The basalts from the Kerguelen Plateau are represented by aphyric and sparsely porphyritic varieties (Coffin, Frey, Wallace, et al., 2000). Tables T3, T4, T5, and T6 summarize the petrographic compositions of the samples studied. Aphyric basalts dominate in most Leg 183 holes. Rarer porphyritic basalts prevail in Holes 1140A and 1137A. Phenocrysts are represented mainly by plagioclase. Phenocrysts of olivine and clinopyroxene are rare (Hole 1137A). The total quantity of phenocrysts within porphyritic basalts varies from single crystals through 25% of the rock volume. Both the degree of crystallinity and vesicularity of the basalts vary greatly. Massive basalts are rare but most common in the northern part of the Kerguelen Plateau (Hole 1140A). Less vesicular basalts are recognized in Holes 1136A and 1138A. Both intergranular and intersertal textures are dominant. Vitrophyric texture is present occasionally (Hole 1140A).
The results of chemical analyses of the basalts from Holes 1136A, 1137A, 1138A, and 1140A (Leg 183) and from Hole 747C (Leg 120) are given in Table T7. To plot petrochemical diagrams, we have recalculated concentrations of major elements to the dry residue (recalculated to 100%). In plotting diagrams, we did not use the following samples: Z-1012 (Sample 183-1137A-34R-3, 18-23 cm), Z-1014 (Sample 39R-2, 114-119 cm), Z-1016 (Sample 43R-2, 84-92 cm), Z-1017 (Sample 45R-1, 121-126 cm), and Z-1018 (Sample 78R-1, 60-64 cm), as they represent litho- and crystalloclastic dacitic tuffs, tuff/basalt contact, and highly altered basalt.
On the (Na2O + K2O)-SiO2 diagram, basalts from the Kerguelen Plateau plot in both tholeiitic and alkaline fields (Fig. F1). Holes 1140A, 749C, and 750B provide exceptions, as they contain only tholeiitic basalts.
On the three-component diagram, Zr-Nb-Y (Fig. F2), the data plot in a compact field in the area of normal mid-ocean-ridge basalt (N-MORB) (Holes 1136A and 1140A), in the area of intraplate tholeiites (Holes 1137A, 1138A, and two samples from Hole 1140A), and the basalts from Hole 747C plot in two fields enriched by MORB (basalts from plume-influenced regions [P type]) and intraplate oceanic tholeiites.
In most cases, basalts from each hole (except Hole 1140A) demonstrate similar REEs in both concentration and distribution patterns, indicative of a common mantle source (Fig. F3). REE distribution indicates that basalts from Holes 1136A, 1137A, 1138A, and 747C are enriched in light rare earth elements (LREEs) in comparison with normal tholeiitic basalts of mid-ocean ridges.
The variational REE diagram for basalts from Hole 1140A (Fig. F3) seems to show two groups. The trend for low-Ti basalts (group 1) is similar to that typical for normal oceanic tholeiites. The trend for high-Ti basalts (group 2) is enriched in LREEs. This suggests that melts, which formed the basalts recovered from Hole 1140A, belong to two sources from different depths.
Figure F4 demonstrates average REE distribution in basalts from holes drilled on the Kerguelen Plateau. Judging from both the REE values and the distribution pattern, basalts from Hole 750B in the southern part of the Kerguelen Plateau are similar to N-MORB. This provides evidence that their initial melts derived from depleted mantle sources. The same is also evident in group 1 basalts from Hole 1140A. Basalts from Holes 1136A and 749C are similar in REE values and distribution of average composition of enriched mid-ocean-ridge basalts (E-MORBs). Basalts from Holes 1138A and 1140A (group 2), and especially those from the Elan Bank (Hole 1137A), are derived from significantly enriched (compared to N-MORB) sources. The basalts from Hole 747C (central part of the Kerguelen Plateau) and tholeiites from Kerguelen Island are similar in REE values and distribution. They demonstrate both LREE and LREE/heavy rare earth elements (HREE) values higher than E-MORB. Alkaline basalts from the Kerguelen and Heard Islands demonstrate the highest values of these parameters.
