Leg 187 was unusual in the large number of holes (23 holes at 13 sites) drilled into basaltic basement and the wide range of crustal ages sampled. The array of drilled sites approximates three ~14- to 28-Ma flow lines that encompass the known contrasts in morphology and geochemical diversity along the SEIR. When combined with the existing suite of young (0–7 Ma), well-characterized dredge samples, these samples provide a rare opportunity to document the progressive alteration (weathering) of seafloor basalts as they age and move away from the spreading axis.
No evidence for pervasive hydrothermal alteration of basalts was encountered during Leg 187.
Thorseth et al. (2003) used conventional microscopy and scanning electron microprobe (SEM) to investigate the mineralogy of glass alteration and the role of microbes in this alteration, using both Leg 187 core material and glasses from dredge samples of younger (0–2.5 Ma) seafloor.
For 0- to 2.5-Ma samples from the AAD, there is ample evidence of ongoing microbial activity along fracture surfaces in the basalt glass and on zeolite crystal surfaces within hairline fractures. Alteration halos on glasses from Leg 187 cores are much thicker adjacent to fractures, but glassy rim thicknesses do not change significantly with age. Microbial morphologies are rare within these glassy rims, although some Mn-rich spherical forms, interpreted as having a microbial origin, were detected in more porous regions within altered glass and in some fractures. At least one style of diffuse, irregular alteration front that is present only in the Leg 187 glasses is inferred to have formed postburial and in the absence of microbial activity (Fig. F3).
Miller and Kelley (this volume) examined the mineralogy and major and trace element chemistry of altered (discolored) basalt whole-rock samples from all 13 Leg 187 sites. They found a uniform and ubiquitous suite of alteration products dominated by Fe oxyhydroxides, smectite-group clays, and palagonite. The alteration assemblages occur in one of four modes: (1) as replacements of irregular patches in the groundmass, (2) as partial or complete replacements of phenocrysts, (3) as vesicle linings and fillings, or (4) as vein or fracture linings and infill. These modes occur in varying combinations in different samples. At Site 1162, some breccia clasts contained high-temperature greenschist facies alteration assemblages that include talc, actinolite, chlorite, and albite.
The dominant chemical indicators of alteration appear to be decreased MgO and increased loss on ignition; for both these indicators, the maximum values at a given site increase systematically with secondary mineral abundance. Both indicators also increase with distance off axis over most of the drill sites, with the notable exception of the oldest three sites, 1152, 1153, and 1154 (25–28 Ma), which are visually and chemically less altered than the younger sites, 1155, 1156, and 1157 (20–25 Ma) (Fig. F4). In three composite flows at Site 1160, we documented pervasive MgO loss from pillow interiors, relative both to their glassy margins and to the underlying massive potions of the composite flows (see Shipboard Scientific Party, 2001). A systematic onshore analysis of alteration chemistry further reveals that, relative to Al2O3, which appears to have been immobile, MgO, Fe2O3 (total), MnO, and SiO2 are lost from altered samples, whereas CaO and K2O are added during the alteration process (Miller and Kelley, this volume).
S. Krolikowska-Ciaglo and F. Hauff (pers. comm., 2003) compared trace element and isotopic compositions of discolored, visibly weathered outer surfaces of basalt pieces with their less altered interiors and, where possible, with coexisting fresh glass. They report a chemical alteration signature that is uniform in nature and extent across the range of depth and latitude encompassed by the drilling. This behavior contrasts with that of the major elements as reported by Miller and Kelley (this volume) and suggests a contrary conclusion—that chemical exchange with seawater had ceased or at least slowed significantly prior to the ~14-Ma age of the youngest Leg 187 site. Krolikowska-Ciaglo and Hauff report that alteration is characterized by loss of MgO (in agreement with Miller and Kelley and onboard studies) and by pervasive addition of boron plus the high field–strength cations Cs, Rb, U, and K. Concentrations of these added elements correlate well with one another and with visual estimates of the degree of alteration. Predictably, Nd contents and 143Nd/144Nd isotopic ratios are not affected by alteration. Sr content does not increase with degree of alteration, but isotopic exchange with seawater increased 87Sr/86Sr significantly, with the most altered samples tending to have the most radiogenic Sr (Fig. F5). Krolikowska-Ciaglo and Hauff postulate that this exchange occurs primarily at sites within the interlayers of hydrated smectites produced by the alteration process.
