*3.2. Samples Pretreatment*

All labware used in the analytical procedures were soaked in 10% HNO3 for at least 48 h, and then thoroughly rinsed with deionized water. Precautions were taken during all stages in order to avoid contamination.

The water samples were filtered under a clean laminar bench. The filtration was carried out in succession through 8 μm and 0.45 μm Millipore filters in order to separate the dissolved (<0.45 μm) and the particulate phase (>0.45 μm) and determine the suspended particulate matter (SPM) in two grain-size fractions (0.45 μm < diameter (d) < 8 μm and d > 8 μm). The filtrates for the determination of dissolved metals were acidified to pH < 2 by HNO3 prior to pre-concentration on Chelex-100 (100–200 mesh, BioRad Laboratories) and subsequent elution of trace metals [27].

The filters for particulate metals (0.45 μm < d < 8 μm and d > 8 μm) were dried to constant weight and then digested with concentrated HNO3 on a hot plate (85 ◦C approximately) overnight [27].

Sediment samples were freeze-dried in a Labconco apparatus. Then the silt and clay fraction (<63 μm) was separated by means of sieving. This approach has been widely used to reduce the physical variability of metal content due to grain size effects [28–30], although recently its effectiveness has been questioned, particularly if coatings of metal scavengers exist in coarse-grained particles [31]. Total metals contents were determined after complete dissolution of sediments with an acid mixture of HNO3-HClO4-HF acids [32]. The 0.5 N HCl extractable metals were determined after the method of Agemian and Chau [33].
