Sulphur trioxide

Long-term toxicity to fish

Link to relevant study record(s)

Reference

Reference 1

Endpoint:

fish short-term toxicity test on embryo and sac-fry stages

Type of information:

experimental study

Adequacy of study:

key study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

Published investigative non-guideline, non-GLP study.

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

read-across source

Qualifier:

no guideline followed

Principles of method if other than guideline:

Brook trout (Salvelinus fontinalis), embryos and fry from three sources (an acidic watershed pH 4.7 to 5.3, a neutral watershed pH 7 and a hatchery, pH 7) were exposed separately to lethal and sub-lethal levels of acidity (pH 3.9, .3, 4.7, 5.2 and 7.0), beginning at fertilisation under flow-through conditions in the laboratory.

GLP compliance:

no

Remarks:

no older, published study

Analytical monitoring:

yes

Details on sampling:

pH was monitored daily during the fall and spring when temperatures were warm (>3 degrees C) and weekly during the cold winter period. Measurements were made using an Accumet 156 pH meter (Fisher Scientific Inc.).

Vehicle:

no

Details on test solutions:

Flow was maintained at 9 L/minute. pH was adjusted by dripping stock solutions of either sodium bicarbonate or undiluted reagent grade sulphuric acid into flowing water to produce the nominal pH values required.

Test organisms (species):

Salvelinus fontinalis

Details on test organisms:

Three strains were used:

fish from Peskowesk Brook, an acidic watershed pH 4.7 to 5.3,
fish from Flat Lake, a neutral watershed pH 7
fish from Fraser Mills Hatchery, pH 7

Wild strains were caught and kept in appropriate conditions until eggs were required. Eggs were dry fertilised and then hardened in static test water (previously adjusted to the appropriate pH, 3.9, 4.3, 4.7, 5.2 and 7.0. Fertilised eggs were then disinfected for 10 minutes in 1% "Argentine". Eggs from both the hatchery and Flat Lake were transported to the laboratory whilst surrounded by ice.

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Remarks on exposure duration:

adult fish were held for apporximately 1 month prior to spawning, eggs were maintained until hatch and for approx. 45 days after

Post exposure observation period:

Not applicable

Hardness:

The Yarmouth water, in which the embryos were raised is a soft water. Test water was soft 8.5 mg/L, specific conductance <65 umhos/cm, alkalinity was 2.8 mg/L.

Test temperature:

Temperatures fell from a high of 9 degrees C in November to a low of 1 degrees C during the winter months and rose again near the end of the experiments to a high of 15 degrees C.

pH:

See measured test concentrations for details.

Dissolved oxygen:

Not reported.

Salinity:

Not applicable

Nominal and measured concentrations:

Nominal test pHs were 3.9, 4.3, 4.7, 5.2 and 7.0.

pH was monitored daily during the fall and spring when water temperatures were warm (>3degress C) and weekly during the cold winter period in the trays containing eggs. Mortalities were scored when embryos turned white.

Measured pH levels during the study were 6.99 (calculated by averaging H+ concentration and then transforming to pH, n=48), 5.22 (n=46), 4.32 (n=46) and 3.94 (n=5)

Details on test conditions:

Not applicable

Reference substance (positive control):

no

Dose descriptor:

NOEC

Effect conc.:

0.31 mg/L

Nominal / measured:

meas. (not specified)

Conc. based on:

test mat.

Basis for effect:

larval development

Remarks on result:

other: NOEC duration not specified

Details on results:

A large number of hatchery embryos died shortly atfer fertilisation at the high pHs (11% at pH 5.2 and 23% at pH 7). To remove this artifact, only embryos which were alive one day after fertilisation were included in the survivorship records.

Results with reference substance (positive control):

Not applicable

Reported statistics and error estimates:

Lethal acidity was defined as the pH level where mortality was greater than 90% in all three strains.

At pH 3.9, most embryos were dead within a few degree-days, therefore results for this acidity have not been shown.

The strains differed in their tolerance to acidity with survival following a consistent pattern: Peskowesk>Flat Lake>Hatchery (see table 1). The separation amongst the three strains increased in proportion to the acidity levels and differences in their survivorship became statistically significant in the low pH treatments.

Table 1 - survival at test end

pH	Survival (%)			Significance
	Peskowesk	Flat Lake	Hatchery
7.0	93.1	91.4	73.6	ns
5.2	93.7	88.2	70.2	ns
4.7	75.5	63.6	16.3	P <0.001
4.3	3.7	0.0	0.2	P <0.05

The results of the survival of at the end of the experiment of fish exposed from fertilisation or from 213 degree days is presented in Table 2. For comparison purposes, survival was adjusted to 100% in both groups at 213 -degree days.

Table 2 - Results of fish survival for fish exposed from fertilisation or after 213 degree days.

pH	Survival (%)		Significance
	Treated from fertilisation	Treated from 213 degree days
7.0	89.2	87.9	ns
5.2	83.6	84.6	ns
4.7	63.8	88.7	P <0.05
4.3	2.1	16.1	P <0.01

The strongest acidities, pH 3.9 and 4.3, were lethal to most embryos. At pH 4.3, less than 4% of the original embryos were still alive at test termination. Embryos appeared to be most susceptible to acid stress immediately after fertilisation. At sublethal acidities (pH 4.7 and above) mortality was most pronounced up to 150 to 200 degree-days after fertilisation. Afterwards survival stabilised and mortality rates were similar among the three strains. At pH 4.3, high mortality also occurred during hatching and swim-up but at sublethal pH levels no increase in mortality occurred at these stages. Deformities were infrequent in all cases.

Because embryos appear to be most susceptible immediately after fertilisation, hatchery eyed embryos were obtained from the Fraser Mills Hatchery and added to the pH treatment st at 213 degree days. In the lower pH treatments (pH 4.3 and 4.7) the eyed embryos introduced had significantly better survival at the end of the experiments than did the Hatchery group exposed from fertilisation. Hatching was delayed in all groups as acidity increased.

The no-observed effect concentration (NOEC) for brook trout (based on survival) exposed to sulphuric acid was observed to be pH 5.2, equivalent to 0.31 mg/L.

Validity criteria fulfilled:

not applicable

Remarks:

There are no appropriate study guidelines.

Conclusions:

Based on the reported information an acceptable assessment of toxicity on embryos and sac-fry stages was achieved.

Executive summary:

A chronic toxicity study with the brook trout (Salvelinus fontinalis) was conducted. Fertilised eggs from three sources (wild strain, acidic watershed, pH 4.7 to 5.3, wild strain from  neutral watershed , pH 7.0, and a hatchery strain, pH 7.0) were exposed to a series of test media, with pH adjustment using sulphuric acid, to varying pHs (3.9, 4.3, 4.7, 5.2 and 7.0) under flow-through conditions. 

