Journal of Conference Abstracts

Session F02:2A

F02 : 2A/09 : F6

The Magofond 2 Cruise (Part 1): the Central Indian Ridge at 19°S

Jerome Dyment¹

Yves Gallet² &

The Magofond 2 Scientific Party

¹ UMR 6538, IUEM, UBO, 1 place Copernic, 29280 Plouzane, France

² UMR 7577, IPGP, 4 place Jussieu, 75005 Paris, France

In October and November 1998, R/V Marion Dufresne has investigated the Central Indian Ridge (CIR) as part of the Magofond 2 experiment. Together with magnetics and gravity, a complete coverage of good quality bathymetric and imagery data were collected using Thomson Marconi Sonar TSM 5265 multibeam system over two large areas. The first area corresponds to the CIR axis in the vicinity of the Rodrigues Ridge, between 18°30'S and 20°S, and extends off-axis up to 75 km, i.e. more than 3 Ma for a half spreading rate of about 22 km/Ma. It provides a good opportunity to study a presumably hot section of a slow to intermediate spreading center and possibly the interaction of the CIR with the hotspot responsible for the creation of the Rodrigues Ridge. The ridge axis is made of a shallow valley with floor at about 3000 m and crests at about 2200 m. The off-axis pattern of abyssal hills is quite regular, with hills as long as 130 km. Although the valley floor presents a succession of highs and lows at intervals of about 20 km which have previously been interpreted as segment centers and ends, the off-axis bathymetry shows no clear evidence of a stable segmentation, which would have left continuous traces of discontinuities. The only stable segmentation unit represents a more than 130 km part of the ridge and is limited to the north by a oblique discontinuity. On the African flank, the Rodrigues Ridge is continued by a series of progressively smaller linear volcanic ridges up to the near vicinity of the CIR axis. Immediately east of the Rodrigues Ridge, three ridges, tentatively names the Three Magi Ridges, 1500 m high above the surrounding seafloor and 20 to 40 km long, lie on oceanic crust created between 6.5 and 4 Ma (anomalies 3A-3). Further east, the Gasitao Ridge, 700 m high and 50 km long, and possibly a parallel line of volcanic edifices located 30 km to the north, lie on oceanic crust created between 3.5 and 0.6 Ma (anomalies 2A-1). All these ridges trend N70°E to N80°E and have no conjugate on the Indian plate. These observations suggest that hotspot-ridge interaction is still active in this area, as also suggested by the enriched MORB previously collected on the CIR axis. Samples from two dredge hauls on the newly discovered ridges will provide insights on their age and enrichment.

F02 : 2A/10 : F6

The Magofond 2 Cruise (Part 2): Spreading Reorganization on the Past Central Indian Ridge

Jerome Dyment (jerome@univ-brest.fr)¹ &

Yves Gallet (gallet@ipgp.jussieu.fr)² &

The Magofond 2 Scientific Party

¹ CNRS UMR 6538, IUEM, UBO, 1 place Copernic, 29280 Plouzane, France

² CNRS UMR 7577, IPGP, 4 place Jussieu, 75005 Paris, France

In October and November 1998, R/V Marion Dufresne has investigated the Central Indian Ridge (CIR) as part of the Magofond 2 experiment. Together with magnetics and gravity, a complete coverage of good quality bathymetric and imagery data were collected using Thomson Marconi Sonar TSM 5265 multibeam system over two large areas.

The second area corresponds to oceanic crust created at the CIR between 48 and 38 Ma (anomalies 21-18) during a major, 50° clockwise change of spreading direction from N35°E to N85°E. In the old spreading system, the western part of the area was characterized by short compartments and large offsets whereas the eastern part presented a long section of linear ridge. In the western part, the change of spreading direction is marked by a 50 to 250 km ridge jump, implying the extinction of a part of the ridge, the breakup and opening of a new ridge section, and consequently the capture of a fragment of the Indian plate. These events are well documented by our magnetic and bathymetric data and confirmed by similar data collected by Indian colleagues on the conjugate area. The initial spreading on the new ridge section seems focussed on bathymetric highs separated by deep discontinuities, which may result from obliquity and low magma budget. In the eastern part, the change of spreading direction is more progressive. The bathymetric features are dominated by three directions: N125°E (roughly perpendicular to the old spreading direction), N175°E (roughly perpendicular to the new spreading direction), and an intermediate N150°E direction which suggests that the change of direction may have actually occurred in two steps. The rotation of the spreading patterns, with locally traces of ridge propagation, results in the creation of various discontinuities, among which one evolved to a fracture zone. Local spreading patterns are very variable within this area, from regularly spaced sets of abyssal hills to very complex and atypical combination of bathymetric highs and lows, and may be related to the reorganization of the magma supply system. Interpretation of the newly collected and older magnetic anomaly data will constrain the timing of the change of direction.

F02 : 2A/11 : F6

Formation of a Massive Sulfide Deposit on a Sediment-Starved Intermediate Spreading Ridge, Central Indian Ocean

Ute Münch &

Peter Halbach

Fachbereich Geowissenschaften; FR Rohstoff- und Umweltgeologie, Freie Universität Berlin, D - 12249 Berlin, Germany

The spreading rate of mid-oceanic ridges influences the ridge morphology as well as the position, extent and structure of related massive sulfide deposits. For slow to intermediate spreading rates, the topography of ridges is characterized by a more rough and deep valley morphology with very pronounced walls, which contrasts to a rather smooth topography along fast spreading ridges. Compared to axial mineral deposits on fast spreading ridges in the eastern Pacific hydrothermal activity is better focused on slow spreading ridges, which also explains the bigger size of individual deposits in the Atlantic, for example the TAG hydrothermal mound. The MESO zone is an inactive vent area at the intermediate spreading Central Indian Ridge, located on a neovolcanic ridge in a water depth of about 2850 m. The sulfide deposit has a mound like shape and is covered by chimney relicts, sulfidic and basaltic talus and altered pillow fragments. Extinct sulfide chimneys occurred mostly along fissures and cracks, which implies that high temperature hydrothermal activity was favored by tectonic structures in the central part of the MESO zone. Low temperature precipitates like hydrothermal influenced sediments and oxide crusts were sampled or observed in the peripheral area, respectively. Although there are no information about the underlying stockwork mineralizations the morphology as well as the characteristic mineral zonation reminds on the active TAG mound deposit. To build up a genetic model for the MESO zone we assume also a wider area of distal venting as well as a comparable stockwork mineral distribution underneath. Besides geological and mineralogical coincidences between the MESO zone and TAG field the chronology of the evolution of both hydrothermal areas are comparable. Age dating using ²³⁰Th/²³⁴U method show several cycles of high temperature hydrothermal activity started possibly 140 ka ago and terminates 10 ka ago; lower temperature precipitates were formed between 13 and 10 ka BP ago in the MESO zone.

F02 : 2A/12 : F6

Extremely REE Enriched Fe-Ti Melts Injected into Strongly LREE Depleted Oceanic Peridotites, Central Indian Ridge

Eric W. Hellebrand

(ehelle@mpch-mainz.mpg.de),

Jonathan E. Snow &

Albrecht W. Hofmann

Max-Planck-Institut fuer Chemie, Abt. Geochemie, Mainz, Germany

Out of 39 abyssal peridotite samples from 10 locations along the Central Indian Ridge, 8 contain sub-mm to cm (rare) thick magmatic dikelets. Their compositions range from primitive clinopyroxenitic and gabbroic to strongly evolved, oxide-rich ferrogabbroic. Furthermore, 8 samples of the vein-free peridotites contain disseminated plagioclase. We obtained electron and ion microprobe data on clinopyroxenes of these mantle-derived rocks and their magmatic injections. Vein- and plagioclase-free peridotites show very little variation in their primary mineral composition and have LREE depleted patterns ((Ce/Yb)_N = <0.005 - 0.02; (Sm/Yb)_N = 0.1 - 0.5; YbN = 1.5 - 5). These values are similar to cpx compositions of other strongly trace element depleted oceanic peridotites (Johnson et al., 1990).

The evolved dikelets contain Cr-poor cpx (Mg# 0.65-0.75), titanomagnetite, ilmenite and other completely altered phases. The magmatic cpx is extremely REE-enriched and has pronounced negative Sr and Ti anomalies as depicted in the extended REE diagram in Fig. 1. We attribute this to significant olivine, plagioclase, and Fe-Ti oxide fractionation of the melt prior to injection. This melt injection led to modification of the host peridotite in a contact zone less than 2 cm thick. Mantle cpx located at 5 mm from the dikelet is REE-enriched (Fig.1).

The presence of such extremely fractionated melts in mantle rocks raises several questions about magma transport at shallow depths. Did the parent magma of the evolved melt fractionate in the upper mantle or did the injection take place off-axis? Why is there not a more extensive reaction zone? The samples were collected at the Vema Fracture Zone, 80 km away from the present CIR axis. Similar evolved dikelets were reported in peridotites from the Mid-Atlantic Ridge (e.g. Tartarotti et al, 1995).

