University of South Florida Scholar Commons

Graduate Theses and Dissertations Graduate School

12-8-2009

Detailed Stratigraphy and Geochemistry of Lower Mount Rogers Formation Metavolcanic Units Exposed on Elk Garden Ridge, VA

Meghan Marie Lindsey University of South Florida

Follow this and additional works at: https://scholarcommons.usf.edu/etd

Part of the American Studies Commons

Scholar Commons Citation Lindsey, Meghan Marie, "Detailed Stratigraphy and Geochemistry of Lower Mount Rogers Formation Metavolcanic Units Exposed on Elk Garden Ridge, VA" (2009). Graduate Theses and Dissertations. https://scholarcommons.usf.edu/etd/1698

This Thesis is brought to you for free and open access by the Graduate School at Scholar Commons. It has been accepted for inclusion in Graduate Theses and Dissertations by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. Detailed Stratigraphy and Geochemistry of Lower Mount Rogers Formation Metavolcanic Units Exposed on Elk Garden Ridge, VA.

by

Meghan Marie Lindsey

A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Geology College of Arts and Sciences University of South Florida

Major Professor: Jeffrey Ryan, Ph.D. Diana Roman, Ph.D. Paul Wetmore, Ph.D.

Date of Approval: December 8, 2009

Keywords: failed continental rift, bimodal volcanic suite, rare earth elements, neoproterozoic, iapetus ocean © Copyright 2010, Meghan Marie Lindsey Acknowledgments

First, I would like to thank Michael Cook, Michael Lucas, and Mary Beck for their many hours of help with field work and their constant support during this project. Second, I would like to thank Zachary Atlas for his help in every step that it took to obtain the geochemical data. I would also like to thank my committee members: Jeff Ryan, Diana Roman, and Paul Wetmore for their willingness to share their knowledge, as well as their unwavering guidance. Table of Contents

List of Tables ii List of Figures iii Abstract v Introduction and Background 1 Introduction 1 Background 2 Regional Geological Setting and Related Units 3 Local Geological Setting 6 Stratigraphy and Petrography 10 Methods 13 Field Mapping 13 Petrography 17 Elemental Analysis 17

Results and Discussion 21 Geologic Map and Stratigraphy 21 Petrography 28 Geochemistry 33 LMRF Ignimbrite and UMRF Rhyolites Trace Element Systematics 36 LMRF Basalts Trace Element Systematics 39 Relationship Between LMRF Basalts and UMRF Rhyolites: Examples From Other Rift Settings 47 Conclusion 53

References 55 Appendices 63 Appendix 1: Direct Current Plasma Spectrometer (DCP) Geochemical Data 64

i List of Tables

Table 1. GPS locations of rock samples and their corresponding stratigraphic unit. 13 Table 2. Geochemical standard values used for calibration in elemental analysis. 14 Table 3. Measured elemental values for lower Mt. Rogers Formation samples. 16 Table 4. Normative results for the LMRF samples collected in this study. 29

ii List of Figures

Figure 1. Location map with field site outlined on a DEM of the entire Mount Rogers area. 3 Figure 2. Map showing the physiographic provinces of Virginia. 4 Figure 3. Map showing the location of Mount Rogers in relation to the Bakersville dikes, Grandfather Mountain Formation, Crossnore Complex plutons, and Catoctin Formation. 8 Figure 4A. Map showing the Holston Mountain-Iron Mountain fault and Stone Mountain fault in the Mt. Rogers area. The Elk Garden Ridge study area is outlined, as well as the triangle type area for the Mt. Rogers Formation. 9 Figure 4B. Key for Figure 4A. 10 Figure 5. Stratigraphic column of the newly defined LMRF units, plus the rhyolites of the UMRF. 22 Figure 6A. Detailed geologic map of the newly defined units in the LMRF exposed on Elk Garden Ridge overlain on a Digital Elevation Map (DEM). 23 Figure 6B. Legend for Figure 6A. 24 Figure 7. Left: A small-scale outcrop picture of the Swallowtail Plagioclase unit (Zmst). Right: Large-scale outcrop picture of the Swallowtail Plagioclase unit (Zmst). 25 Figure 8. A hand sample-scale picture of the Vesicular Basalt unit (Zmv). 25

Figure 9. Outcrop of the Sheet Flow Basalt unit (Zms) with interbedded light colored sedimentary layers. 26 Figure 10. Picture of a preserved pillow basalt outcrop in the Zmp unit. 26 Figure 11. Left: Hand sample-scale picture of the Ignimbrite unit (Zmig). Right: A large rounded clast seen in an outcrop of the Ignimbrite unit (Zmig). 27 Figure 12. Picture of a float block with a possible contact between a basalt and ignimbrite unit. 27

iii Figure 13. Hand sample-scale picture of the Fees Rhyolite unit (Zmf) which has been mapped in the LMRF (Rankin, 1993), but does not outcrop in the study area. 28 Figure 14. Photomicrograph of sample BQ2 (vesicular basalt unit). 30 Figure 15. Photomicrograph of sample Green1 (pillow basalt unit). 31 Figure 16. Photomicrograph of sample Pill1b (pillow basalt unit). 31

Figure 17. Photomicrograph of sample V1 (vesicular basalt unit). 32 Figure 18. Photomicrograph of sample S1 (sheet flow basalt unit). 32 Figure 19. Photomicrograph of sample IG1(ignimbrite unit). 33 Figure 20. Total Alkali concentration vs. Silica concentration (TAS) plot. 34

Figure 21. MnO and TiO2 vs. MgO for LMRF basalts. 35

Figure 22. Al2O3 and K2O vs. SiO2 for the LMRF ignimbrite and UMRF rhyolites. 36 Figure 23. REE abundances of three samples taken from the LMRF ignimbrite, plus the Cranberry gneiss basement. 38 Figure 24. REE abundance of the LMRF ignimbrite, Cranberry gneiss basement, and UMRF rhyolites. 38 Figure 25. REE abundances of the LMRF ignimbrite, Cranberry gneiss basement, and UMRF rhyolites with a field of granites from the plutonic Crossnore Complex. 39 Figure 26. Ti-Zr-Y tectonic discrimination diagram for the LMRF basalts. 40 Figure 27. LMRF basalts plotted on a olivine-silica-diopside ternary diagram. 40 Figure 28. REE abundances of the LMRF basalts with the MORB field. 41

Figure 29. REE abundances of the LMRF basalts and Catoctin Formation metabasalts. 42

Figure 30A. Top: La/Yb (normalized) vs. K2O and Bottom: La/Sm (normalized) vs. TiO2 for low-silica rift-related rocks in the Appalachians. 44

Figure 30B. La/Sm (normalized) vs. TiO2 for rift-related low silica rocks from the Appalachians, East African rift, and Rio Grande rift. 45 Figure 31. Depth of melting plot indicated by Sm/Yb vs. Ce/Yb. 46 Figure 32. La/Yb (normalized) vs. Sm/Yb (normalized) plot for the MRF. 48

iv Detailed Stratigraphy and Geochemistry of Lower Mount Rogers Formation Metavolcanic Units Exposed on Elk Garden Ridge, VA.

Meghan Lindsey

ABSTRACT The lower Mount Rogers Formation (LMRF) is described by Rankin (1993) as a sequence of intercalated metabasalts and volcanogenic sediments with minor metarhyolite. We have chosen to examine the sequence of the LMRF units exposed along Elk Garden Ridge, a high shoulder between the summits of Whitetop Mountain and Mount Rogers in the Mount Rogers National Recreation Area in SW Virginia. This sequence represents an uplifted block of LMRF units enclosed by exposures of Whitetop and Wilburn metarhyolites. In the field, progressive lithologic changes can be observed walking up-section along Elk Garden ridge that are indicative of changes in lava compositions and eruptive environments. From the bottom of the section, massive basalts with distinctive 1-2 cm long swallowtail plagioclase phenocrysts grade into vesicular basalts, then into sheet flow basalts, followed by a thick sequence of aphyric and amygdaloidal pillow basalts.

Further up section, eruptive products transition into rhyolitic ignimbrites and ash and lapilli tuffs. Boulders of cobble conglomerates near the middle of the sequence and sedimentary layers in between individual sheet flows suggest short periods of relative eruptive quiescence. The only unit broken out in the LMRF by Rankin (1993), Fees Rhyolite, is not observed in the field area, suggesting local differences in topography, eruptive products and eruptive styles across the outcrop area during the deposition of these eruptive products.

v Petrographically, the rocks reflect the regional greenschist facies metamorphic conditions with chlorite and epidote as primary metamorphic minerals, and unakite-like zones of mineralization. Relict plagioclase and pyroxene phenocrysts persist, as do primary igneous textures and structures. Compositionally, all of the rocks in the Elk

Garden Ridge sequence are strongly enriched in alkali metals, with elevated Na2O and

K2O contents, and high TiO2 in the basalts. Major and trace element systematics suggest that the chemical signatures of the metabasalts are primary controlled by shallow-level crystallization processes. The LMRF metabasalts share many compositional affinities with later (~570 Ma) rift-related basalts preserved in the Appalachians, suggesting that all of these lavas were formed by melting of a compositionally uniform mantle source, followed by shallow crystallization, despite being separated from one another by large stretches of time and space.

vi Introduction and Background

Introduction The Mount Rogers area in southwestern Virginia has been studied by numerous workers over the last 40 years to help determine the evolution of the rifted eastern margin of Laurentia. In this study, focus on examining in detail the types of volcanic events and products that occur in the lower Mount Rogers Formation (LMRF) to reconstruct its volcanic and petrogenetic history. We chose to concentrate our study area on Elk Garden Ridge (Figure 1) because this is one of the few locations in the LMRF outcrop area where a continuous section is relatively well-exposed, permitting detailed mapping. Past workers (Badger and Sinha, 1988; Rankin, 1993; Aleinikoff et al., 1995; Walsh and Aleinikoff, 1999; Cawood et al., 2001; Tollo et al., 2004) have suggested that the volcanism preserved in the Mount Rogers Formation (MRF) is the result of an aborted rifting event that happened around 760 Ma. The MRF was produced during the first magmatic pulse of a two-phase rifting event that led to the opening of the Iapetus Ocean. The second pulse of rifting, which followed ~180 Ma later, produced extensive flood volcanism exposed in the Catoctin volcanic province of the Shenandoahs, and several other units exposed along the margins of the Blue Ridge. In this study, we make use of detailed stratigraphic and lithologic information, and geochemical analyses of the metavolcanic samples to (1) determine the environments of deposition seen in the LMRF exposed in the field area and relate any changes in deposition to tectonic environment, (2) conclude if the low grade of metamorphism in the Elk Garden Ridge field area preserves primary volcanic features, (3) find any connections between the MRF and the later rift- related volcanic episodes in the region, and lastly (4) find any connections between the