At pH 3.9, most embryos were dead within a few degree-days. Significant differences in mortality between the strains at low pH were observed and these suggested a genetic component to acid tolerance. Mortality in the strain from the acidic watershed was the lowest, followed by the second wild strain. Survival in both wild strains at low pH was much better than the hatchery embryos. These differences in survival at sublethal acidity (pH 4.7 to 7.0) were principally the result of high mortality shortly after fertilisation. After this period, mortality stabilised. Only at pH 4.3 did substantial mortality occur at hatching. The early embryonic strain therefore appears to be the most susceptible to sublethal acid stress in brook trout. Hatchery strain embryos were also introduced at the eyed stage at 213 degree days. Subsequent survival of this group was better at low pH than that of hatchery embryos introduced at fertilisation.  Higher acidity retarded hatch in all cases. 

The no-observed effect concentration (NOEC) for brook trout (based on survival) exposed to sulphuric acid was observed to be pH 5.2, equivalent to 0.31 mg/L.

Reference 2

Endpoint:

fish, juvenile growth test

Type of information:

experimental study

Adequacy of study:

supporting study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

Published investigative non-guideline, non-GLP study.

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

read-across source

Qualifier:

no guideline available

Principles of method if other than guideline:

See study design details

GLP compliance:

no

Remarks:

older, published study

Analytical monitoring:

yes

Details on sampling:

Not reported

Vehicle:

no

Details on test solutions:

Concentrated sulphuric acid was added to lower the pH in the acidic groups at intervals of 3 - 4 days first from neutral to pH 6.0 and then by 0.5 pH units until the respective nominal pH was reached. After this initial period, pH was maintained by the addition of known amounts of sulphuric acid after the daily water change. During the remainder of the day, the target pH was maintained by dripping diluted sulphuric acid at known rates into the holding tanks.

Test organisms (species):

Salvelinus fontinalis

Details on test organisms:

11 month old brook trout were purchased from a commercial hatchery (approx. 18 cm fork length, 100 g body weight). The fish were maintained in fibre glass tanks under constant flow. Prior to use the water was carbon-filtered, then aerated for 24 hours and mechanical agitation in a 1500L capacity plastic tank. The water in the fish holding tanks was constantly aerated and recirculated through nylon wool or a plastic sieve at a rate of 16 - 32 L/min. One third of the water was replaced daily and the tanks were cleaned monthly. Fish were fed once daily to satiation with a commercial trout food. Lighting was maintained to mimic the natural photoperiod.

Water temperature was controlled by heat exchange with chilled water (3-6 degrees C) circulated through plastic tubing. Water pH 7.01 - 7.61, alkalinity TIP 78 mg/L: aluminium 0.095 mg/L, copper 0.003 mg/L.

Fish were acclimated for two weeks to the laboratory water conditions before they were divided into one neutral and three acidic groups of equal size.

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

10 mo

Post exposure observation period:

Not applicable

Hardness:

Not reported

Test temperature:

In the control group temperature ranged between 7.9 °C in March to a maximum of 15.7°C in June to a low of 8.5°C in November. In the pH 4.5 grouptemperature ranged between 8.8 °C in March to a maximum of 15.5°C in June to a low of 8.2°C in November

pH:

pH was checked every 1-2 hours from 08:30 to 18:00 daily. See measured concentrations.

Dissolved oxygen:

Oxygen levels were checked periodically. Mean overall oxygen concentration was 6.9 ± 0.3 mg/mL (12 measurements) with no apparent difference between the pH groups.

Salinity:

Not applicable

Nominal and measured concentrations:

Nominal pH values were 4.5, 5.0 and 5.5 with a control of pH 7.3

Mean measured pH values were control of 7.34 ± 0.01 (297 measurements), 5.56 ± 0.02 (312 measurements), 5.16 ± 0.03 (290 measurements) and 4.48 ± 0.02 (292 measurements)

Details on test conditions:

Not applicable

Reference substance (positive control):

no

Duration:

10 mo

Dose descriptor:

NOEC

Effect conc.:

0.13 mg/L

Nominal / measured:

meas. (not specified)

Conc. based on:

test mat.

Basis for effect:

weight

Remarks:

juvenile fish weight

Remarks on result:

other: NOEC was 5.56 pH units (equivalent to 0.13 mg/L)

Details on results:

Food consumption rate in all groups was about the same at 128.9 ± 5.6 g trout chow per 100 fish per day in March. Thereafter the quantity of food was no longer accurately measured, no difference in food consumption was noticeable during the subsequent months. Throughout the experiments all fish mortalities appeared to be unrelated to pH exposure.

Somatic growth - within each pH group body weight significantly increased during the experimental period (ANOVA, p<0.001). Between group comparison revealed that towards the latter half of the experimental period the mean body weight of the brook trout in the pH 4.5 group was significantly lower (ANOVA, p<0.05 to <0.001) than that of the control. The mean body weight of pH groups 5.0 and 5.5 were mostly between the control and pH 4.5. At the end of the experiment, mean body weight of the males and females in the pH 4.5 group was 73.18 and 72,82% of the same sex in the control group. See table 1 for details.


Gonad growth, egg production and atresia - the gonado-somatic index (GSI) is the ratio of gonadal to body weight, expressed as milligrams per body weight. In males and females, there were no significant differences (ANOVA) between any group at any time of the experimental period. Gonadal weight began to increase in June and reached maximal values in September and November.
The mean ovulation times of the various groups are expressed relative to October 31 (day 0) to allow comparison. Ovulation in pH groups 4.5 and 5.0 occurred very significantly later than the control. No delay was observed in the males, as spermatozoa could be obtained from approximately 50% of the individuals in all groups by the end of October. Correlation analysis showed that there was no correlation between body weight and the mean weight of the ovulated eggs in any experimental group. Thus, larger fish did not have larger eggs. There was no significant difference in egg weight between the control and any of the pH groups. However, eggs from the pH 4.5 group were significantly heavier than those of pH group 5.0 (ANOVA, p <0.02). There was significant correlation between the total number of mature, yolky eggs and body weights in the control, pH 4.5 and pH 5.0 groups. The number of eggs produced per female increased in a linear fashion with increase in body weight. See Tables 2 and 3 for details.

The extent of atresia was significantly higher in the pH 5.0 and 5.5 than in the control and pH 4.5 groups in September (ANOVA, p < 0.01). In November, only pH group 5.0 had more atric eggs than the other groups (p = 0.004).

Results with reference substance (positive control):

Not applicable

Reported statistics and error estimates:

Analysis of variance (ANOVA) and Student-Newman-Keuls (SNK) a posteriori procedure were used to detect significant differences among body weight, body length, gonadosomatic index, average egg weight, and ovulation time. Where appropriate, Student's t-test was also used to test for significant difference between a pair of means. The Kruskal-Wallis ranked ANOVA was used to detect significant differences in the extent of atresia between fish. Confidence limit was taken at the 5% level.

Table 1 - Regression analysis of changes in body weight in brook trout maintained at pH 7.34 (control) and pH 4.5

Group	Slope			Intercept
	Female	Male	p	Female	Male	p
Control	33.0	38.7	NS	54	33	NS
pH 4.5	47.4	24.8	0.001	-182	67	0.05
p	NS	0.02		0.001	0.001

NS - not significantly different

Table 2 - Mean ovulation time expressed as number of days. 