Johnson, KTM, Dick, HJB & Shimizu, N, In: Vissers & Nicolas (Eds): Mantle and Lower Crust Exposed in J. Geophys. Res, 95, 2661-2678, (1990).

Tartarotti, P, Cannat, M & Mevel, C, Mantle and Lower Crust Exposed in Oceanic Ridges and in Ophiolites, 35-70, (1995).

F02 : 2A/13 : F6

Tectonic Study of the South-West Indian Ridge (58°20'E to 60°10'E) from TOBI Side Scan Sonar Imagery

Olga Gomez (gomez@petrel.cst.cnes.fr)¹,

Anne Briais (Anne.Briais@cnes.fr)¹,

Lindsay Parson (lmp@soc.soton.ac.uk)²

Daniel Sauter (Daniel.Sauter@eost.u-strasbg.fr)³ &

The Fuji Scientific Party

¹ LEGOS-GRGS, 14, Ave E. Belin, 31400 Toulouse, France

² SOC, Southampton, UK

³ EOST, Strasbourg, France

The Southwest Indian Ridge (SWIR) is an ultra-slow spreading center (total opening rate 15 mm/yr) which is oblique to the N-S spreading direction. In October 1997, the MD-108/FUJI cruise studied the SWIR from 58°20'E to 60°10'E and between 63°30' and 65°50'E. These two boxes, respectively west and east of the Melville fracture zone, were chosen because the available bathymetry data reveal highly contrasted accretionary volcanic and tectonic characteristics. Data collected during the cruise were multibeam bathymetry, high resolution, deep-tow side scan sonar (TOBI), magnetics and gravimetry. We present the interpretation of the TOBI images of the western box (58°20'E to 60°10'E) and an analysis of the variations of the tectonics along axis. This box includes three second-order segments, segments 17, 16 and 15 (west to east). The limits of the segments, defined as bathymetric lows, do not really correspond to the non-transform discontinuities (NTD) that can be defined using TOBI images, that is zones where seafloor is relatively deprived of volcanic activity. For instance, the NTD defined from the sonar textures between segments 15 and 16 corresponds to a bathymetric high. Segment 17 mostly displays large faults perpendicular to spreading, which split hummocks and a few seamounts and/or lava flows. Segment 16 is a very short segment characterized by series of small faults which split the hummocks. Those faults are usually perpendicular to the spreading direction in the western extremity and parallel to the regional axial valley direction in the central and eastern parts. In the southern part of the image, we observe a very peculiar feature characterized by north-south striations, which could be a detachment fault. Faults splitting hummocks are particularly numerous in the eastern half of segment 15 (east). The poorly faulted western extremity displays the largest extent of lava flows of the entire box, whether sedimented or not. Faults that are parallel to the regional direction of the axial valley tend to be observed in bathymetric lows, and faults that are perpendicular to the spreading direction tend to cluster near bathymetric highs. This, however, is not always true, specially in segment 16, where the western end, in a bathymetric low, mainly shows spreading-perpendicular faults. The two NTDs (as defined by TOBI imagery) are characterized by faults, eroded "blocks" and wasted material. The NTD between segments 15 and 16 is much shorter than that between segments 16 and 17. All those differences between segments may indicate that they are at different periods of magmatic evolution.

F02 : 2A/14 : F6

The Architecture of Slow-Spread Lower Ocean Crust: Wireline Rock Drilling Around ODP Hole 735B (SW Indian Ridge)

Christopher J. MacLeod (MacLeod@cardiff.ac.uk)¹,

Henry J. B. Dick (hdick@whoi.edu)²,

Simon Allerton (sa@robresint.co.uk)³ &

Paul T. Robinson (robinso@is.dal.ca)⁴

¹ Dept. Earth Sciences, Cardiff University, PO Box 914, Cardiff, United Kingdom

² Woods Hole Oceanographic Inst., Woods Hole, MA02543, U.S.A.

³ Robertson Research International, Llanrhos, Llandudno, United Kingdom

⁴ Centre for Marine Geology, Dalhousie University, Halifax NS, Canada

In order to understand better the tectonic and magmatic architecture of the ocean crust at very slow spreading ridges, we have recently carried out a detailed geological investigation of Atlantis Bank, a shallow platform adjacent to the Atlantis II fracture zone (SW Indian Ridge). The 1500 m-deep ODP Hole 735B is located on the summit of this platform and forms the most complete and well-preserved section of in situ oceanic basement ever sampled. However, the ODP hole forms only a 1-D view of the plutonic foundation of the crust, and virtually nothing else is known of the geology of the surrounding area. The lower crust beneath slow spreading ridges is believed to be very heterogeneous, and it is vital that we constrain the length scale and nature of the variability in both magmatism and tectonism.

In mid 1998, on RRS James Clark Ross, we carried out a high-resolution geological study of a 25km² area surrounding Hole 735B. A seafloor video survey was made using the ROPOS ROV, and the surrounding area (~700 sq km) was dredged in order to provide a broader scale geological context. Detailed sampling of the platform summit was made using the British Geological Survey's wireline hard-rock corers. We used two drills: one that takes cores up to 5 m long in ~2000 m water depths, and the new BRIDGE drill, which takes 1.1 m-long, oriented cores in water depths up to 4000 m. Sampling operations were extremely successful, and 42 sites drilled. Cores included not only gabbros but also dolerites, basalts and serpentinised peridotites. We identify and were able to trace for several km a crust-mantle boundary and several ductile shear zones within the crust. Using the oriented cores we are able to study the geometry of deformation, and examine the directions of magnetisation of samples either side of a magnetic reversal boundary.

Session F02:2B

F02 : 2B/21 : F6

Tectonic Versus Lithological Controls on the Alteration History of Upper to Lower Crustal and Mantle Rocks in a Very Slow Spreading Regime: Atlantis Bank, SW Indian Ocean ­ An O-Isotope Study

Arlene G. Hunter (A.Hunter@earth.leeds.ac.uk)

School of Earth Sciences, University of Leeds, Leeds, West Yorkshire, England. U.K.

It is currently believed that very slow spreading ridge systems are compositionally and rheologically different to fast spreading regimes. Hole 735B represents the longest, continuous core of lower oceanic crust from such a regime, located on the western side of the Atlantis Bank, east of the Atlantis II Fracture Zone, on the SW Indian Ridge. Although this core is used as a universal reference for the lithological structure and evolution of lower oceanic crust, the limitations imposed from a single site prevent a greater understanding of the regional magmatic evolution and alteration history of how oceanic crust forms and evolves in a very slow spreading environment. Preliminary results from cruise JR31 (Spring 1998), reveal that the Atlantis Bank is more complex than previously believed, consisting of juxtaposed peridotites, gabbros, dolerites and basalts, dissected by a series of shear zones and faults, parallel to the main spreading direction.

Interaction between the magmatic and tectonic structure of the oceanic crust and upper mantle, will play an important role in governing the style, nature and duration of fluid-rock interaction that occurs on a temporal and spatial basis across the Atlantis Bank, and in very slow spreading regimes in general. Across the Bank, petrological evidence for low to high temperature, and oxidative to non-oxidative alteration in crustal and upper mantle rocks, is apparent. The exact extent, style and location of this alteration can be attributed to a range of controls, including the magmatic (e.g. grain-size, mineralogy, lithology) and/or deformation (e.g. plastic, brittle, intercrystalline, intracrystalline) characteristics of these rocks. Less apparent, is the local to regional tectonic influence on fluid-rock interaction (e.g. the location of large scale faults and shear zones). For example, it is unclear whether the shear zones and faults crossing the platform, act as major fluid-flow conduits, enhancing the extent and duration of fluid-rock interaction in adjacent crustal and mantle sequences, and hence exert a spatial control on alteration, or whether the dominant control is temporal, with the extent of alteration increasing as the crust moves away from the axis.

This study will present a combination of O-isotope and petrological data from across discrete tectonic segments, in order to establish a relationship between different magmatic, tectonic and alteration events within the crust and upper mantle (i.e. depth of alteration vs. current depth, overprinting of alteration stages, etc.). Furthermore, the nature of the alteration fluids in different crustal and mantle sections will be discussed, noting in particular whether the fluid source (e.g. magmatic-deuteric, hydrothermal or metamorphic) varied temporally or spatially, or whether variations in the temperature or duration of specific alteration events occurred. These results will ultimately be combined to form a systematic model for fluid-rock interaction in a very slow spreading regime.

