Environmental Toxicology and Chemistry, Vol. 34, No. 8, pp. 1793–1798, 2015 # 2015 SETAC Printed in the USA
Short Communication DO WATER QUALITY CRITERIA BASED ON NONNATIVE SPECIES PROVIDE APPROPRIATE PROTECTION FOR NATIVE SPECIES? XIAOWEI JIN,*y ZIJIAN WANG,z YEYAO WANG,y YIBING LV,y KAIFENG RAO,z WEI JIN,x JOHN P. GIESY,k#yyzz§§ and KENNETH M. Y. LEUNGkk
yDepartment of Analytical Technique, China National Environmental Monitoring Center, Beijing, China zState Key Laboratory of Environmental Aquatic Chemistry, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China xShijiazhuang Environmental Monitoring Center, Shijiazhuang, China kDepartment of Veterinary Biomedical Sciences and Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan, Canada #State Key Laboratory of Pollution Control and Resources Reuse, School of the Environment, Nanjing University, Nanjing, China yyDepartment of Zoology and Center for Integrative Toxicology, Michigan State University, East Lansing, Michigan, USA zzSchool of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong, China xxDepartment of Biology and Chemistry and State Key Laboratory in Marine Pollution, City University of Hong Kong, Kowloon, Hong Kong, China kkThe Swire Institute of Marine Science and School of Biological Sciences, The University of Hong Kong, Pokfulam, Hong Kong, China (Submitted 20 January 2015; Returned for Revision 7 March 2015; Accepted 10 March 2015) Abstract: The potential use of toxicity data for nonnative species to derive water quality criteria is controversial because it is sometimes questioned whether criteria based on species from one geographical region provide appropriate protection for species in a different region. However, this is an important concept for the development of Chinese water quality criteria or standards. Data were assembled on 38 chemicals for which values were available for both native and nonnative species. Sensitivities of these organisms were compared based on the 5% hazardous concentration values and the species sensitivity distribution from a literature review. Results of the present study’s analysis showed that there is approximately 74% certainty that use of nonnative species to generate water quality criteria would be sufficiently protective of aquatic ecosystems in China. Without applying any assessment factor to the water quality criteria generated from nonnative species, the uncertainty would be 26% when the native Chinese species might be under protection. Applying an assessment factor of 10 would offer adequate protection to native Chinese species for approximately 90% of tested chemicals and thus reduce the uncertainty from 26% to 10%. Environ Toxicol Chem 2015;34:1793–1798. © 2015 SETAC Keywords: Water quality criteria
Chinese native species
Species sensitivity distribution
approach. The species sensitivity distribution approach uses a statistical distribution estimated from a sample of toxicity data and visualized as a cumulative distribution function. Species sensitivity distributions are used to calculate the concentration at which a specified proportion of species will be affected, such as the hazardous concentration for 5% of species (HC5). Species sensitivity distributions are dependent on available data sets and can differ in type of distribution, taxonomic diversity, and sample size. A substantial amount of toxicity data from several taxonomic groups is required to obtain a robust HC5 [3]. In the species sensitivity distribution, the probability of a threshold for effect being exceeded during either acute or chronic exposures is developed. For instance, in the species sensitivity distribution approach, the probability of a particular proportion of species being affected can be estimated. This approach requires data on the concentration–response relationships for a relatively large number of species. As an example, if data for only 2 species were available, each species would represent 50% of the frequency distribution, and resolution or prediction would be poor. As a rule of thumb, it is suggested that to use the species sensitivity distribution approach, data for a minimum of 20 species should be available [3,4]. In this case, each species would represent 5% of the total, and the resolution of the predictive power of the analysis would be 5%. The species sensitivity distribution approach assumes that the species tested represent a random selection of all the possible sensitivities. If, for instance, the data available were for similar species with similar sensitivities that did not represent the entire range of
METHODOLOGIES FOR DERIVING WATER QUALITY CRITERIA
Water quality criteria are defined as levels of individual characteristics or descriptions of conditions of a water body that, if met, will generally protect the designated use(s) and water quality standards. The methods to conduct these assessments vary among jurisdictions but, in general, have the same elements and suffer from the same limitations and uncertainties. Probabilistic approaches have been used to describe among-species variation in sensitivities [1]. Two basic methods for derivation of water quality criteria are in use or proposed for use throughout the world. These include the assessment factor method and statistical extrapolation methods, such as the species sensitivity distribution approach. Single-point assessment factors are recognized as a conservative approach for dealing with uncertainty in assessing risks posed by chemicals. Yet, current applications of safety factors are based on policy rather than on empirical scientific evidence, and they result in values that are protective but not predictive [2]. The present study focuses on the derivation of aquatic life criteria using the species sensitivity distribution method because it can deliver greater statistical confidence to the risk-assessment process compared with the single-point * Address correspondence to [email protected] or [email protected]. Published online 12 March 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2985 1793