.

pH	Proportion of fish ovulated	Ovulation time (days after 31 October)
7.34 (control)	3/4	14.3 ± 0.3 (3)
5.16	7/14	30.0 ± 1.1 (7)
4.48	11/16	19.6 ± 0.8 (8)

Results are based on 34 females, all of which possessed mature, yolky eggs

Table 3 - Average egg weight and total number of eggs calculated for each female of average body weight at the end of the ovulation period

pH	Average female weight (g)	Average egg weight (mg)^a	Number of eggs per female of average body weight^b
7.34 (control)	471	29.8±2.7 (7)	2020 (100)
5.16	333	27.6±2.1 (8)	1290 (63.86
4.48	343	33.3±1.1 (13)	1343 (66.49)

^a – group 1 was not significantly different from 4.48 or 5.16 (ANOVA)
^b – number of eggs expressed as % of those in the control in parentheses

Validity criteria fulfilled:

not specified

Remarks:

There are no appropriate study guidelines.

Conclusions:

Based on the reported information an acceptable assessment of pH effects on growth, ovulation and egg production was achieved.

Executive summary:

A chronic toxicity study with the brook trout (Salvelinus fontinalis) was conducted. Hatchery reared fish were maintained in the laboratory at mean pH values of 7.34, 5.56, 5.16 and 4.48 from early February to December 1984.  At pH 4.8, the mean growth rates of males were uniformly lowered during the entire experimental period.  Among females, growth was inhibited during the first 5 months, but their rate of weight gain recovered during the period of rapid oocyte development.  At the end of the experiment, the body weights of both male and female fish in pH 5.16 and 4.48 were only 71 – 77% of the control fish at ph 7.34.  Growth was not affected by exposure to pH 5.56. Rapid oocyte development occurred simultaneously over all pH groups in June, suggesting that the initiation of gametogenesis was not affected over the range of pH tested.  The number of eggs produced was significantly correlated to body weight; consequently the number of eggs produced by the smaller ph 5.16 – 4.48 females was reduced.  Ovulation was also significantly delayed in the acidic groups.

The overall NOEC for this study was determined on the weight of juvenile trout produced in 10 months and was pH 5.56 (equivalent to 0.13 mg/L).

Reference 3

Endpoint:

long-term toxicity to fish, other

Type of information:

experimental study

Adequacy of study:

key study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

Published, investigative non-guideline, non-GLP study

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

read-across source

Qualifier:

no guideline available

Principles of method if other than guideline:

Groups of sexually mature flagfish (Jordanella floridae) were exposed to sub-lethal levels of acidity, pH 4.5, 5.0, 5.5 and 6.0. Spawning success, fry growth and survival were monitored.

GLP compliance:

no

Remarks:

older, published study

Analytical monitoring:

yes

Details on sampling:

Not reported

Vehicle:

no

Details on test solutions:

The pH of incoming lake water was adjusted by additoon of 0.1 N sulphuric acid injected by an automatic pipettor into the mixing cell. The diluter discharged 0.5 L litre of water per test vessle every two minutes

Test organisms (species):

Jordanella floridae

Details on test organisms:

Mature flagfish, less than one year old, were purchased from a commercial supplier and held in laboratory dechlorinated water for two months before being transferred to the Lake Panache testing location. On arrival at the lake site male and female fish were seperated and held in tanks for seven days in holding tanks.

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

65 d

Remarks on exposure duration:

Spawning and fry growth

Post exposure observation period:

Not applicable

Hardness:

Lake water - 28 mg CaC03/L.

Test temperature:

26.5 ± 0.5 °C

pH:

Nominal pH values were ciontrol (6.8), 4.5, 5.0, 5.5 and 6.0

Dissolved oxygen:

Dissolved oxygen was always maintained at over 90% saturation.

Salinity:

Not applicable

Nominal and measured concentrations:

Nominal pH values were control (6.8), 4.5, 5.0, 5.5 and 6.0

pH of each treatment group was reported to have been maintained within 0.13 pH units at nominal values over all exposure periods.

Details on test conditions:

The study was conducted with a 15 hour light photoperiod.

Reference substance (positive control):

no

Remarks:

not applicable

Duration:

65 d

Dose descriptor:

NOEC

Effect conc.:

0.025 mg/L

Nominal / measured:

nominal

Conc. based on:

other: pH

Basis for effect:

other: fry growth

Details on results:

Adult mortality principally occurred within the final 20days of exposure and was about 7% (one female) in the control group, increasing to 21 and 14%at pH levels 5.5 and 5.0, respectively and 64% at pH 4.5, see Table 1.

All treatment groups commenced spawning within three to four days of one another and egg production and fertility at the end of both the 10 day acclimation period and the five day pH depression period was equivalent among all groups and bore no correlation(p>0.05) to pH. Exposure to depressed pH reduced the total number of spawnings (see Table 1) in each group compared to the controls (P<0.05) except for pH 6.0 (p=0.15). No spawnings occurred at pH 4.5. Spawning frequency overall was observed to decrease directly as pH was reduced (P<0.05).

A significant direct correlation (P<0.05) between total eggs produced and pH was found (See Table 1). Mean daily egg output was significantly impaired (P<0.05) at each pH level compared to the controls. Combining the data for spawning frequency and total eggs produced shows that mean spawning size correlated directly with pH. Mean spawning size was significantly reduced compared to the control (P<0.05) at the pH 5.5 treatment.

The number of fertile eggs was reduced (P<0.05) at every exposure level compared to the controls and this related well (P<0.05) with a decline in pH (see Table 1). Hatchability of eggs to the different pH's was variable, ranging from 17-56% and not indicative of overall trends, yet hatching success was somewhat less during pH 5.0 exposure. Eggs collected from control tanks and incubated in pH 5.0 and 4.5 hatched with 82 and 87% success. The control fry hatched under these pH conditions died within 24-56 hours.

The final mean weights of the 45 day old fry were significantly less (P<0.05) after exposure to all pH treatments compared to the controls and the decrease was directly related (P<0.05) to pH depression (see Table 2). The correlation coefficient (r=0.99) of the final mean fry weight and the final mean fry dorsal image area at the different pH treatments was significantly (P<0.05). Therefore, a change in the dorsal image area was considered representative of a change in growth. The growth rates as indicated by changing dorsal image area of the fry held under control and pH 6 conditions were similar (P>0.05) while the growth rates of fry held under pH 5.5 and 5.0 conditions were considerably reduced (P<0.05) compared to the controls. A significant decrease in growth rate (P<0.05) between fry in pH 5.5 and 5.0 was also evident.

The breeding communities that were allowed to spawn for 16 days and were then exposed to pH 4.5 for seven days abruptly stopped spawning for that entire period. One spawning occurred seven days after normal lake water conditions resumed and a second was observed on day 9 of the final 14 day period. The mean daily output was reduced from 78 eggs after 22 spawnings before pH depression to 4.7 eggs from two spawnings after depression. The total number of eggs produced was 1 249 before depression compared to 66 after.

