Convergent Margin Tectonics in the North American Cordillera:
Implications for Continental Growth and Orogeny
by
Joseph M. English
B.A., Trinity College Dublin, 2001
A dissertation submitted in partial fulfillment of the requirements for the degree of
DOCTOR OF PHILOSOPHY in the School of Earth and Ocean Sciences
O Joseph M. English, 2004 University of Victoria
All rights reserved. This dissertation may not be reproduced in whole or in part, by photocopy or other means, without the permission of the author.
ABSTRACT Supervisor: Dr. Stephen T. Johnston
Continental growth may be accomplished at active convergent margins through tectonic accretion and orogeny. Accretionary processes believed to add material to continents include the collision of island arcs and other unsubductable crustal blocks. Using the Intermontane belt as a case study for assessing accretionary processes, it is concluded that island-arc collision and accretion was the principal mechanism for continental growth with relatively minor contributions from 'sliced-off oceanic seamounts and/or plateaux. Fold and thrust belt formation in the northern Intermontane belt records a Middle Jurassic orogenic event that can be attributed to the collision of island-arc highlighting the importance of island-arc collision for causing orogenesis in the North American Cordillera. However, not all orogenic events in the North American Cordillera can be readily attributed to a collisional event. The leading model for driving Laramide orogenesis in the United States is flat-slab subduction, and thermal modelling indicates that subduction of a relatively buoyant oceanic plateau/aseismic ridge may have been responsible for the shallow trajectory. In the Canadian and Mexican portions of the Laramide, the coeval development of a magmatic arc within 300 km of the trench refutes the existence of flat-slab subduction in these regions, and therefore the processes responsible for this orogeny remain enigmatic and require resolution.
TABLE OF CONTENTS
. .
ABSTRACT...
11...
TABLE OF CONTENTS...
111 LIST OF TABLES...
v LIST OF FIGURES...
vi ACKNOWLEDGEMENTS...
xi CHAPTER 1: INTRODUCTION INTRODUCTION...
2 CONTINENTAL GROWTH...
3 Accretion...
3 Orogeny...
4 PRIMARY OBJECTIVES...
10 CONTRIBUTIONS...
11 METHODOLOGY...
12 REFERENCES...
15CHAPTER 2: ACCRETIONARY PROCESSES IN THE NORTHERN CACHE CREEK TERRANE. CANADIAN CORDILLERA INTRODUCTION
...
18GEOLOGICAL BACKGROUND
...
20...
GEOLOGY OF THE NAKINA AREA 23 Mantle rocks...
24Mafic Intrusive rocks
...
25Mafic Volcanic rocks
...
26Hemipelagite/Siliciclastic rocks
...
27Carbonate rocks
...
28Yeth Crcek formation
...
29COMPOSITION OF MAFIC ROCKS
...
30Methods
...
30Geochemistry
...
32. .
Rock classification diagrams...
32. . .
Th-Hf-Nb discrimmation diagram...
33T h N b versus N b N b plot
...
34Rare earth element (REE) plots
...
37Trace element spider diagrams
...
39PETROGENETIC INTERPRETATIONS
...
41COMPARISONS WITH OTHER DATA FROM NORTHERN CACHE CREEK
...
TERRANE 45 ACCRETIONARY PROCESSES...
49What is accretcd?
. .
...
49lmplicat~ons for continental growth
...
50CONCLUSIONS
...
52REFERENCES
...
54CHAPTER 3: STRATIGRAPHY AND STRUCTURE OF THE CENTRAL WHITEHORSE TROUGH: FOLD AND THRUST BELT FORMATION IN THE NORTHERN CANADIAN CORDILLERA INTRODUCTION
...
61WHITEHORSE TROUGH STRATIGRAPHY
...
63CENTRAL WHITEHORSE TROUGH STRATIGRAPHY
...
66CENTRAL WHITEHORSE TROUGH STRUCTURE
...
73Middle Jurassic NW-SE trending thrust faults and folds
...
73Younger folds and faults
...
77SUMMARY
...
78REFERENCES
...
79CHAPTER 4: COLLISIONAL OROGENESIS IN THE NORTHERN CANADIAN CORDILLERA: IMPLICATIONS FOR CORDILLERAN CRUSTAL STRUCTURE. OPHIOLITE EMPLACEMENT. CONTINENTAL GROWTH AND THE TERRANE HYPOTHESIS INTRODUCTION
...
84UPPER TRIASSIC . MIDDLE JURASSIC ISLAND-ARC SYSTEM
...
86Stikine magmatic arc
...
86Whitchorse Trough forearc basin
...
88Cache Creek 'ophiolite'
...
89Cache Crcek subduction complex
...
90EVOLUTION AND MIDDLE JURASSIC COLLISIONAL OROGENESIS
...
91IMPLICATIONS
...
95Cordilleran crustal structure
...
95Ophiolite emplacement
...
97Continental growth
...
97Terrane hypothesis
...
98REFERENCES
...
100CHAPTER 5: THERMAL MODELLING OF THE LARAMIDE OROGENY: TESTING THE FLAT-SLAB SUBDUCTION HYPOTHESIS INTRODUCTION
...
106 THERMAL MODEL...
109 MODEL RESULTS...
114 DISCUSSION...
120 CONCLUSIONS...
126 REFERENCES...
128CHAPTER 6: INTRACONTINENTAL DEFORMATION BELTS OF THE
LARAMIDE OROGENY: WHAT ARE THE DRIVING FORCES?
INTRODUCTION
...
135Retroarc thrusting
...
135'Orogenic float' tectonics
...
138Flat-slab subduction
...
139. .
Cordilleran transpressional colllslon...
140DISCUSSION
...
141REFERENCES
...
145CHAPTER 7: CONCLUSIONS AND SUGGESTIONS FOR FUTURE WORK CONCLUSION
...
149SUGGESTIONS FOR FUTURE WORK
...
152Appendix I: Whole rock major, trace and REE abundances for mafic rocks in the Nakina area of the northcrn Cache Crcek terrane
...
154Appendix 11: Hydrocarbon Potential of the central Whitehorse Trough, northern Canadian Cordillera
...
162Appendix 111: Rock-Eval VI programmed pyrolysis data from the central Whitehorse Trough
...
183LIST OF TABLES CHAPTER 2 Table 2.1: Normalised percentages of the mapped areas of each unit in the northern Cache Creek terrane (based on Mihalynuk et al., 1996; Massey et al., 2003)
...
22CHAlTER 3 Table 3.1: Ammonite identifications from Atlin Lake.
'
Identifications by Gary G. Johannson. Identification by T. Poulton (GSC); Fossil Report J2-2004-TPP 75 APPENDIX I1 Table 11.1: Vitrinite reflectance data from the central Whitehorse Trough...
174LIST OF FIGURES
CHAlTER 1
Figure 1.1: Map of British Columbia showing the distribution of the primary components of the Intermontane belt in the northcrn Canadian Cordillera (top right) and a regional geologic map (main). This geology map does not include post-Middle Jurassic rock units. Thc study area is outlined by a dashed box
...
5 Figure 1.2: Map showing areas of pre-Laramide and Laramide magmatism, the extent of the Laramide-age thin-skinned fold-and-thrust belt and thick-skinned block uplifts, and the approximate location of the Laramide flat-slab according to Saleeby (2003). Note that Cenozoic extension has not been restored in this figure. This figure was drafted using GMT software (Wessel and Smith, 1995)...
8 Figure 1.3: Distribution of field stations taken within the study area in the northern Canadian Cordillera. See Figure 1.1 for regional location map...
1 3 CHAPTER 2Figure 2.1: Map of British Columbia showing the distribution of the primary components of the Intermontane belt in the northern Canadian Cordillera (top right) and a regional geologic map (main). This geology map does not include post-Middle Jurassic rock units. The Nakina area is outlined by a dashed box
...
