33. GEOCHEMISTRY OF IGNEOUS ROCKS RECOVERED FROM A TRANSECT ACROSS THEMARIANA TROUGH, ARC, FORE-ARC, AND TRENCH, SITES 453 THROUGH 461,

DEEP SEA DRILLING PROJECT LEG 601

D. A. Wood,2 N. G. Marsh,2 J. Tarney,2 J.-L. Joron,3 P. Fryer,4 and M. Treuil3

ABSTRACT

Major and trace element analyses are presented for 110 samples from the DSDP Leg 60 basement cores drilled alonga transect across the Mariana Trough, arc, fore-arc, and Trench at about 18°N. The igneous rocks forming breccias atSite 453 in the west Mariana Trough include plutonic cumulates and basalts with calc-alkaline affinities. Basaltsrecovered from Sites 454 and 456 in the Mariana Trough include types with compositions similar to normal MORB andtypes with calc-alkaline affinities within a single hole. At Site 454 the basalts show a complete compositional transitionbetween normal MORB and calc-alkaline basalts. These basalts may be the result of mixing of the two magma types insmall sub-crustal magma reservoirs or assimilation of calc-alkaline, arc-derived vitric tuffs by normal MORB magmasduring eruption or intrusion.

A basaltic andesite clast in the breccia recovered from Site 457 on the active Mariana arc and samples dredged froma seamount in the Mariana arc are calc-alkaline and similar in composition to the basalts recovered from the MarianaTrough and West Mariana Ridge. Primitive island arc tholeiites were recovered from all four sites (Sites 458-461)drilled on the fore-arc and arc-side wall of the trench. These basalts form a coherent compositional group distinct fromthe Mariana arc, West Mariana arc, and Mariana Trough calc-alkaline lavas, indicating temporal (and perhapsspatial?) chemical variations in the arc magmas erupted along the transect.

Much of the 209 meters of basement cored at Site 458 consists of endiopside- and bronzite-bearing, Mg-richandesites with compositions related to boninites. These andesites have the very low Ti, Zr, Ti/Zr, P, and rare-earth-element contents characteristic of boninites, although they are slightly light-rare-earth-depleted and have lower MgO,Cr, Ni, and higher CaO and AI2O3 contents than those reported for typical boninites. The large variations in chemistryobserved in the lavas recovered from this transect suggest that diverse mantle source compositions and complexpetrogenetic process are involved in forming crustal rocks at this intra-oceanic active plate margin.

INTRODUCTION

On DSDP Legs 59 and 60, 15 sites v/ere drilled alongan east-west transect at about 18°N from the WestPhilippine Basin to the Mariana Trench (Fig. 1) in orderto study the nature of the back-arc and inter-arc basins,the remnant and active arcs, and the fore-arc of theregion. One of the many aims of this drilling programwas to elucidate the processes involved in crustal genera-tion and the composition variation of active marginvolcanics in an intra-oceanic environment, free of thecomplications associated with arc systems close to con-tinental margins. On Leg 60, we recovered igneousrocks from the Mariana Trough, Sites 453, 454, and456; the Mariana Ridge, Site 457; the Mariana fore-arc,Sites 458 and 459; and the arc-side wall of the MarianaTrench, Sites 460 and 461.

Detailed geophysical investigations and dredgingoperations have established that most back-arc or inter-arc basins are of extensional origin and are floored bybasaltic crust, broadly comparable in structure and

Initial Reports of the Deep Sea Drilling Project, Volume 60.2 Department of Geological Sciences, University of Birmingham, P.O. Box 363, Bir-

mingham, B15 2TT, U.K. Dr. Wood's present address: Phillips Petroleum Co. Europe-Africa, Glen House, Stag Place, London SW1E 5DA, U.K. Dr. Marsh's present address:University of Leicester, Department of Geology, Leicester, LEI 7R4, U.K.

Laboratorie de Géochimie Comparee et Systematique, UER des Sciences de la Terre etInstitut de Physique du Globe, LA 196 CNRS, Université Pierre et Marie Curie, 4 PlaceJussieu, 75230 Paris Cedex 05, France, and Laboratoire Pierre Sue CNRS, CEN Saclay, B. P.No. 2, 91190 Gif-sur-Yvette, France.

4 Hawaii Institute of Geophysics, 2525 Collea Road, Honolulu, Hawaii 96822.

composition to those of the major ocean basins (Karig,1971; Hart et al., 1972; Barker, 1972; Hawkins, 1974;Gill, 1976). Recent studies have shown that some of thebasalts flooring back-arc basins are enriched in volatileand mobile elements (e.g., H2O

+ , K, Rb, Ba, Sr, U, andCs) relative to normal mid-ocean ridge basalts (N-typeMORB), and have some affinities with active marginmagma types (Gill, 1976; Tarney et al., 1977; Hawkes-worth et al., 1978; Saunders and Tarney, 1979; Weaveret al., 1979). In contrast, the basalts drilled in theShikoku, West Philippine, and Parece Vela Basins onLegs 58 and 59 show no unequivocal chemical affinitieswith active margin magma types (Marsh et al., 1980,Mattey et al., 1980; Wood et al., 1980a, 1980b). The ba-salts recovered from the Mariana Trough provide uswith an opportunity to compare the compositions ofmaterial from this relatively recently developed, narrowinter-arc basin with those of the broader, more matureback-arc basins of the West Pacific Ocean.

There is as yet no general consensus on the petro-genetic processes of active arc magma types. There aretwo main compositional end members of arc magma-tism: a tholeiitic type showing iron enrichment trendsand a calc-alkaline type showing enrichment in alka-lis, corresponding broadly to Kuno's (1968) pigeoniticand hypersthenic series, respectively. In addition, thereare very rare Mg-rich andesites or boninites, namedafter their type locality in the Bonin Islands (Kurodaand Shiraki, 1975) but also recently dredged from theMariana Trench near the island of Guam (Dietrich et

611

D. A. WOOD ET AL.

130 140° 150=

# Leg 60 Sites + Leg 58 SitesO Leg 59 Sites < •̂ Leg 31 Sites Ridges (all types)

Figure 1. Location of drill sites from Legs 59 and 60.

Trenches with related subductionProposed spreading centers(dashed where no well-defined rift)

612

GEOCHEMISTRY OF IGNEOUS ROCKS

al., 1978). These lavas are thought to be produced dur-ing the early stages of island arc development and aretherefore important in understanding arc magmatism.Similar Mg-rich andesites were drilled at Site 458. Theselavas, together with the primitive island arc tholeiitesand calc-alkaline basalts recovered from the Kyushu-Palau, West Mariana, Mariana Ridges and the Marianafore-arc and Trench during Legs 59 and 60, should helpto constrain petrogenetic models for this tectonic en-vironment.

REGIONAL AND TECTONIC SETTINGThe southern Philippine Sea is divided into three ma-

jor basins (Fig. 1) which are elongated in a north-southdirection. From west to east the basins are: the WestPhilippine Basin, the Parece Vela Basin, and the Mari-ana Trough. The basins are separated by the Kyushu-Palau and West Mariana Ridges, respectively. The east-ern boundary of the Southern Philippine Sea consists ofthe Mariana Arc and fore-arc, which overlie the Pacificplate presently being subducted at the Mariana Trench(Katsumata and Sykes, 1969).

The Mariana Trough is crescent-shaped, open at itssouthern end but closing at the junction between theMariana and West Mariana Ridges in the north. Thetrough is approximately 1500 km long and 250 kmacross at its widest section. The Mariana Ridge is ap-proximately 100 to 120 km across, and the fore-arc, orarc-trench gap, is approximately 150 to 180 km wide.

The crust flooring the Mariana Trough is character-ized by rough topography (Hart et al., 1972) and highheat flow (Sclater et al., 1972). It is underlain by a man-tle zone of high seismic wave attenuation (Barazangi etal., 1975). Poorly developed magnetic lineations havebeen reported for the Mariana Trough (Hussong, Uyeda,et al., 1978). The reasons for the poorly developedmagnetic lineations are suspected to be the rough base-ment topography and low geomagnetic latitude of theregion (Karig et al., 1978). On the basis; of the sedimentthickness, depth, the character of the acoustic base-ment, and the presence of a topographic axial high andmean heat flow, Karig (1971) suggested that the Mari-ana Trough is actively spreading, unlike the other basinsof the Philippine Sea. Estimates for the spreading ratevary from slow, 1 cm yr~1 (Hart et al., 1972) and 2 cmyr~1 (Hussong et al., 1978), to fast, and 4 cm yr~1

(Karig et al., 1978). The uppermost crust in the axialregion of the trough consists of basalts broadly similarto ocean floor tholeiites (Hart et al., 1972; Hawkins,1977). The opening of the Mariana Trough appears tobe a continuation of the processes involved with theopening of the Parece Vela and Shikoku Basins, to thewest, during the middle to late Tertiary. However, therelationship of these eastern basins and ridges to theWest Philippine Basin remains uncertain (Scott et al.,1980).

The Mariana arc can be divided into two sections: anorthern section which is presently active and underlainby a steeply dipping Benioff zone, with seismic activitydown to at least 600 km; and a southern section which ispresently inactive and underlain by a poorly defined

seismic zone less than 200 km deep (Katsumata andSykes, 1969; Isacks and Barazangi, 1977). The twozones overlap between 16° and 17°30'N, the ridge ofthe southern section lying to the trench side. The north-ern ridge carries 10 subaerial volcanic islands and pos-sibly several submarine volcanoes whose extrusive prod-ucts are dominated by quartz-normative basalts and ba-saltic andesites. The lack of iron enrichment and highBa and Sr contents in the basalts suggests that they havecalc-alkalic affinities (Meijer, 1976; Dixon and Batiza,1979). The volcanic islands of the southern ridge consistmainly of basalts and basaltic andesites interbeddedwith and overlain by, predominantly calcareous sedi-ments (Stark, 1963), although more evolved igneousrocks occur, for example, dacite on Saipan (Cloud etal., 1956; Schmidt, 1957; Taylor et al., 1969). Theselavas also lack a trend of strong iron enrichment andplot as a hypersthenic series on an AFM diagram (Stark,1963). In contrast to the Mariana arc and trough, thefore-arc has not been studied in detail. It consists ofcrust which is seismically oceanic in character (Mur-auchi et al., 1968). Sedimentation on the fore-arc ap-pears to be strongly controlled by tectonic activity,mainly block faulting in a tensional regime with sub-sidence since the early Tertiary. It is unlikely that slicesof the main Pacific ocean floor have been imbricatedinto the fore-arc of the Mariana system (Hussong et al.,1978).

The Mariana Trench has been dredged previouslyduring Cruise 17 of the RV Dmitry Mendeleev south ofthe Leg 60 transect, at approximately 12°N (Anony-mous, 1977). The dredges recovered flysch-like sedi-ments, gabbros, metagabbros, serpentinized and mylon-itized periodotites, boninites, basalts, and metabasalts(Dietrich et al, 1978; Sharaskin et al., 1980). The bonin-ites and flysch-like sediments suggest the presence ofarc-derived material in the trench.

ANALYTICAL METHODS

We have analyzed 110 samples from the Leg 60 cores at Birming-ham by X-ray fluorescence (XRF) for major and trace elements by themethod given in Tarney et al. (1978). However, TiO2, MgO, CaO, andNa2O were analyzed using an Rh anode tube rather than a Cr anodetube. Instrumental conditions for major element analysis using an Rhanode X-ray tube are given by Marsh et al. (1980). We have alsoanalyzed 63 samples from the Leg 60 sites at Saclay by instrumentalneutron activation (INA) for additional trace elements by the methodof Chayla et al. (1973).

RESULTSThe XRF data, together with CIPW norms and some

trace element ratios, are given in Tables 1 through 6, 8,and 10. Some of the INA data have been presented byBougault et al. (this volume), and only representativeanalyses of the main lithological units are given here(Table 11). The geochemical variations of the Leg 60sites will be dealt with in two parts: first, the MarianaTrough and Ridge sites (Sites 453, 454, 456, and 457);and secondly, the Mariana fore-arc and Trench sites(Sites 458 through 461). This will be followed by a brief,essentially qualitative, discussion of the implications ofthe geochemical data for the processes involved in the

613

D. A. WOOD ET AL.

petrogenesis of this diverse suite of active plate marginvolcanics.

The Mariana Trough and Ridge

Site 453(17°54.42'N, 143°40.95'E)

This site was located 120 km west of the centralgraben of the Mariana Trough in a water depth of 4693meters. The 150 meters of igneous and metamorphicpolymict breccias recovered are more than 5.4 m.y. oldand are predominantly gabbros with subordinate noritesand metabasalts cemented by a calcite, clay and/or ironoxide matrix. The breccia sequence can be split intothree units based on lithologic changes with depth in thehole. These are:

1) From 455.5 to 541 meters, sub-bottom depth(Cores 49 through 57), an igneous polymict breccia oflarge gabbroic clasts (> 10 cm) with a few diabase andbasalt clasts in a predominantly red-brown matrix. Thecore from Section 55-4 (Piece 4) through Section 57-2(Piece 16), however, has a green matrix and the gabbrosin it have been demagnetized as a result of hydrothermalalteration at elevated temperatures.

2) From 541 to 569.5 meters, sub-bottom depth(Cores 58 through 60), an igneous polymict breccia ofpredominantly metavolcanic clasts. Quartz veining iscommon, and pyrite is abundant in the matrix.

3) From 569.5 to 605 meters, sub-bottom depth(Cores 61 through 64), a breccia composed of sheared,mylonitized and serpentenized metagabbro clasts, withabundant pyrite set in a matrix of serpentine.

The analyzed samples cover virtually the full spec-trum of gabbroic clasts recovered, which include gab-bros, anorthositic gabbros, gabbro pegmatites, noriticgabbros, and hornblende gabbros. Hand-specimen de-scriptions of the analyzed clasts are given in the Appen-dix. Replacement of the original igneous mineral assem-blage of Plagioclase ± clinopyroxene ± orthopyroxene± olivine ± brown hornblende ± accessory magnetiteis common. Replacement minerals include green horn-blende, chlorite, epidote, sericite, carbonates, clays, andalkali feldspars with accessory amounts of zeolites,prehnite, pumpellyite, and stilpnomelane. The volcanicclasts recovered are mainly doleritic in texture andaphyric to sparsely Plagioclase phyric with rare smallvesicles (<0.6 mm). Alteration is moderate to heavywith clinopyroxene being replaced by chlorite, stilpno-melane, and iron oxides, and Plagioclase being replacedby sericite, albite, and clays.

The major and trace element chemistry of the plu-tonic rocks (Table 1) is consistent with many of thembeing cumulates (Figs. 2 and 3), which accounts for thevery low abundances of several of the incompatible orhygromagmatophile (HYG) trace elements (e.g., Zr andY) in some of the samples. The basaltic clasts vary fromaphyric to sparsely Plagioclase, and clinopyroxenephyric with rare pseudomorphs of olivine micropheno-crysts. The compositions of these basaltic clasts are typi-cal of arc magma types with high Sr and Ba but low Ta,Cr, and Ni contents. Of the two basalts analyzed by

INA (Table 10), Sample 453-52-1, 30-36 cm has a moreprimitive arc tholeiite Composition (lower Th and Laand a depletion in the light rare earth elements (REE)relative to the heavy REE (i.e., chrondite-normalizedLa/Tb ratio less than 1).

The high Sr and Ba contents of the plutonic rocksfrom this site (Table 1) are consistent with an originin the deep-seated portion of an island arc. Table 2presents major and trace element analyses of plutonicrocks from the Masirah ophiolite, Indian Ocean crust(Abbotts, 1979) and calc-alkaline plutonic rocks fromChile (Marsh, 1977; Wells, 1978). Note the lower Sr andBa of the plutonic rocks from oceanic crust relative tothe rocks from Site 453. However, the very high Ba andK2O contents of some Site 453 samples (e.g., Sample453-52-4, 0-4 cm) are likely to be due to the presence ofsecondary potassium feldspar. The most plausible ex-planation for these breccias is a derivation from theWest Mariana Ridge during the uplift resulting from theinitial rifting of the Mariana Trough. We note that thebasalt clasts in the breccia drilled on Site 451 (Leg 59) onthe West Mariana Ridge also have calc-alkaline affin-ities (Mattey et al., 1980; Wood et al., 1980a). However,calc-alkaline magmas may have been erupted elsewherein the Mariana Trough (see the following).

Site 454(18°00.78'N, 144°31.92'E)

This site was located 28 km west of the central grabenof the Mariana Trough in a water depth of 3819 meters.A basement section of 80 meters of basalts interbeddedwith sediments was recovered from Hole 454A. The old-est sediments are between 1.2 and 1.6 m.y. old. Fivelithological units and three paleomagnetic units weredistinguished by the shipboard party. The basalts are allconsiderably more vesicular than mid-ocean ridge ba-salts (MORB) erupted in similar water depths (e.g.,Moore and Schilling, 1973).

The geochemical units correspond with the paleo-magnetic units (Fig. 4). The basalts of GeochemicalUnit 1 are olivine phyric (about 10% phenocrysts) withchrome spinel microphenocrysts and contain up to 14weight percent MgO. The basalts in Geochemical Unit2 are sparsely Plagioclase, olivine, and clinopyroxenephyric, and those in Unit 3 are sparsely olivine phyric(up to 7% phenocrysts) with associated chrome spinel.The major element chemistry of the basalts (Table 3) issimilar to normal or N-type MORB with high CaO(about 11 wt.%), but low K2O and TiO2 (about 1wt.%). However, these basalts have higher Sr, Ba, Th,and light REE contents relative to Zr, Ti, Y, and theheavy REE than N-type MORB. In addition they alsodisplay a depletion of Ta relative to Th and La. Thesetrace element characteristics are similar to those ob-served in island arc magmas. Their high vesicularity alsoindicates a volatile rich magma (Garcia et al., 1979).

Site 456(17°54.7'N, 145° 10.8 E)

This site was located 37 km east of the central grabenof the Mariana Trough in a water depth of 3590 meters.

614

GEOCHEMISTRY OF IGNEOUS ROCKS

The sediment above the basement had an age of about1.8 m.y. In two holes, short basement sections (Hole456 about 35 meters; Hole 456A about 40 meters) ofaltered pillow basalt were recovered. The top of thebasement sections in both holes has suffered intense hy-drothermal alteration, under reducing conditions, withabundant pyrite mineralization and quartz veining. To-ward the bottom of the basement sections in both holes,the basalts are only slightly to moderately altered, underoxidizing conditions, and contain fresh glass.

