Late Pleistocene Upwelling and Productivity Variations in the Northwest Indian Ocean Deduced from Spectral Analyses of Geochemical Data from Sites 722 and 724

Abstract

The Oman Margin and Owen Ridge underlie a seasonal upwelling system driven by the Southwest Monsoon. Variations in sediment composition partly reflect changes in the intensity of upwelling and productivity. Samples spanning the last 350 k.y. were obtained from ODP Site 724 on the Oman Margin and Site 722 on the Owen Ridge. Inorganic geochemical time series were investigated with spectral and cross-spectral techniques in order to investigate the long-term variability of the upwelling system. The Oman Margin records show no strongly cyclic patterns, probably due to bioturbation and the input of fluvial material. Conversely, the Owen Ridge records are strongly cyclic being composed predominantly of pelagic carbonate and wind-blown terrigenous material. The Ridge records have permitted reconstruction of the upwelling history. The amount of calcium carbonate on the Owen Ridge varies at three of the Milankovitch orbital periods (100, 41, and 23 k.y.) and is coherent and in phase with global ice volumes. Maximum accumulation rates of wind-blown terrigenous material are inversely related to carbonate contents. The flux of terrigenous material is probably determined by continental aridity. Thus, minimum carbonate contents correspond to maximum aridity and glacial conditions. Other studies (this volume) have demonstrated that wind velocity controls the sorting and grain size of the terrigenous material deposited on the ridge. Wind velocities vary at the three orbital periods, with the 23 k.y. orbital precession period dominant, reflecting the direct link between orbital insolation patterns and regional atmospheric pressure (i.e., global ice volumes are not considered as important). The proportions of heavy minerals in the non-carbonate fraction are recorded by Ti/Al and Cr/Al. These parameters vary at the three orbital periods, and are dominated by 23 k.y. forcing, suggesting that the amount of heavy minerals transported to the Ridge is controlled by wind speed. However, maximum heavy mineral contents are in phase with maximum terrigenous flux rather than maximum grain size. This suggests that aridity determines the timing of heavy mineral transport. Because the heavy minerals are coarser and denser than average grains, the proportion making up the wind-blown sediment depends on the wind speed during arid periods. Wind velocity also influences upwelling intensity, yet ocean productivity records (Ba/Al, P/Al, Ni/Al) are dominated by 100 k.y. cycles, rather than the 23 k.y. precession periodicity, suggesting an indirect link with global ice volumes. Productivity records are coherent with carbonate, but not ice volumes at the 23 k.y. period. Additionally, productivity and %CaCO3 are in phase at the 100 and 41 k.y. periods. We suggest that at these frequencies, productivity is related to the supply of nutrients from continental runoff and hence aridity. At the 23 k.y. period, productivity variations lagged %CaCO3 by about 40° suggesting that as well as runoff there was another, delayed source of nutrients, possibly intermediate water upwelled by the effect of the Southwest Monsoon.