33. PETROLOGY AND ORIGIN OF BASALTS OF THE MIDDLE AMERICA TRENCH,SOUTHERN MEXICO TRANSECT1

Y. I. Dmitriev, Institute of Geology and Ore Deposits, Petrology, Mineralogy and Geochemistry,USSR Academy of Sciences, Moscow, USSR

INTRODUCTION

The volcanism of Central America, according to cur-rent theory (Pichler and Weyl, 1973; Stoiber and Carr,1974; Hey, 1977), is related to the subduction of theCocos Plate under the North American lithosphericplate and the melting of ocean crust material in the sub-duction zone (Green and Ringwood, 1968; Dickinson,1970, Fitton, 1971).

Since Cocos Plate subduction occurs at the rate ofmore than 7 cm/y. (Hey et al., 1977), basalts underlyingupper Miocene sediments of the Middle America Trenchouter slope, penetrated in Hole 487 (Fig. 1) during Leg66 (Moore et al., 1979), should have formed far fromtheir present position if current theory is accurate.

Present manifestations of basaltic magmatism in ad-jacent areas of the Pacific derive from the axial part ofthe East Pacific Rise, the Galapagos spreading center,and transform fracture zones.

The question arises: Are there analogs of the MiddleAmerica Trench basalts among magmatic cock associ-ated with these modern features, or do the trench ba-salts have some other origin?

HOLE 487 BASALTSTo answer this question, let us compare the petrologi-

cal data of Hole 487 basalts, samples of which were pro-vided by the Deep Sea Drilling Project repository atScripps Institution of Oceanography, with my own andother scientists' published material on magmatic basicrocks of different structural elements in the Pacific seg-ment of the Earth's crust.

Hole 487 was drilled 11 km from the axis of the Mid-dle America Trench zone, at 15°51.21'N, 99°10.52'Wand at a water depth of 4764 meters. They reached ba-salt at 172 meters sub-bottom, under a layer of hemi-pelagic gray green mud and pelagic brown clay. I stud-ied Samples 487-20-1, 172-173.5 m and 487-20-2, 173.5-181.5 m.

Based on their petrographic and petrochemical char-acteristics, I subdivide Hole 487 basalts into two groups,olivine-plagioclase phyric and aphyric basalts (Samples487-20-1, 2-9 cm, 29-32 cm, 73-78 cm; 487-20-2, 62-64cm, 80-84 cm) and plagioclase phyric basalts (Samples487-20-1, 15-19 cm, 113-115 cm; 487-20-2, 34-36 cm).

16°45'N

99°30'W

Figure 1. Location map for Hole 487.

98°20'W

Initial Reports of the Deep Sea Drilling Project, Volume 66.

Rocks of both groups, according to the ratio Na2O +K2O/SiO2s derive from tholeiitic basalt magma.

Olivine-plagioclase phyric and aphyric basalts as wellas plagioclase phyric basalts have a fine-grained ground-mass consisting of a cryptocrystalline matrix, grains ofplagioclase, clinopyroxene, olivine, titanomagnetite, spi-nel, and patches of such secondary minerals as calciteand smectite (Table 1).

The main mineralogical differences between therocks studied are type and content of phenocrysts.Olivine-plagioclase phyric basalts contain as much as 11to 12% plagioclase phenocrysts and olivine. Plagioclasephyric basalts contain only 2 to 3% plagioclase pheno-crysts. Although aphyric basalts have no phenocrysts,or contain only a few grains, they are considered to-gether with olivine-plagioclase phyric basalts because oftheir virtual identity in chemical composition.

The texture of the groundmass of basalts is mainlymicrolitic and pilotaxitic.

The comparison of rock-forming mineral composi-tions of olivine-plagioclase phyric and plagioclase phy-ric basalts (Tables 2-6) shows that plagioclase phyric ba-salts have more calcic plagioclase in the groundmass,somewhat more ferruginous clinopyroxene, clearly moreferruginous olivine, titanomagnetite with higher TiO2

703

Y. I. DMITRIEV

Table 1. Modal composition of basalts, Hole 487 (vol.9/o).

Sample(interval in cm)

Phenocrysts:

plagioclaseolivinespinel

Ground mass:

plagioclaseclinopyroxeneolivinetitanomagnetitesecondary mineralsmatrix

487-20-12-9

4.52.3

r g.

