Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Quaternary volcanism in the Wells Gray-Clearwater area, east central British Columbia Hickson, Catherine Jean 1986

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

Item Metadata


831-UBC_1987_A1 H52.pdf [ 38MB ]
JSON: 831-1.0052659.json
JSON-LD: 831-1.0052659-ld.json
RDF/XML (Pretty): 831-1.0052659-rdf.xml
RDF/JSON: 831-1.0052659-rdf.json
Turtle: 831-1.0052659-turtle.txt
N-Triples: 831-1.0052659-rdf-ntriples.txt
Original Record: 831-1.0052659-source.json
Full Text

Full Text

QUATERNARY VOLCANISM IN THE WELLS CRAY-CLEARWATER AREA, EAST CENTRAL BRITISH COLUMBIA by CATHERINE JEAN HICKSON B. Sc. (Honours Geology) University of British Columbia, 1982 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DECREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Geological Sciences We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA December 1986 © Catherine Jean Hickson, 1986 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the The University of British Columbia, ! agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Geological Sciences The University of British Columbia 2075 Wesbrook Place Vancouver, Canada V6T 1W5 Date: December 1986 ABSTRACT Basaltic volcanism in the form of small-volume, subaerial and subaqueous eruptions have occurred in the Wells Cray—Clearwater area of east central British Columbia. These eruptions have been dated by the K-Ar method and by relationships to dated glaciations. The oldest known eruption may be as old as 3.2 Ma, but is more likely 2 Ma or less. The youngest eruptions are less than 7560 ±110 radiocarbon years. The most extensive basalts are valley-filling and plateau-capping flows of the Clearwater unit, which are Pleistocene in age and greater than 25 km 3 in volume. The deposition of flows of the Clearwater unit has overlapped at least three periods of .glaciation. The interaction of glacial ice and basaltic magma has been recorded in the form of tuyas, ice ponded valley deposits and subglacial mounds (SUGM). In a few place glacial till has been preserved beneath basalt flows. Flows of Wells Gray—Clearwater suite appear to have erupted from vents that are both spatially and temporally separated. The individual eruptions were of low volume (<1km 3) and chemically distinct from one another. Major element composition is variable but the lavas are predominantly alkalic. Olivine is the predominant phenocryst phase. Plagioclase and augitic clinopyroxene rarely occur as phenocrysts, but both minerals are ubiquitous in the groundmass. Orthopyroxene was not seen in any of the samples. Flows appear to have erupted with minimal crystal fractionation or crustal contamination. The range of compositions seen in the suite is best explained by a process of partial melting and the progressive depletion of the mantle source by earlier melts. Progressive depletion of the mantle source was coupled with enrichment of parts of the mantle in K as well as some lithophile and siderophile elements. Increasing alkali content may have triggered the highly enriched eruptions of Holocene age that, despite very low degrees of partial melting, were capable of reaching the surface. Overprinting the effects of partial melting are inherited heterogeneities in the source zone of the magmas. Based on whole-rock chemistry the magma source appears to be a highly depleted region similar to that which produces the most depleted mid-ocean ridge basalts (MORB). The zone is, however, capable of producing large volume (== 15%) partial melts and has not been isotopically depleted to the same extent as MORB source regions. Isotope analyses of 8 7 S r / 8 6 S r , 1 4 3 Nd/ 1 4 4 Nd and whole-rock Pb indicate that the magmas may be derived from a remnant of subducted oceanic lithosphere which has been variously depleted by the prior generation of basaltic melts. Isotopic enrichment above the level seen in MORB's is due in part to crustal contamination. The isotopic results are very different than those obtained from samples erupted through thin, allochthonous crust in the Intermontane Belt and may be explained in part by generation of the magmas in oceanic material which was subducted when allochthonous crust lay against the parautochthonous rocks underlying the Wells Cray—Clearwater area. The alkali olivine basalts of the Wells Cray—Clearwater area have erupted onto a tectonically active surface. A peneplain (erosion surface), formed in Eocene-Miocene time has been uplifted since the Miocene and uplift may be continuing. This uplift is in response to an elevated geothermal gradient which may be due to crustal extension. This crustal extension may be similar to that which occurred in the Eocene. The elevated geothermal gradient and reduced pressures attendant with recent uplift and erosion may have initiated basaltic volcanism in the region, rather than a fixed mantle hot spot as proposed in earlier work. i i i Table of Contents ABSTRACT i i F O R W O R D x i v A C K N O W L E D G E M E N T S x i i i 1. INTRODUCTION 1 2. PHYSIOGRAPHY, GEOMORPHOLOGY and QUATERNARY GLACIAL RECORD 5 2.1 PHYSIOGRAPHY 5 2.2 GEOMORPHOLOGY 8 2.2.1 Peneplanation 8 2.2.2 Development of the drainage system 9 2.2.3 The Clearwater depression 12 2.3 GLACIATIONS 14 2.3.1 Pre-Holocene glaciation 15 2.3.2 Hoiocene glaciation 19 2.4 CONCLUSIONS 20 3. GEOLOGY of QUATERNARY VOLCANIC FLOWS and CENTRES 21 3.1 INTRODUCTION 21 3.2 VALLEY-FILLING and PLATEAU-CAPPING UNITS 23 3.2.1 Clearwater unit 23 Distribution 25 Basal deposits 26 Morphology 27 The Flatiron 33 Whitehorse Bluffs 39 Source of the lava flows 44 3.2.2 Wells Gray unit 45 Stevens Lakes 45 3:2.2.2 Fight Lake 46 3.3 ICE-CONTACT FEATURES 47 3.3.1 Tuyas 47 3.3.2 Subglacial mounds and other unclassifiable centres 51 i v Ray Mountain 51 Pyramid Mountain 54 3.3.3 Ponded, valley-edge deposits 55 Sheep Track Bench 55" Jack's Jump 61 3.3.4 Magma and water interaction 62 3.4 PLEISTOCENE and HOLOCENE CINDER CONES 64 3.4.1 Holocene cones 65 Kostal Cone 65 Flourmill Cones 67 Spanish Cones , 68 3.4.2 Dragon Cone 71 3.4.3 Pre-Holocene cones 72 Buck Hill Cone 72 Other glaciated cones 74 3.5 CONCLUSIONS 75 4. ISOTOPIC DATING 76 4.1 INTRODUCTION 76 4.2 POTASSIUM-ARGON GEOCHRONOLOCY 76 4.2.1 Procedure 76 4.2.2 Results 78 Fiftytwo Ridge 81 Jack's Jump : 81 McLeod Hill 82 4.3 RADIOCARBON DATING 83 4.4 CONCLUSIONS 83 5. PETROGRAPHY 84 V 5.1 INTRODUCTION 84 5.2 DESCRIPTIONS 85 5.2.1 Valley-filling, plateau and ice-contact units 85 Type A 85 Type B 88 Type C 92 Type D 92 Trachytic texture 94 5.2.2 Young Flows and Cones 94 Type E .• 94 5.3 CONCLUSIONS 96 6. WHOLE-ROCK CHEMISTRY 98 6.1 INTRODUCTION 98 6.2 METHOD 99 6.3 RESULTS 100 6.3.1 Compositional variations within the suite 103 6.3.2 Petrogenesis 107 Crystal fractionation 110 Partial melting 121 6.4 CONCLUSIONS 132 7. Pb, Sr, and Nd ISOTOPIC RESULTS 135 7.1 INTRODUCTION 135 7.1.1 Techniques 136 7.1.2 8 7 S r / 8 S S r 139 7.1.3 1 " 3 N d / ' **Nd 139 .7.1.4 Whole-rock lead 142 7.2 RESULTS and DISCUSSION 143 v i 7.2.1 Crustal contamination of depleted mantle 146 7.2.2 Old subcontinental lithosphere 152 7.2.3 Sediment contamination of the mantle source 152 7.3 CONCLUSIONS 155 8. THE WELLS GRAY—CLEARWATER AREA IN A REGIONAL CONTEXT 156 8.1 INTRODUCTION 156 8.1.1 Mantle plumes 158 8.1.2 Plume tracks 160 8.2 PLATE VELOCITY 161 8.2.1 Data 161 8.2.2 Results 161 8.3 EULER POLES 163 8.3.1 Method 163 8.3.2 Results 165 8.3.3 Conclusions from Euler pole and velocity data 167 8.4 THE CASE for CONTINENTAL RIFTING 169 8.4.1 Cenozoic crustal movement in the Wells Gray—Clearwater Area> 171 8.4.2 Geophysical characteristics of the Cordilleran lithosphere 172 8.5 THE BASIN and RANGE ANALOGY 174 8.6 CONCLUSIONS 175 9. CONCLUSIONS 176 REFERENCES 179 APPENDIX A: SAMPLE PREPARATION 193 APPENDIX B: X-RAY FLUORESCENCE 205 APPENDIX C: ATOMIC ABSORPTION PROCEDURE 255 APPENDIX" D: WHOLE-ROCK CHEMICAL DATA 264 APPENDIX E: COMPUTER PROGRAMS 295 v i i LIST O F T A B L E S Table Page Table 4-1: K-Ar data table 79 Table 7-1: Sr isotopic results 138 ^ Table 7-2: Sm, Nd and l 4 3 N d / 1 * 4 N d isotopic results 142 Table 7-3: Whole-rock Pb isotopic results 143 Table 8-1: Age and velocity data for volcanic centres, Anahim Volcanic Belt 162 Table 8-2: Calculated Euler poles 164 Table 8-3: Comparison of calculated Euler poles 166 Table A-1: Apparatus used for sample processing 196 Table A-2: Sample preparation '. .....197 Table A-3: Analytical results for prepared samples 200 Table A-4: Elemental contamination 201. Table B-1: Machine parameters for XRF 209 Table B-2: Variables and equations 211 Table B-3: Results of sixteen replicate analyses 226 Table B-4: Probability results for duplicate analyses 226 Table B-5: Results of mixed samples 231 Table B-6: Trace element regression analyses 236 Table B-7: Major element calibration 247 Table B-8: Probability results of duplicate analyses (outside laboratories) 251 Table C-1: AA operating conditions for Sm, Eu, Dy, Er and Yb 259 Table C-2: Determination of Sm, Eu, Dy, Er and Yb in standard rocks 260 Table D-1: Trace and Major element chemistry, normative mineralogy 270 Table D-2: Alphabetic listing of samples numbers and locations 270 v i - i i LIST OF FIGURES Figure Page Figure 2-1: Generalized physiographic regions of British Columbia 6 Figure 2-2: Physiographic regions in the Wells Gray—Clearwater area 7 Figure 2-3: Faults and Eocene sediments 13 Figure 2-4: Wisconsin to Holocene glacial history 16 Figure 2-5: Quaternary stages 17 Figure 3-1: Generalized geologic map of the Wells Gray—Clearwater area 24 Figure 3-2: Clearwater River as it crosses the Murtle Plateau 26 Figure 3-3: Cliffs of the lower Clearwater Valley 26 Figure 3-4: Basal colonnade with "chisel" marks 28 Figure 3-5: Ropy flow top 29 Figure 3-6:. Well developed pillows 29 Figure 3-7: Finely laminated clay squeezed up between pillows 31 Figure 3-8: Lag gravel consisting of basalt boulders 32 Figure 3-9: Diamicton exposed in the Clearwater Valley 34 Figure 3-10: Exposure of irregularly oriented columns at the Flatiron 35 Figure 3-11: Subaerial/subaqueous contact downstream from The Flatiron 35 Figure 3-12: South-west face of Whitehorse Bluffs 40 Figure 3-13: Sedimentary structures exposed at Whitehorse Bluffs 40 Figure 3-14: Disturbed bedding around a subaerial bomb 42 Figure 3-15: Mount Hyalo, a tuya 49 Figure 3-16: Ray Mountain 52 Figure 3-17: Pyramid Mountain, a SUGM 52 Figure 3-18: Well indurated hyaloclastite (pillow breccia) 57 Figure 3-19: Vertical cliffs of hyaloclastite 58 Figure 3-20: High level dyke cutting hyaloclastite 59 Figure 3-21: Kostal Cone 66 Figure 3-22: Rafted blocks on flow surface 66 i x Figure 3-23: Crus ta l xenol i ths incorpora ted in f l ows f r om Flourmi l l C o n e s 70 Figure 3-24: Exposu re of f low base in the v ic in i ty of Span ish C o n e s 70 Figure 3-25: D r a g o n C o n e 73 Figure 3-26: Fus i fo rm b o m b s incorpora ted in to a till 73 Figure 4 -1 : Loca t i on map for isotopical ly d a t e d samp les . . .77 Figure 4-2 : K-Ar dates wi th error bars 80 Figure 5-1: P h o t o m i c r o g r a p h ; type A sec t ion 86 Figure 5-2: P h o t o m i c r o g r a p h ; resorbed p lag ioc lase crystal 87 Figure 5-3: P h o t o m i c r o g r a p h ; type B , sec t i on 87 Figure 5-4: P h o t o m i c r o g r a p h ; type B 2 sec t i on 89 Figure 5-5: P h o t o m i c r o g r a p h ; type B 3 sec t i on 89 Figure 5-6: P h o t o m i c r o g r a p h ; type B , sec t i on 90" Figure 5-7: P h o t o m i c r o g r a p h ; type B 5 sec t i on 90 Figure 5-8: P h o t o m i c r o g r a p h ; type C sec t ion 93 Figure 5-9: P h o t o m i c r o g r a p h ; type D sec t ion 93 Figure 5-10: P h o t o m i c r o g r a p h ; Trachytic texture 95 Figure 6-1: A F M d iagram 101 Figure 6-2: A lka l i vs . si l ica diagram 102 Figure 6-3: P/K vs. Ti/K diagram 105 Figure 6-4: Ba, N b , and N i vs. nephe l ine . . . .109 Figure 6-5: Pressure and temperatures of equ i l i b r i um assemb lage minerals 111 Figure 6-6: Ca l cu la ted equi l ibr ium olivine c o m p o s i t i o n s 113 Figure 6-7a: Ba , Sr, and N b vs. M g O 118 Figure 6-7b: N i , C r , and V vs. M g O 119 Figure 6-8: C h o n d r i t e norma l i zed REE patterns 122 Figure 6-9: Z r / N b vs . La/Yb 124 Figure 6-10: Ba , Rb, and N b vs. C a / M g . J 2 6 Figure 6-11: Ba, Rb , and N b vs. A l / C a : 127 Figure 6-12: Ni and Cr vs. Al/Ca; Ni and Cr vs. Ca/Mg 128 Figure 6-13: Zr/Nb vs. K z O 133 Figure 7-1: Terrane map of British Columbia 137 Figure 7-2: Terrane boundaries and sample locations for isotopic analyses 140 Figure 7-3: Zr/Nb vs. K 2 0 141 Figure 7-4: 1 4 3 N d / 1 4 4 N d vs. 8 7 S r / 8 6 S r '. 144 Figure 7-5: Plot of whole-rock lead data 145 Figure 7-6: Plot of f Nd vs. f Sr 147 Figure 7-7: B 7 S r / 8 6 S r vs. Rb/Sr , 147 Figure 7-8: Rb/Sr vs. Ti/K; Sm/Nd vs. Ti/K, and 1 0 3 Nd/ 1 * 4 Nd vs. 1 4 7 Sm/ 1 4 4 Nd 148 Figure 7-9: Crustal contamination 150 Figure 7-10: 8 7 S r / 8 6 S r vs. whole-rock Pb 154 Figure 8-1: Quaternary volcanic centres and Euler pole tracks 157 Figure 8-2: Transcurrent faults and Eocene basins in British Columbia 170 Figure A-1: XRF scan for W and Ta peaks 198 Figure B-1: Mean values of duplicate sets of analyses 223 Figure B-2: "Working curves" for XRF analyses 232 Fiqure B-3: Comparison of XRF results with other laboratories 248 Figure C-1: Chondrite normalized REE patterns for test samples 261 Figure D-1: Major, minor and trace elements against MgO 265 Figure E-1: Flow chart for data processing 296 x i LIST OF PLATES Plate , „ f . Location, Plate 1 ^ * C U A ^ ' ^ n a t e 1 -Sfeek—peeket -x i i ACKNOWLEDGEMENTS M a n y p e o p l e are respons ib le fo r the p r o d u c t i o n of a thesis. T h o u g h not i nvo l ved in the d i rec t wr i t ing , the i r w o r d s of e n c o u r a g e m e n t , mora l o r f inancial suppo r t result in a f in ished p roduc t . But a thesis is really just the cu lm ina t i on of o n e ' s s tudent days. As s u c h , I o w e great deb t t o the many teachers I have had du r ing m y s c h o o l i n g . N o less of a d e b t is o w e d to m y parents w h o car ted my b ro the r and m e a r o u n d to ' see the w o r l d ' and pu t up w i th m y large c o l l e c t i o n of pebb les—I am sure they never t hough t anyth ing usefu l m igh t c o m e of t h e m . M y h u s b a n d , Paul , has also b e e n espec ia l l y suppor t i ve . I am not sure wha t h e though t the first s u m m e r I d i sappeared ' up no r th ' for four m o n t h s , but his he lp and suppo r t t h r o u g h o u t has b e e n unfa i l ing. M y s inceres t grat i tude g o e s to m y thesis superv isor , W . H . M a t h e w s and to the o the r m e m b e r s of m y c o m m i t t e e ; R.L. A r m s t r o n g , W . C . Barnes, and J .C . Sou the r for thei r he lp and suppor t a l ong the way. I have also bene f i t ed great ly f rom d i scuss ions w i th many o the r p e o p l e . In part icular, L .C. Struik and D. M u r p h y en l i gh tened m e o n the intr icacies of t ec ton i cs in the Shuswap M e t a m o r p h i c C o m p l e x , and J.K. Russel l h e l p e d w i th the chemis t ry of these hype rs thene normat ive , m o d a l alkali o l i v ine basalts. S. J. Juras p o i n t e d ou t that o l d (Pa leozo ic ) vo lcan ics can a lso be of interest and there migh t even be s o m e e c o n o m i c i m p o r t a n c e in a ' b u n c h of basal ts ' . S.T. B i s h o p (1984) and I. G e r t z (1983) p r o v i d e d amiab le and capab le assistance in the f ie ld . D raw ing the M a s o n — D i x o n l ine w i th P. Me tca l f e was always great fun . Thanks also t o F. N o o n e for her t r e m e n d o u s suppo r t and assis tance s ince ' the early days ' . A n u m b e r of p e o p l e have b e e n a great h e l p in var ious aspec ts of the chemis t ry . In part icular, I w o u l d l ike to thank H. Baadsgaard , P. M e t c a l f e , A . Smi th , and P. Cave l l f r o m the Univers i ty of A lber ta , for tak ing m e by the hand and lead ing m e in to the w o n d e r f u l w o r l d of i so topes . S. Horsky , at U B C , was always there w h e n a p r o b l e m d e v e l o p e d ; w h e t h e r it was a p l u g g e d c o l u m n , a b r o k e n XRF unit, or a pesky in jector for the A A . Parks staff of the We l l s C ray and N o r t h T h o m p s o n Prov inc ia l Parks were m o s t he lp fu l t h r o u g h o u t the f ie ld wo rk and p r o v i d e d logist ical suppo r t and radio con tac t as n e e d e d . I w o u l d espec ia l ly l ike to thank the super in tenden t , P. Rogers ; the park natural ist (of Bu fo Inc.), T. C o w a r d ; and ). Br iggs, for their h e l p and l ively conversa t ion . The au thor was s u p p o r t e d by a Natura l Sc iences and Eng ineer ing Research C o u n c i l Post Gradua te Scho la rsh ip . Fund ing for f ie ld w o r k du r ing the s u m m e r s of 1983 and 1984 was p r o v i d e d largely by the G e o l o g i c a l Survey of C a n a d a th rough J .G. Souther . A d d i t i o n a l suppor t for s u m m e r wo rk , chemica l analyses, o t h e r data aqu is t ion and thes is prepara t ion was t h rough Natura l Sc iences and Eng ineer ing C o u n c i l grant, A 1 1 0 7 , t o W . H . M a t h e w s . x i i i FORWORD Major and trace element chemical analyses on samples from the Wells Cray—Clearwater area were carried out in conjunction with similar work by S.j. Juras on the Paleozoic rocks of the Sicker Croup on Vancouver Island. Samples used for the error analyses presented in Appendix B are from both rock suites. A joint investigation of sample contamination stemmed from this analytical work and has been published in the Canadian Mineralogist (Hickson and Juras 1986). A modification of this paper appears as Appendix A. A second joint study was undertaken into the analyses of the rare earth elements by atomic absorption, graphite furnace techniques. Funding for this work came jointly from W.H. Mathews (NSERC grant A1107) and C.I Godwin (Energy Mines and Resources research agreement). A modification of Appendix C will appear in Chemical Geology (Juras ef a/., in press). xiv Fire and Ice S o m e say the w o r l d wi l l e n d in f ire, S o m e say in ice. F r o m what I've tasted of desi re I h o l d wi th those w h o favor fire. But if I had to per ish tw ice , I think I k n o w e n o u g h of hate T o say that for des t ruc t ion ice Is a lso great A n d w o u l d suff ice. — Rober t Frost St. Ronan's Well A n d s o m e rin up hill and d o w n dale, knapp ing the c h u c k y stanes to p ieces w i ' hammers , l ike sa m o n y road-makers run daft. They say 'tis to see h o w the w o r l d was made ! - S i r Wa l te r Scot t XV 1. INTRODUCTION If something is important it straddles the boundary between two maps. If it is very important it lies at the corners of four maps. —W.H. Mathews The Wells Cray—Clearwater area, in east-central British Columbia, is a region of moderate relief between the Interior Plateau and the Columbia Mountains, defined as the Shuswap—Quesnel Highland (Fig. 2-1) (Holland 1964; Mathews in press). The geology of the area was first outlined by Selwyn (1871). Detailed study along the North Thompson River began in 1921 with the work of Uglow (1921) but it was not until 10 years later that mapping was done in the lower Clearwater Valley by Walker (1930) and regional mapping did not commence until the 1960's (Campbell 1963a, 1963b, 1967; Campbell and Tipper 1971). Though Uglow (1921) first reported the presence of young volcanic rocks from the Mann Creek area, a tributary of the North Thompson River, it was not until the regional work of Campbell (1963a, 1963b, 1967) and Campbell and Tipper (1971) that the extent of the volcanic deposits in the Wells Gray—Clearwater area was realized. Campbell and Tipper (1971) correlated them with flat lying lavas (now referred to as Chilcotin Group basalts, Bevier 1983a) of Miocene-Pliocene age in the Interior Plateau. Several young cinder cones were also recognized at that time. As geological mapping continued in the Cordillera, additional centres of late Tertiary volcanic activity were recognized. Souther (1977) produced a synthesis of the work to that date and delineated several late Tertiary to Quaternary volcanic belts. One of these was the east-west trending Anahim Volcanic Belt. In 1982 J.C. Souther, of the Geological Survey of Canada, began a reconnaissance study of volcanic centres in the Anahim Volcanic Belt, previously interpreted to be the manifestation of a mantle hot spot (Bevier et al. 1979). The project was to map and describe some of the volcanic centres not previously 1 2 studied in detail and to put them into the regional context. This study of the Wells Cray—Clearwater area was one of the projects undertaken. Preliminary finding were published in 1984 (Hickson and Souther 1984) based on four weeks field work in 1981 and 1982. Detailed mapping began in 1983 and continued in 1984 (Plates 1 and 2). Basaltic rocks of the Wells Cray—Clearwater area span Quaternary time and include subaerial, subaqueous, and subglacial deposits. These deposits are locally intercalated with glacial till, fluvial sands and gravel, and rarely paleosols. With the exception of the Holocene cones and flows, all other deposits have been modified by regional glaciations and deeply dissected by fluvial erosion. Claciation has left a thick blanket of till over most of the volcanic deposits and outcrop is largely limited to cliff-forming exposures in the valleys. Hickson and Souther (1984) subdivided the lavas into four morphological groups: a post-glacial assemblage, a late glacial assemblage, a valley-filling assemblage, and an early glacial assemblage. Detailed mapping and dating carried out for this study has resulted in the amalgamation the glacial assemblages and shown the importance and extent of some of the upper elevation, plateau-capping volcanics. As a consequence the valley-filling assemblage has been split into the informal Wells Cray unit (above 1200 m) and the Clearwater unit (below 1000 m). Both of these units cover or form plateau surfaces in addition to occurring as valley-filling deposits. In the present study, chemical analyses of basaltic samples are used to determine the chemical affinity of the magmas. These chemical characteristics are used to determine the possible origin and evolution of the basaltic melt in the mantle and crust respectively. On the basis of K-Ar dating, it was concluded that the the Wells Cray—Clearwater suite was younger than the chemically transitional to alkali olivine 3 basalts of the Chilcotin Croup and were not correlative (Hickson and Souther 1984). It was also concluded on the basis of age, petrography, and mantle origin (Fiesinger 1975; Fiesinger and Nicholls 1977) that inception of Quaternary volcanism in the Wells Gray—Clearwater area was consistent with melting produced by a mantle hot spot as proposed by Bevier et al. (1979) (Rogers 1981; Hickson and Souther 1984). The hot spot hypothesis is critically examined in light of further isotopic dating from the Wells Gray—Clearwater area and from elsewhere in the Anahim Volcanic Belt. A possible alternative hypothesis, crustal rifting, is suggested in light of evidence of late Tertiary crustal movement in the region and of regional work on basement rocks and structures. Work to the south and east of the Wells Gray—Clearwater area has identified metamorphic core complexes. Exposure of these core complexes during the Eocene is thought to coincide with crustal extension, rising geothermal gradients, uplift and tectonic denudation and Late Cenozoic developments may continue this earlier history. In the Basin and Range province of the United States, the region of exposure of core complexes is a region of vigorous volcanic activity. Similarity of the Basin and Range province with the Canadian Cordillera will be examined. The Wells Gray—Clearwater area is of interest from a volcanological point of view and from a structural and metamorphic viewpoint. The region has become the focus of a number of metamorphic and structural theses (Pigage 1978; Pell 1984; Getsinger 1985; Montgomery 1985; Murphy 1985; Fillipone 1986; M.A. Bloodgood in preparation; S. Carwin in preparation; P. Lewis in preparation; D. McMullen in preparation), as well as a regional study by L.C. Struik of the Geological Survey of Canada. This interest has been sparked in part by the interpretation of the Canadian Cordillera in terms of displaced terranes proposed by J. C. Monger and others 4 (Monger ef al. 1982; Monger and Berg 1984; Price ef al. 1985). In the work of these and others the basement rocks in the Wells Cray—Clearwater area are interpreted as ranging from far travelled, allochthonous Mesozoic rocks to parautochthonous (North American affinity) Paleozoic to Precambrian rocks. The rocks of North American affinity are interpreted to have a Precambrian basement and based on geophysical data, crustal thicknesses diminish to the west, coincident with the proposed boundary with the allochthonous terranes. The Wells Cray—Clearwater area basalts are among very few late Tertiary volcanic rocks to have erupted through this presumed older and thickened crust. The effect that the Precambrian crust has had on the volcanics will be assessed principally in terms of radiogenic isotopes. Zones of weakness through which conduits might develop are also assessed. If volcanic eruptions are confined to a terrane boundary, there may be some indication of an inherited weakness along that zone if the boundary is steep. 2. PHYSIOGRAPHY, CEOMORPHOLOCY AND QUATERNARY GLACIAL RECORD Nothing under heaven is softer or more yielding than water, but when it attacks things hard and resistant there is not one of them that can prevail. — "Tao Te Ching" 2.1 PHYSIOGRAPHY The Wells Gray—Clearwater area straddles the boundary between the Columbia Mountains and the Interior Plateau of the physiographic system defined by Bostock (1948). Bostock noted that an erosion surface (or peneplain) extended from the Interior Plateau at an elevation of 1200 m to an elevation of at least 1800 m in the Columbia Mountains. Holland (1964), refined Bostock's physiographic divisions for southern British Columbia and added the Shuswap and Quesnel Highlands in the Wells Gray—Clearwater area (Fig. 2-1). The highlands were defined as belts, approximately 80 km wide, which were more elevated and dissected than the Interior Plateau surface. The western boundary of the Quesnel Highland was arbitrarily placed at 5000 ft. (approximately 1525 m) on the gentle eastward rise of the erosion surface. Between the Thompson Plateau and the Shuswap Highland, the boundary coincides with the Louis Creek Fault and North Thompson River (Fig. 2-2). The eastern boundary of both highlands was arbitrarily set along the Hobson, Clearwater, and Murtle lakes and Blue River where elevations reach 2100 m. The peneplain surface is still present east of this point as long slopes leading up the high peaks (Fig. 2-2; Plate 1). This upwarped erosion surface has been substantially dissected by the Clearwater, North Thompson and other rivers. W.H. Mathews (in press) has made minor modifications to the boundary between the Quesnel Highland and the Interior Plateaus (Fig. 2-2). In the revision of the physiographic map of Holland (1964), Mathews (oral communication 1986) identified neotectonic features, such as areas of upwarping and anomalous 5 Figure 2-2: Physiographic regions in 'the Wells Cray—Clearwater area after Mathews (in press). The north and eastward extent of the peneplain (or erosion) surface is shown by the dashed line. The eastern limit of the peneplain relics coincides with the boundary between the Cariboo and Monashee Mountains and the Shuswap Highland. Solid lines outline the physiographic regions and the Clearwater Depression. Contours are at 1220 m (4,000 ft), 1830 m (6,000 ft) and 2440 m (8,000 ft). Elevations between 1220 - 1830 m are shown as a • , and those above 2440 m are shown as solid triangles. 8 depressions. The Murtle Plateau (Clearwater Depression) was recognized as a relatively depressed area, but for reasons of scale, is not depicted in the revised 1:5,000,000 physiographic map. In terms of the neotectonics of the Wells Cray—Clearwater area both the warping of the peneplain and the Clearwater Depression are important features for documenting recent crustal motion. 2.2 GEOMORPHOLOGY 2.2.1 PENEPLANATION Development of the peneplain surface began at the earliest in the Eocene. Cairnes (1931), mapping in the vicinity of Okanagan Lake, showed Eocene Kamloops Group rocks dip 20 to 25° W and are truncated by a conspicuous erosion surface (Cairnes 1931; his Plate IB, p. 113). The erosion appears to have been largely completed by the Miocene because plateau basalts were deposited on comparatively flat terrain. Where the basalts overlie Eocene, Kamloops Group rocks as much as 2 0 ° of angular discordance is found (Campbell and Tipper 1971). Uplift probably resumed after the Miocene and was accompanied by doming to the south and west between the North Thompson and Deadman rivers, just south of the region of Figure 2-2 (W.H. Mathews, oral communication 1986). Uplift in the region south and east of the Clearwater and North Thompson rivers probably contributed to the dissection of the mature erosion surface noted by Campbell and Tipper (1971). Uplift of the peneplain surface is greatest along the eastern boundary of the region but the age and relationship with mapped faults is unknown. Along the eastern margin of the Interior Plateau and into the Quesnel Highlands numerous normal faults (Campbell and Tipper 1971; their Fig. 5, p. 78) have been mapped. These cut Kamloops Croup sedimentary and volcanic rocks which are of Eocene age. In this region, Kamloops Group rocks belong to the Chu Chua and Skull Hill 9 formations, respectively and are interpreted as having been preserved in grabens and half grabens. Some of these structures are several tens of kilometres wide and have several hundred metres of relief. These grabens and half grabens are believed to have formed during a period of major crustal heating (Ross 1974; Ewing 1980; Mathews 1981). Although faulting of the Eocene strata was both syn-depositional and post-depositional, Campbell and Tipper (1971) state that the faulting can extend to no younger than late Miocene as the Deadman River Formation is not cut by late normal faults. However, the Deadman River Formation does not occur in the area of block faulting shown on their map and regions of the plateau and highlands are covered by thick glacial drift that may obscure other fault traces. Peneplanation must have been well underway by the Oligocene. The initiation of erosion was probably in response to uplift associated with the regional thermal event and regional crustal movements which have preserved rocks of the Eocene Kamloops Croup. After this disturbance the land surface acquired its mature form (peneplain). Uplift of the peneplain began after the deposition of the Miocene-Pliocene Chilcotin Croup basalts, but the locus of uplift may have been farther to the east than during the Eocene event. Uplift of the peneplain caused incision by the major rivers as dissection of the landscape began. This more recent uplift may be in response to the same process which earlier, has uplifted and exposed high grade metamorphic rocks (metamorphic core complexes). The exposure and tectonic denudation associated with core complexes in the Okanagan area and elsewhere are considered to be Eocene events (Templeman-Kluit 1984; Parkinson 1985; Parrish 1985; Parrish et al. 1985). 2.2.2 DEVELOPMENT OF THE DRAINAGE SYSTEM The regional drainage system in the Wells Gray—Clearwater area reflects large scale patterns in topography. These patterns may reflect post Eocene diastrophism 10 that for reasons of erosion are not evident in other types of features. Evidence of features, such as fault scarps or sag basins, tends to be removed in regional glaciations or are buried under thick deposits of sediment. The Wells Gray—Clearwater area is drained by a major north-south system comprising the Clearwater and North Thompson rivers (Fig. 2-2). This is cut in an east-west direction by the Murtle and Mahood rivers. Tributary drainage to the Clearwater River is well developed on the eastern side of the region and major, glacially scoured valleys such as Spahats, Moul, and upper Hemp creeks and Murtle River, drain westward into the Clearwater River. On the west side, no streams of similar magnitude drain eastward, with the exception of the Mahood River whose drainage pattern is controlled by Canim Lake to the west. North of Mahood Lake, streams drain westward, then southward into the lake. South of Mahood Lake, streams drain southeasterly to the North Thompson Valley. Streams to the north of Canim Lake flow southeasterly into the lake and those to the south, on the same trend, largely flow into the North Thompson River (Fig. 2-2). Control of the North Thompson—Clearwater River drainage pattern may be structural. The valleys are coincident with faults mapped in the North Thompson region by Campbell and Tipper (1971) and extended into the Clearwater Valley by Schiarizza and Preto (1984). Evidence in Hemp Creek valley north of this point corroborates extension of the fault northwards. There, sediments of Eocene Kamloops Group (correlated with Chu Chua Formation by Campbell and Tipper 1971) are found. These sediments consist of laminated fine to coarse sands, pebble conglomerate, laminated siltstone and shale with coal seams up to 20 cm in thickness. The sequence is consistent with a low gradient, swampy environment. After deposition, the sediments have been sheared and, in some locations, rotated and tilted to vertical. The disruption of the sediments was probably in response to faulting, similar to that seen elsewhere in Eocene sediments which have been 11 preserved in downdropped structures. Basement rocks of the Shuswap Metamorphic Complex exposed in Hemp Creek valley are also highly disrupted. Downstream from the confluence of Hemp Creek and the Clearwater River, rocks of the Raft Batholith are exposed and are more fractured and altered than similar rocks visible higher up on the west side of the river. Faulting appears to have produced a weakened zone in which downcutting is actively occurring. The present north-south drainage may have intersected an original (Eocene?), westwardly directed low gradient drainage system. Whether late diastrophism or local erosion of rock shattered during earlier distrubances is responsible for the the development of Clearwater River and Hemp Creek valleys from the Eocene to the Quaternary cannot be determined. During the Quaternary uplift and/or tilting of a triangular biock, bounded on the east side by Hemp Creek and on the west by the Clearwater River (Green Mountain), has raised the contact between subaerial/subaqueous volcanic deposits 60 m above the equivalent contact on the east side of Hemp Creek (see Chapter 3). A basalt flow upstream of this point has been dated by the K-Ar method (Chapter 4) at 1.4 Ma. Downstream from Hemp Creek, evidence of vertical displacement in the Clearwater Valley is inconclusive. Although the contact between subaerial flows and pillows varies in elevation on either side of the river, these variations can be explained by deposition of the lava flows into water that was running on top of, or cutting down through, older flows. The Quaternary river gradient can be determined at Clearwater where subaerial flows occur at river level (425 m) overlying laminated sands and gravels with up to 40% basalt clasts. Some 14 km upstream, a polymictic deposit of lag gravels, with boulders up to 2 m in diameter, is preserved beneath the flow. Dating of flows above these deposits gives ages between 0.3 and 0.6 Ma (Chapter 4). Correlating these deposits gives a gradient of 9 m/km compared to 5 m/km for the present river valley. The 9 m/km gradient is consistent with the sizes of clasts 1 2 seen in the lag gravels underlying the flows. All of these lag gravels contain basalt clasts eroded from the drainage system prior to the mid-Quaternary. Thus, little change in elevation in a north-south direction is indicated, but east-west changes are not precluded. Regional uplift has elevated areas to the east along an axis parallel to the Southern Rocky Mountain Trench. Tributary drainage to the Clearwater River is well developed on the westward tilted slopes. Orographic effects and local structural trends have either amplified or modified the effect. This is especially apparent on the west side of the Clearwater River where streams are parallel with underlying lithologic and structural trends in Tertiary and older bedrock, and in many cases, to the trends of block faults shown by Campbell and Tipper (1971). 2.2.3 THE CLEARWATER DEPRESSION The Clearwater Depression is synonymous with the Murtle Plateau and appears as an anomalous widening of the Clearwater Valley (Fig. 2-2). The Clearwater Valley extends northward from its confluence with the North Thompson River where it is about 3.5 km in width. North of the confluence with Hemp Creek it widens abruptly and forms the Murtle Plateau (Plate 1; Fig. 2-2). If hills rising above 1000 m are excluded (three are constructional volcanic forms and the fourth, Green Mountain, may have been uplifted in the Quaternary) from the Murtle Plateau, it forms a rhombus approximately 17 km on a side (Fig. 2-2). The Wells Gray—Clearwater area appears to be the locus of transition from northerly trending stuctures such as the Fraser Fault zone and major north-south trending valleys, to northwesterly directed ones (Fig. 8-2). The mapped distribution of metamorphic core complexes (Okulitch 1984; Templeman-Kluit 1984; Parkinson 1985; Brown 1986; Brown and Read 1983; Parrish et al. in preparation) culminates in a wedge at this point. The locus of active uplift may extend northward into the Figure 2-3: Regional map of the Wells Cray—Clearwater area. Eocene Kamloops Croup rocks are shown in a stipple pattern. Areas of highly disrupted Chu Chua sediments are shown by stars. Traces of known post-metamorphic faults are shown as heavy lines (M.A. Bloodgood in preparation; S. Carwin, oral communication 1986; P. Lewis, oral communication 1986; D.L. Murphy, oral communication 1986; Schiarizza 1986; Fillipone 1985; Struik 1985b, in preparation; Campbell 1963a, 1963b, 1967; Schiarizza and Preto 1984; Campbell and Tipper 1971). Topography is as specified in Figure 2-2. 