Mechanisms of magnesium isotope fractionation in volcanic soil weathering sequences, Guadeloupe

Earth and Planetary Science Letters 341-344 (2012) 176–185

Contents lists available at SciVerse ScienceDirect

Earth and Planetary Science Letters journal homepage: www.elsevier.com/locate/epsl

Letters

Mechanisms of magnesium isotope fractionation in volcanic soil weathering sequences, Guadeloupe S. Opfergelt a,b,n, R.B. Georg c, B. Delvaux b, Y.-M. Cabidoche d, K.W. Burton a,1, A.N. Halliday a a

Department of Earth Sciences, University of Oxford, South Parks Road, Oxford OX1 3AN, Oxford, United Kingdom Earth and Life Institute, Universite´ catholique de Louvain, Croix du Sud 2 bte L7.05.10, 1348 Louvain-la-Neuve, Belgium c Trent University, Water Quality Centre, 1600 West Bank Dr., Peterborough, Ontario, Canada d INRA, UR 1321 ASTRO Tropical Agrosystems, Petit Bourg, Guadeloupe b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 March 2012 Received in revised form 22 May 2012 Accepted 7 June 2012 Editor: J. Lynch-Stieglitz Available online 11 July 2012

Magnesium (Mg) stable isotopes are increasingly used as a weathering proxy in soils and rivers, but the impact of the mineralogy of secondary phases on isotope fractionation remains obscure. A better understanding of the behaviour of Mg isotopes during weathering processes is a mandatory step toward deployment of this new tracer for understanding chemical fluxes exported from the critical zone. Here we investigate isotopic variations in d26Mg in bulk soils and clay fractions relative to their parent andesite in three soil weathering sequences from Guadeloupe formed under contrasting climatic conditions. Soils formed in drier conditions (low precipitation) contain smectite, whereas soils formed under wet conditions (high rainfall) are characterized by halloysite and Fe-oxides or kaolinite. All clay fractions have Mg isotopic compositions (d26Mg  0.41% to  0.10%) similar to or heavier than their parent andesite (d26Mg  0.47%) supporting the preferential incorporation of heavy Mg isotopes in secondary Mg-bearing clay minerals with the first direct measurements on clay fractions. Soils with lighter Mg isotope compositions have greater quantities of exchangeable Mg. The data support a contribution from sea spray to the exchangeable Mg pool correlated to the soil weathering degree. This study highlights for the first time that the soil d26Mg not only depend on d26Mg of the parent rock, and on any fractionation that might occur, but also on the Mg retention on the exchange complex, which could in turn be controlled by external inputs such as sea spray. & 2012 Elsevier B.V. All rights reserved.

Keywords: andesite weathering volcanic soils magnesium isotopes clay minerals exchangeable magnesium Guadeloupe

1. Introduction The chemical weathering of continental Ca–Mg silicate rocks influences global climate by consuming atmospheric CO2 (Berner, 1995). Chemical weathering of the host regolith dominantly occurs at the soil–rock interface (White et al., 1998; Brantley, 2010). Therefore, investigating the processes responsible for elemental transfers in the critical zone is fundamental for improving estimates for continent derived input fluxes to the ocean and thus the long-term atmospheric C-budget. Magnesium (Mg) is the eighth most abundant element in the continental crust (Taylor and Mclennan, 1985) and the fourth most abundant species in seawater (Millero, 1974). It is transferred from the continents to the oceans via rivers. During weathering reactions,

n Corresponding author at: Earth and Life Institute, Universite´ catholique de Louvain, Croix du Sud 2 bte L7.05.10, 1348 Louvain-la-Neuve, Belgium. Tel.: þ32 10 47 36 38; fax: þ32 10 47 45 25. E-mail address: [email protected] (S. Opfergelt). 1 Now at Department of Earth Sciences, Durham University, Durham DH1 3LE, United Kingdom

0012-821X/$ - see front matter & 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.epsl.2012.06.010

Mg derived from dissolution of primary minerals may be incorporated into secondary minerals (Borchardt, 1989). Magnesium is also retained on the soil exchange complex, the loss from which constitutes an important source of cations to rivers (Markewitz et al., 2001), representing up to 30% of the annual cation export (Miller et al., 1993). Crucially, Mg is also an essential plant nutrient used for the synthesis of chlorophyll (Epstein and Bloom, 2005), and is released back to the soil via litterfall degradation. The recent advent of high-precision measurement of stable Mg isotopes offers an unprecedented opportunity to obtain new insights into the sources and processes that control the release of chemical elements from the critical zone and therefore, for better understanding of riverine fluxes (Gaillardet et al., 2010). Magnesium isotopes (24Mg, 25Mg, 26Mg) are fractionated during uptake by plants (Black et al., 2008; Bolou-Bi et al., 2010), and through incorporation into secondary silicate minerals. One current view is that heavy Mg isotopes are incorporated into secondary silicate phases, producing lighter riverine signatures in silicate catchments (De Villiers et al., 2005; Brenot et al., 2008; Tipper et al., 2008a; Wimpenny et al., 2011). This is based on isotope measurements of bulk soils that were found to be

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

isotopically heavier than their parent silicate bedrock (Teng et al., 2010; Tipper et al. 2006a), or of detrital smectite Mg heavier than the bulk silicate Earth (Tipper et al., 2010). In contrast, it has also been suggested from an experimental study that fluids can become heavier following precipitation of secondary phases (Wimpenny et al., 2010), although the Mg isotope composition of the secondary phase was not obtained in latter study, leaving some uncertainty. Also, one soil derived from basalt is reported to be isotopically lighter than the basalt (Pogge Von Strandmann et al., 2008). Thus far, the mechanisms controlling Mg retention in the soil and the secondary phases driving the isotope fractionation have not been identified. Here, a direct comparison between Mg isotope compositions in secondary clay minerals and their parent silicate material should provide new insights to better constrain the behaviour of Mg isotopes in weathering processes. Weathering sequences of soils developed on volcanic ash in humid tropical regions are well suited for studying how the degree of weathering affects soil constituents and nutrient distribution (Chadwick et al. 2003, 1999). Volcanic arc settings such as Guadeloupe, characterized by high precipitation rates, are amongst the fastest eroding regions on Earth (Gaillardet et al., 2011). Better understanding silicate weathering processes in these regions is a mandatory step to refine estimates of their CO2 consumption rates and the impact of weathering on the global C budget. In soils from Guadeloupe, secondary Mg-bearing phases such as smectite and hydroxy-Al interlayered smectite/ hydroxy-Al interlayered vermiculite (HIS/HIV) have been identified (Colmet-Daage and Lagache, 1965) and were formed from a typical local andesite (Samper et al., 2007; Sak et al., 2010). Here we investigate Mg isotope variations in bulk soils and clay fractions relative to the parent andesite in three soil weathering sequences from Guadeloupe in order to provide a direct in situ measurement of the fractionation as a function of the degree and detailed processes of weathering.

