IGNEOUS PETROLOGY AND GEOCHEMISTRY

Basement recovered at Site 1140 consisted of 86.7 m of submarine igneous rocks and upper Eocene interbedded pelagic sediments at depths from 234.5 to 321.2 mbsf, with 50% recovery (see Fig. F15; Table T6). In contrast to previous sites (Sites 1136 to 1139), where the principal criteria for division of basement sequences into units were volcanological, at Site 1140 we used (1) the presence of interbedded sedimentary layers and (2) notable downhole changes in phenocryst abundance or type. On this basis, we divided the section into five basaltic units and one sedimentary unit. Unit thicknesses are presented in Table T5 (see "Physical Volcanology"), and we report a summary of the mineralogy, petrology, and igneous texture in Table T6. Information on the volcanic features and on alteration are provided in other sections (see "Physical Volcanology" and "Alteration and Weathering").

All igneous units are pillow basalts. Recovered parts of Units 2 and 3 consist entirely of small (40-100 cm) basaltic pillows. Units 1 and 6 also have <1-m pillows but also contain large ~5-m massive lobes. Unit 5 is more massive and probably composed of only three large lobes (>3 m thick). Unit 4 consists of pale brown to orange dolomite, of which we recovered 0.8 m, including both upper and lower contacts (see "Lithostratigraphy").

An ~3-cm-thick piece of pale green, poorly laminated, sheared nannofossil chalk separates Units 2 and 3 and another ~4-cm-thick piece of pink to orange, well-indurated, fine- to medium-grained dolomite separates Units 5 and 6. These two thin sedimentary layers are grouped with the underlying basaltic units. We interpret the units separated by sedimentary interbeds (Units 2 and 3 and Units 5 and 6) to represent separate eruptions. Even though there was no recovery in the cores that separate Units 1 and 2 (Cores 183-1140A-29R and 30R), a probable sedimentary interlayer was recorded in logging data, and magnetic polarity changes (see "Downhole Measurements" and "Paleomagnetism"). We therefore interpret Units 1 and 2 as separate eruptions. All recovery from Unit 2 was in a single 140-cm-thick section (Section 183-1140A-31R-1), but it is likely that this is only the basal portion of a thicker unit (see Fig. F15). Some units may, however, contain more than one eruption of lithologically similar lava, as suggested, for example, by isolated pockets of sedimentary material between several pillow margins in Unit 1.

With the exception of regions adjacent to pillow margins, particularly where sediment is intercalated or in regions of veining, the basalt recovered is very fresh. The glass at pillow margins shows little sign of hydration or devitrification, and minerals in pillow interiors are commonly unaltered. In the freshest parts of Units 1 and 2, alteration is limited to replacement of interstitial glass by clay and the filling of vesicles by clay and carbonate (see "Alteration and Weathering"). The Site 1140 basalts are unique among those recovered during Leg 183 in that they contain fresh olivine and glass.

Anorthite-rich plagioclase (An_60-70) is the dominant phenocryst phase in all units; it is accompanied by olivine (Fig. F16), except in Unit 6, and by clinopyroxene in Units 2 and 6. Units 1 and 2 are petrographically distinct: Unit 1 contains <1% phenocrysts (except in pillow margins), whereas Unit 2 contains ~20% phenocrysts (Table T6). Many phenocrysts of plagioclase and clinopyroxene in these units are strongly zoned (Figs. F17A, F17B, F18A), and clinopyroxene also displays sector zonation (Fig. F18B). Although most of the olivine phenocrysts are partially to completely altered to brown green clay (Fig. F19), some are fresh.

Although the mineralogy does not differ greatly from that of basalts recovered at previous sites, submarine eruption has produced distinctive textures. To preserve glass for shore-based study, only two thin sections of pillow margins were made. The outermost 1-2 cm of pillow rims are composed of dark brown glass, which in thin section appears unaltered, except adjacent to interpillow sediment, where it is replaced by brown clay (Fig. F20A, F20B). Pillow margins are sparsely porphyritic with a very fine grained to glassy (vitrophyric) groundmass (Figs. F16, F21, F22) that contains extremely fine grained plagioclase microlites or small laths with ragged or swallow-tailed terminations (Fig. F21). Olivine forms subequant euhedral phenocrysts. In the glassy rinds, some of the olivine phenocrysts contain numerous inclusions of unaltered glass and rare inclusions of chromite (Figs. F16, F22). Irregular patches of laminated, foraminifer-bearing dolomitic chalk, and thin (1-2 mm wide) carbonate veins, fill sutures and interstices between some pillows (Figs. F6, F20B, F23).

The main mineral phases of the massive interiors of pillows and lobes are similar to those of the chilled margins (Table T6). The massive interiors of pillows and lava lobes are, however, more crystalline and coarser grained than the pillow margins, reflecting slower cooling rates in the flow interiors. The textures range from intersertal to intergranular, but unlike other Leg 183 sites, subophitic (Fig. F24A, F24B) to locally ophitic textures are present in some of the thicker lobes and pillows. Subophitic texture is most apparent in interiors of the thick lava lobes in Units 5 and 6, where the maximum clinopyroxene and plagioclase grain size reaches 0.7 mm. Flow interiors appear less porphyritic than the margins, probably because phenocrysts of the type preserved in fine-grained pillow margins became surrounded by late-formed crystals of similar size during slow cooling. Clinopyroxene and plagioclase are major groundmass phases in crystalline areas; minor phases are titanomagnetite (5%) and intersertal glass, some of which is still fresh. In some samples, glass has partially crystallized to cryptocrystalline clinopyroxene and opaque minerals. The olivine in the groundmass is partially altered (Fig. F25). Titanomagnetite is generally skeletal. Sulfide of possible magmatic origin is present as minute (<0.01 mm) inclusions in primary minerals and fresh glass (Fig. F26A). Although the small size precludes a positive identification, its optical characteristics suggest that it is pentlandite or pyrite with minor chalcopyrite. Units 2, 3, 5, and 6 (Fig. F8) contain secondary pyrite in veins and as small disseminated grains that replace residual glass (Fig. F26B). Further information on the pyrite is given in "Alteration and Weathering".

