Relationship among parameters of lake polluted sediments evaluated by multivariate statistical analysis

Chemosphere 55 (2004) 1323–1329 www.elsevier.com/locate/chemosphere

Relationship among parameters of lake polluted sediments evaluated by multivariate statistical analysis A. De Bartolomeo, L. Poletti, G. Sanchini, B. Sebastiani, G. Morozzi

*

Dipartimento di Scienze Biochimiche e Biotecnologie Molecolari, Sezione di Scienze Igienistiche ed Ambientali, Universita degli Studi di Perugia, Via del Giochetto, Perugia 06126, Italy Received 26 April 2002; received in revised form 18 August 2003; accepted 22 December 2003

Abstract In the attempt to assess the relationship and interdependency among sediment toxic pollutants, in particular heavy metals, polycyclic aromatic hydrocarbons (PAH), and linear alkyl sulfonates (LAS) and some of the sediment typical components: inorganic carbon (IC), organic material (OM) and acid volatile sulphides (AVS), multivariate techniques of statistical analysis have been applied to a set of chemical data obtained by the analysis of the sediments of the Trasimeno Lake, a central Italy lake characterized by a large surface (128 km2 ) and a low mean depth (about 4.5 m). The results of principal component analysis (PCA) show interrelationships between: OM content and PAH, Pb, and Cu concentrations of the sediments, LAS and AVS, and AVS and IC. The effect of the different sampling periods on sediment composition and contamination level, and the clustering of the sampling sites as a consequence of pollutant load are also shown. The principal component bi-plot of the variables and samples indicates that PAH have the greatest influence on the separation of samples in the different sampling periods.
2004 Elsevier Ltd. All rights reserved. Keywords: Lake sediments; Heavy metals; PAH; AVS; LAS; Principal component analysis

1. Introduction Sediments are an important component of lake ecosystems in which toxic compounds accumulate through complex physical and chemical adsorption mechanisms depending on the nature of the sediment matrix and the properties of the adsorbed compounds (Leivouri, 1998; Ankley et al., 1992; Maher and Aislabie, 1992). The adsorption process, involving a dynamic exchange between the absorbed materials and water, is influenced by several chemical–physical and chemical

*

Corresponding author. Tel.: +39-75-5857332; fax: +39-755857328. E-mail address: [email protected] (G. Morozzi).

parameters like: pH, oxidative–reductive potential, dissolved oxygen, organic and inorganic carbon content. Another important factor is represented by the presence, in water phase, of some anions and cations that can bind or co-precipitate the water-dissolved or suspended pollutants (Di Toro et al., 1991; Calmano et al., 1993; Wen and Allen, 1999). This last factor is indicative of the complicated phenomena involved in the dynamic and fate of the water pollutants not yet completely understood. Some aspects such as the role of the sediments organic material content on the adsorption of non-ionic hydrophobic pollutants (Adams et al., 1995; Van Hattum et al., 1998) and the influence of sulphide concentration on metal ions binding have been clarified (Di Toro et al., 1990). However, metal adsorption is a very complicated phenomenon that can be influenced by all of the above reported chemical and physico-chemical

0045-6535/$ - see front matter
2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.chemosphere.2003.12.005

1324

A. De Bartolomeo et al. / Chemosphere 55 (2004) 1323–1329

parameters as well as by the bacterial communities living in the sediments, which, through oxidation/reduction reactions, may alter the valence state of the metals affecting their release from the sediment matrix (Burton, 1991). Multivariate data analysis techniques can be used to assess the complex eco-toxicological processes by showing the relationship and interdependency among the variables and their relative weights. In particular, the principal component analysis (PCA) has been widely used in eco-toxicology (Simpson et al., 1998; Sparks et al., 1999) since it provides the most relevant information about the structure of multivariate data and the relationships among the objects considered for a general overview of the problem. Information on the role of different variables can also be obtained, indicating those that are strictly correlated giving overlapping information. In the latter case, such variables can be eliminated in the subsequent analyses. In this research, multivariate techniques were used to analyse a set of data obtained from the study of Trasimeno Lake sediments. This laminar lake, located in central Italy, has a large surface area (128 km2 ) in relation to the size of the basin (396 km2 ), and a long water retention-time (24.4 years); it is shallow (4.5 m of mean depth) because the high evaporation rate (155 · 106 m3 /year), the water over-exploitation for the agriculture and the low precipitation levels (700–800 mm/year) in the basin (Regione dell’Umbria, 1992). The partially treated sewage effluent, the transport of raw materials with the run off from the hydraulic basin, and the sedimentation of dead aquatic vegetation in late autumn form thick beds of settled material in some lake areas.