2. Environmental setting The Guadeloupe archipelago is located in the French Indies of the Caribbean (161N, 611W). The two main islands of Guadeloupe are Basse-Terre and Grande-Terre. Basse-Terre is a volcanic island dominated by the volcano La Soufrie re (1467 m above sea level (asl)). Volcanic ash soils cover extensive areas on the slopes surrounding the volcano, developed from andesitic ash deposits of Pliocene to Holocene age, dominated by plagioclase, pyroxene and ferromagnesian volcanic glasses (Dagain et al., 1981; Ndayiragije, 1996; Ndayiragije and Delvaux, 2003). Grande Terre comprises a coral basement covered by volcanic pyroclasts contemporary to the volcanic deposits from the north of Basse-Terre. Contrasting climatic conditions characterise the western (Ws), eastern (Es) and northern (Ns) slopes of la Soufrie re volcano (Fig. 1). The eastern slopes are exposed to rain-bearing Northeast trades from the Atlantic Ocean, with a mean annual rainfall (MAR) exceeding 5000 mm at 500 m asl, and decreasing to 2500 mm downslope at sea level (Chaperon et al., 1985). The western slopes are drier with MAR decreasing from 3200 mm at 450 m asl to 850 mm at sea level. On the northeastern slopes, annual rainfall decreases from 4000 mm on the northeastern slope tops of BasseTerre to 1200 mm on the eastern coast of Grande-Terre. The age of the soils increases from Ws, which has young soils containing a reserve of weatherable primary minerals, to older soils in Es (103–104 yr), to the oldest soils from Ns where weathering has exhausted the pool of primary minerals (105 yr) (Cabidoche et al., 2009). The distribution of soils in Guadeloupe with contrasting degree of weathering and clay mineralogy is related to climate and to the

177

age of the soils as chrono-climo-topo-sequences (Colmet-Daage and Lagache, 1965; Ndayiragije, 1996; Ndayiragije and Delvaux, 2003, 2004). This is typical in the South of Basse-Terre for soils from Ws and Es (Fig. 2). The Ws soil pattern developed in dry conditions involves a weathering sequence Andosol–Cambisol corresponding to the mineralogical sequence ash-allophanehalloysite/smectite (Colmet-Daage and Lagache, 1965; PinerosGarcet, 1994). The Es soil pattern developed in wetter conditions at the same elevation and produced a weathering sequence Perhydrated Andosol–Andosol–Nitisol, corresponding to the mineralogical sequence ash-gibbsite, allophane-halloysite, Fe-oxide (Colmet-Daage and Lagache, 1965; Ndayiragije, 1996; Ndayiragije and Delvaux, 2003, 2004). In the North of Basse-Terre and Grande Terre, the Ns soil pattern produced a weathering climo-sequence Ferralsol–Vertisol from wet to dry conditions yielding the mineralogical sequence kaolinite, Fe-oxide-smectite.

3. Methods 3.1. Sampling Soil samples representative of the contrasting chrono-climotopo-sequences in Guadeloupe were collected during fieldwork in January 2006 (Henriet et al., 2008a, 2008b). A total of seven sites from three soil weathering sequences were selected, including Andosol–Cambisol (Ws), Perhydrated Andosol–Andosol–Nitisol (Es), Ferralsol–Vertisol (Ns) (Figs. 1 and 2). The topsoil (0–20 cm) was collected in each site. A fresh rock sample representative of the protolith andesite was collected in November 2009 from the Piton Tarade, the local basement of the La Soufrie re volcano (altitude 1090 m asl) (Fig. 1). The soils were sampled from sites that have been used for intensive banana cropping for a long time (Colmet-Daage and Lagache, 1965; Ndayiragije, 1996; Dorel et al., 2000). The agricultural system is based on monoculture with successive cropping cycles of 7–9 months. At harvesting time, leaves and pseudostems are cut and left on the soil surface, which significantly limits the mineral export by the vegetation. Before plantation and in a non-recurring process, lime (coral limestone or dolomite stone, 1T/ha; Sansoulet, 2007) is applied to the banana fields for fertilisation. To avoid a direct bias from this initial lime treatment we have chosen sampling sites that are located on  10 yr old plantations, where lime vanished over the years, as supported by similar pH values in the sampled soils (soil pHH2O ¼6.270.5; Henriet et al., 2008b) relative to pH values measured in uncultivated soils from the same areas in Guadeloupe (soil pHH2O ¼5.570.6; Pineros-Garcet, 1994; Ndayiragije, 1996). 3.2. Soil characterisation The soils samples were characterised in a previous study (Table 1; Henriet et al., 2008b). Briefly, soil samples were airdried and sieved at 2 mm to recover the fraction representative for the bulk soil ( o2 mm). The recovery of clay fractions ( o2 mm) was achieved after sonication and sieving at 50 mm to remove sand fractions, and prolonged dispersion with Na þ -resin (Rouiller et al., 1972). Clay ( o2 mm) was separated from silt (2– 50 mm) by successive 24 h cycles of decanting and pipetting in deionised water. Clay fraction mineralogy was assessed by XRD after K þ and Mg2 þ saturation, ethylene glycol solvation and thermal treatments at 300 and 550 1C (Robert and Tessier, 1974). Bulk soils, clay fractions and the andesite sample were analysed for major elements by ICP-AES (Universite´ catholique de Louvain, Belgium) after Li borate fusion at 1000 1C (Chao and

178

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

Fig. 1. Location of sampling sites in Guadeloupe (Basse-Terre and Grande-Terre) with isohyets (mm yr  1; adapted from Chaperon et al., 1985). Parent andesite (triangle)—soils from the western slope sequence: Andosol (Ws-An), Cambisol (Ws-Ca)—soils from the eastern slope sequence: Perhydrated Andosol (Es-An1), Andosol (Es-An2), Nitisol (Es-Ni)—soils from the northern slope sequence: Ferralsol (Ns-Fe), Vertisol (Ns-Verti). The transect A-B is represented in Fig. 2.

Fig. 2. Schematic transect (A–B; represented on Fig. 1) of Western (Ws) and Eastern (Es) rainfall pattern (mm) in relation to elevation (m above sea level (asl)) and major soil types in the island of Basse-Terre, Guadeloupe. Arrows indicate the location of experimental sites from the western and eastern slope sequences (legend as in Fig. 1). Soils from the northern slope sequence are not represented on this transect (adapted from Colmet-Daage and Lagache, 1965; Chaperon et al., 1985; Henriet et al., 2008b).

Sanzolone, 1992). Exchangeable cations in clay fractions were exchanged for by strontium (Sr) using a SrCl2 solution for ion exchange (Hinckley and Bates, 1960) prior to elemental analysis in clay fractions. The exchangeable Mg content was measured on

bulk soils by percolation of ammonium acetate 1 N at pH7 (Page et al., 1982). Bulk soil samples and andesite were also analysed for trace elements after dissolution in HF:HNO3 acid mixture for trace element

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

179

Table 1 Major characteristics of soils from Guadeloupe and of the parent andesite material: clay content, total Mg content in bulk soils (Mgsoil o 2 mm) and clay fractions (Mgclay o2 mm), exchangeable Mg (Mgexch), and clay mineralogy. Western slope sequence (Ws), eastern slope sequence (Es) and northern slope sequence (Ns). Soil data from Henriet et al. (2008b). Name

Soil typea

Location

MARb (mm yr  1)

Altitude (m)

Clay content Mgsoil (%) (g kg  1)

Mgclay (g kg  1)

Mgexch (cmolc kg  1)

Clay mineralogy

Ws-An Ws-Ca Es-An1 Es-An2 Es-Ni Ns-Fe Ns-Verti

Andosol Cambisol P. Andosolc Andosol Nitisol Ferralsol Vertisol

Grand Marigot Belle Vue Moı¨se Neufchateau Changy Feneteau Saint-Julien

3444 1931 4764 3916 2549 3295 730

470 222 424 274 32 164 25

41.5 34.9 59.1 52.3 80.7 81.6 85.0

10.9 9.8 7.9 8.8 1.9 1.6 6.7

2.52 2.32 2.63 2.63 1.18 1.44 6.68

1.2 4.2 4.9 2.3 1.4 2.2 6.6

Allophane Halloysite, smectite Gibbsite, allophane, HIS/HIVd Gibbsite, allophane, HIS/HIVd Halloysite, Fe-oxide Kaolinite, Fe-oxide, gibbsite, HIS/HIVd Smectite

