Embed Size (px)
Citation preview
29. PETROLOGY OF BASALTS OF HOLES 447A, 449, AND 450,SOUTH PHILIPPINE SEA TRANSECT, DEEP SEA DRILLING PROJECT LEG 59
G. S. Zakariadze,1 Janelidze Institute of Geology, Academy of Science of Georgian S.S.R., Tbilisi, U.S.S.R.and
L. V. Dmitriev, A. V. Sobolev, and N. M. Suschevskaya, Vernadsky Institute of Geochemistry andAnalytical Chemistry, U.S.S.R. Academy of Science, Moscow, U.S.S.R.
ABSTRACT
Basalts studied represent different stages of basement formation of the Philippine Sea plate, including the WestPhilippine Basin (middle Oligocene—Hole 447A) and the Parece Vela Basin (upper Oligocene to middle Miocene-Holes 449 and 450 and Dredge Site 1398, on the seventeenth cruise of the Dmitry Mendeleev). All of these basaltsdisplay compositions similar to that of normal mid-ocean ridge (MOR) tholeiites, and despite the age differences, theyseem to be differentiation products of an initial tholeiitic melt of essentially the same composition. According to thesuggested petrologic scheme, this initial liquid is derived from dry melting of mantle lherzolite at 10 kb andtemperatures about 1270°C. Basalts of Hole 447A were formed after a rapid magma injection to shallow depths andcotectic olivine-plagioclase crystallization at 1 atm, whereas Parece Vela basalts were derived from anchieutecticclinopyroxene-plagioclase-olivine crystallization at intermediate pressures (from 5 kb-1 atm). In each case, extrusionand crystallization of the magma were preceded by fractionation of excessive olivine from primary tholeiitic liquid. Thepetrologic and geochemical features of these basalts do not indicate any influence of subduction zone activity. Withintholeiitic compositional limits, there is a slight progressive increase of K, Ti, P, and Ba in the basalts and volcanicglasses from the West Philippine Basin to the Parece Vela Basin and thence to the Mariana Trough; this may indicateslightly increasing depths of origin of the primary tholeiitic liquids from west to east along the South Philippine Seatransect.
INTRODUCTION
Petrologic peculiarities provide a key to understand-ing whether the formation of basalts of marginal seabasins is completely analogous to that of mid-oceanridge (MOR) tholeiites or whether there are some dis-tinctions related to subduction-zone influence or toother geodynamic factors (Kang, 1971; Packham andFalvey, 1971; Moberly, 1972; Sclater et al., 1976;Toksöz and Bird, 1977; Molnar and Atwater, 1978).
Available data do not present a simple picture. Ridleyet al. (1974) and Hawkins (1976; 1977) find marginal seatholeiites and mid-ocean ridge basalts (MORB) com-pletely analogous; other researchers (Hart et al., 1972;Gill, 1976; Tarney et al., 1977; Dietrich et al., 1977)point out distinctions between these two types ofbasalts. To some extent the petrologic and geochemicalfeatures of basement basalts may depend on the genetictype and the stage of evolution of marginal sea basins(Hart et al., 1972; Tarney et al., 1977; Dietrich et al.,1977).
New data obtained from DSDP Legs 58, 59, and 60shed light on problems of the basement origin of variousmarginal basins within the Philippine Sea plate. Threemajor stages of basement formation have been estab-lished for the Philippine Sea plate: 50 to 35 m.y. (WestPhilippine Basin), 30 to 13 m.y. (Parece Vela andShikoku basins), and 5 to 0 m.y. (Mariana Trough).This episodic evolution is the major geodynamic distinc-
Present address: Vernadsky Institute of Geochemistry and Analytical Chemistry,U.S.S.R. Academy of Science, Moscow, U.S.S.R.
tion between the Philippine Sea plate and the largeoceanic plates (Karig, 1975). Thus it is possible to com-pare basalts of the Philippine Sea basins with MORbasalts for each stage and also to examine the evolutionof basaltic volcanism from stage to stage.
The major objective of this chapter is to present theresults of comparative petrologic and geochemical in-vestigation of basement basalts obtained on Leg 59from the West Philippine Basin (Hole 447A) and theParece Vela Basin (Holes 449 and 450). These basaltswere formed during the first two stages of basin forma-tion of the Philippine Sea plate. Comparison of thesebasin basalts is of special interest because their geo-dynamic history may not be completely similar. If theParece Vela is a good example of an extensional back-arc basin, evolution of the West Philippine Basin ismuch more complex; the several alternative models ofits formation are still in dispute (Karig et al., 1973; Karig,1975; Hilde et al., 1977; and Scott et al., this volume).
SUMMARY OF PETROGRAPHY ANDMAJOR PETROCHEMICAL UNITS
Basements of Holes 447A, 449, and 450 are com-posed of aphyric, microphyric, and sparsely phyric ba-salts showing exceptional petrographic similarity toMOR basalts. Many established petrographic units arecomposed of well-defined cooling units: (1) single mas-sive lava flows, (2) pillow lava flows, and (3) pillowedmassive flows composed of an upper pillow lava under-lain by a massive flow. The phenocryst contents arerather limited, ranging from 2% to lθ°7o; rarely do theyamount to 25°7o. Dominating plagioclase phenocrysts
669
G. S. ZAKARIADZE, L. V. DMITRIEV, A. V. SOBOLEV, N. M. SUSCHEVSKAYA
(An87_60) are represented by slightly resorbed prismaticand tabular crystals (0.5-3 mm) rarely oscillatoryzoned. Subordinate anhedral olivine (0-5%, 0.5-2 mm)and clinopyroxene (0-5%, 0.5-2.5 mm) are often glom-eromorphed with plagioclase. Inner parts of lava pil-lows and massive flows display ophitic to subophitictextures and are composed of plagioclase (35-50%, 0.4-0.8 mm, An67_4o), clinopyroxene (30-45%, 0.4-1 mm),olivine (0-5%, 0.1-0.5 mm), titanomagnetite (0-5%,0.01-2 mm), and glass (2-10%). Less crystallized pillowand flow margins enriched with glass reveal texturesvarying from intersertal to hyalopilitic and variolitic.Plagioclase and olivine phenocrysts in some cases con-tain small inclusions of brown chromic spinel. Micro-phenocrysts, microinclusions, and microxenocrysts ofchrome spinel also occur in the groundmass. Scatteredspherical and irregular vugs ( — 5%) and vesicles (~3%)range in size from 0.5 to 5 mm and from 0.05 to 0.1mm, respectively; these are partially or completely in-filled with carbonates (sometimes with zeolites) rimmedby smectite. The basalts commonly are dissected byveins (0.5-1 mm) filled mainly by carbonates.
Almost all of the basalt samples studied were more orless subjected to low-temperature secondary alteration.The most unstable constituents are glass (replaced bysmectite, zeolite, and, partially, by carbonate) andolivine (replaced by the same mineral assemblage and byserpentine, saponite, and iron hydroxides as well). Themajor constituents—plagioclase and pyroxene—exhibitonly slightly secondary alteration.
Analytic results (Tables 1-6) present clear evidencethat all of the basalts are normal MOR tholeiites (Kay etal., 1970; Bougault and Hekinian, 1974; Kay and Hub-bard, 1978; Sun et al., in press). Nevertheless, withinthese compositional limits several obvious petrochemi-cal units are distinguished, based on the composition ofvolcanic glasses, aphyric rock types, and phyric rocktypes with a low degree of posteruptive differentiation(Fig. 1—note that unit numbers referred to in thischapter are listed in Column B; these differ from theunit numbers in Column A, which are used in the sitereport).