Two lithological units in Hole 456 and four litho-logical units in Hole 456A were distinguished by theshipboard party. Unit 1 of both holes includes intenselyaltered pillow basalts which are aphyric or sparsely pla-gioclase phyric (about 1 or 2% phenocrysts). Unit456A-25 is a coarsely Plagioclase phyric basalt (greaterthan 20% resorbed phenocrysts) with pseudomorphs ofrare olivine phenocrysts. Unit 456A-3 is a highly vesicu-lar (up to 30% vesicles of 0.2 to 2 mm diameter), aphy-ric basalt with up to 3 percent olivine microphenocrysts.Unit 456A-4 varies from aphyric to sparsely Plagioclasephyric basalts with rare clinopyroxene phenocrysts.Unit 456-3 is also aphyric and similar to Units 456A-3and 4. Units 456-3 and 456A-4 are vesicular but less sothan Unit 456A-3. The shipboard party have groupedUnits 456A-3 and 456-3 together, but from the geo-chemical data it is apparent that basalts correspondingto Units 456A-2 and 3 are not present in Hole 456 andthat Unit 456-3 can probably be equated with Unit456A-4 (see below and Fig. 5). Unit 456A-5 is petro-graphically similar to Unit 456A-3, although less vesicu-lar, but chemically it is indistinguishable from Unit456A-4.

Most of the basalts from both holes have major andtrace element chemistry (Table 4) similar to N-typeMORB, although with slightly higher Sr and Ba con-tents. However, the porphyritic Unit 456A-2 and therelatively Mg-poor Unit 456A-3 are significantly en-riched in Th, Sr, Ba, K, and light REE relative to Zr, Ti,Y, and the heavy REE when compared to N-type MORB.They also show a depletion of Ta relative to Th and La(Table 11). The basalts from these units therefore havetrace element characteristics that are quite similar toisland arc tholeiites.

Site 457(17°49.99'N, 145°49.02'E)

This site was located on the active Mariana arc in awater depth of 2630 meters. The hole penetrates onlycoarse sand and volcaniclastic breccias. We have ana-lyzed one sample of a basaltic andesite clast from thebreccia (Tables 5 and 11). There are no petrographicdata available for the analyzed sample, but the crystal-vitric tuff clasts with which it is associated in the brecciaincluded glass fragments with sparse euhedral Plagio-clase, clinopyroxene, and amphibole phenocrysts. Theanalysis indicates that the sample has calc-alkaline af-

Denotes lithologic unit, not to be confused with Core 456A-2.

finities with high Th, K, Ba, Sr, and light REE relativeto Zr, Ti, Y, and heavy REE contents.

Inter-Site Relationships of Basement Chemistry

The chemical data for the samples from Sites 454,456, and 457 have been plotted on a series of major andtrace element variation diagrams (Figs. 6 through 13).The main feature of the chemical variations in the basiclavas from these sites is the systematic gradation inchemistry from basalts similar to N-type MORB to ba-salts with calc-alkaline affinities. At Sites 454 and 456,which have penetrated inter-arc basin crust, both basalttypes are found interlayered within a single hole. Geo-chemical Units 454A-1, 454A-2, 456A-2, 456A-3, and457 have calc-alkaline affinities, whereas GeochemicalUnits 454A-3, 456-1 (and 456A-1?—extensive hydro-thermal alteration), and 456A-4 have compositions quitesimilar to N-type MORB.

Figure 6 shows that for the same MgO contents theMORB-like lavas have higher Zr and Ti but lower Srthan the units with calc-alkaline affinities. Biaxial plotsof HYG elements (Figs. 7 and 8) confirm that the dif-ferent magma types cannot be related by simple crystalfractionation processes. On such diagrams, magmas re-lated by crystal fractionation should conform to straight-line trends passing through the origin (Weaver et al.,1972; Treuil and Varet, 1973). On the contrary, theMariana Trough basalts can be seen to follow severaldiverse trends.

The occurrence of lavas with MORB-like and arc-likecharacteristics interbedded in the crust of the MarianaTrough at two localities has important implications forthe magmatic processes controlling the formation ofthis inter-arc basin. One possibility is that the two basalttypes were co-magmatic. If so, such small scale litho-stratigraphic variations indicate that eruptions are re-lated to small, short-lived magma reservoirs. This hasalso been proposed for the slow-spreading Mid-AtlanticRidge to explain similar interlayering of geochemicallydifferent basalt types which cannot be related by crystalfractionation processes (Wood et al., 1979a). Rocks pre-viously dredged from the Mariana Trough also show arange of compositions (Hart et al., 1972), and, althoughsome are similar to MORB, others have some geochem-ical affinities with arc magma types. Hart et al. (1972)distinguished those lavas from arc magmas on the basisof their low 87Sr/86Sr (less than 0.7030) and K/Ba (lessthan 85) ratios, despite them having Ba (25-50 ppm), K(2350-4200 ppm), and Sr (155-208) contents,consider-ably higher than average MORB (Ba ~ 12 ppm, K ~1060 ppm, Sr ~ 124 ppm—Wood, 1979; Sun et al.,1979). Also, the dredged basalts studied by Hart et al.(1972) have flat or slightly light-REE-enriched, chon-drite-normalized REE ratios.

Wood et al. (1980a) have recently noted that theabundance of Ba relative to K, Th, U, and Rb variessignificantly in arc basalts from the West Pacific, withthe Japanese arcs (Wood et al., 1980a) being mostenriched in Ba. Hart et al. (1972) suggest that island arctholeiites have K/Ba ratios invariably less than 30 to 40,based on the work of Philpotts et al. (1971) on Japanese

615

D. A. WOOD ET AL.

Table 1. Major and trace element analyses of igneous rocks from Hole 453.

Sample(interval(in cm)

SiO2

Tiθ2A12O3

tFe 2 O 3

MnOMgOCaONa 2OK 2 O

P2O5

Total

NiCrZnGaRbSrYZrNbBaLaCeNdPbTh

Zr/NbTi/ZrY/ZrCe/ZrBa/Zr(Ce/Y)N

Fe/MgK/RbBa/Sr

QOrAbAnNeDiHyOl

Mt11Ap

42-1,16-18

56.00.82

13.49.980.205.725.973.130.710.10

95.99

9368320

81872991

1266

815116

< l

91.054.0

0.320.162.921.272.02

740.01.42

10.54.4

27.621.2

0.07.5

24.20.01.81.60.3

47-1,99-101

50.20.94

11.113.470.19

10.101.302.533.690.25

93.75

1527861835724652

1288

15211611

< l

52.0108.0

0.880.405.541.121.55

875.04.00

0.023.322.8

5.20.00.0

31.110.22.51.90.6

49-1,56-58

46.20.07

23.85.760.077.18

15.210.750.540.00

99.60

2425

8155

41839

< l126

31331

>9.045.0

0.330.11

14.000.820.93

903.00.30

0.03.26.4

60.20.0

12.66.09.9LO0.10.0

49-3,52-54

55.90.74

17.19.570.154.717.823.510.820.17

100.55

48

4320

8384

2541

< l231

911952

>41.0108.0

0.610.275.631.082.36

852.00.60

5.54.8

29.528.4

0.07.7

19.80.01.71.40.4

50-2,37-39

45.10.21

14.08.080.11

15.1016.12

0.380.180.00

99.28

58304

2111

< l204

35

< l40

94261

>5.0253.0

0.600.808.003.270.62

>1461.00.20

0.01.00.5

36.21.5

35.50.0

22.81.40.40.0

52-1,96-98

51.01.00

15.711.370.155.719.862.930.560.31

98.67

111634179

5112582

< l14820271684

>82.073.0

0.300.331.802.652.31

513.00.29

0.73.3

25.128.5

0.015.920.8

0.02.01.90.7

52-2,94-96

46.00.18

19.38.930.18

10.719.520.872.430.00

98.09

36146581125

240< l

6< l

101873321

>6.0176.0<0.17

0.50169.67

—0.97

807.04.24

0.014.67.5

42.40.04.93.8

24.21.60.30.0

53-3,0-2

45.30.17

19.17.040.10

10.3717.170.430.360.00

100.08

3912220125

31418

< l72

2531

< l

>8.0131.0

0.130.639.00

12.280.79

603.00.23

0.02.10.6

49.21.7

28.90.0

15.41.20.30.0

54-1,40-42

46.60.21

21.36.270.097.88

15.971.180.590.00

100.10

2012327144

39829

< l297

2< l

132

>9.0143.0

0.22<O.ll33.00

—0.92

1233.00.75

0.03.56.7

51.01.8

22.80.0

12.21.10.40.0

54-2,84-86

52.40.81

16.110.840.187.045.833.911.660.18

98.99

513481718

3382565

< l429

91914

< l3

>65.074.0

0.380.296.601.871.79

766.01.27

0.09.9

33.421.8

0.05.3

13.611.2

1.91.50.4

54-3,6

51.41.00

15.112.25

0.46.039.322.231.180.13

98.87

621751817

4292155

< l326

12191275

>55.O109.0

0.380.355.932.222.36

576.00.76

2.17.1

19.128.1

0.015.023.3

0.02.21.90.3

55-2,14-16

45.91.08

14.714.480.328.575.793.300.580.16

94.84

3139220

7322

2356

< l144

111711

12

>56.0116.0

0.410.302.571.821.96

688.00.45

0.03.6

29.424.8

0.03.79.2

22.72.72.20.4

Note: All CIPW norms calculated assuming an Fe2θ3/FeO ratio of 0.15.

lavas. However, island arc tholeiites recently recoveredfrom DSDP Site 448 have K/Ba ratios greater than 90(Mattey et al., 1980). We note also that the low-tem-perature sea-water alteration could significantly in-crease the K/Ba ratios of the lavas. Thus, the high K/Baratios of the dredged basalts do not necessarily distin-guish them from arc magmas. The basalts recoveredfrom the Leg 60 drill sites have suffered variable degreesof alteration, and consequently the K/Ba ratios varyconsiderably. Nevertheless, the K/Ba ratios of mostsamples from Units 457, 454A-1, and 454A-2 are lessthan 50, but in Unit 454A-3 K/Ba varies from 64 to 94.The calc-alkaline arc lavas drilled at Site 451 (Mattey etal., 1980) have K/Ba ratios less than 40.

In Figures 9 and 10, Ni versus Zr and Cr versus Zr,the units which have calc-alkaline affinities and thoseMORB-like units follow distinct trends. This is mainlydue to the different Zr contents of the different magmatypes. The plot of Zr versus Ti (Fig. 11) shows that mostof the lavas have Ti/Zr ratios between 75 and 100,whilst N-type MORB generally have Ti/Zr ratios in therange 90 to 120. Significantly, Units 456A-3 and 457have Ti/Zr ratios less than 90. The main HYG elementdifferences between the two magma types have beensummarized in Figure 12. Here the HYG element abun-dances of three representative samples have been nor-malized to estimated primordial mantle abundances(Wood, 1979) and arranged in approximate order of in-

616

GEOCHEMISTRY OF IGNEOUS ROCKS

Table 1. (Continued).

55-3,25-27

41.80.04

21.78.290.11

12.8112.820.510.290.00

98.32

588

31133

413< l

7< l68

94232

>7.035.0

<0.140.579.71

—0.75

808.00.16

0.01.81.8

56.91.46.30.0

29.51.50.10.0

55-3,90-92

44.60.23

19.08.360.10

11.4816.920.320.070.00

101.09

462342514

< l311

110

< l138

< l1

< l< l

>IO.O137.0

0.10<O.IO

1.30—

0.84> 623.0

0.04

0.00.30.0

49.61.5

27.00.0

19.01.40.40.0

55-4,144-146

43.70.05

27.65.990.096.99

12.820.961.200.00

99.38

172

281711

456< l

7< l547

4432

< l

>7.044.0

<0.140.57

78.14—

0.99906.0

1.20

0.07.15.6

64.21.40.00.0

18.61.00.10.0

56-2,33-35

42.60.05

24.78.150.149.06

11.690.970.730.00

98.05

41< l50167

418< l

7< l301

2631

< l

>7.039.0

<0.140.86

43.00—

1.04869.0

0.72

0.04.47.7

59.30.40.00.0

25.01.40.10.0

56-2,85-87

48.00.05

23.05.900.098.249.762.651.340.00

98.99

221

261414

354< l10

< l409

2< l

13

< l

>IO.O31.0

<O.IO<O.IO40.90

—0.83

795.01.16

0.08.0

19.747.4

1.61.40.0

20.31.00.1ö.o

57-1,30-32

48.30.83

15.211.670.178.34

13.471.690.210.03

99.99

132663182

3871221

< l806655

< l

>21.0234.0

0.570.293.811.231.62

884.00.21

0.01.3

14.333.40.0

27.410.68.42.01.60.1

57-3,8-10

42.40.05

24.76.940.09

11.2314.340.380.160.00

100.32

516219121

427< l

6< l42

8312

< l

>6.055.0

<0.170.507.00

—0.72

1353.00.10

0.01.00.9

64.91.24.90.0

25.11.20.10.0

57-4,36-38

43.00.16

17.910.730.12

14.4111.940.440.090.00

98.78

7726542131

3641

10< l5811335

< l

>IO.O95.00.100.305.807.370.86

780.00.16

0.00.63.8

47.10.0

10.46.4

28.51.90.30.0

57-4,123-125

42.60.03

29.14.790.067.44

15.260.560.530.00

100.34

32< l

16164

523< l10

< l148

3312

< l

>IO.O20.0

<O.IO0.30

14.80—

0.751100.0

0.28

0.02.10.0

75.02.50.50.0

17.80.80.10.0

58-1,6-8

43.71.28

14.115.300.39

14.241.023.000.050.20

93.22

940

12720

< l222

3277

< l112

1719146

< l

>77.0100.0

0.420.251.451.461.25

>390.00.50

0.00.3

27.24.10.00.0

36.616.12.92.60.5

59-1,13-15

70.50.46

12.16.880.362.890.664.980.240.07

99.14

< l< l300

122

1083084

< l104

71284

< l

>84.033.00.360.141.240.982.76

988.00.96

31.21.4

42.52.90.00.0

16.60.01.20.90.2

59-1,47-49

61.00.73

12.86.650.111.948.807.500.030.17

99.77

716196

< l1241344

< l324

15842

>44.099.00.300.340.732.833.98

> 282.00.26

1.30.2

63.61.20.0

25.60.00.01.21.40.4

63-1,0-2

48.00.82

13.913.060.26

12.632.663.340.210.10

95.00

102067173

2751645

< l118

615835

>45.0110.0

0.360.332.622.301.20

592.00.43

0.01.3

29.713.30.00.0

34.611.72.41.60.3

compatibility with the major mantle mineral phases.The enrichment of Cs, Rb, Ra, Th, U, and K relative toHf, Zr, Ti, Y, and the heavy REE and the markeddepletion in Ta and Nb for the calc-alkaline magmatypes clearly distinguish them from MORB-like lavas.

In Figure 13 the lavas from Sites 454, 456, and 457have been plotted in the Th-Hf/3-Ta discriminationdiagram (Wood et al., 1979b; Wood, 1980) with thefields of N-type MORB (A), E-type MORB (B), withinplate lava series (C), and active plate margin lava series(D) distinguished. The field of active margin lava series(D) has been enlarged relative to that originally pro-posed (Wood et al., 1979b, fig. 4) to include additionalanalyses of arc tholeiites from Japan and DSDP Site 448as well as the fore-arc lavas from DSDP Sites 458

through 461 (described later). All the basalts from Site456 except Unit 456A-3 plot within the field of N-typeMORB, and Unit 454A-3 plots between fields A and D(Fig. 13). Units 456A-3, 454A-1, 454A-2, and 457 plotin field D as expected.

Interestingly, the Site 454 lavas plot along a trend inFigure 13 between N-type MORB and calc-alkalinemagma types. We interpret this as a mixing line. Thereseem to be two possible explanations for this mixingrelationship: (1) the two magma types were comagmaticin the crust of the Mariana Trough, being derived fromsmall, short-lived magma chambers, in which mixingbetween the magma types could occur in some in-stances. Mixing due to episodic injection of magma intoa small, sub-crustal reservoir is a common process at the

617

D. A. WOOD ET AL.

30

20

10

n

\\

X \,_ Plagioclase s

(An7O-Ango) \\\•

- D Δ$^J p l

•

× X

XX

I

N\

Λ\

i

•

X

®

N\

\\

\

Basaltic

Plutonics

Volcanics

Metavolcanics

MasirahGabbros

ChileanGabbros

\ ~\

\

Olivine \(Fo70-Fogo) \ N

\\ >^

\

hornblendes x ^

\N

Ca-poor pyroxene

_L10 20 30

MgO (wt. %)

Figure 2. A12O3 versus MgO for igneous rocks from Site 453, showingthat most of the plutonic rocks contain cumulate Plagioclase and aferromagnesian mineral.

Mid-Atlantic Ridge (Rhodes et al., 1979). Resorbedphenocrysts are present in the porphyritic Unit 456A-2and could support a magma mixing hypothesis. Thepresently available data for petrography and mineralogyof the basalts from this site do not provide additionalevidence of magma mixing (e.g., resorbed phenocrystsout of equilibrium with the bulk rock; differences in thecomposition of glass inclusions in the phenocrysts andbulk rock), but a more detailed study would clearly beworthwhile. (2) An N-type MORB magma type couldhave assimilated small but variable amounts of arc-de-rived sediments. Basalts from Sites 454 and 456 are as-sociated with arc-derived vitric tuffs. One basalt at Site456 actually contains a large recrystallized tuffaceousxenolith (see Site 456 summary, this volume). Low seis-mic velocities at Site 454 suggest that sediments may beinterbedded with basalts for several hundred meters andsome of the more massive basalts could be intrusive.This would provide a suitable environment for sedimentassimilation. The fact that from the same cooling unitsome samples have trace element characteristics of arctholeiites, and other samples have trace element charac-teristics of N-type MORB, tends to support a heteroge-neous assimilation process rather than a more homoge-neous magma mixing process. Both processes could ex-plain the lower 87Sr/86Sr and higher K/Ba ratios of theselavas (e.g., dredge samples of Hart et al., 1972) thanmost calc-alkaline basalts.

Magma mixing and/or sediment assimilation pro-cesses may be important in the initial stages of openingof an inter-arc basin and could explain the chemical

1.0

0.5

600

400

200

-

•

-

Δ

-

_

Si-rich *differentiates

•

ΔΔ

Δ

, -" '

• ~ • - ^

XX

Δ

ΔX

X

X

x ×

•

•# .# ** × \ .

"– -> x~- -~.α ~~" - - _ ^

J3 - - - - - " "

1

β

-

• Plutonics

× Volcanics® Metavolcanics

g MasirahGabbros

^ ChileanGabbros

-

-

10MgO (wt.

