobs, 1992). Our new collection and its evaluation confirm the latter identifications and the concomitant age assignment to the earliest Toarcian Kanense Zone. It is further supported by the finding of Lioceratoidespropinquum, a guide fossil of the Pliensbacifloarcian transition in North America, at a lower level. Fossiliferous sedimentary intercalations become sparse within a thick sequence of

andesitic volcanic rocks that overlie the dated tuff. Two fossil collections are used to provide an upper age bracket.

One contains bivalves and an ammonite originally identified as Polyplectus sp., on which a Middle Toarcian age

assignment was based (H.W.Tipper, pers. comm., 1996). The other one, a new collection farther upsection, yielded

Daciylioceras sp., whose range is mutually exclusive with PolypIectu. Assuming no structural repetition, this

finding suggests an Early to early MiddIeToarcian (Kanense to Planulata Zone) age for the entire succession, that

cari be reconciled with the first collection if its diagnostic ammonite is re-interpreted as an indeterminate

harpoceratid.

Sampte 7 yielded abundant, clear, equant to elongate zircons that are relatively couse for volcanic rocks and

have a hi& U content (up to 850 ppm). Two unabraded Fractions (A and B) and two abraded fractions (C and D)

mutually overlap (Fig. 5.3G). The slight discordance and younger 2a~bP8~age of the unabraded fractions indicate

that only mild Pb loss aFfected these zircons. Fraction D touches the concordia curve at a '06~blL?8~age of

180.4*0.4 Ma. Fraction C is slightly discordant but overlaps fraetion D. which allow their 206~b/38~ages to be

regarded as an estimate of the unit's age. Despite the small difference in "P~/~'uages of abraded and unabraded

fractions, al1 four fractions fit on a well-defined discordia Iine that intercepts concordia at an age of 188.Ji2.5 Ma.

This is significantly older than the '"P~/~'u age of the nearly concordant fraction D. The dilemma of this age

interpretation is therefore to reconcile these two ages. Our solution is to take the 2w~b/38~age of D and extend its error to the weighted mean '"~b/'~~bage, yielding a 180.4+8.O/-û.4 Ma age for the tuK

Sample 8. Tufflayer in Srnithers Formation, Diagonal Mountain, McConneH Creek mup ureu

A thick succession of colour-banded fine clastics of the Hazelton Group near Diagonal Mountain resem bles and is partly correlative with uni& known fiom elsewhere in the Hazelton Trough, including the Yuen Member of the Srnithers Formation (Tipper and Richards, 1976) and the Quock Formation (Thomson et al., 1986). The geology of the area is discussed by Evenchick and Porter (1993). Ammonite biostratigraphy of the Diagonal

Mountain area is subject of an on-going study by G. lakobs who carried out extensive field work and detailed collecting £tom measured sections (lakobs, 1993). The dated zircon samples are cafibrated to her sections and we use her unpublished data complemented by Our own results fiom collections made in the immediate vicinity of U-

Pb samples. Publication of the ammonite biostratigraphy and taxonomy is forthcoming by G. Jakobs. Sarnple 8 was collected hma weathered and altered tuff layer within the thick succession of colour-banded fieclastics (infonnally known as "pyjama beds") of the Hazelton Group 4 km SSW from the peak of Diagonal

Mtn. A richly fossiiiferous bed some 25 rn below the dated tuff yielded Yakounia cf. silvae (Jakobs. pers. comrn..

1995) and Pleydellia sp,, clearly indicating the uppermost Toarcian Yakounensis Zone. From approximately 8 m above the sampled tuff, sonninid ammonites were recovered (Jakobs, pers. comm., 1995). Similar, poorly preserved sonninids were also found in the talus. While the poor preservation precludes a precise identification, an

Early Bajocian age is suggested. The biochronologic age of the sampled tuff layer is thus bracketed between the latest Toarcian and Early Bajocian.

SarnpIe 8 yielded a small number of subequant to stubby pnsmatic, euhedral to subhedral zircon crystals. It was spIit into tfiree srnall îkctions (A-C) and one larger fraction (D) that were al1 strongly abraded. Results from the smallest fraction (B) are reversely discordant, analytically infeior, and not considered further. Fractions A. C. and D define a discordia line with good fit (MSWD=0.03) that defines an upper intercept at 179.8 Ma (Fig. 5.3 H).

The discordance is attributed to Pb Ioss, a cornmon phenomenon in the Diagonal Mm. samples. Our best estirnate of the unit's age is I79.tk6.3 Ma, the weighted mean '07~bl'06~bage of fractions A, C. and D.

5.4.4. Aalenian

Sample 9. Rhyoliric tuflin Srnithers Formation, Troitsa Ridge, Whitesail Lake rnap area

A ridge West of Troitsa Peak, Whitesail Lake map area. is the only known section in the Canadian Cordillem where a succession of al1 three Aalenian ammonite zones is present (Poulton and Tipper, 199 1). The upper part of the Hazelton Group (Srnithers Formation) exposed here contains minor felsic volcanic units interbedded in a predominantIy sedimenmy sequence (Fig. 5.4B). A description of the general stratigraphy of the area is given by

Diakow and Mihalynuk (1987). Our attempt to isotopically date a volcanic unit from al1 three Aalenian ammonite zones only gave meaningful results, albeit of poor quality, for the highest zone. Sample 9 is collected fiom an 8 rn thick, soft, well-bedded, felsic, waterlain cryxtal-lithic tuff interbedded with brown, fossiliferous siltstone and fine- grained sandstone. Ammonites found immediately below include Erycitoides howelli and E. cf. reres whereas E. cf spinosum and E. sp occur some 25 m above. These levels also fall into the vertical range of Tmerocera.scimtint and T. kirki. These assemblages unequivocally establish the late Aalenian Howelli Zone as the age of the tuff. Sample 9,25 kg in weight, yielded a srnall amount of fine-grained, clear, prismatic to needle-like zircons. It was subdivided into four Fractions, of which only one (fiaction A) was abraded. This Fraction is only slightly discordant whereas the three unabraded and very fine-grained fractions exhibit significant Pb loss (Fig. 5.31). The

207 ~b/'~~Pbages of Fractions B and C suggest a nearly complete resetting in Cretaceous time, presumably due to the emplacement of nearby plutons of the Coast Plutonic Complex. The '*~b/'~~Pbages of Fractions A and D, however, are within error and allow the calculation ofa weighted mean age of 174.8*16.0 Ma which can be further refined fiom the '06~bP8~age of fraction A as a minimum age to derive a bat age estimate of 174.8+16.0/-12.4

Ma.

5.4.5. Bajocian

Sample 10. Tufllayer in Srnithers Formation, Diagonal Mountain. McConnelI Creek map area

On a ridge 4.5 km SSW of Diagonal Mm., color-banded fine clastics interbedded with rare tuff layers are exposed. These strata, informally referred to as "pyjama beds", are assigned to the upper part of the Hazelton

Croup. (Refer to Sample 8 above for details of the geology of the area) A fossiliferous bed some 30 m below the zircon sample was first located by G. Jakobs who identified Leptosphincres (Prorsisphincfes) cf. meseres, L. cf. ciifensis, and Stephanoceras sp. (Jakobs, pers. comm.. 1996). Elsewhere in the section, in which ammonites occur only sparsely, poorly preserved sonninids and stephanoceratids were found. The age of the sampled tuff layer is early Late Bajocian based on the CO-occurrenceof perisphinctids and stephanoceratids, characteristic of the

Rotundum Zone (Hall and Westermann, 1980).

Sample 10 was collected from a 5 cm thick, blue-grey, cIay-rich tuff layer. Only a small amount of zircon, cornprising mostly fine, subequant to stubby prisrntic crystals, was recovered from a 25 kg sample. Four very small

Fractions were analyzed. A and D were strongly abraded whereas B and C were only moderately abraded. As in the other samples from the Diagonal Mountain area, Pb loss is evidenced by the younger '06~bP8~ages of the less abraded fractions (Fig. 5.35). Due to the small size of the fractions and the relatively large amount of common Pb present, the 206~b/ZOJ~bratio is less than 600 and al1 analyses are rather imprecise. The '07~b/'06~bages are within error, mainly because of their large individual errors. The '"~b/'~Pbages of fractions A and D are in good agreement at about 170 Ma. The 207~b/206~bage of Fraction C is sornewhat older, and that of B is significantly older, perhaps indicating minor inheritance. A three-point regression (involving fractions A. C, and D) yielded a

I77.ïIlI .3 Ma upper intercept age, whereas a line fit through onIy fractions A and D yielded 170.51 15.8 Ma. 90 Fraction D is concordant with a 206~b/?J age of 167.2i0.4 Ma. We combine the -?b/'38~ age of the concordant fraction with the more conservative three-point weighted rnean '07pb/'d6pb age to assign a best age estimate of

1672+ [OS/-0.4 Ma

5.4.6. Bathonian

Sample II. Reworked tuflin Ashman Formation,McDonell Lake, Srnithers map area

At the east end of McDonell Lake, a logging roadcut exposes a short section of tuffaceous sandstone of the

Ashman Formation (Fig. SAC). The locality is about 20 km SE of Ashman Ridge, the type section of the formation

(Tipper and Richards, 1976). Outcrops of Bajocian to Oxfordian fossiliferous strata are disnipted by block faults in the vicinity of McDonell Lake (Frebold and Tipper, 1973). Within the sequence of well-bedded, fossiliferous. green to buR, Iithic sandstone, a 2 m thick interval was identified as wateriain, strongly reworked tuff or tuffaceous sandstone. based on the abundance of feldspar and the presence of glas shards. Two samples. (Field No. 95JP78 and 95JP79) were taken korn adjacent beds that have no demonstrable difference in biochronologic age.

Iniskinites sp. and perisphinctids (PI. S. 1, Figs. 6,9) occur both irnrnediately below and above the isotopicalIy dated tuff. Ammonites recovered fiom above the tuff include Kepplerites ex gr. rychonis (PI. 5.1, Fig. 5).

Keppleriles sp. (PI. 5.1, Fig. I), Parareineckeia? sp. (PI. 5.2, Fig. 4), Lilloetlia cf. Iilloefenris (PI. 5.2. Fig. IO), and

Xenocephaiites cf vicariw (PI. 5.2, Fig. 6). The assemblage best agrees with Fauna B6 of Callomon (1984).

Perhaps the most stratigraphicaIly diagnostic form among them is K. ex gr. pchonis that allows correlation with the basal Upper Bathonian where the genus fint appears, represented by similarly tinely ribbed forms throughout its area (Callomon, 1984). Parareineckeia was thought to be restricted to the Bajocian (Callornon, 1984) but its co- occurrence with Iniskinites, Liiloettia, and Kepplerites suggests that it ranges up to the Bathonian.

Approximately 25 kg samples fiom sites 95JP78 and 95JP79 yielded enough zircon to analyze three and five fractions, respectively. To better resolve the age of the rock, another 25 kg sample from site 95JP78 was processed and three additional fractions were selected and analyzed based on the experience gained fiom the earlier analyses.

Analytical results from the hvo samples are combined in Sample 1 1 as they represent adjacent and essentially coeval beds. EXPLANATION OF PLATES 5.1 AND 5.2 (on following two pages)

Ail figures are 95% of natural size. Figured specimens are housed in the type collection of the GeologicaI Suwey of Canada (GSC). For each specimen the type number is followed by the field number of IocaIity (referred to in text and figures) and the GSC locality number.

Plate 5.1. (On p. 93) Late Bathonian ammonoids of the McDonell Lake and Zymoetz (Copper) River sections.

Fig. 1. Kepplerifes sp., McDoneIl Lake, GSC 1 1 1520,95JP82, GSC Loc. C-176587. Fig. 2. Perisphinctid sp. A, Zymoetz (Copper) River, GSC 1 1 152 1,95JP83. GSC Loc. C- 176588. Fig. 3. "Choflafia"sp., Zymoetz (Copper) River, GSC 1 1 1522,95JP85, GSC Loc. C- 176590. Fig. 4. "Choflatia" sp., Zymoetz (Copper) River, GSC 1 1 l523,95JP87, GSC Loc. C- 176592. Fig. 5 Kepplerifes ex gr. rychonis (Ravn), McDonell Lake, GSC 1 1 1524,95JP82, GSC Loc. C- 176587. Fig. 6. Perisphinctid sp. B, tvlIcDonell Lake, GSC 1 1 1525, 95JP8 1, GSC Loc. C-176586. Figs. 7-8. Kepplerifes sp., Zymoetz (Copper) River, GSC 1 1 1526,95JP87. GSC Loc. C- 176592. Fig. 9. Perisphinctid sp. B, McDonell Lake, GSC 1 1 l527,95JP8 1. GSC Loc. C- 176586. Fig. 1 O. Kepplerim ex gr. logonianus (Whiteaves), Zyrnoetz (Copper) River, GSC 1 1 1528, 9SJP87, GSC LOC.C- 176593.

Plate 5.2. (On p. 94.) Late Bathonian ammonoids of the McDonell Lake and Zymoetz (Copper) River sections.

Fig. 1. Iniskinites cf. martini (Imlay), Zymoetz (Copper) River, GSC I 1 1529,95JP83, GSC Loc. C- 176588. Figs. 2-3. Xenocephalires cf. vicarius Imlay, Zymoetz (Copper) River, GSC 1 1 1530.95JP87. GSC LOC.C- 176592. Fig. 4. Parareineckeia? sp., McDonell Lake, GSC 1 1 153 1,95JP82, GSC Loc. C- 176587. Fig. 5 Iniskinites or Lilloertia sp. (inner whorls), McDonell Lake, GSC 1 1 1532,95JP8 1, GSC LUC.C- 176586. Fig. 6. Xenocephalim cf. vicarius Imlay, McDonell Lake, GSC 1 1 1533,95JP81, GSC Loc. C- 176586. Fig. 7. Cadoceras cf. moffili May, Zymoetz (Copper) River, GSC 1 1 1534,95JP83, GSC Loc. C- 176588. Fig.8. Lifloetfiacf. lilloetensis Cnckmay, Zymoetz (Copper) River, GSC I 1 1535.95JP86. GSC LOC. C- 1 7659 1. Fig.9. Iniskinifes or Lilloeftia sp. (inner whorls), McDonell Lake, GSC 1 1 1536,95JP8 1, GSC LOC.C- 176586. Fig. 10. Lifloettia cf. lilloetemis Crickmay, McDonell Lake, GSC 1 1 1537, 95JP8 1, GSC LOC.C- 176586. Plate Plate 5 Zircon populations from the two samples were homogenous, dorninated by morphologically simple, euhedral

crystals ranging in aspect ratio from subequant to elongate, needle-like grains. Apart from fraction E, al1 other

fractions were moderately to sirongly abraded. The discordance of unabraded fraction E demonstrates that minor

Pb loss affected the zircons (Fig. 5.3K). Fraction A is an inferior analysis with an unusually high arnount of

common Pb and is not considered further. The other fractions fom a cluster centered around the most concordant

fraction G that yielded a '06~b/?.Jage of 158.20.4 Ma. It is within error with the '06~b/"8~ages of fractions F and H that together with G,were also subjected to the greatest amount of abrasion. Slight discordance and reverse discordance of other fractions with similar '06~b/'38~ages suggest that the measurement of '07~b/';5~ratio in small

fractions is less reliable. To account for the possibility of any surface-correlated Pb loss not completely removed by abrasion, we extend the error range of the '06~b/U8~age of fraction G to include the oldest apparent '"pb/'js~ age of fraction D which suggests an interpreted age of 158.2+1.9/-0.4 Ma.

Sample 12. Tufllayer in Smithers Formation, Diagonal Mountain, McConnell Creek map area

Sample 12 was collected fiom a 10 to 20 cm thick, crearn-colored tuff layer interbedded in a thick succession of dark mudstone of the Ashrnan Formation approx. 2.4 km SW of Diagonal Mountain. (Refer to Sample 8 above for the geology of the area.) Ashman Formation strata are tightly folded in the sampling are* rendering smtigraphic observations difficult. hiskinites sp. was collected both below and above the zircon sample. Earlier collections yielded hiskinites cf. robuslus below and Adabofolocerm or Lilloeftia above (G,Jakobs, pers. comm..

1996). hiskinites is taken to indicate the Late Bathonian as the most likely age of the sample although precise correlation around the BathonianICallovian transition is dificult (Callornon, 1984; Poulton et al.. 1994).

This sample yielded a srnall amount of fine-grained zircon. Only two fractions yielded usable analyses.

Fraction A is discordant with an old '07~b?O"~bage suggesting inheritance whereas fraction C lies near the concordia curve (Fig. 5.3L). A two point discordia line yields a lower intercept age of 158.410.8 Ma and an upper intercept age of 1.33I0.14 Ga. As the experience with the other samples from Diagonal Mountain (Sample 8 and

10) dernonstrated, the possibility of Pb loss not rernoved by moderate abrasion of fraction C cannot be ruled out.

Thus the lower intercept age is interpreted as a minimum estirnate of the unit's age. The maximum estimated age is deterrnined by the '0'~b/206~bage of this fraction, 175.2I8.2 Ma. Sample 13. Reworked tufllayer in Ashman Formation, Zymoetz River, Smithers map area

Richly fossiliferous shallow marine sandstone of the Ashman Formation is exposed along a new logging road

2.5 km northeast of the confluence of Limonite Creek and Zymoetz (Copper) River. According to Tipper and

Richards (1976), this locality lies close to the southern limit of BathonianKallovian marine deposits in the Bowser basin. Another section at Tenas Creek occupies an analogous stratigraphie and paleogeographic position and its stratigraphy and ammonite fauna is discussed in detail by Tipper and Richards (1976) and Frebold (1978), respectively. Only one fossil collection frorn Zymoetz River is mentioned by Frebold (1978) and that outcrop of the

"Copper River fossil beds" was described in detail by Jakobs (l986).The new roadcut provides a fresh outcrop of the fossiliferous marine sedimen- rocks that overlie (presurnably with discomfomity) rnaroon tuffs that likely belong the Red ïuff Member of the Nilkitkwa Formation. An 8 cm thick, pale blue, tuffaceous, clay-rich layer containing rare marine fossils was identified as a reworked volcanic ash, It occurs 8.5 rn above the base of the sandstone and it was sampled for U-Pb dating.

Figure 5.4F shows the rneasured section and the vertical distribution of amrnonoids. The assemblage undoubtedly indicates the BathonianKallovian transition. In the framework of North American faunal successions given by Callomon (1984), the assamblage recovered imrnediately below the tuff, including Cadoceras cf.'rnagTri

(PI. 5.2, Fig. 7), lniskinires cf. marrini(PI.5.2, Fig. l), and Xenocephalites cf. vicarizu, clearly corresponds to

Fauna B7(a) that is correlated with the Late Bathonian. Among othen, "Choflaria" sp. (PI. 5.1, Fig. 3,4),

Kepplerites ex gr. logunianus (PI. 5.1, Fig. IO), K. sp. (PI. 5.1, Fig. 7-8), and Lilloetria cf. Iilloefensis (PI. 5.3, Fig.

8) were coIIected €rom above the tuff layer. These can be accommodated in either the latest Bathonian or the Early

Caliovian (Faunas B7-B8 of Callomon (1984). see also Poulton et al. (1994)). Recent studies in Mexico (Sandoval et al., 1990) and South Arnerica (Riccardi et al., 1991 and references therein) suggest that Lilloertia. represented by species closely similar to L. lilloerensis, is rnost abundant in the latest Bathonian Steinmanni Zone and ranges up to the basal Callovian Vergarensis Zone. The South Arnerican biochronology employs common taxa with the

Mediterranean province for correlation with the northwest European standard zonation. The Boreal genera

Caduceras and Kepplerites offer another means of correlation via Alaska, northern Canada, and Greenland, confming the Late Bathonian-Early Callovian age.

An approxirnately 40 kg sarnple (Sarnple 13) yielded less than 1 mg of fine-grained zircon. The grains were clear, colourtess, containing only a few srnaII, clear inclusions, and ranging in morphology from subequant to elongate with a majority of stubby prismatic crystals. The available material was split into nvo very small fractions.

Fraction A marginally intersects concordia whereas the analysis of Fraction B strongly indicates the presence of 96 inherited zircon component (Fig. 5.3M). A line drawn through the two points yieIds a poorly consmined

Proterozoic upper intercept age of 1.30i0.69/-0.55 Ga. The lower intercept at 162.6+2.9/-7.0 Ma is interpreted as

the best estimate of the age of the tuff. It should be noted, however, that the relatively imprecise analyses of small

amounts of zircon are sensitive to the blank Pb correction which strongly affects the apparent "'Pb/% and

107~b/"~Pb ages. Moreover, the slight discordance of fraction A could alternatively be explained by minor Pb loss

not eliminated by the strong abrasion. More analyses, requiring re-sampling of an even larger amount of rock,

would be needed to bener resolve the age of the tuff layer.

5.5. DISCUSSION

Several recent studies demonstrate that U-Pb dating of volcaniclastic rocks intercalated with fossiliferous strata offers one of the best techniques for tirne scale calibration (Tucker and McKerrow, 1995; Mundil et ai.. 1996). Data presented here underscores that voIcanic arc terranes provide a suitable geoIogical setting for such studies. In volcanosedimentary sequences. the ratio of volcaniclastic and marine sedimentan, rocks is a function of proximity to the emptive centre. We found that integrated isotopic and biochronologic dating can be used successfully in moderately proximal sections (as in Samples 4,5,6, 7, and 1 1, see also Samples 7 and 3 for pitfaIls) where of substantial voIcaniclastic units and ammonite-bearing sedimentary rocks are equaiIy abundant. Dating of thin ash beds deposited in a distal setting is promising as there is often a continuous fossil record for zona1 biochronology. However, U-Pb dating is oflen plagued by low zircon yields (see Samples 1.8, 10. 12, and 13). In these cases the capability of analyzing single grains, not attempted in this study, is essential to produce sufficicntly precise ages (Mundil et al., 1996; P6IQ et al., 1997).

Interpretation of U-Pb ages has often proved difficult due to complex isotopic systernatics characterized by Pb loss and inheritance. Pb Ioss was rnost pronounced in samples in proximity to the Coast Plutonic CornpIex and in the thin ash layers. inheritance patterns in Early and Middle Jurassic sarnples from Stikinia reveal the presence of slightly older zircon components in Sample 2 and strong Proterozoic (approx. 1.3 Ga) inheritance in Mo upper

Middle Jurassic samples (12 and 13). In the first case the rninor inherited component may have derived From underlying volcartic strata of the Upper Triassic Stuhini Group or its comagrnatic plutons (Thorkelson et al.. 1995).

Our finding of Proterozoic inheritance is compatible with rnost tectonic models that suggest accretion of Stikinia to

North America by late Middle Jurassic tirne. We did not tind any Proterozoic component in older Hazelton arc volcan ic rocks. A measure of reliability of our data is their internai consistency whereby isotopic and biochronologic age relationships are in agreement. The two most precise isotopic dates (samples 5 and1 1) have less than 1.5% Za error ranges. These new calibration points far exceed the quality of those in most recent time scales (Harland et al., 1990;

Gradstein et al., 1994). The next class comprises sarnples 4,6, 7, and IO where a relatively large positive error results from the conservative use of the weighted mean 207~b/'W~bages to complement the age of a concordant fraction if that could not be duplicated. The 2serror range in these cases is 3 to 7%. Comparable precision is obtained using a weighted mean '07~b/'W~bage for Sample 8 and a lower intercept age for Sample 13. These six items meet or exceed the average precision of isotopic dates used in previous time scales in addition to being superior to them in their zona1 biochronologic constraints. Four of the rernaining five isotopic dates reported here are minimum ages only (Samples 1,2,3, and 12) and another one is of poor precision (Sample 9). However. on the strength of their precise stratigraphie age, they also provide useful constraints for the time scale.

The Early and Middle Jurassic portion of the time scales of Harland et al. (1 990) and Gradstein et al. ( 1994), the ones that employ the most comprehensive isotopic dataset, are based on 25 and 30 ages, respectively. The 13 new dates reponed here along with many more obtained recently by other Cordilleran workers therefore represenrs a substantial addition to the database. A revision of the Jurassic time scale is presented in Chapter 7. U-Pb DATING THE TRIASSIC-JURASSIC

MASS EXTINCTION BOUNDARY

6.1. INTRODUCTION

The Triassic-Jurassic (T-J) boundary is marked by one of the five biggest mass extinction events (Raup and

Sepkoski, 1982) but it is, perhaps, the most poorIy understood (HaIIam, 1990). In addition to the widely debated causation, its timing is aIso poorly constrained. T-J boundary age estimates in widely used time scales Vary

between 2 lm3.5 (Haq et al., 1988) and 203k3 Ma (Odin, 1994). Such disagreement is attributed to the paucity of calibration points. Precise U-Pb data near the boundary exist only from basalts of the continental Newark

Supergroup (Dunning and Hodych, 1990; Hodych and Dunning, 1992). This was used to postulate the boundary at

20 1-202 Ma (Olsen et al., 1996) but correlation of marine and continental sequences has not been independently and unequivocally demonstrated (Hallam, 1990). The quandary is avoided if precise isotopic dating is applied in conjunction with biochronology of marine fossils that allow unarnbiguous correlation with the T-J system boundary.

