The relationship between riverine lithium isotope composition and silicate weathering rates in iceland

Earth and Planetary Science Letters 287 (2009) 434–441 Contents lists available at Earth and Planetary Science Letters The relationship between riverine lithium isotope composition and silicateweathering rates in Iceland N. Vigier , S.R. Gislason K.W. Burton , R. Millot , F. Mokadem a CRPG-CNRS, Nancy-Université, 15 rue ND des Pauvres, 54501 Vandoeuvre les Nancy Cedex, Franceb Univ. of Iceland, Icelandc The Open University, Milton Keynes, UKd BRGM, Metrology, Monitoring, Analysis Division, 3 Av. Claude Guillemin, BP 6009, 45060 Orleans Cedex 2, France This study presents lithium isotope and elemental data for the dissolved phase and suspended and bedload Received 23 July 2009 sediments of the major Icelandic rivers. For the dissolved phase, δ7Li values range between 10.1‰ and Accepted 20 August 2009 23.8‰, while river sediments display lower and much more homogeneous values (δ7Li = 3.1‰–4.8‰), close Available online 9 September 2009 to the composition of unweathered Mid-Ocean Ridge Basalt (MORB). High δ7Li values are associated with Editor: M.L. Delaney high K/Li, Na/Li and Mg/Li ratios, in waters draining mainly old and weathered basalt catchments, whereaslow δ7Li rivers are located in younger parts of the island. Simple mixing between precipitation, Li-rich hydrothermal springs and basalt weathering is unable to explain the entire range of δ7Li values. Instead, a simple model of Li uptake by secondary minerals, associated with clay–water Li isotope fractionation (Δ7Li ranging from −1‰ to -7.5‰) can explain both water and sediment δ7Li values. A negative correlation is silicate weathering observed between basalt chemical erosion rates and δ7Li measured in Icelandic rivers, and an empirical law is inferred. Comparison with literature data suggests that this relationship may be applicable at a more global scale, and, if confirmed, could be of particular use for estimating the evolution of continental weatheringpreserved in marine sedimentary records. However, more data are now needed for rivers draining silicatestypical of the continental crust, in order to refine large scale modelling.
2009 Elsevier B.V. All rights reserved.
Secondly, the δ7Li composition of the ocean (31.2‰, e.g. ) is significantly higher than any Lithium has two stable and light isotopes (7.52% 6Li and 92.48% 7Li) presently known Li source (the global riverine composition is that fractionate significantly during silicate chemical erosion (weath- estimated to possess a δ7Li value of 23.4‰ and that for hydrothermal ering). The few studies of rivers undertaken thus far have suggested exchange is 9‰, ). Two processes that dissolved Li is mainly derived from silicates, even in large basins have been proposed to account for Li isotope fractionation during of mixed lithology, and that the Li isotope composition of waters weathering: (i) rock or mineral leaching, (ii) the formation of varies with indices of chemical erosion ( secondary minerals. Some soil studies suggest that 7Li is preferentially released into solution from fine grained sediments sedimentary records may therefore preserve unique information on This is a priori surprising since 6Li past silicate chemical erosion rates, particularly if the relationship diffuses much quicker than 7Li and other studies suggest little isotope between δ7Li in waters and silicate chemical erosion rate can be fractionation during Li loss from a basalt ( quantified at a continental scale.
In contrast, both experiments and observa- The main evidence for Li isotope fractionation accompanying tions provide evidence for Li isotope fractionation during clay silicate chemical erosion is firstly, that measured δ7Li ratios of formation, at both low and high temperatures continental waters (δ7Li = ((7Li/6Li) / (7Li/6Li)LSVEC − 1) ⁎ 1000) are systematically higher than the rocks they drain and the sediments they carry (suspended and bed load) Small catchments or monolithological basins are characterized by fewer variables, and provide a means to deconvolve the primarycontrols on silicate weathering rates. The volcanic island of Iceland isone of the few regions where river water chemistry and sediment ⁎ Corresponding author.
E-mail address: (N. Vigier).
fluxes have been extensively studied and monitored for more than 0012-821X/$ – see front matter 2009 Elsevier B.V. All rights reserved.
doi:

N. Vigier et al. / Earth and Planetary Science Letters 287 (2009) 434–441 30 years. Geology and rock composition are also well-constrained and rivers have been recognized in Iceland: (i) spring-fed rivers, mainly anthropogenic inputs are minimal. Weathering in Iceland has then located in the central volcanic part where the high rock permeability been extensively studied (e.g. precludes significant surface runoff, (ii) glacier-fed rivers, and (iii) direct runoff rivers draining mainly older basalts, where compaction and ). Recent studies have demonstrated a large sealing by secondary minerals reduces the permeability. Soils are range in chemical and physical erosion rates ( generally thin and the main secondary minerals in soils are allophane Runoff and the age of the basalt and ferrihydrite ().
appear to be two key controls on weathering. In addition, glass may Quaternary volcanic rocks formed under ice during the last 3 Myr play a major role because it can dissolve up to 6 times faster than are glassy (hyaloclastites) but those formed when the island was ice- crystalline basalt (e.g. free are mostly in the form of crystalline lava flows. Thus, the most and is abundant in the hyaloclastites of the central and recent basalts are characterized by a glassy texture, in contrast with glaciated part of the island.
the older northern lava flows which are mainly crystalline.
