T33F-01
A Type Sequence Across a Magma-Poor Ocean Continent Transition: the Example of the Alpine Tethys
The ophiolites from the Alpine Tethys are incompatible with the definition of the classical 3-layered Penrose ophiolite sequence, but they also show features that are inconsistent with ultraslow-spreading ridge sequences or transform settings. The existence of pre-rift contacts between subcontinental mantle and continental crust, the association of top-basement detachment faults with continent derived blocks (extensional allochthons) and tectono-sedimentary breccias overlying subcontinental mantle, and a post-rift sedimentary evolution identical to that of the adjacent distal margin enable to characterize some of the Alpine Tethys ophiolites as remnants of a former Ocean-Continent Transition (OCT). Therefore, we propose that most of the Alpine Tethys ophiolites represent remnants of an ancient Magma-Poor Ocean Continent Transition, referred to as a MP-OCT sequence. A type sequence can be reconstructed based on observations made in the Platta, Tasna and Chenaillet ophiolite units, the former two representing the OCT of the ancient Adriatic and European/Briançonnais conjugate rifted margins, the latter representing a more developed "oceanic" domain. All three units escaped Alpine subduction and preserve pre-Alpine contacts between exhumed basement and a volcano-sedimentary cover sequence. These units preserve the structural, magmatic, hydrothermal and sedimentary record of continental break-up and early seafloor spreading. The observations compare well with those made along the magma-poor Iberia-Newfoundland rifted margins, which are the only example in an OCT where drill holes penetrated into basement. At present, magma-poor rifted margins form up to 50 percent of all rifted margins world-wide. We argue that MP-OCT sequences are more common in the geological record but were, in part mistaken as either Mid Ocean Ridge or tectonically dismembered Penrose-type ophiolite sections.
T33F-02 INVITED
Assessing the conditions of continental breakup at magma-poor rifted margins: what can we learn from slow-spreading mid-ocean ridges ?
In classical plate tectonic models, the transition from rifting to seafloor spreading was assumed to be abrupt in time and space. This enabled to define a precise age for continental breakup and to map an ocean- continent boundary that not only separated continental from oceanic crusts but also two communities of researchers working with different concepts and models. The discovery of mantle exhumation at both ultraslow spreading ridges and magma-poor ocean continent transitions canceled this simple picture. This discovery not only scrutinizes paradigms, concepts and terms previously used to describe the limits between oceanic and continental crusts, but also ask the scientific community to cross the borders. This is the approach we adopt in this presentation, based on a review of observations and concepts, concerning slow and ultraslow ridges. This leads us to emphasize the role of the thermal regime, as a key parameter for tectonic, magmatic and hydrothermal processes associated with both mid-ocean ridges and continental breakup. We propose that the thermal regime of magma-poor OCTs evolves in the following principal phases : the onset of decompression mantle melting during major crustal thinning, followed by localized exhumation of thinned crust and mantle along detachment faults and installation of a localized and « symmetric (on a large scale)» ridge-type thermal regime (active heat balance, no inheritance), which we argue is the most practical definition for continental break-up, and, possibly, the onset of ridge-type, focused mantle upwelling. We propose this poly-phased evolution as a testable frame for our future research, which aims to combine our understanding of rifting and seafloor spreading, in order to better quantify the processes that control continental breakup.
T33F-03 INVITED
Role of the mantle exhumation channel in the formation of ultramafic seafloor
Mantle peridotites from ocean-continent transition zones (OCT's) illustrate the distribution of the scale of upper mantle heterogeneity in extensional systems that evolve from rifting to (ultra-) slow seafloor spreading. We summarize research on mantle processes of the conjugate Iberia-Newfoundland rift and from the Alps that show that the basement of OCT's consists of 3 mantle domains. Thermally undisturbed, cold subcontinental mantle formed the ocean floor next to thinned continental crust. This 'subcontinental domain' is separated by ductile shear zones from an infiltrated (e.g. hot) domain dominated by plagioclase peridotite. The footwall of these mantle shear zones display complex refertilization processes and high-temperature deformation. These rocks are highly heterogeneous and are juxtaposed with depleted lherzolites and dunites (extraction domain). Upwelling of partial melts that enter the conductive lithospheric mantle inevitably leads to freezing of the melt and to the formation of a chemical and rheological barrier, which we term mantle exhumation channel. We will discuss an example that displays km-scale refertilization with active deformation (and melt focusing?) on top, and the formation of a domain that represents focused melt extraction at the bottom. We show that an actively deforming refertilization front in Alpine plagioclase peridotites and in the Iberia Newfoundland rift moved ahead of a melting front. Melt lubricated shear zones (or melt bands) focus melt flow. Continuous uplift leads to crystallization and deformation will prevail in the subsolidus state. Final exposure of infiltrated peridotites on the seafloor is accommodated by faults in which hydrous phases crystallize (chlorite, serpentine, talc). We combine petrologic data and numerical models to illustrate that these processes play a key role in the rejuvenation and erosion of the lithospheric mantle and ultimately form ultramafic seafloor in ocean-continent transitions and ultra-slow spreading ridges. In such settings preservation of ancient mantle blobs is the rule rather than the exception and a key to understand mantle heterogeneity.