1090

–

22.6

-

-

-

Andesite

Piton Tarade

a

IUSS, 2006. Mean annual rainfall. c Perhydrated Andosol. d Hydroxy-Al interlayered smectite (HIS) and hydroxy-Al interlayered vermiculite (HIV). b

analysis. Trace element contents (including Zr) were measured by quadrupole ICP-MS (Open University, UK) in 2% HNO3. The accuracy was assessed by using a water reference material SLRS-4 (Yeghicheyan et al., 2001). The analytical precision is 77%, with a detection limit better than 0.01 mM. 3.3. Magnesium isotope analysis 3.3.1. Sample preparation and chromatography Magnesium isotope compositions were measured on parent andesite, bulk soils ( o2 mm) and clay fractions ( o2 mm). Exchangeable cations in clay fractions were replaced by strontium (Sr) prior to Mg isotope analysis using a SrCl2 solution for ion exchange (Hinckley and Bates, 1960), which prevents a contribution of Mg from the soil exchange complex and is considered to provide a sample with Mg mostly included in mineral structure. Prior to Mg isotope measurement, soil samples were ashed at 450 1C to remove organic matter. Then  100 mg of silicate sample powder were dissolved by acid digestion in a HF:HNO3 mixture (4:1 volume ratio). Dissolved samples were purified for isotopic measurements by ion-chromatography on a cation exchange resin with a quantitative recovery of 498%. The chromatographic separation was adapted from that of Wombacher et al. (2009) for Biorad poly-prep columns (10 ml reservoir) with a 1.2 ml resin bed of BioRad AG50W-X8 (200–400 mesh). Typically 10 mg Mg is loaded onto the column for each sample. The separation is divided in two steps. The first step removes most of the Fe and Ca using 10 M HCl as eluent. The second step separates Mg from the rest of the matrix with successively 0.4 M HCl, 0.15 M HF, 95% acetone/0.5 M HCl, 1 M HCl as eluents. The eluent volumes for Step 1 and Step 2 and the corresponding elution curves are presented in Suppl. Mat. 1. For most of the bulk soils, and clay fractions highly rich in Fe from Feoxides, Step 1 was repeated several times. The purity of samples after column chemistry was confirmed by ICP-MS (University of Oxford, UK, 72%), and column chemistry was repeated if necessary to obtain a ratio of the concentration of any cation to that of Mg below 0.05 (Teng et al., 2007). The procedural blank was 8 ng, comparable with that of Teng et al. (2007). 3.3.2. Mass spectrometry Magnesium isotope ratios were determined by MC-ICP-MS (Nu Plasma HR, University of Oxford, UK) in dry plasma mode and low resolution (Dm/m  400). Samples containing 100 ng/g Mg were introduced into an ARIDUS-2 desolvation unit via an ESI PFA

nebulizer at 100 ml/min flow rate. The instrumental mass bias was corrected for by the sample-standard bracketing technique, and data are expressed in relative deviations of 26Mg/24Mg ratios from DSM3 standard using the common d-notation (%):[(26Mg/24Mg)sample/(26Mg/24Mg)DSM3  1]  1000, with an external reproducibility for DSM3 standard (Galy et al., 2003) of 70.15%, 2SD, for d26Mg. Each sample was analysed 9 times, where each single d-value (n) represents one sample run and two bracketed standard runs.

3.3.3. Precision and accuracy Long term precision and accuracy of the MC-ICP-MS measurements was assessed from multiple measurements over 18 months on the reference materials Cambridge-1 with a d26Mg of  2.6070.15% (2SD, n ¼576), and DSM-3 with d26Mg of  0.01 70.08% (2SD, n¼ 52) (Galy et al., 2003). Cambridge-1 was also processed though chemistry and gave the same d26Mg signature ( 2.59 70.11%). In addition, two Mg in-house standards, Mg Specpure and Mg OXLFG, were repeatedly measured over 18 months as reference materials. The Mg Specpure is an ICP-MS standard mono-elemental solution from Alfa Aesar (stock solution 1000 ppm Mg). The Mg OXLFG is a Mg oxide from Merck dissolved in Milli-Q water (stock solution 1230 ppm Mg in 2 N HNO3), from which the purity was checked by HR-ICP-MS (Thermo Fisher Element 2, University of Oxford, UK). The Mg isotope compositions of Mg Specpure and Mg OXLFG are d26Mg  2.85 70.15% (2SD, n¼153) and 0.3170.16% (2SD, n ¼119), respectively (Table 2). The chemical purification of samples, especially with complex matrices such as silicates, has been validated to ensure that there is no residual analytical artefact. The method was validated on the USGS rock standards (Table 2): BHVO-2 with d26Mg of  0.3170.19% (2SD, n¼30), BCR-2 with d26Mg of  0.3070.19% (2SD, n¼31), AGV-2 with d26Mg of -0.2470.24% (2SD, n¼28), BIR-1a with d26Mg of -0.2770.33% (2SD, n¼14), GSP-2 with d26Mg of -0.0270.31% (2SD, n¼14). These values are in agreement with published values (Baker et al., 2005; Bizzarro et al., 2005; Teng et al., 2007; Wiechert and Halliday, 2007; Tipper et al., 2008b; Pogge Von Strandmann et al., 2008; Wombacher et al., 2009; Huang et al., 2009; Bolou-Bi et al., 2009; Bourdon et al., 2010) (Suppl. Mat. 2). The method was also validated using a standard addition technique adapted to isotope ratios (following Tipper et al., 2008b) on the USGS basalt standard BCR-2 and on a basaltic soil sample from Iceland (Table 2), which is fully described in Suppl. Mat. 3.

180

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

Table 2 Magnesium isotope compositions (%) of reference materials and USGS rock standards, of standard-sample mixtures from the standard addition procedure, of parent andesite, bulk soils and clay fractions (o2 mm) from Guadeloupe: western slope sequence (Ws), eastern slope sequence (Es) and northern slope sequence (Ns).

d26Mg

2SD

d25Mg

2SD

n

d26Mg

2SD

d25Mg

2SD

n

Mg isotope compositions of reference materials and USGS rock standards Cambridge-1  2.60 0.15  1.34 0.09 Mg specpure  2.85 0.15  1.46 0.09 MgOXLFG 0.31 0.16 0.15 0.10 DSM-3  0.01 0.08  0.01 0.06 Cambridge-1 processed  2.59 0.11  1.35 0.05

576 153 119 52 9

BHVO-2 BCR-2 AGV-2 BIR-1a GSP-2

 0.31  0.30  0.24  0.27  0.02

0.19 0.19 0.24 0.33 0.31

 0.16  0.16  0.14  0.18  0.05

0.11 0.11 0.13 0.18 0.21

30 31 28 14 14

Mg isotope composition of the standard 0% BCR-2  3.01 20% BCR-2  2.44 40% BCR-2  1.83 60% BCR-2  1.25 80% BCR-2  0.81 100% BCR-2  0.28 BCR-2 predicted  0.22 Standard predicted  2.95

addition method 0.07  1.56 0.08  1.27 0.09  0.94 0.10  0.66 0.09  0.42 0.10  0.13 0.15 – 0.15 –

0.03 0.04 0.03 0.08 0.04 0.06 – –

5 5 5 5 5 5 – –

0% Icelandic soil 20% Icelandic soil 40% Icelandic soil 60% Icelandic soil 80% Icelandic soil 100% Icelandic soil Icelandic soil predicted Standard predicted