15 20

Figure 3. TiO2 and Sr versus MgO for igneous rocks from Site 453.Gabbros from the Masirah Ophiolite, Oman (Abbotts, 1979), alsoPlagioclase cumulates (cf. Table 2), considered to represent thecumulate layer of oceanic crust, have lower Sr than the Site 453rocks.

variation in basalts dredged from several such basins,which seem to have compositions related to both N-typeMORB and arc magma types, but cannot be classed asone or the other. For example, those from the Lau Basin(Gill, 1976, Hawkins, 1976), the East Scotia Sea (Saun-ders and Tarney, 1979; Hawkesworth et al., 1978), andthe Mariana Trough (Hart et al., 1972; this study).Some basalts from those basins have affinities with arcmagma types, sharing with them higher K, Rb, Ba, Th,Sr, and 87Sr, and lower Ti, Zr, Hf, Ta, and Nb relativeto N-type MORB. In contrast, basalts recovered fromthe broader, more developed marginal basins of theWest Pacific—i.e., the Shikoku, West Philippine andParece Vela Basins—show no clear chemical affinitieswith destructive margin magma types (Marsh et al.,1980; Mattey et al., 1980; Wood et al., 1980a, 1980b).Tarney et al. (1977) noted similar relationships in themarginal basins of the Scotia arc region. This may bedue to the progressive decrease in arc magmatic activityor arc-derived volcaniclastic sediments in the basins asthey become more developed. If so, it has important im-plications for the type of marginal basins sampled byophiolites. For example, the Late Cretaceous Mediter-

618

GEOCHEMISTRY OF IGNEOUS ROCKS

Table 2. Major and trace element XRF analyses of basic plutonic rocks from different tectonic environments.

SampleNumber C378 C379

Major Element Oxides (wt.

SiO2TiO2A12O3tFe2O3MnOMgOCaONa2OK2OP2O5Total

50.50.68

19.99.010.144.05

11.273.560.500.17

99.77

49.51.61

20.09.570.183.43

11.123.620.230.89

100.13

Trace Elements (ppm)

NiCrZnGaRbSrYZrNbBaLaCePbTh

681236022

1591

1176

< l179

10< 1

8< l

1820

10825

< l722

21774

152193063

Gabbros from the Chilean Cordillera

C387

°7o)

46.50.80

21.48.890.114.53

13.972.710.090.01

98.99

41705920

< l590< 135

359

34

< l< l

(Wells,

C389

51.51.15

18.69.820.144.40

10.203.430.410.09

99.79

58676521

5541

1080

4141

764

< 1

1978; Marsh, 1977)

C397

49.80.22

16.18.100.138.86

13.331.170.120.01

97.83

174510

6317

< l482

330

< 1337

< l< l< l

C62

46.90.16

19.16.440.139.82

13.901.110.090.04

97.70

2591935615

1333

534

243

31112

< 1

C184

46.50.79

18.410.20.177.05

12.71.450.450.04

97.84

2449552121

3438

374

595

116

< l

C71

55.40.88

17.47.520.135.749.143.760.150.17

100.29

56472223

4671

20190

6175

1226

< l8

C117

50.00.53

20.43.300.115.60

16.32.150.080.05

98.50

35922321

1540

1454

1533

145

< l

Gabbros from the MasirahOphiolite, Oman (Abbotts, 1979)

X288

42.90.07

24.71.080.026.85

22.220.490.030.00

98.27

91265

158

< 1116< 1

8< 1

723

< ln.d.

MA65

46.40.18

17.24.87

n.d.13.8913.621.610.090.01

97.48

237705

2414

< l175

38

< l16

< l< 1< 1< l

MA205

49.10.25

20.94.290.007.24

11.143.890.280.02

96.77

70355

9152

2133

194

41< l< 1< 1< 1

MA164

50.40.45

20.24.050.075.6812.453.800.080.02

97.20

40512020

< l294

428

533

24

< 1< 1

MA417

59.00.55

18.82.870.001.804.049.970.050.10

97.18

5< l

424

< 118357

71152874488

< l17

Notes: C378, C379, C387, C397 = plagioclase-clinopyroxene leucogabbros from Puerto Aisen, Southern Chile. C389 = plagioclase-clino-pyroxene leucogabbro with minor primary hornblende from Puerto Aisen. C62, C184, C71, C117 = plagioclase-clinopyroxene leuco-gabbros from Atacama Province, Northern Chile.

X288 = leuco-olivine gabbro (cumulus Plagioclase > olivine = diopside).MA65 = clinopyroxene gabbro (Plagioclase = diopside).MA205 = clinopyroxene gabbro (screen to sheeted dykes).MA164 = pegmatitic clinopyroxene gabbro.MA417 = plagiogranite.

ranean ophiolites, which contain both MORB and arc-type lavas in their crustal sequences (Pearce, in press),may represent crust formed in narrow, poorly de-veloped inter-arc basins, such as the Mariana Trough orLau Basin, rather than broad, well-developed back-arcbasins.

For comparison, Tables 6 and 7 present the majorand trace element compositions of calc-alkaline basaltsand andesites dredged by P. Fryer and D. Hussong froma seamount in the northern part of the active MarianaArc (21°58'N, 143°28'E) following seismic activitysuggesting a recent eruption in the area. These lavas arecompositionally similar to the sub-aerially erupted lavasfrom the Mariana Islands (Dixon and Batiza, 1979;Stern, 1979; Chow et al., 1980) and the basalts fromDSDP Sites 451, 453, and 457. They have been plottedin the Th-Hf/3-Ta discrimination diagram (Fig. 13)and plot close to the other calc-alkaline lavas. However,they are significantly more evolved (i.e., low MgO: 2 to4 wt.%) than the drilled samples. This is consistent withthe extensive near-surface crystal fractionation thoughtto prevail along the Mariana Arc (Stern, 1979).

Mariana Fore-arc and Trench

Site 458(17°51.85'N, 146°56.06'E)

This site was located on the southeast flank of a 40mgal. positive gravity anomaly, 120 km east of the ac-

tive Mariana arc, on the fore-arc midway between thearc and the trench in a water depth of 3447 meters. Thesediments overlying the basement were approximately35 m.y. old. Hole 458 penetrated 209 meters of base-ment rocks including glassy pillow lavas and massiveflows (Fig. 14). The shipboard party divided the base-ment section into five lithological units. Units 1, 2, and3 are petrographically similar but distinguished by theirform: massive, coarse-grained lavas of Unit 2 separatethe vesicular, glassy, fine-grained pillows of Units 1 and3. These three units make up 123.5 meters of the coredinterval. They are petrographically unusual in havingendiopside and bronzite phenocrysts (with no olivine),and significant quantities of Plagioclase are only pres-ent in the more holocrystalline samples. The shipboardparty noted that the two pyroxene, glassy samples withno Plagioclase resembled the boninites or Mg-rich ande-sites previously reported from elsewhere on Western Pa-cific fore-arcs, for example, the Bonin Islands (Kurodaand Shiraki, 1975; Kuroda et al., 1978; Shiraki et al.,1979), the Mariana Trench (Dietrich et al., 1978; Cam-eron et al., 1979), and Cape Vogel in southeast PapuaNew Guinea (Dallwitz, et al., 1966; Dallwitz, 1968).

Unit 4 consists of about 75 meters of vesicular, glassyto fine-grained, sparsely Plagioclase and augite phyricpillow and massive lavas, relatively rich in Fe-Ti oxides.The top 10 or so meters of this unit (Sections 41-1 and41-2) are highly fractured and, although not distin-guished petrographically, have a distinct chemistry fromthe rest of the unit (see below). We refer to this part as

619

D. A. WOOD ET AL.

4 ---_-_-_-----: Vitric Mud

3900-

MagneticInclination

-90c 0 +90°

Zr (ppm) Ti (wt. %) Sr (ppm) Cr (ppm)

60 70 80 0.8 1.0 1.2 150 200 250 200 400

GeochemicalUnits

3920-

-£ 3940-

t

3960-

3980-

10

S=SSL

Unit 12 Massive FlowsOl. Flag. Basalt

Graded DrillingBreccia' 'Unit 2

Med. Grained\Aphvr ic Basal'

Unit 3Ol.-Plag. PillowBasalts andInterbeddedVitric Mud

Unit 4Sparsely OL-Plag.-PhyricMassive BasaltFlows

Vitric Mud

Unit 5OlivineMicrophyricPillow Basalts

>

\

4000

Figure 4. Geochemical and lithological stratigraphy of the basement sections of Hole 454A.

i >

\

> 1

> 3

Unit 4a. Another highly fractured zone (-0.5 m) formsthe lower part of Section 44-1 in Unit 4 and may be com-positionally similar to Unit 4a but has not been sampledin this study. Unit 5 forms the lower 30 meters of thehole and consists of massive, fine-grained sparsely pla-gioclase and augite phyric lavas relatively rich in Fe-Tioxides. All the units have suffered low-temperature(zeolite facies) alteration. Most of the glassy matricesare recrystallized to clays, although a few fresh glassypillow rinds are preserved.

The lavas from Units 1, 2, and 3 show coherentchemical compositions (Table 8 and Fig. 14) with highSiO2 (52 to 60 wt.%), moderate to high MgO (5 to 10wt.%), high K20 (0.5 to 1.5 wt.%), and very low TiO2(about 0.3 wt.%), P2O5 (about 0.03 wt.%), Zr, Hf,REE, and Y contents. Although they are chemicallyrelated to boninites, these Site 458 rocks are moreevolved: previously described boninites (e.g., Table 9)generally have MgO contents >9 percent, Mg-numbers>0.7, Ca <8 percent, A12O3 < 14 percent, Ni > 100ppm, and Cr >600 ppm. Most of the Site 458 Units 1,2, and 3 lavas have MgO <9 percent, Mg-numbers<0.7, CaO >8 percent, A12O3 > 14 percent, Ni < 100ppm, and Cr < 300 ppm, and are probably best termedMg-rich andesites. Unit 458-4 lavas are similar to those

of the upper three units but with slightly higher abun-dances of Fe and the HYG elements and lower Mg-numbers and Cr and Ni contents. They are best termedandesites, but the relatively low abundances of the HYGelements in these lavas indicate that they are related tothe Mg-rich andesites and boninites described earlier.

Units 458-4a and 458-5 are chemically different fromthe other Site 458 lavas and have the major elementcompositions of basaltic andesites. Both units havehigher TiO2 (about 1.1 wt.%) and HYG elements con-tents with low CaO (<7.5 wt.%) contents. Unit 458-4ahas low MgO (<4 wt.%) and Unit 458-5 has very highFe contents (EFe2O3 about 13.75 wt.%) and could betermed a ferrobasalt. Units 458-4a and 458-5 both havecompositions similar to those of primitive island-arctholeiites.

Site 459(17°51.75N, 147°18.09'E)

This site was located on the easternmost fore-arcbasin, adjacent to the trench slope break in a waterdepth of 4120 meters. The basement was dated as pre-late Eocene (74.5 m.y.) by the overlying sediments. Hole459B penetrated 132.5 meters of basaltic basementdivided into four lithologic units. The basalts are gener-

620

GEOCHEMISTRY OF IGNEOUS ROCKS

Table 3. Major and trace element analyses of igneous rocks from Hole 454A.

Sample(intervalin cm)

SiO2

Tiθ2A12O3

tFe2θ3MnOMgOCaONa2θK 2 O

P2O5

Total

NiCrZnGaRbSrYZrNbBaLaCeNdPbTh

Zr/NbTi/ZrY/ZrCe/ZrBa/Zr(Ce/Y)N

Fe/MgK/RbBa/Sr

QOrAbAnNeDiHy

OlMt11Ap

5-1,2-4

49.30.89

14.78.500.178.85

14.222.730.330.11

99.80

117223

53194

1762161

< l25

99756

>61.087.00.340.150.411.051.11

689.00.14

0.02.0

17.927.0

2.834.8

0.011.3

1.51.70.3

5-3,106-108

49.90.95

14.98.580.10

10.5411.97

3.130.17

0.10

100.32

114228

4820

11782363

141

612811

63.090.0

0.370.190.651.280.94

1411.00.23

0.01.0

24.325.9

1.126.2

0.017.1

1.51.80.2

5-4,15-17

49.30.86

14.49.420.12

13.3010.972.360.180.09

100.97

382532

52162

1431858

< l49104561

>58.089.00.310.070.840.550.82

730.00.34

0.01.0

19.827.8

0.020.5

6.620.0

1.61.60.2

5-4,16-18

49.30.82

14.19.370.12

14.1711.052.300.140.08

101.42

426594

5315

1137

1851

143

99762

51.097.0

0.350.180.841.230.77

1179.00.31

0.00.8

19.227.3

0.021.1

5.522.0

1.61.50.2

6-2,102-104

50.30.94

14.98.990.11

10.8711.742.930.130.10

101.10

108218

:>018

11612258

1421010

954

58.097.00.380.170.721.120.96

1087.00.26

0.00.8

24.526.9

0.024.2

2.217.0

1.51.80.2

8-1,2-4

49.30.88

14.29.410.14

10.1711.802.660.200.07

98.83

253424

5418

< l1602255

1101076

< l< l

55.096.0

0.400.130.180.781.07

>1652.00.06

0.01.2

22.826.6

0.026.1

4.314.8

1.71.70.2

10-1,81-83

50.90.91

14.88.340.129.96

11.373.210.220.09

99.90

151211

5715

< l221

2362

< l34

910

77

< l

>62.088.0

0.370.160.551.070.97

>1835.00.15

0.01.3

27.225.2

0.024.7

3.214.3

1.51.70.2

11-1,40-42

51.50.96

15.28.520.128.90

11.413.360.230.11

100.27

1231966220

2234

1973

< 1191

789

< l

>73.O79.0

0.260.102.620.901.11

959.00.82

0.01.4

28.425.6

0.024.4

4.211.7

1.51.80.3

11-4,71-73

52.20.88

14.98.690.12

12.2010.702.520.290.11

102.59

174232

5516

< l214

1962

< l64

7108

< l< 1

>62.085.0

0.310.161.031.290.83

>2382.00.30

0.01.7

20.827.8

0.018.716.910.1

1.51.60.2

12-1,100-102

51.80.96

13.98.830.119.89

11.242.780.17

0.10

99.75

116192

5619

< l221

2368

< 152

5149

< l< l

>68.085.0

0.340.210.761.501.04

>1436.00.24

0.01.0

23.625.0

0.024.416.6

5.01.51.80.2

12-2,26-28

51.61.04

15.29.080.138.53

11.533.320.340.13

100.81

1141786125

< l239

2572

1134

7874

< l

72.087.00.350.111.860.791.23

>2806.00.56

0.02.0

27.925.2

0.024.9

4.011.4

1.62.00.3

14-1,2-4

51.51.17

15.79.440.146.55

10.752.680.790.10

98.80

128203

6322

81792782

2

829

1211

< l< l

41.086.00.330.151.001.091.67

816.00.46

0.64.7

23.028.8

0.020.117.90.01.72.20.2

15-1,13-15

50.81.16

15.69.450.147.49

11.042.690.620.09

99.03

133212

6224

51832678

2701611

8< 1< l

39.089.00.330.140.901.041.46

1026.00.38

0.03.7

23.028.8

0.021.114.9

3.61.72.20.2

16-1,32-35

52.01.20

16.09.230.146.64

11.082.890.640.11

99.95

134222

6322

3185

2381

i941596—

< 1

81.089.00.280.111.160.961.61

1760.00.51

0.13.8

24.528.8

0.020.917.10.01.62.30.3

16-1,111-113

50.21.12

14.59.920.168.73

10.422.640.440.09

98.26

269390

66173

1702377

2105

15139—2

39.087.00.300.171.361.391.32

1229.00.62

0.02.7

22.726.9

0.020.516.8

5.41.82.20.2

ally sparsely clinopyroxene and Plagioclase phyric withmore than 10 percent vesicles (and in some cases morethan 30%). The petrological and geochemical differ-ences between the four units are subtle, and they havebeen distinguished by grain size, vesicularity, and mag-netic intensity. The lavas contain up to about 3 percentFe-Ti oxides and patches of quartz and alkali feldsparin the mesostasis. They are relatively rich in Fe with be-tween 10 and 14 weight percent Fe2O3. Some of the fine-grained basalts contain spherulitic zones next to the ves-icles with abundant needlelike opaque phases thought tobe ilmenites.

The Site 459 basalts are quartz-normative and similarto primitive island arc tholeiites (Table 10 and Fig. 15)with low Ti and Zr, high K and a depletion of Tarelative to La and Th. The high vesicularity implies thatthe magmas from which the Site 459 basalts formedwere rich in volatiles. Their low Sr (< 150 ppm) and Ba(generally < 50 ppm) contents and depletion of the lightREE (chondrite-normalized La/Tb < l ) distinguish

them from the calc-alkalic lavas from the MarianaTrough and Ridge.

Site 460(17°40.14 N, 147°35.92E)

This site is located 25 km from the center of theMariana Trench in a sediment pond on the arc-side wallof the trench in a water depth of 6445 meters. Basementwas not reached, but basaltic clasts were recovered froma conglomerate of early Miocene or late Oligocene age(25-27 m.y.). Three such basalt clasts have beenstudied. Samples 460-9,CC, 11-13 cm and 460-9,CC,74-76 cm are sparsely Plagioclase (about 2°7o), olivine(rare pseudomorphs), and clinopyroxene (trace) phyricbasalts with abundant vesicles (about 20%). Sample460A-11-1, 29-31 cm is an aphyric, coarse-grained ba-salt with abundant secondary amphibole replacing relictclinopyroxene. All three samples are primitive island arctholeiites (Table 5) similar to those from Site 459 andUnits 458-4a and 458-5 (Table 11).

621

D. A. WOOD ET AL.

Hole 456Geochemical

Units (PP"1) Sr (ppm)

100 200 300 150 250

Zr (ppm)

70 80 9C

•l•

1

(A

/

•

f

•

SITE 456

Petrography

Hole 456A

Zr(ppm) Sr (ppm) Cr (ppm)

70 80 90 150 250 100 200 3001 r i

•

••

\

ė

•

1 1— I

•

TI\

•

•

•

J •J

1

2

> 3

>• 4

Figure 5. Geochemical and lithological stratigraphy of the basement sections of Holes 456 and 456A.

Site 461(17°46.01N, 147°41.26 E)

This site was located on a small patch of sediment ona local high in the arc-side wall of the Mariana Trench ina water depth of 7030 meters. Basement was not reached,but cobbles of igneous and metamorphic rocks were re-covered in the few meters of sediment (about 20 m) ofuncertain age. One metabasalt sample of unknown pe-trography has been sampled and has a composition(Table 5) similar to that of the primitive arc tholeiites ofthe other fore-arc and trench sites (Table 11).

