The history of the Venice Lagoon contamination is presented here, based on sediment records and on information about industrial activities and past management choices. We used polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) as tracers since the relative abundance of their congeners can help in locating their sources. The most important PCDD/F source to the canals of the first industrial area (built in the 1920s) can be ascribed to processes involving pyrite to obtain sulfuric acid and recovery of copper and other metals. Indeed, homologue profiles and documents show that the PCDD/Fs contamination is mainly related to the use and disposal of both these industrial wastes and materials dredged from the most contaminated canals. In particular, these latter constitute the ground under the second industrial area (built after World War II). Furthermore, hazardous materials were stored until 1992 in an islet exposed to erosion, whereas tracts of industrial canal banks eroded until 2000. Similar situations are frequent in transition and coastal areas worldwide.
The Venice Lagoon (Figure 1) is a shallow aquatic ecosystem located along the north-eastern Italian coast. Its morphology is characterized by a network of canals of various depths, mud flats, tidal marshes and islets.
In time, many human interventions have deeply changed the natural evolution of the lagoon in order to preserve its unique environment. In particular, the diversions of rivers Adige and Brenta to the sea were realized between the 15th and the 16th centuries with the purpose of preventing the silting of navigable areas. As a consequence, fluvial sediment inputs decreased, thus enhancing the marine characteristics of the lagoon environment. In the early 20th century, after World War I, the development of economic activities on the mainland led to the construction of the first industrial area of Porto Marghera (Figure 1). Other activities deeply changed the lagoon environment during the 20th century, such as the building of the trans-lagoonal bridge in 1934, the construction of the second industrial area after World War II, and the excavation of the navigation canals Vittorio Emanuele III and Malamocco Marghera, completed in 1930 and 1969, respectively.
The lagoon has long suffered from the combined effects of natural and anthropogenic influences, worsened by increased hydrodynamics and illegal intensive clam fishing (which involves dredging of the sediments to sift out the clams). These factors have increased erosion with an annual net loss of 800,000 m3 of sediment (Sarretta et al., 2010; Rolinski and Umgiesser, 2005).
A number of authors have studied dated sediment cores to understand the history of this environment, with a special emphasis on pollution trends. A series of papers (Marcomini et al., 1997; Cochran et al., 1998; Frignani et al., 2001a,b, 2005; Bellucci et al., 2000, 2002; Dalla Valle et al., 2005) have provided the necessary information about the distribution of contaminants such as metals, polychlorinated biphenyls (PCBs), polychlorinated dibenzo-p-dioxins and dibenzofurans (PCDD/Fs) and polycyclic aromatic hydrocarbons (PAHs) in sediments of the canals of the industrial area and in the lagoon mud flats. The most recent papers have shown high levels of contaminants within the industrial canals, the major sources being those facilities located within the first industrial area, which has released untreated effluents since the early 1920s. Furthermore, it was possible to assess the prevailing PCDD/F sources within the system (Bellucci et al., 2000; Frignani et al., 2001a): combustion for octachloro dibenzo-p-dioxin (OCDD), vinyl chloride monomer (VCM) production for octachloro dibenzofuran (OCDF) and metal recovery for the most common “chlorine fingerprints” (mainly furans).
The purpose of this work is to demonstrate how most of the contamination presently found in lagoon sediments was spread through the system by virtue of past management choices. This is done by coupling scientific evidence from sediment cores with historical information. We use PCDD/Fs as tracers since previous papers have underlined the informative potential of the distribution of these contaminants (Swerev and Ballschmiter, 1989; Hagenmaier et al., 1994; Cleverly et al., 1997). Other contaminants, in particular As and Al, are considered to provide further hints to support our results.
Materials and Methods
We have reviewed a series of published papers (Fattore et al., 1997; Cochran et al., 1998; di Domenico et al., 1998; Marcomini et al., 1999; Bellucci et al., 2000, 2002, 2007; Frignani et al., 2001a,b, 2005; Dalla Valle et al., 2005) and documents issued by public authorities (e.g. Regione Veneto, Consorzio Venezia Nuova, Magistrato alle Acque). Moreover, a limited number of analyses, including some metals derived from specific industrial processes, were performed on samples from erodible layers of canal banks.
