Light element and lithium isotope signatures of the emii reservoir - the society islands, french polynesia

Light element and lithium isotope signatures of the emii reservoir - the society islands, french polynesia

Material Information

Light element and lithium isotope signatures of the emii reservoir - the society islands, french polynesia geochemical results and an educational application
Harden, Judy Ann
Place of Publication:
[Tampa, Fla.]
University of South Florida
Publication Date:


Subjects / Keywords:
Ocean island basalts
Quantitative literacy
Dissertations, Academic -- Geology -- Masters -- USF ( lcsh )
government publication (state, provincial, terriorial, dependent) ( marcgt )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


ABSTRACT: The purpose of this thesis is to examine the abundance systematics of Li, Be and B, and Li isotopic systematics in lavas from the Society Islands, an enriched mantle (EMII) intraplate site, to further characterize the chemical signatures in the sources for ocean island basalts that may result from subduction-related processes and mantle entrainment. The goal is to see how light-element and Li-isotope systematics vary during ocean-island volcanic evolution and during tropical weathering.B/K, B/Be and Li/V ratios in basaltic Moorea lavas are 0.0001-.0002, 0.6-2.0 and 0.01-0.05 respectively, and the more evolved samples are somewhat higher. These ratios are similar to those for other Society Island lavas, and lower than those for lavas from St. Helena, Erebus, Hawaii, Gough and Reunion, as well as analyzed mid-ocean ridge basalts (MORBs).Li values for Moorea cluster at +3 +5 percent for the freshest lavas, and 0+2 percent for more weathered rocks.These new data from Moorea are consistent with earlier survey results from the Society Islands and indicate a mantle source that includes B-poor (subducted) materials. 7Li values for the freshest Moorea samples are similar to those of other Society Island lavas, suggesting that the EMII isotopic end-member records a Li-isotopic signature similar to that of MORBs. Dilution by entrainment of upper mantle material is unlikely due to differing B/K ratios and similar Li values for the Society and Hawaiian plumes.
Thesis (M.S.)--University of South Florida, 2005.
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 112 pages.
Statement of Responsibility:
by Judy Ann Harden.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
001681053 ( ALEPH )
62724109 ( OCLC )
E14-SFE0001069 ( USFLDC DOI )
e14.1069 ( USFLDC Handle )

Postcard Information



This item has the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001681053
003 fts
005 20060215071155.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 051221s2005 flu sbm s000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001069
QE26.2 (Online)
1 100
Harden, Judy Ann.
0 245
Light element and lithium isotope signatures of the emii reservoir the society islands, french polynesia
h [electronic resource] :
b geochemical results and an educational application /
by Judy Ann Harden.
[Tampa, Fla.] :
University of South Florida,
Thesis (M.S.)--University of South Florida, 2005.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 112 pages.
ABSTRACT: The purpose of this thesis is to examine the abundance systematics of Li, Be and B, and Li isotopic systematics in lavas from the Society Islands, an enriched mantle (EMII) intraplate site, to further characterize the chemical signatures in the sources for ocean island basalts that may result from subduction-related processes and mantle entrainment. The goal is to see how light-element and Li-isotope systematics vary during ocean-island volcanic evolution and during tropical weathering.B/K, B/Be and Li/V ratios in basaltic Moorea lavas are 0.0001-.0002, 0.6-2.0 and 0.01-0.05 respectively, and the more evolved samples are somewhat higher. These ratios are similar to those for other Society Island lavas, and lower than those for lavas from St. Helena, Erebus, Hawaii, Gough and Reunion, as well as analyzed mid-ocean ridge basalts (MORBs).Li values for Moorea cluster at +3 +5 percent for the freshest lavas, and 0+2 percent for more weathered rocks.These new data from Moorea are consistent with earlier survey results from the Society Islands and indicate a mantle source that includes B-poor (subducted) materials. 7Li values for the freshest Moorea samples are similar to those of other Society Island lavas, suggesting that the EMII isotopic end-member records a Li-isotopic signature similar to that of MORBs. Dilution by entrainment of upper mantle material is unlikely due to differing B/K ratios and similar Li values for the Society and Hawaiian plumes.
Adviser: Jeffrey G. Ryan, Ph.D.
Co-adviser: H. Leonard Vacher, Ph.D.
Ocean island basalts.
Quantitative literacy.
Dissertations, Academic
x Geology
t USF Electronic Theses and Dissertations.
4 856


Light Element and Lithium Isotope Signatu res of the EMII Reservoir The Society Islands, French Polynesia: Geochemical Results and an Educ ational Application by Judy Ann Harden 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 Co-Major Professor: Jeffrey G. Ryan, Ph.D. Co-Major Professor: H. Leonard Vacher, Ph.D. Charles B. Connor, Ph.D. Date of Approval: March 24, 2005 Keywords: Boron, geochemistry, mantle, o cean island basalts, quantitative literacy Copyright 2005, Judy Ann Harden


Dedication I would like to dedicate this thesis to a husband and child ren who constantly encouraged and supported me, to parents who in stilled a love for travel and a fascination for rocks and volcanoes, to the professors at Hillsborough Community College that helped launch my dreams, and to the faculty and staff at the University of South Florida who helped make those dreams come true.


Acknowledgments I would like to thank Dr. Jeff Ryan and Dr. Len Vacher for guiding, training, and teaching me and putting up with my continua l harassment while working on this thesis and Dr. Connor for his insight, encouragement, and support. I also want to thank Dr. Ivan Savov and Dr. Rob Watts who so freel y shared their knowledge, companionship, time, and encouragement. I owe a great de al to the wonderful staff of the Gump Research Station who made me feel so welcome. A special thank you goes to Dr. William White who donated samples from his pe rsonal collection for other islands in the Society Chain; although, I probabl y missed out on a really nice tr ip due to his generosity. I would also be amiss if I neglected to tha nk Professor Henry Aruffo for encouraging me to take his “Geography of Tahiti and Moorea cla ss” and sharing his love for an island and its people with so many students.


i Table of Contents List of Tables iv List of Figures v Abstract vii Prologue ix Part I B, Be, Li and Li isotopic systematic s of the Society Islands: Insights into the nature of EMII Mantle sources 1 Introduction 1 Geologic Setting 5 Age/Plate Movement 6 Volcanism 7 Eruptive Products and Lava Composition 9 Previous Work 10 Mantle Plumes and Hotspot Volcanism 10 Geochemistry of Hotspot Volcanics 12 Previous Work on B, Be, Li in Ocean Islands 13 Sampling and Analysis 17 Sample Collection 17 Petrology/Petrography 18


ii Geochemical Methods 23 Major and Trace Elements 23 Light Elements 23 Lithium Isotopes 24 Results 26 Major Element Variations 26 Trace Elements 27 B-Be-Li Concentrations 28 Li Isotopes 33 Discussion and Conclusions 35 Part II – Instructional Modules 39 Introduction 39 Module JAH1A & JAH1B – Fractional Crystallization 43 Module JAH2A & JAH2B – Partial Melting 46 Evaluation of Fractional Cr ystallization Modules 48 Comments on Evaluation Design 51 Results 53 Conclusions 54 References 55 Appendix I 65 Appendix II 68 Fractional Crystalli zation Module JAH1A 68


iii Fractional Crystalli zation Module JAH1B 77 Partial Melting Module JAH2A 86 Partial Melting Module JAH2B 92


iv List of Tables Table 1. Major and trace element data from senior thesis (Harden, 2002) for Moorea samples collected in 2001 65 Table 2. Major and trace element data for Moorea samples collected in 2003. 66 Table 3. Li, Be, and B analysis for 20 samples from Moorea and 13 samples from other Society Islands as indicated. 67 Table 4. 7Li values for 6 samples from Moorea and 6 samples from other Society Islands. 67


v List of Figures Figure 1. Location map showing study area of Society Islands, French Polynesia. 5 Figure 2. Society Island positions, age, and distance from hot spot (after Maury, 2000). 7 Figure 3. Effusive lava flows from Pu`u O`o crater, Kilauea, Hawaii representative of most eruptions at an ocean island hot spot (photo taken March, 2003). 8 Figure 4. Total Alkali-Silica classification of volcanic rocks collected from the island of Moorea.. 9 Figure 5. Island of Moorea with sample collection sites (red dots) and sample names. 17 Figure 6. Primitive basalt (M01Fish2D) w ith subhedral olivine and euhedral pyroxene with reaction rim phenocrysts. 18 Figure 7. M01 B2B hawaiite in plain & crossed-polarized light, plagioclase microphenocrysts with euhedral o livine glomerocry st intergrowth (40X). 19 Figure 8. M01 PK 25N Benmoreite with trachytic texture (plagioclase, spinel, pyroxene microphenocrysts) (40X). 20 Figure 9. Pyroxenite xenolith on le ft, right side is groundmass with microphenocrysts of plagioclase, olivine, and pyroxene, plain & crossed-polarized (40X). 21 Figure 10. All samples plot within Alka lic field (Modified from Macdonald & Katsura, 1964). 26 Figure 11. TiO2 contents increase with decreasing MgO. At ~5% Magnesium, Titanium decreases sharply indicatin g the crystalliza tion of magnetite in the magma chamber. 27 Figure 12. Samples from Moorea follo w predicted model for fractional crystallization for a range of F from 1.0-0.5. 28


vi Figure 13. Li vs MgO (a) and Be vs Mg O (b) contents for Moorea (turquoise squares), and Society Islands (navy s quares) with similar trajectories suggestive of similar parental source. 29 Figure 14. B/Be ratios range from 0.6-2.0. 31 Figure 15. B/K2O ratios range from 0.7-2.0 32 Figure 16. Society samples fall within a range of values for 7Li similar to those of MORBs, Erebus, and other OIBs. 33 Figure 17. Society samples fall within the upper values of Nishio et al., (2004) data for the EMI reservoir and the se rpentinites of Benton et al., (2004) but do not exhibit a ny range variation. 34 Figure 18. A) Basalts erupted at th e Society Island plume are alkalic in composition due to lower temperatures, higher pressure, and/or smaller degree of partial melting. 37 Figure 19. Results of questions posed to students both preand post-lecture regarding the process of fr actional crystallization. 53


vii Light Element and Lithium Isotope Signatu res of the EMII Reservoir the Society Islands, French Polynesia: Geochemical Results and an Educational Application Judy Ann Harden ABSTRACT The purpose of this thesis is to examin e the abundance systematics of Li, Be and B, and Li isotopic systematics in lavas from the Society Islands, an enriched mantle (EMII) intraplate site, to further characteri ze the chemical signatures in the sources for ocean island basalts that may result from subduction-related processes and mantle entrainment. The goal is to see how light-ele ment and Li-isotope systematics vary during ocean-island volcanic evolution and during tropical weathering. B/K, B/Be and Li/V ratios in basa ltic Moorea lavas are 0.0001-.0002, 0.6-2.0 and 0.01-0.05 respectively, and the more evolved samp les are somewhat higher. These ratios are similar to those for other Society Island la vas, and lower than those for lavas from St. Helena, Erebus, Hawaii, Gough and Reunion, as well as analyzed mid-ocean ridge basalts (MORBs). 7Li values for Moorea cluster at +3 +5‰ for the freshest lavas, and 0 +2‰ for more weathered rocks. These new data from Moorea are consistent with earlier survey results from the Society Islands and indicate a mantle source th at includes B-poor (subdu cted?) materials. 7Li values for the freshest Moorea samples ar e similar to those of other Society Island lavas, suggesting that the EMII isotopic e nd-member records a Li-isotopic signature


viii similar to that of MORBs. Dilution by entrainment of upper mantle material is unlikely due to differing B/K ratios and similar 7Li values for the Society and Hawaiian plumes. A more likely explanation is that recycled crust or sediments have minimal influence on the Li isotope signature s of hotspot plumes. Using the Moorea data and geochemical data from other sources, I created a set of Power Point instructional modules for use in petrology classes to ai d in teaching students about the effects of fractiona l crystallization a nd partial melting. I tested the module on fractional crystallization in two upper-level geology cla sses to assess its value in increasing student understanding. Both cla sses received a lecture about fractional crystallization. One class worked through th e module as a homework exercise, while the other did not use the module. Students who worked through the module in addition to the lecture showed an increased understanding of the concept of fractio nal crystallization.


ix Prologue This thesis consists of two parts. The first section is a geochemical examination of ocean island basalts from the island of M oorea, French Polynesia, and other islands of the Society Islands chain. The second part us es this collected data to produce a set of instructional modules for use in classrooms to further student understanding of processes of fractional crystallization and pa rtial melting in the EarthÂ’s mantle.


1 PART I B, Be, Li and Li isotopic systematics of th e Society Islands: Insights into the nature of EMII Mantle sources Introduction Radiogenic isotope ratios have been us ed for the past 40 years to answer questions about processes with in the Earth’s interior. Qu estions posed by Hart (1988), Hofmann (1988) and others include: How many discrete geochemical domains exist in the mantle? How do these domains form? Wher e are they located within the Earth? The deep mantle plumes responsible fo r the generation of ocean island basalts globally have been characterized in terms of four main “endmembers” defined by Pb, Sr, and Nd, isotope signatures. DMM is the depleted mantle source for mid-ocean ridge basalts (MORBs). HIMU mantle has elevated U/Pb ratios (), as indicated by high 206/204Pb and 207/204Pb. EMI is enriched mantle with very unradiogenic 206Pb/204Pb and the lowest 143Nd/144Nd present in the oceans. EMII is enriched mantle and contains the highest 87Sr/86Sr in the ocean and intermediate 206Pb/204Pb and 143Nd/144Nd (Hart, 1988). Because radiogenic isotope ratios are not modified during partial melting and magma chamber processes, data for lavas can be us ed to characterize the mantle source regions of basaltic magmas. The HIMU and the EM isotopic reservoirs have been attributed to subductionrelated origins in the past (Hofmann a nd White, 1982; Zindler and Hart, 1986; Hart,


2 1988; Hauri and Hart, 1993; Reisberg et al ., 1993; Thirwall, 1997). However, such inferences are not equivocal because the pr ocesses of subduction profoundly modify the composition of slab materials as they desce nd into the mantle, invalidating comparisons with the original surface materials (i.e. sedime nts and ocean crustal rocks) that are carried into trenches (Bebout et al., 1993; 1999; Sc hmidt and Poli, 2003). Tracers are required that are both sensitive to the process in que stion and well documented in terms of their terrestrial distribution to confirm the role of subduction or any other terrestrial geochemical process in creating a mantle domain. The systematics of the light elements Li, Be, and B are well understood in subduction-zone processes and are used to ch aracterize volcanic ro cks in all tectonic settings (see Ryan and Langmuir, 1987; 1988; 1993; Ryan et al., 1996; Leeman and Sisson, 1996; Ryan, 2002; Morris and Ryan, 2003; and references therein). Boron systematics in intraplate lava s globally point to a substantia l B depletion in these mantle sources and boron isotopic ratios lower th an those of MORBs (Ryan et al., 1996; Chaussidon and Marty, 1995). In contrast, Be abundances in intr aplate lavas are markedly elevated (Ryan, 2002). Recently the stable-isotope system of lithium (consisting of its two isotopes 6Li and 7Li), expressed as per-mille variations of 7Li from the value of the NIST standard reference material L-SVEC (7Li/6Li = 12.01), has been used su ccessfully to characterize a variety of Earth reservoirs and processe s (Chan et al., 1992, 1994, 1999; 2003; Tomascak et al., 2000; 2002; Pistiner and Henderson, 2003; Rudnick and Nakamura, 2004). This system is useful in studying geologic pro cesses involving lowto moderate-temperature fluid-rock exchanges because of the large mass difference between 6Li and 7Li (~17%)


3 (Tomascak et al., 1999) and the potentially large mass fractionati ons of the two Li isotopes in nature ( 7Li oceans = +32.3 ; sediment up to + 20 ‰; mantle rocks: -17‰ to +12‰, Chan and Edmond, 1988; Chan et al., 1992; Nishio et al., 2004, Rudnick and Nakamura, 2004). The ongoing development of multi-collector inductively coupled plasma-source mass spectrometry (ICP-MS) ha s facilitated the study of this isotopic system, as it permits the rela tively rapid determination of 7Li values on large numbers of samples. Lithium isotopic compositions of lavas accu rately represent their sources because isotope fractionation does not appear to occur during hightemperature crystal-liquid fractionation processes (Tomascak et al., 1999) Lithium isotopes have been used in defining the role of subducted sediment (C han et al., 1999; 2002), basaltic crust (MORB or eclogites) (Chan et al., 2000; Tomascak et al., 2002; Zack et al ., 2004; Bouman et al., 2004), seawater, continental crust (Teng et al ., 2004) and combinations of these in arc magma sources. As a step in more rigorously assessing the role of subduction in intraplate mantle sources, I characterized a suite of samples from the island of Moorea, of the Society Islands in French Polynesia, for their B, Be Li abundances and Li-i sotope signatures. I also analyzed alkali basalts from the other Society Islands to defi ne the overall lightelement signature of the Society chain and to s ee if temporal variati ons are evident. The radiogenic-isotope signatures of these EMII-t ype, hotspot-derived lava suites preserve a signature of past subduction (s pecifically a subducted sedime nt signature; White et al., 1982, 1996). Because the behaviors of light elements during subduction are known in


4 great detail, we may be able to say with greater confidence that an enrichment or depletion in B/Be or 7Li in these lavas is related to a process that happened during an ancient subduction event.


