4,4'-isopropylidenediphenol

Long-term toxicity to fish

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

fish, juvenile growth test

Type of information:

experimental study

Adequacy of study:

key study

Reliability:

1 (reliable without restriction)

Qualifier:

according to guideline

Guideline:

OECD Guideline 215 (Fish, Juvenile Growth Test)

GLP compliance:

yes

Analytical monitoring:

yes

Vehicle:

no

Test organisms (species):

Oncorhynchus mykiss (previous name: Salmo gairdneri)

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

28 d

Nominal and measured concentrations:

Measured: 106, 288, 1200, 3640, 11,000 µg/L

Reference substance (positive control):

no

Key result

Duration:

28 d

Dose descriptor:

NOEC

Effect conc.:

3.64 mg/L

Nominal / measured:

meas. (arithm. mean)

Conc. based on:

not specified

Basis for effect:

growth rate

Duration:

28 d

Dose descriptor:

LOEC

Effect conc.:

11 mg/L

Nominal / measured:

meas. (arithm. mean)

Conc. based on:

not specified

Basis for effect:

growth rate

Validity criteria fulfilled:

not specified

Conclusions:

In an OECD 215 Guideline Study the effect Bisphenol A on Rainbow trout (Oncorhynchus mykiss) was assessed. A NOEC = 3.64 mg/L and LOEC = 11 mg/L on growth rate were derived.

Executive summary:

In an OECD 215 Guideline Study the effect Bisphenol A on Rainbow trout (Oncorhynchus mykiss) was assessed. A NOEC = 3.64 mg/L and LOEC = 11 mg/L on growth rate were derived.

Reference 2

Endpoint:

fish early-life stage toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

June 25 to July 30, 1998

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

comparable to guideline study

Qualifier:

equivalent or similar to guideline

Guideline:

OECD Guideline 210 (Fish, Early-Life Stage Toxicity Test)

Principles of method if other than guideline:

Deviations or differences to OECD 210 not specified.

GLP compliance:

yes

Specific details on test material used for the study:

Bisphenol A (4,4'-isopropylidene diphenol, CAS RN 80-05-7)
> 99 % purity

Analytical monitoring:

yes

Details on sampling:

Samples were collected at least weekly during each phase of testing.

Vehicle:

no

Details on test solutions:

This study was run with a dilution water control and a positive control together with nominal Bisphenol A concentrations of 1.0, 10, 100, 320, and 640 μg/L. A 256 mg/L stock concentrate was prepared by dissolving a nominal 4.16 g of Bisphenol A in 1000 ml of dechlorinated water. This level of Bisphenol A is approaching the solubility limit, but solubility of Bisphenol A is pH dependent. Therefore, to ensure complete solubilisation, the stock concentrate was basified during preparation. To achieve this basification the stock concentrate was stirred for approximately 5 minutes, during which period 15 ml of 1M NaOH were added, followed by 10 minutes ultrasonic treatment. The resulting solution was added to approximately 14 litres of dechlorinated water and the volume made up to 16.25 litres. Overhead mechanical stirring was introduced and left to run overnight. The resulting solution of the stock concentrate was clear and colourless. A series of stock solutions was prepared by addition of the required volume of stock concentrate (256 mg/L) to 10 L glass jars. The positive control stock solution was prepared by adding a nominal 380 μg of 17β-estradiol to 1000 mL of dechlorinated water. The solution was stirred for approximately 30 minutes, followed by 10 minutes ultrasonic treatment, followed by another 30 minutes of stirring. The resulting clear and colourless solution was then diluted to 9.6 litres in the 10 L glass stock jar and placed on the test apparatus. These stock solutions were continuously stirred during the test and were delivered at a nominal flow rate of 0.625 mL/min to the mixing chambers. The flow rates of the stock solutions, and of the dilution water, were measured on day 0 and three times per week thereafter. The nominal dilution achieved at this stage, immediately before delivery to the exposure concentrations and solvent control test tanks, was 400 times. The stock solutions were replenished at 10 day intervals (on exposure days -5, 5, 15, 25, and 35).

Test organisms (species):

Pimephales promelas

Details on test organisms:

The fathead minnow (Pimephales promelas) embryos (batch no. 98/44) used in this study were obtained from two brood stocks (batch nos 98/27 and 96/86) held at Brixham Environmental Laboratory. The adult brood stock (batch no. 98/27) were purchased from Aquatic Research Organisms, PO Box 1271, 1 Lafayette Road, Hampton, New Hampshire 03842, USA and were held at Brixham Environmental Laboratory since 08 April 1998. The batch of fish showed no evidence of disease, therefore, no treatment was necessary. The fish were acclimated to 25 ± 2°C on 8 April 1998 and were held at that temperature until the start of the study. The adult brood stock (batch no. 96/86) were bred on 20 November 1996 from stock held at the Brixham Environmental Laboratory. The batch of fish showed no evidence of disease, therefore, no treatment was necessary. The fish were acclimated to 25 ± 2°C and were held at that temperature until the start of the study. The adult brood stocks were fed daily on a basic diet of Keystart®, (a proprietary brand of fish food). This diet was supplemented with frozen brine shrimp (Artemia salina) for the batch of fish purchased from Aquatic Research Organisms and with live Artemia followed by frozen brine shrimp (Artemia salina) for the Brixham bred batch of fish. The mortality of the adult brood stock was recorded as 1% within seven days prior to starting this study. The brood stocks were held in 45 L glass tanks under artificial lighting and were contained in water as described for the test dilution water, except that the water was only filtered to 10 μm before use. The temperatures of the dilution water in the holding tanks were monitored daily. The pH, conductivity, alkalinity, hardness, free and residual chlorine and total ammonia concentration of the Laboratory dechlorinated water supply were monitored periodically. The brood stocks were separated into breeding groups with 5 to 10 fish per tank. The eggs from the spawnings in 15 separate brood stock tanks were collected and pooled in a dish filled with dilution water. Each batch of eggs was less than 24 hours old, which includes the stage of development up to, and including, the early-embryo stage. Sets of five eggs were randomly selected, microscopically examined for viability, and placed in incubation cups. This process was repeated until each incubation cup contained a nominal 25 randomly selected eggs.

