1.10Issues Arising 1.10.1Limit of Detection (LOD) values
In many of the datasets that were examined the activity concentrations were often reported as less than the limit of detection (LOD). Substituting LOD data with a value to enable inclusion into a larger dataset has been applied in a number of reviews and the approach adopted varies. One approach that has been utilised in the development of some databases on radionuclide transfer to wildlife is LOD/2 substitution that is applied to datasets where no more than 20% of the values are below the LOD (Beresford et al., 2011; Hosseini et al., 2008). Although this approach can be justified on the grounds of simplicity or pragmatism it is best if it is only applied when it has a negligible effect on the end result (Wood et al., 2010). It should also be noted that there is little, if any, statistical rationale behind the substitution method and the approach performs worst in situations where there are multiple detection limits (Helsel, 2005). This is a situation often encountered in radioecological datasets and especially those obtained from gamma spectrometry.
Wood et al. (2010), when assembling data on radionuclide transfer to reptiles, applied survival analysis techniques to handle datasets that included
Johansen & Twining, (2010) included ‘less than minimum detectable activity’ (MDA) values (in this case MDA’s are meant as the same as LOD) only when the MDA’s were less than the average value of the dataset with the intent being that the
ERISS has chosen to only include data that provides a measurement that is equal to or greater than the LOD, i.e. it is an actual measurement of the activity contained within a sample. The approach used by ERISS has also been adopted for the purposes of this work and any CR values calculated are based on data that was reported as being equal to or greater than the limit of detection. All data that was reported as
Given that some of the source data that was considered during this review was the same as that analysed by Johansen & Twining (2010) and Wood et al. (2010) there may be some variation in both the sample sizes (n values) and also the resultant concentration ratios. That noted, the variation in CR values is generally within the same magnitude.
It is recommended that these variations be reconciled in the future and discussed more widely within the Australian radioecology community to ensure a consistent approach is taken when
1.10.2Terrestrial plants from arid and desert regions
When initially applying the data manipulation process to terrestrial plants from arid/desert regions of Australia the data was mostly reported as a dry weight activity concentration. This required the use of dry:fresh weight ratios to convert to a fresh weight activity concentration. In some of the earlier reports reviewed there were no dry:fresh weight ratios reported, so the assumed dry:fresh weight ratio of 0.1 for shrub (other parts) reported in Beresford et al. (2008b) was utilised in the absence of anything more appropriate. Subsequently, when reviewing later datasets dry:fresh weight ratios were reported for similar species from arid/desert regions of Australia. These ratios were consistently and substantially different (average dry:fresh weight ratios in shrub/grass foliage of 0.6) to the assumed values reported in Beresford et al. (2008b). The decision was therefore made to recalculate the earlier plant data and apply the dry:fresh weight ratio of 0.6 to estimate the fresh weight activity concentrations in these plants.
Noting this difference in the dry:fresh weight ratios for the plants from arid/desert regions, and that many of these plants exhibit xerophytic or halophytic characteristics to assist them to deal with the arid, sometimes saline and often extreme heat conditions, it is recommended that development of a set of Australian specific ash:fresh and dry:fresh weight ratios, particularly for terrestrial plants from arid and desert regions of Australia be considered for future work.
In the absence of any appropriate dry:fresh weight ratios being identified during the data review some CRs have only been calculated on a dry weight basis. For example, activity concentration data was identified in a number of reports for leaves of trees from the Eucalyptus and Melaleuca species. It is recommended that further research is undertaken to confirm an appropriate dry:fresh weight ratio for these leaves prior to conversion of this data to a fresh weight concentration.
1.10.3Duplication of data and access to data
During the review of the Australian data that had been submitted to the WTD there was one occurrence of data being included twice (from different sources) and one occurrence of data from a summary journal publication being included, that would have had a different CR for polonium and N (number of samples) value had the data been drawn from the original research report (n=1 from the summary journal publication; n=7 from the original research report). Whilst these occurrences are ultimately not avoidable and once identified can be rectified, it is recommended that a more coordinated approach is taken when collating and submitting future Australian non-human biota concentration to the WTD. This initiative could potentially be led by ARPANSA in cooperation with ERISS and ANSTO, and possibly others.
A strict process for quality assurance and review of data is important. An essential key to these processes is having access to original research/analysis reports. The original source data or reports often provide more detailed information on sampling, sample size, analysis techniques and raw data that may not be available in more easily accessible formats such as journal publications or publically available environmental impact statements (EIS). Some journal publications and particularly EIS documents present data in a more summarised format, and in some cases this lacks information such as whether activity concentrations are on a dry weight or wet weight basis. Establishing a cooperative working group across a range of organisations within Australia would enhance the quality and consistency in the data collated on Australian non-human biota.