T33F-04
Locating hyperextended passive margins based on plate reconstructions and limits of oceanic crust derived from potential fields data.
Recent advances in understanding of passive margins have emphasized that there is a spectrum of margin styles, ranging from volcanic to hyperextended. All extensional margins will eventually develop sea floor spreading if continental separation continues long enough; the differing margin styles reflect local response of the lithosphere and asthenosphere to ongoing extension. Hyperextended margins can be viewed as extensional systems where continental separation has progressed to a point where there is no more continental crust left, but the asthenosphere has not reached the melting conditions necessary for creation of oceanic crust. The result is that the lithosphere starts to delaminate and mantle is exhumed. The trend then is to form hyperextended margins with exhumation where there is a large amount of continental separation before sea floor spreading begins. One documented area is the Iberia-Newfoundland system, where plate reconstructions suggest that separation between Iberia and Newfoundland was more than 500 km before onset of sea floor spreading. In the South Atlantic, seismic data suggests that mantle exhumation did occur in some areas of the salt basins. Plate reconstructions here suggest more than 300 km of movement between South America and Africa before sea floor spreading was able to start. To try and predict other passive margins where exhumed mantle may underlie significant portions of the margins we have compared extension amounts calculated from plate reconstructions to widths of the margins. This requires mapping of the limits of oceanic crust in areas with inadequate seismic or difficult seismic imaging of deep crustal structure. A regional tool for mapping this boundary is the Bouguer gravity anomaly, with the horizontal gradient of the Bouguer anomaly being a refinement of the method. Studies of the Gulf of Mexico, Gulf of Aden and the Australia – Antarctica conjugate margins are presented here. These studies show that it is possible to predict areas of mantle exhumation based on plate reconstructions and mapping of the limits of oceanic crust.
T33F-05
Lithosphere-Mantle Interactions at Continental Rifts
Continental rifts are sites of lithosphere thinning, with sometimes quite significant topography of the lithosphere-asthenosphere boundary. The thinning lithosphere results in mantle upwelling beneath the rift, even in the absence of a mantle plume. The topography of the lithosphere-asthenosphere boundary creates secondary convection in the rift, and, at the sides of the rift, instabilities of the base of the lithosphere that sink into the mantle. These processes are expected to influence the further development of the rift including its margins, partial melting, and mantle flow. Lithosphere-mantle interactions during continental rifting are studied with an upper mantle flow model that includes a high-viscosity lithosphere. Extension of continental lithosphere induces flow in the upper mantle and with a series of experiments the characteristics of this flow are investigated. First, the basic characteristics of the flow (lateral and depth extent, etc) are explored. The upwelling mantle beneath the rift brings warmer mantle material to shallow depths and this creates a low seismic wave velocity perturbation. Our synthetic seismic wave velocity models are compared with seismic tomography from several rift zones. Shallow upper mantle low seismic wave perturbations that are often observed beneath continental rifts can be explained by passive mantle upwelling beneath the rift. The models further predict cases in which the lithosphere at the rift margins are unstable, and base lithospheric material sinks into the upper mantle. Such convective instabilities are visible as high velocity zones in seismic tomography, and they have been found in the Rio Grande rift.
T33F-06
Gakkel Ridge: A window to ancient asthenosphere
We are accustomed to thinking of the ambient mantle as being a well-stirred reservoir, which contains at most regions of stored subducted slabs and "plums" containing lithophile trace element enrichments. What is forgotten in all of this is that the main process of formation of heterogeneities is a negative one – generating 10x more depleted mantle at any given moment than it does oceanic crust. Because the volume of lithosphere subducted over Earth history is so large, it has always been assumed that the process of subduction and convective mixing re-homogenizes the depleted and enriched reservoirs about as fast as it produces them. What if it doesn't? Our primary means of studying mantle heterogeneity however is basalts. Direct study of the mantle entails observations on xenoliths, ophiolites and orogenic lherzolites, and abyssal peridotites. The latter have the inherent problems of being melting residues, associated with fracture zones, are highly serpentinized and rare. The arctic ridge system gives us a unique perspective on the mantle, and samples we have recovered there are relatively free from these problems. Due to the slow spreading rate, which apparently severely limits the melt productivity, the thickest crust in the Arctic ridge system is approximately "normal". The most common crust is about half thickness and there are large expanses with no crust at all, in the sense of Hess, 1962, exposing mantle peridotite in the floor of extensive rift zones. We have shown Os isotopic evidence for the survival of ancient depletion signatures in Gakkel abyssal peridotites that apparently were not destroyed by subduction, convective stirring or resetting during magma genesis (Liu, et al., 2008). Additionally, preliminary Nd isotopic evidence suggests at least a 400Ma intact prehistory for these samples. Apparently, the low melt productivity on Gakkel Ridge has allowed the Gakkel mantle rocks to escape significant resetting due to melt interaction. This implies a very different picture of the mantle from the one above, one where nearly every part of the mantle has an ancient history prior to its incorporation into the lithosphere, and the distribution of heterogeneities (enriched and depleted) into small scale regions that only approximate the bulk mantle on average (Meibom and Anderson, 2004). Sampling of that mantle by basalts cannot test this hypothesis. Sampling of mantle directly may do so. What this means is that every region of mantle sampled on magma starved ridges may contain traces of a previous history of depletion going far back in geologic time.