Definite compositional variation is related to sub-marine low-temperature alteration of basalts. Figure 2illustrates the increase of K2O content with progressiveoxidation (a factor of 2-3 relative to unaltered samples).The most altered rocks of Hole 447A are represented atthe upper (Cores 14-18) and lower (Cores 28 and 35)levels of the basalt section. Altered tholeiites of Sites449 and 450 display less potassium enrichment. An in-crease of potassium is accompanied by an increase ofH2O+ and A12O3 and a decrease of MgO and SiO2 con-tents (Table 1, Fig. 3, A-F). On the whole, similartrends are characteristic for alteration of basaltic glassesof Hole 447A (Table 2, Fig. 6).
Let us consider data for the separate sites.
Hole 447A (West Philippine Basin)
Basaltic basement of Hole 447A (113-296.5 m sub-bottom) is represented by a continuous section of
tholeiitic flows and pillow lavas which are overlain by 28meters of middle Oligocene volcaniclastic breccias andvitric tuffs. On the basis of petrograph and petrochem-ical data, the basement section is subdivided into fivemajor petrochemical units (Fig. 1; the unit designationsused in this chapter are in Column B of this figure; thosein Column A are used in the site report).
Unit 1 is 28 meters thick (113-141 m sub-bottom) andconsists of four flows. The upper three flows are fromthe top down, respectively, a massive flow (2 m), pillowlava (8 m) and a composite flow (pillow lava gradingdown into a massive flow, 8.5 m), all consisting ofplagioclase-phyric tholeiite; the lowermost flow ismassive aphyric tholeiite (9.5 m). Plagioclase pheno-crysts account for 1% to 3%, rarely attaining 10%.Low-temperature alteration is rather extensive, mainlyat the expense of groundmass olivine and glassy meso-stasis, causing a relative decrease in SiO2 and MgO andincrease in A12O3 and K2O contents (Figs. 2 and 3, A, C,and E).
Unit 2 consists of a composite tholeiitic flow 13.5meters thick (141-154.5 m sub-bottom; Fig. 1) and iscomposed of an upper pillow lava (12.1 m) underlain bya massive flow. Phenocrysts are represented by the uni-form assemblage of plagioclase (An72_69, 3-10%),clinopyroxene (3-6%), and olivine (1-2%). The ground-mass has a high content of titanomagnetite (0.5-7%),which increases total iron and titanium contents of thebasalt as compared to the other units in Hole 447A. Asin Unit 1, olivine phenocrysts and olivine in the ground-mass and glass are considerably altered; the basalts arerespectively low in magnesium and silica and enriched inpotassium. Because of superimposed low-temperaturealteration, the basalts of Units 1 and 2 have an unusualcomposition that is off the main petrochemical trend ofHole 447A tholeiites (Fig. 3, A, C, and E).
Although the basalts of Units 3 and 5 show somepetrographic distinctions, petrochemically they form acommon group. Unit 3 (154.5-184.5 m sub-bottom;Fig. 1) consists of three flows: an upper plagioclase-clinopyroxene-olivine microphyric tholeiite (12.5 m); amiddle tholeiite similar to the one above but im-poverished in phenocrysts and enriched with glass (up to80 or 90%, 4 m); and a lower plagioclase-clinopyroxene-olivine phyric dolerite (13.5 m). Each consists of an up-per pillow lava underlain by a massive flow. In mostcases, phenocrysts are represented by the three-mineralassemblage: plagioclase (An82_70), clinopyroxene (2-3%), and olivine (1%). Microprobe analysis of some ofthese phases are given in Tables 3 through 5.
Unit 5 is composed of six flows with a total thicknessof 88 meters. Down-section three composite flows 19.5,7, and 5 meters thick, respectively, are underlain bythree pillow lava flows, respectively, 6.5, 35, and 14.5meters in drilled thickness. Petrographically the flowsare plagioclase (An85_72, 2-10%), olivine (1-5%), phyricbasalts containing small inclusions of chrome spinel(0.01-0.02 mm) in phenocrysts and in the groundmass.At the base of the section, the basalts contain xenocrystsof spinel and plagioclase associated with small (5 mm)
670
PETROLOGY OF BASALTS
113.0
122.0
131.0
140.0
149.0152.0
158.0
167.0
176.0
180.0
_ 185.0
~ 191.0
IQ 198.5ε2o 207.5.α
" 216.5
225.5
234.5
243.5
252.5
261.5
270.5
275.0
284.0
295.0296.5
20
21
6a: Flow6b: Pillow lava6c: Pillow lava/flow6d: Pillow lava/flow
PlagioclasePhyricBasalt
(113.0-131.5 m)
7: Pillow Lava
8a: Pillow lava/flow
8b: Pillow lava/flow
8c: Pillow lava/flow
8d: Pillow lava/flow,doleritic base
\
PlagioclasePhyric Basalt
(131.5-141.0 m)
PlagioclaseClinopyroxeneOlivinePhyric Basalts
(141.0-185.0 m)
Flow, doleritic base
Aphyric to Olivine-Spinel PhyricBasalt
(185.0-194.0 m)
10a: Pillow lava/flow10b: Pillow lava/flow
11a: Pillow lava/flow
11b: Pillow lava/flow
11c: Pillow lava/flow
11d: Pillow lava
OlivinePlagioclaseClinopyroxenePhyric Basalt
(194.0-209.0 m)
11e: Pillow lava
PlagioclaseOlivine SpinelPhyric Basalt
(209.0-296.5 m)
11f: Pillow lava
Plagioclase Phyric Tholeiites and Aphyric Tholeiite
Low in SiOo and MgO, enriched with AUOo and r^O.
(113.0-141.0 m)
, Plagioclase Clinopyroxene Olivine Phyric TholeiiteSlightly Enriched with Titanomagnetite
Significant alteration, low in SiOo and MgO,enriched with KUO. Primary enrichment withFeO* and T i0 o .
(141.0-154.5 m)
Plagioclase Clinopyroxene Olivine Phyric Tholeiite
High in Siθ2 and low in AUOg.
(154.5-185.0 m)
. Aphyric Dolerite Enriched with Olivine (withChrome Spinel Inclusions) and Olivine PlagioclaseClinopyroxene Tholeiites
Enriched with MgO and low in SiOo, FeO*
and TiOo•(185.0-209.0 m)
, Plagioclase Olivine Phyric Tholeiites (ChromeSpinel Inclusions in Plagioclase, Olivine andGroundmass; Xenocrysts of Chrome Spinel andPlagioclase; Xenoliths of Gabbroic Anorthosites)
High in SiOo and low in AUOo.
(209.0-296.5 m)
Figure 1. Major petrochemical units of basement section of Hole 447A. (The unit numbers in the A column are those used in the site report; notethat they differ from the numbers in the B column, which represents the unit divisions referred to in this chapter.)
xenoliths of gabbroic anorthosites. Despite these petro-graphic differences, the rocks of Units 3 and 5 petro-chemically form an indivisible group that distinctlydominates the basement section of Hole 447A. Basaltsof this group are the most siliceous and are relativelylow in magnesium and alumina (Fig. 3). The composi-tion of the whole rocks and the glasses of Units 3 and 5show no significant differences and vary within rathernarrow limits, which is evidence for negligible posterup-tive fractionation of phenocrysts.