14 Premier Range where the highest elevations in the Cariboo Mountains are recorded. Normal faults, developed in response to this uplift, are coincident with Clearwater Lake, and the Clearwater and North Thompson rivers. The sense of movement is west side down (Schiarizza 1986; Schiarizza and Preto 1984). Topographically this is especially evident in the vicinity of Trophy and Raft mountains (Plate 1; Fig. 2-2; Fig. 2-3) where long ridges lead up to peak heights 400 m above those on the west side of the river. This change in elevation, when viewed in cross section, gives a stepped appearance to the topography. The faults running the length of the Clearwater River system form an en echelon set which overlap in the vicinity of the Clearwater Depression. The Murtle Plateau (Clearwater Depression) has formed in the region of overlap (Fig. 2-3). The fact that the north and south sides of the depression parallel other faults in the region, may result from coincidence of the recent features with older structures. When the Clearwater Depression developed cannot definitely be determined but the surrounding mountains have steep vertical fronts which extend from an elevation of 700 m to 1600 m at which point there is an abrupt break in slope (Plate 1; Fig. 2-3). This observation, coupled with offset of Quaternary lavas, leads to the conclusion that faulting and formation of the basin may have extended into Holocene time. 2.3 CLAC1ATIONS Most of the valleys in the Wells Gray—Clearwater area are fault controlled, but modification by glacial ice has undoubtedly been significant. During the Wisconsin Glaciation, ice reached elevations of 2200 m and covered most ridges and many of the peaks, leaving a veneer of glacial till over most of the land surface (Maxwell 1985). Ice advanced from the north-east and the area was inundated by ice from at least 20,000 y.b.p. until 11,000 y.b.p. The ice reached an 15 elevation of 2400 m (Clague 1981) and this glaciation, termed the Fraser Glaciation in British Columbia, was preceded by at least three glacial periods of similar magnitude recorded in sediments in the lower Fraser Valley (Fig. 2-4). These glacial periods, the Semiahmoo, Westlynn glaciations (Fig. 2-4) and a recently discovered one documented by a glaciomarine deposit (dated at greater than 0.1 Ma); the last unnamed period may be Pleistocene or late Tertiary (J.E. Armstrong 1981). Classical Quaternary, glacial periods have come under attack in studies such as that of Easterbrook and Boellstorff (1981) who showed that Kansan and lllinoian drift from the mid-western United States is much older than thought. In one case two 'Quaternary' tills were found to be greater that 2.2 Ma. Mathews and Rouse (in press) have documented a till underlying lava flows in the area of Dog Creek on the Interior Plateau at 1.2 Ma. and Souther (oral communication 1982) has found several late Tertiary tills in the Mount Edziza area. Work with oceanic sediment oxygen isotopic ratios (Shackleton and Opdyke 1976) concludes that major ice caps have occurred somewhere on the earth at least 17 times in the last 2 million years (Fig. 2-4). In the Wells Gray—Clearwater area volcanic features record the effects of ice throughout the Quaternary. 2.3.1 PRE-HOLOCENE GLACIATION Pre-Holocene glacial advances have been recorded in tuyas and other subglacial volcanic centres in the Wells Gray—Clearwater area. Till has also been preserved beneath the flows in at least two locations. Potassium-argon dating has been used to date these features and the dates can be compared to the classic North American Quaternary stages and 1 s O data. The dating is not precise enough to tie directly to the 1 a O data or to the classic North American glacial statigraphy (Fig. 2-5). 16 YEARS BJ>. C X 1 0 3 ) (scale varies) TIME -S T R A T I G R A P W C UNITS GEOLOGIC -CLIMATE UNITS LITHOSTRATIGRAPHIC UNITS WG-CLW Werred age relatlonsrnqa — . _ . H O L O C E N E Raft- Spa hats > ^ o o < a > -i Dunn Peak * Q SALtSH SEDMENT3 AND FRASER RIVER S E D M E N T 3 Kostal Cone Dragon Cone 10 - , ffl > " Q oALfon s t U M c n i i Spanish. Fburmia 3UMA3 DRIFT Cones 12 13 / C A P I L A N O SEDIMENTS FORT LANGLEY FORMATION Buck Hid Pyramid Mt. Pointed Stick Cone 31 Fifty two Ridge 15 18 20 28 30 LATE WISCONSH 09 m a o VA3HON DRIFT Ray Mt. ? r-> o > H 5 Z OUADRA SAND 1 C~OQUTTLAM L______DB1FT ML Hyakj £ o Mann Creek 35 41 o r-m j * m a OLYMPIA NONGLACIAL N T E R V A L COWICHAN HEAD FORMATION 3nook w a Creek 50 o z CO 2 Canim Fata S O 6 2 EARLY WISCONSIN AND PRE-WISCONSIN 3EMIAHMOO GLACIATION SEMIAHMOO D R F T Ida Ridge HIGHBURY NONGLACIAL INTERVAL HIGHBURY 3EDMENTS WESTLYNN GLACIATION WESTLYNN D R F T OLDER 3EDMENT3 Figure 2-4: Wisconsin to Holocene glacial history after J.E. Armstrong (1981) (Lower Fraser Valley) and Duford and Osborn (1978) (Wells Gray—Clearwater area) showing the relationship to ice-contact and subaeriai volcanic features in the Wells Gray-Clearwater area (WG-CLW). 17 Ma 0.0 1.72 1.38 1 PALEO— MAGNETIC TWE-3CALE 0 MARME CLIMATIC STAGES CLASSICAL NORTH AMERICAN STAGES ( A GLACIAL) Wells Gray Clearwater Area K-Ar Data BRUNHE3 f-—> WISCONSINIAN A o I <> I T I S J } 1 1 T ° . L JL i L O Subaeriai A Ic* Contact A Pondad Water 3ANGAMONIAN 1 ' ILLINOIAN A y. . YARMOUTHIAN «,* ' KANSANU • KANSAN 1 • JASAUt_LO i 1 MATUYAMA MATUYAMA y < LOW HIGH — s'*o— c o l d warm AFTONIAN MATUYAMA NEBRA3KAN • OLQOVAI Quate rnary MATUYAMA MATUYAMA Pl iocene I I NORMAL Figure 2-5: Quaternary Stages after Harland et al. (1982) with reference to the K-Ar dates in the Wells Cray—Clearwater area. See Chapter 4 for details of the dating . technique. 18 Despite the lack of correlation with glacial features elsewhere, the tuyas and valley-ponded deposits indicate at least, three periods of glaciation (Chapter 3). The contact between subaerial and subaqueous volcanic deposits in an ice ponded deposit does not necessarily represent the upper elevation of the ice but rather reflects the upper limit of the water table confined by the ice; the ice level may have been much higher. At the bluff called Jack's Jump (1.9 Ma; Chapter 4), the subaerial/subaqueous contact is at 1700 m elevation, 700 m above the valley floor. At Sheep Track Bench (0.27 Ma) the break between subaerial and subaqueous flows occurs at 1400 m (700 m above the valley floor). The three tuyas in the Clearwater Depression (McLeod Hill, 3.5? Ma; Gauge Hill, 0.27 Ma; Mosquito Mound, 0.38 Ma) rise between 1000 and 1200 m above the valley floor which is at 700 m elevation around Mosquito mound in the west and increases to 1000 m elevation near Gauge Hill (Plate 1). These three volcanic centres have suffered the most extensive erosion which may have lowered summit heights by several hundred metres. The level of a former lake is recorded in an exposure along the west bank of the North Thompson River Valley at Blackpool by a deposit of subaerial flows, pillows and pillow breccias. The pillows and pillow breccias form a foreset bedded delta which had a water/air contact at 518 m elevation. These flows give K-Ar dates of 0.35-0.50 Ma (Chapter 4) and indicate a lake was present in the valley of the North Thompson River into which the pillow delta was built. Although it is possible that the lake may have been due to disruption of drainage by landsliding or tectonism, it seems more likely that water was dammed by residual glacial ice. Damming of the valley has been documented in the Kamloops area by Fulton (1967), and in the course of this study varved silts were found at 455 m in the North Thompson Valley and at an elevation of 550 m in tributary valleys of the Clearwater River. During the Fraser Glaciation a classic Gilbert type delta, 100 m in 19 thickness, was built into the North Thompson Valley at its confluence with the Clearwater River so it is possible that this scenario was repeated may times during the Quaternary. 2.3.2 HOLOCENE GLACIATION Even for the Holocene, the number and timing of alpine glacial advances is controversial. The Holocene advances occurred within existing cirques and four have been suggested (Alley 1976, 1980; Duford and Osborn 1978, 1980) in the Wells Gray—Clearwater area (Fig. 2-5). The oldest, the Harper Creek advance, is regarded as a resurgence of alpine glaciation after valley glaciers had begun downwasting. In some cases alpine glaciers of this advance may have connected to valley glaciers (Duford and Osborn 1978). The later, Dunn Peak advance, may have contributed to cirque development, but the Raft-Spahats advance and Little Ice Age probably were insignificant in terms of erosive power. The effects of these Holocene advances are recorded in some of the volcanic deposits and can be used to date the deposits where very young and consequently uncertain K-Ar dates were obtained. Mount Hyalo and Ray Mountain both have minimum K-Ar dates and the latter is thought to be older, based on a larger proportion of radiogenic argon in the sample (J. Harakal, oral communication 1985) and the greater degree of erosion. However, the erosion of Ray Mountain may be anomalous as it lacks capping, resistant flows, occurs on a sharp ridge, and is high enough that cirque glaciers from both the Harper Creek and Dunn Peak advances may have occupied or carved northeast and northward facing cirques. A tree-covered moraine, identified on air photos, occurs at 1800 m but active rock falls have obliterated any moraines above this elevation. Dragon Cone, dated by the 1 *C method (Chapter 4), may shed light on the age of the Dunn Peak advance. The cone is at a sufficiently high elevation to 20 have been influenced by the accumulation of ice during the Dunn Peak advance (Alley 1976, 1980; Duford and Osborn 1978, 1980) but there is no evidence of modification of the cone, nor substantial erosion of the flows. This corroborates the suggestion of Duford and Osborn (1978, 1980; Fig 2-4) that the Dunn Peak advance may predate 7400 years. Figure 2-4 gives the relative ages, based on erosion, of all the young centres. 2.4 CONCLUSIONS Peneplanation probably started in the Eocene and continued through the Oligocene to the Miocene. This erosion, through a period of quiescence, produced a mature, low relief surface, upon which Middle Miocene to Pliocene Chilcotin Group lavas were deposited. Uplift of the surface and dissection of the peneplain followed. Rivers flow both south and west in response to regional elevations which are greatest to the east and north. The southward trend of the Clearwater and North Thompson rivers is controlled by late normal faults which form an en echelon pair, overlapping in the vicinity of the Murtle Plateau. The Plateau surface is inferred to have dropped down in the late Tertiary in the area of overlap to form the Clearwater Depression, probably coeval with the uplift of the peneplain surface. Displacement is at least as recent as the mid-Quaternary and may be ongoing. 3. GEOLOGY OF QUATERNARY VOLCANIC FLOWS AND CENTRES Touch the earth, love the earth, honor the earth, her plains, her valleys, her hills, and her seas; rest your spirit in her solitary places. — Henry Beston 3.1 INTRODUCTION Tertiary or younger volcanic rocks were first identified in the Wells Cray—Clearwater area by Uglow (1921). He mapped the North Thompson Valley to the south of the study area, but also made observations in the Clearwater region. He described a section in the lower Mann Creek valley, where olivine basalts rest on unconsolidated fluvial gravels, and suggested the basalts were Holocene in age. Walker (1930) mapped in the vicinity of Clearwater and expanded on Uglow's observations of the valley-filling basalts, formally classifying them as the Mann Creek Formation. He revised Uglow's age estimate downward to Tertiary, but suggested the basalts were younger than the sheared and faulted Eocene Skull Hill volcanics. Rocks of the Skull Hill Formation were identified in several locations along the North Thompson Valley by both Uglow and Walker. Further mapping of the region did not occur until the 1960's. This may have been in response to the troublesome terrain upon which A.R.C. Selwyn commented in 1871 (p. 18). He described the difficulty "in penetrating the dense and pathless forest and jungle which prevails almost unbroken except by swamps and rivers from Kamloops to the Leather Pass in the Rocky Mountains." Walker (1930; p. 140A) echoed Selwyn's words and in his advice to prospectors said " would seem wisest that prospecting should be carried on by residents of the district who can devote some time to the search for minerals but have some other means of gaining a living." Regional mapping was completed by Campbell (1963a, 1963b, 1967) and Campbell and Tipper (1971) in the sixties. In these works the formation name, Mann Creek for basaltic rocks in the region (Uglow 1921) was abandoned and the rocks ' 21 22 correlated with the Miocene plateau lavas (Chilcotin Group; Bevier 1983a, 1983b), except for the Holocene cones and flows. The present work is the first detailed study of all the Quaternary volcanic rocks in the area and attempts to elucidate the relationships and timing of the various flows and volcanic vents in the area. The volcanic rocks are basaltic, very fresh and generally pale gray to black. The darkest flows are very dense with a fine grained, glassy texture; lighter coloured flows have a more open, diktytaxitic texture. The Quaternary basalts are easily distinguished from Eocene Skull Hill basalts and andesites by their lack of alteration and deformation. Olivine phenocrysts are present in virtually all flows. Plagioclase and/or pyroxene phenocrysts, as well as olivine phenocrysts, occur in only a few flows. Some of the flows contain both crustal and mantle xenoiiths but others have only crustal material. Using thin sections it has been possible to divide the rocks into five subgroups (Chapter 5); however, no one group was found to be distinctive or restricted enough in distribution to be used to differentiate individual flows, or groups of flows, on a regional basis. On the basis of normative mineralogy the the suite plots from the field of basanites to alkali basalts to olivine tholeiites. Based on combined chemical and petrologic grounds, the suite is classified as alkali olivine basalts (Chapter 6). Most samples can be classified as mildly alkaline, high aluminous, olivine basalts, but a small subpopulation is characterized as strongly alkaline to basanitic with normative nepheline values as high as 27 cation %. This distinctive chemistry appears to coincide mainly with the youngest flows and cones. The lack of distinct petrologic units required the assemblage be subdivided on the basis of morphology, elevation, and in the case of Holocene cones, on the basis of age. The formation name, 'Mann Creek' assigned by Walker (1930), is not used as it includes basalts from all of the subdivisions. The following descriptions are thus convenient groupings of particular flows and do not necessarily have 23 genetic or age implications (except for the Holocene cones). Basalts of the Wells Gray—Clearwater area have subdivided into 3 main categories (Fig. 3-1, Plate 1): 1) Valley-filling and plateau-covering flows which are subdivided into the informal Clearwater and Wells Gray units on the basis of elevation. 2) Ice and water contact features which have evidence of a discrete source separate from the valley-filling flows. 3) Holocene and older cinder cones and associated flows. Flows in categories 1 and 2 are Pleistocene in age. Morphologic characteristics are given for each unit but detailed petrography and petrology are given in Chapters 5, 6 and 7. Nomenclature for naming of the fragmental material follows that of the lUCS classification given in Schmid (1981). 3.2 VALLEY-FILLING AND PLATEAU-CAPPING UNITS 3.2.1 CLEARWATER UNIT The Clearwater unit consists of valley-filling flows as well as flows that blanket parts of the adjacent Murtle Plateau (Fig. 3-1, Plate 1). The flows are predominantly subaeriai but are associated with deposits of hyaloclastites consisting of pillows, block and lapilli breccias (composed of broken pillow fragments—pillow breccias) and tuff breccias. Intercalated with the flows are paleosols, fluvial, glacial and colluvial deposits. Where the upper surface of the Clearwater unit is exposed, evidence of glacial erosion is present in the form of striations and chatter marks. Most commonly the upper surface is covered with glacial, alluvial or colluvial material. How much of the lava may have been removed by erosion is not known, but in the vicinity of Spahats Creek (Plate 1), the upper surface of the flows has a 24 G E N E R A L I Z E D G E O L O G Y OF THE W E L L S G R A Y - C L E A R W A T E R A R E A , B . C . , 1 2 0 ° 1 5' W i a o ° 0 0,W 1 1 9° •» 5 ' w  H O L O C E N E A S S E M B L A G E |;X-l-j S u b a e r i a i basal t P L E I S T O C E N E A S S E M B L A G E S | i j 11  j S u b a e r i a i b a s a l t •W Pseudo-craters I L a c u s t r i n e d e p o s i t s •jf- Subaeriai pyroclastics •y.£>"\ Ice c o n t a c t d e p o s i t s Figure 3 - 1 : Main morphological units in the Wells Cray—Clearwater area. Thick, dashed lines are the traces of paleo-river channels preserved by infilling with predominantly subaeriai flows. 25 stepped appearance which may indicate at least 50 m of basaltic flows have been stripped. A conservative estimate of the total pre-erosion volume of the flows is 25 k m 3 . Distribution Exposures of the unit are limited to cliff-forming remnants along the river and lake valleys (Plate 1, Fig. 3-1). The flows are interpreted to form an intermittent mantle across the Murtle Plateau where evidence for underlying volcanic rocks is restricted to aeromagnetic data and to boulder counts in the overlying glacial debris. Based on aeromagnetic data, flows on the plateau surface have infilled pre-existing topographic irregularities and paleo-river channels (Fig. 3-1) The flow are not in-themselves the cause for the broad plateau surface, being absent or forming only a thin veneer in many places. Thick infillings of basalt in paleo-valleys are shown on aeromagnetic maps as magnetic 'highs', several orders of magnitude above the magnetically 'low' rocks of the Shuswap Metamorphic Complex and Raft Batholith. This difference in magnetization is because of the high concentration of magnetite in the basaltic rocks as compared to the more silicic metamorphic and plutonic rocks in the region. Present-day rivers have formed alongside the flow margins, downcutting through the flows into the basement rocks. As a result the thicknesses of basaltic rocks exposed in the cliffs are not necessarily indicative of thicknesses elsewhere in the unit. At the present position of Helmcken Falls, the Murtle River appears to have cut directly into the axis of a paleo-valley (Fig. 3-1). As a consequence 150 m of flows are exposed here, the greatest thickness known within the region. 26 Figure 3-2: View southward down the Clearwater Valley as it crosses the Murtle Plateau. The Clearwater unit is exposed in cliffs along the river. Figure 3-3: View eastward of cliffs in the lower Clearwater Valley. Flows are thin and pinch out laterally along the cliff face. Distinctive horizontal lines are intercalated sediments. The grottos present along the cliff face form by spring seepage along the base of the flows. 27 Basal deposits Exposures of the basal surfaces of the flows are commonly associated with springs and seepages. Extensive tufa deposits have been built around some springs, such as those in the upper Hemp Creek valley and at Red Spring, Ray Spring, Meadow Falls and 3rd Canyon Creek. Where springs exit from the base of a cliff, centralized erosion around the seepage has formed large grottos in the cliff face. A prime example is The Shadden, but others occur elsewhere (Fig. 3-3, Plate 1) Morphology Cooling units are generally 1-3 m thick and show poor to moderately well developed columnar jointing (Fig. 3-3). Most flows greater than 3 m in thickness have a lower colonnade of stout basal columns (Fig. 3-4), an entablature and an upper colonnade of more slender vertical columns. Pipe vesicles and spiracles are found near the base of many flows and vesicle trails are present in the middle, massive portions. Vesicles, abundant near the base of the flows, tend to be smaller, spheroid and more regular than those near the upper cooling surface. These upper vesicles are irregular and gas cavities up to 10 cm across are found. Individual flows or cooling units appear to have limited lateral extent. When viewed in cross section the flows appear to pinch and swell. Flow top or flow bottom breccias are not extensively developed, but when present they can be up to 40 cm in thickness. In a few locations well developed ropy textures (Fig. 3-5) are present but, in general, an aa or blocky surface is most common. Thick sequences of pillows, poorly sorted ash-lapilli-block breccias (pillow breccia), and lapilli tuff breccias are found beneath the flows. In two areas, The Anvil and Whitehorse Bluffs, these deposits are extensive and will 28 29 Figure 3-5: Ropy flow surface exposed under the base of an overlying fl Exposure is 1 km upstream from Spahats Falls (Plate 1). Figure 3-6: Typical exposure of well developed pillows. 30 be discussed separately. In other areas the pillows and pillow breccias occur beneath as well as intercalated with the flows. The pillows range in diameter from 20 to 50 cm, and occur as both closely packed units (with less than five percent interpillow hyaloclastite) (Fig. 3-6), and as deposits where only sparse pillows are present in a matrix of Iapilli tuff or pillow breccia. In a few locations, the pillows form an open network of solid tubes and little or no interpillow hyaloclastite is present, ln these locations individual tubes can be traced for several metres in the cliff face. Peperite consists of sand, pebble and cobble size material of non-basaltic rock mixed with hyaloclastite (Snyder and Fraser 1963a). This mixture occurs as an interpillow matrix and in some cases forms laminated deposits within the pillowed sequence. In one unusual occurrence, clay appears to have been squeezed up from below after deposition of the 1 pillows (Fig. 3-7). The pillows were however, still hot, and the clay was baked onto the cooling pillows. The pillows all show well developed glassy margins, in places up to 3 cm in thickness. The interpillow hyaloclastite is rich in glassy fragments (sideromelane) up to a centimetre in size. The sideromelane weathers to palagonite. The base of the pillowed sections can be examined at a number of locations (Plates 1 and 2). The most common basal deposits consist of unconsolidated boulder conglomerates (lag gravels) and laminated silts and sands, intercalated with cobbles. Most clasts appear to be locally derived from schistose, gneissic and granitic rocks of the Shuswap Metamorphic Complex and granitic-granodiorite rock of the Raft Batholith. Clasts from several unnamed plutonic bodies that outcrop north of the Murtle River are also present. Some of the basal deposits consist of basaltic cobbles and boulders, petrographically similar to the overlying flows (Fig. 3-8). These older 31 Figure 3-7: The base of a pillow deposit in which underlying finely laminated clay has been squeezed up between the pillows during their deposition. Heat from the cooling pillows has locally 'fired' the clay, forming resistant knobs baked to the pillow surface. 32 Figure 3-8: Rounded basalt boulders exposed below a flow in the Clearwater Valley downstream from Hemp Creek. 33 flows once occupied the Blackwater— Hemp Creek valleys but have now been largely removed by erosion. At a location 1 km upstream from Spahats Creek, the underlying deposits appear to consist of colluvium in the form of a talus cone shed from the sides of the paleo-valley. Less than a kilometre north of this location the underlying deposit consists of a matrix supported diamicton composed of striated and faceted cobbles (Fig. 3-9) and interpreted to be a glacial till. The Flatiron The Flatiron is an eroded feature in the Hemp Creek valley that consists of at least one cooling unit with a total observed thickness of 50 m. The basalt is nonvesicular and there is no indication of breccia layers within the unit. The top surface has been glaciated and the basal contact, seen in 2 locations, rests on bedded fluvial material. The basal material, composed of at least 50% basaltic clasts, rests on highly disrupted and altered blue-green phyllites of the Kaza group (as defined by Campbell "and Tipper 1971). The flow(s) that forms The Flatiron has perfectly developed columns, 10 to 40 cm in diameter, which form complex patterns. The columns change orientation from vertical to horizontal in the space of a few metres or centimetres (Fig. 3-10). In one location perfectly developed 10 - 15 cm diameter, vertical columns were traced over a distance of 20 m. The columns show no 'chisel' marks and horizontal fractures are sparse. On the fringes of the area thinner (2-3 m) cooling units with 40 - 50 cm diameter columns are found. In some locations these units have a well developed horizontal (platy) fracture. Downstream from The Flatiron, thick deposits of pillows and pillow breccias underlie subaerial flows. The hyaloclastite forms a foreset bedded 34 Figure 3-9: Exposure of a diamicton which occurs below a pillowed deposit and subaeriai flows (not shown) in the Clearwater Valley. 35 Figure 3-10: View of east facing cliff face at the The Flatiron. Variation in orientation of the columns is marked by the horizontal benches. The irregular surface directly above the talus slope are horizontal columns oriented perpendicular to the cliff face. Figure 3-11: Contact between subaerial and subaqueous bedded pillow delta. The bedding in the pillow delta dips southward, directly downstream from the Flatiron. 36 deposit which, in the region of The Flatiron, dips southward (Fig. 3-11). Farther downstream the beds dip to the north-west, swinging to the south-west in the area of Moul Creek. These foreset-bedded deposits are exposed continuously from below The Flatiron to a point in the Clearwater Valley 2 km downstream from its junction with Hemp Creek. The base of the pillowed section is visible at a number of locations. At Moul Creek, the pillow deposit rests directly on phyllite on the south side of the creek and on well-bedded sands and gravels on the north side. Basal lag gravels are found south of this point but the pillowed sequence becomes thin and discontinuous, gradually being replaced by pillowed sections within the subaerial flows. At 3rd Canyon Creek a channel, more than 5 m deep, is cut into granodiorite of the Baldy Batholith. The channel fill consists of a lag gravel composed of well rounded boulders of basalt, metamorphic and plutonic rock up to 50 cm in diameter. The lag gravel fines upward into cobble size, well rounded to subangular material, in a coarse granular matrix. These fluvial gravels are overlain by a layer of pillows a few metres thick, and a flow at least 10 m in thickness. Several interpretations of the The Flatiron are possible but the most favoured is that the basalt is an erosional remnant of an exceptionally thick flow (for the region). Based on the exposures of basement rock and basal deposits of fluvial gravels, the flow probably filled a paleo-stream channel or topographic low that was present on the west side of Hemp Creek valley. Exposures of stouter columns represent the lower colonnade of a flow in excess of 50 m in thickness. A similar relationship was described from the Yakima Basalt subgroup by Swanson and Wright (1978). This interpretation is supported by aeromagnetic data which does not indicate any unusually high concentration of magnetite rich material in the subsurface as would be 37 expected if the area was an eruptive centre and basaltic rock was present below the surface. The evidence suggests the following scenario for the evolution of The Flatiron. Tertiary sediments and volcanics (correlative with the Skull Hill and Chu Chua formations and mapped by Campbell and Tipper 1971) formed a topographic high in the paleo-Hemp Creek valley. The Flatiron lavas were ponded behind this barrier or constriction in the valley. This portion of the Hemp Creek valley remained a barrier and Hemp Creek was re-established along the east side of the valley where maximum runoff from the Battle Mountain area was channelled. With establishment of Hemp Creek, erosion of the lavas commenced and clasts were transported and deposited downstream in the Clearwater Valley. The paleo-Hemp Creek valley, excavated through the 1.4 Ma lavas, was then filled by flows during eruptions about 0.55 Ma. Flows filled the paleo-channel and blocked the Clearwater River forming a lake upstream from the position of present day Moul Creek. Lava then entered the paleo-Clearwater Valley (the axis of which was to the east of the present channel) and flowed both northward, into the growing lake, and southward down the valley. As flows built upward, they spilled over the topographic high in the vicinity of the present day Flatiron and built a pillow-lava delta in the growing lake. The situation was dynamic; as the eruption(s) progressed, water would have continued to enter the drainage system from the valley walls and upstream sources. The lake, backed up in the Clearwater Valley, may have overtopped or broken through the lava dam, resulting in pillowed sequences intercalated with the subaerial flows at points downstream. Subsequent to these eruptions, streams were established on both sides of the valley (Plate 1) but the most vigorous of the two is Hemp 38 Creek, on the west side of the valley. Maximum runoff in the valley is, however, from the east side. Streams established along this side have incised deep channels that cut from east to west across the valley to join Hemp Creek (Plate 1). The re-establishment of Hemp Creek as the principal drainage may have been facilitated by fault motion parallel to the valley where weakened and disrupted Tertiary and older rocks are exposed in the stream bed. Recent faulting in the valley is suggested by the contact between subaeriai flows and pillowed material. These deposits are well exposed on the west side of Hemp Creek valley (Plate 1) where the contact between subaeriai and subaqueous deposits occurs at an elevation of 700 m. On the opposite side of the valley, in an equally good exposure, the contact is at 640 m. Thus, the south-east side appears to have been dropped down relative to the north-west side. A similar effect is possible if the lake created by the lava dam was progressively filled with hyaloclastite from the east side. Rising water levels would shift the subaerial/subaqueous contact from lower elevations in the east to a higher one in the west. However, this scenario is inconsistent with dip directions and with the uniformity in elevation of the contact from Moul Creek to The Flatiron. The deposits on the east and west sides of Moul Creek may, however, represent two generations of flows. Subaeriai basalts are found up to an elevation of 700 m between Hemp Creek and Moul Creek. These flows may correlate with the 1.4 Ma flows near The Flatiron and with those on the west side of Hemp Creek valley. Flows in the vicinity of Moul Creek and the lower reaches of Hemp Creek may postdate the earlier flows by some 1 million years. Other suggestions of Cenozoic faulting are present within the the Hemp Creek valley. Eocene sediments (Chu Chua Formation; Campbell and 39 Tipper 1971) have been rotated from subhorizontal to vertical attitudes, and Paleozoic and Precambrian rocks of the Kaza Group are highly fractured and disrupted. Several extensive exposures of bluish-green phyllite are present in the area. The phyllite is broken, shows no consistent orientation and is highly altered, with calcite on most fracture surfaces. This material has been interpreted as phyllonite (D. McMullen, oral communication 1986). Quartzite breccias have also been found. Although these features may be the result of much older movement, it is possible that reactivation in the Quaternary has occurred. Whitehorse Bluffs Whitehorse Bluffs is a promontory that projects westward into the Clearwater Valley between Mahood River and Hemp Creek (Plate 1; Fig. 3-12). The southwest face is a near-vertical cliff 275 m high. The predominant rock type is well-bedded lapilli tuff breccia composed chiefly of quenched basalt fragments, sideromelane, olivine crystal fragments, and 1-2% clasts of plutonic and metamorphic rock. The breccia, predominantly a yellowish colour (2.5Y8/2-6/4), consists of layers, a few centimetres to tens of centimetres in thickness, that both fine and coarsen upward. Some beds show small-scale, cut-and-fill structures and low-amplitide ripples (Fig. 3-13). Slumping is also evident, and one clastic dyke was noted. The tuff breccia is friable in the upper parts of the section where it is composed of upward coarsening layers, 10-30 cm thick, of reddish frothy basalt lapilli. These layers are intercalated with 3-4 cm thick black layers of denser basalt granules. Below these laminations the top few metres of buff coloured material contains ropy or scoriaceous basalt bombs that show sag structures (Fig. 3-14). Deeper in the pile alteration and zeolitization has occurred. The presence of phillipsite, smectite and analcite was determined by x-ray 40 Figure 3-12: View looking eastward toward Whitehorse Bluffs from the Clearwater River. Dark bedded layers at the top are subaerial lapilli breccia and rest conformably on well-bedded subaqueous lapilli tuff breccia which is intruded by high level dykes. ' 6 Figure 3-13: Graded beds and small scale cut-and-fill structures are present in the lapilli tuff breccia which makes up the bulk of the deposit that forms Whitehorse Bluffs. 41 diffraction. The zeolitized lower section is well lithified and fractures break across grains rather than around them. Jointing occurs as vertical planes spaced a few metres apart. Erosion along these joint planes forms the prominent cliffs. The bedded lapilli tuff breccia is cut by numerous dykes, up to 10 m thick, which form an anastomosing network throughout much of the pile and create a spectacular array of columnar jointing. Some dykes show evidence of late emplacement fracturing. All dykes have quenched margins and many have small apophyses. Siliceous basement rock is present as angular xenoliths up to 40 cm in diameter, within the dykes. Basal fluvial sediments are exposed in one location. These sediments are made up of at least 6 m of poorly sorted sands and gravels. The maximum clast size noted was 15 cm. Sand-size grains are angular to subangular and are predominantly quartz and feldspar, but some rock fragments are present. Granule, pebble and cobble size clasts consist of rounded to subrounded fragments of metamorphic, granitic and basaltic material. The overlying contact with the tuff breccias is a scoured surface. The sandstone is more consolidated at this point than lower in the section, possibly because of heating by the overlying hyaloclastite. A quenched dyke cuts the sediments and numerous sand and pebble size material is embedded in the surface of the dyke. At the south side of the bluffs close to Green Mountain, two flows of subaeriai basalts are exposed. These flows can be traced south-eastward from the bluffs between 790 and 700 m elevation, but the base is exposed only at the southern end where the units wedge out against the paleo-valley wall. There the overlying flow contains large spiracles and the quenched base rests on a chaotic deposit that appears to be composed of subaeriai bombs 42 ,h"14K D ' ? U r b e d , b e d d , ' n g a r ° U n d 3 c^oriaceous, subaeriai bomb developed when the bomb was dropped onto the unconsolidated lapilli tuf!f breccia 43 as well as pillows. Under the pillow/bomb layer, a clay-rich deposit of granular material is exposed comprising basalt lapilli and scattered, 2-3 cm, elliptical ultramafic nodules. This deposit overlies a crudely bedded, clast supported, layer of metamorphic rock fragments which in turn overlies a clast-supported diamicton composed of angular metamorphic fragments and subrounded granitic clasts. Below 700 m elevation, pillows and pillow breccia are intermittently exposed. The Whitehorse Bluffs area is interpreted to be the result of phreatomagmatic eruptions. The 700 m elevation of the pillow/subaerial flow contact on the north-west side of Hemp Creek is coincident (within 10 m) with the pillow/subaerial contact at the south end of Whitehorse Bluffs. The K-Ar date from a sample from Whitehorse Bluffs and from a flow downstream of Hemp Creek (Chapter 4), are the same within the analytical error. It is possible that the deposits which form Whitehorse Bluffs, were erupted into a lake dammed by lava downstream at Hemp Creek. Water in the subsurface entered the conduit system. This water initiated explosive eruptions that blasted ejecta into the underlying sediments and water. This material, composed mainly of quenched basaltic fragments, was deposited as turbidity flows and normal lacustrine sediments, built a growing pile of ejecta. As the debris pile built upward, explosions became less intense and block sized material survived to be flung out and dropped into the unconsolidated sediments, deforming them. Finally, water was excluded from some, but not necessarily all, of the conduits feeding the eruption. From this point, a mixture of phreatomagmatic and normal strombolian eruptions occurred and the edifice was built a further 60 m above the water surface. Lava flows were then fed by the network of dykes that had penetrated the pile. Streams entering the valley along the slopes of 44 Green Mountain may have reworked some of the marginal deposits and introduced a greater proportion of accessory clasts along the east side than is typical of the pile in general. Source of the lava flows With the exception of the Whitehorse Bluffs body, the flows of the Clearwater unit do not appear to have originated from a single centralized vent. Flow indicators within the cooling units are erratic and give no consistent flow pattern. Topographic considerations dictate a common source would have to be either at the south end of Clearwater Lake or the east arm of Murtle Lake. Neither of these seem likely as no flows appear to be present northward, along the edges of Clearwater Lake, or on the eastern shores of Murtle Lake. Just to the north of Pyramid Mountain (Plate 1) a volcanic centre is inferred from topography and by large basalt boulders downslope. These boulders contain abundant ultramafic nodules up to 7 cm in size, and megacrysts of pyroxene. The presence of xenoliths indicates that the lava is close to the source, as the large dense nodules drop out of low viscosity basaltic flows within a short distance. This centre could have fed flows into paleo-Blackwater Creek and possibly the Helmcken Canyon area of the Clearwater River, but it does not appear to be large enough to have erupted flows over a time span of several hundred thousand years. The flow of the Clearwater unit may have erupted along fissures within the valleys. No evidence of dykes cutting the flows was found, with the exception of the Whitehorse Bluffs area and the upper Wells Gray unit (see section 4.2). These fissure eruptions, if they occurred, might have been similar to those described by Rittman (1962) from the eruption of Threngslaborgir in Iceland. The fissure may have been only a few kilometres long, but the ratio of fluid lava to fragmental material was considerable. 