Development and evaluation of a regression-based model to predict cesium concentration ratios for freshwater fish

John E. Pinder, David J. Rowan, Joseph B. Rasmussen, Jim T. Smith, Thomas G. Hinton, F. W. Whicker

    Research output: Contribution to journalArticlepeer-review

    114 Downloads (Pure)

    Abstract

    Data from published studies and World Wide Web sources were combined to produce and test a regression model to predict Cs concentration ratios for freshwater fish species. The accuracies of predicted concentration ratios, which were computed using 1) species trophic levels obtained from random resampling of known food items and 2) K concentrations in the water for 207 fish from 44 species and 43 locations, were tested against independent observations of ratios for 57 fish from 17 species from 25 locations. Accuracy was assessed as the percent of observed to predicted ratios within factors of 2 or 3. Conservatism, expressed as the lack of under prediction, was assessed as the percent of observed to predicted ratios that were less than 2 or less than 3. The model's median observed to predicted ratio was 1.26, which was not significantly different from 1, and 50% of the ratios were between 0.73 and 1.85. The percentages of ratios within factors of 2 or 3 were 67 and 82%, respectively. The percentages of ratios that were <2 or <3 were 79 and 88%, respectively. An example for Perca fluviatilis demonstrated that increased prediction accuracy could be obtained when more detailed knowledge of diet was available to estimate trophic level.
    Original languageEnglish
    Pages (from-to)89-98
    JournalJournal of Environmental Radioactivity
    Volume134
    Early online date1 Apr 2014
    DOIs
    Publication statusPublished - 1 Aug 2014

    Fingerprint

    Dive into the research topics of 'Development and evaluation of a regression-based model to predict cesium concentration ratios for freshwater fish'. Together they form a unique fingerprint.

    Cite this