Copper speciation in continental inputs to the Vigo Ria: Sewage discharges versus river fluxes

1

2

3

4

5

67

8

9

10

11

12

13

14

15

16

17

18

1920

21

22

23

24

25

26

27

28

29

30

31

32

Available online at www.sciencedirect.com

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

www.elsevier.com/locate/marpolbul

Marine Pollution Bulletin xxx (2007) xxx–xxx

OO

F

Copper speciation in continental inputs to the Vigo Ria:Sewage discharges versus river fluxes

Juan Santos-Echeandia a,b,*, Luis M. Laglera a, Ricardo Prego b,Constant M.G. van den Berg a

a Department of Earth and Ocean Sciences, University of Liverpool, Liverpool L69 3GP, UKb Marine Biogeochemistry Research Group, Instituto de Investigaciones Marinas (CSIC), C/Eduardo Cabello 6, Vigo 36208, Spain

CTED

PR

Abstract

Continental inputs of copper via rivers and sewage into the Vigo Ria were evaluated. The main fluvial input is not contaminated andthe most degraded discharges occur on the southern margin of the middle ria. Continental inputs of copper and ligands to the ria aredominated by sewage treatment plants (136 mol Cu day�1, 124 mol L day�1) supported by rivers (15 mol Cu day�1, 21 mol L day�1).The dissolved fraction is the main channel of discharge for rivers (66%) with particulate matter being predominant in sewage (63%). Dis-solved copper is organically complexed both in rivers (99.8%) and sewage (99.9%). This minor difference may be attributed to the factthat the stability of sewage complexes is greater than those in rivers. Moreover, ligand concentrations are higher in sewage than in rivers.Thus, the natural continental inputs of copper and ligands into the ria are magnified by anthropogenic inputs (5–15 and 3–5 times higherfor copper and ligands, respectively).� 2007 Published by Elsevier Ltd.

Keywords: Copper; Speciation; Rivers; Sewages; Contamination; Vigo Ria

33

34

35

36

37

38

39

40

41

42

43

44

45

NC

OR

RE

1. Introduction

Estuaries are transition areas between the land and theocean where vast amounts of materials, such as trace met-als, are discharged. These estuaries are generally denselypopulated areas with industrial activity, which adds ananthropogenic factor to the natural contributions andmay, to some extent, change the biogeochemical cyclesof trace metals such as copper. These elements reach theestuaries through various routes, some of which may beatmospheric inputs, continental inputs, or uncontrolleddischarges of a punctual nature. Continental inputs maybe canalized as rivers and sewage, or they may reach

47

48

49

50

51

52

0025-326X/$ - see front matter � 2007 Published by Elsevier Ltd.

doi:10.1016/j.marpolbul.2007.10.021

* Corresponding author. Address: Marine Biogeochemistry ResearchGroup, Instituto de Investigaciones Marinas (CSIC), C/Eduardo Cabello6, Vigo 36208, Spain. Tel.: +34 986 231 930; fax: +34 986 292 762.

E-mail addresses: [email protected] (J. Santos-Echeandia), [email protected] (L.M. Laglera), [email protected] (R. Prego), [email protected] (C.M.G. van den Berg).

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

the brackish water bodies in an uncontrolled manner inthe form of runoff or industrial wastes (Clark, 2001). Inthe case of rivers, the amount of copper discharged willdepend on land use and the size of the population itpasses through. Copper can enter rivers and streams froma number of sources – both natural and anthropogenic-within the catchment and directly from effluent discharges(Hart and Hines, 1995). Under most circumstances, oneof the major contributions of the anthropogenic copperload that shows up in the sea comes from urban andindustrial developments along the rivers and estuaries(Ridgway et al., 2003). With respect to sewages, the metalflow depends mainly upon the type of water treated(domestic or industrial), the level of treatment used inthe sewage, and the size of the population using the sew-age treatment plant (Lester et al., 1979; Laxen and Harri-son, 1981).

Quantification of land-derived copper fluxes to thecoastal systems is therefore a key factor in determiningthe extent to which these inputs influence the natural bio-

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

T

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

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

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

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

2 J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

UN

CO

RR

EC

geochemical processes of the elements in the sea.Although it is important to compare river and sewageinputs to investigate the main source of metals to an estu-ary (Helz et al., 1975; Huh, 1996), there have been nocomplete studies that focus on the two sources. Most ofthe papers center on riverine inputs (Apte et al., 1990;Pettine et al., 1996; Loubet et al., 2003; Cleven et al.,2005) or sewage inputs (Sedlak et al., 1997; Van Veenet al., 2002), but there has been only one study comparingthe two (Helz et al., 1975). This is especially critical incoastal areas where riverine/land-based and urban andindustrial sewage inputs are discharged into semi-enclosedembayments, which is the case in some of the GalicianRias (Cobelo-Garcia et al., 2004).

From a toxicological and biogeochemical point of view,metal speciation may be more important than the totalamount of a metal discharged. Most of the copper in nat-ural waters is bound to organic and inorganic ligands, leav-ing only around 1% of the total copper as ‘‘free’’ copperthat is considered to be bioavailable, or toxic (Allen andHansen, 1996). The speciation of copper in riverine andsewage inputs is thus critical to the evaluation of the watersreaching an estuary. In terms of defining the impact of thedischarge on the receiving water, another major factor isthe speciation of the metal within the final effluent (Laxenand Harrison, 1981). Some studies on copper speciation inrivers have been carried out (Apte et al., 1990; Xue et al.,1996; Pettersson et al., 1997; Buykx et al., 1999), butdespite the importance of metal speciation in sewages (Ster-rit and Lester, 1984), only a limited number of publicationshave focused on copper speciation (Sedlak et al., 1997; Sar-athy and Allen, 2005).

As has been pointed out in a recent revision (Prego andCobelo-Garcia, 2003), there is a dire lack of studies dealingwith continental fluxes of metals in the Galician Rias. Onlytwo investigations have been carried out in this direction –the first in the Pontevedra Ria (Cobelo-Garcıa and Prego,2003) and the second in the Ferrol Ria (Cobelo-Garciaet al., 2004). However, they are not as complete as theone presented in this manuscript in terms of number andcomparison of different inputs. If we consider the impor-tance of the Galician Rias from the standpoint of industry,aquaculture (15% of the global mussel production androughly 50% of the European production; Smaal, 2002),port activities, fishing and tourism, it is imperative thatsome kind of control be exercised over these continentalinputs and their possible negative impact, which is notcompatible with the normal use of these rias.

