Simon M.F. Sheppard

Use of 13C to trace origin and cycling of inorganic carbon in the Rhône river system

Dissolved inorganic carbon (DIC) and particulate inorganic carbon (PIC) were sampled at 20 sites on the Rhone and its main tributaries from Switzerland to the Mediterranean, in March, September and July (flood event) 1996. Flux estimation and DIC and PIC values are used to trace watershed inputs and fluvial processes affecting the inorganic carbon budget. DIC concentration and values vary between ∼1 mM at −5‰ near the Rhones source and ∼3 mM at −10‰ near its mouth, with tributaries draining lowland, carbonate-dominated areas at ∼3.8 mM and −11‰. There is an inverse relationship between river DIC concentration and DIC . The headwaters are characterized by high DIC values and low DIC, indicating a minor input of respired CO2. DIC should be mainly produced through the reaction of carbonate with atmospheric CO2 or organic/sulfuric acids. In lowland, carbonate-dominated areas, values of ∼−11‰ are consistent with production of DIC by weathering of limestone C (≈−1‰) with CO2 derived from the oxidation of organic matter (≈−21.5‰). The evolution of the Rhones DIC concentration and appear to be largely determined by mixing between the Rhone and its main tributaries. The CO2 fluxes and DIC mass-balance, calculated for the different segments of the Rhone, point to a net, but very limited input of fluvial respired CO2 from the oxidation of organic C (<5%). The PIC represents a small but quite variable fraction of the inorganic carbon (<3%), except on Alpine catchments and during floods. In these cases, it can be up to 10% of the DIC concentration and export. The carbon and oxygen isotopic composition of PIC from the Rhone during a flood and from Alpine tributaries falls within the δ-range of limestones (≥−2‰; ≥−6‰). Relatively light PIC isotopic composition indicate mixing of sedimentary rocks and carbonate precipitated in situ. Data for the Rhone system are typical of watersheds presenting high chemical weathering rates for HCO3, and Alpine/lowland conditions.

Sulphur isotope variations in the mantle from ion microprobe analyses of micro-sulphide inclusions

Abstract 21 samples of sulphide trapped either as liquid globules or grains in various minerals (olivine, pyroxenes, ilmenite and garnet) or rocks (basalt glasses, peridotites, eclogites and kimberlites) of mantle origin, have been analysed for their sulphur isotope, and their Cu, Ni, Fe compositions by ion microprobe. The results show a wide range of δ 34 S values between −4.9 ± 1 and +8 ± 1‰. Sulphides with high nickel contents (up to 40% pentlandite), corresponding mostly to residual peridotites, have δ 34 S values ranging from −3.2‰ to +3.6‰ with a mode of +3 ± 1‰, compared to low Ni content sulphides, mostly contained in pyroxenites, OIB and kimberlites, ranging from −3.6‰ to +8‰ with a mode of +1 ± 1‰. The δ 34 S of sulphides originating from within the mantle are variable. The sulphide globules with high Ni contents and δ 34 S values close to +3‰, are probably produced by 10–20% partial melting of a mantle source containing 300 ppm sulphur as an upper limit and having a δ 34 S value of +0.5 ± 0.5‰. This difference in δ 34 S values suggests a high-temperature S-isotope fractionation of ≈ +3‰ between liquid sulphide and the sulphur dissolved in the silicate liquid. The sulphur isotopes balance in the system upper mantle + oceanic crust + continental crust + seawater requires a mean δ 34 S value of the primitive upper mantle of +0.5‰, slightly but significantly different from that of chondrites (+0.2 ± 0.2‰) [1].

Ice age at the Middle–Late Jurassic transition?