45 What little fragmental material that might have built up around the fissure would have soon been eroded away. Two uncertainties remain in the case for fissure eruptions. First, eruptions during the intervening time appear to have been localized, forming the major tuyas of the area (see Section 3-3). Second, flows from the Holocene cinder cones are of low viscosity and far travelled (up to 14 km from the vent area). These Holocene eruptions produced small volumes of flows (<1 km 3 ) but because the cones are very susceptible to erosion and the flows are not, the flows are preferentially preserved. The deposits of the Clearwater unit may then have built up gradually from similar small, localized eruptions, that infilled topographic lows in the vicinity of the eruption and under favorable conditions flowed downstream for some greater distance. 3.2.2 WELLS CRAY UNIT The Wells Cray unit occurs topographically above the Clearwater unit but the two lavas are morphologically and chemically indistinguishable. The Wells Cray unit occurs as discrete plateau surfaces and erosional remnants along valley walls at elevations above 1200 m. The basalt flows underlie the broad upland plateau surfaces of Stevens Lakes, Fight, and Philip lakes, and areas to the west of Kostal Cone. The unit also occurs as erosional remnants along the valley walls of Mann, Battle and Philip creeks. In the plateau areas, flows are exposed only at nickpoints where outlet streams drop over the edge of the plateau surface. Stevens Lakes The Stevens Lakes area is underlain by flows of the Wells Gray unit that are exposed at Hemp and Snookwa creeks. The plateau area is notable for its myriad of small lakes and marshes with no outlets. The irregular shorelines of the main lakes are controlled by heaps of glacial debris 46 composed of angular to subangular basalt blocks. Islands in the lakes consist of large basalt blocks. At Snookwa Creek, where a minimum K-Ar date was obtained, two or three, 1 - 2 m thick flows rest on a glaciated granitic surface. The flows have also been glaciated and are covered by bouldery till made up of granitic and basaltic clasts. In the upper Hemp Creek Valley at least five cooling units, each 1 - 3 m thick, are exposed. The base of the flows is not seen in this valley, but numerous mineral springs mark the contact at 1130 m. Intercalated subaeriai flows and pillows are exposed between McLeod Hill and Hemp Creek at 975 m. Though exposure is poor, these are interpreted as the downvalley continuation of the flows from Hemp Creek. No source for the flows in the Stevens Lakes area has been found, but on the aeromagnetic map a strong anomaly exists under a promontory that is elevated 75 m (Plate 1) above the general level of the lakes. No outcrop could be seen in the area. The entire Stevens Lakes region shows a fairly strong aeromagnetic anomaly that is confined to the valley bottom. This strong signature may either represent considerable infilling of a paleo-valley by basaltic flows or suggest that the flows were fed by fissure eruptions within the valley. Fight Lake South of the Stevens Lakes, additional flows of the Wells Cray unit are poorly exposed in the vicinity of Fight Lake and to the south. There is no apparent relationship between the Stevens Lakes area and this area. The area is covered by glacial debris that contain cobble-size, subrounded clasts of basalt, but these make up less than 50% of the deposit. Several magnetic anomalies coincide very well with the area of mapped basalt flows. The exception is an anomaly on Round Mountain. In a search for possible 47 source regions, this area was traversed, but no evidence of flows or dykes could be found. It seems most likely that flows in this area are all related to a source around Fight Lake that flowed into a paleo-valley at Cariboo meadows continuing downslope into the Philip Creek drainage. 3.3 ICE-CONTACT FEATURES Ice contact features can be grouped into three categories: tuyas, subglacial mounds (SUCM) and ponded, valley-edge deposits (VED). All of these are interpreted to have formed by the interaction of basaltic magma with melting glacial ice as opposed to water ponded by bedrock topography, landslides or flows. Chemically, lavas from these centres span compositions similar to those from the Holocene cones as well as from the Clearwater and Wells Cray units. Petrographically, the flows from the ice-contact features contain more quenched material. In thin section, swallow-tail crystals are common, as well as dendrites and amorphous opaques (type B, B 2 , and B 3 Chapter 5; Hickson and Souther 1984). Intersertal glass is prevalent and sideromelane rims on pillows are up to 2 cm thick. Several of the centres contain abundant ultramafic xenoliths and crustal xenoliths were found at all centres. 3.3.1 TUYAS Volcanic hills composed of thick successions of hyaloclastite are present at several locations within the study area (Plate 1; Fig. 3-1, Fig. 3-15). They include flat topped hills of hyaloclastite capped with subaerial flows and conical mounds with no associated flows. The flat-topped structures are tuyas as defined by Mathews (1947) from the Tuya-Teslin area (Watson and Mathews 1944) of British Columbia or table mountains described from Iceland by Sigvaldason (1968), Van Bemmelen and Rutten (1955) and others. The conical mounds are more enigmatic. 48 Gauge Hill, Mosquito Mound, and McLeod Hill are flat topped structures, heavily vegetated, rising 245 to 345 m above their surroundings and are one to two km in diameter. Limited exposures on the lower slopes reveal hyaloclastites. These deposits are capped by subaeriai flows which form escarpments, a few 10's of metres in height, around the upper surface of the edifice. Small enclosed lakes and marshes are present on the upper, drift covered, surface. Mosquito Mound and McLeod Hill are elongate in a north-south direction and Gauge Hill is roughly circular. The elongation is subparallel to the trend of glacial flow indicators from the Fraser Glaciation (Campbell and Tipper 1971). The best exposed tuyas are Mount Hyalo and Fiftytwo Ridge. Mount Hyalo (Fig. 3-15) has undergone sufficient erosion that the internal structure of the centre is exposed along the eastern side and in an eastward facing cirque. The lower portion of the deposit consists of laminated lapilli tuff breccia and pillow breccia interlayered amongst massive, intact pillows. Some of the material appears very similar to flow till and consists of a high proportion of striated and faceted accessory clasts as well as striated, and subangular basalt clasts. Most of the basalt clasts are dense, but others show typical quench textures. The matrix consists of a granular mixture of rock and mineral fragments, and sideromelane. The amount of basaltic material varies in these outcrops. In some cases till grades into typical hyaloclastite where all clasts are angular and show evidence of quenching. Two dykes are exposed along the eastern flank. Each has quenched margins of up to 3 cm of sideromelane, and well developed columns 15 cm in diameter. Above 1800 m elevation at least four, 3 - 5 m thick, subaeriai flows are exposed. These flows contain up to 7% accessory clasts of quartzite and gneissic material. The xenoliths are up to 20 cm in length. In most cases mafic crystals in the xenoliths have been completely melted. Below the lowermost flow, finely laminated, dusky red to reddish-brown (2.5YR3/2-5/4) material, 40 cm in thickness, is exposed. 49 Figure 3-15: Mount Hyalo, to the west of the Murtle Plateau, is the best preserved tuya in the Wells Cray—Clearwater area. It displays the classic tuya features of a flat top, formed by subaerial basalt flows, and flanks of bedded, steeply dipping, deposits of hyaloclastite. 50 It consists of glassy granules and frothy chunks of basalt. The lowest flow is very dense, shows flow banding with accessory xenoliths aligned parallel to the flow banding. Above this flow, a 15 m layer of bright red, frothy lapilli tuff is exposed. Two more flows, poorly exposed and eroded, cap the peak. The topmost of these flows is highly oxidized and frothy. Fiftytwo Ridge is the smallest of the tuyas in the region, being only 125 m thick. At the base of the escarpment along its northern edge (Plate 1) is a well-bedded mixture of hyaloclastite with clasts of accessory metamorphic and plutonic rocks. Overlying the hyaloclastite very dark grey flow banded and mottled basalt is exposed. The appearance of the surface of this basalt is probably the result of variably devitrified glass. Sparse crustal and ultramafic xenoliths are present in the basalt which is very dense and fractures conchoidally. The ultramafic xenoliths are 5 mm or less in diameter, but crustal material measuring up to 7 cm in long dimension was seen. The upper surface of Fiftytwo Ridge is vegetated, but sparse boulders and cobbles of hyaloclastite, as well as metamorphic and plutonic material, show through the cover. Craters, up to 200 m across, punctuate the surface. In the north-facing crater walls, frothy, dark red (10R3/6) agglutinate and irregular bombs, up to 40 cm in size, are exposed. Many bombs are cored by clasts of basement material up to 10 cm in diameter. Small (5 mm size) ultramafic xenoliths are also evident in the more massive material. Fiftytwo Ridge is made up predominantly of flow material in contrast to the other tuyas within which hyaloclastite predominates. Mathews (1947), draws attention to the importance of maintaining an englacial lake which produces the hyaloclastite. Subaeriai flows form only after the pile of hyaloclastite tops the level of the englacial lake. Fiftytwo Ridge is located in a saddle on the flank of Battle Mountain (Plate 1) where there would be difficulty in ponding meltwater generated from the ice by the heat of erupting lava. Water would pond locally in topographic 51 depressions but would soon establish an easy, subglacial exit, into the Clearwater Valley. The glacial ice may also have been thin or quickly melted to form a large cavity above the eruption site. Into this cavity the eruption continued, producing subaerial agglutinate and flows ponded along margins with the ice. No single vent appears to have been established, but rather the eruption occurred from 2 or more fissures or associated vents, aligned along the long dimension of the ridge. 3.3.2 SUBGLACIAL MOUNDS AND OTHER UNCLASSIFIABLE CENTRES Four centres in the Wells Gray—Clearwater area cannot be classified as tuyas, but their morphological form appears to have been strongly influenced by glacial ice. The first of these, Ray Mountain (Fig. 3-16), appears to be the result of a subaerial/subaqueous eruption on a ridge crest at 2000 m elevation. Pyramid Mountain, a conical hill on the Murtle Plateau (Fig. 3-17), was first reported to be a cinder cone (Campbell and Tipper 1971) but lacks several characteristics of subaerial cones. Spanish Mump and Spanish Bonk were not visited by the author but work by P. Metcalfe (in preparation) indicates that Spanish Mump consists of hyaloclastite texturally similar to Pyramid Mountain. Spanish Bonk consists of massive basaltic material that may represent the eroded neck of a small eruptive centre. Ray Mountain Ray Mountain (Plate 1; Fig. 3-16), is the result of an eruption at 2000 m elevation on a ridge composed of gneiss and pegmatite. The volcanic material consists of hyaloclastites, agglutinate and subaerial flows. At the northern end of the centre, glacially polished pegmatite underlies a laminated lapilli tuff deposit 1 m thick. The matrix of the deposit consists of glassy basalt fragments. Clasts in the deposit range up to 10 cm in diameter; those greater than 2 cm are predominantly faceted and striated clasts of gneiss and pegmatite. On Ray Mountain, the lapilli tuff is overlain Figure 3-16: Looking up at the summit area of Ray Mountain (person for scale). Bedded subaerial and subglacial deposits are intercalated with ponded flows visible to the left of the picture. Light coloured material in foreground is altered metamorphic rock. Figure 3-17: View from Green Mountain northward across the Murtle Plateau. Pyramid Mountain, a conical subglacial mound (SUGM), is the prominent feature in the foreground. The notch on the east side represents the uppermost portion of a failed block which has been slumped and rotated. 53 by 3 m of laminated lapilli tuff-breccia/agglomerate composed of bombs up to 10 cm in size. Both of these units display the yellowish brown colour (2.5Y6/4) of sideromelane weathered to palagonite. Above this level, subaeriai flows (?) are interspersed between layers of agglomerate that contain spindle bombs up to 70 cm in diameter, and reddish black lapilli tephra. The flows are up to 5 m thick and are crudely jointed. The tephra and flows dip dominantly to the east at a low angle but the ridge was apparently irregular at the time of the eruption and infilling by flows and tephra occurred between high points to the south and north. Subsequent erosion and the formation of a cirque along the western half of the complex has exposed the interior of the volcanic centre. Dykes cut the tephra and appear to emanate from a massive, elongate, 10 m high, exposure of nonvesicular basalt that may represent the main vent. The basalt contains up to 4% crustal xenoliths, which range from angular fragments that show no evidence of incipient melting to clasts that are largely disaggregated from melting of hydrous mafic minerals. A minimum K-Ar date (0.03 ± 0 . 0 3 ) was obtained from the feeder dyke. The centre is interpreted to have formed sometime during the early advance of ice from the Fraser or a previous glaciation (Chapter 2). Ice covered the ridge crest, glaciating the surface and depositing a thin layer of till. A typical tuya did not form because of the narrowness and slope of the ridge upon which the eruption occurred. The initial eruptions quickly melted through the capping ice, producing only a thin layer of hyaloclastite, in a manner similar to Fiftytwo Ridge. Subsequent activity was subaeriai, producing tephra, blocks and bombs. Flows, however, were ponded by ice on either side of the ridge as the depths and dips of the flows are inconsistent with the topography. After the eruption substantial erosion 54 occurred and several cirques have been formed. Pyramid Mountain Pyramid Mountain is a conical mound, 240 m high, which is a prominent landmark on the Murtle Plateau (Plate 1; Fig. 3-17). It is constructed of crudely bedded layers of frothy, glassy clasts and angular pillow fragments. Intercalated with the pyroclasts are layers of sediment which consists of coarse sandy to granular, subangular to subrounded, crystals and metamorphic and plutonic rock fragments. More massive basalt clasts within the hyaloclastite contain ultramafic nodules, the largest found being 4 cm in diameter, as well as plagioclase and pyroxene xenocrysts (probable mantle origin) and scarce crustal xenocrysts and xenoliths. The present slope of the pile is nearly parallel to the dip of the beds which averages 26° . The exception is on the southeast side where the dip is up to 42° . The east facing slope is steep and scalloped and an area of hummocky terrain occurs directly downslope. North of Pyramid Mountain eskers occur. These ridges are made up of laminated coarse sands and diamictons which have a maximum clast size of 7 cm. Both glassy and massive clasts of basalt are found among the subangular to subrounded rock fragments which make up the diamicton. Polymictic till is exposed along the east side of the Murtle River and is similar to tills seen elsewhere, but is unusually well indurated. The fine size fraction of the till consists of rock fragments, broken crystals of quartz, feldspar, mica and glassy basalt shards. Pyramid Mountain is interpreted to have formed as a subglacial mound (SUCM) during the waning stages of the Fraser Glaciation. The initial eruption occurred subglacially. The englacial water depth may have been shallow as clasts appear to have been ejected subaerially, but deposited and 55 rapidly chilled in a subaqueous environment. When the water drained from the englacial lake, the unsupported slopes slumped on the north and east sides. The SUCM was then at least partially overrun by ice which cut a network of ice marginal channels around the base of the pile and deposited till to the south and east. The eskers were probably deposited at this time. From the number of ice stagnation features present in the area and the good preservation of the SUCM, glacial ice disappeared shortly thereafter. The Murtle River probably originated as one of a number of ice marginal streams. The Murtle River may have maintained its course because of the well indurated nature of the surrounding till (probably due to hydration of the glass shards). Its rather tenuous course across the Murtle Plateau may also be influenced by tilting of the plateau surface in the vicinity of Green Mountain, forcing more recent drainage northward, out of the older more established Blackwater Creek system. 3.3.3 PONDED, VALLEY-EDGE DEPOSITS Volcanic eruptions close to the side of a valley-filling glacier occurred at two locations in the Wells Gray—Clearwater area. Volcanic debris from the eruptions appears to have been confined to the side of the valley by glacial ice. These two locations, Jack's Jump and Sheep Track Bench, are among only five sites in the Wells Gray—Clearwater area where high level dykes were observed. Sheep Track Bench Volcanic deposits of Sheep Track Bench lie on the lower slopes of Trophy Mountains where subaeriai flows, 150 to 350 m thick, are intercalated with deposits of massive to well-bedded tuff breccia and pillows. This sequence overlies 350 m of well-indurated and cliff-forming, bedded to massive pillows, pillow breccia (Figs. 3-18 and 3-19) and lapilli tuff breccia 56 which rest unconformably on flows of the Clearwater unit and underlying basement rock. The beds dip westerly along the entire length of the exposure. Hyaloclastite consists dominantly of angular broken pillow fragments and isolated pillows in a matrix of granular sideromelane. Sparse subangular, faceted, and striated clasts of metamorphic and plutonic rocks make up <1% of the hyaloclastite. Layers of pillows, 40 to 60 cm in diameter, are found at both the base and the top of the sequence as well as intercalated within the lowermost subaerial flows of the overlying unit of intercalated flows and hyaloclastite. In the lower reaches of 1st Canyon Creek megapillows (diameter >2 m) and sills are surrounded by bedded hyaloclastite. Occasional scattered pillows preserved within the breccia cap hoodoos (Fig. 3-19). A striking array of such erosional forms occur in 1st Canyon Creek where the breccia is also intruded by numerous dykes. The entire Sheep Track Bench is cut by dykes, but the greatest concentration occurs in the upper reaches of 1st Canyon Creek. The number of dykes decreases from 1st Canyon Creek to 3rd Canyon Creek and no dykes are exposed to the south between 1st Canyon and Spahats creeks. The dykes range from 0.2 to 1 m in width. All dykes examined have quenched margins in which fragments of hyaloclastite are embedded. Some of the dykes have bulbous appendages and dykeiets and most have well developed columns 10 to 20 cm in diameter. A dyke exposed in outcrop at 2nd Canyon Creek displays all of these features (Fig. 3-20) and is locally referred to as the 'Garter Belt'. In the area of 2nd Canyon Creek, pillowed layers are more numerous, but unlike many of the pillowed deposits seen in the Clearwater unit, none in this area are of an open-framework type. Well-bedded hyaloclastite occurs within the massive to crudely bedded pillows and pillow 57 Figure 3-18: Well indurated hyaloclastite consisting of broken fragments of pillows in a granular matrix of sideromelane. 58 Figure 3-19: Well indurated deposits of hyaloclastite form vertical cliffs and hoodoos Uitf is 30 m high and the exposure is at 2nd Canyon Creek. 59 Figure 3-20: High level dyke cutting hyaloclastite near 2nd Canyon Creek. The dyke is locally known as the 'Carter Belt' and shows well developed columnar jointing as well as quenched margins. ' 6 60 breccia sequence. The well-bedded material consists of coarse sand to granule size grains in fining-upward layers 40 cm in thickness. The contact of the hyaloclastite and underlying rocks is exposed in two places. In 3rd Canyon Creek, the contact is composed of 50 cm of laminated sands and gravels that rest disconformably on a surface of smooth greenstone. These sediments appear to be fluvial and contain subangular cobbles of metamorphic and plutonic clasts, but no basalt clasts were found. The contact of Sheep Track Bench rocks with the Clearwater unit is exposed in a landslide scarp in the Clearwater Valley. The contact is unconformable and no secondary deposits appear to be present between the two units. The eruption of the Sheep Track Bench rocks occurred from dykes that were confined to the eastern side of the Clearwater Valley. The most plausible explanation to produce the thick sequence of tuff breccia is to confine the eruption products along the eastern side of the Clearwater Valley by a large valley-filling glacier. The first eruptions were close to the break in slope along the valley wall at the top of the Clearwater unit. These eruptions formed meltwater pools into which the quiet effusion of basalt lead to the formation of pillows. Sill-like units were intruded into the growing pile of hyaloclastite (Snyder and Fraser 1963a, 1963b) as dykes feeding the eruption penetrated the deposit. Dykes continued to penetrate the pile and later eruptions occurred at the top of the growing mound of debris. New pillows were formed as magma continued to be extruded, but these pillows rolled down the slopes of the growing pile, breaking up and forming the bedded pillow breccia deposits. Entrained debris, liberated from the overlying glacier as it was melted by the eruption, was incorporated with the volcanic clasts. The pile of pillows and breccia continued to grow until it exceeded the upper level of the englacial lake. At this point, subaeriai 61 flows were erupted but fluctuating water levels led to the formation of pillowed layers within the flows. The surface of the volcanic deposits still dipped to the west and well developed lava tubes formed in the lowermost subaeriai flows. These tubes probably fed pillows forming downslope at the water/air interface. The subaeriai flows form the prominent bench on the slopes of Trophy Mountains and have protected the underlying hyaloclastite from erosion. Overlying these deposits are debris flows, flow till and bedded material thought to originate from the eruption of Buck Hill Cone (Section lack's Jump lack's Jump is an ice-marginal pile similar to the Sheep Track Bench deposits. Here however, the eruption intially built a self-supporting pile of debris. As the pile grew, it overlapped the west slope of Killpill Mountain but maintained an independent summit area (Plate 1; Fig. 3-1). Jack's Jump is dominantly bedded tuff breccia, 600 m in thickness, which dip outward in a radial pattern around the summit area. Above 1370 m elevation where traverses were carried out (Plate 2), the deposit is a mixture of well bedded coarse sand to cobble sized clastic material, and massive to poorly bedded pillows and pillow breccia. The bedded material is generally poorly sorted, normally graded and in some cases cross bedded. It contains up to 6% accessory clasts of metamorphic and plutonic rocks. The maximum size accessory fragment noted was a subangular clast 10 x 5 x 3 cm. The contact between the bedded hyaloclastite layers and the pillow layers is sharp. Pillows and megapillows, up to 4 m in diameter, are isolated in a matrix of broken pillows and granular sideromelane. Dykes encountered above 1500 m elevation are very contorted. They have poor to well developed columns, quenched margins and ragged vesicles up to 4 cm 62 in diameter. Surrounding pillows are sparsely vesicular with vesicles only a few mm in size. Between 1700 and 1770 m elevation, exposures are scarce. Where outcrops were seen they are hyaloclastites containing a larger proportion of pillows than is present lower in the deposit. Above 1770 m elevation exposures are poor. Subaerial basalt flows are exposed in low escarpments. The total thickness of subaerial flows may be as much as 150 m. The eruptions at Jack's Jump may have been partly phreatomagmatic, which would account for a greater porportion of fine fragmental material to intact pillows than was noted at other locations such as Sheep Track Bench. A higher proportion of accessory clasts is also present in deposits at Jack's Jump than in the hyaloclastites at Sheep Track Bench. This may be the result of the eruption of Jack's Jump occurring under a more basal portion of the ice sheet. In this position the amount, or supply, of entrained glacial debris was greater than in other areas and the melted out material was incorporated into the volcanic deposit. 3.3.4 MAGMA AND WATER INTERACTION A generalized subglacial/subaqueous volcanic eruption sequence, similar to that proposed by Allen et al. (1982), can be envisaged from well exposed and dissected sections at Jack's Jump, Pyramid Mountain, Whitehorse Bluff, Trophy Mountains and Mount Hyalo. Their model for subglacial eruptions starts with quiet effusion of magma to form pillows at an ice or water depth in excess of 200 m. As the water or ice depth diminishes, generation of hyaloclastite tuff by phreatomagmatic eruptions occurs. This phase is followed by quiet, subaerial effusion of basalt flows when the water level in the englacial lake is exceeded. These subaerial flows enter the water at the margin of the volcanic pile and form a bedded pillow sequence. 63 In the Wells Gray—Clearwater area all of the above mentioned subglacial/subaqueous volcanic centres, except Pyramid Mountain, erupted at a water or ice depth in excess of 200 m (as determined from the contact between subaerial/subaqueous deposits), but a consistent pattern of eruption as envisioned by Allan (1980) and Allen et al. (1982) is not evident. The model developed for the Wells Gray—Clearwater area tuyas and ponded valley-edge deposits starts with quiet effusion of pillows. As the pillow pile increases in height, pillows fomed near the vent roll downslope and break up, starting to build a mixed pile of pillows and pillow breccia. Dykes, penetrating the pile form megapillows and sills. Sudden mixing of magma and water may occur in any one of the feeder dykes, resulting in phreatomagmatic explosions. The explosions generate angular glass shards and quenched basalt fragments that are ejected into the overlying water column. This fragmental material then settles out as either well-bedded sideromelane breccia or as turbidity flows. Melting of the glacial ice liberates englacial debris which is incorporated along with volcanic clasts into the deposit. During a period of eruptive quiescence, which may involve all or part of the volcanic centre, glacial ice can encroach upon the still cooling volcanic debris. In some places overriding ice will erode material and may deposit till. The englacial lake level may fluctuate markedly as subglacial meltwater channels are either newly formed or old ones choked off by ice movement or debris. This fluctuation in water level will result in intercalated hyaloclastite between subaeriai flows. In the case of Whitehorse Bluffs, magma in the conduit system was mixed with water, possibly from groundwater sources rather than the overlying lake. The resulting phreatomagmatic eruptions built a pile of volcanic debris within a few tens of metres of the lake surface. At this point water was completely or partly excluded from the conduit system and volcanic clasts were erupted through the water and into the air where they degassed, forming frothy, oxidized tephra mixed 64 with quenched lithic fragments. During the eruption of Pyramid Mountian a similar situation occurred in which shallow water depth produced mixed frothy tephra and quenched lithic fragments. At any one eruptive centre, the situation would be dynamic. The type and amount of juvenile volcanic material found in the deposits of an eruption will be more a function of water tables, vent or dyke configuration and magma expulsion rates, rather than strictly related to water or ice depth. 3.4 PLEISTOCENE AND HOLOCENE CINDER CONES Four Holocene pyroclastic cones are present in the Wells Cray—Clearwater area. Evidence of at least 6 others was found, but these have suffered moderate to severe erosion suggesting they may be interglacial. Of the Holocene cones, Spanish and Flourmill Cones consist of at least 4 nested pyroclastic cones but Kostal and Dragon Cones, are monogenetic (Basaltic Volcanism Study Project 1981). These young centres form the main subject of a Ph.D. thesis by P. Metcalfe (in preparation); Flourmill Cones formed the basis of the B.Sc. thesis of S.T. Bishop (1985). For this study, mapping and chemical analyses of these young volcanic centres was limited to that necessary to characterize the eruptive products in relationship to the older basalts of the region. The Holocene cones are enriched in rare earth and incompatible elements and have high Ni and Cr contents (Chapter 6, Appendix D). The basalts contain abundant ultramafic and crustal xenoliths, along with pyroxene megacrysts and scarce plagioclase megacrysts. On the basis of normative mineralogy, the samples plot from the field of alkali basalts to basanites in the basalt tetrahedron of Yoder and Tilley (1962). Despite the very high normative nepheline values, and in one case the presence of normative leucite, feldspathoids were not seen in any thin section. However, because of their high alkali content, these samples have been classified as basanitoid by Fiesinger (1975). 65 3.4.1 HOLOCENE CONES Kostal Cone Kostal Cone appears to be the youngest cinder cone in the Wells Cray—Clearwater area (Plate 1; Figs. 3-1 and 3-21). The present cone appears monogenetic; evidence of at least one older cone is present to the north, covered by >1 m of tephra. The base of the older cone has been overwhelmed by lava from more recent eruptions. At Kostal Cone, fire fountaining built a cone 150 m high. This phase of the eruption was followed by the quiet effusion of lava flows with scoriaceous, biocky tops . both from a breach in the cone wall and from the base of the north side of the cone. Early flows appear to have been confined to the vicinity of the cone but then overtopped topographic barriers and spilled into the McDougall Creek valley, flowing 5 km toward Murtle Lake. The collapsed east side of the cone appears to have been rafted eastward by the flows. Flows in the immediate vicinity of the cone are at least 12 m thick, based on depth sounding of Kostal Lake (Coombes 1985). A small constructional dome occurs in the crater. The age of the last eruptive activity at Kostal Cone is not known. Mature cedar trees, greater than 200 years old (C. Ted, oral communication 1985), cover the upper flanks of the cone (Fig. 3-21). Thick tephra close to the flows is vegetated, but the flows themselves show very little reforestation or soil development. Reports, by local hunters and trappers, of hot gases rising from the flows in McDougall Creek could not be substantiated. No material for radiocarbon dating could be found but the preservation of the flows is similar to that seen at Dragon Cone where a maximum age of 7 5 6 0 ± 1 1 0 yr by the 1 "C method was obtained (see Chapter 4). 66 Figure 3-21: Areal view of Kostal Cone showing the breached eastern section of the cone and surrounding flows. A small constructional dome is visible as the tree covered area in the centre of the cone. Figure 3-22: Rafted and rotated chunks of crust, up to 4 m in size, are common on the surface of flows from Flourmill Cones. 67 Basalt boulders, similar to the valley-filling basalts of the Clearwater and Wells Cray units, are found on the west ends of Kostal and Murtle lakes. A K-Ar age of 0.15 Ma was obtained from basalts exposed at McDougall Falls. These flows may have originated in the vicinity of Kostal Cone and represent a very early phase of activity, the source for which has been eroded away or buried by the Holocene flows. An older, eroded cone remnant occurs northwest of Kostal Cone (Plate 1) on the south facing flank of Goat Peaks. Flows produced by eruptions at this point may also be the source of the flows on the west side of Kostal Lake. Flourmill Cones Flourmill Cones consist of a cluster of three nested lapilli block/bomb agglomerate cones, the highest of which is 250 m (Plate 1; Fig. 3-1) (Bishop 1985). Remnants of a fourth cone are also present. The eruptions that produced the cones appear to have been sequential; each subsequent eruption partially destroyed or obliterated the previously developed cone. Tephra, up to 3 m thick, blankets a large area to the east of the cones and was found to rest directly on till. Flows appear to largely postdate fire fountaining and the formation of small constructional domes in two of the cone remnants culminated the eruptive activity. Large blocks of agglutinate are found south-east of the cones. One of these blocks is 50 metres long, 15 m in height, and 20 m wide. These blocks have been surrounded by more recent lava flows. No evidence of spatter or other tephra is found on the surface of the surrounding flows. The blocks of agglutinate may represent either an older buried cinder cone in situ or blocks rafted from the main cones. The flow surface has considerable relief, composed in part of rafted crust up to 4 m thick, rotated and upturned on the flow surface (Fig. 68 3-22). Well developed levees, up to 3 m high, are present in several areas. The flow margin is up to 10 m high in proximal areas, but diminishes to 1-3 m at distal points. The flows were extruded onto a hummocky, but relatively flat, surface of till, ice ablation deposits and fluvial sediments. Ice may have been present during the initial stages of the eruption and deflected the flows from the topographically lower, Flourmill Creek valley. A small region in the centre of the flows (Plate 1), is a kettle hole surrounded by finely laminated silt and gravel beds. Intercalated with these sediments are reworked lenses of tephra. Flows appear to have been channelled to either side of this location. At this same site, a flow overran areas of water-saturated, clay-rich sediment (C. Shook, oral communication 1984). Streams have now formed along the outer margins of the flow but nowhere could the flow base be seen. Spanish Cones Like Flourmill Cones, Spanish Cones are a series of nested pyroclastic cones. These cones are more subdued than Flourmill Cones and multiple ridges of agglomerate indicate that activity was episodic about the same vent. Lava welled up in the cones and flowed through breaches in the east and south walls, sending lava cascading down the slope (Plate 1; Fig. 3-1). The flows came to rest a kilometre downstream of the present outlet of Spanish Lake (which has formed due to damming by the flow). The margins of the flows are irregular and short runout segments terminate in blocky flow fronts. The flows stopped on comparatively steep slopes, and thus appear to have been more viscous than those from either Kostal or Flourmill cones. The basalt is extremely rich in crustal xenoliths. Some exposures have up to 10% angular crustal fragments ranging in size to 7 cm. Most of the clasts are unmelted gneissic and schistose material (Fig. 3-23). Ultramafic 69 nodules and megacrysts of pyroxene (Fiesinger 1975) are also present. The last phase of eruptive activity appears to have formed a 30 m deep explosion pit (Hickson and Souther 1984; their Fig. 8, p. 274). This may have coincided with the formation of three pseudo-craters (Basaltic Volcanism Study Project 1981) to the north-west (Hickson and Souther 1984; their Fig. 8, p. 274). There is no evidence of volcanic material in or around the craters but they probably formed by magmatic interaction with groundwater. The largest, Peridot Lake, may be in part a collapse feature due to magma withdrawal from a shallow chamber. Angular clasts of metamorphic rocks, up to 50 cm across, are found scattered about the surface of the lava flows within Spanish Cones. These clasts may have been ejected by the explosion that formed the pseudo-craters. Limited fumarolic activity followed these eruptions, leaving sulfur and calcareous deposits in at least two locations. The steep slopes on which flows from Spanish Cones rest has accelerated erosion in comparison to the valley bottom which is floored by flows from Flourmill Cones. The base of flow from Spanish Cones could be studied at several locations and was found to rest directly on finely laminated silt and sand, or on glacially polished bedrock and till. In one section, intercalated silty clay, ash-lapilli airfall material and reworked tephra rest directly on polished and striated gneissic basement rock (Fig. 3-24). Immediately north of Spanish cones stream erosion has exposed a 4 m thick section. An underlying till is overlain in sequence by; >40 cm of airfall lapilli-block/bomb tephra, 4 mm of silty sand, 45 cm of airfall ash-lapilli tephra consisting of at least 4, 5 to 10 cm thick, coarsening-upward layers, overlain by 30 cm of oxidized lapilli tephra and capped by a 10 cm, silty soil layer, with mixed bombs. Figure 3-23: Flows from both Spanish and Flourmill cones have an abundance of ^nL3rA: T h i n l y b 6 d d e d t 6 p h r a ( v i s i b l e b e h i n d t r o w e l>< underlying flows from Spanish Cones, rests on a glacially polished surface of gneiss. 71 Eruption of Spanish and Flourmill cones may have been simultaneous. Tephra, covering flows along the southern flank of Spanish Cones, appears to have originated from eruptions at Flourmill Cones. Multiple tephra layers are also found beneath flows from Spanish Cones, in close proximity to Flourmill Cones and may have originated from that source. As tephra from the individual centres cannot be distinguished, this evidence is inconclusive, but eruption of both centres during withdrawal of glacial ice from the valley, or shortly thereafter is indicated. 