Figure 1 shows the Venice Lagoon with sampling locations. Sediment cores E1 and M3 represent a site near the Tresse Islet and a salt marsh of the northern lagoon, respectively. The former was sampled in May 1996 using a manual piston corer, and the latter in August 1998 by insertion of a Plexiglas tube into the marsh sediment. Cores were immediately extruded to obtain 2–4 cm thick sections, with greater detail at the top, and the outer part was discarded to avoid smearing effects. Each slice was put in labeled glass vessels and kept frozen until the analyses of organic micro-contaminants. A separate core was taken at each location for the determination of metals, grain size, organic carbon and radiotracers. These cores were extruded and sectioned at intervals of 2 cm; the slices were put in clean, previously weighed, polyethylene vessels. Core C9 was collected in May 1996 from the Brentelle canal, whereas cores 1 and 5 were obtained in May 1999 from the Nord industrial canal, together with surficial samples (1–3 cm) from erodible banks of the canals Sud (ES1, ES2, ES3, ES4 and ES5) and Ovest (ES6).
The analysis of PCDD/Fs was carried out using established HRGC-HRMS methods (U.S. EPA 1613/94/revision b) for the determination of 17 congeners 2,3,7,8 substituted. Sediment samples were freeze-dried, Soxhlet-extracted (U.S. EPA 3540), and then analyzed through separation and detection procedures. Efficiencies of these analytical steps were evaluated using 13C labeled internal standards. Details of the analytical procedure were provided by Bellucci et al. (2000). Toxic Equivalent Quantities (TEQs), which sum the seventeen 2,3,7,8-substituted congeners on the basis of their toxic equivalence to 2,3,7,8-TCDD (tetrachlorinated dibenzo-p-dioxin), were calculated from PCDD/F concentrations using the international toxic equivalent factors (I-TEFs from NATO/CCMS, 1998). The composition of PCDD/F mixtures was calculated as permil (‰) contribution of each homologue group with respect to the total.
Metals (Al and As) in erodible banks were analyzed by atomic absorption spectrometry (AAS) after total dissolution (Bettiol et al., 2008). Accuracy was checked by analysing the Standard Reference Material NIST 2710 (Montana Soil). Uncertainties were 2 and 7% for Al and As, respectively. Standard deviations on repeated measures were lower than 10% for both metals.
Sediment chronologies were assessed through activity-depth profiles of excess 210Pb (210Pbex), obtained by alpha counting of 210Po, under the assumption of secular equilibrium between the two isotopes. 210Po was extracted from the sediment using hot HNO3 and H2O2, plated onto silver discs and counted with a surface barrier detector. 137Cs was measured by non-destructive gamma spectrometry. All concentrations and activities refer to dry weight.
Results and Discussion
Levels of PCDD/Fs in the Venice Lagoon and the effect of environmental regulations
Sediment chronologies derived from 210Pbex and 137Cs activity depth profiles were discussed by Frignani et al. (2001a,b, 2005) and Bellucci et al. (2007). Figure 2 (top) shows the pattern of PCDD/F contamination recorded by the salt marsh core M3. Cochran et al. (1998) and Bellucci et al. (2007) demonstrated that these sediments can be significant recorders of the atmospheric inputs of contaminants. It is evident that PCDD/F concentrations reached the highest values between 1940 and 1970, and then decreased significantly. This pattern was attributed to a strong reduction of inputs, due to the closure of the most polluting activities within the first industrial area (Frignani et al., 2001a). The homologue downcore distributions show OCDD, attributed to combustion processes, as the most important congener at depth (around 1920s, not shown in Figure 2, top). Moving upward, furans become predominant, with the typical proportion (the so called “chlorine fingerprint”; sample M3(12–15) in Figure 2) that is the index of prevailing industrial contamination. At surface (sample M3(0–1.5), Figure 2, top) OCDD is again the most abundant congener. This trend toward a proportionally higher presence of OCDD in recent lagoon and salt marsh sediments means a higher contribution from combustion, which may be attributed to urban sources, after the reduction of the industrial inputs. This pattern is also consistent with the evolution of the environmental regulations in the Venice area, in particular, an Italian law, approved in 1976 (n. 319), that regulated the composition of industrial effluents. Finally, in 1998, a Ministerial Decree, referred also as “special law for Venice” and further revised in 1999, established very strict limits for industrial effluents to lagoon waters, banning the discharge of several toxic, persistent and bioaccumulative contaminants (As, Cd, PCDD/Fs, PAHs, Hg, Pb, PCBs, chlorinated pesticides, cyanides and tributyltin compounds, TBTs). During this period, several wastewater treatment plants were also put into service.
