R2 comments

[beckybanbury]

Comments to address:

• [x] L166. The definition of precipitation variability is unclear.

• [ ] In the method, maybe indicate the method to calculate PET. PET is an indicator quite dependent on available energy, I am not sure how the authors distinguish the PET influence and radiation influence on carbon fluxes.

• [x] Is it possible to add the number of samples (n), R2 and p value for each regression line? That might be convenient for readers to interpret.

• [x] L276 “little variation” is a touchy qualitative. How little is little?

• [x] L303. Are “temperature seasonality = 0” and “at an annual temperature range of 15 or lower” two different conditions? I am a bit confused by the statement.

• [x] L410-432. I see this as a major concern. The discussion is necessary definitely, though I found the logic presented in this paragraph is a hard to follow. First the authors seem to suggest C fluxes are very related to growing season length in Line 412 and 413. But in L419 and 420, they say that “no relationship” between growing season climate and C fluxes. In between L417 and L418, I do not understand why positive relationships between C fluxes and T, PET and Rad mean no adaption? (why is it relevant to the discussion may also need clarification). L425 mentions “partition this variation”, while I am not sure what is needed to be partitioned and what the variation refers to? L432 at the end the “seasonality” is suddenly thrown in, but I think the discussion is all about growing season length and peak values. Again, I found the whole paper is quite a pleasant reading, just this paragraph makes me scratch my head ;(

[beckybanbury]

@teixeirak what do you think about adding n, rsq and p values to the plots? For me, I'm a little worried that this will clutter the plots as it's 3 - 4 lines of additional information per plot if we do that.

N samples is the same across the majority of analyses + is presented in the summary table, so I think we could leave that out. I don't know how much information p values will add - the plots already show significant/non-significant relationships by dotted or solid lines (perhaps I could make that clearer in the captions) + I don't know what p values would add on top of that. I could perhaps add rsq (I think originally we did do this + took out rsq values) , or we could state rsq in the captions.

[teixeirak]

Yeah, I'm not sure how you this would even be possible. I'd just add a sentence to the legend referring the reader to appropriate tables for this info, and noting that dashed/solid lines indicate significance.

Keep in mind that the paper is officially accepted, not contingent upon changes. We just need to make the changes that we consider improvements.

[beckybanbury]

Thanks for clarifying!

I think I'm finished with these comments then - just made some minor changes. I've pushed the edits but I can't knit for some reason - are you able to knit please?

I'll have a final read through, but pretty much there with this - just the repo to sort!

[teixeirak]

Wonderful; thanks! I will look at it soon, and try to knit. I'm using a different laptop right now, though, as mine had to go in for repair.

[teixeirak]

Do we currently site the ERL review? It's on the verge of being accepted, so I think it's safe to add it. It would be nice to point readers to that.