22. GEOCHEMISTRY, MINERALOGY, AND <M>K-̂ Ar RADIOMETRIC DATING OF LEG 84BASALTS—GUATEMALA TRENCH1

H. Bellon, and R. C. Maury, Université de Bretagne Occidentale,and

J. L. Joron, J. Bourgois, and J. Aubouin, Université Pierre et Marie Curie2

ABSTRACT

Geochemical (atomic absorption, neutron activation analyses), mineralogical (microprobe), and radiometric C40!̂ -40Ar) data are presented for five basalts from the Guatemala Trench area (Deep Sea Drilling Project, Leg 84).

Strong geochemical and mineralogical differences distinguish two types among these basalts: (1) One basalt (Sample567A-19.CC), recovered below Upper Cretaceous limestone has the following characteristics: it is quartz normative andhas low TiO, content, as well as low amounts of Cr, Ni and other transition metals, an LREE depleted pattern, and af-finities of clinopyroxene phenocryst plotted into the field of tholeiitic and calc-alkalic pyroxenes. (2) Four alkaline ba-salts, recovered from the mafic and ultramafic acoustic basement, are nepheline normative and show high TiO2 content,high amounts of Cr, Ni and so on, an LREE enriched pattern and compositions of clinopyroxene phenocryst plottedclose to or within the field of alkali basalt pyroxenes. These basalts are comparable to those recognized in the lower partof the Santa Elena complex and are clearly different from the oceanic basalts of the Cocos Plate.

The radiometric age of the orogenic basalt seems to be close to 80 Ma. The alkaline basalts are clearly older, even if adiscrepancy appears between the results of different analyses because of the secondary effects of alteration.

INTRODUCTION

Hole 567A was drilled only 100m from Hole 494A(DSDP Leg 67, Fig. 1). From the top to the bottom, thefollowing succession was encountered: (1) Pleistoceneand Pliocene slope sediments; (2) Miocene sediments;(3) Upper Cretaceous (upper Campanian - lower Maes-trichtian) limestones; and (4) an acoustic basement ofmafic and ultramafic rocks (Bourgois et al., this vol-ume). A few pieces of basalt were recovered from thishole; five of these have been studied with one majorgoal—to determine their possible affinities with the Leg67 basalts (Maury et al., 1982) by identifying their mag-matic parentages and their ^K-^Ar ages.

LITHOLOGY

Sample 567A-19,CC is a core-catcher sample recov-ered below Upper Cretaceous limestones; its stratigraphicposition is similar to that of the andesitic samples recov-ered from the bottom of Hole 494A, DSDP Leg 67. Thisporphyritic basalt contains millimetric phenocryst of al-tered olivine as well as fresh calcic clinopyroxene in amicrolitic groundmass consisting of Plagioclase, clino-pyroxene, and titanomagnetite; the residual glass hasbeen converted to chlorite and calcite.

The few other samples (567A-25-2, 90-96 cm; 567 A-25-3, 74-78 cm; 567A-26-1 (Piece 1, 3-4 cm) and 567 A-29-2, 147-149 cm represent basaltic blocks that are con-

von Huene, R., Aubouin, J., et al., Init. Repts. DSDP, 84: Washington (U.S. Govt.Printing Office).

2 Addresses: (Bellon and Maury) Université de Bretagne Occidentale, GIS "Océanologieet Géodynamique" No. 410012, Avenue Le Gorgeu, 29283 Brest Cedex, France; (Joron) Groupedes Sciences de la Terre, Laboratoire de Pierre Sue, CEN Saclay BP2, 91190 Gif-sur-Yvette,France; (Bourgois and Aubouin) Department de Géotectonique, Université Pierre et Marie Curie,4, Place Jussieu, 75230 Paris, France.

sidered to be tectonically emplaced in the mafic and ul-tramafic plutonic basement (Bourgois et al., this vol-ume). Sample 567A-26-1 (Piece 1, 3-4 cm) shows a pro-gressive transition from a texture made of skeletal mi-crolites of Plagioclase and clinopyroxene to one com-posed of a fine-grained doleritic texture. This observa-tion suggests that this block could be interpreted to be apillow fragment. This sample probably represents theouter boundary (quenched facies) and the inner un-quenched boundary of a pillow. Sample 567A-29-2, 147-149 cm shows a quenched texture. Finally, Samples567A-25-2, 90-96 cm and 567A-25-3, 74-78 cm arecoarse-grained dolerites.

