Nineteen basement units have been defined in Hole 1139A, including an upper succession of felsic volcanic and volcaniclastic rocks (Units 1-4) and a series of subaerial lava flows (trachybasalt to trachyte, Units 5-19), many with brecciated flow tops (see "Igneous Petrology" and "Physical Volcanology"). Fluid/rock interaction has variably altered all basement units after emplacement, as indicated by secondary minerals that partly replace primary minerals, partly to completely replace mesostasis, and partly to completely fill veins, vesicles, and open spaces between clasts within breccias. We recorded the distribution of secondary minerals in the alteration and vein/structure logs (see the "Supplementary Materials" contents list) for Hole 1139A (Fig. F68). Shipboard time constraints precluded the detailed interpretation of structural data, but orientation measurements were made for the majority of features observed. In addition, we performed XRD analyses to identify secondary minerals (Table T12). These data show that rocks recovered from Hole 1139A display unique alteration patterns, relative to both the intensity of alteration and secondary mineralogy, when compared to other Leg 183 sites. In this section, we describe these unique alteration patterns and qualitatively estimate mass transfer of major elements associated with the most highly altered intervals.
The upper part of the basement section in Hole 1139A (Units 1-4) comprises highly altered and variably oxidized felsic volcaniclastic rocks. Despite the high degree of alteration, primary igneous features such as flow banding and perlitic textures are still visible. Brecciated textures are extremely common and appear to have diverse origins that include primary volcanic processes, faulting, reworking, and secondary hydrothermal processes. Cataclastic fabrics are particularly common, and slickensides are a ubiquitous feature on broken clay-rich surfaces. In general, breccia clasts include pumice altered to light green clay and angular felsic and mafic rocks altered to brown or dark red clay in a matrix of variably oxidized green to brown clay. Other brecciated intervals have red-brown clasts with thin (<2 mm) rims altered to light green clay within a red, variably oxidized matrix (Fig. F24). Alkali feldspar phenocrysts are common in clasts and are particularly abundant within the welded rhyolite composing Unit 4; some phenocrysts are remarkably fresh or only slightly altered. In addition, patchy silicification of clasts and matrix is common in many brecciated intervals (e.g., interval 183-1139A-52R-1, 104-150 cm). We also observed a hydrothermal breccia within Unit 1D (Fig. F18), which is more highly silicified than other brecciated intervals. The matrix is composed of a distinct yellow clay, and the clasts are variably altered felsic and mafic volcanics. Veins and fractures are rare in Units 1-4, although subvertical fractures occur with light brown alteration halos. We also noted rare clay-filled veins that crosscut red alteration bands, suggesting that oxidation predates the veining event (e.g., interval 183-1139A-55R-1, 18-71 cm).
The underlying trachybasalt through trachyte flows composing Units 5-19 (see "Igneous Petrology") are heavily brecciated as a result of volcanic flow-top processes and faulting. Most flows have only thin massive interiors and bases. As in all other Leg 183 holes, the brecciated flow tops (Fig. F36) are much more highly altered compared to the less permeable and more massive flow interiors (Fig. F69). Slickensides are common on broken surfaces, and most clasts are angular. The proportion of matrix and clasts varies widely among different breccias, from ~20% to 80% clasts. Breccia clasts are variably oxidized and vary from black to brick red to pinkish gray; some intervals have clasts with light green rims (e.g., Sample 183-1139A-61R-1 [Pieces 1A-1C, 0-43 cm]), similar to those in Unit 4. In other cases, breccia clasts are completely altered to green clay, suggesting a locally progressive alteration to green clay following oxidation. The matrix of these flow-top breccias is variably oxidized and consists of small lithic clasts replaced by secondary minerals that also fill open spaces between clasts. Red-brown and light green saponite are common in the matrix (Table T12), although red-brown clay is more abundant, even in cases where breccia clasts have been altered to light green clay (e.g., Sample 183-1139A-61R-3 [Piece 1, 0-12 cm]). Calcite and siderite are common secondary minerals in the matrix of breccias, with some intervals composed of ~30% carbonate (e.g., Sample 183-1139A-66R-4 [Piece 1, 0-101 cm]). Amorphous silica and quartz are rarely present.
The thin and more massive flow interiors of Units 5-17 are generally moderately altered. Color varies from gray to green to red, reflecting variable oxidation and different alteration processes in groundmass and phenocrysts. Dark green clays replace and accentuate the groundmass and mesostasis, whereas mafic and feldspar phenocrysts are variably altered to red-brown and green clay, respectively. Carbonate is absent from Unit 5 but is a common secondary mineral in the remainder of the basement of Hole 1139A. Rarely, these basement units have a pale gray hue, and the groundmass is bleached because of the replacement of primary igneous minerals by secondary calcite and siderite (e.g., Sample 183-1139A-64R-2 [Pieces 1-5, 0-143 cm]).
