Trace elements in biodeposits and sediments from mussel culture in the Ría de Arousa (Galicia, NW Spain)

Environmental Pollution 136 (2005) 119e134 www.elsevier.com/locate/envpol

Trace elements in biodeposits and sediments from mussel culture in the Rı´ a de Arousa (Galicia, NW Spain) X.L. Oteroa,*, P. Vidal-Torradob, R.M. Calvo de Antac, F. Macı´ asc a

Departamento de Edafoloxı´a e Quı´mica Agrı´cola, Escola Polite´cnica Superior de Lugo, Universidade de Santiago de Compostela, 27002 Lugo, Spain b Departamento de Solos e Nutric¸a˜o de Plantas, Escola Superior de Agronomı´a Luiz Queiroz, Universidade de Sa˜o Paulo, Piracicaba, Brazil c Departamento de Edafoloxı´a e Quı´mica Agrı´cola, Facultade de Bioloxı´a, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Spain Received 28 May 2004; accepted 22 November 2004

Geochemical behaviour of trace metals in mussel biodeposits and sediments were studied. Abstract The trace elements present at highest concentrations were Cr and Zn, which probably originated from the dumping of effluent from a tanning factory. High proportions of these two elements were associated with the residual fraction. Biodeposits and sediments showed high concentrations of Cd and Pb in the reactive fraction, with a high proportion of the concentration in the reactive fraction being associated with carbonates. Nickel showed a higher degree of pyritization than the previous elements, although most of the Ni was associated with the residual and reactive fractions. Arsenic, Hg and Cu showed high degrees of pyritization, particularly below a depth of 5 cm. The results demonstrate that those elements with a high degree of pyritization may be released into the water through oxidation of the metal sulphides that they form when in suspension in oxic sea water, with the subsequent risk of increased bioavailability to benthic fauna. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Aquaculture; Mussel; Biodeposits; Sediments; Trace elements; Pyritization; Rı´ a; Galicia

1. Introduction Aquaculture has undergone great development in the last 20 years, currently accounting for around 10% of the world’s total fish production, and it is expected that the industry will become more important because of the gradual depletion of natural fish stocks (Barg, 1995). In Galicia, despite increasing development of the cultivation of different fish species (turbot, salmon, bass, etc.), Abbreviations: AVS, Acid Volatile Sulfide; DOP, degree of pyritization of Fe; DTMP, degree of trace metal pyritization. * Corresponding author. Fax: C34 981 596904. E-mail address: [email protected] (X.L. Otero). 0269-7491/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2004.11.026

cultivation of the mussel (Mytilus galloprovincialis) is the most important activity, with an annual production of 190 000 t, which represents 40% of mussel production in Europe and 95% of aquaculture production in Galicia (Xunta de Galicia, 2001). Mussels are cultivated using floating platforms (mussel rafts, known locally as bateas) of approximately 500 m2, from which some 500 ropes are hung and to which the mussels attach, in numbers ranging from 600 to 1000 individuals per m of rope (Macı´ as and Mora, 2001). The total number of rafts installed in Galicia is 3242 (Xunta de Galicia, 2001). Despite the huge economic benefits that this activity generates in Galicia (78 million euros and some 11 500 job positions), concentration of a large number of rafts in

120

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

a relatively small space and the associated handling lead to a significant impact on the prevailing marine environmental conditions. Amongst the effects described in previous studies are an increase in sedimentation due to the sweeping effect of the ropes and to the metabolic activity of the mussels themselves, which can produce between 129e190 kg of faecal material (dry weight) day
1 raft
1 (Cabanas et al., 1979; Collazo et al., 1993), which represents an annual input of some 69.3 t of sediment per raft (Cabanas et al., 1979). This produces changes in benthic metabolism, transforming the sea floor into an anoxic environment where mineralization of organic matter occurs by sulphate reduction and methanogenesis (Macı´ as and Mora, 2001). In addition, it must be taken into account that mussels are powerful filter feeders and each individual filters around 15 m3 of water every year. This means that they can accumulate toxic substances dissolved in sea water or that is associated with material in suspension, and because of this they are used as bioindicators to monitor marine contamination (see e.g. Besada et al., 2002). Some studies have shown significant amounts of heavy metals in the soft tissues of mussels (Otero and Ferna´ndez-Sanjurjo, 2000; Besada et al., 2002) and in their shells (Puente et al., 1996). These metals become incorporated into the sediment after release from the mussels themselves or through the faecal matter that they produce. In the specific case of the Rı´ a de Arousa, a study has been carried out of the heavy metal content of mussel faeces in relation to mussel size, (Collazo and Pascual, 1997). From the results obtained by these authors, taking into consideration the heavy metal contents of medium-sized mussels, we estimate average inputs per raft of: 1342 g Cu year
1, 2495 g Pb year
1, 6449 g Cr year
1 and 979 kg of Fe year
1. Despite the importance of the previous results, few studies of the heavy metal contents of biodeposits and sediments underneath mussels rafts have been carried out to date. Moreover, recent studies have highlighted a serious lack of information about the content and geochemical behaviour of trace metals in sediments in Galician Rı´ as, particularly of Ag, Se, As and Hg (Prego and Cobelo-Garcı´ a, 2003). The aims of the present study were to determine the quality of the sediments in the Rı´ a de Arousa, a natural ecosystem of great ecological and productive interest, and to investigate the geochemical behaviour of eight trace elements (Hg, As, Pb Cu, Zn, Pb, Cd and Ni).

2. Materials and methods 2.1. Study area Galicia is one of the Spanish Autonomous communities with the longest coastline (w1700 km), which

from a geomorphological point of view, is one of the most complex littoral environments of the Iberian Peninsula, where the Rı´ as are the most unusual features (Fig. 1). The origin of the different Rı´ as is complex and has been the subject of many studies since they were first described in by von Richthofen (1886) (for more detail see VidalRomani, 1984). It is generally accepted that the Galician Rı´ as are very special geological formations that exist in few parts of the world (Ireland, China, Great Britain). The Rı´ as were formed in the Tertiary era as a consequence of the reactivation of ancient hercinic faults, giving rise to tectonic sunken valleys that were later invaded by the sea (Torre, 1958). From an economic point of view the Rı´ as constitute a highly productive ecosystem due to coastal upwelling of deep, nutrient rich water, which generally occurs between May and September. The Rı´ a de Arousa is the largest of the Galician Rı´ as, comprising an area of 230 km2; it is also the Rı´ a with the highest concentration of mussel rafts (2332), which represents more than 70% of the total number of rafts in Galicia. Geologically it is fundamentally comprised of plutonic (granites and granodiorites) and metamorphic rocks (mainly gneisses) (Fig. 1). The drainage system basically consists of the rivers Ulla and Umia. The river Ulla flows into the bottom of the Rı´ a; the corresponding drainage area is 2804 km2 and the absolute flow, 79.3 m3 s
1. The river Umia flows into the Rı´ a near the town of Cambados on the southern coast, with a drainage area of 440 km2 and an absolute flow of 16.3 m3 s
1 (Rı´ o Barja and Rodrı´ guez, 1992).