Here we present such results fiom the Queen Charlotte Islands, western Canada, where a new U-Pb date was obtained fiom immediately below the boundary. This date is pooled with 12 other recently obtained and pertinent

U-Pb dates to provide a tightly constrained estimate of the age of the T-J boundary. The updated tirne frarne help focus discussions concerning the end-Triassic mass extinction event and the spacing of mass extinctions. Figure 6.1. Locality index map of Triassic-Jurassic boundary sections in the Queen Charlotte Islands, British Columbia, Canada.

6.2. A NEW U-Pb AGE NEAR THE T-J BOUNDARY IN THE QUEEN CHARLOTTE ISLANDS

The preferred method for dating a stratigraphie boundary is to locate isotopically datable volcanic or voIcaniclastic rocks interbedded in a continuous and fossiliferous section across the boundary. The Sandilands

Formation in the Queen Charlotte Islands is a sequence of thin-bedded shale and siltstone (Cameron and Tipper,

1985) that ranges in age from the Norian to the early Pliensbachian (Desrochers and Otchard, 199 1; PAlfy et al.,

1994). it records an apparently uninterrupted T-J transition exposed at two locales: Kunga Island and Kennecou Point (Tipper et al., 1994) (Fig. 6.1). Light-colored layers causing a characteristic banded appearance were suggested of volcanic ash ongin (Cameron and Tipper, 1985). Immediately below the T-J boundary on Kunga

[siand, a 3.5 cm thick tuff layer was located and sarnpled (Sample KI- I), The undulatory but sharp base and top, somewhat graded texture, and the presence of crystals and small mudstone ripup clam indicate a reworked ash bed.

U-Pb zircon geochronometry was carried out in ihe Geochronology Laboratory of the University of British

Columbia utilizing standard procedures in minera1 separation, zircon chemistry, mass spectrometry. and data reduction (Mortensen et al., 1995). U and Pb Iaboratory blanks were approximately 1 and 8 picograms. respectively.

Zircons were subdivided into four fractions (Table 6.1). Fraction A is slightly reversely discordant, fractions .

B and C are slightly discordant and overlapping, while fiaction D is squarely concordant (Fig. 6.2). The lack of additional concordant analyses hinden an unarnbiguous interpretation of the age. Some Pb loss is indicated by the

195 to 200 Ma range of calculated '06~b/'38~ages. The clustering of 2M~b/38~ages of three out of four fractions around 198 and 200 Ma indicates that in these cases Pb los was almost completely eliminated by the strong

Figure 6.2. U-Pb concordia diagram of zircons fiom a 3.5 cm thick tuff layer immediately below the T-J boundary, Sandilands Formation, Kunga Island. A homogenous population of ca. 0.5 mg zircon was recovered from a 10 kg sample. A11 grains were smaller than 104 Fm, colorless, of excellent clarity without visible cores or zoning, and containing very rare inclusions oniy. Morphologies ranged continuously fiom anhedral and slightly resorbed, oval to elongate grains to euhedral, simple, stubby prismatic to elongate needle-like or tabular crystals. IO1 Wt U Pù*' Pb' LFotopic ratios (*lm.%)* Calculad aga (*Iu.Ma)" Sample 1 Fmction'

mg ppm PP~ pg .h "ebW~ MPbP"U "'PblWPb ""PW"U "'Pbt2'U '"PblSaPb

Table 6.1. U-Pb zircon analytical data of sample KI-[.collected on the southeast shore of Kunga Island, 150 000 map area NTS lO3B/l3- 14, UTM 327280E, 5848350N. [AH zircon fractions. Listed in brackets: Grain size range in microns; Side dope of Franz magnetic separator (in degrees) at which grains are non-magnetic (N) or magnetic (M), using 20" front dope and 1.SA field strength; Grain character: eh=euhedral, sh=subhedral. ah=anhedral, n=needles, sp=stubby prismatic, Ftabular, el=elongate, ov=oval; Degree of air abrasion: s-a=strongly abraded. 2~adiogenicPb 3~easuredratio corrected for spike and Pb hctionation of O.O036/arnui 20% (Daly collecter). 4~otalcommon Pb in analysis based on blank isotopic composition 5~adiogenicPb 6~orrectedfor blank Pb, U. and initial common Pb based on the Stacey and Kramers (1 975) mode1 at 200 Ma.

abrasion. Therefore 199.6k0.4 Ma, the oldest 'O"PblU8~age given by the concordant fraction D. can be regarded as

the crystallization age of the sample. An alternative interpretation relies on the calculated '07~bl'"~bages that FaIl

between 183 and 208 Ma. If fiaction A is dismissed due to its reverse discordance, the other three fractions yield a

weighted mean '07~b/206~bage of 205.714.8 Ma.

The two interpretations disagree beyond their respective error. The weighted rnean '07~b/'06~bage is regarded

as less reliable because in Jurassic zircons the abundance ofZmpbis more than 20 times greater than that of ?O7pb.

Consequentiy, the rneasurement of the latter is inherently less precise and the calculared '07~b/x5~and 207~b/'w~b

ages are strongly dependent on the composition of blank and initial cornman Pb used in corrections. Also, the

discordia line and its upper intercept are not well constrained due to the small spread of fractions B. C, and D.

There are additional geological arguments to help interpret the U-Pb systernatics. The themial history of the

Sandilands Formation on Kunga Island is determined frorn conodonr alteration index values benveen 4.5 and 5.5.

indicative of temperatures in the 245-5 15'C range (Orchard and Forster, 199 1) that is compatible with the results of

vitrinite reflectance studies (Vellutini and Bustin, 1991). The likely event responsible for heating is the

emplacement of nearby Middle Jurassic plutons of the Burnaby Island Plutonic Suite that consists of high-level

plutons with rapid cooling (Anderson and Reichenbach, 1991). The Sandilands Formation remained well below the 102 >800°C closure temperature of zircon (Haeman and Pmish, 199 1) throughout its history rhus any Pb los should

be attributed to recent processes. The low U content (140-1 70 ppm) and the excellent quality of zircons supports

the interpretation that the surface-correlated Pb los was mild and, in case of fraction D, completely removed by

abrasion. To de-emphasize the apparent but not reproduced precision, we round up the age and error to assign

20WI Ma as the age of the tuff layer.

6.3. BIOCHRONOLOGY OF THE T-J BOUNDARY IN THE

QUEEN CHARLOTTE ISLANDS

The T-1 boundary sections on Kunga [siand and Kennecott Point in the Queen Charlotte [stands are

recognized through integrated biochronology using ammonites. radiolarians, and conodonts (Orchard. 199 1 a;

Carter, 1993; Carter, 1994; Tipper et al.. 1994; Carter et al., in press). A surnrnary of these detsiled studies help

provide a biochronologic frarnework for the isotopically dated tuff. SampIed levels, their diagnostic fossils and

zonai assignments are shown on Fig. 6.3. In the critical interval, radiolarian biochronology is the most cornpiete.

The tuff occurs very near the top of the latest Triassic Globolaxtorum towi Zone (equivalent of the ammonite

standard Crickmayi Zone), based on an abundant and diverse fauna (Carter, 1993). The last appearance of the zona1

index (at site R4, Fig. 6.2) is 4 m above the tuff (total thickness of the zone is 26 m) whereas the next higher

collection (RS) 2.5 m Farther upsection yields an impoverished and markedly different assemblage with

Pantaneiiium tanueme, c1early indicative ofthe Early Hettangian (Carter, 1994). Conodont faunas recovered From

Cl-C5 are of low diversity but diagnostic to the Rhaetian. The highest collection. 4 rn above the tuff, yielded

Neogondolella which in the Kennecott Point section occurs slightly above Misikella poszhernsreini and the

ammonite Choristocerm nobile, both guide fossils of the terminal Triassic (Orchard, 199 1 a; Tipper et al., 1994).

No Late Triassic ammonoids have been found yet in the Kunga Island section. The oldest lurassic ammonites

recovered 18 m or more above the niff (A 1-A3, Fig. 6.2) include Waehneroceras and Discamphiceras su~gesting

early Middle Hettangian (Tipper in Carter et al., in press). In surnmary, the tuff lies immediately below the T-J

boundary defined by the disappearance of conodonts and a marked change in radiolarian faunas. 50 m (from base)

Figure 6.3. The U-Pb dated tuff layer in the T-J boundary section on Kunga Island. Key radiolarian (R), conodont (C), and ammonite (A) taxa and section measurernents are cornpiled frorn Carter (1993), Tipper et al. (1994), and Carter et al. (in press).

6.4. OTHER U-Pb AGES AROUND THE T-J BOUNDARY

Several other U-Pb dates help date the T-J boundary. Some of them (items 2 and 5 below) have been used in previous time scales (e.g., Harland et al., 1990; Gradstein et al., 1994) whereas items 1.3, and 4 are used as calibration points for the first the.

(1) Puale Bay, Alaska Peninsula: Three U-Pb ages from tuffs within a well-documented ammonite-bearing basal Jurassic section are reported in Chapter 4. A Middle Hettangian tuff layer is dated at 200.8+2.7/-2.8 Ma and

Middle to Late Hettangian tuffs yield crystallization ages of I97.8cl .Y-0.4 Ma and 197.81 1 .O Ma. The significance of these calibration points is discussed in detait in Chapter 4 and not repeated here.

104 (2) Newark Supergroup: Three U-Pb dates for mafic, rift-related volcanic and hypabyssal rocks of eastem

North America. Dunning and Hodych (1990) report a zircon and baddeleyite age of 200% 1 .O Ma on the Palisades sill and a zircon age of 20 1.311.0 Ma on the Gettysburg sill in the Newark Basin. The sills are thought to be feeden to the Orange Mountain basalt, the lowetmost extrusive rocks which immediately overlies the purponed T-J boundary based on palynological and vertebrate faunal evidence (OIsen et al., 1987; FoweII and Olsen. 1993).

From the Fundy Basin, Hodych and Dunning (1992) obtained a U-Pb zircon age of 10 1.7+1.4/- 1. l Ma from the

North Mountain basalt whose base lies some 20 m above the T-1 boundary, also defined by palynology and vertebrate biostratigraphy (Olsen et ai., 1987).

(3) Goldslide intrusions, northwest British Columbia: The Biotite porphyry, the otdest of three phases of the mineralized Goldslide intrusions on Red Mountain near Stewart (Rhys et al., 1995). yielded a 20 1 -8I0.5 Ma U-Pb date on zircon (Greig et al., 1995). Only this oIder phase is affected by pervasive cleavage related to a regional deformation event near the T-J boundary (Greig et ai., 1995). A few tens of km to the nonh, the youngest deformed strata contain the Rhaetian ammonoid Chori~ocerm(Jakobs and Pal@, 1994). Triassic radiolaria ranging in age fiom Ladinian-Carnian to Norian were recovered from strata cut by the Biotite porphyry (F. Cordey. pers. comm. 1996). Rhys et al. (1995) obtained a U-Pb zircon age of 197.6II -9 Ma from the Goldslide porphyry. the youngest post-kinematic phase. A tuff 8 km to the south that is considered correlative to the top of the Red

Mountain succession yielded a U-Pb age of 199rt2 Ma (Greig and Gehrets. 1995). On Red Mountain, abundant peperitic structures and strongly disnipted country rocks with evidence of soft-sediment deformation suggest that ar least some of the Goldslide intrusions intermingled with unlithified, wet sediments and are nearly coeval with their country rocks (Greig and Gehrels. 1995; Rhys et al., 1995). We recovered an indeterminate ammonite of Early

Jurassic aspect along with the bivalve ûxyfomastratigraphically above the syn-sedirnentary intrusions. In Europe.

Uxytoma is known as low as the Upper Triassic but in the Eastern Pacific it first appears in the Lower Jurassic with common occurrences in the Hettangian (Aberhan, 1994).

(4) Griffith Creek volcanics, northwest British Columbia: Thorkelson et al. ( 1995) obtained two U-Pb ages

(205.8*0.9 and 205.8+1.51-3.1 Ma) from a volcanic unit that is tightly folded and ovedies a conglomerate containing Upper Triassic lirnestone pebbles and Norian clastic strata of the Stuhini Group. Folding predated deposition of the younger, Lower Jurassic Cold Fish volcanics and records another exampie of the regional deformational event near the T-J boundary (Thorkelson et al., 1995) discussed above. (5) Guichon Creek batholith, south-central British Columbia: K-Ar and RbSr dates obtained from this composite intrusion influenced the placement of the T-J boundary in most time scales compiled since 1959. The current best estimate of the crystallization age is a U-Pb zircon age of 2l0I3 Ma (Mortimer et al., 1990). The dated sample is not closely associated with fossiliferous strata but sedirnentary intercalations within the Nicola Group that are crosscut by the batholith range up to the Middle Norian (Frebold and Tipper. 1969). A minimum age for the intrusion is no older than Pliensbachian based on ammonites From the onlapping Ashcrofi Formation. Therefore the age of the Guichon Creek batholith is not considered to provide a first-rate calibration point.

6.5. THE AGE OF THE T-J BOZTNDARY

The chronogram method (described in detail by Harland et al.. 1990) was used to provide an unbiased estimate for the age of the T-J boundary. The calculation is based on the eight U-Pb ages from above the boundary and five ages (including the newly reported one) from below as discussed earlier. The chronogram yields

200.9k0.6 Ma as the age of the boundary (Fig. 6.4). Because the new date reponed here heavily controls the chronogram and may marginally err on the young side, we round this value to 10 k1Ma. This estimate is younger than most previous ones or lies at the young end of their error range. It is also more precise than any other pre-

Cretaceous boundary estimate.

204 203 202 201 200 199 198 Trial age (Ma) Figure 6.4. Chronogram estimation of the T-J boundary age based on 13 U-Pb dates from the terminal Triassic and basal Jurassic (see text for a list and discussion). Previous estimates given by widely used time scales are 208k7 (Palmer, 1983), 21h3.5 (Haq et al., 1988), 20817.5 (Harland et al., 1990), 20313 (Odin, 1994), and 205.7k4.0 (Gradstein et al., 1994). 6.6. TIMING THE END-TRIASSIC MAS$ EXTINCTION

To establish synchroneity of extinctions in the marine and terrestrial realms is critical to assessing the end-

Triassic event. Correlation of a continental T-J boundary defined by palynomorphs and vertebrates (Olsen et al..

1987; Fowell and Olsen, 1993) with the marine, ammonite-based chronology is generally assumed but not infallible (Van Veen, 1995). The Kunga Island and Newark Supergroup U-Pb ages fiom immediately below the marine and above the continental boundary, respectively, agree within error. This provides strong independent evidence that the end-Triassic extinctions are indeed synchronous. This concIusion is implicit in Our chronogram calculation which incorporates marine and terrestrial biochronologic ages at face value.

Fixing the T-J boundary at 201*1 Ma implies that the Norian-Rhaetian interval may be longer than previously thought: The Cordilleran dates suggest a duration of at least 4 Ma or perhaps as rnuch as 9 Ma. There are several fossil groups, mainly shelf dwellers such as ammonites. bivalves. and conodonts, whose severe end-

Triassic extinction was preceded by an apparently protracted decline in divenity and abundance (Hallam, 1996).

Such trends are cIear in the North American Cordillera (Newton, 1983; Orchard, 199 1 b; Tozer, 1994) although the role of the Signor-Lipps effect (Signor and Lipps, 1982) has not been thoroughly evaluated yer. On the other hand. marine microplankton (radiolarians and calcareous nannofossils) shows an abrupt faunal turnover at the boundary

(Carter, 1994; Bown, 1996), similar to the continental record of vertebrates and palynornorphs (Olsen et al,. 1987:

Fowell and Olsen, 1993). This duality in temporal extinction patterns, emphasized by the relative length of the

"twilight zone" for the shelf-dwelling groups, may imply a complex interplay of causal agents in the extiction.

Parailel biological responses to a "press event" and a "pulse evenr" would suggest that a long-term environmental crisis selectively affecting shelf habitats was punctuated by a geologically rapid event that decimated the already stressed taxa and marine planktic and terrestrial ecosystems atike.

6.7. SPACING OF MASS EXTINCTIONS

The controversial hypothesis of a 26 Ma periodicity in recurring mass extinction events claimed by Raup and

Sepkoski (1984) is often criticized on the basis of inadequate time scaIes (Hofian, 1985; Heisler and Trernaine.

1989). An update of the time series analysis using the time scale of Harland et al. (1990) eliminates periodicity in some Jurassic extinction peaks while confidence in remaining perïodicities is low due to the large uncertainties of

Triassic and Jurassic time scale (Sepkoski, 1996).

Based on our estimates, the age of the end-Triassic mass extinction is 201k 1 Ma. The preceding extinction peak is at the end-Pemian. Tuff layen at the Pennian-Triasic boundary in southem China are "~r-'S\r dated at

249.W1.5 and 250.W 1.6 Ma (Renne et al., 1995) and U-Pb dated at 25 120.4 Ma (Claoué-Long et al.. 199 1 ) that suggest ca. 250 Ma as the boundary age. The estimated interval between the PST and T-J events is 49 Ma only slightly shoner than the double phase length (52 Ma) predicted by the periodic extinction rnodel. The next extinction peak in the Jurassic, originally recognized in the Pliensbachian (Raup and Sepkoski. 1984), is now interpreted as a protracted extinction event that culminated in the Early Toarcian (Little and Benton, 1995). In the

Canadian Cordillera, ammonite biochronologically controlled U-Pb dates bracket the interval of elevated extinction rate from 186I1 (Johannson and McNicoll, in press) and 184.7s0.5 (MacIntyre et al., 1997) (Late

Pliensbachian) to 18 1.4I1.2 Ma (Pal@ et al., 1997) (mid-Toarcian). Therefore the next extinction event was over no more than 20 Ma afier the T-J event, at odds with the predictions of the 26 Ma periodicity.

6.8. CONCLUSIONS

The end-Triassic mass extinction is one of the five most important Phanerozoic mass extinction events.

Deciphering the dynamics of this mass extinction requires an accurate time fiame but available age estimares of the

Triassic-Jurassic system boundary in various time scales are conflicting and imprecise. A new U-Pb zircon age of

20WI Ma from a tuff layer immediately below the Triassic-Jurassic boundary in the Queen CharIone islands, western Canada, is well-constrained by integrated biochronology of radiolarians, conodonts, and ammonites in an apparently continuous, marine succession. The new date, pooled with 12 other U-Pb ages from uppermost Triassic and basal Jurassic rocks, is used to construct a chronogram that yields 20 l* 1 Ma for the age of the Triassic-

Jurassic boundary, several million years younger than suggested by most time scales. The close agreement of isotopic ages from marine (Queen Charlotte Islands) and continental (Newark basin) boundary sections suggests chat extinctions in the marine and terrestrial realms were synchronous. The newly recognized extended duration of the terminal Triassic emphasizes the duality of extinction dynamics: Decline of some shelf-dwelling fossil groups

(ammonoids, conodonts, and bivalves) which were decimated by the mass extinction event may have occurred over a few million years interval, suggesting a protracted environmental crisis whereas the marine microplankton (e.g., radiolarians) and terrestrial vertebrates and plants show an abrupt turnover at the boundary. Recent advances in dating the Petmian-Triassic boundary and the Early Jurassic permit re-evaluation of the spacing of the end-Triassic and neighbouring mass extinction events. The interval between the end-Permian and end-Triassic events is close to two phase lengths as predicted by the 26 Ma periodicity model. However, the Late Pliensbachian-Early Toarcian event followed less than 20 Ma afier the end-Triassic crisis. CHAPTER 7

A U-Pb AND *'~r-~'~rTIME SCALE FOR THE JURASSIC

Geochronologic or time scales express the estimated numerical ages of chronostratigraphic units in millions of years. Jurassic chronostratigraphic units are defined on ammonite biochronology based on the weI1-rsiablished zonal standard ftom northwest Europe. in contrast to the high resolution of biochronology, the lurassic cime scaIe remained less well-calibrated than most other periods. Conflicting and imprecise estirnates of the Jurassic time scales derive from the scarcity of biochronologically weH-constrained isotopic ages and the preponderance of low- temperature K-Ar and Rb-Sr dates of poor accuracy and precision (PAIS, 1995). In the North American

Cordillera, systematic effort was made to generate new calibration points by integrating ammonite biochronoIogy of marine sediments and U-Pb zircon dating of interbedded volcanic or volcaniclastic rocks (Pal@ et ai., 1995).

The result is 18 new calibration points (PiiIfy et al., 1997, and Chapters 3-6). Additional radiometric ages, also useful for time scale calibration, are reported in other recent studies. The revised Jurassic time scale presented here differs fi-om the previous because it (1) is more selective in database compilation by employing high precision U-

Pb and "~r-~~~rages only; (2) uses zona1 IeveI biochronology and attempts to estimate chron boundary ages; (3) rejects scaling based on the assumption of equal duration of biochronologic units and minimizes the use of interpolation.

The proposed thescale is compared with previous ones, of which the most frequently cited scales are abbreviated as follows: NDS (Odin. 1982), DNAG (Palmer, 1983; Kent and Gradstein, l985), EX88 (Haq et al.,

1988), GTS89 (Harland et al., 1990), 0D94 (Harland et al., 1990), and MTS94 (Gradstein et al., 1994; Gradstein et al., 1995). 7.2. METHODS

The revised Jurassic time scale is the result of combining methods that were successfiilly employed in

previous work with new approaches made possibie by recent advances in geochronometry and biochronology.

Typically, previous time scale calibrations have utilized three different aproaches (Odin, 1994): ( 1 ) manual

construction that considers each relevant isotopic date individuatly and weighs them subjectively to arrive at best

boundary estimates (as in OD94); (2) statistically-oriented rnethods that treat each accepted date equally and derive

the boundary estimates mathematically (as in GTS89 and MTS94); and (3) reIiance on a select set of a few dates judged most reliable and determination of the intervening boundaries by interpolation assuming either an equal

duration of biochronologic unit. or a constant spreading rate deduced fiom oceanic magnetic anomalies (cg.,

EXX88). Some combination of the above is more cornmon in recent works (GTS89, MTS94). Manual construction

lacks ngourous rnethodology and reproducibility but offers flexibility- This rnethod would sufice if each boundary

had available isotopic age constraints, a situation not yet attainable for the Jurassic. Statistical rnethods offer the

advantage of handling large numbers of dates efficiently and producing reproducible results. Two methods are

tested and availabfe: the chronogram method (Harland et al., 1990) and the maximum likelihood method

(Agterberg, 1988). They are similar in assuming a random distribution of isotopic ages and their results converge

for densely sampled intervals (Agterberg, 1988)- We prefer to use the semi-rigorous chronogram method as it can accommodate, aAer a slight modification, isotopic ages with asymmetric error that are common among the U-Pb ages. The maximum likelihood method, although elegant, statistically solid and rigorous in error propagation, is too restrictive in requiring Gaussian errors for the source data.

In previous time scales, interpolation arose fiom the sparseness of availabIe isotopic ages. The combination of magnetochronology and biochronology provides a powerful tool that allows the interpolation of boundaries if some magnetochrons are directly dated isotopicalIy and a constant spreading rate is assurned during the formation of oceanic magnetic anomalies (Hailwood, 1989). This assumption appears to be tenable for the Late Jurassic

(Channel1 et al., 1995). As no oceanic crust older than Callovian is preserved, the method is not applicabIe for the

Early and Middle Jurassic. In that interval, most scales use some form of biochronologically-based interpolation, assuming equal duration of chrons or subchrons (Harland et al., 1990; Gradstein et al., 1994). This method is inadequate based on three independent lines of evidence which suggest widely disparate durations for Mesozoic ammonite zones: (1) direct high-resolution isotopic dating of Triassic (Mundil et al., 1996) and Cretaceous

(Obradovich, 1993) ammonite zones; (2) MiIankovitch cyclostratigraphy of Jurassic (Smith, 1990) and Cretaceous (Gale, 1995) ammonite zones; and (3) the width of oceanic magnetic anomalies calibrated to Late Jurassic ammonite zones (Ogg et al., 1991; Ogg and Gutowski, 1996).

The need for interpolation is eliminated if diable isotopic ages are available from each zone or stage. The most promising new development in time sale studies is the use of high-precision U-Pb and "~r-~~~rdating on volcanic flows and volcaniclastic layers from biochronologicaIly well-dated sections (e.g., Obradovich, 1993;

Mundil et al., 1996). The present siudy uses this approach in a systematic effort to generate critical U-Pb ages for the Jurassic (Pal@ et al., 1997, Chapters 3-7). A conservative, all-inclusive database compilation is advocated in GTS89 "... that does not exclude any generally accepted data [and therefore] introduces considerable stability into the time scale" (Harland et al.. 1990).

We note that GTS89 and MTS94 incfude K-Ar ages that were produced as early as 1959, whereas there is an emerging consensus among geochronologists that U-Pb and .'OA~-'~A~ systems are the most reliable and precise geochronometers currently available. We chose to emphasize accuracy over stability, therefore we give priority to

U-Pb and 'OA~-~'A~ ages and omit ages derived from materials with low closure temperature. Such a radical approach was already pursued for the Cretaceous (Obradovich, 1993).

