Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Lithofacies, provenance, and diagenesis of jura-cretaceous strata of the Northern Bowser Basin, British… Cookenboo, Harrison O. 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_fall_phd_cookenboo_harrison.pdf [ 10.18MB ]
JSON: 831-1.0052850.json
JSON-LD: 831-1.0052850-ld.json
RDF/XML (Pretty): 831-1.0052850-rdf.xml
RDF/JSON: 831-1.0052850-rdf.json
Turtle: 831-1.0052850-turtle.txt
N-Triples: 831-1.0052850-rdf-ntriples.txt
Original Record: 831-1.0052850-source.json
Full Text

Full Text

LITHOFACIES, PROVENANCE, AND DIAGENESIS OF JURA-CRETACEOUS STRATA OFTHE NORTHERN BOWSER BASIN, BRITISH COLUMBIAbyHARRISON OWEN COOKENBOOB.Sc., Duke University, 1981M.Sc., The University of British Columbia, 1989A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THEDEGREE OF DOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(Department of Geological Science)We accept this thesis as conformingto the required standardTHE UNI SITY OF BRITISH COLUMBIAOctober 1993© Harrison Owen Cookenboo, 1993In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department of  Geological SciencesThe University of British ColumbiaVancouver, CanadaDate^15 October 1993DE-6 (2/88)ABSTRACTLithofacies, provenance, and diagenetic studies of more than 3 km of Late Jurassic to mid-Cretaceous silicilastic sediments exposed in the northern Bowser Basin (northern British Columbia)record the tectonic development of the Canadian Cordillera. Strata are divided into the undivided BowserLake Group and overlying Currier, McEvoy, and Devils Claw formations Lithofacies include marinemudstone, coarsening upward mudstone, fining upward sandstone, coarsening upward sandstone, chertpebble conglomerate, and coal. Common lithofacies associations are interpreted as a progression (fromolder to younger) of shallow marine, lower delta plain, upper delta plain, and alluvial braid-plaindepositional environments. A subsidence model based on sediment compaction and isostatic loadaccounts for the necessary accommodation space.The composition of the sandstone suggests an obducted island arc and oceanic crust asprovenance. Three petrofacies have been identified by modal analysis of framework grains. Petrofacies 1(P1), which occurs in undivided Bowser Lake Group and Currier Formation strata, (Q tFL = 34-14-52;QmFLt = 9-14-77) is volcanic lithic rich with subequal to minor chert, minor monocrystalline quartz(generally <10%), and 10-25 % feldspar. Petrofacies 2 (P2) occurs in lower McEvoy Formation, andhas higher concentrations of chert reflecting a recycled component in the sandstones, but also retainssignificant portions of volcanics (QtFL = 62-5-33; QmFLt = 5-5-89). Petrofacies 3 (P3) occurs in theupper McEvoy and Devils Claw formations, and is chert rich like P2, with less volcanic lithics and asmall but significant portion of metamorphic lithic fragments (Q t =64 %, F=5%, L=31 %; Qm=7%,F=5%, Lt= 88 %). Paleocurrent directions indicate transport from northeast to southwest.Microprobe analysis of detrital chromian spinel accessory grains demonstrates alpine typeperidotite occurs in the provenance. No spinels typical of mid-ocean ridge or Alaskan type complexeswere found. The petrofacies and chromian spinel chemistry are consistent with a provenance from islandarc and marginal basin lithosphere obducted onto the western margin of North America.iiDiagenetic history of the sediments provides insight into depositional and post depositionalprocesses in the basin. Seven stages of cement paragenesis are recognized in the sandstones: 1) chlorite;2) illite; 3) kaolinite; 4) dead oil; 5) quartz; 6) chlorite dissolution; and 7) calcite. Estimatedprecipitation temperatures begin below 80°C for chlorite, and increase to approximately 100°C to200°C for quartz, and to above roughly 200°C for calcite. Fluid inclusions in quartz cements supportsuch temperature estimates. The succession of cements is interpreted to record replacement of originalconnate seawater by acid pore waters derived from organic matter maturation that were forced out bycompaction of the interbedded muds. Carbon isotopes in carbonate concretions from the mudstones areconsistent with formation during a late stage of methanogenesis. Oxygen isotopes from the sameconcretions suggest pore fluids in the muds at the time of formation were meteoric to brackish waterstypical of cool temperate climates. Organic maturation was modelled by using vitrinite reflectance valuesfrom interbedded coals and mudstones, and assuming progressive heating as inferred from the sandstonecement paragenesis. Results of the model indicate that a high paleogeothermal gradient, similar to someback arc basins, best explains the diagenetic history of the northern Bowser Basin.Constraints from the lithofacies, provenance, and pore water evolution studies suggest theBowser Basin began as a deep marine basin which was filled by sediment derived from island arc andmarginal basin crust obducted earlier onto the western margin of North America.i i iivTABLE OF CONTENTSABSTRACT^iiTABLE OF CONTENTS^ivLIST OF FIGURES ixLIST OF TABLES^ xivACKNOWLEDGEMENT xvPREFACE^ xviChapter 1: INTRODUCTION^ 1GEOLOGIC SETTING AND STRATIGRAPHY^ 4Stratigraphy in the study area^ 7REFERENCES^11Chapter 2: LITHOFACIES IN THE NORTHERN BOWSER BASIN RECORDS FILLING OF ADEEP MARINE BASIN AND SUBSEQUENT ISOSTATIC ADJUSTMENT, BRITISHCOLUMBIA, CANADA^ 14ABSTRACT^ 14INTRODUCTION 15FACIES DESCRIPTIONS^ 15UNDIVIDED BOWSER LAKE GROUP^17Marine mudstone facies (MM) 17Coarsening upward sandstone facies (CUS)^21Fining upward sandstone facies (FUS) 23Depositional environments of undivided Bowser Lake Group facies^23CURRIER FORMATION^25Black claystone facies (BC)^26Interbedded claystone, siltstone and very fine grained sandstone^27VFining upward sandstone facies^ 27Coarsening upward sandstones 28Chert pebble conglomerate^ 28Thick coal^ 32Depositional environments of the Currier Formation^ 32MCEVOY FORMATION^ 35Coarsening upward mudstone facies (CUM)^ 36Chert Pebble-Cobble Conglomerate^ 38Coal^ 38Depositional environments of the McEvoy Formation^ 40DEVILS CLAW FORMATION^ 41Fining upward chert pebble-cobble conglomerate (FUPCC)^44Depositional environments of the Devils Claw Formation 48SUMMARY OF DEPOS1TIONAL HISTORY^ 49CLIMATE^ 52SEDIMENT COMPACTION AND ISOSTATIC LOADING AS MECHANISMS OFBOWSER BASIN FORMATION^ 54REFERENCES^ 58Chapter 3: RECORD OF OROGENY IN MESOZOIC SANDSTONES OF THE CANADIANCORDILLERA^ 62ABSTRACT 62INTRODUCTION^63PREVIOUS WORK ON PROVENANCE^ 64COUNTING PROCEDURE^65Grain types counted 66RESULTS^ 69Petrofacies 1^ 71Provenance interpretation^ 74Petrofacies 2^75Provenance interpretation^ 77Petrofacies 3^ 77Provenance interpretation^ 80OTHER INFLUENCES ON PETROFACIES COMPOSITION^ 84PROVENANCE TERRANE TECTONICS^91DISCUSSION^91Bowser Basin tectonic implications^ 93Implications for Cordilleran tectonics 96Implications to rifting of Pangea and the opening of the Atlantic^100CONCLUSIONS^101REFERENCES 102Chapter 4: DETRITAL CHROMIAN SPINEL COMPOSITIONS USED TO RECONSTRUCTTHE TECTONIC SETTING OF PROVENANCE: IMPLICATIONS FOR OROGENY INTHE CANADIAN CORDILLERA^ 107ABSTRACT^ 107INTRODUCTION 108METHODS^ 112PETROGRAPHIC DESCRIPTIONS^ 114RESULTS^114Type of ultramafic source^ 115Origin of the alpine-type peridotite source^ 121PROVENANCE: CANDIDATES IN THE CANADIAN CORDILLERA^126CONCLUSIONS^129REFERENCES 131viChapter 5: DIAGENESIS AND PORE WATER EVOLUTION IN GROUNDHOG COALFIELDSTRATA, NORTHERN BOWSER BASIN, BRITISH COLUMBIA, CANADA^135ABSTRACT^ 135INTRODUCTION 136SANDSTONE CEMENT PARAGENESIS^ 137CHLORITE CEMENTATION STAGE 141Chlorite cementation stage pore water chemistry^141ILLITE CEMENTATION STAGE^ 146Mite Cementation Stage Pore Water Chemistry^ 147KAOLINITE CEMENTATION STAGE^ 147Kaolinite Cementation Stage Pore Water Chemistry^148OIL MIGRATION STAGE^ 148Oil migration stage pore water chemistry^ 148CHLORITE DISSOLUTION STAGE^ 152Chlorite dissolution stage pore water chemistry^ 153QUARTZ CEMENTATION STAGE^ 153Fluid inclusions in fracture-filling quartz cements^155Fluid inclusion implications to pore water evolution 157Quartz cementation stage pore water chemistry^ 159CALCITE CEMENTATION STAGE^ 160Calcite cementation stage pore water chemistry^ 161SUMMARY OF PORE WATER EVOLUTION^ 162PORE WATER EVOLUTION AND GREYWACKE GENESIS^ 164ECONOMIC IMPLICATIONS OF PORE FLUID EVOLUTION 165MUDSTONE PORE WATER EVOLUTION^ 166CARBON AND OXYGEN STABLE ISOTOPES^ 169viiviiiCOMPARISON OF EARLY DIAGENESIS IN SANDSTONES ANDMUDSTO1VES^ 179ORGANIC DIAGENESIS 180BASIN MATURATION MODELLING^ 180PARAMETERS OF THE FAVOURED PALEOGEOTHERMAL GRADIENT ANDBURIAL HISTORY MODEL^ 182RESULTS OF BASIN MATURATION MODMING^ 184IMPLICATIONS FOR DIAGENETIC HISTORY 187IMPLICATIONS OF MATURATION MODEL TO REGIONAL TECTONICS 187CONCLUSIONS^ 188REFERENCES 190Chapter 6: CONCLUDING REMARKS^ 195REFERENCES^ 200APPENDIX 1^ 202REFERENCES^ 215ixLIST OF FIGURESFigure 1-1: Location of the study area in the Bowser Basin and adjacent features of the CanadianCordillera.^ 5Figure 1-2: Location of stratigraphic exposures measured in this study within the northernBowser Basin.^ 6Figure 1-3: Stratigraphic column.^ 8Figure 2-1: Facies described in text.^ 16Figure 2-2: Geological map based on sections measured in this study.^ 18Figure 2-3: Marine mudstones are interbedded with coarsening upward sandstones in section 49,interpreted as a succession of repeatedly stacked delta front lobes.^19Figure 2-4: Bedding surface of marine mudstone facies from the undivided Bowser Lake Groupextensively burrowed by Helminthopsis.^ 20Figure 2-5: Normally graded sandy siltstone layers separated by blue-black shale. These layers aretypical of extensively burrowed, normally graded siltstone and sandstone layers from themarine mudstone facies. 22Figure 2-6: Cross-section of Currier Formation strata.^ 29Figure 2-7: Location map for Currier Formation cross-section (Fig. 2-6).^ 30Figure 2-8: Marine mudstones interfinger with coarsening upward sandstones capped by thickcoals in the lower Currier Formation (section 33 located east of the Klappan River).^31Figure 2-9: Four steps of delta lobe progradation, subsidence, inundation andre-establishment.^ 34xFigure 2-10: Cliff face exposure composed of mostly CUM facies, from the lower part ofsection 12 in the McEvoy Formation. Stratigraphic column for section 12 illustratedin figure 2-11.^ 37Figure 2-11: Stacked coarsening upward mudstones from section 12 in the McEvoy Formation. 39Figure 2-12: Cross-section of Devils Claw formation exposures described in text. Sectionlocations given in map figure 2-13.^ 42Figure 2-13: Location map for Devils Claw Formation cross-section (fig. 2-12)^43Figure 2-14: The upper 420 m of section 13 exposed on an east west trending ridge line westof the Skeena River.^ 47Figure 2-15: Diagrammatic representation of LDPFA depositional environments typical of theCurrier Formation.^ 51Figure 2-16: Thin section from fossil wood, showing abundant early wood and a sharp changeto late wood.^ 53Figure 2-17: Calculation parameters and results for isostatic load plus sediment compactionmodel of Bowser Basin tectonic development.^ 55Figure 3-1: Photomicrographs of typical volcanic lithic and chert grains with poikilitic calcitecement.^ 68Figure 3-2: Composition of Petrofacies 1. Detrital modes of nine grain types plotted againstpercent of sample.^ 72Figure 3-3: Petrofacies 2 composition.^ 76Figure 3-4: Petrofacies 3 composition.^ 78Figure 3-5: Metamorphic grain from P3, probably from either a phyllite or schist. Plane andcrossed polar transmitted light micrographs.^ 79Figure 3-6: Polycrystalline quartz grain of probable metamorphic origin.^ 81Figure 3-7: Muscovite grain (crossed polarized light) from Petrofacies 3.^ 82Figure 3-8: Relative stability of grain types.^ 88Figure 3-9: Evolution of provenance as interpreted from the succession of petrofacies.^90Figure 3-10: Ternary diagrams of sandstone composition.^ 92Figure 3-11: Model of Cordilleran tectonic development consistent with provenance history.^99Figure 4-1: Location of the study area in the Intermontane Belt of the Canadian Cordillera.^111Figure 4-2: Stratigraphic column of the northern Bowser Basin. Horizons sampled are markedwith an asterisk.^ 113Figure 4-3: Ternary plot of the major trivalent cations in chromian spinels.^118Figure 4-4: Plot of Mg/(Mg +Fe2+) against ratio of trivalent cations Fe 3 +/(Fe3 + +Al + Cr) fordetrital spinel data set compared to worldwide occurrences from Irvine (1974).^119Figure 4-5: Cr/(Cr +Al) versus Mg/(Mg+Fe 2+).^ 120Figure 4-6: Cartoon illustrating typical spinel compositions from different sea-floor (potentialalpine-type ophiolite) and continental crust origins.^ 122Figure 4-7: TiO2 versus Cr2O3 plot, demonstrating the restriction of high TiO2 values tosandstones from the McEvoy Formation (petrofacies 2).^ 125xix i iFigure 4-8: A model illustrating closure of marginal seas and fringing island arcs after initiationof rifting of the Atlantic Ocean in the Early Jurassic.^ 127Figure 5-1: Stages of sandstone cementation in the northern Bowser Basin.^139Figure 5-2: Authigenic chlorite coats on sandstone clasts of varying compositions in planepolarized light.^ 142Figure 5-3: Sketch of a microscope view of isopachous chlorite cement with earlier pyriteframboids.^ 143Figure 5-4: Phase diagram showing that chlorite precipitation is favoured by relatively highMg2+ and low H+ activities in solution compared to illite or kaolinite.^145Figure 5-5: Three successive cements in Bowser Basin sandstones occur as a result of changingpore water chemistry described in this phase diagram from Jahren and Aagaard (1989). 149Figure 5-6: Micro-Fourier transform infrared analysis plots of opaque material in pores andfractures of Bowser Basin sandstones demonstrates that the material is organic matterwith very low aliphatic component.^ 150Figure 5-7: Chlorite cemented sandstone saturated with reservoir bitumen: a) oil in intergranularporosity; b) oil in chlorite cemented microfracture.^ 151Figure 5-8: Authigenic quartz has filled a pore lined first by isopachous chlorite cement and thencoated by oil.^ 154Figure 5-9: Figure 5-9: a) Hydrocarbon fluid inclusions in a single crystal of quartz cement froma fracture in silicified wood.; and b) opaque bitumen between quartz crystal surfaces.^158Figure 5-10: Four stages of pore water chemistry shown as a function of changing alkalinitydue to organic maturation and cation concentrations.^ 163Figure 5-11: Range of concretion d180 ratios in siderite cement and their equilibrium formationtemperature assuming three different pore water isotope composition models.^173Figure 5-12: Range of concretion d 180 ratios in ferroan dolomite cement and their equilibriumformation temperature assuming three different pore water isotope composition models. 174Figure 5-13: Comparison of predicted temperatures for siderite, ferroan dolomite and calciteconcretion cements, assuming different modelled pore water compositions.^176Figure 5-14: Range of concretion d180 ratios from calcite cement and their equilibrium formationtemperature assuming three different pore water isotope composition models.^178Figure 5-15: Vitrinite reflectance increases significantly with stratigraphic depth, implying a steepthermal gradient.^ 181Figure 5-16: Burial history curve, assuming ages, stratigraphic thickness, and burial depths offavoured maturation model.^ 185Figure 5-17: Maturation model runs with assumed paleogeothermal gradients of 40 °C/lcm,50°C/km, and 65°C/km for three stratigraphic horizons.^ 186Figure A1-1: The location of the Currier Formation type section. Map A is from MacLeod andHills (1990a) where the Currier type section plots in what are called "McEvoy" strata,and map B is a detail from field mapping by Moffat (1985). 204Figure A1-2: Stratigraphic nomenclature used by various authors for the coal-bearing sectionof the northern Bowser Basin.^ 206Figure A1-3: The upper Jurassic ammonite Amoeboceras recovered from the coal-bearing sectionin the northern Bowser Basin (1.5x).^ 208x ivLIST OF TABLESTable 3-1: Point count results for sandstones from the northern Bowser Basin.^70Table 4-1: Representative analyses of detrital spinels from the northern Bowser Basin.^116ACKNOWLEDGEMENTCompletion of this dissertation has only been possible due to the cooperative and indulgentefforts of numerous friends and colleagues. First, I must acknowledge the support provided by my thesissupervisor Dr. R. M. Bustin, who untiringly read and re-read both this and my earlier Master's thesis.Next, I am happy to have an opportunity to thank M. Dawson of the Institute of Sedimentary andPetroleum Geology, B. Ryan of the British Columbia Ministry of Mines and Energy Resources, E.Swanbergson of Gulf Minerals, and G. Cave of Esso Minerals, who graciously provided logisticalsupport for parts of my field work, and helpfully discussed vagaries of Bowser Basin coal occurrences.Finally, I thank inclusively all the staff (especially Brian Cranston), faculty and my fellow graduatestudents who provided advice, assistance, and instruction too many times to acknowledge individually,but without whom my research would have been stymied.Partial funding for this research originated from NSERC (grant A7337 to R. M. Bustin), theGeological Survey of Canada, Esso Minerals, and the British Columbia Ministry of Energy, Mines andPetroleum Resources, and we thank each organization for their support. Electron microprobe analyseswere made at the University of British Columbia, using facilities supported by NSERC infrastructuregrant #9924. Valuable insights into the chemistry of chromian spinel were provided by conversationswith J. K. Russell, G. T. Nixon, and P. R. Roeder. J. Grigsby, A. R. Basu, and an anonymousreviewer provided constructive comments to an earlier version of the detrital chromian spinel chapter, forwhich I am particularly grateful.Dedicated and intrepid field assistance was provided by C. Bryan, M. Gant, E. Bergeson and A.Toma, and equally dedicated lab assistance and organization was provided by D. JacklinXVPREFACEPresentation of this dissertation is in the format of four stand-alone chapters, following common practiceat the University of British Columbia and most other universities. In order to reduce the repetition that isinevitable with this format, the geologic setting and stratigraphy sections have been deleted from chapters2, 3, 4 and 5, and are combined under one heading in the introductory chapter (chapter 1). Inaccordance with the University of British Columbia regulations, I briefly describe below my contributionto published portions of this thesis. Chapter 4, in slightly edited form, has been submitted to the Journalof Sedimentary Petrology for review under the title "Detrital chromian spinel compositions used toreconstruct the tectonic setting of provenance: implications for orogeny in the Canadian Cordillera" byH. 0. Cookenboo, R. M. Bustin, and K. R. Wilks. As principal author, I attest that I am responsible forall aspects of the concept, data collection, interpretation and writing of the article, with the exception ofrunning the microprobe during data collection, which was done by Dr. Wilks.Appendix 1 consists of a formal discussion of MacLeod and Hills (1990). This discussionentitled "Conformable Late Jurassic (Oxfordian) to Early Cretaceous strata, northern Bowser Basin,British Columbia: a sedimentological and paleontological model: Discussion." was published in 1991 inthe Canadian Journal of Earth Sciences, (volume 28, p. 1497-1502) with Drs. R. M. Bustin and I. W.Moffat as co-authors. As senior author, I wrote the discussion, conceived the arguments, and drew theconclusions. Drs. Moffat and Bustin initiated stratigraphic work in the Groundhog Coalfield thatultimately led to our age constraints, including collection of the pivotal ammonite data mentioned in thediscussion, and additionally provided valuable editorial contributions to the final document.CHAPTER 1INTRODUCTIONSediments are multifaceted records of geologic history. Intrabasinal conditions prior to andduring sedimentation are preserved by the succession of lithofacies and depositional environments. Forexample, cratonic basins commonly begin with alluvial sequences followed by lacustrine, evaporite andshallow to deep marine successions, depending on extent of rifting, climate, and rate of sedimentaccumulation. Extrabasinal conditions are reflected in the provenance record, which depends largely onsediment composition and paleoclimate, both preceding and during deposition. Syn- to post- depositionaltectonic conditions within the basin are recorded by authigenic minerals, organic maturation, andstructural deformation.The broad range of geologic data available from sediments makes their study indispensable tounderstanding evolution of the Canadian Cordillera. Sediments most relevant to orogeny in the CanadianCordillera are those deposited within or adjacent to the Cordillera, during orogeny and accretion. Suchsyn-orogenic basins occurred in the Intermontane Belt of the Canadian Cordillera, where they are locatedbetween the two principal metamorphic and plutonic belts, the Coast Plutonic Complex to the west andthe Omineca Belt to the east. The largest of the Intermontane Belt basins is the Bowser Basin of northernBritish Columbia, which is the subject of this thesis.The problem addressed by this thesis is to use detailed sedimentology of northern Bowser Basinstrata to constrain postulated accretionary and orogenic processes that resulted in formation of theCordillera. The basis of this study are forty-nine stratigraphic sections that were measured and describedin the vicinity of the Groundhog coalfield in the northern Bowser Basin. Samples collected from thosesections were later analyzed in the laboratory by methods including transmitted and reflected lightmicroscopy, scanning electron microscopy, electron microprobe, fluid inclusion microthermometry,micro Fourier transform infrared spectroscopy, x-ray diffraction, stable isotope geochemistry, vitrinitereflectance, and computer modelling of organic maturation. Three principal approaches were employed1in this study: 1) description of lithofacies and interpretation of depositional environments leading to aplausible model of basin origin and development; 2) quantification of sandstone composition leading to adescription of provenance and a model of tectonic origin for the sediment source rocks; and 3) diageneticstudies to constrain pore water evolution, leading to a well-constrained thermal maturation model for thebasin.Lithofacies in the northern Bowser Basin are described in detail in Chapter 2. The lithofaciesare derived from detailed measurement and description of 49 stratigraphic sections selected on the basisof exposure and continuity as determined from air photos. The depositional environment model buildsupon earlier stratigraphic work which established a general trend to shallower marine deposition upwardfrom the base of the Bowser Basin strata, into overlying deltaic and alluvial plain strata of the GroundhogCoalfield (Malloch, 1914; Buckham and Latour, 1950; Eisbacher, 1974; Bustin and Moffat, 1983;Moffat, 1985; Cookenboo, 1989; Ricketts, 1990). In this study, lithofacies and lithofacies associationsare interpreted to represent marine shelf, prodelta, delta, and alluvial plain depositional environments. Inaddition, the succession of lithofacies associations indicates a depositional history which involved a deepmarine basin filled first by progressively shallower marine deposits, and subsequently by near sea-leveldeposition of deltaic, delta plain, and alluvial braid-plain sediments.Chapter 3 describes the methods and results of a provenance study based on modal analysis ofthe sandstones. This study extends provenance interpretations originating in the early part of thiscentury, which associated the abundant chert pebbles with chert rich strata of the Cache Creek terranelocated north of the Bowser Basin (Malloch, 1914; Eisbacher, 1981). The most significant observationin this study is that the sandstones are lithic rich, changing from dominantly volcanic lithic fragments inolder strata, to dominantly chert in overlying younger strata. These changing compositions are split intothree petrofacies, and each petrofacies is interpreted in terms of likely composition of its provenance.Petrofacies 1 was derived from obducted island arc and oceanic crust. Petrofacies 2 was derived fromsimilar source rocks, but increased chert content suggests an additional component of recycled sediments.Petrofacies 3 is interpreted as derived from obducted island arc and oceanic crust, with a large recycled2component and an additional small but significant component of metamorphic and plutonic source.A question remaining from the framework grain modal analysis is - what kind of oceanic crust isincluded in the provenance? This question is addressed in Chapter 4 by using microprobe analyses ofdetrital chromian spinels that occur as accessory grains in most northern Bowser Basin sandstones.Chromian spinel compositions are distinct for a number of ultramafic petrogenetic origins, includingstratiform complexes, Alaskan type peridotites, back-arc basins, and mid-ocean ridges. Detrital spinelcompositions in Bowser Basin sandstones are most compatible with an unmetamorphosed island arc andmarginal basin crust origin, and suggest that mid-ocean ridge or Alaskan-type complexes did notcontribute (at least significantly) to the northern Bowser Basin.Chapter 5 describes pore water evolution in northern Bowser Basin sediments based onobservations of sandstone cement paragenesis, mudstone concretion mineralogy, and stable isotopegeochemistry. Previous analyses of basin maturation have centred on vitrinite reflectance studies (Bustinand Moffat, 1983; Moffat, 1985; Bustin, 1984; Bustin and Moffat, 1989) which yield interpretedmaximum temperatures, but are ambiguous as to whether maximum temperatures were reached early as afunction of progressive heating during burial, or later in response to possible tectonic events. Diageneticobservations in this study enhance the maturation models based on vitrinite reflectance alone byclarifying thermal history and pore water evolution prior to the sediments reaching maximumtemperatures. Seven stages of sandstone cement paragenesis were found: 1) chlorite; 2) illite; 3)kaolinite; 4) dead oil; 5) quartz; 6) chlorite dissolution; and 7) calcite. From these sandstone cementstages, pore water evolution is interpreted as initially saline to brackish marine water being replaced firstby acid waters and migrating oil associated with organic maturation, and later by alkaline watersfollowing thermal breakdown of organic acids at roughly 200°C. The successive cement stages andinterpreted pore water evolution are consistent with progressive heating during burial to temperatures inexcess of 200°C. Geochemistry of mudstone cements, as preserved in carbonate concretions, suggestsinterbedded organic rich muds were the source of the acid pore waters, derived probably by dewateringduring burial compaction. Thermal maturation modelling consistent with the pore water history and3vitrinite reflectance values suggests burial in a very high paleogeothermal gradient regime (60° to 65°C/km).GEOLOGIC SETTING AND STRATIGRAPHYSedimentary strata examined in this study were deposited in the Bowser Basin, which is thelargest basin in the Intermontane Belt of the Canadian Cordillera. The Bowser Basin is located innorthern British Columbia between the two principal metamorphic and plutonic belts of the CanadianCordillera, the Coast Plutonic Complex to the west and the Omineca Belt to the east (Fig. 1-1). Thestudy area is located in the northern part of the Bowser Basin, approximately coincident with thetraditional and informal limits of the Groundhog Coalfield (Fig. 1-2). The sediments were depositedduring the Middle Jurassic through Early (and perhaps earliest Late) Cretaceous, and today are bound onall sides by older volcanic and plutonic rocks except at the eastern margin where younger sediments ofthe Sustut Basin cover the Bowser Basin. The northern and southern limits of Bowser Basin clastics arethe Stikine (north) and Skeena (south) arches. Although the present limits of the Basin are definedsharply, only an eastern boundary is known during deposition, and the basin may have been open to thewest (Ricketts, 1990).Clastic sedimentation in the Bowser Basin was initiated during the Middle Jurassic with deep-water turbidites and shales of the Bowser Lake Group (Tipper and Richards, 1976). Progressivelyshallower marine facies were deposited until the Late Jurassic when interbedded shallow marine, deltaic,and fluvio-deltaic facies began accumulating. The youngest sediments preserved in the Bowser Basin arefluvial and fluvio-deltaic strata deposited during the Early to mid-Cretaceous. The minimum totalthickness of Bowser Basin strata, calculated by adding data in Ricketts (1990) to measurements from thestudy area is in excess of 5 kilometres (Cookenboo and Bustin, 1990). The true total thickness couldsignificantly exceed this minimum estimate, due to complex structural development following4Figure 1-1: Location of the study area in the Bowser Basin and adjacent features of the CanadianCordillera.5Figure 1-2: Stratigraphic exposures measured in this study are located near the Groundhog Coalfield,within the northern Bowser Basin.6deposition. These sediments were subsequently buried by as much as three to five kilometres (Bustin,1984) of now eroded rocks. The eroded beds were presumably at least in part equivalent to UpperCretaceous and Tertiary alluvial strata preserved east of the Bowser in the Sustut Basin. An angularunconformity separates Sustut rocks from clastic and volcaniclastic rocks equivalent to the lower part ofthe Bowser Basin section (Eisbacher, 1974).Structural deformation in the Bowser Basin has included at least two and possibly three phasesof deformation beginning in the Early to mid-Cretaceous (Moffat, 1985; Moffat and Bustin, 1993). Thiscomplex deformational history has led to an estimated 35% to 45 % shortening across the basin (Moffat,1985; Evenchick, 1991).Basement rocks are not exposed in the study area. Along the southern margin of the basin,however, thick island are volcanic sequences of the Upper Triassic Takla Group and Lower to MiddleJurassic Hazelton Groups underlie elastic marine strata which comprise the base of the Bowser Basinsection (Tipper and Richards, 1976). Hazelton Group volcanics are also exposed along the northeasternbasin margin where they are separated from the elastic Bowser Basin strata by a relatively thin (<250metres) volcaniclastic and thin bedded siltstone unit termed the Spatsizi Group (Thomson et al. , 1986).Pillow basalts and basinal sediments of the Lower to Middle Jurassic Salmon River Formation and olderstrata are exposed at the western margin of the basin (Lewis et al., 1993).A window to older basement rocks exists at Oweegee Dome near the western margin of theBowser Basin (Fig. 1-1), where Greig (1991) reports Paleozoic limestones and cherts exposed beneathHazelton and earlier Mesozoic volcanic strata. No older basement is known.Stratigraphy in the study areaFour lithostratigraphic units are recognized in the study area (Cookenboo and Bustin, 1989).From oldest to youngest, they are undivided Bowser Lake Group, and the Currier, McEvoy and DevilsClaw Formations (Fig. 1-3). Formal definitions and detailed descriptions of the Currier,7Figure 1-3: Stratigraphic column. Lithostratigraphic units used in this study are listed under the heading"Northern Bowser Basin".8McEvoy, and Devils Claw Formations are provided in Cookenboo and Bustin (1989), with extensionsand further information resulting from three subsequent years of fieldwork described in Cookenboo andBustin (1990a), Cookenboo and Bustin (1990b), and Cookenboo and Bustin (1991). Summarydescriptions of the stratigraphic units and nomenclature used in this thesis are given below, based onintegration of Cookenboo and Bustin (1989) with subsequent years fieldwork as described in thefieldwork summaries mentioned above Cookenboo and Bustin (1990a), Cookenboo and Bustin (1990b),and Cookenboo and Bustin (1991) and Cookenboo (1989). Age of the stratigraphic units are derivedfrom data presented in Bustin and Moffat (1983), Moffat (1985), Moffat et al. (1988), Cookenboo andBustin (1989), MacLeod and Hills (1990) and Cookenboo (1989). A fuller discussion of the ages ispresented in Cookenboo et al. (1991: Appendix 1), but conservatively the McEvoy can be called EarlyCretaceous and possibly in part middle Albian in age.The base of the exposed section in the sampled area is termed undivided Bowser Lake Group inthis thesis, and consists of a dark grey to black marine mudstone unit that coarsens in the upper part toinclude prodelta sandstones and siltstones. Thickness is poorly constrained because the base is notknown, but estimates range upwards of 2000 metres (Moffat, 1985). The upper part of this unit containsBuchia concentrica, which suggests an age of late Oxfordian to early Kimmeridgian (Poulton et al.,1991). Progressively older marine beds north of the study area suggest that undivided Bowser LakeGroup is as old as the Middle Jurassic (Poulton et al., 1991). Similar dark grey to black marinemudstones of the Ashman Formation in the southern Bowser Basin span the late Bajocian to the top of theearly Oxfordian (Tipper and Richards, 1976). The undivided Bowser Lake Group in the study area islargely equivalent in lithology and age to the Ashman Formation, but because of inconsistencies in usageof the term 'Ashman Formation' in different parts of the basin (Cookenboo and Bustin, 1991), and a lackof demonstrated continuity with the Ashman Formation strata in the southern Bowser Basin, the lessspecific name 'undivided Bowser Lake Group' is used in this thesis.The Currier Formation consists of as much as 1000 m of interbedded shales, siltstones,sandstones, coals and authigenic carbonate layers (Cookenboo and Bustin, 1989; 1990a) that overlie the9marine beds of the undivided Bowser Lake Group. The Currier Formation is exposed from south ofMaitland Creek at the northeast edge of the study area to the Groundhog Range in the south (Fig. 1-2).Based on the ammonite Amoeboceras collected from the lower portion of the coal bearing section (Moffatet al., 1988), and Late Jurassic dinoflagellate suites (Rouse, written communication, 1989) the CurrierFormation is suggested to be in part Late Jurassic. MacLeod and Hills (1990) suggest the CurrierFormation may be in part earliest Cretaceous as well, based on their recovery of latest Jurassic species ofBuchia from beneath coal-bearing strata, and probably Neocomian the astariid bivalve Herzogina fromwithin the formation.The McEvoy Formation consists of 600 m to approximately 1000 m of fluvio-deltaic siltstonesand shales with minor sandstones, thin coals, authigenic carbonate layers, and conglomerates (Bustin andMoffat, 1983; Cookenboo and Bustin, 1989; Cookenboo and Bustin, 1990a; Cookenboo and Bustin,1990b; and Cookenboo and Bustin, 1991). The McEvoy Formation is exposed between the Skeena andNass rivers from south of Mount Klappan in the north to Devils Claw Mountain in the south. Strataexposed south of Sweeney Creek and between the Nass River and Konigus Creek are also assigned to theMcEvoy Formation, and in this area the formation may be as much as 1500 m thick Cookenboo andBustin, (1990b). The McEvoy lacks marine body fossils but has been dated based on palynoassemblagesas late Barremian or Aptian to late middle Albian (Moffat et al., 1988; Cookenboo and Bustin, 1989).Plant macrofossils have been used recently to suggest an earlier (pre-Albian) Cretaceous age (MacLeodand Hills, 1990) for the McEvoy Formation.The Devils Claw Formation consists of up to 600 m of fluvio-deltaic and coastal plain stratacharacterized by thick pebble and cobble conglomerates, with associated coarse sandstones and minorsiltstones and shales. Exposure of the Devils Claw Formation is limited to the area the Skeena and Nassrivers south of Mount Klappan and north of Devils Claw Mountain (Fig. 1-2). The conglomeratic strataof the Devils Claw overlie the McEvoy Formation, and are Albian (Early Cretaceous) to perhaps earliestLate Cretaceous (Cenomanian; Moffat et al., 1988; Cookenboo and Bustin, 1989).10No younger strata are preserved above the Devils Claw Formation. To the east and south east ofthe study area, alluvial strata of the Sustut Group (reported age range of mid-Cretaceous to Paleocene)unconformably overlie rocks of the Bowser Lake Group (Eisbacher, 1974).REFERENCESBuckham, F., and Latour, B. A., 1950. The Groundhog coalfield, British Columbia. Geological Surveyof Canada, Bulletin 16, 82 p.Bustin, R. M., 1984. Coalification levels and their significance in the Groundhog Coal field, north-central British Columbia. International Journal of Coal Geology, v. 4, p. 21-44.Bustin, R. M., and Moffat, I. 1983. Groundhog Coalfield, British Columbia: reconnaissance stratigraphyand structure. Bulletin of Canadian Petroleum Geology, 31: 231-245.Bustin, R. M., and Moffat, I. W., 1989. Semianthracite, anthracite and meta-anthracite in the centralCanadian Cordillera: their geology, characteristics and coalification history. InternationalJournal of Coal Geology, v. 13, p. 303-326.Cookenboo, H. 0., 1989. Lithostratigraphy, palynostratigraphy, and sedimentology of the northernSkeena Mountains and their implications to the tectonic history of the Canadian Cordillera.MSc. thesis, University of British Columbia, Vancouver British Columbia. 131 p.Cookenboo, H. 0., and Bustin, R. M., 1989. Jura-Cretaceous (Oxfordian to Cenomanian) stratigraphyof the north-central Bowser Basin, northern British Columbia: Canadian Journal of EarthSciences, v. 26, p. 1001-1012.Cookenboo, H. 0., and Bustin, R. M. 1990a. Stratigraphy of coal occurrences in the Bowser Basin. inGeological Fieldwork 1989. British Columbia Ministry of Energy, Mines and PetroleumResources, Paper 1990-1, pp. 473-477.Cookenboo, H. 0., and Bustin, R. M. 1990b. Lithostratigraphy of the northern Skeena Mountains,British Columbia: Current Research, Part F, Geological Survey of Canada Paper 90-1F, p. 151-156.Cookenboo, H. 0., and Bustin, R. M. 1991. Coal-bearing facies in the northern Bowser Basin (104A,H): British Columbia Ministry of Energy, Mines and Petroleum Resources, GeologicalFieldwork 1990, Paper 1991-1, p. 415-418.11Cookenboo, H. 0., Bustin, R. M., and Moffat, I. W., 1991. Conformable Late Jurassic (Oxfordian) toEarly Cretaceous strata, northern Bowser Basin, British Columbia: a sedimentological andpaleontological model: Discussion. Canadian Journal of Earth Sciences, v. 28, p. 1497-1502.Eisbacher, G. H., 1974. Sedimentary and tectonic evolution of the Sustut and Sifton Basins, north-central British Columbia. Geological Survey of Canada Paper 73-31, 57 p.Eisbacher, G., 1981. Late Mesozoic - Paleogene Bowser Basin molasse and Cordilleran tectonics,western Canada, In Miall, A. D. (ed.), Sedimentation and Tectonics. Geological Association ofCanada, Special Paper 23, p. 125-151.Evenchick, C. A., 1991. Geometry, evolution and tectonic framework of the Skeena fold belt, northcentral British Columbia. Tectonics, v. 10, p. 527-546.Greig, C. J., 1991. Stratigraphic and structural relations along the west-central margin of the BowserBasin, Oweegee and Kinskuch areas, northwestern British Columbia. Geological Survey ofCanada Current Research Part A, Paper 91-1A, P. 197-205.Lewis, P. D., Thompson, J. F. H., Nadaraju, G., R. G. Anderson, and G. G. Johnson, 1993. Lowerand Middle Jurassic stratigraphy in the Treaty Glacier area and geological setting of the TreatyGlacier alteration system, northwestern British Columbia. Geological Survey of Canada CurrentResearch Part A, Paper 93-1A, p. 75-86.MacLeod, S. E., and Hills, L. V. 1990. Conformable Late Jurassic (Oxfordian) to Early Cretaceousstrata, northern Bowser Basin, British Columbia: A sedimentological and paleontological model.Canadian Journal of Earth Sciences, v. 27, p. 988-998.Malloch, G. S., 1914. Groundhog coal field. Summary report of the Geological Survey Department ofMines, p. 69 to 101.Moffat, I. W. 1985. The nature and timing of deformational events and organic and inorganicmetamorphism in the northern Groundhog Coalfield: implications for the tectonic history of theBowser Basin. Ph.D thesis, University of British Columbia, Vancouver, B.C.Moffat, I. W., and Bustin, R. M., 1993. Deformational history of the Groundhog Coalfield,northeastern Bowser Basin, British Columbia: styles, superposition and tectonic implications.Bulletin of Canadian Petroleum Geology, v. 41, p. 1-16.12Moffat, I. W., Bustin, R. M., and Rouse, G. E. 1988. Biochronology of selected Bowser Basin strata:tectonic significance. Canadian Journal of Earth Sciences, v. 25 p. 1571-1578.Poulton, T. P., Callomon, J. H., and Hall, R. L., 1991. Bathonian through Oxfordian (Middle andUpper Jurassic) marine macrofossil assemblages and correlations, Bowser Lake Group, west-central Spatsizi map area, northwestern British Columbia. Current Research, Part A, GeologicalSurvey of Canada Paper 91-1A, p. 59-63.Ricketts, B. D. 1990, A preliminary account of sedimentation in the lower Bowser Lake Group, northernBritish Columbia, In Current Research, Part F, Geological Survey of Canada, Paper 90-1F, p.145-150.Thomson, R. C. Smith, P. L. and Tipper, H. W. 1986. Lower to Middle Jurassic (Pliensbachian toBajocian) stratigraphy of the northern Spatsizi area, north-central British Columbia. CanadianJournal of Earth Sciences, v. 23, p. 1963-1973.Tipper, H. W. and Richards, T. A., 1976. Jurassic stratigraphy and history of north-central BritishColumbia. Geological Survey of Canada, Bulletin 270, 73 p.13CHAPTER 2LTTHOFACIES IN THE NORTHERN BOWSER BASIN, BRITISH COLUMBIA: FILLING OF ADEEP MARINE BASIN AND SUBSEQUENT ISOSTATIC ADJUSTMENTABSTRACTThick Jura-Cretaceous siliciclastic strata of the northern Bowser Basin contain a record of Cordilleranorogeny that can be understood, in part, by analyzing their lithofacies and reconstructing depositionalhistory. The strata examined in this study belong to four stratigraphic units exposed in the vicinity of theGroundhog coalfield. The units are, from oldest to youngest, undivided Bowser Lake Group strata, andthe Currier, McEvoy, and Devils Claw formations. Lithofacies recognized in this study are defined ongrain size, biotic features, bedding geometry, sedimentary structures, and composition (forconglomerates). They consist of marine mudstone, black to dark gray claystone, coarsening upwardmudstone, fining upward sandstone, coarsening and thickening upward sandstone, chert pebbleconglomerate, and coal. Associations of lithofacies differ in each of the four stratigraphic units exposedin the northern Bowser Basin: 1) marine facies association (undivided Bowser Lake Group); 2) lowerdelta plain facies association (Currier Formation); 3) upper delta plain facies association (McEvoyFormation) and 4) alluvial braid plain facies association (Devils Claw Formation). As the designationsimply, these four lithofacies associations are interpreted as having genetic significance, and the verticalsuccession reflects development of the basin.The succession of lithofacies associations records filling of a pre-existing deep marine basin,first with thick basinal and slope sediments (exposed outside the study area), followed by more than 3000m of 'near sea-level' (shallow marine to deltaic to coastal plain) sediments. Calculation of compactionand lithostatic loading effects demonstrate that merely filling a pre-existing marine basin more than 3000m deep is sufficient mechanism to explain the origin of the Bowser Basin and subsequent creation ofaccommodation space necessary for the entire near sea-level stratigraphic column. By this model,compaction and lithostatic loading of the marine basin fill create accommodation space near sea-level thatis filled first by shallow marine shelf sediments, later by delta plain sediments, and finally by alluvialbraid plain sediments. This constitutes a sediment driven model that requires no other tectonic factors,14such as thrust fault loading, to explain the subsidence of the basin.INTRODUCTIONSediment accumulation closely reflects tectonic setting. Cratonic sedimentary basins tend toaccumulate shallow water and/or subaerial sediments relatively slowly. Many foreland basins accumulatethick shallow marine to alluvial sediments in response to tectonic thrust loading. Continental marginbasins, in contrast to both foreland and cratonic basins, may accumulate siliciclastic sediments indepositional environments as diverse as deep ocean fans to subaerial alluvial fans as rapidly as sedimentsupply allows. Because sedimentation is directly controlled by tectonic conditions, understandinglithofacies and lithofacies successions may permit inferences to be made concerning tectonic controls onthe basin.The purpose of this chapter is to provide lithofacies descriptions and interpretations ofdepositional environments for the siliciclastic succession that in turn forms the basis of inferences aboutthe tectonic setting of the basin.FACIES DESCRIPTIONSMeasured exposures were divided into facies based on their lithologic and biologic components,and some of the facies were further divided into subfacies (Fig. 2-1). The facies and subfacies aredescribed below for each of the four stratigraphic units in the study area (Chapter 1). Because somefacies occur in more than one unit some repetition is inevitable. An interpretation of the depositionalhistory follows the facies descriptions for each unit.Terminology used in descriptions of fine-textured sediments vary widely in the literature. Inthis study, terminology is based on Blatt et al., (1972) and Collinson and Thompson (1989). Mudstoneis applied to lithified sediments in which clays and silts are mixed between 1/3 and 2/3, or moregenerally in unspecified amounts. Claystone is applied more precisely to lithified sediments dominated15Fining Upward Chat Pebble-Cobble Conglomerate (FUCPCC)Coarsening Upward Mudstone (CUM)Black claystone (BC homogeneous + carbonaceous)Thin coalCoarsening Upward Mudstone (CUM)Black daystone (BC homogeneous + carbonaceous)Fining Upward Sandstone (FUS)Coarsening and Thickening Upward Sandstone (CUS)Chert Pebble-Cobble Conglomerate (CPCC)Thin coalBlack daystone (BC laminated + homogeous + carbonaceous)Fining Upward Sandstone (FUS)Coarsening and Thickening Upward Sandstone (CUS)Interbedded claystone, siitstone and very One sandstone (ICSS)Thick coalChat Pebble Conglomerate- well sorted (CPCws)^—17; Marine Mudstone (MM)Coarsening and Thickening Upward Sandstone (CUS)Fining Upward Sandstone (FUS)Figure 2-1: Facies described in text (subfacies in parentheses). Within each formation heading, the faciesare listed in approximate order of decreasing abundance.16by ( > 2/3) clay, and siltstone is applied to lithified sediments that appear in the field to be dominated by(>2/3) silt.UNDIVIDED BOWSER LAKE GROUPUndivided Bowser Lake Group crops out surrounding the Groundhog coalfield, and wasexamined in 10 sections (Fig. 2-2), including four sections in which the contact with the overlyingCurrier Formation was encountered. The thickest apparently continuous stratigraphic section measuredwas 617 m (section 49 located south of the Klappan River Fig. 2-2 and 2-3). The total thickness of theundivided Bowser Lake Group is poorly constrained because no base to the unit occurs in the study area.Similar facies that probably correlate with those described below occur to the north of the study area(Ricketts, 1990; Poulton et al, 1991), west of the Klappan River (Evenchick and Green, 1990), atNotchtop Peak in the Slamgeesh Mountains (Jeletzky, 1976; Cookenboo and Bustin, 1990), and alongthe southern margin of the Bowser Basin (Tipper and Richards, 1976; Richards and Jeletzky , 1974).The age of the unit ranges from late Bajocian to the end of the early Oxfordian in the southern part of theBowser Basin (Tipper and Richards, 1976), and to as young as late Oxfordian or early Kimmeridgian inthe northern part of the basin (Poulton et al., 1991).The undivided Bowser Lake Group strata exposed in the study area are characterized by threefacies described below (Fig. 2-1). The dominant facies are marine mudstone and coarsening andthickening upward sandstones. The third and relatively minor facies is fining upward sandstone.Marine mudstone faciesThe marine mudstone (MM) facies consists of recessive weathering, dark blue-black claystone;silty claystone in light to medium brown laminae; and thin interbeds of siltstone or very fine sandstone.The MM facies is typically lOs to 100s of metres thick, and is pervasively burrowed by a variety of tracefossils (abundant Helminthopsis (Fig. 2-4), with less abundant Teichichnus, Chondrites, Zoophycus,Skolithos, and Rhizocorallium). The brown laminae are normally graded from 3 to 10 cm thick, with the17128° -^12_OUTCROP pATTERN , ,57° 00'- -gE . 