A detailed record of sea surface temperatures in the Northern Hemisphere based on migration of marine invertebrate fauna (ammonites) and isotopic thermometry (δ18O values of shark tooth enamel) indicates a severe cooling at the Middle–Late Jurassic transition (MLJT), about 160 Ma ago. The magnitude of refrigeration (1–3°C for lower middle latitudes) and its coincidence in time with an abrupt global-scale fall of sea level documented through sequence stratigraphy are both suggestive of continental ice formation at this time. Ice sheets may have developed over the high-latitude mountainous regions of Far-East Russia. The drastic cooling just post-dated the Middle–Late Callovian widespread deposition of organic-rich marine sediments (e.g. northwestern Europe, Central Atlantic, and Arabian Peninsula). This thermal deterioration can thus be ascribed to a downdraw in atmospheric CO2 via enhanced organic carbon burial which acted as a negative feedback effect (i.e. the inverse greenhouse effect). The glacial episode of the MLJT climaxed in the Late Callovian, lasted about 2.6 Myr, and had a pronounced asymmetrical pattern composed of an abrupt (∼0.8 Myr) temperature fall opposed to a long-term (∼1.8 Myr), stepwise recovery. The glacial conditions at the MLJT reveal that atmospheric CO2 levels could have dropped temporarily to values lower than 500 ppmv during Mesozoic times.

Oxygen isotope exchange between dissolved phosphate and water at temperatures ≤135°C: inorganic versus biological fractionations

Oxygen isotope fractionations have been measured experimentally between dissolved orthophosphate ions H2PO4− and water from 50°C to 135°C. The derived equilibrium fractionation equation reveals that biogenic apatites are systematically 18O-depleted (8‰ at 20°C) compared with inorganic dissolved phosphate. The activation energy is relatively high at 134 kJmol−1 and the kinetics of the homogeneous inorganic exchange reaction are very slow (k = 10−13 to 10−11 s−1) when extrapolated to surface temperatures. In seawater, the oxygen isotope composition of dissolved phosphate is not in equilibrium with water; its composition should directly reflect source contributions or biological recycling. Any deviation from the composition of the inorganic sources may be attributed to biological recycling of this nutrient. The original isotopic compositions of fossils may be modified in less than 105 years at a temperature of 20°C through mechanisms of dissolution and precipitation in the presence of crustal aqueous fluids. The differences in oxygen isotope fractionations between biogenic or inorganic phosphate and water are applied to re-evaluate the significance of δ18O values of Phanerozoic and Precambrian phosphate deposits.

Experimental hydrogen isotope studies—I. Systematics of hydrogen isotope fractionation in the systems epidote-H2O, zoisite-H2O and AlO(OH)-H2O

HD Fractionation factors between epidote minerals and water, and between the AlO(OH) dimorphs boehmite and diaspore and water, have been determined between 150 and 650°C. Small water mineral ratios were used to minimise the effect of incongruent dissolution of epidote minerals. Waters were extracted and analysed directly by puncturing capsules under vacuum. Hydrogen diffusion effects were eliminated by using thick-walled capsules. HD Exchange rates are very fast between epidote and water (and between boehmite and water), complete exchange taking only minutes above 450°C but several months at 250°C. Exchange between zoisite and water (and between diaspore and water) is very much slower, and an interpolation method was necessary to determine fractionation factors at 450 and below. For the temperature range 300–650°C, the HD equilibrium fractionation factor (αe) between epidote and water is independent of temperature and Fe content of the epidote, and is given by 1000 In αepidote-H2Oe = −35.9 ± 2.5, while below 300°C 1000 In αepidote-H2Oe = 29.2(106T2) − 138.8, with a ‘cross-over’ estimated to occur at around 185°C. By contrast, zoisite-water fractionations fit the relationship 1000 In αzoisite-H2Oe = − 15.07 (106T2) − 27.73. All studied minerals have hydrogen bonding. Fractionations are consistent with the general relationship: the shorter the O-H -- O bridge, the more depleted is the mineral in D. On account of rapid exchange rates, natural epidotes probably acquired their H-isotope compositions at or below 200°C, where fractionations are near or above 0%.; this is in accord with the observation that natural epidotes tend to concentrate D relative to other coexisting hydrous minerals.

