Biota monitoring and the Water Framework Directive-can normalization overcome shortcomings in sampling strategies?

Abstract

We compare the results of different monitoring programs regarding spatial and temporal trends of priority hazardous substances of the European Water Framework Directive (WFD). Fish monitoring data for hexachlorobenzene (HCB), mercury (Hg), and perfluorooctane sulfonic acid (PFOS) sampled in German freshwaters between the mid-1990s and 2014 were evaluated according to the recommendations of the 2014 adopted WFD guidance document on biota monitoring, i.e., normalization to 5 % lipid content (HCB) or 26 % dry mass (Hg, PFOS) and adjustment to trophic level (TL) 4. Data of the German Environmental Specimen Bank (ESB) (annual pooled samples of bream) were compared to monitoring data of the German federal states (FS), which refer to individual fish of different species. Significant decreasing trends (p < 0.01) were detected for Hg in bream (Abramis brama) sampled by both, the ESB and the FS between 1993 and 2013 but not for FS samples comprising different fish species. Data for HCB and PFOS were more heterogeneous due to a smaller database and gave no consistent results. Obviously, normalization could not compensate differences in sampling strategies. The results suggest that the data treatment procedure proposed in the guidance document has shortcomings and emphasize the importance of highly standardized sampling programs in trend monitoring or whenever results between sites have to be compared.

Keywords: Biota monitoring; Environmental Specimen Bank; HCB; Hg; Normalization; PFOS; Water Framework Directive.

Fig. 1

Temporal trends in HCB concentrations (μg/kg; wet weight-based) in muscle tissue of freshwater fish in Germany sampled by the German Environmental Specimen Bank (ESB) and the federal states (FS). Data are adjusted to a standard fish of trophic level 4.0 and 5 % lipid content. Shown are the antilog data of annual means of log-transformed concentrations after normalization. The lines represent the results of the linear regression (solid for significant linear trend, dashed for not significant) and the LOESS smoother (solid for significant non-linear trend, dashed for not significant).The shaded areas mark the 95 % confidence intervals of the LOESS function. Data basis: ESB: annual pool samples of approximately 20 bream each; FS: individual fish of different species. Red line EQS_Biota (10 μg/kg wet weight)

Fig. 2

Temporal trends in Hg concentrations (μg/kg; wet weight-based) in muscle tissue of freshwater fish in Germany sampled by the German Environmental Specimen Bank (ESB) and the federal states (FS). Data are adjusted to a standard fish of trophic level 4.0 and 26 % dry mass. Shown are the antilog data of annual means of log-transformed concentrations after normalization. The lines represent the results of the linear regression (solid for significant linear trend, dashed for not significant) and the LOESS smoother (solid for significant non-linear trend, dashed for not significant). The shaded areas mark the 95 % confidence intervals of the LOESS function. Data basis: ESB: annual pool samples of approximately 20 bream each; FS: individual fish of different species. Red line EQS_Biota (20 μg/kg wet weight)

Fig. 3

Temporal trends in PFOS concentrations (μg/kg; wet weight-based) in muscle tissue of freshwater fish in Germany sampled by the German Environmental Specimen Bank (ESB) and the federal states (FS). Data are adjusted to a standard fish of trophic level 4.0 and 26 % dry mass. Shown are the antilog data of annual means of log-transformed concentrations after normalization. The lines represent the results of the linear regression (solid for significant linear trend, dashed for not significant) and the LOESS smoother (solid for significant non-linear trend, dashed for not significant). The shaded areas mark the 95 % confidence intervals of the LOESS function. Data basis: ESB: annual pool samples of approximately 20 bream each; FS: individual fish of different species. Red line EQS_Biota (9.1 μg/kg wet weight)