Drilling at Holes 899A and 899B penetrated a 562.5-m-thick succession in which four lithostratigraphic units were recognized (Fig. 4; Table 2). Units I and II are similar to those encountered at Sites 897 and 898, and Unit III is broadly comparable to Unit III at Site 897. However, Unit IV at Site 899, although consisting of a mixture of serpentinized peridotite and Early Cretaceous sediments, contains a distinctive upper breccia subunit that is not well-developed in Unit IV at Site 897.
Subsequent drilling at Site 900 suggested the regional stratigraphic correlation shown in Figure 5. At Sites 897 through 900 the gross lithostratigraphy consists of a lower carbonate-rich contourite-turbidite-pelagite sequence and an upper turbidite-pelagite sequence. The two sequences contrast sharply in terms of evidence for reworking by contour currents (which is present only in the lower sequence) and in the abundance of siliceous allochems (which are virtually absent in the upper sequence).
The ages, averaged lithologic compositions, overall colors, facies and depositional environments, boundary depths, and cored intervals of Units I through IV are summarized in Table 2; Table 3 and Table 4 show the color variations exhibited by the lithologies in Units I and II and Unit III, respectively.
Rotary coring (RCB) was employed in both holes. Hole 899A was washed down to 81.5 mbsf, where drilling commenced in Unit I and terminated at 235.5 mbsf in Subunit IIB. Coring in Hole 899B started at 230.5 mbsf in Subunit IIB and terminated at 562.5 mbsf in Unit IV.
Unit I contains greenish gray siliciclastic turbidites capped by light gray nannofossil-rich hemipelagic/pelagic sediments. Subunit IIA consists of intensely bioturbated hemipelagic/pelagic brown clay stone and nannofossil claystone with scattered turbidites. Subunit IIB, dominated by greenish-gray upward-darkening sequences of nannofossil silty claystone and clayey siltstone, is interpreted as turbiditic sediments reworked by contour currents. Unit III (9.3 m thick) comprises an upper Subunit IIIA of brown claystone and organic-rich black concretions and a lower subunit (IIIB) of sandy silty claystone, yellow sandstone and dark reddish brown clayey conglomerate. The clayey conglomerate may represent a layer of highly weathered material that was derived from the immediately underlying serpentinite breccia.
Unit IV contains serpentinite breccias and blocks intercalated with early Aptian to Barremian sediments. Subunit IVA consists of an unusual serpentinite breccia intercalated with calcareous claystone, claystone, and basalt pieces. Subunit IVB consists of similar sedimentary lithologies interbedded with discrete unbrecciated sections of serpentinite, serpentinized peridotite, and gabbros.
Figure 6 is a plot of ages-vs.-depths (see Table 5) for the sedimentary sequence penetrated at Site 899. Sediment accumulation rates, generalized from Figure 6, vary less than rates observed for the corresponding lithostratigraphic units at Sites 897 and 898: Unit I, 35 m/m.y.; Unit II, 10 m/m.y. Currently available biostratigraphic information does not permit one to evaluate sediment accumulation rates for Units III and IV.
Cores 149-899A-1R through -6R-2, 45 cm
Depth: 81.5-131.65 mbsf
Age: late to early Pliocene
In Hole 899A, 21.68 m of Unit I was recovered, which represents an average recovery of 43.6%. The unit consists of turbidites and associated pelagic/hemipelagic sediments. The turbidites exhibit classical Bouma sequences (Bouma, 1962), and range in thickness from about 5 (e.g., Section 149-899A-5R-1) to 150 (e.g., Section 149-899A-3R-1) cm; these turbidites are comparable to those described in Unit I at Sites 897 and 898.
Core disturbance in Unit I is moderate to soupy. In the first three cores, sediment flow-in resulted from the mobilization of the thick (>20 cm), water-saturated, fine- to medium-grained, basal sand layers of the turbidites. Sedimentary structures in these layers are destroyed or disrupted.
Sediment color ranges from olive black to light gray and light olive gray, with lighter-colored sediments containing a higher carbonate content (Table 3). Medium thickness color-banding of olive gray and light olive gray occurs in Section 149-899A-5R-2. Major lithologies are olive gray silty clay, which changes below Core 149-899A-5R to dark yellowish brown, and very light gray to white calcareous clay to nannofossil ooze. Minor lithologies include greenish black and olive gray sand to silty sand and dark greenish gray to very light gray foraminifer-rich sand (e.g., Sections 149-899A-5R-1, -5R-2, -5R-3). Interbedded silt laminae in silty clay are present in Section 149-899A-4R-2.
The turbidite sequences in Unit I consist of a basal sandy silt to silty sand layer that passes upward into silty clay that has been overlain by nannofossil clay. The sand layers decrease in thickness from about 30 cm in Core 149-899A-1R to 3 to 5 cm in Cores 149-899A-4R to -6R. Lamination and cross-lamination occur in a few of the sands. Bioturbation commonly mixes the light gray pelagic calcareous clay and nannofossil ooze into the underlying darker green silty clay; no distinct ichnofauna is visible. Thin reddish purple (5RP 2/2) bands, which cross-cut bedding, are visible in Sections 149-899A-3R-1, -4R-1, and -6R-1.
