 |
||||||
       |
||||||
 |
|
|
|||||||
|
Water Quality Derivation
of Water Quality Criteria to Protect Aquatic Life
Canadian Cataloguing
in Publication Data
This document has been prepared to outline clearly the procedure used to derive water quality criteria in BC. This procedure will help to identify data gaps, encourage research and provide a sound basis for defensible criteria. It will also provide a consistent format for the derivation of criteria and serve as a checklist to ensure the appropriate information has been considered. The following is a brief overview of the procedure used to derive water quality criteria to protect aquatic life in BC. Substances of concern at the provincial level are identified and ranked for criteria development after consultation within the Ministry of Environment, Lands and Parks (now called Ministry of Water, Land and Air Protection). Substances are then selected for criteria development after consultation with federal and other provincial jurisdictions to avoid duplication of efforts. For each substance selected, a literature search is conducted to obtain information on the following:
To proceed with criteria derivation, certain minimum toxicity and environmental fate data requirements should be met. In cases where there is insufficient information to set criteria, interim criteria can be derived providing that a less stringent data set is available. Key toxicity studies found in the literature search are evaluated to ensure that acceptable laboratory practices were used in the design and execution of the experiments. Each key study is judged on its scientific acceptability. When available, the lowest reliable LC50 or EC50 from an acute toxicity test and the lowest-observed-effect level (LOEL) from a reliable chronic exposure study, on sensitive native BC species, are selected. These values are then multiplied by an appropriate safety factor to derive an acute and a chronic criterion. For certain substances, only a single criterion is set which is based on the LOEL from a chronic exposure study or on bioaccumulation. Other factors taken into account include no-observed-effect levels (NOEL) and ambient background concentrations for naturally occurring substances. Alternatively, the most sensitive LC50 or EC50 from an acute exposure study is multiplied by an acute/chronic ratio or appropriate application factor to determine an interim criterion concentration.
BC Environment (now called Ministry of Water, Land and Air Protection) is developing province-wide ambient water quality criteria for substances or physical attributes that are important in both fresh and marine surface waters of British Columbia. This work has the following goals:
The definition adopted for criterion is: A maximum and/or minimum value for a physical, chemical or biological characteristic of water, sediment or biota, applicable province-wide, which should not be exceeded to prevent specified detrimental effects from occurring to a water use, including aquatic life, under specified environmental conditions. The criteria are province-wide in application, but use-specific, and are being developed for the following water uses:
This document describes how criteria are derived to protect aquatic life, and applies to toxic chemical substances more than the physical properties of water (e.g., temperature, pH, suspended solids). Derivation of water quality criteria to protect other water uses will be described under separate cover. Neither criteria, nor objectives which are derived from them, have any legal standing. They are intended as a tool to provide policy direction to those making decisions affecting water quality provided that they do not allow legislated effluent standards to be exceeded. The objectives can be used to establish the allowable limits in waste discharges. These limits are set out in waste management permits, plans, or operating certificates which do have legal standing. The objectives are not usually incorporated as conditions of the permit.
Background This document is required to:
The approach used by the province to derive water quality criteria is similar in many respects to that used by the Canadian Council of Ministers of the Environment (CCME) to derive national water quality guidelines (which are analogous to BC criteria). The latter is outlined in "Protocol for the Derivation of Water Quality Guidelines for the Protection of Aquatic Life"; April, 1991 (CCME, 1991). This similarity is due, in part, to input by BC Environment scientists as members of the CCME review team. The CCME document was chosen as a template upon which to build and refine these Provincial procedures. Certain fundamental aspects of criteria derivation in BC differ from the CCME approach. One change worth noting is that, for some water quality variables, both acute and chronic criteria are recommended by the Province to address acute and chronic toxicity independently, whereas the CCME specifies only a single concentration to address all conditions. In developing the BC approach, a review of all existing approaches (including the US EPA Water Quality Criteria, and the Ontario and CCME Water Quality Guidelines) was performed. In addition, we recognized that scientific judgement is an important and vital part of the process for deriving practical and useable water quality criteria. Accordingly, this document is designed to serve as a guide for those involved in deriving water quality criteria; it is not intended as a protocol to be followed rigidly in every respect. Water quality criteria to protect aquatic life have been prepared for many substances of concern and criteria for other substances continue to be prepared including other priority substances judged to be most urgently needed for water quality assessments and objectives in BC. Until criteria for certain substances are approved by the Ministry Executive, the Water Quality Branch is using what is termed 'working' criteria for water quality, many of which have been recommended by the CCME - formerly known as the Canadian Council of Resource and Environment Ministers (CCREM). This is consistent with Ministry policy that the CCME Guidelines are to be used in developing water quality criteria and objectives and in assessing water quality, but recognizes that site-specific factors may necessitate modification of the CCME Guidelines (Pommen, 1991).
