Irene Sammartino

Fingerprinting sedimentary and soil units by their natural metal contents: A new approach to assess metal contamination

One of the major issues when assessing soil contamination by inorganic substances is reliable determination of natural metal concentrations. Through integrated sedimentological, pedological and geochemical analyses of 1414 (topsoil/subsoil) samples from 707 sampling stations in the southern Po Plain (Italy), we document that the natural distribution of five potentially toxic metals (Cr, Ni, Cu, Zn and Pb) can be spatially predicted as a function of three major factors: source-rock composition, grain size variability and degree of soil weathering. Thirteen genetic and functional soil units (GFUs), each reflecting a unique combination of these three variables, are fingerprinted by distinctive geochemical signatures. Where sediment is supplied by ultramafic (ophiolite-rich) sources, the natural contents of Cr and Ni in soils almost invariably exceed the Italian threshold limits designated for contaminated lands (150 mg/kg and 120 mg/kg, respectively), with median values around twice the maximum permissible levels (345 mg/kg for Cr and 207 mg/kg for Ni in GFU B5). The original provenance signal is commonly confounded by soil texture, with general tendency toward higher metal concentrations in the finest-grained fractions. Once reliable natural metal concentrations in soils are established, the anthropogenic contribution can be promptly assessed by calculating metal enrichments in topsoil samples. The use of combined sedimentological and pedological criteria to fingerprint GFU geochemical composition is presented here as a new approach to enhance predictability of natural metal contents, with obvious positive feedbacks for legislative purposes and environmental protection. Particularly, natural metal concentrations inferred directly from a new type of pedogeochemical map, built according to the international guideline ISO 19258, are proposed as an efficient alternative to the pre-determined threshold values for soil contamination commonly established by the national regulations.

Background levels of potentially toxic metals from soils of the Pisa coastal plain (Tuscany, Italy) as identified from sedimentological criteria

Identification of reliable background values of potentially toxic metals in sediments requires detailed integration of geochemical data with accurate sedimentological studies. Through analysis of 60 soil samples from the Pisa coastal plain, this study shows to what extent sediment provenance and facies characteristics may influence the natural distribution of potentially toxic metals (Cr, Ni, Cu, Zn, Pb) within alluvial and coastal sediments. Metals supplied to the alluvial plain are mostly concentrated within the finest sediment fraction (floodplain clays), while coarser crevasse and overbank deposits exhibit invariably lower metal contents. Beach-ridge sands display the lowest metal concentrations. Transport of ophiolitic detritus by the longshore drift may account for locally high Cr concentrations within beach deposits. Geochemical fingerprinting of individual facies associations in terms of natural metal contents results in the construction of a geologically-based geochemical map. This map offers a more reliable depiction of spatial distribution of background levels than interpolation techniques based uniquely upon statistical methods. Matching background values against metal concentrations from topsoil samples leads to the reliable assessment of the pollution status of Pisa coastal plain. Metal contents exceeding the threshold values designated for contaminated areas (Cr) simply reflect catchment geology, and are not the product of artificial contamination. On the other hand, anthropogenic disturbance may be detected even where metal contents (Pb, Cu) lie below the threshold values. The use of sedimentological criteria is presented here as a pragmatic tool to enhance predictability of natural metal contents in sediments, with obvious positive feedbacks for legislative purposes and environmental protection.

Assessing natural contents of hazardous metals in soils by different analytical methods and its impact on environmental legislative measures

A variety of analytical techniques are commonly used to assess natural metal concentrations in soils. Although the choice of the analytical procedure may appear to be an unimportant issue, through an example from the Po Plain we document that for specific hazardous metals the arbitrary choice between separate analytical techniques may have contrasting legislative implications. Comparative analyses of 160 soil samples obtained with X-ray Fluorescence (XRF) and aqua-regia Inductive Coupled Plasma Mass Spectrometry (ICP-MS), respectively, reveal a systematic lower efficiency of chromium determination by ICP-MS (45–60% of XRF determinations), irrespective of the laboratory chosen. Maximum permissible concentrations of chromium for residential and recreational sites were invariably exceeded by sediments deriving from chromium-rich parent rocks after XRF analyses. In contrast, replicate determinations following aqua regia digestion were generally below the threshold values, defining the same soils as unpolluted...

Tracing provenance and pathways of late Holocene fluvio-deltaic sediments by heavy-metal spatial distribution (Po Plain–Northern Apennines system, Italy)

Abstract The bulk geochemistry of 435 near-surface sediment samples from the southern Po Plain was used to identify the major sources of sediment delivered through distinct tracts of the routing system, from the Apenninic catchments to the Po Delta and the Adriatic coast. Sediment composition from the downstream reaches of the Po River and 23 Apenninic channel-levee river systems is fingerprinted by distinctive heavy metal (chromium and nickel) concentrations, which vary primarily as a function of the local ultramafic rock contribution. For any constant provenance domain, fine-grained (floodplain) sediments are invariably enriched in trace metals relative to their coarser-sized, channel-related counterparts, thus reflecting hydraulic sorting by crevasse and overbank processes. Once the geochemical signatures of fluvial endmembers are established, the relative contribution of the individual detrital sources to the downstream segments of the system can be assessed. Using an example from a multisourced supplied system, we outline the reconstruction of source-rock lithology and sediment pathways by combined sedimentological and geochemical studies as the basis for reliable estimates of sediment budgets in a source-to-sink context.

Factors Controlling Trace-Metal Distribution in Alluvial and Coastal Deposits: Implications for Hydrocarbon Exploration

Reliable estimates of sediment transport and storage from a multi-sourced system across a variety of depositional environments require accurate identification of the sediment sources and their role in modulating sediment supply to the down-dip components of the dispersal system. Through a modern example from the Po Plain, involving three segments (drainage area, alluvial plain, and coastal plain/delta) of the system, this study provides insights into the usefulness of a combined sedimentological and geochemical approach, which should also be applicable to the ancient record. Unlike ancient successions, where individual causes of compositional variability are generally very difficult to unravel, recent (late Holocene) sedimentary successions provide an excellent opportunity to understand the relative impact on trace-metal distribution of source-rock composition and changes in grain size, helping to predict whether and to what extent these two variables might be combined or superposed in the rock record. Among the variety of facies associations that compose the sediment routing system, channel-related deposits, for which connection between river transport and hinterland sediment supply can reasonably be inferred, represent the building block for a comprehensive approach to sediment transfer and storage. For trace-metal characterization, levee and crevasse fine to medium sand appears to be more suitable for provenance analysis than does its coarser-grained (bar sand) counterpart. Identification of end members for individual catchment-river systems with sufficient precision may help disentangle provenance mixing from lithologically homogeneous alluvial (floodplain), deltaic (bay, lagoon, delta front), and nearshore (beach) facies associations, for which the relative contributions from individual detrital sources are virtually unknown.