2.2. Sample collection A total of 44 samples collected from different zones of the Rı´ a de Arousa were analysed in the present study. The sampling locations and types of corer used are shown in Table 1. Sampling was carried out in September 1999 and February 2000. For this, the Rı´ a was divided into three zones in terms of bathymetric characteristics and oceanographic processes. The inner zone corresponds to the narrowest part of the Rı´ a, characterized by an average depth of 10e15 m and a predominantly estuarine environment due to the mixing of fresh and salt water. The middle and outer areas are zones of water exchange between the Rı´ a and the Atlantic Ocean. Circulation takes place in two layers with a surface outgoing current and a deeper, inflowing current (Fraga and Margalef, 1979). The age of the rafts also differs in each sector. The oldest are those situated in the inner and mid zones, which were installed more than 20 ago, whereas the rafts in the outer zone were installed between 5 and 15 years ago (Macı´ as and Mora, 2001).

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

121

Fig. 1. Map showing location, geology and sampling sites of the study area.

2.3. Analysis of the biodeposits and sediments Separation of the surface layer corresponding to the biodeposits from the deeper sediment layers was carried out by visual inspection, by X-ray analysis of the sediment, and using the abundance of mussel shell fragments as an indicator of the biodeposits. Samples of mussel biodeposits and of sediments were cut into 2 and 5 cm segments immediately after their extraction and were maintained frozen at
18  C until their analysis.

The pH and redox potential (Eh) of the surface layer (5 cm) were determined at the time of sampling. The pH was measured with a combination pH electrode and the Eh with a platinum combination electrode, the Eh values being recorded when the variation was less than 2 mV min
1. The values were corrected to the standard hydrogen electrode by adding 244 mV to the measured values. In the laboratory the pH of the dried samples corresponding to different depths was also measured using a ratio of sample to water of 1:2.5. The contents of total organic carbon (TOC) and carbonates (TIC) were

122

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

Table 1 Location of the sampling points, type of corer use and depth of overlying water Code

Site

GCAR1 Inner GCAR2 GCAR3 BCAR1 BCAR2 BCAR3 RA3-1 RA3-2 RA3-3 RA3-4 RA3-5 RA3-6 RA3-7 RA3-8 RA3-9 RA3-10

Coordinates

Type of corer

42  34.470#N, 8  50.325#W Gravity corer 42  36.214#N, 8  53.413#W Gravity corer 42  36.389#N, 8  54.835#W Gravity corer 42  34.455#N, 8  50.516#W Box corer 42  36.204#N, 8  53.440#W Box corer 42  36.377#N, 8  54.844#W Box corer 42  35.890#N, 8  52.807#W Box corer 42  35.284#N, 8  52.992#W Box corer 42  35.912#N, 8  53.092#W Box corer 42  35.927#N, 8  53.442#W Box corer 42  36.043#N, 8  53.329#W Box corer 42  36.125#N, 8  53.155#W Box corer 42  36.201#N, 8  53.448#W Box corer 42  36.380#N, 8  53.306#W Box corer 42  36.473#N, 8  53.493#W Box corer 42  36.198#N, 8  53.851#W Box corer

GCAR4 Middle 42  33.468#N, 8  54.517#W Gravity corer GCAR5 42  34.624#N, 8  55.920#W Gravity corer BCAR4 42  33.436#N, 8  54.542#W Box corer BCAR5 42  34.620#N, 8  55.928#W Box corer RA2-1 42  34.082#N, 8  53.991#W Box corer RA2-2 42  33.819#N, 8  54.269#W Box corer RA2-3 42  33.657#N, 8  53.930#W Box corer RA2-4 42  33.461#N, 8  53.568#W Box corer RA2-5 42  33.158#N, 8  53.237#W Box corer RA2-6 42  33.885#N, 8  54.027#W Box corer RA2-7 42  33.138#N, 8  54.116#W Box corer RA2-8 42  33.099#N, 8  54.116#W Box corer RA2-9 42  33.418#N, 8  54.532#W Box corer RA2-10 42  33.607#N, 8  54.651#W Box corer GCAR6 Outer GCAR7 BCAR6 BCAR7 RA1-1 RA1-2 RA1-3 RA1-4 RA1-5 RA1-6 RA1-7 RA1-8 RA1-9 RA1-10

42  33.215#N, 8  57.531#W Gravity corer 42  29.940#N, 8  55.079#W Gravity corer 42  33.240#N, 8  57.537#W Box corer 42  29.953#N, 8  55.073#W Box corer 42  33.356#N, 8  57.559#W Box corer 42  33.200#N, 8  57.460#W Box corer 42  33.093#N, 8  57.525#W Box corer 42  33.005#N, 8  57.465#W Box corer 42  33.008#N, 8  57.438#W Box corer 42  32.918#N, 8  57.557#W Box corer 42  32.996#N, 8  57.657#W Box corer 42  33.217#N, 8  57.680#W Box corer 42  33.276#N, 8  57.726#W Box corer 42  33.115#N, 8  57.959#W Box corer

Depth of water (m) 14 17 16 14 17 16 15 26 26 26 23 22 23 21 21 23 35 15 35 15 20 38 31 25 19 25 28 43 37 43 36 35 36 35 34 35 36 36 36 36 35 34 32 23

measured using a Leco CNH-1000 analyser, whereas the total S content was measured with a Leco 100 S-C 144DR analyser. Particle size was determined using the pipette method, after Gee and Bauder (1986). The concentration of acid volatile sulphides (AVS) was

determined using 0.50 to 1.0 g of wet sample, according to the method described by Kostka and Luther (1994). Sulphide was released from AVS with 20 ml 1 N HCl, previously deaerated for 40e50 min. The sample was digested in a gas-tight reaction flask during 40e50 min under a continuous flow of nitrogen, which was bubbled through the flask at the slowest possible rate. The evolved H2S was then received in a trap which contained 25 ml of 3% Zn acetate, 1 ml of concentrated H2SO4 and 4 ml of diamine reagent, and precipitated as ZnS. Sulphide was then measured colorimetrically with a UV-VIS spectrophotometer (Vitatron model MCP) at a wavelength of 670 nm, using the methylene blue method of Cline (1969). Total trace metals and Fe, Mn, and Al were extracted from the samples by adding 8 ml of a mixture of HNO3/ HF (3:5 v/v) in a 120-ml Teflon bomb containing 0.5 g of previously dried, ground sediment, and heating the mixture in an Ethos Plus microwave lab station. The efficiency of the extraction process (O90%) was determined using the certified reference marine sediment MESS-1. Enrichment factors (EF) were calculated by dividing the ratio [total metal(ppm)]/[total Al(%)] in the samples, by the same ratio calculated for the deepest section of the sediment core. Trace metals e Fe and Mn e associated with the operationally-defined reactive phase, as well as trace metals associated with the pyrite phase were determined using the sequential extraction method of Huerta-Dı´ az and Morse (1990). Briefly, the method consists of the extraction of four operationally-defined fractions: (1) reactive, essentially consisting of metals associated with carbonates and Fe and Mn oxyhydroxides, extracted with 20 ml of 1 N HCl during 16 h of continuous shaking; (2) silicate, essentially consisting of metals associated with clays and other aluminosilicates, extracted with 30 ml of 10 M HF during 16 h of continuous shaking; (3) organic, essentially consisting of metals associated with organic matter, extracted with 10 ml of concentrated H2SO4 during 2 h of continuous shaking; and (4) pyrite, essentially consisting of metals associated with pyrite, extracted with 10 ml of concentrated HNO3 during 2 h of continuous shaking (for further details on the method see Huerta-Dı´ az and Morse, 1990, 1992). In the present study, the residual fraction was obtained as the difference between the total content minus the metal associated with the reactive and pyritic fractions. The degree of Fe pyritization (DOP), a term proposed by Berner (1970) was calculated in order to establish the percentage of reactive iron incorporated into the pyrite fraction. It was calculated using eq. (1) 

