Phase Chemistry and Petrogenesis of Dacitic Lavas from the Southern Juan De Fuca Ridge

MISSING IMAGE

Material Information

Title:
Phase Chemistry and Petrogenesis of Dacitic Lavas from the Southern Juan De Fuca Ridge
Physical Description:
1 online resource (86 p.)
Language:
english
Creator:
Werts, Kevin R
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.S.)
Degree Grantor:
University of Florida
Degree Disciplines:
Geology, Geological Sciences
Committee Chair:
Perfit, Michael R
Committee Members:
Mueller, Paul A
Panning, Mark Paul
Smith, Mattew

Subjects

Subjects / Keywords:
dacite -- jdfr -- mixing -- mor -- pyroxene
Geological Sciences -- Dissertations, Academic -- UF
Genre:
Geology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Andesites, dacites and rhyodacites were collected from two constructional mounds on a series of hummocky and hooked ridges at the ridge transform intersection of the southern Juan de Fuca Ridge and Blanco Fracture Zone, during the ROV Tiburon’s dive T735. Unlike other evolved lavas collected from ridge axis discontinuities which are generally aphyric, those from dive T735 are moderately phyric, containing pyroxene, plagioclase, fayalitic olivine, quartz and zircon crystals that display evidence for magma mixing. Plagioclase and pyroxene compositions span wide compositional ranges from An27-72 and Mg# 17-84, respectively. Whereas individual plagioclase crystals are fairly homogenous, pyroxenes show strong compositional zoning and are characterized into three types. Type 1 pyroxenes have Mg-rich cores similar to pyroxenes in ferrobasalts collected near the constructional domes. Type 2 pyroxenes have Fe-rich cores, sometimes exhibiting exsolution lamellae. Type three pyroxenes constitute the groundmass and have intermediate compositions (Mg# ~ 55) similar to rims of type 1 and 2 pyroxenes. Complex zoning in pyroxene and the presence of basaltic xenoliths with chilled margins suggests that magma mixing played a role in their evolution. The disequilibria between dacitic liquids and evolved mineral phases indicate that these minerals were in equilibrium with a more siliceous end member magmatic composition. Mixing between an extremely fractionated basaltic andesite and a rhyolitic partial melt from a plagiogranite source explains most major elemental chemical trends and the presence of highly evolved mineral phases. However, the similarity between plagioclase and pyroxene compositions in the basaltic xenoliths and surrounding ferrobasalts suggests that a second batch of magma is responsible for their mobilization.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Kevin R Werts.
Thesis:
Thesis (M.S.)--University of Florida, 2012.
Local:
Adviser: Perfit, Michael R.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-02-28

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044737:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 PHASE CHEMISTRY AND PETROGENESIS OF DACITIC LAVAS FROM THE SOUTHERN JUAN DE FUCA RIDGE By KEVIN RICHARD WERTS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENT S FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Kevin Werts

PAGE 3

3 To Amanda

PAGE 4

4 ACKNOWLEDGMENTS I would like to thank my advisor Dr. Michael Perfit for his advice, support and guidance on this project and Dr. Matt Smith for his daily encouragements and advice. I would also like to thank Dr. Paul Mueller and Dr. Mark Panning for their helpful input and patience. The United States Geological Survey in Denver, CO, Dr. Ian Ridley and Dr. Heather Lowers are thanked for their helpf ulness and support in mineral analyses. I thank the Monterey Bay Aquarium Research Institute and Dr. David Clague for providing me a unique experience on the Juan de Fuca Ridge and for teaching me about sample collection using remotely operated vehicles. I would like to acknowledge Katie Garman and Dr. Rachel Walters for their helpful discussions and advice. I would like to thank my wife, Amanda, for all of her support and understanding. The students, faculty, and staff of the geology department are all tha nked for their friendship and support

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FIGURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ............................. 9 ABSTRACT ................................ ................................ ................................ ................... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 12 2 BACKGROUND ................................ ................................ ................................ ...... 17 Geological Setting ................................ ................................ ................................ ... 17 Pre vious Studies ................................ ................................ ................................ ..... 18 3 ANALYTICAL METHODS ................................ ................................ ....................... 22 4 GLASS GEOCHEMISTRY ................................ ................................ ...................... 25 Basalts and Basaltic Andesites ................................ ................................ ............... 25 Andesites and Dacites ................................ ................................ ............................ 27 5 PETROGRAPHY AND MINERAL CHEMISTRY ................................ ..................... 38 Basalts ................................ ................................ ................................ .................... 38 Andesites and Dacites ................................ ................................ ............................ 38 Microphenocrysts and Phenocrysts ................................ ................................ .. 39 Pyroxene ................................ ................................ ................................ .... 39 Plagioclase ................................ ................................ ................................ 41 FeTi Oxides ................................ ................................ ................................ 42 Olivine ................................ ................................ ................................ ........ 42 Accessory P hases ................................ ................................ ..................... 42 Crystal Clots ................................ ................................ ................................ ..... 43 Basaltic Xenoliths ................................ ................................ ............................. 43 6 DISCUSSION ................................ ................................ ................................ ......... 56 Fractional Crystallization ................................ ................................ ......................... 56 Evidence Supporting Magma Mixing ................................ ................................ ....... 57 Mineral Melt Equilibrium ................................ ................................ ......................... 59 Magma Mixing ................................ ................................ ................................ ........ 61

PAGE 6

6 Petrogenesis ................................ ................................ ................................ ........... 64 7 CONCLUSIONS ................................ ................................ ................................ ..... 79 LIST OF REFERENCES ................................ ................................ ............................... 80 BIOGRAPHICAL SKETCH ................................ ................................ ............................ 86

PAGE 7

7 LIST OF TABLES Table page 3 1 Chemical Analyses Performed on Dive T735 Samples. ................................ ..... 24 4 1 T735 Major Element Glass Compositions ................................ .......................... 29 4 2 T735 Trace Element Glass Compositions ................................ .......................... 31 5 1 Representative Pyroxene Analyses ................................ ................................ .... 45 5 2 Representative Plagioclase Analyses ................................ ................................ 46

PAGE 8

8 LIST OF FIGURES Figure page 1 1 Map of the Juan de Fuca Ridge ................................ ................................ ......... 14 4 1 Major element variation diagrams for dive T735 lavas and Cleft segment lavas ................................ ................................ ................................ ................... 34 4 2 Chlorine variations in dive T735 lav as and Cleft segment lavas ......................... 36 4 3 Primtive mantle n ormalized trace element diagram ................................ ........... 37 5 1 Plane polarized light images of characterized pyroxene types ........................... 47 5 2 Pyroxene and olivine compositions in T735 andesites and dacites .................... 48 5 3 Pyroxene core to rim chemical variations ................................ ........................... 51 5 4 Plagioclase core to rim An# variations ................................ ............................... 53 5 5 Plane polarized light images of basaltic xenolit hs ................................ .............. 54 7 1 Major element differenti ation trends for JdFR, GSC and EPR lavas .................. 66 7 2 Pyroxene quadrilateral comparing T735 pyroxene trends wi th other tholeiitic suites ................................ ................................ ................................ .................. 68 7 3 Histogram s of pyroxene core and rim Mg#s ................................ ....................... 69 7 4 Pyroxene equilibria diagram ................................ ................................ .............. 70 7 5 Olivin e equilibria diagram ................................ ................................ .................. 71 7 6 Mixing lines between sample G12 and various b asaltic lavas ........................... 72 7 7 Mixing lines between sample G12 and various basaltic lavas for MgO (wt.%) vs. Al 2 O 3 (w t.%). ................................ ................................ ................................ 78

PAGE 9

9 LIST OF ABBREVIATION S AMC Axial magma c hamber BFZ Blanco Fracture Zone EPR East Pacific Rise FTIR Fourier transform infrared GSC Galapagos Spreading Center HAM High amplitude magnetic ICP MS Inductively coupled plasma mass spectrometry JdFR Juan de Fuca Ridge LLD Liquid line of descent MBARI Monterey Bay Aquarium Research Institute MOR Mid ocean ridge MORB Mid ocean ridge basalt RAD Ridge axis discontinuity RTI Ridge transform intersection USGS United States Geological Survey wt.% Weight percent XRF X ray fluorescence

PAGE 10

10 Abstract of Thesis Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science PHASE CHEMI STRY AND PETROGENESIS OF DACITIC LAVAS FROM THE SOUTHERN JUAN DE FUCA RIDGE By Kevin Werts August 2012 Chair: Michae R. Perfit Major: Geology Andesites, dacites and rhyodacites were collected from two constructional mounds on a series of hummocky and hooked ridges at the ridge transform intersection of the southern Juan de Fuca Ridge and Blan T735. Unlike other evolved lavas collected from ridge axis discontinuities which are generally aphyric those from dive T735 a re moderately phyric containing pyro xene, plagioclase, faya litic olivine, quartz and zircon crystals that display evidence for magma mixing Plagioclase and pyroxene compositions span wide compositional ranges from An 27 72 and Mg# 17 84, respectively. Whereas individual plagioclase crys tals are fairly homogenous, pyroxenes show strong compositional zoning and are characterized into three types. Type 1 pyroxenes have Mg rich cores similar to pyroxenes in ferrobasalts collected near the constructional domes. Type 2 pyroxenes have Fe rich cores, sometimes exhibiting exsolution lamellae Ty pe three pyroxenes constitute the groundmass and have intermediate composition s (Mg# ~ 55) similar to rims of type 1 and 2 pyroxenes. C omplex zoning in pyroxene and the presence of basaltic xenoliths with chilled margins suggest s that magma mixing played a role in their evolution. T he disequilibria between dacitic liquids and evolved mi neral phases indicate that these

PAGE 11

11 minerals were in e quilibrium with a more siliceous end member magmatic composition. Mixing between an extremely fractionated basaltic andesit e and a rhyolitic partial melt from a plagiogranite source explains most major element al chemical trends and the presence of highly evolved mineral phases. However, the similarity between plagioclase and pyroxene compositions in the basaltic xenoliths and surrounding ferrobasalts suggests that a second batch of magma is responsible for their mobilization.

PAGE 12

12 CHAPTER 1 INTRODUCTION The mid ocean ridge (MOR) system is largely dominated by tholeiitic basalts; however, it is not restricted to mafic lavas. High si lica andesites, dacites and rhyodacites have been recovered from ridge axis discontinuities (RADs) in the Galapagos spreading center (GSC), East Pacific Rise (EPR) and Juan de Fuca Ridge (JdFR) (Byerly et al., 1976; Clague and Bunch, 1976; Byerly, 1980; Ch ristie and Sinton, 1981; Fornari et al., 1983; Perfit and Fornari, 1983; Perfit et al., 1983; Langmuir et al., 1986; Regelous et al., 1999; White et al., 2009; Wanless et al., 2010) and from areas of ridge hotspot interaction such as Iceland, the Pacific A ntarctic Ridge and Axial seamount on the JdFR (Hekinian et al., 1997; Hekinian et al., 1999; Chadwick et al., 2005; Haase et al., 2005). Previous studies have considered lower magma supplies and/or cooler crust at RADs and the propagation of ridge tips through ridge segment ends to explain the occurrence of these high silica MOR lavas (Christie and Sinton, 1981; Sinton et al., 1983; Fornari et al., 1983; Perfit et al., 1983; Juster et al., 1989, Rubin and Sinton, 2007; Wanless et al., 2010). Extreme cry stal fractionation, magma mixing, partial melting and assimilation of basaltic and/or gabbroic crust have all been hypothesized as processes that contribute to their development (Byerly et al., 1976; Byerly, 1980; Clague et al., 1981; Perfit et al., 1983; Juster et al., 1989; Hekinian et al., 1997; Haase et al., 2005; Chadwick et al., 2005; Wanless et al., 2010). This study focuses on a wide compositional range of lavas that were collected in 2004, using the ROV Tiburon during a cruise of the Monterey Bay Aquarium Research Institute (MBARI) to the southern end of the JdFR. Highly evolved mid ocean ridge

PAGE 13

13 basalt (MORB), andesite, dacite and rhyodacite were sampled during dive T735 at the ridge transform intersection (RTI) between the southern Cleft segment of the JdFR and the Blanco Fracture Zone (BFZ) (Figure 1 1). The moderately phyric nature of dacitic lavas from dive T735 make this sample set unique among those previously investigated asaltic xenoliths and the presence of highly evolved mineral phases indicate that the petrogenetic history of the dacites may be more complex than their glass chemistry alone suggests. This study aims to further assess the processes involved in the petroge nesis of MOR dacites at RADs using the petrography and phase chemistry of lavas collected from dive T735 on the southern JdFR.

