(Table 1) Radiolarians in the Oligocene-Miocene at DSDP Hole 72-515B

The Neogene of the southwestern Atlantic is virtually barren of biogenic silica. Of the four sites drilled on Leg 72, only two contained identifiable radiolarian specimens. In the southwestern Brazil Basin (Site 515), radiolarians are present only from the upper Oligocene (Anomaly 8, about 28 Ma) to the middle Miocene (Zone NN8, about 11.5 Ma). On the Rio Grande Rise (Site 516), radiolarians are present only within a short interval of the lower Miocene (Zones N5-N6, about 18-20 Ma). The abrupt cessation of silica deposition in the upper middle Miocene is characteristic of many drill sites in the tropical and temperate Atlantic and implies that a major oceanographic "threshold" was exceeded at this time, allowing the Atlantic waters to become either less productive or relatively silica deficient. Siliceous microfossils are notably more abundant in Oligocene-Miocene sediments of deep regions where carbonate preservation is poor (Site 515) than in equivalent carbonate-rich strata nearby (Site 516). This discrepancy suggests that the presence of calcareous microfossils may act to enhance post-depositional dissolution of biogenic silica tests by elevating the pH of the surrounding pore waters. Carbonate-free clays, by contrast, may provide a more favorable chemical environment for silica preservation.

Late Miocene carbon isotope and paleoproductivity record

We examine whether or not a relationship exists between the late Miocene carbon isotope shift (~7.6-6.6 Ma) and marine productivity at four sites from the Indian and Pacific Oceans (Ocean Drilling Program Sites 721, 1146, 1172, and 846). We use a multiproxy approach based on benthic foraminiferal accumulation rates, elemental ratios, and dissolution indices, and we compare these data to benthic foraminiferal d13C values measured on the same samples. Although some of these sites have been targeted previously in studies of either the late Miocene/early Pliocene "biogenic bloom" (Sites 721 and 846) or the late Miocene carbon isotope shift (Site 1172), our records are the first to establish paired proxy records of carbon isotopes and paleoproductivity allowing a direct assessment of a potential link. Our results indicate that at all sites, productivity increased sometime during the d13C shift; at three sites (721, 1146, and 846), productivity increased at the beginning of the shift. The correlation coefficients derived from linear regression between micropaleontologically derived productivity and foraminiferal d13C values are relatively high during the time interval containing the late Miocene d13C shift (and statistically significant at three of the sites). Carbon flux and isotope mass balance considerations illustrate that transfer of organic matter between the terrestrial and marine reservoirs together with enhanced oceanic upwelling best approximates observed changes in carbon isotope records and paleoproductivity. We note that long-term trend in the Site 846 paleoproductivity record can be correlated to the long-term trend in the Site 848 eolian flux reconstructions of Hovan (1995, doi:10.2973/odp.proc.sr.138.132.1995) hinting at a link between strengthened wind regime and productivity during the late Miocene.

Taxonomic lists and abundances of the Lower Miocene bivalve assemblages of the Vives quarry (Meilhan, SW France)

Biodiversity estimates through geological times are difficult because of taphonomic perturbations that affect sedimentary records. Pristine shell assemblages, however, allow for calibration of past diversity. Diversity structures of two exceptionally preserved Miocene bivalve assemblages are quantitatively determined, compared with recent communities and used as paleoenvironmental proxy. The extremely rich assemblages were collected in Aquitanian (Early Miocene) carbonate sands of the Vives Quarry (Meilhan, SW France). Both paleontological and sedimentological data indicate a coral patch-reef environment, which deposits were affected by transport processes. Among two samples more than 28.000 shells were counted and 135 species identified. Sample Vives 1 is interpreted as a proximal debris flow and Sample Vives 2 as a sandy shoreface/foreshore environment influenced by storms. The two Vives assemblages have a similar diversity structure despite facies differences. Rarefaction curves level off at ~600 shells. The rare species account for more than 80 % of the species pool. The high values of PIE diversity index suggest a relatively high species richness and an even distribution of abundance of the most common species within the assemblages. The fossil data are compared to death shell assemblages (family level) of a modern reefal setting (Touho area, New Caledonia). The shape of the rarefaction curves and PIE indices of Meilhan fossil assemblages compare well to modern data, especially those of deep (>10 m water depth), sandy depositional environments found downward the reef slope (slope and pass settings). In addition to primary ecological signals, the similarity of the Vives samples and the Recent deep samples derives from taphonomic processes. This assumption is supported by sedimentological and paleontological observations. Sediment transports gather allochthonous and in situ materials leading to mixing of various ecological niches. Such taphonomic processes are recorded in the diversity metrics. Environmental mixing and time-averaging of the shell assemblages disturb the preservation of local-scale diversity properties but favour the sampling of the regional-scale diversity.