This study presents Li isotope data for waters and sediments of the More than twenty rivers were sampled across Iceland in June/July main Icelandic rivers. The samples are the same as those used for 2001 (). Catchments from the volcanically active zone in the constraining chemical erosion rates in from centre of Iceland, where the hydrothermal contribution to rivers can dissolved major elements, allowing a direct comparison of chemical be significant, were avoided. Glacier-fed rivers have also been erosion and riverine δ7Li.
sampled, notably the Skaftafellsa River (#23) which is sourced directlyby the large Vatnajokull glacier. More details on the samples and their locations can be found in and River monitoring by the University of Iceland and the Hydrological The island of Iceland is located on the Mid-Atlantic ridge. A Division of the National Energy Authority provides precise estimates combination of sea-floor spreading and associated volcanism has of discharge and corresponding runoff for each watershed studied produced a symmetrical SW–NE zonation in the age of the basalts, here. In addition, the Total Suspended Sediment (TSS) load has been from the centre of the island towards the east and north-west coasts.
regularly recorded for a number of Icelandic rivers over the last four The older basalts (>3 Ma) are located in the north and east of the island, while the most recent basalts are mainly located in the central ). Both records provide valuable information and volcanically active zone.
that can be used to estimate the physical and chemical erosion rates of The climate in Iceland is oceanic boreal, and less than 25% of the most of the watersheds of the sampled rivers. For 2001, a wide range island is vegetated and about 12% is covered by glaciers. Three types of of physical erosion rates were inferred for Icelandic catchments, Fig. 1. Map of the Icelandic rivers sampled during June 2001 for Li analyses. Sample numbers refer to sample designation as given in .
N. Vigier et al. / Earth and Planetary Science Letters 287 (2009) 434–441 between 21 and 4864 t/km2/yr, with an average of 519 t/km2/yr sequence of analyses blanks are measured before and after each which compares well with previous estimates This sample and standard for background correction. Blank values are low, is about twice the world average estimated by typically 3–4 mV for the 7Li (i.e. 0.2%), and 5 min wash time is . Chemical erosion rates, which can be estimated from the flux sufficient to achieve a stable background value. Since blanks are very of dissolved elements exported by rivers, are lower than physical stable, the background correction applied to samples and standards is erosion rates, ranging from 13 to 333 t/km2/yr. Chemical erosion made using the average value of the two bracketing blanks.
rates correlate positively with physical erosion rates and runoff, and The accuracy of this procedure has been checked using Li solutions negatively with the mean age of the drained basalts (e.g. made from LSVEC powder and 6Li and 7Li spikes of known composition (Li6-N and Li7-N respectively, ),and reference materials. Two analyses of seawater gave δ7Li values of 3. Analytical procedure 31.25 ± 0.16 and 31.31 ± 0.12 respectively, which closely correspondto published values (e.g. see review in The JB-2 basalt reference material has also been analyzedand gave δ7Li values of 5.65 ±0.12 and 5.73 ± 0.12, which are within All river samples were filtered on the field using acetate cellulose the published range for this basalt filters (0.2 μm) which separated the dissolved phase from suspended particles, collected on the filter with ultra-pure water. Li concentra-tions (along with other major and trace cations) were determined using ICP-MS. For the suspended and bed load sediments, Li contentswere calibrated against a number of International rock reference materials including AC-E (granite), WS-E (dolerite), and BIR-1, BCR-2and BHVO-1 (basalts). The external reproducibility, determined from The Li contents of Icelandic river waters range between 20 ng l−1 repeat measurements of BCR-2, is better than 2% (1σ), where the for the Heioarvatn and Geithellnadalur Rivers (#19 and #21), draining measured concentration of Li in BCR-2 of 9.7 ± 0.2 ppm (1σ) is old basalts in the eastern part of the island (basalt mean age 5.9 Ma), indistinguishable from the certified value of 9 ± 2 ppm (USGS and 2.2 μg l−1 for the Vestari Jökulsa River. This range is similar to that Certificate of Analysis). For the dissolved load Li was calibrated observed for rivers in a small watershed located in western Iceland against an in-house standard (Scottish river water). The external (). The mean Li content of Icelandic reproducibility, determined by repeat measurement of the natural rivers is 0.6 μg l−1. This is significantly lower than both the Li content river water certified reference material SLRS-4, is better than 3%. The of world-wide rivers (and the Li contents of rivers measured Li content of SLRS-4 of 0.48 ± 0.01 ppb is similar to the value draining mixed lithology basins of the Himalayas ( of 0.54 ± 0.07 ppb obtained by for the same This is likely to be due to low Li contents in basalts relative to river water.
granites and gneisses, and also to high runoffs in Iceland during the icemelting season. The Li contents of two samples from Geysir hot spring, located in central Iceland, have also been determined and are 389 μg/l and 360 μg/l respectively.