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

36 d

Remarks on exposure duration:

Length of test: 32 days post-hatch (36 days overall)

Hardness:

The alkalinity and hardness of the dilution water ranged from 17.4 to 25.0 and from 40.3 to 47.3 mg/L as CaCO3 respectively.

Test temperature:

24.1 to 25.4°C

pH:

7.25 to 8.06

Dissolved oxygen:

7.4 to 8.2 mg l-1

Salinity:

0.5 to 1‰

Conductivity:

Conductivity ranged from 210 to 225 μS cm-1.

Nominal and measured concentrations:

nominal: 1, 10, 100, 320, 640 µg/L
mean, time-weighted measured conc.: 1, 9.3, 84, 254, 579 µg/L (= 80-100 % of nominal)
positive control: nominal 100 ng/L, measured < 0.44 µg/L

Details on test conditions:

The test was conducted at a temperature of 25 ± 1°C with a photoperiod of 16 hours light, followed by 8 hours of dark with a 20 minute dawn/dusk transition period. The light intensity was measured by cosine receptor during the study. These measurements were made midway between the front and back of the apparatus, using a Skye Instruments unit SKL300 photometer with a matching sensor head type SKL310. Readings were taken at the centre and near the two ends of the test rig to determine the range of values along the entire length.

Reference substance (positive control):

yes

Remarks:

17B-estradiol

Key result

Duration:

36 d

Dose descriptor:

NOEC

Effect conc.:

>= 640 µg/L

Nominal / measured:

meas. (not specified)

Conc. based on:

test mat.

Remarks:

based on nominal concentrations.

Basis for effect:

other: hatchability, survival, growth (length and weight)

Results with reference substance (positive control):

Hatchability, larval survival, length: no significant effects.
weight: The positive control data showed a significant effect (P=0.05) when compared against the dilution water control.

Reported statistics and error estimates:

The percentage hatch and survival data were analysed by 2 × 2 contingency table tests to compare the exposure concentrations and positive control against the dilution water control, looking for differences at the 5% significance level (Ref 6). The relative standard deviation (RSD) was defined as the standard deviation expressed as a percentage of the mean. The RSDs of the lengths and weights of the larvae in the dilution water controls were calculated to determine the acceptability of the data according to the United States EPA Environmental Effects Guidelines (Ref 1). The 32 day post-hatch larval length and weight data for the dilution water, positive control and exposure concentration replicates were entered into electronic data files and analysed using statistical procedures in the Brixham Environmental Laboratory computer program ‘LIFESTATS’ (programme version 3.3). The data (all replicates) for larval length and weight were tested for normality (Ref 7) and homogeneity of variance (Ref 8). If the data had a normal distribution and the variance was homogeneous then the individual replicate data were analysed using analysis if variance techniques. If the data did not meet the assumptions for analysis for variance, then the individual replicate data were analysed using a non-parametric procedure such as Wilcoxon’s Rank Sum Test.

Nominal concentrations of 1.0, 10, 100, 320, and 640 ug/L corresponded to mean time-weighted measured concentrations of 1.0, 9.3, 84, 254, and 579 ug/L, respectively. The study is reported based on the nominal concentration. Embryo hatch varied from 88 to 100 %. Survival of the larvae, 32 days post-hatch, varied from 72 to 95 %. NOEC values were determined for each endpoint and showed the NOEC for hatchability, survival (32 days ph), length (32 day ph), and wet weight (32 days ph) to be 640 ug/L in all cases. Analysis of the vtg concentrations showed only a significant increase at the 320 ug/L, and a high degree of within treatment variation. No dose-response was evident in the vitellogenin concentration although it was statistically higher in the 320 ug/L dose. There was great variability in the vtg measurement.

Validity criteria fulfilled:

not specified

Conclusions:

The study of Caunter, 1999, follows OECD 210, is GLP documented and is fully reliable (Klimisch 1). In the FELS study Pimephales promelas fish were exposed to BPA 1, 10, 100, 320, and 640 µg/L, to 17ß-estradiol as positive control, and a negative control. The endpoints hatch success, survival, and growth are relevant. No effects of BPA were observed up to the highest test concentrations: NOEC = 640 µg/L. The study is fully reliable and appropriate for use in regulatory processes.

Executive summary:

A fathead minnow early life stage (ELS) study was performed in general according to OECD 210 guideline. During the ELS, a flow through test apparatus was used to deliver five concentrations of Bisphenol A ranging from 1 to 640 µg/L. Embryo hatch and survival of the larvae were recorded until 32 days post hatch. The fish were then sacrificed and their length and wet weight were recorded. Nominal concentrations of 1.0, 10, 100, 320, and 640 µg/L corresponded to mean time-weighted measured concentrations of 1.0, 9.3, 84, 254, and 579 µg/L, respectively. Analysis of the vitellogenin concentrations showed only a significant increase at the 320 µg/l, with no dose-response and a high degree of within treatment variation. The study is reported based on the nominal concentration. NOEC values were determined for each endpoint and showed the NOEC for hatchability, survival (32 days ph), length (32 day ph), and wet weight (32 days ph) to be 640 µg/L in all cases.

Reference 3

Endpoint:

fish life cycle toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

From 29 September 2004 to 3 March 2006

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

equivalent or similar to guideline

Guideline:

EPA OPP 72-5 (Fish Life Cycle Toxicity)

GLP compliance:

yes

Analytical monitoring:

yes

Vehicle:

no

Test organisms (species):

Pimephales promelas

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

164 d

Remarks on exposure duration:

The fish were unpaired during a 42-day pre-spawn period.