These four samples contain phenocrysts of altered ol-ivine, fresh clinopyroxene and Plagioclase in a ground-mass consisting of Plagioclase, clinopyroxene, titanomag-netite, and hemoilmenite. Alteration minerals (see ra-diometric section) are also present, consisting of abun-dant chlorite and calcite in all samples analyzed as wellas celadonite and albite in the groundmass of Sample567A-26-1 (Piece 1, 3-4 cm) (Table 5).

CHEMISTRY

Major Elements

Sample 567A-19,CC is a quartz normative (1%) ba-salt with low TiO2 (0.64 wt. %) and high A12O3 (16.41wt. %) contents (Table 1). These results are similar tothose obtained for the stratigraphically equivalent ba-salts and andesites from Hole 494A (Maury et al., 1982;Table 1). Chemically, the four other samples seem verysimilar and contrast strongly with Sample 567A-19,CC.They have a low SiO2 content (42.3 to 46.1 wt. %) and ahigh TiO2 content (2.3 to 2.6 wt. %). Three are nephe-line normative basalts; the fourth is silica oversaturated

655

H. BELLON ET AL.

14°0091 00' 90° 00'

13° 00

GUATEMALA

Middle AmericaTrench Axis

Figure 1. Location of Legs 84 and 67 sites off Guatemala. (Bathymetry is in meters.)

but shows normative corundum and a high water con-tent, indicating its high level of alteration.

Trace Elements

Trace element composition of Sample 567A-19,CC issignificantly different from that of the four other sam-ples (Table 2). The former contains low amounts of Cr,Ni, Sr, Zr, Ba, and hygromagmaphile elements. In par-ticular, its ratio Ti/V 20 is close to that in orogeniclavas, according to Shervais (1982). Therefore, its traceelement chemistry is nearly identical to that of Hole494A samples previously studied by Maury et al. (1982);as these authors suggested for the Hole 494 A samples,an orogenic parentage of the basalt of Sample 567 A-19,CC seems to be indicated. Its extended Coryell-Ma-suda plot of REE (rare earth elements) and other hygro-magmaphile elements (Fig. 2) shows the typical LREE(light rare earth elements) depleted pattern, similar tothat of Hole 494A samples.

High amounts of Cr ( = 200 ppm), Ni ( 100 ppm),Sr (242-425 ppm), Ba (113-265 ppm), and hygromag-

maphile elements characterize the four other basalts(Table 2). Their Ti/V ratios range from 57 to 102; sohigher than 50, these ratios are symptomatic of an alka-line affinity (Shervais, 1982). The Coryell-Masuda pat-terns of two samples (Fig. 2) are typically LREE en-riched. These results indicate that these rocks are clearlydifferent from typical oceanic tholeiites from the CocosPlate (Sites 495, 499, 500, DSDP Leg 67) as well as fromHole 494A samples.

The very high contents of Rb (43 ppm) and Cs (2.7ppm) in Sample 567A-29-2, 147-149 cm (Table 2) areprobably linked to secondary alteration processes.

MINERALOGY

Four samples (567A-19,CC; 567A-25-2, 90-96 cm;567A-25-3, 74-78 cm, and 567A-26-1 (Piece 1, 3-4 cm)were studied with a Camebax-type automated microprobe(Microsonde Ouest, Brest—working conditions: 15 kV,10 nA; counting time 6 s; oxide concentrations lowerthan 0.3% were not considered as representative).

656

GEOCHEMISTRY, MINEROLOGY, AND ^K-^Ar RADIOMETRIC DATING

Table 1. Major elements analysis (wt. <%) and CIPW norms of Hole 567A basalts compared with an orogenic basalt fromHole 494A.

Element

Siθ2Tiθ2AI2O3Fe2θ3MnOMgOCaONa2θK2OP2O5

567A-19.CC

50.800.64

16.418.490.136.797.753.210.510.00

Loss on ignition1050°C

H2O-

Total

QorabanneCO

C wodi \ en

l f se n ^fsfofamtilap

S.I.D.I.