Thin sections show that the basalts from Hole 1136A are weakly to moderately altered (Table T3). According to whole-rock chemical analyses, altered basalts contain 0.90-1.58 wt% H2O+ (Table T7). The degree of rock oxidation is moderate. Both olivine and interstitial glass are completely replaced by smectite-chlorite aggregate. Plagioclase is partly replaced by smectite-chlorite aggregate. Vesicles are filled mainly with smectite (Table T8). Veins are filled with calcite. Secondary minerals identified in basalts from Hole 1136A, from vesicles and veins (Table T8) and from groundmass (Table T3), indicate a low-temperature environment for water-rock interaction.
Thin section examination indicated that basalts from Hole 1137A are more altered than those from Hole 1136A. We estimate the degree of basalt alteration as 10%-60% (Table T4). Chemical analyses show that the altered basalts contain 0.19-3.03 wt% H2O+ (Table T7). The Fe2O3/FeO ratio varies from 0.88 to 3.57 (Table T7) and also indicates various degrees of basalt oxidation. Tuff (Sample 183-1137A-43R-2, 84-92 cm) is strongly altered (80%; estimate based on the thin section analysis) (Table T4) and highly oxidized (Fe2O3/FeO = 10.14) (Table T7).
Thin sections show that ferromagnesian minerals in phenocrysts, groundmass, and intersertal glass are completely replaced by chlorite and chlorite-smectite aggregates. Plagioclase phenocrysts in porphyritic basalt (Sample 183-1137A-45R-1, 121-126 cm) are replaced with K-feldspar (adularia?), smectite-chlorite aggregates, and carbonate (Table T4). Plagioclase from the matrix also is replaced with K-feldspar. The degree of alteration is high (40%) (Table T4). Highly altered tuff (Sample 183-1137A-43R-2, 84-92 cm) contains pseudomorphs of chlorite and carbonate derived from alteration of a prismatic minerals. Biotite is chloritized. Interstitial glass is replaced by smectite-chlorite aggregate.
In basement Subunit 2A and Unit 4 (Hole 1137A), smectite is the most common mineral filling vesicles (Table T9). Often, it is present with some admixture of clinoptilolite and heulandite. An admixture of a 7-Å mineral (probably dickite) and calcite is also present. In Subunits 7A and 8A, vesicles also contain chlorite phases (Table T9). Subunit 7A (Sample 183-1137A-37R-5, 73-79 cm) contains a mixed-layer smectite-chlorite mineral. Chlorite, a dominant mineral, is present below (Sample 183-1137A-39R-3, 6-14 cm) and is also accompanied by defective chlorite (Sample 39R-2, 114-119 cm). Defective chlorite was identified according to Drits and Tchoubar (1990). Clinoptilolite, hydromica, and quartz are also present with smectite and chlorite.
We studied two veins in the basalt section of Hole 1137A (Table T9). Veinlets contain smectite, calcite, and clinoptilolite and quartz in trace amounts.
The alteration zone between the basalt and tuff (green matter; Sample 183-1137A-34R-3, 18-23 cm) contains mixed-layer hydromica-smectite minerals and hydromica and quartz in trace amounts (Table T9). The appearance of hydromica phases probably indicates water migration in an oxidizing environment.
The entire complex of secondary minerals in basalts from Hole 1137A (Tables T4, T9) indicates the low-temperature conditions of alteration and shows absence of vertical zonation in secondary minerals in the basalt section in total. The presence of dickite in trace amounts in the upper part of Subunit 2A and Unit 4 provides evidence of subaerial weathering.
Examination of the basalt section from Hole 1138A shows that alteration varies from slight to intense (from 10% to 50%) (Table T5). By chemical analysis, they contain 0.36-4.96 wt% H2O+ (Table T7). The basalts are mostly nonoxidized or only slightly oxidized as suggested by dark gray color, study of thin sections, and Fe2O3/FeO ratio. The latter varies from 0.40 to 2.06 (Table T7). In contrast, tuff (Sample 183-1138A-78R-1, 60-64 cm) is strongly oxidized, with an Fe2O3/FeO ratio of 25.64 and altered (60% in thin section) (Table T5).
Olivine is completely replaced by pale green chlorite. Plagioclase is partly replaced by chlorite. The interior of the plagioclase is replaced by a micaceous mineral (Sample 183-1138A-84R-5, 20-25 cm). Interstitial glass is completely replaced by chlorite and smectite-chlorite aggregates (Table T5).
