Fate of Cd in agricultural soils: A stable isotope approach to anthropogenic impact, soil formation and soil-plant cycling

Martin Imseng1, Matthias Wiggenhauser2, Armin Keller3, Michael Müller3, Mark Rehkämper4, Katy Murphy4, Katharina Kreissig4, Emmanuel Frossard2, Wolfgang Wilcke5, Moritz Bigalke1*

1Institute of Geography, University of Bern, Hallerstrasse 12, CH-3012 Bern, Switzerland 2Institute of Agricultural Sciences, ETH Zurich, Eschikon 33, CH-8315 Lindau, Switzerland 3Swiss Soil Monitoring Network (NABO), Agroscope, Reckenholzstrasse 191, CH-8046 Zürich, Switzerland

4Department of Earth Science & Engineering, Imperial College London, SW7 2AZ London, U.K. 5Institute of Geography and Geoecology, Karlsruhe Institute of Technology (KIT), Reinhard-Baumeister-Platz 1, 76131 Karlsruhe, Germany

*Corresponding author: Moritz Bigalke, moritz.bigalke@giub.unibe.ch, tel. +41(0)316314055

1

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

123

Abstract

The application of mineral phosphate fertilizers leads to an unintended Cd input into agricultural systems, which might affect soil fertility and quality of crops. The Cd fluxes at three arable sites in Switzerland were determined by a detailed analysis of all inputs (atmospheric deposition, mineral P fertilizers, manure and weathering) and outputs (seepage water, wheat and barley harvest) during one hydrological year. The most important inputs were mineral P fertilizers (0.49 to 0.57 g Cd ha-1 yr-1) and manure (0.20 to 0.91 g Cd ha-1 yr-1). Mass balances revealed net Cd losses for cultivation of wheat (-0.01 to -0.49 g Cd ha-1 yr-1) but net accumulations for that of barley (+0.18 to +0.71 g Cd ha-1 yr-1). To trace Cd sources and redistribution processes in the soils, we used natural variations in the Cd stable isotope compositions. Cadmium in seepage water (δ114/110Cd = 0.39 to 0.79‰) and plant harvest (0.27 to 0.94‰) was isotopically heavier than in soil (-0.21 to 0.14‰). Consequently, parent material weathering shifted bulk soil isotope compositions to lighter signals following a Rayleigh fractionation process (ε ≈ 0.16). Furthermore, soil-plant cycling extracted and moved isotopically heavy Cd from the subsoil to the topsoil. These long-term processes and not recent decreasing anthropogenic inputs determined the Cd distribution in our soils.

2

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

456

TOC Art

3

39

40

789

Introduction

The application of mineral P fertilizers leads to an unintended Cd input into agricultural soils. This Cd can be stored in the soil, leached with seepage water or taken up by crops and thus enter the human food chain.1 However, Cd is toxic for plants and humans and accumulates in human bodies.2 Even low Cd concentrations in edible plant parts can pose a risk for human health because its biological half-life is 10-30 years.3

Cadmium is a natural constituent of soil parent material, which is physically and chemically weathered during soil formation. In crustal rocks, Cd concentrations vary between 0.01 and 2.6 mg kg-1 with typically higher abundances in sedimentary than in igneous rocks.4 Weathering of parent material depends on the soil forming factors5 and is the quantitatively most important natural Cd source to soils. Typical Cd concentrations in uncontaminated soils range from 0.1 to 1.0 mg kg-1.6

Another Cd source to soils is atmospheric deposition. Cd in the atmosphere can originate from natural sources like local transport of particles, long-range transport of dust, and volcanic activities but also from anthropogenic emissions.7 Anthropogenic Cd emissions were strongly correlated with both, air Cd concentrations and atmospheric Cd deposition rates to terrestrial surfaces.8 Moreover, a strong correlation between industrial Cd uses and environmental Cd concentrations was revealed by peat cores.7,9 In industry, Cd is used in Ni-Cd batteries, pigments, coatings and in stabilizers for plastic and nonferrous alloys.10 Total industrial Cd emissions peaked in the 1960s and decreased thereafter, in Europe.8

Cd is additionally added to agriculturally used soils with mineral P fertilizers, manure and sewage sludge application. These fertilizers were increasingly applied during the 20th century, after the intensification of agricultural practices.11 As a result, Cd inputs to agricultural soils increased.12,13 In mineral P fertilizers, Cd concentrations vary and reflect the different Cd concentrations in rock phosphates.14 Imports of such fertilizers peaked in 1980 and decreased afterwards by a factor of ~4 till

4

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

101112

2008.15 Still, Cd inputs through mineral P fertilizers are a relevant soil pollution pathway, depending on the Cd concentrations and application rates.12,16,17 In manure, Cd is less concentrated than in mineral P fertilizers and reflects the Cd concentrations of the animal diet including crops, pasture grasses and herbs, and feed additives.18,19 However, also high manure application rates can considerably increase Cd inputs to soils.20,21,22

Sewage sludge is an additional relevant Cd source and Cd concentrations depend on its origin and quality. Sewage sludge application to agricultural soils was prohibited in Switzerland in 2006; nevertheless, earlier application might have contributed considerably to the Cd content of agricultural soils.

The most important Cd outputs from arable soils are with seepage water and crop harvest. First, the output with seepage water is determined by the Cd concentration and the amount of water. Soil solution Cd concentrations are primarily controlled by sorption processes.6 The pH of the soil is thereby the main factor determining soil solution Cd concentrations, followed by the bulk soil Cd concentration.5,23,24 The amount of seepage water depends on the water balance of a soil (precipitation, evapotranspiration and soil water content change). Previous studies assumed constant Cd leaching fluxes25,26 or calculated them based on laboratory adsorption experiments and meteorological data.27 In contrast, this study presents, to our knowledge, the first Cd leaching fluxes calculated with in-situ measured water flux data. Second, output with crop harvest is controlled by crop Cd concentrations,19,26,27 which vary because of the different acquisition and sequestration strategies of plants.6,18,19 It has been shown, that the most important driver of Cd concentrations in plants is the soil Cd concentration, followed by soil pH.6

In soils, Cd derives partly from geogenic sources and partly from anthropogenic inputs of the past. Hence, tools are needed to better distinguish between these different sources. A well-established biogeochemical tool for tracing metal contaminants in the environment is the stable isotope composition.28 The δ114/110Cd values of terrestrial rocks

5

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

131415

and minerals show only limited variability (-0.4 and 0.4‰).29,30,31,32 In contrast, industrial processes can generate substantially larger Cd isotope fractionations (-2.3 to 5.8‰),30,33,34 mainly through partial evaporation and condensation of the metal.30,33,35,36 Notably, soils and sediments near smelters are commonly enriched in anthropogenic Cd. In such environments, stable isotopes have been used to differentiate between anthropogenic and geogenic Cd.35,37,38,39 Moreover, despite smaller isotopic variabilities, a recent study successfully used stable isotopes to trace Cd from mineral P fertilizers in agricultural soils.40

Beside different sources, natural processes can also produce pronounced variations in Cd isotope compositions of agricultural soils. First, processes between solid-phases and liquid-phases lead to the enrichment of heavy isotopes in solutions. For example, after natural weathering, river sediments were more enriched in heavy isotopes than riverbank soil (∆114/110Cdsoil-stream sediment ≥ -0.50‰).41 Similarly, after simulated weathering, Cd in leachates was isotopically heavier than Cd in Pb-Zn ores (∆114/110CdPb-

Zn ore-leachate = -0.53 to -0.36‰).41 Other studies examined Cd isotope fractionation during Cd adsorption to Mn oxyhydroxides and Cd co-precipitation with calcite.42,43 In both cases, the dissolved Cd was isotopically heavier than the adsorbed Cd (∆114/110Cdsolid-liquid = -0.54 to -0.24‰)42 and the co-precipitated Cd (∆114/110CdCaCO3-Cd(aq) ≈ -0.45‰).43 Second, also biological processes cause isotopic Cd fractionation. For example, phytoplankton is preferentially taking up light Cd isotopes,44 leaving residual seawater Cd isotopically heavier (~0 to 3.8‰).45 Similarly, Cd-tolerant plants were more enriched in light isotopes than hydroponic solutions (∆114/110Cdplant-solution = -0.70 to -0.22‰).46 However, Cd in plants was isotopically heavier than Cd in bulk soils (∆114/110Cdsoil-wheat = -0.39 to -0.13).47 This might be an effect of the isotopically heavier Cd in liquid-phases compared to solid-phases;41,42,43 because plants take mainly up Cd from soil solutions.

