MIAMI UNIVERSITY the Graduate School Certificate for Approving The

MIAMI UNIVERSITY

The Graduate School

Certificate for Approving the Dissertation

We hereby approve the Dissertation

Ninad R. Bondre

Candidate for the Degree:

Doctor of Philosophy

William K. Hart, Director

Craig M. White, Reader

Elisabeth Widom, Reader

Jason A. Rech

Brian S. Currie

Michael W. Crowder, Graduate School Representative ABSTRACT

FIELD AND GEOCHEMICAL INVESTIGATION OF BASALTIC MAGMATISM IN THE WESTERN UNITED STATES AND WESTERN INDIA

by Ninad R. Bondre

This dissertation consists of three sub-projects, each of which integrates field and geochemical information to address specific questions pertaining to magmatism in three different continental basaltic provinces. It is expected that these sub-projects will contribute to a deeper understanding of various facets of basaltic magmatism. The first sub-project involves the Sangamner mafic dike swarm in the western Deccan Volcanic Province (DVP). This study, which is the first study of its kind from this province, uses a combination of field and geochemical characteristics to understand the relationship of the dikes with associated lava flows, and their position in the established Deccan stratigraphy. The results indicate that several dikes are geochemically similar to younger formations, and that NE-SW trending dikes in the central part of the area possibly represent a fissure system that could have fed lava flows of the Poladpur and / or Khandala formation. The second sub-project is the first attempt at integrating morphological, textural and compositional data to understand the emplacement of basaltic lava flows from the Steens Basalt in southeastern Oregon, and its eruptive history. The study reveals substantial morphological diversity within the Steens package. Typical compound pahoehoe flows are abundant – these were probably sourced from small to moderate-sized shield volcanoes, and grew as a network of overlapping, inflated lobes. Interbedded a’a and transitional flows suggest that physical parameters varied throughout the eruptive episodes. Geochemical data, in conjunction with morphological data hints at complex eruptive episodes that probably reflect complex processes at depth. These deserve to be investigated in greater detail. The final sub-project focuses on the newly-recognized Jordan Valley Volcanic Field (JVVF) in southeastern Oregon. An integration of field, geochemical and geochronologic information has been used to understand the evolution of this field through space and time, and to better understand the origin of other monogenetic volcano-fields. Salient results include the recognition of two major pulses of volcanism, long-lived vent alignments, and considerable chemical and isotopic diversity. This information is evaluated in terms of its implications for understanding issues pertaining to small-scale mantle heterogeneity, differentiation processes, and melt transport.

FIELD AND GEOCHEMICAL INVESTIGATION OF BASALTIC MAGMATISM FROM THE WESTERN UNITED STATES AND WESTERN INDIA

A DISSERTATION

Submitted to the Faculty of

Miami University in partial

fulfillment of the requirements

for the degree of

Doctor of Philosophy

Department of Geology

Ninad R. Bondre

Miami University

Oxford, Ohio

2006

Dissertation Director: William K. Hart, Ph.D.

TABLE OF CONTENTS

Chapter1: Introduction 1 References 4

Chapter 2: Geology and Geochemistry of the 5 Sangamner Mafic Dike Swarm, Western Deccan Volcanic Province, India: Implications for Regional Stratigraphy

Abstract 5 Body Text 6 References 23

Chapter 3: Morphological and Textural Diversity of 49 the Steens Basalt Lava Flows, Southeastern Oregon, USA: Implications for Emplacement Style and Nature of Eruptive Episodes

Abstract 49 Body Text 50 References 73

Chapter 4: Geology, Geochronology and Compositional 104 Diversity of The Jordan Valley Volcanic Field (JVVF), Southeastern Oregon: Implications for Small-Volume, Monogenetic Volcanism

Abstract 104 Body Text 105 References 145

Chapter 5: Concluding Remarks and Suggestions for Future Work 205

Appendix 1: Analytical Methods 210

Appendix 2: Summary of Methods Used in Ar-Ar Geochronology 215

Appendix 3: Sample Locations and Descriptions 224

Appendix 4: Geochemical Data for Samples From the JVVF 243

ii LIST OF TABLES

Chapter 2: Geology and Geochemistry of the 5 Sangamner Mafic Dike Swarm, Western Deccan Volcanic Province, India: Implications for Regional Stratigraphy

1 – Stratigraphic classification of the Southwestern Deccan 29 Volcanic Province 2 – Major and Trace Element Data for the Sangamner Samples 30 3 – Rare Earth Element and Isotopic Data for the 33 Sangamner Samples 4 – Results of Discriminant Function Analysis for the 34 Sangamner Samples

Chapter 3: Morphological and Textural Diversity of 49 the Steens Basalt Lava Flows, Southeastern Oregon, USA: Implications for Emplacement Style and Nature of Eruptive Episodes

1 – Location and Approximate Stratigraphic Context of 80 Studied Steens Sections 2 – Major and Trace Element Data for the Steens Samples 81

Chapter 4: Geology, Geochronology and Compositional 104 Diversity of The Jordan Valley Volcanic Field (JVVF), Southeastern Oregon: Implications for Small-Volume, Monogenetic Volcanism

1 – Geochronologic Data for Selected Samples from the JVVF 156 2 – Major, Trace Element (including REE) and Isotopic Data for 157 Selected JVVF Samples 3 – Parameters Used in Fractional Crystallization Modeling Using 160 PELE

Appendix 2: Summary of Methods Used in Ar-Ar Geochronology 215

1 – Age Data for the Dated JVVF Samples 216

Appendix 4: Geochemical Data for Samples From the JVVF 243

iii

LIST OF FIGURES

Chapter 2: Geology and Geochemistry of the Sangamner Mafic 5 Dike Swarm, Western Deccan Volcanic Province, India: Implications for Regional Stratigraphy

1 – DVP map with study area, principal flow types and principal dike 35 swarms

2 – Geological Map of the Sangamner Area 37

3 – Binary Plots involving elements and elemental ratios 39

4 – Results of Discriminant Function Analysis 41

5 – Primitive Mantle-Normalized Multi-element Patterns 43

6 – Chondrite-Normalized REE patterns 45

7 – Sr Isotopic Composition versus εNd 47

Chapter 3: Morphological and Textural Diversity of 49 the Steens Basalt Lava Flows, Southeastern Oregon, USA: Implications for Emplacement Style and Nature of Eruptive Episodes

1 – Steens Location Map 84

2 – Binary Plots that Help Determine Stratigraphic Context 86

3 – Measured Sections 88

4 – Field Photographs 90

5 – Field Sketches of Compound Pahoehoe Flows 92

6 – Field Sketches and Photograph of Pahoehoe Flow Lobes 94

7 – Photomicrographs Showing Textural Characteristics 96

8 – Field Sketches of A’a and Transitional Flows 98

iv 9 – Field Photographs of A’a Flows 100

10 – Binary Plots that Link Morphology and Texture with Composition 102

Chapter 4: Geology, Geochronology and Compositional 104 Diversity of The Jordan Valley Volcanic Field (JVVF), Southeastern Oregon: Implications for Small-Volume, Monogenetic Volcanism

1 – Regional Context of the JVVF 161

2 – Satellite Image of the JVVF with Locations of Vents and Vent 163 Alignments

3 – Simplified Geological Map of the JVVF 165

4 – Total Alkalis vs. Silica Diagram and Plot of Zr/TiO2 vs. Nb/Y 167

5 – Binary Plots of SiO2, TiO2, Al2O3 and Total Alkalis vs. Mg number 169

6 – MgO*0.5-K2O*5-TiO2*1.5 Triangular Diagram Showing Principal 171 Magma Types

7 – Binary Plots of Sr, Nb, Rb and Zr vs. TiO2 Revealing Two 173 Principal Groups

8 – Chondrite-Normalized REE diagram 175

9 – Sr vs. Nd Isotopic Compositions 177

10 – Pb Isotopic Compositions 179

11 – Sr and Nd Isotopic Compositions vs. Pb Isotopic Composition 181

12 – Satellite Image Showing the Spatial and Temporal Context of 183 Geochemical Data

13 – Binary Plots Showing Salient Characteristics of Groups 1 and 2 185

14 – Nd Isotopic Composition and Zr vs. Age 187

15 – Results of Batch Partial Melting Modeling of REE for HAOT 189 Generation

16 – La/Sm(N) and Sm/Yb(N) vs. TiO2 Indicating Depth of Melting 191

v 17 – Al-Factor and CaO/Al2O3 vs. Fe-Factor Indicating Depth of 193 Crystallization and Phases Involved

18 – Results of Modeling of Deep Fractional Crystallization of JC-30B 195

19 – Results of Binary Mixing Between Fraction Derived from JC-30B 197 and Melt of Rhyolitic Composition

20 – Results of Binary Mixing Between HAOT (Cow Vent Complex) 199 and SROT-like (V 4569) compositions

21 – End-member Plumbing System Scenarios for Monogenetic 201 Volcano-Fields

22 – Diagram for Conceptualizing Important Parameters in 203 Monogenetic Volcanism

Appendix 2: Summary of Methods Used in 40Ar-39Ar 215 Geochronology

1-7 – Plateau/Isochron Plots for the Dated JVVF Samples 217- 223

Appendix 4: Geochemical Data for Samples From the JVVF 243

1 – JVVF Sample Locations on Satellite Image 252

ACKNOWLEDGEMENTS

It gives me great pleasure to be able to explicitly express a deep sense of gratitude towards everyone that was involved in this dissertation, either directly or indirectly. Right at the outset, I apologize to those whom I might have inadvertently forgotten to acknowledge here. The support and guidance of a number of people was instrumental in the successful completion of this research. My advisor, Bill Hart leads this list. Bill was gracious enough to accept me into his research group and introduce me to the fascinating geology of the northwestern United States. I have learnt immensely from the numerous stimulating discussions that we have had for the past four years and thoroughly enjoyed our time together in the field. He deserves a great deal of credit for initiating me into the world of geochemical analysis and interpretation – acquiring this skill has set me on the path to becoming a well-rounded geologist. Bill encouraged me to think independently and I am indebted to him for being open and receptive to my ideas and interpretations. I would also like to thank everyone on my committee (Craig White, Elisabeth Widom, Jason Rech, Brian Currie and Mike Crowder) for their assistance and encouragement. Their feedback and suggestions at various stages of this dissertation, including the comprehensive examination and proposal defense are greatly appreciated. I especially thank Craig White for agreeing to be my external reader and for being present for my defense. I am very grateful to John Morton for his patience in assisting me with laboratory work as well as for the interesting conversations that we had from time to time. Many thanks also to Stan Mertzman (F&M) for his assistance with XRF analyses and for his promptness. I also owe much gratitude to Darin Snyder, with whom I have been sharing my office for the past few months. Darin has helped me in many different ways over the past few years, but especially in acquiring isotopic data. I have had many interesting discussions with him on all kinds of topics, ranging from geology to history to global politics. Thank you Darin. I also thank Matt Brueseke and Stephen Pasquale for their help in the field. I have enjoyed the numerous discussions that Matt and I have had on geological and non-geological topics. Discussions with John Rakovan and Mike Brudzinski have also been very stimulating. Conversations with Deb Jaisi provided

vii welcome relief during the hectic final few months of the research. Thanks also to Cathy Edwards for help with many practical issues during this dissertation and also to Jeanne Johnston. Zu Watanabe has been a source of encouragement and support throughout my stay in Oxford. I am especially indebted to her for being open and welcoming to me at a time when I was new to the United States. I thank her for the many happy times that we have shared. Zu also assisted me in fieldwork in India as part of my Sangamner sub-project. Special thanks to Cooper Brossy and Lisa Ely of Central Washington University for initiating collaboration regarding geomorphic aspects of the Jordan Valley area. I have enjoyed fieldwork with both of them, and it has been great to discuss the geology of that area with Cooper. Collaboration with Hetu Sheth has also been very fruitful and I thank him sincerely for his assistance with Discriminant Function Analysis. The late Prof. G. P. L. Walker has been a source of inspiration to me – his help at the initial stages of my geologic career is gratefully acknowledged. These acknowledgements would be incomplete without mentioning the profound role that my teachers from India have played in my growth as a geologist and a person. I owe a special debt to Drs. Vivek S. Kale, V. V. Peshwa, S. M. Chitale, P. K. Sarkar, and Sandhya Joshi for their insights and guidance. Thanks also to all my teachers from Karnataka High School for their role in providing me with a solid foundation. My colleagues Raymond Duraiswami, Gauri Dole, Shilpa Patil-Pillai and Vinit Phadnis have always been great company and intellectually stimulating. My family in India has been a constant source of support throughout my life. I am what I am today largely because of them. Thank you so much for your love and openness. Lotta, my girlfriend, has stood behind me like a rock during the past two years. Not only has she shared my joy and elation, but also moments of grief and frustration with utmost empathy. Thank you for believing in me. Finally, this research would not have been possible without financial support from numerous organizations, including the National Science Foundation (to Bill Hart) and the Geological Society of America and Sigma-Xi (to Ninad Bondre). Support from Miami University is also gratefully acknowledged.

viii CHAPTER 1

INTRODUCTION

Basaltic magmatism is a ubiquitous component of geologic activity on the earth and occurs in diverse tectonic settings. It often results in surface volcanism, which is manifested in numerous forms including Shield Volcanoes, Cinder Cones, and Strato-Volcanoes. The visible edifice of a volcano and associated lavas, however, are only one part of a volcanic system (Walker, 1993), which also consists of the magma source, magma chambers, and intrusives. A thorough understanding of a volcanic system requires an integration of field, petrologic, geochemical, and geochronologic information. Such information (in addition to geophysical information), if used judiciously, has the potential to shed light on the entire set of processes ranging from generation of mafic magmas at the source, to their possible differentiation, transport, and eventual eruption. This dissertation consists of three sub-projects, each of which focuses on a particular aspect of a continental basaltic field. Distinct issues of geologic significance are addressed in each of the regions; yet, these sub-projects are united by their emphasis on answering questions common to basaltic magmatism. Large flood basalt flows are often fed by linear vent systems that are subsequently exposed as dike swarms. Understanding their relationship with stratigraphically recognized flow packages is crucial in understanding the eruptive history of such provinces. Chapter 2 presents the results of sub-project 1, which focuses on the Sangamner mafic dike swarm from the western Deccan Volcanic Province (DVP). At the time when this work was undertaken, many questions pertaining to the Deccan magmatic episode remained unanswered. Some of these, which were raised by Peng et al. (1998) and Mahoney et al. (2000), include (a) Which were the principal vent areas for the various formations of the Deccan stratigraphy? (b) Were the eruptions of individual flows of the different formations largely monocentric (feeder dikes concentrated in a particular region) or polycentric (feeder dikes more widely distributed)? (c) How, if at all, did eruptive and intrusive activity fluctuate through space and time? In order to attempt to answer such questions, a combination of field, geochemical and isotopic (Sr and Nd) characteristics was used in order to understand the relationship of the Sangamner dike swarm with the associated lava flows and their position in the established Deccan stratigraphy. The results (Bondre et al. 2006) indicate that several dikes are geochemically similar to younger formations, and that NE-

1 SW trending dikes in the central part of the area possibly represent a fissure system that could have fed lava flows of the Poladpur and / or Khandala formation. Understanding the eruptive history of Continental Flood Basalt provinces has an important bearing on topics pertaining to mass extinctions, magma supply rates, and planetary volcanism. Information pertaining to the morphology and emplacement of the constituent flows is critical to such an endeavor. Chapter 3 discusses the second sub-project, which deals with the morphology and emplacement of lava flows belonging to the Steens Basalt in southeastern Oregon. The Steens Basalt constitutes the main sequence of mid-Miocene tholeiitic flood basalt lava flows found on the Oregon plateau. Although it has received extensive attention in terms of its geochemistry, geochronology and stratigraphy (e.g. Gunn and Watkins, 1970; Carlson and Hart, 1987; Camp et al. 2003), its morphology had been only cursorily studied. Yet, such a study was essential, particularly in light of recent calls for greater emphasis on lesser-studied flood basalts in order to understand fundamental questions pertaining to their morphology and emplacement (e.g. Bondre et al. 2004). This study finds considerable morphological diversity within the Steens Basalt. Compound pahoehoe flows (often plagioclase phyric) are abundant, but a’a and transitional flows are not uncommon either. The abundance of pahoehoe flows is significant in that it hints at slow but sustained eruptions, probably through small to moderate- sized shield volcanoes. There is some evidence that certain eruptions produced pahoehoe as well as a’a flows. This suggests that certain physical variables (such as volatile content, effusion rate) might have changed significantly during single eruptions. Similarly, the alternation of relatively aphyric as well as strongly plagioclase phyric lavas in some sections indicates at a link to deeper processes, and should be pursued through an integration of morphological and geochemical studies. The third sub-project (Chapter 4) integrates field, geochemical and geochronologic information in order to understand the physical make-up and magmatic evolution of the newly- proposed Jordan Valley Volcanic Field (JVVF). This monogenetic volcano-field provides an excellent opportunity to investigate the spatio-temporal evolution of small-volume basaltic volcanism and the structural and geomorphic controls on, as well as effects of such volcanism. Compositional diversity also permits the investigation of source and process-related effects, including length-scales of mantle heterogeneity and various differentiation processes affecting relatively primitive basalts in a continental setting. The former aspect has been a matter of

2 increased interest in the recent past (e.g. Kogiso et al. 2004). Similarly, given its location at the nexus of the High Lava Plains and Yellowstone-Eatern Snake River Plain trends, this volcanic field has the potential to enhance our understanding of the complex tectono-magmatic evolution of the northwestern United States (Shoemaker, 2004 and references therein). The results of this study, discussed in chapter 4, highlight the important role of long-lived vent alignments (reflecting regional structural trends) in controlling the location of vents. Two major pulses of volcanic activity, one around 2 Ma and another at 0.25 Ma are inferred from the geochronologic data. Chemical and isotopic data help recognize two overall chemostratigraphic groups, and considerable between vent and within vent diversity. Compositional diversity in the JVVF, when interpreted in the context of previous work (e.g. Hart et al. 1997; White et al. 2002), hints at mantle heterogeneity at the small scale, notwithstanding the complications imposed by fractional crystallization and potential contamination. It is proposed that monogenetic volcano-fields such as the JVVF are ideal settings for understanding the complexity of physical and chemical processes involved in continental basaltic magmatism, and length scales of mantle heterogeneity in continental settings.

3 References

Bondre, N. R., Duraiswami, R. A., and Dole, G (2004). Morphology and emplacement of flows from the Deccan Volcanic Province, India. Bull. Volcanol, v. 66, pp. 29-45.

Bondre, N. R., Hart, W. K., and Sheth, H (2006). Geology and geochemistry of the Sangamner mafic dike swarm, western Deccan Volcanic Province, India: implications for regional stratigraphy. J. Geol, v. 114, pp. 155-170.

Camp, V. E., Hanson, W. E., and Ross, M. E (2003). Genesis of flood basalts and Basin and Range volcanic rocks from Steens Mountain to the Malheur River gorge, Oregon. Geol. Soc. Am. Bull, v. 115, pp. 105-128.

Carlson, R. W., and Hart, W. K (1987). Crustal genesis on the Oregon Plateau. J. Geophys. Res, v. 92, pp.6191-6206.

Gunn, B. M., and Watkins, N. D (1970). Geochemistry of the Steens Mountain Basalts, Oregon. Geol. Soc. Am. Bull, v. 81, pp. 1497-1516.

Hart, W. K., Carlson, R. W., and Shirey, S. B (1997). Radiogenic Os in primitive basalts from the northwestern U.S.A.: Implications for petrogenesis. Earth. Planet. Sci. Lett, v. 150, pp. 103- 116.

Kogiso, T., Hirschmann, M., and Reiners, P (2004). Length scales of mantle heterogeneities and their relationship to coean island basalt geochemistry. Geochim. Cosmochim. Acta, v. 68, pp. 345-360.

Mahoney, J. J., Sheth, H C., Chandrasekharam, D., and Peng, Z. X (2000). Geochemistry of flood basalts of the Toranmal section, northern Deccan Traps, India: Implications for regional Deccan stratigraphy. J. Petrol, v. 41, pp.1099-1120.

Peng, Z. X., Mahoney, J. J., Hooper, P. R., Macdougall, J. D., and Krishnamurthy, P (1998). Basalts of the northeastern Deccan Traps, India: Isotopic and elemental geochemistry and relation to southwestern stratigraphy. J. Geophys. Res, v. 103, pp. 29,843-29,865.

Shoemaker, K. A (2004). The tectonomagmatic evolution of the late Cenozoic Owyhee Plateau, northwestern United States. Ph.D. Dissertation, Miami University, 276 p.

Walker, G. P. L (1993). Basaltic-volcano Systems. In: Pritchard HM, Alabaster T, Harris NBW, Neary CR (eds) Magmatic processes and plate tectonics. Geol. Soc. Spec. Publ 76, pp. 3-38.

White, C. M., Hart, W. K., Bonnichsen, B., and Matthews, D (2002). Geochemical and Sr- isotopic variations in Western Snake River Plain basalts, Idaho. In: Bonnichsen, B, White, CM and McCurry, M (eds) Tectonic and Magmatic Evolution of the Snake River Plain Volcanic Province. Idaho Geol. Surv. Bull 30, pp. 329-342.

CHAPTER 2

GEOLOGY AND GEOCHEMISTRY OF THE SANGAMNER MAFIC DIKE SWARM, WESTERN DECCAN VOLCANIC PROVINCE, INDIA: IMPLICATIONS FOR REGIONAL STRATIGRAPHY

N. R. Bondre1, W. K. Hart1, and H. C. Sheth2

1 Department of Geology, Miami University, Oxford, Ohio – 45056, USA; 2 Department of Earth Sciences, Indian Institute of Technology (IIT) Bombay, Mumbai 400 076, India.

Published in the Journal of Geology, v. 114, 2006

ABSTRACT

Numerous large, NE-SW to E-W-trending mafic dikes outcrop around Sangamner in the western Deccan Volcanic Province (DVP). This area is part of a broader region postulated to be a shield- like feature and a major eruption center. A combination of field, geochemical and isotopic (Sr and Nd) characteristics is used here to understand the relationship of this dike swarm with the associated lava flows and their position in the established Deccan stratigraphy. Many dikes are compositionally similar to the Khandala and Poladpur Formations belonging to the Lonavla and Wai Subgroups respectively, while one dike is similar to the Ambenali formation. One dike has a composition distinct from all other dikes in this area as well as from most stratigraphic units, although there are many similarities in composition with the Bushe Formation as well as the Boyhare Member of the Khandala Formation. While several dikes are geochemically similar to specific flows/members within certain formations, their isotopic composition is often different, sometimes significantly so. This implies that there is either a greater range in isotopic composition for those members than previously realized, or that magmas with different isotopic compositions underwent broadly similar petrogenetic evolution leading to similarities in elemental composition. NE-SW trending Poladpur and/or Khandala-like dikes are concentrated

in the central part of the area; these dikes appear to represent a vent system that could have fed southern, western or eastern exposures of these younger formations. It is also possible, however, that some or many of the dikes along this system were simply late-stage intrusions of magmas representing the younger formations.

Introduction

The intrusive component of continental flood basalt provinces is often a significant part of the total magmatism (e.g. Walker 1993). This includes sills, dikes and plugs occurring at all levels within the crust. Dikes form an important part of the magmatic plumbing system and can also be related to the prevalent tectonic stresses. Lavas constituting individual flow fields in such provinces often erupt from discrete fissure systems, exposed subsequently as dike swarms. Linear vent systems, represented by dike swarms and accumulation of near-vent deposits, have been identified for several stratigraphic units of the Columbia River Basalt Province (e.g., Swanson et al. 1975). This has aided not only stratigraphic work, but has also had a bearing on physical volcanological studies involving magmatic output and long-distance transport. With a presently exposed area of over 500,000 km2 and straddling the latest Cretaceous- early Tertiary time period, the Deccan Volcanic Province (DVP) of India is a major flood basalt province. Detailed geochemical (including isotopic) and palaeomagnetic studies by numerous research groups over the past two decades have led to the establishment of a geochemical stratigraphy for the southwestern DVP (table 1), and its extension to the northern and eastern parts of the province has also been attempted (e.g., Cox and Hawkesworth 1984, 1985; Beane et al. 1986; Lightfoot et al. 1990; Khadri et al. 1988; Peng et al. 1998; Mahoney et al. 2000; Sheth et al. 2004). While much of the previous work pertains to lava flows, a few workers (e.g., Deshmukh and Sehgal 1988; Bhattacharji et al. 1996; Melluso et al. 1999; Subbarao et al. 1999) have also studied the two principal dike swarms in this province where dikes occur with a high frequency. The West Coast Dike Swarm (WCDS; fig. 1), trending N-S to NNW-SSE, consists of dikes with both tholeiitic and alkaline compositions (e.g., Melluso et al. 2002). The Narmada- Tapi Dike Swarm (NTDS; fig 1), also contains tholeiitic as well as alkaline dikes, and has a predominant ENE-WSW trend (e.g., Sheth 1998; Melluso et al. 1999). In the WCDS south of Mumbai (Bombay), dikes are reported to intrude flows of the youngest (Wai) Subgroup (Hooper

1990) and thus cannot have fed the bulk of the Deccan lavas. However, basaltic dikes that outcrop farther south in Goa, have been argued to be the feeder dikes of some of the younger formations (Widdowson et al. 2000). Based on the common occurrence of alkaline dikes in the two principal swarms and their wider span of ages, Hooper (1990) argued against these being the principal vents for the Deccan eruptions and ascribed this role to the dikes outcropping in the Western Ghats region, roughly between Nasik and Pune (fig. 1). On the contrary, Bhattacharjee et al. (1996), based on geochemistry and K-Ar dates for several tholeiitic dikes from both of these swarms and their associated lava flows, argued that these dikes were feeders to the lavas. The frequency of dikes in the Mumbai-Nasik-Pune region, the third important zone of dikes, is quite variable, both in a lateral and vertical sense. These dikes do not display the strong preferred orientation typical of the other two swarms. Beane et al. (1986) observed that their compositions are similar to those of the associated flows. In conjunction with the specific disposition of the chemostratigraphic formations, they suggested that such random orientation is more probably associated with development of a central shield-type volcanic edifice rather than with true fissure eruption as exemplified by the Columbia River Basalts. Bhattacharjee et al. (1996) considered the random orientations to be the result of a stress regime dictated by large crustal magma chambers. Sheth (2000) argued that true feeder dikes in central volcanoes usually have a radial, not random, arrangement. To the south and southeast of Pune, dikes are virtually absent and no dikes are observed even in areas such as Mahabaleshwar (fig. 1) with excellent vertical exposures. This paucity of dikes in the region dominated by the thick Wai Subgroup (south and south-east of Pune) is perplexing. Lavas along the southeastern fringe of the DVP are well over 200 km away from the nearest exposed dikes. Irrespective of which of the three principal dike swarm regions was the eruptive focus, it seems highly likely that some Deccan lavas flowed distances comparable to flows from the CRB, i.e., several hundred kilometers.

Objectives of this study. In spite of considerable previous work, many questions regarding the nature of the Deccan magmatic episode remain to be satisfactorily answered, some of which were raised by Peng et al. (1998) and Mahoney et al. (2000). (a) What were the principal vent areas for the various formations of the Deccan stratigraphy? (b) Were the eruptions of individual flows of the different formations largely monocentric (feeder dikes concentrated in a

particular region) or polycentric (feeder dikes more widely distributed)? (c) How, if at all, did eruptive and intrusive activity fluctuate through space and time? In order to attempt to answer these questions, an integration of several kinds of information is required. Such information includes the distribution of dikes representing various chemical types and field evidence of feeder dikes, in addition to details of the thickness and distribution of various chemostratigraphic units. Most previous studies have looked at dikes from the DVP only on a broad, regional scale. Although valuable in many respects, they cannot provide the information and insights that a focused field and geochemical investigation of a specific, sub-regional dike swarm can. The present study is one such focused investigation that documents the field relations and major, trace, rare-earth element and Sr and Nd isotope geochemistry of the dikes around Sangamner (fig. 1). These data are used to evaluate these dikes in a stratigraphic context.

Field geology and petrography

Figure 2 is a geological map of the study area, which is situated close to the Western Ghats Escarpment (WGE; fig. 1). This area has moderate relief and is drained by tributaries of the Pravara and Mula rivers (fig. 2). The basalt flows are nearly flat-lying (the sequence has a regional southerly dip of 0.5-1°) and mainly belong to the Thakurvadi Formation (Fm.) of the Kalsubai Subgroup (Khadri et al. 1988; Subbarao and Hooper 1988). Some isolated, high peaks in the westernmost part of the area are formed by basalt lavas of the overlying Bhimashankar Fm. The flows are intruded by a minimum of 25 basaltic dikes. Extensive colluvio-alluvial deposits (locally up to 30 m thick) of the late Quaternary Pravara Formation (Bondre 1999) overlie the basalts along the Pravara River and its tributaries. Patches of these sediments are also found along the Mula River. The basalt flows are classic compound pahoehoe, ranging in thickness from few tens of meters to well over 50 m, and are made up of individual flow lobes ranging in thickness from a few cm to 20 m (Bondre et al. 2000; 2004). The Thakurvadi Fm. is characterized by a wide range of MgO contents (3.5-17 wt.%); however, flows with MgO contents of 6-8% are most common (Khadri et al. 1988). The southwestern and northwestern part of the map area is remote, rugged, and highly vegetated; dikes are harder to spot in this region. Exposure is also poor in the region dominated by the alluvium. However, extensive field work aided by multi-spectral satellite data (from the

Indian Remote-Sensing Satellite IRS-1B) helped in identifying a majority of the dikes. These typically form positive relief features such as prominent humps or spines owing to their resistance to erosion. Only two dikes display negative relief. The dikes range in width from 1-18 meters (m), with a mean width of 7.5 m and a mode of 6 m. Most dikes are simple and have a chilled margin on either side. Four dikes, however, appear to be multiple intrusions. For example, dikes Ch20 and Ch25 show more than two chilled margins. Many of these dikes can be traced over several kilometers. They form aligned, discontinuous humps, and their outcrop width shows significant variation along strike. Off-shoots and apophyses are very common and indicate opportunistic exploitation of fractures by the intruding magma. There is a curious tendency for dikes to occur in pairs (fig. 2). The two dikes forming each pair are separated by a few meters to a few tens of meters. As discussed below, closely spaced dikes do not necessarily have identical chemical composition. Many dikes display well-developed columnar jointing, the disposition of which helps identify their dips; most dikes are not vertical and dip at angles of 70-80°. Two dikes show evidence of deformation during cooling in the form of twisted columns, although this might also be a result of disturbed geotherms due to seepage of water, something that is commonly observed in lava flows. At one locality, dike Ch12 shows near-vertical columns and appears to have spread out locally as a sill. A distinct NE-SW trend for most dikes, with a subsidiary E-W trend is quite apparent from the geological map and the rosette diagram (fig. 2). Field evidence for feeder dikes is quite limited in the DVP; however, excellent evidence of near- vent material (stratified spatter and breccia) was observed at one location (fig. 2) in the area. Plagioclase, olivine, clinopyroxene and iron oxides are the principal mineral phases in the dikes. Most dikes are distinctly porphyritic, and plagioclase is the only phenocryst phase in many. Plagioclase glomerocrysts 1-5 mm in size are extremely common, while one dike has plagioclase megaphenocrysts (> 2 cm). Microscopic observations indicate that almost all plagioclase phenocrysts show zoning, sometimes in complex patterns. Some of the more magnesian dikes have olivine phenocrysts, as well as infrequent glomerocrysts consisting of olivine and clinopyroxene (optically determined to be sub-calcic augite in most cases, rarely pigeonite). While the groundmass of some dikes is markedly inequigranular with intergranular or intersertal textures, it is holocrystalline and equigranular in others with plagioclase and clinopyroxene in a sub-ophitic relationship. Iron oxides are abundant, particularly in the more evolved dikes, which also show an interstitial late-stage residuum or devitrified glass. The

margins of most dikes show a much finer texture than the central regions, with plagioclase phenocrysts more abundant in the central parts.

Geochemical and isotopic characteristics

Analytical methods. Fresh, texturally diverse samples were selected for analysis. Most dikes in the area appear to be quite fresh with no or little evidence of alteration in hand sample. The more mafic dikes have undergone some alteration as evidenced by discolored patches and iddingsitization of olivine. However, some relatively fresh olivine cores can be observed even in these. Concentrations of major elements in all dike samples were measured by Direct Current Argon Plasma Atomic Emission Spectroscopy (DCP-AES) at Miami University following the procedures of Katoh et al. (1999). Trace elements were determined by XRF techniques at Franklin and Marshall College, following the methods of Boyd and Mertzman (1987). Errors are

< 2 % for most major elements (~ 1% for SiO2, 5% for K2O and 10% for P2O5) and 1-5% for most trace elements measured by XRF. The accuracy of the XRF Pb data at the low concentrations measured in this study must be considered. For example, the measured value for USGS standard BHVO-2 is 4 ppm while the accepted value is 3 ppm. Samples from this study that show more or less identical concentrations of other trace elements (e.g. Ch11 and Ch22) show a difference in their Pb values (5 and 7 ppm respectively; table 2). Since a difference in Pb values of 1 ppm can make a significant difference in the form of primitive mantle normalized plots, interpretations involving the presence or absence of Pb anomalies in such diagrams must be treated with caution. Selected rare earth elements (REE) were analyzed by ICP-MS at Miami University. Sample dissolution is achieved using 100 mg of sample powder thoroughly mixed with 300 mg of Li metaborate flux. This mix is placed in a graphite crucible and fused in a furnace at 950 °C for 20 minutes. The resultant molten bead is dissolved in 50 ml of 5% HNO3 and the solution is loaded onto a 1 cm diameter quartz glass column containing 22 g of AG50W-X8 cation exchange resin. The non-REE fraction of the sample is removed by using 210 ml of a mixed

(0.1M Oxalic and 2M HNO3) acid. After this step, 250 ml of quartz-distilled water is passed

through the column, and REE are collected in Teflon beakers using 200 ml of 5M HNO3. After

drying down on a hot plate at low heat, the REE fraction is re-dissolved in 1 ml of 5M HNO3 and

5 ml quartz-distilled water, transferred to 20 ml Teflon beakers and dried down again. The

residue is dissolved in 18 ml of 1% HNO3 and the weight of this solution is recorded. The REE were analyzed from this solution using a Varian ICP-MS (plasma power of 1.37 kW and pump rate of 5 rpm). Measurements for each sample include 5 replicates, with 40 scans per replicate. Calibration curves (linear regressions) for each element were generated from three standard solutions with known concentrations (5, 160, and 500 ppb respectively) and an acid blank, that were run along with the unknowns. Concentrations of unknown samples were calculated from these curves, with 115In serving as the internal standard to monitor and correct for instrumental drift. Two additional blank solutions and a 500 ppb solution were also run as unknowns. Measured concentrations for the 500 ppb solution were within 3% of this value. Concentrations for each element were corrected using the average blank values; however, the blank concentrations were quite low and these corrections did not make a significant difference to the values. Whole-rock Sr and Nd isotopic compositions of nine dikes (covering the spread in elemental compositions) and one lava flow were measured in static mode using a Thermo- Finnigan Triton multi-collector TIMS at Miami University. Separation of Sr and a bulk LREE fraction was achieved using methods similar to those described in Walker et al. (1989) and Snyder (2005). Sr isotopic ratios were corrected for fractionation using 86Sr/88Sr=0.1194. A 2 SD (standard deviation) external reproducibility of 1.4 x 10-5 based on sixty eight measurements of standard NBS 987, which resulted in an average 87Sr/86Sr=0.710236, is quoted for all ratios. Nd was separated from the remaining LREE using an EiChrom Ln-Spec resin, following methods similar to Pin and Zalduegui (1997). Nd isotopic ratios were corrected for fractionation using 143Nd/144Nd=0.7219. A 2 SD external reproducibility of 7 x 10-6 based on sixty one measurements of the La Jolla standard, which resulted in an average 143Nd/144Nd=0.511846, is quoted for all ratios. The isotopic analysis (including column separation) was repeated for sample Ch1 in order to check the reproducibility of the obtained values. The fractionation corrected, measured ratios were age-corrected using whole-rock Rb and Sr (XRF) and Sm and Nd (ICP-MS) concentrations. It is appropriate at this stage to demonstrate that changes to Rb and Sr concentrations well outside the analytical errors make only a small difference to the age-corrected Sr-isotopic ratios. For example, the present day Sr isotopic ratio of Ch1 is 0.70684 (table 3). Based on the Sr

and Rb concentrations of 212 and 35 respectively, the age corrected ratio obtained is 0.70641 (for 66 Ma). Changing the Rb concentration to 25 changes the corrected ratio to 0.70653. Changing the Sr concentration to 180 changes the corrected ratio to 0.70634. These changes are very small as compared to the spread in values for any given formation in the Deccan, and hence the age corrected isotopic data can be considered to be quite robust.

Results. Geochemical data for the samples and some derived parameters such as Magnesium Number (Mg#) are listed in Table 2, while REE and Sr and Nd isotopic data are given in Table 3. Low values (< 1.0 wt.%) of loss on ignition (LOI) for all samples except one (table 1) suggest that the dikes have suffered only low levels of alteration and their elemental abundances thus reflect primary values. Previous studies in the Deccan (e.g., Mahoney et al. 1985, 2000; Beane et al. 1986) have indicated that K and Rb can be affected even at modest levels of alteration, while at higher levels, Ba can also be mobile. Rb values for widely separated dikes that show very similar concentrations of other trace elements (e.g., Ch11 and Ch22) are also similar, suggesting

that this element might not have been very mobile in most samples. Uncertainties on K2O measurements by DCP-AES at the concentrations present in the dikes are relatively high and it is difficult to discriminate alteration-related effects from those related to instrumental precision.

Hence, K2O has not been used as a discriminant for stratigraphic correlation in this study (except as one of the many variables in one run of the discriminant function analysis). The repeat analysis of sample Ch1 yielded 87Sr/86Sr=0.706850 and 143Nd/144Nd=0.512610, within the quoted 2 SD external reproducibility mentioned earlier. Although the range in composition of the dikes is not large, discrete groups can nevertheless be identified on the basis of many major and trace elements and their ratios. The propensity of dikes to occur in a particular group does not seem to display a clear relationship to their location. Two dikes separated by several kilometers may show virtually the same composition (e.g., Ch11 and Ch22; fig. 2, table 2). On the other hand, dike Ch10 which is just a few meters away from Ch11 and is parallel to it shows a markedly different texture and a somewhat different composition (table 2). The highly evolved compositions of some dikes and high iron contents indicate that they were derived from a more fractionated magma (e.g. Ch1, Ch26a).

Chemostratigraphic affinities

Previous studies of lava piles exposed in parts of the DVP lacking an established stratigraphy have employed several tools in order to aid comparisons with the southwestern Deccan formations (Peng et al. 1998; Mahoney et al. 2000; Sheth et al. 2004). These tools include binary discriminant diagrams using elements or isotopes, normalized multi-element patterns, and statistical methods such as discriminant function analysis (DFA). As this multi-pronged approach has proved to be reasonably or very successful for chemostratigraphic correlation, this investigation has also adopted the same general approach. It must be borne in mind that intrusives present some unique problems in stratigraphic correlation. Just like geochemical and magnetic polarity data, relative position in a vertical section is both helpful and critical in correlating flows from two widely spaced areas. Such relationships can often be difficult to ascertain for dikes and sills. In the present case, the only certainty in terms of stratigraphic position is that the dikes are younger than or similar in age to the Thakurvadi Formation which they intrude. Cross-cutting dikes can sometimes be recognized, but poor exposure in regions of low to moderate relief often hampers this. Any attempt at correlating dikes with established stratigraphic units must therefore keep this limitation in mind.

Binary Diagrams. Figure 3 shows plots constructed using concentrations and ratios of some key elements, which have been used by previous workers (e.g. Mahoney et al. 2000; Sheth et al. 2004) to discriminate between various formations. Although there is some diversity within the composition of the dikes (table 2), they define a relatively tight cluster against the backdrop of fields for the southwestern Deccan formations, as seen in Figure 3. This figure shows that many dikes fall well within the fields for younger formations, notably the Khandala, Bushe, Poladpur and Ambenali. It is apparent that there is substantial overlap in the compositional characteristics of several formations. Therefore, such plots by themselves are insufficient to discriminate between them. However, it is certainly possible to eliminate from any consideration formations that do not show significant overlap, such as the Mahabaleshwar. While some dikes plot within the Ambenali field in these plots, only a few of them exhibit other geochemical characteristics that are quite distinctive of the Ambenali, such as low Ba concentrations (typically < 100 ppm; Cox and Hawkesworth 1984, 1985; Beane et al. 1986), thus potentially ruling out an Ambenali

affinity for the majority. Similarly, the affinity of most dikes with the Bushe Fm. can be ruled out because of the lack of distinctive characteristics such as low TiO2 contents (typically < 1.5 wt.%). Thus, based on these simple plots and some general characteristics, most dikes seem to correlate with the Khandala and Poladpur Fms. The more magnesian dikes (samples Ch5 and

Ch20) remain problematic. The low TiO2 and Zr values for dike Ch17 coupled with high Rb and

K2O hint at a Bushe or Boyhare (a member of the Khandala Fm.) affinity for this dike, although some aspects of its composition are unlike any other stratigraphic unit.

Discriminant function analysis. Discriminant function analysis (DFA) was performed in order to quantitatively evaluate chemical affinities of the Sangamner dikes to individual southwestern Deccan formations. For this purpose, a data set consisting of 623 samples from all the southwestern formations except the Panhala was processed using the SPSS 7.5 for Windows (Student Version) software. The methodology utilized is essentially the same as followed by earlier workers (Peng et al. 1998; Mahoney et al. 2000; Sheth et al. 2004). No derived variables (Zr/Y ratio, Mg Number, etc.) were used. Few Pb-Th-U-REE data exist for the southwestern Deccan basalts and therefore these elements were not used as discriminating variables. Cr was not used because of contamination problems with some of the southwestern Deccan Cr data, and

Na2O and MnO were not used because of their limited range of variation in the southwestern data set, low precision and variable analytical quality, besides alteration effects. The southwestern Deccan and Sangamner data were first transformed to standardized values (Z-scores). The Z-score for a sample for any element is the number of standard deviations it is from the mean. The program calculated the F-statistic (essentially the ratio of the between- group variability to the within-group variability) for each variable, and also the discriminant functions, group centroids, and Mahalanobis distance of each sample from the nearest formation centroid. A lower value for the Mahalanobis distance indicates a greater probability of a sample belonging to a particular formation. For the southwestern Deccan data set, eight canonical discriminant functions were obtained. These functions from one to eight account for progressively decreasing percentages of the total variance in the southwestern Deccan data set. This means that function 1 is a more effective discriminator between the groups (i.e., stratigraphic formations) than function 2, which is better than function 3, and so on. It is the two most effective discriminant functions (1 and 2) that are used in this paper (Fig. 4).

Table 4 and Figure 4 show the results of two different runs of the DFA, one using major

(excluding Na2O and MnO) as well as trace elements (Ni, Sc, V, Ba, Rb, Sr, Zr, Y, and Nb) and another using only trace elements. In the first run, 84.6% of original grouped cases were correctly classified and Functions 1 and 2 (out of the eight calculated by the program) account for 69.4% (50 and 19.4 % respectively) of the total variance in the southwestern Deccan data set.

Most samples were matched with the Igatpuri-Jawhar Fms. Progressive exclusion of K2O, then

K2O, MgO and Ni, and finally K2O, MgO, Ni and Fe2O3* did not make any significant difference to the result. This result is obviously incorrect since the dikes intrude the Thakurvadi Fm. which overlies the Igatpuri-Jawhar Fms (table 1). There is substantial overlap in the major element composition of all southwestern Deccan formations, which might have contributed to this anomalous result. In other words, major elements did not prove to be useful discriminants. Significantly different results were obtained, however, when DFA was performed using only the trace elements and no major elements. In this run, 68.4% of original grouped cases were correctly classified, and functions 1 and 2 together account for 69.4 % (42.3 % and 27.1 % respectively) of the total variance in the southwestern Deccan data set. None of the samples was now matched with the Igatpuri-Jawhar Fms; rather, many samples are classified with the Thakurvadi, Bhimashankar, and Khandala Fms. Ch4b, a stubby flow associated with dike Ch4c is matched with the Thakurvadi Fm. Results of this run attest to the much greater variation in the southwestern Deccan lavas in terms of trace elements as compared to major elements. Previous experience (e.g., Sheth et al. 2004) suggests that DFA by itself is insufficient for correlation and must be used in conjunction with other types of evidence (such as isotopic data and normalized multi-element patterns) to aid correlation.

Normalized multi-element patterns and isotopic composition. Comparisons using binary diagrams and DFA provided preliminary matches of the dikes to various formations. In order to further evaluate their stratigraphic affinities, selected dikes were compared with individual flows or members from various southwestern Deccan formations using primitive-mantle-normalized multi-element patterns, chondrite-normalized REE patterns, and Sr and Nd isotopic data (figs. 5, 6 and 7). At places in this section differences in patterns based on the behavior of Pb are recognized. However, due to the generally low Pb concentrations as measured by XRF, we favor using the shape of the overall pattern, rather than individual features such as Pb peaks, in the

ensuing correlations. Few previous studies have used REE patterns specifically for the purpose of chemostratigraphic correlation, as has also been pointed out by Widdowson et al. (2000). One of the reasons for this is that REE data for the Deccan are relatively few. However, as shown by Widdowson et al. (2000), these can prove to be quite useful in discrimination and are therefore used here as supporting evidence for inferences drawn from other multi-element patterns. Ch1 and Ch11 (and dikes similar to them) appear to be similar to the Dhak Dongar (DD) Member of the Khandala Fm. based on their multi-element and REE patterns (fig. 5c). The Sr concentration of Ch1, however, is much lower than typical DD values. In terms of their Sr and Nd isotopic compositions, these two dikes also plot in the Poladpur field, somewhat away from the DD Member as seen in Figure 7. The patterns for Ch18 resemble that of the Rajmachi Member of the Khandala Fm. (figs. 5a, 6c). The Rajmachi Member shows an isotopic composition distinct from other members of the Khandala Fm. (e.g. Peng et al. 1994), and actually plots in the Poladpur field. Ch18 also plots within the Poladpur field in Figure 7, but away from the Rajmachi Member. In light of the excellent match of multi-element patterns of Ch1, Ch11 and Ch18 with the respective Members, this discrepancy in isotopic composition could indicate that these Members from the Khandala Fm. have a wider range in isotopic composition than previously documented. Nevertheless, on the basis of the isotopic compositions and some aspects of the elemental compositions, a Poladpur affinity cannot be discounted. Dike Ch17 has a multi-element pattern quite similar to the Bushe Chemical Type of the Bushe Fm. (fig.5b); however, it differs in having a pronounced depletion in the HREE as seen in

comparison of REE patterns (fig. 6b). The high Rb and K2O, and low TiO2 of Ch17 are strikingly similar to the Bushe Fm. Ch17 is also similar in many ways to the Boyhare Member of the Khandala Fm., which bears several similarities to lavas from the Bushe Fm. Figure 6b reveals that the slope of the REE pattern for Ch17 is more similar to that of the Boyhare member, as compared to the Bushe Chemical Type. Isotopically, this dike plots in between the fields defined 87 86 by the Khandala and Bushe Fms. (fig. 7). Its Sr/ Sr(T) = 0.7116 is slightly lower than typical values of the Bushe Fm. (>0.7120, except for the distinctive Hari member; Lightfoot et al. 1990; Beane 1988; Peng et al. 1994) and slightly higher than the reported value for the Boyhare member (0.7102; Peng et al. 1994). The exact affinity of Ch17, therefore, is difficult to determine and it could also represent a hitherto unsampled composition from the Deccan. Multi- element patterns of dike Ch6 and Ch16 are very similar (not shown) and these were compared

with that of a flow from the Bhimashankar Fm. (JEB 366; Beane 1988), based on their DFA match. The match is not very close for elements such as the LILE. Their REE patterns are, however, almost identical to that of the Bhimashankar sample (Fig. 6a). Both dikes plot in the Poladpur field in Figure 7 and well away from the Bhimashankar field. Once again, there is disparity between affinities inferred from elemental composition and isotopic composition. Dike Ch13 was classified with the Khandala Fm. based on DFA. It may be noted, however, that dike Ch25, which is compositionally quite similar to Ch13 was matched with the Poladpur formation by the DFA. The multi-element patterns for Ch13 and Ch25 are very similar and seem to agree reasonably well with that of the Kusgaon member of the Poladpur Fm. (fig. 5e). Ch13 also plots in the Poladpur field in Figure 7, which strongly suggests that a group of dikes similar to Ch13 (Ch9, Ch14a, Ch15, Ch23, Ch25 and Ch30) belong to the Poladpur Fm. Dike Ch4c also has a multi-element pattern similar to the Kusgaon member (fig. 6e). Ch4c has 87 86 the lowest Sr/ Sr(T) = 0.7045 among all dikes analyzed during this study, which puts it at the upper end of the range exhibited by the Ambenali Fm (fig. 7) and within that for the Mahabaleshwar Fm. Its rather low Ba concentration of ~62, though, favors a match with the Ambenali or Poladpur Formations (Cox and Hawkesworth 1984, 1985; Beane et al. 1986). However, the prominent Pb peak in its multi-element pattern is quite unlike the typical Ambenali patterns which lack any Pb peak at all. Its affinity is provisionally determined to be Poladpur or Ambenali. Interestingly, flow Ch4b, which shows field evidence of being fed by dike Ch4c has a multi-element pattern identical with that of dike Ch20 (fig. 5e). Ch4b and Ch20 have virtually identical Sr and Nd isotopic composition (fig. 7), which is not surprising given their very similar multi-element patterns. These patterns are similar to those for dike Ch4c and the Kusgaon member, but are depleted in REE and elements such as Ti, Zr and Y. There is excellent field evidence for near-vent explosive activity around dike Ch4c, and the base of Ch4b is welded to the spatter, suggesting emplacement when the spatter was hot. In this context, the discordance in isotopic composition between the dike (Ch4c) and flow (Ch4b) creates a difficult situation to explain. A possible explanation for this discrepancy between field and elemental data and the isotopic data might be that the dike is multiply intrusive. The same dike fracture was probably used by two magmas that were compositionally somewhat different. Material belonging to only one of the pulses was sampled in the field, while the other pulse/pulses presumably remain

unsampled at that location. The other pulse appears to have had a composition similar to dike Ch20 which is strikingly similar to Ch4b, although geographically not close to it. In the absence of other data or explanations, this seems to be a reasonable working hypothesis.