Reference substance (positive control):

no

Key result

Duration:

164 d

Dose descriptor:

NOEC

Effect conc.:

130 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

other: male growth (length)

Key result

Duration:

164 d

Dose descriptor:

NOEC

Effect conc.:

130 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

other: male growth (wet weight)

Key result

Duration:

164 d

Dose descriptor:

NOEC

Effect conc.:

>= 567 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

other: female growth (length)

Key result

Duration:

164 d

Dose descriptor:

NOEC

Effect conc.:

>= 567 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

other: female growth (wet weight)

Remarks:

i.e. male survival

There were no statistically significant effects at any treatment level in males or females with respect to survival, growth, fecundity, hatchability, and gonadosomatic index, with the exception of reduced survival in males at 640 ug/L. With respect to supplemental endpoints evaluated in this study, there was a statistically significant increase in vitellogenin at 64 ug/L and higher in both males and females, compared to controls. Gonadal histopathology showed a statistically significant increase in intravascular proteinaceous fluid in females, with minimal to mild changes at 640 ug/L, and in males with minimal to moderately-severe changes at 160 and 640 ug/L. A statistically significant shift towards less mature gametogenic cell types (relative cell frequency as compared to controls) was observed in females at 640 ug/L and in males at 160 ug/L and 640 ug/L. Analytical confirmation of dose established the geometric time-weighted mean measured concentrations during the 164 day study to be <0.293 (MQL), 1.19, 13.4, 52.8, 130, and 567 ug/L.

Validity criteria fulfilled:

yes

Conclusions:

After a 164d-study on fathead minnows, a NOEC of 160 µg/l (based on the active ingredient) on the basis opf adult mortality was determined.

Executive summary:

To evaluate potential effects of chronic exposure, fathead minnow (32 pairs/treatment) were exposed for 164 days to 0, 1, 16, 160, and 640 µg/L Bisphenol A in a flow-through treatment system. The fish were unpaired during a 42-day pre-spawn period. At 42 days, male and female fish were paired and separated within the tanks, with 4 pairs per tank and 8 tanks per treatment group. Eggs were monitored daily and hatchability trials were performed mid-phase (between days 85 and 105) and late phase (between days 135 and 155) with 50 eggs from at least three separate spawning pairs from separate replicate chambers within a treatment. During sacrifice fish were weighed and measured. Blood was taken caudally for vitellogenin measurement. Fish were necropsied and ovaries and testes were individually weighed and preserved. Endpoints examined were survival, length, weight, number of eggs, number of spawns, eggs per spawn, hatchability, vitellogenin, gonadosomatic index, gonadal histopathology and gametogenic cell counting. Analytical confirmation of dose established the measured concentrations during the 164 day study to be <0.293 (MQL), 1.19, 13.4, 52.8, 130, and 567 µg/L. Concentrations are reported on the nominal levels used in the study.

There were no statistically significant effects at any treatment level in males or females with respect to survival, growth, fecundity, hatchability, and gonadosomatic index, with the exception of reduced survival in males at 640 µg/L. With respect to supplemental endpoints evaluated in this study, there was a statistically significant increase in vitellogenin at 64 µg/L and higher in both males and females, compared to controls. Gonadal histopathology showed a statistically significant increase in intravascular proteinaceous fluid in females, with minimal to mild changes at 640 µg/L, and in males with minimal to moderately-severe changes at 160 and 640 µg/L. A statistically significant shift towards less mature gametogenic cell types (relative cell frequency as compared to controls) was observed in females at 640 µg/L and in males at 160 ug/L and 640 µg/L.

In summary, population relevant reproduction endpoints of survival, growth, fecundity and hatchability were not impacted at any concentration. Contrary to previous study results (Sumpter, 2001), changes to the distribution of testicular cell types only occurred at the highest treatment levels tested in this robust study. The ecologically relevant NOEC based on male survival was 160 µg/L.

Reference 4

Endpoint:

fish life cycle toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

not detailed

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

guideline study

Qualifier:

equivalent or similar to guideline

Guideline:

EPA OPP 72-5 (Fish Life Cycle Toxicity)

GLP compliance:

yes

Analytical monitoring:

yes

Vehicle:

no

Test organisms (species):

Pimephales promelas

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

444 d

Reference substance (positive control):

no

Key result

Duration:

444 d

Dose descriptor:

EC10

Effect conc.:

0.089 mg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Remarks:

reported as nominal

Basis for effect:

other: hatchen (F2 generation)

Key result

Duration:

431 d

Dose descriptor:

NOEC

Effect conc.:

16 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Remarks:

reported as nominal

Basis for effect:

number hatched

Remarks:

F2 generation

Remarks on result:

other: The NOEC of 16 µg/L was determined in Caunter 2000 and Sumpter 2001 (which included the histology) study reports but not reported in the Sohoni, 2001 publication.

Nominal Test Concentrations: 1, 16, 160, 640 and 1280 ug/L. Mean Analyzed Test Concentrations ranged between 70 and 96 % of nominal and are reported as nominal concentrations in the report. BPA exposure had no effect on survival in the F0 generation fish. There was no lasting effect on weight or length in females exposed to BPA while males exposed to concentrations of 640 and 1280 ug/L had significantly (p<0.05) reduced body length and weight compared to the control group. Linear regression found a negative dose-response relationship between BPA exposure and somatic growth (length and weight). Vitellogenin, histology, and GSI were reported in the Sumpter et al., 2001, summary of the multigeneration study as those endpoints were evaluated at Brunel University. Plasma vitellogenin was significantly elevated in females only at the final day of testing in the 640 ug/L treatment group. The induction of vitellogenin production in males after BPA exposure was much more pronounced. Vitellogenin concentrations in males exposed to 160, 640, and 1280 ug/L were significantly elevated over the control group at all assessment days with the exception of 160 ug/L after day 43. The GSI indicated a dose-related inhibition of ovary growth in females with significant inhibitions at 640 and 1280 ug/L. After 164 days, the GSI in males exposed to 640 and 1280 ug/L were significantly reduced compared to the control groups. A dose-response effect was seen in the number of different sex cell types in male testis. BPA exposure increased the number of the most immature cells (spermatogonia) and decreased the number of mature, spermatozoa, cells. No instances of intersex cells were observed. There were no deleterious effects on fecundity in the F0 generation females exposed to BPA concentrations of 1, 16, 160, or 640 ug/L, however those exposed to the highest concentration, 1280 ug/L, were significantly reduced compared to the dilution water control. Hatching rates were unaffected in the F1 generation in eggs exposed to concentrations at or below 160 ug/L but those exposed to 640 and 1280 ug/L were significantly reduced. Eggs in the 1280 ug/L treatment group had a hatching rate of 0%. In conclusion, the LOEC for 164 days of BPA exposure in males with respect to vitellogenin induction was 160 ug/L. Vitellogenin levels in males exposed to 640 and 1280 ug/L were 3300-3900 fold higher than levels found in the control groups. At high concentrations, gonadal growth was inhibited in both male and female fathead minnows. The LOEC for sex cells in male testes after 164 days of exposure was 16 ug/L. This effect may not be estrogenic, but instead a sublethal effect of general toxicity. Sex cell inhibition did not appear to impair the ability of males to fertilize eggs. A high variability was reported in egg production for all treatment groups, including the control, yet a distinct reduction in egg fecundity was observed in the highest BPA treatment group. Hatchability of the F1 generation was more sensitive to BPA than the F0 generation. Based on these results, the No Observed Effect Concentration (NOEC) and Lowest Observed Effect Concentration (LOEC) for hatchability of the F1 generation are 160 and 640 μg/L respectively. The NOEC and LOEC for hatchability of the F2 generation are 16 and 160 μg/L respectively. BPA concentrations that negatively affected reproductive performance were between 3 and 6 orders of magnitude higher than those typically found in natural waters, leading the Sumpter et al., 2001 authors to conclude that BPA is weakly estrogenic but is not likely to cause a major environmental problem for fish in most aquatic systems.

