Review of RSPB-2022-2532
Sensory integration of danger and safety cues explains the fear of a quiet coyote

This study used an in-field experiment to investigate how auditory and olfactory cues of a predator, when presented as solitary or combined, affected the behaviour of a mammalian prey species. I really liked the study itself, and I think the results were interesting and are definitely publishable. I had two main concerns. The first is that I think the introduction needs expanding to include context for all the hypotheses being explored. The second is that there are some details missing from the methods that could influence the interpretation of results. Neither of these I think preclude the study from being published, but I would suggest revisions as per below. I should also say that the manuscript was overall well written, I thought, so it was a nice read!

Line 15 – “Alternative options” is too vague. Please clarify?

Line 16 – 22 – This sentence has too much in it, making it very confusing. Can you break it up with respect to the studies etc, or restructure so that you provide examples after you list the consequences of non-redundant signals please?

Line 59 – By dog, I assume you mean coyote? Or do you mean any canids (including wolves, domestic dogs etc). Could you either be consistent (i.e. coyote) or specify wrt wording please, including (if you mean domestic dogs) latin names for any other species being considered in this context.

Line 64 – Your second aim comes out of the blue a little bit for me. You haven’t talked at all in the introduction about (non-predator or conspecific) environmental effects on information processing, decision making or anti-predator behaviour/vigilance etc. “Pristine” and “disturbed” sites can mean many things, and although I know you go on to describe what these mean, they are still being introduced for the first time at the point where we should already have more of an understanding of context in which to consider your aims and hypothesis. How comparable is domestic dog urine to coyote urine? Do domestic dogs (or non-domestic, free ranging dogs) chase, disturb or hunt ground squirrels? This would counteract the habituation hypothesis. There just needs to be more background leading up to this particular aspect of the study.

Line 70 – Again, you have not introduced the aspect of how age class or sex influences responses to predator cues prior in the introduction, so we have no context to place expectations or previous findings. Why would breeding season matter? I know you give references, but you can’t really use references to replace information that the reader needs as context.

Line 116 – I assume this means you re-caught individuals and gave them a different test each capture (except upon first capture?) Or did you keep individuals that you captured in a holding area for a few days? How long did you let the individuals acclimate to the arena (or did you?).

Line 120 – Few details missing for the acoustic stimuli: Was the speaker (box, odor etc) inside the arena or outside – I know you said it was in a box but was that box an object inside the arena? ? How loud were the calls (i.e. a dB level at a certain distance away), how did you decide on this level (i.e. knowing how loud coyote calls are at certain distances, and what those distances might mean to a squirrel)? Do coyotes have different calls in different regions – did you use local calls (it may not matter, but it’s good to know just in case). I also just wonder because if the calls sound close, but there is no visual confirmation, then it may not be a robust stimulus. More information here would clarify this. Also, when did you play the calls?

Line 130 – Were the videos analysed muted so that the acoustic stimuli timings were unknown, and/or otherwise de-identified so that the overall stimulus test category was unknown (i.e. blind scoring to stimulus/test)?

Line 131 – Would this result in an overall distance moved? What would that tell us? Or is is also about orientation? Just a bit more detail here (again, I know you have a reference, but I personally like to read papers without necessarily having to check other references for small details like this over and over again!).

Line 143 – Did each video only get one score? Or could one video/test have a suite of scores over time?

Paragraph at line 217 – How do these predictions fit with your actual study? You didn’t really study spatial or temporal scales so I’m not exactly sure how this paragraph informs us regarding your results or their interpretation/contextualisation. (i.e. is urine not that informative? But then why would the squirrels be more reactive to that stimulus?)

Line 229 – See my comment above about sound cues without visual confirmation – it may signal safety but it also might sound distant? OR if it sounds close you would expect to have visual confirmation. So the calls coming from a speaker might just be confusing/completely non-informative.

Line 272 – Alternatively (not mutually exclusive, but could be an additional component) there could be a selection component coming across here. Juveniles that tend not to react to predators get eaten, thus leaving a subset of adults that are highly reactive (there would be more variability in juveniles as well in this case).

Figure 4 – I like this figure, but why does S1 appear to have 5 individual datapoints? Similarly, it’s hard to tell if S11 and S12 have 3 and 6 datapoints respectively, or if one of the lower-positioned points belong to S11 (in which case it doesn’t overlap with the box, I know, this might be really annoying!). Even if this is the case, S12 appears to still have too many datapoints?