3.4.2 DRAGON CONE Dragon Cone (Fig. 3-25), directly east of Ray Mountain (Plate 1; Fig. 3-1), is a monogenetic pyroclastic cone formed from an eruption that occurred at an elevation of 2000 m. Flows, emanating from this cone, followed the course of Falls Creek to the Clearwater River a distance of some 14 km. Only the cone and distal sections of the flow were traversed (Plate 2). The cone was built of pyroclastic material from vigorous fire fountaining that lobbed bombs upward into the valley to the west, a vertical height of 250 m. The basalt is rich in both crustal and mantle xenoliths, but especially prevalent are megacrysts (xenocrysts?) of plagioclase up to 2 cm in size. The distal end of the flow has a scoriaceous blocky surface and prominent ridges and levees. Tree casts, up to a metre in diameter, are found in an upright attitude at the surface of the flow and horizontal at the base of the flow. A bark imprint in one of the upright, surface moulds, indicates that it was produced by a deciduous tree, but no carbonized remains could be found in this mould or in any of the others. Stream erosion along Falls Creek has exposed the flow base in several locations; everywhere the flow rests directly on fluvial laminated silts, sands and 72 gravels. Gas-escape structures are prevalent in the flow and some spiracles are over a metre in maximum dimension. In one location a peat layer is exposed below the flow base. It forms a layer 2 to 5 cm in thickness overlying at least 9 cm of laminated fine sand. The peat is overlain by 5 cm of intercalated clay, and fine to coarse sand. A 1 "C date on the peat gives an age of .7560 + 110 yrs. 3.4.3 PRE-HOLOCENE CONES Buck Hill Cone Buck Hill Cone, at an elevation of 1500 m, overlies basalts of the Sheep Track Bench deposits. The cone stands 100 m high and has been breached on the west side. Only sparse evidence for flows persist. The cone has been glaciated and erratic cobbles, boulders and till are found on the slopes. The till is well exposed on the lower slopes and consists of fusiform bombs (Fig. 3-31) in a matrix of clay and silt sized grains of mica, quartz, feldspar and rock fragments. Despite the fragile nature of the bombs, many are intact within the till. Downslope from Buck Hill Cone, extensive polymictic flow tills occur that are best exposed in the area of 2nd and 3rd Canyon Creek. Some of the layers contain significant amounts of subangular basalt clasts and quenched fragments but others contain no basaltic material at all. Samples from Buck Hill contain crustal and ultramafic xenoliths and pyroxene megacrysts, but they are not as prevalent as in samples from some of the other young volcanic centres. In the vicinity of Fage Creek (Plate 1), there are isolated remnants of a subaeriai flow which, on the basis of ultramafic nodules and pyroxene megacrysts, is thought to have originated from Buck Hill Cone. igure 3-25: Aerial view of Dragon Cone to the east of Ray Mountain. 74 The eruption that produced Buck Hill Cone appears to have occurred during the waning stages of the Fraser Glaciation. Ice, occupying the valley to a depth of 1500 m, appears to have quenched and disrupted any flows that were produced. Flow tills were being deposited on the slopes and quenched volcanic material was incorporated into these tills. Resurgence of glacier ice later overran the cone, forming a meltwater channel on the east side, possibly eroding the 'breach' in the cone on the west side and depositing the bomb bearing till. Other glaciated cones Several eroded cones (Plate 1) have been identified by airphoto interpretation and correlation with aeromagnetic anomalies. Of these, only Pointed Stick Cone and Ida Cone were visited. Pointed Stick Cone forms a subdued mound, 40 m high, at an elevation of 1550 m. The surface of the cone is vegetated and sparsely covered with erratic boulders of metamorphic and granitic rock. Weathered ribbon and fusiform bombs, in a sandy matrix are exposed on the surface of the cone. The ridge north of the cone exposes a glaciated flow, 5 m in thickness, with crude, 70 cm, columns. Upslope, at the northern end of the ridge from Pointed Stick Cone (Plate 1), two isolated exposures of basalt occur which are coincident with a large magnetic anomaly. Below these flows a basal layer of hyaloclastite was found and the flows themselves contain ultramafic nodules and olivine megacrysts. The presence of the nodules and the coincidence with a large magnetic anomaly indicates that these exposures may be close to a vent which has undergone extensive erosion. An area of weathered agglutinate, with abundant ultamafic nodules and crustal xenoliths on Ida Ridge is thought to be the remnants of a cone 75 (Plate 1). The largest known ultramafic nodules, up to 25 cm in diameter, in the Wells Cray—Clearwater area are found here. This locality was originally discovered by Ida DeCalvert who also reported a second location to the south. A search for this second outcrop was unsuccessful. 3.5 CONCLUSIONS Basaltic flows in the Wells Gray—Clearwater area appear to have emanated from spatially separated vents. Flows were fluid enough to flow down river valleys and cover plateau surfaces of low relief, but little pyroclastic material built up around the vent areas except when the region was inundated by ice. The ice served to confine the flows to a restriced area, producing tuyas, subglacial mounds and ponded, valley-edge deposits. Eruptions were sporadic and individually of small volume. Topographic depressions were filled rapidly and there is little evidence of a long hiatus in any one eruptive sequence. Paleosols are found only beneath basalt flows in the Clearwater Valley because its topographic position represents an area of accumulation rather than erosion. The ongoing nature of the volcanism produced a considerable volume of basaltic rock, estimated at >25 km 3 over several hundred thousand years. The presence of crustal xenoliths and the coincidence of the eruptions with the sides of valley walls and with lineations, suggests that the eruption of small batches of magma may have been facilitated by weakened zones (faults) in crustal rocks. These weaknesses provided ready access for small batches of magma to rise to the surface. 4. ISOTOPIC DATING Age, I do abhor thee, youth, I do adore thee. —William Shakespeare, The Passionate Pilgrim 4.1 INTRODUCTION Twenty-two potassium-argon (K-Ar) whole-rock age determinations were carried out on basalt samples from the Wells Cray—Clearwater area as well as one radiocarbon date on peat. Samples chosen for K-Ar dating are from the entire area of study (Fig. 4-1) and include at least two samples from each of the three morphological units (Chapter 3) as well as samples from areas of substantial exposure but no stratigraphic control. 4.2 POTASSIUM-ARGON GEOCHRONOLOGY 4.2.1 PROCEDURE Samples of alkali olivine basalt chosen for dating (Chapter 6; Appendix D) were reduced from 20-30 cm size blocks to a fist-sized core using a sledge hammer. The resulting fresh samples were then sectioned, and checked for alteration. Those chosen for dating had no evidence of devitrification of the matrix or alteration of olivine, the phenocryst phase. As basalts in the study area lack phenocrysts appropriate for dating, whole-rock samples were used. The validity of this procedure is shown in the work of Dalrymple and Lanphere (1969), Armstrong ef al. (1975) and Souther et al. (1984). The whole-rock samples were crushed using a jaw crusher and ground in a steel disk mill. The resultant material was sieved and the 60 - 80 mesh size retained for dating. Throughout the procedure, care was taken to avoid cross contamination or external contamination of the samples. The dating was carried out under Geological Survey of Canada contract to the geochronology laboratory of R.L. Armstrong at the University of British Columbia. 76 77 G E N E R A L I Z E D G E O L O G Y O F T H E W E L L S G R A Y - C L E A R W A T E R A R E A , 1 2 0 ° 1 5 ' W I Z O ' O O ' W  B . C . 5 2 u 1 5 ' 5 2 ° 0 0 1 1 9 4 5 ' W (o .S2 ) K - A r a g e . Ma (0 .30 ) C l e a r w a t e r (o.ia) 4 K i l o m e t r e s 5 2 15 N 5 2 0 0 ' N 5 1 4 5 ' H H O L O C E N E A S S E M B L A G E |;l;X-j S u b a e r i a l basa l t m Pseudo-c raters •j^ Subaerial pyroclastics P L E I S T O C E N E A S S E M B L A G E S S u b a e r i a l b a s a l t L a c u s t r i n e d e p o s i t s Ice c o n t a c t d e p o s i t s Figure 4-1: Generalized geology map showing locations of samples collected for K-Ar and 1 a C dating. Numbers beside the sample locations refer to Table 4-1. 78 Potassium determinations were done in duplicate by K. Scott on a AA4 spectophotometer. The argon determinations were done by J. Harakal. The whole rock samples were baked at 130°C for approximately 15 hours to reduce atmospheric argon contamination. The samples were then fused using a 60KW Phillips RF induction heater and spiked with high purity argon 38. The gas mixture obtained from the fusion was then purified by passing it over titanium "getter" furnaces. The argon isotopic ratios were then measured on an AEl MS-10 mass spectometer. The results are shown in Figure 4-2 and Table 4-1. Spike calibrations are performed against known aliquotes of "purified" air and against interlaboratory mineral standards such as the CL-O glauconite. The errors which are reported in Table 4-1 are for one standard deviation or for the standard error of the mean as in the case of the potassium analyses. 4.2.2 RESULTS Within the analytical error, there appears to have been almost continuous volcanism in the Wells Gray—Clearwater area for the last 600,000 years (Fig. 4-2). A greater number of young dates occur within the Wells Gray unit as opposed to the Clearwater unit, but this may be a sampling bias. Likewise dates in the Clearwater unit from the Murtle River northward are younger than those to the south. Only four dates of the 22, are older than 1 million years. Three of these samples are from deposits interpreted to be of ice-contact origin (Chapter 3). Samples S20104a (0 .50±0 .05 ) and S20104c ( 0 . 3 5 ± 0 . 0 9 ) are from the same location (Fig. 4-1) and were chosen to test for retention of radiogenic argon in quenched material. Sample "c" represents the glassy rim of a quenched pillow and "a" a sample of a subaeriai flow directly overlying the pillowed deposit. Though the samples give the same date within a 2 error, sample "a" contains more radiogenic argon than sample "c" (Table 4-1). This result is inconsistent with the findings of T A B L E 4 - 1 : K - A r A G E D E T E R M I N A T I O N S . S a m p l e K A r * A r * A g e ( M a ) L o c a t i o n L a t . L o n g . U T M E l e v . (wt .%) A r Z o n e 11 (m) 1) S 2 0 1 0 4 a 0 . 502 0 . 0 0 9 7 10 . 9 0 . 5 0 + 0 . 0 5 B1 a c k p o o l 5 1 ' ' 37 . 3 ' 1 2 0 ' 0 6 . 1 ' F N 9 9 4 2 3 3 1 2 8 0 2 ) S 2 0 1 0 4 C 0 . 4 9 4 0 . 0 0 6 7 4 . 3 0 . 3 5 + 0 . 0 9 B l a c k p o o l 5 1 ' ' 37 . 3 ' 1 2 0 ' 0 6 1 ' F N 9 9 4 2 3 3 1 2 8 0 3 ) S 2 0 5 0 1 0 . 6 2 3 0 . 0 0 0 4 7 0 . 3 0 . 0 2 + 0 . 0 2 M a n n C r e e k 5 1 ' ' 57 . 0 ' 1 2 0 ' 3 6 . 0 ' F N 8 1 5 3 1 9 5 0 3 4 ) S 2 0 1 0 2 0 . 6 9 4 0 . 0 0 5 0 2 . 2 0 . 18 + 0 . 11 M a n n C r e e k 5 1 ' ' 4 0 . 1 ' 1 2 0 ' 1 7 . 2 ' F N 8 7 7 2 7 4 1 2 4 0 5 ) S 1 0 7 0 6 d 0 . . 8 5 7 0 . 0 1 8 7 5 . 5 0 . 5 6 1 + 0 . 106 C l e a r w a t e r R i v e r 5 1 ' ' 53 . 8 ' 1 2 0 ' 0 7 . 0 ' F N 9 8 1 5 3 2 731 6 ) 3 2 2 0 6 0 . 7 0 0 0 . 0 0 7 3 6 . 4 0 . 2 7 + 0 . 0 5 S h e e p T r a c k B e n c h 5 1' ' 48 . 7 ' 1 19" 51 . 9 ' K H 9 4 2 4 4 2 14 15 7 ) S 1 0 8 0 3 0 . 9 12 0 . 0 1 8 4 6 . O 0 . 5 4 7 + 0 . 102 W h i t e h o r s e B l u f f s 5 1 ' ' 46 . 7 ' 1 2 0 ' O O . 8 ' G N 0 6 0 4 0 5 6 4 8 8 ) S 2 0 5 0 3 0 . 5 9 6 0 0 0 0 2 9 0 . 4 0 . 0 1 + 0 .01 C a n i m F a l l s 5 1 ' ' 52 . 0 ' 1 2 0 ' 3 7 . 0 ' F N 6 6 0 4 9 0 7 6 3 9 ) S 2 0 3 0 3 0 . 564 0 . . 0 0 6 1 6 . 2 0 . 2 8 + 0 0 5 H e l m c k e n F a l l s 5 1 ' ' 58 . 0 ' 1 2 0 ' 1 1 . 3 ' F N 9 3 2 6 0 7 748 1 0 ) S 2 0 3 0 7 0 . 6 8 0 0 . 0 0 5 3 2 . 0 0 . 2 0 + 0 .11 He 1 m c k e n F a l l s 5 1 ' 57 . 5 ' 1 2 0 ' 11 . 2 ' F N 9 3 3 5 9 9 6 7 5 1 1 ) 4 0 7 0 2 a 0 . . 7 5 3 0 . 0 4 0 9 12 . .6 1 . 4 + 0 2 F1 a t i r o n 5 1 ' 52 . . 7 ' 1 2 0 ' 0 2 . 8 ' G N 0 3 4 5 1 4 7 2 0 1 2 ) 3 4 6 0 5 1. 22 0 . 0 7 5 0 36 . 9 1 . 6 + 0 . 2 F i f t y t w o R i d g e 5 1 ' 56 . 1 ' 1 1 9 ' 5 2 O ' L H 0 2 3 5 7 6 2 0 6 0 1 3 ) 1 0 5 0 3 b 0 . 9 5 0 0 . 0 1 4 0 4 . . 4 0 . 3 8 + 0 . 10 M o s q u i t o M o u n d 5 2 ' O O . . 7 ' 1 2 0 ' 1 0 . 4 ' F N 9 4 2 6 5 7 9 1 5 14) 1 0 1 0 5 1. 18 0 . 1604 4 . 9 3 . 5 + 0 . 8 M c L e o d H i l l 5 2 ' 01 . 0 ' 120* 01 2 ' G N 0 4 4 6 6 9 1 2 2 0 1 5 ) S 1 0 9 0 1 0 . 6 9 5 0 . 0 0 2 4 3 . 0 0 . 0 9 + 0 . 0 3 Hemp C r e e k 5 1 ' 5 9 . 6 ' 1 1 9 ' 5 3 . 4 ' L H 0 0 8 6 4 0 1440 1 6 ) 1 0 2 0 2 c 0 . 5 6 3 0 . 0 0 0 1 8 0 . 3' 0 . 0 1 + 0 . 01 S n o o k w a C r e e k 5 2 ' 01 . 8 ' 1 1 9 ' 3 8 . 5 ' L H 1 8 8 6 7 3 1464 1 7 ) S 1 0 3 0 9 0 . 8 6 5 0 . 0 0 5 1 5 . 9 0 . 15 + 0 . 0 3 M c D o u g a 1 1 F a l l s 5 2 ' 0 5 . 7 ' 1 1 9 ' 5 1 . 6 ' L H 0 3 6 7 5 4 1037 1 8 ) S I 0 3 0 6 1. 17 0 . 0 1 2 6 6 . 1 0 . 2 7 + 0 . 0 5 G a u g e H i l l 5 2 ' 0 3 . 8 ' 120" 0 0 . 5 ' G N 0 5 1 7 2 3 8 7 2 1 9 ) 4 3 9 0 7 1. 0 5 0 . 0 0 0 5 8 0 . 2 0 . 0 1 + 0 . 01 M o u n t H y a l o 5 2 ' 0 7 . 2 ' 1 2 0 ' 2 1 . 4 ' F N 8 1 2 7 7 3 1 9 0 0 2 0 ) S 1 0 4 0 1 0 . 399 0 0 3 0 0 15 . 3 1 . 9 + 0 . 2 J a c k ' s J u m p 5 2 ' 0 6 . 2 ' 1 2 0 ' 0 3 . 7 ' G N 0 1 2 7 8 3 1768 2 1 ) 4 1 5 0 8 0 . 6 5 4 0 . 0 0 7 5 6 . 6 0 . 3 0 + 0 . 0 6 C l e a r w a t e r L a k e 52" 12 . 8 ' 1 2 0 ' 1 2 . 4 ' F N 9 0 8 8 8 3 6 8 0 2 2 ) 4 4 8 1 0 1. 02 0 . 001 1 0 . 3 0 . 0 3 + 0 . 0 3 R a y M o u n t a i n 5 2 ' 14 . 9 ' 1 2 0 ' 0 6 . 7 ' F N 9 7 1 9 2 3 > 1965 A r * = R a d i o g e n i c A r g o n ; X 10 6 c c / g m . A r = T o t a l A r g o n i n s a m p l e . R a t i o o f A r * / A r i s g i v e n a s a % . T T C o n s t a n t s u s e d : K'" = 0 . 5 8 1 X 1 0 ' ° y 1 ; K « " = 4 . 9 6 2 X 1 0 ' 0 y " ' : K'°/K = 0 . 0 1 1 6 7 % X e V.0 80 o o o o" o o LU < o o o o" o o CD < 0 • 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1 • 0.2 0.3 0.5 0.8 1.0 1.1 1.2 1.3 1.4 1.5 1.8 1.7 1.8 1.9 2.0 • 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 T 1 i 1 K-Ar Dates from the Wells Grey - Clearwater Area - i Note: Scale increments change i i T I I Clearwater Unit Weils Gray Unit Ice Contact Unit Subaeriai flows associated witn substantial deoosits ot tcreset bedded pillow breccia Figure 4-2: K-Ar dates wi th error bars. 81 Faure (1977) for oceanic basalts where it was shown that excess argon is trapped in the quenched rim of the pillows. However, the amout of radiogenic argon in sample S20104a is consistent with other samples yielding a similar age. It is possible that sample S20104c has undergone argon loss from the glassy matrix at atmospheric pressures and temperatures. Fiftytwo Ridge The date obtained from a sample from Fiftytwo Ridge (1.6 Ma) appears substantially too old when compared to the age suggested by the morphology of the ridge. Fiftytwo Ridge is interpreted to be a tuya whose constructional form is essentially unmodified by erosion (Chapter 3). It has little or no surface covering of till but glaciated boulders are scattered about the surface as well as within eruption craters which are surrounded by loosley cemented agglutinate. These observations are inconsistent with the 1.6 Ma date. It is possible that the sample has either not degassed due to rapid quenching, in a manner similar to that cited by (Faure 1977) for oceanic material, or contains excess argon due to the inclusion of radiogenically enriched crustal material in the sample. Crustal and mantle xenoliths and xenocrysts are prevalent in the basalt and many of these fragments are a few mm in size. As crustal material in the region has been reset to Eocene K-Ar ages and mantle xenoliths give Precambrian or older dates, inclusion of a small fragment may adversely affect the dating. Jack's Jump Jack's Jump (1.9 Ma) provides few morphological clues as to its age. It represents one of the most voluminous eruptive centres in the region, but has undergone extensive erosion. The sample chosen for dating is from what appears to be a subaeriai flow at the top of the volcanic pile. The sample 82 contains interstitial glass that may have affected the date (Faure 1977). If the date for Jack's Jump is considered invalid, a more plausible age may be contemporaneous with the eruptions of Mosquito Mound and Gauge Hill. During the 200,000 years suggested by dates from these centres, eruptions occurred while glacial ice occupied the valleys to a level of at least 200 m above the valley floor. McLeod Hill McLeod Hill is the most enigmatic of the "old" centres. There is a large analytical uncertainty in the date and the sample may also have suffered argon loss in a manner similar to sample S10104C as the Ar*/Ar is lower than would be expected for the age in comparison to other samples (Table 4-1) (J. Harakal, oral communication 1986) Thin sections reveal interstitial glass. McLeod Hill has the • stratigraphy and morphology of a tuya (Chapter 3), but it has been substantially modified by erosion. The dated sample is from a flow that was interpreted to have been ponded on the upper surface of the tuya. If the date correctly reflects the time of eruption, the temporal isolation of the centre may be the result of erosion. If other volcanic centres existed, both older and younger than McLeod Hill, they may simply have been eroded, leaving McLeod Hill, because of its bulk and resistance to erosion, to persist after the onslaught of multiple glaciations. Gibbons et al. (1984) calls on a process of "obliterative overlap" to explain the presence or absence of moraines in a glacial sequence. On the basis of probability analysis, they predict that for 10 successive glaciations, only three morainal deposits will survive. This concept might be equally well applied to volcanic flows in an area of multiple glaciation. Obliterative overlap, as well as sampling problems, undoubtably accounts for some of the 83 other gaps in the volcanic record in the Wells Cray—Clearwater area. 4.3 RADIOCARBON DATING Four volcanic cones of Holocene age occur within the Wells Cray—Clearwater area. They are forested but show little or no erosional modification. Exposure of material under the base of the flows was seen only at Spanish Cones and Dragon Cone. No carbonized material was found at Spanish Cone but along Falls Creek, which parallels the flow from Dragon Cone, undercutting of the flows has exposed a well compacted peat layer, 3-7 cm thick. It is underlain by a minimum of 50 cm of clean fluvial sands and overlain by 10 cm of intercalated silts and clay. These sediments are in turn overlain by a 5 m thick lava flow. Recently slumping has exposed an unvegetated surface of the peat-bearing section. The peat was dated in the Radiocarbon Laboratory of the Geological Survey of Canada (sample GSC-3944) and yields a date of 7560±110 date after correction based on 5 1 3 C % o = -28.1. 4.4 CONCLUSIONS Volcanism may have begun in the Wells Gray—Clearwater area some three million years ago, but it seems more likely that the McLeod Hill date is anomalous and that volcanism in the region started some million years later. The lack of dates between 1.4 Ma and 0.6 Ma may be due to sampling bias or to the principle of obliterative overlap. It seems certain, however, that the remaining dates represent more or less continuous volcanism over the last 600,000 years. These dates also support the hypothesis that no one eruptive centre fed all the flows, but rather, eruptions were temporally as well as spatially separated. 5. PETROGRAPHY All things visible and invisible. — Book of Common Prayer 5.1 INTRODUCTION Basaltic rocks of the Wells Gray—Clearwater area appear fairly uniform in hand samples and the characteristics of individual flows are not distinctive enough to separate lavas of differing ages or from separate sources. Holocene cones and flows are exceptions, but even they display some overlap in characteristics with the older flows. Thus, the basaltic rocks could not be subdivided petrographically in the field, but rather have been separated into three informal units on the basis of elevation, morphology and age (in the case of the younger cones and flows) (Chapter 3). Examination of thin sections indicates that all the units are petrographically similar; the modal mineralogy is that of alkali olivine basalts. Most of the samples are porphyritic with forsteritic olivine phenocrysts in a matrix of andesine to labradorite laths, indistinct to ophitic augitic pyroxene, granular to euhedral magnetite and ilmenite, and interstitial acicular apatite. Plagioclase and clinopyroxene rarely occur as phenocrysts. The Holocene cones and flows differ in their abundance of mantle and cumulate xenoliths and xenocrysts (which are ubiquitous); the presence of megacrysts, the more granular nature of the matrix and the greater abundance of opaque material. Crustal xenoliths are present in Pleistocene flows and ultramafic xenoliths, although present in a few locations, are rare. Many of the Pleistocene flows are diktytaxitic. Samples from the ice-contact unit share characteristics with both the Holocene and Pleistocene units. This is probably indicative of the age span of these centres. Five main petrographic categories are distinguishable on the basis of phenocrysts, texture and crystallinity. These are informally referred to as types A 84 85 through E. Samples from the plateau and valley-filling units span types A through D; the young cones and flows are predominantly type E; the ice-contact centres span the entire range but are dominantly type B. Descriptions of each of the types follows using nomenclature as defined by Williams, Turner and Gilbert (1982). Samples from which thin sections were cut, are listed in Appendix D, Table D-1, with the informal classification which was assigned to each. The first line of each type gives the principal distinquishing characteristic of thin sections in that group. 5.2 DESCRIPTIONS 5.2.1 VALLEY-FILLING, PLATEAU AND ICE-CONTACT UNITS Type A Nonporphyritic Samples of type A are commonly holocrystalline, diktytaxitic, and intergranular to subophitic (Fig. 5-1). Plagioclase is the dominant mineral and forms a meshwork of subhedral laths decreasing uniformly in size from a maximum of 5 x 0.5 mm. Larger laths are usually normally zoned and show albite and Carlsbad twinning. Smaller laths are unzoned. Anhedral, unzoned and untwinned patches of feldspar occur interstitially. Optical determination of compositions range from A n 5 0 to A n 5 5 . A second population of more equant plagioclase commonly shows resorption, strong zoning and is Carlsbad twinned or untwinned (Fig. 5-2). Ferromagnesiam minerals occupy the interstices between plagioclase laths, resulting in an intergranular texture. Pyroxene is pale brown, augitic in composition and may be twinned and zoned. Olivine ranges from euhedrai to subhedral granules. Opaques occur as euhedrai cubes (magnetite), to subhedral elongate laths (ilmenite). High relief, acicular needles of apatite occur interstitially. 87 Figure 5-2: Resorbed and zoned plagioclase crystal from a type A section (plane polarized light). Figure 5-3: Type B, section. Opaque glass with microlites of plagioclase surrounding olivine phenocrysts (plane polarized light). 88 Type B Olivine Phenocrysts Ninety percent of all thin sections examined fall in this category which has been subdivided into five sub-types (Figures 5-3 to 5-7) of which the majority of samples are confined to the latter two. The separation of the samples into sub-types is based upon their crystallinity. Type B, are the least crystalline and have extensive, cloudy to opaque, intersertal glass; crystallinity increases through type B 2 to type B a and type B 5 sections are diabasic, usually diktytaxitic, and contain well developed ophitic to subophitic pyroxenes. Samples that have been quenched form a clear subdivision of type B , . They are vitrophyric in texture with well defined olivine phenocrysts and microlites of plagioclase and olivine in a glassy groundmass. Olivine phenocrysts range in concentration from 2 to 15%. There is no field or petrographic evidence to suggest that samples containing more than 10% olivine phenocrysts are cumulates. The olivine in all samples is generally euhedrai to subhedral and inclusions of glass and spinel are present. No strain lamella was seen but slight compositional zoning is present. Phenocrysts range from 1 to 3 mm in size, but their size is fairly uniform within single samples. Some of the larger phenocrysts show "fracture cleavage", thought to originate by rapid cooling and depressurization of the magma. All of the olivine phenocrysts have highly birefringent rims of more fayalitic composition (J. Nicholls, oral communication 1985). Plagioclase forms a mesh of microlites to microscopic laths. Zoning is either absent or faint. Compositions range from A n 3 0 to A n 6 S and some sections display almost this entire range. The compositional range of plagioclase in types B, - B 3 is limited to Iabradorite. The vast majority of sections show both albite and Carlsbad twins. In type B„ and B 5 sections, 89 Figure 5-5: Type B 3 section (plane polarized light). Clinopyroxene is now discernable in the groundmass. There is no systematic variation in either the size or abundance of olivine phenocrysts between petrographic types. 90 Figure 5-6: Type B„ thin section in which clinopyroxene is now a prominent constituent of the groundmass. Holes (diktytaxitic texture), show up as white areas in the thin section (plane polarized light) Figure 5-7: Type B 5 thin section. Clinopyroxene now forms a clearly subophitic to ophitic texture, surrounding laths of plagioclase (plane polarized light). 91 anhedral, untwinned, low relief feldspar is intergranular and forms a meniscus bounding diktytaxitic cavities. Plagioclase concentrations range in mode from 15% to 20% in type B, sections, to 50% in type B 5 sections. Pyroxene ranges from indistinct microgranules and dendrites in type By sections, to subhedral ophitic crystals in type B 5 sections. All pyroxene crystals that are optically distinct, are pale brown in colour. In type B f t and B 5 sections there is a range in coloration, from pale to mid brown with slight pleochroism. Inclined extinction, 2V angles and z^c extinction angles are consistent with an augitic pyroxene. The brown color indicates high Ti content. Microprobe analyses done by Bishop (1985) and Fiesinger (1975) on pyroxenes give a maximum Ti content of 3.6 wt.%. In holocrystalline, type B 5 samples, pyroxene constitutes up to 25% of mode. Opaque material radically changes from type B, to B 5 sections. In type B! , opaques and undifferentiated material form a splotchy, network within intersertal brown glass. The opaque crystals progress from networks of chains, dendrites, and fuzzy grains, to well formed cubes (magnetite) and elongate prisms (ilmenite). Opaque material constitutes 5-7% of the mode in holocrystalline samples. Closely associated with the opaques are acicular needles of apatite. The continuum represented by type B sections appears to reflect the cooling history of the flows. Wherever possible, samples were collected from the centre of the cooling unit. Flows range between 1 and 3 m in thickness, but there is no correlation between flow thickness and the petrographic category into which it was assigned. It is suggested that flows of type B! were extruded as single units that cooled rapidly and were completely crystallized before being covered by further flows. Areas where diabasic textures are best developed were probably rapidly extruded as a 92 series of cooling units, one on top of the other. This retarded cooling in the lower flows and produced the diabasic, diktytaxitic textures. This hypothesis is substantiated in sections from Holocene flows, all of which would be categorized as types B, - B 3 based on crystallinity. Type C Olivine and Plagioclase Phenocrysts These type C samples are holocrystalline, diktytaxitic, and intergranular (Fig. 5-8). The intergranular texture results from pale brown pyroxene, olivine and opaque minerals which occur between the feldspar laths. Plagioclase phenocrysts occur in two populations, one of strongly zoned, subhedral stubby crystals and another of moderately zoned to unzoned Carlsbad and albite twinned laths. Phenocrysts of the first type are rare (<1 %) and the predominant groundmass and phenocryst feldspars in the section are of the second type. Optical determinations of the compositions of the phenocryst laths range from An 4 7 to An 7 8 . Type D Olivine, Plagioclase and Clinopyroxene Phenocrysts Samples in this group have phenocrysts of olivine, plagioclase and clinopyroxene and are holocrystalline and intergranular (Fig. 5-9). Phenocrysts of plagioclase occur as individual laths and as glomeroporphyritic clots with olivine and clinopyroxene. Olivine phenocrysts are euhedrai to subhedral whereas pyroxene is largely subhedral, pale brown, equant grains with faint zoning. Some pyroxene is twinned and few larger grains, visible in hand sample, comprise an aggregate of smaller anhedral grains. As in the other thin section types, plagioclase consists of Carlsbad twinned and moderately zoned euhedrai laths and subhedral Carlsbad twinned or untwinned, strongly Figure 5-9: Type D section. Olivine, plagioclase and clinopyroxene form distinctive dots in a fine grained groundmass (plane polarized light). 94 zoned stubby prisms. Plagioclase in the groundmass forms a meshwork of Carlsbad and albite twinned, lath-shaped microlites. The anorthite content of the microlites could not be determined, but the phenocrysts range from A n 5 2 to A n 5 7 . Feldspars of low relief form anhedral intersertal patches, particularly around the margins of diktytaxitic vesicles. Mafic and opaque minerals form an intergranular network around the plagioclase. The opaque minerals are both granular and irregular squares in plane view. Trachytic texture Trachytic textures were occasionally seen, but are rare (Fig. 5-10). These sections represent samples in which plagioclase was crystallizing in large volumes during flowage and movement ceased such that the original alignment of feldspar laths was preserved within the sample. 5.2.2 YOUNG FLOWS AND CONES Type E Olivine and Clinopyroxene Phenocrysts Samples from the young flows and cones are texturally similar to Type B T to B 3 sections, but contain both pyroxene and olivine phenocrysts. Commoniy these phenocrysts represent a high pressure assemblage (Chapter 6) that is out of equilibrium with the melt and show strong resorption textures (P. Metcalfe in preparation; Bishop 1985; Fiesinger and Nicholls 1977). Olivine megacrysts with strain lamelli and zoning also occur. Fiesinger and Nicholls (1977) thought these represented xenocrysts, but Bishop (1985) concluded that they may have been in equilibrium with the melt at some stage. Equilibrium calculations (Chapter 6) show that they are high pressure phases. Despite the high normative nepheline values generally over 10 cation Figure 5-10: Sections showing well developed trachytic texture in a holocrystallin sample (above) and a fine grained sample with intersertal glass (below), (plane polarized light) 96 % (Appendix D; Table D-1) feldspathoids were not seen in thin section or hand sample. Following MacDonald and Katsura (1964), Fiesinger (1975) and Fiesinger and Nicholls (1977) classified these samples as basanitoid to indicate that they contain more that 5% normative nepheline but petrographically are most similar to alkali olivine basalt. They further classified lavas at Flourmill Cones (which they refer to as Spanish Creek Cones) as ankaramites because of their basanitoid characteristics and abundant pyroxene and olivine phenocrysts. Detailed petrography of Kostal and Flourmill Cones can be found in Fiesinger (1975); Flourmill Cones in Bishop (1985); and all the young centres in P. Metcalfe (in preparation) 5.3 CONCLUSIONS The principal phenocryst phase seen in all sections from the Wells Cray—Clearwater area is olivine. This phenocryst appears to represent the only mineral phase which was crystallizing during the upward transport of the magma, ln the pre-Holocene centres it is unstrained, usually less than 5 mm in size and, although it has an iron rich rim, the cores appear to be a fairly uniform Fo content, = F o 8 0 . Other mineral phases appear to have completed crystallization in situ and show no breakage of crystals from post emplacement movement. Crystal concentration prior to eruption, was low in the majority of flows as trachytic textures are scarce. Clinopyroxene occurs as phenocrysts in a few flows, principally those of Holocene age, but it appears to be out of equilibrium with the melt and often has a kelyphitic rim. It is, however, a major constituent of the matrix in holocrystalline sections and forms well developed subophitic textures. Orthopyroxene was not seen in any of the examined thin sections. 97 Plagioclase is a prominent part of the matrix in all samples but occurs as a phenocryst phase in only a small percentage of flows. Microlites are seen in quenched samples, so crystallization may have begun during transport of the magma to the surface, but the crystallization period prior to eruption was short, llmenite, magnetite and apatite are minor phases in the matrix which occupies the interstitial spaces between plagioclase laths. The matrix is often dominated by intersertal glass. Textural changes in the samples seem to be controlled by the cooling time of the flow. Samples which are cooled quickly contain intersertal glass, slower cooling gives an intergranular texture and those most slowly cooled have a diktytaxitic, diabasic texture. As the majority of flows are 3 m or less in thickness, retardation of the cooling period may result if several flows were emplaced one on top of the other, in a short period of time. 6. WHOLE-ROCK CHEMISTRY In solving problems of this sort, the grand thing is to be able to reason backward. That is a very useful accomplishment, and a very easy one, but people do not practise it much.... —Sherlock Holmes (Sir. A.C. Doyle, A Study in Scarlet) 6.1 INTRODUCTION Basaltic volcanoes are probes of the earth's mantle. The composition of the eruption products reveals magmatic processes that are occurring in the subsurface during the formation and ascent of the magma. These same chemical characteristics have been used to classify specific types of basaltic rocks. Basalts are of two principal types — subalkaline (tholeiitic and calcalkaline) and alkaline. The most voluminous basaltic rocks are tholeiitic in composition; alkalic basalts make up less than 1% of all erupted basalts. Despite the low volume of alkalic basalts, they hold a special fascination for petrologists because the processes that result in their composition are poorly understood. The composition of a magma erupted at the earth's surface is the result of the magma's initial composition and any process, such as crystal fractionation, that the magma has undergone during transport to the surface. All basalts are thought to originate in the upper mantle by the partial melting of peridotite (Bowen 1928; Ringwood 1975; Yoder 1976). Variations in the degree of partial melting and the pressure at which melting occurs are thought to result in the diversity of basaltic melts found around the world (Green and Ringwood 1967; Kushiro 1968, 1972; Ito and Kennedy 1967, 1968). Fractionation of the magma can result in further compositional changes, as can assimilation of crustal rocks, mixing of two magmas, or source heterogeneity. In the Wells Gray—Clearwater area a total of approximately 30 km 3 of basaltic lava has erupted over a 600,000 year period. The region extends 70 km north-south and 30 km east-west and contains isolated cones, as well as more 98 99 extensive valley-filling and plateau-capping flows (see Chapter 3). The major element chemistry of these basalts is fairly uniform but striking differences occur in a number of the trace elements. This whole-rock chemistry will be used to characterize the probable source material for the basaltic magma and what processes, if any, have acted upon that magma during its ascent to the surface. 6.2 METHOD One hundred and thirty one rocks were analyzed for 9 major and minor elements and for the trace elements Ba, Rb, Sr, Nb, Y, Zr, and V. In addition, a subset of these samples were analyzed for Cr, Ni, Cu, Zn and the rare earth elements (REE). Choice of specimens and sample preparation methods are outlined in Appendix A. Most analyses were obtained from whole-rock, pressed powder pellets, using X-ray fluorescence spectroscopy (XRF). Details of this method and the procedure for error analysis is given in Appendix B. Atomic absorption (AA) methods were used to determine the rare earth elements from Sm to Yb. A complete discussion of the technique is given in Appendix C. Iron determinations were made using the method of Wilson (1955) on approximately two thirds of the samples. These same samples were also analyzed by loss on ignition (LOI) for volatile content. The results are listed in tabular form in Appendix D. Graphical presentation of all the major, minor and trace elements can also be found in Appendix D. Computer programs used to reduce the raw data are given in Appendix E. Also included in this appendix are the method and computer programs used to present the data in tabular and graphic form. 100 6.3 RESULTS The suite has uniform or systematically varying concentrations of most major and minor elements. The rocks range in silica content between 42 and 53 wt.%. The differentiation index (Dl; Thorton and Tuttle 1960) of the suite ranges from 29 to 57, but the majority of samples have a Dl of between 30 and 40. The Mg* ((MgO/MgO + FeO)100) number varies from 32 to 56. On an AFM diagram (Fig. 6-1) the suite plots as a restricted field within that outlined by Schwarzer and Rogers (1974) for alkali basalts and below the tholeiitic iron enrichment trend of the Skaergaard complex. On the alkali vs. silica diagram (Fig. 6-2) the suite straddles the boundary between the alkaline and subalkaline fields as defined by Irvine and Baragar (1971) and McDonald (1968). In the new classification scheme presented by Le Bas et al. (1986) the suite spans 6 different fields from tephrite basanite to basaltic andesite (Fig. 6-2). Normative mineralogy was calculated for all samples (Appendix D). Some 97% of the samples are olivine normative and on the basis of normative mineralogy the suite plots from the field of basanites, to alkali olivine basalts to olivine tholeiites as defined by the basalt tetrahedron of Yoder and Tilley (1962). Normative plagioclase ranges from oligoclase to labradorite. Normative quartz is present in three samples (Appendix D; Table D-2). About one quarter of the suite is hypersthene normative. These are exclusively samples from the Clearwater unit (Chapter 3). The classification of the suite based on normative mineralogy is in conflict with the petrographic classification. Normative nepheline concentrations range up to a maximum of 27 cation % (predominantly in samples from young cones and flows), but no feldspathoids were seen in thin section (Chapter 5) or in probe work (Fiesinger 1975). Petrography of the pyroxenes indicate that they are clinopyroxenes, augitic in composition. Using probe data, similar conclusions for 101 Figure 6-1: AFM plot (wt.%) of the Wells Cray—Clearwater suite of basaltic rocks. Young cones and flows are shown as squares, others as triangles. Dashed line is the Skaergaard trend, dotted line is from Thingmuli. Stippled field is for alkali basalts given by Schwarzer and Rogers (1974). 102 5 O CM CO Z o 04 8 -7-6-4 -T E P H R I T E B A S A N I T E B A S A L T I C T R A C H Y -A N D E S I T E A L K A L I N E ^ / . 0 V 0 B A S A L T 4'0 42 4'4 4*6 4^ 5^" S i 0 2 wt.% T B 0 L E 1 1 T 1 C -0 B A S A L T I C A N D E S I T E 52 54 56 0 Holocene cones and flows v Pleistocene Clearwater and Wells Gray units A Ice contact basalts Figure 6-2: Plot of all basalts from the Wells Cray—Clearwater area in terms of total alkalies ( K 2 0 + N a 2 0 wt.%) versus silica content (Si0 2 wt.%). Dashed line (Irvine and Baragar 1971) and dot-dash line (McDonald 1968), separates the fields of alkaline and subalkaline compositions. Solid lines separate the fields of rock types given by Le Bas et al. (1986). 