Applying the classification of Folk (1980), sands and silts of Unit I are subarkoses, arkoses, and lithic arkoses. Grain types in Unit I indicate derivation from a source area that exposed mostly sedimentary, metamorphic, and, possibly, granitic rocks. The detailed petrography of the sediments in Unit I is identical to that described for Site 898 (see "Site 898" chapter, "Lithostratigraphy" section, this volume).
The depositional setting and processes for Unit I are similar to those interpreted for the same unit at Site 897. Turbidity current deposition periodically interrupted pelagic sedimentation.
Cores 149-899A-6R-2, 45 cm, to 149-899B-15R-1, 42 cm
Depth: 131.65-360.6 mbsf
Age: early Pliocene to late Eocene
The major lithologies in Unit II are silty clay/claystone with silt and nannofossil clay/claystone (Table 2). Minor lithologies include clay/claystone, siltstone, clayey silty sandstone, silty sand/sandstone, foraminiferal silty sandstone, and nannofossil chalk.
The transition from Units I to II is marked by a change from gray green siliciclastic turbidites to a sequence dominated by uniformly mottled brownish clay. The boundary between the two units has been placed at the last, clearly identifiable turbidite having a silty sand base at 45 cm in Section 149-899A-6R-2.
Unit II is divided into two subunits on the basis of a gradational change in sediments from poorly consolidated to lithified. Starting from the top of Core 149-899A-14R, and for all the cores in Hole 899B, a saw was used to split the cores. Therefore, all the lithologies below Subunit IIA are described as "stones."
Subunit IIA is dominated by burrow-mottled brownish clay and nannofossil clay. Subunit IIB consists predominantly of grayish green, darkening-upward sequences of nannofossil claystone, and claystone with silt and siltstone with clay. The subunit contains a significant amount of biogenic siliceous material. The boundary between Subunits IIA and IIB is not observed because nothing was recovered in Core 149-899A-12R, and only 1% recovery in Core 149-899A-13R comprising yellowish gray clay. Therefore, the top of Subunit IIB is placed at the top of Core 149-899A-14R, which contains the first sediments showing upward-darkening sequences.
The recovery in cores from Subunit IIA (0%-46%) averaged 25%. This subunit is dominated by dark yellowish brown to very pale orange clay, calcareous clay, and nannofossil clay. Minor lithologies include nannofossil ooze and silty sand/sandy silt, which locally contain abundant foraminifers. The lighter-colored sediments have higher carbonate contents. The lithologies generally are mixed by moderate-to-intense bioturbation, which created mottled gradations between relatively carbonate-rich and carbonate-poor sediments. No identifiable ichnofauna could be recognized. Some thin (1-5 cm) intervals of laminated silty sand/sandy silt with sharp bases and gradational tops occur. In Cores 149-899A-7R and -8R and Interval 149-899A-9R-1, 0-70 cm, these are overlain by brown clay, but elsewhere, lighter-colored nannofossil clays occur above them.
The average recovery in cores from Subunit IIB was 78% and ranged from 28% to 101%. Most of this recovery is characterized by an alternation of clay- and carbonate-rich lithologies that form upward-darkening sequences ranging in thickness between 10 and 30 cm. These sequences usually have sharp bases and tops, and the lithologies are mixed by moderate-to-intense bioturbation. These are mostly grayish green, but are brown at their tops in Cores 149-899B-10R, -12R, -13R, and -14R.
The upward-darkening sequences are absent in places. Olive gray silty claystones occur in Intervals 149-899A-16R-1, 88 cm, to -16R-CC. Brown claystones with silt occur (e.g., Intervals 149-899B-10R-3, 55-105 cm, and -13R-2, 30 cm, to -13R-4, 80 cm). In places within such intervals, the boundary of the brown coloration is not parallel to bedding and sometimes cuts across burrow traces, which suggests that this coloring is the result of diagenesis.
Major lithologies in Subunit IIB are olive-gray, yellowish brown, and moderate brown silty claystone or clayey siltstone, and light greenish, olive gray, and grayish-orange nannofossil or calcareous claystone. Minor lithologies include claystone, calcareous fine sandstone to siltstone, and silty nannofossil claystone. In sandy and silty lithologies, siliceous allochems (diatoms, sponge spicules, and radiolarians) are common (up to 5% by volume; Fig. 7). Grayish red-purple (5RP 4/2) and dusky blue (5PB 3/2) laminae or blebs (possibly manganese oxide) are scattered throughout.
The typical upward-darkening sequence in Subunit IIB is similar to that described in Subunit IIB at Site 898 (see "Lithostratigraphy" section, "Site 898" chapter, this volume). It begins with a basal, thin bed (up to 5 cm thick, but more commonly 0.5 to 3 cm thick) of mixed biogenic (calcareous and siliceous) and terrigenous fine sandstone to siltstone. Parallel lamination, cross-lamination, ripple cross-lamination, and subtle normally graded bedding are frequent in these sandstones and siltstones. Foraminifer-rich fine sandstones also show scattered, subtle reversely graded bedding, with concentrations of large foraminifer tests on bed tops.