1 MRF and modern day continental rifting.

Background Mount Rogers, named after Virginia's first state geologist, W.B. Rogers, is the highest mountain in Virginia at 5,729 feet elevation. Stose and Stose (1944) carried out the first major study on the rocks found in the Mount Rogers area and named them the Mount Rogers Series. The name was later changed by Rodgers (1953) to the Mount Rogers Volcanic Group and then again to the Mount Rogers Formation by Rankin (1970). The geology of the Mount Rogers area is typified by rift-related intercalated metavolcanic and metasedimentary rocks. In the following text, the prefix “meta” is omitted, but should be assumed for these rocks. The volcanic rocks found in the Mount Rogers area are bimodal (i.e., both basalts and rhyolites occur, with no rocks of intermediate composition) and have undergone greenschist facies metamorphism. The age of the Mount Rogers volcanics was initially estimated based on U-Pb zircon dating of the UMRF rhyolites to 820 Ma (Rankin et al., 1969), but more recent measurements on the UMRF rhyolites, specifically the Whitetop Rhyolite member, by Aleinikoff et al., (1995) yield dates of 758 ± 12 Ma. These results are comparable to dates for other suites of rocks associated with the failed first rifting event, for example the Bakersville metagabbro suite exposed to the south (Goldberg et al., 1986), but are younger than most reported ages for nearby granitoids of the Crossnore Plutonic Sequence (Su et al., 1994)

(Figure 3). Rankin (1993) distinguished a sequence of glaciogenic sediments overlying MRF volcanics as the Konnarock Formation, and divided the remaining MRF units into a lower and upper sequence (UMRF). The lower sequence consists of intercalated sedimentary and volcanic rocks, and the upper section contains mostly high silica volcanics. Based on the geologic map of Rankin (1989), the lower section of the MRF mostly outcrops south of the Mount Rogers/Grayson Highlands area, but is also shown to 2 outcrop on Elk Garden Ridge, a shoulder between the summits of Whitetop Mountain and Mount Rogers, in the Stone Mountain thrust sheet. Rankin only briefly describes the rocks found in the lower section of the MRF, and breaks out only one major rhyolite unit, the Fees Rhyolite, which outcrops in and near Grayson Highlands State Park. The purpose of this study is to provide more detail on the stratigraphy and geochemistry of the LMRF to aid in the investigation of the geologic history of the rocks found in the Mount Rogers area.

Figure 1. Location map with the Elk Garden Ridge field site outlined on a DEM of the entire Mount Rogers area.

Regional Setting and Related Units The study area is located in southwestern Virginia in the Blue Ridge Province, within the Appalachian Mountains (Figure 2). The Blue Ridge is an anticlinorium that exposes billion year old crystalline basement rocks. The Blue Ridge extends from southern Pennsylvania to northern Georgia and is typified by Precambrian crystalline gneiss/gneissic granites overlain by deformed and metamorphosed sedimentary and volcanic rocks (Bryant and Reed, 1970). The Blue Ridge rock sequences are overlain on 3 the northwest side by Early Cambrian sandstones and conglomerates of the Chilhowee Group, and the Precambrian metavolcanic and metasedimentary rocks of the Mount Rogers and Catoctin Formations. The southeast side of the Blue Ridge sequences are overlain by Precambrian metamorphosed sedimentary and volcanic rocks from the Ashe, Lynchburg, and Catoctin Formations (Rankin, 1975). The entire Blue Ridge has been subjected to regional low-grade metamorphism associated with the Appalachian/Alleghenian collisional event (approx. 320-300 Ma).

Figure 2. Map showing the physiographic provinces of Virginia. Modified from Fenneman and Johnson, 1946.

Rankin (1976) noted that, in the Appalachians, mafic volcanic rocks are common in the Neoproterozoic cover sequence, volcanic rocks of intermediate composition are sparse, and felsic volcanic rocks are largely restricted to sites of major deflection in Appalachian structural trends. Rankin (1976) attributed this geographic distribution of rock units to hot spots under triple junctions of intersecting continental rifts. A more recent model involves formation of these volcanic units during two pulses of rifting separated by >100 m.y., with spatial overlaps in rift-related units in North Carolina and southern Virginia (Badger and Sinha, 1988; Rankin, 1993; Aleinikoff et al., 1995; Walsh and Aleinikoff, 1999; Cawood et al., 2001; Tollo et al., 2004). The first pulse of rifting, between 765-680 Ma, has been described as a failed rift (aulocogen). The rocks produced during this first rifting event are represented by the MRF, units of the Grandfather Mountain Formation, the Bakersville metagabbro dikes, as well as abundant 4 silicic plutonism of the Crossnore plutonic suite (Figure 3). The second pulse, around 577-550 Ma is represented by the Catoctin Formation (Figure 3), the Camel’s Hump Group in New England, and rocks in New Brunswick and Nova Scotia, and is associated with the rifting that opened the Iapetus ocean. The observation by Rankin (1976) that mafic volcanic rocks are common in the Neoproterozoic cover sequence largely relates to the younger pulse of rifting (Badger and Sinha, 1988; Rankin, 1993; Aleinikoff et al., 1995; Walsh and Aleinikoff, 1999; Cawood et al., 2001; Tollo et al., 2004). The MRF is part of the Crossnore Complex (formerly the Crossnore Plutonic- Volcanic Complex of Rankin, 1970) along with the Grandfather Mountain Formation (742 ± 2 Ma), the Bakersville metagabbro dike swarm (730 Ma), the Catoctin Formation (564 ± 9 Ma) and a suite of plutons and dikes (702 ± 2 Ma) that intrude the Grenville (~1.1-1.3 Ga: Fullagar and Odom, 1973) basement in the French Broad massif (Rankin, 1993; Fetter and Goldberg, 1995; Goldberg et al., 1986; Tollo and Hutson, 1996: Aleinikoff et al., 1995). The Crossnore Complex is an assemblage of granite and peralkaline granite, including the Beech pluton and felsic and mafic volcanics from the Catoctin, Mount Rogers, and Grandfather Mountain Formations. The Grandfather Mountain Formation is exposed in the Grandfather Mountain Window, NC, and partially preserves the stratigraphic record of a rift basin. The stratigraphy is truncated, as a result of the overthrusting of the composite Blue Ridge thrust sheet (Bryant and Reed 1970;

Boyer 1984). The Grandfather Mountain Formation is a 5000-9000 m thick sequence of metaarkose, metasiltstone, and metaconglomerate with interbedded bimodal metavolcanic flows and tuffaceous rocks, cut by mafic dikes and sills of Linville metadiabase (Bryant and Reed, 1970). Major faulting and uplift along the basin margin during rifting are interpreted as leading to the deposition of the rock sequences seen in the Grandfather Mountain Formation (Neton, 1992). The Bakersville metagabbro is an 8 km wide, 200 km long dike swarm of predominantly tholeiitic, variably amphibolitized

5 metagabbros, intruding Precambrian metamorphic rocks of the Blue Ridge terrane. The swarm is exposed from Roan Mountain, TN to Yancey County, NC. (Wilcox and Poldervaart, 1958). The Catoctin volcanic province is an approximately 750 m thick suite of bimodal metavolcanic rocks that extends from near Roanoke, VA to South Mountain, PA (Allen, 1963). The Catoctin Formation is composed mostly of tholeiitic flood metabasalts/greenstones with interbedded clastic metasedimentary rocks (Reed and Morgan, 1971; Badger and Sinha, 1988). Unlike the MRF, the Catoctin Formation outcrops on both the northwest and southeast limbs of the Blue Ridge anticlinorium. On the northwest limb, subaerial eruptive conditions are inferred based on the evidence of columnar jointed lava flows and breccias (Badger, 1986; Espenshade, 1986). On the southeast limb rare pillow metabasalts have been observed by Badger (1986), indicating subaqueous eruptions.

Local Geological Setting The MRF crops out on the northwest side of the allochthonous French Broad massif (Figure 3) that carries Grenville-age rocks and is composed of ~70% metavolcanic rocks and ~30% clastic metasedimentary rocks. The MRF noncomformably overlies the Grenville-age Cranberry gneiss and is paraconformably overlain by the Late Neoproterozoic Konnarock Formation (Rankin, 1989). The Konnarock Formation is paraconformably overlain by the Late Neoproterozoic-Early Cambrian Unicoi Formation of the Chilhowee Group. Together, the Konnarock Formation, Chilhowee Group, and the overlying Cambrian Shady Dolomite are believed to record the transition from continental rifting to the opening of the Iapetus Ocean (Rankin, 1993; Novak and Rankin, 2004). The MRF is exposed within a series of thrust sheets on the northwestern side of the composite Blue Ridge crystalline thrust. The Blue Ridge thrust sheet experienced its last movement in the late Paleozoic during the Alleghanian/Appalachian orogenic event 6 (approx. 320-300 Ma). The sole thrust in the Mount Rogers region is the Holston Mountain-Iron Mountain fault, which bounds the Shady Valley thrust sheet (Figure 4A) (Rankin, Drake, and Ratcliffe, 1989). The Elk Garden Ridge study area is located within the Stone Mountain thrust sheet, where a semi-continuous section of the MRF is exposed, from the non-conformable contact with the Cranberry gneiss through the second youngest known unit of the UMRF, the Whitetop Rhyolite. The type area for the MRF is defined by a triangle with the three corners at Whitetop Mountain, Pine Mountain, and Nella, N.C. (Figure 4A). The thickness of the MRF has been estimated at about 3,000 m in the type area, but this is only a rough estimate given the structural and stratigraphic complexities in the lower part of the formation (Rankin, 1993).

7 Figure 3. Map showing the location of Mount Rogers in relation to the Bakersville dikes, Grandfather Mountain Formation, Crossnore Complex plutons, and Catoctin Formation. Modified from Tollo et al., 2004.

8 Figure 4A. Map showing the Holston Mountain-Iron Mountain fault and Stone Mountain fault in the Mt. Rogers area. The Elk Garden Ridge study area is outlined, as well as the triangle type area for the Mt. Rogers Formation. See Figure 4B for key. Map modified from Figure 3A in Rankin, 1993.

