AGE MODELS AND SEDIMENTATION RATES

One of the discoveries of DSDP Leg 94, which explored the Feni and Gardar sediment drifts in the North Atlantic Ocean, was the presence of remarkable continuous stratigraphic sections, with few, if any, discernible hiatuses through the Pliocene-Pleistocene (Kidd and Hill, 1986). With the drilling of Site 1123, in hemipelagic drift sediments on north Chatham Rise, scientists on Leg 181 have now recovered similar continuous sections, not only for the Pliocene-Pleistocene, but also for most of the Miocene, starting near 20 Ma. Thus, Site 1123 offers the potential to improve the early and middle Miocene time scale (Hilgen et al., 1995), particularly if the carbonate/clastic cycle succession can be reconstructed for that stratigraphic interval. The quality of this carbonate record will particularly depend on continuous detail in the GRAPE density data, the downhole logging, and the hole splices. If it is possible to establish the ~20-m.y. Neogene cycle record by several parameters (reflectance, magnetic susceptibility, and others) it may be possible to correlate this to a calculated insolation tuning target (Laskar et al., 1993) to establish an astronomical time scale back into the late early and middle Miocene. The first order ties of the magnetostratigraphic and biostratigraphic datums to the cycle chronology can then be compared with those established for pieces of its record from other sites around the world (Hilgen et al., 1991a, 1991b, 1995; Shackleton et al., 1995; Shackleton and Crowhurst, 1997; Naish et al., 1997), and used for a Neogene time scale standard for southern Pacific stratigraphies. A remarkable set of data already exists, using information available from onboard studies, yielding a robust age-depth model based on magnetostratigraphic and biostratigraphic datums.

The combined nannofossil, foraminifer, diatom, and radiolarian biostratigraphy at Site 1123 yielded 112 event levels with a preliminary age assignment, using the shipboard stratigraphic framework (see "Biostratigraphy"  in the "Explanatory Notes" chapter). The 112 levels are shown in Table T15, and consist of 51 FO events, 11 acme, FCO, or LCO events, and 50 LO events. Note that a tentative age of the FO of Globoquadrina dehiscens in Sample 181-1123C-28X-CC of 25 Ma cannot be verified without a continuous stratigraphic section across this interval. Hence, this level was not taken into account in the calculation of average sedimentation rates (see below). In Figure F29, all levels are plotted according to their observed depths at Site 1123 and ages defined in "Biostratigraphy" in the "Explanatory Notes" chapter. The position of the arrows in Figure F29 may be extended uphole (downhole) for last occurrences (first occurrences) because of the limited sampling density for shipboard analysis. For comparative purposes, the magnetostratigraphic age model (see "Paleomagnetism") is shown as a line in Figure F29. Because of the size of the plot, strongly or fully overlapping event positions may be obscured in the graphs. All depths are reported in meters composite depth (mcd; not mbsf) values.

Scrutiny of Figure F29 allows the following preliminary observations to be made. The event distributions in time show no obvious outliers, except for several first occurrences of taxa near 80, 205, and 450-460 mcd. A principal break in sedimentation exists near 560 mbsf. This level also marks a change from a low to a much higher rate, above which level the average rate is fairly steady, without any long-term trend. The plot of fossil events vs. sampling depth in the Miocene shows gaps between 14 and 15 Ma, 8.5 and 9.5 Ma, and 6 and 7 Ma. The gaps may change somewhat from changes in event age calibration, but also may have an underlying paleoceanographic meaning. Colder water-mass incursions might have limited immigration of taxa during these relatively short periods.

In the mid-Pliocene interval, between 80 and 120 mcd, LO events seem to "bunch up," as if terminated suddenly. A possible reason is the onset of severe paleoclimatic change, that is, drastic water-mass cooling and severe changes in surface circulation patterns, which affected evolutionary change. An increase in sampling resolution would not change the above observations, but changes in age calibration would do so.

Scatter of the event distribution in the vertical sampling depth scale will diminish with higher resolution sampling. FO events reported in the core catchers may have been estimated to be too shallow by several meters, which may pertain to the FOs of Globorotalia amuria (~16 Ma) at 457.69 mcd, of Cibicidoides wuellerstorfii (16.4 Ma) at 469.15 mcd, of Orbulina suturalis (15.1 Ma) at 440.35 mcd, of Zeagloboquadrina nepenthes (11.8 Ma) at 366.83 mcd, of Sphaeroidinellopsis paenedehiscens (~8 Ma) at 261.36 mcd, of Sphaeropyle langii (6.0-6.2 Ma) at 203.31 mbsf, and of Globorotalia tosaensis (3.2 Ma) at 75.87 mcd (Fig. F29). In particular, the obvious and relatively large offset between the polarity chron trend and the six foraminifer FO events between 15 and 17 Ma may require adjustment, once more samples between the core catchers are studied. On the other hand, the G. amuria and O. suturalis events, among these six, plus the G. tosaensis LO event, are relatively rare, defined by few specimens per sample, which hampers a search for the "top." Thus, no correction may be forthcoming for these event positions.