Guiding Principles for the Development of Water Quality Criteria for Aquatic Life There are several over-riding principles used in developing water quality criteria in BC. These are:
Data Requirements for Criteria Derivation To set water quality criteria, certain basic data should be available. Where insufficient data are available to set criteria, interim criteria may be set. The interim criteria may be upgraded to full criteria status when the data gap is filled. While minimum data requirements have been recommended for both criteria and interim criteria, it is important to emphasize that these are intended as a guide, not as a strict requirement. Flexibility and the use of scientific judgement as well as innovative new approaches are recognized as necessary and important components of the derivation process. For example, consideration must be given to the nature of the substance such as its mode of toxic action, its bioaccumulation potential, or if it exhibits delayed toxicity. Exemptions from the minimum data requirements may be considered on a case-by-case basis provided they are documented and scientifically justified. The final decision of whether criteria or interim criteria are recommended is based, in part, on the confidence the authors have in the criteria. If interim criteria are set, then it is the responsibility of the authors to justify their position and to recommend the information needed to elevate interim criteria to full criteria status.
The goal of freshwater aquatic criteria is the protection and maintenance of all forms of aquatic life and all aquatic life stages in the freshwater environment. Therefore, it is essential that data for fish, invertebrates, and plants be included in the criteria derivation process. For this purpose, minimum data requirements have been recommended. In the derivation process (see Evaluation of Toxicity Data), criteria or interim criteria may be derived from studies involving species not required in the minimum data set (e.g., amphibians, protozoa, bacteria), when reasonable justification exists.
Fish
The reduced requirements for plant toxicity studies were deemed necessary because fewer studies on plants have been conducted (Swanson and Peterson 1988). The minimum data requirements for plants could be increased in the future if data availability improves. In cases where the minimum data requirements for criteria derivation are not met, interim water quality criteria may be developed provided the minimum data set requirements are met.
Fish
If a toxicity study indicates that a plant species is the most sensitive species in the data set, then this study shall be used in the interim criteria derivation process. However, in the absence of data on plants, interim criteria can be derived provided that this data gap is noted. The information that is required to elevate interim criteria to full criteria status needs to be clearly identified to stimulate research that will generate the necessary data.
Recognizing that toxicants may react differently in marine water than in fresh water, and that different species are involved, the data requirements are different to reflect the need for separate criteria for the marine situation. This need for separate marine criteria has been demonstrated by the US EPA and supported by the CCME. For most substances, however, there are fewer data available for marine species, particularly phytoplankton and macroalgae, than are available for the fresh water environment (Hansen 1989). Since the goal of marine aquatic criteria is the protection and maintenance of all forms of aquatic life and aquatic life stages in the marine environment, it is recommended that data for marine fish, invertebrates, and plants be included in the criteria derivation process. As with the requirements for fresh water aquatic life criteria, minimum data requirements have been recommended. In this data set, marine species include those species found in estuarine, coastal, and open-ocean habitats, any of which may be used to derive a criterion or interim criterion.
Fish
Invertebrates
In cases where the minimum data requirements are not met, interim water quality criteria can be derived provided the minimum data requirements are met.