The upper part of Unit 4 is of special interest. It con-sists of a massive aphyric doleritic flow (185-194.0 m
sub-bottom, Fig. 1). The distinctive feature of this unitis marked enrichment in olivine (up to 15%, especiallyat the base of the flow), containing inclusions of brownchrome spinel (0.1-0.2 m). The olivine is extensivelyresorbed to skeletal forms and obviously is the earliestphase that crystallized before plagioclase. Petrochemi-cally the doleritic flow is the most magnesian rock[MgO/(MgO + FeO) = 0.52], the least siliceous(45.55% SiO2), the least ferriferous (2.88% Fe2O3 and5.66% FeO), and the least titaniferous (0.79% TiO2) inHole 447A. The alumina content is rather high for amid-ocean ridge tholeiitic basalt. In addition, the high
671
G. S. ZAKARIADZE, L. V. DMITRIEV, A. V. SOBOLEV, N. M. SUSCHEVSKAYA
o
αΔ
-1-2-3- 4
O23-1®
1
029 *ו
22O -. ^
25O O24
15 ^
5® ×"
®35
I
II
17®
®28/
Δ36-3
-
17DΔ36-2
-
Figure 2. Distribution of K2O contents of tholeiites of Holes 447A,449, and 450 in relation to the state of alteration of rocks. (In thekey 1 and 2 represent fresh and altered tholeiites of Hole 447A,respectively; 3 and 4 are tholeiites of Holes 449 and 450, respec-tively. Numbers refer to cores for respective holes.)
magnesian group includes two other underlying flows ofolivine-plagioclase-clinopyroxene tholeiites (9.5 m and5.5 m, 194.0-209.0 m sub-bottom).
Hole 449 (Western part of the Parece Vela Basin)
Basaltic basement of Site 449 was cored from 111 to151.5 meters sub-bottom, beneath upper-Oligocenenannofossil ooze. The upper part of the basalt section(23.5 m) consists of plagioclase-olivine phyric tholeiite(An8O_7o, 5-7<7o; olivine 2-3%), which is glass-rich be-tween 111 and 124 meters sub-bottom. The underlying17 meters are composed of a massive plagioclase phyrictholeiitic flow, with plagioclase phenocrysts up to 30970.
Hole 450 (Eastern part of the Parece Vela Basin)
Only one flow of plagioclase-olivine phyric pillowedtholeiite was encountered and cored (330-340 m sub-bottom) at Hole 450. Plagioclase phenocrysts in thisflow are common (8-16%, Ang<µ̂ core composition,0.5-6 mm) and contain numerous glass inclusions. Oli-vine phenocrysts are less common (1-2%) and are com-pletely pseudomorphed by iddingsite, clays, and zeo-lites. The flow lies beneath a thick volcaniclastic apronof middle-Miocene fine vitric tuffs thermally metamor-phosed in contact with the basalt. Seismic-reflectionprofiles show that these basalts form a rather smallprominence (less than 1 km across) and that true base-ment is significantly deeper nearby; based on theseobservations, the basaltic flow of Hole 450 is inter-preted as an intrusion that most likely represents a latestage of basement formation (see the Site 450 report,this volume). The considerable petrographic similarityof tholeiites of Sites 450, 449, and 447 favors thisassumption. For more detailed petrography, see the sitereports.
Other Sites in the Parece Vela Basin
Data on the composition of Parece Vela Basin base-ment basalts have also been obtained on DSDP Leg 6(Site 54, Ridley et al., 1974) and on the seventeenthcruise of the Soviet Dmitry Mendeleev, on which thewestern slope of the Parece Vela Rift was dredged.(Initial Report of the scientific party 1977, Dietrich etal., 1977). Basalts at Site 54 are microphyric and aphyricplagioclase-olivine tholeiites, petrographically similarto basalts of Holes 449 and 450 (detailed description ofthese basalts is given by Ridley et al., 1974). At the base
Table 1. Chemistry and normative mineralogy of West Philippine and
Hole
Unit
Sample(intervals in cm)
SiO2
TiO 2
A12O3
Fe 2 O 3
FeOMnOMgOCaONa 2OK2OH 2 O "H 2 O +
P 2 θ 5
Total
QOrAbNeAn
WON
En füiFs 'En IH
Fs l H y
F o lF a l O 1
11Mt
14-2,41-43
47.321.04
17.315.364.610.155.90
11.982.500.371.061.580.13
99.30
2.2621.87
—36.1310.216.363.246.243.181.811.022.044.71
1
15-1,24-26
47.341.05
16.964.505.710.145.60
12.132.700.411.341.730.13
99.74
2.2223.59
37.2111.387.023.195.022.640.980.611.975.56
16-2,87-89
47.800.93
16.284.455.210.146.81
11.982.660.331.022.070.13
99.81
2.0123.26
32.5811.687.563.324.952.173.511.701.824.60
17-3,81-86
47.821.19
15.116.385.560.126.11
11.322.390.711.621.390.16
99.78
0.303.72
20.89
29.6711.396.794.008.925.25
2.335.64
2
18-1,71-76
47.961.19
16.096.545.710.156.48
10.792.510.701.321.210.20
99.85
0.404.25
21.81
31.429.285.283.608.735.95
2.325.75
22-3,32-40
49.461.04
14.712.887.100.157.77
12.212.290.170.990.810.14
99.72
1.181.02
19.78
29.9912.918.554.42
11.204.48
2.014.78
447A
3
23-1,51-58
50.111.05
14.792.437.820.147.84
11.462.160.470.690.450.13
99.54
2.642.78
18.35
29.7611.387.533.03
12.354.09
2.125.56
Parece Vela basin basalts, DSDP
23-3,73-83
50.301.04
14.643.177.280.177.46
11.552.220.100.860.540.13
99.46
3.610.60
19.15
30.2811.397.373.25
11.575.10
2.014.99
24-2,25-39
45.550.79
16.372.885.660.148.92
11.922.130.071.110.340.11
99.09
0.4319.04
36.8810.377.272.216.041.847.112.381.584.21
4
25-1,104-111
47.430.79
16.862.576.680.149.71
11.402.220.070.591.340.12
99.92
0.4219.16
36.588.485.941.839.312.886.592.241.534.43
28-3,52-58
47.860.93
16.015.104.580.157.17
12.033.240.501.280.970.14
99.96
03.0223.672.37
28.3213.288.723.62——6.693.061.804.53
Leg 59.
5
29-3,32-40
49.601.10
14.703.856.890.178.37
11.881.740.240.501.010.07
100.12
3.481.11
14.68
32.2611.507.832.77
13.354.88
2.126.01
35-2,70-75
46.980.79
16.985.124.630.125.05
11.903.791.131.222.060.14
99.91
6.9118.427.99
26.9013.888.194.99——3.372.261.554.62
449
17-2,41-51
47.521.05
17.005.564.090.166.28
11.182.420.422.012.240.10
100.06
0.62.22
20.97—
36.448.595.622.37
10.744.61——2.125.56
450
36-2,146-150
47.010.60
15.985.443.860.145.64
13.012.580.391.392.650.21
99.90
2.4022.76
—32.2114.069.203.873.071.291.660.773.174.41
36-364-75
47.941.26
16.164.934.400.126.56
12.172.670.311.421.330.22
99.49
1.8923.34
—32.25
1.977.933.166.302.521.830.802.474.41
672
PETROLOGY OF BASALTS
0.3 0.5
12
11
Sio
9
8
i
G \
1 8 © ^
\
35® *1 6
-
Or22CC
28
29 2^―- £
20)22
37
30
s 2 <
‰×0.3 0.4 0.5 0.6
V
x×>A
\
0.3 0.4 0.5 0.6
0.3 0.5 0.6
E
2 9 C ^
/O22
O25O24
©37»302—20
23-30^^23 116®®28
17®®1418© © ^
©35i i _i 1
10
0.3 0.4
Y
×××I
0.6MgO
0.3 0.4 0.6
1.4
O 1.2
1.0
0.8
0.6
1.6
1.4
1.0
0.8
y\
O 1© 2Φ 3π 4ffl 5
67
08$ aX 10
0.3 0.4 0.6MgO
0.3 0.4 0.5 0.6
MgO+FeO* MgO+FeO*
Figure 3. Variation of major-element contents of rocks and glasses of Holes 447A, 449, 450, and 54 in relation to the magnesian number. (Key: 1-3= Hole 447A—fresh whole rocks, altered whole rocks, glasses; 4-5 = Hole 449—whole rocks, glasses; 6-7 = Hole 450—whole rocks, glasses;8-9 = Hole 54—whole rocks, glasses; 10 = Parece Vela Rift tholeiites from Dredge Site 1398, seventeenth cruise of the Dmitry Mendeleev.)
