deep sea drilling project initial reports volume 89
TRANSCRIPT
17. PETROLOGY AND GEOCHEMISTRY OF OCEANIC INTRAPLATE SHEET-FLOW BASALTS,NAURU BASIN, DEEP SEA DRILLING PROJECT LEG 891
P. A. Floyd, Department of Geology, University of Keele2
ABSTRACT
Reentry of Hole 462A during Leg 89 resulted in the penetration of a further 140 m of basalt sheet-flows similar tothose found during Leg 61 at the same site. Twelve volcanic units (45 to 56) were recognized, comprising a series of rap-idly extruded, interlayered aphyric and poorly clinopyroxene-plagioclase-olivine phyric, nonvesicular basalts. All ex-hibit variable, mild hydration and oxidation, relative to fresh oceanic basalts, produced under reducing, low-CO2-activ-ity conditions within the zeolite facies. Secondary assemblages are dominated by smectites, zeolites, and pyrite, pro-duced by low-temperature reaction with poorly oxygenated seawater. No systematic mineralogical or chemical changesare observed with depth, although thin quenched units and more massive hypocrystalline units exhibit slightly differentalteration parageneses.
Chemically, the basalts are olivine- and quartz-normative tholeiites, characterized by low incompatible-elementabundances, similar to mildly enriched MORB (approaching T-type), with moderate, chrondite-normalized, large-ion-lithophile-element depletion patterns and generally lower or near-chrondritic ratios for many low-distribution-coeffi-cient (KD) element pairs. In general, relative to cyclic MORB chemical variation, they are uniform throughout, al-though 3 chemical megagroups and 22 subgroups are recognized. It is considered that the megagroups represent sepa-rate low-pressure-fractionated systems (olivine + Plagioclase ± clinopyroxene), whereas minor variations within them(subgroups) indicate magma mixing and generation of near-steady-state conditions. Overall, relatively minor fractiona-tion coupled with magma mixing produced a series of compositionally uniform lavas. Parental melts were produced bysimilar degrees of partial melting, although the source may have varied slightly in LIL-element content.
INTRODUCTION
Extensive areas of the central and western Pacific Oceanfloor are covered by the volcanic products of Cretaceousintraplate volcanism (Winterer, 1973; Larson and Schlan-ger, 1981; Haggerty et al., 1982). The Nauru Basin isone such area (Fig. 1), and drilling at Site 462 duringDSDP Leg 61 demonstrated the existence of a thick LowerCretaceous basaltic sill-flow complex (Larson, Schlang-er, et al., 1981). Magnetic lineation patterns for the west-ern Pacific indicate that the complex rests on Jurassicoceanic basement (Larson, 1976; Hilde et al., 1977; Lar-son and Schlanger, 1981). On the basis of seismic reflec-tion data, the Nauru intraplate volcanic province extendsover some 400,000 km2 (Fig. 1) and probably has a vol-ume comparable to many continental flood basalt prov-inces (Tokuyama and Batiza, 1981).
Hole 462A was reentered during Leg 89 and penetrat-ed farther into the sill-flow complex, but did not reachJurassic oceanic basement below. Combining data forLegs 61 and 89, the sill-flow complex has a minimumthickness of 650 m and is composed predominantly of alower unit of basaltic sheet-flows and an upper unit ofsupposedly intrusive doleritic sills. Although sedimenta-ry interlayers with suitable fossil debris are few, faunalevidence, coupled with radiometric dating, indicates em-placement of some of the complex within the 115-110Ma interval (Schlanger and Premoli Silva, 1981; Ozimaet al., 1981), with the stratigraphically lowest part sam-pled during Leg 89 being lower Aptian or older (see Cas-
Cretaceous submarinesheet-flow complexes
Japan
;• Bonin Is.
• Marianas Is.
/G. E
Australia
Moberly, R., Schlanger, S. O., et al., Init. Repts. DSDP, 89: Washington (U.S. Govt.Printing Office).
2 Address: Dept, of Geology, University of Keele, Staffordshire, ST5 5BG U.K.
Figure 1. Locality map of the western Pacific showing position ofDSDP Site 462 and extent of Cretaceous intraplate sheet-flows (afterWinterer, 1973).
tillo et al., and Schlanger and Moberly, this volume, fordiscussions of age of the basalts).
Chemically and mineralogically, the basalts are notunlike some types of mid-ocean ridge basalts (MORB),with LIL-element patterns generally depleted relative tochondrites, and low SR isotope ratios, but they are suf-ficiently different in a number of respects to set themapart (Batiza et al., 1980; Batiza, 1981; Tokuyama and
471
P. A. FLOYD
Batiza, 1981; Fujii et al., 1981; Saunders, and Castilloet al., this volume). The sheet-flows form a voluminous,chemically distinctive form of submarine intraplate vol-canism, generated on preexisting ocean crust and de-rived from different parent magmas (and/or sources)from MORB.
This chapter describes the petrography and geochem-istry of 75 basalt samples collected during Leg 89 (Table 1)from the deepest part of the sill-flow complex at Hole462A, between 1072 and 1209 m sub-bottom depth.
ANALYTICAL METHODS
Major- and trace-element analysis was performed using a PhilipsPW1212 X-ray fluorescence (XRF) spectrometer calibrated against in-ternational and internal University of Keele standards of appropriatecomposition. Major elements were analyzed in a fused glass disc (1:5mixture of ignited rock powder and Li metaborate flux) using methodand interelement correction procedures modified after Norrish and Hut-ton (1969). Trace elements and Na2O were analyzed in a pressed pow-der pellet using the method of Leake et al. (1969). Ferrous Fe was de-termined by titration with potassium dichromate on a separate aliquotof sample. H 2 O + and CO2 were determined together by HNC Analyz-er. Th, U, Ta, and Hf data were obtained for about half the samples byinstrumental neutron activation analysis (Duffield and Gilmore, 1979)at the Universities Research Reactor, Risley, U.K.
Analytical precision (measured by the coefficient of variation; Ta-ble 2) was evaluated by analysis of 10 separate samples of an internalstandard (KUM-1), and gave ±1-4% for major and minor elements(including FeO) and < ± 5 % for most trace elements, except La, Ce,and Nb (± 10%). Light rare-earth elements (REE) analyzed by XRFspectrometry were checked by inductively coupled plasma emission (ICP)spectrometry at King's College, London (Walsh et al., 1981), and gavereasonable agreement at low concentrations, whereas XRF La values> 5 ppm and XRF Ce values > 15 ppm were about 20% higher thanthe corresponding ICP values. Full REE data will be published else-where; only the XRF light REE results are presented in the analyticaltables.
Analytical accuracy was monitored against international basalt stan-dards BCR-1 and BOB-1 (Table 2), and showed reasonable agreement.Selected Leg 89 samples were also analyzed in other laboratories bysimilar XRF techniques, and showed that the Keele data were gener-ally comparable, except that Ba was erratic and probably unreliable atlow concentrations, whereas Sr, TiO2, and total Fe (as Fe2O3) were sys-tematically high by about 15, 4, and 2%, respectively. Ta abundanceswere found to show greater variability than expected, relative to otherincompatible elements; high values probably reflect contamination dur-ing crushing in a tungsten carbide swing mill.
Normative calculations were performed on analyses recalculated toan anhydrous and carbonate-free basis and a standard oxidation ratioof 0.15 (Brooks, 1976) to reduce the effects of oxidation on the norm.
Modal analysis by automatic point counter was carried out on rep-resentative samples; about 2000 points were counted in most thin sec-tions. An error of about ± 3 % on each mineral constituent (chart inBarringer, 1953) is probably typical, except for some of the very fine-grained quench-textured samples, where identification was difficult.
GENERAL CHARACTERISTICS OF VOLCANICUNITS
Additional to the 44 volcanic units recognized duringLeg 61 (Larson, Schlanger, et al., 1981), a further 12 ba-saltic units (numbered 45 to 56) were recorded withinthe 140-m section from Cores 462A-93 to 462A-109 ofLeg 89 (Fig. 2). Although further examination has con-firmed the presence of subunits, mineralogical changesare relatively minor, and no change has been made inthe number of units previously recorded (Site 462 chap-ter, this volume), except that Unit 51 is no longer consid-ered to represent pillow lavas. Petrography of the mainbasaltic types is described in a subsequent section.
Table 1. Keele sample numbers, corresponding DSDP samplenumbers, and related data of chemically analyzed basalts fromNauru Basin, Hole 462A.
Keelenumber
462A-1462A-2462A-3462A-4462A-5462A-6462A-7462A-8462A-9462A-10462A-11462A-12462A-13
462A-14462A-15462A-16462A-17462A-18462A-19462A-20462A-21
462A-22
462A-23462A-24462A-25462A-26462A-27462A-28462A-29462A-30462A-31462A-32462A-33462A-34462A-35462A-36462A-37462A-38462A-39462A-40462A-41
462A-42462A-43462A-44462A-45462A-46
462A-47462A-48462A-49
462A-50462A-51462A-52462A-53462A-54
462A-55462A-56462A-57462A-58462A-59462A-60462A-61462A-62462A-63462A-64462A-65462A-66462A-67
462A-68462A-69
462A-70462A-71
462A-72462A-73462A-74
462A-75
Sample(interval in cm)
93-1, 35-4093-2, 43-4894-1, 8-1394-2, 24-2994-2, 79-8494-3, 126-13094-5, 21-2694-6, 85-9095-1, 93-9895-3, 20-2595-4, 75-8095-5, 45-5095-5, 61-63
95-5, 76-8095-5, 99-10495-7, 38-4396-2, 122-12797-1, 71-7697-5, 69-7498-6, 5-1098-6, 41-46
99-2, 6-11
100-2, 61-66101-1, 6-10101-1, 108-113101-2, 43-47101-2, 117-122101-3, 35-40101-3, 123-128101-4, 57-61101-4, 115-120101-5, 52-56102-1, 35-39102-1, 112-116102-2, 54-59102-2, 134-139102-3, 82-86102-3, 143-147102-4, 11-15102-4, 51-55102-4, 59-63
102-4, 95-100102-4, 131-136102-5, 5-9102-5, 33-36103-1, 24-29
103-1, 68-73103-1, 107-112104-1, 30-34
104-1, 40-44104-1, 67-71104-1, 81-85104-2, 2-3104-2, 128-133
105-1, 52-56105-1, 98-102105-2, 58-63105-2, 122-126105-3, 50-54106-1, 36-40106-1, 129-133106-2, 106-110106-3, 4-8106-3, 47-51108-1, 17-22108-1, 79-84108-2, 40-44
108-2, 66-70108-2, 125-128
108-3, 28-30108-3, 75-78
109-1, 113-117109-2, 39-43109-2, 144-149
109-3, 112-116
Sub-bottomdepth(m)
1072.081073.661076.711078.371078.921080.881082.841084.981086.661088.931090.981092.181092.32
1092.481092.721095.111097.641104.741110.721120.781121.14
1123.89
1133.641134.581135.601136.451137.201137.881138.761139.591140.171141.041143.271144.041144.961145.771146.741147.351147.531147.931148.01
1148.381148.741148.971149.241152.27
1152.711153.101161.52
1161.621161.891162.511162.731164.01
1170.841171.301172.401173.041173.821174.681175.611176.881177.361177.791189.701190.321191.42
1191.681192.27
1192.791193.27
1199.851200.611201.67
1202.84
Vole.unit
45
46
47
48
49
50
51
52
53
CAj 4
55
56
Texture( modal data)
*G.G.G.
•G.G.G.G.G. Vr.G.
•G. Vr.G.G. Vr.
•Q.
Q.•QG. I.G,
*G.G.
•G.•Q.
•Q.
•I. V.II1
•I
V.
Q. V.Q.I.
*ilI
•IIII
•G.G.
•Q.*Q.
Q•Q. V.Q. S.Q
•Q.
Q v.•Q. v.G. Vr.
Q. V.•Q. S.•Q. V.Q. V.
•Q. V.
•Q. V.•Q. V.Q. V.I.1.I.
*Q.V.
•V. G.V.I.
•V.V. G.
Q. V.•G.
Q. V.•G.
G.*G.Q. V.
*Q.
Lithology
ol-pl-cpx s-phyric Bol-pl-cpx s-phyric Bol s-phyric Bol-cpx-pl s-phyric Bol s-phyric Bpi s-phyric Baphyric Baphyric Bol-s-phyric Bpi s-phyric Bol-pl s-phyric Baphyric Bpl-cpx phyric B
pl-cpx phyric Bol-pl-cpx phyric Bol m-phyric Bcpx-pl m-phyric Baphyric Bcpx-pl-ol s-phyric Bcpx-pl-ol s-phyric Bpl-cpx phyric B
pl-ol-cpx phyric B
aphyric Baphyric Baphyric Baphyric Baphyric Bcpx s-phyric Baphyric Baphyric Baphyric Baphyric Baphyric Baphyric Baphyrix Baphyric ol (?) Baphyric Baphyric ol (?) Bol(?)-cpx s-phyric Bol-pl-cpx m-phyric Bcpx-pl-ol m-phyric
ol-pl-cpx m-phyric Bol-pl-cpx s-phyric Bol-pl-cpx phyric Bol-pl-cpx s-phyric Bol-pl-cpx s-phyric B
ol-pl-cpx phyric Bpl-cpx m-phyric Bcpx s-phyric B
pl-cpx m-phyric Bol-pl-cpx m-phyric Bpl-cpx-ol s-phyric Bcpx-pl m-phyric Bcpx-pl m-phyric B
pl-cpx m-phyric Baphyric Baphyric Baphyric Baphyric Baphyric Baphyric Baphyric Daphyric Daphyric Daphyric Baphyric Baphyric B
pl-cpx m-phyric Bpl-cpx s-phyric B
pl-cpx m-phyric Bpl-cpx m-phyric B
pl-cpx m-phyric Bpl-cpx s-phyric Bpl-cpx phyric B
pl-cpx m-phyric B
Note: pi - Plagioclase, cpx = clinopyroxene, ol = olivine; s-phyric = sparsely phyric, m-phyric= moderately phyric. B = basalt, D = dolerite. Textures: Vr. = vesicular, I. = intersertal,G. = granular, V. = variolitic, S. = spherulitic, Q. = quench (skeletal and/or plumosecrystals ± altered glass).
472
PETROLOGY AND GEOCHEMISTRY
Table 2. Accuracy, precision, and detection limits of Keele internalstandard (KUM-1) and determined values for international stan-dards BCR-1 and BOB-1.
SiO2
TiO2
A12O3
Fe2O3
FeOMnOMgOCaONa2OK2OP2<KH2O +
BaCeCrLaNbNdNiRbSeSrYZr
X
46.402.61
14.921.318.760.258.386.683.601.440.553.47
61981
259373827
1682825
51331
209
KUM-1
S. d.
0.130.050.210.020.040.010.880.110.160.010.020.13
13.28.03.23.12.02.43.61.50.54.81.23.9
C. v.
0.31.91.41.50.42.80.91.64.40.83.63.7
210
1859252142
D. 1.
———————_——
—
3571525233427
BCR-1
S. r.
54.832.21
13.573.498.830.183.566.883.201.790.350.75
6785820291128174531
34938
189
R
54.502.20
13.613.688.800.183.466.923.271.700.360.77
6755418261329164733
33037
185
BOB-1
S. r.
50.381.35
16.193.125.490.157.63
11.683.010.390.150.40
5217
25086
11108
532
20826
105
R
50.601.45
14.413.015.360.178.45
11.233.170.310.15
-
6015
28076
1055
20128
115
Note: x = mean of 10 replicates; S. d. = standard deviation; C. v. = % coefficientof variation; D. 1. = detection limits; S. r. = single result; R = recommendedvalues.