19 Figure 2.2: Generalised bedrock geology map of the northcrn Cache Creek terrane in thc...
Nakina map-area with a 10 km UTM grid superimposed (NTS 104Nl1, 2 and 3) 21 Figure 2.3: Nakina area lithogeochemical data plotted on rock classification diagrams. Volcanic rocks from the carbonate assemblage are alkaline, and range from alkali basalts and basanites to trachytes, while rocks from the oceanic crustal assemblage are subalkaline and dominantly basaltic and basaltic-andcsitic in composition. (a) Na20+K20 versus SiOz (TAS) diagram (from Cox et al., 1979). Abbreviations: P-N - phonolite-
nephelinite, P-T - phonolite-tcphrite, B+T - basanite
+
tephrite, B-A - basaltic andesite.(b) Immobile trace element abundances are also used for rock classification (from Winchester and Floyd, 1977), particularly in altcred volcanic rocks where elemental mobility is suspected. Abbreviations: Bsn/Nph - BasaniteINephelinite; Com/Pant -
ComenditeIPantellerite
...
31 Figure 2.4: Th-Hf-Nb tectonic discrimination diagram after Wood (1980). Thc fields on the diagram can be defined as: (A) N-typc MORB, (B) E-type MORB and tholeiitic within-plate basalts and differcntiatcs, (C) Alkaline within-plate basalts and differcntiates, and (D) Destructive plate-margin basalts and differentiates. Magmatic rocks from the oceanic crustal assemblage can be separated into four components: calc- alkaline arc rocks, island-arc tholeiitcs, BABBIN-MORB and E-MORB. Pillow basalts from the Yeth Creek assemblage are geochemically identical to BABBIN-MORB from the oceanic crustal assemblage of the Cache Crcek terrane. Volcanic rocks from thc carbonate unit are classified as alkalinc within-plate basalts and differentiatcs. Key for. .
vii
Figure 2.5: Nakina area lithogeochemical data plotted on a T h N b versus N b N b diagram (modified from Pearce et al., 1995; Metcalf et al., 2000). The source depletion vector indicates increasing levels of mantle source depletion produced by prior melt extraction. The subduction component vector indicates the compositional effect of the addition of subduction-derived fluids to the melt. Liquid- and cumulate-fractionation vectors parallel the source depletion vector. The majority of samples from the oceanic crustal assemblage and the Yeth Creek assemblage are characterised by low Th/Yb and Nb/Yb ratios indicating that they were derived from a depleted mantle source similar to that for N- MORB's. The island-arc tholeiitcs and calc-alkaline arc rocks of the oceanic crustal assemblage have higher T h N b ratios indicating the importance of subduction-derived fluids in their genesis. Calc-alkaline arc rocks from the oceanic crustal assemblage plot along the liquid-fractionation vector suggesting that these calc-alkaline rocks may have evolved from the more primitive island-arc tholeiites. Volcanic rocks from the carbonate unit are characterised by high T h N b and N b N b ratios indicating that they were dcrived from an enriched mantle source similar to that for OIB's. Key for symbols uscd in Figure
Figure 2.6: Chondrite-normalised rare earth element (REE) plots for samples from the Nakina area. Normalising values are from Taylor and McLennan (1985). Key for
. .
symbols used in F~gure 2.3...
38 Figure 2.7: Trace element spider diagrams for samples from the Nakina area. Data are normalised to N-MORB using the values from Sun and McDonough (1989). Key for.
.
symbols used in F ~ g u r e 2.3...
40 Figurc 2.8: Schematic diagrams depicting the evolution of the Stikine magmatic arc and the Cache Creek terranc in the northern Intermontane belt. Note that the Stikine terrane may have rotated during its evolution in order to enclosc the exotic elements of the Cache Creek tcrrane (e.g. Mihalynuk et al., 1994)...
44 Figurc 2.9: Comparison of Nakina lithogeochemical data with other data from the northern Cache Creek terrane using: (a) the Th-Hf-Nb tectonic discrimination diagram (after Wood, 1980); (b) a T h N b versus N b N b diagram (modified from Pearce et al., 1995; Metcalf et al., 2000); (c) Chondrite-normalised REE plots using the normalising values of Taylor and McLennan (1985); and (d) trace element spider diagrams normalised to N-MORB using the values from Sun and McDonough (1989). Data from the Hall Lake, French Range and Kutcho Crcek areas is from Mihalynuk and Cordey (1997)...
46CHAPTER 3
Figure 3.1: Map of British Columbia showing the distribution of the primary components of the Intermontane belt in the northern Canadian Cordillera (top right) and a regional geologic map (main). This geology map does not include post-Middle Jurassic rock units. The central Whitehorse Trough study area is outlined by a dashed box
...
62Figure 3.2: Generalised stratigraphic framework of the central Whitehorse Trough. Note that the Tanglefoot formation refers to Bajocian chert-pebble conglomerates in the Yukon portion of the Whitehorse Trough, but this formation name has not yet been applied in British Columbia. Timescale from Okulitch (2001)
...
64 Figure 3.3: Generalised stratigraphic column of the Inklin Formation at Atlin Lake (from Johannson, 1994). Stratigraphic thicknesses shown are not estimates of total thickness but represent minima based on measured sections. Not to scale...
67 Figure 3.4: Bedrock geology map of the central Whitehorse Trough; this map is based in part on previous published bedrock geology maps of Aitken (1959) and Souther (1971). Equal area stereonet projections show contoured poles to bedding from the Inklin Formation in the central Whitehorse Trough from: (a) the fold-dominated Atlin Lake region, and (b) the thrust-dominated Taku River region. Region boundary is shown on geologic map...
71 Figure 3.5: Cross-sections for the central Whitehorse Trough from: (a) the fold- dominated Atlin Lake region, and (b) the thrust-dominated Taku River region. Section lines are shown on Fig. 3.4. Some units in cross-section appear thicker than measured stratigraphic thicknesses due to internal imbrication and thrust faulting. Biochronological data from the Atlin Lake area is from Johannson et al. (1997), plus additional data is included in Table 3.1. (UP: Upper Pliensbachian; LP: Lower Pliensbachian; S: Sinemurian)....
74CHAPTER 4
Figure 4.1: Map of British Columbia showing the distribution of the primary components of the Intermontane belt in the northern Canadian Cordillera (top right) and a regional geologic map (main). This geology map does not include post-Middle Jurassic rock units. The study area is outlined by a dashed box
...
85 Figure 4.2: Generalised time-space plot showing the principal components of the Stikine and Cache Creek terranes in the northern Intermontane belt. Abbreviations: HG =Hazclton Group; SG = Stuhini Group
...
87 Figure 4.3: Schematic diagrams depicting the evolution of the Stikine magmatic arc and the Cache Creek terrane in the northern Intcrmontane belt. Note that the Stikine terrane may have rotated during its evolution in order to enclose the exotic elements of the Cache Creek terrane (e.g. Mihalynuk et al., 1994)...
92 Figure 4.4: Schematic crustal cross-section across the northern Intermontane belt...
96CHAPTER 5
Figure 5.1: Map showing areas of pre-Laramide and Laramide magmatism, the cxtent of the Laramide-age thin-skinned fold-and-thrust belt and thick-skinned block uplifts, and the approximate location of the Laramide flat-slab according to Saleeby (2003). Note that Cenozoic extension has not been restored in this figure. This figure was drafted using GMT software (Wessel and Smith, 1995)
...
107 Figure 5.2: Thermal structure of subduction zones with a convergence velocity of 5cmlyr. A: Geometry of a pre-Laramide normal subduction zone. B: 10 Myr flat slab at 60 km depth with no shear-heating. C: 50 Myr flat slab at 60 km depth with no shear- heating. D: 10 Myr flat slab at 60 km depth with shear-heating. E: 50 Myr flat slab at 60 km depth with shear-heating. F: 50 Myr flat slab at 90 km depth with no shear-heating. Note that subduction of young oceanic lithosphere results in slab dehydration prior to its descent into the asthenosphere
...