During the 20th century, the Cd concentrations of European soils have increased, by about a factor of 1.3 to 2.6.13 In contrast, models predict a

6

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

161718

reversal of this trend, such that Cd concentrations are expected to remain constant17 or even decrease,13,48 in European soils over the next 100 years. However, mass balances based on in-situ measured data are lacking. Furthermore, there is only one study which used stable isotopes to trace Cd sources in agricultural soils.40

Here, we used in-situ measured data to establish Cd mass balances for three arable study sites. Soil Cd concentrations and all Cd inputs (atmospheric deposition, mineral P fertilizers, manure and parent material) and outputs (seepage water, wheat and barley harvest) were determined during one hydrological year, from May 2014 to May 2015. In addition, novel approach that uses Cd stable isotope compositions was applied to evaluate the importance of anthropogenic Cd inputs and to investigate Cd cycling in the soils. The aims were to (i) determine, if Cd accumulates in soils under the current agricultural practice, (ii) differentiate between anthropogenic and natural Cd in the soil, and (iii) understand Cd redistribution processes within the soils.

7

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

192021

Materials and Methods

See more details in the SI.

Study sites

More details on the materials and methods can be found in the SI. The study was carried out at two arable monitoring sites (Figure S1) of the Swiss Soil Monitoring Network (NABO)49 in Oensingen (OE) and Wiedlisbach (WI), situated on the Swiss Plateau. In addition, one arable monitoring site was chosen from the cantonal soil monitoring network Basel-Land in Nenzlingen (NE), located in the Swiss Jura. These three sites were selected because of contrasting geology, soil properties and Cd concentrations in the soils. The lowest Cd concentrations were found in WI (0.13 to 0.17 mg kg-1) and the highest in NE (0.97 to 1.66 mg kg-1, Table S1). The soils developed on calcareous alluvial deposits (OE), mixed calcareous, siliceous moraine material (WI) and limestone (NE), respectively. At OE, the soil is classified as a stagnic calcaric eutric fluvic Cambisol, WI is a eutric Cambisol and NE a leptic calcaric eutric Cambisol.

Sampling

Soil samples were taken from four fixed depths (0-20 cm, 20-50 cm, 50-75 cm and >75 cm). Inputs and outputs (Figure 1) were sampled during one hydrological year between May 2014 and May 2015, barley harvest samples were taken after that period, in July 2015. Soil parent material was obtained at each site. The C horizon was sampled at OE (240 to 270 cm depth) and WI (110 to 130 cm depth), at NE, limestone samples were collected from the soil surface. Mineral P fertilizers were obtained from the farmers for each application while liquid cattle manure was sampled once at OE and WI. No manure was sampled at NE but formerly reported Cd concentration data were used for calculations.19 Atmospheric deposition and seepage water were sampled cumulatively every second week, whilst the volumetric water content of the soil was determined with 1-h

8

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

222324

resolution by time domain reflectometry at 50 cm soil depth. Plants were sampled during two cropping seasons (wheat harvest in summer 2014, barley harvest in summer 2015), with roots and shoots (aboveground plant material, consisting of straw and grains) of plants harvested at full maturity.

Laboratory analysis

Basic soil properties including pH, cation-exchange capacity (CEC), texture, C, N, and S concentrations and bulk density were determined and the soils were characterized according to the World Reference Base for Soil Resources.50 Soil, parent material, plant, manure and mineral fertilizer samples were digested using a microwave oven (ETHOS, MLS, Leutkirch, Germany). Cadmium concentrations were determined for the sample digests, atmospheric deposition and seepage water by inductively-coupled mass spectrometry (ICP-MS, 7700x, Agilent Technology, Waldbronn, Germany). Titanium (Ti) concentrations were additionally measured in digests of soil and parent material samples to calculate Cd mass gains or losses per unit volume of soils relative to parent materials (τCd values, Equations S1 and S2).51

The stable Cd isotope compositions of all samples were determined as described in detail in the SI using a double-spike technique by multiple collector inductively-coupled plasma mass spectrometry (MC-ICP-MS, Nu Plasma HR, Nu Instruments Ltd, Wrexam, UK).47,52,53 The total procedural Cd blank (n=11) for the isotopic analyses ranged from 110 to 1011 pg. This is equivalent to less than 2.5% of the smallest indigenous Cd mass among the samples, whilst the typical blank proportion was about 0.4%. Hence, no blank corrections were required for the isotopic data. Several standard reference materials (SRMs) were analyzed together with the samples for quality control and the results showed good agreement with published values (Table S3). The double-spike method also yielded precise Cd concentrations.54 For the SRMs, the measured Cd concentrations were

9

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

252627

slightly lower than the certified values but our data are in in line with the results of other recent studies.55,56,57

The Cd isotope compositions of the samples are reported relative to the NIST 3108 Cd isotope reference material using a δ notation based on the 114Cd/110Cd ratio (Equation S2). Two samples were considered significantly different in their isotopic composition if the results differed by more than 2x the standard deviation of each sample. The ∆114/110Cd values, which denote the apparent isotopic fractionation between two reservoirs and/or two fluxes (e.g., between soil and seepage water) were calculated according Equation S3.

Mass balance calculations

Individual Cd abundance mass balances were calculated for each study site soil, considering inputs from weathering, atmospheric deposition, mineral P fertilizers, manure, and outputs through seepage water and crop harvest (wheat and barley). Input from weathering was thereby calculated from dissolution of the coarse soil (>2 mm) which introduces Cd to the bulk soil. Separate mass balances for wheat and barley cultivation were calculated for each soil (Figure 1, Table S2).

Additionally, stable isotope mass balances were calculated. The isotope composition of a mass balance reservoir or flux (input or output) is composed of several fractions (e.g., wheat harvest = straw harvest + grain harvest) and therefore the mean isotopic composition had to be calculated. The mean value for each reservoir or flux was calculated with Equation 1.

Equation1 :δ 114 /110Cd=∑f=1

n

δ114 /110Cd f ∙mCd f

∑f=1

n

mCd f

δ114/110Cd: isotopic composition of the reservoir or fluxδ114/110Cdf: isotopic composition of the fraction of the reservoir or flux

10

213

214

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

234

235

236

237

238

239240

282930

mCdf: Cd mass in the fraction of the reservoir or flux

The isotope mass balance of each study site was calculated with the same method, with arbitrarily defined mCdf > 0 for bulk soil and inputs and mCdf

< 0 for outputs, and this yields new predicted δ114/110Cd values for the bulk soils. The budgeted unit was the 0-50 cm soil layer. To estimate long term changes in bulk soil isotope compositions, the balances were extrapolated over 1 to 100 hydrological years, with alternating wheat and barley cultivation. Because Cd inputs from fertilizer use and atmospheric deposition might have been higher than today for a significant time period in the last century, an additional scenario was calculated. This assumes the highest possible Cd inputs for the last century and examines the impact on bulk soil isotope compositions. Error propagation was calculated for each step.

Model calculations

During soil formation, parent material was physically and chemically weathered and isotopically heavy Cd was leached with seepage water. This closed-system kinetic Cd isotope fractionation was described with the Rayleigh model (Equation 2) and named as parent material weathering.