F02 : 2B/22 : F6

Vema Fracture Zone (Central Atlantic): Temporal Variations of Mantle Composition and of Accretion Processes at Ridge Axis

Daniele Brunelli (brunelli@igm.bo.cnr.it)¹,

Anna Cipriani (acipriani@iol.it)¹,

Luisa Ottolini (ottolini@crystal.unipv.it)² &

Enrico Bonatti (bonatti@igm.bo.cnr.it)¹

¹ C.N.R.-IGM, Via P. Gobetti, 101, 40129 Bologna, Italy

² CNR-CSCC, Via Abbiategrasso, 209, 27100 Pavia, Italy

The Vema Fracture Zone in the Central Atlantic offers a unique opportunity for studying how the processes of formation of oceanic lithosphere vary with time. An uplifted, relatively complete and undisturbed section of oceanic upper lithosphere is exposed on the southern wall of the transform valley along a seafloor spreading flow line for a distance of about 300 km corresponding to almost 20 m.y. (if we assume an average spreading half-rate of 15 mm/y). Close spaced sampling (at 6-8 km horizontal intervals) of the basal portion of this exposed lithospheric section, for a stretch of almost 200 km, produced a set of mantle-derived peridotites serpentinized to various degrees. These samples preserve relicts of mantle-equilibrated "primary" phases such as olivine, opx, cpx and spinel and give us the opportunity of studying temporal variations of the composition of the mantle for a ~10 m.y.-long interval of time. The peridotites are mostly protogranular or porphyroclastic harzburgites and lherzolites. In a portion of the lithospheric sliver, corresponding to a 2-m.y. time interval, the ultramafic rocks are strongly deformed with amphibole-bearing mylonites prevailing. Preliminary electron and ion probe data on major and trace element chemistry of the mantle-equilibrated minerals in porphyroclastic peridotites, show significant compositional variations with age. For example, the Cr/Al ratio of spinel and the Al₂O₃ content of opx vary systematically with distance from ridge axis, suggesting both long- (several m.y.) and short- (< 1 m.y.) wavelength systematic variations in the composition and degree of melting of the upwelling mantle. These variations probably relate to systematic temporal changes of the thermal structure of the mantle below the Mid-Atlantic Ridge. Stretches of the lithospheric sliver where amphibole-bearing ultramafic mylonites are dominant may correspond to intervals of prevalent a-magmatic extension at ridge axis.

F02 : 2B/23 : F6

The Interaction between Extensional Segments at the Icelandic Mid-Oceanic Ridge

Valerio Acocella (acocella@uniroma3.it)¹,

Agust Gudmundsson (agust.gudmundsson@geol.uib.no)² &

Renato Funiciello (funiciel@uniroma3.it)¹

¹ Dip. Scienze Geologiche. Università Roma TRE. Largo S.L. Murialdo, 1. Roma, Italy

² Geological Institute, University of Bergen, Allegaten 41, N-5007, Bergen, Norway

The Iceland part of the Mid-Atlantic Ridge is composed of a sequence of extensional echelon segments with a power-law length size distribution. These segments interact at various scales, from metres (between small, mode I cracks) to more than hundred kilometres (between spreading centres). In order to study the interaction of extensional segments at various scales in Iceland, detailed field studies, as well as aerial-photograph studies, were performed. The geometry and kinematics of 90 systems of interacting mode I and II cracks were measured in the rift zone of Iceland. The results show that interacting mode I and II cracks have identical geometric and kinematic features, suggesting an analogous mechanical behaviour. For all these fracture systems, the crack-crack interaction generates overlapping zones, with the adjacent, propagating cracks becoming hook-shaped on approaching each other. There is a scale-independent, constant overlap/overstep ratio of the overlapping zones. The aspect ratios A (A=L/W, where L is the length and W is the width of the overlapping zones) show a normal distribution, with 87% of values being between 2 and 6. The remaining 13%, with aspect ratios larger than 6, are attributable to the initial configuration of the cracks being collinear. Horizontal shear is present in the cracks bordering and occurring inside the overlapping zones; the trend of the component of horizontal shear has a normal distribution, from 0° to 60° (where 0°= no horizontal shear, 90°= total strike-slip component), with 82% of the shear comprised between 20° and 50°. The sense and the trend of the component of shear depend upon the direction of the cracks in the overlap zone with regard to the overall direction of extension. A "local" extension direction, perpendicular to the curved cracks and consistent with the stress trajectories predicted from photoelastic experiments, often superimposes the "regional" extension direction in the overlap zone. The overall geometric and kinematic deformation patterns recognized in the field are consistent with the results from mechanical and analogue models of interacting cracks. At a larger scale, the same geometrical and kinematical pattern is found in South Iceland, where there are overlapping spreading rift-zone segments and spreading centres. The results suggest several geometric and kinematic parameters associated with overlapping zones along mid-oceanic ridges may be scale independent.

F02 : 2B/24 : F6

Architecture of the Knipovich Ridge at 76°50'N

Rolf B. Pedersen (rolf.pedersen@geol.uib.no)¹,

Kathleen Crane (kathyc@qur.nrl.navy.mil)²,

Georgiy Cherkashov

(HYDROTH@g-ocean.spb.su)³,

Yury Bogdanov⁴,

Andrey Gebruk⁴ &

Eric Anderson⁵

¹ Geologisk Institutt, Universitetet i Bergen, N-5007, Norway

² Hunter College, 695 Park Avenue, NY 10021, U.S.

³ VNIIOkeangeologia, 1 Angliyski Prospekt, St Petersburg, 190121 Russian Federation

⁴ Shirshov Institute of Oceanology, 23 Krasikova, Moscow, 117218 Russian Federation

⁵ JLBSmith Institute of Ichthyology, StreetPrivate Bag 1015, Grahamstown, 6140 South Africa

The rift valley of the Knipovich Ridge at 76°50'N was investigated using manned submersibles during the summer of 1998. The Knipovich Ridge, which is the northernmost extension of the Mid-Atlantic Ridge, is characterized by ultra-slow and highly oblique spreading and the lack of any major ridge offsets. One of the goals of this four-nation expedition, were to investigate the nature of the ubiquitous non-transform shallow offsets which separate pairs of apparently en-echelon deeps along the rise. Five dives to around 3500 meters were carried out during three days at the site and these dives provided new information on the architecture of the rift. The preliminary dive results show that the bathymetric high that separate the two deeps is covered by pillow basalts dusted by a thin layer of sediment. The transition from the high to the axial deep to the south was marked by recent tectonic activity along faults trending in three directions: 000-010°, 030-040°, and 120-130°. Along the northern wall of the investigated deep, large piles of brecciated lava, fresh sheet flows, drain-back craters were discovered. Most of the lava flows are covered by varying amounts of sediments. The thickness of the sediment cover ranges from a very thin layer, to a thick layer that cover most of the volcanic features. Lava flows that had been extruded onto older, partly sediment-covered flows were observed several places. Intermittent volcanic activity appears therefore to have taken place along this ridge for a considerable period of time. Further to the south, the in active basin floor was covered by a thick layer of sediment. The volcanic sequence that underlies this basin was observed along fault scarps associated with a marginal high that defines the southeastern margin of the basin. The constructional volcanic ridge, the in active sedimented basin and the marginal high seem all to be controlled by structures running NE-SW. This orientation is oblique to the rift valley, which runs north-south, but it is about perpendicular to the spreading direction. The orientation of the sheet flows suggest that the lavas erupted out of the NE-NW trending faults/fissures near their intersection with the NW-SE trending eastern rift valley wall. A suite of samples were collected from the volcanic ridge and from fault scarps associated with the marginal high. The geochemistry and the isotope geochemistry of the sampled volcanic rocks will be reported.

F02 : 2B/25 : F6

Are There Asthenospheric Diapirs beneath Propagating Spreading Centers?

