Frequently Asked Questions
1. The BLMs are too complicated to implement and interpret in a regulatory framework
Simplified BLM tools have been developed and are simple Excel spreadsheets that can either be used on any PC or alternatively the underpinning calculations can be embedded into laboratories or the regulatory compliance checking process.
2. The BLMs would require dedicated staff to run them and understand them
If used as stand-alone tools, many samples can be run through the Simplified BLM tools in a batch process that just requires data entry of the monitoring results. Simplified BLM tools can also be integrated into laboratory systems for automated production of outputs. The outputs are readily interpretable, including a simple risk characterization ratio.
3. The calculation of backgrounds is much easier to do
The use of natural background concentrations is not a replacement for bioavailability consideration. The calculation of backgrounds has much less scientific and technical pedigree than accounting for bioavailability using BLMs and so bioavailability should be given a greater influence than backgrounds when checking compliance, permitting etc. Backgrounds can be considered in the tiered approach following account being taken of bioavailability.
4. The BLMs or User-friendly BLM tools would be expensive to run and require more trained staff
In addition to being technically robust, the user-friendly BLM tools are freely available and simple to use, so minimal training is required to use them. In many cases, the help pages should provide sufficient information for users. In some cases it is possible for the estimation of bioavailability corrections to be automated within laboratory information management systems, thus reducing the need for resources.
5. The BLMs or user-friendly BLM tools require too many additional inputs that we do not measure
The user-friendly tools require a maximum of 4 inputs. These are the dissolved metal concentration, pH, calcium and dissolved organic carbon (DOC). Data for pH and calcium are usually more readily available than DOC. If site specific monitoring data for the key parameters are not available, default values based on historic data may be used. The full BLMs do require more inputs but these can be estimated from calcium using a freely available Excel calculation (from a peer-reviewed journal article) if there are no monitoring data.
6. We have no DOC data. Therefore we cannot implement the method.
DOC has an important influence on bioavailability so it is preferable to use actual monitoring data. However, in the absence of DOC monitoring data it is possible to use precautionary default values based on read across from similar catchment types or to estimate DOC concentrations from other data that is available such as UV absorbance or dissolved iron (Peters 2011, ).
7. The science on ‘bioavailability’ is not well developed
The science underpinning the understanding of bioavailability and BLMs is well studied. More than 500 papers have been published in the scientific literature on BLMs since 2000. SCHER Opinions over the last 4 years have supported the use of bioavailability-based approaches in ESR metals risk assessments and recent EQS guidance documents. The EU Technical Guidance for Deriving Environmental Quality Standards (TGD-EQS - EC 2011), supported by SCHER includes the use of BLMs for setting EQSs for metals. It is also notable that REACH guidance recognizes the use of BLMs in establishing Generic Exposure Scenarios for metals.
8. Where is the technical evidence to support this approach
There are a number of technical reviews of BLMs within the literature and the evidence is also presented and reviewed in the relevant ESR metals risk assessments. One key piece of evidence is that the predictions of toxicity from BLMs match what is observed in the field remarkably well and usually within a factor of 2.
9. These models don’t cover all aquatic species, what about the species for which there are no BLMs?
Studies on different species have shown that the models used are broadly applicable between different species (the binding constants for both toxic metals and competing ions show remarkable consistency between different species) the BLMs are therefore applied to additional species by defining the sensitivity to the toxic metal (which is expressed as the fractional occupancy of the biotic ligand at the threshold level, e.g. EC10.)
10. The BLMs or user-friendly BLM tools do not account for dietary uptake of metals
For those metals where BLMs have been developed, the evidence is that direct toxic action of the metal on a receptor or ‘biotic ligand’ is the most sensitive endpoint. Therefore, the protection provided by using bioavailability and BLMs is more important and relevant than dietary uptake.
11. According to De Laender et al., (2005) & De Schamphelaere, (2003) toxicity for both Cu and Zn can be underestimated when applying BLMs on fresh waters with elevated levels of humic substances, Al, Fe and low pH. As these conditions are common here in Sweden what is the solution used in this model to prevent this?
It is not clear which specific papers are referred to here, although it is customary when performing BLM calculations for the purpose of EQS compliance assessment to assume that only 50% of the DOC is actually active. This is to ensure that DOC which may be inactive with respect to metal binding does not result in unprotective estimates. This approach was agreed under ESR some years ago.
12. Which geochemical model is used to calculate the chemical speciation in this version of chronic BLM?
The speciation codes are the same as those used in WHAM (Tipping 1994 Computers and Geosciences 20:973). However, for a chronic BLM, other type of speciation model (e.g., Visual Minteq) can also be used.
13. What about the influence of other metals present in the waters? Most important in Scandinavia Al & Fe?
Truly dissolved Fe and Al can compete for binding sites on DOC, although due to their tendency to precipitate any effects may be limited. The conservative assumption that only 50% of DOC is “active” is likely to result in the concentration of available binding sites being overestimated (rather than underestimated) even where there are appreciable levels of these metals in true solution. Boundary conditions for Al & Fe have been set in the chronic CuBLM.
14. These models are based on species that are not representative for our waters. Do we need to develop BLMs specifically for our waterbodies?
Studies show that the models are capable of predicting toxicity to species that are endemic to specific regions, e.g., Scandinavia (Deleebeeck et al 2007) Comparison of nickel toxicity to cladocerans in soft versus hard surface waters (Aquatic Toxicology 84:223.).
15. Not all DOC is created equal. We have special DOC in our waterbodies that is not considered in the BLM development.
Studies show that the models are capable of predicting metal toxicity in wide ranges of natural waters that exhibit ranges of DOC types.