A summary of the features defining the upper andlower boundaries of the volcanic units is given in Table 3;criteria for the recognition of separate units are basedon (1) volcaniclastic sediment interlayers, (2) develop-ment of quenched glassy or chilled margins, and (3) mi-neralogical and textural variations. The presence of threethin reworked hyaloclastite layers between Units 45 and46, 46 and 47, and 47 and 48 provides clear evidence ofminor lulls in eruptive activity, confirmed by significantbreaks in stable magnetic inclinations at the same bound-aries. In general, the observed volcanic units correlatewell with the determined magnetic units, as shown inFigure 2. The nongrouping in Magnetic Unit 2 probablyreflects the presence of a number of subunits observedat the top of Volcanic Unit 46.
All the volcanic units are basaltic and considered torepresent sheet flows rather than pillow lavas or sills(Site 462 chapter, this volume), and as such are a contin-uation downward of the type B flows of Leg 61 (Batiza,1981). In general terms, they show a crude alternationof relatively thick aphyric holocrystalline and hypocrys-talline basalt units and thinner, variably phyric, gener-ally quench-textured units. The aphyric units invariablyincrease in grain size toward the center, and are charac-terized by intersertal and granular textures, althoughwedge-shaped skeletal or serrated Plagioclase is commonthroughout. Any interstitial glass is always altered to pa-lagonite, and may be crowded with magnetite granules.These units have cooled relatively slowly, not only be-cause they are thick and massive, but because they wererapidly blanketed and insulated by successive flows, asthe lack of reworked or oxidized tops suggests. On theother hand, the phyric units contain glass, skeletal cli-
nopyroxene, and Plagioclase microlites, and exhibit var-ious quench textures throughout, indicative of rapid butvariable rates of cooling. Variolitic textures composedof clinopyroxene and Plagioclase intergrowths are typi-cal, whereas true spherulites are only occasionally pres-ent, preserved in glassy margins. Glomerocrystic groupsof clinopyroxene and Plagioclase are common; isolatedmicrophenocrysts of olivine are rarer, and appear to beconcentrated more in the upper volcanic units than theother phenocryst phases. The phyric units represent pack-ets of rapidly extruded thin lava flows derived from amagma chamber or chambers that had undergone somedegree of crystal fractionation. Mineralogically, they aredominated by clinopyroxene, which may exhibit threemorphologies within the one flow: large subhedral crys-tals in glomerocrystic groups; smaller, equigranular grains;and quenched plumose or sheaf variolites. Taken in or-der, these forms may represent three crystallization epi-sodes—in the magma chamber, during subsequent up-ward transport toward the surface in feeder channels,and finally on rapid quenching by seawater.
Overall, the paucity of sediment interlayers, thick flowbreccias, and hyaloclastites, together with the relativelyfresh nature of the lavas, suggests rapid effusion of com-paratively fluid basalt that produced a lava pile at a mi-nimum rate of about 130 m/m.y. The widespread devel-opment of the lavas in the Nauru Basin and their rela-tively uniform character suggest that they are sheet flowsderived from fissures rather than central volcanoes.
MINERALOGICAL AND CHEMICAL ASPECTSOF ALTERATION
This section describes the general petrographic fea-tures and overall chemical effects of alteration, as re-lated to the mode of occurrence, depth in the lava pile,and physical parameters. The chemistry of selected sec-ondary phases is outlined by Floyd and Rowbotham (thisvolume); the lack of extensive and progressive alterationprecludes detailed comment on chemical effects often seenin altered ocean-floor basalts (e.g., Hart et al., 1974;Humphris and Thompson, 1978).
General Comments
The degree of alteration in the basalt flows is gener-ally low. Variably colored smectites and palagonite arethe main alteration products; zeolites, celadonite, car-bonate, quartz, and K-feldspar are not very common,and Fe oxides are uncommon or absent (Table 4). Pyriteis abundant as fracture fillings, and is invariably associ-ated with a dark green smectite. Other vein material con-sists of zeolites, calcite, and quartz. The general lack ofvisible patchy discoloration caused by alteration, or ofhaloes and zones adjacent to fractures, is principally aconsequence of the nonoxidative nature of the basalt al-teration and thus the paucity of red Fe oxides. Chemi-cally the basalts are mildly oxidized and hydrated, where-as the CO2 content is very low and reflects the lack ofsecondary replacing carbonate in the rock matrix. Mi-neralogically and chemically the alteration is typical oflow-temperature, post-solidification reaction with coldseawater under largely nonoxidative, zeolite-facies con-
473
P. A. FLOYD
Lithology and main features
ag
ne
iu
nit
1 0 8 0 -
1 2 0 0 -
Aphyric to sparsely clinopyroxene—Plagioclase phyric basalt• Glomerophyric glassy basalt near top of unit• Olivine becomes phenocryst phase toward base in
subvariolitic glassy basalt• Numerous closely spaced horizontal fractures at base
Moderately olivine—clinopyroxene phyric basalt• At top of unit below hyaloclastite layer is quench-textured
moderately phyric basalt• Patchy development of glass-rich zones (glass altered to smectite)
throughout suggest unit may be composed of a number of smallerflow units
• Plagioclase phyric quench-textured subunit at base withchilled top
Olivine—clinopyroxene phyric basalt• Oxidized reworked volcaniclastic sediments at top and bottom of unit• Top of flow represented by spalled glass in smectite—zeolite matrix
Aphyric basalt• Often glassy, now represented by smectite patches throughout• Grades downward into sparsely olivine—clinopyroxene phyric basalt• Fine-grained thin subunit near base
Clinopyroxene—Plagioclase phyric basalt• Glomerophyric texture; glassy quench-textured subunits present
Plagioclase—clinopyroxene—olivine phyric basalt• Glassy throughout and quench-textured• Vesicular or vuggy base on irregular surface of unit below
Moderately clinopyroxene—olivine ± Plagioclase phyric basalt• Number of small cooling units• Quench-textured throughout
Aphyric glassy basalt• Quench-textured top and bottom to unit• Various sparsely phyric variants in central portion• Grain-size increases downward from basalt to fine-grained dolerite
Clinopyroxene—Plagioclase phyric basalt
Clinopyroxene—Plagioclase phyric basalt
• Number of thin quench-textured units present
Moderately clinopyroxene—Plagioclase phyric basalt
Moderately clinopyroxene—Plagioclase phyric basalt• Quench-textured; grain-size variation suggests thin subunits
I7I
3
µ
15
16
Λ Λ Λ
Basalt
Dolerite
ReworkedHyaloclastite
Figure 2. Lithologic and magnetic log, Hole 462A.
× × Clinopyroxene-phyric
o o Plagioclase-phyric
• Olivine-phyric
Quench-textured
474
PETROLOGY AND GEOCHEMISTRY
Table 3. Thickness, location, and nature of the boundaries between volcanic units, Hole 462A.
Sub-bottom Boundary locationThickness depth (Core-Sec,
Unit (m) (m) interval)
Nature ofboundary3
Boundary characteristics £ _
45 >20.4
46
47
48
49
50
51
52
53
54
55
31.1
8.1
16.5
4.6
9.0
8.8
21.3
0.6
7.1
3.0
1092.3
1123.4
1131.5
1148.0
1152.6
1161.6
1170.4
1191.7
1192.6
1199.7
1202.7
95-5, 65 cm
99-1, 112 cm
99-2/100-1
102-4, 64 cm
103-1, 55 cm
104-1, 42 cm
105-1, 7 cm
108-2, 71 cm
108-3, 12 cm
109-1, 104 cm
109-2/109-3
t = glassy basalt (Unit 44, Leg 61, med.-gr. dolerite)b = quench texture, phenocryst density decrease
t = hyaloclastite, quench texture in basalt belowb = quench texture
Reworked volcaniclastic sedimentt = spalled glassy flow top to basaltb = ?
Reworked volcaniclastic sediment
56 >0.3
t = ?b = quench texture, cpx-ol phyric basalt
t = quench texture, cpx-plag phyric basaltb = fine-grained ol-cpx phyric basalt
t = quench texture, three-phase phyric basaltb = vesicular base
Curved irregular contactt = chilled margin, quench textureb = phyric basalt, quench texture
t = aphyric basalt, quench textureb = quench texture
t = glass-coated quenched basalt pebble (breccia?)b = phyric glassy basalt
t = quench texture, grain-size decrease downwardb = chilled margin
Inclined contactt = quench textureb = ?
t = rapid grain-size variation, finer than Unit 55
A = boundary visible within a single portion of core;(poor recovery between sections or cores).
B = adjacent but individual pieces within same section; C = inferred
ditions. Similar low-grade submarine "weathering" atseveral DSDP sites drilled in oceanic basement has beendescribed, although the degree of oxidation is often morevariable and extensive (e.g., Andrews et al., 1977; Prit-chard et al., 1979; Humphris, Thompson, and Marriner,1980).
The low degree of alteration, even at some flow junc-tions, is probably a consequence of effusion of lava sorapid and continuous that successive flows effectivelyinsulated those below from prolonged reaction with sea-water after quenching. Also, the lack of channelways,such as clastic sediment interlayers, breccia zones, or ex-tensive fracturing in the basalt pile, would restrict thecirculation of large quantities of seawater. But seawatermust have initially penetrated the pile via lava flow bound-aries and subsequently through subhorizontal and near-vertical open fracture systems until they became filledwith deposited material.
Chemical Features
The distribution of chemical parameters typifying al-teration (Fe2O3/FeO, H 2 O + , CO2) is shown in Figure 3,and demonstrates the more hydrated and carbonated na-ture of the upper portion of the lava pile sampled onLeg 61. The average Fe2O3/FeO ratio (0.5) and H2O +
content (1.2 wt.%) of the Leg 89 samples are higher thanin fresh MORB (<0.3 Fe2O3/FeO [Miyashiro et al., 1969]and 0.2-0.4 wt.°7o H2O [Moore, 1970; Moore and Schil-ling, 1973]) or Hawaiian intraplate basalts (0.5-0.7 wt.%
H2O [Delaney et al., 1978]). As shown in Figure 4, thereexists between the oxidation ratio and hydration a crudepositive relationship more typical of low-temperature sub-marine alteration than of higher-temperature hydrother-mal metamorphism, which is characterized more by high-er and progressive hydration (Miyashiro et al., 1971).
The highly variable K/Rb ratio and its trend relativeto K (Fig. 5) are also characteristic of low-grade altera-tion, and are mirrored by young ocean crust (Floyd andFuge, 1982). The low-temperature uptake of K by oce-anic basalts and, in particular, by glassy pillow lava sel-vages, is well documented (e.g., Hart and Nalwalk, 1970;Hart et al., 1974), and occurs mainly in altered glass,abundant clays, and, to a lesser extent, in minor K-bear-ing zeolites and K-feldspar. The highest K values in theLeg 89 basalts correspond to plagioclase-rich alterationdomains within holocrystalline basalt near the top ofUnit 52, and reflect the development of much brownsmectite, celadonite, and minor secondary K-feldspar.
Extensive hydration under low-temperature conditionscauses many elements to be mobile and to exhibit pro-gressive gains or losses relative to fresh materials (Hum-phris and Thompson, 1978). The main chemical effectsof alteration in Leg 89 basalts are the development of er-ratic or scattered nonmagmatic distributions of K, Rb,Ba, Sr, and possibly U, and, to a lesser extent, of totalFe, MgO, and Ni. Other elements appear to be stable,with element pairs showing good coherence or distribu-tions that can be attributed to magmatic processes. The
475
P. A. FLOYD
15
α>S 10
30 -
0.36 0.48 0.60
Fe2O3/FeO
50
30
^ Leg 61 data
Q Leg 89 data
Figure 3. Histograms comparing distribution of alteration parameters in Leg 61 (data from Larson, Schlanger, et al., 1981) and Leg 89 (thiswork) samples.
oCM
X 1
Hydrothermalalteration trend
Submarineweathering trend
Fresh MORB
0.5
Fe2O3/FeO
1.0
Figure 4. Hydration (H 2 O + ) and oxidation (Fe2O3/FeO) of Hole 462Abasalts relative to typical trends caused by low-temperature subma-rine weathering and high-temperature hydrothermal alteration ofoceanic basalts (Floyd and Tarney, 1979).
degree of alteration is too low and elemental variationtoo limited to permit recognition of any definite trendswith progressive hydration for the mobile elements. It isthe development of smectite that causes hydration (to-gether with minor increases in K), and when smectite re-places olivine, the Ni is redistributed. Other major ox-ides are generally undisturbed, although the FeOVMgO
000
800
600
400
200
Λ/ \
/ \/ \
//I Reykjanes Ridge lavas
••J*•
\\\\\
\ '^ \\ ̂
^ \
J\
i
—
\\ -
\\
I1/
1500 1000 5000 10000
K (ppm)
Figure 5. Typical low-temperature variation of K/Rb and K in Hole462A basalts relative to Reykjanes Ridge basalts (low-grade altera-tion) and eastern Iceland lava flows (higher-grade hydrothermal al-teration) (Floyd and Fuge, 1982).
ratio, often used as differentiation index, decreases withprogressive hydration (especially in Units 46 and 48), suchthat the most altered basalts are apparently characterizedchemically as the least fractionated with low FeOVMgO.
At such low grades of alteration, net gains and lossesare probably relatively minor, since many mobile elements
476
PETROLOGY AND GEOCHEMISTRY
will be redistributed into stable secondary phases, suchas clays, zeolites, and carbonates (cf. Wood et al., 1976;Floyd and Fuge, 1982). Elemental variation is essential-ly nonsystematic in Leg 89 basalts, and reflects not theoverall alteration intensity, but more the development ofspecific secondary minerals, especially smectites.
Secondary Mineral Distribution
Secondary minerals are developed in the following lo-cations in Leg 89 basalts: glass in quenched flow mar-gins; interstitial glass in hypocrystalline units, pheno-crysts, plagioclase-rich interiors of holocrystalline units,as infilling in vesicles, and as vein material in fractures.The distribution and composition of secondary miner-als in both rock matrix and veins encountered on Leg 61have been commented on by Kurnosov et al. (1981).
The quenched margins of both massive and thin flowunits invariably show the development of yellow-brownpalagonite and/or dark brown smectite between plumosevariolites of clinopyroxene; fresh glass is generally notpresent. At the top of Unit 46, brown smectite is partial-ly replaced by coarse green flakes of celadonite and car-bonate, although this feature is rare, unless the marginalzone is fractured and traversed by vein material. Inter-stitial glass shows various alteration effects, and is al-ways replaced by palagonite and/or smectite, the colorof which may vary according to the absence or presenceof magnetite granules and other crystallites. Glass, lack-ing associated late magnetite, is replaced by clear honey-yellow or lime-yellow palagonite, which may sometimeshave a thin outer rim of replacing reddish-brown or brownsmectite. In a few cases (e.g., Unit 48), dark brown smec-tite "vermicules" may be present in palagonite, suggest-ing diffusive replacement via crude perlitic cracks in theoriginal glass. Palagonite containing magnetite is gener-ally dark yellow-brown throughout and more granularin appearance. Some interstitial glass is totally replacedby dark brown, almost opaque smectite that often masksthe presence of numerous magnetite granules.