Special caution was taken for ensuring complete dissolution of all samples before the Li separation chemistry. River waters were evaporated in Teflon beakers and the residues were dissolved severaltimes in concentrated nitric acid. Bulk sediments and suspended Li isotope compositions have been determined for all river waters particles were digested in a mixture of HF–HNO3–HCl and HClO4, and some suspended sediment and bedload samples. The δ7Li values following the procedure described in and of the waters range between 10.1‰ for the Skaftafellsa (#23), a previously used for measuring Th isotopes. Cation exchange resin glacier-fed river in the south of Iceland, and 23.8‰ for the Vididalsa was used to separate Li from the sample matrix, following a procedure River (#6) located in northern Iceland. This range is similar to those modified from that of The 100% recovery published for world-wide and Himalayan rivers was checked for each sample, considering analytical uncertainties. Li The δ7Li of Geysir hot spring has also been isotopes were measured on pure Li fractions, using the Open measured and is 5.5‰. The range for sands and suspended sediments University Nu-Instruments MC-ICP-MS for waters and the BRGM is narrow (3.1‰–4.8‰) and close to the values estimated for Neptune MC-ICP-MS for sediments. A Cetac Aridus desolvating unweathered MORBs (e.g. but contrasts with the nebuliser was used with the Nu Instruments while a quartz dual values published by for other spray chamber and a low flow PFA microconcentric nebuliser were Iceland river sediments (−1.3‰ to 8.9‰).
used with the Neptune. More details concerning the analyticalprocedure are described in and in . In brief, analyses were performed on 2–3% HNO3 solutionswith Li concentrations of 30 ng/ml, typically yielding a current 5.1. Sources of dissolved Li intensity of 1.5 to 2 × 10−11 A for 7Li ion (using 1011 Ω resistors).
The total duration of data acquisition did not exceed 5–6 min, In order to constrain the controls on the Li isotope signatures of including sample uptake and peak centering. The analytical protocol Icelandic river waters, it is important to first determine the sources of employed involves the acquisition of 15 ratios with a 16 s integration dissolved Li. The Li contents of Icelandic rivers are relatively low time per ratio, and yields an in-run precision of 0.15‰ (2σm). The (0.02–2.2 μg/l) and even though these rivers mainly drain basalts, sample introduction rate is approximately 80–100 µl/min and the alternative Li sources must be considered, in particular hydrothermal total volume of sample used for each measurement is less than waters and atmospheric input.
600 µl, corresponding to about 15–20 ng of Li. Sample are bracketed The rivers studied here are mainly located away from the central by the L-SVEC RM 8545 standard in order to correct isotopic and active ridge where most of the hydrothermal activity is known to compositions for instrumental mass fractionation. The measured 7Li/ occur. Nevertheless, Li concentrations are several orders of magnitude 6Li ratio of a sample is normalized to the mean 7Li/6Li ratio of the two greater in hydrothermal waters than in river waters. In the ocean, Li standards run immediately before and after. In addition, during each contents of hydrothermal fluids range from 3000 to 9000 μg/l. In N. Vigier et al. / Earth and Planetary Science Letters 287 (2009) 434–441 Iceland, the Li content of geothermal waters ranges from 35.9 μg/ sources for these elements. In mixing diagrams, river waters all plot l to 6600 μg/l ( between the three endmembers considered: precipitation (or glacier- ). In the thermal waters of Geysir, which represents the most fed waters), hydrothermal springs and basalt weathering ).
famous, and one of the largest, hydrothermal fields of Iceland, two However, as discussed above, atmospheric inputs are likely to be aliquots sampled during different seasons yield Li contents of 389 μg/ small when compared to the amount of Li released by basaltic l and 360 μg/l. In contrast with Li concentrations, the δ7Li measured in lithologies during weathering. The waters possessing high δ7Li and K/ hydrothermal waters of mid-ocean ridge systems are relatively Li values are located in the older and drier parts of the island. These homogeneous, around 8.7‰ ± 1.4‰ (The δ7Li areas are characterized by a significant amount of clays and secondary measured at Geysir spring is 5.5‰. Taking these values into account, it minerals that have replaced primary minerals in the weathered lava can be calculated that a hydrothermal source contribution of less than flows. Li uptake by secondary minerals could also be responsible for 1% would impose similar δ7Li values for all the rivers studied here.
elevated K/Li, Na/Li and Mg/Li ratios and it is therefore not possible, Thus, the wide range of δ7Li values displayed by Iceland river waters a based on concentrations only, to distinguish between this effect and a priori suggests that the contribution from hydrothermal inputs must mixing process involving the endmembers described above ().
be significantly less than 1%.
