13C values known for the entire Cenozoic (cf. Fig. F19).
Prior to Leg 181, only one APC-cored drill site, DSDP Leg 90 Site 594, was located in the ENZOSS region and its record only extends back to the early Miocene. In composite, Leg 181 drilling retrieved an almost complete stratigraphic succession of largely deep-marine sediment back to the late Eocene (37 Ma), together with two high-quality cores (Sites 1121 and 1124) of Late Cretaceous to Paleocene age (67-56 Ma) (Fig. F6) (McCave et al., in press; Carter, L., et al., in press b). Although it was unsurprising that the drilling intersected sediments of a wide range of ages, drilling through the Cretaceous/Tertiary (K/T) boundary at Site 1124 and almost reaching it at Site 1121 were perhaps particularly unexpected. Overall, sedimentation rates range widely from a low of 0.001 cm/k.y. during the late Neogene for Site 1121, situated under the energetic ACC-DWBC current system, to highs >50 cm/k.y. on the turbidity current-supplied abyssal Bounty Fan (Site 1122) and >120 cm/k.y. on the Canterbury upper slope (Site 1119). Significant paraconformities with attendant hiatuses lasting up to many millions of years occur at most sites and indicate phases of erosive bottom flow both in the pre-modern (Eocene) Pacific and, after 33.5 Ma, under the ACC-DWBC system.
Including the results of earlier drilling, which sampled sediments of early-middle Eocene and early Oligocene age (Leg 29, Sites 277 and 278), after Leg 181 drilling an almost complete Late Cretaceous to Holocene deep-marine stratigraphic record is now available for the New Zealand region for the first time. Detailed study of this data set has already led to significant advances in regional southwest Pacific stratigraphy, micropaleontology, and paleoceanography. We summarize below research completed using Leg 181 samples up until mid-2002. As proved to be the case for the earlier Leg 90 cores, however, we anticipate that many further studies will be based on Leg 181 data sets over the next decade, and even beyond.
The primary source for stratigraphic information about sites drilled during Leg 181 is the Leg 181 Initial Results volume (Carter, R., McCave, Richter, Carter, L., et al., 1999). Since completion of that volume, however, more refined age models have been prepared for the upper 100 meters composite depth (mcd) of Site 1119 (Carter, R., et al., in press) and for Sites 1121 (Graham et al., in press), 1122 (Wilson et al., 2000a), and 1123 (Wilson et al., 2000b). Detailed paleontologic zonation studies have been completed on Paleocene radiolarians from Site 1121 (Hollis, 2002), Eocene-Oligocene nannofossils from Sites 1123 and 1124 (McGonigal and Di Stefano, this volume), late Miocene bolboforms and planktonic foraminifers from Site 1123 (Crundwell, in press), and Pliocene-Pleistocene phytoplankton (Wei et al., submitted [N2]; Fenner and Di Stefano, in press). These studies are summarized below.
The high terrigenous content of the upper slope muds at Site 1119 is not ideal for microfossil studies. Furthermore, the mineralogy also proved to be unfavorable for preserving a depositional remanent magnetism (Carter, R., McCave, Richter, Carter, L., et al., 1999). Thus the shipboard age model was tentative and predicted an age of ~0.7 Ma at 100 mcd. Utilizing accelerator mass spectrometric radiocarbon dates and the onboard multisensor track (MST) gamma radiation record, Carter, R., et al. (in press) erected a detailed age model for the upper part of the core based on matching climate cycles between Site 1119 and the Vostok ice core (Petit et al., 1999, but using the amended age model of Shackleton, 2000). The upper 86.19 mcd of Site 1119 was deposited in the last 0.252 m.y. during marine isotope Stages (MIS) 1-8. The underlying sediments to 100 mcd, beneath a ~25-k.y.-long intra-MIS 8 unconformity at 86.19 mcd, represent MIS 8.5-11 (0.278-0.370 Ma) (Fig. F7).
Southernmost Site 1121 is located on a body of sediment that runs along the base of the Campbell Plateau in water depths of ~4500 m. Rather than the anticipated Neogene drift succession, drilling revealed that the extensive sediment body visible on seismic records comprises a Paleocene siliceous biopelagic sediment apron overlain by a thin Fe-Mn nodule-bearing "skin drift" of foraminiferal sand and light to dark brown clay that encompasses the Neogene history of the DWBC. Shipboard paleontological and paleomagnetic data were interpreted to indicate slow sedimentation rates, ages of ~3.6 Ma at ~3.0 meters below seafloor (mbsf) and ~56 Ma at ~30 mbsf, and Paleocene (>55 Ma) radiolarians possibly in situ for all depths below 19.4 mbsf.
Postcruise measurements (Graham et al., in press) showed that 10Be/9Be values both for the sediments and for entrapped Fe-Mn nodules decrease systematically with depth, consistent with radioactive decay. However, the 10Be/9Be sediment and nodule values diverge from ~3 mbsf to the top of the core, allowing several alternative geochronological models. The preferred age model assumes that the measured 10Be/9Be ratios of the nodule rims reflect the initial 10Be/9Be ratio of contemporary seawater, and these ratios are then used to derive the true age of the sediment where the nodules occur. Graham et al. (in press) thus infer an age of ~17.5 Ma (early Miocene) at only ~7 mbsf in the Site 1121 core (Fig. F8). Calculated sedimentation rates range from 8 to 95 cm/m.y., with an overall rate to 7 mbsf of 39 cm/m.y. The lowest rates coincide with the occurrence of trapped nodules and are interpreted to reflect periods of increased bottom current flow that caused net sediment loss. The growth rates of individual nodules decrease toward the top of the core, which is suggestive of an overall increase in the vigor of the DWBC from ~10 Ma to the present. The lithologic Subunit IA/IB boundary at 15.2 mbsf, below which the sediment is uniform yellow clay with nodules and broken layers of chert, therefore probably marks the start of DWBC activity and represents the abyssal manifestation of the Marshall Paraconformity.
Hollis (2002; see further description below) described the detailed radiolarian biostratigraphy of the underlying in situ Paleocene sediments (~40-140 mbsf) at this site that represent the eroded residuum of the postrift pelagic apron of the New Zealand Plateau.
The Bounty Fan developed across the path of the DWBC at the mouth of the Bounty Trough, 800 km east of eastern South Island. Site 1122 (water depth = 4432 m) was cored ~627 m into a left bank fan levee succession that was deposited under the influence of both east-traveling (and north-overspilling) turbidity currents emanating from the Bounty Trough and the north-flowing DWBC. Core recovery was complete using the APC to 104 mbsf but averaged only 47% between 104 and 627 mbsf because of loss of turbidite sands during rotary drilling (Carter, R., McCave, Richter, Carter, L., et al., 1999). Wilson et al. (2000a) showed that above ~520 mbsf shipboard paleomagnetic results are mostly confirmed by measurements of stepwise-demagnetized discrete paleomagnetic samples. However, between ~520 and 610 mbsf, discrepancies between discrete sample and shipboard measurements required the revision of the shipboard magnetostratigraphy.