4.22

0.64

99.59

1.023.21

28.8830.70

4.252.531.50

15.459.19

1.961.290.00

37.1433.11

494A-29.CC (0-4)

51.880.35

14.608.050.148.688.432.980.470.10

3.32

0.79

99.79

. Sample(hole-core-section, interval in cm)

567A-25-2, 90-96

42.852.62

14.8011.810.177.749.973.130.470.14

4.68

2.15

100.53

2.9919.4126.914.95

10.666.403.69

10.086.412.775.370.36

34.9627.36

567A-25-3, 74-78

42.352.59

14.9511.770.187.018.163.490.430.14

5.14

3.02

99.23

2.8226.0826.48

3.62

7.324.262.72

10.597.452.845.460.37

32.3032.52

567A-26-1 (Piece 1, 3-4)

44.202.57

14.0211.300.206.569.853.350.570.35

2.71

3.89

99.57

3.6625.4723.39

2.87

11.456.634.30

7.795.572.675.300.90

31.4632.00

567A-29-2, 147-149

46.102.29

17.2511.840.112.075.063.291.690.51

5.76

3.46

99.43

2.9211.1831.1624.75

2.14

5.7712.98

2.884.871.35

11.6245.26

Note: Fe2θ3* = total iron as Fe2θ3. The CIPW norms are calculated with 85% total iron as FeO and 15% as Fe2θ3. D.I. = differentiationindex; S.I. = solidification index. Analyses were done by J. Cotten, Brest. Blank spaces in CIPW norms mean absence of normative min-erals or of calculated CIPW norm (Sample 494A-29.CC [0-4 cm]).

Table 2. Trace element analyses of Hole 567A samples (ppm).

Sample(interval in cm)

567A-19.CC567A-25-2, 90-96567A-25-3, 74-78567A-26-1 (# 3-4)567A-29-2, 147-149

Li

AA

132320

29

Ti

AA

384015,71015,53015,41013,730

V

AA

198253248268135

Cr

AA

46198182193209

Mn

NA

18302395253528201550

Fe

AA

10,632514,790014,740014,151514,8280

Co

AA

3540353052

NA

314138

54

Ni

AA

3394799585

NA

46I B100

97

Zn

AA

59928890

150

AA

442

1143

Rb

NA

2.53.41.8

47.3

Sr

AA

91355425289242

Zr

NA

40176181

152

Sb

NA

0.040.060.07

1.06

Cs

NA

0.090.030.04

2.70

B

AA

83144217113265

a

NA

74163191

268

La

NΛ

0.8616.216.1

16.2

Ce

NA

2.635.628.2

29.7

Eu

NA

0.471.802.01

1.78

Tb

NA

0.290.860.85

0.83

Hf

NA

0.934.103.90

3.46

Ta

NA

0.0171.741.82

1.44

Th

NA

0.091.621.70

1.42

U

NA

0.040.400.39

0.38

Note: AA = atomic absorption analysis (J. Cotten, Brest); NA = neutron activation analysis (J. L. Toron, Saclay). Blank spaces indicate no data available.

Pyroxenes

All the analyzed pyroxenes are Ca-rich clinopyrox-enes (Table 3). Most of the phenocryst from Sample567A-19,CC plot into the endiopside field (Fig. 3) andshow high Mg/Fe ratios, high SiO2, and low A12O3 andTiO2 contents (Table 3). The corresponding groundmasscrystals restricted the iron-enrichment trend. In the dis-crimination diagram Ti/(Ca + Na) proposed by Leter-rier et al. (1982) (Fig. 4A), the phenocryst plot into thefield of tholeiitic and calc-alkalic pyroxenes. In the (Ti4- Cr)/Ca diagram (Fig. 4B) of Leterrier et al. (1982),the clinopyroxenes clearly plot in the field of orogenicbasalts, well outside the area of the field that overlaps

with the field of nonorogenic tholeiites. All these resultsare very similar to those obtained from Hole 494A py-roxenes (Maury et al., 1982).

The clinopyroxenes from the three other samples areclearly different: they are richer in iron relative to 567 A-19,CC clinopyroxenes and plot near the augite/saliteboundary or even in the salite field; their A12O3 andTiO2 contents are high. The phenocrysts show only re-stricted zoning patterns, and their rims and the ground-mass crystals are slightly iron-enriched in comparisonwith their cores. In the discrimination diagram Ti/(Ca+ Na) (Fig. 4A), the clinopyroxene phenocrysts plotwithin or near the field of pyroxenes from alkali basalts.The pyroxenes from the alkali basalts from the lower

657

H. BELLON ET AL.

Table 3. Representative clinopyroxene microprobe analyses (wt.%).