1794
Environ Toxicol Chem 34, 2015
X. Jin et al.
Figure 1. The distribution of the log-transformed ratio of hazardous concentration for 5% of species (HC5 ratio) between native and nonnative species in China (open circles), The Netherlands (closed squares), Australia (triangles), and the Arctic (diamonds; i.e., Palearctic vs Nearctic species). Shaded area represents a zone with the HC5 ratio varying from –2 to 2. Chemicals having 2 different HC5 ratios are in bold letters. 2,4-DCP ¼ 2,4-dichlorophenol; 2,4,6-TCP ¼ 2,4,6trichlorophenol.
possible sensitivities, then predicted probabilities would not be accurate [5,6]. In natural ecosystems there are ranges of species, with different natural histories and in different feeding guilds, that might be differentially exposed or have different sensitivities to stressors. It is impossible to test all of the species that make up the biological community of an ecosystem. For this reason, surrogate species are tested and then used as representatives of sensitive species in the environment. In the semiprobabilistic approach used by the US Environmental Protection Agency (USEPA) to develop water quality criteria, sensitive species in different families are selected to represent aquatic plants, invertebrates, and vertebrates [7]. The USEPA guidelines state that only aquatic species resident in North America can be used as test species to derive the water quality criteria, such as the criteria continuous concentration, for the protection of freshwater and marine ecosystems [8–11]. In Australia, species sensitivity distributions and protective concentrations for 95% of local species are being used to derive water quality guidelines for toxicants [12,13]. It is commonly thought that different ecosystems contain different biological constituents and that a concentration threshold that would be harmless in an ecosystem might lead to irreversible toxic effects in others. The potential use of toxicity data for nonnative species to derive water quality criteria is controversial because it is sometimes questioned whether criteria based on species from a geographical region provide appropriate protection for species in a different region [14]. This uncertainty could not be resolved previously, in large part because of the paucity of toxicity data applicable for local species. In particular, toxicity data, especially for chronic effects on resident species in China, are sparse; and selection of resident species for use in the development of aquatic life criteria values is confounded by variations among ecosystems throughout China. However, this is an important concept for the development of Chinese water quality criteria or standards. There might be differences between species in temperate regions and those in tropical regions [15] or the Arctic and, of course, between freshwater and marine environments [16]. In
fact, tropical species were found to be slightly more sensitive to certain chemical substances than species from temperate regions, so an application factor of 10 was suggested to ensure protection of communities in tropical regions [15]. As a result of conservation of biochemical and physiological mechanisms for maintaining homeostasis in freshwater organisms, it is likely that sensitivities of species within a particular class of organisms would not vary significantly among countries within a climatic region. Therefore, it is unlikely that there would be significant differences between the overall sensitivities of organisms between countries within the temperate zone. WATER QUALITY CRITERIA TO PROTECT NATIVE SPECIES IN CHINA
As China has developed economically and become more industrialized, more attention has been paid to the development of a system to protect the environment [17]. Previously, because of a lack of resources, China often adopted water quality criteria promulgated by other jurisdictions [18,19]. Historically, this lack of data on toxicities of chemicals on Chinese species pragmatically led to the use of toxicological data from foreign toxicity databases of nonnative species (e.g., data extracted from the USEPA’s ECOTOX database) to develop chemical water quality criteria [18]. However, as environmental regulation and monitoring have become more sophisticated, the adequacy of this approach has been