We list X-ray fluorescence (XRF) analyses of major and trace elements in 11 basalts in Table T7. Basalt compositions are all tholeiitic (Fig. F27). They are olivine normative in Unit 1 and quartz normative in Units 2, 3, 5, and 6 (Table T7). The low loss on ignition values (<1.8 wt%) support the petrographic evidence that the pillow interiors and the massive portions of these units are relatively unaltered. Except for samples from Units 2 and 3, major element abundances are relatively uniform, with narrow ranges in MgO (6.3 to 8.1 wt%; Mg# from 0.54 to 0.63) and SiO₂ (48.9 to 50.7 wt%) (Fig. F28). However, systematic changes downhole are illustrated in major element abundances. Overall, SiO₂ and Na₂O increase slightly downhole (Fig. F29). In addition, there is a very strong break at Units 2 and 3, which have much higher K₂O, P₂O₅, and TiO₂ and lower Mg#. The major element trends are mirrored by variations in trace element abundances and ratios (Fig. F30). Units 2 and 3 have higher Ba, Zr, Nb, and lower Cr contents and also exhibit the highest [Nb/Zr]_N and [Zr/Y]_N ratios (Figs. F30). Units 1, 5, and 6 define downhole trends of increasing Na₂O content and [Nb/Zr]_N and decreasing Cr, Ni, [Zr/Ti]_N, and [Zr/Y]_N (Figs. F30, F31).

Although the compositional variations of Site 1140 lavas reflect variable degrees of crystal fractionation, essentially olivine and plagioclase as observed in thin sections, variations in Nb/Zr and Zr/Y ratios indicate that the whole sequence cannot have been derived from fractional crystallization of a single parental magma composition. In particular, the enrichment in incompatible element abundances and high Nb/Zr and Zr/Y of lavas from Units 2 and 3 could not be derived by fractional crystallization of the higher Mg# lavas in Units 5 and 6. Finally, the magnetic field reversal between Units 1 and 2 (see "Paleomagnetism") indicates a significant time break between eruptions of Units 1 and 2.

Units 1, 5, and 6 have relatively low concentrations of incompatible trace elements and almost chondritic ratios of Nb/Zr and Zr/Y (Fig. F31). In a primitive mantle-normalized diagram, Units 1, 5, and 6 are between Site 749 and 750 basalts for most elements, but they are distinguished by Rb and K depletion (or Ba enrichment?). Relative to Units 1, 5, and 6, Units 2 and 3 are more enriched in incompatible elements by a factor of 5 (Fig. F31). Site 1140 lavas are among the most incompatible element-depleted lavas from the Kerguelen Plateau (Figs. F27, F31).

Trace element discriminants such as (Zr/Ti)_N vs. Zr illustrate that the Site 1140 basalts fall within the range of central and southern Kerguelen Plateau basalts (Fig. F32). The (Zr/Ti)_N values for Units 1, 5, and 6 are within the range of those for Site 749 basalts and are lower than those of basalts from Sites 1136, 1137, and 1138. With regard to Nb/Y vs. Zr/Y, basalts from Units 1, 5, and 6 are again similar to the Site 749 basalts, whereas Units 2 and 3 basalts are similar to the lower series lavas of Site 1138 (Fig. F33). Site 1140 tholeiites plot within the field defined by basalts derived from the Iceland mantle plume being distinct from mid-ocean ridge basalts (MORB) (Fitton et al., 1997, 1998). In this respect, they are similar to basalts from Sites 749 and 750, which do not show other evidence of continental crustal assimilation (Fig. F33). There is no evidence from shipboard data that Site 1140 magmas assimilated any continental crustal material.

The following features distinguish Site 1140 basalts from those of other Leg 183 sites:

1. Submarine eruption, as indicated by their pillowed structure and intercalated nannofossil-bearing calcareous marine sediments;
2. Relative absence of posteruptive alteration, as indicated by the preservation of fresh glass at pillow margins and the lack of alteration in massive sections;
3. Tholeiitic basalts with higher MgO (up to 8.1 wt%) and Ni (up to 100 ppm) than other basalts from Leg 183 drill sites on the Kerguelen Plateau; and
4. An age between 33.1 and 35.4 Ma, as indicated by the paleomagnetic and biostratigraphic time constraints (see "Paleomagnetism" and "Biostratigraphy").

Site 1140 basalts exhibit some important geochemical features. Two of the five basalt units were derived from a geochemically distinct parental magma that had higher abundances of highly incompatible elements. Despite large differences in age and location, basalts from Units 1, 5, and 6 at this NKP site and basalts from Site 749 and 750 on the southern Kerguelen Plateau (SKP) have some significant geochemical similarities; they are slightly enriched tholeiites (i.e., with higher Nb/Y and Zr/Y than MORB). Site 749 and 750 basalts have the highest ¹⁴³Nd/¹⁴⁴Nd and lowest ⁸⁷Sr/⁸⁶Sr ratios of the Kerguelen Plateau (see Fig. F9 in the "Leg 183 Summary" chapter). These isotope ratios have been interpreted as reflecting the presence of a MORB component in Kerguelen plume-derived magmas. It will be interesting to establish with shore-based isotope studies if Site 1140 basalts show the same features. In contrast to basalts from Site 1137 on Elan Bank, there is no indication that Site 1140 basalts were contaminated by continental crust.