Corinne Fleutelot (fleutelo@ccr.jussieu.fr)¹,

Jean-Philippe Eissen (eissen@orstom.fr)² &

Laure Dosso (dosso@ifremer.fr)³

¹ UBO-UMR 6538 and UPMC, Laboratoire de Pétrologie, case 110, 75252 Paris cedex 05, France

² Centre ORTOM de Brest, BP 70, 29280 Plouzané Cedex, France

³ CNRS UMR 6538, c/o Ifremer, BP 70, 29280 Plouzané Cedex, France

Previous studies on propagating rifts (Sinton et al., 1983; Nicolas, 1989; Gente et al., 1995) showed that segmentation along the axis could represent the superficial mark of the activity of a mantle diapir beneath magma reservoirs. Nicolas (1989) describes also different lines of flow in a mantle diapir from fossil ridge segments of the Oman ophiolite. The North-South Propagating Spreading Center (NS-PSC) from the North Fiji Basin (SW Pacific) displays an interesting geochemical evolution, that gives insights to this debate.Basaltic samples from the central part of the NS-PSC present a contrasted geochemical signature from those of the northern tip : 1) a variability of the Sr and Nd isotopic (still within N- MORB field); 2) a variability of the incompatible REE ratios; 3) a constancy of the compatible elements concentration (Ni, Cr,...); 4) a restricted range of MgO variations shifted towards high values. Fractional crystallisation does not describe the evolution of such non cogenetic samples. But partial melting process could explain this relative constancy of compatible elements associated with variable concentrations of incompatible elements.Variations observed along the NS-PSC in the Ca₈/Al₈ ratio which reflect the melting rate (Niu & Batiza, 1991) and in the initial pressure of melting (Po; Albarède 1992) show values much more variable and higher for the central part (respectively 13-20% and 8-12 kb) compared to those of the northern tip (respectively 13-16% and 7-9 kb). Moreover, the mineralogy of the mantle melting beneath the NS-PSC seems homogeneous (in the spinel-lherzolite domain).A relative variability of isotopic ratios and some highly incompatible elements ratios between the central and the northern tip of the propagator, implies a weak heterogeneity of mantle source. Klein & Langmuir (1989) proposed a model of dynamic partial melting for a thermally heterogeneous mantle : its "hottest" part crosses the solidus at high pressure then its "coldest" part with lower partial melting rates. For the NS-PSC, the column of partial melting of a spinel-lherzolite would have some local geochemical heterogeneities and the tapped liquids would record this feature as a slight variability of isotopic and incompatible elements ratios. This suggests the presence of a diapir focused beneath the central part of the NS-PSC. Is the column of partial melting spreading out in the upper asthenosphere from the center of the NS-PSC towards its tip? It might be the case as geochemical compositions are more variable near its center and flowing towards the tip, the melt gradually loses its geochemical heterogeneities and only the most depleted fraction of the melt is extracted and driven to the tip.

Albarède F, J Geophys Res, 97, 10997-11009, (1992).

Gente P, Pockalny RA, Durand C, Deplus C, Maia M, Ceuleneer G, Mével C, Cannat M, Laverne C, Earth Planet Sci Lett, 129, 55-71, (1995).

Klein EM, Langmuir CH, J Geophys Res, 94, 4241-4252, (1989).

Nicolas A, Kluwer Academic, 367 pp, (1989).

Niu YL, Batiza RJ, J Geophys Res, 99, 19719-19740, (1994).

Sinton JM, Wilson DS, Christie DM, Hey RN, Earth Planet Sci Lett, 62, 193-207, (1983).

F02 : 2B/26 : F6

A new Example of Oblique Rifting During the Slowing Phase of Spreading in the West Philippine Basin

Anne Elisabeth Deschamps (deschamp@dstu.univ-montp2.fr) &

Serge Lallemand (lallem@dstu.univ-montp2.fr)

Laboratoire de Geophysique et Tectonique, Universite Montpellier 2 - UMR CNRS - UM2, Place E. Bataillon, case 060, Montpellier, France

The Tertiary history of the West Philippine Basin is still enigmatic. This basin probably formed by back-arc spreading from the "Central Basin Fault" in the Paleocene-Eocene time. During its formation, the basin underwent a major clockwise rotation, between 50 and 25 Ma (Hall et al., 1995). However, detailed data on this basin were lacking. Bathymetric swath mapping and side-scan sonar imagery were carried out in 1996 with R/V L'Atalante, on a single NW-SE transect across the West Philippine Basin, between Taiwan and the Palau-Kyushu Ridge. This survey revealed detailed structures of the fossil spreading axis. We will present new images of this spreading center, which allow us to question the previous models. Indeed, in most places along the axis, no E-W rift valley offset by N-S transform faults is visible, as we could expect. On the contrary, oblique and rhomboidal structures are observed. This strongly suggests oblique rifting during the last spreading stage. Mostly, the spreading axis and the escarpments marking the normal faults bordering the axis, trend N130°E-N135°E, whereas several roughly N-S valleys are observed. By analogy with the structures accompagnying the oblique rifting at the Mohns Ridge (Norwegian Sea) (Dauteuil and Brun, 1993), we suggest that oblique rifting occured in the West Philippine Basin. The last direction of stretching was N-S as shown by the valleys likely marking some transfer zones. But the old inherited N130°E-trending rift valley which was created during an earlier spreading episode, has accomodated the last N-S spreading episode, after the basin underwent a clockwise rotation. Indeed, as the spreading slowed down, it was probably mechanically too much difficult to create a new E-W rift-transform system. This is consistent with a rotation of the plate contemporaneous with the extinction of the spreading near 35 Ma.

Dauteuil O & Brun JP, Nature, 361, 145-148, (1993).

Hall R, Ali JR, Anderson CD and Baker SJ, Tectonophysics, 251, 229-250, (1995).

F02 : 2B/29 : F6

Mechanics of the Oceanic Rifting: From a Diffuse to a Localised Deformation

Olivier Dauteuil (dauteuil@univ-rennes1.fr) &

Florence Nicollin (nicollin@univ-rennes1.fr)

Geosciences Rennes, campus Beaulieu, Rennes, France

The models of oceanic rifting imply a periodic bathymetry whose the wavelength depends on both volcanic activity (intensity and/or periodicity) and on deformation pattern. Bathymetric profiles of the FAMOUS area, (Mid-Atlantic Ridge), were analysed using wavelet processing to constrain this periodicity. The first result is a lack of an symmetry across axis as well at the segment extremities, as at the centre of the segment. This indicates that the rift morphology results from a combination of different processes acting at different scales. This numerical processing made in the 1-21 km wavelength range, shows that the whole ridge bathymetry is dominated by two wavelength sets: short wavelengths around 4 km and long wavelengths around 15 km, located differently along and across the axis. Along ridge axis, the two wavelength sets are present (around 5 and 13 km); whereas, only one wavelength was pointed out off axis (15-18 km). Along the axis, the values of the short wavelengths tend to decrease from centre (5 km) to extremities (3 km) of a segment.

These topographic wavelengths correspond to periodic instabilities induced by a necking of the lithosphere. Previous theoretical analysis permit to deduce the rheological layering of the lithosphere using the shape of the boudinage and the wavelength values. It was determined that: 1) across the axis, the rheological structure of lithosphere evolves from the four layers to a two layers system; 2) along the axis, the rheological structure corresponds to a four layers system with an increase of the whole lithospheric strength and a thickness decrease of the upper plastic layer. These results are consistent with both the strength profiles estimated with the 3D thermal structure of a segment, and its petrologic layering. At the extremities, the presence of a weak zone at shallow depth is attributed to the top of serpentinised peridotites.

A rifting model is proposed taken into account for a deformation partitioning induced by the 3D rheological structure. This model integrates both this morphological analysis and field data showing that the axial floor is affected by numerous fissures and small faults, while the rift shoulders were generated by few large faults. The deformation evolves from a diffuse mode in the axial zone, where the upper brittle layer is thin, to a localised mode on the valley walls where the brittle layer thickness increases. Farther away from the ridge axis, the larger strength of the thicker lithosphere prevents deformation to develop. This evolution is related to the increase of the lithosphere strength and to an evolution of the strength profile away from the rift axis.

F02 : 2B/30 : F6

Axial Magnetic Anomaly Amplitudes at Slow Spreading Centers: Insight from 3D Numerical Modelling

Sebastien Gac (sgac@sdt.univ-brest.fr),

Jerome Dyment (jerome@univ-brest.fr) &

Chantal Tisseau (tisseau@univ-brest.fr)

UMR 6538, IUEM, UBO, 1 place N. Copernic, 29280 Plouzane, France

Many segments of the Mid Atlantic Ridge systematically show axial magnetic anomalies amplitudes about twice higher at segment ends than at segment centers. Several processes have been proposed to explain those variations, including (1) deepening Curie isotherm from segment centers to ends, (2) serpentinisation of outcroping peridotites at segment ends and (3) increasing fractionation and iron content from segment centers to ends. To test these three processes, we have developped a 3D numerical model to simulate magnetic anomalies at a slow spreading center segment and compare the results to the observations. The distribution of magnetization depends on the thermal state and evolution of the segment. We have therefore developped a steady state thermal model of a slow spreading ridge segment considering a permanent thermal intrusion focused on the center of the segment. The resulting thermal state and evolution, the associated petrological structure of the segment, and the magnetic properties of of the oceanic rocks allow the computation of the distribution of magnetization within the segment. Several models have been tested, with different magnetic structure and properties of the oceanic lithosphere. Models accounting for the presence of serpentinised peridotites at segment ends result in synthetic magnetic anomaly amplitudes in general agreement with the observations, whereas models only considering the hypotheses of a deepening Curie isotherm and the variation of fractionnation and iron content yield magnetic anomaly amplitude variations weaker than the observed ones. This modelling suggests that serpentinised peridotites may have a significant contribution in the marine magnetic anomalies at slow spreading centers.