The fine sandstone beds have sharp and flat bases and sharp-to-gradational tops (Fig. 8A). Many of these beds show planar or cross-lamination (Fig. 8B-8C). Where tops are gradational, the transition from the fine calcareous sandstone to the upper nannofossil claystone is sometimes characterized by a wavy, parallel, or lenticular lamination, as well as by burrow mottling between them. The sandstone and siltstone beds make up about 5% of the total thickness of Subunit IIB.
The basal silt- and sand-rich intervals of the upward darkening sequences are overlain by nannofossil claystones. The boundary with the overlying darker silty claystones and claystones is transitional as a result of mixing by bioturbation. A clearly identifiable ichnofauna (Zoophycos, Planolites, and Chondrites) has been concentrated (or is more clearly visible) in the light-colored, carbonate-rich lithologies in the lower part of each individual sequence. Usually, burrowing extends down to the basal calcareous sandstone bed.
In Hole 899B, slump structures involving claystones and nannofossil claystones are present in several cores (Intervals 149-899B-3R-4, 85-110 cm; 149-899B-5R-3, 30-80 cm; 149-899B-7R-4, 90-95 cm; and 149-899B-8R-3, 100-145 cm); some of these are illustrated in Figure 9.
Sand- and silt-sized detritus in Unit II includes components similar to the assemblage observed in Unit I. Unit II also contains clay-rich lithologies that are similar to those in Unit I, except that the Type 2 clay-rich lithologies (oriented clays in carbonate-poor lithologies; defined in "Lithostratigraphy" section, "Site 897" chapter, this volume) are a more persistent and volumetrically significant part of the lithologic assemblage.
Continuous, slow accumulation by settling through the water column accounts for the clay and nannofossil ooze and for some of the nannofossil clay in this subunit. The few sharp-based, normally graded sequences suggest scattered turbidity flows that transported siliciclastic sand and silts or nannofossil clay material. The moderate-to-intense bioturbation indicates relatively slow deposition, as confirmed by the sediment accumulation rate (Fig. 6) in dysaerobic or normal oxygenated conditions. The relatively high content of biogenic carbonate components indicates that deposition occurred above the CCD.
The position of Subunit IIA, between the terrigenous sandy turbidites of Unit I and the distal, muddy current-deposited facies of Subunit IIB, suggests deposition in an abyssal plain setting that was dominated by hemipelagic and pelagic sedimentation, but that occasionally received siliciclastic and carbonate materials from mud-dominated turbidity flows.
Although repetitive, upward-darkening sequences can be seen through most of Subunit IIB, these do not exhibit the clear, normally graded turbidite signature characteristic of Unit I. Thin silty sand and silt intervals at the bases of many of the upward-darkening sequences lack clear, normal grading (Fig. 8). Silty sands and silts that contain small-scale cross-stratification and parallel lamination indicate bottom current activity (Fig. 8B). Some show reverse grading or have sharp tops as well as sharp bases. The tops often show "lag deposits" of large foraminifer tests. All these features point to reworking by contour currents, as described by Stow and Piper (1984). Possibly, some of the homogeneous, carbonate-poor terrigenous silty claystones and claystones may be mud turbidites or contourites.
The high carbonate content and well-preserved nannofossils (see "Biostratigraphy" section, this chapter) indicate deposition above the CCD, and the ubiquitous bioturbation suggests at least a dysaerobic, if not normally oxygenated, environment. The slumped intervals (Fig. 9) indicate penecontemporaneous downslope movement of sediments.
The presence of turbidites, contourites, and slump deposits in Subunit IIB indicates an abyssal plain setting near the continental rise, consistent with the site's location on the continental side of the Iberia Abyssal Plain, and the relatively low sediment accumulation rate.
Cores 149-899B-15R-1, 42 cm, to -16R-1, 7 cm
Depth: 360.6-369.9 mbsf
Age: late Eocene to Campanian/early Maastrichtian
The boundary between Units II and III was placed at the first occurrence of bands containing disseminated black organic material within claystones. Above the boundary, there is a gradual downward change in color from reddish brown to yellowish brown.
Unit III occurs almost entirely within Core 149-899B-15R and is divided into two subunits. Subunit IIIA is entirely claystone with bands of disseminated black organic material, and Subunit IIIB contains variegated sandy silty claystone, sandstone, and clayey conglomerate. The top of Subunit IIIB is marked by the first sandstone layer at 141 cm in Core 149-899B-15R-3. The colors of the lithologies in Subunit III are summarized in Table 4.
Subunit IIIA consists of mottled, dark yellowish brown and pale yellowish brown (Table 4) claystone with grayish orange lenses of siltstones in the upper part of Section 149-899B-15R-1. The claystones contain several black (N1) organic-rich layers, 2 to 10 cm thick, as well as very hard, brownish black (5YR 2/1) concretions (up to 0.5 by 1 cm). Bioturbation decreases from the top to the bottom of the subunit.
The clay minerals in the carbonate-poor claystone and silty claystone in Subunit IIIA are highly oriented. Silt grains, where present, generally have coatings of yellow birefringent clay. These petrographic characteristics correspond to Type 2 claystones (see definition in "Lithostratigraphy" section, "Site 897" chapter, this volume).