73. CHRONOSTRATIGRAPHIC FRAMEWORK

The Jurassic chronosuatigraphy is traditionally based on the sequence of northwest European ammonite faunas. An apparently well-established, hierarchic scheme of stages, zones, subzones and, for many intervals. horizons has been developed, although none of the Jurassic stage boundaries has been formally defined yet by rneans of Global Boundary Stratotype Section and Point (Page and Melendez 1995; Remane, 1996). Nevertheless, considerable consensus exists regarding the placement of most boundaries.

This study relies heavily on North American data, mainly because the European successions generally lack isotopically datable horizons. It is therefore practical to use a regional ammonite biochronologic scheme, that is readily applicable to constrain the majority of the isotopic dates and correlatable to the northwest European standard. Such North American ammonite biochronologic standards have recently been developed for the

Pliensbachian (Smith et al., 1988), Toarcian (Jakobs et al., 1994), Aalenian (Poulton and Tipper, 1991). and

Bajocian (Hall and Westennann, 1980; HiIlebrandt et al., 1992) stages. Well-documented local zonations also exist for the Hettangian (Tipper and Guex, 1994) and Sinemurian (PhlQ et al., 1994). Because their non-standard units (i.e., infonnally defmed assemblages) have proved to have widespread applicability, herein we mat them equivalent to zones (chrons).

Correlation to the northwest European standard zonation was carefully considered by the respective authon ofNorth Amencan stage zonations and is compiled here in Fig. 7.1. A case study of the chronostratigraphic error

Întroduced by interregional correlation demonstrates it to be no more than one subchron (Piil@ et al., 1997).

Endemism of ammonite faunas increases from the Late Bajocian onward. The Bathonian through Oxfordian faunal succession is increasingly well-understood (Callomon, 1984: Poulton et al., 1994) even though no formal regional zonation has been proposed yet. Precise correlation, however, is hampered by significant differences between North American and European faunas, thetefore only substage level subdivision is applied here. The

Kirnmeridgian and Tithonian is not subdivided here due to their general scarcity of ammonites in North Amenca.

7.4. THE ISOTOPIC AGE DATABASE

Our database of critical isotopic ages is derived from three sources: (1) U-Pb ages from the North Arnerican

Cordillera (either pmduced as part of this project or obtained by other workers and the biochronologic constraints critically reviewed andlor revised by us); (2) U-Pb or "~r-~~~raga culled from the databases of previous time scales (mainly from GTS89 and MTS94); and (3) recently reported U-Pb or ''~r-~~~rages hm outside the

CordiHem that have not been used in time scale calibration before. OnIy ages with adequately documented analytical data are included (Le., dates appearing only in abstracts are not included), with the exception of some recently obtained Cordilleran ages that were made available by our colleagues through personal communications and are currently being prepared for publication.

The isotopic ages were screened for accuracy, precision, and quality of chronostratigraphic consmints.

Accuracy is adequate if reproducibility is demonstrated andor the error assignment is conservative. A few unresolved dates that exhibit loss of radiogenic daughter producs contribute only minimum ages for boundary estimation. Of the U-Pb ages, only datasets incorpotating multiple fraction analyses are considered. The threshold NW EUROPEAN AMMONITE CHRONS Parkinson1 Garantiana Rotundum Subfurcatum Oblaîum

Cassicostatus Laeviuscula Widebayense Concavum Howelll Murchlsonae

Westernanni Yakounensis Levesquel Thouarsense Cassicosta Variabills Planulata Bifrons Falclferum Tenuicostatum Cariottense Kunae Freboldi Davoei 1bex Jamesonl

Plesechioceras? harbledownense Oxynoîum Asterocetas varians Obtusum

Amloceas amouldl

Bucklandi Canadensis Angulata

Llasicus -- Planorbis

Figure 7.1. Early and Middle Jurassic ammonite biochronological units of North America and heir correlation with the northwest European standard. Compiled fiom Tipper and Guex (1994),Pal@ et al. (1994), Smith et al. (1988), Jakobs et al. (1994), Poulton and Tipper (199 1), Hall and Westermann (19801,and Hillebrandt et al. (1992). 114 of acceptable precision is generally *5 Ma (2~51,less precise dates are only exceptionaily included where warranted by the lack of better data. Our maximum allowable ermr is lower han the 6 Ma (2~)average precision in the dataset of MTS94, the most recently produced time scale. Chronostratigraphic constraints are obtained using a variety of methods but preference is given to ammonite biochronology. Assignment of age brackets is based on the tightest possible biochronologic constraints and they may be complernented with reasonable geologic age inferences. Isotopic ages are considered critical to the boundary estimation if they are at ieast bracketed by adjacent stages. Through the use of zona1 biochronology, the average chronostratigraphic precision is significantly improved, although exceptionally a few loosely constrained ages need to be included where better dates are unavai lab le.

Table 7.1 Iists the selected critical ages arranged by stages. AIso included are several Late Triassic ages and one Early Cretaceous age that help constrain the lower and upper boundary of the Jurassic. bringing the total number of critical dates used to 56. A detailed discussion of the isotopic age and chronostratigraphic constraint of each database item is given below, followed by an annotated list of U-Pb and ''~r-~~~rages that were used in

GTS89 anaor MTS94 but are rejected in this study.

7.5. COMMENTS ON ITEMS USED IN THE ISOTOPIC DATABASE

The time scale significance of newly determined items that were generated through this project are discussed in detail in previous chapters. These iclude items 5 (Chapter 6), 9- 1 1 (Chapter 4), 14- 18,22,29.3 1.36.40

(Chapter 5), and 30 (PiiIfy et al., 1997, Chapter 3).

Item I - Guichon Creek batholith

A U-Pb zircon age of 2lW3 Ma was obtained by Mortimer (1990) and interpreted as the crystallization age of the Guichon Creek batholith. See discussion on p. 106 (Chapter 6). K-Ar and RMrminimum ages averaged around 205-209 Ma from the Guichon Creek batholith appear as

NDS 177 (Armstrong in (Odin, 1982). The dates are subsequently used as 205k2.5 Ma (1 sigma) with Norian-

Hetangian brackets in GTS89. MTS94 cites the same data without considering the more precise and accurate U-Pb age used here. Table 7.1

ITEM SHORT NAME AGE ERROR BASE TOP METHOD REFERENCE (Ma) +20 -20 Guichon Creek batholith 210 Mortimer et al,, 1990 Griffith Creek sill 205.8 Thorkelson et al., 1995 Griffith Creek flow 205.8 Thorkelson et al., 1995 Red Mtn.: Biotite porphyry 20 1.8 Greig et al,, 1995 Kunga Island 200 this study North Mîn basait 20 1.7 Hodych and Dunning, 1992 Gettysburg sill 201.3 Dunning and Hodych, 1990 Palisades sill 200.9 Dunning and Hodych, 1990 Puale Bay 1 200.8 this sîudy Puale Bay 2 197.8 this study Puale Bay 3 197.8 this study Cambria Icefield 199 Greig and Gehrels, 1995 Red Mtn.: Goldslide porphyry 197.6 Rhys et al., 1995 Rupert Inlet* 191 this study Ashrnan Ridge I * this study Ashrnan Ridge 2* this study Telkwa ange I this study Telkwa Range 2 this study Joan Lake Thorkelson et al., 1995 Chuclii intrusion Nelson and Bellefontaine, 1996 Atlin Lake (East shore) Mihalynuk and Gabites, unpublished Todagin Mtn. this study Ath Lake (Sloko Island)** Johannson and McNicoll, in press Atlin Lake (Copper Island) Johannson and McNicoll, in press Skinhead Lake Maclntyre et al., 1997 Whitehorse Hart, in press Eskay porphyr~ Macdonald et al., 1992; Childe, 1996 McEwan Creek pluton Evenchick and McNicoll, 1993 Mt, Brock range this study Yakoun River Plilfy et al., 1997 Diagonal Mtn. I this study Julian Lakc Monensen and Lewis, unpublished 33 Treaty Ridge Friedtiiati and Anderson, unpublished

1 Ih Table 7.1 continued

ITEM SHORT NAME AGE ERROR BASE TOP METHOD REFERENCE (Ma) +2~ -20 Eskay rhyolite west U-Pb Childe, 1996 Eskay rhyolite east U-Pb Childe, 1996 Diagonal Mtn. 2 U-Pb this study Gunlock Ar-Ar Kowallis et al,, 1993 Burnaby Island Plutonic Suite U-Pb Anderson and McNicoll, 1995; Anderson and Reichenbach, 1991 Harrison Lake U-Pb Mahoney et al., 1995 McDonell Lake U-Pb this study Copper River U-Pb this study Diagonal Mm. 3* U-Pb this study Chacay Melehué I * U-Pb Odin et al., 1992 Chacay Melehué I U-Pb Odin et al., 1992 Tsatia Mtn. ** U-Pb Ricketts and Parrish, 1992 Josephine ophiolite U-Pb Harper et al., 1994 Rogue Formation* U-Pb Harper et al., 1994; Saleeby, 1984 Hotnarko volcanics U-Pb van der Heyden, 199 1 Tidwell Member Ar-Ar Kowallis et al., in press Brushy Basin Member 1 Ar-Ar Kowallis et al., in press Brushy Basin Member 2 Ar-Ar Kowallis et al., in press Brushy Basin Member 3 Ar-Ar Kowallis et al., in press Brushy Basin Member 4 Ar-Ar Kowallis et al., in press Brushy Basin Member 5 Ar-Ar Kowallis et al., in press Brushy Basin Member 6 Ar-Ar Kowallis et al., in press Grindstone Creek U-Pb Bralower et al., 1990

Table 7.1. Listing of selected critical isotopic ages. Code numbers for constraining stratigraphie units (base and top) are given in Table 7.2. * - minimum age; ** - maxirnum age. Items 2-3 -GriBth Creek volcanics

See p. 105 (Chapter 6).

Items 4 and 13 - Goldslide intrusions, Red Mountain, Item 12 - Cumbria Icefield

See p. 105 (Chapter 6).

Items 6 - 8 -Newark Supergroup basalts

See p. 105 (Chapter 6).

Item 19 -Joan Lake

The U-Pb age of 193.9+1 .Il-4.5Ma was obtained from the Joan Lake area (northwestem British Columbia) by Thorkelson et al. (1995). The sample waç coIlected near the top of a thick weIded rhyolite tuff unit which is disconformably overlain by the fossiliferous Joan Formation that contains an abundant ammonite fauna of Early

Pliensbachian (Whiteavesi Zone) age (Thomson and Smith. 1992). A time gap of undetermined duration is suggested by the erosional surface and the basal conglomerate at the contact between the two formations. The biochronological data therefore provide an upper bracket to the isotopic age: the Cold Fish volcanics at Joan Lake is Whiteavesi Zone (Early Pliensbachian) or older.

Thorkelson et al. (1995) dated three other samples from the Cold Fish Volcanics in the Spatsizi River rnap area. Rhyolite sills and a dyke yielded U-Pb ages of 193,711.9, 194.8+1 .O/-3.0, and 196.6I1.6 Ma, respectively. al1 within error with the Joan Lake rhyolite tuff.

In MTS94, Gradstein et al. (1994, Item 276) use a weighted mean of three of these ages arbitrarily simplified to syrnrnetric erron and interpret it as Early Pliensbachian. Although there are limited occurrences of fossiliferous

Lower Pliensbachian sedirnents interbedded with volcanic rocks elsewhere in the map ara (Thomson et al., 1986), the evidence fiom the Joan Lake section itself doesn't justiQ this restrictive interpretation. Item 20 - Chuchi intrusion

In the Mt. Milligan map area in northem Quesnellia, the locally fossiliferous Chuchi Lake Formation of the

Takla Group is intruded by small igneous bodies. One of them, a monzonite inmision near the BP-Chuchi property.

was dated as 188.512.5 Ma by U-Pb method (Nelson and Bellefontaine, 1996). The crystallization age is defined

by two concordant and overlapping titanite fractions and the lower intercept of a discordia Iine through two zircon

fractions containing small amounts of inherited Proterozoic Pb component. Field observations (e-g.. wet sediment

deformation near the intmsive contacts, predominance of sills, cross-cutting relationships) suggest syn-sedirnentl

(Le., pre-lithification) emplacement of this high-level intrusion (Nelson and Bellefontaine, 1996). Sedirnentary beds

in the area, IikeIy correlative with the sedimentuy rocks intruded by the dated monzonite body (Nelson and

Bellefontaine, 1996), yielded ammonites (Leptuleocerm, Arieticeras, and Amahheu) of the Late Pliensbachian

Kunae Zone (Tipper in Nelson and Bellefontaine, 1996). In one section, however, an Early Pliensbachian (Whiteavesi Zone) ammonite fauna was recovered fiom a thin and stratigraphicaIly lower sedimentary unit (Tipper

in Nelson and Bellefontaine, 1996). A conservative interpretation is to bracket the crystallization age between the Whiteavesi and Kunae zones.

Items 21, 23, and 24 -Atiin Lake

The fossiIiferous Lower Jurassic Laberge Group is well exposed on the shores of Atlin Lake in northwestern

British Coiumbia and contains minor volcanic rocks (Johannson et al., in press). A tuff layer (informaIly assigned ' to the Nordenskiold volcanics) yielded a U-Pb zircon date of 187.5~1Ma (Item 2 1, Mihalynuk and Gabites. pers. comm 1996). Ammonite coIlections stratigraphically below (Tropidoceras acmeon) and above (Metuderoceras cf. talkeetnaense, Duburiceras cf. silviesi, Acanthopleuroceras cf. thomsoni) the tuff indicate the Whiteavesi Zone

(Early Pliensbachian) (Johannson, 1994). On Copper [stand, another crystal tuff Iayer stratigraphically higher was

U-Pb dated at 186.M Ma (Item 24, Johannson and McNicoll, in press). This sarnple is also tightly constrained by ammonite biochronotogy. Diagnostic taxa, indicative of the Late Pliensbachian Kunae Zone, include Reynesoceras colubrifrme and Arieticeras from below and Leptuieocem, Arieticeras, and Protogrammoceras from above the tuff (Johannson, 1994). Another U-Pb age of 186.64-0.51-1 .O Ma was obtained hrn a granitoid boufder within a conglomerate on

Sloko Island (Item 23, Johannson and McNicoll, in press). The conglomerate is also of Kunae Zone age based on

the ammonite fauna (Reynesoceras. Leptaleuceras, Protogramrnoceras. Fuciniceras. Arieticeras) recovered from

adjacent fmer-grained sedirnentary rocks (Johannson, 1994). Assuming geologically rapid uplifi and erosion, the

U-Pb date provides a usefil maximum age for the Kunae Zone (Johannson et al., in press).

Item 2.5 -Skinhead Lake

A U-Pb age of 184.710.5 Ma was obtained by M. Villeneuve (in Machtyre et al., 1997) fiorn zircon in a

rhyolite tuff near Skinhead Lake (west of Babine Lake, northwestern British Columbia). Immediately above the dated unit, frorn a lens of fossiliferous sandstone within the volcaniclastic package we collected a small ammonite

fauna consisting of Arieticeras and Fanninoceras? indicative of the Late Pliensbachian Kunae Zone.

Item 26 - Whitehorse

A dacite tuff assigned to the Nordenskiold volcanics is U-Pb dated at 184.1-1-4.21-1.6 Ma near Mitehone. Yukon (Hart, in press). It is interbedded in a section of fossiliferous Laberge Group sediments. Arielkerus occurs a few rneters below the tuff whereas Arieticeras and Amallheus cf. stokesi was collected upsection indicating the

presence of the Upper Pliensbachian Kunae Zone. The section is discussed in detail along with illustration of the ammonite fauna in Pd@ and Hart (1995).

Item 2 7 - Eskay porphyry

A sill-like feldspar porphyry intrusion in near the Eskay Creek gold mine in the Iskut River area

(northwestem British Columbia) was dated by Macdonald et al. (1992) who reported a U-Pb zircon age of 18612 Ma. It was recalculated and reinterpreted as 184+5/-1 Ma (Childe, 1996). The porphyry intrudes fossiliferous mudstone that yielded Lioceratoides propinquum, Protogrammoceras cf. kurrianum and other hildoceratids characteristic to the Pliensbachian/Toarcian transition (Nadaraju, 1993). The amrnonoids range frorn the toprnost Pliensbachian Carlottense Zone to the basal Toarcian Kanense Zone, although the absence of DactyIiocerar favors an assignment to the Carlottense Zone, for which the age of the intrusion is regarded as a minimum age.

Item 28 - McEwan Creek pluton

The quartz monzonite McEwan Creek pluton intrudes Lower Jurassic and older strata in the northwest part

of the Spatsizi River map area (northwest British Columbia) and was dated by U-Pb method (Evenchick and

McNicoll, 1993). The reported age of 183.5*0.5 Ma is based on the '06~b/38~age of the more precise of two concordant and overlapping zircon fractions. Two titanite fractions also analyzed from the same sample are

perfectly concordant and overlapping at a 'M~b/s8~age of I83.Oi0.5 Ma. Although the zircon and titanite aga are

within error and the somewhat younger age of titanite may be explained by its lower closure temperature, we take a more consewative approach in assigning a crystallization age of 18320.7 Ma based on the weighted mean

'06~b/238~age of the three rnost precise and concordant fractions.

The youngest smtigraphic unit inuuded by the pluton is the Mount Brock volcanics (Evenchick and

McNicoll, 1993). the youngest volcanic member of the Hazelton Group in the Spatsizi area (Thorkelson et al.,

1995). Critical fossil localities providing conscraints on the age of the Mount Brock volcanics are found in

intercalated marine sedimentary rocks (Read and Psutka, 1990). Early and middle Toarcian ammonites are reported

in al1 previous studies (Read and Psutka, 1990; Evenchick and McNicoll, 1993; Thorkelson et al., 1995). Our new fossil collections fiorn the area West of Mount Brock indicate that volcanisrn started in latest PIiensbachian tirne

(see also Thomson, 1986), and the evidence for middle Toarcian needs revision. Dactyliaceras, commonly occuring in (but not restricted to) the lower Toarcian was found stratigraphically above the only collection that had suggested a middle Toarcian age in previous work, based on the identification of Polyplectirs sp. As the ranges of

Polyplectus and Dactyfiocerus are mutually exclusive, it is possible that the specimen in question actually represents some other, morphologically similar harpoceratid ammonite. In conclusion, we favor an interpretation that Mount Brock volcanism (and emplacement of the perhaps CO-magmaticMcEwan Creek pluton) is not younger than early Toarcian. Item 32 -Julian Lake dacite

A thick pile of felsic volcanics with locally interbedded manne sedimentary rocks occur in the Salmon River

Formation near Julian Lake (Iskut River ma, nonhwestern British Columbia). A dacite flow near its base yielded a

U-Pb zircon age of 178*1 Ma (P. Lewis and J. Mortensen, personal communication 1996). Hyaloclastite observed at the flow top points to submarine emplacement, in turn suggesting that the sedirnentary rocks are not significantly younger than the dated flow that they directly overlie. These volcanic sandstones are locally fossiliferous and yielded a diverse ammonoid fauna (Yakouniasilvae, Pleydellia cf. maudenîis, P. cf. crussiornuta. Phymatoceras sp,) clearly indicating the uppemost Toarcian Yakounensis Zone. Several hundreds of metres upsection, another dacite flow yielded an age of 172.31 1 .O Ma (P. Lewis and J.

Mortensen, personal communication 1996). No identifiable fossils have been found in the upper part of the section, therefore the latter date can only be used with a latest Toarcian lower bracket and a Bajocian upper bracket. inferred fiom regional geology as the age of cessation of Salmon River felsic volcanism.

Item 33 - Treaty Ridge

The Treaty Ridge section is an important reference section for the Hazelton and overlying Bowser Lake groups in the Iskut River map area (northwestem British Columbia) (Lewis et al.. 1993). Four separate sarnples frorn the upper felsic unit of the Salmon River Formation have been recently dated by the U-Pb method. Cornplex isotopic systematics that suggest presence of Late Triassic xenocrystic zircon andlor Pb-loss hampered interpretation of data for one sarnple (McNicoll and Anderson, personal communication 1995). Only one sarnple yielded concordant analyses giving an interpreted age of 177.3k0.8 Ma (Friedman and Anderson, personal communication 1996). A weighted mean '"~b/'~~bage of NO discordant but apparently inheritance-free fractions gives a 178*12 Ma age for another sarnple (Friedman and Anderson, personal communication 1996).

Fossiliferous sediments yielded ammonites both below and above the felsic volcanic unit (Lewis et al., 1993;

Nadaraju, 1993; Jakobs and Piil@, 1994). The Upper Aalenian Howelli Zone is documented by the presence of

Erycitoides cf. howelli, Pseudolioceras cf. whiteavesi, Tmeroceras cf. kirki, and Leiocerar? sp. in the underlying mudstone. Sonninia? sp., S~ephanocerassp. and Zemisrephanus sp., collected from siltstone overlying the dated volcanic unit, açsign an Early Bajocian upper age limit to the volcanism. A flow-banded rhyolite unit that underlies the locally mineralized argillite at the Eskay Creek gold mine

(hkut River area, northwestem British Columbia) was sampled for U-Pb zircon dating on the east Iimb of the

Eskay anticline. Childe (1996) obtained an age of 174+U-1 Ma. Erycitoides cf. howelli, the index ammonite of the

late Aalenian Howelli Zone, was collected from above the rhyolite on the same limb of the anticline, less than 3 km

away from the U-Pb sarnpling site. The rhyolite is therefore Upper Aalenian or older. No precise lower bracket cm

be assigned. The latest Pliensbachian (or possibly earliest Toarcian) ammonites listed for the Eskay porphyry (item

27) are the youngest fauna fiom underlying uni& in the area. Regionally, felsic volcanism correlative to the Eskay rhyolite is not known to begin prior to the Iatest Toarcian (see Item 32).

Another sample fiom the sarne rhyolite unit on the West limb of the Eskay anticline yielded an age of 175*2

Ma (Childe, 1996). No ammonites have been found near this sample site but radiolarians identified from drill core

obtained hmthe overlying argiliite are dated as Aalenian to possibly early Bajocian (Nadaraju, 1993). The local

correlation of the rhyolite and argillite units is well-documented by detailed geological mapping and further

supported by the statistically indistinguishable U-Pb dates (Childe, 1996) and concordant ammonoid and

radiolarian ages. Therefore the sarne stratigraphie constraints are applied to both isotopic dates.

Item 3 7 - GunIock

The Middle Jurassic Carmel Formation in Utah contains numerous ash beds several of which have been recently dated using the 'O~r-j~~rsingle crystal laser probe method (Kowallis et al., 1996). One of them is published in detail and thus included in our database: a 166.310.8 Ma sanidine age from the upper part of the

Cmel Formation near Gunlock. (Kowallis et al., 1993). Characteristic bivalves allow correlation of the Carmel

Fm. with the more fossiliferous Sliderock and Rich members of the Twin Creek Limestone that yielded ammonites of late Early to Late Bajocian age (Imlay, 1967; Imlay, 1980).

Younger K-Ar (16 1.i%1.8 Ma (la))and Rb-Sr (1626.5 Ma (10)) ages from the Carmel Formation were reported as NDS 102a and NDS 102b and also used in GTS89. Item 38 -Burnaby Island Plutonic Suite

Plutonic rocks assigned to the Burnaby Island Plutonic Suite (BIPS) yielded several U-Pb dates from the

Queen Charlotte islandç. They range in age between 172 and 2 158 Ma (Anderson and Reichenbach, 1991 ). We use

here the four dates (Poole Point: 168+4/-1 Ma; Shields Bay: 168*4 Ma; Rennell Sound: 164+4/-2 Ma (Anderson

and Reichenbach, 1991); and Cumshewa Head: 167I2 Ma (Anderson and McNicoll, 1995)) which form a tight

cluster in the older part of the age range of the BIPS. The youngest country rocks cross-cut by BIPS intrusions are

Bajocian volcano-sedimentary strata of the Yakoun Group. Ammonoid biochronology of the Yakoun Group

elsewhere in the Queen Charlotte Islands suggests an age range of Widebayense to Oblatum zones (Hall and

Westermann. 1980; Poulton et al., 199 I b). It is possible that the Yakoun Group volcanics represent extrusive equivalents of BIPS plutons. In the Future, isotopic dating of the Yakoun Group volcanics holds promise for contributing to bener constraints on the Bajocian tirne scale. At present, we use a pooled age of 168I4 Ma for early

BIPS plutons as a minimum constraint for the undivided Bajocian stage. The late Bathonian and younger age of the

Moresby Group (Poulton et al., 199 lb) which disconformably overlies the Yakoun Group and is not known to be cross-cut by BIPS is used 10 provide a minimum age for the BIPS, although the oldest strata found directly deposited on B[PS rocks are Early Cretaceous in age (Anderson and Reichenbach, 1991).

