Dear Tim Eickensheidt
One of the referees still has serious concerns on the manuscript. The main problem is the lacking NH3 loss data and how this is reflected to test the hypotheses. This must be better analyzed and discussed before the manuscript is acceptable for publication in BG. This point is stressed by the fact that there is no real uncertainty analysis of measurements and how it is reflected to the results.
Please, take seriously the points raised by the referee 2.
Looking forward to see the revised version of this interesting and important manuscript.
Tuomas Laurila
BG Editor
Referee 2 report:
The manuscript with the new title “Short-term effects of biogas digestate and cattle slurry application on greenhouse gas emissions and N availability from high organic carbon grasslands” has improved. The new title is a little bit odd, in particular ‘high organic carbon grassland’. I suggest: “Short term effects of biogas digestate and cattle slurry application on greenhouse gas emission affected by N availability from peat grasslands”. Many passages are more concise and clear in the new version. However, there are still some points which should be addressed to make it acceptable for publication. In particular, the interpretation and extrapolation of the limited NH3 loss data is still lacking.

General points:
The ammonia part is still not convincing. The authors have still not included an explanation why there was such a strong difference between volatilization dynamics of both slurries. This is most striking and in harsh contrast to previous research. This is in particular with respect to the high DM content of the cattle slurry which suggests an extended volatilization process. But the contrary is the case. The results of one measurement cannot be extrapolated for two vegetation periods. The reply of the authors is correct, that the values of the single measurements were in the range of reported values but this range is rather large and ammonia emissions can strongly vary. This in particular applies to differences between relative ammonia losses between the two organic fertilizers, which are much larger as reported in several earlier studies. And this large difference may have a strong effect on NUE. So, I suggest to use absolute emission levels and relative differences from the literature which are based on a much larger data base as this single measurement for application dates without NH3 loss measurements. Another option is to use the ALFAM (Sogaard et al. 2002) empirical model to estimate ammonia emissions for cattle slurry and to take the average increment of relative ammonia emissions from the literature. ALFAM available is under http://www.alfam.dk.

The actual grassland management should be presented in more detail. I guess that the investigated applications were the fertilization for the 2nd and 3rd cuts in the respective years. How was fertilization done for the first cut and how could that have affected the results. Also the yield level of the first cuts would be helpful to have an idea of the productivity of the sites. So, a summary of the complete grassland management over both vegetation periods should be included and be considered for discussion of the results.

Hypotheses : there exists an potential interrelationship between crop growth/NUE and N2O emissions that should be stated here or in the introduction. Otherwise both research aims seem to be unconnected. Hypotheses a) is not tested in this study as the same amounts of NH4+ were applied or were aimed for during application. So this study cannot be used to test hypothesis a) which explicitly states that NH4+ concentrations (amounts) are the cause for higher emissions. Only higher concentrations are an implausible explanation. The same applies to hypothesis c) if the same amount of slurry NH4+is applied there exists no reason for higher ammonium availability, so this study design cannot be used to test hypothesis c) and it is also in contrast to the assumption that there are higher ammonia emissions: same amount of ammonium and higher losses - lower yields (if there are no other factors)

Fig 5: Mean values of N2O emissions should be used in this analysis rather than single measurements in this regression as groundwater level and application rate characterize plot properties while N2O emissions show microsite variation within a plot.

Please rethink frequent use of “ enhanced” and “affect”

Points in detail
L 18 “has resulted”
L 19-20 “..huge amounts of nutrient rich residues, the by-products of the fermentative process, are used as organic fertilizers.”
L21 “..are increasingly cropped..”
Abstract: NUE and N2O emissions are interconnected and should therefore be investigated at the same time.
L 27: ammonia emissions are not studied extensively enough to support this statement. I suggest adding: in addition NH3 emissions ware determined in one trial to obtain first clues with respect to the effects of soil and fertilizer types….or the like
L 29: “The study was conducted at two areas within a grassland site, which differed in their SOC contents. At each area…three plots were established…”
L 32: “On each plot…”
L 34:”For NH3 measurements we used…at one application.”
L 44:”…following the splash plate…”
L48: replace ‘due to’ by “with mean groundwater level and ammonium application rate…”
L 55: “have been operated”
L56: sentence structure not clear, biogas production not necessarily reduces GHG emissions. “Heat and energy from biogas substitute fossil fuels and biogas production can reduce GHG emissions.”
L 62: delete ‘,thus’, “…representing the second…”
L 63: delete “are left over” – “generated/produced as by-products”
L 68-69: delete “additionally” –“both further favoring N2O…”
L 71: delete great interest ---“are of major concern..”
L 74: old figure and not appropriate reference, give actual figure and direct source (e.g. German GHG inventory)
L 106: give references for grassland sites
L 111: give appropriate reference for acidifiying and eutriphying effect.
L 116: Quakernack et al did not investigate emissions after splash plate application
L 132 NH3 should be excluded here as it was investigated by far less intensively. Give a short sentence, that NH3 was additionally investigated at one particular treatment.
L 144 “Study site”
L 163 “at each grassland site three adjacent plots”
L 195 what does higher accuracy imply? give figure/estimate
Not only evenness at the plot level but also amount of slurry applied at each measurement collar is of importance for the measurements!
L 204: pleae add NH4+/Ntot ratio in Tab.2, Tab 2 does not contain characteristics but mainly application rates.
L 278: this procedure for harvesting is rather imprecise, why no hand harvest or combine harvests in the plot - effect of trace gas measurement on grass growth!
L 293: this correction for NH3 losses is not viable and should be changed as proposed above!
L 356: where are presented the soil moisture values which have been monitored?
L 450: this test is not viable as yields should be obtained on a whole plot level. Cuts from the collars should be pooled to mean yields per plot. Such small harvesting areas within measurement rings are not representative for undisturbed growth conditions and plot based yields.
L 454: N balance should be revised as consideration of NH3-emissions is not appropriate.
L 477: the properties referred to are not reported on Tab 2!!
L 482: see comment on hypotheses a) and c)
L 514 ff. do not understand this argument: 1. Minerailization usually is influenced by fertilizer amount and application rate, 2. Mineralization appears always as negative in an N-balance (as lost from the system by N-uptake)
L 532 not in this study (see above)!
L 536: but why? What can be the processes: could it also be that NH4+ is adsorbed in DM of cattle slurry not in contact with the soil or immobilized during decomposition of the much higher amounts of slurry dry matter?
L 547: this argument is not clear: if digestates stimulate microbial soil carbon break down why should it not be immobilized in this process. However, results may look different after correction of the N-balance.
L 639: It cannot be explained by concentrations as ammonium application rates were almost the same.
L 642 ff: but this nitrogen is also taken up by the crop, as shown in the strong differences in the N balances. Is it probably not an increased microbial activity as such which may explain the differences? And it may probably also explain higher N2O emissions from grassland sites which in general have a very high NUE.
L 685 ff: the strong differences in NH3 loss dynamics between the two slurry types call for explanation which is still missing.