103 pyroxene compositions were reached by Bishop (1985) and Fiesinger (1975). Olivine and augitic pyroxene occur as ubiquitous constituents of the groundmass. A reaction relationship, often noted in tholeiitic lavas, between these two minerals was not observed, even in the three sections which are chemically quartz nomative. As normative mineralogy can be influenced by small chemical differences such as the ferric/ferrous ratio, the petrographic results are considered the more reliable criteria for naming the rocks. As such, samples from the Wells Cray—Clearwater suite are classified as alkaline olivine basalts. 6.3.1 COMPOSITIONAL VARIATIONS WITHIN THE SUITE Compositional variations within the suite are most obvious between the young, predominantly Holocene cones and flows, and the Pleistocene basalts of the Clearwater and Wells Gray units. An examination of these compositional data (Appendix D) indicates that samples from the young cones and flows appear to be enriched in T i 0 2 , P 2 0 5 , and N a 2 0 as well as having the lowest silica concentrations. Trace element results indicate that these same samples have elevated concentrations of Ba, Nb, Sr, Rb, and Zr (Appendix D). Copper and V abundances in samples from the young cones and flows may be slightly enriched, but Ni, Cr, Zn and Y concentrations are similar. This compositional trend is also apparent in samples from ice-contact features that have been interpreted, on the basis of morphology, to date from the Fraser Glaciation (Chapter 2 and 3). These ice-contact features include Pyramid Mountain, Fiftytwo Ridge and Mount Hyalo (Plate 1). The apparent bimodality of the suite may bear upon the petrogenetic processes that the samples have undergone. For this reason the spread in compositions was tested using a statistical method. Histograms of each element were made and the results plotted as cumulative curves on probability paper. Using the method of C. Stanley (in preparation), curves representing one or more normal 104 populations were fit to the data using either visual parameter estimates or a maximum likelihood solution. The method confirms the complexity of the data, but the results could not be used to separate, unambiguously, individual populations. The inability to separate subpopulations stems from the involvement of a specific element in a petrogenic process. If the process results in a variation in the concentration of the element, the statistical analysis tends to split the compositional variations into a high concentration population and a low concentration population. These end member populations cannot then be separated from samples that may show 'inherited' high (or low) concentrations. To clarify relationships that may exist among samples from the suite, a Pearce element-ratio diagram (Pearce 1968) was used. For this technique to be successfully applied, a number of criteria must be met by the system being analyzed (Pearce 1968; Russell and Nicholls, in preparation). Of prime importance for the results considered here is that the 3 variables chosen for the axes must be independent of any petrogenetic process that the magma has undergone since formation. For this reason 3 elements, which are incompatible with phenocrysts present in the melt, are chosen as the axes of the plot. Samples that share a common magmatic source (comagmatic) will plot as tight clusters. Variations will be produced when magmas are the result of melts from source regions which differ in composition. The ratios of Ti/K and P/K were chosen for plotting the samples from the Wells Cray—Clearwater suite. The results are shown in Figure 6-3. Titanium is moderately compatible in clinopyroxene structures. Trace to 3.6 wt.% concentrations of Ti are found in phenocrysts from the Spanish Cones area (Bishop 1985). Likewise, K is moderately compatible in the plagioclase structure and up to 4 mole percent orthoclase was found in microphenocrysts from the Spanish Cones area (Bishop 1985). Similar results were obtained on samples from other centres in the 105 0.40 c 0.35 o u ro > 'I 0.30 QJ C a £ 0.25 ro 0.20 A A NORMAL Jack 's Jump ^ Blackpool c 0.35 QJ ra > aj cz a 0.25 rn A Clearwater + Ice contact V Welte Gray unit • Mann Creek area V Enriched Ice contact A Young cones and flows 0.20 -A Ray Mountain region v yum A '# A N7 A A A & A Spanish, Flourmill area ENRICHED 1.0 1.5 2.0 T i / K ( C a t i o n e q u i v a i e n t ) Figure 6-3: Plot of P/K vs. Ti/K for the Wells Cray—Clearwater area basalts. For melts from a similar parent material the ratio of each daughter should be the same. Variations indicate differing sources or variations in the amount of partial melting which will enrich the melt in the most imcompatible element preferentially with respect to other elements. Samples that come from the same geographic area are marked in stipple pattern. The enriched group is late Pleistocene to Holocene in age. The normal group is older and flows in the stratigraphic sections (for example The Flatiron and Jack's Jump) may have a wide variation in age. Solid triangles are 7 samples from a quarry in the lower Clearwater valley. The error bars shown in the lower right hand corner are applicable to both diagrams. 106 Wells Cray—Clearwater area by Fiesinger (1975). However, plagioclase and pyroxene phenocrysts occur in only a small fraction of the samples analysed. Olivine is the only phenocryst phase that is present in virtually all samples. Potassium, Ti and P are incompatible in the olivine structure and are among the least compatible of the major and minor elements in clinopyroxene or plagioclase. Manganese was not considered here because of the low concentrations and analytical uncertainty. Croups of samples that can be related both spatially and temporally (within a few 10's of thousands of years) plot as a tight cluster. These spatially and temporally related centres include samples from Spanish and Flourmill cones as well as the tuya, Mount Hyalo, the eruption of which is only interpreted, on the basis of morphology, to be temporally related to the other cones in the region. Other groups of flows that are spatially related, but which are not tightly constrained in terms of time, plot on the basis of composition as either tight clusters (Blackpool) or show a wide variation (Jack's Jump, quarry samples and The Flatiron) (Fig. 6-3). On the basis P/K vs. Ti/K ratio variations, the separation of two, more or less distinct, groups can be made (Fig. 6-3). The young cones and flows ('enriched' group) plot at lower P/K and Ti/K ratios than samples from the other units ('normal' group). Separation of these two groups of samples is consistent with the results from an analysis of the single element data. Based on this conclusion the samples were subdivided into two groups, an 'enriched' group and a 'normal' group. The enriched group includes all analyzed samples from the young cones and flows and all samples from the ice-contact features which, on morphological grounds, were considered to be late Pleistocene (Chapters 2 and 3). All of these eruption centres, except Mount Hyalo, contain ultramafic mantle nodules and have olivine and pyroxene phenocrysts (+ plagioclase; type E petrography, see Chapter 5). 107 6.3.2 PETROCENESIS The subdivision of the analyses into separable comagmatic groups on the basis of the P/K vs. Ti/K ratios, as well as more broadly defined 'enriched' and 'normal' subdivisions, suggests that samples are the result of separate sources or differing processes or both. Samples with low Ti/K and P/K ratios may originate from a source which is enriched in K, P, Ti, K, Na, Zr, Rb, Sr, Ba and possibly Cu and V, relative to the high Ti/K and P/K ratio group. However, separation of groups of samples on the basis of elements that are classified as incompatible (except for Zr, Cu, V, and Ti) overlooks the possiblity that variations in those same elements may be the result of petrogenetic processes rather than chemical variations within the parental material. Incompatible elements are those elements that are enriched in magmas which originate by limited partial melting. Barium, Nb, Rb, and Sr, are large ion lithophile elements (LIL) and among the most incompatible of elements. Extreme enrichment in these elements may result from small degrees of partial melting as the affinity (compatibility) of these elements for the liquid phase of the system is great. The degree of compatibility of an element for a particular phase in a system is given by the distribution coefficient (K ), which is the ratio of the concentration of an element in a particular mineral phase compared to the concentration of that element in the coexisting liquid phase. If the K^ is very much less than 1, the element is deemed incompatible in that mineral structure and will concentrate in the melt phase. 'Incompatibility' as it is used here, means that the element is not compatible in the structure of the mineral phases present at the point of origin of the magma. Also included in the LIL group, are K, P, Zr, Ta and Th. Tantalum and Th were not analyzed in samples from the Wells Gray—Clearwater suite. Samples from the enriched group are among the most alkalic in composition of the entire suite. In terms of normative mineralogy samples from the enriched 108 group have up to 27 cation % nepheline. According to theories on the generation of alkalic basalts, the alkalinity of a magma can be increased by the fractionation of olivine from a melt. The removal of olivine from a low nepheline normative liquid, drives the residual liquid to more alkalic compositions (Cox e( al. 1979; Morse 1980). As L1L elements are incompatible in the olivine structure, they tend to be concentrated in the melt phase. Several authors have argued that the concentrations of the LIL elements seen in some alkalic rocks cannot be adequately explained by fractionation of olivine from the melt. Instead, these authors have called upon small degrees of partial melting of the mantle source in order to concentrate LIL elements in the melt phase, (see for example Frey ef al. 1978, Clague and Frey 1982). The relationship between incompatible, LIL element concentrations and increasing normative nepheline is graphically shown in Figure 6-4. This can be contrasted with the behavior of the compatible element Ni when plotted in terms of normative nepheline (Fig. 6-4). Differing degrees of partial melting has also been used to explain some of the variations in major element chemistry of the principal basalt types. Kushiro (1972) has shown that by increasing the melt fraction of a dry peridotite at 20 kb pressure, the liquid changes composition from an initial, silica undersaturated, nepheline normative composition to hypersthene normative at large degrees of partial melting. Using a spinel Iherzolite from Hawaii, Takahashi and Kushiro (1983), demonstrated a similar relationship. They achieved a melt of alkali olivine basalt composition at low degrees of partial melting, olivine tholeiite at increasing degrees of partial melting and finally a melt that was a peridotitic komatiite in composition at the highest degree of partial melting. All of these results were obtained at pressures of between 15kb and 25 kb. Major and trace element concentrations can be used to test both the viability of crystal fractionation and partial melting processes. Compositional 109 1400 — 1200 IOOO H g 800-CQ 600 -400 200 A A A A A oA 4S7 A O A Ac, o 120 — 100 — — 80 — <= a. a. 60-40 • 400 -— 300 -200 -100 -•4* o £ A 7 ^ A A j £ A * O A A A 2 0 " S ^ o * A * O A A a A O i i i O A A A 9 IS — r ~ 20 5 10 Norm. Ne (cat. eq.) 25 30 NORMAL ENRICHED A Clearwater unit + ice contact O Enriched Ice contact unit V Well* Gray unit • Mann Creek area A Young conea and flow* Figure 6-4: Plot of normative nepheline (cation equivalent) against the incompatible elements Ba and Nb. Also shown is the compatible element Ni vs. nepheline. 110 differences among the samples from the Wells Gray—Clearwater suite can be modelled as the result of processes of partial melting or crystal fractionation. Crystal fractionation Crystal fractionation is usually considered to occur at lower pressures than those at which the original melt formed. A homogeneous melt rises from the point of origin in the mantle until reduced pressure and/or temperature stabilizes a particular mineral phase or phases in the melt. In magmas of basaltic compositions these phases will vary according to the composition, pressure and temperature of the magma but the stable phenocryst assemblage can be calculated (Russell 1984; Russell and Nicholls 1987). Using the procedure of Russell and Nicholls (1986), the temperature and pressure of phenocryst phases stable at liquidus conditions can be calculated from the bulk composition of the samples. The types of phenocrysts stable in the melt at equilibrium conditions and the range of temperature and pressures over which they are stable are shown in Figure 6-5. For the majority of the samples analyzed, the inferred order of phenocryst crystallization, at low pressures, is olivine followed by plagioclase and then clinopyroxene. At high pressures, clinopyroxene is stable before plagioclase and can even stabilize before olivine, and there is a large temperature drop between clinopyroxene and plagioclase crystallization. The order of high pressure crystallization is olivine, clinopyroxene followed by plagioclase; or clinopyroxene, olivine followed by plagioclase. The crossover point, at which the change in the order of crystallization between plagioclase and clinopyroxene occurs, is between 1.7 and 6 kb, but for the majority of the analyses the change occurs between 3 and 5 kbars (Fig. 6-5a). 111 1 1 0 0 -0 2 4 6 PRESSURE (kbars) Figure 6-5: a) Plot of results of calculation of the equilibrium mineral assemblage from a range of compositions of basalts from the Wells Cray—Clearwater suite. All samples chosen for calculation have only olivine phenocyrsts. The sets of lines for each mineral indicate the maximum and minimum pressure and temperature variation obtained from equilibrium calculations of differing compositions. The solid dots show the range of pressures and temperatures at which the crossover in the order of crystallization between plagioclase and clinopyroxene occurs, b) Diagram shows the results of the equilibrium calculations using the compositions of samples 34708 and 43705 which have modal olivine + plagioclase + clinopyroxene phenocrysts. Plagioclase is the stable mineral in the melt before olivine at low pressure. The arrows mark the proposed crystallization path of the samples determined from petrography. All calculations are for anhydrous conditions. 112 The phenocryst assemblage and the order of crystallization can be determined petrographically and compared with the calculated results, facilitating a determination of the pressure and temperature of crystallization. Samples that contain olivine plus plagioclase (Type C sections, Chapter 5), without clinopyroxene phenocrysts, must have crystallized at low pressures. Samples with olivine and clinopyroxene phenocrysts (Type E sections, Chapter 5), are most likely to have crystallized at high pressures (and temperatures). Samples in which plagioclase appears to be a crystallizing phenocryst phase (type C sections, Chapter 5) are compositionally distinct from samples in which olivine is the first phase stable in the melt. Examples of the results from these samples are shown in Figure 6-5b and indicate that these samples must also have crystallized at low pressure. For samples that contain only a single phenocryst phase, olivine (type B sections, Chapter 5), the forsterite (Fo) content of the olivine must be determined. Olivine which crystallizes at high pressure is usually richer in Fo than low pressure olivines crystallizing from the same melt (Russell and Nicholls 1987). The equilibrium Fo content can be calculated for differing pressures and temperature (Russell and Nicholls 1987). The olivine composition becomes slightly more forsteritic as the pressure increases, but the value is dependent upon the bulk composition of the sample. Using the technique of Russell and Nicholls (1986) the equilibrium forsterite composition of the magma was calculated for several samples (Figure 6-6) for which compositional results of whole-rock and probe analyses were available. In general, at atmospheric pressure, the equilibrium Fo content exceeds 82%, with the exception of the sample from Jack's Jump (BCV11). At atmospheric pressure, the equilibrium Fo content drops sharply with falling temperature. 113 a. 92 Q V 1 6 1100°C S8042B 1100°C S8053B SO J S80603 Figure 6-6: Equilibrium olivine compositions in the binary system Fo — Fa at varying pressures and temperatures. The upper sets of lines in each diagram indicate increasing pressure and increasing temperature. The lower set of lines in each diagram are at constant pressure but decreasing temperature. Solid squares are olivine phenocryst and megacryst compositions. Solid circles are rim and groundmass compositions. Microprobe results are from Bishop (1985) and Fiesinger (1975). The temperature and pressure of the crystallization of the olivine is inferred from the equilibrium calculation. 114 Probe analyses of olivine from Flourmill Cones (Bishop 1985) show that the olivine megacryst and phenocryst composititions are consistent with crystallization of the mineral phases at high pressures, but olivine of the groundmass composition must have crystallized at low pressure and falling temperature (Fig. 6-6). For probe analyses of olivine samples from Sheep Track Bench (BCV2, BCV3, QV16) and Jack's Jump (BCV11, BCV12; Fiesinger 1975), both the groundmass and phenocryst compositions are compatible with crystallization at low pressures. Petrographic analyses of olivines in samples from the normal group indicate that the forsterite content is around 80 (Chapter 5). The majority of the olivine phenocrysts in these samples have highly birefringent edges that may represent as much as a 10% decrease in the Fo content of the olivine (Nicholls, written communication 1985). This result is consistent with low pressure crystallization of olivine in the magma followed by rapid ascent, precluding the crystallization of other phenocryst phases. The rapid ascent of the magma is substantiated by the inclusion of siliceous metamorphic and granitic crustal fragments in the samples. Despite hydrous mafic phases, many of the crustal xenoliths are unmelted or show only incipient melting around the margins of the clast. The surrounding magma may have been several hundred degrees above the melting point of these clasts, so rapid transport would be required to bring them to the surface unmodified. Olivine fractionation at low pressures is a possible mechanism for compositional changes in the magma for most of the samples in the Wells Gray—Clearwater suite. Plagioclase is not a principal phenocryst phase in the magma and clinopyroxene appears important only for magmas that have retained their high pressure assemblage (i.e. the enriched group). 115 In terms of the major element concentrations, Pearce element-ratio diagrams provide a better test of magmatic hypothesis than standard single element wt.% diagrams (Pearce 1968; Russell and Nichols, in preparation). If magmas from the same starting liquid are related by crystal fractionation of olivine then there will be a linear relationship between (Mg + Fe)/K vs. Si/K. If the analytical error in the concentration of the individual elements is known then the resulting relationship can be tested statistically (Russell and Nicholls, in preparation). Likewise, all elements that may be involved in fractionation of a particular mineral phase can be tested (Pearce 1968; Russell and Nicholi, in preparation) Pearce element-ratio diagrams were used to plot the compositional results of samples from the Wells Cray—Clearwater area. If the samples represent magmas that have undergone crystal fractionation of titaniferous clinopyroxene, then some of the variation in titanium could be explained (Ti/K, Fig. 6-3). In a similar manner fractionation of plagioclase may result in slight changes in the K concentration. To test fractionation of the mineral phases present in samples from the suite, the suite was assumed to be comagmatic and elements involved in the fractionation process were tested by using P in the denominator. Elements involved in the fractionation of olivine (0.5*(Fe + Mg)/P), plagioclase (2.0*Na/P+Al/P) and, clinopyroxene (2.0*Ca/P + (Na-Al)/P), olivine + plagioclase (0.5*(Fe + Mg)/P + 2.0*Ca/P + (Na—Al)/P), olivine + clinopyroxene (1.5*Ca/P + 0.5*(Fe + Mg)/P) and plagioclase + pyroxene (2*Ca/P + 3*Na/P) were tested (Russell and Nicholls, in preparation). The data does not fit the hypothesis that the samples are related by the fractionation of any of the tested minerals. In particular, olivine fractionation could not be confirmed. This conclusion is not surprising and confirms the results based on the 116 interpretation of the P/K vs. Ti/K ratios that the samples do not represent a comagmatic suite. Samples, such as those from Blackpool, that plot as a tight cluster in terms of P/K vs. Ti/K were also tested. A linear trend was obtained on the olivine plot but the results were indistinguishable when the analytical error was considered. The suite and several subpopulations were also tested using K in the denominator, as the analytical error in K 2 0 is small and a greater range in concentration exists (Appendices B and D). The results, however, were similar. Croups of samples which have similar P/K vs. Ti/K ratios, were then tested. In this case the results were inconclusive because the spread in the concentration of the elements was minimal. This minimal spread in concentration, coupled with the analytical error, resulted in conclusions which were not statistically significant. A further test of the hypothesis that the suite is the result of fractionation of comagmatic melts is to use mass balance equations to derive a daughter product from a more 'primitive' parental magma by fractionation of specific phenocryst phases. A computer program developed by Albarede and Provost (1977) was used for this purpose. Mineral chemistry, determined by both probe analyses (Bishop 1985; Fiesinger 1975) and optical means, was used. The results for all samples where the proposed parent composition varied more than 2 wt.% with respect to S i0 2 or MgO from the daughter composition were invalid. The statistical fit for almost all the major element concentrations was poor, particularly for T i 0 2 , P 2 0 5 , and S i 0 2 . Modes, determined both visually and by point counting (Bishop 1985; Fiesinger 1975), varied from the calculated modes by several per cent. In cases where the proposed parent and daughter are compositionally close, and the results were 117 significant, they indicate that the parent and daughter can be related by the removal of a few percent olivine phenocrysts. The possible effect of olivine fractionation is probably best demonstrated by the concentration of Ni and Cr in a whole-rock sample. If olivine is the principal phase being fractionated from the melt, Ni abundances (initially high; >250 ppm) will decrease. The K^ of Ni between olivine and the melt is =10 and for Cr, K^ ==0.2, so Cr should show a similar but less pronounced trend. Figure 6-7 includes a plot of the results of Ni and Cr vs. MgO for the entire Wells Cray—Clearwater suite. Samples from both the enriched and the normal group show similar concentrations of Ni and Cr and a sympathetic decrease in Cr and Ni with decreasing MgO concentration. If only olivine is removed from the system then the concentration of Ni should drop 50 times faster than the Cr concentration based on the distribution coefficients. Chromium is a principal constituent of Cr spinel (Ni K^ = 5; Cr K =10) which is not found as a phenocryst phase in samples from the Wells Gray—Clearwater area but does occur as inclusions in olivine. The similar magnitude of the drop in concentration of both Cr and Ni may be the result of Cr-spinel inclusions in olivine which increase the 'scavenging' capability of the olivine crystals and could account for the similar magnitude of the drop in Cr and Ni concentrations. Figure 6-7 shows the spread in the concentrations of Cr, Ni, Ba, Sr, V and Nb in two sets of samples. The enriched sample set is from Spanish Cones. The samples were collected at uniform distances from the distal to the proximal ends of what is thought, on the basis of morphology, to represent a single eruptive unit. These samples plot in a tight cluster in terms of P/K vs. Ti/K ratios (Fig. 6-3) and may be considered comagmatic on 118 N O R M A L G R O U P J l I- I I ' I I L. E N R I C H E D G R O U P -I 1 1 1 1 1 i i i 1400 1200 --ioao -a. -S- 900 -0 3 m 500-100-200' o o * o o 1200 • 1000 -c 800-400 a ? A A 1 °l ^ o o o o SO -=• SO ^ " i f " " 0 -p 3 10 12 I MgO (Cation equivalent) 3 10 12 I* rigO (Catian equivalent) NORMAL A Claarwatar unit + ic* contact V Wolla Gray unit O Mann Creak araa ENRICHED O Enrlchad lea contact unit A Young eonaa and flow a Figure 6-7a: Plot of the incompatible elements Ba, Sr and Nb vs. MgO. Crosses are average error bars for the analyses. The error varies with the concentration. Errors for each analysis are listed in Appendix D. Separation of enriched and normal samples is outlined in the text. Solid triangles in the enriched group are from Spanish Cones, those in the normal group are from a quarry in the lower Clearwater Valley. 119 100-3sa -300 -250 -e c. a. 200 -, 150-100 — 50-N O R M A L G R O U P I I I I I I ! I l I I A A 7 A Q E N R I C H E D G R O U P J 1 I i ' | + i> A A — 300 I J i 200 • 100 -7 a » 230 2iQ 200 160-120 -A A ^ A As + 8 10 12 1 rtgO ( C a t i o n e g i u v a i e n t ) S 9 '0 12 11 is ~g0 'Canon equivalsnt) NORMAL A Clearwater unit + ice contact V Walla Gray unit • Mann Craek araa ENRICHED O Enrichad lea contact unit A Young conaa and flowa Figure 6-7b: Plot of compatible elements Ni and Cr vs. MgO and the moderately compatible element V vs. MgO. Crosses are average error bars for the analyses. The error varies with the concentration. Errors for each analysis are listed in Appendix D. Separation of enriched and normal samples is outlined in the text. Solid triangles in the enriched group are from Spanish Cones, those in the normal group are from a quarry in the lower Clearwater Valley. 120 this basis. The samples were collected from the surface of the flow and the results suggest that olivine fractionation may have occurred while the flow was emplaced or the flow represents the product of a zoned magma chamber. Olivine fractionation, however, could not be modelled and, of the two samples with the highest Ni, Cr and MgO concentrations, one is from the most distal end of the flow and the other is from the cone itself. Also shown in Figure 6-7 are seven samples from a set of flows in the Clearwater Valley (42501 a-c, 42502a-d; solid triangles, 'normal'group) which compositionally fall into the normal group. Three of these samples (42501 a-c) represent the base, middle and top of a flow, 1.4 m in thickness. The other four samples (42502a-d) are from the middle parts of the remaining four flows which are exposed in the section. The flows average one metre in thickness and no paleosols or other erosional features were noted between successive units. In thin section, the samples contain only olivine phenocrysts and are virtually indistinguishable from one another. They do, however, show almost the complete spread in compositions possible in terms of Ti/K vs. P/K and show a significant spread in terms of Ni, Cr, Ba and Nb. Despite the continuous decrease in Ni concentration, the least primitive (lowest Ni and MgO composition) sample could not be derived from the most primitive by the least squares method. There is also no apparent systematic variation in composition of the flows as the top of the lowermost flow and the uppermost flow are compositionally identical, within the analytical error, in terms of MgO, Ni, Cr, Ba and Nb concentrations. If the compositional variation within the suite is the result of plagioclase or pyroxene fractionation, then some variation in Sr and V should be evident. The K^ of Sr for plagioclase is = 2.2 and for V in clinopyroxene =0.2. An inspection of Figure 6-7 shows little variation in Sr 121 within the normal group and a slight suggestion of increasing Sr within the enriched group with respect to decreasing MgO concentrations. Vanadium shows a considerable scatter within the normal group. Samples from the young cones and flows show decreasing V concentrations at lower MgO concentrations, but the lowest values were obtained from samples that are geographically separated (Pointed Stick Cone, Ida Cone and Ray Mountain; Plate 1). The results of the analysis of the major element data indicate that the suite is not comagmatic and the least primitive members of the suite cannot be reasonably derived from the most primitive members. Small subsets within the suite may be comagmatic and have undergone limited low pressure fractionation of olivine. There is no indication from these results of plagioclase or clinopyroxene fractionation. If compositional variations between samples from the suite are not the result of fractionation of melts arising from a homogeneous source, then two other possiblities can be considered: 1) that the magmas may represent small batch melts that have resulted from varying degrees of partial melting of a homogeneous source and 2) that the magmas represent melts of a heterogeneous source. Partial melting Varying degrees of partial melting of a homogeneous source region will result in magmas which vary in concentration of incompatible elements. Rare earth elements (REE) constitute a sequence of elements which increase in compatibility with a mineral phase as the molecular number of the element increases (increasing K ). Light REE (LREE) are the most incompatible with the solid phases and heavy REE (HREE) the most compatible with the solid phases. Once in the melt phase, any process involving pyroxene will 122 Figure 6-8: Plots of selected, analyses of chondrite normalized REE data from samples in the enriched and normal group. The stipple pattern shows the the range of results of the other group for comparison to the one which is plotted. Chondrite values used for normalization are from Boynton (1984). 123 result in variations in the concentration of the HREE with respect to LREE because the is higher for the HREE (more compatible in the pyroxene structure) than the LREE (Haskin 1984). Chondrite normalized results of the REE analyses are shown in Figure 6-8. Samples from the enriched group have the highest LREE concentrations. This variation can be seen in Figure 6-9 which shows the ratio of La/Yb for the enriched and normal groups. There is no variation in the concentration of HREE between the two groups so pyroxene fractionation of one group, and not the other is unlikely. Major element data do not indicate that pyroxene fractionation is a significant process in the petrogenesis of these samples, the variation in the concentration of the LREE between the enriched and normal group is considered to be the result of variations in the degree of partial melting in the source region. The elements K, P and Ti (Appendix D; Fig. 6-3) have varying K . Of these three elements, K is the most incompatible in the solid phases present at mantle depths (Coombs and Wilkinson 1970), followed by P and Ti. Consequently, low degrees of partial melting will enrich the melt in K > P > Ti resulting in increasing P/K and Ti/K ratios with greater degrees of partial melting. This effect can be seen only at the lowest degrees of partial melting, after which the concentrations reflect heterogeneities in the source region (J.K. Russell, oral communication 1986; Pearce 1968; Russell and Nicholls, in preparation). The degree to which the source material has undergone partial melting can be reflected in the ratios of Ca/Mg and Al/Ca (Clague and Frey 1982) of the resulting magma. The ratio of Ca/Mg decreases as the degree of partial melting increases and the Al/Ca ratio increases (Clague and Frey 1982). 124 • o o o < < • < C^ 3S><] <3 4 << < <3 D> O • oo r-- t o L O ro q N / ^ z ! i o 2 e o c W o ? X r 3 O c a 5 « > z LU 0 « c • 3 • < 2 o >- c O c • O I— LO ro I- O ro L O C N I— O CM — tO o L O o Cxi O O Figure 6-9: Zr/Nb vs. La/Yb ratio of samples from the Wells Gray-Clearwater suite of basalts. 125 Ca/Mg and Al/Ca results for samples from the Wells Cray—Clearwater suite are plotted in Figures 6-10, 6-11 and 6-12. against the incompatible elements Ba, Nb and Rb and the compatible elements Ni and Cr. The incompatible element data show a sympathetic increase in the ratio of Al/Ca coupled with decreasing concentrations of the incompatible trace elements. This result is consistent with increasing partial melting of the source region to produce the observed concentrations in samples. A similar result is evident as decreasing Ca/Mg ratios are coupled to diminishing Ba, Nb, and Rb concentrations. These results are consistent with the conclusion that the samples represent varying degrees of partial melt of a homogeneous source region. However, the concentration of Rb, Nb, and Ba in samples from the primitive group plot consistently above the normal group at similar Ca/Mg ratios. The Ca/Mg ratio indicated by these samples represents a high degree of partial melting. Ultramafic nodules have been analyzed from the region (Fiesinger 1975; Min 1985; Fiesinger and Nicholls 1977) and are predominantly spinel Iherzolites in composition. The nodules are presumed to represent mantle material from above the point of generation of the melt, ln a modal analysis of ultramafic nodules from British Columbia by Min (1985), modal spinel varied in concentration between 1% and 5%, olivine concentrations varied between 38% and 70%, orthopyroxene between 14% and 56% and clinopyroxene between 4% and 30%. Samples that are low in clinopyroxene and/or orthopyroxene are conversely high in olivine, attesting" to the refractory nature of this mineral. If these nodules are representative of the mantle below the Wells Cray—Clearwater area, then variations in the modal compositions may be the result of incomplete melting of the phases present. E CL C L O CD 1400 -1200 -A 1000 - A. 300 -A A A 600 -400 -200 - A A " V 120 -100 £ CL CL ao 60 -4 0 -20 -50 • 40 CL - Q ^ 20 J 10 o A A 7 ^ A A v 4? $ o & A A - - i a i A A A A •7 A A ° A A A 7 A V V M n e r a a a i n g p a r t i a l mal t L-L A O A A Q • ^ A A a 126 u i 0.5 I 1.0 C a / M g 1.5 NORMAL ENRICHED A Clearwater unit + lea contact O Enriched lea contact unit V Walla Gray unit O Mann Creek area A Young conaa and flowa Figure 6-10: Ca/Mg ratio (cations) for the Wells Cray—Clearwater area basalts plotted against the incompatible elements Ba, Rb and Nb. Arrow shows trend in the ratio as the amount of partial melting increased. 1 1400 -1200 -1000 -9- 8 0 0 C L O 6 0 0 -CO 4 0 0 -2 0 0 -120 -100 -C. 8 0 -C L C L . 6 0 H 4 0 -2 0 -5 0 -4 0 -Q . C L 3 0 H 2 0 -10 -A A ^ A A A ° ^ OAA A O * A A ^ A A CO A A O A A O A A°°* Increasing partial matt A A A O A A *0 A A fi ^ A • • A ^  A I 1.6 1.8 A l / C o 2.0 NORMAL ENRICHED A Claarwatar unit + ica contact O Enrlchad ica contact unit V Wella Gray unit • Mann Craak araa A Young conaa and llowa Figure 6-11: Al/Ca ratio (cations) for the Wells Cray-Clearwater area basalts plotted against the incompatible elements Ba, Rb, and Nb. Arrow shows trend in the ratio as the amount of partial melting is increased. 128 o o o ro • o < < 1 «3 <fJ <I < a 6 0 0 « D o a > <i o o O i n CN o o CN o m o o o IT) (ujdd) I N • <] •S <3-«3 o 0 p a 4 ° - ° « g <1 „° 4 ^ < o o o o ro o o CN (ujdd) J O o o o . 0 0 O < LcO o o a LU X o E z Ui u c Ui C C O o o c 3 o > 0 * < • 3 • c c u o c • 3 • c a = • • • < > Figure 6-12: Ni and Cr vs. Al/Ca ratio (cation equ.); and Ni and Cr vs. Ca/Mg ratio (cation equ.). 