Taxes & Accounting

Magnesium retention on the soil exchange complex controlling Mg isotope variations in soils, soil solutions and vegetation in volcanic soils, Iceland

Description
Magnesium retention on the soil exchange complex controlling Mg isotope variations in soils, soil solutions and vegetation in volcanic soils, Iceland
Published
of 21
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Magnesium retention on the soil exchange complexcontrolling Mg isotope variations in soils, soil solutionsand vegetation in volcanic soils, Iceland S. Opfergelt a,b, ⇑ , K.W. Burton a, 1 , R.B. Georg c , A.J. West d , R.A. Guicharnaud e, 2 ,B. Sigfusson f  , C. Siebert a , S.R. Gislason g , A.N. Halliday a a Department of Earth Sciences, University of Oxford, South Parks Road, Oxford OX1 3AN, United Kingdom b Earth and Life Institute, Universite´  catholique de Louvain, Croix du Sud bte L7.05.10, 1348 Louvain-la-Neuve, Belgium c Trent University, Water Quality Centre, 1600 West Bank Dr., Peterborough, Ontario, Canada d University of Southern California, Department of Earth Sciences, Los Angeles, CA, USA e Agricultural University of Iceland, Keldnaholt, 112 Reykjavik, Iceland  f  Reykjavik Energy, Baejarhals 1, 110 Reykjavik, Iceland  g Institute of Earth Sciences, University of Iceland, Sturlugata 7, 101 Reykjavik, Iceland  Received 13 June 2013; accepted in revised form 25 September 2013; available online 22 October 2013 Abstract Understanding the biogeochemical cycle of magnesium (Mg) is not only crucial for terrestrial ecology, as this element is akey nutrient for plants, but also for quantifying chemical weathering fluxes of Mg and associated atmospheric CO 2  consump-tion, requiring distinction of biotic from abiotic contributions to Mg fluxes exported to the hydrosphere. Here, Mg isotopecompositions are reported for parent basalt, bulk soils, clay fractions, exchangeable Mg, seasonal soil solutions, and vegeta-tion for five types of volcanic soils in Iceland in order to improve the understanding of sources and processes controlling Mgsupply to vegetation and export to the hydrosphere. Bulk soils ( d 26 Mg =  0.40 ± 0.11 & ) are isotopically similar to the parentbasalt ( d 26 Mg =  0.31 & ), whereas clay fractions ( d 26 Mg =  0.62 ± 0.12 & ), exchangeable Mg ( d 26 Mg =  0.75 ± 0.14 & ),and soil solutions ( d 26 Mg =  0.89 ± 0.16 & ) are all isotopically lighter than the basalt. These compositions can be explainedby a combination of mixing and isotope fractionation processes on the soil exchange complex. Successive adsorption–desorp-tion of heavy Mg isotopes leads to the preferential loss of heavy Mg from the soil profile, leaving soils with light Mg isotopecompositions relative to the parent basalt. Additionally, external contributions from sea spray and organic matter decompo-sition result in a mixture of Mg sources on the soil exchange complex. Vegetation preferentially takes up heavy Mg from thesoil exchange complex ( D 26 Mg plant-exch  = +0.50 ± 0.09 & ), and changes in  d 26 Mg in vegetation reflect changes in bioavailableMg sources in soils. This study highlights the major role of Mg retention on the soil exchange complex amongst the factorscontrolling Mg isotope variations in soils and soil solutions, and demonstrates that Mg isotopes provide a valuable tool formonitoring biotic and abiotic contributions of Mg that is bioavailable for plants and is exported to the hydrosphere.   2013 Elsevier Ltd. All rights reserved. 1. INTRODUCTION Magnesium (Mg) is an essential biological nutrient, as akey component of chlorophyll (Marschner, 1995), and thusit plays an important role in the biogeochemical cycles of ecosystems (Bormann and Likens, 1970; Likens and Bor-mann, 1974). It is the eighth most abundant element in 0016-7037/$ - see front matter    2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.gca.2013.09.036 ⇑ Corresponding author at: UCL/ELIE, Croix du Sud 2 bteL7.05.10, 1348 Louvain-la-Neuve, Belgium. Tel.: +32 10 47 36 32;fax: +32 10 47 45 25. E-mail address:  sophie.opfergelt@uclouvain.be (S. Opfergelt). 