Equation2 :∆110/114Cd soil− parentmaterial=ε ln f

ε: Rayleigh fractionation factor for soil formationf: Remaining Cd fraction in the soil, relative to the parent material (τCd values + 1)

A soil-plant cycling model was used to test the hypothesis that trees, which formerly covered the agricultural soils,58,59,60,61 took up Cd from the deeper and added it to the upper soil horizons, over the whole time of soil formation. First, the soils were subdivided in two boxes (0-35 cm and 35-75 cm), and remaining Cd fractions (τCd values + 1) and δ114/110Cd values were averaged for both boxes (named as “current values in 2015”). Furthermore, the Cd surplus in the upper (0-35 cm) relative to the deeper horizon soil-box (35-75 cm), presumably cycled by trees, was calculated

11

241242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262263264

265

266

267

268

269

270

271

272

313233

with the help of τCd values. This Cd surplus was divided by the age of the soils (13700 years) to get the annually cycled Cd. Afterwards, the annually cycled Cd was subtracted from the upper and added to the deeper soil-box, in 13700 steps, in the reverse direction as trees did it before. Additionally, the Cd isotopic composition change of the two soil-boxes was calculated (Equation 1), by subtracting the Cd isotopic composition of the annually cycled Cd from the upper and adding it to the deeper soil-box in 13700 annual steps. The isotopic composition of the cycled Cd was thereby determined for each calculation step with ∆114/110Cdsoil-trees = -0.25‰. After 13700 calculation steps, the remaining Cd fractions and δ114/110Cd values were averaged for the two soil-boxes and named as “values 2015 without soil-plant cycling”. Error propagation was calculated according Equations S6-S9.

12

273

274

275

276

277

278

279

280

281

282

283

284

285

343536

Results and Discussion

Cd abundance mass balances

Input fluxes

Cd input from weathering was only important at NE where it accounted for ~17% of total Cd inputs; in contrast, this share was less than 1% at OE and WI (Figure 1, Table S2). The high Cd input from weathering at NE can be attributed first to the type of coarse soil (limestone), second to the high coarse soil volumetric content (5-9% for the two upper soil layers) and third to the high Cd concentrations in limestone (Table S1). At OE, weathering was relatively unimportant because the coarse soil volumetric content was below 0.5% (Table S1). At WI, the coarse soil volumetric content (>6%) was higher than at OE. Nevertheless, the Cd input with weathering was negligible for the mass balances. The reason for this was the siliceous parent material at WI for which weathering rates62,63 were three orders of magnitude smaller than those for calcareous material (at OE and NE).64 Comparable weathering rates were found by previous studies.62,63,64

The Cd input (0.11 ± 0.00 g ha-1 yr-1) from atmospheric deposition and the Cd concentration in atmospheric deposition (~0.01 μg L-1) were similar among the three sites. This input contributed between 7 and 11% to the total Cd inputs. Based on air concentration measurements, Keller et al.19

estimated a median Cd atmospheric deposition rate of ~0.7 g ha-1 yr-1 for Swiss soils in 2003, which is ~7 times higher than the deposition rates found in this study. More recent studies reported lower rates between 0.2 and 0.4 g ha-1 yr-1.65,26 The difference between these estimates and our results most likely reflects the further reduction of anthropogenic Cd emissions in Europe in the past decade.8 In addition to this atmospheric deposition, there might be dry deposition, which we did not quantify and which is also not considered in the literature about arable soils.

13

286

287

288

289

290

291

292

293

294

295

296

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

373839

For all study sites, fertilization was the quantitatively most important Cd input. At OE and NE, Cd inputs from mineral P fertilizers were higher than inputs from manure, while the reverse was true for WI. Application of mineral P fertilizers accounted for 32% to 70% of total Cd inputs with fluxes of 0.75, 0.49 and 0.57 g ha-1 yr-1 for OE, WI and NE, respectively. Similar input rates of 0.10 to 0.79 g ha-1 yr-1 were previously reported for other European soils.13,26,65 However, with the exception of one mineral P fertilizer that was applied at OE, the Cd concentrations of the mineral P fertilizers were found to be below the average value of 67 mg Cd (kg P) -1

determined for such fertilizers in Switzerland.66 Due to the highly variable Cd concentration of mineral P fertilizers (<1 to 213 mg (kg P) -1), the choice of fertilizer will have significant impact on the Cd input. If all mineral P fertilizers that were applied on our study sites had featured the mean Swiss Cd concentration of 67 mg Cd (kg P)-1, the Cd inputs from such fertilizers would have been 45% (OE), 173% (WI), and 150% (NE) higher than the values determined here. Manure application (0.20 to 0.91 g ha -1

yr-1) accounted for 18% to 60% of total Cd inputs. In contrast to the mineral P fertilizers, the Cd concentrations of manure were homogeneous at our sites (180-185 ng g-1). However, they can vary considerably and reach much higher concentrations.19

Output fluxes

The highest Cd output with seepage water was observed at WI (0.99 g ha -1

yr-1, Figure 1, Table S2) where it accounted for >65% of total outputs. The reason for this were the seepage water concentrations which were highest at WI (0.156 μg L-1), followed by NE (0.011 μg L-1) and OE (0.003 μg L-1). The high Cd concentrations of seepage water at WI were related to the low pH. The Cd seepage water flux for that site agrees well with the literature (0.4 to 1.6 g ha-1 yr-1).26,27

For OE and NE, the crop harvest was a more important Cd output than leaching and accounted for more than 93% of the total Cd output at both sites and for both crops (wheat and barley). The higher the soil Cd concentration was, the higher was the output with crop harvest with the

14

315

316

317

318

319

320

321

322

323

324

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

404142

highest at NE (1.77 and 0.57 g ha-1 yr-1 for the wheat and barley harvests, respectively) followed by OE (1.47 and 0.33 g ha-1 yr-1) and WI (0.52 and 0.33 g ha-1 yr-1). These results agree well with the findings of previous studies concerning the coupling of soil Cd concentrations with crop Cd concentrations67,68,69 and crop Cd outputs.13,26,27 For all three sites, Cd abundances of wheat and Cd outputs with wheat harvest were higher than the respective values for barley. At WI, the Cd output difference between the harvesting of the two crops was smallest because barley provided higher crop yield than wheat.

Budget

The most influential driver for the soil Cd budgets is the crop species, because wheat and barley cultivation were associated with net Cd losses and net Cd accumulations, respectively, at all three sites (Figure 1, Table S2). Second, mineral P fertilizer Cd concentrations and application rates are other influential variables for the mass balances. The Scientific Committee on Toxicity, Ecotoxicity and the Environment (CSTEE) stated in 200270 that the application of mineral P fertilizers with Cd concentrations below 50 mg (kg P)-1 will most probably not lead to Cd accumulations in soils, which is supported by our findings. However, the use of fertilizers with higher Cd concentrations (these can be up to 213 mg (kg P)-1)66 will probably lead to Cd accumulations in soils (Figure S3). Thus, it is important to enforce the legal limit of 50 mg (kg P)-1 or introduce limits in countries where no such regulation exists. Finally, the Cd accumulation in soils depends also on other properties of the agricultural system like soil pH, which influences Cd output with seepage water, and manure application.

Contributions of natural and anthropogenic Cd sources to the total soil Cd

The Cd distribution in soils is the result of Cd inputs (weathering, atmospheric deposition, mineral P fertilizers and manure), Cd outputs (seepage water, crop harvest) as well as soil formation processes. Cd

15

347

348

349

350

351

352

353

354

355

356

357

358

359

360

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

434445

stable isotopes are used here as a tool to assess the importance of natural and anthropogenic Cd sources in the studied soils.

