Metal accumulation in riverine macroinvertebrates from a platinum mining region

Journal Pre-proofs Metal accumulation in riverine macroinvertebrates from a platinum mining region J.H. Erasmus, W. Malherbe, S. Zimmermann, A.W. Lorenz, M. Nachev, V. Wepener, B. Sures, N.J. Smit PII: DOI: Reference:

S0048-9697(19)34729-1 https://doi.org/10.1016/j.scitotenv.2019.134738 STOTEN 134738

To appear in:

Science of the Total Environment

Received Date: Revised Date: Accepted Date:

15 July 2019 27 September 2019 28 September 2019

Please cite this article as: J.H. Erasmus, W. Malherbe, S. Zimmermann, A.W. Lorenz, M. Nachev, V. Wepener, B. Sures, N.J. Smit, Metal accumulation in riverine macroinvertebrates from a platinum mining region, Science of the Total Environment (2019), doi: https://doi.org/10.1016/j.scitotenv.2019.134738

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

© 2019 Published by Elsevier B.V.

1

Metal accumulation in riverine macroinvertebrates from a platinum

2

mining region

3

Erasmus, J.H.1, Malherbe, W.1, Zimmermann, S.1,2, Lorenz, A.W.2, Nachev, M.2, Wepener,

4

V.1, Sures, B.2 and Smit, N.J.1

5

1

6

University, 11 Hoffman St, Potchefstroom, 2520, South Africa

Water Research Group, Unit for Environmental Sciences and Management, North-West

7 8

2

9

University of Duisburg-Essen, Universitätsstr. 2, 45141 Essen, Germany

Department of Aquatic Ecology and Centre for Water and Environmental Research,

10 11

Corresponding author: J.H. Erasmus – email: [email protected]; ORCID 0000-0001-

12

9056-5424

13

W. Malherbe – email: [email protected]; ORCID 0000-0003-1852-5641

14

S. Zimmermann – email: [email protected]

15

A.W. Lorenz – email: [email protected]; ORCID 0000-0002-3262-6396

16

M. Nachev – email: [email protected]

17

V. Wepener – email: [email protected]; ORCID 0000-0002-9374-7191

18

B. Sures – email: [email protected]

19

N.J. Smit – email: [email protected]; ORCID 0000-0001-7950-193X

20

Metal accumulation in riverine macroinvertebrates from a platinum

21

mining region

1

22

Erasmus, J.H.1, Malherbe, W.1, Zimmermann, S.1,2, Lorenz, A.W.2, Nachev, M.2,

23

Wepener, V.1, Sures, B.2 and Smit, N.J.1

24

1

25

West University, 11 Hoffman St, Potchefstroom, 2520, South Africa

Water Research Group, Unit for Environmental Sciences and Management, North-

26 27

2

28

University of Duisburg-Essen, Universitätsstr. 2, 45141 Essen, Germany

Department of Aquatic Ecology and Centre for Water and Environmental Research,

29 30

Corresponding author: J.H. Erasmus – email: [email protected]

31

W. Malherbe – email: [email protected]

32

S. Zimmermann – email: [email protected]

33

A.W. Lorenz – email: [email protected]

34

M. Nachev – email: [email protected]

35

V. Wepener – email: [email protected]

36

B. Sures – email: [email protected]

37

N.J. Smit – email: [email protected]

2

38

Abstract

39

South Africa is the world’s main supplier of Pt. The Bushveld Igneous Complex in

40

South Africa contains 75% of the world’s Pt resources. Mining of this precious metal

41

requires large volumes of water for production and removal of waste products. Most

42

of this wastewater is discharged into river systems. Although the source of

43

contamination with Pt in aquatic systems due to mining activities is known, little to no

44

information is available about the impact of Pt on aquatic organisms. Additionally,

45

other metals are released as byproducts of Pt mining, which might also be

46

discharged into the environment. Therefore, concentrations of Cr, Ni, Cu, Zn, Cd, Pt

47

and Pb were determined in water, sediment and macroinvertebrate samples from a

48

reference site (Site 1), a highly impacted site (Site 2) and a moderately impacted site

49

(Site 3) along the Hex River, South Africa. Aquatic invertebrate families representing

50

different functional feeding groups i.e. scraper-grazers (Lymnaeidae), collector-

51

gatherers (Potamonautidae, Hydropsychidae, Tubificidae and Chironomidae),

52

shredders (Baetidae) and predators (Coenagrionidae and Libellulidae) were studied.

53

In the sediments, the concentrations of Cr and Pt were significantly higher at Site 2

54

than at Sites 1 and 3, respectively, whereas concentrations of Ni, Cu, Cd, and Pb

55

showed no significant differences between the sites. Depending on the metal, the

56

aquatic invertebrate families showed different grades of bioaccumulation. The results

57

from especially Lymnaeidae, Baetidae, Tubificidae and Chironomidae showed great

58

promise for the use of these taxa for biomonitoring of metal contaminations. The

59

macroinvertebrates accumulated metals associated with Pt mining, with epi-benthic

60

dwelling taxa (Tubificidae) accumulating higher concentrations of Pt and Cr than

61

other families (e.g. Potamonautidae, Coenagrionidae and Lymnaeidae). These

62

results provide valuable information on the behavior of metals related to Pt mining in 3

63

aquatic ecosystems and therefore can contribute to the risk assessment of these

64

intensive mining activities.

65

Keywords:

66

Platinum, chromium, bioaccumulation, benthic invertebrates, South Africa, mining

67

activities

4

68

1. Introduction

69

Platinum group elements (PGE) are precious metals that occur in sufficient

70

quantities for mining, mainly in South Africa, Russia and the United States (Pawlak

71

et al., 2014). Although PGE are mostly used in automotive catalytic converters

72

(Pawlak et al., 2014), from where they are emitted, mining industries and chemical

73

facilities are also one of the primary sources of PGE into the environment (Zereini

74

and Wiseman, 2015). South Africa is the world’s main producer of PGE with 73, 39

75

and 82 % of Pt, Pd and Rh of global production in 2018, respectively (Johnson-

76

Matthey, 2018). These PGE resources are located within the Bushveld Igneous

77

Complex (Cawthorn, 2010). The Bushveld Igneous Complex is subjected to intensive

78

mining activities (PGE and Cr), which requires large volumes of water for production

79

activities, as well as for the removal of waste products (Haggard et al., 2015). Most

80

of this production water is discharged into surface waters. However, the effect of

81

these waste discharges to aquatic ecosystems mainly remains unclear, but can

82

cause disturbances in the surface PGE geochemical cycle (Almécija et al., 2017).

83

Considerable research has been completed on PGE accumulation from car catalysts

84

in road and airborne dust, as well as in soils in the vicinity of roads of different cities

85

worldwide (summarized in Okorie et al., 2015; Zereini and Wiseman, 2015).

86

Furthermore, elevated concentrations of PGE in organisms living in and on

87

contaminated soils were found in South Africa (Maboeta et al., 2006). These

88

catalyst-derived PGE particles are introduced via surface runoff into sediments of

89

rivers and other water bodies (Ruchter et al., 2015). However, according to Rauch

90

and Fatoki (2015), only a few studies have been conducted on PGE discharges from

91

Pt mining and production activities, while data on the occurrence of PGE in biota

92

near PGE production sites is lacking. 5

93

The Hex River in South Africa drains an intensive PGE mining area within the BIC

94

(Rustenburg area), with some of the world’s most productive Pt mines in the vicinity

95

(Johnson-Matthey, 2018). This river is an ideal case study due to the headwaters

96

being in an unimpacted state, while downstream the impacts of mining, industry,

97

urban and informal settlements can be observed. Almécija et al. (2017) found

98

elevated concentrations of Cr, Ni, Pt and Cu within the Hex River’s sediment, while

99

Maboeta et al. (2006) reported that concentrations of Cu and Ni are associated with

100

Pt mining. Rauch and Fatoki (2013) also found an association of the concentrations

101

of Cr and Cu with Pt mining, especially in the Bushveld Igneous Complex, and

102

reported high levels of Pt in soil samples, which were affected by Pt mining.

