Toxicant default guideline values for water quality in aquatic ecosystems

Search results provide the DGVs and information to support them.

Medium — whether the DGV applies to freshwater or marine water. For ground water and brackish and hypersaline surface water, see the guidance provided below in Guideline values for other water types.

Reliability classification — DGVs are classified as very high, high, moderate, low, very low or unknown. Classification is mainly based on the number and type (chronic, acute or a mix of both) of data used to derive the guideline value, as well as the fit of the statistical (SSD) model to the data (refer to details).

The reliability classification of all DGVs in the DGVs search tool that are still based on the ANZECC & ARMCANZ (2000) guidelines (i.e. those with a 2000 publication date) have been updated to reflect the  classification in Warne et al. (2018).

Publication date — year of publication of the DGV.

DGVs for different levels of species protection — where DGVs have been derived using the SSD method, guideline values are provided for 99, 95, 90 and 80% species protection. The DGV that is applicable to your situation depends on the current or desired condition of the ecosystem and the associated level of protection that is assigned. In most cases:

• high ecological/conservation value system — apply 99% species protection DGV
• slightly to moderately disturbed system — apply 95% species protection DGV
• highly disturbed system — apply 95% species protection DGV, although there may be cases where the 90 or 80% species protection DGVs can be applied.

Exceptions to this assignment of DGVs may include (i) bioaccumulative toxicants, for which a higher level of protection may be required, and (ii) where a jurisdiction has made a different policy decision on the required level of protection.

Guideline values derived using the less preferred assessment-factor method cannot be related to a percentage of species protected; they are assigned an ‘unknown’ level of species protection. Refer to Level of protection for additional guidance on determining an ecosystem condition and associated level of protection.

Specific comments and general comments — for some toxicants, important context or guidance helps you to better understand the DGV and, in some cases, its implementation.

Downloadable files — we provide links to:

• toxicant DGV technical briefs for details of the DGV, including background on the toxicant and its toxicity, available and final toxicity data, final DGVs and any associated caveats to their use, supporting information and references
• toxicant data table, showing raw data that were considered acceptable for inclusion in the guideline value derivation dataset, and the associated toxicity test details
• toxicant quality assessment worksheet, which provides the details of the quality assessment scores for each of the toxicity data values assessed for inclusion in the DGV derivation.

Default guideline values have been derived for fresh and marine waters but not for groundwater or brackish or hypersaline surface waters.

Groundwater

Groundwater is often considered as a source of chemical stressors to surface waters and has historically been managed as such, with guideline values for the relevant community values being applied at the point the groundwater expresses at the surface. However, while groundwater plays a significant role in sustaining surface waters and their various community values, groundwater ecosystems themselves may be important and require protection. These ecosystems function differently to surface water ecosystems and, hence, DGVs derived to protect surface water ecosystems may not be applicable to groundwater ecosystems. We provide some additional information and guidance on how the Water Quality Guidelines should be applied to groundwater, but this is an area that needs further development.

Brackish and hypersaline water (including estuaries)

There are generally too few ecotoxicological data for brackish (i.e. of a salinity between fresh and marine) and hypersaline (i.e. of a salinity greater than seawater) systems to be able to derive DGVs for these types of waters.

Estuaries are typically brackish and sometimes even hypersaline. In addition to the lack of relevant ecotoxicological data, estuarine systems are highly variable and dynamic, especially in relation to their physico-chemistry (e.g. salinity, turbidity) and, therefore, it is not possible to derive DGVs that would be appropriately applicable to all estuarine waters all the time.

In most cases, it will be necessary to make a professional judgment about an appropriate guideline value for brackish or hypersaline waters. The appropriateness of the freshwater and marine water DGVs should be carefully assessed and compared relative to the local conditions (e.g. water physico-chemistry, local species). This would include examination of the supporting data used to derive the DGVs. Note that it is possible that neither the freshwater nor the marine water DGVs will be appropriate to use. Nevertheless, some guidance is provided below.