Thus, the objectives of this study are: (i) to quantifycopper fluxes and the speciation of this metal from themain rivers to the ria; (ii) to quantify copper dischargesand their speciation from the sewage treatment plantslocated on the banks of the ria; (iii) to compare the rela-tive importance of both types of copper inputs to the ria;and (iv) to establish a baseline to determine continentalinput copper speciation parameters and ligand fluxes toa European estuary.

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

ED

PR

OO

F

1.1. Study area

The Vigo Ria, an incised valley where the estuarine zonemoves according to climate changes (Evans and Prego,2003), occupies an area of 156 km2 with a volume of3275 km3 of water. Both sides of the estuary are highlypopulated with approximately 400,000 inhabitants, 77%of which belong to the City of Vigo located on the southcentral margin. Over the past few decades, industrial andport activities have come to occupy the banks of this ria,with the most important industrialized area being the Bou-zas shipyard (Fig. 1). The Vigo Ria is the most industrial-ized and populated ria in Galicia, together with the smallFerrol Ria.

The Vigo Ria has an oceanic climate and tends to bevery dry in summer (Perez-Alberti, 1982), with an annualrainfall of 1950 ± 330 mm. January (285 mm) and August(31 mm) are the months with the highest and lowest rain-fall, respectively.

The main continental inputs of materials to the Vigo Riacome from the discharges of six rivers and their six sewagetreatment plants (STP) (Fig. 1). The main tributary is theOitaven River (around 80% of the total freshwater flowto the ria), whose volume ranges from 55 m3 s�1 in Febru-ary to 1 m3 s�1 in August. It is located at the head of the riaand oriented approximately parallel to its longitudinal axis.Several other small transversal rivers, such as the Lagares(10%), Alvedosa (5%), Ullo (3%), Maior and Fraga (1%),flow into the estuary. In spite of the low flow rate of thesesecondary rivers, it is important to note that some mayplay an important role in determining the amount copperflux to the ria, as will be discussed later. In terms of popu-lation, the Lagares is the most populated basin, as it runsthrough the relatively large city of Vigo (Perez-Arluceaet al., 2005). As far as land use is concerned, some of thebasins have large cultivated areas, i.e., approximately43% of the Fraga, Alvedosa and Ullo basins, with the restof the basins having values of under 10% (Pazos et al.,2000). The most important STP is the one located in thecity of Vigo, which treats the equivalent of a populationof 400,000, followed by the plants of Moana and Cangasserving populations of 35,000 and 30,000, respectively.

2. Sampling and methods

2.1. Sampling

2.1.1. Water samples

Twelve surface water samples (from six rivers and sixsewages) were collected all along the coastal margin ofthe Vigo Ria on 14th December 2004 (Fig. 1) to monitorcontinental inputs into the Ria. The sampling was carriedout in winter since riverine fluxes are very low in summer.All samples were collected in polyethylene bottles (Azlon)that were previously acid-washed and rinsed with Milli-Q50 (MQ) water. Prior to filling the bottles with the sample,they were conditioned with the same water three times. On

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

TD

PR

OO

F163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

8º 05’ 8º 54’ 8º 40’W

’02º24

’51º24

N’01º24

sdnalsIseíC

névatiOreviR

reviRasodevlA

seragaLreviR

emoHepaC

ademaLPo ni t

mk5NAIREBI

ALUSNINEP

aiRogiV

PTSogiV

sieTPTS

PTSsagnaC

añaoMPTS

PTSalednodeR

PTSedacrA

reviRollUroiaMreviR

agarFreviR

seragaL

asodevlAnévatiO

arF ga

roiaM ollU

mk01

draypihssazuoBaera

N

Fig. 1. Study area map. Black dots represent sewage treatment plants. River inputs are also located. The little map inside shows the different basins of therivers discharging into the Ria.

J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx 3

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

REC

the same day, the samples were filtered in the IIM-CSIClaboratories of Vigo through a 0.2 lM polypropylenepore-filter (Pall Corporation) inside an ‘‘clean’’ laboratory(HEPA filtered air, class 1000 lab, with a class 100 laminarflow hood) to avoid contamination. Filtrates were frozenand stored at �20 �C until being thawed at room tempera-ture for an analysis of total dissolved copper and coppertitrations at the University of Liverpool laboratories. Fil-ters were kept for suspended particulate matter (SPM)analysis in the Marine Research Institute of Vigo.

200

201

202

203

204

205

206

207

208

209

210

211

212

213

214

215

216

217

UN

CO

R

2.2. Copper analysis in seawater

2.2.1. Total dissolved copper concentrations

Copper analyses were carried out using voltammetricequipment consisting of an Autolab PGStat 10 voltamme-ter connected to a Metrohm VA 663 electrode stand andcontrolled by a computer (PC). The reference electrodewas a double junction, Ag/AgCl, KCl (3 M), saturatedAgCl, with a salt-bridge filled with 3 M KCl, and the coun-ter electrode was a glassy carbon rod. A hanging mercurydrop electrode (HMDE) was used as the working electrode.

The copper concentration in the filtered samples wasdetermined using a procedure similar to the one describedpreviously (Campos and van den Berg, 1994). Seawaterwas UV-digested (1 h) after acidification to pH 2.2 by theaddition of 10 lL 6 M bidistilled HCl (AnalR BDH) per10 mL of sample in acid-cleaned silica tubes. In the caseof sewages, in addition to the 10 lL of HCl, 10 lL of a

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

E

30% H2O2 solution (Merck Suprapur) was added in orderto ensure the complete oxidation of the organic matter. A10 mL sample aliquot was pipetted into the voltammetriccell and ammonia (AristarGrade Merck) was used to ascer-tain the approximate neutralization of the pH; also,HEPES buffer (BDH, final concentration 0.01 M) and Sal-ycildoxime (SA, Sigma, final concentration of 25 lM SA)were added. The solution was deaerated by purging(5 min) with nitrogen. The voltammetric parameters were:deposition 60 s at �1.1 V whilst stirring, 8 s quiescence at�0.1 V, and a potential scan using the square-wave modu-lation: 10 Hz, step height 2.5 mV, pulse height 25 mV, from0 to �0.8 V. The sensitivity was calibrated by incorporat-ing standard copper additions (Spectrosol BDH) to eachsample. A blank with MQ water was made at intervals ofevery three samples and its concentration was subtractedin order to eliminate any copper contribution fromreagents. Blank concentrations were around 0.15 nM.

