The Significance of Coarse Sediments in Metal ... - Springer Link

THE SIGNIFICANCE OF COARSE SEDIMENTS IN METAL POLLUTION STUDIES IN THE COASTAL ZONE MARIA ALOUPI and MICHAEL O. ANGELIDIS∗ Department of Environmental Studies, University of the Aegean, Karadoni 17, 81100 Mytilene, Greece (∗ author for correspondence)

(Received 28 September 1999; accepted 14 December 2000)

Abstract. The role of coarse material (sand fraction) in the distribution of metals in polluted marine sediments was investigated in the harbors and the coastal zone of Mytilene, island of Lesvos, Aegean Sea. It was found that sand fraction contains a relatively significant proportion of the anthropogenic metals and therefore it cannot be neglected in metal pollution studies of coastal sediments. Also the distribution of the anthropogenic metals (Cd, Cu, Pb and Zn) in both silt+clay and sand fractions follow the same pattern indicating similar pollution sources. In the bulk sediment (clay + silt + sand fraction) all anthropogenic metals had a significant amount (> 50%) in the acid-extractable (and potentially bio-available) fraction. Keywords: Aegean Sea, coarse sediments, island, metals, pollution, sand

1. Introduction One of the major factors that affect the metal concentrations in aquatic sediments is the size of the sediment particles. Higher concentrations are found in the finer particles (Goldberg, 1954; Krauskopf, 1956; Gibbs, 1977; Horowitz and Elrick, 1987) due to a synergetic action of physical (surface area) and chemical (mineralogy) characteristics of the particles in each size fraction. First, the smaller particles have a greater specific surface area than the larger ones. Second, most fine particles consist mainly of clay minerals, which have more metal binding sites than the silicate or carbonate minerals, which are the major components of coarser material in marine sediments (Campbell et al., 1988). As a result, finer particles offer much more sites than the coarser ones for chemical reactions. Therefore, although high metal concentrations can be occasionally found in coarse sediment material (Filipek et al., 1981; Robinson, 1982), the silt and clay fractions usually contain higher metal concentrations than the sand fraction (Förstner and Wittmann, 1983). That is the main reason why metal pollution studies are often carried on the finer fraction (less than 63 µm), suggesting that the coarser material is an ‘inert diluent’ without any significant metal contribution. However, in the coastal zone, the coarse material is an important constituent of the sediment and may represent more than 50% of the bulk sediment mass. Therefore it cannot be neglected when assessing the pollution impact on the coastal Water, Air, and Soil Pollution 133: 121–131, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.

122

M. ALOUPI AND M. O. ANGELIDIS

sediments. Furthermore, the vicinity of land based pollution sources which may dispose off large particles with relatively high metal content (urban and industrial effluents and deposits) may lead to enhanced metal concentrations in coarse sediment material (Horowitz, 1991). The present study is an attempt to investigate the relative importance of coarse sediments as metal carriers in a coastal area affected by land-based pollution sources, compared to fine grained. As total metal concentrations in sediments are the resultant of both geological (i.e. natural) and anthropogenic inputs in the marine environment, we used the extraction with 0.5 N HCl to isolate the anthropogenic effect. This technique removes the non-residual, i.e. non-lattice held metals from sediment grains. These metals are not part of the silicate matrix and have been incorporated into the sediments from aqueous solution by processes such as adsorption and organic complexation, including, thus, metals originating from anthropogenic sources (Chester and Voutsinou, 1981). The non-residual metals are considered to represent the bio-available part of total metal content of marine sediments (Loring, 1992). In order to compare the amounts of metals accumulated to both granulometric fractions, we examined separately the silt + clay (i.e. < 63 µm) and sand (63 µm - 1 mm) fractions. The study area is the harbor and the coastal area of Mytilene (pop. 25,000), island of Lesvos, Aegean Sea, Greece, which receives untreated urban effluents through 25 sewage outfalls located along the town’s coastline (see Figure 1). In general, the urban effluents of Mytilene, though not including any significant industrial effluents, contained important loads of organic matter and anthropogenic metals (Angelidis, 1995). Previous investigation on the contamination of surface sediments in the vicinity of the town, revealed highly enriched concentrations of Cd, Cu, Pb and Zn, as well as organic carbon, in the inner and most of the outer harbor. Some enrichment was also recorded in a part of the sediments of ancient harbor located at the northern part of the town (stations 19 and 21), whereas no signs of metal contamination were detected in the wider coastal zone (Angelidis and Aloupi, 1997). The coastal area is characterized by fine grain sediments in the inner and outer harbor (stations 1– 12, Figure 1) turning into coarser texture seawards (stations 13–18). The sediments of the ancient harbor are sandy with very small amounts of fine-grained material (stations 19–21).