Of the three phenocryst phases—olivine, palgioclase,and clinopyroxene—olivine is always completely replacedand Plagioclase may be variably altered, especially in thecrystal core, whereas clinopyroxene is always fresh. Ol-ivine microphenocrysts in the interiors of the more mas-sive units are completely pseudomorphed by greenishbrown or dark brown smectite. Greater variation is seenin quenched flow margins, where olivine phenocrysts dis-play crisscrossing veined microfractures set in a dirty greenor brown smectite "matrix." In other cases, olivine isreplaced by a dark red or reddish brown smectite, whichmay itself be partially replaced either by calcite or, morerarely, by a dark green celadonite. Only in the quenchedmargins of Unit 48 is olivine totally replaced by zeoliteand calcite; this probably represents a further stage ofalteration after initial pseudomorphing by smectite. ZonedPlagioclase phenocrysts have cores replaced by dark brownor green pleochroic smectite or, more rarely, by brightgreen celadonite and calcite. Celadonite may be inter-mixed or may grade into any associated brown smectite,but is always itself replaced by calcite when present. Pla-gioclase microlites in quench zones may also be partiallyreplaced by greenish smectite.
The interiors of flow units are relatively fresh, exceptfor palagonitization of interstitial glass, pseudomorphingof any olivine microphenocrysts, and the intimate asso-ciation of dark to opaque smectitic material blanketingcoarse magnetite grains. However, the coarse-grained in-teriors of the more massive aphyric Units 48 and 52 showthe patchy development of Plagioclase alteration domains,with the variable growth of smectites, celadonite, zeo-lites, albite, and K-feldspar. This type of alteration isheralded by the development of acicular zeolite needlesand the random development of minor green-brown smec-tite in Plagioclase laths. Some crystals may eventually be-come almost totally replaced by smectite crowded withboth acicular and feathery zeolites. A further develop-ment is the irregular replacement of Plagioclase laths byanhedral pools of K-feldspar (and albite) in areas rich insmectite. Extensive smectite development obscures pla-gioclase crystal boundaries, and often exhibits an ap-parent coaxial intergrowth of brown to opaque smectitefibers and dark green platy celadonite or smectite.
Although vesicles and/or vugs are generally lackingin the Leg 89 basalts, these features are considered to berepresented, respectively, by ovoid areas, 0.2-0.3 mm indiameter, and larger, more irregular star-shaped regions.The amygdales are filled with a greenish smectite, or,when zoned, are lined with brown smectite and have acore of dark green smectite or celadonite. Vugs are pres-ent in Unit 55, and show the development of stronglypleochroic (yellow-green-greenish brown), radiate, coarse-grained plates of smectite.
Vein material has not been studied in detail, but thepresence of ubiquitous pyrite in fractures throughout thelavas clearly indicates that reducing conditions were main-tained during the later stages of alteration. Typical veinassemblages include pyrite-dark green smectite, smectite-zeolites, smectite-zeolites-calcite, stilbite, stilbite-quartz.Zeolites are predominantly stilbite with minor phillipsiteand heulandite. Higher in the flow sequence (drilled dur-ing Leg 61), zeolite-bearing quartz-poor, quartz-pyrite-rich, and quartz-celadonite assemblages were identifiedwith increasing stratigraphic height (Windom and Book,1981). Two vein-forming episodes were also recognized,with an initial reductive phase overprinted by a later oxi-dative phase which oxidized pyrite and smectite to Feoxides and celadonite, respectively.
Alteration Relative to Depth and Physical Parameters
Variation of various physical and alteration parame-ters with depth is shown in Figure 6. The relative uni-formity of these parameters indicates that there are nosystematic trends in the intensity or character of altera-tion with depth. The major deviations occur at recog-nized volcanic unit boundaries, especially between Units45 and 46, where the base and top are defined by sharpincreases in H2O + , Fe2O3/FeO, and CO2, coupled witha decrease in basalt specific gravity. Some individual holo-crystalline units show a "sawtooth" variation in the Fe2O3/FeO ratio, which, considering the increase in oxidationat flow tops and bottoms, probably reflects the presenceof thin multiple flow units rapidly produced during oneeruptive event. Rapid extrusion throughout the eruptiveevent is also indicated by the constancy of the magnetic
477
P. A. FLOYD
o'c
S 1
Sonic vel.Vp (km/s)
5 6 7
Specificgravity
2.6 3.0
Porosity
2 6 10
H2Cf
(wt.%)
0 2 4
co2
(wt.%)
0 0.2 0.4
Fe2O3/FeO
0.4 0.6
Magneticinclination
10° 30° 50°
1 0 8 0 -95
1 1 0 0 -
46
1120-
47
Q.
α>100
5 1140-
49
50
1160-
51
105
1180- :
54
1200- 55
56
I I I I I I
\\
V
i \ i \ i I I i i r
(2)
Mainly aphyric units Mainly phyric units
Figure 6. Variation of physical and alteration parameters with depth, Hole 462A. Parentheses in Magnetic unit column indicate that 2 is not strictlya single unit because of variations.
inclination for the volcanic unit. In areas where frac-ture-propogated alteration is minimal, marked decreasesin specific gravity (and sonic velocity), caused by the pres-ence of low-density secondary minerals, can also be usedto distinguish internal boundaries.
Summary and Conclusions
1. Mineralogically and chemically, the alteration is mildthroughout the lava pile and typical of low-temperature,post-consolidation reaction with seawater.
478
PETROLOGY AND GEOCHEMISTRY
2. Secondary mineral phases are dominated by vari-ous smectites in both the rock matrix and fractures, where-as pyrite is characteristically developed in fractures rath-er than the rock host. Secondary assemblages indicatelow zeolite-facies alteration under reducing, acid to neu-tral, conditions. A late, well-developed oxidative phasewith attendant extensive carbonation, as often seen in al-tered ocean-floor basalts, is not recorded in the Leg 89lavas.
3. Two alteration stages are recognized, (a) Initial non-oxidative reactions involve palagonitization of glass, de-velopment of brownish smectite after palagonite, oliv-ine, and Plagioclase, and the (?) production of minorzeolites. The main chemical changes resulted in bulk-rock increases in H2O + , mobility of K, Rb, Ba, Sr, andNi, and disturbance of the FeOVMgO ratio, (b) A mi-nor increase in redox potential followed, with the devel-opment of green smectite, celadonite, and calcite oftenreplacing earlier phases; host-rock zeolites and K-feld-spar may also have developed at, or just before, thisstage. The second stage resulted in increases in H2O + ,Fe 2 (yFeO, CO2, K, Rb, and mobility of Ca. That theconditions were still essentially reducing is illustrated bythe development of pyrite-green smectite in fractures thattraverse the whole lava pile.
4. No systematic changes in alteration intensity arerecorded with depth. In the absence of extensive pene-trative alteration adjacent to fractures, increases in H 2O+ ,Fe2O3/FeO, and CO2, coupled with decreases in specificgravity and sonic velocity, are representative of flowboundaries. Marked random increases in K or K/Rb ra-tios reflect Plagioclase alteration domains within mas-sive flow units.
PETROGRAPHY OF BASALTIC ROCKS
Three basaltic types occur in the sheet flows. Theyare aphyric basalts, aphyric microdolerites, and sparselyto moderately phyric basalts, characterized by the fol-lowing common phenocryst assemblages: clinopyroxene-plagioclase-olivine, clinopyroxene-plagioclase, and cli-nopyroxene—olivine ± minor Plagioclase. Phyric basalts(with > 10% total phenocrysts) are very limited in ex-tent, and generally exhibit clinopyroxene-plagioclase-ol-ivine phenocryst assemblages dominated by clinopyrox-ene. Basalts containing clinopyroxene and Plagioclase phe-nocrysts are present throughout the total depth, whereasthose containing olivine as a phenocryst phase appear tobe concentrated in the upper stratigraphic units. Aphyr-ic basalts constitute about 44% of the total lava pile,and form Volcanic Units 45, 48, and 52 (Fig. 2). Micro-dolerite occurs only in the coarse-grained interior of aphy-ric Unit 52.
Modal proportions based on point-count data for anumber of samples are shown in Table 4. Clinopyroxeneis the dominant phase in the matrix and among the phe-nocryst assemblages, whereas olivine has not been foundin the matrix (unless masked by alteration products) andoccurs only as individual microphenocrysts now replacedby smectite. Magnetite is common in all basalts, and insome cases reaches 10 vol.% of the rock. It is invariablya late crystallization product, associated with interstitial
glass. Pyrite is secondary, and develops only in the hostrock near to pyrite-smectite-filled fractures. Electron mi-croprobe data on some of the primary phases are pre-sented by Floyd and Rowbotham (this volume).
Aphyric Basalts and Microdolerites
In the central portions of flows, these rocks are oliv-ine-free and either hypocrystalline or holocrystalline withgranular to intersertal textures. Any interstitial glass isnow palagonite or smectite, commonly crowded with mag-netite granules or dendrites, or occasionally with Plagio-clase microlites. Polysynthetic twinned Plagioclase lathsform a groundmass lattice within which are dispersedanhedral clinopyroxene grains that crystallized later. Incoarser holocrystalline varieties, large magnetite grainspartly mantle pyroxene and Plagioclase crystals. Someflow interiors, including the microdolerites, exhibit crys-tal habits indicative of quenching at relatively slow rates,with variolitic groups of wedge-shaped, serrated Plagio-clase subophitically enclosed in a central clinopyroxenenucleus. Individual Plagioclase crystals often form long(up to 1.5 mm) serrated curved crystals, or may be asso-ciated together in single or double pairs to form "bow-ties" (up to 2.5 mm across) unrelated to clinopyroxene.Irrespective of whether a flow interior exhibits quench-serrated Plagioclase or well-crystallized subhedral laths,the clinopyroxene habit is invariably anhedral. In thecentral portion of Unit 48, however, clinopyroxene mayoccasionally develop as elongate serrated crystals whichare nearly always twinned.
Quenched flow margins are nonvesicular, have hyalo-pilitic, variolitic, and, more rarely, spherulitic textures,with numerous crystallites and microlites of clinopyrox-ene, Plagioclase, and magnetite set in altered glass. Be-cause the proportion of crystalline material is always highand completely vitreous margins are generally lacking,flow tops were probably insulated by subsequent flowsor glassy rubble removed by the next extrusion. A tra-verse from the outer glassy portion toward the more crys-talline interior of a flow margin shows a number of mor-phological changes in the crystalline phases, as well as adecrease in the proportion of glass. Within the glass-dominated outer portion, dark, variably coalesced cli-nopyroxene spherulites that have nucleated on tuning-fork Plagioclase microlites are developed. The dark colorresults partly from the fine grain-size of the clinopyrox-ene crystallites, and from the presence of minute inter-spersed Fe ore grains. Farther from the margin, dark cli-nopyroxene variolites may form the bulk of the matrix,enclosing glass and magnetite granules between splayedcrystallites. Plagioclase typically forms swallow-tail ortuning-fork microlites that either form nuclei to the va-riolites or are randomly enclosed by them. Next, the grainsize becomes coarser and the clinopyroxene more bire-fringent. Quenched clinopyroxene exhibits curved plu-mose sprays and sheaflike bundles either originating froma point growth-source or nucleated on larger existing cli-nopyroxene grains (microphenocrysts). Magnetite grainsare concentrated on the margins of the variolites, ratherthan within them. Plagioclase, which always appears tohave crystallized first, maintains a microlitic form, and
479
P. A. FLOYD
Table 4. Modal proportions of primary and secondary minerals in selected basalts, Hole 462A.
Keelenumber
Pointscounted
Phenocrysts
cpx plag ol
Matrix Alteration
cpx plag mag pyrite pal smect zeol carb Comments on alteration features
462A-1 2009 1.7 0.4 a 35.7 22.7 7.1 2.0 30.2 0.1 tr Dark golden brown smectite replaces olivine microphenocrysts462A-4 2059 0.4 1.4 a 38.0 24.3 7.9 1.5 26.3 0.2 Opaque smectitic material partly obscures magnetite462A-7 2007 0.1 0.5 a 37.6 28.3 8.3 tr 25.1 0.1 Ore closely associated with dark smectite-replaced interstitial glass462A-10 2084 1.6 42.0 22.5 9.2 tr 24.1 0.6 Some dark smectite (montmorillonite?) partly replaces Plagioclase
microphenocrysts462A-13 2147 9.7 1.3 58.4 13.2 8.6 tr 8.5 0.4 Quench texture, larger granular cpx recorded as "phenocrysts"462A-15 1965 20.3 0.3 a 25.1 18.7 10.9 1.3 23.5 0.2 Quench texture, larger granular cpx recorded as "phenocrysts"462A-18 1983 39.8 27.5 9.2 0.2 22.9 0.3 Zeolite fibers associated with brown smectite462A-20 2061 tr 1.2 a? 41.1 25.8 9.1 0.4 22.3 Glomerocrystic Plagioclase, some with cpx inclusion zones462A-21 2065 10.6 2.7 46.6 14.2 9.2 13.7 3.1 Quench texture, larger granular cpx recorded as "phenocrysts"462A-22 1990 16.4 0.6 a? 29.4 11.8 9.6 25.0 6.9 0.4 Cores of Plagioclase microphenocrysts replaced by green smectite462A-23 2179 tr 46.1 19.8 8.9 1.2 24.1 Magnetite partly obscured by opaque smectite462A-27 2027 tr 38.8 33.7 4.6 tr 22.9 Granular crystallites in smectite-replaced interstitial glass462A-31 2159 40.6 31.9 8.3 0.6 tr 16.6 2.0 Rare islets of golden palagonite in yellow-brown smectite462A-34 2020 40.1 33.6 3.7 tr 21.0 1.6 Zeolite needles in Plagioclase462A-38 2055 45.5 27.0 3.2 3.0 20.0 1.3 Magnetite partly obscured by opaque smectite462A-40 2021 2.7 2.0 a 51.1 22.1 4.4 0.4 3.8 13.4 0.2 Olivine microphenocrysts replaced by yellow-brown smectite and
minor carbonate462A-41 1093 2.4 3.4 a 76.0 3.3 9.7 1.6 3.1 0.5 Matrix cpx also includes altered glass between plumose crystallites462A-43 1994 1.0 0.4 a 50.5 16.9 9.9 13.0 8.3 Dark golden-brown smectite replaces olivine microphenocrysts462A-46 2104 2.9 1.5 49.8 17.6 13.5 8.3 6.5 Pale yellow-brown palagonite and smectite replace interstitial glass462A-48 2041 2.6 2.2 50.8 23.3 7.5 8.6 5.0 Brown smectite rims pale greenish yellow palagonite areas462A-51 1065 2.9 3.0 a 71.5 10.8 10.1 1.8 Matrix cpx also includes glass between crystallites; smectite
replaces olivine462A-52 2033 0.9 1.1 a 58.6 17.2 6.0 11.7 4.4 Lime-green smectite replaces olivine<?) phenocrysts462A-54 2008 2.3 1.6 56.8 12.9 4.9 14.9 6.5 Plagioclase phenocrysts partly replaced by green smectite and rare
analcite462A-55 2038 2.9 2.1 42.9 21.2 10.1 18.8 2.0 Yellow smectite replaces interstitial glass462A-56 2018 0.2 0.7 41.6 28.5 7.9 16.2 4.8 Palagonitized interstitial glass crowded with crystallites462A-61 2044 43.7 31.9 7.6 14.3 2.4 Zeolite needles and blue-green celadonite in Plagioclase laths462A-63 1516 38.1 35.8 6.5 19.1 0.6 Bright green celadonite (1% of smectite) and late K-feldspars
replace Plagioclase462A-66 2019 40.7 38.3 7.9 tr 11.3 1.8 Opaque smectite associated with magnetite and cpx rims462A-69 1877 0.5 0.4 42.0 21.9 11.4 17.8 5.9 Recorded smectite largely replacing Plagioclase phenocrysts462A-71 1703 4.2 0.4 43.9 18.3 9.2 3.2 20.9 Greenish smectite replaces Plagioclase phenocrysts; brown smectite
replaces glass462A-73 2012 0.9 0.7 40.7 33.2 7.7 12.9 3.9 tr Granular brownish smectite replaces interstitial glass462A-75 1992 4.9 2.7 40.8 17.6 10.1 17.6 6.4 Green smectite or celadonite flakes replace Plagioclase microphe-
nocrysts
Note: cpx = clinopyroxene; plag = Plagioclase; ol = olivine; mag = magnetite; pal = palagonite; smect = smectite; zeol = zeolite; carb = carbonate; tr = trace (<O.l modal %).a Present originally, but now totally replaced.
does not develop larger skeletal crystals with serrated edg-es until well inside the flow. All the features describedare clearly related to the rate of cooling of the flow, andare characteristic of many mid-ocean ridge (MOR) pil-low basalts (e.g., Kirkpatrick, 1979; Natland, 1979a).
