Table 191

Sampling stations descriptions

Station name Description

Dantziggat This station is situated in the middle of the Wadden Sea. The area is highly influenced by tidal movement resulting in frequent sedimentation and re-suspension.

Malzwin Situated in the west part of the Wadden Sea. This part is under frequent influence of North Sea water entering the Wadden sea at high tide, followed by a return stream.

Slijkgat This is the only station in the North Sea, although close to the coast. The station is influenced to a varying extent by the outflow of Rhine and Meuse water discharged from the Haringvliet Sluices.

Vlissingen Station in the mouth of the Western Scheldt. Almost fully marine with influence of the Western Scheldt outflow.

Hansweert Upstream in the Western Scheldt. Salinity is still sufficiently high for mussels to survive.

Bommenede De Grevelingen is the largest marine water lake in Europe.

Since it is a lake it has no tidal current. The station Bommenede is situated in the central part of De Grevelingen.

Wissekerke The Eastern Scheldt is essentially a lagoon. It is connected to the North Sea by an open dam that can be closed when necessary because of a spring tide or storm. The station Wissekerke is situated on the west side not far from the dam.

Yerseke This station is also situated in the Eastern Scheldt but in the eastern part. Several mussel farms are situated in the surrounding area.

Moreover, monitoring by PSs has some advantages over the use of living organisms:

• initial concentration of contaminants is negligible;

• PSs do not metabolise pollutants;

• are not mortal and can be deployed in all salinities;

• the uptake process is simple compared with that found in organisms.

Fig. 19.1. Position of sampling stations in the marine waters of The Netherlands. For sampling site names and description see Table 19.1.

The contaminant uptake, or exchange, process of PSs has been widely investigated and can be better analytically quantified than for mussels. Organisms have several uptake/release routes and the exchange rate depends heavily on environmental variables such as temperature, food availability, stress factors, etc. which influence the activity of the mussels [11]. While PSs take up from only the freely dissolved phase, mussels may also take contaminants through feeding. However, in spite of the better-defined uptake of PSs, the field conditions also may have a considerable influence on the results. Temperature, hydrodynamic conditions and biofouling are factors that influence the uptake rate and may cause different concentrations in the sampler for equal contamination levels. The resulting variability may still be lower than data influenced by biological factors but much higher than common quality assurance targets for analytical variability. This is not an issue when applying PSs to compare heavily contaminated sites with reference areas, e.g., differences in contaminant level of a factor 10 or more. However, in the case of monitoring the quality of (marine) surface waters in order to detect changes in contaminant levels, e.g., as result of measures to reduce pollution, it is of utmost importance to reduce variability to a minimum. This can be done in different ways:

• ''Isolate'' the sampler from external turbulences using an increased layer of stagnant water by installing a cover of a more or less hy-drophilic material. Diffusion through this layer is then the uptake rate-limiting step [12,13]. A disadvantage is that the uptake rate is considerably reduced and consequently the detection limits will increase substantially. Moreover, a hydrophilic layer could also act as a substrate for biofouling more than a strongly hydrophobic surface.

• Another possibility is correcting for the influence of current and waves instead of regulating the uptake rate. This is performed through the use of PRCs that are spiked in the sampler and are released to the water phase by the same processes by which uptake occurs [4,5]. In that way the uptake is not limited and a maximum sensitivity is obtained.

• Booij et al. investigated a third method by mechanically rotating the sampler or creating a large flow that could overrule the differences in local hydrodynamic conditions [14]. Basically, this would be the ideal approach as a very high sampling rate is obtained giving high sensitivity and the influence of local flow conditions is strongly reduced if not excluded. Moreover, when using rotating or otherwise moving samplers the risk of biofouling also diminishes. However, the large amount of energy needed for such non-passive sampling approach is generally not available at sampling stations.

Thus, for correcting the influence of hydrodynamic conditions the use of PRCs is preferred over isolation. It allows a simpler sampler design and sample processing, and results in higher uptake, i.e., better sensitivity.

19.2.4 Objectives

Passive sampling in aquatic environments has been proposed as a promising technique since before 1990. The large amount of literature on passive sampling is mainly research orientated, and comprises data from one time/year and a few sites with special pollution histories. So far passive sampling has not been widely applied in routine water quality monitoring programmes. The objectives of the trial implementation of the passive sampling monitoring programme reported here were:

• to build experience and to optimise the procedure through ''learning by doing'';

• that application alongside mussels should validate the environmental relevance of the results obtained;

• to obtain practical data that allow evaluation of the suitability for routine monitoring.

In routine monitoring the methodology must be robust, repeatable and accompanied by a sufficient level of quality assurance. Over the years the sampler design has been adapted, extraction procedures have been improved and clean-up simplified. Several PRCs have been applied and evaluated.

Validation of passive sampling methods is not as straightforward as for classical analytical methods. Further a distinction must be drawn between analytical and environmental validation. For classical analyses of water usually only analytical validation is considered, meaning that a true concentration is determined with a known uncertainty. However, due to sorption of contaminants to, e.g. suspended matter, such a concentration does not necessarily reflect the (bio)available fraction. The conversion into an environmental risk is generally accompanied by a huge uncertainty and is hard to validate. A sound analytical procedure does not automatically mean that environmentally relevant data are obtained [16].

Passive sampling is a chain of actions and calculations using various constants (all with their uncertainty) that yields an estimate of the freely dissolved aqueous-phase concentration. From an analytical point of view this will end in a higher uncertainty than a classical analytical method. However, one should be aware that a more environmentally relevant result is obtained as the freely dissolved concentration is directly proportional to the environmental threat for that contaminant. Moreover, the concentration obtained reflects the exposure over a long period of time. Therefore, in a risk assessment the result of passive sampling is likely to provide a better reflection of the risk than that of a classical analysis of a total water sample. Validation of passive sampling using whole or filtered water analyses will always suffer from the inability to isolate the freely dissolved fraction, especially for hydrophobic compounds [15]. Testing different filtration systems revealed that part of the free dissolved fraction is affected by sorption to the filter material and contaminants bound to dissolved organic matter pass through the filter to a different extent depending on the extent of clogging of the filter [16]. Therefore, instead of trying to show agreement with large volume grab sampling followed by classical analyses, the validation of passive sampling would be better achieved by demonstrating that exposure concentrations experienced by organisms can be predicted sufficiently accurately.

0 0

Post a comment