103

Macroinvertebrates associated with the sediments in aquatic systems are considered

104

good sentinels for pollution. Due to their low mobility, they can provide valuable

105

information about the conditions in the localities where they are found (Kenney et al.,

106

2009). Macroinvertebrates comprise of a variety of different taxa, such as insects,

107

crustaceans, bivalves, oligochaetes, as well as gastropods (Merritt et al., 2008);

108

each of these taxa exhibit different feeding strategies and occupy different trophic

109

levels (Moore, 2006). According to Kenney et al. (2009), these organisms are

110

responsible for several important ecological functions (i.e. nutrient cycling,

111

decomposition), but also represent an important link in aquatic food webs as both

112

consumers and prey. Consequently, these organisms can provide valuable

113

information on the bioavailability and transfer of metals within food webs.

114

The aim of this study was to investigate the behavior of metals associated with Pt

115

mining in a river system by determining the metal concentrations in water, sediment

116

and aquatic macroinvertebrates at different impacted sampling sites along the Hex

6

117

River, South Africa. In terms of biomonitoring aspects, the bioaccumulation of the

118

metals was evaluated for various macroinvertebrate families.

119 120

2. Materials and Methods

121

2.1. Field sampling

122

Water, sediment and aquatic macroinvertebrate samples were collected at five sites

123

during March 2018 in the Rustenburg area, North West Province, South Africa (Fig.

124

1). These included three sites within the Hex River, one site in a settling pond (Mine)

125

located on a Pt mine and one site in a tributary downstream of the City of

126

Rustenburg (Urban). The sites in the Hex River included a reference site (Site 1), a

127

site located directly downstream of the Pt mining activities (Site 2), and a site located

128

further (approximately 6 km) downstream of the mining area, which is additionally

129

influenced by urban effluent and a wastewater treatment works (Site 3).

130

At each sampling site pH, electrical conductivity (EC) and temperature (all by ExStik

131

II EC500, Extech Instruments), as well as dissolved oxygen (DO) and oxygen

132

saturation (both by ExStik II DO600, Extech Instruments) were measured in situ.

133

Water samples for metal analysis were collected in triplicate in pre-cleaned

134

polyethylene containers. Sediment samples from the upper 10 cm were also

135

collected in triplicate in pre-cleaned polyethylene containers. Macroinvertebrates

136

were sampled with a sweep net (1 mm mesh size) and placed into a sorting tray,

137

where they were rinsed with MilliQ® water and sorted according to families in pre-

138

cleaned polyethylene containers and frozen at -20 °C until further analyses. Families

139

collected at all three Hex River sites were the Lymnaeidae (pond snails), Baetidae

140

(mayflies), Potamonautidae (river crab), Coenagrionidae (sprites, damselflies) and 7

141

Libellulidae (skimmers, dragonflies), while Hydropsychidae (caddisflies) and

142

Tubificidae (aquatic earthworms) were only found at Site 2 and Chironomidae

143

(common midges) only at Site 3. A single recently emerged dragonfly (Libellulidae)

144

was collected at the mine settling pond, while no macroinvertebrates were collected

145

in the urban effluent. These taxa represented different functional feeding groups i.e.

146

scraper-grazers

147

Hydropsychidae,

148

predators (Coenagrionidae and Libellulidae).

149

Figure 1: Map of the study area, indicating the sampling sites in the Hex River (Site

150

1 – 3), in the mine settling pond (Mine) and in the urban effluent (Urban), as well as

151

mining activities, and urban and informal settlements in the surrounding area.

(Lymnaeidae), Tubificidae

and

collector-gatherers Chironomidae),

shredders

(Potamonautidae, (Baetidae)

and

152 153

2.2. Sample preparation

154

Water samples were filtered through a cellulose nitrate filter (0.45 µm, Sartorius

155

Stedim Biotech) to determine the suspended solids (mg/100mL) and acidified with

156

HNO3 (65%, supra pure quality, Merck) to an acid concentration of 1 ‰.

157

Sediment samples were first freeze-dried (FreeZone® 6, Labconco), after which

158

organic content and percentage particle size distribution was determined, according

159

to methods described in Wepener and Vermeulen (2005). The following particle size

160

categories were applied: gravel (>4000 μm), very coarse sand (4000-2000 μm),

161

coarse sand (2000-500 μm), medium sand (500-212 μm), fine sand (212-65 μm) and

162

mud (<65 μm). Subsequently, approximately 0.2 g (dry weight) of the < 65 µm grain

163

size sediment were digested in a mixture of 3 mL HCl (37%, supra pure quality,

164

Merck) and 0.5 mL HNO3 (65%, supra pure quality, Merck) adapted from Djingova et 8

165

al. (2003) using a Mars 5 microwave digestion system (CEM Corporation, Kamp-

166

Lintfort). Following digestion, samples were brought to a volume of 5 mL with MilliQ®

167

water.

168

After thawing, each family of macroinvertebrates from each site was homogenized

169

with a dispersing tool (Ultra-Turrax T25, Janke & Kunkel) and freeze-dried. For

170

mollusks

171

respectively, were removed before homogenization, so that only the soft tissue were

172

used for metal analysis. Samples of 0.1 g (dry weight) were weighed into 20 mL

173

TFM® vessels (MarsXpress, CEM) and digested in a mixture of 3 mL HNO3 (65%,

174

supra pure quality, Merck) and 1 mL HCl (37%, supra pure quality, Merck) adapted

175

from Djingova et al. (2003) using a Mars 6 microwave digestion system (CEM

176

Corporation, Kamp-Lintfort). Taxa with enough sample mass were analyzed in

177

replicates (see supplementary data, Table 2). Following digestion, the solution was

178

transferred to a 5 mL volumetric flask and brought to volume with MilliQ® water. All

179

samples were stored at room temperature until metal detection.

180

2.3. Metal detection

181

In all samples, the concentrations of Ni, Cu, Zn, Cd, Pt and Pb were determined

182

using

183

(PerkinElmer, Elan 6000) equipped with an auto-sampler system (PerkinElmer, AS-

184

90). For the quantification, the following mass lines were evaluated:

185

111Cd, 195Pt, 206Pb.

186

Prior

187

macroinvertebrates were diluted 1:5 and the water samples 1:2 with 1% HNO3

(Lymnaeidae)

quadrupole

to

and

inductively

measurements,

the

crabs

(Potamonautidae),

coupled

digested

plasma-mass

solutions

shell

and

spectroscopy

of

the

carapace,

(ICP-MS)

60Ni, 65Cu, 66Zn,

sediments

and

9

188

(supra pure quality, Merck). Yttrium and Tm (both Certipur®, Merck) were added as

189

internal standard at a concentration of 10 µg/L.

190

The ICP-MS operated at 1000 W plasma power, 14 L/min plasma gas flow, 0.95

191

L/min nebulizer gas flow and a sample flow-rate of 1 mL/min regulated by a

192

peristaltic pump. Calibration of the ICP-MS was performed using a series of 11

193

dilutions

194

concentrations of the analytes were calculated in the samples using corresponding

195

regression lines with a correlation factor of ≥ 0.999. As concentrations of Pt

196

determined by ICP-MS can be affected by interferences, correction for HfO+

197

interference was performed according to Djingova et al. (2003). Chromium

198

concentrations were analyzed by means of atomic absorption spectrometry (AAS)

199

using a PerkinElmer AAnalyst 600 equipped with Zeeman-effect background

200

correction, following the method of Nawrocka and Szkoda (2012). Detection limits for

201

the macroinvertebrate samples (related to a mean sample weight of 136 mg) were

202

determined as three times the standard deviation of the blank measurements and

203

were found to be 0.21 µg/g for Cr, 0.02 µg/g for Ni and Cu, 0.38 µg/g for Zn, 0.00063

204

µg/g for Cd, 0.0008 µg/g for Pt and 0.021 µg/g for Pb.

205

Quality control of the element analysis was performed using applicable standard

206

reference materials for all of the metals (see Table 1). Reference materials, IAEA-

207

407 (fish homogenate reference material; International Atomic Energy Agency,

208

Monaco), DORM-3 (fish protein certified reference material, National Research

209

Council, Canada) and DOLT-3 (dogfish liver certified reference material, National

210

Research Council, Canada) were used to validate the analytical procedure of biota

211

samples. Quality assurance for the sediment analysis was ensured by analyzing

212

BCR723 reference material (road dust standard reference material, Institute for

of

a

multi-element

standard

solution.