1 Present address: Department of Earth Sciences, DurhamUniversity, DH1 3LE Durham, United Kingdom. 2 Present address: European Commission, Land Resource Man-agement Unit, 21027 Ispra, Italy. www.elsevier.com/locate/gca Available online at www.sciencedirect.com ScienceDirect  Geochimica et Cosmochimica Acta 125 (2014) 110–130  the continental crust and the fourth most abundant speciesin seawater (Millero, 1974; Taylor and McLennan, 1985). The Mg in the oceans is derived from weathering of thecontinents, and over long timescales it is the chemical reac-tion of Mg and calcium (Ca) from silicate minerals withatmospheric carbon dioxide (CO 2 ) that is thought to regu-late global climate (Berner, 2004).Estimates suggest that 30–35% of the present-day atmo-spheric CO 2  consumption due to silicate weathering on landis attributable to weathering of basaltic rocks, even thoughthey constitute less than 5% of the overall continental area(Dessert et al., 2003). Not only are basaltic rocks rich indivalent cations such as Fe, Ca, and Mg, but their dissolu-tion rates are rapid (Dessert et al., 2003), resulting in cationrelease rates that are about 2 orders of magnitude fasterthan other silicate rocks (e.g. Oelkers and Schott, 1995;Wolff-Boenisch et al., 2006), thereby providing a significantsource of dissolved inorganic species to the ocean (Dessertet al., 2003). The control on magnesium fluxes exportedfrom basaltic catchments is generally attributed to basaltdissolution and secondary mineral formation, e.g. in Ice-land (Sigfusson et al., 2008). However, the presence of veg-etation significantly increases chemical weathering fluxes(Gislason et al., 1996; Moulton et al., 2000; Stefansson and Gislason, 2001), as shown in Iceland where Mg fluxesare four times higher for vegetated terrains than for barrenareas (Moulton et al., 2000). In cold regions with a seasonalpermafrost thaw, the contribution from vegetation to Mgfluxes exported to the hydrosphere is likely to vary overthe year (e.g., Pokrovsky et al., 2006; Zakharova et al.,2007).Distinguishing the relative magnitude of biotic and abi-otic contributions to Mg fluxes exported to rivers is a pre-requisite for precise quantification of continental Mgweathering fluxes, and the associated long-term regulationof atmospheric CO 2  (Berner, 2004). Magnesium stable iso-topes potentially provide a powerful tracer of the sourcesand processes controlling Mg fluxes in natural waters be-cause they are fractionated by both biotic and abiotic pro-cesses (Schmitt et al., 2012). Vegetation takes uppreferentially heavy Mg isotopes (Black et al., 2008; Bo-lou-Bi et al., 2010, 2012; Tipper et al., 2010). Rivers in sil- icate catchments (De Villiers et al., 2005; Brenot et al.,2008; Tipper et al., 2008, 2012; Wimpenny et al., 2011; Bo- lou-Bi et al., 2012) and soil solutions (Tipper et al., 2010;Bolou-Bi et al., 2012; Pogge von Strandmann et al., 2012)generally display lighter Mg isotope compositions thantheir parental silicate material, and this is usually attributedto the preferential incorporation of the heavy Mg isotopesinto secondary silicate phases (Tipper et al., 2006a; Tenget al., 2010a; Opfergelt et al., 2012; Huang et al., 2012), or the adsorption of heavy Mg isotopes onto clay minerals(Huang et al., 2012). However, disentangling the effects of mineral dissolution, secondary mineral neoformation, andadsorption on the Mg isotope budget remains difficult sincethe influence of Mg retention (adsorption–desorption andcation exchange processes) on Mg isotopes remains unclear.In the global cation budget to the ocean, cation exchangeconstitutes an important source of cations to rivers (Marke-witz et al., 2001), representing for