PAGE 14

14 Figure 1 1 Map of the Juan de Fuca Ridge and T735 dive area A) Map of the entire Juan de Fuca Ridge, t he yellow box indic ates the general location of samples collected from dive T735 an d the location of image B B) H ooked and curved ridges at the ridge transform intersection. The yellow box indicates the general location of image C. C) Locations of samples collected on dive T735. A

PAGE 15

15 Figure 1 1. Continued B

PAGE 16

16 Figure 1 1. Continued C

PAGE 17

17 CHAPTER 2 BACKGROUND Geological Setting The intermediate spreading rate (~ 6 cm/yr full rate) Juan de Fuca spreading center is located off the coasts of Washington and Oregon, extending from the Sovanco Ryan 1986). Seven second order ridge segments (West Valley, Endeavour, Cobb, CoAx ial, Axial, Vance, and Cleft) make up the JdFR from north to south, respectively (Delaney et al. 1981; Normark et al. 1983; Embley et al., 1983; Crane et al., 1985; Kappel and Ryan, 1986; Embley et al., 1990). The Clef 27`N where it intersects the BFZ along its southern end (Embley et al., 1983; Embley et al. 1991; Smith et al., 1994). A 3 km wide axial valley hosts an axial summit collapse trough or fissure known as the 15 m d (Normark et al., 1986; Normark et al., 1987; Embley et al., 1991; Stakes et al., 2006). Seismic reflection profiles have imaged an axial magma chamber (AMC) reflector that is present beneath 60% of the Cleft segment bu end of the Cleft are generally more evolved (lower MgO wt.%) than those collected from its northern end (Sinton et al., 1983; S mith et al., 1994; Stakes et al., 2006). The evolved nature of the southern Cleft lavas is in agreement with the absence of an AMC reflector beneath the majority of the southern portion of the Cleft segment, suggesting s and lower temperatures may prevail in the south,

PAGE 18

18 relative to the north (Sinton and Detrick, 1992; Smith et al., 1994; Perfit and Chadwick, 1998; Stakes et al., 2006). A series of 16 dives using the ROV Tiburon investigated the geologic, morphologic and magnetic characteristics of the Cleft segment from 2000 2004 (Stakes et al., 2006). Among these dives was dive T735, which examined the southern terminus of the Cleft segment and a series of hummocky and hooked ridges that extend south of the RTI and appea r to have formed by magmatic propagation through the BFZ into older and cooler crust generated by the Gorda ridge (Embley and Wilson, 1992; Cotsonika, 2006; Stakes et al., 2006; Wanless et al. 2010). Two constructional domes ranging from 20 30 m in height and 200 500 m in width were sampled along these hooked ridges, the compositions of which range from high silica andesite to rhyodacite (62.2 66.9 wt.% SiO 2 ) (Cotsonika, 2006; Stakes et al., 2006; Wanless et al., 2010; Schmitt et al., 2011). Surrounding the andesitic rhyodacitic domes are ferrobasalts and FeTi basalts that are similar in composition to other lavas collected from the Cleft segment (Stakes et al., 2006; Schmitt et al., 2011). A more detailed account of the geology and tectonics of the C left se gment and dive T735 is presented by Stakes et al. (2006). Previous Studies Vogt and Johnson (1973) predicted the occurrence of Fe and Ti rich basalts along the GSC based on the presence of high amplitude magnetic (HAM) anomalies. This hypothesis led Byerl y et al. (1976) to test whether FeTi rich basalts were present along HAM zones in the GSC and to the discovery of MOR andesites and dacites. Since then, a number of studies have focused on the processes of differentiation responsible for the production of these high silica MOR lavas (Byerly et al., 1976; Byerly, 1980; Clague et

PAGE 19

19 al., 1981; Perfit et al., 1983; Juster et al., 1989; Hekinian et al., 1997; Haase et al., 2005; Wanless et al., 2010). Christie and Sinton (1981) suggested that the occurrence of ev ol ved MOR lavas ridge transform intersection. They hypothesized that propagation of a ridge through older and cooler lithosphere would lead to increased cooling rates and thus mo re extreme crystal fractionation. Low pressure fractional crystallization has been further substantiated by others (Byerly et al., 1976; Clague and Bunch, 1976; Schilling et al., 1976; Byerly, 1980; Fornari et al., 1983; Perfit and Fornari, 1983; Perfit et al., 1983) as rhyodacitic lavas, 75 87% fractional crystallization of a MORB parent is required (Byerly, 1980; Clague et al ., 1981; Perfit et al., 1983). However, closed system fractionation is unable to explain the observed 1983). Therefore, small extents of open s ystem behavior such as magma mixing or crustal assimilation may be required to account for these discrepancies. An alternative to extreme crystal fractionation in the production of MOR dacites is partial melting of oceanic crust. Experimental studies hav e demonstrated that dacitic melts can be produced by partial melting of hydrated gabbroic or basaltic crust at temperatures of 850 2004; Kvassnes and Grove, 2008; Wanless et al., 2010). Assimilation and partial melting of alt ered oceanic crust has been proposed to account for low Nb/La and high

PAGE 20

20 Cl/K ratios observed in andesites from the Pacific Antarctic ridge and in dacites from the EPR, GSC and JdFR (Haase et al., 2005; Wanless et al., 2010). Using energy constrained assimil ation frac tional crystallization models, Wanless et al. (2010) demonstrate d that major and trace element compositions of dacites from the on the EPR can be reproduced by >75% crystallization of a MORB parent magma and 5 20% assimilation of altered amphibole bearing oceanic crust. Wanless et al. (2010) also note d the strong similarities between major and trace element compositions of this study from dive T735 on the JdFR, suggesting that their models may also be applicable to these suites. Models describing the petrogenesis of MOR dacites (i.e. crystal fractionation, partial melting and assimilation of hydrated oceanic crust) have largely been b ased on glass chemistry. However, petrographic evidence from each of these suites suggests that a more complicated history involving magma mixing may have played an important role in their development. Basaltic xenoliths are ubiquitous in dacites from the composed of plagioclase, quartz, and rhyolitic glass occurring in dacites from the GSC and JdFR (Byerly et al., 1976; Perfit et al., 1983; Cotsonika, 2006;). Secondl y, pyroxenes from the GSC and JdFR dacites display both normal and reverse zoning, implying earlier crystallization in magmas that were more mafic or felsic than their host d to support magma mixing. For example, plagioclase xenocrysts in basaltic andesites contain FeTi basalt glass inclusions while those in andesites contain higher silica glass

PAGE 21

21 inclusions, suggesting that the crystals originated in magmas that were either mo re mafic or felsic than their host glass (Perfit et al. 1983).

PAGE 22

22 CHAPTER 3 ANALYTICAL METHODS Thirty nine lava samples were collected on dive T735, which explored the constructional volcanic terrain at the ridge transform intersection of the southern JdFR with the Blanco Fracture Zone. Glassy quenched rims of all samples were analyzed for major elements. Whole rock and mine ral chemistry were also measured for a representative group of samples. Table 3 1 provides a complete list of analyses performed for ea ch of the thirty nine samples collected. Major and minor elements were determined for glass and mineral samples using a JEOL 8900 Electron Microprobe. Minerals phases were analyzed on polished thin sections and natural glasses were analyzed on fresh, hand picked separates mounted and polished on glass slides. Care was taken to collect crystal free chips that were ultrasonically cleaned in a solution made from equal parts 2.5N reagent grade HCl and 30% H 2 O 2 to remove any surficial coatings (Cotsonika, 2006). Microprobe analyses were performed at the United States Geological Survey (USGS) in Denver, CO and Florida International University on four separate occasions from 2004 to 2012. On each occasion, USGS mineral standards were used to calibrate the microprob e analyses and secondary normalizations were completed using University of Florida glass standard 2392 2 and USGS standard glass GSC that also served as drift monitors during analysis. An accelerating voltage of 15 keV, beam current of 20 nA and a beam dia meter of 20 m were used for glass analyses. Mineral analyses were performed wi th a beam diameter of 1 m. Replicate analyses of standard glasses indicate that the analytical precision of major elements in glasses with concentration greater than 1 wt%

PAGE 23

23 is p recise to within 1 to 4 relative percent but becomes significantly greater (10 20%) with abundances below 0.2 wt%. Trace element contents of crystal free glass chips were analyzed at the University of Florida using an Element2 inductively coupled plasma mass spectrometer at medium resolution. USGS rock standards were used for calibration and University of Florida MORB standards 2392 2 and ENDV were used as drift monitors and to check for internal precision. Replicate analyses of USGS standard BHVO 1 in dicate the accuracy and precision for all elements is equal to or better than 5 relative percent. A more detailed discussion of sample preparation, dissolution procedures, standards and errors can be found in Cotsonika (2006) and Goss et al. (2010). So me of the more crystalline sectio ns of dacitic samples with substantial crystal and xenolith contents were analyzed by traditional X Ray Fluorescence techniques at Geo labs in Ottawa, Canada. A representative subset of thirteen glass samples was selected f or volatile analyses (Table 3 1). Samples were analyzed for H 2 O and CO 2 concentrations by Fourier transform infrared (FTIR) spectroscopy at the University of Oregon using methods described in Johnson et al. (2009) and Wanless et al (2011).