111 Figure 5.3: Pressure-temperature (P-T) diagram showing the P-T paths of oceanic crust.A: Metamorphic facies and partial melting curves for basaltic compositions from Peacock and Wang (1999) and references therein, and hydrous minerals stable in the eclogite field from Schmidt and Poli (1998). Eclogite field is shown in grey. B: P-T paths of oceanic crust for various flat-slab subduction zones shown in Figure 5.2. P-T ranges of the Farallon lawsonitc eclogite xenoliths from Usui et al. (2003) shown in stippled pattern. Note the early- stage dehydration of the young oceanic plates
...
113 Figure 5.4: Graphs showing the change in temperature of the oceanic crust at the end of the flat-slab segment (stars in Figs. 5.2B, 5.2C, 5.2D, 5.2E, and 5.2F) for various combinations of convergence velocity and slab age. A: Flat-slab segment at 60 km depth with no shear-heating. B: Flat-slab segment at 60 km depth with shear-heating. C: Flat- slab segment at 90 km depth with no shear-heating. Due to a difference in pressure, the metamorphic assemblages present in the 90 km-deep flat-slab segment are different from those in the 60 km-deep flat-slab segment. This figure was drafted using GMT software (Wcsscl and Smith, 1995)...
117 Figure 5.5: Graphs showing the change in temperature of the oceanic crust at the end of the flat-slab segment for different lengths of that segment. The greater the length of the flat-slab segment, the higher the temperature of the oceanic crust at the cnd of that segment, and vice versa...
119CHAPTER 6
Figure 6.1: Map showing areas of pre-Laramidc and Laramide magmatism, the cxtent of the Laramide-age thin-skinned fold-and-thrust belt and thick-skinned block uplifts, and the approximate location of the Laramide flat-slab according to Saleeby (2003). Note that Cenozoic extension has not been restored in this figure. This figure was drafted using GMT software (Wessel and Smith, 1995)
...
136 Figure 6.2: Schematic diagrams illustrating the various models that have been proposed...
APPENDIX I1
Figure 11.1: Map of British Columbia showing the distribution of the primary components of the Intermontane belt in the northern Canadian Cordillera (top right) and a regional geologic map (main). This geology map does not include post-Middle Jurassic rock units. The central Whitehorse Trough study area is outlined by a dashed box
...
164 Figure 11.2: Schematic pyrogram illustrating the liberation of hydrocarbon during heating of the rock s a m ~ l e . Determined parameters include S1, S2, S3, T,,,,,, and the hydrogen and.
.
oxygen ~ n d ~ c e s
...
167 Figure 11.3: Plot of total organic carbon (TOC) versus residual carbon for samples from the central Whitehorse Trough. 'Low-grade' samples form a subset of the Inklin Formation samples, for which S2 > 0.2 and T,,, < 480 'c.....
168 Figure 11.4: Plot of hydrogen index versus oxygen index for samples from the central Whitehorsc Trough. 'Low-grade' samples form a subset of the Inklin Formation samples, for which S2 > 0.2 and T,,, < 480 OC. The organic matter type in a source rock can bedetcrmined from this plot (Espitalii et al., 1977).
...
170 Figure 1L5: Plot of hydrogen index versus T,,, for samples from the central Whitehorse Trough. 'Low-grade' samples form a subset of the lnklin Formation samples, for which S2 > 0.2 andT
,
,
,
c 480 'c. Due to low S2 values, T,,, values for the majority of other lnklin Formation samples are suspect and are dominantly overmature...
172 Figurc 11.6: Contoured thermal maturation map for the central Whitehorse Trough based on T,,, values. Note: T,,, values are poorly constrained for relatively 'high-grade' samples due to thc low (< 0.2) S2 values; dcspite this, all samples with Sz values > 0.03 were plotted as a crude representation of maturation levels in the region. Vitrinite reflectance data shown in Table 11.1....
173ACKNOWLEDGEMENTS
First and foremost, thanks to Steve Johnston and Mitch Mihalynuk for conspiring to drag me across the pond. Your support, encouragement, hard work and good humour will be remembered always. It wouldn't be possible without you. Thanks to the rest of my committee, Dante Canil, Kathy Gillis and Peter Keller, for discussion and for keeping me on the straight and narrow. Thanks to Brendan Murphy. Thanks also to Kelin Wang, Martin Fowler and Gary Johannson for helping to shape the ideas presented within. The support of everyone in ORCa and the School of Earth and Ocean Scicnces at the University of Victoria, in the British Columbia Geological Survey, in the Pacific Geosciencc Centre, in the Geological Survey of Canada and in the Yukon Geology Program is greatly appreciated. This research was funded by the B.C. Ministry of Energy and Mines, hy the Geological Survey of Canada, by a University of Victoria Fellowship and by two Geological Society of America Research Grants.
My research brought me to Atlin in northwest B.C., one of the most beautiful spots in the world, and many friends shared and added to the experience: Fabrice Cordey, Yann Merran, Kyle Larson, Fionnuala Devine, Lucinda Leonard, Adam Bath, Jacqueline Blackwell, Kara Wight, Lee Ferreira, Brian Grant, Norm Graham and Oliver Roenitz.
Thanks to the Tip of the Diapir for the sessions, to Sussi for keeping the cheques and the deadlines coming and to Kaesy for his eternal optimism above all else. Thanks to Lucinda for getting me here; it will always be appreciated. To all other friends I met along the way.
Thanks to my old friends from Trinity College Dublin for instilling a zest for geology and life in general. Thanks to my parents, Matty and Kathleen, for never questioning why. To my late brother and friend John; seven years have passed and still you inspire me. Finally to the rest of my enormous family, to my old friends back on the sod and to Kara for continuous laughter and support, and for lighting the fire within. I thank you all.
CHAPTER 1
Introduction
Continental crust can be distinguished from oceanic crust by its more evolved and differentiated composition, and its enduring nature as evidenced by comparison of the oldest age of continental and oceanic crust (Archean versus Jurassic respectively). However, the rate and mechanisms of continental growth remain controversial (see Rudnick, 1995; Taylor and McLennan, 1995 for review). Continents are believed to grow through processes active at two distinct plate tectonic settings: intraplate and convergent margins. Continental growth may be achieved at convergent margins by accretion of intraoceanic island-arcs or by direct addition of mantle-derived magma during continental-arc magmatism. Continental growth may be achieved at intraplate settings through the eruption of continental flood basalts and by magmatic addition attributable to decompressive mantle melting during rifhng. An intermediate process involves the intraplate development of anomalously thick oceanic plateaux that may be accreted to the continent during collisional orogenesis at a convergent margin. Conversely, continental crust may be removed and recycled back into the mantle during lower crustal delamination in collision zones or during tectonic erosion of the forearc and sediment subduction at convergent margins. The issue of overall net continental growth - the
balance between crustal addition and crustal recycling - will not be addressed herein, and
the term 'continental growth' is used to refer solely to crustal addition and not crustal recycling.
Continental growth
Continental growth is a process that involves: (1) the accretion of material to the continent, and (2) deformation and metamorphism of the added material through orogeny. Both the addition of material to the continent and subsequent orogeny are
accomplished through processes active in convergent margins; these tectonic processes are the focus of this thesis.
Accretion
Accretionary processes believed to add material to continents at a convergent margin include the collision of island arcs with passive continental margins attached to a subducting oceanic slab and the collision of unsubductable crustal blocks femed in on the subducting slab conveyor belt. Island arcs can become tectonically accreted to buoyant continental crust that has been dragged into a subduction zone by an attached slab, leading to arc-continent collision (e.g. the Banda Arc is currently being thrust onto the continental margin of Australia at Timor Island). Subduction polarity can reverse and initiate on the oceanward side of a newly accreted terrane if the tectonic forces that led to ocean closure remain in place after collision (e.g. Hamilton, 1988). An alternative or additional process for continental accretion at a convergent margin involves the collision of oceanic plateaux (e.g. Abbot, 1996; Ben-Avraham et al., 1981; Saunders et al., 1996). Oceanic plateaux are thicker than normal oceanic crust formed at mid-ocean ridges, and are, therefore, buoyant and more difficult to subduct.
