A rift model for the sedimentary diamond deposits of Brazil

Although several hundred kimberlites have been discovered in the past 20 years in Brazil, the Brazilian diamond production has been derived almost entirely from sedimentary deposits: diamond-bearing conglomerates of different ages and recent alluvials or terraces. Conglomerates as a source of diamond production may not be very significant in economic terms but are extremely important in geological terms because they are the main, known sources of the recent alluvial deposits of Brazil. In this sense, Brazil shows a sharp contrast with other large-scale producing countries, such as South Africa, Botswana, Russia and Australia, where the production comes largely from kimberlites. It has similarities with Angola and India. In the former, the Calonda conglomerates and, in the latter, the Banganapalli conglomerates are as important as sources as are the kimberlites, either because diamonds are recovered from them directly or because they feed the alluvials. But Brazil differs from these countries regarding the age of the oldest diamond finds. While in Angola they are restricted to the Cretaceous and in India to the Upper Proterozoic, in Brazil they spread from the Lower/Middle Proterozoic to the Cretaceous. Brazil is thus a very privileged area to study diamond sedimentary deposits, not only because they are numerous and have economic importance, but also and mainly because they occur in sedimentary sequences of at least four different ages. The recognition of common geological features in diamond sedimentary sequences of different ages led to the conclusion that their recurrence through time reflect fundamental processes which can be put together in a rift evolutionary model. In this study seven different diamondiferous districts in Brazil are briefly described and their rift-related features are emphasised. These features include the position in the sedimentary pile both of the diamond-bearing conglomerates and the volcanic events, as well as the proximal character in relation to growth-faults and the horizontal linear distribution of the conglomerates. The proposed model, stems from a sedimentary reasoning. It links primary diamond sources and related volcanism to rifts, and substantiates the idea that kimberlite/lamproite extrusions predate tectonic paroxysm such that, after erosion, diamonds from those extrusions end up in proximal rift-infill sediments. Here the timing of the extrusion of the primary diamond source rocks is similar to the one determined by White et al. who produced structural evidence to establish their model. Received: 13 January 1997 / Accepted: 22 July 1997