Within individual drill holes, 207Pb/204Pb and 208Pb/204Pb ratios remain constant while 206Pb/204Pb ratios increase with the degree of alteration. From fresh glass values of 6–8, similar to upper mantle values, 238U/204Pb (µ) values range up to a maximum of 183 in altered basalts as a result of secondary addition of seawater uranium. This uranium enrichment leads to significant radiogenic ingrowth of 206Pb that, with time, shifts altered basalts along horizontal arrays in correlation diagrams of 206Pb/204Pb vs. both 208Pb/204Pb and 207Pb/204Pb (Fig. F6).
None of the alteration studies indicate any difference in alteration style between Zone A and the AAD or between Indian- and Pacific-type lavas. This is, perhaps, not surprising from the viewpoint of geochemistry because the differences between Indian and Pacific types are subtle and confined for the most part to isotopic and trace element compositions. Of more interest is the implication that seafloor topographic style had no effect on alteration style. The contrast between the chaotic, predominantly extensional terrain of Zone B and the dominantly magmatic abyssal-hill terrain of Zone A (Christie et al., 1998) should be reflected in structure, porosity, and permeability as well as in the nature and distribution of high- and low-temperature seawater circulation, so the absence of discernible alteration effects is surprising. An alternate interpretation is that the topographic contrast within the AAD is stronger at present than it was during the 14- to 28-Ma period.
In terms of the alteration process itself, interesting contrasts have emerged between alteration-related studies that indicate no significant changes from 14 to 28 Ma and those that do record change. In the former category, Thorseth et al. (2003) reported that microbial activity and much of the mineralogical alteration of basaltic glass appears to have terminated before the 14-Ma age of the youngest drilled samples. A particular style of "granular alteration" of glass appears only to develop postburial and to be unrelated to microbial activity. S. Krolikowska-Ciaglo et al. (pers. comm., 2003) reported that isotopic and trace element exchange with seawater also effectively ceased before 14 Ma. In contrast, Shau et al. (this volume; see "Magnetic Properties" below) report that natural remanent magnetization (NRM) values for drilled basalts decrease progressively with increasing age and Miller and Kelley (this volume) report ongoing mineralogical and chemical evolution of altered basaltic rock (but not glass) samples. One possible explanation for these disparate observations is that seawater circulation had effectively ceased by 14 Ma. This would allow for chemical equilibrium to be established in terms of trace element exchange and in the processes of glass alteration, especially in those processes mediated or enhanced by microbial activity. It is a common observation in altered MORB samples that fresh glass fragments can be isolated by clay-rich alteration products and preserved for many millions of years. Perhaps the chemical gradients that lead to reaction at the active inner edges of the alteration rims can only be maintained while there is active circulation and alteration ceases once flow stagnates. In contrast, (partially) crystalline basalts appear to be thoroughly permeated by seawater and may take much longer to reach equilibrium once circulation ceases. These processes may also depend on diffusive or other low-level flow that continues as burial proceeds and after fracture permeability has closed. Ongoing mineralogical changes are also required to explain the progressive loss of magnetic remanence as the lavas age.
The reasons for the inferred cessation of crustal seawater circulation in this area are of interest because of the general absence of sediment cover (Christie, Pedersen, Miller, et al., 2001). Despite the apparent absence of sediment, our early attempts to dredge from seafloor older than ~7 Ma failed to recover rock samples but frequently recovered Fe-Mn oxide crusts. We speculate that the older seafloor in the region is sealed by such crusts.