Secondary minerals filling vesicles in basalt from Hole 1138A (Tables T10, T11) are characteristic and distinct from those of other Leg 183 holes in the abundance of zeolites (heulandite, clinoptilolite, mordenite, stilbite, analcime, and natrolite). Thomsonite is present occasionally. No vertical zonation of zeolite distribution in Hole 1138A is obvious. For example, clinoptilolite and heulandite are present in basalts in various parts of the basalt section recovered from Hole 1138A. Mordenite is present only in basalt from Unit 6 (Sample 183-1138A-81R-1, 34-39 cm, 30 cm below the top of the lava flow). Analcime and stilbite are present in vesicles and veins of basalts from Units 17 and 19. These minerals are absent from other parts of the basalt section. Clay minerals also lack any vertical zonation (Tables T5, T10). All secondary minerals (clay and nonclay minerals) studied in vesicles and veins, as well as in basalt groundmass, show no vertical zonation in their distribution throughout the basalt section in Hole 1138A.
The abundance of zeolites in basalts from Hole 1138A (in comparison with other Leg 183 holes) is probably related to its closer location to the paleoeruptive center. The presence of several varieties of zeolites and great variation in chemical composition is characteristic of rock alteration in hydrothermal systems.
Within individual lava flows, there are no limitations on the presence of smectite. It fills vesicles near the top of basalt flows, for example, at 4, 20, and 30 cm below the top of the lava flows (Samples 183-1138A-88R-2, 91-96 cm; 82R-2, 76-81 cm; and 81R-1, 34-39 cm, respectively), at the bottom of Unit 17 (Sample 86R-3, 72-74 cm), or at 90 cm above the bottom of Unit 10 (Sample 83R-4, 19-24 cm) (Table T10). Besides marginal parts of lava flows, smectite is present in the interior of basalt Units 6, 9, 11, 13, and 19 (Samples 183-1138A-81R-2, 51-57 cm; 82R-5, 19-23 cm; 83R-5, 106-109 cm; 84R-5, 20-25 cm; and 87R-2, 76-81 cm). All determinations of chlorite, defective chlorite, and serpentine(?) were made from interior and basal parts of basalt flows (Table T10).
The absence of any vertical zonation in secondary minerals in basalt sections in total and the presence of zonation within individual flows had been shown in basalts from Suiko Guyot in the Emperor Seamount Chain (Hole 433C, Leg 55) (Kurnosov, 1986), most impressively in Unit 48 (7.5 m thick). Smectite dominates at the top and bottom of the flow, whereas toward centre of the flow, smectite replaced swelling chlorite. Mixed-layer chlorite-swelling chlorite dominates in the flow interior. A similar distribution of secondary minerals and lack of vertical zonation were recognized in the West Pacific Guyots, Hole 865A, Legs 143 and 144 (Kurnosov et al., 1995). Basalt flows in Allison Guyot suffered low-temperature smectitization, mainly in the upper parts of the units, and chloritization (swelling chlorite and mixed-layer smectite-chlorite mineral) in the interior.
The secondary minerals have probably formed under the influence of individual basalt flows in an environment dominated by horizontal migration of water, primarily along contacts between lava flows. The influence of hot waters of interlayer-fissure circulation on the formation of subhorizontal zeolite zones in basalts is well known in Iceland (Walker, 1960; Tomasson and Kristsmannsdottir, 1972; Kristsmannsdottir and Tomasson, 1978).
Hole 747C is located near Hole 1138A. The basalts studied from Hole 747C show various degrees of alteration (H2O+ varies from 0.69 to 5.14 wt%) (Table T7). In altered basalts, dominant secondary minerals are represented by smectite, or smectites with chlorite or swelling chlorite, and chlorite. Zeolites were determined in amygdules, in the groundmass, and in altered plagioclase phenocrysts. Zeolites are represented by chabazite, natrolite, thomsonite, mesolite, stilbite, and heulandite (Sevigny et al., 1992). Comparison of zeolite and clay minerals in Leg 183 and 120 basalts and in Iceland (Kristmannsdottir and Tomasson, 1978) suggests that alteration of basalts from the central part of the Kerguelen Plateau (Holes 1138A and 747C) occurred at a temperature of 120°C.