5 Geologic Setting The Society Islands lie in the south-cent ral Pacific Ocean at 15-18S latitude and 148-155W longitude (Figure 1). The Societ y Island chain, along with the Australs, Tuamotus, and the Marquesas, make up French Polynesia. These four linear chains are arranged parallel to each othe r in the direction of motion of the Pacific Plate and are approximately perpendicular to the East Pacifi c Rise. All four island chains represent the passage of Pacific Ocean plate lithos phere over a set of volcanic hotspots. Figure 1. Location map showing study area of So ciety Islands, French Polynesia.


6 Islands, seamounts, and atolls comprise the Society Chain. The islands extend over 700 km with the youngest is land, Mehetia, lying to th e southeast and the oldest island, Maupiti, to the northwest. Radiomet ric dating shows that the islands of the Society chain become progressively older to the northwest (4.5Ma – present; Okal, 1987). The land area of the Society Islands is 2,095 km2. Each island is highly dissected and consists of a basaltic volcanic core that has undergone erosion and denudation (Williams, 1933). Geographically, the Society Islands form two groupings. Tahiti, Moorea, Maiao, Mehetia and the atoll of Tetiaroa make up the Windward Islands. The older islands Raiatea, Tahaa, Huahine, Borabora, Maupiti and the atoll of Motuiti constitute the Leeward Islands. Moorea is the second largest island in th e Society chain, with an area of 132 km2 (Stearns, 1978). This island, like the others, represents the summit of a large, eroded, alkali-basalt shield volcano that rises approximately 4,000 m from the ocean floor. Mt. Tohiea, the highest peak of Moorea, has an elevation of 1,207 m (or ~5,200 m above the ocean floor). The island includes a deeply er oded caldera, the northern rim of which has largely collapsed, with only an isolated re mnant preserved, Mt. Rotui. Two large bays, Opunohu and Cook’s, on the north side of the island, give the island its tooth-shaped appearance. Age/Plate Movement The volcanism of Moorea and the other Societ y Islands is classically intraplate in nature. Samples collected by Dymond ( 1975) indicate an age of 1.65 0.13 Ma for


7 Moorea and 0.65 0.22 Ma for Tahiti. The di fference in ages of these two islands correlates with the proposed move ment of the Pacific Plate (Figure 2) at 11 cm/yr over a fixed hot spot or deep mantle plume (Dymond, 1975). Figure 2. Society Island positions, age, and distance from hot spot (after Maury, 2000). Volcanism There is no historical record of volcanic activity for any of th e Society Islands. However, from March to December, 1981, Me hetia (the island over current hotspot) experienced over 3,500 earthquakes associated wi th underwater eruptions at a depth of 1,600 m (Binard et al., 1993). Until recently, the common assumption about Pacific hotspot volcanism was that it is dominated by effusive basaltic lava flow s (Figure 3) with occasional fire fountains. Recent discoveries, however, indicate that e xplosive volcanism has occurred during the formation of ocean islands. One such pyroc lastic deposit near the top of Kulanaokuaiki Pali covers approximately 450 km2 of the summit area of K ilauea (Fiske et al., 1999).


8 Figure 3. Effusive lava flows from Pu`u O` o crater, Kilauea, Hawaii representative of most eruptions at an ocean island hot spot (photo taken March, 2003). With similar geochemistry, tectonic setti ng, and form, the possibility exists that explosive volcanism did occur during the form ation of the Society Islands. Neither the rock record nor collected samples from M oorea, however, show ev idence of explosive volcanism. Evidence for explosive volcanism on Moorea may no longer exist due to the much older age of this island and the amount of erosion that has taken place. There is, however, a unit described as a thick, columnar pyroclastic formation on the island of Tahiti, in the upper stage of its second shield. Eroded blocks from this unit, sampled in the Vaitamanu River, constitute an ignimbrite facies (Hildenbrand, 2003). Although most of the Society Islands probabl y formed by effusive volcanism, explosive activity can no longer be ruled out.


9 Eruptive Products and Lava Composition Moorea lavas are typically thin-bedded pa hoehoe and aa flows, usually 6 meters or less in thickness, and dip 5-10 degrees se award from the eruptive centers (Stearns, 1978). The lavas are alkali basalts with olivin e, pyroxene, and plagioclase phenocrysts. Dikes that cross-cut the flows range between 10 and 40 cm in thickness. Compositionally the lavas range from primitive magnesian basalts to trachytes (Figure 4). The presence of abundant in termediate-composition lavas (mugearites and benmoreites) may reflect magma-mixing events (Maury et al., 2000). Figure 4. Total Alkali-Silica clas sification of volcanic rocks co llected from the island of Moorea. Samples range in composition from primitive basalts to Hawaiites, Mugearites, Benmoreites, and trachytes.


10 Previous Work Mantle Plumes and Hotspot Volcanism Mantle plumes that produce intraplate volcanism appear to be around 200C hotter than ambient upper-mantle temperatures (Okal, 1987). Assumed plume viscosities are only slightly lower than those of the surrounding mantle meaning that a vertical, axisymmetric plume cannot ascend freely thr ough the mantle (Duncan et al., 1994). The surrounding mantle will instead be viscously coup led to the plume and ascend with it as a sheath-like boundary layer that can be hundreds of kilometers th ick. Further reduction in the viscosity contrast between plume and mantle is cause d by conductive heat loss from the plume into this boundary layer material, which also has the effect of increasing the buoyancy of the boundary layer. Duncan et al. (1994) suggest that the plume continuously entrains surrounding mantle as it ascends, such that the material closest to the center of the plume will ascend from the deepest levels, while the outer parts of the plume will be entrained at relatively shallow levels. The “hotspot”, then, is concentrically zoned with deep, hot plume material at its core and progressively shallo wer and cooler mantle material approaching its margins. When this rising structure en counters temperature/pr essure conditions for melting in the upper mantle, the entrained materi als melt to a lesser degree than the core plume mantle because the entrained material te mperature is lower. Given that the upper mantle is chemically more depleted than th e plume core (i.e. it has lower abundances of


11 incompatible elements and volatiles), it s hould melt to far lower extents at a given temperature than the core of the plume. The highest extents of melting observed in intraplate settings correlate with the passage of the plume core beneath the hots pot volcano – thus the highly voluminous eruptions observed at Kilauea and Mauna Loa, which sit above the core of the Hawaiian hotspot. As volcanoes are carried away fr om the plume core by plate motion, smallerdegree melts of cooler outer plume rocks are increasingly dominant, and volcanic activity fades. This pattern has been observed in th e eruptive histories of other intraplate sites and is considered a reasonable model for the vo lcanic history of an island in the Society chain. Until now, no one has conducted a comprehensive examination of the petrology of Moorea lavas. However, lavas from Tahiti have been studied in some detail (Duncan et al., 1994). Mean Nd concentrations increase with time in Tahitian lavas, which means that the mean degree of melting must theref ore have decreased with time. Based on Nd abundance and isotope systematics, Duncan et al. (1994) suggest that 5-15% melting produced the earliest magmas, wh ile as little as 1-2% me lting produced late-shield and late-stage magmas. These late-stage sample s with the highest Nd are believed to be derived from a source consisting predominantly of depleted mantle with less than a 10% admixture of material from the Society plume. As noted by Duncan et al. (1994) and othe rs, the Society Islands are dominated by alkali basalts, which are poorer in SiO2 and richer in alkalis th an tholeiites, consistent with lower extents of melting (Green a nd Ringwood, 1966). Pb/Ce ratios in Society basalts are typical to slightly higher than the values of othe r oceanic basalts and appear to


12 be a feature of their mantle source. Hofma nn and White (1982) used this parameter to suggest that the incompatible-element enri chments which characterize the Society plume are due in part to deep recycling of con tinental-crustal material, such as subducted continental sediments. Elevated Sr and Pb isotopic ratios are also consistent with the hypothesis. Geochemistry of Hotspot Volcanics Hotspot volcanics can preserve extreme va riability in Sr, Nd, and Pb radiogenic isotopic ratios and noble gas isotopes (i.e. ra diogenic isotopic ratios of He, Ar, and Xe) within a chain and even within different ma gma stages of a given island (Zindler and Hart, 1986; Hart, 1988; Kurz and Jenkins, 1982; Staudach er and Allegre, 1981). Explanations of radiogenic isotopic systematic s in ocean islands require multiple isotopic reservoirs in their mantle sources. As many as eight reservoirs (Z indler and Hart, 1986) and as few as four (Hart, 1988) are descri bed in the literature. Hofmann and White (1982) were early proponents of the idea that subducted ocean crustal materials incorporate into one or more of these isotopi c reservoirs. Subsequent studies focused on the possible subduction origins of three reserv oirs, named (originally by Zindler and Hart, 1986) HIMU, EMI and EMII. Hart (1988) and others conte nd that the “enriched” radiog enic isotope signature of the EMII reservoir (i.e., high 87Sr/86Sr, low 143Nd/144Nd, high 206Pb/204Pb and 207Pb/204Pb) is consistent with the likely signatures of subducted sediments. The EMII signature is largely limited to the southern Pacific ocean (i.e., Dupre and Allgre, 1981), and studies


13 suggest that it originates from a “graveyard” of subducted slabs in the mantle beneath the SE Pacific region (i.e. Castillo et al., 1996). Previous Work on B, Be, Li in Ocean Islands The relatively broad support for a subducted sediment origin for some ocean island sources aside, workers focusing on the geochemistry of downgoing slabs warn that metamorphic changes associated with pr ogressive subduction profoundly impact the incompatible-element and isotopic ratios of s ubducted materials. Di rect comparisons of the compositions of surficial materials and the geochemical signatures of hotspot lavas are unlikely to lead to useful conclusions, whic h is particularly true for the “fluid mobile” elements (e.g., Leeman, 1996) including B, Cs, Pb and (in some cases) Li. Boron is one of the most powerful indicator s for slab involvement used in studies of arc petrogenesis (Ryan and Langmuir, 1993; Ryan et al., 1996). Ryan et al. (1996), suggest that devolatilization of subducting plates segregates B into crustal reservoirs and returns large volumes of B-deplet ed material to the deep man tle. Materials from the deep mantle may preserve a distinctly depleted B signature and record the effects of ancient subduction events. B content ranges between 3 and 6 ppm in the majority of ocean island alkali basalts and shows greater relative vari ation than Be, Ce, or K, indicating that it behaves less compatibly during melting and crystallization (Ryan et al., 1996). Primitive mantle abundance estimates for B range between ~0.5 and <0.1 ppm (Leeman and Sisson, 1996). Dostal et al. (1996) studie d Li, Be, and B variations in submarine lavas from various islands in French Polyne sia. Their results show that OIBs from this region have


14 higher B contents than MORBs. Looking at Li/Be and B/Be ra tios, Dostal et al. (1996) concluded that Li and B are most lik ely removed from down-going slabs during subduction-related metamorphism and are not in volved in deep-level mantle recycling. Boron values greater than 5 ppm for intraplate lavas as reported by Dostal et al. (1996) are suspect. Because seawater contains ~ 4.5 ppm B, the use of submarine lava samples may skew the range of B contents to higher values. While B contents are higher in MORBs than OIBs, uniformly lower B/K and B/Nb ratios in the lavas indi cate that all intraplate sour ce regions, regardless of their isotopic characteristics, experienced boron de pletion (Ryan et al ., 1996). Chaussidon and Marty (1995) examined the B-isotope systematic s of submarine intraplate lavas, finding values between -14.6 and -4.3‰ (l ight compared to MORBs with values of -6.5 to -1.2 ‰). They presumed these light values were “primitive” and suggested that assimilation of small amounts of altered ba saltic crust may account for the higher boron ratios of MORBs and back-arc basin basa lts (BABBs -8.0 1.5 to +7.5 1.5 ‰). Similar ranges of values have not been encountered in st udies of subaerial in traplate whole rocks (generally 11B in subaerial lavas are somewhat hi gher: S. Tornarini, unpubl.), and the lack of reference samples determined bot h via TIMS and ion microprobe makes this dataset problematic to integrate. In contrast to Chaussidon and Marty (1995), Ryan et al (1996) suggest that all OIBs show B depletions related to a global event (probably continen tal crust formation) that generated two mantle reservoirs with di stinct B abundances. It is likely that fine distinctions between OIB sour ces developed episodically as more B-depleted subducted materials were added to the mantle (Ryan et al., 1996).


15 Lithium is a moderately incompatible tr ace element that is concentrated in sediments and altered oceanic crust relative to mafic igneous rocks, is soluble in hydrothermal fluids, and behaves similarly to Yb during fractional crystallization and V during melting present in ocean ridge settings Peridotite and basalt data suggest a mantle content of 1.9 ppm Li and indicate that significant Li resides in olivine and orthopyroxene (Ryan and Langmuir, 1987; Se itz and Woodland, 2000). The Li content for MORBs, from the most picritic to the most evolved, ranges from 3 to 15 ppm. Basalts usually contain less than 8 ppm Li (Ryan and Langmuir, 1987). Lithium isotopic signatures ( 7Li) distinguish seawater from oceanic basalts and are elevated in altered crustal rocks (C han and Edmond, 1988; Chan et al., 1992). Subduction may fractionate Li is otopes, such that deeply s ubducted slabs of isotopically light Li generate distinct re servoirs that can be sampled by plume-related magmas (Zack et al., 2003). While 7Li in altered MORBs ranges from +4.5 to +14 ‰, eclogites studied by Zack et al. (2003) have dramatically lower values, from -11 to +5 ‰. Low 7Li in eclogites is inferred to be produced by Rayl eigh distillation-style isotope fractionation during the early stages of metamorphism (Zack et al., 2003). Processes that may also contribute to low 7Li in these eclogites are seafloor al teration of their basaltic protoliths, fluid exchanges between the ec logites and their surrounding ga rnet mica schist during high-grade metamorphism, and fluid loss dur ing prograde metamorphism (Zack et al., 2003). Samples of ultramafic xenoliths presumed to reflect an EMII mantle source have 7Li of +4 to +7 ‰ (Nishio et al., 2004). Th ese values are comparable to those reported for terrestrial volcanic rocks. Anhydrous ultr amafic samples presumed to represent EMI-


16 type mantle (based on radiogenic isotopes) have 7Li of <-17 ‰ (Nishio et al., 2004). Serpentinites of the Mariana forearc st udied by Benton et al. (2004) range in 7Li from -6 to +10‰, indicating complex proc esses of Li isotopic excha nge during slab-fluid/mantle exchanges in shallow subduction systems. Ocean-island basalts studied for Li isotopes include samples from Mt. Erebus and other sites examined by Ryan and Kyle (2004) and lavas from several Hawaiian volcanic centers (Tomascak et al., 1999; Chan and Frey, 2003). 7Li in these lavas range from +3 to +7‰, values nearly indistinguishable from MORBs. No OIB samp le studied thus far shows the low values inferred by Nishio et al. (2004) as indicative of the EMI reservoir. An explanation for the MORB-like Li isotopic values offered by Ryan and Kyle (2004) is mixing between plume material and the MORBlike upper mantle results in the dilution of whatever “plume signature” may have been present. The bulk solid/liquid distribution coeffici ent for beryllium during melting of the mantle and crystallization of basalts is 0.03-0.06 making Be a strongly incompatible trace element, similar in its behavior to neodymium (Ryan and Langmuir, 1988). Be abundances in alkaline intraplate basalts (110 ppm) are up to five times greater than those of MORBs (0.15-2.5 ppm) and correlate with higher abundances of Zr, Nd and other incompatible lithophile elements. Ratio s of Be/Nd and Be/Zr in intraplate basalts are very similar to those of MORBs. More evolved alkaline basalts show larger variations in Be/incompatible element ra tios with progressive differentiation (Ryan, 2002).