of the western slope of the Parece Vela Rift, three majortypes of basalts have been dredged (St. 1398, 17°.07'N,139°24'E): microphyric and aphyric plagioclase-oli-vine-clinopyroxene tholeiites, coarsely phyric-plagio-clase basalts, and variolitic basalts. Parece Vela Riftmicrophyric and aphyric tholeiites do not reveal anypetrographic distinction from other Parece Vela tholei-ites. Plagioclase phenocrysts dominate (5-7'%, An80_70
core composition, An51 marginal zones); olivine andclinopyroxene are present as sparse microphenocrysts.The groundmass is intersertal or hollocrystalline ophiticand consists of coarse plagioclase (40-50%, An7O_5O),anhedral clinopyroxene (35-40%), rare olivine, titano-magnetite (3-5%), and partially or completely alteredglassy mesostasis. Variolitic basalts represent glass-richpillow rinds of the tholeiites of the same type. Thephyric-plagioclase basalts are extensively altered; pla-gioclase phenocrysts are often completely pseudo-morphed by fine aggregates of smectite and zeolites.The primary composition of these rocks is considerablychanged by this alteration.
Analytic data show that there is an apparent dif-ference in the composition of basalts and volcanicglasses of the Parece Vela basement (Tables 1 and 2;Fig. 3), although as a whole these compositions arerather similar to that of MOR basalt. We should em-phasize that analyzed glasses were exceptionally fresh,
whereas all basalt samples were more or less subjectedto low-temperature secondary alteration, mainly at theexpense of olivine and glass. Because considerable post-eruptive fractionation of phenocrysts within flow unitsis not observed, noted differences between glasses andcrystalline rock in one unit should be explained by low-temperature alteration processes. Low-temperature alter-ation is also indicated by the reduced silica andmagnesium contents of basalts, similar to altered basaltsof Hole 447A. In contrast, the low magnesium, hightotal iron, high titanium contents, and constant slightoversaturation with silica of Parece Vela Rift tholeiites(Table 1) indicate that these represent the most differen-tiated rock types of the Parece Vela basement so farobtained.
Trace-element data (Table 6) emphasize the similarityof the West Philippine Basin and the Parece Vela Basinbasalts with normal (depleted) MOR tholeiites. Withinthese limits however, meaningful variations of traceelements have been established. Distribution of traceelements through the section of Hole 447A displays acomplex picture, but, as a whole, it is consistent with themajor petrochemical units (Fig. 4). (In addition to theTable 6 data, some trace-element data are used fromanalyses conducted at the University of Birmingham forLeg 59 shipboard scientists.) Parece Vela basalts areslightly enriched with potassium, titanium, and phos-
673
G. S. ZAKARIADZE, L. V. DMITRIEV, A. V. SOBOLEV, N. M. SUSCHEVSKAYA
Table 2. Microprobe analyses of glasses, DSDP Leg 59.
Hole
Core-Section,Interval (in cm)
Siθ2TiO 2
A12O3
FeOMnOMgOCaONa 2OK 2 ONiO
Total
Cr 2 O 3
MgOMgO + FeO*QOrAbNeAnWoEn DiFsE n HvFs H y
F o πiFa O 1
11
20-1,9-21
50.671.01
15.019.830.118.19
12.142.250.08
99.29
0.454
—
0.519.0
30.612.46.65.49.57.73.12.71.9
22-170-76
50.750.97
14.959.760.238.42
12.432.070.08
99.66
0.463
0.517.5
31.612.76.85.4
10.38.22.72.41.8
30-5,
49.670.86
15.549.270.168.58
13.031.960.100.025
99.19
0.481
0.616.6
33.313.17.25.47.15.34.94.01.6
447A
40-42
49.610.86
15.079.170.168.58
12.412.060.12
98.04
0.483
0.717.4
—31.512.56.95.18.26.14.33.51.6
37-1,57-60
49.46a
1.1516.879.95—
3.4412.412.110.27
95.64
0.257
2.821.67
18.35
37.5511.26
3.717.915.229.23—
2.28
37-1,
50.160.91
15.648.940.188.80
12.741.930.05
99.35
0.496
0.316.3
33.912.26.94.8
10.07.03.52.71.7
50-53
50.270.92
16.249.460.188.23
11.662.290.07
99.32
0.465
—
0.419.4
33.810.05.44.39.37.44.03.51.7
15-2
49.720.92
16.589.400.188.08
11.782.310.24
99.21
0.462
—
1.419.5
34.210.15.44.36.45.15.85.11.7
49.850.96
16.539.520.167.89
11.682.310.21
99.11
0.453
—
1.219.5
34.19.95.34.47.46.14.94.501.8
449
17-1, 0-15
48.950.94
16.6210.000.188.15
11.652.370.200.022
99.08
0.449
—
1.220.1
34.19.95.24.44.13.47.77.21.8
49.590.99
16.759.500.188.02
12.142.460.24
99.87
0.458
—
1.420.8
34.011.05.24.73.22.67.76.81.9
i
36-3
48.82b
0.9015.5410.410.189.94
11.552.210.200.030
99.87
0.10
0.488
—
1.318.7
31.810.65.94.33.52.5
10.88.51.7
150
37-42
49.121.13
16.0110.470.107.4611.542.370.14
98.34
0.416
—
0.820.1
32.610.35.24.96.25.85.15.32.1
Note: * Indicates total Fe as FeO.a Altered glass." Melt inclusion in olivine.
Table 3. Microprobe analysis of plagioclase phenocrysts from DSDPLeg 59 basalts.
Table 4. Selective microprobe analysis of olivine phenocrysts fromDSDP Leg 59 basalts.
Hole
Sample(interval in cm)
Siθ2A12O3
FeOMgOCaONa 2OK2O
Total
Si
Al
Fe
Mg
CaNa
K
Total
An mol %
FeOMgO + FeO*
20-1,9-21
48.3432.270.440.32
16.342.00.03
99.74
2.2221.7490.0170.0220.8050.1780.002
4.994
81.9
0.58
447A
30-5,
47.5032.18
0.4
0.3717.3
1.610.02
99.38
2.1981.7550.0150.0260.8580.1440.001
4.997
85.6
0.52
40-42
49.8630.560.570.37
15.382.420.03
99.23
2.2971.6590.0220.0260.7590.2160.002
4.983
77.63
0.61
449
15^2,
50-53
48.2132.950.460.24
16.762.090.03
100.74
2.1981.7710.0180.0160.8190.1850.002
5.009
81.6
0.66
17-1,0-5
49.7631.710.440.23
15.502.280.04
100.00
2.2721.7070.0170.0160.7580.2020.002
4.975
78.96
0.66
46.3932.720.490.30
17.621.33—
98.87
2.1621.7970.0190.0210.8790.121
—
5.00
87.91
0.62
450
36-3, 37-42
49.5930.250.49
—
14.672.88
—
97.91
2.1321.6620.019
0.7330.261—
4.984
73.73
51.2031.050.490.30
14.422.750.02
100.27
2.3231.6610.0190.0210.7010.2420.002
4.968
74.17
0.62
phorus relative to Hole 447A basalts (Figs. 3, D and 8).Possible petrologic aspects of these differences are dis-cussed in following sections.