Moreover, as illustrated in a and b, it is also difficult, based on It has previously been shown that melted ice from central glaciers concentrations alone, to distinguish between the role of basalt can be used as a proxy for atmospheric inputs in Iceland (e.g. dissolution and the potential impact of any hydrothermal springs ). The Li content of melted ice is relatively low (<6 ng/l), drained by these rivers.
consistent with the non-volatile nature of Li during low temperature Since the range in δ7Li values in river waters (10.1‰–23.8‰) lies evaporation processes. However, this is not negligible when com- between those of hydrothermal ( 9‰) and atmospheric inputs pared with the Li contents measured in some of the Icelandic rivers, (31.2‰), it might be argued that a mixing process involving these such as the Heioarvatn and Geithelnadalur Rivers (). It can be two sources could explain the entire range of isotope compositions reasonably assumed that most of the atmospheric Li comes from measured for these waters. In such a case, it would be very difficult to marine aerosols and that the δ7Li composition of any precipitation is infer precise information concerning low temperature basalt weath- close to the seawater value (31.2‰, ering rates from the Li isotope signatures of Icelandic rivers. However, In river waters, the highest δ7Li (>21‰) and K/Li values, illustrates that the δ7Li values in river waters cannot be closest to the composition of precipitation (), are located in the explained by simple binary mixing of these sources. Similarly, mixing area with the lowest runoff and where the percentages of glacial cover between atmospheric inputs and a solution resulting from basalt are the lowest at the basin scale (Consequently, the weathering does not explain the river δ7Li River δ7Li values high δ7Li values and K/Li ratios measured in these rivers cannot are systematically higher than any of the theoretical mixing lines. The simply be explained by a greater contribution from precipitation.
highest δ7Li values correspond to the oldest basins with the lowest The sources of Li in Icelandic rivers can be more precisely runoff and are therefore unlikely to result from a greater contribution constrained by comparing the concentrations of Li and major from the atmosphere. High δ7Li values, associated with high K/Li or elements in river waters with three determined endmembers (basalt Na/Li ratios, are more consistent with a Li uptake process, during the weathering, precipitation, hydrothermal exchange) (a,b,c). To a formation of secondary phases. shows that the initial solution first approximation, basalt weathering in Iceland can be assumed to must have low δ7Li and a high Li content. It might be envisioned that be congruent, as has been demonstrated for some of the youngest this process occurs in soils and that initial soil solutions would either Icelandic basins (e.g. leading to waters with K/Li, have isotopic compositions close to that of the source basalt, with or Mg/Li and Na/Li ratios similar to the average drained basalt without similar K/Li ratios, or would lie on the mixing curve with ). The Li contents of the rivers correlate with K, Na and Mg precipitation. These data also indicate that negligible isotope contents (r2 = 0.60–0.75, not shown here), suggesting identical fractionation occurs during basalt leaching, as has been suggested in Table 1Li isotope composition (δ7Li) of Icelandic river waters (w), suspended sediments (p) and bedload sands (s).
(t km−2 yr−1) Vestari Jokulsa, Goddolum Laxa at Myvatssveit Jökulsa a Fjöllum Jökulsa i Fljotsdal Age and chemical erosion rates are from a Mg, K and Na concentrations for waters are from b Sample taken at the output of the Myvatn lake.

N. Vigier et al. / Earth and Planetary Science Letters 287 (2009) 434–441 a previous study Indeed a leaching values. However, the main positive trend resulting from secondary process associated with preferential release of 7Li in waters (relative mineral formation does not appear to have been significantly disturbed to 6Li) would result in a negative trend on .
by hydrothermal processes, as expected given the sampling strategy A hydrothermal contribution to the rivers cannot be completely adopted in this study (see ).
ruled out, particularly for glacier fed rivers (Skaftafellsa, #23), whilehydrothermal springs may also be located along the course of other 5.2. Assessing the link between riverine δ7Li and silicate erosion rates rivers. Such a contribution, if significant, would decrease river δ7Li As illustrated in it is difficult to account for the δ7Li values measured in river waters by mixing processes alone. In the older partsof Iceland, high δ7Li values in rivers are associated with low Licontents. These older terrains are characterized by low chemicalerosion rates and basalt porosity has been significantly reduced by theprecipitation of secondary minerals (e.g. Incontrast, in the younger areas, chemical erosion rates are greater,corresponding to a more congruent weathering process, due inparticular to the higher susceptibility to weathering of glassyhyaloclastites. Taken together, these observations strongly suggest alink between the nature and the intensity of silicate chemical erosionand Li isotope signature in river waters. In fact, excepted the watersampled at the output of the Myvatn lake, a negative correlation canbe observed between δ7Li and chemical erosion rates estimated basedon major elements measured for the same samples , This is the first direct correlation between both parameters andconfirm the initial suggestion of , that the δ7Licomposition of river waters could yield precious information onchemical erosion rates, both past and present. An empirical lawlinking δ7Li and silicate chemical erosion rate (w) can be inferred fromthe data shown in : δ7Li = −4:8ðF1:8Þ · lnðwÞ + 36ðF5:9Þ If this law fits well the data for silicate chemical erosion rates lower than 100 t km− 2 yr− 1, i.e. for most of the basins world-wide(), it would need to be refined for higherweathering rate areas. In order to use this relationship for global scalebio-geochemical modelling, more data are needed in order to confirmif this law is representative of silicate chemical weathering at a largerscale, and in particular for other rock types and other climate regimes.
