Why Do Continents Break-Up Parallel to Ancient Orogenic Belts? A
Total Page:16
File Type:pdf, Size:1020Kb
Why do continents break-up parallel to ancient orogenic belts? A. Vauchez*, G. Barruol and A. Tommasi? Lahoraroire de Tectonophysique, C'niverritC de Montpellier II & CNRS, F-35095 44ontpelher C~dexOF, Francc ABSTRACT The frequently observed parallelism between rifts and the pre- assembly, a pervasive deformation of the lithosphere induces a existing orogenic fabric of continents suggests that the inherited lattice-preferred orientation of olivine in mantle rocks. Later on, tectonic fabric of the lithosphere influences the rupture of this crystallographic fabric is 'frozen-in' and represents the main continents. We propose that the existence of a pervasive fabric in source of shear wave splitting. This olivine fabric may entail a the lithospheric mantle induces an anisotropic strength in the mechanical anisotropy in the lithospheric mantle. During lithosphere, that guides the propagation of continental rifts. subsequent tectonic events, especially during rifting, Subcrustal mantle mechanical anisotropy is supported by (i) the mechanical anisotropy may control the tectonic behaviour of anisotropicstrength of olivine, (ii) an ubiquitous tectonic fabric in the lithosphere. exposed mantle rocks, and (iii) measurements of seismic and electrical anisotropy. During major episodes of continent Terra Nova, 9, 62-66, 1997 Mozambique belt and terminates From Florida to the Barents Sea, the Rifting parallel to orogenic belts northward into the Afar. The western North Atlantic ocean opening occurred Ocean-opening through rifting and branch follows the Ubendian belt that parallel to the Hercynian and Caledo- continent break-up is frequently related curves from W-SE to NE-SW. Theu- nian orogens (Fig. 2). In castern North to the occurrence of hotspots. There IS, nissen et a/. (1996) showed that: (i) the America, parallelism between the Ap- however, a discrepancy between hot- Ubendian belt. south and west of the palachians. Triassic rifts. the continent spots acting as pinpoint sources of heat Tanzania craton, is a wide and steep margin, and the main magnetic anoma- and the linear extent of rifts over thou- shear belt resulting from Palaeo- and lies that mark the Atlantic opening sands of kilometres. Moreover rifts tend Neoproterozoic tectonic events, (ii) the indicates that the initial rift propagated to parallel pre-existing orogenic fab- west Tanzania rift developed during along the internal domain of the belt. rics, e.g. North Atlantic rift (Wilson, Phanerozoic times through multiphase South of the West .4frican craton, the 1966), Rio Grande rift (Olsen et a/., reactivation of the Precambrian steeply Atlantic rift formed parallel to Neopro- 1987), North-east China rift (Ma and dipping fabric, and (iii) fault geometry terozoic (southern Senegal, Sierra Leone) Wu, IY87), Bai'kal rift (Delvaux et a/., and kinematics are complex and point and Hercynian belts (northern Senegal, 1995). East African rift (Ring, 1994), toward a SE-NW propagation of the rift Mauritania). North of the craton, rifting West African rift (e.g. Fairhead and in a dextral transtensional strain regime. at = 200 Ma reactivated the Hercynian Binks, 1991), Cape Graben (Burke, belt ofMorocco (plque and Lavillc, 1996). 1976), and Eastern Brazilian rift (Chang South of the North American continent. ef al., 1992). From these observations it formation of oceanic lithosphere in the was suggested that structural inheri- Gulf of Mexico was preceded by the tance influences rift propagation. This dcvclopment of triassic extensional ba- paper propounds that the source of sins elongated parallel to the Ouachitas structural inheritance lies in the fabric mountains (Worral and Snelson, 1989). of the lithospheric mantle developed Gondwana Cretaceous break-up be- during major orogenic episodes. tween South America and Africa is The East African rift and the North and South Atlantic oceans provide spectacular examples of parallelism be- tween rifts and older orogenic belts. The East African rift system (Fig. I), > 2000 km long and still active, formed mostly during the Cenozoic in a shield rtable since the Neoproterozoic (Ring, 1994). The southern part of the system trends almost N-S, parallel to the Mo- zambique belt. Northward, the rift sys- Ocean 20" tem splits into two branches that wrap around the Tanzania craton. The east- ern branch remains parallel to the Correspondence: A. Vauchez, E-mail: vau- Fig. 1 Schematic map (modified from [email protected] fr Ring, 1994) illustrating the parallelism Fig. 2 Schematic map of circum-North talso at Lab. Petrologie et Tectonique. between normal faults in the Fast African Atlantic Caledonian-Hercynian belt\ he- Universite Lyon I & CNRS, France Rift and basement structural fabric