For metals that are affected by hardness or other key toxicity modifying factors (e.g. pH, dissolved organic carbon, alkalinity), it is possible to use the freshwater DGVs for brackish waters with salinities up to 2 ppt, by applying the hardness corrections in Accounting for local conditions or other such corrections as may be published in the future for certain metals. The water hardness at this salinity is approximately 350–400 mg/L (as CaCO3), which roughly corresponds to the upper limit of the range of water hardness values for which the hardness corrections were developed (ANZECC/ARMCANZ 2000). The values of the relevant toxicity modifying factors in the water should always be verified before the DGV is corrected. For higher salinities, the marine DGVs for metals are recommended. Where salinity of an estuarine water body changes frequently (i.e. with the tidal cycle), the lowest of the marine and freshwater DGVs for metals is recommended. The above recommendations apply unless available evidence indicates a different approach is needed. For example, decisions would need to take into account the relative reliability of marine and freshwater DGVs, and whether there is or appears to be a difference in toxicity between freshwater and marine water.

For other toxicants, and if there is little relevant information to support a decision, we recommend that you adopt the lower of the two DGVs (freshwater or marine) as an interim measure. The assessment of ecosystem receptor lines of evidence is always recommended as part of a weight of evidence process to improve understanding of the impacts of toxicants (including metals) in estuarine/brackish waters. As a part of this, it might be necessary to derive a locally relevant guideline value.

The primary purpose of the reliability classification scheme for toxicant guideline values is to provide a quick and transparent way of indicating the general level of confidence in a guideline value. The classification scheme can be applied to both DGVs and site-specific guideline values.

Guideline values are assigned a level of reliability based on 4 factors:

• method used to derive the guideline value (SSD or assessment factor)
• number of species for which appropriate quality toxicity data are available (6 to 7, 8 to 14 or ≥ 15)
• type of toxicity data (chronic, a mixture of chronic and converted acute, a mixture of chronic fresh and chronic marine data, or only converted acute values)
• a visual assessment of the fit of the SSD to the toxicity data (good or poor).

There are 6 classes of reliability: very high, high, moderate, low, very low and unknown reliability (Table 1).

^a The sample size is assumed to comprise data from at least four taxonomic groups.

^b This includes all types of chronic data irrespective of whether they are chronic NEC, BEC10, EC10 and NOEC values or estimates of chronic EC10 and NOEC values that were converted from chronic LOEC, MATC or EC50 data. This also includes combined freshwater and marine data where either of the following two criteria have been met: (i) no statistically significant difference between the freshwater and marine toxicity datasets, (ii). based on chemistry and mode of action, no reason to expect differences in toxicity between freshwater and marine water) have been met.

The reliability of a guideline value helps indicate whether it would benefit from the acquisition and incorporation of more toxicity data into the derivation. Higher reliability guideline values can be derived using the method described in Warne et al. (2025).

Ecotoxicity data needed for the derivation can be obtained from:

• more recent data in the literature, including more recent water quality guideline value documents or their equivalent from other jurisdictions
• generating new data.

Refer to Warne et al. (2025) for more details on the reliability classification.

Improved guideline values can be submitted, via the third-party guideline value derivation process, for national consideration and endorsement as DGVs.

Although the Water Quality Guidelines’ DGVs are not mandatory, unless specified as so by the relevant state, territory or local jurisdiction, we provide guidance here on how the DGVs should be applied.

Our DGVs are mostly derived according to risk assessment principles using data from laboratory tests in clean water. They represent the current best estimates of the concentrations of toxicants that should have no significant adverse effects on the aquatic ecosystem. However, and in keeping with the over-riding principle of continual improvement, where the concentration of a contaminant is below the appropriate guideline value, then the over-riding objective should be to continue to improve, or at least maintain, water quality (i.e. not to allow increases in concentration up to the guideline value).

DGVs are based on direct toxic effects of individual toxicants. Where possible, a hierarchy of chemical measurements (that is, total, dissolved, then bioavailable fractions) should be used for comparison with the DGVs. Preferably, you would do this assessment together with the measurement of other lines of evidence, using a weight-of-evidence process.

The reliability of a guideline value helps indicate how you should use that guideline value. Guidance for how to use the DGVs based on their reliability classification is provided in Table 2. This would also apply to site-specific guidelines values.

DGVs in the Water Quality Guidelines classified as very high, high or moderate represent an adequate screening point for assessing water quality in conjunction with other relevant lines of evidence (weight of evidence).

Sometimes, an assessment of water quality using a high or very high reliability DGV in the absence of other lines of evidence may be appropriate. For example, where the DGV is significantly above the measured environmental concentrations.

DGVs classified as either low, very low or unknown reliability should only be used as interim values with a commitment to improving their reliability or deriving site-specific guideline values.

If low, very low or unknown reliability DGVs are used, then the justification for their use must be provided. Due to the lower confidence in the low, very low or unknown reliability DGVs, additional lines of evidence need to be used in conjunction with these DGVs when assessing water quality.