Results with reference substance (positive control):

Not applicable

Reported statistics and error estimates:

Fish measurements (length and dorsal image) were subjected to logarithmic transformation with respect to time and was used for growth analysis.

Spawning frequency, egg fertility and mortality data were subjected to Chi-square analysis, while egg production data was tested for significance by analysis of variance and independent t-distribution. Fry growth rate data were compared by analysis of covariance and the correlation of all reproductive responses with declining ph was tested by regression analysis.

Table 1 – Reproductive response of flagfish exposed to depressed pH for 20 days

Parameter	Nominal pH					Correlation with pH
	Control (6.8)	6.0	5.5	5.0	4.5
Adult mortality	1	0	3	2	11	-
Spawning	32	24	15*	7*	1*	P <0.05
Total eggs	2467	1277	557	319	1	P <0.05
Mean daily output	61.7	31.9*	13.9*	8.0*	0*	P <0.05
Mean spawning size	77.1	53.2	37.1*	45.6	1*	P <0.05
% fertile eggs	90.1	78.6*	68.0*	55.2*	-	P <0.05
% hatching of fertile eggs	44.4	56.5	51.1	17.0	-	P <0.05

* P <0.05 compared to the control

Table 2 – Survival, weights and dorsal image area of flagfish fry exposed to depressed pH for 45 days.

Parameter	Control (6.8)	6.0	5.5	5.0	Correlation with pH
Total surviving fry (N=50)	37	43	20	2	P <0.05
Final mean weight of fry	0.55	0.42*	0.26*	0.075*	P <0.05
Final mean dorsal image area (mm^2)	123.7	100.3*	71.5*	37.7*	P <0.05

* P<0.05 compared to the control

Validity criteria fulfilled:

not specified

Remarks:

There are no appropriate study guidelines

Conclusions:

Based on the reported information an acceptable assessment of pH effects on flagfish reproduction, growth and survival was achieved.

Executive summary:

Breeding communities of flagfish (Jordanella floridae) were exposed to northern Ontario lake water (hardness 28 mg/L) adjusted to pH levels of 6.0, 5.5, 5.0 and 4.5. Control water (pH 6.8) received no acid treatment. Egg production, egg fertility and fry growth was impaired (P0.05) at all exposure levels. Flagfish fry survival was reduced (P0.05) at pH 5.5 and 5.0, no fry survived at pH 4.5. Variability of hatching in all treatments precluded any identifiable hatching response to depressed pH. Reduction in the reproductive processes monitored indicated the following order of sensitivity: egg production > fry survival > fry growth > egg fertility.

The LOEC 20% (lowest observed effect concentration) was pH 6.0 equivalent to 0.049 mg/L. and the NOEC (LOEC/2) was determined to be 0.025 mg/L.

Reference 4

Endpoint:

fish short-term toxicity test on embryo and sac-fry stages

Type of information:

read-across from supporting substance (structural analogue or surrogate)

Adequacy of study:

key study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

Published investigative non-guideline, non-GLP study.

Justification for type of information:

Sulphur trioxide readily reacts with water to form sulphuric acid. The reaction is instantaneous, to the extent that SO3 will react with water vapour in the atmosphere to form fumes of sulphuric acid. This reaction forms the basis of the manufacturing process of H2SO4. The read-across hypothesis is therefore that SO3 will instantaneously transform into H2SO4 upon contact with water (i.e. in aquatic ecotoxicology tests), thus any observed effects will be directly attributable to sulphuric acid. It is therefore justifiable to derive hazard conclusions from sulphuric acid data, with regard to ecotoxicological endpoints.

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

read-across source

Qualifier:

no guideline followed

Principles of method if other than guideline:

Brook trout (Salvelinus fontinalis), embryos and fry from three sources (an acidic watershed pH 4.7 to 5.3, a neutral watershed pH 7 and a hatchery, pH 7) were exposed separately to lethal and sub-lethal levels of acidity (pH 3.9, .3, 4.7, 5.2 and 7.0), beginning at fertilisation under flow-through conditions in the laboratory.

GLP compliance:

no

Remarks:

no older, published study

Specific details on test material used for the study:

Available study data for sulphuric acid is being used for read-across to the target substance, sulphur trioxide.

Analytical monitoring:

yes

Details on sampling:

pH was monitored daily during the fall and spring when temperatures were warm (>3 degrees C) and weekly during the cold winter period. Measurements were made using an Accumet 156 pH meter (Fisher Scientific Inc.).

Vehicle:

no

Details on test solutions:

Flow was maintained at 9 L/minute. pH was adjusted by dripping stock solutions of either sodium bicarbonate or undiluted reagent grade sulphuric acid into flowing water to produce the nominal pH values required.

Test organisms (species):

Salvelinus fontinalis

Details on test organisms:

Three strains were used:

fish from Peskowesk Brook, an acidic watershed pH 4.7 to 5.3,
fish from Flat Lake, a neutral watershed pH 7
fish from Fraser Mills Hatchery, pH 7

Wild strains were caught and kept in appropriate conditions until eggs were required. Eggs were dry fertilised and then hardened in static test water (previously adjusted to the appropriate pH, 3.9, 4.3, 4.7, 5.2 and 7.0. Fertilised eggs were then disinfected for 10 minutes in 1% "Argentine". Eggs from both the hatchery and Flat Lake were transported to the laboratory whilst surrounded by ice.

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Remarks on exposure duration:

adult fish were held for apporximately 1 month prior to spawning, eggs were maintained until hatch and for approx. 45 days after

Post exposure observation period:

Not applicable

Hardness:

The Yarmouth water, in which the embryos were raised is a soft water. Test water was soft 8.5 mg/L, specific conductance <65 umhos/cm, alkalinity was 2.8 mg/L.

Test temperature:

Temperatures fell from a high of 9 degrees C in November to a low of 1 degrees C during the winter months and rose again near the end of the experiments to a high of 15 degrees C.

pH:

See measured test concentrations for details.

Dissolved oxygen:

Not reported.

Salinity:

Not applicable

Nominal and measured concentrations:

Nominal test pHs were 3.9, 4.3, 4.7, 5.2 and 7.0.

pH was monitored daily during the fall and spring when water temperatures were warm (>3degress C) and weekly during the cold winter period in the trays containing eggs. Mortalities were scored when embryos turned white.

Measured pH levels during the study were 6.99 (calculated by averaging H+ concentration and then transforming to pH, n=48), 5.22 (n=46), 4.32 (n=46) and 3.94 (n=5)

Details on test conditions:

Not applicable

Reference substance (positive control):

no

Dose descriptor:

NOEC

Effect conc.:

0.31 mg/L

Nominal / measured:

meas. (not specified)

Conc. based on:

test mat.

Basis for effect:

larval development

Remarks on result:

other: NOEC duration not specified

Details on results:

A large number of hatchery embryos died shortly atfer fertilisation at the high pHs (11% at pH 5.2 and 23% at pH 7). To remove this artifact, only embryos which were alive one day after fertilisation were included in the survivorship records.