Fish
If a toxicity study indicates that a plant species is the most sensitive species in the data set, then this study shall be used in the interim criteria derivation process. However, in the absence of data on plants, interim criteria can be derived provided that this data gap is noted. As with freshwater aquatic life criteria the information that is required to elevate interim criteria to full criteria status needs to be clearly identified to stimulate research that will generate the necessary data.
In addition to the minimum toxicity data requirements outlined above, studies that have investigated the major environmental fate processes and persistence of the substance in water, soil, sediment, air and biota are required. Potential fate processes include volatilization, hydrolysis, oxidation, photolysis, aerobic and anaerobic biodegradation, long-range transport, soil and sediment sorption/desorption, bioconcentration and bioaccumulation. It is not necessary to have information on each potential fate process. Rather, the intent is to be able to identify the major environmental pathways and fate of a substance in the aquatic environment. Specifically, the following should be determined:
Where possible, the persistence of a substance should be expressed in terms of its half-life. Where significant environmental fate information is lacking, interim criteria are set. In these cases, the information required to elevate the interim criterion to full criterion status needs to be clearly identified to stimulate the necessary research.
The following are not required elements of the minimum data set, but should be included when available because they are useful in assessing the potential hazard of a substance:
Since standard protocols for toxicity testing may become outdated or are not always available or followed, a great deal of variability exists in the quality of published data. To ensure a consistent scientific evaluation for each substance, the data included in the minimum data set should meet certain standards. These include information on test conditions/design (e.g., flow-through, renewal, static), test concentrations, temperature, hardness, pH, adjuvants (i.e., synergistic effects), experimental design (controls, number of replicates) and a description of the statistics used in evaluating the data. A variety of standardized test protocols have been developed for fish, invertebrates and plants. When appropriate, these should be consulted during the evaluation process (for example, see BC Ministry of Environment (Ministry of Water, Land and Air Protection) 1982; EPS 1980; ASTM 1980; OECD 1981; Rand and Petrocelli 1985; US EPA 1985a, 1985b, 1985c; Sergy 1987; Swanson and Peterson 1988). Information useful for interpreting toxicity data is also available (Buikema et al. 1982; Rand and Petrocelli 1985, ch. 1-11) and should be consulted when necessary. When consulting test protocols, it is important to be aware of the following limitations:
Therefore, it is essential that the evaluation of toxicity data not follow a rigidly fixed format. Once evaluated, key data are classified as primary, secondary, or unacceptable as described. All data included in the minimum data set should be primary for criteria derivation to proceed. For interim criteria derivation, primary or secondary data may be used. In either case, a weight-of-evidence approach always should be an underlying principle of criteria derivation. Unacceptable data cannot be used in either derivation procedure. Primary Data
Secondary Data
There are four levels or categories of water quality criteria to protect aquatic life in BC. These are:
The choice of which level to apply depends on a number of factors such as the quantity and quality of toxicity data, and the nature of the substance. Criteria derived from chronic studies are preferably based on the lowest-observed-effect level (LOEL), using a non-lethal endpoint for the most sensitive life stage, of the most sensitive aquatic species investigated. However, when these types of data are unavailable, interim criteria can be derived from acute studies by converting short-term median lethal or median effective concentrations (LC50, EC50) to long-term no-effect concentrations using acute/chronic ratios or safety factors. Species not required in the minimum data set (e.g., amphibians) may be used in either derivation procedure provided that the life stage under investigation is completely aquatic. In addition, bioconcentration data may be used to derive criteria to protect the organisms, or consumers of the organisms, from harmful effects. Each study chosen for the criteria derivation procedure must have demonstrated a clear dose/response relationship and the LOEL must be statistically significant (95% confidence level).
