저 시-비 리- 경 지 2.0 한민

는 아래 조건 르는 경 에 한하여 게

l 저 물 복제, 포, 전송, 전시, 공연 송할 수 습니다.

다 과 같 조건 라야 합니다:

l 하는, 저 물 나 포 경 , 저 물에 적 된 허락조건 명확하게 나타내어야 합니다.

l 저 터 허가를 면 러한 조건들 적 되지 않습니다.

저 에 른 리는 내 에 하여 향 지 않습니다.

것 허락규약(Legal Code) 해하 쉽게 약한 것 니다.

Disclaimer

저 시. 하는 원저 를 시하여야 합니다.

비 리. 하는 저 물 리 목적 할 수 없습니다.

경 지. 하는 저 물 개 , 형 또는 가공할 수 없습니다.

이학석사 학위논문

The Pleistocene adakitic rocks

in the western Panamanian arc:

slab melt or evolved mantle-

derived magma?

중앙아메리카 서파나마 지역의

플라이스토세 아다카이트질 화산암:

슬랩 용융체인가 맨틀기원의 마그마인가?

2018 년 8 월

서울대학교 대학원

지구환경과학부

제 갈 유 진

Master’s Thesis of Yujin Jegal

The Pleistocene adakitic rocks

in the western Panamanian arc:

slab melt or evolved mantle-

derived magma?

August 2018

Graduate School of

Earth and Environmental Sciences

Seoul National University

Yujin Jegal

i

Abstract

The Pleistocene volcanic rocks from western Panama and southern Costa Rica,

Central America are characterized by adakite-like geochemical features of high

Sr/Y (>50) and La/Yb (≥20) with low Y (<15 ppm), low Yb (<1.5 ppm) and low

87Sr/86Sr (<0.704). Early studies argue that they are slab melts with various degrees

of mantle interaction during ascent whereas recent studies suggest that they are

formed by high-pressure (≥30 km depth) differentiation of mantle-derived basaltic

arc magma.

In order to better understand their petrogenesis and test the previously

suggested hypotheses, we have investigated major and trace element and Nd, Sr

and Pb isotope geochemistry and mineralogy of seven volcanic rocks of western

Panama where the Cocos ridge subducts.

Our results show that they are characterized by enrichments in LILE with high

LREE abundances and pronounced depletions in the HREE and HFSE, confirming

the large influence of enriched slab components suggested by previous studies. The

Nd, Sr and Pb isotope data are consistent with those of arc basalts that erupted at

about 2 Ma prior to the western Panamanian adakitic rocks in the same region,

which originated from enriched mantle metasomatized by melts from the

subducting Cocos ridge. However, they have slightly more radiogenic Sr isotopic

ratios than the Cocos Ridge basalts and other Galapagos OIBs.

The compositions of amphibole phenocrysts from the adakitic rocks yield

pressure, temperature and fO2 conditions of ~ 220-960 MPa, ~ 721-1009 °C, ~

NNO - 0.1-NNO + 2.2, suggesting the magmas have experienced crystallization at

multiple depths with large variations in temperature and fO2 conditions.

Crystallization pressure and temperature of clinopyroxene and orthopyroxene

phenocrysts from basaltic rocks of 600-1300 MPa and 1020-1170 °C respectively

show early-stage magma differentiation in lower crust or at crust-mantle boundary

beneath the southern Central American volcanic arc (CAVA).

The calculated melts in equilibrium with high pressure amphiboles (≥920

MPa) have a wide range of Mg# (0.31-0.56) and SiO2 (58.4-71.3 wt.%), which

reflects significant degree of magma differentiation within a crustal magma storage

region (crystal mush zones) underneath the western Panamanian arc. The Mg# of

ii

melts in equilibrium with clinopyroxene phenocrysts in the mafic and felsic rocks

range from 0.41 to 0.68 and they are consistent with Mg# (0.51-0.66) of the basalt

and basaltic andesites in this study, suggesting that pyroxene phenocrysts in the

dacite are from a cognate basaltic magma. Petrology, mineral chemistry and whole-

rock data of the western Panamanian adakitic rocks indicate the extensive magma

evolution from basalt to dacite at high pressure (~35-40 km depth).

We performed trace element and isotope modeling to examine the previously

suggested models for the formation of the adakitic rocks from the southern CAVA.

The results show that the compositional variations of the Pleistocene adakitic rocks

from western Panama and southern Costa Rica are best explained by garnet- and

amphibole- dominated fractional crystallization of basaltic arc melts whose mantle

source was metasomatized by slab melts from the subducting Cocos ridge. This is

consistent with the adakitic features of the calculated melts in equilibrium with

high pressure amphiboles and diagnostic trace elements (La/Yb, Yb, Nb/Yb,

Sm/Yb, Dy/Yb) that show co-crystallization of garnet and amphibole in the

adakitic rocks from western Panama and southern Costa Rica during magma

evolution.

The presence of amphibole-rich cumulative enclave, disequilibrium textures

and compositions in the phenocrysts of the evolved adakitic rocks are consistent

with magma mixing and low pressure fractional crystallization models, suggesting

that they may also have influenced on their geochemical variations of the

Pleistocene adakitic rocks from western Panama and southern Costa Rica.

Keyword: adakitic rock; western Panama; amphibole; pyroxene; slab melts;

mantle-derived melts; magma mixing; fractional crystallization

Student Number: 2016-20429

iii

TABLE OF CONTENTS

Abstract...........................................................................................................i

Table of Contents ..........................................................................................iii

List of Figures ................................................................................................vi

List of Tables ..................................................................................................xii

List of Appendices ........................................................................................xiii

Chapter 1. The Pleistocene adakitic rocks in the western Panamanian

arc: slab melt or evolved mantle-derived magma?

1. Introduction ...............................................................................................1

2. Geological Background and Samples ..................................................4

3. Analytical Methods ..................................................................................7

3.1 Mineral chemistry analysis ............................................................. 7

3.2 Whole-rock major and trace element analysis ................................ 7

3.3 Whole-rock Sr-Nd-Pb radiogenic isotope analysis......................... 8

4. Results .........................................................................................................10

4.1 Petrography and mineral chemistry .....................................................10

4.2 Amphibole compositions ......................................................................19

4.3 Whole-rock geochemistry .....................................................................22

5. Discussion ...................................................................................................28

5.1 Magma storage depth and host melts of western Panamanian adakitic

rocks .........................................................................................................28

5.1.1 Pre-eruptive conditions................................................................28

iv

5.1.2 Equilibrium melt composition ............................................... 34

5.2 Formation of Quaternary adakitic rocks in the western Panamanian

arc .............................................................................................................37

5.2.1 Melting of the subducting oceanic crust ...................................37

5.2.2. Fractional crystallization of mantle-derived arc magmas and

magma mixing .............................................................................40

5.2.3. The role of crystal mush in the western Panamanian adakitic

suites .............................................................................................43

5.3 Petrogenesis of the adakite-like magmas in the western Panamanian

and southern Costa Rican arc ...............................................................44

6. Conclusion ..................................................................................................45

Chapter 2. Platinum Group Element recycling in the Central American

volcanic arc: a preliminary study

1. Introduction ...............................................................................................46

2. Geological Background and Samples ..................................................48

3. Analytical Methods ..................................................................................50

3.1 Whole-rock trace element analysis

3.2 PGE analysis

4. Results .........................................................................................................52

4.1. Whole-rock geochemistry ....................................................................52

4.2. Platinum group element geochemistry ...............................................55

v

5. Discussion ...................................................................................................57

5.1. Effects of magma differentiation ........................................................57

5.2. Effects of mantle process and future work ........................................58

Appendix .........................................................................................................60

References ......................................................................................................77

Abstract (In Korean) ...................................................................................91

Acknowledgement ........................................................................................93

vi

LIST OF FIGURES

Figure 1.1. Tectonic setting of Central America modified from Gazel et al. (2009).

Grey shaded area illustrates approximate area for the adakitic rocks from western

Panama and southern Costa Rica (Gazel et al., 2011). Dashed line in the

continental areas indicates the location of the Central American Volcanic Arc

(CAVA). Miocene (>10 Ma) and Pliocene (~3.6 Ma) mantle-derived arc rocks in

southern Costa Rica are from Gazel et al. (2009). The location of DSDP Site 504,

Costa Rica Rift (Hubberten et al., 1983), and Cocos-Nazca Spreading Center

(CNS) crust (Harpp et al., 2004) is presented as star symbols. PFZ: Panama

Fracture Zone, EPR: East Pacific Rise.

Figure 1.2. Microphotographs of basaltic andesite (sample M91aKH). (a)

glomeroporphyritic aggregates of clinopyroxene grains. (b) glomeroporphyritic

aggregates of clinopyroxene with orthopyroxene. (c) subhedral to euhedral olivines

phenocrysts and lath-shaped plagioclases in the groundmass.

Figure 1.3. Pyroxene compositions in basaltic andesite (M91aKH), trachyandesite

(M38KH) and dacite (M63aKH-b). (a) Ca-Mg-Fe Pyroxene classification diagram

(Morimoto, 1988). (b and c) binary plots of Mg# vs. Al (apfu) for clinopyroxene

and orthopyroxene.

Figure 1.4. Microphotographs of trachyandesite (sample M38KH). (a) some

plagioclase crystals display mingling texture. (b) glomeroporphyritic aggregates of

clinopyroxene crystals. (c) quartz surrounded by reaction rims of clinopyroxene.

Figure 1.5. Microphotographs of dacites (sample M63aKH and M63aKH-b). (a)

plagioclase and biotite in M63aKH. (b) plagioclase and biotite in M63aKH-b. (c)

plagioclase co-crystallized with biotite. (d) clinopyroxene in M63aKH-b. (e)

orthopyroxene in M63aKH-b. (f) embayed and rounded quartz in M63aKH.

Figure 1.6. Microphotographs of dacites (sample M63aKH and M63aKH-b). (a, b

and c) amphiboles in M63aKH. (d) amphiboles in M63aKH-b. Red dotted circles

vii

are locations of laser ablation ICP-MS analyses. The values next to red circles are

the beam sizes of LA-ICP-MS analyses. Filled circles are locations of Field-

Emission EPMA. The compositions of mauve and purple circles indicate low and

high pressure amphiboles, respectively.

Figure 1.7. Mg# vs Si classification diagrams for calcic amphiboles (a) Na + K

<0.5 and (b) Na + K ≥0.5. Mauve and purple circles indicate amphiboles formed at

low (220 MPa) and high (≥920 MPa) pressures, respectively, discussed in section

5.1.1.

Figure 1.8. Binary plots of (a) total Al vs stoichiochemical Na+K and (b) total Al

vs Mg# of amphiboles in the western Panamanian adakitic rocks. The fields of

experimental amphibole compositions are obtained at 220 and 400 MPa (Prouteau

et al., 1999; Scaillet and Evans, 1999), 920 and 1200 MPa (Alonso-Perez et al.,

2009) and 960 MPa (Prouteau et al., 1999) and modified from Ribeiro et al. (2016).

Figure 1.9. Chondrite normalized REE patterns of (a) low-P (220 MPa) and (b)

high-P (≥920 MPa) amphibole phenocrysts.

Figure 1.10. (a) Geochemical classification of samples in this study and adakitic

rocks from western Panama and southern Cost Rica from Gazel et al. (2011) and

Hidalgo and Rooney (2014). (b) Primitive mantle (McDonough and Sun, 1995)

normalized multi-element diagrams of the adakitic rocks in this study and adakitic

rocks from western Panama and southern Costa Rica (shaded area). Average

Pliocene mantle-derived arc basalts from southern Costa Rica, spatially associated

with the adakitic rocks, are from Gazel et al. (2009). The composition of average

MORB is from Gale et al. (2013).

Figure 1.11. Binary plots of (a) Sr/Y vs Y and (b) La/Yb vs Yb to discriminate

adakites from calc-alkaline arc lavas.

Figure 1.12. Binary plots of (a) La/Yb vs SiO2, (b) Yb vs SiO2, (c) Sm/Yb vs SiO2,

(d) Dy/Yb vs SiO2 and (e) Nb/Yb vs SiO2 of western Panamanian and southern

viii

Costa Rican adakitic rocks. Solid and dotted arrows represent variations in melt

composition during garnet and amphibole fractionation, respectively.

Figure 1.13. (a and b) Pb-Nd-Sr isotope systematics for the adakitic rocks in this

study and possible sources. The Cocos ridge rocks as possible sources of the

adakitic rocks from western Panama and southern Costa Rica are from Hoernle et

al. (2000). Red solid lines are mixing curves between the mean Cocos ridge basalts

and the subducting sediments. The sediments are assumed to be a mixture of 70%

hemipelagic sediments and 30% carbonate sediments from DSDP Site 495 (Patino

et al., 2000). Pliocene mantle-derived arc basalts (~3.6 Ma) in southern Costa Rica

are from Gazel et al. (2009). Altered oceanic crust (ODP504 MORB) in the eastern

Pacific is from Shimizu (1989). Galapagos oceanic island basalts are from White et

al. (1993).

Figure 1.14. Binary plots of (a) total Al vs stoichiochemical Si+Mg and (b)

tetrahedral Al vs octahedral Al of amphiboles in the western Panamanian adakitic

rocks (M63aKH and M63aKH-b).

Figure 1.15. Binary plots of (a) Si vs tetrahedral Al plus K and Na and (b)

stoichiochemical Ti vs octahedral Al of amphiboles in the western Panamanian

adakitic rocks (M63aKH and M63aKH-b).

Figure 1.16. Calculated pressure and temperature estimated using clinopyroxene-

liquid and clinopyroxene-orthopyroxene pairs which satisfy Fe-Mg equilibrium for

crystals in basaltic andesite (M91aKH), trachyandesite (M38KH) and dacite

(M63aKH-b). Temperature and pressure precisions are about ±~29 °C and ±~300

MPa, respectively.

Figure 1.17. Mg# of whole-rocks and calculated melts in equilibrium with

clinopyroxene cores and amphiboles.

Figure 1.18. (a) Chondrite-normalized REE patterns of the melts equilibrated with

the high-P amphiboles in the dacite (M63aKH) and the host rock (M63aKH). The

ix

chondrite values are from Sun and McDonough (1989). (b) Binary plots of Sr/Y vs

Y for the equilibrium melts. The parental melts were calculated using high-P

amphibole compositions and high pressure (700 MPa) experimental partition

coefficients between hornblende and basaltic andesite (Nandedkar et al., 2016).

Figure 1.19. Binary plots of (a) Sr/Y and (b) La/Yb. The solid line, black and grey

dashed lines indicate partial melting of the Cocos ridge basalts average (Harpp et

al., 2004), normal oceanic ridge basalts produced at Costa Rica Rift (Hubberten et

al., 1983) and the Malpelo ridge (Harpp et al., 2004), respectively. The calculated

melt compositions (residual 50% Cpx and 50% Grt) are illustrated with varying

degree of partial melting (tick marks, per 10%). The location of DSDP Site 504,

Cost Rica Rift, and CNS crust is presented in Figure 1.1.

Figure 1.20. Binary plots of (a) Sr/Y and (b) La/Yb. The solid line, black and grey

dashed lines indicate partial melting of the Cocos ridge basalts average (Harpp et

al., 2004), normal oceanic ridge basalts produced at Costa Rica Rift (Hubberten et

al., 1983) and the Malpelo ridge (Harpp et al., 2004), respectively. The melt

compositions are calculated using the residual mineralogy of a hornblende eclogite,

following the approach of Drummond and Defant (1990) and illustrated with

varying degree of partial melting (tick marks, per 10%).

Figure 1.21. Binary plots of (a) Sr/Y and (b) La/Yb. The black solid line shows

fractional crystallization of a mantle-derived arc magma (Pliocene arc magma; ~3.5

Ma; Gazel et al., 2009), using a high pressure mineral assemblage (Müntener et al.,

2001) and partition coefficients from Nandedkar et al. (2016), Sen and Dunn

(1994) and Green et al. (2000). The grey solid line indicates fractional

crystallization at low pressure, using a low pressure assemblage based on the

phenocrysts in the evolved adakitic rock from western Panama. The fractional

crystallization melt compositions are illustrated with the fraction of melt remaining

(tick marks, per 0.1). The dotted line represents the mixing model between a

mantle-derived magma and dacite magma (tick marks, per 10%).

Figure 1.22. Binary plots of (a) Sr/Y and (b) La/Yb. The black solid line illustrates

x

fractional crystallization of a mantle-derived arc magma (Pliocene arc magma; ~3.5

Ma; Gazel et al., 2009), using the mineral assemblage without garnet (Cpx : Opx :

Grt : Amp = 52.8 : 17.2 : 0 : 30). The partition coefficients are identical with those

in Figure 1.21.

Figure 1.23. (a) Schematic illustration showing the magma plumbing systems for

adakitic rocks in western Panama and southern Costa Rica, Central America. (b)

Pressure and temperature conditions of western Panamanian adakitic magmas

during ascent and cooling. Arrows indicate adiabatic magma ascent and horizontal

lines represent cooling and crystallizing paths. The liquidus and amphibole-in

boundaries for basalt and dacite are modified from Smith (2014).

Figure 2.1. Tectonic setting of Central America modified from Gazel et al. (2009).

PFZ: Panama Fracture Zone, EPR: East Pacific Rise, CNS: Cocos-Nazca

Spreading Center.

Figure 2.2. TAS diagram of the arc lavas and adakitic rocks in this study. Grey

circles indicate volcanic front lavas of the southern Central America (Carr et al.,

2014).

Figure 2.3. Primitive mantle (McDonough and Sun, 1995) normalized multi-

element diagrams of the rocks from (a) Nicaragua and (b) central Costa Rica and

western Panama. The composition of average MORB is from Gale et al. (2013) and

illustrated by dotted lines. Shaded areas indicate (a) the volcanic front lavas of

Nicaragua from (Carr et al., 2014) and (b) central Costa Rican lavas and western

Panamanian adakitic rocks from Carr et al. (2014), Gazel et al. (2011) and Hidalgo

and Rooney (2014).

Figure 2.4 U/Th and Ba/Nb vs distance along the volcanic front (km) diagrams.

Distance along the volcanic front (km) is measured from a point about 22 km NW

of Tacana volcano, northern Guatemala, Central America (Carr et al., 2007). Small

circles and triangles represent volcanic front lavas of the southern Central America

are from Carr et al. (2014). Small squares indicate the adakitic rocks from western

xi

Panama and southern Cost Rica are from Gazel et al. (2011) and Hidalgo and

Rooney (2014).

Figure 2.5. Nb/Yb and La/Yb vs distance along the volcanic front (km) diagrams.

Distance along the volcanic front (km) is and the references are identical with

Figure 2.4.

Figure 2.6. Binary plots of PGE concentrations (ppb) vs MgO (wt.%), indicating

indices of fractionation for the Central American volcanic lavas. Dotted lines

indicate duplicate analyses.

Figure 2.7. Binary plots of (a) Pd/Cu*1000 vs MgO (wt.%), (b) Pd/Pt vs MgO

(wt.%) and (c) Pd/Ir vs MgO (wt.%) for the Central American volcanic lavas. The

volcanic rocks with arrows are minimum values of Pd/Ir ratios. Vertical dotted lines

indicate duplicate analyses.

