Overall –
This is an interesting paper that does indeed address a persistent conundrum in animal behavior studies: Is observed correlations between a behavior and animal condition a cause or consequence of the behavior. Thus, the framework and methods presented are interesting and relevant to many. The manuscript is generally well written, and the hypotheses are stated clearly (too often though). The authors conclude that H1 is supported (association with DFADs causes tuna to have poor condition), but I am left wondering how much support the authors actually demonstrate. The empirical data in Fig 3 is weak (as the authors acknowledge) and the model results that the author’s rely on are buried in the supplemental material. The methods and results presented in S1 are central to the conclusions drawn by the authors. I therefore think this work should be included in the main text, particularly Table S3. If an expert statistical reviewer supports the methods and results (which I am unable to fully evaluate), I do think this paper will be of interest to many of scientists and the methods developed applied broadly. Below are my detailed comments and expanded feedback that I hope improve the manuscript.

Redundancy throughout the manuscript. I have identified some specific areas.

Overall, the writing is clear and the work easy to understand. There are a few places where the grammar is not correct, and I have tried to flag, but a native English speaker should review at some point.

Over use of hyphens “-“ throughout the manuscript. Sometimes made sentence hard to follow. I suggest reducing the use of hyphens, and relying more on colons and commas.

Acronym use was confusing. Many are undefined. Please clarify each term and how they are different from each other (DFADs, FOB, LOG, NLOG).

The term “Phylogenetic consequence” is used often in the manuscript, but I am left wondering what the authors mean by this exactly. Please provide further description or examples in the Introduction.

Title –
Cute, but not sure it is descriptive enough of the manuscript. No mention of tuna or the intent of assessing mechanistic relationships.

Keywords:
I think purse-seine can be removed as this is not a focal point of the study, and the words “FADs” and “mechanistic models” should be added.

Abstract
Can “physiological condition” be described further? This is very broad. After reading the manuscript I believe this phrase refers to nutritional state. I think that in the Abstract and Introduction, the link between the term “condition” and nutritional state should be defined/clarified.

The phrase “demonstrates the relevance” is vague. I suggest the authors improve the emphasis of this study more here. Relevance to whom? How can these methods be applied?

I find the last sentence of the Abstract to be redundant and think that the authors could use the limited space better by adding more information on the work, and more specifics.

Introduction

The Introduction is good and justifies the study well, with clear hypotheses to be evaluated. I do believe that a bit more detail and clarity is needed though, particularly around acronyms and what “physiological condition” represents for a tuna. Please see specific comments below.

Line 22 – “scale” should be plural: “scales”

Line 23 – add year to Hogan reference: Hogan 2015

Line 25-26 – I find this sentence from “focusing” to “considered” awkward. I suggest re-writing this statement.

Line 30-32. This last sentence of the opening paragraph can be improved. Why is determining the function difficult? What are you referring to at the end of the sentence when “without it” is stated? I suggest adding examples here so that readers can connect your statement to the need for your study.

Line 34 – Should this say “observational science” instead of “experimental science”? Experiments can be controlled to better address this question, as in your example below of the work by Armstrong.

Line 38 – Change “A lot of” to “Many”.

Line 41 – change this to “(e.g., controls, replication)”.

Line 47 – what does “NLOGs” stand for?

Line 55 – Strandings of what? Fish or gear?

Line 57 – change this to read “has increased the purse seine fleet efficiency by increasing the time….”.

Line 60 – It is important to explain what aspect of tuna life history/health this associative behavior can impact. Do you believe this behavior could impact population dynamics, vital rates, survival…? Please explain.

Line 69 – What is FOB? Please define.

Line 76 – “increase their condition recovery afterwards”. After what? This is unclear and should be explained so readers can understand the mechanism you are testing.

Line 88 – It is unclear in this sentence what data is used to test the hypotheses. Where does the data come from on the increased DFAD numbers? Or is this a theoretical test? The next sentence provides clarity, but these two sentences should be merged so it is clear where the evaluation data is coming from.

