Experimental design and statistical analysis in aquatic live animal radiotracing studies: A systematic review

Live animal gamma radioisotope tracing enables the monitoring of real-time contaminant uptake and retention in individual animals. It has been employed by ecotoxicologists to understand how animals respond to a variety of aquatic stressors. The use of the technique increases the complexity of the experimental design, resulting in the production of highly detailed and robust longitudinal data of individual animals. The greater complexity of the statistical models that underpin this data create risks from data being treated incorrectly. A systematic review of the literature was undertaken to comprehensively evaluate the experimental design and statistical approaches employed by current aquatic live animal radiotracing studies. The experimental design of current studies was categorized into one of five design "families". Each experimental design could be described by at least one statistical model, which in turn informs a specific statistical approach. Collectively, 7% of studies provided insufficient information to determine the experimental design used, and 24% of studies undertook no formal statistical analysis of results. When the statistical approaches used in current studies were examined, in most cases the correct approach was undertaken, with only 18% of studies using a statistical approach that did not match the chosen experimental design. This review provides clear guidance on the analysis and interpretation of data collected from each experimental design and explains the advantages and disadvantages of each. Additionally, this paper identifies four key suggestions for future researchers, and provides commentary on the ethical, logistical and cost considerations associated with undertaking experiments that use live animal radiotracing.