2.2.2. Copper titrations

The procedure used for copper titrations was similar tothe one described previously (Campos and van den Berg,1994). A 170-ml filtered sample was transferred to a poly-ethylene bottle (Nalgene), and a HEPES buffer (0.01 M)and SA (10 lM) were added. The mix was stirred thor-oughly. Next, 10-mL aliquots were pipetted into 15,30 mL polystyrene vials (Bibby, Sterilin) that were previ-ously spiked with copper to provide a concentration rangeof 0–900 nM added copper (actual range depending on the

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

T

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

235235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

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

4 J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

NC

OR

REC

initial copper concentration and the ligand concentrationdetermined), and left to equilibrate overnight at room tem-perature. Prior to the first titration, the tubes were condi-tioned twice with seawater containing the same range ofcopper concentrations. The equilibrium concentration ofcopper bound by SA was determined by CSV using anadsorption potential of �0.15 V and a deposition time of60 s in the same voltammetric equipment as the onedescribed above.

The complexation of copper by salycilaldoxime isaffected by the mayor ions present in seawater and, there-fore, depends on the salinity (S) of the sample. Values forthe stability constants of the complexes CuSA and Cu(SA)2

at each salinity were calculated using relationshipsobtained in a previous paper (Campos and van den Berg,1994):

log K 0CuSA ¼ ð10:12� 0:03Þ � ð0:37� 0:02Þ log S;

log B0CuðSAÞ2 ¼ ð15:78� 0:08Þ � ð0:53� 0:07Þ log S:

The complexing capacities and conditional stability con-stants of the natural ligands of the samples were obtainedfrom the linearizing plot [CuL]/[Culabile] vs. [Culabile], asdescribed earlier (Ruzic, 1982; Van den Berg, 1982). Withthe slope and Y-axis value of this plot, it is possible to esti-mate the complexing capacity and the conditional stabilityconstant of the copper ligands in the solution. Labile cop-per concentrations ([Culabile]) correspond to the concentra-tion of Cu(SA)2, and are obtained from the voltammetricanalysis and the sensitivity of the method. Organicallycomplexed copper is calculated by subtracting the labileconcentration from the total copper concentration in thealiquot. The sensitivity of the CSV measurements was ini-tially estimated from the linear part of the titration at highcopper concentrations (from a plot of the peak height as afunction of the copper concentration) and corrected forunderestimation using an iterative calculation similar tothe one used previously (Turoczy and Sherwood, 1997).

All the titration data sets showed curved distributions,which are indicative of the presence in the samples of atleast more than one type of ligand characterized by differ-ent stability constants. Copper complexing ligand concen-trations (CL1 and CL2) and values for conditionalstability constants (K 0CuL1 and K 0CuL2) on the basis of Cu2+

were calculated as before (Laglera and van den Berg,2003) using the iterative fitting of the data to two complex-ing ligands.

313

314

315

316

317

318

319

320

321

322

U2.3. Copper analysis in the SPM

Filters obtained from filtered water (12 samples) weredigested in a microwave oven (Milestone MLS-1200 Mega)using Teflon bombs containing 1.8 mL of nitric acid (65%,Panreac Hiperpur) and 0.6 mL of fluorhydric acid (48%,Merck Suprapur) according to the EPA 3052 guidelinemethod (EPA, 1996). Copper determination was carriedout by Electrothermal Atomic Absorption Spectroscopy

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

ED

PR

OO

F

(ETAAS) with Zeeman background correction (Varian220) and Pd–Mg(NO3)2 as the matrix modifier. The analyt-ical procedure was checked using the reference materialPACS-1 (harbour marine sediment), and the resultsshowed good agreement with the certified values: error per-centage that are always lower than 1.1% for Cu.

3. Results and discussion

3.1. Master variables and SPM

The results of the main variables measured (flow, salin-ity, temperature, pH, oxygen saturation and SPM) areshown in Table 1. River flows are in the range of 160–8000 L s�1 while sewage values vary between 10 and1900 L s�1. In terms of temperature, rivers are usually 2–3� colder than sewage effluents. In the case of salinity, witha few exceptions, the values are consistently below 1� ofsalinity, which indicates a freshwater input. Rivers andsewages have a very similar pH, ranging from 6.1 to 6.9.Oxygen saturation values are higher in rivers (89–99) thanin sewages (14–86). These values are common in the Gali-cian rias, as has been reported in previous papers (Cobel-o-Garcia et al., 2004; Gago et al., 2005).

The SPM levels are higher in sewages ranging from 6.7to 116.3 mg L�1 than in the case of the rivers where the val-ues change between 0.33 and 18.7 mg L�1. Similar valueshave been reported by other authors for this ria: valuesof 2–15 mg L�1 in the rivers flowing into the Vigo Ria(Pazos et al., 2000) and, in other Galician rias (such asthe Ferrol), values of 15–20 mg L�1 for rivers and 30–130 mg L�1 for untreated industrial and domestic effluents(Cobelo-Garcia et al., 2004) are obtained.

3.2. River inputs

3.2.1. Levels and speciation

Dissolved copper values vary between 3.3 and 36.6 nM(Table 2). The highest copper concentrations were foundin the Lagares and Alvedosa rivers, both on the southernmargin of the ria (Fig. 1), while the lowest values were mea-sured for the Fraga, Maior, Ullo (on the northern margin)and the Oitaven river – the main freshwater input source –located at the head of the ria. The low levels of copper inthe Oitaven river (4 nM) indicate the natural copper levelin an uncontaminated river in this region. Table 3 showssimilar values for other Galician rivers, and values on thesame order of magnitude as those for other rivers, bothin Europe and worldwide. In addition, with the exceptionof the Alvedosa and Lagares rivers, the levels are withinthe range of pristine river waters (Hart and Hines, 1995).

Ligand concentrations and conditional stability con-stants are summarized in Table 2. The data were fitted totwo types of ligands, L1 and L2. The concentrations ofthe strongest ligand (L1) vary between 6 and 29 nM (Table2). Values for L2, the weaker ligand, are in the same rangeas those corresponding to the strongest one (between 11

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

UN

CO

RR

EC

TED

PR

OO

F

323

324

325

326

327

328

329

330

Table 1Master variables for the continental inputs (river and sewages) reaching the Vigo Ria during sampling in December 2005

Station Flow (L s�1) Temperature (�C) Salinity pH % O2 SPM (mg L�1)

River

Fraga 162 11.5 0.6 6.56 90 6.5Maior 163 12.4 0.9 6.12 96 2.2Ullo 299 11.0 0.3 6.79 99 1.7Oitaven 7970 9.4 0.0 6.50 96 0.3Alvedosa 661 12.6 0.0 6.67 95 2.5Lagares 1490 10.6 0.0 6.70 89 18.7

