(Appendix B) Benthic foraminifera in Indian Ocean surface sediments

Multivariate analysis was performed on percentages of 46 species of unstained deep-sea benthic foraminifera from 131 core-top to near-core-top samples (322-5013 m) from across the Indian Ocean. Faunal data are combined with GEOSECS geochemical data to investigate any relationship between benthic foraminifera (assemblages and species) and deep-sea properties. In general, benthic foraminifera show a good correlation to surface productivity, organic carbon flux to the sea floor, deep-sea oxygenation and, to a lesser extent, to bottom temperature, without correlation with the water depths. The foraminiferal census data combined with geochemical data has enabled the division of the Indian Ocean into two faunal provinces. Province A occupies the northwestern Indian Ocean (Arabian Sea region) where surface primary production has a major maximum during the summer monsoon season and a secondary maximum during winter monsoon season that leads to high organic flux to the seafloor, making the deep-sea one of the most oxygen-deficient regions in the world ocean, with a pronounced oxygen minimum zone (OMZ). This province is dominated by benthic foraminifera characteristic of low oxygen and high organic food flux including Uvigerina peregrina, Robulus nicobarensis, Bolivinita pseudopunctata, Bolivinita sp., Bulimina aculeata, Bulimina alazanensis, Ehrenbergina carinata and Cassidulina carinata. Province B covers southern, southeastern and eastern parts of the Indian Ocean and is dominated by Nuttallides umbonifera, Epistominella exigua, Globocassidulina subglobosa, Uvigerina proboscidea, Cibicides wuellerstorfi, Cassidulina laevigata, Pullenia bulloides, Pullenia osloensis, Pyrgo murrhina, Oridorsalis umbonatus, Gyroidinoides (= Gyroidina) soldanii and Gyroidinoides cf. gemma suggesting well-oxygenated, cold deep water with low (oligotrophic) and pulsed food supply.

Benthic foraminifera in Southwest Indian Ocean surface sediments

The distribution of deep-sea benthonic foraminifera in core top samples from the southwest Indian Ocean is examined. Principal component analysis reveals two major assemblages. One assemblages between 3600 and 4800-m water depth is dominated by Episominella umbonifera and is associated with cold (Theta = -0.3 to 0.8°C), low salinity (34.66 to 34.72 * 10**-3) Antarctic Bottom Water in the Crozet Basin, in fracture zones, and on the flanks of the Southwest Indian Ridge. A second assemblage, dominated by Planulina wuellerstorfi, Globocassidulina subglobasa, Astrononion echolsi and Pullenia bulloides, is between 1600 and 3800 m on the Crozet Plateau, Madagascar Ridge, Central Indian Ridge, and Southwest Indian Ridge and is associated with relatively warm (Theta = 0.8 to 2.6°C), high salinity (34.72 to 34.76 * 10**-3) North Atlantic Deep Water. The third principal component divides the P. wuellerstorfi assemblage into two subgroups. One is dominated by Epistominella exigua, P. bulloides, P. wuellerstorfi, and A. echolsi and a second is dominated by G. subglobosa. The distribution of the E. umbonifera assemblage and previous hydrographic studies suggest that AABW flows as a western boundary contour current in the Crozet Basin and penetrates fracture zones in the Southwest Indian Ridge between 55 and 57°E and near 66°E as it travels northward into the Madagascar and Mascarene basins. The faunal-water mass associations from the southeast Indian Ocean are compared; the most notable faunal difference is the absence of Uvigerina as a dominant taxon in the southwest Indian Ocean. A comparison of dissolved oxygen and Uvigerina data shows that oxygen is not a major influence upon the distribution of Uvigerina. A correlation analysis of the faunal data and water depth, potential temperature, in situ temperature, salinity, dissolved oxygen, and 1 - Omega, an index of calcium carbonate undersaturation, was carried out to determine the relationships between fauna and hydrography. The second principal component has a significant positive correlation at the 99.9% level with temperature and negative correlations with water depth and 1 - Omega. A general faunal-water mass correlation exists, but it is not possible to determine which variable controls the faunal distributions.