Lithologies within Subunit IIIB include variegated sandstone with clayey silty matrix, very compacted moderate brown sandy silty claystone, white highly altered volcanic ash, laminated brown claystone, lighter-colored silty claystone with black concretions (Fig. 10), and a poorly cemented, multicolored, polymictic, clayey conglomerate. Highly altered basaltic clasts are present in the gravel- and sand-sized fractions. Two different types of sandstone are present:
A few localized beds of dark brown to yellowish brown, bioturbated, highly cemented siltstone to claystone are present in Section 149-899B-15R-4.
The lower part of Subunit IIIB consists of a highly weathered, polymictic, poorly cemented, matrix-rich, variegated conglomerate (Table 4). The clasts include black metasediments, white claystone, granule-sized green grains of expandable clay mineral, and highly altered pebble-sized basalt clasts. A yellowish sandstone at 127 cm in Section 149-899B-15R-4 yielded Campanian to early Maastrichtian foraminifers (see "Biostratigraphy" section, this chapter). Silt- and sand-sized material in Subunit IIIB is mostly quartz and alkali feldspar, and therefore, sandstones and siltstones in this subunit have been classified as arkoses (from classification of Folk, 1980).
The carbonate-poor claystones of Subunit IIIA are the result of slow accumulation in an oxygenated environment, probably below
the CCD. This depositional environment is similar to that described for Subunit IIIA at Site 897 (see "Lithostratigraphy" section, "Site 897" chapter, this volume).
Subunit IIIB is interpreted as a mixture of fine-grained hemipelagic/pelagic deposits and high-density turbidity currents or fluidized sand-silt-clay debris flows that developed on relatively gentle slopes, as interpreted for this subunit at Site 897. The basal, poorly sorted, clayey conglomerate interval may have developed as a "weathering" zone, with reworked material having been developed from the top of Unit IV. This zone might have formed sometime during the period between the youngest date (early Aptian) obtained from sediments in the underlying Unit IV, and the Campanian to Maastrichtian age of the sandstone that overlies the red brown conglomerate in Subunit IIIB.
Cores 149-899B-16R-1, 7 cm, to -37R-1, 32 cm
Depth: 369.9-557.9 mbsf
Age: pre-Late Cretaceous to late Barremian
The upper boundary of Unit IV corresponds to the top of acoustic basement (see "Integration of Seismic Profiles with Observations from the Site" section, this chapter). Figure 11 (back-pocket figure) depicts the lithologic sequence observed in Unit IV, which has been correlated to recovered intervals. Three main lithologic associations were observed within this unit:
Unit IV is divided into two subunits (see Fig. 11, back-pocket figure). The top of Subunit IVA was placed at the top of the Upper Breccia Unit (Core 149-899B-16R-1 at 7 cm). This subunit consists of the serpentinite breccias and their associated sedimentary intervals. Core recovery in this subunit averaged 40%. Recovery in cores above and below the top of Unit IV was near 50%, so the change to the breccias characteristic of Subunit IVA occurs over only a few meters.
The Upper Breccia Unit is approximately 95 m thick and constitutes 83% of the recovered interval in Subunit IVA. The Middle and Lower Breccia Units are thinner and are separated from the Upper Breccia Unit and one another by sediments (Fig. 11, back-pocket figure).
The top of Subunit IVB is placed at the base of the Lower Breccia Unit (Fig. 11, back-pocket figure; Core 149-899B-28R-1 at 128 cm). Subunit IVB consists of sedimentary intervals and discrete, boulder- sized intervals (30 cm to >3 m thick) of serpentinized mafic and ultramafic rocks. Core recovery in this subunit averaged 14% and about 80% of the rock recovered consists of pieces of serpentinite, serpentinized peridotite, and microgabbroic rocks. Although much of this material shows evidence of fracturing and other deformation, breccia units similar to those described in Subunit IVA are not present.
The Upper, Lower, and Middle Breccia Units are similar in lithology. The following description, based on study of the thicker Upper Breccia Unit, probably applies to all three units.
The breccias are poorly sorted, polymictic breccias composed of angular clasts with a very restricted range of composition. No obvious distinction can be seen between "clasts" and "matrix," and a continuous range of particle sizes from microscopic fragments to boulder-sized blocks more than 1 m wide is present. The rocks are coherent, although the nature of any cement is uncertain and not obvious under the microscope (see below).
Shape can be estimated only for the smaller (<4 cm) clasts. These are generally equidimensional and have an estimated visual sphericity of about 0.8 (Folk, 1980), probably reflecting the isotropic nature of most of the igneous and metamorphic rocks that formed these clasts. The estimated average roundness of each clast is about 0.2 to 0.3 (Folk, 1980) for clasts contained entirely within the bounds of the core. However, microscopic clasts appear to be even less rounded (0.1-0.2), indicating that these relatively soft serpentinite fragments have suffered minimal transport prior to inclusion in the breccia or significant fracturing during entrainment or emplacement.