questioned. The question has been posed whether it would be more appropriate to develop water quality criteria specific to the situation in China by acquiring data on toxicity of chemicals to species native to China. However, this uncertainty could not be resolved previously, in large part because of the paucity of toxicity data applicable for local species. For the more than 130 chemicals evaluated in the current Chinese Environmental Quality Standards for Surface Water, and even more chemicals needed to reestablish and revise their water quality criteria, data are available for only few species indigenous to China. Currently, the required resources, knowledge, and political
Australian vs non-Australian
Palearctic vs nearctic
Chinese vs non-Chinese
Taxonomic group
a
Ammonia
Phenola
4-Chlorophenol
Endosulfan
Lindane
Diazion
Fenitrothion
Chlorpyntos
Triclosan
Parathion
Nitrobenzene
Cadmium
Mercury
Arsenic (III)
a
Pentachlorophenol
2,4,6-Trichlorophenol
2,4-Dichlorophenol
Chemical
Native Nonnative Native Nonnative Native Nonnative Native Nonnative
Palearctic Nearctic Palearctic Nearctic Palearctic Nearctic Palearctic Nearctic
Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative
Species
17 71 8 6 9 28 13 17
10 36 7 14 6 11 13 6
14 15 13 21 12 60 14 16 47 33 49 50 20 13 32 38 12 12
Data number
Acute Acute Acute Acute Acute Acute Acute Acute
Acute Acute Acute Acute Acute Acute Acute Acute
Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute
Data type
All All All All All All
All All excl. excl. excl. excl. excl. excl.
All All All All All All All All
All All All All All All All All All All All All All All All All All All
fish fish fish fish fish fish
Taxonomic group
Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr
III III III III III III III III
Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal
Log-Burr III Log-Burr III Log-Burr III Log-Burr III Log-Burr III Log-Burr III Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic
Model fit
Table 1. Toxicity parameter of 38 chemicals for native and nonnative species
0.31 0.17 735 3018 8408 2177 96 93
0.09 0.15 0.9 1.6 0.36 0.37 2.07 1.47
1100 1060 790 720 35 57 264 335 3.52 8.69 4.39 2.0 8.86 13.59 0.22 0.02 33.48 48.46
HC5 (mg/L)
–0.51 –0.77 2.86 3.47 3.92 3.34 1.98 1.97
–1.04 –0.82 –0.05 0.20 –0.44 –0.43 0.31 0.17
3.04 3.02 2.89 2.85 1.54 1.75 2.42 2.52 0.54 0.94 0.64 0.30 0.95 1.13 –0.66 –1.69 1.52 1.68
Log HC5
0.97
0.26
4.11
0.55
0.71
1.03
1.78
1.67
1.45
0.09
1.53
0.45
2.47
1.27
1.63
0.91
0.96
Nonnative/native ratio
(continued)
[24]
[24]
[24]
[13]
[19]
[19]
[19]
[19]
[23]
[25]
[25]
[25]
[25]
[25]
[22]
[21]
[14]
Reference
Water quality criteria to protect native species in China Environ Toxicol Chem 34, 2015 1795
Fenitrothion
Diazinon
Atrazine
Heptachlor
Methoxychlor
Azinphos-methyl
Endrin
Carbaryl
Chlorpyrifos
Deltamethrin
Endosulfan
Pentachlorophenol
Cadmium
Phenol
Copper
Trichlorfon
Lindane
Malathion
DDT
Ammonium
Zinc
Chemical Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative
Species 13 7 10 9 8 7 8 8 8 7 8 7 7 7 8 6 6 7 8 5 6 7 8 4 8 4 5 6 5 6 5 6 5 5 4 6 4 6 5 5 4 4
Data number Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute
Data type Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish
Taxonomic group
Model fit Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal
707 288 218 100 1.99 1 10 141 10 10 39.81 79.43 30.20 39.81 5248 2041 10 257 19.95 10 1 0.1 1 3.98 1 3.98 1202 2951 0.60 0.20 0.1 1.99 10 10 3.98 4.07 13 182 5754 39.81 1148 257 1548
HC5 (mg/L) 2.85 2.46 2.34 2 0.3 0 1 2.15 1 1 1.6 1.9 1.48 1.6 3.72 3.31 1 2.41 1.3 1 0 –1 0 0.6 0 0.6 3.08 3.47 –0.22 –0.7 –1 0.3 1 1 0.6 0.61 4.12 3.76 1.6 3.06 2.41 3.19
Log HC5
6.03
28.84
0.44
1.02
1
19.95
0.33
2.45
3.98
3.98
0.1
0.50
25.7
0.39
1.32
1.99
1
14.1
0.50
0.46
0.41
Nonnative/native ratio
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
Reference
Environ Toxicol Chem 34, 2015
a Because there are no fish data available for Australia, the taxonomic groups for Non-Australian exclude (excl.) fish data. HC5 ¼ hazardous concentration for 5% of species; DDT ¼ Dichlorodiphenyltrichloroethane.
Netherlands vs non-Netherlands
Taxonomic group
Table 1. (Continued)
1796 X. Jin et al.
Water quality criteria to protect native species in China
