Leg 178 sediment classification is based on visual core description, smear-slide analysis, and spectrophotometer reflectance data. Data are condensed to ODP standard barrel sheet format and presented using the program AppleCore. Shipboard sedimentologists chose to follow a modified version of the ODP Leg 105 sediment classification (Shipboard Scientific Party, 1987), which is suited to describing glaciomarine sediment in that it distinguishes several categories of siliciclastic sediments (gravel, sand, silt, and clay) facies. The Leg 178 classification shown in Figure F1 is a simplification of the classification shown in figure 7 of the Leg 105 "Explanatory Notes" chapter of Shipboard Scientific Party (1987). Existing ODP classifications, however, do not adequately address nonsorted or poorly sorted admixtures of siliciclastic sediments, such as tills, ice-rafted and ice-turbated deposits, and debris flows. For these types of clastic sediment, we use the descriptive terms diamict (unlithified) or diamictite (lithified).
Major sediment types are distinguished on the basis of the dominant component (>50%), which provides the principal lithologic name (e.g., diatom ooze and silty sand). When a component comprises 25%-50% of the sediment, it is mentioned as a major modifier preceding the principal name (e.g., diatomaceous clay and nannofossil silty sand). Minor constituents (10%-25%) are included using the term -bearing (e.g., diatom-bearing clay; nannofossil-bearing silty sand). The sediment modifiers are ordered so that the minor modifier(s) precede the major modifier(s). Specific nomenclature for each major compositional group is given below.
The terrigenous category refers to sediments having >50% terrigenous component, which are classified on the basis of grain size as shown in Figure F1. This classification differs from that of Leg 105 in that sediments having >75% of a single component are given that name (e.g., 23% silt, 77% sand = SAND; or 24% clay, 76% silt = SILT). The term diamict is employed for the coarse terrigenous sediments recovered during Leg 178 that comprises gravel and granule clasts (>2 mm), matrix-supported by admixtures of sand and mud (see below).
When the biogenic component is between 25% and 50%, a biogenic modifier is used as shown in Figure F2. In this case, the size designation for the terrigenous sediments is based on the composition of the terrigenous components (e.g., 40% diatoms, 40% clay, 20% silt = DIATOM SILTY CLAY). When both siliceous and carbonate biogenic components are present, and one type of biogenic material has more than one component (e.g., silica composed of diatoms, radiolarians, and sponge spicules), the entire biogenic fraction is treated as a single entity (e.g., 11% foraminifers, 15% radiolarians, 20% diatoms, 30% clay, 17% silt, 7% sand = FORAMINIFER-BEARING SILICEOUS MUD), showing the importance of that type of biogenic material rather than the individual components.
When the biogenic component exceeds 50%, the sediment is an ooze (or lithified equivalent), following Figure F2. As with the biogenic modifiers for the terrigenous sediments, the designation as siliceous or carbonate sediment depends upon grouping all of the biogenic components with that composition, and the terrigenous size fraction is determined only on the basis of the terrigenous components (e.g., 30% clay, 25% foraminifers, 15% nannofossils, 30% radiolarians = CLAYEY FORAMINIFERAL SILICEOUS OOZE).
When the terrigenous component is between 10% and 25%, it is followed by the word -bearing. If the terrigenous component is <10%, it is not included in the description.
A common component of the sediments recovered at deep-water sites of Leg 178 are thin-bedded, fine-grained silts and clays showing a consistent succession of sedimentary structures. These were deposited from turbidity currents. Other fine-grained laminated facies were deposited from bottom currents (contourites). What follows is a brief discussion of the criteria used to separate these two facies.