0N8525'pC8etCDCDCD40E1litIVRENV CUS thickMMthickthick^ CUS thick^ CUS thick:f- NggezimaiV^CUS thickMMsteel^CUS thickMMCUS thickMMCUS thickMMMMMMFINING UPVARD SANDSTONE - 3 stacked tia-biditesMM3 m dike trending 001MMCUS• " huinMotity CUSsection 49O nFACIES700 n-6000 In500 rn —400 rn —300 ri200 rn100 n —PRO—DELTA LOBE(arrow length proportionalto coarsening upwardunit thickness)Figure 2-3: Marine mudstones are interbedded with coarsening upward sandstones in section 49, aninterpreted succession of repeatedly stacked prodelta to delta front lobes.19MI2^3Figure 2-4: Bedding surface of marine mudstone facies from the undivided Bowser Lake Groupextensively burrowed by Helminthopsis.20basal portion (up to 50%) of each layer composed of sharp based, normally graded sandstone or siltstone,which locally is rippled or finely cross laminated. The basal sandstone or siltstone is covered byclaystone, which is commonly thoroughly bioturbated by Helminthopsis. A general trend in the MMfacies is change from pervasively burrowed mudstone in stratigraphically low exposures to mudstonewith fewer trace fossils higher in the section. As intensely burrowed mudstones become rare, rippleswithin interbedded sandstones become abundant, and hummocky cross beds occur rarely. Normallygraded silt- or sandstone layers are regularly spaced within each marine mudstone bed, but spacing variesfrom one normally graded layer per 2 metres to as many as five in 15 cm in different areas. Thethickness of the normally graded layers is also nearly constant within a single bed, but varies from thinlaminae less than 0 5 cm thick to thin beds as much as 5 cm thick in different areas. Interbeddednormally graded layers are generally thicker and more abundant in the upper portions of the MM facies.In addition to being generally coarser, normally graded layers in the upper portion of the MM faciescommonly exhibit abundant ripples and contain relatively few trace fossils, in contrast to the lowerportions of the MM facies in which ripples are rare, and biogenic structures abundant (Fig. 2-5). Marineinvertebrate fossils are common in MM facies of the Bowser Basin, and plant debris is mostly absent.Yellow-brown weathered authigenic ferroan dolomite layers spaced 10 to 20 m vertically apart occur nearthe upper limit of MM occurrence.Coarsening Upward SandstoneThe coarsening and thickening upward sandstone facies (CUS) consists of resistant, light tomedium brown weathered, sandstone and siltstone (grey to dark grey on fresh surfaces) that occurs incomposite and generally parallel thin to thick beds. Grain size increases upward with bed thickness,from very fine to fine grained in thin beds near the base up to fine and less commonly medium grainedsandstone near the top. Typically, the base is gradational with underlying mudstones and near the base,siltstone interbeds make up as much as 50% of the bed. In contrast, the top contact is sharp and siltstoneinterbeds become thin and rare near the top. The overall thickness of CUS in the undivided Bowser LakeGroup, from the gradational base to the sharp upper contact, ranges from 8 up to 40 m. Symmetrical21Figure 2-5: Normally graded sandy siltstone layers separated by blue-black shale. These layers aretypical of extensively burrowed, normally graded siltstone and sandstone layers from the marinemudstone facies.22ripples (wave ripples) are common in the thinly bedded sandstones near the base of CUS. Trough crossbeds occur in the thick upper sandstones in some coarsening upward units. Along the northeast marginof the study area (sections 22 and 38 in figure 2-2), some CUS contain disarticulated shell beds, andconcentrations of chert pebbles (less than 5 cm diameter) in layers near their upper contacts. Some ofthese chert pebble layers define low angle (less than 10°) cross beds which dip to the southwest (towardsthe basin). Such shell and pebble layers are rare in sections 25, 26, 45 and 49 to the southwest (locationsgiven in figure 2-2). CUS is associated with invertebrate-bearing mudstones in adjacent facies ofundivided Bowser Lake Group. Thick CUS generally lack leaf fossils or other coarse plant debris, eitherwithin the facies or in adjacent mudstones.Fining upward sandstone faciesFining upward sandstone facies in the undivided Bowser Lake Group (FUS • e) consists offine and medium grained sandstone near the usually erosive basal contact, commonly with lode casts, andfines upward to very fine sandstone near the gradational, or less commonly sharp, upper contact. Thesandstone is well sorted, and near the base, beds are structureless, up to 2 m thick, and commonlycontain rounded rip-up clasts of mudstone that in some beds are as large as boulders. The very finegrained sandstone near the top of each FUS^• e is commonly rippled. Siltstone or claystone layersoccur gradationally above the very fine sandstone. Locally, FUS rine units occur stacked as many asthree times.Depositional environments of undivided Bowser Lake Group facies:The stacked normally graded layers that comprise most of the marine mudstone facies are similarin form to the tempestites described by Aigner (1985) in modem and ancient sediments, and wereprobably similarly deposited from suspension by bottom currents possibly related to storms. Similarnormally graded thin sands also originate from the Yukon Delta during rapidly waning flow followingextreme storm surge, and have spread more than 100 km offshore across the shallow shelf of the NortonSound, Alaska (Nelson, 1982). The graded sand layers of Norton Sound described by Nelson (1982)23exhibit distal to proximal increases in thickness, bioturbation, coarseness, and current structures similarto the vertical changes in MM facies of the Bowser Basin.The relative paucity of trace fossils in the upper portions of MM facies is consistent with astressed biologic community, suggesting rapid sediment aggradation and possibly brackish waterconditions (Pemberton and MacEachem, 1992). The decrease in trace fossil abundance and increase incurrent and oscillatory sediment structures in MM facies from low in the undivided Bowser Lake Groupto near the top of the unit suggest shallowing upward from outer shelf (below wave base) to shallow shelf(above wave base) deposition. In light of this repeated shallowing trend and the fact that deltaic faciesoverlie the MM facies (as described later in this chapter), it is likely that MM was deposited as prodeltashelf muds seaward of a prograding delta system.The MM facies is gradationally overlain by thick CUS, which represent sands that haveprograded into shallow shelf environments Given their great thickness and areal extent, the sands wereprobably delivered by distributaries of a major delta system. The CUS are a continuation of thecoarsening and shallowing trend that was noted in the MM facies, and probably represent the delta mouthbar depositional environment (c.f. Coleman, 1982). The sands have been reworked by waves to varyingdegrees, as evidenced by the shell lags and pebble layers, with the most intense reworking confined to theeastern edge of the study area. Low angle cross bedding defined by pebble layers and dipping towardsthe basin suggest beach deposits, and indicates that some of the thick CUS became at least brieflyemergent. However, lack of terrestrial plant fossils suggests that emergence, where it occurred, wasephemeral. Although deposited as distributary mouth bars, wave reworking has changed the sands todelta front bars, shoals and barrier islands or spits.Fining upward sandstones in the undivided Bowser Lake Group occur as interbeds in the MMfacies (where it is associated with invertebrate fossils, and extensive bioturbation) and in the upper partsof CUS. The lode casts common at the base, the thick structureless lower sandstone layer (commonlycontaining rip-up clasts), finer and commonly rippled upper sandstone layers, and marine association24suggest that the FUS interbedded with MM was deposited by turbidity currents in inner- to outer shelfenvironments These FUS may have been redeposited from the delta front to offshore positions, similarto the thinner normally graded sands that are abundant in MM facies. FUS associated with the upperportions of thick CUS are probably unreworked distributary channel deposits similar to those describedfrom the Rhone Delta (Oomkens, 1970).The upward trend to shallower water depths in the undivided Bowser Lake Group within thestudy area is the culmination of a regional marine shallowing trend that began in the Middle Jurassic withinitial Bowser Lake Group deposition. Initial Bowser Lake Group sediments were deposited in slope andsubmarine fan environments (Ricketts, 1990) where they overlie deep basinal marine claystones of theMiddle Jurassic (Lower Bajocian) Quock Formation (uppermost Spatsizi Group; Thomson, et al., 1986).By the time undivided Bowser Lake Group sediments exposed in the study area were deposited, slopeenvironments had been succeeded by deep to shallow shelf environments, which in turn were overlain bydeltaic deposits as described below. The progressive change from deep basinal deposits to slope faciesand finally upward to shelf and shallow marine environments is interpreted as a record of shelfconstruction. The shelf was constructed by filling a pre-existing ocean or marginal sea basinCURRIER FORMATIONThe Currier Formation was encountered in 20 sections, four of which contain the contact withthe underlying undivided Bowser Lake Group, and four of which contain the contact with the overlyingMcEvoy Formation (Fig. 2-2). The thickness of the Currier Formation is roughly 1000 m (Cookenbooand Bustin, 1990b), although no continuous sections of the entire formation are known. The CurrierFormation is the main coal-bearing unit of the Groundhog coalfield and is exposed from Maitland Creekin the north to the Groundhog Range in the south.Six facies (one divided into three subfacies) are recognized in the Currier Formation (Fig. 2-1)and described below. The most abundant facies are black claystone (homogeneous, carbonaceous, andlaminated subfacies), interbedded claystone-siltstone-very fine grained sandstone, fining upward25sandstone, and coarsening and thickening upward sandstone. Volumetrically minor facies include thickcoal, and well-sorted chert pebble conglomerate.Black claystone faciesThe black claystone (BC) facies consists of recessive-weathering, variably silty, black to darkgray, thin to thick bedded claystone. This facies is characterized by abundant terrigenous organic matterincluding leaf fossils, petrified wood, coal and disseminated plant debris, and a lack of marineinvertebrates. In addition, a diverse assemblage of spores, pollen and dinoflagellate cysts have beenrecovered from some claystone samples (Rouse, personal communication, 1989). The BC faciescomprises three subfacies recognizable in the field:1) Homogeneous black to dark gray claystone subfacies. Homogeneous black to dark gray claystoneBChomogeneous) commonly 3 to 10 metre thick is the most common subfacies. It consists of variablysilty claystone, and contains abundant and varied leaf remains Few trace fossils are seen in outcrop, butslabbed and polished samples commonly reveal pervasive Hehninthopsis burrows. Interbeds of finesandstone lenses 30 to 50 cm thick of limited lateral extent (less than 10 to 20 m) occur within the bedsof BChomogeneous beds. Thick BChomogeneous beds commonly are gradationally overlain bylaminated claystone or siltstone, and underlain by carbonaceous claystone. This subfacies commonlycontains carbonate concretion layers which weather yellow-brown and usually occur at the tops of beds.Most of the concretions are composed of ferroan dolomite, and some contain cores of siderite. Theyellow-brown layers commonly have very well preserved leaf fossils on their surfaces, and appear to becoalesced concretions.2) Carbonaceous black claystone subfacies (BCcarbonaceous). Carbonaceous black claystonesubfacies (BCca h nseeous) is similar to the homogeneous black claystone subfacies, but is distinctlymore organic rich, with abundant plant fossils and occasional coaly layers. Many carbonaceousclaystones that directly overlie coals (roof shales) have been analyzed for palynology, and contain a widevariety of pollen, spores, and dinoflagellate cysts. This facies commonly grades into homogeneous26claystone or coal.3) Laminated black daystone subfacies (BCiam• td). The laminated black claystone subfacies(BCia inat d), consists of black claystone and parallel laminae of buff to yellowish brown siltstone. Thelaminae usually range between 3 and 15 mm thick for each claystone-siltstone pair, and are normallygraded. The laminated facies is up to 3 metres thick, and commonly lacks both macro- and microfossils,although surrounding beds tend to be rich in plant remainsInterbedded claystone, siltstone and very fine grained sandstoneInterbedded claystone, siltstone and very fine grained sandstone facies (ICSS) occurs as agradational, thin to thick bedded facies in the Currier Formation, and dominates some exposures of theupper part of the formation. This facies consists of gradational beds of claystone overlain by siltstoneand/or very fine grained sandstone, that occur in composite units between 0.5 and 20 m thick. Claystoneis similar to the black claystone described above, and is only included in this facies where gradationallyin contact with beds of siltstone or very fine grained sandstone. Locally, ICSS coarsens upward inclaystone-to-siltstone-to very fine sandstone successions, similar to coarsening upward mudstones in theMcEvoy Formation (described below).Fining upward sandstone facies (FUS)In the Currier Formation, the FUS facies is most commonly 1 to 3.5 m thick, although thickerFUS (up to 20 m) also occur. Similar to the undivided Bowser Lake Group, FUS in the CurrierFormation is resistant, weathers dark gray and light brown (gray to dark gray on fresh surfaces), and iscomposed of well sorted very fine, fine and medium grained sandstone. FUS fines upward from ausually erosive basal contact to a gradational or less commonly sharp upper contact. In the CurrierFormation, FUS is generally laterally continuous at outcrop scale with only minor down-cutting visible.Rounded rip-up clasts of mudstone pebbles are common near the base where the sandstone is usually finegrained, structureless, and thick bedded. Beds are usually structureless, but planar cross bed sets 50 to2775 cm thick occur rarely in the basal portion of the fining upward sandstone. Groove casts up to 10 cmacross are common on the base of FUS where well exposed. Higher in the fining upward sandstones,grain size is finer, and planar or trough cross beds sets less than 50 cm thick are common.Coarsening upward sandstonesCoarsening and thickening upward sandstones (CUS) are common in the Currier Formation. Inthe lower Currier Formation, CUS are similar to those described above for the undivided Bowser LakeGroup, except that in the Currier they are commonly capped by BC carbonaceous and/or coals (includingthe thickest coals of the Groundhog coalfield). For example, in sections 27, 28, 33, 41, 44 and 47located along the limits of Currier Formation exposure (cross-section figure 2-6; and location map figure2-7), the first thick coal occurs 10 to 20 m above CUS thick. In some sections, this transition of marinefacies to terrestrial coals is repeated over stratigraphic intervals of 150 to 300 m, suggesting stacked deltalobes prograding into shallow marine waters. Section 33 (Fig. 2-8), for example, has 10 m of coal-bearing strata above an 18 m thick coarsening upward sandstone in the lower 50 to 100 m of the section,and another coal-bearing section (two seams of 1.5 and 1 m thickness) 300 m higher in the section, againwithin 20 m above a 10 m thick coarsening upward sandstone.In the upper portions of the Currier Formation, thinner CUS occur ranging between 2 to 7 mthick. These differ from thick CUS in the lower Currier Formation and underlying Bowser Lake Groupstrata because they are generally associated with abundant plant remains and rare or absent invertebratefossils, and may have rooted upper contacts.Chert pebble conglomerateChert pebble conglomerate (CPC) is a minor facies in the Currier Formation. Theconglomerates are typically clast supported, and most are between 5 and 20 m thick. Clasts are rounded,well sorted pebbles generally 1 to 3 cm in diameter, and the maximum observed pebble28Figure 2-6: Cross-section of Currier Formation strata. Relative stratigraphic positions of sections arepoorly constrained. Interpreted positions are based on facies character (mainly presence orabsence of thick coals) and relative proximity of each section to marine facies of the undividedBowser Lake Group below and fluvio-deltaic facies of the McEvoy Formation above.29181O000UUI129° 00'^ 128° 30'^ 128° 00'0^15KILOMETRES10,0.91,111tS 111.1tratBRITISH COLUMBIAOUTCROP PATTERNFORMATIONSc,`3 gDEVILS^1 10VD °o ° (3-^-^•^-.-_-_-_MCEVOY • - - ------•^-.^- .. - .CURRIERumdivkled1 B 0 WS E R LAKE GROUPConglomerate:0:30° 00 00 ;Sands-toneSiltstone _=Claystone - - CoalMMCUSMMCOALCUSCOALCUSMMMM500 rn400 n300 rn200 r1.100 ri .^ COALCOALCUSCUSc)---cl—Q--)) COALCUS2 FUSM MCUSFUSOFigure 2-8: Marine mudstones interfinger with coarsening upward sandstones capped by thick coals inthe lower Currier Formation (section 33 located east of the Klappan River).31section 33size is 7 cm. The clasts are more than 95% chert, and occur in colors of gray, green, and black. Theconglomerates form laterally extensive sheets 1 kilometre or more across most outcrops, but show localthickening in channels.In some exposures, thick pebble conglomerate grades laterally into sandstone channels Anexample of such lateral gradation is well exposed south of Mt Klappan (section 3). Here, a 15 m thickconglomerate composed of pebbles generally less than 2 cm in diameter grades over a distance of 150 mto the west into fine grained sandstone. Still farther to the west this unit exhibits a sharp basal contactthat cuts more than 2 m down into the surrounding mudstones.Thick coalCoals in the Currier Formation are commonly associated with carbonaceous claystones anddirect overlie CUS. The thickest coals occur in the lower Currier Formation, where coals up to 3 to 5 mthick have been reported from north of Mount Klappan (Gulf Canada LTD open file reports, 1984, 1987)and near Currier Creek (Bustin and Moffat, 1983). Two poorly exposed seams 8 to 10 m thick wereencountered east of the Klappan River (section 33, Fig. 2-8; Cookenboo and Bustin, 1990b). The thickseams east of the Klappan River directly overlie coarsening upward sandstones that in turn overlie marinemudstones. Seams less than 1 m thick occur in the upper Currier Formation where they are associatedwith thin CUS and not associated with marine mudstones. Some of these thinner coals are rooted inunderlying CUS.Currier Formation coal is anthracite to meta-anthracite in rank (Chapter 5), and commonlycontains framboidal or disseminated pyrite, and diagenetic veins of calcite and quartz.Depositional environments of the Currier FormationThe Currier Formation occurs stratigraphically above, and in its lowest parts interfmgered with,the pro-delta muds and delta front sands that characterize the underlying undivided Bowser Lake Group.As might be expected in such a stratigraphic context, the Currier Formation facies are consistent with32prograding deltaic deposition. Thick FUS and CUS facies in the lower Currier Formation occur at thetransition from marine to partly terrestrial deposition. The thick FUS, with their sharp, at least locallyerosive base, well sorted texture, and fining upward nature, probably represent channel deposits of majordistributary channels. As with the CUS in the upper part of the undivided Bowser Lake Group, the thickCUS of the lower Currier Formation were deposited as delta mouth bars delivered to the delta front bythe distributary channels inferred from the FUS facies. In contrast to the underlying marine strata, thickCUS in the lower Currier Formation became emergent for significant periods. This emergence isdemonstrated most directly by the occurrence of thick coals above many of the CUS, whose plantprecursors required an emergent substrate to grow upon. Furthermore, the fine grained and organic richcharacter of the black claystone facies that commonly occurs above the marine to terrestrial transition isconsistent with accumulation in protected waters, such as form shoreward of emergent delta front barrierbars and islands. The laminated black claystone facies, which is largely restricted to the CurrierFormation, and is commonly associated with BChomogeneous and coarsening upward mudstones issimilar to laminated deposits are known from the central parts of tropical and temperate deltaic lakes andlagoons (Tye and Coleman, 1989). Such a depositional environment is consistent with delta depositionin water bodies of large surface area that are subsequently filled by fine grained crevasse splay orlacustrine delta deposits (Tye and Coleman, 1989). Such large water bodies, including interdistributarylakes and lagoons are common in lower delta plain environments.The stacked coarsening upward cycles, first formed in the underlying marine strata (e. g. section49, Fig. 2-3), and continuing during the accumulation of the Currier Formation (e. g. section 33, Fig. 2-8) suggests deposition by lobes of a repeatedly large prograding delta system. The delta lobes are formedin a cycle, as demonstrated in the Mississippi Delta (Fig. 2-9). The cycle begins with initialprogradation and deposition of deltas front sands followed by abandonment and subsidence, leading nextto wave reworking of barrier sands and lagoon fill deposits, and finally to marine inundation anddevelopment of shallow marine conditions (Fig. 2-9). This entire cycle of delta lobe evolution is33Figure 2-9: Four steps of delta lobe progradation, subsidence, innundation and re-establishment.Simplified from Bhattacharya and Walker (1992), based on data from Boyd and Penland (1988).3442-r^5 41§3Simplified from Bhattacharya and Walker (1992)after data from Boyd and Penland (1988).typically preserved in Currier Formation, suggesting subsidence was rapid compared to the rate ofsediment supply.The thin FUS (1 to 3.5 m thick) and CUS (2 to 7 m thick) facies that are common in the upperCurrier Formation were deposited by smaller scale fluvial systems than the thick FUS and CUS of thelower Currier Formation. These thin FUS and CUS are interpreted as crevasse splay or subdelta channeland mouth bar deposits that filled in protected lagoons and interdistributary bays formed shoreward ofthe delta front. This interpretation is consistent both with the inferred smaller size of fluvial system, andwith the abundance of leaf fossils and the general lack of marine macrofossils in the upper CurrierFormation. Preservation of sediments deposited mostly shoreward of the delta front during later Curriertime suggests the shoreline had prograded west of Konigus Creek, out of the study area. Suchpreferential preservation of shallow water, marginal marine deposits also implies that subsidence in thestudy area had diminished relative to sedimentation rate (i.e., less accommodation space was created perunit of delivered sediment) by the time the upper Currier Formation was deposited.MCEVOY FORMATIONThe McEvoy Formation was examined in 9 sections between the headwaters of the Skeena andNass rivers, three of which contain the contact with the underlying Currier Formation, and 2 of whichcontain an exposed contact with the overlying Devils Claw Formation (Fig. 2-2). The thickness of theMcEvoy Formation exceeds 780 m between the Skeena and Nass rivers. East of the Skeena River atDistingue Mountain (location given in figure 2-2), 670 m of McEvoy Formation strata were measured,but no upper or lower contact was encountered. Lithologically similar strata with a composite thicknessof as much as 1500 m were examined in 5 sections east of the Nass River and south of Sweeney Creekand are tentatively assigned to the McEvoy Formation. Part of this section may be time correlative withthe Devils Claw Formation, but it lacks the thick conglomerates by which the Devils Claw Formation isdefined.35Coarsening upward mudstone and chert pebble-cobble conglomerate are the most characteristicfacies of the McEvoy Formation. These two facies and the usually thin coals are described in detailbelow. Black claystone, fining upward sandstone, and coarsening upward sandstone also occur in theMcEvoy Formation. The latter three facies are not further described below because they have essentiallythe same characteristics described earlier for BC, FUS, and CUS facies in the Currier Formation, withthe following exceptions: BCla • ted was not found in McEvoy Formation sections, and CUS is limitedto thin examples with no associated marine facies.Coarsening upward mudstone facies (CUM)Coarsening upward mudstone (CUM) is the dominant facies of the McEvoy Formation. CUMconsists of moderately recessive beds of claystone and siltstone with interbeds of silty very fine to finegrained sandstone. A typical coarsening upward unit begins with black claystone (which may becarbonaceous or coaly near the base) which has a sharp basal contact. The black claystone gradationallycoarsens upward into siltstone, which in turn coarsens upward into very fine sandstone (locallycoarsening further to fine grained sandstone near the top). The sandstone usually has an abrupt uppercontact that may be rooted, and is commonly overlain by another coarsening upward mudstone. Atypical CUM unit is composed of approximately 70% siltstone, and roughly equal parts claystone andsandstone accounting for the rest. However, a range of claystone-siltstone-sandstone ratios exists, fromsome composed only of claystone and siltstone to others composed dominantly of siltstone and sandstone.This facies weathers dark gray to brownish gray and contains abundant leaf fossils, wood, and plantdebris. Many of the best preserved leaf fossils are found on surfaces of authigenic carbonate interbeds,most commonly located at the siltstone to claystone transition. Invertebrate fossils are lacking and tracefossils are rare.CUMs range from 1 to 8 m thick, and are commonly stacked as many as 6 or 7 times insuccession without being broken by another facies. In section 12 (Fig 2-10; location given in figure 2-2), a McEvoy Formation exposure west of the Skeena River, the thickness of CUMs decreases3637Figure 2-10: Cliff face exposure composed of mostly CUM facies, from the lower part of section 12 inthe McEvoy Formation. Stratigraphic column for section 12 illustrated in figure 2-11.upwards in two successive trends (Fig. 2-11). The lower trend has CUMs near the base 5 m thick thatthin to 1 to 2 m each 200 m higher. This thinning upwards trend is overlain by another thinning upwardstrend 80 m thick that varies from 8 m CUMs near the base to 3 to 3.5 m thick CUMs at the top.Although they vary widely in thickness in different locales and stratigraphic levels, adjacent CUMs froma single exposure tend to be of approximately equal thickness.Chert Pebble-Cobble ConglomerateConglomerate in the McEvoy Formation is generally thicker, coarser and more laterallyextensive than in the Currier Formation. Most conglomerate beds are between 10 and 15 m thick, andsome can be traced in outcrop for more than 2 kms. Clasts are more than 95 % chert pebbles andcobbles, with maximum observed clast size of 10 cm. Three of the thickest and most laterally extensiveof the conglomerates occur in 200 m thick zone near the middle of the McEvoy Formation. Two of thesethree conglomerates occur in successions similar to the fining upward conglomerate facies described forthe Devils Claw Formation later. Beds are typically clast supported, with a matrix of medium to finesand and no preferred grain orientation observed. Locally, some conglomerates grade to pebblysandstone, and basal contacts are erosive.CoalCoal in the McEvoy Formation east of the Nass River consists mostly of thin seams ( < 0.5 m)and discontinuous coaly layers that commonly occur in carbonaceous claystone. Some of these thin coalsare rooted in sandstones or siltstones that form the top of underlying coarsening upward sequences (CUMor thin CUS facies). Coal is semianthracite to anthracite rank, with most vitrinite reflectance values(random reflectance in oil) between 2.1 % and 2.9% (Chapter 5).West of the Nass River and south of Sweeney Creek (sections 34 and 35 in figure 2-2), sixseams of 1 to 3 m thickness occur within 300 m of strata that are interpreted to be a generally finergrained facies of the McEvoy Formation than that previously describe in the type area of the Klappan38Figure 2-11: Stacked coarsening upward mudstones from section 12 in the McEvoy Formation.500 n —400 PI300 rn200 n100 ri Legend ECUCUO E'0▪•^410V erSection 1239and Groundhog coalfields east of the Nags River (Cookenboo and Bustin, 1989). Coal in the SweeneyCreek area is semianthracite rank., with a measured vitrinite reflectance value of Ro d=2.2%. to 2. % .Depositional environments of the McEvoy FormationThe dominant facies of the McEvoy Formation is coarsening upward mudstone (CUM). A close modernanalog for this facies (with excellent preservation potential) is the upward-coarsening lacustrine delta ofLake Fausse Pointe which formed on the upper delta plain of the Mississippi River. As described by Tyeand Coleman (1989), the Lake Fausse delta is a 1.5 to 3 m thick coarsening upward succession of clay,silty clay, silty very-fine grained sand, and fine grained sand that has rapidly aggraded and prograded(approximately 100 years since initiation) over organic-rich backswamp clays. Tye and Coleman (1989)also point out that such upward-coarsening sequences tend to be stacked in repeated units, and three suchupward-coarsening units are stacked in the Lake Fausse area in the 30 m of Holocene sediments.Notably, most of the vertical sediment aggradation in the Lake Fausse area occurred during the rapiddelta building events, but most of the time is represented by slowly accumulating backswamp deposits.CUM facies in the study area probably accumulated mostly on the upper delta plain but diverse dinocystsassemblages (Rouse, written communication, 1989) recovered from carbonaceous claystones(BCcarbonaceous) collected near the base of some CUMs in the McEvoy Formation suggest that thebodies of water filled in by these facies were bays at least intermittently connected to the sea.Similar CUMs that occur alone or in repetitively stacked units in ancient rocks have also beenidentified as crevasse splay deposits in Cretaceous coal bearing strata of the Rocky Mountains (Flores,1985). Like the coarsening up units described by Flores (1985), CUMs in the study area are associatedwith fining upward sandstones (1-3.5 m version), which are interpreted as crevasse splay channels (seeabove) and thick black claystone facies deposited in backswamp and lagoon environments. Unlike theRocky Mountain strata, coals associated with CUMs in the study area are thin and argillaceous.Fining upward sandstones occur in a wide range of thicknesses, suggesting the presence of bothmajor distributary channels and smaller crevasse channels, as were interpreted for the Currier Formation.40Coarsening upward sandstones, however, are consistently relatively thin (2 to 7 m) in the McEvoyFormation, suggesting that deep bodies of water were lacking.Compared to the underlying Currier Formation, where the entire cycle of delta lobe evolution isrepeatedly preserved, the McEvoy Formation contains mostly upper delta plain deposits, and the deltafront is not preserved in the study area (Fig. 2-9). Because no delta front sediments occur, marineinundation between sediment aggradation events was incomplete (assuming continual subsidence), andthus sediment filled the accommodation space that was created by subsidence more quickly than duringCurrier Formation accumulation. Therefore, subsidence rates were lower relative to sedimentation ratefor the McEvoy Formation than the Currier Formation.The chert pebble-cobble conglomerates of the McEvoy Formation are probably channel depositsof the same fluvial system that deposited thick FUS. The great lateral extents of these conglomeratessuggests that the fluvial channels migrated unimpeded across the inferred upper delta plain. Theincreasing abundance and coarseness of conglomerates in the McEvoy Formation compared to the CurrierFormation indicates that the fluvial system became more competent through time at delivering gravels tothe study area. Theoretical modelling by Paola (1988), supported by recent field studies (Heller andPaola, 1992; Gordon and Heller, 1993), indicates that conglomerates composed of resistant clasts (suchas the oligomict chert conglomerates of the northern Bowser Basin conglomerates) tend to progradefarther into a basin as a result of slower subsidence rate. The conglomerates, therefore, support theconclusion that subsidence was relatively slow in the study area during deposition of the McEvoyFormation.DEVILS CLAW FORMATIONThe Devils Claw Formation was examined in 2 sections between the Nass and Skeena rivers,and both sections contain the contact with the underlying McEvoy Formation (Fig. 2-12 and Fig. 2-13).41S- S(lines aelween sections correlate Oases of lining upward conglontensle successions in Mt Devils Claw Tormation in sections /3 and 18, and correlative strata in section 3e)Devils ClawFortmat c)nMcEvoyFormationsection 3 4 section 18section 13treaty, bast a/ atConglomerateSandstoneStitstoneClaystone 8 §I129° 00^ 1280 30'^ 126°00'The maximum measured thickness of Devils Claw strata is 600 m but no younger strata are preservedabove the top of the section and therefore the total depositional thickness is not known.Claystone, mudstone, sandstone and thin coal facies broadly similar to those in the McEvoyformation also occur in the Devils Claw Formation. However, the stratigraphic distribution is moreregular in the Devils Claw Formation than in either the Currier or McEvoy formations: facies in theDevils Claw Formation are stacked repeatedly in a pronounced fining upward pattern, beginning at thebase with a conglomerate and commonly grading upward to carbonaceous claystone (and in some localesthin coal). The pronounced fining upward tendency leads to the different lithologies being describedbelow as a single facies characteristic of the Devils Claw Formation. Coarsening upward mudstones andchert pebble-cobble conglomerates similar to those described for the McEvoy Formation (not in welldefined fining upward successions) also occur as minor components of the Devils Claw Formation, mostcommonly in the lower part of the formation.Fining upward chert pebble-cobble conglomerate (FUPCC)Pebble-cobble conglomerate occurs with sandstone, siltstone, clay-rich mudstone and occasionalthin coal in repeatedly stacked, fining upward units through most of the Devils Claw Formation.Individual fining upward units range from 25 to 30 m thick, and conglomerate generally forms between20 and 70% of the fining upward unit. Overlying the conglomerate, there is gradational change tomedium to fine grained sandstone, siltstone, and carbonaceous shale. Coals up to 1 m thick cap somefining upward units. Although dominated by chert pebble-cobble conglomerate, this facies (FUPCC)differs from others described in this chapter in being composed of many different lithologies. However,the pronounced tendency of these lithologies to occur in the same fining upward succession through mostof the Devils Claw strata justifies consideration of these multiple lithologies as a single facies.Conglomerate forms the base of each FUPCC facies, and occurs in laterally continuous beds thattend to form the resistant cliffs which characterize most Devils Claw Formation exposures. Air photointerpretation indicates continuity of individual beds may be on the order of 5 to 10 km (Cookenboo,441989). The bases of the conglomerates are locally erosive, cutting down more than 3 m into underlyingfine sediments over a lateral distance of less than 50 m. Scours 50 cm to 1 m deep occur in some of theseerosive bases. The beds are usually normally graded, with cobbles concentrated near the base.The conglomerate is composed of clast supported chert pebbles and cobbles up to 15 cm inlongest dimension Like the conglomerates from older units, clasts are more than 95% well rounded,gray, green and black chert, and most are spherical, although roller shapes also occur. Most beds exhibitno preferred clast orientation, although large scale trough cross bedding occurs in some beds in sets morethan 3 m thick. The cross bed sets are composed of layers approximately 20 cm thick that fine upward,with cobbles concentrated in the lower 10 cm, and are concave up.The fining upward conglomerate facies is best exposed in section 18 on a ridge west of the NassRiver (figs. 2-12 and 2-13). Here 11 fining upward sequences, each from 22 to 28 m thick, occurstacked vertically in the Devils Claw Formation. From the base of the Devils Claw Formation upwards,six conglomerates occur over an upward interval of 185 m, with only minor CUM facies separating someof the fining upward sequences. At least five more fining upward successions occur higher in thesection, but are not accessible due to steepness. Observations from a distance indicate that the sharpbased and gradational fining upward pattern continues upward from conglomerate in these successions.An additional conglomerate occurs in the upper McEvoy Formation, 300 m below the Devils ClawFormation. In each unit, the chert pebble-cobble conglomerate occurs at the base, and comprises from40% to 90% of each fining upward sequence. Cobbles as large as 15 cm in their longest dimension areconcentrated near the erosive base, where scour marks are common. Higher in the conglomerate, cobblesbecome rare or absent as pebbles increase, and sandstone interbeds become more abundant. Above theconglomerate, sandstones grade upwards from thick bedded medium grained with planar and trough crossbeds, to fine, very fine and finally silty sandstone. The silty sandstone grades into siltstone, which inturn grades into mudstone and carbonaceous claystone, and in one area the fining upward sequence istopped by coal. The most striking aspect of FUPCC in this exposure is the regularity and completenessof each fining upward sequence (conglomerate-sandstone-mudstone-carbonaceous claystone and rarely45coal). In addition, only minor quantities of other facies are preserved in this section.In section 13 located in the Devils Claw Formation west of the Skeena River (Figs. 2-12 and 2-13), cliffs of FUPCC dominate the upper 450 m of exposure (Fig. 2-14). Conglomerates occur in fivefining upward sequences 19 to 30 metres thick, similar to those in section 18. Maximum observed clastdimension is 13 cm in roller shaped cobbles. Two other conglomerates are much thicker (64 and 68metres thick each). These very thick conglomerates are interpreted as stacked fining upwardconglomerates in which the fine grained beds separating the conglomerates are missing due to subsequenterosion, rather than a fundamental difference in sedimentary processes. The 64 metre thick conglomerateis well exposed, and exhibits three fining upward divisions of 18, 21 and 24 m thickness, separated bysandstone or pebbly conglomerate interbeds, and thus is a composite bed of three fining upwardconglomerates. The 68 m conglomerate is poorly exposed (partially talus covered).Near the base of section 13, lens shaped pebble conglomerates (maximum observed clast size of5 cm) occur that cut sharply down as much as 6 m into surrounding fine grained facies (mostly CUM).These conglomerates are not parts of fining upward successions and are up to 17 m thick, and are limitedto the lowest 50 m of the Devils Claw Formation.The third section measured with abundant FUCPCC is located south of the headwaters ofSweeney Creek (section 34 in figures 2-12 and 2-13). In this exposure, eleven fining upward pebble-cobble conglomerates 2 to 38 metres thick occur over a stratigraphic interval of 850 metres. Thedefining characteristics of erosive base, cobbles (maximum observed clast size 7 to 9 cm) concentratednear the base, associated plant-bearing facies, and fining upward succession of overlying beds are allsimilar to sections 18 and 13, but here the conglomerates are generally thinner Interbedded sandstones,mudstones and occasional coal separating the conglomerates dominate the section.46Figure 2-14: The upper 420 m of section 13 exposed on an east west trending ridge line west of the Skeena River. The more resistant cliffs areconglomerates, as are some of the less resistant strata. The upper part of cliffs to the west (left side of field of view) were unreachable, and suggestthat significantly more strata may occur in the Devils Claw Formation than the 600 m maximum measured in this section.48The three exposures described above exhibit a distinct trend from east to west (sections 13 to 18to 34; cross-section figure 2-12; location map figure 2-13) of increasing deposition and/or preservation offine grained fluvial facies between conglomerate beds. Structural complexity and sparse paleontologicalcontrol make bed to bed correlations tentative in widely separated sections such as these, but as shown incross-section (Fig. 2-12; location map Fig. 2-13), sections 18 and 13 both contain the McEvoy to DevilsClaw contact, and can be confidently correlated lithologically. Section 34 may also be correlative, butbecause of structural separation, lower overall concentration of conglomerate (17 % of the section), andthe increased fluvial fine grained facies, this correlation remains uncertain. Limited paleontologicalcontrol supports the facies based correlation, as do the low vitrinite reflectance values, assuming onlyburial related coalification (Moffat, 1985; Fig. 2-12).Depositional environments of the Devils Claw FormationThe Devils Claw Formation is dominated by conglomerate facies. Near the base of the formation, lensshaped chert pebble conglomerates cut down sharply into fine grained facies, suggesting creation ofravinement surfaces in the underlying McEvoy Formation. Creation of such ravinement surfaces may berelated to relative drops in sea-level, and consequent increase in overall channel slope (Reynolds et al.,1991).Above these lens-shaped conglomerates, fining upward chert pebble-cobble conglomerate faciesdominates more 500 m of strata. The erosive bases, normal grading, and fining upward character suggestfluvial deposition, similar to the processes inferred for deposition of McEvoy Formation conglomerates.In the Devils Claw Formation, however, fine grained facies are mostly absent. The three sectionsillustrated in cross-section figure 2-12 suggest increased basinward deposition and/or preservation ofsediments transported by a gravelly fluvial system, similar to the proximal-distal relationship summarizedby Miall (1985) for the Scott, Donjek and South Saskatchewan rivers. The analogy with the three riverslisted above is imperfect, however, because the great lateral extent of Devils Claw conglomerates suggestthat the fluvial system was largely unbound by valley walls. Given the stratigraphic context overlying(and to some extent mixing with) upper delta plain deposits of the McEvoy Formation, an alluvialbraidplain may best describe the Devils Claw depositional environment. Following the model of Paola(1988) mentioned earlier to explain the progradation of McEvoy Formation conglomerates, the DevilClaw Formation may have accumulated during a period of slow basin subsidence, leading to progradationof oligomict chert conglomerate relatively far into the basin.SUMMARY OF DEPOSITIONAL HISTORYSedimentation began in the study area with accumulation of middle to outer shelf muds andturbidites during the Middle to earliest Late Jurassic. The initial shelf sediments are distal prodeltadeposits that prograded over Bajocian to Callovian (Middle Jurassic) deep basin to slope deposits exposednorth and east of the study area (Ricketts, 1990), and they mark the change from primarily aggradationalsediment accumulation that characterized the Bowser Basin since its inception in the Early to MiddleJurassic, to mainly progradational sediment accumulation that characterizes the remainder of BowserBasin history through the mid-Cretaceous. Each prograding delta lobe filled the availableaccommodation space to near sea-level, as shown by the occurrence of current structures and wavereworking in thick CUS in the upper portions of the undivided Bowser Lake Group. Successiveprograding lobes are recorded by the stacked coarsening upward mudstone to sandstone sequences, andeach lobe filled accommodation space as it was created, implying continual subsidence of the underlyingstrata. The tendency of stacked coarsening upward successions to become thinner higher in thestratigraphic column suggests progressive change in the balance between sedimentation rate andsubsidence. A progressive reduction of subsidence rate through time is consistent with the sedimentcompaction and isostatic adjustment model of accommodation space creation described in a later sectionof this chapter (p. 50). An increase in sedimentation rate could cause similar progradation, and cannot beexcluded.Overlying the marine strata, Currier Formation facies record further progradation of the deltaduring the Late Jurassic and possibly Early Cretaceous. Extensive pests formed above delta front barrier49sands and gave rise to the thick lower Currier Formation coals. A complex of crevasse splay, lagoon fill,and distributary channel deposits accumulated behind the delta front sands. These deltaic depositsrepeatedly built up to subaerial delta plains, and just as repeatedly subsided again (Fig. 2-15). As in theunderlying undivided Bowser Lake Group strata, the thickness of coarsening upward successionsgenerally diminishes up in the stratigraphic column. Further reduction of subsidence rate, is thereforeinferred during Currier Formation accumulation.Deposits of the McEvoy Formation overlying the Currier Formation suggest further deltaprogradation during the mid-Cretaceous. Sediments accumulated by repeated aggradation of fluvialdeposits on a relatively slowly subsiding delta plain. The relative slow subsidence of the delta plaincould indicate a further reduction in subsidence rate, or an equivalent increase in sedimentation rate.Some indications of an increase in sedimentation rate exist. Sandstone framework grain composition ismore chert rich (compositionally mature) in the McEvoy Formation than in the Currier Formation,suggesting that recycled alluvial plain equivalents to the Currier Formation may have been added to theprovenance of the lower McEvoy Formation (Chapter 3). Addition of easily eroded alluvial plainsediments to the provenance is consistent with an increase in sedimentation rate. Furthermore, erosion ofbasin margin sediments is confirmed by the sub-Sustut Group unconformity east of the study areareported by Eisbacher (1974). This unconformity may correlate with the depositional hiatus between theCurrier and McEvoy Formations inferred from palynological data (Rouse, written communication, inCookenboo and Bustin, 1989).Chert pebble to cobble conglomerates become more abundant higher in the McEvoy, anddominate the overlying Devils Claw Formation. The conglomerates are interpreted as a record ofprogradation of alluvial braidplain facies over the older deltaic strata. The highly resistant nature ofchert clasts, and the oligomict nature of the conglomerate are, as mentioned earlier, consistent withprogradation into a slowly subsiding basin (Paola, 1988, Gordon and Heller, 1993).50undividedBowser LakeGroup faciesCurrierFm.faciesMcEvoyFm.fades....