will are available for China to develop its own water quality criteria to protect a range of aquatic environments. For social and political reasons, it might be preferable to use species native to China that have been generated under conditions that simulate the Chinese environment. However, it would take a long time and a great expenditure of resources to collect sufficient information, especially for probabilistic approaches, to achieve this goal. Furthermore, standard testing protocols have yet to be established for a range of surrogate species based on species endemic to China. To develop these test protocols would require a large investment simply to acquire information on nutritional requirements and culturing conditions for species to be used in standardized tests. Although this is feasible, it might not be the most timely and efficient method or allocation of resources and might unduly delay implementation of water quality criteria. COMPARISONS OF SENSITIVITIES BETWEEN NATIVE AND NONNATIVE SPECIES
Comparisons of sensitivities between species native to China and nonnative species can be made for only a few chemicals, including 2,4-dichlorophenol, pentachlorophenol, and triclosan [14,20–22]. When this comparison was made, there was no significant difference between sensitivities of Chinese and nonChinese taxa. In Australia, sensitivities of Australian and nonAustralian organisms to endosulfan, ammonia, 4-chlorophenol, and phenol, based on calculated HC5 values, were similar [13,23]. Similarly, among North American and European taxa with different geographic distributions, sensitivities to a range of toxicants have been shown to be similar. Moreover, natural history, habitat type, and geographical distribution of the species used to construct the species sensitivity distribution did not have a significant influence on the assessment of hazard; but the taxonomic composition of the species sensitivity distribution does have a significant effect on resulting estimates of the HC5 [24]. To further verify the similarity in chemical sensitivity between native and nonnative species, we performed a metaanalysis for 38 cases for which the HC5 values of the chemicals are available for both counterparts. The data set includes 9 chemicals for Chinese versus non-Chinese species, 4 chemicals for Palearctic versus Nearctic species, 4 chemicals for Australian versus non-Australian species, and 21 chemicals for Netherlands versus non-Netherlands fish species. The sensitivities of these organisms are compared based on the ratio between the HC5 value of nonnative species and that of native species generated from species sensitivity distributions (i.e., if the ratio is
Short Communication DO WATER QUALITY CRITERIA BASED ON NONNATIVE SPECIES PROVIDE APPROPRIATE PROTECTION FOR NATIVE SPECIES? XIAOWEI JIN,*y ZIJIAN WANG,z YEYAO WANG,y YIBING LV,y KAIFENG RAO,z WEI JIN,x JOHN P. GIESY,k#yyzz§§ and KENNETH M. Y. LEUNGkk
yDepartment of Analytical Technique, China National Environmental Monitoring Center, Beijing, China zState Key Laboratory of Environmental Aquatic Chemistry, Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, Beijing, China xShijiazhuang Environmental Monitoring Center, Shijiazhuang, China kDepartment of Veterinary Biomedical Sciences and Toxicology Centre, University of Saskatchewan, Saskatoon, Saskatchewan, Canada #State Key Laboratory of Pollution Control and Resources Reuse, School of the Environment, Nanjing University, Nanjing, China yyDepartment of Zoology and Center for Integrative Toxicology, Michigan State University, East Lansing, Michigan, USA zzSchool of Biological Sciences, University of Hong Kong, Pokfulam, Hong Kong, China xxDepartment of Biology and Chemistry and State Key Laboratory in Marine Pollution, City University of Hong Kong, Kowloon, Hong Kong, China kkThe Swire Institute of Marine Science and School of Biological Sciences, The University of Hong Kong, Pokfulam, Hong Kong, China (Submitted 20 January 2015; Returned for Revision 7 March 2015; Accepted 10 March 2015) Abstract: The potential use of toxicity data for nonnative species to derive water quality criteria is controversial because it is sometimes questioned whether criteria based on species from one geographical region provide appropriate protection for species in a different region. However, this is an important concept for the development