Zinc

7.1.3.1 Sensitivity to zinc of different aquatic organisms

An overwhelmingly great fraction of published data on zinc toxicity to aquatic organisms unfortunately suffers from poor quality. For example, many of the older studies suffer from such great deficiencies in experimental design and data presentation that it is doubtful whether the results can be regarded as scientifically well-founded (cf. L & L - Zn, Appendix 1). In fact, it is only in the relatively recent ecotoxicological literature on the aquatic toxicity of Zn that information necessary for reasonable interpretation of the data is given. Such crucial information would include, in addition to test duration, toxic end-point(s) assessed, mortality in control groups and other biological background data, a minimum number of physical/ chemical data, such as temperature, pH and measured Zn concentration in the test solution, as well as water hardness, DOC and the possible occurrence of synthetic complexing agents.

Within the scope of this update of the previous monographs on metals, it is not our intention to try to make a comprehensive review of the very abundant new literature on Zn toxicity. In stead, we would like to refer the reader to the summaries of aquatic toxicity data in the recent "Environmental Health Criteria" on Zn (IPCS, 2001) and to the comprehensive compilation of aquatic toxicity data on Zn made as a part of the EU Environmental Risk Assessment (RA) on Zinc (RIVM, 1999). Research carried out in conjunction with this RA has indicated that the sensitivity of the toxic response of a test organism to an essential metal like Zn is a function of the concentrations of essential elements in which it was cultivated. For example, test organisms reared under conditions where the dissolved Zn concentration was <2 (o,g/l exhibited a mean NOEC of 36 ^g Zn/l, while test organisms reared at >2-<50 ^g Zn/l had a mean NOEC of 100 ^g Zn/l as dissolved Zn (Waeterschoot et al., 2003). It was furthermore demonstrated that a dissolved Zn concentration of 17 ^g/l would provide protection of 95% of the tested aquatic species in the water qualities typical for lowland river systems in Europe.

In the zinc monograph (L & L - Zn) the maximum tolerable concentration (MTC) of zinc in soft inland waters was set at 25 ^g dissolved Zn/l to achieve 95% protection of all endpoints in the aquatic ecosystem, and at 15 ^g Zn/l to achieve close to 100% protection.That the regional, continental European PNEC value would be at a level of 15-25 ^g Zn/l is further supported by a recent report by Woodling et al. (2002), where the authors have determined the chronic toxicity of zinc to the mottled sculpin (Cottus bairdi) in soft water (46.3 mg/l as CaCO3) to 27 ^g/l (LOEC) and 16 ^g/l (NOEC), a lower toxic zinc concentration than observed for any other fish species tested to date (according to the authors)..

What is really new in the field of research on aquatic toxicity of zinc is the development of BLMs predicting acute and chronic zinc toxicity to standard test organisms, rainbow trout, Daphnia magna and a green alga. In this work, the various causes of variation in the toxicity values were identified and quantified so that prediction of chronic effect concentrations could be made within a factor 2 of the observed effect concentrations, not only for laboratory waters but for natural surface waters as well. These new advances in zinc ecotoxicity research will be briefly presented in the following section.

7.1.3.2 Toxicity of zinc estimated by means of BLMs

As mentioned in chapter 6 and in section 7.1.1.2, the BLM concept was originally developed by Di Toro et al. (2001) using the conditional stability constants for competing cations (e.g. Ca2+, H+) and trace metals for binding to biotic ligands on fish gills, derived from metal binding experiments with fish gills (Playle et al., 1993). The fraction of binding sites that needs to be occupied for causing an effect was calculated in the early work by fitting the BLM to existing literature data (Santore et al., 2002). However, an improved methodology was developed by De Schamphelaere and Janssen (2002), where they used experimental effects data of univariate toxicity assays, in which one parameter was varied and all other parameters were kept constant. This approach was successfully applied to the development of acute BLMs for copper and zinc with D. magna, and later on, it was used to develop chronic BLMs for copper and zinc with rainbow trout, D. magna and the green alga, Pseudokirchneriella subcapitata (De Schamphelaere et al., 2003a ; b ; Heijerick et al., 2003).