17 Sampling and Analysis Sample Collection In August 2001 and March 2003, I collected over 40 different samples on Moorea, mostly along the coastal road that en circles the island but a few further inland (Figure 5). Dr. William White at Cornell University provided characterized basalt samples from other islands in the Society chain. Figure 5. Island of Moorea with sample name s and locations (small red dots). Rock types are noted if they differ from digiti zed geology. Black stars indicate points of interest on the island.


18 Petrology/Petrography The following descriptions are from twen ty thin sections from the freshest samples representing all the different rock types. Primitive and evolved basalts are aphyric to moderately porphyritic and contain phenocrysts of olivine and pyroxene in a groundmass of olivine, pyroxene, opaques (magnetite, illmenite) and microphenocrysts of plagioclase. Inferred order of crystallization is olivine+plag, followed by clinopyroxene and later opaques. Olivine and pyroxenes range from euhedral to anhedral depending on alteration; some have distinct reaction rims, and some are zoned (Figure 6). Microscopic zeolite crystals (natrolite, phillipsite, analcime, etc.) are present in some samples. Figure 6. Primitive basalt (M01Fish2D) with subhedral olivine and euhedral pyroxene with reaction rim phenocrysts. Zeolite s line the rim of vesicles (40X). Olivine Olivine pyroxene zeolites


19 Hawaiites and mugearites contain ~1-mm plagioclase phenocrysts in a fine groundmass (60% plagioclase, 30% opaques, 10% pyroxene for hawaiites, predominantly plagioclase in mugearites). Some contain c linopyroxenes and zeolite s. Textures range from subophitic to intersertal or felty (Figure 7). Figures 7. M01 B2B hawaiite in plain & crossed-polarized light, plagioclase microphenocrysts with euhedral olivin e glomerocryst intergrowth (40X). olivine Plagioclase


20 The benmoreites (intermediate lavas) are massive and variably vesicular with a trachytic texture (Figures 8). Zeolites are present in the vesicles of a few of the benmoreite samples, indicating hydrous alteration. Figures 8. M01 PK 25N Benmoreite with trach ytic texture (plagiocla se, spinel, pyroxene microphenocrysts) 40X.


21 One benmoreite sample, M01PK-19N, c ontained a pyroxene xenolith, 2 cm x 3 cm (Figures 9). I observed a similar xenolit h measuring 6 cm x 8 cm in the area where I collected the sample. Figures 9. Pyroxene xenolith on left, right side is groundmass with microphenocrysts of plagioclase, olivine, and pyroxene, plain & crossed-polarized (40X). pyroxene


22 Although samples appear fresh, in thin sec tion most samples show some degree of alteration due to weathering (i.e. olivine oxidation and removal, or reddening of the groundmass).


23 Geochemical Methods I prepared and analyzed samples collected in Moorea in 2001 for major and trace elements as part of a senior thesis project. I used similar methods for samples collected in 2003. Major and Trace Elements I utilized a LiBO2 fluxed, fusion digestion procedure modified from that of Tenthorey et al. (1996) to prepare sa mples for major and some trace-element measurements. For samples collected in 2001, I diluted a 5-ml aliquot of the lithium/beryllium solution with an equal amount of 2M HNO3 + 2000 ppm LiCO3. I added a germanium spike to dilution acids to serve as a performance monitor for the plasma spectrometry measurements and performed the analyses for major and trace elements using the Direct Current Plasma -Atomic Emission Spectro meter (DCP-AES) at the University of South Florida. Light Elements I selected 20 of the freshest Moorea samp les for Li, Be, and B analysis along with 13 well-characterized basalt samples from ot her Society Islands (White et al., 1996; Table 3). I prepared samples for Li and Be analysis following the HF-HClO4 acid digestion method outlined in Ryan and La ngmuir (1987). I used two boron digestion protocols: an HF-HCl-mannitol acid dige stion method modified from Ishikawa &


24 Nakamura (1992) for samples collected in 2001, and a Na2CO3 fluxed fusion technique modified from that of Ryan and Langmuir (1993) for samples collected in 2003. The two methods yielded comparable results for the very low B concentrations of the samples. I used standard additions methods to determ ine all light-element abundances. Samples were measured for B, Li and Be abundances by the (DCP-AES) at the University of South Florida. Analytical precision for Li and Be measurements are routinely 5%; uncertainties for boron measurements at the lo w abundance levels of these samples are in the 10-25% range. Lithium Isotopes I analyzed six samples from Moorea and six samples from other islands in the Society chain for Li isotopes. The M oorea samples reflect both the range of differentiation represented in the suite and so me variation in degr ee of weathering, based on the presence/absence of zeolite phases and ot her indicators. The other Society Islands samples chosen were the most primitive samples available (based on high MgO contents), with the intent of trying to defi ne both the "mantle" signature for Li isotopes, and any temporal variation in the signature that may have occurred. Sample preparation for Li isotopic analys is was at the Department of Geology at the University of Maryland and followed an HF:HNO3:HCl digestion. Samples dripped through a three-column separation method modifi ed from that of Moriguti and Nakamura (1998), in particular in that the th ird column is pressurized with N2 to facilitate separation. On occasion, samples are passe d through the third cation column a second time to improve separation of Na from Li, as excess Na in samples results in spurious Li


25 isotopic measurements. Typically, only samples with qualitative Na/Li intensity ratios of 5 or less are analyzed for isot opic ratios: for my samples, the Na/Li ratios were 3 or less. Lithium-isotopic-ratio measurements were conducted using the NU Plasma doubly focusing multi-collector-inductively co upled plasma source mass spectrometer (MC ICP-MS). Sample measurements are brack eted by measurements of the Li isotopic standard L-SVEC to correct for isotopic fractionation and further calibrated to determinations of in-house standards UM D-1 (+55‰) and IRMM-1 (-0.7‰). Lithium isotope values are expressed as per-mille variations from the L-SVEC Li isotopic standard ( 7Li) based on the following formula: 1000 *6 7 6 7 6 7 6 7 LSVEC LSVEC sampleLi Li Li Li Li Li Li Li Typically, measured Li isotopic rati os for L-SVEC lie between 13.1 and 13.5, ~10% higher than the accepted 7Li/6Li value of 12.01. As the measured values for LSVEC vary by less than 2‰ during the course of a run, it is possible to correct directly to determine 7Li values for bracketed unknowns. The accuracy of the measurements is a ssessed through measurements of reference samples JB-2 (+4.5‰) and BHVO-2 (+5‰). In all cases, the measurements are within 1‰ of accepted values in all runs.


26 Results The elemental and isotopic data for the Moorea and Societies samples are in Tables 1-4 and Figures 10-17. Major Element variations Based on the McDonald and Katsura classification scheme (1964), the samples from Moorea plot within the Hawaiian alkalic basalt field (Figure 10). SiO2 ranges from 46 to 60 wt % TiO2 contents increase steadily with decreasing MgO contents up to 5 wt %; however, in the most evolved rock s (Benmoreites and Mugearites), TiO2 is less than 1.5 wt % indicating crystallization of magne tite within the magma chamber (Figure 11). 0 2 4 6 8 10 12 444648505254565860 SiO2 (wt %)Na2O + K2O (wt %) Tholeiitic A kalic Figure 10. All Moorea samples plot within th e alkalic field (Modified from McDonald & Katsura, 1964).


27 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 MgO (wt %)TiO2 (wt %) Societies Moorea PRIMITIVE EVOLVED Figure 11. TiO2 contents increase with decreasing MgO. At ~5% Magnesium, titanium decreases sharply indicating the crystallization of magnetite in the magma chamber. Trace Elements Concentrations of trace transition metals are consistent with those of alkali basalts. Ni and Cr abundances correlate po sitively with MgO. Zn abundances range from 95 to 115 ppm. Cu abundances range fr om 57 to 82 ppm. The Cu/Zn ratio is < 1 (characteristic of alkali basalts; Table 1-2). Magmas with Ni of 200-300 ppm have likely experienced little olivine crystallization or accumulation (Hart a nd Davis, 1978; Sun and Hanson, 1975). Ni analyses of the Moorea samples lie within this range and confirm their primitive character. The Moorea samples form a continuous spectrum from primitive basalts to trachytes. As indicated on a plot of Ni vs. TiO2, fractional crystallization appears to be the dominant magmatic process of the form ation of the suite of rocks (Figure 12).


28 0 100 200 300 400 500 600 051015202530 TiO2 (wt %)Ni (ppm) predicted Cl Moorea samples Societies 20% 50% Figure 12. Samples from Moorea and other So ciety islands follow a predicted model for fractional crystallization for a range of F from 1.0 to 0.5. The most evolved Moorea sample erupted when approximately 50% of the magma chambe r had crystallized. B-Be-Li Concentrations Lithium abundances vary from 2.9 to 13.2 ppm and Be contents from 1.0 to 3.0 ppm. Both elements increase with progr essive differentiation.(Figure 13). Boron contents are 1.0 3.2 ppm. Excluding samples with zeol ite alteration, the range in boron abundances is even narrower (1.0 2.4ppm).


29 0 2 4 6 8 10 12 14 024681012141618 MgO (wt%)Li (ppm) Societies Moorea 00 05 10 15 20 25 30 35 0.02.0406.08010.012014.016.018.0 MgO (wt%)Be (ppm) Societies Moorea Figure 13. Li vs MgO (a) and Be vs MgO (b ) contents for Moorea (turquoise squares), and Society Islands (navy squares) with simila r trajectories suggestive of similar parental source. Both Li and Be increase with progressive differentiation. a b


30 Lithium and beryllium abundance systematic s in the Society lavas are similar to those reported for other ocean island basalts. Li and Be abundances both increase with increasing SiO2 and decrease with increasing Mg O contents, suggesting that both elements behave incompatibly during diffe rentiation processes that occur at ocean islands. Samples from Moorea and the othe r Society Islands show broadly similar trajectories, perhaps indicati ng a similar parental source. The Li/V ratios average ~0.02 consistent ly, indicating that Li and V behave similarly in OIBs, just as they do in MORBs and that the Li/V ratios of primitive MORBs and OIBs are similar (Ryan and Langmuir, 1987: Table 3). B/K2O and B/Be ratios for basaltic lavas are 0.7 1.4 and 0.6 1.3, respectively, while more-evolved samples have somewhat higher values. Comparison to other ocean island basalts (Dostal et al., 1996; Ryan et al., 1996) shows that the ratios are similar to those for other Society Islands lavas and lo wer than those for St. Helena, Hawaii, Gough and Ascension, as well as analyzed MORB. These low B/Be and B/K ratios are consistent with a mantle source that incl udes boron-poor subducted materials (Figures 14 and 15).


0.1 1.0 10.0 100.0 Be (ppm)B/Be Societies Society Ryan Moorea Ascension Gough St. Helena Hawaii Reunion Marquesas Older Society Altered Moorea ARCs MORBs Figure 14. B/Be ratios range from 0.6-2.0. Primitive Moorea and Society samples have lower B/Be ratios than other ocean island basalts. (Marquesas data ar e from Dostal et al., 1996. Other OI B data are from Ryan et al., 1996). 31


0.1 1.0 10.0 100.0 K2O (wt %)B/K2O Societies Society Ryan Moorea Ascension Gough St. Helena Hawaii Reunion MORBs ARCs Figure 15. B/K2O ratios range from 0.7-2.0. Moorea va lues are similar to other Societ y Islands and lower than those for St. Helena, Hawaii, Reunion and Ascension. Ascension, Gough, St. Helena, Hawaii and Reunion data are from Ryan et al., 1996. 32


33 Lithium Isotopes Samples show an overall range of 7Li from +1.5‰ to +5.5‰, with the Moorea sample subset showing the same range in valu es as that of the other Society islands. These samples all fall within the range of samples from other ocean island basalts from Hawaii, Erebus, Crary, St. He lena, Reunion and Iceland (Tomascak et al., 1999; Chan and Frey, 2003; Ryan and Kyle, 2004) and ar e similar to values for MORBs (Tomascak and Langmuir, 1999)(Figure 16). 0 1 2 3 4 5 6 7 8 0.7020.70250.7030.70350.7040.70450.7050.70550.7060.706587/86Sr7Li, ‰ Societies Reunion Pribilof Is. MORBs Erebus Crary Mtns Bullenmerri Hawaii St. Helena Iceland Figure 16. Society samples fall within a range of values for 7Li similar to those of MORBs, Erebus, and other OIBs. Pribilof Erebus, MORB, and Reunion from Ryan (2004). Bullenmerri from Nishio (2004).


34 Society Island samples do not exhibit the range observed in the data for the EMI reservoir by Nishio et al., ( 2004) or for the forearc serpenti nite data from Benton et al. (2004) and have neither exceptionally low or high 7Li signatures (Figure 17). -20 -15 -10 -5 0 5 10 0.7020.7030.7040.7050.7060.7070.7080.7090.7100.71187/86 Sr7Li EMI Nishio Societies FOREARC SERPENTINITES OIBs Figure 17. Society samples fall within the upper values of Nishio et al., (2004) data for the EMI reservoir and the serp entinites of Benton et al., (2004) but do not exhibit any range variation.


35 Discussion and Conclusions No simple correlations between Loss On Ignition (LOI) and Li isotopes are evident, although it is noted that weathering processes preferentially tend to remobilize and leach out 7Li resulting in overall lower 7Li values (Pistiner and Henderson, 2003; Huh et al., 2004; Rudnick et al., 2004). Th e Moorea samples are from a tropical island with high relief, so it is reasona ble to assume that all samples have been exposed to some weathering. Oddly, the sample with the mo st noticeable alteration, M01AIR1A, has one of the highest 7Li values at +4.6 ‰. M01FISH2D, one of the most primitive samples, appears to be very fresh with some surface zeolites, but it has a 7Li value of only +1.8 ‰. As zeolite formation and associated alteratio n is highly localized, it may be that even visibly weathered samples contain fresh horiz ons, which may be preferentially sampled during cutting, crushing and powdering; while ostensibly fresh sample segments may nonetheless preserve evidence of concealed alteration. Clearly, the presence of any zeolites in such samples is cause for con cern in terms of obtaining a "magmatic" Li isotopic signature. The relatively limited ra nge of values for the Moorea samples, and their similarity to those of fresh Society Island basalts, indicates that absent fresher samples, older lavas may be used to define Li isotopic minima. As with other OIBs, boron and boron ratios are very low (the lowest compared with all other OIBs examined thus far) fo r Moorea and other Society Island samples


36 indicating a source region that is depleted in B. Li and Be abundances are also similar to other OIBs pointing to similar behavior and source abundances. Li isotope ratios in fresh lavas are sim ilar to Hawaii, Erebus and several other ocean islands sites, but not to St. Helena or the Pribilof Islands. The total 7Li range of measured intraplate lavas, at ~4‰, is similar to that observed in MORBs, and offset to only slightly higher values. My samples from the Society Islands are very similar in their 7Li to MORBs. A question inherent in these data is w hy Li isotopes suggest little difference between the source regions of the Societ y Islands and MORBs, while B data (in particular B/K ratios) point to significant differences between the Society Islands source mantle and that of MORBs or other intrap late lavas, such as Hawaii (Figure 18). Entrainment of upper mantle material into the plume (see Duncan, 1994) is a viable means for producing largely similar 7Li values in such volcanic systems. The mean 7Li is +3‰ for Society Island samples, while published data for Hawaiian lavas are around +5‰ (Tomascak et al., 1999; Chan an d Frey, 2003). Both are within error of average MORB values (at +4‰: Tomascak and Langmuir 1999). Entrainment means that variable amounts of a MORB source man tle is sampled during hotspot melting. The more upper mantle material entrained, the grea ter the dilution of the plume signature in the resultant melts, and the more MORB-like the lavas become.


37 Figure 18 A) Basalts erupted at the Society Island plume are alkalic in composition due to lower temperatures, higher pressure, and/ or smaller degree of partial melting. B) Typically, tholeiitic basalt s are erupted from the Ha waiian plume due to higher temperatures, lower pressure and/or higher degrees of partial melting. The similar 7Li values imply that the upper mantle and deep mantle have similar Li isotopic compositions. Whereas, the differing B/K ra tios imply that these plumes are sampling mantle sources with different B/K signatures.