 Pyrite-Fe DOPð%ÞZ !100 Pyrite-FeCReactive-Fe

ð1Þ

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

The degree of trace metal pyritization (DTMP), a term proposed by Huerta-Dı´ az and Morse (1990) and which gives an estimation of the amount of a particular metal (Me) incorporated into the pyrite phase, was calculated using the following equation:   Pyrite-Me DTMPð%ÞZ !100 ð2Þ Pyrite-MeCReactive-Me In addition, the exchangeable and carbonate fractions of Fe, Mn, Cd and Pb were determined using the method of Tessier et al. (1979). The first fraction was extracted by digesting 3.0 g of fresh biodeposit/sediment in 1 M MgCl2 at pH 7 for 30 min, with continuous shaking. The metals associated with the carbonates were then extracted after digestion in a solution of 1 M sodium acetate, adjusted to pH 5 with acetic acid, for 5 h with continuous shaking. The difference between the sum of these two fractions and the fraction extracted with 1 N HCl is assumed to be the fraction of sedimentary metal in the form of oxyhydroxide or associated with iron oxyhydroxides (Otero et al., 2003). Mobilization of trace metals by oxidation of metal sulphides was evaluated by maintaining 80 g of biodeposit and fresh sediment in suspension in 400 ml of water collected from the Rı´ a de Arousa and filtered (0.45 mm, Millipore), for a total of 288 h. In the final stage, after 264 h, 40 ml of 15% hydrogen peroxide, pH 5.5 was added. During the experiment, the pH and Eh of the suspension were measured and aliquots were extracted for analysis of Fe, Mn, Cu, As and Cr in solution. 2.4. General analytical procedures Laboratory and field equipment used during collection and analysis of samples was soaked in a solution of 5% HCl overnight and then rinsed twice with distilled water and Milli-Q water. The concentrations were always expressed in terms of dry weight of soil (dw), and the moisture content was determined from two subsamples dried at 105  C to constant weight. The concentrations of Al, Fe and Mn were determined by flame atomic absorption spectrophotometry (AAS) (Perkin-Elmer model 1100B) and those of Cr, Ni, Cu, Zn, As, Cd and Pb were determined in a graphite furnace AAS (Perkin Elmer 4110ZL). The concentration of Hg was measured in a Leco-Altec Ama-254 analyser. Blanks, standards and the different solutions used in the extractions were prepared with analytical grade reagents and Milli-Q water. 2.5. Data analysis Normality was confirmed by the Kolmogorov-Smirnov test. Those variables that were not normally

123

distributed were transformed to natural logarithms, except for the percentage data for total C, N, and S contents, which were subjected to an arcsine transformation, more appropriate to this kind of data (Zar, 1996). Differences in pH and Eh in the sediments and the biodeposits were established using ANOVA, followed by a Tukey test. Data on C, carbonates and heavy metals in biodeposits and sediments were analysed using a two-way ANOVA with the fixed factor being ‘type of material’ (biodeposit or sediment) and the variable factor ‘zone of Rı´ a’ (inner, middle or outer).

3. Results and discussion 3.1. Physicochemical characterization of the biodeposits and sediments Significantly different contents of sand, silt and clay were found in the biodeposits and the sediments from the three zones of the Rı´ a under study. Only the clay contents of the samples from the middle zone did not show any significant differences (Table 2). The silt fraction predominated in the mussel biodeposits (42e 52%), followed by clay (32e35%) and sand (12e25%), whereas the underlying sediments had a coarser texture, with predominance of a sandy fraction, which was between 2 and 4 times more prevalent than in the biodeposits. The finer texture of the biodeposits is consistent with the texture of the mussel faeces and pseudofaeces, which are known to be clayey or silty (Belzunce-Segarra et al., 1997; Macı´ as and Mora, 2001). It must also be taken into account that the installation of the rafts itself generates an important sweeping effect that favours sedimentation of fine particles (Barg, 1995). The pH values measured in situ in the surface layer of the sediment (0e5 cm) were similar in the three zones of the Rı´ a, and ranged from 7.4 to 8 (Fig. 2). The pH values measured in the samples in the laboratory were higher than those obtained for the field samples. In contrast, there were no significant differences between the pH of the biodeposit and the sediment in any of the three zones of the Rı´ a (range pH 8e8.7). The pH of the material is therefore strongly buffered by the high content of calcium carbonate (see below). Further evidence of this is provided by the results of a prior study carried out in our laboratory, in which samples of biodeposit and sediment were maintained in a solution of hydrogen peroxide (w15%, pH 5.5), with shaking, for 24 h. Despite oxidation of the entire AVS fraction and of 95% of the pyritic fraction, there were no changes in pH, although there was a large increase in the concentration of dissolved calcium, which was probably generated during the neutralization of the acidity produced by the oxidation of metal sulphides by carbonates (Otero et al., 2001).

124

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

Table 2 Percentage of sand, silt and clay in biodeposit and sediment layers Average biodepositGSD (%)

Average sedimentGSD (%)

p

Type of test

Average
biodeposit Average
sediment

Inner Rı´ a Sand Silt Clay

12.9G7.6 (nZ41) 52.2G7.4 34.9G7.7

58.8G16 (nZ12) 18.4G12 20.7G9.3

!0.001 !0.001 !0.001

M-W rst M-W rst M-W rst

0.22 2.83 1.68

Mid Rı´ a Sand Silt Clay

21.3G6.9 (nZ43) 46.2G8.0 32.8G6.7

33.7G8.3 (nZ12) 31.5G6.5 34.9G5.2

!0.001 !0.001 0.218

M-W rst t-Test M-W rst

0.63 1.47 0.94

Outer Rı´ a Sand Silt Clay

24.2G9.0 (nZ33) 42.9G7.8 32.8G6.0

49.1G18.2 (nZ15) 23.5G11 27.4G7.7

!0.001 !0.001 0.009

M-W rst M-W rst M-W rst

0.49 1.82 1.20

Differences between biodeposit/sediment layers at the specified confidence level (p) were calculated with a t-test or, when the normality test failed, a Mann-Whitney rank test (M-W rst).