Sample

Siθ2Tiθ2AI2O3Cr2θ3FeO*MnOMgOCaONa2θ

Total

fcaMg

(̂ Fe + Mn

C ( l )

53.340.101.480.084.470.17

19.8820.15

0.14

99.81

39.1853.78

7.04

C(2)

53.320.122.750.204.790.11

18.5320.38

0.13

100.83

40.7751.57

7.65

C(3)

53.240.172.280.224.910.08

18.7820.36

0.11

100.15

40.4151.86

7.73

Sample(hole-core-section, cm interval)

567A-19.CC

C(4)

52.470.183.640.105.950.18

18.4219.180.09

100.21

38.6851.67

9.65

C(5)

51.420.293.830.196.270.27

18.3718.590.14

99.37

37.7551.8910.36

G(6)

51.880.272.380.008.090.13

16.7519.740.12

99.36

39.9247.1012.98

G(7)

50.360.423.050.019.960.21

15.3819.730.14

99.26

40.2143.6116.18

G(8)

49.190.592.880.00

15.480.30

12.4418.060.16

99.10

37.8736.3025.83

C(9)

49.291.954.100.059.680.10

14.4620.46

0.54

100.63

42.4341.7215.85

567A-25-2,

C (10) 1

49.541.913.550.009.740.05

14.2519.990.33

99.36

42.1341.7716.10

90-96

G ( l l )

50.521.322.520.00

10.650.32

14.6119.610.35

99.90

40.4341.9117.66

G(12)

52.010.611.010.00

13.190.71

15.2517.050.81

100.14

34.7143.2022.09

Note: FeO = total iron as FeO; positions are indicated by C = core of a phenocryst, R = rim of a phenocryst, and G = ground-mass crystal; number in parenthesis = analysis number.

100-

50-

10-

0.5-

ThLa

<>o

Ce

Ta

Hf EuZr Ti

O 567A-19.CC

• 567A-25-2, 90-96 cm

• 567A-29-2, 147-149 cm

Tb

Figure 2. Extended Coryell-Masuda plots for selected samples fromHole 567A.

structural unit of the Santa Elena complex of Costa Ri-ca (Azema and Tournon, 1980) have a similar locationin this diagram (Girard, 1981; Maury et al., 1982). Ithas been suggested that the Santa Elena complex is theonshore equivalent of the Hole 567A sequence (Aubouinet al., this volume; Bourgois et al., this volume).

Plagioclases

In Sample 567A-19,CC, groundmass Plagioclase com-position ranges from An7 2 to An5 0. As previously shownfor Hole 494A samples (table 5 in Maury et al., 1982),the groundmass plagioclases do not contain significantamounts of K2O and MgO and their composition differsfrom that of plagioclases from alkali basalts and ocean-ic tholeiites (Table 4).

In the three other basalts, the plagioclases (labradorto oligoclase) contain significant amounts of orthoclase(up to 10%). These characteristics are commonly foundin plagioclases from alkali basalts. The phenocrysts range

in composition from An 6 9 to An 1 2 and show a normalzoning pattern with a progressive decrease of anorthiteaccompanied by an increase of orthoclase componenttoward their rims (Table 4 — Analyses 27 to 31 show thetransition from andesine to potassic oligoclase). Thegroundmass Plagioclase ranges in composition fromAn6 7 to An3 6.

Iron-Titanium Oxides

Chromiferous spinels (Analysis 35, Table 5) occur spo-radically in Samples 567A-25-2, 90-96 cm; 567A-25-3,74-78 cm; and 567A-26-1, (Piece 1, 3-4 cm). In theAl2θ3/Cr2θ3 diagram of Barbot (1983), these chromif-erous spinels plot within the field of chromiferous spi-nels from alkaline basalts. These spinels are depleted inCr and Al relative to spinels from oceanic basalts (Si-gurdsson and Schilling, 1976). The hemoilmenites haveretained much of their primary composition (Table 5,Analyses 38 to 40). On the other hand, the titanomagne-tites have suffered various degrees of secondary oxida-tion and the sum of their oxides is always below 10%,when it is recalculated according to Carmichael's (1967)method. Despite these results, the compositions of coex-isting titanomagnetites and hemoilmenites have been tenta-tively equilibrated following Powell and Powell's (1977)formulation of Buddington and Lindsley's (1964) geo-thermometer: all the available analyses from a givenrock have been treated using a systematic combinationas described by D'Arco et al. (1981). Temperatures rangefrom 1200 to 900°C, with a maximum frequencyaround 1100 to 1000°C (Fig. 5); the corresponding oxy-gen fugacities values lie between the nickel-nickel oxide(NNO) and the fayalite-magnetite-quartz (FMQ) buffers.

