Regional geodynamic context for the Mesoproterozoic Kibara Belt (KIB) and the Karagwe-Ankole Belt: Evidence from geochemistry and isotopes in the KIB

Abstract

The Mesoproterozoic Kibara Belt and Karagwe Ankole Belt formed between the Congo Craton and the Tanzania-Bangweulu Block in central Africa. The subduction model proposed for the Kibara region is difficult to reconcile with the simultaneous intracratonic transtension in Karagwe Ankole at c. 1380Ma. Nonetheless, our whole rock petrochemical and Sr–Nd isotope data reinforces earlier evidence for a subduction setting for the Kibara Belt. Model ages for the mafic samples associated with the 1380Ma event indicate crustal residence times of c. 2.7Ga, suggesting derivation from evolved subcontinental lithospheric mantle with minor juvenile input. The latter is analogous to an active continental margin. In combination with the petrochemical evidence, this favors a subduction setting, the bimodal nature of the magmatism notwithstanding.The Kibara Belt granites are weakly metaluminous to strongly peraluminous, with low Ba – high Rb contents and restitic micaceous enclaves indicating an origin by metapelite dehydration melting. Two previously recognized granite suites around 1380Ma and 1000Ma in the Kibara Belt are, to an extent, distinguishable geochemically. The age-corrected Nd isotope ratios from Kibara Belt metapelites and granites are similar, confirming that the strongly peraluminous Kibara Belt granites are mostly derived from similar metapelites. The Kibara Belt granites show depleted mantle Nd model ages (TDM) of at least 2.4Ga. Some granites have model ages in excess of 3.0Ga, which can be attributed to multi-stage Nd evolution.Finally, a comparison with the contemporaneous magmatism in the adjacent Mesoproterozoic Karagwe-Ankole Belt indicates similar geochemistry of subordinate mafic dykes, yet reveals distinct differences with the 1370±40Ma Lake Victoria Dyke Swarm and the 1403±14Ma–1374±14Ma layered (ultra)mafic intrusives of the Kabanga-Musongati lineament. We propose that this corresponds to a regional transition toward a collisional setting, related to amalgamation of the proto-Congo Craton. In the Karagwe-Ankole Belt, convergence is superimposed on asthenospheric upwelling, as evidenced by the Lake Victoria Dyke Swarm. Magmatism along the Kabanga-Musongati lineament initiates at 1403±14Ma, and these layered (ultra)mafic intrusives, and is here attributed to preferential upwelling along a transcrustal fault zone. Subordinate mafic to intermediate units interlayer the sediments and could relate to mantle delamination or (back) arc magmatism. Subsequent collision then accounts for the strongly peraluminous granites in the western part of this belt. At c. 1000Ma, strongly peraluminous granites intruded in both belts. Together with similar deformation styles in both adjacent belts, this suggests a regional convergent setting between the Congo Craton and the Tanzania-Bangweulu Block, possibly related to the pan-Rodinian Orogeny.