With

this

calibration,

the

10

213

Reference Materials and Measurements, European Commission), which provided

214

certified values for Pt and indicative values for various trace metals. All elements

215

considered in the present study, showed recovery rates within 20% of the certified

216

range (Table 1). As Pt was not certified in the biological reference materials, a

217

spiking experiment was performed by adding Pt to the sample matrix before and

218

after the digestion in a concentration range of 0.1 to 100 µg/L. The recovery rates

219

were evaluated from the ratios of the slopes calculated for the Pt addition series

220

before and after digestion. All recovery rates were within a 10% deviation range. In

221

order to assess the potential mass interferences during the detection of Pt, after the

222

measurement series, a matrix-adapted addition with Hf was performed as described

223

by Ek et al. (2004). However, obtained interferences rates were always less than 2%

224

and therefore no Hf-correction was applied.

225 226

Table 1: Recovery rates (%) of the elements of interest obtained for the different

227

certified reference materials used in the analysis.

228

2.4. Data analysis

229

Data were tested for normality and homogeneity of variance using D’Agostino and

230

Pearson omnibus normality test and Shapiro-Wilk normality test, respectively. One-

231

way analysis of variance with Tukey’s multiple comparison test was used to test for

232

significant differences in metal concentrations between sites. Level of significance

233

was set at p < 0.05. A Principle Component Analysis (PCA) (Canoco v5.12) was

234

constructed to assess the spatial patterns associated with the water and sediment

235

quality, as well as a PCA with supplementary variables to assess the associations

236

between bioaccumulation in macroinvertebrates and abiotic metal concentrations. 11

237

Data for the PCA was log transformed y = log (x + 1). Bioconcentration factor (BCF =

238

cinvertebrate/ cwater) and biota–sediment accumulation factor (BSAF = cinvertebrate/ csediment)

239

were calculated as the ratio of the average metal concentration in the invertebrates

240

per quantified metal concentration of the filtered water and sediment, respectively.

241 242

3. Results

243

3.1. Water and sediment quality

244

The water and sediment quality results collected from the three sites in the Hex

245

River, as well as the mine settling pond and urban effluent are presented in Table 2

246

and supplementary data, Table 1. The data from the three river sites, mine settling

247

pond and urban effluent were subjected to a PCA biplot and Fig. 2 presents an

248

integrated spatial ordination of the abiotic variables. Distinct spatial differences could

249

be observed, while Site 3, mine settling pond and urban effluent displayed more

250

similar water quality characteristics than Sites 1 and 2. Sites 1 and 2 were

251

characterized by higher concentrations of Cu, Zn and Pb in the water, with higher

252

dissolved oxygen and organic content, as well as finer substrate composition. Site 3,

253

mine settling pond and urban effluent were characterized by higher concentrations of

254

Cr, Ni and Pt in the water, as well as higher turbidity and electrical conductivity. From

255

the metal concentrations in the sediment, Site 2 associated with higher

256

concentrations of Cr, Ni, Cu, Cd and Pb, while the mine settling pond and urban

257

effluent associated with high concentrations of Zn and Pt.

258

12

259

Table 2: Water and sediment quality variables analyzed at the three Hex River sites

260

(Site 1 – 3), as well as the mine settling pond (Mine) and urban effluent (Urban). The

261

codes for each variable are used in the PCA biplot.

262

Figure 2: PCA biplot of the water (A) and sediment (B) quality variables measured at

263

sites in the Hex River (Sites 1 – 3), a mine settling pond (Mine) and an urban effluent

264

(Urban) during March 2018. The water (A) biplot describes 93.2 % of the variation

265

with 81.2 % on the first axis and 12.0 % on the second axis. The sediment (B) biplot

266

describes 84.1 % of the variation with 58.0 % on the first axis and 26.1 % on the

267

second axis.

268

3.2. Metal concentrations in the water

269

Among the four sampling sites, the water of the reference site (Site 1) had the

270

highest concentrations of Cu (21.2 µg/L), Zn (20.6 µg/L) and Pb (2.12 µg/L) (Fig. 3).

271

Going further downstream of the Hex River, the concentrations of these metals tend

272

to decrease. However, statistical analysis revealed no significant differences

273

between the different sampling sites for these elements. The water of the mine

274

settling pond (Mine) contained the highest concentrations of Cd (0.16 µg/L) and Pt

275

(0.47 µg/L) (Fig. 3). The urban effluent (Urban) had the highest concentrations of Ni

276

(115.3 µg/L), and significantly increased in the Hex River from Site 2 to Site 3. As

277

expected, the water of the mine settling pond showed significantly higher

278

concentrations of Pt than all of the other sampling sites. The highly impacted site

279

(Site 2) had significantly higher concentrations of Pt than Sites 1 and 3, while Site 3

280

differed significantly from Site 1. The concentrations of Cr in the water of all sampling

281

sites were below the detection limit.

13

282

Figure 3: Mean Ni (A), Cu (B), Zn (C), Cd (D), Pt (E) and Pb (F) concentrations

283

(µg/L) with standard error of the mean in water collected from selected sites in the

284

Hex River (Site 1 – 3), a mine settling pond (Mine) and an urban effluent (Urban).

285

Chromium concentrations were below the detection limit of 0.19 μg/L. Significant

286

differences (p < 0.05) between sites are indicated in table format by means of

287

asterisk.

288

3.3. Metal concentrations in sediment

289

The sediments of the highly impacted site (Site 2) contained the highest

290

concentrations of Ni (256 µg/g), Cu (119 µg/g), Cd (0.39 µg/g) and Pb (41 µg/g).

291

However, statistical analysis revealed no significant differences for these metals

292

between the sites. In contrast, the mine settling pond (Mine) demonstrated the

293

highest concentrations of Cr (1 609 µg/g), Zn (391 µg/g) and Pt (0.61 µg/g) (Fig. 4).

294

In the sediments of the Hex River, the concentrations of both, Cr and Pt, were

295

significantly higher at Site 2 than at Sites 1 and 3, respectively. The urban effluent

296

site (Urban) contribute to high concentrations of Ni, Zn and Pb.

297

Figure 4: Mean Cr (A), Ni (B), Cu (C), Zn (D), Cd (E), Pt (F) and Pb (G)

298

concentrations (µg/g DW) with standard error of the mean in sediments (DW: dry

299

weight) collected from selected sites in the Hex River (Site 1 – 3), a mine settling

300

pond (Mine) and an urban effluent (Urban). Significant differences (p < 0.05)

301

between sites are indicated in table format by means of asterisk.

302

3.4. Metal concentrations in macroinvertebrate families

303

A PCA with supplementary variables to assess the associations between

304

bioaccumulation in macroinvertebrates and abiotic metal concentrations were

305

constructed, where the metal concentrations in the macroinvertebrates varied 14

306

depending on the family and the sampling site (Fig. 5). Most of the

307

macroinvertebrate families accumulated the highest metal concentrations at Site 2,

308

where the concentrations of Cr, Cd and Pt in the water and all of the metals

309

concentrations in the sediment were the highest. The family Coenagrionidae

310

accumulated higher concentrations of Ni, Cr and Pb at Site 1 than the other sites,

311

while Libellulidae accumulated higher concentrations of Pb at Site 1 than the other

312

sites.

313

concentrations of Zn, while Coenagrionidae and Libellulidae accumulated higher

314

concentrations of Pt and Ni, respectively compared to the other sites.

315

Figure 5: A PCA biplot with supplementary variables of the metal concentrations in

316

water, sediment and macroinvertebrate families, respectively, collected from the

317

three sites in the Hex River (Site 1 – 3) during March 2018. The biplot describes 100

318

% of the variation with 59.2 % on the first axis and 40.8 % on the second axis.

At

Site

3,

Lymnaeidae

and

Potamonautidae

accumulated

higher

319 320

The highest concentrations of Cr were found in the families Tubificidae (92 µg/g DW)

321

from Site 2 and the Libellulidae sample (69 µg/g DW) collected from the mine settling

322

pond (Fig. 6A). Chromium concentrations were higher for most taxa at Site 2, only

323

the Coenagrionidae showed higher concentrations at Site 1. The highest

324

concentrations of Ni were determined in the Lymnaeidae (59 µg/g DW) of Site 3 and

325

in the Libellulidae sample of the mine settling pond (59 µg/g DW). Nickel

326

concentrations were higher in most of the families at Site 3, except for Baetidae at

327

Site 2 and Coenagrionidae at Site 1. Both Lymnaeidae and Potamonautidae showed

328

the highest concentrations of Cu regardless of the location (Fig. 6C). Copper

329

concentrations ranged from 16 µg/g DW in Libellulidae of Site 1 and 88 µg/g DW in

15

330

Potamonautidae of Site 2. In contrast, the Lymnaeidae contained the lowest

331

concentrations of Zn within the macroinvertebrate families with the lowest mean

332

value of 22 µg/g DW of Site 2. Highest concentrations of Zn were found in the

333

Libellulidae sample collected at the mine settling pond (180 µg/g DW) (Fig. 6D).