The Cd isotope compositions of the inputs showed significant isotopic variability (δ114/110Cd = -0.15 to 0.38‰, Figure 2, Table S4). First, the parent materials differed in their Cd isotopic compositions, with the lowest δ114/110Cd values recorded at WI (- 0.14 ± 0.10‰) and higher values at OE (0.04 ± 0.06‰) and NE (0.36 ± 0.04‰). The Cd isotope compositions of the bulk soils were not significantly different from the parent material at OE and WI. However, the soils at NE were isotopically lighter than the parent limestone. Second, the Cd isotope compositions of atmospheric deposition at OE and NE were not significantly different. Bridgestock et al.71 found that marine atmospheric aerosols from the Tropical Atlantic Ocean were characterized by a relatively narrow range of Cd isotope composition (-0.19 to 0.19‰) and they were unable to differentiate between anthropogenic and natural Cd sources. The Cd isotope composition of the atmospheric deposition analyzed here is within the range of industrial waste materials (-0.64 to 0.46‰)30,35,39 but also in accordance with data for aboveground plant material (0.20 to 0.57‰)72,47

which might emit organic aerosols, and terrestrial minerals (-0.50 to 0.67‰).29,31 Thus, we were unable to identify the source of Cd in atmospheric deposition. However, a strong correlation between anthropogenic Cd emissions and atmospheric Cd deposition was shown by Pacyna et al.8 Third, the δ114/110Cd values of mineral P fertilizers (-0.15‰ to 0.15‰) were in the range of those of Earth crust minerals and rocks (-0.50 to 0.67‰).29,30,31 These results suggest that Cd is not fractionated during the manufacturing of mineral P fertilizers which is in line with a recent work on mineral P fertilizers in New Zealand.40 The similar isotope ratios of bedrock and mineral fertilizers furthermore render it difficult to trace Cd from mineral P fertilizers in agricultural soils, which was different to the work of Salmanzadeh et al.,40 where topsoils and phosphate fertilizers had clearly distinct Cd isotope compositions. Finally, the enrichment of heavy Cd isotopes in manure of OE and WI (0.35 to 0.38‰) is in line with the origin of the manure Cd. The cattle of the studied farms mainly fed on

16

378

379

380

381

382

383

384

385

386

387

388

389

390

391

392

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

408

409

410

464748

grass produced on the farm (either during grazing or as hay) and concentrated feedstuff (cereal grains). Pasture plants might show similar Cd fractionation patterns as wheat and barley, whose aboveground parts were enriched in heavy isotopes (0.38 to 0.94‰). Similar 114/110Cd values have been found in wheat shoots and grains (0.20 to 0.57‰)47 and birch leaves (0.70‰).72 It is unknown, however, whether Cd isotope fractionation occurs during digestion in the cattle rumen, but the small difference between the Cd isotope compositions of the manure and the plants suggests that any fractionation should be minor.

Cd in outputs was isotopically heavier than Cd in bulk soils (Figure 2, Table S4). First, seepage water from all three sites was enriched in heavy isotopes (0.39 to 0.79‰). These results are in line with findings from simulated and natural weathering,41 Cd adsorption to Mn-oxyhydroxides42 and calcite precipitation43 studies in which the liquid-phase Cd is always isotopically heavier than the solid-phase Cd. Second, the most important Cd output, which is associated with the wheat and barley harvest (straw and grains), was also enriched in heavy Cd isotopes (0.38 to 0.94‰) which is in line with the literature.72,47

Stable isotope mass balances (Equation 1) offer us a tool to assess the importance of the different in- and outputs on the Cd content of the soils and for these calculations, we simplified our agricultural systems and assumed that they were anthropogenic influenced only during the last 100 years. First, atmospheric deposition of Cd is driven mainly by anthropogenic Cd emissions,8 which have increased the Cd content of soils since 1846.12 Thus, before industrialization, atmospheric deposition rates of Cd will have been lower than at present (2015) and can likely be neglected for most of the soil formation period. This is supported by a number of studies which found that Cd enrichment factors in peat cores were ~20 times lower73 and Cd deposition rates at least one order of magnitude lower in preindustrial times compared to the last 20 years of the 20th century.74,75,76,77 Second, agricultural practices were intensified after World War 111 and this coincided with the use of mineral P fertilizers

17

411

412

413

414

415

416

417

418

419

420

421

422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

495051

and concentrated animal food,15 which are associated with a net import of Cd to the agricultural systems. Before that time, the agricultural systems can be considered as closed. Thus, inputs from fertilization did not exceed outputs from harvest78 and these outputs were by far lower than during the 20th century.11

The Cd isotope mass balances show that δ114/110Cd values in the 0-50 cm layer of our soils change less than 0.03‰ during 100 years with current (2014-2015) agricultural practice and atmospheric deposition (Figure 1, Table S2). At OE, δ114/110Cd would change from 0.10 to 0.08‰, with alternating wheat and barley cultivation. At WI, the δ114/110Cd would decrease from -0.18 to -0.21‰. The smallest influence of the different inputs and outputs on the bulk soil isotope composition would occur at NE with a decrease by 0.01‰ in 100 years. Furthermore, calculations on the maximal possible change for the Cd isotope composition of the bulk soils (0-50 cm) during the last century revealed changes of less than 0.05‰, which is smaller than the measurement error. These calculations based on (i) the evolution of European Cd emissions8, (ii) the Cd isotope compositions of industrial waste,30 (iii) Swiss mineral P fertilizer and feedstuff imports,15 and (iv) the Cd concentrations of mineral P fertilizers.66

These findings demonstrate that a much longer time scale is needed to produce significant changes in bulk soil Cd isotope compositions because known annual inputs and outputs are about 3 (OE and WI) to 4 (NE) orders of magnitude smaller than bulk soil Cd pools (0-50 cm). Therefore, not anthropogenic inputs and outputs but long-term fractionation processes during pedogenesis have controlled the Cd isotope compositions of the bulk soil. Consequently, the isotopic compositions of our soils can be used to investigate long-term soil formation processes.

Redistribution of Cd in the soil

Looking at the predominant part of the soil formation period (13700 to 100 years B.P.), i.e. during pre-agricultural and pre-industrial times, we can

18

443

444

445

446

447

448

449

450

451

452

453

454

455

456

457

458

459

460

461

462

463

464

465

466

467

468

469

470

471

472

473

525354

assume our soils to be semi-closed systems, where Cd leaching with seepage water was the only output flux. The τCd values of our soils indicate (except for 0-20 cm at OE and WI) that 25-86% of the initial Cd in the parent material was lost, most likely with seepage water. Seepage water was more enriched in heavy isotopes than the soils (∆114/110Cdsoil-seepage water = -0.59 to -0.69‰). Consequently, bulk soil Cd isotope compositions shifted to lighter values relative to the parent material. Assuming that agricultural inputs/outputs and atmospheric deposition did not significantly influence the Cd isotope compositions of the bulk soils, as outlined above, the evolution of the bulk soil isotope compositions can be described with a Rayleigh fractionation model.28 To this end, the remaining Cd fractions and ∆114/110Cdsoil-parent material values were plotted (Figure 3a). A best fit for the current soil data was thereby achieved with a Rayleigh fractionation factor (ε) of 0.16. At NE, we observed ∆114/110Cdsoil-parent material between -0.22 and -0.32‰. At OE and WI, less of the initial Cd was lost and ∆114/110Cdsoil-parent

material values were between -0.05 and 0.10‰. Consequently, the soil formation effect could be better observed at NE with a better model fit than at the two other sites (Figure 3a).

But, if only weathering and leaching influenced the Cd distribution in the bulk soils, the largest Cd losses and the lightest Cd isotope compositions should be found in the oldest uppermost horizons. The remaining Cd fractions in the soil, however, indicate an inverse distribution with apparently smaller losses of Cd in the upper than in the deeper horizons (Figures 3a). Thus, there must be a process that added Cd to the surface soil. Interestingly, the Cd surplus in the upper relative to the deeper soil horizons correlated with the Cd concentrations of the parent materials (Figure S4). Therefore, the inverse distribution of the Cd depletion, indicated by the remaining Cd fractions in soils, is most likely caused by a Cd redistribution within the soils rather than a net input from the outside. Previous studies already revealed the importance of the plant pump for the distribution of nutrients79,80 and also for Cd between C and O horizons.81,82,83 This pump also seems to be important at our sites.

