Previous   |   Next

doi:10.2204/iodp.proc.324.103.2010

Igneous petrology

A key objective of Expedition 324 was to collect information regarding the nature of "basement" volcanic rocks on Shatsky Rise in the northwest Pacific. This basement was previously cored during Leg 198 at Site 1213 on the southern high (Tamu Massif) and revealed a series of fine-grained massive volcanic flow units, each several meters thick (Shipboard Scientific Party, 2002b).

Hole U1346A was drilled on Shirshov Massif, northeast of Tamu Massif, and established that a stack of highly vesicular basaltic pillow lavas or lava "inflation units" (stratigraphic Unit V) occurs beneath a succession of pelagic nannofossil-bearing chalks and cherts (stratigraphic Unit I), volcanogenic silts and sands, and larger volcaniclastic debris materials (stratigraphic Unit II). The latter volcaniclastic deposit occurs interbedded with sediment and is interpreted as a solidified "blocky" volcanogenic debris flow (stratigraphic Unit II). In this unit soft-sediment deformation occurs around the larger volcanic clasts indicating that the clasts impacted the sediment, either rolling downslope from its eruption sources or as a mass flow deposit generated through posteruptive erosion of the volcanic edifice. The presence of a deposit of volcanogenic sand and turbidites (stratigraphic Unit III) supports this latter interpretation. Within the lava stack, individual pillow (or inflation) units were readily identified by the presence of chilled glassy margins, upper and lower chill zones, characteristic pillow vesicle patterns, and crystal grain size variations. In Unit V, 40 individual "lava cooling units" were recognized in the retrieved core and together are interpreted to represent a single eruptive event.

The basalts from Unit IV appear aphyric macroscopically, but a closer inspection reveals that, where originally present as microphenocrysts, olivine and pyroxene are now totally replaced by calcite. All samples contain large proportions of less altered, very fine grained plagioclase laths set in a variolitic matrix. The pillow basalts are generally vesicular in nature and have zones that are moderately vesicular (30%–50% vesicles). Alteration has left a marked impression on the igneous rocks, is highly variable, and in some places is pervasive, ranging from less altered dark gray basalts in pillow unit cores to near completely altered and oxidized brown vesicular basalts. Green and brown clays, together with pyrite and iron oxyhydroxides, are the main secondary phases of the highly altered basalts and replace the primary phases as well as the entire glassy mesostasis. Minor pyrite and calcite are also observed associated with the groundmass clays, as well as in veins, especially in the upper 10 m of the basalt.

Stratigraphic unit division

In this site report, larger "stratigraphic units" are given Roman numbering (stratigraphic Units I–V), whereas small-scale "lithologic units" are identified by Arabic numbering (lithologic Units 1–58). We identified lithologic units on the basis of lithology changes, volcanological features (e.g., presence of contacts, chilled margins, and flow tops), changes in primary mineralogy (e.g., abundance of plagioclase, pyroxene, olivine, and oxide minerals), vesicle distribution, color, grain size, structure, alteration, and the presence of volcaniclastic or sedimentary interbeds. In the case of volcanic rocks, these lithologic units typically reflect different volcanic inflation or cooling units. Stratigraphic units, however, may combine these lithologic units into a single volcanic succession, or stack, of inflation units (see "Igneous petrology" in the "Methods" chapter for more details).

Hole U1346A was drilled on Shirshov Massif in the northwest of Shatsky Rise (Fig. F1). Igneous basement was reached at 139.2 mbsf, marking the first instance that volcanic basement samples have been sampled on Shirshov Massif. The cored succession consists of 69.7 m of volcaniclastic-rich sediment, volcaniclastics, and pillow basalt units preserved below a poorly recovered sedimentation succession (stratigraphic Unit I) consisting of chalk and chert (Fig. F10).

Volcaniclastic material was first encountered in Sections 324-U1346A-4R-1 and 4R-2 (119.5–122.0 mbsf) and comprises intercalations of fragmental basaltic debris and sediment containing a moderate to high proportion of crystalline and lithic igneous components. Fluidal mixing of sediment and basaltic clast material at specific horizons is indicative of active interaction between unconsolidated sediment and, possibly, still-plastic lava. Below this volcaniclastic debris unit, Core 324-U1346A-5R contains a finely laminated volcaniclastic turbidite sequence (stratigraphic Unit III) underlain by a section of bedded, sand-sized carbonate sediments, which continues to the top of Section 324-U1346A-6R-1 (stratigraphic Unit IV). Below 139.2 mbsf only igneous material was recovered (Sections 324-U1346A-6R-1 through 16R-1), dominated by highly vesicular aphyric micro- to cryptocrystalline basaltic eruptive pillows and lobes that have been pervasively altered (stratigraphic Unit V). This final unit we considered "true" Shirshov Massif basement.