The moderate TiO2 contents of Ch20 and Ch5 (~2 wt.%), along with their high MgO (>8.5 wt.%) hinted at a correlation with the Thakurvadi Fm., which seemed to have been 87 86 confirmed by DFA. However, the Sr/ Sr(T) = 0.7054 of Ch20 is significantly lower than typical values reported for the Thakurvadi (e.g., Peng et al. 1994) and closer to that of the Poladpur/Mahabaleshwar formations. These dikes do not plot anywhere close to the Mahabaleshwar field in any of the binary diagrams used (fig. 3). Ch20 fringes the Poladpur as well as and Mahabaleshwar fields in Figure 7. Most Poladpur Fm. lavas are less magnesian (MgO ~ 6 wt.%) than these dikes; however, a picrite horizon within the Upper Poladpur Fm. has been reported by Cox and Hawkesworth (1985). It is possible that Ch5 and Ch20 represent a more primitive chemical-type from the Poladpur Fm. Dikes Ch7a and Ch10 have near identical multi-element patterns (fig. 5f). While these have the same general form as the pattern for the Dhak Dongar member, they are somewhat depleted in REE as compared to this member (fig. 6d). The patterns for these dikes were also compared with the Bhimashankar Fm; the match is relatively poor for Rb, Ba, and Pb. The REE patterns for these dikes, however, are virtually identical with that of the Bhimashankar sample JEB 366 (Beane, 1988; fig. 6d). Ch7a was correlated with the Bhimashankar Fm. by the DFA, while Ch10 was correlated with the Khandala Fm. In light of the contrasting information provided by the various techniques, the exact affinity of Ch7a and Ch10 remains uncertain.

Discussion

In the DVP, information on aspects such as the spatial distribution, number and frequency of dikes representing various chemical types in the areas of principal dike exposure is scant. The present study is one of the few attempts to present such details. This makes it possible to evaluate these dikes from a chemostratigraphic perspective and to improve our understanding of potential source areas for the Deccan lavas. An integration of several lines of geochemical evidence allows comparisons of the Sangamner dikes with certain southwestern Deccan formations and in some cases, to particular members within these formations. Comparisons with the regional

stratigraphy are not perfect, however. Isotopic ratios for certain dikes are outside of the range typically exhibited by the formation or member to which they can be correlated based on elemental characteristics. This observation is not new; flows from the central and eastern DVP that are geochemically similar to southwestern Deccan formations have been shown to have systematically different isotopic compositions (e.g., Peng et al. 1998). Unlike those flows, however, intrusives from the present study occur in a region where elemental and isotopic ranges were originally established for units in the Deccan stratigraphy. A recent study by Vanderkluysen et al. (2004) also reports that some dikes from the Nasik-Pune region and from the WCDS are similar to some established chemostratigraphic units in terms of their major and trace elemental composition, but have differing isotopic compositions. The Sangamner dikes are compositionally similar to several southwestern Deccan formations, although most of them can be best related to either the Poladpur or Khandala Fms. In terms of their Sr and Nd isotopic composition, almost all analyzed dikes with the exception of two show an affinity with the Poladpur Fm (fig. 7). Interestingly, these dikes encompass a large range of Poladpur isotopic compositions. Such dikes trend almost exclusively NE-SW and are concentrated in a band in the central part of the area (fig. 2), hinting at the possible manifestation of what could previously have been an eruptive fissure system. The nearest exposures of the Khandala Fm. are to the southeast, not too far (~ 10 km) from the southeastern corner of the study area, while the nearest Poladpur exposures are about 50 km to the southeast. Details of the spatial distribution of different geochemical types for dikes south and southeast of the study area are not available. Nevertheless, the Sangamner area could potentially have been an important vent area for lavas of these two formations. This hints at the possibility that the eruption of different chemical types in the DVP was not entirely random in space and in time. Dike Ch4c shows elemental characters and Sr isotopic composition similar to the Ambenali Fm., except for the prominent Pb peak. This formation is exposed dominantly in the region south and southeast of Pune, about 150 km from the location of this dike. If Ch4c indeed represents this formation this suggests that magmas (of an unknown volume) with Ambenali-type composition were intruded and/or erupted much further north of present-day exposures. Ambenali-like flows are quite common in the eastern and northeastern DVP (e.g., Peng et al. 1998), attesting to either the originally extensive nature of this formation, or polycentric eruptions of this magma type.

It is interesting to note the paucity of dikes in this area with a composition similar to the Thakurvadi/Bushe Fms. No dike in the area was correlated with the Thakurvadi (two dikes are similar in terms of elemental characteristics, but isotopically more primitive). This suggests that feeders to the Thakurvadi might occur farther east, north, and northeast in the areas presently exposing the older Igatpuri-Jawhar formations (see Bhattacharji et al. 1996). Dikes with compositions similar to the Thakurvadi Fm. have also been reported from the Tapi region by Melluso et al. (1999). Only one dike from Sangamner, Ch17, possibly correlates with the Bushe Fm. The Bushe lies between the Khandala and Poladpur Fms. and the significance of the scarcity of dikes compositionally similar to it is unclear. This is particularly surprising given that dikes chemically/isotopically similar to the latter two formations are quite abundant. This could suggest that significant volumes of Bushe magmas were not being supplied to this region. Bushe- like dikes and flows have been found in the central DVP (Sheth 1998; Chandrasekharam et al. 1999; Mahoney et al. 2000), although the flows are not necessarily in the same stratigraphic order as in the southwestern Deccan. Previous studies (Beane et al. 1986; Hooper 1990) have emphasized the randomness in dike orientations in the broader Mumbai-Nasik-Pune region. The present study shows that the orientations of dikes in the Sangamner area (which is part of this broader region) are clearly not random (fig. 2). On the other hand, of ten dikes around Igatpuri (south of Nasik, northwest of the present study area), five dikes trend NNW-SSE, while the other dikes show varying orientations (V. M. Phadnis, unpub. data). There is an urgent need for detailed data on dike distribution, frequency and orientation from this part of the DVP. This will help ascertain whether the “randomness” of orientations applies throughout this region (and so is real, not just apparent), or if discrete areas within this region tend to show consistent orientations. Pending such data, any statements regarding dike orientations and the stress regime associated with these dike swarms remain speculative. Most dikes around Sangamner trend NE-SW, but a few also trend E-W. Do these two distinct trends have any temporal significance? The geochemistry of the dikes offers some clues in this regard. Ch11 and Ch22 are compositionally and texturally identical, but while the former trends NE-SW, the latter trends almost E-W (fig. 2; table 2). This is also the case for dikes Ch1 and Ch26a, suggesting that at least some dikes emplaced along both of these trends were of the same chemical types and thus likely contemporaneous. While dike composition in this area is not

necessarily correlated with location, certain compositions do seem to be concentrated in certain regions. Four of the dikes (Ch1, Ch26a, Ch27 and Ch28) are distinctive in that they are geochemically the most evolved in this area, with the highest concentrations of incompatible elements (Rb, Zr), and the lowest MgO contents (<5 wt.%). All of these dikes occur in the southern part of the study area, relatively close to each other (fig. 2), suggesting that that particular magma type was probably supplied to a restricted area.

Conclusions

Dikes of the broad Mumbai-Nasik-Pune region of the Deccan volcanic province have been previously postulated to be major eruptive vents for the associated lava pile. However, there have been few, if any systematic attempts prior to the present study to correlate these dikes to individual flows/formations that make up this lava pile. Similarly, published data for dike orientation and thickness on a local/sub-regional level are few, although these could illuminate several important issues such as the likelihood of this region having been an important eruptive zone and the nature of the prevalent lithospheric stress. This study reports detailed field and geochemical data for a swarm of basaltic dikes occurring around Sangamner. The data allow the comparison of the dikes to various units within the existing Deccan chemostratigraphy and the evaluation of their role as feeders. The geochemical characteristics of dikes combined with their widths, lengths and frequency indicate that many dikes in the Sangamner area could have fed flows of the Poladpur and/or Khandala formations. One dike is compositionally similar to the Bushe Fm., while one is similar to the Ambenali Fm. The implications of this are not entirely clear, but it could suggest that minor volumes of magma types other than the Khandala and Poladpur were being episodically supplied to this area. By analogy with studies of dikes and flows from other basaltic provinces (e.g. Swanson et al. 1975; Martin 1989; Reidel 2005), it appears that the Sangamner dike swarm would have had the potential to feed flood basalt flows of a significant extent. This study also sheds light on some potential complications that may arise when correlating intrusives from the DVP with established stratigraphic units. The implicit assumption in such a correlation is that the intrusives will display compositions that fall within the range defined by flows within the stratigraphy. This is, however, clearly not the case. Dikes around

Sangamner are very similar to certain units in the Deccan lava stratigraphy in their elemental composition, but some of them differ subtly or significantly from their chemically similar counterparts in their isotopic composition. Similar observations have also been reported in other studies attempting to place flows from other parts of the DVP within the established stratigraphic framework of the southwestern DVP (e.g., Peng et al. 1998; Mahoney et al. 2000). Studies that seek to evaluate geochemical data within a framework of physical volcanology are at their infancy in the DVP. As a result, products of individual eruptive episodes (flow- fields/volcanostratigraphic units) have not been unambiguously identified, complicating attempts to understand how magmatic activity within the province varied in space and in time. Focused physical volcanological and geochemical investigations of both lavas and dike swarms on a local/sub-regional scale, similar to the present one, have the potential to address several important questions regarding the magmatic evolution of the DVP.

ACKNOWLEDGEMENTS

We thank two anonymous reviewers and the editor for their helpful comments. We are grateful to Dr. Stan Mertzman for the XRF data and to Drs. John Morton and Darin Snyder for assistance with other geochemical analyses. Dr Vivek S. Kale, Gauri Dole, Vinit Phadnis, Raymond Duraiswami and Shizuko Watanabe are thanked for their assistance in the field at various stages of this project. Dr. John Mahoney kindly shared his unpublished geochemical data for various Deccan formations and provided comments on an earlier version of the manuscript. This work was partially supported by a Geological Society of America Research Grant, a Sigma-Xi Grant In Aid of Research and a Dept. of Science and Technology (India) research assistantship to Ninad Bondre. Hetu Sheth acknowledges an Industrial Research and Consultancy Centre (IRCC, IIT Bombay) research grant.

REFERENCES CITED

Beane, J. E.; Turner, C. A.; Hooper, P. R.; Subbarao, K. V.; and Walsh, J. N. 1986. Stratigraphy, composition and form of the Deccan Basalts, Western Ghats, India. Bull. Volcanol. 48:61-83.

Beane, J. E. 1988. Flow stratigraphy, chemical variation and petrogenesis of Deccan flood basalts from the Western Ghats, India. Ph. D Dissertation, Washington State University, Pullman.

Bhattacharji, S.; Chatterjee, N.; Wampler, J. M.; Nayak, P. N.; and Deshmukh, S. S. 1996. Indian intraplate and continental margin rifting, lithospheric extension, and mantle upwelling in Deccan Flood Basalt volcanism near the K/T boundary: Evidence from mafic dike swarms. J. Geol. 104:379-398.

Bondre, N. R. 1999. Geology of the area around Akole, Maharashtra (using remote sensing techniques). M. Sc Thesis, University of Pune, Pune, India.

Bondre, N. R.; Dole, G.; Phadnis, V. M.; Duraiswami, R. A.; and Kale, V. S. 2000. Inflated pahoehoe lavas from the Sangamner area of the western Deccan Volcanic Province. Current Science. 78:1,004-1,007

Bondre, N. R.; Duraiswami, R. A.; and Dole, G. 2004. Morphology and emplacement of flows from the Deccan Volcanic Province. Bull. Volcanol. 66:29-45.

Boyd, F.R.; Mertzman, S.A.1987. Composition and structure of the Kaapvaal lithosphere, southern Africa. In: Mysen, B. O., ed. Magmatic Processes - Physicochemical Principles. Geochemical Society Spec. Pub.1:13-24.

23 Chandrasekharam, D.; Mahoney, J. J.; Sheth, H. C.; and Duncan, R. A. 1999. Elemental and Nd–Sr–Pb isotope geochemistry of flows and dikes from the Tapi rift, Deccan flood basalt province, India. J. Volcanol. Geotherm. Res. 93:111–123.

Cox, K. G., and Hawkesworth, C. J. 1984. Relative contribution of crust and mantle to flood basalt magmatism, Mahabaleshwar area, Deccan Traps. Phil. Trans. R. Soc. Lond. A310:627-641.

Cox, K. G., and Hawkeshworth, C. J. 1985. Geochemical stratigraphy of the Deccan Traps at Mahabaleshwar, Western Ghats, India, with implications for open system magmatic processes. J. Petrol. 26:355-377.

Deshmukh, S. S., and Sehgal, M. N. 1988. Mafic dyke swarms in Deccan Volcanic Province of Madhya Pradesh and Maharashtra. In: Subbarao, K. V., ed. Deccan Flood Basalts. Geol. Soc. Ind. Mem. 10:323-340.

Hooper, P. R.1990. The timing of crustal extension and the eruption of continental flood basalts. Nature. 345:246-249.

Katoh, S.; Danhara, T.; Hart, W. K.; and WoldeGabriel, G. 1999. Use of sodium polytungstate solution in the purification of volcanic glass shards for bulk chemical analysis. Nature and Human Activities 4:45-54.

Khadri, S. F. R.; Subbarao, K. V.; Hooper, P. R.; and Walsh, J. N. 1988. Stratigraphy of Thakurvadi Formation, western Deccan Basalt Province, India. In: Subbarao, K. V., ed. Deccan Flood Basalts. Geol. Soc. Ind. Mem. 10:281-304.

Lightfoot, P. C.; Hawkesworth, C. J.; Devey, C. W.; Rogers, N. W.; and Van Calsteren, W. C. 1990. Source and differentiation of Deccan Trap lavas: Implications of geochemical and mineral chemical variations. J. Petrol. 31:1165-1200.

24 Martin, B. S. 1989. The Roza Member, Columbia River Basalt Group; Chemical stratigraphy and flow distribution. In: Reidel, S. P., and Hooper, P. R., eds. Volcanism and tectonism in the Columbia River Flood-Basalt Province. Boulder, CO, Geol. Soc. Am. Spec. Pap. 239:85-104.

Mahoney, J. J.; Macdougall, J. D.; Lugmair, G. W.; Gopalan, K.; and Krishnamurthy, P. 1985. Origin of contemporaneous tholeiitic and K-rich alkalic lavas: a case study from the northern Deccan plateau, India. Earth.Plan. Sci. Lett. 73:39–53.

Mahoney, J. J.; Sheth, H C.; Chandrasekharam, D.; and Peng, Z. X. 2000. Geochemistry of flood basalts of the Toranmal section, northern Deccan Traps, India: Implications for regional Deccan stratigraphy. J. Petrol. 41:1099-1120.

Melluso, L.; Sethna, S. F.; Morra, V.; Khateeb, A.; and Javeri, P. 1999. Petrology of the mafic dyke swarm of the Tapti River in the Nandurbar area (Deccan Volcanic Province). In: Subbarao, K. V., ed. Deccan Volcanic Province. Geol. Soc. Ind. Mem. 43(2):735-755.

Melluso, L.; Sethna, S. F.; D’Antonio, M.; Javeri, P; and Bennio, L. 2002. Geochemistry and petrogenesis of sodic and potassic mafic alkaline rocks in the Deccan volcanic province, Mumbai area (India). Mineral. Petrol. 74: 323-342.

Middlemost, E. A. K. 1989. Iron oxidation ratios, norms and the classification of volcanic rocks. Chem. Geol. 77:19-26.

Nakamura, N. 1974. Determination of REE, Ba, Fe, Mg, Na and K in carbonaceous and ordinary chondrites. Geochem. Cosmochem. Acta. 38:757-775.

Peng, Z. X.; Mahoney, J. J.; Hooper, P. R.; Harris, C.; and Beane, J. E. 1994. A role for lower continental crust in flood basalt genesis? Isotopic and incompatible element

25 study of the lower six formations of the western Deccan Traps. Geochem. Cosmochem. Acta. 58:267-288.

Peng, Z. X.; Mahoney, J. J.; Hooper, P. R.; Macdougall, J. D.; and Krishnamurthy, P. 1998. Basalts of the northeastern Deccan Traps, India: Isotopic and elemental geochemistry and relation to southwestern stratigraphy. J. Geophys. Res. 103:29,843- 29,865.

Pin, C., and Zalduegui, J. F. 1997. Sequential separation of light-rare-earth elements, thorium and uranium by miniaturized extraction chromatography: Application to isotopic analysis of silicate rocks. Anal. Chim. Acta. 339:79-89.

Sheth, H. C. 1998. Geochemistry, petrogenesis, stratigraphy and structure of Deccan flood basalts of the western Satpura–Tapi region, India. Ph.D Dissertation, Indian Institute of Technology, Bombay, India.

Sheth, H. C. 2000. The timing of crustal extension, diking, and eruption of the Deccan flood basalts. Int. Geol. Rev. 42: 1007-1016.

Sheth, H. C. 2005. Were the Deccan Flood Basalts derived in part froma ncient oceanic crust within the Indian Continental Lithosphere? Gond. Res 8:109-127.

Sheth, H. C.; Mahoney, J. J.; and Chandrasekharam, D. 2004. Geochemical stratigraphy of Deccan flood basalts of the Bijasan Ghat section, Satura Range, India. J. Asian Earth. Sci. 23:127-139.

Snyder, D. C. 2005. Processes and Time Scales of Differentiation in Silicic Magma Chambers: Chemical and Isotopic Investigations. Ph. D Dissertation, Miami University, Ohio.

26 Subbarao, K. V., and Hooper, P. R. 1988. Reconnaissance map of the Deccan Basalt Group in the Western Ghats. India. In: Subbarao, K. V., ed. Deccan Flood Basalts. Geol. Soc. Ind. Mem. 10: (enclosure).

Subbarao, K. V.; Dayal, A. M.; Gopalan, K.; Hooper, P. R.; and Walsh, J. N. 1999. Narmada dykes. In: Subbarao, K. V., ed. Deccan Volcanic Province. Geol. Soc. Ind. Mem. 43(2):891-902.

Sun, S. S., and McDonough, W. F. 1989. Chemical and isotopic systematics of oceanic basalts: Implications for mantle composition and processes. In: Saunders, A. D., and Norry, M. J., eds. Magmatism in ocean basins. Geol. Soc. Lond. Spec. Pub. 42:313- 345.

Swanson. D. A.; Wright, R. L; and Helz, R. T. 1975. Linear vent systems and estimated rates of magma production and eruption for the Yakima Basalt on the Columbia Plateau. Am. J. Sci. 275:877-905.

Tolan, T.; Reidel, S. P.; Beeson, M. H.; Anderson, J. L.; Fecht, K. R.; and Swanson, D. A. 1989. Revisions to the estimates of the areal extent and volume of the Columbia River Basalt Group. In: Reidel, S. P., and Hooper, P. R., eds. Volcanism and tectonism in the Columbia River Flood-Basalt Province. Boulder, CO, Geol. Soc. Am. Spec. Pap. 239:1-20.

Vanderkluysen, L.; Mahoney, J. J.; and Hooper, P. R. 2004. Implications for the emplacement of the Deccan Traps (India) from isotopic and elemental signatures of dikes. EOS Trans, AGU, Fall. Meet. Suppl. 45: V51B-0561.

Walker, G. P. L. 1993. Basaltic-volcano Systems. In: Pritchard, H. M., Alabaster, T., Harris, N. B. W., Neary, C. R. eds. Magmatic processes and plate tectonics. Geol. Soc.Spec Pub 76: 3-38.

27 Walker, R. J.; Carlson, R. W.; Shirey, S. B.; and Boyd, S. B. 1989. Os, Sr, Nd and Pb isotope systematics of southern African peridotite xenoliths: implications for the chemical evolution of subcontinental mantle. Geochem. Cosmochem. Acta. 53:1583- 1595.

Widdowson, M.; Pringle, M. S.; and Fernandez, O. A., 2000. A post K-T Boundary (early Palaeocene) age for Deccan-type feeder dykes, Goa, India. J. Petrol. 41: 1177-1194.

Wilson, S. A. 1998. Data compilation and statistical analysis of intralaboratory results for AGV-2. U.S. Geological Survey Open-File Report (in progress).

Table 1: Stratigraphic classification of the southwestern DVP (after Subbarao and Hooper 1988) and Sr isotopic values (at 66 Ma) for each of the constituent formations (after Sheth 2005)

87 86 Group Sub-group Formation Sr/ Sr(66Ma)

Wai Desur 0.7072-0.7080 Panhala 0.7046-0.7055 Mahabaleshwar 0.7040-0.7055 Ambenali 0.7038-0.7044 Poladpur 0.7053-0.7110

Deccan Basalt Lonavala Bushe 0.7078-0.7200 Khandala 0.7071-0.7124

Kalsubai Bhimashankar 0.7067-0.7076 Thakurvadi 0.7067–0.7112 Neral 0.7062–0.7104 Igatpuri-Jawhar 0.7085–0.7128

Table 2: Major oxide (wt.%) and trace element (ppm) data (excluding REE) for the Sangamner samples

CH1 Ch4b(flow) Ch4c Ch5 Ch6 Ch7a Ch8 Ch9 Ch10 Ch11 Ch12 Ch13

SiO2 49.62 49.78 49.02 49.53 50.11 49.51 49.46 49.94 49.58 49.32 50.23 50.35 TiO2 3.16 1.96 2.43 1.83 2.32 2.48 2.49 2.56 2.71 2.98 2.35 2.81 Al2O3 12.58 13.18 13.21 12.63 13.90 14.09 13.37 13.41 14.57 13.49 13.71 12.64 t Fe2O3 16.54 12.08 14.95 11.54 13.88 14.87 15.55 15.83 14.13 15.15 14.12 16.79 MnO 0.23 0.18 0.21 0.16 0.20 0.20 0.22 0.22 0.19 0.20 0.20 0.23 MgO 4.76 7.83 6.02 9.41 6.32 4.89 5.19 5.30 5.39 5.86 6.29 5.30 CaO 9.49 11.12 10.55 11.27 10.37 9.95 9.55 9.78 10.05 9.83 10.38 9.82

Na2O 2.52 1.98 2.29 1.83 2.40 2.62 2.40 2.39 2.69 2.54 2.43 2.47 K2O 0.48 0.45 0.20 0.41 0.43 0.41 0.36 0.36 0.41 0.45 0.38 0.40 P2O5 0.32 0.17 0.21 0.15 0.22 0.25 0.29 0.27 0.27 0.28 0.22 0.23 Total 100.39 98.91 99.59 99.54 100.31 99.58 98.88 100.06 100.36 100.57 100.62 101.05 LOI 0.69 0.17 0.51 0.78 0.16 0.31 nd nd 0.37 0.48 0.31 nd Mg# 40.2 60.2 48.5 65.6 51.6 43.5 43.7 43.5 47.1 47.5 51.0 42.5 30 Rb 35 10 10 11 17 21 nd nd 23 23 16 12 Sr 212 225 222 208 257 290 242 250 287 262 258 246 Y 44 25 32 23 30 32 35 35 35 38 31 37 Zr 228 116 144 105 148 173 153 170 189 201 150 179 V 353 338 360 306 339 318 391 489 322 326 322 389 Ni 50 145 74 228 101 58 59 69 93 101 93 53 Cr 51 381 92 557 241 73 56 59 182 198 221 49 Nb 15 8 10 7 10 14 nd nd 14 15 10 12 Ga 24 21 24 19 23 24 nd nd 26 24 23 23 Cu 259 155 230 140 182 440 216 280 442 483 187 657 Zn 141 91 121 82 111 120 115 107 113 119 110 125 Co 47 47 52 53 48 48 nd nd 44 48 47 52 Ba 146 68 62 87 113 164 142 141 176 207 136 159 Sc 29 37 30 34 31 29 35 38 27 30 30 32 Pb 6 5 4 4 5 4 nd nd 6 5 6 5

Ch14a Ch15 Ch16 Ch17 Ch18 Ch19 Ch20 Ch21 Ch22 Ch23 Ch24 Ch25

SiO2 50.20 49.88 52.01 52.31 51.16 51.45 49.82 49.67 49.28 49.94 49.39 50.14 TiO2 2.74 2.54 2.03 1.29 2.24 2.27 1.92 2.28 2.95 2.60 2.53 2.69 Al2O3 12.54 12.93 14.07 12.65 13.44 13.68 12.97 13.72 13.14 12.97 13.39 12.59 t Fe2O3 16.44 16.01 14.08 11.16 14.43 14.42 11.76 13.87 15.17 15.87 14.46 16.42 MnO 0.22 0.22 0.20 0.16 0.20 0.20 0.18 0.20 0.21 0.23 0.21 0.24 MgO 5.38 5.51 5.15 8.12 5.20 5.19 8.97 6.34 5.85 5.36 6.61 5.35 CaO 9.75 9.79 9.78 9.87 9.59 9.75 11.16 10.48 9.82 9.90 10.87 9.93

Na2O 2.44 2.44 2.68 2.18 2.69 2.72 1.93 2.37 2.61 2.53 2.41 2.53 K2O 0.33 0.27 0.39 0.99 0.56 0.54 0.38 0.30 0.43 0.36 0.35 0.39 P2O5 0.22 0.21 0.18 0.14 0.22 0.22 0.16 0.21 0.29 0.24 0.22 0.23 Total 100.33 100.01 101.09 99.67 99.90 100.76 99.89 99.88 100.14 100.21 100.25 100.63 LOI 0.07 0.21 0.52 0.80 0.17 0.31 0.66 0.44 0.40 0.21 -0.17 0.13 Mg# 43.3 44.6 47.9 64.6 45.7 45.7 64.0 51.6 47.4 44.1 51.6 43.2

31 Rb 19 20 25 32 30 29 12 15 21 16 10 17 Sr 250 250 251 269 246 245 215 263 265 247 222 239 Y 36 34 30 17 35 35 24 30 38 35 33 36 Zr 179 169 153 104 179 179 112 147 207 172 153 177 V 379 375 321 246 335 335 302 330 322 361 343 398 Ni 56 63 46 136 59 58 198 98 102 57 92 58 Cr 55 55 59 405 63 62 466 222 197 57 218 60 Nb 12 11 10 7 11 11 8 10 16 11 12 12 Ga 24 24 22 18 23 23 20 23 26 24 23 23 Cu 623 225 511 84 168 170 118 196 251 216 198 238 Zn 128 126 110 79 120 121 85 109 122 121 113 134 Co 50 49 46 54 50 51 48 47 47 50 50 53 Ba 119 107 146 225 201 175 81 117 194 152 86 103 Sc 30 30 28 27 30 31 31 28 26 30 32 34 Pb 5 6 6 7 7 7 5 4 7 5 4 5

Ch26a Ch27 Ch28 Ch30

SiO2 49.70 49.98 49.59 50.06 TiO2 3.20 3.27 3.12 2.67 Al2O3 11.93 12.18 12.19 12.49 t Fe2O3 17.22 17.32 17.02 16.55 MnO 0.24 0.24 0.24 0.23 MgO 4.82 4.86 4.94 5.35 CaO 9.26 9.18 9.41 9.89

Na2O 2.57 2.50 2.56 2.50 K2O 0.38 0.37 0.49 0.39 P2O5 0.31 0.40 0.30 0.23 Total 100.19 100.30 100.24 100.59 LOI 0.56 nd 0.40 0.22 Mg# 39.5 39.5 40.4 43.0

32 Rb 30 nd 32 14 Sr 207 203 210 239 Y 46 41 45 36 Zr 237 228 231 173 V 360 405 349 382 Ni 48 41 47 56 Cr 47 48 54 61 Nb 17 nd 16 12 Ga 24 nd 25 23 Cu 272 242 254 217 Zn 147 126 140 126 Co 46 nd 48 50 Ba 137 156 181 118 Sc 31 35 27 32 Pb 6 nd 6 6

t +2 Fe2O3 – Total iron as Fe2O3. Mg# = 100(Mg)/(Fe + Mg); ferrous to ferric iron ratio after Middlemost (1989); ‘nd’ is not determined.

Table 3: REE (ppm) and isotopic data for the Deccan Samples

87 86 143 144 Sample La Ce Nd Sm Eu Gd Dy Er Yb Lu Sr/ Sr(T) Nd/ Nd(T) εNd(T)

AGV-2 39.2 72.6 32.4 5.03 1.41 4.76 3.51 1.93 1.67 0.25 AGV-2 38.0 68.0 30.0 5.70 1.54 4.69 3.60 1.79 1.60 0.25 (E)† Ch1 21.6 50.7 30.9 7.47 2.32 8.41 8.34 4.78 3.91 0.57 0.70641 (0.70684) 0.512548 (0.512613) -0.1 Ch4b* 9.10 22.5 15.1 4.23 1.50 4.79 4.72 2.66 2.13 0.30 0.70552 (0.70563) 0.512640 (0.512716) +1.7 Ch4c 10.4 26.5 18.0 5.09 1.78 5.86 6.00 3.41 2.79 0.40 0.70451 (0.70463) 0.512707 (0.512784) +3.0 Ch6 14.5 34.2 22.3 5.59 1.96 6.15 6.07 3.41 2.69 0.39 0.70627 (0.70644) 0.512511 (0.512579) -0.8 Ch7a 16.7 39.4 23.4 5.5 1.82 5.73 5.50 3.10 2.45 0.35 Ch9 14.4 35.7 23.5 5.93 2.05 6.85 6.50 3.72 2.98 0.42 Ch10 18.5 42.6 25.7 6.18 2.02 6.81 6.35 3.51 2.87 0.40 Ch11 21.1 48.9 29.7 7.03 2.29 7.66 7.42 4.19 3.30 0.47 0.70644 (0.70687) 0.512497 (0.512561) -1.1 Ch13 16.2 39.7 26.2 6.66 2.29 7.46 7.57 4.23 3.35 0.48 0.70609 (0.70621) 0.512599 (0.512668) +0.9 Ch16 16.4 37.5 21.4 5.24 1.71 5.76 5.61 3.19 2.65 0.38 0.70851 (0.70877) 0.512318 (0.512385) -4.5 33 Ch17 14.1 30.9 15.7 3.51 1.17 3.57 3.36 1.99 1.54 0.22 0.71162 (0.71193) 0.511862 (0.511923) -13.4 Ch18 21.2 46.9 27.1 6.35 2.08 7.00 6.95 4.00 3.18 0.46 0.70938 (0.70971) 0.512231 (0.512295) -6.2 Ch20 9.20 22.1 14.6 4.03 1.48 4.63 4.54 2.57 2.02 0.29 0.70541 (0.70555) 0.512626 (0.512701) +1.5 Ch24 10.8 27.0 18.3 4.89 1.73 5.69 5.69 3.18 2.62 0.37 Ch25 16.1 39.6 26.1 6.65 2.25 7.32 7.41 4.19 3.27 0.47

87 86 143 144 Sr/ Sr(T), Nd/ Nd(T) and εNd(T) are age-corrected values at 66 Ma based on whole-rock Rb, Sr ,Sm and Nd concentrations. Measured values for 87Sr/86Sr and 143Nd/144Nd are in parentheses. † - Expected values for USGS standard AGV-2 (Wilson 1998); * - flow sample

Table 4: Results of two separate runs of the Discriminant Function Analysis discussed in the text. (Ig-J – Igatpuri-Jawhar; Tha – Thakurvadi; Bhi – Bhimashankar; Kha – Khandala; Pol – Poladpur; N – Neral; p – probability; M-dist – Mahalanobis distance). Samples with question marks are those with a Mahalanobis distance >24, approaching a conditional probability of zero.

Sample Best p M- Func. Func. 2 Best p M- Func. 1 Func. 2 Fm. dist. 1 Fm. dist. match match Run 1 Run 2 Ch1 Ig-J 0.113 13.0 -0.125 -3.514 Kha(?) 0.000 30.2 -2.792 2.147 Ch4b Tha 0.851 4.07 0.899 0.510 Tha 0.967 2.39 0.531 -0.484 Ch4c Bhi 0.376 8.61 0.758 0.017 Pol 0.545 6.92 -0.584 -0.756 Ch5 Tha 0.414 8.20 1.261 -0.247 Tha 0.551 6.86 0.072 -0.039 Ch6 Bhi 0.771 4.87 1.034 -0.821 Bhi 0.888 3.65 -0.427 0.441 Ch7a Ig-J 0.529 7.07 1.328 -1.396 Bhi 0.679 5.71 -0.535 0.985 Ch10 Ig-J 0.112 13.0 1.193 -2.499 Kha 0.094 13.5 -1.269 1.989 Ch11 Ig-J 0.251 10.2 1.678 -3.321 Kha 0.297 9.56 -1.454 2.706 Ch12 Bhi 0.458 7.75 1.006 -1.659 Bhi 0.615 6.29 -1.213 0.758 Ch13 Ig-J 0.293 9.62 -0.157 -1.539 Kha 0.828 4.32 -0.818 1.773 Ch14a Ig-J 0.040 16.2 -0.734 -1.108 Kha 0.291 9.64 -0.894 1.072 Ch15 Ig-J 0.067 14.6 -0.570 -0.676 Bhi 0.483 7.51 -0.618 0.664 Ch16 Ig-J 0.010 20.0 -2.397 -1.136 Bhi 0.238 10.4 -1.829 0.435 Ch17 Tha 0.019 18.3 0.268 -2.457 Ner 0.077 14.2 -1.781 0.892 Ch18 Ig-J 0.014 19.2 -1.866 -2.649 Kha 0.472 7.61 -2.264 2.119 Ch19 Ig-J 0.021 18.0 -1.988 -2.373 Kha 0.350 8.91 -2.132 1.882 Ch20 Tha 0.686 5.65 0.869 -0.311 Tha 0.532 7.04 -0.352 -0.364 Ch21 Bhi 0.327 9.18 1.030 -0.963 Bhi 0.528 7.08 -0.719 0.335 Ch22 Ig-J 0.101 13.3 1.160 -3.373 Kha 0.021 18.0 -1.788 2.643 Ch23 Ig-J 0.343 9.00 0.352 -1.913 Kha 0.605 6.38 -1.351 1.423 Ch24 Ig-J 0.957 2.60 0.772 -0.316 Bhi 0.849 4.09 -0.452 -0.273 Ch25 Ig-J 0.213 10.8 -0.495 -0.467 Pol 0.583 6.57 -0.181 0.955 Ch26a Ig-J 0.023 17.7 -1.313 -3.059 Kha(?) 0.000 31.1 -2.470 2.208 Ch28 Ig-J(?) 0.002 24.7 -0.997 -3.877 Kha(?) 0.000 28.5 -3.173 2.575 Ch30 Ig-J 0.401 8.34 -0.015 -0.904 Kha 0.663 5.86 -0.573 0.959

Figure 1: Map depicting some important features of the Deccan Volcanic Province (DVP). The areas dominated by “simple” and “compound pahoehoe” lava flows and the principal areas of dike concentration are shown. The study area is marked with the white box. NTDS – Narmada- Tapi Dike Swarm; WCDS - West Coast dike swarm, WGE – Western Ghats Escarpment. The Narmada-Tapi region is an ancient zone of weakness that has been reactivated at various times in the geologic past.

Figure 1 (Bondre et al. 2006)

73 0 E 79 0 E

Malwa Plateau 24 0 N Kutch be Lo cture la Narmada Stru nd Ma Saurashtra NTDS

Dhule Tapi Rift Nasik 0 Sangamner 20 N Dikes Mumbai Pune W

C Mahabaleshwar Dominantly Simple D S W

G E 0 200 Dominantly km Compound Pahoehoe

Figure 2: Geologic map of the study area showing the distribution of dikes identified during the present study. Dashed lines indicate dikes/dike extensions inferred from satellite imagery and topographic data. Numbers next to dikes refer to sample numbers discussed in the text. Dike dips are shown where significant. The extent of various chemostratigraphic units is from Subbarao and Hooper (1988) and Khadri et al. (1988). The inset (upper left corner of the map) is a rosette diagram of dike orientations. N % refers to the number of dikes trending in a given direction.

Figure 2 (Bondre et al. 2006)

0 0/ 0/ 30 % 74 00 Dapur 74 15

1 h2 10 Ch22 C 270 90

N % 0 2 946 m h C 9 h1 Devthan C Sangvi 8 h1 M C a h a lu 670 m n g 15 2 h A i 1 C a d R h 14 u . C Ch la 3 R h1 C 780 m Ch17 . 0 a 1 h7 Ch C 9 Ch16 Ch 70 11 6 0 Ch h Sangamner Rajur 5 C (560 m) 8 Akole Ch Ch5 Pravara R. Vitha 0/ 19 30 Chandanapuri 1 Ch 860 m Mula R. Ch24 Ch4c Ch25 23 Ch27 Ch Kotul Ch26a Ch28

805 m Dolasne

Ch30

M Brahmanwada an dvi 0 4 R. Ghargaon km

Village/town Roads Location of potential feeder relationship Quaternary Alluvium & colluvium (Pravara Fm)

0 Dikes 5 Upper Thakurvadi Fm.

Bhimashankar Fm. Lower Thakurvadi Fm.

Middle Thakurvadi Fm. 38

Figure 3: Plots of (a) Nb/Zr vs. Ba/Y, (b) Sr vs. Nb/Zr and (c) TiO2 vs. Zr/Y for the Sangamner dike samples. Fields for relevant southwestern Deccan formations are from Sheth et al. (2004) and Peng et al. (1998), based on data from Beane et al. (1986) and Beane (1988). ‘A’ - Ambenali Fm; ‘B’ - Bushe Fm; DD – Dhak Dongar member of the Khandala Fm; K, K1, K2 – Khandala Fm; P – Poladpur Fm; R – Rajmachi member of the Khandala Fm. Note the significant overlap between the fields for many of the formations.

Figure 3 (Bondre et al. 2006)

0.14

Po 0.12 M

r 0.10 A Z / b N 0.08 K

0.06 B

0.04 0 5 10 15 20 Ba/Y

400

350 P B 300 M r

S 250 A 200 K 150

100 0.04 0.06 0.08 0.10 0.12 0.14 0.16 Nb/Zr

DD 3.5 A K

2 2.5 O

i R T P 1.5 K1 B K2 0.5 2 3 4 5 6 7 Zr/Y

Ch1, Ch26a, Ch27, Ch28 Ch4c, Ch8, Ch24 Ch5, Ch20 Ch6, Ch12, Ch21 Ch7, Ch10 Ch11, Ch22 Ch9, Ch13, Ch14a, Ch15, Ch16 Ch23, Ch25, Ch30 Ch17 Ch18, Ch19

Figure 4: Values of the first two canonical discriminant functions for the Sangamner samples, with fields and centroids of their closest southwestern Deccan formations for the two runs

discussed in the text. ‘A’ shows results of the first run (table 4), where function 1 = – 0.492SiO2

– 0.147Al2O3 + 1.387TiO2 + 0.167CaO + 0.107MgO + 1.375P2O5 – 0.116Ni + 0.042Sc – 0.093V + 0.464Ba + 0.180Rb + 0.314Sr – 1.680Zr – 0.747Y – 0.015Nb; and Function 2 =

0.120SiO2 + 0.029Al2O3 – 0.385TiO2 + 0.048CaO – 0.289MgO – 0.315P2O5 + 0.268Ni + 0.035Sc + 0.764V – 0.804Ba + 0.047Rb + 0.235Sr – 0.945Zr + 0.069Y + 1.174Nb. ‘B’ shows results of the second run (table 4), where Function 1 = – 0.294Ba + 0.963Nb + 0.335Ni – 0.045Rb + 0.188Sc + 0.473Sr + 0.854V – 0.619Y – 0.367Zr. Function 2 = 0.798Ba – 0.715Nb + 0.324Ni – 0.277Rb + 0.100Sc + 0.048Sr + 0.086V – 0.436Y + 1.524Zr. The elemental abundances in these equations are Z-score-standardized values (see Sheth et al. 2004 and references therein). Formation centroids (shown by black circles) for the first run have the following abbreviations and function scores: Igatpuri-Jawhar (IJ) 1.165, –1.160; Thakurvadi (Th) 0.947, 0.356; Bhimashankar (Bh) 1.018, 0.342. For the second run, they are: Thakurvadi (Th) 0.746, –0.274; Bhimashankar (Bh) 0.657, –0.081; Khandala (Kh) –0.942, 2.072; Poladpur (Po) 0.580, –0.508.

Figure 4 (Bondre et al. 2006)

Sangamner samples Bhimashankar A Igatpuri-Jawhar Fm. centroids Thakurvadi

2.5

1.5 2

n Ch4b o i 0.5 Th Bh t c n -0.5 u

F Ig-J -1.5

-2.5

-3.5

-4.5 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 Function 1

Sangamner samples Khandala B Thakurvadi Poladpur Bhimashankar Fm. centroids

5.5

4.5 2 3.5 n o i t 2.5 K c n u 1.5 F 0.5 Bh -0.5 Ch4b Th Po -1.5

-2.5 -5.0 -4.0 -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0 Function 1

Figure 5: Comparison of multi-element patterns for some Sangamner samples with those of selected formations/members of the Deccan stratigraphy. Concentrations have been normalized to the primitive mantle (after Sun & McDonough 1989).

Figure 5 (Bondre et al. 2006)

100 100 Kusgaon Rajmachi Ch13 Ch18 Ch25

10 10 A D

Rb Ba Nb K La Ce Pb Sr P Nd Zr Sm Ti Y Rb Ba Nb K La Ce Pb Sr P Nd Zr Sm Ti Y n o i t 100 100 a r

t Bushe CT Poladpur n Boyhare Ch4b e c Ch17 Ch20 n o Ch4c c d e z i l a m r

o 10 10 n - e l B E t n a m e

v Rb Ba Nb K La Ce Pb Sr P Nd Zr Sm Ti Y Rb Ba Nb K La Ce Pb Sr P Nd Zr Sm Ti Y i t i m

i 100 100 r P Dhak Dongar Dhak Dongar Ch11 Ch10 Ch1 Ch7a

10 10 C F

Rb Ba Nb K La Ce Pb Sr P Nd Zr Sm Ti Y Rb Ba Nb K La Ce Pb Sr P Nd Zr Sm Ti Y

Figure 6: Comparison of REE patterns for some Sangamner samples with those of selected formations/members of the Deccan stratigraphy. Concentrations have been normalized to chondritic values (after Nakamura 1974). REE data for the stratigraphic units are from Peng et al. (1994) and references therein.

Figure 6 (Bondre et al. 2006)

Ch6 100 Ch16 Bhimashankar

10 A

Boyhare 100 Bushe CT Ch17 n o i t a r B t 10 n e c n o c d e z i l 1 a m r Rajmachi o n Ch18 - e 100 t i r d n o h C

10 C

Dhak Dongar Ch10 100 Ch7a Bhimashankar

10 D

La Ce Nd Sm Eu Gd Dy Er Yb Lu

Figure 7: Sr and Nd isotopic data for the Sangamner samples. Also shown are fields for the southwestern Deccan formations (after Peng et al. 1994, and references therein). DD – Dhak Dongar Member; RAJ – Rajmachi Member.

Figure 7 (Bondre et al. 2006)

+10 Central Indian Ridge Ch1 Ch4c Ch20 Ch11 Ch4b Boyhare Ambenali Ch13 Ch6 DD Poladpur Ch16 Ch18 RAJ 0 Mahabaleshwar Igatpuri-Jawhar d N Bhimashankar -10

Neral Bushe

-20 Khandala

0.700 0.704 0.708 0.712 0.716 0.720 87Sr 86Sr

48 CHAPTER 3

MORPHOLOGICAL AND TEXTURAL DIVERSITY OF THE STEENS BASALT LAVA FLOWS, SOUTHEASTERN OREGON, USA: IMPLICATIONS FOR EMPLACEMENT STYLE AND NATURE OF ERUPTIVE EPISODES

N. R. Bondre, W. K. Hart Department of Geology, Miami University, Oxford, Ohio-45056, USA

(To be submitted to the Bulletin of Volcanology)

Abstract This study focuses on mid-Miocene tholeiitic flood basalt lava flows from the Oregon Plateau, northwestern USA (Steens Basalt) and is the first to document and evaluate their morphology from a modern perspective. Field observations of flows from several sections within and proximal to the main exposures at Steens Mountain have been supplemented with textural and geochemical data, and are used to offer preliminary insights into their emplacement. Compound pahoehoe flows of variable thickness appear to be common throughout the study area, laterally and vertically. These tend to be plagioclase phyric and the morphology and disposition of constituent flow lobes are quite similar to those from other provinces. Classic a’a flows with brecciated upper and basal crusts are not abundant, but by no means uncommon. Flows with characters different from typical pahoehoe and a’a are also commonly encountered. Such flows display a range in morphology; flows with reasonably preserved upper crusts but brecciated basal crusts, as well as those displaying well-developed flow-top breccias and preserved basal crusts (rubbly pahoehoe) are observed. The Steens Basalt appears to display greater morphological and textural diversity at the outcrop scale than that described for some other flood basalt provinces. The abundant compound pahoehoe flows (often rich in plagioclase phenocrysts) were likely emplaced during slow but sustained eruptive episodes; their constituent lobes show clear evidence for endogenous growth. The relatively aphyric flows with brecciated surfaces (including a’a) hint at higher strain rates and/or higher viscosity, probably caused by higher effusion rates. A couple of sections

49 are characterised by compositionally similar, but morphologically different flows that were possibly part of the same eruption. This suggests that certain physical parameters changed substantially and abruptly during eruption, and were perhaps accompanied by differentiation processes within the plumbing system. It is likely that such observations indicate temporal fluctuations within complex magmatic and eruptive systems, and deserve closer scrutiny.

Keywords Steens Basalt, lava flows, morphology, emplacement, pahoehoe, a’a

Introduction

Once considered to be monotonous stacks of basaltic lava, Continental Flood Basalt (CFB) provinces are now known to display considerable diversity in lava flow morphology. Whereas most initial studies of flood basalt morphology and emplacement focused on younger provinces such as the Columbia River Basalt (e.g., Shaw and Swanson 1970; Long and Wood 1986; Self et al. 1996; Reidel 1998; Thordarson and Self 1998; Walker et al. 1999), subsequent investigations have also targeted older provinces such as the Deccan Volcanic Province (DVP; e.g., Keszthelyi et al. 1999; Duraiswami et al. 2001, 2003; Bondre et al. 2004a). It is becoming increasingly clear that every CFB province is unique in terms of its tectono-magmatic evolution as well as the types of lava flows and their proportions. Bondre et al. (2004a, b) have cautioned against using only the Columbia River Basalt province as an analogue for all CFB provinces. They have stressed the need to document the morphology of lava flows from each individual province in order to obtain insights into their emplacement and eruptive histories. An important reason for documenting flow morphology in older provinces is to uncover lava types and modes of emplacement that are not observed in younger and active volcanic provinces. Indeed, recent studies from CFB provinces have led to the recognition of distinct lava types that have few young analogues (e.g. rubbly pahoehoe; Keszthelyi et al. 2006). The present study focuses on mid-Miocene tholeiitic flood basalt volcanism on the Oregon Plateau (Steens Basalt) and documents the morphology and textural

50 characters of constituent lava flows. Although this basaltic volcanism has been the subject of numerous detailed geochemical, geochronological and geomagnetic investigations (e.g. Gunn and Watkins 1970; Hart and Carlson 1985; Carlson and Hart 1988; Hart et al. 1989; Johnson et al. 1998; Hooper et al. 2002; Camp et al. 2003), no systematic investigation of flow morphology from a modern perspective had hitherto been conducted. Yet, such information in conjunction with geochemical and textural data can prove to be very valuable in understanding the overall eruptive history and in delineating products of single eruptive episodes. It is hoped that this study will provide a foundation on which future studies that integrate morphological and geochemical aspects can build.

