The Birimian event in the Baoulé Mossi domain (West African Craton) : regional and global context

Abstract: The crystalline basement of the West African Craton (WAC) was established during the Siderian to Orosirian (circa 2.35-1.95 Ga) Birimian event through accretion of extensive tracts of juvenile crust that was tectonically juxtaposed with Archean cratons. The Birimian crust is comprised of volcanic belts and sedimentary basins that have been intruded by multiple generations of intrusive rocks and experienced several tectonothermal events. The basement is mainly exposed in two shields in the northern and southern WAC, respectively, both of which are comprised of a western Archean and an eastern Birimian domain. The southern shield is called Man-Leo and includes the Archean Man and Birimian Baoulé Mossi domains. The aim of this thesis has been to create a preliminary regional-global geodynamic model for the Birimian event in the Baoulé Mossi domain using mainly available literature data but also including some new data from Ghana in the SE Baoulé Mossi domain. Based on this compilation, the geodynamic evolution of the Baoulé Mossi domain is divided into four phases; the Eoeburnean (>2.13 Ga), Eburnean I (2.13-2.10 Ga), Eburnean II (2.10-2.07 Ga) and Eburnean III (<2.07 Ga). The Eoeburnean phase likely began around 2.4-2.3 Ga and is characterized by the accretion of juvenile crust formed in island arcs. A rise in magmatic zircon ages after circa 2.25 Ga may be related to an increase in felsic magmatism as a result of crustal thickening and maturation but also increased preservation of accreted island arcs as a decrease in the number of active subduction zones may have reduced the rate of crustal recycling. Intrusive rocks emplaced during this phase were dominantly sodic granitoids but granites and monzogranites also occur. Tectonothermal and magmatic activity indicate that the Birimian crust in the Baoulé Mossi domain experienced both compression and extension during the Eoeburnean phase. By the end of this phase, an eastward dipping subduction zone had been established along the western margin of the Birimian crust in the Baoulé Mossi domain. Several sedimentary basins in central and SE Baoulé Mossi were established during the Eburnean I phase. The opening of the sedimentary basins may have taken place during regional NE-SW dextral shearing leading to block rotation and development of N-S sinistral shear zones. This also coincided with the collision between the Archean crust of the Man domain and the Birimian crust in the SW Baoulé Mossi domain. The collision also affected the extension of the Baoulé Mossi domain in the Guyana shield of the Amazon Craton. The Eburnean II phase is the most complex. Westward-directed slab rollback in NW Baoulé Mossi led to extension within the overriding Birimian crust. This led to the emplacement of high-K intrusive rocks and explosive extrusive magmatism in NW Baoulé Mossi, in what may constitute a siliceous large igneous province. Extension also led to the opening of younger sedimentary basins in central Baoulé Mossi, possibly along NE-SW oriented shear zones established during the Eburnean I phase. Ongoing collision between the Man domain and the Baoulé Mossi domain led to crustal thickening and associated high-P granulite facies metamorphism in the SE Man domain, possibly around 2.10-2.09 Ga. Granulite facies metamorphism is also recorded in the Archean Amapá block in the E Guyana shield at this time. The Amapá block is separated from the Man domain by a wide belt of low-grade Birimian crust. Simultaneous granulite facies metamorphism in both these areas may be explained by lower crustal detachment in hot Birimian crust. This may allow the upper crust to be displaced without significant thickening until it reaches cooler and more rigid crust were thrust belts are developed. Crustal thickening was followed by a switch to post-collisional sinistral transpression coupled with emplacement of extensive leucogranites between 2095-2080 Ma as the sedimentary basins in central Baoulé Mossi were closed. In contrast to other parts of the Baoulé Mossi domain, the NE part did not experience any significant magmatic activity during this phase, but may have been affected by tectonothermal activity. The Baoulé Mossi domain experienced post-collisional extension during the Eburnean III phase that coincided with the formation of the Bakhuis UHT-granulite belt in the Guyana shield and decompression melting in the Archean crust of SE Man domain and the Amapá block. Limited intrusive and extrusive alkalic post-collisional magmatism was present within the Birmian crust, which