40.449.6

—3.11.8—

487-20-129-32

2.90.9—

39.43.3—

Olivine-Plagioclase Phyric and Aphyric Basalts

487-20-la

45-49

——

37.92.8—

in matrix—

53.5—

59.3

487-20-173-78

1.80.7—

34.73.2—

—59.6

487-20-250-56

1.7r•g•

r•g•

28.363.1

0.41.05.5—

487-20-262-64

4.71.2—

26.63.4—

487-20-280-84

11.6r•g•

31.52.5—

in matrix—

34.1—

54.4

Averageof 7

3.90.4

r.g.

34.122.6

0.10.61.0

37.4

487-20-115-19

2.3—0.2

34.72.1—

Plagioclase Phyric Basalts

487-20-1113-115

1.6—

r.g•

44.650.0—

in matrix

60.7 3.8

487-20-234-36

0.6——

26.04.1—

Averageof 3

1.5—0.1

35.118.7—

in matrix

69.3 44.6

Note: r.g. = rare grains.a Aphyric basalt.

Table 2. Plagioclase in basalts, Hole 487. Table 3. Clinopyroxenes in basalts, Hole 487.

Siθ2Tiθ2A12O3

FeOMgOCaONa2θ

Total

SiTiAlFeMgCaNa% Ab<Vo AnName

Olivine-PlagioclasePhyric Basalts

Sample 487-20-2,

Phenocryst

Core

49.4—

33.90.380.17

18.01.1

99.95

2.138—

1.8410.0140.0010.8890.0989.8

90.2anorthite

Fringe

47.1—

32.80.340.28

17.51.7

99.72

50-56 cm

Lath inGroundmass

49.6—

30.30.640.36

15.42.4

98.7

Cations on 8 Oxygen

2.175—

1.7850.0130.0190.8660.152

14.585.5

bitownite

2.298—

1.6550.0250.0250.6810.215

22.777.3

bitownite

Plagioclase Phyric BasaltsSample 487-20-1,

113-115 cm

Phenocryst

45.2—

36.20.350.25

17.11.5

100.6

2.068

2.0620.0130.0170.8380.133

13.386.7

bitownite

Lath inGroundmass

47.80.1

34.31.50.12

16.12.2

102.12

2.1580.0041.8250.0570.0080.7780.193

18.681.4

bitownite

Note: Chemical compositions of rock-forming minerals in basalts in Hole 487were determined on an electron microprobe by G. N. Muravitskaya of theInstitute of Geology of Ore Deposits, Petrography, Mineralogy andGeochemistry of the USSR Academy of Sciences.

content, and spinel with relatively low Mg and Cr con-tent.

As a whole, except for a higher anorthite content inthe plagioclase groundmass, these data suggest a moredifferentiated character for plagioclase phyric basalts inrelation to olivine-plagioclase phyric and aphyric basalts.

Other evidence for greater differentiation of the pla-gioclase phyric basalts is iron enrichment. The (FeO*/FeO* + MgO)/ × 100 ratio (FeO* is total Fe as FeO) inthese rocks averages 51.1, whereas in olivine-plagio-clase phyric basalts the mean value of this coefficient is50.8. The main distinction of plagioclase phyric basalts,however, lies in the significantly higher A12O3 content,which exceeds 17%; that is, plagioclase-phyric basaltsare high-alumina rocks (Table 7). The high-alumina

SiO2

Tiθ2A12O3

FeOMnOMgOCaONa 2O

Total

SiTiAlFeMnMgCaNa%En°Io Wo% FsName

Olivine-Plagioclase

Phyric BasaltsSample 487-20-2,

50-56 cm

49.70.833.3

11.50.19

14.720.1

—

100.32

50.70.854.88.60.25

14.720.0

—

99.9

Cations on 6

2.1250.0270.1660.4110.0070.9370.921—

37.937.424.6augite

1.8790.0240.2100.2660.0080.8120.794

—42.641.715.7augite

PlagioclasePhyricBasalts

Sample 487-20-1,113-115 cm

53.41.03.7

11.70.11

13.318.30.7

102.21

Oxygen

1.9440.0270.1590.3560.0030.7210.7140.049

36.538.724.8augite

51.31.03.3

14.50.31

15.113.30.3

99.11

1.9320.0280.1470.4560.0100.8470.5360.022

42.828.229.0augite

character of plagioclase phyric basalts is also evident inthe chemistry of rock-forming minerals. A higher A12O3

content is typical not only for feldspar of plagioclasephyric basalts but also for spinel, titanomagnetite, andeven for olivine.