The northern Canadian Cordillera includes a central Intermontane belt of arc and oceanic lithotectonic packages, and provides us with an opportunity to investigate and
test the links between convergent margin processes and continental growth due to accretion (Fig. 1.1). Chapter 2 focuses on the northern Cache Creek terrane, one of the main crustal components that constitutes the Intermontane belt, as a case study for assessing accretionaq processes that involve the transfer of material from a subducting plate to an upper plate; the northern Cache Creek terrane includes a subduction complex that records the existence of a Late Paleozoic - Mesozoic ocean basin. Investigation of
the northern Cache Creek terrane reveals that accretion in the northern Intermontane belt occurred principally by the accretion of island arcs and emplacement of forearc ophiolites during collisional orogenesis. The transfer of oceanic sediments and the upper portions of oceanic seamounts from the subducting plate to a subduction complex accounts for only small volumes of growth of the upper plate.
Orogeny
Convergent margin processes thought to result in orogeny (crustal thickening through deformation) include: (1) introduction of an unsubductable, buoyant lithospheric block (typically continental lithosphere, but can also be thickened oceanic lithosphere, such as an oceanic plateau or island arc) into a subduction zone by the consumption of an attached downgoing slab, resulting in collision with the upper plate, (2) collision of an island arc with a passive continental margin, and (3) flat-slab subduction.
The northern Intermontane belt was involved in a Middle Jurassic orogenic event as indicated by the development of a fold and thrust belt. Chapter 3 documents the development of this fold and thrust belt by outlining the stratigraphy and structure of the central Whitehorse Trough (Fig. 1.1). Chapter 4 investigates the origins of this orogenic
Post emplacement Mtd Jurarrtc plut~nr
0
Kutcho Assemblage. L Permian - E Triassic arc cache creek Cornpier. mirtnly hemlwlagltr and basnllCache Creek Complex. alplne ultramehte
Figure 1 . I : Map o f British Columbia showing the distribution of the primary conlponents of the Intermontane belt i n the northern Canadian Cordillera (top right) and a regional geologic map (main). This geology map does not include post-Middle Jurassic rock units. The study area is outlined by a dashed box.
event and attributes orogenesis to the collision and accretion of an island arc. Therefore, collisional orogenesis has been an important. crustal growth process in the North American Cordillera.
The terrane concept was developed for, and widely applied to, Cordilleran-type orogens (Coney et al., 1980), and states that continents grow at accretionary margins by the collision of thickened, buoyant lithospheric blocks, including island arcs, oceanic seamounts, plateaux and aseismic ridges, embedded in subducting oceanic lithosphere. Some of these exotic or far-travelled terranes are commonly inferred to be thin thrust sheets, < 10 km thick, with no deep crustal or mantle roots (e.g. Cache Creek terrane in Snyder et al., 2002). The terrane concept has, however, been criticised for giving rise to a focus on the differences between adjacent crustal packages, at the expense of searching for possible plate tectonic links between disparate assemblages (Hamilton, 1990). For example, accretionary wedge assemblages are commonly included in terranes separate from adjacent magmatic arc terranes. The terrane concept has, therefore, resulted in an over-emphasis on the lithological and structural coherence of crustal packages at the expense of plate tectonic analyses in which crustal packages are placed into a plate tectonic framework.
As much of the material in Cordilleran orogenic belts has been tectonically accreted by tectonic processes active in convergent margins, Chapter 4 provides an interpretation that places the lithotectonic assemblages of the northern Intermontane belt into a plate tectonic framework. This plate tectonic framework has important implications for Cordilleran crustal structure, ophiolite emplacement, continental growth, and the terrane concept in general.
Not all orogenic events in the North American Cordillera can, however, be readily attributed to the collision and accretion of an island-arc or oceanic plateau. The Laramide orogeny is the Late Cretaceous to Paleocene (80 to 55 Ma) orogenic event that gave rise to parts of the Brooks Range in Alaska, the MacKenzie and Rocky Mountain fold and thrust belts in Canada, the Laramide block uplifts in the United States, and the Sierra Madre Oriental fold and thrust belt in Mexico (Fig. 1.2). This orogenic event is believed to post-date the Jurassic and late-Early Cretaceous accretion of the terranes that make up much of the North American Cordillera. A collisional origin for Laramide orogenesis has, therefore, been ruled out (Dickinson and Lawton, 2001; Dickinson and Snyder, 1978; Monger and Nokleberg, 1996; Monger et al., 1982). The leading model for driving Laramide orogenesis in the United States is flat-slab subduction, whereby stress coupling of a subhorizontal oceanic slab to the upper plate transmitted stress eastwards, producing basement-cored block uplifts and arc magmatism in the foreland (Bird, 1988; Coney and Reynolds, 1977; Dickinson and Snyder, 1978). Chapter 5 tests the flat-slab subduction hypothesis for the Laramide orogeny through thermal modelling of a flat-slab configuration. These thermal models indicate that inboard arc magma generation at significant distances inboard from the trench (> 600 km) during flat-slab subduction is problematic; this conclusion is consistent with the coincidence of volcanic gaps and flat- slab subduction at modem convergent margins. P-T constraints from lawsonite eclogite inferred to originate from the subducted Farallon slab (Usui et al., 2003) indicate that the Laramide flat-slab subduction zone was characterised by a cold thermal regime, which may have been produced by subduction of an old oceanic plateadaseismic ridge.
Figure 1.2: Map showing areas of pre-Laramide and Laramide magmatism, the extent of the Laramide-age thin-skinned fold-and-thmst belt and thick-skinned block uplifts, and the approximate location of the Laramide flat-slab according to Saleehy (2003). Note that Cenozoic extension has not been restored in this figure. This figure was drafted using
The Laramide orogeny is unique in that it is believed to have produced a deformational belt along the length of the North American Cordillera 700 to 1500 km from the nearest convergent margin. Although the flat-slab subduction model has been applied to Laramide orogenesis in the United States, the development of a magmatic arc within 300 km of the trench in Canada and Mexico during Laramide time refutes the existence of flat-slab subduction in these regions (Fig. 1.2). Flat-slab subduction cannot, therefore, account for Laramide orogenesis along its entire length. Chapter 6 reviews proposed mechanisms for producing this inboard deformation, including retroarc thrusting, 'orogenic float' tectonics, flat-slab subduction and Cordilleran collision, and illustrates that the processes responsible for this orogeny remain enigmatic and require resolution.
Primary objectives
The primary objectives of this thesis are to:
assess the accretionary processes that involved the transfer of material from the subducting Cache Creek oceanic plate to an upper plate at a convergent margin identify the principal mechanisms responsible for tectonic accretion in the northern Intermontane belt
document a Middle Jurassic fold and thrust belt in the northern Intermontane belt where it is best exposed in the southern Atlin Lake and Taku River areas and elucidate the driving force of this orogenic event
place the various lithotectonic assemblages of the northern Intermontane belt into a plate tectonic framework and discuss the implications for: (a) Cordilleran crustal structure, (b) ophiolite emplacement, (c) continental growth, and (d) the terrane concept
use thermal models to: (a) assess the feasibility of arc magma generation at significant distances inboard from the trench (> 600 km) during flat-slab subduction, and (b) determine the cause flat-slab subduction by comparing the thermal regimes of various flat-slab models with published P-T conditions from eclogite xenoliths
provide a modem review of the proposed mechanisms for producing the inboard deformation belts of the Laramide orogeny, illustrate that the processes responsible for orogeny remain enigmatic, and outline major questions regarding this orogenic event that require resolution
Contributions
The specific contributions of the author to the following chapters are-
Chapter 2: a principal researcher on a geological bedrock mapping program for the Nakina area; responsible for the collection and preparation of lithogeochemical samples, petrographic descriptions, interpretation of lithogeochemical data, and discussion of implications for accretionary processes. This work indicates that the oceanic crustal rocks of northern Cache Creek terrane are of arc-affinity and not part of an oceanic plateau as previously interpreted. Chapter 3: chief geologist in charge of the geological bedrock mapping program for the Atlin Lake area; responsible for the collection of structural data, and the production of structural cross-sections. This work indicates that the Whitehorse Trough is extensively thrust-faulted, an observation not shown on previous bedrock geology maps of the study area.