Distribution of dinoflagellate cycst at ODP Site 123-765 (Table 1)

Site 765 contains a sequence of tropical, middle Miocene to Holocene dinoflagellate cysts. These diverse assemblages are characterized by abundant Polysphaeridium zoharyi and Spiniferites bulloideus. Abundances of Impagidinium spp. and Nematosphaeridium spp. reflect the shelf-to-slope origin of the assemblages. One new genus, Blysmatodinium, and two new species, Nematosphaeridium (?) wrennii sp. nov. and Blysmatodinium argoi, are described.

Pollen abundance in ODP Site 123-765 (Table 1)

Middle Miocene to Holocene pollen assemblages reveal a history of environmental change in northern Australia. Grass pollen appeared, but was rare, in the late Miocene and was consistently present throughout the Pliocene, but did not become abundant until the Pleistocene. Myrtaceae pollen, characteristic of late Cenozoic assemblages in eastern Australia, is poorly represented, and no unequivocal evidence of rain forest was found.

Geochemistry of ODP Sites 123-765 and 123-766

The sediments of the Argo and Gascoyne abyssal plains are generally lean in organic matter, are immature, and contain hydrocarbons trapped during sediment deposition rather than those generated during sediment catagenesis. TOC concentrations in the Argo Abyssal Plain Cenozoic sediments are 0.5 wt%, and organic matter appears to be from mixed marine and reworked, degraded, organic matter sources, with the latter being contributed by turbidity flows from the nearby continental margin. TOC concentrations within the Cenozoic sediments of the Gascoyne Abyssal Plain are mostly undetectable (<0.1 wt%). Biomarker distributions determined by gas chromatography (GC) and gas chromatography-mass spectrometry (GCMS) indicate that organic matter extracted from the Lower Cretaceous sediments from both sites is predominantly marine with varying contributions from terrestrial organic matter. The specific marine biomarker, 24-n-propylcholestane is in relatively high abundance in all samples. In addition, the relatively high abundance of the 4-methylsteranes with the 23,24-dimethyl side chain (in all samples) indicates significant dinoflagellate contributions and marine organic matter. The ratios of n-C27/n-C17 reflect relative contributions of marine vs. terrestrial organic matter. TOC, while generally low at Argo, is relatively high near the Barremian/Aptian boundary (one sample has a TOC of 5.1 wt%) and the Aptian/Albian boundary (up to 1.3 wt% TOC), and two samples from the Barremian and Aptian sections contain relatively high proportions of terrestrial organic carbon. TOC values in the Lower Cretaceous sediments from Gascoyne Abyssal Plain are low (<0.1 wt%) near the Aptian/Barremian boundary. TOC values are higher in older sediments, with maxima in the upper Barremian (1.02 wt%), the Barremian/Hauterivian (0.6 wt%), and Valanginian (1.8 wt%). Sediments from the upper Barremian contain higher amounts of terrestrial organic carbon than older sediments.

Geochemistry of Early Cretaceous basaltic rocks from ODP Sites 123-765 and 123-766

The Early Cretaceous basaltic rocks obtained from Sites 765 and 766 in the eastern Indian Ocean floor were mostly iron-rich normal mid-ocean ridge basalts (N-MORB), which were derived from a depleted mantle source having strongly light rare earth element (LREE)-depleted rare-earth patterns and a high titanium/zirconium (Ti/Zr) ratio. Basaltic rocks in the upper part of the Site 765 basement section include megacrysts and gabbroic fragments of widely varying mineral chemistry. These megacrysts range from An90 plagioclase, including highly magnesian basaltic glass coexisting with augite of Mg# (= 100 Mg/[Fe+Mg]) at 85, to An50 plagioclase coexisting with hypersthene. This varying mineralogy of megacrysts and gabbroic fragments indicates that a considerable degree of fractional crystallization took place in the magma chamber. The unusual negative correlation between incompatible elements (e.g., TiO2) and FeO*/MgO observed among Site 765 basement basalts and fresh volcanic glasses suggest source-mantle heterogeneity in terms of FeO*/MgO. Strontium isotope ratios (87Sr/86Sr) of the basaltic rocks from both sites are between 0.7027 and 0.7033 and are comparable to those of mid-Indian Ocean ridge basalts (MIORB). The basalt pebbles encountered in the sedimentary section may have been transported from the basement highs nearer the Australian continent and include basaltic compositions ranging from primitive N-MORBs to evolved enriched (E)-MORBs. Their mantle source was not as depleted as that of the basement basalts. These rocks may be the products of earlier volcanism that took place during the rifting of the Australian continent.