40K-40Ar RADIOMETRIC DATA

EXPERIMENTAL PROCEDURE

Four samples—567A-19,CC; 567A-25-2, 90-96 cm; 567A-25-3, 74-

78 cm; and 567A-29-2, 147-149 cm—were dated by the conventional

^K-^Ar method. Bulk rock samples were crushed into fragments of

0.15 to 0.30 mm in size and cleaned with distilled water to eliminate

658

GEOCHEMISTRY, MINEROLOGY, AND ^K-^Ar RADIOMETRIC DATING

Table 3. (Continued).

C(13)

47.032.995.850.019.760.13

12.9420.65

0.43

99.79

44.5338.8316.64

567A-25-3 74-78

C(14)

46.603.365.590.00

11.980.18

11.1720.11

0.52

99.51

44.5534.4121.04

G(15)

48.252.695.120.009.600.33

13.2220.32

0.56

100.09

43.7439.5816.68

G(16)

47.302.354.290.00

13.680.42

11.6118.590.48

99.75

40.6235.3124.07

Sample(hole-core-section, cm interval)

C(17)

47.322.565.490.108.810.00

13.7621.30

0.48

99.80

45.0340.4414.53

R(18)

43.142.604.740.02

10.020.16

12.8021.32

0.35

99.33

45.2937.8416.87

567A-26-1

C(19)

47.414.347.330.05

11.490.10

11.0821.04

0.59

99.16

46.2333.8819.89

R(20)

44.511.984.250.00

15.490.429.49

20.230.56

99.83

44.1228.7827.10

(Piece 1, 3-4)

C(21)

45.304.536.390.13

11.610.19

10.8121.21

0.69

100.07

46.6633.0720.27

G(22)

45.303.745.960.04

11.900.18

10.7321.30

0.45

99.60

46.6632.6920.65

G(23)

46.513.104.760.05

12.470.22

11.1121.29

0.60

100.11

45.6333.1321.24

G(24)

45.034.145.740.00

13.160.189.97

20.600.61

99.42

45.9130.9123.18

Fe + Mn

D

α /Phenocrysts

Cores Rims Groundmass

• O Φ 567A-19.CC

T V V 567A-25-2, 90-96 cm

A Δ A 567A-25-3, 74—78 cm

• o ffl 567A-26-1 (Piece 1,3-4 cm)

Fe+ Mn

Figure 3. Clinopyroxene compositions plotted on Ca-Mg-(Fe + Mn) diagram.

the crushing powder adhering to the grains; grains were picked byhand (Sample 567A-29-2, 147-149 cm) when calcite was too abundantin the groundmass. An aluminum-foil target containing a known vol-ume of 3 8Ar buried as positive ions under a 30-kV energy (Bellon etal., 1981) was added to each sample at time of weighing and was usedas a spike. After fusion in a molybdenum crucible heated by induc-

tion, gases released from the sample were cleaned on three titaniumsponge furnaces, and purification was achieved with a titanium-zirco-nium getter. Isotopic analyses were done using isotopic dilution in a180° and 6-cm radius mass spectrometry running in static mode. Thepermanent magnetic field was 3300 gauss, and the ion acceleratingvoltage between 500 and 600 volts. Mass discrimination correction was

659

H. BELLON ET AL.

0.101

0.05-

o567A-19.CC• 567A-25-2, 9 0 - 9 6 cmT567A-25-3, 74 -78 cmA567A-26-1

(Piece 1, 3—4 cm)

Pyroxenes fromtholeiitic and

calc—alkalic basalts

0.5 0.6 0.7 0.8Ca + Na (%)

0.9 1.0

0.05-

~Z 0.03o

0.01-

Pyroxenes fromnonorogenic tholeiites

Pyroxenes from orogenic basalts

0.5 0.6 0.7Ca (%)

0.8 0.9

Figure 4. Plot of clinopyroxene phenocryst composition on Leterrier et al. (1982) diagrams. Gray area in-dicates plot of clinopyroxene phenocryst composition in orogenic lavas from Hole 494A. A. Ti/(Ca +Na) diagram. B. (Ti + Cr)/Ca diagram.

applied after each analysis, using air argon standards within the argoncompositional range of each sample. Table 6 provides age results.