Chapter 4: responsible for interpretation and discussion of implications of northern Intermontane belt geology for Cordilleran crustal structure, ophiolite emplacement, continental growth and the terrane concept.
Chapter 5: compiled geological constraints on Laramide flat-slab subduction from the published literature; constructed, ran and interpreted thermal models; responsible for the discussion of the implications of these models for the Laramide flat-slab subduction hypothesis.
Chapter 6: reviewed the proposed mechanisms for producing the inboard deformation belts of the Laramide orogeny; responsible for the discussion of the major questions regarding this orogenic event that require resolution.
Methodology
Geological bedrock mapping at 1:50 000 scale in the Nakina area, led by Mitchell Mihalynuk, was completed during the months of July and August in 2001 and 2002. Additional principal researchers included Stephen Johnston, Fabrice Cordey and Yann Merran; geological assistants included Adam Bath, Jacqueline Blackwell, Fionnuala Devine, Kyle Larson, Lucinda Leonard and Oliver Roenitz. The study area covers NTS mapsheets 104N/1, 2 and 3, an area of approximately 2400 km2 (28 km from north to south and 84 km from east to west). Access to the Nakina area was achieved using helicopter charter based out of Atlin. Traverses were carried out on foot from a base camp that was moved every three to four days. Base camp locations were chosen to provide opportunities to locate and follow lithological contacts while maximising coverage of the area. Geological bedrock mapping in the Atlin Lake and Taku River areas was carried out by myself and an assistant (Kara Wight) during the months of July and August in 2003. Traverses were conducted from the shores of Atlin Lake, which was accessed by zodiac from a base camp on southern Teresa Island. Additional mapping further to the southeast in the central Whitehorse Trough in 2003 was achieved by helicopter-supported base camps. During all fieldwork, lithological and structural data, as well as samples for later laboratoly analyses, were collected at stations taken along the traverse. Mapping coverage is indicated by the density of stations within the study area (Fig. 1.3). Samples for thin-section were selected and analysed by myself.
Chapter 2 is based on geological mapping in the Nakina area. Samples for geochemical analysis were selected and prepared by myself; methodology is outlined in Chapter 2. Chapter 3 is based on geological mapping in the central Whitehorse Trough in
Figure 1.3: Distribution of field stations taken within thestudy area in the northern Canadian Cordillera. See Figure 1.1 for regional location map.
the Atlin Lake, Nakina and Taku River areas. The geological map of the central Whitehorse Trough (Fig. 3.4) incorporated stratigraphic, structural and fossil data from unpublished geological maps of the Atlin Lake area produced by Gary Johannson. Subsurface interpretations (shown in cross-section diagrams in Fig. 3.5) are inferred from structural data and field relationships. Chapter 5 is based on computer modelling of the thermal regime of flat-slab subduction. This modelling was carried out by myself under the direction of Kelin Wang and the methodology is outlined in Chapter 5. Appendix 11 is based on a geological study of the central Whitehorse Trough. Samples for programmed pyrolysis were selected and prepared by myself These samples were sent to the Geological Survey of Canada (Calgary) for analysis under the direction of Martin Fowler; methodology is outlined in Appendix 11. Petrographic descriptions and assessment of the reservoir potential of the central Whitehorse Trough is based in part on the work of Gary Johannson (M.Sc. thesis).
References
Abbot, D. H., 1996, Plumes and hotspots as sources of greenstone belts: Lithos, v. 37, p. 113-127.
Ben-Avraham, Z., Nur, A,, Jones, D., and Cox, A,, 1981, Continental accretion: From oceanic plateaus to allochthonous terranes: Science, v. 213, p. 47-54.
Bird, P., 1988, Formation of the Rocky Mountains, western United States: A continuum computer model: Science, v. 239, p. 1501-1507.
Coney, P. J., Jones, D. L., and Monger, J. W. H., 1980, Cordilleran suspect terranes: Nature, v. 288, p. 329-333.
Coney, P. J., and Reynolds, S. J., 1977, Cordilleran Benioff zones: Nature, v. 270, p. 403- 406.
Dickinson, W. R., and Lawton, T. F., 2001, Carboniferous to Cretaceous assembly and fragmentation of Mexico: Geological Society of America Bulletin, v. 113, p.
1142-1 160.
Dickinson, W. R., and Snyder, W. S., 1978, Plate tectonics of the Laramide Orogeny: in
Matthews, V., ed., Laramide folding associated with basement block faulting in the western United States: Memoir: Denver, Co., Geological Society of America, p. 355-366.
Hamilton, W. B., 1988, Plate tectonics and island arcs: Geological Society of America Bulletin, v. 100, p. 1503-1527.
Hamilton, W. B., 1990, On terrane analysis: in Dewey, J.
F.,
Gass, I. G., Cuny, G. B., Hanis, N. B. W., and Sengor, A. M. C., eds., Allochthonous terranes: London, Royal Society of London, p. 55-66.Monger, J., and Nokleberg, W. H., 1996, Evolution of the northern North American Cordillera: generation, fragmentation, displacement and accretion of successive North American plate margin arcs: in Geology and ore deposits of the American Corddlera, Nevada, p. 1133-1 152.
Monger, J. W. H., Price, R. A,, and Tempelman-Kluit, D. J., 1982, Tectonic accretion and the origin of the two major metamorphic and plutonic welts in the Canadian Cordillera: Geology, v. 10, p. 70-75.
Rudnick, R. L., 1995, Making continental crust: Nature, v. 378, p. 571-578.
Saleeby, J., 2003, Segmentation of the Laramide slab
-
evidence from the southern Sierra Nevada region: GSA Bulletin, v. 115, p. 655-668.Saunders, A. D., Tamey, J., Kerr, A. C., and Kent, R. W., 1996, The formation and fate of large oceanic igneous provinces: Lithos, v. 37, p. 81-95.
Snyder, D. B., Clowes, R. M., Cook, F. A,, Erdmer, P., Evenchick, C. A,, van der Velden,
A. J., and Hall, K. W., 2002, Proterozoic prism arrests suspect terranes: Insights
into the ancient Cordilleran margin from seismic reflection data: GSA Today, v. 12, no. 10, p. 4-10.
Taylor, S. R., and McLennan, S. M., 1995, The geochemical evolution of continental crust: Reviews of Geophysics, v. 33, p. 241-265.
Usui, T., Nakamura, E., Kobayashi, K., Maruyama, S., and Helmstaedt, H., 2003, Fate of the subducted Farallon plate inferred from eclogite xenoliths in the Colorado Plateau: Geology, v. 3 1, p. 589-592.
Wessel, P., and Smith, W. H. F., 1995, New version of the Generic Mapping Tools released: Eos, v. 76, F329.