38 Farnetani and Richards (1995), however, sugge st that mantle plumes derived from the deep mantle entrain only a very small fraction of surrounding mantle into the region of the plume that undergoes pa rtial melting. Vertical plume tails with a strong viscosity contrast can entrain only a very small percentage of surrounding mantle (Stacey and Loper, 1983; Davies, 1999). The boron data on the Society Island samples also make the entrainment explanation problematic. B/K ratios for the Society Islands are ~1. Hawaii B/K ratios are ~5, and MORBs are ~10. Dilution of th e plume signal by entrainment and melting of upper mantle material should also cause B/ K ratios to become similar to those of MORBs. The differing B/K ratios of these tw o plumes thus imply that these plume sources are sampling mantle sources w ith very different B/K signatures. Instead of dilution by entrainment, a more viable explanation may be that the two possible mantle sources for both the Societ y and Hawaii plumes and the upper mantle have relatively similar Li isotopic compositi ons. Thus, it may be that recycled oceanic crust or sediments have only a minimal infl uence on the Li isotope signatures of the Society Island lavas. Data from the Society Islands suggest that 7Li values for subducted altered crust reported by Bouman et al. (2004) may be a more accurate representation of the subducted 7Li signature than original h eavy values for altered crust reported by Chan et al. ( 1994). The relatively low 7Li (~+5‰) seen in high-boron lavas from Panama (Tomascak et al., 2000) and in the serpentinite muds from the Mariana forearc (~+6‰; Benton et al ., 2004) also support a modest 7Li signature coming from subducted slabs.


39 PART II Instructional Modules Introduction Quantitative literacy as described by the International Life Skills Survey is an aggregate of skills, knowledge, beliefs, di spositions, habits of mind, communication capabilities, and problem solving skills that people need in order to engage effectively in quantitative situations arising in life and work (Briggs, 2004). Hyman Bass, American Mathematical Society President and former ch air of the Mathematical Sciences Education Board, has noted that quantitative literacy must be taught across th e curriculum, stating “while mathematics and statistics contri bute central knowledge and skills, other disciplines provide the contexts which are so important for quantitative literacy (Steen, 2004).”During workshops for geoscience educator s, however, it is repeatedly said that students have poor mathematical skills and te nd to avoid mathematics whenever possible (Vacher, 2001). Developing material that contains mathematics to increase the quantitative literacy of students, therefore, should be a goal fo r all geoscience educators. In my experience as a student and teaching assistant, I found that one of the larger challenges for students in undergraduate petr ology classes is learning how to think quantitatively about magmatic processes in the context of multiple geochemical variables. This ability is necessary to unde rstanding how major and trace elements vary among petrogenetically related igneous rocks. It is also crucial in distinguishing igneous


40 rock suites that are derived from different mantle sources. Student s face a real challenge when they try to learn how to decipher how magmas from and evolve from elementalabundance data. Traditional le cture reading materials and problem sets are often found lacking by students trying to understand th e rationale underlying the interpretive procedures. I have prepared a series of Power Poin t modules for use in Petrology classes to aid in teaching fractional crystallization a nd partial melting processes in the mantle. These modules are patterned after modules of geological/mathematical problem solving developed by H.L. Vacher and posted on th e website of the Washington Center for Improving the Quality of Undergraduate Educ ation (The Evergreen State College) and the Science Education Resource Center (Carleton College). The modules I have developed are inte nded for use in junior/senior level petrology or mineralogy/pet rology courses where magma tic processes are normally covered. The modules are designed to be performed by students as self-paced lab activities or homework assignments. The goal of the modules is to help students grasp concepts that are not easily understood thr ough lecture or reading. The modules ask the student users to graph geochemical data for a suite of rocks and dete rmine whether partial melting, fractional crystallization, or some other process dominated the formation of the suite. Excel spreadsheets are embedded in the m odules to show stud ents the value of solving a problem once, then using the same sp readsheet for rapid reca lculations. Use of spreadsheets does not require any computer programming skills; therefore, only basic computer literacy is needed. A study by Sm ith (1992) suggests thr ee possible outcomes


41 for students using Excel: (1) Reversal of lack of interest in mathematics; (2) improvement of technological literacy and enhancement of career preparation; (3) revitalization of mathematic al skills through problem solv ing. The content of these modules sharpens mathematical skills by inte grating algebra, logarithms, unit conversion, and graphing. Available geochemical modeling software, such as the GeochemistÂ’s Workbench, can be expensive. Such software is typical ly designed for researchers knowledgeable of geochemistry, as opposed to undergraduates le arning geochemistry for the first time. Even more advanced students may be challenged in becoming proficient with these geochemical programs. This software, along with web-based tools such as the MELTS program, is best suited to train students to become geochemists. Excel, on the other hand, is available on most computers, and st udents are often alrea dy acquainted with its operation. The use of Excel in the modules I have developed not only helps students to understand magmatic processes, bu t also helps students to devel op skills that they can use across a variety of courses and disciplines. A study by Fratesi and Vacher ( 2004) of articles in the Journal of Geoscience Education the principal journal of earth-science teachers in the US, identified 38 articles using spreadsheets, while less than a handful discuss the use of more sophisticated mathematics-oriented computer programs such as MATLAB or Mathematica Geology provides the context needed to sharpen and develop mathematical skills. One of the earliest articles by Ousey ( 1986) introduces the idea of spreadsheets for modeling groundwater flow, an important geologic phe nomenon. Spreadsheets allow students to concentrate on the subject matter rather than the software (Beare, 1993). Students who


42 are proficient with Excel and the mathematics within the modules attain life-long skills that transfer to other fields in geology and even other disciplines.


43 Module JAH1A and JAH1B Fractional Crystallization Part JAH1A of the fractional crystalli zation module introduces the process of fractional crystallization through a series of explanations and calculations of partition coefficients, bulk distribution coefficients, co mpatible vs. incompatible elements, and the Rayleigh fractionation equation. The goal is to develop a way by which one can evaluate whether progressive fractional crystallization relates the samples in a suite of volcanic rocks and, if it does, to calculate for each samp le the extent of fractional crystallization that occurred before the lava erupted. Module calculations use trac e-element data. Trace elements simplify looking at magmatic processes quantitativ ely, because their low concentrations (in the ppm range) mean they do not play a role in the stoichio metry of crystallizing mineral phases. Trace elements, therefore, substitute into crystals that are forming as a function of temperature, pressure, and the overall chemical compositi ons of the mineral and melt. Some minor elements, such as Ti, Mn, and K, behave as trace elements in basaltic rocks and are not major stoichiometric constituents of the mine rals that are forming (Best, 1982). It is possible to constrain the nature of the source rock, identify what minerals (and how much of each) melted to form a magma, and/or id entify the proportions of different minerals that may have crystallized by looking at trace -element variations in a suite of volcanic igneous rocks.


44 The data set used in the m odule draws partly from the collected geochemical data in Part I of this thesis because it is important that a suite of rocks from the same source is used when using geochemical data to look at magmatic processes. The module begins by defining fractional crystallization and discussing partition coefficients ( Dmineral/melt) and bulk distribution coefficients ( Dsolid/melt) and incorporates a Ds/l calculation in a spreadsheet (slides 1-7). W ith the introduction of the Rayl eigh fractionation equation and a spreadsheet calculatio n (slides 8-10), students determin e if a single rock has undergone fractional crystallization in its formation. Ca lculating the percentage of fractionation for each element in a rock, however, is extremely time consuming. Looking at the geochemistry of a suite of rocks, presumably from the same source, is a much better idea. In slides 11 and 12, students read a discussion about the compatibility of elements. Whether an element is compatible or incompatible depends completely on what minerals are present in the melt. For example, when olivines and pyroxenes are crystallizing, Sr is an incompatible element. However, as soon as plagioclase begins to crystallize, it incorporates Sr into its crys tal structure and Sr then becomes a compatible element. The students design a spreadsheet and create a graph to help them visualize the effects of compatibility. The last section of part 1A of the module (slides 13-16) discusses forward modeling. The students design a spreadsheet to calculate pred icted values for a model of fractional crystallization and then plot the gi ven data on the model. A list of questions posted at the end of the module can be used as a homework assignment. Part B introduces graphical techniques as a means of identifying and calculating fractional crystallization without making some a ssumptions that are necessary in part A.


45 Graphical methods rather than repetitive calc ulations are an excelle nt way to test for fractional crystallization in a given suite of samples, and so students learn to examine graphs to determine if fractional crystallization is the dominant process in the formation of the suite. In a closer look at the Rayleigh fracti onation equation (slides 3-6), the module applies logarithms and a little algebra to show students that this equation can be configured as a line in log x vs. log y Students then plot the data on a log-log graph. Another spreadsheet calculates the predicted va lues of fractional crys tallization and plots them on the same graph showing that the samp les not only fall along a straight line, but a line that can be predicted. The end of the module includes three data sets with instructions for students to plot the data and determine from the plots if fractional crystallization is the dominant process in the formation of the suite of samples. Included is a set of questions that can be used for assessment to determine student understanding of the pr ocess of fractional crystallization and its identification from scrutinizing graphs.


46 Module JAH2A and JAH2B Partial Melting Part JAH2A of the module introduces th e process of magma generation due to partial melting. Part JAH2B introduces the Shaw equation and ways to identify partial melting. The goal is to develop a way by whic h one can evaluate whether partial melting is the dominant process in a suite of volcanic rocks and, if it is, to calculate how much melting has occurred. A statement at the begi nning of the module ur ges students to first work through the fractiona l crystallization module (JAH1 ) and to have a basic understanding of phase diagrams. Once again, the module uses trace elements because they are an excellent means of looking at magmatic proce sses due to their low concentr ations. Therefore, it is possible to constrain the nature of the elements in the partial melt. The module uses data from the geochemical analyses in part I of this thesis along with other geochemical data from the lite rature. The module begins by discussing how decreased pressure, increased temperature a nd change in chemical composition can each produce partial melting (slides 3-6). A small animation (slide 7) he lps students visualize the process by which incompatible elements are enriched in the melt and compatible elements are enriched in the mantle. Slides 8-10 ask students to recall their know ledge of phase diagrams and calculate the percent composition by using the lever ru le. This part of the module aims for


47 students to see that melting does not move very far off the eutectic; a slight increase in temperature produces a large increase in melt. The second part of the module (JAH2B) in troduces the Shaw equation for partial melting. Graphs walk students through the c oncept of compatibility but in much less detail than in the module on fractional crystalli zation. Several slides explain the type of samples to analyze, the elemental data to us e, and how to determine graphically whether partial melting is the dominant process for a gi ven suite of rocks. The module ends with a few questions. Particularly instructive are questions that refer the students to new data sets. The students must review these data sets and determine if partial melting is an appropriate interpretation for each of them. Students must then calculate the percent of melting for any data set that represents partial melting.


48 Evaluation of Fractional Crystallization Modules Laura Wetzel of Eckerd College tested modules with similar design but different geological content during the Fall 2003 seme ster. (Wetzel et al., 2003). During the Fall 2004 semester, I tested modules JAH1A and JAH1 B at the University of South Florida to assess improvements in student understanding of the subject of fractiona l crystallization. I chose two upper-level Geology courses, So lid Earth 1 (Mineralogy/Petrology) and Computational Geology (a cour se specifically designed to de velop quantitative literacy). At the beginning of the semester, each cl ass of students answered ten questions, four of which we used to evaluate the effec tiveness of the module. The students of the Solid Earth class (control group) received a lecture on frac tional crystallization, but did not review the module or complete any assi gnments. The students in Computational Geology received the same lecture as prepar ation but had to work through the module and write a paragraph evaluating its contents. I informed both classes that they would be held responsible for the ma terial on future exams. One week after the lecture/assignment, I asked the same questions posed prelecture to assess improvements in student unde rstanding of the concepts and application of fractional crystallization. Results are as follows:


49 Question 1: What is fractional crystallization? Class participants correct incorrect percent correct Solid Earth 1 pre-lecture 19 0 19 0% Solid Earth 1 post-lecture 19 2 17 11% Comp. Geology pre-lecture 14 4 10 29% Comp. Geology post-lecture post-module 14 8 6 57% Question 2: What is the difference between an incompatible element and a compatible element? Class participants correct incorrect percent correct Solid Earth 1 pre-lecture 19 1 18 5% Solid Earth 1 post-lecture 19 5 14 26% Comp. Geology pre-lecture 14 0 14 0% Comp. Geology post-lecture post-module 14 7 7 50% Question 3: What is the difference between a partition coefficient a nd a bulk distribution coefficient? Class participants correct incorrect percent correct Solid Earth 1 pre-lecture 19 0 19 0% Solid Earth 1 post-lecture 19 2 17 11% Comp. Geology pre-lecture 14 1 13 7% Comp. Geology post-lecture post-module 14 2 12 14%


50 Question 4: When plotting a suite of samples that are dominated by fractional crystallization, what type of trend would you expect to see on a linear graph? on a logarithmic graph? Class participants correct incorrect percent correct Solid Earth 1 pre-lecture 19 3 16 16% Solid Earth 1 post-lecture 19 7 12 37% Comp. Geology pre-lecture 14 2 12 14% Comp. Geology post-lecture post-module 14 8 6 57%


51 Comments on Evaluation Design Students in the control group (Solid Earth I) are mostly sophomores and juniors. Three students in the Solid Earth class simultaneously enrolled in Computational Geology. Their participation in the evaluation was limited stri ctly to their responses on the questions posed in Solid Earth; I did not use their responses in Computational Geology in the evaluation. Of the 14 students in Computational Ge ology, only one had not yet taken Solid Earth I. Most of the students in this class are at a senior level, and all of them have worked through several modules on various subjects. They are not only familiar with module design, but also with Excel spreadsheets. Only students that attended all three cl asses for the pre-questions, lecture, and post-questions were included in this study. I told both classes at the beginning of each lecture that they would be responsible for knowing the material for upcoming exams. I designed the lecture to cover the material presented in the modules and presented the material in the same manner for both classes. The students in Computational Geol ogy worked through the module and evaluated it, but I did not requ ire them to turn in a homew ork assignment. I told the students to contact me if they had any que stions while working through the module. Only one student asked for assistance. A problem in the approach with the Co mputational Geology class arose when several students commented that they had not had time to work through the module prior


52 to taking the exam. Consideration had to be taken that we inadvertently tested the seriousness of the students rath er than the module. I addr essed this issue by asking the students to fill out a questionna ire that included a check box stating that they had not worked through the module prior to taking the test. Of the si xteen students that filled out the questionnaire, 11 admitted that they had not worked through the module. Of the five students that did work through the module, two were in the Solid Earth I class and their results were not considered. So, only three students actually worked through the module.


53 Results Students in Solid Earth I showed only a slight increase in understanding the material presented in the lecture, while stud ents in Computational Geology demonstrated a higher increase in understanding (Figure 19) The module will be tested further, explicitly in a Min/Pet class with the same approach as th e Computational Geology class, to see how students at this leve l benefit from the information. 0 10 20 30 40 50 60 Q1Q2Q3Q4Percent Correct Solid Earth Pre Solid Earth post Comp Geol pre Comp Geol post Figure 19. Results of questions posed to stud ents both preand post-lecture regarding the process of fractional crystallization. The results of the data above suggest that working through a module can greatly enhance student knowledge of concepts that are hard to understand through lectures or readings. However, as I later found out, onl y three of the studen ts worked through the module. On the other hand, two of these di d not answer any questions correctly prelecture but answered all four questions correctly after wo rking through the module. The third student answered 50% of the questions co rrectly. Results for the three students who carried out the experiment as it wa s intended are certainly encouraging.


54 Conclusions The fractional crystallization module a ppears to increase understanding of the subject matter. However, further testing of the module is required. When retested, students should be given more time to wo rk through the module and an assignment should be given to make the students more acco untable. Telling them that the material would be covered on an exam did not seem to motivate them enough to work through it. Because the modules are designed to be self-paced and worked through individually outside the classr oom, they can be an excellent supplement to lectures and may replace textbook readings. Another aspect noticed duri ng the testing of the module is that modules may be a valuable tool in exposing the deficiencies in student knowledge of basic mathematical skills. I did not test the partial me lting module, but it is scheduled to be tested in Solid Earth I at the Universi ty of South Florida in the Fall of 2005 along with an additional test of the fractional crystallizat ion module. Other modules that may be appropriate for Petrology courses are magma mixing (in the works) and assimilation.


55 References Beare, R., 1993. How spreadsheets can aid a variety of mathematical learning activities from primary to tertiary level. Tech nology in Mathematics Teaching: A Bridge Between Teaching and Learning. B. Jaworski. Birmingham, U.K.:117-124. Bebout, G.E., Ryan, J.G., and Leeman, W.P ., 1993. B-Be systematics in subduction related metamorphic rocks: character ization of the subducted component. Geochimica et Cosmochimica Acta 57:2227-2237. Bebout, G.E., Ryan, J.G., Leeman, W.P., a nd Bebout, A.E., 1999. Fractionation of trace elements by subduction zone metamorphism : significance for models of crustmantle mixing. Earth and Planet ary Science Letters 177:69-83. Benton, L.D., Ryan, J.G., and Savov, I.P., 2004. Lithium abundance and isotope systematics of forearc serpentinites, C onical Seamount, Mariana Forearc: Insights into the mechanics of slab/mantle exchange during subduction. Submitted to Geochemistry Geophysics Geosystems Best, M.G., 1982. Igneous and Metamorphic Petrology. New York: W.H. Freeman and Company. 630 pp. Binard, N., Maury, R.C., Guille, G., Talandier J., Gillot, P.Y., and Cotten, J., 1993. Mehetia Ialand, South Pacific: geology a nd petrology of the emerged part of the Society hot spot. Journal of Volcanol ogy and Geothermal Research 55:239-260. Bouman,C., Elliott, T., and Vroon, P.Z., 2004. Lithium inputs to subduction zones. Chemical Geology (In press).