Chalk and chert (stratigraphic Unit I, lithologic Units 1 and 2)

The first lithologic unit recovered in Hole U1346A was soft gray chalk interbedded with remarkably resistant layers of black and brown chert. The largest fragment recovered was 0.16 m of chert core, with only a very low recovery of the host carbonate. See "Sedimentology" for a more detailed description of this unit.

Volcaniclastic debris (stratigraphic Unit II, lithologic Units 3–7)

Unit II consists of vesicular basalt in the form of lava flow fragments with characteristic green-blue alteration (see "Alteration and metamorphic petrology" for more details) intercalated and intermingled with hemipelagic limestones containing moderate to high proportions of crystalline and lithic igneous components (Fig. F11). Within this 1.6 m thick basalt-limestone sequence, the nature of the contact between these two distinct lithologies is that of disturbed sedimentary structures indicative of fluidal mixing of soft sediment and basaltic clast material at specific horizons (Figs. F12, F13). Primary sedimentary layering within the unconsolidated sediment (carbonate mudstones) was distorted and largely obliterated by the arrival of the clasts. Moreover, detailed inspection of the clast margins indicates that the clasts were solidified prior to their introduction into the soft seafloor sediment. These relationships, therefore, are consistent with volcanic debris having tumbled from a nearby source; a scarp-base talus, for instance.

Volcaniclastic turbidites (stratigraphic Unit III, lithologic Units 8–10)

Below the volcaniclastic debris of Unit II, a well-preserved 0.85 m section of finely bedded volcaniclastic material (clay to fine sandstone grain size) contains numerous small-scale fining-upward laminae (2–5 mm) superimposed upon two broader fining-upward trends (15–25 cm). The presence of flame structures and microscours and an absence of bioturbation (with the exception of one layer) are the basis for interpreting this part of the unit as a normally graded turbidite (Section 324-U1346A-4R-2). These turbidites contain significant concentrations of materials of volcanic origin, including lithic clasts, individual crystal fragments (plagioclase), glass debris (pyroclasts), and chemically immature silt- to sand-sized particles. The small-scale fining-upward laminae themselves consist of concentrations of glassy vesicle shards, delicately preserving their intricate cuspate forms and providing evidence that transport and reworking was minimal prior to their deposition. Unlike basaltic glass, which has a brown tinge, these glass shards are virtually colorless and may be of different provenance to that of the basaltic volcanism in Units II and V. The bulk of the larger volcanic components are recognizable as small lithic fragments of highly vesicular basalt, very similar to the basalts found in Unit II. These clastic fragments are typically angular in shape and often retain their pristine vesicle structures, again indicating relatively minor particle transport or reworking. Accordingly, the volcanic source for the input in this turbidite sequence must have been co-located or else was located in close proximity. We classified Unit III as a "volcaniclastic turbidite," as the fragmental aggregate contains >60% volcaniclastic grains (see "Igneous petrology" in the "Methods" chapter). See "Sedimentology" for a more detailed description of this unit.

Shallow carbonates (stratigraphic Unit IV, lithologic Units 11–18)

Unit IV consists of an estimated 17.1 m thick sediment series with clay-rich volcanogenic sandstones and siltstones, calcareous mudstones, and bioclastic limestone beds. This carbonate-rich sequence of silt- and sandstone contains both authigenic glauconite and volcanogenic components. The volcanogenic component is similar in character to, but less abundant than, the volcanic debris found in the younger finer grained turbidite series of Unit III. The presence of closely packed shelly (possibly inoceramid) fragments, together with shallow-water benthic foraminifers, places strong constraints on the depth of sedimentation in Unit IV, most likely not exceeding ~200 mbsl water depth. See "Sedimentology" for a more detailed description.

Pillow basalts (stratigraphic Unit V, lithologic Units 19–58)

Unit V is 50.1 m thick and represents the top of the volcanic basement of Shatsky Rise as drilled at Shirshov Massif. This unit is composed of a stack of lava inflation units, including basalt pillow lavas and lobes that in the core range considerably in thickness from 0.3 to ~4 m. These pillows and lobes have similar characteristics, with inner cores that are more massive, slightly coarser in grain size, and nonvesicular, with outer zones that are highly vesicular (as much as 50%), and with rims that frequently have chilled glassy margins or pillow contact zones (Figs. F14, F15). All these inflation units are interpreted to be part of a single eruption pillow stack predominantly containing aphyric micro- to cryptocrystalline basalts. Alteration is pervasive throughout these pillows, with multiple generations of alteration almost completely filling cracks and vesicles, making these basalts strongly amygdaloidal and replacing most (if not all) of the volcanic glass (30%–90%), pyroxene, and olivine in these rocks. The final phase of alteration was oxidative and is the most obvious because it transforms intervals of originally dark bluish gray basalts into zones of red-brown basalt. See "Alteration and metamorphic petrology" for a more detailed description of the various alteration regimes observed in this unit.