Turbidites are recognized by reference to the classic descriptive scheme of Bouma (1962) (divisions TA-TE). Fine-grained turbidites recovered during Leg 178 show most commonly the TC-TE divisions corresponding to cross-laminated silt, parallel-laminated silt, and mud components. Bouma's original scheme has been found, however, to be too generalized for application to muddy turbidites. To facilitate detailed descriptions, Piper (1978) further subdivided the TD and TE divisions of Bouma into laminated silt (D), laminated mud (E1), graded mud (E2), ungraded mud (E3), and pelagic and hemipelagic (H) intervals (Fig. F3). The presence of multiple silt laminations in division TD results from repeated cycles of shear sorting of silt grains and clay flocs in the bottom boundary layer. The overlying TE division records suspension deposition from the tail of the turbidity current, followed by hemipelagic deposition, and reworking by bottom (contour) currents (see Pickering et al., 1989; see below). Later workers have erected more detailed classifications of structures within muddy turbidites (e.g., Stow and Piper, 1984), but discrimination of the different divisions is dependent upon quantitative grain-size measurements and X-radiographs, which were not possible during Leg 178. The interpretation of fine-grained deposits was based, in consequence, upon visual and hand-lens examination using Piper's (1978) scheme. During Leg 178 three distinct facies, L1, L2, and L3, are recognized (Fig. F3) on the basis of the presence or absence of silt laminae. These typically form a depositional continuum in which individual facies are related to distance from source (Fig. F4; Stow and Piper, 1984). Each facies provides important information regarding depositional energy.
Hill (1984) showed that fine-grained turbidites and contourites represent end-members of a continuum of deep-sea sediments that accumulate in very low energy settings; in some cases, distinction may be difficult. This is particularly the case where the tail of a muddy low-energy turbidity current discharges any remaining suspended fine-grained sediment into the water column where it is then affected by bottom currents. Such processes are well described by Stow and Wetzel (1990).
End-member muddy turbidites show sharp bases, a well-defined internal structure, and grain-size sorting (e.g., silt to mud) and, as shown above, are categorized using the descriptive schemes of Bouma (1962) and Piper (1978). In contrast, end-member muddy contourites generally lack any consistent internal structure, although they may reveal the presence of diffuse laminae and isolated ripples, with transitions from silt to mud reflecting the variation in strength of a relatively slow-moving, near-continuous bottom current (e.g., Gonthier et al., 1984; Stow et al., 1986; Nelson et al., 1993; see Facies C, Fig. F5). Such facies typically show complete, pervasive bioturbation (e.g., Gonthier et al., 1984). The differences outlined above are essentially the result of high deposition rates and the short recurrence interval between successive turbidity currents, compared to contourite drifts deposited from bottom currents. The internal structure of a turbidite bed reflects deposition from a discrete, decelerating turbidity current; deposition is relatively rapid (e.g., hours to a few days), and a comparatively thick sediment layer can accumulate in a short interval (Nelson et al., 1991).
A repetitive internal structure (the "Bouma" sequence) and a tendency to form thick and repetitively bedded stratigraphic successions are considered the most important characteristics for distinguishing turbidites from contourites. Bottom (contour) currents are characterized by nonsystematic changes in flow velocity resulting in much poorer overall sorting and a lack of any consistent and repetitive characteristic internal structure (Stow and Piper, 1984). Contourites can be associated with lag surfaces that record increased bottom current velocities; in addition, contourites are extensively mixed and burrowed by ichnofauna (Wetzel, 1984), which reflect comparatively low deposition rates. Very similar criteria for distinguishing turbidites from contourites were employed by Leg 105 scientists (Hiscott et al., 1989).
The objectives of Leg 178 are to examine depositional processes and the history of climate change along the glacially influenced continental margin of the Antarctic Peninsula. In such settings, an important sediment type is poorly sorted diamict facies.
Diamict is used here as a nongenetic term for sediment consisting of admixtures of clasts (defined here as fragments larger than 2 mm in diameter), sand, and mud where clasts are matrix supported (Flint et al., 1960). Matrix support distinguishes diamicts from other poorly sorted sediments such as muddy gravels, which are clast supported. Matrix grain size in diamicts can be described using the scheme outlined in Figure F1. The term diamict is synonymous with diamicton (both are used in the literature for unlithified sediment), whereas the term diamictite is employed for lithified sediments. Diamicts are characteristic of depositional environments receiving poorly sorted glacial or volcaniclastic sediment. Universal descriptive schemes for diamict(ite) have been developed regardless of depositional environment or setting and emphasize the presence or absence of internal structure and organization. Diamict facies are either massive or variably stratified.
