We tie our Leg 178 biochronology to the geomagnetic polarity timescale (GPTS) of Berggren et al. (1995) (BKSA95), which uses the GPTS of Cande and Kent (1995) (CK95). BKSA95 use the astronomically calibrated timescale for polarity boundaries from Chron 1n to Subchron 3n.4n (0-5.23 Ma) (Shackleton et al., 1990; Hilgen, 1991a, 1991b). Extension of the astronomical timescale into the Miocene epoch (Shackleton et al., 1995 [Leg 138]; Shipboard Scientific Party, 1995 [Leg 154]) and fine-tuning of the chronology of the astronomical timescale (e.g., Langereis et al., 1994) is an ongoing process. The differences between astronomical and geomagnetic timescales are <0.25 m.y. for the time interval between the late Miocene and the Quaternary.
For this paper absolute microfossil datum ages are from BKSA95 or were converted from published age estimates based on Berggren et al. (1985a, 1985b) to the timescale of Cande and Kent (1995) through linear interpolation between the nearest geomagnetic reversal boundaries. The biostratigraphic zonation for different groups employed on Leg 178 sediments are summarized in Figure F2. Sources of biostratigraphic ages (Table T2) are as follows.
Numerous diatom biostratigraphic studies have been completed for the Southern Ocean (Gersonde and Burckle, 1990; Baldauf and Barron, 1991; Harwood and Maruyama, 1992; Gersonde and Barcena, 1998; Gersonde, Hodell, Blum, et al., 1999; Ramsay and Baldauf, 1999). The Neogene and Quaternary diatom zonal scheme used during Leg 178 was that proposed by Harwood and Maruyama (1992 [Leg 120]) with the modifications of Winter and Iwai (Chap. 29, this volume) and Iwai (2000a, 2000b).
Abundance and state of preservation vary greatly within Leg 178 sites, with intervals of rare siliceous microfossils noted in the middle Pliocene and Pleistocene. Taxonomic concepts for diatoms from Leg 178 sediments are discussed in Iwai and Winter (Chap. 35, this volume).
Radiolarians were initially studied by Weinheimer during the Leg 178 cruise (Barker, Camerlenghi, Acton et al., 1999). More detailed analyses of the upper Miocene to lower Pliocene interval between 4 and 9 Ma (98 samples from Cores 178-1095B-5H through 52X) were completed by Lazarus (Chap. 13, this volume).
The radiolarian biostratigraphic scheme proposed by Lazarus (1990, 1992) was applied to Leg 178 material. This zonation is based on the earlier schemes of Hays (1965), Chen (1975), Weaver (1976), Keany (1979), and Caulet (1991) and refined by Lazarus (1992) using sediments recovered during Legs 119 and 120. Species concepts follow those used by Keany (1979), Lazarus (1990, 1992), and references therein. Stratigraphic constraints of radiolarian events are based on examination of one sample per core (core catcher) for Sites 1096 and 1101 and two to four samples per core for Site 1095 (Lazarus, Chap. 13, this volume).
The abundance of calcareous nannofossils is generally low throughout the section of all three rise sites, except in the Pleistocene-late Pliocene biocalcareous interval of Sites 1096 and 1101 (Shipboard Scientific Party, 1999a). Remarks regarding the biostratigraphic zonations and calcareous nannofossil datums used in this paper can be found in Winter and Wise (Chap. 26, this volume) for Sites 1096 and 1101 and in Iwai et al. (Chap. 28, this volume) for Site 1095 and continental shelf sites.
The zonal schemes used for Leg 178 were those of Martini (1971) and Bukry (1973, 1975, 1978), code numbered by Okada and Bukry (1980). In addition to the traditional use of first/last occurrences of index species, ranges of other taxa were used to improve the stratigraphic resolution of the Pleistocene-late Pliocene interval (Thierstein et al., 1977; Gartner, 1977, 1990; Pujos, 1988; Raffi et al., 1993; Wei, 1993; Takayama and Sato, 1987; Raffi and Flores, 1995).