DETAILED RELATIONSHIPSOF BASEMENT CHEMISTRY

The lavas from these Mariana fore-arc and Trenchsites can be divided into two groups: (1) the Mg-richandesites and andesites from Site 458 which are relatedto a boninite magma type. The Mg-rich andesites areprobably derived from a boninite parental magma bycrystal fractionation of the two pyroxene phases; (2) theprimitive island arc tholeiites recovered from all foursites. Major and trace element variation diagrams forthese rocks are presented in Figures 16 through 24. TheMg-rich andesites and andesites of Site 458 Units 1through 3 follow a trend toward enrichment of Ca andAl with decreasing Mg (Fig. 15). This suggests thatPlagioclase has not fractionated from these magmas, atleast from the more primitive ones. The Zr and Ti con-tents remain approximately constant with decreasingMgO in these units.

The andesites of Unit 458-4 have higher Fe, Ti andZr, but lower Ca and Al contents than the andesites ofUnits 458-1 through 3 with similar MgO contents (Fig.16). Most of the Unit 458-4 andesites also have lowerK/Zr, Zr/Y, Sr/Zr ratios (Fig. 17) than the Units 458-1through 3. These differences suggest that the Unit 458-4andesites cannot be related to the Units 458-1 through 3lavas by crystal fractionation processes. These HYG ele-ment ratios should remain approximately constant dur-ing crystal fractionation of two pyroxene assemblage.Units 458-1 through 4 have similar Ta/Tb, Th/Tb, andLa/Tb ratios (Fig. 18) and have chondrite-normalized

La/Tb and La/Eu ratios less than 1. Most of the pre-viously described boninite-like lavas have been slight-ly enriched in the light REE relative to the heavy,REE—that is, chondrite-normalized La/Tb and La/Euratios greater than 1 (e.g., Sun and Nesbitt, 1978;Hickey and Frey, 1979). Also, Units 458-1 through 4have low Ti/Zr (less than 75), Y/Ti, and Y/Zr (less than0.35) ratios (Figs. 17 and 21). The low Ti/Zr ratios arecharacteristic of boninites from other localities (Sun andNesbitt, 1978) and distinguish them from most otherterrestrial magma types. Relative to estimated primor-dial mantle abundances for a chondritic earth model, Zrand Hf are enriched relative to Eu, Ti, Tb, and Y inthese lavas (Fig. 23)—a point also noted by Bougault etal. (this volume). The significant fractionation of Ti,Zr, and Y relative to each other is difficult to achieve bymineral-melt processes unless garnet is involved (theheavy REE, Y, and Ti are partitional into garnet prefer-entially to Zr).

The primitive island arc tholeiite lavas from thesesites have higher Fe, Ti, P, and HYG elements, butlower Ca and Al contents than the andesites of Site 458with similar MgO contents (Fig. 16). Moreover, theyhave lower Sr/Zr, Zr/Y, Th/Tb, La/Eu, La/Tb, andTa/Tb ratios (Figs. 17 and 18) and Ni and Cr abun-dances (Figs. 19 and 20). All these island arc tholeiiteshave Ti/Zr ratios close to 100 (Fig. 21)—except Unit458-4a, which has low Mg and Fe and has probably re-sulted from crystal fractionation of an Fe-Ti oxidephase which would reduce the Ti/Zr ratios. Also theyfollow coherent trends on the trace element variationdiagrams.

The lavas from the fore-arc and trench sites plot in adistinct and highly restricted field in the Th-Hf/3-Tatriangle (Fig. 22), close to the boundary between activemargin and N-type MORB magma types. The andesitesof Units 458-1 through 4 plot closer to the Th apex ofthe triangle than the arc tholeiites. All the lavas plotclose to other primitive island arc tholeiites from theWest Pacific active margins. However, these lavas areless depleted in Ta relative to the HYG elements thanother active margin volcanics studied to date. This point

622

GEOCHEMISTRY OF IGNEOUS ROCKS

Table 4. Major and trace element analyses of igneous rocks from Site 456.

Sample(intervalin cm)

Siθ2TiO2

A12O3

tFe 2 O 3

MnOMgOCaONa 2OK2OP2O5

Total

NiCrZn

GaRbSrYZrNbBaLa

CeNdPbTh

Zr/NbTi/ZrY/ZrCe/ZrBa/Zr(Ce/Y)N

Fe/MgK/RbBa/Sr

QOrAbAnNeDi

HyOlMt11

Ap

16-1,145-147

49.31.25

14.510.070.17

10.378.663.340.090.09

97.92

9030511821

< l1462480

1109

61110——

80.094.00.300.141.361.131.13

> 747.00.75

0.00.5

28.924.90.0

14.910.215.31.82.40.2

16-2,92-94

46.21.26

12.310.220.29

11.2212.552.400.060.11

96.56

74276

9318

< l141

2482

167

7129—

—

82.092.00.290.150.821.231.06

> 506.00.48

0.00.4

17.523.3

1.932.70.0

18.71.82.50.3

Hole 456

17-1,95-97

49.31.33

13.310.780.19

12.818.073.680.100.09

99.66

90297

8618

< l1462490

187

698—

—

90.089.00.270.100.970.920.98

> 789.00.60

0.00.6

31.219.50.0

16.20.8

26.11.92.50.2

18-1,50-52

50.21.20

14.59.350.188.39

12.143.300.310.14

99.63

99254

6422

< l1842795

122

876

< l< l

95.076.00.280.070.230.641.29

> 2532.00.12

0.01.8

26.323.90.9

29.00.0

13.01.62.30.3

19-1,18-20

48.21.14

15.59.280.146.58

11.642.770.540.08

95.94

85253

6119

11762386

121

798

< l< l

86.080.00.270.100.230.961.64

> 4474.00.12

0.03.3

24.429.6

0.024.6

4.09.11.72.30.2

10-1,25-27

45.70.628.18.420.217.92

11.803.530.750.11

87.21

2614801511

1382554

< l136

710

87

< l

>54.069.00.460.192.520.981.23

563.00.99

0.05.1

21.24.77.1

50.00.07.81.71.30.3

11-1,

76-78

48.21.38

14.011.170.21

15.284.182.440.060.09

96.99

85338198

15< l1002680

< l5417

575

< l

>80.0103.0

0.330.060.680.470.85

>515.O0.54

0.00.4

21.320.80.00.0

39.010.02.02.70.2

12-1,11-13

48.00.77

18.66.860.10

10.1512.592.800.130.06

100.09

109226

3420

< l212

1666

< l117

1176

< l< l

>66.070.0

0.240.111.771.070.78

>1087.00.55

0.00.8

18.537.82.8

19.50.0

17.31.21.50.2

Hole 456A

12-1,41-43

49.71.15

15.510.190.166.09

11.172.790.580.11

97.46

681526824

4246

2777

175111411

< l< l

77.090.00.350.180.971.271.94

1195.00.30

0.03.5

24.228.90.0

22.512.23.51.82.20.3

13-1,23-25

50.71.18

16.69.870.165.23

11.233.270.620.11

98.92

51106

7716

8251

2684

281101712

3< l

42.084.0

0.310.200.961.612.19

647.00.32

0.03.7

28.029.0

0.022.0

6.75.51.72.30.3

14-1,12-14

52.51.30

16.210.480.174.339.523.430.980.15

99.10

2318792212

2802489

287122114

< l< l

45.088.00.270.240.982.152.81

681.00.31

1.15.9

29.326.2

0.0

17.114.90.01.82.50.3

14-1,38-40

48.41.18

15.09.360.146.39

12.183.120.470.10

96.38

71234

6314

618523

901

43151293

< l

90.079.0

0.260.130.481.281.70

647.00.23

0.02.9

25.326.6

1.129.00.0

10.01.72.30.2

15-1,27-29

49.51.12

15.59.650.137.40

12.282.750.400.09

98.84

88214

68144

1732382

14313

871

< l

82.082.00.280.100.520.851.51

836.00.25

0.02.4

23.529.20.0

26.04.39.71.72.20.2

is also apparent from Figure 20. Comparing this figurewith Figure 10, it can be seen that the fore-arc tholeiiteshave a much smaller depletion in Ta relative to La andCe, and the Mg-rich andesites from Unit 458-1 show nodepletion in Ta relative to La. Figure 24 illustrates thelarge range of La/Ta ratios measured in the lavas recov-ered from the arcs and basins of the West PacificOcean; the Mariana fore-arc lavas and those from theMariana Trough have La/Ta ratios less than 40 and, ex-cept for the Site 458 andesites, greater than 15; the lavasfrom Japan, the Mariana arc, the West Mariana Ridgeand the Kyushu-Palau Ridge have La/Ta ratios greaterthan 40. These trace element variations suggest thatseveral magma types, unrelated by crystal fractionationprocesses, have been erupted during the development ofthe island arcs and inter-arc basins of the west Pacific.The geochemical diversity of the lavas implies that the

petrogenetic processes and mantle source compositionsinvolved are complex.

DISCUSSION

The main aim of this chapter has been to provide afactual account of the geochemistry of Leg 60 igneousrocks. A detailed consideration of the petrogeneticimplications of the data will be deferred to a subsequentpaper. However, qualitative assessments can be made ofsome of the recently favored petrogenetic models forisland arc and back-arc basin magmas, and constraintsplaced on them by the tectonic evolution of the Philip-pine Sea.

Most recent trace element and radiogenic isotopestudies of continental margin, island arc, and marginalbasin lavas have favored petrogenetic models invokingthe involvement of fluids, probably of a supercritical

623

D. A. WOOD ET AL.

Table 5. Major and trace element analyses of igneous rocks fromHoles 457, 460, and 461.

Sample(intervalin cm)

SiO2

TiO2

A12O3

tFe 2 O 3

MnOMgOCaONa2OK 2 O

P 2 θ 5

Total

NiCrZnGaRbSrYZrNbBaLaCeNdPbTh

Zr/NbTi/ZrY/ZrCe/ZrBa/Zr(Ce/Y)N

Fe/MgK/RbBa/Sr

QOrAbAnNeDiHyOlMt11Ap

Hole 457

4,CC

53.40.89

15.611.320.145.336.472.731.130.11

97.09

1729801618

3051981

< l255

1617125

< l

>81.066.0

0.230.213.152.202.46

521.00.84

6.46.9

23.827.7

0.04.0

26.30.02.01.70.3

9,CC,74-76

52.50.87

15.212.730.135.038.243.060.910.05

98.75

22519818

8122

1748

< l211110582

>48.0108.0

0.350.210.441.442.93

940.00.17

2.35.4

26.225.5

0.013.122.4

0.02.21.70.1

Hole 460

9,CC,11-13

53.61.34

16.610.570.073.975.455.511.680.23

99.00

25

912522

1545283

166

1216118

< l

83.096.0

0.630.190.800.763.09

634.00.43

0.010.045.815.70.78.50.0

13.41.92.60.5

11-1,29-31

55.50.61

13.312.280.117.955.233.840.570.04

99.40

161059164

1041538

< l37109551

>38.O97.0

0.390.240.971.471.79

1185.00.36

2.63.4

32.717.40.06.9

32.50.02.11.20.1

Hole 461

3-1,60-62

58.71.01

13.214.060.135.463.003.500.220.08

99.43

271560174

1131669

< l36117624

>69.088.0

0.230.100.521.072.99

465.00.32

15.21.3

29.814.50.00.0

31.40.02.51.90.2

saline and siliceous nature, derived from the subductedoceanic lithosphere and infiltrating the mantle wedgeabove the Benioff zone from which the lavas are subse-quently derived (Best, 1975; Hawkesworth et al., 1978;Saunders and Tarney, 1979; Saunders et al., 1980). Thisprocess would cause the mantle wedge above the sub-ducted plate to become enriched in the mobile elements(Cs, K, Rb, Pb, Ba, Th, U, La, and Sr) which can beleached to varying degrees from altered oceanic crustafter subduction. Because the altered oceanic crust ishydrated with sea water, hydrous fluids derived from itwould be enriched in 87Sr relative to radiogenic Nd (Ndcontent of sea water is very low). In addition, subductedsediments would also be expected to contribute certain

mobile elements and radiogenic Pb to the fluids (Sun,1980). Variations in the composition of the sedimentsand the proportions to which they contribute to fluidsemanating from subducted oceanic crust might explainthe geochemical diversity observed in arc magma typesas a whole (e.g., the large range of HYG element ratios,such as Ba/La). However, we note that recent Pb, Sr,and Nd isotopic data for the Mariana arc (Meijer, 1976;DePaolo and Wasserburg, 1977; Dixon and Batiza,1979; Stern, 1979) suggest that sediment contaminationof the mantle source is minimal (<2°Io).

The introduction of a hydrous fluid into the mantlewedge would lower the solidus and induce partialmelting under conditions of high PH2θ The HYG ele-ment variations measured in the active margin volcanicsof the west Pacific are certainly consistent with thismodel. In order to be able to explain the extremely lowTa and Nb abundances in active margin volcanics ingeneral (Wood et al., 1979b), it is necessary to invokethat these elements are held in a mineral phase in themantle during the partial melting (Wood, 1979; Saun-ders et al., 1980). Ta and Nb are strongly partitionedinto titanium mineral phases—for example, rutile andsphene. Recent experimental studies have shown thatthe stabilities of rutile and sphene are enhanced underhydrous conditions and may be refractory phases witheven high degrees of melting of mafic material (Hellmanand Green, 1979).

The major and trace element compositions of theMg-rich andesites from Site 458 and boninite lavas can,in general terms, be reconciled with this model. Theirmajor element compositions are similar to liquids pro-duced experimentally by high degrees of partial meltingof a hydrous peridotite at low pressures, ~ 10 kbars(Green, 1976). They have generally been interpreted asnear-primary magmas derived by greater than 20 per-cent partial melting of a hydrous mantle source atdepths less than 30 km (Sun and Nesbitt, 1978). The lowabsolute abundances of the immobile HYG elements inthese lavas support a model invoking high degrees ofpartial melting. The absence of a negative Ta anomalyin the Site 458 andesites could reflect the total consump-tion of mantle titanium phases during high degrees ofpartial melting.

The Mg-rich andesites do show compositional rela-tionships with other arc magma types for example, en-richment in volatile and mobile elements—yet there aresome significant differences which suggest that they arenot derived from the same mantle sources. For example,the high Zr/Ti and Ti/Y ratios of the Mg-rich andesitesare similar to those observed in calc-alkaline magmasand probably indicate the involvement of garnet in theirgenesis; yet they are depleted in the light REE, whereasthe calc-alkaline magmas are enriched. However, wenote that previously described boninites are light-REE-enriched, and the Site 458 Mg-rich andesites are lessdepleted in the light REE than the island arc tholeiiteswith which they are associated. The Mg-rich andesitesof Site 458 differ from primitive island arc tholeiites inbeing less depleted in the light REE and in having highZr/Ti and Ti/Y ratios. Overall the Mg-rich andesites

624

GEOCHEMISTRY OF IGNEOUS ROCKS

10

1 5

10

"1 10

1.0

0.5

456-1 (CaCo3?)

_ >454A-PM1

454A-PM2

456-1

a 456A-1

456A-3 ©456A-2454A-PM3

456A-4

456-3

-454A-PM1 & 2

S 456A-1

456-1

456-1

456A-3

B456A-1

454A-PM1 & 2456-3

454A-PM3 Φ 456A-2

456-3 •> 456-1

- 4 5 4 A - P M 3

B456A-1

456A-4

454A-PM2

• 454A-PM1

456A-2

456A-3

B456A-1

456 A-4^ V t454A-PM3 Φ456A-2 V N454A-PM1

*l— 456-3

456A-3/?*456-1

B456A-1

• Hole454A

* Hole 456+ Hole 456A

a Unit 456A-1« Unit 456A-2

Δ Hole 457•fr Average N-type Morb

454A-PM3

456-3

456A-3

10MgO (wt. %)

15

454A-PM1

454A-PM2

ffl456A-1

456-1

10MgO (wt.'

15

0.2

100

90

80

70

60

300

200 J

100

400

300

200

100

20

Figure 6. MgO variation diagram of selected major and trace elements for samples from Sites 454 and 456.

625

D. A. WOOD ET AL.

0.8

0.6

- 0.4O

0.2

20

13 /fθO98

40 60Zr (ppm)

80 100

Figure 7. K2O, Ce, and Sr versus Zr for samples from Sites 454, 456,and 457. Average N-type MORB from Wood (1979). (See Fig. 6for legend.)

have more compositional similarities with calc-alkalinemagma types than primitive island arc tholeiites. Thereare two important features of Mg-rich andesites andboninites which need to be considered when developingpetrogenetic models for them: (1) They appear to belimited to the fore-arc of the West Pacific active marginand associated with the early development of an islandarc system. (2) Their chemical compositions apparentlyresult from high degrees of partial melting in the mantle

0.20

— 0.1

0.8

"c 0.6

0.4

0.2

-s 4

456A-4

454A-PM1 456A-3

456-1

H f

Figure 8. Ta, Th, and La versus Tb for samples from Sites 454, 456,and 457. (See Fig. 6 for legend.)

source under hydrous conditions, which suggests certainthermal energy requirements.

The source of the Mg-rich andesites could be in themantle wedge where the addition of excess slab-derivedhydrous fluids temporarily suppresses its solidus belowthe geotherm and initiated extensive partial melting. Asthe descending slab tends to cool the mantle wedge assubduction continues, the solidus of the mantle wedgewould cease to be significantly depressed below thegeotherm. The Site 458 Mg-rich andesites overlie island-arc tholeiitic lavas which are more depleted in the lightREE. The arc tholeiites could be generated during theinitial stages of hydrous metasomatism of the previouslydepleted mantle wedge by small degrees of partial melt-ing. More extensive metasomatism would enrich thesource in a number of HYG elements as well as depress-ing its solidus and increasing the degree of partialmelting so as to produce boninite-like magmas. As

626

GEOCHEMISTRY OF IGNEOUS ROCKS

400

300

200

100

454A-PM1

454A-PM3

456A-4

20 40 60Zr (ppm)

80 100

Figure 9. Ni versus Zr for samples from Sites 454, 456, and 457.(See Fig. 6 for legend.)

600

-c 400

200

454A-PM3

40 100

Figure 10. Cr versus Zr for samples from Sites 454, 456, and 457.(See Fig. 6 for legend.)