Indeed, if secondary mineral formation is the key process, reducing the levels of cation fluxes to the ocean and resulting in high river δ7Livalues, then quantitative application of this system necessitates thedetermination of Li isotope fractionation accompanying the formationof different Li rich secondary minerals formed during continentalweathering. However, as suggested by , Li isotopefractionation is expected to be broadly similar for all tri- and di-octahedral clays, and solution chemical composition has no effect onisotope fractionation during smectite formation. If this is confirmed,then the law described by Eq. (1) may be representative at a muchlarger scale than simply that of Iceland. It is instructive to note that, bycombining silicate weathering rates reported by and δ7Li values reported by , most of the largerivers, draining lithologies typical of continental crust, display highδ7Li and correspond to very low chemical erosion rates. The Amazon,Ganges, Orinoco, Lena, Huanghe, and the Qiantang Rivers all display(at their outlet) δ7Li values greater than 21‰ for corresponding Fig. 2. Dissolved K/Li versus Mg/Li, Na/Li (ppm/ppm) and the mean age of the drainedbasalts for the main Icelandic rivers. The mean compositions of atmospheric input(Atm.), hydrothermal springs (Hydr.) and drained basalts (Basalt) are also reported(data from analyses of melted Icelandic glacier ice, Pogge von Strandmann (pers.
comm.), and respectively). Dashed lines represent theoretical mixingcurves between hydrothermal springs and atmospheric inputs, and plain linesrepresent theoretical mixing curves between a water resulting from the congruentweathering of basalts and atmospheric inputs. Arrows represent Li uptake by secondaryminerals from a water resulting from the congruent weathering of basalts. The slope ofthese arrows and the slope defined by the data are close to unity, implying that when Liis taken up by clays, K, Na and Mg are similarly taken up. The triangle is for the glacierfed river Skaftafellsa (#23).

N. Vigier et al. / Earth and Planetary Science Letters 287 (2009) 434–441 described by Eq. (4), is calculated assuming a steady-state processsuch that the erosion products carried by the river reflect the sourcerock composition (e.g. δ7Lirock = fs · δs + fw · δw where δ7Lirock is constant and equivalent to the value of unweatheredMORB. This steady-state has recently been demonstrated in Icelandfrom U-series isotopes measured in rivers ), andfrom major and trace element data ().
As shown in , this simple model of Li uptake by secondary minerals can explain all the river data (waters and sediments) whenδ7Lisediment–water ranges between −1‰ and −7.5‰, and fw is less than 20%. The precise mechanisms responsible for the isotopic composi-tion of river waters and sediments are likely to be more complex thanthose proposed here, but this model presents an alternative to theleaching model proposed for explaining soil profiles and water data Fig. 3. K/Li versus (e.g. The degree of Li δ7Li measured in the dissolved phases of the main Icelandic rivers, with same end-members as in The black arrow represents the evolution of water isotope fractionation required to explain the data is significantly composition in the case of leaching from a basaltic rock, assuming no associated Li lower than the estimation of (−17‰) for a clay isotope fractionation. Mixing lines do not fit the river data: measured δ7Li are mixture composed of smectite, zeolite, palagonite and oxyhydroxides systematically larger than predicted by mixing processes.
formed at 4 °C on the ocean seafloor. However, isotope fractionationassociated with weathering in Iceland may be influenced by other chemical erosion rates lower than 14 t km−2 yr−1. Thus, most of the types of secondary minerals. Moreover adsorption processes and large rivers draining continental silicates plot on broadly the same coefficients are likely to be different in seawater and continental trend to that defined by the Icelandic rivers (see ). This may environments. Recent experimental study shows that the incorporation imply that Li isotopes do not fractionate during the leaching of Li rich of Li into the octahedral sites of a smectite at temperatures lower than mineral phases present on continents, in particular biotites. However, 90 °C is associated with Li isotope fractionations ranging between 8.8‰ this would need to be checked experimentally. In order to derive a and 10.9‰ In parallel, Li adsorption on smectite may more general law, it would be necessary to measure δ7Li on the same be associated with lower or negligible isotope fractionation ( water samples as those used for estimating chemical erosion rates and ). Further experiments are to perform systematic studies of granitic catchments.
needed to resolve the present lack of knowledge concerning theprocesses and phases that fractionate Li isotopes at low temperatures.
5.3. Modelling clay–water Li isotope fractionation Overall, the Li isotope composition of river waters appears to be related to chemical erosion rates and secondary mineral phase On a regional scale, the δ7Li values of the dissolved phase in formation. Several studies have shown that the stability of secondary Icelandic rivers are highly variable. In contrast, the corresponding phases in Iceland is dependent upon a number of interrelated factors, suspended and bedload sediments display much more homogeneous the principal of which are elemental supply via leaching, water δ7Li values, close to the composition of unweathered MORB. Wheredata for both phases are available, such a relationship, betweendissolved and sediment loads has been observed for most rivers todate, in particular for Himalayan catchments and the Orinoco basinA simple Rayleighfractionation model can explain the relationship between thesephases, taking into account both the steady-state nature of theerosion processes in Iceland and clay formation. Li is first leached fromthe basaltic source rocks (e.g. in soils) without significant isotopefractionation, resulting in a dissolved phase with the same δ7Li as theunweathered parent rock (i.e. δ7Li ≈ 4‰). When secondary mineralsare formed, they incorporate a significant fraction of the dissolved Li.
