Registration Dossier

Administrative data

Hazard for aquatic organisms

Freshwater

Hazard assessment conclusion:
PNEC aqua (freshwater)
PNEC value:
0.001 mg/L
Assessment factor:
5
Extrapolation method:
sensitivity distribution
PNEC freshwater (intermittent releases):
0 mg/L

Marine water

Hazard assessment conclusion:
PNEC aqua (marine water)
PNEC value:
0.001 mg/L
Assessment factor:
5
Extrapolation method:
sensitivity distribution

STP

Hazard assessment conclusion:
PNEC STP
PNEC value:
9.5 mg/L
Assessment factor:
100
Extrapolation method:
assessment factor

Sediment (freshwater)

Hazard assessment conclusion:
PNEC sediment (freshwater)
PNEC value:
4.62 mg/kg sediment dw
Assessment factor:
50
Extrapolation method:
assessment factor

Sediment (marine water)

Hazard assessment conclusion:
PNEC sediment (marine water)
PNEC value:
1.23 mg/kg sediment dw
Assessment factor:
50
Extrapolation method:
assessment factor

Hazard for air

Air

Hazard assessment conclusion:
no hazard identified

Hazard for terrestrial organisms

Soil

Hazard assessment conclusion:
PNEC soil
PNEC value:
2.3 mg/kg soil dw
Assessment factor:
10
Extrapolation method:
assessment factor

Hazard for predators

Secondary poisoning

Hazard assessment conclusion:
PNEC oral
PNEC value:
2.36 mg/kg food
Assessment factor:
30

Additional information

PNEC aquatic

The PNEC aquatic for freshwater and marine water was calculated using a species sensitivity distribution according to Aldenberg & Jaworska (2000) Method. Ninty-five percent confidence limits are as follows:

 

Upper 95% CL (mg/L) 

Lower 95% CL (mg/L) 

PNEC aquatic freshwater   0.001677

0.000147

 PNEC aquatic marine 0.001289 0.00012

PNEC oral

The PNEC oral was derived by converting the NOEL to a NOEC in accordance with ECHA Guidance.

NOECoral, predator= NOAELoral, predatorx CONVpredator

- Conversion factor = body weight/ daily food intake. For Gallus domesticus (chicken) CONVpredatoris 8.

- For Zebra Finch this is calculated as 12 g/ 3.5 g per day = 3.429

NOECoral, predator= NOAELoral, predatorx CONVpredator

NOECoral, predator= 20.632 mg/kg x 3.429 = 70.747 mg/kg

PNECoral, bird= chronic NOEC/ AF of 30 = 2.36 mg/kg

EFFECTS ASSESSMENT SUMMARY OF EURAR 2002 REPORT

General approach:

EU Risk Assessment 4-Nonylphenol (Branched) and Nonylphenol (EURAR 2002) Report classified a study as valid and reliable for use in the risk assessment if the study fully described the test material used, the test organism, the test method and conditions and if the endpoint concentration was based on measured values. If only some of the criteria were met, then studies were noted as to be “used with care” for support of valid studies. A more rigorous approach involving application of the extended Klimisch system (HERAG, 2007) was used to evaluate studies for REACH registration. Study information not only had to be provided, but studies also had to be performed according to or similar to Guidelines to be considered Klimisch 1. Only studies scoring a Klimisch 1 or 2 were used in the risk assessment as a key or supporting study.

REACH Endpoint 9.1.1 Short-Term Toxicity to Invertebrates

In accordance with EURAR 2002, the Brooke (1993) study is selected as a key study as it presented the lowest (in the EURAR 2002 Report) valid and reliable result of 0.085 mg/L NP for acute toxicity toDaphnia magna. This evaluation considered additional reliable studies that had been published after the EURAR 2002 Report was written. LC/EC50results ranged from 0.021 to 0.60 mg/L NP for freshwater invertebrates and 0.04 to 0.2 mg/L for saltwater invertebrates. However, the Brooke (1993) study was well documented and provided results for the preferred test organism,Daphnia magna. As per REACH Guidance, although long-term data are preferred, short-term data can be used to determine PNEC values. Since adequate reliable long-term aquatic toxicity data were available, short-term reliable data was not incorporated in the aquatic PNEC calculation for in the EURAR 2002 Report or the PNEC derivation for REACH.