A common variant occurs in which the flow marginsof some essentially aphyric basalt units are to some de-gree porphyritic, with large zoned plagioclases, looselyclumped groups of anhedral clinopyroxene crystals, andsometimes small smectite-replaced ovoids of olivine. Glo-merocrystic groups of clinopyroxene and Plagioclase arealso relatively common.
Variably Phyric Basalts
Flows composed of variably phyric basalts may havequenched glassy margins and hypocrystalline or holo-crystalline interiors, whereas others are quench-texturedthroughout (e.g., Units 49-51) and show matrix texturesand mineralogy similar to those of the aphyric flows al-ready described. Some units are poorly vesicular withsmectite-filled vesicles (~ 0.2 mm diameter). The granu-lar groundmass of flow interiors is composed of smallcrystals of clinopyroxene (0.05-0.10 mm), Plagioclase
(0.1-0.2 mm), and magnetite (0.01-0.05 mm). Again,some interiors exhibit coarse quench textures, showingnot only serrated skeletal Plagioclase but also variolitescomposed of both individual Plagioclase and clinopy-roxene bowties. In a few cases, Plagioclase and clinopy-roxene may form intimate coaxial intergrowths in thebowtie. Glassy margins are again variolitic and charac-terized by plumose clinopyroxene (typically up to 0.2 mmlong), often peppered with magnetite grains, together withscattered forked or lath-shaped Plagioclase microlites. Far-ther from the flow margin, some clinopyroxene plumesmay be 0.4-0.5 mm long, and the sheaflike bundles areup to 1.0 mm long.
Olivine microphenocrysts (0.2-0.4 mm) are always re-placed by smectite, and tend to be euhedral in quenchedmargins but more irregular and embayed(?) in flow inte-riors. Plagioclase phenocrysts (An70_80) are invariablyCarlsbad twinned, sometimes polysynthetic twinned, andgenerally lath-shaped (0.2-0.5 mm) if present as isolatedcrystals. Some individuals are faintly zoned with coresreplaced by greenish smectite. Hypocrystalline rocks mayshow a mixture of Plagioclase morphologies, with forkedmicrolites and variable sizes of serrated wedges and laths
480
PETROLOGY AND GEOCHEMISTRY
that grade up to a few subhedral crystals of phenocrystsize. The gradation in size suggests that Plagioclase crys-tallization may have been continuous throughout pre-emptive history and final quenching on the ocean floor.Clinopyroxene phenocrysts (~0.2 mm) typically formgranular-textured glomerocrystic groups up to 1 mmacross. Plagioclase may be associated with these groups,although the texture remains granular, and only rarely isit enclosed subophitically. In a manner similar to Plagio-clase, clinopyroxene may display a number of grain sizesin the same rock, with glomerocrysts composed of smallgrains and larger subhedral individual phenocrysts all setin a fine-grained hyalopilitic groundmass of quenchedclinopyroxene plumes.
CHEMICAL VARIATION AND PETROGENESIS
A series of 75 basalt samples from Leg 89 was chemi-cally analyzed (1) to identify the geochemical features ofintraplate submarine sheet-flows, and (2) to determineany coherent chemical groups and their petrogenetic re-lationships. Major- and trace-element data for the ba-salts are presented in Table 5, and the correspondingDSDP sample number, sub-bottom depth, and a brief li-thology of each sample are given in Table 1. Normativedata are listed in Table 6.
Normative and General Chemical Characteristics
The basalts are all hypersthene- (hy) and diopside-(di) rich, and as illustrated by the spread in the basalttrapezohedron (Fig. 7), are exclusively olivine- (ol) orquartz- (qz) normative tholeiites. The phyric basalts (Vol-canic Units 46, 47, 49-51, and 53-56) are all ol-norma-tive, as are the few aphyric basalts within Unit 46, whereasthe aphyric basalts of Units 48 and 52 are qz-normative.This appears to be a real difference and not an oxidationeffect, since all data were normalized to an oxidation ra-tio of 0.15 before calculation. Oceanic basalts and glassesfrom the Mid-Atlantic Ridge (MAR) and East PacificRise (EPR) show a similar spread across the silica satu-ration divide (e.g., Flower et al., 1977; Tarney et al.,1979; Humphris, Thompson, Gibson, et al., 1980), al-though many are strictly ol-normative (Fig. 7). Relativeto low-pressure experimental equilibria (Thompson, 1982),the Leg 89 basalts form a trend parallel to the 1-atm. co-tectic in equilibrium with olivine-plagioclase-pyroxene(Fig. 7), but are displaced towards hy. Thompson et al.(1983) suggest that one possible explanation for the dis-placement may be that originally deep-seated magmaponded in short-lived, high-level magma chambers anddid not have enough time during residence to crystallizeto the 1-atm. cotectic. Possible support for this sugges-tion in the Leg 89 lavas is the total lack of cumulates,the presence of at least two generations of clinopyroxeneand Plagioclase phenocrysts, and often complete chemi-cal overlap between phyric and aphyric units, implyingminimal fractionation.
Chemically the basalts are characterized by the fol-lowing features (Table 5 and 7):
1. They are tholeiites with relatively uniform compo-sitions throughout (as exhibited by the small elementalrange), compared with the variable chemical composi-tion in many ocean basalt sequences.
2. Limited internal differentiation (resulting from dif-ferent magma groups and/or magmatic processes) is sug-gested, however, by variable FeOVMgO ratios (1.6-2.1)and antipathetic relationships between compatible andincompatible elements within some volcanic units.
3. As a group, the majority of basalts have a highFeO* content (> 12 wt.%) similar to that of some oce-anic ferrobasalts (e.g., Galapagos Rift; Natland, 1980),although incompatible-element abundances are generallylower.
4. Incompatible elements, especially La, Ce, U, Th,K, Rb, Ba, Hf, Nb, and Y, have low abundances, al-though marginally higher than N-type MORB. The widerange shown by K, Rb, Ba, and possibly U is a reflec-tion of mobility during mild alteration.
5. On average, siderophile-element contents are rela-tively low and indicate moderately evolved compositionscompared with primary basaltic melts having high MgOof 10-13 wt.% and Ni contents of about 300-400 ppm(Sato, 1977; Hart and Davis, 1978).
6. Chondrite-normalized Ce/Y ratios (-0.9) and REEdata (Saunders, this volume) show that the basalts arelight-REE-depleted, but to nowhere near the same ex-tent as N-type MORB.
7. Ratios of many incompatible elements with simi-lar, low distribution coefficients (KD's) (e.g., Th/U, La/Nb, Ti/Zr, Zr/Nb, Ti/Y, Zr/Y, TiO2/P2O5) have gener-ally lower than chondritic values, although TiO2/P2θ5 ishigher (Table 7).
The type B flow basalts of Leg 61 (Batiza, 1981) alsoexhibit chemical uniformity, selectively depleted incom-patible-element compositions, and an overall dissimilar-ity to N-type MORB. Although some degree of down-hole variation (e.g., increasing TiO2) was apparent dur-ing Leg 61, no specific chemical groups were documented.Saunders (this volume), on the basis of variable Zr con-tent (Legs 61 and 89, new data), has demonstrated thepresence of three chemical groups developed sequential-ly downhole: I (upper sils), and II and III (lower sheet-flows); all the Leg 89 section falls in group III. In gen-eral terms, the basalts in the basal section of the hole(sampled during Leg 89) are more evolved than thosehigher up (sampled during Leg 61), and have higher Zrand lower Ni contents (Saunders, this volume) and, asseen from the distributions in Figure 8, higher TiO2, K2O,FeO*, and FeOVMgO ratios.
Chemical Stratigraphy
Downhole chemical variations for selected elementsand ratios are shown in Figure 9. Although minor fluc-tuations are evident within the small elemental range ex-hibited, some generalized trends with depth are observed:(1) The FeOVMgO ratio increases and Cr and Ni de-crease, whereas (2) Zr, Y, La, Ce, Nd, and Sr increase.Although the crude antipathetic relationship between com-patible and incompatible elements could imply that thebasalt pile represents the products of a single fractionat-ed magma chamber, the generalized trends are to someextent influenced by the relatively high incompatible-el-ement contents of volcanic units at the base of the se-quence. Also, the trends are not completely systematicfor all related elements; for example, compatible Se in-
481
P. A. FLOYD
Table 5. Major- (wt.%) and trace- (ppm) element data on Nauru Basin basaltic lava flows, Hole 462A, by Keele numbers.
SiO 2
TiO2
A12O3
Fe 2O 3
FeOMnOMgOCaONa 2OK2O
P2O5,H 2 O +
CO 2
Total
S.G.a
FeOVMgOFe2O3/FeO
BaCeCrLaNbNdNiRbSeSrYZrHfTaThU
462 A-1
48.971.33
14.514.268.720.226.76
11.922.550.130.110.970.13
100.58
2.951.860.49
813
15966
1185
543
1123271
462A-2
50.391.27
13.994.438.920.206.24
11.332.450.140.110.760.04
100.27
3.102.070.50
511
15946
1081
543
1082867
1.740.400.260.06
462A-3
50.721.31
13.824.199.120.206.73
11.342.490.140.110.650.08
100.90
2.971.920.46
2211
15456
1181
540
1142871
462A-4
49.761.29
14.054.338.110.196.81
11.722.480.150.101.070.08
100.14
2.881.760.53
3910
169467
842
441132862
1.860.550.260.07
462A-5
49.611.33
14.064.289.340.237.15
11.802.340.120.110.830.07
101.27
2.941.850.46
1510
166467
852
431132965
462A-6
50.111.31
14.234.508.470.206.75
11.612.450.130.110.840.04
100.75
2.931.850.53
1011
16048
1086
344
1132866
462A-7
50.721.28
13.974.009.030.206.68
11.652.400.120.110.650.11
100.92
2.941.890.44
1015
16466
1088
242
1082863
1.880.750.260.06
462A-8
50.031.28
13.613.959.200.227.00
11.172.380.120.100.810.10
99.98
2.941.820.43
1410
16245
1188
241
1062867
462A-9
50.511.27
13.593.879.550.226.96
11.352.540.150.100.640.02
100.77
2.971.870.41
516
16556
1288
242
1032869
462A-10
49.471.32
14.273.579.380.226.93
11.762.440.130.110.840.04
100.48
2.951.820.35
1414
166568
859
411083260
1.860.700.270.058
462A-11
50.041.23
13.683.599.570.227.12
10.842.300.320.110.980.20
100.20
2.991.800.38
1016
16365
1185
942
1092958
462A-12
50.191.31
14.034.389.110.227.08
10.512.410.120.110.940.12
100.53
2.961.840.48
514
159549
875
431123261
462A-13
49.421.31
13.834.118.670.227.02
11.572.430.120.101.260.19
100.35
2.971.760.47
1114
162569
883
441113162
1.840.500.280.07
462A-14
49.161.23
13.696.606.740.147.829.021.810.160.093.060.51
100.03
2.631.620.98
2715
179569
882
481223159
462A-15
50.241.32
14.224.238.140.266.74
11.372.730.120.111.050.31
100.84
2.871.770.52
3214
163558
915
481113264
1.950.670.270.08
462A-16
50.291.26
13.873.259.700.206.82
11.342.510.130.110.760.04
100.28
2.951.850.34
2816
15865
1084
442
1073161
1.930.700.270.051
a S.G. = specific gravity.
Table 5. (continued).
SiO2
TiO2
A12O3
Fe 2O 3
FeOMnOMgOCaONa2OK2O
P 2 θ 5 +
H 2 O +
c o 2
Total
S.G.a
FeO /MgOFe2O3/FeO
BaCeCrLaNbNdNiRbSeSrYZrHfTaThU
462A-33
51.121.20
13.893.789.190.206.59
11.492.310.130.100.930.02
100.95
2.971.910.41
511
159468
832
451092660
462A-34
50.551.18
13.823.379.190.206.93
11.382.280.130.101.010.02
100.16
2.971.760.37
58
16445
1084
244
1092361
462A-35
50.191.18
13.983.639.170.206.95
11.452.340.120.100.840.02
100.17
2.971.790.40
267
163366
872
431092361
462A-36
51.521.19
13.933.529.010.207.01
11.462.340.120.100.900.02
101.32
2.971.740.39
1513
16044
1085
444
1102462
1.870.030.280.06
462A-37
50.631.16
13.993.719.300.206.72
11.312.410.130.100.960.02
100.64
2.991.880.40
128
153367
875
461122257
462A-38
50.841.23
13.783.559.330.206.76
11.392.400.120.110.880.02
100.61
2.971.850.38
336
152455
884
461152665
1.860.830.280.11
462A-39
50.951.23
13.783.549.270.206.70
11.522.420.130.100.760.02
100.62
2.971.860.38
156
150365
844
461172557
462A-40
50.881.24
14.064.048.620.206.52
11.402.350.130.101.120.08
100.74
2.931.880.47
4410
151359
832
461102861
462A-41
50.181.24
13.843.489.110.206.36
11.222.340.150.111.060.18
99.47
2.941.920.38
2410
141358
802
481092767
2.040.420.290.06
462A-42
50.181.24
14.003.539.110.236.47
11.382.370.140.101.030.03
99.81
2.931.900.39
1714
14255
1181
648
1122765
1.860.320.260.06
462A-43
49.521.25
14.124.209.130.237.02
11.642.370.170.101.070.02
100.84
2.921.840.46
516
13845
1276
648
1103270
462A-44
49.981.26
14.013.889.370.256.78
11.792.300.210.110.820.02
100.78
2.971.900.41
914
13855
1074
745
1122863
2.040.620.290.05
462A-45
49.671.25
14.034.249.030.237.01
11.892.450.130.101.100.04
101.17
2.941.830.47
616
13945
1275
545
1123273
462A-46
49.881.23
14.003.889.190.236.70
11.682.390.140.111.010.02
100.46
2.921.890.42
515
137549
772
471102761
462A-47
49.881.25
14.043.829.170.256.95
11.662.360.170.111.030.02
100.71
2.931.810.42
515
142559
832
481072760
482
PETROLOGY AND GEOCHEMISTRY
Table 5. (continued).
