Dear Dr Keidel,

Many thanks for submitting the revised manuscript. The two referees were supportive of your work, and you have managed to address the points they raised appropriately. However, I think that more changes are required for a coherent and more concise presentation of your results. Please address my own points below for further revisions before I can accept the manuscript for publication in Biogeosciences.

With best regards,
Jens-Arne Subke


Lines 27/28: The connection between added CO2 and target year for scenario is confusing. At current concentrations, 20% are about 80 ppm, and with an increase of about 2 ppm per year, would not be reached for around 40 years. Please re-phrase, possibly by omitting a year for the scenario, to avoid confusion.

Lines 150 to 155: A grammatical detail: There is no need to phrase hypotheses as predictions for the future – they are statements about conditions irrespective of when they are measured. So I suggest re-phrasing as e.g. “long-term (>10 years) moderate CO2 enrichment causes increased soil respiration”, etc.

Lines 280 and following: The application of a temperature response model comes a little surprising, as this is not in your list of objectives. Please clarify that this was done to fill gaps in the measured data set at the beginning of the section. Also make it clearer what fraction of data on which annual and seasonal estimates are based was from measured or modeled data. (Possibly include an additional table).

Lines 291 to 293: You should clarify whether the data you used to fir the model were independent from the data you used to validate your model against. This is critical, as a lack of independence means that a slope of 1 would be expected, and does not hold any information with regards to the quality of the model.

Line 293: Full-stop needed after “1”.

304: Please clarify how means were calculated. Presumably you calculated annual sums per ring and averaged between rings as true reps (n = 3)? Otherwise repeated measures would have to be used.

The discussion is lengthy and draws on several aspects not directly covered by your research. This is obviously appropriate as far as interpretation is concerned, but it should be shorter. The discussion of N availability and frost effect allows only very indirect inferences, for example and moisture effects have not been investigated systematically, and should in my view be addressed more briefly.

The discussion of temperature sensitivity is in my opinion not well suited to the paper. There are a number of issues regarding changes in sources of CO2 through the seasons, reference temperature depth etc., which would have to be anaylsed in much more detail to warrant this discussion. You did not introduce the issue of temperature sensitivity and have no hypothesis that addresses the issue. In the interest of the coherence of the paper, I recommend that you omit the discussion of the temperature effects completely. Where the use of the temperature function for gap filling purposes is concerned, this is appropriate to include.

Lines 558 to 563: The sentence is awkward to rad. Please simplify, e.g. to: “Previous results from the GiFACE site show that in periods when soil moisture in the main rooting zone was low (0.3 m3 m-3), soil continued to produce N2O from deeper soil layers (20 – 50 cm), where soil moisture remained high (c. 0.6 m3 m-3) (Müller et al., 2004).”

Units should be in standard SI units, so m3 m-3, rather than cm3 cm-3.

Table 1: Please either rexpand column headings, or explain shorthand in the legend.

Table 2 as well as in the text (e.g. lines 345-346 and 356): Please round figures to at most 4 significant digits. (error terms should be rounded to same decimal place as mean values). Explain what the p-value indicates (difference between teratments per year obtained by which test). I also suggest referring to the table in the results section.

Fig. 2: Please separate measured CO2 fluxes from modelled fluxes. I could not find any indication of the proportion of measured and modelled fluxes, and this would be an appropriate place to illustrate this. It may be best to use separate panels rather than changes in colour or symbols that may be hard to read in the figure.

Fig. 3: Explain what the error bars show. Are these the errors associated by averaging across the three years (i.e. n = 3), or are errors of individual estimates propagated?

There is a mismatch between the legends for Figure 4 and 5.

Figure 5 shows one set of data, when it should show both ambient and elevated CO2 data, I presume? The lines in the graph have to be explained in the legend.

Figure 6 shows the same data as Table 6 and should be omitted.

Dear Dr. Keidel,

I thick you ave managed to address most of my concerns, and think that the manuscript is now more concise than previously. I have two comments that I would like you to address before final acceptance:

Table 3: I assume that you have added this following my request to clarify what is modelled and what is measured. What I had meant was that it is clad in the data, where you have filled gaps with the model, and where you used the measured data. It seems to me that you now present data as measured results independently form the modelled data. That is ok, and should illustrate to the reader what you did. It is not strictly gap-filling, which spies that you use measured data whenever available and modelled data for data gaps. This should be adjusted in the text. Table 3 is not in fact needed.

For the figure showing modelled fluxes, I am surprised that you have apparent outliers. Temperature is your only driver, and the presented results show no outliers, so why are calculated fluxes different? Again, this should be clarified before publication.

Best regards,
Jens-Arne Subke

Daer Dr Keidel,

Many thanks for revising the manuscript. I am happy with all edits, and for the paper to be published in this form.

With best regards,
Jens-Arne Subke