Composition of pore water and diagenetic carbonates of ODP Site 123-765

Degradation of organic matter in slightly organic-rich (1 wt% organic carbon) Neogene calcareous turbidites of the Argo Basin at Site 765 by sulfate reduction results in pore-water phosphate, ammonium, manganese, and carbonate alkalinity maxima. Pore-water calcium and magnesium decrease in the uppermost 100 meters below seafloor (mbsf) in response to the precipitation of calcian dolomite with an average composition of Ca1.15Mg0.83Fe0.02(CO3)2. Clear, euhedral dolomite rhombs range from <1 to 40 µm in diameter and occur in trace to minor amounts (<1-2 wt%) in Pleistocene to Pliocene sediment (62-210 mbsf) The abundance of dolomite increases markedly (2-10 wt%) in Miocene sediment (210-440 mbsf). The dolomite is associated with diagenetic sepiolite and palygorskite, as well as redeposited biogenic low-Mg calcite and aragonitic benthic foraminifers. Currently, dolomite is precipitating at depth within the pore spaces of the sediment, largely as a result of aragonite dissolution. The rate of aragonite dissolution, calculated from the pore-water strontium profile, is sufficient to explain the amount of dolomite observed at Site 765. A foraminiferal aragonite precursor is further supported by the carbon and oxygen isotopic compositions of the dolomite, which are fairly close to the range of isotopic compositions observed for Miocene benthic foraminifers. Dolomite precipitation is promoted by the degradation of organic matter by sulfate-reducing bacteria because the lower pore-water sulfate concentration reduces the effect of sulfate inhibition on the dolomite reaction and because the higher carbonate alkalinity increases the degree of saturation of the pore waters with dolomite. Organic matter degradation also results in the precipitation of pyrite and trace amounts of apatite (francolite), and the release of iron and manganese to the pore water by reduction of Fe and Mn oxides. Spherical, silt-sized aggregates of microcrystalline calcian rhodochrosite occur in trace to minor amounts in Lower Cretaceous sediment from 740 to 900 mbsf at Site 765. A negative carbon isotopic composition suggests that the rhodochrosite formed early in the sulfate reduction zone, but a depleted oxygen isotopic composition suggests that the rhodochrosite may have recrystallized at deeper burial depths.

Physical properties of ODP Sites 123-765 and 123-766

During Ocean Drilling Program Leg 123, two sites were drilled in the deep Indian Ocean. Physical properties were measured in soft Quaternary and Lower Cretaceous sediments to relatively fresh, glass-bearing pillow lavas and massive basalts. Porosities ranged from 89% near the seafloor to 1.6% for the dense basalts. This self-consistent set of measurements permitted some descriptive models of physical properties to be more rigorously tested than before. Predictive relationships between porosity and compressional-wave velocity have generally been based upon the Wyllie time average equation. However, this equation does not adequately describe the actual relationship between these two parameters, and many have attempted to improve it. In most cases, models were derived by testing them against a set of data representing a relatively narrow range of porosity values. Similarly, the use of the Wyllie equation has often been justified by a pseudolinear fit to the data over a narrow range of porosity values. The limitations of the Wyllie relationship have been re-emphasized here. A semi-empirical acoustic impedance equation is developed that provides a more accurate porosity-velocity transform, using realistic material parameters, than has hitherto been possible. A closer correlation can be achieved with this semi-empirical relationship than with more theoretically based equations. In addition, a satisfactory empirical equation can be used to describe the relationship between thermal conductivity and porosity. If enough is known about core sample lithologies to provide estimates of the matrix and pore water parameters, then these predictive equations enable one to describe completely the behavior of a saturated rock core in terms of compressional-wave velocity, thermal conductivity, porosity, and bulk density.