Integration of the paleomagnetic stratigraphy with 30 shipboard microfossil datums and indicators provides a reliable chronology and age model for the Site 1122 sequence down to ~520 mbsf (Wilson et al., 2000a). Significant unconformities occur at ~5 Ma (~485 mbsf), between middle Miocene and early Pliocene drift sediments, and at ~2.2 Ma (~440 mbsf), which corresponds to an increased sedimentation rate and the arrival of the first turbidite sediment from the Bounty Trough and Channel.
Deposition of the North Chatham Drift beneath the DWBC began at 20.5 Ma and has continued virtually uninterrupted up to the present day (Carter, R., McCave, Richter, Carter, L., et al., 1999). The drift is ~600 m thick at Site 1123 and overlies a 13-m.y. hiatus (33.5-21.0 Ma) identified as the regionally pervasive Marshall Paraconformity. Core recovery was nearly complete to ~390 mcd, but below that level recovery using the extended core barrel (XCB) was incomplete. Comparison between downhole and core magnetic susceptibility records demonstrates that sediment loss was distributed throughout the length of each core run (Wilson et al., 2000b). Linear stretching of the magnetic susceptibility records from short core recovery runs were matched to downhole logging data, thus allowing the reconstruction of a complete depositional record of the North Chatham Drift. Chronology from shipboard paleomagnetism (Carter, R., McCave, Richter, Carter, L., et al., 1999) was confirmed by measurement of 256 stepwise demagnetized discrete paleomagnetic samples. Integration of the paleomagnetic stratigraphy with 47 shipboard microfossil datums indicates that an almost complete sequence of magnetic polarity reversals (including one new reversal, C5ADn1r) is preserved (Figs. F9, F10).
Site 1123 is located adjacent to Valerie Passage and distant from terrigenous sediment sources. The sediment contains 40-80 wt% carbonate throughout the core. The cyclic variations in sedimentation rates observed at the site result from a combination of the waxing and waning of DWBC flow and variation in biopelagic productivity (Hall et al., 2002). The most dramatic variation occurred in the middle Miocene, when sedimentation rates slowed between 15 and 13.5 Ma and then increased fourfold after 13.5 Ma. Sedimentation was fairly constant between 12 and 7.5 Ma, waned slightly at 7.5 Ma, and decreased by half between 7 and 5 Ma. Sedimentation increased again briefly in the early Pliocene (~3.2-2.8 Ma), after which sediment accumulation remained relatively constant throughout the rest of the Pliocene-Pleistocene. Comparison with benthic oxygen isotope and sortable silt records (see Hall et al., 2001) suggests that these variations in sedimentation rate represent fluctuations in bottom water production and DWBC flow rate, driven by variations in a cooling Antarctic climate and/or in ice volume, processes that were well under way by the early Miocene (Hall et al., 2003).
A 100-m-thick Paleocene sequence of pelagic siliceous and nannofossil-bearing chalky sediment was recovered from Site 1121, at the eastern foot of the Campbell Plateau (Carter, R., McCave, Richter, Carter, L., et al., 1999; Hancock and Dickens, this volume). Hollis (2002) showed that the lower 40 m of this succession contains a low-diversity radiolarian fauna of early to early late Paleocene age (Zones RP4 and RP5; ~63-59 Ma), similar to faunas already described from onland New Zealand and at DSDP Leg 21 Site 208 from the northern Lord Howe Rise (Fig. F11). The upper 60 m contains diverse middle-late Paleocene (Zone RP6) assemblages that differ from their counterparts elsewhere in their richness in plagiacanthids and cycladophorids, perhaps suggestive of cool-water conditions. The 150 Paleocene taxa so far recorded from Site 1121 are estimated to represent only about one-half of the complete radiolarian species diversity at the site.
An age model based on well-constrained nannofossil and radiolarian datums indicates that the rate of compacted sediment accumulation doubled from 1.5 to 3 cm/k.y. at the RP5/RP6 zonal boundary (~59 Ma). This results in large part from a sudden and pronounced increase in accumulation rates for all siliceous fossils; overall, radiolarians and larger diatoms increase from <100 to >10,000 specimens/cm2/k.y. This increase in biosiliceous productivity corresponds to the 59- to 57-Ma middle Paleocene cooling event, which is marked by the heaviest 13C values known for the entire Cenozoic (cf. Fig. F19).
Site 1121 radiolarian faunas provide an exciting new opportunity for linking Cretaceous-Cenozoic transition sediments in the South Pacific to the rest of the world. Cores recovered from the site contain the richest and most diverse late Paleocene radiolarian assemblages known from the Southern Hemisphere. Correlation with the North Atlantic in Zone RP6 is indicated by the presence of Aspis velutochlamydosaurus, Plectodiscus circularis, and Pterocodon poculum. Other species, including Buryella tetradica and Buryella pentadica, are valuable for local correlation but exhibit considerable diachroneity when their ranges are compared between the Pacific, Indian, and Atlantic Oceans.
Holes 1123C and 1124C penetrated middle Eocene-early Oligocene sediments that contain moderately to poorly preserved calcareous nannofossils. Younger DWBC sediments at these sites consist of alternating white clay-bearing nannofossil chalk and light greenish gray clayey nannofossil chalk (late Oligocene at Site 1124 and early Miocene at Site 1123), which overlie alternating white and light gray micritic limestone (late Eocene-early Oligocene) and red, yellow, pink, and brown mudstone (Paleocene and Eocene). McGonigal and Di Stefano (this volume) report that although the nannofossil assemblages show signs of dissolution and overgrowth, key marker species can be identified. The early Oligocene sediments are distinctly separated from the overlying Neogene succession by the Marshall Paraconformity, a regional marker of environmental and sea level change (Carter, R., 1985).
Age-depth models were constructed using nine nannofossil age datums and three magnetostratigraphic datums. There is good agreement between the biostratigraphy and magnetostratigraphy. The age model for Site 1123 indicates that the Marshall Paraconformity here spans ~21-33.5 Ma, suggesting that current speeds precluded deposition at the site and perhaps corroded and eroded older sediments for more than the first 10 m.y. of DWBC flow. At Site 1124, the succession is disrupted by three major paraconformities, of which the youngest equates with the Marshall Paraconformity and spans ~26.1-31.8 Ma. Two older hiatuses are a 3-m.y. gap that separates early Oligocene and middle Eocene sediment (~34-37 Ma) and a ~19-m.y. gap between middle Eocene mudstone and middle Paleocene nannofossil-bearing mudstone (~58-39 Ma).
Nannofossil information from Sites 1123 and 1124 indicates that the Eocene-Oligocene transition in eastern New Zealand was a time of fluctuating biota, in sympathy with the pulsed development of bottom currents. The widespread presence of the Marshall Paraconformity is consistent with other evidence for the development of DWBC and other flows along the New Zealand margin at that time. The hiatuses at Site 1124 that predate the Marshall Paraconformity may have developed under the influence of Paleogene south-flowing currents (see discussion in "Late Cretaceous-Late Eocene Rift-Drift: Start of the Kaikoura Synthem" in "ENZOSS Revisited: The History of Pacific Deep Flow").