56 Briggs, W., 2004. Quantitative Li teracy/Reasoning. URL: Castillo, P.R. 1996. Origin and geodynamic im plication of the Dupal isotopic anomaly in volcanic rocks from the Philippine island arcs. Geology 24,3: 271-274. Chan, L.H., and Edmond, J.M., 1988. Variation of lithium isotope composition in the marine environment, a preliminary repor t. Geochimica et Cosmochimica Acta 52,6:1711-1717. Chan, L.H., Edmond, J.M., Thompson, G., and Gillis, K., 1992. Lithium isotopic composition of submarine basalts; implicati ons for the lithium cycle in the oceans. Earth and Planetary Scie nce Letters 108, 1-3:151-160. Chan, L.H., Zhang, L., and Hein, J.R., 1994. Li thium isotope characteristics of marine sediments. EOS, Transactions, Am erican Geophysical Union 74,44 Suppl.:314. Chan, L.H., Leeman, W.P., You, C.F., 1999. Lithium isotopic composition of Central American Volcanic Arc lavas: implicatio ns for modification of subarc mantle by slab-derived fluids. Chemical Geology 160:255-280. Chan, L.H., and Kastner, M., 2000. Lithium isotopic compositions of pore fluids and sediments in the Costa Rica subduction zone : implications for fluid processes and sediment contribution to the volcanoes. Earth and Planetary Science Letters 183:275-290. Chan, L.H., Leeman, W.P., and You, C.F ., 2002. Lithium isotopic composition of Central American volcanic arc lavas; imp lications for modificiation of subarc mantle by slab-derived fluids; corr ection. Chemical Geology 182,2-4:293-300.


57 Chan, L.H., and Frey, F.A., 2003. Lithium isot ope geochemistry of the Hawaiian plume: Results from the Hawaii Scientific Drilling Project and Koolau Volcano. Geochemistry Geophysics Geosystems 4(3):1-20. Chaussidon, M and Jambon, A., 1994. Boron content and isotopic composition of oceanic basalts; geochemical and cosmochemical implications. Earth and Planetary Science Letters 121, 3-4:277-291. Chaussidon, M. and Marty, B., 1995. Primitive boron isotope composition of the mantle. Science 269:383-386. Davies, G.F., 1999. Dynamic earth plates, plum es, and mantle convection. Cambridge, UK: Cambridge Press, 458 pp. Dostal, J., Dupuy, C., and Liotard, J., 1982. Geoc hemistry and origin of basaltic lavas from Society Islands, French Polyne sia (South Central Pacific Ocean). Volcanology 45-1: 51-62. Dostal, J., Dupoy, C., and Dudoignon, P., 1996. Distribution of boron, lithium and beryllium in ocean island basalts from French Polynesia: Implications for the B/Be and Li/Be ratios as tracers of subducted components. Mineralogical Magazine 60: 563-580. Duncan, R., Fisk, M.R., White, W.M., and Ni elsen, R.L., 1994. Tahiti: Geochemical evolution of a French Polynesian volc ano. Journal of Geophysical Research 99, B12:24,341-24,357. Dupre, B., and Allgre, C.J., 1981. Compar ative Pb-Sr-Nd isotopic studies for North Atlantic and Indian oceans; consequences on the structure of the EarthÂ’s mantle. EOS, transactions, American Geophysical Union 64,17:423.


58 Dymond, Jack, 1975. K-Ar ages of Tahiti and M oorea, Society Islands, and implications for the hot-spot model. Geology 3, 5: 236-240. Farnetani, C.G., and Richards, M.A., 1995. Th ermal entrainment and melting in mantle plumes. Earth and Planetary Science Letters 136, 3-4:251-267. Fratesi, S.E., and Vacher, H.L., 2004. Usi ng spreadsheets in geos cience education: Survey and annotated bibli ography of articles in the Journal of Geoscience Education through 2003. eJSiE 1(3):168-194. Fiske, R.S., Rose, T.R., Swanson, D.A., and McGeehin, J.C., 1999. The Kulanaokuaiki tephra: product of newly recognized pyr oclastic eruptions at Kilauea Volcano, Hawaii. Eos, transactions s upplement, Fall meeting 1999 80,46:1196-1197. Green, T. H. and Ringwood, A. E., 1966. Contrib. Mineral. Petrol. 18:105 Gurenko, A.A., and Chaussidon, M., 1997. Boron concentrations and isotopic composition of the Icelandic mantle: evid ence from glass inclusions in olivine. Chemical Geology 135:21-34. Hart, S. R. and Davis, K. E., 1978. Nickel pa rtitioning between olivin e and silicate melt. Earth and Planetary Science Letters 40:203-219. Hart, S.R., 1988. Heterogeneous mantle dom ains: signatures, genesis and mixing chronologies. Earth and Planet ary Science Letters 90:273-296. Hauri, E.H., and Hart, S.R., 1993. Re-Os isot ope systematics of HIMU and EMII oceanic island basalts from the south Pacific O cean. Earth and Planetary Science Letters 114: 353-371.


59 Hildenbrand, A., Gillot, P., and Le Roy, I., 2003. Volcano-tectonic and geochemical evolution of an oceanic intra-plate vol cano: Tahiti-Nui (French Polynesia). Earth and Planetary Science Letters 6897:1-17. Hofmann, A. W. and White, W. M., 1982. Mant le plumes from ancient oceanic crust. Earth Planet. Scien ce Letters 57,2:421-436. Hofmann, A.W., 1988. Chemical differentiation of the Earth; the relationship between mantle, continental crust, and oceanic cr ust. Earth and Planetary Science Letters 90,3:297-314. Huh, Y., Chan, L.H., and Chadwick, O.A., 2004. Behavior of lithium and its isotopes during weathering of Hawaiian basalt. Ge ochemistry Geophysics Geosystems 5, 9:Q09002, doi:10.1029/2004GC000729. Ishikawa, T., and Nakamura, E., 1992. Boron is otope geochemistry of the oceanic crust from DSDP/ODP Hole 504B. Geochi mica et Cosmochimica Acta 56, 4:16331639. Kurz, M.D., Jenkins, W.J., and Hart, S.R., 1982. Helium isotopic syst ematics of oceanic islands and mantle heterogeneity. Nature 297:43-47. Leeman, W.P., Sisson, V.B., and Reid, M.R ., 1992. Boron geochemistry of the lower crust; evidence from granulite terranes and deep crustal xenoliths. Geochimica et Cosmochimica Acta 56, 2:775-788. Leeman, W.P., 1996. Boron and other fluid mo bile elements in volcanic arc lavas; implications for subduction proce sses. Geophysical Monograph 96:269-276. Leeman, W.P., and Sisson, V.B., 1996. Geochemistry of boron and its implications for crustal and mantle processes. Reviews in Mineralogy 33:645-707.


60 Maury, R. C., LeDez, A., Guillou, H., 2000. Moorea. Editions du BRGM Service geologique national. 1-59. McDonald, G.A., and Katsura, T., 1964. Ch emical composition of Hawaiian lavas. Journal of Petrology 5:82-133. Montelli, R., Nolet, G., Dahlen, F.A., Mast ers, G., Engdahl, E.R., and Hung, S.H., 2004. Finite-frequency tomography reveals a vari ety of plums in the mantle. Science 303:338-343. Moriguti, T., and Nakamura, E., 1998. High-yield lithium separation and the precise isotopic analysis for natural rock a nd aqueous samples. Chemical Geology 145:91-104. Morris, J.D., and Ryan, J.G., 2003. Subduction zone processes and implications for changing composition of th e upper and lower mantle. In Carlson, JR. (Ed.) Treastise on Geochemistry Chapter 2: The Mantle Chap.2.11,451-470. Nishio, Y., Nakai, S., Yamamoto, J., Su mino, H., Matsumoto, T., PrikhodÂ’ko, V.S., and Arai, S., 2004. Lithium isotopic systematics of the mantle-derived ultramafic xenoliths: implications for EM1 origin. Earth and Planetary Science Letters 217:245-261. Okal, K., 1987. Hotspots: The First 25 Years. In Seamounts, Islands, and Atolls. Geophysical Union. 1-11. Ousey, JR., 1986. Modeling steady-state gr oundwater flow using microcomputer spreadsheets. Journal of Geological Education 34:305-311.


61 Pistiner, J.S., and Henderson, G.M., 2003. Lithium-isotope fractionation during continental weathering processes. Eart h and Planetary Science Letters 214:327339. Reisberg, L., Zindler, A., Marcantonio, F., White, W., Wyman, D., and Weaver, B., 1993. Os isotope systematics in ocean island basalts. Earth and Planetary Science Letters 120, 3-4:149-167. Rose, E.F., Shimizu, N., Layne, G.D., and Grove, T., 2001. Melt production beneath Mt. Shasta from boron data in primitive melt inclusions. Science 293, 5528:281-283. Rudnick and Nakamura, 2004. Preface to “Lithium isotope geochemistry”. Chemical Geology (in Press). Ryan, J., and Langmuir, C., 1987. The systematics of lithium abundances in young volcanic rocks. Geochimica et Cosmochimica Acta 51: 1727-1741. Ryan, J., and Langmuir, C., 1988. Beryllium systematics in young volcanic rocks: Implications for 10Be*. Geochimica et Cosmochimica Acta 52: 237-244. Ryan, J., and Langmuir, C., 1993. The systematics of boron abundances in young volcanic rocks. Geochimica et Cosmochimica Acta 57: 1489-1498. Ryan, J.G., Morris, J., Tera, F., Leema n, B.P., and Tsvetkov, A., 1995. Cross-arc geochemical variations in the Kurile Ar c as a function of slab depth. Science 270:625-627. Ryan, J., Leeman, W., Morris, J., and La ngmuir, C., 1996. The boron systematics of intraplate lavas: Implications for crus t and mantle evolution. Geochimica et Cosmochimica Acta 60: 415-422.


62 Ryan, J.G.,2002. The trace-element behavior of beryllium in terrestrial materials. In Grew. E. ed.: Beryllium: Mineral ogy and Geochemistry Reviews of Mineralogy and Geochemistry 50: 121-145. Ryan, J.G., and Kyle, P.R., 2004. Lithium and lithium isotope variations in intraplate mantle sources: Insights from McMurdo Group lavas (Mt. Erebus) and other intraplate volcanic rocks. Chemical Geology 212:125-142. Schmidt, M.W., and Poli, S., 2003. Generati on of mobile components during subduction of oceanic crust. In Rudnick, R. (ed) Treatise on Geochemistry 3 : The Crust, Chapter 3.17:567-591. Schmincke, H.-U., 2004. Volcanism. Springer-Verlag, New York. 342pp. Seitz, H.M., and Woodland, A.B., 2000. The di stribution of lithium in peridotitic and pyroxenitic mantle lithologies: an indicator of magmatic and metasomatic processes. Chemical Geology 166:47-64. Smith, R., 1992. Teaching precalculus with a spreadsheet. Fifth Annual International Conference on Technology in Collegiate Mathematics. Rosemont, Chicago, Illinois, Addison-Wesley. Sleep, N.H., 1990. Hotspots and mantle pl umes: Some phenomenology. Journal of Geophysical Research 95:6,715-6,736. Spivack, A.J., and Edmond, J.M., 1987. Boron isotope exchange between seawater and the oceanic crust. Geochimica et Cosmochimica Acta 51:1033-1043. Stacey, F.D., and Loper, D.E., 1983. The boundary-l ayer interpretation of D” and its role as a plume source. Physics of the Ea rth and Planetary In teriors 33, 1:45-55.


63 Staudacher, T., and Allgre, C.J., 1982. Te rrestrial xenology. Earth and Planetary Science Letters 60:389-406. Stearns, H., 1978. Comparison of the geology of the Society and the Hawaiian Islands Bernice P. Bishop Museum Occasional paper 24: 276-290. Steen, L.A., 2004. Quantitative Literacy: Why numeracy matters for schools and colleges. URL: Sun, S. S. and Hanson, G. N., 1975. Origin of the Ross Island basanitoids and limitations upon the heterogeneity of mantle sources for alkali basalts and nephelinites. Contrib. Mineral. Petrol. 52:77-106. Teng, F.-Z., McDonough, W.F., Rudnick, R.L., Dalp, C., Tomascak, P.B., Chappell, B.W., and Gao, S., 2004. Lithium isotopic composition and concentration of the upper continental crust. Geochimica et Cosmochimica Acta (In press). Tenthorey, E.A., Ryan, J.G., and Snow, E.A ., 1996. Petrogenesis of sapphirine-bearing metatroctolites from the Buck Creek u ltramafic body, Southern Appalachians. Journal of Metamorphic Geology 14, 2:103-114. Thirwall, M.F., 1997. Pb isotopic and elemen tal evidence for OIB derivation from young HIMU mantle. Chemical Geology 139:51-74. Tomascak, P.B., and Langmuir, C.H., 1999. Li thium isotope variability in MORB. EOS, Transactions American Geophysi cal Union Supplement 80, F1086-1087. Tomascak, P.B., Tera, F., Helz, R.T., and Walker, R.J., 1999. The absence of lithium isotope fractionation during basalt diffe rentiation: New measurements by multicollector sector ICP-MS. Geochimica et Cosmochimica Acta 63: 907-910.


64 Tomascak, P.B., Carlson, R.W., and Shir ey, S.B., 1999. Accurate and precise determination of Li isotopic compositions by multi-collector sector ICP-MS. Chemical Geology 158:145-154. Tomascak, P.B., Ryan, J.G., and Defant, M. J., 2000. Lithium isotope evidence for light element decoupling in the Panama sub-arc mantle. Geology 28:507-510. Tomascak, P., Widom, E., Benton, L., Gold stein, S., and Ryan, J., 2002. Control of lithium budgets in island arcs. Earth and Planetary Science Letters 6108: 1-12. Vacher, 2001. Better math, better geology. Geotimes 46, 3:13. White, W.M., and Hofmann, A.W., 1982. Sr a nd Nd isotope geochemistry of oceanic basalts and mantle evolution. Nature 296, 5860:821-825. White, W.M., and Duncan, R.A., 1996. Geoche mistry and geochronology of the Society Islands: New Evidence for Deep Mantle Recycling. Earth Processes: Reading the Isotope Code. Geophysical Monograph 95:183-206. Williams, H., 1933. Geology of Tahiti, Moorea, and Maiao Bernice P. Bishop Museum Bulletin 105: 1-88. Winter, J.D., 2001. An introduction to igne ous and metamorphic petrology. Prentice Hall, New Jersey. 697p. Zack, T., Tomascak, P.B., Rudnick, R.L., Dalpe, C., and McDonough, W.F., 2003. Extremely light Li in orogenic eclogites: The role of isotope fractionation during dehydration in subducted oceanic crust. Earth and Planetary Science Letters 208:279-290. Zindler, G. A. and Hart, S. R., 1986. Ch emical Geodynamics. Ann. Rev. Earth Planet. Science 14: 493-571.


