Early Oligocene to Pleistocene calcareous nannofossils at Sites 1146, 1147, and 1148 in the northern SCS (Fig. F1; Table T1) were analyzed. Sample spacing is ~100 cm for Site 1146 and 150 cm for Sites 1147 and 1148. A total of 1200 samples were examined.
Site 1146 is located at a water depth of 2092 m within a small rift basin on the mid-continental slope of the northern SCS (Fig. F1). Coring at Site 1146 recovered a 600-m-long, lower Miocene through Pleistocene section of relatively carbonate rich, hemipelagic nannofossil clays with a basal age of ~19 Ma. Site 1146 provides one of the most continuous Neogene sections ever recovered by ODP. Sediments recovered at Site 1146 yield abundant nannofossils that are generally well preserved above 531.2 mcd but are moderately overgrown below that level. Many samples exhibited some degree of reworking. For this reason, a number of last occurrence (LO) events at this site were determined based on semiquantitative methods.
Site 1148 is located at a water depth of ~3294 m on the lowermost continental slope off southern China near the continent/ocean crust boundary and is the most offshore site drilled during Leg 184 (Fig. F1). Site 1147 is located at a water depth of ~3245 m ~0.45 nmi upslope from Site 1148 (Fig. F1) and was designed to recover the uppermost section that appeared to be missing on seismic profiles from Site 1148 (Wang, Prell, Blum, et al., 2000). Sediments at Site 148 yielded abundant nannofossils with preservation that varied downhole. Reworked nannofossils are commonly seen at this site, particularly in early Miocene to late Oligocene Cores 184-1148A-44X, 48X, and 49X. Pronounced reworking in these intervals agrees well with sedimentary observation of redeposition and slumping there.
Traditionally, ODP samples have been referenced to meters below seafloor (mbsf). Over the last decade, meters composite depth (mcd) scales have been developed to obtain complete stratigraphic sections for Legs 138, 154, 162, 172, and 177, which were designed for high-resolution paleoceanographic studies (e.g., Curry, Shackleton, Richter, et al., 1995; Gersonde, Hodell, Blum, et al., 1999). Mcd scales were also constructed during Leg 184, through correlation of stratigraphic intervals from two or more holes cored at the same site. Depth offset and mcd splice tables for each site were produced for transferring the original mbsf of samples into the mcd scale in a composite section (Wang, Prell, Blum, et al., 2000).
Standard smear slides were prepared for analysis. Calcareous nannofossils were examined using standard light microscope techniques under crossed polarizers and transmitted light at 1000x magnification. A few of the late Pleistocene samples were examined by means of scanning electron microscope (SEM) at 10,000x magnification to study the first occurrence (FO) of Emiliania huxleyi.
Two semiquantitative methods were used when necessary: (1) Relative abundance very roughly estimated (at 1000x magnification) for an ODP standard five-category scheme: dominant (>50%), abundant (10%–50%), common (1%–10%), few/frequent (0.1%–1%), and rare (<0.1%); and (2) average number of nannofossil specimens per field of view, estimated by counting >20 fields of view (at 1000x magnification).
The biostratigraphic zonations proposed by Martini (1971) and Bukry (1973, 1975) have been the current standard zonations for Cenozoic calcareous nannofossils. Both standards were slightly modified; for example, Martini's zonations were modified by Martini and Müller (1986) and Bukry's were modified to employ code numbers by Okada and Bukry (1980). The zonation of Martini (1971) has been used for a wide geographic range, whereas the zonation of Okada and Bukry (1980) has been more applicable for low latitudes. We employed both zonations in our investigation (Fig. F2). For convenience, the zone of Martini (1971) was described in the first order.
Cenozoic calcareous nannofossil events have been correlated to magnetostratigraphic and oxygen isotopic stratigraphic data (Backman and Shackleton, 1983; Wei, 1989; Backman and Raffi, 1997). The standard zonations of Martini (1971) and Okada and Bukry (1980) were correlated with the geomagnetic polarity timescale of Cande and Kent (1992, 1995). Results of these studies were summarized by Young et al. (1994) and then by Berggren et al. (1995). These data were employed in this paper (Fig. F2).
Ages of Cenozoic chronostratigraphic boundaries were established by Berggren et al. (1995). According to this study, the age of the Oligocene/Miocene boundary is 23.80 Ma, the age of the Miocene/Pliocene boundary is 5.32 Ma, and that of the Pliocene/Pleistocene boundary is 1.77 Ma, as given in Figure F2.
Identification of most calcareous nannofossils primarily follows the compilation of Perch-Nielsen (1985). Only a few of species are discussed here.
Several Gephyrocapsa species are commonly used as biostratigraphic markers. However, morphologic intergradations exist between species, leading to confusion in identification (Su, 1996). A detailed study is needed to find reliable and applicable criteria to separate them. Thus, only one morphological group, Gephyrocapsa (medium) spp., which is relatively easy to identify, was used in this study. This group includes very early Pleistocene forms of Gephyrocapsa magereli and Gephyrocapsa lumina, and their variants, with a maximum coccolith length >3.5 µm.
A number of Reticulofenestra species—Reticulofenestra asanoi, Reticulofenestra pseudoumbilicus, and Reticulofenestra umbilicus, for example—have been used as Tertiary and Quaternary biostratigraphic markers. They are mostly distinguished by coccolith size (cutoff size), showing a great range of variation in these parameters and causing problems in identification (Backman, 1980; Gallagher, 1989; Young, 1990; Su, 1996). To ensure taxonomic consistency with previous work, we followed the commonly accepted cutoff sizes: the coccolith length of R. asanoi is >5 µm (Sato and Takayama, 1992), and that of R. pseudoumbilicus is >7 µm, in accord with the size of the holotype (Gartner, 1967). Typical R. umbilicus can be easily distinguished from other species by its enormous size. However, morphological intergradations exist between this species and other smaller forms, such as Reticulofenestra dictyoda or Reticulofenestra coenura. Backman and Hermelin (1986) suggested using a lower size limit of 14 µm for recognition of R. umbilicus, and this suggestion has been widely accepted. Reticulofenestra hillae, a large elliptical form, has been considered as an eco-phenotype or a variety of R. umbilicus (Backman and Hermelin, 1986; Berggren et al., 1995). In agreement with these studies, we combined R. umbilicus and R. hillae as a group named R. umbilicus with a cutoff size of 14 µm.
In this study, a number of Sphenolithus species or taxa were employed following Perch-Nielsen (1985). Only the separation of a group of morphologically and phylogenetically related species, Sphenolithus predistentus-Sphenolithus ciperoensis lineage, followed Moran and Watkins (1988). Okada (1990) further observed morphological variances of S. ciperoensis and Sphenolithus distentus from the tropical and subtropical Indian Ocean and named them Sphenolithus aff. ciperoensis and Sphenolithus aff. distentus. Sphenolithus aff. ciperoensis resembles S. ciperoensis, but the apical spine is shorter and wider than the type species. Furthermore, the basal spines of the former are thinner and arranged to form a more widely spread base than the latter. Okada (1990) also found that the V-shaped basal line of Sphenolithus aff. distentus is similar to the typical S. distentus, but the apical spine and overall size of the former is larger (7–12 µm compared to ~5 µm). Sphenolithus aff. ciperoensis and Sphenolithus aff. distentus are also present in our Oligocene materials, hampering identification and determination of the FO of S. ciperoensis and S. distentus. To separate these variants, we followed the criteria suggested by Okada (1990).