Sewage

Cangas 43 14.8 0.0 6.86 72 6.7Moana 79 14.3 1.1 6.68 86 8.3Arcade 13 14.3 0.4 6.87 86 8.1Redondela 75 13.3 5.4 6.66 49 63.2Teis 80 13.1 0.6 6.45 14 116Vigo 1855 15.0 1.2 6.74 59 44.8

Table 2Dissolved, particulate and copper speciation variables measured for the rivers and sewages

Station Cupart. (nM) Cudiss. (nM) pCu (nM) L1 (nM) logK1 L2 (nM) logK2

River

Fraga 6.46 8.46 ± 1.2 15.18 16.6 ± 0.4 15.20 ± 0.1 15.27 ± 2.3 12.70 ± 0.3Maior 4.27 9.18 ± 0.2 14.42 6.3 ± 0.2 15.43 ± 0.2 18.13 ± 4.0 13.79 ± 0.4Ullo 0.62 3.31 ± 0.3 15.63 6.1 ± 0.6 15.59 ± 0.1 10.99 ± 3.3 14.25 ± 0.4Oitaven 0.99 4.01 ± 0.2 15.22 7.0 ± 0.3 15.24 ± 0.1 11.18 ± 1.3 13.80 ± 0.2Alvedosa 9.80 30.7 ± 1 13.94 28.3 ± 2.7 14.83 ± 0.1 19.55 ± 10.9 13.54 ± 0.5Lagares 29.40 36.6 ± 1 13.65 28.8 ± 1.5 14.99 ± 0.5 113 ± 1 12.59 ± 0.2

Sewage

Cangas 9.6 108 ± 3 15.54 120 ± 19 16.19 ± 0.6 372 ± 38 13.99 ± 0.4Moana 19.5 63.7 ± 1 14.87 85.9 ± 1.0 15.23 ± 0.1 1127 ± 44 12.41 ± 0.1Arcade 17.1 126 ± 6 15.73 225 ± 5 15.79 ± 0.1 5400 ± 202 12.77 ± 0.1Redondela 419 61.8 ± 1 14.52 74.2 ± 1.3 15.01 ± 0.1 659 ± 63 12.45 ± 0.1Teis 705 39.1 ± 2 15.90 170 ± 14 15.31 ± 0.1 1163 ± 25 13.22 ± 0.1Vigo 476 310 ± 19 13.05 227 ± 13 14.25 ± 0.2 356 ± 60 12.63 ± 0.4

Table 3Ranges of dissolved, particulate and copper speciation variables for Vigo, Galicia, Europe and world rivers

Copper

Dissolved (nM) Particulate (nM) Speciation

L1 (nM) logK1 L2 (nM) logK2 pCu

Ria Vigo riversa 3–37 0.6–30 7–25 15.0–15.6 10–18 12.7–14.8 9.6–15.6Galician riversb 0.6–33 5–20 850–1700 8.2–8.8f

European riversc 5–50 10–250 40–200 13.2–14.8 50–560 10.9–13.1 12.9–15.1World background valuesd 16 20–90Pristine river waterse 2–16

a This study.b Prego and Cobelo-Garcia (2003), Cobelo-Garcıa and Prego (2003), Antelo et al. (1998).c Loubet et al. (2003), Cleven et al. (2005), Laglera and van den Berg (2003), Pettine et al. (1996), Xue et al. (1996), Apte et al. (1990).d Bewers and Yeats (1977), Salomons and Forstner (1984), Bruland and Franks (1983), Kremling (1985).e Hart and Hines (1995).f ASV method.

J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx 5

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

and 20 nM) with the exception of the Lagares (113 nM).These L2 values are much lower than the ones reportedby Antelo et al., 1998 for two Galician rivers, and lowerthan in other European rivers (Table 3). Antelo et al.

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

(1998) employed a different method (ASV); therefore, theresults must be compared carefully, since the pool ofligands measured and the detection window used are notthe same (Van den Berg et al., 1990).

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

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

378

379

380

381

382

383

384

385

386

387

388

389

390

391

6 J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

Values for logK1 are similar for all the rivers, and fluc-tuate between 14.8 and 15.6 (Table 2). The stability con-stants for the second ligand (logK2) are more variable,with values ranging between 12.6 and 14.3. In the case ofGalician rivers, our values are higher than the onesreported by Antelo et al. (1998). However, as mentionedabove, a different technique was used. European river val-ues are similar to (or a little bit lower than) the onesreported in this study (Table 3).

The summed value of ligand concentrations (L1 + L2) inrivers is always higher than that of the dissolved copperconcentrations. As a result of this proportion and the highvalues of the stability constants, the levels of pCu (labilecopper) are very low (Table 2), and most of the copper inrivers (99%) is expected to be organically complexed.

Particulate copper concentrations were measured toshow the importance of this phase in comparison withthe dissolved one. Data are presented in Table 2. Particu-late copper levels vary between 0.6 and 29.4 nM. The high-est concentration was found in the Lagares River. Thelowest levels correspond to the Ullo river. These valuesare similar to the ones reported previously for other Gali-cian rias and are slightly lower than other European rivers,while they remain within the range of world backgroundvalues (Table 3).

T

392

393

394

395

396

397

398

399

400

401

402

403

404

REC

3.2.2. Fluxes: Copper and ligands

Copper fluxes were calculated to estimate their impacton the ria (Table 4). In the case of rivers, dissolved copperfluxes in the winter of 2005 varied between 0.1 and4.7 mol day�1. These values are consistent with the limitednumber of findings reported for one river in another ria(Prego and Cobelo-Garcia, 2003) and in a few, smallFrench estuaries, which have been the object of research(Waeles et al., 2005; Monbet, 2004). They are, however,negligible in comparison with the major systems discharg-ing to the NE Atlantic (Michel et al., 2000; Chiffoleau

UN

CO

R 405

406

407

408

409

410

411

412

413

414

415

416

417

418

419

420

421

Table 4Dissolved, particulate copper and ligand fluxes for the stations sampled inthe Vigo Ria

Station Particulate(mol day�1)

Dissolved(mol day�1)

L1

(mol day�1)L2

(mol day�1)

Rivers

Fraga 0.09 0.12 0.23 0.21Maior 0.06 0.13 0.09 0.25Ullo 0.02 0.08 0.16 0.28Oitaven 0.68 2.76 4.82 7.69Alvedosa 0.56 1.75 1.61 1.11Lagares 3.79 4.71 3.71 14.5

Sewages

Cangas 0.04 0.40 0.45 1.39Moana 0.13 0.43 0.58 7.65Arcade 0.02 0.14 0.25 6.05Redondela 2.73 0.40 0.48 4.28Teis 4.88 0.27 1.24 8.04Vigo 76.3 49.8 36.4 57.0

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

ED

PR

OO

F

et al., 1994; Cotte-Krief et al., 2000), due to the higherwater discharge of the large European rivers (see Table 5).