These secondary minerals possess a light δ7Li isotope composition,thus their formation will both deplete the residual waters in 6Li, anddrive their δ7Li values to heavier compositions. It is possible to verify ifsuch a single process could produce both water and sediment riverdata, using the following the equations: δw = δiw + 1000:ðα−1Þ · ln fw δs = δw + 1000 · ln α s = ½ðδiw 1000 + 1Þ · ð f α 1Þ = ð fw 1Þ−1 · 1000 Fig. 4. Chemical erosion rates estimated from major elements in Icelandic river watersversus δ7Li measured in the same waters (open symbols) and in the corresponding river sediments (black symbols). The lake sample (#14) is not w and δs are the δ7Li of the dissolved phase and sediment represented. The curve and equation represent the best fit for the Icelandic water respectively, fw is the fraction of residual dissolved Li, and α is the data. The published values for river waters draining large areas of continental crust are clay–water Li isotopic fractionation factor (δ7Liclay–water =1000ln(α)).
shown for comparison (shaded field) (Amazon, Orenoque, Ganges, Lena, Huanghe, The composition of the cumulated or "bulk" sediment (δ–s), as Qiantang; ) (see text for more details).

N. Vigier et al. / Earth and Planetary Science Letters 287 (2009) 434–441 global scale and may be of particular use for estimating past changesin continental weathering. Further data are now needed in order todetermine if this law can be applied to different rock types anddifferent climatic regimes.
We would particularly like to thank Christian France-Lanord (CRPG) for fruitful discussion, and Peter Van Calsteren, Louise Thomasand Mabsie Gilmour (OU) for their technical help during wateranalyses. This project has been funded by the Open University and bythe "Reliefs de la Terrre" INSU-CNRS program.
Arnalds, O., 2004. Volcanic soils of Iceland. CATENA 56 (1–3), 3–20.
Arnorsson, S., Gunnarsson, I., Stefansson, A., Andresdottir, A., Sveibjornsdottir, A., 2002.
Major element chemistry of surface- and ground waters in basaltic terrain, N-Iceland. I. Primary mineral saturation. Geochim. Cosmochim. Acta 66 (23), Fig. 5. δ7Li of the dissolved (δ7Liw) and sediment (δ7Lip) loads of the main Icelandic rivers. The curves represent Rayleigh evolution in the case of Li uptake from an initial Carignan, J., Cardinal, D., Eisenhauer, A., Galy, A., Rehkämper, M., Wombacher, F., Vigier, solution with the same isotopic composition of the source basalts (see Eqs. (2)–(5) in N., 2004. A reflection on Mg, Cd, Ca, Li and Si isotopic measurements and related text), and clay–water Li isotope fractionation of (a) −7.5‰, (b) −5‰ and (c) −1‰.
reference materials. Geostand. Geoanal. Res. 28, 139–148.
Carignan, J., Vigier, N., Millot, R., 2007. Three secondary reference materials for Li isotope measurements: Li7-N, Li6-N and LiCl-N solutions. Geostand. Geoanal. Res.
temperature, pH and residence time. For most of the samples studied here, taken from the lower reaches of their respective catchments, a Chan, L.-H., Edmond, J.M., 1988. Variation of lithium isotope composition in the marine steady state is likely to have been achieved such that soil residence environment: a preliminary report. Geochim. Cosmochim. Acta 52, 1711–1717.
Chan, L.-H., Edmond, J.M., Thompson, G., Gillis, K., 1992. Lithium isotopic composition of times are constant and clay formation is compensated for by chemical submarine basalts: implications for the lithium cycle to the ocean. Earth Planet. Sci.
and physical erosion. However, a steady state may not always be Lett. 108, 151–160.
attained, for example in glacier-fed rivers (such as the Skaftafellsa Chan, L.-H., Edmond, J.M., Thompson, 1993. A lithium isotope study of hot springs and metabasalts from mid ocean ridge hydrothermal systems. J. Geophys. Res. 98, River (#23)) or in groundwaters, where high pH results as much from the isolation from atmospheric CO2, as from chemical weathering. In Chan, L.-H., Gieskes, J.M., You, C.F., Edmond, J.M., 1994. Lithium isotope geochemistry of such cases, these may be no simple relationship with chemical sediments and hydrothermal fluids of the Gaymas Basin, Gulf of California.
Geochim. Cosmochim. Acta 58, 4443–4454.
weathering rates, even if secondary minerals such as Fe–Mg–Ca Chan, L.-H., Alt, J.C., Teagle, D.A.H., 2002. Lithium and lithium isotope profile through smectites and zeolites are stable.
the upper oceanic crust: a study of seawater–basalt exchange at ODP Sites 504B These observations indicate that the model presented in this study and 896A. Earth Planet. Sci. Lett. 201, 197–201.