9 Figure 4B. Key for Figure 4A. Key modified from Figure 3B in Rankin, 1993.

Stratigraphy and Petrography

Rhyolite constitutes about 50-60% of the 70% bimodal volcanic rocks of the MRF and are thought to be localized at the sites of three major volcanic edifices that each erupted approximately 500-1000 km3 of material. The three volcanic edifices were named by Rankin (1993) and are, from east to west, the Razor Ridge, Mount Rogers, and Pond Mountain volcanic centers. Rocks of the Razor Ridge and Pond Mountain centers are mapped as LMRF by Rankin (1989; 1993). The Mount Rogers volcanic center, consisting predominantly of rocks of the UMRF, is built upon intermittent occurrences of basalt/greenstone, rhyolite and sedimentary strata of the LMRF. All of the stratigraphic 10 units in the MRF are in unconformable contact with ~1.0 Ga basement gneisses, indicating relatively rugged topography locally when these volcanic and sedimentary units were deposited (Rankin 1993; Novak and Rankin, 2004). The UMRF, as described by Rankin (1993), is comprised of three different metaluminous rhyolite members: the Buzzard Rock Rhyolite, the Whitetop Rhyolite, and the Wilburn Rhyolite. The Buzzard Rock member, which consists of low-silica rhyolite lava flows, is characterized by an aphanitic, maroon-colored groundmass with 5-20% reddish perthitic alkali feldspar (perthite) and greenish-white plagioclase phenocrysts in approximately equal amounts. The Whitetop Rhyolite Member, which consists mostly of phenocryst-poor high-silica rhyolite lava flows, with intercalated lapilli and lithophysae- rich tuffs, is characterized by an aphanitic maroon-colored groundmass with 0-10% quartz and perthite phenocrysts in varying amounts. The Wilburn Rhyolite Member, which consists of high-silica welded ash-flow tuff, is characterized by an aphanitic, maroon-colored to purple-colored groundmass. Phenocrysts are abundant, comprising up to 30% of the rock, with most exposures showing abundant quartz and perthite phenocrysts + minor biotite, with perthite more abundant near the top of the unit and approximately equal amounts of quartz and perthite at the bottom. Early erupted tuffs include riebeckite and aegirine phenocrysts, documenting an initial peralkaline character (Novak and Rankin, 2004). Each of the three rhyolite members include clasts (in the case of the lava flows, rip-up clasts) of the underlying rhyolites and of the Cranberry gneiss basement rocks. The LMRF was described by Rankin (1993) as consisting of intercalated clastic sedimentary and volcanic rocks; i.e., graywacke, phyllite, conglomerate, arkose, greenstone, and rhyolite. The volcanics and clastic sedimentary rocks have about a 1:2 ratio of abundance, making the section mostly clastic sedimentary rocks. The clasts in the conglomerates are mostly porphyritic felsic volcanic rocks and granitic gneisses

11 consistent with the Cranberry gneiss basement. In some areas the volcanic clasts appear more deformed than the granitic clasts, which leads to the idea that the volcanic clasts may have been pumiceous. LMRF greenstone ranges from aphyric to plagioclase-phyric, and some contain amygdules. The mineralogy of the greenstones is consistent with greenschist facies metamorphism; i.e., epidote-albite-chlorite assemblages dominate. The Fees Rhyolite Member is the only LMRF unit broken out by Rankin (1993). The Fees Rhyolite, which is believed by James (1999) to represent a pyroclastic flow, is characterized by an aphanitic, dark maroon-colored to purple-colored groundmass with about 35% quartz, perthite, and plagioclase phenocrysts + trace pyrite. Some of the perthite crystals reach 5 mm across. Rankin (1993) noted that the successful second pulse of rifting signified by the Catoctin Formation is recorded in the MRF via greenstone and rhyolite dikes, and granophyre. These dikes have been observed to intrude the MRF as young as the Wilburn Rhyolite.

12 Methods

Field Mapping Detailed field mapping was carried out during three week-long visits to the Elk Garden Ridge area of the Mount Rogers National Recreation Area, to define the volcanic units of the LMRF at a fine scale (i.e., at a lava flow or eruptive event), to obtain more detailed information on the stratigraphic relationships and unit orientations identified by Rankin (1993) and presented in the Winston Salem Quadrangle geologic map, and to identify original volcanic features and structures in all volcanic units. GPS measurements were made to establish unit contact positions and sample locations (Table 1). Rock samples were collected from each recognized unit for petrographic study, and for measurements of major element, lithophile trace element, and rare earth element (REE) abundances (Table 2; Table 3).

GPS Point Stratigraphic Unit Sample Name -81.564500 36.657983 Swallowtail Basalt Stailbasalt -81.569883 36.652783 Vesicular Basalt BQ2 -81.574717 36.648033 Vesicular Basalt V1 -81.574783 36.655567 Sheet Flow Basalt Sheet1 -81.579883 36.646117 Sheet Flow Basalt Sheet2 -81.577750 36.657800 Pillow Basalt Green1 -81.576967 36.657117 Pillow Basalt Pill1b -81.577917 36.662517 Ignimbrite IG1 -81.577167 36.663567 Ignimbrite IG2 -81.579867 36.652833 Ignimbrite IG3 Table 1. GPS locations of samples and their corresponding stratigraphic unit.

13 Table 2. Standard values used for calibration. Oxides in weight percent, all other elements in ppm. Values from United States Geological Survey, Geological Survey of Japan, and Centre de Recherches Pétrographiques et Géochimiques. Chondrite values from Haskins et al., 1968. 14 Table 2 cont'd. Values for standards run as unknown.

15 Table 3. Measured values for LMRF samples and the Cranberry gneiss basement. Oxides in weight percent, all other elements in ppm. Zr* measured by ICP-OES.

16 Petrography Thin sections were cut from each sample and analyzed via transmitted light polarizing microscope to identify primary and relict mineral assemblages. Thin section photomicrographs were taken on a Nikon Eclipse LV100POL microscope and computer imaging system in the University of South Florida (USF) Geology Department.

Elemental Analysis Major element compositions were measured by plasma emission spectrometry, using a Perkin Elmer Inductively Coupled Plasma Optical Emission Spectrometer (ICP- OES) in the USF Geology Department. The instrument utilizes an Echelle monochromator and employs a CCD detection system to permit rapid resolution of optical interferences. Fresh interiors of each rock sample were cut, and their surfaces were ground on a diamond wheel to remove any contamination from the sawblade. Samples were crushed in a steel jaw crusher, and were powdered by ball mill using alumina vials to minimize trace element contamination. Digestions for major and minor elements followed classical flux fusion procedures using lithium metaborate (LiBO2), preceded by a sample ignition step. 2.000 grams of each sample were weighed into quartz crucibles and heated at

1050oC for 15 minutes in a muffle furnace, cooled in a dessicator and then reweighed to determine Loss on Ignition as a semi-quantitative measure of sample volatiles. 0.2000 ±0.0002 grams of this dessicated sample were weighed with 0.8000 ±0.0008 grams of flux into a graphite crucible, mixed thoroughly, and heated in a muffle furnace at

~1100oC for 15 minutes. The hot fusion beads were poured into 125 ml sample bottles containing 50 ml of 2 M HNO3 + 10 ppm Ge internal standard. A 2.5 ml aliquot of each sample solution was then pipetted and weighed into a 60 ml sample bottle and diluted to

50 ml in 2 M HNO3 + 1000 ppm Li + 10 ppm Ge solution. 17 The ICP-OES instrument operated at an RF power of 1300 watts, with a nebulizer flow of 0.8 ml/minute, auxiliary gas flow of 0.2 ml/minute, and plasma gas flow of 15 ml/min. The optics were optimized on Mn in a 1 ppb standard solution. Any element in high concentration was viewed radially, and any element in low concentration was viewed axially. To maximize the signal level, a known amount of an Li internal standard was added to each sample. The REE and high-abundance lithophile trace element concentrations were collected using the Perkin Elmer PQ Elan quadrupole Inductively Coupled Plasma Mass Spectrometer (ICP-MS) system housed in the USF Geology Department. The ICP-MS was run at an RF power of 1100 watts in normal (non-reaction cell) mode, and optimized for nebulizer gas flow and autolens parameters using standard routines developed by

Perkin-Elmer that take into account the intensities of Mg, In, and U. CeO2/Ce ratios were tracked through instrument tuning to minimize oxides, and Ba ratios were tracked to minimize double-charging. Samples were digested for ICP-MS analysis via a different fluxed fusion digestion procedure. In this case, a sodium carbonate (Na2CO3) flux was utilized to ensure the digestion of zircons and other resistant minor minerals, while at the same time keeping sample total dissolved solids (TDS) low. 0.2000 ±0.0002 grams of each sample was weighed into a Pt crucible along with 1.0000 ±0.0001 grams of Na2CO3 flux. The sample and flux were mixed in the crucible to achieve a homogenous mixture. The crucibles were covered to prevent spattering, and placed in a muffle furnace for a >60 minute period, during which time temperatures were raised slowly to 1050oC. The crucibles were cooled to room temperature in the furnace, and then immersed in approximately 80 ml of deionized (DI) water in 180 ml teflon jars to dissolve overnight. Sample solids were physically scraped and rinsed from the crucibles, and the resulting solute-residue mixtures were centrifuged and rinsed three to four times with DI water to remove all

18 supernate and concentrate the solid carbonate residues. These residues were dissolved in

~7M HNO3 (made by mixing equal portions of Teflon distilled concentrated HNO3 and

DI water) until dissolved. Remaining solids in the crucibles were dissolved in 2 to 3 ml of ~7 M HNO3 and added back to their respective sample solutions. The samples were

o evaporated to dryness at 150 C, and were then resuspended in ~5ml of 2% HNO3. This drydown/resuspension step after digestion in nitric acid encourages the precipitation of any remaining dissolved silica, which can clog the ICP-MS nebulizers. Samples were then centrifuged and rinsed three to four times with 2% HNO3. Supernates were pipetted off of any gelatinous silica solids, placed into 250 ml volumetric flasks, and diluted to volume in 2% HNO3. Before ICP-MS analysis, an aliquot of each sample was spiked with an internal standard solution of 10 ppb Indium and Germanium in 2% HNO3 to optimize instrument performance. We analyzed the major, trace, and rare earth elements of all of the samples, with overlap in the ICP-OES and ICP-MS data for Zr. We chose to neglect all Group 1A metals measured by ICP-MS because these appear to be removed during the digestion process. Among the other elements measured by ICP-MS, Be, Ga, Sr, Y, Zr, Nb, Ba, Ce, Pr, Dy, Er, Tb, Yb, Lu, Hf, Pb, Th, and U occasionally showed high percent errors (>5%), while errors on Ta were so high that they are not reported here. The reference materials BHVO-1, JA-1, JB-3, NBS-688, BIR-1, JB-2, JR-1, G-2, JA-1, JG-2 and JG-3 were used as calibration standards for our ICP-MS measurements, while BHVO-1, BIR-1, JB-3, JB-2, JA-1, JG-2, JG-3, JR-1, GS-N, and JA-1 were used as standards for ICP-OES measurements. Subsets of these standards were used in each analytical run to create working curves for each element, which were used to determine concentrations. Compositionally matched standards were used for rhyolite and basalt measurements, respectively

19 Analytical precision was determined through in-run checks of standards during each run, and replicate sample analyses. GSN, JA-1, and JB-3 were used as run checks for the ICP-MS runs and G-2 and NBS-688 were used as run checks for the ICP-OES runs. Precision for the ICP-MS trace element runs are less than 20% for all of the elements, except for Y (41.17%). Precision for the ICP-OES major element runs are better than 5% for all of the elements, except for P2O5 (8.41%) and Na2O (8.83%).