LO events reported in the core catchers may have been estimated to be too deep by at least several meters and include the last appearance of Sphenolithus heteromorphus (13.57 Ma) at 469.15 mcd, of Globorotalia conica (~11.5 Ma) at 428.6 mcd, the acme (10.2 Ma) and the last occurrence (9.9 Ma) of Cyrtocapsella japonica at 400.34 and 394.14 mcd, respectively, the last occurrence of Discoaster quinqueramus (5.56 Ma) at 232.49 mcd, of Lychnodictium audax (3.7 Ma) at 167.57 mcd, and of Stichocorys peregrina (<3Ma) at 149.38 mcd. Berggren et al. (1995b) discuss at length problems around the diverging age assignments for the last occurrence of the nannofossil S. heteromorphus. If there is no change in sampling depth of the event, the Site 1123 evidence suggests the event was estimated to be too young by 0.5 Ma.

Forty-six calcareous microfossil events that were considered to be relatively reliable, at least for local age calibration, were combined with the magnetostratigraphy. The paleomagnetic age model at Site 1123 is shown by a solid line in Figure F29. No polarity chrons appear to be missing (see "Paleomagnetism"), and the record tracks the central tendency of the "events channel" remarkably well. The already rather tight fit of the age assignments for the microfossil events and the polarity chrons provides an excellent opportunity to use the record in Figure F29 for future biochronologic calibration in the southwestern Pacific realm and possibly beyond. For example, the LOs of Globorotalia puncticulata and of Zeaglobigerina druryi appear too old by 0.6 Ma or more, whereas the FO of Sphaeroidinellopsis paenedehiscens at ~8 Ma might shift to 7 Ma. More detailed studies are needed to resolve these differences. Ultimately, a table may be drawn up of recalibrated event ages in the temperate to subtropical Southwest Pacific, as compared also to the North Atlantic realm, involving all microfossil groups studied onboard ship.

Sedimentation rates through time at Site 1123, using a slightly smoothed data set of polarity chron ages vs. depth for the Neogene and biostratigraphic age interpretation for older strata, are shown in Figure F30. The rates are corrected for compaction in the following manner. First, over 200 porosity measurements from Site 1123 (see "Physical Properties") were analyzed for trends, using the programs DEPOR and BURSUB (Stam et al., 1987; Gradstein et al., 1989). Grain density was averaged at 2700 kg/m³. A clear trend in porosity with depth exists, with porosity exceeding 71% at the top, and decreasing more or less linearly to almost 30% at the bottom of the Hole 1123C. Program runs were executed for the fine-grained lithologies at the site. Next, compacted and restored rates of sedimentation were derived, with the age intervals slightly smoothed, as shown in Table T16. Thickness restoration resulted in a near-doubling of the older Neogene and Paleogene rates, from 5-25 m/m.y. to 15-35 m/m.y., with negligible effects on younger rates; this is in accord with the measured porosity-depth trend. In general, higher sedimentation rates, up to 45 m/m.y. on average in the late Neogene, were preceded by average rates of 35 m/m.y. in the early Neogene, and ~15 m/m.y. in the late Eocene and early Oligocene.

A simple burial diagram, without paleo-water depth but corrected for compaction, is shown in Figure F31. Age levels used are five or more meters apart. Burial was remarkably steady and constant for this drift.

The principal feature of the sedimentation rate and age-depth curves (Figs. F23, F29) is remarkable uniformity over the past 20 Ma. The evidence suggests that the mixture of biopelagic skeletal remains and hemipelagic terrigenous sediment has remained rather constant over this period, testifying to long-term stability in both the productivity regime and the sediment delivery system. The early Miocene rate of ~15 m/m.y. ends at a marked slow-down in sedimentation at 15-13.5 Ma corresponding to the middle Miocene shift in ¹⁸O and growth in the Antarctic Ice Sheet. It remains to be determined whether the effect was to increase the flow speed and reduce sedimentation rate through winnowing or to decrease flow rate and, consequently, sediment delivery. Following this near-hiatus, sedimentation rate increases sharply to 38 m/m.y. and from 12 Ma steadies at 28 m/m.y. until 7 Ma, when there is a further sharp increase to ~47 m/m.y. This probably corresponds to increased delivery of sediment to Bounty Trough caused by sharp uplift of the Southern Alps in response to a pronounced phase of compression deformation along the plate boundary (Walcott, 1998). From 5 Ma to the present a constant sedimentation rate of 36 m/m.y. has prevailed with no apparent interference from climatic or tectonic changes. This probably reflects the constancy of the sediment delivery system dominated by the DWBC. Three phases or pulses of volcanism are recognized: a late Miocene phase (7.5-9.0 Ma; 280-320 mcd) probably coincident with early compression at the New Zealand plate boundary (Walcott, 1998), a phase in the mid-Pliocene (3.1-4.0 Ma; 110-145 mcd), and a recent phase in the Quaternary (<1.7 Ma; <75 mcd) coincident with the opening of the Taupo Volcanic Zone (Shane et al., 1996). Earlier volcanism is marked by tephras at 385 and 485 mcd (~11.3 and 12.0 Ma, respectively).