e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 7 ( 2 0 1 4 ) 24–27
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/etap
Short communication
Comparison of embryo toxicity using two classes of aquatic vertebrates Gunnar Carlsson ∗ , Leif Norrgren Department of Biomedical Sciences and Veterinary Public Health, Swedish University of Agricultural Sciences, Box 7028, 750 07 Uppsala, Sweden
a r t i c l e
i n f o
a b s t r a c t
Article history:
Toxicity tests of musk ketone (MK) and tetrabromobisphenol-A (TBBPA) on embryos were
Received 10 September 2013
conducted in two amphibian species, Xenopus (Silurana) tropicalis and the Swedish native
Received in revised form
species Rana arvalis. TBBPA was also tested on fish embryos of Danio rerio. All species were
24 October 2013
tested in similar experimental setup. Musk ketone caused decreased heart rates at con-
Accepted 27 October 2013
centrations from 10 and 100 g/L in R. arvalis and X. tropicalis, respectively. TBBPA caused
Available online 4 November 2013
effects at 1000 g/L in all three species. The responses were comparable between all three
Keywords:
risk assessment.
species which supports the relevance for using data from non-native species in national Amphibians
© 2013 Elsevier B.V. All rights reserved.
Fish Embryo Toxicity
1.
Introduction
The fish embryo toxicity (FET) test (OECD, 2013) which is based on individual exposure of eggs in well-plates has been frequently used with zebrafish (Danio rerio) in routine effluent monitoring and in testing of single chemicals (Braunbeck et al., 2005; Carlsson and Norrgren, 2004). Studies have shown promising results with comparable data when adapting two other fish species recommended by OECD, the Japanese medaka (Oryzias latipes) and the fathead minnow (Pimephales promelas) to the same protocol (Braunbeck et al., 2005). D. rerio as well as Xenopus frogs are aquatic animals from two different classes of aquatic vertebrates which allows for regular studies of embryo toxicity due to simple ways of obtaining eggs. Young stages of both species have been used
∗
Corresponding author. Tel.: +46 18671145. E-mail address: [email protected] (G. Carlsson). 1382-6689/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.etap.2013.10.015
widely in development and toxicity models and standardised guidelines are available (OECD, 2013; ASTM, 1999). However, performing risk assessment for regional conditions, based on the results from toxicity studies on non-native species, can be questioned. The aim of the present study was to compare the responses and the sensitivity between embryos from two different classes of aquatic vertebrates, fish and amphibians when exposed to two environmentally relevant substances, musk ketone (MK) and tetrabromobisphenol-A (TBBPA). This was accomplished by using the same embryo toxicity method to test both D. rerio and west-African clawed frog (Xenopus (Silurana) tropicalis) in identical experimental setup. Further, to explore the relevance of the result from the two tropical species to a native Swedish species, also embryos of moor frog (Rana arvalis) were exposed to these chemicals. Data for MK
e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 7 ( 2 0 1 4 ) 24–27
on D. rerio were obtained from an earlier study (Carlsson and Norrgren, 2004).
2.