20 Results and Discussion

Geologic Map and Stratigraphy Based on field mapping and GPS control points, a detailed surficial geology map of the LMRF exposed on Elk Garden Ridge was developed, working from Figure 3A in Rankin, 1993 as a base map (Figure 4A; Figure 6A). The LMRF on Elk Garden Ridge is bounded by the Stone Mountain Fault on the west, south and east sides, which places it in contact with Whitetop Rhyolite to the west and the Wilburn Rhyolite on the east. A stratigraphic contact between the LMRF and Buzzard Rock Rhyolite lies to the north of Elk Garden Ridge. Five discrete volcanic units were distinguished and named in the LMRF in this study based on lithologic changes observed in the field. Each unit was named beginning with a “Zm” to maintain continuity with the convention used by Rankin (1993) in identifying these Precambrian age units. Starting at the bottom of the stratigraphic section (Figure 5), massive basalts with distinctive 1-2 cm swallow-tail plagioclase phenocrysts (Swallowtail Plagioclase Basalt, Zmst) (Figure 7) grade into vesicular basalts (Vesicular Basalt, Zmv) (Figure 8). The vesicular basalts are overlain by sheet flow basalts (Sheet Flow Basalt, Zms) (Figure 9), and then by a thick (approx. 480 m) sequence of aphyric and amygdaloidal pillow basalts (Pillow Basalt, Zmp) (Figure 10).

21 Figure 5. Stratigraphic column of the newly defined LMRF units, plus the rhyolites of the UMRF. The thickness of the LMRF units were calculated assuming no elevation change and a uniform dip. The thickness of the UMRF rhyolites are from Rankin, 1993.

At the top of the stratigraphic section (Figure 5), eruptive products transition into rhyolitic ignimbrites (Ignmibrite, Zmig) (Figure 11). It appears that the ignimbrite is in stratigraphic contact with the basalts based on a rare float block with ignimbrite and basalt in contact (Figure 12). It is possible that the contact is not continuous throughout the section, but a lack of outcrop prevents a conclusive determination. The Fees Rhyolite unit (Figure 13) separated out of the LMRF sequence by Rankin (1993) is not observed in the section exposed on Elk Garden Ridge, suggesting local differences in topography, 22 eruptive products and eruptive styles across the MRF field area during the evolution of this volcanic field (e.g. James, 1999).

Figure 6A. Detailed geologic map of the newly defined units in the LMRF exposed on Elk Garden Ridge overlain on a Digital Elevation Map (DEM). The two overturned orientations, and Zmw, Zmwt, Zmb, Yc are from Rankin, 1993.

23 Figure 6B. Legend for Figure 6A.

Rare outcrops of cobble conglomerates near the middle of the sequence and sedimentary layers between individual sheet flows suggest short periods of relative quiescence between emplacement of the units. During field mapping, a relative abundance of volcanic rock (80%) to sedimentary rock (20%) was observed in the Elk Garden Ridge area, in conflict with Rankin's inferences and observations of a high overall proportion of sedimentary rocks relative to volcanic rocks in the LMRF overall. To constrain the orientation of the section, structural data was collected where outcrops permitted. Four bedding orientations, five joint orientations, and one sediment layer orientation were recorded (Figure 6A). Rankin, (1993) recorded only two bedding orientations in the Elk Garden Ridge study area, and mapped these beds as overturned (Figure 6A). His orientations were inferred based on the stratigraphy, and not on primary sedimentary features, and in a different area west of Elk Garden Ridge the bedding appears not to be overturned (D. W. Rankin, pers comm. 2009). In this study, no overturned beds were observed, and using the shape of pillow basalt unit, we believe the section to be upright, with basalts transitioning from subaerial to distinctively subaqueous 24 eruptive styles up section. The transition from subaerial to subaqueous eruptive styles implies a deepening rift basin partly filled with volcanic products.

Figure 7. Left: A small-scale outcrop picture of the Swallowtail Plagioclase unit (Zmst). Right: Large-scale outcrop picture of the Swallowtail Plagioclase unit (Zmst). Epidotes can be seen weathering out of the outcrop.

Figure 8. A hand sample-scale picture of the Vesicular Basalt unit (Zmv). Most of the vesicles have been filled with calcite through hydrothermal activity (Rankin, 1993).

25 Figure 9. Outcrop of the Sheet Flow Basalt unit (Zms) with interbedded light colored sedimentary layers.

Figure 10. Picture of a preserved pillow basalt outcrop in the Zmp unit.

26 Figure 11. Left: Hand sample-scale picture of the Ignimbrite unit (Zmig). Right: A large rounded clast seen in an outcrop of the Ignimbrite unit (Zmig).

Figure 12. Picture of a float block with a possible contact between a basalt and ignimbrite unit. A thin sedimentary layer may be present between the contact.

27 Figure 13. Hand sample-scale picture of the Fees Rhyolite unit (Zmf) which has been mapped in the LMRF (Rankin, 1993), but does not outcrop in the study area.

Petrography Despite evidence for pervasive greenschist facies metamorphism, LMRF basalts preserve relict phenocrysts and textural relationships indicative of their igneous protoliths. The rocks found in the study area have a mineral assemblage that reflects the metamorphic conditions (albite-epidote-actinolite-chlorite) with chlorite and epidote as porphyroblasts. Most of the time the metamorphic mineral assemblage is expressed in the amygdules and in fracture fills. Unakite-type (orthoclase-epidote-quartz intergrowth) zones are seen in fractured or fragmented areas in the metabasalts. A normative calculation with the elemental data collected for each LMRF sample demonstrates the assemblage of primary minerals expected (Table 4).

28 Table 4. Normative results for the LMRF samples collected in this study. Values in weight percent.

From unit to unit, the metamorphic mineral assemblages do not vary much, with the exception of the appearance of serpentine in the vesicular basalt unit. Relict igneous textures such as the variolitic texture are well-preserved in the pillow basalt unit, as well as primary flow features with preferred orientations of minerals. Igneous textures are best preserved in the pillow basalt unit, perhaps because this unit shows less fracturing and layer disruption than the lower units in the basalt sequence, or because the quench rims of the pillows are more effective at preserving primary textures. The swallowtail plagioclase basalt unit contains approximately 10-20% groundmass with 30-40% plagioclase, 20-30% chlorite, 20-25% epidote, and 1% oxide crystals. The vesicular basalt unit contains approximately 10-15% ground mass with 20- 50% feldspar, 15-20% serpentine, 10-15% chlorite, 20-70% pyroxene, 1-5% oxide crystals, and 5-10% vesicles. The sheet flow basalt unit contains approximately 20-25% groundmass with 40-50% feldspar, 20-30% chlorite, 10-15% epidote, 10-15% pyroxene, and 5-10% vesicles. The pillow basalt unit contains approximately 5-10% groundmass with 60-80% feldspar, 10-15% epidote, 5-20% chlorite, and 10-20% vesicles. The ignimbrite unit has approximately 5-10% groundmass with 15-20% quartz, 10-15%

29 feldspar, 5-10% calcite, 1-5% amphibole, 15-20% diopside, 10-15% plagioclase, 15-20% epidote, and 1-2% Zircon. Figures 14-18 are photomicrographs that depict examples of the textures and relict/secondary phenocryst minerals that occur in the different LMRF metabasalt types. Figure 19 is a photomicrograph of the metaignimbrite found in the LMRF displaying the various crystal compositions and sizes.

The ignimbrite unit has a Shands index A/CNK (=Al2O3/CaO+Na2O+K2O, in moles) range of 1.48-1.66 and A/NK (=Al2O3/Na2O+K2O, in moles) range of 1.54-1.81 which qualifies it as peraluminous. Rankin (1993) generally identified the three rhyolites of the UMRF to have a peralkaline to metaluminous affinity. Novak and Rankin (2004) specifically described the Wilburn Rhyolite to have a peralkaline basal layer with a metaluminous upper part. The transition of peraluminous to metaluminous and peralkaline represents a loss of mica over the evolution of the Mount Rogers volcanic center.

Figure 14. Sample BQ2 is from the vesicular basalt unit and contains large ( > 5 mm) chlorite crystals. Most of the feldspar grains and chlorite grains are altering to epidote. Small cubic pyrites, clinopyroxene grains and calcite grains occur sporadically throughout the sample. Picture taken in cross-polarization (XP).

30 Figure 15. Sample Green1 was taken from the pillow basalt unit and exhibits a relatively aphanitic groundmass with a variolitic texture. The lath-like feldspar crystals are altering to epidote and sericite. The grain sizes in the slide range from micron-scale level to 0.5 mm in length. Picture taken in plain-polarization (PP).

Figure 16. Sample Pill1b is from the pillow basalt unit which contains primary olivine grains being “consumed” by epidote grains. Albitic feldspar grains show preferred orientations, perhaps a relict igneous flow texture. Distinct reaction rims around the epidotes can be seen, showing irregular boundaries of alteration. In this slide, late-stage subsolidus deformation may be observed in deformed chlorite grains. Picture taken in cross-polarization (XP).

31 Figure 17. From the vesicular basalt unit, sample V1 contains a large serpentine grain with secondary calcite, as well as large (3-4 mm in length) chlorite grains. The lath-like feldspar grains appear to have no preferred orientation, and red iron oxides are visible on the edges of minerals. Picture taken in cross-polarization (XP).

Figure 18. Sample S1 is from the sheet flow basalt unit and displays a matrix of relict feldspar and pyroxene grains with larger (<.25 mm) chlorite and epidote grains. Some of the chlorites and epidote grains are seen to be clumped together, enclosed in reaction rims of sericite. Picture taken in cross-polarization (XP).