From a chemical viewpoint, aphyric basalts are verysimilar to olivine-plagioclase phyric basalts, and whatdifferences there are are in keeping with the mineralogi-cal composition of the rocks and their position in crosssection. For example, the lower CaO content in aphyricbasalts as compared with olivine-plagioclase phyric ba-salts reflects the absence of plagioclase phenocrysts en-riched with anorthite in the former. Aphyric basalts arealso distinctive because of their higher MgO and Cr2O3

704

PETROLOGY AND ORIGIN OF BASALTS

Table 4. Olivines in basalts, Hole 487.

SiO2

TiO2

A12O3

FeOMnOMgOCaONiO

Total

SiTiAlFeMnMgCaNi°/o Fo°7o FaName

Olivine-Plagioclase

Phyric BasaltsSample 487-20-2,

50-56 cm

Fringe

40.6——

10.40.12

47.80.460.14

99.52

Core

40.8——

10.40.13

48.20.410.22

100.16

PlagioclasePhyric Basalts

Sample 487-20-1113-115 cm

38.30.111.9

15.70.21

41.81.7—

99.73

Cations on 4 Oxygen

1.003——

0.2150.0021.7610.0120.003

88.511.5

1.002——0.2130.0031.7640.0110.004

88.611.4

chrysolite

0.9730.0020.0570.3330.0051.5810.046

79.021.0

chrysolite

by displacement during drilling. In somewhat general-ized form, the distribution of these varieties of basalt inthe lower part of Hole 487 is as follows:

Table 5. Titanomagnetite in basalts, Hole 487.

PlagioclasePhyric

Olivine-Plagioclase BasaltsPhyric Basalts Sample

Sample 487-20-2, 487-20-1,50-56 cm 113-115 cm

SiO2

TiO2

A12O3

FeOMnOMgOCaOV2O5

Total

1.316.32.4

73.00.432.00.360.44

96.11

1.816.13.2

72.70.431.30.420.47

96.88

2.316.13.3

73.00.511.20.480.42

97.63

9.817.54.0

59.50.511.41.8

n.d.

94.51

9.217.56.1

57.00.60.92.1

n.d.

93.4

Note: The determination of the unit cell dimen-sion of magnetic ore mineral from Sample487-20-1, 113-115 cm, performed by M. T.Dmitrieva, Institute of Geology of Ore De-posits, Petrography, Mineralogy and Geo-chemistry of the USSR Academy of Scien-ces, shows that according to the value of thisparameter, a = 8.405 ± 0.005 Å, the min-eral is almost pure magnetite (a = 8.395-8.40 Å).

content and their lower degree of iron oxidation. Agenetic relation between aphyric, olivine-plagioclasephyric, and plagioclase phyric basalts is revealed by therelationship between their mineral and chemical compo-sition and their position in the cross section of the hole(Fig. 2), though the latter may be significantly distorted

Core 20, Section 1

olivine-plagioclase phyric basaltplagioclase phyric basaltolivine-plagioclase phyric and

aphyric basaltsolivine-plagioclase phyric basaltplagioclase phyric basaltdrilling breccia

Core 20, Section 2

drilling brecciaplagioclase phyric basaltolivine-plagioclase phyric basalt

Depth from the Top ofthe Section (cm)

0-1515-23

23-7272-110

110-122122-150

0-2626-4545-90

The relationship between aphyric, plagioclase phyric,and olivine-plagioclase phyric basalts is clearest in theupper part of Core 20, Section 1, from 0 to 72 cm.

In this series (also bottom to top) one can see (1) anincrease in phenocryst and CaO content, (2) enrichmentof rocks in iron [(FeVFeO* + MgO) × 100], and (3) in-crease in the degree of iron oxidization in basalts [(Fe2O3/Fe2O3 + FeO) × 100)].

It is possible that the variations of mineral andchemical composition are related to the differentiationprocesses inside the lava flow, or sheet. Flotation ofplagioclase crystals and their accumulation in the upperparts of basaltic bodies have been discussed with rela-tion to drilling in the Philippine Sea (Dmitriev et al.,1979).

The increased value of the ferruginity (fractionation)coeffficient [(FeOVFeO* + MgO) × 100] in the upperparts of differentiated basaltic bodies is a well-known

Table 6. Spinels in basalts, Hole 487.

Olivine-Plagioclase PlagioclasePhyric Basalts Phyric Basalts

Sample 487-20-2, Sample 487-20-1,50-56 cm 113-115 cm

SiO2

TiO2

AI2O3FeOMnOMgOCaOCr2O3

Total

SiTiAlFeMnMgCaCrFe/Fe + Mg

0.60.37

30.715.80.14

17.60.27

32.8

98.28

Cations on 12 Oxygen

0.0530.0253.2221.1760.0112.3350.0262.3090.335

5.30.36

34.815.0

—14.41.2

30.2

101.38

0.4430.0233.4301.049—1.7740.1081.9950.372

705

Y.I . DMITRIEV

Table 7. Chemical composition of basalts, Hole 487 (wt.%).