Replicate treatment groups were pooled for statistical analysis. Weaknesses in the spermatogenesis portion of this study (eg. number of fields counted, number of cell types assessed) result in not being able to use that data to set a NOEC based on the histology data alone.

Validity criteria fulfilled:

not specified

Conclusions:

The NOEC and LOEC for this study were based on F2 hatchability and were reported as 16 and 160 µg/L, respectively. Based on the dose-response data given by the authors, a robust EC10 was calculated based on the hatching (F2 generation). The EC10 was determined to be 0.089 mg/l as calculated by the Trap software. EC10 values were derived in accordance with Moermond et al. (2016) and ECHA (2008).

Executive summary:

The Sumpter, 2001, study is the result of a multigenerational exposure (444 days; 431 days exposure to BPA) of fathead minnow to Bisphenol A. Nominal concentrations were 1, 16, 160, 640, and 1280 µg/L and the mean analysed concentrations were 70 to 96 % of nominals throughout the study. This study should be considered in tandem with Sumpter, 2001, in which is reported gonadal growth, biochemical measures, and histopathological findings from the same exposure. Bisphenol A exposure had no effect on survival in the F0 generation fish. There was no lasting effect of weight or length in females exposed to Bisphenol A while males exposed to concentrations of 640 and 1280 µg/L had significantly reduced body length and weight compared to controls. There were no deleterious effects on fecundity in the F0 generation females exposed to BPA concentrations of 1, 16, 160, or 640 µg/L, however those exposed to the highest concentration, 1280 µg/L, were significantly reduced compared to the dilution water control. Hatching rates were unaffected in the F1 generation in eggs exposed to concentrations at or below 160 µg/L but those exposed to 640 and 1280 µg/L were significantly reduced. Eggs in the 1280 µg/L treatment group had a hatching rate of 0 %. Overall, the NOEC and LOEC for this study were based on F2 hatchability and were reported as 16 and 160 µg/L, respectively. Based on the dose-response data given by the authors, a robust EC10 was calculated based on the hatching (F2 generation). The EC10 was determined to be 0.089 mg/l as calculated by the Trap software. EC10 values were derived in accordance with Moermond et al. (2016) and ECHA (2008).

Reference 5

Endpoint:

fish life cycle toxicity

Type of information:

experimental study

Adequacy of study:

key study

Study period:

15 January to 10 May 2010

Reliability:

1 (reliable without restriction)

Rationale for reliability incl. deficiencies:

comparable to guideline study

Qualifier:

equivalent or similar to guideline

Guideline:

EPA OPPTS 850.1500 (Fish Life Cycle Toxicity)

Deviations:

yes

Remarks:

4 replicate design and various minor changes based on current practice, laboratory experience and Hansen et al., 1978

GLP compliance:

yes

Analytical monitoring:

yes

Details on sampling:

Prior to initiation of initial definitive, pretest samples were taken from the low, mid and high treatment levels and the control to confirm that the system was functioning properly and the selected exposure concentrations were established and maintained. In addition, samples of the diluter stock solution were analyzed during the pretest period. During the 40-day initial definitive, the diluter had been in operation and was functioning well.

During the study, one replicate of each treatment level was sampled at test initiation, at hatch and weekly thereafter from one alternating replicate of each treatment level and control. Each exposure solution sample was collected from the approximate midpoint of the test aquaria with a disposable serological pipet. During the pretest and the in-life periods of the exposure, a sample of the stock solution being used at the time of analysis was also analyzed.

Three quality control (QC) samples were prepared at each sampling interval and remained with the set of exposure solution samples throughout the analytical process. The QC samples were prepared at bisphenol A (BPA) concentrations similar to the treatment level range. Results of these analyses were used to judge the precision and quality control maintained during the analysis of exposure solution samples.

Vehicle:

no

Test organisms (species):

Cyprinodon variegatus

Details on test organisms:

Sheepshead minnow (Cyprinodon variegatus) embryos (≤ 30 hours old) were received from Aquatic Biosystems, Fort Collins, Colorado. Upon arrival at Springborn Smithers (15 January 2010), the sheepshead minnow were assigned SSL Lot No. 10A04. The embryos were gradually acclimated to 28 ºC dilution water and held at test temperature until they were placed in the exposure system.
The eggs were impartially distributed to the egg incubation cups in the following manner: Twenty-eight labeled incubation egg cups containing control water were placed in a water bath maintained at 28 ºC. A Carolina dish containing the eggs was placed in the same water bath. The eggs were impartially placed in the egg cups five at a time until each cup contained 50 eggs. The eggs in each egg cup were then examined microscopically to determine fertility and all nonviable eggs were replaced. To initiate the study, the incubation cups, each containing 50 viable eggs, were then suspended in the respective exposure aquaria (one cup per replicate vessel). At study initiation, the embryos were < 30 hours old.

Test type:

flow-through

Water media type:

saltwater

Limit test:

no

Total exposure duration:

116 d

Test temperature:

Daily measurements of temperature in the test aquaria and continuous temperature monitoring established that the exposure solution temperatures ranged from 26 to 29 °C in the upper and lower levels during the 116-day exposure period.

pH:

The pH ranged from 6.7 to 8.2.