of Chinese water quality criteria or standards. Data were assembled on 38 chemicals for which values were available for both native and nonnative species. Sensitivities of these organisms were compared based on the 5% hazardous concentration values and the species sensitivity distribution from a literature review. Results of the present study’s analysis showed that there is approximately 74% certainty that use of nonnative species to generate water quality criteria would be sufficiently protective of aquatic ecosystems in China. Without applying any assessment factor to the water quality criteria generated from nonnative species, the uncertainty would be 26% when the native Chinese species might be under protection. Applying an assessment factor of 10 would offer adequate protection to native Chinese species for approximately 90% of tested chemicals and thus reduce the uncertainty from 26% to 10%. Environ Toxicol Chem 2015;34:1793–1798. © 2015 SETAC Keywords: Water quality criteria
Chinese native species
Species sensitivity distribution
approach. The species sensitivity distribution approach uses a statistical distribution estimated from a sample of toxicity data and visualized as a cumulative distribution function. Species sensitivity distributions are used to calculate the concentration at which a specified proportion of species will be affected, such as the hazardous concentration for 5% of species (HC5). Species sensitivity distributions are dependent on available data sets and can differ in type of distribution, taxonomic diversity, and sample size. A substantial amount of toxicity data from several taxonomic groups is required to obtain a robust HC5 [3]. In the species sensitivity distribution, the probability of a threshold for effect being exceeded during either acute or chronic exposures is developed. For instance, in the species sensitivity distribution approach, the probability of a particular proportion of species being affected can be estimated. This approach requires data on the concentration–response relationships for a relatively large number of species. As an example, if data for only 2 species were available, each species would represent 50% of the frequency distribution, and resolution or prediction would be poor. As a rule of thumb, it is suggested that to use the species sensitivity distribution approach, data for a minimum of 20 species should be available [3,4]. In this case, each species would represent 5% of the total, and the resolution of the predictive power of the analysis would be 5%. The species sensitivity distribution approach assumes that the species tested represent a random selection of all the possible sensitivities. If, for instance, the data available were for similar species with similar sensitivities that did not represent the entire range of
METHODOLOGIES FOR DERIVING WATER QUALITY CRITERIA
Water quality criteria are defined as levels of individual characteristics or descriptions of conditions of a water body that, if met, will generally protect the designated use(s) and water quality standards. The methods to conduct these assessments vary among jurisdictions but, in general, have the same elements and suffer from the same limitations and uncertainties. Probabilistic approaches have been used to describe among-species variation in sensitivities [1]. Two basic methods for derivation of water quality criteria are in use or proposed for use throughout the world. These include the assessment factor method and statistical extrapolation methods, such as the species sensitivity distribution approach. Single-point assessment factors are recognized as a conservative approach for dealing with uncertainty in assessing risks posed by chemicals. Yet, current applications of safety factors are based on policy rather than on empirical scientific evidence, and they result in values that are protective but not predictive [2]. The present study focuses on the derivation of aquatic life criteria using the species sensitivity distribution method because it can deliver greater statistical confidence to the risk-assessment process compared with the single-point * Address correspondence to [email protected] or [email protected]. Published online 12 March 2015 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2985 1793