19

474

475

476

477

478

479

480

481

482

483

484

485

486

487

488

489

490

491

492

493

494

495

496

497

498

499

500

501

502

503

504

505

555657

To assess the importance of the plant pump, a soil-plant cycling model was introduced (Figures 4, S6 and S7). In the model, the soils were subdivided in two boxes. Among all sites, the remaining Cd fractions and δ114/110Cd values in 2015, which include the plant pump effect, indicate more and isotopically heavier Cd in the upper (0-35 cm) than in the deeper (35-75 cm) soil-box. As for the four soil depths, we can plot the remaining Cd fractions and the ∆114/110Cdsoil-parent material values of the two boxes and fit the same Rayleigh fractionation model for soil formation (Figure 3b). In the next step, the soil-plant cycling model was applied to reckon back the effect of the plant pump. Without soil plant-cycling, less and isotopically lighter Cd was found in the upper (0-35 cm) than in the deeper (35-75 cm) soil-box, among all sites. This is exactly the Cd distribution which we would expect if parent material weathering was the dominating soil formation process and plants would not have cycled Cd. The remaining Cd fractions and ∆114/110Cdsoil-parent material values were again plotted. The best fit of the Rayleigh fractionation model resulted thereby in a soil formation factor ε of 0.16, like for the situation with “current values in 2015”.

Overall, the Cd distribution in our soils can be explained with two processes. (i) Parent material was physically and chemically weathered, heavy Cd isotopes were leached with seepage water and shifted the bulk soil isotope compositions towards lighter values. Simultaneously, (ii) plants cycled heavy Cd from the deeper to the upper soil horizons and inverted the Cd distribution and isotopic compositions in the soils.

Environmental implications

The Cd mass balances reveal a balanced system with net loss or net accumulation depending on crop type grown and fertilizer Cd concentrations. Currently, Cd does not further accumulate in soils if legal limits of Cd in fertilizers are enforced. The input fluxes to the soils have Cd isotope compositions that are identical within error which hindered Cd source tracing with end-member mixing models in the study. However,

20

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

523

524

525

526

527

528

529

530

531

532

533

534

535

536

585960

isotope mass balances have shown to be a promising tool to estimate the anthropogenic share of Cd contamination in agricultural soils. For the three systems, the long-term natural processes soil formation and soil-plant cycling which acted on ~13700 years dominated over the more recent anthropogenic impacts. These anthropogenic fluxes only became more important during the last century but annual fluxes with industry induced atmospheric deposition, fertilizer applications and crop harvests were still by 3-4 orders of magnitude smaller than the soil Cd pools.

Supporting information

Section 1: Detailed information on the materials and method. Figure S1: Map with location of the study sites. Figure S2: Sampling at the study sites. Figure S3: Cd abundance mass balances as a function of the Cd concentration in mineral P fertilizers. Figure S4: Relationship between Cd surplus in the topsoils and Cd concentration in parent materials. Figure S5: Rayleigh fractionation model for soil formation after removal of the soil-plant cycling effect with alternative ∆114/110Cdsoil-trees values. Figure S6: Soil-plant cycling model results for WI and NE. Table S1: Soil properties. Table S2: Calculated Cd abundance and stable isotope mass balances. Table S3: Standard reference materials. Table S4: Measured isotope compositions.

Acknowledgements

This study was funded by the Swiss Parliament via the National Research Program (NRP) 69 “Healthy Nutrition and Sustainable Food Production” (SNSF grant no. 406940_145195/1). We thank the farmers from the study sites for cooperation, Lorenz Schwab for the characterization of the soils, Barry Coles for the help in the MAGIC laboratories and the plant nutrition group for the support with the plant digestions. Many thanks to the members of the soil science and TrES group at the University of Bern for support in the laboratory and helpful discussions.

21

537

538

539

540

541

542

543

544

545

546

547

548

549

550

551

552

553

554

555

556

557

558

559

560

561

562

563

564

565

566

616263

References

(1) Nagajyoti, P. C.; Lee, K. D.; Sreekanth, T. V. M. Heavy metals, occurrence and toxicity for plants: a review. Environ. Chem. Lett. 2010, 8 (3), 199–216 DOI: 10.1007/s10311-010-0297-8.

(2) Godt, J.; Scheidig, F.; Grosse-Siestrup, C.; Esche, V.; Brandenburg, P.; Reich, A.; Groneberg, D. A. The toxicity of cadmium and resulting hazards for human health. J. Occup. Med. Toxicol. 2006, 1 (1), 22 DOI: 10.1186/1745-6673-1-22.

(3) Nordberg, G. F. Historical perspectives on cadmium toxicology. Toxicology and Applied Pharmacology. 2009, pp 192–200.

(4) Traina, S. J. The Environmental Chemistry of Cadmium. In Cadmium in Soils and Plants; Mclaughlin, M. J., Singh, B. R., Eds.; Springer Verlag: Dordrecht, 2013; pp 11–37.

(5) Scheffer, F.; Schachtschabel, P.; Blume, H.-P.; Brümmer, G. W.; Horn, R.; Kandeler, E.; Kögel-Knabner, I.; Kretzschmar, R.; Stahr, K.; Thiele-Bruhn, S.; et al. Lehrbuch der Bodenkunde, 16. Auflag.; Spektrum Akademischer Verlag: Heidelberg, 2010.

(6) Smolders, E.; Mertens J. Cadmium: In: Heavy Metals in Soils: Trace Metals and Metalloids in Soils and their Bioavailability. Springer 3rd ed.

(7) Shotyk, W.; Krachler, M.; Martinez-Cortizas, A.; Cheburkin, A. K.; Emons, H. A peat bog record of natural, pre-anthropogenic enrichments of trace elements in atmospheric aerosols since 12 370 14C yr BP, and their variation with Holocene climate change. Earth Planet. Sci. Lett. 2002, 199 (1–2), 21–37 DOI: 10.1016/S0012-821X(02)00553-8.

(8) Pacyna, J. M.; Pacyna, E. G.; Aas, W. Changes of emissions and atmospheric deposition of mercury, lead, and cadmium. Atmos. Environ. 2009, 43 (1), 117–127 DOI:

22

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587

588

589

590

591

592

593

594

595

646566

10.1016/j.atmosenv.2008.09.066.

(9) Rausch, N.; Nieminen, T.; Ukonmaanaho, L.; Le Roux, G.; Krachler, M.; Cheburkin, A. K.; Bonani, G.; Shotyk, W. Comparison of Atmospheric Deposition of Copper, Nickel, Cobalt, Zinc, and Cadmium Recorded by Finnish Peat Cores with Monitoring Data and Emission Records. Environ. Sci. Technol. 2005, 39 (16), 5989–5998 DOI: 10.1021/es050260m.

(10) U.S. Geological Survey. Mineral commodity summaries 2016. 202 p.

(11) Frossard E.; Bünemann E.; Jansa J.; Oberson A.; Feller C. Concepts and pratices of nutrient management in agro-ecosystems: Can we draw lessons from history to design future sustainable agricultural production systems? Die Bodenkultur 2009, No. 60(1), 43–60.

(12) Jones, K. C.; Symon, C. J.; Johnston, A. E. Retrospective analysis of an archived soil collection II. Cadmium. Sci. Total Environ. 1987, 67 (1), 75–89 DOI: 10.1016/0048-9697(87)90067-2.

(13) Six, L.; Smolders, E. Future trends in soil cadmium concentration under current cadmium fluxes to European agricultural soils. Sci. Total Environ. 2014, 485–486, 319–328 DOI: 10.1016/j.scitotenv.2014.03.109.

(14) Mar, S. S.; Okazaki, M. Investigation of Cd contents in several phosphate rocks used for the production of fertilizer. Microchem. J. 2012, 104, 17–21 DOI: 10.1016/j.microc.2012.03.020.

(15) Spiess, E. Nitrogen, phosphorus and potassium balances and cycles of Swiss agriculture from 1975 to 2008. Nutr. Cycl. Agroecosystems 2011, 91 (3), 351–365.

(16) Nicholson, F. A.; Jones, K. C.; Johnston, A. E. Effect of Phosphate Fertilizers and Atmospheric Deposition on Long-Term Changes in the Cadmium Content of Soils and Crops. Environ. Sci. Technol. 1994, 28 (12), 2170–2175 DOI: 10.1021/es00061a027.