Materials and methods
Tetrabromobisphenol-A (TBBPA) was kindly provided by Professor Åke Bergman, Stockholm University and Musk Ketone (MK) were obtained from Sigma Aldrich Sweden AB. Stock solutions were prepared by dissolving chemicals in dimethylsulfoxide (DMSO) and stored at 4 ◦ C until mixed with exposure water to a final DMSO concentration of 0.1%. In all tests, exposure concentrations of both chemicals were 1.0, 10, 100 and 1000 g/L including a control with 0.1% DMSO. Fish water was prepared according to ISO (1999) and continuously aerated at a temperature of 26 ◦ C. Reproduction of D. rerio was performed as previously described (Carlsson and Norrgren, 2004). Briefly, adult fish of both sexes were mixed in cone shaped reproduction funnels, provided with nets, separating adults from their eggs. Eggs at homogenous developmental stages were collected 30 min after onset of light and immediately exposed to a concentration series of TBBPA in 50 mL solution. The eggs were examined for selection of eggs that were between the 4 and 16-cell stages. Selected eggs were transferred to 96-well plates, one egg in each well, together with 250 L solution. Each exposure group included 24 eggs. The plates were kept at a temperature of 26 ◦ C and a photoperiod of 12:12 h for 48 h. The eggs were examined at different observation times as summarised in Table 1. The water used for X. tropicalis throughout the experiment, was made according to the FETAX protocol (ASTM, 1999), diluted 1:1 with deionised water and continuously aerated at 26 ◦ C. One female and one male adult frog were injected in the dorsal lymph-sac with 20 international units (IU) of human chorionic gonadotropin, followed by 100 IU injections 2 days later to induce spawning. After spawning, frogs were removed from the aquarium and the eggs were collected. The jelly-coat of the eggs was removed by swirling the eggs in l-cysteine solution (1.25 g l-cysteine in 50 mL frog water). Eggs were then exposed to concentration series of MK and TBBPA. The following procedure was the same as for D. rerio with transfer to 96-well plates. Studied endpoints at different observations times are shown in Table 1. Newly laid R. arvalis eggs from 12 different clutches were collected immediately after spawning from a natural pond where no other species were observed, and stored a week in a refrigerator (+4 ◦ C) until the onset of the experiment. Eggs were carefully separated from each other and individually placed in 50 mL beakers containing the series of TBBPA or MK. The water used for the R. arvalis embryos was made according to the FETAX protocol (ASTM, 1999) and continuously aerated at 17 ◦ C. The test was initiated with 25 individuals in each concentration. Different clutches contributed to the same number of eggs in each treatment. Remaining eggs, not used in the test were left in water in the laboratory for evaluation of fertilisation rate in different clutches. The beakers were placed in a water bath with a temperature of 17 ◦ C and the photoperiod was 12:12 h. Exposure media was renewed after 2 and 4 days of exposure and the study was terminated after 7 days when the tadpoles had reached the corresponding stage
25
in development as in the X. tropicalis experiment. Endpoints and times of observations are shown in Table 1. The heart was not visible due to pigmentation so heart rate was determined by counting pulsation of the external gills after hatching at 4 days of exposure. Eggs recorded as undeveloped after one day, were excluded from the study. Heart rates were analysed using one-way ANOVA followed by Dunnett’s post hoc test, comparing exposed groups with controls. Total number of affected embryos and number of unfertilised eggs were analysed comparing each treatment with respective control using Fisher’s exact-test with Bonferroni adjustment. The significance level was set at 0.95 (p < 0.05). The analyses were made in Minitab 16.
3.
Results and discussion
Nitro musks and brominated flame retardants (BFRs) are examples of chemicals that have been detected in environmental samples in relatively high concentrations (Gatermann et al., 1999; de Wit, 2002). Musk ketone (MK) and tetrabromobisphenol-A (TBBPA), belongs to the nitro musks and BFRs, respectively. MK has previously been studied using embryos from D. rerio with the same method as in the present study (Carlsson and Norrgren, 2004). The two substances studied on the amphibian species, their effects, and the concentrations where these effects where observed, correlated well with the results obtained from D. rerio in the present study and in Carlsson and Norrgren (2004). All three species had a lowest observed effect concentration (LOEC) of 1000 g/L of TBBPA, where oedema, absent circulation and mortality were the main findings at the last observation times. 100% of D. rerio and R. arvalis and 85% of X. tropicalis were affected when exposed to 1000 g/L of TBBPA. In D. rerio, an increased number of embryos (68%) exposed to 1000 g/L TBBPA showed lack of spontaneous movement already after 24 h and a decline in heart rate (61% of the control heart rate, p < 0.01) after 48 h exposure. These findings were not observed in the two frog species. Concentrations of TBBPA are often below limit of detection in natural water (Kuiper et al., 2007). Thus, there seems to be low risk for an acute toxic impact caused by water exposure to TBBPA. Both X. tropicalis in the present study, and D. rerio (Carlsson and Norrgren, 2004) exposed to MK showed oedema and absent blood circulation in the highest tested concentration, 1000 g/L. Further, heart rates were decreased in all three species. For X. tropicalis, heart rate decreased at 100 g/L and higher (Fig. 1). For R. arvalis exposed to MK, a clear concentration-response relationship on decline in heart rate was recorded for individuals exposed to 10 g/L and higher (Fig. 1). MK has been detected in water up to g/L levels but is generally in lower concentrations (European Union, 2005). The fact that R. arvalis show the same low LOEC as D. rerio (Carlsson and Norrgren, 2004) when exposed to MK, raises the concern for the effects of synthetic musk contamination previously discussed by Carlsson and Norrgren (2004). The present study shows that D. rerio and X. tropicalis might be used in the same experimental setup. The opportunity to use aquatic vertebrate species from different classes with the same basic methodology might provide more information of