Several zonal schemes have been developed for the mid and high latitudes of the Southern Hemisphere (e.g., Jenkins and Srinivasan, 1986; Berggren et al., 1995). However, these schemes are not fully applicable to the planktonic foraminiferal fauna in all the sediments recovered during Leg 178 because of the absence or low abundance of foraminiferal species. In general, high-latitude foraminiferal assemblages contain low-diversity and long-ranging species that are of limited biostratigraphic use. Planktonic foraminiferal zonation and classification follow Berggren (1992), developed from Leg 120 Sites.
The polarity zones recorded in the drift sediments can be correlated to the GPTS with little or no ambiguity, particularly for Sites 1096 and 1101, as discussed in the revised magnetostratigraphic interpretations of Acton et al. (Chap. 37, this volume). As noted in the introduction, Acton et al. (Chap. 37, this volume) make use of the revised biostratigraphic events and datums from this study, resulting in an internally consistent interpretation for the biomagnetostratigraphy. They suggest that neither hiatuses nor abrupt large changes in sedimentation rates are needed to match the polarity zones to the GPTS, although such an interpretation gives rise to discrepancies in the biostratigraphic datums for part of the Site 1095 sedimentary section.
The discrepancies may arise for several reasons (see also "Synthesis and Discussion" ), some of which may be related to magnetostratigraphy from Leg 178 or previous drilling. In particular, biostratigraphic datums may have been miscorrelated with the GPTS in prior studies owing to incomplete recovery, hiatuses, and sediments that are less than ideal paleomagnetic recorders, all of which can lead to inaccurate magnetostratigraphies. Such problems could also affect the Leg 178 sites. Thus, we also investigate an alternate interpretation of the paleomagnetic data for Site 1095, which includes a hiatus near the Miocene/Pliocene boundary.
Depths are given in meters below seafloor (mbsf), which is based on drill pipe measurements specific to each drill hole, and, when necessary, in meters composite depth (mcd), which is a depth scale common to all holes at a site. The mcd scale overcomes some of the inaccuracies in the mbsf scale, but more importantly, allows coeval stratigraphic features recovered in more than one hole to be placed at a common depth (e.g., Barker, Chap. 6, this volume). Such features could be misaligned in depth by several meters in the mbsf depth scale as illustrated for polarity reversals in Figure F12 of Acton et al. (Chap. 37, this volume).
We use the mcd scales for Sites 1095 and 1096 constructed by Barker (Chap. 6, this volume). Using tables in his paper, the mbsf depth can be converted to mcd by adding or subtracting the appropriate offset in meters. The offsets vary in the upper part of Sites 1095 and 1096. Because the deeper portions of both sites are single cored, a constant offset is used for most of the section. For example, at Site 1095, only Hole 1095B was cored below ~90 mbsf. Thus, below the top of Core 178-1095B-2H at 92.5 mbsf, a constant offset of -5.5 m is applied to obtain mcd depths (e.g., 92.5 mbsf = 87.0 mcd). Similarly, only Hole 1096C was cored below ~260 mbsf. Thus, below the top of Core 178-1096C-8X at 260.9 mbsf, a constant offset of +5.54 m is applied to obtain mcd depths. Finally, no mcd scale is needed for Site 1101 because only a single hole (1101A) was cored. All depths for Site 1101 are therefore given in mbsf only.
Offsets also occur between depths from wireline measurements made during downhole logging and the drill pipe measurements. For Hole 1095B, Acton et al. (Chap. 37, this volume) determined the offsets by correlating whole-core susceptibility with susceptibility measured on the second logging run of the Geologic High-Resolution Magnetic Tool (GHMT). The depth offsets between whole-core susceptibility events and logging events are fairly constant, with the logging depths being on average 4 to 6 m deeper than mbsf depths and 9 to 11 m deeper than mcd depths.