Bolboforms are calcareous spinose marine microfossils inferred to have been phytoplankton. Their distinctive and changing morphologies and high resistance to dissolution make bolboforms important index fossils with which to supplement calcareous foraminifers and calcareous nannofossils. They occur particularly in middle to high latitudes in both hemispheres and have a stratigraphic record that extends back to the Paleogene (Cooke et al., 2002).
Crundwell (in press) established a high-resolution late Miocene biostratigraphy for the southwest Pacific using bolboforms and planktonic foraminifers from Site 1123, which also has an exceptionally complete magnetostratigraphic and astrochronologic record (Wilson et al., 2000b). The biochronologic model is based on dissolution-resistant species and morphotypes and is constrained by 33 bioevents that also occurred at DSDP Site 593 (Challenger Plateau, northwest of New Zealand). Well-constrained bolboform events include the total range zones of Bolboforma subfragoris s.l. (11.56-10.50 Ma), Bolboforma gruetzmacheri (10.46-10.31 Ma), Bolboforma capsula (10.20-10.13 Ma), Bolboforma pentaspinosa (10.15-10.08 Ma), Bolboforma gracilireticulata s.l. (9.75-9.61 Ma), lower Bolboforma metzmacheri s.s. Subzone (9.54-9.34 Ma), upper Bolboforma metzmacheri s.s. Subzone (9.01-8.78 Ma), Bolboforma metzmacheri ornata (8.45-8.25 Ma), Bolboforma praeintermedia (8.25-8.21 Ma), and distinctive abundance spikes associated with the upper B. subfragoris s.l. Subzone (10.61 Ma) and the lower B. metzmacheri Subzone (9.44 Ma) (Fig. F12). Planktonic foraminiferal events include well-defined intervals of dextral coiling excursions in Globoconella miotumida (10.93-10.82 and 9.62-9.42 Ma), the regional disappearance of Globoquadrina dehiscens (8.88 Ma), an acme zone of Truncorotalia juanai (7.23-6.23 Ma), and the evolutionary appearances of Globoconella conomiozea (~6.87 Ma), Globoconella mons (~5.72 Ma), Globoconella sphericomiozea (~5.53 Ma), Globoconella pliozea (~5.39 Ma), and Globoconella puncticulata (~5.28 Ma).
The close fit of the biostratigraphic data to a linear line of correlation between DSDP Site 593 and ODP Site 1123 and the consistency of the stratigraphic ordering of events suggest that most of the biostratigraphic events recognized are near synchronous. This correlation also shows that late Miocene sedimentation rates were in phase between the two sites, despite their 1400-km separation and accumulation beneath different water masses. Although the age model is based on only two oceanic sites, it establishes for the first time a high-quality late Miocene biochronostratigraphic framework for the temperate southwest Pacific Ocean and Tasman Sea regions.
DSDP Site 594 and ODP Site 1125 both have an established stratigraphy which, together with their respective location on the cooler and warmer sides of the STF, makes them particularly useful for calibration of regional nannofossil stratigraphies in the southwest Pacific. Based on correlation between these and other mid-latitude sites (DSDP Leg 90, Sites 593, 590, and 588), Wei and Chen (submitted [N3]) assessed the Southern Hemisphere synchroneity of eight commonly used nannofossil datum levels. These bioevents and their new age assignments are as follows:
In addition, the timing of two other stratigraphically useful nanno-events remains unchanged from previous determinations, namely
Because of the rarity of some age-diagnostic species, this biozonation is broader than the conventional low-latitude nannofossil biozonation. Nevertheless, the new southwest Pacific biozones are an important aid for correlating sites within and across the transitional water masses and temperatures of the STF.
Fenner and Di Stefano (in press) used population censuses of late Quaternary calcareous nannofossil and diatom assemblages to develop a quantitative calcareous nannofossil stratigraphy for the area near the STF. Five core sites near Chatham Rise and on both sides of the STF were selected for study (NIWA core Q 858, DSDP Site 594, and ODP Sites 1120, 1121, and 1123). Stratigraphic marker species used to establish the general age of the studied cores include Emiliania huxleyi (FAD = 0.28-0.268 Ma), Fragilariopsis reinholdii (LAD = 0.65 Ma), Fragilariopsis fossilis (LAD = 0.70 Ma), and Fragilariopsis doliolus (FAD = 1.8 Ma). The top sections of all cores fall within the stratigraphic range of E. huxleyi and are therefore younger than 0.268 Ma (early MIS 7). Detailed age constraints were developed using oxygen isotope stratigraphy, tephrochronology, and 14C age determinations for two sites north of the Chatham Rise (NIWA piston core Q858 [Fenner et al., 1992] and Site 1123 [Hall et al., 2001]) and one site south of the Chatham Rise (Site 594 [Nelson et al., 1993]). Correlation shows that mid-late Holocene sediment is missing at both Sites 1123 and 594, perhaps due to core-top drilling disturbance.
Based on this stratigraphy, abundance changes were determined for phytoplankton species that are present both north and south of the Chatham Rise. Five nannofossil abundance shifts were identified that apparently occur synchronously both north and south of the Chatham Rise:
In contrast to the nannofossils, the diatom assemblages largely comprise species that are restricted to one side of the Chatham Rise. No stratigraphically useful diatom abundance changes were found among taxa that are common both north and south of the rise. The only consistent event is a peak of Thalassionema nitzschioides in MIS 3 that coincides with the abundance peak of the nannofossil Helicosphaera carteri and may correlate with the MIS 3.2 isotope event (Martinson et al., 1987). Despite this lack of quantitative stratigraphic indicators, T. nitzschioides and Rhaphoneis surirelloides are more abundant in sediments deposited during cooler times (MIS 2-4) and Paralia sulcata is more abundant in sediments deposited during warm intervals.
The newly defined calcareous nannofossil datums are based on prominent species' abundance shifts within the past 130 k.y. and can be applied throughout the subtropical to subantarctic region east of New Zealand. Where a stable oxygen isotope record is not available, the nannofossil datums can be used to correlate between cores and to calculate sedimentation rates.
Leg 181 drilling complemented earlier DSDP work in the New Zealand area, especially the exceptional cores that were retrieved from Leg 29 Site 277 and Leg 90 Sites 593 and 594, to provide for the first time a combined deep-ocean record that spans the entire Cretaceous-Cenozoic tectonosedimentary cycle (Kaikoura Synthem). Substantial advances are thereby being made in our knowledge of Southern Hemisphere mid-latitude biostratigraphy (see previous section) and in the wider application of Leg 181 micropaleontologic analysis. Other postcruise studies, summarized below, apply paleontology to evolutionary and paleoceanographic problems.
Deep-sea foraminifer species have an estimated low background turnover rate of 2%/m.y. (McKinney, 1987). The most severe extinctions of deep-sea foraminifers, with a loss of 30%-50% of species, occurred during the Paleocene/Eocene Thermal Maximum (PETM) (~55 Ma) (van Morkhoven et al., 1986; MacLeod et al., 2000) and is inferred to have resulted from oxygen-poor, warm bottom waters coupled with changes in surface productivity (Katz et al., 1999). Extended periods of slightly enhanced extinction rate have also been identified globally in the late Eocene-early Oligocene (36-30 Ma), the middle Miocene (16-12 Ma) (Weinholz and Lutze, 1989), and the middle Pleistocene (dubbed the "Stilostomella extinction" after the family Stilostomellidae, which died out at this time; Schonfeld, 1996).