The composition of the clasts is a striking feature of these three breccia units. Clasts of only two types are present (see detailed discussion below). The dominant lithology is serpentinitized peridotite (>90% of all fragments). The peridotitic fragments vary widely in composition, and both pyroxene-rich and olivine-rich (dunitic) types are present. Some of the serpentinized peridotite clasts show evidence of deformation prior to inclusion in the breccia (see "Structural Geology" section, this chapter). In the upper part of the Upper Breccia Unit, (Cores 149-899B-16R and -17R), the clasts are dominantly yellow (typically 10YR 4/2 to 10YR 5/4), calcitized, and oxidized; elsewhere in the Upper Breccia Unit, and in the other two breccia units, the serpentinite clasts are greenish black (5GY 2/1). Fragments of metamorphosed Mg-rich igneous rocks (<10% of all clasts) constitute the only other clast type present. Such rock fragments generally are small in size (<10 cm), and white, gray, or even pink. The mineralogy of these metamorphic fragments is discussed in more detail below.
Clast size was studied by systematically measuring the larger clasts in the recovered cores through the Upper Breccia Unit (Table 6). Apparent maximum size shown on the cut surface of the core was recorded. For larger, boulder-sized blocks, the measurement recorded is the length of continuous core that intersected the block. A histogram drawn from 156 clasts having an apparent maximum size greater than 2 cm (Fig. 12) shows that smaller fragments (2-5 cm) dominate and that the size distribution is approximately logarithmic binomial. The median size of clasts >2 cm is 3.75 cm. This size distribution mimics that found in some glacial tills and in the lunar regolith.
Figure 13 shows the apparent maximum size of each measured clast plotted vs. depth, demonstrating that there is a tendency for larger clasts to be restricted to the central part of the Upper Breccia Unit. The significance, if any, of this relationship is uncertain.
Although the bulk composition of these three breccia units is clear from examination of the clasts, three samples were analyzed for major elements. The samples did not contain large clasts and thus are representative of the finer material within the Upper Breccia Unit (Table 7). The ultramafic bulk composition of the unit was confirmed by these analytical data. This conclusion also is supported by analyses of trace elements in a much larger sample set (Table 8).
In summary, the breccia units are poorly sorted, and composed of a suite of ultramafic clasts of a restricted compositional range. The clasts vary widely in size from boulders to microscopic pieces, and the small and microscopic fragments are highly angular in shape.
Only two types of clast occur within the three breccias. Most are serpentinized peridotites, but a minority (<10%) of the clasts are metamorphosed Mg-rich igneous rocks.
The larger blocks (cobbles to boulders) are variable in composition and range from altered serpentinites to pyroxene-rich and pyroxene-poor peridotites. In the upper part of the Upper Breccia Unit (Cores 149-899B-16R to -18R), the serpentinite clasts have been highly altered. Bright green granules and rare larger clasts appear to be made of clay (smectite?), possibly derived from pyroxene crystals. Other fragments are gravel-sized, white to creamy, altered serpentinite, crosscut by purple to dark blue serpentine veins. Some of these contain a similar bright green mineral (e.g., Intervals 149-899B-16R-3, 43-60 cm, and-17R-1, 16-38 cm; Fig. 14).
Below this altered horizon in the Upper Breccia Unit, the dominant clasts are serpentinized peridotites that lack the obvious clay alteration. Several different varieties can be distinguished:
Weakly deformed peridotites are the most frequent lithology. They display coarse-grained textures and are serpentinized, with the mesh structure appearing more fine-grained in serpentinized dunites. The only primary mineral remaining is spinel. In places (e.g., Section 149-899B-19R-2), clasts are composed of highly deformed serpentinite with porphyroclastic to mylonitic textures. Some former olivine-rich bands have been replaced by fine mesh serpentinite. Elsewhere, colorless fibrous amphibole has replaced orthopyroxenes and chlorite has formed within magnetite- and spinel-rich zones. Pumpellyite-rich zones may represent former clinopyroxene. Similar retrograde metamorphic assemblages were observed in Subunit IVB (Interval 149-899B-29R-1, 129-132 cm) in an undeformed former clinopyroxenite band that crosscut serpentinized dunites. These metamorphic replacements are not homogeneously distributed for, in places, the texture and mineralogy have been obscured by later alteration and the rocks have a patchy pale gray color (e.g., Intervals 149-899B-19R-3, 0-15 cm; -20R-2, 13-21 cm; -21R-2, 50-95 cm; -28R-1, 103-112 cm; also Fig. 16). However, the original textures are recognizable.
In thin section, the smaller fragments are mostly serpentinite clasts. Discrete, scattered magnetite, pyroxene, and brown-spinel crystals also occur and represent remnants of unserpentinized peridotites.
Thin sections of pebble- to sand-sized clasts display well-crystallized serpentine (clear areas with uniform extinction). Other areas of serpentine are cloudy and are made of short crystals. Both well-crystallized and cloudy serpentine may coexist in a single fragment and seem to be, respectively, recrystallized and "relict" serpentine. A third type of serpentine has large crystals (up to 2.5 mm) with an obvious and regular cleavage (lizardite derived from bastite?). These crystals often are bent and/or crosscut by the most recent serpentinite crystallization event, which produced the clear serpentine (Interval 149-899B-21R-5, 32-36 cm).