Appendix I Sample SiO2Al2O3FeOMgOMnOCaOK2ONa2O3TiO2totalSrCrBaNiScVZnCu (wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(ppm)(ppm)(ppm)(ppm)(ppm)(ppm)(ppm)(ppm) M01FISH 2D 45989.9312.7714530.159.600.931.832.7098.43420862783953025510060 M01FISH A45.4112.6613.3713330.169.750.952.132.93100685276792113892925210461 M01T1A 45.7814.6312.1611.160.159.311.262.272.9399.64566545270298272359558 M01FISH2B 459612.4413.6913350. M01FISHB1462214.9512.8411.790.159.841.192.313.08102365965713743692724610458 M01FISH2A 462411.9113.1912.400.159.021.352.043.0399.355525053273632825710968 M01B2C 46.7611.8613.9413510.179.231.381.923.00101.775766553003382825910366 M01AR1A 469212.4113.3011350.169.621.272.283.2410056573506296284272599864 M01B1A 475714.5713.737.900.1810.251.342.873.1810159956291268190282119568 M01B2A48.011813012.00.148.651.541.962.95100005375153353292925612159 M01PK9N485415.5712.405.590.139.881.352.953.4499.8657764280100272749565 M01B2B 485715.8712.365.070.179.951.713.064.01100.7766242339105273069982 M0118N B 488316.0312.984.380.159.991.832.824.0110103698-34 M01QB 489813.4812.729.220.188.332.232.722.59100.455525512512852722111561 M01T2A 490413.9412.918.870.168.931.632.923.34101.755553593532352625810360 M01QA 493814.4113.128.430.188.512.172.692.71101605065302492772721811359 M0118N A 49.7117.1511.954.410.158.752.193.053.931012974236 M01MOEA2A 49.7417.0912.614.490.157.472.293.303.721008663795 M01QE54.21968.5390.153.993.754.611.1710000205162149127178716341 M01QD 54.4616.618.373.950.125.524.143.981.1898.34445337 M01FALLS1A 545919.708. 79 M0113S A 556518.228.042.910. M01QC 55.7417.357.963. M01PK13S-B 58.317.77.9290.143.274.654.730.86100531459411479165914323 M01PK19N59.11909.4050.121.845.524.190.8410054783219131153614814 M01PK24N59.91908.2100.162.095.664.630.621012673378943164114918 M01PK25.5N60.218.47.6060.111.925.704.740.6710000761918027142614713 Table 1. Major and trace element data from senior thesis (Harden, 2002 unpublished) for Moorea samples collected in 2001. 65


SampleSiO2 Al2O3 FeOMgOMnOCaO K2ONa2O TiO2TotalSr (ppm)CrBa(ppm)Ni(ppm)Sc(ppm)V(ppm)Zn(ppm)Cu(ppm) (wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(wt%)(ppm)(ppm)(ppm)(ppm)(ppm)(ppm)(ppm)(ppm) M03LN-0094689.814316.00.187990.791652.561004068202885172622312356 M03PK19N46.411013514.10.179.161051922.741004786593204643429212575 M03T2A 47.011512914.00.168580951.732.51994336692954823125412074 M03020 47.010913514.10.138620861.772.56995086813094983026312962 M03011 47.210813314.10.168.771201822.621004577142984013326413058 M03010 47.610.113814.50.168881091822.741014547302445083125012745 M03PK20N49.41421255.70.1410.661.412.403691006731813471123031211678 M03LN-D5021661094.20.098261942983.71995871113741002930012274 M03BELV52.51461098.20.166622583502271016703747432162315712534 Table 2. Major and trace element data for Moorea samples collected in 2003. 66


67ISLANDSAMPLE Li Be B (ppm) (ppm) (ppm) MooreaM01FISH 2D 491.23 1 MooreaM01FISH A5.11.19 1 MooreaM01T1A 5.41.03 1 MooreaM01FISH2B 4.71.52 1 MooreaM01FISHB1581.53 2 MooreaM01FISH2A 6.861.8 1.4 MooreaM01B2C 591.391.8 MooreaM01AIR1A 321.353.2 MooreaM01B1A 6.60.952.24 MooreaM01PK9N5.71.561.8 MooreaM01B2B 621.71.3 MooreaM01QB 11.32.751.8 MooreaM01T2A 781.761.3 MooreaM01QA 13.153 3 MooreaM03PK19N58 0.7 3.3 MooreaM03020 69 1.1 3.6 MooreaM03011 12.32.9 4.2 MooreaM03010 6.7 1.1 1.5 MooreaM03LIN-D 65 1.0 3.1 MooreaM03BELV 4.6 1.1 2.7 BoraBora101287 45 1.0 6.1 BoraBora101288 62 1.1 9.2 Huahine101211 63 1.0 4.9 Huahine101198 93 2.7 8.5 Maiao MAO-113 6.6 1.5 2.2 Maiao MAO 65 3.6 1.5 4.3 Maupiti73-204 59 1.3 1.9 MehetiaMHT-101 3.7 1.1 2.2 MehetiaMHT-155 43 1.2 1.6 Tahaa Ta3F 5.4 0.9 3.0 Tahaa Ta8b 5.1 1.0 3.1 Tahiti 74-422 58 1.1 2.5 Tahiti 74-437 70 2.1 3.1 Table 3. Li, Be, and B analysis for 20 samples from Moorea and 13 samples from other Society Islands as indicated. ISLANDSAMPLE 7 Li MooreaM01FISH2D1.76 MooreaM01FISH A2.29 MooreaM03PK19N4.95 MooreaM01AIR 1A4.59 MooreaM03LIN-D1.47 MooreaM03BELV4.75 Huahine1012113.67 MaiaoMAO-1134.1 Maupiti73-2043.1 MehetiaMHT-1013.52 TahaaTa8b3.55 Tahiti74-4225.47 Table 4. 7Li values for 6 samples from Moorea and 6 samples from other Society Islands. The samples from Moorea we re chosen based on increasing SiO2 to see temporal variations of 7Li within the formation of an island and to note the affects of weathering. Samples from the other Societ y Islands were chosen based on low SiO2 to see temporal variation within th e formation of an island chain.


68 Appendix II Fractional Crystallization Module JAH1A 1 When magma cools and crystallizes in the Earth’s interior, the growing crystals incorporate selected chemical elements of the magma into their structure. As a result, the magma is depleted in those elements. On the other hand, elements that are not incorporated in the growing crystals are enriched in the magma. Separation of crystals from the liquid by settling, floating, oradhering to the magma chamber walls produces a magma with a different chemical composition. This process is called fractional crystallization Quantitative Concepts and Skills Manipulating equations Weighted average Forward modeling Fractional crystallization causes basaltic magmas to evolve into basaltic-andesiticmagmas. Further crystallization pushes the magma toward an even more siliciccomposition. Mantle Processes 1A –Fractional Crystallization Module JAH-1AJudy Harden, University of South Florida


69 2 Part A introduces the process of fractional crystallization through a series of explanations and calculations of partition coefficients, bulk partition coefficients, compatible vs. incompatible elements, and the Rayleighfractionation equation. The goal is to develop a way by which we can evaluate whether fractional crystallization has occurred in a suite of volcanic rocks and, if it has, to calculate how much fractional crystallization has occurred. This module consists of two parts, A and B. Part B introduces graphical techniques as a means of identifyingand calculating fractional crystallization without making some assumptions that are necessary in part A. About this module Slides 4-7 explain the difference between partition coefficients and bulk partition coefficients. Slides 8-10 introduces the Rayleigh fractionation equation. Slides 11-12 explain compatibility. Slides 13-16 uses forward modeling to look for fractional crystallization.Data in this module are from samples collected from the island of Moorea. Moorea is part of the Society Island Chain, French Polynesia. Two of the better-known islands of the chain are Tahiti and BoraBora. Moorea Tahiti PREVIEW


70 4 Partition coefficients The composition of crystallizing minerals changes continuously as the composition of the remaining melt changes. Using major element geochemistry to calculate the effects of fractional crystallization is complicated due to the changing composition of the melt. Different phases of mineral growth preferentially incorporate or exclude trace elements to a greater extent than they do major elements. Therefore, trace elements provide a useful fingerprint to constrain origins of melt systems and their evolutionary processes. concentration of trace element in mineral concentration of trace element in melt KD= The concentration of a trace element in a mineral is proportional to the concentration of the trace element in the liquid from which it grew. This principle is represented by the partition coefficient ( KD), which is defined as follows: Trace elements are elements that represent <1wt.% of the rock composition and are typically in the ppmrange. 5 Partition coefficients in a mineral Suppose you have a huge caldron of vegetables (magma chamber) and you decide to make cucumber and onion salad (olivine). When you pull chopped cucumber and onion from the caldron, you may likely mix in some cherry tomatoes (Nickel) because they fit well (compatible) into the structure of what you are making. Later, however, you decide to make succotash (plagioclase). You pull corn and lima beans from the caldron, and it is unlikely that you will mix in any cherry tomatoes (Nickel) because hey just dont fit as well (incompaible) into the structure of succotash. Sufferin Succotash! KDof nickel in olivine is 10, meaning that nickel is compatible and will fit into the structure of olivine. In contrast, the KDfor nickel in plagioclase is 0.07, meaning that nickel is incompatible and will likely not fit into the structure of plagioclase. Trace elements are incorporated into some minerals more than they are into others. This phenomenon is addressed by the concept of compatibility Compatibility results from the effects of the structure of the mineral and the ionic radius and charge of the element. For example: KDrelates to the concentration of a trace element in a mineral. What about rocks?


71 6 Bulk Partition Coefficient -D To calculate D you need to know the mass fraction ( X ) of the element in each mineral and the KDvalue for the element in each of the minerals. D = XaKD,a+ XbKD,b+ XcKD,cwhere a b ,and c are different minerals in the assemblage. D is the sum of the weighted average of each element in the mineral assemblage. The bulk partition coefficient addresses the amount of trace element in a rock. KDvalues are derived empirically. A good website to find these values is the Geochemical Earth Reference Model (GERM) : concentration of trace element in mineral assemblage concentration of trace element in meltD = 7 Calculate the bulk partition coefficient for nickel ( DNi) in a rock that contains 10% olivine, 42% clinopyroxene, 45% plagioclase, and 3% magnetite. KDvalues are 10, 3, 0.07, and 48 respectively. Once you have set up the spreadsheet to calculate DNifor the mineral assemblage, use the GERM website to find the values for titanium ( KD,Ti) in the same mineral assemblage and calculate DTi. Will Ti stay in the melt or be incorporated into the crystal? Save this spreadsheet, because you will use a similar calculation later in the module. According to your answer, will Ni stay in the melt or be incorporated into the crystal composition of the rock? = cell containing a number. = cell containing an equation. Calculating D Use the sumproduct function in this cell. BCD 2 XKD3Olivine0.110 4Pyroxene0.423 5Plagioclase0.450.07 6Magnetite0.0348 7 8 D =


72 8 RayleighFractionation To see how trace-element concentrations change as a result of crystallization, weuse the Rayleighfractionation equation shown below. Minerals that crystallize out of a melt and the trace elements they contain are removed from chemical contact with the residual liquid. Therefore, the concentration of the trace element in the remaining liquid ( Cl) is lower than the original concentration ( Co) of the trace element. Cois sometimes called the parental source Cldepends on Co, D, and the fraction of remaining liquid ( F ). Geochemical data of rock samples are the values we use for Cl. Cl= CoF(D-1) Suppose we want to track the concentration of a trace element ina rock formed during fractional crystallization. We can show this variation on a graph as a function of crystallization progress using a ratio of Cl/ Co. To see how, rearrange the equation to: John William Strutt, known as Lord Rayleigh, was born on November 12, 1842. His education was interrupted repeatedly due to ill health and his prospects of attaining maturity were uncertain. Despite his educationally impaired youth, he later won fame as an outstanding scientist receiving many awards and honors one of which was the Nobel Prize in 1904 ( Nobel Lectures Physics 1901-1921, Elsevier Publishing Company, Amsterdam, 1967). ) 1 ( D o lF C C F2When D = 3 F When D = 2 1 When D = 1 1/ F when D = O Cl/ Co= See what happens when we insert values for D. 9 Calculating F Problem A sample collected from Mooreacontains 61 ppmCopper (Cu) and 389 ppmNickel (Ni). Assume the parental basalt contained 56 ppmCu and 517 ppmNi. What is the value for F if DCu= 0.17and DNi= 3.73? BCD 2CuNi 3 Cl61389 4 Co56517 5 D 0.173.73 6 D-1 -0.832.73 7 F We know Clfrom geochemical analyses of our rocks. We calculated D. Often oneassumes a value for Coby using calculated values for spinel peridotites(representative of the EarthÂ’s mantle). One can find these values on the GERM website. For this module, we set Coequal to the value of Clin our most magnesium-rich sample. With this assumption, we can develop a spreadsheet to calculate F


73 10 Calculating a value of F for each element in each sample that you have collected would be very time consuming. There is a better way of finding the percent of fractionation for your samples. Calculating F The answer reflects the portion of the original melt present in the magma chamber when the fractionated magma erupted to form this sample. BCD 2CuNi 3 Cl61389 4 Co56517 5 D 0.173.73 6 D-1 -0.832.73 7 F 0.900.90 Identical values testify that (a) fractional crystallization has occurred and (b) your geochemical analysis is accurate. When F = 1, no crystallization has occurred (100% of the melt remains). In contrast, F = 0.1 means that 90% of the original melt has crystallized. You should not expect every element to be exactly 0.90. Why? 11 Before we try to find F for a suite of rocks, we must fully understand compatibility. Create a spreadsheet now to see how Cl/ Cochanges with different values of D and F Then plot your results as a function of F for each D value. Make the y -axis a logarithmic scale. Compatibility The values you calculate are the predicted values for fractional crystallization for any given D. = cell containing intermediate formula BCDEF 2 D 0123 3 D-1 -1012 4 5 F Cl/CoCl/CoCl/CoCl/Co60.1 70.2 80.3 90.4 100.5 110.6 120.7 130.8 140.9 151


74 12 Compatibility What you should discover with your graph is that compatible elements ( D > 1) show depletion in the melt (Cl/Co< 1). Depletion occurs more rapidly the higher the value for D. Incompatible elements ( D < 1 )show enrichment in the melt ( Cl/ Co>1). Compatible elements preferentially enter the solid phase while incompatible elements preferentially remain in the liquid phase. 001 0.10 100 1000 0 1 FCl/ Co D = 0 D = 1 D = 2 D = 3 13 Now we are ready to analyze a suite of samples to test whether fractional crystallization is the dominant process of their formation. Theprocess we use here illustrates forward modeling where values are predicted and thentested to see if real data fit the model. Using previously designed spreadsheets to calculate ratios To see the maximum compositional variation, we use a compatible element and an incompatible element. Use the spreadsheet you created in Slide 11 to predict the Cl/ Covalues for Ti and Ni with D values of 0.32 and 3.73 respectively. (Just type the new values for D in your chart and let Excel do the work!) BCD 2TiNi 3 D 0.323.73 4 D-1 -0.682.73 5 6 F Til/TioNil/Nio70.1 80.2 90.3 100.4 110.5 120.6 130.7 140.8 150.9 161 After you obtain the values for Til/Tioand Nil/Nio, plot them on a graph with the Ni ratio on the y -axis and the Ti ratio on the x -axis.


75 14 What the graph tells us. Your graph should produce a curve that indicates that Ti increases in abundance while Ni decreases. 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0123456 Til/Tio predicted ratios At this point, 90% of the original melt remains, while 10 % of the original melt has crystallized. 80% 70% Now were ready for our suite of samples! 15 Create a new spreadsheet as shown to the right to calculate the concentration ratios for real samples collected from Moorea. The samples were analyzed on a Direct Current Plasma Atomic Emission Spectrometer (DCP-AES) to obtain the values of major and trace elements for each sample. Adding real data Now plot the calculated sample ratios on the graph that you just made with the predicted ratios. Samples that plot on the curve confirm a dominant process of fractional crystallization. Which samples appear to be affected by some process other than fractional crystallization? BCDEF 2TiNiTil/Tio Nil/Nio 3 Co2.56517 4FISH2D2.70395 5LN-0092.56517 6FISH A2.93389 7T1A2.93298 8FISH2B2.89385 9FISHB13.08369 10FISH2A3.03363 11PK19N2.74464 12B2C3.00338 13AR 1A3.24284 14T2A2.51482 15202.56498 16112.62401 17102.74508 18B1A3.18190 19B2A2.95329 20PK9N3.44100 21B2B4.01105 22QB2.59285 23T2A3.34235 24PK20N3.69112 25QA2.71277 26LN-D3.71100 27BELV2.27216 28QE1.17127 29PK13S-B 0.8679 30PK19N0.8431 31PK24N0.6243 32PK25.5N0.6727


76 16 Samples that plot off the curve owe their origins to some other process. This plot allows you to see what samples were erupted after a percentage (blue dots) of fractional crystallization took place. 0.00 0.20 0.40 0.60 0.80 1.00 1.20 Til/TioNil/Nio Predicted values Moorea samples Moorea samples affected by some other process than fractional crystallization These samples were erupted when ~50% of the magma chamber had crystallized. These samples were erupted when less than 10% of the magma in the chamber had crystallized. 17 1.Calculate F for a rock that contains 3.03 (wt %) Ti and 392 ppmNi with a composition of 10% olivine, 45% pyroxene, and 45% plagioclase. Assume that theparental source contained 2.56 (wt %) Ti and 517 ppmNi. Is Ti a compatible or incompatible element? 2.Suppose you calculate values for F on slide 10 and they donÂ’t match. What could be some possible explanations for the discrepancy? 3.How sensitive is D to changing rock composition? 4.What happens when D =1? 5.Since F represents the amount of remaining liquid, ( 1-F) is the amount of crystallization that has occurred. Add a column to the spreadsheet from slide 11 to calculate (1F ). Plot each of the D values vs. (1F ) What observations can you make regarding this graph? 6.What are some possible explanations for the Mooreasamples that do not plot anywhere on the predicted fractionation curve (green dots on graph of slide16)? 7.On the graph of Slide 16, the samples end at the point where ~45% of the original magma had crystallized. Would you expect to find erupted rocks further down the curve? Explain your answer. 8.Look through geologic literature and find geochemical data for asuite of rocks from another island in a chain such as Hawaii, Canaries, Galapagos. Show whether fractional crystallization was a dominant process in the formation of thoserocks. End of the module questions


77 Fractional Crystallization Module JAH1B 1 In module 1A we demonstrated various ways of determining whetherfractional crystallization was a dominant process in the formation of a suite of rocks. This module introduces the use of logarithms as a shortcut method of testing for fractional crystallization. Quantitative Concepts and Skills Manipulating equations Logarithmic graphing This module should be worked only after working through module 1A. Mantle Processes 1B –Fractional Crystallization Module JAH-1B 2 PREVIEW Slides 3-6 Manipulate the Rayleighfractionation equation to produce the equation of a straight line on a log-log graph Slide 7 provides data for use on a log-log graph. Slides 8-10 show a means of double checking your assumptions about fractional crystallization. Slides 11-16 contain end-of-module questions in the form of data sets. Slide 17 Pre-and Post-module questions Note: To save time and to minimize errors, the data sets in this module are in Excel format. This means that when you are in Power Point in normal view (not in the slide show), you can double click on the data and Excel will open. You can then copy the data to a new Excel spreadsheet rather than re-type the entire data set. At that point, you can manipulate the data by sorting or any other process.