Copper is clearly complexed by the strongest ligand (L1)and, when necessary, the second ligand (L2) complexes theexcess free copper (Fig. 2). Thus, dissolved organicallycomplexed copper fluxes reach the ria waters. These arean important contribution of ligands to the Vigo Riawaters (Table 4).

Particulate copper fluxes range between 0.02 and3.79 mol day�1 (Table 4). Similar values have beenreported for the rivers reaching the Morlaix estuary (Mon-bet, 2004). However, from a global perspective, the fluxesreaching the Vigo Ria are negligible in comparison withthe main European estuaries, such as the Seine in France(Chiffoleau et al., 1994).

The relative importance of each river flux and the differ-ent fractions of copper – both dissolved and particulate –are given in Table 3 (labile fluxes have been omittedbecause of their low values). Fifty-six percent of the copperriverine inputs are associated with the Lagares River, fol-lowed by the Oitaven River (24%). It is interesting to notethat the most contaminated river (with regard to copperconcentrations) as well as the most pristine contribute sim-ilar copper fluxes to the ria, which indicates the importanceof concentrations and flows. Sixty-four percent of the cop-per discharged to the Vigo Ria from rivers is in the dis-solved fraction, while 36% is in the particulate fraction.

Considering all the variables measured (master vari-ables, copper and ligands), the rivers flowing into the VigoRia can be divided in two groups: (i) the most degradedtributaries are the Lagares, Alvedosa and Fraga rivers.These inputs present the lowest oxygen saturation percent-ages, the highest SPM levels, and the highest copper andligand concentrations. (ii) There is a second group thatconsists of the Oitaven, Ullo and Maior rivers, which doesnot seem to have been anthropogenically influenced. Highoxygen saturation percentages, low SPM levels, and lowcopper and ligand concentrations are the main characteris-tics of these inputs.

3.3. Sewage treatment plant inputs

3.3.1. Levels and speciation

Sewage dissolved copper levels ranged between 62 and310 nM. (Table 2). The two highest values were measuredin the Vigo and Arcade STP, which are located on thesouthern margin of the ria (Fig. 1). Another remarkablyhigh value was found in the Cangas STP (110 nM), whichis situated on the northern margin (Fig. 1). The rest of thesewages presented copper levels ranging between 40 and64 nM. These concentrations are similar to the onesreported for other untreated Galician sewages, and areon the low end of the worldwide sewage range (Table 4),which has only been the subject of a few studies.

In terms of ligand concentrations, L2, values were nor-mally one order of magnitude higher than L1 (Table 2).These values are on the low end of the range reported by

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

CTED

PR

OO

F422

423

424

425

426

427

428

429

430

431

432

433

434

435

436

437

438

439

440

441

442

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

Table 5Ranges of dissolved, particulate and copper speciation variables for Vigo, Galicia, European and world sewages

Copper

Dissolved (nM) Particulate (nM) Speciation

L1(nM) logK1 L2 (nM) logK2

Ria Vigo sewagesa 60–310 10–700 75–230 14.2–16.2 380–5400 12.4–14.0Galician sewagesb 40–350 70–550Worldwide sewagesc 22–1260 4–1100 30–1300 12.7–13.8 430–2500 9.2–10.4

a This study.b Cobelo-Garcıa and Prego (2003), Cobelo-Garcıa and Prego (2003).c Laxen and Harrison (1981), Goldstone et al. (1990), Sedlak et al. (1997), Van Veen et al. (2002), Sarathy and Allen (2005).

Station

Fraga Maior Ullo Oitavén Alvedosa Lagares Cangas Moaña Arcade Redondela Teis Vigo

log

L f

lux

(nm

ol d

ay-1

)

1e+7

1e+8

1e+9

1e+10

1e+11

L1

L2

Cu Complexed

Fig. 2. Ligand and organically complexed copper fluxes in the different continental inputs (rivers and sewages) to the Vigo Ria.

J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx 7

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

UN

CO

RR

Eother authors for L1 in worldwide sewages, while similarranges have been reported for L2 levels (Table 4).

In the case of sewages, the values of logK1 rangebetween 14.3 and 16.2. With regard to logK2, levels fluctu-ate between 12.4 and 14.0. These values are slightly higherthan the ones reported for other world sewage treatmentplant effluents considered in the few studies focusing onthese measurements (Table 4).

As is the case with rivers due to the high concentration ofligands and strong stability constants, the pCu values in thesewage outfalls are high (13.1–15.9). Thus, more than 99%of dissolved copper is organically complexed, which wouldsuggest that harmful effects from the copper would not beexpected from these complexed copper inputs (Van Veenet al., 2002). No pCu values have been previously reportedin the literature for sewage treatment plant effluents.

Particulate copper levels in sewages range between 10and 705 nM (Table 2). Similar values have been measuredin two other Galician sewages, as well as others found allover the world (Table 4).

3.3.2. Sewage fluxes of copper and ligandsThe fluxes of dissolved copper varied between 0.1 and

49.8 mol day�1 (Table 4). Lower copper fluxes (0.09–

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

8.1 mol day�1) were reported by Prego and Cobelo-Garcia(2003) for untreated sewages in the Pontevedra Ria and inother – although limited – data reported for different sew-age treatment plant effluents around the world (Goldstoneet al., 1990; Sedlak et al., 1997).

Sewage fluxes of the strongest ligand (L1) fluctuatebetween 0.3 and 36.4 mol day�1. The high L1 flux measuredat the Vigo STP (36.39 mol day�1) in comparison with therest of the fluxes, which are in the range of 0.25–1.24 mol day�1, is worth noting. The weakest ligand fluxes(L2) are in the 1.4–57 mol day�1 range. The most remark-able input of L2 is, once again, from the Vigo STP(57 mol day�1). Ligand fluxes (L1 + L2) are higher thandissolved copper; consequently, almost all dissolved copperreaches the estuary organically complexed. These ligandfluxes are consistent with what was reported by Sedlaket al. (1997) for sewage effluents discharging into SanFrancisco.

In the case of particulate copper fluxes, values varybetween 0.02 and 76.3 mol day�1. Vigo sewage is the mostimportant contributor of particulate copper to the ria,while the rest of the fluxes are below 5 mol day�1 similarto the values reported by other authors (Goldstone et al.,1990; Sedlak et al., 1997).