334

However, at all sampling sites of the Hex River the highest concentrations of Zn and

335

Cd were obtained for the family Baetidae. In all families, the mean concentrations of

336

Cd were higher at Site 2 (the site directly downstream of Pt mining activities) in

337

comparison to the other sites. The same trend was revealed for Pt with the exception

338

of the Libellulidae sample collected at the mine settling pond. For both Pt and Pb

339

(Fig. 6F-G), the concentrations were the highest in Tubificidae, whereas Site 2

340

showed the highest metal concentrations as compared with the other sites, except

341

for Coenagrionidae and Libellulidae.

342

Figure 6: Mean Cr (A), Ni (B), Cu (C), Zn (D), Cd (E), Pt (F) and Pb (G)

343

concentrations (µg/g DW) with standard error of the mean in macroinvertebrate

344

families (DW: dry weight) collected from selected sites in the Hex River (Site 1 – 3)

345

and a mine settling pond (LOD: limit of detection).

346 347

3.5. Bioaccumulation in macroinvertebrate families

348

The BCF ranged from 23 (Pt in Potamonautidae) to 27 143 (Zn in Baetidae),

349

whereas the BSAF ranged from 0.012 (Cr in Potamonautidae) to 11.17 (Zn in

350

Baetidae) (Tables 3 and 4). The mean BCF between metals and invertebrate

351

families decreased in the order Zn > Cu > Ni > Cd > Pb > Pt, while the order of mean

352

BSAF was Zn > Cu > Cd > Ni > Pt > Pb > Cr. When considering all the invertebrate

353

families collected within the Hex River, the families with the highest BCF and BSAF 16

354

were Lymnaeidae for Ni and Cu, and Baetidae for Zn and Cd, while Chironomidae

355

had the highest for Cr and Pb, and Tubificidae for Cr, Pt and Pb. In contrast, when

356

considering only the families collected at all the Hex River sites, Lymnaeidae had the

357

highest BCF and BSAF for Ni, Cu and Pt, while Baetidae had the highest for Cr, Zn,

358

Cd and Pb. When comparing the sites, the highest BCF continuously occurred in the

359

invertebrates at Site 1 for Ni, Site 2 for Cd and Pt, and Site 3 for Cu, Zn and Pb,

360

while the highest BSAF continuously occurred at Site 1 for Cr, Zn, Cd, Pt and Pb,

361

and Site 3 for Ni and Cu. The BCF differed between sites. The BCF decreased in the

362

order Ni > Zn > Cu > Cd > Pb > Pt (Site 1), Zn > Cu > Cd > Ni > Pb > Pt (Site 2) and

363

Zn > Cu > Pb > Ni > Cd > Pt (Site 3). On the other hand, the BSAF for Sites 1 and 2

364

showed the same order (Zn > Cu > Cd > Ni > Pt > Pb > Cr), while at Site 3 the BSAF

365

decreased in the order Zn > Cu > Ni > Cd > Pt > Pb > Cr.

366

Table 3: Bioconcentration factors (BCF) calculated as the ratio of the metal

367

concentrations in macroinvertebrate families and metal concentrations in water

368

collected from selected sites in the Hex River (Site 1 – 3) and a mine settling pond

369

(Mine).

370

Table 4: Bioaccumulation factors (BSAF) calculated as the ratio of the metal

371

concentrations in macroinvertebrate families and metal concentrations in sediment

372

collected from selected sites in the Hex River (Site 1 – 3) and a mine settling pond

373

(Mine).

374

4. Discussion

375

4.1. Key metals for mining, byproducts and waste

376

Platinum mining in South Africa provide efficient supplies to the world’s demand for

377

Pt (Johnson-Matthey, 2018), but are a major source of metal contamination within 17

378

aquatic ecosystems associated with these mining regions (Rauch and Fatoki, 2013).

379

The key metals for mining in the Rustenburg area consist primarily of platinum group

380

elements (especially Pt) and Cr, while Ni and Cu are recovered as byproducts from

381

the mined ores. In the ores of the Bushveld Igneous Complex the PGE occur

382

naturally in high concentrations, while the concentrations of Cu, Cr and Ni were

383

found to be 12-, 22- and 22- fold higher than in the normal upper earth crust,

384

respectively (Barnes and Maier, 2002). Although these metals are naturally present

385

in the environment, the intensive mining activities (Site 2 and mine settling pond)

386

within this region can even increase the concentrations within the aquatic

387

ecosystem. Accordingly, in the sediment samples of the present study, the

388

concentrations of both, Cr and Pt, were significantly higher for the settling pond, as

389

well as for the highly impacted site (Site 2) than for the reference site (Site 1). The

390

concentrations of Ni, Cu, Cd and Pb showed no significant differences between the

391

sites. This clearly confirms the entry of mining associated Pt and Cr into the river

392

ecosystem.

393

When comparing the data of the present study with results of previous studies in the

394

Hex River near Rustenburg, the concentrations of Pt in the water of the highly

395

impacted site (Site 2) and the mine settling pond were four fold and 28 fold higher

396

than the concentrations reported by Somerset et al. (2015) (see supplementary data,

397

Table 3). The concentrations of Pt in the sediment at Site 2 were four fold higher

398

than the concentrations recorded by Almécija et al. (2017). Rauch and Fatoki (2013)

399

reported concentrations of Pt of up to 0.653 µg/g in topsoil collected near a Pt

400

smelter within the Hex River catchment, indicating that Pt can also be transported via

401

air.

18

402

Downstream from the reference in the Hex River, concentrations of Cd and Pt in the

403

water increased at Site 2 and decreased at Site 3. These metal concentrations can

404

be attributed to the intensive mining activities, but also to industrial and urban

405

activities from the city of Rustenburg. The decrease in these metal concentrations

406

can be due to a dilution effect as the wastewater treatment plants are located

407

between Sites 2 and 3, and add a large volume of water (31.5 ML/day; DWA, 2011)

408

to the river system. Furthermore, the increase of the concentrations of Ni from Site 2

409

to Site 3 can be attributed to wastewater treatment effluents as domestic wastewater

410

effluents are considered a major source of Ni together with industrial effluent

411

(Cempel and Nikel, 2006). On the other hand, concentrations of Cu, Zn and Pb in the

412

water were the highest at Site 1 and decreased further downstream, which indicates

413

that these metals are geological in origin.

414

The concentrations of Cr, Ni, Zn, Cd and Pt in the sediments were higher at Site 2,

415

than Sites 1 and 3. The concentrations of Cu and Pb followed the same trend, but

416

decreased at Site 3 to almost the same concentrations as at Site 1. This increase in

417

all metal concentrations at Site 2 indicates to a point source of pollution (i.e. mainly

418

mining activities) whereas the pollution source is remediated downstream at Site 3.

419

This decrease in metal concentrations downstream from Site 2 can possibly be

420

ascribed to the metals that precipitate or adsorb to organic matter (Blais et al., 2008)

421

and decrease steadily downstream of the pollution source.

422

Sources of these metals into the environment include natural weathering of the

423

geology, mining activities, industrial and domestic wastewater effluent, as well as

424

urban runoff. According to Kelly et al. (2012), Ni sulphide deposits usually contain

425

high concentrations of PGE and Cu. The high concentrations of Cu can be attributed

426

to natural weathering or dissolution of Cu minerals, mining activities, sewage 19

427

treatment plant effluent, as well as runoff from the surrounding settlements. The

428

sources for Zn in aquatic ecosystems are due to natural weathering and erosion of

429

rocks and ores, but can also enter the system by means of industrial effluent, mining

430

activities, as well as sewage treatment plant effluent (DWAF, 1996). According to

431

DWAF (1996), Cd can usually be found in association with Cu, Pb and Zn sulphide

432

ores, which can be found in the Bushveld Igneous Complex. Although Pb occurs

433

naturally within the geology of this region, the concentrations of Pb within the

434

sediment at Site 2 can be attributed to the intensive mining activities. It is reported by

435

Wittmann and Förstner (1976) that Pb carbonate was used in the PGE refining

436

process, and the high concentrations of Pb at Site 2 may be historical Pb waste that

437

is still present in the environment. However, road runoff from heavy traffic (Rauch

438

and Fatoki, 2009; Hwang et al., 2016) in the adjacent area may also contribute to the

439

Pb contamination.