Massive diamict facies have no internal organization or structure and consist of scattered clasts supported by fine-grained matrix. In glacial and glacially influenced marine settings, such facies are not diagnostic of any one subenvironment and are produced by a wide range of sedimentary processes. Massive diamicts can be deposited subglacially (till), or by postdepositional downslope resedimentation as massive debris flows (debrites); the same debrite facies, however, can be generated in nonglacial settings from the mixing of different sediment populations during downslope mass flow, most commonly by mixing gravelly debris flows or turbidites with fine-grained slope sediments. Massive diamict facies also accumulate in deeper water below wave base, by ice rafting of sand and larger clasts into muddy basinal deposits (rainout facies). In many cases, the internal distribution of clasts within massive diamict may provide diagnostic information (e.g., normal or inverse grading). Another useful descriptor is clast abundance (e.g., clast poor or clast rich), which relies on visual classification of number of clasts/area on the surface of the split core. We adopted the comparison chart for visual percentage estimation presented by Mazzullo et al. (1988; fig. 16). A visual percentage estimate of 20% gravel clasts is taken as the boundary between clast-poor and clast-rich facies.
Stratified diamict facies encompass a range of subfacies from well-stratified through weakly stratified to chaotic. Again, no particular subfacies is diagnostic of any specific environment although a dominance of stratified diamict facies usually indicates a slope setting. Stratified facies can originate by repeated sediment gravity flow, variation in the flux of sediment delivered from suspension settling or rainout from floating ice, or reworking by currents. Chaotic facies can be produced by incomplete mixing of different sediment populations during downslope mass flow and may indicate an ice-proximal location (Eyles, 1993). The same facies, however, may be associated with strata deformed either below glacier ice (glaciotectonism) or by grounding keels of floating icebergs and pack ice (ice turbation).
Bioturbated diamict facies arise from bioturbation of existing diamicts (thereby providing important data regarding the physical properties of the diamict at the time of bioturbation) but also from the disruption of interbedded muddy/sand facies containing ice-rafted debris.
Regardless of type, diamict facies must be interpreted with care, using criteria such as vertical and lateral facies associations (i.e., sequence context), any paleontological data, overall basinal setting and geometry, and acoustic signature as identified from seismic records. For example, on glaciated continental margins, the intimate association of diamict facies with turbidite facies is usually a good indicator of a setting dominated by sediment gravity flows and suggests a debrite origin for diamicts. A channeled geometry and hummocky upper surface is typical of, but not exclusive to, this sedimentary setting. Paleontological data provide critical contextual information. Thus, downslope sediment flow and mixing may be identified by the presence of displaced shallow-water microfauna, the presence of rainout facies, or by an in situ microbiota. As with interpretation of massive diamict facies, the examination of the sequence context of facies, in addition to paleontological information and ichnofacies types, will provide important clues as to depositional environment.
Confusion can arise where ice-rafted debris is present in muddy strata. A massive or laminated mud with ice-rafted clasts is not regarded as a diamict in the absence of a poorly sorted matrix. The general rule of thumb is that a mud with >10% of gravel clasts by volume should be described as a diamict. This number varies from worker to worker and is subjective, but it can be estimated from the comparison chart for visual percentage estimation presented as figure 16 in Mazzullo et al. (1988).
To avoid usage of genetic terms such as till, Eyles et al. (1983) proposed a simple lithofacies scheme for description of diamict facies (Fig. F6). This scheme is now widely employed for sediment description in volcanogenic and glaciogenic environments where large volumes of poorly sorted sediment are produced (e.g., Miall, 1984).