462A-17
50.371.25
13.933.379.230.207.13
11.372.530.130.100.850.07
100.53
2.971.720.37
2916
166659
846
421072858
1.940.760.250.05
462A-18
50.011.24
13.873.888.940.226.84
10.832.430.530.111.100.09
10.09
2.931.820.43
1915
170559
908
431073158
462A-19
50.441.27
13.843.759.740.226.76
11.902.360.120.110.690.08
101.28
2.981.940.39
513
16654
11876
421113167
1.900.650.260.09
462A-20
49.151.29
14.084.029.350.206.89
11.612.440.150.100.880.05
100.21
2.981.880.43
511
159448
864
411113263
462A-21
49.351.30
14.173.989.530.206.91
11.712.350.130.111.010.07
100.82
2.951.900.42
515
164558
843
4510832612.041.060.280.08
462A-22
49.601.23
13.613.729.040.466.74
11.662.220.170.100.830.20
99.58
2.951.840.41
516
170659
876
461092865
1.930.860.2660.071
462A-23
50.031.27
14.194.528.850.206.90
11.742.670.120.110.540.02
101.16
2.921.870.51
911
147478
834
421102868
1.860.870.280.062
462A-24
51.161.19
13.764.658.260.196.59
11.352.380.120.111.030.02
100.81
2.961.890.56
59
152465
827
461082560
462A-25
50.191.20
13.994.158.360.206.58
11.722.470.120.111.200.02
100.31
2.921.840.50
59
157448
842
461082666
1.760.710.260.07
462A-26
51.411.21
13.814.218.200.206.80
11.182.550.120.110.980.02
100.80
2.911.760.51
510
150559
872
441093169
462A-27
50.311.23
14.244.217.950.226.91
11.392.810.120.101.400.02
100.91
2.871.700.53
515
16455
1186
347
1032971
462A-28
50.001.25
14.193.908.290.206.64
11.352.830.130.110.930.02
99.84
2.841.780.47
910
155448
892
481092963
1.800.580.300.07
462A-29
50.761.22
13.824.088.280.206.52
11.262.450.130.110.990.02
99.84
3.011.830.49
511
152348
864
47IU2963
462A-30
49.971.22
14.304.458.600.206.76
11.822.380.120.101.030.02
100.97
3.201.860.52
514
15555
1085
546
1142865
462A-31
50.911.18
13.764.018.840.206.76
11.472.340.120.100.880.02
100.59
2.931.840.45
58
147457
782
441112457
462A-32
50.271.18
13.793.868.960.206.87
11.502.270.130.100.950.02
100.10
2.991.810.43
58
161367
824
451082555
1.850.630.290.09
Table 5. (continued).
462A-48
50.071.26
14.103.628.890.226.47
11.842.510.130.111.130.02
100.37
2.901.880.41
516
13755
1082
347
1172963
1.840.480.290.07
462A-49
49.571.22
14.094.398.650.236.89
11.462.710.130.101.200.02
100.66
2.791.830.51
518
14665
1481
547
1053269
1.960.560.300.04
462A-50
49.361.25
13.944.458.890.236.90
11.612.330.140.111.320.02
100.55
2.871.870.50
818
14365
1479
649
1123073
2.140.930.320.07
462A-51
51.001.25
14.054.247.800.236.18
11.242.990.130.111.070.04
100.33
2.801.880.54
519
14475
15834
511063373
1.870.430.310.08
462A-52
50.281.29
13.724.428.230.236.59
10.972.840.140.111.250.02
100.09
2.891.850.54
1913
14555
1179
247
1052761
1.970.670.330.10
462A-53
49.411.25
13.974.278.420.266.83
11.382.410.140.091.290.02
99.74
2.851.800.51
1813
14055
1077
248
1072962
1.780.320.290.10
462A-54
50.181.26
14.064.278.340.236.75
11.432.550.130.111.100.02
100.43
2.881.800.51
513
14356
1077
248
1092962
1.890.420.290.08
462A-55
49.861.23
13.964.398.820.256.78
11.892.360.130.101.050.03
100.85
2.911.880.50
512
141359
782
471102560
1.960.640.280.06
462A-56
50.691.22
13.894.328.720.206.47
11.722.490.120.100.950.03
100.92
2.931.950.50
912
142359
835
471112560
462A-57
50.371.22
13.944.268.650.196.75
11.562.390.120.100.870.02
100.47
2.901.850.49
513
138556
804
461112765
1.880.700.290.09
462A-58
50.901.21
13.654.058.910.206.52
11.222.440.120.110.960.03
100.32
2.961.930.45
510
138469
793
471112561
462A-59
51.061.24
13.764.358.860.206.53
11.342.330.120.100.960.02
100.87
2.951.940.49
510
141458
782
481122760
462A-60
51.031.22
13.784.358.650.206.66
11.392.350.150.111.070.02
100.98
2.981.890.50
814
14156
1276
249
1102867
1.950.460.290.08
462A-61
50.681.21
13.664.008.990.206.37
10.732.210.430.101.080.02
99.68
2.961.980.44
4015
13865
1477
845
1053273
1.790.460.260.05
462A-62
50.881.23
13.954.028.680.206.51
11.052.460.570.100.970.02
100.64
2.971.890.46
2912
139569
798
451072965
1.770.460.260.05
462A-63
50.751.31
13.614.138.890.226.22
10.752.440.420.111.110.02
99.98
2.992.030.46
2814
13055
1266
646
1093171
483
P. A. FLOYD
Table 5. (continued).
SiO2
TiO2
A12O3
Fe 2O 3
FeOMnOMgOCaONa2OK2O
P 2 θ 5 +
H 2 O +
co2Total
S.G.a
FeOVMgOFe2O3/FeO
BaCeCrLaNbNdNiRbSeSrYZrHfTaThU
462A-64
47.841.15
12.943.438.680.196.16
10.452.430.320.106.050.04
99.78
2.961.910.40
810
139459
787
4410827632.060.440.290.08
462A-65
47.441.15
12.893.188.950.196.05
10.522.340.130.096.160.04
99.13
2.941.950.36
218
147355
826
451122969
462A-66
50.621.22
13.633.579.000.206.35
11.152.350.130.111.100.02
99.45
2.981.920.40
513
144559
815
45IU27662.070.600.290.10
462A-67
51.121.25
13.564.158.880.206.56
11.492.380.130.100.960.03
100.81
3.031.920.47
139
144468
777
461112962
462A-68
50.251.30
14.064.467.810.206.65
11.252.340.120.101.660.02
100.22
2.911.780.57
524
14887
1681
350
11632782.240.460.340.10
462A-69
50.081.30
13.674.718.800.256.33
11.722.530.170.111.140.02
100.83
2.902.060.54
717
12466
1074
447
1152862
462A-70
49.551.29
13.834.318.920.257.15
11.512.540.150.111.140.03
100.78
2.911.790.48
522
11885
1375
247
1133167
1.940.630.300.07
462A-71
49.571.32
13.734.438.890.236.72
11.612.500.160.101.160.02
100.44
2.931.920.50
517
12375
1176
449
1123067
462A-72
49.661.30
13.574.329.000.206.28
11.362.570.130.101.130.03
99.55
2.922.050.48
516
11855
1075
346
1193061
462A-73
50.071.31
14.104.718.870.226.93
11.502.530.130.111.080.03
101.59
2.931.890.53
2717
11765
1073
749
11530682.210.410.290.10
462A-74
50.251.34
14.414.478.050.226.69
11.053.030.130.111.320.03
101.10
2.711.800.56
3218
11585
1174
448
1133273
462A-75
49.871.33
13.884.468.040.206.80
10.523.510.130.111.250.02
100.12
2.801.770.55
2920
11975
1379
649
10728732.260.790.340.10
creases with depth, but TiO2 is lowest at intermediatelevels and some Zr contents are as high at the top of thepile as at the base. In general, the fine detail of the chem-ical variation, relative to the gross trends, precludes asimple genetic relationship between all the volcanic units,or indeed between only the aphyric units as representa-tives of initial melts.
Chemical Megagroups and Petrogenetic Implications
Chemical groups representing different magma batchesinclude rocks of overall similar composition, as well asthose whose members can be related by the low-pressurefractionation (addition and subtraction) of observed phe-nocryst phases from a parental magma. Variation betweenthe parental magmas of each independently fractionatedchemical group may reflect the operation of high-pres-sure fractionation, variable partial melting of a uniformsource, or, possibly, generation from heterogeneous man-tle.
The group discriminant used was TiO2, which showsa marked difference between Volcanic Units 48 to 52, inthe central portion of the hole, relative to those at thetop and bottom, and allows division of the basalts intothree megagroups, A, B, and C, in order of increasingdepth (Figs. 9, 10). Little overlap between the megagroupsis seen even when the analytical error is taken into ac-count, although differences between megagroups A andC are less marked when the standard deviation is consid-ered. Average TiO2 ± standard deviation is 1.28 ± 0.03for group A, 1.22 ± 0.04 for B, and 1.31 ± 0.02 for C.Other chemical distinctions between the three megagroupsare apparent, however, from averaged data (Table 8) andelemental distributions relative to TiO2 content (Fig. 11).
Megagroup A, in particular, is characterized by gener-ally higher siderophile-element contents (but oddly lowerSe), whereas megagroup C is typical higher in many in-compatible trace elements. At a particular TiO2 content,the major distinctions are in the levels of siderophile traceelements, but incompatible-element variations are less welldefined (Fig. 11).
Because the FeO*/MgO ratio shows some variabilitycaused by secondary alteration, the Zr content has beenused as a stable index to demonstrate variable fractiona-tion, even though the range (25 ppm) is very limited(Fig. 12). Variably phyric basalts are present in each ofthe megagroups, and overlap the compositional rangeshown by any associated aphyric basalts, as for exam-ple, aphyric Volcanic Unit 45 and phyric Unit 46 (Fig.12, Zr-Cr plot). The gross variation displayed can bebroadly related via fractionation of the main observedphenocryst phases within each megagroup:
1. Megagroup A (phenocrysts: clinopyroxene-olivine,minor Plagioclase). Clinopyroxene fractionation is indi-cated by the decrease in both Cr and Se with progressiveZr content. Ni also decreases (not shown), and is indica-tive of olivine fractionation, although alteration of oliv-ine microphenocrysts to smectite has caused considera-ble scatter in the Ni data. Y content shows little varia-tion, and thus produces variable Zr/Y ratios, againsuggesting fractionation of clinopyroxene, which can ac-commodate Y in its structure. Partial melting involvinga Y-bearing phase (e.g., pyroxene, garnet) can also causevariation in Zr/Y, but to a much greater degree than cli-nopyroxene fractionation (Pearce and Norry, 1979). Thechemical stratigraphy (Fig. 9) shows that the lower Zr/Yratios are confined to Unit 46 and the phyric base of
484
PETROLOGY AND GEOCHEMISTRY
Table 6. Normative composition of Nauru Basin basaltic lava flows (Hole 462A), calculatedon anhydrous, carbonate-free basis and an Fe2O3/FeO of 0.15.
Keelenumber
462 A-1462A-2462A-3462A-4462A-5462A-6462A-7462A-8462A-9462A-10462A-11462A-12462A-13462A-14462A-15462A-16462A-17462A-18462A-19462A-20462A-21462A-22462A-23462A-24462A-25462A-26462A-27462A-28462A-29462A-30462A-31462A-32462A-33462A-34462A-35462A-36462A-37462A-38462A-39462A-40462A-41462A-42462A-43462A-44462A-45462A-46462A-47462A-48462A-49462A-50462A-51462A-52462A-53462A-54462A-55462A-56462A-57462A-58462A-59462A-60462A-61462A-62462A-63462A-64462A-65462A-66462A-67462A-68462A-69462A-70462A-71462A-72462A-73462A-74462A-75
Qz
_
0.520.10
—
0.45
_——
0.20—
3.57
—
0.18
1.55—
1.11
—1.28
1.050.511.440.92
1.380.400.760.721.461.510.65_——
———
—
0.120.071.231.551.211.85
1.11—
0.611.951.381.47——
_—
—
Or
0.780.830.820.900.710.770.710.720.890.771.920.720.720.990.720.770.773.180.710.900.771.020.710.710.720.710.710.780.780.720.710.770.770.780.720.710.770.710.770.770.910.841.011.250.770.831.010.780.770.840.780.840.840.780.770.710.710.720.710.892.583.382.522.020.830.780.770.121.010.890.960.780.770.770.78
Ab
21.7820.9121.1121.2819.8020.8320.3520.4021.5220.7819.7420.5920.9216.0723.3821.4021.5520.8619.9320.8620.0019.1522.5320.2521.1521.6823.9724.2821.0420.2219.9219.4319.5919.5019.9819.7620.5020.4120.5620.0520.2520.3620.1619.5220.7920.3920.0921.4623.1319.9325.5824.4020.7821.7920.0821.1520.3720.8519.8019.9719.0220.9520.9522.0021.3520.2720.2320.1821.5521.6421.3822.1521.3825.7930.14
An
28.0326.9926.1727.1427.5027.5827.0526.4425.2727.7826.4427.3726.9330.1726.5126.3926.4425.7426.7627.3027.8927.1926.3026.6527.0526.0126.1025.9926.7128.0926.8427.3627.2127.3827.5427.1027.1126.6126.4527.6527.4227.5627.5327.3726.9827.2827.3727.1026.1227.4624.8124.6427.4026.8027.2626.4727.1526.1926.8426.7226.5325.5025.3025.1026.1926.7726.0528.3325.6126.0726.0425.5726.7025.4822.06
Di
24.9423.8724.0125.3324.7324.2724.4023.4525.1324.7921.6019.6924.4410.5323.0424.2724.1422.7325.6624.7824.2324.7225.6224.1225.6123.8724.8325.0024.0924.8424.5124.4524.1623.8923.9923.8623.6124.2924.9523.3522.7923.9524.6225.2425.8825.1024.8026.0025.2024.7825.3424.6324.3224.5025.8625.7524.7024.1223.9524.1722.2623.9323.2024.3824.2223.8425.1822.9426.5625.2426.0125.7324.3423.7724.84
Hy
11.6621.7022.5717.7217.1919.5921.9223.7919.6915.7625.0626.1618.8533.4518.8920.7620.0219.7420.8215.1117.2122.7112.9421.7320.0821.7514.6414.1721.1718.7422.0922.5721.8922.6722.7122.3822.6122.2121.5821.7422.0621.6516.7319.3915.0019.3418.9118.0912.4418.2917.2617.4719.1419.5318.3320.8322.0321.9022.1122.0522.7420.8721.7021.0621.7821.3921.3421.9017.7014.7316.1518.7717.1612.224.54
Ol
7.58——
2.594.781.81
0.022.364.910.13
2.98
2.371.322.092.690.965.844.52
_6.79
0.48
4.894.83
—2.42
0.17__
——
4.852.125.511.992.761.577.343.551.332.912.491.762.67
0.43—
0.45—
2.236.244.221.744.436.83
12.48
Mt
2.422.492.472.332.5324.12.432.472.502.432.492.532.412.552.322.442.372.422.512.512.532.422.472.352.352.312.272.292.322.422.402.412.422.372.402.332.432.412.402.372.402.392.492.472.472.452.432.352.442.502.252.372.392.362.462.422.422.432.462.422.452.372.452.412.442.392.432.312.522.472.492.522.512.332.34
11
2.552.432.492.482.532.502.442.462.422.522.372.512.532.452.542.412.392.392.412.482.482.382.412.352.312.312.362.412.352.332.252.272.282.272.262.262.352.352.342.372.412.392.392.402.382.362.392.422.342.402.402.492.422.422.352.332.332.322.362.332.342.352.522.342.362.362.392.522.492.472.532.512.482.562.56
Ap
0.260.260.260.240.260.260.260.260.240.260.260.260.240.220.260.260.240.260.260.240.260.240.260.260.260.260.240.260.260.240.240.240.240.240.240.240.240.260.240.240.270.240.240.260.240.260.260.260.240.260.260.260.220.260.240.240.240.260.240.260.240.240.260.250.230.270.240.240.260.260.240.240.260.260.26
% Mgendmember
Di
535052545353525353535353545654535553525352525352535456545354535352545454525353535252535253525353535353535454525253525152515251525252525450545250525454
En
504749515050495049505050515351505150484949494949505152515049495049515051494949494949494950495049504949505050494849484849484947494849495147504947495051
Fo
47——494747—47474846—48—48474948464746—47—47—4948—47————48———————47464746474647464747484846—————_47—46——_—44484645474848
Note: Qz = quartz; Or = orthoclase; Ab = albite; An = anorthite; Di = diopside; Hy = hypersthene; Ol =olivine; Mt = magnetite; II = ilmenite; Ap = apatite; En = enstatite; and Fo = fosterite.