T33F-07
Dependence of Mantle Exhumation at Rifted Continental Margins on the Deformation Mode of Breakup Lithosphere Thinning
Mantle exhumation at rifted continental margins requires that rupture of continental crust and the unroofing of mantle occurs before the start of significant melt production. The relative timing of the onset of ocean ridge melt production is sensitive not only to extension rate, mantle temperature and mantle depletion but also the deformation mode of continental lithosphere thinning leading to continental breakup. Two end-member modes of continental lithosphere thinning deformation have been examined: depth-uniform (pure-shear) lithosphere stretching and thinning, and lithosphere thinning by upwelling divergent flow. Horizontal tensile plate forces provide the driving force for the pure-shear deformation. Upwelling divergent flow is assumed to be driven by a combination of horizontal plate boundary forces and thermal and melt buoyancy initiated by pure-shear lithosphere stretching, and predicts a simple transition from pre-breakup lithosphere thinning to sea-floor spreading. For the N. Iberian - N. Newfoundland margins, pure-shear breakup lithosphere thinning model predicts that the onset of melt generation occurs prior to breakup rupture of the continental crust for normal mantle temperature and chemical composition. In contrast the upwelling divergent flow model predicts the onset of melt generation after continental crust rupture leading to ~ 100 km mantle exhumation on each margin. Continental lithosphere thinning leading to continental breakup and sea-floor spreading initiation is most likely achieved by a simultaneous combination of pure-shear and upwelling divergent flow within continental lithosphere and asthenosphere. The relative importance of these deformation modes is dependent on depth, pre-breakup extension rates and mantle temperature. We proposes that beneath 10-15 km depth the dominant mode of continental lithosphere thinning leading to breakup is upwelling divergent flow, while for depths shallower than 10-15 km (corresponding to the cooler upper lithosphere) the dominant thinning mode is pure-shear in the form of brittle faulting.
T33F-08
Sea-floor spreading initiation: constraints from geophysical data of the Thetis Deep, northern Red Sea
A major step in the "Wilson Cycle" is the splitting of a continent and the birth of a new ocean, with the consequent formation of passive plate margins. The transition from a continental to an oceanic rift can be observed today nowhere better than in the Red Sea/Gulf of Aden system. We have carried out during several years a number of expeditions in the axial portion of the Northern Red Sea, in the region where the northernmost nuclei of axial emplacement of oceanic crust can be observed. High resolution multibeam, magnetics, gravity and multichannel seismic reflection surveys from the Thetis Deep revealed rates and modes of initial pulses of sea floor spreading, velocity of S to N axial propagation of the oceanic rift, evolution of initial MORB-type crust and nature of the mantle thermal anomaly that caused the transition from a continental to an oceanic rift. The Thetis deep is made of three en echelon fault-bounded axial basins that are joined together with axial volcanic ridges and a large number of scattered small central volcanoes. The southern basin shows a strong linear magnetic anomaly corresponding to the axial neo-volcanic zone. Two negative symmetric anomalies identified as Matuyama are present in the southernmost part of this basin, suggesting that the emplacement of oceanic crust at this site started roughly 2.5 Ma, with an average half spreading rate of 6 mm/yr. The central sub-basin is also characterized by a strongly magnetic linear neo- volcanic zone that, however, is flanked only by a small, "vanishing" symmetrical negative anomaly suggesting emplacement of oceanic crust not earlier than about 1 Ma. The northern sub-basin does not show a clearly defined linear neo-volcanic zone although it displays a strong central magnetization suggesting initial emplacement of oceanic crust < 0.7 Ma. This pattern implies a south to north time progression of the initial emplacement of oceanic crust within the Thetis system, with a propagation rate of about 20 mm/yr. Gravity data inversions constrained by seismic data reveal that the oceanic crust extends from the axial neo-volcanic ridges toward the master faults of the axial depression with crustal thickness ranging from 4 to 6 km. The increasing thickness of basaltic crust toward the edges of the basin together with higher degree of melting, inferred by the geochemistry of the basaltic glasses, and higher central magnetization of the northernmost and youngest basin suggest a pulse of faster spreading rate at the onset of sea-floor spreading.