forc North Atlantic lifting 62 c 1997 Blackwell Science Ltd Terra Nova, Vol 9, No. 2, 62-66 Continents break-up parallel to orogenic belts A. Vauchez et a/. .. related to the upwelling of the Tristan belt along the N-S-trending Neopro- tends to impede propagation of rifts da Cunha and St. Helena plumes. The terozoic Dahomeyide belt, but the main into cratonic nuclei, and their bound- initial rift system, however, propagated continental breakup occurred through ary with the surrounding lithosphere over more than 3000 km between and wrench faulting between the St. Helena may represent a preferential site for beyond these structures, following the plume and the North Atlantic spread- extensional deformation to localize. Malmesbury-Gariep, Dom Feliciano, ing ridge. This Cretaceous wrench tec- This model is certainly valid for rifts Kaoko, Ribeira and West Congolian tonics (Fairhead and Binks, 1991) formed along old cratonic nuclei as the Neoproterozoic belts (Fig. 3a). The reactivated Neoproterozoic transconti- Bdikal and the Kenya rifts, although it parallelism observed between the South nental E-W-shear zones. remains difficult to evaluate the respec- Atlantic rift system and the continental tive Contribution of rheological hetero- geneity and structural inheritance since fabric is frequently striking since basins Rift propagation and rheologicat follow the curvatures of these orogenic usually orogenic belts wrap around the heterogeneity of the continental cratons and rifts are parallel to both the belts. At the southern tip of Africa, for lithosphere instance, the initial rift bends parallel to craton boundary and the surrounding the Gariep-Malmesbury then the Cape Various interpretations have been pro- belts. Moreover, rheological heteroge- fold belts where it became transten- posed to explain the relationship be- neities fail to explain all cases of rifts sional (Burke, 1976). Along the eastern tween continental lithospheric rifts and developed in an ancient continental coast of Brazil, Chang et al. (1992) the tectonic fabric of continents. Rheo- lithospherc, since in many places rifts showed that the east Brazilian rift is logical calculations (e.g. Dunbar and formed far from any old cratonic nu- remarkably parallel to basement struc- Sawyer, 1989) show that an increase in cleus (e.g. the southern East African tures (Fig. 3b). Even where it propa- crustal thickness can halve the total Rift or the South Atlantic rift). gated in the SBo Francisco craton, the strength of the continental lithosphere south Atlantic rift followed the steep and thus favour localization of rifting. Inherited lithospheric mantle tectonic fabric and large-scale shear In active orogens, crustal thickness is fabric: clues from seismic and zones of the 2Gyr-old Itabuna belt expected to vary and to culminate close electrical anisotropies and Salvador-Esplanada belt of un- to suture zones that represent good known age. Burke (1976) emphasized candidates for the localization of sub- Upper mantle xenoliths and peridotite that grabens oblique to the tectonic sequent rifting. For rifting beginning massifs usually display a well-defined fabric of the continents rapidly aborted. shortly after the end of an orogeny (e.g. crystallographic fabric due to past de- The central Atlantic is more complex North Atlantic in the Appalachians/ formation of the lithosphere (e.g. Mer- since the ocean curves from N-S to Mauritanidesorogen), variation in crustal cier and Nicolas, 1975). Since upper almost E-W and crosscuts various geo- thickness across the belt is likely and mantle rock-forming minerals (espe- tectonic domains. Bonatti (1996) sug- may result in lateral variation of the total cially olivine) are seismically anisotro- gests that a cold equatorial belt impeded strength of the lithosphere and favour pic, their crystallographic preferred the propagation of the southern Atlan- the localization of the rupture process. orientation results in a significant seis- tic rift northward. An aborted branch Local lithospheric weakness asso- mic anisotropy of mantle rocks (e.g. of the rift propagated north of this cold ciated with crustal roots does not satis- Nicolas and Christensen, 1987), and is factorily explain rift propagation within regarded as the main source of teleseis- an old and stable shield (e.g. South mic shear wave splitting (Silver, 1996). Atlantic or East African rifts). The time This supports the notion that shear lag between the last orogeny and rift wave splitting is mostly related to the initiation is of several hundreds of Myr, tectonic fabric of the upper mantle, a duration long enough for large-scale either due to active (asthenospheric) excess crustal thickness to be removed or ‘frozen’ (lithospheric) mantle flow. by surface erosion, isostatic compensa- Short-scale spatial variations of