REACH Endpoint 9.1.3 Short-Term Toxicity to Fish

In accordance with EURAR 2002, the Brooke (1993) study was determined as a key study as it presented the one of the lowest valid and reliable result of 0.128 mg/L NP, based on survival, for acute toxicity to the preferred freshwater test organism,Pimephales promelas. In addition, a more recent study (Dwyer et al., 2005) which provided reliable short-term data for 16 fish species was obtained for the purpose of the REACH dossier. This study was also designated as a key study, although actual test concentrations were not measured, only the stock solution was measured. 

EURAR (2002) presented a 96-hr LC50of 0.017 mg/L for the marine fish,Pleuronectes americanusas presented by Lussier et al. (2000). Additional reliable short-term studies for marine fish were obtained (Dwyer et al., 2005 and Kelly 2000), however, the result presented by Lussier et al (2000) remains the lowest reliable short-term result for toxicity of NP to fish.

REACH Endpoints 9.1.5 Long-Term Toxicity to Invertebrates and 9.1.6 Long-Term Toxicity to Fish

In accordance wtih EURAR 2002, reliable long-term data were utilized for PNEC determination. See below PNEC Comparisons for details regarding long-term data comparisons.

REACH Endpoint 9.1.2 Growth Inhibition in Aquatic Plants (Algae)

EURAR (2002) presented a 72-hr EC50value of 0.0563 mg/L and EC10 value of 0.0033 mg/L forScenedesmus subspicatus(Kopf, 1997) and a 96-hr EC50of 0.027 mg/L for the saltwater alga Skeletonema costatum (Ward and Boeri 1990). EURAR 2002 Report states that EC50results are considered short-term results where the EC10result would be considered equivalent to a long-term NOEC. Kopf (1997) could not be included in this assessment as it is a book where ownership issues precluded it from being included in the dossier submission. For PNEC calculation, REACH Guidance requires NOEC or EC10values to be used. Available reliable NOEC or EC10values of 0.5 mg/L forDesmodesmus subspicatus(Huls 1989)and 0.694 mg/L forSelenastrum capricornutumBrooke (1993) were obtained and used for derivation of REACH PNEC calculations described below. EURAR (2002) determined that the EC10results from the Kopf (1997) study was the lowest reliable long-term result for aquatic plants and was used for PNEC calculation.

PNEC Comparison – Freshwater and Marine

There was insufficient valid data to calculate a PNEC using a species sensitivity distribution extrapolation method in 2002. Therefore, EURAR 2002 Report used the Assessment Factor approach, using both freshwater and marine data. Since long-term data were available for three species representing three trophic levels an AF of 10 was used resulting in a PNECwaterof 0.33 ug/L NP. Sufficient reliable data to calculate a PNEC using a species sensitivity distribution extrapolation method was obtained for the purpose of the REACH dossier. The resulting PNECfreshwateris 0.627 ug/L NP and PNECsaltwateris 0.536 ug/L NP. The difference between the EURAR 2002 and REACH PNECs is due to the statistical approach used for the actual calculation (i.e., extrapolation method versus the assessment factor approach) and the addition of other reliable data found published after the EURAR 2002 Report was written.

 

PNEC Comparison – Sediment

EURAR 2002 Report presented a PNECfreshwatersedimentof 0.039 mg/kg and a PNECmarinesedimentof 0.039 mg/kg calculated by using equilibrium partitioning method. One reliable freshwater study (Bettinetti and Provini, 2002) for two freshwater sediment organisms and one reliable marine study (Zulkosky et al., 2002) for one marine sediment organism was obtained for the purpose of the REACH dossier. ENVIRON used these reliable studies and the assessment factor approach were used to determine the PNECfreshwatersedimentof 4.62 mg/kg and PNECmarinesedimentof 1.23 mg/kg NP for the REACH dossier.