••• ....1.0.•^••••,..... ••••...•••••••• ..... .0" ,.....I'^..... 0.. ,...... „..., ..••••..0".•••••• 0.0%0.- ....• .00" ..I' I.......0".•••••I' .0"*"*" . I*............/.....".0** ../ ...."'....'....../....":„........ .I.••••"'"'"-- I'....."*..."*.••• •••••.I•••*". .I.....0 ........, ,..■...***".I.I",, ...0 „..... -.'ii i ..0." ...**"I-0"......, ..... Iiii .•••••"..."'.0*°••••*"...0°.0":„..051Figure 2-15: Diagrammatic representation of deltaic depositional environments typical of the CurrierFormation prograding over shallow marine shelf deposits of the undivided Bowser Lake Group.During accumulation of the upper McEvoy and Devils Claw Formations, therefore, subsidence rate wasprobably decreasing. Sedimentation in the Bowser Basin ceased some time after Devils Clawaccumulation, and during the Late Cretaceous, some of the Bowser Basin strata were eroded andtransported into the Sustut basin to the east (Eisbacher, 1974; Bustin and McKenzie, 1989).CLIMATEClimate affects many sedimentologic processes, but the record left by climate effects is subtleand difficult to quantify from the rock record. However, understanding climate is of great importance topaleogeographic reconstructions, and therefore it is worth considering the sedimentologic evidenceavailable in a qualitative attempt to reconstruct paleoclimate for the northern Bowser Basin. The mostobvious sedimentologic evidence concerning climate in the northern Bowser Basin is the presence of coalin all facies associations above the undivided Bowser Lake Group. Variation in coal seam thickness andabundance in different facies associations, however, is controlled by changing depositional processes, notnecessarily climate change. Clearly, coal implies humid climatic conditions. Modern peat accumulationsare favoured in two regions where precipitation exceeds evaporation: the tropics, (between 13 °N and7 °S latitude); and the temperate zone (north and south of 38° latitude), where moist air in the Hadleycell descends as rain (Sellers, 1965 in Ziegler et al., 1987). Paleogeographic reconstructions of coaldeposits on a worldwide basis confirm that most coals were deposited within either the tropic ortemperate zones (Habicht, 1979). Other features allow distinction between tropical and temperate zoneorigin for northern Bowser Basin sediments. Most important is the presence of conspicuous tree ringscomprised mostly of earlywood with a sharp and thin transition to latewood (Fig. 2-16) in the abundantfossil wood. Additionally, earlywood cell walls are thin and their tracheid cross-sectional areas arelarge. Conspicuous tree rings form in response to seasonal growth and are pervasive in temperate zonewoods, but are commonly lacking or faintly developed in tropical woods (Creber and Chalconer, 1985).The large thin walled tracheids are comparable to modern conifers not commonly exposed to freezing,and suggest mild winters (Spicer, 1989). False rings also occur in Bowser Basin wood, suggestingdrought conditions during some growing seasons52Figure 2-16: Silicified transverse section of fossil wood from the Currier Formation, showing abundantearly wood and a sharp change to late wood. Field of view is 4 mm in long dimension.late wood53(summer months), although other causes of false rings such as waterlogging, fire or insect attack cannotbe eliminated (Spicer, 1989; Spicer and Parrish, 1990). The combination of coals and conspicuous treerings, therefore, points to a humid temperate climate for the northern Bowser Basin.An extensive collection of Bowser Basin leaf fossils has been described by MacLeod and Hills(1990). From their work, it is clear that composition of the flora remained relatively stable throughoutBowser Basin sediment accumulation, and that the described flora compares well to worldwideoccurrences of humid temperate floras from other Upper Jurassic and Lower Cretaceous strata. Inaddition, Nilssonia, an important component of the Bowser Basin flora, has been shown to be adeciduous plant (Kimura and Sekido, 1975). Deciduous plants, like conspicuous tree-rings, suggest aseasonal climate. The leaf fossils, therefore, support a humid, seasonal temperate climate during BowserBasin sedimentation.Further evidence for a cool temperate rather than equatorial or sub-tropical climate is providedby stable 0 isotope ratios in carbonate mudstone concretions (Chapter 5). The d 180 values suggestconcretion precipitation from meteoric or mixed meteoric and marine waters similar to those of presentday coastal British Columbia and Alaska.In contrast, climate of the continental interior appears to have changed during this period.Macroflora assemblages from strata east of the Cordillera change from the Late Jurassic to the end of theEarly Cretaceous. One of the prominent changes is the introduction of angiosperm leaf fossils in somemid-continent assemblages in the early Albian. The early angiosperms are thought to have beencolonizers of stressed environments, which may have been more common in the interior due totectonically induced climatic change. For example, Ruddiman and Kutzbach (1991) have demonstratedthat late Tertiary uplift of large areas of the Cordillera had caused drying of the North Americancontinental interior and summer drying of the California coast, but not northwest along the coast toAlaska. Similar orogenic separation during the late Mesozoic may have produced a seasonally dry andstressed climate in the continental interior (Lamberson, et al., in preparation), while the Bowser Basinclimate remained humid and stable.SEDIMENT COMPACTION AND ISOSTATIC LOADING AS MECHANISMS OF BOWSER BASINFORMATIONSubsidence is required in order to accommodate the thick shallow marine, deltaic, and other nearsea-level coastal plain facies in the northern Bowser Basin. A plausible cause for the subsidence isprovided by the mechanisms of sediment compaction and isostatic loading, which are inevitably involvedin creating subsidence in any large sediment pile such as the Bowser Basin. By assuming Airy isostaticadjustment, and adopting geologically reasonable values for sediment thickness, compaction, and density,a model of the amount of subsidence attributable to sediment compaction and isostatic adjustment iscalculated below (Fig. 2-17).54INITIALBASINWATERDEPTHDENSITY^) NEAR SEA LEVEL ACCOMMODATION SPACEsediment(swatwater^Atdoc•^mantlew11.3MI a)COMPACTION*Am. °O'"°'0""compleY^2C00ISOSTATICLOADCOMPACTION+ LOAD4300 m 1.0^2.3^3.3 500 m 1800 m 2100 m2.5 945 945•41518741518790 9040 40Total kostatic load and compaction subsides= 3800mI^3000 m^I 1.0^2.4^3.3 so() 1280 1700• 2.5 ••792365792385180 18372 7233 33I Total isostatic load and compaction subsides= 3200e.I^2000 m^I 1.0^2.0^3.3 500 11002.5 500 500225 225101 10145 4520 20I Total Isostatic load and compaction subsidence 2000..„loos im ^1.0^1.8^3.3 250 250 850112 11251 5123 2310 105 5I Total isostatic load and compaction subsidence 830Sample calculation:4000 m Initial water depth compaction+ loadsubsidencedensity - densitysed.^waterXAvailableaccommodationspace+ compaction 011.11111•••••11density mantleline 1: [(2.3 - 1.0)/ 3.3] x 4000 m = 1600 m + 500 m compaction = 2100 mline 2: [(2.5 - 1.0)/ 3.3] x 2100 m compacted = 945 m accommodation spaceNote: available accommodation space = Initial water depth in line 1and compaction + load subsidence in line 2Figure 2-17: Calculation parameters and results for isostatic load plus sediment compaction model ofBowser Basin tectonic development. Total of compaction plus load equals predicted subsidence of basin(created accommodation space) due to sediment accumulation. Following procedures described in Allenand Allen (1990).55The most significant initial condition for a basin filling model is the starting depth of the BowserBasin, which is only loosely constrained. However, lithofacies clearly indicate Bowser Basin depositionbegan in a deep marine environment (Ricketts, 1990). Support for an initial deep marine setting comesfrom: 1) provenance studies which place the Bowser Basin adjacent to obducted oceanic terranes(Chapters 3 and 4); 2) thermal maturation modelling, which requires high paleogeothermal gradientsconsistent with oceanic basement (Chapter 5); and 3) the deep basinal character of presumably underlyingSpatsizi Group sediments (Thomson, et al. , 1986). Considering the long-recognized fundamentaltectonic division of most of the earth's surface into oceanic basins greater than 4000 m deep orcontinental platforms within 100 m of sea-level (c.f. Wegener, 1929), an initial depth as great as 4000 mmay be reasonable. Because of uncertainties, however, subsidence models are calculated for initial basinsdepths of 4000 m, 3000 m, 2000 m, and 1000 m. Other assumptions, including mantle, crustal, andsediment densities, thickness of stratigraphic units and amount of compaction are detailed in thesubsidence calculation described in figure 2-17. From these calculations, an initial water depth of 4000m leads to creation of approximately 3800 m of accommodation space near sea-level, enough to accountfor accumulation of the entire preserved near sea-level stratigraphic section in the Bowser Basin, plusanother 700 m of strata burying the Devils Claw Formation. Three thousand metres initial water depthleads to 3200 m accommodation space, just enough to account for accumulation of the preserved nearsea-level strata. Two thousand metres initial water depth creates 2000 m accommodation space, which issufficient to explain accumulation of the undivided Bowser Lake Group, Currier Formation, and half ofthe McEvoy Formation. One thousand metres initial water depth leads to only 850 m accommodationspace, which can account for accumulation of only MFA and the lower Currier Formation. Based onthese calculations, it can be concluded that initial water depths greater than 3000 m adequately accountfor the subsidence required by the preserved near sea-level sedimentary strata in the northern BowserBasin, merely as a result of isostatic adjustment and sediment compaction. From another perspective,unless the pre-existing basin was shallower than 3000 m, there is no need to invoke tectonic mechanismsother than compaction and isostatic adjustment, such as the thrust loading suggested by Ricketts et a/.(1993), to account for Bowser Basin sediment accumulation.56Subsidence and sediment accumulation in the Bowser Basin implies uplift and erosion of thesediment source area. As described above and in chapters 3 and 4, the provenance consisted of obductedisland arc and marginal marine lithosphere located north and east of the Bowser Basin. Sandstonepetrofacies in successive stratigraphic units, furthermore, suggest that the provenance evolved to greaterdiversity of rock types during later McEvoy and Devils Claw Formation accumulation. The causes ofcontinual provenance uplift are likely myriad. Isostatic rebound as a result of erosion of materialdestined for the Bowser Basin probably contributed to the uplift. The erosion implied by sedimentaccumulation in the Bowser Basin may therefore have contributed to the tectonic evolution of the WesternCanadian Cordillera.57REFERENCESAigner, T. 1985. Storm depositional systems, dynamic stratigraphy in modern and ancient shallowmarine sequences. In Friedman, G. M., Neugebauer, H. I., and Seilacher, A. (eds.), LectureNotes in Earth Sciences. Springer Verlag, Berlin. 174 p.Allen. P. A. and Allen, J. R., 1990. Basin Analysis. Blackwell Scientific Publications, Cambridge,Massachusetts. 451 p.Bhattacharya, J. P. and Walker, R. G., 1992. Deltas, In Walker, R. G., and James, N. P, (eds.) FaciesModels: response to sea level change. Geological Association of Canada p. 157-177.Blatt, H., Middleton, G., and Murray, R., 1972. Origin of Sedimentary Rocks. Prentice-Hall Inc.,Englewood Cliffs, New Jersey. 634 p.Boyd, R., and Penland, S., 1988. A geomorphic model for Mississippi delta evolution: Gulf CoastAssociation of Geological Societies, Transactions, v. 38, p. 443-452Bustin, R. M., and McKenzie, K. J., 1989. Stratigraphy and depositional environments of the SustutGroup, southern Sustut Basin, northcentral British Columbia. Bulletin of Canadian PetroleumGeology, v. 31, p. 231-245.Collinson, J. D., and Thompson, D. B., 1989. Sedimentary Structures. Unwin Hyman, Boston. 207 p.Cookenboo, H. 0. 1989. Lithostratigraphy, palynostratigraphy, and sedimentology of the northernSkeena Mountains and their implications to the tectonic history of the Canadian Cordillera.M.Sc. thesis, University of British Columbia, Vancouver British Columbia. 131 p.Cookenboo, H. 0., and Bustin, R. M. 1990. Lithostratigraphy of the northern Skeena Mountains,British Columbia: Current Research, Part F, Geological Survey of Canada Paper 90-1F, p. 151-156.Coleman, J. M., 1982. Deltas. International Human Resources Development Corporation. Boston. 124P.Creber, G. T. and Chalconer, W., 1985. Tree growth in the Mesozoic and Early Tertiary and thereconstruction of Palaeoclimates. Palaeogeography, Palaeoclimatology, Palaeoecology, v. 52, p.35-60.Eisbacher, G. H., 1974. Sedimentary and tectonic evolution of the Sustut and Sifton Basins, north-central British Columbia. Geological Survey of Canada Paper 73-31, 57 p.58Evenchick, C. A., and Green, G. M., 1990. Structural style and stratigraphy of southwest Spatsizi maparea, British Columbia. In Current Research, Part F, Geological Survey of Canada, Paper 90-1F, p. 135-144.Flores, R. M., 1985. Coal deposits in Cretaceous and Tertiary fluvial systems of the Rocky MountainRegion. In Flores, R. M., Etheridge, F. G., Miall, A. D., Galloway, W. E. and Fouch, T. D.(eds.), Recognition of fluvial depositional systems and their resource potential. Society ofEconomic Paleontologists and Mineralogists Short Course Notes v. 19, p. 167-216.Gordon, I., and Heller, P. L., 1993. Evaluating major controls on basinal stratigraphy, Pine Valley,Nevada: implications for syntectonic deposition. Geologic Society of America Bulletin, v. 105,p. 47-55.Gulf Canada Resources Limited 1987. Lost Fox property. British Columbia Ministry of Energy, Minesand Petroleum Resources, Open File Report 723.Gulf Canada Resources Limited 1984. Mount Klappan property. British Columbia Ministry of Energy,Mines and Petroleum Resources, Open File Report 111.Habicht, J. K. A., 1979. Paleoclimate, paleomagnetism, and continental drift. Studies in Geology v. 9,16 p.Heller, P. L. and Paola, C., 1992. The large-scale dynamics of grain-size variation in alluvial basins, 2:Application to syntectonic conglomerate. Basin Research, v. 4, p. 91-102.Jeletzky, 0. L., 1976. Preliminary report on stratigraphy and depositional history of Middle to UpperJurassic strata in McConnell Creek map-area (94D west half), British Columbia. In Report ofActivities, Part A, Geological Survey of CAnlida, Paper 76-1A, p. 63-67.Kimura, T. and Sekido, S., 1975. Nilssoniocladus n. genus (Nilssoniaceae n. fam), newly found fromthe early Lower Cretaceous of Japan. Palaeontographica Abt. B., v. 153, p. 111-118.Lamberson, M. L., Bustin, R. M., Pratt, K. C., and Kalkreuth, W., in prep. The formation of inertinite-rich pests in the mid-Cretaceous Gates Formation: implications for the interpretation of mid-Albian history of paleowildfire.MacLeod, S. E., and Hills, L. V. 1990. Conformable Late Jurassic (Oxfordian) to Early Cretaceousstrata, northern Bowser Basin, British Columbia: A sedimentological and paleontological model.Canadian Journal of Earth Sciences, v, 27, p. 988-998.Moffat, I. W. 1985. The nature and timing of deformational events and organic and inorganicmetamorphism in the northern Groundhog Coalfield: implications for the tectonic history of theBowser Basin. Ph.D thesis, University of British Columbia, Vancouver, B.C.59Miall, A. D. 1985. Multiple-channel bedload rivers. In Flores, R. M., Etheridge, F. G., Miall, A. D.,Galloway, W. E. and Fouch, T. D. (eds.), Recognition of fluvial depositional systems and theirresource potential. Society of Economic Paleontologists and Mineralogists Short Course Notesv. 19, p. 83-100.Nelson, C. H., 1982. Modern shallow-water graded sand layers from storm surges, Bering shelf: amimic of Bouma sequences and turbidite systems. Journal of Sedimentary Petrology, v. 52, p.537-545.Paola, C., 1988. Subsidence and gravel transport in alluvial basins. In K. L. Kleinspehn, and C. Paola(eds..), New Perspectives in Basin Analysis. Springer-Verlag, New York. p. 231-243.Pemberton and MacEachern, 1992. Trace fossil facies models: environmental and allostratigraphicsignificance. In Walker, R. G. and James, N. P. (eds.), Facies Models, Geological Associationof Canada, p. 47 -72.Poulton, T. P. , Callomon, J. H., and Hall, R. L., 1991. Bathonian through Oxfordian (Middle andUpper Jurassic) marine macrofossil assemblages and correlations, Bowser Lake Group, west-central Spatsizi map area, northwestern British Columbia. Current Research, Part A, GeologicalSurvey of Canada Paper 91-1A, p. 59-63.Reynolds, D. J., Steckler, M. S., and Coakley, B. J., 1991. The role of sediment load in sequencestratigraphy: the influence of flexural isostasy and compaction. Journal of GeophysicalResearch, v. 96, p. 6931-6949.Richards T. A. and Jeletzky, 0. L, 1974. A preliminary study of the Upper Jurassic Bowser Assemblagein the Hazelton west half map-area, British Columbia (93M-W 1/2). Geological Survey ofCanada Paper 75-1A, p. 31-36.Ricketts, B. D. 1990, A preliminary account of sedimentation in the lower Bowser Lake Group, northernBritish Columbia, in Current Research, Part F, Geological Survey of Canada, Paper 90-1F, p.145-150.Ricketts, B. D., Evenchick, C. A., Anderson, R. G., and Murphy, D. C., 1993. Bowser Basin, northernBritish Columbia: constraints on the timing of initial subsidence and Stikinia-North Americaterrane interactions. Geology, v. 20, p. 1119-1122.Ruddiman, W. E. and Kutzbach, J. E., 1991. Late Cenozoic uplift and climate change. Transactions ofthe Royal Society of Edinburgh: Earth Sciences, v. 81, p. 301-319.Sellers, W. D., 1965. Physical Climatology. The University of Chicago Press.Spicer, R. A., 1989. Physiological characteristics of land plants in relation to environment through time60Transactions of the Royal Society of Edinburgh: Earth Sciences, v. 80, p. 321-329.Spicer, R. A., and Parrish, J. T., 1990. Latest Cretaceous woods of the central North Slope, Alaska.Palaeontology, v. 33, Part 1, p. 225-242.Thomson, R. C. Smith, P. L. and Tipper, H. W. 1986. Lower to Middle Jurassic (Pliensbachian toBajocian) stratigraphy of the northern Spatsizi area, north-central British Columbia. CanadianJournal of Earth Sciences, v. 23, p. 1963-1973.Tipper, H. W. and Richards, T. A., 1976. Jurassic stratigraphy and history of north-central BritishColumbia. Geological Survey of Canada, Bulletin 270, 73 p.Tye, R. S. and Coleman, J. M., 1989. Depositional processes and stratigraphy of fluvially dominatedlacustrine deltas: Mississippi delta plain. Journal of Sedimentary Petrology, v. 59, p. 973-996.Wegener, A., 1929. The origin of continents and oceans. Fourth edition. Dover Publishing, Inc., NewYork. 246 p.Ziegler, A. M., Raymond, A. L., Gierlowski, T. C., Horrell, M. A. Rowley, D. B. and Lottes, A. L.,1987. Coal, climate, and climate, and terrestrial productivity: the present and early Cretaceouscompared. In Scott, A. C. (ed.), Coal and Coal-bearing Strata: Recent Advances, GeologicalSociety Special Publication v. 32, p. 25-49.61CHAPTER 3RECORD OF OROGENY IN MESOZOIC SANDSTONES OF ME BOWSER BASINABSTRACTThree compositionally distinct, lithic-rich petrofacies occur in Upper Jurassic and Lower Cretaceoussandstones of the northern Bowser Basin. Petrofacies 1 (Pl; Quartz ^(Qt) =34%, Feldspar (F)=14%, Lithics (L) =52%; Quartzinonmystall., (Q.) =9%, F=14%, Lithictotai (Lt) =77%) is volcaniclithic rich with subequal to minor chert, minor monocrystalline quartz (generally <10%), and 10 to 25%feldspar. P1 is found in the oldest samples studied from Currier Formation and older Bowser LakeGroup sandstones. Stratigraphically above P1, sandstones are chert rich, and contain fewer volcaniclithic components. Petrofacies 2 (Qt=62%, F=5%, L=33%; Q.=5%, F=5%, Lt=89%) is rich inchert and poor in feldspar, and has a strong volcanic lithic component and few metamorphic grains likePl. P2 has been found only in the lower McEvoy Formation. The most chert rich samples come fromthe upper McEvoy and overlying Devils Claw Formations, the youngest stratigraphic units exposed inthe northern Bowser Basin, and are designated petrofacies 3 (P3) in this study. Petrofacies 3 (P3;Qt =64%, F=5%, L=31 %; Q.=7%, F=5%, Lt=88%) is dominantly chert, with a lesser volcaniclithic content. P3 also has lower feldspar content than Pl, and a small but significant component(<10%) of metamorphic lithic fragments. Chromite and serpentine are accessory minerals found in allpetrofacies. Altered muscovite occurs in P3.The provenance for PI is interpreted as obducted sea floor and island arc terranes. P2 isinterpreted having the same source, but with the addition of recycled PI sediments. P3, like P2,combines first cycle ocean floor and island arc provenance with recycled sediments of the same type(such as are preserved in P1). Recycled sediments present in P2 and P3 suggest that uplift of the basinmargins may have occurred after deposition of Currier Formation and older Bowser Lake Group (P1),and prior to accumulation of the lower McEvoy Formation (P2). A partly metamorphic and plutonicsource for middle McEvoy through Devils Claw formation (P3) may record early stages of dissection of62the island arc that earlier provided the source of P1 and P2.Independent geological and sedimentological evidence suggest that change in climate, relief,mechanical abrasion or burial diagenesis are unlikely to have significantly affected sandstonecompositions.Comparison with reports of other Cordilleran sandstones demonstrate that chert and volcaniclithic rich sandstones are widespread features of Mesozoic Cordilleran basins. The widespreadsimilarities suggest similar provenance tectonic setting, and that an extensive region of obducted oceanicterranes centered today around the Omineca geanticline was the sediment source for these basins.INTRODUCTIONOne of the most profound geologic events of the Mesozoic was the rifting of the Atlantic Ocean. Thisrifting broke apart Pangea and caused the American Continents to override great expanses of Pacificocean floor. In the process, island arcs were accreted, marginal seAs were closed, and the westerncontinental margin developed into the extended orogenic belt of the Cordillera (Coney et al., 1980).Climate patterns, depositional systems and even the evolution of life were all significantly changed byevents resulting from rifting of the Atlantic. Record of this paroxysm remains preserved in sedimentarybasins both adjacent to and removed from the active orogenic belt. Detailed study of the sedimentaryrecord is called for because of its potential to illuminate the major orogenic events of Mesozoic geologichistory.Intermontane basins contain the most detailed records of the development of the Mesozoicorogenic belt by virtue of their central location within the Cordillera. However, basins within theIntermontane Belt of the Canadian Cordillera are complexly deformed, and of little economic interest,and have therefore received less detailed study than their sedimentary record deserves. In contrast, theWestern Canada Sedimentary Basin (WCSB) east of the Canadian Cordilleran orogenic belt has beenstudied in detail due to its prolific hydrocarbon potential and relatively undisturbed stratigraphy.63Petrographic characterization of Upper Jurassic and Lower Cretaceous sandstones exposed in thenorthern portion of the Bowser Basin, largest of the basins of the Intermontane Belt of western Canada,was undertaken in an attempt to read part of the record of Cordilleran orogeny. Point-count data andobservations of accessory grains are presented herein, and form the basis for interpretations of sandstoneprovenance. Comparison of the Bowser Basin data with other data sets, including Dickinson andSuczek's (1979) worldwide compendium, sheds light on Cordilleran tectonic development, and allowssome inferences to be made concerning events directly related to the Mesozoic rifting of the Atlanticocean.PREVIOUS WORK ON PROVENANCEMost previous work concerning the provenance of Bowser Basin sediments has concentrated onconglomerate pebble compositions and current directions. Eisbacher (1981), following earlier work ofMalloch (1914), related chert pebbles to erosion of Cache Creek Group strata exposed to the north andnortheast in the Atlin terrane, and makes brief reference to Bowser Lake Group sandstones composed ofapproximately 50% volcanic fragments, 40% chert, and 10% quartz. Such pebbles have yieldedradiolaria as young as Early to Middle Jurassic (Pliensbachian to Bajocian; Cordey, pers. comm. cited inRicketts and Evenchick, 1991), consistent with the youngest known strata of the Cache Creek Group(Cordey et al., 1987). Paleocurrent directions reported by Eisbacher (1981) and supported by data inCookenboo (1989) indicate a generally northeastern source area. More recently, Green (1991) reportedthat almost exclusively chert pebble conglomerate of grey, black and green clast colour occurs in theundivided marine Bowser Lake Group strata exposed north of (and presumably stratigraphically below)rocks sampled in this study. Pebble counts made in the study area reveal similarly oligomictconglomerates of all shades of white, grey, green and black chert.Isotope ratios have been employed to produce a novel type of provenance data. Neodymiumisotopes from Jurassic Bowser Basin elastic strata indicate a predominantly juvenile magma source withminor but steadily increasing content of continental material higher in the section (Samson, et al., 1989).64This isotopic data is consistent with an island arc source mixed either with a few percent continentalmaterial or a larger fraction of intermediate material (Samson, et al., 1989).COUNTING PROCEDUREGrains were point-counted by the traditional method which considers each grain individually,using a step size greater than particle size. Well to very well sorted medium grained sandstones werecounted to reduce the influence of grain size. A small number of sandstones with mean apparent graindiameter outside the medium sand (250 to 500 pm) size class were also counted, and are denoted byasterisk in data presentation. The data should be directly comparable with other data derived by use ofthe Gazzi-Dickinson point count method, because in the northern Bowser Basin, lithic fragments areconsistently very finely crystalline, and do not contain a significant proportion of sand-sized crystals(larger than 62.5 pm).At least three hundreds grains were counted per sample, with grains assigned to one of ninecategories as described in Table 1. Three hundred counts per thin section leaves significant error marginsat the 99 % confidence level, especially when considering percentages of less abundant components. Forexample a count of 15 out of 300 (5%) has error margins at 99 % confidence levels of plus 3 % and minus2%, i.e. there is a 99% chance that the actual population has between 3 and 8% of whatever componentwas counted. To halve the 99% confidence interval for a 5% component to between 4 and 6.5%requires increasing the count by four times to 1200 grains (Cheeney, 1982).Burial compaction has severely altered many sandstones, especially from the lower part of theBowser Basin section, where concavo/convex and sutured grain contacts and stylolitization attributable topressure solution are common. As a result, samples exhibiting signs of extreme compaction wereavoided for point-counting. Samples preferred for point-counting were those (relatively rare) that exhibitpore-filling cementation which preserved depositional texture. The best such examples of pore-fillingcementation are those that contain a poikilitic calcite cement or isopachous chlorite cement whichpreserve original grain geometry.65Samples from the deepest portion of the stratigraphic column in the Bowser Basin tend to bemore affected by burial compaction, and many such samples are unsuitable for provenance determination.Pervasive alteration, replacement by calcite, and pressure solution effects, all become more pronouncedwith increasing depth in the section. Such effects are most destructive in fine and very fine grainsandstones, and the overall finer grain nature of Currier and older Bowser Lake Group rocks severelylimits sample selection. In the oldest rocks examined, provenance signal becomes obscured bydevelopment of slatey fabric in many grains. This fabric is recognizable by preferred alignment ofbirefringent clay minerals and can be pervasive in the most deeply buried samples. This alteration maymark the depth of burial limits to which provenance signal is decipherable in the Bowser Basin.Accessory grains with potential tectonic significance were noted during thin sectionexamination. Compositions were later confirmed if necessary by examination with combined scanningelectron microscopy and energy dispersive x-ray (SEM-EDS) system. Chromian spinel is the bestexample of such a tectonically significant accessory mineral identified in this study. Further study ofchromian spinels from some of these sandstones involving microprobe analysis of chemical compositionsis detailed in Chapter 4 of this thesis.Feldspars were tentatively classified based on twinning styles during point-counting. Referencesections were stained for potassium feldspar using sodium cobaltinitrite. Both methods showed that nopotassium feldspar was present in most samples. Due to a lack of significant potassium feldspar content(except in certain stratigraphic intervals as discussed later), no further effort was deemed useful inattempting to refine the percentages of plagioclase versus potassium feldspar.Grain types countedChert: Grains called chert in this study are micro- or cryptocrystalline quartz with no visible feldsparphenocrysts. Chert occurs in colors ranging from clear to brown in thin section, and is more or lessargillaceous, forming a continuum with siliceous mudstone (argillite), counted as lithic sedimentaryfragments. Crystallites are typically intergrown and sutured, and average less than 30 pm in diameter.66Distinction between chert and siliceous mudstone is made on the criteria of overall low birefringence,and high resistance to compaction deformation for grains counted as chert. This definition of chertdiffers somewhat from Dickinson (1985), who counts all argillaceous siliceous grains as lithicsedimentary fragments. We believe counting argillaceous chert sand grains as chert conforms better toterminology applied outside sandstone provenance studies, such as in works referring to bedded chert orconglomerate pebbles (e.g. Pollack, 1987; Sugitani et al., 1991).Volcanic lithic fragments: Volcanic lithic fragments are generally finely crystalline with includedmicrolites of fine grain plagioclase lathes and a trachytic texture (Fig. 3-1). Less commonly volcaniclithic fragments contain quartz phenocrysts. Grains are more or less altered to chlorite. Very finegrained micro- or cryptocrystalline grains were called volcanic lithic fragments if direct indications ofvolcanic origin such as plagioclase lathes or other microlites were identified. Chloritized grains werecounted as volcanic lithic fragments if any internal volcanogenic texture was recognized.Metavolcanics: Green to yellow-brown grains that commonly have anomalously low birefringence occurpervasively as minor components in Bowser Basin sandstones. These grains include detrital chlorite andrare serpentine and were counted as metavolcanics.Sedimentary lithic fragments: Sedimentary lithic fragments include mudstones, claystones andsiliceous argillite. They form a continuum with low rank metasedimentary lithic fragments from whichthey are distinguished by lack of significant clay alignment There is also a continuum betweensedimentary lithic fragments and argillaceous chert. Sedimentary lithic fragments are distinguished byhigher birefringence and generally lower resistance to compaction (i.e. usually compacted grains).67Figure 3-1: Photomicrographs of typical volcanic lithic and chert grains with poikilitic calcite cement.68No attempt has been made to distinguish intra-basinal from more distantly transported clasts. Locallyderived mudstone rip-up clasts combine with provenance derived sedimentary lithic fragments in widelyvarying degrees in different samples, leading to wide variance in abundance throughout the section.Monocrystalline quartz: Monocrystalline quartz is counted for single crystal quartz grains. Both singleextinction (less than 5°) and undulatory extinction (greater than 5°) are common. Quartz grainscomposed of multiple crystals larger than 62 pm are counted as monocrystalline quartz as well, in orderfor this category to conform with Q m of Gazzi-Dickinson point counts.Polycrystalline quartz: Polycrystalline quartz refers to all lithic quartz grains composed of more than 1crystallite averaging between 30 and 62 pm. Chert combined with polycrystalline quartz areapproximately equivalent to Qp of Gazzi-Dickinson point counts.Feldspar: Feldspar was counted for any low birefringence monomineralic grain displaying twinning,cleavage near 90°, or evidence of replacement characteristic of feldspars. Twinning type was notedwhere present, but estimations of plagioclase to potsissium feldspar ratios depended on examination ofsodium cobaltinitrite stained reference sections.Replaced grains (alterites): A grain was counted as replaced or "alterite" if replacement was toocomplete to allow confident assignment to original grain type. Replacement is commonly by carbonate,megaquartz, microcrystalline quartz, chlorite, illite, sericite or opaques. Replacement counts arepresented as percent of the total grain count in the results section, but are not included in Q mFLt or QtFLcalculations.RESULTSThe results of the point-counts are summarized in table 3-1. All of the samples are rich in chert and/orvolcano-lithic fragments, and relatively lacking in monocrystalline quartz, except for two samples ofSkeena Group green sandstones.69Table 3-1: Point count results for sandstones in the northern Bowser Basin. Abbreviations are describedin caption for figure 3-2. The table is arranged with samples in stratigraphic order, starting with oldestat the base.PETROFACIES 3 (MIDDLE AND UPPER MCEVOY AND DEVILS CLAW FORMATIONS)^QZmono QZPc F(p +k) CHERT VRF^SRF^MVRF^MRF^ALTHOC 18-10^17.3%^5.2%^5.8%^42.9%^4.9%^6.4%^2.4%^2.1%^13.1%HOC 18-6^3.8%^5.0%^3.1%^59.6%^12.3%^6.9%^3.5%^4.7%^1.2%HOC 34-19^5.4%^3.3%^3.7%^46.5%^12.7%^6.4%^2.3%^6.0%^13.7%HOC 35-23^4.4%^6.8%^2.4%^52.7%^11.9%^9.5%^1.7%^5.8%^4.8%HOC 19-8^5.1%^1.9%^4.8%^50.3%^16.8%^11.4%^2.5%^5.4%^1.9%G H7-40^13.0%^6.5%^5.1%^46.9%^5.8%^6.8%^3.8%^9.6%^2.4%HOC 14-8^4.5%^4.8%^2.6%^50.0%^18.5%^7.6%^1.9%^4.1%^6.1%*HOC 8- 1^4.9%^4.2%^5.6%^38.0%^14.1%^13.4%^2.1%^5.6%^12.0%HOC 12-10^4.8%^2.9%^5.2%^46.8% 19.4%^9.4%^3.2%^2.6%^5.8%HOC 33-16^5.5%^6.1%^6.3%^43.3%^11.5%^8.2%^0.3%^4.2%^14.5%PETROFACIES 2 (LOWER MCEVOY FORMATION)^QZmono QZPc F(p + k) CHERT VRF^SRF MVRF MRF^ALTHOC 19-5^7.5%^2.9%^3.3%^47.4% 26.5%^5.2%^2.3%^1.3%^3.6%HOC 15-9^3.2%^5.3%^5.7%^50.9% 27.0%^4.2%^0.0%^1.1%^2.8%HOC 16-1^4.3%^7.4%^6.3%^42.8% 25.8%^2.7%^0.0%^0.7%^10.0%HOC 33-7^5.0%^6.0%^4.6%^51.0% 22.7%^0.3%^2.3%^1.7%^6.3%PETROFACIES 1 (CURRIER FORMATION)^QZmono QZPc F(p +k) CHERT VRF^SRF MVRF MRF^ALTHOC 14-1^18.0%^2.7%^13.0% 27.0% 29.3%^3.7%^4.7%^0.0%^1.7%HOC 8-11^12.1%^3.3%^9.1%^13.4% 32.2%^2.3%^3.3%^3.3%^21.2%HOC 8-8^4.2%^3.8%^7.5%^20.0% 39.6% 14.2%^3.8%^1.7%^5.4%HOC 15-7^7.5%^4.9%^16.7%^13.1% 35.1%^8.9%^2.3%^0.3%^11.1%HOC 30-8^8.6%^1.5%^22.8%^10.5% 41.7%^5.2%^4.0%^1.9%^3.7%HOC 30-6^2.9%^1.9%^11.1%^27.9%^42.5%^5.1%^1.3%^0.3%^7.0%HOC 30-3^4.1%^0.3%^10.4%^14.2% 59.0%^4.7%^2.5%^0.3%^4.4%HOC 27-1^7.2%^1.4%^23.8%^17.5% 22.4%^2.0%^0.6%^1.7%^23.3%HOC 22-6^5.4%^2.4%^2.7%^42.2% 34.8%^10.8%^0.0%^0.0%^1.7%SAMPLES THAT DO NOT FIT PETROFACIES AS DESCRIBED IN TEXT^QZmono QZPc F(p + k) CHERT VRF^SRF MVRF MRF^ALTHOC 19-2^0.6%^2.6%^6.4%^15.2%^59.7%^5.5%^1.9%^1.3%^6.8%HOC 17-7^10.6%^1.6%^16.0%^14.7%^38.8%^2.5%^0.9%^0.3%^15.6%HOC 13-3^8.7%^2.0%^13.4%^22.1%^24.1%^8.1%^7.3%^1.2%^13.1%SKEENA GROUP AT MOSQUE MOUNTAIN^QZmono QZPc F(p + k) CHERT VRF^SRF MVRF MRF^ALT^HOC 30-12 39.2%^13.7%^9.6%^13.2%^7.3%^3.2%^2.3%^5.6%^5.8%HOC 30-10 32.8%^13.2%^12.0%^1.9%^5.0%^0.3%^8.8%^8.5%^17.4%70The most striking result of the point-counts is the identification of three distinct grain typepopulations, named herein petrofacies 1, 2 and 3, each characterizing a particular stratigraphic interval.Composition of the petrofacies is described in the following section and is followed by interpretation ofprovenance. Inherent in the provenance interpretations is the assumption that provenance controlssandstone composition. Most researchers have concluded that provenance exerts the strongest controlover composition, and provenance interpretations have been made accordingly. However, other factorsincluding climate, relief, transport processes, alluvial storage, recycling and burial diagenesis clearly arecapable of influencing composition. Consideration is given to these other factors and their likely effectson compositions of the studied sandstones in a separate section following the petrofacies descriptions andinterpretations.Petrofacies 1Petrofacies 1 (P1) is present in Currier and older undivided Bowser Lake Group rocks, and isdominated by volcanic lithic fragments, with subequal to minor chert, low monocrystalline quartz(<10% except in samples #14-1 and 8-11), and about 10-25 % feldspar (Fig. 3-2). Volcanic lithicfragments typically have trachytic texture with more or less aligned lathe-like plagioclase phenocrysts setin an aphanitic groundmass, and are variously fresh to chloritized. Quartz and vitric shards are generallyabsent from volcanic lithic fragments, although some euhedral monocrystalline quartz of possiblevolcanic origin occur. In the two samples with the greatest concentration of quartz (samples 14-1 and 8-11), mono- and polycrystalline quartz were point-counted separately by the methods outlined by Basu etal. (1975), to better characterize the origin of quartz. Non-undulatory monocrystalline quartz dominatedother types by at least a factor of four to one, and polycrystalline quartz was mostly composed of muchgreater than 4 crystallites per grain.Chert in P1 only rarely contains definitive radiolarian or sponge spicule ghosts, and therefore anorigin from devitrified volcanics for a portion of the chert in P1 can not be discounted. The totalvolcanic signal may thus include some portion of the chert and quartz grains, and therefore may be71Sample number14-1EN= 8-116:= 8-815-7CS:2E3 22-627-1L^I 30-3Jl 125: 30-630-8Media. nValueYYPetrofacies 1r7.7,779F(p+k)^CH^VRF^SRFGRAIN TYPES771.72 27:1i:!XXMRF MVRFQm^Qpc AltFigure 3-2: Composition of Petrofacies 1. Detrital modes of nine grain types plotted against percent ofsample. Qm= monocrystalline quartz; Q pc = polycrystalline quartz (crystallites between 30 and 62 pm);F63+k)---  total feldspar; CH= chert (crystallites less than 30 pm); VRF= volcanic lithic fragments;SgF'= sedimentary lithic fragments not including chert; MRF= metamorphic lithic fragments; MVRF=metavolcanics including chlorite and serpentine; ALT= altered grain (unrecognizable origin).MODAL COMPOSITION OF PETROFACIES 172• 500I—• 4000 30u_0z- 20cr• 10greater than the volcanic lithic fragments alone suggest. Unaltered mafic minerals are notably absent,except for biotite in more or less chloritized tabular grains.Metamorphic fragments are rare in P1, and are limited to slatey grains from low grademetapelites.Chromian spinel is a notable accessory mineral in P1 sandstones. It is easily identified based ondark reddish brown (nearly opaque in some cases) color and isotropic properties in thin section.Chromian spinel commonly occurs in ocean floor peridotites, olivine basalts and dunites and is oftenpreserved during the alteration of ocean floor rocks (Hekinian, 1982). It has also been reported as thedominant heavy mineral in sandstones sourced from ophiolites (Stattegger, 1987). Chromian spinetgrains from selected sandstones were further analysed by microprobe and yielded compositions similar tospinels from alpine-type peridotites associated with marginal basin origin (Chapter 4). Zircon is rare inthe Bowser Basin, although zircon is usually the most abundant type of accessory heavy mineral grain insandstones. Typically less than one zircon grain occurs for every five to ten chromian spinels.Serpentine with an anastomosing network texture typical of alteration from olivine (Shelley, 1975), andin some cases containing small (less than 10 micrometre), dispersed chromian spinel crystals, has alsobeen identified as an accessory mineral in some P1 sandstones.Feldspar exhibits the most pronounced variation in Pl. Most arkosic samples are locatedtowards the south and east edge of the study area, which may reflect proximity to a rugged source area.Feldspar composition supports a proximal source to the southeast. Potassium feldspar is absent fromsamples in the northern and central portion of the study area, but is present in minor quantities near theeastern margin of the basin. Potassium feldspar is a common component of many island arc suites, but israre in ocean floor basalts. Concentration of potassium feldspar in the most arkosic samples supportstheir proximity to an island arc source. Much lower feldspar and exceptionally high chert content isfound in sample 22-6. This sample comes from a wave reworked marine bar sand topped by a shell lag.The relatively more mature composition is attributed to mechanical abrasion by wave reworking and thus73provides an example of the effects of intense reworking on the composition of P1 sandstones.Provenance interpretationThe composition of petrofacies 1, dominated by volcanic lithic fragments and sub-equal or lesser amountsof chert (some of undoubted biogenic origin), points to an oceanic provenance. Two types of oceanicsource rocks are inferred from grain composition observations. The observations that quartz isuncommon, feldspar is largely plagioclase, biogenic chert is relatively abundant, and chromian spinel,serpentine and biotite occur as accessory minerals point to a peridotite, other ultramafic or basalticoceanic terrane such as uplifted sea floor. Support for sea floor involvement can be found in the fact thatmuch of the volcanic lithic content is more or less altered to chlorite, which is a common trait of oceanfloor volcanics (Hekinian, 1982). Localized arkosic samples, as well as a minor fresh volcanic lithiccomponent, suggest a volcanic island arc source in addition to the inferred ocean crust source. Similarquartz-poor, feldspathic to arkosic sands derived from an oceanic island arc occur in Central America(Lundberg, 1991). Although modal quartz values are consistently low, most of the non-chert quartz inthe northern Bowser Basin is monocrystalline, and some is euhedral. This composition of quartz iscomparable to volcanic sands described by Girty et al. (1988), and therefore provides further support fora volcanic source.The commonly altered nature of the volcanic lithic fragments indicates the volcanic signal of P1is largely derived from extrabasinal eroded and transported paleovolcanics (Zuffa, 1985). No clearintrabasinal neovolcanic signal was identified in the sandstones, suggesting active volcanism was absentor minor during PI time. Contrary evidence for active volcanism in Currier rocks is provided by drillcores which have encountered several thin tuffaceous layers. However, these tuffs are fine grained andmay represent long distance aerial transport (Gulf Canada Resources Limited, open file reports, 1984,1987). The paleovolcanic signal is consistent with a provenance of uplifted sea floor and (mostly)inactive island arcs. Concentration of more arkosic samples on the southeastern margin of the study areamay point approximately to the location of the inferred arc. In that direction, good candidates for an74older arc source include the Hogem Batholith and related calc-alkaline plutons. The Hogem batholith is alarge intrusive body of mostly calc-alkaline Late Triassic and Early Jurassic age rocks (Garnett, 1978).Similar arc sources that may have contributed sediment to the Bowser Basin from the northwest includethe HoodIluh Batholith (Moffat et al., 1988). Suitable ocean floor rocks include the Cache Creek Groupin the Atlin terrane to the north as originally suggested by Eisbacher (1981), and Slide Mountain terranerocks lying farther to the northeast.Red chert, which is absent from northern Bowser Basin conglomerates, is common in radiolariancherts that formed above open ocean crust (Jones and Murchey, 1986; Sedlock and Isozaki, 1990). Theassociation of red chert and spreading centers is supported by experimental demonstrations that chertturns red when heated above 230° to 290°C in the presence of oxygen (Rick, 1978). Biogenic chertcolors that occur in the Bowser Basin may also be characteristic of particular sea floor depositionalsettings. Grey and green chert are associated with oceanic island arcs, and black chert is associated withcontinental margins, where argillaceous continent detritus dilutes siliceous material and cherts commonlygrade to claystone (Jones and Murchey, 1986).Petrofacies 2Petrofacies 2 (P2) occurs in lower McEvoy Formation sandstones and is dominated by volcanicEthic grains, but is also rich in chert and low in feldspar (Fig. 3-3). P2 has little or no metamorphiccontent, and accessory grains include rare petrified wood and coal spar, in addition to numerousargillaceous intrabasinal sedimentary lithic fragments. The high volcanic lithic content and lack ofmetamorphic indicators are similar to P1 (Currier Formation and older strata), but the high chert contentis similar to younger McEvoy and Devils Claw formations sandstones (described below as petrofacies 3).Because of P2's stratigraphic position and composition, it can best be considered a transitional petrofaciesbetween the volcanic lithic rich P1 sandstones of the Currier Formation and older strata, and the chertrich sandstones of upper McEvoy and Devils Claw formations.75MRF MVRF AltCH^VRF^SRFGRAIN TYPESSample number■I 19-5MedianValuePetrofacies 2CZ=Om F(p+k)coo605040302010Figure 3-3: Petrofacies 2 composition. Plotted parameters are the same as in figure 3-4.MODAL COMPOSITION OF PETROFACIES 276Provenance interpretationPetrofacies 2 is similar in composition to PI with a high volcanic lithic content, low quartz, and little orno metamorphic signal. P2 differs from P1, however, in having a higher chert content. P2 is thereforecompositionally more mature than P1, which suggests recycled PI as a source. Accessory grains such ascoal spar and petrified wood occur rarely in P2, indicating erosion of previously lithified sediments, andare consistent with a recycled PI provenance. Because P2 lacks a metamorphic input, and is restricted tolower McEvoy Formation strata, metamorphic terrane unroofing apparently commenced during mid-McEvoy time.Petrofacies 3Petrofacies 3 (P3) occurs in sandstones from the middle McEvoy Formation through at least themiddle to upper Devils Claw Formation. It is characterized by >40% chert, <20% volcanic lithicfragments, and a small but significant component ( <10%) of metamorphic lithic fragments (Fig. 3-4).Feldspar content drops compared to P1 to less than 6% of total grains, but like Pl, only plagioclasefeldspar is present in most samples. Monocrystalline quartz is more abundant than in PI, and commonlyhas euhedral outlines with few inclusions characteristic of volcanic origin (and sometimes embaymentsand rarely with negative crystal inclusions which can indicate a volcanic origin as well; Scholle, 1982).Volcanic lithic fragments have a trachytic texture like P1, and are the same size and texture asother grains, and only rarely appear fresh. Most are extensively chloritized, as are accessory biotitegrains. Glass shards are very rare components (only one clearly identified), and easily altered maficminerals are absent except in easternmost areas. The volcanic grains may therefore be described asderived from an extrabasinal paleovolcanic source (Zuffa, 1985).Metamorphic lithic fragments occur more commonly in P3 than in PI and are mostlymetapelites, derived from slates and phyllites (Fig. 3-5). Other grains can also be considered part of77500W 4000 30tL020cr1wa.Petrofacies 3Figure 3-4: Petrofacies 3 composition. Plotted parameters are the same as in figure 3-4.MODAL COMPOSITION OF PETROFACIES 378Om^Qpc^F(p+k)^CH^VRF^SRF^-PARF^MVFIF^AltGRAIN TYPESFigure 3-5: Metamorphic grain from P3, probably from either a phyllite or schist. Plain and cross polartransmitted light micrographs. Grain is approximately 350 pm long dimension.79the total metamorphic provenance signal. Some polycrystalline quartz grains have more than 10 stretchedcrystallites with sutured contacts (Fig. 3-6), which are very good indicators of phyllite, schist or highergrade metamorphic provenance (Scholle, 1982). Muscovite, which is abundant in metamorphic terranes,is present in P3 as an accessory mineral that is commonly partially chloritized (Fig. 3-7).Accessory grains of chromian spinel are only found in some P3 sandstones. In samples thatcontain chromian spinels, the composition is very similar to that of P1 chromite, based on microprobestudy (Chapter 4, this thesis). Accessory grains of serpentine, some with an irregular anastomosingnetwork typical of olivine replacement (Shelley, 1975), have also been identified in some sandstones.Provenance interpretationPetrofacies 3 is similar to P1 in that both are lithic rich, but substantial differences are present,implying some changes in provenance. Primarily, P3 has a much higher chert and lower volcanic lithiccontent. Total quartz (combination of monocrystalline quartz, polycrystalline quartz, and chertequivalent to Qt of Dickinson and Suczek, 1979) is close to 70%, and most of the total quartz is chert(45-60%). Volcanic lithic fragments are pervasively chloritized and less common (5-20%) than in P1.Relatively high total quartz content suggests in part a recycled sedimentary provenance. The high chertcontent, pervasive volcanic lithic component and accessory chromian spinel, serpentine and chloritizedbiotite suggests the origin of the recycled sediments was a sea floor and island arc provenance, much likethat for P1. Most likely P3 is in part recycled alluvial facies equivalent to marine and deltaic faciespreserved in P1, and in part first cycle sediments from the same sea floor source rocks.Interpreting P3 exclusively as recycled P1 is not sufficient, however, to account for othercomponents of the grain composition. The clear metamorphic signal composed of polycrystalline quartzgrains with more than ten sutured, stretched crystallite contacts per grain, increased metapelite grainabundance and rare accessory muscovite indicates a metamorphic terrane as part of P3's provenance.Volcanic lithic fragments are less common than P1, as expected if P2 is partly recycled P1, and80Figure 3-6: Polycrystalline quartz grain of probable metamorphic origin. Grain is approximately 550 pmin long dimension.81Figure 3-7: Muscovite grain (crossed polarized light) from Petrofacies 3. Grain is approximately 300pm long dimension82the volcanic lithic grains are mostly attributable to a paleovolcanic source like P1. However, thepresence of rare volcanic lithic fragments with included glass shards and quartz, and euhedralmonocrystalline quartz of probable volcanic origin, suggests minor input from active, probably siliceous,volcanic sources not evident in Pl.Considered together, the grain composition of P3 indicates a provenance terrane consisting ofunroofed metasedimentary facies (including at least slates, phyllites and schists) mixed with both firstcycle and recycled oceanic lithosphere facies. Uplift of previously deeply buried rocks is supported bycollection of a single calc-alkaline quartz diorite pebble (2 cm by 6 cm) from an otherwise oligomictchert pebble conglomerate of the upper McEvoy Formation. Indications within P3 of uplift and activeacid magmatism in the provenance are supported by the presence of increasingly coarse, thick andabundant chert pebble and cobble conglomerate up through the McEvoy and Devils Claw Formations.As with Pl, the Cache Creek Group and Slide Mountain Assemblage along with magmatic arcrocks associated with the Hogem or related batholiths are likely candidates for source. Potassium-argonages from the Hogem batholith, and stocks to the east of it, indicate early to mid- Cretaceous resumptionof increasingly sialic volcanism (Garnett, 1978). The recycled and metamorphic components may reflectuplift associated with dissection and emergence of the Jurassic island arc.Three McEvoy and Devils Claw Formation samples were counted that do not fit in P3 or P2 andwarrant individual consideration. Each of these samples is enriched in volcanic lithic fragments andrelatively depleted in chert compared to P3 or P2, but also differ significantly from Pl.Sample 19-2 is the most volcanic-rich sandstone in either the McEvoy or Devils ClawFormations. It was collected in a basal pebble-granule conglomerate layer of the McEvoy Formation.This basal conglomerate has been correlated to both the north and south of sample 19-2's location, andmay represent a tectonic pulse associated with inception of McEvoy/Devils Claw deposition (Cookenboo,1989) in an analogous manner to the Cadomin Formation of the Blairmore Group in the Western CanadaSedimentary Basin (Varley, 1984).83Sample 17-7, collected from eastern-most McEvoy strata (at Distingue Mountain east of theSkeena River), has a significant accessory mafic mineral assemblage. Another sampled sandstone fromthe same vicinity that was too fine grained to point-count shows a similar mafic assemblage. Sandstonesfrom this locale are closest to the provenance (as indicated by current directions) of any other P2 or P3samples, and may be compositionally less mature as result.Sample 13-3 is the stratigraphically highest sandstone examined from the Devils ClawFormation. This sample is from a grey-green sandstone with accessory fine grained muscovite and rareglaucony grains (possibly glauconitic minerals based on petrography and SEM-EDS- lack of x-raydiffraction data renders glauconitic mineral determination uncertain; Odin and Fullagar, 1988). Sample13-3 is also unusual among sandstones from the study area because it contains potassium feldspar. Thissample may suggest a younger volcanic/tectonic pulse, but further investigation of the upper Devils ClawFormation is needed to be certain of its implications.Two samples of quartz-rich sandstone were counted. Both are from micaceous green sandstonestypical of the Skeena Group (Tipper and Richards, 1976), and both were collected from the northeasternridge of Mosque Mountain, stratigraphically above volcanic lithic rich fluvial sandstones of Pl. Thesesandstones carry a very strong metamorphic provenance signal. Muscovite is common, as arepolycrystalline quartz grains with aligned sutured crystallite boundaries and in some samples includedfine grained mica (sericite). Some of these grains are interpreted as schists and phyllites (Scholle, 1982).Medium sized quartz grains were sufficiently abundant to warrant counting and plotting on a diamonddiscrimination diagram as described by Basu et al. (1975). The quartz plots within the middle and uppermetamorphic provenance among sands derived from garnet through sillimanite rank metamorphics, whichis consistent with the presence of schistose grains and muscovite.OTHER INFLUENCES ON PETROFACIES COMPOSITIONTo this point, interpretation of sandstone petrofacies has assumed provenance alone controls composition.However, as alluded to earlier, composition actually depends on the interplay of factors as diverse as84facies exposed in the provenance, climate, relief, transport processes, alluvial storage, recycling andburial diagenesis. The point-count characterization of sandstone is a combination of signals from each ofthese diverse factors and extracting the provenance signal alone can be a perilous exercise. Mostresearchers have concluded that the strongest individual signal is generally provenance, but that otherfactors can mimic provenance variations is amply demonstrated by reports of first cycle quartz arenitesand sub-arenites produced in modern humid tropical environments (Dutta and Suttner, 1986).That humid weathering controls on sandstone composition are not limited to tropical climateshas been shown by work in the southeastern United States and California (Grantham and Velbel, 1988;Girty, 1991). Basu (1985) demonstrated humid climate is reflected in increased compositional maturityfor both plutonic and metamorphic provenances. Similarly, preferential destruction of labile grains withincreased burial depth, chemical weathering, mechanical abrasion (as in reworked shallow marine sands)or recycling are well established (Blatt, 1982).All these diverse factors can be divided into two classes distinguished by effect on composition:first, an "additive" class consisting only of provenance facies and diagenesis (in which grains have beenaltered so completely they are counted as another grain type); and second, a "destructive" class consistingof all the other factors, which each tend to preferentially destroy labile grains. Because an attempt ismade during counting to recognize alterites, the assumption is made that the "additive" class is entirelycomposed of provenance signal. The provenance signal then becomes less labile grain rich throughexposure to the "destructive" class of factors. That some alterites are not recognized as such in thecounts is suggested by the elevated monocrystalline quartz values for two quartz cemented P1 sandstones(i.e., some quartz cement was counted as quartz grains), however, this is believed not to be a seriousdefect. By making the assumption that provenance signal becomes less strong in labile grains throughtime, two logical tests can be applied to the previously made provenance interpretations.1) If the petrofacies more affected by a particular "destructive" factor has more labile grains thana less affected petrofacies, provenance changes must contribute to the difference.852) If a compositionally more mature petrofacies has types of labile grains not present in a lessmature petrofacies, provenance control is indicated.Passing the above tests permits confidence that qualitative differences between petrofacies reflectprovenance changes rather than variations induced by "destructive" factors:In order to apply these two tests, decisions must be made as to which grains are relatively morestable and which petrofacies has been exposed to the more severe action of each. Qualitative ranking ofrelative stability is all that can be accomplished, because quantitative data for specific lithic grain types ismostly lacking. Although monocrystalline quartz can be assumed to be most stable grain type in themedium sand fraction (Blatt et al. , 1980), relative stability between chert and volcanic lithic grains areless well known, because little is presently known about relative chemical and mechanical stability ofchert versus fine grained volcanics. Grain size certainly plays a role, as shown most convincingly by theoligomict chert pebble composition of the conglomerates compared to the high volcanic lithic fragmentcontent of co-eval sandstones. As grain size is reduced, chert becomes less stable relative to volcaniclithic fragments (and volcanic lithic fragments become important components of sandstone). Despitebeing relatively unstable at very fine or finer grain size, at medium sand size chert is assumed to berelatively more stable than volcanic lithic grains. Metamorphic lithic fragments are considered less stablethan either chert or volcanic lithic fragments. Polycrystalline quartz grains (with sutured crystallites) arerelatively more stable than other metamorphic grains, and therefore make good markers for metamorphicprovenance. Qualitatively, metapelites, sedimentary lithic fragments, chloritized volcanic lithics, andfeldspar are all considered here to be the most labile fraction, with volcanic lithic probably somewhatmore stable, and monocrystalline quartz and chert the most stable fraction.Given the assumed relative stabilities of the major grain types, (Fig. 3-8) the proportion ofstable grains increases upwards in the succession.Consideration will be given below to each of the "destructive" factors. Based on stratigraphicrelationships and sedimentologic interpretations, each petrofacies with more or less confidence can be86ranked in relation to each destructive factor.The highest confidence can be assigned to relative burial effects. The Currier and older BowserLake Group rocks are stratigraphically beneath the McEvoy and Devils Claw Formations, and thereforeare assumed to have experienced greater burial alteration. The validity of this assumption is supportedby the independent observation that in the older sandstones pressure solution effects including stylolitesand spaced cleavage are pervasive.Applying test #1 with respect to burial effects, PI was most deeply buried and P2 next mostdeeply buried, but compositional stability is in the order P3 > P2 > P1; therefore, provenance differences,not burial diagenesis, are accepted as accounting for differences between Pl, P2 and P3.Similarly, PI is the petrofacies likely to be most affected by abrasion because of its partiallyshallow marine depositional character. Choosing between P2 and P3 as likely more affected by abrasionis difficult because both are fluvio-deltaic deposits, but an increase in conglomerate in P3 bearing stratamay suggest less transport and therefore less abrasion for P3. Again, as with relative burial depth, thepetrofacies most affected by abrasion is the most labile, leading to the conclusion that provenancechanges rather than abrasion account for changes in composition. The same reasoning applies to relief inthe source area, if conglomerate composition of associated strata can be read as an indication of sourcearea relief. P1 strata (Currier and older Bowser Lake Group rocks) are less than 3 % conglomerate (instratigraphic thickness), the McEvoy Formation is approximately 6 % conglomerate, and the Devils ClawFormation is nearly 50% conglomerate.87Figure 3-8: Relative stability of grain types assumed in this study.88Quartz(monoaystalline)ChertQuartz(polycrystalline)Volcanic LithicFragmentsFeldsparMetamorphic LithicFragmentsSedimentary LithicFragmentsMeta-Volcanic LithicFragmentsMafic GrainsRELATIVELY STABLECLAST TYPESLABILEFRACTIONJ._Climate is also difficult to evaluate quantitatively, but three factors suggest seasonally humidconditions prevailed throughout Currier, McEvoy and Devils Claw time. First, at least thin seams ofcoal occur throughout the section as does petrified wood with conspicuous development of annual rings.Conspicuous tree rings are botanical evidence of seasonal climate (Creber and Chalconer, 1985) and coalis clear evidence of humid climate. Second, deposition occurred on the western continental marginadjacent to the ocean and west of mountains, a geographical position that is likely to be humid outsidethe tropics. Finally, macroflora species composition remained quite stable throughout the Currier toDevils Claw section (MacLeod and Hills, 1990), supporting the persistence of a similar climate duringdeposition of all examined stratigraphic units. Given evidence for climatic stability, climatic changes arediscounted as a cause of changing grain compositions.Labile metamorphic indicators not seen in P1 or P2 appear in the otherwise compositionallymost mature P3. P3 therefore passes test #2 by possessing labile grains not seen in the other petrofacies,although P3 is otherwise compositionally most mature. This leads to acceptance of the conclusion thatprovenance changes are indicated by the appearance of metamorphic indicators in P3.In summary, all the "destructive" signals seem to be contra-indicated as root causes ofcompositional changes. This conclusion follows from application of the two tests described earlier in thissection and consideration of independent geological evidence applicable to these rocks. Provenancechanges as interpreted from the grain composition can therefore be accepted with reasonable confidenceas the cause of petrofacies variations through time.Therefore three successive stages of provenance evolution (Fig. 3-9) are interpreted based on thepetrofacies compositions: 1) during P1 deposition, first cycle detritus was eroded from uplifted oceaniclithosphere and island arc facies; 2) during P2 time first cycle and recycled sea floor and island arc facieswere eroded; and 3) during P3 time further uplift led to erosion of metamorphic, minor plutonics, andacid volcanic arc strata in addition to continued erosion of first cycle and recycled oceanic lithosphereand island arc facies.89PETROFACIES 1ISLAND ARCAND SEA-FLOORASEMBLAGESFigure 3-9: Evolution of provenance as interpreted from the succession of petrofacies.EVOLUTION OF PROVENANCE PETROFACIES 2PETROFACIES 2AF1/414ISLAND ARCD SEA-FLOORASEMBLAGESISLAND ARC0 SEA-FLOORASEMBLAGESRECYCLED ALLUVIALEQUIVALENTS TO PIRECYCLED ALLUVIALEQUIVALENTS TO P1MINOR METAMORPHICAND PLUTONIC90DEVILS CLAWFORMATIONMcEVOYFORMATIONCURRIERFORMATIONUNDIFFERENTIATEDBOWSEFI LAKEGROUPco0O05PROVENANCE TECTONIC SETTINGPoint-count categories were regrouped in order to plot the grain composition data on Dickinson-Suczekdiagrams (Fig. 3-10; modified from Dickinson, 1985). Percentages for each parameter were thencalculated for the three petrofacies, and mean sample values were plotted on Dickinson and Suczek(1979) type tectonic provenance discrimination diagrams yielding results consistent with theinterpretations made above. Petrofacies 1 (Qt=34 %, F=14%, L=52%; Qm=9 %, F=14%, Lt=77 %)plots in the magmatic arc field of the Q tFL and QmFLt diagrams. Petrofacies 2 (Qt=62 % , F=5%,L=33%; Qm=5%, F=5%, Lt=89%) and petrofacies 3 (Qt=64%, F=5%, L=31%; Qm=7%,F=5%, Lt=88%) plot in the recycled orogen provenance field of those same diagrams, as expected fromthe provenance interpretations. P1 is very much borderline to the magmatic arc field, however, and isonly slightly within the volcanic rather than plutonic field. On the Q mFLt diagram, both petrofacies plotmore completely within their appropriate fields, probably reflecting the postulated volcanic origin forsome of the chert in P1. The volcanic greater than plutonic signal, and increasing ratio of chert to quartzimplied from the diagrams for P1 and P3, respectively, are entirely appropriate based on graincompositions.On the QmPK diagram (not shown), P1 plots within the family of Circum-Pacificvolcanoplutonic suites, as might be anticipated from the Bowser Basin's position within the circum-Pacific orogenic belt of the Cordillera. P3 plots in the field of increasing maturity/stability fromcontinental provenances, which reflects the addition of metamorphic and plutonic provenance terranes torecycled P1 sediments.DISCUSSIONThe following discussion examines implications from this study at three scales: 1) a local scale whichrelates the changing grain compositions through time to Bowser Basin tectonics; 2) a Cordilleran-widescale which places this study in perspective with other information relating to Cordilleran orogenic91i. Rit'C'YCteaoR CHZ EftPROVENACE4P1•14A4MATIC ARCPROVENANCEDETRITAL MODES(flogs OW Dickinson and Suczsk. 1 g731)QtLFigure 3-10: Ternary diagrams of sandstone composition. After Dickinson and Suczek (1979).Qm92Ltdevelopment; and 3) a world-wide scale which relates the filling of the Bowser Basin to rifting of theAtlantic and break up of Pangea.Bowser Basin tectonic implicationsPetrofacies described in this study both support and expand upon earlier conclusions concerning theprovenance of Bowser Basin sediments. Eisbacher (1981) concluded, based on the oligomict chertcomposition of conglomerate clasts, that ocean floor facies of the Cache Creek Group exposed in theAtlin terrane north and east of the Bowser Basin were the most likely source for Bowser Basin sedimentsThis study supports the conclusion that the Bowser Basin sediments were derived from oceanic suites.The sandstone grain compositions demonstrate that sea floor and island arc igneous suites were alsopresent in the provenance, and that the igneous suites were probably inactive ("ancient" at the time oftheir erosion). Uplift and erosion of both sea floor and island arc suites probably implies obduction ontothe North American continental margin by the Early or Middle Jurassic, soon after initiation of rifting ofthe Atlantic Ocean. Such an obducted island arc and associated sea floor provenance is sufficient todescribe the entire source of Currier Formation and older Bowser Lake Group rocks.For the younger rocks of the McEvoy and Devils Claw Formations, however, the sandstonecompositions document a more complex provenance. At least part of the ocean floor source in P2 and P3is recycled sediment, eroded presumably from alluvial equivalents to Currier and older Bowser T AtieGroup strata, and the additional input of unequivocal metamorphic grains necessitates erosion of ametamorphic terrane for part of the provenance. In other words, provenance of McEvoy and DevilsClaw strata was a mixture of oceanic facies, recycled oceanic facies and minor unroofed metamorphicfacies with associated magmatic activity. The most likely source for the metamorphic signal is unroofingof the Omineca Geanticline and the timing of that unroofing, at least in the vicinity of the northernBowser Basin, most likely was not before middle McEvoy deposition (first occurrence of P3).The amount of material eroded can be estimated roughly from the extent and thickness ofpreserved Bowser Basin strata. The Bowser Basin covers roughly 50 000 square kilometres today93(Wheeler and McFeely, 1987), and restoring 35% to 40% post-depositional compressional deformation(Moffat, 1985; Evenchick, 1991), suggests the area at the time of deposition was in excess of 70 000square kms. True stratigraphic thickness is between 5 and 10 kms (Chapter 2), and porosity is virtuallyabsent, therefore, at minimum, a source area the size of the present day Bowser Basin would have beeneroded to a depth of roughly 7 to 14 kms. Modem sediment budget studies suggest that such an estimatesignificantly under-represents the total eroded material from the Bowser Basin provenance. For example,as much as 40% of Yukon River sediment load may have by passed the Yukon delta and shelf to bedeposited more than 300 kms north in the Chukchi Sea (Nelson and Creager, 1977). If similar sedimentbypass conditions held for the Bowser Basin, then the eroded material could have covered an area the sizeof the present day Bowser Basin, eroded to a depth of between 10 and 20 kms.Erosion of pre-existing sedimentary strata is inferred from the high content of chert in P2 andP3, along with a smaller recycled volcanic lithic signal. The most obvious source for recycled sands ofthis composition are up-dip equivalents of Currier and older Bowser Lake Group rocks. Preserved faciesof sediments associated with P1 are shallow marine and deltaic. Alluvial equivalents are unknown, andmay have been eroded to produce the recycled signal in P2 and P3 sandstones. An angular unconformityrecording erosion of Bowser Lake Group rocks has been reported beneath Sustut Group rocks east(depositionally up-dip) of the Bowser Basin (Eisbacher, 1981), suggesting erosion of these rocks is likelyto have contributed to P2 and P3.The oligomict chert composition of the conglomerate clasts contrasts markedly with the variedand changing composition of the sandstones. Chert is the clast type least susceptible to mechanicalabrasion during transport, and lack of less stable clast types in the conglomerates may reflect somedistance of transport. Estimation of the minimum distance of transport can be made considering the lackof less stable clast types in the conglomerates in comparison with modern rivers. Chert gravel in riversdraining plutonic and metamorphic rocks of the Black hills of South Dakota, for example, dominatesover other lithic types after less than 15 kilometres transport (Plumley, 1948). Extrapolation of thesedata suggests lithic gravels other than chert may be virtually eradicated from gravels that receive as little94as 25 kilometres transport under similar conditions. Direct application of these distances to the BowserBasin is probably not valid, due to its dominantly volcanic rather than plutonic/metamorphic source, andthe likely differing conditions of climate and source area relief. However, 25 kilometres of streamtransport as a minimum estimate is revealing, and probably is the minimum necessary to account for theoligomict chert composition. Conglomerate textures also support at least moderate transport distances,because the pebbles and cobbles are well to very well rounded. Unlike volcanic lithic gravels, volcanicsands can dominate the sediment load of very long rivers. Examples of major modern rivers dominatedby volcanic lithics include the Fraser, Colville, Columbia, Magdelena, and Yukon rivers (Potter, 1979).All of these modem volcanic lithic dominated rivers occur in the circum-Pacific region (Potter, 1979),the same as the rivers that fed the Bowser Basin.Petrofacies documented in this study also shed light on relationships between deposits in thenorthern and southern parts of the basin P1 is volcanic lithic rich, as are the Ashman, Trout Creek, andyounger Bowser Lake Group rocks in the southern Bowser Basin (Tipper and Richards, 1976; Richards,pers. comm. 1991 - unpublished sandstone thin section analyses). Richards' data suggest that chert-richsandstone is encountered earliest in the southern Bowser Basin in Kitsun Creek rocks of the lower SkeenaGroup (?Hauterivian -Albian; Tipper and Richards, 1976). Higher in the Skeena Group (?Albian),micaceous green sandstones compositionally similar to samples 30-10 and 30-12 carry a strongmetamorphic signal and abundant potassium feldspar. If the entry of metamorphic grains occurredapproximately contemporaneously throughout the Bowser Basin and elsewhere in the western CanadianCordillera (see next subheading below), then the Skeena Group green sandstones may be approximatelycontemporaneous with P3. Support for this idea may be found in the composition of the sample 13-3,the highest Devils Claw sandstone examined. As mentioned earlier, this grey-green sandstone containsboth fine grained muscovite and the first potassium feldspar seen in the Devils Claw or older sandstones.Ages of the studied strata based on marine invertebrates and palynology appear consistent withthis interpretation. The Currier and older Bowser Lake Group strata are dated as Late Jurassic andperhaps in part earliest Cretaceous, largely contemporaneous with upper Ashman Formation and younger95Bowser Lake Group facies dated as Late Jurassic in the southern Bowser Basin (Tipper and Richards,1976). The lower McEvoy (P2) are dated by palynology as late Barremian to Aptian age (Moffat et al.,1988; Rouse, pers. comm., in Cookenboo and Bustin, 1989). The middle McEvoy through the middleDevils Claw interval has been dated as middle to late Albian by palynology, and the upper Devils Clawas latest Albian to Cenomanian (Moffat et al., 1988; and Rouse, pers. comm , 1989). Micaceous greensandstones of the Skeena Group exposed on the Skeena arch near the southern margin of the BowserBasin and farther south in the Chilcotin-Nechako region were also deposited in the mid-Cretaceous (midto late Albian for the Chilcotin-Nechako region- Hunt, 1992; and Albian, or possibly as old as Aptian toBarremian for the Skeena Arch region- Palsgrove and Bustin, 1991). Both P3 and Skeena greensandstones show the first significant metamorphic provenance terrane signals in their respective parts ofthe basin.Implications for Cordilleran tectonicsLithic-rich sandstones similar to those of the northern Bowser Basin characterize Mesozoic sandstonesfrom Cordilleran basins. Out of Dickinson and Suczek's (1979) list of 88 separate sandstonecompositions from around the world, only 5 are chert and 2 volcanic lithic rich (>30%) , but 4 of 5chert and both volcanic lithic rich sandstones are from Mesozoic Cordilleran basins (in part, of course, aresult of sample bias). More recent studies and data sets not included in Dickinson and Suczek (1979)strengthen the conclusion that the lithic rich sandstones are characteristic of Mesozoic Cordilleran basins.The widespread extent and commonalty of character between these chert and volcanic lithic richsandstones has implications for Cordilleran tectonics.Even an incomplete listing of occurrences of chert rich sandstones demonstrates that chert ispervasive in northern Cordilleran sandstones during the Cretaceous. Chert-rich sandstones were reportedin Dickinson and Suczek (1979) from the mid-Cretaceous Virginia Ridge Formation in the Methow Basin(Cole, 1973; Tennyson and Cole, 1978), Lower Cretaceous Blairmore Group (Mellon, 1967) of theWestern Canada Sedimentary Basin (WCSB); and the Upper Triassic Vester Formation of Oregon96(Dickinson and Suczek, 1979). Chert and quartz rich sandstones are also reported from the Late Jurassic-Early Cretaceous Kootenay Group (Gibson, 1985) and Early Cretaceous Gates and Moosebar Formations(Carmichael, 1983; Leckie, 1983), of the WCSB. These Lower Cretaceous sandstones from the WCSBgenerally lack an abundant volcanic lithic fraction. By the early to mid-Cretaceous, chert and/or quartzdominated sandstones are ubiquitous in Canadian Cordilleran basins.Gates Formation sandstones contain grains of low to medium rank metamorphics (Leckie,1983), suggesting provenance from a metamorphic terrane during the Albian. Metamorphic andplutonic indicators are generally absent from Kootenay Group sandstones (Gibson, 1985). The first clearindications of a metamorphic provenance in the WCSB closely parallels the Skeena Group greensandstone (Albian, at least in part) and McEvoy Formation (?Albian) in the Bowser Basin.Sandstones dominated by volcanic lithic fragments in the northern Cordillera have been reportedfrom a number of basins. Included in these reports are Middle Blairmore Group sandstones of the LowerCretaceous WCSB (Mellon, 1967), the Methow and Tyaughton intermontane basins and possibly themelange belt of the North Cascades (Garver, 1989). This report adds Late Jurassic to earliest Cretaceoussandstones of P1 in the northern Bowser Basin. Unpublished data (Richards, pers. comm. 1991) indicateLate Jurassic sandstones in the southern Bowser Basin are also dominanted by volcanic lithic fragments.Although specific source areas for lithic sandstones have been the subject of considerablespeculation, in general most authors have argued for a source west of the WCSB and east of theIntermontane Belt, in an area centering around the Omineca geanticline. In some samples, the same rocksmay be sources for detritus shed both to the east and west. This study has suggested the HogemBatholith and associated island arc volcanics as possible sources for Bowser Basin sandstones. TheHogem batholith has also been suggested as part of the source of the Gates Formation east of theOmineca geanticline (Leckie, 1983).The Omineca Geanticline apparently cored an extensive land mass marginal to North Americathat provided a source for clastics in all directions. The land mass apparently began shedding oceanic97lithosphere and island arc detritus into the northern Bowser in the Middle Jurassic.Changes in sandstone composition provide insight into the tectonic history of this Ominecageanticline landmass. The first detritus shed was volcanic lithic rich, reflecting ocean floor and arcsources. Some of those first sediments are preserved in P1 in the Bowser Basin, but others were recycledby uplift in the Early to mid-Cretaceous. These recycled sediments accumulated both in the BowserBasin west of the source, and the WCSB east of the source.The source area tectonic regime deduced from P1 grain composition requires erosion of anundissected arc and obducted of oceanic lithosphere. The extrabasinal and paleovolcanic signal (and lackof neovolcanic indicators) among the volcanic lithic grains suggests that the arc source was ancient (atleast largely inactive) at the time of its erosion. Arc volcanics surrounding (and presumably beneath)Bowser Basin clastics further support proximity to an older arc. The volcanic lithic signal is morepronounced in sediments west of the Omineca geanticline than in the WCSB, suggesting that the arc wasout-board of the obducted sea-floor terrane, and that the obducted ocean floor may have actually beenback-arc seafloor. Uplift and erosion of such ancient island arc and associated sea-floor probablyoccurred by obduction of marginal basin and fringing island arc suites onto the western margin of NorthAmerica prior to the Middle Jurassic. A model involving Early Jurassic closure of marginal seas (not aunique solution) is illustrated in figure 3-11. Obduction of oceanic and island-arc strata of the SlideMountain assemblage, compatible with this model, occurred northeast of the Bowser Basin by the lateEarly Jurassic (Hansen, 1992). A compatible history of Early to Middle Jurassic obduction for SlideMountain and Quesnellia rocks on to North America (Murphy, 1989), and Middle to Late Jurassicobduction of Cache Creek terrane on to Quesnellia (Mortimer et al., 1989) also has been described in thesouthern Canadian Cordillera.98PAEOkid • • ,-•601.14 4 ,W040,!, :LATE PALEOZOIC^EARLY/MIDDLE JURASSIC -01-- SELECTED TERRANES TODAYFigure 3-11: Model of Cordlileran tectonic development consistent with interpretation of provenance.a) Late Paleozoic reconstruction ofwestern North American margin,based on Bowser Basin provenanceconsiderations. Note the marginalsea is fringed by active (ot island arcs.b) By the Early to Middle Jurassic, fringingisland arcs and associated marginalseas have been closed by abduction ontothe western margin of North America.Note that the island arc is no longer active.c) Terranes as they appear today in theCanadian Cordillera. Obducted terranesare discontinuous erosional remnantsof their Mesozoic extents.u:1Added to the source area during the Early to mid- Cretaceous are metamorphic and plutonicindicators, as well as indications of active acidic volcanism and minor plutonics, reflecting the first actualunroofing of rocks belonging to the Omineca Crystalline Belt and dissection of the P1 island arc. Theunroofing is first recorded in the WCSB by the Gates Formation (middle Albian), for which the Hogembatholith plutonics are a suggested source (Leckie, 1983). In the Bowser Basin, first unroofing isrecorded in the middle McEvoy (Early Cretaceous - possibly middle Albian) and Skeena Group(?Hauterivian to Albian). As in the Gates Formation, unroofing of the Hogem batholith or relatedplutons is a possible contributing source for P2.Implications to rifting of Pangea and the opening of the AtlanticThe widespread occurrence of chert and volcanic lithic rich sandstones in Mesozoic Cordilleranbasins, although rare elsewhere in the world, suggests a commonalty of origin in response to unusualconditions. It is proposed that these lithic sands were sourced from an extensivearea of ophiolite andisland arcs obducted shortly after North America began rifting from Pangea in the Early to MiddleJurassic (Hay et al., 1981). A latest Triassic to earliest Jurassic change in North American plate motion(Ekstrand and Butler, 1989) preceded the initial rifting of Pangea and opening of the Atlantic Ocean. Onthe western margin of North America, the initial rifting of the Atlantic appears to be recorded by theobduction of the expansive area of marginal sea floor and island arcs that formed the provenance forBowser Basin sands. The occurrence of similar Jura-Cretaceous lithic suites throughout the Cordillerasuggests that closure of marginal seas and obduction of fringing island arcs was a widespread processassociated with initial rifting of the Atlantic. Remains of these marginal seas and island arc arepreserved in the elongate outcrop occurrences of Cache Creek, Slide Mountain and Bridge River rocksstretching from Washington through British Columbia and into the Yukon.100CONCLUSIONS1) Sandstones of the northern Bowser Basin exhibit changing grain composition in successivestratigraphic units.2) Currier Formation and older Bowser Lake Group sandstones belong to a volcanic lithic rich suitetermed petrofacies 1 (P1) in this report. P I is interpreted as the product of obducted marginal sea-floorand island arc provenance terranes.3) Sandstones in the lower McEvoy Formation are chert rich, with lesser volcanic lithic fragments.These sandstones are assigned to petrofacies 2 (P2), which is interpreted as the product of first cycle andrecycled sea-floor and island arc provenance.4) Upper McEvoy and Devils Claw Formation sandstones are chert rich with lesser volcanic lithicfragments, and also show for the first time in the northern Bowser Basin a metamorphic (phyllite/schist)and minor plutonic provenance signal. These sandstones are termed petrofacies 3 (P3) in this report, andare interpreted as the product of first cycle and recycled sea-floor and island arc provenance, withadditional exposed metamorphic and plutonic facies.5) Chert and volcanic lithic rich sandstones similar to the northern Bowser Basin petrofacies characterizeMesozoic basins located adjacent to and within the Canadian cordillera. The composition and widespreadnature of these sandstones suggests uplift and erosion of a large area of island arc and marginal sea floorpresumably obducted to the western margin of North America as a consequence of initiation of rifting ofthe Atlantic ocean in the Early to Middle Jurassic.101REFERENCESBasu, A., 1985. Influence of climate and relief on compositions of sands released at source areas. InZuffa, G. G. (ed.), Provenance of Arenites. D. Reidel Publishing Company, p. 1-18.Basu, A., Young, S. W., Suttner, L. J., James, W. C. and Mack, G. H., 1975. Re-evaluation of the useof undulatory extinction and polycrystallinity in detrital quartz for provenance interpretation.Journal of Sedimentary Petrology, v. 45, p. 873-882.Blatt, H. 1982. Sedimentary Petrology. W. H. Freeman and Company. 564 p.Blatt, H., Middleton, G. and Murray, R., 1980. Origin of Sedimentary Rocks, 2nd ed.,: Prentice-Hall,Inc., Englewood Cliffs, New Jersey, 782 p.Carmichael. S. M. M., 1983. Sedimentology of the Lower Cretaceous Gates and Moosebar Formations,northeast coalfields, British Columbia. Ph.D. thesis, University of British Columbia,Vancouver, British Columbia. 285 p.Cheeney, R. F., 1983. Statistical methods in geology. George Allen Unwin, Boston, 169 p.Cole, M. R., 1973. Petrology and dispersal patterns of Jurassic and Cretaceous sedimentary rocks in theMethow River area, north Cascades, Washington. PhD thesis University of Washington, 110 p.Coney, P. J., Jones, D. L., and Monger, J. W. H., 1980. Cordilleran suspect terranes. Nature, v. 288,p. 329-333.Cookenboo, H. 0., 1989. Lithostratigraphy, palynostratigraphy, and sedimentology of the northernSkeena Mountains and their implications to the tectonic history of the Canadian Cordillera.MSc. thesis, University of British Columbia, Vancouver British Columbia. 131 p.Cookenboo, H. 0., and Bustin, R. M., 1989. Jura-Cretaceous (Oxfordian to Cenomanian) stratigraphyof the north-central Bowser Basin, northern British Columbia: Canadian Journal of EarthSciences, v. 26, p. 1001-1012.Cordey, F., Mortimer, N., Dewever, P., and Monger, J. W. H., 1987. Significance of Jurassicradiolarians from the Cache Creek terrane, British Columbia. Geology, v. 15, p. 1151-1154.Creber, G. T. and Chalconer, W., 1985. Tree growth in the Mesozoic and Early Tertiary and thereconstruction of Palaeoclimates. Palaeogeography, Palaeoclimatology, Palaeoecology, v. 52, p.35-60.102Dickinson, W. R., 1985. Interpreting provenance relations from detrital modes of sandstones. In Zuffa,G. G. ed., Provenance of Arenites: D. Reidle, Dordrecht, p. 333-361.Dickinson, W. R. and Suczek, C. A., 1979. Plate tectonics and sandstone compositions. AmericanAssociation of Petroleum Geologists Bulletin, v. 63, p. 2164-2182.Dutta, P. K. and Suttner, L. J., 1986. Alluvial sandstone composition and paleoclimate, II. Frameworkmineralogy. Journal of Sedimentary Petrology, v. 56, p. 346-358.Eisbacher, G. H., 1974. Sedimentary and tectonic evolution of the Sustut and Sifton Basins, north-central British Columbia. Geological Survey of Canada Paper 73-31, 57 p.Eisbacher, G., 1981. Late Mesozoic - Paleogene Bowser Basin molasse and Cordilleran tectonics,western Canada, In Miall, A. D. (ed.), Sedimentation and Tectonics. Geological Association ofCanada, Special Paper 23, p. 125-151.Ekstrand, E. J. and Butler, R. F., 1989. Paleomagnetism of the Moenave Formation: Implications forthe Mesozoic North American apparent polar wander path. Geology, v. 17, p. 245-248.Evenchick, C. A., 1991. Geometry, evolution and tectonic framework of the Skeena fold belt, northcentral British Columbia. Tectonics, v. 10, p. 527-546.Garnett, J. A., 1978. Geology and mineral occurrences of the southern Hogem Batholith, BritishColumbia Department of Mines and Petroleum Resources Bulletin, No. 70, 75 p.Garver, J. I., 1989. Tectonic significance of polymodal compositions in melange sandstones, westernmelange belt, North Cascade Range, Washington - Discussion. Journal of SedimentaryPetrology, v. 58 p. 1046-1050.Gibson, D. W., 1985. Stratigraphy, sedimentology and depositional environments of the coal-bearingKootenay Group, Alberta and British Columbia. Geological Survey of Canada Bulletin 37, 108p.Girty, G. H., 1991. A note on the composition of plutoniclastic sand produced in different climatic belts.Journal of Sedimentary Petrology, v. 61, p. 428-433.Girty, G. H., Mossman, B. J. and Pincus, S. D., 1988. Petrology of Holocene sand, peninsular ranges,California and Baja Norte, Mexico: Implications for provenance-discrimination models. Journalof Sedimentary Petrology, v. 58, p. 881-887.103Grantham, J H. and Velbel, M. A., 1988. The influence of climate and topography on rock-fragmentabundance in modern fluvial sands of the southern Blue Ridge Mountains, North Carolina.Journal of Sedimentary Petrology, v. 58, p. 219-227.Green, G. M., 1991. Detailed sedimentology of the Bowser Lake Group, northern Bowser Basin, BritishColumbia. Current Research, Part A, Geological Survey of Canada Paper 91-1A, p. 187-195.Gulf Canada Resources Limited 1984. Mount Klappan property. British Columbia Ministry of Energy,Mines and Petroleum Resources, Open File Report 111.Gulf Canada Resources Limited 1987. Lost Fox property. British Columbia Ministry of Energy, Minesand Petroleum Resources, Open File Report 723.Hansen, V. L., 1992. P-T evolution of the Teslin suture zone and Cassiar tectonites, Yukon, Canada:evidence for A- and B-type subduction. Journal of Metamorphic Geology, v. 10, p. 239-263.Hawkins, J W. Jr., 1980. Ophiolites. Proceedings International Ophiolite Symposium, Cyprus, 1979. p.244-254.Hay, W. W., Barron, E. J. , Sloan, J. L. and Southam, J. R., 1981. Continental drift and the globalpattern of sedimentation. Geologische Rundschau, v. 70, p. 302-313.Hekinian, R. 1982. Petrology of the ocean floor. Elsevier Oceanography Series. v. 33. ElsevierScientific Publishing, Amsterdam. 407 p.Hunt, J. A., 1992. Stratigraphy, maturation and source rock potential of Cretaceous strata in theChilcotin-Nechako region of British Columbia. MSc. thesis, The University of BritishColumbia. 448 p.Jones, D. L. and Murchey, B., 1986. Geologic significance of Paleozoic and Mesozoic radiolarian chert.Annual review of Earth and Planetary Sciences. v. 14, p. 455-492.Leckie, D. A., 1983. Sedimentology of the Moosebar and Gates Formations (Lower Cretaceous). Ph.D.Thesis, McMaster University, 515 p.Lundberg, N., 1991. Detrital record of the early Central American magmatic arc: Petrography ofintraoceanic forearc sandstones, Nicoya Peninsula, Costa Rica. Geological Society of AmericaBulletin, v. 103, p. 905-915.MacLeod, S. E., and Hills, L. V., 1990. Conformable Late Jurassic (Oxfordian) to Early Cretaceousstrata, northern Bowser Basin, British Columbia: A sedimentological and paleontological model.Canadian Journal of Earth Sciences, v. 27: 988-998.104Malloch, G. S., 1914. Groundhog coal field Summary report of the Geological Survey Department ofMines. p. 69 to 101.Mellon, G. B., 1967. Stratigrphy and petrology of the Lower Cretaceous Blairmore and mannvilleGroups, Alberta foothills and plains. Alberta Research CouncilBulletin, v. 21, 270 p.Moffat, I. W., 1985. The nature and timing of deformational events and organic and inorganicmetamorphism in the northern Groundhog Coalfield: implications for the tectonic history of theBowser Basin. Ph.D thesis, University of British Columbia, Vancouver, B.C.Moffat, I. W., Bustin, R. M., and Rouse, G. E., 1988. Biochronology of selected Bowser Basin strata:tectonic significance. Canadian Journal of Earth Sciences, v. 25, p. 1571-1578.Mortimer, N., van der Heyden, P., Armstrong, R. L., and Harakal, J., 1989. U-Pb and K-Ar datesrelated to the timing of magmatism and deformation in the Cache Creek terrane and Quesnellia,southern British Columbia. Canadian Journal of Earth Sciences, v. 27, p. 117-123.Murphy, D., C., 1989. Crustal paleo-rheology of the southwestern Canadian Cordillera and its influenceon the kinematics of Jurassic convergence. Journal of Geophysical Research, v. 94, p. 15723-15739.Nelson, H. and Creager, J. S., 1977. Displacement of Yukon-derived sediment from Bering Sea toChukchi Sea during Holocene time. Geology, v. 5, p. 141-146.Odin, G. S. and Fullagar, P. D., 1988. Geological significance of the glaucony facies. In Odin, G. S.(ed.) Green Marine Clays: Developments in Sedimentology 45, Elsevier, New York. p. 295-392.Palsgrove, R. J. and Bustin, R. M., 1991. Stratigraphy, sedimentology and coal quality of the LowerSkeena Group, Telkwa Coalfield, Central British Columbia, NTS 93L/11. British ColumbiaMinistry of Energy, Mines and Petroleum Resources, Paper 1991-2, 60 p.Pollock, S. G., 1987. Chert formation in a volcanic arc. Journal of Sedimentary Petrology, v. 57, p.75-87.Potter, P. E., 1978. Petrology and chemistry of modern big river sands. Journal of Geology, v. 86, p.423-449.Rick, J. W., 1978. Heat-altered cherts of the lower Illinois Valley; an experimental study in prehistorictechnology. Northwestern University Archeological Program Prehistoric Records, v. 2 p.105Ricketts, B. A. and Evenchick, C. A., 1991. Analysis of the Middle to upper Jurassic Bowser Basin,northern British Columbia. Current Research, Part A, Geological Survey of Canada Paper 91-1A, p. 65-73.Samson, S. D., McClelland, W. C., Patchett, P. J., Gehrels, G. E. and Anderson, R. G., 1989.Evidence from neodymium isotopes for contributions to Phanerozoic crustal genesis in theCanadian Cordillera. Nature, v. 337, p. 705-709.Scholle, P., 1982. A color illustrated guide to constituents, textures, cements, and porosities ofsandstone and associated rocks. American Association of Petroleum Geologists Memoir 28. 