238U - 230Th - 226Ra Disequilibria in Volcanics of Karthala Volcano: Evidences of Plume-Lithosphere Interactions in the Source of the Volcanism

Christelle Claude-Ivanaj Laboratoire de Géochimie et Cosmochimie, URA CNRS 1758, IPG Paris,

4, place Jussieu, 75252 paris cedex 05, France


Claude J. Allègre Laboratoire de Géochimie et Cosmochimie, URA CNRS 1758, IPG Paris,

4, place Jussieu, 75252 Paris cedex 05, France

Use of radioactive disequilibria between the nuclides of the U-Th series in recent volcanic rocks offers a powerful approach in the study of magmatic processes. In effect, these disequilibria can provide a quantitative measurement of both the extent and time scale of incompatible element fractionation during magma formation.

Here we present 238U - 230Th - 226Ra disequilibria of young lavas of the Karthala volcanic series in the Comores islands. The Comore archipelago, along which volcanism appears to be age progressive, is considered to have been formed by the movement of the Somali plate over a hot spot. This volcano, located in the Grande Comore island, is the presently active shield volcano of the archipelago.

Precise thorium and radium isotopic measurements were performed using a new technique developped in our laboratory on a high abundance sensitivity TIMS (Finnigan Mat 262) equipped with a Retarding Potencial Quadrupole.

All the lavas possess significant 230Th and 226Ra
excesses similar to those observed in MORB and in many other OIB and consistent with the determined bulk partition coefficients (D) in the source where D(Ra)<<D(Th)<<D(U) during melting processes. Although (226Ra/230Th) and (230Th/238U) ratios are in many cases positively correlated (Chabaux and Allègre, 1994), here we find an inverse relationship. U-Th-Ra disequilibria are also correlated with the strontium isotopic variability of the lavas and thus the observed negative correlation is likely to be related to recent mixing in the source rather than fractionation processes. This is also supported by the Th/U ratio of the source (2.8), calculated assuming instantaneous melt extraction. This value, lower than the Th/U ratio commonly encountered in the source of oceanic islands (3-3.7) (Newman et al., 1984) suggests that the Karthala plume consists in a mixure of primitive (Th/U=4) and depleted mantle sources (Th/U=2.5) (Allègre et al., 1982). As noted previously by Class et al. (1994), the isotopic composition of the deap plume is probably represented by the most radiogenic lavas whereas the depleted overlying oceanic lithosphere may be a reservoir for generating the lowest 87Sr/86Sr signatures. In terms of radioactive disequilibria, the lithospheric component is characterised by the highest (230Th/238U) ratios and the lowest (226Ra/230Th) ratios. Further work is necessary in order to determine whether this is the signature of very low degrees of melting or reflects trace elements enrichment in this 100 Ma old oceanic lithosphere.


Allègre, C.J., Dupré, B. & Léwin, E., Nature 299, 21-24 (1982).

Chabaux, F. & Allègre, C.J., Earth Planet . Sci. L ett. 126, 61-74 (1994).

Class, C. & Goldstein, S.L., A.G.U Fall Meeting V31D-5 (1994).

Newman, S., Finkel, R.C., Mac Dougall, J.D., Earth Planet . Sci. L ett. 65, 315-324 (1984).