The main purpose of this work was to test the potential of the BLMs to predict copper and zinc speciation, complexation, bioavailability and chronic aquatic toxicity over a relevant range of water chemistry conditions as existing in European inland surface waters.

The importance of various bioavailability modifying factors (water quality parameters) for chronic zinc toxicity to the three test species was quantified and the relative importance of each of these is summarized in Table 7.5. The toxic end-point used for rainbow trout was mortality, for daphnids mortality and reproduction impairment, and for the green alga, growth rate deviating from that of the control.

Table 7.5. Importance of zinc bioavailability modifying factors for toxicity variation: number of times the lowest and highest toxicity differed in a series of univariate chronic toxicity tests with zinc for rainbow trout (RT), D. magna (DM) and the green alga P. subcapitata (PS). After De Schamphelaere et al., 2003a.

Parameter

Calcium

Magnesium

Overall, EC50 d:o , NOEC

18 28

63 127

As shown in Table 7.5, the changes in zinc bioavailability to aquatic organisms can be quantified and predicted. The toxicity differences caused by bioavailability parameters were highest for algae (a factor of 130), and lower for fish (factor 30) and daphnids (factor 4). For algae the pH effect was the most important, but for fish, the effect of calcium was greater than that of pH. In spite of these differences, BLMs for the three organisms were able to significantly reduce the variation; i.e. the chronic effect concentrations were generally predicted within a factor of 2 from the observed values for all three organisms. Thus, it was concluded that the BLMs developed in the study conducted by Schamphelaere et al. (2003a) could be used as efficient tools to reduce the toxicity variation, due to differences in zinc bioavailability, from a factor of more than 100 to a factor of 2, in both laboratory and field waters.

The conditional stability constant for binding of zinc to the BL, log KZnBL, was assumed to be equal to the one reported for the acute Zn-BLM for Daphnia magna by Heijerick et al. (2002a), i.e. 5.31, and equal in all three organisms. The fraction of binding sites occupied by zinc to produce 50% mortality in acute exposures, the /ZrBL was found to be 14% in the case of rainbow trout and about 42% in the case of D. magna. The corresponding / values in the chronic exposures are shown in Table 7.6, together with the chronic toxicity levels, recalculated as ^g dissolved Zn per litre of water.

Table 7.6. Chronic toxicity data for Zn, estimated by means of BLMs for rainbow trout (RT), D. magna (DM) and the green alga P. subcapitata (PS). After De Schamphelaere et al., 2003a.

Parameter

RT

DM

/ZnBL , %

10.4

11.7

/ZnBL , %

5.1

7.7

14.3*

EC50, ^g Zn/l NOEC, ^g Zn/l

110-1970 32-890

110-370 48-170

26-1630 4.8-610*

* for algae, the NOEC value is replaced by EC10

* for algae, the NOEC value is replaced by EC10

The variations in chronic toxicity values given in Table 7.6 are based on laboratory tests with reconstituted waters covering the theoretical range of water qualities that may occur. However, the ranges of physico-chemical data actually observed in European surface waters have also been used to predict chronic toxicity to the three test organisms by means of the BLMs. When this was done using 3-year data for 411 monitoring locations in the GEMS database, representing the whole of Europe, the following results were obtained:

Range of NOEC predictions for the calculated 10th to 90th percentiles for the 50th percentiles of the 3-year GEMS data, representing the whole of Europe:

• Rainbow trout, chronic toxicity, NOEC: 120 - 1080 ^g Zn/l

• Daphnia magna, chronic toxicity, NOEC: 110 - 760 ^g Zn/l

• Green alga, chronic toxicity, NOEC: 23 - 100 ^g Zn/l

It should be stressed, however, that Scandinavian surface waters are not well represented in the GEMS database. A recent compilation of water quality data from Norway, Finland and Sweden (NIVA, 2001) shows that the median values for water hardness in these countries vary in the range 4.0 -11.5 mg/l as CaCO3, the pH between 6.4 and 6.8, while the median TOC levels are 1.9 - 6.1 mg/l. The corresponding 50th percentiles in the GEMS all-European database are: 151 mg/l, 7.8 and 4.1 mg/l, respectively.

0 0

Post a comment