Map showing the Arno river catchment and its major tributaries, the largest towns, and the sampling stations subdivided according to the sampling year. Coordinates refer to UTM32T (ED50) system. 

Contexts in source publication

Context 1

... equivalent of 1.4 10 inhabitants. In addition, the Bisenzio river receives domestic untreated waste waters from the Vaiano locality and those from the northern part of Florence through the Macinante Canal. All the domestic water contributions to the Ombrone river are treated. On the whole, TDS values increase downstream from 290 to 499 mg/l in the Bisenzio river and from 332 to 1057 mg/l in the Ombrone river (Cortecci et al ., 2002). (3) The Usciana river and some minor tributaries re- ceive discharges (95% depurated) from many tanneries in the Santa Croce and Pisa areas, which corresponds to a pollution load of 3.2 × 10 6 inhabitants. In this same area, there are also numerous paper-mills which release their depurated waste waters to the Usciana tributary. TDS values in the Usciana river increase from 309 mg/l to 2179 mg/l roughly 40 km before the confluence with the Arno river (Cortecci et al ., 2002). Upstream from Florence, pollution in the Arno river is basically related to contributions from the Chiana river, which receives waste water from galvanic plants process- ing gold in the Arezzo district and untreated effluents from intensive agricultural-zootechnic activities along the Val di Chiana. The galvanic plant caused a mercury anomaly in the stream sediments in the area, as reported by Dall ’ Aglio (1971). At about 30 km from the confluence, TDS values increase (from 552 to 736 mg/l) and then slightly decrease (684 mg/l). The main conclusions of the hydrogeochemical sur- veys of the waters from the Arno catchment performed by Bencini and Malesani (1993) and Cortecci et al . (2002) are (1) Chiana, Bisenzio, Ombrone and Usciana tributaries considerably influence the chemical composition of the Arno river, contributing valuable amounts of Na ( ± K) chloride and sulphate of anthropogenic origin, (2) Sieve, Pesa, Elsa, Egola and Era tributaries dilute Na and Cl into the Arno river by adding waters rich in Ca ( ± Mg) and HCO 3 (Sieve, Egola and Pesa) and SO 4 (Elsa and Era). The chemical composition of the latter tributaries appears to be basically determined by the lithology of the feeding sub-basins. The mineralogy and heavy metal content of the suspended and bedload sediments from the main river Arno and its tributaries were investigated by Bencini and Malesani (1993). The mineralogy mainly consists of quartz (9 to 44%), feldspars (7 to 32%), calcite (1 to 38%) and clay minerals (11 to 80%), the latter being represented by vermiculite, chlorite, illite, kaolinite. Analysed metals in these sediments (Cu, Pb, Zn, Ni, Cr, Mn) should be predominantly trapped within the mineral structures (92 to 99%), especially in the clay minerals, and to a lesser extent, in metal-organic complexes (0.6 to 8%) and adsorbed on clay particles (0.1 to 0.4%). Sampling campaigns were conducted during low-flow conditions, to ensure maximum recovery of the sediments. Most of the bed sediments were sampled by dredging the centre of the rivers. When cobbles and boulders were present on the river bed, sediments were directly collected by hand in the active channel. Sediments were wet-sieved in the field to <80 mesh (177 μ m), according to the method of Rose et al . (1979). 500 grams of sediments were collected and analyzed for each site. The location of the sampling sites is reported in Fig. 1 and Table 1, subdivided according to the period of sampling. Major and trace elements were determined by X-ray fluorescence spectrometry on pressed powered pellets using a Philips PW 1480 automated spectrometer following the methods of Franzini et al . (1972, 1975), Leoni and Saitta (1976) and Leoni et al . (1982) for matrix corrections. Long term reproducibility for major elements was generally better than 7%, whereas for trace elements, it was on average better than 10%. Absolute accuracy relative to certified values or International Reference Material was generally within the reproducibility range (e.g., Dinelli et al ., 1996). Analytical homogeneity between batches was checked by duplicate analysis of selected samples and found to be better than 5%. Mineralogical analyses were performed with a Philips PW 1130 (Cu K α radiation Ni filtered) by pressing powders into alumina holders in order to avoid preferential orientation of sheet-silicates. Corrections for differential intensity response were made according to the methods proposed by Cook et al . (1975) to provide a semi-quanti- tative estimation of mineral abundances based on XRD scans. The analyses were performed on the samples collected during the 1997 survey, using the same powders prepared for the chemical analysis. Organic carbon and nitrogen were determined in the specimens collected in 1997 using a Carlo Erba EA 1110 CHNS analyser, with a reproducibility better than ± 0.1%. Carbonate was removed from the sediments with hydro- chloric acid within a Ag-capsule and then heated at 70 ° C. The relationships among the various chemical elements were investigated through hierarchical cluster analysis (complete linkage as cluster method, Pearson correlation as similarity measure) using a SPSS ® statis- tical package. The range of composition of the stream sediments sampled in November 1996 and June – July 1997 for the Arno river and its tributaries is reported in Table 2, along with the chemical composition of additional sediments from selected tributaries (Elsa, Era, Ombrone and Bisenzio) sampled in July 2000 (the complete data set is available upon request from the authors). Element distribution in stream sediments is expected to show slight seasonal variations (Chork, 1977), and this can be evaluated through the relative standard deviation of the mean of the measurements (Birch et al ., 2001a, b). The repeated sampling at the same station in the Arno catchment allowed the recognition of systematic variations of the chemical composition for only six stations (A17, A15 and A11, Arno; IX, Sieve; V, Pesa; II, Usciana). The differences were ascribed to marked textural difference, as indicated by the large relative standard deviations of SiO 2 (up to 12%), Al 2 O 3 (up to 16.5%) and Zr/ Rb (up to 56%), despite the fact that the sediments had been sieved to reduce the uncertainty related to grain size (see Birch et al ., 2001b, for a discussion of the problem). Textural differences are related to different hydrological conditions during the sampling periods, namely the re- duction of flow discharge particularly evident in the sum- mer 1997 survey, which produced a relative enrichment in the fine-grained fraction of the sediments in some stations. However, in the following discussion of the spatial distribution of the elements, mean values from repeated sampling will be used. The main mineral phases in the sediments were mica, chlorite, sheet-silicates, quartz, feldspars, plagioclases and calcite (Table 3). In some samples, smectite, kaolinite, dolomite were also present, whereas halite was only observed in the sample A02 at Pisa. According to the lithological distribution in the drainage basin, sheet silicates are abundant in the tributaries of the southern flank: Era, Egola, Elsa all contribute to relatively large quanti- ties of these minerals, but the largest contributor is the Chiana. In addition, Ombrone and Sieve, two northern tributaries, transport large amounts of sheet silicates. The Arno river does not seem to be greatly influenced by all these inputs, but sheet silicates were observed in the sediments at station A19 (Laterina), which is located upstream of a dam, at station A12 (Ponte a Signa) and at station A02 (Pisa ...