Figure 2.8. Primitive mantle-normalized (McDonough and Sun 1995) PGE patterns

for primitive arc lavas from Nicaragua, central Costa Rica and western Panama.

Figure 2.9. Pd concentration (ppb) and Pd/Ir ratio vs distance along the volcanic

front (km) diagram. Vertical dotted lines indicate duplicate analyses.

xii

LIST OF TABLES

Table 1.1. Mg number (#) and SiO2 (wt.%) of the rocks, pyroxenes, and

amphiboles and their equilibrium melts.

Table 1.2. Major element compositions of representative amphiboles in dacites

(M63aKH and M63aKH-b) from western Panama, Central America.

Table 1.3. Major element (wt.%) and trace element (ppm) compositions of western

Panamanian adakitic volcanic rocksa.

Table 1.4. Pb-Nd-Sr isotopic compositions of western Panamanian adakitic

volcanic rocks.

Table 1.5. Representative clinopyroxene analyses and thermobarometry for

clinopyroxene-liquid (whole-rock) pairs.

Table 1.6. Thermobarometry for clinopyroxene-orthopyroxene pairs and their

compositions

Table 2.1. Procedural blanks (ng/g).

Table 2.2. Analytical results of PGE in PGE reference material TDB-1 (ng/g).

Table 2.3. Concentrations of MgO (wt.%), PGE (ng/g) and trace element ratios of

volcanic lavas from the CAVA.

xiii

LIST OF APPENDICES

Appendix 1. Electron microbe data of standards, analyzed at KOPRI.

Appendix 2. LA-ICP-MS analysis results (KOPRI) of NIST 612 glass and BCR-2g.

Appendix 3. LA-ICP-MS analysis results (KBSI) of BCR-2g and AGV-2.

Appendix 4. Trace elements analysis results (KOPRI) of BHVO-2 and AGV-2.

Appendix 5. Major elements analysis results (SNU) of AGV-2.

Appendix 6.1. Electron microbe data of minerals of M91aKH.

Appendix 6.2. Electron microbe data of minerals of M38KH.

Appendix 6.3. Electron microbe data of minerals of M63aKH.

Appendix 6.4. Electron microbe data of minerals of M63aKH-b.

Appendix 7. Major element variation diagrams against SiO2 (wt.%) for the adakitic

rocks from western Panama and southern Costa Rica. Yellow squares and small

lemon squares represent western Panamanian and southern Costa Rican adakitic

rocks in this study and from Gazel et al. (2011) and Hidalgo et al. (2014),

respectively.

Appendix 8.1. Details about fractional crystallization model of a fractionated

magma in section 5.2.2.

Appendix 8.2. Major element variation diagrams for (a) CaO vs MgO (b) Al2O3 vs

MgO and (c) FeOtotal vs MgO in wt% oxide for the adakitic rocks from western

Panama and southern Costa Rica.

xiv

Appendix 9. Trace element variation diagrams against SiO2 (wt.%) for the adakitic

rocks from western Panama and southern Costa Rica.

Appendix 10. Binary plots of (a) Sr/Y vs Y and (b) La/Yb vs Yb to test under-

plated basalt melting. The residual mineral assemblage and partition coefficients

are identical with Figure 1.19. Tick marks illustrate varying degree of partial

melting (per 10%).

1

Chapter 1. The Pleistocene adakitic rocks in the

western Panamanian arc: slab melt or evolved

mantle-derived magma?

1. Introduction

Adakite is an intermediate to felsic plutonic or volcanic rock characterized by

high Na2O (>3 wt.%), Sr/Y (>~10) and La/Yb (>~20), low Y (<~10 ppm) and Yb

(<~1 ppm) (Castillo, 2012; Defant and Drummond, 1990). It initially refers to

partial melts of an eclogitized subducting basaltic slab (Defant and Drummond,

1990). However, alternative processes have been proposed for the formation of

adakite-like magmas in arc environments: 1) lower crustal melts formed under the

pressure and temperature in which garnet amphibolite or amphibole-bearing

eclogite are stable (Atherton and Petford, 1993; Dai et al., 2017; Goss and Kay,

2009; Kay and Kay, 2002; Kay et al., 1994; Petford and Atherton, 1996; Wang et

al., 2004), and 2) garnet±amphibole fractionation of mantle-derived basaltic arc

magmas at high pressure (Castillo, 2008; Castillo et al., 1999; Macpherson et al.,

2006; Rodríguez et al., 2007; Rooney et al., 2011). The adakitic suites are often

closely associated with mafic rocks (Castillo, 2008; Castillo et al., 1999;

Danyushevsky et al., 2008; Kay, 1978; Kelemen et al., 2004; Macpherson et al.,

2006; Martin et al., 2005). They are considered to be slab melts equilibrated with

mantle during ascent (Kay, 1978; Kelemen et al., 2004; Martin et al., 2005), or

partial melts derived from the metasomatized mantle (Castillo et al., 1999;

Danyushevsky et al., 2008; Macpherson et al., 2006; Martin et al., 2005).

Quaternary volcanic rocks from western Panama and southern Costa Rica,

adjacent to the Costa Rican-Panamanian border at the southern extent of the

Central American Volcanic Arc (CAVA) (Fig. 1.1), display the adakite-like

geochemical signatures. Petrogenesis of these rocks has been a matter of debate.

Early studies (de Boer et al., 1991; Defant et al., 1992) argued that they are

slab melts formed by partial melting of a warm and young subducting oceanic crust

as indicated by their high La/Yb and Sr/Y and low Yb and Y. The adakitic

geochemical features of the lavas were explained by partial melting of a MORB

2

source (Defant et al., 1992; Defant et al., 1991b). Tomascak et al. (2000) argued

that the MORB-like δ7Li isotope and low B/Be ratios of the Quaternary adakitic

rocks from western Panama support this hypothesis.

Recent isotopic studies have suggested that the adakitic rocks are derived

from sub-arc mantle metasomatized by slab melts (Bindeman et al., 2005; Gazel et

al., 2009; Gazel et al., 2011; He et al., 2017; Wegner et al., 2011). Bindeman et al.

(2005) and Wegner et al. (2011) showed that δ8OOlivine values of western

Panamanian adakitic rocks higher than those of average mantle peridotites are

coupled with Sr/Y, La/Yb and 87Sr/86Sr of host volcanic rocks. They pointed out

that this correlation can be best explained by the enrichment of sub-arc mantle

source by slab melts. Pd-Nd-Sr isotopes of the western Panamanian and southern

Costa Rican adakitic rocks (Abratis and Worner, 2001; Feigenson et al., 2004;

Gazel et al., 2009; Gazel et al., 2011; Hoernle et al., 2008) exhibit that they have

moderate amounts of the Galapagos hotspot track component (e.g. Cocos/Coiba

ridges), suggesting that the adakitic rocks were formed by melting of the mantle

wedge reacted with melts derived from the subducting Galapagos hotspot tracks

(Gazel et al., 2009; Gazel et al., 2011).

Hidalgo and Rooney (2010) emphasized the role of garnet±amphibole

fractionation at lower crustal depth in the adakitic signatures of western

Panamanian volcanic rocks. They investigated the pressure and temperature

conditions of the magmas using compositions of amphiboles from adakitic rocks

and their cumulative enclaves from Baru volcano, western Panama. The results

show that the Baru magmas have undergone extensive magma differentiation in the

hot crystal mush zone at middle-lower crustal depths (Hidalgo and Rooney, 2010).

Detailed geochemical and chronological studies of the volcanic edifices in Baru

volcano (Hidalgo and Rooney, 2014) showed that the substantial volume (~450

km3) of magma erupted efficiently within ~0.213 My. The high magma production

rate in this region is similar to those of typical arc systems where hydrous melting

of sub-arc mantle occurs and is unlikely to be achieved by dehydration melting of

eclogitized basaltic rocks (Hidalgo and Rooney, 2014).

In this study, we investigated whole-rock major and trace elements, and Nd,

Sr and Pb isotope geochemistry of the Pleistocene adakitic volcanism in western

Panama where the Cocos ridge is subducting beneath Central America (Gazel et al.,

3

2011) to better understand their petrogenesis. We also present major and trace

element geochemistry of amphibole and pyroxene phenocrysts measured by EPMA

and LA-ICP-MS, respectively to investigate their crystallization depth and the melt

composition from which they crystallized. The previous petrogenetic models for

adakitic rocks from western Panama and southern Costa Rica were re-examined

using our new data together with those from the previous studies.

4

2. Geological Background and Samples

The Farallon plate had split into the Cocos and Nazca plate (Lonsdale, 2005;

Mann, 2007) during the early Miocene (~23 Ma), forming the Cocos-Nazca

spreading ridge. The Cocos-Nazca spreading ridge has interacted with the

Galapagos hotspot plume, which generated the Galapagos hotspot tracks; the

Cocos/Coiba ridge and the Seamount Province (Meschede and Barckhausen, 2001).

The onset of the Cocos Ridge subduction underneath southern Central America is

controversial, although most studies suggested that it was earlier than ~8-10 Ma

(Abratis, 1998; Abratis and Worner, 2001) or at least ~3 Ma (MacMillan et al.,

2004; Morell, 2015).

The southern CAVA has been formed by the subduction of the Cocos and

Nazca plates beneath the Caribbean plate (Fig. 1.1) at a rate ranging from ~8 cm/yr

of the Cocos plate (DeMets et al., 2010; Kobayashi et al., 2014) to ~4 cm/yr of the

Nazca plate (DeMets et al., 2010). The subducting crust beneath the southern

CAVA was produced at the Cocos-Nazca spreading center and overprinted by the

Galapagos hotspot traces (Feigenson et al., 2004; Gazel et al., 2011; Hoernle et al.,

2008; Lonsdale and Klitgord, 1978). The Panama fracture zone, an active

transform fault between the Cocos and Nazca plate along 83°W (Fig. 1.1),

intersects with the southern Central American trench (Molnar and Sykes, 1969).

Late Cenozoic magmatism of the western Panamanian arc is characterized by

tholeiitic to calc-alkaline volcanism during the Middle Oligocene to Middle

Miocene (de Boer et al., 1988; de Boer et al., 1991), which resulted from

orthogonal subduction of the Cocos Plate beneath Costa Rica and Panama (Morell,

2015). During the Late Miocene, the arc magmatism changed toward high-K calc-

alkaline compositions (Gazel et al., 2009), suggesting slightly enriched mantle

compositions in the middle Oligocene (Gazel et al., 2009; Wegner et al., 2011). In

the late Miocene, the arc magmatism was displayed by high-K to shoshonitic

compositions, and their magma sources are considered to be comparable to the

present mantle beneath the southern CAVA (Gazel et al., 2009; Wegner et al., 2011).

Quaternary adakitic volcanism along the southern CAVA has occurred

dominantly in southern Costa Rica and western Panama (de Boer et al., 1991;

Defant et al., 1992; Defant et al., 1991a; Hidalgo and Rooney, 2010, 2014; Hidalgo

5

et al., 2011; MacMillan et al., 2004; Morell, 2015; Wegner et al., 2011).

Petrogenesis of the Quaternary adakitic rocks in the easternmost extent of the

CAVA, adjacent to central Panama (Panama Canal) differs from that in southern

Costa Rica and western Panama, close to the Costa Rican-Panamanian border

(Hidalgo et al., 2011; Rooney et al., 2015).

The adakitic magmatism of the easternmost segment of the CAVA has erupted

around ~1.5 Ma (MacMillan et al., 2004; Rooney et al., 2011; Tomascak et al.,

2000; Wegner et al., 2011). Rooney et al. (2015) argued that the Quaternary

adakitic rocks from El Valle volcano, located near the Panama Canal region, are

formed by fluid-fluxed melting of the mantle wedge underneath the Panamanian

arc where the Nazca Plate is subducting obliquely, and fractionation of amphibole

+garnet at lower crustal depth. The abundant faults and fracture zones on the plate

subducting under central Panama may have enhanced fluid flux into the mantle

wedge in this region (Rooney et al., 2011; Rooney et al., 2015).

The Quaternary adakitic volcanism in western Panama and southern Costa

Rica, close to the Costa Rican-Panamanian border, is characterized by the

Galapagos hotspot components (Feigenson et al., 2004; Gazel et al., 2011; Hoernle

et al., 2008). Previous studies (Carr et al., 2003; Feigenson et al., 2004; Gazel et al.,

2009; Gazel et al., 2011; Hoernle et al., 2008; Leeman et al., 1994) suggested that

adakitic rocks from western Panama and southern Costa Rica originated from the

mantle influenced by enriched slab components. The higher La/Yb and more

enriched LREE of the adakitic rocks from western Panama and southern Costa

Rica, compared to those of northern Central America, have been attributed to the

Galapagos hotspot contribution (Carr et al., 2003; Gazel et al., 2009; Gazel et al.,

2011). Metasomatized mantle source, enriched by the interaction with the

Galapagos hotspot component, has also been documented by Pb-Nd-Sr isotopic

studies in the ranges of Galapagos oceanic island basalts and their tracks such as

the Seamount province and the Cocos/Coiba ridge (Feigenson et al., 2004; Gazel et

al., 2009; Gazel et al., 2011; Hoernle et al., 2008).

In this study, we selected seven representative adakitic volcanic rocks from

the study of Gazel et al. (2011), which cover a wide range of rock types including

basalt, basaltic andesite, trachyandesite and dacite. They occur as lava flows or

domes in western Panama along the southern CAVA where the Cocos ridge

6

subducts. The eruption ages of the samples are between 0.01 Ma and 2.6 Ma (Gazel

et al., 2011).

Figure 1.1. Tectonic setting of Central America modified from Gazel et al. (2009).

Grey shaded area illustrates approximate area for the adakitic rocks from western

Panama and southern Costa Rica (Gazel et al., 2011). Dashed line in the

continental areas indicates the location of the Central American Volcanic Arc

(CAVA). Miocene (>10 Ma) and Pliocene (~3.6 Ma) mantle-derived arc rocks in

southern Costa Rica are from Gazel et al. (2009). The location of DSDP Site 504,

Costa Rica Rift (Hubberten et al., 1983), and Cocos-Nazca Spreading Center

(CNS) crust (Harpp et al., 2004) is presented as star symbols. PFZ: Panama

Fracture Zone, EPR: East Pacific Rise.

7

3. Analytical Methods

3.1. Mineral chemistry analysis

Major element compositions of minerals were determined by a Field-Emission

Electron Probe Microanalyzer (FE-EPMA; JEOL JXA-8530F) with voltage and

beam current of 15 kV and 10 nA, respectively at Korea Polar Research Institute

(KOPRI), South Korea. An Excel spreadsheet PROBE-AMPH (Tindle and Webb,

1994) was used to recalculate the raw data and determine the structural formulae of

amphibole. Multiple analyses of an amphibole reference material (the Smithsonian

Microbeam Standard NMNH 111356) and other silicate minerals were conducted

during the EPMA session to check the data quality. Our results on the standards are

consistent with the reference values (Appendix 1).

Trace element contents in amphibole were measured by LA-ICP-MS using an

ArF excimer laser system (193 nm wavelength, ESI NWR 193 model) combined

with an ELAN 6100 quadrupole ICP-MS at KOPRI. The analyses were conducted

at 3.6 J/cm2 beam density, 5 Hz beam frequency using 50 and 75 µm beam sizes. A

NIST 610 glass was used as a primary standard and Ca contents measured by

EPMA were used to correct the yield differences between the standard and

unknown samples. Reference materials of NIST 612 and BCR-2g were analyzed

together with the unknown samples to monitor the data quality (Appendix 2). The

relative differences between our results and the reference values (Jochum et al.,

2009; Jochum et al., 2011) were <4% and <10% for NIST 612 and BCR-2g,

respectively.

3.2. Whole-rock major and trace element analysis

Whole-rock major element compositions of all the studied samples, except for

M63aKH-b, are reported in Gazel et al. (2011). The major element contents of

M63aKH-b and trace element contents of the seven samples were additionally

determined in this study. The samples were crushed into rock chips, washed with

deionized pure water and powdered in an agate mill (PM 400; Retsch) at KOPRI.

For the major element analysis of M63aKH-b, 0.5 g of the sample powder was

weighted into a platinum crucible with 1.0 g of lithium tetraborate and 0.5 g of

8

lithium metaborate. Before fusion, they were mixed thoroughly. The sample and a

reference material (AGV-2, USGS) were fused at 1000°C for 30 minutes, using

Katanax K2 Prime Electric Fusion Machine at KOPRI. The fused discs were

analyzed for major elements by S8 TIGER (Bruker, Germany) XRF installed at the

National Center for Inter-university Research Facilities (NCIRF) at Seoul National

University. The fused discs were recycled for trace element analysis by LA-ICP-

MS at Korea Basic Science Institute (KBSI), South Korea. The laser system is

composed of a NWR193UC 193 nm ArF excimer laser system and a Thermo ICAP-

Q quadrupole ICP-MS. Our results on BCR-2g and AGV-2 are consistent with the

reference value (Jochum et al., 2009) within <10% for most elements (Appendix 3).

For trace element analyses of the other six samples, about 30 mg of each

powdered sample was dissolved in an HF-HNO3-HClO4 mixture and heated at

120°C for one day. The dried samples were dissolved in 2% HNO3 and diluted

4000 times. Indium and Re were used as internal standards to correct instrumental

and signal drift. The final solutions were analyzed by an ELAN 6100 quadrupole

ICP-MS (Perkin-Elmer) at KOPRI. Reference materials (BHVO-2 and AGV-2)

were analyzed together with the samples to check accuracy and precision of the

measurement, and the results are reported in Appendix 4. Our results are consistent

with the values from Jochum et al. (2016) within 10% for most elements, except

for Hf and Ta (< 20%) for AGV-2 and Nb, Hf, Ta, and Pb (< 20%) for BHVO-2.

3.3. Whole-rock Sr-Nd-Pb radiogenic isotope analysis

Sr, Nd, and Pb isotope ratios were measured by a method described in Lee et

al. (2011) and references there in. Briefly, rock chips were washed with Milli-Q

water, crushed to small pieces by a rock crusher and 1-2 mm-sized pieces were

hand-picked. The pieces were washed with 2N HCl in an ultrasonic bath for

approximately 10 minutes and repeatedly rinsed with Milli-Q water. About 50 mg

of washed rock chips were dissolved in concentrated HF and HClO4 mixture at

120°C for three days. Pb was separated using glass columns filled with AG1X8

(100-200mesh) resin (0.5 ml), equilibrated with 0.5N HBr and was collected in 6N

HCl. The collected samples were dried at 100°C for 24 hours. To separate Sr from

the REEs, the column was loaded with 6 ml of 2.5N HCl, using Biorad AG 50W-

X8 (100–200 mesh). The REEs were obtained by loading 7 ml of 6N HCl after Sr

9

collection. Nd was separated from the solution containing the REEs using 0.7 ml of

AG 50W-XB (200-400 mesh) and loaded in 5 ml of 0.22 M HIBA

(2Hydroxyisobatyric acid and Ammonia water). All procedures were performed in

a Class 500 clean room at KOPRI. The Sr-Nd-Pb isotope ratios of the final

solutions were determined by a thermal ionization mass spectrometer (TRITON,

Thermo Fisher Scientific), equipped with nine-adjustable Faraday cups, at KOPRI.

