INTRODUCTION

Objective 3 of Leg 183 was to determine mass transfer between the volcanic crust and atmosphere-hydrosphere-biosphere system. Basaltic volcanism affects the Earth's climate system, both during the constructive stages, through emission of aerosols and greenhouse gases, and also during the destructive stages, through consumption of carbon dioxide by silicate weathering and organic carbon (OC) burial in volcanogenic sediments (Berner, 1990; Reusch and Maasch, 1998). In large igneous provinces such as the Kerguelen Plateau, these effects may be globally significant. Mafic volcanic rocks contain high concentrations of Ca and Mg silicates and are relatively soluble, making them efficient carbon sinks; chemical denudation rates in basaltic catchments are on the order of 14 x 106 g/km2/yr (1.4 kg/cm2/m.y.) (Bluth and Kump, 1994). Organic carbon burial adjacent to eroding volcanic provinces may also be enhanced because of the large surface area of the layer-silicate weathering products (Mayer, 1994). For example, continental margin sediments contain on the order of 1 mg OC/m2 of surface area, and clay-sized fractions yield specific surface areas of 60 m2/g (Bock and Mayer, 2000), so that clay-rich sediments might be expected to contain up to several weight percent OC.

This study contrasts the carbon fluxes associated with erosion of a volcanic terrane (Skiff Bank) with those in the Bengal Fan associated with erosion of middle-crustal sialic material. In the Bengal Fan, organic carbon burial far exceeds carbon sequestration by silicate weathering, in part because of the low concentrations of Ca and Mg silicates in the metamorphic core of the Himalayas and in part because of the highly productive lowlands traversed by the Ganges and Brahmaputra Rivers (France-Lanord and Derry, 1997). In the current study, in addition to measuring the organic carbon flux, I compare the average compositions of terrigenous sediment and volcanic basement to constrain silicate weathering-related carbon dioxide consumption.

An original intent of this study was to characterize weathering of the Kerguelen flood basalt province. At Sites 1136, 1137, and 1138, the tops of subaerially erupted flood basalts are only slightly oxidized, suggesting that little weathering occurred between eruptions. Direct study of weathering profiles at the top of the basement is generally impossible because the regolith has been washed away. (The exception is Broken Ridge, where basalts were pervasively altered to clays following uplift on a rift shoulder and northward translation to a hot and humid climate.)

Site 1139 affords an opportunity to investigate carbon fluxes related to the weathering and erosion of volcanic rocks under variable conditions through the late Oligocene and early Miocene (~30-15 Ma). The site lies in a perched basin on the southwest flank of Skiff Bank, a submerged volcano on the northwest extremity of the Kerguelen Plateau in the southern Indian Ocean (50°11´S, 63°56´E; 1450 m below sea level) (Fig. F1). In Unit II (Shipboard Scientific Party, 2002), volcanogenic materials are mixed with pelagic carbonate, which varies on several scales between 10 and 90 wt%. The bulk of the volcanogenic materials are considered to be terrigenous (i.e., products of terrestrial weathering and erosion) based on presence of clay minerals, predominance of silt, sedimentation rates that are consistent with erosion of a volcanic island, and very rare presence of pyroclastic material such as shards (Shipboard Scientific Party, 2000). Pelagic components record local surface ocean conditions, and benthic foraminifers record bottom conditions including water depth. Ideally, it might be possible to determine how carbon fluxes responded to changing relative sea level and climate. Did increased erosion during glacioeustatic lowstands offset the effect on weathering of low glacial temperatures? How was the organic carbon flux affected by mineral surface area contributed by the terrigenous component?

Hole 1139A cores also bear on the tectonic evolution of the Kerguelen Plateau (Objective 4 of Leg 183) through the Cenozoic. Water depth estimates based on benthic foraminifer assemblages, in combination with glacioeustastic records, are used to reconstruct the subsidence/uplift history of Skiff Bank. The Kerguelen Plateau sits in a critical position with respect to the Antarctic Circumpolar Current (ACC), a major component of the Antarctica-Southern Ocean climate system; in particular, any vertical movements of the plateau may have affected this circulation. Whether or not plateau tectonism affected the ACC, the origin of some carbonate variations in Hole 1139A cores may be rooted in oscillations of this circulation. Unit II sediments potentially contain a detailed record of climate through the Oligocene and Miocene that is analogous to records from other sites receiving mixed terrigenous and pelagic sediment, for example, Ceara Rise near the mouth of the Amazon River (Shackleton, Curry, Richter, and Bralower, 1997) and Site 594 off the east coast of New Zealand (Nelson et al., 1986). At these sites, light oxygen isotopic ratios (interglacials) correspond with high carbonate concentration, high reflectance, and low terrigenous content as indicated by low magnetic susceptibility and low natural gamma activity (and vice versa) (Shackleton, Curry, Richter, and Bralower, 1997; Nelson et al., 1986).

In summary, the major questions posed during this study include

What are the causes of the fluctuations in carbonate concentration?
What is the subsidence/uplift history of Skiff Bank?
What climate signal can be extracted from Hole 1139A cores?
What were the carbon fluxes and how did they respond to changing climate and relative sea level?

NEXT