The manuscript entitled "Transcriptomic profiles of consistent risk-taking behavior across time and contexts in European sea bass" investigates the consistency in a fundamental behavioural tendency across developmental stages in an ecologically and economically relevant fish species. Furthermore, it investigates the effect of chronic stress on risk taking behaviour. Finally, the authors present candidate genes that are differentially expressed in opposing behavioural types.

While reading the manuscript for the first time, I was a bit confused about all the different aspects of the experimental design. It seemed like the authors perhaps put too many things into one experimental design. However, after the second reading I think that the variation of findings has its own value and contributes nicely to the existing personality literature.

I have several comments that I would like to see addressed by the authors before I can recommend publication in Proceedings B.

Abstract
L19 typo: supposed to be consistent
L21 typo: periods of time
L21/22 This statement is too strong. Replace “lacking” by “scarce” or similar.
L24 unclear; I wouldn’t consider stress an environment
L25 what do the authors mean with “validated”? Please be more precise.
L31 typo: personality traits come with

Introduction
L37-40 The authors use risk-taking behaviour and exploration behaviour as they were synonyms. According to the most influential personality papers (eg. Reale et al. 2007), they are not. If the authors decide to use the terms interchangeably they should state this in the beginning, mentioning also the important literature that considers both terms as independent traits.
L54-56 I think the authors should be more precise with their reasoning. How does the consistency of a behavioural trait on an inter-individual level lead to the assumption that it is a conserved trait?

Methods
Why were the fish starved prior to the behavioural assay?
L172 To decide on a group treatment (novel environment, hypoxia, stress etc.) comes with several advantages. It saves experimental time and probably reduces acute stress responses in a swarming fish (as I assume the young fish naturally school). However, the behavioural responses of the individuals tested together likely influence each other. How did the authors account for the non-independence in the experimental design?
L194 Is the number of words of the manuscript close to the allowed maximum? If not, I would recommend not to outsource core statistical models to the supplement. As a reader I find it easier to understand the statistical analyses from the main manuscript if possible.

Results
L258 typo: to exit the sheltered part

Conclusion
L415 I recommend to avoid “clearly” here. I think it is an overstatement given the results.

The authors should be generally more cautious with overstatements regarding their results and uniqueness of their study. I believe this is not necessary given that the results are interesting enough to make this study worth being published.

Best wishes to the authors.
Sincerely,
Carolin Sommer-Trembo

(my name can be included if tis isn't against the Proc B policy)