32 Figure 19. Sample IG1, from the ignimbrite unit contains large (1-2 mm) grains of all types of minerals seen in the other samples, which is to be expected in an ignimbrite. Quartz, K-feldspar, calcite, microcline, amphibole, dioside, pagioclase, and zircon can all be found in this sample. In this slide, exsolution features can be seen on the individual grains. Picture taken in cross-polarization (XP).

Geochemistry LMRF basalts are geochemically consistent with alkaline basalt protoliths, and LMRF ignimbrites resemble to a first order other MRF rhyolites. A Total Alkalies versus

SiO2 classification plot (Figure 20) includes the samples collected in this study, as well as MRF samples analyzed by Hernandez et al., 1994 and the 2008 spring Analytical Techniques in Geology class (Appendix 1). The rocks range from around 40-75 wt. %

SiO2 and 2-13 wt. % Na2O+K2O. The rocks constitute a bimodal volcanic suite of alkaline mafic rocks and rhyolites, with few lavas of intermediate composition.

The basalts of the LMRF range from 40-53 wt. % SiO2 and 3-7 wt. % MgO, and the ignimbrite unit samples have SiO2 contents ranging from 72-77 wt. %. Major oxide variation plots for the MRF show mineral phase crystallization trends with respect to increasing MgO and SiO2 content (Figure 21; Figure 22).

33 Figure 20. Total Alkali concentration vs. Silica concentration (TAS) plot modified from LeBas et al., 1986.

34 Figure 21. MnO and TiO2 vs. MgO for LMRF basalts. TiO2 forms broadly inverse arrays versus MgO, while no clear patterns are evident for MnO.

35 Figure 22. Al2O3 and K2O vs. SiO2 for the LMRF ignimbrite and UMRF rhyolites. Wilburn data from Novak and Rankin, 2004. Wilburn*, Whitetop* and Buzzard Rock* data from Escobar, 1972. Al2O3 shows a strong inverse correlation with increasing SiO2, while K2O trends are more scattered and broadly inverse.

LMRF Ignimbrite and UMRF Rhyolites Trace Element Systematics Rare Earth Element (REE) systematics of the LMRF ignimbrite shows distinct light rare earth (LREE) enrichments, comparatively flat heavy rare earth (HREE) patterns, and high overall REE abundances (between 20 and 400 times chondrite values) (Figure 23). Distinct negative Eu anomalies indicate plagioclase crystallization effects as

Eu2+ is taken up preferentially by plagioclase; while variable Ce anomalies suggest 36 secondary hydrothermal effects, as Ce4+ can be mobilized metasomatically. There is an order of magnitude increase in metaignimbrite HREE abundances relative to those of the Cranberry gneiss, which constitutes the >1.0 Ga basement rocks in the region (Figure 23). Comparing the LMRF ignimbrites to the rhyolites of the UMRF, it is clear that the patterns are roughly parallel overall, (Figure 24) but that the UMRF lavas exhibit a much stronger negative Eu anomaly and overall higher REE abundances. The Eu anomaly differences may be attributed to differing eruptive conditions (i.e., feldspar segregated from rhyolite lavas before eruption, while the ignimbrite erupted both lava and crystal separates), but despite these differences it is apparent that the rhyolite units of the UMRF and the ignimbrite found in the LMRF are geochemically similar. In particular, the Buzzard Rock Rhyolite closely parallels the LMRF ignimbrite samples suggesting that they are similar in source and possibly the same unit expressed as different eruptive styles (Figure 24). In Figure 25, the field for the Crossnore Complex granites from north and central Virginia is overlain on the MRF REE patterns obtained during this study. The MRF REE patterns largely parallel those from the Crossnore suite, but they are slightly less enriched in LREE. This may indicate that lower degrees of fractional crystallization or higher degrees of melting are involved in the formation of the LMRF high-silica eruptive products than for the Crossnore Complex granitoids.

37 Figure 23. REE abundances of three samples taken from the LMRF ignimbrite, plus the Cranberry gneiss basement. Note that the ignimbrite is enriched in HREE relative to the basement. Chondrite values from Haskin et al., 1968.

Figure 24. REE abundance of the LMRF ignimbrite, Cranberry gneiss basement, and UMRF rhyolites. Wilburn data from Novak and Rankin, 2004. Wilburn*, Whitetop* and Buzzard Rock* data from Escobar, 1972. Chondrite values from Haskin et al., 1968.

38 Figure 25. REE abundances of the LMRF ignimbrite, Cranberry gneiss basement, and UMRF rhyolites with a field of granites from the plutonic Crossnore Complex. Wilburn data from Novak and Rankin, 2004. Wilburn*, Whitetop* and Buzzard Rock* data from Escobar, 1972. REE field of the Crossnore Complex from Tollo and Aleinikoff, 1996; Tollo et al., 2004. Chondrite values from Haskin et al., 1968.

LMRF Basalts Trace Element Systematics LMRF basalts are plotted on a Ti/100-Zr-Y*3 discrimination diagram (Pearce and Cann, 1973; Figure 26) to infer their tectonic environments. Figure 26 shows that all samples fall in the “Within Plate Basalt” field, consistent with intraplate or continental rift origins. Some of the samples verge towards the MORB field, possibly indicating to a transition from rifting to oceanic spreading in the LMRF rocks. In Figure 27, the LMRF basalts are plotted on a 1 atm projected phase diagram for alkaline mafic lavas (from Sack et al., 1987) to ascertain the likely crystallizing assemblages for these rocks and to constrain their degree of evolution. All of the samples lie on or close to the Ol+DI(+PL) cotectic boundary, which indicates that all of the samples record three-phase crystallization at low pressures in a shallow magma chamber. Thus, the LMRF basaltic lavas were relatively evolved melts by the time they were erupted.

39 Figure 26. Ti-Zr-Y tectonic discrimination diagram for the LMRF basalts. Plot modified from Pearce and Cann, 1973.

Figure 27. LMRF basalts plotted on a olivine-silica-diopside ternary diagram. The cotectic boundary is the 1 atmosphere plagioclase-saturated liquidus. Values are calculated from the normative components of olivine (Ol), diopside (DI), silica (SIL), and plagioclase (PLAG). Fields labeled on diagram are pigeonite (PIG) and orthopyroxene (OPX). Arrows indicate directions of decreasing temperature. Modified from Sack et al., 1987 and Grove et al., 1982, 1983.

40 REE systematics for the LMRF basalts (Figure 28) show that all of the samples are strongly LREE-enriched, with elevated La/Sm and La/Yb ratios, and slight negative Eu anomalies in some samples. The sub-parallel nature of the patterns indicate that these lavas are cogenetic (i.e., they come from the same magma chamber). The LMRF basalts are markedly enriched in REEs overall as compared to Mid-ocean Ridge Basalts (MORBs). Their strongly LREE enriched patterns and different HREE slopes indicate that the LMRF basalts were derived from a non-depleted mantle source. When comparing the REE patterns for the UMRF rhyolites and LMRF ignimbrite (Figure 24) to the LMRF basalts (Figure 28), it can be said that even though the overall abundances of the REEs are different, the overall slope (average La/Yb for basalts and rhyolites is approx. 4.2 and 6.5, respectively) in both figures are very similar. This slope similarity points to the possibility of the UMRF rhyolites and LMRF ignimbrite having a petrogenetic link to the LMRF basalts by an initial source. Alteration by crystallization and crustal assimilation of an initial source has lead to the present REE patterns for the LMRF ignimbrite and UMRF rhyolites.

Figure 28. REE abundances of the LMRF basalts with the MORB field. Chondrite values from Haskin et al., 1968.

41 To place constraints on the relationships between MRF volcanism and later rift- related volcanic episodes in the region, the LMRF basalts are compared to the ~180 Ma younger Catoctin Formation metabasalts of northern Virginia. Figure 29 plots the REE patterns of both LMRF basalts and Catoctin Formation metabasalts. The shapes of the LMRF and Catoctin REE patterns indicate that these rocks are derived from geochemically similar sources (Figure 29). The relatively higher REE abundances of the LMRF suggest that the LMRF basalts have undergone a greater extent of crystallization before eruption than the Catoctin Formation metabasalts did.

Figure 29. REE abundances of the LMRF basalts and Catoctin Formation metabasalts. Catoctin metabasalt data from Badger and Sinha, 2004. Chondrite values from Haskin et al., 1968.

To elucidate even more the geochemical relationships between the later rift- related volcanic sequences in the Appalachians (i.e., Catoctin Formation and Camel's

Hump Group) and the MRF suite, the La/Yb ratios of these lavas versus K2O concentrations and the La/Sm ratios versus TiO2 concentrations are examined in Figure 30A. On these diagrams, the variation patterns of sample suites will differ as a function of the dominant magmatic process. If partial melting is the primary control on the

42 geochemical variations of Appalachian rift-related rocks, strongly incompatible trace element abundances should vary substantially, while more compatible elements will vary less. Thus, during partial melting the La/Yb and La/Sm ratios should change significantly related to the degrees of melting at the range of values for F, the fraction of melt formed, that are typical for melting on Earth (F = 0-0.1), as will TiO2 and K2O. On the other hand, if fractional crystallization is the dominant effect, La/Sm and La/Yb should not vary much, given the range in value for the fraction of melt remaining

(typically between 0.7 and 0.99), while the K2O and TiO2 content will still change significantly according to the degree of crystallization. In the top of Figure 30A, both melting and crystallization-related trends can be seen in the data. The LMRF basalts and Catoctin metabasalts appear to fall along a crystallization trend at similar percents of partial melting, while the Camel's Hump Group rocks show a crystallization trend, but at a lower degree of melting. In the bottom of Figure 30A, the MRF and all but one of the Catoctin Formation samples again form a single subhorizontal trend, consistent with crystallization. The Camel's Hump Group samples form a separate melting and crystallization trend, but at higher overall TiO2 contents. In Figure 30B the MRF and other Appalachian rift-related volcanic data are compared to suites from the Rio Grande Rift and the East African Rift Zone. In all of these suites, similar melting/crystallization relationships are evident, supporting the contention that the Appalachian rift-related metavolcanic suites, despite being separated from one another by >500 km distance along strike in the mountain range and up to 180 Ma in time, can be explained by variable degrees of partial melting of a chemically similar mantle source, followed by crystallization in a shallow magma chamber.

43 Figure 30A. Top: La/Yb (normalized) vs. K2O and Bottom: La/Sm (normalized) vs. TiO2 for low-silica rift-related rocks in the Appalachians. Catoctin data from Badger and Sinha, 2004. Camels Hump Group data from Coish et al., 1985.