485
P. A. FLOYD
MAR Leg 49
MAR Leg 37
EPR Leg 54
01 Hy
Figure 7. Distribution of Hole 462A basalts in normative basalt trape-zohedron, compared with MORB data from Mid-Atlantic Ridge(MAR), Legs 37 and 49 (Flower et al., 1977; Tarney et al., 1979)and East Pacific Rise (EPR), Ocean Crust Panel (Ridge), and Si-queiros Fracture Zone (Humphris, Thompson, Gibson, et al., 1980).Olivine-plagioclase-clinopyroxene-liquid cotectic at 1 atm. pres-sure from Thompson (1982). Qz quartz, Hy = hypersthene, Di= diopside, Ol = olivine; Ne = nepheline.
Unit 45, both of which contain clinopyroxene as the dom-inant phenocryst phase (Table 4). As seen in Figure 13,the vast preponderance of the basalt suite has a rela-tively uniform Zr/Y ratio (average 2.3; range 2.2-2.5),whereas the slight increase in Zr/Y with progressive Zrcontent for Units 45 and 46 in particular is more indica-tive of fractionation than of partial melting. Plagioclasecan also occasionally occur as a phenocryst phase, butthe Sr distribution (Fig. 12) has been too disturbed byalteration to show any definite fractionation trend.
2. Megagroup B (phenocrysts: clinopyroxene, Plagio-clase, olivine). Overlap again occurs between phyric Units49 to 51 and aphyric Unit 52, although aphyric Unit 48is generally less evolved than the others (Fig. 12, Cr-Zrplot). Cr again decreases, but the lack of significant var-iation in Se suggests that clinopyroxene fractionation wasrelatively minor. The systematic increase in Y and Zr (es-sentially constant Zr/Y ratio) also precludes appreciableclinopyroxene removal, and is typical of olivine and/orPlagioclase fractionation, which on precipitation causesa corresponding increase in both Zr and Y, owing to ex-clusion from these phases. Olivine involvement is cer-tainly indicated by a large decrease in Ni with progres-sive Zr content. Plagioclase was also fractionated, foralthough Sr is again variable, the more evolved basaltstend to have the lower Sr contents.
In this megagroup, unlike A, it seems possible thatphyric Units 49 to 51 may not be in fact directly relatedto the adjacent aphyric Units 48 and 52. The aphyricunits can be linked via olivine-plagioclase and minorclinopyroxene fractionation, as seen by decreasing Ni,Sr, and Cr, respectively. The phyric units, however, haveconstant Cr, and are all slightly enriched in TiO2, Sr, La,and Ce, compared with aphyric Unit 52 at the same stageof chemical evolution. The phyric units also have con-
stant, marginally lower Zr/Ce ratios (4.3) compared withthe higher, more variable, but similar ratios (6.1) of thetwo aphyric units (Fig. 14). The La/Nb ratio is also high-er in the aphyric units. As both of these ratios are unaf-fected by fractionation or partial melting, but reflect thesource composition, two sources of slightly different com-position are indicated for the megagroup B phyric andaphyric units. These features suggest that the phyric unitsof megagroup B are related via olivine and Plagioclasefractionation, but were developed in a magma chamberindependent of the aphyric units. The parental magmaswere probably derived from slightly different sources (thephyric lavas being representative of the more enrichedsource), which would account for the differing Zr/Ceand La/Nb ratios and incompatible-element levels.
3. Megagroup C (phenocrysts: minor clinopyroxeneand Plagioclase). The internal chemistry of this mega-group is not markedly influenced by clinopyroxene frac-tionation, since Cr is essentially constant and Se actu-ally increases with Zr content (Fig. 12). Olivine has notbeen recognized as a phenocryst phase, except that itsinfluence is suggested by an irregular decrease in Ni. Onthe other hand, Sr shows a decreasing trend with in-creasing Zr, suggesting that Plagioclase is the major frac-tionating phase, as it is actually observed to be in thebasalts. The Zr/Ce, La/Nb, and P2O5/Ce ratios, as wellas the levels of stable incompatible elements (Figs. 9 and14 and Table 8), are different from those of the preced-ing megagroups, and probably indicate a slightly light-REE-enriched source compared with that of megagroupsA and B.
If the least evolved basalts of megagroups A and Brepresent, or approximate, "parental" compositions, thenthe only major difference between them is that the me-gagroup A "parent" has higher Cr and Ni, and a corre-spondingly lower FeOVMgO ratio, whereas incompati-ble-element abundances are only marginally lower. Vari-ation in siderophile-element abundances indicates thatthe "parental" magmas could be related via high-pres-sure mafic fractionation before their subsequent low-pres-sure evolution in separate magma chambers. However,the similarity of Se contents in the supposed initial com-positions precludes extensive high-pressure pyroxene frac-tionation, even though Cr is variable. It seems more like-ly that the parental magmas were related (if at all) viaminor high-pressure olivine fractionation. The uniform-ity of Zr/Y ratios between the "parental" compositionsrules out their generation by different degrees of partialmelting, and the source also appears essentially homog-enous in terms of similar La/Nb and Ce/P2θ5 ratios.Another possibility that has not been fully explored, how-ever, is that the two "parental" compositions reflect var-iable degrees of magma mixing of a more primitive melt(with higher Ni content) and a more evolved melt (rela-tively lower Ni), the latter of which again implies olivinefractionation relative to the former.
Trace-Element Fine Structure within the ChemicalMegagroups
Notwithstanding the limited compositional rangeshown by most trace elements within the megagroups,
486
PETROLOGY AND GEOCHEMISTRY
Table 7. Average chemical compositions of Nauru Basin sheet-flow basalts (Legs 89 and 61 compila-tions), various MORB types, and Hawaiian tholeiite.
SiO2
TiO2
A12O3
Fe2O3
FeOMnOMgOCaONa2OK2O
P2O5+H 2O +
CO 2
BaCeCrLaNbNdNiRbSeSrYZr
Ti/ZrLa/NbP 2 O 5 /CeTiO 2 /P 2 O 5
Zr/NbTi/YZr/Y
X
50.171.25
13.904.088.820.226.73
11.372.460.160.101.170.05
1313
14955
10824
461102964
1181.0
7712.512.8
2592.2
1Leg 89
S.d.
0.690.050.260.470.500.030.280.440.210.090.010.880.08
104
1410.72522335
Range
47.44-51.521.15-1.34
12.89-14.513.18-6.606.74-9.740.14-0.466.05-7.829.02-11.921.81-3.510.12-0.570.09-0.110.54-6.160.02-0.51
5-446-24
115-1793-84-85-16
66-912-9
40-51103-12222-3355-78
2Leg 61
X
49.390.99
13.96—
11.280.207.84
11.942.060.07—1.490.43
108
3033
1157
46992679
75
———
2283.0
3N-typeMORB
4.981.67
16.02.07.50.187.5
11.22.80.130.150.210.13
129
393338
1461
4412433
100
1001.0
16711.133.3
3043.0
4T-tvoe* lJFv
MORB
48.151.76
13.594.107.250.187.74
11.892.120.290.201.160.21
5622
235131314894
4314331
108
981.0
918.88.3
3413.5
5P-tvüeMORB
49.013.02
14.046.437.840.245.469.822.780.480.391.520.12
964246192925328
3926944
213
850.7
937.77.3
4124.8
6Hawaiiantholeiite
49.232.50
13.933.098.520.168.40
10.352.160.380.260.610.10
20231
269121421
1268
3135026
161
930.9
849.6
11.5576
6.2
7Chondrite
average
620a
———————
120a
46a
——
6.90.865—0.3280.350.63—0.357.9
11.82.06.84
910.9
1229.8
19.5310
3.4
Note: References: 1—this work (75 samples); 2—compilation calculated from data in Larson, Schlanger, et al. (1981);3—averages from Condie (1976, major elements), Delaney et al. (1978, H 2O + CO2), Sun and Nesbitt (1977) andSun et al. (1979), trace elements, TiO2 , K2O, P 2 Os; 4—compiled average for Reykjanes Ridge, DSDP Holes 407,408, and 409 (Tarney et al., 1979 + Appendix IV; Floyd and Tarney, 1979); 5—compiled average for E. IcelandTertiary upper lava group (Flower et al., 1982; Gibson et al., 1982); 6—compiled average for Hawaiian tholeiites(Macdonald and Katsura, 1964; Easton and Garcia, 1980; Leeman et al. 1980; Schilling and Winchester 1969);7—values from Thompson (1982), based partly on Sun and Nesbitt (1977). — = no data.
a Ti, K, and P values as elements in ppm.
small but sharp jumps in abundance and reversals in sys-tematic trends do occur with depth (Fig. 9). This allowsthe basalt flows to be divided into small chemical sub-groups (22 in all), each of which is apparently composedof internally related members. These features are similarto those seen in many MOR sections (e.g., Flower et al.,1977; Natland, 1979b), although here the cyclicity of che-mical groups, and their variation, are much more sub-dued. How significant these subgroup variations are is dif-ficult to say, but two possible genetic models are suggested:(1) post-magma-chamber fractionation, and (2) magmamixing within a replenished magma chamber. Supportfor the first suggestion comes from the fact that the mostprimitive composition of each subgroup (within any me-gagroup) could be related via mafic fractionation, withthe possible exception of the subgroups in phyric Units49 to 51. It is tentatively suggested that this relationshipcould occur after initial low-pressure fractionation with-in the parent magma chamber, with separate magmapulses expelled undergoing further minor fractionation(or flow differentiation) as they rose toward the oceanfloor. The two generations of clinopyroxene phenocrysts
observed in many basalts lend some support to this, inthat the glomerocrysts probably crystallized within themagma chamber, whereas the isolated granular micro-phenocrysts (on which variolitic clinopyroxene is nucle-ated) crystallized on route in feeder channels near thesurface. On the other hand, a more likely (and preferred)model is that the "sawtooth" subgroup variation repre-sents magma mixing between a continuously fractionat-ing stored magma and influxes of new magma from be-low (cf. O'Hara and Mathews, 1981). Each subgroup re-presents the eruption of a variably fractionated magmabatch, and the chemical jump to a new subgroup indi-cates replenishment of the magma chamber by new meltof generally more primitive composition.
DISCUSSION AND CONCLUSIONS1. Leg 89 intraplate basalts represent a series of rap-
idly extruded sheet-flows composed of thin, interlayeredaphyric and poorly clinopyroxene-plagioclase-olivine phy-ric, olivine- to quartz-normative tholeiites. Twelve vol-canic units have been recognized on the basis of observ-able contacts and textural and lithologic variation. All
487
P. A. FLOYD
30
20
10
0
50
40
30
20
10
0.9 1.1 1.3Ti0o
10.3
1.24 1.56 1.88FeOVMgO r r T , D Leg 61
E2 Leg 89
0.04 0.08 0.12 0.16K00
I(M
o
1.3
1.1
0.9
-
—
-
-
I
0 °o0 0
. °° o•
0 0
•• ***#I 1
D
0
o c
°°o
•
oo o'
&c
1
00
8bo°
.
o
0. °
o°\ Too o o8o
Leg 61Leg 89
i
10 11 12
FeO* (wt.%)
13
Figure 8. Comparison of TiO2, K2O, FeO*, and FeOVMgO distribu-tions in Leg 61 (data taken from Larson, Schlanger, et al., 1981)and Leg 89 (this work) samples.
the units are essentially nonvesicular, or if vesicles are pres-ent they are very small and indicate extrusion under verydeep water. The basalts exhibit only mild hydrous andoxidative alteration effects under reducing, low-CO2-ac-tivity conditions, during low-temperature reaction withcold seawater. Secondary assemblages are dominated byvarious smectites, zeolites, pyrite, with minor celadon-ite, K-feldspar, albite, and quartz, all developed underzeolite-facies conditions. Glass is invariably replaced bypalagonite and/or brown smectite. No systematic chem-ical changes in alteration intensity relative to depth areexhibited.
2. Chemically the basalts are remarkably uniformthroughout, with low incompatible-element abundances,but are moderately evolved as regards primary composi-tions (having low Mg numbers) and their high total Fe.These apparently opposing chemical features probablyreflect a low incompatible-element source coupled witha relatively high degree of fractionation soon after meltsegregation. A high proportion of the samples have chem-ical (> 12 wt.% FeO*) and mineralogical (paucity of ol-
ivine; clinopyroxene-plagioclase glomerocrysts) featurestypical of ferrobasalts from Pacific Ocean rifts (e.g.,Galapagos Rift and EPR; Natland, 1980), although in-compatible elements are by comparison generally low atspecific FeO* levels. Models for the generation of ferro-basalts associated with fast-spreading ridges indicate highdegrees of low-pressure fractionation (-74% of paren-tal magma) in flat-topped, laterally extensive, isolated,unreplenished magma chambers (Clague and Bunch,1976; Natland, 1980). The regionally extensive nature ofsubmarine sheet-flows in the Nauru Basin, and their iso-lation from spreading ridges during the Cretaceous, pre-clude a direct tectonic comparison of these ferrobasalts.Ferrobasalts are also commonly associated with Icelan-dic tholeiites, and form part of a progressively fraction-ated lineage from more primitive compositions (Wood,1978). However, although FeO* values are comparable,Icelandic ferrobasalts are more chemically evolved thanthe Nauru Basin rocks, and exhibit higher FM values(FeO* + MnO/FeO* + MnO + MgO = 74 to 77) andincompatible-element abundances (e.g., TiO2 3.5-4.0wt.%) and lower MgO contents of between 4 and 5 wt.%(Wood, 1978; Flower et al., 1982). By comparison, theNauru Basin ferrobasalts appear to have had a differentgenesis or to have been derived from a source-compo-sition different from that of ridge-associated Fe-richbasalts.
3. Comparison of average chemical compositions (Ta-ble 7; Fig. 15) indicates that, overall, the Nauru Basinbasalts are only mildly enriched in the highly incompati-ble elements, compared with N-type MORB. Many in-compatible-element ratios (Th/U, La/Nb, P2O5/Ce,TiO2/P2O5, Ti/Zr, Zr/Nb, Ti/Y, Zr/Y) have values gen-erally lower than or close to chondritic. A mildly LIL-depleted, MORB-like, approximately chondritic sourcefor the basalts is suggested. It seems unlikely that theminor enhancement of Ba, K, Rb, Th, etc., relative toN-MORB, is a reflection of alteration-enrichment, be-cause this feature is typical of the vast majority of thebasalts, including the least altered. Whether or not theintraplate basalts were contaminated by highly altered,LIL-enriched Jurassic ocean crust, through which themagmas must have passed, is difficult to assess. Sr-iso-tope data for Leg 61 basalts (Fujii et al., 1981) are uni-form but marginally higher (0.7037) than N-type MORB,and could equally reflect minor differences in source or87Sr contamination by Jurassic MORB enriched by sea-water alteration. Possible contamination aside, the Nau-ru Basin intraplate sheet-flows are similar to mildly en-riched types of MORB (e.g., Reykjanes Ridge, Fig. 15).