PETROLOGY
MOR-type tholeiites originate in the upper mantleunder rather narrow limits of temperature and pressure
Hole
Sample(interval in cm)
SiO2
TiO2
A1 2 O 3
C r 2 O 3
FeOMnOMgOCaO
Na 2ONiO
Total
SiTiAlCrFeMnMgCaNaNi
Total
Fe + 2
Fe + 2 + MgO
40.60n.d.a
n.d.n.d.11.730.20
47.900.31n.d.0.21
100.0
0.996_ b
0.2400.041.7520.008
0.004
3.004
0 120
n.d.n.d.n.d.n.d.
11.83n.d.
48.020.30n.d.n.d.
_
0.121
0.8750.004
—
-
0 121
447A
30-5, 40-42
n.d.n.d.n.d.n.d.
11.42n.d.
47.940.29n.d.n.d.
_
0.117
0.8790.004
—
-
0 118
n.d.n.d.n.d.n.d.
11.53n.d.
47.840.31n.d.n.d.
_
0.119
0.8770.004
—
-
0.119
40.33n.d.n.d.n.d.12.620.21
46.450.24n.d.0.22
100.02
1.002
0.2620.0041.7200.0060.050.005
2.999
0.132
39.81n.d.n.d.n.d.13.770.15
46.260.25n.d.0.19
100.39
0.991
0.2870.0031.7180.0070.040.004
3.010
0.143
449
17-1, 0-5
39.940.020.030.04
13.190.16
46.220.31n.d.0.20
100.09
0.9950.0000.0010.0010.2750.0031.7170.0080.040.004
3.0O4
0.138
40.050.020.030.03
13.170.16
46.340.31n.d.0.18
100.20
0.9950.0000.0010.0010.2740.0031.7170.008
0.004
3.003
0.138
a Not determined.b No data.
that are sustained during the entire formation ofOceanic Layer 2 (Kushiro, 1973; Fujii and Kushiro,1976-1977; Bender et al., 1978). The most detailedmodel of tholeiitic volcanism is based on statistical com-
674
PETROLOGY OF BASALTS
Table 5. Selective microprobe analysis of clino-pyroxene phenocrysts DSDP Leg 59 basalts.
Hole
Sample(intervals in cm)
SiO2
TiO2
A12O3
Cr 2 O 3
FeOMnOMgOCaONa 2O
Total
SiTiAlCrFeMnMgCaNa
Total
Fe + 2
Fe + 2 + Mg
447A
20-1, 9-21
52.180.402.660.555.880.12
17.1819.420.26
98.65
1.9320.0110.1160.0160.1820.0040.9490.7710.009
4.000
Λ A g:\
U.lol
450
36-3,
49.680.704.760.836.000.11
17.4619.900.31
99.74
1.8330.0190.2070.0240.1850.0040.9610.7870.022
4.043
(\ 1 C\
U.lol
37-42
n.d. a
n.d.n.d.n.d.6.37n.d.
17.3318.05
_ b
———
0.106—
0.5120.383—
Λ \Π\U.I /I
a Not determined.b No data.
positional data of basaltic lavas and volcanic glasses ofthe Atlantic Ocean floor and is confirmed by petrologicexperiment (Kushiro, 1973; Fujii and Kushiro, 1976-1977; Bender et al., 1978; Dmitriev et al., 1978). Thismodel is used here for comparison with the petrologicfeatures of the Philippine Sea basement tholeiites.
According to the model, primary tholeiitic melts ofthe Atlantic are formed in lherzolitic upper mantleunder dry conditions from spinel-bearing to plagioclase-bearing depths facies. Depths of the origin of primarymelts and compositional variation of their differentia-
tion products are related to the tectonic regime of evolu-tion of different segments of mid-ocean ridges. Owingto the differences of this regime, melts of the centralAtlantic (between the latitudes 25 °S and 36 °N) origi-nate at depths of 15 to 25 km and those of the northernand southern Atlantic at depths of 30 to 40 km. Appar-ently for the same reasons, tholeiites of the centralAtlantic are less differentiated than tholeiites of the nor-thern and the southern Atlantic (Kushiro, 1973; Fujiiand Kushiro, 1976-1977; Bender et al., 1978; Dmitrievet al., 1978; 1979).
Figure 5 is an Al2O3-Tiθ2 diagram that correlates theprimary tholeiitic melts and compositional limits of theAtlantic lavas with the composition of Philippine Seatholeiites (Tables 1 and 2). The figure shows that crystalfractionation of the Philippine Sea basaltic magmas israther restricted, because rocks with titanium contentless than 0.7% and rocks sharply enriched or impov-erished in alumina are completely absent. In contrast,the variation of titanium content is considerable. Thelowest concentrations of titanium are in fresh basalts(Unit 4, Fig. 1) and glasses of the lower part of the base-ment section of Hole 447A, as well as in fresh glasses ofHoles 449 and 54. These compositions fall within rela-tively narrow fields of the cotectic and anchieutecticcompositions of Atlantic tholeiites that originated atpressures near 10 kb and then cooled at low pressures(Dmitriev et al., 1979). Slightly raised titanium contentis characteristic of altered basalts of Hole 447A (Unit 2)of basalts from Holes 449, 450, and 54, and of PareceVela Rift basalts. Relatively high titanium content isestablished as well in altered glass of Hole 447A andglass samples of Hole 450 (Table 2).
Figure 6 shows the composition of the same wholerocks and glasses for the olivine-diopside-plagioclasesystem applied earlier to the model of tholeiitic volcan-ism of the Atlantic (Dmitriev et al., 1978, 1979). Crossesdenote the evolutionary path of tholeiitic liquid formedat 15.5 kb and then crystallized at 1 atm within tempera-ture limits of 1205° to 1270° (Bender et al., 1978). Thefigure illustrates good coincidence of compositionalvariation of whole rocks and glasses of the middle and
Table 6. Trace-element abundances of basalts from Holes 447A, 449, and 450 (ppm).
Hole
Core-Section,Interval (in cm)
RbBaSrYCrCoNiCuZnMnPTiV
14-2,41-43
6.831
12022
37949
10111698.5
1550572
6240248
15-1,24-26
4.927
14029
n.d.46
110129100
1720572
6300230
16-2,87-89
4,30
13029
33052
131116103
1350572
5580233
17-3,81-86
13.5259128
1915272
136110
1550704
7140250
18-1,71-70
12299320
175506466
1151570880
7140240
447A
22-3,32-40
5206033
25649
13310988
1480616
6240308
23-1,51-58
0.9257130
25247
13013496
1680572
6300320
23-3,73-83
1.0836531
2504515011490
1610572
6240264
24-2,25-39
0.326
11023
27554
18412069
1490484
4740170
25-1,104-111
0.425
13023
34560
23111381
1280528
4740250
28-3,52-58
4.5276932
45250
128164105
1370616
5580267
35-2,70-75
14n.d. a
11034
41037
1258576
1020616
4740261
449
15-2
4.5n.d.n.d.n.d.n.d.
36149102103
1450n.d.
6600304
17-2,41-51
643
13022
34036
176108
89124010566300
250
450
36-2,146-150
5.341
12024
32552926489
1590924
9600320
36-3,64-75
538
13023
33352
1168378
1470968
7560210
a Not determined.