440

Dissolved metal concentrations were compared to the Australian and New Zealand’s

441

guidelines for fresh water quality (ANZECC) (ANZECC, 2000), European Union

442

environmental quality standards (EU-EQS) (EU-EQS, 2008), South Africa’s target

443

water quality range (TWQR) (DWAF, 1996), as well as the United States

444

Environmental Protection Agency national recommended water quality criteria for

445

aquatic life (USEPA) (USEPA, 2009). Concentrations of Ni in the water at Site 3 was

446

the highest in the river sites and even higher at the mine settling pond and exceeded

447

ANZECC of 11 µg/L, as well as EU-EQS of 20 µg/L. Concentrations of Cu in the

448

water at all the sites exceeded ANZECC and TWQR of 1.4 µg/L for very hard water,

449

as well as the chronic effect value (CEV) of 2.8 µg/L, while the concentrations of Cu

450

at Site 1 exceeded the acute effect value (AEV) of 12 µg/L for aquatic ecosystems in

451

South Africa. Although these concentrations exceeded these values, it indicates that 20

452

the toxicity of Cu is potentially mitigated due to increased water hardness and

453

dissolved oxygen (Table 1), as well as the presence of high concentrations of Zn

454

(DWAF, 1996). Concentrations of Zn in the river water at all of the sampling sites

455

exceeded the TWQR (2 µg/L) and the CEV (3.6 µg/L), while the Zn concentrations at

456

all the sites, except Site 3, exceeded ANZECC (8 µg/L). However, as discussed

457

before for Cu, also the toxicity of Zn in aquatic systems is reduced with an increase

458

in water hardness (DWAF, 1996). Therefore, although the concentrations of Zn

459

exceeded even the acute effect value, it may not be toxic to the biota within this river

460

system due to the high water hardness. High concentrations of Pb at Site 1 in the

461

water, exceeded the TWQR (1.2 µg/L) and can be attributed to the natural

462

weathering of sulphide ores (DWAF, 1996).

463

4.2. Metal bioaccumulation in macroinvertebrates

464

The trends in metal concentrations of the macroinvertebrate families were similar to

465

that seen for the sediment. Organisms collected at Site 2, generally had the highest

466

metal concentrations that again decreased at Site 3, with the exception of Ni and Zn,

467

which were higher in organisms collected at Site 3. Although the Libellulidae

468

individual was the only macroinvertebrate found in the mine settling pond, its

469

exceptionally high concentrations of Cr, Ni and Pt clearly demonstrate the biological

470

availability of these metals emitted by the mine. Thus, to our knowledge the present

471

study demonstrated for the first time that Pt from mine activities is bioavailable to

472

aquatic macroinvertebrates.

473

However, in contrast to the present work a few field studies focused on the uptake of

474

Pt from automobile traffic by macroinvertebrates in urban aquatic systems. Previous

475

field studies conducted on the accumulation of Pt by freshwater crustaceans and

21

476

clams reported concentrations ranging from below detection limit to 0.0013 µg/g

477

(Ruchter et al., 2015). Additionally, concentrations of Pt of up to 4.5 µg/g have been

478

reported in the asselid Asellus aquaticus from urban river systems in Germany and

479

Sweden (Rauch and Morrison, 1999; Moldovan et al., 2001; Haus et al., 2007).

480

Especially families associated with the sediment (Lymnaeidae, Baetidae and

481

Tubificidae) accumulated Pt at Site 2, with intensive Pt mining activities. At this site,

482

an average concentration of Pt of 0.071 µg/g was recorded in Tubificidae, which was

483

55 times higher than reported values in freshwater clams in urban environments

484

(Ruchter et al., 2015). Furthermore, Haus et al. (2007) determined a BSAF of 0.11

485

for A. aquaticus from an urban pond in Germany, which is in the same range as for

486

the different macroinvertebrate families of the present study (0.02-0.50). Even if

487

different macroinvertebrate families were compared in terms of their metal

488

accumulation patterns, it emerges that traffic-related Pt and Pt from mine activities

489

are biologically available to a similar degree.

490

The high concentrations of Cu in Lymnaeidae and Potamonautidae at all the sites

491

could be explained by the fact that these families have hemocyanin (Cu based

492

metalloproteins) for oxygen transportation rather than hemoglobin (Fe based

493

metalloproteins) (van Aardt, 1993; Brown, 1994). Even though these families have

494

naturally increased concentrations of Cu, their higher concentrations at Sites 2 and 3

495

were most probably due to anthropogenic activities (mining and settlement) that are

496

present in the catchments upstream of these sites and are reflected in the BCF and

497

BSAF of both families. Somerset et al. (2015) found concentrations of Cu of 139.1

498

µg/g DW in Potamonautidae collected from the Hex River, while the highest

499

concentration of Cu recorded in the present study is 88.2 µg/g DW.

22

500

The bioaccumulation factors clearly varied with the sampling site. According to John

501

and Leventhal (1995), the bioavailability of metals in aquatic ecosystems can be

502

influenced by several factors, i.e. among others total concentration and speciation of

503

the metal, pH, total organic content, redox potential, volume and velocity of water.

504

For example, although Site 1 had the lowest concentrations of Ni, the families

505

collected at this site had the highest BCF, while the same trend occurred with Zn,

506

Cd, Pt and Pb in the BSAF. Thus, the metals at this site, even though occurring at

507

lower concentrations in the matrices, might be more bioavailable to the

508

macroinvertebrates as compared with the other sampling sites.

509

4.3. Macroinvertebrates as biomonitoring tool

510

Macroinvertebrates can be used as an effective biomonitoring tool. In the present

511

study,

512

accumulated the highest metal concentrations showing the highest BCF and BSAF,

513

the same was found by Kemp et al. (2017) in the Marico River, South Africa. These

514

families are all associated with the sediment (Harrison, 2002; van Hoven and Day,

515

2002; Barber-James and Lugo-Ortiz, 2003), except for Lymnaeidae which are

516

generally associated with aquatic vegetation (Appleton, 2002). All of these families

517

feed on detritus, by means of either ingesting sediment or scraping it from the

518

substratum (Appleton, 2002; Harrison, 2002; van Hoven and Day, 2002; Barber-

519

James and Lugo-Ortiz, 2003).

520

According to Bervoets et al. (2016), more tolerant families (i.e. Chironomidae and

521

Tubificidae) should rather be regarded as suitable indicator organisms for ecological

522

effects in aquatic ecosystems, than sensitive families (i.e. Ephemeroptera,

523

Plecoptera and Trichoptera). From the results of their field surveys between 1990

the

families

Lymnaeidae,

Baetidae,

Tubificidae

and

Chironomidae

23

524

and 2010 in the Scheldt and Meuse River basin, metal concentrations within the

525

family Tubificidae were generally higher than in the family Chironomidae (Bervoets et

526

al., 2016), which were also the case for the present study. Bervoets et al. (2016)

527

calculated the critical body burden of metals for Tubificidae and Chironomidae (see

528

supplementary data, Table 4) based on accumulated concentrations that were

529

related to macroinvertebrate community changes. In the present study, the

530

concentrations of Cr and Ni of the Tubificidae and Chironomidae exceeded the

531

calculated critical body burdens. Nevertheless, both families were found in high

532

numbers at the sampling sites.

533

The family Baetidae had the highest concentrations of Zn and Cd, as well as

534

relatively high concentrations of Cr. Various authors regard this family as the most

535

tolerant family towards metal pollution within the order Ephemeroptera (Awrahman et

536

al., 2016; Bervoets et al., 2016). Metal concentrations within lower trophic families

537

(i.e. scraper-grazers, collector-gatherers and shredders) were generally higher as

538

compared with predator families (Coenagrionidae and Libellulidae) (Liess et al.,

539

2017). This is supported by the work of Goodyear and McNeill (1999), Gray (2002),

540

Cardwell et al. (2013), Kemp et al. (2017), Liess et al. (2017) and Rodriquez et al.

541

(2018) where these metals were reported to not biomagnify within the food chain.

542

According to Liess et al. (2017), possible reasons for lower metal concentrations in

543

predator families can be that (i) predator families have the ability to regulate their

544

internal metal concentrations better than lower trophic families, and (ii) predator

545

families also have higher body mass to surface ratios, which can account for lower

546

metal uptake rates. Although these metals (Zn, Cd, Cr) are considered as potentially

547

toxic to aquatic biota at the high concentrations found in the present study, we did

548

not observe any obvious lethal effects. This is possibly due to protective 24

549

mechanisms that aquatic invertebrates utilize to reduce metal exposure, e.g.

