Evaluating summarised radionuclide concentration ratio datasets for wildlife

Area of Science:

• Radioecology
• Environmental Science
• Wildlife Toxicology

Background:

• Concentration ratios (CR(wo-media)) are crucial for radioecological models predicting radionuclide transfer from environment to wildlife.
• Current generic CR(wo-media) values lead to significant discrepancies between model predictions and real-world measurements.
• Existing wildlife transfer databases often contain summarized data, hindering the development of more precise, condition-specific values.

Purpose of the Study:

• To develop a statistically robust method for reconstructing underlying datasets from summarized data.
• To enable the calculation of condition-specific CR(wo-media) values and derive appropriate geometric summary statistics.
• To improve the accuracy and reliability of radioecological models for wildlife.

Main Methods:

• A novel statistical approach was developed to reconstruct detailed datasets from aggregated data.
• Geometric summary statistics were employed, acknowledging the lognormal distribution of CR(wo-media) values.
• The method was applied to terrestrial datasets within the Wildlife Transfer Database (WTD).

Main Results:

• The study successfully reconstructed underlying datasets and calculated condition-specific CR(wo-media) values.
• Statistically significant differences were found in CR(wo-media) values across sub-categories (e.g., by feeding strategy in mammals).
• Dataset limitations and biases within the WTD were identified as factors influencing these differences.

Conclusions:

• The developed method provides a statistically defensible way to derive condition-specific CR(wo-media) values from summarized data.
• Identified sub-category differences may enhance specific radioecological assessments, but require careful interpretation due to data limitations.
• The use of CR(wo-media) for estimating transfer should be approached with caution in assessments beyond the screening level due to inherent uncertainties.