675
G. S. ZAKARIADZE, L. V. DMITRIEV, A. V. SOBOLEV, N. M. SUSCHEVSKAYA
100Cr, Ni
200 300 400
K2O
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
113.0
295.0296.5
Figure 4. Variation of some trace-element contents along the basement section of Hole 447A. (Solid lines depict analytical data on Leg 59 basaltsobtained at the University of Birmingham; dashed lines represent data presented in Table 6.)
lower part of the basement section at Hole 447A (withthe variation curve denoted by crosses), except for thelower horizontal part of this curve. This part of thecurve represents the initial stage of liquid evolutionrelated to olivine fractionation from the primary melt,which is in equilibrium with mantle lherzolite at 10.5 kband then is rapidly uplifted to low pressures withoutconsiderable decrease of temperature. Matching experi-mental data with the major compositional trend of ba-salts and volcanic glasses of Hole 447A and the estab-lished distribution of titanium and alumina most likelyindicates that the origin of Site 447 basalts is related tothe differentiation of primary tholeiitic liquid, repre-senting partial melt of the upper mantle lherzolite at 10kb, which then crystallized at shallow depths along thecotectic. In this case, the composition of the volcanicglasses should be close to the composition of low-pres-
sure anchieutectic; and the composition of Mg-richaphyric rocks of Unit 5 should be close to the composi-tion of the melt that most nearly approached that ofprimary liquid. The differences of the last two are de-fined by crystallization of olivine at pressure limits from10.5 kb to 5 kb, corresponding to olivine crystallizationfield expansion (Bender et al., 1978). This suggestedcomplex differentiation trend is shown on Figure 6 asArrow I. The horizontal part of the arrow shows the dif-ferentiation path related to crystallization of excessiveolivine, whereas the vertical part, in line with the nor-mative clinopyroxene apex, is the path of cotectic crys-tallization of olivine and plagioclase at low pressure.
Basalts of Unit 4, with the closest to primary-meltcomposition display, have, as expected, low K, La, Ce,Nd, and Ti contents and high Ni content. Relative Srenrichment of Unit 4 basalts (Fig. 4) is consistent with
676
PETROLOGY OF BASALTS
O 11
• 12
® 13
α 14
• 15
Δ 16
A 17
0 18
f 19
O 20
® 21
Figure 5. Correlation of A12O3 and TiO2 contents of Atlantictholeiites (Dmitriev et al., 1979) and tholeiites of Philippine Seabasement. (Key: 1 = Primary melt of Atlantic tholeiites originatedat 10 kb; 2 = porphyritic basalts with cumulose olivine; 3 = por-phyritic basalts with cumulose plagioclase; 4 = glomerophyricbasalts (Ol-Pl); 5 = glomerophyric basalts (Pl-Ol-Cpx); 6 = fieldof cotectic crystallization; 7 = field of anchieutectic crystalliza-tion; 8 = primary melt of Atlantic tholeiites, originated at 3-5 kb;9 = sparsely phyric basalts (Pl-Ol-Cpx); 10 = field of anchieutec-tic crystallization; 11 = basalts of the lower part of the basementsection of Hole 447A; 12 = glasses from the lower part of thebasement section of Hole 447A; 13 = basalts of the upper part ofthe basement section of Hole 447A; 14 = basalts of Hole 449; 15= glasses of Hole 449; 16 = basalts of Hole 450; 17 = glasses ofHole 450; 18 = basalts of Hole 54; 19 = glasses of Hole 54; 20 =basalts of Dredge Site 1398 (from the seventeenth cruise of theDmitry Mendeleev; and 21 = altered glasses of Hole 447A.)
the early fractionation of olivine in the first stage of dif-ferentiation and with the cotectic crystallization ofolivine and plagioclase that followed. The relatively lowchromium content of Unit 4 basalts is possibly the resultof an earlier stage of irregular fractionation of spinel in-clusions with excessive olivine. It should be emphasized,however, that low chromium content and zirconiumenrichment of magnesium-rich and potassium-poorbasalts of Unit 4 (Fig. 4) is still to be explained.
The compositions of glasses of Holes 449, 450, and54 are distributed along a line that branches from thecotectic part of Trend I toward increase of normativeplagioclase content (Fig. 6, Arrow II). A12O3 and Na2Oare increased and CaO is decreased in this direction.Variation of other major elements is insignificant,although compared to that of Trend I of Hole 447A,glasses of the Parece Vela Basin are enriched with Ti, K,and total Fe and are lower in Mg. In this respect, glasseslying at the end of Arrow II (Fig. 6) are apparentlyrelated to the final low-temperature stages of differen-tiation. Compositional variation along Trend II is pos-sibly caused by the anchieutectic crystallization (Ol +PI + Cpx) of tholeiitic melt. The position of the transi-tional point from cotectic to anchieutectic crystalliza-tion cannot be precisely defined, but the shift of thispoint toward enrichment of glasses with normativeclinopyroxene should be coincident with decrease of
Figure 6. Ol-Pl-Di diagram for compositions of basalts and glasses ofPhilippine Sea floor compared to experimental data (Bender et al.,1978). (The key shows 1 = composition of the liquid phase oftholeiites, originated at 10.5 kb and crystallized at 1 atm (Bender etal., 1978); 2 = average composition of the most magnesian basaltsof the Hole 447A (Unit 4); 3 = average composition of volcanicglasses of Hole 447A; 4 = basalts of the upper part of the base-ment section of Hole 447A; 5 = plagioclase phyric basalts of Hole449; 6 = average composition of glasses of Hole 449; 7 =plagioclase-olivine phyric basalts of Hole 450; 8 = glasses of Hole450; 9 = aphyric basalts of Hole 54; 10 = glass of Hole 54; 11 =plagioclase-olivine-clinopyroxene phyric basalts of Dredge 1398;and 12 = altered glass of Hole 447A; L liquid; see text for ex-planations of Arrows I and II.)
pressure. The initiation of Trend II from approximatelythe middle of Trend I may indicate that the anchieutec-tic relations of Trend II had been attained at moderatepressures, between 10 kb and 1 atm. This case wouldcorrespond to crystallization of the melt originating at10 kb (Trend I) and cooling of the melt at moderatepressures in an intermediate-level magma chamber. Thehigh Al content of clinopyroxene phenocrysts from thebasalt of Hole 450 (Table 5) seems to confirm this sug-gestion.
We estimated the formation temperatures of Leg 59basalts, basing our figures on the well-known geother-mometry of olivine-melt equilibrium (Roeder and Em-slie, 1970). We obtained preliminary experimental dataof phenocryst-melt inclusion thermometry as well(Table 7). As shown in this table, calculated temper-atures of olivine-glass equilibria (Roeder and Emslie,1970) are noticeably lower than homogenization tern-
Table 7. Crystallization temperatures of tholeiites of
Holes 447A, 449, and 450, DSDP Leg 59.
Hole
Sample(intervals in cm)
T°C Ol/La
T°homCb
(inclus. in olivine)
447A
30-5,40-42
1220
449
17-1,0-15
1210
1250
17-1,0-15
1260
1250
450
36-3,32-42
1200
1240
a Calculated temperatures of olivine-glass equilibria."Homogenization temperatures of melt inclusions
olivine phenocrysts.
677
G. S. ZAKARIADZE, L. V. DMITRIEV, A. V. SOBOLEV, N. M. SUSCHEVSKAYA
peratures of melt inclusions of olivine phenocrysts. Thismay be caused by either of the following:
1) Calculated temperatures of olivine-glass equi-libria characterize the temperature of the system immed-iately before quenching and thus give an estimate of itsminimum value. In comparison, homogenization tem-peratures of melt inclusions from the central parts ofolivine phenocrysts represent earlier stages of the crys-tallization process that are higher in temperature.
2) Olivine-glass equilibrium temperatures were cal-culated for 1-atm pressure (Roeder and Emslie, 1970),but these temperatures increase with increasing pressureapproximately at 5°C per 1 kb (Vaganov and Kuznet-sov, 1976). Hence we conclude that the tholeiites westudied crystallized at pressures greater than 1 atm; thusthe calculated temperatures were slightly underestimated.The latter is apparently true for Hole 450 tholeiites con-taining clinopyroxene phenocrysts enriched with alu-mina. In any case, calculated as well as experimentallyobtained temperatures lay in limits established for drymelting conditions.
