2.4. Sediment sampling and analysis
Sediments samples were collected in the Control and FTW ponds 6 times over a 2-year period. Prior to sediment collection a combined Platinum-Ag/AgCl ring electrode linked to a laboratory redox meter (826, Metrohm) (Hinchey and Schaffner, 2005 and Rodríguez et al., 2010) was inserted into the sediment to a depth of ∼2 cm (detailed methodology is provided in supplementary data). 208 mV were added to the reading to calculate the redox potential relative to hydrogen electrode (Eh). Subsequently, grab sediment samples (2–3 cm of the top layer) were collected in polypropylene (PP) containers and kept in a cooler on ice until arrival at the environmental laboratory of the University of Auckland where samples were stored at 4 °C. Limited access from the side banks into each pond meant that only two and four samples were collected during the first and second missions, respectively, for the Control pond, and six samples were collected during the first mission in the FTW pond. Access was subsequently resolved, so eight or nine samples were collected in each pond during the remaining missions. Zn and Cu concentrations were determined as per Jumbe and Nandini (2009) which provides a total digestion of the samples including the residual form generally associated with silicate lattices. Weight lost on ignition (LOI) which corresponds to the content of organic matter of the sample was determined by ashing 2–4 g of sample at 500 °C for 12 h (Chagué-Goff, 2005 and Wang et al., 2010). A three step extraction which enables quantification of Cu and Zn as 1) exchangeable/carbonate form, 2) fraction bound to hydroxides, and 3) bound to organic compounds and sulphides, was performed based on “procedure A” of Ianni et al. (2001). The residual (detrital) fraction was calculated as the difference between the total metal content and the sum of the metals released by the sequential extraction. In contrast to Ianni et al. (2001), the sediments used were not dried before the extraction to avoid changes in chemical speciation which might occur during the drying process (Chagué-Goff, 2005). The equivalent of about 1 g of dried sediment was used for the initial extraction. The extracts from the sequential extraction and the total digestion procedures were analysed for Cu and Zn by flame or graphite (depending on concentration range) AAS (Varian SpectrAA-50/55). One method blank and one field sample duplicate were analysed for Cu and Zn for each total digestion and sequential extraction step (RPD ranged from −19.8 to 19.4% with a median of 1.4%). Detailed methodology is provided in supplementary data.