2. Materials and methods 2.1. Sampling A specimen for each season from five representative points of the lake (Fig. 1) was sampled for a total of 19 samples, considering that it was not possible to collect the sample in the summer in site B. All the sample (0.5 kg) were a pool of four sediments first layer (upper 5 cm) samplings collected by a modified Ekman sampler in the same day at different hours starting at 9.30 a.m. up to 3.30 p.m. Four of the sampling sites (A, B, C, E) were located about 50 m offshore and one was in the middle of the lake (site D). Three of the coastal points were situated near the most populated areas (A, C, E) and the fourth (B) in an area particularly rich in aquatic vegetation. The collected sediments were placed in glass vessels, previously washed with pure HNO3 and rinsed with ultra-pure water (Millipore), and refrigerated at )20 C until analysis. The following parameters were determined: acid volatile sulphides (AVS), organic material (OM), inorganic carbon (IC), cadmium, chromium, nickel, lead, copper, linear alkyl sulfonates (LAS) and polycyclic aromatic hydrocarbons (PAH). 2.2. Acid volatile sulphides Acid volatile sulphides are solid phase sulphides soluble in HCl. The analysis was carried-out according to the method reported by Di Toro et al. (1991). Briefly, the solid phase sulphides are converted to H2 S adding cold HCl 6 N. H2 S, carried by a nitrogen flow, trapped

Fig. 1. Location of sampling sites in the Trasimeno Lake. A (Long. 12110 0500 ; Lat. 43090 0500 ); B (Long. 12110 1800 ; Lat. 43050 4800 ); C (Long. 12040 4500 ; Lat. 43060 4700 ); D (Long. 12060 5600 ; Lat. 43080 1400 ); E (Long. 12030 0700 ; Lat. 43100 1900 ).

A. De Bartolomeo et al. / Chemosphere 55 (2004) 1323–1329

in a solution of AgNO3 , and the precipitated Ag2 S is filtered, dried and weighed. The method, standardized with FeSO4 and Na2 S solutions in three repeated experiments, gave an average H2 S yield of 88% ± 6 with respect to the theoretical value. 2.3. Heavy metals analysis One gram of sediment was dried at 105 C for 1 h, finely ground and digested in a 50 ml flask with a metal free HNO3 :HCl (1:3) solution (Rauret, 1998). For the digestion procedure and analysis the method reported by the Water Research Institute of the National Research Council (1985) was employed with minor modifications. Briefly, the digestion was carried out at the boiling point for 2 h, by swirling every 10 min. Successively, the cooled samples were centrifuged at 8000 rpm to eliminate the inert solid particles, transferred to a 25 ml bottle and diluted with ultra-pure water. 20 ll were analysed by a Perkin–Elmer mod. 560 Atomic Absorption Spectrophotometer equipped with a HGA 400 graphite furnace. 2.4. Heavy metals leaching test The method employed was that reported by the Water Research Institute (1985). Briefly, two grams of sediment, treated as above for the metals analysis, were added to a volume of ultrapure water 16 times higher than their weight in a previously HNO3 -washed and ultra-pure water rinsed vessel. The mixture was continuously stirred by a magnetic stirrer for 24 h at room temperature. The acidity was adjusted to pH 5 with acetic acid (0.5 M) and periodically checked to maintain constant the pH value. At the end of the test, the aqueous phase was separated by centrifugation at 8000 rpm and then analysed for the presence of the metals both in the aqueous and in the solid phase. A blank was treated in the same way of the sample except the sediment was not added. 2.5. PAH analysis The method employed was fundamentally that reported by the Water Research Institute of the National Research Council (1990). An aliquot (5.0 g) of wet sediment was centrifuged at 8000 rpm in a hot cleaned (250 C for 1 h) and preweighed centrifuge tube. The water phase was carefully separated and the residue was gently dried at 35 C in the dark under N2 stream for 24 h and stored overnight in a desiccator. Before the analysis, the centrifuge tube containing the dried residue was re-weighed, and added with 1-methyl pyrene, 2,20 -binaphthyl and indeno-