F02 : 2B/31 : F6

Analysis of the Influence of Hot-Spots on the Segmentation of Mid-Ocean Ridges. Comparison with Models of Mantle Convection

Anne Briais (Anne.Briais@cnes.fr) &

Michel Rabinowicz (Michel.Rabinowicz@cnes.fr)

LEGOS-CNRS UMR5566-GRGS, 14, Ave E. Belin, 31400 Toulouse, France

We analyze the axial segmentation of slow-spreading Mid-Atlantic Ridge (MAR) and Central Indian Ridge (CIR) near the Iceland, Azores and Reunion/Rodrigues hotspots. From satellite-derived gravity maps, we observe two main changes in the segmentation and its evolution near hotspots. First, the amplitude of the gravity signal associated with the segmentation is smaller in ridge sections 500-1000 km from hotspots than in colder ridge sections, even though the wavelength of the segmentation does not systematically vary with distance to hotspots. This is especially clear on the Reykjanes Ridge (RR) where the present-day axis is characterized by en échelon axial volcanic ridges in an unsegmented axial valley. For older lithosphere on the flanks of the RR, a segmentation stable over 10-20 m.y. is observed, with segments similar to those of the MAR farther south. The change in the signature of the segmentation occured earlier in the north than in the south. This recent change in the segmentation style may be a consequence of a change in the activity of the hotspot, or a change in the geometry of the relation between ridge and hotspot. Near the Azores, the present-day axial morphology and segmentation changes within 200 to 300 km of the hotspot only. The traces of the discontinuities, however, are difficult to follow off-axis, suggesting that the geometry of the axis changed rapidly in the last 20 m.y. The 18-20°S section of CIR axis closest to the Reunion hotspot is shallow and has a low-relief axial valley, with very low signature of the segmentation in the gravity anomalies. The change from a segmentation with a strong gravity signature to a segmentation with a low signature occured diachronously from 20 Ma in the north to 10 Ma in the south.The second major observation is that for these three hotspots, the ridge sections at distances up to 1500 km show discontinuities migrating away from the hotspot, at very different rates. The RR show very narrow V shapes pointing south. Near the Azores or the Reunion/Rodrigues hotpots the migration rate is much lower.We compare these observations with the results of 3-D numerical models coupling the axial geometry and crustal production with the small-scale convection of mantle in the low-viscosity zone beneath the axial zone. The migration of the segments away from hotspot might be explained by a sweeping of the small-scale convection beneath the ridge axis by the larger-scale mantle flow from the hotspot. The signature of the segmentation and its stability in time are probably controlled by the rheological behavior of the axial lithosphere.

F02 : 2B/32 : F6

Chemical and Isotopic Constraints to Varied Formation Processes for the Brines of Different Deeps in the Red Sea

Marie-Claire Pierret (pierret@illite.u-strasbg.fr)¹,

Delphine Bosch (bosch@dstu.univ-montp2.fr)²,

Gérard Blanc (blanc@geocean.u-bordeaux.fr)³,

Norbert Clauer (nclauer@illite.u-strasbg.fr)¹ &

Pierre Anschutz

(anschutz@geocean.u-bordeaux.fr)³

¹ CGS/EOST UMR 7517, 1, rue Blessig, 67084 Strasbourg, France

² GGP UMR 5567, place E. Bataillon, 34095 Montpellier, France

³ DGO UMR 5805, avenue des Facultés, 33405 Talence, France

About twenty deeps (assimilated to oceanisation cells) filled with hot brines and/or metalliferous sediments, are located along the Red-Sea axis. They represent an interesting natural laboratory allowing the study of the influence and importance of the hydrothermal activity in a recent and slow-spreading oceanic environment. During the REDSED cruise in september 1992, the Suakin, Port-Soudan, Valdivia, Chain, Atlantis II and Nereus deeps were investigated, and the water columns were sampled and studied by bathysonde recording in each of them. The present study summarizes the results of an approach combining major-, trace elemental and isotopic (oxygen, hydrogen, strontium) analyses. Results show important heterogeneities in temperature, salinity, hydrographic structuration and chemistry, each brine having its own characteristics.

The brines of the Atlantis II and Chain deeps are depleted in SO₄ and Mg (respectively 5 and 2 times less concentrated), enriched in metals, Li, Ba, Si, and have low ⁸⁷Sr/⁸⁶Sr ratios (0,70696 ± 0,0002) relative to the above sea water. This indicates inputs of hydrothermal fluids having exchanged with basaltic rocks at temperatures up to 250°C. In fact, it can be shown that the lower brine of Atlantis II deep fills Chain next to it, where no specific hydrothermal activity is detected. The brine of Nereus deep is enriched in the same elements than the lower brine of Atlantis II. It has a lower ⁸⁷Sr/⁸⁶Sr ratio (0,70656 ± 0,0002), is depleted only in SO₄, and yields a basic pH (7,5 instead of pH=5 for Atlantis II lower brine). Fluids supplying this deep have interacted with the oceanic crust at temperatures ranging from 150 to 250°C. These brines, with a basalitc Sr contribution between 8 and 12%, are the only ones that indicate unequivocally a hydrothermal influence.

The Port-Soudan brine yields evidence for a slight basalt interaction at temperatures lower than 100°C. The brine of Valdivia deep results mainly from evaporite dissolution (especially halite) during shallow circulation. The brines of Suakin deep significantly differ from the others by lower temperatures, salinities and Fe concentrations. They have oxygen and hydrogen isotopic signatures comparable to those of old interstitial waters. They seem to have formed from an old sea water which dissolved evaporites without any input.

In summary, formation of the Red-Sea brines involved three major processes: leaching of Miocene evaporite, convective fluid circulation, and interaction with basaltic crust and old sedimentary levels. Differentiate combinations of these processes allow distinction of the brines in the studied deeps. This outlines the fact that hydrothermal activity is not identical along the whole Red-Sea axis; it does not favour a unique formational model for the brines. A striking feature is the lack of relationships between the position of the deeps along the axis and their evolutive maturity.

F02 : 2B/33 : F6

Trace Elements in Serpentine Minerals and the Role of Seawater (14°45' on the Mid Atlantic Ridge: The Logachev Hydrothermal Field)

Beate Orberger (orberger@geol.u-psud.fr)¹,

Nicole Métrich (metrich@drecam.cea.fr)²,

Catherine Mével (mevel@ccr.jussieu.fr)³,

Michelle Mosbah (mosbah@drecam.cea.fr)⁴ &

Yves Fouquet (fouquet@ifremer.fr)

¹ Laboratoire de Géochimie, Université Paris XI, Batiment 504, 91405 ORSAY Cedex, France

² Laboratoire Pierre Süe, CEA-CNRS, CE Saclay, Gif sur Yvette, France

³ Laboratoire de Pétrologie, Université Paris VI, Paris, 75005

⁴ Laboratoire Pierre Süe, CEA-CNRS, CE Saclay, Gif sur Yvette, France

The black smoker Logachev at 14°45'N on the Mid Atlantic Ridge is the first hydrothermal field which was found on the top of completely serpentinized ultramafic rocks outcropping at the seafloor. Trace element studies on serpentine minerals have been carried out in order to study the role of seawater during serpentinization. Nuclear microprobe analyses have been performed on three serpentine minerals: bastite, the pseudomorphoses after pyroxenes, mesh serpentine which are olivine replacements and tiny serpentine veins developing in bastite cleavages or crosscutting them. Trace elements such as Cl, F, S, Cu, Zn, Ca, K, Ni, Cr and Mn were detected from several tenth to several thousands of ppm. The trace element concentrations and the interelement relationship in serpentines vary a) with the serpentine mineral and b) with the geographic location to the black smoker. Chlorine and in part S originated from seawater, whereas Cu, Zn, Ca, K, Ni, Cr and Fe and the major amount of S were mobilized during serpentization and partitioned between the serpentine and the aqueous solution. The physico-chemical conditions during serpentinization vary from high T (>350°C), neutral to alkaline pH and fO₂ close to the iron magnetite boundary far away from the Logatchev site to lower T (<350°C), acidic and oxidizing conditions near the black smoker.