The Steens Basalt

The extrusive products of a major regional northwestern United States mid-Miocene magmatic event are preserved primarily as mafic flood lavas of the well-known Columbia River Basalt province, the Oregon Plateau to the south, and the intervening Malheur Gorge region of southeastern Oregon. These products include the Steens Basalt, which is here defined as the main sequence of mid-Miocene tholeiitic flood basalt lava flows found on the Oregon Plateau. In this definition we include the “Steens Mountain Basalt” of Fuller (1931); the “Pueblo Mountains basalt” of Avent (1970) and Hart et al. (1989); the “Steens-type basalt” of Carlson and Hart (1985); and the “upper Steens basalt” and “lower Steens basalt” of Camp et al. (2003). The Steens Basalt exposed at and proximal to Steens Mountain, which is the focus of this study, is conservatively estimated to have been emplaced between 16.58 ± 0.2 to 15.5 ± 0.3 Ma (Brueseke et al. in review). Furthermore, following the discussion of volcano types by Walker (1993), we characterize the Steens Basalt as a moderate-sized flood-basalt field. Walker describes such fields as consisting of monogenetic volcanoes erupted from widely scattered vents. The constituent lava flows cover wider areas than monogenetic volcano fields, overlap or are superposed to form parallel-stratified successions, and have much greater volumes. Giant flood-basalt fields such as the Deccan and the Columbia River Basalt often receive considerable attention. However, Walker (1993) cites many moderate to small-sized

51 flood-basalt fields occur around the world, including Harrat Rahat in Saudi Arabia (Camp and Roobol 1989) and the McBride Province in Queensland. The thickest exposures of the Steens Basalt succession (~ 1 km) intruded by numerous feeder dikes are observed along the Steens-Pueblo fault scarp (Fig. 1; inset). This NNE-SSW scarp extends for almost 130 km (e.g., Hart et al. 1989) and is related to younger Basin and Range extension in southeastern Oregon. Steens Basalts along with the nearby basalts of the Malheur Gorge have a volume of approximately 60,000 km3 (Carlson and Hart 1988). The area covered by flows emanating from the Steens Mountain area could have been over 20,000 km2 (Hart and Carlson 1985). Spectacular exposures of thick stacks of flood basalt lava flows occur along glacially carved canyons in the Steens and Pueblo Mountains areas. Excellent (and more accessible) exposures also occur along the Catlow Rim, west of the main eastern scarp. Geochemically and temporally similar flows occur in several other locations in the Idaho-Oregon-Nevada tri-state region including the western Owhyee Mountains (Shoemaker and Hart 2004; Hart and Carlson 1985), the Santa Rosa-Calico Volcanic field (Brueseke et al. in review; Brueseke and Hart, in press) and the Pine Forest Range (Colgan et al. 2006).

Previous work

Some of the earliest descriptions of the morphology and textures of the Steens Basalt flows are found in Fuller (1930, 1931). The textural descriptions are particularly detailed. Fuller reports flows averaging around 3 m in thickness with vesicular tops and pipe vesicle-bearing bases, and highlights the fact that contacts of successive flows often merge when traced laterally. Terminations of flow lobes are clearly visible in a photograph in Fuller (1930; Fig. 79). In the modern context, such flows would be referred to as compound. Flows with scoriaceous and brecciated surfaces are also described, as are geographic variations in flow morphology. Fuller comments on the rarity of flows with columnar jointing in contrast to flows from the Columbia River Basalt province, and attributes the morphological features of Steens flows to their greater fluidity (higher volatile content). He describes the texture of most Steens flows as open or diktytaxitic; this latter term that he proposed has been widely applied to basaltic lavas

52 from around the world. Thin sheets of lava associated with a dike at the base of the Mosquito Creek cirque are postulated by him to be near-vent features. Numerous examples of segregation structures within the lava flows are described, as are marked variations in phenocryst content within a single flow. Detailed descriptions of Steens Basalt flow textures are also provided by Avent (1970). He found diktytaxitic textures to be much less common than Fuller (1931) had suggested; instead, he mentioned that flows often weather to a black mottled appearance, which in thin section can be attributed to ophi-mottling (poikilomosaic texture; Shelly 1993), with clots of pyroxene up to 5 mm in diameter. Most flows are thin (1.5 - 3 m) and holocrystalline; highly porphyritic as well as fine-grained varieties are common. Gunn and Watkins (1970) and Johnson et al. (1998) have briefly described flow characters and textures of flows from sections along the main eastern scarp of Steens Mountain. Brueseke et al. (in review) and Bondre et al. (2004b) have commented on the commonly compound nature of Steens flows, and on the observation that many flow surfaces appear to be close to brecciation. The former authors suggest that many flows might consist of inflated pahoehoe lobes and they also identify several flows capped by breccias or rubble.

The Steens Basalt lava flows

The intent of this study is to provide the first detailed, modern perspective on Steens Basalt flow morphology and textures. In order to acquire information with spatial and chronostratigraphic context, detailed observations of flow morphology were recorded from numerous small sections from widely separated areas and throughout the vertical extent of the Steens Basalt exposed at and proximal to Steens Mountain. Certain regions were better represented than others during this study, primarily due to reasons of accessibility and exposure. At every location, the type of lava (e.g., pahoehoe), compound versus simple nature, textural characteristics, thickness, and disposition of flows/lobes were among the observations that were recorded. In case of compound flows, the nature of individual lobes and their internal structure were also recorded. Quantification of vesicle distribution was not attempted as it was outside the scope of this

53 study; however, qualitative observations pertaining to it were made. Similarly, textures have been qualitatively discussed and no quantification has been attempted. Figure 1 shows the locations where detailed field observations and measurements were recorded and where samples were collected. These sections are placed in a general chronostratigraphic context within the Steens Basalt sequence based on previous stratigraphic, geochemical, and Ar-Ar data (Hart et al. 1989; Brueseke and Hart 2000; Brueseke et al. in review and references therein) and on previously established chemostratigraphy (Gunn and Watkins 1970; Hart et al. 1989; Johnson et al. 1998; Hooper et al. 2002; Camp et al. 2003). This information is summarised in Table 1. The various classification criteria do not necessarily agree with each other. Geochemical data for the ‘upper’ and ‘lower’ Steens basalts often overlap each other (Fig. 2) and it is not always possible to unambiguously classify a sample using these. For example, samples N13/04 and N14/04 (Table 2, Fig. 2) were collected close to the summit of Steens Mountain and would be classified as belonging to the upper Steens basalt (Johnson et al. 1998; Camp et al. 2003) based on their stratigraphic position. However, these two samples are geochemically quite distinct, with only N14/04 plotting within the field for the upper Steens in various binary plots (Fig. 2). N13/04, on the other hand, plots within the region of overlap between the two fields (Fig. 2). In such cases, stratigraphic position and radiometric ages (where available) have been given priority in classifying the sections. Where neither is well-constrained, geochemical criteria have been used. Also, since many of the locations included in the present study are away from the well-studied Steens Mountain section, and since the region is extensively faulted, their stratigraphic context should be taken only as an approximation. The informal term ‘middle’ Steens is used in this study for those locations which appear to occupy a position intermediate between the upper and lower Steens. Selected flows were sampled for petrographic studies as well as to obtain compositional data. Textural observations were made in the field as well as in the laboratory using an optical microscope. The samples were analysed for major elements at Miami University using the methods described in Katoh et al. (1999). Trace element data were obtained at Franklin and Marshall College using methods outlined by Mertzman (2000). Geochemical data are presented in Table 2; note that the various plots used in this

54 paper are based on data that have been normalised to 100% anhydrous. Logs for selected sections have been constructed using measured thicknesses of units and field observations, and are shown in Figure 3. This figure reveals that there is considerable diversity in flow morphology between and within the measured sections. Compound pahoehoe flows are found in most sections (e.g., Fig. 4a), whereas a’a flows are less common. Many sections are characterised by transitional flows or flows that do not easily fall into the established categories of pahoehoe or a’a. The general characteristics of each of these principal morphological types will be described in detail below with reference to specific sections and locations.

Compound pahoehoe flows

Flows consisting of multiple, superposed lobes of pahoehoe lava are termed compound pahoehoe flows (Walker, 1971). Such flows are common in the Columbia River Basalt province and abundant in the DVP (e.g., Walker 1971; Deshmukh 1988; Bondre et al. 2000). Steens Basalt flow lobes also show classic pahoehoe features. Their scale and disposition are very similar to those in the Snake River Plain and Owyhee Plateau regions (Bondre and Hart, unpub data), Hawaii and the DVP (Bondre et al. 2004a). Lobes in the Columbia River Basalt province are often much thicker and significantly more extensive (e.g., Thordarson and Self 1998). The Steens lobes are characterised by a three-part structure common to pahoehoe lobes around the world, with a vesicular upper crust, a central vesicle poor core, and a vesicular basal crust. Many of the Steens Basalt lobes display distinct chilled rinds, often bright red or orange in color, and ropy structure is quite common. The individual lobes measured during this study range in thickness from approximately 0.5 m - 5 m and can be laterally traced up to a few meters to tens of meters in any given outcrop. Basal crusts are usually 0.1-0.3 m in thickness, with cores and upper crusts around 1-2 m each. The Steens pahoehoe flows tend to be plagioclase phyric, sometimes with considerable proportions of mega-phenocrysts (> 2 cm) of plagioclase. During the present study, the best exposures of such flows were observed along the Steens Loop Road, as well as along the Catlow Rim (Fig. 1). Details of the nature of these flows are presented below with the help of specific examples.

55 Location 11 (close to the Jackman campground; Fig. 1) exposes multiple, overlapping lobes (Fig. 5a) that contain plagioclase mega-phenocrysts. Most lobes show well-preserved, vesicular basal crusts. In some lobes, these show a banding of vesicles with vesicle size increasing towards the top (Fig. 6a). Pipe vesicles are absent; however, all other characteristics of these lobes are similar to the P-type lobes described by Wilmoth and Walker (1993) in Hawaii as well as P-type lobes from the DVP (Bondre et al. 2004a). The cores are poorly vesicular and contain bands or zones, some of which are relatively fine grained, while others contain plagioclase mega-phenocrysts. This is probably a result of flow-induced segregation, similar to that described by Fuller (1930). Upper crusts are moderately to highly vesicular, often showing vesicle banding, with vesicle size increasing with depth into the flow lobe. Plagioclase mega-phenocrysts in the upper crusts of some lobes in this section tend to be fairly randomly arranged and do not show preferred alignment (Fig. 6a). On the other hand, the cores tend to have zones of mega-phenocrysts, which show preferred alignment, usually with the long-axes parallel to the lobe margins (Fig. 6a). Cores also sometimes show the presence of segregation veins and vesicle cylinders, although it is hard to discern if these are filled with segregated material or are simply trails of vesicles. The basal crusts are sometimes devoid of phenocrysts, which could be because of floatation of these to the middle or upper parts of flow lobes. Fuller (1930) has described bases rich in phenocrysts and attributed that to their gravitational settling. This suggests that flotation as well as sinking of phenocryts occurred during emplacement, depending on the interplay of phenocryst size and composition, and lava viscosity and composition. Very similar features are also observed at Locations 12 and 14 (Fig. 1, 3; Table 1). Fairly extensive lobes (Fig. 4b) are observed at Location 14; cores in this section are rich in segregation structures and show multiple levels of vesicle cylinders and segregation veins. Bases of the thicker lobes in this section show the development of many small toes/lobes, and some of these lobes show a radial arrangement of plagioclase mega-phenocrysts (Fig. 6b). The bases also have well-developed ropy surfaces and red/orange, chilled rinds. Classic compound pahoehoe flows are also found along the Catlow Rim (e.g., Location 5; Fig. 1, 3). Some flows here differ from those previously described in that their constituting lobes have well-developed, open pipe vesicles in their basal crusts (Fig.

56 5b). Most lobes show chilled rinds as well as ropy surfaces. Some lobes are relatively fine grained, while others are moderately plagioclase phyric. An examination of a thin section of the core of one such lobe reveals a relatively holocrystalline, diktytaxitic texture, indicating slower cooling as compared to the upper and basal crusts (Fig. 7a). Some small, uniformly vesicular S-type lobes (Wilmoth and Walker, 1993) are also noted in these flows (Fig. 5b). A relatively thick section measured along a canyon cutting into the Catlow Rim (Location 6; Fig. 1, 3) reveals that some Steens Basalt pahoehoe flows reach significant thicknesses. This section documents part of what appears to be a single, thick, compound pahoehoe flow. No discernible flow boundaries were observed. Basal parts of the section are dominated by mega-plagioclase phyric lobes, and excellent features such as abutting lobes and smaller lobes filling gaps between larger ones are observed. The upper parts contain finer grained or aphyric lobes. Pipe vesicles are rare. The two uppermost lobes in the measured section do not have well-defined, massive cores. Instead, there are numerous bands of small vesicles alternating with relatively massive ones. Each massive- vesicular pair might represent a pulse of lava injected into the inflating lobe (similar to the process described by Walker et al. 1999). Observations at other locations along the Catlow Rim also indicate that some Steens compound pahoehoe flows attain significant thicknesses (e.g., Fig. 4a). Location 8 along the Catlow Rim (Fig. 1, 3) displays a package of what appear to be thin sheets of clastogenic lava. This exposure is easily visible from the road owing to its distinctive layered appearance. The individual sheets/lobes are quite thin, and have an average thickness of 0.37 m and a mode of 0.13 m. Each unit is mildly plagioclase phyric (with phenocrysts up to 2 mm in size), very vesicular throughout and is laterally discontinuous. The vesicles are near-spherical, 1-2 mm across. Contacts between different units are welded to varying degrees and thin scoriaceous layers are observed at multiple levels within the measured sections. Near the upper part of one of the measured sections (not shown), thicker lobes (up to 9 m) with slightly brecciated surfaces are observed. These lobes possess a complex internal structure composed of thinner units whose margins have been annealed. They also contain altered olivine phenocrysts in addition to plagioclase, and traces of segregated material. All these characteristics

57 indicate proximity to an eruptive locus with thinner clastogenic lava, at places transitioning to thicker, welded agglutinate. These observations are strong evidence for this section being part of or proximal to an eruptive vent (large scoria cone / small shield); similar features were observed at the Mosquito Creek cirque (Fuller, 1931; Fig. 76) and by Brueseke et al. (in review) at the Mickey Butte locality on the northeast side of the Alvord desert (Fig. 1).

A’a flows

Typical a’a flows are characterised by brecciated or clinkery upper and lower crusts, separated by massive, poorly vesicular cores (e.g., Macdonald 1972; Peterson and Tilling 1980; Rowland and Walker 1990). The cores sometimes contain entrains of breccia and highly deformed vesicles. Although not immediately apparent, compound a’a flows are not uncommon. The best candidates for Steens Basalt a’a flows observed during this study are located along the Catlow Rim, although flows from the Mickey Butte section are also most likely a’a flows. At Location 7 along the Catlow Rim (Fig. 1, 3), a’a flows separated by breccia zones are clearly visible from the road, even from a distance. This package of flows appears to be overlain and underlain by compound pahoehoe flows. Three of the a’a flows were measured and described in detail and two were sampled for geochemistry. These flows are mostly aphyric; each flow has a fairly undulating surface with thick flow-top and flow-base breccias (Fig. 8a, 9a). These flows are thicker (6-10 m) than the average pahoehoe flow lobes described earlier. The breccias are also quite thick, averaging 2 m. The Location 7 a’a flows contain sub-rounded to platy, highly vesicular and oxidised fragments that appear to have been derived from previously formed crust (Fig.8b). Many fragments are on the order of 3-10 cm, with some ranging in size up to 0.75 m. The central zones of the flows are crudely columnar and vesicular throughout, although the frequency of vesicles is much lower than in the breccias. They have a slightly platy / banded appearance, which at least in part seems to be because of the arrangement of small, stretched vesicles that are aligned in parallel planes. Measurements

58 for one of the flows (on a vertical surface) revealed that these vesicles have a short axis of around 2 mm and a long axis of around 0.5 mm. Larger vesicles in the central zone are sub-rounded, with some being slightly elongated. The region between the central zones and flow-top breccias contain elongated vesicles and lava blisters. Texturally, the central zones are fine grained and show the presence of two domains – Domain 1 has subophitic/ophimottled and intergranular textures with plagioclase, clinopyroxene and olivine, while Domain 2 consists of a high concentration of large, irregular iron oxides, and lesser plagioclase and clinopyroxene. Domain 2 occurs as patches or bands in Domain 1, and this distribution is more easily visible in hand specimens and under a hand lens. The upper two flows from this section are compositionally almost identical (N11A, N11B; Table 2, Fig. 10), suggesting that they constitute lobes of a compound a’a flow, emplaced during the same eruption. The lower flow (sample N11C; Table 2, Fig. 10) is very similar in composition to the other two samples for many oxides and elements, such as CaO, Na2O and Cr. It has slightly higher titania and lower magnesia, nickel and alumina. It is possible that this flow is also part of the same eruption and taps a slightly more evolved magma, perhaps reflecting removal of olivine and plagioclase. At Location 5 along the Catlow Rim, two a’a flows sandwich a pahoehoe flow. A sample from the lower a’a flow also shows two domains – here, Domain 1 is similar to the sample from the previous location, whereas Domain 2 consists of plagioclase, highly altered olivine (probably chlorophaeite) and diktytaxitic voids (Fig. 7b). The latter are completely absent from Domain 1, and the olivines in Domain 1 are unaltered, showing typical birefringence. Close to the Steens Mountain summit, alternating pahoehoe and a’a flows are observed. On the road close to the Radio Tower (Location 13), a flow displays several classic a’a features. It has a very undulating lower surface (wavelength of undulations of a few meters), with a 0.5 m-1.5 m thick flow-base breccia (Fig. 9b). The central zone is fine grained with a hint of platyness, and contains large (up to a meter across) entrains of breccia. The upper surface is not well-exposed at this location. This flow appears aphanitic in hand specimens and thin section observations confirm the very fine grained texture that appears more andesitic than basaltic. Geochemical data reveal that this flow

59 (N14; Table 2) is a basaltic andesite (containing almost 54% SiO2) and is one of the most evolved flows sampled during this study (Fig. 10a). The section measured at the Mickey Butte (Location 15; Fig. 1, 2) consists of three thick, aphyric flows and one rich in plagioclase mega-phenocrysts; the former are bounded by thick breccias. These are quite likely to be a’a flows, although the plagioclase phyric flow appears to have broken up into lobes toward the top (Fig. 3). The three aphyric flows are distinctly platy and appear andesitic in outcrop. One of these flows has jointing that ramps up toward the top of the flow. Thin section observations on one of the flows reveal an unusual texture with fresh patches of basalt (showing a pronounced ophimottled texture) enclosed in a mosaic defined by altered oxide or glass-rich material (Fig. 7c). Very similar observations regarding surface weathering of Steens flows and their texture are described by Avent (1970). Geochemical data show that these flows are basaltic in composition (samples N20B and N20D; Table 2, Fig. 10b). A feature common to most a’a flows discussed in this paper is their relatively aphyric, fine grained character, as compared to the pahoehoe lobes that tend to be coarser and carry moderate to large plagioclase phenocrysts. One difference from classic a’a flows found in younger provinces is that constituent fragments of the breccias of these Steens flows are more vesicular and subrounded to platy, as opposed to highly spinose and dense. These breccias are akin to those associated with rubbly pahoehoe flows (e.g., Keszthelyi et al. 2006); however, the presence of prominent flow-base breccias is strongly suggestive of an a’a, rather than rubbly pahoehoe character. Similarly, highly contorted vesicles are not found in the interiors of the Steens a’a flows (although stretched vesicles are common), a feature observed in younger provinces.

Transitional flows

Flows that are difficult to characterize as typical pahoehoe or a’a are common in the studied sections. Although several transitional varieties such as slabby pahoehoe, toothpaste pahoehoe and rubbly pahoehoe have been recognised in other provinces, many of the Steens flows are difficult to classify as these. Their characters appear to be intermediate between pahoehoe and a’a; however, the term ‘transitional’ as used in this

60 section is not intended to necessarily imply that these flows transitioned from pahoehoe to a’a during an eruption. Flows with transitional characters were observed in the two sections studied in the northern Pueblo Mountains (Location 9; Fig. 1, 3). These flows are either aphyric or moderately plagioclase phyric. Some flows display preserved upper crusts (without any vesicle zonation), but have brecciated basal crusts (Fig. 8c). The central zones of such flows are either markedly platy or contain stretched vesicles. Upper crusts of some flows also contain less frequent, large (> 3 cm) vesicles and lava blisters. Some flows have a distinctly banded appearance. Underlying the earlier described a’a flow close to the Radio Tower at the Steens Mountain summit (Location 13; Fig. 1) is what initially appears to be a classic, inflated pahoehoe lobe. It has a well-preserved, banded upper crust and a vesicle-poor core with numerous vesicle cylinders and segregation veins. However, this lobe has an irregular, brecciated base, which is quite unlike typical pahoehoe lobes that invariably have preserved basal crusts. In the section at the nearby Location 14, a thick flow with a brecciated base and platy top overlies the compound pahoehoe flow discussed earlier (Fig. 1, 3). A very thick flow with well developed columnar jointing is observed at places along the Catlow Rim (Fig. 4d). Observations using a pair of binoculars reveal that this flow has a smooth base but a fairly thick flow-top breccia, which has eroded at places forming caves (Fig. 4d). This flow appears to be rubbly pahoehoe in character. A small section along the Steens Loop road (Location 10; Fig. 1, 3) contains three lobes overlain by a thick, extensive, mega-plagioclase phyric flow. The lowermost lobe is aphyric with a very undulating basal crust that is brecciated. The central part is relatively vesicle poor and crudely jointed. The upper crust shows an interesting distribution of vesicles. Large vesicles (often > 5 cm across, showing a domed or somewhat irregular appearance) are present close to the upper surface with a moderate to low frequency, and grade into smaller, more frequent vesicles in the lower parts of the crust (Fig. 8d). This distribution is the opposite of what is usually found in typical, inflated pahoehoe lobes. This lobe is exposed at another location further down the road. Here, its upper crust is incipiently brecciated in addition to the same vesicle distribution as at the previous

61 location. Lobes similar to this occur at many locations along this road and some have fairly thick basal breccias. Thick flows/flow units along Oregon Highway 78 (Location 1; Fig. 1, 3) are also hard to classify as a’a or pahoehoe. An upper section with three units is in faulted contact with a lower one containing two units. All units are either aphyric or mildly plagioclase phyric. In the upper section (Fig. 3), a three-part structure with upper crust, core and basal crust is clearly discernible for the lowermost unit, but bases are obscured by talus for the upper two units. These units are quite different from typical pahoehoe found in other Steens sections. Upper crusts display relatively large vesicles and lack banding or size distribution. The central zones are much thicker than upper crusts and are crudely columnar. The middle unit has two distinct platy zones, one just below the upper crust, and one further into the central zone. The part below the lower platy zone is characterised by stretched vesicles. The base of the lowest unit is undulating and brecciated. It is possible that the morphology of these flows is indicative of a more evolved (basaltic andesitic?) composition, although such indications have proven to be unreliable as is clear from the previous section on a’a flows. Sample N1/05 (Table 2) from this section shows a basaltic composition. It does have slightly higher silica (just over 50 wt.%) than many of the other Steens samples, but as discussed earlier, some of the flows represented by those other samples are a’a in morphology. In the lower section at this location (not shown), the two measured flows have brecciated upper crusts, massive cores, and basal crusts that are preserved or incipiently brecciated. The flow-top breccias are highly oxidised and are capped by scoriaceous material. These flows can be classified as rubbly pahoehoe.

Implications for flow emplacement

Pahoehoe flows

Although pahoehoe morphology is not necessarily indicative of slow emplacement (e.g., Solana et al. 2004), compound flows with small, superposed Hawaiian-style flow lobes are strongly suggestive of low effusion rates. It is important to note that it is difficult to

62 derive quantitative estimates of effusion rates based on such comparisons, since flow morphology is also dependent on other variables such as the temperature and composition of lavas. Yet, the striking resemblance of compound pahoehoe flows from this study with those from the DVP, Hawaii and the Snake River Plain suggests that they were emplaced in a similar leisurely manner. The morphology of individual lobes is indicative of inflation (e.g., Walker 1991; Hon et al. 1994; Self et al. 1996; Bondre et al. 2000, 2004a); however, extensive and thick inflated sheet lobes like the Columbia River Basalt province have not been observed during this study. Although Hawaiian compound pahoehoe flows are typically tube-fed, lava tubes have not been observed in the Steens sections. This is very similar to the DVP, where thick, compound pahoehoe flows occur but evidence for lava tubes is scant (Bondre et al. 2004a). It is possible, however, that ongoing studies in these provinces might reveal the existence of tube systems (see Duraiswami et al. 2004). Information obtained during the present study is insufficient to comment on the lateral extent of the compound flows. Additional complications in estimating this are imposed by the faulted nature of the Steens Basalt pile. Yet, based on the fact that the bulk of Steens volcanism is localised around Steens Mountain and that many compound flows are relatively thin (few tens of meters), it appears unlikely that most Steens pahoehoe flows traveled the extremely long distances (>200 km) inferred for many flows from the Columbia River Basalt province. Rather, it is more probable that thick packages of pahoehoe lava accumulated relatively close to eruptive vents forming scutulum-type shields (Walker 1993). This interpretation is bolstered by the presence of near-vent deposits at multiple locations and geochemical indicators of polycentric eruptions (Brueseke et al. in review). Based on the absence of local feeders and similarities in geochemical and isotopic composition, Carlson and Hart (1987) related the flows exposed in the Albert and Poker Jim Rims, in the McDermitt region, and along the western flank of the Santa Rosa range with the main Steens Mountain section. It is now recognized that local feeders are present in the Santa Rosa region, but the western flank Steens Basalt flows are still interpreted to have been sourced from the Steens Mountain vicinity (Brueseke and Hart, in press). All of these distal localities are less than 100 km from the Steens Mountain summit region. Pahoehoe flows emanating from small shields

63 have been known to travel for more than 100 km when the paleotopography is favourable (e.g. the Undara flow in the McBride Province; Atkinson et al. 1975). Thus it is not unlikely that at least some Steens flows travelled considerable distances. Pahoehoe flows are typically the products of low viscosity lavas (viscosity depends on composition, temperature, crystal and volatile content) that are subjected to low/moderate strain rates during flow (e.g., Peterson and Tilling, 1980). The effect of crystals on the rheology of magmas has been studied by numerous workers (e.g., Ryerson et al. 1988; Pinkerton and Norton, 1995; Saar et al. 2001). An important consideration for basaltic lava flow morphology is the onset of yield strength when the volume of crystals in the melt exceeds a certain critical value (e.g., Saar et al. 2001). This marks the transformation from a Newtonian to a Bingham rheology and might trigger transitions in lava flow morphology (e.g., Cashman et al. 1999). It is interesting to note that the Steens compound pahoehoe flows tend to be plagioclase phyric, and often contain a high frequency of plagioclase mega-phenocrysts (often > 20 volume %), which were likely present prior to emplacement. Cooling and crystallizing lava with high initial concentrations of large phenocrysts might be expected to be more viscous than crystal poor lava, and develop significant yield strength. Yet, these lobes are predominantly pahoehoe, which would typically be indicative of a low effective viscosity of the lava. Similar mega-phenocryst bearing flows from the DVP also show pahoehoe morphology, suggesting that such morphology might be controlled more by volatile content, groundmass crystallinity and eruption rates. Fuller (1930) has described many examples of variations in phenocryst content (plagioclase as well as olivine) throughout the vertical extent of Steens Basalt flows, and has stressed that the fluidity of the enclosing lava may have allowed such variations.

A’a flows

A’a flow morphology is a complex function of several variables such as viscosity, strain rate, crystallinity, gradient, and effusion rate (e.g., Macdonald 1953; Peterson and Tilling 1980; Kilburn, 1981, 1990; Rowland and Walker 1990; Cashman et al. 1999). A’a flows tend to be the result of flow in open channels during eruptive episodes characterised by

64 relatively high effusion rates. The generally greater thickness of the Steens a’a flows, their relatively aphyric character, and the presence of elongated vesicles and platy/banded interiors indicate moderate to high viscosity lava that was subjected to high strain rates during emplacement, which was likely faster than that of the Steens pahoehoe flows. The thick a’a flows in the Mickey Butte section are clearly platy; the ramping seen in upper parts of one of the flows also strongly suggests a high viscosity and would be more typical of intermediate and silicic compositions. Textural characters described earlier might offer some insights into the emplacement of these flows. Flow interiors are much finer in texture as compared with interiors of pahoehoe lobes, and are sparsely vesicular. Fragments from the flow-top breccias, on the other hand, tend to be moderately to highly vesicular. This suggests that considerable devolatilization during early stages of crust formation was accompanied by synchronous brecciation of the crusts, possibly leading to enhanced cooling and crystallization of flow interiors. This would probably have led to an increased viscosity and development of considerable yield strength. Preferred orientation of crystals is not apparent in these samples; however, the presence of two domains with differing textures hints at flow related segregation. This segregation might have occurred at an advanced stage of crystallization, thus inhibiting orientation of microphenocrysts. Fuller (1930) mentions that drusy cavities in the finer grained flows are restricted to certain zones. This observation is similar to that for one of the a’a samples from this study, where diktytaxitic voids occur only in one of the domains (labeled ‘2’ in Fig. 7b). This domain is also rich in altered olivine crystals. It is possible that a flow-induced segregation process drove volatile enriched material to low pressure regions associated with incipient shear planes as the flow crystallised. The olivines associated with such regions might have been altered in-situ by the volatile rich material, which can explain the presence of fresh olivines in the other domain (labeled ‘1’ in Fig. 7b) not subjected to fluid infiltration. The field expression of these flows does not indicate that they were emplaced as channelised flows – they appear to have possessed considerably wide fronts. Their exposure is restricted to vertical sections, however, which makes it difficult to obtain three-dimensional structural information and to determine whether they were channelised.

Transitional flows

The transitional Steens Basalt flows have only been studied in cross section. Many of these flows display preserved upper crusts but brecciated basal crusts. The breccias are usually composed of subrounded to platy, vesicular fragments. These characters are different from transitional varieties such as slabby and toothpaste pahoehoe. Slabby pahoehoe flows are transitional to a’a and form when the upper crust breaks up into slabs of varying sizes (e.g., Peterson and Tilling 1980; Cashman et al. 1999; Duraiswami et al. 2003). Toothpaste pahoehoe is another transitional variety (similar to proximal a’a) with longitudinal grooves on the upper crust and a spinose character (Rowland and Walker 1987). Observations pertaining to such types have primarily been restricted to surface exposures in young provinces, with the exception of the study by Duraiswami et al. (2003) who reported a slabby pahoehoe flow from the DVP. As a result, the nature of the basal crust for such flows has not been described in detail. Rubbly pahoehoe is a term used to describe flows with brecciated upper crusts but generally preserved basal crusts (e.g., Keszthelyi et al. 2006; Duraiswami et al. in review). The brecciation of basal crusts as opposed to upper crusts is difficult to explain in the absence of modern analogues, but suggests that these Steens flows might be different from rubbly pahoehoe. One possible reason for basal brecciation might be late stage flow along an irregular surface. Further studies on the nature of these breccias and flow textures are needed to shed additional light on their emplacement. Other transitional flows described during the present study, however, are best interpreted as rubbly pahoehoe. The emplacement of these flows was likely faster than the more typical Steens pahoehoe flows, involving higher strain rates and disruption of pre-existing pahoehoe crusts (e.g., S. Managave, unpub data, 2000; Keszthelyi et al. 2006; Duraiswami et al. in review). Much of the disruption probably occurred during surges in effusion rate as the eruption progressed (Bondre et al. 2004a).

Implications for eruptive history

66 Some CFB provinces such as the DVP display morphological diversity on a regional scale. The lower subgroup of the western Deccan stratigraphy consists almost entirely of compound pahoehoe flows, with simple (rubbly pahoehoe) flows occurring in the upper subgroups. It is apparent from the preceding discussion that the Steens Basalt is characterised by morphological and textural diversity, even at the outcrop scale. The geochemical data, in addition to morphological and textural data obtained during this study, provides insights into and raise interesting questions regarding the nature of the Steens eruptive episodes. These data confirm the compound nature of many Steens eruptions. For example, analyses for two samples (N18A and N18B; Table 2) collected at different heights within the postulated vent section of clastogenic lava (Location 8) show almost identical composition, confirming that they were part of the same eruptive sequence (Fig. 10). Geochemical data also indicate that some of the Steens eruptions formed compound a’a flows, as is clear from the data for location 7 discussed earlier (Table 2; Fig. 10). This study reveals that flows/flow lobes within a sequence can have similar or different composition. At Location 10, for example, the composition of the lowermost unit is different from the one overlying it (Fig. 10). Some of the compositional differences, such as alumina (Fig, 10d), might be a result of variation in plagioclase content – N15A, the lowermost unit is aphyric whereas the overlying unit is moderately plagioclase phyric. However, N15A has much higher Ba (Fig. 10a) and Sr (not shown) than N15B. If plagioclase concentration/dilution were the principal factor leading to compositional differences between N15A and N15B, then it is the latter that would have had higher Ba and Sr. This suggests that these two units belong to different eruptions and that their compositional differences are inherited from pre-emplacement differentiation processes. In the Mickey Butte section (Location 16) a plagioclase phyric flow (N20C) is sandwiched between two aphyric, a’a flows of very similar composition (N20B and N20D; Table 2). These three units are very similar to each other in terms of components such as TiO2, SiO2, CaO, Cr and Nb (Fig. 10b, c). However, N20B has markedly higher

Al2O3 (Fig. 10d) and lower MgO than the other two flows. This raises the question as to whether this flow could be related to the other two by a combination of processes

67 including fractional crystallization and concentration of plagioclase, or assimilation of a plagioclase rich cumulate. A similar process has also been invoked by Gunn and Watkins (1970) for explaining geochemical variations between the Steens flows. Simple mass- balance models attempting to simulate these processes were unable to reproduce the composition of N20B for a number of key elements. However, this possibility is worth further investigation, especially since alternations of aphyric and highly plagioclase phyric Steens flows are quite common. Such work will benefit from the acquisition of samples from thicker sections and detailed mineral composition determinations. Location 5 along the Catlow Rim (Fig. 1) provides an example of how two flows that are morphologically and texturally very different can have a similar composition. In this section, a typical compound pahoehoe flow with variable amounts of plagioclase phenocrysts is sandwiched between two a’a flows. The lower a’a flow (N16A) has a composition very different from the overlying pahoehoe lobes and a’a flow (Table 2), and is clearly part of a different eruption. The composition of lobes within pahoehoe flow (N16B, N17A and N17C) is quite similar to that of the upper a’a flow (N17B) for many oxides and elements, such as TiO2, CaO, Zr and Nb (Table 2; Fig. 10). The higher alumina contents of N16B (as compared to N17A and N17C) likely are a result of scattered plagioclase mega-phenocrysts in this sample (Fig. 10d), although the lower

P2O5 and K2O are a bit perplexing. Nevertheless, observations from this section suggest that these morphologically and texturally distinct flows could have resulted from the same eruptive episode, perhaps because of a change in the rates of effusion and cooling. One of the possible scenarios for this section would involve slower effusion and inflation of phenocryst and volatile rich lobes forming the compound pahoehoe flow with an open texture. Higher effusion rates coupled with rapid degassing, cooling and crystallization would lead to the formation of the upper a’a flow. These observations indicate that some of the Steens eruptive episodes may have been characterised by small to moderate changes in composition, but rapid changes in the physical properties of erupted lava, such as phenocryst and volatile contents. Observations regarding variations within sections have also been recorded by previous workers. Fuller (1930) discusses the periodic repetition of several flow types from a thick section exposed on the north side of the Alvord Basin. A typical sequence here begins

68 with basaltic pyroclastics at the base, followed by dense glassy basalt, more open textured lava, and finally lava bearing plagioclase mega-phenocrysts. The denser basalt displays a more evolved composition than the open textured varieties. He attributes these variations in texture and composition to processes occurring in a differentiating dike. Morphological and textural variations in flow sequences observed during the present study could also be related to pre and syn-eruptive processes occurring at depths, perhaps within differentiating magma chambers. Variations in flow morphology and textures are not necessarily unidirectional as described by Fuller (1930), but also repetitive, suggesting a cyclic nature to the responsible processes. Gunn and Watkins (1970) identified distinct groups of 3-12 flows based on geochemical data for the thick Steens section exposed along the north fork of the Alvord Creek. Brueseke et al. (in review) find that many of the “individual flows” that they sampled appear to represent discrete flow packages of compositionally similar lava. Gunn and Watkins (1970) suggested that the geochemical variations between their flow groups reflect the addition or removal of plagioclase and/or olivine. These authors attributed the cyclic nature of the lavas to successive recharging of a magma chamber, along with feldspar crystallization and flotation. This process may be visualised using the modern ideas presented by Marsh (1996), who conceives of magma chambers as dynamic reservoirs with crystallization proceeding from the sidewalls to the interiors. Such chambers consist of a complex network of intrusions with melt-crystal bearing solidification fronts and central crystal-free melts zones. Solidification fronts can be periodically disrupted by mixing of melt from the source and that resident within the chamber system – if the resultant eruption samples only the crystal-free melt, aphyric eruptions result whereas if crystals from the front are incorporated, the resulting eruptions form porphyritic lavas. In light of the probability of polycentric eruptions (e.g., Brueseke and Hart 2000; Brueseke et al. in review) in the context of the Steens Basalt, attribution of these variations to processes within a single magma chamber is questionable. The different flow packages could have been associated with multiple vents and probably multiple magma chambers. A hypothetical scenario responsible for textural and morphological variations is presented here. It draws from previous work on the Steens Basalt as well as ideas

69 presented in Marsh (1996) and Sen et al. (2006). Plagioclase phenocrysts begin to crystallize within a magma chamber or mush zone (solidification front). This crystallization leads to the generation of volatile-enriched, variably differentiated melt within the chamber system. Eruption triggered either by volatile oversaturation or fresh input from the source could expel either phenocryst-poor or phenocryst-rich magma – this would depend on phenocryst and magma densities and the extent to which the solidification zone was disrupted. In order to attain the largest sizes exhibited by the Steens phenocrysts, the magma chamber/feeder system would have to be active for a considerable duration, perhaps hundreds of years in some cases (see Sen et al. 2006). As discussed below, the field examples of Steens vents are similar to small to moderate– sized monogenetic shields and are unlikely to have had associated chambers that were active for such extended periods. However, these vents might nevertheless have drawn their magma from a more extensive and widespread network of intrusive bodies. From this perspective, flows with plagioclase megaphenocrysts probably represent local vents and feeders tapping long-lived portions of this system. Certain questions, however, remain to be answered. For example, the above scenario does not sufficiently explain why the phenocryst-rich lavas from the Steens tend to be compound pahoehoe, and why effusion rates might have varied as the eruption progressed. Future studies should focus on the details of how the observed morphological, textural and geochemical characteristics might relate to complex differentiation processes and evolution of plumbing systems.

Conclusions

Previous studies have focused on the stratigraphy, paleomagnetism and geochemistry of the Steens Basalt sequence and have sought to evaluate its position in the context of the larger mid-Miocene volcanism in the northwestern U.S. Unlike the abundant geochemical data on various Steens sections, morphological information was scant, a situation similar to that in the DVP until recently. However, such information can be crucial when using geochemical variations to interpret eruptive history. This study is the

70 first detailed, modern look at the morphology of these flows and their textures as they relate to morphology and emplacement. Observations from widely separated locations throughout the apparent vertical extent of the Steens Basalt reveal an abundance of compound pahoehoe flows; such flows tend to be plagioclase phyric and often contain large phenocrysts. The scale, internal structure and disposition of constituent flow lobes are very similar to those of lobes in Hawaii and the DVP, and suggest similar modes of emplacement. In contrast with the DVP, a’a flows are not uncommon in the Steens sequence and some of them are compound. They appear to represent volatile poor and/or more viscous lavas, which were probably emplaced at higher strain rates. Transitional flows with characters different from typical pahoehoe and a’a flows are also very common. They display diverse morphology with basal or upper crustal breccias, and platy zones within their interiors. Some of these are akin to rubbly pahoehoe flows, but others do not satisfy the criteria for such flows. It is hoped that this study will trigger further work aimed at quantifying textural variations and using numerical modeling to simulate cooling and crystallization of these flows. The morphological and textual diversity at an outcrop scale, as well as the relation between morphology, texture and composition hint at a link to deeper processes. Physical factors (such as phenocryst and volatile content, effusion rate) appear to have varied during single eruptions, leading to the formation of morphologically and texturally diverse flows. Systematic changes within crustal magma chambers might be responsible for this, although focused investigations are required to confirm or deny this hypothesis. It is expected that an integration of physical volcanological information with geochemical and textural data will enable the identification of products of single eruptive episodes. Not only will this help in testing the proposed hypothesis, but will also shed additional light on how the Steens eruptions varied through space and time.

Acknowledgements We are grateful to Dr. John Morton for his help with geochemical analyses. We appreciate the help of Stephen Pasquale in the field; discussions with Dr. Matt Brueseke and Raymond Duraiswami were very helpful. Some Steens Basalt sections along the Catlow Rim could easily approached only by traversing ranch land.

71 We thank Mr. Stacy Davies, manager of the Roaring Springs Ranch for allowing us this access. This investigation was partially funded by NSF grants EAR-0106144 and EAR- 0506887 to W.K. Hart.

72 References Cited

Atkinson A, Griffin TJ, Stephenson PJ (1975) A major lava tube system from Undara Volcano, North Queensland. Bull Volcanol 39:266-293

Avent JC (1970) Correlation of the Steens-Columbia River Basalts: some tectonic and petrogenetic implications. In: Gilmour EH and Stradling D (Eds), Proceedings of the Second Columbia River Basalt Symposium, pp133-156

Bondre NR, Dole G, Phadnis VM, Duraiswami RA, Kale VS (2000) Inflated pahoehoe lavas from the Sangamner area of the western Deccan Volcanic Province. Current Science 78 :1004-1007

Bondre NR, Duraiswami RA, Dole G (2004) Morphology and emplacement of flows from the Deccan Volcanic Province, India. Bull Volcanol 66:29-45

Bondre NR, Duraiswami RA, Dole G (2004) A brief comparison of lava flows from the Deccan Volcanic Province and the Columbia-Oregon Plateau Flood Basalts: Implications for models of flood basalt emplacement. In: Sheth HC, Pande K (eds) Magmatism in India through Time. Proc Ind Acad Sci (Earth and Planet Sci) 113:809-817

Brueseke ME, Hart WK (2000) Re-evaluation and new age constraints on the eruptive history of the mid-Miocene Steens Basalt, southeastern Oregon. Geol Soc Am Abstr Prog 32:147

Brueseke ME, Heizler MT, Hart WK, Mertzman SA (in revision) Distribution and geochronology of Oregon Plateau (U. S. A) flood basalt volcanism: The Steens Basalt revisited. J Volcanol Geotherm Res

73 Brueseke ME, Hart WK (in press) Geology and petrology of the mid-Miocene Santa Rosa-Calico Volcanic Field, Northern Nevada. Nevada Bureau of Mines and Geol Bull 112

Camp VE, Roobol M (1989) The Arabian continental alkali basalt province; Part I, Evolution of Harrat Rahat, Kingdom of Saudi Arabia. Geol Soc Am Bull 101:71- 95

Camp VE, Ross ME (2004) Mantle dynamics and genesis of mafic magmatism in the intermontane Pacific Northwest. J Geophys Res 109:B08204

Camp VE, Hanson WE, Ross ME (2003) Genesis of flood basalts and Basin and Range volcanic rocks from Steens Mountain to the Malheur River gorge, Oregon. Geol Soc Am Bull 115:105-128

Cashman KV, Thornber C, Kauahikaua JP (1999) Cooling and crystallisation of lava in open channels, and the transition of pahoehoe lava to aa. Bull Volcanol 61:306- 323

Carlson RW, Hart WK (1987) Crustal genesis on the Oregon Plateau. J Geophy Res 92:6191-6206

Carlson RW, Hart W K (1988) Flood basalt volcanism in the Pacific northwestern United States. In: Macdougall JD (ed) Continental Flood Basalts, Kluwer, Dordecht, pp 35–62

Colgan JP, Dumitru, TA, McWilliams M, Miller EL (2006). Timing of Cenozoic volcanism and Basin and Range extension in northwestern Nevada: New constraints from the northern Pine Forest Range. Geol Soc Am Bull 118:126-139

74 Deshmukh SS (1988) Petrographic variations in compound flows of Deccan Traps and their significance. In: Subbarao KV (ed) Deccan Flood Basalts, Mem Geol Soc India 10, pp 305-319

Duraiswami RA, Bondre NR, Dole G, Phadnis VM, Kale VS (2001) Tumuli and associated features from the western Deccan Volcanic Province, India. Bull Volcanol 63:435-442

Duraiswami RA, Dole G, Bondre NR (2003). Slabby pahoehoe from the western Deccan Volcanic Province: evidence for incipient pahoehoe-aa transitions. J Volcanol Geotherm Res 121:195-217

Duraiswami RA, Bondre NR, Dole G (2004) A Possible lava tube system in a hummocky lava flow at Daund, western Deccan Volcanic Province, India. In: Sheth HC, Pande K (eds) Magmatism in India through Time. Proc Ind Acad Sci (Earth and Planet Sci) 113:819-829

Duraiswami RA, Bondre NR, Managave S (in review) Morphology of rubbly pahoehoe (simple) flows from the Deccan Volcanic Province: implications for emplacement. J Volcanol Geotherm Res

Fuller RE (1930) The petrology and structural relationship of the Steens Mountain Volcanic Series of southeastern Oregon. Ph.D. dissertation, University of Washington, Seattle, pp 1-282

Fuller RE (1931) The geomorphology and volcanic sequence of Steens Mountain in southeastern Oregon. Washington University Publications in Geology 3:1-30

Gunn BM, Watkins N D (1970) Geochemistry of the Steens Mountain Basalts, Oregon. Geol Soc Am Bull 81:1497-1516

75 Hart WK, Carlson RW (1985) Distribution and geochronology of Steens Mountain-type basalts from northwestern Great Basin. Isochron/West 43:5-10

Hart WK, Carlson RW (1987) Tectonic controls on magma genesis and evolution in the northwestern United States. J Volcanol Geotherm Res 32:119-135

Hart WK, Carlson RW, Mosher SA (1989). Petrogenesis of the Pueblo Mountain basalt, southeastern Oregon and northern Nevada. Geol Soc Am Spec Paper 239:367-378

Hon K, Kauahikaua J, Denlinger R, MacKay K (1994) Emplacement and inflation of pahoehoe sheet flows: observations and measurements of active lava flows on Kilauea Volcano, Hawaii. Geol Soc Am Bull 106:351-370

Hooper P, Binger GB, Lees KR (2002) Ages of the Steens and Columbia River flood basalts and their relationship to extension-related calc-alkalic volcanism in eastern Oregon. Geol Soc Am Bull 114:43-50

Johnson JA, Hawkesworth CJ, Hooper PR, Binger GB (1998) Major and trace element analyses of Steens Basalt, Southeastern Oregon. USGS Open File Report 98- 482:1-30

Keszthelyi L, Self S, Thordarson T (1999) Application of Recent studies on the emplacement of basaltic lava flows to the Deccan Traps. In: Subbarao KV (ed) Deccan Volcanic Province, Mem Geol Soc India 43:485-520

Keszthelyi L, Thordarson T (2000). Rubbly pahoehoe: a previously undescribed but widespread lava type transitional between aa and pahoehoe. Geol Soc Am Abstr Prog 32:7

Keszthelyi L, Self S, Thordarson T (2006). Flood lavas on Earth, Io and Mars. J Geol Soc London 163:253-264

76 Kilburn CRJ (1981) Pahoehoe and aa lavas: a discussion and continuation of the model by Peterson and Tilling. J Volcanol Geotherm Res 11:373-389

Kilburn CRJ (1990) Surfaces of aa flow fields on Mount Etna, Sicily: morphology, rheology, crystallisation and scaling phenomena. In: Fink JH (ed) Lava flows and Domes: Emplacement Mechanisms and Hazard Implications. Springer, Berlin, pp 129-156.

Long PE, Wood BJ (1986) Structures, textures and cooling histories of Columbia River Basalt flows. Geol Soc Am Bull 97:1144-1155

Macdonald, G.A., 1972. Volcanoes. Englewood Cliffs (Prentice-Hall), pp 1-492

Macdonald GA (1953) Pahoehoe, aa, and block lava. Am J Sci 251:169–191

Marsh BD (1996) Solidification fronts and magmatic evolution. Min Mag 60:5-40

Mertzman SA (2000) K-Ar results from the southern Oregon - northern California Cascade Range. Ore Geol 62:99-122

Peterson DW, Tilling RI (1980) Transition of basaltic lava from pahoehoe to aa, Kilauea Volcano, Hawaii: field observations and key factors. J Volcanol Geotherm Res 7:271-293

Pinkerton H, Norton G (1995) Rheological properties of basaltic lavas at sub-liquidus temperatures: laboratory and field measurements on lavas from Mount Etna. J Volcanol Geotherm Res 68:307-323

Reidel SP (1998) Emplacement of Columbia River flood basalt. J Geophy Res 103:27393-27410

77 Rowland SK, Walker GPL (1987) Toothpaste lava: characteristics and origin of a lava structural type transitional between pahoehoe and aa. Bull Volcanol 52:631- 641

Rowland SK, Walker GPL (1990) Pahoehoe and aa in Hawaii: volumetric flow rate controls the lava structure. Bull Volcanol 52:615-628

Ryerson FJ, Weed HC, Piwinski AJ (1988) Rheology of subliquidus magmas. I. Picritic compositions. J Geophy Res 93:3421-3436

Saar MO, Manga M, Cashman,KV, Fremouw S (2001) Numerical models of the onset of yield strength in crystal-melt suspensions. Earth Planet Sci Lett 187:367-379

Self S, Thordarson T, Keszthelyi L (1996) A new model for the emplacement of Columbia River basalts as large, inflated pahoehoe lava flow fields. Geophy Res Lett 23:2689-2692

Sen G, Borges M, Marsh BD (2006) A case for short duration of Deccan Trap eruption. EOS, Trans Am Geophy Union 87:197-200

Shelley D (1993) Igneous and metamorphic rocks under the microscope. Chapman & Hall, pp 1-445

Shoemaker, KA and Hart, WK (2004) Temporal controls on basalt genesis and evolution on the Owyhee Plateau, Idaho and Oregon. In: Bonnichsen, B, White, CM and McCurry, M (eds) Tectonic and Magmatic Evolution of the Snake River Plain Volcanic Province. Idaho Geological Survey Bulletin 30: 313-328.