The redox conditions of the biodeposits and the sediments in all cases corresponded to strongly reduced environments, with the Eh ranging between
380 and 29 mV (Fig. 3). The highest values corresponded to the upper 5 cm, which differed significantly from the lower layers of the biodeposit itself and from the sediment. The higher redox potential in the upper layer of the biodeposit may be a consequence of its contact with the layer of oxic water, characteristic of the Galician Rı´ as. In contrast, below 5 cm the values of the redox potential of the biodeposit and the sediment were not significantly different, which appears to indicate that as the biodeposit accumulates on the sea bed, the redox boundary migrates upwards, conferring anoxic conditions to the biodeposit. The total organic carbon (TOC) content was significantly higher in the biodeposit (3.3e3.4%) than in the sediment (1.7e3.5%), except in the inner section of the Rı´ a (Table 3). Location also affects the content of organic C is location. As with the finer fractions of the sediment, the lowest values were found at the installations situated in the outermost part of the Rı´ a. In

contrast, the content of carbonates (TIC) was higher in the sediments than in the biodeposit, with the highest levels in the middle and inner zones of the Rı´ a (Table 3). These results show that the composition of both the sediments and the biodeposits are affected by the processes that characterize each sector of the Rı´ a de Arousa under study (e.g. supply of sediments of continental origin, age of the installations, velocity of the sea currents, etc.). The AVS fraction was present in both the biodeposits and the sediments, indicating the existence of reducing sulphate activity. However, it should be pointed out that the concentrations in the biodeposits were much higher than in the sediments (AVS biodepositZ7:02G13:4 mmol g
1; AVS sedimentZ0:36G0:53 mmol g
1, p!0:01, nZ10; data not shown), possibly partly due to the higher quantity of organic carbon in the biodeposits, and more importantly that it is a more recent organic matter, presumably comprised of labile forms of organic matter (e.g. low weight organic acids), which stimulate the activity of sulphate-reducing bacteria (Howarth, 1984).

Fig. 2. Variation in the pH of the biodeposits and sediments.

125

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

Fig. 3. Redox potentials of the surface layer (0e5 cm) and a deeper layer (>5 cm) of the biodeposit and of the sediment.

3.2. Total content of trace metals Although the area surrounding the Rı´ a de Arousa is neither densely populated nor heavily industrialized, the presence of contamination by different heavy metals has been reported. However, the relevant studies concerned a very limited number of trace elements (basically Cr, Ni, Cu, Pb and Zn) and scarcely considered their geochemical behaviour (see e.g. Barreiro et al., 1988; Real et al., 1993; Calvo de Anta et al., 1999; Prego and Cobelo-Garcı´ a, 2003). The main source of trace element contamination occurs at some 10 km upstream of the mouth of the river Ulla, as a result of the effluent discharged from a tanning factory situated in the town of Padro´n (see Fig. 1b). In this zone the Cr content in the sediments of the river Ulla reached above 1700 mg kg
1 in the !63 mm fraction, and high concentrations of this element were also found in the inner part of the Rı´ a de Arousa, with values ranging between 191 and 620 mg kg
1, whereas in the parts furthest from the mouth of the river, the values were clearly lower (19e73 mg kg
1) (Barreiro et al., 1988; Real et al., 1993; Calvo de Anta et al., 1999; Prego

and Cobelo-Garcı´ a, 2003). Zinc is the second most common element, with concentrations ranging between 123 and 282 mg kg
1 in sediments from the river Ulla and between 125 and 150 mg kg
1 in sediments from the Rı´ a of Arousa. The maximum concentrations coincide with those of Cr and therefore may originate from the waste discharged from the tanning factory and/or from the sewage discharged from the town of Padro´n (Real et al., 1993, Fig. 1). Contamination by Cu and Ni (river Ulla: w50e365 mg Cu kg
1, and w40e130 mg Ni kg
1; Rı´ a of Arousa: w25e55 mg Cu kg
1; w40e50 mg Ni kg
1) appears to originate from the run-off from an abandoned copper sulphide mine which drains into tributaries of the river Ulla (Real et al., 1993; Calvo de Anta and Pe´rez-Otero, 1994). The available data on Pb refer to the concentrations in the !2 mm fraction, which in the sediments range between 2 and 188 mg kg
1, with the highest concentrations corresponding to the areas closest to the main centres of population (e.g. Vilagarcı´ a of Arousa, Cambados, Ribeira) (Barreiro et al., 1988). Finally, Cd was found to be present at much lower concentrations than the previous elements, with concentrations ranging between 0.7e2.5 mg kg-1

Table 3 Mean concentration (GSD) of total organic carbon (TOC), total inorganic carbon (TIC) and total heavy metals in biodeposits and sediments from the Rı´ a of Arousa, and the p values for ANOVA Biodeposit Inner (nZ15) TOC (%) TIC (%) Cu (mg kg
1) Zn Cd Ni Cr Hg Pb As

Sediment Mid (nZ13)

Outer (nZ10)

Inner (nZ11)

p Mid (nZ10)

Outer (nZ8)

Site

Biodeposit/ sediment

Interaction

2.41G0.3 1.60G0.5

3.30G0.9 2.81G2.0

2.51G0.51 3.15G1.03

3.51G1.0 2.64G2.0

1.85G0.8 5.00G2.3

1.70G0.69 5.00G2.68

!0.001 !0.001

!0.001 !0.001

0.007 n.s

32.07G3.3 175.6G113 0.480G0.06 65.80G10.2 153.1G45.9 0.256G0.31 57.73G4.68 25.25G6.39

22.54G4.6 114.8G27.6 0.560G0.15 62.54G15.1 142.2G29.5 0.220G0.19 58.46G7.45 23.97G6.10

19.31G3.0 144.5G79.9 0.480G0.18 55.85G9.89 129.2G27.6 0.130G0.04 61.92G7.78 25.38G6.74

21.40G7.75 96.80G36.3 0.515G0.19 62.60G13.1 111.4G22.7 0.093G0.08 44.30G15.4 21.91G7.61

17.80G5.7 79.80G22.9 0.320G0.09 44.80G13.2 91.00G23.9 0.06G0.03 44.20G10.2 15.98G3.00

12.50G4.04 82.75G23.3 0.360G0.10 45.13G15.1 104.2G39.6 0.060G0.02 48.13G17.9 22.74G10.1

!0.001 n.s. n.s. 0.002 n.s. n.s. n.s. n.s.

!0.001 !0.001 0.006 0.003 !0.001 !0.001 !0.001 0.016

n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s.