201p.Sedlock, R. L. and Isozaki, Y., 1990. Lithology and biostratigraphy of Fransiscan-like chert andassociated rocks in west-central Baja California, Mexico. Geological Society of AmericaBulletin. v. 102, p. 852-864..Shelley, D., 1975. Manual of optical mineralogy. Elsevier, New York, p. 239.Stattegger, K., 1987. Heavy minerals and provenance of sands: modeling of lithological end membersfrom river sands of northern Austria and from sandstones of the Austroalpine Gosau Formation.Journal of Sedimentary Petrology, v. 57. p. 301-310..Sugitani, K., Sano, H., Adachi, M., and Sugisaki, R., 1991. Permian hydrothermal deposits in the MinoTerrane, central Japan: implications for hydrothermal plumes in an ancient ocean basin.Sedimentary Geology, v. 71, p. 59-71.Tennyson and Cole, 1978. Tectonic significance of upper Mesozoic Methow-Pasayten sequence, northeastem Cascade Range, Washington and British Columbia. In Mesozoic paleogeography of thewestern United States. Society of Economic Paleontologists and Mineralogists Pacific SectionPacific Coast Paleogeography Symposium 2, p. 499-508.Tipper, H. W. and Richards, T. A., 1976. Jurassic stratigraphy and history of north-central BritishColumbia. Geological Survey of Canada, Bulletin 270, 73 p.Varley, C. J., 1984. Sedimentology and hydrocarbon distribution of the Lower Cretaceous CadominFormation. In Koster, E. H., Steel, R. J., (eds. ,) Sedimentology of gravels and conglomerates.Memoir 10, Canadian Society of Petroleum Geologists, p. 175-187.Wheeler, J. 0., and McFeely, P., 1987, Tectonic assemblage map of the Canadian Cordillera andadjacent parts of the United States of America. Geological Survey of Canada, Open File 1565.Zuffa, G. G., 1985. Optical analyses of arenites: influence of methodology on compositional results. InZuffa, G. G. (ed.), Provenance of Arenites. D. Reidel Publishing Company, p. 165-189.106CHAPTER 4DETRITAL CHROMIAN SPINEL COMPOSITIONS USED TO RECONSTRUCT THETECTONIC SETTING OF PROVENANCE: IMPLICATIONS FOR OROGENY IN THECANADIAN CORDILLERAABSTRACTDetrital chromian spinel is an important accessory mineral in northern Bowser Basin sandstones.Microprobe analyses show that Bowser Basin detrital chromian spinels have Cr/(Cr +Mg) between 0.21and 0.89; Mg/(Mg+Fe2 +) between 0.24 and 0.70; and Fe3 +/(Fe3 + + Al + Cr) values below 0.12.Comparison with spinels from the literature on ultramafic rock types indicates that the detrital spinelcompositional range closely matches spinels from alpine-type peridotites emplaced by obduction ofmarginal basin crust and island arc complexes. Furthermore, Bowser Basin spinels are compositionallydistinct from spinels derived from mid-ocean ridges, stratiform ultramafic complexes or Alaskan-typeperidotites.Based on a study of detrital spinels, it is concluded that the Bowser Basin provenance includedextensive alpine-type peridotite that originated as marginal basin lithosphere, and that there is noevidence for mid-ocean ridge derived material in the provenance. This provenance interpretation agreeswith earlier interpretations which related chert pebbles in the Bowser Basin to chert-rich strata of theCache Creek terrane and is consistent with the interpretation, based on detrital modes analysis, whichcalls for an obducted oceanic crust and island arc provenance. The chromian spinel chemistry adds,however, the important tectonic details that mid-ocean ridge derived lithosphere, Alaskan-typeperidotites, and stratiform complexes were not exposed in the provenance, and that erosion of the sourcearea occurred prior to any regional greenschist or higher grade metamorphic events. Thus microprobecompositional analysis of detrital spinels adds important detail to the provenance interpretation that is not107available through detrital modes analysis alone. The conclusion that marginal basin lithosphere ratherthan mid-ocean ridge derived material was the sediment source has important implications to the tectonichistory of the Canadian Cordillera. It suggests that the obducted provenance terranes were closelyassociated with North America. The success of microprobe analysis of detrital chromian spinet from thenorthern Bowser Basin suggests that similar methods could be profitably used in other basins derivedfrom complex accreted terranes.INTRODUCTIONSandstone provenance studies utilize a broad array of techniques that conceptually fall into twobasic approaches. The most widely used approach involves characterization of the bulk composition ofthe sandstone, typically by point-count determination of modes for detrital framework grains. Data fromsuch studies are commonly plotted on ternary diagrams following the methods of Dickinson and Suczek(1979) or others, and the ternary plots are used to interpret the tectonic setting of the provenance (e.g.,recycled orogen versus magmatic arc versus continental block). The success of bulk characterizationmethods, and especially detrital modes analysis, in relating sands and sandstones to particular provenancetectonic settings is demonstrated by their widespread application and stems directly from their use of themajor components of sandstone as the basis of the provenance interpretation. However, when usedalone, detrital modal analysis yields sandstone provenance interpretations that are potentially ambiguousbecause many factors other than type of source rock exposed in the provenance affect the finalcomposition of sandstones. These factors include climate, relief, transport mechanisms and post-burialintrastratal solution. Well-documented cases of both ancient and recent first cycle quartz arenites derivedfrom humid tropical areas highlight the potential influence that factors other than source rockcomposition play in determining the final composition of sandstones (Franzinelli and Potter, 1983;Suttner and Dutta, 1986). This ambiguity has been demonstrated statistically by Molinaroli et a/.(1991), who used discriminant function analysis of data from Dickinson et al. (1983) to show that thedetrital modes approach leads to a correct prediction of the tectonic setting of the provenance in only 75to 85% of the cases In order to increase confidence in provenance interpretations and simultaneously108add detail unavailable from conventional detrital modes analyses, a second approach (outlined below) tosandstone provenance analysis is employed with increasing frequency.This second approach utilizes a single (or at most a few) mineral which can be related byappearance, composition or other distinguishing feature to a well-defined petrogenesis. Methods thatexploit such an approach are extremely diverse. Examples include petrographic classification of quartzgrains (Basu et al., 1975), compositional analysis of feldspars (Trevena and Nash, 1981) or heavyminerals, and laboratory methods of radiogenic decay or fission track dating of detrital zircons (c.f.Smith and Gehrels, 1991; Ross and Parrish, 1991). However, the advantages and disadvantages of thissecond approach are also diverse. For example, the quartz grain classification method of Basu et al.(1975) distinguishes plutonic from metamorphic terranes based on relative abundances of undulatory andpolycrystalline grain types, but no quartz types uniquely indicate a single provenance. Lack of a uniquelinkage between quartz grain type and a particular provenance leads to ambiguities due to potentiallymixed provenance terranes, and this method does not address a possible volcanic origin at all.Alternatively, feldspars and heavy mineral suites are susceptible to differential degrees of weathering andintrastratal solution (Pettijohn, 1975) which can mask or even obliterate the provenance signal. Datingdetrital zircons can successfully determine the age of the provenance, but may not reveal much about thetectonic setting of the provenance simply because zircons are common to many petrologic associations.Despite their limitations, each of these methods can add important detail and confidence to provenanceinterpretations based on bulk composition techniques.This study uses electron microprobe compositional analysis of detrital chromian spinel todetermine tectonic setting of provenance. Compositional analysis of chromian spinel is routinely appliedin petrologic studies of spinel-bearing mafic and ultramafic rocks (e.g. Irvine, 1973; Dick and Bullen,1984; Hawkins and Melchior, 1985; Nixon, et al., 1990), because chromian spinel composition is asensitive indicator of parental melt conditions. Chromian spinel also is unusually stable chemicallycompared to other ultramafic minerals, and may be the only mineral unaltered by sub-greenschist faciesserpentinization common to sea floor environments (Hekinian, 1982). Detrital chromian spinel is109important to provenance studies for the same reasons that led to the widespread utilization of spinels inpetrologic studies of mafic and ultramafic rocks. Chromian spinel's unusual chemical stability ensurespreservation of the compositional signature after burial in sedimentary strata, in contrast to most othermafic minerals which alter rapidly at near surface conditions. For provenance studies, chromian spineloffers the further advantages of mechanical stability (lack of cleavage and high degree of hardness;Bowie, 1967), and ease of recognition in both transmitted and reflected light petrography (medium todark red-brown, yellow-brown or green-brown color, high relief, low reflectivity, and isotropic nature).Despite chromian spinels wide use in petrologic studies, its use in provenance studies has beenlimited. Previous use of chromian spinel chemistry to aid provenance interpretations include a study byUtter (1978), who used the morphology and composition of detrital chromite to relate Proterozoic placersands in South Africa to an Archean greenstone origin. In Japan, detrital chromian spinel chemistry hasbeen used to relate serpentine sandstones to adjacent peridotites (Arai and Okada, 1991). Otherapplications of detrital chromian spinel studies include heavy mineral concentrate exploration for depositsof diamonds and other economic minerals (Pasteris, 1984; Mitchell, 1991).The provenance of sandstones examined in this study is of central importance, both in a spatialand a temporal sense, to understanding the history of accretion in the Canadian Cordillera. Theexamined sandstones are from strata exposed in the northern Bowser Basin, located in the Intermontanebelt of the Cordillera (Fig. 4-1), and were deposited during the Late Jurassic and Early Cretaceous, whenmuch of the accretion of the Cordillera occurred. Earliest provenance interpretations in the Bowser Basinassociated the abundance of chert pebbles with chert-rich rocks of the Cache Creek Group (Malloch,1914). This interpretation is reinforced by identification of sandstones as having chert- and volcanic-lithic rich compositions, and paleocurrent measurements suggesting a northerly and northeasterlysediment source (Eisbacher, 1974; Cookenboo, 1989).110Figure 4-1: Location of the study area in the Intermontane Belt of the Omar'inn Cordillera. Terraneboundaries and Alaskan type peridotites after Wheeler and McFeeley (1987).1 11Framework grain detrital modes established that the provenance tectonic setting evolved from primarilyan island arc and oceanic crust terrane to a combined terrane of island arc, oceanic crust, recycledsediments, metapelites, and minor plutonic rocks (Chapter 3). Stable isotope analysis of Bowser Basinshales suggests a similar evolution of provenance from nearly juvenile oceanic crust to increasinglycontinental values during the Early through Late Jurassic interval (Samson et al., 1989). Chromianspinel compositions have the potential to provide a more detailed understanding of provenance tectonicsetting, especially as relates to the type of oceanic lithosphere, and therefore can potentially bothstrengthen provenance interpretations based on detrital modal analyses, and add detail to the accretionaryhistory of the Cordillera.METHODSRoutine thin section examination and point-count analyses revealed that chromian spinel occursas an accessory mineral (<1%) in all stratigraphic units. Samples containing relatively higherconcentrations of chromian spinel were selected during thin section analysis for microprobe study. Sevensandstones were chosen as representative of the four stratigraphic horizons exposed in the study area (fig4-2), and a total of 124 grains were analyzed from these seven chosen sandstones.Major and minor element compositional data for chromian spinels were obtained from polishedthin sections using a fully automated, four spectrometer Cameca SX-50 scanning electron microprobeemploying the PAP matrix correction routine (Pouchou and Pichior, 1984). Analyses were made with anaccelerating voltage of 15 kV, a beam current of 30 nA, and a spot size of 2 pm; TAP, LiF, and PETwere used as the analyzing crystals. Peak counting times were 30 seconds for Si, Al, Ti, Cr, Fe, Mn,Mg, and Ca and 15 seconds for Zn, V, and Ni; background times were half of peak times. Analysis ofstandards as unknowns was done at the beginning, middle and end of analytical runs to ensure propercalibration throughout. All Fe is expressed as Fe203, and atomic concentrations of ferric and ferrousiron were calculated assuming ideal spinel stoichiometry. The assumption of ideal stoichiometry isroutinely made in petrographic studies of spinel-bearing rocks (c.f. Dick and Bullen, 1984; Hawkins and112MPLED HORIZ1DEVILS CLAWFORMATIONMcEVOYFORMATIONCURRIERFORMATIONUNDIFFERENTIATEDBOWSER LAKEGROUPFigure 4-2: Stratigraphic column of the northern Bowser Basin. Horizons sampled are marked with anasterisk.113Melchior, 1985; Duncan and Green, 1987; Nixon et al., 1990), and has been shown to be reasonable fora wide range of natural and synthetic chromian spinel compositions when checked by wet chemistry andMossbauer spectra investigations (Osborne et al., 1981). Atomic ratios are used herein, unless weightper cent oxide is specified. The notation Cr# is used for atomic ratios of Cr/Cr + Al and Mg# is used foratomic ratios of Mg/Mg +Fe 2 + following the common practise in petrologic literature (c.f. Dick andBullen, 1984).Most grains were analyzed twice, once in the core and once near the rim. The only exceptionswere grains too small to permit separate analysis of core and rim. Core to rim variations were measuredboth to identify grains that may have partially re-equilibrated during mineral growth or in response tometamorphic events, and to check whether significant compositional gradients may have been establishedby low temperature alteration, weathering or post-depositional diagenetic processes.PETROGRAPHIC DESCRIPTIONSChromian spine! grains examined in this study generally are medium to dark red-brown (and in afew cases nearly opaque) in transmitted light. Less commonly, yellowish-brown grains occur, and onedark greenish-brown grain was observed. The range of spinel colors suggests a relatively broad range ofCr and Al substitution (Bernier, 1990). Chromian spinels are most abundant in fine to very fine grainedsandstones, and in some samples are concentrated in thin 'amino,- which are also enriched in other heavymineral grains such as zircon. Typical grains are between 75 to 150 pm across. Grain marginscommonly exhibit conchoidal fractures, suggesting mechanical breakage, but some grains are subhedral,suggesting preservation of the original crystal boundaries. Grains are generally visually homogeneousand show no obvious signs of zoning. Several serpentine grains with red-brown spinet inclusions werenoted. A few spine! grains contain inclusions of olivine (0F92-94).114RESULTSDetrital spinels measured in this study exhibit a wide variation in major element concentrations(representative analyses presented in table 4-1). Cr# ranges between 0.21 and 0.89, with most Cr# above0.35. Only four out of 124 grains measured had Cr# less than 0.30. Mg concentration generallydecreases, and Fe increases, with increasing Cr#. Mg# ranges between 0.24 and 0.70. Fe 3 +concentration is consistently low. The maximum Fe 3 +/(Fe3 + +Cr+Al) in the measured spinels is 0.12.Core to rim variations are minor, and no consistent trends were noted.The elements Ti, Mn, V, Zn, and Ni generally occur in only trace amounts (less than 1 0 wt %oxide). Ti is the most abundant of these minor elements, but even TiO2 concentrations generally arelow, and more than 90% of measured grains have TiO2 concentrations less than 0 25%. All of the moreTiO2 enriched grains occur in the youngest stratigraphic units sampled (upper McEvoy and lower DevilClaw formations). A single exceptionally Ti02-rich grain (4.2 wt % oxide), which also contains veryhigh total Fe (58.6 wt % oxide), also occurs in the upper McEvoy Formation. This grain plots outsidethe compositional limits on all the discriminant diagrams discussed in the following section (figures 4-3,4-4, and 4-5).In my detrital grains, V and Zn concentrations consistently are below 0.50 wt. % oxide, and Niconcentrations are less than 0.25 wt. % oxide. One exceptional grain has Zn concentration of 0.57 wt.% oxide in its rim (V in this grain's rim is 0.35 wt. % oxide). This grain, like the high TiO2 graindiscussed above, is from the upper McEvoy Formation.Type of ultramafic sourceMajor element concentrations of chromian spinels from a variety of types of ultramaficcomplexes around the globe have been arranged into fields on several discriminant diagrams, each ofwhich represent one face of the spinel compositional prism (Irvine, 1974; Dickey, 1975; Basu et al.,1975; Wilson, 1982; and Dick and Bullen, 1984). By comparison with these fields, the type ofultramafic source of the detrital spinels can be determined. However, it must be noted that occurrencewithin a field on any one diagram alone does not necessarily imply a match of the detrital spinels with aparticular origin. Because variation occurs between three trivalent and two divalent cations, there is115TABLE 4-1: Representative analyses of detrital spinels from the northern Bowser Basin. Ideal spinelstochiometry was assumed in calculation of Fe 2 + and Fe3 + atomic ratios. Cr# equals atomic ratio ofCr/(Cr +Al) and Mg# equals atomic ratio of Mg/(Mg +Fe2+)REPRESENTATIVE ANALYSES1 2 3 4 5 6SiO2 0.01 0.85 0.05 0.02 0.03 0.07Ai203 33.79 34.66 28.38 23.02 22.46 9.621102 0.06 0.04 0.08 0.11 0.12 0.30Cr2O3 30.87 29.74 40.11 43.04 42.78 53.45Fe2O3 19.88 19.13 16.40 22.30 21.48 28.75MnO 0.08 0.12 0.09 0.11 0.09 0.16MgO 14.60 14.63 13.81 11.72 11.39 8.02CaO 0.00 0.05 0.00 0.03 0.10 0.02ZnO 0.18 0.26 0.19 0.28 0.31 0.11V203 0.12 0.10 0.27 0.24 0.28 0.12NIO 0.19 0.19 0.07 0.02 0.03 0.05TOTAL 99.78 99.77 99.45 100.89 99.07 100.67Atomic Ratios (stoictUometric calculations based on 4 0)Al 1.19 1.22 1.02 0.85 0.85 0.39Ti 0.00 0.00 0.00 0.00 0.00 0.01G 0.72 0.70 0.97 1.07 1.08 1.45Fe2 + 0.35 0.35 0.37 0.45 0.46 0.59Fe3+ 0.09 0.08 0.01 0.08 0.06 0.15Mg 0.65 0.65 0.63 0.55 0.55 0.41Mn 0.00 0.00 0.00 0.00 0.00 0.01Cr# 0.65 0.36 0.49 0.56 0.56 0.79Mg# 0.38 0.65 0.63 0.54 0.54 0.41116overlap among fields on some plots, although the fields may be largely distinct on other plots.A ternary plot of Cr, Al and Fe3 + (Fig. 4-3) is useful for distinguishing alpine-type peridotitesfrom Alaskan-type and stratiform peridotite complexes, as demonstrated by Dick and Bullen (1984). Inalpine-type peridotites, Cr increases with increasing Fe3 +, but Fe3 + concentrations overall remain quitelow. Both stratiform and Alaskan-type complexes exhibit, in part, much higher concentrations of Fe 3 +than the alpine-type peridotites, and greater scatter of Fe3 + concentrations relative to Cr concentrations.For my spinels, chromium concentration increases with increasing Fe 3 + concentrations, and overall Fe3 +concentrations remain low (Fig. 4-3), consistent with an Alpine-type peridotite origin (Dick and Bullen,1984). High Fe3 + spinels, which distinguish many stratiform complexes are entirely lacking from mydata set (Irvine, 1974; Wilson, 1982; Dickey, 1975; and Dick and Bullen, 1984).The importance of Fe-rich spinels in ultramafic bodies formed by fractional crystallization in thecrust (Alaskan-type peridotites and stratiform complexes) is emphasized in figure 4-4, a plot ofFe3 +/(Fe3 + +Cr + Al) versus Mg#. Except for a single grain which is considered separately later, thedata from this study also plots entirely within the field of alpine-type peridotites (Irvine, 1974) in thisdiagram. The field for Alaskan-type complexes plots entirely outside the range of sampled grains, andtherefore can be excluded as a significant contributor to the detritus. The stratiform field from Irvine(1974) partially overlaps the alpine-type peridotite field (Fig. 4-5), but none of the detrital grains exhibitthe high Fe3 + compositions that distinguish at least some stratiform complex chromian spinels (Irvine,1974; Haggerty, 1979; Wilson, 1982; and Dick and Bullen, 1984) from alpine-type peridotites[exclusively low Fe3 +: Fe3 +/(Fe3 + + Al + Cr) less than approximately 0.12]. Although contributionof minor portions of low Fe 3 + stratiform spinels is cannot be completely ruled out, such an explanationrequires ad-hoc assumptions of only low Fe3 + stratiform spinels being contributed, and thereforesignificant contribution from stratiform source is unlikely. Taken together, the Mg# versusFe3 +/(Fe3 + +Al+ Cr) and Cr - Al - Fe3 + ternary plots demonstrate that detrital chromian spinels fromthe northern Bowser Basin have close affinities to alpine-type peridotites and exclude Alaskan-typeperidotites and stratiform complexes as major sediment sources.117Fd+TulameenAlaskan-TypeComplexStratiform \^Abyssalcomplexes •, ultramafic xenolit\ and Alpine-type•eridotrt spinelsCr +Figure 4-3: Ternary plot of the major trivalent cations in chromian spinels. Solid circles are values fromthe detrital spinels in this study. For comparison, dark-gray fill covers nearly 300 spinel analyses froman Alaskan-type peridotite (Tulameen complex), after Nixon, et al. (1990). Hatching denotes field forchromian spinels of mantle melting origin, including ultramafic xenoliths, abyssal dunites, spinel andplagioclase peridotites, and alpine-type peridotites (Basu et al., 1975; and Dick and Bullet', 1984).Stippled field shows compositional range of stratiform complex spinels derived by fractionalcrystallization, including data from the Bushveld, Rhum, Stillwater, Hartley (Great 'Dyke'), and Marumstratiform complexes (Duke, 1982; Wilson, 1982; and Dick and Bullen 1984).118- 0.80- 0.60- 0.40 Alaskan type- 0.20Alp ine •• • • • •I ••,141s4re4-Ii*".` i0.80^0.60 0.40i0.203+Fe3+Fe +Cr+AlFigure 4-4: Plot of Mg/(Mg+Fe2 +) against ratio of trivalent cations Fe 3 ÷/(Fe3 + +Al+ Cr) for detritalspinel data set compared to worldwide occurrences from Irvine (1974).MgMg+Fe2+119MgMg+Fe 2+MgMg+Fe 2+Figure 4-5: Cr/(Cr Al) versus Mg/(Mg+Fe2 +). Fields include mid-ocean ridge-derived chromianspinels (abyssal peridotites, dunites and basalts; Dick and Bullen, 1984); Bowser Basin detrital spineldata set (data points marked by filled circles); and field for alpine-type peridotites. ratios)120Alpine-typeperidotite— 0.80— 0.60StratiformcomplexDetritalSpinetsV0.80^0.60^0.40^0.2011^1^1^1^1^1^1^IOrigin of the alpine-type peridotite sourceThe origin of the alpine-type peridotite source can be constrained by consideration of the Cr#versus Mg# plot (Fig. 4-5). Chromian spinels formed at mid-ocean ridges in both peridotites and basaltsare restricted to Cr# less than 0 60, and typically have high Mg#s (many 0.70 to 0.85; Dick and Bullen,1984). In contrast, spinels in back-arc basin basalts tend to have lower Mg# for a given Cr#, andassociated island-arc spinels exhibit Cr# in excess of 0.60 (Dick and Bullen, 1984; Fig. 4-6). Thedetrital spinels span a much greater range of Cr# compositions than abyssal spinets, and therefore abyssalrocks are excluded as the only source. A better match to spinel chemistry in this study comes from somevery large and complex alpine-type peridotites because they exhibit a broad range of Cr#, includingvalues both below and above the abyssal limit of 0.60. These large ophiolites, which include the Samailin Oman (Pallister and Hopson, 1981), the New Caledonian (Nicolas and Prinzhofer, 1983) andJosephine peridotites, the Bay of Islands Complex (Dick and Bullen, 1984) and the Kanuti Ophiolite inAlaska (Loney and Himmelberg, 1989), are thought to have a "complex multi-stage melting history notfound at mid-ocean ridges" (Dick and Bullen, 1984, p. 73). These ophiolites, and by inference thedetrital chromian spinels in this study, likely originated from a complex of marginal basin lithosphereand island arc-associated rocks. Such a tectonic setting and complex melting history suggests aprovenance in proximity to a continental margin.The possibility exists that the detrital spinels could be a mixture of mid-ocean ridge-derivedseafloor with more depleted marginal basin and island arc suites. Two aspects of the detrital spinetcompositions, however, suggest that a mid-ocean ridge derived abyssal source did not contribute (at leastsignificantly) to the sampled strata in the northern Bowser Basin. High alumina spinels (Cr#= 0.10 to0.30) are common in abyssal peridotites, dunites, and basalts (Dick and Bullen, 1984), but are almostentirely lacking among my detrital spinels (only four grains out of 124 measured less than 0.30, lowestCr#= 0.21). High Mg#s (0.70 to 0.85 out of a range from 0.55 to 0.85) are also common in spinelsfrom abyssal peridotites and basalts (Dick and Bullen, 1984), but like high alumina spinels also aremissing from my detrital spinels (highest Mg#= 0.70). The lack of both high Al and high Mg12140 CONTINENTAL CRUSTMarginal basinseafloorCr#= 0.30 to 0.55Mg# mostly <0.70Mid-ocean ridgeCr#= 0.10 to 0.60Mg#= 0.55 to 0.85j LITHOSPHEREATHENOSPHEREFigure 4-6: Cartoon illustrating typical spinel compositions from different sea-floor (potential alpine-typeophiolite) and continental crust origins. No scale implied.122Cr# and Mg# values for spinels from different tectonic settings Alaskan andStratiformperidotitesCr#= 0.60 to 0.90Mg#=0.10 to 0.65Island-arcCr#= 0.60 to 0.90Mg#=0.40 to 0.65UPPER MANTLEspinels, which are the best indicators of mid-ocean ridge origin, strongly suggests that northern BowserBasin spinels did not originate from a mid-ocean ridge source. The generally lower Mg# of my spinelssuggests that they crystallized at lower temperatures than normal mid-ocean ridge basalts (MORB) orperidotite (Hawkins and Melchior, 1985) because Mg concentration in chromian spinels decreases withlower temperatures of formation and increased partial melting (Hill and Roeder, 1974). One scenario forincreased partial melt is increased water content, and a subducted slab associated with marginalbasin/island arc development can provide the source for such elevated water content. Therefore, fluidsevolved from subducted lithosphere may have contributed to the chemical signature of the ophioliteprovenance as preserved in the detrital spinels.A modern analog for my detrital spinels may in part occur in rocks of the Marianas Trough.Back-arc basin rocks in the Marianas Trough contain spinels with moderate Cr#s (0.30 to 0.55) andMg#s (maximum Mg#= 0.73; average Mg#= 0.62; Hawkins and Melchior, 1985). Like my detritalspinels, the Marianas Trough spinels lack the very high Al and Mg examples common to abyssal suites.High Cr spinels (Cr# greater than 0.60) are lacking in the Marianas Trough suites, but do occur inspinels from Marianas island arc volcanics, including boninites (Cr#= 0.73 to 0.89; Bloomer andHawkins, 1987). Obduction of a complex of rock suites similar to Marianas back-arc crust and islandarc volcanics (including boninites) satisfactorily explains the compositional variation of my detritalspinels.The single exceptionally TiO2 rich spinel, mentioned earlier in the results section, deservesconsideration. Such high TiO2 chromian spinels (4.2 wt. % oxide) are uncommon, but have beenreported from late stage wehrlites of the Samail ophiolite (Pallister and Hopson, 1981). This grain istherefore consistent with the interpreted input from a complex alpine peridotite source, and may furtherreflect addition of more evolved, late stage ophiolitic material to the provenance.Spinel chemistry can record exposure to greenschist and higher grade metamorphism bydevelopment of compositional zoning, changes in major element chemistry, and addition of significant123(less than approximately 1.0 weight per cent oxide) concentrations of certain trace elements. Majorelement changes most clearly indicative of metamorphism include enrichment of Fe 3 +, commonly inopaque ferritchromite rims, which can begin during greenschist metamorphism, and development of veryhigh Cr (Cr# greater than 0 90) and very low Al (Cr# less than 0.05) compositions during greenschistand higher grades of metamorphism (Evans and Frost, 1975; Whittaker and Wadkinson, 1983). Neithercompositional zoning nor very high Cr, Al or Fe3 + content spinels occur among my detrital spinels. Thegenerally low levels of Ni, Zn, and V combined with lack of significant core to rim variations and lackof very high Cr, Al or Fe3 + spinels suggests erosion of the obducted alpine-type peridotite sourceoccurred prior to any greenschist or higher grade regional metamorphic events. Lack of any consistent orappreciable core to rim variation trend also mitigates against the possibility of alteration of the spinelsafter deposition.The same type of ultramafic source appears to have been the source of Bowser Basin sedimentsthroughout accumulation of the sampled strata. No clear changes in spinel chemistry have been identifiedfrom the base to the top of the examined strata. Several subtle changes, however, suggest that a greatervariety of source rocks became exposed through time. All spinels with TiO2 concentrations greater than0.25% were recovered from upper McEvoy Formation strata (Fig. 4-7), as well as a single grain with Znconcentration of 0.57 wt. % in its rim (V in grain's rim is 0.42 wt. %). Zn concentrations above 0.50%are commonly associated with sulfide mineralization (Bernier, 1990), suggesting that this single grainmay record some localized metasomatic activity within the provenance. The slightly more variablecompositions that occur in the McEvoy Formation are consistent with exposure of additional source rocksin basically the same source area. A similar conclusion that sediment source rocks became more variedthrough time is evident from the sandstone framework grain detrital modal study (Chapter 3).124Figure 4-7: TiO2 versus Cr203 plot, demonstrating the restriction of high TiO2 values to sandstones fromthe McEvoy Formation (petrofacies 2). Dashed line has a value of 0.25 wt. % Ti02.1251.0%TiO2wt%0.0%• McEvoy Fm(petrofacies 2)0 Older units•• • • •• •^•• —• ♦^♦^®°aEs."^P•13 •• ••; s,^ .^B. 8.• • ®° ru • aci • ° ;9t1a It : IeV. v 0• 0ED • so 10.0%^Cr 0 wt%^65.0%2 3PROVENANCE: CANDIDATES IN THE CANADIAN CORDILLERAThe earliest provenance interpretations for the northern Bowser Basin concluded that the mostlikely source was oceanic rocks of the Cache Creek terrane, based on the abundance of chert pebbles inconglomerates and paleocurrent direction indicators implying input from the north and east (Malloch,1914; Eisbacher, 1974; Cookenboo, 1989). More detailed interpretation of provenance based on detritalmodal analysis of the sandstones demonstrates that ocean floor and island arc terranes contributed to theCurrier Formation and undivided Bowser Lake Group strata, with additional contribution from dissectedarc and low to medium grade metapelites evident in the younger McEvoy and Devils Claw Formations(Chapter 3). Provenance interpreted from the detrital chromian spinel study is entirely consistent withthe conclusions reached from the detrital modal analysis, but adds details significant to tectonicreconstructions of the Cordillera. In addition to confirming that island arc and oceanic crustal rocks wereinvolved in the provenance, it is inferred from the detrital spinel compositions that the °el-Attic terranewas obducted marginal basin crust, and that the obducted material was eroded prior to any regionalmetamorphic events affecting the source area.Several possible sources of chromian spinel exist to the north and east of the Bowser Basin, thedirection of provenance indicated by paleocurrents. Alaskan-type peridotite bodies are scattered amongrocks north and east of the Bowser Basin (Fig. 4-1), but these potential sources of chromian spinel can beeliminated as significant contributors because this type of peridotite contains Fe3 +-rich spinels unlike anyfound in the detrital spine! suite. Two oceanic terranes containing peridotites are also present to thenorth and east (Fig. 4-1 and Fig. 4-8c): 1) the Cache Creek terrane (Mississippian to Lower Jurassic;Wheeler, et al., 1991) composed mainly of MORB-like tholeiitic and alkaline basalt rocks generally ofopen ocean affinities (Wheeler and McFeely, 1987); and 2) the Slide Mountain terrane (Devonian toUpper Triassic; Wheeler, et al., 1991) composed of rocks with marginal basin and island arc affinities(Wheeler and McFeely, 1987). Both terranes contain alpine-type ophiolites, but few spinet analyses arepresently available from either the Slide Mountain or Cache Creek groups. Based on its marginal basin126DETAILATE PALEOZOIC ^00-- EARLY/MIDDLE JURASSIC -100-- SELECTED TERRANES TODAYi.^v v1:^i, v v : ^\\,-...)"S..r, 1i vvsr ,11<laNE ARy CH., •-:!^v\ vV ,sloom^v vv v v ^,SKEENA ARCH \,..Figure 4-8: Model of Cordilleran tectonic development consistent with interpretation of provenance.a) Late Paleozoic reconstruction ofwestern North American margin,based on Bowser Basin provenanceconsiderations. Note the marginalsea is fringed by active (A) island arcs.b) By the Early to Middle Jurassic, fringingisland arcs and associated marginalseas have been closed by obduction ontothe western margin of North America.Note that the island arc is no longer active.c) Terranes as they appear today in theCanadian Cordillera. Obducted terranesare discontinuous erosional remnantsof their Mesozoic 00 0 tz1.0 ^0 . st08^0 SR^F.?g^8 Zg'`"N g^gR. CI)^S6^).54"(7; F)":RA -(41^gkroo°09. fti%< •51 6.6"0gff'rpCR;''z'z'qvg.01a6.E , P,ta.zg8ttg^§—^o-t; . , 'EP^qg.R2Pc1.9..R.;0'a^)›.nK1MA..4) 5)Fr g .„. 0 ,T,N14,^0 01.Q't 00 5"114G ff'N^@,-,;1 •CD 01I 00 o+,C')(1) ggr. ;071v5:0 " fe,B. 5.01P's*and island arc affinities, the Slide Mountain terrane appears to be a potential source for part of theBowser Basin sediments. Specifically, the Slide Mountain assemblage may have contributed Cr-richspinels to the detrital spinels similar to those reported from the Redfern amphibolite in southeasternBritish Columbia (Radloff, 1989). At first examination, the more abyssal affinities of the Cache CreekGroup would seem to render it incompatible with the detrital spine! chemistry. However, the CacheCreek terrane is only in part mid-ocean ridge-derived, and other parts have island arc and marginal-basinaffinities that would make them suitable sources for the detrital spinels (Wheeler et al., 1991). Chromianspinel compositions from Cache Creek Group north of the Bowser Basin near Atlin have Cr#s between0.23 to 0.89; Mg#s 0.36 to 0.68 (Ash, written communication, 1993) and are comparable to my detritalspinels. Other Cache Creek chromian spinel analyses from five unaltered cores from Mt. Sydney-Williams (Cr#s from 0.31 to 0.53 and Mg#s from 0.48 to 0.73) and Murray Ridge and Chromian Peak(43 km north of Mt. Sydney-Williams; Cr#s from 0.54 to 0.91) southeast of the Bowser Basin also fit thedetrital data set (Whittaker and Wadkinson, 1981; Whittaker, 1982; Whittaker and Wadkinson, 1983).No high Al-high Mg spinels (Cr# less than 0.20; Mg# greater than 0.75) which most clearly indicate amid-ocean ridge origin have been reported from the Cache Creek. These variable spine! compositionssupport the conclusion that at least parts of the Cache Creek Group have close affinities to marginal basinlithosphere and would be suitable provenance for Bowser Basin sediments.The detrital spinel compositions help to clarify understanding of the processes of accretionarytectonics that formed the Cordillera. From the detrital spinels, it is clear that obducted marginal seafloorand island arcs shed large quantities of detritus into the northern Bowser Basin from at least the LateJurassic (Oxfordian and younger) through the mid-Cretaceous (Aptian or Albian; MacLeod and Hills,1990; Cookenboo and Bustin, 1989). The spinel chemistry indicates that portions of marginal basinlithosphere and island arc volcanics formed from mantle with a higher degree of partial melting at lowertemperatures compared to mantle at mid-ocean ridges. Such a chemical environment is expected inmarginal basin floor formed above a subducting slab (e.g. Marianas Basin; Hawkins and Melchior,1985), probably due to its enrichment in water derived from the subducting slab, and possibly because of128repeated cycles of melting. Because northern Bowser Basin sediments lack the low Cr# and high Mg#spinels typical of abyssal seafloor, mid-ocean ridge derived spinels appear to be absent from the sourcearea. Based on the conclusion that the sediments in the Bowser Basin originated as marginal basinlithosphere, the process of accretionary tectonics in the provenance of the northern Bowser Basinprobably involved accretion of terranes that were proximal to North America. A model consistent withthese data (Fig. 8a, 8b, and 8c) calls for closure of marginal seas, and emplacement of remnants of thoseseas and associated island arcs on to the western margin of North America during or before the MiddleJurassic. Timing of this closure event may be associated with the latest Triassic to earliest Jurassicchange in plate motion recorded in paleomagnetic studies of North America (Ekstrand and Butler, 1989),which in turn is related to the initiation of Atlantic rifting.CONCLUSIONS1)Alpine-type peridotite is part of the Bowser Basin provenance.2) Chromian spinel chemistry is similar to those alpine-type peridotites derived from complex mantlemelting histories such as are believed to have been emplaced by obduction of marginal basin lithosphereand island arc complexes.3) Lack of high Al and high Mg spinels, which commonly occur in abyssal bosAlts and peridotites,implies that mid-ocean ridge derived lithosphere did not contribute detritus significantly to the examinedBowser Basin strata. Lack of mid-ocean ridge contribution, combined with clear evidence for obductedmarginal basin and island arc ultramafic suites, suggests the tectonic processes affecting Bowser Basinprovenance involved accretion and obduction of fringing marginal seas and island arcs onto the westernmargin of North America, rather than closure of true open oceans.4) Erosion of the source of Bowser Basin sediments occurred prior to any regional greenschist or highergrade metamorphic events affecting the provenance.1295) The detrital chromian spinels lack high Fe3 + examples that typically occur in Alaskan and stratiformtype peridotite complexes, suggesting that such peridotites did not contribute significantly to LateJurassic and Early Cretaceous sediments in the northern Bowser Basin. This implies that Alaskan-typeperidotites which are exposed today in the likely source area north and east of the of the study area wereprobably not exposed until after Bowser Basin sedimentation.6) The ability of detrital chromian spinels to be used to discriminate plate boundary conditions forprovenance suggests that methods used in this study may have broader application and could helpelucidate the process of accretionary tectonics if applied in other Cordilleran basins, where an ultramaficsource is suspected.130REFERENCESArai, S. and Okada, H., 1991. Petrology of serpentine sandstone as a key to tectonic development ofserpentine belts. Tectonophysics, v. 195, p. 65-81.Basu, A. R. and MacGregor, I. D., 1975. Chromite spinels from ultramafic xenoliths. Geochimica etCosmochimica Acta, v. 39, p 937-945.Basu, A., Young, S. W., Suttner, L. J., James, W. C. and Mack, G. H., 1975, Re-evaluation of the useof undulatory extinction and polycrystallinity in detrital quartz for provenance interpretation.Journal of Sedimentary Petrology, v. 45, p. 873-882.Bernier, L. R., 1990, Vanadiferous zincian-chromian hercynite in a metamorphosed basalt-hostedalteration zone, Atik Lake, Manitoba. Canadian Mineralogist, v. 28, p. 37-50.Bloomer, S. H. and Hawkins, J W., 1987, Petrology and geochemistry of boninite series rocks from theMariana trench. Contributions to Mineralogy and Petrology, v. 97, p. 361-377.Bowie, S. H. U., 1967, Microscopy: Reflected Light, In J. Zussman ed. , Physical Methods inDeterminative Mineralogy. Academic Press, New York.Cookenboo, H. 0., 1989, Lithostratigraphy, palynostratigraphy, and sedimentology of the northernSkeena Mountains and their implications to the tectonic history of the Canadian Cordillera.[unpublished MSc. thesis]: The University of British Columbia. 131 p.Cookenboo, H. 0., and Bustin, R. M., 1989, Jura-Cretaceous (Oxfordian to Cenomanian) stratigraphyof the north-central Bowser Basin, northern British Columbia. Canadian Journal of EarthSciences, v. 26, p. 1001-1012.Dick, H. J. B., and Bullet', T., 1984, Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology, v.86, p. 54-76.Dickey, J. S. 1975. A hypothesis of origin for podiform chromite deposits. Geochimica et CosmochimicaActa, v. 39, p. 1061-1074.Dickinson, W. R. and Suczek, C. A., 1979. Plate tectonics and sandstone compositions: AmericanAssociation of Petroleum Geologists Bulletin, v. 63, p. 2164-2182.Dickinson, W. R., Beard, L. S., Brakenridge, G. R., Erjavec, J. L., Ferguson, R. C., Inman, K. F.,Knepp, R. A., Lindberg, F. A., and Ryberg, P. T., 1983, Provenance of North American131Phanerozoic sandstones in relation to tectonic setting. Geological Society of America Bulletin,v. 94, p. 222-235.Duke, J. M., 1982. Ore deposits models 7. Magmatic segregation deposits of chromite. GeoscienceCanada, v. 10, p. 15-24.Duncan, R. H. and Green, D. H., 1987, The genesis of refractory melts in the formation of oceaniccrust. Contributions to Mineralogy and Petrology. v. 97, p. 326-342.Eisbacher, G., 1974, Evolution of successor basins in the Canadian Cordillera, In Dott, R. H., andShaver, R. H., eds., Modem and Ancient Geosynclinal Sedimentation. Society of EconomicPaleontologists and Mineralogists, Special Publication 19, p. 274-291.Ekstrand, E. J. and Butler, R. F., 1989, Paleomagnetism of the Moenave Formation: Implications forthe Mesozoic North American apparent polar wander path. Geology, v. 17, p. 245-248.Evans, B. W. and Frost, B. R. 1975. Chrome-spinel in progressive metamorphism - a preliminaryanalysis. Geochimica et Cosmochimica Acta, v. 39, p. 959-972.Franzinelli, E. and Potter, P. E., 1983, Petrology, chemistry, and texture of modern river sands,Amazon River system. Journal of Geology, v. 91, p. 23-39.Haggerty, S. E., 1979. Spinels in high pressure regimes. hi Boyd, F. R. and Meyer, 0. A. (eds.), Themantle sample: inclusions in Kimberlites and other volcanics. Proceedings of the secondinternational kimberlite conference, v. 2, p. 183-196..Hawkins, J. W. and Melchior, J. T., 1985, Petrology of Mariana Trough and Lau Basin basalts. Journalof Geophysical Research, v. 90, p. 11431-11468.Hekinian, R. 1982. Petrology of the ocean floor. Elsevier Oceanography Series. v. 33. ElsevierScientific Publishing, Amsterdam. 407 p.Irvine, T. N., 1974, Petrology of the Duke Island ultramafic complex, southeastern Alaska. GeologicalSociety of America Memoir 138, 240 p.Loney, R. A. and Himmelberg, G. R., 1989, The Kanuti Ophiolite, Alaska. Journal of GeophysicalResearch, v. 94, p. 15,869-15,900.MacLeod, S. E., and Hills, L. V. 1990. Conformable Late Jurassic (Oxfordian) to Early Cretaceousstrata, northern Bowser Basin, British Columbia: A sedimentological and paleontological model.Canadian Journal of Earth Sciences, v. 27: 988-998.132Malloch, G.S., 1914. Groundhog coal field Summary report of the Geological Survey Department ofMines. p. 69 to 101.Mitchell, R. H., 1991. Kimberlites and Lamproites: Primary sources of diamond. Geoscience Canada v.18, p. 1-16.Molinaroli, E., Blom, M. and Basu, A., 1991, Methods of provenance determination tested withdiscriminant function analysis. Journal of Sedimentary Petrology, v. 61, p. 900-908.Nicolas, A. and Prinzhofer, A., 1983, Cumulative or residual origin for the transition zone in ophiolites;structural evidence. Journal of Petrology, v. 24, p. 188-206.Nixon, G. T., Cabri, L. J. and LaFlamme, J. H. G., 1990, Platinum-group-element mineralization inlode and placer deposits associated with the Tulameen Alaskan-type complex, British Columbia.Canadian Mineralogist, v. 28, p. 503-535.Osborne, M. D., Fleet, M. E. and Bancroft, G. M., 1981, Fe2+ -Fe3 + ordering in chromite and Cr-bearing spinels. Contributions to Mineralogy and Petrology, v. 77, p. 251-255.Pallister, J. S. and Hopson, C. A. 1981, Samail ophiolite plutonic suite: field relations, phase variation,cryptic variation and layering, and a model of a spreading ridge magma chamber. Journal ofGeophysical Research, v. 86, p. 2593-2644.Pasteris, J. D., 1984, Use of indigenous kimberlite minerals, particularly spinels, in the evaluation ofdiamond potential. In Petruk, W. (ed.), Symposium on Process Mineralogy HI: applications inmetallurgy, coal, concrete, smelting, and exploration. Department of Energy, mines, andResources, Ottawa, p. 157-179.Pettijohn, F. J. 1975, Sedimentary rocks. 