The Stilostomella event was first identified in the Atlantic Ocean, where 10 benthic foraminiferal species from six genera are reported to have disappeared between 1.0 and 0.6 Ma (Lutze, 1979; Thomas, 1987); extinctions occurred 0.1-0.2 m.y. earlier at depths >3000 m and in southern latitudes (Weinholz and Lutze, 1989; Schonfeld, 1996). Middle Pleistocene extinctions of some of these taxa have also been recorded from several sites in the Indian Ocean (Gupta, 1993) and the Pacific Ocean (Schonfeld and Spiegler, 1995). Hayward (2001, 2002) used new samples available from Leg 181 sites to reassess the nature of the Stilostomella event in the southwest Pacific. This event is associated with the disappearance of at least 2 families, 15 genera, and 48 species of dominantly uniserial, elongate foraminifers with distinctive apertural modifications (i.e., ~15%-25% of the fauna). These taxa progressively dwindled in number and became extinct during glacial periods throughout the late Pliocene to middle Pleistocene (~2.5-0.6 Ma), with most extinctions occurring between 1.0 and 0.6 Ma, at the time of the middle Pleistocene transition (MPT) (Fig. F13). Hayward's are the first high-resolution studies of the Stilostomella extinction event, and they indicate that the event was far more significant for deep-sea diversity loss than the 10-species extinction previously reported. The middle Pleistocene extinction of elongate benthic foraminifers was the most dramatic last phase of a worldwide decline in their abundance that began during cooling near the Eocene/Oligocene boundary and continued until at least the middle Pleistocene.
Hayward et al. (2001) used cluster analysis and correspondence analysis to study the distribution of benthic foraminiferal faunas from grab samples, piston core tops, and DSDP and ODP core tops at sites located east of New Zealand over a water depth range of 90-4700 m. A total of 465 benthic taxa were identified, of which 139 are new records for the New Zealand region. The relative abundance of common species, species associations (six of which are delimited by the cluster analysis), upper depth limits of key benthic species, and the relative abundance of planktonic foraminifers are the most useful proxies for estimating paleobathymetry. In a further development, Hayward et al. (2002) reported on the environmental factors that principally determine benthic foraminiferal associations in the southwest Pacific area. Important factors—all of which are related to water depth—include dissolved oxygen content, seasonality of food supply, organic carbon flux, advection of water masses, bottom water carbonate corrosiveness, benthic energy state at the boundary layer, and grain size distribution of the seabed.
Hayward et al. (in press) performed benthic foraminiferal censuses (336 species) and fragmentation estimates on 85 Neogene samples from Site 594 and Sites 1120-1125 (Fig. F14). Sample associations were determined using cluster analysis and canonical correspondence analysis, which showed that the foraminiferal groupings are most strongly influenced by bathymetric depth, reflecting water mass stratification, and age, reflecting biotic evolution. Three intervals of foraminiferal taxonomic turnover occur at 16-15, 11.5-10, and 2-0.5 Ma and are inferred to correspond to intervals of enhanced cooling and increased surface water productivity; the late Pliocene-middle Pleistocene Stilostomella extinction culminated at the MPT. Significant differences between lower bathyal faunas north and south of the Chatham Rise suggest the presence of an oceanic front (predecessor of the STF) along the rise since at least the early Miocene. Modern AAIW benthic associations were established north of the Chatham Rise at 10-9 Ma and south of it at 3-1.5 Ma. Middle-upper bathyal faunas on the Campbell Plateau are dominated by reticulate bolivinids during the early and middle Miocene, indicative of sustained productivity above relatively sluggish, poorly oxygenated bottom waters. Faunal changes and sediment hiatuses indicate increased current vigor over the Campbell Plateau since the latest Miocene.
Downcore studies for planktonic foraminiferal fragmentation indices show significant amounts of abyssal carbonate dissolution throughout most of the Neogene, peaking at upper abyssal depths in the late Miocene (11-7 Ma), with the lysocline progressively deepening thereafter. Peak abundances of Epistominella umbonifera indicate an increased input of cold SCW to the DWBC at 7-6 Ma. Faunal association changes imply establishment of the modern oxygen minimum zone (upper CDW) in the latest Miocene. Faunal assemblage changes are consistent with stepped increases in productivity at 16-15, 3-1.5, and 1-0.5 Ma, whereas benthic taxonomic turnover was concentrated at 16-15, 11.5-10, and 2-0.5 Ma. These microfaunal changes are interpreted in terms of the pulsed, sequential development of southern, and later northern, polar glaciation, with consequent cooling of bottom waters, increased vertical and lateral stratification of ocean water masses, and increased surface water productivity.
Sabaa et al. (in press) conducted planktonic foraminiferal census counts to estimate middle-late Pliocene (4.0-2.37 Ma) sea-surface temperatures (SSTs) in cool subtropical waters at Site 1125, presently just north of the STF. Using the modern analog technique (MAT), SSTs at this location are estimated to have often been cooler than both modern and last glacial SSTs, although there were brief periods when temperatures were 1°-2°C warmer than today. Specifically, summer temperature maxima were 1°-2°C warmer than present day and winter temperature minima were 6°-10°C cooler, but the warm maxima were brief and spasmodic. Major cooling excursions of 6°-10°C occurred at 3.35, 3.0, and 2.8 Ma, and minor coolings of shorter duration and lower magnitude (~4°C cooling) occurred at 2.7 and 2.4 Ma. These results demonstrate episodes of strong cooling in the middle-late Pliocene of the subtropical southwest Pacific. The coolings might be a regional effect or, alternatively, could stem from increased upwellings of cold intermediate water or from northward migrations of the STF. Sabaa et al. (in press) prefer the regional cooling interpretation because of the evidence that exists for mid-Pliocene cooling in other regions of the western Pacific (e.g., Andersson, 1997; Heusser and Morley, 1996), despite the inferred occurrence of an overall global warming at this time (Dowsett et al., 1996).
Site 1123 possesses a robust chronology, the upper part of which is based on the astronomical tuning of the benthic foraminiferal 18O record (Hall et al., 2001, 2002). The ~180-m-thick Pliocene-Pleistocene sequence at the site represents the best continuous record of climate and vegetation change covering the last 5.25 m.y. in the New Zealand region. Mildenhall (2003) and Mildenhall et al. (in press) used censuses of the terrestrial pollen and spore record at the site in order to determine (1) how variations in terrestrial palynomorph assemblages are related to global climate cycles and regional climatic/oceanographic processes and (2) the degree to which New Zealand's terrestrial vegetation was influenced by major climatic changes associated with the MPT at ~0.92-0.62 Ma. Despite the long distance from shore and relatively great water depth (3290 m), palynomorph abundance is more than adequate at Site 1123 to demonstrate a strong correlation between the orbitally tuned marine record and climatically controlled changes in terrestrial vegetation (Fig. F15).