Sources of PCDD/Fs to the Venice Lagoon
The strong gradient of PCDD/F concentrations observed between the two industrial districts (64130-98 ng kg−1 as TEQ, moving from the first to second industrial area, respectively) demonstrates that the most important sources were located within the first industrial area, where many years of uncontrolled discharges led to a very high contamination of the canals Nord, Brentella and Salso (Bellucci et al., 2000; Frignani et al., 2001a). This trend is true not only for PCDD/Fs, but also for metals (Bellucci et al., 2002) and polychlorinated biphenyls (Frignani et al., 2001b). Montobbio and Bernstein (2000) confirmed this pattern, finding the maximum concentrations of all contaminants in the bottom sediments of canals Brentella, Nord, and in part of the Ovest. This means that sediment contamination dates back to the beginning of activities within the first industrial area.
The search for the sources of PCDD/Fs, starting from available published data, was particularly challenging. Fattore et al. (1997) and di Domenico et al. (1998) first identified two kinds of homologue profiles in lagoon sediment samples (3S and 1S, Figure 1) and associated them to as many different sources: the former (3S, Figure 3a), composed mainly by OCDD and attributed to untreated domestic sewage and combustion, dominated in sediments of urban areas, while the latter (1S, Figure 3b), containing mostly furan congeners with a certain amount of OCDD, was widespread all over the lagoon and was associated to the production of VCM in the 2nd industrial area. However, Bellucci et al. (2000) and Frignani et al. (2001a) found that only the sediment samples of the Lusore-Brentelle canal (where VCM plants discharged without any treatment until the 1980s), were characterized by the almost exclusive presence of OCDF among furans, with a variable content of OCDD (surficial level of core C9, Figure 3c). It followed that this was to be considered the PCDD/F composition typically marking VCM plant effluents, as shown by a VCM stripping profile (sample PU4043, Figure 3d; Stringer et al., 1995) and confirmed by Isosaari et al. (2000) in sediments from the Gulf of Finland. Therefore, VCM production in the second industrial area was not responsible for the contamination of the whole Venice Lagoon, as it appears to have affected only a restricted area in the Lusore-Brentelle canal. Thus, an additional PDCC/F source, associated with other industrial activities, had to be identified to explain the pattern displayed by lagoon sediments. Reviewing all available data, the following features were observed: (1) each sample from the first industrial area showed high PCDD/F concentrations with the already mentioned “chlorine fingerprint,” widespread all over the lagoon (see marsh core M3 at 13 cm depth, Figure 2 [top], and sample 1S in Figure 3), and (2) there was a relationship between PCDD/Fs and As vertical profiles in canal cores 1 and 5 (Figure 4; data from Magistrato alle Acque di Venezia (MAV) and Port Authority of Venice 1999). These results suggested that furans originated from a process taking place in the fist industrial area and were somehow connected to As inputs. The likely “culprit” was then identified in the production of sulfuric acid by the roasting of pyrite, where As is contained as an impurity: during this process, pyrite ashes are treated with NaCl (10%) and then roasted at 450–550°C with the production of CuCl2 and FeCl2 (Mariani, 1972), which are known to catalyze the formation of PCDD/Fs. The resulting roasted residue is then leached for the recovery of Cu and, in lower amounts, Au and Zn. In turn, the precipitated iron oxide (the so called “purple ore”) is agglomerated in tiles through a sintering process which could be responsible for PCDD/F formation due to the presence of Cl and C from the previous phases. These plants have been in operation since 1927, were still working in 1968 and have been definitively closed in the early 1970s, in perfect agreement with PCDD/F chronologies recorded by salt marsh core M3 (Figure 2, top). Thus, they can be considered the major source of contaminants to the lagoon. A similar case regarded a German industry which discharged a red siliceous slag (Kieselrot), containing high amounts of PCDD/Fs (up to 64.5 μg kg−1 as TEQ) resulting after the extraction of Cu from a black mineral using the same industrial process described above (Theisen et al., 1993).