Macroscopic description

Volcaniclastic debris (stratigraphic Unit II)

As discussed in "Igneous petrology" in the "Methods" chapter, volcaniclastic material includes a range of materials from rubbly in situ volcanic debris to resedimented materials such as volcanic sands or tuffs (see Figs. F5, F6 in the "Methods" chapter). Volcanic materials of all sizes may be the direct product of eruptive processes (pyroclastic) or may be accumulations through processes involving transport, sorting, and redeposition (epiclastic). Pyroclastic activity includes hydrovolcanic deposits formed by explosive interaction between magma and water, as well as quench fragmentation (e.g., hyaloclastite and peperite). Peperite is a distinct volcanic-sediment facies occurring where submarine basaltic lava flows interact with unconsolidated sediment as they erupt into water bodies or, more commonly, on the seafloor. The mingling of still-molten lava flows and wet sediments produces distinct volcanic textures resulting from the physical interaction between lava and sediment (such as entrainment, baking, and chilling), as well as physical and chemical fragmentation and alteration through steam-rock interaction during flash heating of seawater. The term "peperite" is used particularly in those cases where lava-sediment mingling generated soft-sediment deformation features and when these interactions resulted in quenched margins, plastic deformation, or the in situ "jigsaw-fit" cracking of the volcanic clasts involved. However, because these characteristic peperite features could not readily be identified in the retrieved core and therefore require shore-based petrographic studies for confirmation, we classified the volcanic materials in Unit II simply as rubble or debris flows.

Unit II (lithologic Units 3–7 in Figs. F10, F11) is divided into five intervals in which volcaniclastic material is an important component (see Sections 4R-1 and 4R-2 in "Core 324-U1346A-4R" in "Visual core descriptions" in "Core descriptions"). Three of these five intervals (Units 3, 5, and 7) for the most part consist of light green-blue, highly vesicular basalt mixed with sparsely distributed angular fragments of sedimentary material. The other two intervals (Units 4 and 6) consist of the same materials, but they are dominated by highly disturbed greenish to light brown carbonate mudstones (Figs. F12, F13) with only a few angular clasts (all >7 cm) of greenish vesicular basalt intermixed, in a few places showing soft-sediment deformation at or near the basalt/sediment contacts (intervals 324-U1346A-4R-1, 28–32 cm, and 75–90 cm). The nature of the contact relations between these two distinct lithologies indicates that the basaltic clasts solidified prior to their introduction into the soft seafloor sediment, consistent with debris introduced from a nearby volcanic source.

The sparsely to moderately vesicular basalt has a cryptocrystalline to microcrystalline groundmass. Approximately half of the highly spherical vesicles are filled with secondary minerals (predominantly calcite), and veins within the basalts are filled with dark green clays. Chilled margins are not present in the basaltic clasts, and typically plagioclase crystals and vesicles are broken along the margins of these clasts. The clayey layers in the limestone matrix are deformed parallel to the margins of the basaltic clasts, and sediment in a few cases fills vesicles and small voids in the vesicular basalt clasts.

Pillow lava and inflation pods (stratigraphic Unit V)

Many mafic lava units develop through endogenous growth or "inflation," for example, pahoehoe, massive flood basalt units, and submarine pillow lavas (Self et al., 1998). Pillow lava units (see Figs. F14, F15 for schematic drawings) are a typical product of basaltic eruption in submarine conditions (Batiza and White, 2000). Basaltic lava is internally transported within propagating flows (or within lava tubes) to active flow fronts where the lava flow advances by creating and inflating a lobe that typically has a continuous crust (Walker, 1991; Hon et al., 1994). On contact with water, these lobes chill rapidly, forming a glassy outer crust (Figs. F14, F15C, F16) that entraps gas vesicles that continue to form and grow within the remaining molten core. The surface contraction and stretching and fracturing of the solidifying outer skin provides the first places where vesicles accumulate and get trapped, leading to small radially arranged vesicle trains immediately inward of the glassy chilled pillow margin (Figs. F15C, F16). Within the still-molten core of the lava, continued cooling causes the gases to further exsolve, producing more vesicles, which move toward the surface but become frozen at the base of the solidification front that moves inward as the outer part of the pillow keeps cooling. This is the origin of the concentric zones of vesicles that are distinctive of the internal structure of many pillow lavas (Fig. F15A). Large vertical pipe vesicles as long as a couple of centimeters may form in the cores of larger pillow units (Fig. F15B) when stagnation of lava and slow cooling subsequently allow these larger vesicle trains to develop. Given sufficient time, the largest vesicles within these cores will slowly evacuate toward the upper region of the pillows, resulting in near-nonvesicular pillow cores in the larger inflation units. Distinctive crystal grain size changes can also be observed from pillow margin to core because the development of the outer (insulating) crust arrests any further rapid chilling of the lava within, allowing a coarser, sometimes wholly crystalline internal fabric to develop. Accordingly, the degree of interstitial glass diminishes markedly away from the edges of a pillow unit, increasing in crystallinity inward. During emplacement, glassy chilled zones of pillow units typically spall, accumulating in the interstices between neighboring pillow units (see lithologic Units 55 and 57 for good examples). These glassy outer layers and spalled interpillow hyaloclastites are highly susceptible to alteration and erosion, explaining their relatively poor preservation and recovery during drilling, excepting those instances where the material has been incorporated into zones of secondary cementation.