1325

(1,2,3,-cd) fluoranthene as internal standards. The residue was mixed with anhydrous Na2 SO4 and quantitatively transferred to the glass extraction vessel of the Soxhlet apparatus. After extraction carried out overnight with dichloromethane, the activated copper dust was added to remove sulfur. Successively, the solvent was evaporated by rotary evaporator to 2–3 ml, transferred into a 5 ml conical flask, reduced to the volume of 200–300 ll under N2 stream at 30 C, and chromatographed together with a PAH standard solution on a 20 · 20 cm Silica gelchromatographic plate, 0.25 cm thickness, eluted by benzene–hexane (1:1) mixture to obtain a sharp fluorescent band (Rf ¼ 0:75) containing the purified PAH fraction. The band was scraped and transferred to a 5 · 1 cm column with a fitted glass septum and then eluted by the addition of 2 + 2 + 2 ml of CH2 Cl2 . The solvent was gently evaporated to dryness under N2 stream in a 5 ml conical flask, the residue dissolved in 50 ll of toluene, and 1 ll injected ‘‘on-column’’ in a Fisons gas-chromatograph equipped with a SPB1 HR column (30 m · 0.32 mm; film thickness 0.25 lm) and a FID. The analysis conditions were the following: carrier gas (H2 ) at flow rate 0.8 ml/min. Temperature program: 90–180 C at 15 C/min and 180–270 C/min at 7.5 C/ min. Detector temperature 300 C. The chromatographic peaks, identified by their retention-times, were successively confirmed by GC–MS analysis carried-out on a HP 5890 series II gas-chromatograph equipped with a mass-spectrometer detector HP 5971A. All the solvents used in the analysis were FID grade (Baker). 2.6. LAS analysis Five grams of dried and ground sediment were mixed with anhydrous Na2 SO4 and extracted overnight in Soxhlet by a toluene/methanol mixture (3:2). The extract was gently evaporated to dryness and added with 50 ml of ultra-pure water. After 30 min sonication to allow the solubilization of MBAS in water, the solution was filtered through a 0.45 lm Sartorius membrane to separate and discard the insoluble substances; the procedure was repeated twice rinsing the flask with 25 + 25 ml of ultra-pure water. The three filtrates were joined together, transferred to a separatory funnel, and analysed as reported in the Standard Methods (APHA, 1985). 2.7. Organic material and inorganic carbon analysis The organic material (OM) was evaluated by oxidation with K2 Cr2 O7 in acid medium (concentrated H2 SO4 ) in the presence of Ag2 SO4 as a catalyst. An aliquot of 2.5 g of dried sediment, 20 ml of K2 Cr2 O7 2 N, and 26 ml of concentrate H2 SO4 were placed in 200

The raw data obtained during the sampling period only give information on the level of sediment contamination by toxic substances (Table 1) in the different sampling sites and in the different sampling periods considered, and moreover on the low levels of metals released from the solid matrix (Table 2), these last also

WA AE AD AC AB AA

3. Results and discussion

Table 1 Concentrations of parameters in different seasons and sampling sites

Principal component analysis (PCA) was performed on the whole data set. The cluster analysis plot was carried out to describe the association of samples and the principal component variable loading plot was computed. In the latter the vector lengths indicate the variability associated with the single variable, and the cosine of the angle between the vectors reflects the degree of correlation between variables. The bi-plot shows the samples as points and the variables as vectors in an attempt to correlate the information obtained from the loading plot analysis with those from the principal component analysis.

WB

2.8. Principal component analysis

The first letter indicates the season (A ¼ Autumn, W ¼ Winter, Sp ¼ Spring, S ¼ Summer) and the second the sampling sites (A, B, C, D, E). The heavy dense growth of the aquatic vegetation made it impossible to sample the sediment at site B during the summer.

