Geochemical Evidence for Subduction Fluxes, Mantle Melting and Fractional Crystallization Beneath the South Sandwich Island Arc

The volcanoes of the South Sandwich island arc follow threedistinct series: low-K tholeiitic (followed by Zavodovski, Candlemas,Vindication, Montagu and Bristol), tholeiitic (followed by Visokoi,Saunders and Bellinghausen) and calcalkaline (followed by Leskov,Freehand and part of Cook and Thule). Flux calculations indicatethat the percentage contribution of the subduction componentto the mantle source of all three series varies from undetectable(e.g. Zr) through small (e.g. Nd=20%) and moderate (e.g. La,Ce, Sr=50–80%) to dominant (e.g. Pb, K, Ba, Rb, Cs >90%)with little change along the arc. Isotope systematics (Pb, Nd,Sr) show that this subduction component obtains a greater contributionfrom altered oceanic crust than from pelagic sediment. Elementsfor which the subduction contribution is small show that themantle is already depleted relative to N-MORB mantle (equivalentto loss of an 2•5% melt fraction) before melting beneaththe arc. After addition of the subduction component, dynamicmelting of this depleted mantle then causes the variations inK that distinguish the three series. The estimated degree ofpartial melting (20%) is slightly greater than that beneathocean ridges, though geothermometry suggests that the primarymagma temperature (1225C) is similar to that of primary MORB.About half of the melting may be attributed to volatile addition,and half to decompression. Dynamic melting involving three-dimensional,two-phase flow may be needed to explain fully the inter-islandvariations. KEY WORDS: geochemistry; petrology; fluxes; melting; subduction ^*Corresponding author     

Initiation of Subduction Zones as a Consequence of Lateral Compositional Buoyancy Contrast within the Lithosphere: a Petrological Perspective

Tonga and Mariana fore-arc peridotites, inferred to representtheir respective sub-arc mantle lithospheres, are compositionallyhighly depleted (low Fe/Mg) and thus physically buoyant relativeto abyssal peridotites representing normal oceanic lithosphere(high Fe/Mg) formed at ocean ridges. The observation that thedepletion of these fore-arc lithospheres is unrelated to, andpre-dates, the inception of present-day western Pacific subductionzones demonstrates the pre-existence of compositional buoyancycontrast at the sites of these subduction zones. These observationsallow us to suggest that lateral compositional buoyancy contrastwithin the oceanic lithosphere creates the favoured and necessarycondition for subduction initiation. Edges of buoyant oceanicplateaux, for example, mark a compositional buoyancy contrastwithin the oceanic lithosphere. These edges under deviatoriccompression (e.g. ridge push) could develop reverse faults withcombined forces in excess of the oceanic lithosphere strength,allowing the dense normal oceanic lithosphere to sink into theasthenosphere beneath the buoyant overriding oceanic plateaux,i.e. the initiation of subduction zones. We term this conceptthe ‘oceanic plateau model’. This model explainsmany other observations and offers testable hypotheses on importantgeodynamic problems on a global scale. These include (1) theorigin of the 43 Ma bend along the Hawaii–Emperor SeamountChain in the Pacific, (2) mechanisms of ophiolite emplacement,(3) continental accretion, etc. Subduction initiation is notunique to oceanic plateaux, but the plateau model well illustratesthe importance of the compositional buoyancy contrast withinthe lithosphere for subduction initiation. Most portions ofpassive continental margins, such as in the Atlantic where largecompositional buoyancy contrast exists, are the loci of futuresubduction zones. KEY WORDS: subduction initiation; compositional buoyancy contrast; oceanic lithosphere; plate tectonics; mantle plumes; hotspots; oceanic plateaux; passive continental margins; continental accretion; mantle peridotites; ophiolites     

The anatomy of an early Dinantian terraced floodplain: palaeo-environment and early diagenesis