For each core we considered the total content of trace metals extracted from the surface layer of the biodeposits (5e6 cm depth) and for the sediment the total content of trace metals extracted from the sample taken at 150 cm. When the depth was less than this, the deepest samples were used. n.s., not significant.

126

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

(Barreiro et al., 1988). Unfortunately there are no data available for Hg and As. In concordance with the results of the above-cited studies, our results showed that Cr was the predominant element in the sediment (the layer situated below the biodeposits, with average values ranging between 91 and 111 mg kg
1, followed by Zn (79e96 mg kg
1), Ni

(55e66 mg kg
1), Pb (57e62 mg kg
1), As (23e26 mg kg
1), Cd (0.48e0.56 mg kg
1) and Hg (0.13e0.26 mg kg
1) (Table 3, Fig. 4). The concentrations of Ni, Cr, Cu and Zn were similar to those found in previous studies of surface sediments in the same Rı´ a, but not affected by mussel biodeposits (Real et al., 1993). These concentrations appear to indicate the absence of heavily

Fig. 4. Variation in the total content of trace metals with depth in different parts of the Rı´ a of Arousa.

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

contaminated areas in the Rı´ a of Arousa, with the exceptions of the mouth of the river Ulla e where there are high concentrations of Cr, and to a lesser extent, Zn and Cu e and some small areas close to the main centres of population, where high levels of Pb were found (Real et al., 1993; Prego and Cobelo-Garcı´ a, 2003). Thus, some authors consider that mussel cultivation is the human activity that has the greatest impact on the geochemical cycles and the ecology of this Rı´ a (Figueras, 1989; Tenore et al., 1982). Previous studies have shown high accumulations of trace metals in the faecal material produced by mussels, which may accumulate on the floor of the Rı´ a (Collazo and Pascual, 1997). Our results show that the concentrations of all of the trace elements in the biodeposits were significantly higher than in the underlying sediments, with enrichment factor values ranging from 2 to 25 (Table 3, Fig. 4). Despite the fact that Cr was the predominant element, the highest enrichment factors corresponded to Hg (Fig. 5). Some concern has recently been expressed about an increase in the concentration of this metal, which although has not yet reached a critical level, indicates the existence of a focus of contamination that must be monitored (Macı´ as and Mora, 2001). However, more than 50% of the total Hg is associated with the residual fraction of the sediment (basically silicates), with very low reactivity and bioavailability (Figs. 6, 7). A similar situation was found with other trace metals, except for Cu, which predominantly occurred in the pyritic fraction.

127

As regards spatial variation, several studies have shown a decrease in the total content of trace metals from the innermost areas of the Rı´ a towards the sea (Fo¨rstner and Wittmann, 1983; Fuller, 1990). However, we found that there were no significant differences among the three areas in the total contents of trace metals, except for Ni, which showed a higher content in the inner part of the Rı´ a (Table 3), possibly because of the mixing processes that take place (Real et al., 1993). 3.3. Geochemical behaviour of Fe, Mn and trace elements The highly significant correlations between concentrations of Cr, Zn, Ni, Cu, As and Hg and those of total organic C, Fe and Al show that the presence of the former elements are largely determined by the contents of organic matter, iron oxyhydroxides and clays (Table 4). Furthermore, the highly significant correlations between total S and divalent metals (except Pb and Cd, see below) indicate that in anoxic environments, where sulphate reduction takes place, the formation of metal sulphides also contributes to the total content of these elements (see e.g. Morse and Luther, 1999). In contrast, the negative correlations with carbonates indicate that the presence of shells (mainly mussel shells) contributes to the dilution of metals in both the sediment and in the biodeposit. Lead and Cd showed different behaviour and were not correlated with the previous components. Other authors have obtained

Fig. 5. Enrichment factor (EF) for trace metals extracted from core BCAR2. Enrichment factors were calculated by dividing the ratio [total metal(ppm)]/[total Al(%)] in the samples, by the same ratio calculated for the deepest section of the sediment core.

128

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

Fig. 6. Distribution of trace metals in the reactive, pyrite and residual fractions in core BCAR2 (inner Rı´ a).

similar results for the Rı´ a of Vigo (e.g. Rubio et al., 2000; Evans et al., 2003). One possible explanation for these results is that both Pb and Cd are incorporated in the structure of the biogenic carbonates, as shown by Sturesson (1976, 1978), therefore biogenic carbonates may be a source of these two elements in biodeposits.

Our results support this idea, as a high percentage of Cd (12.6e16.6%) and of Pb (3.4e19.2%) in the reactive fraction is associated with the carbonate fraction in both biodeposits and sediments (Fig. 8). The latter may also help to explain the low degree of pyritization of these metals, despite the fact that divalent metals should form

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

129

Fig. 7. Distribution of trace metals in the reactive, pyrite and residual fractions in core BCAR4 (mid-Rı´ a).

the corresponding metal sulphides or co-precipitate with iron sulphide under anoxic-sulphidic conditions (see e.g. Morse et al., 1987; Huerta-Dı´ az and Morse, 1992; Di Toro et al., 1990). Further explanation for the low degree of pyritization of these two metals, as well as that of Zn may be related to the method applied. In this

respect, Morse and Luther (1999) showed that these three elements precipitate forming their respective sulphides (MeS) before the formation of FeS and FeS2 occurs, as they show faster water exchange than Fe2C, however, the sulphides that these three elements form are soluble in 1N HCl and therefore, on extraction, are

130

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

Table 4 Coefficients of correlation between trace metals and TOC, TIC, total S, Fe and Al %

Cu

Zn

Cd

Ni

Cr

Hg

Pb

As

Biodeposit TOC (nZ41) TIC (nZ41) S total (nZ23) Al (nZ27) Fe (nZ35)

0.589**
0.704*** 0.693*** n.s. 0.444***

0.437***
0.593*** 0.545*** n.s. 0.460***

0.360* n.s. n.s. n.s. n.s.

0.546***
0.784*** 0.615*** 0.507** 0.816***

0.367*
0.559*** 0.516*** n.s. 0.608***

0.737***
0.327* 0.389* n.s. n.s.

n.s. n.s. n.s. n.s. n.s.

0.446**
0.567*** 0.515*** 0.510** 0.675***

Sediment TOC (nZ23) TIC (nZ23) S total (nZ23) Al (nZ28) Fe (nZ39)

0.616**
0.632** 0.549** 0.475* 0.328*

0.439*
0.649** n.s. n.s. n.s.

0.349** n.s. n.s. n.s. n.s.