SE SD SC SA SpE SpD WC

WD

WE

SpA

SpB

SpC

where B is the volume of FeSO4 Æ 7H2 O used to titrate the blank, B1 is the volume of FeSO4 Æ 7H2 O used to titrate the sample, N is the FeSO4 Æ 7H2 O solution normality, P is the amount of sample analysed, 0.003 is the carbon equivalent weight. For the evaluation of the Inorganic Carbon, a sample aliquot (5 g) was burned at 400 C for 3 h to completely eliminate the organic material, placed in a desiccator for 1 h, and weighed. Successively, the same sample was burned at 950 C for 3 h, placed in a desiccator for 1 h and weighed to determine the amount of the Inorganic Carbon by the difference between the two weights, taking into account the ratio between the molecular weight of carbon dioxide and carbon.

0.12 34.5 39.1 0.4 80.4 24.0 2.3 17.3 4.6 63.2

OM ¼ ðB  B1Þ  N  200=20  0:003  1000=P

0.49 1.46 123.5 98.6 78.7 152.6 0.9 0.6 110.0 79.8 27.2 22.5 0.9 9.1 21.2 16.9 11.8 32.3 138.7 203.8

ml cylindrical flask with a 100 mm neck, closed by a glass stopper with an inserted thermometer. The flask was rapidly warmed at 160 C and maintained at this temperature for 10 min. After cooling, the sample was transferred to a 200 ml flask, diluted to 200 ml with ultra-pure water, and left to settle. To 20 ml of the clear solution, previously transferred to a 250 ml Erlenmeyer flask, 8 ml of concentrate H3 PO4 and few drops of 4diphenylamino sulfonate indicator were added and the excess of K2 Cr2 O7 was titrated with FeSO4 Æ 7H2 O 0.2 M. A blank was carried-out in the same conditions except the addition of the sediment sample. The OM content was calculated by the following equation:

0.47 0.84 0.18 0.64 0.04 0.09 0.17 0.07 0.49 0.06 0.11 0.23 0.18 0.02 0.03 0.16 99.0 152.7 139.5 81.7 20.1 85.2 118.2 121.1 132.0 79.2 328.0 122.2 223.7 15.1 196.8 168.1 53.8 56.7 67.5 212.4 34.2 36.6 77.9 87.4 100.9 24.8 28.6 71.1 87.0 32.4 44.0 60.7 0.5 0.7 1.1 2.1 0.9 1.1 1.8 0.5 1.1 0.4 0.8 1.1 0.8 0.4 0.5 0.6 105.4 118.5 63.3 77.1 27.7 76.4 69.5 36.8 67.4 52.5 77.1 79.8 50.1 26.3 84.1 68.1 36.1 38.1 22.6 69.4 12.8 30.2 31.5 54.4 37.5 25.3 65.2 57.3 17.6 15.8 18.1 19.3 3.6 2.0 2.5 15.6 1.3 1.9 2.1 11.0 11.8 3.4 6.7 7.8 8.3 4.5 6.4 6.8 30.3 44.7 35.4 23.4 11.9 29.6 35.8 31.5 17.4 13.3 25.4 31.8 29.6 11.6 43.8 30.7 22.8 8.5 24.6 21.7 6.8 21.7 7.5 13.3 13.2 4.2 6.0 15.0 4.3 3.2 6.0 4.0 104.2 176.1 205.1 196.3 116.5 164.5 146.7 286.4 294.9 281.2 289.9 362.5 303.0 233.0 398.0 368.0

A. De Bartolomeo et al. / Chemosphere 55 (2004) 1323–1329

AVS (g/kg) OM (g/kg) IC (g/kg) Cd (mg/kg) Cr (mg/kg) Ni (mg/kg) Pb (mg/kg) Cu (mg/kg) LAS (mg/kg) PAH (lg/kg)

1326

A. De Bartolomeo et al. / Chemosphere 55 (2004) 1323–1329 Table 2 Percentages of metals released from the sediment samples Toxic metals