example up to 30% of theannual Sr export (Miller et al., 1993).A magnesium isotope fractionation associated withMg adsorption onto soil constituents has been suggestedand is thought to favour the adsorption of heavy Mg rel-ative to the solution (Huang et al., 2012). Additionally, amixing process is suggested from published data since Mgisotope variations in soils or in soil solutions depend onthe amount of exchangeable Mg (Ryu et al., 2011; Bolou-Bi et al., 2012; Huang et al., 2012; Opfergelt et al., 2012;Pogge von Strandmann et al., 2012) and on Mg sourcescontributing to the soil exchange complex (e.g., contribu-tion from sea spray; Tipper et al., 2010; Opfergelt et al.,2012). If correct, since cation exchange on the soil ex-change complex is a faster process than weathering andmineral neoformation, this suggests that the Mg isotopevariations in soil solutions may be partly controlled byion exchange processes in soils. However, a direct assess-ment has not yet been possible since published data pres-ent either the bulk soil, or the soil solution or theexchangeable Mg, but not all of these components fromthe same soils. Better distinguishing whether an isotopefractionation and/or a mixing process is associated withMg retention (cation exchange and adsorption–desorp-tion) is a prerequisite to distinguishing the roles of weath-ering and mineral neoformation from retention processeson Mg release to the hydrosphere.In the present study, combined  d 26 Mg compositions of parent basalt, bulk soils, clay fractions, exchangeable Mg,seasonal soil solutions (before and after the growing sea-son) and vegetation on the same soils (five Icelandic soilsderived from basalt) were measured to directly to assesswhether an isotope fractionation accompanies Mg retentionon the soil exchange complex. Soils have been selected thatrange from well-drained organic-poor to poorly drained or-ganic-rich. Consequently, the chemistry of the soil exchangecomplex is different, being rich in base cations in well-drained soils (neutral), and rich in protons in poorlydrained soils (acidic). These results provide better con-straints on the role of Mg retention in soils on Mg isotopevariations in soils, thereby improving the utility of Mg Table 1Main characteristics of the selected soil profiles: soil type, location name, latitude and longitude, vegetation and drainage.Soil name ID Location Latitude N Longitude W Vegetation DrainageHistic Andosol HA Hestur N64  34 0 28.1 00 W21  35 0 41.9 00 Grass Poorly drainedHistosol H Klettur N64  38 0 36.6 00 W21  28 0 53.2 00 Grass Poorly drainedBrown Andosol BA Arnbjarnarlaekur N64  43 0 32.2 00 W21  30 0 14.2 00 Grass Freely drainedGleyic Andosol GA Korpa N64  9 0 1.0 00 W21  45 0 3.9 00 Grass Top well-drained then poorlyVitrisol V Geita Sandur N63  49 0 38.7 00 W20  12 0 34.1 00 Grass (2% cover) Well-drainedParental basalt Arnbjarnarlaekur N64  43 0 32.2 00 W21  30 0 14.2 00 OutcropS. Opfergelt et al./Geochimica et Cosmochimica Acta 125 (2014) 110–130 111  isotopes as a proxy for tracing the impact of weatheringprocesses on Mg fluxes exported to the hydrosphere. 2. ENVIRONMENTAL SETTING Five typical Icelandic soil types (Histic Andosol, HA;Histosol, H; Brown Andosol, BA; Gleyic Andosol, GA;Vitrisol V) were selected for this study following the Icelan-dic soil classification of  Arnalds (2004). These were sampledin September 2009 (Table 1). Among the profiles, HA, Hand BA are from the Borgarfjo¨r ð ur catchment (West Ice-land; Fig. 1), GA is from an experimental site just northof Reykjavik, and V is located South of Langjo¨kull (SouthWest Iceland; Fig. 1). Profiles HA and H are poorly drainedsoils (wetlands) relative to the freely drained profile BA andwell-drained profile GA and V (Table 1). All sites are cov-ered by grasslands, and peats were developed at the HA andH sites.The climate is temperate, with a mean annual precipita-tion (MAP) of 1017 mm yr  1 and mean annual temperature(MAT) of 4.6   C (station Hvanneyri, 1999–2010; IcelandMeteorological Office, IMO). In 2009 and 2010 (years of sampling), the specific annual precipitations were 901 and640 mm yr  1 , respectively. The studied soils are lowlandIcelandic soils characteristic of the island’s subarctic re-gions, with a warmer mean annual temperature (MAT =2 to 5   C) than the arctic climate of the highland Icelandicsoils (MAT =  2 to 2   C; Arnalds, 2004; Guicharnaud,2009). The biologically dormant cold season extends over7–9 months, whilst the growing season is short (over 3– 5 months) from early May until the end of August (Gui-charnaud, 2009). Due to the maritime winter climate of Iceland, the soils are exposed to significantly more freeze– thaw cycles than any other subarctic regions (Orradottiret al., 2008).The geology of Iceland is dominated by basalts which in-crease in age from the centre of Iceland towards the westernand eastern sides of the island, on both side of the Atlanticridge oriented SW-NE. Basalts from West Iceland in theBorgarfjo¨r ð ur catchment are Tertiary (older than 3.1 My;Hardarson et al., 2008). Soils in West Iceland receive alow influx of aeolian deposition of volcanic ash(  0.1 mm yr  1 ; Sigfusson et al., 2008) relative to the ashdeposition rates on soils in South West Iceland, closer tothe rift zones (  2 mm yr  1 ; Arnalds, 2004).Iceland is covered by   48% Andosols (Brown, Histicand Gleyic Andosols),   40% Vitrisols and Leptosols, and  1% Histosols, while glaciers and lakes comprise   11%(Arnalds, 2004). Areas dominated by vegetation are cov-ered by Andosols, desert areas by Vitrisols and wetlandareas by highly organic Histosols (Arnalds, 2008). Ando-sols cover around 1.9% of the Earth, but contain some5% of carbon that is stored in soils (Eswaran et al., 1993).The andic properties of the soils and the cold climate (meansummer temperature of 12   C) are factors favouring lowrates of organic matter decomposition in Iceland (Guichar-naud, 2009). In addition, oxidation is impaired in poorlydrained areas. Consequently in subarctic and arctic soils,the accumulation of plant debris results in the progressivebuildup of peat in poorly drained soils (Histic Andosoland Histosol). 3. METHODS3.1. Rock and soil sampling and characterisation 3.1.1. Basalt and soil sampling  The parent basalt was sampled at the BA site, and this isrepresentative of the typical Tertiary tholeiitic basalts(Moulton et al., 2000; Hardarson et al., 2008) that are the Fig. 1. Soil map of Iceland with selected soil sites. Five typical Icelandic soil types: Histic Andosol, HA; Histosol, H; Brown Andosol, BA;Gleyic Andosol, GA; Vitrisol, V. The sites of BA, HA and H are in the Borgarfjo¨r ð ur catchment (shown by the square), GA is North of Reykjavik, and V is South of Langjo¨kull. The soil map is based on Arnalds (2004) and Arnalds and Gretarsson (2001). The Histic Andosol studied in Pogge von Strandmann et al. (2012) in Hvalfjo¨r ð ur fjord, South of Borgarfjo¨r ð ur catchment is located for comparison (grey star).112 S. Opfergelt et al./Geochimica et Cosmochimica Acta 125 (2014) 110–130  parent material for HA, H, BA, GA. The horizon C in theVitrisol was considered as unweathered parent volcanic ashfor that soil. Each soil profile was described following FAOguidelines and sampled by horizon (IUSS, 2006). Soil sam-ples were air-dried and sieved at 2 mm to recover the frac-tion representative of the bulk soil for further physical,chemical and mineralogical characterisation at Universite´catholique de Louvain (UCL, Belgium). 