Two types of alteration, oxidative and nonoxidative, are recognized in Hole 747C. Oxidative zones are marked by goethite, Fe hydroxides, calcite, and celadonite(?). This secondary mineral assemblage forms at sites of low-temperature water-basalt interaction (Bass, 1976; Bass et al., 1973; Kurnosov, 1986).
Basalts from pillow lavas of Hole 1140A are less altered than basalts from other Leg 183 holes. Thin sections indicate alteration of basalts of 5% to 20% (Table T6). Only two samples had alteration of a moderate to high degree, from 25% to 35%. The basalts are fresh or scarcely oxidized. The Fe2O3/FeO ratio is low and varies from 0.53 to 1.07 (Table T7). In only two samples, the ratio was 2.41 and 2.31.
Aggregates of chlorite-smectite replace olivine. Plagioclase is completely replaced with chlorite-smectite aggregate or with a mica-type mineral. Interstitial glass is completely replaced by chlorite or chlorite and ore minerals (Table T6). Smectites dominate in the fine fraction that was removed from the basalts (Table T6).
In four samples, vesicles are filled with smectite; the vein sampled from the glass is filled with calcite (Table T12). Joint fissures are covered, in one case, with a thin layer of smectite and defective chlorite and, in the other case, smectite with traces of serpentine(?) and quartz.
Zeolites have not been identified, and this is the principal difference in basalt alteration from Hole 1140A (submarine extrusion) compared with those from Holes 1136A, 1137A, and, especially, Hole 1138A (subaerial extrusion).
Thus, pillow lava basalts from Hole 1140A have a low degree of alteration and most are not oxidized. Secondary minerals indicate a low-temperature alteration environment. Circulation of hotter fluids probably occurred along cracks now filled with smectite, defective chlorite, serpentine(?), and quartz. Nevertheless, these probable fluids did not play a significant role in the alteration of the basalt section of Hole 1140A.
We studied mobility of chemical elements in relation to alteration of basalts under both oxidative and nonoxidative environments (Bass 1976; Bass et al., 1973). To estimate chemical element mobility, we used data on the amount of major and minor elements (grams per 1000 cm3) in basalts (Table T13).
Table T14 shows two examples of mass balance from nonoxidative and oxidative environments of alteration of basalts from the Kerguelen Plateau.
Basalts from Hole 1140A have similar degrees of alteration. Nevertheless, they are favorable for analyzing the mobility of chemical element in a "pure" nonoxidized environment, as they are neither oxidized nor vesicular. For further reference, we chose Sample 183-1140A-35R-1, 35-42 cm, where H2O+ = 0.74 wt%; Fe2O3/FeO ratio = 1.01; and density = 2.93 g/cm3 (Tables T7, T14). For comparison we chose nonoxidized Sample 183-1140A-34R-1, 117-121 cm, where H2O+ = 1.40 wt%; Fe2O3/FeO ratio = 0.69; and density = 2.57 g/cm3. The trend to a decrease under nonoxidative alteration in basalt is most evident in major elements, REEs, Cu, Nb, Zr, Y, Rb, and Sr (Table T14). Ni, V, Co, Zn, and Ba show a weak trend to increase during alteration.
Sample 183-1137A-27R-1, 100-105 cm (Fe2O3/FeO = 3.57 and H2O+ = 2.03 wt%), is oxidized (Table T7) and highly vesicular (Table T14). Hence, for comparison, we chose relatively fresh nonoxidized basalt with similar vesicularity (Sample 183-1137A-39R-2, 114-119 cm; H2O+ = 0.99 wt%). Comparison revealed that oxidizing alteration leads mostly to the accumulation of Fe, Mg, Ca, P, REEs, Co, Zr, and Ba in basalts (Table T14). This is especially evident for Sr. In contrast, Si, Al, Mn, Na, K, Ni, V, Cu, Y, and Rb show a decrease.
This study has shown that the mobility of chemical elements during alteration of basalts from the Kerguelen Plateau is different in oxidizing and nonoxidizing environments. We conclude that generally in nonoxidizing alteration environments, basalts lose most elements. In contrast, in oxidizing alteration, basalts accumulate many elements. The degree of alteration of basalts (selected for estimation of chemical element mobility) is low, so the mobility of elements seems to be at the rudimentary stage.