Chlorine

Long-term toxicity to aquatic invertebrates

Description of key information

Read-across from sodium hypochlorite (justification see IUCLID5 section 6.1 or CSR section 7.1.1): 
In a field study on chlorinated condenser cooling effluents using mollusk bivalves (Liden et al., 1980), the survival of oysters (Crassostrea virginica) and clams (Rangia cuneata) maintained at three TRO concentrations for 15 days was not affected at concentration as high as 62 μg/l, while oyster mean shell deposition was significantly reduced in the treated animals. At the lowest test concentration (14 μg/l) a 14% reduction in shell deposition was observed, so that following TGD the NOEC can be estimated as LOEC/2, i.e. 7 μg/l. 

Key value for chemical safety assessment

Marine water invertebrates

Marine water invertebrates

Effect concentration:

0.007 mg/L

Additional information

Read-across from sodium hypochlorite (justification see IUCLID5 section 6.1 or CSR section 7.1.1):

Freshwater:

The long-term toxicity of hypochlorite to invertebrates has been tested on bivalve mollusks. These organisms are selected as test species because they are biofouling

organisms which need to be controlled. They also play an important ecological role in nutrient cycling and as a food source but are not among the most sensitive species.

Martin et al. (1993) measured mortality in Dreissena polimorfa exposed in a static test at 21 °C for 20 days in a darkened environment to minimize photolytic loss of sodium hypochlorite. The LC50 after 11 days was about 1 mg/l and remained practically unchanged at the end of the experiment. At 0.5 mg/l mortality was <10% after 20 d; this was the highest concentration with no significant effects, so that it can be considered as NOEC. The same authors (Martin et al., 1993) carried out a longer test at 12 °C, using the same experimental procedure. After 15d exposure at 1 mg/l 50% mortality occurred and at the end of the experiment (29d) it reached 70%. No NOEC has been calculated. In these studies the actual exposure chlorine concentrations are likely to be over-estimated: concentrations were measured and maintained by daily adjustments, but authors did not measure the decay curve and do not specify if initial or average concentrations were used in the endpoint calculation. We do not consider these data adequate for risk assessment because the toxicity is likely underestimated, also in consideration that the pulse dosing conditions allow a recovery of animals, in contrast to flow through conditions. The results of other tests with Dreissena (Klerks and Fraleigh, 1991, see below) support this reasoning.

Ramsay et al. (1988) carried out a field study to measure the time required for continuous chlorination (with sodium hypochlorite) to produce 100% mortality in adult Corbicula fluminea (asiatic clam). The TRC was continuously monitored during the flow-through test. At the lowest concentration (0.05 mg/l), 100% of the clams died after 36d exposure. From the effect-time curve we estimated that 50% clams were dead after 8d.

Klerk and Freileigh (1991) investigated the long-term toxicity of NaOCl to zebra mussel (Dreissena polimorfa) in a number of tests (28d, static renewal and intermittent exposure; 28 and 56 d flow-through), carried out using natural water at different temperatures. In the longer flow-through test, after 56d exposure to the

lowest tested concentration (0.08 mg/l as free chlorine), the mortality reached 55%. At this concentration, 86% reduction in the filtering frequency was observed, but this effect was not concentration related. For the three treatments (0.08, 0.26, 1.35 mg/l of free Cl2) the LT50 values ranged from 16 to 54 d.

The study by Kilgour and Baker (1994) was designed to evaluate the influence of many variables (season and place of animals’ collection and experimental protocol) on the toxicity of sodium hypochlorite to Dreissena polimorfa. The test lasted 9 d, during which zebra mussels were exposed in a static system to 5 hypochlorite concentrations that were adjusted daily. The LC50 was calculated using average concentrations estimated on the basis of decay curves. The most sensitive mussels were those collected in late summer, whose 9d LC50 was about 0.5 mg/l (as TRO). It is worth noting that this concentration corresponds to the NOEC in the experiments of Martin et al (1993a). The study by Kilgour and Baker is reliable but no NOEC can be derived; it can only be retrieved the LC50 as an indication of long term toxicity (supportive information).

The studies by Ramsay et al. (1988), Klerks and Fraleigh (1991), and Kilgour and Baker (1994) are well conducted and described but no NOEC/LOEC is reported or can be derived. The end points provided cannot be used as such for risk assessment, the information gained only provides an indication of long term toxicity for an invertebrate species, which, it is important to consider, is neither a standard species nor one of the most sensitive. We used these results as supportive information (rated s).

Marine water:

In the scientific literature, three studies investigating the long-term toxicity of sodium hypochlorite have been found. In a field study on chlorinated condenser cooling effluents using mollusk bivalves (Liden et al., 1980), the survival of oysters (Crassostrea virginica) and clams (Rangia cuneata) maintained at three TRO concentrations for 15 days was not affected at concentration as high as 62 μg/l, while oyster mean shell deposition was significantly reduced in the treated animals. At the lowest test concentration (14 μg/l) a 14% reduction in shell deposition was observed, so that following TGD the NOEC can be estimated as LOEC/2, i.e. 7 μg/l. This data is rated 2 because it was obtained from a non-standard test.

Scott and Middaugh (1977) examined the lethal and sublethal effects of chlorine to adult oyster (Crassostrea virginica) during seasonal chronic exposures (45-75 days) to incoming estuarine sea water, whose temperature, pH and salinity was naturally fluctuating. Survival was reduced at relatively high concentrations of Chlorine- Produced Oxidants (CPO); a rough estimate is LC10≅160 μg/l CPO on the basis of the concentration-effect curves. On the other hand, severe sublethal effects (mean condition index and gonadal index) were observed also at the lowest tested concentration (140 μg/l CPO), so that the NOEC for these effects would be expected to be <<140 μg/l CPO. Because of the lack of a clear endpoint, these data cannot be used in risk assessment.

In another experiment conducted in summer Scott. et al. (1979) found that mortality of C. virginica was not significantly reduced at 110 μg/l but the high mortality observed in the control make this result unreliable. At the same concentration marked sublethal effects were observed, but no direct concentration-effects relationship was found (not valid data).