78 3 Graphical methods rather than repetitive calculations are an excellent way to test for fractional crystallization in a given suite of samples. Using the graph made in slide 16 of module 1A requires that we know Coand D for the trace elements in order to see if our samples follow a predicted curve for fractional crystallization. By using the technique in this part of the module, we no longer need to know Coand D Using logarithms for graphing purposes Logarithms and a little algebra make it possible to test whetherfractional crystallization is a dominant process in the formation of a suite of rocks. ItÂ’s as easy as plotting the data on a log-log graph. Always use a compatible and incompatible element with the largest range in concentration. To see how the method is going to work, we start with the Rayleighfractionation equation. Cl= CoF(D-1)(1) 4 Using logarithms for graphing purposes Set equations 4a and 4b equal to each other and solve for log Nil. Before proceeding, write down your equation. Take the logarithm of equation 1. (2)F D C Co llog ) 1 ( log log Substitute the element of interest into equation 2, and do the same for the second element of interest. (3a) (3b)F D Ni Ni F D Ti TiNi o l Ti o llog ) 1 ( log log log ) 1 ( log log Now, solve equations 3a and 3b for log F ) 1 ( log log log Ti o lD Ti Ti F ) 1 ( log log log Ni o lD Ni Ni F(4a) (4b)


79 5 Does your equation look like the equation below? If it doesnt, review the rules of logarithms and try again. Using logarithms for graphing purposes The website listed below offers easy to understand explanations of logarithm rules. o Ti Ni o l Ti Ni lTi D D NiTi D D Ni log 1 1 loglog 1 1 log (5) 6 Using logarithms for graphing purposes Although this equation 5 may look complicated, it is merely the equation of a straight line. o Ti Ni o l Ti Ni lTi D D NiTi D D Ni log 1 1 loglog 1 1 log y = m x + b Because of this relationship, we can say that the dominant process of formation for a suite of samples that produce a straight line trend on a log-log graph is consistent with fractional crystallization.


80 7TiNiFISH2D2.70395LIN-009256517FISH A2.93389T1A293298FISH2B289385FISHB1308369FISH2A303363PK19N2.74464B2C300338AIR 1A3.24284T2A2514822025649811262401102.74508B1A3.18190B2A295329PK9N3.44100B2B401105QB2.59285T2A334235PK20N369112QA2.71277LIN-D3.71100 10 100 1000 1 10 Ti (wt%)Ni (ppm) Moorea samples Using the following samples from Moorea, create a plot of the samples on a log-log graph. You can see that a straight line trend occurs and we can say that the samples are consistent with fractional crystallization processes during the formation of this suite of samples. We can also do a simple calculation to check if our samples follow a predicted trend. 8 Log-log graphs Modify the spreadsheet you created for slide 13 in part 1A of this module to include the Covalues for Ni and Ti concentrations. You have already calculated F( D-1)(same as Cl/ Co); now multiply it by the given Covalues to solve for Cl. This will allow you to see the predicted sample values (not ratios) for the given parental concentrations that you can then compare with real sample values. BCDEF 2T iN i 3 D 0.323.73 4 D-1 -0.682.73 5Co2.56517 6predictedpredicted 7 F F ( DTi 1) F ( DNi 1) TilNil80.14.790.00 90.22.990.01 110.41.860.08 120.51.600.15 130.61.420.25 1711.001.00 The parental magma ( Co) is the most primitive (highest magnesium, lowest silica) found in an area. It is the one from which we assume all others have been derived. Now, plot the predicted sample values on a logarithmic graph.


81 9 Log-log graphs From the graph, we can see that the Mooreasamples not only form a straight line trend, but they also fall along a line that we predict for fractional crystallization processes. The Moorea samples all plot along the trend of this line except for the six samples that you have already discovered have been affected by some other process. 10 100 1000 110 Ti (wt%)Ni (ppm) predicted sample values Moorea samples What about the slope of the line? 10 Calculating slope of logs Using the predicted values you calculated, or picking two pointsfrom the line on your logarithmic graph (we used the second and third points),calculate the slope of the line. The logarithmic form of the Rayleighequation says the slope is ( DNi–1/ DTi–1 = slope). Using the D values from slide 8, you should calculate a slope of -4.0. If your slopes don’t match, double check your graph points and your math. They must match! Remember the rules of logarithms in your calculation! BCD 212 3 x 2.752.98 4 y 387.77281.14 5 m = Now you’re ready to plot data and decide whether fractional crystallization is the dominant process for the suite of rocks that you are working with!


82 1.You may plot samples on a log-log graph and two different trajectories may be evident as in the graph below of additional data from Moorea. What are some explanations? 2.Add a trend line to the graph you made for slide 9. Display theequation and the R2value. What is the slope of the line? 3.The following slides contain data for several different suites of rocks. For which of the suites of rocks, if any, are the chemical data consistentof fractional crystallization? For which suites of rocks, if any, do the chemical data contradict fractional crystallization as the dominant process? 1 10 100 1000 1101001000 Zn (ppm) End of module questions 12 Sample set 1: The following data from Hawaiian Volcano Observatory represent basalts from Kilauea on the Big Island of Hawaii. This volcano has beenerupting from the Pu`u O`ocrater almost continuously since 1983 adding over 220 hectares to the island.Sample SiO2MgO TiO2Ni S-546.6819.521.8542 S-1546.8218.871.86525 S-748.2213.672.27296 S-948.4113.342.3293 F-1748.7713.132.42269 S-849.1210.452.53170 F-1249.3410.582.69191 S-2549.448.852.67130 S-1449.458.552.62149 S-349.628.852.55168 S-149.918.082.62104 1050.536.333.2394 650.556.193.3190 850.616.613.1292 F-550.766.093.3782 250.995.423.5772 551.065.443.6568 Note that Ni decreases in concentration with increasing SiO2and that TiO2increases in concentration with increasing SiO2. Effusive flow during March 2003 at Kilauea, Hawaii (photo by Judy Harden). Click on the link below to see exciting pictures of current eruptions and to learn more about Kilauea and other Hawaiian Volcanoes.


83 13 Sample set 2: The data set on the following slide is from Colima, Mexico. The Colimavolcano complex is one of the most active volcanic systems in North America and has recently experienced dome-growing eruptions and Vulcanian type explosions. ColimaVolcano degassing 6-2003. (photo and map courtesy of RicMacNeil) 14 Sample set 2: Colima, Mexico. You will need to sort the data before choosing the elements you want to look at. What can you determine about the formation of these samples?(Luhr, J.R. and Carmichael, I.S.E., 1990. Petrologicalmonitoring of cyclical eruptive activity at VolcanColima, Mexico. Journal of Volcanologyand Geothermal Research, 42:235-260).SampleSiO2MgOVCrNiCuZnRbZrBa S- S-8.155.675.719919377355912121382 Col-20457.075.73111330138406120114532 1004-62058.523.771466624256418127457 1004-41060.612.768761676622136540 1004-4116062.7192252596223137541 1004-41260.342.84961518126724140548 1004-40460.063.19123371556624140494 1004-40559.923.161082621146825140514 1004-40960.693.299241687122133543 1004-41360.232.81112616136326130536 1004-41460.323.09110292496521141537 1004-41660.952.76881914106225137505 1004-41860.633.13872629126525144528 1004-44460.382.91131816146322135535 M82-1160.223.291364927197124134472 M82-10602.931062111177122144497 1004-42058.43.961444723186516131422 1004-42156.884.111386026146519125416 1004-41557.544.061463834257120131426 Col- Col-17.260.792.846367224320136556 Col-17.360.642.9570259214920146498 Col-17.460.392.950307134620147559 Col-17.560.432.868864194421138574 Col-17.660.662.8365288233418119526 1004-41760.782.911141318206221145515 Col-9B6082.971071821186525141498


84 15 Sample set 3: The following data set is from SoufriereHills, Montserrat in the Caribbean. After 8 years of activity the volcano has recently quieted but only after destroying the main city of Plymouth and the only airport on the island. Small eruption with a pyroclasticflow at SoufriereHills, Montserrat on April 20, 2003 (photo by Judy Harden). 16SampleSiO2MgOVThScLiBe 13653.73432761.832.67980.62 13950.175.723430.535.15020.49 78848.055.673670.5137.44.420.45 83051.714.823161.0335.26360.56 114848.285.483360.5135.24820.48 120450.185.433140.8535.15.10.5 2559.452.731382.315.112.090.8 12759.692.99107417 15258.64.141062.8218.713.760.76 15459.743.051172.531412.290.88 78563.792.25822.911.21690.86 81959.962.911363.0217.210.090.75 113558.393.11302.4415.88310.76 113658.653.151342.5316.65970.77 113762.962.09853.26101650.89 1865462.632.46853.0712.215.981 4058.732.891282.5616.414.40.86 4759.952.961272.3914.513.180.83 17458.892.861322.241814.220.78 24459.752.941232.4714.916.120.81 28860.682.561092.7813.715.090.85 665L60.282.851222.5414.712.450.78 107862.032.3580313.612.260.84 112258.152.991332.4715.713.160.77 115158.133.041382.4815.412.670.76 6054.763.481692.3616.815.070.67 6250.524.542030.8620.310.690.53 6355.643.41791.917.912.810.79 53152.82431981.5422.211.120.64 55151.474.242241.0223.49330.55 56453.83.681901.5721.911.250.65 66351.784.142101.0119.412.870.56 113349.224.922340.7822.59520.51 Set 3: How is this set of samples different from the other setsthat you have looked at? Why? (Zellmer, G.F., Hawkesworth, C.J., Sparks, R.S.J., Thomas, L.E., Harford, C.L., Brewer, T.S., and Loughlin, S.C., 2003. Geochemical evolution of the SoufriereHills Volcano, Montserrat, Lesser Antilles Volcanic Arc. Journal of Petrology 44:1349-1374.)


85 Geology related questions: What rock types or features of rocks do we see at the surface ofthe Earth that give us insight about the composition of the Earths mantle? What is fractional crystallization? What is the difference between an incompatible and a compatible element? What is the difference between a partition coefficient and a bulk partition coefficient? When plotting a suite of samples that are dominated by fractional crystallization, what type of trend would you expect to see on alinear graph? On a logarithmic graph? Would a magma that undergoes fractional crystallization be depleted or enriched in incompatible elements? Quantitative Literacy related questions: What are the basic rules of logarithms? What is forward modeling? What is a weighted average? Pre-and Post-module questions


86 Partial Melting Module JAH2A 1 We have gained much insight about the interior of the Earth since 1863. We now know that some Dante-like “seat of fires” does not exist and that volcanic eruptions occur when certain processes cause the rocks in the mantle to melt. Students should work through module 1A and 1B (fractional crystallization) and have a basic understanding of phase diagrams before working through this module. Mantle Processes 2A –Partial Melting Module JAH-2A“Igneous rocks are those which have been ejected in a melted state, as from volcanic vents, or from fissures opened to some seat of fires within or below the earth’s crust.” – New Textbook of Geology (Dana, 1863) 2 PREVIEW Slides 3-6 discuss the various known means of melting in the Earth’s interior: decreasing pressure, increasing temperature, and changing the composition of the mantle. Slide 7 contains a simple animation of current ideas of how partial melting occurs and how the concentration of incompatible elements decreases as melting increases. Slides 8-10 present a binary phase diagram and asks students to use the lever rule to calculate the percentage of a melt. Slide 11 presents a calculation illustrating how much melting must occur to completely melt a mineral and change a phase. Slide 12 –end-of-module questions. See module JAH2B for an introduction of the Shaw equation and data sets to look for indications of partial melting. This module is the first part of a set of two (JAH2A and JAH2B).JAH2A was developed to illustrate current ideas of how melting occurs within the mantle. JAH2B introduces ways to identify partial melting using geochemistry. It is highly recommended that each module be worked through for a better understanding of partial melting.


87 3 The Processes of Melting Seismic data indicate that melting does not occur in the mantle under normal circumstances. Although various estimates have been calculated for the oceanic geothermal gradient, none of them approach the solidus of the mantle. Melting, therefore, is not a product of the geothermal gradient. However, melting must occur or we would not have basalts eruptedat the surface of the Earth. There are only three known ways to cause melting in the mantle: 1.Lower the pressure 2.Raise the temperature 3.Change the composition Before looking at samples, lets take a look at these processes. As with fractional crystallization, it is important to look at asuite of samples from the same source, rather than at a single sample, to identify partial melting. 4 Lowering pressure Melting can occur by a decrease in pressure at a constant temperature. One way is to raise mantle rocks rapidly enough to minimize heat loss tothe surroundings. Diverging plates are the prime location for this process to occur. Mantle material flows upward to fill in areas that have been vacated from erupting Mid-ocean ridge basalts (MORB). This process is known as decompression melting. It occurs not only where plates are rifted apart. It also occurs at hot spots. The material inside the area of the bold black lines is typically called mush by geologists. It consists of some solid and some melt.Diagram after Wyllie, P.J., 1981. Plate tectonics and magma genesis. Geol. Rundsch, 70:128-153. 0 5 10 15 20 25 0100020003000 (C)0 100 200 300 00 500 600 700Depth (km) Liquidus solidus Oceanic Geotherm

PAGE 100

88 5 Increasing temperature Melting could occur if the mantle was heated beyond the normal geothermby the decay of radioactive elements such as K, U, and Th. Radioactivity is the only known source of heat in the interior of the earth other than that from the primordial differentiation process. The radioactive elements can produce only ~10-8J g-1 a-1. The specific heat of a typical rock is ~1 J g-1deg-1. With these values, it would take over 107years for radioactivity in the mantle to increase the temperature in a mantle rock 1oC, making radioactivity an unlikely source for melt. 0 5 10 15 20 25 0100020003000 ( C) 0 100 200 300 00 500 600 700 Liquidus solidus O ceanic Geotherm Diagram after Wyllie, P.J., 1981. Plate tectonics and magma genesis. Geol. Rundsch, 70:128-153. 6 Changing composition Although the mantle is much more hydrated in subduction zones, amphibole and phlogopitehave been found in mantle xenoliths. Water, therefore, is present in the mantle. Other fluid inclusions contain liquid CO2. At high pressure, water can dramatically decrease the solidus temperature of the mantle. The water content of normal mantleis only ~0.1 weight %. In order to produce melting, however, additional volatiles need to be added to the mantle. The following slide has a simple animation to show the process of changing the composition of the mantle. Melts within the mantle are thought to be merely droplet sized.