CHAPTER
2
Accretionary Processes in the Northern
Cache Creek Terrane, Canadian Cordillera
Introduction
Accretionary processes involve the transfer of material such as oceanic sediments, seamounts, island arcs, oceanic plateaux and microcontinental fragments from a subducting plate to an upper plate. Evidence of accretion tectonics can be observed at many of the world's active continental margins, as well as along the suture zones of ancient orogenic belts, and this process is an important mechanism for continental growth. Accretion occurs by two principle mechanisms: (1) accretion of island arcs as a continental plate enters and blocks up a subduction zone (e.g. the Banda Arc is currently being thrust onto the continental margin of Australia at Timor Island), and (2) transfer of oceanic sediments, seamounts and ridges from the subducting plate and incorporation into the subduction complex (Cloos, 1993). The northern Cache Creek terrane in the Canadian Cordillera includes a subduction complex that records the existence of a Late Paleozoic - Mesozoic ocean basin (Monger, 1975; Cordey et al., 1991; Orchard, 1991;
Fig. 2. l), and hence provides an opportunity to assess the preservation/accretion potential of various tectonic elements within the oceanic realm.
The objectives of this paper are to:
(a) elucidate the original tectonic setting@) in wh~ch the mafic rocks of the northern Cache Creek terrane were produced, and
(b) discuss the significance of the northern Cache Creek terrane as an analogue for accretionruy processes, the emplacement of ophiolitic assemblages and continental growth at convergent margins.
I/C1
L T i ~ a ~ ~ ~ c - E J U ~ ~ S S ~ C an. ~lastlcWtonrc0
PaieoiUlc strate 2nd &formed Paieozoc arcFigure 2.1 : Map orBritish Columbia showing the distribution oTthe primary components o r the Intermontane belt in the northern Canadian Cordillera (top right) and a regional geologic map (main). This geology map does not include post-Middle Jurassic rock units. The Nakina area is outlined by a dashed box.
Geological Background
The Cache Creek terrane is a belt of Mississippian to Lower Jurassic oceanic rocks (Monger, 1975; Cordey et al., 1991; Orchard, 1991) that occupies a central position within the Intermontane belt of the Canadian Cordillera in British Columbia (B.C.) (Coney et al., 1980; Fig. 2.1). Fossil fauna within carbonate rocks of the Cache Creek terrane are uniquely exotic with respect to the remainder of the Canadian Cordillera as they are typical of the equatorial Tethyan realm, contrasting with coeval faunas in adjacent arc terranes of Stikinia and Quesnellia (Fig. 2.1) that show closer linkages with ancestral North America (Monger and Ross, 1971; Orchard et al., 2001). Rocks comprising the Cache Creek terrane represent two distinctive lithotectonic elements: a Wddle Triassic to Lower Jurassic, subduction-related accretionary complex, and a dismembered oceanic basement assemblage (Terry, 1977; Monger et al., 1982; Ash, 1994; Mihalynuk, 1999). In map view, the northern Cache Creek terrane (N of 58O) is a southeast-ward tapering wedge (Fig. 2.1); this terrane is comprised of tectonically imbricated slices of chert, argillite, volcaniclastic rocks, carbonate and wacke, with a belt of ultramafics, gabbro and basalt along its western margin (Aitken, 1959; Mihalynuk et al., 2002; Mihalynuk et al., 2003a; Figs. 2.1 and 2.2). If the mapped area of exposed units in the northern Cache Creek terrane is representative of the volume of each lithology, the following estimates can be made: ultramafic rocks 7%, gabbroic intrusive rocks 2%, volcanic strata 15%, carbonate strata 6% and hemipelagite/sililiclastic strata 69% (Table 2.1). Rocks of the northern Cache Creek terrane were emplaced to the west over the Stikine terrane in the Middle Jurassic resulting in fold-and-thrust belt formation in the Whitehorse Trough (Thorstad and Gabrielse, 1986; Mihalynuk, 1999; English et al., in
Bedrock
Geology
of
the
Cache
Creek
terrane
in
the
Nakina
Map-area
L'thogeochemica sample sites Early Eocene Sloko Group-
Inyoice to bssall tuff and ksserflow rocks Hem~peiagile~rliciciartic un~t: eneR manly rbboned. Late Caiboniferousto Late Tilarslc Middle Jurassic inlruriens Hornblende~biot8te granodicdte Carbonlkrour - Lower Jurassic Cache Cmekcompier Carbonate unn: carbonate, rnatny Mande unn: harzbur@te rectonite Mddiecarbonlferousio Mdde Perman and rementin~te mClange Camonate unri: Argdlife and MaRc ntrusve unit: gabbro, dime. arg~lleeeovs oiivell beddedcarbonare0
qquamdionte, mnoi piagiogrmte Carbonale unl: Augite porphyv MaRc volcanic unr: baram tun. and related rocks0
o~llow bssaik. volcanciarr~c mcks Figure 2.2: Generalised bedrock geology map of the northern Cache Creek terrane in the Nakina map-area with a 10 km UTM grid superimposed (NTS 104N/1,2 and 3).Table 2.1 : Normalised percentages of the mapped areas of each unit in the northern
Cache Creek terrane (based on Mihalynuk et al., 1996; Massey et al., 2003).
Unit
Mantle unit Mafic intrusive unit Mafic volcanic unit HernipelagitelSililiclastic unit
Carbonate unit
Liihologies
ultramafic rocks, harzburgite, serpentinite gabbroic to dioritic intrusive rocks basaltic volcanic and volcaniclastic rocks
chert, siliceous argillite, sililiclastic rocks limestone, marble, calcareous sedimentary rocks
% of area 7 2 15 69 6
review); the timing of this deformational event is constrained to be younger than the age of the youngest blueschist in the Cache Creek terrane (French Range, -174 Ma: Mihalynuk et al., in press), and older than the age of the oldest post-deformational intrusions (-172 Ma: Mihalynuk et al., 1992; Mihalynuk et al., 2003a; Bath, 2003). Post- collisional chert-pebble conglomerates derived from the Cache Creek terrane were deposited across the Whitehorse Trough and Bowser Basin of the Stikine terrane during Bajocian time (Ricketts et al., 1992; Mihalynuk et al., in press) providing an overlap assemblage.
Geology
of
the Nakina Area
In the Nakina area (Fig. 2.1), rocks of the northern Cache Creek terrane can be assigned to one of 5 generalized units: (1) mantle, (2) mafic intrusive, (3) mafic volcanic,
(4) hemipelagite/siliciclastic, and (5) carbonate (Fig. 2.2). A separate mafic volcanic unit with sparse relict pillows is known as the Yeth Creek formation (informal, Mihalynuk et al., 2004). The Yeth Creek formation crops out south of the Nahlin Fault, a crustal scale structure which marks the southwestern limit of northern Cache Creek terrane strata along most of its length (Fig. 2.1). A number of post-deformational Middle Jurassic quartz-dioritic to granodioritic plutons intrude the Cache Creek terrane within the study area (Fig. 2.2). The Nakina area of the Cache Creek terrane has been metamorphosed to prehnite-pumpellyite facies, although thermal upgrading to biotite-grade is observed around large plutons.
Mantle rocks
Tectonised harzburgite, composed of olivine and orthopyroxene with accessory clinopyroxene and chromite, forms a coherent 1.5 x 15 km, dun-weathering body along the southwestern margin of the Cache Creek terrane in the Nakina area (Fig. 2.2). Orthopyroxene and chromite grains form streaky clusters outlining a high temperature fabric interpreted as having a mantle origin; lineated orthopyroxene grains are up to 2 cm in length. This mantle fabric is cut by
-
2 - 4 mm thick pyroxenite dikelets. There is no deformational fabric that post-dates the intrusion of the pyroxenite dikelets, indicating that strain was highly partitioned and that the harzburgite acted as a rigid body during emplacement.North of the Nakina River, this harzburgite body is structurally disaggregated, and comprises the dominant knocker type within the western part of a serpentinite melange belt. Gabbro and hornblende diorite knockers are subordinate to the harzburgite, but increase in abundance to the east. The knockers within the serpentinite melange, which can be up to several square kilometres across in map-view, are lithologically dominated by basalt and chert along the eastem flank of the m6lange belt. The age of these ultramafic rocks remains unconstrained: a UPb zircon age of 245.4
*
0.8 Ma has been reported from peridotite in the northeastern part of the Cache Creek terrane (Gordey et al., 1998). The relationship between this peridotite body and the ultramafic rocks described here is unknown.Mafic Intrusive rocks
Medium-grained plagioclase and pyroxene gabbro and amphibole-phyric diorite occur as isolated intrusions and as knockers within serpentinite melange in the Nakina area. Primary clinopyroxene is subhedral to euhedral and commonly displays exsolution lamellae and zoning. In more altered samples, especially within the serpentinite melange belt, primary pyroxene and hornblende are preferentially altered to actinolite; secondary minerals mainly form pseudomorphs after the igneous minerals that they replaced. Plagioclase is variably altered, from fresh to totally turbid; authigenic minerals include white mica, prehnite, quartz and carbonate. Pyroxenehornblende is subequal in abundance to plagioclase. Ophitic textures are common.