cooled and stabilized between 2.0-1.9 Ga. Limited reactivation of the Birimian crust during this period may have taken place in response to far-field events. On a global scale, the Birimian event led to the assembly of a continent — here referred to as Atlantica-Midgardia-Ur — that incorporated continental blocks now present in Africa, South America, India, East Antarctica, Western Australia and Eastern Europe, India, Antarctica and Western Australia. The assembly of Atlantica-Midgardia-Ur coincided with rifting and breakup among crustal blocks now present in North America, northern Europe, North and south-central Australia, East Antarctica and northern Asia. These blocks were subsequently assembled along 2.0-1.7 Ga accretionary belts culminating with the formation of the supercontinent Columbia around 1.8-1.7 Ga, which also included Atlantica-Midgardia-Ur. The assembly of Atlantica-Midgardia-Ur has many similarities with the assembly of Gondwana in the Neoproterozoic regarding the timing and duration as well as spatial distribution of tectonothermal and magmatic activity. In addition, many of the crustal blocks which formed part of Gondwana also formed part of Atlantica-Midgardia-Ur. Likewise, the behavior of the crustal blocks in North America, northern Europe, North and south-central Australia, East Antarctica and northern Asia during the assembly of Columbia is equivalent to the behavior of these blocks during the late Paleozoic (0.3 Ga) assembly of the supercontinent Pangea. The assembly of both Atlantica-Midgardia-Ur and Gondwana coincided with distinct positive excursions in 87Sr/86Sr and δ13C in marine carbonates. The peaks of δ13C excursions coincide with accretionary orogenic activity during both the Birimian event and the assembly of Gondwana. Meanwhile, peaks in 87Sr/86Sr during both cycles coincide with collisional orogenic activity related to the assembly of Atlantica-Midgardia-Ur and Gondwana, respectively. The similarities between the assembly of Columbia and Pangea indicate that they represent two iterations of a particular type of supercontinent, here called Pangea-type. The similarities between the tectonic events and the excursions in 87Sr/86Sr and δ13C indicate that the global tectonic evolution during the Paleo- and Neoproterozoic was fundamentally the same. A supercontinent (Kenorland) equivalent to Rodinia should therefore have existed during the Neoarchean-Paleoproterozoic and broken up in a similar manner to Rodinia. For this reason, Kenorland and Rodinia, as well as the next supercontinent Amasia, can therefore be assumed to represent three iterations of another type of supercontinent, here called Rodinia-type. Supercontinent cycles thus record the transition from either a Rodinia- to Pangea-type supercontinent, or vice versa. The Rodinia- to Pangea-type supercontinent cycles coincide with periods during which the oxygen concentration in the atmosphere was significantly increased. Rodinia- to Pangea-type supercontinent cycles therefore appear to be particularly important for the evolution of the atmosphere, biosphere and hydrosphere. If Kenorland was the first supercontinent, then a “true” supercontinent cycle corresponds to the breakup of one Rodinia-type supercontinent and the subsequent assembly of the next Rodinia-type iteration. In this context, Pangea-type supercontinents are only transient stages when enough crust is aggregated to form a supercontinent. The breakup of Kenorland should have mirrored the “inside-out” breakup of Rodinia in the Neoproterozoic and the ongoing assembly of Amasia. This allows for a reverse schematic reconstruction of the continental blocks of Rodinia as they were positioned in Kenorland. The consistent behavior of most continental blocks since the breakup of Kenorland suggests that they may be divided into three “continental cells”. Each cell is characterized by a particular behavior during a Rodinia- to Rodinia-type supercontinent cycle. Transfer of continental blocks between cells may take place during the breakup of a Rodinia-type supercontinent. Transfer seemingly occur in a dynamic fashion in which a given cell “loses” a block to one cell but at the same time “gains” a block from the other cell. As such, the continental blocks are rotated between the cells even as the size of the cells remains unchanged. Although there are differences between successive Rodinia- to Rodinia-type supercontinent cycles — as shown by the apparent absence of an Atlantic-type ocean during the Kenorland-Rodinia cycle — they are still controlled by the same fundamental cyclicity, which was established during the formation and subsequent breakup of Kenorland.