10

4. Results

4.1. Petrography and mineral chemistry

The western Panamanian Pleistocene rocks investigated in this study consist

of basalt (M64fKH), basaltic andesite (M64bKH, M66aKH, M91aKH),

trachyandesite (M38KH) and dacite (M63aKH, M63aKH-b).

Basaltic andesite (M91aKH) is composed of ~13 vol.% phenocrysts in a fine-

grained groundmass, consisting mainly of olivine, clinopyroxene, orthopyroxene

and plagioclase. Olivine phenocrysts are subhedral to euhedral and their

compositions range from Fo83 to Fo85. Clinopyroxene (~4 vol.%) and

orthopyroxene (~2 vol.%) phenocrysts occur as euhedral to subhedral crystals.

Some clinopyroxene grains form glomeroporphyritic aggregates with

orthopyroxene (Fig. 1.2a and 1.2b). Clinopyroxenes range in composition from

diopside to augite (Fig. 1.3a). They show weak normal zoning with increases in Al

and Fe contents from core to rim. Clinopyroxene cores have Mg# [molar

Mg/(Mg+Fe)] of 0.79-0.86 (Fig. 1.3b, Table 1.1). Compositional zonation is little

in orthopyroxene and their cores exhibit a narrow range of Mg# from 0.88 to 0.90

(Fig. 1.3c, Table 1.1). Euhedral to subhedral plagioclase microphenocrysts range

from 100 μm to 300 μm in size (Fig. 1.2b). The groundmass is mainly composed of

lath-shaped plagioclase, clinopyroxene, Fe oxides and glass.

Trachyandesite (M38KH) is porphyritic with plagioclase, clinopyroxene and

quartz phenocrysts (~35 vol.%) in a fine-grained matrix. Plagioclase phenocrysts

are subhedral to anhedral with oscillatory zoning and Carlsbad twin. Some

plagioclase crystals display a rounded shape with sieve texture (Fig. 1.4a), which is

indicative of disequilibrium (Kuşcu and Floyd, 2001). They show a wide range of

compositional variations in An contents from 30 to 59 %. Some phenocrysts show

reversed zoning with increases of An contents toward the rims. Clinopyroxene

phenocrysts occur as weakly zoned euhedral to subhedral crystals. Some form

glomeroporphyritic aggregates (Fig. 1.4b). They have cores with Mg# ranging from

0.69 to 0.82 (Fig. 1.3b, Table 1.1). Quartz phenocrysts are rounded, resorbed and

surrounded by reaction rims of clinopyroxene (Fig. 1.4c). This together with sieved

and resorbed plagioclase phenocrysts infer magma mingling processes. The matrix

11

is intersertal and mainly composed of plagioclase and clinopyroxene dominantly

with minor Fe oxides and glass.

Hornblende-biotite dacites (M63aKH and M63aKH-b) consist of 50-60 vol.%

phenocrysts (Fig. 1.5), which include plagioclase, amphibole, biotite, quartz, K-

feldspar and Fe oxide. Sample M63aKH-b contains clinopyroxene (~3 vol.%) and

orthopyroxene (~2 vol.%) phenocrysts. Plagioclase phenocrysts are dominant (30-

40 vol.%) and occur as euhedral to anhedral crystals. Some show resorbed and

rounded grain boundary with sieve-textured core, indicative of disequilibrium with

host melt. The An contents (%) vary widely from 33 to 71. Some phenocrysts show

revered zoning with An contents, increasing toward the rims. Amphibole

phenocrysts (5-10 vol. %) are euhedral to subhedral and up to 1mm in length (Fig.

1.6). They display pale yellow to dark reddish brown pleochroism. Most of

amphibole phenocrysts exhibit patchy or oscillatory-zoning, characterized by dark

and bright zones in back-scattered electron images (Fig. 1.6b and 1.6d), suggesting

sub-solidus re-equilibration with host melt. The amphiboles from M63aKH-b have

Fe-Ti oxide rims (10-200 μm width), which is indicative of destabilization during

magma ascent or pre-eruptive re-heating (Farver and Brabander, 2001; Rutherford

and Devine, 2003). Few of them occur as inclusions in plagioclase phenocrysts in

both samples. Biotite phenocrysts (5-10 vol.%) occur as euhedral brown flakes

with distinct cleavage. Quartz phenocrysts are rounded and embayed (<10 vol.%,

Fig. 1.5f). Clinopyroxene (~3 vol. %) and orthopyroxene (~2 vol. %) phenocrysts

from M63aKH-b occur as euhedral to subhedral crystals and have augitic, diopsidic

and enstatitic compositions (Fig. 1.3a). Core compositions of clinopyroxene and

orthopyroxene have a range in Mg# of 0.76-0.87 and 0.84-0.90 (Fig. 1.3b and 1.3c,

Table 1.1), respectively. Mg# in the orthopyroxene cores are mostly higher than in

the rims (Fig. 1.3c). A fine-grained cumulative enclave occurs in sample M63aKH.

It is mainly composed of plagioclase, amphibole and biotite. The compositional

variations of amphibole phenocrysts in the cumulates are within the compositional

range of amphibole phenocrysts found in the host rock. The fine-grained (or

microcrystalline) groundmass of M63aKH and M63aKH-b is mainly composed of

plagioclase and glass, and small amounts of interstitial quartz are observed in the

groundmass.

12

Figure 1.2. Microphotographs of basaltic andesite (sample M91aKH). (a) glomeroporphyritic aggregates of clinopyroxene grains. (b)

glomeroporphyritic aggregates of clinopyroxene with orthopyroxene. (c) subhedral to euhedral olivines phenocrysts and lath-shaped

plagioclases in the groundmass.

13

Figure 1.3. Pyroxene compositions in basaltic andesite (M91aKH), trachyandesite

(M38KH) and dacite (M63aKH-b). (a) Ca-Mg-Fe Pyroxene classification diagram

(Morimoto, 1988). (b and c) binary plots of Mg# vs. Al (apfu) for clinopyroxene

and orthopyroxene.

14

Table 1.1. Mg number (#) and SiO2 (wt.%) of the rocks, pyroxenes, and

amphiboles and their equilibrium melts.

The details about calculated melt compositions and subdivision of the amphiboles into two groups are

presented in chapter 5.1.

15

Figure 1.4. Microphotographs of trachyandesite (sample M38KH). (a) some plagioclase crystals display mingling texture. (b)

glomeroporphyritic aggregates of clinopyroxene crystals. (c) quartz surrounded by reaction rims of clinopyroxene.

16

Figure 1.5. Microphotographs of dacites (sample M63aKH and M63aKH-b). (a) plagioclase and biotite in M63aKH. (b) plagioclase and biotite

in M63aKH-b. (c) plagioclase co-crystallized with biotite. (d) clinopyroxene in M63aKH-b. (e) orthopyroxene in M63aKH-b. (f) embayed and

rounded quartz in M63aKH.

17

Figure 1.5. Microphotographs of dacites (sample M63aKH and M63aKH-b). (a) plagioclase and biotite in M63aKH. (b) plagioclase and biotite

in M63aKH-b. (c) plagioclase co-crystallized with biotite. (d) clinopyroxene in M63aKH-b. (e) orthopyroxene in M63aKH-b. (f) embayed and

rounded quartz in M63aKH.

18

Figure 1.6. Microphotographs of dacites (sample M63aKH and M63aKH-b). (a, b and c) amphiboles in M63aKH. (d) amphiboles in M63aKH-b.

Red dotted circles are locations of laser ablation ICP-MS analyses. The values next to red circles are the beam sizes of LA-ICP-MS analyses.

Filled circles are locations of Field-Emission EPMA. The compositions of mauve and purple circles indicate low and high pressure amphiboles,

respectively.

19

4.2. Amphibole compositions

Amphibole phenocrysts from the two dacites are calcic (Ca ≥1.69 apfu; Ti

≤0.29 apfu) and magnesian (0.61≤ Mg ≤0.75 apfu) amphiboles (Fig. 1.7, Table 1.2).

They show large variations in Mg# from 0.54 to 0.77 (Table 1.1). Patchy and

oscillatory-zoned amphiboles are displayed by dark and bright zones in back-

scattered electron images (Fig. 1.6). Bright zones have higher Aliv, Na+K, Fe2+ and

lower Mg# than dark zones.

We subdivided the amphiboles into two groups according to their

compositions: low pressure (220 MPa; low-P) and high pressure (≥920 MPa; high-

P). Details about the subdivision will be discussed in section 5.1.1.

Figure 1.8 shows the co-variation of Altotal with Na+K and Mg# of each group

of amphibole phenocrysts. Their total Al composition increases with increasing

Na+K and has a negative correlation with Mg#. The chondrite-normalized REE

patterns (Sun and McDonough, 1989) of both groups of amphiboles are similar to

each other, showing enrichment in the middle REEs (MREE) relative to LREE and

depletion in the HREEs (Fig. 1.9). It should be noted that the low-P amphiboles

contain higher REE contents than high-P amphiboles with weak Eu negative

anomalies.

Figure 1.7. Mg# vs Si classification diagrams for calcic amphiboles (a) Na + K

<0.5 and (b) Na + K ≥0.5. Mauve and purple circles indicate amphiboles formed at

low (220 MPa) and high (≥920 MPa) pressures, respectively, discussed in section

5.1.1.

20

Figure 1.8. Binary plots of (a) total Al vs stoichiochemical Na+K and (b) total Al

vs Mg# of amphiboles in the western Panamanian adakitic rocks. The fields of

experimental amphibole compositions are obtained at 220 and 400 MPa (Prouteau

et al., 1999; Scaillet and Evans, 1999), 920 and 1200 MPa (Alonso-Perez et al.,

2009) and 960 MPa (Prouteau et al., 1999) and modified from Ribeiro et al. (2016).

Figure 1.9. Chondrite normalized REE patterns of (a) low-P (220 MPa) and (b)

high-P (≥920 MPa) amphibole phenocrysts.

21

Table 1.2. Major element compositions of representative amphiboles in dacites

(M63aKH and M63aKH-b) from western Panama, Central America.

Pre-eruption conditions are estimated using the compositions of experimental amphiboles and the

calibrations of Ridolfi and Renzulli (2012)b and Ridolfi et al. (2010)c. The details about the pre-

eruptive conditions and calculated parental melts are described in section 5.1. Mg# = Mg/(Mg+FeT),

which is a minimum estimate, following the approach of Ribeiro et al. (2016).

22

4.3. Whole-rock geochemistry

The Pleistocene volcanic rocks from western Panama in this study are plotted

in basalt, basaltic andesite, trachyandesite and dacite fields (SiO2 > 51.0 wt.%, K2O

< 3.0 wt.%, Al2O3 > 15.0 wt.%, Na2O = 3.1-3.8 wt.%) in a total alkali-silica

diagram (Fig. 1.10a, Table 1.3). They are characterized by adakite-like

geochemical features of Sr >926 ppb, high Sr/Y >50, La/Yb ≥20, low Y <15 ppm

and low 87Sr/86Sr <0.704 (Fig. 1.11 and 1.13, Table 1.3 and 1.4). They also exhibit

significant enrichment of highly incompatible elements and relative depletion of

HREEs and HFSE compared to the average value of global MORBs from Gale et

al. (2013) (Fig. 1.10b).

Figure 1.13 and Table 1.4 show Pb-Sr-Nd isotopic composition of adakitic

rocks from western Panama and southern Costa Rica. Their Pb isotope ratios are

similar to those of the Cocos ridge basalts (13.0-14.5 Ma; Hoernle et al., 2000) and

mantle-derived arc basalts from southern Costa Rica (<5.3 Ma; Gazel et al., 2009),

which are spatially associated with the western Panamanian and southern Costa

Rican adakitic lavas. Note that the Pb isotope ratios are significantly higher than

the altered oceanic crust of the Pacific plate (ODP hole 504; Shimizu, 1989), which

is suggested to be the subducting crust before the collision of the Galapagos

hotspot traces (Patino et al., 2000). The adakitic lavas from western Panama and

southern Costa Rica have lower 143Nd/144Nd ratios and higher 87Sr/86Sr ratios than

those of the oceanic crust from ODP hole 504 (Shimizu, 1989). They display

similar Nd isotope ratios and relatively higher 87Sr/86Sr ratios compared to the

Cocos ridge basalts (Hoernle et al., 2000). Their Nd and Sr isotope ratios are

consistent with those of the Pliocene Costa Rica arc basalts (Gazel et al., 2009).

The major and trace element and isotopic data of the selected adakitic rocks in

this study are comparable to those reported in previous studies and cover a wide

range of rock types, suggesting they are representative of Quaternary adakite-like

magmas in western Panama and southern Costa Rica.

23

Figure 1.10. (a) Geochemical classification of samples in this study and adakitic

rocks from western Panama and southern Cost Rica from Gazel et al. (2011) and

Hidalgo and Rooney (2014). (b) Primitive mantle (McDonough and Sun, 1995)

normalized multi-element diagrams of the adakitic rocks in this study and adakitic

rocks from western Panama and southern Costa Rica (shaded area). Average

Pliocene mantle-derived arc basalts from southern Costa Rica, spatially associated

with the adakitic rocks, are from Gazel et al. (2009). The composition of average

MORB is from Gale et al. (2013).

Figure 1.11. Binary plots of (a) Sr/Y vs Y and (b) La/Yb vs Yb to discriminate

adakites from calc-alkaline arc lavas.

24

Figure 1.12. Binary plots of (a) La/Yb vs SiO2, (b) Yb vs SiO2, (c) Sm/Yb vs SiO2,

(d) Dy/Yb vs SiO2 and (e) Nb/Yb vs SiO2 of western Panamanian and southern

Costa Rican adakitic rocks. Solid and dotted arrows represent variations in melt

composition during garnet and amphibole fractionation, respectively.

25

Figure 1.13. (a and b) Pb-Nd-Sr isotope systematics for the adakitic rocks in this study and possible sources. The Cocos ridge rocks as possible

sources of the adakitic rocks from western Panama and southern Costa Rica are from Hoernle et al. (2000). Red solid lines are mixing curves

between the mean Cocos ridge basalts and the subducting sediments. The sediments are assumed to be a mixture of 70% hemipelagic sediments

and 30% carbonate sediments from DSDP Site 495 (Patino et al., 2000). Pliocene mantle-derived arc basalts (~3.6 Ma) in southern Costa Rica

are from Gazel et al. (2009). Altered oceanic crust (ODP504 MORB) in the eastern Pacific is from Shimizu (1989). Galapagos oceanic island

basalts are from White et al. (1993).

26

Table 1.3. Major element (wt.%) and trace element (ppm) compositions of western

Panamanian adakitic volcanic rocksa.

27

Table 1.4. Pb-Nd-Sr isotopic compositions of western Panamanian adakitic

volcanic rocks.

28

5. Discussion

5.1. Pre-eruptive conditions and host melts of western Panamanian

adakitic rocks

5.1.1. Pre-eruptive conditions

The Al-in-hornblende geobarometry has been widely used to estimate the

crystallization pressures of hydrous magmas (Anderson and Smith, 1995;

Hammarstrom and Zen, 1986; Hollister et al., 1987; Johnson and Rutherford, 1989;

Schmidt, 1992). Erdmann et al. (2014) argued that the Al-in-hornblende

geobarometry should be used with caution because Al contents in amphiboles are

not only affected by pressure, but also by other factors including temperature

(Erdmann et al., 2014; Putirka, 2016), H2O fugacity (Scaillet and Evans, 1999) and

melt composition (Ridolfi and Renzulli, 2012).

Amphibole phenocrysts occur in adakitic dacites from western Panama. They

show negative correlations between Si+Mg and total Al (Fig. 1.14a), suggesting

that pressure-dependent Al-Tschermakitic exchange (Si + R2+ = [4]Al + [6]Al)

(Anderson and Smith, 1995; Spear, 1981) dominantly controlled the amphibole

compositions despite poor correlations between tetrahedral and octahedral Al (Fig.

1.14b). This is consistent with no correlation of Ti with both tetrahedral and

octahedral Al, which implies that Al contents are not affected by temperature

dependent Ti-Tschermak substitutions (Ti + R2+ = 2[6]Al or Ti + [4]Al = [6]Al + Si)

(Anderson and Smith, 1995; Bachmann and Dungan, 2002; Spear, 1981). However,

decreasing Si trends with increasing tetrahedral Al+K (Fig. 1.15a) may point to

temperature-dependent edenitic substitution (Si + [A]□ = [4]Al + [A](K+Na))

(Anderson and Smith, 1995; Spear, 1981), suggesting that the amphibole

composition is also affected by temperature to some extent. These cation variations

in the amphiboles suggest that the western Panamanian adakitic magmas have

experienced crystallization at multiple depths with large variations in temperature

and probably H2O contents.

The Al-in-hornblende geobarometry has been experimentally calibrated by

silica-oversaturated silicate melts (Schmidt, 1992) in equilibration with amphiboles

29

having Fe# [FeT/(FeT + Mg)] ratios of 0.40-0.65 (Anderson and Smith, 1995).

Therefore, this method is only applicable to melts under such limited conditions.

Although the dacites from western Panama have interstitial quartz in the

groundmass, the amphibole phenocrysts often display disequilibrium textures such

as Fe-Ti oxide reaction rims and patch zonings (Fig. 1.6), implying that they may

not be in equilibrium with the host melt. Moreover, most of the amphiboles have

Fe# of <0.40, which is outside the calibration range (Table 1.2).

Figure 1.14. Binary plots of (a) total Al vs stoichiochemical Si+Mg and (b)

tetrahedral Al vs octahedral Al of amphiboles in the western Panamanian adakitic

rocks (M63aKH and M63aKH-b).

Figure 1.15. Binary plots of (a) Si vs tetrahedral Al plus K and Na and (b)

stoichiochemical Ti vs octahedral Al of amphiboles in the western Panamanian

adakitic rocks (M63aKH and M63aKH-b).

30

Therefore, we used experimental amphibole compositions (Alonso-Perez et al.,

2009; Prouteau et al., 1999; Scaillet and Evans, 1999) to estimate approximate

magma storage depth of the western Panamanian adakitic rocks in this study as a

similar approach used in Samaniego et al. (2010) and Ribeiro et al. (2016) for

adakites elsewhere.

The compositions of the western Panamanian adakitic rocks are comparable

with the melt compositions of the crystallization experiments, including a synthetic

dacite (Alonso-Perez et al., 2009), adakitic dacite and rhyolite (Prouteau et al.,

1999) and synthetic dacite glass (Scaillet and Evans, 1999). Additionally, the

experimental conditions of Alonso-Perez et al. (2009), Prouteau et al. (1999) and

Scaillet and Evans (1999) are within the range of the P-T conditions (T = 750-

1000 °C, P = 220-1200 MPa) suggested for the adakitic magmatism in the western

Panama (T = 780-950 °C, P = 250-850 MPa) by Hidalgo and Rooney (2010).