Some of these intrusive rocks are intensely sheared while others are internally undeformed. A folded, sheet-like gabbroic body displays an intrusive contact with overlying mafk volcanic rocks and fault contact with underlying ultramaf~c, mafic and quartz-rich clastic rocks south of Mount O'Keefe. Gabbroic rocks in the Hard Luck Peaks area display an intrusive contact with basaltic rocks. Hornblende-phyric diorite, gabbro and tonalite knockers within the serpentinite melange belt range in size from a metre to hundreds of metres; these knockers are undeformed compared to the serpentinite matrix. Tonalite and quartz-diorite knockers from the serpentinite melange have produced mid- late Permian ages (Devine, 2002; samples FDE01-3 1-7 and FDE01-3 1-12).
In the southeastern part of the study area, diorite is intruded by an irregular network of comagmatic pegmatitic tonalite dykes less than 0.5 m thick. These tonalite dykes are composed of plagioclase and hornblende with about 10% interstitial quartz, and accessory titanite and zircon. Plagioclase has been partially altered to prehnite, while
homblende only displays traces of chlorite alteration. A 40Arl39Ar age determination (using laser step-heating analysis) on the homblende is consistent with that of extracted zircons which reveal a Upper Permian U/Pb age (261.4 i 0.3 Ma, Mihalynuk et al.,
2003a; sample MM101-27-6); this age is consistent with U-Pb zircon ages of magmatic knockers within the serpentinite melange (Devine, 2002). These ages are similar to one from similar rocks in the southern Cache Creek terrane (257
*
5 Ma: Schiarizza et al., 2000). Hence, it appears that the oceanic crust that is preserved in the Cache Creek terrane was produced during Upper Permian time.Mafic Volcanic rocks
Basalt and mafic volcaniclastic rock is the dominant volcanic unit within the Nakina area. This unit commonly displays well-preserved, aphanitic lapilli and ash-sized fragments despite widespread replacement by prehnite, pumpellyite, calcite and chlorite. Fresh surfaces are a distinctive mint green colour with a grey or pinkish hue caused by filaments (fine shear bands) of clay and iron oxides. Weathered surfaces display a fragmental texture with angular clasts up to 10 cm in size; these clasts commonly contain relict plagioclase and pyroxene phenocrysts. Both the pyroxene and plagioclase range from fresh to extensively altered in thin-section. Locally, pyroxene phenocrysts are euhedral. Disseminated pyrite, pyrhhotite and minor chalcopyrite are common, but comprise less than 1-2% of the unit. Interbeds of chert are Middle Triassic in the northern part of the Nakina area (Mihalynuk et al., 2003a), and Permian in the Hard Luck Peaks area (Mihalynuk et al., 2003b). Extensive exposures of well-formed pillows near Hard
Luck Peaks are fine-grained, vesicular, and in some cases contain medium-grained feldspar laths comprising up to 10% of the rock.
Hemipelagite/Silicklastic rocks
Chert, argillite and fine-grained wacke dominate the northern Cache Creek terrane (69%, Table 2.1). In the Nakina area, chert varies in colour from black to grey to tan and occurs as either massive, featureless argillaceous chert successions or well-ribboned 2-6 cm thick beds of chert with 2-5 mm dark grey argillaceous partings. Chert is commonly interbedded with mafic volcaniclastic strata or wacke containing a large proportion of volcanic quartz grains. Radiolaria extracted from chert in the Nakina area range in age from Permian to Upper Triassic (Mihalynuk et al., 2003a); elsewhere in the northern Cache Creek terrane, radiolaria range in age from Carboniferous to Lower Jurassic (Cordey et al., 1991).
Siliceous mud-rich wacke in the Nakina area is brown and less commonly dark grey, black or blue-grey. This wacke commonly contains chert grains and rare cobbles, and volcanic clasts from ash to lapilli size. Sparse quartz grains may be derived from quartz diorite, clasts of which occur as rare, foliated granules. Locally, these strata grade into chert or volcaniclastic rocks. Some wacke units are dominated by sand or silt-sized grains. Detrital zircons from wacke in the Nakina area are as young as Lower Jurassic (182 i 4 Ma; Mihalynuk et al., 2003a), although most detrital ages are Middle to Upper
Cadonate rocks
Carbonate rocks comprise much of the central part of the Nakina area (Fig. 2.2). Typical lithologies include light grey, well-bedded, bioclastic (turbiditic?) limestone; coarse limestone breccia which occurs as sheets and channels decimetres to several metres thick; indistinctly thick to medium-bedded, cream-coloured limestone 80-120 m thick; and distinctly medium to thin-bedded, dark grey to black, fetid andor argillaceous limestone
-
40 m thick. Massive limestone is the most abundant unit and many cubic kilometres are featureless except for sparse crinoid ossicles and fragments of bivalves, bryozoa rare corallites, pisoids or limestone clasts. Massive limestone probably accumulated in an intra-oceanic platformal setting during Upper Carboniferous and Permian time and interfingered with well-bedded lagoonal facies and talus to turbidite facies on the platform margins (Monger, 1975; Merran, 2002; Mihalynuk et al., 2003a).Isolated volcanic accumulations occur within carbonate units, and the volcanic rocks appear to form the substrate of carbonate reefs (Merran, 2002; English et al., 2002). Flows are brown-weathering, and pinkish maroon fresh, with pillows containing zones of elongate, calcite-filled vesicles and separated by interpillow hyaloclastite. Rounded to angular carbonate blocks comprise irregular interlayers within the volcanics; they are interpreted to be of olistostromal origin. The most extensive framework reef constructed on an accumulation of volcanic rocks occurs on the southwest flank of 'Sideout Mountain' (Fig. 2.2; Monger, 1977; Merran, 2002), where Upper Mississippian reefAagoona1 carbonates were deposited on volcanic breccia.
Feldspar porphyry, in assumed stratigraphic contact with overlying Carboniferous well-bedded limestone, occurs as pillowed flows and volcaniclastic layers less than 10 m
thick. An augite porphyry unit (Fig. 2.2) occurs as a laterally extensive, monomictic matrix-supported breccia reaching up to 250m in thickness, and containing blocks that are up to 1 m across. Interlayered tuffs contain as much as 30% zoned euhedral augite crystals that are up to 1 cm across, as well as less abundant plagioclase phenocrysts. Groundmass is dominated by plagioclase and Fe-oxides. In the upper part of this succession, the volcaniclastic sequence becomes more calcareous and grades upwards into a sequence of limestones.
Yeth Creek formation
Pillow basalts are exposed on the southwestern side of the Nahlin Fault in the Yeth Creek area (Fig. 2.1). Fine-grained, dark green to grey, vesicular pillow basalts, breccia and sheet flows are the principal lithologies. These volcanic rocks are typically pyroxene and plagioclase-phyric; plagioclase crystals are moderately altered and pyroxene crystals are fresh. One interpretation is that the pillow basalts of the Yeth Creek area are basement to the overlying sililiclastic sediments of the Laberge Group in the Whitehorse Trough (Souther, 1971). This stratigraphic relationship was not confirmed, but the basalt was sampled for lithogeochemical analysis.