Results with reference substance (positive control):

Not applicable

Reported statistics and error estimates:

Lethal acidity was defined as the pH level where mortality was greater than 90% in all three strains.

At pH 3.9, most embryos were dead within a few degree-days, therefore results for this acidity have not been shown.

The strains differed in their tolerance to acidity with survival following a consistent pattern: Peskowesk>Flat Lake>Hatchery (see table 1). The separation amongst the three strains increased in proportion to the acidity levels and differences in their survivorship became statistically significant in the low pH treatments.

Table 1 - survival at test end

pH	Survival (%)			Significance
	Peskowesk	Flat Lake	Hatchery
7.0	93.1	91.4	73.6	ns
5.2	93.7	88.2	70.2	ns
4.7	75.5	63.6	16.3	P <0.001
4.3	3.7	0.0	0.2	P <0.05

The results of the survival of at the end of the experiment of fish exposed from fertilisation or from 213 degree days is presented in Table 2. For comparison purposes, survival was adjusted to 100% in both groups at 213 -degree days.

Table 2 - Results of fish survival for fish exposed from fertilisation or after 213 degree days.

pH	Survival (%)		Significance
	Treated from fertilisation	Treated from 213 degree days
7.0	89.2	87.9	ns
5.2	83.6	84.6	ns
4.7	63.8	88.7	P <0.05
4.3	2.1	16.1	P <0.01

The strongest acidities, pH 3.9 and 4.3, were lethal to most embryos. At pH 4.3, less than 4% of the original embryos were still alive at test termination. Embryos appeared to be most susceptible to acid stress immediately after fertilisation. At sublethal acidities (pH 4.7 and above) mortality was most pronounced up to 150 to 200 degree-days after fertilisation. Afterwards survival stabilised and mortality rates were similar among the three strains. At pH 4.3, high mortality also occurred during hatching and swim-up but at sublethal pH levels no increase in mortality occurred at these stages. Deformities were infrequent in all cases.

Because embryos appear to be most susceptible immediately after fertilisation, hatchery eyed embryos were obtained from the Fraser Mills Hatchery and added to the pH treatment st at 213 degree days. In the lower pH treatments (pH 4.3 and 4.7) the eyed embryos introduced had significantly better survival at the end of the experiments than did the Hatchery group exposed from fertilisation. Hatching was delayed in all groups as acidity increased.

The no-observed effect concentration (NOEC) for brook trout (based on survival) exposed to sulphuric acid was observed to be pH 5.2, equivalent to 0.31 mg/L.

Validity criteria fulfilled:

not applicable

Remarks:

There are no appropriate study guidelines.

Conclusions:

Based on the reported information an acceptable assessment of toxicity on embryos and sac-fry stages was achieved.

Executive summary:

Data on the chronic toxicity to freshwater fish is available for sulphuric acid and is considered suitable for read-across (based on the analogue approach) to the target substance, sulphur trioxide. Sulphur trioxide readily reacts with water to form sulphuric acid. The reaction is instantaneous, to the extent that SO₃ will react with water vapour in the atmosphere to form fumes of sulphuric acid. This reaction forms the basis of the manufacturing process of H₂SO₄. The read-across hypothesis is therefore that SO₃ will instantaneously transform into H₂SO₄ upon contact with water (i.e. in aquatic ecotoxicology tests), thus any observed effects will be directly attributable to sulphuric acid. It is therefore justifiable to derive hazard conclusions from sulphuric acid data, with regard to ecotoxicological endpoints.

A chronic toxicity study with the brook trout (Salvelinus fontinalis) was conducted. Fertilised eggs from three sources (wild strain, acidic watershed, pH 4.7 to 5.3, wild strain from  neutral watershed , pH 7.0, and a hatchery strain, pH 7.0) were exposed to a series of test media, with pH adjustment using sulphuric acid, to varying pHs (3.9, 4.3, 4.7, 5.2 and 7.0) under flow-through conditions. 

At pH 3.9, most embryos were dead within a few degree-days. Significant differences in mortality between the strains at low pH were observed and these suggested a genetic component to acid tolerance. Mortality in the strain from the acidic watershed was the lowest, followed by the second wild strain. Survival in both wild strains at low pH was much better than the hatchery embryos. These differences in survival at sublethal acidity (pH 4.7 to 7.0) were principally the result of high mortality shortly after fertilisation. After this period, mortality stabilised. Only at pH 4.3 did substantial mortality occur at hatching. The early embryonic strain therefore appears to be the most susceptible to sublethal acid stress in brook trout. Hatchery strain embryos were also introduced at the eyed stage at 213 degree days. Subsequent survival of this group was better at low pH than that of hatchery embryos introduced at fertilisation.  Higher acidity retarded hatch in all cases. 

The no-observed effect concentration (NOEC) for brook trout (based on survival) exposed to sulphuric acid was observed to be pH 5.2, equivalent to 0.31 mg/L.

Reference 5

Endpoint:

fish, juvenile growth test

Type of information:

read-across from supporting substance (structural analogue or surrogate)

Adequacy of study:

supporting study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

Published investigative non-guideline, non-GLP study.

Justification for type of information:

Sulphur trioxide readily reacts with water to form sulphuric acid. The reaction is instantaneous, to the extent that SO3 will react with water vapour in the atmosphere to form fumes of sulphuric acid. This reaction forms the basis of the manufacturing process of H2SO4. The read-across hypothesis is therefore that SO3 will instantaneously transform into H2SO4 upon contact with water (i.e. in aquatic ecotoxicology tests), thus any observed effects will be directly attributable to sulphuric acid. It is therefore justifiable to derive hazard conclusions from sulphuric acid data, with regard to ecotoxicological endpoints.

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

read-across source

Qualifier:

no guideline available

Principles of method if other than guideline:

See study design details

GLP compliance:

no

Remarks:

older, published study

Specific details on test material used for the study:

Available study data for sulphuric acid is being used for read-across to the target substance, sulphur trioxide.

Analytical monitoring:

yes

Details on sampling:

Not reported

Vehicle:

no

Details on test solutions:

Concentrated sulphuric acid was added to lower the pH in the acidic groups at intervals of 3 - 4 days first from neutral to pH 6.0 and then by 0.5 pH units until the respective nominal pH was reached. After this initial period, pH was maintained by the addition of known amounts of sulphuric acid after the daily water change. During the remainder of the day, the target pH was maintained by dripping diluted sulphuric acid at known rates into the holding tanks.

Test organisms (species):

Salvelinus fontinalis

Details on test organisms:

11 month old brook trout were purchased from a commercial hatchery (approx. 18 cm fork length, 100 g body weight). The fish were maintained in fibre glass tanks under constant flow. Prior to use the water was carbon-filtered, then aerated for 24 hours and mechanical agitation in a 1500L capacity plastic tank. The water in the fish holding tanks was constantly aerated and recirculated through nylon wool or a plastic sieve at a rate of 16 - 32 L/min. One third of the water was replaced daily and the tanks were cleaned monthly. Fish were fed once daily to satiation with a commercial trout food. Lighting was maintained to mimic the natural photoperiod.

