Possible origin of K-rich volcanic rocks from Virunga, East Africa, by metasomatism of continental crustal material: Pb, Nd and Sr isotopic evidence

Abstract

The isotopic compositions of Sr, Nd and Pb together with the abundances of Rb, Sr, U and Pb have been determined for mafic and felsic potassic alkaline rocks from the young Virunga volcanic field in the western branch of the East African rift system. 87Sr/ 86Sr varies from 0.7055 to 0.7082 in the mafic rocks and from 0.7073 to 0.7103 in the felsic rocks. The latter all come from one volcano, Sabinyo. Sabinyo rocks have negative ε Ndvalues ofε Nd = −10. Nd and Sr isotopic variations in the basic potassic rocks are correlated and plot between Sabinyo and previously reported [1] compositions (ε Nd = +2.5; 87Sr/ 86Sr≈ 0.7047) for Nyiragongo nephelinites. The Pb isotopic compositions for Sabinyo rocks are nearly uniform and average 206Pb/ 204Pb≈ 19.4, 207Pb/ 204Pb= 15.79–15.84, 208Pb/ 204Pb≈ 41.2. The basic potassic rocks have similar 206Pb/ 204Pb values but range in 207Pb/ 204Pb and 208Pb/ 204Pb from the Sabinyo values to less radiogenic compositions. Excellent correlations of 87Sr/ 86Sr with Rb/Sr, 1/Sr and 207Pb/ 206Pb for Sabinyo rocks suggest these to be members of a hybrid magma series. However, the nearly uniform Pb compositions for this series points to radiogenic growth of 87Sr in the magma source region following an event which homogenized the isotopic compositions but not Rb/Sr. The Rb-Sr age derived from the erupted Sabinyo isochron-mixing line is consistent with the ∼500 Myr Pb-Pb age from Nyiragongo [1], which suggests that this event affected all Virunga magma sources. The event can again be traced in the Pb-Pb, Pb-Sr and Nd-Sr isotopic correlations for all Virunga rocks, including Nyiragongo, when allowances are made for radiogenic growth subsequent to this mixing or incomplete homogenization event. Inferred parent/daughter element fractionations point to a metasomatic event during which a mantle fluid invaded two lithospheric reservoirs: a +ε Nd reservoir sampled by the Nyiragongo nephelinites and suggested to be the subcontinental mantle and a −ε Nd reservoir sampled by the mafic and felsic potasssic volcanism. Whether this −ε Nd reservoir is the crust, continental crustal material in the mantle or anomalous mantle cannot be decided from the data. The simplest answer, that this reservoir is the continental crust, seems to be at variance with experimental evidence suggesting a subcrustal origin for basic potassic magmas. Partial melting of the ancient metasomatised lithospheric domains and ensuing volcanism seems to be entirely a response to decompression and rising geotherms during rifting and thinning of the lithosphere.