LITHOSPHERIC SHEAR ZONES AND MANTLE-CRUST CONNECTIONS

Abstract A crustal-scale ductile shear zone network in the Precambrian granulite-facies crust of Madagascar is examined to determine the nature of the connections between the mantle and lower crust. Based on three independent data sets — field and satellite mapping, C- and O-isotope geochemistry and gravimetry — this crust is divided into three zones: (1) outside of shear zones; (2) minor shear zones that are 350 km long (up to 1000 km) and 20–35 km wide. The mantle is uplifted by about 10 km beneath the major shear zones. The major shear zones are rooted in and are inferred to be controlled by the mantle; they directly tapped mantle-derived CO 2 . The small-scale minor shear zones were controlled by crustal processes and focused crustally derived H 2 O-rich±CO 2 fluids. The regular distribution of the shear zones on a crustal scale is in agreement with models of buckling of the continental lithosphere in a compressional context. The propagation of these mechanical instabilities promoted and channelled fluid flow. These major Pan-African shear zones thinned the crust and were reactivated during the subsequent drifting of Madagascar and opening of the Indian Ocean during Jurassic to Cretaceous times. They also controlled many of the brittle fault zones in the overlying sedimentary basins. Mantle-rooted large-scale shear zones are inferred to be a general feature of cratonic areas reactivated by shear zone systems.

Rare earth element contents of Jurassic fish and reptile teeth and their potential relation to seawater composition (Anglo-Paris Basin, France and England)

The rare earth element (REE) chemistry of Jurassic shelf seawater from western Europe (Anglo-Paris Basin) was investigated by analyzing the fish and reptile teeth deposited in shallow to deeper water (<200 m) environments. REE patterns in apatites are controlled by the host sediments. Vertebrate teeth sampled from the siliciclastic sediments (calcareous sandstones and shales) show flat shale-normalized REE patterns that reflect the dominant influence of the continental source from which the REE were derived. Carbonate deposits, protected from the clastic sources, contain fish and reptile teeth whose REE patterns reflect more accurately the REE composition of the overlying water column. The REE patterns are similar through the Bajocian to the Oxfordian and are characterized by a depletion in heavy rare earth element (Dy/YbN=1.8–5.0) compared to modern seawater compositions (Dy/YbN=0.8–0.9). The fractionations among HREE in the Jurassic seawater are more efficient than in modern oceans and increases with depth instead of decreasing. These HREE removal processes operated as early as the Paleozoic and disappeared during the Late Cretaceous. The fractionation of HREE relative to LREE is correlated with the increasing depth of the basin as indicated by two independent criteria that are the sedimentary structures and the oxygen-isotope composition of coexisting brachiopods. The Dy/YbN ratio of the marine biogenic phosphates can be used as a proxy of paleo-water depths at the scale of a water mass. Ce anomalies (ΩCe) in biogenic apatite-bearing limestones are variable and systematically negative. Relative to modern surface seawater and Bathonian values (mean ΩCe=−0.63±0.10), weak Ce anomalies during the Callovian (mean ΩCe=−0.21±0.09) suggest the onset of more reducing conditions. This redox change coincided with an increase of seawater temperatures as suggested by the oxygen-isotope compositions measured on the same teeth. We speculated that the Callovian low ΩCe could result from a decrease in the pO2 due to the warming of seawater.

Hydrogen and oxygen isotope variations in the high himalaya peraluminous Manaslu leucogranite: Evidence for heterogeneous sedimentary source