 2.92  2.42  1.87  1.29  0.75  0.19  0.18  2.99

0.18 0.11 0.16 0.24 0.19 0.05 0.02 0.02

 1.51  1.24  0.97  0.67  0.38  0.10 – –

0.10 0.05 0.11 0.13 0.11 0.03 – –

5 5 5 5 5 5 – –

Mg isotope compositions in samples from Guadeloupe Andesite  0.47 0.09  0.26 Ws-An bulk soil  0.19 0.03  0.11 Ws-Ca bulk soil  0.31 0.03  0.17 Es-An1 bulk soil  0.35 0.09  0.19 Es-An2 bulk soil  0.30 0.07  0.16 Es-Ni bulk soil  0.21 0.11  0.14 Ns-Fe bulk soil  0.32 0.19  0.18 Ns-Verti bulk soil  0.49 0.09  0.26

0.06 0.03 0.05 0.07 0.03 0.07 0.07 0.05

9 8 9 9 9 8 8 9

Ws-An clay Ws-Ca clay Es-An1 clay Es-An2 clay Es-Ni clay Ns-Fe clay Ns-Verti clay

 0.16  0.22  0.36  0.35  0.24  0.10  0.41

0.06 0.11 0.06 0.05 0.09 0.08 0.04

 0.11  0.09  0.20  0.17  0.13  0.02  0.21

0.07 0.10 0.03 0.06 0.06 0.05 0.05

9 6 8 5 9 9 8

4. Results 4.1. Soil weathering degree The main characteristics of the bulk soils, clay fractions and andesite samples are summarized in Table 1. The chemical composition (weight% oxide) of the parent andesite from Piton Tarade is 57.3% SiO2, 9.0% Fe2O3, 17.2% Al2O3, 7.3% CaO, 3.8% MgO, 3.3% Na2O, 0.8% K2O, 0.1% MnO, 0.7%TiO2, 0.2% P2O5, which is typical of the andesite from Guadeloupe (Samper et al., 2007; Sak et al., 2010). In the soils, the clay content increases from Ws (35–41%) to Es (52–80%) to Ns (82–85%), consistent with increasing weathering from Ws to Es to Ns (Fig. 3). The mineralogy of the clay fraction evolves from allophane to halloysite/smectite on the drier slope (Ws), and from allophane to halloysite/Fe-oxide on the wetter slope (Es). In the Ns, the mineralogy evolves from kaolinite/Feoxide including gibbsite and hydroxy-Al interlayered smectite (HIS) and hydroxy-Al interlayered vermiculite (HIV) in wet conditions to smectite in dry conditions. Therefore the clay fractions contain secondary Mg-bearing phyllosilicates (smectite, HIS and HIV). The overall weathering intensity in the three weathering sequences can be assessed from the elemental depletion of Mg in soils (weathered w) relative to their protolith andesite (parent p), using the open system elemental mass transfer coefficient t (Brimhall and Dietrich, 1987). The loss or gain of a mobile element x (x ¼Mg) relative to a immobile element Zr is calculated as follows (Brimhall and Dietrich, 1987; Kurtz et al., 2000): tx ¼[(Xw Zrp)/(Xp Zrw)]  1). A value of tx ¼0 represents the unweathered material and a value of tx ¼ 1 stands for a complete removal of x from the parent material. The tMg values decrease progressively from Ws to Es to Ns (Fig. 3), indicating that the Ws and Es are less depleted in Mg than the Ns. Both clay content and tMg values support an increasing weathering degree from Ws, to Es, to Ns, which corresponds to an increasing age of the soils (Cabidoche et al., 2009). This can also

Fig. 3. Distinction between the three weathering sequences using two weathering index: the clay content and the elemental depletion (tMg) in bulk soils (weathering index tx, x¼ Mg; Brimhall and Dietrich, 1987; Kurtz et al., 2000). Western slope sequence (Ws), eastern slope sequence (Es) and northern slope sequence (Ns).

be related to a higher rainfall pattern on the Es slope (MAR 5000– 2500 mm) and Ns-Fe (4000 mm) relative to the Ws slope (MAR 3200–850 mm) (Chaperon et al., 1985). 4.2. Magnesium distribution in soil In soils, Mg is distributed between primary minerals, secondary phases, and exchangeable Mg retained on the soil exchange complex (Fig. 4). These different Mg pools can be quantified: magnesium exchangeable (Mgexch) is a direct measurement by percolation (see methods), Mg in Mg-bearing secondary minerals (Mgsecondary) is a direct measurement of the total amount of Mg in clay fractions (o2 mm) after Sr saturation (see Methods), and Mg in primary minerals (Mgprimary) can be calculated by retrieving Mgsecondary and Mgexch from the total Mg content measured in bulk soils.

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

Fig. 4. Schematic representation of the Mg distribution in soils (no scale) in primary minerals (Mgprimary), Mg-bearing secondary minerals (Mgsecondary), and þ onto the soil exchange complex (Mgþ exch). The soil exchange complex is made of charged secondary clay minerals (open light grey forms) and organic compounds (light grey patches) and retains Mg with other exchangeable cations. Primary minerals are open black forms.

Fig. 5. Magnesium distribution in Guadeloupe bulk soils relative to the andesite between primary (Mgprimary), secondary minerals (Mgsecondary) and exchangeable Mg (Mgexch) pools. Magnesium exchangeable is a direct measurement by percolation of ammonium acetate (see methods), Mgsecondary is a direct measurement of the total amount of Mg in clay fractions (o 2 mm) after Sr saturation (see Methods), and Mgprimary is calculated by retrieving Mgsecondary and Mgexch from the total Mg content measured in soils. For the values given for each soils [X(Y)], X corresponds to the ratio between Mgprimary and Mgsecondary and Y to the proportion of exchangeable Mg in soils (Y¼Mgexch/Mgtotal). Soil names as in Fig. 1.

The distribution of Mg in the soils from the three weathering sequences is presented in Fig. 5. In Ws and Es, with increasing degree of weathering there is a decrease in the Mg content of the primary minerals and an increase in Mg incorporation into secondary minerals and Mg retention on the exchange complex (e.g., in Es-Ni strongly weathered, 28%, 64% and 9%, respectively, relative to Es-An1 less weathered 59%, 33% and 8%, respectively; Fig. 5). In the Ns, the primary mineral reserve is completely weathered (Fig. 5). The ratio between Mg in primary minerals and that included in secondary phases (Mgprimary/Mgsecondary; Fig. 5) can therefore be used as a weathering index to describe soil evolution. A ratio Mgprimary/Mgsecondary 41.5 defines a group of less evolved soils younger than 104 yr with clay content o60% (Ws-An, Es-An1, Es-An2, Ws-Ca), whereas a ratio of Mgprimary/Mgsecondary o0.5 corresponds to a group of more evolved soils older than 105 yr with a clay content 4 80% (Es-Ni, Ns-Fe, Ns-Verti). This ratio is useful for defining weathering processes in a range of soils with