Pollen and spores from a range of terrestrial environments are present in all samples, but overall the recovered palynomorph assemblages are derived from regional podocarp/hardwood forest vegetation from the southern part of North Island, New Zealand. Angiosperm pollen, although generally sparse, are dominated by Nothofagus, whereas robust buoyant spores such as Cyathea and bisaccate pollen such as Podocarpus/Prumnopitys are overrepresented. This overrepresentation probably results from the ability of these taxa to float great distances. Gradual changes occur in the dominant pollen types, with more mesothermal taxa (e.g., Brassospora) in the Pliocene and early Pleistocene and fewer mesothermal taxa (e.g., Fuscospora and podocarp conifers) in the middle and late Pleistocene. This suggests a gradual change from warm, humid, and perhaps cloudy conditions to a cooler, drier climate as the global climate deteriorated. Superimposed on this long-term change are more intense glacial-interglacial cycles in which glacial periods are enriched in Halocarpus, Phyllocladus, Nothofagus fusca type, and Coprosma pollen relative to the interglacial flora with Cyathea, tall tree Podocarpus/Prumnopitys, and Dacrydium cupressinum. Time series analysis indicates that the vegetation record is covariant with marine climate proxies (as represented by carbonate content) and that it is strongly coherent at the 41- and 100-k.y. Milankovitch frequencies. This is the first time that terrestrial pollen changes have been shown to occur almost exactly in phase with a marine climatic signal. A pronounced increase in amplitude and a coeval lengthening of climate cycles from 41 to 100 k.y. occurs between 0.92 and 0.62 Ma and provides a rare vegetation record during the fundamental MPT reorganization of Earth's climate system.
Scott and Hall (in press) used MAT, ordinations, and minimum spanning trees to compare 32 foraminiferal assemblages from Site 1123 with 35 core-top assemblages collected between 35° and 61°S east of New Zealand (Fig. F16). Many Site 1123 faunas in the 2.7-m.y. interval sampled are transitional between colder- and warmer-water assemblages, similar to core-top faunas along the crest of the Chatham Rise. This result is consistent with earlier studies from this region and suggests that the STF was positioned over the Chatham Rise through glacial and interglacial periods at least back to the late Pliocene. Other assemblages are rare, but one at 1.165 Ma closely resembles core-top assemblages between 44° and 48°S and may identify cooler surface water.
The dominance of Globoconella inflata is a principal feature of Site 1123 assemblages, but across the MPT this species is generally present in subordinate numbers to dextral specimens of Neogloboquadrina pachyderma. There are no close core-top analogs for such faunas, but other data show that they develop in high biomass water associated with upwelling or mixing. The proportion of sinistrally coiled N. pachyderma rises to 0.6 between 2.45 and 2.57 Ma, soon after the intensification of Northern Hemisphere glaciation. Although the coiling data indicate subantarctic near-surface water, the species remains rare and the assemblage overall is still of transitional character. Therefore, probably only minor entrainment of subantarctic water occurred.
A new high-resolution census study of planktonic foraminifers across the MPT has begun (M.P. Crundwell and G.H. Scott, unpubl. data). This MIS 27-12 (0.44-1.00 Ma) interval is relatively free from carbonate dissolution, and the coiling signature of N. pachyderma in the census probably identifies further periods in which subantarctic near-surface water was introduced into the subtropical gyre. Such colder-water assemblages also contain increased proportions of sinistral N. pachyderma, to the extent that they are sometimes present in subequal numbers to dextral specimens.
The Quaternary Taupo Volcanic Zone (TVZ) is Earth's most productive modern rhyolitic center (Wilson et al., 1995a, 1995b), but many details of its history—and those of its western predecessor, the Miocene-Pliocene Coromandel Volcanic Zone (CVZ)—remain unknown. Previous research was based mainly on terrestrial records, which are affected by tectonism, erosion, and burial beneath successive volcanic deposits (Shane, 2000; Wilson, 1994). Leg 181 provided the most complete record of major volcanic events available to date in the form of tephra layers preserved at Sites 1122-1125, which in prevailing conditions lie downwind of the CVZ and TVZ. Thus, Leg 181 drilling provided a new standard history of the evolution of these major plate boundary volcanic zones (Carter, L., et al., 2003, in press a). This information, together with that from DSDP Leg 29 Site 284 (Kennett et al., 1979), Leg 90 Sites 590-594 (Nelson et al., 1986b), and existing piston core data (e.g., Ninkovich, 1968; Lewis and Kohn, 1973; Froggatt et al., 1986; Watkins and Huang, 1977; Carter, L., et al., 1995), has given us a firm understanding of ash dispersal for the offshore New Zealand region.
Carter, L., et al. (2003) used glass shard geochemistry and isothermal plateau fission track dating, supported by robust chronologies developed from paleomagnetic stratigraphy, orbitally tuned reflectance profiles, and benthic isotope records, to demonstrate that near-continuous rhyolitic volcanism occurred within the CVZ and TVZ since ~12 Ma (Fig. F17). This history is particularly well preserved at Site 1124, which, although 600 km from the inferred volcanic sources, contains 134 macroscopic tephra layers with individual tephra up to 92 cm thick.
Volcanic activity related to the development of the Australian/Pacific boundary began in western North Island in the Oligocene and early Miocene (e.g., Hayward, 1993). Yet, despite continuous core records that stretch back to the early Miocene (Site 1123) and late Oligocene (Site 1124), the Leg 181 macroscopic tephra record only begins at ~12 Ma. Nearby onland, tephra also appear at ~12-13 Ma in a well-exposed Miocene terrigenous succession at Mahia Peninsula (Schneider et al., 2001). The absence of discrete tephra in sediments east of New Zealand prior to this time is attributed to the following factors:
At Site 1124, the earliest rhyolitic tephra (~12 Ma) predates the earliest known terrestrial ignimbrite in the CVZ by more than a million years. Thereafter, through the Miocene and Pliocene, the data show that episodes of volcanism were punctuated by periods of subdued activity or quiescence. The longest break in activity is 0.7 m.y., with most breaks being <0.5 m.y. The first phase of pronounced tephra deposition occurred at ~7.7-7.0 Ma, which coincides with a major phase of caldera development in the CVZ (Adams et al., 1994). Quiescence between 7.0 and 6.3 Ma was followed by short bursts of activity at 6.30-6.24 Ma and ~5.2 Ma, about the respective ages of the Pumpkin Rock Ignimbrite/Wheuakite Rhyolite and Ahu Ahu Rhyolite eruptions onshore. The ensuing Pliocene ash record indicates 2 m.y. of near-continuous but more intense activity than in the late Miocene. This contrasts with the terrestrial record, which places the end of CVZ volcanism at ~4 Ma and the start of TVZ volcanism at ~2 Ma (Adams et al., 1994).
From ~1.6 Ma onward, major rhyolitic activity was centered in the TVZ (Wilson et al., 1995b). Again, new Leg 181 data show that large rhyolitic eruptions were more frequent than previously known from terrestrial studies. Wilson et al. (1995b) noted that large TVZ eruptions were associated with major caldera-forming periods, although such events were probably interspersed with smaller eruptions. The Site 1124 record demonstrates, however, that tephra of intervening age and of similar thickness to those deposited during the known periods of caldera collapse are present, suggesting that some hitherto unknown calderas in the TVZ (and CVZ) are buried by younger eruptives or occur integrated within larger and partly younger caldera complexes (Wilson et al., 1995b). Another possibility is that other eruptive sources may lie within the offshore extension to the TVZ, or beyond. Given the close similarity between the major oxide glass geochemistry of ashes from the CVZ and TVZ and the near-continuum of eruptions implied by the Leg 181 tephra record, it is difficult to attribute particular ashes or groups of ashes to a definitive source. The two closely related volcanic zones may have succeeded one another in time or, alternatively, there may have been an overlapping interval when both zones were active together.