The lowermost Carboniferous rocks in the Cockburnspath area of east Berwickshire (southern Scotland) are interpreted as coastal floodplain sediments. A lower mudstone-dominated unit is composed of silty mudstones and shales with subordinate sandstones and argillaceous ferroan dolomites (cementstones). These are interpreted as distal floodplain sediments with periodic crevasse-splay deposition. The dark grey colour of the mudrocks suggests deposition in reducing conditions, probably in floodplain lakes. Most of the cementstones are concretionary, some with septarian cracks, suggesting an early diagenetic origin. An immature palaeosol suggests periodic pedogenesis under improved drainage. A synsedimentary erosion surface indicates incision of a valley into the floodplain, presumably in response to base-level lowering. An upper sandstone-dominated unit starts with fine-grained rippled sandstones, cut by small channel sandstones. These are interpreted as floodplain lake deposits fed by crevasse channels. A distinctive conglomerate with cementstone clasts, wood fragments and fish remains is interpreted as a major overbank deposit, dumped into a pre-existing floodplain lake. A bivalve fauna was established in the overlying mudstones, followed by a thin limestone with a restricted marine fossil assemblage, showing that seawater flooding of the lakes occurred at times. Mudrocks throughout the sequence contain no pyrite, except for the marine band which has an organic-carbon/sulphur ratio and degree of pyritization value typical of marine sediments. The concretionary cementstones have δ¹³C values around —4 to —6%^0/00 PDB which are interpreted as indicative of anaerobic oxidation of organic matter. The combined geochemical data suggest a significant involvement of iron reduction in cementstone formation, although the δ¹³C values are ambiguous in assessing the relative involvement of methanogenesis and methane oxidation. Limited seawater inundation of the floodplains might have supplied magnesium and calcium ions for dolomite formation assuming that any H₂S derived as a result of sulphate reduction was oxidized by iron reduction. Alternatively a weathering source for solutes might have been involved.     

U-series Isotope Data on Lau Basin Glasses: the Role of Subduction-related Fluids during Melt Generation in Back-arc Basins

New thermal ionization mass spectrometry U-series disequilibriumdata are presented for 24 basaltic to dacitic glasses from activespreading centres in the back-arc Lau Basin (SW Pacific), togetherwith additional inductively coupled plasma mass spectrometrytrace element analyses and Sr–Nd–Pb isotope data.Valu Fa Ridge samples, adjacent to the arc front, have highU/Th and (²³⁰Th/²³⁸U) <1, implying a recent (<<350ka) addition of a U-rich slab-derived fluid. The Valu Fa datacan be combined with existing ²³⁰Th–²³⁸U data for theCentral Tonga arc to infer a fluid addition event at     

Mapping seabird nesting habitats in Franz Josef Land, Russian High Arctic, using digital Landsat Thematic Mapper imagery

Supervised classification of digital Landsat satellite images was used to locate seabird nesting habitats in the Russian High Arctic archipelago of Franz Josef Land, a region where the avifauna is poorly known and ecologically vulnerable. Major seabird nesting colonies are readily identifiable in Landsat Thematic Mapper (TM) imagery of the region due primarily to the distinctive spectral signature of vegetation on ornithogenically altered soils below bird cliffs. Supervised image classification was used to pinpoint areas displaying spectral characteristics typical of documented seabird nesting habitats. A total of 101 seabird nesting colony locations identified in Russian and Western literature from 1898 to 1996 was used as training sites to develop spectral signatures from a summer TM image mosaic for use in a supervised maximum likelihood classification. The classified image was thresholded and compared to a map of documented nesting locations. Of the 101 field-documented nesting sites, 96 were clearly identified in the classified image. An inventory was produced of all undocumented seabird habitats suggested by the classification, totalling over 300 sites. The methodology used may be applicable to other arctic regions and is intended as a first step when planning ecological protection zones in remote and inaccessible arctic regions.     