44 Figure 30B. La/Sm (normalized) vs. TiO2 for rift-related low silica rocks from the Appalachians, East African rift, and Rio Grande rift. Catoctin data from Badger and Sinha, 2004. Camels Hump Group data from Coish et al., 1985. East African data from Barrat et., 2003; Haase et al., 2000; Barrat et al., 2003. Rio Grande rift data from Gibson et al., 1992; Baldridge, 1979a; Leat et al., 1989.

Another way to elucidate the geochemical relationships between the later rift- related volcanic sequences in the Appalachians and the MRF suite is to infer the depth of melting. Badger and Sinha (2004) used a Ce/Yb and Sm/Yb plot to infer the depth of melting for the Catoctin Formation and Camel's Hump Group basalts. The LMRF basalts from this study were added to this plot in Figure 31. The LMRF suite overlaps the Catoctin field in the diagram, suggesting that they had similar depths of melting. The Camel's Hump Group has higher Ce/Yb and Sm/Yb ratios (i.e., higher depth of melting) and partially overlaps the Catoctin Formation field, but not the LMRF. This trend suggests that as you continue on strike to the north from the MRF in Virginia to the

Camel's Hump Group in Vermont, a deeper source is being tapped as the rift is evolving.

45 It may also be inferred from Figure 31 that since the LMRF and the Catoctin Formation overlap, they may represent melts from the same source, were it not for the fact that their eruptions were separated by over 180 Ma.

Figure 31. Depth of melting plot indicated by Sm/Yb vs. Ce/Yb. Modified from Badger and Sinha, 2004.

46 Relationship Between LMRF Basalts and UMRF Rhyolites: Examples From Other Rift Settings

The REE systematics of LMRF basalts and ignimbrites, as well as those of the UMRF rhyolite units both preserve high REE concentrations and prominent LREE enrichments. The MRF patterns (especially those of the basalts) are similar to the REE patterns of the Catoctin Formation metabasalts and metabasalts of the Camel's Hump Group, but differ markedly from the REE signature of the prevalent basement rocks in the region (Cranberry gneiss). Most models for the formation of silicic lavas involve extensive assimilation and crystallization from a mafic parental magma (i.e., Hildreth, 1981), with the evolved lava moving toward the chemical signature of the primary assimilant. Novak and Rankin (2004) suggest that the evolutionary processes evident in the Wilburn Rhyolite follow those outlined by Hildreth (1981), pointing to extensive assimilation, yet our rare-earth element results do not indicate a transition toward crustal compositions. In Figure 32, we infer mixing of the MRF on a La/Yb (normalized) vs. Sm/Yb (normalized) plot with the Grenvile basement (Cranberry gneiss), N-MORBs, E- MORBs, and spinel peridotite as possible end members in a mixing trend. It is evident from Figure 32 that the basement has anomalously high La/Yb and Sm/Yb ratios compared to the MRF volcanic suite, suggesting that very little (<5%) assimilation of the basement was involved in the genesis of these magmas. A possible linear mixing trend can be seen between the LMRF basalts, Wilburn Rhyolite, and Cranberry gneiss supporting the extensive assimilation suggestion of Novak and Rankin (2004) for the Wilburn Rhyolite. Thus, while assimilation is clearly involved in the genesis of high- silica MRF lavas, their anomalously elevated trace-element signatures, typical of alkaline-series magmas, mask this contribution to a great degree. 47 Figure 32. La/Yb (normalized) vs. Sm/Yb (normalized) plot for the MRF. Cranberry gneiss, N-MORB, E-MORB and Spinel Peridotite added for possible end member in a mixing trend. N-MORB, E-MORB and Spinel Peridotite data from Geochemical Earth Reference Model (GERM).

Bimodal volcanism (i.e., basalt-rhyolite, with little to no intermediate compositions in between) such as is seen in the MRF is a feature seen in both continental rifts (i.e., East African Rift) and continental hot spots (i.e., Yellowstone hot spot). Some authors believe that rifting and hot spots are intimately linked, and that continental rifting may be initiated by mantle plumes (i.e., hot spot). Some models hypothesize that a mantle plume triggers partial melting of the crust, and can sometimes cause continental breakup (Campbell and Griffiths, 1990). Other models portray the mantle plume as a passive participant, only causing extension by an anomalously heated mantle (Peate et al., 1990). As previously noted, Rankin (1976) hypothesized that the bimodal volcanic rocks of the MRF erupted as a result of a hot spot under a triple junction of intersecting continental rifts. The most recent work done by Aleinikoff et al. (1995) and Tollo et al. (2004) argue against the triple junction hypothesis, but do not rule out the involvement of

48 a mantle plume. According to many authors, (e.g., Wilson, 1989; Macdonald et al., 1995; Baker et al., 1996b, 2000; Ayalew, 2000; Cox, 1988; Garland et al., 1995; Kar et al., 1998; Borg and Clynne, 1998; Grunder, 1995) the bimodal volcanism found in continental rifting systems, especially rifting systems with continental flood basalts (CFB), appear to be the result of fractional crystallization plus crustal assimilation and/or decompression-related partial melting of underplated basaltic magmas emplaced during extrusion of flood basalts. To try and identify likely petrogenetic models for the basalt followed-by rhyolite sequence of magmatism observed in the MRF, we have examined models proposed for several modern rifting systems: specifically, the East African Rift (EAR), Rio Grande Rift (RGR) and the North American Basin and Range (BR). The EAR displays basalt sequences followed by or intercalated with voluminous rhyolite/ignimbrite sequences, much like the MRF. Also, like in the MRF, the rhyolites/ignimbrites of the EAR are found to always overlay the basalts, and recurrences of the basalt-rhyolite stratigraphy are seen lower in the section (e.g. the Fees Rhyolite in the LMRF). This specific stratigraphy seen in the EAR is believed to be the result of a CFB-type eruption as the first pulse of event, followed by voluminous rhyolitic/ignimbrite eruptions >105 (e.g., Cawthorn and Walraven, 1998; Hawkesworth et al., 2000 ), with little to no intermediate compositions magmas (Ayalew et al., 2002).

Past workers (Gibson, 1972; Barberi et al., 1975; Walter et al., 1987; Gasparon et al., 1993; Boccaletti et al., 1995) have suggested that the petrogenesis of the EAR bimodal volcanic suite is via fractional crystallization of basaltic magmas that have seen minimal crustal contamination. A model proposed by Ayalew et al. (2002) to explain the change from basaltic to rhyolitic eruptions in the EAR examines stress field changes over the evolution of the volcanic field. They believe that the mantle plume at the base of the lithosphere that

49 produced the initial EAR volcanism caused radial extension, making the vertical lithostatic load above the plume the largest stress component, sigma 1, and in turn, leading to sigma 2 and sigma 3 being horizontal. As a result of the direction of sigma 1, magma storage and transportation in the early stages of magma ascent was vertical (dikes), creating the CFB sequences in the EAR. The vertical diking increased sigma 2 and sigma 3 by affecting the magma pressure and the Poisson coefficient (Vigneresse et al., 1999), which allowed them to exceed sigma 1. This increase of sigma 2 and sigma 3 over sigma 1 would have caused magma storage and transportation to switch from vertical diking to horizontal magma chambers. Thus, the storage time of melts in the magma chambers would have increased and allowed for long enough differentiation times that basaltic melts could evolve toward rhyolitic melts. Simultaneously, as the CFB volcanism continued, lithostatic load pressures increased, making it almost impossible to create horizontal magma chambers. Therefore, in the latter stages of CFB volcanism, the orientation of magma storage and transportation should be vertical. This could have meant that the magma chambers were formed at the onset of basaltic extrusion and were differentiating the entire span of volcanic activity. In the MRF there is no evidence for CFB volcanism, but the model presented above for the EAR may still be used to explain the bimodal volcanism observed in the MRF. If the hot spot hypothesis by Rankin (1976) is true, then it may be possible that even though CFB volcanism is not observed in the MRF like in the EAR, the crust above the mantle plume when the rifting of the Mt. Rogers region was active saw the same stress field changes as have been observed in the EAR. In contrast to the EAR and MRF, bimodal volcanism found in the RGR followed the eruption of a thick suite of intermediate to silicic eruptive products. This stratigraphic sequence is believed to reflect a shift in primary magma source from lithospheric to asthenospheric as a change in the tectonic regime occurred. The intermediate rocks are

50 believed to be the products of pre-extension volcanism caused by the subduction of the Farallon plate, while the bimodal volcanism is a product of crustal extension initiated by an anomalous mantle plume associated with the collapse of the failed Farallon plate (McMillan et al., 2000). The initiation of extension in the RGR may be different than the initiation of extension for the EAR and MRF, but the same bimodal suites are seen indicating that similar magmatic processes must have occurred. The bimodal volcanism expressed in the northern BR is the result of the Yellowstone hot spot and the associated active continental rifting of the BR. Many authors (e.g., Leeman, 1982a and Mabey, 1982) suggest that the track of the Yellowstone hot spot is represented by the Snake River Plain, which has moved through the BR province. Over the evolution of the Yellowstone hot spot, it has erupted a bimodal volcanic suite ranging from olivine tholeiites to rhyolitic domes and ash-flows (Leeman, 1982a,b). Draper (1991) contends that the petrogenetic relationship between the basalt and rhyolite end members of the northern BR bimodal suite do not result from fractional crystallization and crustal assimilation as in the EAR and RGR, but rather that the rhyolites are crustal melts initiated by basaltic injections from the Yellowstone mantle plume. In contrast to the northern BR, the volcanism found in the southern BR is a result of partial melting associated with continued extension, not a mantle plume (Bradshaw et al., 1993).

In the context of modern systems undergoing continental extension, several possible models can be called on to explain the stratigraphic and petrologic characteristics of the MRF. Many of these models involve mantle plumes, which is consistent with the contentions of Ranking (1976), but not all models require plume activity to generate bimodal alkaline magma suites. It is clear from these comparisons that the MRF does possess many of the signatures seen in modern rifts, and that both magma chamber crystallization and assimilation processes were involved in the genesis

51 of these lavas.