129 Calcium (and Na) in the system is present in clinopyroxene; Al is found predominantly in spinel, but is also present in significant amounts in clinopyroxene and orthopyroxene (Fiesinger 1975). The ratio of Ca/Mg in" clinopyroxene found in Iherzolite nodules in the Wells Cray—Clearwater area ranges between 1.2 and 1.5 (Fiesinger 1975). In orthopyroxenes the Ca/Mg ratio varies between 0.02 and 0.04 (Fiesinger 1975). In a whole-rock sample derived by partial melting, variations in the Ca/Mg ratio can result by varying the proportions of clinopyroxene and orthopyroxene being melted. If clinopyroxene is the predominant mineral phase being melted, then the resulting melt should reflect the ratio of Ca/Mg in the 'parental' clinopyroxene. Similarly, if melting orthopyroxene is the major contributor to the magma, then the resulting melt should reflect the ratio of Ca/Mg in orthopyroxene. In a similar manner the ratio of Al/Ca in the magma should reflect the proportions and compositions of the mineral phases being melted. In orthopyroxene the Al/Ca ratio varies between 6 and 2, and in clinopyroxene, between 0.1 and 0.5 The Al/Ca ratio will also be affected by the melting of spinel and the Ca/Mg ratio by melting olivine. An interpretation of the compositional results plotted in Figures 6-10, 6-11 and 6-12 is that compositional variations result from altering the proportions of mineral phases being melted in the mantle source. In the initial melt that is produced, the amount of Ni and Cr is not significant as only miniscule amounts of spinel and olivine are being melted. The melt must also be removed from the system rapidly, so that re-equilibration of the melt with the surrounding minerals does not occur. The whole-rock Ni concentration found in a spinel Iherzolite from the remnants of a cone on Ida Ridge, is 2253 ppm and in the same sample the Cr concentration is 2271 ppm. Nickel and Cr do not necessarily have to be introduced from 130 olivine, as significant amounts of Ni are present in clinopyroxenes. An analysis of a clinopyroxene xenocryst from Kostal Cone gives a Cr concentration of 2057 ppm and a Ni concentration of 335 ppm. Spinel is present as inclusions in the clinopyroxene, hence the high Cr content. Samples in the enriched group may be the result of melting predominantly clinopyroxene. The Ca/Mg ratio is similar to that found in the clinopyroxenes in the spinel Iherzolite nodules (Fiesinger 1975) and the Al/Mg ratio is low, though it is not as low as that seen in the clinopyroxenes from the nodules. If clinopyroxene is the dominant phase being melted to form the initial melts which have given rise to the most silica undersaturated liquids then a second process must be occurring. Melting of pyroxene cannot result in a liquid which is lower in silica than the pyroxene itself. In the field of garnet stability and in the presence of C 0 2 , olivine + clinopyroxene + liquid! •* garnet + orthopyroxene + (CO 3 in liquid2) (Clague and Frey 1982). The resuit of this reaction is a liquid lower in silica and higher in Ca than the initial liquid composition. If the basalts of the Wells Gray—Clearwater suite have been produced in the spinel stability field, as suggested by the xenoliths, a comparable reaction relationship may have occurred, but this relationship has not been documented. Partial meiting can be tested in a qualitative way by using mass balance equations in a manner similar to that used to test crystal fractionation. The computer program developed by Albarede and Provost (1977) was used for this purpose. Various compositions of pyrolite (Basaltic Volcanism Study Project 1981; Ringwood 1975) and Iherzolite (Fiesinger 1975; Min 1985; Fiesinger and Nicholls 1977) were used. Phenocryst compositions were taken from the probe data of Fiesinger (1975) and Min (1985). The 131 most silicic end member of the suite could be derived by 19% partial melting of a source of pyrolite composition. The most mafic compositions, found in the young cones and flows, were derived by 14% partial melting. This small difference in the degree of partial melting is probably insufficient to result in the enrichment in LIL elements seen in the most alkalic, lowest silica members of the suite. If a depleted spinel Iherzolite composition is used as the parental material, the 'enriched' group can be derived by as little as 0.3% partial melting. These results are very similar to those achieved by Fiesinger (1975). Though this small degree of partial melting is sufficient to explain the enrichment of the melt in LIL elements, various authors have argued that such small amounts of melt will not aggregate to produce magmas capable of reaching the surface (see for example Sleep 1974). McKenzie (1985) has however argued that alkalic components depolymerize the melt, decreasing its viscosity and allow the ascent of melt fractions as low as 0.1%. The relationship between the concentration of the elements shown in the previous diagrams can be explained by varying the degree of partial melting and the mineral phases that are dominant in the melt. LIL elements, and the LREE are most enriched in samples that represent the lowest degree of partial melting. Imprinted on this pattern are complications due to variations in the depletion of the source region. The ratio of Zr/Nb is a measure of the depletion of the source material. Zirconium is moderately compatible in mineral phases present at mantle depths, whereas Nb is incompatible. At small degrees of partial melting, Nb will be concentrated in the melt relative to Zr (low ratio), but as the amount of partial melting increases the ratio will increase to some constant value which reflects the concentration of the elements in the source 132 zone, in a manner similar to the P/K and Ti/K ratios. This effect can be seen by plotting Zr/Nb vs. the incompatible element K (Fig. 6-13). If a source zone in the mantle has undergone previous melting events and magma has been extracted, then the source area will be depleted in Nb and Zr. Normal mid-ocean ridge basalts (MORB), which are thought to be the result of large degrees of partial melting ( = 20%), have Zr/Nb ratios which range from 25 - 110. MORB's, termed E-MORB type (Basaltic Volcanism Study Project 1981), which are thought to originate from the most depleted mantle have Zr/Nb values of less than 15 (Basaltic Volcanism Study Project 1981). Figure 6-13 shows the results for the Wells Cray—Clearwater area plotted against the elemental concentrations of the incompatible element K. The incompatibility of these elements ranks K > Nb > Zr. As in the previous diagrams, the concentration of elements in the enriched group is consistent with the hypothesis that these samples represent the lowest degrees of partial melting in the source region. The observed elemental concentrations of the remaining samples indicate that the source region is heterogeneous and has been depleted by the generation of prior partial melts. 6.4 CONCLUSIONS The whole-rock chemical analyses of basalts from the Wells Gray—Clearwater area is consistent with derivation of the samples by the high pressure melting (within the field of spinel stability) of a heterogeneous, but generally depleted source region. The basaltic rocks originated as partial melts of varying percentages; those melts that were the result of the largest degree of partial melting accumulated in crustal reservoirs long enough to re-equilibrate high pressure mineral 133 1.0 1.5 2.0 K 2 0 (cation eq.) 2.5 NORMAL A Clearwater unit + ice contact V Wells Qray unit • Mann Creek area ENRICHED O Enr iched ice contact unit A Young cones and flows Figure 6-13: Plot of the Zr/Nb ratio as a measure of depleation of the source region from prior partial melts, against the incompatible element K as K 2 0 cation equivalent. 1 3 4 phases, drop any mantle nodules, and begin to crystallize low pressure phenocryst phases. Crustal xenoliths were incorporated into the melt above this point and the magma was transported rapidly to the surface. Magmas which are the result of the lowest degrees of partial melting apparently originated in a source region that was already depleted by the generation of prior partial melts but was, however, enriched in LIL elements and other trace and minor elements. The enrichment of the source zone in alkalies may have made it possible for a small degree of partial melting to occur and form a magma that was capable of reaching the surface (McKenzie 1985; Takahashi and Kushiro 1983). Transport of these magmas to the surface was rapid enough that high pressure mineral assemblages were preserved. Magma velocity was also capable of transporting large (greater than 25 cm in diameter), high density, mantle xenoliths to the surface. These magmas represent the youngest eruption centres in the region. This time dependence leads to the conclusion that the magmas which represent a high degree of partial melting formed first and depleted the source zone. Later magmas were the result of smaller degrees of partial melting. A volatile phase may have separated from the first formed magmas and enriched the depleted mantle around the area of magma formation, making partial melts of the depleted source possible at a later time. 7. PB, SR, AND ND ISOTOPIC RESULTS It is more of a business to interpret the interpretations than to interpret the things. — Michel De Montaingne (France, 1533-1592) 7.1 INTRODUCTION Whole-rock determinations of 8 7 S r / 8 6 S r , 1 a 3 Nd/ 1 fta Nd, 2 0 8 Pb/ 2 0 a Pb, 2 0 7 P b / 2 0 1 , P b and 2 0 6 Pb/ 2 0 4 Pb were made on the Quaternary alkali olivine basalts of the Wells Cray—Clearwater area. Pb, Sr and Nd isotopic systematics have been used to evaluate source material and evolution of volcanic rocks (see O'Nions 1984 and references therein) and evolution of the crust and mantle (see jacobsen ef al. 1986 and references therein). Recent work on the isotopic composition of several volcanic assemblages in the Canadian Cordillera points to a heterogeneous mantle beneath the volcanic centres (Bevier 1983b, 1986, in preparation; Smith 1986; Metcalfe and Smith 1986). Mantle heterogeneity has been inferred to result from the extraction of partial melts from the source rock among other reasons. This process has altered the elemental ratios in the source rocks leading to differences in the time-integrated isotopic ratios (Smith 1986; Bevier 1983a, 1983b, 1986, in preparation). In recent work on volcanoes of the western Anahim Volcanic Belt, Bevier' (1986, in preparation) showed that samples from these centres are isotopically similar to the Chilcotin Croup basalts and seamounts on the Juan de Fuca Ridge. Bevier postulates that a widespread area of depleted oceanic mantle underlies the accreted terranes of the Canadian Cordillera. The Wells Cray—Clearwater area lies between rocks that are clearly part of the North American craton (Price ef al. 1985) and allochthonous rocks of the accreted terranes (Struick 1984, 1986a; Price ef al. 1985). The Quaternary volcanic assemblage lies on rocks of the Barkerville Terrane (Struik 1984, 1986a) which is composed of Precambrian to Permian rocks of the Snowshoe Croup. Proterozoic 135 136 Kaza Croup rocks (Campbell and Tipper 1971; Murphy and Rees 1983) occur west of these and are juxtaposed against allochthonous rocks of the Slide Mountian and Quesnel Terranes (Fig. 7-1 and 7-2). The Barkerville Terrane is part of the Kootenay-Monashee Terrane (Monger et al. 1982; Price et al. 1985) which is considered parautochthonous and to have a Precambian root (Monger and Berg 1984; Price et al. 1985). Although the Wells Cray—Clearwater area basalt erupted through thickened Precambrian crust and carries variously melted crustal xenoliths, the major and trace element chemistry does not reflect crustal contamination. Fractionation of the magmas has been minimal and variations in the trace element concentrations are considered to reflect varying degrees of partial melting of a depleted, and possibly heterogenous mantle source (Chapter 6; Metcalfe and Smith 1986). There are indications that some of the magmas (chiefly the most recent) have arisen from a mantle source which has been altered by a metasomatic event, enriching the source region in some trace and minor elements (Chapter 6; P. Metcalfe, in preparation). On the basis of the whole-rock chemistry, it is considered likely that the volcanics represent primary melts and an isotopic study may reveal more about heterogeneities in the mantle (Chapter 6; Metcalfe and Smith 1986) and the lateral extent of the depleted mantle postulated by Bevier (1983b, 1986) to underlie the accreted terranes. 7.1.1 TECHNIQUES Samples from the Wells Cray—Clearwater area were chosen to represent the range of ages (Table 7-1; Chapter 4) present in the assemblage, except for the Holocene centres which are under investigation by P. Metcalfe (in preparation; Metcalfe and Smith 1986). As far as possible, the samples also represent the products erupted from separate centres (Chapter 3; Fig. 7-2), and are chemically characteristic of the suite (Fig. 7-3; Chapter 6). Sample locations are plotted on the 137 Figure 7-1: Generalized terrane subdivisions within the major orogenic belts of the Canadian Cordillera. AX, Alexander Terrane; BR, Bridge River Terrane; CA, Cassiar Terrane; CC, Cache Creek Terrane; CD, Cadwallader Terrane; CC, Chugach Terrane; CN, Chilliwack-Nooksack Terrane; JF, juan de Fuca Oceanic Plate; KO, Kootenay Terrane; MO, Monashee Terrane; MT, Methow. Terrane; OZ, Ozette Terrane; PR, Pacific Rim Terrane; SM, Slide Mountain Terrane; ST, Stikinia Terrane; QN, Quesnellia Terrane; WR, Wrangell Terrane; YA, Yakutat Terrane; YT, Yukon-Tanana Terrane. From Price et al. (1985; their Fig. 3 and Table 1, p.3-3). 138 T A B L E 7-1: Sr I S O T O P I C RESULTS S a m p l e Rb Sr Rb /S r 8 7 S r / 8 6 S r e S r t f (Rb/Sr) t K -A r * N u m b e r ppm ppm ±1a Ma 1) 34605 37 751 0.049 0.703394 + 17 -15.7 0.63 (1.6?) 2) S20102 11 433 0.027 0 . 7 0 3 4 1 7 ± 1 7 -15.4 -0.10 0.18 3) 32206 15 472 0.032 0.703516 + 18 -14.0 0.07 0.27 4) 41508 14 486 0.029 0.703563 + 1-3 -13.3 -0.03 0.30 5) 10503b 20 458 0.044 0 . 7 0 3 6 5 3 ± 1 5 -12.0 0.47 0.38 6) Sl0706d 16 494 0.032 0 . 7 0 3 6 4 3 ± 1 8 -12.2 0.07 0.56 7) SI0309 16 473 0.034 0.703743 + 18 -10.7 0.13 0.15 8) S10401 12 417 0.029 0 . 7 0 3 8 9 5 ± 2 2 -8.6 -0.03 1.9 9) S20303 11 462 0.024 0 . 7 0 3 9 6 2 ± 2 2 -7.6 -0.20 0.28 10) S10306 27 611 0.044 0.704043 + 09 -6.5 0.47 0.27 11) S10803 18 467 0.038 0 . 7 0 4 5 6 0 ± 1 4 0.9 0.27 0.55 12) S10901 19 371 0.051 0 . 7 0 4 8 6 9 ± 1 5 5.2 0.70 0.09 Results in table are listed in order of increasing 8 7 S r / 8 6 S r values. NBS 987 standard was run, along with the unknown samples, and gave 0.71024 + 3. A compilation of 11 runs, run concurrently with this work by Smith (1986), gave 0 . 7 1 0 2 6 ± 3 . Laboratory value is 0.71025. t Normalized to UR (uniform reservoir) of 0.7045 (DePaolo and Wasserburg 1976a, 1976b) * DePaolo and Wasserburg (1976a, 1976b) (Rb/Sr) ° = 0.03 from O'Nions et al. (1978) U R * Details of K-Ar dates are given in Chapter 4, and Table 4-1. 139 generalized geologic map in Figure 7-2 and can be found on Plate 2. 7.1.2 8 7 Sr/8 6 Sr 8 7 S r / 8 6 S r ratios were determined on 12 whole-rock powders using a VC MM 30 mass spectrometer at the University of Alberta, under the guidance of H. Baadsgaard. The procedure used for the extraction and purification of Sr involves Ba co-precipitation and ion exchange (Chaplin 1981) using 0.1 g of sample in a clean air environment. Determination of Sr and Rb was made using XRF (Appendices B and D). Concentrations averaged 20 ppm Rb and 400 ppm Sr, these values are high enough that spiking of the sample was not necessary. Yield was sufficient (approximately 75%) so that only half of the purified Sr precipitate was loaded onto outgassed, double rhenium filaments. 7.1.3 1 " 3 N d / 1 ""Nd Six whole-rock powdered samples (approximately 0.4 gms) were analysed, in a clean air environment, for 1 4 3 Nd/ 1 ft * Nd ratios and by isotope dilution for Sm and Nd values at the University of Alberta under the guidance of H. Baadsgaard using the technique of Cavell and Baadsgaard (1985). Final ignition of the residue in a platinum vessel, as called for by Cavell and Baadsgaard (1985), was not necessary and excess alkalies and organic material were removed by passing the solution through a very small cation ion exchange column (for a total of four times through ion exchange columns). The resulting eluant was evaporated to dryness after adding 3 drops of H N 0 3 , then loaded onto the side filament of outgassed, double rhenium filaments. The isotopic analyses were obtained on a VC MM 30 mass spectrometer. 140 G E N E R A L I Z E D G E O L O G Y O F T H E W E L L S G R A Y - C L E A R W A T E R A R E A , B . C . 1 2 0 ° 1 f W 1 g O ° 0 0 ' W 1 1 9 ° * 5 ' W H O L O C E N E A S S E M B L A G E P L E I S T O C E N E A S S E M B L A G E S S u b a e r i a i basal t 111[11 j S u b a e r i a i b a s a l t m Pseudo-craters L a c u s t r i n e d e o o s i t s •jf Subaeriai pyroclastics | ; j Ice c o n t a c t d e p o s i t s Figure 7-2: Generalized map of the Wells Gray—Clearwater area showing the locations of samples collected for isotopic determinations. Samples for which 8 7 S r / 8 S S r , 1 4 3 N d / i a 4 N d and whole-rock lead data were obtained (•): Samples with only 8 7 S r / 8 6 S r data (•). Also shown are the terrane subdivisions within the area. 141 K o O (cation ea.) N O R M A L A Clearwater unit + ice contact V Wells Gray unit • Mann Creek area ENRICHED O Enr iched Ice contact unit A Young cones and flows Figure 7-3: Zr/Nb vs. K 2 0 for the complete suite of analyses of the Wells Cray—Clearwater area. Samples with isotopic determinations are solid; numbers refer to the order in Table 7-1 and indicate increasing 8 7 S r / 8 S S r initial ratios. 142 T A B L E 7-2: S m , N d a n d 1 4 3 N d / 1 4 4 N d ISOTOPIC RESULTS S a m p l e N u m b e r S m ppm N d 1 ppm 4 7 S m / 1 4 4 N d 1 4 3 N d / 1 4 4 N d * e N d t 1 4 7 ft S m / 1 4 4 N d S10306 6.3 29.4 0.129 0 . 5 1 2 8 3 7 ± 4 7 4.23 -0.344 S10803 6.1 27.9 0.133 0 . 5 1 2 7 6 2 ± 2 8 2.75 -0.324 S10901 (3.8) (22.0) n.d. 0 . 5 1 2 7 0 0 ± 2 0 1.54 n.d. S20303 5.3 23.4 0.136 0 . 5 1 2 9 2 8 ± 2 0 6.01 -0.309 32206 5.1 21.8 0.141 0 . 5 1 2 9 2 2 ± 1 9 5.89 -0.283 41508 5.3 22.9 0.138 0.512836 + 20 4.23 -0.298 La Jolla Nd standard was run, along with the unknown samples, and gave 0.511843 ± 1 9 . A compilation of 21 runs, run concurrently with this work by Smith (1986), gave 0.511841 ± 1 7 . n.d.; not determined. Bracketed Sm and Nd values were determined by AA. * Normalized to 1 4 6 Nd/ 1 4 4 Nd 0.7219 (O'Nions et al. 1979). t Ratio to CHUR (chondritic uniform reservoir) of 0.51262 (O'Nions et al. 1979) following DePaolo and Wasserburg (1976a, 1976b). * Jacobsen and Wasserberg (1980); DePaolo and Wasserburg (1976a, 1976b) ( 1 4 7 S m / i • • N d ) » i m = 0.1967 CHUR 7.1.4 WHOLE-ROCK LEAD Whole-rock powdered samples (approximately 15 mg) were digested by a mixture of HF and H N 0 3 acids in clean, covered teflon beakers in a clean air environment. Lead was separated by anion exchange chromatography using HCl and HBr. Purified samples were loaded onto previously outgassed rhenium filaments using the silica gel-phosphoric acid technique. Analyses were performed by M. L. Bevier at the Geochronology Section, Geological Survey of Canada, Ottawa, on a Finnigan MAT 261 mass spectrometer equipped with fully adjustable multiple collectors. Four of the Faraday collectors were positioned to simultaneously collect ions of lead with masses of 208, 207, 206, and 204. Measured Pb isotopic compositions were corrected for fractionation by factors determined by the average deviation of NBS 981 from its absolute value (Todt et al. , 1984). 143 T A B L E 7-3 : W H O L E - R O C K P b I S O T O P I C R E S U L T S S a m p l e n u m b e r 2 0 8 p b / 2 O 4 p 5 2 0 7 p b / 2 0 4 p b 2 0 6 p b / 2 o a p b S10306 38.639 15.609 18.851 S10803 38.547 15.560 18.621 S10901 38.683 15.632 18.767 S20303 38.672 15.624 19.090 32206 38.515 15.597 18.861 41508 38.591 15.624 18.797 l a within-run precision averages 0.03% for all Pb isotopic ratios. A measured Pb procedural blank of 89 picograms is less than 0.01% of the total sample Pb and thus no blank correction was made. 7.2 RESULTS AND DISCUSSION Results for each isotopic system are given in Tables 7-1, 7-2 and 7-3. Figure 7-4 shows the combined results for 8 7 S r / 8 6 S r and 1 4 3 Nd/ 1 ft " Nd and Fig. 7-5 gives the Pb isotope correlation diagrams. The measured ratios are considered initial ratios because of the young age of the rocks. All three isotopic systems plot away from the field defined for MORB (mid ocean ridge basalts) and do not form linear arrays (Figs. 7-5 and 7-6). The variation in the isotopic values is considered to be greater than can be explained by differing degrees of partial melt of an isotopically homogeneous source region (Barberi ef al. 1980) and indicates that the mantle may be heterogeneous on the scale of a few kilometres. Several models have been put forward to account for spread in the initial ratios of Pb, Sr and Nd isotopes, including: 1) mixing of isotopically distinct mantle sources (see for example Hart 1985; DePaolo 1983); 2) enrichment of depleted mantle by influx of radiogenic volatiles from a 144 Figure 7-4: Plot of 1 " 3 Nd/ 1 * * Nd vs. B 7 S r / 8 S S r . Results from the Wells Cray—Clearwater area are marked by a star. Solid squares are results from Chilcotin Group basalts (Smith 1986). Dark solid line (ORL) is the oceanic regression line of O'Nions et al. (1979). Stipple delineates the field of mixing between a chondritic uniform reservoir and depleted mantle or MORB source. Also outlined is the the trend for sediment recyclying (White and Hoffmann 1982). Field of horizontal lines marks the results from Holocene cones in the Wells Gray—Clearwater area (Metcalfe in preparation; Metcalfe and Smith 1986). EPR, dash-dot field, is from Zindler et al. (1984) for the East Pacific rise; BR — - — is the Basin and Range field from Hart (1985) ; HA, —.. — .. are results from Hawaii compiled by Smith (1986). 145 40.0 15.30 19.5 p b / 2 0 4 P b Figure 7-5: Plot of lead isotopic values for the Wells Cray—Clearwater area basalts (•). CRB, — Columbia River Basalts and the Saddle Mountain unit stippled (Church 1985). The Chilcotin Croup basalts results are shown as dark stipple (Bevier 1983b), the western Anahim Volcanic Belt results as a the field of diagonal lines (Bevier 1986; in preparation). The field of pelagic clays (cross-hatch) is shown from Cohen and O'Nions (1982) and O'Nions (1984). Oceanic regression line (ORL) is from Church and Tatsumoto (1975) and is based on oceanic type depleted mantle from the Juan de Fuca—Corda plates. 146 lower mantle source (see for example Worner ef al. 1985); 3) melting of mantle that has been contaminated by subducted oceanic sediments (see for example R.L. Armstrong 1981; White and Hoffman 1982); 4) contamination of melts from the mantle with continental crustal material (see for example Downes 1984); 5) or a combination of one or more of these (see for example Church 1985; James 1982; Worner et al. 1985). Each of the other individual models is a possibility that may explain the isotopic variations seen in the Wells Gray—Clearwater area and will be discussed in turn. 7.2.1 CRUSTAL CONTAMINATION OF DEPLETED MANTLE The systematic decrease in 1 " 3 N d / 1 * 4 N d coupled with increasing 8 7 S r / 8 6 S r (Fig. 7-4), is characteristic of a source region which has undergone depletion by the generation of prior partial melts. These partial melts have depleted the reservoir in Rb and Nd relative to Sr and Sm (O'Nions et al. 1979; White and Hoffmann 1982; Carlson et al. 1981, Jacobsen et al. 1986). A plot of the chemical fractionation factors (Fig. 7-6) indicates fairly uniform depletion of Nd but varying degrees of enrichment in Rb (or depletion in Sr). If the Rb/Sr variations are inherited from a variously depleted mantle source then 8 7 S r / 8 S S r vs. Rb/Sr (Fig. 7-7) should show correlation as well as a 7 S r / 8 6 S r vs. Rb and 8 7 S r / 8 6 S r vs. Sr. In these data the values are only looseiy correlative. Rb 'enrichment', relative to Sr may also arise through varying degrees of partial melting as Rb is slightly less compatible in the solid phases present in the mantle than Sr (Cox et al. 1979). The elements, Ti and K show a similar relationship and all four of these elements rank in incompatibility as K > Rb > Sr > Ti. Figure 7-8A shows Rb/Sr vs. Ti/K coupled with the 8 7 S r / 8 6 S r results and geographic location. The ratios suggest that control of the elemental abundances is through varying amount of partial melting or heterogeneity 147 +.5 depletion E CO \5 - . 5 • • • enrichment - . 5 0 + .5 f (Rb/Sr) Figure 7-6: Fractionation factors show enrichment of basalts from the Wells Cray—Clearwater suite in Rb and Nd relative to Sm and Sr. f factors are as defined by DePaolo and Wasserburg (1976a, 1976b). .7065 | j . .7080 .7045 .7040 .7036 .7030 Figure 7-7: 8 7 S r / B 6 S r vs. Rb/Sr ratios. Solid line is the 4.55 Ca geochron (Church 1985). OIB , outline field of oceanic island basalts; CRB, is that defined by the Picture Gorge, Imnaha, Steens Mountain, and Grande Ronde members of the Columbia River Basalts (Church 1985). Figure after Church (1985). 1 Figure 7-8: A) Relationship between the ratios Rb/Sr vs. Ti/K. 8 7 S r / 8 S S r results are shown as numbers beside the samples ranked from 1, lowest 8 7 S r / 8 S S r ratio to 12, the highest. General geographic locations are given in Figure 7-1. B) Similar plot for Sm/Nd vs. Ti/K ratios with generalized location names. (Plate 1) C) Isotopic results for Nd. Points do not define a pseudoisochron. 149 in the source region and the isotopic ratios do not reflect the elemental ratios. In terms of 8 7 S r / 8 6 S r , the variations reflect an outside source of 8 7 S r that has overwhelmed the 'inherited' initial ratio of the magma, but has not influenced the trace element abundances. In a plot of 8 7 S r / 8 6 S r vs. Rb/Sr (Fig. 7-7) the data plot to the right of the 4.55 Ga geochron, indicating the source region has a complex history and does not reflect material that has been isolated from mantle forming processes for 4.55 Ga. The Nd isotopic data can be interpreted in a similar manner to the Sr data. High Sm/Nd ratios (K Nd < Sm) coincide with high Ti/K values (Fig. 7-8B) but do not correlate with the isotopic ratios or show pseudoisochrons (Fig. 7-8C). The Sm/Nd ratios are consistent with a moderately depleted mantle source and there is less variation in the 1 4 3 Nd/ 1 fl 4 Nd initial ratios than in 8 7 S r / 8 6 S r intial ratios. The Nd and Sr elemental abundances can be interpreted as the result of derivation of the magmas by varying degrees of partial melting of an initially depleted source which has been contaminated by radiogenic crustal material which has enriched the source in 8 7 Sr and to a lesser extent 1 4 3 Nd. The relative enrichment probably reflects varying abundances of the radiogenic isotopes in the contaminating material. The amount of contamination is, however, still small, and amounts to less than 2% if the source is highly radiogenic crustal rocks (Fig. 7-9), based on the work of Beilski-Zyskind ef al. (1984). This low level of contamination does not result in pseudoisochrons (Fig. 7-7, 7-8A). If assimilation of crustal xenoliths by the magma is the source of the isotopic variation, then 15% contamination by upper crustal material (using the values of Beilski-Zyskind ef al. 1984) is necessary to get from the most isotopically depleted sample to the least depleted sample. Figure 7-5, a plot of the Pb data, shows that data from the Wells Cray—Clearwater area plots above the oceanic regression line as well as the results 150 c CO E PN 3 Figure 7-9: e plot showing the crustal contamination results of Beilski-Zyskind ef al. (1984), K values are the ratio of Sr/Nd , , , /Sr/Nd (depleted source) (crustal) 151 from the western Anahim belt and the Chilcotin lavas (Bevier 1983a, 1986, in preparation), but overlaps the field of the Columbia River Basalts (Church 1985). Bevier (1983b, 1986, in preparation) has interpreted the results from the Anahim Volcanic Belt and the Chilcotin Group basalts as derived from mantle material which is isotopically similar to the Juan de Fuca seamount source and has not undergone modification or contamination during ascent. The Columbia River Croup basalt fields are interpreted as being derived from either mantle material which has been contaminated with oceanic and crustal sediments (Church 1985), or mixing of two distinct mantle derived magmas which have undergone crystal fractionation and limited crustal contamination (Carlson ef al. 1981, 1983). Church (1985) ruled out crustal contamination for most of the Columbia River Group by the lack of crustal xenoliths in the basalts. The abundance of crustal xenoliths in the Wells Gray—Clearwater suite means that this cannot be ruled out as the cause of the isotopic variation, but crystal fractionation is unlikely (Chapter 6) and if mixing of two sources has occurred the result has been obscured by partial melting effects. The high values for 2 0 8 P b / 2 o a P b and 2 0 7 P b / 2 0 I , P b are most consistent with limited assimilation of radiogenic rock (Downes 1984; DePaolo 1983; Francis et al. 1980). From the Nd isotopic data and the whole-rock chemistry (Chapter 6), the amount of contamination is small and should not result in the strong covariation of Pb systematics with B 7 S r / 8 6 S r (Fig. 7-10) or 1 U 3 N d / i a a N d (James 1982). It is possible that the Sr results, and possibly the other isotopic systems, reflect cumulate and/or xenocrystic olivine. In Sr isotopic work by Min (1985), whole-rock mantle xenoliths and mineral separates collected at several late Tertiary volcanic centres in the east-central and southern Cordillera were analysed. It was found that the whole-xenoliths have low present day ratios (0.7025 to 0.7030), but the component minerals, in particular olivine, had ratios up to 0.715. It is possible that the Sr initial ratios may be affected by cumulate and/or xenocrystic olivine but 152 for the rocks examined in this study there is no evidence that they are olivine cumulates or that the olivine phenocrysts are in fact xenocrysts (Chapter 4, 6). Also, the amount of Sr contained in the olivine is very small (less than 3 ppm and usually under 1 ppm; Min 1985) and although the 8 7 S r / 8 6 S r ratio is high, large amounts would have to be present to affect the whole-rock ratios which typically have over 350 ppm Sr (Table 7-1). 7.2.2 OLD SUBCONTINENTAL LITHQSPHERE The Pb results from the Wells Cray—Clearwater area plot well below those of the Saddle Mountain unit of the Columbia River Basalts (Fig. 7-5). The results from the Saddle Mountain lavas require from 15-20% assimilation of continental crustal rocks to explain the isotopic data (Carlson et al. 1981). This degree of contamination is not evident in the major and trace element patterns and it was concluded that the Saddle Mountain lavas originate in underplated upper mantle rocks that have been isolated from mantle convection for 2.6 Ca (Carlson et al. 1981; Church 1985). This 2.6 Ca subcontinental lithosphere extends eastward from the Columbia River Plateau into the Snake River—Yellowstone Plateau (Leeman 1975; Doe et al. 1982), but no evidence of it appears in the data collected in this study, or by others in the Canadian Cordillera (Bevier 1983b, 1986, in preparation; Smith 1986; Metcalfe and Smith 1986). Mineral isochron dates determined by Min (1985) on nodules from the Canadian Cordillera give Late Proterozbic to Mesozoic ages. The dates ranges from 645 Ma to 104 Ma (Min 1985). 7.2.3 SEDIMENT CONTAMINATION OF THE MANTLE SOURCE The possibility of recycling oceanic sediment into the mantle has been discussed and debated by Moorbath (1978) and R.L. Armstrong (1981) among others. In the most recent work on oceanic islands evidence for contamination by 153 oceanic sediments results in a subparallel trend of points away from the mantle array in the Nd isotopic system (Fig. 7-4) (O'Nions 1984; White and Hoffmann 1982). The Wells Cray—Clearwater area data do not show a similar trend and plot close to the mantle array line of O'Nions (1984) (Fig. 7-4). On the 2 0 7 Pb/ 2 0 4 Pb vs. 2 0 6 Pb/ 2 0 4 Pb diagram (Fig. 7-5) pelagic sediments plot slightly above and to the right of the field for the Wells Gray—Clearwater area data and on 2 0 7 p b / 2 0 4 p b _ 2 0 6 p b / 2 0 « p b v $ 8 7 S r / 8 6 S r ( p j g j^Q) t h g d a t a p | Q t b e t w e e n t h e field of MORB and oceanic sediments. Although it is arguable whether or not the data presented here supports the hypothesis of contamination of the mantle by subducted oceanic sediments, this process may explain some of the heterogeneities seen in the trace element abundances. The Zr/Nb ratios (Fig. 7-3) suggest that the mantle is more depleted than a MORB source, yet the abundances of other elements show enrichment and all the isotopic results plot away from the MORB source field. Enrichment of depleted mantle by subducted oceanic sediments may explain this apparent anomaly. On the P/K vs. Ti/K plot (Chapter 6, Fig. 6-3) localized sources plot as clusters, indicating a similar mantle source for each area, but variation in the source composition on a scale of a few kilometres. In terms of isotopic ratios (P. Metcalfe, in preparation; Metcalfe and Smith 1986) (Fig. 7-4) the Holocene centres cannot be distinguished from one another. Contamination of the mantle source by subducted pelagic sediments would lead to many separate reservoirs with varying Th/U, U/Pb, Rb/Sr, and Sm/Nd ratios. Each of these reservoirs would evolve in an isotopically distinct manner from adjoining areas but the heterogeneities must survive more than 10 9 years (Cohen and O'Nions 1982) to be distinguishable. Mantle nodules from the Canadian Cordillera (Smith 1986) have 8 7 S r / 8 6 S r values ranging from 0.7030 to 0.7119 and eNd from -20 to +8. Smith attributed these variations to long lived (>1 Ga ) heterogeneities in the mantle. If the heterogeneities are 154 . Q Q . 40.0 39.5 39.0 o CM J3 38.5 a. co o CV 38.0 37.5 19.0 JO a. * 18.5 o CM J Q a. S 18.0 CM 17.5 I L C R B ^ . . 31 S N A K E RIVER G R O U P O C E A N I C SEDIMENTS CQ OC O 2 CHILCOTIN GROUP B A S A L T S ( v. o J I I I L 0.702 0.704 0.706 0.708 0.710 0.712 0.714 8 7 S r / 8 a S r Figure 7-10: Plot of whole-rock Pb and Sr isotopic data modified after Church (1985). Wells Gray—Clearwater area results •; solid line is the MORB line defined by O'Nions (1984); Chilcotin basalts (Bevier 1983) are shown in stipple pattern; C R B , outline is the Steens Mountain, Picture Gorge, Imnaha, Grande Ronde and Wanapum units of the Columbia River Basalts (Church 1985), also outlined are the fields of pelagic ocean sediments (cross-hatch) and the field of Snake River basalts (diagonal lines) (Church 1985). 155 preserved, the resulting mantle structure may be similar to that envisioned by Zindler et al. (1984; their Fig. *A, p. 188) and result in the isotopic and trace element patterns seen in lavas today. 7.3 CONCLUSIONS Isotopic results from the Wells Cray—Clearwater area are most consistent with derivation of the magmas from moderately depleted mantle. Others have suggested that this depletion is the result of earlier melting events and magma extraction. Variations in the elemental ratios of the magmas is a recent phenomena, related to varying degrees of partial melting and is not correlated with the initial ratios of the isotopes. Enrichment in radiogenic 8 7 S r , 1 t t 3 N d , 2 0 8 Pb and 2 0 7 Pb is the result of contamination of the reservoir with small amounts (less than 2%) of radiogenic, upper crustal rocks. Crustal xenoliths enclosed within the magma are either unmelted or only partially melted such that crustal contamination had minimal effect on the isotopic, major and trace element abundances. 8. THE WELLS CRAY-CLEARWATER AREA IN A REGIONAL CONTEXT If a man will begin with certainties, he shall end in doubts; but if he will be content to begin with doubts, he shall end in certainties. — Francis Bacon 8.1 INTRODUCTION The relationship of the Wells Gray—Clearwater suite of volcanic rocks to other volcanic rocks in the Canadian Cordillera has been the source of much speculation. The area was mapped at a scale of 1:250,000 by Campbell (1963a; 1963b; 1967) and Campbell and Tipper (1971) who correlated the flat lying, valley-filling flows with flat-lying lavas of the Interior Plateau now referred to as the Chiicotin Group basalts (Bevier 1983a). The Chilcotin Group basalts are far-travelled flows in the form of low shield volcanoes with vents now marked by gabbro or dolerite plugs (Bevier 1983a). They are alkaline to transitional in chemical affinity and appear to represent primitive melts which are largely unmodified by crustal contamination and crystal fractionation (Bevier 1983b). The Clearwater basalts were considered to be a distal facies of these lavas, formed where flows at the eastern margins of the plateau were channelled into valleys along the bounding Quesnel and Shuswap Highlands (Campbell and Tipper 1971). The Holocene centres which overlie the valley-filling • flows were inferred to mark the easternmost visible expression of the Anahim Volcanic Belt (Souther 1977; Hickson and Souther 1984). Souther (1977), described the Anahim Volcanic belt (Fig. 8-1) as extending approximately east-west along latitude 52° and is defined by three large bimodal shield volcanoes and 37 Quaternary basaltic centres. Bevier et al. (1979), suggested that this belt may reflect the trace of a mantle plume. W.H. Mathews (oral communication 1984), has suggested that the Nazko area, northwest of Williams Lake, may mark the eastern termination of the Anahim Volcanic Belt and Souther (1986) based on this, suggested that the Clearwater area may represent a second 156 157 Figure 8-1: Quaternary volcanic centres in British Columbia with volcanic centres of the Middle Miocene to Holocene Anahim Volcanic Beit shown as stars. Solid triangles are known Quaternary volcanic centres (Mathews, in press) The numbered lines are the best fit tracks for the Euler poles considered in Table 8-3. Solid lines are for King Island to Clearwater—Wells Cray (excluding Nazko Cone) and the dashed lines are for King Island to Nazko Cone (excluding the Wells Cray—Clearwater area). The line lengths represent the motion of the North American plate over the hot spot in 15 Ma. The dash-dot line extending southward form the WC-CLW area, links other valley-filling Quaternary centres. • • 158 parallel hot spot track. W.H. Mathews (oral communication 1984) has also pointed out that there may be a link between young volcanic centres to the south between Clearwater and Princeton (Fig. 8-1) which form a linear trend along the west side of the Omenica Crystalline Belt (Fig. 7-1). A number of these centres form valley-filling flows similar to those in the Wells Cray—Clearwater area (Preto 1974; Church 1979) and are less than a million years old. As shown by Hickson and Souther (1984) and this work (Chapter 4), valley-filling and piateau-capping basalts are younger than the Chilcotin Group (Bevier 1983a). However, the possibility remains that the entire sequence of basalts defines the eastern end of the Anahim Volcanic Belt. Subsequent to the work of Bevier ef al. (1979), more studies have been undertaken on several of the centres in the Anahim Volcanic Belt and more isotopic dates are available across the breadth of the belt (Souther 1986). This new information can be used to critically test the mantle plume hypothesis by comparing the timing of volcanism along the track and the coincidence of the path to that predicted by the motion of the North American Plate. Continental rifting is explored as a possible alternative explanation for the volcanism in the Wells Gray—Clearwater area and for other young volcanic centres in the Canadian Cordillera to the south and north of the Wells Gray—Clearwater area, but first work on mantle plumes will be reviewed. 8.1.1 MANTLE PLUMES Linear chains of volcanoes have puzzled geologists for some time. With the advent of plate tectonic theories, many of the belts could be explained in terms of subduction (arc volcanism), or the generation of new crust (oceanic ridge crest volcanism). There remained however, a number of chains that appeared to have no relationship to these processes, in particular the Hawaiian Islands. 159 The origin of mid-plate linear chains of volcanoes, such as the Hawaiian Islands, from thermal plumes in the mantle, was first suggested by J. Tuzo Wilson (1973). He recognized their significance in understanding mantle and crustal processes and pointed out that there tends to be a greater concentration of hot spots on slow moving or stationary crust. In experimental modelling of the mantle by Richter and Parsons (1975) and McKenzie and Richter (1976), a series of polygonal cells with central areas of upwelling could be generated. Debate continues as to whether the motion of the plate will disrupt the cell pattern (mantle plume) and form roll structures (Marsh and Marsh 1976) or whether the cell pattern will remain intact and the cell layer will decouple at the low velocity zone (McKenzie and Richter 1976). Thiessen et al. (1979) applied the 'cell' model to the African Plate and pinpointed some 45 topographic high spots which have a pattern and a separation that closely resemble those predicted by the mantle plume experiments. The African high spots are of two types: those that coincide with areas of stable craton which has not been active since the Precambrian; and high areas with central volcanoes on craton which has been tectonically active since the Precambrian. They inferred from these relationships that it is more difficult for mantle plumes to melt through stable craton than active tectonic zones. Recent seismic tomography supports a model in which the plume decouples at the low velocity zone and indicates that the convective pattern may in fact be very deep, involving convection from the core/mantle boundary to the upper mantle (Kerr 1984). Hot spots like Hawaii and Iceland appear to have "hot roots" which extend well into the lower mantle (Kerr 1984). 