Interactive comment on “ Depolymerization and mineralization – investigating N availability by a novel 15 N tracing model ” by Louise C

In the manuscript “Deploymerization and mineralization – investigating N availability by a novel 15N tracing model” by Louise C. Andresen et al. the authors compare an analytical approach and a numerical model based on experimental 15N tracing to estimate soil N dynamics. The main N processes discussed are gross rates of protein depolymerization and NH4 mineralization. The authors also calculate microbial nitrogen uses efficiency from the results of the analytical and the numerical approach. The presented manuscript is clearly written, structured well and the results are presented in a meaningful manner. The approach to compare an analytical method and a numerical tracing model is also of great interest. However there are two major concerns at this point: 1. In contrast to the repeatedly mentioned studies by Schimel and Bennett and


Interactive comment
Printer-friendly version Discussion paper sumption rates compared to the "traditional" analytical approach developed by Kirkham and Bartholomew (1954).However, the work in its present form is not convincing, as it has a fundamental experimental flaw and inconclusive reasoning.
My main concern relates to the high amount of 15N-labelled AA to both soils: approx.600% and 120% of the initial FAA pool was added as 15N-AA to the Podzol and Umbrisol, respectively (based on data given in Table 1).It is known that addition of a given substrate often stimulates its consumption, and thus it is common practice to add only a small fraction of the initial target pool when performing isotope pool dilution assays, in order to minimize such bias on gross rate estimates.Furthermore, the authors erroneously state "Following the recommendation by Wanek et al. (2010), the FAA label addition was 10 to 20 times larger than the initial FAA content in the original substrate (litter or soil)" (Pg 9 L25-26).However, Wanek et al. (2010) clearly state that a maximum of 25% of the initial FAA pool should be added in the form of 15N-tracer, based on a preliminary determination of the size of the FAA pool.The excessive addition of 15N-AA in the present study likely resulted in biased gross rate estimates, in particular CFAA, and thus also affected the calculation of nitrogen use efficiency (NUE).The stimulation of amino acid mineralization rates due to high 15N-AA label addition, are also likely to have resulted in an overestimation of MFAA rates with the numerical approach (and not only rates estimated by the analytical approach), which would explain the high MFAA rates estimated.In Figure 2d and 3d, it becomes also evident that the 15N-AA label was not homogenously distributed, otherwise there would have not been an increase in 15N/14N in the first time interval.Of course, for this time interval, it is not possible to calculate a rate using the analytical approach.The weak dilution of the 15N-AA label during the first time intervals was not due to low depolymerization rates, as suggest by the authors, but rather likely due to high 15N-labelled substrate addition, which resulted in an enrichment of the FAA pool of 60-70 at%.In line with the comment by Referee #1, I thus think that it is necessary to repeat the experiment with appropriate 15N-labelled substrate addition.

Discussion paper
This brings me to the question of how to evaluate which approach yields more realistic rate estimates, which certainly cannot be concluded by simply comparing rates estimated by the two different approaches.I believe that both analytical and numerical models have disadvantages and limitations.For instance, I agree with Referee #1 and #2 that the current numerical model is missing a microbial biomass pool with a different turnover time than the SON pool.The authors also claim that their numerical model overcomes the problem of high label addition by integrating gross rates over a longer time.As already pointed out by Referee #1, the enormous 15N-AA input could result in end-product inhibition of peptidases, leading to lower depolymerization rates during longer incubations.Here, I would like to add that longer soil incubations often result in ammonium accumulation over time, due to the absence of plant roots, which would constantly remove a part of the soil nutrients.For example, such an ammonium accumulation was observed for the Podzol, as the ammonium concentration increased 3-fold during the 240h incubation (Figure 2a).Therefore, I suggest that the authors need to provide some experimental evidence that the numerical model actually overcomes problems such as high label addition.Regarding the comparison between the two different approaches, statistical support should be provided for differences in rates estimated by the two models.
Furthermore, the authors claim that using gross rates computed by their numerical model (IFAA and MFAA) yields more accurate estimates of microbial nitrogen use efficiency compared to the model by Mooshammer et al. (2014).The authors raise two points: (1) the analytical approach yields less accurate rate estimates (but see comment above), and ( 2 compounds in soil and plant litter.Therefore, both NUEFAA and NUE models are conceptually justified and there is no evidence that using MFAA instead of M yields better estimates of microbial NUE.Indeed, estimates of microbial NUE could be improved by including also organic N compounds other than amino acids, which, however, remains a great analytical challenge. I also suggest that the authors should more carefully prepare the manuscript, as there are some inconsistencies and mistakes.Equations of the analytical approach (Eq. 1, 2 and 3) are partially wrong: equation 1 for production rate has a mistake in the numerator of the second term; equation 2 for the consumption rate is actually the production rate; and, equation 3, for the case when there is no change in concentration over time, also seems wrong, at least when compared to their former work (Andresen et al. 2015).Furthermore, the authors used different units: µg N/g, µg FAA/g, µmol N/g, nmol N/g.In some instances, it is not even stated whether it refers to N or FAA (Table 1, Table 2).I suggest using consistent units throughout the manuscript.As I understood it, the initial FAA concentrations presented in Table 1 should correspond to the sum of FAA presented in Figure 4.For Podzol, Table 1 shows 1.3 µg, whereas in Figure 4 the values roughly sum up to 3 µg.In turn, for Umbrisol, both Table 1 and Figure 4 seems consistent: 7.7 µg in Table 1, as well as roughly 7.7 in Figure 4.  Table 1: The initial soil ammonium concentrations of both soils should be stated.There is no need to say that the C:N ratio refers to dry soil.
) Mooshammer et al. (2014) use gross N mineralization (M) instead of amino acid mineralization (MFAA).However, the nitrogen use efficiency model proposed in the present study (NUEFAA) is conceptually different than that by Mooshammer et al. (2014) (NUE).The authors estimate here amino acid-N use efficiency (NUEFAA) based on amino acid immobilization and amino acid mineralization, whereas Mooshammer et al. based their model on gross amino acid consumption rate as proxy for microbial organic N uptake, since proteins are the main N-containing C3 comment Printer-friendly version Discussion paper : From a biological rather than mathematical perspective, I would say that low NUE leads to high N mineralization, and not that high N mineralization leads to low NUE.Pg 2 L23-24: The authors should state what the obvious limitations are.-2: For N mineralization, 13 min for equilibration between added and native ammonium is quite short.Pg 4 L3-4: Why were FAAs also extracted with 1 M KCL? Pg 4: The authors used the protocol developed by Wanek et al. (2010).However, the authors do not even once cite this work in Materials & Methods.Pg 4 L16: The reference is wrong.It should be Husek, 1991.Pg 4 L18-23: The authors should explain why some amino acids have the same m/z: Alanine and Glycine (m/z 116/117); Leucine, Serine, Isoleucine and Threonine (m/z 158/159); Proline and Aspartic acid (m/z 142/143); Pg 5 L5 and L15: The original references(Kirkham and Bartholomew, 1954;Watkins and Barraclough, 1996) are sufficient.Delete reference Andresen et al., 2015.Pg 5 L13: Specify that it is excess 15N abundance.Pg 6 L15: I do not always see a good fit of the model in Figure 2 and 3.For example in Figure 3b, the fit of the model for ammonium concentration seems not to fit the experimental data.Pg 8 L18-19: In Wanek et al. (2010) the samples were plant litter and not organic soil.

Figure 1 B
Figure 1 B: Equation is wrong.