Vesicles generally compose <5% of Units 5-17, although in rare cases, vesicularity approaches 20%. Vesicles are completely filled with a secondary mineral assemblage that includes green clay, siderite, calcite, and, rarely, amorphous silica. Light brown siderite, white calcite, and, more rarely, green clay occur as alternating semicircular bands within vesicles forming a colloform texture. Calcite typically fills the interior of these same vesicles (Fig. F69). Geopetal structures are rare.
Steeply to moderately dipping veins and fractures generally much less than 4 mm wide are common within Units 5-17. Calcite, siderite, and brown and green clay are the most abundant minerals filling veins or partially lining fractures. Green to red oxidation halos as much as 2 mm wide surround veins and fractures. We also observed multiple crosscutting, approximately orthogonal sets of calcite veins in some intervals that locally offset siderite-lined vesicles (Fig. F69). These relationships suggest early siderite formation followed by at least two calcite-forming events. The appearance of siderite as a secondary mineral cementing clasts within breccias and partly filling veins and vesicles is one characteristic that makes Hole 1139A unique when compared to other Leg 183 sites.
Alteration increases abruptly within Units 18 and 19 and is characterized by two processes that have affected these lowermost flow units to different degrees: (1) oxidation characterized by red staining and (2) pervasive replacement of groundmass and phenocrysts by quartz and carbonate minerals, including siderite and calcite. Both types of alteration are most intense along veins, but many intervals appear to have been pervasively altered in the absence of permeability provided by fractures or vesicles. Based on visual core descriptions, rocks that have been only moderately changed are oxidized red gray and probably represent relatively unaltered trachytes in this part of the basement. In the most intensely altered intervals, the rocks are white with some pale pink patches and have a gritty or sandy texture. Our XRD and thin-section analyses of these intervals reveal that the groundmass and sanidine phenocrysts have been variably to completely altered to siderite and microcrystalline quartz with minor clay (Table T12). Based on our observations of crosscutting relationships, we believe that both styles of alteration have been superimposed on these lowermost basement units at different times. Similar processes probably affected Units 5-17 as well, although to a lesser extent.
The top of Unit 18 is a brecciated sanidine-phyric trachyandesite (Fig. F70). Breccia clasts have diffuse margins and are variably flattened with possible welded textures (see "Physical Volcanology"). The color is generally red brown to gray brown, suggesting variable degrees of oxidation and differing amounts of siderite. These relatively fresh rocks gradually become more silicified and more oxidized with depth, as indicated by a change in color to very light gray and the appearance of local 2-mm streaks of brick red (hematite?) alteration (e.g., Sample 183-1139A-70R-2 [Pieces 1A-1D, 30-68 cm]). The intensity of alteration increases abruptly in Section 183-1139A-70R-2 at 68 cm with the first appearance of pale green to white trachyandesite that has been replaced by quartz and siderite (Fig. F71). Sanidine phenocrysts are variably replaced by these same phases, and some fresh phenocrysts are present, even in the most altered rocks. In general, the most intense alteration (white to pale pink) is present as 1- to 2-cm halos around veins partly filled with quartz, siderite, and calcite. We observed this style of alteration to the bottom of Section 183-1139A-70R-4. Within this depth range, completely altered white to light pink trachyandesite is moderately vesicular with siderite- and quartz-filling vesicles (Fig. F72). Also, brick red oxidation halos 1-2 mm wide appear along veins that crosscut the intense white to pale green alteration zones and small patches of relatively fresh red-gray trachyandesite (Fig. F73). These relationships suggest a relatively intense oxidation event after the alteration that replaced the trachyandesite with quartz and siderite.
The remainder of Unit 18 (Sections 183-1139A-70R-5 through 71R-4) exhibits complex paragenetic relationships among different alteration and oxidation events of varying intensity. In general, the most intense white to pale green to pale pink alteration facies composes <10% of the rock over this interval and is completely absent in some sections. Very light gray to yellow-gray variably altered (silicified?) trachyandesite is more abundant within this part of Unit 18 (as much as 30%) and locally replaces the relatively fresh and moderately oxidized red-gray trachyandesite. We interpret these alteration zones to represent somewhat less intense examples of the white to pale green to pale pink horizons in the upper part of Unit 18, although this requires confirmation with more extensive chemical analyses of the bulk rock. Typically, these very light gray to yellow-gray zones are crosscut by prominent red alteration halos around veins filled with hematite, quartz, siderite, and calcite (Figs. F74, F75). Irregular cavities filled with these same secondary minerals also are present (Fig. F76). Intense oxidation accentuates primary igneous textures, particularly the mesostasis. We also noted narrow white to pale pink alteration halos around veins filled with quartz, siderite, and calcite (Fig. F77). These veins crosscut and locally offset dark red oxidation bands and light gray silicified(?) trachyandesite, consistent with an oxidation event before replacement of wall rock with quartz and siderite. Thus, our observations are consistent with a complex alteration history characterized by multiple oxidation events occurring both before and after multiple silicification(?) or silica-mobilization events and replacement of the trachyandesite by quartz and siderite.