Pillow inflation units

All pillow characteristics described above were observed in basaltic basement Unit V recovered at Site U1346. In cases where chilled margins, glassy rinds, or pillow contacts (Fig. F15C, F16) were observed, the interval in the cored material could be confidently described as a separate pillow unit (Fig. F11). However, more commonly, the position of contacts between two pillow units had to be inferred using a combination of vesicle characteristics, changes in grain size and texture, and/or the presence of glassy material (see "Interpillow volcaniclastics" for more details). In instances where a pillow unit could not be unequivocally determined, the generic term "lava inflation unit" was used.

The largest inflation units range in diameter between 1 and ~4 m. The larger units typically contain massive nonvesicular cores with sometimes two or three isolated, centimeter-long, vertical vesicle trains. The smaller units are only ~0.3–1 m in diameter and were typically more readily identifiable as pillow lava, especially when preserved in entirety within two or three, or even in single core pieces (Figs. F14, F15A). These smaller pillows generally have thick (10–30 mm) glassy margins and vesicle patterns typical of submarine lava flows, including outward radiating and banded inward concentric vesicle distributions. Even though recovery of the external parts of pillow units (i.e., glassy rind, chilled zone, and adjacent vesicle-rich region) typically is poor, good examples were encountered in Sections 324-U1346A-8R-1, 13R-1, and 14R-1 and two excellent examples occur in Section 16R-2 (lithologic Units 52 and 56). By contrast, larger fragments or even entire sections of the massive core regions of larger inflation units are well represented in the material recovered in Sections 324-U1346A-9R-2 through 12R-2 (lithologic Units 32–37).

In instances where the outer regions of inflation units were preserved, aphanitic zones (Figs. F14, F15C, F16) were found immediately inside the glassy margins. Pillow interiors often display variolitic texture imparted by acicular plagioclase microcrysts with interstitial clinopyroxene and lesser titanomagnetite and only a modest (0–0.25 mm) variation from crypto- to microcrystalline texture (see groundmass grain size in Fig. F10) toward the cores of the larger pillow units.

General observation of the recovered cores indicates fundamental changes in the degree and type of alteration. The topmost Cores 324-U1346A-6R and 7R display bluish gray color derived partly from primary mineralogy and partly from the replacement of igneous minerals by dark clay-type alteration products. In Cores 324-U1346A-8R through 15R the basalts are pervasively altered to a brownish alteration color, with only the inner cores of the larger inflation units retaining the bluish gray color typical of the alteration in the uppermost units. In these brown oxidative alteration zones, the small pillow units and outer glassy or vesiculated zones are always altered in their entirety. The lowermost Cores 324-U1346A-15R and 16R are characterized by a return to bluish gray altered basalts. In Unit V differences in color therefore reflect the variation in the style of alteration through the pillow stack (for details see "Alteration and metamorphic petrology").

The most remarkable feature of these pillow units is their unusually high degree of vesicularity. Their formation is probably significantly augmented by steam, reflecting shallow-water conditions during eruption. The smallest recognizable inflation unit within the succession of Unit V consists of a single pod or pillow of basalt that often has upper and lower chill zones and glassy contacts. Internally these pods frequently have characteristic vesicle patterns that allow them to be distinguished from one another as well (or from broken up intervals in core when recovery is poor). Because the pervasive alteration makes it very difficult to impossible to observe any petrologic break within Unit V (see "Petrography"), we concluded that these individual lava units are themselves stacked (grouped) together into one "lava package" representing a short eruptive period of continuous build-up of lava units. As no interbedded sediment was recovered, the succession of inflation units in Hole U1346A likely represents a continuous "single eruptive event" (Chenet et al., 2009) during which successive pods or pillows were produced and stacked in a single package one upon another (Table T5).

Vesicle distribution patterns

Systematic variations in the sphericity, roundness, abundance, and size of vesicles occur from the margins to the interiors of the pillow inflation units. These types of vesicle variations (Fig. F14) are particularly well developed in Section 324-U1346A-14R-1 (lithologic Units 40 and 42). A series of distinct vesicle types and patterns have been recognized these may be summarized as follows:

• Type 1: radiating vesicle trains. At the outermost margins small (2–10 mm) elongated vesicle trains radiate outward, truncating at the glassy and cryptocrystalline exterior. These vesicle trains appear to be concentrated along incipient fractures produced during the initial cooling contraction of the chilled "outer skin" of the erupted magma pod (Figs. F14, F15C).