Till(ite) is a specific genetic term employed for a diamict(ite) deposited directly by glacier ice (Eyles and Eyles, 1992). Tills are deposited subglacially by a variety of processes and are correspondingly lumped together as basal tills. Lodgement/melt-out tills are deposited where debris in the ice base is plastered over a stiff substrate (a hard bed). Such tills are predominantly massive diamict facies (with secondary stratified facies) and tend to be thin (<10 m) in a basinal context. In contrast, much thicker basal till strata, up to 50 m or more, accumulate by repeated deposition from debris being transported as a "deforming layer" (referred to as a "soft bed"). Current models assume that resultant deformation tills are deposited by successive "freezing" of debris at the base of the deforming layer. These conditions apply where ice overrides soft sediments that undergo pervasive deformation, mechanical mixing, and homogenization analogous to processes in a cement mixer. These tills consist of stacked beds of massive diamict facies, but rafts of pre-existing sediment can commonly be identified where homogenization was incomplete. "Soft bed" conditions have now been identified across much of the outer margins of the large continental ice sheets in the Northern Hemisphere. These outer margins were highly lobate, and probably fast-flowing ice streams lubricated by soft beds developed over wet sediment or poorly lithified bedrock. Similar conditions probably applied as expanded Antarctic ice sheets moved across shelf sediments during successive glacial maxima. At the shelf edge, such subglacial debris was probably discharged downslope and redeposited as a variety of sediment gravity flows. The term waterlain till has been used in the past to refer to such sediments, but the term is not preferred here because the debris has undergone redeposition by nonglacial processes.
The role of "ice turbation" is now increasingly recognized. At ice-sheet margins, seafloor sediments in water depths perhaps as deep as 500 m are subject to mechanical deformation and sorting by the repeated grounding of iceberg keels (or, in shallower water, by grounding keels of pressure-ice ridges). Seismic data and direct observation of ice scours (on modern shelves and from Pleistocene examples) show deformed ice-turbated sediment akin to diamict.
Nomenclature is that of Miall (1984), Mazzullo et al. (1988), and Pemberton et al. (1992). Key symbols used on ODP visual description forms and barrel sheets are standard (Fig. F7).
Nomenclature remains unchanged from Mazzullo et al. (1988) and uses terms such as thinly laminated (1-3 mm thick), laminated (3 mm-1 cm thick), very thin bedded (1-3 cm thick), thin bedded (3-10 cm), medium bedded (10-30 cm), thick bedded (30-100 cm), and very thick bedded (>100 cm).
These terms follow standard ODP nomenclature (Mazzullo et al., 1988) and that used during Leg 105 (Shipboard Scientific Party, 1987).
Reflectance from cores was routinely measured at evenly spaced intervals of 5 cm downhole using a Minolta Spectrophotometer CM-2002, except for Site 1098, where the measurements were taken at 1- or 2-cm intervals. Measurements were taken as soon as possible after the cores were split to minimize redox-associated color changes that occur when deep-sea sediments are exposed to the atmosphere. We followed the procedure described in the Leg 174A "Explanatory Notes" chapter of Shipboard Scientific Party (1998a). The spectrophotometer records the percentage of reflected light in 10-nm wavelength steps, from 400 to 700 nm. The hue and chroma attributes were used to determine the standard Munsell notation. A tristimulus system is used to describe color; the measured reflectance of a specific spectral energy distribution under standardized conditions is compared with the three primary colors, red, green, and blue. The result of the comparison is expressed as X, Y, and Z, respectively, and called the tristimulus values. The tristimulus values X, Y, and Z can be converted to the CIELAB system, with derived values called L*, a*, and b*, where L* is the lightness parameter, and a* and b* represent the chromaticity parameters. As an aid to defining lithostratigraphic units, we used one of the parameters (L*, a*, b*) of the CIELAB system because these give a maximum response to changes of lithology in the sediments cored during Leg 178 (see also Blum, 1997, Chap. 7).