3rd ed., Harper and Row, New York, 628 p.Pouchou, J. L., and Pichoir, F., 1984. A new model for quantitative analysis: Part I. Application of theanalysis of homogeneous samples. La Recherche Aerospatiale, v. 5, p. 47-65.Radloff, J. K., 1989, Origin and obduction of the ophiolitic Redfern Complex on the Omineca-Intermontane Belts boundary, western Cariboo Mountains, British Columbia. [unpublishedMSc. thesis]: The University of British Columbia 178 p.Ross, G. M., and Parrish, R. R., 1991, Detrital zircon geochronology of metasedimentary rocks in thesouthern Omineca Belt, Canadian Cordillera. Canadian Journal of Earth Sciences, v. 28, p.1254-1270.133Samson, S. D., McClelland, W. C., Patchett, P. J., Gehrels, G. E. and Anderson, R. G. 1989. Evidencefrom neodymium isotopes for contributions to Phanerozoic crustal genesis in the CanadianCordillera. Nature, v. 337, p. 705-709.Smith, M. T., and Gehrels, G. E., 1991, Detrital zircon geochronology of Upper Proterozoic to LowerPaleozoic continental margin strata of the Kootenay arc: implications for the early Paleozoictectonic development of the eastern Canadian Cordillera. Canadian Journal of Earth Sciences, v.28, p. 1271-1284.Suttner, L. J. and Dutta, P. K., 1986, Alluvial sandstone composition and paleoclimate, I. Frameworkmineralogy. Journal of Sedimentary Petrology, v. 56, p. 329-345.Trevena, A. S. and Nash, W. P., 1981, An electron microprobe study of detrital feldspar. Journal ofSedimentary Petrology, v. 51, p. 137-150.Utter, T. 1978, The origin of detrital chromites in the Klerksdorp Goldfield, Witwatersand, SouthAfrica. Neues Jahrb. Mineralogies Abh., v. 133, p. 191-209.Wheeler, J. 0., and McFeely, P., 1987, Tectonic assemblage map of the Canadian Cordillera andadjacent parts of the United States of America. Geological Survey of Canada, Open File 1565.Wheeler, J. 0., Brookfield, A. J., Gabrielse, H., Monger, J. W. H., Tipper, H. W., and Woodsworth,G. J. (compiler), 1991, Terrane map of the Canadian Cordillera. Geological Survey of CanadaMap 1713A, scale 1:2 000 000.Whittaker, P. J., 1982, Chromite occurrences in ultramafic rocks in the Mitchell Range, BritishColumbia. In Current Research. Part A, Geological Survey of Canada, Paper 82-1A p. 239-245.Whittaker, P. J. and Wadkinson, D. H. 1981, Chromitite in some ultramafic rocks of the Cache CreekGroup, British Columbia, In Current Research. Part A, Geological Survey of Canada, Paper 82-lA p. 239-245.Whittaker,P. J., D. H. Wadkinson, 1983, Origin of chromite in dunitic layers of the Mt. Sydney-Williams ultramafic rock complex, British Columbia, in M. J. Gallagher, ed., Metallogeny ofbasic and ultrabasic rocks. Institute of Mining and Metallurgy, London p. 217-228.Wilson, A. H., 1982. The geology of the Great 'Dyke', Zimbabwe: the ultramafic rocks. Journal ofPetrology, v. 23, p. 240-290.134CHAPTER 5DIAGENESIS AND PORE WATER EVOLUTION IN GROUNDHOG COALFIELD STRATA,NORTHERN BOWSER BASINABSTRACTPore water evolution in the northern Bowser Basin of Brititsh Columbia has been studied byintegrating data from sandstone cement paragenesis, mudstone concretion geochemistry, and organicmaturation. This study suggests that diagenesis in shallow marine, deltaic, and coastal plain siliciclasticsediments such as occur in the northern Bowser Basin can be explained by the interaction of seawatermoving through sandstones mixing with acid waters derived from dewatering of interbedded organic richmuds. Sandstones examined in this study are chert and volcanic lithic arenites and wackes derived fromisland arc and sea floor terranes.Sandstone cement paragenesis includes seven discrete cement stages. From earliest to latest thecement stages are: 1) pore lining chlorite; 2) pore lining to pore filling illite; 3) pore filling kaolinite; 4)oil migration through some of the remaining connected porosity; 5) quartz cement; 6) chloritedissolution; and 7) calcite cement. These seven stages of cement record pore water evolution (in thesandstones), which is interpreted as an advection controlled process. Pore water chemistry interpretedfrom the cementation history suggests seawater was the initial pore fluid. Seawater composition changedduring transport through the sandstones, first by loss of Mg2 + and Fe2 + to chlorite precipitation (stage1). Dewatering of interbedded organic rich mudstones probably added Mg2+ and Fe2 + to partiallybuffer the loss to chlorite cementation. Acids produced during breakdown of organic matter mixed intosandstone pore fluids due to further compaction of the muds, leading to reduction of initial alkalinity .Reduction in alkalinity, in turn, favors change from chlorite to illite precipitation (stage 2), and finally tokaolinite (stage 3). Pore waters reached their peak acidity at the time of oil migration (stage 4). Chloritedissolution (stage 5) and quartz precipitation (stage 6) occurred when pores were filled by these acidfluids. Fluid inclusions in fracture filling quartz cements contain petroleum, high-pressure methane, and135methane-rich aqueous solutions. Homogenization temperatures from primary two phase inclusions areconsistent with quartz cementation during progressive heating between approximately 100 ° and 200 °C.Following quartz precipitation, alkaline pore waters were re-established, as evidenced by late stagecalcite cementation (stage 7).Mineralogy of mudstone concretions is mostly siderite and ferroan dolomite, consistent withformation in the methanogenic zone of early diagenesis in organic rich siliciclastics. C isotope ratiosrange from +1 to +5 per mil, consistent with formation during later rather than earlier methanogenesis.0 isotope ratios suggest precipitation from initial meteoric to brackish waters at least as depleted as -10per mil SMOW, although the possibility of later re-equilibration cannot be eliminated.Pore water evolution in the mudstones is interpreted as a diffusion controlled process, withinitial fluids dominated by meteoric waters from a cool, temperate climate. Organic diagenesis leads tohigh dissolved carbonate in the muds. Carbonate combined with Fe 2+ and Mg2 + presumably derived, atleast in part, from dissolution of mafic minerals.Organic maturation, as measured by vitrinite reflectance, records steep paleogeothermalgradients, and high maximum basin temperatures. Forward modelling suggests that the steep thermalgradients and high maximum temperatures were sufficient to account for the interpreted pore waterhistory. The modelling fits best with paleogeothermal gradients of 60° to 65°C/km. Such highgeothermal gradients suggest the Bowser Basin accumulated above oceanic crust in a back-arc basin.INTRODUCTIONDiagenesis of siliciclastic rocks is controlled by temperature, solute characteristics and, to a lesserdegree, pressure of the contained pore fluids. Temperature, pressure, and solute characteristics in turnrelate to sediment and basin geologic history. Analysis of diagenetic effects with the goal of recoveringinformation about the pore fluids is, therefore, a potentially powerful approach to understanding theorigin and development of siliciclastic basins. Unfortunately, the complexity of pore fluids makes136unambiguous determination of temperatures and solute characteristics difficult. Ambiguity in diageneticinterpretations can, however, be reduced by integrating observations and analyses from multiple lines ofevidence.Three lines of diagenetic evidence used in the diagenetic study described in this chapter aresandstone cement paragenesis, mudstone concretion geochemistry, and organic matter maturation. Eachof these data sets records a different aspect of diagenesis in a siliciclastic succession. Sandstone cementparagenesis, for example, records changing compositions of pore fluids controlled by advection, and mayallow inferences regarding depositional conditions and later tectonic or metamorphic events. Mudstoneconcretion geochemistry, in contrast, records conditions of pore fluids that are dominantly affected bydiffusion processes, and may allow inferences about depositional processes and organic matter content.The diagenesis of organic matter, which is primarily controlled by thermal exposure after burial, permitsinferences concerning the maximum temperatures affecting sediments.The siliciclastic succession examined is the Bowser Basin, which is one of a series of Jura-Cretaceous intermontane basins in the Canadian Cordillera. Methods employed in this study includepetrographic and geochemical study of sandstone paragenesis and mudstone concretions, and vitrinitereflectance studies of organic matter. The various diagenetic observations are integrated to form the basisof a basin maturation model which is described in a later section. This model has implications to thedepositional history and tectonic development of the Bowser Basin and other sedimentary basins of theCanadian Cordillera. These implications are discussed in the final section of this chapter.SANDSTONE CEMENT PARAGENESISDifferent mineralogies of cement precipitate in response to changing composition, temperatureand pressure of the pore fluids. The paragenetic sequence for cements described herein has beenestablished by petrographic examination of representative samples collected from all stratigraphichorizons exposed in the study area. Examined samples include very fine, fine and medium grainedsandstones. Original depositional textures and cement paragenesis are obscure in most samples, whichare severely compacted. Rare samples in which abundant early cementation has occurred, however,137preserve depositional textures and cement paragenesis. The paragenetic sequence described below isbased on examination of mostly medium grained sandstones in which intergranular cements areexceptionally well preserved. More severely compacted sandstones were excluded from considerationduring initial establishment of the paragenetic sequence. Following establishment of the parageneticsuccession, more severely compacted sandstones were then examined, and importantly, did not contradictthe established paragenetic sequence. The sandstone cement paragenetic sequence is summarized below(Fig. 5-1). A more detailed description of each cementation stage and its implied pore water conditionfollows this summary1) Isopachous chlorite cement formed on grain surfaces. Crystals are 10 to 25 pmlong, and line pores in continuous coats with individual crystals oriented perpendicularto the grain surface.2) illite cement occurs as pore-lining or pore-filling cement inside some pores earlierlined by authigenic chlorite.3) Kaolinite cement occurs as pore-filling cement in some sandstones previously linedby chlorite, illite, or both chlorite and illite.4) Pore filling reservoir bitumen ("dead oil"), as determined by micro-Fouriertransform infrared analysis, occurs in some chlorite lined pores.5) Chlorite dissolution follows oil migration in sandstones that still have interconnectedporosity. Chlorite dissolution is either partial or complete, and is in partcontemporaneous with quartz cement (see below).6) Quartz cement occurs as syntaxial overgrowths, or micro- and macro-crystalline porefilling cements. Macro-crystalline quartz cement fills fractures during this stage.138Figure 5-1: Stages of sandstone cementation in the northern Bowser Basin.Sandstone ParagenesisDepth.AuthlgenicChlorite ^_ 4crystal rlilite^. . A growth13 st Kaolinite r 1.. ..1- ' Oil migrationkr ,_g8.: Chlorite dissolution ..........T 1 Quartz .. ,^.....4 7 A...^.... 1it! Calcite : ' " "I 1grainrep mendI— 15139Hydrocarbon fluid inclusions in fracture filling quartz cements indicate that oilmigration in part was contemporaneous with the quartz cementation stage.7) Calcite occurs as a late stage pore filling cement, commonly with poikilitic texture,and as late stage grain replacement cement. Calcite also fills fractures, either alone oras a successor to earlier quartz cement in (presumably) reactivated fractures.Not all of the stages of cementation summarized above occur in a single sandstone, but the order of thecements is consistent throughout the examined sample suite. For example, the combination ofisopachous chlorite and calcite is present only in some thin sections, but where present, the chloritecement always precedes the calcite. Similarly, reservoir bitumen never precedes chlorite cement, andquartz cements consistently post-date both chlorite and reservoir bitumen Calcite never precedes quartz,reservoir bitumen, or chlorite cements.Although cement stages are never observed out of order, some cement stages are commonlyabsent from particular sandstones. For example, isopachous chlorite cement is only rarely followed byreservoir bitumen, and is more commonly followed by calcite cement. Observed cement stages may varyacross individual thin sections. For example, chlorite cement may have formed isopachous coats thatcompletely occlude some pore throats, and in those pores no subsequent cementation or dissolutionoccurred. Elsewhere in the same thin section, pore throats were not completely blocked, and chloritedissolution occurred. Subsequent calcite cement in such samples occurs either directly on grain surfaces,or upon a partially formed coat of chlorite, depending on the extent of chlorite dissolution.The paragenetic sequence listed above is an ideal sequence of diagenetic stages that is morecomplete than the cementation record preserved in any individual sandstone. By considering all thecement stages listed above together, however, a more complete record of pore water evolution can bederived. The petrographic appearance of each cementation stage is described below, and the pore waterchemistry implied by each stage is discussed in the following section.140CHLORITE CEMENTATION STAGEChlorite cement forms continuous isopachous coats as much as 25 pm thick, of light green toyellow-green (transmitted light) crystals oriented perpendicular to grain surfaces (Fig. 5-2). The cementsform regardless of substrate composition, as demonstrated by equal thickness of cements around grainsirrespective of mineralogy. In some pores, chlorite has clearly grown out into the pore space and crystalsdisplay regular crystal shape. In other pores, although crystal orientation remains perpendicular, theinterior surface of the chlorite cement is irregular, suggesting subsequent dissolution by pore waters.Chlorite cements subjected to later dissolution are commonly very thin in parts of a thin section, and maybe altogether absent in still other parts where dissolution was apparently complete.Chlorite cement forms early, as demonstrated by its formation directly on grain surfaces, andisopachous habit which implies growth into empty pore space. However, some compaction precededchlorite cementation, as evidenced by the common occurrence of long grain contacts, and general lack oftangential contacts.Pyrite framboids occur rarely in association with isopachous chlorite cements of the northernBowser Basin. In some samples, the framboids are clearly located against grain surfaces and therefore, inthose samples at least, preceded chlorite cementation (Fig. 5-3). Pyrite framboids are well known toform soon after burial (c.f. Gautier and Claypool, 1984) in response to the low oxygen-high sulfate poreconditions characteristic of the sulfate reducing zone (Berner, 1980). Such conditions are usuallyrestricted to the upper few metres of organic rich sediments (Gautier and Claypool, 1985). Theassociation of chlorite cement with pyrite framboids supports early formation of the chlorite.Chlorite cementation stage pore water chemistryPrecipitation of chlorite cement during the earliest sandstone cementation stage has implicationsregarding initial pore fluid composition. Laboratory experiments demonstrate that141Figure 5-2: Authigenic chlorite coats on sandstone clasts of varying compositions in plane polarizedlight. Qtz= quartz; VRF= volcanic rock fragment; Ch= chert. Scale bars are 50 Am.142Figure 5-3: Sketch of a microscope view of isopachous chlorite cement with earlier pyrite framboids.After the chlorite cementation stage, compaction of the pore space caused the chlorite cement to breakaway from parts of the grain surfaces. In the top left quadrant, 60 pm of overlap in the chlorite cementoccurs (in the form of a small reverse fault). In the lower right quadrant, compression of the pore spacecaused the chlorite cement and pyrite framboids contained within it o fold out from the grain surface intothe remaining pore space. Pore-filling poikilitic calcite cement precipitated after the chlorite cementationstage. A and B are optically distinct crystals that both continue beyond the illustrated pore space.14360 micrometresprecipitation of chlorite cement requires pore solutions enriched in Fe 2 + and Mg2 + (Small et al., 1992).Mg2 + activities must be relatively high or precipitation of other authigenic clays (illite or kaolinite) isfavoured (Fig. 5-4; Jahren and Aagaard, 1989). In addition, chlorite precipitation suggests basic waters,because chlorite is unstable when exposed to warm acidic solutions (Foscolos, 1985). This instability isdemonstrated in the laboratory by the common use of weak acid at 80° to 100°C to remove chlorite fromclay mineral mixtures in order to facilitate XRD identification of clay minerals (Foscolos, 1985), and inoil fields by the success of injection of weak acids in dissolving chlorite cements (Almon and Davies,1981). The relationship of chlorite stability to H+ activity is shown quantitatively in the phase diagramin figure 5-4 (Jahren and Aagaard, 1989). The most likely source of large volumes of basic water isseawater, which is typically buffered between pH 8.0 and 8.4 (Brownlow, 1979), and chlorite cementsare commonly associated with marine or brackish water deposits (Thomas, 1981). Normal seawater isalso a source of abundant Mg2+. Fe2 + is not abundant in seawater, but concentrations rise near themouths of major rivers, as demonstrated by the tendency of iron-rich green marine clays to precipitate inshallow marine waters near river mouths (Odin and Masse, 1988). Another potential source for bothFe2 + and Mg2 + ions is the breakdown of Fe-Mg silicates. A combination of river and Fe-Mg silicatessources probably contributed solutes to Bowser Basin pore waters because the sediments were derived inpart from erosion of mafic and ultramafic rocks (Chapters 3 and 4), and were deposited in shallowmarine, deltaic, and fluvio-deltaic settings (Chapter 2).Temperatures during formation of chlorite cements can be constrained only within a broadrange. Textural evidence in this study demonstrates that chlorite cementation occurred early, andtherefore temperatures must have been relatively low. Compacted textures in chlorite cementedsandstones, however, suggest that precipitation did not initiate immediately following deposition. Initialtemperatures of formation were therefore somewhat elevated from depositional temperature and thereforewere probably above 15° to 25°C. An upper end to the chlorite cementation temperature range ofroughly 80° to 100°C is implied by subsequent oil migration (see discussion under oil migration headinglater). Chlorite precipitation is also limited by the production of acid waters due to the maturation144Figure 5-4: Phase diagram showing that chlorite precipitation is favoured by relatively high Mg2+ andlow H+ activities in solution compared to illite or kaolinite. After Jahrens and Aagaard (1989).100° C, 4 x 10 7 Pascals, Qtz sat145K-feldsparIlliteKaolinite1^1^1^I12^4^6^8^10log (log++Chlorite^I ^I1 ^14^16/a 2H+)of organic matter, which initiates and peaks at approximately 80° to 100°C (Surdam and Crossey, 1985;Surdam and MacGowan, 1987). In summary, the pore water chemistry deduced for the chlorite cementstage implies a basic, Mg2 + and Fe2 + rich solution at temperatures above 15° and below about 90°C.Such a pore fluid is compatible with derivation from normal seawater that has elevated Fe 2 + and(perhaps) Mg2 + concentrations.The interpreted temperature range (above 15°C and below 90°C) is reasonable based onprecipitation temperatures of authigenic chlorite reported in the literature. Examples from the literatureinclude clays structurally similar to chlorite which precipitate in warm shallow seas at the sediment waterinterface (Odin and Masse, 1988), and authigenic chlorites from the North Slope of Alaska that formedin sandstones never exposed to temperatures greater than approximately 60°C (Smosna, 1988). In theseNorth Slope sandstones, chlorite cementation precedes quartz overgrowths and later calcite pore-fillingcements (Smosna, 1988). Further examples include subarkoses and sublitharenites of the Spindle Field,Colorado, that contain a similar sequence of early chlorite cement followed by quartz and then calcitecements (Pittman, 1988). Authigenic chlorites formed between 70° and 160°C also have been describedfrom the North Sea by Jahren and Aagaard (1989).ILLITE CEMENTATION STAGEIllite occurs as either a pore lining or pore filling cement in some chlorite cemented sandstones.Illite is distinguished from chlorite in thin section by lack of green color in plane polarized light andhigher birefringence in cross polarized light, and in SEM analysis by characteristic filamentous andribbon-like crystal shapes of illite. Energy dispersive spectrometry of the wavy pore-filling cementsconfirms the occurrence of K± and lack of Mg 2 + which supports the mineral identification as illite.Textural relationships demonstrate that illite consistently forms following chlorite cementation.Some samples with illite also contain later pore-filling quartz cement, but illite has not been noted in anysample with reservoir bitumen. Illite, therefore, follows chlorite cement and precedes quartz cement.146111ite Cementation Stage Pore Water ChemistryIIlite precipitation is favoured over chlorite under conditions of higher H+ activity and lowerMg2+ activity (Fig. 5-4; Jahren and Aagaard, 1989). As shown by the phase diagram in figure 5-4, K+activity need not increase to move a fluid from the chlorite stability field to either the illite or kaolinitestability fields. Mite cementation following chlorite can be explained by a relatively simple model ofpore water evolution. Some combination of lowering Mg 2 + concentrations and increased acidity is allthat is required. Assuming that seawater was the starting pore fluid as inferred above, Mg 2 +concentration should be reduced as a direct result of chlorite cementation. As formation temperaturesrise due to continued burial, the alkaline character of normal seawater should become more acidic due toproduction of organic acids associated with organic maturation. K+ originally present in seawater issubsequently available to precipitate as authigenic illite.Mite precipitation may be limited by K+ available from seawater, because the provenance ofnorthern Bowser Basin sandstones does not favor increasing K+ concentrations in the pore fluids(Chapters 3 and 4). Bowser Basin provenance consists of largely unmetamorphosed oceanic terranes,which generally lack many common K+ bearing mineral phases such as K-feldspar or muscovite. K+bearing phases in Bowser Basin sediments derived from those oceanic terranes are largely limited touncommon biotite, and therefore only limited sources of K+ are available to authigenic clays fromorigins other than seawater.KAOLIMTE CEMENTATION STAGEAuthigenic kaolinite occurs in some Bowser Basin sandstones previously cemented by chlorite orillite. Kaolinite is recognized petrographically in this study based on its vermiform habit, andbirefringence lower than illite. Kaolinite forms as pore filling cement inside pores already lined bychlorite or illite, which is strong petrographic evidence that kaolinite precipitation occurred subsequent toillite or chlorite cementation. Kaolinite is also commonly associated with reservoir bitumen in the pores,which is consistent with kaolinite precipitating after the illite cementation stage described above, and147prior to or contemporaneous with the oil generation and migration stage described below.Kaolinite Cementation Stage Pore Water ChemistryKaolinite precipitation is the next stage in the evolution of pore water that began as seawater,and subsequently precipitated first chlorite and then illite. Pore water in this evolutionary process firstloses Mg2 + and Fe2+ during chlorite precipitation, and then K+ during illite precipitation. Acidity inthe pore waters increases simultaneously, due to creation of organic acids associated with the breakdownof organic matter. The trends of decreasing Mg 2+ and K+ combine with increasing H+ activity andtogether lead to pore water conditions favouring kaolinite precipitation (Fig. 5-5).Off, MIGRATION STAGEOpaque pore filling material occurs in some pores and fractures that were earlier lined withisopachous chlorite cement. Micro-Fourier transform infrared analysis (mF1^ lit) demonstrates that theopaque material is reservoir bitumen (dead oil) that has lost most of its aliphatic components (probablydue to post-migration thermal degradation; Fig. 5-6). The oil migrated into pores which were alreadylined with thick coats of chlorite cement (Fig. 5-7a), indicating that oil migration followed chloritecementation, but preceded any significant chlorite dissolution. Reservoir bitumen also occurs in fracturesin some northern Bowser Basin mudstones (Fig. 5-7b), including fractures in silicified wood, wherehydrocarbons are preserved in fluid inclusions in quartz cements, as described later.Oil migration stage pore water chemistryGeneration and migration of oil occurs as a result of organic maturation processes that areconstrained to relatively narrow temperature ranges. Oil generation initiates after suitable organic matterhas been exposed to sufficient heat for a sufficient length of time (thermal exposure), and148K-feldspar (. 1Illite^ChloriteKaoliniteI^1Figure 5-5: Three successive cements in Bowser Basin sandstones occur as a result of changing porewater chemistry described in this phase diagram from Jahren and Aagaard (1989). Pore water chemistrystarts near position 1, with high Mg2 + and relatively low H+ activities, as expected for fluids derivedfrom seawater. Reduction in Mg2 + activity due to chlorite cementation, and increase in dissolved acidsderived from processes of organic matter maturation shift pore water chemistry towards position two, asrecorded by precipitation of illite subsequent to chlorite (see text for descriptions). Position three isreached in some pores, where kaolinite is precipitated following or contemporaneous with oil migration.100 C, 4 x 10 7Pascals, Qtz sat.2^4^6^8^10^12^14 16log (aMg++ /a 2 + )149Figure 5-6: Micro-Fourier transform infrared analysis plots of opaque material in pores and fractures ofBowser Basin sandstones demonstrates that the material is organic matter with very low aliphaticcomponent. Loss of aliphatics is typical of dead oil, otherwise referred to as reservoir bitumen.119 —[ZSL — ^euaid-joino ogew am gpgMS97,8 — ^150eouecposqyFigure 5-7: Chlorite cemented sandstone saturated with reservoir bitumen: a) oil in intergranularporosity; b) oil in chlorite cemented microfracture. Scale bars are 100 Am.151intense oil generation initiates between temperatures of approximately 60° to 130°C (Hunt, 1979). Oilgeneration is essentially complete in most sediments at temperatures below 150°C (Hunt, 1979),although light oil, condensate and methane production commonly continues to 180°C (Tissot and Welte,1984). Vertical oil migration may cause the hydrocarbons to move into cooler sandstones, but it seemsreasonable to assume that the oil migration stage occurred at temperatures of approximately 90° to130°C. The chlorite cementation stage occurred entirely prior to oil migration and can be restricted,therefore, to temperatures below approximately 80° to 90°C. Elite and kaolinite precipitation followchlorite cementation in some pores and probably occur simultaneously with initial oil migration as porewaters became less alkaline. Textural observations are consistent with such an interpretation becausesome chlorite lined pores are filled with illite that is dark brown in plain polarized light, suggesting thatthe illite is coated by migrating oil. Overlap of oil migration with illite and kaolinite precipitation ischemically reasonable because illite, kaolinite, and oil migration are all associated with increasingly acidpore water conditions (Carothers and Kharaka, 1978; Jahren and Aagaard, 1989).CHLORITE DISSOLUTION STAGEChlorite cement dissolution is evident in many northern Bowser Basin sandstones. The clearestevidence for chlorite dissolution is observed in thin sections in which some pores are completelyoccluded with thick isopachous chlorite cements, whereas pores elsewhere in the same thin sectioncompletely lack (or have only thin) chlorite cements. Such sandstones are interpreted to haveexperienced an early stage of chlorite precipitation that closed some pores completely, but left other porespartially open. Acidic waters subsequently moved through the remaining connected pores, leading todissolution of chlorite cement. Dissolution could not occur in pores that were completely occludedduring early chlorite cementation, presumably because acid waters could not permeate through theblocked pore spaces. The result is that some pores are filled with chlorite and other pores in the samesandstone lack chlorite altogether. Chlorite dissolution occurred prior to quartz and calcite cementation,because both quartz and calcite cements occur in pores with all or part of the earlier cement removed.152Quartz occurs rarely as the final pore filling cement inside a reservoir bitumen coating which isin turn inside pore-lining isopachous chlorite cement (Fig. 5-8). Some of these pores that containchlorite, reservoir bitumen, and quartz cements are missing continuous chlorite coats. Incompletechlorite coats are unusual, and only occur where quartz is present in addition to reservoir bitumen. Thisrelation of quartz with incomplete chlorite coats implies chlorite dissolution occurred after oil migration,but before or at the same time as quartz cementation.Chlorite dissolution stage pore water chemistryThe inference of basic waters for the chlorite cementation stage described earlier suggests asimple and plausible explanation for the chlorite dissolution stage: the pore waters change from basic toacidic. Such a change to acidic conditions continues the trend of decreasing alkalinity inferred for theillite, kaolinite, and oil migration stages described above.QUARTZ CEMENTATION STAGEQuartz cements occur as syntaxial overgrowths and micro- or megacrystalline pore -fillingcements. Quartz overgrowths in some pores form smooth crystal faces, and rarely grow into pore-fillingcements. Micro- or megacrystalline textures appear to depend at least in part, on the substrate.Megacrystalline quartz cements tend to form either on coarsely crystalline quartz grain surfaces, or onother earlier cements. The coarsely crystalline quartz cements tend to be clear and contain only rare andsmall fluid inclusions. Micro-crystalline pore-filling quartz cements generally occur in contact with chertgrains.The relative timing of the quartz cement stage is interpreted petrographically from theobservation that quartz cement occurs in some pores inside thick isopachous chlorite cements. Asmentioned earlier, quartz also rarely occurs upon reservoir bitumen lined isopachous chlorite153Figure 5-8: Authigenic quartz fills a pore lined first by isopachous chlorite cement and then coated byoil. This relation demonstrates that quartz cementation followed oil migration in the sandstones.Chlorite cement in this field of view is irregular, suggesting chlorite has been partially dissolved.Partially to completely dissolved chlorite cements are commonly closely associated with quartzcements, suggesting that the acid conditions favouring quartz cementation lead to dissolution ofchlorite. Scale bar is 50 Am.154grain coats. In fractures containing both quartz and calcite cements, textural relations consistentlyindicate quartz cementation preceded calcite cementation. Quartz cementation, therefore, followed oilmigration, and preceded calcite cementation.Fluid inclusions in fracture-filling quartz cementsQuartz cements occur in fractures as well as in sandstone pore spaces. Fluid inclusionspreserved in the fracture-filling cements are larger and more varied than those in pore-filling cements.Quartz crystals as large as 7 cm across occur in some fractures, and some contain fluid inclusions up to10 pm (although rarely greater than 5 pm) in largest dimension. Observations of fluid inclusions infracture filling quartz cements provide independent constraints on pore fluid composition and evolution.Crushing quartz fracture filling cements releases abundant bubbles. The large size andabundance of bubbles demonstrates that high pressure gas is contained in some of the fluid inclusions.Solubility characteristics of the bubbles are useful for qualitative identification of contained gas phases(Roedder, 1974). The bubbles persist in immersion oil, but they dissolve rapidly in kerosene.Dissolution in kerosene is rapid for bubbles less than 5 pm in diameter, and takes less than two secondsfor bubbles up to 10 pm in diameter. Larger bubbles with lower surface area to volume ratios dissolvemore slowly until they reach diameters less than 10 pm, at which point they disappear rapidly. Relatingbubbles to individual source inclusions is difficult. However, growth zones with many primary singlephase, dark gray gas filled inclusions commonly release large numbers of bubbles. In one such growthzone, a large, dark grey single phase inclusion with an irregular shape formed a bubble that dissolvedrapidly. Bubbles also appear when crushing zones of two phase, liquid-rich inclusions, and aftercrushing the formerly two phase inclusions are single phased, suggesting that the bubbles releasedoriginated as the vapor phase of the two phased inclusions.Quartz cement from fractures in silicified wood, which is common in fine grained facies, wasexamined petrographically and on a USGS-type heating and freezing stage. Silicification of the woodbegan early, as demonstrated by the uncrushed cell walls and fine cellular detail preserved in the wood.155Early silica cementation within fossil wood cells probably depends on chemical conditions related tobiogenic degradation within the cell spaces, and those chemical conditions may be largely unrelated topore water conditions in associated sandstones (Hesse, 1990). Subsequent to the earliest, intercellularcements, however, quartz fracture filling cement formed, which are likely related to advection of porefluids. This cement precipitated in a distinct succession of layers that becomes relatively youngertowards the middle of the fracture. The earliest fracture filling cements occur on fracture walls, and havewell shaped crystal faces projecting into the fractures. Fluid inclusions in this early fracture fill cementare generally small, and include both clear, single phase aqueous inclusions, and yellow-brown to darkbrown coloured (Fig. 5-9a) one or two phase hydrocarbon inclusions (aqueous inclusions are almostalways clear, and brown inclusions are almost certainly hydrocarbons; Roedder, 1974). Opaquereservoir bitumen ("dead oil") covers the interior crystal face surfaces of some of these early quartzcrystals (Fig. 5-9b).A younger phase ("medial phase") of quartz cement occurs inside the earliest fracture cement.This medial phase consists of several layers of quartz crystals which contain two phase aqueousinclusions with irregular to smooth shapes, some of which occur in distinct growth zones. In oneprimary growth zone, homogenization temperatures (Th) for two phase aqueous inclusions withconsistent liquid to vapour ratios (L/V) were measured by temperature cycling. Nine such inclusionsfrom the same growth zone, each less than 3.5 pm in diameter with smooth but irregular shape andcontaining "dancing" bubbles, yielded Th between 90.5° and 97.3°C (plus or minus 1.5°C). Due to thesmall size of the inclusions, final melting could not be directly observed, but reappearance andunhindered movement of the vapor bubble upon heating was determined by using a freeze/refreezetemperature cycling technique. Final melting (Tm) determined for the same set of inclusions yieldedtemperatures between +4.5° and +24.2°C (dropped to +22.8° in repeat run of highest T. inclusion).Because T. is above 0°C, final melting is assumed to be final melting of clathrate. A larger inclusionwith apparently smaller L/V did not homogeni  7it. to much higher temperatures (+120°C), and thereforeis assumed to have leaked. T. for this larger inclusion was +1.9°C, consistent with clathrate melting156under lower pressures than the smaller, consistent Th inclusions. Three fluid inclusions from anothermedial phase quartz crystal yielded less consistent Th, but still suggestive of low temperaturehomogenization. Th for these three inclusion were between 92° and 102°C.Single phase inclusions, which are gas filled based on their dark gray appearance' s (resultingfrom low refractive index; Roedder, 1974; 1984) also occur in medial phase quartz crystals. Some ofthese darker inclusions are clearly primary, and some that have irregular but smooth walled shape occurin the previously described growth zone along with the measured inclusions discussed above withconsistent L/V and Th. Single phase gas inclusions from other, probably later, primary growth zoneshave well developed negative crystal shapes. Late in the medial phase, many large irregular shapedprimary hydrocarbon inclusions formed, apparently as a single event.The youngest quartz cement in the examined fracture ("late stage") precipitated adjacent to thecrystals containing the large hydrocarbon inclusions mentioned above, and consists of clear quartz. Latestage fluid inclusions include primary two phase aqueous inclusions with smooth walls, which tend toexhibit somewhat poorly developed negative crystal shape. Single phase gas inclusions with goodnegative crystal shapes are also common in the same growth zones. In one growth zone, primary, liquid-rich, two-phase, aqueous inclusions with consistent L/V ratios yielded Th between 141° and 162°C.The measured inclusions vary in shape from moderately irregular to negative crystal shape, and the twolargest inclusions are over 7 pm in each dimension. Six primary two phase inclusions (negative crystalfaces) from a different growth zone with apparently consistent L/V ratios yielded rather inconsistent Th,ranging from 161° to 194°C.Fluid inclusion implications to pore water evolutionCrushing quartz cements demonstrates that high pressure gas is abundant, and the gas isprobably methane due to the bubbles' rapid solution into kerosene (Roedder, 1974). Many bubbles157158Figure 5-9: a) Hydrocarbon fluid inclusions (see arrows) in a single crystal of quartz cement from afracture in silicified wood. The inclusions formed in a single growth zone near the edge of this crystal,and are shown in a micrograph combining three successively deeper focus levels. The largest inclusionsin the picture are 5 to 7 pm across; b) Hydrocarbon fluid inclusions, and opaque bitumen (arrows)between quartz crystal surfaces. Fracture is 175 pm wide in silicified wood.emerged from zones with abundant single phase inclusions, and one bubble was observed to emerge froman individual dark gray single phase inclusion, which clearly indicates that the single phase inclusionscontain methane. In addition, methane occurs pervasively in organic rich sediments, (Roedder, 1984;Hanor, 1987) which strongly supports the suggestion that the high pressure gas released during crushingis methane. Complete dissolution of even very large bubbles (formed by coalescence of many smallerbubbles) suggests that little or no CO2 (or other gas species) is mixed with the methane. Althoughnecking of inclusions at reduced temperatures can isolate vapor bubbles that were originally in solution,the occurrence of many single phase gas inclusions with good negative crystal shapes in the same primarygrowth zones as two-phase inclusions with consistent L/V ratios strongly suggests that methane wastrapped both as a separate phase and in aqueous solution (Bodnar et aL , 1985). Observations of liquid-rich two phase inclusions during crushing were more ambiguous, but suggest that many (most or all) ofthe aqueous two phase inclusions probably also contain a high pressure methane vapor phase. Methaneadds many difficulties to quantitative microthermometry measurements in fluid inclusions, includingaffecting melting temperatures of saline solutions and forming a clathrate that is difficult to see in verysmall inclusions (Hanor, 1980). However, the most serious difficulty is probably the potential for in-situproduction of excess methane by post-trapping thermal breakdown of acids or hydrocarbons (Hanor,1980). Valuable qualitative information is available, however. T. of aqueous solutions were +1.9° to+24.2°C, indicating formation of a methane clathrate. Consistent Th from medial phase primaryinclusions suggests minimum trapping temperatures greater than 97°C. Less consistent Th from latephase cements suggest higher trapping temperatures, probably in excess of 161°C to 194°C.Quartz cementation stage pore water chemistrySilica precipitation is favoured by acidic pore waters (Friedman, et al. 1974; 1992).Increasingly acidic pore water conditions were established during oil migration as demonstrated first bythe illite cementation stage and then the subsequent chlorite dissolution stage. The quartz cementationstage in sandstones follows chlorite dissolution and may be explained as a simple continuation of thetrend of increasingly acidic pore water conditions. The temperature range for the quartz cementation159stage is constrained at the low end by temperatures attributed to the onset of oil generation/migration,and the generation of pore water acidity probably occurred between 80° to 130°. Primary hydrocarboninclusions demonstrate that fracture filling quartz cements in silicified wood precipitated in partcontemporaneously with oil migration. The fluid inclusions also demonstrate the occurrence of abundantmethane in the pore waters, both as a separate gas phase and in solution in aqueous fluids. Th forselected primary two phase inclusions suggests that quartz formed contemporaneously with oil migrationformed at temperatures near 100°C. Late quartz formed after oil migration, probably at temperaturescloser to 200°C (Th between 141° and 194°C: minimum trapping temperatures higher dependent onpressure correction). Organic acids are destroyed by thermal reactions between 120° and 200°C,leading to loss of pore water acidity, and presumably terminating quartz cementation (Surdam andCrossey, 1985).CALCITE CEMENTATION STAGEAuthigenic calcite occurs in pores and fractures as coarsely crystalline or poikilitic texturedcement. Poikilitic crystals are up to several millimetres across, completely surrounding detrital grains,and are usually twinned. Textural relationships demonstrate that calcite formed after both chlorite andquartz cement stages. Calcite cementation appears to have continued to great depth in the northernBowser Basin. Some sandstones from lower stratigraphic levels (undivided Bowser Lake Group andCurrier Formation) are calcite cemented and replaced to such a degree that calcite forms more than 30%of the rock volume (visual estimate). In these sandstones, original depositional textures are obscured, butit appears that most of the calcite occurs as grain replacement and that replacement followed extensivecompaction. Other sandstones from higher stratigraphic levels (upper McEvoy and Devils ClawFormations) commonly have extensive poikilitic calcite pore-filling cements. In contrast to thesandstones with extensive replacement cements from lower stratigraphic levels, the pore-filling cementspreserve and visually outline the grain contacts. The grain contacts are dominated by pressure solutionfeatures, suggesting that calcite cements which occur in these pore-filling cements also followedsignificant burial. However, these sandstones are from the younger stratigraphic units and therefore were160never buried as deeply as those sandstones dominated by grain replacement calcite cements.Calcite cementation stage pore water chemistryCalcite precipitation as a late stage cement implies return of alkaline pore water conditions suchas prevailed during the early chlorite cement. However, unlike the alkaline waters that led to chloritecementation, pore waters during the late calcite cementation stage were probably not replenished byrelatively unaltered seawater. Mg 2 + concentrations during the late stage calcite cementation weresignificantly lower than in seawater, because dolomite precipitation is thermodynamically favoured overcalcite at elevated temperatures unless Mg 2 + concentrations are very low relative to Ca2 + (Hardie,1987). Pore filling calcite cementation is assumed to have initiated at temperatures near 200°C, becauseit followed quartz cementation which occurred up to at least 200°C. At such elevated temperatures,acids generated during organic maturation have been thermally broken down (Carothers and Kharaka,1978). Calcite becomes less soluble with rising temperatures, and therefore late stage calciteprecipitation is expected to continue as long as formation temperatures rise. Extensive late stage calcitegrain replacement as experienced in the older strata supports the re-establishment of alkaline pore waters,because silica solution (necessary for silicate grain replacement) requires alkaline conditions (Friedman,et al., 1974; 1992).The source of the Ca2 + is unclear. Ca2 + could remain from connate seawater incorporated withthe sediments during or shortly after deposition, and only during late stage diagenesis be able toprecipitate with the return of alkaline pore water chemistry. Alternatively (or additionally), Ca 2 + couldbe released from minerals in the sediments. The provenance of the Bowser Basin sediments, whichincludes mafic and ultramafic rocks of oceanic and island arc origin (Chapters 3 and 4), suggests thesediments may have contributed to Ca 2 + in the pore waters. Ca2 + occurs in plagioclase feldspar in suchrocks, and is further taken up from seawater in large amounts during some types of hydrothermalalteration (Seyfried, et al., 1988). Probably Ca2 ÷ is released from feldspars into pore fluids during latestage diagenesis in the Bowser Basin. A further potential source of Ca 2 + could be dissolution of161biogenic carbonate buried with the sediments. No primary carbonate beds have been identified within theBowser Basin siliciclastic succession, and shelly fauna is rare to absent above the lower CurrierFormation, and therefore this source of Ca2 + is presumably relatively minorSUMMARY OF PORE WATER EVOLUTIONSandstone pore water evolution in the northern Bowser Basin can be adequately described byinteraction of three factors: 1) initial fluid composition dominated by seawater; 2) dissolution of unstablesiliciclastic sediments leading to mobilization of Mg, Fe, Si, and Ca; and 3) organic matter maturationleading first to the production of acids and later to thermal breakdown of the acids. Interaction of thesethree factors results in evolution of pore water that can be described in four successive stages summarizedin figure 5-10 and below:Stage 1) Alkaline pore waters interact with mafic and ultramafic detritus in Bowser Basinsediments leading to elevated Fe2 + and Mg2 + concentrations. Precipitation ofauthigenic chlorite is favoured from these alkaline, Fe and Mg-rich pore waters attemperatures below approximately 90°C.Stage 2) Pore waters become less alkaline due to breakdown of organic matter, leading first toillite precipitation, which depletes K+ supplied from seawater. Kaolinite precipitationoccurs following K+ depletion in sandstones that have not been completely occluded bypore filling cements.Stage 3) Pore waters become increasingly acidic as a result of continued generation of organicacids. Oil migration, chlorite dissolution and quartz162OUR TO POCK/MTH:I tNrEmenomsClrE TO PEOPITATIO1CATION CONCENTRATIONDEPLETION ENRICHMENT ALKALINITY Pore waterStageChlorite precipitation 1!Hite precipitationKaolinite precipitationORGANIC^oil migrationMATTER^chlorite dissolutionMATURATIONquart precipitation04'calcite precipitationcarbonate replaces silicates342Figure 5-10: Four stages of pore water chemistry shown as a function of changing cation concentrationsdue to rock-water interactions, and alkalinity due to organic maturation.163precipitation occurred under these acid water conditions. Acid waters probablyinitiated at temperatures between approximately 80° and 130°C, and persisted totemperatures near 200°C.Stage 4) Pore water acidity is eliminated due to thermal degradation of organic acids.Thermal breakdown of organic acids in oil field waters begins at approximately 120°C,and is complete by approximately 200°C (Carothers and Kharaka, 1978). Calcitecementation follows re-establishment of alkaline pore waters at temperatures in excessof 200°C.The four stages of pore water evolution described above record advection in Bowser Basin sandstones.When compaction and cementation combine to completely occlude pores, advection ceases and the recordof sandstone cement paragenesis is terminated. As long as advection continues, replenishment of porefluids is required. During the earliest stage of pore water evolution, replenishment was probablydominated by relatively unaltered seawater, which is the most likely source of large volumes of Mg2 +and K+ rich alkaline fluid. The later stages of pore water evolution, in contrast, are increasingly affectedby dewatering of interbedded organic-rich mudstones. Mixing of waters derived from the mudstoneswith seawater is initially recorded by increasing acidity of pore waters. This increasing acidity leads tothe succession of authigenic clay cements preceding oil migration. By stage three in the pore waterevolution, waters derived from the mudstones clearly dominate advection through the sandstones. Stagefour, however, records a return to more alkaline conditions, and may thus record re-establishment ofexternal recharge by seawater or, perhaps more likely, meteoric waters as the dominant pore fluid. Stagefour pore waters probably moved through the sediments only after dewatering of the mudstones due tocompaction was largely complete.PORE WATER EVOLUTION AND GREYWACKE GENESISA general model for pore water evolution that explains a broad variety of sandstone cementhistories can be developed from the interpreted controls on cement paragenesis in the Bowser Basin.164Varying grain composition, pH of depositional waters, organic