Recycled ocean crust and sediment in Indian Ocean MORB

Abstract

Mid-ocean ridge basalts (MORB) from the Indian Ocean are characterized by higher 87Sr/ 86Sr and lower 206Pb/ 204Pb, on average, than Atlantic and Pacific MORB. These distinct isotopic characteristics are thought to be the result of large-scale mantle contamination, either by ancient recycled sediments or delaminated subcontinental lithosphere. To constrain the upper mantle mixing end-members, we have analyzed the concentrations of Nb, U, Th, Pb and 17 other incompatible elements (by HPLC and ID-TIMS methods), as well as Sr, Nd, and Pb isotope ratios in nine hand picked MORB glasses from the Central Indian and Carlsberg Ridge. Nb/U and Ce/Pb ratios are sensitive indicators of crustal contaminants in the mantle, and such contaminations have been identified in EM-type ocean island basalts. Here we show the first case of highly significant correlations of Nb/U, Ba/Nb and Ba/Th with 87Sr/ 86Sr for Indian Ocean MORB and we confirm and strengthen a previously detected correlation with Ce/Pb (or Nd/Pb). These correlations are interpreted as residual heterogeneities that have not been completely rehomogenized by convective stirring of the mantle. The results are explained by a quantitative model involving contamination of an Atlantic-Pacific type upper mantle with a heterogeneous, recycled component consisting of ∼ 1.5 Ga old altered oceanic crust with small, but variable, amounts of ancient pelagic sediment added. This contamination causes higher 87Sr/ 86Sr, lower 143Nd/ 144Nd, generally low but variable 206Pb/ 204Pb and the distinct trace element correlations seen in the Indian Ocean MORB data. Mass balance calculations show that only small amounts of pelagic sediment are necessary to satisfy the criteria of the contamination model, and these can be subducted within as little as 200 Ma. Lead isotopic systematics constrain the recycled sediment component to be older than about 1.25 Ga. Apart from this, the model does not constrain the age of the contamination event since the results are equally valid for (1) “slow” mixing of the recycled component subsequent to subduction at ∼ 1.5 Ga or (2) storage of the material after subduction in a mantle boundary layer followed by relatively recent reinjection into the MORB source and “fast” mixing.