Sample(interval in cm)

Siθ2TiO2

A12O3

Fe2O3FeOMnOMgOCaONa2θK2OH 2 O~H2O +

CO 2

P2O5Li2ORb2OCs2OFClSCr2O3

V2O5

Total

Olivine-Plagioclase Phyric and Aphyric Basalts

487-20-12-9

47.800.86

16.474.823.940.166.97

13.992.210.19—

2.21——

n.d.n.d.n.d.

0.020.140.020.0580.046

99.90

487-20-145-49

48.600.93

16.401.33

7.350.139.23

12.781.910.10_

0.980.150.070.00160.00130.00060.020.030.060.060.04

100.17

487-20-250-56

47.700.86

16.384.354.450.557.54

14.022.120.18

—

1.44—_

n.d.n.d.n.d.

0.020.190.020.0190.02

99.86

487-20-280-84

47.900.88

16.702.247.030.169.44

12.821.870.074_

0.87_

0.090.00130.00070.00040.010.070.060.050.01

100.27

Averageof 7

48.00.88

16.493.195.690.258.30

13.402.030.14—

1.380.040.040.00150.00100.00050.020.110.010.050.03

100.04

Plagioclase Phyric Basalts

487-20-115-19

47.200.89

18.002.756.080.118.43

13.141.870.10_

1.010.390.030.00130.00060.00040.030.040.060.060.05

100.24

487-20-1113-115

48.200.87

17.103.725.200.117.94

12.672.080.100.281.72—

0.040.00570.00040.00060.030.040.02

n.d.n.d.

100.29

Averageof 2

47.70.88

17.63.245.640.118.19

12.911.980.100.141.370.280.040.00350.00050.00050.030.040.040.060.05

100.40

Note: Chemical analyses of basalts in Hole 487 were carried out by Y. V. Dolinina, Institute of Geology of OreDeposits, Petrography, Mineralogy and Geochemistry of the USSR Academy of Sciences.

a Aphyric basalt.

(cm) Sample

-10

0 10 20 30 40 50% 48

flfiflß• G•U

a.QüDO

Kfiflf

§§§;

b.

d.

50 52

FeO*

FeO*+MgO

54%

x100

16 17 18%

AI 2 O 3

12 13 14%

CaO

1.5

N

2.0 2.5%

K2O

0

Fe.

20

,0-,

40

!°3+FeO

60%

xlOO

Figure 2. Variations of basalt composition in cross section in Hole 487. (a = olivine-plagioclase phyric basalts, b = plagioclase phyric basalts, c =aphyric basalts, d = basaltic breccia. Samples: 1 = 487-20-1, 2-9 cm; 2 = 487-20-1, 15-19 cm; 3 = 487-20-1, 29-32 cm; 4 = 487-20-1, 45-49cm; 5 = 487-20-1, 73-78 cm; 6 = 487-20-1, 113-115 cm; 7 = 487-20-2, 34-36 cm; 8 - 487-20-2, 50-56 cm; 9 = 487-20-2, 62-64 cm; 10 =487-20-2, 80-84 cm.)

706

PETROLOGY AND ORIGIN OF BASALTS

fact, as is the enrichment of the same parts of basalticsheets and sills by volatile components. With regard tothe latter, the degree of iron oxidation increases begin-ning at the horizon of aphyric basalts and there areabout 0.6% vesicles (former gas bubbles) in olivine-pla-gioclase phyric basalt in the top of the cross section. Inthe lower part of the cross section, plagioclase phyric,olivine-plagioclase phyric, and aphyric basalts containno such vesicles.

The decrease in quantity of plagioclase phenocrystsand the absence of olivine phenocrysts in plagioclasephyric basalts in the upper part of the presumed basalticbody may be related to the dissolution of phenocrystsdue to their disequilibrium with the more differentiatedmelt of this zone.

On the basis of these data we may conclude thataphyric, olivine-plagioclase phyric, and plagioclase phy-ric basalts are derivatives of a common tholeiitic melt,differentiated in the body of basaltic flow or sheet aftereruption on the seafloor.

The interpretation of the sequence aphyric and oliv-ine-plagioclase phyric basalts, olivine-plagioclase phy-ric basalts, and plagioclase phyric basalts as a series re-flecting differentiation inside the basaltic flow or sheetis contradicted by the thickness (less than 1 m) of thepresumed basaltic body, which should preclude signifi-cant differentiation.