That tephra layers are more frequent and thicker off eastern, compared to western, New Zealand confirms the dominance of prevailing westerly winds on ash dispersal (cf. Frontispiece, part A, inset). For example, Site 1124 contains 134 tephra layers averaging 9.6 cm thick, whereas DSDP Leg 90 Site 593, the closest western drill site to the volcanic sources, bears only four macroscopic tephra layers that average <3.5 cm thick (Nelson et al., 1986a).
Within such a pattern of preferential eastward dispersal, some variability might be expected, resulting from fluctuations in paleowind patterns. Thus the increased thickness and frequency of tephra during the Pliocene may not only reflect increased volcanic activity but also a general strengthening of the westerly wind regime at ~3-4 Ma (Kennett and von der Borch, 1986; Zhou and Kyte, 1992). Other variability may be associated with El Niño-Southern Oscillation (ENSO) events, with some models suggesting that periods of strong ENSO activity lasted for at least 0.5 m.y. (Clement et al., 2001). Vigorous El Niño conditions favor the eastward dispersal of ash, whereas La Niña weather encourages westward transport. Interestingly, the Rangitawa eruption of 0.345 ± 0.012 Ma (Pillans, 1996), which has a wide distribution, coincides with a shutdown of ENSO, as proposed by Clement et al. (2001). The weakened westerlies (or La Niña easterlies) of the shutdown period may explain the (unusual) presence of the Rangitawa (Mt. Curl) tephra on both sides of New Zealand (Froggatt et al., 1986; Shane, 2000). Superimposed on these large but short-scale climatic variations is the regular rhythm of the Milankovitch glacial-interglacial cycles. At the Leg 181 sites, most macroscopic tephra accumulated during glacial periods, suggestive of increased windiness at those times (e.g., Stewart and Neall, 1984; Hesse, 1994). More frequent eruptions during glaciations may also have been caused by the hydrostatic unloading of magma chambers during falls in sea level, leading to decompression melting (Paterne et al., 1990).
Eruption size can play a prominent role in ash dispersal (Nelson et al., 1985). Major events may be accompanied by ash columns that extend through the low-level winds of the tropopause, currently 20-22 km high at the latitude of the TVZ (Trenberth, 1992), and penetrate the stratosphere, especially in the austral spring-summer when stratospheric winds have a strong westward component while lower level winds continue to the east. Examples of inferred stratospheric penetration include the ~41- to 45-km-high eruption column estimated for the 26.17-ka Oruanui eruption (Wilson, 1994) and the ~35- and 50-km-high columns estimated for the 3.47-ka Waimihia and 1.718-ka Taupo eruptions (Pyle, 1989). Nonetheless, eruption columns apparently only rarely penetrated into the stratospheric zone of easterly winds, as 97% of all marine tephra are observed in the east.
The latitudinal limits of tephra distribution from the CVZ/TVZ extend south from 30°S (Ninkovich, 1968; Lewis and Kohn, 1973; Pillans and Wright, 1992; Watkins and Huang, 1977) to somewhere between 46°S (Site 1122) and 50°S (Site 1120) (Carter, R., McCave, Richter, Carter, L., et al., 1999). Although Site 1120 displays no obvious tephra, the presence there of several late Cenozoic hiatuses means that the absence of tephra may be more apparent than real. Nonetheless, the isopachs for several recent large eruptions presented by Carter, L., et al. (1995) show that Site 1120 must be located close to the edge of the tephra distribution area. Such minor ambiguities notwithstanding, it is certain that below 60°S southwest Pacific ash-bearing sediments have an Antarctic rather than a New Zealand provenance (Shane and Froggatt, 1992).
Since completion of the Leg 181 Initial Reports volume (Carter, R., McCave, Richter, Carter, L., et al., 1999), studies have been completed on Site 1123 clay mineralogy (Winkler and Dullo, this volume), Site 1123 major element analyses (Weedon and Hall, this volume, in press), dissolved manganese concentrations (Dickens, this volume), Site 1121 carbonate percentages (Hancock and Dickens, this volume), microbial sulfate reduction (Böttcher et al., this volume, in press), biogenic opal concentrations (Suzuki et al., this volume), the use of color reflectance as a carbonate proxy (Millwood et al., this volume), stable isotope measurements (Hall et al., 2001; Harris, this volume; Carter, R., et al., in press; Yang et al., 2002; Wei et al., submitted [N2]; Mii et al., unpubl. data [N4]) and petrophysical properties of muds from Site 1125 (Kim et al., 2001). These studies are summarized below.
Site 1123, located on the northeastern flank of the Chatham Rise, contains an almost uninterrupted record of sedimentation under the DWBC from the early Miocene onward (~20.5-0 Ma). Systematic mineralogical analyses by Winkler and Dullo (this volume) confirm the very fine grained, carbonate-rich nature of the Site 1123 succession. Clay mineral assemblages are dominated by smectite and illite, with high smectite values in the Eocene decreasing upcore (Fig. F18, right). Accompanying a smectite decrease from 21 Ma, illite and chlorite concentrations progressively increase, with a significant step at 6.4 Ma. An abrupt tenfold increase in the percentage of >63-µm-diameter sediment grains that occurs at ~1.3 Ma is apparently not accompanied by changes in the clay assemblage.
The early Miocene start and gradual nature of the long-term changes in clay mineralogy beneath the DWBC at Site 1123 contrasts with the more recent (middle Pliocene; ~3.5 Ma) change to chlorite-illite-dominated sediments documented at Site 594, which lies under AAIW (Dersch and Stein, 1991) (Fig. F18, left). The Site 1123 change reflects a regionally uniform sedimentation process that was driven by long-term factors, whereas sediment deposited at Site 594 was subjected to a more sudden change to chlorite-illite-rich assemblages, which was probably caused by the arrival of the first turbidity current overspills from the Bounty Channel.
Overall, the mineralogic changes observed at Sites 1123 and 594 are consistent with regional evidence for progressive uplift along the Alpine plate boundary from the early Miocene (~25 Ma) onward, with increased rates of plate collision since the Pliocene (Walcott, 1978; Tippett and Kamp, 1993; Sutherland, 1996; Batt et al., 2000). These events caused progradation of the eastern South Island terrigenous sediment prism and the activation of channel systems that fed detritus onto the slope and, ultimately, into the DWBC (Carter, R., 1988b). Initially, the prism was derived mainly from the erosion of low-grade metagraywackes and chlorite schists, but from the Pliocene onward there has been an increasing contribution from higher-grade schists. Despite previous assertions, no unique date marks the intensification of South Island plate boundary uplift. Rather, the stratigraphic evidence clearly demonstrates a regional early Miocene (~25 Ma) start to the provision of "new" terrigenous sediment (e.g., Finlay, 1953; Lewis et al., 1980), an early middle Miocene (~17 Ma) acceleration of supply (e.g., Cutten, 1979; Turnbull et al., 1993), and, with the precise age depending upon location, a Pliocene (~3-5 Ma) start to the deposition of higher-grade metamorphic detritus, consequent upon marked mountain uplift (e.g., Sutherland, 1996; Batt et al., 2000).