10000

8000-

6000 -

4000 -

2000-

fiy // ×/ ×

//•

Average / x

N-type ^MO RB/

m/ 456-1

/'g£^*^456-3^

/454A-<4^á' ^ */y PM3 456A-3

• •̂ ''

-

20 40 60Zr (ppm)

80

Figure 11. Ti versus Zr for samples from Sites 454, 456, and 457.(See Fig. 6 for legend.)

100

50

10

~\ 1 1 r -i—i 1—i—i 1 1—i—i r

Sample 456A-13-1, 23-26 cm ×Sample 456-16-2, 92-96 cm *

Sample 457-4, CC o

J L J L J L I I LCs Rb Ba Th U K Ta Nb La Ce Sr Nd P Hf Zr Eu Ti Tb Y

Figure 12. HYG element abundances of selected samples from Sites456 and 457 normalized to estimated primordial mantle abun-dances (from Wood, 1979, except that a Ti value of 1500 ppm isused here). Continuous line is average N-type MORB.

metasomatism continued to cool and enrich the mantlewedge, the degree of partial melting would decrease andthe magmas would develop a more calc-alkaline com-position.

Alternatively, the difference between calc-alkalineand primitive island arc tholeiites could result from themantle sources of the latter being metasomatized byhydrous fluids driven off the subducted slab at lowpressures (perhaps during the prograde metamorphicreaction of greenschist to amphibolite), and the formerfrom material infiltrated (veined) by hydrous melts (orfluids) derived from the slab at higher pressures (per-haps during the prograde metamorphic reaction am-phibolite to eclogite). Ringwood (1974) proposed theinvolvement of a liquid component derived by partial

627

D. A. WOOD ET AL.

Th

Hf/3

1 = Japan calc-alkali2 = DSDP Site 4513 = DSDP Site 4484 = Japan arc tholeiite5 = DSDP Sites 458-4616 = Mariana arc

dredge rocks

• = DSDP Site 454

o = DSDP Site 456

= DSDP Site 457(basaltic clast)

Ta

Figure 13. Th-Hf/3-Ta discrimination diagram (Wood et al., 1979b; Wood, 1980) for samples from Sites454, 456, and 457. Areas: A = N-type MORB; B = E-type MORB; C = within plate lavas; D = ac-tive plate margin lavas. The fields of Japanese lava series and DSDP Leg 59 samples are from Wood etal., 1980a, 1980b.

melting of quartz eclogite in the genesis of calc-alkalinemagmas. The partial dehydration of the slab at lowpressures would leach much of the radiogenic Sr addedto the ocean crust during alteration, so that the residueinvolved in the amphibolite to eclogite transformationcould be isotopically similar to MORB. This might ex-plain the low 87Sr/86Sr (~ 0.7032-0.7034) of the Mari-ana arc calc-alkaline lavas (Dixon and Batiza, 1979).

The most recent models for the tectonic evolution ofthe Mariana type plate boundary require that the posi-tion of the trench remains approximately fixed relativeto the mantle. Poehls (1978) proposed that the MarianaTrough and other back-arc basins are caused by inter-action between the subducted plate and the plate behindthe trench in places where trenches are terminated oroffset due to variations in subduction rates along theplate margin. He envisaged the slowing of the subduc-tion rate where aseismic ridges (in this case the CarolineRidge) collide with island arcs (Vogt et al., 1976). In thismodel the strike-slip motions occurring in the southernMarianas have induced the opening of the MarianaTrough, but the Mariana Trench has remained fixed. Intheir recent model, Uyeda and Kanamori (1979) proposethat it is the westward motion of the Eurasian plate inthe Philippine Sea area that has caused back arc exten-sion in the area. In this model the downgoing slab is an-chored to the mantle, and therefore the position of thetrench remains fixed relative to the mantle.

Mattey et al. (1980) have pointed out that if thetrench remains fixed relative to the mantle, the samemantle wedge would remain coupled to the Benioff zoneand be repeatedly infiltrated with fluids or melts derivedfrom the subducted plate. Moreover, these models im-ply that the Kyushu-Palau Ridge and West MarianaRidge would have formed in the same position relativeto the trench and Benioff zone as the active Marianaarc. The Parece Vela Basin (Fig. 1) was the first back-arc basin to open as a response to westerly movement ofthe West Philippine Basin or a strike-slip movement inthe regions of the Palau and Yap Trenches. The arcmagmatism became more calc-alkaline, forming theWest Mariana Ridge followed by back-arc extension inthe Mariana Trough.

Considered in these terms the observed temporalvariations in the geochemistry of the lavas erupted in theMariana system may be able to constrain some of thepetrogenetic models: the lavas from the Mariana fore-arc and Kyushu-Palau Ridge are primitive arc tholeiites(although we note that there are some systematic dif-ferences in the trace element ratios of lavas from the tworegions, e.g., La/Ta, Fig. 24); lavas from the youngerWest Mariana Ridge and active Mariana arc are calc-alkaline, as are some of the lavas in the MarianaTrough. Thus, there is a trend from early primitiveisland-arc tholeiites to subsequent calc-alkaline lavaserupted in the Mariana system. These variations suggest

628

GEOCHEMISTRY OF IGNEOUS ROCKS

Table 6. Major and trace element analyses of Recent basalts and andesites dredged from a seamount (latitude 21 °57'N, longitude 143°27'E) inthe northern Mariana arc.

SiO 2

Tiθ2A1 2O 3

tFe2θ3MnOMgOCaON a 2 OK 2 O

P2O5

Total

NiCrZnGaRbSrYZrNbBaLaCeNdPbTh

Zr/NbTi/ZrY/ZrCe/ZrBa/Zr(Ce/Y)N

Fe/MgK/RbBa/Sr

QOrAbAnNeDiHyOlMt11Ap

AndesitesMV1801

53.10.82

19.18.510.132.75

10.492.570.920.12

98.51

413691720

4571762

< l331

10251344

>62.079.0

0.270.405.343.613.59

382.00.72

6.65.5

22.138.5

0.011.611.70.01.51.60.3

MV1802

51.50.74

18.38.480.134.03

10.712.341.200.15

97.49

36104631828

4771976

< l283

13301866

>76.058.00.250.393.723.882.44

326.00.59

3.96.7

20.337.1

0.013.814.20.01.51.40.4

MV1803

52.80.80

18.68.760.142.91

10.482.520.580.11

97.66

413691812

5011957

< l307

11261483

>57.084.0

0.330.465.393.363.49

401.00.61

7.53.5

21.838.6

0.011.912.50.01.61.60.3

MV1804

54.20.82

19.78.690.132.61

10.392.671.130.18

100.51

610651824

4952068

2337

112816105

34.072.0

0.290.414.963.443.86

391.00.68

6.36.6

22.538.2

0.010.212.00.01.51.50.4

MV15101

55.31.13

15.010.910.182.897.892.731.210.15

97.39

11196

825

3872882

2481

153820128

41.083.00.340.465.873.334.38

402.01.24

11.07.3

23.725.7

0.011.615.10.01.92.20.4

MV15117

60.61.11

14.09.910.191.985.323.641.470.24

98.47

< l< l110

1432

35840

1052

633194425

76

52.063.0

0.380.426.032.705.80

381.01.77

16.58.8

31.317.70.06.6

13.80.01.82.10.6

MV15146

56.71.28

13.811.430.182.866.993.081.290.15

97.81

< l6

991729

3403398

3536

173220109

33.078.0

0.340.335.472.384.63

369.01.58

12.07.8

26.620.5

0.012.015.20.02.02.50.4

MV15188

52.70.79

18.79.380.153.58

10.822.510.890.11

99.61

925741719

4821959

2327

11251376

29.080.00.320.425.543.233.04

389.00.68

4.55.3

21.337.3

0.013.414.00.01.61.50.3

MV15197

56.91.13

15.110.120.162.617.262.921.410.18

97.73

< l9

981637

37031

1053

555174322

87

35.065.0

0.300.415.293.414.50

316.01.50

12.68.5

25.324.4

0.09.7

14.20.01.82.20.4

MV15220

55.31.14

14.811.040.183.178.022.641.200.14

97.56

310991426

3952982

< l476

173720

97

>82.083.00.350.455.803.134.04

383.01.21

10.97.3

22.925.5

0.012.215.70.02.02.20.3

MV15214

55.50.92

16.69.130.152.367.932.711.220.13

96.73

< l4

911529

4142787

< l509

153419106

>87.063.0

0.310.395.853.094.49

349.01.23

12.17.5

23.730.7

0.07.9

13.60.01.61.80.3

MV15268

53.00.88

18.69.820.153.21

10.562.561.020.14

99.99

411921723

5132065

1340

13321695

65.081.00.310.495.233.933.55

368.00.66

4.96.0

21.736.4

0.012.913.60.01.71.70.3

MV15295

55.40.88

16.99.900.173.929.582.501.010.12

100.37

1338821423

3952466

< l388

11271658

>66.080.00.360.415.882.762.93

365.00.98

8.55.9

21.131.8

0.012.515.70.01.71.70.3

that the nature of the mantle wedge and/or the petro-genetic processes have changed with time. One possibil-ity is that the mantle wedge has become more refractory(i.e., more Mg-rich) and depleted in the high-field-strength trace elements, despite being repeatedly re-plenished by a volatile (lithophile element-rich) fluidor melt phase. In this case, the primitive island-arctholeiites could represent early melting of the mantlewedge, having essentially the composition of the N-typeMORB mantle reservoir enriched in the volatile ele-ments. Such mantle would melt to similar or slightlyhigher degrees than that feeding mid-ocean ridges. Thecalc-alkaline magmas would then represent partial meltsderived from the re-enriched mantle residue of primitive

arc-tholeiite magmatism. Such a refractory mantle isunlikely to melt to such high degrees, and the HYG ele-ment composition of the liquids produced would bemuch more influenced by the composition of the vola-tile phase producing the re-enrichment. In this modelthe proportion of the minor titanium phases (sphene orrutile) in the mantle wedge would increase with time;this, together, with a general decrease in the degree ofpartial melting could explain a general increase in themagnitude of the negative Ta anomaly with time: thetholeiites from the fore-arc are less depleted in Ta thanthe other arc lavas (Figs. 22 and 24). The mantleunderlying the fore-arc at the present has been cooledsufficiently to preclude subsequent magmatism, and re-

629

D. A. WOOD ET AL.

Table 6. (Continued).

MV15297

58.00.95

15.49.620.182.527.083.211.200.18

98.40

210901326

3643082

< l500

153720

97

>82.069.00.370.456.103.034.43

383.01.37

13.27.2

27.624.60.08.6

14.00.01.71.80.4

MV15304

54.90.89

16.79.870.163.829.402.471.040.13

99.31

1334951626

3922270

< l395

14291688

>70.076.00.310.415.643.243.00

332.01.018.66.2

21.031.50.0

12.415.70.01.71.70.3

MV15344

55.11.00

17.09.830.152.818.922.701.030.12

98.70

411851623

3742477

< l414

123316

< l2

>77.078.00.310.435.383.384.06

372.01.119.86.2

23.131.60.0

10.613.80.01.71.90.3

MV1502

54.40.71

20.58.000.123.17

11.132.660.690.16

101.53

< l7

581735

5091755

< l215

9241245

>55.O77.00.310.443.913.472.93

164.00.426.24.0

22.241.3

0.010.212.40.01.41.30.4

MV1504

52.30.80

19.28.750.132.87

10.272.391.300.17

98.22

38

702035

5682290

< l339

184021

47

>90.053.00.240.443.774.473.54

308.00.605.37.8

20.638.60.0

10.413.00.01.51.50.4

MV1506

59.60.98

14.99.720.182.636.383.271.330.17

99.20

< l3

981527

3553591

< l562

163922

54

>91.065.00.380.436.182.744.29

409.01.58

14.87.9

27.922.30.07.3

14.90.01.71.90.4

MV1508

57.31.14

13.710.530.172.906.812.851.310.14

96.86

36

921431

3243192

< l574

163821

86

>92.074.00.340.416.243.014.21

351.01.77

14.48.0

24.921.4

0.010.715.20.01.92.20.3

MV1515

54.81.26

14.89.790.143.548.722.171.490.17

96.85

1237821636

35931

107< l612

214623159

>107.071.00.290.435.723.643.21

344.01.70

11.59.1

19.027.10.0

13.914.00.01.82.50.4

MV1520

52.00.71

19.38.140.123.58

10.882.371.180.16

98.48

2865661933

5121977

< l265

12311853

>77.055.00.250.403.444.012.64

297.00.524.17.1

20.439.1

0.012.313.20.01.41.40.4

MV1526

55.40.88

18.67.990.113.449.673.501.170.22

100.97

311481918

4672071

2326

122915

< l7

35.074.00.280.414.593.562.69

540.00.704.16.8

29.331.20.0

12.511.70.01.41.70.5

MV1580

59.71.10

13.49.850.192.005.143.491.480.22

96.58

< l< l1211632

35039

103< l644

214124

610

> 103.064.00.380.406.252.585.71

384.01.84

17.19.1

30.617.20.06.8

14.00.01.82.20.5

MV17

54.00.98

15.410.880.183.288.372.561.320.16

97.18

18

971732

43223832

394173819148

41.071.00.280.464.754.063.85

342.00.918.78.0

22.327.50.0

12.116.20.01.91.90.4

MV15132

60.51.11

13.99.820.191.905.213.671.480.23

98.02

< l< l1121531

35038

104< l644

204525

76

> 104.064.00.370.436.192.915.99

396.01.84

16.78.9

31.717.30.06.6

13.50.01.72.20.6

MV1517

52.50.93

15.710.510.172.858.702.661.260.15

95.42

25

1041833

4302481

< l398

16341866

>81.069.00.300.424.913.484.28

317.00.937.37.8

23.628.50.0

13.514.30.01.91.90.4

Table 7. INA trace element analyses (ppm) of 15 andesites from the Mariana arc.

SampleNumber

RD17MV1502MV1504MV15O8MV1515MV1520MV1526MV1580MV15101MV15117MV15132MV15146MV15220MV180fMV1802

Se

29.823.823.427.228.423.628.724.430.324.824.632.231.424.524.7

Cr

6047495374994947624847666553

134

Co

27.219.520.120.621.623.225.516.426.716.416.627.726.922.325.0

Ni

8678

1636

80.5800.799

1046

Rb

30.429.630.625.631.030.415.130.825.329.329.927.023.518.425.7

Sr

431502551307355508441343392349339340389468487

Zr

9266

10888

13410694

13611412611289967093

Cs

0.681.730.590.82

• 1.050.470.120.890.630.860.870.780.680.480.49

Ba

4042.19371557627298302626462613626527482322278

La

17.49.4

19.416.620.215.714.020.515.219.519.916.416.111.514.6

Ce

29.916.230.627.9 :32.8 :27.722.2 :33.6 :30.9 i33.2 ;33.527.7 :27.6 :18.6 :25.6

Eu

1.381.141.451.341.371.451.321.691.361.721.691.491.401.141.42

Tb

0.690.440.630.780.770.580.580.990.700.960.970.830.730.490.56

Hf

2.161.272.192.682.872.151.793.022.403.032.992.692.421.761.91

Ta

0.110.060.110.140.170.110.090.150.130.150.160.130.120.090.08

Th

2.671.232.662.713.612.001.743.082.463.033.022.652.551.751.76

U

0.760.390.830.781.000.640.610.860.620.870.860.670.720.490.58

La/Ta

158157176119119143156137117130124126134128183

La/Th

6.57.67.36.15.67.98.06.76.26.46.66.26.36.68.3

La/Tb

25.221.430.821.326.227.124.120.721.720.320.519.822.123.526.1

TVHf

1.21.0.];

(:•

:

:

:

:

:

:

1.21.01.3).9[.01.0LOLOLOLOLI1.00.9

Ba/La

23.223.319.133.631.019.021.630.530.431.431.532.129.928.019.0

La/Ce

0.580.580.630.600.620.570.630.610.490.590.590.590.580.620.57

Note: Samples from P. Fryer; sample designations as in Table 6.

630

GEOCHEMISTRY OF IGNEOUS ROCKS

Magnetic Inclination) ) Zr (ppm) T1O2 (wt. %)

- 9 0 0 +90 40 60 80 100 0.4 0.6 0.8 1.0

Sr (ppm)

90 110 130 150

r. 1 , GeochemicalCr (ppm) U n j t

50 100 150 200 250

3710-

3730-

3750-

3770-

3790-

3810-

3830-

3850-

3870-

3890-

3910-

Sandstone,Siltstone, VitricTuff

Unit 1Vesicular, GlassyTwo PyroxenePillow Basalts|_imestoneInterbeds

Unit 2

Two Massive Flowsof Two PyroxeneBasalt. BrecciatedShear Zones ofCarbonate andClay linedFractures

Unit 3Thin, VesicularGlassy, TwoPyroxeneBasalt Flows

Unit 4Heavily AlteredVesicular,CPX-PlagFlow and PillowBasalts

Unit 5MassiveFine-GrainedAugite PlagioclaseBasalts

4a

Figure 14. Geochemical and Hthological stratigraphy of the basement sections of Hole 458.

cent arc magmatism has become restricted to a zoneabove the steeply dipping Benioff zone.

The possible coexistence of N-type MORB and calc-alkaline magmas in the initial period of inter-arc basinformation, as observed in the Mariana Trough (Sites454 and 456), might be expected from this model. Thegeneral absence of basalts with calc-alkaline affinities inthe Parece-Vela Basin and Shikoku Basin implies thatthe influence of the arc mantle sources disappears oncethe basins are well developed.

ACKNOWLEDGMENTS

The analytical work was carried out with financial support fromthe Natural Environment Research Council (NERC), U.K. and theCentre National de la Recherche Scientifique, France. DAW andNGM gratefully acknowledge NERC fellowships supporting work onsamples recovered by DSDP. We thank'Drs. J. H. Natland, J. A.Pearce, and A. D. Saunders for their reviews and comments on themanuscript.

REFERENCES

Abbotts, I. L., 1979. A field petrological and geochemical study of theMasirah Ophiolite, Oman [Ph.D. dissert.]. University of Birming-ham, U.K.

Anonymous, 1977. Initial report of the geological study of oceaniccrust of the Philippine Sea floor. Ofioliti, 2:137-168.

Barazangi, M., Pennington, W., and Isacks, B. L., 1975. Global studyof seismic wave attenuation in the upper mantle behind island arcsusing P waves. J. Geophys. Res., 80:1078-1092.

Barker, P. F., 1972. A spreading centre in the East Scotia Sea. EarthPlanet. Sci. Lett., 15:123-321.

Best, M. G., 1975. Migration of hydrous fluids in the upper man-tle and potassium variation in calc-alkalic rocks. Geology, 3:429-432.

