The agreement between middle and late Miocene diatom ages and 40Ar-39Ar ages reinforces the reliability of several diatom datums in this interval. As discussed by Ramsay and Baldauf (1999), many Neogene diatom datums still require precise calibration to the polarity timescale or strontium isotope stratigraphy, and calibration in multiple drill cores is necessary in order to ascertain the synchroneity of datums between different regions of the Southern Ocean. It appears, however, that most Miocene diatom bioevents at Kerguelen Plateau sites do not display the degree of diachronism interpreted by Ramsay and Baldauf (1999). Censarek and Gersonde (2002) reach a similar conclusion for sites in the Atlantic sector of the Southern Ocean, further reinforcing the synchroneity of Miocene datums and general utility of diatom biostratigraphy in Miocene sequences from the Southern Ocean.
In the present study, a calibration of the LO of N. denticuloides is determined at Site 1138 through 40Ar-39Ar dating of a tephra layer in Core 183-1138A-19R. The LO of N. denticuloides occurs between 171.70 and 170.20 mbsf, and a mean age of 11.45 ± 0.22 Ma is derived from the tephra horizon at 170.81 mbsf (Fig. F7). In Hole 747A, the LO of N. denticuloides occurs between 61.47 and 59.97 mbsf (Harwood and Maruyama, 1992), in association with the Subchron C5An.1n/C5r.3r reversal at 61.30 mbsf (Fig. F7). The age calibration for this reversal is 11.94 Ma (Berggren et al., 1995), and the LO of N. denticuloides is assigned a calibrated age of ~11.7 Ma, due to the sample spacing uncertainty. Therefore, the age calibration in Hole 747A of ~11.7 Ma is in close agreement to that of ~11.5 Ma determined in Hole 1138A.
A middle Miocene tephra horizon that occurs in association with the LO of N. denticuloides is also present in Hole 747A at 60.90 mbsf (Fig. F7). The middle Miocene tephra layer at Site 747 is more trachytic in composition (Morche et al., 1992) than the Site 1138 ash, but it is possible that the two horizons represent the same eruptive event. The LO of N. denticuloides falls immediately below the tephra horizons in both drill cores (Fig. F7), although there is some uncertainty in the placement of the datum levels because of sample spacing gaps. In spite of this uncertainty, the tephra horizons are correlative within the limits of current age control for the two drill cores (Fig. F7). A middle Miocene tephra layer was also reported at Site 737 (Northern Kerguelen Plateau) in Section 119-737B-5R-1 (Morche et al., 1991). This tephra, however, falls well above the LO of Denticulopsis dimorpha (Baldauf and Barron, 1991) and is therefore late Miocene in age and not correlative with the ash horizons at Sites 747 and 1138.
The LO of N. denticuloides is well documented from drill cores on Maud Rise (Leg 113) and Meteor Rise (Leg 177) (Gersonde and Burckle, 1990; Censarek and Gersonde, 2002). In the Maud Rise sections, Censarek and Gersonde (2002) recalibrated the LO of N. denticuloides to the Berggren et al. (1995) timescale. An age of 11.82 Ma is calculated for this datum in Hole 689B, and an age of 11.78 Ma is calculated in Hole 690B. A similar age of 11.86 Ma is also determined at Site 1092, which is located at a more northerly position within the modern PFZ (Censarek and Gersonde, 2002). These ages are in general agreement with those derived from Kerguelen Plateau sections.
The upper Pliocene-Pleistocene diatom zonation for the Southern Ocean is presently defined by datums that are widely separated in time. If the LO of Fragilariopsis barronii is not utilized in the zonal scheme (due to taxonomic difficulties in separating from F. barronii from F. kerguelensis), a large gap is present between the LO of Thalassiosira kolbei (1.9 Ma) and the LCO of Actinocyclus ingens (0.66-0.70 Ma). Future detailed studies on the Kerguelen Plateau will hopefully identify other datums within this interval, such as the first abundant occurrence datum of Thalassiosira elliptipora (~1.1 Ma) utilized by Zielinski and Gersonde (2002). Distinct Pleistocene acmes of Actinocyclus ingens were not identified in Hole 1138A, which would have allowed application of ages derived from northern Southern Ocean sites (Gersonde and Bárcena, 1998). Also, Pleistocene samples from Hole 1138A contained only rare specimens of Fragilariopsis matuyamae, a taxon that has been shown to be biostratigraphically useful at northern Antarctic and Subantarctic sites (Gersonde and Bárcena, 1998; McMinn et al., 2001). An uppermost Pleistocene acme of Hemidiscus karstenii, however, did allow application of the Southern Ocean LO datum (0.21 Ma) for this taxon.
Several diatom events recognized in Hole 1138A are identified as potentially useful datums for future studies on the Kerguelen Plateau and application in a revised Neogene diatom zonation. Useful markers in the upper Miocene include the FO of Hemidiscus karstenii f. 1 (of Ciesielski, 1983), the LO of Denticulopsis crassa, and the FO and LO of Hemidiscus triangularus. The FO of H. triangularus, for example, has already been applied in revised zonal schemes for the Atlantic sector of the Southern Ocean (Censarek and Gersonde, 2002). Additionally, the FO of Thalassiosira lentiginosa var. ovalis and the FO and LO of Fragilariopsis heardensis are potentially useful in the upper Pliocene, and the LO of Rouxia antarctica may be of use in the Pleistocene.