Solana MC, Kilburn CJ, Rodriguez Badiola E, Aparicio A (2004) Fast emplacement of extensive pahoehoe flow fields: the case of the 1736 flows from the Montana de la Nueces, Lanzarote. J Volcanol Geotherm Res 132:189-207

78 Swisher CC, Ach JA, Hart WK (1990) Laser fusion 40Ar/39Ar dating of the type Steens Mountain Basalt, southeastern Oregon, and the age of the Steens geomagnetic polarity transition. EOS, Trans Am Geophy Union 71:1296

Thordarson T, Self S (1998) The Roza Member, Columbia River Basalt Group: A gigantic pahoehoe lava flow field formed by endogenous processes? J Geophy Res 103:27411-27445

Walker GPL (1971) Compound and simple flows and flood basalts. Bull Volcanol 35:579-590

Walker GPL (1991) Structure and origin by injection under surface crust, of tumuli, ‘lava rises’, ‘lava-rise pits’, and lava inflation clefts’ in Hawaii. Bull Volcanol 3: 546- 558

Walker GPL (1993) Basaltic-volcano Systems. In: Pritchard HM, Alabaster T, Harris NBW, Neary CR (eds) Magmatic processes and plate tectonics. Geol Soc Spec Publ 76:3-38

Walker GPL, Canyon-Tapia E, Herrero-Bervera (1999) Origin of vesicle layering and double imbrication by endogenous growth in the Birkett basalt flow (Columbia river plateau). J Volcanol Geotherm Res 88:15-28

Wilmoth RA, Walker GPL (1993) P-type and S-type pahoehoe: A study of vesicle distribution patterns in Hawaiian lava flows. J Volcanol Geotherm Res 55:129- 142

79 Table 1: Locations included in the present study and their approximate stratigraphic context

Location Location description Easting Northing Approximate Inferred stratigraphic position number Age (Ma)† 1 Along Route 78, close to Folly Farm 401938 4771054 16.03±0.30 upper Steens 2 Southwest of Frenchglen, west side of road 338837 4741669 ‘middle’ Steens 3 Along the Catlow Rim (observations made with a pair of 337186 4724766 ‘middle’ to upper Steens binoculars) 4 Close to mouth of canyon along the Catlow Rim 337560 4721983 ‘middle’ to upper Steens 5 Two closely spaced exposures along the Catlow Rim 341014 4713637 ‘middle’ to upper Steens 6 Southern wall of prominent canyon cutting through the 343086 4708178 ‘middle’ to upper Steens Catlow Rim 7 Section along Catlow Rim 347049 4698342 lower Steens* 8 Section along Catlow Rim, close to beginning of Funnel 343441 4687999 lower Steens* Canyon road 9 Two closely spaced locations close to road leading to 356024 4673114 upper Steens* Domingo Pass (Pueblo section) 357017 4672466 10 Along southern section of the Steens Loop Road 362032 4726059 upper Steens

80 11 Along northern section of Steens Loop Road leading from 368299 4729427 upper Steens Jackman Campground to summit 12 Along northern section of Steens Loop Road, south of 370618 4727860 upper Steens intersection with Kaiger Gorge road 13 Steens summit (close to radio tower) 370575 4721697 upper Steens 14 Close to Wildhorse Lake trail west of radio tower 369844 4721620 upper Steens 15 Section close to Miranda Flat 385050 4729150 16.33±0.21 upper Steens 16 Section at the base of Mickey Butte (observations along 389181 4727447 16.58±0.18 lower Steens bedrock stream)

† Ages from Brueseke et al. (in review); * Stratigraphic position based only on geochemical criteria. UTM coordinates are from zone 11T Table 2: Major element (wt.% oxides) and trace-element (ppm) data for Steens Basalt samples from the present study

Sample N1/05 N11A/04 N11B/04 N11C/04 N12/04 N13/04 N14/04 N15A/04 N15B/04 Location 1 7 7 7 14 13 13 10 10

SiO2 49.97 46.18 46.66 46.97 46.87 47.37 52.64 49.99 47.74 TiO2 1.98 1.70 1.73 1.93 2.48 1.70 2.39 2.14 2.40 Al2O3 15.36 18.05 17.48 16.43 16.15 19.24 16.06 16.09 18.29 t Fe2O3 13.04 13.19 13.41 14.43 14.60 12.00 12.04 12.71 12.63 MnO 0.19 0.17 0.18 0.18 0.19 0.15 0.18 0.18 0.17 MgO 5.36 5.97 6.13 5.56 6.08 5.12 2.61 4.26 4.01 CaO 8.82 9.86 9.43 9.54 8.25 9.10 5.00 6.62 8.69

Na2O 3.07 2.95 2.99 3.04 3.31 3.33 4.43 4.08 3.67 K2O 0.68 0.41 0.53 0.77 0.96 0.60 2.26 1.79 1.00 P2O5 0.32 0.22 0.24 0.32 0.25 0.29 0.94 0.71 0.41 LOI 0.50 0.21 0.40 0.25 -0.39 -0.04 0.38 -0.01 -0.25 Total 99.29 98.92 99.19 99.42 98.76 98.86 98.92 98.56 98.75 Rb 15 7 10 13 22 9 45 29 21 81 Sr 438 489 488 517 520 675 696 746 560 Y 26 26 24 26 38 25 38 35 35 Zr 149 90 89 108 193 114 247 193 198 V 265 299 319 325 308 270 228 272 260 Ni 16 140 130 94 112 158 2 44 48 Cr 32 47 49 48 58 48 8 28 36 Nb 14.2 6.6 6.7 7.5 10.8 7.7 12.7 11.4 10.9 Cu 5 162 189 180 270 105 33 86 150 Zn 119 83 91 88 106 90 118 104 89 Ba 397 254 279 334 416 298 896 738 455 La 14 6 8 11 14 10 36 29 21 Ce 33 14 17 20 31 21 68 57 40 Sc 30 32 31 31 26 25 22 23 25 Pb 3 3 2 4 6 3 9 6 5

Sample N16A/04 N16B/04 N17A/04 N17B/04 N17C/04 N18A/04 N18B/04 N19/04 N20B/04 Location 5 5 5 5 5 8 8 9 16 SiO2 46.76 47.53 48.80 47.83 48.25 48.38 48.57 50.17 48.66 TiO2 1.66 2.66 2.68 2.83 2.64 1.52 1.60 1.80 1.90 Al2O3 17.22 15.53 14.32 14.80 14.15 16.49 16.12 16.22 15.32 t Fe2O3 13.22 14.35 14.99 15.88 14.85 11.75 11.69 11.23 11.91 MnO 0.18 0.20 0.22 0.22 0.23 0.18 0.18 0.16 0.17 MgO 7.26 4.92 4.66 4.16 4.91 7.31 7.06 4.34 7.23 CaO 10.00 8.57 8.56 8.36 8.49 11.28 11.14 7.83 9.45 Na2O 2.91 3.39 3.30 3.59 3.23 2.80 2.75 3.61 2.89 K2O 0.47 1.25 1.69 1.57 1.62 0.68 0.91 2.13 0.94 P2O5 0.23 0.45 0.75 0.77 0.71 0.31 0.32 0.81 0.61 LOI -0.16 -0.13 -0.13 0.18 0.50 0.08 0.63 1.17 1.09

82 Total 99.75 98.72 99.82 100.18 99.58 100.77 100.97 99.47 100.16 Rb 6 25 28 30 27 10 11 38 14 Sr 468 482 458 519 452 453 459 583 432 Y 24 40 45 48 43 22 23 29 26 Zr 83 210 232 238 227 84 91 192 123 V 301 334 357 382 376 283 290 242 267 Ni 143 50 48 45 49 116 120 74 80 Cr 49 44 52 30 51 183 185 45 65 Nb 5.9 12.5 13.4 12.8 13.3 5.0 6.1 14.3 10.4 Cu 120 242 262 237 267 134 149 113 48 Zn 96 103 116 121 115 75 78 100 103 Ba 240 464 445 513 445 249 336 635 262 La 10 22 22 18 19 6 9 28 14 Ce 18 41 46 40 40 12 17 52 25 Sc 29 28 32 33 31 33 29 22 26 Pb 1 6 6 5 6 2 2 5 3

Sample N20C/04 N20D/04 Location 16 16

SiO2 48.47 48.15 TiO2 1.98 1.79 Al2O3 17.93 15.55 t Fe2O3 11.07 11.91 MnO 0.17 0.19 MgO 4.34 7.57 CaO 9.56 9.78 Na2O 3.49 2.80 K2O 1.31 0.76 P2O5 0.51 0.22 LOI 1.00 1.07 Total 99.82 99.78 Rb 23 9 Sr 575 437

83 Y 33 25 Zr 158 107 V 271 249 Ni 74 95 Cr 66 82 Nb 9.9 10.1 Cu 200 37 Zn 88 100 Ba 365 212 La 19 10 Ce 36 22 Sc 27 24 Pb 4 3

t Fe2O3 -total iron as Fe2O3; LOI-loss on ignition. Location numbers refer to those from Table 1.

Figure 1: Map showing the locations where detailed observations were recorded during the present study. Solid and dashed lines indicate roads. Inset shows the extent of the spatial extent of the Columbia-Oregon Plateau flood basalt volcanism (after Camp and Ross 2004), and the position of the Steens-Pueblo fault scarp (SPF).

Figure 1

WASHINGTON 0 / Ro 118 40 ute 78

CRB province 1

Diamond OREGON

SPF IDAHO

0 200 km Frenchglen 2 11 Steens Loop 15 Road Mickey Butte 10 12 420 40 / 3 16 4 13, 14

C a Alvord Catlow t l o desert Valley w 5

i m 6

Andrews 7

Alvord Peak 420 20 / 8

Fields

119 0 Pueblo Mountain 1180/20

85 Figure 2: Selected major element plots of geochemical data from the present study. Fields for the ‘upper’ and ‘lower’ Steens basalt based on data from and criteria defined by Johnson et al. (1998) and by Camp et al. (2003). Dashed lines in ‘a’ and ‘c’ drawn at 1 wt.% K2O and 300 ppm Ba respectively, which have been used to demarcate the ‘upper’ and ‘lower’ Steens (e.g. Brueseke et al. in review).

Figure 2 3 a b

Upper Steens 3 2 2 O O 2 i K T Lower Steens 2 1

0 1

c d 0.9 87 800

0.7

600 5 O 2 a P B 0.5 400

200 0.3

0 0.1 2 4 6 8 10 2 4 6 8 10 12

TiO/2PO25 MgO

Location 1 Location 5 Location 7 Location 8 Location 9 Location 10 Location 13 Location 14 Location 16 Figure 3: Selected sections measured and described during the present study. Location numbers refer to those in Figure 1 and Table 1. Only part of the section at Location 16 has been shown. Note the morphological diversity between and within sections.

Figure 3 Location 16

50 m ? Location 6 e n o z

d e l a e n n a Location 9 Location 14 Location 5

25 Location 1 89 Location 10 Location 7 Location 8

? e r u t

? c u r t s

p m a r a i r o c s 0

vesicular regions breccia joints platy zone covered interval

small pahoehoe segregation pipe vesicles larger vesicles flow contact lobes strucures Figure 4: Field photographs of various sections along the Catlow Rim. (a) A thick, compound pahoehoe flow (> 25 m) with numerous overlapping lobes. The arrows demarcate lateral terminations of the lobe 1. Lobe 2 occupies the low created by the undulating surface of the lobe below it. (b) A moderately thick, plagioclase phyric pahoehoe lobe at Location 14, overlain by an aphyric, platy flow. (c) Thick (~ 50 m) section of clastogenic lava and agglutinate. The boundaries between some of the individual clastogenic units have been marked by lines to facilitate their recognition. (d) Thick (> 15 m) columnar-jointed flow exposed along the Catlow Rim (labeled 1). The arrow indicates the position of the flow-top breccia, which has weathered at places to form shallow caves. Overlying this flow is a package of pahoehoe lobes (labelled 2).

Figure 4

a b 2

1 1 2 91

c d

1 Figure 5: (a) Compound pahoehoe flow exposed at Location 11 near the Jackman Campground. Note the absence of pipe vesicles, vesicle banding in the crusts and lobe terminations. (b) Compound pahoehoe flow exposed at Location 5, overlain by an a’a flow. Note the presence of pipe vesicles at the bases of flow lobes and the presence of small toes.

Figure 5

3 m

vesicular breccia upper crusts

pipe vesicles joints

93 Figure 6: (a) Schematic sketch of the internal structure of a pahoehoe flow lobe from Location 11. Note the random orientation of plagioclase mega-phenocrysts in the upper crust (1), as opposed to distinct alignment in the core (2). The core also contains segregation structures (3). The basal crust (4) shows prominent vesicle banding, with vesicle size increasing towards the top. (b) Sketch of the basal crust (1) of the flow lobe in Figure 4b, showing the presence of many smaller lobes/toes (2). Detailed sketch of lobe 1 showing a radial arrangement of plagioclase mega-phenocrysts (3). (c) Photograph of small, P-type pahoehoe lobe at Location 5, showing well-developed open pipe vesicles.

Figure 6

a c 1

2 3 m 3

4 b 1 2 1 m

3 1 m

95 Figure 7: Photomicrographs showing the different textural characteristics of the Steens Basalt flows. (a) Typical diktytaxitic texture in the core of a pahoehoe lobe exposed at Location 5. (b) Two textural and mineralogical domains (labeled 1 and 2) in the a’a flow exposed at Location 5. Note the presence of microvesicles in domain 1 (appear white and irregular) and the strongly altered olivines (dark grey/black), as opposed to a subophitic texture in domain 2. (c) Typical ophimottled texture in an a’a flow exposed at Location 15, with large clinopyroxene crystals enclosing plagioclase laths.

Figure 7

500 m b

500 m c

500 m

97 Figure 8: (a) Field sketch of an a’a flow at Location 7. Note the undulating nature of flow-top and flow-base breccias and sections of the core protruding into those. (b) Detailed sketch of the nature of the breccias bounding the flow in ‘a’. Most clasts lack sharp corners and are moderately to highly vesicular. (c) Field sketch of a transitional flow exposed at Location 9. The upper crust is undisrupted and has relatively spherical vesicles, whereas the basal crust is brecciated. Stretched vesicles characterise the central zone of this flow. (d) Vesicle distribution in the upper crust of a flow from Location 10. Smaller vesicles at the base of the upper crust grade into larger vesicles at the top.

Figure 8

a b

0.1 m 3 m

vesicles c d

3 m 1 m

vesicular regions breccia

larger vesicles joints

99 Figure 9: (a) Photograph showing the breccia zone (demarcated by arrows) between two a’a flows from Location 7. The flow-top breccia of the lower flow has merged with the flow-base breccia of the upper flow. (b) A’a flow exposed at the summit of Steens Mountain (Location 13) displaying clinkery / brecciated upper and lower surfaces, and a massive interior. Note the hammer for scale.

100

Figure 9 a 2

101 Figure 10: Geochemical plots of (a) Ba vs. TiO2 (b) SiO2 vs. TiO2 (c) Nb vs. TiO2 and

(d) Al2O3 vs. TiO2 for samples discussed in the text. The analytical uncertainty for silica is shown by the vertical bar in ‘b’.

102

Figure 10

54 a b N14/04

800 52

600 2 O a 50 i B S

400 48 N11C/04 N17B/04 200 46 20

103 14 c d 19 12 N15B/04 N20B/04 18 3 O

10 2 l b 17 A N

8 N16B/04 N15A/04 16

6 15

4 14 1 2 3 1 2 3

TiO 2 TiO 2

Location 1 Location 5 Location 7 Location 8 Location 9 Location 10 Location 13 Location 14 Location 16 CHAPTER 4

GEOLOGY, GEOCHRONOLOGY AND COMPOSITIONAL DIVERSITY OF THE JORDAN VALLEY VOLCANIC FIELD (JVVF), SOUTHEASTERN OREGON: IMPLICATIONS FOR SMALL-VOLUME, MONOGENETIC VOLCANISM

ABSTRACT

The newly-defined Jordan Valley Volcanic Field (JVVF) is situated at the northernmost part of the Owyhee Plateau. It consists of 14 identifiable vents ranging in age from the latest Miocene to Recent and provides a good opportunity to investigate small-volume, monogenetic basaltic volcanism. This study integrates physical and geochemical data for the JVVF and uses them to understand its evolution as well as of Monogenetic Volcano-Fields (MVF) in general. The JVVF is characterized by considerable between-vent chemical and isotopic diversity. Within-vent heterogeneity is also documented, particularly in the younger eruptive systems. Two overall chemostratigraphic groups are identified primarily based on age and plots of Sr and Nb vs. TiO2. The first group includes younger (<250 ka), mildly alkaline vents with >400 ppm Sr and >18 ppm Nb, whereas the second group is defined by older (ca. 0.5-5.5 Ma) vents that erupted a continuum of tholeiitic basalt compositions from low-K, high-alumina olivine tholeiite to Snake River olivine tholeiite endmembers, with <300 ppm Sr and <18 ppm Nb. Compositional diversity within the JVVF (between and within vents) can be, in principle, attributed to a combination of factors including source heterogeneity, magma mixing, fractional crystallization and crustal assimilation. This study suggests that although differentiation processes such as fractional crystallization at depth did occur, the geochemical variations require a contribution from chemically and isotopically heterogeneous sources. JVVF basalt geochemistry correlates poorly with location, whereas there is a general shift toward less radiogenic Sr isotope ratios and increasing alkalinity through time. If heterogeneous sources are indeed involved, then the data suggest that the scale of such heterogeneity must be small. Furthermore, magma generation processes tapping these sources must also have a temporal aspect. This study clearly reveals the complexity of MVF and demonstrates the usefulness of integrating physical and geochemical data in understanding their evolution.

104 INTRODUCTION

Polygenetic volcanoes, such as Mt. Etna and Mt. St. Helens, are long-lived and characterized by multiple eruptions spanning a substantial period of time. Monogenetic volcanoes (Walker, 1993), on the other hand are relatively small volcanoes that are short-lived and erupt only once during their life span. Such volcanoes have a strong propensity to occur in clusters, forming monogenetic-volcano fields (MVF). They are either associated with flanks of major volcanoes or occur as scattered vents in intracontinental settings, often away from major plate boundaries. Extensional settings seem to be preferred although this is by no means a rule. There is considerable diversity in the nature of MVF – some fields contain hundreds of volcanoes, primarily scattered cinder cones and maars (e.g. Springerville volcanic field in Arizona), whereas others contain tens of volcanoes (e.g. Auckland and Big Pine Volcanic Fields in New Zealand and Eastern California respectively). The characters and formation of such fields have been discussed in some detail by Settle (1979) and Wood (1980). Recent studies include those by Conway et al. (1997) and Connor et al. (2000). The Snake River Plain and the Owyhee Plateau (Shoemaker and Hart, 2002) regions of the western United States, on the other hand, are dominated by ‘scutulum-type’ monogenetic shield volcanoes, leading to a style of volcanism referred to as ‘Plains-style’ by Greeley (1982). These differences are probably a result of variations in the rate of thermal energy supply, with higher rates allowing continuous magma pathways to be established and favoring the formation of shields by Hawaiian-style fire fountaining. Basalts associated with MVF from the Basin and Range Province typically are strongly alkaline and xenolith bearing, and often reflect derivation by small degrees of melting from deeper parts of metasomatized lithospheric mantle (e.g. Big Pine Volcanic Field; Ormerod et al. 1991). On the other hand, those associated with the Owyhee Plateau-Snake River Plain regions are predominantly tholeiitic and indicate a relatively shallow depth of origin (or last equilibration) in the uppermost asthenosphere or variably enriched lithosphere. Most MVF are characterized by considerable between-vent compositional heterogeneity. Vent alignments are quite common and can be long-lived – volcanoes with distinct compositions can be erupted along the same alignment. Until recently, however, heterogeneity within individual monogenetic volcanoes had not been the subject of much attention with the exception of a few studies (e.g. Glazner et al. 1991). The past few years have witnessed an increasing focus

105 on this aspect, leading to the recognition of the fact that compositional heterogeneity within individual monogenetic volcanoes is not uncommon (e.g. Nemeth et al. 2003; Strong and Wolff, 2003; Bondre and Hart, 2004). These studies and ongoing ones are contributing towards answering several questions pertaining to the evolution of MVF, such as: a) What are the causes of compositional heterogeneity within and between monogenetic volcanoes? What do they tell us about the relative contributions of source heterogeneity and its length scales, and differentiation processes? b) What are the time scales involved in the generation, transport, and eruption of magmas in such volcanoes? c) What governs the spatial distribution and relative ages of monogenetic volcanoes? d) What role do faults and fractures play in the ascent and eruption of magmas, and what controls the longevity of vent alignments? This study presents the results of an integrated field, geochemical and geochronological investigation of the Jordan Valley Volcanic Field (JVVF), an MVF located in southeastern Oregon (Fig. 1). The study aims to understand the volcanological evolution of this field and address, at least in part, the questions posed above. More specifically, the primary goals of this study are to: a) Understand the field relationships of the vents and lava flows constituting the JVVF, including vent alignments and relative chronology, b) Comprehensively characterize the vents and associated lava flows in terms of their chemical and isotopic composition, c) Supplement field-based interpretation with radiometric ages for selected samples in order to construct a robust chronology of events, and d) Use the field, geochemical and geochronologic data to understand the compositional make-up and petrologic evolution of this volcanic field. It is not the purpose of this work to come up with a comprehensive petrogenetic model for the JVVF. That project will require mineral-scale geochemical data as well as, more importantly, geophysical and geochemical constraints on the nature of the lithosphere in this region. Similarly, the regional tectonomagmatic implications of Owyhee Plateau basaltic volcanism have recently been extensively dealt with by Shoemaker (2004). The present study, on the other hand, aims at revealing the importance of an integrated physical and geochemical

106 investigation in illuminating various questions posed by MVF and in providing pointers to understanding the physical and petrogenetic evolution of such fields. Fieldwork in the Jordan Valley area was undertaken in 2003 and 2004 with a view to identifying vents and understanding stratigraphic relationships between various flows in the JVVF. Fieldwork in 2006 focused on observing and sampling flows from the canyon of the Owyhee River as part of an ongoing collaboration with researchers from the geology department at Central Washington University. This chapter will provide a regional context, review previous work, and then present field and geochronologic information for the JVVF. This will be followed by a discussion of geochemical data and its spatial and temporal aspects. Broad aspects of the petrogenesis of the JVVF lavas will then be discussed followed finally by implications for the physical and chemical evolution of MVF.

REGIONAL TECTONOMAGMATIC FRAMEWORK AND PREVIOUS WORK

The JVVF lies within the Oregon Plateau (Fig. 1a), in a transition zone between the northern part of the Owyhee Plateau (Shoemaker and Hart, 2002) and the southern edge of the Miocene Oregon-Idaho graben (Cummings et al. 2000), as shown in Figure 1b. It also lies on another, even more important transition between cratonic North America to the east and accreted terranes to the wast (e.g. Leeman et al. 1992). It is close to the nexus of the Yellowstone-Eastern Snake River Plain and Oregon High Lava Plains trends (Shoemaker, 2004). This region of Oregon has been volcanically active since the mid-Miocene, and has witnessed the eruption a variety of products ranging from rhyolite, to fractionated tholeiites, to relatively primitive olivine tholeiites, to mildly alkaline basalts. The earliest pulse of volcanism was characterized by extensive and strongly fractionated tholeiitic flood basalts of the Columbia-Steens episode that initiated around 17 Ma. These have been attributed either to the inception of a mantle plume below the lithosphere (Pierce and Morgan, 1990; Geist and Richards, 1993; Camp and Ross, 2004) or to the initiation of back-arc rifting (e.g. Carlson and Hart, 1987). Such volcanism continued intermittently until around 11 Ma (Hart and Carlson, 1987), when relatively primitive high alumina olivine tholeiiites (HAOT) started erupting throughout the Oregon Plateau. Such basalts mark a change from voluminous and extensive fissure-fed eruptions to those from local fissures and discrete monogenetic shields.

107 An interesting aspect of the basaltic volcanism in the northwestern United States is the varying isotopic composition of the lavas in space and in time. For example, lavas east of approximately 117 degrees longitude have 87Sr/86Sr ratios of ~ 0.706 and higher, while those to the west have lower ratios. Temporally, isotopic compositions show an intriguing relationship. The most radiogenic Sr is associated with lavas erupted during 11-6 Ma while younger and older lavas have less radiogenic Sr (Hart, 1997). The spatial variations in isotopic composition have been attributed to the differing nature of the basement in this region, with higher values corresponding to the cratonic part of North America and lower values corresponding to accreted terranes (Leeman et al. 1992). It has been observed that the same isotopic composition can occur over a wide range of basalt differentiation, suggesting that the first order variations in isotopic composition are a result of variations in relative contributions of different mantle sources (e.g. Shoemaker and Hart, 2002). Just as it has been magmatically active, this region has also been tectonically active. Unlike the Columbia Plateau in the north, the Oregon Plateau has undergone diffuse extension related to the broader Basin and Range extension. However, since it accommodates greater extension in the south and minimal extension to the north, it is characterized by numerous strike- slip zones, such as the Brothers Fault Zone (Lawrence, 1976; Walker and MacLeod, 1991). As mentioned before, the JVVF lies at the intersection of the Miocene Oregon-Idaho Graben (OIG) with the Owyhee Plateau. The OIG was the site of localized E-W extension and syn-extensional volcanism from approximately 15 to 12 Ma. The extensive Owyhee Basalts (actually basaltic andesites) are part of the lavas that were erupted from foci within this graben. This graben ceased to be active around 10 Ma (Cummings et al. 2000) when the western Snake River Plain trend became established. The Owyhee Plateau on the other hand seems to have resisted substantial extension. Much of the volcanism seems to have been focused along its margins (Brueseke et al. 2002). From the previous discussion, it is clear that the Oregon Plateau in general and the vicinity of the JVVF in particular have had a prolonged magmatic history. This must no doubt have preconditioned the lithosphere in this region in a complex manner, and any investigation of the JVVF volcanism must bear this history in mind. Volcanism on the Owyhee Plateau (Fig. 1b) has been comprehensively documented by Shoemaker (2004) and Shoemaker and Hart (2002), and has been evaluated in the context of regional models. Shoemaker (2004) proposed that the Owyhee Plateau region was a discrete

108 tectnono-magmatic entity, and owed its origin to thrusting of accreted lithosphere over a westward projecting shelf of cratonic lithosphere. Subduction of the Farallon plate during and subsequent to the Laramide Orogeny modified the asthenosphere beneath the Owyhee Plateau by the addition of fluids rich in large ion lithophile elements (LILE). This subduction modified wedge melted to large degrees during rollback or foundering of the slab in the Middle Miocene to generate voluminous flood basalt eruptions on the Oregon Plateau (e.g. the Steens Basalt). After around 11 Ma, diffuse extension set in, leading to low degrees of melting of the asthenosphere as well as of the subcontinental lithospheric mantle (SCLM). Shoemaker has discussed in detail the compositional characteristics and petrogenesis of this volcanism, and concluded that differing mantle source regions have to be invoked in order to explain the first order compositional diversity. High Alumina Olivine Tholeiite (HAOT; Hart et al. 1984) magmas resulted from asthenospheric melts interacting with small amounts of mafic material in the lithosphere (e.g. Hart et al. 1997). In order to explain the relatively radiogenic Sr-isotopic compositions of HAOT lavas, the mafic contaminant would have to be radiogenic enough to overprint the depleted-mantle signatures. Snake River Olivine Tholeiite (SROT) magmas resulted largely from melting of the SCLM (including mafic material within this). Various basalt compositions transitional (TB) between these two end-member varieties resulted from various degrees of mixing between the HAOT-like and SROT-like melts. The most recent pulse of volcanism in this area includes vents close to the margins of the Owyhee Plateau, including monogenetic vents from the JVVF that have erupted a combination of HAOT and TB to mildly alkaline lavas. No endmember SROT-like compositions have erupted in the Owyhee Plateau region after around 6 Ma. Shoemaker (2004) proposed that this reflects depletion in the readily fusible components within the main Owyhee Plateau SCLM, and greater input of the asthenospheric mantle and lithospheric mantle associated with the accreted terranes.

THE JORDAN VALLEY VOLCANIC FIELD

Geology and geomorphology This study is the first to recognize the group of monogenetic volcanoes lying to the east of the Owyhee River and north of the Antelope Valley Fault system as part of a distinct volcanic field – the Jordan Valley Volcanic Field (JVVF; Fig. 1-3). Some of these vents have been studied

109 previously by Hart (1982; 1985), Hart and Mertzman (1983), and Hart et al. (1984) who constructed a generalized stratigraphy for this area as well as chronologically and compositionally characterized the basaltic volcanism. These authors observed the occurrence of coevally erupted HAOT and TB to mildly alkaline basalts in close geographic proximity. Shoemaker and Hart (2002) and Shoemaker (2004) included selected basalts from the JVVF and the Oregon-Idaho graben in their regional investigation of the Owyhee Plateau. The present study focuses only on the latest Miocene (~ 6 Ma) to Recent basaltic volcanism that constitutes the JVVF, and the reader is referred to Hart and Mertzman (1983) for a review of the late Cenozoic volcanic stratigraphy of the Jordan Valley area. The JVVF, as recognized in this study, consists of 14 individual basaltic vents and their associated lava flows as seen in Figure 2 - a satellite image of this area. Two of these vents were interpreted from the satellite data but have not been included in the present study. Figure 3 is a simplified geologic map of the JVVF, which is based on field mapping conducted during this study, remotely sensed data, and previous mapping by Hart and Mertzman (1983). Details regarding the vents are provided in Table 1. The reader is advised to refer to Figure 2 and Table 1 for names of individual volcanoes from the JVVF and various other structural and geomorphic features that will be referred to throughout the text. Some of these vents lie on distinct alignments – three such alignments are readily visible on Figure 2. These include the NNW-SSE Coffeepot Crater-Three Mile Hill alignment (A), the E-W West Crater-Skinner Hill alignment (B) and the E-W Owyhee Butte-Three Mile Hill (C) alignment. As will be discussed in detail later, these alignments appear to have been active for extended periods of time. It may be emphasized that the recognition and significance of vent alignments in such fields should be borne out by concrete observations in the field pertaining to the geologic nature of the causative features, such as faults/fractures and/or dikes. Alignments ‘A’ and ‘B’ show clear evidence of a relationship with pre-existing faults/fractures (Fig. 2 and 3). Alignment ‘C’ does not show such evidence; however, based on its parallelism with ‘B’ and with other major E-W features in the area such as the Antelope Valley Fault system (Fig. 2 and 3), it is likely to be geologically significant. All JVVF vents, except two, are low-volume ‘scutulum-type’ shield volcanoes (Noe- Nygaard, 1968). Such shields are extremely common on the Snake River Plain and Owyhee Plateau, and are characterized by low slopes. The Rocky Butte is the youngest preserved shield

110 in the JVVF. It has two craters at the summit with walls of basalt or welded spatter. The western crater appears to be younger, based on the fact that a flow emanating from it entered the eastern crater. The Rocky Butte summit lacks accumulations of loose cinder/scoria. Some older shields do show the presence of spatter cones/ramparts and clastogenic material (e.g. Owyhee Butte, Bogus Bench Vent; Fig. 3). The two remaining vents (the Coffeepot Crater and the West Crater) are better characterized as transitional between shields and cinder/scoria cones. Basaltic material associated with the JVVF volcanoes typically overlies older silicic or intermediate composition material associated with the Oregon-Idaho Graben or Plio-Pleistocene fluvio-lacustrine sediments (e.g. Ferns et al. 1993). In the southernmost part of the area, within the Jordan Creek Canyon, older Miocene basalts underlie the JVVF basalts (Hart and Mertzman, 1983; Shoemaker, 2004). In turn, the JVVF basalts are overlain by younger fluvio-lacustrine sediments and loess. Flows associated with the vents are typical compound pahoehoe, very similar to those found on the Snake River Plain. Evidence of well-developed tube systems is apparent in some of the flows, such as the West Crater flow and the Coffeepot Crater flow field. These compound pahoehoe flows are characterized by numerous inflation features, such as tumuli, pressure ridges and lava inflation pits. A’a flows are quite rare; a notable example is found at the terminus of the northeastern flow exposures of the West Crater vent. Near vent material is invariably clastogenic, although loose cinder is found at places around Coffeepot Crater and the Cow Vent complex. Volcanism in the JVVF has been strongly influenced by and has itself influenced fluvio- lacustrine activity in the area. Many of the flows have been guided by pre-existing topography and stream channels (e.g. the West Crater lava flow; Fig. 2, 3). The terminal portions of many lava flows entered standing water, as is evident from the pillow-palagonite complexes associated with the West Crater and Owyhee Butte lava flows along the Owyhee River canyon. At the same time, many lava flows have also strongly influenced the drainage that existed during their emplacement. For example, the Coffeepot Crater lava flow damned the Cow Creek to create the Cow Lakes (Fig. 2, 3). Several advances of the West Crater lava flow temporarily dammed the Owyhee River, whereas the Skinner Hill/Rocky Butte lava flow appears to have pushed part of the Jordan Creek to the south near Arock (Fig. 2, 3). The Owyhee River and Jordan Creek provide a fascinating perspective on the prolonged interplay of volcanism and fluvial activity in this area.

111 Most of the area occupied by the JVVF is characterized by low relief. Except along the Owyhee River and the Jordan Creek, thick vertical sections through the stratigraphy are difficult to find. Initial correlations of flows to vents depended on a relative chronology of events constructed using the lateral relationships and relative ages of lava flows as inferred from geomorphic expression. This was refined by using radiometric ages for selected vents and flows, as well as their petrographic characteristics and their chemical and isotopic compositions.

Geochronology In order to verify the relative age relationships between various vents and flows as determined on the basis of stratigraphy, 40Ar/39Ar ages were obtained for 6 vents and one associated flow (see Appendix 2 for details regarding methodology). K-Ar ages for three other vents were reported by Hart et al. (1984). The age of the Coffeepot Crater vent is constrained by a radiocarbon age of material from the base of the Cow Lakes that formed in response to damming of the Cow Creek by the Coffeepot Crater flow (Mehringer, 2004). These age data, presented in Table 1, indicate that volcanic activity within the JVVF spanned the interval from around 6 Ma to less than 1.5 ka. What is immediately apparent from these ages is that there were three distinct phases of activity in the JVVF. The earliest phase, around 5.5 Ma is represented only by a single JVVF vent (Vent 4569), although other vents of this general age are identified on the Owyhee Plateau south of the JVVF (Hart et al. 1984, Shoemaker, 2004). The period from around 3 Ma to 1.8 Ma appears to have been one of significant activity, with 5 vents erupting in this duration. What is very interesting is the fact that four vents, three of which are widely separated from each other, erupted simultaneously around 1.9 Ma. As will be seen in the subsequent section, these vents are compositionally diverse, which demands an explanation. The final phase of eruptive activity, represented by five vents, began around 250 ka and continued until less than 1.5 ka. This period of time is roughly equivalent to the maximum errors on ages of vents representing the pulse of activity around 1.9 Ma (0.08-0.23 m.y.). It could be said, in that context, that the eruptions constituting the most recent phase of activity might also be construed to form a pulse of activity. It appears highly likely that this area might experience eruptions in the near future.

112 Petrography and Geochemistry During fieldwork, a comprehensive sampling program was undertaken in order to collect samples for petrographic and geochemical characterization. In light of the possibility of within- vent compositional heterogeneity, numerous samples were collected for most of the volcanoes. These included vent-related material as well as proximal and distal portions of the lava flows. Two phases of eruptive activity, closely associated in time had previously been identified for the youngest vent, Coffeepot Crater. This vent and associated flow field had been subjected to high- density sampling (>30 samples) and analyzed for its chemical composition (Hart et al. 1992; Hart, unpub data). Details of the analytical methods employed are presented in Appendix 1. Sample locations and descriptions are presented in Appendix 3 and geochemical data are in Appendix 4. Data for representative samples from each vent are in Table 2. Figure 1 in Appendix 4 shows the locations of the samples on the satellite image of the JVVF, with numbers corresponding to those in Table 2 and Appendix 4. Geochemical data obtained during this study substantially augment previously available analyses for this area, some of which have been presented in Hart and Mertzman (1983), Hart (1985), Hart et al. (1992), Shoemaker and Hart (2002), and Shoemaker (2004).

Petrography Samples from vents such as Owyhee Butte and Bogus Bench vent and associated lava flows are characterized by a diktytaxitic texture, which is typical of HAOT flows throughout the northwestern United States. These samples are holocrystalline, olivine phyric (with rare plagioclase phenocrysts) and have a groundmass of plagioclase, clinopyroxene, olivine and oxides. Sub-ophitic and intergranular texures are common. The lone SROT-like vent (Vent 4569) shows sparse olivine phenocrysts in a sub-ophitic groundmass of plagioclase and clinopyroxene, along with scattered olivine and oxides. Vents such as Rocky Butte and Coffeepot Crater are also olivine phyric (usually with fresher and slightly larger olivines) although scattered plagioclase phenocrysts are abundant in some flows. Olivine and plagioclase often occur as glomerocrysts. A flow within the younger crater on Rocky Butte contains plagioclase phenocrysts up to 1 cm in length, which are larger than in any other flow from the JVVF. The West Crater has a distinctive glomeroporphyritic texture, which makes it possible to recognize its flows even from isolated outcrops within the Owyhee Canyon. Most West Crater flows are markedly inequigranular with

113 a wide range in sizes of phenocrysts as well as groundmass crystals. Clinopyroxenes in some of these mildly alkaline flows, particularly those belonging to Clarks Butte and Rocky Butte, are distinctly purple-brown in color, and inferred to be titanium-rich (titan-augite). Textural information for specific samples is presented in Appendix. 3.

Major and Trace-element characteristics (excluding Rare Earth Elements) Figure 4 shows the data plotted on a Total Alkalis vs. Silica (TAS) diagram of Le Bas et al.

(1986) as well as on the Zr/TiO2 vs. Nb/Y diagram of Winchester and Floyd (1977). It is apparent that most of the older samples (5.6-0.44 Ma) are subalkaline, whereas the younger samples (< 0.25 Ma) are mildly alkaline. All samples, except one sample lie in the ‘Basalt’ field on the TAS diagram, whereas the ‘Trachybasalt’ field is represented by one sample (Fig. 4a). All the younger samples plot in the alkali basalt field in Figure 4b. The Mg-numbers [Mg2+ / (Mg2++Fe2+)] of the samples range from 51-65. These values indicate that the JVVF basalts have not undergone prolonged or substantial post generation differentiation. This is further corroborated by the fact that silica does not show systematic variation with Mg number, and

several samples with Mg numbers > 60 have widely varying TiO2 contents (Fig. 5a, b). Alumina also does not show significant variation with Mg number (Fig. 5c). The plot of total alkalis vs. Mg number (Fig. 5d) is interesting in that two sub-parallel trends can be discerned. Samples with the same Mg number show different values for total alkalis, which is not consistent with the same/similar magmas undergoing fractional crystallization. It is likely that other processes must have been involved in the petrogenesis of the JVVF basalts. As will be seen later, the trace element concentrations and isotopic compositions lend credence to this contention. The geochemical data help identify the following broad compositional types, as shown in Figure 6: a) HAOT b) SROT-like c) AB (mildly alkaline basalts) d) TB; basalts transitional between HAOT and SROT These types are essentially those identified by Hart and Mertzman (1983) in this area, with the exception of SROT, which is represented in the present study by one vent (Vent 4569; sample N2/03 in Table2). Shoemaker (2004), in his study of the main Owyhee Plateau south of

114 the JVVF also found a close association between HAOT, SROT-like basalts and TB. AB, however do not occur in that region. On the other hand, young, mildly alkaline basalts have been reported from the Melba area in the Western Snake River Plain by White et al. (2002), which erupted between around 2 Ma and 0.4 Ma. The sequence in the Melba area is quite similar to that from the JVVF – strongly tholeiitic lavas as well as mildly alkaline lavas erupted in the same area after 2 Ma. However, in the tholeiites in the Melba area are more strongly differentiated than similar-aged tholeiites in the JVVF. In Figure 7, four trace elements of varying incompatibilities have been plotted against

TiO2. It is quite apparent from this figure that most vents have unique compositions, and data form distinct clusters. Two broad compositional groups can readily be discerned in the plot of Sr vs. TiO2 (Fig. 7a) and without much difficulty in the plot of Nb vs. TiO2 (Fig. 7b). One group (Group 1) is characterized by <300 ppm Sr (except for one outlier with a little over 300 ppm) and

< 18 ppm Nb. Samples from this group have a much wider range in TiO2 contents and define a linear trend in these and many other plots. The second group (Group 2) has > 400 ppm Sr and >

18 ppm Nb, a narrower range in TiO2, and lacks linear compositional trends. In Figure 7a, Group 2 appears to have two sub-groups, one with 450-550 ppm Sr and the other with > 600 ppm Sr, represented by the Coffeepot Crater Phase 1(Table 1). In plots of Rb and Zr vs. TiO2 (Fig. 7c, d), these groups are not immediately apparent – however, a closer look indicates that one set of samples tend to define a linear trend whereas the other set does not. These trace element data are supported by the total alkalis vs. Mg number plot, which also appears to indicate the presence of two groups (Fig. 5d). The geochronological information discussed earlier reveals that Group 1 consists of vents belonging to the two older phases of activity (HAOT, SROT-like and TB) from around 6 Ma to 0.44 Ma (Table 1). Among these, samples associated with the Owyhee Butte and the Bogus Bench Vent show near-identical major and trace element compositions. Group 2 consists of vents from the youngest phase of activity (AB) from around 0.25 Ma to Recent. This indicates a temporal control (among other possible factors) on the compositions constituting each group. Among the Group 2 vents, the Coffeepot Crater and Rocky Butte show considerable internal diversity. For the former, this takes the form of two distinct groups – data for only selected samples for this vent have been plotted, but the two groups are still visible in Figure 7. On the other hand, the Rocky Butte shows a scatter of compositions. It is not clear whether the

115 more limited range in compositions of other vents is a result of limited sampling and whether more extensive sampling would reveal a similar level diversity. However, the randomly sampled and widely distributed West Crater basalts (including two vent samples, one proximal flow and three distal flows) show similar compositions. This could suggest that certain JVVF vents are compositionally more heterogeneous than others. Some flows correlated with the Rocky Butte vent might also have erupted from an older vent that is now covered by younger flows – this interpretation is strengthened by the presence of several flows from the Owyhee Canyon that show a Rocky Butte signature, although flows from this vent cannot be physically traced even close to the canyon. This hypothesis needs to be tested with additional fieldwork supplemented by paleomagnetic and geochronologic data in order to be able to better interpret issues of within- vent heterogeneity.

Rare Earth Element(REE) and isotopic characteristics Selected samples were analyzed for REE as well as for Sr, Nd and Pb isotopic compositions. At least one sample was analyzed for each vent, and more than one sample was analyzed for vents that have heterogeneous compositions in terms of major and trace elements. Figure 8 shows chondrite-normalized plots of REE for the analyzed samples. The patterns range from essentially flat to moderately steep, reflecting LREE enrichment to varying degrees. Except for the lone SROT-like vent (Vent 4569), all samples have near uniform HREE contents. Flat REE patterns with a positive Eu anomaly are shown by the HAOT vents that are chemically the most primitive, viz. the Owyhee Butte and the Bogus Bench Vent (Fig. 8). Such patterns are characteristic of HAOT all over the northwestern United States (e.g. Hart, 1982, McKee et al. 1983), and are strikingly similar to those for back-arc basin basalts (BABB; Fryer et al. 1981). Sinton et al. (2003) mention that such flat REE patterns at total abundances varying from ~ 5 to 20 times chondritic values appear to be characteristic of all back-arc basins, despite widely varying REE patterns in associated island arcs. Hart (1982), Hart et al. (1984), and McKee et al. (1983) pointed out this similarity of HAOT with BABB and MORB, and suggested that it indicates magma genesis in a similar tectonic setting. The AB show varying degrees of LREE enrichment, with La values for most vents ~ 40 times higher than chondritic values. Only one vent, Clarks Butte has more enriched La values (Fig. 8), which is in agreement with its higher Rb and Nb concentrations. HREE enrichment for all vents, on the other hand, ranges from ~ 8 to 10

116 times the chondritic values. Vent 4569 shows a pattern with similar degrees of LREE enrichment as most AB samples, but uniformly higher MREE and HREE contents. TB patterns are slightly shallower than those for the AB, but otherwise similar. Isotopic data for the JVVF samples are presented in Figures 9, 10 and 11. 87Sr/86Sr values range from 0.70382 to 0.70626, although all except one sample have values less than 0.7055 (Fig. 9b). 143Nd/144Nd values, as expected, vary inversely with Sr-isotopic composition. Once again, samples from the Owyhee Butte and Bogus Bench Vent are near identical, forming an outlier with relatively high 143Nd/144Nd as well as 87Sr/86Sr values (Fig. 9b). Vent 4569 has the highest 87Sr/86Sr and lowest 143Nd/144Nd, which lie within the field for basalts from the Snake River Plain (Fig. 9a). It also has a Pb isotopic composition more radiogenic than all other samples (Fig. 10), and such values have not been reported for basalts from either the Owyhee Plateau (Shoemaker, 2004) or from the Snake River Plain (e.g. Hart, 1985). The 206Pb/204Pb value is within the range of values reported for SCLM of the Wyoming Craton (Carlson and Irving, 1994), ocean island basalts (OIB; e.g. Zindler and Hart, 1986), and average upper continental crust (e.g. Zartman and Haines, 1988). The 207Pb/204Pb and 208Pb/204Pb values are higher than those of OIB but within the SCLM field. Note that the range in isotopic compositions of SCLM is quite large, and no specific information exists as to the nature of the SCLM in the JVVF region. The AB vents show slightly increasing 207Pb/204Pb and 208Pb/204Pb with increasing 87Sr/86Sr, but slightly decreasing 206Pb/204Pb with increasing 87Sr/86Sr. Within-vent isotopic heterogeneity has not been observed for HAOT and TB vents, but is displayed by two of the AB vents, viz. Coffeepot Crater and Rocky Butte. 87Sr/86Sr values were measured for 16 Coffeepot Crater samples, and 143Nd/144Nd for 5 samples, representing both phases of activity. Similar to the trace element data discussed earlier, the Sr and Nd isotopic data define two distinct clusters (Fig. 9b; only selected samples shown). The two craters on Rocky Butte also have distinctly different 87Sr/86Sr, 143Nd/144Nd and slightly different 206Pb/204Pb values, but similar 207Pb/204Pb and 208Pb/204Pb values (Fig. 9, 10, 11).