0.698***
0.817*** 0.622*** 0.460* 0.532***

n.s.
0.509* n.s. 0.627*** 0.603***

0.685***
0.676*** 0.480* n.s. n.s..

n.s. n.s. n.s. 0.409* n.s.

n.s.
0.451* n.s. n.s. 0.420**

*p!0:05, **p!0:01, ***p!0:001; n.s., not significant.

incorporated with the reactive fraction and not the pyritic fraction (Cooper and Morse, 1998). Nevertheless, a small proportion of the concentration of these metals, which corresponds to the concentration that we found associated with the pyrite fraction, may co-precipitate with pyrite, because it is not soluble in HCl (Table 5, Fig. 9). Chromium also showed a low degree of pyritization (Table 5, Figs. 6, 7, 9). Under anoxic conditions Cr occurs as Cr3C, the electronic configuration of which makes it kinetically inert to reaction with sulphides and which is therefore, not incorporated into pyrite (Morse and Luther, 1999). Under anoxic conditions Cr either precipitates as an hydroxide [Cr(OH)3, (Cr,Fe)(OH)3] or becomes associated with organic matter (Rai et al., 1989). A high percentage of this metal was also found in the residual fraction (Figs. 6, 7), where it should basically be found as chromite (FeCr2O4), which is the most important natural chromium ore. Slightly higher degrees of pyritization were found for Mn and Ni (Table 5, Fig. 9). In the case of Mn, the average values for the degree of pyritization were 34.2G15% for the biodeposit and 47.3G8% for the sediment, similar to those found in previous studies (Huerta-Dı´ az and Morse, 1992; Otero et al., 2003). The explanation for these results is derived from the fact that

the oxides of Mn(IV) are reduced to Mn2C before reduction of Fe oxyhdroxides and initiation of sulphate reduction. Therefore, when Mn(IV) becomes reduced there are sulphides in the interstitial water. Under these conditions the solubility of Mn2C is determined by carbonate, with precipitation of rhodochrosite (MnCO3) or more probably a mixed carbonate of Ca and Mn, such as kutnahorite (for more detail, see Bo¨ttcher, 1998; Glasby and Schulz, 1999; Otero et al., 2003). The latter aspect may explain the relatively high percentages of Mn that we found associated with the carbonate fraction (biodeposit: 10.95%; sediment: 14.2%; Fig. 8). On the other hand, the incorporation of Mn into the pyrite fraction requires that a large quantity of the Fe oxyhydroxides have previously been pyritized. Previous studies have established that Mn begins to appear in high quantities (DTMP-MnO20%) in the pyrite fraction when more than 85% of the poorly crystalline

Table 5 Average concentration (mg kg
1) of trace metals (GSD) in the reactive fraction (nZ10) and the pyritic fraction (nZ11) of biodeposits and sediment samples from the inner, middle and outer zone

Cd Pb Zn Cr Ni Cu As

Fig. 8. Partitioning of Fe, Mn, Pb and Cd in biodeposits and sediment from core GCAR2.

Hg

Reactive Pyrite Reactive Pyrite Reactive Pyrite Reactive Pyrite Reactive Pyrite Reactive Pyrite Reactive Pyrite Reactive Pyrite

Biodeposit

Sediment

0.200G0.060 0.005G0.001 26.28G2.83 0.050G0.01 20.98G11.41 1.51G2.30 54.23G18.80 2.30G1.08 13.30G0.80 3.80G1.90 9.18G3.37 5.30G3.48 2.80G0.57 3.47G1.26 0.016G0.014 0.098G0.076

0.120G0.010 0.041G0.027 12.21G3.52 0.021G0.007 13.84G2.30 3.44G1.24 26.60G5.28 4.64G1.70 10.13G2.66 6.21G2.12 1.08G1.01 10.73G4.07 1.55G0.67 9.13G3.40 0.001G0.001 0.024G0.013

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

131

concentration of these trace metals (w1000 times higher than the concentration of Cu and 100 000 times higher than the concentration of Hg), it is highly probable that they will form mixed sulphides. However, Cu and Hg sulphides are only soluble in HNO3, therefore unlike Pb and Zn sulphides, they become incorporated in the pyrite fraction (Cooper and Morse, 1998). As regards As, the association of this element with the pyrite groups of minerals (e.g. arsenopyrite) is well known, therefore the high degrees of pyritization are to be expected and are consistent with those found in previous studies (e.g. Huerta-Dı´ az and Morse, 1992).

Fig. 9. Mean (GSD) of degree of pyritization of Fe (DOP), Mn and trace metals (DTMP) in biodeposits (nZ7) and sediments (nZ11) samples from the inner, middle and outer zones.

Fe oxyhydroxides have been pyritized (Otero and Macı´ as, 2003). When this occurs, the concentration of Fe2C in the interstitial water begins to be reduced at the same time as there is an increase in the ratio of Mn2C/ Fe2C in the interstitial water, which allows absorption of dissolved Mn onto FeS and pyritization (Morse and Luther, 1999). The degree of pyritization of Fe was 58.9G7% for the biodeposit and 67.9G10% for the sediment, which indicates that the non-residual Fe was mainly incorporated in the pyrite fraction (Table 5, Figs. 6, 7, 9). However, it should also be pointed out that a large part of the reactive Fe was not incorporated in the pyrite fraction, despite the strongly reducing conditions found in these environments, and was presumably accounted for by Fe associated with crystalline Fe oxyhydroxides or silicates, forms which some authors consider as poorly reactive for forming pyrite (see e.g. Canfield et al., 1992; Postma, 1993; Postma and Jakobsen, 1996). The incorporation of Ni into the pyrite fraction ranged between 33.6G8% for the biodeposit and 38.4G6% for the sediment. The lower degree of pyritization of Ni compared with that of Fe appears to be related to the lower rate of water exchange, because Ni reacts more slowly with sulphides, tending to become incorporated in the pyrite fraction rather than forming sulphide, and therefore the DTMP-Ni tends to increase with the DOP (Morse and Luther, 1999). By contrast, Cu, As and Hg showed generally high degrees of pyritization (Cu: biodeposit 58.6G18%, sediment 90.2G10%; As: biodeposit 68.8G8%, sediment 85.4G6%; Hg: biodeposit 97.8G2%, sediment 94.1G2%, Fig. 9). The high percentages of pyritization of Cu and Hg are consistent with fast reaction kinetics with the sulphides, higher than those of Fe2C (Morse and Luther, 1999), which allows them to form their respective sulphides. However, taking into account that the concentration of reactive Fe is much higher than the

3.4. Mobilization of trace metals by oxidation of metal sulphides Although the total content of trace metals does not represent a high environmental risk, because a high percentage of the metals are associated with the residual fraction (and are therefore rather unreactive), the high degree of pyritization found for the most toxic metals (e.g. Cu, Hg, As) may favour their mobilization, thus making them more bioavailable. A change in the geochemical conditions of the biodeposit towards oxidizing conditions, as may occur if it forms a suspension due to the action of water currents or of bioperturbation generated by the benthic biota, may involve oxidation of the metal sulphides and release of the metals associated with the pyrite fraction into the overlying water of the Rı´ a or into the interstitial water. The increases in Fe, Mn, As, Cu and Cr in solution on maintaining the biodeposit and sediment in suspension in oxic sea water (Eh: 350e500 mV) are shown in Fig. 10. The results show an increase in the concentration of Fe, Mn and to a lesser extent As and Cu. All of these metals show high degrees of pyritization, therefore their release is presumably due to oxidation of metal sulphides. Further support for this idea is given by the fact that Cr, a metal that scarcely appeared in the pyrite fraction, was always below the limit of detection (5 ppb). The concentration of the latter, as with the other elements, increases notably on producing strong oxidation with hydrogen peroxide, possible because of the release of the fraction of Cr associated with the organic matter. These results indicate that the elements with the highest degrees of pyritization (Hg, Cu, As, Ni) may be released into the water and accumulate in the benthic biota, as well as in the mussels. The results are also consistent with those of a previous study in which it was observed that Nereis diversicolor accumulated trace elements that were associated with the pyrite fraction (Cu, Ni) but did not accumulate Cr, despite this being present at high concentrations in the sediment (much higher than those of Cu) in the reactive fraction (soluble en 1 N HCl) and associated with the organic matter (for