Chemical discrimination of the tectonic environment,using either major or trace elements, gives disparate re-sults, with the Nauru Basin intraplate basalts plotting inisland-arc, or straddling island-arc, MORB fields (Fig.16). Clearly, if basalts similar to these were now foundin a tectonically complex setting on continental crust,chemical discrimination of the environment would be er-roneous and would probably designate them as MORB—that is, plate-margin rather than intraplate.
4. Although the basalts are characterized by a gen-eral uniformity of composition, as already stated, varia-tions do occur which allow them to be divided, on the
488
PETROIX>GY AND GEOCHEMISTRY
S <2
S 1FeOVMgO
1.6 1.8 2.0i—i—r
TiO,
1.2 1.3 1.4
Cr
120 160
Ni
80 90
Zr
60 70
Ce
10 20
Zr/Y
2.0 2.6
La/Nb
0.5 1.5T—i—r
> <1080-
95 J1100"
46
1120-
47
100
1140H 8483
49
50
1160-
51
105
1180-
54
1200- 55
56
v B8B Bg
i
i—r
i
I
i—rr
T
T—r
<
/
\
i
x
r
i—r
S
S ; Mainly aphyric units Mainly phyric units
Figure 9. Chemical stratigraphy in Hole 462A (Leg 89), showing the distribution of the megagroups A, B, and C and various subgroups withdepth.
489
P. A. FLOYD
1.3 -
1.2 -
A A
• B
c
//1
1
1
1
1
' I
•
- —
1
A
,1
A
A A~~A
g-J
•
1
"A ^-'°
A +
— - ± _ _ —A /
— A A;-*""""•
I
>- Units 53-56
\£ -
Units 45-47
y/--Units 48-52
60 70 80Zr (ppm)
Figure 10. TiO2-Zr plot of Hole 462A (Leg 89) basalts showing the di-vision of the volcanic units into three chemical megagroups A, B,andC.
basis of TiO2 content, into three chemical megagroups(A, B, C) with depth. Each megagroup is conceived asrepresenting a high-level magma chamber that stores mag-ma for variable periods before eruption. The eruptive
stratigraphy developed reflects the progressive fractionalcrystallization of such magma bodies, together with mix-ing caused by the influx of more primitive magma at thebase.
In general, aphyric and poorly phyric volcanic unitswithin the megagroups overlap in chemical composition,the various basalts being related via mafic and/or pla-gioclase fractionation. The small range of chemical var-iation, the lack of cumulates, and the generally low pro-portion of observed phenocrysts in any megagroup sug-gest that the magma chambers were probably small andshort-lived, such that extensive fractionation did not takeplace. Also, if eruption is related to replenishment ofthe chamber by new magma batches from below (Sparkset al., 1980), then rapid extrusion implies a semicontinu-ous supply of new magma and consequently the oppor-tunity of magma mixing (O'Hara and Mathews, 1981)that periodically interrupts the fractionation process. Insuch a process the extrusive stratigraphy would be ex-pected to show a number of small chemical breaks be-tween eruptive units derived from such a magma cham-ber. The more frequent or larger the replenishment, thesmaller the chemical variation might be, especially if den-sity differences between stored and injected melts wereminor and mixing could readily occur. As shown in the
Table 8. Mean and standard deviation (s. d.) of chemical megagroups A, B, and C in NauruBasin basaltic sheet-flows, Hole 462A (major oxides in wt.%, trace elements in ppm).
SiO 2
TiO 2
A1 2O 3
F e 2 O 3
FeOMnOMgOCaONa 2 OK 2 Op 2 θ 5 ^H 2 O +
C O 2
Total
BaCeCrLaNbNdNiRbSeSrYZr
FeOVMgOTi/ZrTi/YZr/YLa/NbP 2 O 5 / C eT i O 2 / P 2 O 5
(45, 46, 47
Mean
49.931.28
13.954.108.980.226.90
11.342.410.160.110.990.12
100.49
1513
164559
864
431103064
1.84120256
2.131.00
84.611.6
A; 22 samples)
S. d.
0.520.030.240.650.670.060.290.630.170.090.010.490.11
1025111322424
Chemical groups(volcanic units:
B(48, 49, 50, 51, 52
Mean
50.311.22
13.874.008.790.216.65
11.392.440.160.101.250.03
100.42
12 112
146459
814
461102864
1.86114261
2.280.80
83.312.2
; no. of samples)
• 45 samples) (53,
S. d.
0.800.040.270.360.380.020.240.330.160.090.0051.070.03
03810.62422335
C54, 55, 56;
Mean
49.911.31
13.914.488.550.226.69
11.322.690.130.111.240.03
100.59
1419
12375
1276
448
1143069
1.88114262
2.301.41
57.911.9
8 samples)
S. d.
0.290.020.270.150.490.020.290.380.380.050.010.190.01
123
11112321326
490
PETROLOGY AND GEOCHEMISTRY
90
80
-
•• 4
4
•
AA
A AA• A A•
• A• AA» ••>•••••
1
pA
A
A 1
AAAA
\A
Megagroups
A A
• B
+ C
35
I 30
25
1.2Ti0
1.3
•AA A A / A
++ +• • • Aj
••1.4 1.2 1.3
TiO2 (wt.%)1.4
ò
160
120
- l••~-
A
A A
j : .it
i
•
i
L A
I1.2 1.3
O
20
10
_
-•
. f• \A AAA \ t• A A A•• A•• A A• • • AA••1 1
+ + /AA
A
AA
11.4 1.2 1.3 1.4
Ti0 TiO
50
45
40
• oA• ••• •• A• • • • • • • A
• • • A A• A A A A
A A AAAAA A
A
15
T3 10
•
1.2 1.3 1.4 1.2 1.3 1.4
TiO TiO
Figure 11. Distribution of various compatible and incompatible elements relative to TiO2 in chemical mega-groups A, B, and C.
preceding section, the megagroups can be divided intosmall chemical subgroups, the "sawtooth" variation ofwhich probably reflects a mixing process. Some chemi-cal subgroups show internal fractionation, whereas oth-ers are essentially uniform in composition and representcomplete mixing with large quantities of new melt to asteady-state composition. In the previous section it wassuggested that minor fractionation in the feeder conduitsafter expulsion from the magma chamber might also beresponsible for the chemical subgroups. It seems morelikely, however, that the small fluctuations in low-KD el-ement ratios throughout the pile are more satisfactorilyexplained on a magma-mixing model, whereas if frac-tionation were the only factor, such ratios would be con-stant for the megagroup as a whole.
In conclusion, the following magma plumbing systemis envisaged. Magma originates by the relatively high, but
uniform, partial melting of a MORB-like depleted man-tle. On the basis of various low-KD element ratios, thereis some evidence to suggest minor differences in sourcecomposition, with a more LIL-enriched source (relativeto A and B) providing the initial melt for megagroup C.The degree of melting is very similar, however, in bothsources.
If the original magmas were in equilibrium with nor-mal mantle peridotite (e.g., Frey et al., 1978), they musthave undergone some olivine (and minor clinopyroxene)fractionation to produce three parental melts of variablesiderophile- and incompatible-element contents that se-gregated in separate, small, high-level magma chambers.Variable internal fractionation takes place, such that allthe flows originating from one chamber are grossly re-lated via limited clinopyroxene-plagioclase-olivine pre-cipitation. Extensive progressive fractionation is not
491
P. A. FLOYD
M 4 5 "
ü
εα.a.>
E 160 -
120 ~
25 ~
6 0 70
Zr (ppm)
80
Figure 12. Distribution of various elements relative to progressive fractionation (as measured by Zr content)within the three chemical megagroups A, B, and C.
2.5
2.0
•-
••
•
/1+
—r
••
i
•
•
•
•
+.±-
•
-f•
+
+•
1
-» + + • • • i• •! m+ + v
^ Units 45 and 46
60 70
Zr (ppm)80
Figure 13. Zr/Y-Zr plot for Hole 462A (Leg 89) basalts, showing theessentially constant Zr/Y ratio for the majority of samples (•) rela-tive to a progressive increase in Volcanic Units 45 and 46 ( + ) . Sam-ples with Zr/Y at about 2.6-2.7 correspond to the base of Unit 48.
7 -
O 5
N
3 -
A
•
+
-
_ \
ABC
I
• *•••
Units 49-51
I
•# A * A• * A
A A Λ
I
A
-
A A
120 140 160 180
Cr (ppm)
Figure 14. Zr/Ce-Cr plot for Hole 462A (Leg 89) basalts, showing theseparate clustering of the phyric Volcanic Units 49 to 51 relative tothe rest of chemical megagroup B, and also the typically lower Zr/Ce ratio of chemical megagroup C relative to A and B.
492
PETROLOGY AND GEOCHEMISTRY
100
50
10
I I I I I I I I I I I I I I I IBa Rb Th U K Nb Ta La Ce Sr Nd P Sm Zr Hf Ti Y Yb
10
1.0
0.5 ~I I I I I I I I I I I I I I I
Ba Rb Th U K Nb Ta La Ce Sr Nd P Sm Zr Hf Ti Y Yb
IIIIIIH Nauru Leg 89 basalts
A—• N-MORB*—4 T-MORB•—• P-MORB• — Hawaiian tholeiite
Figure 15. Spidergrams for MORB averages and Leg 89 basalts (mega-group averages enclosed within envelope) normalized relative to(A) chondrite abundances (factors from Thompson, 1982) and (B)N-type MORB (factors from Sun and Nesbitt, 1977). Comparativedata taken from Table 7. The high normalized Rb-value for Leg 89basalts is probably analytical.
achieved, however, owing to the frequent influx of newmagma, produced by similar degrees of melting, that mix-es with the stored magma. The magma chambers rapidlyexpel their melts, apparently in sequence, and were pro-bably short-lived.
ACKNOWLEDGMENTS
Thanks are due many shipboard colleagues, especially Dr. S. O.Schlanger, Dr. R. Moberly, and Dr. J. Haggerty, for enlightening dis-cussions on Pacific Cretaceous volcanology and tectonics. Technicalhelp on Glomar Challenger— in thin-section preparation, photogra-phy, and running the HNC Analyser—were much appreciated. Involve-ment in DSDP/IPOD shipboard studies would not have been possibleexcept for the continued support of the Natural Environmental Re-search Council, U.K., which is gratefully acknowledged. I thank DavidEmley and Margaret Aiken for their help and expertise in analyzingthe samples. Dr. A. Saunders provided constructive criticism and help-ful comment on this work, for which I am grateful. Dr. R. Batiza andDr. J. Bender are also acknowledged for their reviews of this chapter.
REFERENCES
Andrews, A. J., Barnett, R. L., MacClement, B. A. E., Fyfe, W. S.,Morrison, G., et al., 1977. Zeolite facies metamorphism, geoche-mistry and some aspects of trace element redistribution in alteredbasalts of DSDP Leg 37. In Aumento, F., Melson, W. G., et al.,Init. Repts. DSDP, 37: Washington (U.S. Govt. Printing Office),795-810.
Barringer, A. R., 1953. The preparation of polished sections of oresand mill products using diamond abrasives and their quantitativestudy by point counting methods. Trans. Am. Inst. Min. Metall.Engr., 63:26-41.
Batiza, R., 1981. Trace-element characteristics of Leg 61 basalts. InLarson, R. L., Schlanger, S. O., et al., Init. Repts. DSDP, 61:Washington (U.S. Govt. Printing Office), 689-696.
Batiza, R., Larson, R. L., Schlanger, S. O., Shcheka, S. A., and To-kuyama, H., 1980. Trace element abundances in basalts of the Na-uru Basin. Nature (London), 286:476-478.
Brooks, C , 1976. The Fe2O3/FeO ratio of basalt analyses: An appealfor a standardized procedure. Bull. Geol. Soc. Denmark, 25:117-120.
Clague, D. A., and Beeson, M. H., 1980. Trace element geochemistryof the East Molokai Volcanic Series, Hawaii. Am. J. Sci., 280A:820-844.
Clague, D. A., and Bunch, T. E., 1976. Formation of ferrobasalt ateast Pacific midocean spreading centres. J. Geophys. Res., 81:4247-4256.
Condie, K. C , 1976. Trace element geochemistry of Archaean green-stone belts. Earth-Sci. Rev., 12:393-417.
Delaney, J. R., Muenow, D. W., and Graham, D. G., 1978. Abun-dance and distribution of water, carbon and sulfur in the glassyrims of submarine pillow lavas. Geochim. Cosmochim. Acta, 42:581-594.
Duffield, J., and Gilmore, G. R., 1979. An optimum method for thedetermination of rare earth elements by neutron activation analy-sis. J. Radioanal. Chern., 48:135-145.
Easton, R. M., and Garcia, M. O., 1980. Petrology of the Hilina For-mation, Kilauea Volcano, Hawaii. Bull. Volcanol., 43:657-673.
Flower, M. F. J., Pritchard, R. G., Brem, G., Cann, J. R., Delaney, J.,et al., 1982. Chemical stratigraphy, Iceland Research Drilling Proj-ect Reydarfjordur, eastern Iceland. J. Geophys. Res., 87:6489-6510.
Flower, M. F. J., Robinson, P. T, Schmincke, H.-U., and Ohnmacht,W., 1977. Petrology and geochemistry of igneous rocks, DSDPLeg 37. In Aumento, F , Melson, W. G., et al., Init. Repts. DSDP,37: Washington (U.S. Govt. Printing Office), 653-679.
Floyd, P. A., and Fuge, R., 1982. Primary and secondary alkali andhalogen element distribution in Iceland Research Drilling Projectbasalts from eastern Iceland. J. Geophys. Res., 87:6477-6488.
Floyd, P. A., and Tarney, J., 1979. First-order alteration chemistry ofLeg 49 basement rocks. In Luyendyk, B. P., Cann, J. R., et al., In-it. Repts. DSDP, 49: Washington (U.S. Govt. Printing Office),693-708.
Frey, F. A., Green, D. H., and Roy, S. D., 1978. Integrated models ofbasalt petrogenesis: A study of quartz tholeiites to olivine meliliti-tes from southeastern Australia utilizing geochemical and experi-mental petrological data. J. Petrol., 19:463-513.
Fujii, N., Notsu, K., and Onuma, N., 1981. Chemical compositionand Sr isotopes of Deep Sea Drilling Project Leg 61 basalts. InLarson, R. L., Schlanger, S. O., et al., Init. Repts. DSDP, 61:Washington (U.S. Govt. Printing Office), 697-700.
493
P. A. FLOYD
Gibson, I. L., Kirkpatrick, R. J., Emmerman, R., Schmincke, H.-U.,Pritchard, G., et al., 1982. The trace element composition of thelavas and dikes from a 3-km vertical section through the lava pileof eastern Iceland. J. Geophys. Res., 87:6532-6546.
Haggerty, J. A., Schlanger, S. O., and Premoli Silva, I., 1982. LateCretaceous and Eocene volcanism in the southern Line Islands andimplications for hotspot theory. Geology, 10:433-437.
Hart, S. R., and Davis, K. E., 1978. Nickel partitioning between oliv-ine and silicate melt. Earth Planet. Sci. Lett., 40:203-219.