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

TED

PR

OO

F

469

470

471

472

473

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

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

LAGARES OITAVEN ALVEDOSA OTHERS0

10

20

30

40

50

60

RIVERS

% R

iver

cop

per

flux

VIGO OTHERS0

20

40

60

80

100

SEWAGES

Dissolved

Particulate

STP RIVERS0

20

40

60

80

100

CONTINENTAL INPUTS

% S

TP

copp

er f

lux

% T

otal

cop

per

flux

Fig. 3. Percentage of each copper flux (dissolved and particulate) forrivers, sewage treatment plants and both.

8 J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

UN

CO

RR

EC

Ninety-three percent of the copper sewage inputs areassociated with the Vigo STP, while the rest of the sewagesdischarge the remaining 7%. The predominant dischargephase is the particulate (63%), followed by the dissolvedphase (38%) (Fig. 3b).

Like the rivers, the sewages can also be divided into twogroups: (i) the Vigo and Teis STP present the lowest oxy-gen saturation percentages, the highest SPM levels, andconsiderable copper and ligand concentrations. (ii) The sec-ond group, consisting of Redondela, Arcade, Moana andCangas, has higher oxygen saturation percentages andlower levels of the other variables (SPM, copper andligands). These differences between the groups can be asso-ciated with the equivalent population size treated by eachplant and the efficiency of the different levels of treatment.

3.4. Comparison of river and sewage inputs

Dissolved copper concentrations in sewages (39–310 nM) are one order of magnitude higher than in rivers(3–37 nM). As a consequence, and considering the differentflows (Table 1), dissolved copper fluxes entering from sew-ages are higher than the fluxes entering from rivers. How-ever, except for the Vigo STP, sewage fluxes aredeceptive. In both cases, a high percentage of the dissolvedcopper is organically complexed: the values range fromaround 99.8% in the case of rivers to more than 99.9% insewages. No previous data on complexation percentageshave been reported for sewage outfalls, nor have any com-parative studies been carried out with other types of conti-nental inputs, such as rivers.

Copper complexing ligands, which are present in riversand sewages originate mainly from dissolved organic mat-ter (DOM), formed by a mixture of many compounds(Cabaniss and Shuman, 1988; Xue et al., 1996; Sarathyand Allen, 2005). In the case of rivers, the DOM is domi-nated by humic substances resulting from soil erosion,decomposition of vegetation, and phytoplanctonic exu-dates (Mantoura et al., 1978; Livens, 1991; Kogut andVoelker, 2001). In sewages, due to anthropogenic influence(which is usually greater than in rivers), extra sources oforganic matter have to be considered. In addition, syn-thetic complexing agents, biopolymers, surfactants, pro-teins, EDTA and NTA join the natural sources (humicsubstances and exudates) to enter sewage treatment plants(Stiff, 1971; Sedlak et al., 1997; Sarathy and Allen, 2005).

Concentrations of the strongest complexing ligand (L1)are again one order of magnitude higher in sewages thanin rivers. With regard to L2, a similar situation can beobserved, but here the difference is around two orders ofmagnitude. As a result, sewage ligand fluxes are higherthan river ligand fluxes. It must be pointed out that, withthe exception of the input from the Vigo STP, L1 fluxesare higher in rivers, while sewages continue to be the maininput in L2 fluxes.

The destination of a very high percentage of copper andligand fluxes is the southern margin of the Vigo Ria (99%

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

of the dissolved copper and 93% of the ligands), whichwas to be expected, considering the large population andindustrial activity along the southern margin.

The conditional stability constants are quite similarbetween rivers and sewages. The values of logK1 changebetween 14.8–15.6 in rivers and 14.3–16.2 in sewages.The levels of logK2 vary between 12.6–14.8 in riversand 12.4–14.0 in sewages. Log K1 values are, in general,higher in sewages, while logK2 levels are higher in rivers.

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

T

533

534

535

536

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

567

568

569

570

571

572

573

574

575

576

577

578

579

580

581

582

583

584

585

586

587 Q1

588

589

590

591

592

593

594

595

596

597

598

599

600

601

602603604605606607608609610611612613614615616617618619620621622623624625626627628629630631632633634635636637638639640641642643644645646647648649

J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx 9

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

UN

CO

RR

EC

These minor differences could be associated with thediverse nature of the complexing agents present in thewaters. A comparison of a sewage-extracted FA samplewith a natural soil sample reveals very few differences inthe binding sites, which would explain the relatively minordifferences in the stability of the complexes (daSilva andOliveira, 2002). A similar situation would occur withhumic acids, since they are of like origin; thus, the extraorganic complexes from anthropogenic sources wouldexplain these discrepancies.

Despite the importance of the speciation controllingcopper biogeochemical cycles and metal bioavailabilityin the receiving waters, little attention has been paid tocopper speciation parameters and ligand fluxes reachingestuaries or coastal areas (Kogut and Voelker, 2001)from the continent. Only one paper makes a clear refer-ence to these types of inputs (Shank et al., 2004). Hence,the data presented in this study (Tables 2–4) could be avery useful baseline for future research on metal specia-tion in coastal areas. The high complexation capacitiesobserved in the waters entering the Vigo Ria (Table 2and Fig. 2) are very important in terms of toxicity andbioavailability because the excess ligands could act ascomplexing agents when metal inputs of a punctual nat-ure occur.

Overall, the relative importance of river versus sewagefluxes and the phases (dissolved and particulate) of eachflux are shown in Fig. 3c. Ninety-one percent of the cop-per inputs reach the Vigo Ria through sewages while theriver contribution of copper to the ria is only 9% of thetotal. Most of the copper discharged from sewages is inthe particulate phase while, in the case of rivers, mostof it reaches the ria in the dissolved phase. The importantrole of sewage as a source of copper to the ria increasesduring the dry season due to diminishing fluvial flows.No similar research has been carried out along these linesfor estuaries.

The total copper fluxes (rivers and sewages) into theVigo Ria during early winter conditions were calculated.The result obtained, 9.54 kg day�1, is very similar to thefindings of Cobelo-Garcia et al. (2004) for a small, butindustrialized Northern ria (Ferrol) where water treatmentdoes not exist. Due to the effect of sewages, copper dis-charges are 5 times higher, L1 is 3 times higher, and L2 is5 times greater as compared to an idyllic situation consist-ing of only natural inputs. With regard to particulate cop-per fluxes, sewage discharges are 15 times greater than innatural inputs. These ‘‘enrichment factors’’ with respectto rivers could have important effects on the waters andsediments of the Vigo Ria.