2. Materials and Methods Surface sediment samples were collected from 21 stations in the harbor and the coastal area of Mytilene, island of Lesvos, Greece, during July 1995 (Figure 1). Grain size distribution was measured by wet sieving and the following fractions were determined: silt+clay (< 63 µm), sand (63 µm < × < 1 mm) and gravel (> 1 mm).

SIGNIFICANCE OF COARSE SEDIMENTS IN METAL POLLUTION

123

Figure 1. Study area and sampling stations.

The total metal (Cd, Cr, Cu, Fe, Mn, Pb and Zn) content of the samples was measured after total decomposition of 200 mg of dried ground sediment (combined clay + silt + sand (< 1 mm) sample) with 1 mL of aqua regia and 6 mL of HF, in Savillex Teflon bombs heated in a microwave oven (Loring and Rantala, 1992). The most-available metal content of the sediments was determined separately in the silt + clay fraction and the sand fraction of the sediments using a one- step weak acid extraction method (0.5 N HCl at room temperature [approximately 20 ◦ C] overnight - Agemian and Chau, 1976). The metals extracted with weak acid represent the most biologically available fraction (Luoma and Bryan, 1981) and

124


may be used as an indicator of the anthropogenic contamination of the sediments (Chester and Voutsinou, 1981). Metal determinations were performed using a Perkin-Elmer 5100ZL Atomic Absorption Spectrometer with Zeeman background correction. Chromium, Cu, Fe, Mn, Pb and Zn were determined by FAAS, while Cd was determined by GFAAS with a mixture of 50 µg NH2 H2 PO4 and 3 µg Mg(NO3 )2 as matrix modifier (Aloupi, 1999). The quality assurance (accuracy and precision) of the analytical results for the total metal concentrations was controlled with replicate analysis (n = 6) of Reference Materials certified by NRCC (BCSS-1 marine sediment, PACS-1 harbor sediment) and IAEA (SDM2TM marine sediment) (Aloupi and Angelidis, 2001). Since no certified Reference Material is available for the acid extractable metals, only precision was tested (analysis of 10 sub-samples) and the following relative standard deviations were calculated: Cd 9.9%, Cr 1.5%, Cu 1.9%, Fe 2.1%, Mn 0.6%, Pb 3.8%, Zn 1.5%. In the determination of both total and extractable metal concentrations the 20% of the samples were analysed in duplicates. Field homogeneity was also tested by duplicate sampling in 5 stations. Statistical analysis of data was performed with SPSS Ver. 7.5 for Windows software.

3. Results and Discussion The total metal concentrations and the silt + clay content in the bulk sediments (clay + silt + sand fraction) of the study area are presented in Table I. Table II shows the acid extractable concentrations of the metals in the silt + clay and sand fractions of the sediments. As in the case of total concentrations (Angelidis and Aloupi, 1997), acid extractable concentrations of the anthropogenic metals (Cd, Cu, Pb and Zn) were higher in the sediments of the harbor of the town (stations 1–12) compared to the sediments of the wider coastal area, for both the fine (silt + clay) and coarse (sand) fractions. For all metals (with the exception of Mn), concentrations in the silt + clay fraction were higher than in the sand, as expected. Statistical analysis with the use of paired samples t-test showed statistically significant differences in the means of metal concentrations between the two grain size fractions (Table III). However, the sand fraction had also important metal concentrations, especially in the stations from the harbor of Mytilene (Table II). Metal enrichments in coarse sediments have already been recorded in polluted areas (Krumgalz et al., 1992) and may be explained by the formation of coatings on the surface of the particles. Carbonates, hydroxy-oxides and organic matter, play an important role in the development of the specific surface area of particles that may lead to the increase of their adsorption capacity (Horowitz and Elrick, 1987). On the other hand, coating formation may lead to completely different results when occurring in fine sediments. In such cases coatings may lead to the consolidation of fine particles and to the formation of