26
e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 7 ( 2 0 1 4 ) 24–27
Table 1 – Time after onset of exposure for observations of studied endpoints and criteria for endpoints in embryo tests using Danio rerio, Xenopus tropicalis and Rana arvalis. Endpoints
Criteria
Danio rerio
Xenopus tropicalis
Rana arvalis
Mortality Spontaneous movement Eye development Tail extension Blood circulation Pigmentation Heart rate
Coagulated egg/embryo or lack of heart beat, yes/no Movement within 30 s, yes/no Normal eye development, yes/no Tail normally extended, yes/no Flow of blood cells are visible in caudal artery in tail, yes/no Dark eyes, dark pigmentation spots on body, yes/no Time for 30 heart beats are registered and converted to beats per minute
24 h, 48 h 24 h 24 h 24 h 48 h 48 h 48 h
24 h, 48 h 24 h 24 h 24 h 48 h 48 h 48 h
48 h, 96 h, 7d – – – – – 96 h
the general impact of a chemical than studies using only one animal species or one animal class. With the approach of two species in the same experimental setup, we can directly follow differences in e.g. heart rate and blood circulation without having to take experimental differences into consideration. In ecotoxicological risk assessment, results from tests on a very limited number of species have to serve as representatives for all species. Further, the tested species are often non-native and thus might be of low relevance for local situations. Therefore, comparisons of responses between common laboratory species and native species are important. In the present study, R. arvalis was used as an example of how a native species respond in comparison with laboratory species. This species has previously been used in embryo studies using lower temperatures and showing slower developmental rate compared with X. tropicalis (Sagvik et al., 2008). Observed endpoints for R. arvalis were basically the same as for the tropical species, although some endpoints were not possible to observe due to heavy pigmentation in R. arvalis embryos. Further, temperature was lower, exposure volume was higher and exposure time was increased. In the present study, the tropical species corresponded well to the native R. arvalis which suggests that it is proper to use these as reliable model species. However, it must be emphasised that these comparisons is very limited based on only two chemicals and three species.
There was a large variation in fertilisation rate among egg clutches collected for the R. arvalis study ranging from 3% to 93% when examining the eggs not included in the study. In the actual test, between 32% and 40% of the eggs in the different treatment groups were found to be undeveloped after one day of exposure. These eggs came from the same clutches in all treatment groups and there were no differences in number of undeveloped eggs among treatments. Eggs from these clutches were also to a high degree undeveloped among the eggs not included in the study. Thus, these undeveloped eggs were considered to be unfertilised and excluded from the study. The mortality further in the test was assessed on the individuals that were still alive at day one (n = 15–17). No chemical analyses were performed of the exposure solutions which is a lack of the study. However, measured concentrations have previously been shown to be close to nominal in test systems similar as the present ones for both MK (Breitholtz et al., 2003) and TBBPA (Norman Haldén, 2010) indicating stability of both compounds. In conclusion, the present study suggests that the method used, developed for D. rerio, can be applied also for X. tropicalis and after modifications such as lover temperature and longer duration to R. arvalis. Also, the responses are comparable between species when embryos are exposed to MK or TBBPA. Combining multiple species in toxicity tests with the
D. rerio
**
1.2
**
**
1.0
R. arvalis
Relative heart rate
X. tropicalis 0.8
** **
** **
0.6
**
0.4 0.2 0.0
Control
1.0
10
100
1000
Concentration (µg/l) Fig. 1 – Mean heart rate relative to the mean value of respective control group for Danio rerio and Xenopus tropicalis after 48 h of exposure to musk ketone and for Rana arvalis after four days of exposure. Bars and error bars represents means and standard deviations, respectively. Data on Danio rerio are from Carlsson and Norrgren (2004). **represents p-values of
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.elsevier.com/locate/etap
Short communication
Comparison of embryo toxicity using two classes of aquatic vertebrates Gunnar Carlsson ∗ , Leif Norrgren Department of Biomedical Sciences and Veterinary Public Health, Swedish University of Agricultural Sciences, Box 7028, 750 07 Uppsala, Sweden
a r t i c l e
i n f o
a b s t r a c t
Article history:
Toxicity tests of musk ketone (MK) and tetrabromobisphenol-A (TBBPA) on embryos were
Received 10 September 2013
conducted in two amphibian species, Xenopus (Silurana) tropicalis and the Swedish native
Received in revised form
species Rana arvalis. TBBPA was also tested on fish embryos of Danio rerio. All species were
24 October 2013
tested in similar experimental setup. Musk ketone caused decreased heart rates at con-
Accepted 27 October 2013
centrations from 10 and 100 g/L in R. arvalis and X. tropicalis, respectively. TBBPA caused
Available online 4 November 2013
effects at 1000 g/L in all three species. The responses were comparable between all three
Keywords:
risk assessment.