In Figure 1.8, the amphiboles in the dacites are mostly plotted in the

compositional fields of the low pressure (220 MPa; low-P) and high pressure (≥920

MPa; high-P) experimental amphiboles (Alonso-Perez et al., 2009; Prouteau et al.,

1999; Scaillet and Evans, 1999). These results suggest that the adakitic magmas

have stalled and underwent at least two stages of amphibole crystallization at

depths of ~8 km and ~35 km. It is worth noting that some patchy-zoned

amphiboles have compositions corresponding to both high-P and low-P amphiboles,

indicating that they initially crystallized at high pressure and re-equilibrated with

the host melt at low pressure.

To estimate crystallization temperature and oxidation state of amphiboles in

the dacites, we have used the calibrations of Ridolfi and Renzulli (2012) and

Ridolfi et al. (2010), following the evaluation of Erdmann et al. (2014). The results

are presented in Table 1.2. The crystallization temperatures of amphiboles in the

western Panamanian adakitic rocks range between ~721 and 1009 °C using the

calibrations of Ridolfi and Renzulli (2012) and between ~780 and 979 °C using

those of Ridolfi et al. (2010). The calculated fO2 conditions range from NNO - 0.1

to + 2.2, using the calibrations of Ridolfi et al. (2010). These results correspond

well with those found in Hidalgo and Rooney (2010). They suggested that the

compositional variations of amphiboles in Quaternary adakitic magmas from Baru

volcano, adjacent to our study area, indicate fractionation at the upper and lower

31

crustal depth (250-800 MPa and 800-900 °C).

We also applied clinopyroxene-melt and two pyroxene geothermobarometry

(Putirka, 2008) to constrain their crystallization pressure and temperature using

compositions of clinopyroxene and orthopyroxene phenocrysts and their host rocks.

Only the equilibrated pairs of clinopyroxene-melt and clinopyroxene-

orthopyroxene whose Fe-Mg exchange coefficients are KD(Fe-Mg)cpx-liq = 0.25-

0.30 and KD(Fe-Mg)cpx-opx = ~1.16, respectively, were selected. The pressure and

temperature were calculated using Equation 30 of Putirka (2008) and the model of

Putirka et al. (2003) for clinopyroxene-melt geothermobarometry and Equation 39

and 36 of Putirka (2008) for two-pyroxene geothermobarometry. The pressure and

temperature of pyroxene crystallization are ~810 MPa and ~1170 °C for basaltic

andesite (M91aKH), ~600 MPa and ~1020 °C for dacite (M63aKH-b) and ~1300

MPa and ~1120 °C for trachyandesite (M38KH) (Fig. 1.16, Table 1.5 and 1.6).

Figure 1.16. Calculated pressure and temperature estimated using clinopyroxene-

liquid and clinopyroxene-orthopyroxene pairs which satisfy Fe-Mg equilibrium for

crystals in basaltic andesite (M91aKH), trachyandesite (M38KH) and dacite

(M63aKH-b). Temperature and pressure precisions are about ±~29 °C and ±~300

MPa, respectively.

32

Table 1.5. Representative clinopyroxene analyses and thermobarometry for clinopyroxene-liquid (whole-rock) pairs.

Clinopyroxenes satisfy Fe-Mg equilibrium for clinopyroxene-liquid thermobarometry.

33

Table 1.6. Thermobarometry for clinopyroxene-orthopyroxene pairs and their compositions.

Fe-Mg equilibrium for clinopyroxene-orthopyroxene thermobarometry is satisfied.

34

5.1.2. Equilibrium melt composition

Amphibole Mg# is largely controlled by melt composition (Grove et al., 2003),

although their Mg# can also be affected by temperature (Alonso-Perez et al., 2009)

and fH2O or fO2 (Grove et al., 2003). The experimental results of Grove et al.

(2003) show that amphiboles in equilibrium with primitive high Mg andesites and

basaltic andesites mostly have Mg# of ~0.83. On the contrary, experimental

amphiboles derived as a fractionated phase of an evolved andesitic melt have low

Mg# of ~0.65 (Alonso-Perez et al., 2009).

The amphibole phenocrysts from the dacites display Mg# of 0.54-0.77 (Table

1.1 and 1.2), reflecting that their equilibrated melts have experienced

compositional changes during magma differentiation. The Mg# of the equilibrated

melts is estimated using KD(Fe-Mg)hbl-liq (0.38±0.06) (Alonso-Perez et al., 2009;

Sisson and Grove, 1993) and Fe/Mg of amphiboles. The calculated melt

compositions in equilibrium with high-P and low-P amphiboles have a wide range

of Mg# of 0.36-0.56 and 0.31-0.50, respectively. It is worth noting that low Mg# (≤

0.41) of the calculated melts is in close agreement with those found in the study of

Alonso-Perez et al. (2009), suggesting that experimental andesite to dacite liquids

in equilibrium with amphiboles, fractionated from andesite magmas, have Mg # of

0.30-0.41. Also, a range of Mg # of the calculated melts is consistent with those of

basaltic andesite to dacite in this study (Mg# = 0.50-0.65). This suggests that the

western Panamanian adakitic magmas underwent extensive differentiation from

basalt to dacite.

The SiO2 contents (wt.%) of melts in equilibrium with amphiboles in the

dacite (M63aKH, M63aKH-b) are calculated using the model of Zhang et al.

(2017). The SiO2 (wt.%) of melts in equilibrium with high-P and low-P amphiboles

are 58.4-71.3 and 66.8-80.8, respectively (Table 1.2). These compositional

variations in the calculated melt compositions reflect significant degree of magma

differentiation within a crustal magma storage region underneath the western

Panamanian arc.

The Mg# of melts in equilibrium with clinopyroxene phenocrysts in the

basaltic andesite (M91aKH) and dacite (M63aKH-b) are calculated assuming a

KD(Fe-Mg)cpx-liq of 0.23 for clinopyroxene and melt (Sisson and Grove, 1993) and

XFe2O3/XFeO of 0.17 at given conditions (T = 1100 °C, P = 900 MPa and NNO = +

35

1), using the equation of Kress and Carmichael (1991). The results show that the

clinopyroxenes in the basaltic andesite and dacite are equilibrated with melts that

have Mg # of 0.49-0.68, consistent with Mg# (0.51-0.66) of the basalt and basaltic

andesites in this study (Fig. 1.17). These suggest that the clinopyroxene

phenocrysts were derived from basalt or basaltic andesite magma in the lower

crustal magma reservoirs (~35-40 km).

We also calculated REE contents of the melt equilibrated with high-P

amphiboles which have high Mg# of 0.68-0.75, using high-pressure (700 MPa)

experimental partition coefficients between hornblende and basaltic andesite melt

(Nandedkar et al., 2016). The REEs of the calculated melts are normalized to

chondritic abundances (Sun and McDonough, 1989) in Figure 1.18a. The

calculated melt compositions display adakite-like geochemical features with

La/Yb >40 and strong depletion in HREEs (Fig. 1.18). The REE patterns are

similar to those of the host dacite (M63aKH). These results suggest prolonged

magma evolution from basaltic andesite to dacite mainly driven by fractionation

of high-P amphiboles and garnet at lower crustal depth (~35 km), which may play

an important role in the adakitic features of the western Panamanian Quaternary

volcanism.

Despite lack of garnet phenocrysts in the samples, lines of geochemical

evidence support for garnet fractionation. First, the adakitic rocks from western

Panama and southern Costa Rica mostly have higher Sm/Yb ratios (≥4; Fig.

1.12c) than arc volcanic rocks that underwent shallow magma evolution (Goss

and Kay, 2006, 2009; Haschke et al., 2006; Kay et al., 2010), suggesting fractional

crystallization of HREE-rich garnet. Also, we confirm co-crystallization of garnet

and amphibole in the adakitic rocks of western Panama and southern Costa Rica

based on trace element variations during magma evolution. HREEs are strongly

compatible to garnet hence garnet-dominant fractionation will increase both

LREE/HREE and MREE/HREE ratios and decrease HREE with increasing SiO2.

The Panamanian adakitic rocks show increases in La/Yb and Nb/Yb and a

decrease in Yb as SiO2 increases while Dy/Yb and Sm/Yb remain in constant (Fig.

1.12). This trends can be accounted for by fractionation of garnet and amphibole

(Davidson et al., 2007; Profeta et al., 2015; Richards, 2011) because MREE are

more compatible to amphibole than LREE and HREE (Blundy and Wood, 2003).

36

Figure 1.17. Mg# of whole-rocks and calculated melts in equilibrium with

clinopyroxene cores and amphiboles.

Figure 1.18. (a) Chondrite-normalized REE patterns of the melts equilibrated with

the high-P amphiboles in the dacite (M63aKH) and the host rock (M63aKH). The

chondrite values are from Sun and McDonough (1989). (b) Binary plots of Sr/Y vs

Y for the equilibrium melts. The parental melts were calculated using high-P

amphibole compositions and high pressure (700 MPa) experimental partition

coefficients between hornblende and basaltic andesite (Nandedkar et al., 2016).

37

5.2. Formation of Quaternary adakitic rocks in the western

Panamanian arc

Adakitic magmas with high to low silica variety can be formed by slab

melting with various extent of mantle interaction during ascending (Kay, 1978;

Kelemen et al., 2004; Martin et al., 2005) or high-pressure fractionation of mantle-

derived basaltic magma (Castillo, 2008; Castillo et al., 1999; Danyushevsky et al.,

2008; Macpherson et al., 2006; Rodríguez et al., 2007; Rooney et al., 2011). The

adakitic nature of Quaternary volcanic rocks from the southern CAVA has been a

matter of debate. Some argued that they are slab melts. The other suggested that

they originated from mantle-derived basaltic magma. The petrology and mineral

chemistry data obtained in this study showed that western Panamanian adakitic

magmas experienced extensive magma evolution from basalt to dacite in the lower

crust (~35-40 km depth), which support for the latter model. Here, we performed

trace element and isotope model to examine the validity of the two models for

generation of Quaternary adakite-like lavas from the southern CAVA.

5.2.1. Melting of the subducting oceanic crust

We performed melting models for several possible subducting oceanic crusts

in this region (Fig. 1.19): i) Cocos ridge basalts (13.0-14.5 Ma; Harpp et al., 2004)

as the subducting Galapagos hotspot trace, ii) typical oceanic ridge tholeiites from

Costa Rica Rift (DSDP Site 504; Hubberten et al., 1983) and iii) the Cocos-Nazca

ridge basalts (tholeiitic lavas from the Malpelo ridge), suggested as normal oceanic

crust by Harpp et al. (2004) (Fig. 1.1).

The batch melting models were carried out following the approach of

Macpherson et al. (2006). We use partition coefficients from Sen and Dunn (1994)

and assume a residual mineral assemblage of 50% garnet and 50% clinopyroxene,

resulted from partial melting of amphibolites at 1050-1100 °C and 1.6-2.0 GPa in

the experimental study of Rapp and Watson (1995). The results show that batch

melting of all the possible subducting crusts cannot produce the compositional

variations of the adakitic rocks in southern CAVA (Fig. 1.19). The batch melting

models with different garnet to clinopyroxene ratios from 1:9 to 9:1 in the residue

did not change the results significantly.

According to a model conducted by Drummond and Defant (1990),

38

Panamanian adakitic rocks can be produced by melting of a MORB source with the

residual mineral assemblage of a hornblende eclogite. It appears that the melting

depths (~80-100 km) of the subducting slab underneath the southern Central

America subduction zone agree well with the residual phases of a hornblende

eclogite (Hebert et al., 2009; Husen et al., 2003; Syracuse and Abers, 2006).

Following the approach of Drummond and Defant (1990), we performed batch

melting model of the possible subducting slab using the residual mineralogy of a

hornblende eclogite (Cpx : Hbl : Grt : Qz = 0.60 : 0.15 : 0.15 : 0.10) with partition

coefficients of Sen and Dunn (1994). The result also shows that the adakitic

signatures of the volcanic rocks in western Panama and southern Costa Rica cannot

be generated from slab melting (Fig. 1.20). Moreover, if the western Panamanian

and southern Costa Rican adakitic lavas are slab melts, the mafic associated with

adakitic rocks in this study would be expected to be slab melts interacted with sub-

arc mantle during ascent. Rapp et al. (1999) showed that experimental melts

derived from the reaction between eclogitic melts and sub-arc mantle at 1100 °C

and 3.8 GPa have Mg# of 0.52-0.55 and SiO2 of 64.4-65.4 wt.%, inconsistent with

the compositional variations (Mg# = 0.51-0.66, SiO2 = 51.5-56.8 wt.%) of basalt

and basalt andesites in this study.

The results of trace element models are consistent with isotope data. The Nd

and Sr isotope data of the adakitic rocks from western Panama and southern Costa

Rica are consistent with those of mantle-derived arc basalts that erupted about 2

Ma prior to the adakitic rocks in the same region (Pliocene Costa Rica arc lavas;

~3.6 Ma; Gazel et al., 2009) whereas the 87Sr/86Sr ratios are relatively higher than

the subducting slab, the Cocos ridge (Fig. 1.13b). This suggests that they are not

direct products of slab melting.

It may be possible that the isotopic composition of the slab melts is modified

by subducted sediments (Kay, 1978; Sajona et al., 2000; Singer et al., 2007;

Yoshikawa et al., 2016). To confine the Sr, Nd and Pb isotopic contribution from

subducted sediments, we present mixing models between the subducting slab (the

Cocos ridge basalt; Hoernle et al., 2000) and sediments in Figure 1.13. For possible

subducting sediments, we use a mixture of 70% hemipelagic sediments and 30%

carbonate sediments from DSDP Site 495 (Patino et al., 2000), off the coast of

Guatemala on the subducting Cocos Plate (Fig. 1.1). The results show that the

39

mixture of the Cocos ridge basalt and the subducting sediments may partly explain

Pb-Nd-Sr isotopic compositions of the western Panamanian and southern Costa

Rican adakitic rocks. However, sediment contributions up to 3% and 15% for Nd-

Sr and Pb isotopic compositions, respectively, appear to be too high, considering

that ≤0.1 % sediment input is required for central Costa Rican volcanic front lavas

(~0.3 Ma; Carr et al., 2007), produced by partial melting of sub-arc mantle

metasomatized by slab melts derived from the subducting Galapagos hotspot track

(the Seamount Province; Gazel et al., 2009). The flux of subducted sediments into

a sub-arc mantle wedge underneath the southern CAVA is suggested to be limited

along the Central America subduction zone based on their low B/La, Ba/La and

10Be/9Be ratios (Carr et al., 2003; Leeman et al., 1994).

Figure 1.19. Binary plots of (a) Sr/Y and (b) La/Yb. The solid line, black and grey

dashed lines indicate partial melting of the Cocos ridge basalts average (Harpp et

al., 2004), normal oceanic ridge basalts produced at Costa Rica Rift (Hubberten et

al., 1983) and the Malpelo ridge (Harpp et al., 2004), respectively. The calculated

melt compositions (residual 50% Cpx and 50% Grt) are illustrated with varying

degree of partial melting (tick marks, per 10%). The location of DSDP Site 504,

Cost Rica Rift, and CNS crust is presented in Figure 1.1.

40

Figure 1.20. Binary plots of (a) Sr/Y and (b) La/Yb. The solid line, black and grey

dashed lines indicate partial melting of the Cocos ridge basalts average (Harpp et

al., 2004), normal oceanic ridge basalts produced at Costa Rica Rift (Hubberten et

al., 1983) and the Malpelo ridge (Harpp et al., 2004), respectively. The melt

compositions are calculated using the residual mineralogy of a hornblende eclogite,

following the approach of Drummond and Defant (1990) and illustrated with

varying degree of partial melting (tick marks, per 10%).

5.2.2. Fractional crystallization of mantle-derived arc magmas and magma

mixing

Fractional crystallization model of Pliocene (~3.6 Ma) regional arc basalts

(Gazel et al., 2009) are performed to examine if the adakitic rocks of western

Panama and southern Costa Rica can be formed by fractionation of a mantle-

derived basaltic magma (Fig. 1.21). We use a high pressure mineral assemblage

(Cpx : Opx : Grt : Amp = 52.8 : 17.2 : 12.3 : 17.4) obtained by high pressure (1200

MPa and 925 °C) experiments (Müntener et al., 2001) because petrology and

mineral chemistry data suggest fractionation in the lower crust (~35-40 km) played

an important role in geochemical evolution of the Quaternary adakitic rocks from

western Panama. The results show that geochemistry of the adakitic rocks is best

explained by fractional crystallization (~10-35 %) of the Pliocene Costa Rican arc

magma in the same region. The bulk partition coefficients are 3 for Y and 3.1 for

Yb to elucidate the trends in the fractional crystallization model. If amphiboles

were the only phase to retain Yb and Y, the bulk partition coefficients for Y and Yb

drop in 1.7 and 0.7, respectively (Fig. 1.22). These results are consistent with the

geochemical evidence for garnet fractionation in the western Panamanian and

southern Costa Rican adakitic rocks (Fig. 1.12).

41

We also attempt a low pressure fractional crystallization model of a basaltic

andesite magma, fractionated from a mantle-derived arc magma (Pliocene Costa

Rican arc basalt average; Gazel et al., 2009), because the evolved adakitic rocks in

this study consist of low pressure assemblage, which are low-P amphiboles,

plagioclase and Fe-Ti oxide. The details about the model are described in Appendix

8.1. The result shows that fractional crystallization (~50 %) of a basaltic andesite

magma did not change the results of the high pressure fractional crystallization

model significantly (Fig. 1.21), suggesting that low pressure fractional

crystallization may also have an influence on their geochemical variations.

The evolved adakitic rocks in this study display disequilibrium textures in the

minerals (Fig. 1.4 and 1.5), indicative of magma mixing (Kuşcu and Floyd, 2001).

Particularly, sample M63aKH-b (dacite) have euhedral to subhedral pyroxene

phenocrysts and thick Fe-Ti oxide rims in amphibole phenocrysts (Fig. 1.5 and 1.6)

in contrast to sample M63aKH (dacite), suggesting that they may have undergone

different evolution paths. Clinopyroxene cores in the dacite have high Mg# of 0.76-

0.87, which represent antecrysts from a cognate basaltic magma. These results may

support the magma mixing hypothesis for the western Panamanian adakitic rocks.

For magma mixing modeling, we use a mantle-derived arc magma (Pliocene Costa

Rican arc basalt average; Gazel et al., 2009) and an evolved magma (sample

M63aKH) as end members (Fig. 1.21). The results show that mixing between the

mantle-derived melt and the dacite magma reproduce the compositional variations

of the adakitic rocks from western Panama and southern Costa Rica (Fig. 1.21).

Reversed zoning of plagioclase with An-rich rims and patchy- or oscillatory-zoned

amphiboles in the dacites may be resulted from variations in conditions (e.g.

temperature, fO2) in magma chamber, mixing with incoming batches of more mafic

magma.