Water temperature was controlled by heat exchange with chilled water (3-6 degrees C) circulated through plastic tubing. Water pH 7.01 - 7.61, alkalinity TIP 78 mg/L: aluminium 0.095 mg/L, copper 0.003 mg/L.

Fish were acclimated for two weeks to the laboratory water conditions before they were divided into one neutral and three acidic groups of equal size.

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

10 mo

Post exposure observation period:

Not applicable

Hardness:

Not reported

Test temperature:

In the control group temperature ranged between 7.9 °C in March to a maximum of 15.7°C in June to a low of 8.5°C in November. In the pH 4.5 grouptemperature ranged between 8.8 °C in March to a maximum of 15.5°C in June to a low of 8.2°C in November

pH:

pH was checked every 1-2 hours from 08:30 to 18:00 daily. See measured concentrations.

Dissolved oxygen:

Oxygen levels were checked periodically. Mean overall oxygen concentration was 6.9 ± 0.3 mg/mL (12 measurements) with no apparent difference between the pH groups.

Salinity:

Not applicable

Nominal and measured concentrations:

Nominal pH values were 4.5, 5.0 and 5.5 with a control of pH 7.3

Mean measured pH values were control of 7.34 ± 0.01 (297 measurements), 5.56 ± 0.02 (312 measurements), 5.16 ± 0.03 (290 measurements) and 4.48 ± 0.02 (292 measurements)

Details on test conditions:

Not applicable

Reference substance (positive control):

no

Duration:

10 mo

Dose descriptor:

NOEC

Effect conc.:

0.13 mg/L

Nominal / measured:

meas. (not specified)

Conc. based on:

test mat.

Basis for effect:

weight

Remarks:

juvenile fish weight

Remarks on result:

other: NOEC was 5.56 pH units (equivalent to 0.13 mg/L)

Details on results:

Food consumption rate in all groups was about the same at 128.9 ± 5.6 g trout chow per 100 fish per day in March. Thereafter the quantity of food was no longer accurately measured, no difference in food consumption was noticeable during the subsequent months. Throughout the experiments all fish mortalities appeared to be unrelated to pH exposure.

Somatic growth - within each pH group body weight significantly increased during the experimental period (ANOVA, p<0.001). Between group comparison revealed that towards the latter half of the experimental period the mean body weight of the brook trout in the pH 4.5 group was significantly lower (ANOVA, p<0.05 to <0.001) than that of the control. The mean body weight of pH groups 5.0 and 5.5 were mostly between the control and pH 4.5. At the end of the experiment, mean body weight of the males and females in the pH 4.5 group was 73.18 and 72,82% of the same sex in the control group. See table 1 for details.


Gonad growth, egg production and atresia - the gonado-somatic index (GSI) is the ratio of gonadal to body weight, expressed as milligrams per body weight. In males and females, there were no significant differences (ANOVA) between any group at any time of the experimental period. Gonadal weight began to increase in June and reached maximal values in September and November.
The mean ovulation times of the various groups are expressed relative to October 31 (day 0) to allow comparison. Ovulation in pH groups 4.5 and 5.0 occurred very significantly later than the control. No delay was observed in the males, as spermatozoa could be obtained from approximately 50% of the individuals in all groups by the end of October. Correlation analysis showed that there was no correlation between body weight and the mean weight of the ovulated eggs in any experimental group. Thus, larger fish did not have larger eggs. There was no significant difference in egg weight between the control and any of the pH groups. However, eggs from the pH 4.5 group were significantly heavier than those of pH group 5.0 (ANOVA, p <0.02). There was significant correlation between the total number of mature, yolky eggs and body weights in the control, pH 4.5 and pH 5.0 groups. The number of eggs produced per female increased in a linear fashion with increase in body weight. See Tables 2 and 3 for details.

The extent of atresia was significantly higher in the pH 5.0 and 5.5 than in the control and pH 4.5 groups in September (ANOVA, p < 0.01). In November, only pH group 5.0 had more atric eggs than the other groups (p = 0.004).

Results with reference substance (positive control):

Not applicable

Reported statistics and error estimates:

Analysis of variance (ANOVA) and Student-Newman-Keuls (SNK) a posteriori procedure were used to detect significant differences among body weight, body length, gonadosomatic index, average egg weight, and ovulation time. Where appropriate, Student's t-test was also used to test for significant difference between a pair of means. The Kruskal-Wallis ranked ANOVA was used to detect significant differences in the extent of atresia between fish. Confidence limit was taken at the 5% level.

Table 1 - Regression analysis of changes in body weight in brook trout maintained at pH 7.34 (control) and pH 4.5

Group	Slope			Intercept
	Female	Male	p	Female	Male	p
Control	33.0	38.7	NS	54	33	NS
pH 4.5	47.4	24.8	0.001	-182	67	0.05
p	NS	0.02		0.001	0.001

NS - not significantly different

Table 2 - Mean ovulation time expressed as number of days. 

.

pH	Proportion of fish ovulated	Ovulation time (days after 31 October)
7.34 (control)	3/4	14.3 ± 0.3 (3)
5.16	7/14	30.0 ± 1.1 (7)
4.48	11/16	19.6 ± 0.8 (8)

Results are based on 34 females, all of which possessed mature, yolky eggs

Table 3 - Average egg weight and total number of eggs calculated for each female of average body weight at the end of the ovulation period

pH	Average female weight (g)	Average egg weight (mg)^a	Number of eggs per female of average body weight^b
7.34 (control)	471	29.8±2.7 (7)	2020 (100)
5.16	333	27.6±2.1 (8)	1290 (63.86
4.48	343	33.3±1.1 (13)	1343 (66.49)

^a – group 1 was not significantly different from 4.48 or 5.16 (ANOVA)
^b – number of eggs expressed as % of those in the control in parentheses

Validity criteria fulfilled:

not specified

Remarks:

There are no appropriate study guidelines.

Conclusions:

Based on the reported information an acceptable assessment of pH effects on growth, ovulation and egg production was achieved.

Executive summary:

Data on the chronic toxicity to freshwater fish is available for sulphuric acid and is considered suitable for read-across (based on the analogue approach) to the target substance, sulphur trioxide. Sulphur trioxide readily reacts with water to form sulphuric acid. The reaction is instantaneous, to the extent that SO₃ will react with water vapour in the atmosphere to form fumes of sulphuric acid. This reaction forms the basis of the manufacturing process of H₂SO₄. The read-across hypothesis is therefore that SO₃ will instantaneously transform into H₂SO₄ upon contact with water (i.e. in aquatic ecotoxicology tests), thus any observed effects will be directly attributable to sulphuric acid. It is therefore justifiable to derive hazard conclusions from sulphuric acid data, with regard to ecotoxicological endpoints.