PAGE 24

24 Table 3 1 Che mical Analyses Performed on D ive T735 Samples. Glass major and mineral = electron microprobe, glass trace = Element2 ICP MS, whole rock = XRF, volatiles = FTIR spectroscopy Sample G1 G2 G3 G4 G5 G6 G7 G8 Glass Major Glass Trace Whole Rock Volatiles Minerals G9 G10 G11 G12 G13 G14 G15 G16 Glass Major Glass Trace Whole Rock Volatiles Minerals G17 G18 G19 G20 G21 G22 G23 G24 Glass Major Glass Trace Whole Rock Volatiles Minerals G25 G26 G27 G28 G29 G30 G31 G32 Glass Major Glass Trace Whole Rock Volatiles Minerals G33 G34 G35 G36 G37 G38 G39 Glass Major Glass Trace Whole Rock Volatiles Minerals

PAGE 25

25 CHAPTER 4 GLASS GEOCHEMISTRY Samples collected from dive T735 are discussed here in the context of those previously analyzed from the Cleft suite, which has been extensively sampled both on and off axis (Smith et al., 1994; Stakes et al, 2006). Basa ltic lavas from the Cleft suite are normal, incompatible element depleted MORB with MgO values ranging from 8.1 wt.% in the most primitive basalts to 3.8 wt.% in the more evolved basaltic andesites. A strong correlation between distance along the ridge and MgO values of Cleft lavas has previously been cited, generally decreasing from north to south (Smith et al ., 1994; Stakes et al. 2006) Major element variations of the Cleft suite resemble typical tholeiitic differentiation trends (Figure 4 1). Decreasing values of MgO are associated with increasing values of TiO 2 FeO T (total FeO as Fe 2+ ), Na 2 O, K 2 O and P 2 O 5 and decreasing values of SiO 2 Al 2 O 3 and CaO in the basaltic lavas (Figure 4 1). The trends are typically smooth; however, significant deviations fro m the curvilinear paths and greater variability occur in SiO 2 TiO 2 FeO T Al 2 O 3, Na 2 O, K 2 O and P 2 O 5 at ~ 6 to 4 MgO wt.% in basaltic lavas from Stakes et al. (2006). These inflections vary from the original curvilinear paths in the following ways: (1) val ues of SiO 2 Al 2 O 3 and K 2 O are higher at a given MgO (2) values of FeO T TiO 2, Na 2 O and P 2 O 5 are lower at a given MgO and (3) inflections of each oxide diverge toward lower MgO values or in the direction of the compositions of T735 ande sites and dacites ( Figure 4 1). Basalts and Basaltic Andesites Basalts and ferrobasalts from d i ve T735 span a limited range of compositions in comparison to the entire Cleft suite (Table 4 1). The MgO values (6.29 7.55 wt%) are

PAGE 26

26 intermediate, falling between the most primitiv e and evolved samples collected from the Cleft. Most major and minor elements (SiO 2 CaO, Na 2 O, K 2 O, P 2 O 5 and MnO) do not vary greatly at given MgO values. Two exceptions to this are FeO T and TiO2, which span a broad range from 11.3 12.9 wt.% and 1.67 2 .26 wt.% respectively at ~7 wt% MgO. Most major element oxides of dive T735 basalt samples plot centrally within the LLD defined by the entire Cleft suite at a given MgO. However, Al 2 O 3 and TiO 2 plot above and below other Cleft suite samples respectively, at a given MgO value. The major element variations of basaltic lavas can be modeled by relatively large amounts (> 55%) of low pressure (<100 MPa) fractional crystallization of a relatively primitive melt (Smith et al. 1994; Tierney, 2003; Stakes et al. 2 006). Decreasing values of Al 2 O 3 with decreasing values of MgO coupled with increasing values of CaO/Al 2 O 3 suggest that plagioclase or plagioclase + olivine were the first phases in the crystallization sequence (Smith et al. 1994; Cotsonika, 2006). Clinopy roxene likely begins to fractionate at ~7.5 wt% MgO, when CaO/Al 2 O 3 begins to decrease (Smith et al. 1994; Cotsonika, 2006). Volatiles (H 2 O and CO 2 ) were analyzed from two basalts collected on dive T735. Samples T735 G23 and T735 G35 have H 2 O values of 0. 27 and 0.17 wt% and CO 2 values of 118 and 101 ppm, respectively. Chlorine contents in these two samples and those collected by Smith et al. (1994) are less than 0.05 wt.%. The more evolved basaltic lavas (MgO < ~6 wt.%) collected by Stakes et al. (2006) ha ve significantly higher values, up to 0.31 wt.%, that increase with decreasing MgO (Figure 4 2 a ). Values of Cl/K 2 O in T735 basalts vary between 0.05 and 0.23 wt.%, increasing with decreasing MgO values (Figure 4 2b).

PAGE 27

27 Trace element values in T735 basalt sam ples (Table 4 2) are similar to those in differentiated basalts from the Cleft segment (Smith et al., 1994). They are typical normal incompatible element depleted MORB with negative Sr and Eu anomalies on primitive mantle normalized diagrams (Figure 4 3). Trace elements from T735 basalts show variations similar to those described by Smith et al. 1994, but extent to lower MgO values. Andesites and Dacites Dive T735 andesites and dacites appear to extend the tholeiitic differentiation trends to low MgO (1.94 0.6 wt.%) and high SiO 2 (62 67 wt.%) compositions. However, a clear break in the data set occurs between 2 to 3.8 wt.% MgO (Figure 4 1). Major element variations of SiO 2 CaO, Na 2 O, K 2 O and CaO/Al 2 O 3 follow and extend the trends defined by the basaltic la vas, whereas, TiO 2, FeO T and P 2 O 5 have marked inflection points, decreasing in abundance with decreasing MgO (Figure 4 1). The Al 2 O 3 trends in samples containing ~ 0 to 2 wt.% MgO generally parallel t hose from ~ 4.5 to 6.5 wt.% MgO. Whole rock Al 2 O 3 value s for the evolved lavas are significantly lower than their corresponding glass values and appear to follow the same curvilinear path produced by the basaltic glasses. Water and CO 2 concentrations were determined in a number of evolved samples (Table 4 1) Values of H 2 O and CO 2 range from 0.35 to 2.0 wt.% and 0.28 ppm to below detection limits, respectively. Lower values of CO 2 are likely related to degassing, given the highly vesicular nature of the samples, whereas higher concentrations of H 2 O may either be related to increasing degrees of evolution or to assimilation of hydrated crustal material (Wanless et al. 2011). Chlorine concentrations range from 0.41 to 0.61 wt.% (Figure 4 2a) This extends the curvilinear trend in Figure 4 2 a to values over 12 t imes

PAGE 28

28 that measured in the m ost chlorine rich T735 basalt. The Cl/K 2 O values are lower (0.4 to 0.5 wt.%) than those of several of the basaltic andesites from Stakes et al. (2006) which approach values of 0.8 wt.% (Figure 4 2b). Evolved lavas exhibit increas ed incompatible trace element abundances with decreasing MgO values (Figure 4 3) However, Sr values are similar to those of the basalts, showing little change with fractionation and thus producing large negative Sr anomalies on mantle normalized diagrams There are generally no crossing patterns indicating the lavas may be related by fractional crystallization (Figure 4 3). Sample G20 however, does exhibit crossing patterns in Zr and Eu which have similar values with the two andesites, G18 and G19.

PAGE 29

29 Table 4 1. T735 Major Element Glass Compositions Sample SiO 2 TiO 2 Al 2 O 3 FeO T MnO MgO CaO Na 2 O K 2 O P 2 O 5 Cl Total (wt.%) H 2 O CO 2 (ppm) G1 50.48 2.12 13.31 12.31 0.21 6.87 11.27 2.53 0.17 0.22 0.02 99.81 G2 50.39 2.06 13.32 12.3 0 0.2 0 7.04 11.18 2.51 0. 17 0.22 0.02 99.69 G3 50.42 2.12 13.33 12.32 0.21 6.94 11.17 2.54 0.17 0.2 0 0.02 99.74 G4 50.59 1.87 13.48 11.96 0.22 6.86 11.25 2.69 0.18 0.23 0.03 99.68 G5 50.47 2.09 13.38 12.21 0.2 0 6.94 11.24 2.56 0.17 0.22 0.02 99.8 0 G7 50.48 2.12 13.12 12.29 0.2 0 6.78 11.17 2.52 0.17 0.24 0.03 99.41 G8 50.38 2.1 0 13.13 12.22 0.2 0 6.92 11.13 2.54 0.17 0.21 0.02 99.33 G9 63.73 1.2 0 12.42 9.45 0.17 1.04 4.39 4.76 1.14 0.33 0.55 99.47 1.86 G10 63.98 1.17 12.43 9.35 0.17 1 .00 4.29 4.76 1.15 0.33 0.55 99.48 1.96 G11 63.36 1.22 12.38 9.87 0.19 1.03 4.43 4.79 1.1 0 0.4 0 0.51 99.56 1.73 G12 66.89 0.81 12.23 7.95 0.14 0.6 0 3.41 5.01 1.3 0 0.18 0.61 99.44 1.98 G13 62.82 1. 39 12.35 9.81 0.16 1.4 0 4.85 4.52 1.09 0.34 0.52 99.55 1.63 G14 62.9 0 1.33 12.36 9.8 0 0.19 1.38 4.8 0 4.57 1.08 0.37 0.53 99.59 0.35 224.88 G15 62.7 0 1.33 12.33 10.25 0.19 1.15 4.6 0 4.76 1.07 0.42 0.49 99.58 2 .00 G16 66.43 0.85 12.44 8.02 0.1 6 0.6 0 3.57 5.07 1.28 0.3 0 0.61 99.63 1.77 G17 64.33 1.21 12.56 9.16 0.18 0.76 4.03 5.04 1.17 0.35 0.55 99.63 G18 62.16 1.26 13.32 8.89 0.16 1.94 5.4 0 4.36 1.02 0.18 0.41 99.43 1.97 G19 61.95 1.34 13.59 8.94 0.16 1.73 5.38 4.62 1.03 0.17 0.42 99.65 2.01 G20 65.75 0.98 12.88 7.28 0.12 1.33 4.32 4.71 1.28 0.11 0.52 99.6 0 1.51 G21 50.63 2.26 12.89 12.7 0 0.22 6.68 11.11 2.58 0.18 0.25 0.02 99.8 0 G22 50.09 2.22 12.77 12.79 0.22 6.69 11.03 2.62 0.16 0.23 0.03 99.12

PAGE 30

30 Tabl e 4 1 Continued Sample SiO2 TiO2 Al2O3 FeOT MnO MgO CaO Na2O K2O P2O5 Cl Total (wt.%) H2O CO2 (ppm) G23 50.53 2.19 13.02 12.66 0.22 6.74 11.17 2.62 0.16 0.23 0.02 99.85 0.27 117.74 G24 50.54 2.22 12.91 12.78 0.21 6.67 11.11 2.57 0.17 0.23 0.02 99. 72 G25 50.46 2.22 13.02 12.72 0.22 6.68 11.1 0 2.63 0.17 0.2 0.03 99.73 G26 50.65 2.18 13.08 12.64 0.23 6.92 11.19 2.6 0.16 0.22 0.02 100.16 G27 50.46 2.13 13.05 12.44 0.24 7.05 11.19 2.53 0.16 0.19 0.02 99.76 G28 50.47 2.13 13.11 12.42 0.22 7.13 11.2 0 2.54 0.15 0.2 0.02 99.88 G29 50.64 2.13 13.04 12.9 0 0.24 6.82 10.94 2.62 0.16 0.23 0.02 100.02 G30 50.66 2.11 12.93 12.58 0.21 6.69 10.86 2.57 0.17 0.2 0.02 99.28 G31 51.57 1.67 13.38 11.3 0 0.2 0 6.78 10.98 2.65 0.2 0.16 0.04 99.22 G32 50.32 2.15 12.81 12.77 0.24 6.29 10.92 2.63 0.15 0.21 0.02 98.78 G34 50.46 1.61 13.6 0 10.78 0.19 7.43 11.81 2.45 0.12 0.16 0.01 98.93 G35 50.67 1.61 13.76 10.86 0.19 7.55 11.84 2.47 0.12 0.14 0.01 99.54 0.17 101 .00 G37 50.67 2.11 13 .05 12.76 0.23 6.68 10.95 2.63 0.15 0.18 0.02 99.71 G38 50.72 2.09 13.05 12.69 0.22 6.7 0 10.94 2.61 0.15 0.18 0.02 99.66