One section contains what appear to be small sedimentary particles (Interval 149-899B-26R-1, 30-34 cm). These fragments, with sizes between 0.2 to 1.8 mm, have very fine-grained dusty cores rimmed by clear zones, 0.05 to 0.1 mm thick, that resemble serpentine. Other more angular fragments, classified as serpentinite in plane polarized light, are isotropic. A few fine-grained clasts of serpentinite display the characteristic fabric of mylonitized serpentinite (Interval 149-899B-20R-3, 50-53 cm).
Shipboard examination of the breccias failed to identify any fragments of continental basement, and no clasts containing quartz were recovered. Only one 2-cm feldspar-bearing clast was found. It shows porphyroclastic feldspar, granulated epidote, a brown mica, and zircon. The type of feldspar is unclear and needs further study.
The most common metamorphic clast type found in the breccia unit is gray to white (rarely pink and green), very fine-grained, and displays no preferred mineral orientation (Fig. 17). Some clasts have been crosscut by veinlets restricted to the clast or extending into the matrix, the blocks may be fractured and the pieces slightly displaced (Interval 149-899B-20R-3, 43-52 cm). These fine-grained clasts appear to be low-grade metamorphic Mg- and/or Ca-rich rocks. Few of them have an intersertal texture. The clasts display several mineral associations, which may be grouped as follows:
Several features of the breccia units constrain potential models for the mode of emplacement of the breccia. Some of the more important of these features are the composition, texture, structure, and stratigraphic framework. These aspects of the units are discussed here with a view to identifying the processes leading to the formation of the units and their origin.
The serpentinite bulk composition of the breccia units significantly constrains their mode of origin. Possible basaltic and microgabbroic rocks are also present among the metamorphic clasts. They are identified by their unmodified texture. No fragments of sedimentary rocks, pelitic metamorphic rock, or granitic clasts are included.
As noted above, most of the clasts and boulder-sized blocks within the breccia units are serpentinized peridotite. The small fraction of the clasts (<10%) that are metamorphic contain a mineral assemblage characteristic of the prehnite-pumpellyite facies. Any model for the origin of the breccias must generate this specific mineral assemblage.
Texture is a striking feature of these breccias. Initial shipboard studies suggested that the fragments range in size from submicroscopic to boulder-sized blocks. All the smaller (1-20 mm) and microscopic (<l mm) clasts appear to be angular; and there is complete size gradation between "clasts" and "matrix" that makes this distinction arbitrary. Examples of fragments that appear to have been broken and granulated into elongated ribbons are common and visible at both the hand-specimen and microscopic scale (Fig. 18). Textures of this type are not characteristic of normal sedimentary mass flows, while volcaniclastic, submarine ash flows of this composition are very rare.
A third general feature of the unit is the overall absence of structure and sorting within the breccia. Within the Upper Breccia Unit, only a very few examples of a primary planar fabric were seen (e.g., penetrative deformation, zone of sorting, and so forth) that crosses the core. Most examples were short, discontinuous shear zones related to the margins of boulder-sized blocks (Fig. 16; see "Structural Geology" section, this chapter). However, a weak, irregular penetrative shear fabric within the breccia was observed in places at the top and base of the Upper Breccia Unit and the base of the Lower Breccia Unit. Overall, the breccia units are best described as unsorted and without planar structures (but see "Structural Geology" section, this chapter). We interpret this as indicating essentially instantaneous formation of each of the three breccia units.
The stratigraphic framework presented above (Fig. 11, back-pocket figure) also significantly constrains the origin of the breccias. The sequence of biostratigraphic ages of intercalated fine-grained sediments within Unit IV (Table 2 and Table 3; Fig. 11, back-pocket figure) suggests a normal stratigraphic succession. It is for this reason that the Lower Breccia Unit may be considered the oldest of the three tabular breccias. Any tectonic model for the origin and emplacement of the breccias must generate this apparent stratigraphy.
Finally, low temperatures of emplacement/formation for the breccia units are indicated by the absence of thermal metamorphism in the underlying sediments (see below). Certainly, the bases of the breccias were never exposed to elevated temperatures (? 50°>100°C). The center of the thicker, Upper Breccia Unit might have reached higher temperatures without affecting the adjacent sediments, but no evidence is seen that the clasts themselves have been altered by post-emplacement thermal metamorphism. The clasts are not welded together, and the Upper Breccia Unit displays none of the welding features characteristically developed following the emplacement of hot ash flows.
In conclusion, the observations outlined here significantly constrain models for the emplacement of the breccias. Tentatively, we assume that the three breccia units are some form of mass-flow deposit, laid down in a subaqueous environment. However, the origin of the fragmentation that generated the clasts remains obscure.
All three breccia units show clear evidence of significant post-emplacement alteration. The two most obvious features resulting from this alteration are the color variation within the Upper Breccia Unit and the calcite veining of all three units. These two features are associated and, in the case of the upper unit, are related to large-scale chemical changes in the bulk composition.