42

Figure 1.21. Binary plots of (a) Sr/Y and (b) La/Yb. The black solid line shows

fractional crystallization of a mantle-derived arc magma (Pliocene arc magma; ~3.5

Ma; Gazel et al., 2009), using a high pressure mineral assemblage (Müntener et al.,

2001) and partition coefficients from Nandedkar et al. (2016), Sen and Dunn

(1994) and Green et al. (2000). The grey solid line indicates fractional

crystallization at low pressure, using a low pressure assemblage based on the

phenocrysts in the evolved adakitic rock from western Panama. The fractional

crystallization melt compositions are illustrated with the fraction of melt remaining

(tick marks, per 0.1). The dotted line represents the mixing model between a

mantle-derived magma and dacite magma (tick marks, per 10%).

Figure 1.22. Binary plots of (a) Sr/Y and (b) La/Yb. The black solid line illustrates

fractional crystallization of a mantle-derived arc magma (Pliocene arc magma; ~3.5

Ma; Gazel et al., 2009), using the mineral assemblage without garnet (Cpx : Opx :

Grt : Amp = 52.8 : 17.2 : 0 : 30). The partition coefficients are identical with those

in Figure 1.21.

43

5.2.3. The role of crystal mush in the western Panamanian adakitic suites

In our model, crystallization of amphibole and garnet fractionation at lower

crustal reservoir is the key component to produce adakite-like signatures in the

western Panamanian and southern Costa Rican arcs. Recently, crystal mush zones

in lower crusts are proposed to elucidate the magmatic differentiation in arc crusts

and compositional variation among arc magmas (Bachmann and Bergantz, 2004,

2008; Davidson et al., 2007; Smith, 2014).

Lines of geochemical evidence support for garnet and amphibole-dominant

fractionation in the crystal mush zones at lower crust beneath western Panama and

southern Costa Rica (Fig. 1.12). The compositional variations in the high-P

amphibole phenocrysts with patchy or oscillatory zoning in the dacites and their

cumulative enclave suggest that they may be derived from chemically

heterogeneous magma batches at the lower crustal magma reservoir. A wide range

of compositional variations (Mg# and SiO2 contents) in the calculated melts in

equilibrium with the high-P amphiboles can indicate that they have undergone

progressive differentiation within a diverse source region at lower crustal depth.

Estimated pre-eruptive conditions (variable pressure, temperature and fO2) from

amphibole and pyroxene phenocrysts suggest that there may be thermally and

chemically variable crystal mush zones in the western Panamanian arc crust. This

model is consistent with the results of Hidalgo and Rooney (2010), suggesting that

adakitic magmas at Baru volcano, adjacent to our study area, have undergone

magma stalling and amphibole-dominant fractionation in the crystal mush zone at

lower crustal depth. Taken together, we suggest that the role of crystal mush at

lower crustal depth is possibly significant to produce overall adakite-like

magmatism in western Panamanian and southern Costa Rican arcs, where the

Cocos ridge subducts.

44

5.3. Petrogenesis of the adakite-like magmas in the western Panamanian and

southern Costa Rican arc

We propose a petrogenetic model for the western Panamanian and southern

Costa Rican adakitic rocks based on the results of this study (Fig. 1.23).

The basalt magmas with elevated La/Yb and Sr/Y and low Yb and Y are

derived from sub-arc mantle metasomatized by enriched slab melts. They stalled in

the lower crust close to the crust-mantle boundary (~35-40 km depth) and

underwent various degrees of fractionation. Early stage fractionation was

dominated by olivine, orthopyroxene and clinopyroxene, which formed basaltic

andesite. This was followed by late stage fractionation predominated by amphibole

and garnet in the crystal mush zones at lower crustal level, forming adakitic

andesite and dacite. These evolved magmas stalled in the upper crust (~8 km),

experiencing further evolution before eruption. Mixing between dacitic to andesitic

magmas and basaltic magmas involved in adakite-like andesite and dacite magmas

in the western Panamanian and southern Costa Rican arc, considering magma

mixing textures observed in the trachyandesite and dacite and the presence of high

Mg# pyroxene phenocrysts in the dacite as antecrysts from a cognate basaltic

magma.

Figure 1.23. (a) Schematic illustration showing the magma plumbing systems for

adakitic rocks in western Panama and southern Costa Rica, Central America. (b)

Pressure and temperature conditions of western Panamanian adakitic magmas

during ascent and cooling. Arrows indicate adiabatic magma ascent and horizontal

lines represent cooling and crystallizing paths. The liquidus and amphibole-in

boundaries for basalt and dacite are modified from Smith (2014).

45

6. Conclusion

The Pleistocene volcanic rocks from western Panama and southern Costa Rica

display adakite-like geochemical features. We investigated petrology, mineral

chemistry, whole-rock major and trace element and isotope geochemistry of the

Pleistocene adakitic volcanic rocks in western Panama to constrain their

petrogenesis.

1. Pre-eruptive pressure and temperature estimated using amphibole and

pyroxene phenocrysts show that the adakitic lavas from western Panama have

experienced fractionation at the crust-mantle boundary (~35-40 km) and upper

curst (~8 km).

2. High-P amphiboles (≥920 MPa) have Mg# ranging from 0.54-0.77,

indicating the adakitic basalt magma has undergone moderate degree of

fractionation forming andesitic and dacitic magmas in the lower crust. Presence of

the low-P amphibole phenocyrsts suggests that further fractionation in the upper

crust. The patchy-zoned amphiboles showing both high and low-P amphibole

compositions may represent high-P amphiboles partially re-equilibrated with host

melts at low pressure.

3. Various disequilibrium textures observed in trachyandesite and dacite

suggest magma mixing involves in their evolution. The high Mg# (0.76-0.87)

pyroxene phenocrysts in the dacite may represent antecrysts from a cognate

basaltic magma. A mixing model shows that mixing with incoming batches of a

mafic magma into a more evolved magma is responsible for the compositional

variations of the adakitic rocks from western Panama and southern Costa Rica.

4. The trace element and isotope geochemistry of western Panamanian and

southern Costa Rican adakitic rocks are best explained by the following processes.

The basalt magmas with adakitic signatures are derived from sub-arc mantle

metasomatized by enriched slab melts and, stalled at the crust-mantle boundary

(~35-40 km depths). Basaltic andesites are formed by early stage fractional

crystallization of mantle-derived basaltic melts in the lower crust where they

fractionated olivine, pyroxene and garnet. Finally, late stage fractionation

predominated by garnet and amphibole in the crystal mush zone at lower crustal

level forms a wide compositional range of evolved magmas including adakitic

andesite and dacite.

46

Chapter 2. Platinum Group Element recycling in the

Central American volcanic arc: a preliminary study

1. Introduction

Platinum group elements (PGEs) are indicators of parting melting in the

mantle or sulfide saturation in a melt because they are fractionated into sulfide

phases (Barnes et al., 2015; Kiseeva et al., 2017; Lorand and Luguet, 2016;

Rehkämper and Halliday, 1997). PGEs are subdivided into two groups on the basis

of their melting temperature (O’Driscoll and González-Jiménez, 2016): i) IPGE

(Iridium group; Os, Ir, Ru; melting T >2000°C) and ii) PPGE (Palladium group; Pd,

Pt, Rh; melting T <2000°C). IPGEs tend to be compatible whereas PPGEs are

more incompatible during mantle melting (Bockrath et al., 2004; Pitcher et al.,

2009).

There have been numerous studies on the behavior of lithophile and high-field

strength elements (LILEs and HFSEs) in subduction zones, suggesting that

aqueous fluids and/or slab melts from the subducting slab transfer efficiently

subducting components (LILEs and HFSEs) to the sub-arc mantle (Ayers, 1998;

Class et al., 2000; Hermann and Rubatto, 2009; Plank and Langmuir, 1993; Singer

et al., 2007). However, there have been fewer studies on the behavior of

chalcophile elements (PGEs, Au and Cu) at subduction zones. Previous studies on

the subduction-related mantle wedge xenoliths demonstrate that PPGEs are

significantly fluxed into the sub-arc mantle wedge relative to IPGEs (Kepezhinskas

et al., 2002; Rehkämper and Halliday, 1997). Kepezhinskas et al. (2002)

documented that re-fertilization by slab-derived fluids and melts caused

enrichments of Pd and Pt relative to the Ir group in the sub-arc mantle beneath the

Kamchatka arc. Also, Dale et al. (2009) suggested that Pd and Pt are efficiently

fluxed into the mantle wedge from the subducting slab compared to Ir and Ru.

These are consistent with the results of the studies on the arc lavas that PPGE are

more mobile in slab-derived hydrous fluids and/or sediment melts, forming Pt and

Pd enrichments in arc lavas (Dale et al., 2012; Woodland et al., 2002).

47

In this study, we investigated PGE geochemistry of primitive volcanic rocks

from the Central America volcanic arc (CAVA) system to understand the effect of

subduction input on PGE geochemistry of subduction zone magmas. Previously,

the subducting input and melting processes in the CAVA have been reported by

many studies (Bolge et al., 2009; Carr et al., 1990; Carr et al., 2003; Carr et al.,

2007; Feigenson and Carr, 1993; Feigenson et al., 2004; Gazel et al., 2009; Gazel

et al., 2011; Goss and Kay, 2006; Leeman et al., 1994; Patino et al., 2000; Saginor

et al., 2013; Whattam and Stern, 2016). Nicaraguan volcanic front lavas were

formed by melting of sub-arc mantle metasomatized by aqueous fluids and

sediment melts derived from the subducting normal slab, East Pacific Rise crust

(Bolge et al., 2009; Carr et al., 2007; Gazel et al., 2009; Saginor et al., 2013). Sub-

arc mantle beneath central Costa Rica and western Panama is metasomatized by

slab melts from the subducting slab overprinted by the Galapagos hotspot traces;

the Seamount Province and the Cocos ridge (Feigenson et al., 2004; Gazel et al.,

2009; Gazel et al., 2011), respectively.

We chose six Nicaraguan and seven central Costa Rican basalts with high Mg

contents (≥ 8.0 wt. % MgO). The six western Panamanian adakitic rocks are

representative of the Pleistocene adakitic volcanism in western Panama where the

Cocos ridge subducts. Their whole-rock major and trace elements were determined

to confirm their representative of the CAVA. PGE concentrations of them were

measured using a Ni-sulfide fire assay-isotope dilution method.

48

2. Geological Background and Samples

The southern Central American arc volcanism began to form during the east-

dipping subduction of the Farallon plate at the western margin of the Caribbean

Large Igneous Province (139-83 Ma) (Hoernle et al., 2004) at ~70 Ma (Loewen et

al., 2013). During the early Miocene (~23 Ma), the Farallon plate had split into the

Cocos and Nazca plate (Lonsdale, 2005; Mann, 2007) and subducted underneath

the Caribbean plate at a rate ranging from ~8 cm/yr of the Cocos plate (DeMets et

al., 2010; Kobayashi et al., 2014).

Beneath Nicaraguan arc, the oceanic crust formed at the East Pacific Rise has

been subducting at the spreading rate of ~65 mm/yr (Mann, 2007) (Fig. 2.1). The

subducting slab underneath Nicaragua consists of upper hemipelagic sediment,

lower carbonate sediment and normal mid-ocean ridge basalt (Patino et al., 2000).

Nicaraguan volcanic front lavas were formed by melting of sub-arc mantle

metasomatized by aqueous fluids and sediment melts from the subducting slab

(Bolge et al., 2009; Carr et al., 2007; Gazel et al., 2009; Saginor et al., 2013).

On the other hand, arc volcanism beneath central Costa Rica and western

Panama has been significantly affected by the Galapagos hotspot components

(Feigenson et al., 2004; Gazel et al., 2009; Gazel et al., 2011). The interaction

between Greater Antilles subduction zone and Galapagos plume had initiated at 90-

95 Ma (Loewen et al., 2013). The Cocos-Nazca spreading ridge, formed by the

division of the Farallon plate, has been interacted with Galapagos hotspot plume

and this interaction has generated Galapagos hotspot tracks; the Cocos ridge and

the Seamount Province (Loewen et al., 2013; Meschede and Barckhausen, 2001).

Sub-arc mantle beneath southern Costa Rica and western Panama is metasomatized

by slab melts from the subducting slab overprinted by the Seamount Province and

the Cocos ridge, respectively (Feigenson et al., 2004; Gazel et al., 2009; Gazel et

al., 2011).

We have investigated 19 young arc volcanic rock samples (<2.58 Ma) from

Nicaragua, central Costa Rica and western Panama, Central America. Samples

from Nicaragua and central Costa Rica are of primitive basaltic compositions with

high MgO contents (≥8.0 wt.%). The petrogenesis of the adakitic rocks from

western Panama were investigated in Chapter 1.

49

Figure 2.1. Tectonic setting of Central America modified from Gazel et al. (2009).

PFZ: Panama Fracture Zone, EPR: East Pacific Rise, CNS: Cocos-Nazca

Spreading Center.

50

3. Analytical Methods

3.1. Whole-rock trace element analysis

For trace element analyses of the samples (NIC-CN4 from Nicaragua, CR-

PP8 from central Costa Rica and six samples from western Panama), about 30 mg

of each powdered sample was dissolved in an HF-HNO3-HClO4 mixture and

heated at 120°C for 1 day. The dried samples were dissolved in 2% HNO3 and

diluted 4000 times. Indium and Re were used as internal standards to correct

instrumental and signal drift. The final solutions were analyzed by an ELAN 6100

quadrupole ICP-MS (Perkin-Elmer) at KOPRI. Reference materials (BHVO-2 and

AGV-2) were analyzed together with the samples to check accuracy and precision

of the measurement, and the results are reported in Appendix 4. Our results are

consistent with the values from Jochum et al. (2016) within 10% for most elements,

except for Hf and Ta (< 20%) for AGV-2 and Nb, Hf, Ta, and Pb (< 20%) for

BHVO-2.

3.2. PGE analysis

PGE concentrations of the Central American arc lavas were measured using a

Ni-sulfide fire assay-isotope dilution method following the procedure presented in

Park et al. (2012). Briefly, 3-5 g of sample powder was mixed with Ni, S and

sodium borax powder in the ratio of sample : Ni : S : borax = 10 : 1 : 0.5 : 10. The

mixtures were then spiked with a solution enriched in 99Ru, 105Pd, 191Ir, 195Pt. After

drying the mixture for 1 hour at 100°C, we fused them in a preheated furnace at

1150°C for 30 minutes. Quenched Ni-sulfide beads were dissolved in 6M HCl and

then, the solution was filtered through a filter paper (0.45 lm cellulose membrane,

Millipore). The collected filter paper was digested in 4 ml of aqua regia and

refluxed at 140°C for 2 hours. After digestion of the filter is completed, the solution

was dried down to ~0.1 ml. The solution is then diluted with 6 ml of 2% HNO3 and

refluxed at 120°C for 2 hours. The concentrations of the final solution were

measured by an iCAP Q quadrupole ICP-MS (Thermo Fisher) and ELAN 6100

quadrupole ICP-MS (Perkin-Elmer) at KOPRI. Procedural blanks were determined

by using 10g of sodium borax, 0.5 g of Ni and 0.25 g of S or 6g of sodium borax,

51

0.3 g of Ni and 0.15 g of S without sample powder (Table 2.1). TDB-1 is used as

the reference material to evaluate the accuracy of analyses (Table 2.2). All the PGE

data result from subtracting the average of blanks.

Table 2.1. Procedural blanks (ng/g).

Table 2.2. Analytical results of PGE in PGE reference material TDB-1 (ng/g).

52

4. Results

4.1. Whole-rock geochemistry

The samples in this study range from basalt to dacite (SiO2 >48.0 wt.%,

K2O+Na2O <3.0 wt.%, Al2O3 >10.0 wt.%, Na2O = 1.1-3.7 wt.%) (Fig. 2.2).

Nicaragua arc lavas are enriched in the LILEs and the LREEs compared to the

HFSEs (Fig. 2.3a). In contrast, arc lavas from central Costa Rica and western

Panama have significant enrichment of highly incompatible elements and depletion

of the HREEs and the HFSE compared to average value of global MORBs from

Gale et al. (2013) (Fig. 2.3b), indicating that subduction of enriched Galapagos

hotspot tracks (the Seamount Province and the Cocos ridge) have influenced arc

lavas in central Costa Rica and adakitic rocks in western Panama (Gazel et al.,

2009; Gazel et al., 2011; Saginor et al., 2013).

The volcanic rocks from the southern Central American arc were plotted on

the pre-defined subduction input components (Fig. 2.4). U/Th ratio is used as a

proxy of the subducting sediment input (Patino et al., 2000; Saginor et al., 2013).

Nicaraguan lavas have the highest U/Th ratios (Fig. 2.4), indicating the subduction

of U-rich hemipelagic sediments under the Nicaraguan arc (Leeman et al., 1994;

Patino et al., 2000). U/Th ratios decrease toward central Costa Rica and western

Panama (Fig. 2.4) and this trend results from the subduction of Galapagos hotspot

traces under the region (Gazel et al., 2009; Saginor et al., 2013). Also, Ba/Nb ratios,

suggested to be a tracer and magnitude of subduction input by Pearce and Stern

(2006), decrease with the distance along the volcanic front (Fig. 2.4). The highest

U/Th and Ba/Nb ratios of Nicaragua lavas indicate maximum subduction input in

the Nicaraguan arc, consistent with their high B/La, Ba/La and 10Be/9Be ratios

(Carr et al., 2003; Hildreth and Moorbath, 1988; Leeman et al., 1994; Patino et al.,

2000). Increases of La/Yb and Nb/Yb ratios toward central Costa Rica and western

Panama (Fig. 2.5) indicate source enrichments or high mantle fertility by slab melts

derived from the subducting Galapagos hotspot tracks (Gazel et al., 2009; Gazel et

al., 2011).

We confirm that the samples in this study are representative of the regional

geochemical features of the southern Central America subduction zone using major

and trace element data.

53

Figure 2.2. TAS diagram of the arc lavas and adakitic rocks in this study. Grey

circles indicate volcanic front lavas of the southern Central America (Carr et al.,

2014).

Figure 2.3. Primitive mantle (McDonough and Sun, 1995) normalized multi-

element diagrams of the rocks from (a) Nicaragua and (b) central Costa Rica and

western Panama. The composition of average MORB is from Gale et al. (2013) and

illustrated by dotted lines. Shaded areas indicate (a) the volcanic front lavas of

Nicaragua from (Carr et al., 2014) and (b) central Costa Rican lavas and western

Panamanian adakitic rocks from Carr et al. (2014), Gazel et al. (2011) and Hidalgo

and Rooney (2014).

54

Figure 2.4. U/Th and Ba/Nb vs distance along the volcanic front (km) diagrams.

Distance along the volcanic front (km) is measured from a point about 22 km NW

of Tacana volcano, northern Guatemala, Central America (Carr et al., 2007). Small

circles and triangles represent volcanic front lavas of the southern Central America

are from Carr et al. (2014). Small squares indicate the adakitic rocks from western

Panama and southern Cost Rica are from Gazel et al. (2011) and Hidalgo and

Rooney (2014).

Figure 2.5. Nb/Yb and La/Yb vs distance along the volcanic front (km) diagrams.

Distance along the volcanic front (km) is and the references are identical with

Figure 2.4.