23

596

597

598

599

600

601

602

603

604

605

606

607

608

609

610

611

612

613

614

615

616

617

618

619

620

621

622

623

624

676869

(17) Bigalke, M.; Ulrich, A.; Rehmus, A.; Keller, A. Accumulation of cadmium and uranium in arable soils in Switzerland. Environ. Pollut. 2017, 221, 85–93 DOI: 10.1016/j.envpol.2016.11.035.

(18) Wiersma, D.; van Goor, B. J.; van der Veen, N. G. Cadmium, lead, mercury and arsenic concentrations in crops and corresponding soils in the Netherlands. J. Agric. Food Chem. 1986, 34 (6), 1067–1074 DOI: 10.1021/jf00072a033.

(19) Keller, A.; Rossier, N.; Desaules, A. Schwermetallbilanzen von Landwirtschaftsparzellen der nationalen Bodenbeobachtung: NABO - Nationales Bodenbeachtungsnetz der Schweiz; Schriftenreihe der FAL; FAL: Zürich, 2005; Vol. 54.

(20) Nicholson, F. A.; Smith, S. R.; Alloway, B. J.; Carlton-Smith, C.; Chambers, B. J. An inventory of heavy metals inputs to agricultural soils in England and Wales. Sci. Total Environ. 2003, 311 (1–3), 205–219 DOI: 10.1016/S0048-9697(03)00139-6.

(21) Keller, A.; Schulin, R. Modelling heavy metal and phosphorus balances for farming systems. Nutr. Cycl. Agroecosystems 2003, 66 (3), 271–284 DOI: 10.1023/A:1024410126924.

(22) Keller, A.; Schulin, R. Modelling regional-scale mass balances of phosphorus, cadmium and zinc fluxes on arable and dairy farms. Eur. J. Agron. 2003, 20 (1–2), 181–198 DOI: 10.1016/S1161-0301(03)00075-3.

(23) Degryse, F.; Smolders, E.; Parker, D. R. Partitioning of metals (Cd, Co, Cu, Ni, Pb, Zn) in soils: Concepts, methodologies, prediction and applications - a review. Eur. J. Soil Sci. 2009, 60 (4), 590–612 DOI: 10.1111/j.1365-2389.2009.01142.x.

(24) Boekhold, A. E.; van der Zee, S. E. A. T. M. Significance of soil chemical heterogeneity for spatial behaviour of cadmium in field soils. Soil Sci. Soc. Am. J. 1992, No. 56, 747–754.

24

625

626

627

628

629

630

631

632

633

634

635

636

637

638

639

640

641

642

643

644

645

646

647

648

649

650

651

652

653

707172

(25) Jeng, A. S.; Singh, B. R. Cadmium status of soils and plants from a long-term fertility experiment in southeast Norway. Plant Soil 1995, 175 (1), 67–74 DOI: 10.1007/BF02413011.

(26) Sternbeck J.; Eriksson J.; Österaas A.H. The role of mineral fertilisers for cadmium in Swedish agricultural soils and crops. In Commission E; Vol. 2011.

(27) Moolenaar, S. W.; Lexmond, T. M. Heavy-metal balances of agro-ecosystems in the Netherlands. Netherlands J. Agric. Sci. 1998, 46 (2), 171–192.

(28) Wiederhold, J. G. Metal Stable Isotope Signatures as Tracers in Environmental Geochemistry. Environ. Sci. Technol. 2015, 49 (5), 2606–2624 DOI: 10.1021/es504683e.

(29) Wombacher, F.; Rehkämper, M.; Mezger, K.; Münker, C. Stable isotope compositions of cadmium in geological materials and meteorites determined by multiple-collector ICPMS. Geochim. Cosmochim. Acta 2003, 67 (23), 4639–4654 DOI: 10.1016/S0016-7037(03)00389-2.

(30) Shiel, A. E.; Weis, D.; Orians, K. J. Evaluation of zinc, cadmium and lead isotope fractionation during smelting and refining. Sci. Total Environ. 2010, 408 (11), 2357–2368 DOI: 10.1016/j.scitotenv.2010.02.016.

(31) Schmitt, A.-D.; Galer, S. J. G.; Abouchami, W. Mass-dependent cadmium isotopic variations in nature with emphasis on the marine environment. Earth Planet. Sci. Lett. 2009, 277 (1–2), 262–272 DOI: 10.1016/j.epsl.2008.10.025.

(32) Rehkämper, M.; Wombacher, F.; Horner, T. J.; Xue, Z. Natural and Anthropogenic Cd Isotope Variations. In Handbook of environmental isotope geochemistry; Baskaran, M., Ed.; Advances in isotope geochemistry; Springer: Heidelberg, 2011; pp 125–154.

25

654

655

656

657

658

659

660

661

662

663

664

665

666

667

668

669

670

671

672

673

674

675

676

677

678

679

680

681

682

737475

(33) Martinková, E.; Chrastný, V.; Francová, M.; Šípková, A.; Čurík, J.; Myška, O.; Mižič, L. Cadmium isotope fractionation of materials derived from various industrial processes. J. Hazard. Mater. 2016, 302, 114–119 DOI: 10.1016/j.jhazmat.2015.09.039.

(34) Cloquet, C.; Carignan, U.; Libourel, G.; Sterckeman, T.; Perdrix, E. Tracing Source Pollution in Soils Using Cadmium and Lead Isotopes. Environ. Sci. Technol. 2006, 40, 2525–2530 DOI: 10.1021/es052232+.

(35) Cloquet, C.; Carignan, J.; Libourel, G.; Sterckeman, T.; Perdrix, E. Tracing source pollution in soils using cadmium and lead isotopes. Einvironmental Sci. Technol. 2006, 40 (8), 2525–2530 DOI: 10.1021/es052232+.

(36) Wombacher, F.; Rehkämper, M.; Mezger, K. Determination of the mass-dependence of cadmium isotope fractionation during evaporation. Geochim. Cosmochim. Acta 2004, 68 (10), 2349–2357 DOI: 10.1016/j.gca.2003.12.013.

(37) Gao, B.; Zhou, H.; Liang, X.; Tu, X. Cd isotopes as a potential source tracer of metal pollution in river sediments. Environ. Pollut. 2013, 181, 340–343 DOI: 10.1016/j.envpol.2013.05.048.

(38) Wen, H.; Zhang, Y.; Cloquet, C.; Zhu, C.; Fan, H.; Luo, C. Tracing sources of pollution in soils from the Jinding Pb-Zn mining district in China using cadmium and lead isotopes. Appl. Geochemistry 2015, 52, 147–154 DOI: 10.1016/j.apgeochem.2014.11.025.

(39) Chrastný, V.; Čadková, E.; Vaněk, A.; Teper, L.; Cabala, J.; Komárek, M. Cadmium isotope fractionation within the soil profile complicates source identification in relation to Pb--Zn mining and smelting processes. Chem. Geol. 2015, 405, 1–9 DOI: 10.1016/j.chemgeo.2015.04.002.

(40) Salmanzadeh, M.; Hartland, A.; Stirling, C. H.; Balks, M. R.; Schipper, L. A.; Joshi, C.; George, E. Isotope Tracing of Long-Term Cadmium

26

683

684

685

686

687

688

689

690

691

692

693

694

695

696

697

698

699

700

701

702

703

704

705

706

707

708

709

710

711

712

767778

Fluxes in an Agricultural Soil. Environ. Sci. Technol. 2017, 51 (13), 7369–7377 DOI: 10.1021/acs.est.7b00858.

(41) Zhang, Y.; Wen, H.; Zhu, C.; Fan, H.; Luo, C.; Liu, J.; Cloquet, C. Cd isotope fractionation during simulated and natural weathering. Environ. Pollut. 2016, 216, 9–17 DOI: 10.1016/j.envpol.2016.04.060.

(42) Wasylenki, L. E.; Swihart, J. W.; Romaniello, S. J. Cadmium isotope fractionation during adsorption to Mn oxyhydroxide at low and high ionic strength. Geochim. Cosmochim. Acta 2014, 140, 212–226 DOI: 10.1016/j.gca.2014.05.007.