181

differing age but similar clay content (Es-Ni and Ns-Fe). In the Es, Es-Ni is likely older than Es-An1 and Es-An2, as supported by a higher clay content and a lower Mgprimary/Mgsecondary ratio in EsNi. This supports that soils in Es are not only distributed along a climo-topo-sequence but along a chrono-climo-topo-sequence. The proportion of exchangeable Mg can be evaluated with the ratio between exchangeable Mg and the total Mg content in soils (Mgexch/Mgtot; Fig. 5). A high proportion of exchangeable Mg (NsVerti, Mgexch/Mgtot 40.06) are found in smectite-rich soils, since smectite is a double-layer phyllosilicate characterized by a high cationic retention capacity. Similarly, Es-An1 and Ns-Fe containing interlayered HIS and HIV (Table 1) also possess a high proportion of exchangeable Mg (Mgexch/Mgtot 40.06). Es-Ni also displays a high proportion of exchangeable Mg (Mgexch/ Mgtot 40.06) which can be attributed to the high clay content in this soil (80% clay; Table 1). By contrast, soils with a lower clay content (35–52%; Table 1) and an absence of double-layer phyllosilicates display low proportion of exchangeable Mg (Mgexch/Mgtot o0.06 in Ws-An, Ws-Ca, Es-An2; Fig. 5). 4.3. Magnesium isotopic compositions of soils and clay fractions The Mg isotope compositions of the andesite, soils and clay fractions are presented in Table 2 and Fig. 6. The d26Mg ratios of bulk soils is either similar to the parent andesite (  0.47%) within uncertainty (Ns-Verti, Es-An1), or significantly heavier (  0.31% to  0.19%). Secondary clay fractions (o2 mm) are isotopically similar (Ns-Verti, Es-An1, Es-An2) or significantly heavier (  0.24% to  0.10%) than the parent andesite (Fig. 6). Bulk soils containing Mg-bearing secondary phases (smectite or HIS/HIV) such as Ws-Ca, Es-An1, Es-An2, Ns-Fe and Ns-Verti tend to be isotopically lighter than bulk soils with allophane, halloysite or Fe-oxide (Ws-An and Es-Ni). Clay fraction Mg isotope compositions are not significantly different from their corresponding bulk soils, except in Ns-Fe where the clay fraction is significantly heavier (  0.10%) than the bulk soil (  0.32%). In the Ws weathering sequence, d26Mg decreases from Andosol to Cambisol, whereas in the Es weathering sequence, d26Mg increases from Perhydrated Andosol to Nitisol (Fig. 6). In the Ns, d26Mg decreases from Ferralsol to Vertisol (Fig. 6).

Fig. 6. Magnesium isotope compositions (%) in bulk soils (full symbols) and clay fractions (open symbols) relative to the parent andesite (full line with uncertainty as a shaded area) in the three soil sequences. Soil names as in Fig. 1. Error bars on d26Mg (2SD) correspond to the external precision on standards ( 7 0.15%) or the internal precision of the sample if larger.

182

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

5. Discussion 5.1. Weathering impact on magnesium isotope variations in soils Soils in Guadeloupe are distributed along chrono-climo-topo sequences and clay mineralogy in soils is strongly depending on climatic conditions of soil formation (Colmet-Daage and Lagache, 1965; Ndayiragije, 1996; Ndayiragije and Delvaux, 2003, 2004). In those soils, secondary Mg-bearing phases such as smectite and HIS/HIV have been identified (Table 1) and were formed from a local andesite (Samper et al., 2007; Sak et al., 2010). A minor proportion of the HIS/HIV minerals might also be derived from the weathering of hydrothermal smectites that originated from the pyroclastite (Cabidoche et al., 1987), as reported in Ferralsol and Nitisol (van Oort, 1988). The Mg isotope compositions of clay fractions from Guadeloupe are similar to or heavier than their parent silicate material (Fig. 6), which supports a preferential incorporation of heavy Mg isotope in secondary Mg-bearing phases during weathering processes with the first direct d26Mg measurements on clay fractions relative to their parent silicate material. This is in good agreement with previous measurements in bulk soils that are isotopically heavier than parent silicate bedrock (Teng et al., 2010; Tipper et al., 2006a, 2010; Fig. 7). A significant Mg pool in soils is retained on the soil exchange complex (1–17%; Fig. 5). Interestingly, the Mg isotope compositions in bulk soils are correlated with the amount of exchangeable Mg (R2 ¼0.84; Fig. 8a). This correlation is the first evidence that the retention of Mg on the soil exchange complex contributes to control Mg isotope variations in soils. More precisely, Ns-Verti, Es-An1, Ws-Ca bulk soils display lighter Mg isotope compositions, and contain a larger amount of exchangeable Mg. These soils are characterized by the presence of smectite or HIS/HIV in their clay fractions, which are double-layer phyllosilicates characterised by a high cationic retention capacity. Although silicate weathering is the most important source of Mg in these soils, some Mg in soils may also be derived from atmospheric inputs (aeolian dust, sea spray), and/or from banana culture (lime amendment, biological recycling). The impact of external inputs to soil surface layers is evaluated in topsoil samples (top 20 cm).

the top 20 cm of the soil. Given that the Mg isotope composition of quartz and mica is within the range of the parent silicate material in the soil, this is not likely to significantly affect the soil Mg isotope composition. 5.3. Potential contribution from banana culture It is expected that plants preferentially take up heavy Mg isotopes (Black et al., 2008; Bolou-Bi et al., 2010). In these plantations, the vegetation is returned to the soil at harvesting time, so the vegetation export is very limited. Assuming for all soils a 25 T yr  1 crop yield with a density of 2100 banana plants ha  1, the total dry matter production of a banana crop is estimated to be 14 T ha  1 yr  1 from which 10 T ha  1 yr  1 dry matter of residues are left in the field (Twyford and Walmsley, 1973). Assuming a Mg content in banana plant of 3 g kg  1 (Henriet et al., 2008a), this would account for an input of 30 kg ha  1 yr  1 of Mg from plant residues, which is less than 1% of the Mg budget of the top 20 cm of the soil. Considering a plant with a Mg isotope signature of  0.9% (average value for plants from Bolou-Bi et al., 2009), a contribution of 40% from plant residues to the Mg budget would be necessary to produce a significant change in the soil Mg isotope composition. This is much higher than the plant return budget calculated (less than 1%), which suggests that the biological cycling of Mg only plays a minor role on the control of the Mg isotope compositions of the soils. Lime amendment on topsoil before banana plantation is an additional Mg contribution. The amount of lime amendment is 1 T ha  1 and is not recurring. Taking into account MgO content in lime of 4%, total lime Mg would account for 24 kg ha  1of Mg, which is less than 1% of the Mg budget of the top 20 cm of the soil. With Mg isotope composition of limestone (  3%; Tipper et al., 2006b) or dolomite (   2%; Galy et al., 2002; Carder et al., 2005), an isotope mass balance calculation indicates that a minimum of 15% contribution from the lime to the Mg budget would be necessary to significantly change the Mg isotope composition of the soil. This is much higher than the calculated contribution (less than 1%), which demonstrates that the initial lime amendment ( 10 yr old banana culture) does not significantly affect the soil Mg isotope composition.

5.2. Potential contribution from aeolian dust 5.4. Potential contribution from sea spray to exchangeable Mg Aeolian dust over the Caribbean Islands originates from Sahel and the Sahara (Petit et al., 2005) with a mineralogy dominated by quartz and mica (Muhs et al., 2010). Taking into account Mg deposition of 10 kg ha  1 yr  1 from Saharian dust (Boy and Wilcke, 2008), this would account for 0.3% of the Mg budget of

Fig. 7. Magnesium isotope compositions (%, 2SD) in Guadeloupe bulk soils and clay fractions relative to their parent andesite compared with data from the literature. Bulk soils or clay fractions are represented by symbols, and the parent silicate material is represented by a vertical full line. [1] This study; [2] Brenot et al. (2008); [3] Teng et al. (2010); [4] Tipper et al. (2010); [5] Tipper et al. (2006a); [6] Pogge Von Strandmann et al. (2008); BSE ¼bulk silicate Earth.