Results and Discussion

Sample 567A-19.CC indicates an age of 78.7 ± 3.9m.y. Beyond the formation of secondary calcite and chlo-

rite in glassy parts of the sample, our geochemical andmineralogical studies detected no major alteration prod-ucts in this basalt, such as K2O enrichment or/and alarge developed secondary mineralogical paragenesis,both linked to seawater weathering. Therefore the age of78.7 m.y. may be considered valid or at least close to the

660

GEOCHEMISTRY, MINEROLOGY, AND RADIOMETRIC DATING

Table 4. Plagioclase analyses (microprobe) for Sample 567A-25-2, 90-96 cm (wt. % ) .

Siθ2AI2O3Fe2θ3CaONa2θK2O

Total

CaNaK

C (25)

54.5427.69

0.6410.275.550.32

99.01

49.748.5

1.8

R(26)

61.2023.53

0.705.298.180.70

99.60

25.370.7

4.0

C(27)

66.4927.27

0.629.366.370.31

100.42

44.154.2

1.7

Position (analysis number)

C(28)

57.1526.21

0.717.946.770.36

99.14

38.559.4

2.1

C(29)

59.1125.04

0.486.707.560.62

99.41

31.965.2

2.9

C(30)

61.2823.45

0.395.118.470.72

99.42

24.072.0

4.0

R(31)

65.2920.110.382.608.971.67

99.02

12.578.0

9.5

G(32)

53.1128.97

0.8211.694.770.20

99.56

56.842.0

1.2

G(33)

54.6427.70

0.649.935.840.33

99.08

47.550.6

1.9

G(34)

58.1025.49

0.827.526.930.50

99.36

36.460.7

2.9

Note: Position is indicated by C = core of phenocryst, R = rim of phenocyrst, and G = groundmass crys-tal; number in parenthesis = analysis number. Fe2θ3* = total iron as FeO. The sizes of Plagioclase phe-nocrysts 25 and 26 and 27 through 31 are 150 µm and 450 µm, respectively.

Table 5. Analyses of chromite (CHR), titanomagnetite (MT), and coexisting hemoilmenite (IL), and of secondaryminerals, chlorite (CHL), celadonite (CE), and albite (AB), (wt. °/o).

Mineral

Siθ2Tiθ2AI2O3Cr2θ3Fe2θ3FeOMnOMgOCaONa2θK2O

Total

Usp (%)He (%)FM

567A-26-1(Piece 1, 3-4)

CHR (35)

0.005.43

16.2126.1517.4020.02

0.306.760.000.060.00

101.35

0.71

MT (36)

0.0522.00

1.350.00

22.8447.73

3.010.000.070.000.00

97.05

65.8

Sample (interval in cm)

567A-25-2, 90-96

MT (37)

0.0823.32

1.230.05

20.6648.71

3.370.000.030.000.02

97.47

69.3

IL (38)

0.0848.55

0.150.008.56

89.490.692.000.070.060.03

99.68

8.1

IL (39)

0.0048.29

0.100.00

10.3139.19

0.652.040.090.000.02

100.69

9.6

IL (40)

0.0049.42

0.260.008.07

38.750.702.900.070.090.03

100.29

7.6

567A-25-2, 90-96

CHL (41)

33.680.01

14.820.00

19.360.00

18.760.450.160.06

87.30

0.37

567A-26-1 (Piece 1

CE (42)

54.440.025.590.00

17.230.006.110.080.04

10.51

94.02

0.61

CE (43)

54.280.137.870.00

15.930.006.020.220.06

10.72

95.23

0.60

,3-4)

AB (44)

67.820.00

20.800.020.08

0.000.040.24

10.090.03

99.12

Note: Numbers in parentheses = analysis numbers. Total iron as Fe2θ3 for albite, FeO for chlorite and celadonite; distributed byR3O4 and R2O3 stoichiometry for spinel and ilmenite, respectively. For ulvöspinel (Usp) and hematite (He), molar percentagesare calculated following CarmichaePs (1967) method. FM = FeO + MnO/MgO + FeO* + MnO.

true crystallization age of the orogenic lava. This age isin good agreement from the same stratigraphic positionat the bottom of upper Campanian - lower Maestrichti-an limestones—around 72 ± 1 m.y., according to Odinand Kennedy (1982).