Sampling sites A

B

C

D

E

Cd Cr Ni Pb Cu

28.5 2.5 6.9 10.6 6.1

29.9 1.6 4.8 11.4 11.9

20.9 1.4 9.1 11.4 10.3

28.0 1.7 6.7 11.4 6.7

24.3 1.8 5.3 5.7 5.0

indicative of their low mobility. To obtain more reliable information about the relationships among the variables, the eventual clustering of the samples and the seasonal effect, the principal component analysis was applied. Three components, whose eigenvalues were higher than 1, accounted for a cumulative variance of 69.7% (Table 3). The effect of different sampling periods and the differences among some of the sampling sites are well shown by the PCA analysis results (Figs. 2 and 4). The autumn and winter samples, with positive score along the principal component 2, are well separated from spring and summer samples that show a negative score. About the sampling sites, site C, characterised by negative scores in the first component, is well separated from the others because of the low pollutant load and in clear opposition to site B, which shows the highest

Table 3 Eigenvalues, total and cumulative % of variance in the factor analysis of sediment loading with chemical parameters over sampling periods Factor

Eigen values

% total variance

Cumulative eigenvalues

Cumulative %

1 2 3

2.94 2.23 1.79

29.4 22.3 17.9

2.94 5.18 6.97

29.4 51.8 69.7

Position B PCA - scores of component 3

SC

Position C Position A

WC

WB

SpA

SpB

SpC SE

SA

AB

AC

SpE WE

WA SpD

SD WD

Position D

AA AD

W=Winter Sp=Spring S=Summer A=Autumn

AE PCA - scores of component 1

Fig. 2. Clusters of the sampling sites by PCA. Plot of principal component 1 vs 3.

1327

pollutants concentrations, especially in autumn and in winter (Table 1). The apparent sandy nature of the sediments of site C, along with the relatively high organic material content of site B, could explain these results. Each of the other sampling sites (A, D, E) is not well separated from the others, so suggesting that the sampling period may be, in this case, more important than the site location. The cluster analysis of the variables (Fig. 4) shows the relationship between the organic material (OM) content and polynuclear aromatic hydrocarbon (PAH) concentration in the sediment samples. This confirms the results reported in the simple correlation matrix (Table 4) and agrees with the literature data concerning the role of the organic fraction of the sediments in binding the hydrophobic compounds (Adams et al., 1995; Van Hattum et al., 1998). Another important relationship concerns the linear alkyl sulfonates (LAS) and the acid volatile sulphides (AVS) (Table 4, Fig. 2). This kind of relationship cannot be considered a simple association; on the contrary, when the mechanism of AVS formation is considered, a causality correlation becomes evident. LAS, in fact, can be completely degraded into sulphates (Van Guikel, 1996; Sigoillot and Nguyen, 1997), which are successively reduced to sulphides by sulphate-reducing bacteria (Thamdrup et al., 1993) and precipitate as iron or manganese salts to form AVS. In a previous study (Regione dell’Umbria, 1992) these metals were found in sediments from Trasimeno Lake at concentrations of 2–3 g/kg for iron and of 0.560–1.518 g/kg for manganese. The relationships among toxic metals and the other variables is not so clear. Ni and Pb are associated with IC, and in a lesser extent with AVS, while copper with organic material (Fig. 4). If the variable loadings of components 1 and 3 are plotted (Fig. 3), an interesting association may be observed between lead and PAH, that could indicate the predominant anthropogenic origin of PAH. The pointed out association between PAH and Pb on one hand, and between Cu and OM content on the other, can be explained when the anthropic activities in the lake basin are considered. The complete absence of industrial plants, if some copper handcraft activities are excluded, and the contemporarily presence of small or medium sized urban centres, together with agricultural activities, although not intensive, represent the characteristics of the area that can be mainly defined tourist and recreational. It can be then hypothesised that PAH and Pb derived both from urban liquid wastes in which exhausted oils and/or other combustion process derivatives were discharged and not completely removed by biological treatment plants. The association between Cu and OM can be similarly explained by the inlet in the urban sewage of incompletely purified wastes from copper factories. It should also been observed that the occurrence of Cu in the organic fraction of sediments

1328

A. De Bartolomeo et al. / Chemosphere 55 (2004) 1323–1329

Table 4 Correlation coefficient for the indicated variables AVS OM IC Cd Cr Ni Pb Cu LAS PAHtot *

AVS

OM

IC

Cd

1

)0.03205 0.643779 1 )0.05109 1



Cr

)0.12157 )0.11665 )0.06151 1

Ni 

0.502636 0.27297 0.130436 0.061579 1

Pb

Cu

LAS

0.124841 0.270471 )0.02304 0.317371 0.209589 0.509314  0.403031 0.732339 )0.02974 0.047597 )0.15807 0.102918 0.240177 )0.16623 0.468982 1 0.548786 0.175661 1 )0.05202 1

PAHtot 

0.628847 )0.15102 0.563559 0.30823 0.291695 0.233703 0.237888 0.056861 1

)0.20828 0.539557 )0.00429 )0.17256 )0.20787 0.13452 0.498969 0.270996 )0.24656 1

p < 0:05. p < 0:01.