In the lower part of the cross section one can also seethe sequence (from top to bottom) olivine-plagioclasephyric basalts, plagioclase phyric basalts, and olivine-plagioclase phyric basalts (without aphyric basalts), butthe regular variations in rock composition observed inthe upper part of the cross section are absent. It is possi-ble that the original relationship between rocks in thelower part of the cross section was disturbed duringdrilling, when a rather thick (54 cm) layer of basalt thathad disintegrated during drilling (Samples 487-20-1,122-150 cm and 487-20-2, 0-26 cm) was formed.

Thus the data seems to indicate that basalts in Hole487 comprise at least two lava flows or sheets.

OTHER EASTERN PACIFICOCEAN FLOOR BASALTS

To explain the genesis of basalts exposed in Hole 487,on the seaward side of the Middle America Trench, letus compare data about their composition with similarmaterials about basalts of other morphostructures in theEastern Pacific.

Let us begin with a comparison between rock associa-tions in the Middle America Trench, which consists ofnormal and high-alumina basalts, and those of tholeiiticbasalts of different morphostructures in the Pacific; Iomit consideration of basaltic series in ocean islandswith an alkaline affinity.

In the Galapagos spreading center, according to mymaterial on Holes 424 and 425 of Leg 54 (Dmitriev, inpress) and the data of other investigators (Anderson etal., 1975; Rudnik, 1976; Yeats et al., 1973), the follow-ing rocks in the tholeiitic series occur: high-alumina ba-salts, normal basalts, Fe-basalts. There are also alkalic

basalts on seamounts in the Galapagos zone of spread-ing.

In the Siqueiros transform fracture zone, accordingto my study of samples from Hole 427, Leg 54 (Dmitriev,in press) and to other data (Batiza et al., 1977), there arehigh-alumina, normal, and Fe-basalts of tholeiitic seriesas well as alkalic basalts.

On the flanks of the East Pacific Rise, according tomy investigation (Dmitriev, in press) of Holes 419, 420,422, and 428, Leg 54, and the results of Leg 16 (Yeats etal., 1973), only normal and Fe-basalts of tholeiitic seriesoccur.

All these findings indicate that no direct correspond-ence exists between associations of rocks in Hole 487and the basaltic associations of other morphostructuresof adjacent regions of the Pacific. In zones of spreadingand zones of transform fractures, basaltic associationsare more diverse then in Hole 487, for they comprisealong with normal and high-alumina basalts, tholeiiticFe-basalts and alkalic basalts as well. On the flanks ofthe East Pacific Rise, on the other hand, high-aluminabasalts were not observed; instead normal tholeiiticbasalts associate with Fe-basalts.

The peculiarity of the associations of normal andhigh-alumina basalts in Hole 487 becomes more obviouswhen we compare their average chemical compositionwith corresponding data on normal and high-aluminabasalts of the Galapagos spreading zone, Siqueiros trans-form fracture zone, and the western flank of the EastPacific Rise (Table 8).

There is a distinct difference between normal basaltsof the Middle America Trench and other morphostruc-tures of the East Pacific. As it is clear from Table 8,normal basalts of the Middle America Trench have sig-nificantly lower SiO2, TiO2, and Na2O + K2O content;a minimum value of (FeOVFeO* + MgO) × 100; andhigher concentrations of A12O3 and CaO than normaltholeiitic basalts from other morphostructures.

Table 8. Average composition of normal and high-alumina tholeiiticbasalts of the Middle America Trench, Hole 487, and other mor-phostructures of the East Pacific.