One effect has been observed with ligands: L1 fluxes aremultiplied by a factor of 3 and L2 fluxes by a factor of 5 ascompared to natural values.

4. Uncited references

Brand et al. (1986).

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

ED

PR

OO

F

Acknowledgements

The authors would like to thank Unai Otxotorena forhelping with the sampling, Clemente Trujillo for carryingout the SPM copper analyses, and Paula Ferro for hertechnical assistance. The visit of J. Santos-Echeandia wasfinancially supported by a EU Marie Curie Training Fel-lowship (contract No HPMT-CT-2001-00218) to the Mar-ine Electrochemistry Group of the Department of Earthand Ocean Sciences (University of Liverpool). This workis a contribution to the Spanish LOICZ program andwas supported by CICYT under the project ‘‘Biogeochem-ical Budget and Fluxes of Metals in a Ria (REN2003-04106-C03)’’.

References

Allen, H.E., Hansen, D.J., 1996. The importance of trace metalspeciation to water quality criteria. Water Environment Research 68(1), 42–54.

Antelo, J.M., Arce, F., Penedo, F.J., 1998. Effect of seasonal changes onthe complexing of Cu(II) by dissolved organic matter in river water.Water Research 32 (9), 2714–2720.

Apte, S.C., Gardner, M.J., Gunn, A.M., Ravenscroft, J.E., Vale, J., 1990.Trace metals in the Severn estuary. A Reappraisal. Marine PollutionBulletin 21 (8), 393–396.

Bewers, J.M., Yeats, P.A., 1977. Oceanic resident times of trace elements.Nature 268, 595–598.

Brand, L.E., Sunda, W.G., Guillard, R.R.L., 1986. Reduction of marinephytoplankton reproduction rates by copper and cadmium. Journal ofExperimental Marine Biology and Ecology 96, 225–250.

Bruland, K.W., Franks, R.P., 1983. Mn, Ni, Cu, Zn and Cd in the westernNorthAtlantic. In: Trace metals in Seawater. NATO Conf. Ser. 4:Mar. Sci. V. 9. Plenum Press, New York, pp. 395–414.

Buykx, S.E.J., Cleven, R.F.M.J., Hoegee-Wehmann, A.A., van den Hoop,M.A.G.T., 1999. Trace metal speciation in European River waters.Fresenius’ Journal of Analytical Chemistry 363 (5–6), 599–602.

Cabaniss, S.E., Shuman, M.S., 1988. Copper-binding by dissolvedorganic-matter. 1. Suwannee River fulvic-acid equilibria. Geochimicaet Cosmochimica Acta 52 (1), 185–193.

Campos, M.L.A.M., van den Berg, C.M.G., 1994. Determination ofcopper complexation in sea water by cathodic stripping voltammetryand ligand competition with salicylaldoxime. Analytica Chimica Acta284, 481–496.

Chiffoleau, J.F., Cossa, D., Auger, D., Truquet, I., 1994. Trace metaldistribution, partition and fluxes in the Seine estuary (France) in lowdischarge regime. Marine Chemistry 47 (2), 145–158.

Clark, R.B., 2001. Marine Pollution. Oxford University Press, Oxford.Cleven, R., Nur, Y., Krystek, P., Van den Berg, G., 2005. Monitoring

metal speciation in the rivers meuse and Rhine using DGT. Water, Airand Soil Pollution 165 (1–4), 249–263.

Cobelo-Garcıa, A., Prego, R., 2003. Land inputs, behaviour andcontamination levels of copper in a ria estuary (NW Spain). MarineEnvironmental Research 56 (3), 403–422.

Cobelo-Garcia, A., Prego, R., Labandeira, A., 2004. Land inputs of tracemetals, major elements, particulate organic carbon and suspendedsolids to an industrial coastal bay of the NE Atlantic. Water Research38 (7), 1753–1764.

Cotte-Krief, M.H., Guieu, C., Thomas, A.J., Martin, J.M., 2000. Sourcesof Cd, Cu, Ni and Zn in Portuguese coastal waters. Marine Chemistry71 (3–4), 199–214.

da Silva, J.C.G.E., Oliveira, C.J.S., 2002. Metal ion complexationproperties of fulvic acids extracted from composted sewage sludgeas compared to a soil fulvic acid. Water Research 36 (13), 3404–3409.

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.

T

650651652653654655656657658659660661662663664665666667668669670671672673674675676677678679680681682683684685686687688689690691692693694695696697698699700701702703704705706707708

709710711712713714715716717718719720721722723724725726727728729730731732733734735736737738739740741742743744745746747748749750751752753754755756757758759760761762763764765766767

10 J. Santos-Echeandia et al. / Marine Pollution Bulletin xxx (2007) xxx–xxx

MPB 3573 No. of Pages 10, Model 5+

14 November 2007 Disk UsedARTICLE IN PRESS

UN

CO

RR

EC

EPA, 1996. Method 3052, Microwave assisted acid digestion of siliceousand organically based matrices. Available from <http://www.epa.gov/epaoswer/test/3052.pdf/>.

Evans, G., Prego, R., 2003. Rias, estuaries and incised valleys: is a ria anestuary? Marine Geology 196 (3–4), 171–175.

Gago, J., Alvarez-Salgado, X.A., Nieto-Cid, M., Brea, S., Piedracoba,S., 2005. Continental inputs of C, N, P and Si species to the Ria deVigo (NW Spain). Estuarine, Coastal and Shelf Science 65 (1–2),74–82.

Goldstone, M.E., Kirk, P.W.W., Lester, J.N., 1990. The behaviour ofheavy metals during wastewater treatment I. Cadmium, chromium andcopper. The Science of The Total Environment 95, 233–252.

Hart, B.T., Hines, T., 1995. Trace elements in rivers. In: Salbu, B.,Steinnes, E. (Eds.), Trace Elements in Natural Waters. CRC Press,Boca Raton, FL, pp. 203–221.

Helz, G.R., Huggett, R.J., Hill, J.M., 1975. Behavior of Mn, Fe, Cu, Zn,Cd and Pb discharged from a wastewater treatment plant into anestuarine environment. Water Research 9 (7), 631–636.

Huh, C.A., 1996. Fluxes and budgets of anthropogenic metals in the SantaMonica and San Pedro Basins off Los Angeles: review and reassess-ment. Sciences of the Total Environment 179 (1–3), 47–60.

Kogut, M.B., Voelker, B.M., 2001. Strong copper-binding behaviour ofterrestrial humic substances in seawater. Environmental Science &Technology 35 (6), 1149–1156.