A chronic toxicity study with the brook trout (Salvelinus fontinalis) was conducted. Hatchery reared fish were maintained in the laboratory at mean pH values of 7.34, 5.56, 5.16 and 4.48 from early February to December 1984.  At pH 4.8, the mean growth rates of males were uniformly lowered during the entire experimental period.  Among females, growth was inhibited during the first 5 months, but their rate of weight gain recovered during the period of rapid oocyte development.  At the end of the experiment, the body weights of both male and female fish in pH 5.16 and 4.48 were only 71 – 77% of the control fish at ph 7.34.  Growth was not affected by exposure to pH 5.56. Rapid oocyte development occurred simultaneously over all pH groups in June, suggesting that the initiation of gametogenesis was not affected over the range of pH tested.  The number of eggs produced was significantly correlated to body weight; consequently the number of eggs produced by the smaller ph 5.16 – 4.48 females was reduced.  Ovulation was also significantly delayed in the acidic groups.

The overall NOEC for this study was determined on the weight of juvenile trout produced in 10 months and was pH 5.56 (equivalent to 0.13 mg/L).

Reference 6

Endpoint:

long-term toxicity to fish, other

Type of information:

read-across from supporting substance (structural analogue or surrogate)

Adequacy of study:

key study

Reliability:

2 (reliable with restrictions)

Rationale for reliability incl. deficiencies:

study well documented, meets generally accepted scientific principles, acceptable for assessment

Remarks:

Published, investigative non-guideline, non-GLP study

Justification for type of information:

Sulphur trioxide readily reacts with water to form sulphuric acid. The reaction is instantaneous, to the extent that SO3 will react with water vapour in the atmosphere to form fumes of sulphuric acid. This reaction forms the basis of the manufacturing process of H2SO4. The read-across hypothesis is therefore that SO3 will instantaneously transform into H2SO4 upon contact with water (i.e. in aquatic ecotoxicology tests), thus any observed effects will be directly attributable to sulphuric acid. It is therefore justifiable to derive hazard conclusions from sulphuric acid data, with regard to ecotoxicological endpoints.

Reason / purpose for cross-reference:

reference to same study

Reason / purpose for cross-reference:

read-across source

Qualifier:

no guideline available

Principles of method if other than guideline:

Groups of sexually mature flagfish (Jordanella floridae) were exposed to sub-lethal levels of acidity, pH 4.5, 5.0, 5.5 and 6.0. Spawning success, fry growth and survival were monitored.

GLP compliance:

no

Remarks:

older, published study

Specific details on test material used for the study:

Available study data for sulphuric acid is being used for read-across to the target substance, sulphur trioxide.

Analytical monitoring:

yes

Details on sampling:

Not reported

Vehicle:

no

Details on test solutions:

The pH of incoming lake water was adjusted by additoon of 0.1 N sulphuric acid injected by an automatic pipettor into the mixing cell. The diluter discharged 0.5 L litre of water per test vessle every two minutes

Test organisms (species):

Jordanella floridae

Details on test organisms:

Mature flagfish, less than one year old, were purchased from a commercial supplier and held in laboratory dechlorinated water for two months before being transferred to the Lake Panache testing location. On arrival at the lake site male and female fish were seperated and held in tanks for seven days in holding tanks.

Test type:

flow-through

Water media type:

freshwater

Limit test:

no

Total exposure duration:

65 d

Remarks on exposure duration:

Spawning and fry growth

Post exposure observation period:

Not applicable

Hardness:

Lake water - 28 mg CaC03/L.

Test temperature:

26.5 ± 0.5 °C

pH:

Nominal pH values were ciontrol (6.8), 4.5, 5.0, 5.5 and 6.0

Dissolved oxygen:

Dissolved oxygen was always maintained at over 90% saturation.

Salinity:

Not applicable

Nominal and measured concentrations:

Nominal pH values were control (6.8), 4.5, 5.0, 5.5 and 6.0

pH of each treatment group was reported to have been maintained within 0.13 pH units at nominal values over all exposure periods.

Details on test conditions:

The study was conducted with a 15 hour light photoperiod.

Reference substance (positive control):

no

Remarks:

not applicable

Duration:

65 d

Dose descriptor:

NOEC

Effect conc.:

0.025 mg/L

Nominal / measured:

nominal

Conc. based on:

other: pH

Basis for effect:

other: fry growth

Details on results:

Adult mortality principally occurred within the final 20days of exposure and was about 7% (one female) in the control group, increasing to 21 and 14%at pH levels 5.5 and 5.0, respectively and 64% at pH 4.5, see Table 1.

All treatment groups commenced spawning within three to four days of one another and egg production and fertility at the end of both the 10 day acclimation period and the five day pH depression period was equivalent among all groups and bore no correlation(p>0.05) to pH. Exposure to depressed pH reduced the total number of spawnings (see Table 1) in each group compared to the controls (P<0.05) except for pH 6.0 (p=0.15). No spawnings occurred at pH 4.5. Spawning frequency overall was observed to decrease directly as pH was reduced (P<0.05).

A significant direct correlation (P<0.05) between total eggs produced and pH was found (See Table 1). Mean daily egg output was significantly impaired (P<0.05) at each pH level compared to the controls. Combining the data for spawning frequency and total eggs produced shows that mean spawning size correlated directly with pH. Mean spawning size was significantly reduced compared to the control (P<0.05) at the pH 5.5 treatment.

The number of fertile eggs was reduced (P<0.05) at every exposure level compared to the controls and this related well (P<0.05) with a decline in pH (see Table 1). Hatchability of eggs to the different pH's was variable, ranging from 17-56% and not indicative of overall trends, yet hatching success was somewhat less during pH 5.0 exposure. Eggs collected from control tanks and incubated in pH 5.0 and 4.5 hatched with 82 and 87% success. The control fry hatched under these pH conditions died within 24-56 hours.

The final mean weights of the 45 day old fry were significantly less (P<0.05) after exposure to all pH treatments compared to the controls and the decrease was directly related (P<0.05) to pH depression (see Table 2). The correlation coefficient (r=0.99) of the final mean fry weight and the final mean fry dorsal image area at the different pH treatments was significantly (P<0.05). Therefore, a change in the dorsal image area was considered representative of a change in growth. The growth rates as indicated by changing dorsal image area of the fry held under control and pH 6 conditions were similar (P>0.05) while the growth rates of fry held under pH 5.5 and 5.0 conditions were considerably reduced (P<0.05) compared to the controls. A significant decrease in growth rate (P<0.05) between fry in pH 5.5 and 5.0 was also evident.

The breeding communities that were allowed to spawn for 16 days and were then exposed to pH 4.5 for seven days abruptly stopped spawning for that entire period. One spawning occurred seven days after normal lake water conditions resumed and a second was observed on day 9 of the final 14 day period. The mean daily output was reduced from 78 eggs after 22 spawnings before pH depression to 4.7 eggs from two spawnings after depression. The total number of eggs produced was 1 249 before depression compared to 66 after.