The use of appropriate conversion factors was discussed above in regard to dry:fresh and ash:fresh weight ratios. Similarly, in some instances there are no conversion factors available for some data manipulations. For example, tissue concentration ratios for thorium have been calculated for Australian mammals. However, suitable whole-organism to tissue CRs such as those presented in Yankovich et al. (2010) have not yet been established to convert tissue CRs to whole-organism CRs for thorium in mammals. Whilst less relevant to the uranium mining industry, this may be one data gap that is relevant to other industries (i.e. mineral sands) and by examining the Australian datasets in more detail suitable thorium conversion factors could be developed.
1.10.5Data gaps, Maralinga, and other environments 1.10.5.1Data gaps
The tables below summarise the broad IAEA wildlife groups for terrestrial (Table 2) and freshwater (Table 3) habitats relevant to uranium mining for which data is available for Australian non-human biota. These tables identify wildlife groups for which whole-organism CRs for Australian biota are available and the groups for which additional information was confirmed during this review. For terrestrial habitats (predominantly arid/desert mining areas) there is an increase in data from two (mammal and reptile) to six wildlife groups with new CRs available for bird, grasses and herbs, shrubs and trees. For freshwater habitats (predominantly tropical mining areas) there has been an increase from five to six groups. Additional data have been identified for most of the five groups, and limited new data identified for algae. Once published, this data will benefit the Australian uranium mining industry by consolidating datasets, providing a comparison to the international values, and will assist in supporting more robust radiological assessments, particularly in the screening phase of assessments.
These two tables provide a snapshot of potential data gaps. Whether or not data is required for additional wildlife groups is something that should be considered in a strategic manner or on a case by case basis, given the variation in species that can be seen in the wide range of Australian environments. Additionally, ongoing involvement in international research programs (such as the IAEA program MODARIA: Modelling and Data for Radiological Impact Assessments, that will run from 2012-2015) will be important to ensure a continuing involvement and understanding of the evolution of international databases such as the WTD. Other information may also be available for additional wildlife groups by examining datasets for stable isotope measurements that may have been captured within the ecotoxicology research community. These alternative data sources may provide information that can be used to estimate CRs for some radionuclides.
1.10.5.2Maralinga
Maralinga is located in central South Australia. Between 1956 and 1963 the Maralinga lands were used by the British to conduct nuclear weapons testing, including both critical and non-critical nuclear test trials. From a climatic perspective Maralinga is in an arid desert area which is a common environment to many uranium deposits in Australia. However, in relation to Maralinga data, the isotopes reported were the anthropogenic radionuclides Cs 137, Am 241 and Pu-239/240 which are not relevant to the uranium mining and milling industry. For the purposes of this review the Maralinga data was not included or reviewed further.
1.10.5.3Other environments
There are a wide range of practices that release radioactivity to the environment. Whilst this review only considered data for non-human biota inhabiting uranium mining and milling environments it is recommended that any future publication of Australian CRs, including those from this project, be incorporated with all available data for Australian environments.
Table 2: IAEA Broad Terrestrial Wildlife groups and relationship to ICRP RAPs and Australian data for uranium mining areas
IAEA Broad Terrestrial Wildlife Groups
|
Potential appropriate ICRP RAP
|
Australian data (Yes/No)
|
Currently in WTD
|
Identified during project
|
Amphibian
|
Frog
|
No
|
No
|
Arachnid
|
–
|
No
|
No
|
Arthropod
|
Bee
|
No
|
No
|
Bird
|
Duck
|
No
|
Yes
|
Annelid
|
Earthworm
|
No
|
No
|
Fern
|
–
|
No
|
No
|
Fungi
|
-
|
No
|
No
|
Grasses and herbs
|
Wild Grass
|
No
|
Yes
|
Lichens and Bryophytes
|
–
|
No
|
No
|
Mammal
|
Rat or Deer
|
Yes
|
Yes
|
Mollusc - gastropod
|
–
|
No
|
No
|
Reptile
|
–
|
Yes
|
Yes
|
Shrub
|
–
|
No
|
Yes
|
Tree
|
Pine Tree
|
No
|
Yes
|
Table 3: IAEA Broad Freshwater Wildlife groups and relationship to ICRP RAPs and Australian data for uranium mining areas
IAEA Broad Freshwater Wildlife Groups
|
Potential appropriate ICRP RAP
|
Australian data (Yes / No)
|
Currently in WTD
|
Identified during project
|
Algae
|
–
|
No
|
Yes
|
Amphibian
|
Frog
|
No
|
No
|
Bird
|
Duck
|
No
|
No
|
Crustacean
|
–
|
Yes
|
No
|
Fish
|
Salmonid
|
Yes
|
Yes
|
Insect
|
–
|
No
|
No
|
Insect larvae
|
–
|
No
|
No
|
Mammal
|
–
|
No
|
No
|
Mollusc
|
–
|
Yes
|
Yes
|
Phytoplankton
|
–
|
No
|
No
|
Reptile
|
–
|
Yes
|
Yes
|
Vascular Plant
|
Wild Grass
|
Yes
|
Yes
|
Zooplankton
|
–
|
No
|
No
| |
Yüklə 0.94 Mb.