Even if exchangeable cations such as Mg, Ca, K, Na are primarily derived from rock weathering, external inputs to soils such as sea spray can significantly affect the exchangeable cation budget in soils (Whipkey et al., 2000), especially in strongly weathered soils characterised by an extreme cation depletion (Kennedy et al., 1998). A potential contribution from sea spray to the budget and isotopic composition of exchangeable Mg in soils from Guadeloupe can be evaluated by mass balance. Assuming that all the Mg in the soil exchange complex originates from sea spray, the soil Mg isotope composition can be corrected for a seawater contribution (d26Mg¼  0.82%; see review by Foster et al., 2010 and references therein). The difference between the soil d26Mg corrected value and the uncorrected values (d26Mg seaspray 26 corr  d Mg uncorr.) varies between 0.01% and 0.1% (Fig. 8b), which is within the analytical error. This suggests that with exchangeable Mg originating from sea spray, the reserve in exchangeable Mg in soil is not large enough to significantly affect the bulk soil Mg isotope composition. But a sea spray contribution to the Mg exchange complex cannot be ruled out. Interestingly, it appears that the difference d26Mg seaspray corr  d26Mg uncorr. increases with an increasing

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

183

Fig. 8. (a) Plot of Mg isotope compositions (%) in bulk soils as a function of the amount of exchangeable Mg in bulk soils (cmolc kg  1; Table 1). (b) Plot of the difference between the d26Mg value in bulk soil corrected for sea spray contribution and the uncorrected value (d26Mg seaspray corr - d26Mg uncorr. in %; see Section 5.4 for calculation) as a function of the amount of exchangeable Mg in bulk soils. (c) Plot of d26Mg seaspray corr  d26Mg uncorr. (%) as a function of the elemental depletion (tMg). (d) Plot of Mg isotope compositions (%) in clay fractions as a function of the amount of exchangeable Mg in bulk soils. Western slope sequence (Ws), eastern slope sequence (Es) and northern slope sequence (Ns). Error bars on d26Mg (2SD) correspond to the external precision on standards ( 7 0.15%) or the internal precision of the sample if larger.

amount of exchangeable Mg in bulk soil, and is larger for Es-Ni and Ns-Fe (Fig. 8b). This supports that the contribution from sea spray to the exchangeable Mg on the complex is larger in the two more weathered soils of the studied sites (soils with high clay content and low tMg value; Fig. 3 and Table 1). The contribution from sea spray to the exchangeable Mg in soil is thus inversely correlated to the soil weathering degree, as represented by the Mg depletion in soils (Fig. 8c). This is in good agreement with a larger contribution from sea spray to nutrient budget in cation depleted soils (Kennedy et al., 1998; Chadwick et al., 1999). Based on these results, the significantly lighter d26Mg ratio in Ns-Fe bulk soil relative to the corresponding clay fraction (Fig. 6) can be attributed to the contribution from the exchangeable Mg pool (17%) to the bulk soil composition. It is likely that sea spray contributes to the exchangeable Mg pool in that soil (Ns-Fe), as this is a strongly weathered soil (81.6% clay; Table 1). The d26Mg in the clay fraction in Ns-Fe (Sr-saturated) reflects Mg incorporated into phyllosilicate minerals (HIS/HIV). The Mg onto the soil exchangeable complex is available for secondary mineral neoformation. This raises the question whether Mg originating from sea spray might be incorporated into structural sites (octahedral sites) of clay minerals. The Mg isotope compositions of the clay fractions (reflecting Mg incorporated into the

mineral structure) are lighter with an increasing amount of exchangeable Mg in soils (Fig. 8d). This suggests that an increasing contribution from sea spray to the soil exchange complex may contribute to provide Mg for secondary clay neoformation. These data highlight how complex the Mg stable isotope ratio in a soil is. It depends not only on the isotopic composition of the parent rock, and on any fractionation that might occur, but also on the d26Mg composition of the exchange complex, which could in turn be controlled by external inputs such as sea spray. 5.5. Implications This study provides a direct evidence for the incorporation of heavy Mg isotopes in secondary clay fractions relative to the parent andesite. This contributes to the lighter riverine signatures in silicate catchments (e.g., De Villiers et al., 2005; Brenot et al., 2008; Tipper et al., 2006b, 2008a; Pogge Von Strandmann et al., 2008, Wimpenny et al., 2011). The data highlight that a substantial Mg pool from the soil exchange complex is partly controlling the soil Mg isotope composition. This exchangeable Mg reservoir potentially impacts Mg isotope compositions in soil solutions and Mg isotope budget exported from soils, acting as a buffer for soil solutions (Tipper et al., 2010).

184

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

Since increasing amount of exchangeable Mg appears to drive the bulk soil isotope composition to lighter values (Fig. 8a), the Mg retention process might explain why soils may be lighter than their parent silicate material (Pogge Von Strandmann et al., 2008).

6. Concluding remarks This study supports a preferential incorporation of heavy Mg isotopes in secondary Mg-bearing clay minerals, with the first Mg isotope measurements on clay fractions relative to the parent andesite. In addition, the data provide evidence that Mg retained on the soil exchange complex contributes to the shift to lighter Mg isotope compositions in soils, and support a contribution from sea spray to the exchangeable Mg. This study represents a first step in identifying the processes of Mg retention in the soil and the secondary phases that drive Mg isotope fractionation. However, further investigations of depth profiles are needed to better understand the overall weathering processes and to more precisely exclude the potential impact from atmospheric and agricultural input. More investigations will also be required to constrain the amplitude of the isotope fractionation associated with Mg incorporation in secondary phases, and to identify if an isotope fractionation is associated with the Mg retention on the soil exchange complex.

Acknowledgements We would like to thank A. Iserentant, C. Givron, A. Lannoye, P. Populaire, C. Henriet, N. de Jaeger, L. Bodarwe´ (Universite´ catholique de Louvain, Belgium) and M. Dorel (CIRAD, Guadeloupe) for their contribution in the completion of this project, J. West, K. Hendry, C. Siebert, N. Belshaw, T. Krastev, F. Mokadem, A. Mason, S. Wyatt for their help in the Isotope geochemistry lab of the University of Oxford (UK), and S. Hammond (Open University, UK) for ICP-MS analyses. The manuscript greatly benefited from the comments of E. Tipper and three anonymous referees. S.O. is funded by the ‘‘Fonds National de la Recherche Scientifique’’ (Belgium) and acknowledges a funding from FSR (Fonds Special de Recherche, UCL, Belgium). Funding for isotopic analyses at Oxford was provided by a grant to A.H. from the European Research Council.

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.epsl.2012.06.010.

References Baker, J., Bizzarro, M., Wittig, N., Connelly, J., Haack, H., 2005. Early planetesimal melting from an age of 4.5662 Gyr for differentiated meteorites. Nature 436, 1127–1131. Berner, R.A., 1995. Chemical weathering and its effect on the atmospheric CO2 and climate. In: White, A.F., Brantley, S.L. (Eds.), Chemical Weathering Rates of Silicate Minerals. Rev. Mineral.. Mineral. Soc. Am., Washington, D.C. (USA), pp. 565–583. Bizzarro, M., Baker, J.A., Haack, H., Lundgaard, K.L., 2005. Rapid timescales for accretion and melting of differentiated planetesimals inferred from 26Al–26Mg chronometry. Astrophys. J. 632, L41–L44. Black, J.R., Epstein, E., Rains, W.D., Yin, Q.Z., Casey, W.H., 2008. Magnesium-isotope fractionation during plant growth. Environ. Sci. Technol. 42, 7831–7836. Bolou-Bi, B.E., Vigier, N., Brenot, A., Poszwa, A., 2009. Magnesium isotope compositions of natural reference materials. Geostand. Geoanal. Res. 33, 95–109. Bolou-Bi, E.B., Poszwa, A., Leyval, C., Vigier., N., 2010. Experimental determination of magnesium isotope fractionation during higher plant growth. Geochim. Cosmochim. Acta 74, 2523–2537. Borchardt, G., 1989. Smectites. In: Dixon, J.B., Weed, S.B. (Eds.), Minerals in Soil Environments. Soil Science Society of America, Madison, pp. 675–727.