Abstract The Manaslu granite belongs to the High Himalaya leucogranitic belt which was produced by melting of the crust during postcollisional thrusting. δD and δ18O values have been determined for whole rock and coexisting minerals from the ~8 km thick Manaslu massif and its 50 km long dyke sheet, its country rocks and the Formation 1 (F1) paragneisses which are the source of the granite. For the granite, δDmusc range from −70 to −85%. and δ18O W.R. from 10.9 to 12.8%. H and O-isotope fractionations among minerals are consistent with high temperature equilibrium and, for oxygen, closed system evolution. A few samples, coming mainly from the dyke swarm, have very low δD values (down to −188%.) and biotite-muscovite H-isotope fractionations indicative of disequilibrium; the D H ratios of associated magmatic tourmaline are essentially unmodified. From the distribution of δD values in the granite and its country rocks, circulation of very low deuterium meteoric hydrothermal waters was extremely localized. Because these waters are depleted in deuterium by up to 50%. relative to modern meteoric waters, the Manaslu area was either at an altitude substantially higher than that of today (2500–6000 m for the analyzed samples) or a mountain chain once existed to the south. The F1 gneisses have δ18Oquartz between 12 and 14.3%. which confirms that the granite was generated from F1, but δD values are ≈20%. higher than in the granite. Such a difference can be a result of degassing of the magma and/or introduction of fluid in the melting zone. Infiltration of low δD fluid (≈−90%.) into the hot but dry F1 probably triggered partial melting; these fluids could have come from the dehydration of the Midlands sediments which are separated from the overlying F1 by the Main Central Thrust. The correlations among δ18O, ( 87 Sr 86 Sr) 20 Ma and ϵNd values in both F1 and the granite indicate that the variations of these isotopic ratios in the Manaslu are inherited from those in F1 at the time of melting. In turn, these ratios in F1 are related to the proportion of quartz and phyllosilicates for the isotopic ratios of Nd and O, and to the quantity of radiogenic Sr generated within the sediment, which is a function of age and Rb content (amount of phyllosilicate). Some other Himalayan leucogranites require either other crustal source rocks or the δ18O and 87 Sr 86 Sr ratios of F1 vary along the Himalaya.

The Amazon River: behaviour of metals (Fe, Al, Mn) and dissolved organic matter in the initial mixing at the Rio Negro/Solimoes confluence

We studied the changes in major elements and organic carbon concentrations during the initial stage of the mixing of the black (Rio Negro) and the White (Rio Solimoes) waters in the Amazon River basin to understand the geochemical processes that could control the redistribution between particulate and dissolved fractions. Water samples were collected at six stations including the Rio Negro and the Rio Solimoes and four stations downstream from the confluence. The relative contributions of the two tributaries were determined using a triple tracer approach (d 18 O, dD, Cl � ). Particulate (>0.2 Am) and dissolved (<0.2 Am) concentrations of major elements (Ca, Mg, Fe, Al, Si) and organic carbon (POC and DOC) were measured. Major elements in the particulate fraction were found to have a nonconservative behaviour in the initial stage of the mixing due to mineral removal. In the dissolved fraction, only the DOC, Fe, and Mn behaved nonconservatively. The Fe losses could be due to preferential removal of Fe bound to N-rich organic matter (OM) and/or to preferential removal of Fe oxyhydroxides. The increasing dissolved manganese content in the dissolved phase is explained by a reductive dissolution of manganese oxides due to massive inputs of phenolic-rich OM from the Rio Negro. The amount of DOC removed from the water column in the initial stage of the mixing would represent 4% of the total annual DOC flux of the Amazon River at the reference gauging station of Obidos. D 2002 Elsevier Science B.V. All rights reserved.

East Pacific Rise at latitude 21°N: isotopic composition and origin of the hydrothermal sulphur

The sulphur isotope composition of 16 pyrite and chalcopyrite samples from recent sulphide deposits (“Cyana”—project RITA) and active sulphide mineralisation (“Alvin”—project RISE) associated with hydrothermal sources at 380±30°C on the East Pacific Rise at latitude 21°N have been measured. The34S/32S ratios are relatively uniform and essentially identical for both sites: δ34S=+1.4to3.0%. (CDT), mean +2.1‰. The sulphides were analysed after the majority of the very numerous micro-inclusions of anhydrite had been removed. Two independent physico-chemical analyses of the data demonstrate that about 90% of the sulphur was leached from the basaltic host rocks by the circulating seawater-hydrothermal fluids.