3.1.2. Physical properties The bulk density was determined by collecting andweighing a fixed soil volume using the core method (Blake,1965). The particle size distribution was determined byquantitative recovery of clay (<2  l m), silt (2–50  l m) andsand (>50  l m) fractions after sonication and dispersionwith Na + -saturated resins (Rouiller et al., 1972). Recovered clay fractions were pre-treated prior to further characterisa-tion: organic matter was removed by H 2 O 2  treatment andexchangeable cations were replaced by SrCl 2  saturation(Hinckley and Bates, 1960). 3.1.3. Chemical properties Soil pH values were measured on bulk soils in deionisedwater with a ratio of 5 g of soil for 25 ml of solution. Thetotal soil organic carbon content was determined by gaschromatography after dry combustion, using a ThermoFinnigan CHN autoanalyzer. The exchangeable acidity of the soil (sum of Al 3+ and H + exchangeable) was quantifiedby percolation with 1 M KCl following the Bremer method(Page et al., 1982). Exchangeable cations (Ca, Mg, K, Na)were collected from bulk soils by percolation with ammo-nium acetate 1 M at pH 7 (Page et al., 1982) and then quan-tified by atomic absorption spectroscopy; this techniqueused to collect the exchangeable Mg was specifically testedfor Mg isotope measurements (Bolou-Bi et al., 2012). Thecation exchange capacity of the soil (CEC, expressed in cen-timole of charge per kilogram, cmol c  kg  1 ), reflecting theamount of negative charges available on the surface of soilparticles (organic and clay mineral) potentially available toretain cations, was determined on bulk soils by quantifyingthe desorption of the ammonium acetate 1 M at pH 7 (Pageet al., 1982). 3.1.4. Mineralogical properties Major elements were determined in basalt, bulk soilsand pre-treated clay fractions by atomic emission spectros-copy (ICP-AES, UCL, Belgium) after Li-borate fusion(Chao and Sanzolone, 1992). The total contents of alkalineand alkaline-earth cations were summed as the total reservein bases (TRB = [Na] + [Mg] + [Ca] + [K]) to estimate thecontent of weatherable minerals in soil horizons (Herbillon,1986); the TRB value in soils decreases with increasingweathering.Selective extraction techniques of Si, Al, and Fe in bulksoils were used for the characterisation of poorly crystallinephases, undetected or poorly identified by X-ray diffraction(XRD). Separate and independent extractions using darkoxalate (Si o , Al o , Fe o : ammonium oxalate 0.2 M pH 3;Blakemore et al., 1981) and DCB (Si d , Al d , Fe d ; Mehraand Jackson, 1960) were performed, followed by ICP-AES determinations. Oxalate generally extracts phases thatinclude short-range ordered minerals (i.e., poorly crystallineFe-oxides and aluminosilicates such as ferrihydrite andallophane, respectively), whereas dithionite-citrate-bicar-bonate (DCB) extracts poorly crystalline and crystallineFe-oxides (Schwertmann and Taylor, 1989; Borggaard,1990; Shoji et al., 1993; Cornell and Schwertmann, 1996).The allophane content was determined by multiplying Si o by 6 (Parfitt, 1990), but the oxalate extraction data is con-sidered with caution since volcanic glass might also bepartly dissolved especially at pH below 6 (Oelkers and Gis-lason, 2001; Arnalds and Gı´slason, 2002; Wolff-Boenischet al., 2004).The mineralogy of the crystalline phases was investi-gated by XRD (Cu K a , Bruker D8). For primary minerals,bulk soil powders were analysed after organic matter re-moval and oxalate extraction, and for secondary phases,DCB-treated clay fractions were analysed after K + andMg 2+ saturation, ethylene glycol solvation and thermaltreatments at 300 and 550   C (Robert and Tessier, 1974). 