Results with reference substance (positive control):

Not applicable

Reported statistics and error estimates:

Fish measurements (length and dorsal image) were subjected to logarithmic transformation with respect to time and was used for growth analysis.

Spawning frequency, egg fertility and mortality data were subjected to Chi-square analysis, while egg production data was tested for significance by analysis of variance and independent t-distribution. Fry growth rate data were compared by analysis of covariance and the correlation of all reproductive responses with declining ph was tested by regression analysis.

Table 1 – Reproductive response of flagfish exposed to depressed pH for 20 days

Parameter	Nominal pH					Correlation with pH
	Control (6.8)	6.0	5.5	5.0	4.5
Adult mortality	1	0	3	2	11	-
Spawning	32	24	15*	7*	1*	P <0.05
Total eggs	2467	1277	557	319	1	P <0.05
Mean daily output	61.7	31.9*	13.9*	8.0*	0*	P <0.05
Mean spawning size	77.1	53.2	37.1*	45.6	1*	P <0.05
% fertile eggs	90.1	78.6*	68.0*	55.2*	-	P <0.05
% hatching of fertile eggs	44.4	56.5	51.1	17.0	-	P <0.05

* P <0.05 compared to the control

Table 2 – Survival, weights and dorsal image area of flagfish fry exposed to depressed pH for 45 days.

Parameter	Control (6.8)	6.0	5.5	5.0	Correlation with pH
Total surviving fry (N=50)	37	43	20	2	P <0.05
Final mean weight of fry	0.55	0.42*	0.26*	0.075*	P <0.05
Final mean dorsal image area (mm^2)	123.7	100.3*	71.5*	37.7*	P <0.05

* P<0.05 compared to the control

Validity criteria fulfilled:

not specified

Remarks:

There are no appropriate study guidelines

Conclusions:

Based on the reported information an acceptable assessment of pH effects on flagfish reproduction, growth and survival was achieved.

Executive summary:

Data on the chronic toxicity to freshwater fish is available for sulphuric acid and is considered suitable for read-across (based on the analogue approach) to the target substance, sulphur trioxide. Sulphur trioxide readily reacts with water to form sulphuric acid. The reaction is instantaneous, to the extent that SO₃ will react with water vapour in the atmosphere to form fumes of sulphuric acid. This reaction forms the basis of the manufacturing process of H₂SO₄. The read-across hypothesis is therefore that SO₃ will instantaneously transform into H₂SO₄ upon contact with water (i.e. in aquatic ecotoxicology tests), thus any observed effects will be directly attributable to sulphuric acid. It is therefore justifiable to derive hazard conclusions from sulphuric acid data, with regard to ecotoxicological endpoints.

Breeding communities of flagfish (Jordanella floridae) were exposed to northern Ontario lake water (hardness 28 mg/L) adjusted to pH levels of 6.0, 5.5, 5.0 and 4.5. Control water (pH 6.8) received no acid treatment. Egg production, egg fertility and fry growth was impaired (P0.05) at all exposure levels. Flagfish fry survival was reduced (P0.05) at pH 5.5 and 5.0, no fry survived at pH 4.5. Variability of hatching in all treatments precluded any identifiable hatching response to depressed pH. Reduction in the reproductive processes monitored indicated the following order of sensitivity: egg production > fry survival > fry growth > egg fertility.

The LOEC 20% (lowest observed effect concentration) was pH 6.0 equivalent to 0.049 mg/L. and the NOEC (LOEC/2) was determined to be 0.025 mg/L.

Description of key information

A number of non-standard chronic fish studies with sulphuric acid are available; studies were performed in various species and using various exposure times and endpoints.

Key value for chemical safety assessment

Fresh water fish

Fresh water fish

Dose descriptor:

NOEC

Effect concentration:

0.025 mg/L

Additional information

The first study is a chronic toxicity study with the brook trout (Salvelinus fontinalis). Fertilised eggs from three sources (wild strain, acidic watershed, pH 4.7 to 5.3, wild strain from  neutral watershed , pH 7.0, and a hatchery strain, pH 7.0) were exposed to a series of test media, with pH adjustment using sulphuric acid, to varying pHs (3.9, 4.3, 4.7, 5.2 and 7.0) under flow-through conditions. At pH 3.9, most embryos were dead within a few degree-days. Significant differences in mortality between the strains at low pH were observed and these suggested a genetic component to acid tolerance.  Mortality in the strain from the acidic watershed was the lowest, followed by the second wild strain.  Survival in both wild strains at low pH was much better than the hatchery embryos.  These differences in survival at sublethal acidity (pH 4.7 to 7.0) were principally the result of high mortality shortly after fertilisation.  After this period, mortality stabilised.  Only at pH 4.3 did substantial mortality occur at hatching.  The early embryonic strain therefore appears to be the most susceptible to sublethal acid stress in brook trout.  Hatchery strain embryos were also introduced at the eyed stage at 213 degree days.  Subsequent survival of this group was better at low pH than that of hatchery embryos introduced at fertilisation.  Higher acidity retarded hatch in all cases. The no-observed effect concentration (NOEC) for brook trout (based on survival) exposed to sulphuric acid was observed to be pH 5.2, equivalent to 0.31 mg/L.

The second study reported is also a chronic toxicity study with the brook trout (Salvelinus fontinalis). Hatchery reared fish were maintained in the laboratory at mean pH values of 7.34, 5.56, 5.16 and 4.48 from early February to December 1984.  At pH 4.8, the mean growth rates of males were uniformly lowered during the entire experimental period.  Among females, growth was inhibited during the first 5 months, but their rate of weight gain recovered during the period of rapid oocyte development.  At the end of the experiment, the body weights of both male and female fish in pH 5.16 and 4.48 were only 71 – 77% of the control fish at ph 7.34.  Growth was not affected by exposure to pH 5.56. Rapid oocyte development occurred simultaneously over all pH groups in June, suggesting that the initiation of gametogenesis was not affected over the range of pH tested.  The number of eggs produced was significantly correlated to body weight; consequently the number of eggs produced by the smaller ph 5.16 – 4.48 females was reduced.  Ovulation was also significantly delayed in the acidic groups. The overall NOEC for this study was determined on the weight of juvenile trout produced in 10 months and was pH 5.56 (equivalent to 0.13 mg/L).

The third study investigated breeding communities of flagfish (Jordanella floridae) were exposed to northern Ontario lake water (hardness 28 mg/L) adjusted to pH levels of 6.0, 5.5, 5.0 and 4.5. Control water (pH 6.8) received no acid treatment. Egg production, egg fertility and fry growth was impaired (P0.05) at all exposure levels. Flagfish fry survival was reduced (P0.05) at pH 5.5 and 5.0, no fry survived at pH 4.5. Variability of hatching in all treatments precluded any identifiable hatching response to depressed pH. Reduction in the reproductive processes monitored indicated the following order of sensitivity: egg production > fry survival > fry growth > egg fertility. The LOEC 20% (lowest observed effect concentration) was pH 6.0; equivalent to 0.049 mg/L. and the NOEC (LOEC/2) was determined to be 0.025 mg/L.