Late Weichselian ice-sheet sensitivity over Franz Josef Land, Russian High Arctic, from numerical modelling experiments

A numerical ice-sheet model was run in order to produce reconstructions of the Late Weichselian ice coverage of Franz Josef Land, Russian High Arctic. The model grid covers the archipelago and surrounding shelf, but does not include the whole Barents-Kara region or the extensive ice cover that may have built up there. One experiment, where rates of iceberg calving at the grounded margin were curtailed because of the assumed presence of permanent thick sea ice, yielded a single I.8 km-thick ice dome which covered the entire archipelago and surrounding sea. If, however, iceberg calving were included in the model's environmental input, the extent of the ice sheet would be limited to the periphery of the archipelago. If a large ice sheet existed over Franz Josef Land, the deglaciation of the islands may have been linked to the decay of the adjacent Barents-Kara Sea Ice Sheet, permitting iceberg calving (enhanced by relative sea-level rise) to occur. The introduction of a water-depth-related iceberg calving function at 15 000 yr ago forced an initial rapid rate of ice-sheet decay of 30 000 km³ 1000 yr'. However, as the ice sheet thinned, and isostatic rebound began, the calculated rate of iceberg calving was reduced such that ice remained over the archipelago at 8000 yr ago. The model's failure to simulate complete ice-sheet decay by 8000 yr ago is at variance with radiocarbon-dated raised terraces on Franz Josef Land, which indicates the complete deglaciation of the islands at this time.     

Remote sensing of ice cap outlet glacier fluctuations on Nordaustlandet, Svalbard

Landsat multispectral scanner images and 1:50, 000 scale aerial photographs are used to measure marginal fluctuations in 22 outlet glaciers of the Nordaustlandet ice caps, Svalbard, for all or parts of the period 1969 to 1981. Little was previously known about the behaviour of these glacier termini. Digital analysis of Landsat computer compatible tapes yielded measurement errors of less than ±150m, whereas data extracted from aerial photographs had an accuracy of ±25 m. Of the 22 glacier termini examined using aerial photographs, 15 were retreating, four were static and three were advancing. Retreat was usually in the order of hundreds of metres during the period of observation. For any outlet glacier, retreat was probably a result of either (1) glacier response to climatic wanning since the early part of the 20th century, or (2) stagnation and thinning during the quiescent period between surges. Short term iceberg calving events may also be responsible for retreat in a few cases. All observed outlet glaciers of Austfonna were static or retreating between 1969 and 1981, but analysis of Landsat imagery and aerial photographs showed that five outlet glaciers from Vestfonna and one from Vegafonna were advancing for all or part of that period. Bodleybreen advanced by a mean of 440 and a maximum of 580m a⁻¹ between 1976 and 1981. Bodleybreen, Søre Rijpbreen, and Palanderbreen are identified as surging, based on increases or changes in the pattern of surface crevassing. The ice streams Aldousbreen, Frazerbreen, and Idunbreen also advanced between 1976 and 1981, but surface crevasse patterns remained largely unaltered and surging is not inferred.     

Maximum extent of the Eurasian ice sheets in the Barents and Kara Sea region during the Weichselian

Based on field investigations in northern Russia and interpretation of offshore seismic data, we have made a preliminary reconstruction of the maximum ice-sheet extent in the Barents and Kara Sea region during the Early/Middle Weichselian and the Late Weichselian. Our investigations indicate that the Barents and Kara ice sheets attained their maximum Weichselian positions in northern Russia prior to 50 000 yr BP, whereas the northeastern flank of the Scandinavian Ice Sheet advanced to a maximum position shortly after 17 000 calendar years ago. During the Late Weichselian (25 000-10 000 yr BP), much of the Russian Arctic remained ice-free. According to our reconstruction, the extent of the ice sheets in the Barents and Kara Sea region during the Late Weichselian glacial maximum was less than half that of the maximum model which, up to now, has been widely used as a boundary condition for testing and refining General Circulation Models (GCMs). Preliminary numerical-modelling experiments predict Late Weichselian ice sheets which are larger than the ice extent implied for the Kara Sea region from dated geological evidence, suggesting very low precipitation.