550

elimination (excretion), detoxification (e.g. metallothioneins) and sequestration

551

(Hoffman et al., 2002; Ahearn et al., 2004; Cardwell et al., 2013; Mugwar &

552

Harbottle, 2016; Kemp et al., 2017). Kemp et al. (2017) found that metallothioneins

553

were even present in macroinvertebrates from a region where metal contamination

554

was not present, but naturally high metal concentrations from geological background,

555

which indicates to the inherent natural protective role of metallothioneins in

556

macroinvertebrates.

557 558

5. Conclusion

559

The aim of this study was to determine the concentrations of metals in water,

560

sediment and aquatic macroinvertebrates associated with Pt mining in the Hex River,

561

South Africa. Based on the bioaccumulation in the biota, the biological availability of

562

metals was evaluated at river sites affected by mining activities. From the present

563

results, it can be concluded that Pt mining activities are a major source of Cr and Pt

564

contamination within the Hex River system. However, urban and informal

565

settlements, as well as the geological background also contribute to the metal

566

concentrations found in this system. These results provide valuable information and

567

bridge the gap regarding the concentrations associated with Pt mining and

568

processing activities, as well as the concentrations within biota occurring within these

569

systems. The results also demonstrated bioaccumulation of these metals in

570

macroinvertebrates in relation to the metal concentrations in water and sediment,

571

respectively, at specific sites. Although Pt mining in South Africa provide efficient

572

supplies to the world’s demand for Pt, they are a major source of metal

25

573

contamination within aquatic ecosystems associated with these mining regions.

574

These aquatic ecosystems can be monitored by using macroinvertebrates as

575

biomonitoring tool, especially the families Lymnaeidae, Baetidae, Tubificidae, as well

576

as Chironomidae. Although high metal concentrations were recorded within selected

577

families, there were no obvious detrimental effects on them, leading to the

578

assumption that they possess strategies to cope with these concentrations.

579

Recommendations for future studies include the consideration of (i) a wider spectrum

580

of monitoring organisms such as transplanted native freshwater clams (Corbicula

581

fluminalis) and vertebrate species e.g. fish, (ii) passive sampling devices, (iii) effect

582

indication (e.g. molecular biomarkers) to evaluate whether these environmental

583

concentrations have negative effects on the biota and compare it with the effects

584

found in a laboratory study (Brand et al., 2019), as well as (iv) other river systems

585

draining Pt mining regions to assess whether similar trends occur.

586 587

Acknowledgements

588

This work is based on the research and researchers supported by the National

589

Research Foundation (NRF) of South Africa (NRF Project GERM160623173784;

590

grant 105875; NJ Smit, PI) and BMBF/PT-DLR (Federal Ministry of Education and

591

Research, Germany, grant 01DG17022; B Sures, PI). Opinions, findings,

592

conclusions and recommendations expressed in this publication are that of the

593

authors, and the NRF and BMBF/PT-DLR accepts no liability whatsoever in this

594

regard. Miss Anja Greyling, NWU-Water Research Group, is thanked for the creation

595

of the study area map. Dr. Ruan Gerber and Miss Marelize Labuschagne, NWU-

596

Water Research Group, are thanked for their assistance in the field. This is

26

597

contribution number 348 of the North-West University (NWU) Water Research

598

Group.

599

27

600

References

601

Ahearn, G.A., Mandal, P.K., Mandal, A., 2004. Mechanisms of heavy-metal

602

sequestration and detoxification in crustaceans: a review. J. Comp. Physiol. B.

603

174(6), 439-452. https://doi.org/10.1007/s00360-004-0438-0.

604

Almécija, C., Cobelo-Garcia, A., Wepener, V., Prego, R., 2017. Platinum group

605

elements in stream sediments of mining zones: The Hex River (Bushveld Igneous

606

Complex,

607

https://doi.org/10.1016/j.jafrearsci.2017.02.002.

608

Appleton, C.C., 2002. Mollusca, in: Day, J.A., de Moor, I.J. (Eds.), Guides to the

609

Freshwater Invertebrates of Southern Africa, Araneae, Water Mites and Mollusca,

610

WRC Report No. TT 182/02. Water Research Commission, Pretoria, pp. 42-125.

611

Australian and New Zealand Environment and Conservation Council (ANZECC).

612

2000. Australian and New Zealand Guidelines for Fresh and Marine Water Quality,

613

Australian Water Association, Canberra.

614

Awrahman, Z.A., Rainbow, P.S., Smith, B.D., Khan, F.R., Fialkowski, W., 2016.

615

Caddisflies Hydropsyche spp. as biomonitors of trace metal bioavailability thresholds

616

causing disturbance in freshwater stream benthic communities. Environ Pollut. 216,

617

793-805. https://doi.org/10.1016/j.envpol.2016.06.049.

618

Barber-James, H.M., Lugo-Ortiz, C.R., 2003. Ephemeroptera, in: de Moor, I.J., Day,

619

J.A., de Moor, F.C. (Eds.), Guides to the Freshwater Invertebrates of Southern

620

Africa, Ephemeroptera, Odonata and Plecoptera, WRC Report No. TT 207/03. Water

621

Research Commission, Pretoria, pp. 16-142.

South

Africa).

J.

Afr.

Earth

Sci.

129,

934-943.

28

622

Barnes, S.J., Maier, W.D., 2002. Platinum-group elements and microstructures of

623

normal Merensky Reef from Impala platinum mines, Bushveld complex. J. Petrol. 43,

624

103-128. https://doi.org/10.1093/petrology/43.1.103.

625

Bervoets, L., de Jonge, M., Blust, R., 2016. Identification of threshold body burdens

626

of metals for the protection of the aquatic ecological status using two benthic

627

invertebrates.

628

https://doi.org/10.1016/j.envpol.2015.12.005.

629

Blais, J.F., Djedidi, Z., Cheikh, R.B., Tyagi, R.D., Mercier, G., 2008. Metals

630

Precipitation from Effluents: Review. Pract. Period. Hazard. Toxic Radioact. Waste

631

Manage. 12(3), 135-149. https://doi.org/10.1061/(ASCE)1090-025X(2008)12:3(135).

632

Brand, S.J., Erasmus, J.H., Labuschagne, M., Grabner, D., Nachev, M.,

633

Zimmermann, S., Wepener, V., Smit, N.J., Sures, B., 2019. Bioaccumulation and

634

metal-associated

635

polymorpha, following short-term platinum exposure. Environ Pollut. 246, 69-78.

636

https://doi.org/10.1016/j.envpol.2018.11.061.

637

Brown, D.S., 1994. Freshwater Snails of Africa and their Medical Importance, CRC

638

Press, London.

639

Cardwell, R.D., DeForest, D.K., Brix, K.V., Adams, W.J., 2013. Do Cd, Cu, Ni, Pb

640

and Zn Biomagnify in Aquatic Ecosystems? Rev. Environ Contam T. 226, 101-122.

641

https://doi.org/10.1007/978-1-4614-6898-1_4.

642

Cawthorn, R.G., 2010. The platinum group element deposits of the Bushveld

643

Complex

644

https://doi.org/10.1595/147106710x520222.

in

Environ

biomarker

South

Pollut.

responses

Africa.

in

Platinum

a

210,

freshwater

Metals

Rev.

mussel,

54(4),

76-84.

Dreissena

205-215.

29

645

Cempel, M., Nikel, G., 2006. Nickel: A Review of Its Sources and Environmental

646

Toxicology. Pol. J. of Environ. Stud. 15(3), 375-382.

647

Department of Water Affairs (DWA). 2011. Green Drop Report 2011: Waste Water

648

Service Regulation, The Government Printer, Pretoria.

649

Department of Water Affairs and Forestry (DWAF). 1996. South African Water

650

Quality Guidelines. Volume 7: Aquatic Ecosystems, The Government Printer,

651

Pretoria.

652

Djingova, R., Heidenreich, H., Kovacheva, P., Markert, B., 2003. On the

653

determination of platinum group elements in environmental materials by inductively

654

coupled plasma mass spectrometry and microwave digestion. Anal. Chim. Acta.

655

489(2), 245-251. https://doi.org/10.1016/S0003-2670(03)00716-5.