125


TABLE I Total metal concentrations and silt + clay content in surface sediments from Mytilene coastal area Stations

H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20 H21

Silt + Clay

Cd

Cr

Cu

Fe

Mn

Pb

Zn

%

µg g−1

µg g−1

µg g−1

%

µg g−1

µg g−1

µg g−1

85.79 71.29 95.37 70.75 96.66 82.30 90.06 11.04 67.42 66.88 66.73 68.86 64.80 87.80 75.96 41.42 42.00 35.45 3.20 0.40 5.95

0.364 0.395 0.495 0.166 0.409 0.232 0.206 0.173 0.130 0.142 0.173 0.170 0.131 0.085 0.082 0.119 0.076 0.062 0.125 0.052 0.048

150 144 154 108 146 138 103 86.3 111 120 116 83.0 122 132 120 86.5 81.8 97.5 71.2 44.6 52.6

66.0 63.5 86.2 43.1 67.6 48.8 41.5 21.0 34.5 41.3 42.0 30.5 27.3 23.4 30.5 24.0 16.7 20.4 24.6 9.39 14.9

2.69 2.39 2.81 2.19 2.66 2.56 2.34 1.43 2.71 2.74 2.58 2.00 2.40 2.70 2.75 1.65 1.74 1.74 1.34 1.29 0.77

258 266 257 242 260 262 248 360 271 259 263 249 280 300 302 222 229 261 228 223 201

82.5 73.6 93.0 52.0 78.2 58.0 63.3 32.7 38.1 54.3 56.5 51.7 44.2 39.8 42.5 39.0 34.5 37.1 44.7 30.5 54.9

173 184 230 112 191 128 134 83.5 74.1 112 116 95.1 69.1 73.1 72.9 76.7 45.8 55.0 83.0 38.8 45.1

larger particles with smaller specific surface area and, thus, smaller adsorption capacity. The particle consolidation may also lead to an ‘artificial’ enhancement of metal concentrations in the coarser sediment fractions, if the consolidated particles are not disaggregated during grain size separation, and therefore analyzed together with the coarse particles. High correlation coefficients were calculated for the human-related metals Cd, Cu, Pb and Zn, in each grain-size fraction separately (Table IV), indicating similar distribution processes. Also, the concentrations of each of the anthropogenic metals in the silt+clay and sand fraction were highly correlated (correlation coefficients for Cd: 0.86, Cu: 0.66, Pb: 0.64 and Zn: 0.84, all significant at 0.05 level). Such correlations suggest that the metal enhancement in the sediments of a polluted area is simultaneously recorded in both silt + clay and sand fractions.

TABLE II

0.388 0.492 0.513 0.188 0.420 0.236 0.219 0.299 0.118 0.194 0.186 0.160 0.164 0.106 0.127 0.113 0.072 0.074 0.175 – 0.136

silt + clay

Cd

0.195 0.322 0.156 0.106 0.202 0.146 0.103 0.160 0.145 0.119 0.126 0.139 0.105 0.072 0.062 0.078 0.037 0.059 0.119 0.064 0.058

sand

a 100% sandy sediment.

H1 H2 H3 H4 H5 H6 H7 H8 H9 H10 H11 H12 H13 H14 H15 H16 H17 H18 H19 H20a H21

Stations

from Mytilene coastal area

18.0 17.7 19.3 15.8 18.8 17.1 16.4 14.6 15.1 15.6 15.8 13.7 14.7 20.5 16.0 14.2 15.2 13.2 13.3 – 10.9

silt + clay

Cr

9.36 11.4 6.58 3.22 8.48 7.82 2.50 6.84 3.35 9.68 6.45 2.44 6.01 2.56 7.18 5.84 7.21 6.93 8.81 5.87 3.07

sand

42.0 44.2 58.9 30.7 44.1 30.4 24.6 26.7 16.7 26.9 28.1 19.4 17.5 10.7 16.1 17.2 9.4 12.6 32.7 – 20.1

silt + clay 19.8 23.8 11.5 3.22 23.4 14.1 1.19 6.44 3.92 19.2 20.3 3.79 8.13 3.96 6.40 4.81 2.87 3.99 10.3 3.11 3.04