PNEC Comparison – STP

There is limited reliable data available concerning the effects of NP to STP organisms. EURAR 2002 reported a PNEC for micro-organisms of 9.5 mg/L NP. No additional reliable data for the toxicity of NP to micro-organisms were obtained. Therefore, the PNEC of 9.5 mg/L NP for STP organisms developed for the purpose of the REACH dossier is in agreement with the value presented in the EURAR (2002).

 

PNEC Comparison – Soil

The EURAR 2002 derived a PNECsoilof 0.3 mg/kg NP based on long-term NOECs for three trophic levels of soil organisms. The PNEC was based on the most sensitive organism (an earthworm) with an assessment factor of 10 applied to the NOEC of 3.44 mg/kg. A number of terrestrial studies published since the 2002 Report were obtained for the purpose of REACH which also meant that three trophic levels, with long-term endpoints, were represented in the dataset. Soil invertebrates were the most sensitive organisms which is consistent with the EURAR 2002 Report. Based on reliable and relevant studies that met adequacy criteria, and application of an assessment factor of 10 to the resulting lowest NOEC/EC10of 23 mg/kg for a Collembolan species, a PNECsoilof 2.3 mg/kg was provided.

ENDOCRINE EFFECTS ASSESSMENT SUMMARY OF EURAR 2002 REPORT

These notes provide a summary of the literature review of studies of the potential for endocrine-mediated effects in aquatic organisms following exposure to nonylphenol. The review is presented in chronological order from the first review for the CSR submitted to ECHA in 2010; an update in 2017 following a review of literature between 2010 and 2017; and a recent update in 2019 following the 6 June 2018 Decision of the Board of Appeal.

1.      Summary of Oestrogenic Effects

It has been shown that nonylphenol can exhibit oestrogenic effects on aquatic organisms. European Union Risk Assessment Report 4-Nonylphenol and Nonylphenol (branched) Risk Assessment Final Report 2002 (EURAR 2002) found that reliable data indicate oestrogenic effects of nonylphenol can occur around 10-20 µg/L. The following discussion is a summary of the studies presented in the peer-reviewed EURAR 2002 Report for nonylphenol. In addition, the Environment Agency for England and Wales made available recent work for nonylphenol, which included a detailed literature review. A summary of the results of oestrogenic effects studies is provided below. Also, a literature search was performed to obtain any reliable relevant data for endocrine disrupting effects of nonylphenol published after September 2008. A summary of those studies are also included.  In vitro studies and studies involving injection-type exposures were not considered environmentally relevant (based on exposure route) and not discussed here. A further update was included in 2019 in response to the 6 June 2018 Decision of the Board of Appeal to include the Schwaiger et al (2002) study on Rainbow trout.

1.1 Fathead Minnow (Pimephales promelas)

Harrieset al (2000) found that serum vitellogenin (VTG) induction occurred at 8.1 – 57.7 µg/L in a paired-breeding experiment. Giesyet al (2002) determined a 42-day NOEC of 3.4 µg/L when measured elevation of plasma E2. Miles-Richardsonet al (1999) found there was no effect on tubercle or fat pad size or survival at 3.4 µg/L nonylphenol when fish were exposed in flow through test for 42 days. Schoenfusset al (2008) determined a 7-day NOEC of 11 µg/L and LOEC of 15 µg/L when measuring VTG induction.