Elliott, T., Thomas, A., Jeffcoate, A., Niu, A., 2006. Lithium isotope evidence for enables the link between secondary mineral formation, chemical subduction enriched mantle in the source of mid-ocean-ridge basalts. Nature 443, erosion rates and the Li isotope composition of residual waters to be quantified, and the data also suggest that secondary mineral Eiriksdottir, E.S., Louvat, P., Gislason, S.R., Óskarsson, N., Hardardóttir, J., 2008. Temporal formation acts as a key control on the cation flux to the ocean. The variation of chemical and mechanical weathering in NE Iceland: evaluation of asteady-state model of erosion. Earth Planet. Sci. Lett. 272, 78–88.
relationship between water chemistry and secondary mineral Gaillardet, J., Dupré, B., Louvat, P., Allègre, C.J., 1999. Global silicate weathering and CO2 formation is consistent with the U-series data from the same rivers, consumption rates deduced from the chemistry of large rivers. Chem. Geol. 159, where soil development and sediment residence time have also been Gannoun, A., Burton, K.W., Vigier, N., Gislason, S.R., Rogers, N., Mokadem, F., Sigfusson, shown to be a major influence on chemical erosion and river water B., 2006. The influence of weathering process on riverin osmium isotopes in a basaltic terrain. Earth Planet. Sci. Lett. 243, 732–748.
Georg, R.B., Reynolds, B.C., West, A.J., Burton, K.W., Halliday, A.N., 2007. Silicon isotope variations accompanying basalt weathering in Iceland. Earth Planet. Sci. Lett. 261, 6. Summary and conclusion Gislason, S.R., Eugster, H.P., 1987a. Meteoric water–basalt interactions. II: a field study The Li isotope compositions of the dissolved phase and suspended in N.E. Iceland. Geochim. Cosmochim. Acta 51 (10), 2841–2855.
Gislason, S.R., Eugster, H.P., 1987b. Meteoric water–basalt interactions. I: a laboratory and bedload sediments of the main Icelandic rivers have been study. Geochim. Cosmochim. Acta 51 (10), 2827–2840.
determined by MC-ICP-MS. For the dissolved phase, δ7Li range Gislason, S.R., Arnorsson, S., Armannsson, H., 1996. Chemical weathering of basalt in between 10.1‰ and 23.8‰, while river sediments display a narrower southwest Iceland: effects of runoff, age of rocks and vegetative/glacial cover. Am. J.
Sci. 296, 837–907.
range of values (3.1‰–4.8‰), close to that of unweathered MORB.
Gislason, S.R., Oelkers, E.H., Snorrason, A., 2006. Role of river-suspended material in the High δ7Li values are associated with high K/Li, Na/Li and Mg/Li ratios, global carbon cycle. Geology 34 (1), 49–52.
and characterise the waters draining mainly old and weathered Henley, R.W., Ellis, A.J., 1983. Geothermal systems ancient and modern: a geochemical review. Earth Sci. Rev. 19, 1–50.
basalts. Rivers with low δ7Li are located in the younger parts of the Huh, Y., Chan, L.-H., Zhang, L., Edmond, J.M., 1998. Lithium and its isotopes in major island. It is shown that source-mixing alone, between precipitation, world rivers: implications for weathering and the oceanic budget. Geochim.
Li-rich hydrothermal springs and basalt weathering, cannot explain Cosmochim. Acta 62, 2039–2051.
the entire range of δ7Li values. Instead, the Li isotope compositions of Huh, Y., Chan, L.-H., Edmond, J.M., 2001. Lithium isotopes as a probe of weathering processes: Orinoco River. Earth Planet. Sci. Lett. 194, 189–199.
both waters and sediments can be explained by a steady-state erosion Huh, Y., Chan, L.-H., Chadwick, O., 2004. Behavior of lithium and its isotopes during process where the formation of secondary minerals is associated with weathering of Hawaiian basalts. Geochem. Geophys. Geosys. 5. doi:10.1029/ significant Li isotope fractionation. The data obtained in this study, James, R.H., Palmer, M., 2000. Marine geochemical cycles of the alkali elements and combined with chemical erosion rates previously determined on the boron: the role of sediments. Geochim. Cosmochim. Acta 64, 3111–3122.
same samples provide the first direct evidence for a relationship James, R.H., Rudnicki, M.D., Plamer, M., 1999. The alkali element and boron between δ7Li and chemical erosion rates. An empirical law between geochemistry of the Escanaba Trough sediment-hosted hydrothermal system.
Earth Planet. Sci. Lett. 171, 157–169.
chemical erosion rate and δ7Li can be defined, and comparison with James, R.H., Allen, D.E., Seyfried, Jr, 2003. An experimental study of oceanic crust and literature data suggests that this law may be applicable at a more terrigenous sediments at moderate temperatures (51 to 350 °C): insights as to N. Vigier et al. / Earth and Planetary Science Letters 287 (2009) 434–441 chemical processes in near-shore ridge-flank hydrothermal systems. Geochim.
Stefansson, A., Gislason, G.R., 2001. Chemical weathering of basalts, SW Iceland: effect Cosmochim. Acta 67, 681–691.
of rock crystallinity and secondary minerals on chemical fluxes to the ocean. Am. J.