PAGE 31

31 Table 4 2 T735 Trace Element Glass Compositions Sample Sc V Cr Co Ni Cu Zn Ga Rb Sr Y G2 41.2 382 173 37.6 57.2 55.4 9 9.2 18 .0 1.2 102 50.4 G4 42 .0 340 188 38 .0 53.8 64.8 89.2 17.4 1.2 118 44.2 G5 41.8 391 180 38.5 59.7 56.2 101.9 18.3 1.2 104 51.2 G7 40.4 378 170 37.2 57.4 54.8 98.3 17.6 1.2 100 49.7 G8 41.1 384 173 37.8 59.3 55 .0 99.7 18 .0 1.2 102 50.4 G9 18.1 73 5 12.6 4.5 15.9 137.9 26.6 10.3 86 214.7 G10 21.1 109 21 15.9 10.4 23.2 136.2 25.9 9.4 90 197.2 G11 20.1 67 4 13.6 3.4 15.9 152.3 28 .0 10.6 93 230 .0 G12 15.9 48 2 10 .0 1.7 13.5 144.5 27.6 12 .0 80 240.4 G13 19.6 110 11 14.9 7.9 18.7 134.3 25.8 10.3 86 20 2.6 G14 19.9 113 11 15.2 8 .0 19 .0 134.6 26.1 10.1 87 201.2 G15 19.5 71 7 13.9 4.3 15.4 146 .0 27.3 9.9 93 212.6 G16 16.1 48 3 10.4 2.2 13.6 144.4 27.8 11.8 82 238.8 G17 18.2 66 6 12.3 3.7 14.2 142.4 27.2 10.7 88 220 .0 G18 22.6 178 44 19.1 18.9 27.5 116 .1 23.3 9.4 79 160 .0 G19 23.4 180 54 19.7 20.4 27.9 116.7 23.1 9.2 79 158.4 G20 17.1 123 32 14.1 13.8 21.7 113.4 24.3 12.7 67 194.2 G21 41.5 391 152 38.8 58.5 54.8 102.3 18.2 1.3 102 52.9 G22 41.1 389 148 38.2 55.2 54.7 102.7 18.2 1.3 102 53.4 G23 42 0 395 159 38.4 54.7 55.5 103 .0 18.6 1.3 105 53.1 G28 41.6 392 169 38 .0 55.9 55 .0 100.7 18.3 1.2 104 51.7 G29 41.1 388 101 38.3 45.6 53.6 103.4 18.3 1.3 109 52.3 G32 40.7 392 98 39.1 46.5 53.5 105.2 18.5 1.1 101 51.7 G35 40.5 330 287 38 .0 70 .0 63.3 84.7 16.6 0.7 101 37.6

PAGE 32

32 Table 4 2. Continued Sample Zr Nb Ba La Ce Pr Nd Sm Eu Gd Tb G2 135.8 4.47 15.1 0 4.93 15.69 2.51 13.8 0 4.68 1.51 5.46 1.17 G4 125.1 4.34 16.69 4.75 14.74 2.36 12.59 4.18 1.36 4.73 1.04 G5 136.9 4.59 15.44 5.02 15.94 2.54 13.99 4.7 2 1.52 5.51 1.19 G7 133.9 4.42 14.92 4.87 15.47 2.48 13.46 4.55 1.47 5.34 1.15 G8 134.7 4.49 16.23 4.98 15.63 2.49 13.78 4.64 1.48 5.34 1.17 G9 637.2 19.29 82.57 29.97 82.13 11.33 64.59 18.89 3.94 20.89 4.46 G10 585.3 17.4 0 77.4 0 26.79 74.5 0 10.43 59.1 0 17.38 3.67 19.08 4.11 G11 672 .0 20.1 0 87.69 30.59 84.39 11.84 68.05 20.19 4.2 0 22.01 4.75 G12 663.7 21.2 0 95.39 33.78 91.2 0 12.79 72.82 21.15 4.15 23.23 4.99 G13 590.1 17.82 83.83 28.28 77.5 0 10.8 0 61.17 17.96 3.67 19.72 4.2 0 G14 580.3 17.72 79.17 28 .15 77.34 10.74 61.05 17.73 3.65 19.55 4.16 G15 636.4 18.62 78.15 29.13 80.08 11.22 64.26 18.89 4.1 0 20.77 4.43 G16 671.6 21.01 93.54 33.47 91.83 12.6 0 71.67 20.97 4.14 22.82 4.91 G17 641.9 19.47 84.95 30.18 83.32 11.59 66.59 19.47 4.05 21.31 4.55 G18 440.8 14.09 70.55 22.6 0 61.4 0 8.51 47.36 13.86 2.81 15.1 0 3.27 G19 439.8 13.96 69.2 0 22.36 60.66 8.41 46.82 13.76 2.79 14.96 3.24 G20 396.8 17.03 88.7 0 29.38 78.75 10.71 59.18 16.98 3.01 18.4 0 3.97 G21 140.9 4.68 16.25 5.21 16.28 2.6 0 14.31 4.78 1.52 5. 48 1.2 0 G22 142 .0 4.76 16.5 0 5.24 16.39 2.61 14.55 4.87 1.54 5.52 1.21 G23 142.3 4.71 15.99 5.17 16.28 2.61 14.48 4.86 1.53 5.51 1.21 G28 137.8 4.59 16.17 5.06 15.89 2.54 14.01 4.73 1.5 0 5.38 1.19 G29 142 .0 4.68 16.48 5.19 16.44 2.63 14.4 0 4.82 1.53 5. 49 1.19 G32 134.3 4.14 14.65 4.67 15.04 2.44 13.52 4.64 1.48 5.34 1.18 G35 101.7 2.88 11.21 3.4 0 11.43 1.92 9.98 3.49 1.18 4 .00 0.88

PAGE 33

33 Table 4 2. Continued Sample Dy Ho Er Tm Yb Lu Hf Ta Pb Th U G2 7.74 1.8 0 5.13 0.69 4.97 0.77 3.8 0 0.284 0.372 0.27 0 0.111 G4 6.78 1.56 4.37 0.61 4.34 0.67 3.4 0 0.241 0.349 0.268 0.105 G5 7.82 1.83 5.21 0.7 0 5.01 0.77 3.83 0.28 0 0.373 0.276 0.111 G7 7.57 1.77 5.01 0.67 4.83 0.75 3.72 0.275 0.361 0.268 0.109 G8 7.66 1.79 5.11 0.68 4.94 0.76 3.73 0.269 0.397 0.272 0.11 2 G9 29.37 7.19 21.99 2.46 20.64 3.19 18.32 1.273 2.48 0 2.818 1.067 G10 27.02 6.61 20.16 2.29 19.23 2.97 16.83 1.142 2.291 2.544 0.97 0 G11 31.08 7.67 23.25 2.59 22 .00 3.41 19.06 1.346 2.627 2.83 0 1.074 G12 32.88 8.09 24.87 2.78 23.63 3.65 19.62 1.45 0 2 .915 3.285 1.24 0 G13 27.73 6.81 20.82 2.35 19.92 3.08 17.18 1.217 3.105 2.796 1.054 G14 27.59 6.77 20.67 2.34 19.55 2.99 16.77 1.165 2.347 2.725 1.03 0 G15 29.32 7.14 21.87 2.46 20.71 3.16 18.21 1.23 0 2.43 0 2.629 1.017 G16 32.59 8.02 24.58 2.75 23.37 3. 56 19.71 1.398 2.85 0 3.208 1.214 G17 30.13 7.4 0 22.67 2.55 21.52 3.29 18.68 1.302 2.59 0 2.85 0 1.089 G18 21.69 5.31 16.26 1.88 15.73 2.4 0 13 .00 0.962 2.141 2.325 0.866 G19 21.56 5.29 16.13 1.86 15.57 2.38 12.96 0.927 2.075 2.287 0.851 G20 26.49 6.53 19. 96 2.29 19.15 2.9 0 13.1 0 1.169 2.739 3.14 0 1.148 G21 7.95 1.86 5.26 0.7 0 5.16 0.8 0 3.89 0.274 0.398 0.287 0.118 G22 8.01 1.88 5.33 0.71 5.22 0.81 3.92 0.293 0.406 0.29 0 0.118 G23 8.02 1.87 5.31 0.7 0 5.21 0.81 3.89 0.271 0.389 0.285 0.117 G28 7.79 1.82 5.21 0.69 5.07 0.79 3.76 0.266 0.389 0.279 0.114 G29 7.94 1.85 5.22 0.7 0 5.16 0.79 3.89 0.275 0.414 0.287 0.115 G32 7.76 1.83 5.22 0.69 5.07 0.79 3.73 0.237 0.371 0.25 0 0.1 00 G35 5.88 1.35 3.73 0.54 3.72 0.58 2.82 0.15 0 0.248 0.173 0.069

PAGE 34

34 Figu re 4 1 Major element variation diagrams for dive T735 lavas and Cleft segment lavas from Smith et al. 1994 and Stakes et al. 2006. MgO is plotted against A) SiO 2 B) TiO 2 C) Al 2 O 3 D) FeO(t) E) CaO F) Na 2 O G) K 2 O and H) P 2 O 5

PAGE 35

35 Figure 4 1 Continue d

PAGE 36

36 Figure 4 2 Chlorine variations in dive T735 lavas and Cleft segment lavas from Smith et al. 1994 and Stakes et al. 2006 A.) MgO v Cl and B) MgO v Cl/K 2 O

PAGE 37

37 Figure 4 3 Prim i tive mantle normalized trace element diagram of T735 basalts (blue sym bols), andesites (purple symbols) and dacites rhyodacites (red symbols).

PAGE 38

38 CHAPTER 5 PETROGRAPHY AND MINE RAL CHEMISTRY Basalts Dive T735 basalts and ferrobasalts ar e slightly phyric containing < 1 0% crystals. They contain plagioclase and clinopyroxene pheno crysts (up to 2 mm in size) that are subhedral to euhedral and display subophitic cotectic intergrowth relationships. Smaller microphenocrysts < 1 mm typically have skeletal or swallowtail shapes, indicative of rapid quenching. FeTi oxides are absent as ph enocrysts in both the basalts and ferrobasalts collected. Analyses of plagioclase produce An values ranging from 64 to 69. Clinpyroxene compositions are augitic with Mg#s [Atomic Mg/(Mg+Fe 2+ )*100] of ~81. There is no evidence of xenolithic fragments or re sorbed crystals in the basaltic lavas Andesites and Dacites The presence of basaltic xenoliths and mingling between dark and light colored glass in the andesitic and dacitic lavas demonstrates that they have clearly been contaminated and possibly mixed. Their history is complex and descriptions of the dacitic lavas (hereby including andesites unless stated otherwise) are only generalized here due to their heterogeneous nature. They are moderately phyric, containing between ~ 10 30% microphenocrysts (<1 mm ) and phenocrysts (up to 2 mm). Glass is tan in color but ranges from dark brown to opaque in areas that have undergone significant amounts of devitrification or are in close proximity to basaltic xenoliths. Vesiculation is significant in the dacites, repr esenting up to 15% compared to < 2 3% in the basalts. The vesicles have rounded or elliptical shapes up to several cm in size, which are commonly flow aligned in the direction of the groundmass crystals.