PAGE 101

89 7 Melting begins when a flux is added to the system (mantle). Flux is a substance that reduces the melting point of the mixture when it is added to a mixture Water is one of the most dominant fluxes. It is added to the mantle during subduction processes. Initial melts are contained within the intergranularspace in a rock and are adsorbed to grain surfaces. Press the enter key now to begin the animation. Water added Melting begins containing highest concentration of incompatible elements Liquid rises due to density difference As drops of melt accumulate, concentration of incompatible elements decreases Mantle Residue depleted in incompatible elements 8 Melting, 1 The binary phase diagram below will help us see that melting in the mantle happens at the solidus, above which crystals and melt exist together for a large range of temperatures. At the eutectic point (A), the temperature is 1542oC. For a different phase to occur, the temperature needs to increase to 1557oC (point B). Diagram after Bowen, N.L., 1956. The Evolution of Igneous Rocks Dover Publications, Inc., New York. 332 pp. 1500 1600 1700 1800 1900 2000 0102030405060 MgO (wt %)T (oC)1500 1600 1700 1800 1900 2000 A BSiO2Mg2SiO4MgSiO3 Qtz + liquid Qtz + enstatite liquid enstatite +liquid

PAGE 102

90 9 Melting, 2 This diagram is an enlargement of part of the previous one. Print this page and use the lever rule to calculate the percent of enstatite(MgSiO3) and quartz present at the location of the red dot. Only solids exist at this point. If you need help with the lever rule, click here 1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600 010203040 MgO (wt %)T (oC)1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600 ASiO2MgSiO3 Qtz + enstatite Qtz + liquidEnstatite + liquid Liquid 10 Melting, 3 By increasing the temperature by a degree or two, we have shifted the position of the red dot from a solid into an area of solid and melt. Now use the lever rule again to calculate the percent of enstatiteand liquid present at the new location. You should see that a slight increase in temperature will greatly increase the amount of liquid present. 1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600 010203040 MgO (wt %)T (oC)1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600 ASiO2MgSiO3 Qtz + enstatite Qtz + liquidEnstatite + liquid Liquid

PAGE 103

91 11 Consider a mantle rock with a composition of 74% olivine, 15% orthopyroxene (opx), 10% clinopyroxene(cpx), and 1% spinel. How much melting occurs? From BowenÂ’s reaction series, we know that the pyroxenes will bethe first to melt. If they melt equally at 50% cpxto 50% opx, how much melting must occur before all of the cpxis consumed and a phase change takes place? We know that olivine forms as orthopyroxene melts, but the amount is small enough to disregard it in this calculation. BCD 2% mineral% melt 3olivine0.74 4opx0.150.1 5cpx0.10.1 6spinel0.01 7ratio1:1 8% melt0.2 A value larger than 20% melting is much higher than can occur according to experiments. This means that real-world melts do not get past the eutectic point and that an entire mineral phase will not melt. 12 1.How can we obtain mantle-derived samples such as the one described on slide 5? (ophiolites, dredge samples, xenoliths in basalts, xenoliths in kimberlites, possibly stony meteorites.) 2.Define the terms liquidusand solidus. 3.Determine the temperature in oFfor the eutectic (point A) on slide 8. 4.Can hotspots be a means of adding heat to the mantle? (They can add heat to the mantle, but they are local perturbations and cannot produce all the basalt seen at the surface of the Earth.) 5.How do we know the mantle is predominantly solid? (Geophysical studies show that S-waves cannot propagate through a liquid; therefore, we know that the mantle is predominantly solid and the outer core is liquid.) 6.Using the values on slide 5, how long would it take for radioactivity in the mantle to increase the temperature in mantle rock 1oC? (1J/g deg / 10-8J/g a = 1 deg C / 107a) End-of-module questions Continue with partial melting module JAH2B.

PAGE 104

92 Partial Melting Module JAH2B 1 The advances in geochemistry, geophysics, and volcanology have increased our understanding of the earth and the processes that shape and formit. It is recommended that students work through JAH2A before working through this part of the module. Mantle Processes 2B –Partial Melting Module JAH-2B“The only way we can form any notion of what goes on at greater depths, is through volcanoes; they, therefore, deserve the careful study ofany one who wishes to know the little that can be learned of the vast unknown region of the earth’s interior.”First Book in Geology (Shaler, 1884) 2 PREVIEW Slide 3 introduces the Shaw equation used for partial melting. Slides 4-6 ask the students to develop a spreadsheet to calculate the Cl/Coratio for given D values and then to graph those values versus F to see the range of concentration of incompatible and compatible elements. Slide 7 uses a spreadsheet mentioned in the Fractional Crystallization module (JAH1A) to calculate bulk partition coefficients. Slides 8-9 ask students to use the Shaw equation to calculate and graph Cl/Covalues for incompatible trace elements to see how concentrations vary with progressive melting. Slides 10-11 offer some guidelines on what type of samples to choose, whatelements to plot, and how to determine if partial melting has played a role in the formation of a given suite of rocks. Slides 12-17 end-of-module questions and related data.

PAGE 105

93 3 Because the concentration of trace elements with D ’s 0vary inversely with the amount of liquid diluting it, their concentration in the liquid reflects the proportion of liquid at a given state of melting. The batch-melting model is a simple model in which the melt and solid remain in equilibrium until it is released and ascends as an independent system. The batch-melting equation listed below was derived by H. R. Shaw: Co= trace element concentration in original assemblage Cl= trace element concentration in the liquid D = bulk partition coefficient F = weight fraction of melt produced F F D C Co l ) 1 ( 1 PARTIAL MELTING –The Shaw Equation Now that we’ve looked at processes of melting in the mantle, we can study a model that detects when samples have been affected by partial melting. 4 The graph should help you visualize how dilution of the concentration of elements occurs as melting progresses. Create a spreadsheet using the formula for batch melting as shown below. Plot your results as Cl/Cofor each D value vs. F. Use a logarithmic scale for the y -axis. BCDEFGHI 2 D 00.30.6139 3 Melting: Cl/Co = 1/((1-F)D + F) 4F1-FD = 0D = 0.3D = 0.6 D = 1D = 3D = 9 50.01 60.05 70.1 80.2 90.3 100.4 110.5 120.6 130.7 140.8 150.9 161 MELTING CALCULATION

PAGE 106

94 5 When D >> 1, the concentration of compatible elements in the melt show only small ranges in abundance during initial melting. For small values of F (low degrees of melting), the concentration of highly incompatible elements ( D << 1) varies greatly. The highly incompatible elements concentrate in the first few drops of melt and become progressively more dilute as F increases. PARTIAL MELTING MELTING0.1 1.0 10.0 100.0 FCl/Co D = 0 D = 0.3 D = 0.6 D = 1 D = 3 D = 9 6 LetÂ’s see how this technique works. If we know the concentration of a trace element in a magma (Cl) and D, we can use the Shaw equation to calculate Coenabling us to characterize and constrain the source region of magmas. When Fapproaches zero, the equation reduces to: D C Co l1 By knowing the concentration of a highly incompatible element (D 0)in the magma and the source rock, we can calculate F, the fraction of partial melt produced. F C Co l1 PARTIAL MELTING Only those elements with DÂ’s <<1 show a wide range in abundance for a given range of F. For example, the maximum enrichment for an element with D= 0.01 is 100 while an element with D= 0.3 the maximum enrichment is 3.3.

PAGE 107

95 7 Suppose we have a sample of basalt derived from melting in the mantle with a composition of 70% olivine, 17% orthopyroxene (opx), 12% clinopyroxene (cpx), and 1% spinel. Use the spreadsheet you created in slide 7 of module 1A and KDvalues for Rband Li from the Geochemical Earth Reference Model (GERM) website listed below to calculate D for each element. You should discover that Rbis highly incompatible and that Li is only slightly more compatible than Rb(moderately incompatible). Now we can calculate Cl/ Cofor different F values. COMPATIBILITY BCD 2 XKD3Olivine 0.70.04 4opx 0.170.0006 5cpx 0.120.01 6spinel0010 7 8 DRb = BCD 2 XKD3Olivine 0.70.35 4opx 0.170.26 5cpx 0.12 6spinel 0.01 7 8 DLi = 8 Using the Shaw equation, calculate the Cl/ Coratio for Rband Li. Add another column to calculate the Rb/Li ratio. After completing the spreadsheet, plot the Cl/ Coratios for the two elements vs F on a linear graph with arithmetic scales on both axes. BCDE 2RbLi 3 D 0.030.29 4 5 F Rbl / RboLil / LioRb/Li 60.05 70.1 80.15 90.2 100.25 110.3 120.4 130.5 140.6 150.7 160.8 170.9 181

PAGE 108

96 9 You should see from your graph (1) that Rb, a highly incompatible element, is concentrated in the earliest stages of melting; (2) that it is strongly partitioned from the source rock; and (3) that the Rb/Li ratio decreases as melting increases. When looking at a data set, we donÂ’t know what process formed the suite of rocks. They may have formed from fractional crystallization or partial melting. So, what must we do to identify partial melting? The concentration of incompatible elements becomes more dilute as melting continues. Partial Melting0 2 4 6 8 10 12 14 FCl/Co Rb Li Rb/Li Li on the other hand, is only moderately incompatible and does not vary much with progressive melting. 10 1.You must look at a suite of rocks, not just one rock. 2.The rocks should have MgOcontents > 7 wt. %. However, MgOcontents > 15 wt % are probably due to olivine accumulation. 3.The rock should be glassy and should not contain excess olivine phenocrysts. CHOOSING SAMPLES 1.The major element compositions should not vary much among the samples. 2.Incompatible elements will vary significantly. Highly incompatible elements will vary even more. 3.Compatible elements will show at most a slight variation. EXAMINING THE GEOCHEMISTRY IDENTIFYING PARTIAL MELTING

PAGE 109

97 11 A steeply-sloped, tight linear trend (A) indicates partial melting as a dominant process while a shallower slope with a more scattered trend (B) indicates some other process such as fractional crystallization (but not necessarily). EXAMINING THE GRAPHS Because the concentration of highly incompatible elements has such a large range, and moderately incompatible elements vary only slightly, a plot on a linear graph of the ratio of a highly incompatible element to a moderately incompatible element vs the moderately incompatible element is best (Example: Be/Li vs.Be) What elements are the best to plot? 0 2 4 6 8 10 12 14 020406080100120 highly incompatiblehighly incompatible/mod. incompatible 0 1 2 3 4 5 6 01020304050 highly incompatiblehighly incompatible/mod. incompatible AB 12 1.As F approaches 1, what can we say about the Cl/ Coratio in the Shaw equation? (The concentration of every trace element in the melt equals the concentration in the source rock, it approaches 1.) 2.Make plots of the data from various settings on the following slides (You will need to use the spreadsheet and rock compositions from slide 7 and KDvalues from the GERM website to calculate D values so that you will know which trace elements to plot). Which suite of rocks, if any, indicates that partial melting was a dominant process of formation. 3.After plotting all the data, create a spreadsheet like the example below to calculate F for each sample in any suite of rocks that appears to have been formed from partial melting. Be sure to use element(s) with a D value close to zero (see slide 6): End-of-module questions BCD 2Co30.02ppm 4 5ClF60.20510 70.20710 80.2578 90.2578 100.16412 110.2428 120.3825 Cofor Be ~ 0.01 Cofor Ba~ 1.1 Cofor Li ~ 1.9 Cofor Zr~ 9.0 Cofor Y ~ 4.0

PAGE 110

98 13East Pacific Rise 5-13 N sampleSiO2MgOTiO2LiBeBaZrY ch123-849.77.472.22717449 ch21-949.759.351.173.9412.586.425.4 ch19-249.78.531.295.070.4179128.7 ch2-649.57.821.6127.114034.3 ch112-149.767.561.516.360.54413.811133.7 ch38-150.756.632.177.550.9531.8177.544.4 ch82-360.52.581.3318.52.154747384.9 ch41-1350.5562.12722258.5 ch11-750.58.151.338.995.531.3 ch122-551.57.051.7226.812438.7 ch122-650.87.531.411297.132.8 ch116- CH96-649.287.371.6112.111336.3 ch73-256.63.651.4516.92.4380.546481.4 ch103-651.53.792.9716.61.9244.1380 ch116- ch26- ch4- ch19-5498.791.245.070.415.910531 ch21-1049. The East Pacific Rise is a divergent boundary. Data: Langmuir, C.H., Bender, J.F., and Batiza, R., 1986. Petrologic and tectonic segmentation of the East Pacific Rise, 5o30-14o30N. Nature 322:422-429. Basalt photos: 14FAMOUS Area MORBS (Mid-Atlantic Ridge) sampleSiO2MgOTiO2LiBeBaZ r Y fam 528-4 47.0410.840.673.650.20520.253.622.3 fam530-348.3611.080.743.50.20720.8 fam525-52 48.411.20.814.180.25729.5 fam525-52 48.411.20.814.180.25725.545 fam527-148.4211.040.653.70.1642047.823.7 fam528-12 48.810.490.778 29.25825.9 fam518-22 49.110.420.823.60.24221.95420.9 fam526-2a49.369.6851.054.75 3975.526.2 fam 527-449.628.21.415.7 6910432 fam523-3a49.758.331.425.78 739532.2 fam529-449. fam523-250.048.311.4 1009828.5 A ii77-76-6 6 50.386.741.85 60.7 The FAMOUS (French American Mid-Ocean Undersea Study was performed in an area around the 45oN latitude where ALVIN could be used to collect samples. DATA: Langmuir, C.H., 1980. A major and trace element approach to basalts. Ph.D. thesis, S.U.N.Y at Stony Brook. Langmuir, C.H., Bender, J.F., Bence, A.E., Hanson, G.N. and Taylor, S.R., 1077. Petrogenesisof basalts from the FAMOUS area: Mid-Atlantic Ridge. Earth and Planetary Science Letters 36:133-156. Bryan, W.B., and Moore, J.G., 1977. Compositional variations of young basalts in the Mid-Atlantic Ridge rift valley near lat. 36o49N. Geological Society of America Bulletin 88:556-570. Bryan, W.B., Thompson, G. and Michael, P.J., 1979. Compositional variation in a steady state zoned magma chamber: Mid-Atlantic Ridge at 36o50N. Tectonophysics 55:63-85. PHOTOS: Upper right: acoustics/shipops.html Lower right: Left:

PAGE 111

99 15Juan De Fuca Ridge sampleSiO2MgOTiO2LiBeBaZrY jfd01-0450.813.4172.99716.751.6647.45403101 jfc02-0150.65.082.4910.51.24229475 aii84-wf74 49.696.052.12 26.6205.557.09 aii84-wf12 49.486.961.827.520.58616.7127.842 aii84-wf13 49.527.711.446.10.41214.595.633.9 1461-7r49.746.91.83 17.6135.142.2 1411-b148.867.451.465.270.49827.710131 1410-2a508.61.23 138128.7 jfd-250.146.751.867.80.652215948 JFD03-0350.917.31.8 2012942 1405-B549.4981.38 9.6395.134.8 1415-B5-249.127.541.59 1410938.5 JFD02-0150.97.21.77 1712340.3 JFD11-0248.37.91.41 2810430 aii84-WF2 50.971.8 2014745.5 1406-2B-749.68.631.16 137628.5 1410-3B49.48.61.23 138829.5 The Juan de Fucaridge is a divergent boundary off the coast of western Canada. Kappel, E., 1985. Evidence for volcanic episodicityand a non-steady state rift valley. Ph.D. thesis, Columbia University. 16Society Islands SampleSiO2 MgO TiO2LiBeBaZrY 1012874.51.0 1012886.21.1 1012112.906.31.0513.8210119852.14.32.839.32.7 MAO-1133.216.61.5355.25MAO 65 3.61.5433.2273-20445.712.52.765.91.3410.39MHT-10144.215.73.323.71.1304.58MHT-15544.814.63.564.31.2344.85Ta3F50. The Society Islands, French Polynesia, were formed by a hot spot like the islands of Hawaii. White, W.M., Harden, J.A., 2004. Light Element and Lithium Isotope Signaturesof the EM II Reservoir: The Society Islands, French Polynesia: Geochemical Results and an Educational Appication. M.S. thesis Universit y of South Florida.

PAGE 112

100 17 N Moorea is the third youngest island in the Society Chain.Moorea SampleSiO2 MgO TiO2LiBeBaZrY M01FISH2D46.014.52.704.91.278 M01FISH A45.413.32.935.11.2211 M01T1A45.811.22.935.41.0270 M01FISH2B46.013.42.894.71.5222 M01FISHB146. M01FISH2A46. M03PK19N46.414.12.745.80.7320 M01B2C46.813. M01AIR 1A46.911. M0302047.014.12.566.91.1309 M0301147.214.12.621232.9298 M0301047.614.52.746.71.1244 M01B1A47.6793.186.61.0268 M01QB49.0922.591132.8251 M01T2A49.0893.347.81.8353 M01QA49.48.42.711323.0249 To the right are thin sections of basalts from the island. The middle photo is of a xenolith.Harden, J.A., 2004. Light Element and Lithium Isotope Signaturesof the EM II Reservoir: The Society Islands, French Polynesia: Geochemical Results and an Educational Application. M.S. thesis, University of South Florida.


Download Options

Choose Size
Choose file type
Cite this item close


Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.


Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.


Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.


Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.