1794
Environ Toxicol Chem 34, 2015
X. Jin et al.
Figure 1. The distribution of the log-transformed ratio of hazardous concentration for 5% of species (HC5 ratio) between native and nonnative species in China (open circles), The Netherlands (closed squares), Australia (triangles), and the Arctic (diamonds; i.e., Palearctic vs Nearctic species). Shaded area represents a zone with the HC5 ratio varying from –2 to 2. Chemicals having 2 different HC5 ratios are in bold letters. 2,4-DCP ¼ 2,4-dichlorophenol; 2,4,6-TCP ¼ 2,4,6trichlorophenol.
possible sensitivities, then predicted probabilities would not be accurate [5,6]. In natural ecosystems there are ranges of species, with different natural histories and in different feeding guilds, that might be differentially exposed or have different sensitivities to stressors. It is impossible to test all of the species that make up the biological community of an ecosystem. For this reason, surrogate species are tested and then used as representatives of sensitive species in the environment. In the semiprobabilistic approach used by the US Environmental Protection Agency (USEPA) to develop water quality criteria, sensitive species in different families are selected to represent aquatic plants, invertebrates, and vertebrates [7]. The USEPA guidelines state that only aquatic species resident in North America can be used as test species to derive the water quality criteria, such as the criteria continuous concentration, for the protection of freshwater and marine ecosystems [8–11]. In Australia, species sensitivity distributions and protective concentrations for 95% of local species are being used to derive water quality guidelines for toxicants [12,13]. It is commonly thought that different ecosystems contain different biological constituents and that a concentration threshold that would be harmless in an ecosystem might lead to irreversible toxic effects in others. The potential use of toxicity data for nonnative species to derive water quality criteria is controversial because it is sometimes questioned whether criteria based on species from a geographical region provide appropriate protection for species in a different region [14]. This uncertainty could not be resolved previously, in large part because of the paucity of toxicity data applicable for local species. In particular, toxicity data, especially for chronic effects on resident species in China, are sparse; and selection of resident species for use in the development of aquatic life criteria values is confounded by variations among ecosystems throughout China. However, this is an important concept for the development of Chinese water quality criteria or standards. There might be differences between species in temperate regions and those in tropical regions [15] or the Arctic and, of course, between freshwater and marine environments [16]. In
fact, tropical species were found to be slightly more sensitive to certain chemical substances than species from temperate regions, so an application factor of 10 was suggested to ensure protection of communities in tropical regions [15]. As a result of conservation of biochemical and physiological mechanisms for maintaining homeostasis in freshwater organisms, it is likely that sensitivities of species within a particular class of organisms would not vary significantly among countries within a climatic region. Therefore, it is unlikely that there would be significant differences between the overall sensitivities of organisms between countries within the temperate zone. WATER QUALITY CRITERIA TO PROTECT NATIVE SPECIES IN CHINA