The upper part of Unit 19 (Sections 183-1139A-71R-4 [Piece 1B, 19 cm] through 71R-6, Piece 1D) is highly brecciated and variable in color, reflecting differing degrees of oxidation of both matrix and clasts. The brecciated intervals are uniformly clast dominated with clasts representing 75%-90% of the rock. Clasts and matrix are very light greenish gray near the top of the unit (Fig. F78) and gradually become more oxidized with depth. Within some intervals (e.g., interval 183-1139A-71R-5, 0-45 cm), clasts are red with light green rims, similar to the alteration patterns we observed in basement units at much shallower depths in Hole 1139A (e.g., Unit 5). At deeper intervals, color changes gradually with clasts becoming very pale brown (Fig. F79) and finally very pale green to light red (Fig. F80) within a red oxidized matrix. Thin-section and XRD analyses of these different breccias reveal no significant change in secondary mineralogy despite the color variations. Quartz is a common phase replacing groundmass in clasts and partly replacing sanidine phenocrysts. Relatively small amounts of clay (illite) were also identified in thin section and by XRD (Table T12). It is noteworthy that XRD analyses of the breccia clasts did not reveal any significant carbonate minerals; however, moderate amounts of carbonate-replacing groundmass and phenocrysts could be identified in thin section. This may be a result of sampling heterogeneous breccias that have been less pervasively altered to carbonate minerals. Thus, carbonates may be less abundant in some parts of Unit 19. Silicification (or silica mobilization), however, is just as pervasive as that observed in the completely altered white to pale pink alteration zones in Unit 18.
Unit 19 is more massive from Section 183-1139A-71R-7 to the bottom of the hole in Section 183-1139A-73R-3. The color of the flow varies from red brown to gray, with gray intervals representing some of the freshest rocks encountered in Hole 1139A (e.g., interval 1139A-72R-2, 0-85 cm). Sanidine phenocrysts are variably altered and rarely replaced by calcite. Minor carbonate is also found in glassy vesicular areas, and the groundmass is locally mottled and possibly silicified. Otherwise, vesicles are absent. Variable oxidation of the trachyte gives the rock a red-banded appearance (Fig. F81), and both the mesostasis and mafic phenocrysts with acicular textures are locally oxidized. Veins and fractures are numerous in Unit 19, are generally <2 mm wide, and are filled with brown clays, calcite, white clay(?), and rare siderite. White alteration halos, as wide as 10 mm, are common around veins filled with brown clay and calcite. We also observed numerous slickensides on fractured surfaces lined with clay.
Downhole variations in CO2 and H2O concentrations of basement units (see "Organic and Inorganic Geochemistry" and "Igneous Petrology") indicate that alteration in the bottom part of the basement in Hole 1139A is not associated with extensive hydration of volcanic protoliths but is instead characterized by carbonate metasomatism. Indeed, thin-section and XRD analyses do not reveal any abundant hydrous secondary phases, but siderite is ubiquitous. Thus, alteration was likely produced by the interaction of relatively siliceous volcanic rocks with hydrothermal solutions that have unusual compositions compared to those normally associated with alteration of typical oceanic crust. Moreover, it is not clear whether alteration occurred in submarine or subaerial environments.
We have compared the chemical compositions of the most highly altered "bleached" rocks in Units 18 and 19 with the least-altered compositions to qualitatively determine the extent of mass transfer during alteration (Fig. F82). For this analysis, we used the chemical compositions of highly altered rock close to relatively fresh rock (based on visual examination of the core) with similar TiO2 concentrations. Titanium is assumed to be immobile during alteration, and rock pairs with similar TiO2 concentrations should have had similar compositions before alteration. The results demonstrate that alteration is accompanied by the removal of alkali and alkaline earth elements and aluminum, presumably related to replacement of glass, groundmass, and sanidine phenocrysts by secondary minerals. Surprisingly, there is no net gain in silica. In fact, silica may have decreased slightly during alteration of Unit 19, despite clearly visible quartz partly replacing groundmass and sanidine phenocrysts in thin section. This suggests that silica may simply be remobilized from relatively siliceous protoliths during alteration. Water exhibits a net loss from Unit 18 during alteration but a prominent net gain in the bleached horizon from Unit 19. This is consistent with the presence of illite or possibly some other clay minerals in Unit 19, as determined from thin-section and XRD analyses (Table T12). In contrast, CO2 shows large net mass gains in the bleached horizons from both Units 18 and 19. Total iron appears to have been added to the rock during alteration of Unit 19, whereas the bleached horizon in Unit 18 shows no significant increase when compared to the protolith composition. Manganese shows the opposite behavior, with a net gain in Unit 18 and no change in Unit 19. Although iron-rich solutions were likely involved, iron may be locally transferred from silicate minerals into siderite during interaction with carbonate-rich solutions. Clearly, a more quantitative analysis of this type requires trace element and isotope data to better constrain protolith composition, the temperature of alteration, and the provenance of hydrothermal fluids. This will be the subject of postcruise research.