matter content, and rate of burial withingeologically reasonable limits permits prediction of pore water evolution paths leading to diversecementation histories. For example, initial acidic meteoric waters might yield an initial cementationstage of kaolinite, rather than chlorite as in the Bowser Basin If located in an intracratonic basin, whichwould typically have more abundant K± bearing species, an initial illite or K-feldspar stage is morelikely. Low primary production or slow sedimentation rate could reduce the amount of buried organicmatter compared to the Bowser Basin. This could eliminate the acid water stage associated with thermalmaturation, and thus eliminate oil migration, chlorite dissolution, and quartz cementation. The differentcementation histories outlined above will result in pseudo-matrix rich sandstones with the widely variableauthigenic clay, quartz, carbonate, and organic matter content that typifies greywackes.ECONOMIC IMPLICATIONS OF PORE FLUID EVOLUTIONThe changes from alkaline to acidic and back to alkaline pore fluid chemistry have severalimportant implications to economic mineral potential in the Bowser Basin. First, hydrocarbon potentialis low, although the acidic pore water stage clearly records that oil and gas generation occurred, andfurthermore, sandstones saturated with bitumen (dead oil) prove that oil pooled in reservoir rocks.However, fluid inclusions in quartz and the subsequent alkaline pore fluid stage both attest to laterformation temperatures greatly in excess of the oil window. Vitrinite reflectance data discussed later inthis chapter confirm that maximum temperatures reached roughly 250°C to 315°C (depending onstratigraphic position), which greatly exceeds the oil death line of roughly 165 °C (Tissot and Welte,1984). Multiple stages of structural deformation followed sediment accumulation in the Bowser Basin(Moffat and Bustin, 1993), and the preservation of saturated bitumen reservoirs may imply that the oilwas thermally degraded prior to compressional deformation. Gas reservoirs that could have survivedpost-oil migration thermal exposure would likely have been breached by later structural deformation.Remaining hydrocarbon potential may be limited to gas from coal, because there is presently little data tosuggest whether adsorbed methane might survive post-generation structural deformation.165Although the high temperatures implied by the later alkaline pore fluid stage make hydrocarbonpotential low, the change from acidic to alkaline conditions at temperatures near 200°C suggestspotential for accumulation of economic metals may be high. Gold is relatively soluble in acidic solutionsand therefore potentially mobilized as either a chloride or sulphide complex. Solubility drops by a factorof as much as 30 times at 200°C when pH is changed from 6.0 to 9.4 (Seward, 1973). Such a variationin pH occurred in the Bowser Basin between the quartz and calcite cementation stages, and thereforemetals mobilized by acid pore fluids may have been precipitated simultaneous with the initiation ofcalcite precipitation. Thus, if gold or other acid soluble metals were present in the sediments at the timeof burial, then the diagenetic pore water evolution of the Bowser Basin appears to be an effective methodof mobilization and concentration.MUDSTONE PORE WATER EVOLUTIONAuthigenic carbonate occurs in resistant yellow-brown weathering concretions or concretionarylayers in mudstones of the northern Bowser Basin. Concretions commonly exceed 10 cm in diameter,and some layers are more than 1 m thick. The concretions are lithified, have homogeneous black to darkgray interiors, but lack concentric layering. When cut and polished, finely detailed sedimentarystructures such as trace fossils and laminations can be observed. Fossil leaves, and less commonly shells,are very well preserved on the outer surfaces of many concretions, but similar preservation is rare in theinterior of the concretions.Authigenic layers with an early diagenetic origin are described in the literature from manysedimentary basins (Raiswell, 1971; Gautier and Claypool, 1984). Like these examples from theliterature, authigenic layers in the northern Bowser Basin are assumed to have a diagenetic origin, andthus may record information relevant to the pore water history of the mudstones. The concretions arefurther assumed to have precipitated from the inside out over time, in a manner similar to that ofauthigenic carbonates described in other elastic basins. Because the concretions formed in the mudstonesduring diagenesis, their record of pore water chemistry should generally reflect the pore water evolution166determined for the sandstones. However, whereas advection controlled pore water circulation in thesands, dewatering compaction probably controlled fluid movement in the muds, and therefore somedifferences in pore water chemistry may also be expected.Authigenic carbonates vary in thickness and abundance with stratigraphic level. In shallowmarine mudstones of the undivided Bowser Lake Group, (the lowest stratigraphic level examined),authigenic layers consist of poorly lithified concretions or bands of concretions typically less than 6 cmthick, and usually separated vertically by more than 10 m of mudstone. Up section, coincident with theappearance of organic rich deltaic facies, authigenic layers are more common and better developed. Welllithified ovoid concretions 10 to 20 cm in diameter occur in layers commonly separated vertically by lessthan 5 m of strata. Continuing up section, the concretion layers become more continuous and moreclosely spaced vertically. Within the upper portion of the Currier Formation, some layers have growntogether into continuous resistant beds up to 10 cm thick. In the overlying McEvoy Formation, somecontinuous authigenic layers exceed 1 m in thickness. Raiswell (1988) suggested that the thickness anddegree of continuity of authigenic carbonates layers is in part controlled by the temporal continuity ofdeposition. Discontinuous sedimentation is needed to give continuous authigenic layers time to develop.If such a relation holds in the Bowser Basin, then the increasing abundance, thickness, and continuity ofauthigenic layers higher in the section suggests that sedimentation became more intermittent,commensurate with changes from shallow marine to deltaic, and finally to fluvio-deltaic depositionalenvironments.Nine concretions from Bowser Basin strata were sampled for mineralogical and geochemicalanalysis. The stratigraphically lowest sample analyzed (concretion sample 22-13) was collected fromdeltaic black shales deposited 110 metres above marine shell beds near the transition from undividedBowser Lake Group strata to the Currier Formation. Six of the nine analyzed concretions were collectedwithin the lower delta plain deposits of the lower Currier Formation, one from higher in the CurrierFormation, and one from the overlying McEvoy Formation. Each of these concretions was in turnsampled three, or rarely four, times from the middle to the rim. This sampling pattern was designed to167detect changes in pore water chemistry that may have occurred during the growth of the concretion. Thesamples were analyzed by XRD to determine mineralogy, and acid dissolution/gas chromatography-massspectrometry to determine C and 0 stable isotope ratios.Carbonate minerals identified in the concretions are siderite, ferroan dolomite (ankerite), andcalcite. Concretions in which two carbonate minerals occur contain siderite and ferroan dolomite.Siderite probably formed prior to ferroan dolomite, based on siderite's greater abundance in the interiorof two of the examined concretions. Mostly ferroan dolomite concretions with only minor sideriteexhibit no trends of changing mineralogical abundance from core to rim. Ferroan dolomite also occursalone in two other analyzed concretions. Calcite occurs in one monomineralic concretion, and may haveformed either late, after the ferroan dolomite, or early before the formation of siderite.Siderite precipitation requires anoxic pore waters depleted in sulfate (Curtis, 1967), andcommonly occurs very early in diagenesis (Gautier and Clatpool, 1984). Siderite concretions analyzed inthis study occur in mudstones associated with thick lower delta-plain coals (Chapter 2). In lower deltaplain environments of the Mississippi River, early siderite concretions form in the upper few metres ofsediment (Coleman, 1982). The occurrence of siderite cores inside ferroan dolomite concretions suggeststhat siderite also may have begun precipitating very early in the northern Bowser Basin mudstones.The concretions have a thick (up to 1 to 2 cm) yellow-brown weathering rind on their surfaces,presumably derived from oxidized Fe out of the ferroan dolomite structure. Ferroan dolomite isprobably the most widespread authigenic cement in the basin, as well as among analyzed samples,because a similar yellow-brown weathering rind occurs in most well developed carbonate layers inoutcrop. Ferroan dolomite, like siderite, is an authigenic mineral characteristic of the methanogenic zoneof early diagenesis. Given the inferred very early diagenesis origin for the siderite concretions, and thefact that ferroan dolomite follows siderite (in some concretions at least), it is concluded that ferroandolomite was probably a later rather than earlier precipitation product of the methanogenic zone.168CARBON AND OXYGEN STABLE ISOTOPESCarbon and oxygen stable isotope ratios were determined for ten of the examined concretions, inan attempt to better understand the source of the authigenic carbonate and the temperatures at which itformed. Carbon isotope ratios in authigenic carbonates primarily reflect the source of the carbon.During the earliest stages of diagenesis in organic rich muds, isotopically light carbon is released bybacterial degradation of organic matter, because bacteria preferentially consume the lighter carbonisotopes first. As a result, very early authigenic carbonates tend to have very light d 13C values.Carbonates formed in the aerobic and sulfate reducing zones of North Sea sandstones, for example, haved13C values ranging from -35 to -20%0 (Kantorowicz, et al. , 1987). As diagenesis progresses, bacteriaare forced to consume progressively heavier carbon, and authigenic carbonates become progressivelyheavier as a result (Gautier and Claypool, 1985).Carbon isotope values in authigenic carbonates from northern Bowser Basin mudstones arerestricted to a relatively narrow set of values, with most d 13C values occurring between +0.8 and+6.0%o PDB. Such values are considerably heavier than most very early formed authigenic carbonatesfrom mudstones, and are typical of later stages of methanogenesis. The C isotope ratios, therefore,support a later rather than earlier origin for the concretions within the methanogenic stage of earlydiagenesis. The d13C ratios also suggest authigenic carbonate precipitation in the muds occurredconcurrently with precipitation of authigenic clays in the sandstones, because authigenic clays were beingprecipitated in sandstones during early diagenesis (as described in the sandstone paragenetic cementsection of this chapter).Kantorowicz et al. (1987) concluded that ferroan carbonate cements with the positive d 13C ratioscharacteristic of late methanogenesis, correlate well with later hydrocarbon generation. Clearly oil wasgenerated in northern Bowser Basin sediments, because dead oil remains in some sandstones (as discussedearlier). Therefore, organic matter capable of generating oil must have remained in the muds afterbacterial activity ceased. Kantorowicz et al.'s (1987) association of positive d 13C ratios with oil169generation is, therefore, supported in the rocks of the northern Bowser Basin.Oxygen isotope ratios were also determined for the mudstone concretions from the northernBowser Basin. c/180 values in authigenic carbonates are a function of temperature and the (0 80 ratio ofthe pore fluid at the time of precipitation. Additional factors that affect resultant /0 80 values include themineralogy of the carbonate, and the concentrations of other dissolved ions that may occur in the porefluid. The fractionation effects attributable to different minerals encountered in this study can beaccounted for by use of experimentally derived fractionation equations in carbonate species-water systems(calcite: Friedman and O'Neil, 1977; siderite: Carothers et al., 1988; ferroan dolomite: Fisher and Land,1986; Longstaffe, 1989). However, little data is available concerning fractionation effects in complexsolutions such as might be expected in actual pore waters. The effect of other ions in solution onfractionation in reactions, therefore, remains unknown, and interpretations made in this chapter regarding0 isotope data are made with the caveat that such effects are not accounted for. Even with such a caveat,the d180 data are inherently ambiguous due to their dependence on both formation temperature and thec/180 composition of the pore waters. If neither temperature of formation or pore water di 80composition can be tightly constrained, as is the case in the Bowser Basin, then interpretations mustremain somewhat speculative. There are three different models of plausible pore water compositions,and some important inferences concerning Bowser Basin pore water evolution can be reached. The threemodels are described below:1) The -10 %o SMOW meteoric model assumes a meteoric origin for pore fluids in the muds at the timeof concretion precipitation. This model further assumes that the meteoric water was isotopically similarto modern meteoric water which precipitates in the cool humid temperate climate of western BritishColumbia and southern Alaska (-16 %o to -10%o SMOW; Yurtsever, 1975; Gat, 1980). Such values arein accord with climatic and paleogeographic reconstructions that point to a cool humid temperate climateon the western margin of North America for Bowser Basin sediments (Chapters 2 and 3). Notably thismeteoric model is inconsistent with a low latitude position for the Bowser Basin during or shortly afterdeposition, as suggested from some paleomagnetic and paleontologic studies. Such a low latitude origin170would imply meteoric d180 ratios of -4 %o or higher, and would be more closely consistent with the 0960seawater model described later.The -10 %o SMOW meteoric model is geologically plausible for the northern Bowser Basinbecause of the paleogeographic considerations alluded to earlier, and because of the abundance of coaland plant remains preserved in the sediments. -10 %0 is at the high end of likely meteoric rechargevalues, and yields a maximum set of temperatures for plausible meteoric waters. Such a maximummeteoric temperature is convenient for comparison with the other models, and also takes into account thetendency of pore waters to become isotopically heavier after burial (Hanor, 1987). -10 %o SMOW is alsoa reasonable model for initial brackish waters in the muds, for example if the meteoric water mixed withd180 ratio of -16% SMOW mixed with approximately 1/3 seawater by volume.This meteoric pore water model contrasts with the seawater origin inferred for the sandstonesbased on the paragenetic cement succession described earlier. This contrast may reflect the dominance ofdiffusion processes during burial and dewatering of the muds, at the same time that advection rechargedby seawater dominates diagenetic processes in the sandstones.2) The 0%0 seawater model assumes an unevolved seawater origin for the pore fluids. This model isgeologically plausible if seawater (or tropical rain water) dominates pore fluids in the muds, and if thisseawater is largely unaltered by dissolution of detrital minerals or breakdown of plant material within thesediments. Given the relatively high temperatures this model points to, as discussed below, anassumption that detrital dissolution is unimportant and that plant derived carbon is not contributed insignificant amounts is difficult to support geologically.3) The +4%o evolved seawater model assumes a seawater origin for the pore fluids, and accounts for thegeologically likely condition of the pore fluid becoming isotopically heavier after burial. The magnitudeof the shift to positive d 180 values is uncertain, but most mafic and ultramafic rocks such as form at leastpart of the provenance for Bowser Basin sediments (Chapters 3 and 4) have d180 values between +5 and+7960 (Brownlow, 1979). +4%0 d 180 for the pore waters, therefore, makes a reasonable model that171accounts for dissolution by such detrital grains. Given the chemically unstable nature of most detritalmaterial derived from mafic and ultramafic terranes, and the relatively high temperatures implied by thismodel (as discussed below), this +4 %o SMOW evolved seawater model is probably more plausiblegeologically than the 0%0 SMOW seawater model described above for pore fluids that were originallyderived from seawater.The three pore fluid d 180 models outlined above yield three distinct temperature ranges forequilibrium precipitation of each of the different authigenic carbonate mineralogies. For sideritecements, which range in d 180 values from + 11.5 %o to + 15.0 %o SMOW, the indicated temperatureranges are approximately 60° to 80°C assuming the -10%0 SMOW meteoric model; 140° to 180°C forthe 0%0 SMOW seawater model; and 190° to 260°C for the +4%0 SMOW evolved seawater model(Fig. 5-11). The -10 %o SMOW meteoric model is the only model to yield geologically reasonableprecipitation temperatures, given the presumed methanogenic zone origin of the siderite cements, if weassume that the d 180 values reflect original values and not open system re-equilibration values acquiredsome time well after concretion precipitation. The high temperatures yielded by both the 0%0 and +4%o SMOW seawater models are only plausible assuming re-equilibration at high temperatures.However, due to the high temperatures apparently reached in the Bowser Basin following deposition(Bustin, 1984; and discussion later in this chapter), neither seawater model can be eliminated fromconsideration based on the siderite d180 values alone.Ferroan dolomite cement d 180 values indicate higher temperature ranges for each model than thetemperatures deduced from the siderite d 180 values. d180 values for most of the concretions, whichwere collected in deltaic strata of the lower Currier Formation, range from 9 %o to 13 %o SMOW. Thecore of concretion sample 22-13, collected from the top of the undivided Bowser Lake Group at thetransition from marine to deltaic strata, yielded higher d 180 ratios of 16.1 %o. The 9 to 13 %o . d180range corresponds to a temperature range of approximately 75° to 110°C for the -10%o SMOW meteoricmodel; 190° to 280°C for the 0%o SMOW seawater model; and 290° to in excess of 400°C for the +4evolved seawater model (fig 5-12). As was the case for the siderite d 180 values, the only model to172Figure 5-11: Range of concretion d180 ratios in siderite cement and their equilibrium formationtemperature assuming three different pore water isotope composition models. The three models aredescribed in text. D8000^ 081730000Oa0cnR E0072UOC0c330OO8^8^8Figure 5-12: Range of concretion d 180 ratios in ferroan dolomite cement and their equilibrium formationtemperature assuming three different pore water isotope composition models. The three models aredescribed in text.174yield geologically reasonable formation temperatures is the -10%o SMOW meteoric model. The +1696oSMOW value for the core of sample 22-13 is interpreted using this model as either indicating concretionformation initiated at a somewhat lower temperature of about 50°C from similar -10 %o SMOW porewater or that some seawater mixed with meteoric water (brackish water), raising the d 180 of the porefluid. Pore water c/180 of -7 %o would raise predicted equilibrium precipitation temperature to 70°C.Because this sample was collected just above shell-bearing marine beds, the seawater mixing explanationappears most likely.The 0%0 SMOW seawater model is tenable for ferroan dolomite cements only if re-equilibrationoccurred during deepest burial (maximum temperatures generally believed to be less than approximately300°C; Bustin, 1984; and discussion later this chapter). The +4%0 SMOW evolved seawater model isbarely tenable even assuming re-equilibration at the highest postulated formation temperatures. Unlikewith the -10 %o SMOW meteoric model, the +16 %o SMOW core value for sample 22-13 is not readilyexplicable under either of these models.The higher temperatures indicated for ferroan dolomite compared to siderite cements areconsistent with the -10 %o SMOW meteoric model, but are largely uninterpretable under the othermodels. Assuming that temperatures predicted by the -10%o SMOW meteoric model are correct, sideriteprecipitation occurred at temperatures between 60° and 80°C, and ferroan dolomite precipitationfollowed at temperatures of 75° to 110°C (Fig. 5-13). These temperature ranges are at least partiallywithin the methanogenic zone, and the change from siderite to ferroan dolomite follows the trend ofrelatively more siderite in the cores than near the rim indicated by two of the concretions.The much higher formation temperatures for ferroan dolomite compared to siderite precipitationderived from the two seawater models is not readily explicable based on any obvious geologic process.Under these models, siderite would have re-equilibrated over a broad range of temperatures175200 °Cv.:11Vittetle,(f.*" A-Tve•R•43.•%t300°C —100°C•-•+4%.SMOW evolved seawater mode•-10%.SMOW meteoric modelFigure 5-13: Comparison of predicted temperatures for siderite, ferroan dolomite and calcite concretioncements, assuming different modelled pore water compositions.Siderite Calcite FerroanDolomite176from 10° to perhaps 100°C below the even broader range of ferroan dolomite re-equilibrationtemperatures (Fig. 5-13). There is no obvious geologic reason for siderite to re-equilibrate atsignificantly lower temperatures than the ferroan dolomite, especially where siderite occurs in cores offerroan dolomite concretions. The entirely ferroan dolomite sample 22-13 is also inexplicable underthese models. Assuming the 0 %o SMOW seawater model, the ferroan dolomite core of sample 22-13must have re-equilibrated at approximately 150°C, and the rim (d 180=12.9 %o SMOW) atapproximately 195°C.One calcite concretion was analyzed for 0 isotopes. This concretion was collected from theMcEvoy Formation, approximately 500 to 1000 m stratigraphically above the other discussed samples.Assuming the parameters of the -10%o SMOW meteoric model, precipitation occurred between 75° and80°C (Fig. 5-14). As with siderite and ferroan dolomite, the other models yield significantly higherpore fluid temperatures (190° to 200°C for 0%0 SMOW seawater, and 280° to 300°C for the +4%0SMOW evolved seawater model). Given the lack of other analyzed concretions with either similarmineralogy or stratigraphic origin, more specific discussion of this sample is unwarranted, other than tonote that re-equilibration is required to explain any but the -10%0 SMOW meteoric model for calcite aswell as siderite or ferroan dolomite.The -10 %o SMOW meteoric model is the only one of the three models that yields geologicallyreasonable precipitation temperatures. Meteoric water from a cool temperate climate, is therefore,required to explain the origin of the analyzed concretions, if the d 180 values approximate values at thetime of precipitation. Both of the seawater models require re-equilibration of the authigenic carbonates atnear maximum burial temperatures to explain observed d 180 values. Because of the high temperaturesthe Bowser Basin strata were exposed to after burial, re-equilibration cannot be excluded as a possibility.However, re-equilibration fails to explain the observed core to rim variations and the lower sideritecompared to ferroan dolomite temperatures, and therefore is a very unsatisfactory explanation. Thesuperior ability of the -10 %o SMOW meteoric model to explain all the observed d180 data indicates that177I^I^I^1^1^I^I^I^I^r^I^I^0 o 00 0 00Cao8Figure 5-14: Range of concretion c/ 180 ratios from calcite cement and their equilibrium formationtemperature assuming three different pore water isotope composition models. The three models aredescribed in text.178cool temperate meteoric waters controlled mudstone pore fluid compositions in the Bowser Basin duringearly diagenesis. Meteoric water in muds is also consistent with the coal and flora found in associationwith the mudstones. In fact, the -10%o SMOW meteoric model favoured in this discussion may assumestarting pore fluid d180 values that are too high. Meteoric waters as low as -16 %o SMOW, which yieldinitial ferroan dolomite and siderite precipitation at temperatures of 20° to 30°C are consistent withobservations of very early initiation of concretion growth reported from some modern deltas (Tye andColeman, 1989).COMPARISON OF EARLY DIAGENESIS IN SANDSTONES AND MUDSTONESAuthigenic cements precipitated during early diagenesis differ between the mudstones andsandstones of the northern Bowser Basin. For example, authigenic carbonates precipitated as concretionsin mudstones, and authigenic clays precipitated in sandstones. Although the differences in cementmineralogy between mudstones and sandstones are striking, some similarities in the pore waters aresuggested. Mg and Fe are important elemental components both of the chlorite cements in thesandstones, and of the ferroan dolomite cements in the mudstones. In addition, relatively alkaline watersare required for either chlorite or carbonate cement stability. However, some chemical conditions mustbe different to cause chlorite to precipitate in the sandstone at nearly the same time that ferroan dolomiteis precipitating in the mudstone. One possible factor is that sulfate may inhibit dolomite formation(Baker and Kastner, 1981), although that point is questioned by Hardie (1987). Perhaps sulfateconcentrations were higher in the sandstones, in which seawater (an excellent source of sulfate) was beingreplenished by advection. Alternatively (or in addition to), HCO3 - (and CO32-) concentrations wereprobably higher in the mudstones, because this is where the organic matter was being consumed bybacteria. Higher HCO3 - concentration clearly favors carbonate precipitation.The difference between pore fluids in the sandstones and mudstones is further emphasized by thed180 ratios in the mudstone concretions, which strongly favor meteoric (and in at least one samplebrackish) water as the original pore fluid in the muds. This striking difference in pore water for the179muds and sands during early diagenesis probably reflects the fundamental difference in groundwatertransport processes that prevail in the different lithologies. Diffusion and dewatering dominate in mudsand advection, possibly recharged by seawater, dominates in sands. Hence cementation in the mudsappears to record a unidirectional expulsion of meteoric waters whose chemistry is largely controlled bybreakdown of organic matter, while cementation in the sands records mixing of meteoric waters drivenout of the muds with connate saline waters. Similar replacement of marine or brackish waters bymeteoric waters in a marginal marine succession is recorded by strongly negative c/ 180 (-8.3 %o to -18.2%0) values in early formed carbonate concretions in sandstones from Scotland (Wilkinson, 1993).ORGANIC DIAGENESISVitrinite reflectance (random in oil: Rormd) was measured from coals and carbonaceous mudstones fromeach stratigraphic unit within the study area to provide a maximum temperature for diagenesis. The dataset shows that the oldest strata have the highest reflectance values, and that values become markedlylower in progressively younger strata (Fig. 5-15). The lowermost strata, consisting of the marineundivided Bowser Lake Group, yield vitrinite reflectance values in excess of Ro d= 4.1%, that isequivalent to a coal rank of anthracite and meta-anthracite (Bustin, et al., 1990). The Currier Formationyields Rormd values between 2.8% and 4.1%. The thickest and most economically interesting coals inthe Currier Formation occur in the lower part of the unit, and have Ro rmd values between 3.1% and4.1 %, that corresponds to anthracite rank. McEvoy Formation Ro rmd values span the range between1.9 and 2.9%, equivalent to semianthracite coal rank (Bustin, et al., 1990). Like the Currier Formation,the McEvoy Formation tends to exhibit lower Ro rmd values upward in the formation. The Devils ClawFormation yields Ro rmd values between 1.1% and 1.8%, equivalent to medium and low volatilebituminous coal.BASIN MATURATTON MODFLLINGA thermal maturation model that employs Arrhenius type equations to represent organicmaturation was to used to determine maximum temperatures in the basin. This model is a forward180AAAA500 m.14, AA A -H-Figure 5-15: Vitrinite reflectance increases significantly with stratigraphic depth, implying a steepthermal gradient.181%R 01.0MT TERRAZMOSQUE MAAA• AAAAA AACURRIER^MCEVOYFORMATION FORMATIONDEVILSCLAW FMburial depth AAA^ %R 0^A♦* A+rand11 %R0A^max-4+2.0-3.0-4.0-5.0-6.0-modelling type that requires input based on assumptions of paleogeothermal gradient and burial history(Chonchawalit, 1993). Output is in the form of predicted vitrinite Ro d values. When output valuesapproximate the measured values, then the assumptions used are concluded to be a plausible model ofpaleogeothermal gradient and burial history.PARAMETERS OF THE FAVOURED PALEOGEOTHERMAL GRADIENT AND BURIAL HISTORYMODELThe paleogeothermal gradient and burial history model used in this analysis is constrained bygeologic parameters derived from a combination of stratigraphic (Cookenboo, 1989 and Appendices 1, 2,and 3), lithofacies (Chapter 2), paleontologic (Appendix 1), provenance (Chapters 3 and 4), anddiagenetic data (this chapter). Stratigraphic data, as summarized in chapter 1 (based on Cookenboo andBustin, 1989; Cookenboo, 1989; and Appendixes 1, 2, and 3) yields the following estimated thicknessesfor units in the study area: 1000 m for the Currier Formation; 1000 m for the McEvoy Formation; and600 m for the Devils Claw Formation. Lithofacies demonstrate progressive shallowing of the marinebasin fill to near sea-level by the end of undivided Bowser Lake Group deposition, and subsequentaccumulation of Currier, McEvoy, and Devils Claw formations near sea-level.Paleontologic data described in Appendix 1 are the basis for ages used in this model: CurrierFormation was deposited between the Oxfordian (Late Jurassic) and the Earliest Cretaceous (Berriasian);the McEvoy Formation was deposited during the latest Barremian or Aptian to the mid to late Albian(mid-Cretaceous); and the Devils Claw Formation in latest Albian or earliest Cenomanian (LateCretaceous) time (Bustin and Moffat, 1983; Moffat et al., 1988; MacLeod and Hills, 1990; Cookenbooet al., 1991). Numerical ages for the stages are taken from (Palmer, 1983).Provenance analysis (Chapters 3 and 4) constrains the basin to some type of oceanic settingbecause the basin was filled by detritus derived from island arc volcanics and marginal basin lithosphere,and because that detritus was derived from the east (to northeast). In other words, the Bowser Basin wasoutboard of obducted oceanic terranes, and therefore also likely accumulated in an oceanic terrane. The182paleogeothermal gradient favoured for this setting is 65°C/km, similar to some western Pacific marginalbasins(e.g. the central Sumatra basin with a gradient of 65° to 90°C/km; North, 1985), and higher thanmost continenetal values. Simulations using lower geothermal gradients are also reported here as a testof the validity of the model.The pore water evolution determined from sandstone cement paragenesis and concretiongeochemistry is consistent with progressive burial leading to steadily increasing formation temperaturesfrom burial to over 200°C, which provides the final constraint for the thermal maturation model. Thisconstraint is that the thermal model should use a linear, constant geothermal gradient. In other words,there is no need to postulate a separate thermal event sometime late in basin evolution.The depth to which the Devils Claw Formation was buried is one further geologic parameter thatmust be considered. This depth is difficult to estimate, because strata that buried the Devils ClawFormation are now eroded entirely away. The depositional thickness of these eroded strata has been thesubject of much speculation (Bustin, 1984, Bustin and Moffat, 1989), because the thickness controls themaximum depth to which Bowser Basin sediments were buried. Bustin (1984) used approximately 1500m of post-Devils Claw burial by strata equivalent to the Sustut Group that are presently exposed east ofthe basin (Eisbacher, 1981), and high geothermal gradients (in excess of 50°C/km) to account for thehigh rank coals in the Groundhog Coalfield. Bustin and Moffat (1989) demonstrated that an additional2000 m (beyond the 1500 m accounted for by Sustut Group equivalents) of burial were required toachieve anthracite rank coals in the Currier Formation at more nearly typical continental geothermalgradients of 30 to 40°C/km. Fifteen hundred metres of post Devils Claw burial is favoured in thethermal model in this study, because: 1) 1500 m is compatible with known thickness of Sustut Groupsediments deposited prior to Late Cretaceous erosion of the Bowser Basin (Eisbacher, 1974); and 2) 1500m is compatible with preservation of significant intergranular porosity in upper McEvoy Formationsandstones until after late stage calcite cementation (Galloway, 1974; Trevena and Clark, 1986).183Due to the uncertainties in actual burial depth, models assuming deeper burial were also run,and are described later.RESULTS OF BASIN MATURATION MODE! 1,INGThe best fit of the calculated values to measured vitrinite reflectance values from the northernBowser Basin is the model run using 65°C/km as the geothermal gradient and assuming 1500 m of burialof the Devils Claw Formation (see burial history curve Fig. 5-16). Selected model runs using 65°C/kmand lower geothermal gradients and yielding output in the form of calculated vitrinite reflectance valuesare illustrated in figure 5-17. Calculated maximum temperatures for the best fit run (65°C/km) are215°C for basal McEvoy Formation and deeper stratigraphic horizons. Temperatures in excess of200°C were inferred from the pore water evolution described earlier, and therefore this model generallyagrees with sandstone paragenesis observations. The maximum temperature reached by the upperMcEvoy or lower Devils Claw Formation horizon is 155°C. Lower formational temperatures thanpredicted by the sandstone cements can be explained as the result of vertical migration of pore fluidsfrom greater depths. Geothermal gradients less than 60° to 65°C/km are insufficient to explainmeasured vitrinite reflectance values assuming 1500 m of post Devils Claw Formation burial. Forexample, the 50°C/km run predicts semianthracite (Ro d = 2.4%) in the lower Currier coal measures(Fig. 5-17), in contrast to actual ranks of anthracite and meta-anthracite. Notably, the 65°C/kmgeothermal gradient run matches observed vitrinite reflectance values in the upper stratigraphic horizonsas well as in the lower units. This matching at all stratigraphic levels is lacking in runs described below.Maturation model runs using a geothermal gradient of 40°C/km, but assuming 5000 m burial ofthe Devils Claw Formation (instead of 1500 m used above) results in anthracite in lower Currier andbelow, which is compatible with measurements. However, the 40°C/km and 5000 m burial run alsopredicts anthracite occurs in the upper McEvoy/basal Devils Claw, which is incorrect. 3000 m burialand 40 ° Cfkm yields compatible reflectance values in the upper McEvoy/basal Devils Claw Formation,1840Figure 5-16: Burial history curve, assuming ages, stratigraphic thickness, and burial depths of favouredmaturation model.TIME185145 Ma 130 Ma^ 1000 ^2000115 Ma^95 Ma 85 Ma^55 Ma^ 10 Ma 0STRAT1GRAPHIC MILLIONS OF YEARS BEFORE PRESENTHORIZONS PREDICTEDCurrier Formation(lower)145 MYA^130^115Om^+1000m^+0m^1030m^1000m95+1800m260085+1500m4100m55-1500m2800m10-1000m1030m0 MYAPresentTopographyAGE (MYA) %R °rand65°/km+ 65°C^+0°C10°C^75°C^75°C+105°Clao d+100°C280°C-100°C180°C-105°C75°C-65°C10°CTemp. changeInterval temp. 4.0%+50°C^+0°C +80C +75°C -75°C -80°C -50°C Temp. change50o/km10°C^60°C^60°C 140°C 21 5P c 140°C 60°C 10°C Interval temp. 2.4%+40°C^+0°C10°C^50°C^50°C4043/1<rn+65°C105°C+60°C165°C-60°C105°C-65°C50°C-40°C10°CTemp. changeInterval temp. 1.3%-.4 MILLIONS OF YEARS BEFORE PRESENT115 105^95 85 55 10 0 MYA AGE (MYA)Base of McEvoy Orn +500m^+1100m +1500m -1500m -500m PresentFormation 500m^1600m 3100m 1600m 500m Topography6e/km^10°C+35°C^+70C45°C^115°C+100°C215°C-100°C115°C-70°C45°C-35°C10°CTemp. changeInterval temp. 2.4%50°/km^10°C+25°C^+55°C35°C 96 c+75°C16.5°C-75°C90°C-55°C35°C-25°C10°CTemp. changeInterval temp. 1.9%+20°C^+45°C +60°C -60°C -4st -20°C Temp. change4Cf/kM^10°C 30°C^75°C 135°C 75°C 30°C 10°C Interval temp. 0.8%MILLIONS OF YEARS BEFORE PRESENT-.4100^95 85 55 10 0 MYA AGE (MYA)Top of McEvoy/ _ +700m +1500m -1500m -500m presentBase of Devils Claw""^703m 2200m 700m 200rn Topography65°/km+45°C10C^55°C+100°C155°C-100°C55°C-5°C50°C-40°C10°CTemp. changeInterval temp. 1.1%507k m+35°C10 C^45°C+75°C120°C-75°C45°C-5°C40°C-30°CtotTemp. changeInterval temp. 0.7%4e/km+30°C10C^40°C+60°C1 cec-60°C40°C-5°C35°C-25°C1ecTemp. changeInterval temp. 0.6%but predicts the deepest modelled horizon (500 m below the Currier Formation) does not even reachsemianthracite rank.The results of the computer maturation simulation are resistant to changes in timing withingeologically reasonable bounds. Changes in timing of burial and uplift have negligible influence on themodel output, because Road is related to the highest temperatures reached by a stratigraphic unit andtherefore termination of burial effectively stops Ro d increase. Runs using alternative age assignmentssimilar to those suggested by leaf fossil studies (MacLeod and Hills, 1990) also do not differsignificantly in final Rorand.IMPLICATIONS FOR DL4GENETIC HISTORYThe thermal maturation model yields geothermal gradients and maximum temperaturesconsistent with the proposed pore water evolution. Most importantly, maximum temperatures deducedfrom the maturation model exceed the minimum temperatures inferred for the late stage of calcitecementation (in excess of 200°C). The compatibility of the diagenetic history with the maturation modeldoes not exclude other possible explanations, but it does permit confidence that the preferred models ofboth maturation and diagenetic history are geologically reasonable histories for the northern BowserBasin.IMPLICATIONS OF MATURATION MODEL TO REGIONAL TECTONICSThe maturation model used above implies that the geothermal gradient must be higher thannormal continental crustal averages to result in the high maturation gradient that occurs in the northernBowser Basin. Low gradients will not produce anthracite, except with very deep burial, and if burial isthat deep, low gradients would cause the entire preserved stratigraphic column to be anthracite. Thecompatibility of a high geothermal gradient with both the maturation model and previously determinedindependent geologic data strongly supports accepting the favoured basin development model.187A high geothermal gradient is consistent with the inferred accumulation of the Bowser Basinsediments in an oceanic or back-arc basin (Chapter 4) west of the North American continental margin.Similar basins in the southwest Pacific typically have high heat flows. The proposed high heat flow alsofits well with other basins in Cordillera. Maturity gradients (and coal ranks) for Mesozoic strata in theWestern Canada Sedimentary Basin (WCSB) are lower than in comparable aged strata of the BowserBasin, but increase westward towards the mountains (Bustin, 1991). This trend of increasing geothermalgradient continues further west to the Bowser Basin. Heat flow also increases westward, roughlydoubling from 40 mW/m2 in the Foreland Belt to approximately 80 mW/m2 in the Intermontane Belt(Sweeney, et al., 1992). Increasing geothermal gradient and heat flow are probably due to variation intype of basement, because Paleozoic and older North American continental crust thins from the cratontowards the deformed belt, and probably changes to oceanic crust farther west (Thompson, 1989).Sediment accumulation in the Bowser Basin occurred above this oceanic crust outboard of NorthAmerica. The Bowser Basin sediments were therefore subjected to high geothermal gradients followingdeposition.CONCLUSIONS:1) Pore water evolution of the northern Bowser Basin is recorded in the paragenetic sequence ofsandstone cements. That sequence of cements records the following diagenetic stages: i) early isopachouschlorite cement; ii) illite cementation;^kaolinite cementation; iv) oil generation and migration; v)chlorite dissolution; vi) quartz cementation; and vii) late stage calcite pore-filling and grain replacementcement.2) Pore water chemical conditions inferred from the sandstone cement paragenesis suggest that chloriteprecipitated from an alkaline fluid such as seawater. The initial seawater probably was enriched in Feand Mg due to dissolution of mafic minerals.3) The progression of sandstone cement stages was primarily controlled by changes in pore fluid pH, andsecondarily by rock-water interactions. Acidity increased in pore waters at the time of oil migration as aresult of thermal maturation of organic matter. lllite, kaolinite, and quartz cements precipitated from188these increasingly acidic pore waters. Precipitation of illite removed K + from solution (rock-waterinteraction), favoring the change from the illite to kaolinite cement stage. Acid water conditions alsocaused chlorite dissolution. Calcite cementation began after the thermal breakdown of organic acids ledto reduced acidity in the pore waters, probably at temperatures in excess of 200°C.4) Pore fluids in the mudstones were strongly depleted relative to seawater in d 180 duringmethanogenesis as recorded in authigenic carbonate concretions. The highly negative c/ 180 valuessuggest that meteoric waters (i.e., fresh or brackish) from cool temperate climates were the primarysource of the pore waters in the mudstones.5) Maturation modelling assuming heating of the sediments due to deposition and progressive burial in ahigh geothermal gradient regime such overlying back-arc crust accounts for measured vitrinite reflectancedata. There is no need for later thermal events associated with structural deformation or tectonic eventsto explain the diagenetic history of the northern Bowser Basin, although such thermal events cannot beexcluded.189REFERENCESAlmon, W. R., and Davies, D. K. 1981. Formation damage and the crystal chemistry of clays. In F. J.Longstaffe (ed.), Clays and the resource geologist; Mineralogical Association of Canada ShortCourse Handbook, v. 7, p. 81-103.Baker, P. A., and Kastner, M., 1981. Constraints on the formation of sedimentary dolomite. Science, v.213, p. 214-216.Bodnar, R. J., Reynolds, T. J., and Kuehn, C. A., 1985. Fluid-inclusion systematics in epithermalsystems. In Berger, B. R., and Bethke, P. M., (eds.) Geology and geochemistry of epithermalsystems. Reviews in Economic Geology v. 2, p. 73-97.Brownlow, A. H., 1979. Geochemistry. Prentice-Hall, Inc. Englewood Cliffs, New Jersey. 498 p..Bustin, R. M., 1984. Coalification levels and their significance in the Groundhog Coal field, north-central British Columbia. International Journal of Coal Geology, v. 4, p. 21-44.Bustin, R. M. 1991. Organic maturity in the western Canada sedimentary basin. International Journal ofCoal Geology, v. 19, p. 319-358..Bustin, R. M., and Moffat, I. 1983. Groundhog Coalfield, British Columbia: reconnaissance stratigraphyand structure. Bulletin of Canadian Petroleum Geology, 31: 231-245.Bustin, R. M., and Moffat, I, W., 1989. Semianthracite, anthracite and meta-anthracite in the centralCanadian Cordillera: their geology, characteristics and coalification history. InternationalJournal of Coal Geology, v. 13, p. 303-326.Bustin, R. M., Barnes, M. A., and Barnes, W. C., 1990. Determining levels of organic diagenesis insediments and fossils fuels. In McIlreath, I. A. and Morrow, D. W. (eds.) Diagenesis.Geoscience Canada Reprint Series 4, p. 205-226.Carothers, W. W., Adami, L. H., and Rosenbauer, R. J., 1988. Experimental oxygen isotopefractionation between siderite-water and phosphoric acid liberated CO2. Geochimica andCosmochimica Acta, v. 52, p. 2445-2450.Carothers, W. W., and Kharaka, Y. K., 1978. Aliphatic acid anions in oilfield waters-implications forthe origin of natural gas. American Association of Petroleum Geologists Bulletin, v. 62, p.2431-2441.190Chonchawalit, A., 1993. Basin analysis of Tertiary strata in the Pattani basin, Gulf of Thailand. Ph.D.thesis, The University of British Columbia, 366 p.Cookenboo, H. 0., 1989. Lithostratigraphy, palynostratigraphy, and sedimentology of the northernSkeena Mountains and their implications to the tectonic history of the Canadian Cordillera.MSc. thesis, University of British Columbia, Vancouver British Columbia. 131 p.Cookenboo, H. 0., and Bustin, R. M., 1989. Jura-Cretaceous (Oxfordian to Cenomanian) stratigraphyof the north-central Bowser Basin, northern British Columbia: Canadian Journal of EarthSciences, v. 26, p. 1001-1012.Eisbacher, G. H., 1974. Sedimentary and tectonic evolution of the Sustut and Sifton Basins, north-central British Columbia. Geological Survey of Canada Paper 73-31, 57 p.Eisbacher, G. H., 1981. Late Mesozoic-Paleogene Bowser Basin Molasse and Cordilleran tectonics,Western Canada. In A D. Miall (ed.), Sedimentation and Tectonics in Alluvial Basins.Geological Association of Canada, Special Paper 23, p. 123-151.Fisher, R. S., and Land, L. S., 1986. Diagenetic history of Eocene Wilcox sandstones, south-centralTexas. Geochimica and Cosmochimica Acta, v. 50, p. 551-561.Foscolos, A. E., 1985. Catagenesis of argillaceous rocks. In Mclireath, I. A. and Morrow, D. W. (eds.)Diagenesis. Geoscience Canada Reprint Series 4, p. 177-188.Friedman, G. M., Amiel, A. J., and Schneidermann, N. A., 1974. Submarine cementation in reefs:example from the Red Sea. v. 44, p. 816-825.Friedman, G. M., Sanders, J. E., and Kopaska-Merkel, D. C., 1992. Principles of sedimentary deposits.Maxwell Macmillan Company, Toronto. 717 p.Friedman, I., and O'Neil, J. R., 1977. Compilation of stable isotope fractionation factors of geochemicalinterest. In . Fleischer, M. (ed.), Data of Geochemistry, sixth edition. United States GeologicalSurvey Professional Paper 440-kk, 12 p.Galloway, W. E., 1974. Deposition and diagenetic alteration of sandstone in northeast Pacific arc-relatedbasins: implications for graywacke genesis. Geological Society of America Bulletin, v. 85, p.379-390.Gat, J. R. 1980. The isotopes of hydrogen and oxygen precipitation. In Handbook of EnvironmentalIsotope Geochemistry Fritz, P. and Fontes, J. Ch., (eds), v. 1, p 21-48.Gautier, D. L., and Claypool, G. E., 1984. Interpretation of methanic diagenesis in ancient sediments byanalogy with processes in modern diagenetic environments. In McDonald, D. A., and Surdam,191R. C., (eds.) Clastic Diagenesis. American Association of Petroleum Geologists Memoir 37, p.111-123.Hanor, J. S., 1980. Dissolved methane in sedimentary brines: potential effect on the PVT properties offluid inclusions: Economic Geology, v. 75, p. 603-609.Hanor, J. S., 1987. Origin and migration of subsurface brines. Society of Economic Paleontologists andMineralogists. Short Course Notes v. 21, 247 p.Hardie, L. A., 1987. Dolomitization: a critical view of some current views. Journal of SedimentaryPetrology, v. 57, p. 166-168.Hesse, R. 1990. Silica diagenesis: origin of inorganic and organic replacement cherts. In McIlreath, I. A.and Morrow, D. W. (eds.) Diagenesis. Geoscience Canada Reprint Series 4, p. 253-275.Hunt, J. M., 1979. Petroleum Geochemistry and Geology. W. H. Freeman and Company, SanFrancisco. 