F02 : 2B/34 : F6

Fluid-Sediment Interactions Related to Hydrothermal Circulation in the Eastern Flank of the Juan de Fuca Ridge

Martine D Buatier

(martine.buatier@univ-fcomte.fr)¹,

Monnin Christophe (monnin@lucid.ups-tlse.fr)²,

Delphine Charpentier³,

Anne-Marie Karpoff (amk@illite.u-strasbg.fr)³ &

Gretchen Früh-Green⁴

¹ Université de Franche Comte, 16 Route de Gray, 25660 Besancon, France

² Lab. de Géochimie, 38 rue des 36 Ponts, 31400 Toulouse, France

³ EOST-CGS, CNRS, 1 rue Blessig, 67084 Strasbourg, France

⁴ Sonneggstr. 5, ETH-Z,CH-8092 Zurich, Switzerland

Fluid circulation within the oceanic crust can be traced through the study of water-rock interactions. The preliminary results of ODP Leg 168 (eastern flank of the Juan de Fuca Ridge, Northeast Pacific) have shown that low temperature hydrothermal fluids circulate in a direction perpendicular to the ridge axis. This circulation, which seems to be mostly restricted to the upper part of the basaltic crust, is still active about 100 km from the ridge axis. A variably thick sedimentary cover acts as a hydrological barrier. Fluid discharge to the deep ocean occurs on basement highs where the thickness of the sedimentary cover does not exceed 40 meters. The present study focused on the characterisation of the chemistry and mineralogy of the sediments in a discharge area (Sites 1030/1031). Our observations show that mineralogical reactions related to fluid-sediment interactions occur in sediments located immediately above the basement. This sediment alteration is particularly intense at Site 1031 where vertical fluid advection is occurring. Stable isotope data on carbonates and thermodynamic calculations on the carbonate-pore water equilibrium suggest that the observed reactions (dissolution of biogenic calcite and precipitation of Mg-Fe smectite and Ca-zeolites) can be related to the present-day active fluid advection through the sedimentary column. A comparison between altered and non-altered sediment chemistry suggest an uptake of Mg by the sediment and a release of Ca to the pore waters. These data are consistent with the pore water compositions, i.e. pore waters at the basement-sediment contact are enriched in Ca and depleted in Mg at 1030/1031 compared to other sites drilled during Leg 168. These data suggest that the advecting fluids consist of slightly altered seawater that has reacted with sediments during its advection and which has been chemically modified during dissolution of carbonates and precipitation of smectite and zeolites. This study shows that fluid-sediment interactions should be taken into account in the calculation of the chemical fluxes related to hydrothermal processes in oceanic ridge flanks.

Session F02:3P

F02 : 3P/01 : PO

Kinematic Evolution of the Western Gulf of Aden

Laurence Audin (audin@ipgp.jussieu.fr)¹,

Paul Tapponnier (tappon@ipgp.jussieu.fr)¹,

Philippe Patriat (patriat@ipgp.jussieu.fr)² &

Isabelle Manighetti (manig@ipgp.jussieu.fr)¹

¹ IPGP-Tectonique, 4,place jussieu, 75252 Paris Cedex 05, France

² IPGP-Geophysique, 4,place jussieu, 75252 Paris Cedex 05, France

Detailed study of the marine magnetic anomalies in the western part of the Gulf of Aden (Tadjouraden survey) associated with new paleomagnetic and K/Ar sampling of Tertiary volcanics (Ethiopia) provides new insights to quantitatively analyze the deformation mechanisms related to the propagation of the Aden ridge toward the Afar depression. Reconstructions and Euler poles proposed for the Arabia/Somalia motion, did not succeed in fitting back together the surface of the Ethiopian traps which are associated with the onset of the rifting in the Afar region (20 Ma). Since much of the Afar depression is floored by the 2 Ma Stratoid basalts, evidence for previous deformation processes is only accessible in and around the Aïsha/Ali Sabieh block onland and in the Gulf of Aden in the marine area. Indeed, marine magnetic anomalies could only be retrieved back to anomaly 5 (10 Ma) in the whole area. East of the NE trending Shukra El Sheik fault, continuous seafloor spreading can account for the anomaly series back to 10 Ma. But, on the western side, clear magnetic anomalies are reliable only back to anomaly 2. This sudden loss of the oldest series of anomalies implies that a major boundary lies along the Shukra El Sheik fault. Extension must have occurred north or south of that boundary, and indications for such a deformation process are absent from the studied structures of Yemen. Onland in Ethiopia, 2 to 20 Ma volcanics have recorded such deformation, revealed by morphotectonic studies of regional faults. It provides new constraints on the beginning of rifting and the opening of the southern Afar depression since ~20 Ma in connection with the propagation of the Aden ridge. Taking into account all these features, a kinematic model of propagation and transfer of extension within southern Afar is proposed.

F02 : 3P/02 : PO

The Southeastern Lewis Hills (Bay of Islands Ophiolite): Geology of a Deeply Eroded, Inside-Corner, Ridge-Transform Intersection

Günter Suhr (gsuhr@mpch-mainz.mpg.de)¹ &

Peter A. Cawood (p.cawood@info.curtin.edu.au)²

¹ Max Planck Insitut für Chemie, Postfach 3060, 55020 Mainz, Germany

² Tectonics Special Research Center, School of Applied Geology, Curtin University, GPO BOX U1987, Perth WA6001, Australia

The eastern part of the Lewis Hills (Bay of Islands Ophiolite) is considered a classic locality for studying the accretion of oceanic crust to the inactive part of an oceanic transform fault, i.e an outside corner, ridge-transform intersection setting (Karson and Dewey, 1978). Detailed mapping in the SE Lewis Hills has revealed a number of new features which suggest an inside-corner locality. These are: (1) two near vertical, highly extensional dextral strike slip zones of mainly amphibolite grade; (2) a highly atypical, extremely depleted and low strain harzburgite containing a core of late dunite which underwent brittle extensional deformation and which represents mantle lithosphere; (3) a mylonitic low angle normal fault near the top of the lithospheric mantle which shows early anhydrous and late hydrous fabrics and is overlain by mafic-ultramafic cumulates showing non-mylonitic, high temperature plastic strain; and (4) the presence in the NE Lewis Hills of asthenospheric mantle which suggests a strong gradient in the lithospheric thickness from north to south.

We interpret this association as an oceanic detachment fault ("core complex") which nucleated at the rheologically weakened crust-mantle boundary due to a strongly uplifted mantle lithosphere. The field evidence suggests that the origin of the uplift in the south was by dynamic forces as a result of extensional plate stresses acting on a lithosphere of variable thickness (Phipps Morgan et al., 1987). Vertical ascent was facilitated by sets of transform-related extensional strike slip faults.

In addition, there is strong evidence that the lithospheric mantle in the SE Lewis Hills was originally part of the opposing plate. Its involvement in the spreading history suggests that ridge propagation occurred along the transform boundary. We suggest a model involving an anticlockwise rotation of the plate spreading direction to explain the extensional nature of the transform and the inferred ridge propagation. Aspects of the enormously complex geology in the SE Lewis Hills should serve as good analogues for the exposure of plutonic rocks at ridge-transform intersections, now also a major target of ODP drilling.

Karson, JA & Dewey, JA, Geol. Soc. Am. Bull, 89, 1037-1049, (1978).

Phipps Morgan, J, Parmentier, EM, & Lin, J, J. Geophys. Res, 92, 12823-12836, (1987).

F02 : 3P/03 : PO

First Sulfides Found in Sediments of the Deep Marmara Basin ­ are They Prove of Hydrothermalism in the Sea of Marmara?

Thomas Kuhn (tkrumgeo@zedat.fu-berlin.de)¹,

Grossmann Rico¹,

Halbach Peter (hbrumgeo@zedat.fu-berlin.de)¹,

Kuscu Ismail (environ@crescent.mta.gov.tr)² &

Algan Oya (debien@superonline.com)³

¹ Free Univers. Berlin, Malteserstr. 74-100, D-12249 Berlin, Germany

² MTA Ankara, 06520 Ankara, Turkey

³ University of Istanbul, Müsküle Sok. 1, 34470 Istanbul, Turkey

The Sea of Marmara is a marine basin in northwest Turkey that connects the Black Sea with the Aegean Sea. It is located at the North Anatolian fault. The fault zone splits into several fault strands so that the Sea of Mamara segments into five blocks, three pull-apart basins separated by two push-up structures. South of Sea of Marmara a lot of geothermal sites and hotwater springs have been identified on the mainland. We supposed, that hydrothermal activity continues into the Deep Marmara Basin, one of the pull-apart basins. Thus, the sediments could contain hydrothermal precipitates. In gravity cores collected during the R/V Sismik I cruise LEG 9807 in July 1998 clay sediments with sandy turbiditic layers of some centimeters thickness were sampled. Two of the sand layers were cemented by pyrites. First geochemical element analyses of this layers reveal high concentrations of Fe, Cu, S, As and Sb compared to the surrounding sediments. The high concentrations of Fe, Cu and S are due to the pyrite content, As- and Sb- concentrations could result from hydrothermalism or from ascenting diagenetic solutions. During the meeting we want to present results of mineralogical (XRD & microspcopy) and geochemical (ICP & SEM) investigations of the sulfides which are currently under work.

