abstract
- Short-term adult fish reproductive tests are widely used to assess the toxicity of chemicals and waste streams. However, these tests often have low power to detect differences in egg production among treatments, due to high variance and small sample sizes, limiting their effectiveness for informing regulatory decisions. A protocol for a fish reproductive test using mummichog (Fundulus heteroclitus) was refined to increase statistical power. Three studies using the original protocol were compared with three studies using the refined protocol. Tank pre-selection and sample size increased the a priori power from 11.2% to 85.7%. After exposure, average power levels were 62.0%, a more than five-fold increase compared to studies that used the original protocol (power of 15.0%). There was a high level of consistency compared to the original protocol; differences >33% in female and male gonad size and egg production could be detected among treatments. This study demonstrates that a refinement process can address shortcomings in short-term adult fish reproductive protocols, creating a solid foundation for further standardization and possible regulatory use.