1.2 Rainbow trout (Oncorhynchus mykiss)

Joblinget al (1996) measured VTG concentration and spermatogenesis in adult males exposed to 30 µg/L (nominal) nonylphenol for three weeks. Results indicate VTG increased 100-1000 times more than the control while spermatogenesis was slightly delayed. The 21-day NOEC for reduced testis weight is reported at 20.3 µg/L. Pedersonet al (1999) found a significant induction of VTG when exposing Rainbow trout for 9 days in a flow through system. Ackermanet al (2002) exposed Rainbow trout in a flow through system for one year covering embryonic, larval and juvenile stages for one generation only. A LOEC for VTG was calculated of 1.05 µg/L, however NOECs for higher biological endpoints (sex reversal, intersex and hatching rates) were all >10 µg/l.Ashfieldet al (1998) exposed Rainbow trout in a flow-through system from hatch to early sexual maturity and assessed gonado(ovo)somatic index (OSI) of females. Elevated OSI was significant at the 30 µg/L nonylphenol exposure. Tremblay and Van Der Kraak(1998) found increased VTG in blood plasma at 50 µg/L nonylphenol after juvenile fish were exposed for 21 days. An increase in VTG mRNA was found at 14.14 µg/L byLechet al (1996) when fish were exposed for 72 hrs. Harries et al (2001) measured the ovasomatic index for 2 year old fish exposed to nonylphenol for 18 weeks. Results indicate the NOEC and LOEC to be 8.3 and 85 µg/L nonylphenol, respectively. 

Schwaigeret al (2002) exposed groups of male and female Rainbow trout to nonylphenol at 1 and 10 µg/l for intermittent periods during 4 months prior to spawning. Spawning then took place and the F1 generation were grown on to sexual maturity, 3 years later, with no exposure of the offspring. A range of different endpoints for reproductive effects were measured. VTG in the plasma was measured for both the F0 and the F1 generation. VTG concentrations in the males from the F1 generation was found to be greater than the control, but the same was not true of the F0 generation. A NOEC of <1 µg/l for VTG induction is inferred under these particular conditions. The viability of the embryos and mortality of the embryos prior to the egg eyed stage were affected at both exposure concentrations, with a NOEC of <1 µg/l. Hatching success was significantly reduced at 10 µg/l, but not at 1 µg/l, with a NOEC determined of 1 µg/l. No differences were found between the histology of the gonads of the exposed and control fish in terms of maturation and alterations (cell necrosis or inflammation), with a NOEC of >10 µg/l.Van den Beltet al (2003) measured VTG concentrations in rainbow trout exposed to concentrations of nonylphenol of 20, 100 and 500 µg/l for three weeks. Significant induction was measured at 100 µg/l, but not at the lower concentration, with a NOEC determined of 20 µg/l.Madigouet al (2001) exposed egg eyed staged rainbow trout embryos and for an ongoing ten days to nonylphenol at high test concentrations of 44,000 and 4,400 µg/l. A significant effect on sex reversal based on samples of the gonads assessed at 9 months of age was observed at the highest concentration tested. A NOEC between >4,000 and <44,000 µg/l was determined.

1.3 Other fish species

Allenet al (1999) exposed adult male Flounder (P. flesus) to nonylphenol for three weeks. The 21-day NOEC for reduced testis weight is >24.5 µg/L and reduced liver weight is 7.2 µg/L. Japanese Medaka (Oryzias latipes) were exposed to nonylphenol for 100 days. The NOEC for secondary sex characteristics assessed was 10 µg/L , the NOEC for male papillary processes was 100 µg/L, and the NOEC for testis-ova was 30 µg/L (Balch and Metcalfe, 2003). Gray and Metcalfe(1997) exposed post-hatch Medaka to nonylphenol for 90 days and found 50% of males exposed to 50 µg/L nonylphenol with testis-ova. Yakotaet al (2001) found that secondary male sex characteristics were eliminated when Medaka were exposed to 51.5 µg/L nonylphenol, in a life-cycle test, but not at 17.7 µg/L nonylphenol. Testis-ova were significantly higher in the parent generation exposed at 17.7 µg/L nonylphenol, but not at 8.2 µg/L nonylphenol. Kashiwadaet al (2002) found female-specific protein induction occurred in adult males when exposed to 0.1 – 100 µg/L nonylphenol for 5 weeks. The draft Environment Agency report also lists eight additional studies which assessed the oestrogenic effects of nonylphenol on Medaka. LOECs ranged from 0.1 µg/L for detection of female-specific proteins (Fsp) (Tabataet al, 2001) to 100 µg/L for abnormal gonad and testis-ova (Balch and Metcalfe 2006). However, it should be noted that Tabata et al (2001) did not state the levels of Fsp that were detected in the male fish and whether it was significantly different from controls.  The draft Environment Agency report lists two studies which assessed the oestrogenic effects of nonylphenol on Carp (Cyprinus carpio) and one study with Swordtail fish (X. helleri). No oestrogenic effects were found when carp were exposed to 5.36 µg/L nonylphenol for 28-31 days (Villeneuveet al, 2002). VTG was induced at 4-100 µg/L nonylphenol in a three day exposure to Swordtail fish (Kwaket al, 2001).