52 Conclusion

The study presented in this thesis was carried out to place constraints on the eruptive settings of the Precambrian lower Mount Rogers Formation, and on the relationship of Mount Rogers Formation volcanism to later Precambrian rift-related volcanic episodes in the Appalachians. The following conclusions can be made on the basis of this study:

1) The Elk Garden Ridge study area exposes an upright sequence of volcanic rocks from the lower Mount Rogers Formation. Subaerially erupted basalts on the bottom of the section transition into distinctively subaqueous eruptive styles (pillow lavas) up section, overlain by a ignimbrite that is compositionally similar to the upper Mount Rogers Formation rhyolites. The sequence appears to record a basin filling volcanic event during continental rifting of Laurentia in the Neoproterozoic.

2) Compositionally, all of the rocks in the field area are strongly enriched in alkali metals, with elevated Na2O and K2O contents, and high TiO2 in the basalts, consistent with “anorogenic” magmatism associated with the earliest stages of Iapetan rifting.

3) Even though there is approximately 180 my and >500 km between the Mount Rogers Formation and the Catoctin Formation and the other latest Precambrian rift-related volcanic sequences preserved in the Appalachians (i.e., Camel's Hump Group), they have remarkably similar geochemical signatures, suggesting a shared origin due to melting of a compositionally uniform sub-continental mantle source, followed by shallow-level

53 fractional crystallization. A more detailed study is needed to petrogenetically relate the UMRF to the LMRF. Also, a study relating the MRF to the dike swarm associated with the successful pulse of rifting that intruded the MRF may allow for some more insight into the petrogenetic relationships between the failed first pulse of rifting and the successful second pulse of rifting, and maybe the evolution of the Laurentian margin.

54 References

Aleinikoff, John N., Robert E. Zartman, Marianne Walter, Douglas W. Rankin, Peter T. Lyttle, and William C. Burton. 1995. U-pb ages of metarhyolites of the catoctin and mount rogers formations, central and southern appalachians: Evidence for two pulses of iapetan rifting. The American Journal of Science. 295, (4): 428. Allen, Rhesa M.,Jr. 1963. Geology and mineral resources of greene and madison counties. Virginia Division of Mineral Resources Publication: 102. Ayalew, D., 2000. Origin by fractional crystallization of transitional basalt for the Asela- Ziway pantellerite. Chemical Geology 168, 1-3. Ayalew, D. P. Barbey, B. Marty, L. Reisberg, G. Yirgu, and R. Pik, 2002. Source, genesis, and timing of giant ignimbrite deposits associated with Ethiopian continental flood basalts. Geochimica et Cosmochimica Acta 66, 8: 1429-1448. Badger, R.L. 1989. Geochemistry and petrogenesis of the catoctin volcanic province, central appalachians.Blacksburg, Virginia Polytechnic Institute. Badger, R.L., and A.K. Sinha. 1988. Age and sr isotopic signature of the catoctin volcanic province; implications for subcrustal mantle evolution. Geology (Boulder) 16, (8) (Aug): 692-5. Badger, Robert L., A. Krishna Sinha, Richard P. Tollo, Louise Corriveau, James McLelland, and Mervin J. Bartholomew. 2004. Geochemical stratigraphy and petrogenesis of the catoctin volcanic province, central appalachians; proterozoic tectonic evolution of the grenville orogen in north america. Memoir - Geological Society of America 197,435-58. Baker, J.A., M.F. Thirlwall, and M.A. Menzies, 1996b. Sr-Nd-Pb isotopic and trace element evidence for crustal contamination of plume-derived flood basalts: Oligocene flood volcanism in western Yemen. Geochimica et Cosmochimica Acta 60, 2559-2581. Baker, J.A., C.G. Macpherson, M.A. Menzies, M.F. Thirlwall, M. Al-Kadasi, and D.P. Mattey, 2000. Resolving crustal and mantle contributions to continental flood volcanism, Yemen: Constraints from mineral oxygen isotope data. Journal of Petrology 41, 1805- 1820. Baldridge, W.S. 1979. Mafic and ultramafic inclusion suites from the rio grande rift (new mexico) and their bearing on the composition and thermal state of the lithosphere. Journal of Volcanology and Geothermal Research 6, (3-4) (Nov): 319-51. Barberi, F., G. Ferrara, R. Santacroce, M. Treuil, and J. Varet, 1975. A traditional basalt- pantellerite sequence of fractional crystallization, the Bonia Centre (Afar Rift, Ethiopia). Journal of Petrology 16, 22-56. 55 Barrat, J.A., J.L. Joron, R.N. Taylor, S. Fourcade, R.W. Nesbitt, and B.M. Jahn. 2003. Geochemistry of basalts from manda hararo, ethiopia: LREE-depleted basalts in central afar. Lithos 69, (1-2) (JUL): 1-13. Boccaletti, M., A. Getaneh, R. Mazzuoli, L. Tortorici, and T. Trua, 1995. Chemical variations in a bimodal magma system: The plioQuarternary volcanism in the Dera Nazret area (Main Ethiopian Rift, Ethiopia). African Geoscience Reviews 2, 37-60. Borg, L.E., and M.A. Cylnne, 1998. The petrogenesis of felsic calc-alkaline magmas from the southernmost Cascades, California; Origin by partial melting of basaltic lower crust. Journal of Petrology 39, 1197-1222. Boyer, Steven E. 1984. Origin and significance of compositional layering in late precambrian sediments, blue ridge province, north carolina, U.S.A; planar and linear fabrics of deformed rocks; a selection of papers delivered at an international conference held at ETH. Journal of Structural Geology 6, (1-2): 121-33. Bradshaw, T.K., C.J. Hawkesworth, and K. Gallagher, 1993. Basaltic volcanism in the Southern Basin and Range: no role for a mantle plume. Earth and Planetary Science Letters 116, 45-62. Bryant, Bruce, and Reed, J.C. Jr. 1970. Geology of the grandfather mountain window and vicinity, north carolina and tennessee: 190. Cambell, I.H. And Griffiths, R.W., 1990. Implications of mantle plume structure for the evolution of flood basalts. Earth Planetary Science Letters, 99: 79-93. Cawood, P.A., P.J. McCausland, and G.R. Dunning. 2001. Opening iapetus: Constraints from the laurentian margin in newfoundland. Bulletin of the Geological Society of America 113, (4) (Apr): 443-53. Cawthorn, R.G., and F. Walraven, 1998. Emplacement and crystallization time for the Bushvld complex. Journal of Petrology 39, 1669-1987.

Coish, R. A., F. S. Fleming, M. Larsen, R. Poyner, and J. Seibert, 1985. Early rift history of the proto-atlantic ocean; geochemical evidence from metavolcanic rocks in vermont. American Journal of Science 285, (4) (Apr): 351-78. Cox, K.G., 1988. The Karoo Province. In Continental Flood Basalts (ed. J.D. Macdougall), 239-271, Kluwer, Dordrecht. Draper, D.S., 1991. Late Cenozoic bimodal magmatism in the northern Basin and Range Province of southern Oregon. Journal of Volcanology and Geothermal Research 47, 299- 328. Escobar, Leopoldo Lopez. 1972. Appalachian rhyolites; geochemical data concerning their origin. Master's thesis, Massachusetts Institute of Technology, Cambridge, MA, United States (USA). Espenshade, G.H.,1986. Geology of the Marshall quadrangle, Fauquier County, Virginia. 56 United States Geological Survey Bulletin. 1560, 60. Fenneman, N.M., and Johnson, D.W., 1946. Physical divisions of the United States: U.S. Geological Survey, scale 1:7,000000, 1 sheet. Fetter, Allen H., and Steven A. Goldberg. 1995. Age and geochemical characteristics of bimodal magmatism in the neoproterozoic grandfather mountain rift basin. Journal of Geology 103, (3) (May): 313-26. Fullagar, Paul D., and A. Leroy Odom. 1973. Geochronology of precambrian gneisses in the blue ridge province of northwestern north carolina and adjacent parts of virginia and tennessee. Geological Society of America Bulletin 84, (9): 3065-79. Garland, F., C.J. Hawkesworth, M.S.M. Mantovani, 1995. Description and petrogenesis of the Parana rhyolites, Southern Brazil. Journal of Petrology 36, 1193-1227. Gasparon, M., F. Innocenti, P. Manetti, A. Peccerillo, and A. Tsegaye, 1993. Genesis of the Pliocene to Recent bimodal mafic-felsic volcanism in the Debre Zeyt area, central Ethiopia: Volcanological and geochemical constraints. Journal of African Earth Science 17, 145-165. Geochemical Earth Reference Model (GERM). http://earthref.org/cgi-bin/er.cgi?s=germ- s0-main.cgi. Gibson, I.L., 1972. The chemistry and petrogenesis of a suite of pantellerites from the Ethiopian rift. Journal of Petrology 13, 31-44.

Gibson, S.A., R.N. Thompson, P.T. Leat, A.P. Dickin, M.A. Morrison, G.L. Hendry, and J.G. Mitchell. 1992. Asthenosphere-derived magmatism in the rio grande rift, western USA; implications for continental break-up; magmatism and the causes of continental break-up. Geological Society Special Publications 68, 61-89. Goldberg, S.A., J.R. Butler, and P.D. Fullagar. 1986. The bakersville dike swarm: Geochronology and petrogenesis of late proterozoic basaltic magmatism in the southern appalachian blue ridge. American Journal of Science 286, 403-430. Grove, T.L., D.C. Gerlach, and T.W. Sando. 1982. Origin of calc-alkaline series lavas at medicine lake volcano by fractionation, assimilation and mixing. Contributions to Mineralogy and Petrology 80, (2): 160-82. Grove, Timothy L., and Michael B. Baker. 1983. Effects of melt density on magma mixing in calc-alkaline series lavas. Nature (London) 305, (5933) (29 Sep): 416-8. Grunder, A.L., 1995. Material and thermal roles of basalt in crustal magmatism: Case study from eastern Nevada. Geology 23, 952-956. Haase, Karsten M., Richard Muehe, and Peter Stoffers. 2000. Magmatism during extension of the lithosphere; geochemical constraints from lavas of the shaban deep, northern red sea. Chemical Geology 166, (3-4) (May): 225-39. Haskin, Larry, M.A. Haskin, F.A. Frey, and T.R. Wildeman. 1968. Relative and absolute 57 terrestrial abundances of the rare earths, in Ahrens, L.H., ed., Origin and distribution of the elements: 889-912. Hawkesworth, C.J., S. Blake, P. Evans, R. Hughes, R. Macdonald, L.E. Thomas, S.P. Turner, and G. Zellmer, 2000. Time scale of crystallization in magma chambers- Integrating physical, isotopic and geochemical perspectives. Journal of Petrology 41, 991-1006. Hernandez, David, Chad Gunter, and Jeffrey G. Ryan. 1994. Geochemical affinities of metavolcanic rocks from the mt. rogers formation, SW virginia; geological society of america, southeastern section, 43rd annual meeting. Abstracts with Programs - Geological Society of America 26, (4) (Mar): 19. Hildreth, E.W., 1981, Gradients in silicic magma chambers: Implications for lithospheric magmatism. Journal of Geophysical Research 86, 10,153-10,192. James, Sara C. 1999. Field, petrographic, and geochemical constraints on the geologic history of the fees rhyolite member: Mount rogers formation, blue ridge province, VA. Bachelor's thesis, College of William and Mary. Kar, A., B. Weaver, J. Davidson, and M. Colucci, 1998. Origin of differentiate volcanic and plutonic rocks from Ascension Island, South Atlantic Ocean. Journal of Petrology 39, 1009-1024. Leat, P.T., R.N. Thompson, M.A. Morrison, G.L. Hendry, and A.P. Dickin. 1988. Compositionally-diverse miocene-recent rift-related magmatism in NW colorado; partial melting, and mixing of mafic magmas from three different asthenospheric and lithospheric mantle sources; oceanic and continental lithosphere; similarities and differences. Journal of Petrology 1988, 351-77. LeBas, M.J., R.W. Lemaitre, A. Streckeisen, and B. Zanettin. 1986. A chemical classification of volcanic-rocks based on the total alkali silica diagram. Journal of Petrology 27, (3) (JUN): 745-50. Leeman, W.P., 1982a. Development of the Snake River Plain-Yellowstone Plateau province, Idaho and Wyoming: An overview and petrologic model. In; B. Bonnichsen and R.M. Breckenridge (Editors), Cenozoic Geology of Idaho. Idaho Bur. Mines Geology Bulletin 26, 155-178.

