Abstract

Re–Os results are reported for a suite of spinel peridotite xenoliths from the Sidamo region of Ethiopia, near the southernmost tip of the Main Ethiopian Rift (MER). The sulfide petrography and platinum group element (PGE) characteristics, as well as the major and trace element geochemistry of these samples have been investigated in previous studies [Earth Planet. Sci. Lett. 153 (1997) 67; Chem. Geol. 196 (2003) 57], and shown to reflect the effects of melt percolation, probably related to the impingement of the mantle plume that initiated rifting. The Os isotopic ratios correlate roughly with indices of melt depletion in the peridotites, providing evidence for a long period of radiogenic ingrowth subsequent to an ancient melt extraction event. Nevertheless, this correlation has been perturbed, probably during the melt percolation process documented by the earlier studies. Consideration of the Os data in conjunction with the sulfide petrography, PGE spectra, and geochemical and textural variations of the rocks indicates that the mechanism of Os mobility changed as the melts migrated upwards through the percolation column. In the high porosity region near the bottom of the column, now represented by the peridotites with granular textures, Os from the passing melts was scavenged by monosulfide solid solution phases. The apparent modification of the Os isotopic ratios in these rocks suggests that large quantities of melt, perhaps 2 to 5 g per gram of rock, passed through the granular facies peridotites. In the shallower, low porosity region represented by peridotites with deformed textures, the melt was highly evolved and volatile rich. These volatiles may have served as the vector for Os mobility evident in the high 187Os/ 188Os ratios of the harzburgitic samples. Despite the effects of recent melt percolation, the main feature of the Os systematics is the general correlation of 187Os/ 188Os with indices of melt depletion. This demonstrates that Os isotopes may still provide useful chronological information, even in regions where the lithosphere has been strongly perturbed by the impingement of a mantle plume. Model ages obtained from this correlation, as well as the Re-depletion age of the most depleted sample, indicate that melt extraction occurred between 2.4 and 2.8 Ga. This is much older than the Pan-African age of the overlying crust [J. African Earth Sci. 26 (1998) 207], and could suggest a major decoupling between the crust and the lithospheric mantle in this region, with the mantle representing an extension of the nearby Tanzanian cratonic root. However, the geochemistry of the Sidamo peridotites does not resemble that of typical cratonic mantle [Earth Planet. Sci. Lett. 96 (1989) 15]. Furthermore, several recent studies [Earth Planet. Sci. Lett. 177 (2000) 319; Ninth Annual V.M. Goldschmidt Conference, Abstract #7389, LPI Contribution No. 971, Lunar and Planetary Institute, Houston (CD-ROM), 1999; Geochim. Cosmochim. Acta 55 (1991) 1421; Science 281 (1998) 2011] have shown that large masses of depleted material can survive in the convecting mantle for hundreds of millions of years. Thus it may be unwise to equate the time of melt extraction with the time of lithosphere stabilization in a given region.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call