F02 : 3P/04 : PO

The Zoning of MORB Composition from the Juan-De-Fuca Ridge and the Nature of This Zoning Occurrence

Aleksey S. Zhmodik (zhmodik@uiggm.nsc.ru) &

Viktor N. Sharapov (vik@uiggm.nsc.ru)

Institute of Geology SB RAS, pr.Koptuga, 3, Novosibirsk, 630090, Russia

We used a wealth analytical data (from literary sources and original(Niu and Batiza, 1993; Smith et al., 1994) on the chemical composition and trace elements in basalts of Juan-de-Fuca ridge. Investigated analytical data for all segments of the ridge. Data of statistical processing sugests that the Juan-de-Fuca ridge has zonal structure that similar to the MAR segments zoning. In central its part dominate more magnesia and less titaniferous melts. Also the most primitive lava were obtained here. This situation the most is clearly looked through on distributions of the relations of isotopes a strontium (Eaby and Claque, 1984), iron, magnesium and calcium. Consideration of petrochemical data of basalts from Juan-de-Fuca ridge shows that in its scope are wide-spread two main petrochemical group of MORB: 1) «type NMORB» (dominate); 2) ferrobasalts;

Geochemical factors indicate that amongst specified rocks there is melts, that appearing from different primary mantle «sources», but having cognate trends of differentiation in low-deep chambers. Problem of estimation of primary melts generation depths can be not correctly determined only on the base of petrochemical and geochemical data, in which is not take into account a sequesterring the phases in geterophase systems in the field of decompressional melting in the mantle diapirs or «hot spots» (M^cKenzie and Bickle, 1988; Klein and Langmuir, 1987; Emley and Chadwick, 1994). The data received by us about dynamics of partial melting zones formation while nor permit to evaluate the thermodynamic conditions of different initial melts occurrence under MOR. It is possible to confidently estimate only conditions of occurrences of porphyric phenocrysts or temperature of liquidus. They can correspond to the depth of intermediate chambers. Such estimation we have for lavas MAR. The similar estimations we have received for the basalts of all segments of the Juan-de-Fuca ridge. On mentioned parameter the nature of zoning is similar to described for segments of MAR.

The petrogenetic data permit to make the series of preliminary conclusions, which permit to realize the reasons of heterogeneous development of ore-magmatic systems under axial zones of MOR: 1) in all investigated segments of MOR we see the general feature - availability of two main petrochemical type of magmas, which was melted on different depth from relatively depleted and not depleted substances; 2) trends of fractionation of these smelting in near-surface chambers are similar, that induces the similar tendencies in distributions of petrogenic components in series of allocated petrochemical groups; 3) we have realize, that in low-spreading ridges one tendencies in chemical trends primary smelting, and in fast-spreading - other ones prevail; 4) in smelting from less depleted substance the abnormal contents of water and other volatile are fixed, as well as trace components, which are source of development of magmatic ore mineralisation at differentiation of such magmas in near-surface magmatic chambers.

This work was supported by UIGGM SB RAS VMTK grant 1 1736.

Niu Y & Batiza R, J. Geophys. Res, 98, 7887-7902, (1993).

Smith MC, Perfit MR & Jonasson IR, J. Geophys. Res, 99, 4787-4812, (1994).

Eaby J & Claque DA, J. Geophys. Res, 89, 7883-7890, (1984).

M^cKenzie D & Bickle MJ, J. Petrol, 29, 625-679, (1988).

Klein EM & Langmuir CH, J. Geophys. Res, 92, 8089-8115, (1987).

Embley RW & Chadwick WW, J. Geophys. Res, 99, 4741-4760, (1994).

F02 : 3P/05 : PO

Mineralogy and Sulphur Isotope Geochemistry of the Logatchev Sulphides, Mid Atlantic Ridge 14°45'N

Olivier Rouxel (rouxel@crpg.cnrs-nancy.fr)¹,

Yves Fouquet (fouquet@ifremer.fr)²,

Stuart Boyd (sboyd@crpg.cnrs-nancy.fr)¹ &

Marc Chaussidon (chocho@crpg.cnrs-nancy.fr)¹

¹ CRPG-CNRS, BP 20, 54501 Vandoeuvre, France

² IFREMER, BP 70, 29280 Plouzané, France

Massive sulphides and black smokers from the newly discovered Logatchev hydrothermal field at 14°45'N have been examined to determine their mineralogical, chemical and sulphur isotope characteristics. Hydrothermal sulfide deposits were sampled by the Nautile submersible during the MICROSMOKE (1995) cruise. Sulfide mineralisations at Logatchev contrast strongly with other submarine hydrothermal fields by the high copper content, original chimney zonation and ultrabasic substratum. In situ and conventional techniques were used to determine the sulfur isotopic composition of sulfur-bearing phases. These analysis were performed at the CRPG on a Cameca IMS3f ion microprobe and a VG602D mass spectrometer and are given in ³⁴S notation according to the CDT international standard. Chemical data were obtained by electron microprobe (Cameca SX50) at IFREMER. Chalcopyrite is the dominant phase, followed by anhydrite and a typical weathering assemblage including bornite/digenite. Hematite with minor magnetite is locally enriched in chimney walls. Unusual occurrence of Co-pentlandite was identified and occurs locally as small (less than 10µm) irregular grains in chalcopyrite veinlets. The mineralogy and geochemical composition of black smoker chimneys at the Logatchev hydrothermal field were studied in order to understand the growth mecanism and to caracterise the sulphur isotope variation along a chimney cross section. The mineralogy, Se concentrations and sulphur isotope values of chalcopyrite change gradually in response to changes in fluid temperature and composition and suggest a relatively simple model of chimney growth. The high ³⁴S content of chalcopyrite (³⁴S=5-6‰) corresponds to the high value of Se (2000 ppm) and may be caused by local sulfate reduction by the Fe²⁺ bearing hydrothermal fluid. The low value of ³⁴S (³⁴S=0-1‰) may be due to a decrease of rate of sulphate reduction with a decrease of temperature and/or a increase of pH of vent fluid near chimney environment. ³⁴S values are scattered at a small scale from 0 to 8.2‰ with mean value for chalcopyrite 4.7‰ (n=20) wheras bulk isotopes values for active and inactive chimneys range from 4.3 to 5.4‰ (n=6). This enhances local sulphur isotope fractionation importance during hydrothermal deposits growths. Previous studies of Mid Atlantic hydrothermal fields (Snake Pit and Broken Spur) reported low ³⁴S values for sulphides (from -0.8 to 2.8‰) and were interpreted as reflecting a smaller contribution of seawater sulphate to the deposits. The range of sulphur isotopes values of chalcopyrite at Logatchev is significantly higher than the range reported from other hydrothermal field at MAR. We propose that this difference reflects the influence of the ultramafic substratum in sulphur isotope geochemistry of sulphides rather than fluid flows paths and mixing in tectonic structures. Interaction of seawater with olivine during serpentinization could provide a strong potential for reduction of sulfate to sulfide and H₂S with high water/rock ratio. Acknowledgments: We thank T.K. Kyser for the inter-laboratory comparisons of sulphur isotopes analyses.

F02 : 3P/06 : PO

Pb Isotopic Compositions of Basalts, Hydrothermal Fluids and Deposits from the Menez Gwen, Lucky Strike and South Amar Segments of the Northern Mid-Atlantic Ridge (South of the Azores)

Laure Dosso (laure@ifremer.fr),

Claire Bollinger,

Jean-Luc Charlou (charlou@ifremer.fr),

Yves Fouquet (fouquet@ifremer.fr),

Henri Bougault (Henri.Bougault@ifremer.fr),

Germain Bayon,

Anne Françoise Normand,

Sophie Parisot &

Laure Sevin

CNRS/IFREMER, BP70, 29280- Plouzane, France

The isotopic composition of lead has been measured in basaltic samples from the Menez Gwen, Lucky Strike and South Amar segments of the Mid-Atlantic ridge located south of the Azores triple junction. Hydrothermal fluids and deposits (sulfides in Lucky Strike site and Rainbow site -South Amar segment-, sulfates in Menez Gwen site) have been sampled in each of these three segments and measured for their Pb isotopic composition. The Pb isotopic compositions of the deposits are comparable to those of the basalts but for all three sites, the Pb isotopic ratios measured in the fluids are less radiogenic than those reported in the deposits and the basalts. In particular there is a significant difference (up to 0.05) in ²⁰⁷Pb/²⁰⁴Pb between the fluid and the deposit within a single segment. The difference is much less apparent in ²⁰⁶Pb/²⁰⁴Pb or ²⁰⁸Pb/²⁰⁴Pb ratios. The interpretation of this systematic difference between the Pb isotopic signature of the fluid and its associated deposit will be discussed.

F02 : 3P/07 : PO

Near-Bottom Magnetic Study in the Mid-Atlantic Ridge Segment at 21^o 40'N

Chie Honsho (honsho@ori.u-tokyo.ac.jp)¹,

Jerome Dyment (jerome@univ-brest.fr)²,

Pascal Gente (gente@univ-brest.fr)² &

Kensaku Tamaki (tamaki@ori.u-tokyo.ac.jp)¹

¹ Ocean Research Instityte, Univ. of Tokyo, Minamidai 1-15-1, Nakano-ku, Tokyo 164-8639, Japan

² UMR CNRS Domaines oceaniques, Universite de Bretagne Occidentale, 6 avenue Le Gorgeu, BP 809, 29285 Brest Cedex, France

Near-bottom geomagnetic fields were measured with a submersible on the Mid-Atlantic Ridge segment located at 21^o 40'N during the TAMMAR cruise in 1996 (Gente et al., 1996).

