Fig. 1. Locations of Sites 1088, 704 and 1172 in the Southern Ocean with respect to Southern Hemisphere summer (A) and winter (B) temperatures (Levitus and Boyer, 1994). See Table 1 for a summary of the precise location and water depths.

Fig. 2. Derivation of the late Miocene portion of the age model for Site 1088 (B) based on a comparison to orbitally tuned benthic foraminiferal δ¹³C records (A). The horizontal grey bars highlight the timing of the late Miocene δ¹³C maximum (7.6 Ma) and two of the δ¹³C minima that are used as age control points (see also Table 2). The long term trend in the published Site 1172 Globigerina bulloides δ¹³C record (Tedford and Kelly, 2004) (C) agrees and no revisions to the published age model are justifiable. The geologic boundaries reflect the ages according to the Gradstein et al. (2004) time scale.

Fig. 3. Derivation of the Pliocene portion of the age model for Site 1088 based on comparison to orbitally tuned benthic foraminiferal δ¹⁸O record from the deep Pacific (Mix et al., 1995, Hodell and Venz–Curtis, 2006). Four δ¹⁸O maxima in the Site 1088 δ¹⁸O record (A) can be easily adjusted to Site 849 (D). Accordingly we make a few minor adjustments to the Globigerina bulloides δ¹⁸O record from Site 704 (B). The Site 1172 record is not long enough to justify any adjustments to the Pliocene portion of the record. The geologic boundaries reflect the ages according to the Gradstein et al. (2004) time scale.

Fig. 4. ODP Leg 177 shipboard biostratigraphic datums versus depth (solid circles) and additional age control points based on the high resolution Cibicidoides δ¹³C (late Miocene) and Globigerina bulloides δ¹⁸O (Pliocene) records (open squares). For comparison, the aged derived from the lower resolution Cibicidoides record are indicated with the crosses. The geologic boundaries reflect the ages according to the Gradstein et al. (2004) time scale.

Fig. 5. Comparison of stable isotope records from Globigerina bulloides and Cibicidoides at Site 1088 (A and B, respectively), Site 704 (C and D), Site 1172 (E, and F). Note that for ease of comparison with the δ¹³C records, the δ¹⁸O values are plotted with higher values up in this figure. The geologic boundaries reflect the ages according to the Gradstein et al. (2004) time scale. Heavy lines reflect a Gaussian smoothing function using a 1% sampling portion.

Fig. 6. Comparison of the long term δ¹³C (A) and δ¹⁸O (B) trends from Globigerina bulloides at Sites 1088, 704, and 1172 to the long term trend recorded in the deep Pacific by benthic foraminifera Cibicidoides (Site 849, [Mix et al., 1995] and [Hodell and Venz-Curtis, 2006]). Long term trends are derived from a Gaussian smoothing function using a 5% sampling portion. Note that in (A), in order to visualize the relative changes that occurred in the surface versus the deep records, the long term trends are plotted as to overlay the δ¹³C maxima at 7.6 Ma. Horizontal lines in (B) indicate the δ¹⁸O values of late Holocene samples available for Site 1088 from nearby RC11-83 (Charles et al., 1996); for Site 704 published by Hodell and Venz (1992), and for Site 1171 from Weaver et al. (1997).

Fig. 7. Scatter plots of planktic and benthic foraminiferal δ¹⁸O versus δ¹³C values at Sites 1088 (A), 704 (B) and 1172 (C). The correlation coefficients from the linear regression are provided for each data set (see text for details).

Site summary

Summary of revised age control points for Site 1088