The responsibility of past management choices in the contamination of the Venice Lagoon
The use of industrial wastes and dredged material from industrial canals for marsh land filling has been documented since the late 1920s. This practice was continued until the 1970s, to reach average thickness of 2.3–3 m above sea level (Regione Veneto, 2000; Bernstein and Bonsembiante, 2000). In particular, during the 1920s and 1930s, the infilling residues came from the distillation of coke and the productions of glass, sulfuric acid, phosphate fertilizers and pesticides. During the late 1930s and 1940s the discharged waste materials originated from the industries of aluminum, zinc and ammonia, along with some sludges produced by thermoelectric plants (Regione Veneto, 2000; Bernstein and Bonsembiante, 2000). The industrial canals Sud and Ovest were excavated in the 1960s through those embankments of industrial wastes. Consequently, all these interventions contributed to the diffusion of contaminants across the lagoon. Another bad management choice, which contributed to the worsening of the situation, consisted in leaving extended parts of the above mentioned canal banks unprotected by walls for a long time, causing a systematical erosion that dispersed contaminated mud within the canals and into the lagoon. To gain insight for our historical reconstruction, we sampled such banks, including the collection of both black layers (dominant in sample ES5, Figure 1, and containing residues of pyrite ashes), and red ones (samples ES1, ES2, ES3, ES4 and S6, Figure 1, probably originating from bauxite treatment for Al production). Results showed that black layers are enriched in As (up to 10,400 mg kg−1), whereas the red ones are characterized by high Al concentrations (11,300–43,700 mg kg−1) and, rather surprisingly, by elevated content
s of PCDD/Fs (particularly sample ES2, 3230 ng kg−1 TEQ) with the same furans’ profiles, widespread all over the lagoon and originating within the first industrial area (1S, Figure 3b). The problem of canal bank erosion was faced, and resolved, by the Regione Veneto in 2000 with the approval and realization of a master plan for the complete confinement of banks in the whole area of Porto Marghera, which was declared of “National Interest.”
Another indirect source of contaminants to the lagoon was identified in wastes stored in the Tresse Islet since the 1930s. Like the industrial canals Sud and Ovest, the unconfined banks of the islet eroded for years, thus releasing contaminated mud into the lagoon (Montobbio and Bernstein, 2000). The Tresse Islet was completely sealed from 1993 to 1996 by a reconstructed perimeter dike system. Different engineering solutions have been adopted to reduce the wave action due to ship traffic and to prevent the transfer of contaminants from soils and groundwater aquifers to the lagoon. The sealed perimeter dikes resulted in a large area suitable for placement of new material dredged from canals and channels (Palermo and Averett, 2000).