Material and Methods

Line 103 – Make it clear that phase 3 stems from phase 2. State “If individuals experience starvation in Phase II for too long,…”

Line 112 – Are these terms written correctly? I don’t see them in the equations. nμp, nμm

Line 130 – Is the Stationary model represented as Fig 1A? This is called the “General model” in the caption. I am confused about what this stationary model represents. Is this no change in condition? Please explain what this model represents.

Line 142 – What is the range of phase angle values? And do tuna in bad condition have low values and tuna in good condition have high values? It is important to explain the significance of this metric to readers not familiar with it.

Line 148 – I am guessing that the “IOTC dataset” that is mentioned here is derived from the “3-BU” form previously mentioned. This should be clarified and described fully.

Line 151 – I see here FOB is defined. This needs to happen at first use of the term in the manuscript.

Line 155 – What does “LOG” stand for and represent?

Line 155-157 – I found this very confusing. I could not follow what the terms are representing. Likely because LOG has not been defined.

Line 158 – Why is the scale increased to 2 degrees when the raw data was 1X1?

S1 Supplemental – Non-linear regressions:

The methods and results in these models appear to be central to the conclusions drawn by the authors. I therefore think this work should be included in the main text, particularly Table S3.

Line 469 – please define quarters.

Line 475-476 – I am not a statistician, but the role of β, δ, ζ appear to be central to the model and results (“used to fit the density of FOBs”), especially since when ζ is removed from the model, the YFT2 model allow the authors to “reject H2”. Can these terms be explained further about they represent?

Line 479 – “points” should be “point”.

Line 487 – “which had a smaller than the first model”. Smaller than what?

Table S3 caption – Please explain what the difference between YFT1 and YFT2 is. And the grammar here should be revised: “the last columns present…”

Results:

Lines 193-197 – Good summary.

Line 202 – “a significant decreasing trend” – in what? PA?

Lines 200-205 – I believe the authors are explaining here that although we don’t see a relationship in the data plotted in Fig. 3, the statistical analyses (non-linear models) show a relationship. If I am understanding this correctly, the work presented in S1 must be included in the main text.

Additionally, I believe the authors are trying to make the point that the data plotted in Fig 3 looks like the curve plotted in Fig 2 for H1, and thus H1 is most likely. This point should be made more clearly. Furthermore, the curve looks like a negative exponential distribution. Can the relationship be assessed with this fit?

Discussion:

Lines 207-214: While this is a good summary of the work, it is highly redundant of the material presented in the Introduction.

Line 217-218: This presentation of the hypotheses is repetitive.

Line 224-239: This is a good description of the study caveats and needed future work, but it should go after the description of the study results are presented.

Line 242-243: This presentation of the hypotheses is repetitive.

Line 253 – Unless I missed it, this is the first time 2 species are mentioned. Work on skipjack needs to be clarified earlier.

Lines 257-260: This is a good summary of results, but I am left wondering how confident the findings are, and thus how confident the authors should be in their statements and conclusions. The data shown in Fig 3 are less than convincing (not much of a trend, as the authors themselves state), and the results the authors rely on to support their conclusions are hidden in S1. I believe S1 should be moved to the main text. There are many places in the manuscript that are redundant, so with some editing, there will be room for the added material from S1.

Line 269-270: This presentation of the hypotheses is repetitive.

Line 272-274: Good point.

Line 277-278: Is it a direct 1:1 relationship? Looks more like a negative exponential distribution?

Line 284: Should FADs be DFAD? Terms are confusing and hard to differentiate.

Line 291 – “This study determines an ontogenetic consequence of yellowfin tuna associative behavior.” Does it? I think I need more convincing, although I do like the discussion in the rest of this paragraph.

Line 308 – remove second “increase”.

Line 312 – change to “allows us to”.

Line 315-318 – I find this closing statement repetitive again. If other repetitive statements are removed, this closing point would be ok here.