SPATIAL AND TEMPORAL ASPECTS OF THE COMPOSITIONAL DATA

The relationship between composition, location and ages of the individual volcanoes was briefly alluded to in the previous sections, and is explored in detail in order to determine whether

117 there are any systematic trends. If such trends or patterns do indeed exist, they could shed additional light on the structural and geochemical evolution of the JVVF, as well as on length scales of source and/or process heterogeneity. In general, the eruption of the various compositional types (HAOT, SROT-like, TB, and AB) appears to be random in space. However, it is interesting to note that all AB vents have erupted along two distinct alignments, whereas the other types of vents have not necessarily done so. The temporal relationship has already been discussed in the geochronology section, and indicates that the AB belong to the younger phase of activity. Locations of vents and vent alignments on the satellite image are shown in Figure 2, whereas Figure 12 shows ages, and salient compositional characteristics on the same base image. The NNW-SSE trending Coffeepot Crater-Three Mile Hill alignment is the most long-lived of the three recognized alignments, and consists of four vents that range in age from 1.86 Ma to 0.013 Ma. It includes one TB vent (Three Mile Hill) and three AB vents (Rocky Butte, Clarks

Butte and Coffeepot Crater). These vents are compositionally diverse, with K2O ranging from 0.67 to 2.07 wt.%, Nb from 18-50 ppm, Sr from 305-675 ppm and 87Sr/86Sr from 0.70382 to 0.70543. The ages indicate that volcanism along this alignment is generally younger to the north but was not progressive. Rather, it occurred in a staggered fashion. In contrast with this alignment, the E-W trending West Crater-Skinner Hill alignment seems to have been shorter lived, with vents ranging in age from 0.06 Ma to <0.03 Ma (although the possibility of future eruptions occurring along it is not precluded). The age of the Skinner Hill vent is not known; however, the geomorphic expression of its flows is similar to those of West Crater, indicating a similar to slightly older age. The three vents along this alignment are all AB (West Crater, Rocky Butte and Skinner Hill), and compositionally less diverse than the earlier alignment. Their K2O contents range from 0.95-1.27 wt.%, Nb from 21-37 ppm, Sr from 445-542 ppm and 87Sr/86Sr from 0.70450-0.70498. It should be noted that if some Owyhee Canyon-filling flows are ultimately determined to belong to West Crater/Rocky Butte, the range of K2O extends up to 1.45 wt.%. Vents along the final recognized alignment (the E-W trending Owyhee Butte-Three Mile Hill alignment) range in age from 2.78-1.86 Ma. This alignment was thus active for an interval intermediate between the other two alignments. What is rather interesting is that two of the vents, viz. the Owyhee Butte and Three Mile Hill erupted at essentially the same time within error at

118 ~1.86 Ma. K2O contents of the constituent vents range from 0.07-0.34 wt.%, whereas Nb and Sr concentrations range from 5-16 ppm and 186-305 ppm respectively. 87Sr/86Sr values are quite similar, from 0.70527-0.70542, but their 143Nd/144Nd values are significantly different (0.512740 and 0.512617 respectively), precluding their being related to each other simply by fractional crystallization. It is appropriate here to mention that vents aligned in a NW-SE direction, similar to the Coffeepot Crater-Three Mile Hill alignment discussed above have also been recognized from the Owyhee Plateau by Shoemaker (2004). Shoemaker notes that two of these alignments, the NW- SE Grassy Mountain (GM) and Northern Nevada (NN) alignments, contain vents with essentially indistinguishable eruption ages (i.e., within analytical error), but from which basalts of HAOT, TB and SROT compositions were erupted. These vents range in age from ~ 10 Ma to 7 Ma. This, in conjunction with the ages of vents along the Coffeepot Crater-Three Mile Hill alignment indicates that volcanism along such NW-SE to NNW-SSE alignments has been active since ~ 10 Ma. This has resulted in the eruption of a wide range of magma compositions. Only HAOT and SROT end-members appear to have been available in the earlier phase of activity on the Owyhee Plateau, whereas the younger phase in the JVVF involves HAOT, SROT as well as AB compositions. These vents could be rooted on deep lithospheric fractures (e.g. Korme et al. 1997) as noted by Shoemaker (2004), which in turn could reflect regional tectonic stresses. Apart from alignments in the JVVF, other relationships are also interesting and noteworthy. The Owyhee Butte and the Bogus Bench Vent are separated by ~ 16 kilometers, and do not lie along any obvious alignment (Fig. 2, 3). However, both vents erupted within error at around 1.86 Ma and have identical elemental as well as isotopic compositions (Fig. 8-12). Although relatively primitive HAOT lavas have very similar elemental compositions throughout the northwestern United States, they are not necessarily isotopically identical. The near identical ages and compositions of these two vents suggest two possible scenarios. The first is that magmas for each vent were derived independently from near identical sources and underwent the same processes en route to the surface. The second is that only one magma batch was generated but was erupted at two different locations depending on easiest access to the surface. It is of course possible that there was an interval of a few hundred or few thousand years between the formation of these volcanoes – it is not possible to distinguish this on the basis of available age data / field relationships. This does not affect the first scenario discussed above. The feasibility

119 of the second scenario, however, could potentially be affected, particularly if the volume of magma generated or intruded at depth was small. This is so because small volumes of magma are prone to cool and crystallize (the specific times depend on the geothermal gradient, latent heat of crystallization, and thermal diffusivity) relatively rapidly. For example, consider the simple conductive cooling equation of Jaegar (1968): t = τ a2/31.5, where ‘t’ is the time required for a given amount of cooling, ‘τ’ is nondimensional time (t a2/31.5), and ‘a’ is the radius of a cylinder or sphere, or half thickness of a sheet. For τ = 0.1, cooling will have penetrated to the center of the body and for τ = 1, there is substantial cooling at the center of the body (Best and Christiansen, 2001). Using this to model cooling of a 200 m thick sheet of magma suggests that it should cool substantially in just over 300 years. However, even when cooling has just penetrated the center (τ = 0.1), which will happen in just over 30 years, the magma should have crystallized and the composition of the residual liquid (i.e. eruptible liquid) will have changed. Note that if this sheet were 100 km2 in areal extent, it would amount to a volume of 5 km3. Assuming 1/10th of this volume were erupted, about 1 km3 of lava would result. This is similar to the maximum eruptive volume of the Owyhee Butte, which is probably in the range of 1.25 km3. This modeling is undoubtedly simplistic, particularly since little is known about what volume of magma is generated under monogenetic volcanoes and what percent of it is extruded. Is there, for example, an extensive zone of melting and intrusive complex below the JVVF which is randomly tapped by individual volcanoes? Or is there a limited zone of melting below each volcano? It is difficult to comment on this in the absence / extreme paucity of geophysical data from this region of the northwestern United States. Similarly, the generation of latent heat of crystallization will increase the duration of cooling. The pulse of activity around 1.9 Ma also includes the Little Owyhee Butte. This vent is <5 kilometers from Owyhee Butte; yet, it is compositionally distinct. It is TB in character, with 87 86 higher TiO2, Sr and Nb (Fig. 5-7) than Owyhee Butte, but interestingly is very similar in Sr/ Sr (0.70528) to Owyhee Butte (0.70527). However, its 143Nd/144Nd (0.512620) is much lower than that of Owyhee Butte (0.512740), precluding fractional crystallization of the Owyhee Butte magma in its genesis. The Little Owyhee Butte is also isotopically almost identical to the Three Mile Hill. The possibility of these two being related to each other by fractional crystallization will be discussed in the section on petrogenesis.

120 IMPLICATIONS AND DISCUSSION

Petrogenesis As mentioned at the outset, a comprehensive petrogenetic evaluation of the JVVF lavas is not the primary aim of this study. However, a discussion of petrogenesis is expected to be useful from the perspective of the evolution of MVF in general, particularly in illuminating source and process related effects on monogenetic volcanism. This requires an evaluation of the chemical and isotopic data in their spatial and temporal context. Salient aspects of field, compositional and geochronologic data presented in the previous sections can be summarized as follows: • The JVVF consists of numerous monogenetic volcanoes (vents and associated flows), some of which were erupted along distinct alignments • Volcanism in the JVVF ranges in age from ~ 5.5 Ma to < 1.5 Ka, and occurred in three distinct phases. The first phase of activity is represented only by one vent, whereas the two subsequent phases are represented by six and five vents respectively. • Two overall compositional and temporal groups are recognized – an older (~5.5 Ma-0.44 Ma) group of HAOT, SROT-like and TB vents, and a younger (<0.25 Ma) group of AB vents (Fig. 5-11). This suggests that three fundamental magma types (HAOT, SROT-like, and AB) were involved in the JVVF volcanism. • Most vents show distinct chemical and/or isotopic compositions; within-vent chemical and isotopic diversity is exhibited by two of the AB vents (Coffeepot Crater and Rocky Butte; Fig. 7-11). • There is no systematic relationship between composition and location (Fig. 12). Diverse compositions were erupted along the various alignments. A pair of vents separated by a few kilometers may show diverse compositions, whereas another pair of vents separated by > 10 km may show identical compositions. A pulse of activity around 1.9 Ma generated four volcanoes with diverse compositions.

Figures 13 and 14 present additional plots that help highlight the key issues that must be considered while discussing the petrogenesis of the JVVF lavas. Figures 13a and 13b are plots of La/Nb vs. Ba and La/Nb vs. Sr/Y respectively. Both La and Nb are incompatible elements,

121 and the ratios of such elements can provide insights into the characteristics of sources. However, in order to derive such insights, it is important to ascertain that the elements in consideration had near identical incompatibility. This can be gauged from plots of log of one element versus log of the other element (Sims and DePaolo, 1997). On such a plot, elements with identical incompatibility form linear trends with slopes of 1. Plots of log La vs. log Nb satisfy this condition only for Group 2 samples (slope of 0.95), but not for Group 1 samples (slope of 1.4). This suggests that La/Nb of the Group 2 samples could be inherited from the source. Fig 13a shows that the La/Nb ratio for Group 2 vents remains nearly constant, but these vents show a wide range in Ba concentrations. On the other hand, Group 1 samples show a crude positive correlation between La/Nb and Ba. If these characteristics are inherited from the sources and have not been modified by differentiation processes, then this suggests that the Group 2 magmas were derived from a source / sources with relatively uniform La/Nb ratios. The range in Ba would then reflect either variable enrichment of the source in Ba, or a melting / differentiation process that leads to Ba enrichment. Fig. 13b shows that Group 1 samples have relatively constant Sr/Y ratios, but Group 2 samples have widely varying Sr/Y ratios. This is primarily because Sr and Y in Group 1 samples vary systematically, whereas they don’t in the case of Group 2 samples. These characteristics are consistent with other plots presented earlier (e.g. Fig. 7) where Group 1 samples show well-defined trends, whereas Group 2 samples do not. Figures 13c and 13d show plots of Zr vs. Nb and Zr vs. Y. The sub-parallel trends in Figure 13d and the trends with distinctly different slopes in Figure 13c are strong evidence for the existence of two petrogenetically distinct groups, which must have been derived from fundamentally different sources / processes. The Group 2 samples do appear to show a systematic trend in Figures 13e and 13f where Rb concentrations increase with increasing 87Sr/86Sr, while Ni concentrations decrease. Such trends often are considered to point to the involvement of assimilation accompanying crystallization (AFC) or assimilation of chemically and isotopically evolved crustal components. Interestingly many Group 1 samples (HAOT lavas represented by the Cow Vent Complex, TB and SROT-like lavas) show a systematic increase in Sr concentration with increasing 87Sr/86Sr. This agrees with nearly linear trends observed in Figure 5, 7 and 13b, all of which point to the important role of source or magma mixing in the generation of the Group 1 suite (as previously suggested by Hart, 1985).

122 Figure 14a shows the variation in 143Nd/144Nd with age. While there appears to be a general trend of increasing 143Nd/144Nd with decreasing age of samples, this relationship is by no means simple. For example, at ~ 2Ma, HAOT and TB vents with very different 143Nd/144Nd were erupted. Similarly, in the 0.5-0.25 Ma interval, a HAOT and an AB vent with very different 143Nd/144Nd were erupted. Group 2 vents, however, show a systematic increase in 143Nd/144Nd with decreasing age. These vents also show systematic decreases in Rb and 87Sr/86Sr from oldest (Clarks Butte) to youngest (Coffeepot Crater) (not shown). Such trends in a single volcano could be interpreted to result from the tapping through time of a stratified magma chamber that had undergone open system (AFC) evolution. The question in this case is whether such a process is likely to lead to the eruption of discrete monogenetic volcanoes. Certain other elements such as Zr, however, do not show any systematic relationship with age (Fig. 14b), even for Group 2 vents. Such apparently contradictory indications are taken to suggest that the Group 2 AB volcanoes are not simply linked to one long-lived magmatic system. The relationships illustrated in Figures 13 and 14 serve to emphasize that source hetereogeneity and complex melting processes, coupled with open system differentiation processes involving discrete magma batches, are responsible for generating compositional diversity within the JVVF lavas. While both source and process effects are implicated, evaluation of the data presented thus far in the context of previous studies (e.g. Hart et al. 1984; Hart, 1985; Leeman et al. 1992; Hart et al. 1997; Shoemaker and Hart, 2002) strongly suggests that the two identified JVVF magmatic groups (Group 1, HAOT-TB-SROT and Group 2, AB) are, to a first order, present because of the availability of three endmember primary magma types; HAOT, SROT, and AB. It is clear, however, that HAOT and SROT-like magma types were erupted in the earlier period of volcanism in the JVVF, whereas the AB magma type has erupted only during the most recent pulse. It should be noted that the source reservoirs and processes involved in the production of these endmember magma types are heterogeneous at least with respect to trace element and isotope parameters, but appear to be capable of producing three fundamentally similar bulk compositions over time. For example, HAOT and SROT endmembers have been feeding Owyhee Plateau and JVVF monogenetic systems since approximately 11 Ma. Below, the generation of each of these dominant magma types is briefly discussed.

123 Primary Magma generation HAOT Experimental work (Bartels et al. 1991) shows that primitive HAOT magmas could have been generated by melting of an anhydrous spinel peridotite mantle source at relatively shallow depths of 30-35 km or crystallization of a more primitive magma en route through the lithosphere (Carlson and Hart, 1987; Hart et al. 1997). This is further strengthened by the presence of positive Eu anomalies in almost all HAOT lavas (e.g. McKee et al. 1983; Hart, 1985; this study). Such positive Eu anomalies could indicate a role for derivation of some volume of melt from plagioclase peridotite – a process that could occur near the transition between the plagioclase and spinel stability zones. Elkins et al. (2001), on the other hand in their study of HAOT from a transect across the Cascades determine depths of melting ranging from 36 km-66 km. The overall composition of HAOT is rather similar to BABB irrespective of their geographic location (e.g. McKee et al. 1983; Hart et al. 1984). Such BABB signatures could be related to asthenospheric mantle enriched by slab-derived fluids, or SCLM enriched in a similar fashion. It is possible that melt was generated within the shallowest parts of the asthenosphere, but finally equilibrated within the SCLM (thus giving rise to the plagioclase signature). The radiogenic Sr isotopic compositions of HAOT lavas throughout the Oregon Plateau require some contribution from the lithospheric mantle. Indeed, Hart et al. (1997) on the basis of radiogenic Os isotopic signatures within primitive HAOT invoked a similar scenario. The component within the SCLM contributing the radiogenic Os, according to them, is mafic in nature – such a component could result from intrusion over time of small volumes of asthenosphere-derived magmas within the SCLM, probably at a deep level of neutral buoyancy. Hart et al. (1997) propose that this interaction with the SCLM was not accompanied by significant fractional crystallization, at least for the more primitive HAOT. Hart (1982, 1985) suggested that HAOT originate by relatively high degrees of partial melting (10-30%) of lherzolitic upper mantle. Given the similarity of HAOT compositions with BABB and MORB, and their remarkable chemical homogeneity (see Hart et al. 1997), this study used REE to model partial melting of an estimated depleted MORB mantle (DMM). Such an estimate as well as distribution coefficients involved have been compiled recently by Workman and Hart (2005). Using a modal batch melting model, REE patterns similar to N8/03 (JVVF HAOT) were produced by 7.5% melting of this DMM. However, the modeled La and Ce

124 concentrations are much lower than in N8/03. This is not surprising given the fact that BABB sources are likely to be enriched in LREE, just as they are in Ba and Sr. When La and Ce values for ‘Enriched’ DMM provided by Workman and Hart (2005) are used, the match is much better, except for the Eu anomaly in N8/03, which either indicates the presence of plagioclase in the source, or shallow last equilibration as discussed earlier (Fig. 15).

SROT The petrology and petrogenesis of SROT magmas have been discussed in detail by workers such as Leeman (1974, 1982a, 1982b), Hart (1985), and Reid (1995). Hart (1985) suggested that SROT originate by partial melting degrees similar to HAOT (10-30%) of lherzolitic upper mantle (probably SCLM), which is enriched in iron as compared with that generating HAOT. The lone SROT-like vent from the JVVF (Vent 4569, sample N2/03) has the (T) highest Fe2O3 contents of any vent from the JVVF, consistent with this contention. It also has a uniformly enriched REE pattern, as opposed to LREE enrichment but HAOT-like HREE concentrations in TB and AB. Hart and Mertzman (1981) showed that different degrees of partial melting of the same source cannot reproduce the incompatible element variations between HAOT and SROT end members. The highly radiogenic isotopic composition of N2/03 (Figures 9-11) is strong evidence for generation from a different source.

AB Young AB are rare in the Oregon Plateau area; in fact, the JVVF AB are among the only such lavas from this region and their petrogenesis has not been looked at in detail. Shoemaker (2004) highlighted the fact that their 87Sr/86Sr values are lower, while Rb/Y ratios are higher than older tholeiitic lavas. On the basis of their generally less radiogenic isotopic composition (tending towards MORB values), he suggested that they reflect derivation from a sublithospheric source at depths greater than those involved in HAOT genesis. The alkaline nature of these lavas as well as enrichment in LREE and LILE (Fig. 8) hint at low degrees of partial melting. Their HREE concentrations are very similar to the HAOT, leading to the question as to whether garnet was involved in stabilizing these during melting. The involvement of garnet, however, should

lead to high chondrite-normalized La/SmN and Sm/YbN ratios (e.g. Wilbold and Stracke, 2005).

Some JVVF lavas have relatively high La/SmN, but low Sm/YbN values, which do not point to

125 the role of garnet in any of the JVVF lavas (Fig. 16). The source is therefore likely to be in the spinel stability field. Shoemaker suggested that their mildly alkaline nature and LILE enrichment could reflect melting of asthenospheric mantle that has been enriched in LILE by metasomatic fluids. As noted previously, a range in AB chemical and isotopic compositions exists. Relatively high Mg# and low 87Sr/86Sr, among other characteristics, of the very young eruptions at Coffeepot Crater suggest that these compositions are most representative of an endmember JVVF mildly alkaline (AB) magma type.

Factors producing compositional diversity

The above discussion, as well as geochemical data presented earlier, suggests that the contribution of different mantle sources was an important factor in the petrogenesis of the JVVF lavas. However, a host of processes have been invoked for other MVF, including fractional crystallization (Nemeth et al. 2003) and lower crustal contamination (Glazner et al. 1991; Ramos and Reid, 2005). It is therefore prudent to evaluate the effects of and relative contributions of the following factors in generating compositional diversity in the JVVF lavas: (a) fractional crystallization, (b) different degrees of partial melting, (c) assimilation and fractional crystallization (AFC), and (d) source heterogeneity (including mixing of compositionally different sources or magmas from such sources). In view of the fact that many of the lavas are isotopically distinct, they cannot therefore be related to each other simply by different degrees of partial melting of the same source. However, it is evident from regional studies of isotopic composition and efforts to link these to mantle structure (e.g. Leeman et al. 1992) that the SCLM beneath the JVVF is complex with respect to age, and chemical and isotopic composition. It is the precise spatial disposition of this complexity at smaller scales that is not clear. Nevertheless, it is likely to have a vertical as well as lateral component. In light of this, it is possible to visualize a rising melt column that aggregates melts generated from this complex SCLM. Reiners (2002), for example, has discussed in detail the mixing of large degree partial melts of pyroxenite (which melts at greater depths) with small degree partial melts of peridotite (which melts at shallower depths) in order to explain isotopic diversity within monogenetic and short- lived polygenetic volcanoes. Such a process is able to reproduce the decreasing incompatible

126 element concentrations but increasingly radiogenic 87Sr/86Sr through time in the eruptions Reiners studied. Such a process, however, is more akin to mixing of melts generated from different sources, and will be discussed in a subsequent section dealing with that. In the following discussion, geochemical data for the JVVF will be evaluated in the context of each of the remaining factors.

Fractional crystallization As discussed earlier, geochemical data suggest that most basalts from the JVVF are relatively primitive and their major and trace element characteristics clearly indicate that they did not evolve from fractional crystallization of the same magma. None of the basalts have MgO and Ni contents that would indicate their being primary mantle melts. Some of the vents, such as the Owyhee Butte, Bogus Bench Vent, and Coffeepot Crater with relatively high MgO (>9%), Ni (>150 ppm), and Cr (>250 ppm) could represent mantle-derived melts that underwent minimal

amounts of olivine fractionation. Decreases in CaO/Al2O3 with Mg-number (not shown) for the JVVF basalts hint at removal of clinopyroxene. Clinopyroxene phenocrysts are absent or rare in these lavas as mentioned earlier, although this could be the result of efficient removal of the phenocrysts during fractional crystallization at substantial depth, for example. In order to derive additional insights into the depths of differentiation of the JVVF magmas and to determine whether clinopyroxene fractionation is feasible, the ALFE projection of Reid et al. (1989) and the companion plot of CaO/Al2O3 versus Fe-factor are used. The ALFE projection is an X-Y plot depicting Fe-factor (defined as molar [FeO*/ (FeO*+MgO)], where FeO* has been reduced by

the amount of FeO in ilmenite) versus Al-factor (molar [(Al2O3-CaO*-Na2OK2O) / (FeO*+MgO)], where CaO* has been reduced by the amount of CaO in apatite), which can eliminate the effects of additions or subtractions of feldspar of any composition. The CaO/Al2O3 versus Fe-factor companion plots can discriminate the nature of fractionation: decreasing

CaO/Al2O3 with increasing Fe-factor in cogenetic magmas indicates clinopyroxene fractionation at either high or low pressure, while increasing CaO/Al2O3 with increasing Fe-factor indicates olivine + plagioclase fractionation. Figure 17 shows the JVVF data plotted on these diagrams. It is apparent from Figure 17a that the data project between the high and low pressure cotectics but favoring the higher pressure (up to 10 kb) scenario. While broadly similar relationships were noted by Shoemaker (2004) for

127 the Owyhee Plateau basalts south of the JVVF, a couple of key differences are noted: (a) The two distinct JVVF groups (1 and 2) project approximately as sub-horizontal groups offset by Al- factor, and (b) 10 kb cotectic crystallization trends are defined less well by the JVVF data. The observation that few JVVF samples parallel the 10 kb trend of decreasing Al-factor and that materials erupted from individual vents tend to form data “clusters” could indicate that magmas feeding individual vents crystallized from distinct parental melts that had lost olivine. Trends of crystallization of different source compositions (enriched in iron, for example), could be displaced towards higher Fe-factors. Regardless of these potential complexities, the position of the JVVF samples on this plot argues for primary crystallization of these magmas at depth

(Group 2 deeper than Group 1). Furthermore, the decreasing CaO/Al2O3 ratio with increasing Fe- factor (Fig. 17b) is suggestive of extraction of clinopyroxene at depth. Systematic within-vent decreases in Al- and Fe-factors are also observed (e.g. Rocky Butte), though, which could point at a role for continued cotectic crystallization at shallower levels. Olivine phenocrysts are common in these basalts, and the decreasing Ni with decreasing MgO (not shown) does suggest an important role for fractional crystallization of olivine. Plagioclase phenocrysts and glomerocrysts are also common. However, while Sr concentrations decrease with Mg number, alumina is either constant or even increases for the younger vents (Fig. 5c). Fractional crystallization dominated by plagioclase (e.g. shallow crustal), therefore, does not appear to have been a significant process. Numerous previous studies have discussed the fact that TB compositions cannot be generated from HAOT magmas by fractional crystallization (e.g. Hart and Mertzman, 1981; Hart et al. 1984), and this will not be discussed here. The incompatible element concentrations of AB are even higher than the TB and therefore cannot have been derived from HAOT by realistic degrees of fractional crystallization. This conclusion is supported by the distinct trends and groups illustrated in various plots of Figure 13. It was mentioned in the discussion of spatial and temporal aspects of the geochemical data that the TB vents Little Owyhee Butte and Three Mile Hill are isotopically nearly identical (Fig. 9-11) and erupted simultaneously around 1.86 Ma. The (T) latter shows a more evolved composition, with lower MgO and higher TiO2 and Fe2O3 (Fig. 5) It also has slightly higher incompatible trace element concentrations (Fig. 7). However, major element modeling suggests that the bulk composition of Three Mile Hill is not reproduced well by deep fractional crystallization from a Little Owyhee Butte parental magma (at 8 kb).

128 Although there are no indications of significant shallow crystallization, it is worth mentioning that models simulating 1-5 kb fractional crystallization fare even worse, leading to significantly iron enriched compositions. Bernie’s Butte is an older vent (2.78 Ma) but nearly identical in composition to Little Owyhee Butte. It has a lower Mg number, slightly higher TiO2, lower alumina (Fig. 5) and lower Sr (Fig. 7) than Little Owyhee Butte. Removal of olivine and plagioclase can help explain these variations – however, the fact that a million years passed between the eruption of these lavas makes a scenario involving fractional crystallization of the same magma highly unlikely. Compositionally similar end-member primary magmas were likely available in the JVVF area throughout its eruptive history; thus slightly different amounts of olivine and plagioclase fractionation from such magmas could produce the characteristics observed at Bernie’s Butte and Little Owyhee Butte.

Assimilation and fractional crystallization (AFC) Monogenetic basaltic volcanism, until recently, was considered to involve relatively rapid transport of mantle-derived magma to the surface. However, several studies have shown that such magmas may pond at various levels within the lithosphere (e.g. Glazner et al. 1991; Klugel, 1998; Nemeth et al. 2003) and hence are in principle subject to crustal assimilation. This is likely to be accompanied by fractional crystallization, which has been recently demonstrated to have been an important process in some MVF (e.g. Nemeth et al. 2003). A well-known example of the operation of crustal assimilation during monogenetic volcanism is Paricutin – a calk-alkaline volcano in Mexico that erupted lavas ranging in composition from early basaltic andesites to late andesites for a period of nine years (Wilcox, 1954; McBirney et al. 1987). The latter andesites show geochemical evidence for assimilation of granitic crust. McBirney et al. (1987) invoked preferential tapping of dense but hotter underlying liquid prior to the overlying andesitic cap from a stable, zoned magma chamber. Such zonation would have existed long before actual eruption. However, Dungan (2005) suggests that the assimilation could have occurred as the eruption progressed, by incorporation of crustal material along the margins of the reservoir or conduit. This hints at relatively high rates of digestion of granitic xenoliths. Lower crustal assimilation was invoked by Glazner et al. (1991) for the Pisgah Crater cinder cone and associated lava flows to explain within-vent compositional heterogeneity. This vent was recently subjected to a detailed study by Ramos and Reid (2005). Using Sr isotopic

129 measurements on single crystals and groundmass, these authors demonstrated that the geochemical trends through time at the Pisgah Crater involving decreasing incompatible element concentrations with increasingly radiogenic 87Sr/86Sr could be explained by lower crustal assimilation. White et al. (2002) suggest that the youngest lavas in the Melba area of the western Snake River Plain (fed from monogenetic vents) show evidence for small amounts of crustal assimilation accompanied by varying degrees of fractional crystallization. In light of these studies, as well as considering some of the plots discussed earlier (e.g. Fig. 13e and f), it is appropriate to evaluate whether compositional diversity within the JVVF lavas might be a result of AFC. Modeling by Hart (1985) showed that it is improbable that AFC was responsible for the generation of evolved TB and SROT-like magmas from HAOT magmas. Several key elements and oxides (and especially the lower silica in SROT as compared with HAOT) were not reproducible, even though a variety of fractionating phases and assimilants were used. The uniform chemical characteristics of HAOT irrespective of the underlying lithosphere also make it difficult to invoke AFC involving crustal material. The Coffeepot Crater eruptive products define two distinct chemical and isotopic groups (Fig. 7, 9-11). However, the compositional variations between these groups cannot be related to each other by simple FC or AFC (Hart et al. 1992; Bondre and Hart, 2004) – although crustal contamination and fractional crystallization are consistent with the data they are insufficient to explain the variation. A contribution from two different sources or magma types is necessary. The AB lavas from the JVVF when considered together, however, show certain characteristics that hint at the involvement of AFC in their evolution. These include increasing Rb (Fig. 13e) 87 86 and K2O, and decreasing MgO and Ni (Fig. 13f) with increasing Sr/ Sr. Other incompatible elements such as Nb and Ba also generally increase with increasing 87Sr/86Sr. The earlier discussion on the nature of fractional crystallization of the JVVF magmas strongly suggests that this process took place at depths corresponding to pressures of around 8 kb, i.e. at approximately 25 km. If fractional crystallization at this depth were accompanied by assimilation, then the assimilant is likely to have been lower crust or melts thereof. In the case of bulk mafic or pelitic components from the lower crust, the erupted lavas should show trends of decreasing incompatible elements with increasing 87Sr/86Sr (e.g. Ramos and Reid, 2005). This is the opposite of what is observed for the Group 2 vents, and the substantial involvement of bulk

130 mafic lower crust is therefore unlikely. Trace element and isotopic modeling using IgPet (Terra Softa, Inc.) shows that shallow AFC can generate the most evolved AB (Clarks Butte) from the least evolved one (Coffeepot Crater) only if high degrees of assimilation (30%) or fractional crystallization (>60%) are involved. Modeling using the PELE software (Boudreau, 1999) shows that such high degrees of fractional crystallization (~ 60%) at shallow depths result in a very different major element composition. Similarly, assimilation of large quantities of shallow bulk crustal lithologies would lead to enrichment in silica and K2O, which is not observed in the JVVF samples. Hence shallow AFC does not appear to be a feasible process for generating the full range of elemental and isotopic diversity in the Group 2 lavas. An independent line of reasoning for this is the fact that the youngest and oldest AB vents are separated by almost 0.25 M.y. In polygenetic systems, a relationship with the oldest material being the most evolved (Clarks Butte) and the youngest being the least evolved (Coffeepot crater) would be highly suggestive of eruption of the most fractionated and/or contaminated magmas first, followed by the most primitive ones. However, it is unclear whether such a scenario would operate in small- volume monogenetic systems unless large volumes of magma were continuously available at depth. There is no indication at present that such volumes are available beneath the JVVF. Similarly, in settings where AFC is suspected, more evolved products such as latites, dacites or rholites are often observed (e.g. Craters of the Moon) in close affinity with less differentiated basalts. There is no such young and evolved material in the JVVF. On the other hand, assuming that endmember primary AB magma was available beneath the JVVF for the past 250 ka, it is reasonable to suggest that individual batches of such magma could undergo variable differentiation histories leading to the heterogeneous compositions erupted from distinct AB vents. To test this idea two scenarios are modeled: 1) high pressure fractional crystallization (i.e. lower crustal depths) of the most primitive AB magma (similar to Coffeepot Crater), and 2) mixing of this fractionated liquid with silicic compositions taken to represent possible melts derived from the lower crust or from the walls of the magma conduit system. Combined PELE and IgPet modeling programs were used to simulate these differentiation histories. Using PELE, varying degrees of fractional crystallization of the Coffeepot Crater Phase 2 composition was modeled at 8 kb pressure (parameters in Table 3). The trend for the resulting liquid line of descent for model ‘A’ (Table 3) is depicted in the plots of Figure 18. It is apparent that the modeled deep fractional crystallization process can lead to

131 derivative liquids with broadly similar bulk compositions as those observed at, for example, Rocky Butte, but cannot lead to the more evolved Clarks Butte compositions. As previously discussed, isotopic distinctions also preclude a simple, crystallization-only differentiation history. The second modeled scenario uses the fractionated (residual) liquid composition derived above as the mafic end-member for simple binary mixing with melt of a rhyolitic composition (rhyolite from the JVVF area). This modeling is used in order to examine the possibility of producing the range of AB characteristics, particularly those of the oldest and most evolved AB lavas from Clarks Butte, via a coupled crystallization and mixing process. As illustrated in Figure 19, between 20-25% addition of this silicic melt to the fractionated Coffeepot Crater magma reproduces many aspects of the Clarks Butte eruptive products. While certainly by no means a precise fit to the data, the modeling results for major and trace elements suggest that open system processes acting on a common starting AB magma may contribute to the overall diversity of the JVVF Group 2 volcanism. However, the modeling fails to reproduce the Sr isotopic composition of Clarks Butte. In order to arrive at the Sr isotopic composition of this vent, the evolved mixing end-member would have to have >200 ppm Sr and more radiogenic 87Sr/86Sr than the rhyolitic melt used in the model. Evolved rhyolitic and rhyo-dacitic material from the Santa Rosa-Calico Volcanic Field just south of the Owyhee Plateau is also determined not to be an appropriate end-member. Using other compositions for crustal melts and varying crystallization conditions might lead to better results. However, it needs to be emphasized that in the absence of crustal xenoliths or petrographic evidence for mixing as well as good geophysical data, it is difficult to test the validity of this hypothesis more rigorously at this time.

Source heterogeneity, source mixing or magma mixing Basaltic magmas generally owe their origin to partial melting of ultramafic rocks in the mantle (although some studies show that in certain settings, mafic lithologies might be involved). Unlike in oceanic settings, however, identification and characterization of mantle sources in continental settings is complicated by the fact that these magmas originate in or pass through lithosphere that is variable in thickness and extremely variable in composition. Continental lithosphere includes the SCLM, lower continental crust and upper continental crust. Recognition of mantle heterogeneity requires that the magmas have had none or minimal interaction with the latter components, or that such interaction can be accounted for and “subtracted”. The erupted

132 compositions can also indicate mixing of two or more sources or of magmas derived from these sources. In such cases, contribution of each source needs to be quantitatively determined. Similarly, specific information regarding the nature of the SCLM needs to be available in order to distinguish magmas derived from it from those derived from the underlying asthenospheric mantle. Complete information of this kind is not yet available for the Oregon Plateau region. Nevertheless, certain interpretations regarding source heterogeneity and potential mixing can be made on the basis of the observed compositional diversity in the JVVF lavas. Shoemaker (2004) invoked the interaction of three different sources for the Owyhee Plateau basalts that he investigated. These sources correspond to the C1, C2 and C3 sources originally proposed by Carlson (1984) and discussed by Carlson and Hart (1987, 1988). C1 has low 87Sr/86Sr and high 143Nd/144Nd, and is similar to the source of northern hemisphere intraplate oceanic basalts. The 207Pb/204Pb of most JVVF samples is higher than for most intraplate oceanic basalts, although values for the Coffeepot Crater (Fig. 10) fall within or very close to the oceanic array as well as close to the C1 source composition of Carlson and Hart (1987; see their Fig. 7). C2 is the source component with significantly higher 206Pb/204Pb values than C1, and also more radiogenic 87Sr/86Sr (Fig. 11). This component likely represents the contamination of C1-type mantle with pelagic sediments in a back-arc setting, although this signature could also arise from crustal contamination. The C3 component is defined primarily by SROT compositions, and is characterized by the highest 87Sr/86Sr and lowest 143Nd/144Nd ratios. It also has high 207Pb/204Pb and 208Pb/204Pb, but relatively low 206Pb/204Pb and is presumed to represent SCLM. Although the enriched signature of the SCLM has been considered to be related to modification by subduction processes, Reid (1995) suggests that the magnitude of U enrichment is too high to be explained in this manner. She suggests that this could arise from metasomatism by alkaline melts and CO2 rich fluids in the wake of the Yellowstone plume. The existence of three fundamental magma types in the JVVF (viz. HAOT, SROT-like and AB) also indicate the contribution of different mantle sources. While these sources are generally consistent with those discussed above, there are important differences. HAOT are chemically fairly uniform, but they show a wide variation in their isotopic characteristics with location. HAOT from the western part of the Oregon Plateau have relatively unradiogenic 87Sr/86Sr and high 143Nd/144Nd, and could represent melts generated from C1-like sources. A chemically depleted source, similar to the depleted asthenospheric mantle, also appears to have

133 been important in generating the HAOT magmas from the JVVF. However, the relatively radiogenic 87Sr/86Sr and 143Nd/144Nd favor the Hart et al. (1997) model which invokes mixing of C1 melts with those derived from mafic material from the lithosphere. Melts derived from such material were presumably not significantly different from the C1 melts in their incompatible element concentrations or were of a limited volume, and hence did not substantially affect the elemental character of the mixed melt. A chemically and isotopically enriched source, perhaps in the SCLM, probably generated the lone SROT-like magma (Vent 4569). Its 87Sr/86Sr and 143Nd/144Nd are very similar to those proposed for the C3 source. However, its Pb isotopic characteristics are significantly different – Vent 4569 shows extremely radiogenic 206Pb/204Pb (almost 20), higher than even the value proposed for the C2 component. It also has the highest 207Pb/204Pb and 208Pb/204Pb values among all vents from the JVVF, and forms a distinct outlier in such plots. If this source component is indeed SCLM, it is unclear why it should show a Pb isotopic signature so different from that ascribed to the C3 component and magmas derived from it (see Carlson and Hart, 1987). Melting of a mantle source contaminated with sediment-derived material could lead to such characteristics – this contamination would have had to be ancient, rather than modern, given the highly radiogenic Pb isotopic composition. This composition, however, is also within the range of some measured Wyoming Craton SCLM compositions (spinel peridotite, pyroxenite and glimmerite xenolith data from Montana; Carlson and Irving, 1994). In principle, this vent could also reflect melting of a pod of SCLM that owes its highly radiogenic Pb isotopic composition to ancient metasomatism by sediment-derived fluids or by fluids related to the Yellowstone plume (as suggested by Reid, 1995). The uniqueness of the Pb isotopic composition of this vent deserves further investigation. A source enriched in LILE, perhaps similar to OIB sources appears to have been involved in the generation of the younger (Group 2) magmas. The Coffeepot Crater, the youngest vent in this group shows the lowest 87Sr/86Sr and highest 143Nd/144Nd of all vents from the JVVF. Isotopically, it is closest in composition to the C1 source. It has been suggested by Hart and Carlson (1987) that the JVVF AB could represent deeper, relatively uncontaminated melts of asthenospheric mantle. Shoemaker (2004), as mentioned before, suggests that these represent melts of fluid metasomatized asthenospheric mantle. What is interesting is that HAOT lavas with low 87Sr/86Sr from the western part of the Oregon Plateau, presumably derived from depleted

134 asthenospheric mantle with minimum interaction with the SCLM, are not enriched in LILE. On the other hand, the Coffeepot Crater lavas with equivalent MgO contents and low 87Sr/86Sr as the western Oregon Plateau HAOT are enriched in LILE, Sr, and Nb (Fig. 7). These characteristics cannot be explained easily by smaller degrees of partial melting, since those would increase the HREE concentrations unless garnet was present in the source. As seen earlier, this does not appear to be the case. Similarly, AFC involving continental crust would lead to increases in 87Sr/86Sr, something that is not observed. It follows therefore, that HAOT magmas uncontaminated by lithospheric melts (as found in the western Oregon Plateau) and the least contaminated AB (such as the Coffeepot Crater samples) from the JVVF must be derived from different sources. The mildly alkaline nature of the Coffeepot Crater lavas and the fact that it is the youngest vent in the JVVF could reflect recent tapping of an OIB-like asthenospheric source. Furthermore, the fact that within-vent elemental and isotopic diversity is documented for younger AB vents such as Coffeepot Crater and Rocky Butte demonstrates that this source itself might be heterogeneous at a small scale. Shoemaker (2004) demonstrated that several Owyhee Plateau basaltic compositions can be explained as mixtures of sources of endmember HAOT and SROT in various proportions. Such mixing was initially proposed by Hart (1985) for basalts from the northwestern United States. Transitional Basalts are especially likely to have been generated by mixing between C1 and C3 components. These source components also appear to be geographically distinct. C3, i.e. SCLM appears to dominate beneath the Snake River Plain, whereas C1 (depleted, oceanic type mantle) dominates west of the Sr706 line (e.g. Leeman et al. 1992). The presence of both source components is however implicated in the Owyhee Plateau region (Shoemaker, 2004; Shoemaker and Hart, 2002). Given the potential for source heterogeneity, it is likely that many of the JVVF transitional basalts were derived by variable amounts of mixing between HAOT and SROT sources or of magmas derived from these sources. Mixing processes should be identifiable by systematic variations and trends on binary elemental and isotopic plots. It was mentioned before that Group 1 samples do indeed show such a relationship on several plots (e.g. Fig. 5, 7). The two end-members on elemental plots appear to be HAOT vents such as the Owyhee Butte, and the SROT-like vent (Vent 4569), which would mix in various proportions to generate TB vents such as the Three Mile Hill and Bernie’s Butte. However plots of 87Sr/86Sr and 143Nd/144Nd vs. several major and trace elements reveal that the Owyhee Butte and similar compositions cannot

135 be the depleted end-member. Instead, that end-member appears to be the Cow Vent Complex, which is isotopically more primitive but chemically slightly more evolved than the Owyhee Butte (Fig. 5, 6, 7). This raises the question as to why the Owyhee Butte-type HAOT magmas were not involved in mixing with the SROT-like end-member. Whether this is fortuitous or has broader implications is something that future work could address. Figure 20 illustrates that the elemental and isotopic characteristics of JVVF TB vents can be reproduced by mixing ~60% of the Cow Vent Complex end-member (HAOT) with ~40% of the vent-4569 end-member (SROT). However, the Pb-isotopic compositions of TB vents are not reproduced by mixing between these two end-members. This is primarily because of the extremely radiogenic Pb isotopic composition of vent-4569. This vent might, however, be an exceptional case of melting of a pod or pocket of SCLM with highly radiogenic Pb isotopic composition. This does not preclude the existence of a mixing end-member at depth with more of a “typical” SROT-like Pb isotopic composition as originally suggested for the greater JVVF and Owyhee Plateau areas by Hart (1985).

Summary The above discussion on petrogenesis of the JVVF eruptive products can be summarized as follows: • Much of the observed compositional diversity in the JVVF appears to result from the contribution of heterogeneous sources in the production of the three main magma types; HAOT, SROT-like and AB. • The availability of these three main magma types beneath the JVVF has resulted in the production of two distinct groups of lavas; the older, tholeiitic Group 1 and the younger mildly alkaline Group 2. • None of the erupted lavas have compositions indicative of their being primary mantle melts. The most primitive lavas can be generated by the removal of olivine and/or low Ca pyroxene at depths of 8-10kb. • Shallow fractional crystallization appears to not have been an important process affecting the prominent between vent JVVF compositional diversity. However, some of the within- vent compositional diversity could be the result of very shallow fractionation of olivine and plagioclase in lava ponds or during temporary storage beneath the eruptive vents.

136 • Certain geochemical characteristics hint at the operation of open system differentiation processes, particularly in the Group 2 AB. Deep fractional crystallization and mixing of crustal melts could explain some, but not all of the observed compositional characteristics. Shallow AFC, however, is inconsistent with the bulk of the physical and chemical considerations. Although it is not possible to entirely discount AFC in light of the limited modeling, it does not appear to have left unambiguous imprints on the erupted compositions.

Implications for the evolution of MVF and small-scale mantle heterogeneity Compositional variations within the products of individual monogenetic volcanoes have are recently been reported by several workers (e.g. Hughes et al. 2002; Nemeth et al. 2003; Strong and Wolff, 2003; Bondre and Hart, 2004). The trace element and isotopic variations observed by Strong and Wolff (2003), for example, are difficult to explain by differentiation processes alone. They tested various scenarios that could potentially generate the observed variations, including fractional crystallization, variable degrees of partial melting, and crustal contamination. None of these were able to explain all the observed geochemical characteristics, although increasing LREE/HREE ratios suggested some role for varying degrees of partial melting. They were therefore compelled to invoke two different sources (OIB-like and N-MORB like) for each of the vents that they studied. In addition to this, the variation in LILE compositions between different vents led them to postulate varying degrees of slab fluid input as a third component. The JVVF data are also consistent with contribution from multiple source reservoirs, some or all of which have received variable input from slab-derived components. Reiners (2002) compiled compositional data for monogenetic or short-lived polygenetic volcanoes from a variety of tectonic settings in order to investigate issues related to source heterogeneity. Based on radiogenic Os isotopic compositions of later erupting lavas, he invoked a scenario involving pyroxenite domains within a peridotitic matrix. Melt generated from pyroxenite (and which erupts later in the eruptive cycle) would have more radiogenic Os isotopic composition as compared with that generated from peridotite. This implies that melts generated from geochemically distinct sources mix subsequently to lead to observed compositional variations. The distinct sources may be in the form of pods of one type of source embedded in the matrix of another at small spatial scales. On a larger scale, this type of process is similar to

137 the one envisioned by Hart et al. (1997) to explain regional isotopic variations within HAOT magmas. Bryce and DePaolo (2004) presented results of Pb, Sr and Nd isotopic and trace-element compositions of phenocrysts, groundmass and whole rock of basaltic lavas from diverse tectonic settings. These included the Inmaha basalts of the Columbia River Basalt Group, a MORB and a lava flow from Mt. Etna. Their observations do not relate directly to MVF (although flood basalt eruptions are essentially monogenetic), but are insightful and relevant to the present discussion. In each of these settings, they found considerable Pb isotopic heterogeneity between plagioclase phenocrysts, magnetite separates, and the groundmass. Their results indicated that magmas of distinct composition must have mixed just before eruption. This implies the existence of heterogeneous source regions and also that the processes of melting and segregation are insufficient to homogenize the differences. The present study is one among a very few studies (e.g. Hughes et al. 2002) that explicitly address the diversity between monogenetic volcanoes within a field rather than within single volcanoes. Its results suggest that such observations also provide important constraints on the issues involved. The presence of vents with distinct chemical and isotopic compositions in the JVVF and, more importantly, their spacing could be used to place constraints on the nature and scale of source heterogeneity. If such diversity is largely the result of melting of heterogeneous sources, which is indicated by the discussion on petrogenesis, then this suggests that various source components could be intimately mixed within the mantle. Their sizes could be in the range of a kilometer to several kilometers, and could occur as pods or veins of one material in another. For example, the fact that the Owyhee Butte and Little Owyhee Butte are around 5 km apart and show diverse compositions that cannot easily be related to each other by differentiation processes could indicate kilometer scale heterogeneity. Very similar observations have been recorded from the southern Owyhee Plateau region by Shoemaker (2004), where HAOT, SROT and TB vents occur in close proximity. Such a configuration is very different in scale than the major, large scale heterogeneity related to sub-cratonic mantle to the east and that associated with accreted terranes to the west. Within-vent chemical and isotopic diversity as observed for the Coffeepot Crater suggests that the sizes of heterogeneities might be even smaller, perhaps on the order of a few hundred meters, although processes other than source heterogeneity cannot yet be completely ruled out. Observations of small-scale heterogeneity are

138 not new (e.g. Zindler and Hart, 1986) – however, they have largely had to rely on observations derived from ophiolites or xenoliths. What is significant about the present study is that it allows for inferences about the scale of heterogeneity to be made in the absence of such data. On the other hand, some studies report evidence that suggests that differentiation processes rather than source heterogeneity are responsible for compositional variations within monogenetic volcanoes. Nemeth et al. (2003) relate the observed compositional variations solely to fractional crystallization. White et al. (2002) model the compositions of the youngest basalts in the Melba area by a combination of upper crustal assimilation and fractional crystallization. Glazner et al. (1991), Glazner and Farmer (1992), and Ramos and Reid (2005) invoke cryptic lower crustal contamination. Such scenarios involve crustal storage of the magma, although there are few constraints on the typical residence times. Some constraints are provided by diffusion modeling of single crystals (e.g. Klugel, 1998), which suggest multiple, short-lived storage zones. Small bodies of magma are prone to cool and crystallize rapidly at such zones; this suggests that prolonged residence is not feasible. Marsh (1996) invoked a scenario involving magmatic mush columns forming the plumbing system of volcanoes; if this scenario operates beneath MVF, then a system of interconnected mush zones / magma chambers could be visualized. Such a model could allow prolonged residence of magma in the lithosphere and thereby facilitate fractional crystallization, magma mixing and AFC. The review of previous studies reveals the existence of two principal hypotheses, viz. (a) the observed compositional diversity in MVF is the result of source heterogeneity, preserved due to limited or no crustal residence and incomplete mixing, and (b) differentiation processes such as fractional crystallization and crustal contamination are important in effecting the observed compositional diversity in MVF. As in many other instances in geological research, there is a natural tendency for “either-or” scenarios to dominate the discussion (e.g. whole mantle vs. layered mantle convection). However, in light of the fact that MVF occur in diverse tectonic settings with different tectonomagmatic histories, it is not surprising to envision both hypotheses being valid in specific contexts. As discussed in the introduction, MVF have diverse characteristics. For example, in MVF such as those on the Snake River Plain, which are dominated by shield volcanoes, extensive crustal magma storage has been inferred (e.g. McQuarrie and Rodgers, 1998). In such cases, even if magmas were derived from heterogeneous sources, this residence might allow greater time for mixing and homogenization. This might also

139 permit AFC processes to operate to varying extents, thus masking the effects of source heterogeneity. On the other hand, in MVF dominated by small-volume cinder cones and lava flows, which are often alkaline and xenolith bearing and interpreted to be sourced from greater depths, crustal residence might be minimal leading to the preservation of source-related compositional diversity. Work by Geist et al. (2002) on a single episode of Snake River Plain magmatism suggests the operation of AFC processes. However, it should be noted that work by Hughes et al. (2002) on monogenetic volcanism in the Eastern Snake River Plain indicated a negligible role, if at all, for crustal contamination. Rather, varying degrees of partial melting of heterogeneous sources were invoked. It is obvious that no single model appears to explain all the observations from various MVF. Future studies involving single crystal isotopic analysis might help in identifying potential contributions of contamination, especially from the lower crust. The two possibilities discussed above form end-members for magmatic plumbing systems beneath MVF, and are conceptualized in Figure 21. What is needed now is an integration of several types of information in order to determine which of these end-members are most likely, whether they depend on the specific tectonic environment, and whether both end- members can co-exist within the same setting. Such information includes geophysical and geochemical data, as well as thermal and numerical modeling of the generation, segregation, interaction and residence of small magma batches. Integrated studies of the kind proposed above can only enhance the understanding of such issues. Numerical modeling by Bons et al. (2004) offers some interesting insights into melt accumulation/segregation, and the production of chemically distinct melt batches. They propose a model for melt extraction from the source based on a system of Self Organized Criticality. This implies that transport and accumulation is not a steady process, but occurs in many small and some occasionally large bursts; i.e. these processes are punctuated. A key to this is the formation and transport of numerous small melt batches, each of which may therefore have a different chemistry. Each melt batch might not equilibrate with its source, which may lead to disequilibrium partitioning of trace elements and even isotopes. At first sight, this model is consistent with observations of compositional variations within basalts from MVF in general and from the JVVF in particular. In fact, the JVVF appears to be a “type-example” of this concept. The between-vent and within-vent compositional diversity in the JVVF suggests that melt generation in this setting might also have been dominated by small melt batches, which coalesced only much later to produce the larger

140 outpourings of basaltic lava. The inherent “randomness” associated with a system of Self Organized Criticality might provide a good framework to understand the lack of apparent pattern to the locations and compositions of the JVVF vents, and those in other MVF. Figure 22 is intended to help conceptualize the types of processes that might be expected in MVF magma generation and evolution – the contribution of each needs to be evaluated with the kind of integrated work alluded to earlier. How might small-scale mantle heterogeneities be generated? Ultimately, this must have to do with mantle convection and the resultant recycling of oceanic crust (including ocean islands) and the processes associated with subduction related and other types of fluid and melt metasomatism. McKenzie et al. (2004) carried out numerical experiments of convection and found that any heterogeneity introduced into the mantle is stretched out into thin sheets within a few overturns, thus indicating that the mantle should be extremely inhomogeneous. Diffusion, in the absence of melting is inefficient at homogenization. Thus, it seems likely that subducted slabs might lose their integrity over geologic time and disaggregate into small pods and sheets. The “texture” of the mantle containing such inhomogeneities is not well known yet; however, the results of the reviewed studies clearly indicate that detailed geochemical studies of basalts from a given region, coupled with physical modeling can shed considerable light on this aspect. Thorkelson and Breitsprecher (2005) discuss an interesting way of creating long-term (and potentially small-scale) mantle heterogeneities during the subduction process. They discuss the generation of adakitic melts during ridge subduction in slab-window environments. During this process, the edge of the slab undergoes thermal erosion and disintegrates, transferring slab restite to the ambient mantle. This restite is dominantly eclogite and individual restite fragments can range from boulder-size blocks to individual mineral grains. These restite components, depending on ambient conditions of temperature, pressure and mantle flow may become part of either the lithospheric or asthenospheric reservoirs. In the latter, convection will cause them to “streak-out” (as in the model of McKenzie et al. 2004) and form pyroxenitic veins or pods. This appears to be one thermally and physically feasible way of reintroducing components from the oceanic crust to the mantle which, in addition to slab melts, can lead to mantle heterogeneity at various spatial scales.