132

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

Fig. 10. pH, Eh and concentration of trace metals in solution obtained after maintaining biodeposit and sediment samples in suspension in oxic sea water. L. D: below detection limit.

more details see Otero et al., 2000a,b; Otero and Macı´ as, 2002). Furthermore, it was also demonstrated that following the sharp increase in the concentration of Fe and Mn in solution, caused by the input of hydrogen peroxide, the concentration then decreased rapidly because of the low solubility of the oxidized forms of these metals, which precipitate forming oxyhydroxides.

4. Conclusions The results show that the mussel deposits under study were enriched with trace elements compared with the underlying sediments. However, the concentrations

found correspond to a sedimentary environment that is not affected by serious contamination. Furthermore, a high percentage of the total content of trace metals was associated with the residual fraction, therefore the corresponding reactivity and bioavailability were very low. However, the high degree of pyritization found for some of the most toxic trace metals may favour their release by oxidation of the sulphides that they form, thus making them bioavailable to benthic fauna. The results indicate the need for strict monitoring of water and sediment quality in Galician Rı´ as, with special attention being paid to contamination by mercury. In addition, further studies are required on the geochemical behaviour of trace metals in relation to their bioavailability.

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

Acknowledgements The present study was financed by the Consellerı´ a de Pesca, Marisqueo e Acuicultura of the Xunta de Galicia (Autonomous government of Galicia) and the Spanish (Direccio´n General de Universidades del Ministerio de Educacio´n y Ciencia, HBP-2002-0040) and Brasilian government (CAPES). We thank Marı´ a Santiso for its laboratory assistance.

References Barg, U.C., 1995. Guidelines for the Promotion of Environmental Management of Coastal Aquaculture Development. 0FAO Fisheries Technical Paper 328. FAO, Rome. Barreiro, R., Carballeira, A., Real, C., 1988. Metales pesados en los sedimentos de cinco sistemas de Rı´ a (Ferrol, Burgo, Arousa, Pontevedra y Vigo. Thalassas 6, 61e70. Belzunce-Segarra, M.J., Bacon, J.R., Prego, R., Wilson, M.J., 1997. Chemical forms of heavy metals in surface sediments of the San Simo´n inlet, Rı´ a de Vigo, Galicia. Journal of Environmental Science and Health A32, 1271e1292. Berner, R.A., 1970. Sedimentary pyrite formation. American Journal of Science 268, 1e23. Besada, V., Fumega, J., Vaamonde, A., 2002. Temporal trends of Cd, Cu, Hg, Pb and Zn in mussel (Mytilus galloprovincialis) from the Spanish North-atlantic coast 1991e1999. The Science of the Total Environment 288, 239e253. Bo¨ttcher, M.E., 1998. Manganese(II) partitioning during experimental precipitation of rhodochrosite-calcite solid solutions from aqueous solutions. Marine Chemistry 62, 287e297. Cabanas, J.M., Gonza´lez, J.J., Marin˜o, J., 1979. Estudio del mejillo´n y de su epifauna en los cultivos flotantes de la Rı´ a de Arousa. II. Observaciones previas sobre la retencio´n de partı´ culas y la biodeposicio´n de una batea. Boletin del Instituto Espanol de Oceanografia 268, 45e80. Calvo de Anta, R.M., Pe´rez-Otero, A., 1994. Soils affected by acid mine waters in Galicia. Water, Air, and Soil Pollution 73, 247e263. Calvo de Anta, R.M., Quintas, Y., Macı´ as, F., 1999. Caracterizacio´n de materias para la recuperacio´n de suelos degradados. I. Sedimentos bioge´nicos de las Rı´ as de Galicia. Edafologı´ a 6, 47e58. Canfield, D.E., Raiswell, R., Bottrell, S., 1992. The reactivity of sedimentary minerals towards sulfide. American Journal of Science 292, 659e683. Collazo, J.L., Pascual, C., 1997. Contenido en cobre, plomo, hierro y cromo en los biodepo´sitos del mejillo´n cultivado en Rı´ a de Arousa (NO Espan˜a). In: Prego, R., Ferna´ndez, J.M. (Eds.), Procesos Biogeoquı´ micos En Sistemas Costeros Hispano-Lusos. CSICDiputacio´n de Pontevedra, pp. 73e77. Collazo, J.L., Tilves, F., Vin˜as, J.M., Pascual, C., 1993. Estudio in situ de la biodeposicio´n del mejillo´n (Mytilus galloprovincialis) cultivado en batea: produccio´n y composicio´n quı´ mica elemental. In: Cervin˜o, A., Landin, A., de Coo, A., Guerra, A., Torre, M. (Eds.), Proceedings of Fourth National Congress on Aquaculture. Centro de Investigaciones Marin˜as de Pontevedra, pp. 437e442. Cooper, D.C., Morse, J.W., 1998. Extractability of metal sulfide minerals in acidic solutions: Applications to environmental studies of trace metals contamination. Environmental Science and Technology 32, 1076e1078. Cline, J.E., 1969. Spectrophotometric determination of hydrogen sulfide in natural waters. Limnology and Oceanography 14, 454e458.

