Brosse tektoniek in de Lufilische plooi-breukgordel en zijn voorland. Inzichten in het archief van het spanningsveld in verband met tektonische plaatbewegingen (Katanga, DRC)

摘要

The Lufilian fold-and-thrust belt – also known as the Lufilian Arc – and the Kundulungu foreland in the Katanga region (Democratic Republic of Congo) have attracted the attention of several generations of geologists thanks to the discovery of world class Cu-Co ore deposits. Its geological context, tectonic evolution and metallogenesis are relatively well known, in particular for the Neoproterozoic to early Paleozoic, Katangan sedimentary sequences that have been folded and faulted during the Lufilian orogeny as a result of the interaction between the Kalahari and the Congo-Tanzania cratons in the context of the Pan-African amalgamation of Gondwana. The Lufilian Arc and its foreland, as the adjacent Mesoproterozoic Kibara belt, also show signs of active tectonics, with incipient rift basins, diffuse seismicity and thermal springs. This region is presently undergoing continental extension in the context of the poorly defined, embryonic south-western branch of the East-African Rift System. But the tectonic evolution of the Lufilian Arc and its foreland between the paroxysmal deformation stages of the Lufilian orogeny at ~ 550 Ma and the late Neogene to Quaternary development of the south-western branch of the East African rift system, remains poorly understood, although it can be related to an important Cu-dominated mineral remobilization leading to world-class ore deposits. This long period is essentially characterized by brittle tectonics.This research focuses on this long period of brittle tectonics. The outer part of the Lufilian Arc is suitable for this study thanks to well exposed fault-rocks in open mines spread over the entire arc and foreland. In the arc, observations of mesoscopic faults lead to the definition of three different domains (south-eastern, central and north-western), all with diversely oriented faults with breccia and megabreccia. Megabreccia are widespread at regional scale and can be interpreted as a tectonic unit formed during the major Lufilian faulting event. In the south-eastern part of the outer Lufilian Arc, breccia are bordered by thrust faults parallel to the Lufilian trend. In the central part, however, megabreccia are delimited by subvertical faults with variable orientations, at high angle to the trend of the belt. They show either an oblique thrust or a dip-slip reactivation of bedding planes. In the north-western domain, thrust sheets are dominant features. Before this work, the curved architecture of the belt, which reflects the lateral zonation from SE to NW, has been interpreted by some as sourced from a crustal indentation at the last stage of the Lufilian orogeny. We sampled and analysed brittle structures in the Lufilian Arc and its foreland. Our work confirmed the unequal distribution of fault characteristics in the entire outer belt. At each site, dissimilar outcrops provided fault-slip data obeying to a consistent cross-cutting relationship that allowed a chronologic separation. This was used as a criterion for manual separation of large data sets into homogeneous subsets. When computed, coeval sets gave comparable stress tensors. In this work, we reinterpreted the dissimilar geometrical characteristics of the diversely oriented mesoscopic faults that were neglected in previous tectonic models (D1-D3). The presence of large tectonic basins in the arc, its foreland and the Kibara basement that can be correlated with seismotectonic features such as earthquake epicentre and thermal springs, is also taken into account. The modern techniques of fault-kinematic analysis and tectonic stress inversion have been used to resolve these questions. We present new fault-kinematic field observations and paleostress results computed from a database of 1889 fault-slip data at 22 sites by interactive stress tensor inversion and data subset separation. They have been assembled and correlated into 8 major brittle events, their relative succession established primarily from field-based criteria and interpreted in function of the regional tectonic context. The first brittle structures observed were formed during the Lufilian compressional climax, after the transition from ductile to brittle deformation (stage 1). They have been re-oriented during the orogenic bending that led to the arcuate shape of the belt (stage 2). Unfolding the stress directions allowed to reconstruct a well-defined N-S to NNE-SSW direction of compression, consistent with the stress directions recorded outside the belt. Constrictional deformation occurred in the central part of the arc, probably during orogenic bending. After the bending, the Lufilian Arc was affected by a NE-SW transpression of regional significance (stage 3), inducing strike-slip reactivation, dominantly sinistral in the Lufilian Arc and dextral in the Kundelungu foreland. The next two stages were recorded only in the Lufilian Arc. Late-orogenic extension was induced by a σ1–σ3 stress-axis permutation in a more transtensional regime (stage 4). Arc-parallel extension (stage 5) marks the final extensional collapse of the Lufilian orogen. Stages 2 to 5 are accompanied by important Cu-Co deposits due to fluid circulation in the related brittle mesostructures. In early Mesozoic, NW-SE transpressional inversion (stage 6) was induced by far-field stresses generated at the southern active margin of Gondwana. Finally, this region was affected by – still active – rift-related extension, successively in a NE-SW direction (stage 7, Tanganyika trend) and NW-SE direction (stage 8, Moero-Upemba trend).The variation of tectonic stresses trough time reflects several first-order geodynamic events that affected a much wider region than the one investigated. The brittle data illustrate several stages of a first geodynamic event related to the Lufilian orogeny, since the onset of the brittle realm, at about 550 Ma ago: from the paroxysm of orogenic compression (stage 1), oroclinal bending (stage 2) to orogenic collapse (stage 5). The oroclinal bending, resulting in the Lufilian Arc, is considered to be constrained by the Kibara belt to the northwest and the Bangweulu block to the east. The related constrictional deformation also lead to accentuated salt tectonics and the formation of tectonic megabreccia prior to the Lufilian transpressional inversion (stage 3). A second geodynamic event (stage 6) was recorded as a transpressional inversion, which is interpreted as a far-field effect of the Gondwanide orogeny – as recorded in the Cape fold belt (South Africa) – during the early Mesozoic. The effects of the widely recognised late Santonian and late Maastrichtian regional inversion events have not been found. After the early Mesozoic inversion, the Lufilian region has been affected by extensional tectonics related to the break-up of Gondwana. It is not clear when the extensional conditions started, but two major directions of extension apparently succeeded in time (stages 7 & 8), with the last (stage 8) one fitting the current extension directions deduced from earthquake focal mechanisms. Fault-kinematic analysis of brittle structures and the reconstruction of related stress states allowed contributing to an overall geodynamic model of the Lufilian Arc and its foreland. The D1-D3 Lufilian phases are redefined. The new model suggests that only two orogenic phases – the D1 (Kolwezian) and D2 (Monwezian) phases – are part of the Lufilian orogeny. The D3 (Chilatembo) phase, recorded in orogen-orthogonal structures, is interpreted post-orogenic, related to the previously undocumented far-field stress induced by the Gondwanide collisional transpressional inversion (stage 6). D1 occurred under N-S compression (stage 1) between North and South-Gondwana, Congo-Tanzania and Kalahari cratons respectively. D2 occurred under an NE-SW transpressional inversion (stage 3) during the closure of the Mozambique Ocean caused by the collision between East- and West-Gondwana. Finally, by working on sites with several sub-sites where a large data set can be gathered and which are characterised by a polyphase brittle deformation history, we advanced the methodology of fault-kinematic analysis and tectonic stress inversion by adapting the Win-Tensor program to the context of fold-and-thrust belts. With our experience of the Lufilian Arc, some methodological add-ons such as the interactive data sorting have been developed.