Summary

The paper by Namekawa et al introduces a scalable behavioral approach to study odor discrimination learning in adult zebrafish. To test odor discrimination learning, this study used food reward conditioning and applied 3D quantitative analysis of fish behavioral response to odors. The results show fish learning capability to associate presence of odors with a food reward.

Strengths and Impacts

1) The study is relevant. In the introduction the authors highlighted the limitations of conventional approach for studying sensory associative learning behavior, which are overcome by their improved approach.

2) The experimental setup is designed to monitor the progress of fish associative learning behavior, with minimal or no disturbances. The authors applied computer-aid automation for odors and food deliveries, added water circulation to the experiment tank, and used several tanks for parallel testing. This allows running the protocols of acclimatization, training, and testing in the same tank throughout the experiment.

3) The study is reliable. Training and testing with two types of odor (one for CS+ and another CS-) on a fish show that fish do learn to associate the presence of an odor with a food reward. Since the authors used adult fish from both sexes and of wild type or transgenic line, the odor discrimination behavior observed is reproducible regardless of gender or genetic background. Besides, similar discrimination behavior observed in separate experiments dismisses possible technical and analytical errors. Moreover, blinding may not be necessary in the study because the process from video recording to analysis was automated.

4) 3D behavioral study is quite informative. The 3D quantitative analysis of zebrafish behavior using several parameters (e.g., relative z-position, residence in the reward zone, surface sampling events, etc.) captured behavioral details that reflect fish responses to odors. Furthermore, the composite behavioral scoring used shows a robust difference in fish responses between CS+ and CS-.

5) Using the behavioral approach, odor associative learning in the zebrafish is well-demonstrated. The behavioral results support that fish was able and quickly learned to discriminate odors during the first day. Food deliveries in 30 s after odor application was sufficient for fish to associate the presence of an odor with a food reward but delaying the food deliveries unpaired such association.

Weaknesses

1) Some methodological details in the study are missing. It would be better to have an explanation on how the authors determined sample sizes and statistical power. Besides, there is no information on randomization method used for assigning test fish. The authors reasoned that the video analysis was automated and so blinding was not performed. Nonetheless, the authors used the Wilcoxon test, which is proper for analyzing effects between two paired groups having repeatedly measured on individual samples.

2) Using only reward conditioning paradigm to establish the behavioral approach in the present study is understandable. However, adding fear conditioning paradigm would allow comparison between fish responses to a reward or punishment. This could be useful information for future study of the underlying mechanism.

3) The degree of odor discrimination or associative learning of fish might vary with different odor types or preferences. The discrimination scores between conditioning with Ala/Trp (attractive odor) and that with Ala/Cys (Cys is an aversive odor), conditioning with Ala/Cys seems to show a lower discrimination score (Figure 5). However, comparison between the odor types was not tested statistically.

Other minor issues are not found.

Overall, the study is technically sound, and the behavioral approach is robust. It can be useful for studying the neural basis of sensory associative learning in the zebrafish.