(43) Horner, T. J.; Rickaby, R. E. M.; Henderson, G. M. Isotopic fractionation of cadmium into calcite. Earth Planet. Sci. Lett. 2011, 312 (1–2), 243–253 DOI: 10.1016/j.epsl.2011.10.004.

(44) Lacan, F.; Francois, R.; Ji, Y.; Sherrell, R. M. Cadmium isotopic composition in the ocean. Geochim. Cosmochim. Acta 2006, 70 (20), 5104–5118 DOI: 10.1016/j.gca.2006.07.036.

(45) Ripperger, S.; Rehkämper, M.; Porcelli, D.; Halliday, A. N. Cadmium isotope fractionation in seawater --- A signature of biological activity. Earth Planet. Sci. Lett. 2007, 261 (3–4), 670–684 DOI: 10.1016/j.epsl.2007.07.034.

(46) Wei, R.; Guo, Q.; Wen, H.; Liu, C.; Yang, J.; Peters, M.; Hu, J.; Zhu, G.; Zhang, H.; Tian, L.; et al. Fractionation of Stable Cadmium Isotopes in the Cadmium Tolerant Ricinus communis and Hyperaccumulator Solanum nigrum. Sci. Rep. 2016, 6, 24309 DOI: 10.1038/srep24309.

(47) Wiggenhauser, M.; Bigalke, M.; Imseng, M.; Müller, M.; Keller, A.; Murphy, K.; Kreissig, K.; Rehkämper, M.; Wilcke, W.; Frossard, E. Cadmium isotope fractionation in soil-wheat systems. Environ. Sci. Technol. 2016, 50 (17), 9223–9231 DOI: 10.1021/acs.est.6b01568.

(48) FitzGerald R.; Roth N. Cadmium in mineral fertilizers - human and environmental risk update: SCAHT report for BLW.

27

713

714

715

716

717

718

719

720

721

722

723

724

725

726

727

728

729

730

731

732

733

734

735

736

737

738

739

740

741

798081

(49) Gubler, A.; Schwab, P.; Wächter, D.; Meuli, R. G.; Keller, A. Ergebnisse der Nationalen Bodenbeobachtung (NABO) 1985-2009; 2015.

(50) World reference base for soil resources 2014: International soil classification system for naming soils and creating legends for soil maps; World soil resources reports; FAO: Rome, 2014.

(51) Brimhall, G. H.; Chadwick, O. A.; Lewis, C. J.; Compston, W.; Williams, I. S.; Danti, K. J.; Dietrich, W. E.; Power, M. E.; Hendricks, D.; Bratt, J. Deformational mass transport and invasive processes in soil evolution. Science 1992, 255 (5045), 695–702 DOI: 10.1126/science.255.5045.695.

(52) Murphy, K.; Rehkamper, M.; Kreissig, K.; Coles, B.; van de Flierdt, T. Improvements in Cd stable isotope analysis achieved through use of liquid-liquid extraction to remove organic residues from Cd separates obtained by extraction chromatography. J. Anal. At. Spectrom. 2016, 31 (1), 319–327 DOI: 10.1039/c5ja00115c.

(53) Xue, Z.; Rehkamper, M.; Schonbachler, M.; Statham, P. J.; Coles, B. J. A new methodology for precise cadmium isotope analyses of seawater. Anal. Bioanal. Chem. 2012, 402 (2), 883–893 DOI: 10.1007/s00216-011-5487-0.

(54) Ripperger, S.; Rehkämper, M. Precise determination of cadmium isotope fractionation in seawater by double spike MC-ICPMS. Geochim. Cosmochim. Acta 2007, 71 (3), 631–642 DOI: 10.1016/j.gca.2006.10.005.

(55) Goix, S.; Point, D.; Oliva, P.; Polve, M.; Duprey, J. L.; Mazurek, H.; Guislain, L.; Huayta, C.; Barbieri, F. L.; Gardon, J. Influence of source distribution and geochemical composition of aerosols on children exposure in the large polymetallic mining region of the Bolivian Altiplano. Sci. Total Environ. 2011, 412–413, 170–184 DOI: 10.1016/j.scitotenv.2011.09.065.

(56) Wiseman, C. L. S.; Zereini, F.; Puttmann, W. Traffic-related trace

28

742

743

744

745

746

747

748

749

750

751

752

753

754

755

756

757

758

759

760

761

762

763

764

765

766

767

768

769

770

771

828384

element fate and uptake by plants cultivated in roadside soils in Toronto, Canada. Sci. Total Environ. 2013, 442, 86–95 DOI: 10.1016/j.scitotenv.2012.10.051.

(57) Jochum, K. P.; Nohl, U.; Herwig, K.; Lammel, E.; Stoll, B.; Hofmann, A. W. GeoReM: A New Geochemical Database for Reference Materials and Isotopic Standards. Geostand. Geoanalytical Res. 2005, 29 (3), 333–338 DOI: 10.1111/j.1751-908X.2005.tb00904.x.

(58) Mailänder, R.; Veit, H. Periglacial cover-beds on the Swiss Plateau: Indicators of soil, climate and landscape evolution during the Late Quaternary. CATENA 2001, 45 (4), 251–272 DOI: 10.1016/S0341-8162(01)00151-5.

(59) Burga, C. A.; Perret, R.; Vonarburg, C. Vegetation und Klima der Schweiz seit dem jüngeren Eiszeitalter: Vegetation and climate history in Switzerland during the later Pleistocene and Holocene; Ott: Thun, 1998.

(60) Ammann, B.; Lotter, A. F.; Eicher, U.; Gaillard, M.-J.; Wohlfarth, B.; Haeberli, W.; Lister, G.; Maisch, M.; Niessen, F.; Schlüchter, C. The W{ü}rmian Late-glacial in Iowland Switzerland. J. Quat. Sci. 1994, 9 (2), 119–125 DOI: 10.1002/jqs.3390090205.

(61) Lotter, A. F.; Eicher, U.; Siegenthaler, U.; Birks, H. J. B. Late-glacial climatic oscillations as recorded in Swiss lake sediments. J. Quat. Sci. 1992, 7 (3), 187–204 DOI: 10.1002/jqs.3390070302.

(62) Buss, H. L.; Sak, P. B.; Webb, S. M.; Brantley, S. L. Weathering of the Rio Blanco quartz diorite, Luquillo Mountains, Puerto Rico: Coupling oxidation, dissolution, and fracturing. Geochim. Cosmochim. Acta 2008, 72 (18), 4488–4507 DOI: 10.1016/j.gca.2008.06.020.

(63) White, A. F.; Brantley, S. L. The effect of time on the weathering of silicate minerals: Why do weathering rates differ in the laboratory and field? Chem. Geol. 2003, 202 (3–4), 479–506 DOI: 10.1016/j.chemgeo.2003.03.001.

29

772

773

774

775

776

777

778

779

780

781

782

783

784

785

786

787

788

789

790

791

792

793

794

795

796

797

798

799

800

801

858687

(64) Emmanuel, S.; Levenson, Y. Limestone weathering rates accelerated by micron-scale grain detachment. Geology 2014, 42 (9), 751–754 DOI: 10.1130/G35815.1.

(65) Belon, E.; Boisson, M.; Deportes, I. Z.; Eglin, T. K.; Feix, I.; Bispo, A. O.; Galsomies, L.; Leblond, S.; Guellier, C. R. An inventory of trace elements inputs to French agricultural soils. Sci. Total Environ. 2012, 439, 87–95 DOI: 10.1016/j.scitotenv.2012.09.011.

(66) Gisler A.; Schwab L. Marktkampagne Dünger 2011/2012: Kennzeichnung und Schwermetalle. Bern.

(67) Gray, C. W.; McLaren, R. G.; Roberts, A. H. C.; Condron, L. M. Cadmium phytoavailability in some New Zealand soils. Aust. J. Soil Res. 1999, 37 (3), 461 DOI: 10.1071/S98070.