• Type 2: concentric vesicle zones. Inward of the Type 1 vesicles is a zone of high-sphericity, rounded to subrounded vesicles. These are concentrated into a series of 2 to 5 concentric bands resulting in a wide highly vesiculated zone, the thickness of which depends on the size of the inflation unit (Figs. F14, F15A, F15C). The concentric banding most probably records the inward progression of the solid/liquid interface during solidification of the pillow unit and generally is more intensely developed in the upper parts of the larger inflation units.

• Type 3: gas blisters. Coalescence of small (0.5–2 mm) vesicles is common in the innermost part of the banded zone of Type 2 vesicles, and in some instances this coalescence results in larger (30–50 mm) "blister-" type vesicles (Fig. F15A). These oversized vesicles can have an irregular and ragged internal morphology and typically form in the later stages of solidification when the remaining magma was becoming too viscous to allow significant vesicle movement.

• Type 4: pipe vesicles. Vertically and subvertically oriented trains of (subrounded) elongate vesicles occur in the core regions of some larger inflation units (Fig. F15B). These vesicles record the stagnation of the magma during its cooling and solidification once the unit was emplaced.

The remainder of the core is typically nonvesicular or else contains only a sparsely uniform distribution of very small microvesicles (<0.2 mm).

Interpillow volcaniclastics

In the pillow stack of Unit V some highly altered intervals of carbonate-cemented fragmentary volcanogenic material occur. In the drill core these were typically recovered as individual isolated pieces, but they invariably occurred at the junction between two or more successive inflation units. These are interpreted as pockets of glass-shard material that accumulated in the interstices of the newly formed lava stack as the result of surface spallation from nascent pillow units. Accumulations and pockets of these kinds of materials are also reported in onshore studies of LIPs containing similar stacks of pillow lavas, as for instance preserved in accreted terrains (Greene et al., 2010). In Hole U1346A profound recrystallization makes determination of the morphology and size of the fragments difficult, but the larger volcaniclastic fragments appear to be 3–5 cm in length. As seawater and hydrothermal fluids have passed pervasively through the volcanic basement at this site, in particular using the interpillow spaces as conduits, much of these void spaces now is occupied by calcite cement. Good examples of this interpillow volcaniclastic material occur in Sections 324-U1346A-14R-1 (lithologic Units 31 and 44) and 8R-1 (lithologic Unit 30).

Petrography

Coring in Hole U1346A penetrated >52 m of volcanic rock, most of it in a sequence of particularly vesicular and strongly altered pillow lava units. The primary vesicularity elevated the porosity of the rock and contributed to its ease of fracturing, allowing seawater and hydrothermal fluids to penetrate and pervasively alter the volcanic basement. As a result, the basement rocks are strongly overprinted by the effects of these fluids. Patterns of alteration are readily apparent in the strongly contrasting green, brown, and (ostensibly fresher) gray coloration of the rock (see "Alteration and metamorphic petrology").

This pervasive attenuation prevents a full and clear description of the primary igneous petrography and mineralogy of the basalt, at least for the present. Most of the rock consists of 60%–80% secondary minerals, dominated by clays, Fe oxyhydroxides, calcite, and pyrite. Calcite alone is so pervasive a replacement mineral that judging the original extent of occurrence of olivine and even clinopyroxene from thin sections is difficult in many of the pillow basalt specimens in Unit V. Distinctions between aphyric rocks and those containing (remnants of) microphenocrysts therefore cannot easily be made. Even secondary calcite in the veins and vesicles of some portions of the core has been substantially recrystallized and stained brown with Fe oxyhydroxides in a later alteration phase.

Most of the rocks appear aphyric because no phenocrysts can be readily discerned in hand specimens. However, attention to the chilled margins of pillows gives the impression that they originally contained small (<1 mm) euhedral microphenocrysts of plagioclase and olivine. Nevertheless, some of the rocks are porphyritic because they contain a few larger phenocrysts (>1 mm) or glomerocrysts (of mostly plagioclase), as shown in Table T6. Groundmass textures reveal no obvious differences in the size and crystal morphologies of marginal spherulites or acicular pillow interiors, which might suggest that chemical differences occur within the pillow sequence of Unit V. However, the volcanic clasts found in the shallow sediment of Unit II may represent a slightly different composition since they are more strongly plagioclase-phyric when compared to the pillow basalts from Unit V.

Glassy rinds are present on a number of the smaller pillow units, although they are typically completely altered. In rare instances, where glass is preserved in small amounts, it offers a starting point for developing a crystallization history for these basalts. For instance, fresh glass was found in two small fragments, each ~0.5 cm wide in a pillow chilled margin in Thin Section 28 (Sample 324-U1346A-9R-1, 56–60 cm) (Fig. F17). In this thin section, the fresh glass is partly surrounded by an orange palagonite rim and contains tiny crystals of plagioclase with branching or dendritic morphologies and long slender needles of larger plagioclase crystals. Adjacent to one glass fragment, an altered olivine microphenocryst containing small, dark brown Cr spinel is partly intergrown with, and also plated by, small acicular plagioclases. Although no other glass was found in the 60 thin sections studied, spherulitic margins elsewhere in the core appear to have the same mineralogy and crystal morphologies. The rocks in general, therefore, may be described as a very sparsely microphyric olivine-plagioclase basalt with accessory spinel.

