This is a very relevant study to all field ecologists, although the conclusions are quite scary given how we normally design our field surveys... definitively this is something we need to think more about! I have really enjoyed reading the manuscript. It is nicely presented, although I think some bits would benefit from some clarification, mostly the parts relating to the description of the study survey and definitions of study units (see my detailed comments below). In all, this is a timely contribution that calls for rigour when designing field experiments, which is especially important when trying to draw conclusions from studies conducted at different sites. I outline a few comments below that I hope will be helpful to the authors.
isabel c barrio
Abstract L8. Applied to do what?
L16: “by great complexity”?
L28: maybe could add more details here – why is it particularly relevant to community ecology?
L74: I find this a bit confusing – but maybe it is just a matter of wording. Your research question is if both approaches to define sampling units lead to the same conclusions when measuring plant community characteristics in mesic and snowbed habitats. Then, you defined sampling units by one or other approach, and measured different parameters of those communities using the same methods. Right?
L121: maybe it would be helpful if in figure 1 you could somehow show this hierarchy, maybe adding labels for the terms you use in the text: sampling region > herding districts > landscape areas > sampling units > habitats (> transects > plots).
L126: what do you mean exactly by ‘landscape areas’? each of those 2x2 km pixels? ‘Sampling units’ are then the 25x25 m pixels within the landscape areas, right?
L138: but the two approaches (formal and subjective) were applied to define the habitats, not the sampling units (that were defined according to the criteria presented in the previous paragraph)…? Also, habitats are defined within each sampling unit – how do you use topographic criteria in GIS if your pixel size is 25x25 m? (actually what you state in L145).
L149: should it be ‘habitat patch’ instead of ‘sampling unit’ here? Did the two habitats have to occur within the same ‘sampling unit’?
L155: was this ‘expert knowledge’ always come from the same observer(s)? When were the assessments conducted? (would this affect selection if plants were in different phenological state?)
L171: how did you account for different sampling effort?