species which supports the relevance for using data from non-native species in national Amphibians
© 2013 Elsevier B.V. All rights reserved.
Fish Embryo Toxicity
1.
Introduction
The fish embryo toxicity (FET) test (OECD, 2013) which is based on individual exposure of eggs in well-plates has been frequently used with zebrafish (Danio rerio) in routine effluent monitoring and in testing of single chemicals (Braunbeck et al., 2005; Carlsson and Norrgren, 2004). Studies have shown promising results with comparable data when adapting two other fish species recommended by OECD, the Japanese medaka (Oryzias latipes) and the fathead minnow (Pimephales promelas) to the same protocol (Braunbeck et al., 2005). D. rerio as well as Xenopus frogs are aquatic animals from two different classes of aquatic vertebrates which allows for regular studies of embryo toxicity due to simple ways of obtaining eggs. Young stages of both species have been used
∗
Corresponding author. Tel.: +46 18671145. E-mail address: [email protected] (G. Carlsson). 1382-6689/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.etap.2013.10.015
widely in development and toxicity models and standardised guidelines are available (OECD, 2013; ASTM, 1999). However, performing risk assessment for regional conditions, based on the results from toxicity studies on non-native species, can be questioned. The aim of the present study was to compare the responses and the sensitivity between embryos from two different classes of aquatic vertebrates, fish and amphibians when exposed to two environmentally relevant substances, musk ketone (MK) and tetrabromobisphenol-A (TBBPA). This was accomplished by using the same embryo toxicity method to test both D. rerio and west-African clawed frog (Xenopus (Silurana) tropicalis) in identical experimental setup. Further, to explore the relevance of the result from the two tropical species to a native Swedish species, also embryos of moor frog (Rana arvalis) were exposed to these chemicals. Data for MK
e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 7 ( 2 0 1 4 ) 24–27
on D. rerio were obtained from an earlier study (Carlsson and Norrgren, 2004).
2.