Volcaniclastic debris (stratigraphic Unit II)

The basaltic clasts of Unit II are generally sparsely to moderately plagioclase phyric (as much as ~5%). All basaltic clasts are vesicular (5%–20% of vesicles) with the vesicles completely filled by secondary minerals, dominated by calcite (see "Alteration and metamorphic petrology"). Plagioclase phenocrysts and microphenocrysts (0.5–1.2 mm) in this unit are typically subhedral to euhedral (Figs. F18, F19, F20). Groundmass microlites range from microcrystalline (0.1–0.2 mm) to cryptocrystalline (<0.1 mm), exhibit spherulitic and intersertal textures (Fig. F20), and consist of plagioclase (~20%) that typically is subhedral to euhedral and acicular (Fig. F19). Opaque minerals are minor (2%–7%).

Pillow lava and inflation pods (stratigraphic Unit V)

Since the rocks are pillow lavas, considerable variation in cooling rate and crystallization occurred between pillow rim and pillow interior. Thin section examination confirms that traverses from the pillow margin to interior show textural gradations from glassy pillow rims to a microcrystalline margin and an innermost very fine grained crystalline interior. Degrees of alteration depend strongly on proximity to veins and sometimes vesicles (see "Alteration and metamorphic petrology"). All pillow basalts show some level of vesicularity, ranging from sparsely vesicular in core regions to highly vesicular at the outer margins (~50%).

The pillow basalts of Unit V are aphyric (Fig. F18), but they also show sporadic traces of completely altered olivine microphenocrysts (0.1–0.4 mm) pseudomorphed to calcite (Fig. F21). Clinopyroxene has been altered to a similar degree, whereas plagioclase is significantly less affected. It seems that in the majority of cases that clinopyroxene is likely to have constituted a significant proportion of the groundmass mineralogy. However, these minerals now are virtually completely replaced by calcite, and only traces of strongly altered titanomagnetite can be seen in either transmitted or reflected light. In some instances large clinopyroxene crystals (0.2–1.0 mm) are still somewhat intact and subophitically enclose plagioclase (Fig. F22), which, from the pillow margin inward, transition into dense networks of interlocking acicular plagioclase (Fig. F20A–F20C). These tabular plagioclase microcrystals also may plate the margins of pseudomorphed olivine that often contains small brown Cr spinel and melt inclusions (Figs. F20D, F23).

Groundmass in pillow rims is typically glassy or cryptocrystalline (<0.1 mm) and shows spherulitic or hyalophitic texture. On the other hand, groundmass in pillow interiors is microcrystalline (0.1–0.2 mm) or very fine grained (0.2–0.5 mm) and typically exhibits spherulitic and intersertal texture (Fig. F19). Plagioclase microlites are the most common and typically subhedral to euhedral, often in a markedly acicular texture (Fig. F20). Although plagioclase is the least altered mineral in these pillow basalts, alteration is often significant and has in particular attacked the cores of these crystals (Fig. F19). Olivine microphenocrysts are euhedral to subhedral (Fig. F21) and completely altered to calcite. Spinels are observed in the cores of altered olivine pseudomorphs (Fig. F21), and sometimes remnants of previous melt inclusions are observed in these pseudomorphs (Fig. F23). Clinopyroxene microlites and microphenocrysts are anhedral to subhedral, but often the clinopyroxene has grown in fibers (Figs. F20E, F22E, F22F). Because of the almost complete alteration to calcite, the modal abundance of both olivine and clinopyroxene are not easily estimated; nevertheless, their relative abundance is summarized in Table T6.

Phenocrysts

Basalts of Unit V are considered to be of a single petrographic type, carrying microphenocrysts of olivine and spinel, plus microphenocrysts of plagioclase and, where it has not been obliterated by alteration, clinopyroxene. Table T6 categorizes the occurrence (as not present, rare, present, or abundant) and distribution in all thin sections of phenocrysts and microphenocrysts of olivine, plagioclase, and clinopyroxene. Rocks that visually are discerned as aphyric may in thin section contain microphenocrysts. Microphenocrysts are crystals that are <1 mm in length and tabular to euhedral and faceted in outline, and thus were present in the molten basalt upon eruption. Cr spinel is a consistent accessory mineral that is typically enclosed in olivine phenocryst hosts but is so tiny that it cannot be observed in hand specimens. Under the microscope it is usually engulfed in groundmass crystallization. Nevertheless, if Cr spinel is present in olivine with clear crystal outlines, it must have been present in the melt upon eruption as well. In Table T6, spinel is simply noted when it is present or absent.

