Trace element pollution in Italian volcanic soils: the case study of the Solofrana river valley

The Solofrana river valley is an agriculture based area of southern Italy characterised by a huge concentration of tanning plants (~160). Despite the recently observed decline in Cr content in river water (Adamo et al., 2001) the soils from the valley still contain this and other trace elements in concentrations above usual levels, mainly as a consequence of the long use of the polluted river waters as a source of irrigation, of the frequent river overflow as well of the widespread use of metal-rich agricultural materials. The high metal retention properties of the volcanic soils of the valley might play a key role to maintain the elements in soil for a long time. A suspect that an increase of soil contamination could cause severe damage to crops suggested soils have to be supervised (Adamo et al., 2003).A multidisciplinary study, taking into consideration pedoenvironmental, micromorfological, chemical, hydrological, biological and microbiological analyses, was carried out in 2001 restricting attention to overflooded soils. Crossing in a GIS environment different2 land information (soil, hydrology, flood and land use maps) three study sites of about 700 m with the same soil type, Humic Haplustand (Soil Survey Staff, 1998), not cultivated by 5 years and affected by different numbers of floods were selected: 1) E1, flooded in 1981; 2) E2, flooded in 1981 and 1993; 3) E3, flooded in 1981, 1993 and 1998. At the E3 site, flooded again in 2002, soil and sediments were sampled soon after the event as E4 site. A site without flooding (C site) as control was also individuated. The selection of not cultivated sites was produced in order to avoid the influence of agriculture treatments over microbe activities, hindering the influence of the trace metals.Pedological profiles were opened on the two extreme cases: P1, the not polluted control site (C site), and P2, the site subjected to 3 flood events (E3 site). Soil sampling was carried out following different operative methods and at different depths, according to the analyses to be performed.-1Soils were classified as sandy-loam (clay content 127-205 g kg ), with neutral-subalkaline pH according with carbonates presence. All the soils had a good amount of organic C (23-52 g kg-1), total N (2.2-4.1 g kg-1) and available P (19-45 mg kg-1). In general, soil properties showed some grade of variability, especially in E3 and E4 soils, probably because of the disturb caused from the recent overflowing events.Copper and Chromium were the main soil contaminants: their total content in many cases was higher than the limits established by the current Italian legislation (DL 92/99: Cu 100 mg kg-1; DM 471/99: Cu 120, Cr 150 mg kg-1). The distribution of total content among studied soils suggested for Cu and Cr different sources. Only Cr accumulation was related to overflowing events. Cu accumulation was most likely attributable to past agricultural use of fertilizers and pesticides.Sequential chemical extractions indicated for all polluted soils preferential association of Cr and Cu with oxidizable forms, whereas in non polluted soils both elements mostly occurred in residual mineral forms of a silicate and oxide nature. For both metals the soluble and exchangeable forms made always a small contribution to the total. Significant amounts of Cr and Cu were recovered in the acid ammonium oxalate extraction, suggesting association of46metals with short-range-order aluminosilicates and organo-mineral complexes. Cr(III) isotherm adsorption curves obtained at pH 6 were adequately described by the Langmuir equation and showed significative difference between control and flodded soils. High adsorption capacity of Cr(III) ions from aqueous solutions on the control and flooded E3 soil was observed. A significant presence of short-range ordered mineral phases, shown by selective chemical extractons and by the x-ray diffractometry of the clay fraction, was likely involved in soil metal retention.Total microbial biomass and fungal mycelium were generally higher in soil E1. Similarly, in this soil was observed the greatest bacterial diversity, as compared to soils E2 and E3. In terms of bacteria community similarity (Sørensen index) the studied soils could be ranked as follow: C ? E1 ? E2 ? E3. Only some of the measured enzymatic activities (acid phosphatase, arylsulphatase, ?-glucosidase, dehydrogenase) decreased with the number of flooding events, whereas others (urease and FDA hydrolase) were not changed or increased. In some cases the control soil showed lower enzymatic activities than those of flooded soils. The high metal content of soils did not appear to have a predominant influence on the studied biological and biochemical parameters possibly as a result of both low metal mobility and high C content. Infact, soil E4 sampled immediately after a new flooding event, although its high Cr content (536 mg kg-1), had higher biological and biochemical activities. The large input of Cr from the sediments (1012 mg kg-1) was masked by their high levels of organic carbon (61 g kg-1).Soil hydraulic properties and solute transport behaviour, determined by the Wind's method and by a miscible flow experiment on P1 and P2 profiles, confirmed the high Cr(III) retention capacity of the soil with high values of dispersivity and showed occurrence of preferential flow paths. The analysis of the soil water retention curves showed a decrease of large pores in the flooded soil compared to control. Frequent occurrence of clay and silt coatings along elongated pores in the surface and subsurface soil horizons was showed by optical microscopy observations. This suggested a possible transfer of metal-rich sediments along the soil pore network during water movement.