Item 39 - Harrison Lake

Mahoney et al. (1995) obtained a U-Pb zircon age of 166.0I0.4 Ma from a rhyolite near the top of the

Weaver Lake Member (Harrison Lake Formation) exposed on Echo Island in Harrison Lake. southwestern British

Columbia. No age diagnostic fossils have been recovered from sedimentary interbeds within the Weaver Lake

Member (Arthur et al., 1993). The youngest fossil known from the underlying Francis Lake Member is

Erycitoides? sp. and Tmetoceras scissum of Late Aalenian age. Both Arthur et al. (1993) and Mahoney et al. (1995) speculate that the Weaver Lake Member is as young as Early Bajocian or possibly even younger. The Harrison

Lake Formation is unconfomably overlain by the Mysterious Creek Formation which yielded a rich Early

Callovian ammonite fauna (Arthur et al., 1993). The unconformity separating the units represents a regional deformation event and strongly suggests that the Weaver Lake Member is significantly older than Early Callovian. The isotopic age is further supponed by another two. nearly identical U-Pb ages From related rocks in the

area. A rhyolite dyke fmm strata correlative to the upper part of the Weaver Lake Member near the Seneca mineral

deposit yielded an age of 165.9+6.4/4.3 Ma (McKinley, 1996) and a comagmatic quartz-feldspar porphyry stock

(Hernlock Valley stock) was dated at i66.0i0.4 Ma (Mahoney et al., 1995).

Items 43-44 - Chacay Melehué

Two mff layers from the Chacay Melehué section in the Neuquén Basin, Argentina, were dated at L60.510.3

Ma and 161 .&OS Ma by zircon U-Pb method (Odin et al., 1992). Both dates are lower intercept ages, the first one

is corroborated by one concordant fraction. The section has a well-documented ammonite fauna that is dominated

by endemic South American forms but ailows correlation with the European standard. The lower sarnple is very

near to the boundary of the regional Steinmanni and Vergarensis zones that is equated to the Bathonian/Callovian

boundary (Riccardi et al., 199 1). The upper sample was collected near the boundary between the Bodenbenderi and Proximum zones that approximately corresponds to the middle of the Boreal Calloviense Zone or the

Mediterranean Gracilis Zone in the upper part of the Lower Callovian (Riccardi et al., 199 1).

45 - Tsatia Mountain

Lenses of boulder conglomerate within the Bowser Lake Group in the Bowser Basin (northwestem British

Columbia) locally contain dacite clasts, one of which was dated by LI-Pb zircon method. Ricketts and Parrish

(1992) obtained a precise age of 160.7I0.7 Ma based on analysis of three concordant fractions. The dated clast was

recovered From a biostratigraphically well-constrained section on Tsatia Mountain. An Early Callovian Cadoceras

fauna was found below the conglomerate whereas overlying deposits yielded Sfenocadocerusof Middle Calovjan age and still younger assemblages higher upsection (Poulton et al., 1991a; Poulton et al., 1994). The dacite clast therefore cannot be younger than Middle Callovian. Item 16 -Josephine ophiolite

U-Pb ages fiom plagiogranites in the Josephine ophiolite have been used in previous time scales. GTS89

(Item HS 1) uses an age of 157*2 Ma as Oxfordian or older. The original data appear in Saleeby ( 1982) as analyses of two different samples (one single-fraction and one based on two fractions of unabraded zircons). A later revison of one of these dates to 1621 Ma (Saleeby, 1987) is not considered in GTS89 or MTS94 (Item 253, Iabelled as

HS2). A full documentation is now available (Harper et al.. 1994), showing that the interpreted age is the '06~bP%J- age (with 1 sigma error) ofthe subsequently analyzed, concordant, abraded fraction. We assign a more conservative age estirnate to samples when no duplicate concordant fraction is available and there is evidence of

Pb-loss. A weighted mean 207~b/Z06~bage from the two fractions (sample A88Z) combined with the 2a~b/3?Jage of the concordant one as a minimum age gives 162+7/-2 Ma as the crystallization age. The choice of a larger positive error is also warranted when an independent "~r-~~~rplateau age of 165*3 Ma ( l sigma) obtained from a high-level gabbro intrusion 1 km from the U-Pb dated plagiogranite (Harper et al., 1994) is considered.

Furthemore, Wright and Wyld (1986) reported a U-Pb age of 164i1 Ma from the Devils Elbow ophiolite remnant thought to be correlative to the Josephine ophiolite.

The Josephine ophiolite is overlain by the Galice Formation which yielded bivalves (Buchia concentrica).

. perisphinctid ammonites, and radiolarians (e.g., Mirifisus) indicating an Oxfordian age (Imlay, 1980; Pessagno and

Blome, 1990).

Item 4 7 - Rogue Formation f uff breccia

A U-Pb zircon age of 157* 1.5 Ma (1 a) was obtained f?om tuff breccia in the Rogue Formation (KIamath

Mountains, northem California) (Saleeby, 1984; Harper et ai., 1994). Analytical data have not been published.

Considering the scarcity of Late Jurassic data, we include this date with a qualifier as a minimum age. It is warranted as Pb-loss is documented as a common phenomenon affecting the U-Pb systematics of Jurassic zircons from the area (Saleeby, 1987). The Rogue Formation is overlain by the spanely fossiliferous Galice Formation which yielded bivalves (Buchia concentrica) and radiolarians (e.g., Mirr/iliur) indicating an Oxfordian age (Imlay.

1980; Pessagno and Blome, IWO). Item 48 - Hotnmko volcanics

Van der Heyden (199 1) obtained a preliminary U-Pb age of 154.4*1.2 Ma on a sample from the Hotnarko volcanics in the eastem Coast Belt in west-central British Columbia. Volcaniclastic rocks near the base of the succession yielded Andifrigoniaaff. plummemis (identification by T. Poutton) which ranges from the Callovian through the rniddIe Oxfordian. It thus provides a lower stratigraphic bracket for the dated volcanic rocks whereas geologic evidence suggested that the entire succession rnay have been deposited within a brief volcanic episode

(van der Heyden, 199I ).

Item 49 - Tidwell Member (Morrison Formation)

Three sanidine 'O~r-j~~rsingle-crystal laser fusion ages (1 54.8k 1.2, 154.8k 1.1, and 154.822.8 Ma) and one plateau age (154.911.0 Ma) are reported by (Kowallis et al., in press) hman ash bed in the Tidwell Member

(lower part of the Momson Formation) sampled in hvo sections of Utah. Ali analyses determine the age of the same ahbed and are remarkably concordant therefore we use the most precise of them in our database. The youngest cardioceratid ammonites hmthe underlying Redwater Shale (Stump Formation) are Middle Oxfordian

(Irnlay, 1980; Callomon, 1984). The base of the Morrison Formation marks a regional unconfomity. Ostracods and charophytes (Schudack et al., in press) as well as palynomorphs (Litwin et al., in press) suggest that the lower pan of the Morrison Formation, including the Tidwell Member, is Kirnmeridgian (likely Early Kimmeridgian) in age.

Magnetosmtigraphic studies also reveal correlation between the reversa1 sequence in the Morrison Formation and the Kimmeridgian oceanic magnetic anomaly pattern (Steiner et al., 1994).

Items 50-55 - Brushy Basin Member (Morrison Formation)

Six m~r-39~rlaser fusion ages for plagioclase From volcanic ash layers in the Brushy Basin Member

(Colorado Plateau) range between 153 and 148 Ma (Kowallis et al., 199 I). This dataset is now superseded by seven more precise 'OA~-'~A~ sanidine laser fusion ages from different ash layers in five section of the upper part of the

Brushy Basin Member (Kowallis et al., in press). Six of thern (150.3rt0.5, 150.7I1 .O, 149.31 1.1, 149.31 1.0,

149.0I0.8, and 148.1I1.0 Ma) are consistent with their relative stratigraphic position within the unit and are preferred. A11 of the dated ash layers occur in the upper part of the unit, above the level rnarking a characteristic change in clay mineralogy (Kowallis et al., in press).

Correlation based on ostracods and charophytes suggests the isotopically dated interval to be Kimmeridgian in age in its lower part, possibly ranging to the Tithonian near the top (Schudack et al., in press). Palynological results are [ess concIusive but corroborate the placement of these strata to the Kimmeridgian and Tithonian (tinvin et al., in press). We infer an age restricted to the Kimmeridgian for the lowest two dated ashes.

Item 56 - Grindstone Creek tuf

A precise and well-docurnented U-Pb zircon age of t 37.1 + t .6/-û.6 Ma was obtained fiorn two altered crystal tuff layers within a fossiliferous Upper Jurassic to Lower Cretaceous mudstone succession (Great Valley

Sequence) in California by Bralower et al. (1990). The dated levels are assigned to Buchia mciroidm and B. pacijica bivalve zones and the Assiperra infracrefacea (NK-2A) nannofossil subzone, both indicating a Late

Berriasian age. Correlation with DSDP sites containing integrated magnetostratigraphic and nannofossil records pemited their assignmnent ro the CM16 or CM16n magnetochrons (Bralower et al., 1990). Channel1 et al. (19953 revised the ranges of critical nannofossil taxa and suggst that the dated level can represent the CM 16-CM I5 magnetochrons, still within the Upper Berriasian. This date is used in MTS94 as Item 232.

7.6. COMMENTS ON ITEMS USED IN EARLIER TIME SCALES BUT REJECTED

IN THIS STUDY

Item HL BI (GTS89) and Item 269 (MTS94)

The quoted 'OA~-'~A~ age is 185.0Tt3.0 Ma (1 sigma) with Pliensbachian-Toarcian stage brackets. The original source (Hess et al., 1987) reports 1 1 ages, ranging hm190 to 180 Ma, from a 2000 km' area in the northem Caucasus underlain predorninantly by thick volcanosedimentary sequences. No detailed stratigraphy was reported but sedimentary rocks generally associated with the volcanics were said to contain rare Pliensbachian bivalves and brachiopods as well as Toarcian ammonites. We conclude that the coarse stratigraphie resolution of this dataset does not warrant its inclusion in the time scde calibration and the simple averaging of the 1 I different ages is inadequate for inclusion in Our database.

Item 251 (MTS94)

MTS94 quotes an age of 155.3i3.4 Ma as an "~r-'~Ar age for Oxfordian oceanic cmst recovered from

DSDP Site 765. However, it is clear from the originaI source (Ludden, 1992) that the age in question was obtained using the K-Ar method, therefore we exclude it from Our database. See also remarks in section 7.10.

Item 258 (MTS94)

Gradstein et al. (1994) quote an "~r-~~~rage of 166.8I4.5 Ma obtained by PringIe (1992) as an age for

Callovian oceanic cmst recovered fiom DSDP Site 801. In Pringle's (1992) account, this is the age of tholeiitic

MORB basalt recovered fiom the base of the drillhote. lt is overlain by alkaline off-ridge basalt that yielded an 40 ~r-~'Arage of 157.4i0.5 Ma. Magnetostratigraphic evidence suggests that the site was drilled into the Jurassic

Quiet Zone some 450 km away fiom anomaly M37, the oldest known preserved oceanic magnetic lineation thought to be Callovian in age. At DSDP Site 801, original radiolarian biostratigraphy from the oldest overlying sedimentary rocks indicated a latest Bathonian to earliest Callovian age (Matsuoka, 1992). This interpretation was challenged by Pessagno (1996) who considered the fauna as middie Oxfordian and argued that the younger isotopic age supports this assignment. Considering the controversy that surrounds this dataset, we chose not to include it in the present compilation.

NDSI 84 (revised in GTS89) and Item 2 72 (MTS91)

Both GTS89 and MTS94 quoted an ''~r-~~~rage of 197.612.5 Ma from the Toodoggone volcanics in

British Columbia, based on an undocumented revision of NDSl84 (Armstrong in Odin, 1982). The stratigraphie age given is Sinemurian to Toarcian but the date is rejected on the bais of a lack of analytical documentation.

Clark and Williams-Jones (1991) recently obtained a hlly documented 'OAT-'~AI- age of 1%. 111.6 fiom the

Toodoggone volcanics but there is no fossil data available to bracket the age of these mainly subaerial volcanic

129 rocks. Although it can be reasonably correlated with the early phase of Hazelton Group volcanism recorded elsewhere (e.g., Telkwa Formation, Cold Fish volcanics), it is unuseable for time scale calibration.

Item SCHI (GTS89) and Item 245 (and 217?) (HS94)

GTS89 used a U-Pb age of 153+1 Ma in the Kimmeridgian. MTS94 revised the emr to *3 Ma and lists apparently the same age twice, the second tirne with Oxfordian-Kirnmeridgian brackets. In fact this is not a U-Pb age itseIf but rather a "conservative estimate of the age of the Nevadan deformation" (Schweikert et al., 1984). It is based on severai, poorly docurnented or single-fiaction early-determined U-Pb ages pooled together with K-Ar dates. These items are rejccted as a composite of questionable veracity and validity to time scale calibration.

Item HS2 (GTS889) and Item 234 (labelled as HSI in MT'S94

GTS89 quoted a &Pb age of I50.5=k2 Ma as Kimmeridgian or younger. Evidently the quoted date derived fiom a combination of two single-fraction U-Pb ages of unabraded zircons (Saleeby et a[., 1982; Harper, 1984): a

15 1*3 Ma age obtained from a dyke intruding the Josephine ophiolite and a second 15WMa age from a sill intruding the overlying Galice Formation. îhese ages are rejected as suspect because Pb-loss is documented from several other samples frorn the same dataset in subsequent work on abraded zircons (Saleeby, 1987; Harper et al..

1994).

Item HMP2 (GTS9) and Item 239 (MTS94)

The Tithonian or older U-Pb age of 152.5I2 Ma used in GTS89 and MTS94 presumably represents a mean of '06~b/238~and '07~blUS~ages reponed by Hopson (1981). This date doesn't satisfi our criteria for inclusion in the database as it is based on the analysis of a single fraction. Item HMP3 (GTS9) and Item 240 (MTS94)

The Tithonian or older O-Pb age of 15412 Ma used in GTS89 and MTS94 is apparently the "'P~/"*uage of a slightly revenely discordant fraction reported by Hopson (1 98 1). We reject this date as being based on the analysis of a single fraction where reverse discordance likely indicates analytical problems.

Ikm HMPl (GTS89) and Item 255 (MTS94)

The quoted Oxfordian or older U-Pb age is 16UMa, apparently the younger 'W~b/a8~age of the two unabraded fractions analyzed by Hopson (198 1). This fraction is slightly reversely discordant and the other fraction - yielded a '"P~J~~P~age of 1 84 * 10 Ma which is outside the error of the '"P~/?J age of l65*2 Ma. This discrepancy of apparent ages points to complexities in the &Pb systematics. This date is rejected as the true crystaIlization age cannot be unambiguously resoIved fiom the published analyses.

Item HMPl (GTS89) and Item 254 (MT.94)

'The quoted Oxfordian or older U-Pb age is 16 152 Ma based on data of Hopson (198 1). Of the two analyzed unabraded fractions, one yielded a significantly older apparent "'~b/'~~bage outside the error of the 'Ob~b/?J age. We interpret this as an indication of Pb-loss, inheritance, or both. On the same grounds as in the previous item. we reject this date as the tme crystallization age cannot be unambiguously resolved from the published analyses.

7.7. DIRECT DATING OF STRATIGRAPHIC BOUNDARIES

Ideally, the age of a biochronologically-defined boundary is determined by isotopic dating of a volcanogenic layer situated at or in the immediate vicinity of the boundary. This situation occurs rarely in the Jurassic. The

Triassic-lurassic boundary is directly dated in eastem North America and the Queen Charlotte Islands (Chapter 6).

In the Newark Basin. U-Pb ages of 200.9+1 .O Ma (Item 8) and 20 1.3* 1.O Ma (Item 7) (Dunning and Hodych,

1990) were obtained on sills that are thought to be feeders to the Orange Mountain basalt, the lowermost extmsive rock which lies irnmediately above the Triassic-lurassic boundary defined by palynology and vertebrate biostratigraphy (Olsen et al., 1987; FowelI and Olsen, 1993). From the Fundy Basin, a U-Pb zircon age of

201.7+1.4/-1.1 Ma (1992) is reported From the North Mountain basalt whose base lies about 20 m above the

Triassic-Jurassic boundary. A tuff layer immediately below the conodont- and radiolarian-defined Triassic- Jurassic boundary in the Queen Charlotte Islands yielded a U-Pb age of 20Wl Ma (Item 5) (PAIS. et al., in review).

These four dates are reconciled to a 200.910.6 Ma boundary estimate, that is rounded to 20 1*1 Ma to de-

emphasize apparent precision in the presence of slightly confl icting ages (see Chapter 6).

A volcanic ash layer directly above the base of the Middle Toarcian Crassicosta Zone. a regional standard

ammonite zone for North Amenca, is dated by U-Pb method at I8 1,4il.2 Ma in the Queen Charlotte Islands (Pilm et al., 1997).

A third directly dated level is the boundary of the South American Steinmanni and Vergarensis chrons that is

equated to the Bathonian-Callovian boundary (Riccardi et al., 1991). A tuff layer located at this boundary in the

Chacay Melehué section (Neuquen Basin, Argentina) yielded a U-Pb zircon date of 16 1 .&O5 Ma (Odin et al., 1992).

7.8. CHRONOGRAM ESTIMATION OF BOUNDARIES

In the absence of direct isotopic dating, the ages of a stratigraphic boundary can be estimated using dates

from adjacent units. The chronogram method, described in derail in GTSS9 (Harland et al., 1990), calculates the E error function value for trial ages in a time window scanned for possible boundary ages. Taking into account the relevant isotopic dates, their stratigraphic position below or above the boundary in question, and their plus and minus errors, it provides a semi-rigorous measure of compatibility of the data with the trial ages. Following

GTS89, the best chronogram estimate is defined by the minimum value of E or the mean of a range where E = 0. whereas the endpoints of the error range around the best estimates are taken where E = Emi,,+I.We use here a slightly modified formula that accommodates asymmemc errors, which are common among interpreted U-Pb ages:

E = L(Y~Q~/S'~~*+ C(O~-&)~/SO~~ where t, is the trial age, Yiare the isotopic ages stratigraphicallyyounger than the boundary for which Yi> te. Seyi are their plus 2a error, Oi are the isotopic ages stratigraphically older than the boundary for which Oi < t,, Ski are their minus 2a error.

We calculated chronograrns for each chron boundary in the Hettangian through Bajocian and each substage boundary in the Bathonian through Callovian. As not al1 isotopic dates have zona1 or substage resolution

132 constraints (not even attempted for the Oxfordian through Tithonian), stage boundary chronograms were aIso calculated (Table 72).These rnay be different fiom the chronograrn of the earliest chron of the stage if significant dates lack zona1 constraints (e.g. the Hettangian). The chronogram method assumes a randorn distribution of dates within the stratigraphic intervals whose boundaries are sought. Therefore the directly determined boundary ages tisted above are omitted frorn the calculation of those boundaries and accepted if they overlap with the error range of the respective chronograrn.

If a unit lacks critical dates confined to if its chronogram tends to converge to the next younger unit. Fig. 7.2 thus shows only those meaningfül chronograms that have no identical counterpart for a younger stratigraphic boundary.

The assumption of randorn distribution may prove to be unfounded for stage chronograms if dates are crowded in some parts and are missing fiorn other parts of the unit. This is demonstrated for the Sinemurian.

Pliensbachian, Aalenian, Bajocian, and Bathonian and its consequences are discussed next. 1-Code initial Maximum Minimum +Errer -Errer Error STAGE boundary (20) (20) range Psiloceras/Euphyllites Franziceras Pseudoaetomoceras Canadensis Coroniceras/Amouldi Varians Harbledownense Tetraspidoceras Imlayi Whiteavesi Freboldi Kunae Carlottense Kanense Planulata Crassicosta Hillebrandti Yakounensis Westermanni~Scissum Howelli Widebayense Crassicostatus Kirschneri Oblaturn Rotundurn Epizigzagiceras MiddleLate Bathonian Early Callovian Middle Callovian Late Callovian HETTANGIAN SMEMURIAN PLIENSBACHIAN TOARCIAN AALENIAN BAJOCIAN BATHONIAN CALLOVIAN OXFORDIAN KIMMERIDGIAN TITHONIAN BERRIASIAN

Table 7.2. Chronogram estimates of the initial boundary of chrons, substages and stages. Chrons with poor stratigraphic record (Euphyllites, Coroniceras, and Westernanni) were combined with adjacent chrons. Bold face designates good quality chronograms that are different from the next younger stratigraphic boundary and the error range is less than 5 Ma.

7.9. ADJUSTED STAGE BOUNDARY ESTIMATES

Stage boundaries are of particular interest. therefore their age estimates are considered individually.

Chronogram estimates may be biased if no sufficientdata exist in the vicinity of the boundary. This is indicated if

the chronogram of the earliest chron is identical to any or both adjacent chrons. While chronogram maxima and minima are retained as valid, reasonable adjustments in the best estimates are made with respect to the cntical data.

As discussed above, iwo stage boundaries are dated directly. The base of the Hettangian fixed at 200.9=0.6

Ma based on the four direct ages is in good agreement with the chronogram age derived frorn the other pertinent dates. However, the base of the CalIoviari is pegged by a single date (16 1.&OS Ma, Item 43) that is consistent with a slightly younger date (hem 44) produced by the same authors (Odin et al., 1992) from the same section but is in conflict with another relatively precise age (Item 40) From the Bathonian. As a result, the chronogram minimum takes an unusually high value (>21) and in this case we choose to define the error range where E = 2Emi,. Such a chronoam age of IS9.ï*l. l Ma overlaps with the directly detennined boundary age and is our preferred estirnate.

In the Early Jurassic, the chronograrn age of the Pliensbachian-Toarcian boundary (1 83.6+ l.6/- 1.1 Ma) is tightly controlled as there is a series of good quality chronograms for the neighbouring chrons. The Hettan,'-tan-

Sinemurian and the Sinemurian-Pliensbachian boundaries are less well constrained. The Canadensis Zone, Iikely to span Hettangian-Sinemurian the boundary, is here arbitrarily classified as Sinemurian but this does not affect the chronogm. Early and early Late Sinemurian ages are al1 minimum ages causing an asymrnetric chronogram. The chronogram estimate of 195.8+2.4/-3.0 is adjusted by shifiing the best estimate up to 197.0+ 1-2/42Ma. Such re- apportioning is also supported by observed sediment thicknesses in many Lower lurassic sections.

The chronograrn of the Sinernurian-Pliensbachian is identical to the initial Whiteavesi chron boundary and is not well constrained. On the other hand, the chronogram of the Late Sinemunan Harbledownense chron only differs from the initial Sinemunan boundary by a lower minimum. No data exist From the Tetraspidoceras and

Imlayi chrons adjacent to the Sinemurian-Pliensbachian boundary. A more balanced allocation of time requires downward adjustment of the Harbledownense chron and upward adjustment of the initial Pliensbachian. A possible less conservative interpretation of Item 17, the main control on the Late Sinemurian, would aIso suggest a younger age (192-194 Ma) for the Harbledownense chron. Therefore we pick 192 Ma (with errors adjusted to +3.8/-5.2 Ma) for the best estimate of the age of the Sinemunan-Pliensbachian boundary. In the Middle lurassic, the tight Toarcian-Aalenian boundary chronogram is identical to the middle-late

Aalenian boundary chronogram calling for an upward shift in age. We propose a boundary age of i 78.0 Ma (with errors adjusted to +1 .O/-1.5 Ma).

The Aalenian-Bajocian and Bajocian-Bathonian boundary chronograms are less tightly constrained and are also controlled by dates from the younger half of the stages. Their best estimates and maximum ages are only marginally older than those of the Laie Bajocian Rotundum Zone and the Middle-Late Bathonian, respectively.

Hence an upward adjument of boundary ages is likely to produce more realistic allocation. We propose 174 Ma for the initial Bajocian boundary, corresponding to the oldest trial age with an error function value of zero. The duration of the late Aalenian Howelli Chron thus appears to be 3.6 Ma, conspicuously longer than any other chron.

There are three controlling dates from the Howelli Zone (items 33-35) and none from the adjacent zones. therefore the unusually long chron duration rnay be an artifact of a siight inaccuracy in some of the isotopic dates. The preferred estimate for the Bajocian-Bathonian boundary is 166 Ma.

7.10. LATEST JURASSIC STAGE BOUNDARY ESTIMATES THROUGH

MAGNETOCHRONOLOGIC LNTERPOLATION

The lack of isotopic ages from definitively dated Tithonian rocks renders chronogram estimation of the

Kimmeridgian-Tithonian and the Tithonian-Berriasian boundaries extremely imprecise. Consequently, we employ magnetochronologic interpolation similar to that discussed in detail in GTS89 and MTS94. For the Jurassic-

Cretaceous transition, we use the magnetic anomaly block mode1 derived from the Hawaiian lineation set that was shown to best approximate a constant spreading rate (Channell et al., 1995). The oldest Cretaceous isotopic date available to anchor the magnetochronoIogy is 137.1+1.6/-0.6 Ma (Item 56) that is indirectly correlated to the Late

Berriasian Ml6 chron using nannofossils (Bralower et al., 1990). A subseqent revision suggests that a broader correlation 6th the Ml64415 interval is more appropriate (Channell et al., 1995). Anchoring the Jurassic side OF the lineation set is more controvenial. MTS94 uses direct dating of M26r at 155.3k3.4 Ma in the Argo Abyssal

Plain (Ludden, 1992). This in fact is a minimum K-Ar age of a celadonite vein therefüre we did not include it in our dataset (see p. 129). €rom the same site, incremental heating M~r-39~rdating of basalt from M25-26 did not yield a plateau age, only a disputable total fusion age of 15516 Ma (20) was obtained [Ludden, 1992). Instead we rely on the chronogram estimate of the Oxfordian-Kimmeridgian boundary (1 54.7+ 1.O/-0.9 Ma).

