Isotopic indicators of deformation in the Red Sea

Abstract

Palaeostress analysis of spreading oceanic ridges requires data on extension over time and at different locations along strike so that mechanical models can be formulated and tested. The seafloor palaeomagnetic record provides average spreading rates but as the last (normal) chron measures 790,000 years an alternative method is required for identifying episodes of spreading at time scales of <10 3 yr which can be fused with seismic and geodetic sequences. On the assumption that the level of hydrothermal circulation on the sea floor is linked with tectonic activity, palaeotemperature and geochemical changes in the seawater determined from cores of the sea floor sediments should serve as proxies for episodes of extension provided they can be dated and alternative explanations for them ruled out. The Red Sea provides a study area for evaluating this proposal because it is a marginal sea with a shallow threshold which limits mixing with the open ocean especially during periods of low sea level, it receives little runoff from the arid lands that surround it, and its median ridge is hydrothermally active. Chemical evidence has been reported for two broadly synchronous late Quaternary hydrothermal episodes in the Atlantis II and Thetis deeps; 87Sr/ 86Sr ratios in the Atlantis II deep are thought to reflect variations in hydrothermal activity. Published δ 13C and δ 18O data from cores both within brine pools and outside them indicate fluctuations during the Holocene for which no climatic counterparts on bordering land areas or in the Arabian Sea are documented; moreover our study has revealed a major (positive) δ 18O excursion older than 40,000 yr in Globigerinoides ruber in Core CH10003-3 PC at a depth of 1374 m which is not associated with hardgrounds or a δ 13C fluctuation and which may have resulted from a hydrothermal pulse. There is no such signal in the core for events of magnitude up to 8.3 that are recorded in the palaeoseismic record of the last 50,000 yr in the Lisan Marls, which suggests that the δ 18O pulse from Core CH10003-3 corresponded with an event of M>8.3 in the RedSea-Dead Sea rift system Other isotopes may prove to be more sensitive and therefore responsive to smaller deformation.