Stratigraphic distribution of marine palynomorph species recorded for the Miocene of ODP Hole 105-645E (Table 1)

A total of 145 samples were analyzed for palynology, and all were found to be productive. Residues are dominated by pollen, terrestrial spores, and land plant tissues. Marine palynomorphs occur in all samples, which allowed us to recognize five Miocene dinocyst assemblage zones. Dinocyst assemblages indicate cool-water conditions and suggest a neritic rather than fully oceanic environment, with not only North Atlantic and Norwegian Sea affinities, but also containing both notable protoperidiniacean and possible endemic elements. Dinocyst assemblages indicate an early Miocene age for the bottom of Hole 645E and an age no younger than early late Miocene (Sample 105-645E-24R, CC) near the top of the interval studied. These age assignments provide an estimated initiation of ice rafting in Baffin Bay at between 7.4 and 9.5 Ma. Increased terrigenous influx and apparent disappearance of certain dinocyst taxa occur in the middle to late Miocene and may be related to oceanographic changes or climatic deterioration. Spores and pollen indicate a climate that varied within a temperate regime during the early and middle to early late Miocene, followed by climatic deterioration. Four new dinocyst species are described: Batiacasphaera gemmata, Impletosphaeridium prolatum, Operculodinium vacuolatum, and Selenopemphix brevispinosa. The acritarch genus Cyclopsiella Drugg and Loeblich is emended, and two new combinations have been created: Cyclopsiella granosa (Matsuoka) and Cyclopsiella? laevigata (Chateauneuf). Cyclopsiella granosa (Matsuoka) n. comb. is considered a subjective junior synonym of Cyclopsiella granulata He and Li. Ascostomocystis granulatus Chateauneuf has been provisionally allocated to Cyclopsiella and renamed Cyclopsiella? chateauneufii. Two new acritarch species are described: Cyclopsiella spiculosa and Cymatiosphaera! baffinensis.

Stable carbon and oxygen isotope ratios of Late Miocene and Early Pliocene foraminifera from the Southwest Pacific

The Late Miocene-Early Pliocene paleoclimatic history has been evaluated for a deep drilled sediment sequence at Deep Sea Drilling Project Site 281 and a shallow water marine sediment sequence at Blind River, New Zealand, both of which lay within the Subantarctic water mass during the Late Miocene. A major, faunally determined, cooling event within the latest Miocene at Site 281 and Blind River coincides with oxygen isotopic changes in benthonic foraminiferal composition at DSDP Site 284 considered by Shackleton and Kennett (1975) to indicate a significant increase in Antarctic ice sheet volume. However, at Site 281 benthonic foraminiferal oxygen isotopic changes do not record such a large increase in Antarctic ice volume. It is possible that the critical interval is within an unsampled section (no recovery) in the latest Miocene. Two benthonic oxygen isotopic events in the Late Miocene (0.5 ? and 1 ? in the light direction) may be useful as time-stratigraphic markers. A permanent, negative, carbon isotopic shift at both Site 281 and Blind River allows precise correlations to be made between the two sections and to other sites in the Pacific region. Close interval sampling below the carbon shift at Site 281 revealed dramatic fluctuations in surface-water temperatures prior to a latest Miocene interval of refrigeration (Kapitean) and a strong pulse of dissolution between 6.6 and 6.2 +/- 0.1 m.y. which may be related to a fundamental geochemical change in the oceans at the time of the carbon shift (6.3-6.2 m.y.). No similar close interval sampling at Blind River was possible because of a lack of outcrop over the critical interval. Paleoclimatic histories from the two sections are very similar. Surface water temperatures and Antarctic ice-cap volume appear to have been relatively stable during the late Middle-early Late Miocene (early-late Tongaporutuan). By 6.4 m.y. cooler conditions prevailed at Site 281. Between 6.3 and 6.2 -+ 0.1 m.y. the carbon isotopic shift occurred followed, within 100,000 yr, by a distinct shallowing of water depths at Blind River. The earliest Pliocene (Opoitian) is marked by increasing surface-water temperatures.