10% glass particles
(EL in Table 3). Volcanic
glass occurs as light to dark brown, ± opaque (tachylitic), and
colorless fragments in ash horizons with contents of glass
particles varying from ~10% to 90%. Three types of ash deposits
can be distinguished on the basis of structure and modal
compositions: (1) discrete ash layers contain >80% glass
particles and have a sharp base and a gradational top (AL in
Table 3); (2) disseminated ash layers contain 10%-80% glass
particles (DA in Table 3); and (3)
ash-enriched layers are deposits with 10% glass particles
(EL in Table 3).
Ash layers from Site 982 were sampled in lithologic Subunits IIA-IID, which range from the middle Miocene to the lower Pliocene (~18.8-~5 Ma). These sediments comprise a sequence of marine nannofossil oozes and chalks with minor amounts of clay, clayey nannofossil mixed sediments, and clays, with variable amounts of nannofossils and silt-all deposited above the calcite compensation depth.
Tertiary ash layers of Site 985 are distributed throughout lithologic Units II-V, which range in age from late Oligocene to early Pliocene (<24-~5 Ma). The sediments are predominantly fine-grained siliciclastics (silty clays; clays with silt and clays). Biogenic carbonate in significant amounts (>10%) is restricted to the upper part of the sequence (Units I-III). The increasing silt content of Unit III sediments may mark the transition from the relatively ice-free conditions of the Miocene to the weak Northern Hemisphere glaciation from the late Pliocene to the Holocene (Shipboard Scientific Party, 1996b).
The oldest sampled ash layers belong to the Oligocene (Hole 985A, Unit V). These ashes were not analyzed because of their high degree of alteration. The oldest fresh ash layers occur in Hole 983A (Subunit IVB) and Hole 982B (Subunit IID) and are middle Miocene in age (Fig. 2).
Nearly three-quarters of the ash deposits studied here are discrete ash layers (tephra) with sharp bases and gradational or diffuse upper boundaries (AL in Table 3). The layers are <1-24 cm thick and consist of 80%-98% of colorless and/or brown glass particles. Rock fragments, crystals (mostly feldspar, with minor amounts of pyroxene, amphibole, olivine, quartz, mica, zeolites, glauconites, and opaque minerals), and fossils are subordinated. The colorless glass is dominated by highly vesicular fragments, Y-shaped bubble-wall shards, and pumice fragments. Colored glass mostly has a blocky shape. Compared to ash layers from Legs 151 (Lacasse et al., 1995, 1996; Werner et al., 1996) and 152 (Werner et al., 1998; Lacasse et al., 1998), greater amounts of highly vesicular and/or pumiceous brown glass shards are present (Table 3). The high glass content suggests that these ash layers are primary deposits or are volcanic deposits that are minimally reworked.
About one-quarter of the studied ash deposits are ash-bearing to ash-rich sediments, with 5%-80% glass content mixed with nonvolcanic material (DA in Table 3). These layers are interpreted as epiclastic deposits, transported and deposited as turbidites.
Unique when compared with other ODP sites in the North Atlantic (e.g., Bitschene et al., 1989; Lacasse et al., 1996; Lacasse et al., 1998; Werner et al., 1996; Werner et al., 1998) is the low abundance of pure mafic ashes (~10% of all analyzed ash layers) and the relatively high abundance of bimodal ashes (~45%) in the Miocene sequence of Sites 982 and 985. Also remarkable is the high amount of strongly altered ash at Site 985.
The analytical totals of major-element analyses (electron microprobe) for the lower Pliocene-middle Miocene ash layers from Sites 982 and 985 range from ~88 to 96 wt% for felsic glasses and from 96 to 100 wt% for mafic and intermediate glasses (with iron calculated as FeO* = total Fe as [FeO + Fe2O3]). The low totals of felsic glasses probably result from the high content of initial volatiles of up to 7 wt% H2O (Bitschene and Schmincke, 1990) and/or are a result of hydration and alteration with a significant loss of alkali elements (Na2O, and K2O). Therefore, analyses with totals <93 wt% (felsic glasses) and <98 wt% (mafic and intermediate glasses), respectively, are not shown in the diagrams presented in this report.
Figure 3 shows the total alkali vs. silica distribution of the analyzed samples. The geochemical spectrum ranges from basalt to low-alkali and high-alkali andesites, andesites, dacites, and rhyolites. Except for two layers depleted in alkali elements (Samples 162-985A-26X-6, 0-3 cm, and 28X-5, 22-27 cm), the alkali vs. silica distributions of the ash layers of Sites 982 and 985 are very similar. More than 90% of brown glass fragments of the two sites are basaltic to basaltic-andesitic, and nearly 100% of the colorless glasses are rhyolitic. Two layers from Site 982 and one from Site 985 are andesitic; three fragments of one layer of Site 985 are dacitic.
At Site 982, ~60% of ash layers consist of one or two chemically homogeneous fractions. Half of these layers contain one to three fragments with a different chemical composition (indicated by an asterisk in Table 4 and Table 5). The remaining ash layers show a wide variability in chemical composition. In one of these layers (interval 162-982B-18H-6, 49-50 cm), a trend from mafic to felsic compositions is recognized.
The amount of chemically more or less homogeneous tephra is somewhat higher at Site 985 (~70%). A trend from mafic to felsic compositions is visible in two layers (intervals 162-985A-23X-11, 12-14 cm, and 26X-6, 55-57 cm; Table 5). The remaining ash layers are highly variable in chemical composition.
Figure 4 shows the analyses of individual mafic to intermediate glass shards; SiO2, Al2O3, CaO, FeO*, K2O, and TiO2 are plotted vs. MgO, an index of differentiation. All analyses show increasing SiO2 and K2O and decreasing CaO, FeO*, and TiO2 as MgO decreases.
Most of the rhyolitic glass shards are compositionally homogeneous. Figure 5 depicts the major-element composition of only the rhyolitic glasses; Al2O3, CaO, FeO*, and K2O are plotted vs. SiO2. With a few exceptions, Al2O3 and FeO* decrease with increasing SiO2, CaO is relatively constant and low (<3 wt%), but K2O ranges between ~2 and ~5 wt% and has no significant trend. A few glasses plot outside of the major groups. Most of the ash belongs to the low-potassium series (<3.8 wt% K2O) defined by Sigurdsson and Loebner (1981) for North Atlantic ashes. All rhyolites have an Icelandic affinity (Fig. 3, Fig. 6). Moreover, most analyses match the field for highly evolved rocks of the Tertiary Icelandic central volcanoes Thingmuli and Breiddalur (Fig. 6A, Fig. 7; Walker, 1963; Carmichael, 1964). The basaltic to andesitic and dacitic glasses also correspond to mafic and intermediate rocks of these Tertiary Icelandic central volcanoes, at least in major-element composition (Fig. 6A-D, Fig. 7). Furthermore, the glasses clearly follow the Thingmuli tholeiitic trend (Carmichael, 1964) as shown in the FeO* vs. MgO plot (Fig. 6D). Even though the comparison of whole-rock analyses (data from Carmichael, 1964, and Walker, 1963) and analyses of matrix glasses (this study) is difficult because of the effect of crystallization, the correspondence in concentrations of incompatible elements (e.g., TiO2 and K2O) indicates an Icelandic source area.
