Registration Dossier

Ecotoxicological information

Long-term toxicity to fish

Currently viewing:

Administrative data

Link to relevant study record(s)

Description of key information

The chronic toxicity to fish of 1,3-diethylthiourea was assessed by Van Leeuwen et al. (1990). However, these studies were disregarded for many reasons (significant methodology deficiences and documentation insufficient for assessment).

Key value for chemical safety assessment

Additional information

Two studies published within the same article investigated the long-term effects of 1,3 -diethylthiourea (DETU) on the embryo and larval developments of rainbow trout (Oncorhynchus mykiss) and zebrafish (Danio rerio). LOEC were determined for some regulatory endpoints. However, the article is not reliable enough to consider its results as relevant for regulatory purposes. We detail below the reasons why we conclude for disregarding these studies and classify them as Klimisch 3a/3b.

I – Critical analysis of van Leeuwen et al. (1990) Fish embryos as teratogenicity screens: a comparison of embryotoxicity between fish and birds; Environmental and Ecotoxicology Safety 20, 42 -52.

Experiment with Oncorhynchus mykiss

Study description

Van Leeuwen et al. (1990) carried out a 60-days ELS test on rainbow trout Oncorhynchus mykiss (formerly Salmo gairdneri) eggs and larvae under semi-static conditions, similar to the OECD 210 (2013) fish early-life toxicity test guideline. Freshly spawned eggs were challenged with graded concentrations of 1,3 -diethylthiourea (DETU) dissolved in water and separated by a 1.8 spacing factor. Eggs were incubated in relative darkness at 10°C and larvae hatched after 32 days. Photoperiod was then set at 12L:12D and kept constant until the end of the experiment. “Embryotoxicity”, mortality, weight and length of surviving larvae at the end of the experiment were recorded. This study is quoted Klimisch 3 for a multiplicity of reasons that are detailed below.

Inappropriate energy status of fish

Guideline 210 requires to feed rainbow trout larvae from the swim-up stage, a step that usually occurs 10 to 14 d after hatching (2015 FAO), i.e. at days 42 to 46 of the study, which was not done in the study of van Leeuwen et al. (1990). The absence of feeding for the 14 to 18 remaining days certainly resulted in lower energy status and weakened organisms, which is in opposition with the principle of the test.

Photoperiod was set at 12L:12D just after hatching, while a 7-day darkness period is demanded in the OECD 210 guideline. As for feeding, the 7-day darkness period is set to limit the energy loss due to the movements of the larvae under light conditions. The larvae are indeed more active under light than under darkness, which induces supplemental energy consumption and faster yolk sac depletion.

These two biological limitations surely affected the results by increasing the sensitivity of fish to DETU in an inappropriate way for regulatory purposes.

Methodological deficiencies

Dissolved oxygen was not monitored throughout the study. This clearly is a major experimental flaw when working with fish in general. A hundred rainbow trout larvae kept in 10 L water is indeed a significant loading rate requiring care to water oxygen level. Even if the authors state that aeration was provided, this is not sufficient proof that adequate oxygen concentrations were maintained throughout the 60 days of the study. This lack of information is considered as a methodological deficiency.

Loading rate was a hundred larvae per 10 L of solution in a semi-static system. Solution renewal was every two or three days. Assuming a larval mean weight of 50 mg at hatching as a lower case (From and Rasmussen, 1991), this makes a loading rate of 0.5 g/L over a 72-h period, i.e. a loading rate of 1.5 g/L per 24-h period, which is far above the recommended 0.5 g/L per 24 hours (OECD 210, 2013). This raises concerns about water quality (not monitored) and also dissolved oxygen levels, as discussed above.

Other methodological deficiencies include (i) duplicate instead of the quadruplicate testing recommended by the OECD 210 (2013) guideline, (ii) absence of water temperature monitoring and (iii) absence of any analytical monitoring of the substances.

Documentation insufficient for assessment

The article only presents statistical conclusions. No figure, table or data about the analysed variables (mortality, embryotoxicity, larval weight, larval length) in function of the tested doses are presented. Therefore, no judgement about the severity of the dose effect can be made (e.g., do the significant effects detected by mean comparison correspond to 3, 20% or any other percentage difference with the control group?).

The range of tested concentrations is not given and cannot be accurately guessed (although we would agree using a LOEC if the study and article were of sufficient quality for regulatory purposes). The only information given is the spacing factor which is 1.8

The quality (hatching and survival rates) of the control group cannot be assessed. The authors recognized that they had some mortality in the control groups without giving any figure.

The monotonicity of the dose-response and statistical regression quality cannot be checked.

The LOEC for length and weight are different, which is abnormal since both parameters are normally highly correlated in fish.

We consider that the documentation provided by van Leeuwen et al. (1990) is insufficient for reliability assessment.

Conclusions

For all these reasons, this study is classified Klimisch 3a (documentation insufficient for assessment) and 3b (significant methodological deficiencies). It is therefore disregarded for regulatory purposes in the framework of the REACH regulation. We consider that the results of Van Leeuwen et al. (1990) should not be taken into account for PNEC derivation.

II – Critical analysis of van Leeuwen C.J. et al. (1990) Fish Embryos as Teratogenicity Screens: A comparison of Embryotoxicity between Fish and Bird; Ecotoxicology and Environmental Safety 20 ,42-52.

Experiment with Danio rerio

Study description

Van Leeuwen et al. (1990) carried out a 7-days ELS test on zebrafish Danio rerio. This study on zebrafish is not a real long-term toxicity to fish experiment but we still examined the protocol description and results. Freshly spawned eggs of zebrafish were challenged with different doses of 1,3 -diethylthiourea (DETU) dissolved in water and separated by a 3.2 spacing factor. Sixty eggs were incubated in 50 mL solution for each concentration. Photoperiod was then set at 12L:12D and kept constant until the end of the experiment. Temperature was set at 25°C. Treatments were not replicated. “Embryotoxicity” and larval mortality were recorded.

We consider that this study should be quoted Klimisch 3 for a multiplicity of reasons that are detailed below.

Methodological deficiencies

Dissolved oxygen was not monitored throughout the study. This clearly is a major experimental flaw when working with fish in general. Sixty zebrafish larvae confined in 50 mL solution is indeed a significant loading rate requiring care to water oxygen level. Even if the authors state that aeration was provided, this is not sufficient proof that adequate oxygen concentrations were maintained throughout the 7-day period of the study. This is considered as a methodological deficiency.

Treatments were not replicated which is a major flaw. We wonder how mean comparisons could be performed without any replicate (e.g. how can a LOEC for mortality rate be calculated without any tank replicate ?). This is probably one of the highest concerns that can be raised about this study.

Other methodological deficiencies include (i) absence of water temperature monitoring and (ii) absence of any analytical monitoring of the substances.

Documentation insufficient for assessment

In this section, we address the same critical remarks as in part I. We consider that the documentation provided by van Leeuwen et al. (1990) is insufficient for reliability assessment of the study on zebrafish.

Conclusions

For all these reasons, this study is classified Klimisch 3a (documentation insufficient for assessment) and 3b (significant methodological deficiencies). It is therefore disregarded for regulatory purposes under REACH regulation. We consider that the results of Van Leeuwen et al. (1990) should not be taken into account for PNEC derivation.

III - Overall conclusions

As a conclusion of the critical reviewing of Van Leeuwen et al. (1990), we consider that both studies should be both classified Klimisch 3a (documentation insufficient for assessment) and 3b (significant methodological deficiencies). They are therefore disregarded for regulatory purposes under REACH regulation.