Magnetochronologic correlation of land-based, ammonite-constrained sections with the oceanic magnetic anomalies allow the placement of this boundary at the base of M2Sn (Ogg and Gutowski, 1996). As the isotopic ages controlling the chronogram are not precisely constrained (Items 48-55), allowance needs to be made for correlation uncertainty in the chronogram age. We regard M29 near the base of the Late Oxfordian (Ogg and

Gutowski, 1996) as a maximum, based on the maximum age of the Momson Formation by ostracods (Schudack et al., in press) and ammonoids from underIying strata (Imlay, 1980; Callomon, 1984). The interpolation yields

15 1.5+1.O/- 1.4 Ma for the Kimmeridgian-Tithonian boundary and 144.8+2.6/-3.7 Ma for the Tithonian-Berriasian boundary. The error limits reflect a combination of the numeric errors of the Oxfordian-Kimmeridgian boundary chronogram and the Bemasian isotopic age and the associated biochronologic/magnetochronologiccorrelation uncertainties.

7.11. THE SCTRASSIC TIME SCALE

In summary, the initial boundaries of Jurassic stages are proposed as follows: Bemasian (Jurassic-Cretaceous)

Tithonian

Kimmeridgian Oxfordian Callovian

Bathonian

Bajocian Aalenian Toarcian Pliensbachian

Sinemurian

Hettangian (Triassic-Jurassic)

In addition, EarIy and Middle Jurassic chron boundary ages were also estirnated. Although this list remains incomplete, the following initial chron boundaries (other than those coinciding with stage boundaries) are known with less than 5 Ma range of error:

139 Howelli (Aalenian)

Yakounensis (Toarcian) Crassicosta (Toarcian)

PIanuIata (Toarcian)

Carlottense (Pliensbachian) Kunae (Pliensbachian)

Freboldi (PIiensbachian)

Franziceras (Hettangian)

7.12. DISCUSSION

The Surassic thescale developed here differs fiorn its predecessors in several aspects of methodology. it is

based excIusively on U-Pb and 'OA~-'~A~ ages. Although it was not an explicit criterion for inclusion in the

isotopic database, al1 ages used were reported afier 1990. Despite such restrictions. there are 50 Jurassic isotopic

ages used in this study, slightly exceeding the similar nurnber in previous scales (e-g., 45 in GTS89 and 43 in

MTS94). Both GTS89 and MTS94 entered the errors at the Io confidence level whereas the more conservative Za

level is used throughout this study.

irnprovernent is most significant in the Early Jumsic portion of the time scale. Through direct estimates of

the base of the Hettangian and the Crassicosta Chron and eight additional, precise (Le.. error range is less than 5

Ma) chronogram boundary ages, nearly half of the 20 chrons now have acceptable boundary estimates. Direct estimation of zonal duration was attempted for parts of the Pliensbachian and Toarcian only. From the Freboldi to the Crassicosta chrons, a sequence of five well-constrained chronograms and a directly dated boundary permit us to

propose the following best esthates for zonal duration: Freboldi: 1.1 Ma, Kunae: 1.6 Ma, Carlottense: 0.5 Ma,

Kanense: 0.8 Ma, and Planulata: 1.4 Ma. Although the errors of boundary ages clearly indicate that chron durations are also subject to emr, the observed more than three-fold difference among the chron durations is unlikely to be accounted for by randorn error. Therefore Jurassic ammonite chrons appear to be of disparate length. Such a conclusion seriously undemines earlier tirne scales which were founded on interpolation based on the assumption of equal duration of ammonite chrons.

Fig. 7.3 presents a cornparison of the proposed time scale with the five scales most widely used in recent works. Key observations are as follows: (1) The Jurassic Period had a duration of approxirnately 56 Ma, shorter 140 than previously estimated; (2)The Triassic-Jurassic boundary (20 1 Ma) is younger than previously thought; (3)

There is a good on the Jurassic-Cretaceous boundary age among the proposed scale ( 145 Ma) , DNAG.

GTS89, and MTS94, whereas the same boundary is too young in EXX88 and OD94 which use K-Ar glauconite ages; (4) The Late Jurassic appears to be significantly shorter (1 1.8 Ma) han the Early and Middle Jurassic epochs

(23 and 21.4 Ma, respectively); (5) Re-apportioning of time in the Early and Middle Jurassic is a direct consequence of not using interpolated stage durations.

As dernonstrated here, the new generation of time scales need to be based exclusively on U-Pb and "'~r-

39 Ar ages to benefit fiom the improved precision and accuracy offered by these two dating methods. lmproved statistical techniques can be sought to refine age estimates while accommodating asymmetric, non-Gaussian errors fiequently encountered in U-Pb dating. However, the goal of systematicaI1y defining chron boundary ages can only be achieved through the acquisition of still more catibration points, ultimately eliminating the need for interpolation. The dating of volcanic flow or pyroclastic uni6 within fossiliferous sequences is proved to be the most success~lway of obtaining calibration points. An increasingly more refined time scale offering consistently high precision and resolution at the zona1 Ievel is shown in this snidy to be a reajistic goal. Ma DNAG EXX88 GTS89 0094 MTS94 THIS STUDY

6 TTH *'II-1 T KIM a6 i"if TTH TTH OXF fl.0 KIM u.2 CLV OXF +2 i3.6 CLV BTH u . 9.8 BTH BAJ 54.0 *41-3 AAL BAJ t4.0 TOA AAL *4, TOA PLB i4.a PL0 $39 SIN w-'l SIN i3.n HET fi HEL

Figure 7.3. Cornparison of the proposed Jurassic time scale with other major time scales (DNAG, EXX88, GTS89, MTS94,OD94). Stage abbreviations follow those in GTS89. Error bars for stage boundaries in the new scale are shown on the right. CHAPTER 8

TIME CONSTRAINTS ON JUIUSSIC MASS EXTINCTIONS

AND RECOVERIES

8.1. INTRODUCTION

Mass extinction events (MEE) have been in the focus of heightened paleontologicat research since the theory of extraterrestrial impacts leading to extinctions was documented by Alvarez et al. (1980). in atternpting to understand the dynamics of MEE, attention and research effort has also been directed to the processes of biotic recovery following MEE (e.g., Hart, 1996). Basic questions concerning the dynamics of extinction and recovery, such as rates of divenity change, taxon origination, and extinction, can only be quantified using an accurate time scale. The derection of fine-scale patterns were previously prohibited by tirne sales that were based on the assumption of equal duration for biochronoiogic units. True zonal resolution time scales such as the one attempted here are needed to model recovery processes. The assessrnent of extinction periodicity, a pattern first documented by Raup and Sepkoski (19841, is also dependent on the tirne scale used.

Stage level time series anaiysis of a compiled global taxonomie dataset dernonstrates extinction peaks that surpass background extinction among both marine animal genera and families (Sepkoski, 1996). The end-Triassic crisis represents a first-order MEE whereas second order peaks corresponding to the Piiensbachian, Caltovian, and

Tithonian were also identified. We discuss below the implications of the amended Jurassic time scale, outIined in the previous chapter, for the end-Triassic and Surassic MEE. Revised timing constraints help provide insights into the following problems: (1) the timing and gradua1 vs. sudden nature of MEE;(2) the timing of biotic recoveries;

and (3) the spacing of MEE within the Jurassic and, in a broader context, the Mesozoic.

8.2. THE END-TRIASSIC MASS EXTINCTION AND EARLIEST JURASSIC

RECOVERY

The end-Triassic ME€ is one of the five most severe biotic crises dunng the Phanerozoic. Recent surnrnaries are given by Hallam (1990; 1996a). The age of the Triassic-Jurassic boundary, marked by this MEE, is shown to be 201*l Ma (see chapters 6 and 7). A detailed discussion of this event and the implications of the revised age is also given in Chapter 6.

The earliest Jurassic recovery is the most significant evolutionary radiation during the Jurassic.

Unfortunately, the temporal framework outlined here remains inadequate to unambiguousIy describe the repopulation of niches vacated after the end-Triassic MEE. Besides the age of the Triassic-Jurassic boundary. only the initia1 boundary of the Franziceras chron (i.e., the beginning of the Middle Henangian) is defined satisfactorily with the chronogram method as 198.7k1.3 Ma. The duration of the Early Hettangian (i.e., Planorbis standard chron) is therefore best estimated at 2.3 Ma but the error ranges indicate that this chron could also possibly be very short. Control over the younger Hettangian and Sinemurian chrons rernains too weak to assess the recovery pattern.

Owing to the sparsity of basal Jurassic rocks worldwide, earliest Jurassic faunas are not weI1 known but appear to be cosmopolitan and of low diversity. Compilations of diversity changes exist for the ammonoids (Hallam. lW6k

Smith and Liang, in review), bivalves, and brachiopods (Hallam, 1987). Ammonoids, the best studied fossil group, suffered near-extinction at the Triassic-Jurassic boundary and show extremely low diversity at al1 taxonomic levels in the Planorbis Chron. The same holds mefor the other groups as well. Thus, this zone can be regarded as the post-extinction lag period. Delayed recovery, known to be especially pronounced in the Early Triassic (Hallam,

1991), therefore may also characterize the Planorbis Chron. One of the simple and popular models describing evolutionary radiations is the logistic growth model predicting initial exponential growth in diversity, Iater dampened to approach an equilibrium state (Walker, 1985). Graphically it is rnanifest in a sigmoidaI growth curve.

Assuming that rates of taxonomic evolution, as expressed in diversity change, are reft ected in the duration of empiricd biochronological units, the duration of the first few chrons is cntical to determine the shape of the

diversity growth curve. ïhe greatest increase in supraspecific ammonoid diversity is confined to the Middle

Hettangian which is therefore expected to be shorter than the Early Hettangian. Isotopic age constraints are not

precise enough yet to substantiate this model. In fact, our dating results rernain also compatible with alternative

results of independent zonal duration estirnates derived from Milankovitch cyclostratigraphy in southern England.

Based on presurned cycle counts, Smith (1990) found the Planorbis Chron only about half as long as the next younger Liasicus Chron. One caveat is that the Triassic-Jurassic boundary as defined by the incorning of

Psiloceras in England may be facies controlled thus this biozone may not represent the entire earliest Jurassic

locally (Hallam, 1995).

8.3. THE PLIENSBACHIAN-TOARCIAN EVENT

Raup and Sepkoski (1984) onginally recognized a second order extinction peak in the Pliensbachian. HaIlam

(1986) argued that this event in fact occurred during the Early Toarcian and was regional in extent, in response to the anoxic event in epicontinental northwest Europe and the western Tethys (Jenkyns, 1988). Aberhan and Fürsich

(1995) noted that the extinction is recorded in South America thus it is not restricted to a single region.

Recent detailed analyses at zonal level temporal resolution suggest that the Late Pliensbachian-Early

Toarcian event spans five standard ammonite chrons across the Pliensbachian-Toarcian boundary (Little and

Benton, 1995; Liale, 1996). This coincides with the best controlled interval in our time scale which indicates that the elevated extictintion rate was sustained for 4.3+1.8/-2.0 Ma. Backward smearing in the fossil record (Signor and Lipps, 1982) is an alternative explanation to account for a protracted and apparently gradua1 extinction pattern but, in this case, the broad extinction plateau suggests a tnily long-term event.

There appears to be an immediate recovery after the Pliensbachian-Toarcian event, even if it failed to restore the diversiiy of ammonites and other groups to the Late Pliensbachian level (Hallam, 1996b; Smith and Liang, in review). Although not very well constrained, the average zonal duration for the Middle and Late Toarcian recovery interval (1.1 +O.9/-O.ï Ma) is statistically indistinguishable from that of the preceding crisis (0.9+0.3/-0.4 Ma). 8.4. LATER WRASSIC EVENTS

The percent extinction curves of Sepkoski (1996) contain significant maxima in the Callovian and the

Tithonian, sirnilar in magnitude to the Pliensbachian. However, significant extinction in the Callovian is only discernible in filtered global data containing only "well-preserved" genera and it has not yet been recognized in local or regional snidies (Sepkoski, 1996). There is one study which suggests elevated Ir levels at the Callovian-

Oxfordian boundary in Poland (Brochwicz-Lewinski et al., 1986). Our best age estimate for the Callovian-

Oxfordian boundary is 156.6+2.0/-2.7 Ma.

The Tithonian extinction peak is displayed at both genus and farnily level, in filtered or unfiltered global data

(Raup and Sepkoski, 1984; Sepkoski, 1996). NevertheIess, it remains controversial because marked extinction appears to be reçtricted to bivalves and some oîher benthic groups in Europe. Its global extent has not been demonstrated. Changes in ammonoid faunas are less pronounced at the Jurassic-Cretaceous boundary (Hallam.

1986). Our estimate for the Tithonian-Berriasian boundary is 144.8+2.6/-3.7 Ma.

Clearly, the significance and dynarnics of these events are poorly understood. Aggravating the problem is the lack of finer, zona1 resolution in our later Jurassic time scale. Much more work is needed to establish a better temporal framework for the purported Callovian and Tithonian extinction events.

8.5. SPACING OF JURASSIC AND OTEER hlESOZOIC MASS EXTINCTIONS

The spacing of the end-Triassic and the Pliensbachian-Toarcian events was briefly discussed in Chapter 6.

Here we develop a broader framework to discuss al1 Jurassic events discussed above, along with other major events in the Mesozoic or at its boundaries. Liberating the underlying tirne scale from the effect of interpolation based on the assumption of equal duration for zones/subzones removes any possible autocorrelation of the aileged extinction periodicity and biochronologies. Apart fiom our Jurassic time scale, we use estirnates fiom direct dating of the following mass extinction boundaries: Permian-Triassic: 250 Ma (Claoue-Long et al., 199 1; Renne et al.. 1995).

Cenomanian-Turonian: 93 Ma (Obradovich, 1993; Kowallis et al., 1995), Cretaceous-Tertiary: 65 Ma (Sharpton et al., 1992; Swisher et al., 1992; Krogh et al., 1993). Cretaceous-Tertiary

Callovian (CLV) ( 92 [1041 1 64 [781 Pliensbachian- end-Triassic (14)

- -- endPermian (P-T) 185 157

lTH CLV PL& TJ P-1 TOA 145 157 182-185 201 250 Ma

Cretaceous-Tertiary I I

Callovian (CLV) 1 92 [IO41 1 64 (781 1 12 [26] 1 + 1

K-T CE& TTH CLV PL& TJ P-T TUR TOA 65 93 145 157 182-185 201 250 Ma

Figure 8.1. Periods between mass extinction events in the Mesozoic and at its boundaries. For cornparison, predicted values derived from the 26 Ma periodicity hypothesis (Raup and Sepkoski, 1984) are given in brackets. Bold face font in shaded box denotes good agreement between the mode1 and actual values. A: Model periods (in brackets) based on a three period lengths duration of the Jurassic. B: Model periods (in brackets) based on a two period lengths duration of the Jurassic. See tex1 for discussion. A compmkon of the actual period between extinction events with their predicted spacing based on the 26 Ma periodicity mode1 (Raup and Sepkoski, 1984) is presented in Fig. 8.1. It allows a simple test of the mode[, without resorting to sophisticated mathematical tools. At both the young and the old ends of the studied interval, the

Tithonian and Cretaceous as weIl as the end-Permian and end-Triassic events appear to fit the periodic rnodel. The spacing of the two intra-Jurassic events, the Pliensbachian-Toarcian and the Callovian is also approximately 26

Ma. However, when compared to the younger and older events, these two extinctions appear consistently out of phase compared to the predicted periodicity. Curiously, equating the whole Jurassic to two period lengths instead of three restores the apparent periodicity among the pre-Jurassic and terminal to post-Jurassic events (see Fig 8.1 B).

The statistical significance of this pattern rernains to be evaluated. CHAPTER 9

There are more than a dozen differenr time scales published in the last 15 years that provide age estimates for

Jurassic stage boundaries. Their discrepancies and the large uncertainties hamper their use and warrant a new anempt to produce a refined time scale. A major weakness of the available scales is the small number and often Iow quality of isotopic ages used, among which K-Ar and RbSr dates of questionable reliability predominate. A fiirther problem is the Frequently imprecise stratigraphic constraints on many plutonic rocks dated.

Recent advances in radiomemc dating established the U-Pb and 'O~r-j~~rmethods as the two most precise and accurate dating methods currently availabie. The use of these techniques to date volcaniclastic rocks of precisely known stratigraphic age promises to yield usefiil new calibration points, as proved by recent studies from other systems.

Sarnples for U-Pb zircon dating were collected from Stikinia, Wrangellia, and Peninsular ternes of British

Columbia and Alaska. In these Jurassic arc assemblages, the sampled volcaniclastic rocks Vary from proximal pyroclastic flows to dista1 air-faII tuffs and occur interbedded in fossiliferous marine sediments that are dated by ammonite biochronology. The chaIlenge of such an integrated dating approach is that thin ash layets, frequently found in abundantly fossiliferous sections, generally contain only a small amount of zircon, whereas proximal pyroclastics are more likely to yield abundant zircon but the surrounding sedimentary facies are often poorly fossiliferous.

We report 18 new Li-Pb ages fiom Jurassic rocks integrated with zona1 biochronologic dating of under- and overlying strata. In the Queen Charlotte Islands, a tuff layer imrnediately below the Triassic-Jurassic boundary was dated at 20W 1 Ma Lowermost Jurassic (middle and upper Hettangian) volcaniclastics from Alaska yielded ages of

200.8+2.7/-2.8 Ma, 197.U1.0 Ma, and 197.8+1.2/-0.4 Ma. An ash layer near the base of the Crassicosta Zone

(Middle Toarcian) in its type section in the Queen Charlotte Islands was dated at 18 1.4* 13 Ma The other new dates also furnish important time scale calibration points.

The ammonite biostratigraphy from sections measured in the course of this snidy provide important contributions for the North American Jurassic, especially from the Hettangian of Alaska. Theoretical biochronological investigation is centered around the quantification of correlation uncertainty. Statistical range extension based on the number and density of occurrence levels in the observed ammonoid ranges was considered but the 95% confidence level is empirically shown to be too high for a well-studied ammonite locality. The cornputer-assisted Unitary Association rnethod was used to identiQ global maximum ranges of Toarcian ammonoid taxa and determine the most Iikely and the maximum permissible correlation of a North American regional ammonite zone with the northwest European standard. The Unitary Associaton method is not cormpted significantly by the taxonomie noise in the database. The correIation uncertainty between the two regions is not more than F 1 standard subtone.

A radiornetric age database was built for the construction of the revised Jurassic time scale. Apart frorn the 18 newly obtained U-Pb ages, sampling sites of several recently reported Cordilleran ages were studied in order to improve biochronologic constraints. Additional dates were compiled from previous time scales and recent liteninire. The database consists of 50 U-Pb and '~r-~~~rages with a precision of 15 Ma (20) or better that are confined to no more than two adjacent stages, preferably expressed at tonal resolution.

For the first the for the Jurassic, the calibration of chron boundaries was attempted. Direct dates are available for the Triassic-Jurassic boundary and the initial boundaries of the Crassicosta chron and the Callovian stage. The chronogram method was used to estimate ail Early and early Middle Jurassic chron boundaries, late

Middle Jurassic substage boundaries, and Late Jurassic stage boundaries. Chronogram estimates are meaningful only when controHing ages are available for a given stratigraphic unit, therefore the chronogram is different from those of the adjacent units. The most significant improvernent concerns the Pliensbachian and Toarcian, where six consecutive chron boundaries are determined. The derived chron durations are disparate, varying between 0.4 and

1.6 Ma. The assumption of equal duration of chrons or subchrons, used for interpolation in several previous time scales, is therefore not considered valid. For Early and Middle Jurassic stage boundaries, we use the appropriate chronogram estimate if it is unique and corresponds to the chronogram of the earliest chron of the given stage, otherwise adjuments are made and justified on a case by case basis. The latest Jurassic isotopic database remains too sparse, therefore chronogram estimates can be improved using interpolation based on magnetochronology. The initial boundaries of Jurassic stages are proposed as follows:

Berriasian (Jurassic-Cretaceous) 144.8 t2.61-3.7 Ma

Tithonian

Kimmeridgian

Ox fordian

Callovian

Bathonian

Bajocian

Aalenian

Toarcian

Pliensbachian

Sinemurian

Hettangian (Triassic-Jurassic)

In addition, the following initial chron boundaries (other than those coinciding with stage boundaries) are known with better than 5 Ma range of error:

Howelli (Aalenian) 177.6 t1.41-1.1 Ma

Yakounensis (Toarcian) 180.1 +0.71-3.0 Ma

Crassicosta (Toarcian) 181.4k1.2 Ma

PlanuIata (Toarcian) 182.8 +2.3/-2.6

Carlottense (Pliensbachian) 184.1 + 1 .O/-1.6 Ma

Kunae (Pliensbachian) 185.7 +0.6/-0.8 Ma

Freboldi (Pliensbachian) 186.8 11.8 Ma

Franziceras (Hettangian) 198.7A1.3 Ma This study demonstrates that a new generation of time scales should onIy employ reliable and precise U-Pb

and 40~r-39~rdating integrated with zonal biochronologic consûaints. Although still more calibration points are

needed, a time scale for chron boundaries is a realistic goal, partiaIly accomplished herein.

One application of the revised time scale is a re-evahation of the timing of mass extinction events and

subsequent biotic recoveries. The Triassic-Jurassic boundary at 20 1 Ma is younger than suggested by most

previous time scales. The end-Triassic marine and terrestrial extinctions appear to be simultaneous. The latest

Triassic (Norian-Rhaetian) was a several million years long period of decline for ammonoids. bivalves, and

conodonts, while marine rnicroplankton and terresiria1 vertebtates and paIynomorphs show an abrupt change at the

boundary. The duality in temporal extinction patterns may imply an interplay of of long-term environmental ctisis

severdy affecting shelf habitats and an abrupt terminal Triassic event. The timing of the recovery is still not well

consuained. The Planorbis Chron, interpreted as a post-extinction lag period of global low diversity. has an

estimated duration of 2.3 Ma but its error range also permits a much shorter lengih. The second-order

Pliensbachian-Toarcian mass extinction, recently shown to span £ive standard ammonite zones (Little and Benton,

1995), coincides with the best controlled interval of the new time smle. Elevated extinction rate was sustained for

4.3+1.8/-2.0 Ma, between 185.7 and 18 1.4 Ma. It appears to be followed by an immediate recovery, and average

zonal durations before, during, and after the event are indistinguishable. The nature of the Callovian and Tithonian

mas extinctions remains controversial. The 26 Ma periodicity of mass extinctions (Raup and Sepkoski. 1984). a

theory criticized on the grounds of inaccurate lime scales, was assessed using the revised Jurassic time scale and

ment dating of oher Mesozoic mas extinctions. The approximately 26 Ma periodicity is plausible between the end-Tnassic and end-Pemian, between the Pliensbachian-Toarcian and Callovian, and among the Tithonian,

Cenomanian-Turonian, and end-Cretaceous evenb. However, the spacing of the end-Triassic and Pliensbachian-

Toarcian events is less than 20 Ma, significantly less than predicted by the periodicity model. Aberhan, M. 1994. Early Jurassic Bivalvia of northem Chile. Part 1. Subclasses Palaeotaxodonta, Pteriomorphia, and Isofilibranchia. Beringeria, 13: 1-1 15.

Aberhan, M., and Fürsich, F. 1995. Divenity analysis of Lower Jurassic bivalves of the Andean basin and the Pliensbachian-Toarcian mas extinction. Canadian Paleontology Conference, Program and Abstracts, 5: 1.

Afanasyev, G.D. 1987. Phanerozoic Geochronologicat Scale and the Problem of Geological Tirne. Nedra, Moscow, 144 p.

Afanasyev, GD., and Zykov, S.I. 1975. Phanemzoic Geochronological Time Scale in the Light of Significantly New Decay Constants. Nauka, Moscow, 100 p.

Agterberg, F.P. 1988. Quality of tirne scales: a statistical appraisaf. In Current Trends in Geomathematics. Edired by D.F. Memam, Plenum, New York, pp. 57-103.

Agterberg, F.P. 1990. Automated sn-atigraphic correlation. Developments in Palaeontology and Stratigraphy. v. 13. Eisevier, New York, 424 p.

Agterberg, F.P. 1994. Estimation of the Mesozoic geological time scale. Mathernatical Geology, 26: 857-876.

Alvarez, L., Alvarez, W., Asaro, F., and Michel, H. 1980. Extraterrestrial cause for the Cretaceous-Tertiary extinction. Science, 208: 1095-1 108.

Anderson, KG., and McNicoll, V.J. 1995. A note on U-Pb dating of Middle Jurassic plutonic suites: Cumshewa Head pluton, southeastem Moresby Island, Queen CharIone Islands, British Columbia. In Current Research, Part A. Geological Survey of Canada Paper 1995- I A, pp. 9 1-96.