PAGE 39

39 Plagioclase is absent from the groundmass which is d ominated by elongated and skeletal pyroxenes (Figure 5 1d).However, larger crystals of pyroxene, plagioclase, FeTi oxides, olivine, quartz, and zircon are present in smaller proportions. Several populations of crystal clots and basaltic xenoliths in the da cites are outlined below on s that may have only a thin rim of opaque glass surrounding them. Mineral populations of both individual crystals and those in crystal clots are also characterized here on the basis of their chemical compositions and textures, as described below. Microp henocrysts and Phenocrysts Pyroxene Microphenocrysts and phenocrysts of pyroxene are subdivided into three types on the and textural features. Pyroxene compositions range from high Ca diopside and hedenbergite to low Ca pigeonite and ferropigeonite (Figure 5 2). A representative list of chemical analyses from core rim pairs for each population is given in Table 5 1. Type 1 pyroxenes are normally zoned and are characterized by a Mg rich clinopyroxene or low Ca pyroxene core, with Mg#s ranging from 68 80 (Figure 5 3a). They are subhedral to euhedral, less than 1 mm in size, and lack resorption features. A thin (<5 microns) brown rim commonly surrounds colorless cores (Figure 5 1b). The rims are sharp and may be significantly more ev olved than the cores, with Mg#s as low as 42 (Figure 5 3a). However, their relatively small size makes it difficult to obtain precise analyses of the rim in many cases.

PAGE 40

40 Type 2 pyroxenes are reversely zoned and are characterized by a Fe r ich clinopyroxene o r Fe rich low Ca pyroxene core, with Mg#s ranging from 17 54 (Figure 5 3a). The crystals are generally similar in size to type 1 pyroxene, but may be as large as 2 mm. Most type 2 pyroxenes are anhedral and are commonly embayed with resorbed edges. Like t ype 1 pyroxenes, a sharp brown colored rim often surrounds the original core (Figure 5 1c). Rim compositions span a wide range from Mg#s of 22 73 with average Mg#s of ~50 (Figure 5 3a). However, the rims are always less evolved than their core compositi on. The brown colored rims surrounding type 2 pyroxenes may be wider (up to 20 microns) than those surrounding type 1. Wider reaction rims are sometimes patchy, containing both small amounts of glass and higher Mg# pyroxene. Some type 2 ferroaugites contai n exsolved lamellae of low Ca pyroxene that are <5 microns wide, suggesting that they have had a protracted cooling history. The third type of pyro xene characterized here occurs in the groundmass and is distinguished by its brown color and smaller grain s ize (<20 microns in width). It occurs as either euhedral crystals that are < 20 microns in size or as elongated skeletal grains up to 1mm in length but generally less than 5 10 microns in width (Figure 5 1d). Some of the elongated crystals are more approp riately termed crystallites, exhibiting a needle like morphology, often developing fasicular textures. The cores of type 3 pyroxenes may contain a colorless nuclei (1 than the rims which have ave rage Mg#s of 55 (Figure 5 3a). These rounded nuclei are similar to type 1 pyroxene in both composition and lack of color, suggesting that the nuclei may have originated from a relatively mafic magma. However, distinctions between analyses with colorless nu clei and those without are difficult due to their

PAGE 41

41 smaller grain size and difficulty of distinguishing the two using backscatter electron imaging. A number of analyses plot within the pyroxene miscibility gap (Figure 5 2).These analyses may be metastable p yroxenes that crystallized rapidly during quenching of the liquid (Mazzullo and Bence, 1976). Alternatively, this may reflect analysis of a combination of both pigeonite and augite, similar to pyroxenes analyzed in an desites from the Galapagos Spreading Ce nter (Perfit et al. 1983). This is clearly the case for type 2 pyroxenes where thin exsolved lamellae of ferropigeonite occur within the ferroaugites. Low Ca pyroxene has been analyzed in all pyroxene types (Figure 5 3b), however, analyses are rare and petr ographic distinctions cannot be m ade between clinopyroxene and low Ca pyroxenes in thin sections. Plagioclase Plagioclase compositions range from An 25 72 and are separated here into two distinct populations, neither of which appear as a groundmass phase. Although plagioclase spans a wide range of compositions, individual crystals are usually only weakly zoned. Representative plagioclase compositions are provided in Table 5 2. Type 1 plagioclase is ty pically less than 0.5 mm in length and euhedral to subh edral with cores that have An# > 60. They generally lack resorption features, similar to type 1 pyroxene. Normal zoning occurs in most crystals, however, reverse zoning may occur in crystals that span smaller ranges of An content (<10%) from core to rim (F igure 5 4). Type 2 plagioclases are generally larger than type 1 (up to 2 mm), anhedral, and have core compositions with An# < 53. They are reversely zoned, although they may show normal zoning if the core and rim vary by only small amounts of An content

PAGE 42

42 (Figure 5 4). Type 2 plagioclase is commonly resorbed and may contain melt inclusions that are similar in composition to the host glass. FeTi Oxides Most FeTi oxides analyzed are titanomagnetites with ilmenite being extremely rare and often too small to precisely analyze. A single titanomagnetite ilmenite pair has been analyzed from sample T735 G12 (the T735 div e number is removed from all pre ceeding sample names) yielding temperatures of ~800 titanomagnetites analyzed ar e composed of 19 24 TiO 2 wt.% and 68 75 FeO T wt.%. Some oxides appear to be embayed while others display skeletal textures. Individual populations could not be discerned on the basis of texture and chemical composition from the collected analyses. Olivine Olivine is rare in the dacites. It is generally < 1 mm in size, yellowish green, anhedral, and is commonly highly resorbed. It is fayalitic in composition ranging from Fo 7 Fo 27 with most analyses below Fo 15 (Figure 5 2). Zoning is not significant in olivi ne, although a thin reaction rim (Fo 27 ) mantles a Fo 7 core in one instance. Accessory Phases Quartz crystals up to 0.5 mm in size are present in thin section G20, which is one of the highest SiO 2 (65.75 wt.%) glasses analyzed They are rounded, highly res orbed, embayed and typically contain melt inclusions that are similar in composition to the dacitic lavas. Myrmekitic intergrowths occur in thin sections G19 and G20. These intergrowths occur as mm sized resorbed fragments. Small (<100 micron) euhedral or skeletal zircons are also present in the dacites. They are sometimes embayed and overgrown by euhedral rims (Schmitt et al., 2011). The zircons were previously

PAGE 43

43 analyzed by Schmitt et al. (2011), who obtained a U Th crystallization age of 29.34.8 ka and cr in zircon thermometery. Crystal Clots Four types of crystal clots are identified here, each of which is classified on the basis of their mineral assemblage, chemistry and texture. Type A clots are composed o f T1 pyroxene + T1 plagioclase +/ FeTi oxides that appear to be a late crystallizing phase. Individual clots are generally less than 1 mm and are irregularly shaped. A quenched rim of dark glass sometimes surrounds the clot of crystals. Unlike the individ ual microphenocrysts of T1 pyroxene and plagioclase, these crystals are typically skeletal and commonly demonstrate subophitic intergrowth textures (Figure 5 1a). Type B clots include T2 pyroxene + T2 plagioclase +/ FeTi oxides (Figure 5 1c). They are ge nerally larger than type A clots, up to a few mm in size. Unlike type A clots, textural features such as morphology and size cannot be used to distinguish individual microphenocrysts of T2 plagioclase and pyroxene from those in clots. Type C clots are def ined by T2 plagioclase and fayalitic olivine. Type 2 plagioclase compositions (An 30 35 ) are more restricted than those found in type B clots. A single clot (type D) of skeletal olivine (Fo 80 ), plagioclase (An 66 ) and clinopyroxene with Mg# of 77 is found in sample G20. The clot is ~ 2 mm in size with individual crystals up to 0.5 mm in 80 ) is rimmed by Fo 49 indicating it equilibrated in or reacted with a significantly more evolved liquid after initial crystallization. Basaltic Xenoliths Basaltic xenoliths are made up of plagioclase + augite +/ pigeonite +/ FeTi oxides + basaltic glass. They are easily distinguished from crystal clots by the significant amounts of opaque glass that surround the groundmass and pheno crysts (Figure 5 5a).

PAGE 44

44 Plagioclase (An 64 72 ) and pyroxene (Mg# 71 82) are compositionally similar to T1 pyroxene and T1 plagioclase (Table 5 2). Two types of basaltic xenolith are distinguished here on the basis of morphology and crystallinity. The first t ype is characterized by its rounded to subrounded morphology. It ranges from holohyaline to hypocrystalline with a quenched glassy rind surrounding each xenolith. In some cases the rounded basaltic xenoliths are intruded by lighter colored dacitic glass (F igure 5 5d). Grain size generally coarsens towards the center of the xenoliths. Subophitic crystal clots similar in mineralogy, size and shape to type A clots may be located in the xenoliths interior (Figure 5 5b). The second type of basaltic xenolith dist inguished here is characterized by its angular or jagged morphology (Figure 5 5c). It is usually larger, up to several centimeters in size, is generally aphyric and does not typically have a glassy rind surrounding its ri m.

PAGE 45

45 Table 5 1. Representative Pyro xene Analyses Sample 735 9 CPX3 1 735 9 CPX3 4 735 14 Pyx12 735 14 Pyx12 735 11 Pyx1 Type T1 T1 T2 T2 T3 Location Core Rim Core Rim Core Mg# 80.2 61.92 23.81 53.55 57.57 SiO 2 50.72 50.46 48.5 50.74 48.64 TiO 2 1.08 0.95 0.21 0.65 1.57 Al 2 O 3 3.39 1.99 0.26 1.56 3.28 FeOt 9.01 15.38 38.21 21.63 16.45 MnO 0.2 0.41 0.98 0.61 0.4 MgO 16.46 13.38 6.7 13.99 11.95 CaO 18.45 16.53 4.25 9.53 16.2 Na 2 O 0.26 0.27 0.04 0.15 0.33 K 2 O 0 0 0 0.01 0.01 P 2 O 5 0.01 0.08 0.03 0 0.07 Cr 2 O 3 0.2 0.07 0 0.01 0 NiO 0.03 0 0 Total 99.77 99.52 99.21 98.89 98.89 Sample 735 11 Pyx1 735 9 Pyx2 735 9 Pyx2 735 19 xeno 735 19 xeno Type T3 T3 T3 BX BX Location Rim Core Rim Core Rim Mg# 51.97 61.17 53.22 80.78 71.16 SiO 2 49.38 47.54 49.33 53.85 46.83 TiO 2 0.91 2.18 0.98 0.56 2.51 Al 2 O 3 1.55 3.96 1.89 1.84 7.67 FeOt 20 15.4 18.5 7.32 11.05 MnO 0.53 0.38 0.48 0.21 0.21 MgO 11.84 12.76 11.19 17.26 13.92 CaO 13.88 15.6 16.17 18.11 16.56 Na 2 O 0.19 0.32 0.27 0.26 0.4 K 2 O 0.02 0 0.02 0.01 0.03 P 2 O 5 0 0.1 0 .05 Cr 2 O 3 0.01 0 0.02 NiO 0.01 0 0.02 Total 98.32 98.24 98.9 99.42 99.17