Hancock and Dickens (this volume) analyzed 39 samples from the Paleocene biopelagic sediments at Site 1121. Carbonate concentrations vary widely between 3.7 and 51.4 wt% (average = 31 wt%). Site 1121 was at ~3800 m water depth or deeper during the deposition of these sediments. The occurrence of such relatively high carbonate percentages here indicates that the Paleocene carbonate compensation depth (CCD) in the southwest Pacific was significantly deeper than previous studies suggested (e.g., van Andel, 1975).
Weedon and Hall (this volume, in press) report on a high-resolution analytical study of mid-late Pleistocene drift sediments from Site 1123. In order to capture Milankovitch cyclicity, ~1000 samples were taken at a close spacing of 5-10 cm from four selected intervals: 0-1.2 Ma (Pleistocene), 13.9-15.4 Ma (middle Miocene), 20.0-20.6 Ma (early Miocene), and 32.8-33.1 Ma (early Oligocene). Samples were analyzed for elemental concentrations by inductively coupled plasma-atomic emission spectrophotometry. After analysis, results were normalized using aluminium concentrations to provide proxies for nutrient levels, siliciclastic and volcaniclastic sediment composition, and bottom water redox conditions.
The results of Ba/Al for the Pleistocene interval were discussed by Hall et al. (2001), and a geochemical interpretation of the additional elemental ratios and the more extensive age range of host sediment is given by Weedon and Hall (in press). One important result is that samples located close to macroscopic tephra layers, presumed therefore to contain bioturbated ash, are characterized by relatively high Si/Al and K/Al, low Ti/Al, and, in some cases, low calcium carbonate contents. Such tephra-bearing samples were removed from the Pleistocene data set prior to time-series analysis for determination of the pelagic sediment history. None of the pre-Pleistocene samples were detected to be contaminated by tephra.
In the late Pleistocene, productivity variations at the 41-k.y. orbital obliquity frequency are apparent in the Ba/Al and P/Al time series, which are coherent but not in phase (Ba/Al leading P/Al). In the middle Miocene, a trend of increasing carbonate, Ba/Al, P/Al, and Si/Al may reflect gradually increasing surface water nutrient supply (i.e., higher productivity). Carbonate and P/Al are here highly coherent but do not reflect orbital cyclicity. However, Ba/Al does show 41-k.y. cyclicity, so orbital-scale changes may have affected productivity too. No evidence for regular cyclicity exists in any of the variables for the older early Miocene and early Oligocene intervals sampled. Based on comparative values of carbonate, Si/Al, Ba/Al, and P/Al and the presence of biogenic silica, the early Oligocene was more highly productive than the younger sampled intervals or today.
The simplest explanation of the long-term trend is that surface water nutrient levels systematically increased during the middle Miocene, perhaps because of the growth of ice in Antarctica. However, it is an important point that the highest inferred nutrient levels and productivity at Site 1123 occurred prior to DWBC activity during the early Oligocene.
Shipboard pore water and gas analyses show that sediments from the seven Leg 181 sites span an exceptional range of chemical environments. Pore water alkalinity, NH4+, SO42-, and PO43- concentrations as well as headspace CH4 concentrations indicate significant differences in sediment redox conditions across the region (Carter, R., McCave, Richter, Carter, L., et al., 1999). The distribution of solid and dissolved manganese plays an important role in geochemical interpretations of such sedimentary environments, and Dickens (this volume) therefore measured the dissolved Mn2+ concentrations of pore waters at Sites 1119, 1122, 1123, and 1125. He reports Mn2+ concentrations ranging between 0.1 and 26.5 µM and averaging 1.8, 3.3, 3.8, and 1.0 µM at the four sites, respectively. Mn2+ concentrations are relatively high at the deepwater sites (1122 and 1123) and relatively low at the shallow sites (1119 and 1125), perhaps reflecting higher inputs of reducible solid Mn phases at the deeper locations.
Böttcher et al. (this volume, in press) analyzed 79 interstitial water samples from Sites 1119-1125 for stable isotopes of dissolved sulfate (34S) and for major and minor ions. Sulfate from the interstitial fluids had 34S values between +20.7
and +60.0 (i.e., were enriched in 34S with respect to modern seawater; 34S = ~+20.6). Microbial sulfate reduction is therefore inferred to have occurred at all sites with an intensity that depended upon the availability of organic matter. This availability is controlled by sedimentation rate, which is itself related to factors such as productivity and the presence or absence of turbidity or bottom currents. Böttcher et al. also showed that the amount of reduced inorganic sulfur (essentially pyrite, which is the product of microbial sulfate reduction) varied commensurately between 0.05 and 0.63 wt%.
Suzuki et al. (this volume) measured biogenic opal concentrations from bulk sediment samples using wet alkaline extraction of freeze-dried bulk sediment. Samples were taken throughout Sites 1123 and 1125 and from the uppermost 100 mcd of Site 1124. The first and last sites lie within the influence of the DWBC, whereas Site 1125 lies within AAIW. Site 1124 showed opal contents of 2-8 wt%, which is relatively high compared to the other two sites, where values mostly range 1-4 wt%. A subbottom maximum in biogenic opal content occurring between 1.0 and 1.5 mcd at all three sites represents the position of the Last Glacial Maximum (LGM) and is suggestive of enhanced silica productivity during glacials (cf. Fenner et al., 1992; Nelson et al., 1993). The sampling resolution used by Suzuki et al. did not allow capture of the full detail of the climatic signature below the LGM.
It is well established that color reflectance measurements can provide an estimate of the component mineralogy of sediment cores, especially with respect to calcium carbonate content (Mix et al., 1995; Balsam et al., 1999). Leg 181 was the first ODP voyage during which a new automated Minolta spectrophotometer system was used for routine core logging. Because of several teething problems, the data collected at Site 1119 were of poor quality. Accordingly, Millwood et al. (this volume) collected postcruise reflectance data from the Site 1119 core. Empirical regression relationships between reflectance (400- and 550-nm wavelength) and carbonate content were then formulated separately for each Leg 181 site using the available shipboard and laboratory carbonate determinations from Carter, R., McCave, Richter, Carter, L., et al. (1999). These regressions were then applied to calculate a model carbonate percentage curve for each site based on the reflectance data.
Because of the primary role that the marine oxygen isotope stages play in stratigraphic correlation and the richness of their information content, stable isotope analysis has become "master-proxy" for paleoceanographic studies. Thus, a prime intention of Leg 181 drilling was to recover multicored, continuous sequences that were suitable for high-resolution foraminiferal stable isotope studies. Sites 1119, 1121, and 1123 have so far yielded such results (Figs. F19, F21).