Cameron, W. E., Nisbet, E. G., and Dietrich, V. J., 1979. Boninites,komatiites and ophiolitic basalts. Nature, 280:550-553.

Chayla, B., Jaffrezic, H., and Joron, J.-L., 1973. Analyse par activa-tion dans les mentions épithermiques. Application à la determina-tion d'elements en trace dans les roches. Comptes Rendus Acad.Sci. Paris, 277:273-275.

Chow, T. J., Stern, R. J., and Dixon, T. H., 1980. Absolute and rela-tive abundances of K, Rb, Sr and Ba in circum-Pacific island-arcmagmas, with special reference to the Marianas. Chem. Geol.,28:111-121.

Cloud, P. E., Jr., Schmidt, R. G., and Burke, H. W., 1956. Geologyof Saipan, Mariana Islands. U.S. Geol. Surv. Prof. Paper, 280A.

Dallwitz, W. B., 1968. Chemical composition and genesis of clino-enstatite-bearing volcanic rocks from Cape Vogel, Papua: a dis-cussion. Rept. 23rd Int. Geol. Congr., 2:229-242.

Dallwitz, W. B., Green, D. H., and Thompson, I. E., 1966. Clinoen-statite in a volcanic rock from Cape Vogel, Papua. J. Petrol., 7:375-403.

DePaolo, D. J., and Wasserburg, C. J., 1977. The sources of islandarcs as indicated by Nd and Sr isotopic studies. Geophys. Res.Lett., 4:465-468.

Dietrich, V., Emmermann, R., Oberhansli, R., et al., 1978. Geochem-istry of basaltic and gabbroic rocks from the west Mariana basinand the Mariana Trench. Earth Planet. Sci. Lett., 39:127-144.

Dixon, T. H., and Batiza, R., 1979. Petrology and chemistry of recentlavas in the Northern Marianas: Implications for the origin ofisland arc basalts. Contrib. Mineral. Petrol., 70:167-181.

631

D. A. WOOD ET AL.

Garcia, M. O., Liu, N. W. K., and Muenow, D. W., 1979. Volatiles insubmarine volcanic rocks from the Mariana Island arc and trough.Geochim. Cosmochim. Acta, 43:305-312.

Gill, J. B., 1976. Composition and age of Lau Basin and Ridge vol-canic rocks: implications for evolution of an interarc basin andremnant arc. Geol. Soc. Am. Bull., 87:1384-1395.

Green, D. H., 1976. Experimental testing of "equilibrium" partialmelting of peridotite under water-saturated, high pressure condi-tions. Can. Mineralogists, 14:255-268.

Hart, S. R., Glassley, W. E., and Karig, D. E., 1972. Basalts and seafloor spreading behind the Mariana Island arc. Earth Planet. Sci.Lett., 15:12-18.

Hawkesworth, C. J., O'Nions, R. K., Pankhurst, R. J., et al., 1978.A geochemical study of island arc and back-arc tholeiites from theScotia Sea. Earth Planet. Sci. Lett., 36:253-262.

Hawkins, J. W., 1974. Geology of the Lau Basin, a marginal basinbehind the Tonga Arc. In Burke, C. A., and Drake, C. L. (Eds.),The Geology of Continental Margins: New York (Springer), pp.505-520.

, 1976. Petrology and geochemistry of basaltic rocks of theLau Basin. Earth Planet. Sci. Lett., 28:283-297.

., 1977. Petrology and geochemical characteristics of mar-ginal basin basalts. In Talwani, M., and Pitman III, W. C. (Eds.),Island Arcs, Deep Sea Trenches and Back-Arc Basins: MauriceEwing Series 1: Washington (American Geophysical Union),355-365.

Hellman, P. L., and Green, T. H., 1979. The role of sphene as an ac-cessory phase in the high pressure partial melting of hydrous maficcompositions. Earth Planet. Sci. Lett., 42:191-201.

Hickey, R., and Frey, F. A., 1979. Petrogenesis of high-Mg andesites:Geochemical evidence. EOS, 60:413.

Hussong, D. H., Uyeda, S., et al., 1978. Near the Philippines—Leg60 ends in Guam. Geotimes, 23(10): 19-22.

Isacks, B. L., and Barazangi, M., 1977. Geometry of Benioff zones:Lateral segmentation and downwards bending of the subductedlithosphere. In Talwani, M., and Pitman III, W. C. (Eds.), IslandArcs, Deep Sea Trenches and Back-Arc Basins: Maurice EwingSeries I: Washington (American Geophysical Union).

Karig, D. E., 1971. Origin and development of marginal basins in thewestern Pacific. J. Geophys. Res., 76:2542-2561.

Karig, D. E., Anderson, R. N., and Bibee, L. D., 1978. Characteristicsof back arc spreading in the Mariana Trough. J. Geophys. Res.,83:1213-1226.

Katsumata, M., and Sykes, L. R., 1969. Seismicity and tectonics ofthe western Pacific: Izu-Mariana-Caroline and Ryuku-Taiwanregion. J. Geophys. Res., 74:5923-5948.

Kuno, H., 1968. Differentiation of basalt magmas. In Hess, H. H.,and Poldervaart, A. (Eds.), Basalts (Vol. 2): New York (Wiley),623-688.

Kuroda, M., and Shiraki, K., 1975. Boninite and related rocks of Chi-chi-Jima, Bonin Islands. Japan Rep. Fac. Sci. Shizuoka Univ.,10:145-155.

Kuroda, M., Shiraki, K., and Urano, H., 1978. Boninite as a possiblecalc-alkalic primary magma. Proc. Tokyo Conf., pp. 280-281.

Marsh, N. G., 1977. The geochemistry of plutonic rocks from Ata-cama Province, North Chile [M.Sc. thesis]. University of Birming-ham, U.K.

Marsh, N. G., Saunders, A. D., and Tarney, J., 1980. Geochemistryof basalts from the Shikoku and Daito Basins, DSDP Leg 58. InKlein, G. deV., Kobayashi, K., et al., Init. Repts. DSDP, 58:Washington (U.S. Govt. Printing Office), 805-842.

Mattey, D. P., Marsh, N. G., and Tarney, J., 1980. The geochem-istry, mineralogy and petrology of igneous rocks from the WestPhilippine and Parece Vela Basins, and from the Kyushu-Palauand West Mariana Ridges, IPOD Leg 59. In Kroenke, L., Scott,R., et al., Init. Repts. DSDP, 59: Washington (U.S. Govt. Print-ing Office), 753-800.

Meijer, A., 1976. Pb and Sr isotopic data bearing on theorigin of vol-canic rocks from the Mariana island arc system. Bull. Geol. Soc.Am., 87:1358-1369.

Moore, J., and Schilling, J-G., 1973. Vesicles, water and sulfur inRejkjanes Ridge basalts. Contrib. Mineral. Petrol., 41:105-118.

Murauchi, S., Den, N., Asaro, S., et al., 1968. Crustal structure of thePhilippine Sea. J. Geophys. Res., 73:3143-3171.

Pearce, J. A., in press. Geochemical evidence for the genesis anderuptive setting of lavas from Tethyan Ophiolites. Proc. Symp, onOphiolite Problems. Geol. Survey Cyprus.

Philpotts, J. A., Martin, W., and Schnetzler, C. C , 1971. Geochem-ical aspects of some Japanese lavas. Earth Planet. Sci. Lett.,12:89-95.

Poehls, K. A., 1978. Inter-arc basins: a kinematic model. Geophys.Res. Lett., 5:325-328.

Rhodes, J. M., Dungan, M. A., Blanchard, D. P., et al., 1979. Magmamixing at mid-ocean ridges: evidence from basalts drilled near22°N on the mid-Atlantic Ridge. Tectonophysics, 55:35-61.

Ringwood, A. E., 1974. The petrological evolution of island arc sys-tems. J. Geol. Soc. London, 130:183-204.

Saunders, A. D., and Tarney, J., 1979. The geochemistry of basaltsfrom a back-arc spreading centre in the East Scotia Sea. Geochim.Cosmochim. Acta, 43:555-572.

Saunders, A. D., Tarney, J., and Weaver, S. D., 1980. Transversegeochemical variations across the Antarctic Peninsula: Implica-tions for the genesis of calc-alkalic magmas. Earth Planet. Sci.Lett., 46:344-360.

Schmidt, R. G., 1957. Geology of Saipan, Mariana Islands, ChapterB. Petrology of the volcanic rocks. U.S. Geol. Surv. Prof. Paper,28OB:127-175.

Sclater, J. G., Hawkins, J. W., Mammerickx, J., et al., 1972. Crustalextension between the Tonga and Low Ridges: petrologic and geo-physical evidence. Bull. Geol. Soc. Am., 83:505-517.

Scott, R. B., Kroenke, L., et al., 1980. Init. Repts. DSDP, 59:Washington (U.S. Govt. Printing Office).

Sharaskin, A. Ya, Dobretsov, N. L., and Sobolev, N. V., in press.Marianites: the clinoenstatite bearing pillow lavas associated withophiolite assemblage of the Mariana Trench. Proc. Int. OphioliteSymp. Cyprus.

Shiraki, K., Kuroda, N., and Urano, H., 1979. Clinoenstative-bear-ing boninite of Muko-jima, Bonin Islands. J. Geol. Soc. Japan,85:591-594.

Stark, J. T., 1963. Petrology of the volcanic rocks of Guam. U.S.Geol. Surv. Prof. Paper. 403C, 32.

Stern, R. J., 1979. On the origin of andesite in the northern MarianaIsland Arc: Implications from Agrigan. Contrib. Mineral. Petrol.,68:207-219.

Sun, S-S., 1980. Lead isotopic study of young volcanic rocks frommid-ocean ridges, ocean islands and island arcs. Philos. Trans.Roy. Soc. Lond. Ser. A., 297:409-445.

Sun, S-S., and Nesbitt, R. W., 1978. Geochemical regularities andgenetic significance of ophiolitic basalts. Geology, 6:689-693.

Sun, S-S., Nesbitt, R. W., and Sharaskin, A. Y., 1979. Chemicalcharacteristics of mid-ocean ridge basalts. Earth Planet. Sci. Lett.,44:119-138.

Tarney, J., Saunders, A. D., and Weaver, S. D., 1977. Geochemistryof volcanic rocks from the island arcs and marginal basins ofScotia Arc region. In Talwani, M., and Pitman III, W. C. (Eds.),Island Arcs, Deep Sea Trenches and Back-Arc Basins: MauriceEwing Series, 1: Washington (American Geophysical Union),367-377.

Tarney, J., Saunders, A. D., Weaver, S. D., Donnellan, N. C. B., andHendry, G. L., 1978. Minor element chemistry of basalts from Leg49, North Atlantic Ocean. In Luyendyk, B. P., Cann, J. R., et al.,Init. Repts. DSDP, 49: Washington (U.S. Govt. Printing Office),567-691.

Tarney, J. Wood, D. A., Saunders, A. D., et al., 1980. Nature ofmantle heterogeneity in the North Atlantic: evidence from deep seadrilling. Philos. Trans. Roy. Soc. Lond. Ser. A., 297:179-202.

Taylor, S. R., Capp, A. C , Graham, A. L., et al., 1969. Trace ele-ment abundances in andesite 11. Saipan, Bougainville and Fiji.Contr. Mineral. Petrol., 23:1-26.

Treuil, M., and Varet, J., 1973. Critères volcanologiques, pétrolog-iques et géochimiques de la genèse et de la differentiation desmagmas basaltiques: exemple de 1'Afar. Bull. Soc. Geol. France,15:506-540.

Uyeda, S., and Kanamori, H., 1979. Back-arc opening and the modeof subduction. J. Geophys. Res., 84:1049-1061.

Vogt, P. R., Lowrie, A., Bracey, D. R., et al., 1976. Subduction ofaseismic ocean ridges: Effects on shape, seismicity, and othercharacteristics of consuming plate boundaries. Geol. Soc. Am.Special Paper 172.

632

GEOCHEMISTRY OF IGNEOUS ROCKS

Weaver, S. D., Saunders, A. D., Pankhurst, R. J., et al., 1979. A geo-chemical study of magmatism associated with the initial stages ofback-arc spreading: the Quaternary volcanics of Bransfield Strait,South Shetland Islands. Contr. Mineral. Petrol., 68:151-169.

Weaver, S. D., Sceal, J. S. C , and Gibson, I. L., 1972. Trace elementdata relevant to the origin of trachytic and pantelleritic lavas in theEast African Rift System. Contrib. Mineral. Petrology, 36:181-190.

Wells, P. 1978. The geochemistry of the Patagonia batholith between45°S and 46°S latitude [M.Sc. thesis]. University of Birmingham,U.K. (unpublished).

Wood, D. A., 1979. A variably veined sub-oceanic upper mantle: geneticsignificance for mid-ocean ridge basalts from geochemical evidence.Geology, 7:499-503.

, 1980. The application of a Th-Hf-Ta diagram to problemsof tectonomagmatic classification and to establishing the nature ofcrustal contamination of basaltic lavas of the British TertiaryVolcanic Province. Earth Planet. Sci. Lett., 50:11-30.

Wood, D. A., Joron, J. L., Marsh, N. G., et al., 1980a. Major andtrace element variations in basalts from the North Philippine Seadrilled during DSDP Leg 58: A comparative study of back-arcbasin basalt with lava series from Japan and mid-ocean ridges. InKlein, G. deV., Kobayashi, K., et al., Init. Repts. DSDP, 58:Washington (U.S. Govt. Printing Office), 873-894.

Wood, D. A., Joron, J. L., and Treuil, M., 1979b. A re-appraisal ofthe use of trace elements to classify and discriminate betweenmagma series in different tectonic settings. Earth Planet. Sci.Lett., 45:326-336.

Wood, D. A., Mattey, D. P., Joron, J. L., et al., 1980a. A geo-chemical study of 17 selected samples from the basement coresrecovered at Sites 447, 448, 449, 450, and 451 DSDP Leg 59. InKroenke, L., Scott, R., et al., Init. Repts. DSDP, 59: Washington(U.S. Govt. Printing Office), 743-752.

Wood, D. A., Tarney, J., Varet, J., et al., 1979a. Geochemistry of ba-salts drilled in the North Atlantic by IPOD Leg 49: implicationsfor mantle heterogeneity. Earth Planet. Sci. Lett., 42:77-97.

633

D. A. WOOD ET AL.

Table 8. Major and trace element analyses of igneous rocks from Hole 458.

Sample(intervalin cm)

Siθ2TiO2

A12O3

tFe 2O 3

MnOMgOCaONa2OK2OP 2 O 5

Total

NiCrZnGaRbSrYZrNbBaLaCcNdPbTh

Zr/NbTi/ZrY/ZrCe/ZrBa/Zr(Ce/Y)N

Fe/MgK/RbBa/Sr

QOrAbAnNeDiHyOlMt1!Ap

27-1,130-132

54.60.157.9

11.920.42

16.351.832.671.640.00

97.45

3151303

575

2155

< 124

25911

< 1< l

3< 1

12.037.0

<0.04<0.04

2.462.460.85

648.01.07

0.09.9

23.24.80.03.6

49.35.62.10.30.0

28-1,58-60

52.50.34

18.09.310.115.09

11.172.770.630.03

99.97

80214

71147

1228

31< l41

66362

>31.066.0

0.260.191.321.842.12

747.00.34

1.33.7

23.434.90.0

16.816.60.01.60.60.1

28-1,125-127

52.80.34

18.38.550.115.35

11.322.670.960.02

100.41

79229

64157

1196

36< l22

7< 1

23

< l

>36.057.0

0.17<0.02

0.61

1.851141.0

0.18

0.85.7

22.535.00.0

17.216.00.01.50.60.1

28-1,140-142

56.10.28

13.79.320.146.119.591.970.440.02

97.70

7418954148

927

282

336733

< l

14.059.0

0.250.251.182.461.77

452.00.36

12.32.6

17.128.0

0.016.920.0

0.01.70.50.0

29-1,54-60

53.10.33

15.49.160.108.979.262.601.230.01

100.13

91244

741713

1063

31< 1

137532

< l

>31.065.0

0.100.160.424.091.18

785.00.12

0.07.3

22.026.6

0.015.622.2

3.31.60.60.0

29-2,37-39

52.00.36

14.410.230.129.718.702.630.880.01

99.01

76216

76155

1032

30< l

157224

< l

>30.071.0

0.070.070.502.461.22

1453.00.15

0.05.2

22.525.2

0.014.923.5

5.31.8

0.70.0

30-1,56-58

51.50.30

17.89.360.176.327.993.131.530.01

98.17

81201

559

1169

527

< l54

752

< l7

>27.066.0

0.190.192.002.461.72

1155.00.78

0.09.2

27.030.6

0.08.0

14.77.41.70.60.0

31-1,39-41

53.20.33

16.19.240.107.479.492.651.120.01

99.73

73245

65119

992

30< 1

6664

< 19

>30.067.0

0.070.200.207.371.43

1033.00.06

0.56.6

22.528.7

0.015.023.5

0.01.60.60.0

31-1,131-133

57.10.33

15.29.380.115.19

10.452.810.850.05

101.48

56168481446

1168

33< l35

6< l

222

>33.O59.0

0.24<0.03

1.06—

2.10154.0

0.30

7.35.0

23.425.9

0.020.614.70.01.60.60.1

32-1,116-118

58.80.30

14.29.320.105.029.602.361.200.04

100.94

59199501661

1066

30< 149

6211

< l

>30.060.0

0.200.071.630.822.15

163.00.46

11.87.0

19.824.3

0.018.715.30.01.60.60.1

32-3,34-36

59.00.30

13.99.340.114.629.242.501.160.04

100.24

591684814

113105

835

< l29

6< l< l

2< l

>35.O52.0

0.23<0.03

0.83—

2.3485.0

0.28

12.66.9

21.123.3

0.018.414.6

0.01.60.60.1

33-2,19-21

56.90.30

13.28.980.128.138.932.240.610.03

101.73

67191521513

1006

28< l51

64352

>28.O64.0

0.210.141.821.641.28

391.00.51

11.33.6

18.623.8

0.015.923.8

0.01.50.60.1

34-1,34-41

59.80.30

14.78.950.125.53

10.762.611.120.04

101.02

65200

521524

1079

30< 147

6434

< l

>30.060.0

0.300.131.571.091.88

387.00.44

6.96.6

21.924.8

0.022.914.00.01.50.60.1

35-1,42-44

53.30.30

13.88.750.115.239.742.330.830.04

100.97

60181

551413851236

< l197

< l1

< l< l

>36.O50.0

0.33

<0.030.53

—1.94

531.00.22

14.24.9

19.524.5

0.019.114.90.01.50.60.1

35-2,69-71

52.40.35

16.59.200.127.129.762.810.790.01

99.93

101282

76178

1243

39< l46

722

< l< 1

>39.O54.0

0.080.051.181.641.50

817.00.37

0.94.7

23.830.1

0.015.022.5

0.01.60.70.0

36-1,27-29

52.40.35

15.110.130.148.737.903.041.050.01

98.88

97267

73135

1075

34< l30

8532

< l

>34.062.0

0.150.150.88

—1.35

1743.00.28

0.06.3

26.024.8

0.012.121.7

5.81.80.70.0

37-1,52-54

0.3715.19.360.107.6310.132.871.380.05

99.40

621845716331191236< l545232

< l

>36.061.00.330.061.500.411.42

347.00.45

0.08.2

24.424.40.021.111.76.91.60.70.1

634

GEOCHEMISTRY OF IGNEOUS ROCKS

Table !