160 8.1.2 PLUME TRACKS Mantle plume tracks are sensitive indicators of plate motion, as well as mantle motion. However, tracks on continental plates tend to be poorly defined and scarcer than those on oceanic plates. The North American Plate has a greater rate of motion than the African Plate (considered to be stationary by Minster and Jordan 1978), but its motion is still slow compared to the Pacific Plate (2.7 cm/yr compared to 8.9 cm/yr). If an analogy can be made between the North American and African continents, some 36 hot spots should exist beneath North America. The fact that only two continental hot spots are recognized is probably a consequence of a number of factors, including the possibility that motion of the plate may diminish the effect of the plume, particularly if the plate is continental. Other hot spot effects may be subtle and not easily recognized. The altitude of a particular segment of crust is thought to be related directly to the length of time since that segment has overridden a mantle plume (Crough 1983). The vast exposures of Precambrian basement rocks in the Canadian Shield are thought to result from the North American Plate overriding the Great Meteor hot spot some 200 million years ago (Vink et al. 1985). The western margin of North America contains abundant Late Cenozoic volcanic centres, some occurring in well-defined linear Belts. The Garibaldi—Cascade belt and the Aleutian Islands are thought to be directly related to subduction along the western margin of North America. Others, such as the Chilcotin Croup and Columbia River Group basalts, appear to have resulted from back-arc volcanism related to the subduction of the Farallon, Kula and the present-day Juan de Fuca Plates. The Stikine Belt appears to be related to trans-tensional rifting (Souther et al. 1984; Souther and Hickson 1984). Both the Snake River-Yellowstone Volcanic Belt and the Anahim Volcanic Belt have been suggested to have mantle plume origins (Vink ef al. 1985; Bevier ef al. 1979) based on linearity and diminishing ages from 161 west to east. 8.2 PLATE VELOCITY 8.2.1 DATA Table 8-1 presents the oldest obtained isotopic ages for several centres in the Anahim Volcanic Belt (King Island, Rainbow, llgatchuz and Itcha Ranges, and the Wells Cray—Clearwater area (WC—CLW)). The calculated rate of plate motion is given along a great circle from each centre to the next. The alternative eastern termination for the Anahim Volcanic belt, a young cinder cone at Nazko (Fig. 8-1; Souther 1986; W.H. Mathews, oral communication 1984) is also tested. Nazko appears to be an isolated cinder cone and although it has had at least two phases of activity (Hickson, unpublished data 1985), these occurred within a period of some hundreds or thousands of years. The table lists the calculated rate of motion from the Itcha Range to Nazko and from the Itcha Range to the Wells Cray—Clearwater area. To minimize possible errors in the dating, average rates are also given for the full lengths of the belts (Anguita and Hernan 1975). The Clearwater area is deeply dissected and has numerous dated flows. Subaqueously quenched basalts make interpretation of some of the dates difficult (Chapter 4), but the consistency of the remaining dates indicates onset of a major volcanic phase approximately 600,000 years ago. Using the older K-Ar dates (Chapter 4) would increase the apparent velocity between the Itcha Range and the Wells Gray—Clearwater area. 8.2.2 RESULTS The centre-to-centre velocities (Table 8-1) are slightly lower than those predicted by Minster and Jordan (1978) from eastern North America but are T A B L E 8 - 1 : P O S I T I O N S AND R E L A T I V E V E L O C I T I E S B E T W E E N V O L C A N I C C E N T R E S , A N A H I M B E L T L o c a t i o n L a t . L o n g . A g e G r . C i r c l e A g e D i f f . A p p . V e l . ( M a ) D i s t . ( k m ) ( M a ) ( c m / y r ) G r . C i r c l e C u m . T i m e C u m . C u m . ( M a ) A v e . V e l . D i s t . ( k m ) ( c m / y r ) K i n g I s l a n d 5 2 ' 0 6 ' 1 2 7 ' 4 8 ' 1 3 . R a i n b o w R a n g e 5 2 ' 4 1 ' 1 2 5 ' 4 9 ' r 8 . 7 ! I l g a c h u z R a n g e 5 2 ' 4 5 ' 125 ' 1 8 ' L 6 . 1 1 I t c h a R a n g e 5 2 ' 4 2 ' 1 2 4 ' 5 2 ' r 3 . 5 W G - C L W I t c h a R a n g e N a z k o C o n e 5 2 ' 0 0 ' 120 ' 1 4 ' 0 . 6 52 5 6 ' 123 41 149 35 3 0 3 2 3 8 0 4 . 3 2 . 6 2 . 6 3 . 5 3 . 5 3 . 5 1 . 3 1 . 2 9 . 2 2 . 3 149 183 2 0 9 5 1 5 291 4 . 3 6 . 9 9 . 5 1 2 . 4 13 . 3 . 5 2 . 7 2 . 2 4 . 1 2 . 2 1 S o u t h e r ( 1 9 8 6 ) ' B e v i e r ( 1 9 7 8 ) 163 reasonable, given the difficulty of dating the oldest material present at each of the volcanic centres. The slightly higher velocity, between King Island and the Rainbow Range, would be reduced to 2.8 cm/yr if the surface volcanism at King Island predated the crystallization of the pluton by some million years, a not improbable figure. The alternate segment, from the Itcha Range to Nazko, is consistent with these results, but serious problems are encountered if the Wells Cray—Clearwater area is considered the termination point. In this case, the calculated rate is in excess of 9 cm/yr, an improbably high figure. The average velocities, taken along great circle paths from King Island and terminating at each centre in turn, are more in line with the predicted velocity of 2.7 cm/yr for the North American Plate. The King Island to Wells Cray—Clearwater area velocity is substantially lowered from the last segment velocity. This reflects the southerly position of the Wells Cray—Clearwater area (Fig. 8-1), which considerably diminishes the arc length between the end members. A Nazko 'end point' is still more consistent with the topology than the Wells Cray—Clearwater area. 8.3 EULER POLES 8.3.1 METHOD To further test the consistency of the mantle plume track hypothesis, a computer program was developed to calculate the Euler pole for the track. The Euler pole that best fits the volcanic centres is found by a maximum likelihood technique, details are given in P. Hickson and C.J. Hickson (in preparation). As well as computing the Euler pole for a given track, the program will compute a maximum likelihood solution for the track to a given Euler pole position, such as that determined from other mantle plume track data. 164 T A B L E 8-2 : C A L C U L A T E D EULER P O L E S E n d p o i n t Lat . L o n g . D i s . to R M S e r r o r L o c a t i o n N P o l e (km) 1 0 k m a s s i g n e d sca t te r Nazko Cone 34°N 77°E 91.55° 7.65 3.8250 W G - C L W 47°N 124°W 5.58° 7.86 3.9309 Yellowstone 38°N 110°E 90.00° 3.48 .1100 3 0 k m a s s i g n e d sca t te r Nazko Cone 35°N 77°E 90.12° 7.91 1.3188 W G - C L W 33°N 52°E 94.63° 31 .14 5.1895 Yellowstone 38°N 110°E 89.45° 2.42 0.4849 1 0 0 k m a s s i g n e d sca t te r Nazko Cone 35°N 77°E 90.01° 7.74 0.3870 W G - C L W 37°N 51°E 90.38° 31.26 1.5629 Yellowstone 38°N 110°E 90.00° 3.63 0.2176 fx 2 in bold type indicate probabilities of >50% that the track fits the calculated pole. Since the track is assumed to be a curve, the degrees of freedom are (n-3). n = 5 for Nazko and WG —CLW segments and n = 6 for Yellowstone The initial problem in the determination of the pole to a given track, is to estimate the expected width of the mantle plume track. A figure of 100 km is commonly used and even a width of 200 km has been suggested (Thiessen ef al. 1979). After 10 Ma at a rate of 2 cm a" 1 , a track would be only 200 km in length and should not be clearly discernible. This is in contrast to the apparently narrow, well defined track that both the King Island to Itcha Range segment (Fig. 8-1) and the Yellowstone track exhibit (Table 8-2). An expected RMS (root mean square) scatter of 10 km (95% of the centres falling within 19.6 km of the track), 30 km (95% of the centres falling within 58.5 km of the track) and 100 km (95% 165 of the centres falling within 196 km of the track) has been assigned to the data and the poles calculated for each case (Tabie 8-2). These data were also tested by comparing each of the calculated poles to that calculated from the data for the Yellowstone track; to the pole calculated by Duncan (1984) for the New England-Great Meteor track, and to that obtained by Minster et al. (1973, 1974) and Minster and Jordan (1978). The poles of Minster et al. (1973, 1974) and Minster and Jordan (1978) are based on a solution that simultaneously solves for the pole that best fits data from many plates. The results are tabulated in Table 8-3. 8.3.2 RESULTS As can be seen in Table 8-2, calculated RMS error for the case of King Island to Nazko and for the Yellowstone track is better than the assigned error. The resulting x 2 values give probabilities of greater than 25% and 50% respectively, for the most tightly constrained case (10 km assigned RMS error). As the RMS assigned scatter for the track increases, the x 2 values drop significantly and the calculated position of the pole remains fixed. This drop in x 2 is also evident in the King Island to WG—CLW segment, but it is immediately evident that greater scatter in the data exists (RMS calculated error of >30 km) and the x 2 value does not drop to a probable level until an RMS assigned error of 100 km. As the RMS assigned error is changed, the pole shifts position significantly. The position of the tightly constrained 10 km pole reflects the extreme southward curvature of the track. The effect of this curvature is diminished in the less tightly constrained calculations. In Table 8-3, the fit of the tracks to various Euler poles is listed. The x 2 values indicate a poor fit of the data in most cases. None of the data fits the Yellowstone pole particularly well and even at 100 km RMS error, the probability is 166 • T A B L E 8-3: C O M P A R I S O N O F C A L C U L A T E D E U L E R P O L E S T rack U t . L o n g . D i s . to R M S x 2 t P o l e (km) 10 k m 30 k m 100 k m Yellowstone pole N a z k o C o n e 37.80°N 109.7°E 77.67° 39.19 19.5932 6.5321 1.9596 W G - C L W 37.80°N 109.7°E 79.20° 122.26 61.1315 20.3772 6.1132 S e a m o u n t 37.80°N 109.7°E 104.21° 49.61 29.7634 9.9211 2.9763 Minster et al. (1974) N a z k o C o n e 48.1°N 97.9°E 72.71° 19.35 9.6741 3.2247 0.9674 W G - C L W 48.1°N 97.9°E 73.07° 94 .22 47.1092 15.7031 4.7109 Y e l l o w s t o n e 48.1°N 97.9°E 84.46° 29.99 17.9933 5.9978 1.7993 S e a m o u n t 48.1°N 97.9°E 92.05° 32.20 19.3220 6.4407 1.9322 Minster and Jordan (1978) N a z k o C o n e 58.31°N 130.3°E 49.64° 41.59 20.7962 6.9321 2.0796 W G - C L W 58.31°N 130.3°E 50.06° 126.39 63.1946 21.0649 6.3195 Y e l l o w s t o n e 58.31°N 130.3°E 65.56° 4.58 2.7500 0.9167 0.2750 S e a m o u n t 58.31 °N 130.3°E 83.19° 84.01 50.4077 16.8026 5.0408 Duncan (1984, seamount data) N a z k o C o n e 61.8°N 85.7°E 62.92° 8.97 4.4826 1.4942 0.4483 W G - C L W 61.8°N 85.7°E 63.21° 64.34 32.1722 10.7241 3.2172 Y e l l o w s t o n e 61.8°N 85.7°E 73.60° 55.58 33.3476 11.1159 3.3348 * Tracks for the four poles are plotted in Figure 8-1. fx 2 in bold type indicate probabilities of >50% that the track fits the calculated pole. Since the track is assumed to be a curve, the degrees of freedom are (n-3). n = 5 for Nazko and W G — C L W segments and n = 6 for the Yellowstone and seamount tracks. slightly less than 5% that the King Island-Nazko track results from rotation around 167 that pole. The results are better when the Minster ef al. (1974) pole is used. The result for the Nazko track gives a better than 25% probability at the RMS error level of 30 km. WC—CLW, on the other hand, has a less than a 10% probability at RMS error of 100 km. The fit of most of the data to the Minster and Jordan pole published in 1978 is poor; the Nazko and seamount tracks have a <30% probability, and the WC—CLW segment a <5% probability at an RMS level of 100 km. The fit of the Yellowstone track to the same pole is, however, excellent. 8.3.3 CONCLUSIONS FROM EULER POLE AND VELOCITY DATA The apparent incompatibility of most of the data to the tested poles is an interesting result. Twenty mantle plumes were used in the calculations of Minster ef al. (1973, 1974) and nine in the later paper by Minster and Jordan (1978). Three hot spots were included from the North American Plate in the 1973 and 1974 calculations, but only those mantle plumes with "migration pertinent" to the last 10 Ma were used in the 1978 paper. In their 1978 paper, Yellowstone was the only pole included from the North American Plate and the statistical weighting assigned to it was second only to that assigned to the Galapagos-Carnegie Ridge. This has apparently strongly influenced their calculated pole, which is reflected by the excellent fit of the Yellowstone track to the pole, but the poor fit of the other data. The 1974 pole fits more of the data, probably reflecting the greater number of points that were used in the calculation. With more tracks, the result is less influenced by any single track that may have been incorrectly interpreted as a mantle plume. The New England track fits none of the multi-plate poles well, nor the Yellowstone track very well. The program used herein calculates poles in good agreement with those of Duncan (1984). Somewhat surprisingly, the King Island-Nazko track fits the 0-21 Ma pole of Duncan best, giving the lowest RMS value (8.97 km) and a better than 80% probability at an assigned RMS value of 168 100 km. The poor fit of the proposed tracks to what appears to be the better constrained euler poles bears consideration. It may be that some presumed mantle plumes (i.e. hot spot tracks) used to constrain the poles have alternate origins. Possible alternatives may be such things as propagating fractures (Anguita and Hernan 1975); leaky transform faults (Ross 1983); crustal flexure over a subducting slab (Stacey 1974); show strong control by prior tectonic fabrics; or the plumes are not fixed in the mantle (Chase and Sprowl 1984). The results for the King Island to Nazko Cone segment of the Anahim Volcanic Belt are consistent with the fixed mantle plume hypothesis. Given the above evidence, it seems unlikely that the Wells Gray—Clearwater area had its origins in the mantle plume that has created the western portion of the belt. Souther (1986), has suggested that if the McNaughton Lake seismic swarm (Rogers 1981) and Wells Cray—Clearwater area are related to a mantle plume then they may be the result of a second, parallel plume. If this is true then it must be of very recent origin and a track 'per se' has not as yet developed. Nevertheless, the fact remains that young volcanic rocks have erupted in the Wells Cray—Clearwater area and, though they may not be the result of a mantle plume in the same sense as the western Anahim Volcanic belt, mantle inhomogeneities and/or upwelling have been sufficient to produce basaltic melts. The presence of these melts, along with geophysical evidence, indicates that anomalous upper mantle (Rogers 1981; Gough 1986) exists in this region. A possible alternative hypothesis for volcanism in the region is mantle upwelling resulting from continental rifting or rifting that may be the reflection of a broader plume marked by major central volcanoes. 169 8.4 THE CASE FOR CONTINENTAL RIFTING Gabrielse (1985) has proposed that a series of faults in the north-central Cordillera form an anastamosing network between the northern Rocky Mountain Trench and the lntermontane belt (Fig. 7-1). He has documented cumulative dextral transcurrent displacements on these faults of at least 750 km. This movement appears to have occurred from mid-Mesozoic to Cenozoic time, particularly during the late Eocene to Oligocene. The mid-Cenozoic movement occurred contemporaneously with high heat flow, plutonism, volcanism, intrusion of lamprophyre dykes, sedimentation in grabens, and rapid uplift of northwesterly trending elongate ranges. Preservation of sediments and volcanics in the grabens and half grabens along the lineaments (Figs. 8-2 and 2-3) suggests normal faulting, related to extension, may have been contemporaneous and associated with dextral transcurrent faulting (Gabrielse 1985; R.L. Brown et al. in preparation). Gabrielse (1985) also noted that faults with Cenozoic displacement have marked topographic expression. Evidence of transcurrent faulting in south-central British Columbia is not as well documented, but continental rifting during the Eocene has been established by the work of Ewing (1980, 1981), Monger (1985), Mathews and Rouse (1984) and others. During Eocene time, numerous fault-bounded grabens and half grabens formed associated with basaltic and silicic volcanism. Ewing postulates that the volcanism resulted from subduction along the southwestern margin of the Cordillera (Ewing 1980, 1981), over which was superposed transcurrent and extensional motion. This combined regime, termed "splinter tectonics" by Ewing, resulted in concentric folding, enclosed basins, and uplift. Ewing (1980) maintains that this system was controlled by the Fraser-Straight Creek Fault system from which a splay network extended eastward to the Okanagan Valley and the Columbia Valley of Washington. Thick sequences of sediments, as well as volcanics, were deposited in the basins 170 Figure 8-2: Trancurrent faults (heavy lines) and Eocene basins (rectangles), in British Columbia after R.L Brown ef al. (in preparation) and Gabrielse 1985. 171 and preserved as downdropped blocks (Mathews and Bustin 1986) (Fig. 8-2). Dynamic 'wrench' fault models have been suggested that link transcurrent and normal faulting (Harding 1974; McKenzie 1978; Wilcox et al. 1973). These models involve ductile spreading of the deep crust accompanied by high heat flow. Volcanism is coincident with the rise in geothermal gradient and thinning of the continental crust. The rise in geothermal gradient may be a direct consequence of the rifting (Posavec et al. 1973, McKenzie 1978, Yorath and Hyndman 1983). 8.4.1 CENOZOIC CRUSTAL MOVEMENT IN THE WELLS CRAY-CLEARWATER AREA> Within the Wells Cray—Clearwater area there is considerable evidence of faulting, but timing and offset of the faults is poorly constrained. Both dextral strike-slip and normal faults have been observed or inferred (Schiarizza 1986; Struik 1984, 1985a, 1985b, 1986a, 1986b; Schiarizza and Preto 1984) (Fig. 2-3). Struik (1986a) has suggested that the transcurrent faults in the area north of Azure Lake may be the continuation of the northern Rocky Mountain Trench system, as shown by Gabrielse (1985), or form an en echelon set joining the Fraser River—Straight Creek Faults to the northern Rocky Mountain Trench system (Price 1984; Price and Carmichael 1986). From the Azure Lake region, Struik (1985b) reports fault zone features such as "flaser grit, phyllonite, distended competent beds and quartz veins in ductile pelitic matrix, extension faults, mineral-filled tension gashes and lineations of elongate trains and inclusions...." (p. 308). In the North Thompson River valley in the vicinity of Clearwater (Fig. 2-3), Schiarizza (1986) interprets the most recent motion on mapped normal faults, to be west-side-down. Further evidence of normal faulting is found to the northwest in the vicinity of Quesnel Lake (Fig. 2-3) (S. Garwin, oral communication 1986; P. Lewis, oral communication 1986) and to the west around Crooked Lake (M.A. Bloodgood, in preparation; Fillipone 1986). The age and amount 172 of displacement on these faults is poorly constrained, as the faults cut predominantly metamorphosed and disrupted stratigraphy in which geochronologic control is lacking. W.H. Mathews (oral communication 1986) reports an area of doming south of Bonaparte Lake which has affected 15 Ma Chilcotin Croup basalts. Downwarping occurs along the Bonaparte River, which results in Chilcotin Croup basalts dipping gently westward (W.H. Mathews, oral communication 1986). An interpretation of the timing of the formation of the Clearwater Depression, and uplift of the peneplain formed in late Eocene to early Miocene (Chapter 2), suggests that both may result from post Miocene crustal movement. Drainage in the area also has strong structural control (Chapter 2). Though reports of Holocene faulting in British Columbia are scarce, Recent fault scarps have been noted cutting glacial deposits in the Pemberton area (W.H. Mathews, oral communication 1986) and north of Bridge River (Eisbacher and Clague 1983). W.H. Mathews (oral communication 1986), reports the presence of a linear feature, interpreted to be a sag basin several metres deep, in drift over late Tertiary lava in the vicinity of the Fraser River Fault Zone. In the southern Rocky Mountain Trench Clague (1974) reported Quaternary fault scarps and Clowes (1980), states that the main control on this portion of the trench is en echelon normal faults, west-side-down. Fault facets have been reported from along the Tintina fault in northern British Columbia. Directly east of the Wells Gray—Clearwater area, a 1918 earthquake of magnitude 6, was located in the Southern Rocky Mountain trench by Rogers and Ellis (1979). 8.4.2 GEOPHYSICAL CHARACTERISTICS OF THE CORDILLERAN LITHOSPHERE Geophysical evidence for thin crust and high heat flow includes magnetic structures, reflection and refraction data, conductance, and direct heat flow measurements. Early work by Caner (1969, 1970), on magnetic fields, indicated that 173 much of the upper crust in the southern Omineca Crystalline belt may be at or above the Curie temperature for magnetic minerals. This finding is consistent with high heat flow and recent crustal uplift. Coles ef al. (1976) suggested that this magnetically quiet zone may also indicate that crystalline basement is not present, or that magnetic minerals and structures have been largely remobilized and their magnetic signatures obscured or lost. Magnetic data throughout the southern Cordillera prompted Caner (1970) to postulate an abnormally thin lithosphere and partial melting in the lower crust. From geomagnetic work carried out near Clearwater, Dragert and Clarke (1977), postulated that a conductivity heterogeneity at 40-50 km depth may be the result of hydration or partial melting. This conclusion is supported by Wickens (1971) whose work on seismic refraction indicated that the low velocity layer approaches the base of the crust over the same area studied by Caner (1970). Stacey (1973), using seismic velocities, states that the upper mantle must have a lower-than-normal density, or must be hot, to reconcile gravity and refraction results. The latest work on geomagnetic data, following similar work by Dragert and Clarke (1977), is by Gough (1986) who suggests that the results indicate a thinned lithosphere under the Cordillera. Some of the most compelling evidence for a thin lithosphere under the Cordillera comes from heat flow data. Work by Jessop ef al. (1976), Jessop ef al. (1984), Davis and Lewis (1984) and Lewis ef al. (1985) established that the southern part of the Intermontane Belt is an extension of the Cordilleran Thermal Anomaly Zone as defined by Blackwell (1969). The southeastern Omineca Crystalline Belt also appears to have elevated heat flux which Judge (1977) has attributed to deep groundwater circulation. Lewis ef al. (1985) concludes that the Intermontane and Omineca belts are part of the same heat flow province and that this province is most similar to the Basin and Range heat flow province. The heat flow seen at the present time could be the result of heating the entire crust of the Intermontane 174 Belt some 50 million years ago. However, if the elevated heat flow was not maintained after this time, considerable subsidence would occur, for which there is no evidence. Rogers (1981) reported a concentration of seismic events in the McNaughton Lake area. The events, seismic swarms, are characteristic of those occurring in areas of volcanism and rifting, and Rogers (1981) interpreted the quakes as possible evidence for the Anahim hot spot, being "activated by the regional uplift due to thermal expansion of the crust in the hot spot region." (p. 826). The initiating force for these earthquakes may have been the elevated thermal gradient (or 'hot spot') and associated crustal extension and thinning. 8.5 THE BASIN AND RANGE ANALOGY The scenario for the northern Cordillera is remarkably similar to that of the Basin and Range Province of the southwestern United States, west of the Rocky Mountains and east of the San Andreas Fault. North-trending transcurrent faults and high angle normal faults have produced pull-apart basins, high heat flow and volcanism (Lachenbruch 1979; Livaccarno 1979; McKenzie and Jackson 1984; Robyn and Hoover 1982; Suneson and Lucchitta 1983; Lachenbruch and Sass 1978). The intensity of Basin and Range volcanic activity, appears greater here than in the northern Cordillera but this is due in part to a lack of glacial erosion which has permitted volcanic cones and flows millions of years old, and fault facets recording hundreds of thousands of years of motion, to be preserved. The Basin and Range, also appears to have undergone a more prolonged and vigorous extension than is apparent in the northern Cordillera where it appears that vigorous extension in the Eocene was followed by a quiescent period (Chapter 2). Late Tertiary rifting in the Canadian Cordillera has often been dispensed with because of the apparent lack of deformation in Chilcotin Group lavas but if a 175 region is bounded on all sides by faults, the faults may served to isolate sections of crust as stable blocks. Displacement is taken up along the edges of each block (Robyn and Hoover 1982) and the blocks tend to rotate in a clockwise fashion (in a right lateral system), and undergo little internal deformation. Such a process has been inferred to have occurred in late Tertiary time for the western Cascade Range and the Klamath Mountains (Magill and Cox 1981; McKee et al. 1983); the Coast Ranges of Washington, Bates et al. (1981); the Sierra Nevada Mountains (Frei 1986; Frei ef al. 1984). The rotation of sections of the Intermontane Belt, as coherent blocks, could be tested by paleomagnetic work on the Chilcotin Croup basalts. 8.6 CONCLUSIONS The Anahim Volcanic Belt may be the trace of fixed a mantle plume if the eastern termination is considered to be represented by the Nazko Cone, but seems inadequate to explain volcanism in the Wells Cray—Clearwater area. Evidence of crustal deformation in the Canadian Cordillera since the Miocene is limited. Crustal extension cannot be documented, but uplift and broad warping is clearly shown. Certainly, volcanism along 'leaky' normal faults, which may be reactivated older transcurrent structures, in a thinned lithosphere, accompanied by an elevated geothermal gradient, is a viable explanation for the seemingly unrelated and scattered remnants of Quaternary volcanics that stretch the length of the Intermontane and Omineca belts in the Canadian Cordillera. More work is required on a number of the isolated remnants of volcanic rocks of unknown age. Their eruptive history, timing and relationship to basement structures must be documented. Detailed work on rates and timing of uplift are necessary, along with paleomagnetic studies, to determine if crustal rotation and/or tilting has occurred. Studies of geomorphic features such as peneplains, and shorelines of glacial lakes, are necessary to document the most recent crustal movements. 9. CONCLUSIONS Believing where we cannot prove. —Alfred, Lord Tennyson An episode of crustal heating and uplift in the Eocene initiated an erosion cycle which ultimately led to a low relief surface (peneplain), upon which Miocene-Pliocene Chilcotin Croup lavas were deposited. Uplift of this surface occurred in post Miocene time with attendant dissection of the landscape by rivers. Rivers flow both south and west in response to regional elevations, which are greatest to the east, adjacent to the Rocky Mountain Trench, and to the north in the Premier Range. The trend of the Clearwater and North Thompson rivers is controlled by late normal faults which form an en echelon pair, overlapping in the vicinity of the Murtle Plateau. The plateau surface remained relatively low in the area of overlap to form the Clearwater Depression during the late Tertiary, probably coincident with uplift and dissection of the peneplain surface. Faulting is at least as recent as the Pleistocene and may be continuing. Uplift may be the result of recent crustal heating but this heating does not appear to be related to the fixed mantle plume that may be generating the volcanoes of the western Anahim Volcanic Belt. The dissection of the peneplain surface and formation of the depression created broad valleys in which lava flows were ponded. Volcanism in the Weils Gray—Clearwater area may have begun as early as three million years ago, but it seems more likely that intense, ongoing volcanism, started some million years later. The lack of K-Ar dates between 1.4 Ma and 0.6 Ma may be due to sampling bias, or to the removal of early flows by erosion. It seems certain however, that the remaining dates represent more or less continuous volcanism for at least the last 600,000 years. These dates support the hypothesis that no one centre fed all the flows but rather, individual eruptions were temporally as well as spatially separated, and of low volume. 176 177 The largest volume of flows are preserved as flat lying valley-filling or plateau-capping remnants. The flows are thin, 3 m or less but built up to a thickness of 160 m in topographic depressions. Much greater thicknesses of volcanic deposits are preserved in the tuyas found in the region. The tuyas formed from subglacial eruptions and record up to three periods of glaciation that are pre-Fraser Glaciation. The tuyas consist of thick successions of pillow breccia, tuff breccia, with intercalated pillows, all capped by one or more subaeriai flows. Chemistry of the Wells Gray—Clearwater area basalts is consistent with variable degrees of high pressure melting of a heterogeneous, and generally depleted source region. Magmas which are the result of the largest degree of partial melting accumulated in crustal reservoirs long enough to re-equilibrate high pressure mineral phases, drop any mantle nodules, and begin to develop low pressure olivine phenocrysts. Transport of the magma to the surface was sufficiently rapid that no other phenocryst phases formed and crustal xenoliths, incorporated into the magma, suffered only partial melting. The magmas which result from the lowest degrees of partial melting, originated in a source region that appears to have been enriched in LIL elements, and other trace and minor elements, such as Ti, relative to the source generating larger volume partial melts. This enrichment may have made it possible for very small volume partial melts to occur, and for the melt to reach the surface. Transport of these magmas to the surface was rapid, preserving the high pressure mineral assemblage, and transporting large (>25 cm diameter), high density, mantle xenoliths. These 'enriched' magmas, are the youngest in the suite, which leads to the conjecture that the intial, high volume (large partial melts) magmas formed first and as the source zone was depleted, later magmas were the result of smaller degrees of partial melting. 178 The isotopic results from the Wells Cray—Clearwater area are consistent with the trace element results which require derivation of the magmas from moderatly depleted mantle material. Variations in the elemental ratios of the magmas is a recent phenomenon, related to variations in the degree of partial melting, and are not reflected in the initial isotopic ratios. Enrichment in radiogenic 8 7 Sr, 1 ' 3 Nd, 2 0 8 Pb and 2 0 7 Pb is the result of contamination of the reservoir with a small volume of radiogenic, probably upper crustal rock. Further work on the phenocryst assemblage may help to better define the nature of the mantle heterogeneities seen in the whole-rock chemistry. The presence of a MORB type source may be tied to subduction of oceanic crust during the accretion of the displaced terranes during the Mesozoic. Work on the isolated remnants of volcanics along the eastern boundary of Slide Mountain and Quesnel Terranes may better define the mantle heterogeneities. Work on these scattered, and seemingly isolated, remnants is also important to document their eruptive history, timing and relationship to basement structure, and to determine their tectonic control. Detailed work on the rates and timing of uplift are necessary, along with paleomagnetic studies, to determine whether crustal rotation and/or tilting has occurred. Studies of geomorphic features, such as peneplains and the shorelines of glacial lakes, are necessary to document the most recent crustal movements. REFERENCES Alberade, F. and Provost, A. 1977. Petrological and geochemical mass-balance equations: An algorithm for least-square fitting and general error analysis. Computers and Ceoscience, 3 , pp. 309-326 Allen, C.C. 1980. Icelandic subglacial volcanism: thermal and physical studies. Journal of Geology, 88, pp. 108-117. Allen, C . C , Jercinovic, M.J., and Allen, J.S.E5. 1982. Subglacial volcanism in north-central British Columbia and Iceland. Journal of Geology 90 , pp. 699-715. Alley, N.F. 1976. Post Pleistocene glaciations in the interior of British Columbia [abstract]. Geological Association of Canada (Cordilleran Section), Program and Abstracts, p. 6. Alley, N.F. 1980. Holocene and latest Pleistocene cirque glaciations in the Shuswap Highland, British Columbia: discussion. Canadian Journal of Earth Sciences, 17, pp. 797-798. Anguita, F., and Hernan, F. 1975. A propagating fracture model versus a hot spot origin for the Canary Islands. Earth and Planetary Science Letters, 27 , pp. 11-19. Armstrong, J.E. 1981. Post-Vashon Wisconsin Glaciation, Fraser Lowland, British Columbia. Geological Survey of Canada Bulletin, 3 2 2 , 34 p. Armstrong, R.L. 1981. Radiogenic isotopes: the case for crustal recycling on a near-steady-state no-continental-growth Earth. Philosophical Transactions of the Royal Society of London, Series A, 3 0 1 , pp. 443-472. Armstrong, R.L. 1982. Cordilleran metamorphic core complexes—from Arizona to southern Canada. Annual Review of Earth and Planetary Sciences, 10 , pp. 129-154. Armstrong, R.L., Leeman, W.P., and Malde H.E. 1975. K-Ar dating, Quaternary and Neogene volcanic rocks of the Snake River Plain, Idaho. American Journal of Science, 2 7 5 , pp. 225-251. Barberi, F., Civetta, L., and Varet, J. 1980. Sr isotopic composition of Afar volcanics and its implication for mantle evolution. Earth and Planetary Science Letters, 50 , pp. 247-259. Basaltic Volcanism Study Project. 1981. Basaltic volcanism on the Terrestrial Planets. Pergamon Press, Inc., New York. 1266 p. Bates, R.G., Beck, J.E., and Burmester, R.F. 1981. Tectonic rotations in the Cascade Range of southern Washington. Geology, 9, pp. 184-189. Bevier, M.L. 1978. Field relations and petrology of the Rainbow Range shield volcano, west-central British Columbia, unpublished M.Sc thesis, University of British Columibia, Vancouver, 100 p. Bevier, M.L. 1983a. Regional stratigraphy and age of Chilcotin Group basalts, 179 180 south-central British Columbia. Canadian Journal of Earth Sciences, 20, pp. 515-524. Bevier, M.L 1983b. Implications of chemical and isotopic compositions for petrogenesis of the Chilcotin Croup basalts, British Columbia. Journal of Petrology, 24, pp. 207-226. Bevier, M.L 1986. Isotopic evidence for depleted mantle beneath southwestern British Columbia [abstract]. Geological Association of Canada, Program with Abstracts, 11, p. 45. Bevier, M.L , Armstrong, R.L, and Souther, J.C. 1979. Miocene peralkaline volcanism in west-central British Columbia—its temporal and plate-tectonics setting. Geology, 7, pp.' 389-392. Beilski-Zyskind, M., Wasserburg, J.C., and Nixon, P.H. 1984. Sm-Nd and Rb-Sr systematics in volcanics and ultramafic xenoliths from Malaita, Solomon Islands, and the nature of the Ontong Java Plateau. Journal of Geophysical Research, 89, pp. 2415-2424. Bishop, S.T. 1985. The petrology of the Flourmill volcanic centre in the Wells Gray—Clearwater area, east-central British Columbia, unpublished B.Sc. thesis, University of British Columbia, Vancouver, B.C., 49 p. Blackwell, D.D. 1969. Heat-flow determinations in the northwestern United States. Journal of Geophysical Research, 74, pp. 992-1007. Bostock, 1948. Physiography of the Canadian Cordillera, with special reference to the area north of the fifty-fifth parallel. Geological Survey of Canada, Memoir 247, 106 p. Bowen, N.L 1928. The evolution of the igneous rocks. Princeton University Press, Princeton, New Jersey, 332 p. Boynton, M.V. 1984. Cosmoschemistry of the rare earth elements: meteorite studies. in Developments in geochemistry 2: Rare earth element geochemistry. Elsevier Pub. Co., New York, 510 p. Brown, R.L. 1986. Metamorphic core complexes in southern British Columbia [abstract]. Geological Association of Canada, Program with Abstracts, 11, p. 49. Brown, R.L., and Read, P.B. 1983. Shuswap Terrane of British Columbia: a Mesozoic "core complex". Geology 11, p. 164-168. Cairnes, C E . 1931. Mineral resources of northern Okanagan Valley, British Columbia. Geological Survey of Canada Summary Reports, pp. 66A-109A. Campbell, R.B. 1963a. Adams Lake, British Columbia. Geological Survey of Canada, Map 48-1963. Campbell, R.B. 1963b. Quesnel Lake (east half), British Columbia. Geological Survey of Canada, Map 1-1963. Campbell, R.B. 1967. Canoe River, British Columbia. Geological Survey of Canada, Map 15-1967 181 Campbell, R.B., and Tipper, H.W. 1971. Geology of Bonaparte Lake map-area, British Columbia. Geological Survey of Canada, Memoir 363. Caner, B. 1969. Long aeromagnetic profiles and crustal structure in western Canada. Earth and Planetary Science Letters, 7, pp. 3-11. Caner, B. 1970. Electrical conductivity structure in western Canada and petrological interpretation. Journal of Geomagnetism and Geoelectricity, 22, pp. 113-129. Carlson, R.W., Lugmair, G.W., and Macdougall, J.D. 1981. Columbia River volcanism: the question of mantle heterogeneity or crustal contamination. Geochimica et Cosmochimica Acta, 45, pp. 2483-2499. Carlson, R.W., Lugmair, G.W., and Macdougall, J.D. 1983. "Columbia River volcanism: the question of mantle heterogeneity or crustal contamination" (reply to a comment by D.J. DePaolo). Geochimica et Cosmochimica Acta, 47, pp. 845-846. Cavell, P.A., and Baadsgaard, H. 1986. Geochronology of the Big Spruce Lake alkaline intrusion. Canadian Journal of Earth Sciences, 23, pp. 1-10. Chase, C.G. and Sprowl, D.R. 1984. Proper motion of hot spots: Pacific Plate. EOS [abstract], 65, p. 1099. Chaplin, C.E., 1981. Isotope geology of the Gloserheia granite pegmatite, south Norway. M.Sc. thesis, University of Alberta, Dept. of Geology, Edmonton, Alta. Church, B.N. 1979. A survey of Cenozoic magnetostratigraphy in south-central British Columbia. Geological Fieldwork 1979, paper 1980-1, Province of B.C. Church, S.E. 1985. Genetic interpretation of lead-isotopic data from the Columbia. River Basalt Group, Oregon, Washington, and Idaho. Geological Society of America Bulletin, 96, pp. 676-690. Church, S.E., and Tatsumoto, M. 1975 Lead isotope relations in ocean ridge basalts from the Juan de Fuca-Gorda Ridge area, northeastern Pacific Ocean, Contributions to Mineralogy and Petrology, 53, pp. 253-279. Clague, J.J. 1974. The St. Eugene Formation and the Development of the Southern Rocky Mountain Trench. Canadian Journal of Earth Sciences, 1 1 , pp. 916-938. Clague, J.J. 1981. Late Quaternary geology and geochronology of British Columbia. Part 2: Summary and Discussion of radiocarbon-dated Quaternary history. Geological Survey of Canada Paper 80-35, 41 p. Clague, J.J. and Frey F.E. 1982. Petrology and trace element geochemistry of the Honolulu volcanics, Ohau: Implications for the oceanic mantle below Hawaii. Journal of Petrology, 23, pp. 447-504. Clowes, R.M. 1980. Compilation of geophysical/geological data near Golden, B.C. Earth Physics Branch, Seismological Service of Canada, Open File Number 80-14. Cohen, R.S., and O'Nions, R.K. 1982. Identification of recycled continental material in 182 the mantle from Sr, Nd, and Pb isotope investigations. Earth and Planetary Science Letters, 61, pp. 7-84. Coles, R.L., Haines, G.V., and Hannaford, W. 1976. Large scale magnetic anomalies over western Canada and the Arctic: a discussion. Canadian Journal of Earth Sciences, 13, pp. 790-802. Coombes D.M. 1985. A reconnaissance survey of Kostal Lake. A.S.A.P. Reference No. 345001, Fisheries Branch, Ministry of Environment, Province of British Columbia, Coombes D.S. and Wilkinson J.F.G. 1969. Lineages and fractionation trends in undersaturated volcanic rocks from the east Otaga volcanic province (New Zealand) and related rocks. Journal of Petrology, 10, pp. 440-501. Cox, K . C , Bell, J.D., and Pankhurst, R.J., 1979. The interpretation of igneous rocks. Allen and Unwin, Boston, Mass. 450 p. Crough, S.T. 1983. Hotspot swells. Annual Review of Earth and Planetary Sciences, 11, pp. 165-193. Dalrymple, C.B., and Lanphere, M.A., 1969. Potassium-argon dating. Published by W.H. Freeman, San Francisco, Calif. 