Qualifications and Setting Criteria Another condition that must be met to qualify for this dual approach is that sufficient acute and chronic data must be available to set both an acute and chronic criterion (see Minimum Aquatic Toxicity Data Requirements for Freshwater Criteria and Minimum Aquatic Toxicity Data Requirements for Marine Criteria). This decision is not always possible at this stage, especially if the toxicity of a substance is affected by some environmental factor such as water hardness or pH. However, there is usually some indication of such a relationship in the scientific literature. To test this relationship, the toxicity data are plotted against the modifying environmental factor. The acute and chronic data are identified by different symbols. This graphical presentation summarizes the toxicity data and serves several useful purposes in the process of criteria derivation and evaluation, as well as during their application. These are:
Once the key acute and chronic data have been identified (i.e., the relevant lower concentrations that induce acute and chronic toxicity in the most sensitive species tested), they are evaluated in terms of their scientific soundness and rated as primary, secondary, or unacceptable (see Data Classification). Appropriate safety factors (typically between 0.1 and 0.5) are then applied to the primary key acute and chronic data to derive acute and chronic criteria. If NOELs for sensitive life stages of sensitive species fall within this safety range for chronic data (i.e., between the LOEL and the calculated safe value), then the NOEL may be adopted as the chronic criterion. It should be noted that the magnitude of the safety factor may vary from substance to substance depending upon the quality and quantity of toxicity data (the toxicity of some substances is well-defined so that the safety factor need not be as large as for other substances less well understood). The actual size of the safety factor is decided on a case-by-case basis and involves the use of scientific judgement to maintain some flexibility in the derivation process. Ambient background concentrations for substances that occur naturally may also play a role in the size of a safety factor. Criteria set far below levels that occur naturally in BC waters, and in which aquatic life thrive, would be impractical and unusable for assessing the environmental impact of anthropogenically generated substances. When there is a relationship between the toxicity of the substance and some modifying environmental factor, then the criterion may be specified in terms of a regression equation and shown on a graph.
The averaging period for the chronic criterion may differ depending upon the substance under investigation and is somewhat arbitrary (e.g., five to 30 days have been used for BC water quality objectives). These times were chosen as reasonable and practical durations to address chronic effects and to fit into monitoring timetables for provincial agencies. Five samples are considered the minimum needed to calculate the average; however, in some cases where the concentrations fluctuate widely in nature, more than five samples may be necessary. On the other hand, if concentrations are uniform and rarely exceed the chronic criterion, then less frequent monitoring may be justified. In this case, failure of any individual sample to meet the chronic criterion would serve as an alert signal to increase the monitoring. For some substances, such as residual chlorine, the BC criteria are time-related whereby the averaging periods for ambient monitoring are based on the toxicity exposure-duration data (Singleton 1989). The minimum duration of the averaging period for residual chlorine is set at the threshold of chronic toxicity. For freshwater this threshold is four days, but for marine waters it is only two hours.
This category is employed for those substances that defaulted from the Derivation of Acute and Chronic Criteria from Acute and Chronic Studies for such reasons as:
The derivation process for this category is basically the same as that for the chronic criterion in the foregoing category (Derivation of Acute and Chronic Criteria from Acute and Chronic Studies). This single criterion typically is based on the LOEL using a non-lethal endpoint and multiplied by an appropriate safety factor (usually between 0.1 and 0.5). A NOEL may be used if it falls within this range provided it is based on the most sensitive life-stage of a sensitive species native to BC waters. This approach is used to derive a preliminary water quality criterion regardless of whether bioconcentration is a concern. When bioconcentration of a particular substance is a concern, then an additional assessment must be made. If the bioconcentration assessment results in a safe limit lower than the preliminary water quality criterion, then the preliminary criterion can be adjusted accordingly. This derivation process is described more fully in Derivation of Water Quality Criteria from Bioconcentration Data which follows.