Materials and methods
Tetrabromobisphenol-A (TBBPA) was kindly provided by Professor Åke Bergman, Stockholm University and Musk Ketone (MK) were obtained from Sigma Aldrich Sweden AB. Stock solutions were prepared by dissolving chemicals in dimethylsulfoxide (DMSO) and stored at 4 ◦ C until mixed with exposure water to a final DMSO concentration of 0.1%. In all tests, exposure concentrations of both chemicals were 1.0, 10, 100 and 1000 g/L including a control with 0.1% DMSO. Fish water was prepared according to ISO (1999) and continuously aerated at a temperature of 26 ◦ C. Reproduction of D. rerio was performed as previously described (Carlsson and Norrgren, 2004). Briefly, adult fish of both sexes were mixed in cone shaped reproduction funnels, provided with nets, separating adults from their eggs. Eggs at homogenous developmental stages were collected 30 min after onset of light and immediately exposed to a concentration series of TBBPA in 50 mL solution. The eggs were examined for selection of eggs that were between the 4 and 16-cell stages. Selected eggs were transferred to 96-well plates, one egg in each well, together with 250 L solution. Each exposure group included 24 eggs. The plates were kept at a temperature of 26 ◦ C and a photoperiod of 12:12 h for 48 h. The eggs were examined at different observation times as summarised in Table 1. The water used for X. tropicalis throughout the experiment, was made according to the FETAX protocol (ASTM, 1999), diluted 1:1 with deionised water and continuously aerated at 26 ◦ C. One female and one male adult frog were injected in the dorsal lymph-sac with 20 international units (IU) of human chorionic gonadotropin, followed by 100 IU injections 2 days later to induce spawning. After spawning, frogs were removed from the aquarium and the eggs were collected. The jelly-coat of the eggs was removed by swirling the eggs in l-cysteine solution (1.25 g l-cysteine in 50 mL frog water). Eggs were then exposed to concentration series of MK and TBBPA. The following procedure was the same as for D. rerio with transfer to 96-well plates. Studied endpoints at different observations times are shown in Table 1. Newly laid R. arvalis eggs from 12 different clutches were collected immediately after spawning from a natural pond where no other species were observed, and stored a week in a refrigerator (+4 ◦ C) until the onset of the experiment. Eggs were carefully separated from each other and individually placed in 50 mL beakers containing the series of TBBPA or MK. The water used for the R. arvalis embryos was made according to the FETAX protocol (ASTM, 1999) and continuously aerated at 17 ◦ C. The test was initiated with 25 individuals in each concentration. Different clutches contributed to the same number of eggs in each treatment. Remaining eggs, not used in the test were left in water in the laboratory for evaluation of fertilisation rate in different clutches. The beakers were placed in a water bath with a temperature of 17 ◦ C and the photoperiod was 12:12 h. Exposure media was renewed after 2 and 4 days of exposure and the study was terminated after 7 days when the tadpoles had reached the corresponding stage
25
in development as in the X. tropicalis experiment. Endpoints and times of observations are shown in Table 1. The heart was not visible due to pigmentation so heart rate was determined by counting pulsation of the external gills after hatching at 4 days of exposure. Eggs recorded as undeveloped after one day, were excluded from the study. Heart rates were analysed using one-way ANOVA followed by Dunnett’s post hoc test, comparing exposed groups with controls. Total number of affected embryos and number of unfertilised eggs were analysed comparing each treatment with respective control using Fisher’s exact-test with Bonferroni adjustment. The significance level was set at 0.95 (p < 0.05). The analyses were made in Minitab 16.
3.
Results and discussion
Nitro musks and brominated flame retardants (BFRs) are examples of chemicals that have been detected in environmental samples in relatively high concentrations (Gatermann et al., 1999; de Wit, 2002). Musk ketone (MK) and tetrabromobisphenol-A (TBBPA), belongs to the nitro musks and BFRs, respectively. MK has previously been studied using embryos from D. rerio with the same method as in the present study (Carlsson and Norrgren, 2004). The two substances studied on the amphibian species, their effects, and the concentrations where these effects where observed, correlated well with the results obtained from D. rerio in the present study and in Carlsson and Norrgren (2004). All three species had a lowest observed effect concentration (LOEC) of 1000 g/L of TBBPA, where oedema, absent circulation and mortality were the main findings at the last observation times. 