55

4.2. Platinum group element geochemistry

The PGE concentrations of the volcanic lavas from Nicaragua, central Costa

Rica and western Panama are given in Table 2.3. The volcanic rocks in this study

were analyzed in duplicate (n=2-3) and reproducibility are revealed as 42-97% for

Ir, 38-89% for Ru, 44-89 % for Rh, 34-100% for Pt and 82-99 % for Pd. Ir and Ru

concentrations of them are variable (Fig. 2.6a and 2.6b), though Rh contents of

Nicaraguan lavas are higher than lavas from central Costa Rica and Panama (Fig.

2.6c). Pd contents are significantly enriched in Nicaraguan lavas, compared to

central Costa Rican lavas and western Panamanian adakitic lavas (Fig. 2.6e).

Figure 2.6. Binary plots of PGE concentrations (ppb) vs MgO (wt.%), indicating

indices of fractionation for the Central American volcanic lavas. Dotted lines

indicate duplicate analyses.

56

Table 2.3. Concentrations of MgO (wt.%), PGE (ng/g) and trace element ratios of volcanic lavas from the CAVA.

57

5. Discussion

5.1. Effects of magma differentiation

The purpose of this section is to sort out primitive rocks to investigate PGE

recycling during the mantle process. The results show that PGE concentrations

decrease with decreasing MgO contents in the Central American arc samples with

≤9 wt.% MgO (Fig. 2.6). This suggests that onset of crystallization of PGE-bearing

minerals is at ~9 wt.% MgO.

Formation of an immiscible sulfide melt causes sequestration of chalcophile

elements (Cu, Au, PGEs) with depletion of Cu, Pd and Au (Brenan et al., 2016). Pd

partition preferentially into sulfide melts relative to Cu (Peach et al., 1990), leading

to low Pd/Cu ratios by sulfide fractionation (Dale et al., 2012; Park et al., 2013).

Pt-rich alloy saturation in the residue results in fractionation Pt from Pd, and it can

cause high Pd/Pt ratios of lavas (Lorand and Luguet, 2016; Park et al., 2013).

Residual Ir-Ru alloys (e.g. laurite) and/or chromite may increase Pd/Ir ratios of

lavas (Brenan et al., 2016; Kepezhinskas et al., 2002). Also, Pd/Ir ratios correspond

to Cu-rich base metal sulfide fractionation (Lorand and Luguet, 2016). Using these

indicators, we evaluated that several volcanic rocks in this study have been

undergone sulfide saturation and excluded to minimize the effect of magma

differentiation (Fig. 2.7 and Table 2.3). Thus, the most primitive rocks were

selected for investigating mantle process only.

Figure 2.7. Binary plots of (a) Pd/Cu*1000 vs MgO (wt.%), (b) Pd/Pt vs MgO

(wt.%) and (c) Pd/Ir vs MgO (wt.%) for the Central American volcanic lavas. The

volcanic rocks with arrows are minimum values of Pd/Ir ratios. Vertical dotted lines

indicate duplicate analyses.

58

5.2. Effects of mantle process and future work

The primitive-normalized PGE patterns along the southern Central American

arc are featured by high PPGE (Pd, Pt, Rh)/IPGE (Ru, Ir) ratios, typical signature

of subduction zones (Dale et al., 2012; Woodland et al., 2002) (Fig. 2.8).

Pd (7.2-9.4 ppb) and Pt (5.0-5.6 ppb) contents are significantly enriched in

Nicaraguan lavas, compared to central Costa Rica and western Panama lavas,

which result in high Pd/Ir (89-458) in Nicaragua (Fig. 2.9). These results are

consistent with those of the Tonga arc lavas that generally have high Pd contents

and Pd/Ir ratios (Dale et al., 2012; Park et al., 2015). The low Pd/Ir ratios of the

lavas from Costa Rica (12.9-13.4 of Pd/Ir) and western Panama (32.4-69.9 of Pd/Ir)

are similar to those of OIBs (Chazey and Neal, 2005; Ely and Neal, 2003; Ireland

et al., 2009). These differences could result from two possible processes: i) more

efficient transfer of Pd and Pt into the sub-arc mantle via fluids (and possibly

sediment melts) than slab melts, ii) different mantle melting conditions such as

oxygen fugacity, degree of melting and melting depth. Future work will focus on

the examination of mantle melting modeling to investigate which process can

produce the PGE features of the Central American arc system.

Figure 2.8. Primitive mantle-normalized (McDonough and Sun 1995) PGE

patterns for primitive arc lavas from Nicaragua, central Costa Rica and western

Panama.

59

Figure 2.9. Pd concentration (ppb) and Pd/Ir ratio vs distance along the volcanic

front (km) diagram. Vertical dotted lines indicate duplicate analyses.

60

Appendix

Appendix 1. Electron microbe data of standards, analyzed at KOPRI.

Oxide concentrations are in wt.%.

61

Appendix 2. LA-ICP-MS analysis results (KOPRI) of NIST 612 glass and BCR-2g.

Trace element concentrations are in ppm. Si and Ca are in wt.%.

62

Appendix 3. LA-ICP-MS analysis results (KBSI) of BCR-2g and AGV-2.

Trace element concentrations are in ppm. Si and Ca are in wt.%.

63

Appendix 4. Trace elements analysis results (KOPRI) of BHVO-2 and AGV-2.

Concentrations are in ppm.

Appendix 5. Major elements analysis results (SNU) of AGV-2.

Oxide concentrations are in wt.%.

64

Appendix 6.1. Electron microbe data of minerals of M91aKH.

Major element concentrations are in wt.%.

65

Appendix 6.2. Electron microbe data of minerals of M38KH.

66

Major element concentrations are in wt.%.

67

Appendix 6.3. Electron microbe data of minerals of M63aKH.

68

Major element concentrations are in wt.%.

69

Appendix 6.4. Electron microbe data of minerals of M63aKH-b.

70

Major element concentrations are in wt.%.

71

Appendix 7. Major element variation diagrams against SiO2 (wt.%) for the

adakitic rocks from western Panama and southern Costa Rica. Yellow squares and

small lemon squares represent western Panamanian and southern Costa Rican

adakitic rocks in this study and from Gazel et al. (2011) and Hidalgo et al. (2014),

respectively.

72

Appendix 8.1. Details about fractional crystallization model of a fractionated

magma in section 5.2.2.

We attempt fractional crystallization modeling of a basaltic andesitic magma

to account the results of this study that high-P and low-P amphiboles may be

fractionated from cognate andesitic melts and evolved dacitic melts, respectively

(section 5.1.2).

First, we calculate the degree of fractional crystallization to produce a

differentiated melt with ~5 wt.% of MgO, coinciding with MgO contents of the

basaltic andesite in this study, with the help of Petrolog3 software (Danyushevsky

and Plechov, 2011) and the percent is ~20%. The calculations are performed at a

QFM +3 buffer, based on the results of Fe3+/∑Fe data from He et al. (2017), and at

~1 GPa, which is an estimated pre-eruptive pressure using pyroxene phenocrysts in

this study. To assess for the crystallization phases, major element trends in CaO,

Al2O3, and MgO and FeOtotal are examined and show the evidence for olivine,

pyroxene and plagioclase fractionation (Appendix 8.2). The mineral-melt

equilibrium models were selected considering liquid compositions and the P-T

conditions conducted by the models in Petrolog3 (Ariskin et al., 1993; Beattie,

1993; Ford et al., 1983). For the starting melt composition, the basalt (sample

M64fKH) with 7.8 wt.% MgO has been assumed to be representative of parental

magmas with ~3 wt.% of pre-eruptive water contents constrained by Benjamin et al.

(2007).

Based on the extent of fractional crystallization estimated from Petrolog3,

melt compositions, produced from ~30% fractional crystallization of the Pliocene

Costa Rican arc basalts, are selected. The value of ~30% is approximate because

Petrolog3 cannot model amphibole fractionation. Thus, we assumed that the

percent of fractional crystallization is higher than ~20%. The partition coefficients

for dacitic melts (Arth and Barker, 1976; Fujimaki et al., 1984; Mahood and

Hildreth, 1983; Nash and Crecraft, 1985; Pearce and Norry, 1979) were used. The

mineral assemblage is based on the phenocrysts in the most evolved sample

(M63aKH-b) in this study (Pl : Amp : Bt : Qz = 0.56 : 0.16 : 0.20 : 0.08). The

results reveal that Sr/Y ratios of a basaltic andesite magma, obtained by ~30%

fractional crystallization of Pliocene arc basalt, decrease with a decline in Y

contents, whereas they have nearly increasing La/Yb ratios with a rise of Yb

contents.

73

Appendix 8.2. Major element variation diagrams for (a) CaO vs MgO (b) Al2O3 vs

MgO and (c) FeOtotal vs MgO in wt% oxide for the adakitic rocks from western

Panama and southern Costa Rica.

74

Appendix 9. Trace element variation diagrams against SiO2 (wt.%) for the adakitic

rocks from western Panama and southern Costa Rica.

75

Appendix 10. Binary plots of (a) Sr/Y vs Y and (b) La/Yb vs Yb to test under-

plated basalt melting. The residual mineral assemblage and partition coefficients

are identical with Figure 1.19. Tick marks illustrate varying degree of partial

melting (per 10%).

Adakitic signatures of volcanic rocks in arc regions can be produced by

melting at the base of thickened crust where garnet is stable and thermal state is

sufficient to form lower crustal melts (Atherton and Petford, 1993; Dai et al., 2017;

Goss and Kay, 2009; Kay and Kay, 2002; Kay et al., 1994; Petford and Atherton,

1996; Wang et al., 2007).

In the western Panamanian and southern Costa Rican arc, the arc crust was

thickened by magmatic accretion since the Miocene (~22Ma) with the growth rate

of ~10 km3/km/Ma (MacKenzie et al., 2008). The thickness of the arc crust has

been geophysically determined to be ~40 km (Dzierma et al., 2010; Gazel et al.,

2015; Hayes et al., 2013; MacKenzie et al., 2008; Protti et al., 1994), containing

mafic crust formed by underplating and fractionation of basaltic magmas (Sallarès

et al., 2001).

Delamination of lower crust also can trigger to yield adakitic magmas (Kay

and Kay, 1993; Kay et al., 1994; Wang et al., 2007), while the evidence for

delamination, such as missing a deep crustal root and uplifting or increasing heat

flow, is lack at the Costa Rican arc (Hayes et al., 2013).

Based on the tectonic regime of western Panama and southern Costa Rica, we

attempt to perform melting modeling of under-plated arc basalts using regional arc

basalts from southern Costa Rica, Central America. The subducting slab

components under the Southern Central America volcanic front changed from

normal altered oceanic crust to the crust overprinted by the Galapagos hotspot

76

tracks (Gazel et al., 2009; Saginor et al., 2013), controlling the compositional

variation of mantle-derived magma at ~8-10 Ma (Abratis, 1998; Abratis and

Worner, 2001) or at least ~3 Ma (MacMillan et al., 2004; Morell, 2015). Thus,

partial melting of Pliocene (~3.6 Ma) and Miocene (10.5-14.1 Ma) arc basalts

(Gazel et al., 2009) are carried out as they are assumed to be under-plated basaltic

crust within the arc region. Batch melting of Pliocene regional arc basalts can

produce adakitic melts, but the extents of melting are extremely high (>80%).

77

REFERENCES Abratis, M. (1998) Geochemical variations in magmatic rocks from southern Costa Rica as

a consequence of Cocos Ridge subduction and uplift of the Cordillera de Talamanca.

Unpublished PhD. Thesis, Georg–August–Universität, Göttingen, 1-148.

Abratis, M. and Worner, G. (2001) Ridge collision, slab-window formation, and the flux of

Pacific asthenosphere into the Caribbean realm. Geology 29, 127-130.

Alonso-Perez, R., Müntener, O. and Ulmer, P. (2009) Igneous garnet and amphibole

fractionation in the roots of island arcs: experimental constraints on andesitic liquids.

Contributions to Mineralogy and Petrology 157, 541.

Anderson, J.L. and Smith, D.R. (1995) The effects of temperature and fO₂ on the Al-in-

hornblende barometer. American Mineralogist 80, 549-559.

Ariskin, A.A., Frenkel, M.Y., Barmina, G.S. and Nielsen, R.L. (1993) COMAGMAT: a

Fortran program to model magma differentiation processes. Computers & Geosciences

19, 1155-1170.

Arth, J.G. and Barker, F. (1976) Rare-earth partitioning between hornblende and dacitic

liquid and implications for the genesis of trondhjemitic-tonalitic magmas. Geology 4,

534-536.

Atherton, M.P. and Petford, N. (1993) Generation of sodium-rich magmas from newly

underplated basaltic crust. Nature 362, 144.

Ayers, J. (1998) Trace element modeling of aqueous fluid–peridotite interaction in the

mantle wedge of subduction zones. Contributions to Mineralogy and Petrology 132,

390-404.

Bachmann, O. and Bergantz, G.W. (2004) On the origin of crystal-poor rhyolites: extracted

from batholithic crystal mushes. Journal of Petrology 45, 1565-1582.

Bachmann, O. and Bergantz, G.W. (2008) Rhyolites and their source mushes across tectonic

settings. Journal of Petrology 49, 2277-2285.

Bachmann, O. and Dungan, M.A. (2002) Temperature-induced Al-zoning in hornblendes of

the Fish Canyon magma, Colorado. American Mineralogist 87, 1062-1076.

Barnes, S.J., Mungall, J.E. and Maier, W.D. (2015) Platinum group elements in mantle

melts and mantle samples. Lithos 232, 395-417.

Beattie, P. (1993) Olivine-melt and orthopyroxene-melt equilibria. Contributions to

Mineralogy and Petrology 115, 103-111.

Benjamin, E.R., Plank, T., Wade, J.A., Kelley, K.A., Hauri, E.H. and Alvarado, G.E. (2007)

High water contents in basaltic magmas from Irazu Volcano, Costa Rica. Journal of

78

Volcanology and Geothermal Research 168, 68-92.

Bindeman, I., Eiler, J., Yogodzinski, G., Tatsumi, Y., Stern, C., Grove, T., Portnyagin, M.,

Hoernle, K. and Danyushevsky, L. (2005) Oxygen isotope evidence for slab melting in

modern and ancient subduction zones. Earth and Planetary Science Letters 235, 480-

496.

Blundy, J. and Wood, B. (2003) Partitioning of trace elements between crystals and melts.

Earth and Planetary Science Letters 210, 383-397.

Bockrath, C., Ballhaus, C. and Holzheid, A. (2004) Fractionation of the platinum-group

elements during mantle melting. Science 305, 1951-1953.

Bolge, L.L., Carr, M.J., Milidakis, K.I., Lindsay, F.N. and Feigenson, M.D. (2009)

Correlating geochemistry, tectonics, and volcanic volume along the Central American

volcanic front. Geochemistry, Geophysics, Geosystems 10.

Brenan, J.M., Bennett, N.R. and Zajacz, Z. (2016) Experimental results on fractionation of

the highly siderophile elements (HSE) at variable pressures and temperatures during

planetary and magmatic differentiation. Reviews in Mineralogy and Geochemistry 81,

1-87.

Carr, M., Feigenson, M. and Bennett, E. (1990) Incompatible element and isotopic evidence

for tectonic control of source mixing and melt extraction along the Central American

arc. Contributions to Mineralogy and Petrology 105, 369-380.

Carr, M., Feigenson, M., Patino, L. and Walker, J. (2003) Volcanism and geochemistry in

Central America: Progress and problems. Inside the subduction factory, 153-174.

Carr, M.J., Feigenson, M.D., Bolge, L.L., Walker, J.A. and Gazel, E. (2014)

RU_CAGeochem, a database and sample repository for Central American volcanic

rocks at Rutgers University. Geoscience Data Journal 1, 43-48.

Carr, M.J., Saginor, I., Alvarado, G.E., Bolge, L.L., Lindsay, F.N., Milidakis, K., Turrin,

B.D., Feigenson, M.D. and Swisher, C.C. (2007) Element fluxes from the volcanic front

of Nicaragua and Costa Rica. Geochemistry, Geophysics, Geosystems 8.

Castillo, P.R. (2008) Origin of the adakite–high-Nb basalt association and its implications

for postsubduction magmatism in Baja California, Mexico. Geological Society of

America Bulletin 120, 451-462.

Castillo, P.R. (2012) Adakite petrogenesis. Lithos 134, 304-316.

Castillo, P.R., Janney, P.E. and Solidum, R.U. (1999) Petrology and geochemistry of

Camiguin Island, southern Philippines: insights to the source of adakites and other lavas

in a complex arc setting. Contributions to Mineralogy and Petrology 134, 33-51.

Chazey, W.J. and Neal, C.R. (2005) Platinum-group element constraints on source

79

composition and magma evolution of the Kerguelen Plateau using basalts from ODP

Leg 183. Geochimica et Cosmochimica Acta 69, 4685-4701.

Class, C., Miller, D.M., Goldstein, S.L. and Langmuir, C.H. (2000) Distinguishing melt and

fluid subduction components in Umnak Volcanics, Aleutian Arc. Geochemistry,

Geophysics, Geosystems 1.

Dai, H.-K., Zheng, J., Zhou, X. and Griffin, W. (2017) Generation of continental adakitic

rocks: Crystallization modeling with variable bulk partition coefficients. Lithos 272,

222-231.

Dale, C., Burton, K., Pearson, D., Gannoun, A., Alard, O., Argles, T. and Parkinson, I.

(2009) Highly siderophile element behaviour accompanying subduction of oceanic

crust: whole rock and mineral-scale insights from a high-pressure terrain. Geochimica

et Cosmochimica Acta 73, 1394-1416.

Dale, C.W., Macpherson, C.G., Pearson, D.G., Hammond, S.J. and Arculus, R.J. (2012)

Inter-element fractionation of highly siderophile elements in the Tonga Arc due to flux

melting of a depleted source. Geochimica et Cosmochimica Acta 89, 202-225.

Danyushevsky, L.V., Falloon, T.J., Crawford, A.J., Tetroeva, S.A., Leslie, R.L. and

Verbeeten, A. (2008) High-Mg adakites from Kadavu Island Group, Fiji, southwest

Pacific: Evidence for the mantle origin of adakite parental melts. Geology 36, 499-502.

Danyushevsky, L.V. and Plechov, P. (2011) Petrolog3: Integrated software for modeling

crystallization processes. Geochemistry, Geophysics, Geosystems 12.

Davidson, J., Turner, S., Handley, H., Macpherson, C. and Dosseto, A. (2007) Amphibole

“sponge” in arc crust? Geology 35, 787-790.

de Boer, J.Z., Defant, M., Stewart, R., Restrepo, J., Clark, L. and Ramirez, A. (1988)

Quaternary calc-alkaline volcanism in western Panama: regional variation and

implication for the plate tectonic framework. Journal of South American Earth Sciences

1, 275-293.

de Boer, J.Z., Defant, M.J., Stewart, R.H. and Bellon, H. (1991) Evidence for active

subduction below western Panama. Geology 19, 649-652.

Defant, M., Jackson, T., Drummond, M., De Boer, J., Bellon, H., Feigenson, M., Maury, R.

and Stewart, R. (1992) The geochemistry of young volcanism throughout western

Panama and southeastern Costa Rica: an overview. Journal of the Geological Society

149, 569-579.