AB

AVS LAS IC

Cd Cr

Pb

OM

PAH

Cu

PCA - scores of component 2

PCA - loadings of component 3

**

WC

AC

WD AA

WE SpC

Cd

SE

Ni

has been widely reported (Loska and Wiechula, 2003). However, the lack of clustering between the heavy metals and the typical sediment components (AVS and IC), seems to confirm the complicate behaviour of these pollutants, that can be influenced by many factors (Burton, 1991). The low levels of metals release from the sediments in acid medium (pH ¼ 5), apart the considerations concerning their mobility, could lead to the conclusion that the metals are bound to the solid matrix not only as sulphides or carbonate but also in more stable forms. The metal adsorption onto iron and manganese oxide in the sediments has been widely investigated (Burton, 1991; Wen and Allen, 1999). Since it is unlikely that the redox potential of the sediments is high enough to allow the co-precipitation of metals with iron or manganese oxides or in some other oxidized form, it cannot be excluded that the oxidative process may take place during sampling, storage and pre-treatment for the metal-release test. This could cause the initial metal release with the successive precipitation in a more stable form. This hypothesis is supported by the fact that the oxidation and the release of sulphide-bound metals with the subsequent re-adsorption onto oxyhy-

WA

WB

Cu

AE SpB Cr Pb AVS Ni LAS IC SpE SpA SA

SD SpD

PCA - loadings of component 1

Fig. 3. Principal component variable loading plot. Component 1 vs 3.

AD PAH OM

SC

W=Winter Sp=Spring S=Summer A=Autumn

PCA - scores of component 1

Fig. 4. Principal component bi-plot of the variables and sampling sites.

droxides seems to occur in a very short time, about 1 h (Hirst and Aston, 1983). The bi-plot (Fig. 4), in which the samples from the different seasons are reported together with the chemical parameters, shows that spring and summer samples are associated with the OM and PAH, while autumn and winter samples are more closely associated with the IC, AVS and LAS concentrations. These findings indicate that the OM and PAH on one hand, and the AVS, IC and LAS concentrations, on the other, seem to be the parameters most responsible for the separation of the sampling sites with respect to the sampling periods. About this last point, however, a longer more in depth analysis is needed to confirm the associations and form reliable hypothesis.

4. Conclusions The analysis of raw data or a simple correlation analysis gives a limited amount of information about the dynamics and interdependency of the variables impli-

A. De Bartolomeo et al. / Chemosphere 55 (2004) 1323–1329

cated in the chemical and toxicological characterisation of lake sediments. Multivariate techniques of statistical analysis seem to provide an important tool for a better understanding of the complex dynamics of pollutants in this very complex matrix.

Acknowledgement We are grateful to the Local Public Administration (Provincia di Perugia––Italy) for the financial support.