Dissolved oxygen:

The dissolved oxygen concentration ranged from 5.1 to 8.3 mg/L.

Salinity:

Salinity measurements ranged from 19 to 22 ‰.

Nominal and measured concentrations:

Nominal concentrations of 9.4, 19, 38, 75, 150 and 300 µg a.i./L, corresponded to time-weighted mean measured concentrations of 7.1, 17, 31, 66, 130 and 250 µg a.i./L. In addition, there was a negative control.

Details on test conditions:

Natural seawater was used as dilution and control water during this study. Seawater was pumped from the Cape Cod Canal, Bourne, Massachusetts from about 1 to 4 meters offshore at a depth of approximately 0.5 meters. The seawater was then transferred by a pump (fiberglass reinforced thermoplastic housing) through polyvinyl chloride (PVC) pipes and transported to the laboratory in a 3400-L fiberglass holding tank. In the laboratory, the seawater was adjusted to a salinity of 20 ± 3 ‰ with laboratory well water and recirculated within an epoxy-coated concrete holding reservoir prior to use. The seawater was pumped under constant pressure through aged PVC pipes to the intermittent flow diluter system. The seawater used for this study had a salinity range of 18 to 22 ‰ and a pH range of 6.9 to 8.7. Routine analyses were conducted periodically on representative samples of the water source for the presence of pesticides, PCBs, and toxic metals by GeoLabs, Inc., Braintree, Massachusetts. None of these compounds have been detected at concentrations that are considered toxic in any of the water samples analyzed, in agreement with ASTM (2002) standard practices. In addition, representative samples of the dilution water source were analyzed monthly for total organic carbon (TOC) concentration. The TOC concentration of the dilution water ranged from 0.92 to 0.98 mg/L for the months of January through May 2010.
An intermittent-flow proportional diluter (Mount and Brungs, 1967) with a dilution factor of 50 % was used to deliver six concentrations of Bisphenol A (BPA) and a dilution water control to four replicates of exposure aquaria. Stock solutions with a nominal concentration of 2.5 mg a.i./mL were prepared and the pH of the stock solution was adjusted to a final pH of 11 by dropwise addition of sodium hydroxide and was sonicated at approximately 40 ºC for approximately one hour. The resulting stock solution was observed to be clear and tan in color with no visible undissolved test substance present following preparation.
An FMI pump was calibrated to deliver 0.477 mL per diluter cycle of the 2.5 mg a.i./mL diluter stock solution to a total volume of` 3.937 L of dilution water contained in the diluter system's mixing chamber. This solution was continuously stirred using a magnetic stirrer and a Teflon®-coated stir bar. The concentration of the solution contained in the mixing chamber was equivalent to that of the highest nominal test concentration (300 µg a.i./L) and was subsequently diluted to provide the remaining nominal treatment levels (150, 75, 38, 19 and 9.4 µg a.i./L). Flow-splitting chambers were used to promote further mixing and to equally distribute the test solutions through glass delivery tubes to the replicate exposure aquaria. The negative control consisted of dilution water only. During the test, 0.5 L per cycle of test solution was delivered to each aquarium at a rate of approximately 410 cycles per 24 hours. This flow provided turnover rate of 7.6 volume replacements every 24 hours, or a 90% replacement time of 7 hours (Sprague, 1969).
The exposure system was two-tiered, consisting of an upper and a lower level water bath (one positioned over the other, Figure 1). Each water bath contained fourteen exposure aquaria which included two of the four replicates for each concentration and the control, or a total of 28 aquaria in the two levels combined. The upper level water bath contained replicates A and B, and the lower water bath contained replicates C and D. The exposure aquaria measured 60 cm x 30 cm x 30 cm. The positions of the 14 aquaria in each water bath were assigned by a random selection of numbers relating to test concentration in each replicate section (designated A, B, C and D) of the system. The test aquaria were fabricated from glass with silicone sealant seams, and were equipped with a 15-cm high end-drain to maintain the test solution volume at 27 liters.
The embryo incubation cups were 5-cm diameter, 8-inch high glass jars with nylon screen bottoms (475-mesh Nitex®). Each aquarium also accommodated a larval growth chamber which measured 30 x 13 x 25 cm and was constructed with a glass bottom and ends and Nitex® screen sides. During the spawning phase, each aquarium was divided into three equal sized compartments. The center compartment of each aquarium contained a spawning chamber which measured 24.5 x 24 x 15 cm and was constructed with glass sides and a 3 x 3 cm mesh bottom that extended up the sides to approximately one third of the total height. Each spawning chamber sat on a removable tray constructed of glass with 3-cm high sides and a 475-µm mesh Nitex® bottom. These trays were designed to secure the eggs produced by the parental fish in the spawning baskets above and provided a means of easily consolidating and collecting spawns.
The temperature of the exposure solutions was maintained at 28 ± 1 °C using heated water baths. A photoperiod of 16 hours of daylight and 8 hours of darkness was maintained throughout the exposure period. Illumination was provided by Sylvania Octron® fluorescent bulbs at a light intensity of 51 to 120 footcandles (550 to 1200 lux). Newly hatched larvae were fed live brine shrimp nauplii three times daily for approximately the first 28 days. A small amount of flaked food was also offered daily after day 14. Juvenile-adult fish were fed twice daily, one feeding of frozen brine shrimp and one feeding of flaked fish food.

Reference substance (positive control):

no

Duration:

4 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

number hatched

Remarks on result:

other: LOEC > 250 µg/L

Duration:

28 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

mortality

Remarks:

F0

Remarks on result:

other: LOEC > 250 µg/L

Duration:

28 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

length

Remarks:

F0

Remarks on result:

other: LOEC > 250 µg/L

Key result

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

66 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

reproduction

Duration:

116 d

Dose descriptor:

LOEC

Effect conc.:

130 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

reproduction

Remarks:

F0

Duration:

55 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

mortality

Remarks:

F0

Remarks on result:

other: LOEC > 250 µg/L

Duration:

55 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

length

Remarks:

F0

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

mortality

Remarks:

F0

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

length

Remarks:

F0 male

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

weight

Remarks:

F0 male

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

length

Remarks:

F0 female

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

weight

Remarks:

F0 female

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

number hatched

Remarks:

F1

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

time to hatch

Remarks:

F1

Remarks on result:

other: LOEC > 250 µg/L

Duration:

28 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

mortality

Remarks:

F1

Remarks on result:

other: LOEC > 250 µg/L

Duration:

28 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

length

Remarks:

F1

Remarks on result:

other: LOEC > 250 µg/L

Duration:

28 d

Dose descriptor:

NOEC

Effect conc.:

250 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

weight

Remarks:

F1

Remarks on result:

other: LOEC > 250 µg/L

Duration:

116 d

Dose descriptor:

other: MATC

Effect conc.:

93 µg/L

Nominal / measured:

meas. (TWA)

Conc. based on:

act. ingr.