100% of D. rerio and R. arvalis and 85% of X. tropicalis were affected when exposed to 1000 g/L of TBBPA. In D. rerio, an increased number of embryos (68%) exposed to 1000 g/L TBBPA showed lack of spontaneous movement already after 24 h and a decline in heart rate (61% of the control heart rate, p < 0.01) after 48 h exposure. These findings were not observed in the two frog species. Concentrations of TBBPA are often below limit of detection in natural water (Kuiper et al., 2007). Thus, there seems to be low risk for an acute toxic impact caused by water exposure to TBBPA. Both X. tropicalis in the present study, and D. rerio (Carlsson and Norrgren, 2004) exposed to MK showed oedema and absent blood circulation in the highest tested concentration, 1000 g/L. Further, heart rates were decreased in all three species. For X. tropicalis, heart rate decreased at 100 g/L and higher (Fig. 1). For R. arvalis exposed to MK, a clear concentration-response relationship on decline in heart rate was recorded for individuals exposed to 10 g/L and higher (Fig. 1). MK has been detected in water up to g/L levels but is generally in lower concentrations (European Union, 2005). The fact that R. arvalis show the same low LOEC as D. rerio (Carlsson and Norrgren, 2004) when exposed to MK, raises the concern for the effects of synthetic musk contamination previously discussed by Carlsson and Norrgren (2004). The present study shows that D. rerio and X. tropicalis might be used in the same experimental setup. The opportunity to use aquatic vertebrate species from different classes with the same basic methodology might provide more information of
26
e n v i r o n m e n t a l t o x i c o l o g y a n d p h a r m a c o l o g y 3 7 ( 2 0 1 4 ) 24–27
Table 1 – Time after onset of exposure for observations of studied endpoints and criteria for endpoints in embryo tests using Danio rerio, Xenopus tropicalis and Rana arvalis. Endpoints
Criteria
Danio rerio
Xenopus tropicalis
Rana arvalis
Mortality Spontaneous movement Eye development Tail extension Blood circulation Pigmentation Heart rate
Coagulated egg/embryo or lack of heart beat, yes/no Movement within 30 s, yes/no Normal eye development, yes/no Tail normally extended, yes/no Flow of blood cells are visible in caudal artery in tail, yes/no Dark eyes, dark pigmentation spots on body, yes/no Time for 30 heart beats are registered and converted to beats per minute
24 h, 48 h 24 h 24 h 24 h 48 h 48 h 48 h
24 h, 48 h 24 h 24 h 24 h 48 h 48 h 48 h
48 h, 96 h, 7d – – – – – 96 h
the general impact of a chemical than studies using only one animal species or one animal class. With the approach of two species in the same experimental setup, we can directly follow differences in e.g. heart rate and blood circulation without having to take experimental differences into consideration. In ecotoxicological risk assessment, results from tests on a very limited number of species have to serve as representatives for all species. Further, the tested species are often non-native and thus might be of low relevance for local situations. Therefore, comparisons of responses between common laboratory species and native species are important. In the present study, R. arvalis was used as an example of how a native species respond in comparison with laboratory species. This species has previously been used in embryo studies using lower temperatures and showing slower developmental rate compared with X. tropicalis (Sagvik et al., 2008). Observed endpoints for R. arvalis were basically the same as for the tropical species, although some endpoints were not possible to observe due to heavy pigmentation in R. arvalis embryos. Further, temperature was lower, exposure volume was higher and exposure time was increased. In the present study, the tropical species corresponded well to the native R. arvalis which suggests that it is proper to use these as reliable model species. However, it must be emphasised that these comparisons is very limited based on only two chemicals and three species.
There was a large variation in fertilisation rate among egg clutches collected for the R. arvalis study ranging from 3% to 93% when examining the eggs not included in the study. In the actual test, between 32% and 40% of the eggs in the different treatment groups were found to be undeveloped after one day of exposure. These eggs came from the same clutches in all treatment groups and there were no differences in number of undeveloped eggs among treatments. Eggs from these clutches were also to a high degree undeveloped among the eggs not included in the study. Thus, these undeveloped eggs were considered to be unfertilised and excluded from the study. The mortality further in the test was assessed on the individuals that were still alive at day one (n = 15–17). No chemical analyses were performed of the exposure solutions which is a lack of the study. However, measured concentrations have previously been shown to be close to nominal in test systems similar as the present ones for both MK (Breitholtz et al., 2003) and TBBPA (Norman Haldén, 2010) indicating stability of both compounds. In conclusion, the present study suggests that the method used, developed for D. rerio, can be applied also for X. tropicalis and after modifications such as lover temperature and longer duration to R. arvalis. Also, the responses are comparable between species when embryos are exposed to MK or TBBPA. Combining multiple species in toxicity tests with the
D. rerio
**
1.2
**
**
1.0
R. arvalis
Relative heart rate
X. tropicalis 0.8
** **
** **
0.6
**
0.4 0.2 0.0
Control
1.0
10
100
1000
Concentration (µg/l) Fig. 1 – Mean heart rate relative to the mean value of respective control group for Danio rerio and Xenopus tropicalis after 48 h of exposure to musk ketone and for Rana arvalis after four days of exposure. Bars and error bars represents means and standard deviations, respectively. Data on Danio rerio are from Carlsson and Norrgren (2004). **represents p-values of