37-2,62-64

54.20.34

15.09.020.107.349.892.791.270.05

100.02

65185531834

1159

32< 144

6231

< l

>32.064.00.280.061.370.551.43

310.00.38

1.07.5

23.624.60.0

19.720.5

0.01.60.60.1

1. (Continued).

38-1,51-53

54.90.30

14.78.920.126.36

10.682.661.140.04

99.81

75219

531619

1098

31< l396442

< 1

>31.058.00.260.131.261.231.63

498.00.36

3.46.7

22.524.90.0

22.916.50.01.60.60.1

39-1,37-39

54.80.28

14.49.340.156.429.402.140.540.01

97.49

7117260127

866

29< l25

6< l< 1

3< 1

>29.059.00.21

<0.030.86—1.69

644.00.29

9.23.3

18.628.7

0.015.621.5

0.01.70.60.0

39-2,94-96

52.50.36

15.79.430.159.168.392.630.860.01

99.20

99217

51143

102< l35

< l204

< 1< 1

11< l

>35.O61.0

<0.03<0.03

0.572.461.19

2382.00.20

0.05.1

22.428.9

0.010.628.1

1.61.70.70.0

40-2,90-92

52.80.35

14.910.240.098.977.982.471.030.01

98.86

97290

831612

1042

38< l18752

< l< l

>38.055.00.050.130.476.141.32

713.00.17

0.76.2

21.126.9

0.010.731.0

0.01.80.70.0

41-1,22-24

53.21.13

15.59.660.063.736.125.701.520.20

96.83

1111882525

16931

1011

799

15116

< l

101.067.0

0.310.150.781.193.00

505.00.47

0.09.3

45.712.52.2

14.60.0

10.41.72.20.5

42-1,61-63

52.70.58

15.511.110.086.768.512.940.830.05

99.03

65176821815

1342058

< l197963

< l

>58.O60.0

0.340.160.331.111.91

462.00.14

1.05.0

25.126.9

0.012.725.1

0.02.01.10.1

42-1,86-88

54.80.49

13.610.370.166.128.352.540.570.02

96.98

7517059136

941144

< l42

5326

< l

>44.067.00.250.070.950.671.97

784.00.45

8.73.5

22.224.7

0.014.722.4

0.01.91.00.1

43-1,118-120

54.50.51

14.210.610.166.257.882.480.570.02

97.22

8215765127

941051

< l49

63411

>51.061.0

0.200.060.960.741.97

681.00.52

8.53.5

21.626.70.0

11.124.80.01.91.00.0

44-1,53-55

54.70.49

14.310.190.155.787.972.450.530.02

96.66

7616460138

1011144

< l296844

< 1

>44.066.00.250.180.661.792.04

550.00.29

9.93.2

21.427.50.0

11.023.2

0.01.81.00.1

45-1,8-10

54.10.49

12.910.290.155.648.202.540.510.02

94.81

7716760148

1001151

< 141

5< l

242

>51.057.00.22

<0.020.80_2.12

533.00.41

9.73.2

22.723.50.0

15.921.1

0.01.91.00.1

45-1,24-26

52.80.56

15.310.690.086.719.092.950.970.07

99.16

71165681719

1392547

< l80

996

< l< l

>47.072.00.530.191.700.881.85

425.00.58

0.55.8

25.225.8

0.015.822.9

0.01.91.10.2

46-1,75-77

52.81.13

13.613.820.106.175.883.410.750.06

97.74

2311

1091813

1442065

< l631010712

>65.0104.0

0.310.150.971.232.60

479.00.44

2.74.5

29.520.1

0.07.8

29.30.02.52.20.1

46-1,83-85

52.61.16

14.013.740.096.336.063.460.690.09

98.26

2810

1112012

1492668

< 143

9128

< l2

>68.0102.0

0.380.180.631.132.52

479.00.29

1.94.2

29.821.0

0.07.5

29.50.02.42.20.2

47-1,67-69

52.01.11

14.313.050.106.857.423.550.480.14

98.99

21109226

81512569

< 144

87546

>69.096.00.360.100.640.692.21

501.00.29

0.02.9

30.321.9

0.011.925.7

1.42.32.20.3

47-2,90-92

51.81.09

14.213.530.107.326.783.200.490.10

98.66

3516

10224

71383464

< l3711763

< l

>64.0102.0

0.530.110.580.512.14

575.00.27

0.72.9

27.423.3

0.08.4

31.30.02.42.10.2

48-1,79-81

52.21.07

13.813.880.165.516.652.590.720.02

96.58

241492159

1111759

< l777532

< 1

>59.0109.0

0.290.081.310.722.92

665.00.69

6.24.4

22.724.7

0.07.8

28.30.02.52.10.1

49-1,136-138

50.61.04

13.913.730.097.966.513.631.360.11

98.95

1913692120

1381947

< l83

8873

< l

>47.0133.0

0.400.171.771.032.00

564.00.60

0.08.1

31.017.90.0

11.58.6

16.82.42.00.3

635

D. A. WOOD ET AL.

Table 9. Representative major element analyses ofboninites selected from the literature.

SiO2

Tiθ2A12O3

Fe 2 O 3

FeOMnOMgOCaONa 2OK 2 O

P2O5H 2 O +

Total

1

53.760.24

12.530.636.480.16

11.936.441.900.680.065.13

99.94

2

56.880.28

13.761.596.410.225.447.992.170.490.074.40

99.70

3

56.460.29

14.655.073.060.136.928.192.691.280.052.42

101.21

4

55.560.21

10.252.855.860.18

13.575.911.510.690.033.65

100.29

5

54.090.308.393.656.540.15

13.035.460.750.410.077.4

100.20

Notes:1 = boninite, Bonin Islands, Kuroda et al. (1978).2 = bronzite andesite, Bonin Islands, Kuroda et al.

(1978).3 = 1403-24 contact rock, Mariana Trench, Dietrich

et al. (1978).4 = 1403-34 enstalite porphyritic boninite, Mariana

Trench, Dietrich et al. (1978).5 = LB105 clinoenstatite-bearing, high-Mg andes-

ite, Dabi Volcanics, Cape Vogel, Papua, Dall-witz et al. (1966).

Table 10. Major and trace element analyses of igneous rocks from Hole 459B.

Sample(intervalin cm)

Siθ2T1O2A12O3

tFe 2O 3

MnOMgOCaONa2OK2OP2O5Total

NiCrZnGaRbSrYZrNbBaLaCeNdPbTh

Zr/NbTi/ZrY/ZrCe/ZrBa/Zr(Ce/Y)N

Fe/MgK/RbBa/Sr

QOrAbAnNeDiHyOlMt11Ap

60-2,80-82

54.60.74

14.710.130.145.55

10.272.980.470.07

99.69

43131601627

1211538

< l38

7< l

2< l< l

>38.O117.0

0.39<0.03

1.00—

2.12144.0

0.31

5.02.8

25.325.5

0.020.816.40.01.81.40.2

61-1,115-117

53.80.66

14.99.900.146.40

10.652.960.440.07

99.97

57159651724

1201538

< l24

7321

< 1

>38.O104.0

0.390.080.630.491.79

153.00.20

2.72.6

25.126.1

0.021.617.90,01.71.30.2

61-2,47-49

54.40.67

14.29.830.145.40

11.532.670.540.07

99.48

54141601615

1161536

< l20

83321

>36.0111.0

0.420.080.560.492.11

301.00.17

5.43.2

22.725.3

0.026.213.10.01.71.30.2

63-1,53-55

51.10.83

14.810.020.149.378.773.300.680.08

99.12

686790186

1272053

< l195544

< l

>53.O94.0

0.380.090.360.611.24

938.00.15

0.04.0

28.223.8

0.015.911.212.5

1.81.60.2

64-1,41-43

51.30.82

13.811.310.149.908.173.320.570.09

99.38

52568722

51233752

< l86

5872

< l

>52.O95.0

0.710.151.650.531.32

955.00.70

0.03.4

28.321.2

0.015.514.412.52.01.60.2

65-1,59-61

52.61.21

13.913.630.166.076.063.730.890.06

98.35

2413

1152513

1252768

< 177

6111023

>68.0107.0

0.400.161.131.002.60

566.00.62

0.55.3

32.118.90.09.4

27.70.02.42.30.2

66-1,73-75

56.20.99

13.411.850.155.537.353.270.630.08

99.41

2417722228

1092456

< l33

6443

< l

>56.0106.0

0.430.070.590.412.49

186.00.30

7.83.7

27.820.1

0.013.422.0

0.02.11.90.2

66-2,147-149

56.50.98

13.211.980.154.807.383.120.860.08

99.03

2113701946

1072254

131

796

< l2

54.0109.0

0.410.170.571.002.89

156.00.29

9.35.1

26.719.50.0

14.120.10.02.11.90.2

67-1,118-120

57.40.93

12.811.840.144.957.023.141.450.08

99.70

2416

1071822

1092250

< 137124522

>50.0112.00

0.440.080.740.452.77

547.00.34

8.58.6

26.616.60.0

14.919.80.02.11.80.2

68-1,51-53

58.60.72

12.211.410.125.176.822.891.890.07

99.90

3022851726

1151946

< l38105551

>46.094.0

0.410.110.830.652.56

603.00.33

9.911.224.514.70.0

15.519.70.02.01.40.2

69-1,144-146

56.80.76

14.410.230.115.878.433.400.690.08

100.81

2823662012

1441951

< 128

97641

>51.089.0

0.370.140.550.902.02

477.00.19

6.14.0

28.521.9

0.015.719.50.01.81.40.2

69-3,83-85

56.00.94

14.311.180.115.177.933.831.000.10

100.59

2220742128

1312461

< l40105521

>61.093.0

0.390.080.660.512.51

295.00.31

3.45.9

32.218.90.0

16.318.40.01.91.80.2

70-1,49-51

55.40.88

14.210.860.115.628.573.580.820.09

100.13

3055652111

1392757

175

99554

57.092.0

0.470.161.320.822.24

619.00.54

3.44.8

30.320.3

0.017.918.60.01.91.70.2

71-2,38-40

57.21.14

12.812.290.115.076.713.641.030.10

100.07

1114822116

1213463

< l48111072

< l

>63.O108.0

0.540.160.760.722.81

534.00.40

7.26.1

30.815.60.0

14.220.5

0.02.12.20.2

71-2,105-107

57.21.12

13.411.930.115.377.023.860.840.11

100.89

1413892211

1233073

151

87842

73.092.0

0.410.100.700.572.58

633.00.41

5.94.9

32.416.50.0

14.420.5

0.02.12.10.2

73-1,38-40

53.11.14

14.012.030.135.778.174.020.980.11

99.44

2022832213

1342766

< l46

99841

>66.0104.0

0.410.140.700.822.42

623.00.34

0.05.8

34.7.17.30.0

18.913.94.42.12.20.3

73-3,118-120

52.91.11

15.311.380.104.547.674.571.220.14

98.88

1114802319

1593875

< l37111711

23

>75.O89.0

0.510.230.491.102.91

533.00.23

0.07.3

39.117.70.0

16.54.59.32.02.10.3

636

GEOCHEMISTRY OF IGNEOUS ROCKS

Magnetic Inclination Zr (ppm)

-90° 0 +90° 40 50 60 70π—i—i—r

TiO2 (wt. %) Sr (ppm)

0.8 1.0 1.2 120 140 160 50 100 150

_ , . GeochemicalCr (ppm)

U n j t

4680-

4700-

4720-

4740-

4760

4780-

4800-

4820-

62

63

64

65

66

67

68

69

SilicifiedClaystone

Unit 1Medium-GrainedAphyricVesicularBasalt Flow(Dolerite)

Unit 2HeavilyFracturedFine-GrainedSparsely PhyricVesicularCP×-Plag.Basalt

Unit 3Medium-GrainedSparselyVesicularCP×-Plag.Basalt(Dolerite)Sill

Unit 4Fine-GrainedSparsely PhyricVesicularCP×-Plag.Pillow BasaltsHeavilyFractured

4a

4b

Figure 15. Geochemical and lithological stratigraphy of the basement sections of Hole 459B.

637

D. A. WOOD ET AL.

Table 11. Representative major and trace element analyses of basement rocks recovered on DSDP Leg 60. Major element oxides in weight percent,trace elements in parts per million.

LocationLithological Unit

SiO2

TiO2

A12O3

Fe 2O 3

FeO a

MnOMgOCaONa 2 OK 2 OP 2 θ 5H2O +

Total

Mg/(Mg + Fe2 + )Total Fe as FeOSeCrCoNiZnGaRbSrYZrCsBa1 aCeEuTbHfTaThU

453-1

51.01.00

15.71.359.010.155.719.862.930.560.31n.d.

97.66

0.53010.2330.516

33.1113417

9511

25

82

0.08148

12.5024.3

1.120.622.260.112.380.63

453-2

48.30.83

15.21.399.250.178.34

13.471.690.210.03

n.d.

98.96

0.61610.5049.721

45.66

75

18

17429

21

55

0.03326

1.32

0.560.340.630.020.14~

Mariana Trough

454A-pml

49.30.86

14.41.127.470.12

13.3010.972.360.180.09n.d.

100.14

0.7608.48

27.1532

49.9382

52

16

2143

18

58

0.0149

1.986.6

0.860.481.350.110.240.12

454A-pm3

50.21.12

14.51.187.860.168.73

10.422.640.440.09n.d.

97.38

0.6648.93

28.739042.4

269

6617

3170

2377

0.02105

4.909.2

1.010.642.010.160.32

-

456-1

46.21.26

12.31.228.100.29

11.2212.552.400.060.11n.d.

95.66

0.7119.20

31.1276

35.474

93

18

< l141

24

82

0.0267

2.678.1

1.180.621.910.160.200.05

456A-3

50.71.18

16.61.177.820.165.23

11.233.270.620.11

n.d.

98.05

0.5438.88

32.8106

32.35177

16

8251

26

840.16

81

5.1513.1

1.250.672.020.160.630.18

MarianaRidge

457

53.40.89

15.61.358.970.145.336.472.731.13

0.11n.d.

96.02

0.51310.1924.129

27.317

16

18

305

19

81

0.41255

4.209.7

0.710.471.660.080.730.25

458-1

52.50.34

18.01.117.380.115.09

11.172.770.630.03n.d.

99.15

0.5518.38

36.9.214

28.180

71

14

7122

8310.09

41

1.182.3

0.280.200.820.060.15

—

Mariana Fore-Arc

458-4a

53.21.13

15.51.157.660.063.7361,25.701.520.20n.d.

95.98

0.4648.69

21.41114.21188

25

25

169

31

1010.44

79

3.037.5

1.030.572.230.120.230.13

458-5

52.81.13

13.61.64

10.960.106.175.883.410.750.06

n.d.96.52

0.50012.4327.71133.723

109

18

13

144

20

65

0.1863

2.806.1

0.880.501.700.090.160.10

459B-1

53.80.66

14.91.187.850.146.40

10.652.960.440.07

n.d.

99.10

0.5928.91

35.2159

39.657

65

17

24

120

1538

1.1424

1.132.9

0.520.341.010.050.100.03

459B-4a

57.21.14

12.81.469.740.115.076.713.641.030.10

n.d.98.99

0.48011.0627.214

32.91182

21

16

121

34

630.25

48

1.936.0

0.910.551.820.070.180.17

Mariana Trench

460-1

53.61.34

16.61.268.380.073.975.455.511.680.23

n.d.98.07

0.4579.51

23.85

20.62

91

25

22154

52

83

0.3866

3.998.4

1.340.842.850.130.320.13

460-2

52.50.87

15.21.51

10.090.135.038.243.060.910.05

n.d.97.62

0.47011.4531.951

26.822

98

18

8122

1748

0.0621

2.021.3

0.700.381.390.090.210.26

461A

58.71.01

13.21.67

11.150.135.463.003.500.220.08

n.d.

98.19

0.46512.6523.915

29.227

60

17

4113

1669

0.2536

1.814.0

0.300.321.900.080.250.15

SiO2

TiO2

A12O3F e 2 O 3

FeO a

MnOMgOCaONa2OK2OP 2 O 5

H2O +

Total

Mg/(Mg + F e 2 + )Total Fe as FeOSe (1NA)Cr (XRF)Co (1NA)Ni (XRF)Zn (XRF)Ga (XRF)Rb (XRF)Sr (XRF)Y (XRF)Zr (XRF)Ca (1NA)Ba (XRF)La (1NA)Ce (INA)Eu (INA)Tb (INA)Hf (INA)Ta (INA)Th (INA)U (INA)

Notes:a Fe calculated as Fe2θ3/FeO = 0.15.

453-1 = Sample 453-52-1, 96-102 cm; 453-2 = S456-1 Sam458-4a = Sa:460-1 = Sa

638

GEOCHEMISTRY OF IGNEOUS ROCKS

15i r i r

10

• = Unit 458-1o = Units 458-2 and 3π = Unit 458-4a× = Unit 458-4* = Unit 458-5+ = Hole 459Bv = Hole 460• = Hole 461A o o

15

10

10

1.0

0.5

v;

D + VO

o' o

' 5 ^ * ^> 1 **-458-5

+ +H+++ + V

* + ^ r ^x λ f ‰ - 458-4

v •^S^l- —-tfo-^rD V ^o *"D °c£L_ o — — 9 ^

458-1,2,3

458-4a P 458-5

458-4

I i J I

458-1,2,3

I I I

1 I I I I I I T

+ k H

+ +f +×vo × o

O OO O V• o o

• A KK XX

O ,O i O O

A

I 458-4

459B-1

°458-1 to 3

+ * A k+ k ×

××O XX

7• o

XX

o

' m++ *

+ *

I I I I

0.2

0.1

0

100

50 3

150

100

50

150

50 3

10 10MgO (wt. %) MgO (wt. %)

Figure 16. MgO variation diagram of selected major and trace elements for samples from Sites 458 through 461.