Defant, M.J., Clark, L.F., Stewart, R.H., Drummond, M.S., de Boer, J.Z., Maury, R.C.,

Bellon, H., Jackson, T.E. and Restrepo, J.F. (1991a) Andesite and dacite genesis via

contrasting processes: the geology and geochemistry of El Valle Volcano, Panama.

80

Contributions to Mineralogy and Petrology 106, 309-324.

Defant, M.J. and Drummond, M.S. (1990) Derivation of some modern arc magmas by

melting of young subducted lithosphere. Nature 347, 662.

Defant, M.J., Richerson, P.M., De Boer, J.Z., Stewart, R.H., Maury, R.C., Bellon, H.,

Drummond, M.S., Feigenson, M.D. and Jackson, T.E. (1991b) Dacite genesis via both

slab melting and differentiation: petrogenesis of La Yeguada volcanic complex, Panama.

Journal of Petrology 32, 1101-1142.

DeMets, C., Gordon, R.G. and Argus, D.F. (2010) Geologically current plate motions.

Geophysical Journal International 181, 1-80.

Drummond, M.S. and Defant, M.J. (1990) A model for trondhjemite‐tonalite‐dacite genesis

and crustal growth via slab melting: Archean to modern comparisons. Journal of

Geophysical Research: Solid Earth 95, 21503-21521.

Dzierma, Y., Thorwart, M.M., Rabbel, W., Flueh, E.R., Alvarado, G. and Mora, M. (2010)

Imaging crustal structure in south central Costa Rica with receiver functions.

Geochemistry, Geophysics, Geosystems 11.

Ely, J.C. and Neal, C.R. (2003) Using platinum-group elements to investigate the origin of

the Ontong Java Plateau, SW Pacific. Chemical Geology 196, 235-257.

Erdmann, S., Martel, C., Pichavant, M. and Kushnir, A. (2014) Amphibole as an archivist of

magmatic crystallization conditions: problems, potential, and implications for inferring

magma storage prior to the paroxysmal 2010 eruption of Mount Merapi, Indonesia.

Contributions to Mineralogy and Petrology 167, 1016.

Farver, J.R. and Brabander, D.J. (2001) Magma ascent rates from mineral reaction rims and

extension to teaching about volcanic hazards. Journal of Geoscience Education 49, 140-

145.

Feigenson, M.D. and Carr, M.J. (1993) The source of Central American lavas: inferences

from geochemical inverse modeling. Contributions to Mineralogy and Petrology 113,

226-235.

Feigenson, M.D., Carr, M.J., Maharaj, S.V., Juliano, S. and Bolge, L.L. (2004) Lead isotope

composition of Central American volcanoes: Influence of the Galapagos plume.

Geochemistry, Geophysics, Geosystems 5.

Ford, C., Russell, D., Craven, J. and Fisk, M. (1983) Olivine-liquid equilibria: temperature,

pressure and composition dependence of the crystal/liquid cation partition coefficients

for Mg, Fe2+, Ca and Mn. Journal of Petrology 24, 256-266.

Fujimaki, H., Tatsumoto, M. and Aoki, K.i. (1984) Partition coefficients of Hf, Zr, and REE

between phenocrysts and groundmasses. Journal of Geophysical Research: Solid Earth

81

89.

Gale, A., Dalton, C.A., Langmuir, C.H., Su, Y. and Schilling, J.G. (2013) The mean

composition of ocean ridge basalts. Geochemistry, Geophysics, Geosystems 14, 489-

518.

Gazel, E., Carr, M.J., Hoernle, K., Feigenson, M.D., Szymanski, D., Hauff, F. and van den

Bogaard, P. (2009) Galapagos‐OIB signature in southern Central America: Mantle

refertilization by arc–hot spot interaction. Geochemistry, Geophysics, Geosystems 10.

Gazel, E., Hayes, J.L., Hoernle, K., Kelemen, P., Everson, E., Holbrook, W.S., Hauff, F.,

Van Den Bogaard, P., Vance, E.A. and Chu, S. (2015) Continental crust generated in

oceanic arcs. Nature Geoscience 8, 321.

Gazel, E., Hoernle, K., Carr, M.J., Herzberg, C., Saginor, I., Van den Bogaard, P., Hauff, F.,

Feigenson, M. and Swisher, C. (2011) Plume–subduction interaction in southern Central

America: Mantle upwelling and slab melting. Lithos 121, 117-134.

Goss, A. and Kay, S. (2006) Steep REE patterns and enriched Pb isotopes in southern

Central American arc magmas: Evidence for forearc subduction erosion? Geochemistry,

Geophysics, Geosystems 7.

Goss, A. and Kay, S. (2009) Extreme high field strength element (HFSE) depletion and

near-chondritic Nb/Ta ratios in Central Andean adakite-like lavas (~ 28 S,~ 68 W).

Earth and Planetary Science Letters 279, 97-109.

Green, T., Blundy, J., Adam, J. and Yaxley, G. (2000) SIMS determination of trace element

partition coefficients between garnet, clinopyroxene and hydrous basaltic liquids at 2–

7.5 GPa and 1080–1200 C. Lithos 53, 165-187.

Grove, T.L., Elkins-Tanton, L.T., Parman, S.W., Chatterjee, N., Müntener, O. and Gaetani,

G.A. (2003) Fractional crystallization and mantle-melting controls on calc-alkaline

differentiation trends. Contributions to Mineralogy and Petrology 145, 515-533.

Hammarstrom, J.M. and Zen, E.-a. (1986) Aluminum in hornblende: an empirical igneous

geobarometer. American Mineralogist 71, 1297-1313.

Harpp, K.S., Wanless, V.D., Otto, R.H., Hoernle, K. and Werner, R. (2004) The Cocos and

Carnegie aseismic ridges: A trace element record of long-term plume–spreading center

interaction. Journal of Petrology 46, 109-133.

Haschke, M., Günther, A., Melnick, D., Echtler, H., Reutter, K.-J., Scheuber, E. and Oncken,

O. (2006) Central and southern Andean tectonic evolution inferred from arc magmatism,

The Andes. Springer, pp. 337-353.

Hayes, J.L., Holbrook, W.S., Lizarralde, D., Avendonk, H.J., Bullock, A.D., Mora, M.,

Harder, S., Alvarado, G.E. and Ramírez, C. (2013) Crustal structure across the Costa

82

Rican volcanic arc. Geochemistry, Geophysics, Geosystems 14, 1087-1103.

He, Y., Wu, H., Ke, S., Liu, S.-A. and Wang, Q. (2017) Iron isotopic compositions of

adakitic and non-adakitic granitic magmas: Magma compositional control and subtle

residual garnet effect. Geochimica et Cosmochimica Acta 203, 89-102.

Hebert, L.B., Antoshechkina, P., Asimow, P. and Gurnis, M. (2009) Emergence of a low-

viscosity channel in subduction zones through the coupling of mantle flow and

thermodynamics. Earth and Planetary Science Letters 278, 243-256.

Hermann, J. and Rubatto, D. (2009) Accessory phase control on the trace element signature

of sediment melts in subduction zones. Chemical Geology 265, 512-526.

Hidalgo, P.J. and Rooney, T.O. (2010) Crystal fractionation processes at Baru volcano from

the deep to shallow crust. Geochemistry, Geophysics, Geosystems 11.

Hidalgo, P.J. and Rooney, T.O. (2014) Petrogenesis of a voluminous Quaternary adakitic

volcano: the case of Baru volcano. Contributions to Mineralogy and Petrology 168,

1011.

Hidalgo, P.J., Vogel, T.A., Rooney, T.O., Currier, R.M. and Layer, P.W. (2011) Origin of

silicic volcanism in the Panamanian arc: evidence for a two-stage fractionation process

at El Valle volcano. Contributions to Mineralogy and Petrology 162, 1115-1138.

Hildreth, W. and Moorbath, S. (1988) Crustal contributions to arc magmatism in the Andes

of central Chile. Contributions to mineralogy and petrology 98, 455-489.

Hoernle, K., Abt, D.L., Fischer, K.M., Nichols, H., Hauff, F., Abers, G.A., Van Den

Bogaard, P., Heydolph, K., Alvarado, G. and Protti, M. (2008) Arc-parallel flow in the

mantle wedge beneath Costa Rica and Nicaragua. Nature 451, 1094.

Hoernle, K., Hauff, F. and van den Bogaard, P. (2004) 70 my history (139–69 Ma) for the

Caribbean large igneous province. Geology 32, 697-700.

Hoernle, K., Werner, R., Morgan, J.P., Garbe-Schonberg, D., Bryce, J. and Mrazek, J.

(2000) Existence of complex spatial zonation in the Galápagos plume. Geology 28,

435-438.

Hollister, L.S., Grissom, G., Peters, E., Stowell, H. and Sisson, V. (1987) Confirmation of

the empirical correlation of Al in hornblende with pressure of solidification of calc-

alkaline plutons. American mineralogist 72, 231-239.

Hubberten, H., Emmermann, R. and Puchelt, H. (1983) GEOCHEMISTRY OF BASALTS

FROM COSTA-RICA RIFT SITE-504 AND SITE-505 (DEEP-SEA DRILLING

PROJECT LEG-69 AND LEG-70). INITIAL REPORTS OF THE DEEP SEA

DRILLING PROJECT 69, 791-803.

Husen, S., Quintero, R., Kissling, E. and Hacker, B. (2003) Subduction-zone structure and

83

magmatic processes beneath Costa Rica constrained by local earthquake tomography

and petrological modelling. Geophysical Journal International 155, 11-32.

Ireland, T.J., Walker, R.J. and Garcia, M.O. (2009) Highly siderophile element and 187 Os

isotope systematics of Hawaiian picrites: implications for parental melt composition

and source heterogeneity. Chemical Geology 260, 112-128.

Jochum, K.P., Brueckner, S.M., Nohl, U., Stoll, B. and Weis, U. (2009) Geostandards and

Geoanalytical Research bibliographic review 2008. Geostandards and Geoanalytical

Research 33, 501-505.

Jochum, K.P., Weis, U., Schwager, B., Stoll, B., Wilson, S.A., Haug, G.H., Andreae, M.O.

and Enzweiler, J. (2016) Reference values following ISO guidelines for frequently

requested rock reference materials. Geostandards and Geoanalytical Research 40, 333-

350.

Jochum, K.P., Weis, U., Stoll, B., Kuzmin, D., Yang, Q., Raczek, I., Jacob, D.E., Stracke, A.,

Birbaum, K. and Frick, D.A. (2011) Determination of reference values for NIST SRM

610–617 glasses following ISO guidelines. Geostandards and Geoanalytical Research

35, 397-429.

Johnson, M.C. and Rutherford, M.J. (1989) Experimental calibration of the aluminum-in-

hornblende geobarometer with application to Long Valley caldera (California) volcanic

rocks. Geology 17, 837-841.

Kay, R. (1978) Aleutian magnesian andesites: melts from subducted Pacific Ocean crust.

Journal of Volcanology and Geothermal Research 4, 117-132.

Kay, R.W. and Kay, S.M. (1993) Delamination and delamination magmatism.

Tectonophysics 219, 177-189.

Kay, R.W. and Kay, S.M. (2002) Andean adakites: three ways to make them. Acta

Petrologica Sinica 18, 303-311.

Kay, S.M., Coira, B. and Viramonte, J. (1994) Young mafic back arc volcanic rocks as

indicators of continental lithospheric delamination beneath the Argentine Puna plateau,

central Andes. Journal of Geophysical Research: Solid Earth 99, 24323-24339.

Kay, S.M., Coira, B.L., Caffe, P.J. and Chen, C.-H. (2010) Regional chemical diversity,

crustal and mantle sources and evolution of central Andean Puna plateau ignimbrites.

Journal of Volcanology and Geothermal Research 198, 81-111.

Kelemen, P.B., Yogodzinski, G.M. and Scholl, D.W. (2004) Along‐strike variation in the

Aleutian island arc: Genesis of high Mg# andesite and implications for continental crust.

Inside the subduction factory 138, 223-276.

Kepezhinskas, P., Defant, M.J. and Widom, E. (2002) Abundance and distribution of PGE

84

and Au in the island-arc mantle: implications for sub-arc metasomatism. Lithos 60, 113-

128.

Kiseeva, E.S., Fonseca, R.O. and Smythe, D.J. (2017) Chalcophile elements and sulfides in

the upper mantle. Elements 13, 111-116.

Kobayashi, D., LaFemina, P., Geirsson, H., Chichaco, E., Abrego, A.A., Mora, H. and

Camacho, E. (2014) Kinematics of the western Caribbean: Collision of the Cocos Ridge

and upper plate deformation. Geochemistry, Geophysics, Geosystems 15, 1671-1683.

Kress, V.C. and Carmichael, I.S. (1991) The compressibility of silicate liquids containing

Fe 2 O 3 and the effect of composition, temperature, oxygen fugacity and pressure on

their redox states. Contributions to Mineralogy and Petrology 108, 82-92.

Kuşcu, G.G. and Floyd, P.A. (2001) Mineral compositional and textural evidence for

magma mingling in the Saraykent volcanics. Lithos 56, 207-230.

Lee, M.J., Lee, J.I., Kwon, S.-T., Choo, M.K., Jeong, K.-S., Cho, J.-H. and Kim, S.-R.

(2011) Sr-Nd-Pb isotopic compositions of submarine alkali basalts recovered from the

South Korea Plateau, East Sea. Geosciences Journal 15, 149.

Leeman, W.P., Carr, M.J. and Morris, J.D. (1994) Boron geochemistry of the Central

American volcanic arc: constraints on the genesis of subduction-related magmas.

Geochimica et Cosmochimica Acta 58, 149-168.

Loewen, M.W., Duncan, R.A., Kent, A.J. and Krawl, K. (2013) Prolonged plume volcanism

in the Caribbean Large Igneous Province: new insights from Curaçao and Haiti.

Geochemistry, Geophysics, Geosystems 14, 4241-4259.

Lonsdale, P. (2005) Creation of the Cocos and Nazca plates by fission of the Farallon plate.

Tectonophysics 404, 237-264.

Lonsdale, P. and Klitgord, K.D. (1978) Structure and tectonic history of the eastern Panama

Basin. Geological Society of America Bulletin 89, 981-999.

Lorand, J.-P. and Luguet, A. (2016) Chalcophile and siderophile elements in mantle rocks:

Trace elements controlled by trace minerals. Reviews in Mineralogy and Geochemistry

81, 441-488.

MacKenzie, L., Abers, G.A., Fischer, K.M., Syracuse, E.M., Protti, J.M., Gonzalez, V. and

Strauch, W. (2008) Crustal structure along the southern Central American.

MacMillan, I., Gans, P.B. and Alvarado, G. (2004) Middle Miocene to present plate tectonic

history of the southern Central American Volcanic Arc. Tectonophysics 392, 325-348.

Macpherson, C.G., Dreher, S.T. and Thirlwall, M.F. (2006) Adakites without slab melting:

high pressure differentiation of island arc magma, Mindanao, the Philippines. Earth and

Planetary Science Letters 243, 581-593.

85

Mahood, G. and Hildreth, W. (1983) Large partition coefficients for trace elements in high-

silica rhyolites. Geochimica et Cosmochimica Acta 47, 11-30.

Mann, P. (2007) Overview of the tectonic history of northern Central America. Geological

Society of America Special Papers 428, 1-19.

Martin, H., Smithies, R., Rapp, R., Moyen, J.-F. and Champion, D. (2005) An overview of

adakite, tonalite–trondhjemite–granodiorite (TTG), and sanukitoid: relationships and

some implications for crustal evolution. Lithos 79, 1-24.

McDonough, W.F. and Sun, S.-S. (1995) The composition of the Earth. Chemical geology

120, 223-253.

Meschede, M. and Barckhausen, U. (2001) The relationship of the Cocos and Carnegie

ridges: age constraints from paleogeographic reconstructions. International Journal of

Earth Sciences 90, 386-392.

Molnar, P. and Sykes, L.R. (1969) Tectonics of the Caribbean and Middle America regions

from focal mechanisms and seismicity. Geological Society of America Bulletin 80,

1639-1684.

Morell, K.D. (2015) Late Miocene to recent plate tectonic history of the southern Central

America convergent margin. Geochemistry, Geophysics, Geosystems 16, 3362-3382.

Morimoto, N. (1988) Nomenclature of pyroxenes. Mineralogy and Petrology 39, 55-76.

Müntener, O., Kelemen, P.B. and Grove, T.L. (2001) The role of H 2 O during

crystallization of primitive arc magmas under uppermost mantle conditions and genesis

of igneous pyroxenites: an experimental study. Contributions to Mineralogy and

Petrology 141, 643-658.

Nandedkar, R.H., Hürlimann, N., Ulmer, P. and Müntener, O. (2016) Amphibole–melt trace

element partitioning of fractionating calc-alkaline magmas in the lower crust: an

experimental study. Contributions to Mineralogy and Petrology 171, 71.

Nash, W. and Crecraft, H. (1985) Partition coefficients for trace elements in silicic magmas.

Geochimica et Cosmochimica Acta 49, 2309-2322.

O’Driscoll, B. and González-Jiménez, J.M. (2016) Petrogenesis of the platinum-group

minerals. Reviews in Mineralogy and Geochemistry 81, 489-578.

Park, J.-W., Campbell, I.H. and Arculus, R.J. (2013) Platinum-alloy and sulfur saturation in

an arc-related basalt to rhyolite suite: Evidence from the Pual Ridge lavas, the Eastern

Manus Basin. Geochimica et Cosmochimica Acta 101, 76-95.

Park, J.-W., Campbell, I.H. and Eggins, S.M. (2012) Enrichment of Rh, Ru, Ir and Os in Cr

spinels from oxidized magmas: evidence from the Ambae volcano, Vanuatu.

Geochimica et Cosmochimica Acta 78, 28-50.

86

Park, J.-W., Campbell, I.H., Kim, J. and Moon, J.-W. (2015) The role of late sulfide

saturation in the formation of a Cu-and Au-rich magma: Insights from the platinum

group element geochemistry of Niuatahi–Motutahi lavas, Tonga Rear Arc. Journal of

Petrology 56, 59-81.

Patino, L.C., Carr, M.J. and Feigenson, M.D. (2000) Local and regional variations in

Central American arc lavas controlled by variations in subducted sediment input.

Contributions to Mineralogy and Petrology 138, 265-283.

Peach, C., Mathez, E. and Keays, R. (1990) Sulfide melt-silicate melt distribution

coefficients for noble metals and other chalcophile elements as deduced from MORB:

Implications for partial melting. Geochimica et Cosmochimica Acta 54, 3379-3389.

Pearce, J.A. and Norry, M.J. (1979) Petrogenetic implications of Ti, Zr, Y, and Nb

variations in volcanic rocks. Contributions to mineralogy and petrology 69, 33-47.

Pearce, J.A. and Stern, R.J. (2006) Origin of back‐arc basin magmas: Trace element and

isotope perspectives. Back-Arc Spreading Systems: Geological, Biological, Chemical,

and Physical Interactions, 63-86.

Petford, N. and Atherton, M. (1996) Na-rich partial melts from newly underplated basaltic

crust: the Cordillera Blanca Batholith, Peru. Journal of petrology 37, 1491-1521.

