Review:

A Synthesis of Sphagnum Litterbag Experiments: Initial Leaching Losses Bias Decomposition Rate Estimates

The manuscript points out additional and valuable information about the influence of leaching losses on decomposition modeling, as they can lead to notable errors over longer periods of time. From my point of view, the results from the topic under investigation underline the importance of additional research in that field. Therefore, after carefully revising the manuscript and clearing some open questions (see further comments), I believe that the manuscript has potential to be published.

The abstract section gives some background information, but the problem statement is missing. The reason why the study has been conducted should be mentioned or further elaborated to underline the importance of the present study. More background information would be necessary to understand what initial leaching is. Generally, I am a fan of short abstracts, but I would suggest including more information in the Intro and Results part of the abstract. This would increase the readers interest to read further.

The authors have used enough related and up-to-date works. Together with more settled works, the authors give a nice synthesis of different litterbag studies and sum up some important obstacles. However, the paragraphs are sometimes disjointed, which can make reading difficult. Also, the wording needs to be revised in some cases to be more precise. In addition, typing errors should be corrected throughout the manuscript. The number of figures and tables is ok, the layout or graphic design could be improved.

From my point of view, the present manuscript does not fulfil all requirements for being published. A precise revision, especially of the abstract, the introduction section and the discussion is needed (see comments below). After addressing these points, the manuscript has potential to be published.

L5: In the first part of the sentence, you argue that you want to test if there is a bias in k0 due to l0, while in the second part of the sentence you already argue that they do so – as your aim is to quantify error estimates.

L20: the sentence is not clear to me

L27: the sentence is not clear to me

L36: “in reality…” delete

L 41:” and came to the conclusion” concluded that…

L65: leach

L65: it is unclear which plants will be investigated. Do you focus on Sphagnum species or include others? Especially as you mention other, also vascular plants and lichens afterwards (L 75ff)

L69: again, litterbag experiments in general or Sphagnum litter only?

L147: why only 15% not 18% as mentioned above?

L154f: check typos

Figure 4 (a): what does the grey horizontal line indicate?

L355: “for several hours”, what was the time span? Within 48h? or less?

L442ff: the aim is to improve litterbag experiments in the field, and you point out that the collection of litterbags shortly after the start of the experiment is necessary. Could you also give a suggestion which parameters should be measured to describe initial decomposition rates? Could that also help to make future studies more comparable?

Could you sum up the take home message? How can initial leaching effects be prevented in further studies?

General notes:

• Abbreviation C carbon
• Check spelling and typesetting
• Air-dried/ air dried, keep it uniform
• One-pool/one pool
• Check tables and figures
• Your discussion points out clearly, that many studies lack in information about the methodology, including sample preparation, corrections and calculations. This could also be included in the Conclusion.

Review for Teickner et al, Biogeosciences.

First, sincere apologies for the delays in this review. I started reviewing this paper multiple times, and tim got I got stuck while attempting to fully understand the large number of Bayesian models that the authors performed and the underlying assumptions in these models. To be honest, I’m still not sure if I understood all aspects of these models, the details of which are somewhat beyond my expertise.

The manuscript by Teickner and coauthors reports on a reanalysis of Sphagnum litterbag experiments conducted to estimate litter decomposition rates in peatlands. They posit that the decomposition rates inferred from such experiment are overestimated if initial leaching is not taken into account. They provide a detailed analysis of how different experimental procedures may have caused particularly high or low fractions of initial leaching, and provide guidance for future litter bag experiments.

This is a timely study of an important topic relevant to simulating carbon storage in peatlands. The study applies state of the art methods and the conclusions are well supported by the study results. The manuscript is clearly written and reads easily (well, with the exception of the underlying mathematics, but I guess that’s unavoidable).

I have some thoughts that could be incorporated into the manuscript, although I do not think that these are critical for the publication:

• The authors put emphasis on how litterbag experiments can be improved to more accurately calculate mass lot rates. Over the last decade, there has been some substantial criticism of the litterbag approach (in upland studies). Catrufo et al 2015 (Nature Geosciences), for example, used uniformly isotope-labelled plant litter to study the persistence and vertical translocation of litter derived C into soil. I think a publication that focuses on how litter bag experiments best done should address the question _if_ litterbag experiments are still the best approach to study decomposition rates (e.g., because isotope labelling of Sphagnum plants is not feasible).
• Regarding the bias of l0 on k0: While reading this study, I was wondering why these are inferred at the same time using a complex model. Would is not be easier to just exclude the initial mass from the dataset, and calculate mass loss rates based on the mass from litter retrieved at different time points?
• Finally, I have some doubt about extrapolating decomposition models fitted to data from <5 years decomposition to long time scale (up to 100 years)? Regardless if l0 is correctly quantified or not, I have doubts that decomposition over the next 95 years follows the same trends measured initially?