Anderson, R.G., and Reichenbach, 1. 199 1. U-Pb and K-Ar mework for the Middle to Late Jurassic (1 72-2 158 Ma) and Tertiary (46-27 Ma) plutons in Queen Charlotte Islands. British Columbia. In Evolution and hydrocarbon potential of the Queen Charlotte Basin, British Columbia. Edited by G.J. Woodsworth. Geological Survey of Canada Paper 90- 10, pp. 59-87.

Arkell, W.J. 1956. Jurassic Geology of the WorId. Oliver and Boyd, Edinburgh & London, xv + 806 p.

Armstrong, R.L. 1978. Pre-Cenozoic Phanerozoic time scale - Cornputer file of critical dates and consequences of new and in-progress decay-constant revisions. ln Contributions to the Geologic Tirne Scale. Edited by G.V. Cohee, M.F. Glaessner and H.D. Hedberg. Arnerican Association of Petroleum Geologiçtç, Studies in Geology 6, pp. 73-9 1.

Armstrong, R.L. 1982. Late Triassic-Early Jurassic time scale calibration in British Columbia, Canada. In Numerical Dating in Stratigraphy. Edifeà by G.S. Odin, Wiley, Chichester, pp. 509-5 13.

Armstrong, R.L. 199 1. A brief history of geochronornetry and radiogenic isotopic studies. In Applications of Radiogenic Isotope Systems to Problems in Geology. Edited by L. Heaman and J.N. Ludden. Mineralogical Association of Canada Short Course Handbook 19, pp. 1-26.

Arthur, A., Smith, P.L., Monger, J.W., and Tipper, H.W. 1993. Mesozoic stratigraphy and Jurassic paleontology West of Harrison Lake, southwestern British Columbia. Geological Survey of Canada Bulletin 441,62 p. Ash, CH., Macdonald, RWJ., and Friedman, RM. 1997. Stratigraphy of the Tatogga Lake area, northwestern British Columbia (1MW 12&l3 104G/9&16). In Geological Fieldwork 1996. British Columbia Geological Survey Paper 1997- 1, pp. 29 1-296.

Ash, C.H.,Stinson, P.K., and Macdonald, R.W.J. 1996. Geology of the Todagin Plateau and Kinaskan Lake Area hiorthwestern British Columbia (104 W 12, 1 O4Gf9). In Geological Fieldwork 1995. Edired by B. Grant and LM. Newell. British Columbia Geological Survey Paper 1996- 1, pp. 155- 1 74.

Barker, F., and 14 others. 1994. Some accreted volcanic rocks of Alaska and their elernental abundances. In The Geology of Alaska Edired by G. Plafker and H.C. Berg. The Geology of North America, v. G-I. Geological Society of America, Boulder, Colorado, pp. 555-587.

Barrell, J. 19 17. Rhythm and the measurements of geologic time. Geological Society of America Bulletin, 28: 745-904.

Baumgartner, P.O. 1984. Cornparison of unitary associations and probabilistic ranking and scaling as applied to Mesozoic radiolarians. Computers & Geosciences, 10: 16% 183.

Baumgartner, P.O. 1993. EarIy Cretaceous radiolarians of the Northeast Indian Ocean (Leg 123: Sites 765,766 and DSDP site 26 1): The Antarctic-Tethys connection. Marine Micropaleontology, 21 : 329-352.

Bayer, U. 1987. Chronometric calibration of a comparative time scale for the Mesozoic and Paleozoic. Geologische Rundschau, 76: 485-503.

Bengtson, P. 1988. Open nomenclature. Palaeontology, 31 :223-227.

Bloos, G. 1983. nie zone of Schlotheimia murmorea (Lower Lias) - Hettangian or Sinemurian? Newsletters on Stratigraphy, 12: t 23- 1 3 1.

Bloos, G. 1988. Ammonizes murmoreus Oppel (Schlotheimiidae) im unteres Lias (angulata-Zone. depressa- Subzone) von Wnrttemberg (Sildwestdeutschland).Stuttgarter BeiMge mr Naturkunde. Serie B, 141: 1-47.

Bloos, G. 1994. Frllhe Arietitidae (Ammonoidea) aus dem Hettangium (Angulata-Zone, Unt. Lias) von Wurtternberg (SW-Deutschland). Stuttgarter Beitriige zur Naturkunde, Serie B, 219: 1-67.

Bloos, G. 1996. The Hettangian/Sinernurian stage boundary - Funher observations. In Advances in Jurassic Research. Edired by A.C. Riccardi. GeoResearch Forum, 1-2. Transtec Publications, Zurich, pp. 1 19- 128.

Bown, P.R. 1996. Recent advances in Jurassic caIcareous nannofossil research. In Advances in Jurassic Research. Edired by A.C. Riccardi. GeoResearch Forum 1-2. Transtec Publications. Zurich, pp. 55-66.

Bralower, TL,Ludwig, K.R., Obradovich, J.D., and Jones, D.L. 1990. Berriasian (Early Cretaceous) radiometric ages fiom the Grindstone Creek Section, Sacramento Valley. Califomia. Earth and Planetary Science Letters, 98: 62-73.

Brochwicz-Lewinski, W. et al. 1986. Anornalia irydowa na granicy juiy srodkowej i gomey. Pneglad Geologicny, 33: 83-88.

Callomon, J.H. 1984a. A review of the biostratigraphy of the post-Lower Bajocian ammonites of western and northem North America. In luiassic-Cretaceous biochronology and paleogeography of North America. Edired by G.E.G. Westermann. Geological Association of Canada Special Paper 27, pp. 143- 174.

Callomon, J.H. 1984b. Biostratigraphy, chronostratigraphy and al1 that-again! In International Symposium on Jurassic Stratigraphy, Erlangen. Edired by O. Michelsen and A. Zeiss, v. 3. Geological Survey of Denmark, pp. 6 1 1-624. Callomon, J.H. 1995. Time fiom fossiis: S. S. Buckman and Jurassic high-resolution geochronology. In Milestones in Geology. Edited by M.J. Le Bas. Geological Society of London Memoir 16, pp. 127-150.

Cameron, B.E.B., and Tipper. H.W. 1985. Jumsic stratigraphy of the Queen Charlotte Islands, British Columbia. Geological Survey of Canada Bulletin 365.49 p.

Capps, S.R 1923. The Cold Bay district Alaska. United States Geological Survey BuIletin, 739-C: 77-1 16.

Cm, P.F., Jones, B.G., Quinn, B.G., and Wright, A.J. 1984. Toward an objective Phanerozoic time scale. Geology, 12: 274-277.

Carter, E.S. 1993. Biochronology and paleontology of uppermost Triassic (Rhaetian) radiolarians, Queen Charlotte Islands, British Columbia, Canada. Mernoires de Géologie (Lausanne), Il: 1-1 75.

Carter, ES. 1994. Evolutionary trends in latest Norian through Hettangian radiolarians from the Queen Charlotte Islands, British Columbia. Gbobios, Mémoire Spécial, 17: 1 1 1-1 19.

Carier, E.S., Whalen, P.A., and Guex, S. in press. Biochronology and paleontology of Lower Jurassic (Hettangian and Sinemurian) radiolarians fiorn the Queen Charlotte Islands, British Columbia. Geological Survey of Canada Bulletin.

Channell, J.E.T., Erba, E., Nakanishi. M., and Tamaki, K. 1995. Late Jurassic-Early Cretaceous time scales and oceanic magnetic anomaly block models. In Geochronology, Time Scales and Global Snatigraphic Correlation. Edited by W.A. Berggren, D.V. Kent, M.-P. Aubry and J. Hardenbol. SEPM Special Publication 54, pp. 5 1-63.

Childe, F. 1996. U-Pb geochronology and Nd and Pb isotope characteristics of the Au-Ag-rich Eskay Creek volcanogenic massive sulfide deposit, British Columbia. Economic Geology, 91 : 1209- ISX.

Claoué-Long, J-C., Zichao, Z., Gougan, M.. and Shaohua, D. 1991. The age of the Permian-Triassic boundary Earth and Planetary Science Letters, 105: 182-190.

Clark, J.R., and Williams-Jones, A.E. 1991. 40~r/39~rages of epithermal alteration and volcanic rocks in the Toodoggone Au-Ag district, north-central British Columbia (94E). In Geological Fieldwork 1990. British Columbia Geological Survey Branch, Paper 199 1 - 1, pp. 207-2 16.

Cope, J.W.C., Getty, T.A., Howarth, M.K., Morton. N., and Torrens, H.S. 1980. A correIation of Jurassic rocks in the British Isles, Part 1: Introduction and Lower Jurassic. Geological Society of London Special Report 14. 73 P-

Cowie, J.W., and Bassett, M.G. 1989. Global Stratigraphic Chart. Episodes. 12: supplement.

Cox, A.V., and Dalrymple, G.B. 1967. Statistical analysis of geomagnetic reversal data and the precision of potassium-argon dating. Journal of Geophysical Research, 72: 2603-26 14.

Cresta, S., Pallini, G., and Venturi, F. 1989. Jurassic ammonite assemblages in the Valdorbia section. Memorie Descrittive della Carta Geologica d'ltalia, 39: 89-94.

Crickmay, C.H. 1928. The sîratigraphy of Parson Bay, British Columbia. University of California Publications, Bulletin of the Depamnent of Geological Sciences, 18: 51-70.

Dean, W.T. 1954. Notes on part of the Upper Lias succession at Blea Wyke, Yorkshire. Proceedings of the Yorkshire Geological Society, 29: 161- 179.

Dean, W.T., Donovan, D.T., and Howarth, M.K. 1961. The Liassic ammonite zones and subzones of the northwest European province. Bulletin of the British Museum (Natural History), Series A, Geology, 4: 438-505. DeBari, S.M., and Sleep, N.H. 1991. High-Mg, low-Al bulk composition ofthe Talkeetna arc, Alaska: Implications for primary magmas and the nature of arc cm. Geological Society of Amenca Bulletin, 103: 37-47.

Desjardins, P., Machtyre, D.G., Hunt, J., Lyons, L., and Pattenden, S. 1990. Geology of the Thautil River map area (93U6). in Geological Fieldwork, 1989. British Columbia Ministry of Energy, Mines, and Petroleum Resources, Paper 1990-1, pp. 91-99.

Desroctiers, A., and Orchard, M.J. 199 1. Stratigraphic revisions and carbonate sedirnentology of the Kunga Group (Upper Triassic-Lower Jurassic), Queen Charlotte Islands, British Columbia. In Evolution and hydrocarbon potential of the Queen Charlotte Basin, British Columbia. Edited by G. Woodsworth. Geological Suwey of Canada Paper 90-10, pp. 163-172.

Detteman, R.L., Miller, J.W., and Case, S.E. 1985. Megafossil locality map, checklists, and pre-Quaternary stratigraphie sections of Ugashik, Bristol Bay, and part of KarIuk quadrangles. Alaska. United States Geological Suwey Miscellaneous Field Studies Map, MF-1 539-8.

Diakow, L., and Mihalynuk, M. L987. Geology of Whitesail Reach and Troitsa Lake map areas (93UIOW. 1 1E). In Geological FieIdwork 1986. British Columbia Ministry of Energy, Mines, and Petroleum Resources, Paper 1987-1, pp. 171-179.

Dommergues, J.-L., and Meister, C. 1987. La biostratigraphie des ammonites du Carixien (Jurassique inférieur) d'Europe occidentale: un test de la méthode des associations unitaires. Eclogae Geologicae Helvetiae. 80: 9 19- 938.

Donovan, D.T., Callomon, J.H., and Howarth, M.K. 198 1. Classification of the Jurassic Ammonitina. In The Ammonoidea. Edited by M.R. House and J.R. Senior. Systernatics Association Special Publication 18, pp. 101-155.

Dunning, G.R., and Hodych, J.P. 1990. U/Pb zircon and baddeleyite ages for the Palisades and Gettysburg sills of the northeastem United States: [mplications for the age of the Triassic/Jurassic boundary. Geology. 18: 795- 798.

Edwards, L.E. 1982. Quantitative biosu-atigraphy: The method should suit the data. In Quantitative Stratigraphic Correlation. Edited by J.M. Cubitt and R.A. Reyment, John Wiley & Sons. New York. pp. 45-60.

Elmi, S., Atrops, F., and Mangold, C. 1974. Les zones d'ammonites du Domérien-Callovien de l'Algérie occidentale - PremKre partie: Domérien-Toarcien. Documents des Laboratoires de Géologie de la Faculté des Sciences de Lyon, 61: 1-84.

Elmi, S., GabilIy, J., Mouterde, R., Rulleau, L., and Rocha, R.B. 1994. L'étage Toarcien de l'Europe et de la Téthys; Divisions et correlations. Geobios Mémoire Spécial, 17: 149- 159.

Evenchick, C.A., and McNicolI, V.I. 1993. U-Pb age for the Jurassic McEwan Creek pluton, north-cenual British Columbia: regional setting and implications for the Toarcian stage boundary. In Radiogenic Age and Isotopic Studies: Report 7. Geological Survey of Canada Paper 93-2, pp. 91-97.

Evenchick, C.A., and Porter, J.S. 1993. Geology of west McConnell Creek map are& British Columbia. In Current Research, Part A. Geological Survey of Canada Paper 93- 1A, pp. 47-55.

Fischer, H., and Gygi, R. 1989. Numerical and biochronological time scales correlated at the ammonite subzone level; K-Ar, RMrages, and Sr, Nd, and Pb sea-water isotopes in an Oxfordian (Late Jurassic) succession of northem Switzerland. Geological Society of America Bulletin, 101: 1584- 1597.

Fitch, F.J., Forster, SC, and Miller, J.A. 1974. Geological time scale. Reports on Progress in Physics. 37: 1433- 1496. Fowell, S.J., and Olsen, P.E. 1993. Time calibration of TriassidJurassic microfloral turnover, eastern North Amenca. Tectonophysics, 222: 36 1-369.

Fowell, S.J., Cornet, B., and Olsen, P.E. 1994. Geologically rapid Late Triassic extinctions: Palynological evidence from the Newark Supergroup. In Pangea: Paleoclimate, Tectonics, and Sedimentation Dunng Accretion, Zenith, and Breakup of a Supercontinent. Edired by G.D. Klein. Geological Society of America Special Paper 288, pp. 197-206.

Frebold, H. 1967. Hettangian ammonite faunas of the Taseko Lakes area, British Columbia. Geological Survey of Canada Bulletin 158,35 p.

Frebold, H. 1978. Ammonites from the Late Bathonian "fniskinires fauna" of central British Columbia. Geological Survey of Canada Bulletin 307,27 p.

Frebold, H., and Tipper, H.W. 1969. Lower Jurassic rocks and fauna near Ashcrofl, British Columbia, and their relation to some granitic plutons. Geological Suwey of Canada Paper 69-23,20 p.

Frebold, H., and Tipper, H.W. 1973. Upper Bajocian - Lower Bathonian ammonite fauna and stratigaphy of Srnithers area, British Columbia. Canadian Journal of Earth Sciences, 10: 1 109-1 13 1.

Gabilly, J. 1976. Le Toarcien à Thouars et dans le centre-ouest de la France. Les Stratotypes français, 3. Centre Nationale de la Recherche Scientifique, Paris, 217 p.

Gale, A.S. 1990. A Milankovitch scale for Cenomanian time. Terra Nova, 1: 420-425.

Gale, A.S. 1995. Cyclosuatigraphy and correlation of the Cenomanian stage in Western Europe. In Orbital Forcing Timescales and Cyclostratigraphy. Edited by M.R. House and A.S. Gale. Geological Society of London Special Publication 85, pp. 177- 197.

Gallitelli Wendt, M.F. 1970. Ammoniti e stratigrafia del Toarciano umbro-marchigiano (Appennino centrale). Bolletino della Societa Paleontologica Italiana, 8: 1 1-62.

Goy, A., and Martinez, G. 1990. Biozonacih del Toarciense en el hrea de La Almunia de Dofia Godina-Rida (Sector Central de la Cordillera Ibérica). Cuadernos de Geologia ibérica, 14: 1 1-53.

Gradstein, F.M., Agterberg, F.P., Ogg, J.G., Hardenbol, J., van Veen, P., Thierry, J., and Huang, 2. 1994. A Mesozoic time scale. Journal of Geophysical Research, B, 99: 24,05 1-24.074,

Gradstein, F.M., Agterberg, F.P., Ogg, J.G., Hardenbol, J., van Veen, P., Thierry, J., and Huang, Z. 1995. A Triassic, Jurassic, and Cretaceous time scale. In Geochronology, Time Scales and Global Süatigmphic Correlation. Edited by W.A. Berggren, D.V. Kent, M.-P.Aubry and J. Hardenbol. SEPM Special Publication 54, pp. 95-126.

Greig, C.J., and Gehrels, G.E. 1995. U-Pb zircon geochronology of Lower Jurassic and Paleozoic Stikinian strata and Tertiary intrusions, northwestern British Columbia. Canadian Journal of Earth Sciences, 33: 1 155-1 17 1.

Greig, C.J., McNicoll, V.J., Anderson, R.G., Daubeny, P.H., Harakal, J.E., and Runkle, D. 1995. New K-Ar and U- Pb dates for the Cambria Icefield area, nonhwestern British Columbia. In Current Research, Part A. Geological Survey of Canada Paper 1 995- 1 A, pp. 97- 1 03.

Gudrin-Franiatte, S. 1990. Ammonites du Lias inferieur de France II: Psiloceratidae, Schlotheimiidae et premiers Arietitidae. Mémoires GERGH, 29: 1-207.

Guex, J. 1972. Répartition biostratigraphique des ammonites du Toarcien moyen de la bordure sud des Causses (France) et revision des ammonites décrites et figurées par Monestier (193 1). Eclogae Geologicae Helvetiae, 65: 61 1-645. Guex, S. 1975. Description biostratigraphique du Toarcien supérieur de la bordure sud des Causses (France). Eclogae Geologicae Helvetiae, 68: 97429.

Guex, S. 1980. Remarques préliminaires sur la distribution smtigraphique des ammonites hettangiennes du New York Canyon (Gabbs Valley Range, Nevada). Bulletin de Géologie, Lausanne, 250: 127-140.

Guex, J. 1987. Sur la phylogenése des ammonites du Lias inférieur. Bulletin de Géologie, Lausanne, 292: 1-'7- 140.

Guex, J. 1989. Note sur le genre Frawicerm Buckman (Ammonoidea, Cephalopoda). Bulletin de Géologie, Lausanne, 305: 347-354.

. Guex, J. 1991. Biochronological Correlations. Springer-Verlag, Berlin, 252 p.

Guex, J. 1995. Ammonites hettangiennes de la Gabbs Valley Range (Nevada, USA). Mémoires de Géologie (Lausanne), 27: 1- 13 1.

Guex, S., and Rakus, M. 1991. Les Discamphiceratinae (Psiloceratidae), une nouvelle sous-famille d'ammonites (Cephalopoda)du Jurassique inférieur. BuIIetin de GéoIogie, Lausanne, 31 1: 309-3 16.

Guex, S., and Taylor, D. 1976. La Iimite Hettangian-Sinemurien, des Préalpes romandes au Nevada. Eclogae geologicae Helvetiae, 69: 521-526.

Haggw J.W., and Tipper, H.W. 1994. New results in Jura-Cretaceous stratigraphy. northem Vancouver Island. British Columbia. In Current Research, Part E, Geological Survey of Canada Paper 19944, pp. 59-66.

Hailwood, €.A. 1989. The role of magnetostratigraphy in the development of geological time scales. Paleoceanography, 4: 1- 18.

Hall, R.L., and Westemann, G.E.G. 1980. Lower Bajocian (Jurassic) cephalopod faunas frorn western Canada and proposed assamblage zones for the Lower Bajocian of North America. Palaeontographica Americana, 9: 1-93.

Hallam, A. 1986. The Pliensbachian and Tithonian extinction events. Nature, 319: 765-768.

Hallam, A. 1987. Radiations and extinctions in relation to environmental change in the marine Lower Jurassic of northwest Europe. Paleobiology, 13: 152- 168.

Hallam, A. 1990. The end-Triassic rnass extinction event. In Global catastrophes in Earth history; An interdisciplinary conference on impacts, volcanisrn, and mass mortality. Edited by V.L. Sharpton and P.D. Ward. Geological Society of Arnerica Special Paper 247, pp. 577-583.

Hallam, A. 1991. Why was there a delayed radiation after the end-Palaeozoic extinctions? Historical Biology, 5: 257-262.

Hallam, A. 1995.Oxygen-restricted facies of the basal Jurassic of northwest Europe. Historical Biology, 10: 247- 57.

Hallam, A. 1996a. Major bio-events in the Triassic and Jurassic. In Global Events and Event Stratigraphy in the Phanerozoic. Edited by O.H. Walliser, Springer, Bedin, pp. 265-283.

Hallam, A. 1996b. Recovery of the marine huna in Europe after the end-Triassic and early Toarcian mass extinctions. In Biotic Recovery fiorn Mass Extinction Events. Edifed by M.B. Hart. Geological Society of London Special Publication 102, pp. 23 1-236.

Hallam, A., Hancock, LM., LaBrecque, J.L., Lowrie, W., and Channell, J.E.T. 1985. Junssic to Paleogene: Part 1 Jurassic and Cretaceous geochronology, and Jurassic to Paleogene magnetostratigraphy. In The Chronology of the Geological Record. Ediied by N.J. Snelling. Geological Society of London Memoir 10. pp. 1 18-140. Haq, B.U., Hardenbol, J., and Vail, P.R. 1987. Chronology of fluctuating sea levels since the Triassic. Science, 235: 1 156-1 167.

Haq, B.U., Hardenbol, J., and Vail, P.R 1988. Mesozoic and Cenozoic chronomtigraphy and cycles of sea level change. Society of Economic Paleontologists and Mineralogists Special Publication 42. pp. 7 1 - 108.

Harland, W.B. 1978. Geochronologic scales. In Contributions to the Geologic Time Scale. Edired by G.V. Cohee, M.F. Glaessner and H.D. Hedberg. Arnerican Association of Petroieum Geologists, Studies in Geology 6, pp. 9-32.

Harland, W.B., and Francis, H. (Editors) 1 97 1. The Phanerozoic Time-Scate - a Supplement. GeologicaI Society of London Special Publication 5,356 p.

Harland, WB., Armstrong, R.L., Cox, A.V., Craig, L.E., Smith, A.G., and Smith, D.G. 1990. A Geologic Time Scale 1989. Cambridge University Press, Cambridge, 263 p.

Harland, W.B., Cox, A.V., Llewellyn, P.G., Pichon, C.A.G., Smith, AG., and Walters, R. 1982. A Geologic Time Scale. Cambridge University Press, Cambridge, 13 1 p.

Harland, W.B., Smith, AG., and Wilcock, B. (Editors) 1964. The Phanerozoic the scale. (A symposium dedicated to Professor Arthur Holmes). Quarterly Journal of the Geological Society of London 120 S. 458 p.

Harper, G.D. 1984. The Josephine ophiolite, northwestern California. Geological Sociecy of America BuIIetin. 95: 1009- 1026.

Harper, GD., Saleeby, J.B., and Heizler, M. 1994. Formation and emplacement of the Josephine ophiolite and the Nevadan orogeny in the Klamath Mountains, California-Oregon: UIPb zircon and m~r?9~rgeochronology. Journal of Geophysical Research, 99: 4293-432 1.

Hart, C.J.R. in press. A transect across northern Stikinia: Geology of the northem Whitehorse map area, Yukon Territory (lO5Df 13-16). Yukon Geology Program, Bulletin 8.

Hart, M.B. (Editor) 1996. Biotic Recovery from Mass Extinction Events. Geological Society of London Special Publication 102,394 p.

Hearnan, L., and Parrish, R. 1991. U-Pb geochronology of accessory minerais. In Applications of radiogenic isotope systems to problems in geology. Edited by L. Heaman and J.N. Ludden. Mineralogical Association of Canada Short Course Handbook 19, pp. 59- 102.

Heisler, J., and Tremaine, S. 1989. How dating uncertainties affect the detection of periodicity in extinctions and craters. [karus, 77: 2 13-2 19.

Hess, J.C., Lippolt, HL, and Borsuk, A.M. 1987. Constraints on the Jurassic time-scale by '*~r/''~rdating of North Caucasian volcanic rocks. Journal of Geology, 95: 563-57 1.

Hillebrandt, A.v. 1987. Liassic ammonite zones of South America and correlations with other provices, with descriptions of new genera and species of ammonites. In Biostratigmfia de los sistemas regionaIes del Jurisico y Cretacico en América del Sur. v. 1. Mendoza, pp. 11 1-157.

Hillebrandt, A.v. 1990. The Triassic/Jurassic boundary in northern Chile. Les Cahiers de l'université Catholique de Lyon, Sdrie Scientifique, 3: 27-53.

Hillebrandt, A.v. 1994. The TriassidJurassic boundary and Hettangian biostratigraphy in the area of the Utcubarnba Valley (northern Peru). Géobios Mémoire Spécial, 17: 297-307. Hillebrandt, A.v., and Schmidt-Effing, R. 1981. Ammoniten aus dem Toarcium (Jura) von Chile (SUdamerika). Ziaeliana, 6: 1-74.