Because the measurements were made close to the seafloor, the observed magnetic field contains short-wavelength variations produced by the topography and the up-and-down motion of the submersible. We calculated the synthetic anomalies by assuming the uniform magnetization of the seafloor and compared the amplitudes between the synthetic and observed profiles to estimate the magnetization of the seafloor along the dive tracks.

Fifteen dives made in the segment provided three cross sections of the axial valley, one of which reaches the Brunhes/Matuyama boundary. In the spreading direction, the alternation of relatively high (10-20 A/m) and low (< 10 A/m) magnetization is recognized. The magnetization on the axis is generally high, but not particularly. Though some of the low magnetizations seem to correspond to fault zones, we consider that the variation in magnetization have existed since the formation, and that the most likely reason is the palaeointensity variation.

Including other four dives made in the inner floor, the magnetization on the axis was obtained at six different locations from the center to the southern end of the segment. It resulted in a fairly constant value of about 13 A/m. On the other hand, the magnetization distribution calculated from the surface magnetic data shows much higher values at the ends of the segment than at the center, as commonly recognized in the mid-oceanic ridges in the world. Because a constant thickness ofthe magnetic layer is assumed in the calculation, it means either the more highly magnetized basalts or the thicker magnetic layer (or both of them) at the ends of the segment. Our results of near-bottom magnetics support the latter explanation of thickened magnetic layer. We consider the most probable source for the additional magnetization would be the serpentinized peridotites, which is expected from the thinned crust and the penetration of the sea water into the mantle at the segment ends.

Gente P, Ceuleneer G, Dauteuil O, Dyment J, Honsho C, Laverne C, Le Turdu C, Mitchell NC, Ravilly M & Thibaud R, Interidge News, 5-2, 27-31, (1996).

F02 : 3P/08 : PO

Volcanic and Bathymetric Segmentation of the Southwest Indian Ridge from TOBI Sidescan Imagery West of Melville Fracture Zone

Daniel Sauter (Daniel.Sauter@eost.u-strasbg.fr)¹,

Lindsay Parson (Lindsay.M.Parson@soc.soton.ac.uk)²,

Veronique Mendel (vxm@soc.soton.ac.uk)²,

Olga Gomez (Olga.Gomez@cnes.fr)³ &

The Fuji Scientific Party

¹ IPG CNRS/ULP, 5 rue Rene Descartes, 67084 Strasbourg cedex, France

² SOC, Southampton, UK

³ GRGS, Toulouse, France

Bathymetric and gravimetric data of the ultra-slow spreading Southwest Indian Ridge (SWIR, 16 mm/yr) allow to define two contrasting segmentation patterns on each side of the Melville fracture zone (60°42'E). During FUJI (French, U-K, Japan InterRidge) cruise, conducted in october 1997, high resolution TOBI sidescan sonar data have been acquired to document the interplay between magmatic and tectonic processes in these two areas. We present an interpretation of TOBI images combined with swath bathymetry data that provides new insights into the segmentation characteristics and evolution of the SWIR axial valley between 58°27'E and 60°13'E. A wide range of volcanic features, from large seamounts to small scale hummocks and sheet flows are observed on the TOBI images. Surprisingly, these volcanic features occur almost everywhere along the axial valley floor, onto the bathymetric swells corresponding to second-order segment centers as well as in the deep and large oblique extensional basins corresponding to non-transform discontinuities. Fresh looking volcanic hummocks are forming a set of 200-800 m-high AVRs with various orientations in segment centers. They also appear fresh, in one instance forming a 200 m-high hummocky ridge in a 5000 m-deep sedimented basin and in another burying the axial valley floor in a large and intensively faulted area between two segments. Unfaulted sheet flows and seamounts have a more peculiar setting: they occur mainly at segment ends.We interpret the occurrence of different settings and kinds of volcanic constructions in terms of volcano-tectonic cycles and in terms of modes of magmatic supply within the shallow crust. We also discuss the implications of this distribution of the volcanism on the evolution of the small-scale segmentation of the SWIR.

F02 : 3P/09 : PO

Structures and Petrophysics of Rocks in the Vicinity of an Active Detachment Fault System, Western Woodlark Basin (SW Pacific)

Sybille Roller (rollers@sun2.ruf.uni-freiburg.de)¹,

Achim Kopf (kopf@obs-vlfr.fr)²,

Jan H. Behrmann

(behrmann@sun2.ruf.uni-freiburg.de)¹ &

ODP Leg 180 Shipboard Scientific Party

¹ Geologisches Institut, Universität Freiburg, Albertstr. 23 B, D-79104 Freiburg, Germany

² Géosciences Azur, B.P. 48, F-06235 Villefranche-sur-Mer Cedex, France

Within the scope of drilling at ODP (Ocean Drilling Program) Leg 180 in the Western Woodlark Basin it has been possible to sample the rocks adjacent to a major, tectonically active low-angle detachment fault. The western Woodlark Basin is a region of strong N-S extension presently leading to crustal break and the formation of an oceanic backarc basin.

Offset drilling at three sites (ODP Sites 1108, 1114 and 1117) permits to compose a synoptic section of the footwall and hangingwall, and serves as a base to evaluate contact deformation adjacent to major structures, infer deformation mechanisms, and characterize the physical properties of deformed and undeformed sediments and hard rocks.

The turbiditic syn-rift sediments of the hangingwall show brittle deformation of variable intensity, brecciation, and steeply to moderately dipping minor faults. There are ubiquitous slickensides and extensional fractures in the coarser grained lithologies. In fine grained sediments, shearing is localised in zones of scaly fabrics. One major detachment fault intersected at Site 1117 contains several meters of fault gouge, and soft ultra- cataclasite showing signs of alteration. A complete transition from ductile mylonitic to brittle deformation fabrics can be studied in the retrogressed mafic basement rocks of the tectonically exhumed footwall.

Sets of crosscutting veins mainly containing quartz, calcite and epidote group minerals are witnesses of a complex history of syntectonic fluid-rock interaction. Porosities in the sediments cored at Site 1108 generally decrease downhole in an exponential fashion. In the hangingwall of faults intersected at 160-170 m and 380 m depth below sea floor strain enhanced compaction leads to porosity loss and decrease in permeability by at least one order of magnitude. The fault zones themselves have porosity maxima. In this way, fault hangingwalls are sealed, and fluid flow is effectively channelled. Porosity and permeability of structurally intact basement rocks are very low, but increase markedly with increasing brecciation and deformation intensity.

F02 : 3P/10 : PO

Trace Elements in Serpentine Minerals A N the Role of Seawater (14°45' on the Mid Atlantic Ridge: the Logachev Hydrothermal Field)

Beate Orberger (orberger@geol.u-psud.fr)¹,

Nicole Métrich (metrich@drecam.cea.fr)²,

Catherine Mével (mevel@ccr.jussieu.fr)³,

Michelle Mosbah (mosbah@drecam.cea.fr)² &

Yves Fouquet (fouquet@ifremer.fr)⁴

¹ Laboratoire de Géochimie, Université Paris XI, Batiment 504, 91405 ORSAY Cedex, France

² Laboratoire Pierre Süe, CEA-CNRS, CE Saclayclay, Gif sur Yvette, France

³ Laboratoire de Pétrologie, Université Paris VI, Paris, 75005

⁴ IFREMER, BREST

The black smoker Logachev at 14°45'N on the Mid Atlantic Ridge is the first hydrothermal field which was found on the top of completely serpentinized ultramafic rocks outcropping at the seafloor. Trace element studies on serpentine minerals have been carried out in order to study the role of seawater during serpentinization. Nuclear microprobe analyses have been performed on three serpentine minerals : bastite, the pseudomorphoses after pyroxenes, mesh serpentine which are olivine replacements and tiny serpentine veins developing in bastite cleavages or crosscutting them. Trace elements such as Cl, F, S, Cu, Zn, Ca, K, Ni, Cr and Mn were detected from several tenth to several thousands of ppm. The trace element concentrations and the interelement relationship in serpentines vary a) with the serpentine mineral and b) with the geographic location to the black smoker. Chlorine and in part S originated from seawater, whereas Cu, Zn, Ca, K, Ni, Cr and Fe and the major amount of S were mobilized during serpentization and partitioned between the serpentine and the aqueous solution. The physico-chemical conditions during serpentinization vary from high T (>350°C), neutral to alkaline pH and fO₂ close to the iron magnetite boundary far away from the Logatchev site to lower T (<350°C), acidic and oxidizing conditions near the black smoker