141 Implications for Regional Magmatism The main thrust of this study was to comprehensively characterize the JVVF in terms of its geology, geochemistry and geochronology, and to evaluate its petrologic evolution within the context of significant questions pertaining to MVF. Shoemaker (2004) has comprehensively addressed the significance of basaltic volcanism in the Owyhee Plateau area for understanding the Cenozoic tectonomagmatic evolution of the northwestern United States. The JVVF occurs at the northern fringe of the Owyhee Plateau, but many observations from the present study are broadly consistent with his interpretations. Salient among these include: (a) heterogeneous mantle sources are responsible for the first order compositional diversity, (b) SROT magmas stopped erupting in this region after around 5 Ma, which reflects the exhaustion of fusible components from the lithospheric mantle shelf represented by the Owyhee Plateau. However, the eruption of the vents with TB compositions well after 5 Ma suggests that this end-member source / magma remained available after 5 Ma, but only as a mixing end-member because of the limited volumes available, (c) The bulk of volcanism in the JVVF occurred after 2 Ma as opposed to older volcanism on the Owyhee Plateau, suggesting that major eruptive activity shifted to the fringes of the Plateau. Comparison of the timing and nature of volcanism in the Melba area (White et al. 2002) and that in the JVVF reveals some interesting similarities between the two regions. The Melba area is characterized by an older phase of tholeiitic volcanism (9-4 Ma) and a younger phase of tholeiitic to alkaline volcanism (2-0.4 Ma). This is strikingly similar to the youngest pulses of JVVF volcanism ranging from ~2 Ma to Recent, which also include tholeiitic as well as mildly alkaline lavas. The younger alkaline vents from both areas show characteristics that indicate some role for AFC processes. This temporal relationship is significant in that the JVVF is ~ 100 km west of the Melba area of the western Snake River Plain. What is even more interesting is that a very similar shift from tholeiitic to more alkaline compositions has been documented from the Smith Prairie region north of the Snake River Plain (Vetter and Shervais, 1992). These authors noted that the mildly alkaline lavas from that area were younger than 0.7 Ma. It is unlikely that this trend of tholeiitic to alkaline volcanism from widely separated areas is coincidental. As White et al. (2002) also suggest, this probably indicates a recent regional change in the nature of the mantle source. Perhaps this signifies depletion of the easily fusible

142 components in the lithospheric mantle beneath this region, and greater input of asthenospheric magmas.

Concluding Remarks While convection in the earth’s mantle is capable of stretching and disaggregating larger heterogeneities (e.g. coherent subducted slabs, delaminated lithosphere) into smaller streaks and blobs, diffusion is incapable of further homogenization on time scales pertinent to the age of the earth. Therefore, the existence of small-scale mantle heterogeneity is not surprising. However, the fact that the effects of such heterogeneity are seen in basalts from diverse environments has an important bearing on processes of melt generation and transport through the earth’s mantle and crust. Similarly, knowledge of the exact spatial disposition of such heterogeneities within the peridotitic mantle (i.e. the fabric of the mantle) can provide new insights into the nature and efficacy of mantle convection. In continental settings, MVF appear to be useful laboratories for understanding the length-scales of heterogeneities as well as the dynamics of the melting process. In such settings, the region of partial melting may be spatially restricted and thus such volcanoes are potentially uniquely suited to recording source related compositional diversity. At the same time, melt batches probably escape substantial mixing and homogenization in the absence of large crustal magma chambers, thus allowing this heterogeneity to be faithfully recorded. Further detailed studies of such settings, coupled with advances in modeling the physics of melt generation and transport have a great potential for enhancing our understanding of geodynamics in general. As a step in this direction, the present study focused on the physical, geochemical and geochronologic characterization of the JVVF. Field mapping led to the clarification of stratigraphic relationships, which in conjunction with age data led to the identification of three major pulses of activity. A database of major and trace element compositions as well as isotopic compositions of the constituent vents and lava flows is now available. A significant contribution of this study is the provision of a robust spatial and temporal context to the interpretation of geochemical data. This is very important if this field is to provide insights into the interplay of source and process related effects on compositions of monogenetic volcanoes. Analysis of the geochemical data in its spatio-temporal context, accompanied by limited modeling suggests the following:

143 (a) Similar to other MVF, the JVVF is characterized by between-vent as well as within-vent compositional diversity (b) Much of this diversity is strongly suggestive of contribution from heterogeneous sources and mixing between magmas derived from various sources. Furthermore, such heterogeneity is likely to be on the scale of kilometers. (c) Broadly speaking, three main magma types were erupted in this region, viz. HAOT, SROT- like and AB. Only AB lavas have erupted in the post 0.25 Ma period, which indicates contribution from a previously untapped source/sources. (d) Differentiation processes have undoubtedly acted – among these processes deeper fractional crystallization appears to have been the most important. Some of the characteristics of the AB are consistent with open system processes whereas others are not. Crustal contamination, however, appears to not have significantly affected HAOT magmas.

As discussed earlier, studies from other MVF have urged caution in the recognition of source heterogeneity on the basis of compositional variations in such fields. This is primarily because in those studies, crustal contamination has been demonstrated to have played an important role. The present study has not attempted detailed quantitative modeling and it lacks high-resolution data such as isotopic compositions of groundmass and mineral separates. Similarly, geophysical data for the Oregon Plateau are sparse. These types of data are required to better evaluate the potential role of differentiation processes, particularly open system modification of magmas in generating the observed compositional diversity. It is important to keep these limitations in mind while interpreting evidence of source heterogeneity beneath the JVVF.

144 References

Anders, E. and Grevesse, N. (1989). Abundances of the elements: meteoric and solar. Geochimica et Cosmochimica Acta 53: 197-214. doi: 10.1016/0016-7037(89)90286-X

Best, M. G., and Christiansen, E. H (2001). Igneous Petrology. Blackwell Science, 458 p.

Bondre, N. R., and Hart, W. K (2004). Nature and causes of compositional variations within individual monogenetic volcanoes: insights from the Coffeepot Crater, Jordan Valley Volcanic Field, Oregon. Abstract volume of the 2nd International Maar Conference, Occasional papers of the Geological Institute of Hungary, v. 203, p. 44.

Bons, P. D., Arnold, J., Elburg, M. A., Kalda, J., Soesoo, A., and van Milligen, B. P (2004). Melt extraction and accumulation from partially molten rocks. Lithos, v. 78, pp. 24-42.

Boudreau, A. E (1999). PELE; a version of the MELTS software program for the PC platform. Computers & Geosciences, Vol. 25, pp. 201-203.

Brueseke, M.E., Hart, W. K., and Shoemaker, K. A. (2002). New tectonic implications derived from southern Oregon plateau basaltic volcanism; defining the Owyhee Block. Geol. Soc. Am. Abst. Prog., v. 34, p. 38.

Bryce, J. G., and DePaolo, D. J (2004). Pb isotopic heterogeneity in basaltic phenocrysts. Geochim. Cosmochim. Acta, v. 68, pp. 4453-4468.

Camp, V. E and Ross, M. E (2004). Mantle dynamics and genesis of mafic magmatism in the intermontane Pacific Northwest. J. Geophys. Res, v. 109. B08204 DOI:10.1029/2003JB002838.

145 Carlson, R. W (1984). Isotopic constraints on Columbia River flood basalt genesis and the nature of the subcontinental mantle: Geochim. Cosmochim. Acta, v. 48, pp. 2357- 2372.

Carlson, R. W., and Hart, W. K. (1987). Crustal genesis on the Oregon Plateau. J. Geophys. Res., v. 92, pp. 6191-6206.

Carlson, R.W., and Hart, W. K (1988). Flood basalt volcanism in the northwestern United States. In: Macdougall, J. D (ed.) Continental Flood Basalts: Kluwer Academic Publishers, pp. 35-61.

Carlson, R. W., Lugmair, G. W., and Macdougall, J. D (1981). Columbia River volcanism: the question of mantle heterogeneity or crustal contamination. Geochim. Cosmochim. Acta, v. 45, pp. 2483-2499.

Connor, C. B., Stamatakos, J. A., Ferrill, D. A., Hill, B. E., Ofoegbu. G. I., et al. (2000). Geologic factors controlling patterns of small-volume basaltic volcanism: Application to a volcanic hazards assessment at Yucca Mountain, Nevada. J. Geophys. Res., v. 105, pp. 417-432.

Conway, M. F., Ferrill, D. A., Hall, C. M., Morris, A. P., Stamatakos, J. A., et al. (1997). Timing of basaltic volcanism along the Mesa Butte Fault in the San Fransisco Volcanic Field, Arizona, from 40Ar/39Ar dates: Implications for longevity of cinder cone alignments. J. Geophys. Res, v. 102, pp. 815-824.

Cummings, M. L., Evans, J. G., Ferns, M. L., and Lees, K. R. (2000). Stratigraphic and structural evolution of the middle Miocene synvolcanic Oregon-Idaho graben. Geol. Soc. Am. Bull., v. 112, pp. 668-682.

DePaolo, D. J (1981). Trace element and isotopic effects of combined wallrock assimilation and fractional crystallization. Earth. Planet. Sci. Lett, v. 53, pp. 189-202.

146 Ferns, M. L., Evans. J. G., and Cummings, M. L (1993). Geologic map of the Mahogany Mountain 30x60 minute quadrangle, Malheur County, Oregon and Owyhee County, Idaho. Dept.of Geology and Mineral Industries, Oregon.

Geist, D., and Richards, M. (1993). Origin of the Columbia Plateau and Snake River Plain: Deflection of the Yellowstone Plume. Geology, v. 21, pp. 789-792.

Geist, D. J., Hughes, S, S., and Sims, E. N (2002). Open-system evolution of a single episode of Snake River Plain magmatism. Geol. Soc. Am. Spec. Pap 353, pp. 193-204.

Glazner, A. F., Farmer, G. L., Hughes, W. T., Wooden, J. L., and Pickthorn, W (1991). Contamination of basaltic magma by mafic crust at Amboy and Pisgah Craters, Mojave Desert, California. J. Geophys. Res, v. 96, pp. 13,673-13,691.

Glazner, A. F., and Farmer, G. L (1992). Production of isotopic variability in continental basalts by cryptic crustal contamination. Science, v. 255, pp. 72-75.

Greeley, R (1982). The Snake River Plain, Idaho : representative of a new category of volcanism. J. Geophys. Res, v. 87, pp. 2705-2712.

Hart, W. K (1982). Chemical, geochronologic and isotopic significance of low K, high- alumina olivine tholeiite in the northwestern Great Basin, U.S.A. Ph.D. dissertation, Case Western Reserve University, 410 p.

Hart, W. K (1985). Chemical and isotopic evidence for mixing between depleted and enriched mantle, northwestern U.S.A. Geochim. Cosmochim. Acta, v. 49, pp. 131-144.

Hart, W. K (1992). The development of the Oregon Plateau: Perspectives from Neogene mafic to intermediate volcanism: Geol. Soc. Am. Abstr. Prog, v. 24, p. 32.

147 Hart , W. K. (1997). Plume-asthenosphere-lithosphere interactions through time: A comparision of the Ethiopian and northwestern USA basalts. Geol. Soc. Amer. Abst. Prog., v. 29, p. 298.

Hart, W. K., and Mertzman, S. A (1981). Source mixing; an example from basalts of the northwestern U.S. Geol. Soc. Am. Abstr. Prog, v. 13, p. 168.

Hart. W. K., and Mertzman, S. A (1983). Late Cenozoic volcanic stratigraphy of the Jordan Valley area, southeastern Oregon: Oregon Geol. v. 45, p. 15-19.

Hart, W. K., and Carlson, R. W. (1987). Tectonic controls on magma genesis and evolution in the northwestern United States. J. Volcanol. Geotherm. Res., v. 32, pp. 119- 135.

Hart, W. K., Aronson, J. L., and Mertzman, S. A (1984). Areal distribution and age of low-K, high-alumina olivine tholeiite magmatism in the northwestern Great Basin. Geol. Soc. Am. Bull, v. 95, pp. 186-195.

Hart, W. K., Hackett, W. R., and Mertzman, S. A (1992). Eruptive and petrogenetic history of the Jordan Craters volcanic field, southeastern Oregon. Geol. Soc. Am. Abstr. Prog, v. 24, p. 32.

Hart, W.K., Carlson, R. W., and Shirey, S. B (1997). Radiogenic Os in primitive basalts from the northwestern U.S.A.: Implications for petrogenesis: Earth. Planet. Sci. Lett, v. 150, p. 103-116.

Hegner, E. and Tatsumoto, M (1987). Pb, Sr, and Nd isotopes in basalts and sulfides from the Juan de Fuca Ridge. J. Geophys. Res, v. 92, pp. 11,380-11,386.

Hofmann, A. W (1988). Chemical differentiation of the earth: The relationship between mantle, continental crust, and oceanic crust. Earth. Planet. Sci. Lett, v. 90, pp. 297 - 314.

148

Hooper, P. R. and Conrey, R. M (1989). A model for the tectonic setting of the Columbia River basalt eruptions. In: Reidel, S. P. and Hooper, P. R (eds). Volcanism and Tectonism in the Columbia River Flood-Basalt Province. Geol. Soc. Am. Spec. Paper 239, pp. 293-306.

Hughes, S., Wetmore, P. H., and Casper, J. L (2002). Evolution of Quaternary tholeiitic basalt eruptive centers on the Eastern Snake River Plain, Idaho. In: Bonnichsen, B., White, C. M., McCurry, M. (eds). Tectonic and magmatic evolution of the Snake River Plain Volcanic Province, Idaho Geol. Surv. Bull. 30, pp. 363-386.

Irvine, T. N. and Baragar, W. R. A (1971). A guide to the chemical classification of the common volcanic rocks. Canadian. Jour. Earth. Sci, v. 8, pp. 523-548.

Ito, E., White, W. M., and Gopel, C (1987). The O, Sr, Nd and Pb isotope geochemistry of MORB. Chem. Geol, v. 62, pp. 157-176.

Jaegar, J. C (1968). Cooling and solidification of igneous rocks. In: Hess, H. H., and Poldervaart, A (eds). The Poldervaart treatise on rocks of basaltic composition. New York, Willey Interscience, pp. 503-536.

Klugel, A (1998). Reactions between mantle xenoliths and host magma beneath La Palma (Canary islands): constraints on magma ascent rates and crustal reservoirs. Contrib. Min. Pet, v. 131, pp. 237-257.

Kogiso, T., Hirschmann, M. M., and Reiners, P. W (2004). Length scales of mantle heterogeneities and their relationship to ocean island basalt geochemistry. Geochim. Cosmochim. Acta, v. 68, pp. 345-360.

149 Korme, T., Chorowicz, J., Collet, B., and Bonavia, F. F (1997). Volcanic vents rooted on extension fractures and their geodynamic implications in the Ethiopian Rift: J. Geophys. Res, v. 79, pp. 205-222.

Lambert, R. St. J., Chamberlain, V. E., and Holland, J. G (1995). Ferro-andesites in the Grande Ronde Basalt: Their composition and significance in studies of the origin of the Columbia River Basalt Group. Canadian Jour. Earth. Sci, v. 32, pp. 424-436.

Lawrence, R. D (1976). Strike-slip faulting terminates the Basin and Range province in Oregon: Geol. Soc. Am. Bull, v. 87, pp. 846-850.

Le Bas, M.J., Le Maitre, R. W., Streckeisen, A., and Zanettin, B (1986). A chemical classification of volcanic rocks based on the total alkali-silica diagram: J. Petrol, v. 27, pp. 745-750.

Leeman, W. P (1974). Petrology of basaltic rocks from the Snake River Plain, Idaho: Part 1. Ph.D. dissertation, Univ. of Oregon.

Leeman, W.P (1982a). Development of the Snake River Plain-Yellowstone Plateau province, Idaho and Wyoming: an overview and petrologic model. In: Bonnichsen, B. and Breckenridge, R.M. (eds.). Cenozoic Geology of Idaho: Idaho Bureau of Mines and Geology Bulletin, v. 26, pp. 155-177.

Leeman, W. P (1982b). Olivine tholeiite basalts of the Snake River Plain, Idaho. In: Bonnichsen, B. and Breckenridge, R.M. (eds.). Cenozoic Geology of Idaho: Idaho Bureau of Mines and Geology Bulletin, v. 26, pp. 181-191.

Leeman, W. P., Oldow, J. S., and Hart, W. K. (1992). Lithospheric scale thrusting in the western U. S. Cordillera as constrained by Sr and Nd isotopic transitions in Neogene volcanic rocks. Geology, v. 20, pp. 63-66.

150 MacLeod, N. S., Walker, G. W., and McKee, E. H (1975). Geothermal significance of eastward increase in age of Upper Cenozoic rhyolitic domes in southeastern Oregon: U.S.G.S Open-File Report 75-348, 21 p.

Mark, R. K., Bowman, H. R., Asaro, F., McKee, E. H., and Coats, R. R (1975). A high 87Sr/86Sr mantle source for low alkali tholeiite, northern Great Basin. Geochim. Cosmochim. Acta, v. 39, pp. 1671-1678.

Marsh, B. D (1996). Solidification fronts and magmatic evolution. Mineral. Mag, v. 60, pp. 5-40.

McBirney, A. R., Taylor, H. P., and Armstrong, R. L (1987). Paricutin re-examined: a classic example of crustal assimilation in calc-alkaline magma. Contrib. Mineral. Petrol, v. 95, pp. 4-20.

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. Geol. Soc. Am. Bull, v. 94, pp. 292-304.

McKenzie, D., Stracke, A., Blichert-Toft, J., Albarede, F., Gronvold, K., and O’Nions, R. K (2004). Source enrichment processes responsible for isotopic anomalies in oceanic island basalts. Geochim. Cosmochim. Acta, v. 68, pp. 2699-2724.

McQuarrie, N., and Rodgers, D. W (1998). Subsidence of a volcanic basin by flexure and lower crustal flow: The Eastern Snake River Plain, Idaho. Tectonics, v. 17, pp. 203-220.

Mehringer, P (2004). Radiocarbon dates and microfossil analyses: Upper Cow Lake, Oregon. Progress Report 2 for BLM, Vale District, Oregon. Secrets of the Past Consulting, 10 p.

151 Nemeth. K., White, J. D. L., Reay, A., and Martin, U. (2003). Compositional variation during monogenetic volcano growth and its implications for magma supply to continental volcanic fields. Jour. Geol. Soc. Lond., v. 160, pp. 523-530.

Nielsen, R. L (1992). BIG-D.FOR: A FORTRAN program to calculate trace element partition coefficients for natural mafic and intermediate composition magmas. Computers and Geosciences, v. 18, pp. 773-788.

Noe-Nygaard, A (1968). On extrusion forms in plateau basalts. Shield volcanoes of ‘Scutulum’ type. Science in Iceland Anniversary Volume, 10-13. Societas Scientarium Islandica, Rekjavik.

Ormerod, D. S., Rogers, N. W., and Hawkesworth, C. J (1991). Melting in the lithospheric mantle: Inverse modelling of alkali-olivine basalts from the Big Pine Volcanic Field, California. Contrib. Mineral. Petrol, v. 108, pp. 305-317.

Pierce, K. L., and Morgan, L. A (1990). The track of the Yellowstone hotspot, volcanism, faulting, and uplift. USGS Open File Report 90-0415.

Plank, T. and Langmuir, C. H (1998). The chemical composition of subducting sediment and its consequences for the crust and mantle. Chem. Geol, v. 145, pp. 325-394.

Ragland, P. C (1989). Basic analytical petrology. Oxford Univ. Press, 369 p.

Ramos, F. C., and Reid, M. R (2005). Distinguishing between melting of heterogeneous mantle sources from crustal contamination: insights from Sr isotopes at the phenocryst scale, Pisgah Crater, California. J. Petrol, v. 46, pp. 999-1012.

Reid, M. R (1995). Processes of mantle enrichment and magmatic differentiation in the eastern Snake River Plain: Th isotope evidence: Earth. Planet. Sci. Lett, v. 131, pp. 239- 254.

152 Reiners, P. W (2002). Temporal-compositional trends in intraplate basalt eruptions: Implications for mantle heterogeneity and melting processes. Geochem. Geophy. Geosys, v. 3, DOI: 10.1029/2001GC000250.

Rollinson, H (1993). Using geochemical data: evaluation, presentation, interpretation. Longman Scientific and Technical, Harlow, England, 352 p.

Settle, M. (1979). The structure and emplacement of cinder cone fields. Am. Jour. Sci., v. 279, pp. 1089-1107.

Shoemaker, K. A (2004) The tectonomagmatic evolution of the Late Cenozoic Owyhee Plateau, Northwestern United States. Ph.D. Dissertation, Miami University, 276 p.

Shoemaker, K. A., and Hart, W. K. (2002). Temporal controls on basalt genesis and evolution on the Owyhee Plateau, Idaho and Oregon. In: Bonnichsen, B., White, C. M.,

McCurry, M. (eds). Tectonic and magmatic evolution of the Snake River Plain Volcanic Province, Idaho Geol. Surv. Bull. 30, pp. 313-328.

Sims, K. W .W., and DePaolo, D. J (1997). Inferences about mantle magma sources from incompatible element concentration ratios in oceanic basalts. Geochim. Cosmochim. Acta, v. 61, pp. 765-784.

Sinton, J. M., Ford, L. L., Chappell, B., and McCullough, M. T (2003). Magma genesis and mantle heterogeneity in the Manus Back-Arc Basin, Papua New Guinea. J. Petrol, v. 44, pp. 159-195.

Strong. M., and Wolff, J. (2003). Compositional variations within scoria cones. Geology, v. 31, pp. 143-146.

153 Sun, S. S., and McDonough, W. F (1989). Chemical and isotopic systematics of oceanic basalts: Implications for mantle composition and processes. In: Saunders, A. D., and Norry, M. J (eds). Magmatism in ocean basins. Geol. Soc. Lond. Spec. Pub 42, pp. 313- 345.

Thorkelson, D. J. and Breitsprecher, K (2005). Partial melting of slab window margins: genesis of adakitic and non-adakitic magmas. Lithos, 79:25-41.

Vetter, S. K., and Shervais, J. W (1992). Continental basalts of the Boise River Group near Smith Prairie, Idaho. J. Geophys. Res, v. 97, pp. 9043-9061.

Walker, G. P. L. (1993). Basaltic-volcano systems. In : Pritchard, H. M., Alabaster, T., Harris, N. B. W., Neary, C. R. (eds). Magmatic processes and plate tectonics. Geol. Soc. Spec. Publ. 76, pp. 3-38.

White, C., Hart, W. K., Bonnichsen, B., and Matthews, D (2002). Geochemical and Sr- isotopic variations in Western Snake River Plain basalts, Idaho. In: Bonnichsen, B., White, C. M., McCurry, M. (eds). Tectonic and magmatic evolution of the Snake River Plain Volcanic Province, Idaho Geol. Surv. Bull. 30, pp. 329-342.

Wilbold, M., and Stracke, A (2005). Trace element composition of mantle end-members: implications for recycling of oceanic and upper and lower continental crust. Geochem. Geophys. Geosys, v. 7, doi:10.1029/2005GC001005.

Wilcox, R. E (1954). Petrology of Paricutin volcano, Mexico. USGS Bull 965, pp. 281- 353.

Winchester, J. A and Floyd, P. A (1977). Geochemical classification of different magma series and their differentiation products using immobile elements. Chem. Geol, v. 20, pp. 325-343.

154 Wood, C. A. (1980). Morphometric evolution of cinder cones. J. Volcanol. Geotherm. Res., v. 7, p. 387-413.

Workman, R. K, and Hart, W. R (2005). Major and trace element composition of the depleted MORB mantle (DMM). Earth. Planet. Sci. Lett, v. 231, pp. 53-72.

Zartman, R. E., and Haines, S. M (1988). The plumbotectonic model for Pb isotopic systematics among major terrestrial reservoirs; a case for bi-directional transport. Geochim. Cosmochim. Acta, v. 52, pp. 1327-1339.

Zindler, A., and Hart, S (1986). Chemical geodynamics. Ann. Rev. Earth. Planet. Sci, v. 14, pp. 493-571.

155 Table 1: Geochronologic data for samples from the JVVF

Sample Vent Age ± 2 σ (Ma)* Technique

JC-36A Coffeepot Crater (Phase 1) 0.0013-0.0017 Radiocarbon (on underlying organic material)

JC-30B Coffeepot Crater (Phase 2) 0.0013-0.0017 Radiocarbon (on underlying organic material)

JV99-1 West Crater 0.061 ± 0.023 40Ar-39Ar (this study) N2/03 Vent 4569 5.64 ± 0.43 40Ar-39Ar (this study) N6/04 Bogus Bench Vent 1.92 ± 0.22 40Ar-39Ar (this study) N9/03 Owyhee Butte 1.86 ± 0.23 40Ar-39Ar (this study) N9/04 Little Owyhee Butte 1.87 ± 0.08 40Ar-39Ar (this study) N17/03 West Crater/Rocky Butte? (flow) 0.086 ±0.0 33 40Ar-39Ar (this study) N18/03 Bernie’s Butte 2.78 ± 0.17 40Ar-39Ar (this study) H8-70 Clarks Butte 0.25 ± 0.05 K-Ar (Hart et al. 1984) H9-49 Three Mile Hill 1.86 ± 0.19 K-Ar (Hart et al. 1984) H8-57 Rocky Butte (flow) 0.03 (max) K-Ar (Hart et al. 1984) H9-42 Cow Vent Complex (flow) 0.44 ± 0.16 K-Ar (Hart et al. 1984)

156 Table 2: Geochemical data for representative samples from the JVVF

Sample JC-36A JC-30B N10/03 N11/03 JV99-1 H8-60A H8-70 Vent‡ CC CC RB RB WC SH CB Map ID* 1 5 8 9 21 28 29

SiO2 47.58 48.43 47.53 47.22 48.74 48.02 48.79 TiO2 2.25 1.72 1.93 1.96 1.62 1.91 1.94 Al2O3 15.80 15.97 16.64 15.55 16.63 16.23 16.75 † t Fe2O3 11.69 10.33 11.90 12.17 10.87 11.21 11.53 MnO 0.16 0.15 0.18 0.17 0.16 0.17 0.17 MgO 8.76 9.13 7.34 7.81 8.31 8.22 6.69 CaO 9.90 9.85 9.27 8.98 9.99 9.65 8.82

Na2O 2.94 3.07 3.17 3.12 2.86 3.12 3.08 K2O 0.72 0.96 1.27 0.96 1.00 1.12 1.79 P2O5 0.35 0.43 0.45 0.39 0.45 0.40 0.54 LOI 0.62 0.31 -0.35 0.52 -0.54 0.63 0.92 Total 100.77 100.35 98.41 97.74 100.09 100.68 101.02 Rb 13.7 19.2 30.5 21.5 21.9 21.6 43.6 Sr 662 528 519 484 446 542 489 Y 22 21 25 25 21 21 24 Zr 120 135 153 143 115 146 199 V 217 215 207 217 230 205 182 Ni 148 149 97 124 130 139 83 Cr 184 264 102 151 270 202 127 Nb 19.1 31.8 30.8 29.0 25.3 36.9 43.6 Ga 17.5 16.9 18.5 18.0 16.7 19.4 17.9 Cu 61 43 55 43 47 57 51 Zn 90 79 86 92 87 95 83 Co 50 46 47 46 48 45 43 Ba 279 234 423 231 320 322 594 U 0.60 <0.5 <0.5 <0.5 <0.5 2.10 2.15 Th 0.8 2.7 2.3 2.0 0.6 4.8 3.3 Sc 26.0 28.0 20.0 25.0 31.0 26.7 23.0 Pb 4 4 4 4 1 3 --- La 12.0 17.3 17.2 18.1 13.6 18.0 26.2 Ce 28.1 34.4 38.3 45.1 29.8 36.3 51.3 Nd 18.1 19.2 21.2 19.0 16.8 18.8 23.7 Sm 4.52 4.18 4.77 4.49 3.89 4.26 4.96 Eu 1.64 1.49 1.74 1.57 1.44 1.50 1.63 Gd 4.48 4.42 4.91 4.58 4.10 4.45 4.98 Dy 3.89 3.85 4.30 4.11 3.68 3.89 4.37 Er 2.07 2.00 2.46 2.11 2.11 1.96 2.21 Yb 1.81 1.90 2.20 2.21 1.89 2.14 2.24 Lu 0.28 0.31 0.33 0.29 0.27 0.29 0.34 87Sr/86Sr 0.70383 0.70408 0.70498 0.70450 0.70470 0.70443 0.705192 143Nd/144Nd 0.512853 0.512787 0.512691 0.51274 0.512699 0.51271 0.512644 206Pb/204Pb 18.804 18.830 18.668 18.862 18.601 18.833 18.672 207Pb/204Pb 15.563 15.575 15.591 15.586 15.585 15.606 15.604 208Pb/204Pb 38.435 38.547 38.643 38.661 38.486 38.687 38.723 Age (Ma) 0.002 0.002 0.03? 0.03? 0.06 0.06? 0.25

157

Sample N23/03 H9-49 N8/03 N9/04 N6/04 N18/03 N2/04 Vent‡ CVC TMH OB LOB BBV BB WCS Map ID* 34 36 37 42 44 49 53

SiO2 47.99 46.97 47.12 47.18 47.80 47.55 46.72 TiO2 1.31 1.92 1.02 1.65 0.96 1.74 1.45 Al2O3 16.56 15.89 16.38 16.57 17.06 15.61 17.01 † t Fe2O3 11.14 12.90 10.99 12.45 10.93 12.37 11.94 MnO 0.18 0.17 0.18 0.19 0.19 0.19 0.18 MgO 8.64 7.88 9.20 8.76 8.89 7.95 8.88 CaO 11.60 10.51 11.45 9.98 11.91 10.67 10.78

Na2O 2.55 2.63 2.44 2.70 2.46 2.54 2.53 K2O 0.22 0.72 0.20 0.45 0.12 0.59 0.46 P2O5 0.15 0.34 0.15 0.24 0.08 0.34 0.26 LOI -0.40 0.85 0.01 -0.33 -0.31 -0.32 -0.39 Total 99.45 100.78 98.74 99.83 100.08 99.25 99.83 Rb 4.9 15.0 3.6 12.8 3.9 13.8 8.2 Sr 247 305 213 299 178 262 264 Y 23 27 18 27 20 25 22 Zr 86 131 59 110 44 113 84 V 269 268 248 225 257 268 218 Ni 130 124 169 153 124 113 128 Cr 252 253 316 247 259 252 229 Nb 7.8 16.0 6.0 12.6 4.4 14.4 12.2 Ga 17.9 18.3 16.5 20.8 17.4 18.0 17.8 Cu 77 60 83 86 104 74 55 Zn 84 124 69 90 73 91 81 Co 51 60 53 51 52 51 52 Ba 204 404 107 286 132 218 227 U <0.5 --- <0.5 1.50 0.90 <0.5 0.70 Th <0.5 1.5 <0.5 2.0 0.7 <0.5 <0.5 Sc 32.0 31.5 31.0 28.0 34.0 27.0 27.0 Pb 3 --- 3 4 1 5 3 La 8.3 13.5 3.9 8.9 3.4 10.6 10.0 Ce 13.5 28.4 9.5 21.0 7.8 24.4 21.6 Nd 10.6 16.2 6.8 13.4 6.2 15.1 12.5 Sm 2.93 4.02 2.03 3.55 1.95 3.85 3.26 Eu 1.07 1.43 0.87 1.70 0.85 1.46 1.20 Gd 3.49 4.50 2.54 4.00 2.57 4.39 3.57 Dy 3.60 4.49 2.96 4.04 3.07 4.22 3.56 Er 1.93 2.30 1.86 2.40 1.99 2.54 1.92 Yb 2.05 2.38 1.79 2.23 1.88 2.31 2.05 Lu 0.29 0.34 0.26 0.34 0.28 0.34 0.29 87Sr/86Sr 0.70446 0.70535 0.70527 0.70528 0.70534 0.70542 --- 143Nd/144Nd 0.512794 0.512617 0.512740 0.512620 0.512762 0.512597 --- 206Pb/204Pb 18.854 18.757 18.842 18.800 18.887 18.811 --- 207Pb/204Pb 15.596 15.630 15.621 15.646 15.617 15.654 --- 208Pb/204Pb 38.809 38.887 38.910 39.051 38.909 39.100 22 Age (Ma) 0.44 1.86 1.86 1.87 1.92 2.78 ?

158 Sample N2/03 Vent‡ V 4569 Map ID* 56

SiO2 46.61 TiO2 2.04 Al2O3 15.25 † t Fe2O3 13.70 MnO 0.19 MgO 8.13 CaO 10.01

Na2O 2.56 K2O 0.43 P2O5 0.47 LOI -0.38 Total 99.01 Rb 9.2 Sr 293 Y 32 Zr 155 V 256 Ni 131 Cr 204 Nb 17.2 Ga 19.9 Cu 51 Zn 112 Co 57 Ba 228 U <0.5 Th 1.5 Sc 29.0 Pb 7 La 17.5 Ce 38.4 Nd 22.9 Sm 5.48 Eu 1.93 Gd 6.01 Dy 5.49 Er 3.11 Yb 2.76 Lu 0.42 87Sr/86Sr 0.70626 143Nd/144Nd 0.512483 206Pb/204Pb 19.745 207Pb/204Pb 15.736 208Pb/204Pb 39.964 Age (Ma) 5.64

† Total iron as Fe2O3; * Map ID refers to sample locations (see Figure 1; Appendix 4); ‡ For Vent abbreviations see Figure 3 159 Table 3: Parameters used in modeling of fractional crystallization using PELE

Model Pressure % of solids Phases† Temperature Temperature Number Sources for D- removed interval increments of steps values A 8 kb 15 P, LCAP, 1291º C - -5º C 13 Rollinson (1993) CPX, SP 1231º C and references therein; Ragland (1989) B 8 kb 33 LCAP, P, 1291º C - -5º C 16 Same as above CPX, SP 1216º C

† P – plagioclase, LCAP – low Ca pyroxene, CPX – clinopyroxene, SP – spinel. Phases are listed according to their abundance

160 Figure 1: (A) Map of the northwestern United States depicting the location of the Jordan Valley Volcanic Field (JVVF) with respect to principal tectonomagmatic provinces (B) Digital elevation model of the Oregon-Idaho-Nevada tri-state region. Location of the JVVF in the context of important regional tectonic features is shown. OIG – Oregon-Idaho Graben; SM – Steens Mountain; NNR – northern Nevada Rift; MT – Midas Trough; OWP – Owyhee Plateau; OM – Owyhee Mountains; WSRP – Western Snake River Plain.

161 Figure 1

A Cascades Columbia Plateau Oregon Plateau Snake River Plain High Lava Plains Owyhee Plateau Jordan Valley Volcanic Field

122 120 118 116 114 Degrees W. Longitude

B Jordan Valley Volcanic Field OIG WSRP

SM OM

OWP

OR ID NV NV

NNR 50 km MT

162 Figure 2: Satellite image of the JVVF showing the principal vents and vent alignments (A, B and C) discussed in the text.

163 Figure 2

N < 0.25 Ma 5 km ~ 1.9 Ma 0.5-5 Ma A

? 164

Owyhee River

Antelope Valley Fault system Figure 3: Simplified geologic map of the JVVF depicting the principal vents, associated lava flows/flow fields and important structural features. Note the complex inter-fingering of the flows, typical of coalescing shield volcanoes. WCS – West Crater Shield; BB – Bernie’s Butte; BBV – Bogus Bench Vent; OB – Owyhee Butte; LOB – Little Owyhee Butte; TMH – Three Mile Hill; CVC – Cow Vent Complex; CB – Clarks Butte; SH – Skinner Hill; WC – West Crater; RB – Rocky Butte; CC – Coffeepot Crater. Deer Butte and Table Mountain are postulated to be vents but have not been sampled.

165 Figure 3

5 km Owyhee River Rivers/creeks Towns/roads

Faults/fractures Vents ?

Deer Butte CC ? Table Mountain

Vent 4569 Cow Lakes

BBV 4770000N Cow Creek CVC

WC 166 RB WCS SH

BB Jordan Valley OB 5 TMH . 9 . S Danner U ? LOB Arock Jordan Creek ?

4750000N ek re Antelope Valley Antelope Reservoir C an Fault system rd Jo

440000E Rome 460000E 480000E

Vent 4569 Bogus Bench Vent Cow Vent Complex RB/CB/SH ? Undifferentiated

WC Shield (?) Owyhee Butte Clark’s Butte Rocky Butte Lakes/reservoirs

Bernie’s Butte Little Owyhee Butte Skinner Hill Coffeepot Crater

Table Mountain Three Mile Hill West Crater QuaternarySediments Figure 4: (A) JVVF samples plotted on a Total Alkalis vs. Silica diagram of LeBas et al. (1986). Line dividing alkaline and subalkaline rocks is from Irvine and Baragar (1971). (B) JVVF samples plotted on the discriminant diagram of Winchester and Floyd (1977). Note the clear distinction between subalkaline and alkaline basalts in the latter.

167 Figure 4

8 Trachyandesite 7 Basaltic Trachy- Trachy-andesite 6 basalt O 2

K 5 + O 2 a 4 Andesite N

3 Basaltic Andesite 2 Basalt 44 46 48 50 52 54 56 58 60

SiO2

Rhyolite 2 O i T / r .1 Rhyodacite/Dacite Z

TrachyAndesite Andesite

.01 Andesite/Basalt

Alk-Basalt .001 SubAlkaline Basalt

.01 .1 1 10 Nb/Y

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

168 Figure 5: Variation of (A) SiO2, (B) TiO2, (C) Al2O3 and (D) Na2O+K2O with Mg number for basalts from the JVVF. Note the different TiO2 contents for samples with the same Mg number in ‘B’ and two sub-parallel trends in ‘D’.

169 Figure 5 51 A B 50 2

49 2 2 O i O i S 48 T

1 47

Mg # Mg # 6 170 C D 5 17 O 2 3 K

O + 2

l 4 O A 2 a

16 N 3

15 2 50 55 60 65 70 50 55 60 65 70

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2) Figure 6: MgO-K2O-TiO2 ternary diagram with fields for principal magma types from the JVVF.

171 Figure 6

MgO*0.5

Most HAOT

Some HAOT, TB, and SROT-like

K2O*5 TiO2*1.5

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

172 Figure 7: Plots of (A) Sr, (B) Nb, (C) Ba, and (D) Zr versus TiO2 for the JVVF basalts. The presence of two groups (1 and 2) is apparent in ‘A’ and ‘B’, which are highlighted by fields for clarity. Arrows in ‘C’ and ‘D’ show the direction of systematic variation of Ba and Zr respectively with TiO2 for Group 1 samples. No such systematic variation is evident in case of Group 2 samples.

173 Figure 7

700 60 A B 600 Group 2 Group 2 50

500 40

Sr 400 30 Nb

300 20 Group 1 200 Group 1 10

TiO2 TiO2 174

40 C D 200

30 1 Rb up Zr ro G 20 1 100 up Gro

0 0 0.5 1.5 2.5 1.5 2.5

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2) Figure 8: Chondrite-normalized Rare Earth Element (REE) plot for selected JVVF samples. Note the widely varying values for LREE but relatively constant values for HREE, and the positive Europium anomaly for two of the HAOT samples.

175 Figure 8

REE - Nakamura, 1974 Rock/Chondrites 100

SROT-like AB and TB

OB and BBV (HAOT)

La Ce Nd Sm Eu Gd Dy Er Yb Lu

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

176 Figure 9: (A) Plot of 143Nd/144Nd vs. 87Sr/86Sr for the JVVF basalts (modified after Shoemaker, 2004) in the context of fields for data from other regions and mantle components and reservoirs. NMORB - Juan De Fuca Ridge, East Pacific Rise, and Galapagos Ridge basalts (Hegner and Tatsumoto, 1987; Ito et al, 1987); SED - Pacific Ocean trench sediments (Plank and Langmuir, 1998); CRB - Columbia River Basalt (Carlson et al, 1981; Carlson, 1984; Hooper and Hawkesworth, 1993; Lambert et al, 1995); SRP - Snake River Plain basalts (Leeman and Manton, 1971; Noble et al, 1973; Mark et al, 1975; Leeman, 1982; Leeman et al, 1992; and Leeman, unpublished data); HAOT - western Oregon Plateau high alumina olivine tholeiites (Hart, 1985). HIMU, EM1, and EM2 - ocean island basalt mantle components from Zindler and Hart (1986). C1, C2, and C3 - northwestern United States mantle components from Carlson (1984) (B) The same plot as in ‘A’ at a different scale. Note the two distinct Coffeepot Crater as well as Rocky Butte phases. The Owyhee Butte and Bogus Bench Vent show near identical composition as do the TB vents.

177 Figure 9 0.5135 A NMORB

C1 CRB HIMU

d 0.5130 N 4 4 1 / d HAOT C2 N 3 4 1 EM2 0.5125 SED

SRP EM1 C3

0.5120 0.700 0.705 0.710 0.715 87Sr/ 86Sr

0.5129

B Coffeepot (Phase 1)

0.5128

Owyhee Butte, Bogus Bench Vent

d Coffeepot (Phase 2) N 4 0.5127

4 Rocky Butte 2 1 / d

N Rocky Butte 1

3 TB vents 4

1 0.5126

0.5125

0.5124 0.703 0.704 0.705 0.706 0.707 87Sr/ 86Sr 178 Figure 10: Plots of (A) 207Pb/204Pb and (B) 208Pb/204Pb vs. 206Pb/204Pb for the JVVF basalts. Fields and mantle components as in Figure 9a.

179 Figure 10 15.8 EM2 CRB A U IM 15.7 SED H C3 b P 4

0 SRP 2 C2 /

b 15.6 P 7 0 2 C1

15.5 HAOT EM1 NMORB

B C3 EM2 U CRB IM H

b 39 P

4 SRP C2 0 2

b P 8 0 2 EM1 C1 38 SED HAOT NMORB

37 17 18 19 20 206Pb / 204Pb

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

180 Figure 11: Plots of (A) 87Sr/86Sr and (B) 143Nd/144Nd vs.206Pb/204Pb for the JVVF basalts. Fields and mantle components as in Figure 9a.

181 Figure 11 0.715 A SED

0.710

r SRP S 6 8

/ EM2

r C3 S 7 8 EM1 0.705 C2 CRB HIMU C1 NMORB HAOT

B NMORB HAOT

0.5130 CRB C1 HIMU d N 4

4 SED 1 / d C2 N 3 SRP 4 1 0.5125 EM2

EM1 C3

0.5120 18 19 20 206Pb/204Pb

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

182 Figure 12: Satellite image of the JVVF depicting the location of the vents, their ages and salient compositional characteristics. Note the near-identical age and compositions of OB and BBV, and the significantly different composition of BBV and V 4569. See Figure 3 caption for vent abbreviations.

183 Figure 12

N CC < 0.25 Ma ~ 1.5 ka ~ 1.9 Ma 5 km TiO2=1.72-2.28 Nb=18-34 0.5-5 Ma 87Sr/86Sr=0.7038- 0.7041

V 4569 5.64 Ma

TiO2=2.06 Nb=20 ppm BBV 87Sr/86Sr=0.7063 1.92 Ma

TiO2=0.96 Nb=4 CVC 87 86

184 Sr/ Sr=0.7053 0.44 Ma

TiO2=1.32 CB Nb=8 87 86 0.25 Ma Sr/ Sr=0.7045 TiO =1.94-2.15 Owyhee River 2 Nb=44-51 WC 87Sr/86Sr=0.7052 0.061 Ma RB 0.03 Ma TiO2=1.59-1.70 Nb=25 TiO2=1.95-2.19 SH 87Sr/86Sr=0.7047 Nb=25-32 0.06Ma? 87 86 Sr/ Sr=0.7045- TiO2=1.91 BB 0.7050 Nb=31-37 OB 2.78 Ma 87Sr/86Sr=0.7044 1.86 Ma TiO =1.79 TMH 2 1.86 Ma TiO2=1.02 Nb=14 87 86 TiO =1.92 Nb=6 Sr/ Sr=0.7054 2 87Sr/86Sr=0.7053 Nb=16 87Sr/86Sr=0.7054

LOB 1.87 Ma

TiO2=1.64 Nb=13 87Sr/86Sr=0.7053 Figure 13: Binary plots involving various trace elements, trace element ratios and 87Sr/86Sr for the JVVF basalts. (A) La/Nb vs. Ba, (B) La/Nb vs. Sr/Y, (C) Zr vs. Nb, (D) Zr vs. Y, (E) 87Sr/86Sr vs. Rb and (F) 87Sr/86Sr vs. Ni. The two JVVF chemostratigraphic groups are easily recognizable in these plots. Note the constant La/Yb of Group 2 in ‘A’ and constant Sr/Y of Group 1 in ‘B’. Trends of Group 2 samples in ‘E’ and ‘F’ (dashed lines with arrows) are indicative of a role of open system processes in the evolution of these basalts.

185 Figure 13

2.0 2.0

A Group 1 B 1.5 Group 1 1.5 b b

N Group 2 N / 1.0 1.0 / a a L L

0.5 0.5 Group 2 0.0 0.0 0 100 200 300 400 500 600 10 20 30 Ba Sr/Y 300 300 C D Group 2 p 2 ou Gr 200 1 200 p u Zr ro Zr G 100 Group 1 100

0 0 0 10 20 30 40 50 20 25 30 35 40 Nb Y 0.707 0.707

E Group 1 F Group 1 0.706 0.706 r r S S 6 6 8 8 / / r r

0.705 0.705 S S 7 7 8 8 Gro p 2 up 0.704 ou 2 0.704 Gr

0.703 0.703 0 10 20 30 40 50 100 150 200 250 Rb Ni

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

186 Figure 14: Variation of (A) 143Nd/144Nd and (B) Zr versus age for the JVVF basalts. Decreasing 143Nd/144Nd with age is apparent, but Zr concentrations show no systematic trend with age.

187 Figure 14 0.5129 A 0.5128

d 0.5127 N 4 4 1 / d N 3 4 1 0.5126 Group 2

0.5125

B 250

200

Group 1 Zr 150

100

0 .001 .01 .1 1 10 Age (Ma)

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

188 Figure 15: Results of modeling of batch melting of a depleted MORB mantle (DMM) for REE. 7.5% melting of DMM (with La and Ce values similar to enriched DMM) produces REE concentrations similar to the most primitive JVVF HAOT. The positive Europium anomaly is, however, not reproduced indicating either a more plagioclase rich source or shallower final equilibration of the magmas. Compositions of DMM and enriched DMM and distribution coefficients used in the modeling are from Workman and Hart (2005).

189 Figure 15

Rock/ Chondrites REE-Nakamura, 1974

100

N8/03 (Owyhee Butte)

DMM (with La and Ce of ‘E’ DMM) Modeled

La Ce Nd Sm Eu Gd Dy Er Yb Lu

190 Figure 16: Plots of (A) La/Sm(N) and (B) Sm/Yb(N) versus TiO2 for selected JVVF samples. Average values for N-MORB (from Hofmann, 1988) and several OIB (from Wilbold and Stracke, 2005 and references therein) are shown for reference. Normalizing values are from

Anders and Grevesse (1989). None of the JVVF samples have high Sm/Yb(N) values similar to most OIB; thus the presence garnet in the JVVF mantle source(s) is not indicated.

