Application of BLMs

In the first few years of the development of the BLM concept, a series of predictions of the toxicity of copper and silver to freshwater fish were made and compared with results from bioassays conducted in a great variety of test water compositions. It was shown that a wide range of 96h LC50 values were obtained simply from adjustments of pH, DOC, alkalinity and the hardness of the test water. The BLM could be used to predict these effects. Input data to the BLM included measured water chemistry (pH, DOC, Ca, Mg, Na, K, Cl, SO4, CO3). In 'expressis verbis', BLM predicted LC50 values representing the amount of trace metal (e.g. copper) necessary to accumulate to lethal BL concentrations. Comparison of BLM results with measured values usually showed an excellent agreement, within a factor of two, of the measured values.

Using the BLM concept, De Schamphelaere and Janssen (2002) developed a method to estimate stability constants (values are given) for the binding of Cu2+, CuOH+, Ca2+, Mg2+, Na+, and H+ to the BL, solely based on toxicity data. By using these constants, the authors were able to predict acute Cu toxicity for D. magna as a funtion of just these water characteristics. Equilibrium equations for the binding of cations to the BL-sites could be formulated according to:

where kCuBL = stability constant for Cu2+ binding to BL sites

By combining mass balances and equilibrium equations for all cations that compete with Cu for binding sites, an equation was obtained for the fraction of the total number of Cu binding sites (ligands) occupied by Cu, which finally determines the magnitude of toxic effects independent of water quality characteristics. From this, 48h-EC50 values can be predicted, when Ca2+, Mg2+, Na+ and H+ activities are known, but without knowing measured Cu concentrations on the BL. Increases in Ca, Mg, Na (not K) resulted in an increasing 48h-EC50 for Cu. In contrast, bioassays showed limitations when the FIAM was used to predict Cu toxicity, since corresponding 48h-EC50s (here as Cu2+) differed by more than a factor of 10. Part of these differences may be explained by the linear relationship observed between EC50s (Cu2+) and Ca2+, Mg2+, Na+ and H+ activities and competition between metal ions and cations and hydrogen for binding sites on the biological surface. The fact that also Mg, beside Ca, reduced Cu toxicity suggests species-specific differences. In turn, similar effects of MgSO4 and MgCl2 indicate that inorganic ligands had no effect. The authors also observed differences in EC5o when pH was varied and explain this by proton competition at the BL site (linear relation), or by different pH values in the microenvironment of the BL compared to the ambient bulk solution leading to misinterpretation of the Cu-H interaction (De Schamphelaere and Janssen, 2002).

Also CuOH+ toxicity has to be considered, resulting in less Cu2+ needed to provoke toxicity at higher pH. A linear regression analysis suggested that indeed CuOH+ co-toxicity rather than proton competition may be the cause for the observed toxicity changes. The good agreement reported by others between the BLM constants of D. magna and P. promelas (Santore et al., 2001) suggests that daphnid BLs have the same affinity for metals and cations as BLs of fish, which may be incorrect, however, at pH values where CuOH+ co-toxicity becomes relevant. De Schamphelaere and Janssen (2002) found that the BL binding constant, log KCuBL for D. magna was 8.02 per M, while the constant for CuOH+ , log KCuOHBL, was 7.45 per M.

When the stability constants were included into the BLM computer program designed to predict 48h-EC50s (as dissolved Cu) for the water characteristics measured in the study, the predicted values differed from the observed EC50s only by a factor < 1.5. However, this excellent agreement between predicted and observed toxicity values was found only when tests were performed at pH <8, but not at all for tests performed at pH >8. Additional experiments (De Schamphelaere et al., 2002) demonstrated that this was due to toxicity of the CuCO3 complex, which is the most abundant inorganic copper species at pH >8. The log BL binding constant for this complex was determined to be 7.01 per M.

Later work on chronic copper toxicity to D. magna by De Schamphelaere (2003) showed that the log BL binding constants for the hydroxyde and carbonate complexes to gills were higher in chronic tests than in acute tests, 7.75 and 7.38 per M, respectively, indicating that other copper species, in addition to free cupric ions play a substantial role in the manifestation of chronic toxicity to daphnids. Also the effect of pH was more accentuated in the chronic experiments with daphnids compared with the acute tests (cf. review by Van Sprang, 2003).

In an elegant series of experiments Meyer et al. (1999) demonstrated that in laboratory tests, measured concentrations of nickel in gills [Ni g;ll ] and calculated concentrations of copper in gills [Cu g;ll ] were constant predictors of acute toxicity of Ni and Cu to fathead minnows (Pimephales promelas) when water hardness varied up to 10-fold, whereas total aqueous concentrations and free-ion activities of Ni and Cu were not. Thus, the BLM, which simultaneously accounts for (a) metal speciation in exposure water and (b) competitive binding of transition-metal ions and other cations to biotic ligands predicts acute toxicity better than does free -ion activity of Ni and Cu. The LA50 value or the "critical BL concentration", i.e. the concentration of metal bound to the fish gill at 2-3 h corresponding to 50% mortality at 96 h, was found to be: for nickel about 250 nmol Ni per g wet weight (ww) and for copper about 12 nmol/g (ww).