Hillebrandt, A.v., Smith, P.L., Westemann, G.E.G., and CalIomon, J.H. 1992. Ammonite zones of the circum- Pacific region. In The Jurassic of the Circurn-Pacific. Edited by G.E.G. Westermann, Cambridge University Press, Cambridge, pp. 247-272.

Hodych, J.P., and Dunning, G.R 1992. Did the Manicouagan impact trigger end-of-Triassic mas extinction? Geology, 20: 5 1-54.

Hofiinan, A. 1985. Patterns of family extinction: dependence on definition and geologic time scale. Nature, 315: 659-662.

Holmes, A. 1937. The Age of the Earth. Nelson, London, 263 p.

Holmes, A. 1947. The construction of a geological time-scale. Transactions of the Geological Society Glasgow. 32: 117-152.

Holrnes, A. 1959. A revised geological tirne-scale. Transactions of the Edinburgh Geologicai Society, 17: 183-2 16.

Hopson, C.A., Mattinson, J.M., and Pessagno, E.A.J. 1981. Coast Range ophiolite, western California. In The Geotectonic Development of California. Edited by W.G. Ernst, Prentice Hall, New Jersey, pp. 418-5 10.

House, M.R. 1985. A new approach ta an absolute timescale from rneasurements of orbital cycles and sedimentary rnicrorhythms. Nature, 31 5: 72 1-n5.

Howarth, M.K. 1962. The Jet Rock Series and the Alum Shale Series of the Yorkshire Coast. Proceedings of the Yorkshire Geological Society, 33: 38 1-422.

Howarth, M.K. 1992. ïhe ammonite family Hildoceratidae in the Lower Jurassic of Britain. Monoagaph of the Palaeontographical Society (London) 145-146 (Publ. No. 586 and 590). 200 p.

Imlay, R.W. 1967. Twin Creek Limestone (Jurassic) in the Western lnterior of the United States. United States Geological Su~eyProfessional Paper 540, 105 p.

Imlay, R.W. 1980. Jurassic paleobiogeography of the conterminous United States in its continental setting. United States Geological Survey Professional Paper 1062, 133 p.

Imlay, R.W. 198 1. Early Jurassic ammonites from Alaska. United States Geological Survey Professional Paper 1 148,49 p.

Imlay, R.W., and Dettermann, R.L. 1977. Some Lower and Middle Jurassic beds in Puale Bay-Alinchak Bay area. Alaska Peninsula. Amencan Association of Petroleum Geologists Bulletin, 61: 607-6 1 1.

Jakobs, G.K. 1986. Ammonite biosfratigraphy of the Bathonian-Callovian of Copper River, northemBritish Columbia. B.Sc. thesis, University of British Columbia, Vancouver.

Jakobs, G.K. 1992. Toarcian (Lower Jurassic) ammonite bioshatigraphy and ammonite fauna of North America. Ph-D. thesis, University of British Columbia, Vancouver.

Jakobs, G.K. 1993. Jurassic stratigraphy of the Diagonal Mountain area, McConnell Creek map am, north-centml British Columbia. In Current Research, Part A. Geological Survey of Canada Paper, pp. 43-46.

Jakobs, G.K. in press. Toarcian ammonite faunas from the North American Cordillera. Geological Survey of Canada Bulletin 428. Jakobs, G.K., and PhKy, J. 1994. Upper Triassic to Middle Jurassic biostratigraphic and facies studies in the Iskut River map area, northwestern British Columbia. In Current Research, Part E. Geological Survey of Canada Paper 94- 1E. pp. I 7-28.

Jakobs, G.K., Smith, P.L., and Tipper, H.W. 1994. An ammonite zonation for the Toarcian (Lower Jurassic) of the North American Cordillera. Canadian Journal of Earth Sciences, 31: 9 19-942.

Jelet*, LA. 1976. Mesozoic and ?Tertiary rocks of Quatsino Sound, Vancouver Island, British Columbia. Geological Survey of Canada Bulletin 242.

Jenkyns, H. 1988. The early Toarcian (Jurassic) anoxic event: Stratigraphic, sedimentary, and geochemical evidence. American Journal of Science. 288: 10 1- 151.

Johannson, G.G. 1994. Biochronology and sedimentology of the Laberge Group: Provenance constraints on Early Jurassic evolution of the northern Stikinian arc, Atlin Lake, northwestern British Columbia. M.Sc. thesis, University of British Cotumbia, Vancouver, 271 p.

Johannson, G.G., and McNicoll, V.J. in press. New U-Pb data from the Laberge Group, northwest British Columbia: ImpIications for Stikinian arc evolution and Early Jurassic time-scale calibrations. In Radiogenic Ages and Isotopic Studies. Geological Survey of Canada Paper 1997-2.

Johannson, G.G., Smith, P.L., and Gordey, S.P. in press. Provenance constraints on Early Jurassic evolution of the northern Stikinian arc: Evidence fiom the Laberge Group, Whitehone Trough, northwestern British CoIumbia. Canadian JournaI of Earth Sciences.

Jones, C.E., Jenkyns, H.C., Coe, A.L., and Hesselbo, S.P. 1994a. Strontium isoropic variations in Jurassic and Cretaceous seawater. Geochimica et Cosmochimica Acta, 58: 3061 -3074.

Jones, C.E., lenkyns, H.C., and Hesselbo, S.P. 1994b. Strontium isotopes in Early Jurassic seawater. Geochimica et Cosmochimica Acta, 58: 1285- 130 1.

Kellum, L.B., Davies, SN.. and Swinney, C.M. 1945. Geology and oil possibilities of the southwestern part of the Wide Bay anticline, Alaska. United States Geological Survey Open-File Report: 1-17.

Kennedy, W.J., and Odin, G.S. 1982. The Jurassic and Cretaceous time scale in 198 1. In Numerical Dating in Stratigraphy. Ediled by G.S. Odin, Wiley, Chichester, pp. 557-592.

Kent, D.V., and Gradstein, F.M. 1985. A Cretaceous and Jurassic geochronology. Geological Society of Arnerica Bulletin, 96: I419-1427.

Kent, D.V., and Gradstein, F.M. 1986. A Jurassic to recent chronology. In The Western North Atlantic Region. Edired by P.R. Vogt and BE. Tucholke. The Geology ofNorth America, v. M. Geological Society of America, Boulder, pp. 45-50.

Koaek, A.V. 1966. Die Ammonitenabfolge des griechischen Toarcium. Annales Géologiques des Pays Helléniques, 17: 1-157.

Kowallis, B.J., Christiansen, E.H., and Deino, A.L. 199 1. Age of the Brushy Basin Member of the Morrison Formation, Colorado Plateau, western USA. Cretaceous Research, 12: 483493.

Kowallis, B.J., Christiansen, E.H., Deino, A.L., Kunk, M.J., and Heaman, L.M. 1995. Age of the Cenomanian- Turonian boundary in the Western Interior of the United States. Cretaceous Research, 16: 109- 129.

Kowallis, B.J.. Christiansen, E.H., Deino, A.L., Peterson, F., Turner, C.E., and Blakey, R.C. 1996. Implications of new high prccisisr; rn~r/3~~r laser-microprobe ages for the Jurassic time scale. In The Continental Jurassic. Edited by M. Morales. Museum of Northem Arizona Bulletin, pp. 7-8. Kowallis, B.J., Christiansen, E.H., Deino, A.L., Peterson, F., Turner, C.E.. Kunk, M J., and Obradovich, J.D. in press. The age of the Morrison Formation. In ïhe Upper luraçsic Morrison Formation: An interdisciplinary snidy. Edired by K. Carpenter, D.J. Chure and LI. Kirkland. Geological Society of America Special Paper.

Kowallis, B.J., Christiansen, EH., Everett, B.H., Crowley, K.D., Naeser, C.W., Miller, D.S., and Deino. A.L. 1993. Possible secondary apatite fission track age standard hmaltered volcanic ash beds in the Middle Jurassic Came1 Fomation, southwestern Utah. Nuclear Tracks and Radiation Measurernents, 21: 5 19-524.

Krogh, T.E. 1982. impmved accuracy of U-Pb zircon ages by the creation of more concordant systems using an air abrasion technique. Geochimica et Cosmochimica Acta, 46: 637-649.

Krogh, T.E., Karno, S.L., and Bohor, B.F. 1993. Fingerprinting the WT impact site and determining the tirne of impact by U-Pb dating of single shocked zircons frorn disial ejecia. Earth and Planetary Science Letten, 119: 425-429.

Kulp, J.L. 1961. GeoIogic time scale. Science, 133: 1105-1 I 14.

Lambert, RJ. 197 1. The pre-Pleistocene Phanerozoic tirne-scale -a review. in The Phanerozoic Tirne-Scale - a Supplernent. E&ed by W.B. Harland and H. Francis. Geological Society of London, Special Publication 5. pp. 9-3 1.

Lange, W. 1952. Der untere Lias am Fonsjoch und seine Ammonitenfauna. Palaeontographica (A), 100: 50-162.

Larson, R.L., and Hilde, T.W.C. 1975. A revised tirne scale of magnetic reversais for the Early Cretaceous and Late Jurassic. Journal of Geophysical Research, 80: 2586-2594.

Lewis, P.D., Thompson, J.F.H., Nadaraju, G.. Anderson, RG., and Johannson, G.G. 1993. Lower and Middle Jurassic stratigraphy in the Treaty Glacier area and geological setting of the Treaty Glacier altemtion systern, northwestern British Columbia. In Current Research, ai^. ~eolo~icalSurvey of Canada Paper 93-IA, pp. 75-86.

Liang, B. 1994. A Lower Jurassic ammonite image database and its applications. Ph.D. thesis, University of British Columbia, Vancouver, 43 1 p.

Little, C.T.S. 1996. The Pliensbachian-Toarcian (Lower Jurassic) extinction event. In The Cretaceous-Tertiary Event and Other Catastrophes in Earth l-iistory. Ediredby G. Ryder, D. Fastovsky and S. Garmer. Geological Society of Arnerica Special Paper 307, pp. 505-5 12.

Little, C.T.S., and Benton, M.J. 1995. Early Jurassic mass extinction: A global long-term event. Geology. 23: 495- 498.

Lihvin, R.J., Peterson, F., and Turner, C.E. in press. Palynological evidence on the age of the Morrison Fomation, Western Interior U.S.: a preliminq report. In The Upper Jurassic Monison Formation: An interdisciplinary study. Ediled by K. Carpenter, D.J. Chure and J.I. Kirkland. Geological Society of America Special Paper.

Ludden, LN. 1992. Radiornetric age determinations for basement from sites 765 and 766, Argo abyssal plain and northwestem Australian margin. Edired by F.M. Gradstein, J.N. Ludden et al. Proceedings of the Ocean Drilling Program, Scientific Results, v. 123, pp. 557-559.

Macdonald, J.A., van der Heyden, P., Lefebure, D.V., and Alldrick, D.3. 1992. Geochmnometry OFthe Iskut River area - an update. In Geological Fieldwork 1991. British Columbia Ministry of Energy, Mines and Petroleum Resources Paper 1992- 1, pp. 495-50 1. MacIntyre, D.G., Webster, I.C.L., and Villeneuve, M. 1997. Babine Porphyry Belt Project: Bedrock geology of the Old Fort Mountain area (93M/I), British Columbia. In Geological Fieldwork 1996. Edired by D. Lefebure, W. McMillan and J. McArthur. British Columbia Ministry of Employment and Investment, Paper 1997-1. pp. 47- 67.

Mahoney, J.B., Friedman, RM., and McKinley, S.D. 1995. Evolution of a Middle Jurassic volcanic arc: stratigraphic, isotopic, and geochemical characteristics of the Harrison Lake Formation. southwestern British Columbia. Canadian Journal of Earth Sciences, 32: 1759-1 776.

Mann, K.O., and Lane, H.R. (Editors) 1995. Graphic correlation. Society of Economic PaIeontoIogists and Mineralogists Special Publication 53,263 p.

Marsden, H., and Thorkelson, D.J. 1992. Geology of the Hazelton volcanic belt in British Columbia: Implications for the Early to Middle Jurassic evolution of Stikinia. Tectonics, 11: 1266-1287.

Marshall, C.R. 1990. Confidence intervals on stratigraphic ranges. PaleobioIogy, 16: 1-1 0.

Marshall, C.R. 1994. Confidence intervals on stratigraphic ranges: partial relaxation oFthe assumption of randomly distributed fossil horizons. Paleobiology, 20: 459-469.

Martin, G.C. 1926. The Mesomic sratigraphy of Alaska. United States Geological Survey Bulletin 776,493 p.

Matsuoka, A. 1992. Jurassic and Early Cretaceous radiolarians fkom Leg 129, Sites 800 and 80 1, western Pacific Ocean. Edited by R.L. Larson, Y. Lancelot et al. Proceedings of the Ocean DriIling Program, Scientific Results, v. 129, pp. 203-2 1 1.

McDougall, I., and Harrison, T.M. 1988. Geochronology and TbennochronoIogy by the '*~r/"~rMethod. Oxford University Press, New York, 2 12 p.

McKinley, S.D. 1996. Volcanic stratigraphy and lithogeochemistry of the Seneca Zn-Cu-Pb prospect, southwestern British Columbia. M.Sc. thesis, University of British Columbia, Vancouver. 192 p.

Menning, M. 1989. A synopsis of numerical time scales 19 17- 1986. Episodes, 12: 3-5.

Memihue, C., and Turner, G. 1966. Potassium-argon dating by activation with fast neutrons. JoumaI of Geophysical Research, 71: 2852-2857.

Monger, J.W.H. et al. 1991. Cordilleran termes. in: Upper Devonian to Middle Jurassic assemblages. In Geology of the Cordilleran Orogen in Canada. Edited by H. Gabrielse and C.J. Yorath. Geology O€ Canada (also The Geology of North America), v. 4 (also G-2). Geological Survey of Canada, (alsa GeologicaI Society of America), pp. 28 1-327.

Monger, J.W.H., and McMillan, W.J. 1984. Bedrock geology of Ashcroft (92 i) map area. Geological Survey O€ Canada Open File 980.

Mortensen, J.K., Ghosh, D.K., and Ferri, F. 1995. U-Pb geochronology of intrusive rocks associated with copper- gold porphyry deposits in the Canadian Cordillera. ln Porphyry Deposits of the Northwestern Cordillera of Nonh America. Edited by T.G. Schroeter. Canadian lnstitute of Mining, Metallurgy and Petroleum Special Volume 46, pp. 142- 158.

Mortimer, N., van der Weyden, P., Armstrong, R.L., and Harakal, J. 1990. U-Pb and K-Ar dates related to the timing of magmatism and deformation in the Cache Creek terrane and Quesneltia, southern British Columbia. Canadian Journal of Earth Sciences, 27: 1 17-123.

Muller, S.E., Northcote, K.E., and Carlisle, D. 1974. Geology and mineral deposits of Aler&Bay - Cape Scott map area, Vancouver Island, British Columbia. Geological Suwey of Canada Paper, 74-8. Mundil, R, Brack, P., Meier, M., Rieber, H., and Oberli, F. 1996. High resolution O-Pb dating of Middle Triassic volcaniclastics: Tirne-scale calibration and verification of tuning parameten for carbonate sedimentation. Earth and Planetary Science Letiers, 141: 137- 15 1.

Nadaraju, G.T. 1993. Triassic-Jurassic biochronology of the eastern Iskut River map area, northwestern British Columbia. M.Sc. thesis, University of British Columbia, Vancouver, 223 p.

Nelson, J.A., and Bellefontaine, K. 1996. The geology and mineral deposits of north-central Quesnetlia; Tezzeron Lake to Discovery Creek, central British Columbia. British Columbia Ministry of Employment and Investment, Geological Survey Branch Bulletin 99, 1 12 p.

Newton, C.R. 1983. Paleozoogeographic affinities ofNorian bivalves fiom the WrangelIian, Peninsular, and Alexander terranes, western North America. In Pre-Jurassic rocks in western North American suspect terranes. Edired by C.H. Stevens, Society of Economic Paleontologists and Mineralogists, Pacific Section, Los Angeles. California, pp. 37-48.

Newton, C.R. 1989. Triassic-Jurassic boundary section at Puale Bay, Alaska Peninsula: Comparative diversity patterns of skeletal faunas and trace fossils. American Association of Petroleum Geologistç Bulletin, 74: 730.

Nixon, G.T., Harnmack, J.L., Koyanagi, V.M., Payie, G.J., Panteleyev, A., Massey, N.W.D., Hamilton, S.V., and Haggart, J.W. 1994. Preliminary geology of the Quatsino-Port McNeiIl map areas, northern Vancouver Island (92U12, 1 1). In Geological Fieldwork 1993. British Columbia Ministry of Energy, Mines and Petroleum Resources Paper 1994-1, pp. 63-85.

Nokleberg, W.J., Plafker, G., and Wilson, F.H. 1994. Geology of south-central Alaska. In The Geoiogy of Alaska. Edited by G. Plafker and H.C. Berg. The Geology of North America, v. G-1 . Geological Society of Amerka. Boulder, Colorado, pp. 3 1 1-366.

Obradovich, J.D. 1988. A different perspective on glauconite as a chronometer for geologic time scale studies. Paleoceanography, 3: 757-770.

Obradovich, J.D. 1993. A Cretaceous time scale. In Evolution of the Western Interion Sasin. Edired by W.G.E. Caldwell and E.G. Kauffman. Geological Association of Canada Special Paper 39, pp. 379-396.

Odin, G.S. (Editor) 1982. Numerical Dating in Stratigraphy. Wiley, Chichester, 1040 p.

Odin, G.S. 1992. Numencal time scale in 1988. in The Jurassic of the Circum-Pacific. Editedby G.E.G. Westermann, Cambridge University Press, Cambridge, pp. 3-1 1.

Odin, G.S. 1994. Geological time scale. Comptes rendus de l'Académie des sciences, Série II, 318: 59-71.

Odin, G.S., and Odin, C. 1990. Échelle numérique des temps géologiques. Gtochronique, 35: 12-21.

Odin, G.S., Baadsgaard, H., Hurford, A.J., and Riccardi, A.C. 1992. U-Pb & fission track geochronology of Bathonian-Callovian "tuffs" fiom Argentina. In Phanerozoic Time Scale. Edited by G.S. Odin. Bulletin de Liaison et Information 1 1, pp. ! :-17.

Ogg, J.G., and Gutowski, J. 1996. Oxfordian and Lower Kirnmeridgian magnetic polarity time scaIe. In Advances in Jurassic Research. Ediled by A.C. Riccardi. GeoResearch Forum 1-2, Transtec Publications, Zurich, pp. 406-4 14.

Ogg, J.G., Hasenyager, R.W., Wimbledon, W.A., Channell, J.E.T., and Bralower, T.J. 199 1. Magnetostratigraphy of the Jurassic-Cretaceous boundary interval-Tethyan and English faunal realms. Cretaceous Research, 12: 455-482. Olsen, P.E., Kent, D.V., Cornet, B., Witte, W.K., and Schlische, RW, 1996a. High-resolution stratigraphy ofthe Newark rifi basin (early Mesozoic, eastern North Amenca). Geological Society of America Bulletin, 108: 40- 77.

Ofsen, P.E., Schlische, RW., and Fedosh, M.S. 1996b 580 ky duration of the Early Jurassic flood basalt event in eastern North America estimated using Milankovitch cyclostratigraphy. In The Continental Jurassic. Edired by M. Morales. Museum ofNortheni Arizona Bulletin 60, pp. 11-22.

Olsen, P.E.. Shubin, N.H.,and Anders, M.H. 1987. New Early Jurassic tetrapod assemblages constrain Triassic- Jurassic tetrapod extinction event. Science, 237: 1025-1029.

Oppei, A. 1856-1858. Die Juraformation Englands, Frankreichs und slidwestlichen Deutschlands. Wlimemberger Naturforschende Jahreshefie, 12-14. Ebner & Seubea Stuttgart, 857 p.

Orchard. M.J. 199Ia. Late Triassic conodont biochronology and biostratigraphy of the Kunga Group, Queen Charlotte Islands, British Columbia. In Evolution and hydrocarbon potential of the Queen Charlotte Basin. British Columbia. Edited by G.J. Woodsworth. Geological Survey of Canada Paper 90- 10, pp. 173- 194.

Orchard, M.I. 199 1b. Upper Triassic conodont biochronology and new index species fiom the Canadian Cordillera. In Qrdovician to Triassic Conodont Paleontology of the Canadian Cordillera. Edired by M.J. Orchard and A.D. McCracken. Geological Survey of Canada Bulletin 417, pp. 399-335.

Orchard, M.J., and Forster, P.J.L. 1991. Conodont colour and thermal maturity of the Late Triassic Kunga Group. Queen Charlotte Islands, British Columbia. In Evolution and hydrocarbon potential of the Queen Charlotte Basin, British Columbia. Edired by G.J. Woodsworth. Geological Survey of Canada Paper 90- 10, pp. 453-464.

Page. K., and Melendez, G. 1995. Protecting the Jurassic: global boundary stratotypes and conservation. Geology Today, (1995): 226-228.

Pal@. J. 1995. Development of the Jurassic geochronologic scale. Hantkeniana, 1: 13-25.

PhlQ, J., and Hart, C.J.R. 1995. Biostratigraphy of the Lower and Middle Jurassic Laberge Group, Whitehorse map area (105D), southeni Yukon. In Yukon Exploration and Geology 1994. Exploration and Geological Services Division, Yukon, Indian and Northern Affairs Canada, pp. 73-86.

Pal@, J., and Schmidt, K.L. 1994. Biostratigraphic and facies studies of the Telkwa Formation (Lower Jurassic), Srnithers rnap area, British Columbia. In Current Research, Part E. Geological Survey of Canada Paper 94-IE, pp. 29-38.

Pile, J., Parrish, RR, and Smith, P.L. 1997. A U-Pb age fiom the Toarcian (Lower Jurassic) and its use for time scale calibration through error analysis of biochronologic dating. Earth and Planetary Science Letters, 146: 659-675.

Pd@, J., Smith, P.L., and Tipper, H.W. 1994. Sinemurian (Lower Jurassic) ammonoid biostratigraphy of the Queen Charlotte Islands, western Canada. Géobios Mdmoire Spécial, 17: 385-393.

Pblfy, J., Smith, P.L., Mortensen, J.K., Friedman, R.M., McNicoll, V.J.. Parrish, R.R., and Anderson, R.G. 1995. Pmgress on Cordilieran Jurassic time scale calibration. Geological Association of Canada Program with Abstracts. 20: A-79.

Palmer, A.R. 1983. The Decade of the North American Geology 1983 geologic time scale. Geology, 11: 503-504.

Parrish, R,R 1990, O-Pb dating of monazite and its application to geological problems. Canadian Journal oFEarth Sciences, 27: 143 1 - 1450. Parrish, RR., and Krogh, T.E. 1987. Synthesis and purification of 2O5Pb for U-Pb geochronology. Chemical Geology, 66: 103- 1 10.

Parrish, RR., Roddick, J.C., Loveridge, W.D., and Sullivan, R.W. 1987. Uranium-lead analytical techniques at the geochronology laboratory, Geological Survey of Canada. In Geological Survey of Canada Paper. 87-2. pp. 3- 7.

Paul, C.R.C. 1982. The adequacy of the fossil record. In Problems of Phylogenetic Reconstruction. Edited by K.A. Joysey and A.€. Friday. Systernatics Association Special Volume 2 1, pp. 75- 1 17.

Pessagno, E.A.J., and Blorne, C.D. 1990. implications of new Jurassic stratigraphie, geochronometric, and paleolatitudinal data from the western Klarnath terrane (Smith River and Rogue Valley subterranes). Geology, 18: 665-668.

Pessagno, E.A.J., and Meyerhoff Hull, D. 1996. "Once upon a tirne in the Pacific": Chronostratigraphic misinterpretation of basal strata at ODP Site 80 1 (central Pacitic) and its impact on geochronology and plate tectonics. In Advances in Jumsic Research. Ediced by A.C. Riccardi. GeoResearch Forum 1-2. Transtec Publications. Zurich, pp. 79-92.

Polubotko, I.V., and Repin, Y.S. 1972. Ammonity i zonal'noye raschleneniye nizhnego leyasa severo-vostoka SSSR [Ammonites and zonal subdivision of the Lower Lias of northeastem USSR]. Materialy po geologii i polenyrn iskopayemym severo-vostoka SSSR, 20: 97-1 16 [In Russian].

Poulton, T.P., and Tipper, H.W. 1991. Aalenian ammonites and strata of western Canada. Geological Survey of Canada Bulletin 41 1, 71 p.

Poulton, T.P., Callomon, J.H., and Hall, R.L. 1991a. Bathonian through Oxfordian (Middle and Upper Jurassic) marine macrofossil assemblages and correlations, Bowser Lake Group, west-central Spatsizi map are& northwestern British Columbia. In Current Research, Part A. Geological Survey of Canada Paper 91-I A, pp. 59-63.

Poulton, T.P., Hall, R.L., and Callornon, J.H. 1994. Ammonite and bivalve assemblages in Bathonian through Oxfordian strata of northern Bowser Basin, northwestem British Columbia, Canada. Geobios Mémoire Spécial, 17: 4 15-42 1.