Composition
of
Mafic Rocks
Samples of 59 intrusive and extrusive mafic rocks were analyzed for major and trace element abundances. Isotropic gabbro samples were collected to avoid obvious cumulate layers. All mafic units described above are represented in the suite analyzed.
Methods
Samples were prepared using steel jaw crusher and disk mill at the British Columbia Ministry of Energy and Mines rock preparation facility. Tungsten carbide grinding surfaces were not used in order to minimize trace element contamination, particularly Ta and Nb. Major oxides were determined for 18 samples by LiB02 fusion and ICP-ES analysis at ACME Analytical Laboratories, Vancouver. Minor and tracehare earth element geochemistry was determined on all 59 samples by ICP MS techniques, at ACME Analytical Laboratories and Memorial University, Newfoundland. Resultant geochemical data are presented in Appendix I. A key to the symbols used in the various geochemical plots is provided in Figure 2.3; this classification is based on a complete analysis of the geochemical data.
The primary goal of this geochemical analysis is to determine the magma source and, hence, the paleotectonic environment of the igneous rocks. The success of this analysis hinges on obtaining an original geochemical signature that has not modified by subsequent hydrothermal alteration and metamorphism. Major elements, especially alkalis, should be considered mobile during alteration and metamorphism ( e g Smith and Smith, 1976), although major element compositional classifications can be compared with immobile element compositional classifications in order to test major element
Leaend
Oceank crustal assemblage Carbonate unR
Island-arctholeiites 0 Ocean-island type
X Calc-alkaline Global rveragsdmodern analogues
A Back-arc basin basalt
.
IAT(Jennereld.. 1987)0 E-MORBlype A N-MORB (sun and ~ c m a u g h . 7989)
Yeth Creek formation
+
E-MORE (Sun arm Mcmoouph, 7989)+
Back-an bas" basalte 0 OiB (Sunarm McDonough. 1980)0 35 45 55 65 75 SiO, 5 , . . . . , ,
:
B
1 :-
0 0X
0.1 i 0- :5
0.01 ; AndesiteiBasalt Alkali 0 Basalt Subalkaline Basalt N b NFigure 2.3: Nakina area lithogeochemical data plotted on rock classification diagrams. Volcanic rocks from the carbonate unit are alkaline, and range from alkali basalts and basanites to trachytes, while rocks from the oceanic crustal assemblage are subalkaline and dominantly basaltic and basaltic-andesitic in composition. (a) Na,O+K,O versus SiO, (TAS) diagram (from Cox et al., 1979). Abbreviations: P-N phonolite-nephelinite, P-T phonolite- tephrite, B+T basanite
+ tephrite, B-A basaltic andesite. (b) Immobile trace element
abundances are also used for rock classification (from Winchester and Floyd, 1977), particularly in altered volcanic rocks where elemental mobility is suspected. Abbreviations: Bsn/Nph Basanitemephelinite; ComiPant ComenditeIPantellerite.mobility. Low-field strength elements (LFSE) such as K, Rb, Cs, U, Pb, Ba and Sr should be considered mobile, while Th is the only LFSE that can be considered immobile in most circumstances (e.g. Jenner, 1996). High-field strength elements (HFSE) such as Ti, Zr, Hf, Nb and Ta, and rare earth elements (REE) can be considered immobile in most cases (e.g. Jenner, 1996) and, hence, they are the most useful group of elements for determining paleotectonic environments.
Geochemistry
On the basis of field mapping criteria, rocks from the mafic intrusive and mafic volcanic units belong to an oceanic crustal assemblage. From a geochemical perspective, it will be demonstrated that this oceanic crustal assemblage consists of island-arc tholeiites (IAT), island-arc calc-alkaline rocks, back-arc basin basalts (BABB) and enriched mid-ocean ridge basalts (E-MORB). It will also be demonstrated that the Yeth Creek formation consists of BABB and that the carbonate-hosted volcanic rocks consist of ocean-island basalts (OIB) and more evolved alkaline volcanic rocks.
Rock classification diagrams
Analysed samples from the Nakina area are both alkaline and subalkaline (Fig. 2.3A). On the total alkalis versus silica (TAS) diagram of Cox et al. (1979), the volcanic rocks interbedded with Carboniferous carbonate are classified as alkaline, whereas the rest of the samples are subalkaline. Most of the samples reveal a basaltic to basaltic andesite composition. Compositional groupings based on major element profiles are consistent with those based on immobile trace element compositions, although classification varies slightly. For example, rock classification based on the TAS diagram
compares well with that based on the immobile element Zr/TiOz versus NbN diagram
(Fig. 2.3B; Winchester and Floyd, 1977). The only major difference between these plots is that a porphyritic trachyte plots in the subalkaline field (rhyolite) in the major element classification of Figure 2.3A, but in the alkaline field in the trace element classification of Figure 2.3B. This sample is plagioclase-rich, and contains centimetre-sized alkali feldspar phenocrysts, which may skew the analyses due to their high Si and Na contents, and very low Mg, Fe and Ca concentrations.
Th-Hf-Nb discrimination diaaram
The Th-Hf-Nb discrimination diagram (after Wood, 1980) can be used to discriminate between volcanic rocks produced in a suprasubduction zone setting and those produced in a mid-ocean ridge setting, because these elements are less susceptible to mobility during low-grade metamorphism and their concentrations, therefore, relate to original magmatic processes. Arc magmas display depletion in HESE (especially Nb and Ta) coupled with enrichment in LFSE (e.g. Th) relative to other HFSE (Wood, 1980). Calc-alkaline rocks are the common products of arc magmatism and tend to be characterised by a greater enrichment in Th relative to Hf (HE/Th < 3); they can, therefore, be distinguished from more primitive island-arc tholeiites using a Th-Hf-Nb diagram (Wood, 1980). Non-arc rocks can also be classified using this diagram because normal mid-ocean ridge basalts (N-MORB) have low N b contents and high H f m ratios, while OIB have higher
Nb
contents and lower HfINb ratios. Basaltic rocks from marginal basins (e.g. BABB) can plot within the normal mid-ocean ridge field (Pearce, 1996), and hence discrimination between these two tectonic environments may be difficult.Based on the Th-Hf-Nb diagram (Fig. 2.4), magmatic rocks from the oceanic crustal assemblage can be separated into four components: calc-alkaline arc rocks, island- arc tholeiites, N-MORBIBABB and E-MORB. The calc-alkaline arc rocks are dominated by evolved intrusive rocks such as tonalites and hornblende-phyric diorites, while more mafk rocks from the oceanic crustal assemblage are dominantly classified as island-arc tholeiites and BABB (Appendix I). Although the BABB of the oceanic crustal assemblage plot in the N-MORB field (Fig. 2.4), it will be demonstrated that trace element concentrations for these rocks are more consistent with a back-arc setting. Pillow basalts from the Yeth Creek formation are geochemically identical to BABB from the oceanic crustal assemblage of the Cache Creek terrane (Fig. 2.4). Volcanic rocks from the carbonate unit are classified as alkaline within-plate basalts and differentiates (Fig. 2.4).
T W b versus NbNb olot
Trace element ratio plots such as ThIYb versus NbiYb can be used to assess the level of depletion in the mantle source (e.g. Pearce, 1982; Pearce et al., 1995). Each ratio consists of the concentration of a more incompatible element divided by the concentration of a less incompatible element such that previous melt extraction from the mantle source will result in a decrease in both ratios. N-MORB is derived from a depleted mantle source that experienced depletion in incompatible elements during previous melt extraction, whereas OIB is derived from an enriched mantle source (Fig. 2.5). The presence of subduction-derived fluid prior to magmagenesis increases the TWYb ratio, but not the Nb/Yb ratio, in the resulting melt (Fig. 2.5).
Based on the ThA'b versus NbiYb plot, the majority of samples from the oceanic crustal assemblage are characterised by low ThA'b and NbiYb ratios indicating that they