656

Ek, K.H., Rauch, S., Morrison, G.M., Lindberg, P., 2004. Platinum group elements in

657

raptor eggs, faeces, blood, liver and kidney. Sci. Total Environ. 334-335, 149-159.

658

https://doi.org/10.1016/j.scitotenv.2004.04.067.

659

European

660

Environmental Quality Standards for Priority Substances and certain other

661

Pollutants, Official Journal of the European Union, Strasbourg.

662

Goodyear, K.L., McNeill, S., 1999. Bioaccumulation of heavy metals by aquatic

663

macro-invertebrates of different feeding guilds: a review. Sci. Total Environ. 229(1-

664

2), 1-19. https://doi.org/10.1016/S0048-9697(99)00051-0.

665

Gray, J.S., 2002. Biomagnification in marine systems: the perspective of an

666

ecologist. Mar. Pollut. Bull. 45, 46-52. https://doi.org/10.1016/S0025-326X(01)00323-

667

X.

Parliament,

Council

of

the

European

Union

(EU-EQS).

2008.

30

668

Haggard, E.L., Sheridan, C.M., Harding, K.G., 2015. Quantification of water usage at

669

a

670

http://dx.doi.org/10.4314/wsa.v41i2.14.

671

Harrison, A.D., 2002. Chironomidae, in: Day, J.A., Harrison, A.D., de Moor, I.J.

672

(Eds.), Guides to the Freshwater Invertebrates of Southern Africa, Diptera, WRC

673

Report No. TT 201/02. Water Research Commission, Pretoria, pp. 110-158.

674

Haus, N., Zimmermann, S., Wiegand, J., Sures, B., 2007. Occurrence of platinum

675

and additional traffic related heavy metals in sediments and biota. Chemosphere. 66,

676

619-629. http://dx.doi.org/10.1016/j.chemosphere.2006.07.097.

677

Hoffman, D.J., Rattner, B.A., Burton, G.A., Cairns, J., 2002. Handbook of

678

Ecotoxicology, second ed. Lewis Publishers, New York.

679

Hwang, H., Fiala, M.J., Park, D., Wade, T.L., 2016. Review of pollutants in urban

680

road dust and stormwater runoff: part 1. Heavy metals released from vehicles. Int. J.

681

Urban Sci. 20(3), 334-360. http://dx.doi.org/10.1080/12265934.2016.1193041.

682

John, D.A., Leventhal, J.S., 1995. Bioavailability of metals, in: Du Bray, E.A. (ed.),

683

Preliminary Compilation of Descriptive Geoenvironmental Mineral Deposit Models.

684

U.S. Department of the Interior. U.S. Geological survey.

685

Johnson-Matthey. 2018. PGM Market Report 2018. Johnson-Matthey, London.

686

Kelly, M., Allison, W.J., Garman, A.R., Symon, C.J., 2012. Mining and the

687

Freshwater Environment. Elsevier Science Publishers Ltd., Essex.

688

Kemp, M., Wepener, V., de Kock, K.N., Wolmarans, C.T., 2017. Metallothionein

689

Induction as Indicator of Low Level Metal Exposure to Aquatic Macroinvertebrates

South

African

platinum

processing

plant.

WaterSA.

41(2),

279-286.

31

690

from a Relatively Unimpacted River System in South Africa. Bull Environ Contam

691

Toxicol. 99, 662-667. http://dx.doi.org/10.1007/s00128-017-2201-8.

692

Kenney, M.A., Sutton-Grier, A.E., Smith, R.F., Gresens, S.E., 2009. Benthic

693

macroinvertebrates as indicators of water quality: The intersection of science and

694

policy.

695

http://dx.doi.org/10.1163/187498209X12525675906077.

696

Liess, M., Gerner, N.V., Kefford, B.J., 2017. Metal toxicity affects predatory stream

697

invertebrates less than other functional feeding groups. Environ Pollut. 227, 505-512.

698

http://dx.doi.org/10.1016/j.envpol.2017.05.017.

699

Maboeta, M.S., Claassens, S., van Rensburg, L., Jansen van Rensburg, P.J., 2006.

700

The Effect of Platinum Mining on the Environment from a Soil Microbial Perspective.

701

Water Air Soil Pollut. 175(1-4), 149-161. http://dx.doi.org/10.1007/s11270-006-9122-

702

1.

703

Merritt, R.W., Cummins, K.W., Berg, M.B., 2008. An Introduction to the Aquatic

704

Insects of North America, fourth ed. Kendall Hunt Publishing, Dubuque.

705

Moldovan, M., Rauch, S., Gómez, M., Palacios, M.A., Morrison, G.M., 2001.

706

Bioaccumulation of palladium, platinum and rhodium from urban particulates and

707

sediments by the freshwater isopod Asellus aquaticus. Water Res. 35(17), 4175-

708

4183. http://dx.doi.org/10.1016/S0043-1354(01)00136-1.

709

Moore, J.W., 2006. Animal ecosystem engineers in streams. BioSci. 56, 237-246.

710

http://dx.doi.org/10.1641/0006-3568(2006)056[0237:AEEIS]2.0.CO;2.

Terr.

Arthropod

Rev.

2,

99-128.

32

711

Mugwar, A.J., Harbottle, M.J., 2016. Toxicity effects on metal sequestration by

712

microbially-induced carbonate precipitation. J. Hazard. Mater. 314, 237-248.

713

http://dx.doi.org/10.1016/j.jhazmat.2016.04.039.

714

Nawrocka, A., Szkoda, J., 2012. Determination of Chromium in Biological Material by

715

Electrothermal Atomic Absorption Spectrometry Method. Bull Vet Inst Pulawy. 56,

716

585-589. https://doi.org/10.2478/v10213-012-0103-4.

717

Okorie, I.A., Enwistle, J., Dean, J.R., 2015. Platinum group elements in urban road

718

dust. Curr. Sci. 109, 938-942. https://doi.org/10.18520/v109/i5/938-942.

719

Pawlak, J., Łodyga-Chruścińska, E., Chrustowicz, J., 2014. Fate of platinum metals

720

in

721

https://doi.org/10.1016/j.jtemb.2014.03.005.

722

Rauch, S., Fatoki, O.S., 2009. Platinum and lead in South African road dust, in:

723

Rauch, S., Morrison, G., Monzón, A. (Eds.) Highway and Urban Environment.

724

Springer, Dordrecht.

725

Rauch, S., Fatoki, O.S., 2013. Anthropogenic platinum enrichment in the vicinity of

726

mines in the Bushveld Igneous Complex, South Africa. Water Air Soil Poll. 224,

727

1395-1406. https://doi.org/10.1007/s11270-012-1395-y.

728

Rauch, S., Fatoki, O.S., 2015. Impact of Platinum Group Element Emissions from

729

Mining and Production Activities, in: Zereini, F., Wiseman, C.L.S. (Eds.), Platinum

730

Metals in the Environment. Springer, New York, pp. 19-29.

731

Rauch, S., Morrison, G.M., 1999. Platinum uptake by the freshwater isopod Asellus

732

aquaticus

733

https://doi.org/10.1016/S0048-9697(99)00203-X.

the

environment.

in

urban

J.

rivers.

Trace

Sci.

Elem.

Total

Med.

Environ.

Bio.

28,

235(1-3),

247-254.

261-268.

33

734

Rodriquez, P., Méndez-Fernández, L., Pardo, I., Costas, N., Martinez-Madrid, M.,

735

2018. Baseline tissue levels of trace metals and metalloids to approach ecological

736

threshold concentrations in aquatic macroinvertebrates. Ecol. Indic. 91, 395-409.

737

https://doi.org/10.1016/j.ecolind.2018.04.004.

738

Ruchter, N., Zimmermann, S., Sures, B., 2015. Field studies on PGE in Aquatic

739

Ecosystems, in: Zereini, F., Wiseman, C.L.S. (Eds.), Platinum Metals in the

740

Environment. Springer, New York, pp. 351-360.

741

Somerset, V., van der Horst, C., Silwana, B., Walters, C., Iwuoha, E., 2015.

742

Biomonitoring and Evaluation of Metal Concentrations in Sediment and Crab

743

Samples from the North-West Province of South Africa. Water Air Soil Poll. 226, 43-

744

68. https://doi.org/10.1007/s11270-015-2329-2.

745

United

746

Recommended Water Quality Criteria – Aquatic Life Criteria, Office of Science and

747

Technology, Washington, DC.