Basis for effect:

reproduction

Details on results:

The objective of this 116-day study was to evaluate the long-term (chronic) effects of exposure to Bisphenol A (BPA) on the marine fish, sheepshead minnow (Cyprinodon variegatus). Data were compiled on the effects of exposure on hatching success, survival, growth (total length and wet weight) and reproductive success of first generation (F0) fish (eggs/female/day) and the hatching success, survival and growth (total length and wet weight) of their progeny (F1).

F0 embryos hatched by the fourth day of incubation at all treatment levels and the control. Mean hatching success for embryos exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels averaged 86, 95, 90, 91, 87 and 86 %, respectively. Statistical analysis (Fisher’s Exact Test) determined no significant difference in hatching success among embryos exposed to any of the treatment levels tested compared to the control (94 % hatch).

Following 28 days post-hatch exposure, F0 mean larval survival at all treatment levels ranged from 94 to 99 %, which was not statistically different from the control (96 %), based on Fisher’s Exact Test. The mean total length of larval fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 21.2, 21.7, 22.2, 21.8, 21.7 and 21.7 mm, respectively. Statistical analysis (Williams’ Test) determined no significant difference in mean total length among larval fish exposed to any of the treatment levels tested compared to the control (21.3 mm).

Following 55 days post-hatch, mean larval survival ranged from 92 to 99 % for all treatment levels, which was not statistically different from the control (96 %), based on Fisher’s Exact Test. The mean total length of larval fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 33.4, 33.5, 33.6, 34.1, 32.9 and 33.8 mm, respectively. Statistical analysis (Williams’ Test) determined no significant difference in mean total length among larval fish exposed to any of the treatment levels tested compared to the control (33.9 mm).

At test termination, F0 mean survival at all treatment levels ranged from 93 to 99 %, which was not statistically different from the control (96 %), based on Fisher’s Exact Test.

Mean male total length among fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 45.9, 46.7, 46.0, 47.1, 46.7 and 46.2 mm, respectively. Statistical analysis (Williams’ Test) determined no significant difference in mean male total length compared to the control (46.8 mm). Mean female total length among fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 39.5, 40.4, 39.2, 41.0, 39.4 and 41.5 mm, respectively. Statistical analysis (Williams’ Test) determined no significant difference in mean female total length compared to the control (40.4 mm).

Mean wet weight of males among fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 1.87, 2.01, 1.84, 2.02, 1.98 and 1.95 g, respectively. Statistical analysis (Williams’ Test) determined no significant difference in mean male wet weight compared to the control (1.97 g). Mean wet weight of females among fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 1.09, 1.17, 1.06, 1.22, 1.11 and 1.39 g, respectively. Statistical analysis (Williams’ Test) determined no significant difference in female wet weight compared to the control (1.16 g).
The number of eggs per female per day was 12, 13, 12, 14, 8 and 9 among fish exposed to the 7.1, 17, 31, 66 130 and 250 µg a.i./L treatment levels, respectively. Statistical analysis (Williams’ Test) determined a significant difference in the number of eggs per female per day among females exposed to the 130 and 250 µg a.i./L treatment levels compared to the control (15 eggs/female/day).

The mean percent hatching success ranged from 76 to 94 %. Statistical analysis (Williams’ Test) determined no significant difference in hatching success among F1 fish exposed to any of the treatment levels tested compared to the control (87 %). All embryo groups incubated during this study hatched within 5 to 6 days. Statistical analysis (Kruskal-Wallis’ Test) of time-to-hatch indicated no toxicant-related difference at any treatment level tested compared to the control (5 days).

Following 28 days of post-hatch exposure, F1 larval survival ranged from 89 to 98 % and was unaffected at all treatment levels tested compared to the control (98 %) based on Fisher’s Exact Test. Mean total length of larval fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 22.7, 22.8, 23.1, 22.9, 22.9 and 22.2 mm, respectively. Statistical analysis (Williams’ Test) determined no significant difference in mean total length among larvae exposed to any of the treatment levels tested compared to the control (22.6 mm). Mean wet weight of larval fish exposed to the 7.1, 17, 31, 66, 130 and 250 µg a.i./L treatment levels was 0.173, 0.184, 0.188, 0.179, 0.179 and 0.165 g, respectively. Statistical analysis (Williams’ Test) determined no significant difference in mean wet weight among larvae exposed to any of the treatment levels tested compared to the control (0.170 g).

No effects were noted in any of the parameters measured at the highest concentration tested expect for reproductive success measured as eggs/female/day. Based on F0 reproductive success (eggs/female/day), the MATC of bisphenol A (BPA) and sheepshead minnow was estimated to be 93 µg a.i./L, while the study NOEC was 66 µg a.i./L and the study LOEC was 130 µg a.i./L.

Reported statistics and error estimates:

At test termination, data obtained for each endpoint were analyzed to identify significant reductions in the treatment organisms compared to the control organisms. Analyses were performed using the mean organism response in each treatment group rather than individual response values. All statistical analyses were conducted at the 95 % level of certainty except in the case of Shapiro-Wilks' and Bartlett's Tests, in which the 99 % level of certainty was applied. The 99 % level of certainty is preferred for qualifying tests. Shapiro-Wilks’ Test for normality (Weber, et al., 1989) was used to compare the observed sample distribution with a normal distribution for all endpoints. As a check on the assumption of homogeneity of variance, data for each endpoint were analyzed using Bartlett’s Test (Sokal and Rohlf, 1981).

