Silver

Long-term toxicity to aquatic invertebrates

Description of key information

Seven chronic NOECs were taken forward for use in the derivation of the freshwater PNEC by statistical extrapolation. These were an EC10 of 6.13 µg/L Ag for Ceriodaphnia dubia (geometric mean of four EC10 values reported by Kolts et al. (2009) and Naddy et al. (2007)), an EC10 of 2.14 µg/L Ag for Daphnia magna (Bianchini and Wood 2008), an EC10 of 1.0 µg/L Ag for Ceriodaphnia reticulata (Elnabarawy et al. 1986), an EC10 of 1.48 µg/L Ag for Isonychia bicolor (Diamond et al. 1990), an EC10 of 1.0 µg/L Ag for Stenonema modestum (Diamond et al. 1992), an EC10 of 1.68 µg/L Ag for Corbicula manilensis (Diamond et al. 1990) and an EC10 of 14.43 µg/L Ag for Chironomus tentans (Call et al. 1999). A NOEC of 8.6 µg/L Ag for Arbacia punctulata (Ward et al. 2006) is taken forward for the derivation of the marine PNEC.

Additional information

Data on the toxicity of silver were available for several different species of freshwater invertebrates (crustaceans, insects and molluscs).

 

Key long-term data from three species of crustacean were available:Ceriodaphnia dubia, Ceriodaphnia reticulata and Daphnia magna. C. dubia data were selected from two of the three available reliable studies: Kolts et al. (2009) and Naddy et al. (2007). Both studies were conducted according to US EPA guidelines under a range of water conditions and measured concentrations of dissolved silver in the test media. As the water conditions did not appear to influence the results of the tests, the geometric mean of four available seven-day EC10 values for inhibition of reproduction (2.48, 10.1, 6.48, 8.69 μg/L measured dissolved silver) was used in the SSD (6.13 μg/L dissolved silver). C. dubia data from Rodgers et al. (1997), whilst considered generally reliable, were not selected as key data for the calculation of the PNEC as they were generated from a static exposure regime. Key data for C. reticulata were available from a single seven-day semi-static study (Elnabarawy et al. 1986). This study reported a NOEC for reproduction of 1.0 µg/L total silver (nominal). Elnabarawy et al. (1986) also report the results of similar reproduction tests usingDaphnia pulex. However, as this test reported effects on reproduction at the lowest test concentration it was not suitable for PNEC derivation. Key long-term Daphnia magna data were available from five studies: Nebeker et al. (1983), Elnabarawy et al. (1986), Naddy et al. (2007), Bianchini and Wood (2008) and Schlich et al. (2017c). From these studies, key data were selected from Bianchini and Wood (2008) as, of the studies assigned a Klimisch 1 score (Bianchini and Wood (2008) and Schlich et al. (2017c)), this study reported the lowest EC10 values. Within this study a 21-day EC10 value of 2.14 μg/L dissolved silver (measured) for inhibition of growth in moderately hard water was preferred to results from hard water (2.33 μg/L) and hard water with added sulfide treatments (3.19 μg/L), as these appeared to reduce the bioavailability and toxicity of silver.

Key data for three insect species representing two insect orders were available. An EC10 of 1.48 μg/L total silver (measured) was calculated for effects on moult production inIsonychia bicolour(a mayfly) in a 20-day semi-static exposure (Diamond et al. 1990). Exposure of Stenonema modestum, a second species of mayfly, for 14 days resulted in a comparable NOEC of 1.0 μg/L measured dissolved silver, also for effects on moult production (Diamond et al. 1992). Both of these studies were assigned Klimisch scores of 2 as they were conducted using non-standard species and measured a non-standard endpoint. However, both of these results are considered to be valid and sufficiently robust for PNEC derivation. Key data on Chironomus tentans were available from a 10-day flow-through water-only exposure conducted by Call et al. (1999). An EC10 of 14.43 μg/L was calculated from the growth data presented in the paper. Rodgers et al. (1997) also conducted exposures with C. tentans and reported a 10-day NOEC of 125 μg/L dissolved silver (measured); these data were not selected for use in thecalculation of the PNECas they were undertaken under static exposure conditions. Data from a single mollusc species are available (Corbicula manilensis). Diamond et al. (1990) report a 21-day EC10 of 1.68 μg/L total silver (measured) for growth inhibition derived from a semi static test.

The lowest reliable long-term result for marine crustaceans is a 28-day developmental NOEC of 19 µg/L dissolved silver for Americamysis bahia (Ward et al. 2006). The authors conducted tests at three salinity levels: 10, 20 and 30 ‰. Whilst increases in salinity generally require higher concentrations of dissolved silver to cause a chronic effect, this was not seen in this study which produced NOECs of 6, 34 and 19 µg/L dissolved silver at 10, 20 and 30 ‰ salinity, respectively.

The lowest reliable long-term result for marine molluscs is a 24-month reproduction NOEC of 5 μg/L dissolved silver for Crepidula fornicata (Nelson et al. 1983). Larval releases were significantly reduced at 10 μg/L in the parental generation and whilst the number of larvae released by the second generation was reduced at 5 μg/L it was not statistically significant. This test was carried out at 25±2 ‰ salinity.

The lowest long-term result for echinoderms is a 30-day NOEC of 8.6 µg/L dissolved silver for Arbacia punctulata for sublethal effects, which included loss of spines or spine turgor and an inability to adhere to surfaces (Ward et al. 2006). This study was initiated with adult sea urchins and at study termination an unsuccessful attempt was made to collect gametes from each surviving adult, indicating that the sea urchins were not conditioned to spawn.

The normal salinity range for seawater is 30 - 35 ‰ and therefore the PNEC should be based on the most sensitive result from a test carried out within this range. A long-term saltwater PNEC for silver should therefore be based on the NOEC for effects on the sea urchin, Arbacia punctulata (8.6 µg/L).