PAGE 46

46 Table 5 2. Representative Plagioclase Analyses Sample 735 20 Plag1 735 20 Plag1 735 9 Plag10 735 9 Plag10 Type T1 T1 T2 T2 Location Core Rim Rim Core An# 68.86 7 0 46.49 43.39 SiO 2 50.86 50.02 55.93 57.54 TiO 2 0.05 0.05 0.03 0.04 Al 2 O 3 30.26 29.92 26.32 26.49 FeOt 0.52 0.58 0.49 0.38 MnO 0.01 0.02 0.01 MgO 0.19 0.23 0.04 0.04 CaO 14 14.15 9.49 9.11 Na 2 O 3.47 3.31 5.95 6.47 K 2 O 0.05 0.06 0.14 0.15 P 2 O 5 0.01 0.02 Cr 2 O 3 0.01 0.01 NiO 0.04 0.01 0.01 0 Total 99.45 98.37 98.42 100.24 Sample 735 20 Plag8 735 20 Plag8 735 20 Pl3BX 735 20 Pl3BX Type T2 T2 BX BX Location Core Rim Core Rim An# 31.39 28.88 69.75 66.47 SiO 2 60.56 60.98 50.8 4 51.47 TiO 2 0.06 0.1 Al 2 O 3 24.5 23.82 30.33 29.7 FeOt 0.33 0.33 0.66 0.84 MnO 0 0.02 0.01 0.01 MgO 0.02 0.2 0.22 CaO 6.5 6.01 14.23 13.48 Na 2 O 7.7 8.02 3.4 3.72 K 2 O 0.22 0.25 0.02 0.05 P 2 O 5 0.03 0.03 Cr 2 O 3 0.01 NiO 0.02 0.01 Total 99.82 99.43 99.81 99.65

PAGE 47

47 Figure 5 1 Plane polarized light images of characterized pyroxene types A) Image of type A crystal clot with T1pyroxene + T1 plagioclase, note the elongated fibers of pyroxene that have crystallized on plagiocla se B) Image of type A crystal clot with subhedral T1 pyroxene + T1 plagioclase + FeTi oxides Large white (clear) oblong features are vesicles C) Image of T2 pyroxene with brown rim + T2 plagioclase + FeTi oxides in type B clot D) Image of type 3 pyroxenes the red box indicates the rounded colorless nuclei discussed in text. A B C D

PAGE 48

48 Figure 5 2 Pyroxene and olivine compositions in T735 andesites and dacites A) All data B) Type 1 pyroxene C) Type 2 pyroxene D) Type 3 pyroxene. A

PAGE 49

49 Figure 5 2 C ontinued B C

PAGE 50

50 Figu r e 5 2 Continued D

PAGE 51

51 Figure 5 3 Pyroxene core to rim chemical variations in A) Mg# and B) CaO for type 1, 2 and 3 pyroxenes. Open squares indicate rims and closed squares are cores A

PAGE 52

52 Figure 5 3 Continued B

PAGE 53

53 Figure 5 4 Plagioclase core to rim An # variations for individual crystals of type 1 and 2 plagioclases. Open squares indicate rims and closed squares are cores.

PAGE 54

54 Figure 5 5 Plane polarized light images of basaltic xenoliths. A) Type 1 rounded basaltic xenolith with chilled margins B) Type 1 basaltic xenolith with coarser crystals in its interior and C) Type 2 basaltic xenolith with a more angular morphology and jagged margins. D) Image of type 1 basaltic xenolith being intruded by dacitic glass. A B

PAGE 55

55 Figure 5 5 Continued C D

PAGE 56

56 CHAPTE R 7 DISCUSSION Fractional Crystallization southern JdFR have yielded lavas ranging from relatively primitive basalt to rhyodacite that follow typical tholeiitic differentiation trend s (Perfit et al. 1983; White et al. 2009; Wanless et al. 2010) (Figure 7 1). As discussed by Wanless et al. (2010) the major and trace element compositions of these suites are remarkably similar. Although ther e is a compositional gap in JdFR data from ~ 2 to 4 MgO wt.%, there is evidence suggesting that some of the southern JdFR lavas may be products of extreme crystal fractionation as has been concluded for other MOR dacitic suites For example, high amounts (~ 75 85%) of crystal fractionation have been p roposed to explain the occurrence of high silica 1983; Perfit et al. 1983, Juster et al. 1989). Furthermore, pyroxenes from the southern JdFR dacites produce trends similar to large igneous intrusions like Bushveld and Skaergaard complexes, that increase in Fe content as fractional crystallization proceeds (Taylor, 1964; Atkins 1969; Nwe 1975) (Figure 7 2). In particular, low Ca pigeonite is absent beyond Fs 65 similar to bot h the Skaergaard and Bushveld suites (Figure 7 2). However, fractional crystallization has been extensively modeled by Wanless et al. (2010) who show that although ~75 85% crystallization of a ferrobasaltic parent can produce most major element trends, it does not produce all of the chemical variations found in the dacitic lavas. In particular values of K2O, Al 2 O 3 and Cl are higher and P2O5 lower in the dacitic magmas than the predicted values (Wanless et al. 2010). Furthermore, incompatible trace elements require even greater degrees (>90%) of

PAGE 57

57 fractional crystallization, suggesting that other processes must be taking place. Petrographic evidence indicates that magma mixing may play a role, which is evaluated below. Evidence Supporting Magma Mixing Severa l petrographic features in the dacitic lavas suggest that magma mixing may have played an important role in their evolution. These features include (a) the presence of basaltic xenoliths, mafic clots and mafic xenocrysts, (b) resorbed mineral phases that a re in disequilibrium with their host glass, (c) olivine reaction rims (Fo 80 to Fo 49 ) in type D clots and (d ) complexly zoned pyroxenes with rim values that approach intermediate compositions between the two end members Two types of basaltic xenoliths and two mafic crystal clots (types A and D) were incorporated into the dacitic lav as. Type 1 xenoliths, exhibit rounded morphologies, chilled margins and decreasing grain size from the center of the xenolith to the rim, which indicate that they were liquid pr ior to being incorporated into the dacites. The irregular shapes, jagged edges, and lack of a chilled margin in type 2 basaltic xenoli ths indicate that they may instead be pieces of wall rock that have been mechanically transferred and partially assimilate d into the dacitic magma. Type A crystal clots are similar in both mineralogy and texture (e.g. acicular pyroxene, swallow tail plagioclase) to the coarser crystals located at the cores of some type 1 basaltic xenoliths. It is likely that type A clots orig inated from the same source as the type 1 basaltic xenoliths based on their similar chemistry and textural features. The presence of skeletal olivine in type D crystal clots suggests that more than one magma type was involved in mixing with sample G20.

PAGE 58

58 The more evolved mineral phases (e.g. iron rich T2 pyroxene, sodic T2 plagioclase, fayaltic olivine, FeTi oxides and quartz) commonly display resorbed margins. The effect of assimilating lower temperature, more evolved mineral phases into high temperature (le ss evolved) magmas was first described by Bowen (1928). Lower temperature phases will react by dissolving in higher temperature magma. If the mineral phase is saturated within a magma a reaction rim may develop and the new rim will be compositionally equiv alent to that which is in equilibrium with the magma (Bowen 1928). Type two (T2) pyroxenes demonstrate this phenomenon, their corroded cores typically are surrounded by a reaction rim that is less evolved (higher Mg#) and occasionally has a sieve like text ure within the core Resorption may also occur due to changes in pressure. Plagioclase for example will become unstable if it undergoes decompression and has a more sodic composition than that which is stable in the melt (Pearce et al. 1987). Although d ecompression may account for resorption of plagioclase it does not easily explain resorption features in other mineral phases like pyroxene, olivine, or quartz whose stabilities are less sensitive to small changes in pressure. Higher temperature xenocryst s do not dissolve in lower temperature melts (Bowen, 1928). Instead they will continue to grow if the phase is saturated in the melt, rimmed by the composition which is in equilibrium with the melt (Bowen 1928, Shelley 1983). This is consistent with T1 pyr oxenes, which lack resorption features and are typically rimmed by more Fe rich compositions. Unlike T1 pyroxene, T1 plagioclase does not appear to have a reaction rim. However, plagioclase does not appear to be a stable phase in the dacitic lavas because it does not appear as a microphenocryst or

PAGE 59

59 phenocryst of the groundmass. Recent experimental studies on dacitic lavas collected from the Pacific Antarctic Ridge, with similar compositions to dacites in this study, show that p lagioclase crystallizes after c linopyroxene in a water saturated system a nd may include mixing driving plagioclase off the liquidus, suppression of plagioclase due to high H 2 O or possibly rapid quenching prohibiting nucleation of plagioclase in the more viscous dacites. However, the lack of zoning in plagioclase relative to the pyroxenes suggests that kinectics has not played a major role and that plagioclase is instead not saturated in the melt. The documented compositional variations in pyroxene p rovide a compelling argument in support of magma mixing. The cores of T1 and T2 pyroxenes appear to be bimodal, with average Mg#s of 75 and 33, respectively. However, the rims of T1 and T2 pyroxenes converge with those of intermediate composition T3 pyroxe nes at Mg# ~ 55 consistent with being in equilibrium with several of the dacitic glasses (Figure 7 3b). T2 pyroxenes crystallized from a melt more evolved than that of the dacites. The phase chemical data (Figure 7 3a) also indicate that T3 cores are more magnesian than their rim compositions, similar to those of T1 pyroxenes. The colorless nuclei of many T3 cores and their relatively magnesian compositions indicate that some T3 cores originally crystallized in basaltic magmas. Mineral Melt Equilibrium The petrographic and chemical arguments supporting magma mixing discussed above favor the interpretation that many of the mineral phases are xenocrysts. The minerals that display resorption features (T2 pyroxene, T2 plagioclase, fayalitic olivine, FeTi ox ides and quartz) likely crystallize d in a magma even more evolved than the

PAGE 60

60 dacitic host glass. An attempt to constrain the composition of that end member magma is made here using mineral melt equilibria. The Mg# of melts which are in equilibrium with t he cores and rims of T1, T2, and T3 pyroxenes can be constrained using mineral melt equilibria. The Fe/Mg K D values for clinopyroxene liquid are relatively constant in basalts and andesites, with average values of 0.27 0.03 (Grove and Juster, 1989; Kinzl er and Grove, 1992). These values are less well constrained for high silica dacites and rhyolites, however, this value is believed to increase to > 0.35 in magmas with > 5 wt % H2O (Gardner et al. 1995; Grove et al. 2003; Frey and Lange 2011). Although the dacitic magmas are relatively hydrous for MOR lavas, they are generally < 2wt% H2O, thus typical Fe/Mg K D values of 0.27 0.03 for clinopyroxene liquid equilib ria are used here (Figure 7 4). Pyroxene cores and rims that lie within the error envelope in F igure 7 4 are considered to be in equilibrium, while those that lie outside are in disequilibrium with their host glass. The cores of the most primitive T1 pyroxenes (Mg# ~ 80) and several of the pyroxenes from basaltic xenoliths are in equilibrium with m low as 47 using a Kd value of 0.24 and as high as 58 using a Kd value of 0.30 (Figure 7 4). This is consistent with the ferrobasalts collected from dive T735 which have a mean Mg# of 52.5. Most of the pyroxene rim compositions lie within the equilibrium 13.1. T2 pyroxene cores that lie below the equilibrium boundaries are in equilibrium with and G16, which are the most