For the three other lavas with alkaline affinities, theradiometric results are scattered and probably reflect theeffects of alteration and its contribution to the potassi-um budget of each sample, either through a secondaryK2O enrichment in the groundmass or on the rim of pla-gioclase phenocryst or through a secondary K2O deple-tion of the original whole rock linked to the develop-ment of K2O-poor minerals. A correlation appears be-tween the loss on ignition (1050°C) and the age of eachsample. The youngest radiometric age (91 ±4.5 m.y.)was obtained for Sample 567A 29-2, 147-149 cm, whichgave the highest L.O.I, (loss on ignition—1050°C)(5.76%). Sample 567A-25-3, 74-78 cm with a L.O.I, of

5.14%, has an intermediate radiometric age of 132 ±6.6 m.y. The oldest age (169 ± 8 . 4 m.y.), of Sample567A-25-2, 90-96 cm corresponds to a L.O.I, of 4.68%.

In the latter two samples, the K2O contents (0.43 and0.47—see Table 6) and the trace elements data (Rb, 2and 4 ppm; Cs, below 0.05 ppm) indicate no strong vari-ations. A large modification of the primary potassiumbudget seems unlikely, but considering the volume ofsecondary K2O-poor minerals in the samples, it appearsthat the oldest radiometric age is obtained from the ba-salt in which the secondary paragenesis is less devel-oped, compared to the youngest one where chlorite isabundant. The youngest age (91 m.y.) of Sample 567A29-2, 147-149 cm which is a probable pillow-rim, is at-tributed at least in part to a strong enrichment in K2O(1.69%) accompanied by high values of Rb and Cs con-tents (43 and 2.7 ppm respectively), and perhaps to con-tinual loss of a radiogenic argon because of the low re-

661

H. BELLON ET AL.

Table 6. 4 0 K - 4 0 A r radiometric analyses.

SampleDescription (interval in cm)

Porphyritic 567A-19.CCbasalt

Coarse- C 567A-25-2, 9-96grained jdolerites (. 567A-25-3, 74-78

Quenched 567A-29-2, 147-149texture

Age (m.y.)(± uncertainty)

78.7 ± 3.9

169 ± 8.4

77.679.9

168170

132 ± 6.691 ± 4.5

40 Ar*( 1 0 - ? c m 3 / g )

13.1213.5226.8427.2819.1751.29

K 2 O(%)

0.510.510.470.470.431.69

Weight

(g)

0.98390.97731.00531.00211.01270.9941

o/n ^Ar*

4 θ A r T

71.769.168.370.682.650

Exp.number

B.77B.76B.74B.75B.78B.80

Note: 40 Ar* signifies radiogenic argon. Ages are calculated following the relation: time (in m.y.) = 4153.9 logjQ (1 +142.33 4θAr*/K) calculated with the constants by Steiger and Jager (1977). Uncertainties are estimated to be ± 5 % ofthe calculated age. Ages in braces are the results of two independent determinations carried out on the sample. In the4 0Ar*/40Arj relation, ^Ar-p signifies total 40Ar (i.e., radiogenic 40Ar* and atmospheric 40Ar). Exp. number =experiment number.

CM

• 567A-25-2, 90-96 cm• 567A-25-3, 74-75 cm

900 1000 1100T(°C)

1200

Figure 5. Temperature-oxygen fugacity (T-fo2) diagram for coexistingiron-titanium oxides. NNO = nickel-nickel oxide; FMQ = fayali-te-magnetite-quartz buffer; HM = hematite-magnetic buffer.

tention of some of the potassium sites. In this sample,the radiogenic/total argon ratio (Table 6) is significantlylower (0.50), whereas its K2O content is three timesgreater than that of the two other samples. Their argonratios average 0.7 to 0.8.

We may retain the significant difference between theage of the orogenic lava and the ages (132-168 m.y.) ofthe alkaline lavas, which we must consider with caution.These must be analyzed for comparison with ages ob-tained from numerous samples of the onshore complexof Santa Elena, and substantiated by paleontological da-ta. In particular, some radiometric ages around 130 m.y.have been determined on potassic alkaline rocks fromthe lower structural unit of the complex (Bellon et al.unpublished data).

CONCLUSIONS

1. Sample 567A-19.CC, which is stratigraphically anal-ogous to the Hole 494 A basalts, is similar to these ba-salts, and most of their mineralogical and geochemical

characteristics favor an orogenic affinity. Thus the ra-diometric age of 78.7 ± 3 . 9 m.y. may represent the ageof cooling.