16. Are the chronic and acute BLMs interchangeable?
We would strongly recommend that considerable caution is exercised when trying to make comparisons between different BLMs, even when they are for the same potentially toxic metal. This is especially important in the case of comparisons between acute and chronic BLMs, and particularly so in the case of copper due to the fact that whilst there is a protective effect of calcium on acute copper toxicity there is no protective effect of calcium on chronic copper toxicity, as your expert review of the copper BLM will no doubt have revealed. Your expert review will also no doubt have revealed that in addition to Cu2+, CuOH+ and CuCO3 are also included as potentially toxic copper species in the chronic copper BLM, which is a further difference between the acute and chronic models.
17. According to Van Genderen et al. (2005) and Sciera (2004) using the acute copper BLM on waters with lower hardness then 50mg CaCO3 may underestimate the toxicity of copper, in this case to larval fathead minnow. Is this addressed in the current version of the chronic BLM?
Calcium does not have a competitive effect on chronic copper toxicity, as can be readily observed by performing calculations in which Ca is varied but all other conditions remain constant with the chronic Cu BLM.
18. Studies have shown that BLM is systematically giving discrepancies in calculating reliable toxicity data for soft waters (a factor of 8) (Schamphelaere & Jansen,2004; Boeckman & Bidwell, 2006)
Boeckman & Bidwell 2006 - The effects of temperature, suspended solids and organic carbon on copper toxicity to 2 aquatic invertebrates - Water Air and Soil Pollution 171: 185. This study refers to acute tests, so uses a different BLM to the chronic Cu model, although the abstract states that LC50 values based on total copper concentrations were significantly greater than free ion LC50s for both species, suggesting that the BLM principles still apply.
19. The underpinning requirement of the BLMs is that the system is at equilibrium condition…but this never happens in nature!
Many natural systems exist in, or close to, a steady state pseudo-equilibrium, and in the vast majority of cases the assumption that the waterbody is close to equilibrium will be appropriate for bioavailability calculations.
20. What metals have BLMs and user-friendly tools?
There are BLMs and user-friendly tools for Cu, Ni, Mn and Zn. Under development are BLMs for, Co, Pb, Al and Fe. The BLM concept would not be appropriate for metalloids, such as Hg, because of the importance of volatility of inorganic forms and the key exposure route being an organic Hg form.
21. If a BLM is available for lead should I use this or the DOC correction to evaluate compliance with the WfD EQSbioavailable
The latest WfD EQSbioavailable was derived before the lead BLM was available. Therefore for compliance purposes an availability correction based on dissolved organic carbon should be employed (see section 3.3 of this guidance)
22. The User-friendly tools are not the BLM.
The results of the user friendly models have been validated against the full BLMs and also use the same datasets.
23. The user-friendly tools show predictions that are mostly below the 1:1 line when plotted against the full-BLMs
The predictions are all within a factor of 2. For some metals, such as Cu, the exclusion of some parameters (especially Na) that may have a protective effect means that the predictions are overprotective. However, within a tiered risk based framework this is acceptable.
24. It is not possible to do a compliance assessment using these methods.
Compliance assessments using User-friendly BLMs have been undertaken in a number of countries including The Netherlands and the UK. If a generic bioavailability based EQS is in place then compliance assessments using User-friendly BLMs need not be any more complex than any other calculated parameter.
25. The pH/DOC/Ca do not cover the sensitive waters I’m interested in.
The applicable range of water chemistry conditions which the BLMs can be applied to is limited by the ability of standard test species to survive and reproduce under extreme conditions of pH and Ca. Typically low pH (,6) and low Ca concentrations (<5 mg l-1) cannot be tolerated by standard test species (such as Daphnia magna), although the limited available testing on soft water species does not suggest that they are likely to be more sensitive than other sensitive species which have been tested. Furthermore, studies show that the models are able to predict toxicity to organisms from soft waters.
26. Is it possible to extrapolate beyond the boundaries of the models?
Following further testing, the datasets for Cu, Ni and Zn have validated boundaries for the BLMs that have been modified (see below). Options for what to do outside the boundaries are discussed further in Section 6.1.
27. The upper Ca range on the model for Ni is only 88 mg L-1 we have waters that have much greater Ca. What is the implication of having a high Ca exceedance?
At Ca concentrations above this there are no additional protective effects (see above). So, while Ca concentrations may be higher in the waters the positive influence of ecotoxicity mitigation is limited.
28. How do I summarize the input data for the tool for the calculation of annual average compliance?
The ideal situation would be to have matched dissolved metal data and Ca, DOC and pH for every site for every sampling occasion. However, the difference between doing this and using annual average pH and Ca and an annual median for DOC is very limited. A median DOC value should be used as DOC may be more variable in waters and using the median is a more appropriate statistic. It is probably possible in most cases to not need to measure values every time, for example hardness or Ca. But, for DOC it is difficult to tell immediately and generally needs a few years of monitoring to get default concentration for a waterbody, although some waterbodies this would not be appropriate. For very variable aquatic systems matched data should always be used in assessing annual average compliance.
References
Peters A, Crane M, Adams W. 2011 Effects of iron on benthic macroinvertebrate communities in the field. Bulletin of Environmental Contamination and Toxicology 86:591-595.
Tipping E. Corbishley HT. Koprivnjak JF. Lapworth DJ. Miller MP. Vincent CD. Hamilton-Taylor J. 2009. Quantification of natural DOM from UV Absorption at Two Wavelengths. Environmental Chemistry. Rapid Communication. October.