Kremling, K., 1985. The distribution of cadmium, copper, nickel,manganese and aluminium in surface waters of the open Atlanticand European Shelf area. Deep-Sea Research Part A. OceanographicResearch Papers 32 (5), 531–555.

Laglera, L.M., van den Berg, C.M.G., 2003. Copper complexation by thiolcompounds in estuarine waters. Marine Chemistry 82 (1–2), 71–89.

Laxen, D.P.H., Harrison, R.M., 1981. The physicochemical speciation ofCd, Pb, Cu, Fe and Mn in the final effluent of a sewage treatmentworks and its impact on speciation in the receiving river. WaterResearch 15 (9), 1053–1065.

Lester, J.N., Harrison, R.M., Perry, R., 1979. Balance of heavy metalsthrough a sewage treatment works. 1. Lead, cadmium and copper.Science of the Total Environment 12 (1), 13–23.

Livens, F.R., 1991. Chemical-reactions of metals with humic material.Environmental Pollution 70 (3), 183–208.

Loubet, M., Baque, D., Priscia, O., Dupre, B., 2003. Metal content inGaronne basin rivers: evidences for a general non point contaminationof anthropogenic origin. Journal de Physique PHYSIQUE IV 107,793–796 (Part 2).

Mantoura, R.F.C., Dickson, A., Riley, J.P., 1978. Complexation of metalswith humic material in natural-waters. Estuarine and Coastal MarineScience 6 (4), 387–408.

Michel, P., Boutier, B., Chiffoleau, J.F., 2000. Net fluxes of dissolvedarsenic, cadmium, copper, zinc, nitrogen and phosphorus from theGironde Estuary (France): seasonal variations and trends. EstuarineCoastal and Shelf Science 51 (4), 462–541.

Monbet, P., 2004. Dissolved and particulate fluxes of copper through theMorlaix river estuary (Brittany, France): mass balance in a smallestuary with strong agricultural catchment. Marine Pollution Bulletin48 (1–2), 78–86.

Pazos, O., Nombela, M.A., Vilas, F., 2000. Continental contribution ofsuspended sediment to an estuary: Ria de Vigo. Scientia Marina 64 (3),295–302.

Perez-Alberti, A., 1982. Xeografıa de Galicia. Instituto Geologico yMinero de Espana, Madrid.

768

Please cite this article in press as: Santos-Echeandia, J. et al., CopperBull. (2007), doi:10.1016/j.marpolbul.2007.10.021

ED

PR

OO

F

Perez-Arlucea, M., Mendez, G., Clemente, F., Nombela, M., Rubio, B.,Filgueira, M., 2005. Hydrology, sediment yield, erosion and sedimen-tation rates in the estuarine environment of the Ria de Vigo, Galicia,Spain. Journal of Marine Systems 54 (1–4), 209–226.

Pettersson, C., Allard, B., Boren, H., 1997. River discharge of humicsubstances and humic-bound metals to the Gulf of Bothnia. Estuarine,Coastal and Shelf Sciences 44 (5), 533–541.

Pettine, M., Bianchi, M., Martinotti, W., Muntau, H., Renoldi, M.,Tartari, G., 1996. Contribution of the Lambro River to the totalpollutant transport in the Po watershed (Italy). Science of the TotalEnvironment 192 (3), 275–297.

Prego, R., Cobelo-Garcia, A., 2003. Twentieth century overview of heavymetals in the Galician Rias (NW Iberian Peninsula). EnvironmentalPollution 121 (3), 425–452.

Ridgway, J., Breward, N., Langston, W.J., Lister, R., Rees, J.G., Rowlatt,S.M., 2003. Distinguishing between natural and anthropogenic sourcesof metals entering the Irish Sea. Applied Geochemistry 18 (2), 283–309.

Ruzic, I., 1982. Theoretical aspects of the direct titration of natural watersand its information yield for trace metal speciation. Analytica ChimicaActa 140, 99–113.

Sarathy, V., Allen, H.E., 2005. Copper complexation by dissolved organicmatter from surface water and wastewater effluent. Ecotoxicology andEnvironmental Safety 61 (3), 337–344.

Salomons, W., Forstner, U., 1984. Metals in the Hydrocycle. Springer,Berlin.

Sedlak, D.L., Phinney, J.T., Bedsworth, W.W., 1997. Strongly complexedCu and Ni in wastewater effluents and surface runoff. EnvironmentalScience & Technology 31 (10), 3010–3016.

Shank, G.C., Skrabal, S.A., Whitehead, R.F., Avery, G.B., Kieber, R.J.,2004. River discharge of strong Cu-complexing ligands to southAtlantic bight waters. Marine Chemistry 88 (1–2), 41–51.

Smaal, A.C., 2002. European mussel cultivation along the Atlantic coast:production status, problems and perspectives. Hydrobiologia 484 (1–3), 89–98.

Sterrit, R.M., Lester, J.N., 1984. Significance and behaviour of heavymetals in wastewater treatment processes. 3. Speciation in wastewatersand related complex matrices. Science of the Total Environment 34 (1–2), 117–141.

Stiff, M.J., 1971. Chemical states of copper in polluted fresh water and ascheme of analysis to differentiate them. Water Research 5 (8), 585–599.

Turoczy, N.J., Sherwood, J.E., 1997. Modification of the van den Berg/Ruzic method for the investigation of complexation parameters ofnatural waters. Analytica Chimica Acta 354 (1–3), 15–21.

Van den Berg, C.M.G., 1982. Determination of copper complexation withnatural organic ligands in seawater by equilibration with MnO2 I.Theory. Marine Chemistry 11 (4), 307–322.

Van den Berg, C.M.G., Nimmo, M., Daly, P., Turner, D.R., 1990. Effectsof the detection window on the determination of organic copperspeciation in estuarine waters. Analytica Chimica Acta 232 (1), 149–159.

Van Veen, E., Burton, N., Comber, S., Gardner, M., 2002. Speciation ofcopper in sewage effluents and its toxicity to Daphnia magna.Environmental Toxicology and Chemistry 21 (2), 275–280.

Waeles, M., Riso, R.D., Le Corre, P., 2005. Seasonal variations ofcadmium speciation in the Penze estuary, NW France. Estuarine,Coastal and Shelf Science 65 (1–2), 143–152.

Xue, H.B., Oestreich, A., Kistler, D., Sigg, L., 1996. Free cupric ionconcentrations and Cu complexation in selected Swiss lakes and rivers.Aquatic Sciences 58 (1), 69–87.

speciation in continental inputs to the Vigo Ria: ..., Mar. Pollut.