To complete the scenario of poor past management choices, it is worth mentioning the role of the Nord Industrial canal dredging, carried out following urgent procedures related to navigation purposes (minimum draft 9 meters). In these cases, the dredging was carried out without a prior characterization of the amount of sediment to be removed, and without a monitoring plan during and after the completion of work. The effects of these activities can be observed in the PCDD/F vertical profiles of Figure 4 (cores 1 and 5, MAV and Port Authority of Venice 1999) and Figure 2, bottom (core E1; Frignani et al., 2001a). The record shown by core 5 is complete, whereas that of core 1 appears truncated due to the dredging of the topmost sediment (likely up to 200 cm, Figure 4), thus exposing a contaminated sediment layer at the water-sediment interface. In turn, PCDD/F homologue profiles at E1 (Figure 2, bottom) show the “chlorine fingerprint” that originates within the first industrial area (Bellucci et al., 2000), confirming that this lagoon site has always been influenced by the accumulation of material resuspended within the Nord Industrial Canal. In addition, PCDD/F concentrations in core E1 display a decrease in contamination dated after the suspension of dredging operations (year 1987). Nevertheless, the terminal part of the Nord Industrial canal (corresponding to the “darsena Fincantieri”) was further dredged in 1997 to allow the passage of a large vessel, and this may explain a less significant contaminant decrease at site E1 than at other lagoon sites (Frignani et al., 2001a).
These examples show how different sources and processes operating in the past were responsible for the contamination of the Venice lagoon, originating from Porto Marghera industrial areas. In addition, the lack of regulation and environmental protection caused the inappropriate use of contaminated materials, the spreading of contaminants due to erosion of unconfined banks, and the resuspension during dredging operation without a prior physical-chemical characterization.
In 2001 maintenance dredging operations in highly contaminated Venice canals were suspended, thus causing a natural silting with the consequent reduction of water depth. Consequently, in 2004, the Port Authority of Venice issued a decree to reduce the draft for ships transiting the Malamocco-Marghera canal from 9.60 m to 9.14 m.
This particular situation, in which environmental issues (no availability of disposal sites for contaminated sediment) overlap with socio-economic ones (i.e. a severe decrease of merchant shipping), had forced several local institutions, the Port Authority, and the Regione Veneto to ask for a declaration of the state of emergency. In this situation, the estimated amount of sediment to be dredged ranged from about 5 million to more than 9 million m3 for scenarios of −9.5 m and −12 m, respectively: the most contaminated sediments, that cannot be disposed inside the lagoon without prior treatment, ranged from 1,067,128 m3 and 1,709,220 m3, respectively (CCPV, 2006).
In summary, the activities arranged for the solution of the emergency have considered (CCPV, 2011):
1) the increase of the total capacity of the Tresse Islet for contaminated sediments, through the construction of embankments up to +9.50 m and the creation of new confined nearby areas;
2) the setting up of plants for the treatment of highly contaminated sediments and their disposal in a confined area besides the Nord Industrial Canal (Molo Sali, for not dangerous wastes) or in the landfill Vallone Moranzani (both for dangerous and not dangerous wastes) (Figure 1);
3) a dredging project of the Malamocco-Marghera canal up to −11 m, with continued maintenance to ensure the intended depth and consequent disposal of dredged sediments;
4) the dredging of portion of the Ovest Industrial Canal and the Sud Industrial Canal up to −10.5 m.
These activities are presently in progress. Different dredging technologies are used according to the material to be removed. Turbidity curtains are used to prevent the spreading of materials in the surrounding area during dredging.
The history of the Venice Lagoon management should provide a warning, mostly addressed to those developing countries where intense industrialization and economic growth have not reached their peaks yet, and where the environmental regulation on sediment management is still lacking. Presently, in those countries, it is not unusual to use dredged sediments for land reclamation without a previous characterization of their potential threat to the environment and human health. In these cases, specific studies aimed at an environmentally sound characterization of dredging sediments are strongly recommended.
Funds were provided by the Projects “Sistema Lagunare Veneziano,” “Orizzonte 2023” and by EniChem S.p.A. The authors thank F. Colombo for his valuable assistance, C. Carraro and S. Raccanelli for PCDD/F analyses. This is contribution No. 1777 from the Istituto di Scienze Marine, Bologna.