Diatom biostratigraphy of the Kerguelen Plateau and Prydz Bay, Southern Ocean

Samples were examined for diatoms from 22 holes at 11 sites cored by ODP Leg 119 on the Kerguelen Plateau and in Prydz Bay, East Antarctica. Diatoms were observed in Oligocene through Holocene sediments recovered from the Kerguelen Plateau. The diatom flora from the Kerguelen Plateau is characterized by species such as Azpeitia oligocenica, Rocella gelida, Rocella vigilans, and Synedra jouseana in the Oligocene and Crucidenticula nicobarica, Denticulopsis hustedtii, Nitzschia miocenica, and Thalassiosira miocenica in the Miocene. This somewhat cosmopolitan assemblage gives way to a Pliocene and Holocene assemblage characterized by species such as Nitzschia kerguelensis, Thalassiosira inura, and Thalassiosira torokina, which are endemic to the Southern Ocean region. Samples examined from Prydz Bay are generally devoid of diatoms. The exception is Site 739, where diatoms occur sporadically in lower Oligocene and upper Miocene through Quaternary sediments. The Leg 119 diatom biostratigraphic results allow the development of a stratigraphic framework for the Indian sector of the Southern Ocean. This diatom zonation integrates diatom zonations developed previously for other sectors of the Southern Ocean. The zonation proposed here is based on biostratigraphic events of both geographically widespread and endemic species calibrated to the paleomagnetic stratigraphy. As such, this zonation has application throughout the Southern Ocean and allows correlation from the southern high latitudes to the low latitudes.

Stepwise consolidation of glacigenic sediments of the Prydz Bay

A series of samples from the five sites drilled across the continental shelf and upper slope in Prydz Bay during ODP Leg 119 were consolidation tested in an oedometer. Preconsolidation stresses increase downcore at Sites 739 and 742 in a stepwise manner, and the steps are interpreted to represent periods of increased action of grounded glaciers covering the entire shelf. By the use of theoretical ice sheet surface profiles giving the range of possible ice thicknesses, sediment loading and subsequent erosion seem to be the most important factor for increasing the overconsolidation ratios, and a total glacial erosion exceeding 1 km is possible. Four separate steps in consolidation, here termed "load events" have been identified. The lowermost load event, 1, is correlated to the onset of glaciations reaching the shelf edge and an early period of extensive glaciations, starting in early Oligocene or possibly earlier. Glacial activity related to the buildup of ice in West Antarctica in the late Miocene is tentatively correlated to load event 2. Event 3 is the trace of relatively extensive glacial erosion probably in the Pliocene, whereas the upper step in preconsolidation stress, load event 4, results from the last glaciation reaching the shelf edge, possibly during the late Weichselian. Correlations to other data related to Antarctic glacial history are, however, hampered by the poor age control of the cored diamictites. Consolidation tests may provide a tool for finding the position for hiatuses and unconformities formed subglacially and obscured by subglacial reworking.

(Table 1) Ice floe size and other characteristics during aerial observations in Prydz Bay, Antarctica

On the basis of aerial photographs of sea ice floes in the marginal ice zone (MIZ) of Prydz Bay acquired from December 2004 to February 2005 during the 21st Chinese National Antarctic Research Expedition, image processing techniques are employed to extract some geometric parameters of floes from two merged transects covering the whole MIZ. Variations of these parameters with the distance into the MIZ are then obtained. Different parameters of floe size, namely area, perimeter, and mean caliper diameter (MCD), follow three similar stages of increasing, flat and increasing again, with distance from the open ocean. Floe shape parameters (roundness and the ratio of perimeter to MCD), however, have less significant variations than that of floe size. Then, to modify the deviation of the cumulative floe size distribution from the ideal power law, an upper truncated power-law function and a Weibull function are used, and four calculated parameters of the above functions are found to be important descriptors of the evolution of floe size distribution in the MIZ. Among them, Lr of the upper truncated power-law function indicates the upper limit of floe size and roughly equals the maximum floe size in each square sample area. L0 in the Weibull distribution shows an increasing proportion of larger floes in squares farther from the open ocean and roughly equals the mean floe size. D in the upper truncated power-law function is closely associated with the degree of confinement during ice breakup. Its decrease with the distance into MIZ indicates the weakening of confinement conditions on floes owing to wave attenuation. The gamma of the Weibull distribution characterizes the degree of homogeneity in a data set. It also decreases with distance into MIZ, implying that floe size distributes increase in range. Finally, a statistical test on floe size is performed to divide the whole MIZ into three distinct zones made up of floes of quite different characteristics. This zonal structure of floe size also agrees well with the trends of floe shape and floe size distribution, and is believed to be a straightforward result of wave-ice interaction in the MIZ.