Context 2

... as sink and/or carriers for a large number of pollut- ants that can either be transported away from source areas according to fluvial dynamics or be stored in the solid fraction of the bed sediments. In such case, sediments can become a source of pollution if the environmental conditions change within the sedimentary column or in the river course, and if the solids are removed and re- suspended. In some cases, “natural pollution” can over- print the anthropogenic pollution, the former being related to the weathering of ore deposits or to the occurrence of peculiar types of rocks. The present work focuses on the bulk chemistry of active bedload sediments from the Arno river catchment in northern Tuscany (Italy). The samples were collected both from the main river and from its principal tributaries during low flow conditions in November 1996 and June–July 1997, and in July 2000, for some selected tributaries (Elsa, Era and Usciana). The analyses were performed for the inventory, survey, assessment and monitoring of metal pollution, and for geochemical mapping to delineate the spatial distribution of chemical elements. The results are compared to available geological data in order to evaluate the effects of lithology on sediment com- position and to suggest background values. The role of organic matter in controlling heavy metal distribution is also discussed. The Arno river is 242 km long and its catchment covers an area of 8228 km 2 . The river flows into the Tyrrhenian Sea about 10 km downstream from Pisa (Fig. 1). Florence, the biggest town of Tuscany, is situ- ated about 135 km from the headwaters of the Arno river. Other towns include Arezzo, located in the upper section of the river, and Pistoia and Prato, located north of Florence in the catchment area of the Ombrone river and Bisenzio river, respectively, both right-hand tributaries of the Arno river. Sedimentary rocks dominate the Arno river catchment (Fig. 2). Sandstones (Cervarola and Macigno Formations) are widespread in the headwaters of the Arno river and near the northern tributaries in the upper catchment, such as Salutio and Sieve. Marls, clays and minor sandstones (Pliocene and Pleistocene marine, continental and lacustrine deposits) occur in the Chiana valley, the most important tributary of the Arno river in its upper section. Sandstones, shales and calcareous turbidites interbedded with chaotic clays along with scattered ophiolitic blocks are the prevailing terrains in the central part of the basin around Florence. Clastic deposits occur in the plain area between Florence and Pistoia, which consisted of a lacustrine basin during the late Pleistocene (Capecchi et al ., 1975; Bossio et al ., 1993). Downstream from Florence, the right-hand tributaries (Bisenzio, Ombrone and Usciana) drain an area composed of sandstones belong- ing to the Macigno Formation and limestones and cherty limestones. The left-hand tributaries (Greve, Egola, Pesa, Elsa and Era) flow over Plio-Quaternary fine-grained marine and lacustrine clayey and sandy deposits. Triassic limestone and gypsum-anhydrite along with Messinian gypsarenite occur in the headwaters of Elsa and Era rivers which are also characterized by the occurrence of ophiolitic rocks, mainly serpentinites. At the end of the basin, north of Pisa, metamorphic rocks (marbles and metapelites) occur in the Monti Pisani area, on the right- side of the Arno river. The coastal plain around Pisa is composed of a graben filled by alluvia. In the Arno river catchment, the water discharge variations follow those of the precipitations, although with a slight delay, since the majority of the basin is composed of low-permeability rocks. Historical records of discharge indicate a distribution with two maxima (December and March), with a major minimum in August. As reported by the Consorzio Pisa Ricerche (1998), major anthropogenic inputs in the Arno river occur downstream from Florence. They originate from: (1) The direct discharge of the whole Florence city domestic black and white water system, with a pollution load equivalent to 10 6 inhabitants. This increases COD from 10 to 30 mg/l, N-NH 4 from trace to 8 mg/l and phosphates from 0.1 to 1 mg/l. (2) The Bisenzio and Ombrone tributaries carry waste waters from industrial settlements and nurseries around Pistoia and Prato. Pollution load due to waste waters (about 80% depurated) from textiles in the Prato area ...