639

D. A. WOOD ET AL.

150

458-4a

iB-4aJu459B-1

458-1 to3.• / U 459B-3

458-4

J I I L20 40 60

Zr (ppm)

J L80 100

Figure 17. K2O, Y, and Sr versus Zr for samples from Sites 458through 461. (See Fig. 16 for symbols.)

0.15

0.05

458-4 × /458- /

- 1, 2 and 3 /

0.4 0.6Tb(ppm)

Figure 18. Ta, Th, and La versus Tb for samples from Sites 458through 461. (See Fig. 16 for symbols.)

640

GEOCHEMISTRY OF IGNEOUS ROCKS

100

80

B• 60

40

20

β= 458-3

458-1

458-2b

458-4

458-2a

458-5 459B-4b

458-4a

20 40 60Zr(ppm)

80 100

Figure 19. Ni versus Zr for samples from Sites 458 through 461. (SeeFig. 16 for symbols.)

300

200

100

0 458-1,2, and 3

458-4

459B-1

459B-4b458-4a

20 40 4 5 8 - 5 60"Zr (ppm)

80

Figure 20. Cr versus Zr for samples from Sites 458 through 461. (SeeFig. 16 for symbols.)

10000

8 0 0 0 -

6000 -

4 0 0 0 -

2000 -

* = Unit 458-1o = Units 458-2 and 3a = Unit 458-4a« = Unit 458-4* = Unit 458-5+ = Hole 459B* = Hole 460• = Hole461A

40 60Zr (ppm)

80

Figure 21. Ti versus Zr for samples from Sites 458 through 461. (SeeFig. 16 for symbols.)

641

D. A. WOOD ET AL.

Hf/3

a = DSDP Site 458

+ = DSDP Site 459

7 = DSDP Site 460

• = DSDP Site 461

Th Ta

Figure 22. Th-Hf/3-Ta discrimination diagram for samples from Sites 458 through 461. See Figure 11for description of the marked areas.

300

100

50

10

0.5

1 -

• = Sample 458, 28-1, 58-64 cm (458-1)D = Sample 458, 46-1, 75-81 cm (458-5)* = Sample 459B, 64-1, 4 1 - 4 4 cm (459B-2)Δ = Sample 460, 9,CC. 11-13 cm

I I I I I I I I I I I I I I I ICs Rb Ba Th U K Ta Nb La Ce Sr Nd P Hf Zr Eu Ti Tb Y

Figure 23. HYG element abundances of selected lavas from Sites458, 459, and 460 normalized to estimated primordial mantleabundances (from Wood, 1979, except that a Ti value of 1500 ppmis used here).

0.15

-£ 0.1

0.05

2.0 3.0La (ppm)

5.0

Figure 24. La versus Ta for Leg 60 samples. The fields of Japaneselava series and DSDP Leg 59 samples are from Wood et al., inpress a and b. La/Ta ratios for E-type and N-type MORB are fromWood et al. 1979a.

642

GEOCHEMISTRY OF IGNEOUS ROCKS

APPENDIX

Table 1. XRF sample descriptions (hand-specimen), Leg 60.

Sample(interval in cm)

Hole 453

42-1, 16-18 Mudstone-siltstone, dark greenish gray.47-1, 99-101 Bioturbated mudstone, brownish green.49-1, 56-61 Coarse-grained leuco gabbro, 70 percent

feldspar, ferromagnesian minerals replacedby clays/chlorite/serpentine ± amphibole(?).

49-3, 52-58 Fine- to medium-grained, leucocraticgabbro(?), pink coloration around feldspars,60 percent felsic, altered ferromagnesianminerals [clay/chlorite(?)].

50-2, 37-42 Coarse-grained leucocratic gabbro, 70 per-cent feldspar, altered ferromagnesianminerals [clays/chlorite/serpentine(?)].

52-1, 96-102 Sparsely plagioclase-phyric basalt.52-2, 94-97 Heavily altered, leucocractic gabbro, 60 per-

cent feldspar, altered ferromagnesianminerals [clays/chlorite/serpentine(?)].

53-3, 0-5 Gabbro, 50 percent feldspar, 50 percentaltered ferromagnesian minerals [clays/chlorite/serpentine(?)].

54-1, 40-47 Badly altered, leucocratic(?) gabbro.54-2, 84-91 Arkose, various lithic fragments, one half of

sample dark gray, possibly plagioclase-phyricbasalt; the other half, brownish gray matrixwith felsic grains.

54-3, 6-10 Basalt (dolerite) Plagioclase and possiblyolivine or clinopyroxene-phyric.

55-2, 14-18 Aphyric basalt clast, dark gray clay altera-tion products, fracture lined by red-brownclays.

55-3, 25-29 Leuco-gabbro, 60 percent feldspar, 40 per-cent altered ferromagnesian minerals.

55-3, 90-94 Leuco-gabbro, 65 percent feldspar, 40 per-cent altered ferromagnesian minerals.

55-4, 144-148 Leuco-gabbro, 70 percent feldspar, 30 per-cent chloritic alteration products.

56-2, 33-35 Leuco-gabbro, 70 percent feldspar, 30 per-cent chloritic alteration products, similar toprevious sample.

56-2, 85-90 Anorthosite, interstitial chlorite/clay.57-1, 30-35 Gabbro, dark-gray feldspar, red-brown

alteration products after ferromagnesianminerals.

57-3, 8-12 Leuco-gabbro, 70 percent feldspar, 30 per-cent dark-brown and red-brown alterationproducts after ferromagnesian minerals.

57-4, 123-130 Coarse-grained leuco-gabbro, 70 percentfeldspar, 30 percent interstitial altered ferro-magnesian minerals.

57̂ 4-, 36-81 Equigranular gabbro, 50 percent altered fer-romagnesian minerals, 50 percent feldspar.

58-1, 6-10 Greenish gray, plagioclase-phyric metabasalt(15% phenocrysts), occasional vesicles, somecontaining pyrite.

59-1, 13-15 Sparsely feldspar-phyric (1-2%),metavolcanic rock, aphanitic, Prussian graygroundmass, a few vesicles; quartz and car-bonate vein cuts sample; quartz coating onone side of sample.

59-1, 47-52 Gray-purple metavolcanic rock, carryingpyrite.

63-1, 0-5 Prussian gray, plagioclase-phyricmetabasalt(?), a few vesicles, some of whichare carbonate-lined.

Table 1. (Continued).

Sample(interval in cm)

Site 454

454A-5-1, 2-8

454A-5-3, 106-110

454A-5A 15-18

454-6-2, 102-106

454A-8-1, 2-7

454A-10-1, 81-84

454A-11 -1,40-44

454A-1M, 71-75

454-12-1, 100-104

454A-12-2, 26-29

454A-14-1, 2-4

454-15-1, 13-16

454A-16-1, 33-36

454A-16-1, 111-114

Site 456

456-16-1, 145-148

456-16-2, 92-96

456-17-1, 95-99

456-18-1, 50-53

456-19-1, 18-21

456A-10-1, 25-27456A-11-1, 76-79

456A-12-1, 11-15

456A-13-1, 23-26

Medium-grained aphyric, altered greenishgray basalt, 2 percent vesicles.Medium-grained, altered, gray-green basalt,10 percent vesicles.Medium-grained, altered, aphyric, gray-greenbasalt, 5 percent vesicles.Medium- to coarse-grained basalt (dolerite),2 percent vesicles; outer surface of samplecovered by light-green clays.Gray, aphyric, aphanitic basalt, 2 percentvesicles lined by dark gray-brown clay.Gray, medium-grained, altered basalt, 10percent vesicles, 2 percent large vesicles (1-2mm across), often lined by light-green clay,other vesicles <0.5 mm.Gray, medium-grained, altered basalt, 10percent vesicles, 2 percent (1-2 mm) lined bylight-green clay, other vesicles <0.5 mm,similar to previous sample.Gray, medium-grained, altered basalt, 5 per-cent vesicles (<0.5 mm).Gray, medium-grained, altered basalt, 5 per-cent vesicles (<0.5 mm) similar to previoussample.Gray, fine-grained, aphyric basalt, 10 per-cent vesicles (<2 mm), larger vesicles linedby blue-gray clay.Gray, fine-grained, aphyric basalt, 10 per-cent vesicles (<2 mm).Gray, fine-grained, aphyric basalt, 10 per-cent vesicles, slightly altered.Olive-green-gray, fine-grained, aphyricbasalt, 5 percent vesicles, slightly altered.Olive-green-gray to gray, aphanitic, aphyricbasalt, 2 percent vesicles.

Olive-green-brown, aphanitic, aphyric basalt,3 percent large vesicles (1-2 mm), some withcalcite and/or pyrite lining.Fine-grained, blue-gray, plagioclase-phyricbasalt. (5% phenocrysts, 0.5-2 mm across),1 percent vesicles.Fine-grained, blue-gray, plagioclase-phyric(5% phenocrysts, 0.5-2 mm across) basalt, 1percent vesicles, some containing calcitecrystals.Medium-grained, olive-gray basalt, 3 percentvesicles, some lined by calcite and green clay.Sparsely plagioclase-phyric, dark gray,medium-grained basalts, 5 percent vesicles,sample cut by fracture lined by Fe oxides/hydroxides.Purple-gray mudstone.Green-gray, aphyric, aphanitic metavolcanicrock, 10 percent vesicles (<2 mm), somecontain pyrite.Green-gray, fine-grained basalt, 2 percentvesicles, patches of dark-gray clay alterationproducts.Dark-gray, aphyric, aphanitic basalt, 5 per-cent vesicles (<2 mm), 5 percent vesicles(2-10 mm) lined by or filled by light-brownclays.

643

D. A. WOOD ET AL.

Table 1. (Continued). Table 1. (Continued).

Sample(interval in cm)

Sample(interval in cm)

456A-14-1, 12-16 Dark-gray, aphyric, aphanitic basalt, 5 per-cent vesicles ( < 2 mm), 5 percent vesicles(2-10 mm), lined by yellow, red-brown andpurple alteration products.

456A-14-1, 38-42 Olive-gray, aphyric, aphanitic basalt, 2 per-cent vesicles (<0.5 mm), concentric zoningof alteration colors through sample suggestspart of pillow interior.

456A-15-1, 27-31 Gray, aphyric, fine-grained basalt, 4 percentvesicles concentrated in "pipes," vesiclesoften lined by gray or light-brown clays.

Hole 457

4,CC Tan-colored piece of pumice, very soft.

Hole 458

27-1, 130-132 Olive-green, fine-grained sandstone, cut bycarbonate-lined fracture.

28-1, 58-64 Heavily altered, light-green-gray, spheruliticbasalt, one large vesicle ( 5 × 3 mm).

28-1, 125-130 Heavily altered, light-green-gray, spheruliticbasalt.

28-1, 140-146 Dark gray, aphyric, fine-grained basalt, 3percent vesicles ( < 2 mm), a few largervesicles lined by blue-gray clay. •

29-1, 54-57 Olive-green, fine-grained, aphyric basalt, 1percent vesicles (< 1 mm).

29-2, 37-42 Dark-green-gray, fine-grained aphyric,altered basalt, < 1 percent vesicles (< 1 mm).

30-1, 56-59 Heavily altered, olive-green, spheruliticbasalt, 5 percent vesicles (2-5 mm), lined bygray-blue clay and white mineral (zeolite?).

31-1, 39-41 Heavily altered, light olive-green, fine-grained, aphyric basalt, one vesicle (2 + 2mm).

31-1, 131-136 Light-gray (green tinges), fine-grained,aphyric basalt, 3 percent vesicles (< 1 mm).

32-1, 116-119 Light-gray-green, medium-grained, aphyricbasalt, 1 percent vesicles (< 1 mm).

32-3, 34-36 Light-gray-green, aphyric, aphanitic basalt, 5percent vesicles (< 1 mm).

33-2, 19-22 Gray-green, medium-coarse grained (ca. 1mm) basalt, few vesicles (< 1 mm).

34-1, 39-42 Light-green-gray, medium-grained, basalt, 2percent vesicles (< 1 mm), one surface ofsample is a fracture covered by a light-olive-green clay and carbonate.

35-1, 42-45 Light-green-gray, medium-grained, aphyricbasalt, 5 percent vesicles (< 1 mm); one sur-face of sample is a fracture; basalt heavilyaltered, but no coating of secondaryminerals.

35-2, 69-73 Light gray, aphyric, aphanitic basalt, severaltrains of small vesicles (<0.5 mm).

36-1, 27-31 Light gray-green, aphyric, aphanitic basalt,few vesicles, cut by carbonate andclay/chlorite(?)-lined vein.

37-1, 52-55 Gray-green, fine-grained, aphyric basalt.37-2, 62-64 Gray-green, medium-grained, aphyric basalt,

few vesicles.38-1, 51-53 Battleship-gray, fine-grained basalt, 2 per-

cent vesicles ( < 2 mm), odd plagioclase(?)phenocrysts(?).

39-1, 37-40 Gray, very fine-grained basalt, 5 percentvesicles, often lined or infilled by gray clayand calcite.

40-2, 90-93 Battleship-gray, aphyric, aphanitic basalt, 1percent vesicles (<0.5 mm), sample cut bycarbonate-lined fracture.

41-1, 22-26 Heavily altered, olive-gray, aphyric, aphaniticbasalt, 3 percent vesicles (<0.5 mm).

42-1, 61-64 Green-gray, fine-grained aphyric basalt, fewvesicles, slickenside surface on sample coveredby chlorite.

42-1, 86-88 Dark-gray, aphyric, aphanitic basalt, 2 percentvesicles often surrounded by light-gray altera-tion zone and lined by light-blue clay andzeolites or chalcedony. Sample cut by zeolite orchalcedony-lined vein.

43-1, 118-121 Gray, fine-grained basalt, 2 percent vesicles,heavily altered to green clays, a few zeolites(?)in vesicles.

44-1, 53-60 Gray, aphyric, aphanitic basalt, 2 percentvesicles lined by zeolites and blue-gray clay.

45-1, 8-12 Gray, aphyric, aphanitic basalt, 2 percentvesicles lined by zeolites and blue-gray clay,similar to previous sample.

45-1, 24-27 Olive-green-gray, aphyric, fine-grained basalt,10 percent vesicles (<3 mm).

46-1, 75-81 Gray-green, fine-grained, aphyric basalt, fewvesicles.

46-1, 83-86 Gray-green, aphyric, aphanitic basalt, fewvesicles, one vesicle filled by a zeolite(?), trainsof fine vesicles (<0.5 mm).

47-1, 67-71 Light-green-gray, aphyric, aphanitic basalt;three parallel trains of fine vesicles (<0.5 mm)cut sample, plus a few larger vesicles (1-2 mm).

48-1, 79-82 Gray-green, aphyric, aphanitic basalt, lustrous,black glassy margin on one side of sample, veryfresh, but coating of clays/chlorite on outersurfaces.

49-1, 136-138 Olive-green-gray, fine-grained, aphyric basalt, 5percent vesicles, gray-blue clay lining tovesicles.

Hole 459

60-2, 80-83 Gray-green, medium-grained inequigranularbasalt, 3 percent vesicles (< 1 mm).

61-1, 115-118 Light-gray-green, medium-grained inequi-granular basalt, 15 percent vesicles ( < 2 mm),specks of Fe111 oxides present.

61-2, 47-50 Green-gray, medium-grained inequigranularbasalt, 2 percent vesicles ( < 2 mm), mesostasisaltered to light-olive-green and tan clays.

63-1, 53-54 Olive green-gray, fine-grained aphyric basalt, 2percent vesicles, generally <0.5 mm, a few 1-2mm.

64-1, 41-44 Dark green-gray, fine-grained, aphyric basalt, 3percent vesicles (< 1 mm).

65-1, 59-63 Heavily altered, dark green-gray, aphyric,aphanitic basalt, 1 percent vesicles ( < l mm);sample also includes part of glassy margin(black, lustrous, relatively fresh), and is cut bycarbonate and clay and/or chlorite(?)-linedvein.

66-1, 73-77 Heavily altered, olive-green-brown, medium-grained, inequigranular basalt, 3 percentvesicles ( < 2 mm), brown clay lining to somevesicles.

66-2, 147-150 Heavily altered, olive-green-brown, medium-grained, inequigranular basalat, 1 percentvesicles ( < 2 mm), one large vesicle (5 mm).

67-1, 118-120 Olive-green, medium-grained, inequigranularbasalt, 3 percent vesicles, bright olive-greensecondary mineral—malachite? Infilling andreplacing mesostasis near some vesicles.

644

GEOCHEMISTRY OF IGNEOUS ROCKS

Table 1. (Continued).

Sample(interval in cm)

68-1, 51-54

69-1, 144-146

69-3, 83-86

70-1, 49-53

71-2, 38-40

71-2, 105-108

73-1, 38-40

73-3, 118-120

Site 460

460-9,CC, 7-8 (#2)

460A-9,CC, 11-13

460A-11-1, 29-33

Hole 461A

3-1, 60-62

Olive-green, fine- to medium-grained, ine-quigranular basalt, bright-olive-green mineral(malachite?), replacing mesostasis and liningfracture with carbonate.Gray, fine-grained, aphyric basalt, 2 percentvesicles, sample surface covered by greenand brown clays.Olive-green, very fine grained aphyric basalt,2 percent vesicles (<2 mm).Dark-olive-green-gray, fine-grained, aphyricbasalt, 4 percent vesicles (<0.5 mm), 1 per-cent vesicles (1-2 mm); sample cut byslickensided, chlorite-lined fracture.Heavily altered, olive-green-gray, fine-grained basalt, 3 percent vesicles (< 1 mm).Heavily altered, olive-green-gray, fine-grained basalt, 3 percent vesicles (< 1 mm);similar to previous sample, cut by slicken-sided chlorite-lined fracture.Heavily altered, olive-green-gray, fine-grained, aphyric basalt, 3 percent vesicles,similar to previous two samples.Heavily altered, olive-green-gray, fine-grained, aphyric basalt, 2 percent vesicles(<2mm).

Heavily altered, olive-green-brown, fine-grained, plagioclase-phyric (5%) basalt, 1percent vesicles (0.5-2 mm).Olive-green-gray, aphanitic, aphyric basalt, 2percent vesicles (<5 mm).Green-gray, medium-grained basalat, oddvesicle (<2 mm), lined by dark-green clay.

Green-brown, fine-grained, aphyric basalt,light-tan and dark-green clays ± chloritecoating surface of sample, cut by fracturelined by Fe111 oxides.

645