Hart, S. R., Erlank, A. J., and Kable, E. J. D., 1974. Sea floor basaltalteration: Some chemical and Sr isotopic effects. Contrib. Miner-al. Petrol., 44:219-230.
Hart, S. R., and Nalwalk, A. J., 1970. K, Rb, Cs and Sr relationshipsin submarine basalts from the Puerto Rico Trench. Geochim. Cos-mochim. Acta., 34:145-156.
Hilde, T. W. C , Uyeda, S., and Kroenke, L., 1977. Evolution of thewestern Pacific and its margins. Tectonophysics, 38:145-165.
Humphris, S. E., and Thompson, G., 1978. Hydrothermal alterationof oceanic basalts by seawater. Geochim. Cosmochim. Acta, 42:107-126.
Humphris, S. E., Thompson, R. N., Gibson, I. L., and Marriner, G.E, 1980. Comparison of geochemistry of basalts from the East Pa-cific Rise, OCP Ridge and Siqueiros Fracture Zone, Deep Sea Drill-ing Project, Leg 54. In Rosendahl, B. R., Hekinian, R., et al., In-it. Repts. DSDP, 54: Washington (U.S. Govt. Printing Office),635-669.
Humphris, S. E., Thompson, R. N., and Marriner, G. F , 1980. Themineralogy and geochemistry of basalt weathering, Holes 417Aand 418A. In Donnelly, T., Francheteau, J., Bryan, W., Robinson,P., Flower, M., Salibury, M., et al., Init. Repts. DSDP, 51, 52, 53,Pt. 2: Washington (U.S. Govt. Printing Office), 1201-1218.
Kirkpatrick, R. J., 1979. Process of crystallization in pillow basalts,Hole 396B, DSDP Leg 46. In Dmitriev, L., Heirtzler, J., et al., In-it. Repts. DSDP, 46: Washington (U.S. Govt. Printing Office),271-282.
Kurnosov, V. B., Kholodkevich, I. V., and Ya Shevchenko, A., 1981.Secondary minerals of basalts from the Nauru Basin, Deep SeaDrilling Project, Leg 61. In Larson, R. L., Schlanger, S. O., et al.,Init. Repts. DSDP, 61: Washington (U.S. Govt. Printing Office),653-671.
Larson, R. L., 1976. Late Jurassic and Early Cretaceous evolution ofthe western central Pacific ocean. /. Geomag. Geoelectr., 28:219-236.
Larson, R. L., and Schlanger, S. O., 1981. Geological evolution of theNauru Basin and regional implications. In Larson, R. L., Schlang-er, S. O., et al., Init. Repts. DSDP, 61: Washington (U.S. Govt.Printing Office), 841-862.
Larson, R. L., Schlanger, S. O., et al., 1981. Init. Repts. DSDP, 61:Washington (U.S. Govt. Printing Office).
Leake, B. E., Hendry, G. L., Kemp, A., Plant, A. G., Harvey, P. K.,et al., 1969. The chemical analysis of rock powders by automaticX-ray fluorescence. Chern. Geol., 5:7-86.
Leeman, W. P., Budahn, J. R., Gerlach, D. C , Smith, D. R., and Po-well, B. N., 1980. Origin of Hawaiian tholeiites: Trace elementconstraints. Am. J. Sci., 280A:794-819.
Macdonald, G. A., and Katsura, T., 1964. Chemical composition ofHawaiian lavas. /. Petrol., 5:82-133.
Miyashiro, A., Shido, F., and Ewing, M., 1969. Diversity and originof abyssal tholeiite from the Mid-Atlantic Ridge near 24° and 30°North latitude. Contrib. Mineral. Petrol., 23:38-52.
, 1971. Metamorphism in the Mid-Atlantic Ridge near 24°and 30°N. Phils. Trans. R. Soc. London, A268:589-603.
Moore, J. G., 1970. Water content of basalt erupted on the oceanfloor. Contrib. Mineral. Petrol., 28:272-279.
Moore, J. G., and Schilling, J. G., 1973. Vesicles, water and sulphur inReykjanes Ridge basalts. Contrib. Mineral. Petrol., 41:105-118.
Mullen, E. D., 1983. MnO/TiO2/P2O5: A minor element discrimi-nant for basaltic rocks of oceanic environments and its implica-tions for petrogenesis. Earth Planet. Sci. Lett., 62:53-62.
Natland, J. H., 1979a. Crystal morphologies in basalts from DSDPSite 395, 23°N, 46°W, Mid-Atlantic Ridge. In Melson, W. G., Ra-binowitz, P. D., et al., Init. Repts. DSDP, 45: Washington (U.S.Govt. Printing Office), 423-445.
, 1979b. Comparison of chemical and magnetic stratigraphyof basement rocks at DSDP Sites 332 and 395. In Melson, W. G.,
Rabinowitz, P. D., et al., Init. Repts. DSDP, 45: Washington (U.S.Govt. Printing Office), 657-677.
, 1980. Effect of axial magma chambers beneath spreadingcenters on the compositions of basaltic rocks. In Rosendahl, B.R., Hekinian, R., et al., Init. Repts. DSDP, 54: Washington (U.S.Govt. Printing Office), 833-850.
Norrish, K., and Hutton, J. T., 1969. An accurate X-ray spectrograph-ic method for the analysis of a wide range of geological samples.Geochim. Cosmochim. Acta, 33:431-453.
O'Hara, M. J., and Mathews, R. E., 1981. Geochemical evolution inan advancing, periodically replenished, periodically tapped, con-tinuously fractionated magma chamber. J. Geol. Soc, London,138:237-278.
Ozima, M., Saito, K., and Takigami, Y., 1981. ̂ A r / ^ A r geochrono-logical studies on rocks drilled at Holes 462 and 462A, Deep SeaDrilling Project, Leg 61. In Larson, R. L., Schlanger, S. O., et al.,Init. Repts. DSDP, 61: Washington (U.S. Govt. Printing Office),701-703.
Pearce, J. A., 1975. Basalt geochemistry used to investigate past tec-tonic environments on Cyprus. Tectonophysics, 25:41-68.
, 1976. Statistical analysis of major element patterns in ba-salts. J. Petrol., 17:15-43.
Pearce, J. A., and Cann, J. R., 1973. Tectonic setting of basic volcanicrocks determined using trace element analyses. Earth Planet. Sci.Lett., 19:290-300.
Pearce, J. A., and Norry, M. J., 1979. Petrogenetic implications of Ti,Zr, Y and Nb variations in volcanic rocks. Contrib. Mineral. Pet-rol., 69:33-47.
Pritchard, R. G., Cann, J. R., and Wood, D. A., 1979. Low-tempera-ture alteration of oceanic basalts, DSDP Leg 49. In Luyendyk, B.P., Cann, J. R., et al., Init. Repts. DSDP, 49: Washington (U.S.Govt. Printing Office), 709-714.
Sato, H., 1977. Nickel content of basaltic magmas: Identification ofprimary magmas and a measure of the degree of olivine fractiona-tion. Lithos, 10:133-120.
Schilling, J.-G., and Winchester, J. W , 1969. Rare earth contributionto the origin of Hawaiian lavas. Contrib. Mineral. Petrol., 23:27-37.
Schlanger, S. O., and Premoli Silva, I., 1981. Tectonic, volcanic andpaleogeographic implications of redeposited reef faunas of LateCretaceous and Tertiary age from the Nauru Basin and the Line Is-lands. In Larson, R. L., Schlanger, S. O., et al., Init. Repts.DSDP, 61: Washington (U.S. Govt. Printing Office), 817-827.
Sparks, R. S. J., Meyer, P., and Sigurdsson, H., 1980. Density varia-tion amongst mid-ocean ridge basalts: Implications for magmamixing and the scarcity of primitive lavas. Earth Planet. Sci. Lett.,46:419-430.
Sun, S. S., and Nesbitt, R. W., 1977. Chemical heterogeneity of theArchaean mantle, composition of the earth and mantle evolution.Earth Planet. Sci. Lett., 35:429-448.
Sun, S. S., Nesbitt, R. W., and Sharaskin, A. Y., 1979. Geochemicalcharacteristics of mid-ocean ridge basalts. Earth Planet. Sci. Lett.44:119-138.
Tarney, J., Saunders, A. D., Weaver, S. D., Donnellan, N. C. B., andHendry, G. L., 1979. Minor-element geochemistry of basalts fromLeg 49, North Atlantic Ocean. In Luyendyk, B. P., Cann, J. R., etal., Init. Repts. DSDP, 49: Washington (U.S. Govt. Printing Of-fice), 657-691.
Thompson, R. N., 1982. Magmatism of the British Tertiary volcanicprovince. Scott. J. Geol., 18:49-107.
Thompson, R. N., Morrison, M. A., Dickin, A. P., and Hendry, G.D., 1983. Continental flood basalts...arachnids rule OK? In Hawk-esworth, C. J., and Norry, M. J. (Eds.), Continental Basalts andMantle Xenoliths: Nantwich (Shiva Publishing Ltd., U.K.), pp.158-185.
Tokuyama, H., and Batiza, R., 1981. Chemical composition of igne-ous rocks and origin of the sill and pillow basalt complex of NauruBasin, southwest Pacific. In Larson, R. L., Schlanger, S. O., et al.,Init. Repts. DSDP, 61: Washington (U.S. Govt. Printing Office),673-688.
Walsh, J. N., Buckley, F., and Barker, J., 1981. The simultaneous de-termination of the rare earth elements in rock using inductivelycoupled plasma source spectrometry. Chern. Geol., 33:141-153.
Windom, K. E., and Book, P., 1981. Vein minerals in basalt, Hole462A, Leg 61 of the Deep Sea Drilling Project. In Larson, R. L.,
494
PETROLOGY AND GEOCHEMISTRY
Schlanger, S. O., et al., Init. Repts. DSDP, 61: Washington (U.S.Govt. Printing Office), 647-651.
Winterer, E. L., 1973. Regional problems. In Winterer, E. L., Ewing,J. I., et al., Init. Repts. DSDP, 17: Washington (U.S. Govt. Print-ing Office), 911-922.
Wood, D. A., 1978. Major and trace element variations in the Tertiarylavas of eastern Iceland and their significance with respect to theIceland geochemical anomaly. J. Petrol., 19:393-436.
Wood, D. A., Gibson, I. L., and Thompson, R. N., 1976. Elementmobility during zeolite facies metamorphism of the Tertiary ba-salts of eastern Iceland. Contrib. Mineral. Petrol., 55:241-254.
Date of Initial Receipt: 8 March 1984Date of Acceptance: 4 July 1984
0.1
r10.3 0.5
-1.3
-1.5
-1.7
— 5000
MORB1000
Ocean-floor basalts
Island-arc tholeiites
50 100
Cr (ppm)
500 1000
Ti 10 V/' Nauru basalts(Leg 89)
Y × 3 MnO × 10 P2O5 × 10
Figure 16. Distribution of Hole 462A (Leg 89) basalts in various tectonic discrimination diagrams. Diagrams fromPearce and Cann (1973), Pearce (1975, 1976), and Mullen (1983). F, and F2 are major element discriminant func-tions (Pearce, 1976). MORB = mid-ocean ridge basalt field; IAT = island arc tholeiite field.
495
P. A. FLOYD
0.1 mm 0.1 mm 0.1 mm
0.1 mm 0.1 mm 0.1 mm
0.1 mm 0.1 mm 0.5 mm
0.1 mm 0.1 mm 0.1 mm
Plate 1. Photomicrographs illustrating petrographic and textural features of Hole 462A (Leg 89) submarine sheet-flow basalts (ppl = plane polar-ized light; xp = crossed polars). 1. Dark magnetite-rimmed pyroxene spherulites nucleated on skeletal Plagioclase microphenocrysts. The origi-nal vitreous matrix is replaced by clear palagonite and granular smectite, Sample 462A-103-1, 60-61 cm (ppl). 2. Fibrous pyroxene spherulitenucleated on clinopyroxene microphenocryst core, Sample 462A-102-5, 5-9 cm (xp). 3. Skeletal tuning-fork Plagioclase microlites in an essen-tially opaque hyalopilitic groundmass, Sample 462A-102-4, 59-63 cm (ppl). 4. Plumose clinopyroxene variolite nucleated on terminations ofPlagioclase microphenocrysts, Sample 462A-103-1, 60-61 cm (ppl). 5. Plumose clinopyroxene bundles, Sample 462A-102-4, 131-136 cm (xp).6. Well-spaced clinopyroxene blades containing small dark inclusions in an open-structured variolite, Sample 462A-101-2, 117-122 cm (xp). 7. Ske-letal serrated Plagioclase partly enclosed in a central, twinned clinopyroxene "nucleus," Sample 462A-101-4, 115-120 cm (xp). 8. CruciformPlagioclase bowties subophitically enclosed in a clinopyroxene "nucleus," Sample 462A-101-4, 57-61 cm (xp). 9. Open-structured variolite ofskeletal twinned clinopyroxene and Plagioclase, Sample 462A-106-2, 3-4 cm (xp). 10. Palagonite-replaced interstitial glass with a margin ofdark smectite, Sample 462A-103-1, 107-112 cm (ppl). 11. Interstitial palagonite replaced by a thin marginal zone and inward-growth hemi-spheres of fibrous smectite, Sample 462A-105-2, 4-5 cm (ppl). 12. Interstitial palagonite replaced by vermicules of dark smectite along relictperlitic cracks, Sample 462A-101-4, 115-120 cm (ppl).
496
PETROLOGY AND GEOCHEMISTRY
0.2 mm 0.2 mm 0.5 mm
0.5 mm 0.5 mm 0.5 mm
0.5 mm 0.5 mm 0.2 mm
0.1 mm 0.1 mm 0.2 mm
Plate 2. Photomicrographs illustrating petrographic and textural features of Hole 462A (Leg 89) submarine sheet-flow basalts (ppl = plane po-larized light; xp = crossed polars). 1. Hyalopilitic texture of quenched clinopyroxene plumes and magnetite grains, Sample 462A-98-6, 41-46cm (ppl). 2. Slightly coarser quench fabric than in Fig. 1, with clinopyroxene plumes, Plagioclase microlites, and minor glass, Sample 462A-104-2, 128-133 cm (ppl). 3. Variolitic clinopyroxene plumes and dark smectite-replaced interstitial glass, Sample 462A-102-4, 131-136 cm(ppl). 4. Coarser-grained quench fabric of clinopyroxene plumes and serrated Plagioclase blades, Sample 462A-100-2, 61-66 cm (ppl). 5. In-tersertal texture with interstitial glass, Sample 462A-108-2, 40-44 cm (ppl). 6. Coarse-grained intersertal to granular texture typical of holocrys-talline flow interiors, Sample 462A-106-3, 4-8 cm (ppl). 7. Carlsbad-twinned Plagioclase glomerocrysts, Sample 462A-95-5, 99-104 cm (xp).8. Glomerocrystic group of granular clinopyroxene and Plagioclase, Sample 462A-95-5, 61-63 cm (ppl). 9. Finely zoned Plagioclase microphe-nocryst, Sample 462A-95-7, 38-43 cm (xp). 10. Plagioclase microphenocryst partially replaced by flaky smectite, Sample 462A-99-1, 119-122cm (ppl). 11. Acicular zeolite needles and dark smectite in a plagioclase-rich area of coarse holocrystalline flows, Sample 462A-101-4, 115-120cm (ppl). 12. Plagioclase alteration domain with replacive albite, K-feldspar, and dark smectite, Sample 462A-102-1, 35-39 cm (xp).
497