Components

Tiθ2A12O3

F e 2 O 3

FeOMnOMgOCaONa 2 OK2OH2O

P2O5

TotalFeO*/(FeO* + MgO) × 100

1

49.781.31

15.779.70

0.167.33

12.332.620.19

0.12

99.3157.0

Normal Basalts

2

49.511.30

14.832.177.550.178.69

12.253.000.12

0.16

99.7552.1

3

49.521.49

14.7610.39

7.7111.812.460.191.240.15

99.7260.0

4

48.00.88

16.493.195.690.258.30

13.302.030.141.380.04

99.7950.8

High-Alumina Basalts

5

47.670.65

18.442.085.830.139.61

12.642.120.110.320.07

99.6344.6

6

49.200.88

17.71.096.750.148.84

12.572.470.05

0.10

99.7946.75

7

47.70.88

17.63.245.640.118.19

12.911.980.101.37

51.1

Note: 1, 5—Galapagos spreading zone: 1—average of 19 analyses; 5—average of 4analyses (on the basis of analyses from Anderson et al., 1975; Dmitriev, in press,analyst, Y. V. Dolinina; Rudnik, 1976; Yeats et al., 1973). 2, 6—Siqueiros Zone oftransform fracture: 2—average of 17 analyses: 6—average of 2 analyses (on the basisof analyses from Batiza et al., 1977). 3—Western flank of the East Pacific Rise(9° - l l °N, 1O5°-15O°W): average of 16 analyses (on the basis of analyses fromDmitriev, in press, analyst, Y. V. Dolinina: Yeats et al., 1973). 4, 7—Middle AmericaTrench, Hole 487: 4—average of 4 analyses; 7—average of 2 analyses.

707

Y. I. DMITRIEV

The difference between high-alumina trench basaltsand basalts of the same type from other ocean morpho-structures is less distinct. High-alumina basalts of theMiddle America Trench, like the normal basalts of thismorphostructure, have a higher CaO and a lower Na2O+ K2O content, than high-alumina basalts of spreadingzone, transform fault or flanks of the mid-oceanicridge, but the main difference is a significantly higher(FeOVFeO* + MgO) × 100 value in trench basalts.

All these data on the chemistry of Middle AmericaTrench basalts support the conclusion that these rocksare not entirely identical to rocks of the same kind inspreading zones, transform faults, and on the flank ofthe mid-oceanic ridge. Thus the question arises: Is itpossible that Middle America Trench basalts originatein the formation of trench—that is, that they eruptedlocally rather than being brought into the trench zonemechanically by the movement of the lithosphericplates?

If this is the case, we expect that the composition oftrench basalts will possess characteristics related to thespecific geodynamic conditions of magmatism in thesemorphostructures which distinguish areas of oceanictholeiitic magmatism from zones of island arc calc-alkaline magmatism.

Let us examine the position of Middle AmericaTrench basalts in Hole 487 with regard to the chemicalcomposition of oceanic tholeiitic basalts, including ba-salts of other deep sea trenches and island arcs of thePacific. On the TiO2 /(FeOVFeO* + MgO) × 100diagram (Fig. 3) one can see that basalts of the MiddleAmerica Trench, the Mariana Trench, and the YapTrench occupy a position between basalts of oceanicspreading zones, zones of transform faults, and slopesof the East Pacific Rise on the one hand and basalts ofisland arcs on the other hand. It is also obvious fromthis diagram that whereas high-alumina basalts of zonesof spreading and transform faults belong to the mostprimitive rocks of these morphostructures, high-alu-mina basalts of deep sea trenches, according to the valueof (FeOVFeO* + MgO) × 100, are more differen-tiated members of the tholeiitic series.

Deep sea trench basalts occupy the same intermediateposition on the A12O3 /(FeOVFeO* + MgO) × 100diagram (Fig. 4), but whereas on the previous diagramthe distinction of trench basalts is the lower TiO2 con-tent as compared with spreading zone basalts of asimilar stage of differentiation, transform fault zones,flanks of the East Pacific Rise, in Figure 4 the distin-guishing feature of trench basalts is a higher A12O3 con-centration.

CONCLUSIONS

By plotting the chemical analyses of basalts belong-ing to different morphostructures on the MgO/FeOVNa2O + K2O diagram (Fig. 5) one can obtain a generalnotion about the place of normal and high-aluminabasalts in deep sea trenches in the basic processes ofoceanic tholeiitic magma evolution.

Evidently, the tholeiitic basalts of all Pacific morpho-structures form a single trend of evolution in the direc-

tion of iron enrichment, which is typical of the tholeiiticbasaltic magma in general.

Normal and high-alumina basalts of the MiddleAmerica Trench, as well as of the Yap and Marianatrenches, occupy a transitional position between normaltholeiitic basalts and the point at which iron-rich dif-ferentiates (Fe-basalts) begin to dominate.

Points reflecting the chemistry of trench basalts arein the right portion of the diagram, which is evidence ofan alkali deficit in these rocks as compared with basaltsof other oceanic morphostructures.

It is necessary to note that the high-alumina basaltsof island arcs even more advanced in the evolutionaryseries of oceanic tholeiitic magma than high-aluminabasalts of the deep sea trenches. This fact confirms thetransitional position of tholeiitic trench basalts betweenbasalts of typical oceanic morphostructures and islandarc basalts.

