This manuscript presents GrasslandTraitSim.jl, a novel process-based model designed to simulate the effects of land use and climate on plant functional composition and ecosystem services in grasslands. Building on existing methods, the model links plant morphological traits with demographic and physiological processes, providing a comprehensive tool for researchers and land managers. As noted ("This article has a strong focus on model description"), the paper provides a detailed description of GrasslandTraitSim.jl, including its technical specifications and features (see Section 3 on "Technical details of the GrasslandTraitSim.jl model"), and seeks to position it as a tool for advancing ecological research. However, the manuscript vacillates between a technical and a scientific focus, leading to some ambiguity regarding its overall contribution. While GrasslandTraitSim.jl shows promise for advancing ecological research, particularly in understanding grassland dynamics, certain aspects of the paper may require further consideration by the journal editor to determine its suitability for publication.

Firstly, although the model is described in great detail, the paper contains only a limited number of case studies. While these examples are useful, they do not sufficiently demonstrate the application of the model to real-world scenarios or its potential for practical decision-making in grassland management. The authors could strengthen the manuscript by providing more robust examples of how the model works in different environments or how it can inform specific management practices. These additions would be of particular benefit to readers looking for tangible applications.

Second, the paper presents several simplifications in the model that are not fully justified. Understanding the rationale behind these simplifications is crucial, as it will help readers to assess the reliability and limitations of the model. The authors should provide a clearer explanation of why these assumptions were made and how they might affect the results of the simulations. This would help to improve the overall credibility and robustness of the paper.

Furthermore, although the paper mentions that the model can be used for grassland management, it does not elaborate on how it could be integrated into real-life management strategies or policies. A more detailed discussion of the practical applications of the model - particularly in terms of its potential impact on conservation or land-use practices - would make the work more relevant to practitioners in the field. Including such practical considerations would help to bridge the gap between model development and its application in real-world scenarios.

Finally, the paper does not include a performance comparison with other similar models, nor does it assess the performance of the model in different settings. A comparative analysis with other existing models would provide useful context for readers and establish the position of GrasslandTraitSim.jl within the current body of ecological modelling tools. Such a comparison would also give readers a better understanding of the relative strengths and weaknesses of the model.

In light of these points, the journal editor may wish to assess the relevance of this paper to the journal, given its strong technical focus on model development and limited exploration of its scientific and practical implications. Addressing these concerns could make the paper a more complete and valuable contribution to the field. The authors may also wish to incorporate the many suggestions made by this reviewer directly into the manuscript, as this may serve to improve its scientific depth and practical value.

The manuscript presents a new model for grasslands, that allows simulations of the resposes of grasslands to land use and environmental drivers. Overall, I agree with the authors that grassland models need to be improved, particularly in the context of large-scale/global models that often focus on forests. I appreciate the trait-based approach that allows analyses of trait diversity and how it responds to grazing and other drivers. Thereby, a detailed respesentation of individual species in not required. Understanding grassland dynamics is important and timely due to increasing pressures by land use and climate change.

The model description is detailed and comprehensive and it is published together with examples and tutorials. This makes the model accessible to the modelling community, and results are reproducible. Nonetheless, some aspects of the model could be better motivated and justified. Thefore, I suggest being more specific about these points, particularly by adding supporting references where possible (see also minor comments below for examples).

I understand the argument that the manuscript is about presenting the model and that including full case studies would make the manuscript extremely long. A caveat of this approach is that it is more challenging to convince readers that the model is a step forward and to highlight the benefits compared to other models. In addition, a discussion of model uncertainties and deviations between the presented model result and data remains open. I recommend adding a short discussion on these points and therefore shorten elsewhere. I think there is potential to shorten by being more concise.

Minor comments:

l 27: "soils that are neiter too dry nor too wet" this is a bit vague. Another factor in this context is fire, even though it's not so relevant in the grasslands considered in the study.

l 49: "focussing on different parts of the model" I would rather argue that different model focus on different components and aspects of the ecosystem such as grazing, diversity, fire impacts, etc.

l 50, 51 remove "for example" in one of the sentences?

l 71: most DGVMs such as LPJmL include grass PFTs or herbaceous PFTs for C3 and C4 photosynthetic pathways, but their representation is typically simplified compared to trees.

l 118: "possible to run simulation with just one patch" I suggest rewording, it should also be possible do do this in a model with interaction, even though interactions don't play a role in this case.

l 124: 5000 kg/ha, is this a common biomass in the study system and a reference?

l 127: can negative values occur in the model? Does that indicate caveats in the model design?

l 129: based on the list of processes, is ther also plant mortality or is this implicitly considered in senescence?

Fig 1: This scheme is very helpful, but I was uncertain where to start and why the last step is at the top. So maybe add some letters or numbers to guide readers through the graph (and refere to the letters in the figure caption) or restructure that the first step is at the top. 

l 175: "easiear to think about growth reducer" this is vague and I recommend clarification.

l 183: in my experience, it's better to calculate total LAI by caculating the total leaf area in a patch divided by patch size, instead of summing LAI. But maybe this is the same in your model.

l 184: if abp is abovground per total biomas - shoudn't it be multiplied in the equation?

l 190: "we want to distribute..." I recommend rewording, you don't distribute the growth but this describes the rule how it's done in the model.

l 245: nutrient deficit instead of "too litle nutrients"

l 397: "distributes the grazed biomass" do you mean that the amount of biomass removal by grazers is species-specific?

l 413-414: why does fodder supply not fully compensate the requirements of animals? Is there are reference for this assumption?

l. 437: "die along the stem" is unclear to me, please clarify.

l. 475: use SI units instead of bar (ie Pa)

l. 500-507: you could add a table in the supplement to provide more details for the sites

l. 510: does this simplification influence the model results? Time between grazing periods allows vegetation to recover such that the grazing impacts may differ compared to a long grazing period.

l. 535: "more trait data" than? Not clear to me what more refers to.

l. 537: "callibration data from 2010 to 2021" why was the end of the entire period used for callibration? Often, the first period is used for callibration and the second for forecasting/model evaluation.

l. 538-539: are the sites in the north and the south similar enough such that a model callibrated sith northen sites can be applied to the southern sites? Why did you not use a randomized design for callibration and evaluation sites?

l. 551: replace "make sense" by "are plausible", can the ranges also be backed up with references?

l. 576: "we do not interpret the results" I suggest removing this statement or - even better - interpreting the resuls, even if this is the aim of a follow-up study. As it is now, I think this is unsatisfying for the readers to see model results but no interpretation.

l. 587: are consumed more instead of "fed more"?

l. 591: "we think ... model can be used..." This does not sound very convincing, I recommend a more convincing statement highlighting the novelty and potential of the model.

l. 595: climate change was not studied, is this an aim of a subsequent study?

l. 613: maybe provide examples what data sets you have in mind.