DGVs should always be used in the context of all accompanying supporting information, for example:

• In most cases, DGVs classified as either moderate, low, very low or unknown will have certain data limitations that could be addressed by acquiring more chronic toxicity data
• Very high and high reliability DGVs can have specific limitations, which should have been documented, and that may need to be taken into account. Examples of this are provided in Warne et al. (2025).

See also:

Additional guidance on how a DGV should be applied:

• toxicant DGV technical briefs
• toxicant guideline value derivation method (Warne et al. 2025)
• accounting for local conditions.

You should note that jurisdictions may specify alternative requirements for the use of DGVs, or even site-specific guideline values, based on the guideline value reliability or other information.

DGVs should not be used as blanket values for Australia and New Zealand because ecosystem types vary widely, even on a smaller scale. Variability amongst ecosystems and associated water quality can affect toxicant transport and degradation, bioavailability and toxicity.

In some cases, you can refine DGVs to account for local water quality characteristics, such as water hardness, dissolved organic carbon concentration and naturally elevated background concentrations.

We provide guidance on how to compare DGVs or site-specific guideline values to monitoring data.

The guidance provided here is general and jurisdictions should always be consulted for any additional specific requirements for determining exceedances of toxicant guideline values for aquatic ecosystem protection.

Determining exceedances of DGVs should be done as part of a weight of evidence approach for water/sediment quality assessment, where relevant pressure, stressor and/or ecosystem receptor indicators are measured to make an overall assessment of whether water/sediment quality is acceptable.

The Monitoring section of the ANZG website provides guidance on comparing monitoring data to guideline values. This guidance recommends that the 95th percentile of the monitoring data for a toxicant be compared with the DGV. Thus, a DGV shall be regarded as being exceeded if the 95th percentile of the monitoring data, having been rounded according to the Australian Standards SAA 2706-2003 to the same number of significant figures in the DGV, exceeds the DGV, taking into account the accuracy of the chemical analysis for the toxicant.

This approach is more stringent than for guideline values derived using the reference site approach (e.g. the physical and chemical stressor DGVs), because the toxicant DGVs are based on actual biological effects data and, so, by implication, exceedance of a DGV indicates a potential for ecological harm. Because the proportion of values required to be less than the DGV is very high (95%), a single observation greater than the DGV would be legitimate grounds for determining that an exceedance has occurred in most cases, even early in a sampling program. However, the need or otherwise for subsequent management action would depend on an overall assessment of risk/impact based on all the lines of evidence.

As noted above, the accuracy of the chemical analysis for a toxicant in a water sample should also be taken into account when determining an exceedance of a DGV. It is essential that analytical measurements are as accurate as possible given available analytical techniques. Commercial analytical laboratories will report the uncertainty of an analysis, such as a 95% confidence interval around the measurement, upon request. Consideration may need to be given as to whether the lower 95% confidence interval of the measurement does or does not exceed the DGV. The implications of this may differ depending on the purpose of the assessment (e.g. regulatory compliance versus general ecosystem health/condition assessment) and also between jurisdictions.

Finally, and in keeping with the over-riding principle of continual improvement, where the concentration of a toxicant is below the appropriate guideline value, then the over-riding objective should be to continue to improve, or at least maintain, water quality (i.e. not to allow increases in concentration up to the guideline value).

Why are significant figures important for default guideline values?

Water quality guideline values for toxicants, including default guideline values (DGVs), have uncertainty associated with them that is due to numerous factors including (but not necessarily limited to): 

• Toxicity tests from which data are used to derive guideline values:
  • Analytical measurements from the toxicity tests
  • Variability of the test organisms’ responses in the toxicity tests
  • Model uncertainty in the concentration-response models used to characterise toxicity
• The species sensitivity distribution:
  • Uncertainty associated with using non-preferred estimates of chronic toxicity (e.g. NOEC values, converted acute values)
  • Uncertainty associated with the fit of the selected distribution
• Extrapolations:
  • Extrapolating effects from the laboratory to the field
  • Extrapolating from few species and taxonomic groups to the community/ecosystem. 