In all statistical analyses, p values of less than or equal to 0.05 were indicative of a statistically significant effect. Statistical analyses of continuous data (growth and reproduction) were performed using the mean biological response for each treatment level. These data were compared to the control using Williams' Test (Williams, 1971, 1972) or Kruskal-Wallis’ Test (Sokal and Rohlf, 1981) following ANOVA. TOXSTAT® Version 3.5 (Gulley et al., 1996) was used to perform the computations. Statistical analyses of binomial data (survival and hatching success) were performed using Fisher's Exact Test (two-tailed) (SPSS, Inc., 1999).

The Maximum-Acceptable-Toxicant Concentration (MATC) was established by using the most sensitive indicator of toxicity during the full life-cycle exposure of sheepshead minnow. The MATC was expressed as the geometric mean of the lowest test concentration causing adverse effects (LOEC) for that endpoint, and the highest test concentration causing none (NOEC).

Validity criteria fulfilled:

yes

Conclusions:

No effects were noted in any of the parameters measured at the highest concentration tested expect for reproductive success measured as eggs/female/day. Based on F0reproductive success (eggs/female/day), the MATC of Bisphenol A and sheepshead minnow was estimated to be 93 µg a.i./L, while the study NOEC was 66 µg a.i./L and the study LOEC was 130 µg a.i./L.

Executive summary:

The objective of this 116-day study was to evaluate the long-term (chronic) effects of exposure to Bisphenol A on the marine fish, sheepshead minnow (Cyprinodon variegatus).  Data were compiled on the effects of exposure on hatching success, survival, growth (total length and wet weight) and reproductive success of first generation (F₀) fish (eggs/female/day) and the hatching success, survival and growth (total length and wet weight) of their progeny (F₁). No effects were noted in any of the parameters measured at the highest concentration tested expect for reproductive success measured as eggs/female/day. Based on F₀reproductive success (eggs/female/day), the MATC of Bisphenol A and sheepshead minnow was estimated to be 93 µg a.i./L, while the study NOEC was 66 µg a.i./L and the study LOEC was 130 µg a.i./L.

Description of key information

There are five key studies (all reliability 1, reliable without restriction) available which address the long-term toxicity of Bisphenol A to fish. The most sensitive endpoint was egg hatchability in F2 of a 444-day multigeneration study with fathead minnow (Pimephales promelas) according to EPA OPP 72-5 method which  reported a NOEC of 16 µg/L (Sumpter et al., 2001, published in Staples et al., 2011, Ecotox Environ Safety 74:1548-1557) Here, van Sprang et al (2022) also evaluated data to give an EC10 value of 0.089 mg/L based on dose-response data. Results reported from a fish early life stage study (OECD 210) with P. promelas led to significantly higher NOECs based on hatchability, survival, and growth at 640 µg/L in all cases (Caunter et al., 1999, published in Staples et al., 2011, Ecotox Environ Safety 74:1548-1557). With the same test species there were no statistically significant effects in males or females with respect to growth, gonad weight, gonadosomatic index (GSI), or reproduction variables (e.g., number of eggs and spawns, hatchability) at any Bisphenol A concentration in the third key study by Rhodes et al. (2008, published in Mihaich et al., 2012, Environ Toxicol Chem, 31: 2525-2535) which was conducted in accordance with EPA OPP 72-5 method. However, Rhodes et al., 2008, reported of reduced survival with male fish at 640 µg/L (NOEC = 160 µg/L). Bayer AG, 1999, reported a NOEC of 3.64 mg/L in an OECD 215 study with Oncorhynchus mykiss. For marine species, a NOEC for reproductive success (egg/female/day) was observed in a key study investigating the life-cycle exposure to sheepshead minnow (Cyprinodon variegatus) at 66 µg /L and equivalent to EPA OPPTS 850.1500 method (York, 2010, published in Mihaich et al., 2018, Environ Toxicol Chem, 37:398-410).

There are three crucial freshwater life-cycle fish toxicity studies that were retained for PNEC derivation, with the fathead minnow (Pimephales promelas), the zebrafish (Danio rerio), and the Medaka fish (Oryzias latipes).

Sumpter et al. (2001) reported on a multigeneration study exposing the fathead minnow (P. promelas) for 444 days, through two generations. Daily observations of mortality, behaviour and appearance were made and any abnormal effects recorded for the F0, F1, and F2 generation fish. Survival, growth, reproduction, gonadal size, vitellogenin levels, and gonadal histology were evaluated. Overall, the author-reported NOEC based on F2 hatchability in this multigenerational exposure was 16 µg/L, while an EC10 of 89 µg/L has been calculated from the reported dose-response data. Previously, the NOEC of 16 µg/L was used for PNEC derivation purposes. However, Sumpter et al. (2001) also reported a LOEC of 160 µg/L, meaning that the ‘true’ NOEC is between 16 and 160 µg/L. The selection of the NOEC of 16 µg/L is therefore very conservative. Because of such a huge spacing factor in the test concentrations (i.e., factor 10), preference is given to the EC10 value, following the criteria as developed in Chapter 2.

The Japanese Ministry of Environment (2006) also published a multigeneration study exposing the the Medaka fish (O. latipes) to BPA and reported effects on hepatosomatic index, vitellogenin, hatching, mortality, weight and length for the F0 and F1 generation fish. Overall, the NOEC based on the most sensitive biological endpoint, i.e., mortality was 247 µg BPA/L.

Segner et al. (2003) report a full life cycle study in which zebrafish (D. rerio) were exposed to 94 - 1500 μg/L BPA, with analytical confirmation of exposure concentrations. The LOEC for vitellogenin induction and changes in gonad histology was 375 μg/L (the NOEC was 188 μg/L). The LOEC for juvenile growth, time to spawning, mating behaviour, eggs per female and fertilisation success was 1500 μg/L while a NOEC of 750 μg/L was reported for these endpoints. There was no effect on the hatching rate of offspring.