Bourdon, B., Tipper, E.T., Fitoussi, C., Stracke, A., 2010. Chondritic Mg isotope composition of the Earth. Geochim. Cosmochim. Acta 74, 5069–5083. Boy, J., Wilcke, W., 2008. Tropical Andean forest derives calcium and magnesium from Saharan dust. Global Biogeochem. Cycles, GB1027. Brantley, S.L., 2010. Weathering: rock to regolith. Nat. Geosci. 3, 305–306. Brenot, A., Cloquet, C., Vigier, N., Carignan, J., France-Lanord, C., 2008. Magnesium isotope systematics of the lithologically varied Moselle river basin, France. Geochim. Cosmochim. Acta 72, 5070–5089. Brimhall, G.H., Dietrich, W.E., 1987. Constitutive mass balance relations between chemical composition, volume, density, porosity, and strain in metasomatic hydrochemical systems; results on weathering and pedogenesis. Geochim. Cosmochim. Acta 51, 567–587. Cabidoche, Y.-M., Achard, R., Cattan, P., Clermont-Dauphin, C., Massat, F., Sansoulet, J., 2009. Long-term pollution by chlordecone of tropical volcanic soils in the French West Indies: a simple leaching model accounts for current residue. Environ. Pollut. 157, 1697–1705. Cabidoche, Y.-M., Feller, C., Larque, P., Sobesky, O., 1987. Sur un double me´canisme d’acidification des sols sous l’influence de cendres volcaniques re´centes. Le cas de la Soufrie re de Guadeloupe apre s les e´ruptions de 1976–77. C.R. Acad. Sci. Paris 304, se´rie II, 935–938. Carder, E.A., Galy, A., Mckenzie, J.A., Vasconcelos, C., Elderfield, H., 2005. Magnesium isotopes in bacterial dolomites: a novel approach to the dolomite problem. Geochim. Cosmochim. Acta 69, A213. Chadwick, O.A., Derry, L.A., Vitousek, P.M., Heubert, B.J., Hedin, L.O., 1999. Changing sources of nutrients during four million years of ecosystem development. Nature 397, 491–497. Chadwick, O.A., Gavenda, R.T., Kelly, E.F., Ziegler, K., Olson, C.G., Elliott, W.C., Hendricks, D.M., 2003. The impact of climate on the biogeochemical functioning of volcanic soils. Chem. Geol. 202, 195–223. Chao, T.T., Sanzolone, R.F., 1992. Decomposition techniques. J. Geochem. Explor. 44, 65–106. Chaperon, P., L’Hote, Y., Vuillaume, G., 1985. Les ressources en eau de surface de la Guadeloupe. Editions de l’ORSTOM, Collection Monographies Hydrologiques 7, France. Colmet-Daage, F., Lagache, P., 1965. Caracte´ristiques de quelques sols de´rive´s de roches aux Antilles Franc- aises. Cah. ORSTOM Ser. Pedol 3, 91-121. Dagain, J., Paterne, M., Westercamp, D., 1981. La mise en place du massif volcanique Madeleine-Soufrie re, Basse-Terre de Guadeloupe, Antilles. C.R. Acad. Sci. Paris 292, 921–926. De Villiers, S., Dickson, J.A.D., Ellam, R.M., 2005. The composition of the continental river weathering flux deduced from seawater Mg isotopes. Chem. Geol. 216, 133–142. Dorel, M., Roger-Estrade, J., Manichon, H., Delvaux, B., 2000. Porosity and soil water properties of Caribbean volcanic ash soils. Soil Use Manage. 16, 133–140. Epstein, E., Bloom, A.J., 2005. Mineral Nutrition of Plants: Principles and Perspectives, 2nd ed. Sinauer Associates, Inc., Sunderland, MA, USA 380 pp. Foster, G.L., Pogge von Strandmann, P.A.E., Rae, J.W.B., 2010. The boron and magnesium isotopic composition of seawater. Geochem. Geophys. Geosyst. 11, Q08015. Gaillardet, J., Louvat, P., Lajeunesse, E., 2011. Rivers from volcanic island arcs: the subduction weathering factory. Appl. Geochem. 26, S350–S353. Gaillardet, J., Viers, J., Dupre´, B., 2010. Trace elements in river waters. In: Holland, Heinrich D., Turekian, Karl K. (Eds.), Geochemistry of Earth Surface Systems. Academic Press, Elsevier, London, pp. 293–339. Galy, A., Bar-Mathews, M., Halicz, L., O’Nions, R.K., 2002. Mg isotopic composition of carbonate: insight from speleothem formation. Earth Planet. Sci. Lett. 201, 105–115. Galy, A., Yoffe, O., Janney, P.E., Williams, R.W., Cloquet, C., Alard, O., Halicz, L., Wadhwa, M., Hutcheon, I.D., Ramon, E., Carignan, J., 2003. Magnesium isotope heterogeneity of the isotopic standard SRM980 and new reference materials for magnesium isotope ratio measurements. J. Anal. Atom. Spectrom. 18, 1352–1356. Henriet, C., Bodarwe´, L., Dorel, M., Draye, X., Delvaux, B., 2008a. Leaf silicon content in banana (Musa spp.) reveals the weathering stage of volcanic ash soils in Guadeloupe. Plant Soil 313, 71–82. Henriet, C., De Jaeger, N., Dorel, M., Opfergelt, S., Delvaux, B., 2008b. The reserve of weatherable primary silicates impacts the accumulation of biogenic silicon in volcanic ash soils. Biogeochemistry 90, 209–223. Hinckley, D.N., Bates, T.F., 1960. An X-ray fluorescent method for the determination of small amounts of montmorillonite in kaolin clays. Am. Mineral. 45, 239–241. Huang, F., Glessner, J., Ianno, A., Lundstrom, C., Zhang, Z., 2009. Magnesium isotopic composition of igneous rock standards measured by MC-ICP-MS. Chem. Geol. 268, 15–23. I.U.S.S. Working Group WRB (2006) World Reference Base for Soil Resources 2006, 2nd ed. World Soil Resources Reports No. 103. FAO, Rome. Kennedy, M.J., Chadwick, O.A., Vitousek, P.M., Derry, L.A., Hendricks, D.M., 1998. Changing sources of base cations during ecosystem development, Hawaiian Islands. Geology 26, 1015–1018. Kurtz, A.C., Derry, L.A., Chadwick, O.A., Alfano, M.J., 2000. Refractory element mobility in volcanic soils. Geology 28, 683–686. Markewitz, D., Davidson, E.A., Figueiredo, R.O., Victoria, R.L., Krusche, A.V., 2001. Control of cation concentrations in stream waters by surface soil processes in an Amazonian watershed. Nature 410, 802–805. Miller, E.K., Blum, J.D., Friedland, A.J., 1993. Determination of soil exchangeablecation loss and weathering rates using Sr isotopes. Nature 362, 438–441.