The higher A12O3 content in deep sea trench basaltsand in island arc basalts cannot be the result of the com-mon differentiation of tholeiitic basaltic magma, be-cause in the process of differentiation A12O3 content de-creases and, accordingly, on the MgO/FeOVNa2O +K2O diagram the high-alumina basalts of spreadingzones and zones of transform faults are in associationwith the least differentiated rocks with the highest MgOcontent.

Thus parental melts of deep sea trench basalts, and inan even larger degree island arcs, evidently are deriva-tives of oceanic tholeiitic magma, which evolved underconditions resulting either in fractionation and accumu-lation of Al2O3-rich phases in some parts of the magmachamber or in assimilation by magma of the aluminouscrust material.

The transitional (to judge by the number of features)chemical composition of normal and high-alumina ba-salts in Hole 487 from typical ocean basalts to island arcbasalts indicates, from my point of view that the basaltsin Hole 487 were formed in the transitional zone be-tween the ocean and the island arc volcanic belt of theAmerican Cordilleras—that is, in the Middle AmericaTrench.

REFERENCESAnderson, R. N., Claque, D. A., Klitgord, K. D., et al., 1975. Mag-

netic and petrologic variation along the Galapagos spreading cen-ter and their relation to the Galapagos melting anomaly. Geol.Soc. Am. Bull., 86:683-694.

Batiza, K., Rosendahl, B. R., and Fisher, R. L., 1977. Evolution ofoceanic crust Part III. Petrology and chemistry of basalts from theEast Pacific Rise and Siqueiros fracture zone. /. Geophys. Res.,82:265-276.

Dickinson, W. R., 1970. Relations of andesites, granites and deriva-tive sandstones in arc-trench tectonics. Rev. Geophys. SpacePhys., 8:813-860.

Dmitriev, Y. I., in press. Basalts from the East Pacific Rise near 9°Ndrilled on DSDP Leg 54 compared with marginal-basin and ocean-island basalts. In Hekinian, R., Rosendahl, B., et al., Init. Repts.DSDP, 54: Washington (U.S. Govt. Printing Office), 695-304.

Dmitriev, Y. I., Solovova, I. P., Dolinina, Y. V., et al., 1979. Petro-logia basitov Philippinskogo morya po dannim glubokovodnogoburenia. Iz. Akad. Nauk. SSSR Ser. Geol., N4:19-33.

Ewart, A., and Bryan, W. B., 1972. Petrography and geochemistryof the igneous rocks from Eua, Tongan Islands. Geol. Soc. Am.Bull., 83:3281-3298.

708

PETROLOGY AND ORIGIN OF BASALTS

Fitton, J. G., 1971. The generation of magma in island arcs. EarthPlanet. Sci. Lett., 11:63-67.

Green, T. H., and Ringwood, A. E., 1968. Genesis of the calc-alka-line igneous rock suite. Contrib. Mineral. Petrol., 18:105-162.

Hey, R., 1977. Tectonic evolution of the Cocos-Nazca spreading cen-ter. Geol. Soc. Am. Bull., 88:1404-1420.

Hey, R., Johnson, G. L., and Lowrie, A., 1977. Recent plate motionsin the Galapagos area. Geol. Soc. Am. Bull., 88:1385-1403.

Larson, E. E., Reynolds, R. L., Merrill, R., et al., 1974. Major-ele-ment petrochemistry of some extrusive rocks from the volcanicallyactive Mariana Island. Bull. Volcanol., 38:361-377.

Leonova, L. L., Popolitov, E. I., Volinets, O. N., et al., 1978. Tipichetvertichnich basaltov Kamchatki v svyasi s problemoi pervich-nich magm. Petrologicheskie issledovaniya basitov ostrovnichdug: Moscow (Institute of the Earth's Physics), pp. 157-176.

Lopez-Escobar, L., Frey, F. S., and Vergaza, M., 1977. Andesites andhigh-alumina basalts from the central south Chile High Andes:Geochemical evidence bearing on their petrogenesis. Contrib.Mineral. Petrol., 63(No. 3):203.

Moore, J. C, Watkins, J. S., Shipley, T. H., et al., 1979. Progressiveaccretion in the Middle America Trench, Southern Mexico.Nature, 281:638-642.

Pichler, H., and Weyl, R., 1973. Petrochemical aspects of CentralAmerican magmatism. Geol. Rundsch., 62(2):357-396.

Rudnik, G. G., 1976. Magmaticheskie i metamorphicheskie porodivpadini Chessa. Geologo-Geophisicheskie issledovaniya v yugo-vostochnoi chasti Tichogo okeana: Moscow (Nauka), pp.116-125.