Cu sand

0.383 0.372 0.423 0.374 0.408 0.386 0.372 0.231 0.447 0.399 0.367 0.251 0.283 0.352 0.375 0.267 0.346 0.317 0.233 – 0.178

silt + clay

Fe

0.184 0.281 0.114 0.018 0.258 0.179 0.008 0.311 0.072 0.346 0.168 0.056 0.264 0.003 0.262 0.055 0.164 0.126 0.138 0.156 0.017

sand

117 120 125 107 108 110 112 330 73.8 108 103 116 127 81.7 131 119 121 180 115 92.0 109

Mn sand

107 108 99.7 111 110 114 133 120 103 101 104 120 95.4 105 134 105 133 124 92.9 – 87.7

silt + clay 54.0 61.3 81.8 51.1 65.8 48.4 52.7 49.2 28.2 40.8 69.6 50.8 35.3 28.4 36.6 69.2 32.9 37.3 74.5 – 57.3

silt + clay

Pb

40.4 48.7 38.6 5.72 58.6 28.9 0.40 16.1 12.1 5.32 31.2 3.10 14.1 1.08 11.4 22.7 8.60 13.8 32.4 14.5 5.88

sand

110 137 156 75.0 131 82.1 79.8 105 27.8 54.5 72.4 61.9 32.1 27.9 32.5 52.4 31.1 35.7 89.3 – 68.8

silt + clay

Zn

64.0 92.7 58.6 28.5 77.1 40.5 18.6 31.5 19.4 49.3 43.1 24.7 17.1 5.72 10.7 47.2 18.3 23.5 58.3 26.6 30.9

sand

0.5 N HCl extractable metal concentrations (in µg g−1 except for Fe in %) in the silt + clay and sand fractions of surface sediments

126 M. ALOUPI AND M. O. ANGELIDIS


TABLE III Paired-samples t-test for the comparison of means of metal concentrations extracted with 0.5 N HCl from silt + clay and sand fractions of the sediments

Pairs of variables

Difference of means

t

df

p (2-tailed)

Cdsilt+clay – Cdsand Crsilt+clay – Crsand Cusilt+clay – Cusand Fesilt+clay – Fesand Mnsilt+clay – Mnsand Pbsilt+clay – Pbsand Znsilt+clay – Znsand

0.098 9.48 16.8 0.191 -6.94 32.1 34.5

4.779 12.324 7.017 8.052 -1.504 9.466 6.392

17 17 17 17 17 17 17

0.000 0.000 0.000 0.000 ns 0.000 0.000

ns: non significant.

TABLE IV Correlation coefficients of 0.5 HCl extractable metal concentrations in each of the silt + clay and the sand fractions of the sediments

Cr Cu Fe Mn Pb Zn

silt + clay sand silt + clay sand silt + clay sand silt + clay sand silt + clay sand silt + clay sand

Cd

Cr

Cu

Fe

Mn

Pb

0.59a 0.56a 0.94b 0.79b 0.52a ns ns ns 0.57a 0.74b 0.94b 0.84b

0.47a 0.77b 0.81b 0.83b ns ns 0.02 0.70b 0.42 0.71b

ns 0.69b ns ns 0.72b 0.79b 0.97b 0.82b

ns ns ns ns ns ns

ns ns ns ns

0.77b 0.87b

a correlation significant at 0.05 level (2-tailed test). b correlation significant at 0.05 level (2-tailed test).

ns: non significant.

127

128


Figure 2. Metals projection in the first three Principal Components for the silt+clay (a) and sand (b) fraction of sediments.