191 Figure 16 4 A

I O ) N (

m 2 S / a L

N-MORB average

OIB (> 4) ) N ( b 2 Y / m S

N-MORB average

1 1 2 TiO2

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

192 Figure 17: Plots of (A) Al-factor and (B) CaO/ Al2O3 versus Fe-factor for the JVVF basalts. Most samples plot between the 1 atm and 10 kb cotectics, but closer to the latter in ‘A’.

Decreasing CaO/ Al2O3 with increasing Fe-factor for the younger Group 2 basalts in ‘B’ is suggestive of clinopyroxene fractionation. See text for details.

193 Figure 17 -0.12 A -0.16 approx. 10 kb -0.20 crystallization trend

r -0.24 o t c a f

- -0.28 l Cpx extraction A -0.32 n regio Hypothetical ctic cote -0.36 Parent cpx plg+ Olv olv+ extraction 1 atm -0.40 0.8 B

0.7

3 Plag O 2 extraction l A / 0.6 O a C Olv extraction 0.5

Cpx extraction

0.4 0.22 0.26 0.30 0.34 0.38 0.42 0.46 0.50 0.54 Fe-factor

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

194 Figure 18: Results of modeling of step-wise fractional crystallization of JC-30B (composition

‘A’ in Table 3) at 8 kb pressure on plots of (A) Al-factor vs. Fe-factor, (B) MgO vs. TiO2 (C)

CaO/ Al2O3 vs. TiO2, (D) K2O vs. TiO2 and (E) SiO2 vs. TiO2. Each increment in the model corresponds to a temperature decrease of 5ºC. For details see text and Table 3. Also shown are the Group 2 samples.

195 Figure 18

-0.10 A

-0.15 r

o JC-30B t c

a -0.20 f - l A

-0.25 Final step

-0.30 0.30 0.35 0.40 0.45 0.50 Fe-factor

10 0.7 B C 9 JC-30B

0.6 3 O

8 2 l O A g / M O

7 a

0.5 C 6

D E 2 49 2 O O 2 i K 1 S 48

0 47 1.5 2.0 2.5 1.5 2.0 2.5

TiO2 TiO2

V. 4569 WCS BB OB BBV LOB TMH

CVC C B SH WC RB CC (1) CC (2)

Steps in 8 kb fractional crystallization of JC-30B

196 Figure 19: Results of calculations of simple binary mixing between a composition derived from deep fractionation of JC-30B (composition ‘A’ in Table 3) and a rhyolitic melt (composition of a rhyolite from the JVVF) for selected elements and oxides. (A) Zr vs. Sr, (B) Rb vs. Sr, (C) Ba vs. Zr, (D) MgO vs. CaO, (E) K2O vs. SiO2, and (E) Spidergram of Primitive Mantle-normalized concentrations of selected trace elements (Normalizing values from Sun and McDonough, 1989).

197 Figure 19

500 200 A B 400 150 300 Zr 100 Rb 200 50 100

0 0 0 100 200 300 400 500 600 0 100 200 300 400 500 600 700 Sr Sr 2000 10 C D 8 1500 6 Ba 1000 MgO 4 500 2

0 0 0 100 200 300 400 500 600 700 0 5 10 15 20 Zr CaO 6 1000

5 E F

4 100

K2O 3 2 10

0 1 50 60 70 80 RbBa K Nb La Ce Sr P Nd ZrSm Ti Y

SiO2

Rhyolitic melt Clarks Butte Group 2 (except Clarks Butte)

Residual liquid A Coffeepot Crater, phase 2 Group 1

Residual liquid B Modeled mixture Mixing curve (20% increments)

198 Figure 20: Results of calculations of simple binary mixing between a sample from the Cow Vent Complex (HAOT) and Vent 4569 (SROT-like). The results indicate that mixing approximately 60% of the former with approximately 40% of the latter reproduces many of the elemental and 87 86 87 86 isotopic characteristics of the TB suite. (A) Sr/ Sr vs. Zr, (B) Sr/ Sr vs. TiO2 and (C) Spidergram of Primitive Mantle-normalized concentrations of selected trace elements (normalizing values from Sun and McDonough, 1989).

199 Figure 20 3 200 A B SROT-like Group 2 (V 4569) Group 2 150 2 2 O

Zr TB i 100 (TMH, LOB, BB) T

HAOT (CVC) 50 HAOT (OB&BBV) 1

200 0.703 0.704 0.705 0.706 0.707 0.703 0.704 0.705 0.706 0.707 87Sr/ 86Sr 87Sr/ 86Sr 100 C Modeled mixture SROT-like (V 4569)

HAOT (CVC) TB (BB)

Rock/Primitive Mantle Sun/McDon. 1989-PM 1 Rb Ba K Nb La Ce Sr P Nd Zr Sm Ti Y Figure 21: Two end-member scenarios for the plumbing systems of Monogenetic Volcano- Fields (MVF). The first scenario involves generation of magmas in a heterogeneous mantle and transport to the surface (rapid?) without significant residence at lithospheric / sublithospheric levels. Evidence for this scenario comes from some MVF with abundant mantle xenoliths in their constituent basalts. Chemical and isotopic diversity within and between individual monogenetic volcanoes is likely to be preserved due to the rapid transport of magma. The second scenario involves magma generation in a heterogeneous mantle, but also significant residence of magmas at one or multiple levels during transport to the surface. This scenario could perhaps be operational in those MVF whose constituent lavas lack mantle xenoliths and which show considerable evidence for open system processes. There is a greater probability for chemically and isotopically heterogeneous magma batches to mix and homogenize, as well as for digestion of crustal xenoliths. The JVVF plumbing system probably lies somewhere in between these two end-members.

201 Figure 21

Scenario 1

Scenario 2 202 HeterogeneousHeterogeneous mantlesource source mantle

Level of neutral buoyancy (MOHO/ Brittle-Ductile transition?)

Heterogeneous mantle source Figure 22: Conceptual diagram showing various processes that are likely to be involved in MVF magma generation and evolution. Heterogeneities (pyroxenite / eclogite veins or pods?) in the mantle source region begin melting earlier than the ambient dominant constituent (peridotite?). Whether the distinct isotopic and elemental character of such melts is retained depends on how fast melt segregates (thus escaping diffusive equilibration; Kogiso et al., 2004). Accumulation and transport of melt might occur in the form of separate, perhaps compositionally distinct melt batches (Bons et al. 2004). Mixing can occur in the transport system (e.g. dikes) or in a magma chamber, but as long as this time scale is rapid, heterogeneity can be preserved in erupted compositions (Bryce and DePaolo, 2004).

203 Figure 22

Eruption

Temporary Mixing of individual sub-crustal melt batches and crystal /crustal storage growth; assimilation?

Transport

Segregation

Diffusion dominant component Heterogeneities (pods or veins?)

meters to kilometers

204 CHAPTER 5

CONCLUDING REMARKS AND SUGGESTIONS FOR FUTURE WORK

CONCLUDING REMARKS The general approach of this dissertation was profoundly inspired by the concept of a “volcanic system” introduced by the late Prof. G. P. L Walker. Although each sub-project focused on answering specific questions and tackling specific aspects, what united them was the broader attempt to apply such a “systems approach”. At every stage of this work, an attempt was made to not lose track of the fact that basaltic volcanism does not occur in isolation from other geologic processes – it is significantly influenced by concurrent processes such as fluvial incision and deposition, and faulting. In turn, these processes are modified to various extents by basaltic volcanism. Although the focus of this work was volcanological, the author sought to bear in mind the fundamental unity and connectedness of geologic processes. An important objective in undertaking this research was also to demonstrate the rich potential of studies that integrate field-based and geochemical information from basaltic provinces. It is a strong conviction of this author that such integration, mindful of the multitudinous scales and interrelationships of geologic processes, can provide fresh perspectives on many outstanding issues. The first sub-project of this dissertation involved dikes from the Sangamner region of the DVP. This focused investigation was the first of its kind in the DVP and is significant because this region is postulated to be a major eruptive area. It revealed that most mafic dikes around Sangamner (western Deccan Volcanic Province) are compositionally similar to the Poladpur / Khandala Formations of the younger Wai Subgroup. These dikes clearly define a swarm, thus necessitating a revaluation of previous work that attributed a random orientation to dikes from this part of the DVP. This swarm could have fed flows belonging to these formations (one dike shows field evidence), although the possibility that they represent late stage intrusions cannot be discounted. Some dikes show an affinity to other formations such as the Ambenali and Bushe suggesting that small volumes of other magma-types were also supplied to this area. Results of the second sub-project highlight the morphological and textural diversity of the Steens Basalt lava flows. This is the first work to focus on morphological aspects of the Steens

205 flows, rather than their geochemistry, age or paleomagnetism. The presence of packages of near- vent material and strongly compound pahoehoe flows suggests that the Steens eruptions were sourced from small to moderate-sized shield volcanoes, similar to those occurring in the Eastern Snake River Plain. These flows were likely emplaced at low effusion rates during sustained eruptive episodes, and grew as a network of overlapping, inflated lobes. The presence of a’a and transitional varieties (such as rubbly pahoehoe) interbedded with the pahoehoe flows indicates that physical parameters changed frequently through the eruptive episodes. These types of flows probably resulted from higher effusion rates and / or higher viscosities. The limited geochemical data obtained during this study, when integrated with the morphological data, clearly reveals the potential of such integration for understanding eruptive history. The Jordan Valley Volcanic Field (JVVF) was recognized as a distinct entity in southeastern Oregon based on work done as part of the third sub-project. Fieldwork that built on the foundations of previous mapping led to the clarification of stratigraphic relationships in this monogenetic-volcano field (MVF) and provided a clear context in which to understand the compositional and age data. The newly obtained 40Ar-39Ar ages, previously available K-Ar ages, and a recently reported radiocarbon age reveal that recognized vents in the JVVF span a range of around 6 Ma to 1.5 ka. Major activity occurred around 2 Ma and in the 0.25 Ma-1.5 ka interval. Chemical and isotopic diversity between and within vents from the JVVF appears dominantly to be the result of small-scale mantle heterogeneity and mixing of melts derived from various sources. Qualitative evaluation of geochemical data and limited modeling suggests that JVVF magmas underwent fractional crystallization at depths corresponding to ~8 kb pressure. Some of the chemical and isotopic data hint at the involvement of crustal contamination in the evolution of the suite of younger, mildly alkaline basalts. Modeling suggests that although concentrations of some elements are reproduced well by a combination of deep fractionation and mixing with evolved silicic melts, the Sr isotopic composition is poorly reproduced. The models are not consistent with the operation of shallow AFC processes. Observations such as the near simultaneous eruption of compositionally diverse lavas from closely spaced vents clearly highlight the complexity of plumbing systems in monogenetic-volcano fields. If future high resolution studies (e.g. mineral-scale isotopic compositions) confirm that open system processes such as crustal contamination were not a significant process in evolution of the JVVF basalts, then the composition of vents and their disposition could provide an estimate of the scale of

206 source-heterogeneity – this is particularly significant in the absence of mantle xenoliths and high-resolution geophysical data for this part of the northwestern United States.

SUGGESTIONS FOR FUTURE WORK

Sub-project 1 Important insights into the relationship of dikes from the DVP to established chemostratigraphic formations are provided by results of this sub-project. However, it was not possible to classify several dykes unambiguously as belonging to a particular formation (e.g the Khandala / Poladpur-type dikes). In some previous studies, the Pb isotopic composition of flows has been found to be an important tool in discriminating between formations that have similar Sr and Nd isotopic compositions. It would be highly desirable to obtain Pb isotopic data for the Sangamner dikes in order to determine whether more precise correlation is possible. As mentioned in Chapter 2, however, several other types of information are also needed in order to better understand the relationship of dike swarms from the DVP with exposed flow packages. The most critical of these is the identification of individual eruptive units (flow fields) on the basis of integrated field and geochemical studies. Another fruitful line of study could be to subject the Sangamner dikes to a detailed petrogenetic analysis. This analysis will require acquisition of mineral compositions, additional isotopic data and quantitative modeling but should help in further understanding the generation and evolution of the Deccan magmas. Many of the dykes have well-preserved chilled margins and fresh phenocrysts – glass composition and melt inclusion studies are hence possible and deserve to be undertaken.

Sub-project 2 Results of this sub-project clearly reveal the morphological and textural diversity in the Steens Basalt lava flows. Flows transitional between pahoehoe and a’a appear to be quite common – these deserve further scrutiny. Future work could subject these to detailed textural studies using the Scanning Electron Microscope (SEM) as well as Computer Aided Tomography (CAT) – such techniques are increasingly being used to understand the textural characteristics of lava flows (crystallinity, crystal orientation, etc), which have an important bearing on their emplacement. These studies, particularly on glassy margins, will help in estimating the pre-

207 eruptive viscosity and yield strength. Textural analysis and thermal modeling of the unique mega-plagioclase phyric flows should also be undertaken in order to determine how such crystal- rich lavas can behave in a relatively fluid manner. The relationship between morphology, texture and composition at a couple of locations hints at links to deeper processes. The alternation of different morphological and textural types (e.g. aphyric and plagioclase phyric) is also quite enigmatic. It will be very interesting to characterize 2 or 3 thick sections of the Steens Basalt in terms of their morphology and composition in order to recognize products of single eruptive episodes. This is likely to shed light on the spatial and temporal nature of individual Steens eruptions, and help understand whether the textural variations could be linked to processes in shallow magma chambers or mush columns.

Sub-project 3 The JVVF provides numerous insights into the make-up and evolution of monogenetic basaltic volcanism. Small-scale source heterogeneity is strongly implicated by the geochemical data in their spatio-temporal context. In the future, it will be important to extensively sample a couple of vents other than the Coffeepot Crater and Rocky Butte in order to determine whether within-vent chemical and isotopic heterogeneity is observed. Whole rock compositional data, however, have limitations in terms of the resolution at which they help understand geologic processes. So, for example, cryptic crustal contamination is difficult to recognize but its role needs to be evaluated in light of studies from other MVF that call upon such a process. Similarly, the contribution of multiple mantle sources, although suggested by whole rock data, can be difficult to characterize in detail because of possible overprinting by differentiation processes. On the other hand, such aspects are often recorded faithfully by individual minerals, groundmass, and melt inclusions. Acquisition of mineral-scale chemical and isotopic data for the JVVF samples will be very fruitful in corroborating the existence of small-scale heterogeneity and recognizing if cryptic contamination did indeed occur, and if so, whether the contaminant was crustal material or the lithospheric mantle. Integration of the JVVF dataset with that for monogenetic volcanoes on the southern Owyhee Plateau will also be a very useful endeavor in shedding additional light on basaltic volcanism in this area. Fieldwork during this study along the Owyhee Canyon and previous work along the Jordan Creek reveals that numerous lava flows entered lakes or dammed rivers temporarily – this could have had a significant effect on the

208 drainage in this area, and provides an opportunity to understand the sub-regional uplift and incision history. Geomorphologists from the Central Washington University are already looking at the Late Quaternary lava flows in this regard, but it would be very interesting to expand such work to include the Pliocene flows and associated material. This would help understand the complex interplay of volcanism, sedimentation and fluvial incision that this area has witnessed since at least the late Miocene.

209 APPENDIX 1: ANALYTICAL METHODS

SAMPLE PREPARATION All rock samples were prepared for analysis through the following methods. Fist-sized or larger hand samples were reduced using a hydraulic rock splitter and a representative sample of each was saved for reference. Split fragments for selected samples were cut for thin sections and standard (27 x 46 mm) thin sections were obtained commercially. Weathering rinds were then trimmed from the remaining sample and postmagmatic mineralization in void space within the sample was removed by the saw. A silicon carbide belt grinder was used to remove metal shavings from trimmed fragments prior to crushing through a jaw crusher. Prior to being passed through a Braun Chipmunk brand jaw crusher (steel crushing plates), rock fragments were washed in deionized water and kept at room temperature until dry. Once through a primary crush, samples were handpicked when necessary to remove any other post-magmatic mineralization and passed through again. The entire jawcrushed sample was then passed through the alumina ceramic crushing plates of a Braun brand pulverizer (“disk-mill”). An approximately 30 ml aliquot of this was then run through a Spex brand alumina ceramic shatterbox for 10 minutes to ensure a fine grained powder for major and trace element flux fusions. Final powders were placed in borosilicate glass vials, dried uncapped at 110º C overnight, and then stored in a desiccator (capped).

LOSS ON IGNITION (LOI) LOI was determined for almost all rock samples analyzed for major and trace element chemistry. Approximately 1 gram of sample was weighed into a ceramic crucible and the weight was recorded. The powder and crucible were heated to 950º C for one hour, cooled at room temperature in a dessicator, and then weighed. LOI (in weight percent) was calculated by using the following equation: LOI = (Initial weight – final weight)/initial weight)*100.

MAJOR AND TRACE ELEMENT ANALYSIS The whole rock powders were analyzed for major and selected trace elements by Direct Current Argon Plasma Spectroscopy (DCP-AES) and X-Ray Fluorescence (XRF) at Miami University and Franklin and Marshall College respectively.

210 Samples analyzed by DCP-AES analysis were prepared by manually mixing 200 mg of

sample with 600 mg of purified LiBO2 flux (Spectroflux 100A, manufactured by Johnson Mathey Materials Technology). This mixture was then fused in a graphite crucible at 950º C for 20 minutes. The molten bead generated during the fusion process was dropped into 50 ml of a

6% HNO3 solution spiked with 3000 ppm Li, 10 ppm Ge, and 20 ppm Cd and shaken vigorously until dissolved. This solution stood at room temperature for at least 12 hours and was then ready for analysis (the trace element stock solution; TESS). A 1 ml aliquot of the TESS was added to

25 ml of a 6% HNO3 solution spiked with 3000 ppm Li and 30 ppm Ge. Li was present as a plasma enhancer and matrix suppressant, while the Ge (major) and Cd (trace) spikes are used as internal references that within-run background corrected element intensities are normalized to for reduction of in-run machine/plasma drift. In every major element run, one procedural blank

(LiBO2 only) and 9-10 international rock standards were prepared and analyzed with the batch of unknown samples. Throughout the procedure, precise masses of all measured solids and solutions were recorded so that dilution factors and final concentrations could be determined accurately. During each analytical run, each sample solution was analyzed three times and the standard solutions and blank were analyzed four times. A multi-element cassette was used to allow simultaneous determination of major elements. Data collection and reduction were performed by specific software written for the DCP-AES at Miami University. Major element concentrations were determined by comparison of the intensity ratio if an unknown to calibration curves generated by the rock standards run with a given set of samples. The reader is referred to Katoh et al. (1999) for further details regarding the DCP methods employed. XRF trace element analyses (Rb, Sr, Ba, Y, V, Zr, Nb, Ni, Cr, Ga, Cu, Zn, U, Th, Co, Pb, Sc, La and Ce) were performed on all samples in this study by Dr. Stan Mertzman at Franklin and Marshall College, following the methods described in Mertzman (2000). These are outlined and described at the following Internet site: http://server1.fandm.edu/departments/earthandenvironment/facilities/xrf/index.html. Note that some elements, particularly La, Ce, U, Th and Pb were not analyzed for some of the samples.

211 RARE EARTH ELEMENT ANALYSIS A suite of 10 rare earth elements (REE) were analyzed by ICP-MS at Miami University. Two different procedures were used: (A) An older procedure involving column separation for selected Sangamner samples and a subset of selected Jordan Valley samples, and (B) A newer procedure that did not involve column separation for the remaining subset of the Jordan Valley samples. Each is described below. (A) Sample dissolution was achieved using 100 mg of sample powder thoroughly mixed with 300 mg of Li metaborate flux. This mix was placed in a graphite crucible and fused in a furnace

at 950 °C for 20 minutes. The resultant molten bead was dissolved in 50 ml of 5% HNO3 and the solution was loaded onto a 1 cm diameter quartz glass column containing 22 g of AG50W-X8 cation exchange resin. The non-REE fraction of the sample was removed by using 210 ml of a mixed (0.1M Oxalic and 2M HNO3) acid. After this step, 250 ml of quartz-distilled water was passed through the column, and REE were collected in Teflon beakers using 200 ml of 5M

HNO3. After drying down on a hot plate at low heat, the REE fraction was re-dissolved in 1 ml

of 5M HNO3 and 5 ml quartz-distilled water, transferred to 20 ml Teflon beakers and dried down again. The residue was dissolved in 18 ml of 1% HNO3 and the weight of this solution was recorded. The REE were analyzed from this solution using a Varian ICP-MS (plasma power of 1.37 kW and pump rate of 5 rpm). Measurements for each sample included 5 replicates, with 40 scans per replicate. Calibration curves (linear regressions) for each element were generated from three standard solutions with known concentrations (5, 160, and 500 ppb respectively) and an acid blank, that were run along with the unknowns. Concentrations of unknown samples were calculated from these curves, with 115In serving as the internal standard to monitor and correct for instrumental drift. Two additional blank solutions and a 500 ppb solution were also run as unknowns. Measured concentrations for the 500 ppb solution were within 3% of this value. Concentrations for each element were corrected using the average blank values; however, the blank concentrations were quite low and these corrections did not make a significant difference to the values. The USGS standard AGV-2 was measured along with the samples to monitor accuracy. Differences in measured and accepted values are <10% for all elements except Sm, for which the difference is 13%. (B) Sample dissolution was achieved using 50 mg of sample powder, thoroughly mixed with 50 mg of flux (3:2 mixture of Sodium Tetraborate and Potassium Carbonate). This mixture was

212 placed in a graphite crucible and fused in a furnace at 950 °C for 20 minutes. The resultant

molten bead was dissolved in 25 ml of 5% HNO3, and this solution was subsequently diluted by adding 100 ml Quartz-distilled water. Precise weights were recorded at each stage. The REE were analyzed from this solution using a Varian ICP-MS (plasma power of 1.37 kW and pump rate of 5 rpm). Measurements for each sample included 5 replicates, with 40 scans per replicate. Calibration curves (linear regressions) for each element were generated from 10 international standards and an acid blank that were run along with the unknowns. High-silica standards were not considered for La and Ce, since they were observed to skew concentrations to higher values. Solutions for the international standards were made at the same time and using the same methods as for the samples described above. Concentrations of unknown samples were calculated from these curves, and an internal standard correction was made using an interpolated value between 115In and 185Re to correct for instrumental drift. The 2 SD errors on limited replicate analyses of selected samples are < 6%. Accuracy was tested by analyzing USGS standard AGV-2 as well as comparing measured values of samples analyzed by the column method described above. The results for most elements are similar to those for the procedure described above.

ISOTOPIC ANALYSIS Whole-rock Sr, Nd and Pb isotopic compositions of selected samples (chosen based on their relevance to the particular sub-project) were measured in static mode using a Thermo- Finnigan Triton multi-collector TIMS at Miami University. For the selected Sangamner samples, only Sr and Nd isotopic compositions were measured. Separation of Sr and a bulk LREE fraction was achieved using methods similar to those described in Walker et al. (1989) and Snyder (2005). Nd was separated from the remaining LREE using an EiChrom Ln-Spec resin, following methods similar to Pin and Zalduegui (1997). For the selected Jordan Valley samples, Sr, Nd and Pb isotopic compositions were measured. Separation of Pb was achieved using methods similar to those described in Walker et al. (1989). Material eluted before Pb was kept for further separation of Sr and Nd. Sr was separated using procedures similar to Deniel and Pin (2001). Nd was separated from the remaining LREE as described above. The isotopic analysis (including column separation) was repeated for certain samples in order to check the reproducibility of the obtained values.

213 Sr isotopic ratios were corrected for fractionation using 86Sr/88Sr=0.1194. A 2 SD (standard deviation) external reproducibility of 1.4 x 10-5 based on sixty eight measurements of standard NBS 987, which resulted in an average 87Sr/86Sr=0.710236, is quoted for all ratios. Nd isotopic ratios were corrected for fractionation using 143Nd/144Nd=0.7219. A 2 SD external reproducibility of 7 x 10-6 based on sixty one measurements of the La Jolla standard, which resulted in an average 143Nd/144Nd=0.511846, is quoted for all ratios. For the Sangamner dike samples, the fractionation corrected, measured ratios were age-corrected using whole-rock Rb and Sr (XRF) and Sm and Nd (ICP-MS) concentrations. 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb were fractionation corrected by 0.1% per amu based on measured ratios in NBS 981 from values in Todt et al. (1996). Errors on measured values of 206Pb/204Pb, 207Pb/204Pb, and 208Pb/204Pb (2 SD) based on fifty-eight measurements of NBS 981 were ± 0.01, ± 0.02, and ± 0.06 respectively.

References

Katoh, S., Danhara, T., Hart, W. K., and WoldeGabriel, G (1999). Use of sodium polytungstate solution in the purification of volcanic glass shards for bulk chemical analysis. Nature and Human Activities, v. 4, pp. 45-54.

Mertzman, S. A (2000). K-Ar results from the southern Oregon - northern California Cascade Range. Ore. Geol, v. 62, pp. 9-122.

Deniel, C., and Pin, C (2001). Single-stage method for the simultaneous isolation of lead and strontium from silicate samples for isotopic measurements. Anal. Chim. Acta, v. 426, pp. 95- 103.

Todt, W., Cliff, R. A., Hanser, A., and Hoffman, A. W (1996). Evaluation of a 202Pb-205Pb double spike for high-precision analysis: In: Hart, S. R., and Basu, A (eds). Earth processes: Reading the isotopic code. Geophys. Monograph, AGU, v. 95, pp. 429-437.

Walker, R. J., Carlson, R. W., Shirey, S. B., and Boyd, S. B (1989). Os, Sr, Nd and Pb isotope systematics of southern African peridotite xenoliths: implications for the chemical evolution of subcontinental mantle. Geochim. Cosmochim. Acta, v. 53, pp. 1583-1595.

214 APPENDIX 2: SUMMARY OF METHODS USED IN 40AR-39AR GEOCHRONOLOGY

40Ar/39Ar ages were obtained at Oregon State University (OSU) for seven samples representing previously undated vents from the Jordan Valley Volcanic Field, and one additional flow (Table 1). Minimally altered portions of these samples were used for this purpose. Except for N17/03, the age for which was obtained on groundmass, all other ages were obtained on whole-rock samples. Samples were irradiated at the OSU TRIGA reactor for 7 hours at 1MW power. The Neutron flux measured using FCT-3 biotite monitor (Renne et al. 1994) and the data were reduced by ArArCALC software (Koppers, 2002). Plateau ages were preferred for all samples except JV99-1 and N17/03. For these latter samples, normal isochron ages were preferred. A summary of these results is presented in Table 1. Also attached are the relevant age spectra and isochron diagrams for each dated sample (Figures 1-7). Details of the procedures used at Oregon State University are provided in Jordan et al. (2004).

References

Jordan, B. T., Duncan, R. A., Grunder, A. L (2004). Geochronology of age-progressive volcanism of the Oregon High Lava Plains; implications for the plume interpretation of Yellowstone, J. Geophys. Res, v. 109, doi:10.1029/2003JB002776.

Koppers, A. P (2002). ArArCALC; software for (super 40) Ar/ (super 39) Ar age calculations. Computers &. Geosciences, v. 28, pp. 606-619.

Renne, P. R., Becker, T., Curtis, G. H., Deino, A. L., Jaouni, A. R., Sharp, W. D., Swisher, C. C. III., Turrin, B. D., and Walter, R. C (1994). Intercalibration of astronomical and radioisotopic time. Geology, v. 22, pp. 783-786.

215 Table 1

Sample Material Plateau Age Steps* %39Ar MSWD Isochron Age 40Ar/36Ar (± 2 σ) Total Fusion Age (Ma; ± 2 σ) (Ma; ± 2 σ) (± 2 σ)

JV99-1† whole rock 88 ± 28 Ka 5/7 81.2 2.60 61 ± 23 Ka 334.4 ± 17.8 179 ± 41 Ka N2/03 whole rock 5.64 ± 0.43 8/9 96.7 1.50 5.92 ± 1.18 293.7 ± 6.7 5.86 ± 0.42 N6/04 whole rock 1.92 ± 0.22 7/7 100.0 0.53 1.86 ± 0.33 297.2 ± 8.2 2.01 ± 0.31 N9/03 whole rock 1.86 ± 0.23 7/8 97.0 0.32 1.84 ± 0.43 295.8 ± 7.0 2.16 ± 0.36 N9/04 whole rock 1.87 ± 0.08 8/8 100.0 0.72 1.86 ± 0.12 295.9 ± 3.6 1.84 ± 0.15 N17/03† groundmass 125 ± 24 Ka 4/7 89.0 0.17 86 ± 33 Ka 305.4 ± 4.4 201 ± 39 Ka N18/03 whole rock 2.78 ± 0.17 6/8 93.2 0.72 2.75 ± 0.31 295.8 ± 3.9 3.91 ± 0.47

* Plateau age data includes number of steps in the plateau (steps in plateau / total steps) and % 39Ar in plateau; † Isochron ages

216 preferred over plateau ages for these samples

Figure 1

Sample no. JV99-1

800 Ar-Ages in Ka

700 Weighted Plateau 90.8 ± 37.3 Total Fusion 600 178.8 ± 40.7 Normal Isochron 61.4 ± 23.4 500 Inverse Isochron 217

Ar 61.6 ± 22.4

36 400

Ar / MSWD

40 0.45 300

Sample Info 200 Material: whole rock 100 Irradiation: OSU1D06 J-value: 0.002089

0 0 1250 2500 3750 5000 6250 7500 8750 10000 39Ar / 36Ar

Figure 2

Sample no. N2/03

24 Ar-Ages in Ma

21 Weighted Plateau 5.64 ± 0.43 Total Fusion 18 5.86 ± 0.42 Normal Isochron 5.92 ± 1.18 15 Inverse Isochron

) 5.94 ± 1.15 a M 12 e ( 218 MSWD Ag 5.64 ± 0.43 Ma 1.50 9

Sample Info 6 Material: whole rock 3 Irradiation: OSU1D06 J-value: 0.002075

0 0 102030405060708090100 Cumulative 39Ar Released (%)

Figure 3

Sample no. N6/04

8.0 Ar-Ages in Ma

7.0 Weighted Plateau 1.92 ± 0.22 Total Fusion 6.0 2.01 ± 0.31 Normal Isochron 1.86 ± 0.33 5.0 Inverse Isochron 1.86 ± 0.33 a)

(M 4.0 e 219 MSWD Ag 1.92 ± 0.22 Ma 0.53 3.0

Sample Info 2.0 Material: whole rock 1.0 Irradiation: OSU1D06 J-value: 0.002063

0.0 0 102030405060708090100 Cumulative 39Ar Released (%)

Figure 4

Sample no. N9/03

16 Ar-Ages in Ma

14 Weighted Plateau 1.86 ± 0.23 Total Fusion 12 2.16 ± 0.36 Normal Isochron 1.84 ± 0.43 10 Inverse Isochron 1.84 ± 0.43 a)

(M 8 e 220 MSWD Ag 0.32 6

1.86 ± 0.23 Ma Sample Info 4 Material: whole rock 2 Irradiation: OSU1D06 J-value: 0.002053

0 0 102030405060708090100 Cumulative 39Ar Released (%)

Figure 5

Sample no. N9/04

8.0 Ar-Ages in Ma

7.0 Weighted Plateau 1.87 ± 0.08 Total Fusion 6.0 1.84 ± 0.15 Normal Isochron 1.86 ± 0.12 5.0 Inverse Isochron 1.86 ± 0.11 a)

(M 4.0 e 221 MSWD Ag 1.87 ± 0.08 Ma 0.72 3.0

Sample Info 2.0 Material: whole rock 1.0 Irradiation: OSU1D06 J-value: 0.002038

0.0 0 102030405060708090100 Cumulative 39Ar Released (%)

Figure 6

Sample no. N17/03

560 Ar-Ages in Ka

490 Weighted Plateau 133.9 ± 52.1 Total Fusion 420 200.5 ± 39.1 Normal Isochron 85.5 ± 32.8 350 Inverse Isochron

Ar 85.9 ± 31.2

36 280 222 Ar / MSWD

40 0.60 210

Sample Info 140 Material: 70 groundmass Irradiation: OSU1D06 0 J -value: 0.001987 0 225 450 675 900 1125 1350 1575 1800 39Ar / 36Ar

Figure 7

Sample no. N18/03

40 Ar-Ages in Ma

35 Weighted Plateau 2.78 ± 0.17 Total Fusion 30 3.91 ± 0.47 Normal Isochron 2.75 ± 0.31 25 Inverse Isochron 2.75 ± 0.31 a) M 20 e ( 223 g MSWD A 0.72 15

Sample Info 10 2.78 ± 0.17 Ma Material: whole rock 5 Irradiation: OSU1D06 J-value: 0.002026

0 0 102030405060708090100 Cumulative 39Ar Released (%)

APPENDIX 3: SAMPLE LOCATIONS AND DESCRIPTIONS

The general format of the sample description is as follows: Location and geologic context, followed by petrographic description at the hand-specimen scale and then at the thin- section scale (where available).

DECCAN DIKES SUB-PROJECT

Sample ID: CH1 Chemostratigraphic Affinity: Khandala Fm (Dhak Dongar Member) Latitude: ? Longitude: ? Description: Sample from large hump/spine of the NE-SW trending, 11 m wide dike exposed to the south side of the road leading to Pimpalgaon Matha. Fine grained aphyric texture.

Sample ID: Ch4b Chemostratigraphic Affinity: Poladpur Fm? Latitude: 19º 27/ 43.8// Longitude: 74º 08/ 24.3// Description: Stubby, columnar flow overlying the E-W trending dike at Pimpalgaon Matha. Base of flow welded to the pyroclastic material associated with the dike.

Sample ID: Ch4c Chemostratigraphic Affinity: Ambenali Fm. Latitude: 19º 27/ 43.8// Longitude: 74º 08/ 24.3// Description: Sample of E-W trending, 5 m wide dike exposed at Pimpalgaon Matha. Location shows strong evidence for this dike being a feeder. The dike appears to intrude crudely bedded spatter and appears to form a stubby columnar flow. This dike is also exposed intermittently in the Chandanapuri Valley, and has abundant off-shoots.

Sample ID: Ch5 Chemostratigraphic Affinity: Poladpur Fm. Latitude: 19º 32/ 01.9// Longitude: 74º 03/ 08.9// Description: Sample of N-W trending, 6 m wide Thugaon Budruk dike. Dike is visibly altered; sample was taken from the core of a spheroidally weathered column. Sample shows abundant olivine phenocrysts altered to iddingsite. Thin section shows highly inequigranular with phenocrysts of plagioclase, olivine and rare clinopyroxene. Groundmass consists of plagioclase, clinopyroxene, and abundant oxides. Olivines are euhedral but altered to iddingsite along fractures. Plagioclase crystals are interestingly anhedral.

Sample ID: Ch6 Chemostratigraphic Affinity: Poladpur / Bhimashankar Fm. Latitude: 19º 32/ 34.7// Longitude: 73º 58/ 13.7// Description: Sample of NE-SW trending, 6 m wide Unchhakhadak dike. This dike is seen to dip around 22º to the East. Thin section shows a relatively fine texture with several glomerocrysts of plagioclase and clinopyroxene, along with plagioclase phenocryts in a groundmass of plagioclase, 224 clinopyroxene, patches of microcrystalline residuum, and scattered oxides. There appears to be a wide variation in the sizes of plagioclase phenocrysts and microphenocrysts.

Sample ID: Ch7a Chemostratigraphic Affinity: Khandala (Dhak Dongar Member) / Bhimashankar Fm. Latitude: 19º 33/ 24.9// Longitude: 73º 58/ 28.7// Description: Sample of NE-SW trending, 3 m wide dike just west of the Unchhakhadak dike. It is explosed along the channel of the Pravara River and at one location becomes slightly sinuous, with twisted columns suggesting syn-intrusion deformation. Sample is plagioclase phyric with a fine grained groundmass.

Sample ID: Ch8 Chemostratigraphic Affinity: ? Latitude: 19º 33/ 23.0// Longitude: 73º 55/ 36.5// Description: Sample of NE-SW trending, 10 m wide dike at Mhaladevi forming a prominent spine. Dike dips 50º to the South. Plagioclase phyric with a fine grained groundmass.

Sample ID: Ch9 Chemostratigraphic Affinity: ? Latitude: 19º 33/ 20.5// Longitude: 73º 55/ 34.9// Description: Sample of NE-SW trending, approximately 8 m wide dike parallel to above dike at Mhaladevi. Plagioclase phyric with a fine grained groundmass.

Sample ID: Ch10 Chemostratigraphic Affinity: Khandala Fm. (Dhak Dongar Member) / Bhimashankar Fm. Latitude: 19º 32/ 44.3// Longitude: 73º 56/ 05.0// Description: Sample of NE-SW trending, 12 m wide dike at Rhumbodi forming a prominent spine. Dike dips 50º to the South. Highly plagioclase phyric with large (giant) plagioclase phenocrysts with long axes aligned parallel to the axis of the dike. Thin section shows some large plagioclase phenocrysts in a relatively coarse groundmass. The latter contains clinopyroxene and plagioclase in a sub-ophitic relationship and numerous patches of microcrystalline residuum and brown glass.

Sample ID: Ch11 Chemostratigraphic Affinity: Khandala Fm. (Dhak Dongar Member) Latitude: 19º 32/ 44.3// Longitude: 73º 56/ 05.0// Description: Sample of NE-SW trending, 1-2 m wide dike at Mhaladevi, parallel to above dike. Sample contains some moderately large phenocrysts of plagioclase, but much less than the above dike.

Sample ID: Ch12 Chemostratigraphic Affinity: Poladpur/Bhimashankar Fm. Latitude: 19º 34/ 58.6// Longitude: 74º 00/ 37.1// Description: Sample of NE-SW trending, 17 m wide dike exposed at Chaphewadi near Gardani. The dike dips steelply to the East (75-80º) and has a glomeroporphyritic texture. Thin section

225 shows plagioclase glomerocrysts that are much larger than groundmass crystals of plagioclase and are distinctly zoned. Very few phenocrysts of olivine are present, altered to iddingsite. Near this location, the dike appears to pass locally into a sill, and shows columns with widely varying orientations.

Sample ID: Ch13 Chemostratigraphic Affinity: Poladpur Fm. Latitude: 19º 34/ 44.9// Longitude: 74º 00/ 43.8// Description: Sample of NE-SW trending, 6 m wide dike exposed at Chaphewadi, just south of above dike. Sparsely plagioclase phyric with a fine grained groundmass. Thin section shows 1 or 2 glomerocrysts of plagioclase in a hypocrystalline groundmass of sub-ophitic plagioclase and clinopyroxene, and abundant glass. Oxides are extremely sparse; presumably the dike cooled before they could crystallize from the residual liquid.

Sample ID: Ch14 Chemostratigraphic Affinity: Poladpur/Bhimashankar Fm. Latitude: 19º 34/ 49.6// Longitude: 74º 02/ 19.1// Description: Sample of NE-SW trending, c. 4 m wide dike exposed at Tambol forming a prominent hump. Plagioclase phyric with a relatively fine grained groundmass. Thin section shows that the groundmass consists of plagioclase and clinopyroxene in a sub-ophitic relationship, but a few pockets of brown glass are also seen. Oxides are scattered throughout the thin section.

Sample ID: Ch15 Chemostratigraphic Affinity: Poladpur/Bhimashankar Fm. Latitude: 19º 34/ 56.4// Longitude: 74º 02/ 15.1// Description: Sample of NE-SW trending plagioclase phyric dike exposed just north of above dike at Tambol. Thin section shows that the groundmass (coarser than the above dike) consists of plagioclase and clinopyroxene in a sub-ophitic relationship. This dike also has some pockets of granophyric material. Abundant, relatively large oxide grains are present.

Sample ID: Ch16 Chemostratigraphic Affinity: Poladpur Fm. / Bhimashankar Fm. Latitude: 19º 33/ 41.9// Longitude: 74º 05/ 43.7// Description: Sample of NE-SW trending, 6 m wide glomeroporphyritic dike exposed as a large hump at Wadgaon Landga. Thin section shows plagioclase glomerocrysts that are much larger than groundmass crystals. Some of these show inclusions of what appear to be oxides, aligned along the lamellae. Groundmass consists of sub-ophitic plagioclase and clinopyroxene with scattered oxides.

Sample ID: Ch17 Chemostratigraphic Affinity: Bushe Fm. / Boyhare Member of Khandala Fm. Latitude: 19º 33/ 49.5// Longitude: 74º 05/ 44.6// Description: Sample of NE-SW trending, 3 m wide dike exposed at Wadgaon Landga just north of above dike. It appears highly altered and contains olivine phenocrysts. Thin section reveals

226 abundant olivine phenocrysts, sometimes in clumps, altered to iddingsite but preserving their euhedral form.

Sample ID: Ch18 Chemostratigraphic Affinity: Khandala Fm. (Rajmachi Member?) Latitude: 19º 37/ 44.3// Longitude: 74º 10/ 59.4// Description: Sample of NE-SW trending dike exposed at Sonoshi. Plagioclase phyric with a fine grained groundmass. At contact with bedrock, it appears that small intrusions (c. 0.5 m wide) are associated with the dike. They look like squeeze-ups in pahoehoe flows; one also has pipe vesicles along the margins. Dike columns are inclined at places, suggesting warping during cooling.

Sample ID: Ch19 Chemostratigraphic Affinity: Khandala Fm. (Rajmachi Member?) Latitude: 19º 37/ 50.8// Longitude: 74º 11/ 05.7// Description: Sample of NE-SW trending, 10 m wide dike exposed just north of above dike at Sonoshi.

Sample ID: Ch20 Chemostratigraphic Affinity: Poladpur Fm? Latitude: 19º 38/ 51.3// Longitude: 74º 09/ 43.1// Description: Sample of NNE-SSW trending, 4 m wide dike exposed in Karhe Ghat. Dikelets and squeeze-ups are associated with this dike. Squeeze-ups feed flow lobes, but it is not clear whether they are sourced from the dike.

Sample ID: Ch21 Chemostratigraphic Affinity: Poladpur / Bhimashankar Fm. Latitude: 19º 42/ 12.6// Longitude: 74º 08/ 57.3// Description: Sample of NE-SW trending, c. 8 m wide dike exposed just north of above dike. Moderately large, scattered plagioclase phenocrysts in a fine grained groundmass. Dike appears to have more than one chilled margin, suggesting that it might be multiply intrusive. The fine grained margins have been eroded into gullies at places.

Sample ID: Ch22 Chemostratigraphic Affinity: Khandala Fm. (Dhak Dongar Member?) Latitude: 19º 42/ 06.0// Longitude: 74º 07/ 01.7// Description: Sample of E-W trending dike exposed on the way to Chas from Nandur Shingote. Width uncertain, plagioclase phyric with a very fine grained groundmass. Thin section shows plagioclase phenocrysts and some olivine phenocrysts altered to iddingsite. Groundmass is fine grained with plagioclase, clinopyroxene, oxides and glass. At places, the glass has devitrified.

227 Sample ID: Ch23 Chemostratigraphic Affinity: Khandala Fm. (Rajmachi Member?) Latitude: 19º 37/ 50.8// Longitude: 74º 11/ 05.7// Description: Sample of NE-SW trending, 10 m wide plagioclase phyric dike exposed just north of above dike at Sonoshi. Texture very similar to above dike.

Sample ID: Ch24 Chemostratigraphic Affinity: Ambenali Fm.? Latitude: 19º 27/ 24.4// Longitude: 74º 01/ 48.1// Description: Sample of E-W trending, 5.5 m wide dike exposed on the way to Lingdev from Kotul Junction. Fine grained, no apparent glomerocrysts of plagioclase.

Sample ID: Ch25 Chemostratigraphic Affinity: Poladpur Fm. Latitude: 19º 27/ 13.5// Longitude: 74º 01/ 53.6// Description: Sample of ENE-WSW trending narrow (3.5 m) dike, just south of above location on the way to Lingdev. Sample taken from edges, which are fine grained and fresh without obvious phenocrysts. Central part of the dike has plagioclase glomerocrysts; however, this part was altered and hence was not sampled. Dike also has several small off-shoots.

Sample ID: Ch26a Chemostratigraphic Affinity: Khandala Fm. (Dhak Dongar Member?) Latitude: 19º 25/ 16.6// Longitude: 74º 03/ 07.7// Description: Sample of E-W trending, 3 m wide dike exposed south of Lingdev. Sample from fine grained, non porphyritic margin. Central part of the dike has plagioclase glomerocrysts – this was sampled (Ch26b) but not analyzed.

Sample ID: Ch27 Chemostratigraphic Affinity: Khandala Fm. (Dhak Dongar Member?) Latitude: 19º 25/ 11.6// Longitude: 74º 03/ 10.8// Description: Sample of E-W trending, 3 m wide dike exposed south of Lingdev. Sample from fine grained, non porphyritic margin. Central part of the dike has plagioclase glomerocrysts – this was sampled (Ch26b) but not analyzed.

Sample ID: Ch27 Chemostratigraphic Affinity: Khandala Fm. (Dhak Dongar Member?) Latitude: 19º 25/ 11.6// Longitude: 74º 03/ 10.8// Description: Sample of NE-SW trending dike exposed just south of above dike. Sample fine grained, from margin of dike.

228 Sample ID: Ch28 Chemostratigraphic Affinity: Khandala Fm. (Dhak Dongar Member?) Latitude: 19º 24/ 53.9// Longitude: 74º 03/ 22.6// Description: Sample of E-W trending dike (or pair of dikes?) just south of the above location. It appears as though there are two dikes separated by a narrow bedrock exposure, although it could be the same dike that bifurcated.

Sample ID: Ch29 Chemostratigraphic Affinity: ? Latitude: 19º 22/ 55.5// Longitude: 74º 05/ 25.4// Description: Sample of coarse grained, olivine phyric material just south of Chas on east side of road. Material has a bouldery outcrop and could be part of a sill. Thin section shows large, distinctly elongated olivine crystals (>2 mm), altered along cracks to iddingsite. Groundmass with plagioclase and clinopyroxene in an intergranular/sub-ophitic relationship, and scattered, sparse oxides.

Sample ID: Ch30 Chemostratigraphic Affinity: Poladpur Fm. Latitude: 19º 22/ 43.0// Longitude: 74º 05/ 36.2// Description: Sample of NE-SW trending dike exposed some kilometers south of Chas. Thin section shows numerous olivine phenocrysts (1-2 mm), completely altered to iddingsite. Groundmass with numerous, small clinopyroxene grains between plagioclase laths, and scattered oxides.

Sample ID: Ch31 Chemostratigraphic Affinity: ? Latitude: 19º 22/ 25.5// Longitude: 74º 06/ 06.8// Description: Sample of potentially the same sill as earlier (Ch29) – has an expression different from flows from this area. Crudely columnar, coarse grained, olivine phyric; being mined for rock. Thin section shows abundant, large, and often elongated olivine phenocrysts (4-5 mm) in a groundmass with plagioclase and clinopyroxene in an intergranular relationship and scattered oxides.

STEENS SUB-PROJECT (UTM ZONE 11T)

Previous workers have divided the type Steens sequence into two groups, viz. ‘upper Steens basalt’ and ‘lower Steens basalt’ based on age, stratigraphic position and geochemical characteristics. Samples from this study have also been assigned to these two chemostratigraphic groups using similar criteria (see chapter 4 for details). The informal term ‘middle Steens’ has been used to classify certain samples from sections geographically removed from the type section based on their stratigraphic position.

229 Sample ID: N11A/04 Chemostratigraphic Group: lower Steens? Easting: 347079 Northing: 4698342 Description: Sample of uppermost, ~10 m thick a’a flow collected from a section along the Catlow Rim. Fine grained, aphyric with compact and platy appearance. Thin section shows two domains: one domain has subophitic/ophimottled and intergranular plagioclase, clinopyroxene and olivine. The other domain has dominantly altered olivine/iron oxides, microvesicles, and less plagioclase and clinopyroxene.

Sample ID: N11B/04 Chemostratigraphic Group: lower Steens? Easting: 347079 Northing: 4698342 Description: Sample of ~ 4 m thick middle a’a flow collected from a section along the Catlow Rim. Fine grained, aphyric with compact and platy appearance.

Sample ID: N11C/04 Chemostratigraphic Group: lower Steens? Easting: 347079 Northing: 4698342 Description: Sample of lowermost ~ 4 m thick a’a flow collected from a section along the Catlow Rim. Fine grained, aphyric with compact and platy appearance.

Sample ID: N12/04 Chemostratigraphic Group: upper Steens Easting: 369840 Northing: 4721600 Description: Uppermost flow from a section close to the Steens Mountain summit, close to where the Wildhorse Trail begins. Flow is ~ 8 m thick with a generally platy appearance and brecciated base.

Sample ID: N13/04 Chemostratigraphic Group: upper Steens Easting: 370570 Northing: 4721690 Description: Sample of an almost 5 m thick transitional pahoehoe lobe close to the Steens Mountain summit (on the road to the Radio Tower). Oxidized, brecciated base and core with numerous segregation features.

