Basic reporting

I found the paper very well presented and clearly written, with only a few minor problems with organisation in the Materials/Methods and Results/Discussion sections. Specifically, I feel that the statements of line 120-122 (Ruckstuhl & Kokko 2002) would be better positioned around line 80 where behavioural dichotomy was first mentioned.

Line 43: "(Šárová et al. 2007)" not hyperlinked with bibliography
Line 99: "even." should be replaced with "event"
Line 123: Please be more specific for the statement "considering the entire dataset..." - perhaps remind the reader that this pertains to all separate scores from focal, two nearest-neighbour and control animals.

Experimental design

The selection of the focal animal for each scored observation was well conceived and explained, however I am left with some doubt as to the selection of the "control" animal. I am left with the impression that there may have been unconscious bias in its selection - perhaps the control was selected as it was most visible, thus being more peripheral to the herd, thus (conceivably) engaged in more vigilant/active sets of behaviours. It is not probably possible to gauge the centrality in the herd group of the focal/control animal as no photography or filming took place, but the lack of detail in the selection of the control is the only aspect that weakens the presented argument - but this is not a fatal flaw.
The control animal was also labelled "random" in Figure 1 - perhaps one label be consistently used to avoid confusion in the reader.
Some detail was omitted - how many discarded antlered males were in the herd group, to provide a more precise number of females/young in the observation group rather than "c. 99" (Line 63).
How often was the herd disturbed - there is mention that some 20-min observation periods were discarded due to disruption by humans (lines 88-89), but it would be more helpful to know what percentage of observation periods were aborted.

Validity of the findings

Validity of the findings is good such that the methods can be replicated by others and collective behaviour models further tested with photography/filming/proximity analyses - as the authors discuss.

Comments for the author

The points and queries I raised in the previous comments are provided only for the author's guidance should they wish to further improve the paper. I am satisfied that the paper could be otherwise accepted as presented.