1.4 Water Flea (Daphnia magna)

Baldwinet al (1997) investigated the effects of nonylphenol on testosterone metabolism and resulting effects on reproduction in a three week test. After 48 hr nonylphenol exposure to adults and after 3 week exposure to nonylphenol for neonates, daphnids were exposed to14C-labelled testosterone for additional 16 hours. Results indicate nonylphenol concentrations of <25 µg/L nonylphenol could significantly affect androgen metabolism, which may contribute to effects to reproduction.

1.5 Other aquatic organisms

The draft Environment Agency report summarizes studies performed using a crab (Carcinus aestuarii) and Pacific oyster (Crassostrea gigas). The NOEC for VTG induction were 50 µg/L nonylphenol for the crab (Ricciardiet al, 2008). Nice(2005) exposed 3 month old juvenile Pacific oysters (Crassostrea gigas) to 1 and 100 µg/L nonylphenol for 72 hours. Oysters were then removed, rinsed and grown to sexual maturity in a flow-through system. At test termination, shell length and body weight was determined not significantly different from controls. However, study author found significant effects to sperm motility at 1 µg/L nonylphenol exposure. It should be noted that assessment of sperm motility (i.e., length of time, movement) in fish has been criticized for subjectivity byKime and Nash(1999). The study author believed the method of assessment used in the study (motile or non-motile only) was adequate, however. No other reliable studies assessing sperm motility for nonylphenol exposed organisms were available for comparison. Therefore, study results should be taken cautiously.

2.      Post-September 2008 Literature Review

Lyeet al (2008) exposed intermoult male Shore crabs (Carcinus maenas) to measured concentrations of 10 and 100 µg/L nonylphenol for 12 weeks in a static-renewal system. Although no significant mortality occurred, significant effects were detected for gonad weight at 10 µg/L nonylphenol exposure. Significant increase in the hepatosomatic index was seen in the 10 µg/L exposure and a significant decrease in ecdysone equivalents in the 100 µg/L nonylphenol exposure. However, no induction of VTG was seen at either concentration.Kortneret al (2009) exposed immature Atlantic salmon (salmo salar) to nonylphenol for 72 hrs. It was found that VTG mRNA in the liver significantly increased and MRNA levels of Cyp19a, step involved in estrogen production, was significantly decreased when exposed to 50 µg/L nonylphenol. This study indicates that oestrogenic effects can be observed on mRNA. The recentWatanabeet al (2017) study of nonylphenol exposure to Japanese medaka (Oryzias latipes) followed the OECD 240 test guideline Medaka extended one generation test (MEOGRT). The population-relevant endpoints are reported and relied upon in the species sensitivity distribution in aquatic PNEC development. Additional results were available for biomarker responses with the most sensitive endpoints (lowest NOECs) being reported for the hepatosomatic index (HSI), gonadosomatic index (GSI), VTG, secondary sexual characteristics (SSC), intersex and gonad phenotype, as follows:

- HSI NOECs of 1.27 µg/l for F0 females, F1 embryo and F1 adult males;

- GSI NOECs of 2.95 µg/l for F0 females and F1 adult males;

- VTG NOECs of 2.95 µg/l for F0 and F1 embryo males;

- SSC NOECs of 9.81 µg/l for male F1 embryo and adults;

- Intersex NOECs of 9.81 µg/l for male F1 embryo and adults; and

- Gonad phenotype NOEC of 9.81 µg/l for F1 adults.