Olivine: Examples of olivine phenocrysts are shown in Figure F21. All examples are from the pillow stack sequence of Unit V. Each olivine crystal is entirely replaced by calcite, but in even the most delicate features of the original crystals, including enclosed Cr spinel, outlines of melt inclusions, fracture patterns, and delicate dendritic extensions are preserved. The largest olivine phenocryst in all the thin sections is still only ~1.5 mm in longest dimension; all others are on the order of 0.1–0.2 mm. Tiny circular melt inclusions, now crystallized to clay minerals and invariably encased in calcite, are quite common (Fig. F23).

Plagioclase: Some fairly large plagioclase phenocrysts and glomerocrysts occur in the basalt clasts of the volcanic breccia of Unit II (Fig. F22A, F22B). Most thin sections of the pillow sequence of Unit V are devoid of even the smallest plagioclase microphenocrysts. Only one microphenocryst was found in a sample from near a pillow rim.

Clinopyroxene: In most thin sections, acicular plagioclase is not intergrown with clinopyroxene. This is true even in samples where the plagioclase forms an interconnected network (Fig. F19). Existence of such clinopyroxene, however, is difficult to judge in samples where calcite has replaced most interstitial material between the plagioclase crystals. A few samples are sufficiently fresh and coarse grained to preserve intergrowths of needlelike plagioclase and irregularly shaped but often bladed clinopyroxene showing patterns of dendritic crystal growth (Fig. F22). Only one sample contains fresh clinopyroxene microphenocrysts (Table T6), but even these crystals seem to consist of coalesced bladelike domains that are intergrown with acicular plagioclase needles (Fig. F22E, F22F).

Spinel: Spinel most commonly occurs as tiny (~5–10 µm) isolated cubic or octahedral crystals within altered olivine microphenocrysts (Fig. F24A, F24E, F24F). The small crystals are dark brown in transmitted light (Fig. F24B), quite often occur in clusters or clumps (Fig. F24A, F24C), and are strongly reflective and never altered, despite complete replacement of adjacent olivine by calcite. Usually these crystals occur singly or in small numbers within individual olivine crystals (Fig. F24D). Only one larger semi-pyramidal crystal of spinel (Fig. F24F) occurs in all of the thin sections examined. The presence of spinel proved to be the key to recognizing many associated olivine crystals otherwise completely pseudomorphed by calcite.

Segregation vesicles

Perhaps the most striking physical feature of the basalts, especially in the pillow lava stack of Unit V, is their vesicularity. Some rocks have as much as 50% vesicles that are now amygdules filled with calcite and other secondary minerals. Whereas vesicularity clearly indicates shallow eruption, it remains to be determined whether the magmas themselves were sufficiently rich in volatiles to produce so many large vesicles or whether the erupting pillows partially ingested seawater to form steam.

In the more coarsely crystalline pillow interiors we also find segregation vesicles. These occur when adjacent molten material develops crystals, particularly sharply pointed acicular plagioclase, which can puncture the walls of the vesicles. This allows a little bit of interstitial melt to leak (segregate) into the vesicle voids as the rock cools and gas within the vesicle contracts. Figure F25A shows a vesicle within an outer spherulitic portion of a pillow margin that simply was quenched and not yet punctured. The vesicle walls are now lined with brown clays. Another small vesicle shown in Figure F25B was found further into a pillow interior, showing that its walls first were punctured by acicular plagioclase needles and then partially filled with melt introduced through the severed walls. The introduced melt often forms a curving meniscus within the vesicle, against which the shrinking volume of gas typically forms a boundary. Later, alteration minerals were introduced and formed the orange clays and other fine-grained crystalline materials in this vesicle.

Since the injected melt originally was interstitial between the crystals of the surrounding rock, it is more differentiated than the bulk composition of these pillow lavas. In the case of basaltic magma crystallizing mainly plagioclase, this means that the segregate melts are richer in TiO₂ and total iron (Fe₂O₃^T). The result is that the vesicle menisci comprise intergrowths of dendritic clinopyroxene with consistent extinction directions over several adjacent elongate needles of the mineral (Fig. F25C, F25D). Tiny skeletal titanomagnetite occurs between the pyroxene dendrites (Fig. F25E). These are about the only places where titanomagnetite can be seen even (in reflected light) in the otherwise strongly altered rock.

Some vesicles are very large (0.5 cm to several centimeters wide) and thus can contain very large segregation menisci. The injected melt clearly continued to vesiculate even in the segregation vesicles (Fig. F25F). In this case, these second-generation vesicles are now lined with green clay, and the larger, original vesicle space, sequentially, was lined with clear clay, green clay and calcite.