Site 1119 (water depth = 396 m) penetrated 495 m of micaceous, terrigenous mud with interbeds of bottom current-emplaced sand, generally <1 m thick. Carter, R., et al. (in press) provided a high-resolution (~2 k.y./sample) planktonic stable isotope record through the ~47-m-thick MIS 1-5 interval at this site, based on Globigerina bulloides. A lower-resolution (~5 k.y./sample) data set for MIS 6-10 extends the isotope record to 100 mcd, but the record is interrupted by a ~25-k.y. unconformity that cuts out part of MIS 8. Although the Site 1119 MIS 1-9 oxygen isotope record broadly matches the established SPECMAP pattern, site-specific oscillations correspond to movements of the STF across the site during periods of rapid climatic and sea level change. When combined with the color reflectance and natural gamma ray logs, the paleoceanographic proxies from the upper part of Site 1119 define a climatic record that accords closely to Antarctic continental air temperature, as represented by the Vostok deuterium record (Petit et al., 1999). Thereby, the Site 1119 record demonstrates a close correlation between changes in Southern Hemisphere mid-latitude oceanographic proxies and south polar air temperatures, consistent with strong intra-hemispheric atmospheric coupling.
Sites 1121 and 1124 penetrated a pre-DWBC succession corresponding to the postrifting Cretaceous-Oligocene sediment apron. Mii et al. (unpubl. data [N4]) and Wei et al. (submitted [N2]) made stable isotope measurements on bulk nannofossil-rich sediment samples, 64 from Site 1121 (early-late Paleocene) and 58 from Site 1124 (Late Cretaceous-Oligocene) (Fig. F19). Despite the age and (for Site 1124) burial depth of the sediments sampled, the isotope values do not vary systematically with carbonate content or with burial depth, nor do the carbon and isotope values show any systematic linear relationship with each other. Any diagenetic or dissolution effect is therefore minimal, except possibly for the deepest sample from Site 1121 (133.52 mbsf), which displays an anomalous 2 negative excursion in 18O accompanied by a smaller negative change in 13C. This sample apart, the Site 1121 samples span ~62.2-56.2 Ma and display a slight overall positive trend in both 13C (~2.7-3.3) and 18O (~0.0-0.5). Fluctuations in magnitude as large as 1 occur on either side of this trend, with a particularly marked 13C depletion spike of 2.26 at 58.51 mbsf, just above the Chron C26n/C25r boundary. At Site 1124, 13C values range 1.7-3.4 and 18O values range -1.3-0.9. Prominent in the 13C record is a 1.1 negative shift across the K/T boundary, followed immediately by a 1.0 increase in the early Paleocene. Values of 13C then decrease by 1.3 at the Paleocene/Eocene boundary, increase by 0.5 in the late Eocene, and finally decrease by 0.8 at the Eocene/Oligocene boundary. Values of 18O decrease by 1.0 in the latest Cretaceous and remain around -1.0 into the early Paleocene before increasing again by 1.4 later in the early Paleocene. Values of 18O then fluctuate around -0.3 until they increase to 0.7 near the Paleocene/Eocene boundary. A 1.4 decrease occurs in the early Eocene, after which 18O values increase to 0.9 at the Eocene/Oligocene boundary, before settling at ~0.8 in the Oligocene.
Although the Site 1121 record covers a relatively short time span and the Site 1124 record is punctuated by three major paraconformities, these results show throughout a pattern that is closely comparable to previously published oceanic foraminiferal isotopic records (e.g., Zachos et al., 2001). Both records show negative carbon and isotope shifts across the K/T boundary followed by a gentle cooling trend through the Paleocene, and Site 1121 displays well the established Cenozoic carbon isotope maximum value at ~57 Ma. Site 1124 also displays notably enriched isotope values during the middle Eocene and—albeit within a lengthy paraconformity—contains a sharp cooling enrichment across the Eocene/Oligocene boundary. These results show that stable isotope ratios within Paleogene bulk sediments from the southwest Pacific mimic the global oceanic pattern. The waters of the southwest Pacific Ocean were therefore in free communication with those of the world ocean, and isotope measurements from the region provide a good means of chemostratigraphic correlation (Wei et al., submitted [N2]).
Site 1123 contains an almost continuous early Miocene-Holocene (0-20.5 Ma) record of sedimentation under the DWBC. A benthic stable isotope record with a resolution of 3-5 k.y./sample was generated for the site to 109 revised mcd (rmcd), or ~3.0 Ma (early Pliocene), based on Cibicides spp. (Harris, this volume; Hall et al., 2001) (Fig. 21B). The isotope measurements were evaluated in the context of the complete composite record available for the site to ~4.7 Ma, which has been tuned to an orbital timescale (Hall et al., 2001). The stable isotope measurements record the influence of North Atlantic and Southern Ocean deepwater masses on water properties at Site 1123. In addition, the Site 1123 record closely matches that of ODP Leg 138 Site 849 in the eastern equatorial Pacific (Mix et al., 1995), and the resulting age model agrees with the excellent paleomagnetic reversal record established for Site 1123 (Carter, R., McCave, Richter, Carter, L., et al., 1999; Wilson, 2000b).
For Site 1125, Yang et al. (2002) determined the oxygen and carbon isotope values of Globigerina bulloides and Uvigerina spp. for 207 samples spaced at ~24-k.y. intervals through the upper ~200 mbsf (0-5 Ma). The 18O results delineate two episodes of middle Pliocene warmth between 4.7 and 3.2 Ma and the start of a cooling trend at ~2.9 Ma, a little earlier than the 2.4 Ma accepted as the initiation of Northern Hemisphere glaciation (e.g., Raymo, 1994). During the middle Pliocene warm interval, the 13C profile of Uvigerina spp. shows a significant depletion trend, whereas the planktonic G. bulloides record displays relatively heavy values compared to the glacial times. These results are consistent with an increase in shallow-water and shelf biomass, perhaps supplied to Site 1125 via an invigorated East Cape Current system. During the late Pliocene global cooling, an enrichment of 12C in the surface ocean results in a decreasing trend of 13C in G. bulloides. The 13C profile of Uvigerina spp., however, shows a different trend from that of G. bulloides and indicates stronger production of AAIW during glaciations.
The effect of consolidation on sediment microfabrics and petrophysical properties is of considerable interest. In order to study these changes, Kim et al. (2001) subjected samples from Site 1125 to scanning electron microscope (SEM) observation both prior to and after laboratory consolidation tests. Preliminary X-ray diffraction and grain size analyses demonstrated that the samples studied were mineralogically and texturally similar and thus ideal for study. Porosity was measured before and after each consolidation test, and permeability was estimated based on the theoretical model of Bryant et al. (1986).
SEM images show that as porosity decreases the geometry and distribution of intergranular pores change significantly in concert. Clay plates with randomly distributed voids and abundant edge-to-edge and edge-to-face contacts in unconsolidated sediment lose these characteristics on consolidation. After consolidation, bedding-parallel fabrics and bedding-elongate voids are characteristic. Visual estimates of the porosity loss imposed by consolidation from SEM images generally agree with measured values. The results of this study therefore confirm the classic descriptions of clay fabric changes and porosity reduction during consolidation and show that consolidation is the major control on the petrophysical properties of buried sediments.