2. The four other samples are clearly different from567-19,CC orogenic basalt and from the oceanic basaltsfrom the Cocos Plate. Their geochemical tendencies (en-richment in light rare earths, Ba, Sr, and Ti, and thepresence of normative nepheline) and their mineralogi-cal features (pyroxenes and plagioclases) indicate thatthese rocks are alkaline basalts. They may be comparedto those occurring in the lower unit from the SantaElena complex. Their radiometric ages are older, al-though scattered results may reflect the effects of sec-ondary alteration, and their ages might actually bemuch greater.

3. It is suggested (Aubouin et al., this volume) thatLeg 84 basement sequences are equivalent to those ofthe Santa Elena complex. Our study indicates that base-ment ages may support this interpretation.

ACKNOWLEDGMENTS

R. J. Stern and Stephan Nelson reviewed this chapter; the authorswish to thank them for suggesting improvements in the manuscript.

REFERENCES

Azema, J., and Tournon, J., 1980. La péninsule de Santa-Elena, Cos-ta-Rica: un massif ultrabasique charrié en marge pacifique de l'A-mérique Centrale. CR. Acad. Sci. Paris, 290:9-12.

Barbot, J., 1983. Le nickel dans les olivines et les spinelles des rochesbasiques et ultrabasiques [Thesis]. Université de Bretagne Occiden-tale, Brest.

Bellon, H., Quoc Buü, N., Chaumont, J., and Philippet, J. C , 1981.Implantation ionique d'argon dans une cible-support: applicationau traçage isotopique de 1'argon contenu dans les minéraux et lesroches. CR. Acad. Sci. Paris, 292:977-980.

Buddington, A. F., and Lindsley, D. H., 1964. Iron-titanium oxideminerals and synthetic equivalents. J. Petrol., 5:310-357.

Carmichael, I. S. E., 1967. The iron-titanium oxides of salic volcanicrocks and their associated ferromagnesian silicates. Contrib. Min-eral. Petrol., 14:36-64.

D'Arco, Ph., Maury, R. C , and Westercamp, D., 1981. Geothermo-metry and geobarometry of a cummingtonite-bearing dacite fromMartinique, Lesser Antilles. Contrib. Mineral. Petrol., 77:177-184.

Girard, D., 1981. Pétrologie de quelques series spilitiques mésozoiquesde la zone caraibe et des ensembles magmatiques de 1'ile de Toba-go: implications géodynamiques [Thesis]. Université de BretagneOccidentale, Brest.

Leterrier, J., Maury, R. C , Thonon, P., Girard, D., and Marchal, M.,1982. Clinopyroxene compositions as a method of identification

662

GEOCHEMISTRY, MINEROLOGY, AND "<>K-̂ Ar RADIOMETRIC DATING

of the magmatic affinities of paleo-volcanic series. Earth Planet.Sci. Lett., 4:33-38.

Maury, R. C , Bougault, H., Joron, J. L., Girard, D., Treuil, M.,Azema, J., and Aubouin, J., 1982. Volcanic rocks from Leg 67sites: mineralogy and geochemistry. In Aubouin, J., von Huene, R.et al., Init. Repts., DSDP, 67: Washington (U.S. Govt. PantingOffice), 557-576.

Odin, G. S., and Kennedy, W. J., 1982. Mise à jour de 1'échelle destemps mésozoiques. CR. Acad. Sci. Paris, 294:383-386.

Powell, R., and Powell, M., 1977. Geothermometry and oxygen ba-rometry using coexisting iron-titanium oxides: a reappraisal. Min-eral. Mag., 41:257-263.

Shervais, J. W., 1982. Ti-V plots and the petrogenesis of modern andophiolitic lavas. Earth Planet. Sci. Lett., 59:101-118.

Sigurdsson, H., and Schilling, J.-G., 1976. Spinels in Mid-AtlanticRidge basalts: chemistry and occurrence. Earth Planet. Sci. Lett.,29:7-20.

Steiger, R. H., and Jager, E. 1977. Subcommission on geochronology:convention on the use of decay constants in geo and cosmochro-nology. Earth Planet. Sci. Lett., 26:359-362.

Date of Initial Receipt: 7 April 1984Date of Acceptance: 23 August 1983

663