617 p.Jahren, J. S. and Aagaard, P., 1989. Compositional variations in diagenetic chlorites and illites, andrelationships with formation-water chemistry. Clay Minerals, v. 24, p. 157-170.Longstaffe, F. J., 1989. Stable isotopes as tracers in elastic diagenesis. In: Hutchinson, I. E. (ed.), ShortCourse in Burial Diagenesis, Mineralogical Association of Canada Short Course Series, v. 15,p. 201-277.MacLeod, S. E., and Hills, L. V. 1990. Conformable Late Jurassic (Oxfordian) to Early Cretaceousstrata, northern Bowser Basin, British Columbia: A sedimentological and paleontological model.Canadian Journal of Earth Sciences, v. 27: 988-998.Moffat, I. W., and Bustin, R. M., 1993. Deformational history of the Groundhog Coalfield,northeastern Bowser Basin, British Columbia: styles, superposition and tectonic implications.Bulletin of Canadian Petroleum Geology, v. 41, p. 1-16.Moffat, I. W., Bustin, R. M., and Rouse, G. E. 1988. Biochronology of selected Bowser Basin strata:tectonic significance. Canadian Journal of Earth Sciences, v. 25, p. 1571-1578.North, F. K., 1985. Petroleum Geology. Allen and Unwin, London. 607 p.Odin, G. S. and J. P. Masse, 1988. The verdine facies from the Senegalese continental shelf. In Odin, G.S. (ed.) Green Marine Clays. Developments in sedimentology v. 45. Elsevier, New York, p.83- 106.192Palmer, A. R., 1983. The decade of North american geology 1983 geological time scale. Geology, v.11, p.503-504.Pittman, E. D., 1988. Diagenesis of Terry Sandstone (Upper Cretaceous), Spindle Field, Colorado.Journal of Sedimentary Petrology, v. 58, p. 785-800.Raiswell, R., 1971. The growth of Cambrian and Liassic concretions. Sedimentology, v. 17, p. 147-171Raiswell, R., 1988. Chemical model for the origin of minor limestone-shale cycles by anaerobic methaneoxidation. Geology, v. 16, p. 641-644.Roedder, E., 1974. Composition of Fluid Inclusions. United States Geological Survey Professional Paper44011, 164 p.Roedder, E., 1984. Fluid Inclusions. In Ribbe, P. H., (ed.), Reviews in Mineralogy, MineralogicalSociety of America, v. 12, 644 p.Seyfried, W. E., Bendt, M. E., and Seewald, J. S., 1988. Hydrothermal alteration processes at mid-ocean ridges: constraints from diabase alteration experiments, hot-spring fluids, andcompositions of the oceanic crust. Canadian Mineralogist, v. 26, p. 787-804.Small, J S , Hamilton, D. L., and Haber h, S., 1992. Experimental simulation of clay precipitationwithin reservoir sandstones 1: Techniques and examples. Journal of Sedimentary Petrology, v.62, p. 520-529.Smosna, R., 1988. Low-temperature, low-pressure diagenesis of Cretaceous sandstones, Alaskan NorthSlope. Journal of Sedimentary Petrology, v. 58, p. 644-655.Surdam, R. C. and Crossey, L. J., 1985. Mechanisms of organic/inorganic interactions insandstone/shale sequences. In Gautier, D. L., Kharaka, Y. K., and Surdam, R. C. (eds.)Relationship of organic matter and mineral diagenesis. Society of Economic Paleontologists andMineralogists short course v. 17, p. 177-232.Surdam, R. C., and MacGowan, D. B., 1987. Oilfield waters and sandstone diagenesis. AppliedGeochemistry, v. 2, p. 613-619.Sweeney, J. F., Stephenson, R. A., Currie, R. G., and DeLaurier, J. M., 1992. Tectonic FrameworkPart C. Crustal Geophysiscs. In Gabrielse, H., and Yorath C. J. (eds.), Geology of theCordilleran Orogen in Canada. Geological Society of America's Geology of North America, v.G-2, p. 39-58.193Thomas, J. B., 1981. Classification of clay minerals in tight gas sandstones: case studies in which clayminerals are crucial to drilling fluid selection, formation evaluation, and completion techniques.In F. J. Longstaffe (ed.), Clays and the resource geologist; Mineralogical Association of CanadaShort Course Handbook, v. 7, p. 104-118.Tissot, B. P. and Welte, D. H., 1984. Petroleum Formation and Occurrence. Springer-Verlag, NewYork,. 699 p.Trevena, A. S., and Clark, R. A., 1986. Diagenesis of sandstone reservoirs of Pattani Basin, Gulf ofThailand. American Association of Petroleum Geologists Bulletin, v. 70, p. 299-308.Tye, R. S. and Coleman, J. M., 1989. Depositional processes and stratigraphy of fluvially dominatedlacustrine deltas; Mississippi delta plain. Journal of Sedimentary Petrology, v. 59, p. 973-996.Wilkinson, M. 1993. Concretions of the Valtos Sandstone Formation of Skye: geochemical indicators ofpalaeo-hydrology. Journal of the Geological Society, London, v. 150, p. 57-66.Yurtsever, Y. 1975. Worldwide survey of stable isotopes in precipitation. Rep. Sect. Isotope Hydrology,International Atomic Energy Agency, 40 p.194CHAPTER 6CONCLUDING REMARKSLithofacies, provenance, and diagenetic studies (chapters 2, 3, 4, and 5) support a deep basin fill modelof Bowser Basin development. This model suggests the sediments themselves may have largelycontrolled the tectonic development of the Bowser Basin until after deposition of the youngest preservedstrata (Devils Claw Formation). By this model, Bowser Lake Group sedimentation began in a deepmarine basin, possibly of typical ocPAnic depths (3000 to 4000 m), and sediments built up to near sea-level, forming a shelf by the Late Jurassic. Sediments exposed in the study area accumulated above theshelf and deeper marine basin fill. Accommodation space was created by subsidence primarily due to thecombined action of sediment compaction and isostatic adjustment. The major conclusions from eachchapter that pertain to this basin model are reviewed below, and later discussed in relation to a collapsedmargin model of Cordilleran accretion.Lithofacies successions demonstrate that Bowser Basin sediments accumulated as fill in a pre-existing deep oceanic basin. North of the study area, Bathonian and Callovian (Middle Jurassic) basinaland slope elastics (Ricketts, 1990) overlie Bajocian (Middle Jurassic) and older deep marine elastic andvolcaniclastic equivalents to are volcanic rocks exposed along the Stikine Arch. Similar Lower to MiddleJurassic oceanic arc volcanics and basinal sediments are exposed on the western (Lewis, et al., 1993),southern (Tipper and Richards, 1976) and eastern margins of the Bowser Basin. Within the study area,more than 3000 m of Oxfordian (Upper Jurassic) to Albian (Lower Cretaceous) sediments accumulated,overlying the deep marine sediments exposed to the north. Lithofacies associations described in thisstudy indicate that these sediments accumulated in shelf, deltaic and alluvial coastal plain environments.Notably, all the sediments in the study area are interpreted to be near sea-level deposits.Accumulation of more than 3000 m of shelf, delta, and coastal plain deposits requires creationof near sea-level sediment accommodation space. Because sediments exposed in the study areaaccumulated over deep basinal fill, sediment compaction and isostatic subsidence of the basin must have195contributed to the accommodation space. The amount of subsidence attributable to sediment compactionand isostatic subsidence largely depends on the initial depth of the oceanic basin, which is poorlyconstrained. Calculations were made of sediment compaction and isostatic subsidence assuming variousinitial depths for the Bowser Basin (Chapter 2). These calculations suggest that by filling a pre-existingbasin with initial depths of 3000 to 4000 m (typical of modern open ocean, back-arc and interarc basins;Shupe, 1992), accommodation space sufficient to contain the entire preserved near sea-level sedimentaryrecord exposed in the study would be created by compaction and isostatic adjustment. Initial depths of2000 m result in roughly 2000 m of compaction and isostatic subsidence, enough to accommodate themarine shelf, lower delta plain and one half of the upper delta plain facies associations. Filling a 1000 minitial depth basin only accommodates 850 m of near sea-level deposits, enough to account for the shelfand part of the lower delta plain accumulation. A sediment-controlled model consistent with thesecalculations and the observed lithofacies can, therefore, explain the tectonic development of the BowserBasin until after the youngest preserved strata (Devils Claw Formation) were deposited. By thissediment-controlled model, the entire sediment record preserved in the study area was accommodated bysediment compaction and isostatic adjustment resulting from filling a typical marine basin of 3000 to4000 m initial depth. Alternatively, outside tectonic forces such as thrust loading (c.f. Ricketts et al.,1993) or thermal subsidence must be invoked if initial water depths were less than 2000 m.Sandstone provenance described in chapters 3 and 4 supports an oceanic setting for the BowserBasin. Interpretation of framework grain modal analysis (chapter 3) suggests the provenance wasobducted oceanic lithosphere and (at least mostly) inactive island arc volcanics. Recycled sediment andminor plutonic and metasedimentary indications not observed in Upper Jurassic sandstones appear in theLower Cretaceous sediments, but the original (first cycle) sediment source remained the same island arcand oceanic lithosphere throughout the preserved strata. Microprobe analysis of detrital chromian spinelsin the sandstones (chapter 4) supports an oceanic origin for the strata, and suggests that the obductedoceanic lithosphere originated in a marginal basin setting (suprasubduction zone) rather than at a mid-ocean ridge. Furthermore, erosion of the obducted oceanic strata probably occurred prior to any high196grade (greenschist) metamorphic events affected the source rocks.In summary, the provenance was obducted island arc assemblages and marginal basinlithosphere. Current directions indicate the obducted source rocks were north and northeast of the studyarea (Eisbacher, 1981; Cookenboo, 1989), and therefore between the North American craton and theBowser Basin. The setting of the Bowser Basin was seaward of obducted oceanic terranes.Accumulation of Bowser Basin sediments in a pre-existing basin of typical oceanic depths, as suggestedby the lithofacies analysis in chapter 2, is reasonable given the oceanic provenance. Subsidencecontrolled by compaction/isostatic load as modelled in chapter 2 is compatible, therefore, with both thelithofacies and the provenance studies.Diagenetic studies (chapter 5) are the basis of a model of pore water evolution for the BowserBasin that is consistent with accumulation in a deep ocean basin. Pore fluid chemistry was interpretedfrom sandstone cement paragenesis and mudstone concretion geochemistry as a record of pore waterevolution. Pore waters in the sandstones evolved from alkaline marine or brackish waters at or soon afterburial, to acidic waters associated with organic maturation occurring interbedded mudstones (80° to200°C), and back to alkaline waters after thermal breakdown of organic acids at temperatures above200°C. Anthracite coals in the study area suggests a relatively high paleogeothermal gradient.Maturation modelling described in chapter 5 is compatible with a paleogeothermal gradient of 65°C/km,and maximum temperatures of slightly more than 300°C for the lower part of the exposed strata (500 mbelow the base of the Currier Formation). The progressive increase of temperature from initial burial toin excess of 200°C (as recorded by the pore water evolution), and peaking near 300°C (as recorded byvitrinite reflectance), suggests heating due to progressive burial in a high paleogeothermal gradientregime such as typifies oceanic settings.Progressive burial in a high paleogeothermal gradient regime is compatible with the deepoceanic basin setting for the Bowser Basin postulated from the lithofacies and provenance considerationsdescribed earlier. Other settings with high paleogeothermal gradients, such as some continental rift197basins, could satisfy the maturation model in chapter 5, but lithofacies in such settings tend to changefrom older alluvial deposits to younger marine facies, rather than the deep to marginal marine andalluvial succession in the Bowser Basin. Sandstone compositions in continental rift basins are also likelyto be incompatible with those of the Bowser Basin, because rift basins tend to have at least in part acratonic provenance.The model of Bowser Basin origin and tectonic development postulated earlier in this chapter iscompatible with the multiple lithofacies, provenance and diagenetic conclusions of this thesis. Themodel is also compatible with a collapsed margin model of Cordilleran development. As put forwardhere, such a model for Cordilleran tectonic development suggests that marginal seas and fringing islandarcs to the west of the North American craton accreted to the continental margin after the latestTriassic/earliest Jurassic change in plate motion (Ekstrand and Butler, 1989). Obduction of the islandarcs and parts of the marginal basin lithosphere exposed the Bowser Basin provenance to erosion by theMiddle Jurassic. Such a history of accretion has been described for the southern Canadian Cordillera,with the Slide Mountain and Quesnellia terranes obducting onto the continental margin during the Earlyand Middle Jurassic (Murphy, 1989), and the Cache Creek terrane obducting onto Quesnellia in theMiddle and Late Jurassic (Mortimer et al. , 1990). Following obduction of eastern oceanic terranes,continued North American plate motion towards the Pacific Ocean (as recorded by the continued riftingof the Atlantic Ocean) requires an arc to the west of the Bowser Basin. The sandstone provenance forGravina-Nutzotin sediments of southeast Alaska record such an active island arc west of the BowserBasin during the Late Jurassic and Early Cretaceous (Cohen and Lundberg, 1988). The Coast PlutonicComplex (CPC) may be the magmatic record of the required arc to the west of the Bowser Basin, becauseportions of the CPC were forming above east dipping subduction during the Late Jurassic and EarlyCretaceous (van der Heyden, 1989). Compressional deformation of the Bowser Basin began in the LateCretaceous following deposition of the Devils Claw Formation (Moffat and Bustin, 1993),contemporaneous with major Late Cretaceous magmatic activity in the CPC (van der Heyden, 1989). ByCampanian and Maastrichtian time (latest Cretaceous), the Bowser Basin was apparently uplifted and198shedding sediments to the east into the Sustut basin (Eisbacher, 1981; Bustin and McKenzie, 1989).199REFERENCESBustin, R. M., and McKenzie, K. J., 1989. Stratigraphy and depositional environments of the SustutGroup, southern Sustut Basin, northcentral British Columbia. Bulletin of Canadian PetroleumGeology, v. 31, p. 231-245.Cohen, H. A., and Lundberg, N., 1988. Sandstone petrology of the Seymour Canal formation (Gravina-Nutzotin Belt): Implications for the accretion history of southeast Alaska [abs.,]:. GeologicalSociety of America Abstracts with Programs, v. 20, p. 163.Eisbacher, G. H., 1981. Late Mesozoic-Paleogene Bowser Basin Molasse and Cordilleran tectonics,Western Canada. In A D. Miall (ed.), Sedimentation and Tectonics in Alluvial Basins.Geological Association of Canada, Special Paper 23, p. 123-151.Ekstrand, E. J. and Butler, R. F., 1989, Paleomagnetism of the Moenave Formation: Implications forthe Mesozoic North American apparent polar wander path: Geology, v. 17, p. 245-248.Lewis, P. D., Thompson, J. F. H., Nadaraju, G., R. G. Anderson, and G. G. Johnson, 1993. Lowerand Middle Jurassic stratigraphy in the Treaty Glacier area and geological setting of the TreatyGlacier alteration system, northwestern British Columbia. Geological Survey of Canada CurrentResearch Part A, Paper 93-1A, p. 75-86.Moffat, I. W., and Bustin, R. M., 1993. Deformational history of the Groundhog Coalfield,northeastern Bowser Basin, British Columbia: styles, superposition and tectonic implications.Bulletin of Canadian Petroleum Geology, v. 41, p. 1-16.Mortimer, N., van der Heyden, P., Armstrong, R. L., and Harakal, J., 1990. U-Pb and K-Ar datesrelated to the timing of magmatism and deformation in the Cache Creek terrane and Quesnellia,southern British Columbia. Canadian Journal of Earth Sciences, v. 27, p. 117-123.Murphy, D., C., 1989. Crustal paleo-rheology of the southwestern Canadian Cordillera and its influenceon the kinematics of Jurassic convergence. Journal of Geophysical Research, v. 94, p. 15723-15739.Ricketts, B. D. 1990, A preliminary account of sedimentation in the lower Bowser Lake Group, northernBritish Columbia, in Current Research, Part F, Geological Survey of Canada, Paper 90-1F, p.145-150.Ricketts, B. D., Evenchick, C. A., Anderson, R. G., and Murphy, D. C., 1993. Bowser Basin, northernBritish Columbia: constraints on the timing of initial subsidence and Stikinia-North Americaterrane interactions. Geology, v. 20, p. 1119-1122.200Shupe, J. F., 1992. Worlds oceans floors: Pacific Ocean and Indian Ocean. Graves, W. (ed.), map.National Geographic Society.Tipper, H. W. and Richards, T. A., 1976. Jurassic stratigraphy and history of north-central BritishColumbia. Geological Survey of Canada, Bulletin 270, 73 p.van der Heyden, P., 1989. U-Pb and K-Ar geochronometry of the Coast Plutonic Complex, 53°N to54°N, British Columbia, and implications for the Insular-Intermontane superterrane boundary.Ph.D. thesis, the University of British Columbia, Vancouver British Columbia. 392 p.201APPENDIX 1CONFORMABLE LATE JURASSIC (OXFORDIAN) TO EARLY CRETACEOUS STRATA,NORTHERN BOWSER BASIN, BRITISH COLUMBIA: DISCUSSIONMacLeod and Hills' (1990a) article entitled "Conformable Late Jurassic (Oxfordian) to EarlyCretaceous strata, northern Bowser Basin, British Columbia: A sedimentological and paleontologicalmodel" makes an important contribution to understanding the complex geologic history of the northernBowser basin. They have reported insightful sedimentologic descriptions and comprehensive plantmacrofossil collections. Errors in stratigraphic nomenclature and incompleteness in their treatment ofpreviously published paleontologic data, however, led them to propose revisions to the age andconformity of the strata that I feel warrants further discussion. I am pleased to have this opportunity todiscuss these points in light of recent research in the area and to be allowed to relate my bestunderstanding of the stratigraphic relationships.Three specific concerns with MacLeod and Hills (1990a) that I wish to discuss are: (1)erroneous application of published stratigraphic nomenclature (for example, inadvertently assigning theCurrier Formation type section to the McEvoy Formation); (2) incomplete discussion of previouslypublished paleontologic data with direct bearing on the age of the strata (most importantly omission ofreference to the late Oxfordian to early ICimmeridgian ammonite Amoeboceras recovered from within thecoal measures, calling into question their conclusion that the Currier Formation is entirely EarlyCretaceous); and (3) comparisons with plant macrofossils collected elsewhere in western Canada whichdo not always utilize the most current age estimates of those strata (at least five out of seven"Neocomian" and "Aptian" floras from outside the northern Bowser basin in Figure 9 from MacLeod andHills (1990a) occur in stratigraphic units that extend into the early or middle Albian, calling intoquestion their conclusion of a pre-Albian age for strata in the northern Bowser basin). In the followingdiscussion, I will explore these points more completely, and then conclude by presenting my bestresolution of the data, based on work by myself and Moffat et al. (1988), and my understanding ofMacLeod and Hills' data.2021) Errors in stratigraphic nomenclatureThe first point I will discuss is the misapplication of previously published stratigraphic names(Cookenboo and Bustin 1989). This point may be the most important in as much as, without consistencyof usage, no comparison of work between authors can be meaningful. For this reason, formaldescriptions, including type section localities were published in Cookenboo and Bustin (1989) for each ofthe formalized stratigraphic units. The location of the Currier type section described in Cookenboo andBustin (1989, p. 1007) as "southwest of Mount Klappan between the headwaters of the Little KlappanRiver, Tahtsedle Creek, and Didene Creek (lat. 57°11'30"N, long. 129°0'50"W)" and shown in mapview (Figure 2 of the same article) plots in the middle of "McEvoy" on MacLeod and Hills' (1990a) map(Fig. Al-la). A map resulting from detailed field work in the area shows that McEvoy Formation strataare restricted to only the top of the ridge where the Currier Formation type section is located (Fig. Al-lb; Moffat, 1985). Although none of MacLeod and Hills' section locations are included in their recentarticle, additional photographs were presented by MacLeod and Hills (1990b) which clearly show that theCurrier type section had been extensively sampled, but MacLeod and Hills (1990a, 1990b) incorrectlyincluded those samples in their McEvoy floral collections. One effect of this error in stratigraphicnomenclature is to mix Currier and McEvoy plant collections, and thereby mask any age differences thatmay exist between the stratigraphic units. This mixing of stratigraphic units is compounded in MacLeodand Hills' (1990a) comparison of flora with other occurrences in western Canada. As shown in theirFigure 9 (MacLeod and Hills 1990a, p. 996), the McEvoy and Devils Claw Formations are treated asone unit ("McEvoy-Devils Claw flora"), but as I have already remarked, plants collected from part of theCurrier were included in their McEvoy flora. MacLeod and Hills (1990a, pp. 994, 995) conclude that"There are no floral breaks either within or between the Currier and McEvoy formations" and that "Theconformable nature of the entire Jackson to Devils Claw stratigraphic sequence is further supported bythe apparent continuity of macrofloras across the Currier-McEvoy boundary." Given their mixing ofdefined stratigraphic nomenclature, I feel such conclusions are unsupported. I believe this error may betraced to MacLeod and Hills' attempt to directly equate informal stratigraphic units developed by GulfCanada Resources T imited (1984)203Figure A1-1: The location of the Currier Formation type section. Map A is from MacLeod and Hills(1990a) where the Currier type section plots in what are called "McEvoy" strata, and map B is a detailfrom field mapping by Moffat (1985). Mapped lithostratigraphic units are: Kdc= Devils Claw; Kmu=upper McEvoy; Kml= lower McEvoy; Jc= Currier; and Jj= Jackson.( a )%( b )Q = Recent cover^`■ ..... fault5fLocation ofCurrier Formation0tkilometres type sectionLocation oflierAkAMIP Q„...--^/Km, --,--,„. .,Currier Formation^riN"ir \.. .4:type section -----eir-------\-^r \""--.... jc /f^-A% ',..^/.,%., ^'-‘,Jc^C";N.---.:'McEvoy • I'M \^•,. ...;.,.--:-.--u_7...,,, I\-e.Zw, i v^.^. ' ^/I •44 .^• N. '^•Jj^' ..,^, s\,,.. N,N. \A^-1 i \^.i'lC--),_ -,,"S.<^'k•-4 i^‘^:+k...".;^-..„jc-,...: t^cs,‘.....‹.....^,1 /l,ieut^I I-z.^k--NNco^r^,9^'eaNass Lake \ 'N. l?-NS.(\:;\\*11\ '^.2.`.1\ ..".I ‹....4i1• 11.^Km .^'r.1 N..^f:—.^•■^I 'st-N 't.1...1 1^/---^-- s ^N f—...,1/4^1Jc '^‘^t-',"!t,^ ,:t^•,. _,—',•-• •^\ Ki'N ''' -`-1 v0^201 110^100,,, Q^C\\-1(m'...,^....--;:- 1 -9 ....)- N..kilometres 2-.c)., Kn.^ ‘"128°30'20457°00' —56°40'129°30'during their exploration of the Klappan coalfield to published formations (Cookenboo and Bustin 1989).Gulf used a similar four unit division of the stratigraphy, but the stratigraphic boundaries chosen ascontacts were not equivalent to ours. Gulfs coal-bearing unit (referred to as the Klappan unit inunpublished coal reports; Gulf Canada Resources Limited 1984, 1987) started with the first thick coaland ended with the last thick coal. The Currier's contact with the McEvoy, however, was described as"...the appearance of thin- to thick- bedded light grey, fine sandstones of limited lateral extent, thicksiltstones, shales, and occasional pebbly conglomerates, as well as a marked reduction in coal"(Cookenboo and Bustin 1989, p. 1005). The upper part of the Currier Formation, which is well exposedin the type section, in fact has few thick coals.Coal-bearing strata that previously were treated as part of the informal upper Jackson unit(Moffat 1985; Moffat et al. 1988; Cookenboo and Bustin 1989) were included in the Currier Formationby MacLeod and Hills (1990a). The same revision in nomenclature (effectively lowering the basalcontact of the Currier to include all of the coal-bearing strata (Fig. A1-2), some of which werepreviously included in the upper portion of the underlying Jackson unit) was anticipated by Cookenbooand Bustin (1990), leaving me in complete agreement regarding the most useful definition of the base ofthe Currier Formation. This lower part of the Currier Formation, which is poorly exposed in the typesection, contains much of the coal, and may be more closely equated with Gulfs informal Klappan unit.Equating the informal Klappan unit with the Currier Formation as defined in Cookenboo and Bustin(1989) may have led MacLeod and Hills to inadvertently exclude the Currier type section (andpresumably other equivalent strata) from the Currier Formation, and to include those strata in theMcEvoy Formation. MacLeod and Hills' "Currier" designation therefore may correctly represent aconsistent and mappable lithostratigraphic unit, but it is invalid because it incorrectly uses a previouslydefined formation name.I wish to emphasize that differences I have pointed to between MacLeod and Hills' (1990a)stratigraphy and that found in earlier publications about the area (Cookenboo and Bustin 1989; Moffat etal. 1988) relate to nomenclature, specifically the misuse of defined stratigraphic names I do not disputetheir geologic descriptions or interpretations, rather I point out that in their comparison205Figure A1-2: Stratigraphic nomenclature used by various authors in the coal-bearing section of thenorthern Bowser basin.Bustin and^Moffat (1985)Moffat (1983) Moffat et al.(1988)Cookenbooand Bustin(1989)andcookeebooand Bustin(1990)This discussion^Discussion^GulfPaPer andMacleod andHis (1990)CanadaResources Ltd.(1984, 1987)Devils ClawunitDevils ClawunitDevils ClawFormationDevils ClawFormationDevils ClawFormationDevils ClawMcEvoyunitMcEvoyunitMcEvoyFormationCurrierFormationTMcEvoyFormationCurrierFormationMcEvoyFormationMaHochCurrierunitCurrierunitupper Jacksonunitupper JacksonunitCurrierFormationKUPParlJacksonunitJacksonunitJacksonunitAshmanFormationJacksonunitSpate206of results with previous work, they are in part referring to the same rocks by different names. I believe itis important to point this out, or great confusion may result among interested readers.Perhaps renaming MacLeod and Hills' "Currier" the Klappan member of the lower CurrierFormation would eliminate much of the stratigraphic confusion. Their usage of "McEvoy" should bediscontinued because it includes both the upper strata of the Currier Formation and the strata that wereactually defined as the McEvoy Formation.2) Omission of relevant paleontological dataThe second point I wish to raise is the omission in MacLeod and Hills (1990a) of relevantpreviously published paleontologic data. Perhaps the most significant omission is that of the recovery ofthe Late Jurassic ammonite Amoeboceras identified by P. Smith (personal communication, in Moffat andBustin 1984) and H. Tipper (personal communication, in Moffat et al. 1988). The photographreproduced in figure A1-3 was not included in the article by Moffat et al. (1988) because one reviewersuggested it added little. This ammonite was recovered from within the coal-bearing strata 20 m belowthe top of the informal Jackson unit, as the term was applied by Moffat (1985). A coal seam 1 to 1.5 mthick crops out immediately below, placing the strata within the Currier Formation, as used by MacLeodand Hills (1990a). Elsewhere in Canada, Amoeboceras occurs in the black shale facies of the UpperBowser Lake Group (Tipper and Richards 1976), in the Upper Oxfordian Passage beds in the FemieGroup near Jasper, Alberta, and in the Upper Oxfordian green beds of the Femie Group near CarbondaleRiver, Alberta, where it is associated with Buchia concentrica (Frebold and Tipper 1970; Hall 1984). Inthe Arctic, Amoeboceras is known from strata only as young as the early Kimmeridgian (Frebold 1961),strongly suggesting the Currier is not entirely Early Cretaceous.Seemingly dismissed by MacLeod and Hills (1990a) are previously reported palynomorphassemblages recovered from the McEvoy and Devils Claw Formations. Three assemblages, assigned aBarremian age (G. E. Rouse, personal communication, in Cookenboo and Bustin 1989), a middle Albianage (Moffat et a/. 1988) and a Cenomanian age (G. E. Rouse personal communication, in Cookenbooand Bustin 1989) all contained angiosperm microflora (figured in Moffat et al. 1988). The oldest(Barremian) assemblage was recovered 35 m from the base of the McEvoy Formation207Figure A1-3: The upper Jurassic ammonite Anweboceras recovered from the coal-bearing section in thenorthern Bowser basin (1.5x). The tabulate venter (not seen here) bears a serrated keel.208from rocks just above the Currier type section, and 200 in above the last sample of Jurassicdinoflagellates recovered from the Currier type section. Barremian palynomorphs 35 m above the baseof the McEvoy Formation (and directly overlying the Currier type section) and a middle Albianpalynomorph assemblage in the overlying Devils Claw Formation (Moffat et al. 1988) led to thesuggestion that the McEvoy was Barremian to mid-Albian (Cookenboo and Bustin 1989). A suggestedCenomanian age for a palynomorph assemblage from higher in the Devils Claw Formation (G. E. Rouse,personal communication, in Cookenboo and Bustin 1989) led to an assigned age range of late middleAlbian to Cenomanian for the Devils Claw Formation (Cookenboo and Bustin 1989). Based onpublished palynological assemblages, the McEvoy and Devils Claw Formations together range fromBarremian to Cenomanian in age (Moffat et al. 1988; Cookenboo and Bustin 1989). This age range isolder, in part, than "the Albian to Cenomanian age" MacLeod and Hills (1990a, p. 995) mistakenlyattributed to Moffat et al. (1988) for these units. Further unpublished work has caused me to amend therange slightly to late Barremian or Aptian to latest Albian or Cenomanian.3) Comparisons with other plant macrofossil collectionsThe third and final point concerns the ages of leaf collections from elsewhere in western Canadathat MacLeod and Hills (1990a) compare to the Currier, McEvoy and Devils Claw flora. MacLeod andHills (1990a) closely follow Bell (1956) who assigned Neocomian and Aptian ages to leaf fossilcollections from a number of units. More recently published ages for the same stratigraphic units basedon criteria other than plant macrofossils, including palynology, ammonites, trigonids, buchias, and K-Ardates, have yielded a new perspective on the age of some of these units. Five of the seven stratigraphicunits containing the floras (excluding the mixed Currier, McEvoy and Devils Claw floras) listed inFigure 9 of MacLeod and Hills (1990a) as being Aptian or Neocomian in age, extend to the early ormiddle Albian (specifically the Gething, Lower Blairmore, Jackass Mountain, Tantalus, and Skeenafloras; Stott 1963, 1968, 1982; Jeletzky and Tipper 1968; Tipper and Richards 1976; Kleinspenh 1985;Lowey and Hills 1988).209The earliest and perhaps best documented claim that Bell's (1956) "Aptian" flora occurs in earlyor middle Albian age strata was made by Stott (1963) for the Gates Formation in the WCSB. Stott(1968, pp. 76-77) later named 41 plant forms that were identified by Bell, D. C. McGregor, and F. M.Hueber as "Luscar (Aptian)" flora. Bell (1956, pp. 10-11: quoted in Stott 1968) concluded that "theextreme rarity of dicotyledons within it, together with the survival of many species occurring in theKootenay flora, certainly favors the Aptian age.", but Stott showed conclusively that the Gates Formationoverlies marine shales of the Moosebar, which contains marine fauna of middle Albian age. Jeletzky andTipper (1968), writing about the Jackass Mountain Group, recognized the controversy about whetherBell's "Aptian" flora is really Aptian or Albian and concluded "In this report, the strata containing thisflora will be considered Aptian fully realizing that some or all of these rock units may be middle or lowerAlbian in age." (Jeletzky and Tipper 1968, p. 47). Caldwell (1984) also recognized the "Aptian" floraage difficulty, and concluded "A survey of the literature conveys the impression that too great a reliancehas been placed on the dominantly pre-Albian age suggested originally by the flora, despite anindistinguishable flora having been recovered from younger rocks of known Albian age." (Caldwell1984, p. 187).MacLeod and Hills' (1990a) conclusion that all of the McEvoy and Devils Claw Formations arepre-Albian in age rests largely on their assumption that leaf collections lacking angiosperm leaves arerestricted to Aptian or older strata. Persistence until the late Albian of a flora lacking in angiosperms iswell documented in northern Alaska (Smiley 1969a, 1969b; Spicer 1987). An estimated 10 000specimens from 250 localities in the 3 000 m thick Nanushuk and Colville Groups of the north slope ofAlaska were divided by Smiley (1969a, 1969b) into seven floral zones. The seven floral zones are datedby marine invertebrates and microfossils which were present in interbedded marine sediments. The twooldest floral zones range from middle Albian or older to late Albian in age and are dominated bypteridophytes, ginkgophytes, cycadophytes, and conifers. The first angiosperms appear in the thirdfloral zone of late Albian to early Cenomanian age (Smiley 1969a; Inoceramus dunveganensis first occurswithin the upper part of this floral zone). Angiosperms only become dominant components of the florain the overlying fourth floral zone. A notable similarity with the McEvoy and Devils Claw Formations210is the intermittent occurrence of rare angiosperm palynomorphs in strata deposited before the firstappearance of angiosperm leaves (May and Shane 1985). The occurrence of such a similar flora in othermiddle and late Albian rocks largely refutes MacLeod and Hills' argument that the composition of theBowser basin flora precludes an Albian age for the McEvoy and Devils Claw Formations.The difficulty in using plant macrofossils as a guide to age is partly due to the long-rangingnature of plant macrofossils. This long-ranging character is well demonstrated in MacLeod and Hills(1990a) Figure 4 which lists the worldwide and North American ranges of 16 "dominant plant fossils".As figured, only 4 species do not extend through the Albian. Furthermore, two of those four,Sagenopteris williamsii (Newberry) Bell and Ctenis borealis (Dawson), previously have been reportedfrom the early Albian Gates Formation (Stott 1968). Contrary to MacLeod and Hills' (1990a, p. 994)statement that "Most of the 19 species that are common to both the Currier and McEvoy formations arealso long-ranging, pre-Albian, Mesozoic taxa... ", I conclude, based in part on their own data, that mostor perhaps all sixteen species common to the Currier and McEvoy formations are long-ranging Throughthe Albian.Because the plant macrofossils are long-ranging, age assignments based on them must be basedon paleofloristic abundance trends rather than extinction events. MacLeod and Hills demonstrate, basedon their extensive macrofloral collections, that the Currier through Devils Claw floras are rich incycadophytes and gingkophytes and contain only rare Sagenopteris leaves. Additionally, in all theirfloral collections no angiosperm leaves have been found. MacLeod and Hills (1990a) use these facts tosuggest paleofloristic similarities to collections of leaves from the Western Canada Sedimentary Basin.Based on my understanding of the ages of these floras and the Alaskan North Slope flora described bySmiley (1969a, 1969b), I suggest that MacLeod and Hills' leaf collections are consistent with an age asyoung as middle or late Albian. A middle or late Albian age for part of the McEvoy and Devils ClawFormations is entirely consistent with the ages deduced from palynological data (including angiospermpollen and dinoflagellate cysts figured in Moffat et al. 1988), and does not necessitate a reappraisal of theages as MacLeod and Hills (1990a) propose.211Although MacLeod and Hills (1990a) assert that paleofloristic abundance trends suggest theMcEvoy and Devils Claw are pre-Albian in age, I suggest that a better explanation of the paleofloristictrends they identify may be variations in climate. This seems especially likely because none of the"dominant plant macrofossils" demonstrably become extinct before the end of the Albian, and similarfloras as young as late Albian are known. Vahkrameev (1978) recognized an abundant ginkgo and cycadflora, similar to that described by MacLeod and Hills, as indicative of a warm, temperate and humidclimatic province. Perhaps the western margin of North America remained more temperate and humid(or perhaps simply less seasonal) longer into the Albian than did the Western Canada Sedimentary Basin,which was apparently separated from the Pacific during the Early Cretaceous by mountains of the mid-Columbian orogen (Stott 1982).My resolution of the dataConsidering the points I have raised, it seems incumbent upon me that I give my best resolutionof the complete spectrum of published evidence bearing on the ages of the Currier, McEvoy and DevilsClaw Formations. I will consider each unit separately, summarizing the data that I believe most relevantand trying, where possible, to achieve the most compatible interpretation of the varied data sets.There are two apparently incompatible data sets that bear directly on the age of the CurrierFormation. The ammonite Amoeboceras (Moffat et al. 1988) and dinoflagellates includingNannoceratopsis sp. , Pareodinia minuta, and Pareodinia cf. ceratophora (Cookenboo and Bustin1989), recovered from coal-bearing strata, suggest that the Currier is at least in part Late Jurassic.MacLeod and Hills (1990a) have produced specimens of Buchia from strata below the Currier,suggesting a Late Jurassic (Oxfordian to Tithonian) age, and the astariid bivalve Herzogina, previouslyknown only from Neocomian strata, in the Currier Formation (MacLeod and Hills 1990a). Resolution ofthe paradox of latest Jurassic Buchia occurring stratigraphically below Kimmeridgian to Oxfordianammonites appears possible in four ways: (1) some or all of the invertebrates and microfossils were mis-identified; (2) ammonite and dinoflagellate cysts were redeposited in younger strata; (3) age relationsbetween Late Jurassic Buchias and Amoeboceras are less well known than previously thought; or (4) the212stratigraphy and structure is more complex than previously suspected. No evidence has been offered tosuggest that any of the first three options apply. Given the complexity of the structure and stratigraphy inthe northern Bowser basin, the fourth possibility seems most likely. Because the locations of the Buchiacollections were not provided in MacLeod and Hills (1990a), suggestions to resolve this apparentcontradiction may be premature, but perhaps the Buchia beds are not stratigraphically below all the coal-bearing strata and record a post- or late Currier marine transgression.Resolution of the age of the Currier rests on several questions that are still open, but at present,it seems strongly indicated (based on the ammonite Amoeboceras and dinoflagellates recovered from thecoal-bearing strata) that the Currier is in part Late Jurassic, and (based on the species of Buchias andHerzogina) may also be in part latest Jurassic and earliest Cretaceous.The age of the McEvoy and Devils Claw formations can be resolved somewhat moresatisfactorily by realizing that the "lower Blairmore-Luscar-Gething flora extends upward into rocks ofmiddle Albian age." (Stott 1968, p. 77). A Barremian to middle Albian age for the McEvoy is consistentboth with similar macroflora occurrences elsewhere in western North America and with published agesbased on palynology. Palynomorphs recovered from the Devils Claw Formation are consistent with anage range of late middle Albian to latest Albian or CenomanianFinally, I wish to give my view of the evidence for and against the existence of anunconformable contact between the Currier and McEvoy Formations. Cookenboo and Bustin (1989)suggested an age range of Oxfordian to Kimmeridgian or Tithonian for the Currier and Barremian tomiddle Albian for the McEvoy. MacLeod and Hills (1990a, p. 996) conclude that "The presence ofHerzogina (Neocomian) near the base of the Currier Formation and a macroflora with a distinctly EarlyCretaceous, pre-Albian age component in the Currier and overlying McEvoy and Devils ClawFormations eliminates the need for a major Kimmeridgian-Barremian [Kimmeridgian or Tithonian toBarremian is a more accurate representation of the proposal] unconformity in this area as proposed byCookenboo and Bustin (1989)." I concur with MacLeod and Hills that their evidence fails to support anunconformity, but due to the long-ranging nature of the macroflora and the mixing of stratigraphic unitsby MacLeod and Hills, I believe that the data presented by MacLeod and Hills (1990a) are insufficient to213deny the existence of such an unconformity. Ammonite and palynological evidence for a Late JurassicCurrier and palynological evidence for a Barremian to Albian or Cenomanian McEvoy and Devils Clawis neither conclusively supported nor contradicted by the evidence MacLeod and Hills have so farpresented. Even if the Currier is in part Neocomian as suggested by MacLeod and Hills' Buchias andHerzogina, the Currier may only be as young as the earliest Cretaceous (Berriasian), allowing most of theNeocomian for an hiatus or unconformity. The occurrence of such an unconformity is consistent with thewell established regional unconformity known from Lower Cretaceous strata elsewhere in the CanadianCordillera, both along the margins of the Bowser basin (Eisbacher 1974; Tipper and Richards 1976) andwithin the Western Canada Sedimentary Basin (Stott 1982). Because there is no compelling reason tochange views, I elect to stand by the proposal that a unconformity probably separates the Currier andMcEvoy Formations in the study area (although perhaps of somewhat briefer duration thanKimmeridgian or Tithonian to Barremian span originally proposed.).214REFERENCESBell, W. A. 1956. Lower Cretaceous floras of western Canada. Geological Survey of Canada, Memoir285.Bustin, R. M., and Moffat, I. 1983. Groundhog Coalfield, British Columbia: reconnaissance stratigraphyand structure. Bulletin of Canadian Petroleum Geology, v. 31, p. 231-245.Caldwell, W. G. E. 1984. Early Cretaceous transgressions and regressions in the southern interiorplains. In The Mesozoic of middle North America. Edited by D. F. Stott and D. J. Glass.Canadian Society of Petroleum Geologists, Memoir 9, p. 173-203.Cookenboo, H. 0., and Bustin, R. M., 1989. Jura-Cretaceous (Oxfordian to Cenomanian) stratigraphyof the north-central Bowser Basin, northern British Columbia. Canadian Journal of EarthSciences, v. 26, p. 1001-1012.Cookenboo, H. 0., and Bustin, R. M., 1990. Stratigraphy of coal occurrences in the Bowser Basin inGeological Fieldwork 1989. British Columbia Ministry of Energy, Mines and PetroleumResources, Paper 1990-1, p. 473-477.Eisbacher, G. H. 1974. Sedimentary and tectonic evolution of the Sustut and Sifton Basins, north-centralBritish Columbia. Geological Survey of C: I Ada, Paper 73-31.Frebold, H. 1961. The Jurassic faunas of the Canadian Arctic. Geological Survey of Canada, Bulletin74.Frebold, H., and Tipper, H. W. 1970. Status of the Jurassic in the Canadian Cordillera of BritishColumbia, Alberta, and southern Yukon. Canadian Journal of Earth Sciences, v. 7, p. 1-21.Gulf Canada Resources Limited 1984. Mount Klappan property. British Columbia Ministry of Energy,Mines and Petroleum Resources, Open File Report 111.Gulf Canada Resources Limited 1987. Lost Fox property. British Columbia Ministry of Energy, Minesand Petroleum Resources, Open File Report 723.Hall, R. L. 1984. Lithostratigraphy and biostratigraphy of the Fernie Formation (Jurassic) in thesouthern Canadian Rocky Mountains. In The Mesozoic of middle North America. Edited by D.F. Stott and D. J. Glass. Canadian Society of Petroleum Geologists, Memoir 9, p. 233-247.Jeletzky, J. A., and Tipper, H. W. 1968. Upper Jurassic and Cretaceous rocks of Taseko Lakes map areaand their bearing on the geological history of southwestern British Columbia. Geological Survey215of Canada, Paper 67-54.Kleinspenh, K. L. 1985. Cretaceous sedimentation and tectonics, Tyaughton-Methow Basin,southwestern British Columbia. Canadian Journal of Earth Sciences, v. 22, p. 154-174.Lowey, G. W., and Hills, L. V. 1988. Lithofacies, petrography and environments of deposition,Tantalus Formation (Lower Cretaceous) Indian River area, west-central Yukon. Bulletin ofCanadian Petroleum Geology, v. 36, p. 296-310.MacLeod, S. E., and Hills, L. V. 1990a. Conformable Late Jurassic (Oxfordian) to Early Cretaceousstrata, northern Bowser Basin, British Columbia: A sedimentological and paleontological model.Canadian Journal of Earth Sciences, v. 27, p. 988-998.MacLeod, S. E., and Hills, L. V. 1990b. Conformable Late Jurassic (Oxfordian) to Early Cretaceous(Aptian) strata, northern Bowser Basin. Geological Association of Canada/MineralogicalAssociation of Canada Annual Meeting, Program with Abstracts, v. 15, p. A80.May, F. E., and Shane, J. D. 1985. An analysis of the Umiat delta using palynological and other data,North Slope, Alaska. In Geology of the Nanushuk Group and related rocks, North Slope,Alaska. Edited by A. C. Huffman. United States Geological Survey, Bulletin 1614, p. 97-120.Moffat, I. W. 1985. The nature and timing of deformational events and organic and inorganicmetamorphism in the northern Groundhog Coalfield: implications for the tectonic history of theBowser Basin. Ph.D thesis, University of British Columbia, Vancouver, B.C.Moffat, I. W., and Bustin, R. M. 1984. Superposed folding in the northern Groundhog coalfield;evidence for polyphase deformation in the northeastern corner of the Bowser basin. In Currentresearch, part B. Geological Survey of Canada, Paper 84-1B, p. 255-261.Moffat, I. W., Bustin, R. M., and Rouse, G. E. 1988. Biochronology of selected Bowser Basin strata:tectonic significance. Canadian Journal of Earth Sciences, v. 25 p. 1571-1578.Smiley, C. J. 1969a. Cretaceous floras of the Chandler-Colville region, Alaska: stratigraphy andpreliminary floristics. American Association of Petroleum Geologists Bulletin, 53:482-592.Smiley, C. J. 1969b. Floral zones and correlations of Cretaceous Kukpowruk and Corwin Formations,northwestern Alaska. American Association of Petroleum Geologists Bulletin, v. 53, p.2079-2093.Spicer, R. A. 1987. Late Cretaceous floras and terrestrial environment of northern Alaska. In AlaskanNorth Slope geology. Edited by I. Tailleur and P. Weimer. Society of EconomicPaleontologists and Mineralogists, v. 50, p. 497-512.216Stott, D. F. 1963. Stratigraphy of the Lower Cretaceous Fort St. John Group, Gething and CadominFormations, Foothills of northern Alberta and British Columbia. Geological Survey of Canada,Paper 62-39.Stott, D. F. 1968. Lower Cretaceous Bullhead and Fort St. John Groups between Smoky and PeaceRivers, Rocky Mountain Foothills, Alberta and British Columbia. Geological Survey of Canaria,Bulletin 152.Stott, D. F., 1982. Lower Cretaceous Fort St. John Group and Upper Cretaceous Dunvegan Formationof the Foothills and Plains of Alberta, British Columbia, District of Mackenzie and YukonTerritory. Geologic Society of Canada, Bulletin 328.Tipper, H. W., and Richards, T. A. 1976. Jurassic stratigraphy and history of north-central BritishColumbia. Geological Survey of Canaria, Bulletin 270.Vachrameev, V. A., 1978. The climates of the northern hemisphere in the Cretaceous in light ofpaleobotanical data. Paleontological Journal, V. 12, 143-154.217


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items