PAGE 61

61 respectively. While the most evolved T735 lavas collected may be in equilibrium with some T2 cores, a more evolved melt is required to explain the l ower Mg# cores. Fayaltic olivine xenocryst compositions can be used to further assess the range of equilibrium melt compositions required for the evol ved mineral phases. The Fe/Mg K D between olivine and liquid has a constant value of 0.30 0.03 that is independent of changes in composition and temperature (Roeder and Emslie, 1970; Ussler III and Glazer, 1989). Fayaltic olivine Mg#s ranging from ~2 to 29 are in equilibrium with melt Mg#s of ~ 1 to 11, spanning the entire range of Kd values from 0.27 to 0 .33 (Figure 7 5). This is consistent with the predictions of a more evolved end member, as illustrated by clinopyroxene liquid equilibria calculations. However, the olivine compositions require an even more evolved end member (possibly as low as Mg# ~ 1) a nd are too Fe rich and or Mg poor to be in equilibrium with any of the dacitic glasses. This supports the high amounts of resorption and general scarcity of fayalitic olivine. Furthermore, it is only found in lavas with more than 63 wt.% SiO2 and Mg# < 17. 9 Magma Mixing Although physical and mineralogic evidence indicate magma mixing clearly played a role in the development of the andesites and dacites, the extent to which mixing is responsible for their chemical variability and whether they can be re lated through mixing of only two end members has not yet been estab lished. Differences in source s degrees of partial melting of these sources and/ or extents of fractionation can produce a wide range of evolved end member compositions, leading to an array of possible mixing lines. Furthermore, the extent of fractional crystallization of the mafic end member may also lead to an array of potential mixing lines. This appears to be true for several basaltic andesites from Stakes et al. (2006), which diverge fro m the main LLD between

PAGE 62

62 ~ 6 to 4 MgO wt.%. Because dacitic glasses and their whole rock equivalents follow fairly linear chemical trends and their trace element abundance patterns are very similar (Figure 7 6), it is likely that heterogeneities and varying degrees of partial melting play only a minor role in their development and that the evolved end member melts had a narrow range of compositions. indicates they must have origin ated from a lower Mg# melt than the most evolved rhyodacite collected from dive T735 (G12) with Mg# of 13. Since the composition of sample G12 is known and because it generally follows the main linear trend produced by the dacitic lavas in most variation d iagrams, (Figure 7 6) it is used here as the evolved end member composition. If the dacitic lavas can be related through mixing, the most evolved end member (Mg# ~ 1) can be constrained by extending mixing lines to lower MgO and FeO(t) values that approach Mg# of ~ 1. Basaltic compositions ranging from 3.76 to 7.55 MgO wt.% are evaluated here to constrain the mafic end member compositions involved in mixing. Basalts surrounding the dacitic constructional mounds are an obvious choice for the potential mafi c end members 69) are similar to those found in the basaltic xenoliths (Mg# = 71 82 and An# = 64 72) The most mafic (G35), evolved (G32) and an intermediate (G7) ferrobasalt sample from dive T735 were chosen for mixing considerations, ranging from 6.29 to 7.55 MgO wt.%. The three mixing lines separate the evolved T735 lavas into three distinct groups (a) the main group of dacites which require mixing with a more evolved composition (b) the two andesite s G18 & G19, whose trends are conflicting on variation diagrams (i.e. do not fall

PAGE 63

63 on the same mixing lines in multiple plots) and (c) G20 which requires mixing with a more primitive magma (Figure 7 6). Sample G20 is the only sample containing type D clots that include skeletal olivine (Fo80), plagioclase (An66) and clinopyroxene (Mg# = 77), supporting interaction with a more primitive magma than the T735 basalts. Because the two andesites (G19 & G20) do not follow the main linear trend and mixing lines requ ire conflicting mafic end member compositions (depending on the variation diagram used) it is unli kely that the two andesites were produced along mixing lines through sample G12. In order to successfully model mixing between sample G12 and basaltic magm as to produce the linear trend of the dacites, the mafic end member composition must be more evolved than those collected from dive T735. Therefore, samples T460 G24, T736 G 29 and T736 G11 from other areas in the southern Cleft segment are also considered as possible end member compositions. Dives T460 and T736 are north of dive T735 but are still located within the southernmost portion of the Cleft segment. Samples T460 G24, T736 G29 and T736 G11 were collected from the Crestal Boundary Ridge described by Stakes et al. (2006). Their MgO values range from 3.76 to 5.30 wt.% with the most evolved sample (T736 G11) being a basaltic andesite. Mixing lines become increasingly more successful at intersecting the dacitic compositions as the degree of evolution in t he basaltic lavas increases, with FeTi basalts and basaltic andesites providing the best fit (Figure 7 6). This is true in all cases except for Al 2 O 3 (Figure 7 7) which cannot successfully be modeled by two component mixing of any possible Cleft end member composition.

PAGE 64

64 High Al 2 O 3 values in MOR dacites have previously been attributed to the preferential assimilation of amphibole, which can be found in hydrated oceanic crust (Wanless et al. 2010). This is supported by high Cl/K2O and H2O values in the dacit ic lavas that also cannot be explained by fractional crystallization alone. Dissolution of amphibole during mixing cannot be ruled out, however, amphibole has not been observed in any of the MOR dacites. An alternative hypothesis is that the high Al 2 O 3 val ues are related to crystallization of pyroxene in the mixed dacitic melts. Whole rock compositions have lower Al 2 O 3 values than their glasses and they lie below the proposed mixing lines. Because basaltic xenoli ths will either increase the Al 2 O 3 further or decrease them by only minor amounts it is unlikely that they a re responsible for the lower Al 2 O 3 values of the whole rocks. However, Al 2 O 3 poor xenocrysts or the abundant pyroxene in the groundmass may play a role. The influence of xencorysts should be s mall compared to T3 pyroxene, which may account for up to 20% of some thin sections. Mass balan ce calculations suggest that Al 2 O 3 MgO and FeO(t) variation between glass and whole rock samples can be attributed to ~ 5 to 7% crystallization of T3 pyroxene ( see vectors in Figures 7 6 and 7 7). Although several thin sections contain > 7% pyroxene, this number may be influenced by (a) the wide range of crystallinity (b) minor influence of xenocrysts and (c) the presence of basaltic xenoliths. Therefore, it is d ifficult to quantify the amount of pyroxene crystallization necessary for the increased values but pyroxene crystallization is consistent with observed differences between the whole rock and glass compositions. Petrogenesis Wanless et al. (2010) hypothes ize that extreme crystal fractionation of MORB and partial melting and assimilation of altered basaltic or gabbroic crust is responsible for

PAGE 65

65 the development of MOR dacites. An alternative model is proposed here that accounts for both the highly evolved min eral phases in the JdFR dacites and evidence supporting magma mixing. Plagiogranite is the most likely source of the highly evolved phases, supported by the presence of myrmekite and exsolved pyroxene that are produced only by slow cooling. Intrusion of ba saltic magma would supply enough heat to cause partial melting of plagiogranite, formin g a rhyolitic melt. This rhyolitic melt may then mix (10 40%) with the highly fractionated FeTi basalt or basaltic andesites to produce dacitic magma The similarity bet ween pyroxene and plagioclase from type 1 basaltic xenoliths and those found in ferrobasalts indicates that a second injection of ferrobasaltic magma remobilized the dacitic magmas but did not extensively mix with them. This is also supported by mingling b etween ligh t and dark glasses in many instances (Figure 5 5d) High Cl/K2O in MOR dacites has been attributed to melting of hydrothermally altered crust or mixing with brines (Wanless et al. 2010, 2011). Wanless et al. (2011) favor hydrothermally altered c rust over assimilation of saline brines because it supports the major and trace element compositions of the dacitic lavas. However, partial melting of a plagiogranite source could also explain the trace element enrichments and crystallization of p yroxene m ay explain the high Al 2 O 3 wt.%. Therefore, either a hydrothermally altered plagiogranite source or contamination of saline brines may attribute to the high Cl/K2O in the dacitic lavas.

PAGE 66

66 Figure 7 1 Major element variations al. 2006 and this study). MgO is plotted against A) SiO 2 B) TiO 2 C) Al 2 O 3 D) FeO(t) E) CaO F) Na 2 O G) K 2 O and H) P 2 O 5

PAGE 67

67 Fi gure 7 1 Continued

PAGE 68

68 Figure 7 2 Pyroxene quadrilateral comparing T735 pyroxene trends with other tholeiitic suites. These suites include the GSC (Perfit et al. 1983 and Perfit unpublished) lavas and the Skaergaard (Nwe, 1975) and Bushveld (Atkins, 196 9) complexes. Several pyroxenes plot within the pyroxene miscibility gap; possible reasons are discussed in the text.

PAGE 69

69 Figure 7 3 Histogram s of pyroxene core and rim Mg#s. Analyses of type 1, 2, and 3 pyroxenes in A) cores and B) rims A B

PAGE 70

70 Figure 7 4 Pyroxene e quilibria diagram based on Fe Mg partitioning between clinopyroxene and liqud. The dashed lines represent the outer Kd boundaries of 0.24 and 0.30 with the solid line representing a Kd of 0.27 Pyroxene cores generally fall outside of the equ ilibrium envelope, however, many of the rims appear to be in equilibrium.

PAGE 71

71 Figure 7 5 Olivine e quilibria diagram based on Fe Mg partitioning between olivine and liquid. The dashed lines represent the outer Kd boundaries of 0.27 and 0.33 with the solid line representing a Kd of 0.3.

PAGE 72

72 Figure 7 6 Mixing lines between sample G12 and various basaltic lavas for MgO vs. A) SiO 2 B) TiO 2 C) FeO(t) D) Na 2 O E) K 2 O and F) P 2 O 5. The black arrow in C) indicates the direction of pyroxene fractionation. A

PAGE 73

73 Figure 7 6 Continued B

PAGE 74

74 Figure 7 6 Continued. C

PAGE 75

75 Figure 7 6 Continued D

PAGE 76

76 Figure 7 6 Continued D E

PAGE 77

77 Figure 7 6 Continued F

PAGE 78

78 Figure 7 7 Mixing lines between sample G12 and various basaltic lavas for MgO (wt.%) vs. Al 2 O 3 (wt.%) The black arrow ind icates the direction of pyroxene fractionation

PAGE 79

79 CHATER 8 CONCLUSIONS Dacitic lavas collected from mid ocean ridge axis discontinuities have recently been attributed to extreme crystal fractionation with small amounts (5 20%) of partial melting and assi milation of hydrated oceanic crust. However, dacitic lavas from the ridge transform intersection of the Blanco Fracture Zone and southern Juan de Fuca Ridge indicate that magma mixing plays an important role in their evolution. The phyric nature and preser vation of highly evolved mineral phases make this suite of dacites source is required to explain the presence of highly evolved mineral phases (e.g. fayalitic olivine, quart z, Fe rich pyroxene) and the presence of exsolved pyroxene and myrmekite. Mixing of 10 40% of a highly evolved basaltic magma such as FeTi basalt or basaltic andesite and partial melts from a plagiogranite source are able to reproduce most major element co mpositions. High Al 2 O 3 values are attributed to the crystallization of pyroxene in the absence of plagioclase. In order to explain the high Cl/K 2 O ratios in the dacitic lavas; the plagiogranite source must either be hydrothermally altered or some assimilat ion must take place with saline brines in the oceanic crust