(68) Gaw, S. K.; Kim, N. D.; Northcott, G. L.; Wilkins, A. L.; Robinson, G. Uptake of ΣDDT, arsenic, cadmium, copper, and lead by lettuce and radish grown in contaminated horticultural soils. J. Agric. Food Chem. 2008, 56 (15), 6584–6593 DOI: 10.1021/jf073327t.

(69) Adams, M. L.; Zhao, F. J.; McGrath, S. P.; Nicholson, F. A.; Chambers, B. J. Predicting cadmium concentrations in wheat and barley grain using soil properties. J. Environ. Qual. 2004, 33 (2), 532–541 DOI: 10.2134/jeq2004.0532.

(70) Scientific Comittee on Toxicity, E. and the E. (CSTEE). OPINION OF THE CSTEE ON “Member State assessments of the risk to health and the environment from cadmium in fertilizers”; Brussels, 2002.

(71) Bridgestock, L.; Rehkämper, M.; van de Flierdt, T.; Murphy, K.; Khondoker, R.; Baker, A. R.; Chance, R.; Strekopytov, S.; Humphreys-Williams, E.; Achterberg, E. P. The Cd isotope composition of atmospheric aerosols from the Tropical Atlantic Ocean. Geophys. Res. Lett. 2017, 44 (6), 2932–2940 DOI: 10.1002/2017GL072748.

(72) Pallavicini, N.; Engström, E.; Baxter, D. C.; Öhlander, B.; Ingri, J.;

30

802

803

804

805

806

807

808

809

810

811

812

813

814

815

816

817

818

819

820

821

822

823

824

825

826

827

828

829

830

888990

Rodushkin, I. Cadmium isotope ratio measurements in environmental matrices by MC-ICP-MS. J. Anal. At. Spectrom. 2014, 29 (9), 1570–1584 DOI: 10.1039/C4JA00125G.

(73) Martínez-Cortizas, A.; Pontevedra-Pombal, X.; Muñoz, J. C. N.; García-Rodeja, E. Four Thousand Years of Atmospheric Pb, Cd and Zn Deposition Recorded by the Ombrotrophic Peat Bog of Penido Vello (Northwestern Spain). Water. Air. Soil Pollut. 1997, 100 (3), 387–403 DOI: 10.1023/A:1018312223189.

(74) Coggins, A. M.; Jennings, S. G.; Ebinghaus, R. Accumulation rates of the heavy metals lead, mercury and cadmium in ombrotrophic peatlands in the west of Ireland. Atmos. Environ. 2006, 40 (2), 260–278 DOI: 10.1016/j.atmosenv.2005.09.049.

(75) Schell, W. R.; Tobin, M. J.; Novak, M. J. V; Wieder, R. K.; Mitchell, P. I. Deposition History of Trace Metals and Fallout Radionuclides in Wetland Ecosystems using 210Pb Chronology. Water. Air. Soil Pollut. 1997, 100 (3), 233–239 DOI: 10.1023/A:1018332727732.

(76) Wong, H. K. T.; Nriagu, J. O.; Coker, R. D. Atmospheric input of heavy metals chronicled in lake sediments of the Algonquin Provincial Park, Ontario, Canada. Chem. Geol. 1984, 44 (1–3), 187–201 DOI: 10.1016/0009-2541(84)90072-X.

(77) Norton, S. A.; Dillon, P. J.; Evans, R. D.; Mierle, G.; Kahl, J. S. The History of Atmospheric Deposition of Cd, Hg, and Pb in North America: Evidence from Lake and Peat Bog Sediments. In Acidic Precipitation: Sources, Deposition, and Canopy Interactions; Lindberg, S. E., Page, A. L., Norton, S. A., Eds.; Springer New York: New York, NY, 1990; pp 73–102.

(78) Historisches Lexikon der Schweiz HLS. Historisches Lexikon der Schweiz HLS; 2014.

(79) St. Arnaud, R. J.; Stewart, J. W. B.; Frossard, E. Application of the “Pedogenic Index” to soil fertility studies, Saskatchewan. Geoderma

31

831

832

833

834

835

836

837

838

839

840

841

842

843

844

845

846

847

848

849

850

851

852

853

854

855

856

857

858

859

860

919293

1988, 43 (1), 21–32 DOI: 10.1016/0016-7061(88)90052-3.

(80) Jobbagy, E. G.; Jackson, R. B. The distribution of soil nutriments with depth : Global patterns of the imprint of plants. Biogeochemistry 2001, 53, 51–77.

(81) Goldschmidt, V. M. The principles of distribution of chemical elements in minerals and rocks. The seventh Hugo M{ü}ller Lecture, delivered before the Chemical Society on March 17th, 1937. J. Chem. Soc. 1937, 0 (0), 655–673 DOI: 10.1039/JR9370000655.

(82) Reimann, C.; Englmaier, P.; Flem, B.; Gough, L.; Lamothe, P.; Nordgulen, Ø.; Smith, D. Geochemical gradients in soil O-horizon samples from southern Norway: Natural or anthropogenic? Appl. Geochemistry 2009, 24 (1), 62–76 DOI: 10.1016/j.apgeochem.2008.11.021.

(83) Reimann, C.; Arnoldussen, A.; Englmaier, P.; Filzmoser, P.; Finne, T. E.; Garrett, R. G.; Koller, F.; Nordgulen, Ø. Element concentrations and variations along a 120-km transect in southern Norway – Anthropogenic vs. geogenic vs. biogenic element sources and cycles. Appl. Geochemistry 2007, 22 (4), 851–871 DOI: 10.1016/j.apgeochem.2006.12.019.

32

861

862

863

864

865

866

867

868

869

870

871

872

873

874

875

876

877

878

879

949596

Figure 1: Cadmium abundance and stable isotope mass balances of the three arable soils at OE (a), WI (b) and NE (c) for one hydrological year (May 2014 – May 2015). Mass balances were calculated for wheat (I) and barley cultivation (II). System inputs are shown in red, system losses in green. Sizes of the boxes are proportional to the size of Cd fluxes (compared to the reference box for 1 g Cd ha -1 y-1). Sizes of the bulk soil boxes had to be reduced and would be 100x (OE), 50x (WI), and 500x (NE) bigger to proportionally represent real values. Net losses and net accumulations represent the mass balance values after one hydrological year for the two crops. Calculated δ114/110Cd values of inputs, outputs and bulk soil (0-50 cm) are shown next to the boxes. The bulk soil Cd isotope compositions after 100 hydrological years were calculated with current fluxes (current)* and with maximal inputs through atmospheric deposition, mineral P fertilizers and manure (max)** during the 100 model years.

33

880

881

882

883884885886887888889890891892

979899

Figure 2: Cd isotope compositions of the inputs, outputs and different depths of the bulk soils at the study sites OE ( ), WI () and NE (). Mineral fertilizers are not site specific (). Error bars represent 2 x standard deviations of sample replicates where n>1 and measurement replicates where n=1. Isotope values of wheat and barley harvest were calculated according to Equation 1 and error propagation according Equations S6-S9.

34

893

894895896897898

100101102

Figure 3: Relationships between the remaining Cd fraction in the soils (τCd values + 1) from the parent material (pm) and the apparent fractionation between the soils and parent materials (114/110Cdsoil-pm). a: values in 2015 for the 4 soil horizons and 3 sites; ε = 0.16. b: current values in 2015, averaged for the two boxes of the soil-plant cycling model (0-35 cm and 35-75cm); ε = 0.17. c: values in 2015 without soil-plant cycling; ε = 0.16.

35

899

900

901902903904905

103104105

Figure 4: Results of soil-plant cycling model at Oensingen. The remaining Cd fractions (τCd values + 1) and the isotope compositions of the two soil boxes (0-35 and 35-75 cm) in grey indicate values in 2015 and include soil-plant cycling over the whole soil formation period. Input parameters for the soil-plant cycling model (in green) were the cycling time (i.e. age of the soil), ∆114/110Cdsoil-plant and the cycled Cd. The remaining Cd fractions and isotope compositions of the two soil boxes (0-35 and 35-75 cm) in red indicate values in 2015 if soil-plant cycling would not have occurred.

36

906907908909910911912

106107108