As China has developed economically and become more industrialized, more attention has been paid to the development of a system to protect the environment [17]. Previously, because of a lack of resources, China often adopted water quality criteria promulgated by other jurisdictions [18,19]. Historically, this lack of data on toxicities of chemicals on Chinese species pragmatically led to the use of toxicological data from foreign toxicity databases of nonnative species (e.g., data extracted from the USEPA’s ECOTOX database) to develop chemical water quality criteria [18]. However, as environmental regulation and monitoring have become more sophisticated, the adequacy of this approach has been questioned. The question has been posed whether it would be more appropriate to develop water quality criteria specific to the situation in China by acquiring data on toxicity of chemicals to species native to China. However, this uncertainty could not be resolved previously, in large part because of the paucity of toxicity data applicable for local species. For the more than 130 chemicals evaluated in the current Chinese Environmental Quality Standards for Surface Water, and even more chemicals needed to reestablish and revise their water quality criteria, data are available for only few species indigenous to China. Currently, the required resources, knowledge, and political
Australian vs non-Australian
Palearctic vs nearctic
Chinese vs non-Chinese
Taxonomic group
a
Ammonia
Phenola
4-Chlorophenol
Endosulfan
Lindane
Diazion
Fenitrothion
Chlorpyntos
Triclosan
Parathion
Nitrobenzene
Cadmium
Mercury
Arsenic (III)
a
Pentachlorophenol
2,4,6-Trichlorophenol
2,4-Dichlorophenol
Chemical
Native Nonnative Native Nonnative Native Nonnative Native Nonnative
Palearctic Nearctic Palearctic Nearctic Palearctic Nearctic Palearctic Nearctic
Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative
Species
17 71 8 6 9 28 13 17
10 36 7 14 6 11 13 6
14 15 13 21 12 60 14 16 47 33 49 50 20 13 32 38 12 12
Data number
Acute Acute Acute Acute Acute Acute Acute Acute
Acute Acute Acute Acute Acute Acute Acute Acute
Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute
Data type
All All All All All All
All All excl. excl. excl. excl. excl. excl.
All All All All All All All All
All All All All All All All All All All All All All All All All All All
fish fish fish fish fish fish
Taxonomic group
Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr Log-Burr
III III III III III III III III
Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal
Log-Burr III Log-Burr III Log-Burr III Log-Burr III Log-Burr III Log-Burr III Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic Log-logistic
Model fit
Table 1. Toxicity parameter of 38 chemicals for native and nonnative species
0.31 0.17 735 3018 8408 2177 96 93
0.09 0.15 0.9 1.6 0.36 0.37 2.07 1.47
1100 1060 790 720 35 57 264 335 3.52 8.69 4.39 2.0 8.86 13.59 0.22 0.02 33.48 48.46
HC5 (mg/L)
–0.51 –0.77 2.86 3.47 3.92 3.34 1.98 1.97
–1.04 –0.82 –0.05 0.20 –0.44 –0.43 0.31 0.17
3.04 3.02 2.89 2.85 1.54 1.75 2.42 2.52 0.54 0.94 0.64 0.30 0.95 1.13 –0.66 –1.69 1.52 1.68
Log HC5
0.97
0.26
4.11
0.55
0.71
1.03
1.78
1.67
1.45
0.09
1.53
0.45
2.47
1.27
1.63
0.91
0.96
Nonnative/native ratio
(continued)
[24]
[24]
[24]
[13]
[19]
[19]
[19]
[19]
[23]
[25]
[25]
[25]
[25]
[25]
[22]
[21]
[14]
Reference
Water quality criteria to protect native species in China Environ Toxicol Chem 34, 2015 1795
Fenitrothion
Diazinon
Atrazine
Heptachlor
Methoxychlor
Azinphos-methyl
Endrin
Carbaryl
Chlorpyrifos
Deltamethrin
Endosulfan
Pentachlorophenol
Cadmium
Phenol
Copper
Trichlorfon
Lindane
Malathion
DDT
Ammonium
Zinc
Chemical Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative Native Nonnative
Species 13 7 10 9 8 7 8 8 8 7 8 7 7 7 8 6 6 7 8 5 6 7 8 4 8 4 5 6 5 6 5 6 5 5 4 6 4 6 5 5 4 4
Data number Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute Acute