Later work, summarized by Di Toro et al. (2001), provided slightly modified values of the "critical BL concentration" in gills of fathead minnows: for Cu - 10 nmol/g (ww), for Ag - 17 nmol/g (ww), and for Ni -239 nmol/g (ww). In the same paper, critical BL concentrations were given for Ceriodaphnia dubia (Cu - 0.19 nmol/g) and for Daphnia magna (Ag -2.26 nmol/g). The fraction of sites at the BL occupied by the trace metal to produce a 50% effect in fish, i.e. the fCuBL50% in the case of copper, was 33% for fish (49% in the case of silver). However, in the crustaceans, the critical fraction of sites occupied by toxicants was only 0.6% in the case of C. dubia and Cu, and 6.5% in the case of D. magna and Ag. On the other hand, the BL binding constants for both Cu and Ag were almost identical (log values varied only in the range 7.3-7.4 per M) over the three investigated species and so were the site densities (30-35 nmol/g ww) (Di Toro et al., 2001). De Schamphelaere's and Janssen's (2002) determination of BL constants resulted in a value of 39% for the critical fraction of sites at the BLfCuBL50%) for D. magna, that was occupied by Cu to produce 50% effect.

In the extensive work on zinc (Heijerick et al., 2002a; Janssen, 2003) the BL binding constant, log KZnBL , was determined at 5.31 for both rainbow trout gills and for D. magna. In the case of acute toxicity of Zn, the critical fraction of BL sites occupied by Zn was 14% in the case of rainbow trout, but as much as 42% in the case of D. magna. Critical fractions to produce chronic toxicity of Zn (expressed as NOEC values) were 7.4% for rainbow trout and 7.7% for D. magna.

Heijerick et al. (2002b) established stability constants (KBL) for the binding of cations to algal cells by using a mathematical approach, which was recently developed by De Schamphelaere and Janssen (2002). By means of adsorption experiments they could show that K+ is the only cation not significantly changing Zn toxicity. Toxicity tests performed at various pH indicated a pH-dependent change of stability constants, due to competition between H+ and Zn2+, which was reducing Zn toxicity in a non-linear way. In addition, pH affected also the physiology of the BL (algae). According to the BLM, Zn toxicity to algae can be modelled as a function of key water characteristics. The results of Heijerick et al. (2002b) confirm that the statement that "the binding characteristics of BLs are independent of the test medium charcteristics" is at least not valid for algae. The critical fraction to give chronic effects to algae (ECio) was determined to be 14.3% for the green algal species used.

As a part of the research conducted to produce baseline information for the EU Risk Assessment for Zinc, BLMs were developed to predict chronic zinc toxicity to three standard test organisms, i.e. rainbow trout, Daphnia magna and the green alga, Pseudokirchneriella subspicata. The work consistently illustrated the importance of bioavailability parameters for chronic toxicity of zinc to rainbow trout, daphnids and algae and demonstrated that changes in zinc bioavailability to aquatic organisms can be quantified and predicted. The differences in toxicity, caused by bioavailability parameters were highest for the alga (a factor of 100), somewhat lower for fish (a factor of 2o) and lowest for daphnids (a factor of 4). For the green alga, the pH effect was the most important, while the effects of Ca, Mg and Na were negligible. For the daphnid, hardness and pH appeared to be equally important, whereas for rainbow trout, the effect of Ca was more important than that of pH. The effect of DOC seemed to be most pronounced for the alga, and of similar magnitude for the daphnid and the fish. In spite of these differences, the BLMs for all three organisms were able to significantly reduce the variation in effect concentrations, i.e. chronic effect concentrations were generally predicted within a factor of two from the observed values for all organisms, both when using laboratory waters and waters collected in the field (Janssen, 2003).

It certainly would be wise to keep in mind that, although BLMs seem to function well for organisms such as phytoplankton, zooplankton and fish, which normally take up substances from the surrounding water via well defined biological surfaces, e.g. the cell membrane (in the case of unicellular algae) or the gill membrane (in the case of multi-cellular organisms), thus fitting the GSIM model, it cannot be taken for granted that organisms taking up metals from the surrounding media via other routes will fit any standard BLM. An example clearly demonstrating the existence of such exceptions from the "BLM-rule" was given by Guo et al. (2001), who worked with American oysters (Crassostrea virginica). These, and most probably other marine bivalves, which are filter-feeding animals, are capable of concentrating metals from the large volumes of seawater they filter. Since trace metals are often associated with potentially nutritious DOC in the water, it was hypothesized that the uptake of DOC by bivalves may also enhance the uptake of trace metals. It was found that exactly this was the case, especially in the presence of high-molecular-weight DOC at high concentrations, the uptake of metals in the oysters was generally greatly enhanced, possibly due to direct ingestion and digestion of colloidally complexed metals (Guo et al., 2001).

It must be understood, however, that uptake of trace metals into the body tissues of filter-feeding animals does not necessarily result in toxic effects, since metals may be sequestered in not readily available forms in the tissues. In fact, the issue of dietary metal toxicity to filter-feeding organisms such as cladocerans is extremely important. It is obvious that, for example, the BLM developed by De Schamphelaere and Janssen (2003) to predict chronic waterborne copper toxicity to Daphnia magna is based on the assumption that there is no contribution of dietary copper to toxicity in the tests used for model development. In order to investigate the validity of this assumption, De Schamphelaere (2003) carried out a great number of chronic tests with D. magna, where toxicity was measured at three regimes of exposure to copper : (1) waterborne only, (2) dietary only, and (3) waterborne + dietary. It was found that the 21-d NOEC and EC50 for reproduction were nearly identical in the waterborne and in the waterborne + dietary exposure, being 95 and 110 ^g Cu/l (as dissolved copper), respectively. From this result it was concluded that the contribution of dietary copper to toxicity was unimportnat and did not interfere with BLM-predicted chronic copper toxicity (De Schamphelaere, 2003). Exposure to dietary copper resulted in an increased copper body burden in the adult daphnids, but no additional toxicity was observed. On the contrary, the dietary exposure route resulted in a 75% increase in growth and reproduction in the highest exposure level. Thus, when cells of green algae, used to feed the daphnids during the tests, were enriched with copper, a significant stimulatory effect was observed in the daphnids.

0 0

Post a comment