PAGE 80

80 LIST OF REFERENCES Journal of Petrology 10 ( 2 ): 222 249. d water saturated melting Journal of Petrology 32: 365 401. Bowen, N.L. (1928). The Evolution of the Igneous Rocks. Dover Publications, New York. Byerly, G. R., Melson, W. G., & V ogt, P. R. (1976). "Rhyodacites, andesites, ferro basalts and ocean tholeiites from the Galapagos spreading center." Earth Planet. Sci. Lett 30 : 215 221. the Galapagos Sp Journal of Geophysical Research 85 ( 87 ): 3,797 3,810. Canales, J. P., Detrick, R. S., Carbotte, S. M., Kent, G. M., Diebold, J. B., Harding, A., Babcock, J., Nedimovic, M. R. & Ark, E. V. (2005). "Upper crustal structure and axial topograph y at intermediate spreading ridges: Seismic constraints from the southern Juan de Fuca Ridge." Journal of Geophysical Research 110 Chadwick, J., Perfit, M., Ridley, I., Jonasson, I., Kamenov, G., Chadwick, W., Embley, R., Le Roux, P., & Smith, M. (2005) "Magmatic effects of the Cobb hot spot on the Juan de Fuca Ridge." Journal of Geophysical Research 110 Christie, D. M. & Sinton, J. M. (1981). "Evolution of abyssal lavas along propagating segments of the Galapagos spreading center." Earth Planet. Sci Lett 56 : 321 335. Clague, D. A. & Bunch, T. E. (1976). "Formation of Ferrobasalt at East Pacific Midocean Spreading Centers." Journal of Geophysical Research 81 ( 23 ): 4247 4256. Clague, D. A., Frey, F. A., Thompson, G., & Rindge, S. (1981). "Minor and t race element geochemistry of volcanic rocks dredged from the Galapagos spreading center: Role of crystal fractionation and mantle heterogeneity." Journal of Geophysical Research 86 : 9469 9482. Cotsonika, L. (2006). Petrogenesis of Andesites and Dacites fro m the Southern Juan de Fuca Ridge. Geology Gainesville, University of Florida: 188 p. Journal of Geophysical Rese arch 90 : 727 744.

PAGE 81

81 hot spot Journal of Geophysical Research 86 : 11,747 11,750. Embley, R. W., Hammond S., Malahoff, A., Ryan, W. B. F., Crane, K., & Kappel, E. Eos Trans. AGU 64 : 853. Embley, R. W., Chadwick, W., Perfit, M. R., & Baker, E. T. (1991). "Geology of the northern Cleft segment, Juan de Fuca Ridge: Recent lava flows, sea floor spreading, and the formation of megaplumes." Geology 19 : 771 775. Zone Marine Geophysical Res earches 14 : 25 45. Embley, R. W., Chadwick, W., Perfit, M. R., Smith, M. C., & Delaney, J. R.(2000). "Recent eruptions on the CoAxial segment of the Juan de Fuca Ridge: Implications for mid ocean ridge accretion processes Journal of Geophysical Research 105 : 16501 16525. Fornai, D. J., Perfit, M. R., Malahoff, A., & Embley, R. (1983). "Geochemical studies of abyssal lavas recovered by DSRV Alvin from Eastern Galapagos Rift, Inca Transform, and Ecuador Rift 1. Major element variations in natural glasses a nd spacial distribution of lavas." Journal of Geophysical Research 88 : 10,519 10,529. Frey, H. M. & Lange, R. A. (2011). "Phenocryst complexity in andesites and decites from the Tequila volcanic field, Mexico: Resolving the effects of degassing vs. magma m ixing." Contrib. Mineral. and Petrol 162 : 415 445. constraints on pre eruptive water contents and changing magma storage prior to explosive eruptions of Mount St. Helens volc Bull Volcano 157 :1 17. Goss, A. R., Perfit, M. R., Ridley, W. I., Rubin, K. H., Kamenov, G. D., Soule, S. A., Fundis, A., & Fornari, D. J. (2010). "Geophysical of lavas from the 2005 2006 eruption at the East Pacific Rise, 9'46'N 9'56'N: Implication s for ridge crest plumbing and decadal changes in magma chamber compositions." Geochemistry, Geophysics, Geosystems 11 ( 5 ): 1 35. Ca pyroxene stability and olivine pyroxene liquid equil ibria at 1 atm in natural basaltic and Contrib. Mineral. Petrol 103: 287 305. Grove, T.L., Elkins Tanton, L.T., Parman, S.W., Chatterjee, N., Muntener, O., & melting controls on calc Contrib. Mineral. Petrol 145 : 515 533.

PAGE 82

82 Haase, K. M., Stroncik, N. A., Stroncik, N. A., Hekinian, R., & Stoffers, P. (2005). "Nb depleted andesites from the Pacific Antarctic Rise as analogs for early continental crust." Geology 33 ( 12 ): 921 924. N., & Maia, M. (1997). "Intraplate versus ridge volcanism on the Pacific Antarctic Ridge near 37'S 111'W." Journal of Geophysical Research 102 : 12,265 12,286. Hekinian, R., Stoffers, P., Ackermand, D., Revillon, S., Maia, M., & Bohn, M. (1999). "Ridge hotspot interaction: the Pacific Antarctic Ridge and the foundation seamounts." Marine Geology 160 : 199 223. Jo hnson, E.R., Wallace, P., Delgado Granados, H., Manea, V.C., Kent, A.J.R., Bindeman, I.N., Donegan, C.S. (2009). Subduction related volatile recycling and magma generation beneath Central Mexico: insights from melt inclusions, oxygen isotopes and geodynam ic models Journal of Petrology 50 1729 1764. Juster, T. C., Grove, T. L., & Perfit, M. R. (1989). "Experimental constraints on the generation of FeTi basalts, andesites, and rhhydacites at the Galapagos spreading center, 85'W and 95'W." Journal of Geoph ysical Research 94 : 9,251 9,274. Steady State Rift Valley Along Northeast Pacific Spreading Centers: Evidence from Sea MARC Journal of Geophysical Research 91 (B14): 13,925 13,940. ocean ridge basalts: 1. Journal of Geophysical Research 97 : 6,885 6,906. Koepke, J., Feig, S. T., Snow, J., & Freise, M. (2004). "Petrogenesis of oceanic plagiogranites by partial melting of gabbros: an experimental study." Contrib. Mineral. Petrol 146 : 414 432. Kvassnes, A. J. S. & Grove, T. L. (2008). "How partial melts of mafic lower crust affect ascending magmas at oceanic ridges." Contrib. Mineral. Petrol 156 : 49 7 1. Nature 322 : 422 429 Journa l of Geophysical Research 81 : 4,327 4,351. Normark, W. R., Morton, J. L., Koski, R. A., Clague, D. A., & Delaney, J. R. (1983). Geology 11 : 158 163. Normark et al. (1986). southern Juan de Fuca Ridge: Preliminary observations from the submersible Geology 14 : 823 827.

PAGE 83

83 Normark, W. R., Morton, J. L., & Ross, S. Journal of Geophysical Research 92 (B11): 11,283 11,290. of the Skaergaar Contrib. Mineral. Petrol 49 : 285 300. Pearce, T. H., Russel, J. K., & Wolfson, I. (1987). "Laser interferences and Normarski interference imaging of zoning profiles and plagioclase phenocrysts from the May 18, 1980 eruption of Mount St. Helens, Washington." Am. Mineral 72 : 1,131 1,143. Perfit, M. R. & Fornai, D. J (1983). "Geochemical studies of abyssal lavas recovered by DSRV Alvin from eastern Galapagos Rift, Inca Transform, and Ecuador Rift 2. Phase chemistry and crystal lization history." Journal of Geophysical Research 88 : 10,530 10,550. Perfit, M. R., Fornari, D. J., Malhoff, A., & Embley, R. W. (1983). "Geochemical studies of abyssal lavas recovered by DSRV Alvin from eastern Galapagos Rift, Inca Transform, and Ecuador Rift 3. Trace element abundances and petrogenesis." Journal of Geophysical Research 88 : 10,551 10,572. Perfit, M. R., and W. W. Chadwick, Jr., Magmatism at mid ocean ridges: Constraints from volcanological and geochemical investigations, in Faulting and Magmatism at Mid Ocean Ridges Geophys. Monograph 106 : 59 116. Am Geological Society American Bulletin 72 : 1,267 1,270. Regelous, M., Niu, Y., Wendt, J. I., Batiz, R., greig, A., & Collerson, K. D. (1999). "Variations in the geochemistry of magmatism on the East Pacific Rise at 10' 30'N since 800 ka." Earth Plant. Sci. Lett 168 : 45 63. Contr. Mineral. and Petrol 29 : 275 289. ocean ridge thermal and magmatic s Earth Planet. Sci. Lett. 260 : 257 276. Schmitt, A. K., Perfit, M. R., Rubin, K. H., Stockli, D. F., Smith, M. C., Cotsonika, L. A., active m id Earth Plant. Sci. Lett 302 : 349 358.

PAGE 84

84 Shelley, David, 1993, Igneous and Metamorphic Rocks Under the Microscope, Chapman and Hall, London, UK ocean ridge magma chambers Journal of Geophysical Research 97 197 216. Sinton, J. M., Wilson, D. S., Christie, D. M., Hey, R. N., & Delaney, J. R. (1983). Earth Plant. Sci. Lett. 62 : 193 207. Smith, M. C. basalts from the southern Juan de Fuca Ridge: Controls on the spatial and temporal evolution of mid Journal of Geophysical Research 99 : 4,787 4,812. Stakes, D. S. Perfit, M. R., Tivey, M. A., Caress, D. W., Ramirez, T. M., & Maher, M. Geochemistry, Geophysics, Ge osystems 7 ( 4 ): 1 32. Taylor, R. B. (1964). Geology of the Duluth Gabbro Complex near Duluth, Minnesota. The University of Minnesota Press, Minneapolis. Tierney, S. E. (2003) Distribution and composition of lavas from the southern Cleft segment of the Juan de Fuca Ridge: Tectonomagnetic evolution of a ridge transform intersection University of Florida: 195. Ussler III, W., Glazner, A.F. (1989) Phase equilibria along a basalt rhyolite mixing line: implications for the origin of calc alkaline intermediate ma gmas Contrib. Mineral. Petrol. 101 : 232 244. Science 150 : 485 489. Natu re. 245 : 373 375 Wanless, V. D., Perfit, M. R., Ridley, W. I., & Klein, E. (2010). "Dacite petrogenesis on mid ocean ridges: Evidence for oceanic crustal melting and assimilation." Journal of Petrology 51 ( 12 ): 2377 2410. Wanless, V.D., Perfit, M.R., Ridl ey, W.I., Wallace, P.J., Grimes, C.B., & Klein, E.M. mid Geology 287 : 54 65. sform faults, oceanic ridges and magnetic anomalies Science 150 : 482 485.

PAGE 85

85 White, S. M., Mason, J. L., Macdonald, K. C., Perfit, M. R., Wanless, D. V., & Klein, E. tions over the past two million years for delivery of magma to the overlapping spreading centers at 9'N Earth Plant. Sci. Lett 280 : 175 184.

PAGE 86

86 BIOGRAPHICAL SKETCH Kevin Werts grew up in Lubbock, Texas and graduated from Frenship High Sc hool in the spring of 2004. He then attended South Plains Community college from the fall of 2004 to the spring of 2006. Afterwards he attended Texas Tech University beginning in the summer of 2006 where he majored in geological sciences. He earned his B.S. in geology in the summer of 2010. Then he began graduate school at the University of Florida in the fall of 2010 and earned his M.S. in geology in the summer of 2012.