Basalts of Holes 449, 450, and 54 inherit composi-tions of respective glasses and are shown as parallel linesin Figure 6. For each of these lines Ti, Al, Ca, Na, andK increase and Si and Mg decrease in the direction fromglass to basalt. Total iron displays an irregular patternalong this direction; i.e., it may increase, decrease, orremain constant, although the FeOVMgO ratio dis-tinctly increases toward basalts. Established composi-tional variation from glasses to rocks cannot be ex-plained by phenocryst fractionation models. This varia-tion may be explained by low-temperature submarinealteration of basalts.
In addition to the basalts of Holes 449, 450, and 54,submarine alteration processes have considerably af-fected the upper part of the basement section at Hole447A (Units 1 and 2). Similar compositional changeshave been demonstrated recently by Honnorez et al.(1978) for the submarine alteration process of Atlantictholeiites. An analogous trend is characteristic for thetransition from fresh to altered volcanic glass at Hole447A (Fig. 6).
Fresh basalts and glasses of the Parece Vela basementreveal a tendency to relative enrichment with Ti, Ba,and P, compared to those of Hole 447A (Table 6; Figs.3, 5, 7, and 8). Marked enrichment of fresh glasses withpotassium (by a factor of approximately 2) is estab-lished from Hole 447A to Parece Vela sites. Variationlimits of other trace elements, and Ba/Sr, Ce/Y, andZr/Nb ratios, display no marked differences for PareceVela and West Philippine basin tholeiites, and all arewell within the variation limits of normal MORB. Hartet al. (1972) described the nearby Mariana Troughtholeiites (latitude 17°N), which had high titanium con-tent (1.25-1.66% TiO2) and exhibited a distinct ten-dency toward enrichment with K, Rb, Sr, Ba, and rareearth elements (REE). Thus the available data allow usto state that along the transect under discussion,tholeiitic basalts of the basement show a slight increasein alkaline tendencies from the western to the easternbasins of the Philippine Sea.
0.20
0.10
•V 5
17
\
\\
36
O 1
α 2IS 3
Δ 4
i 20 22
\p o
\\\\
37 J
0.40 0.42 0.44 0.46
MgO
0.48 0.50 0.52
MgO+FeO*
Figure 7. Variation of K2O in glasses of Holes 447A, 449, and 450.(The key is the same as for Fig. 2, except that the square with thedot designates a glass inclusion in an olivine phenocryst [Hole449].)
1000
900
800
700
600
500
× on-
Δ 36'-3
Δ36-2
O18-1
X
1 7 O X *
XX *
35 O * ® 2 2
14 1615OO Q O23
O 2 4
I 1 I 1 1 •_
0.3 0.4 0.5
MgO
0.6 0.7
MgO+FeO*
Figure 8. Distribution of phosphorus in West Philippine Basin andParece Vela Basin tholeiites. (The key is the same as for Fig. 3;asterisks indicate data from Hart et al. [1972].)
678
PETROLOGY OF BASALTS
DISCUSSION
The results of our petrologic investigation of thebasement basalts of the West Philippine and PareceVela basins show that these basalts are essentially in-distinguishable from normal (depleted) MOR tholeiites,although the basin basalts are slightly enriched with Ba(25-43 ppm) and display lower Zr/Nb ratios (10.7-22)(Kay et al, 1970; Kay and Hubbard, 1978, Sun et al., inpress). The process of evolution of tholeiitic magmatismduring formation of Oceanic Layer 2 of such basins oc-curs under conditions of dry basaltic systems and doesnot display any petrologic or geochemical evidence ofsubduction-zone influence (presence of water-bearingphases, enrichment with water and large-ion-lithophyle[LIL] elements, decrease of crystallization tempera-tures, etc.). In this respect, the geodynamic features ofbasement formation of these basins in the PhilippineSea plate are, for the most part, very similar to that ofmid-ocean ridges.
It is to be emphasized, however, that these petrologicdata characterize two different stages of basement for-mation of the Philippine Sea plate: the West PhilippineBasin (50-35 m.y.) and the Parece Vela Basin (30-13m.y.). It is interesting that according to geologic andgeophysical data, the geodynamic evolution of thesetwo basins may not be completely similar. If the PareceVela Basin is a good example of an extensional back-arcbasin, the history of the West Philippine Basin is muchmore complex (Karig, 1975; Scott et al., this volume).Several alternative models of development of the WestPhilippine Basin are still disputed: different versions ofrapid back-arc spreading; MOR-type spreading fol-lowed by entrapment to form marginal basins, andmodification of the basement by later intrusion or ex-trusion from a mosaic of crustal fissures are allpostulated models of formation (Karig et al., 1973;Karig, 1975; Louden, 1976; Hilde et al., 1977).
Although the number of basement sections in the pre-sent study is rather limited, the petrologic data we ob-tained do not support the idea of considerable dif-ferences in the geodynamic evolution of these two ba-sins. The distinctions between the evolutionary trends ofWest Philippine and Parece Vela tholeiites are mainlythe result of differences in upwelling and differentiationdynamics of primary tholeiitic liquids rather than theresult of change in the geodynamic situation. On onehand, the composition of Hole 450 basalts is completelyconsistent with the differentiation trend of Parece Velatholeiites and hence apparently reflects the process ofbasement formation. On the other hand, it is note-worthy that the consecutive stages of basement forma-tion of the Philippine Sea plate (West Philippine Basin,Parece Vela Basin, and Mariana Trough) are character-ized by slight but distinct increases of alkaline tenden-cies of basaltic volcanism. This phenomenon is not ex-plained by the differentiation schemes suggested earlierand possibly reflects the tendency toward increase ofdepth of the origin of the primary tholeiitic liquids or adecrease in the degree of partial melting of mantle lherz-holite. At the first approach, this suggestion seems to be
in accord with the tendency of decrease of silica contentand increase of normative olivine content of glassesfrom the West Philippine Basin to the Parece Vela Basin(Green and Ringwood, 1967), although the aluminacontent does not change considerably. Such a conclu-sion, however, may be true only for these few samplesof the South Philippine Sea transect that we studied. In-deed, the Philippine Sea basalts may display too great adiversity in composition to be consistent with the simplescheme suggested earlier. In particular, Marsh et al.(1978) came to the conclusion that a considerablechemical diversity of mantle sources fed the basalts ofShikoku and Daito basins.
CONCLUSIONS1) All the analyzed basement basalts of the basins of
the Philippine Sea (Hole 447A—West Philippine Basin—and Holes 449 and 450—Parece Vela Basin) are essen-tially indistinguishable from MOR-type tholeiites. Vol-canism of other types known for large oceanic plates oractive continental margins is completely absent in the in-vestigated basement sections.
2) All basalts studied represent differentiation pro-ducts of primary tholeiitic liquid originating from a drylherzholitic mantle at temperatures close to 1270°C andpressures at about 10 kb.
3) At the earliest stage of fractionation, within pres-sure limits from 10 to 5 kb, the primary liquid lost aconsiderable amount of the olivine that was excessive tocotectic crystallization. Further separation of crystal-lized phases did not take place, and because of this, allthese rocks represent the products of cotectic or an-chieutectic crystallization at pressures from 5 kb to 1atm.
4) The basalts of Hole 447A of the West PhilippineBasin are interpreted to be products of cotectic crys-tallization (Ol + PI) of tholeiitic melt at 1 atm, afterseparation of excessive olivine and rapid uplift of themelt at shallow depth without considerable decrease oftemperature. These basalts are represented by a rela-tively wide range of compositions.