Pitcher, L., Helz, R.T., Walker, R.J. and Piccoli, P. (2009) Fractionation of the platinum-

group elments and Re during crystallization of basalt in Kilauea Iki Lava Lake, Hawaii.

Chemical Geology 260, 196-210.

Plank, T. and Langmuir, C.H. (1993) Tracing trace elements from sediment input to

volcanic output at subduction zones. Nature 362, 739.

Profeta, L., Ducea, M.N., Chapman, J.B., Paterson, S.R., Gonzales, S.M.H., Kirsch, M.,

Petrescu, L. and DeCelles, P.G. (2015) Quantifying crustal thickness over time in

magmatic arcs. Scientific Reports 5, 17786.

Protti, M., Gu, F. and McNally, K. (1994) The geometry of the Wadati-Benioff zone under

southern Central America and its tectonic significance: Results from a high-resolution

local seismographic network. Physics of the Earth and Planetary Interiors 84, 271-287.

Prouteau, G., Scaillet, B., Pichavant, M. and Maury, R. (1999) Fluid-present melting of

ocean crust in subduction zones. Geology 27, 1111-1114.

Putirka, K. (2016) Amphibole thermometers and barometers for igneous systems and some

implications for eruption mechanisms of felsic magmas at arc volcanoes. American

Mineralogist 101, 841-858.

Putirka, K.D. (2008) Thermometers and barometers for volcanic systems. Reviews in

mineralogy and geochemistry 69, 61-120.

87

Putirka, K.D., Mikaelian, H., Ryerson, F. and Shaw, H. (2003) New clinopyroxene-liquid

thermobarometers for mafic, evolved, and volatile-bearing lava compositions, with

applications to lavas from Tibet and the Snake River Plain, Idaho. American

Mineralogist 88, 1542-1554.

Rapp, R., Shimizu, N., Norman, M. and Applegate, G. (1999) Reaction between slab-

derived melts and peridotite in the mantle wedge: experimental constraints at 3.8 GPa.

Chemical Geology 160, 335-356.

Rapp, R.P. and Watson, E.B. (1995) Dehydration melting of metabasalt at 8–32 kbar:

implications for continental growth and crust-mantle recycling. Journal of Petrology 36,

891-931.

Rehkämper, M. and Halliday, A. (1997) Development and application of new

iondashexchange techniques for the separation of the platinum group and other

siderophile elements from geological samples. Talanta 44, 663-672.

Ribeiro, J.M., Maury, R.C. and Grégoire, M. (2016) Are Adakites slab melts or high-

pressure fractionated mantle melts? Journal of Petrology 57, 839-862.

Richards, J.P. (2011) High Sr/Y arc magmas and porphyry Cu±Mo±Au deposits: just add

water. Economic Geology 106, 1075-1081.

Ridolfi, F. and Renzulli, A. (2012) Calcic amphiboles in calc-alkaline and alkaline magmas:

thermobarometric and chemometric empirical equations valid up to 1,130° C and 2.2

GPa. Contributions to Mineralogy and Petrology 163, 877-895.

Ridolfi, F., Renzulli, A. and Puerini, M. (2010) Stability and chemical equilibrium of

amphibole in calc-alkaline magmas: an overview, new thermobarometric formulations

and application to subduction-related volcanoes. Contributions to Mineralogy and

Petrology 160, 45-66.

Rodríguez, C., Sellés, D., Dungan, M., Langmuir, C. and Leeman, W. (2007) Adakitic

dacites formed by intracrustal crystal fractionation of water-rich parent magmas at

Nevado de Longaví volcano (36· 2 S; Andean Southern Volcanic Zone, Central Chile).

Journal of Petrology 48, 2033-2061.

Rooney, T.O., Franceschi, P. and Hall, C.M. (2011) Water-saturated magmas in the Panama

Canal region: a precursor to adakite-like magma generation? Contributions to

Mineralogy and Petrology 161, 373-388.

Rooney, T.O., Morell, K.D., Hidalgo, P. and Fraceschi, P. (2015) Magmatic consequences

of the transition from orthogonal to oblique subduction in Panama. Geochemistry,

Geophysics, Geosystems 16, 4178-4208.

Rutherford, M.J. and Devine, J.D. (2003) Magmatic conditions and magma ascent as

88

indicated by hornblende phase equilibria and reactions in the 1995–2002 Soufriere Hills

magma. Journal of Petrology 44, 1433-1453.

Saginor, I., Gazel, E., Condie, C. and Carr, M.J. (2013) Evolution of geochemical variations

along the Central American volcanic front. Geochemistry, Geophysics, Geosystems 14,

4504-4522.

Sajona, F.G., Maury, R.C., Prouteau, G., Cotten, J., Schiano, P., Bellon, H. and Fontaine, L.

(2000) Slab melt as metasomatic agent in island arc magma mantle sources, Negros and

Batan (Philippines). Island Arc 9, 472-486.

Sallarès, V., Dañobeitia, J.J. and Flueh, E.R. (2001) Lithospheric structure of the Costa

Rican Isthmus: Effects of subduction zone magmatism on an oceanic plateau. Journal of

Geophysical Research: Solid Earth 106, 621-643.

Samaniego, P., Robin, C., Chazot, G., Bourdon, E. and Cotten, J. (2010) Evolving

metasomatic agent in the Northern Andean subduction zone, deduced from magma

composition of the long-lived Pichincha volcanic complex (Ecuador). Contributions to

Mineralogy and Petrology 160, 239-260.

Scaillet, B. and Evans, B.W. (1999) The 15 June 1991 eruption of Mount Pinatubo. I. Phase

equilibria and pre-eruption P–T–f O2–f H2O conditions of the dacite magma. Journal of

Petrology 40, 381-411.

Schmidt, M.W. (1992) Amphibole composition in tonalite as a function of pressure: an

experimental calibration of the Al-in-hornblende barometer. Contributions to

mineralogy and petrology 110, 304-310.

Sen, C. and Dunn, T. (1994) Dehydration melting of a basaltic composition amphibolite at

1.5 and 2.0 GPa: implications for the origin of adakites. Contributions to Mineralogy

and Petrology 117, 394-409.

Shimizu, H. (1989) REE, Ba, and Sr abundances and Sr, Nd, and Ce isotopic ratios in Hole

504B basalts, ODP Leg 111, Costa Rica Rift, Proc. ODP, Sci. Results. Ocean Drilling

Program, pp. 401-413.

Singer, B.S., Jicha, B.R., Leeman, W.P., Rogers, N.W., Thirlwall, M.F., Ryan, J. and

Nicolaysen, K.E. (2007) Along‐strike trace element and isotopic variation in Aleutian

Island arc basalt: Subduction melts sediments and dehydrates serpentine. Journal of

Geophysical Research: Solid Earth 112.

Sisson, T. and Grove, T. (1993) Experimental investigations of the role of H 2 O in calc-

alkaline differentiation and subduction zone magmatism. Contributions to mineralogy

and petrology 113, 143-166.

Smith, D.J. (2014) Clinopyroxene precursors to amphibole sponge in arc crust. Nature

89

Communications 5, 4329.

Spear, F.S. (1981) Amphibole-plagioclase equilibria: an empirical model for the relation

albite+ tremolite= edenite+ 4 quartz. Contributions to Mineralogy and Petrology 77,

355-364.

Sun, S.-S. and McDonough, W.-s. (1989) Chemical and isotopic systematics of oceanic

basalts: implications for mantle composition and processes. Geological Society, London,

Special Publications 42, 313-345.

Syracuse, E.M. and Abers, G.A. (2006) Global compilation of variations in slab depth

beneath arc volcanoes and implications. Geochemistry, Geophysics, Geosystems 7.

Tindle, A. and Webb, P. (1994) PROBE-AMPH—a spreadsheet program to classify

microprobe-derived amphibole analyses. Computers & Geosciences 20, 1201-1228.

Tomascak, P.B., Ryan, J.G. and Defant, M.J. (2000) Lithium isotope evidence for light

element decoupling in the Panama subarc mantle. Geology 28, 507-510.

Wang, Q., Wyman, D.A., Xu, J.-F., Zhao, Z.-H., Jian, P. and Zi, F. (2007) Partial melting of

thickened or delaminated lower crust in the middle of eastern China: Implications for

Cu-Au mineralization. The Journal of Geology 115, 149-161.

Wang, Q., Xu, J.-F., Zhao, Z.-H., Bao, Z.-W., Xu, W. and Xiong, X.-L. (2004) Cretaceous

high-potassium intrusive rocks in the Yueshan-Hongzhen area of east China: Adakites

in an extensional tectonic regime within a continent. Geochemical Journal 38, 417-434.

Wegner, W., Wörner, G., Harmon, R.S. and Jicha, B.R. (2011) Magmatic history and

evolution of the Central American Land Bridge in Panama since Cretaceous times.

Bulletin 123, 703-724.

Whattam, S.A. and Stern, R.J. (2016) Arc magmatic evolution and the construction of

continental crust at the Central American Volcanic Arc system. International Geology

Review 58, 653-686.

White, W.M., McBirney, A.R. and Duncan, R.A. (1993) Petrology and geochemistry of the

Galápagos Islands: Portrait of a pathological mantle plume. Journal of Geophysical

Research: Solid Earth 98, 19533-19563.

Woodland, S., Pearson, D. and Thirlwall, M. (2002) A platinum group element and Re–Os

isotope investigation of siderophile element recycling in subduction zones: comparison

of Grenada, Lesser Antilles Arc, and the Izu–Bonin Arc. Journal of Petrology 43, 171-

198.

Yoshikawa, M., Tamura, A., Arai, S., Kawamoto, T., Payot, B.D., Rivera, D.J., Bariso, E.B.,

Mirabueno, M.H.T., Okuno, M. and Kobayashi, T. (2016) Aqueous fluids and

sedimentary melts as agents for mantle wedge metasomatism, as inferred from

90

peridotite xenoliths at Pinatubo and Iraya volcanoes, Luzon arc, Philippines. Lithos 262,

355-368.

Zhang, J., Humphreys, M.C., Cooper, G.F., Davidson, J.P. and Macpherson, C.G. (2017)

Magma mush chemistry at subduction zones, revealed by new melt major element

inversion from calcic amphiboles. American Mineralogist: Journal of Earth and

Planetary Materials 102, 1353-1367.

91

국 문 초 록

중앙아메리카 섭입대 서파나마 지역의 플레이스토세 아다카이트질

마그마는 높은 Sr(>926 ppm), Sr/Y(>50), La/Yb(≥20)과 낮은 Y(<15

ppm)과 Yb(<1.5 ppm), 87Sr/86Sr (<0.704)의 아다카이트질 암석의

지화학적 특징을 보이며, 아다카이트 분류 도표(Castillo 2012)에서

아다카이트 영역에 도시된다. 이전 연구에서 이 아다카이트질 마그마는

섭입하는 슬랩의 직접적인 용융 작용 및 맨틀과의 반응으로

형성되었다고 주장하였으나 최근 연구는 맨틀에서 기인한 현무암질

마그마의 고압(≥30 km 깊이) 분별결정작용에 의해 형성되었다는 그

형성과정에 대해 이견이 있다.

본 연구는 코코스 해령이 섭입하는 서파나마 지역의 플라이스토세

아다카이트질 마그마의 성인을 규명하기 위해 기존에 연구된 시료 중

대표적인 시료에 대해 추가적으로 광물 분석과 미량원소 및 동위원소

분석을 통한 암석지구화학적 연구를 수행하였다.

원시맨틀 값(McDonough and Sun, 1995)과 비교한 미량원소

패턴에서 경희토류 원소가 중희토류 원소에 비하여 비교적 부화되어

있으며 이전 연구에 의해 부화된 슬랩 성분의 영향을 받았다는 것이

제안되었다. 서파나마의 아다카이트질 암석의 Pb 및 Nd 동위원소비는

코스타리카 화산호 현무암(<5.3 Ma; Gazel et al., 2009)과 파나마

지역을 섭입하는 슬랩인 코코스 해령 현무암과 유사한 값을 보이지만,

Sr 동위원소비는 코코스해령 현무암보다 큰 값을 보인다.

마그마의 정치 심도 및 결정화 조건을 추정하기 위해 각섬석의

성분을 이용하였으며 압력(~ 220-960 MPa), 온도(~ 721-1009 °C),

산소 퓨가시티(~ NNO - 0.1-NNO + 2.2), 평형용융물의 성분 (0.31-

0.56 melt Mg# and 58.4-80.8 wt.% melt SiO2)을 계산하였다. 이

결과는 서파나마 지역의 마그마 기원은 불균질한 결정 부유대(crystal

mush zone)일 가능성을 시사하며 약 35km 와 15km에서 정치한

것으로 사료된다. 현무암질 암석의 단사휘석과 사방휘석의 결정화

압력은 600-1300 MPa 이며 온도는 1020-1170 °C 이고 이 결과는

초기 마그마가 중앙아메리카 섭입대의 하부지각 또는 하부지각-맨틀

경계에서 정치하고 분화하였음을 지시한다.

고압 각섬석(≥920 MPa)과 평형상태의 용융물의 성분은 비교적

92

넓은 Mg# (0.31-0.56)와 SiO2 (58.4-71.3 wt.%) 범위를 가지며 이

결과는 서파나마 화산호 아래 하부 지각에 위치한 마그마 저장소(결정

부유대)에서 상당한 마그마 분화 작용이 있었다는 것을 지시한다.

고철질암석과 규장질암석의 단사휘석과 평형상태의 용융물은 0.41-

0.68의 Mg# 범위를 가지고 이 범위는 본 연구의 현무암과 현무암질

안산암의 Mg# 범위(0.51-0.66)와 일치하며 석영안산암의 휘석 반정은

하부 지각에 정치한 동원성 현무암질 마그마 기원이였음을 지시한다. 본

연구의 암석학적, 광물화학적 그리고 전암데이터 결과는 중앙아메리카

섭입대에 서파나마와 남코스타리카에서 고압(~35-40 km 깊이)에서

현무암부터 석영안산암까지 광범위한 마그마 진화 작용이 있었음을

암시한다.

이전 연구에서 제안된 중앙아메리카 섭입대의 아다카이트질

마그마의 형성 기원 중 어떤 과정을 통해 이 지역의 아다카이트질

마그마 성분을 형성시킬 수 있는지 규명하기 위해 Sr/Y 대 Y 과 La/Yb

대 Yb 변화도를 이용하여 미량원소 및 동위원소 모델링을 실시하였다.

그 결과 섭입하는 코코스 해령으로부터 기인한 슬랩 용융체에 의해 교대

작용을 겪은 맨틀에서 기인한 현무암질 마그마가 석류석이 안정한

하부지각에서 석류석과 각섬석이 우세한 분별결정작용을 거칠 때

서파나마와 남코스타리카의 아다카이트질 암석이 형성될 수 있음을

보여준다. 고압 각섬석과 평형상태의 고압평형용융물의 아다카이트질

지화학적 특성과 사파나마와 남코스타리카의 아다카이트질 마그마의

석류석 및 각섬석의 동결정화(co-crystallization)를 지시하는

미량원소비가 이를 뒷받침하는 증거이다.

본 연구의 시료 중 분화된 시료에서 관찰되는 각섬석이 우세한

집적암 조직, 반정에서 관찰되는 비평형(disequilibrium) 조직과 이를

지시하는 성분은 마그마 혼합 모델과 저압 분별결정작용 모델의 결과와

일치한다. 이 결과는 마그마 혼합 과정과 저압 분별결정작용이

서파나마와 남코스타리카의 플라이오세 암석의 아다카이트질 특성을

형성시키는데 영향을 끼쳤음을 지시한다.

주요어: 아다카이트질 암석, 서파나마, 각섬석, 휘석, 슬랩 용융물, 맨틀

기원 용융물, 마그마 혼합 작용, 분별결정화작용

학 번: 2016-20429

93

ACKNOWLEDGEMENT

I am sincerely indebted to many individuals and groups in completing the work

presented here successfully. First, I wish to express my hearty thanks to my

supervisor, Professor Jung-Woo Park, for his support, guidance, teaching, advice

and encouragement throughout the completion of this work as well as my graduate

career. I could learn a lot not only about the geological science but also the attitude

of mind toward living to become a researcher.

I would like to deeply thank to Prof. Sang-Mook Lee (Seoul National

University) and Dr. Mi Jung Lee (Korea Polar Research Institute) for their critical

reviews and advice on this work.

My sincere thanks are also extended to Prof. Kaj Hoernle (GEOMAR) and Prof.

Esteban Gazel (Cornell University) for providing the rock samples and advices on

this work.

I would like to express my thanks to Seunghee Han (Korea Polar Research

Institute) for her assistance and teaching of sample preparation and analytical

assistance in the use of LA-ICP-MS and ICP-MS. I wish to thank Jihye Oh (Korea

Institute of Ocean Science & Technology) for her assistance and teaching of

sample preparation and analytical assistance in the use of ICP-MS. I am grateful to

Seung-Ah Ha, Jongmin Beak, Dr. Sun Young Park and Jung Jin Moon, Sangbum

Park and Yun Seok Yang (Korea Polar Research Institute) for their kind assistance

in sample preparation and the use of TIMS and EPMA. I would like to thank Dr.

Changkun Park (Korea Polar Research Institute) for his comments and advice on

the use of EPMA. I wish to thank Hyun-jae Maeng (ASP) for his advice on setting

up cleaning system and clean facilities in Marine Petrology and Geochemistry

Laboratory at Seoul National University.

I am sincerely grateful to Prof. Jaehyung Yu (Chungnam National University)

for his advice and giving me an opening to study Geology and do research. I also

deeply thank to Dr. Jiwhan Ahn (Korea Institute of Geoscience and Mineral

Resources) for her advice and encouragement to decide to study Geology for a

Master’s. My hearty thanks go to Prof. Sung Hi Choi for her advice and leading me

to have a dream of a Geochemist.

94

I will never forget my memories of members of the Marine Petrology and

Geochemistry Laboratory, Jihyuk Kim, Sunki Choi, Jiwon Hwang, Junhee Park,

Gyuseung Park, Sarang Choi, Choi Seongjun, Sungwhan Lim and Gyuchan Park

for their help, concern and helpful advice. My thanks also go to Inkeong Moon,

Yuri Lee and the colleagues of the laboratory of Marine Drug Discovery (Seoul

National University) for their help and kindness. My thanks also go to Stacy

Murdock, Flora Ukoli, Mihyun Lee, Mi Kyung Jeong, Sohee Kim, Hanna Park,

Eunsol Lee, Jin A Yang and Gyeong-ja Park for their concern and encouragement

throughout my school days.

I would like to express my hearty thanks to my parents for their endless love,

trust, encouragement and devotion. I am deeply grateful to my sisters, Sungjin

Jegal and Dajeong Jegal, for their endless love. I also deeply thank to my

grandmothers, Sun Ja Yang, Hoyeong Lee, Mi Ja Yang, and my late grandfathers,

Jong Gwon Kim, Bok Jegal for their support and endless love. I sincerely thank to

my parents-in-law and their family for their love and support. Finally, I heartily

thank to my husband, Yonghwi Kim, for his warm-hearted love, help, hearty advice

and support, with my love.