Poulton, T.P., Hall, R.L., Tipper, H. W., Cameron, B.E.B., and Caner, E.S. 199 1 b. Current status of Middle Jurassic biostratigraphy of the Queen Charlotte Islands, British Columbia. In Evolution and hydrocarbon potential of the Queen Charlotte Basin, British Columbia. Ediced by G.J. Woodsworth. Geological Survey of Canada Paper 90- 1O, pp. 237-252.

Pringle, M.S. 1992. Radiornetric ages of basaltic basernent recovered at Sites 800. 801. and 802, Leg 129, western Pacific Ocean. Edited by RL. Larson, Y. Lancelot and et al. Proceedings of the Ocean Driliing Program, Scientific Results, 129: 389-398.

Quinzio Sinn, L.A. 1987. Sûatigraphische Untenuchungen im Unterjura des SUdteils der Provinz Antofagasta in Nord-Chile. Berliner geowissenschaîtliche Abhandlungen, Reihe A, 87, 105 p.

Rakiis, M. 1993a. EarIy Liassic ammonites from the Steinplane-Kammerkohnlm area (Nonhern Calcareous AlpsISalzburg). Jahrbuch der Geologische Bundesanstalt, 136: 9 19-932.

Rakus, M. 1993b. Lias ammonites of the West Carpathians Part one: Hettangian. Zapadné Karpaty, séria paleontoldgia, 17: 7-40.

Raup, D.M.,and Sepkoski, J.J., Jr. 1982. Mass extinctions in the marine fossil record. Science, 215: 150 1- 1503. Raup, D.M., and Sepkoski, J.J., Jr. 1984. Periodicity of extinctions in the geotogic past. National Academy of Science Proceedings, 81: 80 I-805.

Read, P.B., and Psutka, J. 1990. Geology of the Ealue Lake east-half (104HA3E)and Cullivan Creek (104H/14) map areas, British Columbia Geological Survey of Canada Open File 224 1.

Remane, 1. 1996. The revised guidelines of ICS and their bearing on Jurassic chronostratigraphy. In Advances in Jurassic Research. L4ledbyA.C. Riccardi. GeoResearch Forum 1-2, Transtec Publications. Zurich, pp. 19-22.

Renne, P.R., Zichao, Z., Richards, M.A., Black, M.T., and Basu, A.R. 1995. Synchrony and causal relations between Permian-Triassic boundary crises and Siberian flood volcanisrn. Science, 269: 14 i 3- 14 16.

Repin, Y.S. 1988. Hettangian and Sinemurian. In The Jurassic Ammonite Zones of the Soviet Union. Ediredby G.Y. Kryrnholts, M.S. Mesezhnikov and G.E.G. Westermann. Geological Society of America Special Paper 223, pp. 3-9.

Rhys, D.A., Sieb, M., Frostad, S.R, Swanson, CL., Prefontaine, M.A., Mortensen, J.K., and Smit, H.Q. 1995. Geology and setting of the Red Mountain gotd-silver deposits, northwestern British Columbia. In Porphyry Deposits of the Northwestern Cordillera of North America. Edited by TG. Schroeter. Canadian lnstimte of Mining, MetalIurgy and Petroleum Special Volume 46. pp. 8 1 1-828.

Riccardi, AC, Damborenea, S.E., Manceaido, M.O., and Ballent, SC. 199 1. Hettangian and Sinemurian (Lower Jurassic) biostratigraphy of Argentina. Journal of South American Earth Sciences, 4: 159-170.

Riccardi, AC, Westerrnann, G.E.G., and Elmi, S. 1991. Biostratigraphy of the upper Bajocian-rniddle Caltovian (Middle Jurassic), South America. Journal of South American Earth Sciences. 4: 149-157.

Ricketts, B.D., and Parrish, R.R. 1997. Age and provenance of felsic clam in Bowser Basin, northern British Columbia. Geological Survey of Canada Paper 92-2, pp. t 4 t - 144.

Ricketts, B.D., Evenchick, C.A., Anderson, R.G., and Murphy, D.C. 1992. Bowser basin, northern British Columbia: Constraints on the timing of initial subsidence and StikiniaNorth America terrane interactions. Geology, 20: 1 1 19- 1 122.

Roddick, J.C. 1987. Generalized numerical error analysis with applications to geochronology and thermodynamics. Geochimica et Cosmochimica Acta, 51: 2 129-2 135.

Roddick, J.C., Loveridge, W.D.,and Parrish. R.R. 1987. Precise U-Pb dating of zircon at the sub-nanogram Pb level. Chemical Geology, Isotope Geoscience Section, 66: 1 1 1 - 12 1.

Roden, M., Panish, RR, and Miller, D.S. 1989. The absolute age of the Eifelian Tioga ash bed, Pennsylvania. Journal of Geology, 98: 282-285.

Sadler, P.M., and Strauss, D.J. 1990. Estimation of completeness of stratigraphic sections using empiricaI data and theoretical models. Journal of the Geological Society London, 147: 47I-485.

Saleeby, J.B. 1984. Pb/U zircon ages Frorn the Rogue River area, western Jurassic belt, Klarnath Mountains. Oregon. Geological Society of Amerka Abstracts with Programs, 16: 33 1.

Saleeby, J.B. 1987. Discordance patterns in Pb/U zircon ages of the Sierra Nevada and Klamath Mountains. EOS (Transactions, American Geophysical Union), 68: 15 14-15 15.

Saleeby, J.B., Harper, G.D.,Snoke, A.W., and Shatp, W.D. 1982. Time relations and smctural-stratigraphie patterns in ophioloite accretion, west central Klamath Mountains, California. Journal of Geophysical Research, 87: 383 1-3848. Salvador, A. 1985. Chronostratigraphicand geochronornetric scales in COSLJNA stratigraphic correlation charts of the United States. American Association of Petroleum Geologists Bulletin, 69: 18 1-1 89.

Sandoval, J., Westermann, G.E.G., and Marshall, MC1990. Ammonite fauna, stratigraphy and ecology of the Bathonian-Callovian (Jurassic) Tecocoyunca Group, south Mexico. Palaeontographica (A), 210: 93-149.

Savary, J., and Guex, J. 1991. BioGraph: un nouveau programme de construction des corrélations biochronologiques basees sur les associations unitaires. Bulletin de Gdologie, Lausanne, 313: 3 17-340.

Schudack, M.E., Peterson, F., and Turner, C.E. in press. Biostratigraphy, paleoecology. and biogeography of ostracods and charophytes fiom the Upper Jurassic Morrison Formation, Western Interior, U.S.A. in The Upper Jurassic Monison Formation: An interdisciplinary study. Edired by K. Carpenter, D.J. Chure and J.I. Kirkland. Geological Society of America SpeciaI Paper.

Schwanacher, W. 1993. Cyclostratigraphy and the Milankovitch Theory. Elsevier, Amsterdam, 225 p.

Schweikert, R.A., Bogen, N.L.. Girty, G.H.,Hanson, R.E., and Merguerian, C- 1984. Timing and structural expression of the Nevadan orogeny, Sierra Nevada, California. Geological Society of Arnerica Bulletin, 95: 967-979.

Sepkoski, J.J., Jr. 1996. Patterns of Phanerozoic extinction: a perspective from global data bases. In Global Events and Event Stratigraphy in the Phanerozoic. Edired by O.H. Walliser, Springer, Berlin, pp. 35-51.

Sharpton, V.L., Dalrymple, G.B., Marin, L.E., Ryder, G., Schuraytz B.C., and Uruttia-Fucugauchi. J. t992. New links between the Chicxulub impact structure and the CretaceousITertiary boundary. Nature. 359: 8 19-82 1.

Signor, P.W., and Lipps, J.H. 1982. Sarnpling bias, graduai extinction patterns and catastrophes in the fossiI record. in Geological Implications of Impacts of Large Asteroids and Cornets on the Earth. Eded &y L.T. Silver and P.H. Schiltz. Geological Society of Arnerica Special Paper 190, pp. 291-296.

Smith, A.G. 1993. Methods for improving the chronometric time-scale. In High Resolution Stratigraphy. Edited by E.A. Hailwood and R.B. Kidd. Geological Society of London Special Publication 70, pp. 9-25.

Smith, D.G. 1990. Stratigraphie correlation of presumed Milankovitch cycles in the Blue Lias (Hettangian to earliest Sinemurian), England. Terra Nova, 1: 457460.

Smith, P.L., and Liang, B. in review. Covariation and morphologie diversity patterns in Early Jurassic ammonites. Paleobiology.

Smith, P.L., Beyers, J.M., Carter, ES., Jakobs, G.K., Piilfy, J., Pessagno, E., and Tipper, H. W. 1994. North America, Lower Jurassic. In Jurassic taxa ranges and correlation charts for the Circum Pacific. Edired by G.E.G. Westermann and AC. Riccardi. Newsletters on Stratigraphy 3 1, pp. 33-70.

Smith, P.L., Thomson, R.C., and Tipper, H.W. 1984. Lower and Middle Jurassic sediments and volcanics of the Spatsizi map area, British Columbia. In Current Research, Part A. Geological Suwey of Canada Paper 84-1 A. pp. 117-120.

Smith, P.L., Tipper, H.W., Taylor, D.G., and Guex, J. 1988. An ammonite zonation for the Lower Jurassic of Canada and the United States: the Pliensbachian. Canadian Journal of Earth Sciences. 25: 1503-1 523.

Smith, W.R, and Baker, A.A. 1924. The Cold Bay-Chignik district, Alaska. United States Geologicaf Survey Bulletin, 755: 15 1-2 18.

Snelling, N.J. 1985. An interim tirne scale. In The chronology of the geological record. Edited by N.J. SnelIing. Geological Society of London Memoir 10, pp. 26 1-265. Springer, M., and Lilje, A. 1988. Biostratigraphy and gap analysis; the expected sequence of biostratigraphic events. Journal of Geology, 96: 228-236.

Stacey, J.S., and Kramers, J.D. 1975. Approximation of terrestrial lead isotope evolution by a two-stage model. Earth and Planetary Science Letten, 26: 207-22 1 .

Steiger, R.H., and Jager, E. 1977. Subcommission on geochronology: Convention on the use of decay constants in geo- and cosmochronology. Earth and Planetary Science Letters, 36: 359-362.

Steiner, M.B., Lucas, S.G., and Shoemaker, E.M. 1994. Correlation and age of the Upper Jurassic Morrison Formation fiom magnetostratigraphic analysis. In Mesozoic Systems of the Rocky Mountain Region. Edited by M.V. Caputo, J.A. Peterson and K.J. Franczyk, SEPM Rocky Mountain Section, Denver. pp. 3 15-330.

Strauss, D., and Sadler, P.M. 1989. Classical confidence intervals and Bayesian probability estimates for ends of local taxon ranges. Mathematical Geology, 21: 41 1-427.

Swisher, C.C. et al. 1992. Coeval 40~r/39~rages of650 Million years ago fiom Chicxulub Crater rnelt rock and Cretaceous-Tertiary boundary tektites. Science, 257: 954-958.

Taylor, DG. 1986. The Hettangian-Sinemurian boundary (Early Jurassic): repIy to Bloos I983. Newsletters on Stratigraphy, 16: 57-67.

Taylor, DG. 1990. Two species of Paracaloceraî fiom the Canadense Zone (Hettangian-Sinemurian stages) in Nevada (USA). Bulletin de Géologie, Lausanne, 309: 2 1 1-2 19.

Thomson, R.C., and Smith, P.L. 1992. Pliensbachian (Lower Jurassic) biostratigraphy and ammonite fauna of the Spatsizi area, north-central British Columbia. Geological Survey of Canada Bulletin 437.87 p.

niomson, R.C., Smith, P.L., and Tipper, H.W. 1986. Lower to Middle Jurassic (Pliensbachian to Bajocian) stratigraphy of the northern Spatsizi area, north-central British Columbia. Canadian Journal of Earth Sciences, 23: 1963-1973.

Thorkelson, D.J., Mortensen, J.K., Marsden, H., and Taylor, R.P. 1995. Age and tectonic setting of Early Jurassic episodic volcanism along the northeastem margin of the Hazelton Trough, northern British Columbia. fn Jurassic Magmatism and Tectonics of the North American Cordillera. Edited by D.M. Miller and C. Busby. Geological Society of America Special Paper 299, pp. 83-94.

Tilmann, N. 1917. Die Fauna des unteren und mittleren Lias in Nord- und Mittel-Peru. Neues Jahrbuch tilr Mineralogie, Geologie, und Pal3ontologie, 41: 628-7 12.

Tipper, H.W., and Guex, J. 1994. Preliminary remarks on the Hettangian ammonite succession in Queen Charlotte Islands, British Columbia. Geobios Mémoire Spécial, 17: 477-483.

Tipper, H.W., and Richards, T.A. 1976. Jurassic stratigraphy and history of north-central British Columbia. Geological Survey of Canada Bulletin 270.73 p.

Tipper, H.W., Carter, E.S., Orchard, M.J., and Tozer, E.T. 1994. The Triassic-Jurassic (T-J) boundary in Queen Charlotte Islands, British Columbia defined by ammonites, conodonts and radiolarians. Geobios Mémoire Spécial, 17: 485-492.

Tozer, E.T. 1971. One, two or three connecting links benveen Triassic and Jurassic ammonoids? Nature, 232: 565- 566.

Tozer, E.T. 1994. Canadian Triassic arnmonoid faunas. Geological Survey of Canada Bulletin 467,663 p. Tucker, RD., and McKerrow, W.S. 1995. Early Paleozoic chronology: A review in light of new U-Pb zircon ages hmNewfoundland and Britain. Canadian lournal of Earth Sciences, 32: 368-379. van der Heyden, P. 1991. Prelirninary U-Pb dates and field observations frorn the eastern Coast Belt near 52"N. British Columbia. In Current Research, Part A. Geological Survey of Canada Paper 9 1-A, pp. 79-84.

Van Eysinga, F.W.6.c. 1975. Geological Tirnetable. Elsevier, Amsterdam.

Van Hinte, J.E. 1976. A lurassic time scale. Amerîcan Association of Petroleum Cieologists Bulletin, 60: 489-497.

Van Veen, P.M. 1995. Tirne calibration of Triassicllurassic microfloral turnover, eastern North Arnerica: Comment. Tectonophysics.

Vellutini, D., and Bustin, R.M. 1991. Source rock potential of Mesozoic and Tertiary strata of the Queen Charlotte Islands, British Columbia. In Evolution and hydrocarbon potential of the Queen CharIone Basin. British Columbia. Edired by G.I. Woodsworth. Geological Survey of Canada Paper 90- 10. pp. 38 1-409.

Venturi, F. 1975. Rapponi filetici e stratigrafici dei generi toarciani Mercariceras, Brodieia. Hifdoceras, Phymaroceras, Chartronia dell'Appennino Centrale. Rivista italiana di Paleontologia e Stratigrafia 81 : 195- 246.

W&ner, F. 1882-1898. Beitriige zur Kenntnis der tieferen Zonen des unteren Lias in den nord-ostlichen Alpen, I- VIII. Beitrage mr Paliontologie und Geologie des Osterreich-~ngarnsund des Orients, 2-1 l: 1-29 l.

W(Lhner, F. 1886. Zur heteropischen Differenziemng des alpinen Lias. Verhandlungen der kaiserlich-koniglichen geologischen Reichsanstalt, (1 886): 168-1 76.

Walker, T. 1985. Diversification functions and the rate of taxonornic evolution. ln Phanerozoic Diversity Patterns: Profiles in Macroevolution. Edited by J. Valentine, Princeton Univerity Press, Princeton, pp. 3 1 1-334.

Wang, 1.. Newton, C.R., and Dunne, L. 1988. Late Triassic transition fiorn biogenic to arc sedimentation on the Peninsular terrane: Puale Bay, Alaska Peninsula. Geological Society of Arnerica Bulletin, 100: 1466-1478.

Weedon, G.P. 1985. Hemipelagic shelf sedimentation and climatic cycles: the basal Jurassic (Blue Lias) of South Britain. Earth and Planetary Science Leners, 76: 321-335.

Westennann, G.E.G. 1984. Gauging the duration of stages: a new approach for the lurassic. Episodes. 7: 26-28.

Westennann, G.E.G. 1988. Duration of Jurassic stages based on averaged and scaled subzones. In Recent Advances in Quantitative Stratigraphic Correlation. Edited by F.P. Agterberg and C.N. Rao, Hindustan, Delhi. pp. 90- 100.

Wilson, F.H., and Shew, N. 1992. Map and tables showing geochronology and whole-rock geochernistry of selected samples, Ugashik and part of Karluk quadrangles, Alaska. United States Geological Survey MiscelIaneous Field Studies Map, 1539-E.

Wilson, F.H., Detterman, R.L., and Case, J.E. 1985. The Alaska Peninsula Terrane; a definition. United States Geological Suwey Open-File Report 85-450, 17 p.

Wright, J.E., and Wyld, S.J. 1986. Significance of xenocrystic Precambrian zircon contained within the southern continuation of the losephine ophiolite: Devils Elbow ophiolite remnant, Klarnath Mountains, northern California. Geology, 14: 67 1-674. INPUT OF COMPUTER-ASSISTED BIOCHRONOLOGIC

CORRELATION USING BioGraph

Al.l. LIST OF AMMONOID TAXA AND THlEiR CODES

OOBa Brodieia alticarinatus Hammatoceras porcarellense O1 Bb Brodieia bayani Hammatoceras praefalllca OlOBc Brodieia clausa Hammatoceras speciosum 02Bg Brodieia gradata Hammatoceras 038p Brodieia primaria 048 Brodieia Harpocerasfalcifer Harpoceras suberaratum 05Cc Catacoeloceras crmm Harpoceras subplanafum 06Cg Catacoeloceras ghinii Harpoceras 07C Catacoeloceras Haugia illustris 09Cg Collina gemma Haugia navis 090Cl Collina linae Haugia phillipsi 091Cm Collina mucronata Haugia variabilis i OC Collina Haugia vitiosa Haugia 1 lDa Dactylioceras athiezicum 12Dc Dactylioceras commune Hilahites propeserpentinus 13D Dactylioceras Hildaites serpentiniformis Hildaites serpentinus 14Dm Denckmannia malagrna Hildaites l5Dt Denckmannia tumefacta 16D Denckmannia Hildoceras aperrum Hildoceras bifrons 17Ef Hildoceras carerinii Hildoceras crassum 18Fs Frechiella subcarinata Hildoceras lusitanicum Hildoceras semipolifum 19Gs Grammoceras srriafuitcm Hildoceras sublevisoni 20Gt Grammoceras thouarsense Hildoceras tethysi 21G Grammoceras Hildoceras

22Hc Hammaroceras cosrarum Leukadiella ionica 23Hi Hammatoceras insigne Leu kadieila Mercaticeras diIatum Phymatoceras pseudoerbaeme Mercaticeras hellenicum Phymatoceras robustum Mercariceras mercati Phymatoceras rude Mercaticeras tyrrhenicum Phymatoceras speciosum Mercaticeras umbilicatum Phymatoceras venusrulum Mercaticeras Phymatoceras

Nodicoeloceras crassoides Podagrosites Nodicoeloceras Podagrosites latescem

Paroniceras sternale Pseudogrammoceras aratum Paroniceras Pseudogrammoceras bingmanni Pseudogramrnoceras doerntense Peronoceras desplacei Pserrdogrammoceras/aIlaciosum PeronocerasJbularum Pseudogrammoceras subregale Peronoceras moerickei Pseudogrammoceras Peronoceras planivenier Peronoceras pacificum Pseudolillia emiliana Peronoceras subarmarum Peronoceras verticosum Pseudomercaticerasfi.antzÏ Peronoceras vortex. Pseudomercaticeras Iatum Peronoceras vorticellum Pseudomercaticeras rotaries Peronoceras Pseudomercaticeras

Plymatoceras crassicosta (=cornucopia) Rarenodia planulata Phymatoceras elegans Phymatoceras erbaense Zugodacrylites braunianus Phyrnaroceras hillebrandti Phymatoceras narbonense A1.2. AMMONOID DISTRIBUTION IN REPRESENTATIVE TOARCIAN SECTIONS

Taon codes are followed by range expressed in bed or leveI nurnben (Sources, additionai locality information, and notes are listed in Table 3.2, p. 35)

Section 1: Yakoun River Base: 19 Top: 130 Section 5: Djebel-es- OOBa Section 3: Valdorbia Saffeh (2) 15Dt Base: 1 Top: 20 Base: 60 Top: IO3 16D 20Gt OOBa 21G OlOBc 23Hi OlBb 27H O2Bg 30Hsp 0413 31H 05Cc 50Li 06Cg 5IL 07C 56M 090CI 64Pm 09Cg 66Pve 1oc 69P 18Fs 70Pc 230Hp 720Ph 27H Section 4: Djebel-es- 72Per 28Hf Saffeh (3B) 74Pp 29Hsx Base: IO Top: 54 76h 31H Section 6: Paghania 78P 39Hs OlOBc Base: 1 Top: 78 790P1 40H 02Bg 79P 41Ha 04B OOBa 93 Pf 42Hb 09Cg O 1Bb 96P 44Hcr 10C 04B 97Rp 45HI 13D oscc 46Hsp 27H 06Cg Section 2: Joan 47Hsl 31H 07C Lake 48Ht 42Hb 09Cg Base: 1 Top: 7 49H 45H1 1OC 50Li 46Hsp 13D 51L 47Hsl 27H 52Md 49H 28Hf 540Mt 52Md 2lH 54Mm 53Mh 37Hp 5SMu 54Mm 40H 56M 55Mu 46Hsp 58Na 56M 47Hsl 59N 59N 49H 60Ps 61P 52Md 61P 75Pro 53Mh 62Pd 78P 54Mm Section 10: Ricla and La Almunia Base: 55 Top: 307 Section 11: Aios Saint-Nicolas Base: 44 Top: 1 1 8

Section 7: Monte di Civitella Base: 1 Top: 32

OlBb 28-29 048 28-29 18Fs 11-11 27H 12-3 1 28Hf 24-26 31H 24-26 40H 3-16 42Hb 25-25 46Hsp 14-29 47Hsl 2-20 540Mt 21-21 54Mm 16-27 55Mu 17-25 56M 7-27 59N 1-18 69P 11-11 7lPel 9-9 72Per 28-3 1 77OPv 31-31 77Ps 31-31 Section 9: Camplong 78P 9-32 Base: I Top: 68 93Pf 19-30 94Pl 31-3 1 95Pr 15-24 96P 15-31

Section 8: St-Paul- des-Fonts Base: 1 Top: 18

OlBb 1616 03Bp 15-18 048 14-18 050Cd 12-16 07C 10-18 090Cl 14-15 091Cm 16-18 09Cg 7-1 1 10C 7-18 13D 1-3 l5Dt 17-17 Section 12: Ravenscar and Section 14: Wbitby Quebrada Yerbas Base: 16 Top: 78 Buenas Base: 7 Top: 1 O

Section 13: Section 16: Rio del Quebrada El Bolito Toro Base: 6 Top: 9 Base: 7 Top: 9

Section 15: Quebrada Larga Base: 8 Top: I 1 APPENDIX 2

CORRELATION TABLES USING UNITARY ASSOCIATIONS

(BioGraph OUTPUT)

First column: level or bed nurnber; second column: UA range (see Fig. 3.5 for definition of UA)

Section 1: Yakoun River

130: 38-38 129: 38-38 128: 38-38 127: 38-38 126: 38-38 125: 38-38 124: 38-38 123: 38-38 122: 38-38 121: 38-38 120: 38-38 119: 38-38 118: 38-38 117: 38-38 116: 38-38 Section 2: Joan 115: 38-38 Lake 114: 38-38 113: 38-38 112: 38-38 111: 38-38 1 IO: 38-38 109: 37-38 108: 37-38 107: 37-38 106: 37-38 105: 37-38 Section 3: Valdorbia 104: 37-38 103: 37-38 102: 37-38 101: 37-38 17: 2û-28 Section 6: Paghania 24: 10-15 16: 2627 23: 10-15 15: 26-27 7s: 10-10 14: 24-24 21: 10-15 13: 24-24 20: 10-15 12: 23-24 19: 10-15 11: 19-19 18: lû-15 IO: 18-18 Section 5: Djebei-es- 17: 10-15 9: 18-18 Saffeh (2) 16: 10-15 8: 14-14 15: 10-15 7: 14-14 14: 10-15 6: 12-13 13: 10-15 5: 7-8 12: 10-15 4: 7-7 11: 10-15 3: 4-4 IO: 10-15 2: 4-14 9: 10-15 1: 1-1 8: 10-15 7: 10-15 Section 4: Djebel-es- 6: 10-15 Saffeh (36) 5: lû-15 4: 10-15 3: 1615 2: 2-15 1: 2-2

Section 7: Monte di Civitelia Section 8: St-Paui- des-Fonts

Section 9: Camplong

Section 10: Rida and La Almunia Section II: Ains Saint-Nicolas

Section 12: Ravenscar and Whitby Section 15: Quebrada Larga Section 13: Quebrada El Bolito

Section 16: Rio del Toro Section 14: Quebrada Yerbas Buenas IIVIMUL Lvnwn~~uf~ TEST TARET (QA-3)

APPLIEO IMAGE. lnc - 1653 East Main Street