Key value for chemical safety assessment

Fresh water fish

Fresh water fish

Dose descriptor:

EC10

Effect concentration:

0.089 mg/L

Marine water fish

Marine water fish

Dose descriptor:

NOEC

Effect concentration:

0.066 mg/L

Additional information

There are three key long-term toxicity studies for fathead minnow (Pimephales promelas), one with Oncorhynchus mykiss and one life-cycle study with an estuarine fish, the sheepshead minnow (Cyprinodon variegatus). Caunter et al. (1999, published in Staples et al., 2011, Ecotox Environ Safety 74:1548-1557) reported the results of a 36-day early life stage study with fathead minnow which were exposed to 1.0, 10, 100, 320, and 640 µg/L nominal concentration. The NOEC determined in that study based on hatchability, survival and growth was 640 µg/L. Sumpter et al. (2001, published in Staples et al., 2011, Ecotox Environ Safety 74:1548-1557) reported on a multigeneration study exposing the fathead minnow (Pimephales promelas) for 444 days, through two generations, to a dilution water control and nominal Bisphenol A concentrations of 1.0, 16, 160, 640, and 1280 µg/L. Daily observations of mortality, behaviour and appearance were made and any abnormal effects recorded for the F0, F1, and F2 generation fish. Survival, growth, reproduction, gonadal size, vitellogenin, and gonadal histology were evaluated. Gonadal evaluations and vitellogenin concentrations were reported in the companion study (Sumpter et al., 2001) but the results did not reveal an even more sensitive outcome than observed in the main Sumpter et al. (2001) study which was hatchability in F2. Due to weaknesses of the spermatogenesis assessment in the supporting study by Sumpter et al. (2001) (e.g. quality of histopathology sections/slides, number of fields counted, number of cell types assessed) it was not possible for the authors to derive a NOEC based on the histology data alone and was rated as reliability 2 study. The reported NOEC of 16 µg/L for vitellogenin in males (Sumpter et al., 2001, companion study) nevertheless supports the same NOEC value for hatchability in F2 (Sumpter et al., 2001, main study, published in Staples et al., 2011, Ecotox Environ Safety 74:1548-1557).

Rhodes et al., 2008, the third key study, is a follow-up study to the Sumpter et al. (2001) studies and aimed to evaluate the possible effects of Bisphenol A on gonadal cell growth (Rhodes, 2008). In this 164-day study with exposure to 0, 1, 16, 64, 160, and 640 µg/L BPA, Rhodes et al. (2008, published in Mihaich et al., 2012, Environ Toxicol Chem, 31: 2525-2535) reported of no statistically significant effects at any treatment level in males or females with respect to growth, gonad weight, gonadosomatic index, or reproduction variables (e.g., number of eggs and spawns, hatchability), with the exception of reduced survival in males at 640 µg/L (NOEC = 160 µg/L). With respect to supplemental endpoints evaluated in this study, there was a statistically significant increase in vitellogenin at 64 µg/L and higher in both males and females, compared to controls. Gonadal histopathology showed a statistically significant increase in intravascular proteinaceous fluid in females, with minimal to mild changes at 640 µg/L, and in males with minimal to moderately-severe changes at 160 and 640 µg/L. A statistically significant shift towards less mature gametogenic cell types (relative cell frequency as compared to controls) was observed in females at 640 µg/L and in males at 160 µg/L and 640 µg/L. In summary, population relevant reproduction endpoints of growth, fecundity and hatchability were not impacted at any concentration. Contrary to previous study results (Sumpter et al., 2001), changes to the distribution of testicular cell types only occurred at the highest treatment levels tested in this robust study. The ecologically relevant NOEC based on male survival was 160 µg/L.

Bayer AG, 1999, reported a NOEC of 3.64 mg/L in an OECD 215 study with Oncorhynchus mykiss.

There is one key estuarine/marine life-cycle study with the sheepshead minnow (Cyprinodon variegatus; York, 2010, published in Mihaich et al., 2018, Environ Toxicol Chem, 37:398-410). The objective of this 116-day study was to evaluate the long-term (chronic) effects of exposure to Bisphenol A on marine water fish, the sheepshead minnow. Data were compiled on the effects of exposure on hatching success, survival, growth (total length and wet weight) and reproductive success of first generation (F0) fish (eggs/female/day) and the hatching success, survival and growth (total length and wet weight) of their progeny (F1). No effects were noted in any of the parameters measured at the highest concentration tested except for reproductive success measured as eggs/female/day. Based on F0 reproductive success (eggs/female/day), the study NOEC was 66 µg/L and the study LOEC was 130 µg/L.

Further supporting studies (Klimisch 1 or 2) were identified. Sumpter et al., 2001, performed a study according to EPA OPPTS 850.1500 method with P. promelas and reported a NOEC of 16 µg/L (already mentioned before). Yokota et al., 2000, determined a NOEC of 355 µg/L in an OECD 210 study with Oryzias latipes. Segner et al., 2003, reported a 100-d NOEC of 0.75 mg/L for Danio rerio. Kang et al., 2002, in an exploratory non-guideline study with O. latipes reported of a LOEC of  837 µg/L. Honkanen et al., 2004, in an exploratory study with Salmo salar reported of no effects following exposure with 10, 100, and 1000 µg/L (NOEC > 1000 µg/L). Metcalfe et al., 2001, reported of a NOEC of 50 µg/L with O. latipes. Japan MOE, 2004, reported a NOEC of 247 µg/L, also for Oryzias latipes. For the same species, Sun, 2014, reported a NOECs of 60 µg/L. Finally, Kinnberg and Toft, 2003, determined no effect up to 500 µg/L with Poecilia reticulata. Thus, there are many supporting studies which report NOEC in the range of 16 to > 1000 µg/L.

In contrast, several other chronic fish studies which are listed and discussed in this chapter were rated as Klimisch 3 (not reliable) due to major short-comings or Klimisch 4 (not assignable) due to e.g. insufficient documentation and disregarded in the risk assessment. Full justifications for disregard are provided in the endpoint study records and the respective robust study summaries (e.g. Lahnsteiner et al., 2005, Chen et al., 2015, Kwak et al., 2001, deKermoysan et al., 2013, Hatef et al., 2012, Keiter et al., 2012).

In conclusion, there were five key and several supporting studies. The key studies which reported of the lowest effect levels were Sumpter et al., 2001, who reported a NOEC of 16 µg/L for the freshwater compartment and Rhodes et al., 2008, who reported a NOEC of 66 µg/L for the marine water compartment.