Preindustrial regional background levels of metals in the sediment in waters surrounding Stockholm

In order to classify the quality of aquatic sediments, a set of "natural" or "comparative values" is required, which can be expected for systems without any human impact. However, it is a well established fact that these "natural", rather geologically derived background values are not identical for an entire country, due to variations in bedrock geology. For this reason, the "natural background" concentration of metals must also differ from region to region. In this context it is difficult to determine regional background concentrations for metals in areas that have been exposed to contaminants released from human activities over long periods of time. Moreover, natural processes, such as bio-turbation, sediment resuspension and elevation of deposition bottoms, may overlap and confuse the build-up of stable sediments. Thus, sampling and analysis of the metal content at a certain sediment depth might not automatically provide unbiased data on "pre-historical" or natural levels of metals.

This dilemma can be resolved by looking for metals in the sediment, which exhibit very low concentrations at a certain sediment depth, and which are known at the same time to be typical anthropogenic contaminants, but untypical for the geological formation of the study area. This would strongly indicate that the sediment layer in consideration is uninfluenced by human activities. Consequently, the concentrations of other trace elements could also be interpreted as "natural background levels". Examples of such indicator elements, to be used for selecting sediment samples with natural background concentrations of other trace metals, could be mercury and cadmium.

The Swedish SQCs for both freshwater and marine sediments establish, as a basis for sediment quality classification, a set of "natural" or "comparative values" that are to be expected if there was no human impact. These "natural" or "original" metal concentrations are set as uniform levels for the whole country (and its coastal waters). However, it is a well established fact that these rather geologically derived background values are not identical for the entire country, due to the varying bedrock geology. For this reason, the "natural background" concentration of the different trace metals must also vary from region to region.

It is sometimes argued that it is very difficult to determine the regional background concentration of trace metals in sediment, at least in areas that have been exposed to contaminants released from human activities over very long periods of time. Moreover, a number of natural phenomena, such as bioturbation, sediment resuspension and elevation of deposition bottoms, may confuse the regular build-up of stable sediments (compare section 5.4.6). Thus, sampling and analysis of the metal content at a certain sediment depth might not automatically provide unbiased data on "pre-historical" or natural levels of metals.

A possible way of resolving this dilemma may come from the fact that there are metals not necessarily related to a specific geological formation. Therefore, if certain trace elements, known to be typical anthropogenic contaminants and, at the same time untypical for the geological formation in the study area, exhibit very low concentrations at a certain sediment depth, this may be a strong indication that the sediment layer in consideration is uninfluenced by human activities, and that the concentrations of other trace elements at the same time could be interpreted as "natural background levels". Typical candidates of such indicator elements, to be used for selecting sediment samples with natural background concentrations of other trace metals, could be mercury and cadmium.

This approach has been tried on a material presented by Ostlund et al. (1998), covering trace metal data, obtained by means of digestion and analysis according to Swedish Standard, from altogether 394 samples taken from 117 sediment cores collected in the close surroundings of the city of Stockholm, both from Lake Malaren, the inner archipelago (Baltic Sea) and from some smaller lakes within the borders of Greater Stockholm. The criteria for selection of samples from the database were:

S the deepest layer analysed in each specific sediment core (usually at sediment depths of more than 40 cm, but occasionally at around 30 cm depth);

S LOI (loss on ignition) in general >7% (mostly in the range 7-10%, in a few small lakes up to 49%);

S total concentration of Cd <0.4 ^g/g DS and total concentration of Hg<0.1 ^g/g DS.

The justification for choosing these cut-off values for Cd and Hg was a previous set of nation-wide trace metal background concentrations in lake sediments, proposed by SEPA in 1991.

Altogether 15 sediment samples fulfilled the above criteria, and these were examined with respect to the concentrations of a certain number of trace metals, (see Table 5.16).

Table 5.16. Possible regional background concentrations of some trace metals in uncontaminated sediment layers from 15 sediment cores, collected in the Stockholm area. Concentrations are based on digestion of samples in 7 N nitric acid (Swedish Standard) and are expressed as ^g/g dw (from Landner, 2002).

Table 5.16. Possible regional background concentrations of some trace metals in uncontaminated sediment layers from 15 sediment cores, collected in the Stockholm area. Concentrations are based on digestion of samples in 7 N nitric acid (Swedish Standard) and are expressed as ^g/g dw (from Landner, 2002).

Trace metal

Conc

. range

Corrected* range

Arithmetr. Mean

Median

Cadmium

0.10

- 0.35

0.22

0.20

Mercury

0.009

-0.074

--

-

0.033

0.036

Copper

34

- 47

34 -

42

38

37

Chromium

40

- 61

40

- 58

52

52.5

Nickel

27

- 40

27

- 37

33

33

Lead

17 -

- 51

17

- 31

21

20.5

Zinc

100 -

- 180

100 ■

- 160

122

120

* Correction means that one single outlier has been omitted. Mean values are computed from the corrected range.

* Correction means that one single outlier has been omitted. Mean values are computed from the corrected range.

As can be seen in Table 5.16, the metal concentration ranges in the 15 samples are quite narrow, and in most cases, the mean or median values are close to the minima. This makes it possible to draw the conclusion that it is likely that the real regional background level of Cu in the Stockholm area is 37-38 ^g/g DS (i.e. not 15 ^g/g DS, as indicated in the nation-wide background data). The corresponding background levels for Ni is 33 ^g/g DS, for Pb about 20 ^g/g DS and for Zn about 120 ^g/g DS. Therefore, it seems reasonable to make assessments of the degree of metal contamination of the inner water-ways of Stockholm by comparing measured concentrations with the above proposed "regional background levels", at least until some more reliable background data for the geological domain of eastern Svealand are available.

The above proposed regional background levels for Cu and Zn in Lake Malaren sediments turn out to be quite similar to very recently established levels for "preindustrial sediments from the Strangnas area" (Strangnas being a city on Lake Malaren upstream of Stockholm) by Lithner et al. (2003), of 35 ^g Cu/g DM and 130 ^g Zn/g DM. These authors also indicate that the Cu and Zn concentrations in "unaffected superficial sediments upstream Stockholm" may be estimated at around 50 ^g Cu/g DM.

0 0

Post a comment