Results of these studies are in agreement with studies reported and summarised in the EURAR (2002) and the draft Environment Agency report. Reliable studies investigating oestrogenic effects of nonylphenol to terrestrial organisms were not found.

3.      Response to 6 June 2018 Decision of the Board of Appeal

Part D of Section II of the Contested Decision on ‘Information for environmental (predicted no effect concentration, PNEC), requires that the transgenerational effects as described by Schwaiger et al (2002) on Rainbow trout (Oncorhynchus mykiss) are taken into account in derivation of a long-term NOEC. In response to this request, information on several studies have been added to this review of potential endocrine-mediated effects in published studies, including Rainbow trout species in Schwaiger et al (2002) and two other studies Van den Belt et al (2003) and Madigou et al (2001) that were also found in the literature.

Nine studies were available for the assessment of potential endocrine effects in rainbow trout (Oncorhynchus mykiss) (section 1.2 of this note). The studies investigated effects across the different biological pathways from cellular (e.g. induction of vitellogenin) to organism level (e.g. reduced fecundity). The reported NOECs are variable, ranging from <1µg/lto >4,400µg/l.The weight of evidence is such that the majority of studies report that no adverse effects would be expected at the PNEC freshwater (0.644µg/l) for nonylphenol. Most of the studies relate to the organ or cellular level and have not led to an adverse outcome that could influence PNEC derivation. Only two studies demonstrate NOECs in the vicinity of the PNEC; these are Schwaiger et al, 2002 and Ackermann et al, 2002. Ackermann et al (2002) is a Klimisch 1 study covering embryonic, larval and juvenile stages for one generation only. VTG induction occurred at <1 µg/l (a biomarker response only) and all endpoints assessed (sex reversal, intersex, hatching rates) resulted in NOECs of >10 µg/l. Schwaiger was a co-author on the Ackermann et al (2002) study. The Schwaiger et al (2002) study was multi-generational with adverse effects observed for the viability of the second-generation embryos at <1 µg/l. This aspect was not covered in the Ackermann et al (2002) study. These NOECs are unbounded, which introduces uncertainty, but where there is uncertainty the weight of evidence is with the other studies that show that the PNEC is protective of potential endocrine-mediated effects. It is considered that the PNEC of 0.644 µg/l is protective and provides a margin of safety for the effects observed in the Schwaiger et al (2002) and Ackermann et al (2002) studies.

4.      Overall conclusion

EURAR 2002 found that reliable data indicate oestrogenic effects of nonylphenol can occur around 10-20 µg/L. Although the subsequent individual test data (Nice 2005, Tabata et al, 2001) presents effects seen at lower concentrations study, however reliability is questioned by Ramboll. The Schwaiger et al (2002) study indicates a NOEC for the viability of second generation embryonic effects at <1µg/l, however lower concentrations were not investigated so a bounded NOEC was not determined. It is considered that the PNEC of 0.644 µg/l offers sufficient margin of safety to be protective of potential effects, such as those observed in the Schwaiger et al (2002) study, given the weight of the evidence for the rainbow trout indicates reproductive effects only occur at higher concentrations (section 1.2).

The calculated PNEC freshwater and PNEC marine as presented in this CSR Report are 0.644 µg/L and 0.548 µg/L and therefore protective of potential oestrogenic effects related to nonylphenol exposure.

 

Conclusion on classification

Nonylphenol does not fulfil the screening criteria for persistence (P-criterion) and bioaccumulation (B-criterion). Although the T-criterion is fulfilled for aquatic organisms (but not for bird toxicity) the overall conclusion is that nonylphenol does not meet the PBT or vPvB criteria. No further testing or an emission characterisation and risk characterization for PBT/vPvB substances in accordance with REACH Article 14(4) is required.