Color changes are obvious only in the Upper Breccia Unit. The finer-grained matrix of this unit is predominantly dark yellowish brown (10YR 4/2) to moderate yellowish brown (10YR 5/4) in its upper part (Cores 149-899B-16R and -17R; Table 6). However, the recovered rocks from Core 149-899B-18R are darker (10YR 2/2) and gradually become dark green/gray or gray (N4) toward the base. Rocks from the lower parts of the unit (Cores 149-899B-20R through -26R) generally are dark green (5GY4/1) or greenish black (5G 2/1), with only short brownish sections (e.g., Sections 149-899B-20R-1 and -21R-2). While these color variations generally are gradational, only at 47 cm in Section 149-899B-25R-3 is there a sharp boundary between a medium gray and a medium light gray matrix. The sharpness of this boundary and the fact that it crosses a clast and displays no relationship to other primary features of the breccia demonstrate clearly the post-emplacement nature of the alteration.
Calcite alteration, perhaps reflecting a combination of veining and replacement, is more pervasive in the upper part of the Upper Breccia Unit (Cores 149-899B-16R and -17R; Fig. 19 and Fig. 20). Large-scale features that are clearly the result of replacement by calcite are absent, but in thin section, both vein filling and possible replacement fabrics are apparent. Calcite veins (Table 9; Fig. 19), broadly distributed in Cores 149-899B-18R through -25R, exhibit a slight tendency toward greater abundance in the central part of the Upper Breccia Unit. The larger veins (>1 cm wide) often occur as part of multivein complexes at the borders of boulder-sized serpentinite pieces, which are themselves more abundant in the central part of the Upper Breccia Unit (Fig. 13). These boulders may have generated a ductility contrast that facilitated veining. Some of these larger veins contain fibrous serpentine, which suggests that they have replaced or overprinted a previous serpentine-veining event (e.g., Interval 149-899B-18R-5, 58-62 cm). The larger veins show clear evidence of multiple stages of vein filling, and cross-cutting relationships can be observed in some intervals (e.g., Interval 149-899B-18R-5, 60-65 cm). Smaller veins (<1 mm), apparently spawned by the larger veins, are widely distributed throughout the upper part of the unit and typically form an anastomosing network.
The calcite veining (Fig. 19) clearly post-dates the emplacement of the Upper Breccia Unit and is among the youngest events modifying the Upper Breccia Unit, because veins clearly transect clast boundaries. Only very rare veins of a fine-grained brownish material were seen cutting the set of calcite veins (Interval 149-899B-16R-2, 66-78 cm).
Bulk chemical analysis of 21 samples from the Upper Breccia Unit (Table 8; Fig. 20) confirm that the upper part of the unit is indeed enriched in CaCO3. This enrichment is particularly marked in samples from Cores 149-899B-16R and -17R. Below Core 149-899B-18R, calcite alteration is much more unevenly distributed, and only limited zones have been pervasively altered.
Data for Sr and Ba show that these two elements are also significantly enriched in the upper part of the Upper Breccia Unit, with the enrichment being clearly related to the calcite alteration (Table 8 and Fig. 20). Conversely, elements such as Ni and Cr, abundant in the unaltered breccia, have been depleted in the altered zone.
At this preliminary stage, the alteration has been ascribed to interaction with circulating seawater during the period immediately following the emplacement of the unit. This alteration follows a similar pattern to that observed in the upper part of the massive, serpentinized peridotite recovered from Hole 897C.
Basalts occur in close association with the sedimentary lithologies in Unit IV. Basalts are present as isolated pebble- to cobble-sized pieces; they are rounded pebble-sized pieces within claystone and calcareous claystone (Table 10). Most of the basalts have no vesicles. Other basalts have vesicles filled with a white, fine-grained chlorite or a pink mineral, probably calcite. Undeformed pieces of microgabbro (or coarse dolerite) are present (Intervals 149-899B-34R-1, 37-72 cm; -35R-1, 0-60 cm). These rocks have an intergranular texture with plagioclase laths (2.6 mm long) and euhedral altered olivine. Although mylonites occur in Interval 149-899B-34R-1, 0-37 cm, and although some of these may be mylonitized microgabbros, although the only analyzed sample is not, most of the fine-grained mafic igneous rocks from this site have not been deformed.
The first occurrence of basaltic rocks is in Section 149-899B-26R-1 (Fig. 21). This piece is distinctly brecciated and has been interpreted as an autobrecciated lava unrelated to the serpentinite breccia units. Dark and fine-grained fragments of basalt (up to 2 cm), outlined by thin white veins, are dispersed in a dark matrix having flow texture similar to that of the included fragments. The fragments are rimmed by chlorite (0.2-1 mm thick). The shape of the breccia fragments is not well defined, as the basalt has been highly altered. Nearly all of the plagioclase and glass have been replaced by chlorite and smectite. Similar basalts were found in Interval 149-899B-28R-2, 8-10 cm. The basalt recovered from Section 149-899B-27R-1, 19-20 cm, is porphyritic, with about 20% relatively unaltered plagioclase phenocrysts. Olivine phenocrysts, replaced by chlorite or serpentine, also are present. Two other basalt pieces exhibit variolitic texture (e.g., Interval 149-899B-31R-1, 9-14 cm). This mode of crystallization and the alteration of these rocks clearly suggest that these are pieces of submarine lavas.
