Faktorer som styr skillnader mellan totalhalter och lösta halter metaller i ett antal svenska ytvatten

Results of the analysis of metals in the Swedish environmental monitoring refers to the acid soluble content of metals in the water phase. According to the definition of dissolved components in water only metals in the dissolved form (“Melöst“) pass a 0,45 µm filter while metals in the particulate form (“Mepart”) are trapped in the filter. Results from the Swedish environmental monitoring program therefore covers both forms of the metals. Because only the dissolved phase is considered to be directly bioavailable the current method is bound to overestimate metal content. To estimate the proportion of metals present in particulate form a study was performed during which the proportion metals in dissolved form (“Mefrac” = “Melöst”/”Metotal”) were examined. Metals lead (Pb) , cadmium (Cd) , cobalt (Co) , zinc (Zn) , nickel (Ni) , copper (Cu) , arsenic (As) , vanadium (V), chromium (Cr) and aluminum (Al) , manganese (Mn) and iron (Fe ) were studied. Results from 31 stations around the country as well as a smaller number of samples in different lakes around the country are included. This report studied I) how metals differ in their behavior, II ) how to estimate the level of metals in dissolved form (“Melöst”) based chemical variables or chemical models , III) how different catchments differ, IV) it was tested if land can affect how metals are distributed at finally it was tested V) whether a classification of sites with high metal would be affected if one only had access to the total concentrations instead of of dissolved metals. I) While metals Cu, As, Ni, Cd are mainly (>85%) in the dissolved phase ( “Melöst”) both Zn, Pb , V, Cr and Co can be bound to particles and are thus partly trapped in a 0,45 µm filter . The proportion of metals in particulate form (“Mepart“) increases with increasing pH, total metal concentration, turbidity, percentage of particulate iron and manganese and conductivity while it decreases with lower pH, and high values for both filtered absorbance ( abs420_F) and TOC. pH is the variable that alone explains the largest proportion of the variation. Reliable estimates of “Mefrac” thus require reasonably constant pH with a small margin for error and little variation over time. In catchments with higher pH variation especially in the pH range from 5.5 to 7.0 where the effect is greatest, estimates of “Mefrac” will have larger uncertainty. II) Equations have been developed to calculate the proportion of metals in dissolved form in which other known chemical parameters are used as input. Based on total concentrations these models allow to estimate the proportion of metals in particulate form within reasonable margins for almost all metals. The regressions for metals in dissolved form (“Melöst”) that have been developed in this report seem to be fairly universal for all studied areas. The models explain between 70-98 % of the observed variation. For the chemical modeling is not the case. The chemical models perform much worse in antropogenically affected areas . The reason for this may be that the formation of other particles than ferrihydrite (Fe (OH )3) in these waters that bind up metals like lead . Because the fields are limed it is close to assume that calcite or aragonite could conceivably occur. The chemical patterns that determine the presence of metals in dissolved or particulate form is for most metals in accordance with the model calculations. The presence of particles such as clay or dead organisms is probably the greatest source of error for the discrepancy between the calculated and measured conditions. For some metals, good correlations between “Mefrac” of metals and “Mefrac” of Al, Fe or Mn were observed. This may in some cases be used to obtain better estimates of the proportion of particulate metal. This applies primarily to the metals Pb, Cr and Co and could be accomplished by simply measuring the proportion of Fe, Al and Mn in particulate form. III) This analysis reveals that there are significant differences for “Mefrac” for the different metals and the different areas. There exist no simple clear pattern among the 31 different areas that were included in this study. In acidic waters in the catchments Lill-Fämtan, Svartberget, Lommabäcken, Ringsmobäcken it may be noted that almost all metals are in the dissolved form (“Mepart” near zero ). In the more alkaline and larger areas such as stations Dalälven Älvkarleby, Stockholm järnvägsbron , a much greater proportion metals is in particulate form. Even in affected areas the proportion of metals in particulate form increases (example of these areas are Nissan, Emån, Forsån). Larger areas usually have higher pH values as the contributions of well-buffered groundwater is greater. This also applies to lakes with longer residence times. Areas with a high proportion of arable land also exhibit higher proportion of metals in particulate form (te.g. Holen).This is presumably related to a larger mobilization of particles than in the other streams. IV) Land use explains up to 50 % of the variation of “Mefrac” for Zn and Fe but less than 30% for the two metals (Pb , Co) which have the highest proportion of metals in particulate form. The variables that contribute significantly ( + increasing proportion metals in dissolved form , - declining proportion metals in dissolved form ) to the models for “Mefrac” for metals Fe , Zn, Pb and Co are percentage of lakes in the area (-) , drainage area size (-) , percentage of logged forest (-) , proportion of arable land (-) and percentage of forest (+ ), volume pinus contorta (+ ), volume pine (+) and percentage of forest outside marshlands (+). Factors that control the residence time in the water or proportion of particles increases the proportion metals in particulate form , while the presence of forest leads to a higher proportion of metals in dissolved form. V) The fraction of metals in particulate form has a significant impact on the assessment in areas where elevated metal concentrations occur. Around 6117% of the median values for metals in the 28 different stations that are not affected by point sources are 10 times above background levels. When just considering the dissolved proportion of metals with the same criterion only 36 approx. 11% of the median values valuessites exceed the background levels 10 times. In areas with high concentrations of metals compared to background levels (e.g. 10 times above and higher) and in areas with known point sources, it is suggested to measure the dissolved from of metals more regularly.