Leeman, W.P., 1982b. Olivine tholeiitic basalts of the Snake River Plain, Idaho. In; B. Bonnichsen and R.M. Breckenridge (Editors), Cenozoic Geology of Idaho. Idaho Bur. Mines Geology Bulletin 26, 181-191. Mabey, D.R., 1982. Geophysics and tectonics of the Snake River Plain, Idaho. In; B. Bonnichsen and R.M. Breckenridge (Editors), Cenozoic Geology of Idaho. Idaho Bur. Mines Geology Bulletin 26, 139-154. Macdonald, R., G.R. Davies, B.G.J. Upton, P.N. Dunkley, M. Smith, and P.T. Leat, 1995. Petrogenesis of Silali volcano, Gregory Rift, Kenya. Journal of Geological Society of London 152, 703-720 McMillan, N.J., A.P. Dickin, D. Haag, 2000. Evolution of magmas source regions in the 58 Rio Grande Rift, southern New Mexico. Geological Society of America Bulletin 112, 10: 1582-1593. Neton, M.J., 1992, Late Proterozoic rifting of Laurentia: source and deposition of conglomerate units of the Grandfather Mountain Formation, North Carolina Blue Ridge: Unpub. M.S. Thesis, University of Tennessee, Knoxville, 180 p. Novak, S.W., and D.W. Rankin. 2004. Compositional zoning of a neoproterozoic ash- flow sheet of the mount rogers formation, southwestern virginia blue ridge, and the aborted rifting of laurentia. GSA Memoir, (197): 571-600.

Pearce, J.A., and J.R. Cann. 1973. Tectonic setting of basic volcanic rocks determined using trace element analyses. Earth Planetary Science Letters 19, 290-300. Peate, D.W, C.J. Hawkesworth, M.S.M Mantovani, and W. Shukowsky, 1990. Mantle plumes and flood basalt stratigraphy in the Parana, South America. Geology, 18, 123- 1226.

Peterson, Virginia, Jeffrey G. Ryan, and 1997-1998 REU Site Program. 2009. Petrogenesis and structure of the buck creek mafic-ultramafic suite, southern appalachians: Constraints on ophiolite evolution and emplacement in collisional orogens. Geological Society of America Bulletin 121, (3-4) (MAR-APR): 615-29. Rankin, D.W. 2009. Personal Communication. ———. 1970. Stratigraphy and structure of precambrian rocks in northwestern north carolina, in Fisher, G.W., Pettijohn, F.J., Reed, J.C., Jr., and Weaver, K.N., eds., Studies of appalachian geology-central and southern. New York, Interscience Publishers: 227- 245. Rankin, D.W., A.A. Drake, and N.M. Ratcliffe. 1989. Geologic map of the U.S. appalachians showing the laurentian margin and the taconic orogen, in Hatcher, R.D., Jr., Thomas, W.A., and Viele, G.W., eds., The appalachian-ouachita orogen in the united states. F-2 of The Geology of North America, . Rankin, D.W. 1976. Appalachian salients and recesses - late precambrian continental breakup and opening of iapetus ocean. Journal of Geophysical Research 81, (32): 5605- 19. ———. 1975. Continental-margin of eastern north-america in southern appalachians - opening and closing of proto-atlantic ocean. American Journal of Science A275, : 298,&. ———. 1968. Stratigraphic and structural position of the upper precambrian mount rogers volcanic group, virginia, north carolina, and tennessee. Special Paper - Geological Society of America: 495. Rankin, D.W., T.W. Stern, J.C. Reed Jr, and M.F. Newell. 1969. Zircon ages of felsic volcanic rocks in the upper precambrian of the blue ridge, appalachian mountains. Science 166, (3906): 741-4. Rankin, Douglas W. 1993. The volcanogenic mount rogers formation and the overlying 59 glaciogenic konnarock formation : Two late proterozoic units in southwestern virginia. Washington, D.C.; Denver, CO: U.S. G.P.O. ;U.S. Geological Survey, Map Distribution.

Rankin, Douglas W., Leopold Lopez-Escobar, and Fred A. Frey. 1974. Rhyolites of the upper precambrian mount rogers (virginia) volcanic center; geochemistry and petrogenesis. EOS, Transactions, American Geophysical Union 55, (4): 475. Rankin, Douglas W., Julia M. G. Miller, and Edward L. Simpson. 1994. Geology of the mt. rogers area, southwestern virginia blue ridge and unaka belt; fieldguides to southern appalachian structure, stratigraphy, and engineering geology. Virginia Tech Department of Geological Sciences 10, (Apr): 127-76. Rankin, Douglas Whiting, and Carolina Geological Society. 1967. Guide to the geology of the mt. rogers area, virginia, north carolina and tennessee. [Raleigh]: Carolina Geological Society. Reed, John C.,Jr, and Benjamin A. Morgan. 1971. Chemical alteration and spilitization of the catoctin greenstones, shenandoah national park, virginia. Journal of Geology 79, (5): 526-48. Reed, John Calvin, Jr. 1955. Catoctin formation near luray, virginia. Geological Society of America Bulletin 66, (7) (Jul): 871-96. Reed, M.F., R.C. Bartholomay, and S. S. Hughes. 1997. Geochemistry and stratigraphic correlation of basalt lavas beneath the idaho chemical processing plant, idaho national engineering laboratory. Environmental Geology (Berlin) 30, (1-2) (Mar): 108-18. Roden, M.F., F.A. Frey, and D.A. Clague. 1984. Geochemistry of tholeiitic and alkalic lavas from the koolau range, oahu, hawaii: Implication for hawaiian volcanism. Earth and Planetary Science Letters 69, (1): 141-58. Roden, Michael F., Frederick A. Frey, and David A. Clague. 1984. Geochemistry of tholeiitic and alkalic lavas from the koolau range, oahu, hawaii; implications for hawaiian volcanism. Earth and Planetary Science Letters 69, (1) (Jul): 141-58. Rodgers, John. 1953. Geologic map of east tennessee with explanatory text. Tennessee Division of Geology Bulletin 58(2): 550.

Sack, R.O., D. Walker, and I.S.E. Carmichael. 1987. Experimental petrology of alkalic lavas - constraints on cotectics of multiple saturation in natural basic liquids. Contributions to Mineralogy and Petrology 96, (1): 1-23. Stose, G.W., and A.J. Stose. 1944. The chilhowee group and ocoee series of the southern appalachians. American Journal of Science 242, (8): 401-16. ———. 1944. The chilhowee group, and ocoee series of the southern appalachians. American Journal of Science 242, (7): 367-90. Su, Qi, Steven A. Goldberg, and Paul D. Fullagar. 1994. Precise U-pb zircon ages of neoproterozoic plutons in the southern appalachian blue ridge and their implications for the initial rifting of laurentia. Precambrian Research 68, (1-2) (Jun): 81-95. 60 Tollo, Richard P., and John N. Aleinikoff. 1996. Petrology and U-pb geochronology of the robertson river igneous suite, blue ridge province, virginia; evidence for multistage magmatism associated with an early episode of laurentian rifting. American Journal of Science 296, (9) (Nov): 1045-90. Tollo, Richard P., John N. Aleinikoff, Mervin J. Bartholomew, and Douglas W. Rankin. 2004. Neoproterozoic A-type granitoids of the central and southern appalachians; intraplate magmatism associated with episodic rifting of the rodinian supercontinent. Precambrian Research 128, (1-2) (05 Jan): 3-38. Tollo, Richard P., and Frederick E. Hutson. 1996. 700 ma rift event in the blue ridge province of virginia; a unique time constraint on pre-iapetan rifting of laurentia. Geology (Boulder) 24, (1) (Jan): 59-62. Vigneresse, J.L., B. Tikoff, and L. Amegilo, 1999. Modification of the regional stress field by magma intrusion and formation of tabular granitic plutons. Tectonophysics 302, 203-224. Walsh, Gregory J., and John N. Aleinikoff. 1999. U-pb zircon age of metafelsite from the pinney hollow formation; implications for the development of the vermont appalachians. American Journal of Science 299, (2) (Feb): 157-70. Walter, R.C., W.K. Hart, and J.A. Westgate, 1987. Petrogenesis of a basalt-rhyolite tephra from the west-central Afar, Ethipoia. Contributions to Mineralogy and Petrology 95, 462- 480. Wilcox, Ronald E., and Arie Poldervaart. 1958. Metadolerite dike swarm in bakersville- roan mountain area, north carolina. Geological Society of America Bulletin 69, (11) (Nov): 1323-67. Wilson, M., 1989. Igneous Petrogenesis: A Global Tectonic Approach. Harper Collins, London.

61 Appendices

62 Appendix 1: Direct Current Plasma Spectrometer (DCP) Geochemical Data

Data from the USF Spring 2008 Analytical Techniques in Geology class and Hernandez et al., 1994. For DCP procedure and analytical precision, see Peterson and Ryan, 2009.

63