Jeffcoate, A., Elliott, T., Thomas, A., Bouman, C., 2004. Precise, small sample size Sci. 6 (513–556), 2001.
determinations of lithium compositions of geological reference materials and Taylor, S.R., Urey, H.C., 1938. Fractionation of the lithium and potassium isotopes by modern seawater by MC-ICPMS. Geostand. Geoanal. Res. 28, 161–172.
chemical exchange with zeolites. J. Chem. Phys. 6, 429–438.
Kisakürek, B., Widdowson, M., James, R.H., 2004. Behaviour of Li isotopes during Tomasson, H., 1986. Glacial and volcanic shore interactions. Part 1: on land. In: continental weathering: the Bidar laterite profile, India. Chem. Geol. 212, 27–44.
Sigbjarnarson, G. (Ed.), Iceland Coastal and River Symposium Proceedings.
Kisakürek, B., James, R.H., Harris, N.B.W., 2005. Li and δ7Li in Himalayan rivers: proxies University of Iceland, Reykjavik, pp. 7–16.
for silicate weathering? Earth Planet. Sci. Lett. 237, 387–401.
Vigier, N., Bourdon, B., Turner, S., Allègre, C.J., 2001. Erosion timescales derived from U- Louvat, P., 1997. Etude géochimique de l'érosion fluviale d'îles volcaniques à l'aide des decay series measurements in rivers. EPSL 193, 549–563.
bilans d'éléments majeurs et traces, Doctorat Thesis, IPGParis.
Vigier, N., Burton, K.W., Gislason, S.R., Rogers, N.W., Duchene, S., Thomas, L., Hodge, E., Millot, R., Guerrot, C., Vigier, N., 2004. Accurate and high-precision measurement of Schaefer, B., 2006. The relationship between riverine U-series disequilibria and lithium isotopes in two reference materials by MC-ICP-MS. Geostand. Geoanal. Res.
erosion rates in a basaltic terrain. Earth Planet. Sci. Lett. 249, 258–273.
28, 153–159.
Vigier, N., Decarreau, A., Millot, R., Carignan, J., Petit, S., France-Lanord, C., 2008.
Milliman, J.D., Meade, R.H., 1983. World-wide delivery of river sediment to the oceans.
Quantifying Li isotope fractionation during smectite formation and implications for J. Geol. 91, 1–21.
the Li cycle. Geochim. Cosmochim. Acta 72, 780–792.
Négrel, P., Allègre, C.J., Dupré, B., Lewin, E., 1993. Erosion sources determined by Williams, L.B., Hervig, R.L., 2005. Lithium and boron isotopes in illite–smectite: the inversion of major and trace element ratios and strontium isotopic ratios in river: importance of crystal size. Geochim. Cosmochim. Acta 69, 5705–5716.
the Congo Basin case. Earth Planet. Sci. Lett. 12, 59–76.
Wolff-Boenisch, D., Gislason, S.R., Oelkers, E.H., 2006. The effect of crystallinity on Palsson, S., Vigfusson, G.H., 1991. Nidurstödur svifaurswmaelinga 1963–1990, dissolution rates and CO2 consumption capacity of silicates. Geochim. Cosmochim.
Acta 70, 858–870.
Pistiner, J.S., Henderson, G., 2003. Lithium isotope fractionation during continental Yeghicheyan, D., Carignan, J., Valladon, M., Bouhnik Le Coz, M., Le Cornec, F., Castrec- weathering processes. Earth Planet. Sci. Lett. 214, 327–339.
Rouelle, M., Robert, M., Aquilina, L., Aubry, E., Churlaud, C., Dia, A., Deberdt, S., Pogge von Strandmann, P.A.E., Burton, K.W., James, R.H., van Calsteren, P., Gislason, S.R., Dupré, B., Freydier, R., Gruau, G., Hénin, O., de Kersabiec, A.-M., Macé, J., Marin, L., Mokadem, F., 2006. Riverine behaviour of uranium and lithium isotopes in an Morin, N., Petitjean, P., Serrat, E., 2001. A compilation of silicon and thirty one trace actively glaciated basaltic terrain. Earth Planet. Sci. Lett. 251, 134–147.
elements measured in the natural river water reference material SLRS-4 (NRC– Rudnik, R.L., Tomascak, P.B., Njo, H.B., Gardner, L.R., 2004. Extreme lithium isotopic CNRC). Geostand. Newsl. Geost. Geoanal. Res. 25, 465–474.
fractionation during continental weathering revealed in saprolites from South Zhang, L., Chan, L.H., Gieskes, J.M., 1998. Lithium isotope geochemistry of pore waters Carolina. Chem. Geol. 212, 45–57.
from Ocean Drilling Program Sites 918 and 919, Irminger Basin. Geochim.
Seyfried Jr., W.E., Chen, X., Chan, L.-H., 1998. Trace element mobility and lithium isotope Cosmochim. Acta 62, 2437–2450.
exchange during hydrothermal alteration of seafloor weathered basalt: anexperimental study at 350 °C, 500 bars. Geochim. Cosmochim. Acta 62, 949–960.

Source: http://www.crpg.cnrs-nancy.fr/spip.php?action=acceder_document&arg=223&cle=6817cf05d7afe4dc41475845d4009033a0caa575&file=pdf%2F2025.pdf