133

Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Hicks, M.B., Mayr, S.M., Redmond, M.S., 1990. Toxicity of cadmium in sediments: The role of acid volatile sulfide. Environmental and Toxicological Chemistry 9, 1487e1502. Evans, G., Howarth, R.J., Nombela, M.A., 2003. Metals in the sediments of ensenada de San Simo´n (inner Rı´ a of Vigo), Galicia, NW Spain. Applied Geochemistry 18, 973e996. Figueras, A.J., 1989. Mussel culture in Spain and France. World Aquaculture 20, 8e17. Fo¨rstner, U., Wittmann, G.T.C., 1983. Metal Pollution in the Aquatic Environment. Springer-Verlag, Berlin. Fraga, F., Margalef, R., 1979. Las Rı´ as gallegas. In: Estudio y Explotacio´n Del Mar En Galicia. Universidad de Santiago de Compostela, Santiago de Compostela, pp. 176e182. Fuller, C.C., 1990. Distribution and transport of sediments bound metals contaminants in the Rı´ o Grande de Tarcoles, Costa Rica (Central America). Water Research 24, 805e812. Gee, G.W., Bauder, J.W., 1986. Particle-size analysis. In: Klute, A. (Ed.), Methods of Soil Analysis. Part 1. Physical and Mineralogical Methods. American Society of Agronomy, Inc. and Soils Science of America, Madison, Wisconsin, USA, pp. 383e411. Glasby, G.P., Schulz, H.D., 1999. Eh-pH diagrams for Mn, Fe, Co, Ni, Cu, and As under seawater conditions: application of two new types of Eh pH Diagrams to the study of specific problems in marine geochemistry. Aquatic Geochemistry 5, 227e248. Howarth, R.W., 1984. The ecological significance of sulfur in the energy of salt marsh and coastal marine sediments. Biogeochemistry 1, 5e27. Huerta-Dı´ az, M.A., Morse, J.W., 1990. A quantitative method for determination of trace metal concentrations in sedimentary pyrite. Marine Chemistry 29, 119e144. Huerta-Dı´ az, M.A., Morse, J.W., 1992. Pyritization of trace metals in anoxic marine sediments. Geochimica Cosmochimica Acta 56, 2681e2702. Kostka, J.E., Luther III, G.W., 1994. Partitioning and speciation of soil phase iron in saltmarsh sediments. Geochimica Cosmochimica Acta 58, 1701e1710. Macı´ as, F., Mora, J., 2001. Procesos biogeoquı´ micos y comunidades bento´nicas en fondos de Rı´ as bajo bateas de mejillo´n (Subproyecto 4). Xunta de Galicia. Informe ine´dito. Morse, J.W., Millero, F.J., Cornwell, J.C., Rikard, D., 1987. The chemistry of the hydrogen sulfide and iron sulphide system in natural waters. Earth-Science Reviews 24, 1e42. Morse, J.W., Luther III, G.W., 1999. Chemical influences on trace metals-sulfide interactions in anoxic sediments. Geochimica Cosmochimica Acta 63, 3373e3378. Otero, X.L., Ferna´ndez-Sanjurjo, M.J., 2000. Mercury in faeces and feathers of yellow-legged gulls (Larus cachinnans) and in soils from their breeding sites (Cı´ es Islands-NW Spain) in the vicinity of a chlor-alkali plant. Fresenius Environmental Bulletin 9, 56e63. Otero, X.L., Macı´ as, F., 2002. Variation with depth and season in metal sulfides in salt marsh soils. Biogeochemistry 61, 247e268. Otero, X.L., Macı´ as, F., 2003. Spatial variation in pyritization of trace metals in salt marsh soils. Biogeochemistry 62, 59e86. Otero, X.L., Huerta-Dı´ az, M.A., Macı´ as, F., 2000a. Heavy metal geochemistry of saltmarsh soils from the Rı´ a of Ortigueira (mafic and ultramafic areas, NW Iberian Peninsula). Environmental Pollution 110, 285e296. Otero, X.L., Sa´nchez, J.M., Macı´ as, F., 2000b. Bioaccumulation of heavy metals in thionic fluvisols by a marine polychaete (Nereis diversicolor): The role of metal sulfide. Journal of Environmental Quality 29, 1133e1141. Otero, X.L., Quintas, I., Macı´ as, F., 2001. Caracterizacio´n fı´ sicoquı´ mica de los biodepo´sitos de mejillo´n en las Rı´ as Baixas (GaliciaNE de la Penı´ nsula Ibe´rica). Nova Acta Cientı´ fica Compostelana 11, 197e206. Otero, X.L., Huerta-Dı´ az, M.A., Macı´ as, F., 2003. Influence of a turbidite deposit on the extent of pyritization of iron, manganese

134

X.L. Otero et al. / Environmental Pollution 136 (2005) 119e134

and trace metals in sediments from the Guaymas Basin, Basin of California (Mexico). Applied Geochemistry 18, 1149e1163. Postma, D., 1993. The reactivity of iron oxides in sediments: A kinetic approach. Geochimica Cosmochimica Acta 57, 5027e5034. Postma, D., Jakobsen, R., 1996. Redox zonation: Equilibrium constraints on the Fe(III)/SO4-reduction interface. Geochimica Cosmochimica Acta 60, 3169e3175. Prego, R., Cobelo-Garcı´ a, A., 2003. Twentieth century overview of heavy metals in the Galician Rias (NW Iberian Peninsula). Environmental Pollution 121, 425e452. Puente, X., Villares, R., Carral, E., Carballeira, A., 1996. Nacreous shell of Mytilus galloprovincialis as a biomonitor of heavy metals pollution in Galiza (NW Spain). The Science of the Total Environment 183, 205e211. Rai, D., Eary, L.E., Zachara, J.M., 1989. Environmental chemistry of chromium. The Science of the Total Environment 86, 15e23. Real, C., Barreiro, R., Carballeira, A., 1993. Heavy metals mixing behaviour in estuarine sediments in the Rı´ a of Arousa (NW Spain). Differences between metals. The Science of the Total Environment 128, 51e67. Rı´ o Barja, F.X., Rodrı´ guez, F., 1992. Os Rı´ os Galegos. Concello da Cultura Galega, Santiago de Compostela, 333 pp. Rubio, B., Nombela, M.A., Vilas, F., 2000. Geochemistry of major and trace elements in sediments of the Rı´ a de Vigo (NW Spain):

an assessment of metal pollution. Marine Pollution Bulletin 40, 968e980. Sturesson, U., 1976. Lead enrichment in shells of Mytilus edulis. Ambio 5, 253e262. Sturesson, U., 1978. Cadmium enrichment in shells of Mytilus edulis. Ambio 7, 122e125. Tenore, K.R., Cal, R.M., Hanson, R.B., Lo´pez-Jamar, E., Santiago, G., Tietjen, J.H., 1982. Coastal upwelling in the Rı´ as Bajas, NW Spain. Constrasting the benthic regimes of the Rias de Arosa and Muros. Journal of Marine Research 40, 701e772. Tessier, A., Campbell, P.G., Bisson, M., 1979. Sequential extraction for the speciation of particulate trace metals. Analytical Chemistry 51, 844e851. Torre, E., 1958. Estado actual del conocimiento de las Rı´ as gallegas. In: Homenaje a Otero Pedrayo. Ed. Galaxia, Vigo, pp. 237e250. VidaleRomani, J.R., 1984. A orixe das Rı´ as Galegas. Estado da cuestio´n (1986e1983). Cuadernos da Area de Ciencias marin˜as de Estudos Galegos 1, 13e25. von Richthofen, F., 1886. Fu¨hrer fu¨r Forschungsreisende. Jenecke, Hanover, 734 pp. Xunta de Galicia, 2001. Acuicultura. www.xunta.es/conselle/pe/ sector_acui.htm. Zar, J., 1996. Biostatistical Analysis. Prentice-Hall, Upper Saddle River, NJ.