Stark, J. T., 1963. Petrology of the volcanic rocks of Guam. Geol.Sur. Prof. Paper 403-C.

Stoiber, R. E., and Carr, M. J., 1974. Quaternary volcanic and tec-tonic segmentation of Central America. Bull. Volcanol., 37(No.3):304-325.

Yeats, R. S., Forbes, W. C, Heath, G. R., et al., 1973. Petrology andgeochemistry of DSDP Leg 16 basalts, Eastern Equatorial Pacific.In van Andel, Tj. H., Heath, G. R., et al., Init. Repts. DSDP, 16:Washington (U.S. Govt. Printing Office), 617-640.

1.5-

.0-

0.5-

_

O161

OCD1

X

SD8o

Φ D536

• 5 3 6 •536• 536D536

• 536

l

O428A

O422O422

DD2A

.__.. OQBROD17 OSD4 v

OSD4 <>CD C 3 2

OSD4DD7,-,„,. DD7ADD5 D532

O428AOSD5

D530D530

O428A

OSD6 R T § π T 6 D O 161AD 1 5 8 D158 D T 6 163 O

H536

C D 1 π 5 3 6 Q 1 6 3 D 4 2 5V DD3B O163 O 1 6 3 D D 3 A

+ M17OSD3 SD3 OQBR D +Y17 UD7 XM1

Λ A. n7 π<=>̂nO SD3O / " > .526 1+487

v —'

×M2

XCH1 DD7

×M1

XTg1

162O

D424

XM1

-

i

40 45a: Db:•

O,50

O 4FeO*

55

x100

60 65

FeO +MgO

Figure 3. TiO2 - (FeO*/FeO* + MgO) × 100 in Pacific basalts of various morphostructures. (1 = zones of spreading, 2 = transform faults, 3 =flanks of the East Pacific Rise, 4 = deep sea trenches, 5 = island arcs, 6 = basalts in Hole 487. a (upper line) = normal tholeiitic basalts, b(lower line) = high-alumina tholeiitic basalts. Analyses [some of the following symbols appear only in Figs. 4 and 5]: 1 = zones of spreading,Galapagos spreading zone: D2A, D3B, D5, D7, D7A, D17, T6 [Anderson et al., 1975], 424, 425 [Dmitriev, in press, analyst, Y. V. Dolinina],530, 532, 536 [Rudnik, 1976], 158 [Yeats et al., 1973]. 2 = transform faults, Siqueiros Zone fault: CD1, Dl, D3, D4, D5, D6, D8 [Batiza et al.,1977]. 3 = flanks of the East Pacific Rise: 422, 428A [Dmitriev, in press, analyst, Y. V. Dolinina], 161, 162, 163 [Yeats, et al., 1973]. 4 = deepsea trenches: M17—Mariana Trench, Y17—Yap Trench [Dmitriev, Leg 17 of Dmitriy Mendeleev, analyst, Y. V. Dolinina]. 5 = island arcs:Kl—Kamchatka [Leonova et al., 1978] Ml (Larson et al., 1974], M2 [Stark, 1963]—Mariana Arc Tgl (Ewart and Bryan, 1972)—Tonga [LopezEscobar et al., 1977] CHI—Chilian Cordilera.

709

Y. I. DMITRIEV

21

20

19

18

CO

16

15

14

13

O161

1536×Tg1

X C M X M 1

• 536

• 5361536

XM2•536

• ( + 487* S D 3 ^ ' + Y 1 ?

Y17

+M17

1536

D536

OCD1

OSD3 \ ™ 532a" +M17D7D-. OD3B

^ DD5 QD7 D536G532

OD17 ^cr.,. O428A\cπo VθL)4 O1 63

OQBR<>001

D <JQBR O S D 6 Q 1 6 3

OSD4 yj|| ° O163 ^OSD3 <>D158 S

OSD5

×K1

DD7

M1D T 6

×DD3A T 6 O163 ODDT6 162

DT6D530-ID530

O428A161A

424DO

^ O 4O157 D17

D424

40 45 50

1 o, 23 +4FeO

55

x100

60 65

\^y FeO +MgO

Figure 4. A12O3 + FeOVFeO* + MgO × 100 in Pacific basalts of various morphostructures. (Legend and analyses same as in Fig. 3.)

FeO FeO

O160

O159

CDV

D536 QBR •536 5 3 ^

4030

MgO

80 70 60 50

Na2O+K2O J6

Figure 5. MgO/FeO*/Na2O + K2O in Pacific basalts of various morphostructures. (Legend and analyses same as in Fig. 3.)

710