To further understand the relationships between metal concentrations, reflected by the correlation coefficients discussed above, Principal Component Analysis was performed to the data sets of the metal concentrations in the two granulometric fractions. The three first components of the correlation coefficients matrix, after Varimax rotation, were used to describe the relationships between the variables. The results are presented in Figure 2. The three first components explained 94% of the total variation in the silt+clay fraction and 92% in the sand fraction. In both grain-size fractions, metals were grouped following a similar pattern. The anthropogenic metals Cd, Cu, Pb and Zn have their higher loadings in the first Principal Component, which explains the greater part of the total variance (57% for silt + clay and 64% for sand). So, it can be assumed that the first Component represents the human impact to the system, which is responsible for the greater part of the data variance. It also seems that this human impact, carried out through the urban effluents discharge to the marine environment, is recorded to the coarser as well as to the finer material of the adjacent sediments. The second Principal Component carries the higher loading of Cr and Fe. This second Component represents the natural origin of sediment components. The grouping of Cr with Fe suggests that Cr like Fe is mainly of natural origin. This finding, along with the small extractable percentage of both metals (see below), is supported by previous results from the investigation of surface sediments from the study area (Angelidis and Aloupi, 1997). Finally, the third Principal Component carries the higher loading of Mn, an indication of the special behavior of the metal in the marine environment. Unlike all the other metals, extractable Mn concentrations show a uniform spatial distribution in the studied sediments, whereas these concentrations are similar in both granulometric fractions (see Tables II and III). The lack of a preference of nonresidual forms of Mn for a specific grain-size fraction has been assessed in other


129

Figure 3. Percentage of extractable Cd, Cu, Pb, and Zn from the sand fraction in relation to the mean sand content of the sediments of the inner, outer and ancient harbors and the wide coastal area of Mytilene.

marine environments, where coatings of Mn oxides were found on grains of sand (Loring and Nota, 1973). The acid extractable concentrations in the bulk sediment (silt+clay+sand) were calculated by summing the partial concentrations weighted to the percentage of the relative granulometric fraction in the sediment. Afterwards the sums were compared to the total metal concentrations, to estimate the percentage of the acid extractable, i.e. non-residual, metal. The non-residual character of the metals follows the order: Cd » Pb > Zn > Cu > Mn > Fe Cr. Cadmium appears to be totally extractable and bioavailable (104% ± 17%), in accordance to the literature (Campbell et al., 1988). The other anthropogenic metals had also high extractable percentages (48% ± 12% for Cu, 69% ± 20% for Pb and 55% ± 13% for Zn) and were placed in the first positions of the sequence. On the other hand Fe and Cr were the less acid-extractable metals (12% ± 3% and 11% ± 2% respectively), indicating their association mostly to the lattice of minerals of the sediments. These results indicate that an important part of the metals accumulated in fine, as well as in coarser sediments of the marine coastal environment of Mytilene are potentially bio-available and may be released to the marine trophic chain.

130


The percentage of total metal concentrations (Cd, Cu, Pb and Zn) extracted by 0.5 N HCl from the sand fraction in 4 sub-areas (inner harbor, outer harbor, wider coastal area and station 19 of the ancient harbor), were calculated and are presented in Figure 3. For all metals examined the percentage of total concentrations extracted from the sand fraction increases with the sand content of the sediments, reflecting the increasing relative importance of the coarser fraction in the sediment, as expected. Therefore, since contamination is restricted to the harbor area, the real importance of sand as anthropogenic metal carrier seems negligible. However, this is not the case in the sediments of the ancient harbor, which are considered contaminated by anthropogenic metals (Aloupi and Angelidis, 2000). In the ancient harbor sediments, sand (at the absence of fine-grained sediments) is the major anthropogenic metal carrier (92% of Cd, 41% of Cu, 70% of Pb and 68% of Zn). Thus, although the fine material is efficiently dispersed from the ancient harbor by the high hydrological regime of the area, contamination is still recorded in the sandy sediments.

4. Conclusions The study of the acid extractable metals in the surface sediments of the coastal area of Mytilene revealed the contamination of the harbor sediments by Cd, Cu, Pb and Zn. Although concentrations were higher in the silt + clay fraction than in the sand fraction, percentages of total concentrations extracted from the coarse fraction, especially in the sandy northern harbor, showed that sand is also an important carrier of extractable metals. The high correlation coefficients found between the acid-extractable concentrations of the metals Cd, Cu, Pb and Zn in both grain-size fractions, as well as for each metal in the two granulometric fractions, indicate their common anthropogenic source. Also, it seems that the metal contamination is recorded in both silt + clay and sand fractions of the polluted sediments. It is therefore suggested that coarse sediments should not be disregarded in pollution studies of the coastal zone, especially in areas where an important part of the sediment material may have a texture characterized as ‘sand’.