Data type Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish Fish
Taxonomic group
Model fit Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal Log-normal
707 288 218 100 1.99 1 10 141 10 10 39.81 79.43 30.20 39.81 5248 2041 10 257 19.95 10 1 0.1 1 3.98 1 3.98 1202 2951 0.60 0.20 0.1 1.99 10 10 3.98 4.07 13 182 5754 39.81 1148 257 1548
HC5 (mg/L) 2.85 2.46 2.34 2 0.3 0 1 2.15 1 1 1.6 1.9 1.48 1.6 3.72 3.31 1 2.41 1.3 1 0 –1 0 0.6 0 0.6 3.08 3.47 –0.22 –0.7 –1 0.3 1 1 0.6 0.61 4.12 3.76 1.6 3.06 2.41 3.19
Log HC5
6.03
28.84
0.44
1.02
1
19.95
0.33
2.45
3.98
3.98
0.1
0.50
25.7
0.39
1.32
1.99
1
14.1
0.50
0.46
0.41
Nonnative/native ratio
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
[26]
Reference
Environ Toxicol Chem 34, 2015
a Because there are no fish data available for Australia, the taxonomic groups for Non-Australian exclude (excl.) fish data. HC5 ¼ hazardous concentration for 5% of species; DDT ¼ Dichlorodiphenyltrichloroethane.
Netherlands vs non-Netherlands
Taxonomic group
Table 1. (Continued)
1796 X. Jin et al.
Water quality criteria to protect native species in China
will are available for China to develop its own water quality criteria to protect a range of aquatic environments. For social and political reasons, it might be preferable to use species native to China that have been generated under conditions that simulate the Chinese environment. However, it would take a long time and a great expenditure of resources to collect sufficient information, especially for probabilistic approaches, to achieve this goal. Furthermore, standard testing protocols have yet to be established for a range of surrogate species based on species endemic to China. To develop these test protocols would require a large investment simply to acquire information on nutritional requirements and culturing conditions for species to be used in standardized tests. Although this is feasible, it might not be the most timely and efficient method or allocation of resources and might unduly delay implementation of water quality criteria. COMPARISONS OF SENSITIVITIES BETWEEN NATIVE AND NONNATIVE SPECIES
Comparisons of sensitivities between species native to China and nonnative species can be made for only a few chemicals, including 2,4-dichlorophenol, pentachlorophenol, and triclosan [14,20–22]. When this comparison was made, there was no significant difference between sensitivities of Chinese and nonChinese taxa. In Australia, sensitivities of Australian and nonAustralian organisms to endosulfan, ammonia, 4-chlorophenol, and phenol, based on calculated HC5 values, were similar [13,23]. Similarly, among North American and European taxa with different geographic distributions, sensitivities to a range of toxicants have been shown to be similar. Moreover, natural history, habitat type, and geographical distribution of the species used to construct the species sensitivity distribution did not have a significant influence on the assessment of hazard; but the taxonomic composition of the species sensitivity distribution does have a significant effect on resulting estimates of the HC5 [24]. To further verify the similarity in chemical sensitivity between native and nonnative species, we performed a metaanalysis for 38 cases for which the HC5 values of the chemicals are available for both counterparts. The data set includes 9 chemicals for Chinese versus non-Chinese species, 4 chemicals for Palearctic versus Nearctic species, 4 chemicals for Australian versus non-Australian species, and 21 chemicals for Netherlands versus non-Netherlands fish species. The sensitivities of these organisms are compared based on the ratio between the HC5 value of nonnative species and that of native species generated from species sensitivity distributions (i.e., if the ratio is