Sample ID: N14/04 Chemostratigraphic Group: upper Steens Easting: 347079 Northing: 4698342 Description: A’a flow collected from the road to the Radio Tower close to the Steens Mountain summit. This flow is ~6 m thick and overlies the pahoehoe lobe mentioned above (N13/04). Sample appears banded and aphyric. Thin section shows a very fine grained texture with abundant plagioclase and oxides. The groundmass appears to be microcrystalline oxides and perhaps a feldspar that occupies spaces between plagioclase laths.

230 Sample ID: N15A/04 Chemostratigraphic Group: upper Steens Easting: 362032 Northing: 4726059 Description: Lowermost flow/flow lobe exposed in a section along the southern section of the Steens Loop Road. Preserved upper crust but extremely undulating and brecciated base. Sample appears aphyric and fine grained. Thin section shows abundant plagioclase laths arranged in sheafs, with intergranular olivine and clinopyroxene. The latter two constituents are volumetrically much lower than plagioclase.

Sample ID: N15B/04 Chemostratigraphic Group: upper Steens Easting: 362032 Northing: 4726059 Description: Sample of flow lobe overlying the above flow/flow lobe (N15A/04) in the section along the southern section of the Steens Loop Road. Preserved upper and basal crusts; sample is moderately plagioclase phyric.

Sample ID: N16A/04 Chemostratigraphic Group: lower to ‘middle’ Steens? Easting: 341014 Northing: 4713637 Description: Lower a’a flow in a section along the Catlow Rim. Flow is distinctly platy and appears altered. uppermost a’a flow collected from a section along the Catlow Rim. Fine grained, aphyric with compact and platy appearance. The thin section shows that the texture is quite similar to N11A/04, with two domains.

Sample ID: N16B/04 Chemostratigraphic Group: ‘middle’ to upper Steens? Easting: 341014 Northing: 4713637 Description: An approximately 5 m thick pahoehoe flow lobe overlying the a’a flow described above (N16A/04), with scattered large plagioclase phenocrysts (~5 mm). Thin section shows a typical dikytaxitic texture, with plagioclase phenocrysts in a groundmass of subophitic plagioclase and clinopyroxene.

Sample ID: N17A/04 Chemostratigraphic Group: ‘middle’ to upper Steens? Easting: 341051 Northing: 4713538 Description: Flow lobe of compound pahoehoe lobe overlying the a’a flow at the section along the Catlow Rim described earlier (N16A/04).

Sample ID: N17B/04 Chemostratigraphic Group: ‘middle’ to upper Steens? Easting: 341051 Northing: 4713538 Description: A’a flow at the same location as the one above, overlying the compound pahoehoe flow. Extensive, > 8m thick unit with a fine grained, platy appearance.

231 Sample ID: N17C/04 Chemostratigraphic Group: ‘middle’ to upper Steens? Easting: 341051 Northing: 4713538 Description: Flow lobe of compound pahoehoe lobe at above location, some meters north of N17A/04.

Sample ID: N18A/04 Chemostratigraphic Group: lower Steens? Easting: 343441 Northing: 4687999 Description: Sample of ~9 m thick agglutinate (?) in a vent-related section exposed along the Catlow Rim north of Funnel Canyon. Olivine and plagioclase phyric material with segregation structures. Olivine iddingsitized to varying degrees.

Sample ID: N18B/04 Chemostratigraphic Group: lower Steens? Easting: 343441 Northing: 4687999 Description: Sample of ~0.2 m thick, S-type pahoehoe lobe in the section discussed above (N18A/04). Highly vesicular and plagioclase phyric.

Sample ID: N19/04 Chemostratigraphic Group: upper Steens? Easting: 356024 Northing: 4673114 Description: < 3 m thick flow with brecciated base exposed close to the road leading to the Domingo Pass (Pueblo Mountain Section). Sample has a sugary texture with sparse plagioclase (up to 1 cm long) and olivine phenocrysts.

Sample ID: N20B/04 Chemostratigraphic Group: lower Steens Easting: 389181 Northing: 4727449 Description: Thick (~14 m) a’a flow exposed along an incised stream cutting through the Mickey Butte. Aphyric with a platy appearance. Thin section reveals patches with an ophimottled texure (clinopyroxene and plagioclase) enclosed in a mosaic defined by altered oxide or glass rich material.

Sample ID: N20C/04 Chemostratigraphic Group: lower Steens Easting: 389181 Northing: 4727449 Description: 9 m thick, highly plagioclase phyric flow underlying above flow (N20B/04). Appears to break up into lobes towards the top. Sample shows aligned, large plagioclase phenocrysts (>1 cm) and some iddingsitized olivine phenocrysts in a dense, black groundmass.

232 Sample ID: N20D/04 Chemostratigraphic Group: lower Steens Easting: 389181 Northing: 4727449 Description: Thick (~14 m) a’a flow underlying the above flow (N20C/04). Distinctly platy appearance.

JORDAN VALLEY SUB-PROJECT (UTM ZONE 11T)

Samples from the Jordan Valley Volcanic Field define two distinct chemostratigraphic groups. Group 1 consists of HAOT, transitional basalt and SROT-like vents, whereas Group 2 consists of mildly alkaline vents. Details regarding this can be found in chapter 4.

Sample ID: JC-36A Chemostratigraphic Group: 2 Easting: 462405 Northing: 4777035 Description: Flow field associated with spatter cones in main crater area. Sample from massive portion of middle flow bench of collapse pit exposing multi-level (at least three) lava tube including lava levees, flow structures on inner walls, and lava stalactites on undersides of benches. Hypocrystalline, fine grained with microphenocrysts of olivine > those of plagioclase.

Sample ID: JC-4 Chemostratigraphic Group: 2 Easting: 462689 Northing: 4777093 Description: Sample from thick section of massive, highly welded agglutinate in northeastern portion of main crater wall; material can be traced laterally to vertical wall exhibiting vertical grooves created by fluctuating lava pond crust. Dense, fine grained with fresh olivine phenocrysts.

Sample ID: JC-5 Chemostratigraphic Group: 2 Easting: 462816 Northing: 4776972 Description: Sample from surface shelly pahoehoe flow <75m southeast of crater rampart

Sample ID: JC-30B Chemostratigraphic Group: 2 Easting: 462816 Northing: 4776972 Description: Northeastern edge of main Coffeepot Crater outflow field near Groundhog Reservoir. Sample from massive portion of upper flow bench (surface flow) of collapse pit exposing multi-level (at least two) lava tubes including lava levees, flow structures on inner walls, and lava stalactites on undersides of benches. Massive with microphenocrysts and phenocrysts of olivine and plag.

233 Sample ID: H97-1A and B Chemostratigraphic Group: 2 Easting: 461460 Northing: 4776765 Description: Samples from isolated spatter ridge and associated very local flow, approximately 35 m long by 10 m wide, located along the SW extension of main Coffeepot Crater-Spatter Cone linear. Approximately 1.2 km from main crater.

Sample ID: N1/03 Chemostratigraphic Group: 2 Easting: 462289 Northing: 4771813 Description: Collected from western fringe of the Coffeepot Crater flow. Fresh, grey basalt with abundant olivine phenocrysts.

Sample ID: N2/03 Chemostratigraphic Group: 1 Easting: 456515 Northing: 4772828 Description: Sample from rim of vent southwest of Coffeepot Crater. Not much exposure around, except for this rim basalt. Sparsely olivine phyric, olivine altered at places to iddingsite. Thin section shows a holocrystalline, plagioclase and clinopyroxene are the dominant mineral constituents and define a typical sub-ophitic texture. Olivine is less abundant and occurs as scattered phenocrysts or between plagioclase laths.

Sample ID: N3/03 Chemostratigraphic Group: 1? Easting: 456226 Northing: 4772071 Description: Single, thick flow/flow lobe exposed in the rim to the south of the above vent. Large vesicles, platy appearance, with abundant glomerocrysts of olivine and plagioclase. Olivine iddingsitized at places. Thin section reveals a markedly inequigranular texture, with bundant olivine phenocrysts and some glomerocrysts of plagioclase and olivine. Most olivine crystals are iddingsitized to varying degrees. Groundmass has large clinopyroxene grains showing varying optical orientation, partially or fully enclosing plagioclase laths (ophi-mottling).

Sample ID: N4/03 Chemostratigraphic Group: 1? Easting: 455894 Northing: 4771457 Description: Flow/flow lobe across the valley to the south of above flow. > 30 feet thick, compact flow with plagioclase and olivine clumps in a dark, massive matrix.

Sample ID: N5/03 Chemostratigraphic Group: 1? Easting: 456697 Northing: 4771762 Description: Flow/flow lobe across the valley from above exposed along hillslope, apparently stratigraphically below N4/03. Open, microvesicular texture with olivine (fresh) and plagioclase phenocrysts.

234 Sample ID: N6/03 Chemostratigraphic Group: 2 Easting: 454504 Northing: 4767362 Description: Tumulus of young-looking lava (appeared to be distal flow of Clarks Butte). Sample fresh and vesicular, with olivine and plagioclase phenocrysts.

Sample ID: N7/03 Chemostratigraphic Group: 2 Easting: 449746 Northing: 4766231 Description: West Crater flow in the Bogus Creek canyon, close to the Bogus Rim. Collected from an elongated inflated lobe/tumulus. Diktitaxytic, distinctly porphyritic with prominent olivine and plagioclase phenocrysts. Thin section shows a markedly inequigranular texture with widely varying sizes of phenocrysts as well as groundmass crystals. Glomerocrysts and phenocrysts of plagioclase and olivine in a groundmass of clinopyroxene, plagioclase and some olivine.

Sample ID: N8/03 Chemostratigraphic Group: 1 Easting: 448626 Northing: 4755775 Description: Collected from the top of a small butte on the east side of the main Owyhee Butte. Appears to be a spatter cone with clastogenic lava/agglutinate. Thin section shows two textural domains. In the first, few olivine phenocrysts occur in a groundmass of clinopyroxene, plagioclase and olivine. Clinopyroxene and plagioclase are in a sub-ophitic relationship whereas groundmass olivine crystals occupy spaces between plagioclase laths. The second domain is dominated by scattered olivine and plagioclase microphenocrysts. Plagioclase laths enclose a large number of very small crystals, probably clinopyroxene that has crystallized late.

Sample ID: N9/03 Chemostratigraphic Group: 1 Easting: 447901 Northing: 4755507 Description: Summit of Owyhee butte, clastogenic lava/agglutinate with obvious plagioclase crystals. Sample is holocrystalline and thin section shows few, scattered plagioclase phenocrysts in a groundmass dominated by sub-ophitic plagioclase and clinopyroxene; few olivine crystals in intergranular relationship with plagioclase.

Sample ID: N10/03 Chemostratigraphic Group: 2 Easting: 465688 Northing: 4760894 Description: Sample from upper lava lake on the Rocky Butte summit (upper crater). Thick, glomeroporphyritic basalt with glomerocrysts of fresh, subhedral olivine and plagioclase. Thin section shows a markedly inequigranular texture with numerous glomerocrysts of plagioclase and olivine, in addition to individual phenocrysts of these minerals. Groundmass with sub-ophitic plagioclase and a purple, slightly pleochroic clinopyroxene (Titan-augite), and olivine occupying spaces between plagioclase laths. Abundant, scattered oxides.

235 Sample ID: N11/03 Chemostratigraphic Group: 2 Easting: 466112 Northing: 4760858 Description: Clastogenic lava/agglutinate from the rim of lower crater on the Rocky Butte summit. Sample is vesicular, fine-grained and appears aphyric to the naked eye; markedly different from N10/03. Thin section shows fresh, sub-hedral olivine phenocrysts in a groundmass of plagioclase laths enclosing clinopyroxene, olivine and abundant oxides. Texture suggestive of fairly rapid crystallization.

Sample ID: N12/03 Chemostratigraphic Group: 1? Easting: 457542 Northing: 4761446 Description: Sample collected from E-W fault scarp exposed to the east of West Crater, close to topographic saddle. Open texture with plagioclase and olivine phenocrysts. Thin section shows plagioclase phenocrysts (strongly zoned) and glomerocrysts, rarely with olivine phenocrysts. Groundmass with clinopyroxene, plagioclase and minor olivine. Some olivine grains enclose plagioclase laths in a poikilitic relationship.

Sample ID: N13/03 Chemostratigraphic Group: 2 Easting: 456338 Northing: 4762280 Description: A’a flow of West Crater, close to eastern edge of the West Crater flow-field. Sample vesicular with olivine and plagioclase phenocrysts. In thin section, sample shows markedly inequigranular texure with glomorocrysts of plagioclase and olivine in a much finer groundmass of plagioclase, clinopyroxene, olivine and oxides.

Sample ID: N14/03 Chemostratigraphic Group: 2 Easting: 456837 Northing: 4762627 Description: Rocky Butte flow close to its western fringe (across the road from N13/03). Sample from massive part of flow-front, with fresh olivine and plagioclase phenocrysts.

Sample ID: N15/03 Chemostratigraphic Group: 2 Easting: 456716 Northing: 4763568 Description: Sample from cleft elongate tumulus/pressure ridge showing large, domed vesicles. Massive/compact appearance, olivine microphenocrysts visible. Thin section shows olivine phenocrysts and glomerocrysts in an intersertal groundmass with plagioclase, clinopyroxene, olivine, oxides and glass. Couple of plagioclase crystals appear boxy.

Sample ID: N16/03 Chemostratigraphic Group: 1 Easting: 446889 Northing: 4767970 Description: Sample from diktytaxitic flow lobe belonging to lower basalt flow along the Bogus Rim. This flow is overlain by a thick sediment package and finally by the Bogus Rim basalt flow.

236 Thin section shows typical diktytaxitic, holocrystalline texture with olivine phenocrysts and glomerocrysts of plagioclase and olivine. Groundmass of clinopyroxene, plagioclase, olivine and oxides. Olivine altered to iddingsite along cracks and margins, otherwise fresh.

Sample ID: N17/03 Chemostratigraphic Group: 2 Easting: 446960 Northing: 4767983 Description: West Crater / Rocky Butte (?) flow on the Bogus Creek canyon floor. Sample from tumulus, porphyritic with olivine and plagioclase phenocrysts. Thin section shows a vesicular, porphyritic texture with dominant plagioclase and subsidiary olivine phenocrysts. Groundmass contains microphenocrysts of clinopyroxene and plagioclase, glass, and skeletal oxides.

Sample ID: N18/03 Chemostratigraphic Group: 1 Easting: 461946 Northing: 4756507 Description: Flow from the summit of butte due west of Three Mile Hill (“Bernie’s Butte”). Abundant plagioclase and olivine phenocrysts, sometimes in clumps. Olivine considerably iddingsitized. Thin section confirms the presence of clumps of plagioclase and olivine. Plagioclase clumps show two types of arrangement – radial and parallel. Olivine crystals show a range in sizes and appear fresh except for rims of iddingsite. Clinopyroxene occurs typically as large crystals partially or fully enclosing plagioclase laths in the groundmass. Oxides are not abundant and are scattered throughout the thin section.

Sample ID: N19A/03 Chemostratigraphic Group: 1 Easting: 462480 Northing: 4754653 Description: Olivine phyric flow lobe of compound pahoehoe flow along the Merrill Springs Rim, south of Bernie’s Butte. Thin section shows a diktytaxitic texure with glomerocrysts of plagioclase and olivine in a groundmass of plagioclase, clinopyroxene, olivine and oxides.

Sample ID: N19B/03 Chemostratigraphic Group: 1 Easting: 462480 Northing: 4754653 Description: Apparently olivine-poor flow lobe of compound pahoehoe flow along the Merrill Springs Rim, south of Bernie’s Butte. Thin section shows abundant large glomerocrysts of plagioclase and plagioclase and olivine in a groundmass coarser than N19A/03. Groundmass contains plagioclase, olivine and clinopyroxene in intergranular and subophitic relationships.

Sample ID: N20/03 Chemostratigraphic Group: 1 Easting: 471962 Northing: 4771714 Description: Single, thick, pahoehoe flow lobe overlying sediments along the rim of Cow Lakes, close to where Coffeepot Crater flow damms Cow Creek. Thin section shows a diktytaxitic, holocrystalline texture with scattered olivine phenocrysts in a groundmass of clinopyroxene and plagioclase showing an ophi-mottled texture.

237 Sample ID: N21/03 Chemostratigraphic Group: 1 Easting: 473778 Northing: 4767599 Description: Fine-grained, apparently aphyric material from vent-remnant just east of Cow Butte. This remnant is marked by banded, clastogenic lava. Thin section shows a highly vesicular texture with scattered olivine phenocrysts in a fine grained groundmass similar to N20/03. Oxides are concentrated around edges of vesicles or occur as groups of anhedral crystals at various places throughout the groundmass.

Sample ID: N22/03 Chemostratigraphic Group: 1 Easting: 473130 Northing: 4767690 Description: Summit of Cow Butte, dominated by oxidized cinder and agglutinate. Sample from float on west side of Butte, more prominent olivine and plagioclase phenocrysts than N21/03.

Sample ID: N23/04 Chemostratigraphic Group: 1 Easting: 472017 Northing: 4768529 Description: Sample from clastogenic material west of Cow Butte with sparse, iddingsitized olivine phenocrysts and plagioclase laths. Thin section shows a vesicular texture with few, scattered large plagioclase phenocrysts and glomerocrysts of plagioclase and olivine. Groundmass is fine grained groundmass with plagioclase and clinopyroxene in a subophitic relationship.

Sample ID: N24/03 Chemostratigraphic Group: 1 Easting: 472411 Northing: 4769027 Description: Sample from mound of diktytaxitic material a little north of above location. Thin section shows a texture similar to N23/03, but with abundant olivine in the groundmass.

Sample ID: JV99-1 Chemostratigraphic Group: 2 Easting: 455121 Northing: 4760985 Description: Sample from massive in-situ block in southern crater wall, likely a portion of a shield-building flow. Slightly vesicular with abundant plag microphenoscrysts and laths, fresh olivine phenocrysts, and scattered glomeroporphyritic clumps of olv+plg.

Sample ID: JV99-2 Chemostratigraphic Group: 2 Easting: 455121 Northing: 4760985 Description: Sample from block/rubble of collapsed lava tube/channel at base of northern crater wall. Petrography same as JV99-1.

238 Sample ID: H8-58 Chemostratigraphic Group: 2 Easting: 469860 Northing: 4757750 Description: Black, vesicular, olivine and plagioclase phyric Rocky Butte flow southwest of the main vent.

Sample ID: N1/04 Chemostratigraphic Group: 2 Easting: 458548 Northing: 4764951 Description: Olivine phyric basalt collected from tumulus/pressure ridge of presumed Clarks Butte flow, close to the northwestern fringe of the Rocky Butte flow. Thin section shows olivine phenocrysts in a groundmass comprising of plagioclase, olivine and glass.

Sample ID: N2/04 Chemostratigraphic Group: 1 Easting: 455760 Northing: 4760140 Description: Sample from eastern flank of potential broad shield into which West Crater is nestled. Has an old, eroded appearance with not much exposure of basalt. Diktytaxitic, olivine phyric basalt. Thin section shows a holocrystalline texture with plagioclase, olivine and clinopyroxene. Rare plagioclase glomerocrysts, olivine with iddingsitized rims.

Sample ID: N3/04 Chemostratigraphic Group: 1? Easting: 454550 Northing: 4760980 Description: Olivine phyric basalt with a less open texture than N2/04, collected from E-W fault scarp just west of West Crater.

Sample ID: N5/04 Chemostratigraphic Group: 3 Easting: 452798 Northing: 4767205 Description: Flow lobe of a compound pahoehoe flow exposed along the Bogus Rim, just ahead of above location (N4/04). Rim appears abruptly lower in elevation here. Sample looks similar to N4/04.

Sample ID: N6/04 Chemostratigraphic Group: 1 Easting: 454973 Northing: 4769801 Description: Clastogenic material on high point on the Bogus Bench. Probably remnant of a broad shield. Olivine phyric basalt with a fine-grained groundmass; olivine partially iddingsitized.

239 Sample ID: N7/04 Chemostratigraphic Group: 1 Easting: 450684 Northing: 4766440 Description: Olivine phyric basalt with an open texure, sampled from lobe of a compound pahoehoe flow exposed along the Bogus Rim, close to the beginning of the Bogus Creek canyon.

Sample ID: N8/04 Chemostratigraphic Group: 1 Easting: 444666 Northing: 4758102 Description: Sample taken from highly elongate tumulus/pressure ridge of probably Owyhee Butte flow exposed close to the Owyhee Canyon rim northwest of Owyhee Butte.

Sample ID: N9/04 Chemostratigraphic Group: 1 Easting: 446802 Northing: 4752270 Description: Sample from rim of crater of Little Owyhee Butte vent. Rampart of welded spatter/clastogenic lava, mostly dipping into the crater. Olivine phyric basalt; olivine partially iddingsitized. Thin section shows few olivine phenocrysts in a groundmass of abundant olivine, along with plagioclase, clinopyroxene and oxides.

Sample ID: N21/04 Chemostratigraphic Group: 1 Easting: 452293 Northing: 4758076 Description: Sample from the Merrill Springs Rim north of Arock. 2 m thick flow lobe belonging to compound pahoehoe flow, shows the presence of segregations. Diktytaxitic with plagioclase and olivine phenocrysts.

Sample ID: OWY-12 Chemostratigraphic Group: 1 Easting: 441534 Northing: 4762229 Description: Lower of two flow units of diktytaxitic “sagebrush” flow, which underlies “cheat grass” flow – affinity to Saddle Butte or Owyhee Butte. Collected by Cooper Brossy.

Sample ID: OWY-15 Chemostratigraphic Group: 1 Easting: 444712 Northing: 4769326 Description: Grey, diktytaxitic flow of Bogus Grade, south of parking lot. Collected by Cooper Brossy.

Sample ID: OWY-21A Chemostratigraphic Group: 2 Easting: 458417 Northing: 4784246 Description: First intracanyon lava upstream of Birch Creek. Collected by Cooper Brossy.

240 Sample ID: OWY-21B Chemostratigraphic Group: 2 Easting: 457463 Northing: 4782980 Description: Third intracanyon lava upstream of Birch Creek. Collected by Cooper Brossy.

Sample ID: OWY-22 Chemostratigraphic Group: 2 Easting: 442379 Northing: 4771320 Description: Uppermost Lambert Rocks flow unit just upstream (on river) of Bogus Falls. Collected by Cooper Brossy.

Sample ID: OWY-23 Chemostratigraphic Group: 2 Easting: 442501 Northing: 4771642 Description: Basal flow unit at Bogus Falls. Collected by Cooper Brossy.

Sample ID: OWY-31 Chemostratigraphic Group: 2 Easting: 444361 Northing: 4768220 Description: Lambert Rocks lava at upstream-most point, near Bed Stream Gate and trail to river. Collected by Cooper Brossy.

Sample ID: OWY-35 Chemostratigraphic Group: 2 Easting: 442847 Northing: 4775644 Description: Upper lava outcrop at Whistling Birds Rapid; downstream of Hoot Owl Springs drainage. Collected by Cooper Brossy.

Sample ID: OWY-36 Chemostratigraphic Group: 2 Easting: 443022 Northing: 4775740 Description: Lower lava outcrop at Whistling Birds Rapid; downstream of Hoot Owl Springs drainage. Collected by Cooper Brossy.

Sample ID: N8/06 Chemostratigraphic Group: 2 Easting: ? Northing: ? Description: Intracanyon lava outcrop at Hole in the Ground, downstream of Ranch House; Collected by Jim O’Connor.

241 Sample ID: H8-24 Chemostratigraphic Group: 2 Easting: ~ 457833 Northing: ~ 4750428 Description: Sample slightly vesicular with glomerocyrsts of plagioclase and olivine. Although this flow appears to be derived from Rocky Butte, geochemical data and expression on satellite image suggest an affinity with Skinner Hill. Further description in Hart (1982).

Sample ID: H8-60A Chemostratigraphic Group: 2 Easting: ~ 478595 Northing: ~ 4759761 Description: Massive flow material near top of Skinner Hill, distinctly olivine and plagioclase phyric. Further description in Hart (1982).

Sample ID: H8-70 Chemostratigraphic Group: 1 Easting: ~ 465535 Northing: ~ 4764858 Description: Sample from Clarks Butte summit. Summit area gives an appearance of a large, collapsed shield, with a crater of approximately 50 m diameter with walls of welded pyroclastic material. Sample vesicular, with iddingsitized olivine phenocrysts and abundant laths of plagioclase. Further description in Hart (1982).

Sample ID: H8-71 Chemostratigraphic Group: 1 Easting: ? Northing: ? Description: Sample from Clarks Butte flow exposed between the summit and Dry Lakes Reservoir. Vesicular, similar in texture to H8-70. Further description in Hart (1980).

Sample ID: H9-42 Chemostratigraphic Group: 1 Easting: ~ 474761 Northing: ~ 4769142 Description: Eroded tumulus of flow belonging to the Cow Vent Complex. Somewhat diktytaxitic with clumps of olivine and plagioclase. Further description in Hart (1982).

Sample ID: H9-49 Chemostratigraphic Group: 1 Easting: ~ 467238 Northing: ~ 4755357 Description: Summit of Three Mile Hill – broad, eroded shield with scattered oxidized pyroclastic material. Sample is glomeroporphyritic with olivine and plagioclase phenocrysts.

References

242 APPENDIX 4: GEOCHEMICAL DATA FOR SAMPLES FROM THE JVVF

Sample JC-36A JC-4 H97-1A JC-5 JC-30B N1/03 N6/03 Vent‡ CC CC CC CC CC CC RB Map ID* 1 2 3 4 5 6 7 SiO2 47.58 47.53 47.47 48.34 48.43 48.29 47.18 TiO2 2.25 2.162.17 1.78 1.72 2.16 2.07 Al2O3 15.80 15.98 16.07 16.25 15.97 16.05 16.20 † t Fe2O3 11.69 11.81 11.50 10.46 10.33 11.29 11.93 MnO 0.16 0.17 0.16 0.16 0.15 0.17 0.18 MgO 8.76 9.29 9.18 8.62 9.13 8.00 7.86 CaO 9.90 9.939.31 9.98 9.85 9.28 9.33 Na2O 2.94 3.08 3.06 2.67 3.07 3.35 3.14 K2O 0.72 0.690.66 1.00 0.96 0.86 0.94 P2O5 0.35 0.290.44 0.41 0.43 0.42 0.41 LOI 0.62 0.15 -0.34 0.80 0.31 --- -0.51 Total 100.77 101.08 99.68 100.47 100.35 99.86 98.73 Rb 13.7 12.017.0 19.3 19.2 17.5 17.8 Sr 662 655608 486 528 600 521 Y 22 2319 23 21 23 25 Zr 120 117115 152 135 141 147 V 217 208213 231 215 201 220 Ni 148 156173 118 149 128 117 Cr 184 188175 248 264 163 166 Nb 19.1 17.9 --- 29.531.8 25.2 27.3 Ga 17.5 18.1 --- 16.516.9 18.9 18.4 Cu 61 6549 51 43 69 60 Zn 90 8494 77 79 83 93 Co 50 49 --- 7946 45 48 Ba 279 262246 238 234 267 277 U 0.60 0.35 --- 0.70<0.5 <0.5 <0.5 Th 0.8 --- 1.22.7 1.2 <0.5 Sc 26.0 27.9 --- 27.028.0 25.0 26.0 Pb 4 --- 34 3 4 La 12.0 11.9 --- 18.2 17.3 --- 19.0 Ce 28.1 27.0 --- 39.8 34.4 --- 38.0 Nd 18.1 17.4 --- 20.2 19.2 ------Sm 4.52 4.42 --- 4.54 4.18 ------Eu 1.64 1.63 --- 1.52 1.49 ------Gd 4.48 4.57 --- 4.82 4.42 ------Dy 3.89 3.84 --- 3.91 3.85 ------Er 2.07 1.87 --- 2.30 2.00 ------Yb 1.81 1.89 --- 2.20 1.90 ------Lu 0.28 0.28 --- 0.31 0.31 ------87Sr/86Sr 0.70383 0.70382 --- 0.70398 0.70408 ------143Nd/144Nd 0.512853 0.512850 --- 0.512804 0.512787 ------206Pb/204Pb 18.804 18.957 --- 18.838 18.830 ------207Pb/204Pb 15.563 15.578 --- 15.567 15.575 ------208Pb/204Pb 38.435 38.545 --- 38.561 38.547 ------Age (Ma) 0.002 0.002 0.002 0.002 0.002 0.002 ?

243 Sample N10/03 N11/03N14/03 N15/03 N17/03 N1/04 H8-58 Vent‡ RB RB RB RB RB RB RB Map ID* 8 9 10 11 12 13 14 SiO2 47.53 47.22 47.24 47.08 49.71 47.10 48.47 TiO2 1.93 1.962.01 2.01 2.11 2.09 2.23 Al2O3 16.64 15.55 15.79 15.53 16.34 16.69 16.09 † t Fe2O3 11.90 12.17 12.75 11.94 11.82 12.35 13.35 MnO 0.18 0.17 0.18 0.17 0.17 0.18 0.19 MgO 7.34 7.81 7.68 7.73 6.36 6.99 7.22 CaO 9.27 8.989.24 9.75 9.49 9.42 9.11 Na2O 3.17 3.12 3.02 3.01 3.31 3.20 3.19 K2O 1.27 0.961.06 0.97 1.16 0.96 1.00 P2O5 0.45 0.390.37 0.37 0.41 0.39 0.38 LOI -0.35 0.52-0.61 0.05 -0.53 -0.55 -0.43 Total 99.34 98.84 98.74 98.61 100.36 98.83 100.80 Rb 30.5 21.525.6 22.7 26.4 17.9 13.0 Sr 519 484428 478 496 500 435 Y 25 2527 24 27 25 26 Zr 153 143149 138 142 140 147 V 207 217229 225 230 232 225 Ni 97 124109 115 80 109 109 Cr 102 151120 164 155 168 160 Nb 30.8 29.027.0 24.6 25.2 26.0 --- Ga 18.5 18.018.8 17.1 18.6 18.4 --- Cu 55 4353 54 62 61 54 Zn 86 9299 96 97 95 100 Co 47 4652 50 42 50 --- Ba 423 231310 298 458 345 355 U <0.5 <0.5<0.5 1.10 <0.5 <0.5 --- Th 2.3 2.02.8 1.3 1.0 2.4 --- Sc 20.0 25.027.0 26.0 24.0 29.0 25.8 Pb 4 4 4 4 4 2 --- La 17.2 18.118.0 15.0 17.0 12.0 19.8 Ce 38.3 45.137.0 32.0 35.0 27.0 40.0 Nd 21.2 19.0 ------18.5 Sm 4.77 4.49 ------3.84 Eu 1.74 1.57 ------1.36 Gd 4.91 4.58 ------4.23 Dy 4.30 4.11 ------4.23 Er 2.46 2.11 ------2.59 Yb 2.20 2.21 ------2.30 Lu 0.33 0.29 ------0.36 87Sr/86Sr 0.70498 0.7045 ------0.70486 143Nd/144Nd 0.512691 0.512740 ------0.512699 206Pb/204Pb 18.668 18.862 ------18.806 207Pb/204Pb 15.591 15.586 ------15.632 208Pb/204Pb 38.643 38.661 ------38.880 Age (Ma) 0.03 max 0.03 max 0.03? 0.03? 0.086 0.03? 0.03?

244

Sample OWY-21A OWY-21B OWY-22 OWY-31 OWY-35 N8/06 JV99-1 Vent‡ RB/CB? RB/CB? RB/CB? RB/CB? RB/CB? RB/CB? WC Map ID* 15 16 17 18 19 20 21 SiO2 47.62 47.82 48.20 48.52 47.78 48.03 48.74 TiO2 2.00 2.011.94 1.95 1.99 1.99 1.62 Al2O3 16.28 16.64 16.19 16.34 16.43 16.25 16.63 † t Fe2O3 11.95 12.07 11.48 11.41 11.76 11.92 10.87 MnO 0.18 0.18 0.17 0.16 0.17 0.17 0.16 MgO 7.37 7.41 7.68 7.56 7.37 7.26 8.31 CaO 8.86 9.029.36 9.39 9.21 8.84 9.99 Na2O 3.16 3.23 3.22 3.19 3.21 3.13 2.86 K2O 1.39 1.431.02 1.04 1.41 1.45 1.00 P2O5 0.50 0.470.36 0.37 0.49 0.47 0.45 LOI --- -0.63 -0.52 -0.39 -0.32 -0.54 Total 99.30 99.65 99.62 99.42 99.43 99.17 100.09 Rb 32.2 --- 20.9 ------21.9 Sr 517 --- 530 ------446 Y 25 --- 23 ------21 Zr 171 --- 124 ------115 V 211 --- 200 ------230 Ni 93 --- 121 ------130 Cr 119 --- 192 ------270 Nb 36.2 --- 22.9 ------25.3 Ga 18.6 --- 18.7 ------16.7 Cu 38 --- 58 ------47 Zn 87 --- 86 ------87 Co 47 --- 45 ------48 Ba 494 --- 475 ------320 U 0.80 --- 0.70 ------<0.5 Th 2.5 --- 0.9 ------0.6 Sc 22.0 --- 24.0 ------31.0 Pb 1 --- 1 ------1 La 19.0 --- 13.0 ------13.6 Ce 35.0 --- 28.0 ------29.8 Nd ------16.8 Sm ------3.89 Eu ------1.44 Gd ------4.10 Dy ------3.68 Er ------2.11 Yb ------1.89 Lu ------0.27 87Sr/86Sr ------0.70470 143Nd/144Nd ------0.512699 206Pb/204Pb ------18.601 207Pb/204Pb ------15.585 208Pb/204Pb ------38.486 Age (Ma) 0.03? 0.03? 0.03? 0.03? 0.03? 0.03? 0.06

245

Sample JV99-2 N7/03 N13/03 OWY-23 OWY-36 H8-24 H8-60A Vent‡ WC WC WC WC? WC? SH? SH Map ID* 22 23 24 25 26 27 28 SiO2 48.79 47.74 48.36 48.45 47.50 47.87 48.02 TiO2 1.65 1.671.66 1.70 1.66 1.94 1.91 Al2O3 16.50 16.50 16.40 16.70 16.52 16.18 16.23 † t Fe2O3 10.98 11.08 11.11 11.21 10.94 12.05 11.21 MnO 0.17 0.17 0.17 0.16 0.16 0.17 0.17 MgO 8.19 7.55 8.54 8.20 7.81 8.44 8.22 CaO 9.97 10.15 10.10 9.98 10.31 9.46 9.65 Na2O 2.88 2.81 2.80 3.04 2.87 3.19 3.12 K2O 1.02 1.031.02 0.92 0.94 0.96 1.12 P2O5 0.48 0.340.36 0.31 0.34 0.36 0.40 LOI -0.23 -0.38-0.47 -0.35 -0.51 0.47 0.63 Total 100.40 98.66 100.06 100.31 98.53 101.09 100.68 Rb 30.0 21.9 24.7 16.6 --- 18.6 21.6 Sr 446 459 455 510 --- 515 542 Y 19 23 22 19 --- 23 21 Zr 124 127 130 101 --- 143 146 V 225 226 233 220 --- 206 205 Ni 130 91 120 131 --- 133 139 Cr 239 182 267 217 --- 143 202 Nb --- 24.9 26.4 21.0 --- 31.5 36.9 Ga --- 17.7 17.1 18.0 --- 17.7 19.4 Cu 49 68 63 46 --- 54 57 Zn 77 84 86 79 --- 83 95 Co --- 44 49 46 --- 48 45 Ba 360 330 324 340 --- 292 322 U --- <0.5 <0.5 1.60 --- 0.50 2.10 Th --- 1.8 1.5 2.1 --- 0.6 4.8 Sc 25.0 27.0 29.0 25.0 --- 24.0 26.7 Pb --- 3 4 1 --- 2 3 La --- 14.0 14.0 12.0 --- 17.0 18.0 Ce --- 31.0 30.0 26.0 --- 32.0 36.3 Nd ------18.8 Sm ------4.26 Eu ------1.50 Gd ------4.45 Dy ------3.89 Er ------1.96 Yb ------2.14 Lu ------0.29 87Sr/86Sr ------0.70434 0.70443 143Nd/144Nd ------0.512742 0.512710 206Pb/204Pb ------18.833 207Pb/204Pb ------15.606 208Pb/204Pb ------38.687 Age (Ma) 0.06 0.06 0.06 0.06 0.06 0.06? 0.06?

246

Sample H8-70 H8-71H9-42 N20/03 N21/03 N23/03 N24/03 Vent‡ CB CBCVC CVC CVCCVC CVC Map ID* 29 30 31 32 33 34 35 SiO2 48.79 49.02 46.77 47.61 47.68 47.99 47.57 TiO2 1.94 2.151.30 1.31 1.28 1.31 1.28 Al2O3 16.75 16.63 16.58 16.01 16.62 16.56 16.74 † t Fe2O3 11.53 11.71 11.46 11.46 11.06 11.14 11.10 MnO 0.17 0.17 0.17 0.19 0.18 0.18 0.18 MgO 6.69 5.74 8.68 8.61 8.71 8.64 8.77 CaO 8.82 8.0211.19 11.25 11.88 11.60 11.49 Na2O 3.08 3.19 2.55 2.45 2.47 2.55 2.51 K2O 1.79 2.080.26 0.24 0.19 0.22 0.20 P2O5 0.54 0.850.17 0.24 0.15 0.15 0.14 LOI 0.92 0.220.85 -0.14 0.28 -0.40 -0.39 Total 101.02 99.77 99.98 99.24 100.49 99.93 99.58 Rb 43.6 42.64.4 5.6 5.7 4.9 5.2 Sr 489 435242 252 245 247 236 Y 24 2923 24 20 23 21 Zr 199 23788 95 84 86 84 V 182 211254 259 267 269 267 Ni 83 73137 144 134 130 134 Cr 127 122240 340 262 252 248 Nb 43.6 50.67.3 8.5 7.6 7.8 6.5 Ga 17.9 18.318.7 17.4 17.7 17.9 18.6 Cu 51 3579 57 75 77 75 Zn 83 9175 79 78 84 76 Co 43 3950 51 50 51 52 Ba 594 585116 154 68 204 80 U 2.15 <0.5 --- <0.5<0.5 <0.5 <0.5 Th 3.3 1.3 --- 1.2<0.5 <0.5 <0.5 Sc 23.0 24.035.6 33.0 33.0 32.0 33.0 Pb --- 1 --- 4 4 3 4 La 26.2 31.08.2 8.0 6.0 8.3 4.0 Ce 51.3 52.021.2 17.0 12.0 13.5 8.0 Nd 23.7 ------10.6 --- Sm 4.96 ------2.93 --- Eu 1.63 ------1.07 --- Gd 4.98 ------3.49 --- Dy 4.37 ------3.60 --- Er 2.21 ------1.93 --- Yb 2.24 ------2.05 --- Lu 0.34 ------0.29 --- 87Sr/86Sr 0.70519 --- 0.70435 ------0.70446 --- 143Nd/144Nd 0.512644 ------0.512794 --- 206Pb/204Pb 18.672 ------18.854 --- 207Pb/204Pb 15.604 ------15.596 --- 208Pb/204Pb 38.723 ------38.809 --- Age (Ma) 0.25 0.25 0.44 0.44 0.44 0.44 0.44

247

Sample H9-49 N8/03 N9/03 N3/04 N8/04 OWY-12 N9/04 Vent‡ TMH OB OB OB? OB OB LOB Map ID* 36 37 38 39 40 41 42 SiO2 46.97 47.12 47.83 47.53 47.68 47.57 47.18 TiO2 1.92 1.020.97 1.09 1.03 0.90 1.65 Al2O3 15.89 16.38 16.43 17.37 17.15 16.54 16.57 † t Fe2O3 12.90 10.99 10.65 10.68 11.14 10.64 12.45 MnO 0.17 0.18 0.17 0.18 0.18 0.17 0.19 MgO 7.88 9.20 9.02 8.83 8.90 9.31 8.76 CaO 10.51 11.45 11.51 11.76 11.70 11.40 9.98 Na2O 2.63 2.44 2.36 2.61 2.43 2.50 2.70 K2O 0.72 0.200.21 0.25 0.16 0.21 0.45 P2O5 0.34 0.150.10 0.13 0.09 0.07 0.24 LOI 0.85 0.01-0.11 0.63 -0.14 -0.33 Total 100.78 99.13 99.15 101.05 100.31 99.31 99.83 Rb 15.0 3.6 6.1 6.2 5.0 3.7 12.8 Sr 305 213191 228 186 194 299 Y 27 1818 20 19 17 27 Zr 131 59 57 66 49 42 110 V 268 248230 251 248 249 225 Ni 124 169138 131 127 157 153 Cr 253 316243 232 247 280 247 Nb 16.0 6.0 5.9 8.3 6.0 4.9 12.6 Ga 18.3 16.517.2 16.7 17.3 16.3 20.8 Cu 60 8391 88 100 84 86 Zn 124 69 71 64 71 68 90 Co 60 5349 47 49 48 51 Ba 404 10798 124 107 65 286 U --- <0.5<0.5 <0.5 1.10 <0.5 1.50 Th 1.5 <0.5<0.5 0.5 <0.5 1.0 2.0 Sc 31.5 31.028.0 34.0 33.0 31.0 28.0 Pb --- 3 3 2 2 1 4 La 13.5 3.9 5.0 5.0 3.0 5.0 8.9 Ce 28.4 9.510.0 12.0 7.0 9.0 21.0 Nd 16.2 6.8 ------13.4 Sm 4.02 2.03 ------3.55 Eu 1.43 0.87 ------1.70 Gd 4.50 2.54 ------4.00 Dy 4.49 2.96 ------4.04 Er 2.30 1.86 ------2.40 Yb 2.38 1.79 ------2.23 Lu 0.34 0.26 ------0.34 87Sr/86Sr 0.70535 0.70527 ------0.70528 143Nd/144Nd 0.512617 0.512740 ------0.512620 206Pb/204Pb 18.757 18.842 ------18.800 207Pb/204Pb 15.630 15.621 ------15.646 208Pb/204Pb 38.887 38.910 ------39.051 Age (Ma) 1.86 1.86 1.86 1.86? 1.86? 1.86? 1.87

248

Sample N16/03 N6/04 N7/04 OWY-15 N3/03 N12/03 N18/03 Vent‡ BBV BBV BBV BBV/OB? BB? BB? BB Map ID* 43 44 45 46 47 48 49 SiO2 47.63 47.80 47.72 47.09 46.79 47.53 47.55 TiO2 0.96 0.960.97 0.91 1.80 1.67 1.74 Al2O3 17.07 17.06 17.43 16.89 15.90 15.87 15.61 † t Fe2O3 10.41 10.93 11.00 10.39 13.02 12.70 12.37 MnO 0.17 0.19 0.18 0.17 0.19 0.19 0.19 MgO 8.95 8.89 9.57 9.21 7.82 8.10 7.95 CaO 11.50 11.91 11.50 11.47 10.29 10.99 10.67 Na2O 2.47 2.46 2.45 2.36 2.68 2.44 2.54 K2O 0.24 0.120.16 0.22 0.41 0.37 0.59 P2O5 0.12 0.080.09 0.10 0.32 0.29 0.34 LOI -0.32 -0.31-0.42 0.28 -0.17 -0.38 -0.32 Total 99.20 100.08 100.63 99.08 99.03 99.79 99.24 Rb 5.4 3.93.6 4.3 6.0 7.8 13.8 Sr 215 178192 339 284 240 262 Y 19 2018 18 28 27 25 Zr 66 4450 53 119 117 113 V 221 257234 222 281 257 268 Ni 145 124156 147 150 104 113 Cr 259 259283 263 206 216 252 Nb 6.8 4.45.8 6.5 12.3 9.7 14.4 Ga 17.1 17.417.5 15.8 18.7 17.1 18.0 Cu 83 10478 89 63 63 74 Zn 63 7370 67 101 89 91 Co 51 5250 48 54 49 51 Ba 101 13293 86 177 324 218 U <0.5 0.900.60 <0.5 <0.5 0.80 <0.5 Th <0.5 0.7<0.5 1.3 1.2 1.6 <0.5 Sc 29.0 34.032.0 31.0 29.0 30.0 27.0 Pb 3 12255 5 La 4.0 3.43.0 7.0 13.8 15.0 10.6 Ce 9.0 7.88.0 16.0 27.9 30.0 24.4 Nd --- 6.2 ------14.8 --- 15.1 Sm --- 1.95 ------3.63 --- 3.85 Eu --- 0.85 ------1.33 --- 1.46 Gd --- 2.57 ------4.21 --- 4.39 Dy --- 3.07 ------4.45 --- 4.22 Er --- 1.99 ------2.76 --- 2.54 Yb --- 1.88 ------2.45 --- 2.31 Lu --- 0.28 ------0.38 --- 0.34 87Sr/86Sr --- 0.70534 ------0.70542 143Nd/144Nd --- 0.512762 ------0.512597 206Pb/204Pb --- 18.887 ------18.811 207Pb/204Pb --- 15.617 ------15.654 208Pb/204Pb --- 38.909 ------39.100 Age (Ma) 1.92 1.92 1.92 1.92? 2.78? 2.78? 2.78

249

Sample N21/04 N19A/03N19B/03 N2/04 N4/03 N5/03 N2/03 Vent‡ BB WCS WCS WCS ? ? V 4569 Map ID* 50 51 52 53 54 55 56 SiO2 47.66 47.60 48.04 46.72 49.77 48.51 46.61 TiO2 1.78 1.411.39 1.45 1.12 1.19 2.04 Al2O3 15.73 16.46 16.74 17.01 15.62 15.73 15.25 † t Fe2O3 12.35 11.71 11.56 11.94 10.59 11.08 13.70 MnO 0.20 0.18 0.18 0.18 0.18 0.18 0.19 MgO 7.71 8.69 8.75 8.88 8.47 8.42 8.13 CaO 10.88 10.96 11.18 10.78 10.18 10.58 10.01 Na2O 2.63 2.46 2.46 2.53 2.57 2.51 2.56 K2O 0.78 0.400.38 0.46 0.65 0.54 0.43 P2O5 0.44 0.230.22 0.26 0.17 0.19 0.47 LOI -0.25 -0.45-0.32 -0.39 -0.19 -0.15 -0.38 Total 99.90 99.65 100.57 99.83 99.14 98.77 99.02 Rb 13.9 9.7 8.7 8.2 13.5 10.3 9.2 Sr 260 235237 264 235 235 293 Y 27 2323 22 25 26 32 Zr 113 90 86 84 112 108 155 V 271 246238 218 235 249 256 Ni 93 142135 128 158 142 131 Cr 236 260266 229 324 286 204 Nb 14.8 10.710.5 12.2 8.6 9.3 17.2 Ga 19.2 17.117.8 17.8 16.5 17.1 19.9 Cu 75 7262 55 77 73 51 Zn 90 8980 81 98 82 112 Co 48 5553 52 50 50 57 Ba 269 148182 227 272 188 228 U <0.5 <0.5<0.5 0.70 <0.5 <0.5 <0.5 Th 2.4 0.8<0.5 <0.5 0.7 1.2 1.5 Sc 31.0 30.029.0 27.0 34.0 35.0 29.0 Pb 3 54374 7 La 10.0 8.0 9.0 10.0 11.0 9.0 17.5 Ce 23.0 18.020.0 21.6 26.0 19.0 38.4 Nd --- 10.7 --- 12.5 ------22.9 Sm --- 2.63 --- 3.26 ------5.48 Eu --- 0.99 --- 1.20 ------1.93 Gd --- 3.16 --- 3.57 ------6.01 Dy --- 3.57 --- 3.56 ------5.49 Er --- 2.29 --- 1.92 ------3.11 Yb --- 2.07 --- 2.05 ------2.76 Lu --- 0.32 --- 0.29 ------0.42 87Sr/86Sr ------0.70626 143Nd/144Nd ------0.512483 206Pb/204Pb ------19.745 207Pb/204Pb ------15.736 208Pb/204Pb 39.964 Age (Ma) 2.78 >2.78? >2.78? >2.78? ? ? 5.64

250 Sample N5/04 Vent‡ V 4569? Map ID* 57 SiO2 44.85 TiO2 2.19 Al2O3 14.27 † t Fe2O3 14.89 MnO 0.21 MgO 10.82 CaO 9.10 Na2O 2.40 K2O 0.56 P2O5 0.58 LOI -0.35 Total 99.53 Rb 13.1 Sr 275 Y 35 Zr 161 V 243 Ni 207 Cr 344 Nb 17.2 Ga 19.1 Cu 58 Zn 114 Co 62 Ba 410 U <0.5 Th 3.2 Sc 28.0 Pb 3 La 21.4 Ce --- Nd 25.1 Sm 5.95 Eu 1.92 Gd 6.27 Dy 5.75 Er 2.94 Yb 2.87 Lu 0.38 87Sr/86Sr --- 143Nd/144Nd --- 206Pb/204Pb --- 207Pb/204Pb --- 208Pb/204Pb --- Age (Ma) 5.64?

† Total iron as Fe2O3; * Map ID refers to the Figure 1; ‡ For vent abbreviations refer to Chapter 4, Figure 3

251 Figure 1

5 km 20 15 16

1, 2, 3 5 19, 26

56 6 47 27 17, 25 54 55

44 46 35 31 34 252 43 18 12 3332 53 7 2345 13 30 29 11 10 21&2224 41 48 39 8 9 53 28

40 50 14 49 37 38 36 51&52