References Agemian, H. and Chau, A. S. Y.: 1976, ‘Evaluation of Extraction Techniques for the Determination of Metals in Aquatic Sediments’, Analyst 101, 761–767. Aloupi, M.: 1999, ‘Study of the Influence of Urban Discharges on the Geochemistry of Heavy Metals in the Marine Coastal Zone of Mytilene’, Ph.D. Thesis, Department of Environmental Studies, University of the Aegean, Mytilene, Greece, 259 pp, (in Greek). Aloupi, M. and Angelidis, M. O.: 2001, ‘Geochemistry of Natural and Anthropogenic Metals in the Coastal Sediments of the Island of Lesvos, Aegean Sea’, Environ. Pollut. (in press). Angelidis, M. O.: 1995, ‘The Impact of Urban Effluents on the Coastal Marin Environment of Mediterranean Islands’, Wat. Sci. Technol. 32, 85–94.


131

Angelidis, M. O. and Aloupi, M.: 1997, ‘Assessment of Metal Contamination in Shallow Coastal Sediments Around Mytilene, Greece’, Int. J. Environ. Anal. Chem. 68, 281–293. Campbell, P. G. C., Lewis, A. G., Chapman, P. M., Crowder, A. A., Fletcher, W. K., Imber, B., Luoma, S. N., Stokes, P. M. and Winfrey, M.: 1988, ‘Biologically Available Metals in Sediments’, No NRCC 27694, National Research Council of Canada, Ottawa, 298 pp. Chester, R. and Voutsinou, F. G.: 1981, ‘The Initial Assessment of Trace Metal Pollution in Coastal Sediments’, Mar. Pollut. Bull. 12, 84–91. Filipek, L., Chao, T., Carpenter, J.: 1981, ‘Factors Affecting the Partitioning of Cu, Zn and Pb in Boulder Coatings and Stream Sediments in the Vicinity of a Polymetallic Sulfide Deposit’, Chem. Geol. 33, 45–64. Förstner, U. and Wittmann, G. T. W.: 1981, Heavy Metals in the Aquatic Environment, Springer Verlag, Berlin, 486 pp. Gibbs, R. J.: 1977, ‘Transport Phases of Transition Metals in the Amazon and Yukon River’, Bull. Geol. Soc. America 88, 829–843. Goldberg, E.: 1954, ‘Marine Geochemistry I – Chemical Scavengers of the Sea’, J. Geol. 62, 249– 265. Horowitz, A. J.: 1991, A Primer on Sediments – Trace Element Chemistry, Lewis Publ., Michigan, U.S.A., 136 pp. Horowitz, A. J. and Elrick, K.: 1987, ‘The Relation of Stream Sediment Surface Area, Grain Size and Composition to Trace Element Chemistry’, Appl. Geochem. 2, 437–451. Krauskopf, K. B.: 1956, ‘Factors Controlling the Concentrations of Thirteen Rare Metals in Seawater’, Geochim. Cosmochim. Acta 9, 1–32B. Krumgalz, B. S., Fainshtein, G. and Cohen, A.: 1992, ‘Grain Size Effect on Anthropogenic Trace Metal and Organic Matter Distribution in Marine Sediments’, Sci. Total Environ. 116, 15–30. Loring, D. H.: 1992, ‘Factors Controlling the Bioavailability of Heavy Metals in Sediments’, Report for the ICES Working Group on Biological Contaminants, ICES, Copenhagen, 70 pp. Loring, D. H. and Nota, D. J. C.: 1973, ‘Morphology and Sediments in the Gulf of St. Lawrence’, Bulletin 182, Fisheries and Marine Service, Ottawa, 147 pp. Loring, D. H. and Rantala, R. T. T.: 1992, ‘Manual for the Geochemical Analysis of Marine Sediments and Suspended Particulate Matter’, Earth Sci. Rev. 32, 235–283. Luoma, S. N. and Bryan, G. W.: 1981, ‘A Statistical Assessment of the Trace Metals in Oxidized Estuarine Sediments Employing Chemical Extractants’, Sci. Total Environ. 17, 165–196. Robinson, G.: 1982, ‘Trace Metal Adsorption Potential of Phases Comprising Black Coatings on Stream Pebbles’, J. Geochem. Explor. 17, 205–219.