Most of the uncertainty in a guideline value is unquantified and, where it is quantified (see below), it is often not reported. However, it is important that the level of precision to which a guideline value is reported is consistent with the magnitude of the uncertainty in the value. Reporting a guideline value to too many significant figures will give a misleading impression of the precision of the estimate. Although not all the uncertainty in a guideline value is quantified, the uncertainty in the fit of the selected distribution can be quantified and is likely to be at least as large as the other sources of uncertainty. Thus, this uncertainty is used as the basis for determining the number of significant figures for toxicant DGVs, and can also be used as such for site-specific guideline values.

Determining the appropriate number of significant figures for default guideline values

From a statistical modelling perspective, the estimate of uncertainty of an estimated x% species protection (PCx) value (note that it is the PC80, PC90, PC95 and PC99 values that ultimately become the DGVs) is the standard error (SE). Normal scientific practice is that the significant figures to which a value is reported should be no more than the first significant figure in the SE of the value. Therefore, the following rule for determining the number of significant figures for DGVs should be applied: 

The significant figure to which a DGV is reported must correspond to the place of the first significant figure in the SE of the DGV. Where the SE is greater than the DGV, the DGV must be reported to one significant figure.

The standard errors for the PCx values can be obtained from the approved toxicant guideline value derivation software, shinyssdtools. To assist users with the rule, several worked examples are provided below:

• For a DGV ± SE of 5.6738 ± 0.4536 µg/L: the first significant figure of the SE is at the 10^-1 place; thus the DGV is reported to the 10^-1 place, as 5.7 µg/L, corresponding to two significant figures.
• For a DGV ± SE of 1713 ± 1200 µg/L: the first significant figure of the SE is at the 10³ place; thus the DGV is reported to the 10³ place, as 2000 µg/L, corresponding to one significant figure.
• For a DGV ± SE of 0.335 ± 3.53 µg/L: the SE is greater than the DGV; thus the DGV is reported to one significant figure, as 0.3 µg/L.

As a further example, Table 1 provides the outcome of an assessment of the number of significant figures for ANZG (2021) boron in freshwater DGVs. 

^a PCx level: The percent of species predicted to be protected.

Rounding for default guideline values and measured environmental concentrations

Default guidelines values need to be rounded to the appropriate number of significant figures, while measured environmental concentrations need to be rounded to the same number of significant figures as the DGVs when they are being compared to the DGVs. All rounding of DGVs and measured environmental concentrations should be done in accordance with the “rounding to evens” rule set out in Australian Standard SAA 2706-2003 (Standards Australia 2003). Notably, when environmental data are reported independently of comparisons with guideline values, they should be reported to the number of significant figures that is commensurate with the precision of the analysis.

The “rounding to evens” rule adopts the most common convention for rounding – that is, rounding up when the value following the last value to be retained is >5 and rounding down when the value following the last value to be retained is <5. However, when the value following the last value to be retained is exactly 5 (i.e. there are no non-zero numbers to the right of this 5), then the value should be rounded to the closest even number. For example, when rounding to two significant figures, 7.450 (and also 7.45) would be rounded down to 7.4, while 7.550 (and also 7.55) would be rounded up to 7.6. In contrast, 7.4501 would be rounded up to 7.5, while 7.5501 would still be rounded up to 7.6. More examples can be found in Standards Australia (2003).

What if two adjacent DGVs are rounded to the same value?

Where uncertainty in the DGVs is high (i.e. the SE is large), it is possible that two adjacent DGVs (e.g. PC95 and PC90) would be rounded to the same final value. For example, if PC95 and PC90 values of 1.6 and 2.4 µg/L, respectively, needed to be rounded to one significant figure, it would result in both final values being 2 µg/L. This indicates that the uncertainty in the PCx estimates is too high to be able to differentiate between the 95% and 90% species protection levels at the selected number of significant figures. Ideally, additional toxicity data would be generated to add to the dataset and reduce the uncertainty in the DGVs, but this is often not possible in the short to medium term.

Where two DGVs are rounded to the same value, decisions on which DGV to adopt or other action to take should be made in consultation with the local jurisdiction. Some options on what to do in such a situation are provided below, based on the above hypothetical case where the rounded PC95 and PC90 values are the same (referred to below as the PC95/90 value). In considering the below options, it should be remembered that the Water Quality Guidelines recommend the use of a weight of evidence approach to water and sediment quality assessment that does not rely on the DGVs alone.