258 p. Davis, E.E. and Lewis, T.J. 1984. Heat flow in a back-arc environment, Intermontane and Omineca Crystalline belts, southern Canadian Cordillera. Canadian Journal of Earth Sciences, 21, pp. 715-726. DePaolo, D.J. 1983. Comment on "Columbia River volcanism: the question of mantle heterogeneity or crustal contamination" by R.W. Carlson, G.W. Lugmair and J.D. Macdougall. Ceochimica et Cosmochimica Acta, 47, pp. 841-844. DePaolo, D.J., and Wasserburg, G.J. 1976a. Inferences about magma sources and mantle structure from variations of 1 4 3 Nd/ 1 4 * Nd. Geophysical Research Letters, 3, pp. 743-746. DePaolo, D.J., and Wasserburg, C J . 1976b. Nd isotopic variations and petrogenetic models. Geophysical Research Letters, 3, pp. 249-252. Doe, B.R., Leeman, W.P., Christiansen, R.L., and Hedge, C E . 1982. Lead and strontium isotopes and related trace elements as genetic tracers in the upper Cenozoic rhyolite-basalt association of the Yellowstone Plateau volcanic field. Journal of Geophysical Research, 87, pp. 4785-4806. Downes, H. 1984. Sr and Nd isotope geochemistry of coexisting alkaline magma series, Cantal, Massif Central, France. Earth and Planetary Science Letters, 69, pp. 321-334. Dragert H., and Clarke, C.K.C. 1977. A detailed investigation of the Canadian Cordillera geomagnetic transition anomaly. Journal of Geophysics. 42, pp. 373-390. Duford, J.M., and Osborn, G.D. 1978. Holocene and latest Pleistocene cirque glaciations in the Shuswap Highland, British Columbia. Canadian Journal of Earth Sciences, 15, pp. 865-873. 183 Duford, J.M., and Osborn, C D . 1980. Holocene and latest Pleistocene cirque glaciations in the Shuswap Highland, British Columbia: Reply. Canadian Journal of Earth Sciences, 17, pp. 799-800. Duncan, R.A. 1984. Age progressive volcanism in the New England Seamounts and the opening of the central Atlantic Ocean. Journal of Geophysical Research, 89 , pp. 9980-9990. Easterbrook, D.J., and Boellstorff, J. 1981. Paleomagnetic chronology of "Nebraskan-Kansan" tills in midwestern U.S.A. International geological correlation program, project 73/1/24, Quaternary glaciations in the northern hemisphere, report no. 6, Aug. 16-25, 1979. Eisbacher, G.H., and Clague, J.J. 1983. Slope stability, southern Coast Mountains and Fraser Lowland. Geological Association of Canada, Cordilleran Section, Field Trip 1983, 46 p. Ewing, T.E. 1980. Paleogene tectonic evolution of the Pacific Northwest. Journal of Geology, 88 , pp. 619-638. Ewing, T.E. 1981. Regional stratigraphy and structural setting of the Kamloops Group, south-central British Columbia. Canadian Journal of Earth Sciences, 18, pp. 1464-1477. Faure, G. 1977. Principles of isotope geology. John Wiley & Sons, New York, NY, 464 p. Fiesinger, D.W. 1975. Petrology of the Quaternary volcanic centers in the Quesnel Highlands and Garibaldi Provincial Park areas, British Columbia, unpublished Ph.D. thesis, University of Calgary, Calgary, Alta. Fiesinger, D.W., and Nicholls, j. 1977. Petrography and petrology of Quaternary volcanic rocks, Quesnel Lake region, east-central B.C. The Geological Association of Canada, Special Paper no. 16., pp. 28-38 . Fillipone, J.A. 1985. Structure and metamorphism of the Boss Mountain area, southwestern Cariboo Mountains, British Columbia, unpublished M.Sc. thesis, University of British Columbia, Vancouver, British Columbia, 130 p. Francis, P.W., Thorpe, R.S., Moorbath, S., Kretzschmar, G.A.,and Hammill, M. 1980. Strontium isotope evidence for crustal contamination of calc-alkaline volcanic rocks from Cerro Galan, northwest Argentina. Earth and Planetary Science Letters, 48 , pp. 257-267. Frei, L.S. 1986. Additional paleomagnetic results from the Sierra Nevada: further constraints on Basin and Range extension and northward displacement in the western United States. Geological Society of America Bulletin, 97 , pp. 840-849. Frei, L.S., Magill, J.R., and Cox, A. 1984. Paleomagnetic results from the central Sierra Nevada: constraints on reconstructions of the western United States, Tectonics, 3, pp. 157-177. Frey, F.A., Green, D.H., and Roy, S.D. 1978. Integrated models of basalt petrogenesis: a study of quartz tholeiites to olivine melilitites from southeastern Australia utilizing geochemical and experimental petrological data. 184 Ceochimica et Cosmochimica Acta, 3 8 , pp. 1023-1059. Fulton, R.J. 1967. Deglaciation studies in Kamloops region, an area of moderate relief, British Columbia. Geological Survey of Canada Bulletin, 1 5 4 , 36 p. Gabrielse, H. 1985. Major dextral transcurrent displacements along the northern Rocky Mountain Trench and related lineaments in north-central British Columbia. Geological Society of America Bulletin, 9 6 , pp. 1-14. Getsinger, J.S. 1985. Geology of the Three Ladies Mountain/Mount Stevenson area, Quesnel Highland, British Columbia, unpublished Ph.D. thesis, University of British Columbia, Vancouver, B.C. 239 p. Gibbons, A.B., Megeath, J.D., and Pierce, K.L. 1984. Probability of moraine survival in a succession of glacial advances. Geology, 1 2 , pp. 327-330. Gough, D.I. 1986. Mantle upflow tectonics in the Canadian Cordillera. Journal of Geophysical Research, 9 1 , pp. 1909-1919. Green, D.H., and Ringwood, A.E., 1967. The genesis of basaltic magmas. Contributions to Mineralogy and Petrology, 1 5 , pp. 103-190. Harding, T.P. 1974. Petroleum traps associated with wrench faults. Geological Association of Petroleum Geologist Bulletin, 5 8 , pp. 1290-1304. Harland, W.B., Cox A.V., Llewellyn, P.C., Pickton, C.A.C., Smith, A.C., and Walters, R. 1982. A geologic time scale. Cambridge University Press, Cambridge, Great Britain, 131 p. Hart, W.K. 1985. Chemical and isotopic evidence for mixing between depleted and enriched mantle, northwestern U.S.A. Ceochimica et Cosmochimica Acta, 49, pp. 131-144. Haskin, L.A. 1984. Petrogenetic modelling—use of rare earth elements, in Developments in geochemistry 2: rare earth element geochemistry, P. Henderson, ed. Elsevier, pp. 115-152. Hickson, C.J., and Juras, S.J. 1986. Sample contamination by grinding. Canadian Mineralogist. 2 4 , pp. 585-589. Hickson, C.J., and Souther, J.G. 1984. Late Cenozoic volcanic rocks of the Clearwater-Wells Gray area, British Columbia. Canadian Journal of Earth Sciences, 2 1 , pp. 267-277. Holland, 1964. Landforms of British Columbia: a physiographic outline. British Columbia Department of Mines and Petroleum Resources Bulletin, 4 8 , pp. 45-55. Irving, T.N., and Baragar, W.R.A. 1971. A guide to the chemical classification of the common volcanic rocks. Canadian Journal of Earth Sciences, 8 , pp. 523-548. Ito, K., and Kennedy, C.C. 1967. Melting and phase relations in a natural peridotite to 40 kilobars. American Journal of Science, 2 6 5 , pp. 519-538. Ito, K., and Kennedy, G.C., 1968. Melting and phase relations in plane tholeiite-lhezolite-nepheline basanite to 40 kilobars with geological implications. 185 Address to Institute of Geophysics and Planetary Physics, U.C.L.A. Jacobsen, S.B., and Wasserburg, G.J. 1980. Sm-Nd isotopic evolution of chondrites. Earth and Planetary Science Letters, 5 0 , pp. 139-155. Jacobsen, S.B., Quick, J.E., and Wasserburg, G.J. 1986. A Nd and Sr isotopic study of the Trinity peridotite; implications for mantle evolution. Earth and Planetary Science Letters, 6 8 , pp. 361-378. James, D.E. 1982. A combined O, Sr, Nd and Pb isotopic and trace element study of crustal contamination in central Andean lavas. I. Local geochemical variations. Earth and Planetary Science Letters, 5 7 , pp. 47-62. Jessop, A.M., Hobart, M.A., and Scalter, J.G., 1976 The world heat data collection-1976. Earth Physics Branch, Energy Mines, and Resources Canada, Ottawa, Geothermal Series, No. 5 , 125 p. Jessop, A.M., Souther, J.G., Lewis, T.J., and Judge, A.S. 1984. Geothermal measurements in northern British Columbia and southern Yukon Territory. Canadian Journal of Earth Sciences, 2 1 , pp. 599-608. Judge, A. 1977. Terrestrial heat flow and the thermal structure of the eastern Canadian Cordillera [abstract]. Geological Association of Canada, Program with Abstracts, 2 , p. 28. Juras, S.J., Hickson, C.J., Horsky, S.J., Godwin, G.I., and Mathews, W.H. (in press). A practical method for the analysis of rare earth elements in geological samples by graphite furnace atomic absorbtion and x-ray fluorescence. Chemical Geology. Kerr, R.A. 1984. Developing a big picture of earth's mantle. Science, 2 2 5 , pp. 702-703. Kushiro, I. 1968. Compositions of magmas formed by partial zone melting of the earth's upper mantle. Journal of Geophysical Research, 7 3 , pp. 611-633. Kushiro, I. 1972. Origin of some magmas in oceanic and circum-oceanic region. Tectonophysics. 17 (3 ) , pp. 211-222. Lachenbruch, A.H. 1979. Heat flow in the Basin and Range Province and thermal effects of tectonic extension. Pure and Applied Geophysics, 1 1 7 , pp. 34-51. Lachenbruch, A.H., and Sass, J.H. 1978. Models of an extending lithosphere and heat flow in the Basin and Range Province. Geological Society of America, Memoir 152, pp. 209-250. Le Bas, M.J., Le Maitre, R.W., Streckeisen, A. and Zanettin, B. 1986. Chemical classification of volcanic rocks. Journal of Petrology, 2 7 , pp. 746-750. Leeman, W.P. 1975. Radiogenic tracers applied to basalt genesis in the Snake River Plain-Yellowstone National Park region—evidence for a 2.7-b.y.-old upper mantle keel. Geological Society of America, Abstracts with Programs, 7 , p. 1165. Lewis, T.J., Jessop, A.M., and Judge, A.S. 1985. Heat flux measurements in southwestern British Columbia: the thermal consequences of plate tectonics. 186 Canadian journal of Earth Sciences, 22, pp. 1262-1273. Livaccarno, R.F. 1979. Late Cenozoic tectonic evolution of the western United States. Geology, 7, pp. 72-75. McKee, E.H., Duffield, W.A., and Stern, R.J. 1983. Late Miocene and early Pliocene basaltic rocks and their implications for crustal structure, northeastern California and south-central Oregon. Geological Society of America Bulletin, 94, pp. 292-304. McKenzie, D. 1978. Some remarks on the development of sedimentary basins. Earth and Planetary Science Letters, 40, pp. 25-32. McKenzie, D. 1985. The extraction of magma from the crust and mantle. Earth and Planetary Science Letters, 74, pp. 81-91. McKenzie, D., and Jackson, J. 1984. The geometric and thermal consequences of the extension of continental lithosphere [abstract]. The Geological Society of America, Abstracts with Programs, 16, p. 589. McKenzie, D., and Richter, F. 1976. Convection currents in the earth's mantle. Scientific American, 235, pp. 72-89. MacDonald, G.A. 1968. Composition and origin of Hawaiian lavas. Geological Society American Memorandum, 116, pp. 477-522. MacDonald, A., and Katsura, T. 1964. Chemical composition of Hawaiian lavas. Journal of Petrology, 25, pp. 653-682. Magill, J., and Cox, A. 1981. Post-Oligocene tectonic rotation of the Oregon Western Cascade Range and the Klamath Mountains. Geology, 9, pp. 127-131. Marsh, B.D., and Marsh J.G. 1976. On global gravity anomalies and two-scale mantle convection. Journal of Geophysical Research, 81, pp. 5257-5280. Mathews, W.H. 1947. "Tuyas," flat-topped volcanoes in northern British Columbia. American Journal of Science, 245, pp. 560-570. Mathews, W.H. 1981. Early Cenozoic resetting of potassium-argon dates and geothermal history of North Okanagan area, British Columbia. Canadian Journal of Earth Sciences, 18, pp. 1310-1319. Mathews, W.H. (in press). Physiographic map of the Canadian Cordillera. Geological Survey of Canada, Map 1701 A. Mathews, W.H., and Bustin, R.M. 1986. Vitrinite reflectances from Eocene rocks of southern British Columbia, a regional reconnaissance. Canadian Journal of Earth Sciences, 23, pp. 259-261. Mathews, W.H., and Rouse, G.E. (in press). An Early Pleistocene proglacial succession in south-central British Columbia. Canadian Journal of Earth Sciences, 23. Mathews, W.H., and Rouse, G.E. 1984. The Gang Range—Big Bar area, south-central British Columbia: stratigraphy, geochronology, and palynology of the Tertiary beds and their relationship to the Fraser Fault. Canadian Journal of Earth 187 Sciences, 21, pp. 1132-1144. Maxwell, R.E. (project coordinator) 1985. Wells Cray Biophysical—South half. Thematic Mapping Unit, Surveys and Resource Mapping Branch, Ministry of Environment, Victoria, British Columbia. Metcalfe P., and Smith A. 1986. Apatite control on Sm and Nd in Quaternary basalts of Wells Grey Provincial Park, British Columbia, [abstract] Geological Association of Canada, Annual meeting, May 1986, Abstracts with Programs, 11, pp. 101. Min S., 1985. Sr isotopic study of ultramafic nodules from neogene alkaline lavas of B.C., Canada, and Josephine Peridotite, southwestern Oregon, U.S.A. unpublished M.Sc. thesis, University of British Columbia, Vancouver, B.C. Minster, J.B., Jordan, T.H., Molnar, P., and Haines, E. 1973. Numerical modelling of instantaneous plate tectonics. Geophysical Journal of the Royal Astronomical Society, 35, pp. 351-352. Minster, J.B., Jordan, T.H., Molnar, P., and Haines, E. 1974. Numerical modelling of instantaneous plate tectonics. Geophysical Journal of the Royal Astronomical Society, 36, pp. 541-576. Minster, J.B., and Jordan, T H . 1978. Present-day plate motions. Journal of Geophysical Research, 83, pp. 5331-5354. Monger, J.W.H. 1985. Structural evolution of the southwestern Intermontane Belt, Ashcroft and Hope map areas, British Columbia, In Current research, Part A, Geological Survey of Canada, Paper 85-1A, pp. 349-358. Monger, J.W.H., and Berg, H.C. 1984. Part B, Lithotectonic terrane maps of the North America Cordillera, Lithotectonic terrane map of Western Canada and Southeastern Alaska. Miscellaneous Field Studies, United States Geological Survey. Monger, J.W.H., Price, R.A., and Tempeiman-Kluit, D.J. 1982. Tectonic accretion and the origin of the two major metamorphic and plutonic welts in the Canadian Cordillera. Geology, 10, pp. 70-75. Montgomery, J.R. 1985. Structural relations of the southern Quesnel Lake Gneiss, Isosceles Mountain area, southwest Cariboo Mountains, British Columbia, unpublished M.Sc. thesis, University of British Columbia, Vancouver, B.C. 96 P-Moorbath S. 1978. Age and isotope evidence for the evolution of continental crust. Philosophical Transactions of the Royal Society of London, 288, pp. 401-413. Morse S.A. 1980. Basalts and phase diagrams. Springer-Verlag, New York, NY. 493 p. Murphy, D.C 1985. Stratigraphy and structure of the east-central Cariboo Mountains, British Columbia and implications for the geological evolution of the southeastern Canadian Cordillera, unpublished Ph.D. thesis, Carlton University, Ontario. 187 p. Murphy, D.C. and Rees C.J. 1983. Structural transition and stratigraphy in the Cariboo Mountains, British Columbia. In Current research, Part A, Geological 188 Survey of Canada Paper, 83-1A. pp. 245-252. Okulitch, A.V. 1984. The role of the Shuswap Metamorphic Complex in Cordilleran tectonism: a review. Canadian Journal of Earth Sciences, 21, pp. 1171-1193. O'Nions, R.K. 1984. Isotopic abundances relevant to the identification of magma sources. Philosophical Transactions of the Royal Society of London, Series A, 310, pp. 591-603. O'Nions, R.K., Evensen, N.M., Hamilton, P.J., and Carter, S.R. 1978. Melting of the mantle past and present: isotope and trace element evidence. Philosophical Transactions of the Royal Society of London, Series A, 258, pp. 547-559. O'Nions, R.K., Carter, S.R., Evensen, N.M., and Hamilton, P.J. 1979. Geochemical and cosmochemical applications of Nd isotope analysis. Annual Review of Earth and Planetary Sciences. 7, pp. 11-38. Parkinson, D.L. 1985. U-Pb geochronometry and regional geology of the southern Okanagan valley, British Columbia: the western boundary of a metamorphic core complex, unpublished M.Sc. thesis, University of British Columbia, Vancouver. 149 p. Parrish, R.R. 1985. Metamorphic core complexes of southern British Columbia: distinctions between extensional or compressional origins. Geological Association of America, Cordilleran Section Annual Meeting, Program with Abstracts, 17, p. 399. Parrish, R.R., Carr, S. and Parkinson D.L. 1985. Metamorphic complexes and extensional tectonic, Southern Shuswap Complex, southeastern British Columbia. In Field Guides to Geology and Mineral Deposits in Southern Cordillera, Geological Society of America, Cordilleran Section Meeting, May 1985. pp. 12-1-12-15. Pearce, T.H 1968. A contribution to the theory of variation diagrams. Contributions to Mineralogy and Petrology, 19, pp. 142-157. Pell J. 1984. Stratigraphy, stucture and metamorphism of Hadrynian strata in the southeastern Cariboo Mountians, British Columbia, unpublished Ph.D. thesis, University of Calgary, Alberta. 185 p. Pigage, L.C. 1978. Metamorphism and deformation on the northeast margin of the Shuswap Metamorphic Complex, Azure Lake, British Columbia, unpublished Ph.D. thesis, University of British Columbia, Vancouver, B.C. 289 p. Posavec, M., Taylor, D., Leeuwen, T. van and Spector, A. 1973. Tectonic controls of volcanism and complex movements along the Sumatran fault. system. Geological Society of Malaysia Bulletin, 6, pp. 43-60. Preto, V.A. 1974. Geology of the Allision Lake-Missezula Lake area. Province of B.C., Department of Mines and Petroleum Resources. Map Number 17. Price, R.A. 1984. Cordilleran tectonic accretion-the North American connection as illustrated in continental margin transect B-2 [abstract]. The Geological Society of America, Program with Abstracts, 16, p. 628. Price, R.A., and Carmichael, D.M. 1986. Geometric test for Late Cretaceous-Paleogene 189 intracontinental transform faulting in the Canadian Cordillera. Geology, 14, pp. 468-471. Price, R.A., Monger, J.W.H., and Roddick, J.A. 1985. Cordilleran cross-section; Calgary to Vancouver. Geological Society of America, Cordilleran Section Annual Meeting, May 1985, Field Trip Guide Book, pp. 3-1 to 3-49. Read, P.B., and Brown, R.L. 1981. Columbia River fault zone: southeastern margin of the Shuswap and Monashee complexes, southern British Columbia. Canadian Journal of Earth Sciences, 18, pp. 1127-1145. Richter, F.M., and Parsons, B. 1975. On the interaction of two scales of convection in the mantle. Journal of Geophysical Research, 80, pp. 2529-2541. Ringwood, A.E. 1975. Composition and petrology of earth's mantle. McGraw-Hill, 618 P-Rittmann A. 1962. Volcanoes and their activities. Wiley-interscience, NY, 305 p. Robyn, T.L., and Hoover, J.D. 1982. Late Cenozoic deformation and volcanism in the Blue Mountains of central Oregon: microplate interactions. Geology, 10, pp. 572-576. Rogers, G.C. 1981. McNaughton Lake seismicity— more evidence for an Anahim hotspot. Canadian Journal of Earth Sciences, 18, pp. 826-828. Rogers, G.C. and Ellis, R.M. 1979. The eastern British Columbia earthquake of February 4, 1918. Canadian Journal of Earth Sciences, 16, pp. 1484-1493. Ross, J. V. 1974. A Tertiary thermal event in south-central, British Columbia. Canadian Journal of Earth Sciences, 11, pp. 1116-1122 Ross, J.V. 1983. The nature and rheology of the Cordilleran upper mantle of British Columbia: inferences from peridotite xenoliths. Tectonophysics, 100, pp. 321-357. Russell, J.K. 1984. Petrology of diamond craters, S.E. Oregon, unpublished Ph.D. thesis, University of Calgary, Calgary, Alta., 156 p. Russell J.K. and Nicholls J. 1987 Early crystallization history of alkali olivine basalts, Diamond Craters, Oregon. Geochimica et Cosmochimica Acta, 51, pp. 143-154. Schiarizza, P. 1986. Geology of the Vavenby area, NTS 82M/5,11,12. Ministry of Energy, Mines and Petroleum Resources, Province of British Columbia, Open File Map 1986/5. Schiarizza, P.A., and Preto, V.A. 1984. Geology of the Adams Plateau—Clearwater area. Ministry of Energy, Mines and Petroleum Resources, Province of British Columbia, Map Number 56. Schmid, R. 1981. Descriptive nomenclature and classification of pyroclastic deposits and fragments: recommendations of the IUCS subcommission on the systematics of igneous rocks. Geology, 9, pp. 41-43. Schwarzer, R.R. and Rogers, J.J.W. 1974. A worldwide comparison of alkali olivine 190 basalts and their differentiaton trends. Earth Planetary Science Letters, 23, pp. 286-296. Selwyn R.C. 1871-1872. Journal and report of preliminary explorations in British Columbia. Geological Survey of Canada, Report of Progress, pp. 16-72. Shackleton, N.J., and Opdyke, N.D. 1976. Oxygen-isotope and paleomagnetic stratigraphy of Pacific Core V28-29, late Pliocene to latest Pleistocene. Geological Society of America Memoir, 145, pp. 449-464. Sigvaldason G.E. 1968. Structure and products of subaquatic volcanoes in Iceland. Contributions to Mineralogy and Petrology, 18, pp. 1-16. Sleep, N.H., 1974. Segregation of magma from a mostly crystalline mush. Geological Society of America Bulletin, 85, pp. 1225-1232. Smith, A., 1986. Isotopic and geochemical studies of Terrane I, south-central British Columbia, unpublished Ph.D. thesis, University of Alberta, Edmonton, Alta., 113 p. Snyder, G.E.,and Fraser, G.D., 1963a. Pillowed lavas, I: Intrusive layered lava pods and pillowed lavas, Unalaska Island, Alaska. Geological Survey Professional Paper 454-B,C, U.S.A. government printing office. Snyder, C.E.,and Fraser, G.D., 1963b. Pillowed lavas, II: a review of selected recent literature. Geological Survey Professional Paper 454-B,C, U.S.A. government printing office. Souther, J.C. 1977. Volcanism and tectonic environments in the Canadian Cordillera—a second look. The Geological Association of Canada Special Paper 16, 24 p. Souther, J.C. 1986. The western Anahim Belt: root zone of a peralkaline magma system. Canadian Journal of Earth Sciences, 23, pp. 895-908. Souther, J.G., and Hickson, C.J. 1984. Crystal fractionation of the basalt comendite series of the Mount Edziza Volcanic Complex, British Columbia: major and trace elements. Journal of Volcanology and Geothermal Research, 21, pp. 79-106. Souther, J .C, Armstrong, R.L., and Harakal J. 1984. Chronology of the peralkaline, late Cenozoic Mount Edziza Volcanic Complex, northern British Columbia,-Canada. Geological Society of American Bulletin, 95, pp. 337-349. Stacey, R.A. 1973. Gravity anomalies, crustal structure and plate tectonics in the Canadian Cordillera. Canadian Journal of Earth Sciences, 10, pp 615-628. Stacey, R.A. 1974. Plate tectonics, volcanism and the lithosphere in British Columbia. Nature, 250, pp. 133-134. Struik, L .C 1984. Stratigraphy of Quesnel Terrane near Dragon Lake, Quesnel map area, central British Columbia. In Current research, Part A, Geological Survey of Canada, Paper 84-1 A, pp. 113-116. Struik, L.C. 1985a. Pre-Cretaceous terranes and their thrust and strike-slip contacts, Prince George (east half) and McBride (west half) map areas, British 191 Columbia. In Current research, Part A, Geological Survey of Canada, Paper 85-1A, pp. 267-272. Struik, L C . 1985b. Dextral strike-slip through Wells Gray Provincial Park, British Columbia. In Current research, Part A, Geological Survey of Canada, Paper 85-1A, pp. 305-309. Struik, L C . 1986a. A regional east-dipping thrust places Hadrynian onto probable Paleozoic rocks in Cariboo Mountains, British Columbia. In Current research, Part A, Section 2, Geological Survey of Canada, Paper 86-1A, pp. 589-594. Struik, L C , 1986b. Imbricated terranes of the Cariboo gold belt with correlations and implications for tectonics in southeastern British Columbia. Canadian Journal of Earth Sciences. 23, pp. 1047-1061. Suneson, N.H., and Lucchitta, I. 1983. Origin of bimodal volcanism, southern Basin and Range Province, west-central Arizona. Geological Society of America Bulletin, 94, pp. 1005-1019. Swanson, D.A., and Wright, T.L 1978. Bedrock geology of the Northern Columbia Plateau and adjacent areas. In The Channeled Scabland (V.R. Baker and D. Nummedal, eds.). NASA, p. 35-37. Takahashi, E. and Kushiro, I. 1983. Melting of a dry peridotite at high pressures and basalt magma genesis. American Mineralogist, 68, pp. 859-879. Templeman-Kluit, D. 1984. Meteoric water model for gold veins in a detached terrane. The Geological Society of America, Program with Abstracts, 16, p. 674. Thiessen, R., Burke, K., and Kidd, W.S.F. 1979. African hotspots and their relation to the underlying mantle. Geology, 7, pp. 263-266. Thorton, C P . and Tuttle O.F. 1960. Chemisty of igneous rocks. I. Differentiation Index. American Journal of Science, 258, pp. 664-684. Todt, W., Cliff, R.A., Hansen, A. and Hofmann A.W. 1984. 2 0 2 Pb + 2 0 5 Pb double spike for lead isotopic analyses. Terra Cognita, 4, 209 p. Uglow, W.L 1921. Geology of the North Thompson Valley map-area, British Columbia. Geological Survey of Canada Summary Reports, pp. 72A-121A. Van Bremmelen, R.W., and Rutten, M.G. 1955. Tablemountains of northern Iceland. Brill, Leiden. 217 p. Vink, C.E., Morgan, W.J., and Vogt, P.R. 1985. The earth's hot spots. Scientific American, 252, pp. 50-57. Walker, J.F. 1930. Clearwater River and Foghorn Creek map-area, Kamloops District, British Columbia. Geological Survey of Canada, Summary Reports, Part A, pp. 125A-153A. Watson, K. DeP., and Mathews, W.H. 1944. The Tuya-Teslin area, northern British Columbia. British Columbia Department of Mines Bulletin, 19, 52 p. White, W.M., and Hofmann, A.W. 1982. Sr and Nd isotope geochemistry of oceanic 192 basalts and mantle evolution. Nature, 296, pp. 821-825. Wickens, A.J. 1971. Variations in lithospheric thickness in Canada. Canadian Journal of Earth Sciences, 8, pp. 1154-1162. Wilcox, R.E., Harding, T.P. and Seely, D.R. 1973. Basic wrench tectonics. American Association of Petroleum Geologists Bulletin, 57, pp 74-96. Williams, H., Turner, F.J., Gilbert, C M . , 1982. Petrography: an introduction to the study of rocks in thin sections. W.H. Freeman and Co., San Fransisco. 625 P-Wilson, A.D. 1955. A new method for the determination of ferrous iron in rocks and minerals. Bulletin of the Geological Survey of Great Britain, 9, pp. 56-58. Wilson, J.T. 1973. Mantle plumes and plate motions. Tectonophysics, 19, pp. 149-164. Worner, C , Staudigel, H., and Zindler, A. 1985. Isotopic constraints on open system evolution of the Laacher See magma chamber (Eifel, West Germany). Earth and Planetary Science Letters, 75, pp. 37-49 Yoder, H.S. Jr. 1976. Generation of Basaltic Magma. National Academy of Sciences, Washington, D . C , 265 p. Yoder, H.S. Jr. and Tilley, C E . 1962. Origin of basalt magmas: an experimental study of natural and synthetic rock systems. Journal of Petrology, 3 , pp. 342-532. Yorath, C.J., and Hyndman, R.D. 1983. Subsidence and thermal history of Queen Charlotte Basin. Canadian Journal of Earth Sciences, 20, pp. 135-159. Zindler, A., Staudigel, H., and Batiza, R. 1984. Isotope and trace element geochemistry of young Pacific seamounts: implications for the scale of upper mantle heterogeneity. Earth and Planetary Science Letters, 70, pp. 175-195. APPENDIX A: SAMPLE PREPARATION Roses have thorns, and silver fountains mud; Clouds and eclipses stain both moon and sun, And loathsome canker lives in sweetest bud. All men make faults. —William Shakespeare SAMPLE SELECTION The objective of this study is to characterize a suite of basic volcanic rocks in terms of chemical differences among volcanic rocks within the suite, possible mantle source areas, possible fractionation paths of the magma, and contamination of the magma by crustal components. The analytically determined composition of the rock must be as close as possible to that found in the original magma; therefore, all rocks chosen for analysis must be as fresh as possible. Any rocks showing signs of alteration were eliminated from the samples selected for chemical analysis. Alteration was considered in the form of oxidation of the iron compounds, sericitization of the silicate minerals or removal of elements through groundwater or hydrothermal systems developed within the flow rocks. The interaction of solutions with the surrounding rock may introduce new elements or leach existing elements and/or produce conditions suitable for the growth of new mineral species such as zeolites, clays or carbonates. The original chemical balance of the rock may therefore undergo substantial change. Thin sections were checked for visible alteration of the minerals, and by major element analysis for unusual amounts or ratios of alkali elements, large loss on ignition (LOI), and in an abnormal F e 2 0 3 / F e O ratio. If the F e 2 0 3 value exceeds FeO the rock is considered to have been oxidized. Fine grained or glassy volcanic rocks are thought to be more representative of the original magmatic composition and may be more homogeneous than rocks with phenocrysts (Russell 1984). As phenocrysts may represent a cumulative phase in the magma (Stout and Nicholls ef al. 1985; Souther and Hickson 1984) a sample 193 194 containing them may not be representative of the bulk magmatic composition. Depending on the magma viscosity, phenocrysts that are on the liquidus at the time of eruption may be dense enough to settle downward within the flow (Mathews ef al. 1964) as well as in the magma chamber, forming a cumulate layer that may not be obvious as such if the exposure is poor. Phenocrysts may also show lateral variation away from the vent, with olivine dropping out of the flow close to the vent and the feldspars carried further. If the magma contains mantle and/or crustal nodules these may also show vertical and horizontal variation within the flow. These effects tend to complicate and mask a chemical trend shown by lavas from different vent areas or erupted at different times. To assess the effect of these processes on the chemical composition of the magma, samples were collected laterally from the vent area of a young cinder cone out to the distal reaches of a flow. Samples (SB0302 to SB410B; Table D-2, Appendix D) were also collected across the flow surface at 0.5 km, 1.5 km, 3 km and 4 km (the distal end of the flow). In a second location with good vertical exposure, samples were collected from the base, middle and top of two cooling units and from the middle of a further three cooling units (42501A to 42502D; Table D-2, Appendix D). The results are discussed in Chapter 6. For all other samples an attempt was made to collect from what was perceived in the field as the centre of the flow or cooling unit. TEST O F S A M P L E C O N T A M I N A T I O N BY G R I N D I N G Once the samples were chosen they were reduced to a powdered form appropriate for the analytical technique chosen. Before proceeding with this step a study of sample contamination was undertaken. Large, difficult-to-detect errors can result in contamination during sample preparation. This study examines the contamination contributed by five types of sample-preparation apparatus commonly used in geological laboratories. 195 G r i n d i n g M e t h o d Table 1 lists the specifications of the five stage grinding sequence used: tungsten carbide and chrome steel shatterboxes, an agate mortar, a corundum-ceramic handmill, and hardened steel disk grinding plates. Each apparatus was further cleaned by (1) grinding a sample of silica sand (Thompson & Bankston 1970), (2) blowing out the powdered silica with compressed air, and (3) scrubbing with a clean dry nylon brush. Then 50 g of Ottawa Sand Standard (Fisher Scientific, > 99% S i 0 2 , 20 to 30 mesh) were ground to less than 200 mesh. Minimum grinding times required to reduce the samples to less than 200 mesh were used (Table 2) as determined by test sieving samples after sequential grinding periods of 10 seconds; these samples were then discarded. Samples used for analysis were not sieved due to the possibility of contamination from the stainless steel or brass sieves that were available (Lavergne 1965; Thompson & Bankston 1970). After grinding, the samples were kept in 18.5 ml polyvinylchloride (PVC) vials (PVC is a potential low level source of Ti, Zn, Na and Cd; Scott and Ure 1972). The powders were then made into three gram pellets for x-ray fluorescence (XRF) analysis. The pellet press uses a tungsten carbide piston and has a stainless steel housing. Polyvinyl alcohol was used as a binder. Two XRF units were used; a Philips PW 1400 with a Rh tube for the major elements and most of the minor and trace elements, and a Philips PW 1410 with a Mo tube for La, Ce, and Nd (Table 3). Tungsten concentrations could not be determined because of the lack of standards with known concentrations; however, the W La ^, 2 peaks were scanned to determine the presence of this element (Fig. A-1). The Ta La! peak was also scanned but no Ta could be detected due to the high detection limit inherent in the XRF method. T A B L E A - 1 . A P P A R A T U S USED FOR SAMPLE P R O C E S S I N G S U R F A C E 1 FORM MANUFACTURER B r a z i l i a n a g a t e : S 1 0 i 99 .91%, m o r t a r a n d p e s t l e ; o u t e r F r i t s h " p u 1 v e r i s e t t e - 2 " , t y p e A I J O I 0 .02% N a . O 0 .02%, F e i O i d i a m e t e r 18 cm, i n n e r d i a m e t e r 0 2 . 0 0 1 .01%. K , 0 0 .01%, MnO 0 . 0 1 % , C a O 1 3 . 5 c m . p e s t l e 7 cm. 0 .01%. MgO 0.01% C o r u n d u m - c e r a m i c ' : a A l ; O i w i t h h a n d m i l l ( b u c k b o a r d ) ; 30 cm u n k n o w n p o s s i b l e t r a c e a m o u n t s o f K . s q u a r e p l a t e a n d manua l N a , S i . C a , C u , F e , M g , P b , B , a n v i l - s h a p e d r o c k e r . C r , L i , M n , a n d N i T u n g s t e n C a r b i d e : W 88%. C 6%. Co 6%, s h a t t e r b o x / r i n g g r i n d e r ; o u t e r d i a m e t e r 17 cm X 8 cm. p u c k a n d 1 r i n g Spex I n d u s t r i e s s h a t t e r b o x N o . 8 5 0 0 C h r o m e s t e e l : H i g h C r , h i g h C s t e e l t y p e A 1 S 1 . 0 3 ; C 1.93%, C r 13.21%. Cu 0.03%. Mn 0.46%, Mo 0.02%, N i 0.08%, P 0 .019%. S i 0 .38%. S 0 .005%. W 0 .01%, o t h e r m e t a l s <0.01%. s h a t t e r b o x / r 1 n g g r i n d e r ; o u t e r d i a m e t e r 2 1 . 5 cm X 6 c m , p u c k a n d two r i n g s . Rock l a b r i n g g r i n d e r #137 H i g h C a r b o n S t e e l : S p e c i a l i r o n a l l o y , c l a s s 3 0 / 3 5 ; S i 2.30%. F e 93 .4%, Mn 0.70%. P 0.12%, C 3.45% d i s c g r i n d e r ; 20 cm d i a m e t e r w i t h o n e s t a t i o n a r y a n d one r o t a t i n g d i s c . B t c o - B r a u n UA P u l v e r i z e r e q u i p p e d w i t h UA 5 1 a n d UA 52 s t a n d a r d g r 1 n d i n g p i a t e s . ' ) T h e c o m p o s i t i o n s o f t h e s u r f a c e s a r e t h o s e r e p o r t e d b y t h e m a n u f a c t u r e r , e x c e p t a s n o t e d . ' ) T h e e x a c t c o m p o s i t i o n o f t h i s a p p a r a t u s i s n o t k n o w n . 197 TABLE A-2. SAMPLE PREPARATION S A M P L E A P P A R A T U S TIME T O <200 m e s h OS-Ag agate mortar 10 min. OS-MUL corundum-ceramic handmill 15 min. OS-W2 tungsten carbide shatterbox 120 s. OS-CR chrome steel shatterbox 45 s. OS-DSK steel-disk mill two passes with touching plates 1000 . Figure A-1: XRF peak scan between 42° and 46° 28 for W and Ta La peaks. ACV-1 and BCR-1 are USGS standard rock samples with 0.55 ppm and 0.4 ppm respectively. Sample OS-Ag was prepared using an agate mortar and OS-W2 prepared using a tungsten carbide shatterbox. Analytical conditions for the Philips PW 1410: 40 mA, 60 kV and LiF200 analyzing crystal. 199 Five rare earth elements were analyzed by using a graphite furnace, flameless atomic absorption (AA) (Table 3) after dissolution of the rock powder in hydrofluoric and perchloric acids (Horsky and Fletcher 1981; Juras ef al. in preparation). Results Analytical results are given for each of the grinding techniques (Table 3). The results are semi-quantitative and, because of the highly abrasive properties of the Ottawa Sand, reflect maximum contamination from the grinding apparatus. The relative values for W are shown in Figure 1. St. Louis (1984), in a study using neutron activation, found W contamination from a tungsten carbide shatterbox was high enough that counting could not be undertaken for more than five days following irradiation of the sample. Joron ef al. (1980) showed W contamination on the order of 100's of ppm, and Nb and Ta contamination of three to five ppm results from use of the tungsten carbide shatterbox. Nisbet ef al. (1979) also indicate W, Co and Ta contamination from the use of tungsten carbide equipment. Further work at the University of British Columbia by P. J. Michael (oral communication) found consistent contamination of approximately 0.5 ppm Nb after grinding basaltic rock for 50 seconds in a tungsten carbide shatterbox. St. Louis (1984) found Sc contamination at a level of 0.05 ppm when grinding quartz sand in a tungsten carbide shatterbox. Joron ef al. (1980) also analyzed for Sc but found no contamination when rocks of basaltic composition were ground. The results of these studies are combined with our work in Table 4; elements are listed alphabetically with the relative degree of contamination indicated. Conclusions As the number of steps undertaken in sample preparation should be minimized a jaw crusher was used to reduce samples directly to a size suitable for the shatterbox. The disk mill, although rapid, was not used. From this work and that of Thompson & Bankston (1970) corundum, mullite, and other forms of ceramic 200 TABLE A - 3 : ANALYTICAL RESULTS FOR PREPARED SAMPLES S A M P L E (wt.%) OS l -Ag O S - M U L O S - W 2 O S - C R O S - D S K S A N D 1 D E T . LIMIT 1 (wt.%) M A C H