Parece Vela tholeiites represent the products of an-chieutectic crystallization (Cpx + PI + Ol) that was ini-tiated at intermediate pressures (between 5 kb and 1atm). This may have happened as a result of delay ofupwelling magma in the intermediate magma chamberat depths from 3 to 5 km.
5) Basement basalts show a slight increase of alkalinetendencies from the West Philippine Basin to the PareceVela Basin, and thence to the Mariana Trough. Thismay reflect a tendency toward increasing depth of pri-mary liquid generation or a decreasing degree of partialmelting of the upper mantle source. Such a conclusion,however, may be true only for the South Philippine Seatransect studied on DSDP Leg 59.
REFERENCES
Bender, J. F., Hodges, F. N., and Bence, A. E., 1978. Petrogenesis ofbasalts from the project FAMOUS area: experimental study from0 to 15 k bars. Earth Planet. Sci. Lett., 41:277-302.
679
G. S. ZAKARIADZE, L. V. DMITRIEV, A. V. SOBOLEV, N. M. SUSCHEVSKAYA
Bougault, H., and Hekinian, R., 1974. Rift valley in Atlantic Oceannear 36°5O'N: Petrology and geochemistry of basaltic rocks.Earth Planet. Sci. Lett., 24:249-261.
Dietrich, V., Emmerman, R., Keller, G., et al., 1977. Tholeiiticbasalts from the Tyrrhenian Sea floor. Earth Planet. Sci. Lett.,36:285-296.
Dietrich, V., Emmerman, R., Oberhànsli, H., et al., 1978. Geochemi-stry of basaltic and gabbroic rocks from the West Mariana Basinand the Mariana Trench. Earth Planet. Sci. Lett, 39:127-144.
Dmitriev, L. V., Sobolev, A. V., and Suschevskaya, N. M., 1978.Primary melt of oceanic tholeiite and composition of upper mantleof ocean. Dokl. Akad. Nauk USSR, 240:177-180.
, 1979. Conditions of formation of primary melts ofoceanic tholeiites and variation of its composition. Geokhimiya,2:163-178.
Frey, F. A., Bryan, W. B., and Thompson, G., 1974. Atlantic Oceanfloor. Geochemistry and petrology of basalts from Legs 2 and 3 ofthe Deep Sea Drilling Project. J. Geophys. Res., 79:5507-5527.
Fujii, T., and Kushiro, I., 1977. Melting relations and viscosity of anabyssal tholeiite. Carnegie Inst. Washington, Yearb., 76:461-465.
Gill, J. B., 1976. Composition and age of Lau Basin and Ridgevolcanic rocks. Implications for evolution of an interarc basin andremnant arc. Bull. Geol. Soc. Am., 37:1384-1395.
Green, D. H., and Ringwood, A. E., 1967. The genesis of basalticmagmas. Contrib. Mineral. Petrol., 15:103-190.
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.
Hawkins, J. W., 1976. Petrology and geochemistry of basaltic rocksof the Lau Basin. Earth Planet. Sci. Lett., 28:283-297.
, 1977. Petrologic and geochemical characteristics ofmarginal basin basalts. In Talwani, M., and Pitman, W. C , III(Eds.), Island Arcs, Deep Sea Trenches, and Back-Arc Basins:Washington (Am. Geophys. Union), pp. 355-365.
Hilde, T. W. C , Uyeda, S., and Kroenke, L., 1977. Evolution of theWestern Pacific and its margin. Tectonophysics, 38: 145-165
Honnorez, J., Bohlke, I. L., and Honnorez, B. M., 1978. Petrologicaland geochemical study of the low temperature submarine altera-tion of basalt from Hole 396B, Leg 46. In Dmitriev, L., Heirtzler,J., et al., Init. Repts. DSDP, 46: Washington (U.S. Govt. PrintingOffice), 299-311.
International Working Group on the IGCP Project "Ophiolites"(Dmitry Mendeleev, June-August, 1976), 1977. Initial reports ofthe geological study of oceanic crust of the Philippine Sea floor.Ofioliti, 2:137-68.
Karig, D. E., 1971. Origin and development of marginal basins in theWestern Pacific. J. Geophys. Res., 76:2542-2561.
, 1975. Basin genesis in the Philippine Sea. In Karig, D. E.,Ingle, J. C , Jr., et al, Init. Repts. DSDP, 31: Washington (U.S.Govt. Printing Office), 857-879.
Karig, D., Ingle, J. C , Bouma, A. H., et al., 1973. Origin of the WestPhilippine Basin. Nature, 246:458-461.
Kay, R. W., and Hubbard, N. J., 1978. Trace elements in oceanridge basalts. Earth Planet. Sci. Lett., 38:95-116.
Kay, R., Hubbard, N. J., and Gast, P. W. 1970. Chemical character-istics and origin of oceanic ridge volcanic rocks. /. Geophys. Res.,75:1585-1613.
Kushiro, I., 1973. Origin of some magmas in oceanic and circum-oceanic regions. Tectonophysics, 17:211-222.
Louden, K. E., 1976. Magnetic anomalies in the West PhilippineBasin. In Sutton, G. H., Manghnani, M. H., and Moberly, R.(Eds.), The Geophysics of the Pacific Ocean Basin and Its Margin:Geophys. Monogr. Ser., 19: Washington, D.C. (Am. Geophys.Union), 253-267.
Marsh, N. G., Tarney, J., and Saunders, A. D., 1978. Geochemistryof basalts from the Shikoku and Daito Basins: IPOD Leg 58.Trans. Am. Geophys. Union, 59:1186. (Abstract)
Moberly, R., 1972. Origin of lithosphere behind island arcs, withreference to the Western Pacific. Geol. Soc. Am. Mem., 132:35-55.
Molnar, P., and Atwater, T., 1978. Interarc spreading and cordillerantectonics as alternates related to the age of subducted oceaniclithosphere. Earth Planet. Sci. Lett., 41:330-340.
Packham, G. H., and Falvey, D. A., 1971. An hypothesis for theformation of marginal seas in the Western Pacific. Tectono-physics, 11:79-109.
Ridley, W. I., Rhodes, I. M., Reid, A. M., et al., 1974. Basalts fromLeg 6 of the Deep Sea Drilling Project. Part 1. /. Petrol.,15:140-159.
Roeder, P. L., and Emslie, R. F., 1970. Olivine-liquid equilibrium.Contrib. Mineral. Petrol., 2:275-289.
Sclater, J. G., Karig, D., Lawver, L. A., et al., 1976. Heat flow,depth, and crustal thickness of the marginal basins of the SouthPhilippine Sea. /. Geophys. Res., 81:309-318.
Sun, S. S., Nesbitt, R., and Sharaskin, A., in press. Geochemicalcharacteristics of mid-ocean ridge basalts. Earth Planet. Sci. Lett.
Tarney, J., Saunders, A. D., and Weaver, S. D., 1977. Geochemistryof volcanic rocks from the island arcs and marginal basins of theScotia arc region. In Talwani, M., and Pitman, W. C , III (Eds.),Island Arcs, Deep Sea Trenches, and Back-Arc Basins: Wash-ington (Am. Geophys. Union), pp. 367-377.
Toksöz, M. W., and Bird, P., 1977. Formation and evolution ofmarginal basins and continental plateaus. In Talwani, M., and Pit-man, W. C. Ill (Eds.), Island Arcs, Deep Sea Trenches, and Back-Arc Basins: Washington (Am. Geophys. Union), pp. 379-393.
Vaganov, V. I., and Kuznetsov, I. E., 1976. Determinations of tem-peratures of crystallization of olivine for ultrabasic and basicrocks. Modern Methods of Petrologic Investigations: Moscow(Nauka), pp. 142-157.
680
