Precambrian Research 231 (2013) 218–235 Contents lists available at SciVerse ScienceDirect Precambrian Research journal h om epa ge : www.elsevier.com/locate/precamres A new lithostratigraphic subdivision and geodynamic model for the Pan-African western Saldania Belt, South Africa Hartwig E. Frimmel a,b,∗ , Miguel A.S. Basei c , Vinicius X. Correa c , Ndawedapo Mbangula a,d a Department of Geological Sciences, University of Cape Town, Rondebosch 7701, South Africa b Institute of Geography and Geology, University of Würzburg, Am Hubland, D-97074 Würzburg, Germany c Instituto de Geociências, Universidade de São Paulo, São Paulo, Rua do Lago 562, SP, Brazil d Rosh Pinah Zinc Mine, Rosh Pinah, Namibia a r t i c l e i n f o Article history: Received 11 September 2012 Received in revised form 15 February 2013 Accepted 19 March 2013 Available online xxx Keywords: Saldania Belt Malmesbury Terrane Provenance analyses Pan-African Ediacaran a b s t r a c t A provenance study involving U–Pb and Lu–Hf isotope analyses of detrital zircon grains as well as whole rock geochemical, Rb–Sr, Sm–Nd and Pb isotope analyses of siliciclastic metasedimentary rocks from the principal tectonostratigraphic units of the Pan-African western branch of the Saldania Belt in South Africa revealed age relationships and source domains that are not compatible with current stratigraphic schemes and geodynamic models for the evolution of this belt. The previous subdivision into three ter- ranes cannot be upheld. The sediments in both the former Tygerberg and Swartland terranes seemingly have the same provenance but differ from those in the former Boland Terrane. The latter is considered a para-autochthonous zone on top of basement rocks of the late Mesoproterozoic Namaqua-Natal Belt. On the basis of our new data it is recommended to unify the former Tygerberg and Swartland terranes into a single terrane for which the name Malmesbury Terrane is proposed. The Colenso Fault, previously regarded as a terrane boundary, is reinterpreted as reactivated old structure inherited from the basement within the Malmesbury Terrane, whereas the major terrane-bounding fault is the Piketberg–Wellington Fault, which can be linked most likely with the Schakalsberge Thrust further north in the Gariep Belt. The age of sedimentation is constrained within a narrow time window between 557 and 552 Ma in the Malmesbury Terrane and between 609 and 532 Ma in the Boland Zone. No evidence of older, Cryo- genian sedimentary rocks could be found, except for the volcanic-dominated Bridgetown Formation, which is presumably a tectonic sliver. The Cambrian Klipheuwel Group is undoubtedly younger than, and unrelated with regard to its provenance to, the other Saldanian sediments. All geochemical evidence points to an active continental margin as source for the sediments in the Malmesbury Terrane and Boland Zone. While the ages of the detritus in the former are strongly dominated by Neoproterozoic, especially Cryogenian, ages, those in the Boland Zone are mainly early Tonian and late Mesoproterozoic. The most likely source of the Cryogenian detritus is the Dionisio Cuchilla-Pelotas magmatic arc in southeastern Brazil and Uruguay, which formed within a splinter of Kalahari Craton, i.e. Arachania. A back-arc position relative to that arc is inferred for the depo-basin in which the redefined Malmesbury Group, making up the Neoproterozoic succession of the Malmesbury Terrane, was deposited. Consequently, our new data support the hypothesis that the Neoproterozoic successions of the western Saldania Belt and their northern equivalents in the western Gariep Belt do not reflect a full Wilson Cycle but only the waning stages of an Ediacaran back-arc basin. The main suture between the Kalahari and Rio de la Plata cratons must be located, therefore, to the west of the Dionisio Cuchilla-Pelotas magmatic arc, probably along the Major Gercino-Sierra Balena Lineament within southeastern South America. This implies that the opening of the modern South Atlantic followed to a large extent the axis of the postulated former Ediacaran back-arc basin. © 2013 Elsevier B.V. All rights reserved. ∗ Corresponding author at: Institute of Geography and Geology, University of Würzburg, Am Hubland, D-97074 Würzburg, Germany. Tel.: +49 931 3185420; fax: +49 931 3187344. E-mail address: hartwig.frimmel@uni-wuerzburg.de (H.E. Frimmel). 1. Introduction The Saldania Belt along the southwestern margin of South Africa remains one of the most poorly understood Neoproterozoic/Early Cambrian Pan-African fold belts of Gondwana. This is in spite of a number of relatively recent papers dealing with its structure, sedi- mentology and tectonic setting (e.g. Rozendaal et al., 1999; Belcher 0301-9268/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.precamres.2013.03.014