GEOCHEMISTRY

Site 1163 is located on ~17-Ma crust within Segment B4 of the AAD. Two lithologic units are identified in Hole 1163A, based on macroscopic and microscopic examination (see "Igneous Petrology"). Two glass samples—one from a pillow basalt fragment (Sample 187-1163A-1W-CC) and one from a glass clast in carbonate-cemented breccia (Sample 187-1163-8R-1, 41-43 cm)—were analyzed for major and trace elements by ICP-AES (Table T3). Three whole-rock powders were analyzed for major and trace elements by XRF only.

Both Units 1 and 2 basalt glasses from Hole 1163A have ~9.0 wt% MgO; in contrast, whole rocks have 0.5-1.0 wt% less MgO (Fig. F20), as has been observed at previous sites during Leg 187. The glasses and whole rocks overlap in all other major and trace element concentrations, except for Ni, which is noticeably higher in whole-rock samples. Average analyses suggest that Unit 1 and Unit 2 glasses have different major element characteristics, but the ranges reported in Table T3 overlap substantially (Fig. F21), indicating that some differences are analytical. However, analysis of trace elements was less problematic, so we are confident that the trace element difference observed between the Unit 1 and Unit 2 glasses is significant. In particular, Unit 2 is higher in all trace elements except Cr (Fig. F21). Unit 2 also has lower CaO/Al₂O₃ ratios, suggesting a lower degree of melting, consistent with its elevated trace element contents. Unit 1 and Unit 2 whole rocks show the same compositional contrasts as their associated glasses.

The glasses from Hole 1163A are more primitive (i.e., ~9.0 wt% MgO) than any basalt yet recovered within Segment B4. A hallmark of AAD Segment B4, the present focus of the depth anomaly, is its compositional diversity, driven by variations in mantle melting processes rather than by low-pressure crystal fractionation. Simple data trends with decreasing MgO are rare. This generalization also applies to Site 1163 lavas, which display a relatively large range in trace element concentrations at a given MgO content. Y, Zr, Cr, and TiO₂ contents are higher, and Ba and Na₂O are lower, than in younger Segment B4 lavas, increasing the compositional diversity within this segment beyond that previously documented. The slightly higher CaO/Al₂O₃ ratios and lower Na₂O contents of Site 1163 glasses indicate a higher degree of melting at 17 Ma than beneath the present Segment B4 axis. Higher Y, Zr, and TiO₂ contents also suggest a mantle source higher in these constituents relative to the mantle source for present-day Segment B4 lavas.

The Zr/Ba systematics of Site 1163 basalts indicate an Indian- to Transitional Pacific-type mantle source beneath Segment B4 of the AAD at ~17 Ma (Fig. F22A). Both Unit 1 and Unit 2 glasses have higher Zr/Ba ratios and lower Ba contents than present-day Segment B4 lavas and, consequently, plot closer to the Pacific-type field. In spite of plotting within the Pacific field defined by zero-age SEIR basalt glass, the Na₂O/TiO₂ systematics suggest an Indian-type mantle affinity (Fig. F22B) based on the distribution of Leg 187 glass samples. In addition, low Na₂O has been a consistent characteristic of Leg 187 basalts (see "Sodium and Titanium" in the "Leg Summary" chapter). The lower Na₂O and often higher TiO₂ contents of Site 1163 lavas offset the data to a position below the original division based on 0- to 7-Ma SEIR mid-ocean-ridge basalt (MORB) glass data, even for rocks whose Ba/Zr values give a clear indication of Indian mantle provenance. We believe this indicates a significant contrast in the major and trace element character of the mantle and its melting properties in this region between the time Leg 187 lavas were produced and the present day.