• For a high ecological value system – the PC99 would typically still be applied
• For a slightly to moderately disturbed system:
  1. err on the side of caution and adopt the PC99 value instead of the PC95/90 value, to provide greater confidence that the ecosystem will be protected;
  2. derive site-specific guideline values that are more reflective of local conditions; or
  3. adopt the PC95/90 value but implement additional ecosystem receptor monitoring to provide greater assurance that the ecosystem is still being protected. Notably, local stressor and ecosystem receptor data collected over a sufficient period of time will provide knowledge that can potentially be used to refine guideline values to local conditions, thus, reducing the reliance on the DGVs.
• For a highly disturbed system – the PC95/90 or PC80 would typically be applied depending on the system and nature of the issue.

Decisions might also be guided by the purpose of the assessment (e.g. general ecosystem health monitoring, wastewater discharge licence compliance). Also, users should refer to the Levels of Protection web page for more guidance on how to apply the Water Quality Guidelines to different ecosystem conditions.

Updates to the guideline value derivation method

The approved toxicant guideline value derivation method documented by Warne et al. (2025) includes the new rules for the number of significant figures and rounding. 

Deviations from the guidance

Any deviations outside the significant figures and rounding guidance for toxicant guideline values need to be based on best professional judgment that is technically sound and defensible, and fully documented.

It might be deemed appropriate to derive a site-specific guideline value where there is no DGV for a toxicant. The Water Quality Guidelines provide guidance on deriving site-specific guideline values, while complementary and additional guidance can also be found in van Dam et al. (2019) and Huynh & Hobbs (2019). State and territory governments may also have their own guidance for deriving and applying site-specific guideline values, and you should always consult with them at the outset on appropriate methods.

You could derive a site-specific guideline value using either reference-site data (for instance  80th percentile of background or reference-site data) or biological-effects data.

In some cases, it might be more beneficial and cost-effective to derive a DGV that can be applied to your site, and then propose it for inclusion in the Water Quality Guidelines.

ANZECC & ARMCANZ 2000, Australian and New Zealand Guidelines for Fresh and Marine Water Quality, Australian and New Zealand Environment and Conservation Council and Agriculture and Resource Management Council of Australia and New Zealand, Canberra.

Batley, GE, van Dam, RA, Warne, MStJ, Chapman, JC, Fox, DR, Hickey, CW & Stauber, JL 2018, Technical Rationale for Changes to the Method for Deriving Australian and New Zealand Water Quality Guideline Values for Toxicants, Australian Government Department of Agriculture and Water Resources, Canberra.

Dalgarno, S 2018. ssdtools: A shiny web app to analyse species sensitivity distributions. Prepared by Poisson Consulting for the Ministry of the Environment, British Columbia. https://bcgov-env.shinyapps.io/ssdtools/.

Huynh T & Hobbs D. 2019.  Deriving site‐specific guideline values for physico‐chemical parameters and toxicants. Report prepared for the Independent Expert Scientific Committee on Coal Seam Gas and Large Coal Mining Development through the Department of the Environment and Energy. Canberra (AU).

OECD 1992, Report of the OECD Workshop on Extrapolation of Laboratory Aquatic Toxicity Data to the Real Environment, OECD Environment Monographs No. 59, Organisation for Economic Co-operation and Development, Paris.

OECD 1995, Guidance Document for Aquatic Effects Assessment, OECD Environment Monographs No. 92 (Series on Testing and Assessment: Ecotoxicity Testing No. 3), Organisation for Economic Co-operation and Development, Paris.

Warne, MStJ, Batley, GE, van Dam, RA, Chapman, JC, Fox, DR, Hickey, CW & Stauber, JL 2018, Revised Method for Deriving Australian and New Zealand Water Quality Guideline Values for Toxicants, Australian Government Department of Agriculture and Water Resources, Canberra.

Standards Australia 2003. Numerical values - Rounding and interpretation of limiting values (Reconfirmed 2018). AS 2706-2003. Standards Australia, Canberra, ACT, 15 pp.

Warne MStJ, Batley GE, van Dam RA, Chapman JC, Fox DR, Hickey CW, Stauber JL and Fisher R 2025. Method for deriving Australian and New Zealand water quality guideline values for protecting aquatic ecosystems from toxicants – update of 2018 version. Prepared for the Australian and New Zealand Guidelines for Fresh and Marine Water Quality. CC BY 4.0. Australian and New Zealand Governments and Australian state and territory governments, Canberra, ACT, Australia.

van Dam RA, Hogan AC, Humphrey CL & Harford AJ 2019, How specific is site-specific? A review and guidance for selecting and evaluating approaches for deriving local water quality benchmarks, Integrated Environmental Assessment and Management 15: 683–702.​​