748

van Aardt, W.J., 1993. The influence of temperature on the heamocyanin function

749

and oxygen uptake of the freshwater crab Potamonautes warreni. Comp. Biochem.

750

Physiol. A Physiol. 106(1), 31-35. https://doi.org/10.1016/0300-9629(93)90034-2.

751

van Hoven, W., Day, J.A., 2002. Oligocheata, in: Day, J.A., de Moor, I.J. (Eds.),

752

Guides to the Freshwater Invertebrates of Southern Africa, Non-Arthropods, WRC

753

Report No. TT 167/02. Water Research Commission, Pretoria, pp. 203-236.

754

Wepener, V., Vermeulen, L.A. 2005. A note on the concentrations and bioavailability

755

of selected metals in sediments of Richards Bay Harbour, South Africa. Water SA,

756

34: 589-595.

States

Environmental

Protection

Agency

(USEPA).

2009.

National

34

757

Wittmann, G.T.W., Förstner, U., 1976. Heavy Metal Enrichment in Mine Drainage: I.

758

The Rustenburg Platinum Mining Area. S. Afr. J. Sci. 72, 242-246.

759

Zereini, F., Wiseman, C.L.S., 2015. Platinum Metals in the Environment. Springer,

760

New York.

761 762

Table 2: Recovery rates (%) of the elements of interest obtained for the different

763

certified reference materials used in the analysis. Elements

764

Recovery rates (%) BCR-723

IAEA-407

DORM-3

DOLT-3

Cd

88

101

110

96

Cr

80

84

96

119

Cu

n.c.

97

100

110

Ni

94

90

118

93

Pb

118

108

91

85

Pt

104

n.c.

n.c.

n.c.

Zn

94

86

86

96

n.c.: not certified

765

35

766

Table 2: Water and sediment quality variables analyzed at the three Hex River sites

767

(Site 1 – 3), as well as the mine settling pond (Mine) and urban effluent (Urban). The

768

codes for each variable are used in the PCA biplot. Code

Site 1

Site 2

Site 3

Mine

Urban

Dissolved oxygen (mg/L)

DO

7.3

5.5

1.5

n.m.

3.8

Electrical conductivity (μS/cm)

EC

375

548

1646

1775

3950

pH

pH

8.2

8.3

7.8

8.0

7.6

Suspended solids (mg/100mL)

SS

1.8

1.5

3.6

9.9

1.7

Temperature (°C)

TEMP

23.8

24.9

23.8

25.6

23.0

Turbidity (FAU)

TURB

12

13

14

12

17

Organic content (%)

OC

7.7

5.7

3.2

n.m.

1.4

Gravel (%)

GR

16.0

13.4

3.4

n.m.

6.7

Very course sand (%)

VCS

6.7

11.4

2.5

n.m.

27.4

Course sand (%)

CS

22.6

35.1

20.3

n.m.

52.8

Medium sand (%)

MS

19.0

18.5

43.8

n.m.

9.7

Fine sand (%)

FS

16.9

14.7

25.1

n.m.

3.0

Mud (%)

M

18.8

7.0

5.0

n.m.

0.4

Water quality variable

Sediment quality variable

769

n.m.: not measured

36

770

Table 3: Bioconcentration factors (BCF) calculated as the ratio of the metal

771

concentrations in macroinvertebrate families and metal concentrations in water

772

collected from selected sites in the Hex River (Site 1 – 3) and a mine settling pond

773

(Mine). Met

Sit

al

e Sit e1 Sit

Ni

e2 Sit e3 Min e Sit e1 Sit

Cu

e2 Sit e3 Min e Sit e1 Sit

Zn

e2 Sit e3 Min e Sit e1 Sit

Cd

e2 Sit e3 Min e Sit e1

Pt

Sit e2 Sit

Macroinvertebrate families Lymnaei

Baetid

Hydropsych

Chironomi

Tubifici

Potamonaut

Coenagrion

Libelluli

dae

ae

idae

dae

dae

idae

idae

dae

9 031

8 239

-

-

-

4 761

8 446

5 323

5 387

2 541

1 695

-

4 315

2 288

1 008

1 681

1 979

377

-

765

-

830

482

698

-

-

-

-

-

-

-

1 241

2 285

1 533

-

-

-

2 038

1 060

748

7 304

3 535

2 639

-

2 705

8 052

2 221

3 149

16 435

3 234

-

5 271

-

14 845

3 283

4 189

-

-

-

-

-

-

-

6 330

1 079

3 326

-

-

-

2 035

3 523

2 696

1 774

10 421

4 980

-

9 268

3 208

5 948

4 969

6 176

27 143

-

10 529

-

8 816

11 380

9 905

-

-

-

-

-

-

-

15 629

872

3 522

-

-

-

333

820

294

1 275

8 927

2 073

-

3 389

1 065

3 062

1 081

302

303

-

251

-

224

119

129

-

-

-

-

-

-

-

155

89

155

-

-

-

23

108

117

283

170

45

-

972

61

31

66

99

47

-

82

-

36

47

36

37

e3 Min e Sit e1 Sit Pb

e2 Sit e3 Min e

-

-

-

-

-

-

-

108

565

777

-

-

-

229

597

653

1 209

1 702

1 742

-

3 338

519

563

844

2 011

1 301

-

3 692

-

814

1 435

1 963

-

-

-

-

-

-

-

2 052

774

38

775

Table 4: Bioaccumulation factors (BSAF) calculated as the ratio of the metal

776

concentrations in macroinvertebrate families and metal concentrations in sediment

777

collected from selected sites in the Hex River (Site 1 – 3) and a mine settling pond

778

(Mine). Met

Sit

al

e Site 1 Site

Cr

2 Site 3 Min e Site 1 Site

Ni

2 Site 3 Min e Site 1 Site

Cu

2 Site 3 Min e Site 1 Site

Zn

2 Site 3 Min e Site 1 Site

Cd

2 Site 3 Min e

Macroinvertebrate families Lymnaei

Baetid

Hydropsychi

Chironomi

Tubificid

Potamonauti

Coenagrioni

Libellulid

dae

ae

dae

dae

ae

dae

dae

ae

0.05

0.07

-

-

-

0.02

0.07

0.06

0.03

0.05

0.03

-

0.07

0.01

0.02

0.03

0.02

0.02

-

0.06

-

0.01

0.02

0.03

-

-

-

-

-

-

-

0.04

0.88

0.81

-

-

-

0.47

0.83

0.52

0.97

0.46

0.31

-

0.78

0.41

0.18

0.30

2.27

0.43

-

0.88

-

0.95

0.55

0.80

-

-

-

-

-

-

-

0.82

4.34

2.91

-

-

-

3.87

2.02

1.42

3.82

1.85

1.38

-

1.42

4.21

1.16

1.65

8.81

1.73

-

2.83

-

7.96

1.76

2.25

-

-

-

-

-

-

-

1.50

3.38

10.42

-

-

-

6.37

11.04

8.44

1.48

8.71

4.16

-

7.75

2.68

4.97

4.15

2.54

11.17

-

4.33

-

3.63

4.68

4.07

-

-

-

-

-

-

-

1.25

1.73

6.99

-

-

-

0.66

1.63

0.58

0.62

4.32

1.00

-

1.64

0.52

1.48

0.52

0.23

0.23

-

0.19

-

0.17

0.09

0.10

-

-

-

-

-

-

-

0.15

39

Site 1 Site 2

Pt

Site 3 Min e Site 1 Site 2

Pb

Site 3 Min e

0.18

0.30

-

-

-

0.05

0.21

0.23

0.14

0.09

0.02

-

0.50

0.03

0.02

0.03

0.17

0.08

-

0.14

-

0.06

0.08

0.06

-

-

-

-

-

-

-

0.10

0.12

0.17

-

-

-

0.05

0.13

0.14

0.05

0.08

0.08

-

0.15

0.02

0.03

0.04

0.09

0.06

-

0.16

-

0.04

0.06

0.09

-

-

-

-

-

-

-

0.17

779 780 781 782

Highlights 

macroinvertebrates

783 784

Quantification of Cr, Ni, Cu, Zn, Cd, Pt, Pb in water, sediment and



Metal contamination increased in water and sediment downstream of mining activities

785 786



Metal bioaccumulation was significant in macroinvertebrate families

787



Macroinvertebrate families showed different grades of bioaccumulation

788



Urban and informal settlements also contributed to metal pollution

789

40

790

791

41

792

42

793

43

794

44

795 45

796

46