S. Opfergelt et al. / Earth and Planetary Science Letters 341-344 (2012) 176–185

Millero, F.J., 1974. The physical chemistry of seawater. Annu. Rev. Earth Planet. Sci. 2, 101–150. Muhs, D.R., Budahn, J., Skipp, G., Prospero, J.M., Patterson, D., Bettis III, E.A., 2010. Geochemical and mineralogical evidence for Sahara and Sahel dust additions to Quaternary soils on Lanzarote, eastern Canary Islands, Spain. Terra Nova 22, 399–410. Ndayiragije, S., 1996. Caracte´risation d’une se´quence d’alte´ration de sols de´rive´s de mate´riaux pyroclastiques sous climat tropical humide des Antilles, Guadeloupe. Ph.D. Thesis, Universite´ catholique de Louvain, 181 pp. Ndayiragije, S., Delvaux, B., 2003. Coexistence of allophane, gibbsite, kaolinite and hydroxy-Al-interlayered 2: 1 clay minerals in a Perudic Andosol. Geoderma 117, 203–214. Ndayiragije, S., Delvaux, B., 2004. Selective sorption of potassium in a weathering sequence of volcanic ash soils from Guadeloupe, French West Indies. Catena 56, 185–198. Page, A.L., Miller, R.H., Keeney, D.R., 1982. Methods of Soil Analysis: part 2, Chemical and Microbiological Properties, 2nd ed. Am. Soc. Agron. and Soil Sci. Soc. Am., Madison. Petit, R.H., Legrand, M., Jankowiak, I., Molinie´, J., Asselin de Beauville, C., Marion, G., Mansot, J.L., 2005. Transport of Saharan dust over the Caribbean Islands: study of an event. J. Geophys. Res. 110, D18S09. Pineros-Garcet, J.D., 1994 Etude d’une climotopose´quence de sols de´rive´s de mate´riaux volcaniques re´cents, Coˆ te sous le vent de la Basse-Terre, Guadeloupe. MSc Thesis, Universite´ catholique de Louvain, p. 86. Pogge Von Strandmann, P.A.E., Burton, K.W., James, R.H., Van Calsteren, P., Gislason, S.R., Sigfusson, B., 2008. The influence of weathering processes on riverine magnesium isotopes in a basaltic terrain. Earth Planet. Sci. Lett. 276, 187–197. Robert, M., Tessier, D., 1974. Me´thode de pre´paration des argiles des sols pour des e´tudes mine´ralogiques. Ann. Agron. 25, 859–882. Rouiller, J., Burtin, G., Souchier, B., 1972. La dispersion des sols dans l’analyse granulome´trique. Me´thode utilisant les re´sines e´changeuses d’ions. ENSAIA Nancy 14, 194–205. Sak, P.B., Navarre-Sitchler, A.K., Miller, C.E., Daniel, C.C., Gaillardet, J., Buss, H.L., Lebedeva, M.I., Brantley, S.L., 2010. Controls on rind thickness on basaltic andesite clasts weathering in Guadeloupe. Chem. Geol. 276, 129–143. Samper, A., Quidelleur, X., Lahitte, P., Mollex, D., 2007. Timing of effusive volcanism and collapse events within an oceanic arc island: Basse-Terre, Guadeloupe archipelago (Lesser Antilles Arc). Earth Planet. Sci. Lett. 258, 175–191. Sansoulet, J., 2007 Water flow and transport of potassium and nitrate ions in a soil with anionic exchange capacity under rainfall redistributed by crop. Ph.D. Thesis, Institut national agronomique Paris Grignon, p. 160. Taylor, S.R., Mclennan, S.M., 1985. The Continental Crust: its Composition and Evolution: an Examination of the Geochemical Record Preserved in Sedimentary Rocks. Blackwell Science, Oxford 312 pp.. Teng, F.-Z., Li, W.Y., Rudnick, R.L., Gardner, L.R., 2010. Contrasting lithium and magnesium isotope fractionation during continental weathering. Earth Planet. Sci. Lett. 300, 63–71.

185

Teng, F.Z., Wadhwa, M., Helz, R.T., 2007. Investigation of magnesium isotope fractionation during basalt differentiation: implications for a chondritic composition of the terrestrial mantle. Earth Planet. Sci. Lett. 261, 84–92. Tipper, E.T., Galy, A., Bickle, M.J., 2006a. Riverine evidence for a fractionated reservoir of Ca and Mg on the continents: Implications for the oceanic Ca cycle. Earth Planet. Sci. Lett. 247, 267–279. Tipper, E.T., Galy, A., Bickle, M.J., 2008a. Calcium and magnesium isotope systematics in rivers draining the Himalaya–Tibetan-Plateau region: lithological or fractionation control? Geochim. Cosmochim. Acta 72, 1057–1075. Tipper, E.T., Louvat, P., Capmas, F., Galy, A., Gaillardet, J., 2008b. Accuracy of stable Mg and Ca isotope data obtained by MC-ICP-MS using the standard addition method. Chem. Geol. 257, 65–75. Tipper, E.T., Gaillardet, J., Louvat, P., Capmas, F., White, A.F., 2010. Mg isotope constraints on soil pore-fluid chemistry: evidence from Santa Cruz, California. Geochim. Cosmochim. Acta 74, 3883–3896. Tipper, E.T., Galy, A., Gaillardet, J., Bickle, M.J., Elderfield, H., Carder, E.A., 2006b. The magnesium isotope budget of the modern ocean: Constraints from riverine magnesium isotope ratios. Earth Planet. Sci. Lett. 250, 241–253. Twyford, I.T., Walmsley, D., 1973. The mineral composition of the Robusta banana plant. I. Methods and plant growth studies. Plant Soil 39, 227–243. van Oort, F., 1988. Pre´sence et e´volution des mine´raux argileux accessoires de type 2:1 dans les sols ferrallitiques d’origine volcanique de la Guadeloupe. Conse´quences physico-chimiques. C.R. Acad. Sci Paris Serie 2 (307), 1297–1302. Whipkey, C.E., Capo, R.C., Chadwick, O.A., Stewart, B.W., 2000. The importance of sea spray to the cation budget of a coastal Hawaiian soil: a strontium isotope approach. Chem. Geol. 168, 37–48. White, A.F., Blum, A.E., Schulz, M.S., Vivit, D.V., Stonestrom, D.A., Larsen, M., Murphy, S.F., Eberl, D., 1998. Chemical weathering in a tropical watershed, Luquillo Mountains, Puerto Rico: I. Long-term versus short-term weathering fluxes. Geochim. Cosmochim. Acta 62, 209–226. Wiechert, U., Halliday, A.N., 2007. Non-chondritic magnesium and the origins of the inner terrestrial planets. Earth Planet. Sci. Lett. 256, 360–371. Wimpenny, J., Burton, K.W., James, R.H., Gannoun, A., Mokadem, F., Gislason, S.R., 2011. The behaviour of magnesium and its isotopes during glacial weathering in an ancient shield terrain in west Greenland. Earth Planet. Sci. Lett. 304, 260–269. Wimpenny, J., Gislason, S.R., James, R.H., Gannoun, A., Pogge Von Strandmann, P.A.E., Burton, K.W., 2010. The behaviour of Li and Mg isotopes during primary phase dissolution and secondary mineral formation in basalt. Geochim. Cosmochim. Acta 74, 5259–5279. Wombacher, F., Eisenhauer, A., Heuser, A., Weyer, S., 2009. Separation of Mg, Ca and Fe from geological reference materials for stable isotope ratio analyses by MC-ICP-MS and Double-Spike TIMS. J. Anal. At. Spectrom. 24, 627–636. Yeghicheyan, D., Carignan, J., Valladon, M., Bouhnik Le Coz, M., Le Cornec, F., Castrec-Rouelle, M., Robert, M., Aquilina, L., Aubry, E., Churlaud, C., Dia, A., Deberdt, S., Dupre´, B., Freydier, R., Gruau, G., He´nin, O., de Kersabiec, A.-M., Mace´, J., Marin, L., Morin, N., Petitjean, P., Serrat, E., 2001. A compilation of silicon and thirty one trace elements measured in the natural river water reference material SLRS-4 (NRC-CNRC). Geostand. Newsletter: J. Geostand. Geoanal. 25, 465–474.