References Adams, W.J., Kimerle, R.A., Mosher, R.G., 1995. Aquatic safety assessment of chemical sorbed to sediments. In: Cardwell, R.D., Purdy, R., Bohner, (Eds.), Aquatic Toxicology and Hazard Assessment: Seventh Symposium. STP 854. American Society for Testing and Materials, Philadelphia, PA, pp. 429–453. Ankley, G.T., Lodge, K., Call, D.J., Balcer, M.D., Brooke, L.T., Cook, P.M., Kreis Jr., R.G., Carlson, A.R., Johnson, R.D., Niemi, G.J., et al., 1992. Integrated assessment of contaminated sediments in the lower Fox River and Green Bay Wisconsin. Ecotoxicol. Environ. Saf. 23 (1), 46–63. APHA, AWWA, WPCF, 1985. Standard Methods for the Examination of Water and Wastewater, 16th ed. Burton Jr., G.A., 1991. Assessing the toxicity of freewater sediments. Environ. Toxicol. Chem. 10, 1585–1627. Calmano, W., Hong, J., Forstner, U., 1993. Binding and mobilization of heavy metals in contaminated sediments affected by pH and redox potential. Water Sci. Technol. 28 (8–9), 223–235. Di Toro, D.M., Mahony, J.D., Hansen, D.J., Scott, K.J., Hicks, M.B., Mayr, S.M., Redmond, S.M., 1990. Toxicity of cadmium in sediments: the role of Acid Volatile Sulfide. Environ. Toxicol. Chem. 9, 1487–1502. Di Toro, D.M., Zarba, C.S., Hansen, D.J., Berry, W.J., Swartz, R.C., Cowan, C.E., Pavlou, S.P., Allen, M.E., Thomas, N.A., Pasquin, R.P., 1991. Technical basis for establishing sediment quality criteria for non-ionic organic chemicals by using equilibrium partitioning. Environ. Toxicol. Chem. 10, 1541–1583. Hirst, J.M., Aston, H.A., 1983. Behaviour of copper, zinc, iron and manganese during experimental resuspension and

1329

reoxidation of polluted anoxic sediments. Estuar. Coast. Shelf Sci. 16, 549–558. Leivouri, M., 1998. Heavy metal contamination in surface sediments in the Gulf of Finland and comparison with the Gulf of Bothnia. Chemosphere 36, 43–59. Loska, K., Wiechula, D., 2003. Application of principal component analysis for the estimation of source of heavy metal contamination in surface sediments from the Rybnik Reservoir. Chemosphere 51, 723–733. Maher, W.A., Aislabie, J., 1992. Polycyclic aromatic hydrocarbons in nearshore marine sediments of Australia. Sci. Total Environ. 112 (2–3), 143–164. Rauret, G., 1998. Extraction procedure for the determination of heavy metals in contaminated soil and sediment. Talanta 46, 449–455. Regione dell’Umbria, 1992. Piano per la Gestione ed il Controllo del Bacino del Lago Trasimeno Finalizzato al Contenimento dell’Eutrofizzazione. Rapporto Finale. Deliberazione Giunta Regionale, 19 Maggio 1992, n. 3405. Sigoillot, J.C., Nguyen, M.H., 1997. Complete oxidation of linear alkybenzene sulphonates by bacterial communities selected from coastal seawater. Appl. Environ. Microbiol. 58 (2), 1308–1312. Simpson, C.D., Harrington, C.F., Cullen, W.R., Brigth, D.A., Reimmer, K.J., 1998. Polycyclic aromatic hydrocarbons contamination in marine sediments near Kitimat, British Columbia. Environ. Sci. Technol. 32, 3266–3272. Sparks, T.M., Scott, W.A., Clarks, R.T., 1999. Traditional multivariate techniques: potential for use in ecotoxicology. Environ. Toxicol. Chem. 18 (2), 128–137. Thamdrup, B., Finster, K., Wurgler-Hansen, J., Bak, F., 1993. Bacterial disproportion of elemental sulphur coupled chemical reduction of iron and manganese. Appl. Environ. Microbiol. 59, 101–108. Van Guikel, C.G., 1996. Complete biodegradation of xenobiotic surfactants by consortia of aerobic microorganisms. Biodegradation 7 (2), 151–164. Van Hattum, B., Pons, M.J.C., Montanes, J.F.C., 1998. Polycyclic Aromatic Hydrocarbons in fresh water isopods and field-partitioning between abiotic phases. Arch. Environ. Contam. Toxicol. 35 (2), 257–267. Water Research Institute (IRSA), 1985. Analytical Methods for Sludge. Physico-Chemical Parameters, IRSA-CNR, Rome. Water Research Institute (IRSA), 1990. IPA––Idrocarburi Policiclici Aromatici. Determinazione in GC/FID previa estrazione e purificazione, Q 64, no. 25, IRSA-CNR, Roma. Wen, X., Allen, H.E., 1999. Mobilization of heavy metals from Le An River sediments. Sci. Total Environ. 227, 101–108.