Only four samples in this unit (three basalts and one diabase) were analyzed for major elements (Table 11) and three for minor elements. The high loss-on-ignition indicates the highly altered nature of the analyzed rocks. The high magnesium content (16.4% MgO) of Sample 149-899B-27R-1, 33-40 cm, also may reflect alteration. All three basalts have high, but variable, values of K2O, reflecting either alteration or a non-MORB-type composition, and the diabase (Sample 149-899B-35R-1, 7 cm) has a very high K2O content (5.67%). The average nickel and chromium for two samples of basalt and microgabbro (Samples 149-899B-28R-2, 10-13 cm, and -35R-1, 7-11 cm; 154 and 230 ppm, respectively, for Ni and Cr) are normal for mafic rocks.
In summary, the basalts are extremely variable texturally, mineralogically, and compositionally, which precludes an interpretation that relates the pieces to a common origin. The variations among the basalts are consistent with the interpretation of these pieces as clasts.
Four recurring sedimentary lithologies are present in cores beneath the Upper Breccia Unit (Fig. 11, backpocket figure; Table 12). These are intimately associated with the basalt pieces described above and also have been interspersed with the unbrecciated segments of serpentinite, serpentinized peridotite, mylonite, and gabbro. The sedimentary lithologies are as follows:
Three other sedimentary lithologies occur only once each in Unit IV. One of these is a limestone (Interval 149-899B-27R-1, 94-103 cm) composed almost exclusively of loosely aggregated sand- and siltsized, subhedral, monocrystalline and polycrystalline calcite. Poorly preserved nannofossils are present in minor amounts. Another lithology is that of a gray laminated siltstone with prominent soft-sediment faulting (e.g., Interval 149-899B-35R-1, 101-121 cm). This lithology is a striking compositional and textural contrast to the other sedimentary lithologies in Subunit IVB (Fig. 26). It is a subarkose (classification of Folk, 1980) and contains the only occurrence of greater-than-trace amounts of quartz and alkali feldspar below the base of Unit III. The third lithology is a black radiolarian(?) chert with veins filled by spherules of length-slow chalcedony (Interval 149-899B-29R-2, 24-32 cm; Fig. 25B). This chert is dark-colored because of a large amount (?10%) of opaque organic matter and pyrite. Allochems have been highly recrystallized, and their identification is problematic.
At Site 897, no unit clearly analogous to the serpentinite breccia at Site 899 is well developed. Nevertheless, certain broad similarities can be identified between lithologic associations in Unit IV at the two sites (see "Site 897" chapter, this volume).
Except for the serpentinite breccias of Subunit IVA, all of the nonsedimentary lithologies observed in Unit IV at Site 899 are interpreted as clasts. Several lines of evidence support our contention that the basalt pieces, serpentinite, serpentinized peridotite, mylonite, gabbro, as well as some claystone, and the radiolarian chert represent sedimentary clasts of pebble to boulder size. The rounded shape of pieces within the claystones and apparent weathering rinds on some pieces support this interpretation. Several isolated rounded, pebble-sized clasts, most notably Section 149-899B-29R-2, Pieces 1 (claystone) and 2 (basalt with an apparent weathering rind; Fig. 26) are partially coated with a coherent layer of claystone that resembles armoring, which may have been acquired during repeated transport and deposition.
The intercalation of sedimentary lithologies with these intervals of nonsedimentary rock is the most obvious feature shared by sequences designated as Unit IV at Sites 897 and 899. The specific sedimentary lithologies present within Unit IV vary somewhat in their proportions, but several lithologies are common to both sites. Site 897 was characterized by a larger amount of calcareous material, including limestones. Ages obtained from nannofossils (late Hauterivian to late Aptian at Site 897, Barremian to early Aptian at Site 899) are similar at both sites and, in both cases, indicate a normally ordered sequence that gets younger upward. This supports the idea that the calcareous claystones, from which the ages were obtained, in some way may be present in locally deposited sediments into which the clasts of nonsedimentary material were introduced.
Another recurrent sedimentary lithology in Unit IV is the highly deformed, dark gray, organic-rich mudstone associated with deformed serpentine masses. At Site 897, the Shipboard Scientific Party considered (see "Site 897" chapter, this volume) that this material might be the product of drilling disturbance. However, the distinctive organic content and age of this material strengthened the interpretation that it is sedimentary in origin.
The laminated, gray, subarkosic siltstone, which displays soft sediment deformation, also appears at both Sites 897 and 899. The great compositional contrast between this siltstone and the associated sediments supports the idea that it may represent a clast in a manner analogous to the basalt fragments. However, the laminations are subhorizontal, an orientation that is highly improbable for an allochthonous clast.
As with the breccia units, some variation of a mass flow origin is also proposed for the mixed assemblage of igneous and sedimentary lithologies that are found both intercalated between the breccia bodies in Unit IVA and in Unit IVB. Such an interpretation is drawn by analogy to Unit IV at Site 897 because of the similarities in lithologic types and associations observed in the Units IV between the two sites, exclusive of the serpentinite breccia (see "Site 897" chapter, this volume). Again, the presence of serpentinite and serpentinitized peridotite within an upward-facing sedimentary sequence would seem to indicate exposure of mantle-derived rocks to submarine weathering and transport, beginning at least as early as Barremian.