3.2. Soil solutions sampling and characterisation Soil solutions from these profiles (except in Vitrisol)were sampled over two different seasons (in September2009 and June 2010), using macro rhizon soil water sam-plers (length 9 cm, diameter 4.5 mm, porosity 0.2  l m; Eijk-elkamp). June (spring) and September (late summer) werechosen in order to sample the soil solutions before and afterthe growing season, respectively. Soil solution pH was mea-sured in the field. Major and trace element analyses wereobtained by ICP-MS (Open University, UK) in 2%HNO 3 . The accuracy was assessed by using the water refer-ence material SLRS-4 (Yeghicheyan et al., 2001). The ana-lytical precision was ±6% for major and ±7% for traceelements, with a detection limit <0.01 mM for major and<0.01  l M for trace elements. Silicon concentrations weremeasured by photospectrometry (±2%, University of Ox-ford, UK). Anion concentrations were determined by ion-chromatography (IC, ±3%; University of Oxford, UK for2009 solutions, and University of Southern California,USA for 2010 solutions). 3.3. Plant sampling and characterisation Grass-type vegetation (hummocky grassland) was sam-pled at each site. A bulk sample of shoot parts was madefrom five sub-samples taken at   2 m from the soil profile(360   C around the profile). Plants were dried at 60   C forone week. Major elements in plants were determined byICP-AES after Li-borate fusion at 1000   C and dissolutionof fusion beads in 10% HNO 3  (Chao and Sanzolone, 1992). 3.4. Mg isotope measurements Magnesium isotope compositions were determined forthe parental basalt, bulk soils (<2 mm), pre-treated clayfractions (<2  l m), exchangeable Mg, soil solutions, andvegetation. Prior to Mg isotope measurement, bulk sampleswere ashed (soils and plants) or reacted with H 2 O 2 S. Opfergelt et al./Geochimica et Cosmochimica Acta 125 (2014) 110–130 113  (exchangeable Mg and soil solutions) to remove organicmatter, then dissolved by acid digestion in suprapurHF:HNO 3  mixture. Dissolved samples were purified forMg isotope measurements through cation exchange resinfollowing the method described in Opfergelt et al. (2012)adapted from Wombacher et al. (2009). The proceduralblank was 8 ng, which is comparable to that of  Tenget al. (2007).Magnesium isotope ratios were measured by MC-ICP-MS (Nu Plasma HR, University of Oxford, UK) in dryplasma mode at low resolution. The instrumental mass biaswas corrected for by sample-standard bracketing, and thedata are expressed in relative deviations of   26 Mg/ 24 Mg fromthe DSM3 standard (Galy et al., 2003) using the common  d -notation ( & ): [( 26 Mg/ 24 Mg) sample /( 26 Mg/ 24 Mg) DSM3  1]  1000. The external reproducibility for  d 26 Mg for theDSM3 standard is ±0.15 & , 2SD. Each sample was ana-lysed 9 times, for which each single  d -value represents onesample run and two bracketing standard runs. The accu-racy of the mass bias correction was assessed using a threeisotope plot ( d 25 Mg vs.  d 26 Mg), and all samples lie on amass-dependent fractionation array. Long term precisionand accuracy were assessed from multiple measurementsof the reference material Cambridge-1 with a  d 26 Mg of   2.60 ± 0.15 &  (2SD,  n  = 576) over 18 months. In addi-tion, USGS rock standard BHVO-2 was analysed repeat-edly ( d 26 Mg of    0.31 ± 0.19 &  (2SD,  n  = 30)). Bothresults are indistinguishable from previously published val-ues (Galy et al., 2003; Pogge von Strandmann et al., 2008;Huang et al., 2009). 4. RESULTS4.1. Key properties of basalt and soils Based on their chemical (Table 2) and mineralogical(Table 3) properties, the five soil types can be divided intotwo groups: the freely drained soils V–BA–GA are charac-terised by a neutral soil pH and a low C content (pH H 2 O 6.7 ± 0.7; 5.4 ± 2.9% C), and the poorly drained soilsHA–H are acidic and organic-rich (pH H 2 O  4.8 ± 0.6;21 ± 9% C) (Fig. 2a). The CEC increases from 18 to 41to 53 cmol c  kg  1 from V to BA–GA to HA–H. This is re-lated to an increase in organic carbon content from 0.3%in V, to 6.4% in BA–GA, and to 21% in HA–H, directlyincreasing the negative charges associated to organic mat-ter. The distribution of base cations on the soil exchangecomplex is dominated by Ca and Mg in all soils (69.0%Ca > 25.4% Mg > 4.1% Na > 1.5% K; Table 2). The soil ex-change complex in soils is not entirely saturated by basecations: the base saturation (BS = ratio of the sum of exchangeable bases [Ca + K + Mg + Na] to CEC; Table 2)is higher in V (50%) than in other soils (24%). This is mainlyexplained by the soil pH decreasing from V to HA–H,thereby increasing the amount of protons on the soil ex-change complex in acidic soils and limiting the retentionof base cations (soil exchange acidity of 1.7 cmol c  kg  1 inHA–H against 0 cmol c  kg  1 in other soils; Table 2).The mineralogy of primary minerals in these soils mea-sured by XRD analysis (augite, Ca-rich plagioclase, andmagnetite) directly reflects that of the tholeiitic basalt, typ-ical of the Tertiary basalts of West Iceland (Moulton et al.,2000; Hardarson et al., 2008). Based on selective extractionswith oxalate and DCB (Table 3), secondary weatheringproducts are dominated by short-range ordered minerals(allophane and ferrihydrite), which is in good agreementwith previous work on soils from this area (Crovisieret al., 1992; Wada et al., 1992; Moulton et al., 2000; Ste- fansson and Gislason, 2001; Arnalds, 2004; Sigfussonet al., 2008). The allophane content is higher in V–BA– GA (12.3%) than in HA–H (5.4%; average values from Ta-ble 3). Allophane forms above pH 4.9 when Al is not form-ing a complex with humus (Mizota and van Reeuwijk,1989). Therefore, the allophane content is higher in the neu-tral V–BA–GA soils than in the more acidic HA–H soils.The presence of allophane at pH < 4.9 in HA–H is best ex-plained by the aeolian deposition of allophanic materialsand/or by higher pH conditions earlier in the developmentof the soil (Arnalds, 2004). The proportion of Fe-oxides islarger in HA–H than in V–BA–GA (ratio Fe DCB overthe total Fe content: Fe d /Fe t  = 0.75 and 0.42, respectively;average values from Table 3). Crystalline clay minerals,such as smectite and kaolinite, were only detected byXRD analysis in the HA soil (Table 4). Depending on thereaction progress, weathering of basaltic glass can resultin the formation of different secondary mineral assemblages(Crovisier et al., 1992; Stefansson and Gislason, 2001): the progressive dissolution of basaltic glass is first associatedwith secondary mineral assemblages dominated by Fe andAl-hydroxides, and later, with increasing Si concentrations,the formation of crystalline clays kaolinite and smectite.Relative to the basalt (TRB = 733 cmol c  kg  1 ; Table 3),HA and H are more depleted in base cations(TRB = 166 cmol c  kg  1 ; Table 3) than V, BA and GA(TRB = 418 cmol c  kg  1 ; Table 3). HA–H soils also displayhigher clay content (48%; Table 2) than V, BA and GA(30%; Table 2). Overall, this supports a higher degree of weathering in HA–H relative to V–BA–GA soils(Fig. 2b). This is consistent with a lower bulk soil densityin HA–H than in BA–GA (average of 0.3 and 0.8 g cm  3 ,respectively; Table 2), and with the higher proportion of Fe-oxides in HA–H (Fe d /Fe t  = 0.75 in HA–H and 0.42 inV–BA–GA; Table 3). 4.2. Magnesium distribution in soils The average value of the total Mg content (Mg t ) in soilsis higher in V–BA–GA (20 g kg  1 ) than in HA–H (6 g kg  1 ;Table 3), and is distributed between primary minerals, clayfractions (<2  l m), and the soil exchange complex (Fig. 3a).The amount of Mg in the clay fraction (Mg clay fraction ) is di-rectly measured as the total amount of Mg in pre-treatedgranulometric clay fractions (<2  l m; Table 4). The quantityof exchangeable Mg (Mg exch ) is a direct measurement thatincludes the exchangeable Mg retained on secondary alumi-nosilicates, humus and fresh organic matter (Table 2). Theamount of Mg in primary minerals (Mg primary ) can be cal-culated by subtracting the amount of Mg clay fraction  andMg exch  from the total Mg content measured in the bulksoils (Mg t ; Table 3). This calculation indicates that magne- 114 S. Opfergelt et al./Geochimica et Cosmochimica Acta 125 (2014) 110–130
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x