Assessment

A variety of features indicate a petrographic (and thus likely compositional) uniformity of the pillow sequence of Unit V. These are

1. Persistence of small olivine microphenocrysts and occluded tiny Cr spinel;

2. Rarity or absence of plagioclase microphenocrysts;

3. Presence of acicular plagioclase in the groundmass of almost all samples examined, forming interlocking crystalline networks in coarser grained rocks; and

4. A high degree of vesicularity.

The original work of Miyashiro et al. (1969) recognized two principal types of basalt from the Mid-Atlantic Ridge, namely aphyric and strongly plagioclase-phyric variants of what we now recognize as mid-ocean-ridge basalt (MORB). Spinel occurs in these rocks, whereby the plagioclase-phyric MORB variant often has a few olivine phenocrysts, dark brown Cr spinel, and at times phenocrysts of clinopyroxene (Natland et al., 1983). When the pillow sequence of Unit V is taken in conjunction with the plagioclase-phyric basalts of Unit II, the rocks of Hole U1346A have some resemblance to both types of basalt recognized by Miyashiro et al. (1969). In addition, the basalts seem to be less differentiated (based on the absence of clinopyroxene in the mineralogy) than those of Site 1213 from the southern extremity of Shatsky Rise. Other comparisons to MORB and the other Shatsky Rise basalts may arise from geochemistry, but unless immobile elements and element ratios are employed, the original igneous geochemical stratigraphy may be difficult to obtain because of the far-reaching alteration.

Conclusions

Coring in Hole U1346A penetrated >52 m of volcanic rock, most of it a seemingly continuous sequence of particularly vesicular and strongly altered pillow lava and larger inflation units. The volcaniclastic material encountered in Unit II is interpreted as solidified volcanogenic debris introduced onto a shallow marine substrate (less than ~200 m deep) consisting of mainly unconsolidated mud. The soft-sediment deformation around these large volcanic clasts indicates that the material fell into the sediment, possibly rolling downslope from a higher eruption source. Unit II also appears to intervene between the formation of a volcanic stack of pillow lavas (Unit V) and the onset of sedimentation in a progressively deepening water column (Units IV, III, and I). Within that context Unit II may instead include volcanogenic fragments displaced from an existing volcanic edifice (i.e., a talus or debris flow) following cessation of volcanism. The presence of volcanogenic sand and turbidites described in Unit III further supports this interpretation. The delicate cuspate glass shards within the fining-upward lamellae of this unit may, however, have a different origin. Their preservation precludes significant transportation by ocean floor turbidity currents. They therefore may represent subaerial ash fall material that has settled through the water column, or else these glass shards are derived from a contemporaneous hyaloclastite source (e.g., an erupting submarine vent) located nearby. Similar clouds of vesiculated submarine glass debris have been observed settling on the seabed near ongoing phreatomagmatic activity supplied from Pu'u O'o on Hawaii (Umino et al., 2006). Curiously, the pale to colorless character of the glass shards is inconsistent with a mafic source, requiring further analysis to determine the composition of these glass fragments and their origin.

The actual volcanic basement of Shatsky Rise at the Shirshov Massif consists of a stack of highly vesicular basaltic pillow or inflation units grouped into Unit V. Individual pillows were readily identified by the presence of chilled glassy margins, upper and lower chill zones, characteristic pillow vesicle patterns, and crystal grain size variations. In total, 40 of these individual inflation units were recognized (and reported in this site report as lithologic Units 19–58). Downhole logging results indicate levels at which the volcanic stack is less dense, and may be more susceptible to fragmentation, in particular as these are often correlated with intervals of poor recovery. A plausible explanation is that the lava units are smaller, more vesicular, or more strongly altered in these intervals. However, the absence of any intercalated sedimentary material in the cores within Unit V suggests that the succession exists of a near-continuous build up of lava lobes at this particular location and thus likely represents a single lava package or even an eruptive event of pillow basalt (Table T5).

Based on information from sediments cored just above the lava, the volcanic rocks of Units II and V erupted in water certainly no deeper than a few hundred meters (and possibly even shallower). Assuming that the primary water content of Shirshov Massif magma was similar to MORB and Ontong Java Plateau (OJP) magmas (0–0.5 wt%), we estimated that the volatilization depth must have been <300 m (Newman and Lowenstern, 2002). This is an important factor that no doubt contributed to the relatively high vesicularity in the lavas of Units II and V. The close association with carbonate-rich sediment also may have controlled the high percentage of calcite lining fractures and calcite filling the myriad of vesicles in the rock. The vesicularity in turn elevated the porosity of the rock and its ease of fracturing, which almost certainly contributed to the ability of seawater and hydrothermal fluids to penetrate the rock and alter it. Finally, the drilled sequence of igneous rock also must have been close to a low-temperature discharging hydrothermal vent, similar to those found near submarine hot springs on mid-ocean spreading ridges. As a result, the volcanic basement rocks of Units II and V are strongly overprinted by the effects of meandering fluids, and patterns of alteration are readily apparent in the strongly contrasting green, brown, and (ostensibly fresher) gray rock in the core.

Top of page   |   Previous   |   Next