When a substance bioconcentrates from the water into the tissues of an organism, a separate assessment of bioconcentration in the criteria derivation process is required. To derive a water quality criterion from bioconcentration data, some basic information is necessary:
To derive a water quality criterion based on the above information, the lowest tissue residue level that induces a harmful effect in the exposed organism or its consumers should be determined. This value is then divided by the highest reliable BCF to derive a LOEL for the water. To derive a water quality criterion (WQC) to protect the organism from accumulating harmful body burdens, the LOEL is multiplied by an appropriate safety factor (typically 0.1 to 0.5 - see Derivation of Acute and Chronic Criteria from Acute and Chronic Studies) as follows: WQC = (Lowest Harmful Tissue Residue Level x Safety Factor)/ BCF If a no-harmful-effect-tissue-residue is available, then this value may be used to replace the lowest harmful tissue residue level provided the effects studied involve sensitive chronic endpoints of sensitive species. If this no-effect value is used, then a safety factor may not be required. To determine the final water quality criterion, the value calculated here is compared to the preliminary water quality criterion determined in Derivation of a Single Criterion from Chronic Studies. The final water quality criterion should be the most scientifically defensible of the two values. However, given the variability of BCFs for many substances (e.g., laboratory-derived BCFs ranging over three orders of magnitude have been measured for some substances), and the subjective nature of application factors, a statistical value, such as the geometric mean, may be considered as the final criterion if there is a wide range between the values derived by each method. The rationale for this alternative statistical approach is that if the values derived by the two methods are similar, then there is a high level of confidence in the criterion. However, when the range between the two values is wide for a particular substance, there is less confidence that either of the values are accurate. Hence, the assumption was made that the safe level (criterion) probably lies somewhere between the two values.
The procedure for the derivation of interim water quality criteria is similar to that used to derive criteria, except that the minimum data requirements are not as rigorous (see Minimum Aquatic Toxicity Data Requirements for Freshwater Criteria and Minimum Aquatic Toxicity Data Requirements for Marine Criteria). In addition, secondary data (Data Classification) are acceptable for the derivation of interim criteria, but unacceptable data should not be used. Chronic data are preferred over acute data as a basis to derive an interim water quality criterion. When available, acute/chronic ratios (ACR) can be used to convert the median lethal results of a short-term study to an estimated long-term no-effect concentration for the most sensitive species for which chronic results are unavailable(Kenaga 1982). An ACR is calculated by dividing an LC50 or EC50 by the no-observed-effect level (NOEL) from a chronic exposure test for the same species (i.e., LC50/NOEL). It is important to note that an ACR should only be used from studies that were designed for this purpose to avoid complications arising from different test conditions or different test populations. Further, the use of an ACR needs to be carefully rationalized since the available evidence indicates that for a given substance, ACRs may vary between species with different sensitivities, and across major taxonomic groupings (Mount 1977; Stephan 1985). The interim criterion is derived by dividing the most sensitive LC50 or EC50 by the most appropriate ACR. In the event that acute/chronic ratios are not available, the alternate method of choice to derive an interim criterion value from an acute study is to multiply the LC50 or EC50 value by a universal application factor. At present, ACRs are not available for all substances and, to meet this situation, universal application factors have been widely used (US EPA 1972). The application factor (AF) for non-persistent substances (half-life in water less than 8 weeks) is 0.05; for persistent substances, the AF is 0.01. These application factors are now endorsed by the majority of Canadian jurisdictions involved in developing water quality criteria, guidelines, or objectives (e.g., CCME, International Joint Commission, Ontario, Manitoba, and Saskatchewan). However, it must be emphasized that, although the universal application factors have been empirically tested and supported (e.g., Kenaga 1982), several studies (Mount 1977; Buikema et al. 1982; Mayer et al. 1986) have suggested that these factors may be inappropriate for several substances (e.g., diazinon, zinc). Therefore, universal application factors for deriving an interim criterion should be used only in the absence of chronic data and in the absence of ACRs for acute data. The information that is required to elevate interim criteria to full criteria status needs to be clearly identified to stimulate research that will generate the necessary data.
The following steps must be followed to establish criteria as Ministry policy:
The review time should be limited to about one month for each of the first and second drafts. Copies of the report are made available through mailing lists, the internet, libraries, and requests to the Water Quality Branch of the Ministry of Environment, Lands and Parks (Water, Land and Air Protection) in Victoria. On occasion, the proposed Water Quality Criteria may need to be incorporated into site remediation plans or an Operational Certificate prior to formal approval through the above process. The proposed Water Quality Criteria should still be put through the formal review and approval process so that it can be treated as policy for future use within the Ministry. The criteria are subject to review and revision as new knowledge becomes available, or as circumstances dictate.
|
||||||||
|
|