This manuscript aims to investigate roles of abiotic stress and competitive interactions in shaping plant community assembly by using dynamic vegetation model with plant functional types. The approach seems interesting however, I found several methodological uncertainties, potential flaws, styles and errors throughout the manuscript that should be addressed to make it convincing.

Introduction: The foundation of the story is week for me. The knowledge gap or research questions and hypothesis are missing. Introduction is a bit general for me. Wise to make it more appealing by synthesizing key ecological aspects. Basically, it needs to present the significance of this study. For instance, what are the knowledge gaps based on previous studies in the Himalayan regions. How this study helps to advance our understanding on plant community assembly. There are several trait-based (morphological to elemental scale), field observation-based studies along the elevation gradients or focused on particular ecotones, which are overlooked in the manuscript. Those studies might be helpful to improve it.

The methodology needs several clarifications. There are several issues on data (parameters) used in the model simulation (see specific comments as well) which weakens the robustness of the model. For instance, Avolio et al. (2019) developed equation based on grassland plot data including species removal approach, it is not clear how authors link it with carbon masses. It is not clear, how this approach linked with observed traits data by Maharjan et al. 2023? Similarly, as Maharjan et al. (2021) collected the traits using common tree species, several deciduous species did not sample as they lost their leaves at the time of fieldwork (please see Maharjan et al. 2021). I am very curious, is these data really applicable to know the local species richness and evenness?

It was not mentioned if the occurrence pixels.  For example, sometimes the pixels be in the forest, doesn’t fall to the forest but an open area, which is clearly an erroneous or inaccurate observation. The data point checking and cleaning could be done to ensure their correctness

Result section is too comprehensive and very difficult to figure out the key findings. Better to present figures with key results in the main text and other can be transferred to the supporting information. I would suggest just to present the result. Don’t mix it with some explanations.

The overall discussion is not well written for me. Mostly, authors just present their results and compare them with other similar studies and fail to provide scientific evidence or possible ecological mechanisms to support their results. Thus, it needs to synthesize the results rather than presenting results directly. Also, key results should be highlighted and justified with scientific evidences. Deeper discussion needed including the mechanisms why and how you obtained such results? What are the implications of these results under climatic changes? How PFTs drive ecological niche formation, how traits explain it, what are the ecological mechanisms and what are their ecological implications. It warrants deeper discussion and wide literature review.

L130: please cross-check the sentence. Generally, precipitation peaks above 1500-2500 m and it decreases from around 3000 m.

L198: how do you define the wet days or dry days? What are the criteria?

L215-220: how do you define these functional groups? What are the criteria? It should be well described and methods should be reproducible.

Table 1: what are the sources of these data? source should be provided.

As I know authors used tree traits data ranging from about 100 m to 3800 m (see Figure 1 Maharjan et al. 2021), how model is simulated up to around 5800 m? Generally, trees are found up to around 4000 m, do this model is also applicable to simulate different ecosystems (such as alpine grasslands). This may create several issues about the robustness of the model. Since authors highlighted of the significance of study is comparing simulated and observed data (see L112-113), there is large gaps.

Figure 3: above-ground carbon mass or above-ground biomass? some places authors mentioned carbon mass, some places biomass, it makes lots of confusions. The terminologies should be consistent. As I see in Khanal and Boer (2023), it might be above-ground biomass. As Forest lines normally up to around 4000 m and Nepal’s forest inventory covers only forested areas, I am surprising to see the some observed data close to 6000 m. It indicates there are some errors using the plot-level data.

L335: How did you simulated LAI? it is not mentioned in the methodology section.

L342: How do you defined C4 and C3 grasses, as both types grasses found abundantly along the gradient.

Figure 4: For other parameters such as above-ground biomass and bole height, but just showing simulated LAI don’t provide any insights whether it works well or not.

Figure 5 caption: source of observed data should be acknowledged.

L451: Is there any pine species up to 4500 m? In my understanding, it is totally wrong.

L467: As authors performed compare the patterns separately, thus, in my understanding, not well evaluated the complex interactions. To evaluate the overall interactions, it is wise to quantify the relative importance of different biotic and abiotic variables on PFTs.

L521-522: Actually, high elevations, particularly above 3000 m deciduous Betula utilis (Himalayan birch) is one of the most dominant species. Thus, this statement could be problematic.

L633: Not well discussed whether simulated results synchronized with observed data or not and reasons behind these.

Overall, this manuscript makes a valuable contribution to understanding how competitive interactions shape plant community assembly across a large elevational gradient in the Central Himalayas, using a dynamic vegetation modelling framework that explicitly represents multiple plant functional types (PFTs). The integrative approach combining productivity, vegetation structure, and competition indices to examine vegetation–environment interactions is both timely and scientifically relevant. The study addresses an important gap in regional-scale ecosystem modelling for complex mountain systems. With some targeted revisions and clarifications, the manuscript could be further strengthened.

One point that may benefit from refinement concerns the statement that PFT functional diversity declines monotonically with elevation while compositional evenness increases (L38–39). This pattern may hold under a strictly linear interpretation of the elevational gradient; however, empirical evidence from the Himalayas often suggests a mid-elevation peak in species and functional diversity, with lower diversity in both lowland plains and high alpine zones. This well-documented hump-shaped relationship could be acknowledged and discussed, particularly to clarify how the model results align with, or diverge from, observed diversity patterns.

Regarding Figure 1, the source of the vegetation type data is unclear. The referenced web link does not appear to host the dataset used, and the map seems more consistent with vegetation layers produced under the Land Resource Mapping Project of the 1980s. Clarifying the exact data source and citation would improve transparency and reproducibility.

In Lines 201–212, the allometry dataset is referred to as “BADD,” whereas the correct name is the Biomass and Allometry Database (BAAD). This appears to be a minor typographical issue but should be corrected.

For Figure 2, I found the datasets and labels somewhat difficult to interpret. Lines 303–305 indicate that two estimates of GPP are being compared, but it is not entirely clear whether one of the point types represents GPP estimates from Bi and Zhou. If my reading of the figure is correct, the mean latitudinal trends appear similar at lower latitudes, with increasing divergence at higher latitudinal bands. It may be helpful to explicitly plot and compare the mean GPP trends for both datasets, or alternatively to clarify in the text whether the intention is simply to demonstrate broad similarity in overall trends.

In Lines 318–320, the manuscript attributes underestimated biomass regions primarily to dominance by the Rhododendron PFT. However, recent studies report that these regions are often dominated by Fir and Brown Oak in addition to Rhododendron. Including these species in the discussion would strengthen the interpretation, especially given that generalized PFT representations may not fully capture plot-level heterogeneity. These species achieve high biomass through contrasting strategies Brown Oak through high wood density, Fir through large stem volume despite lower wood density, and Rhododendron through high stem density which could help explain discrepancies between simulated and observed biomass.

For Figure 3 and the associated text, it is important to note that the observed carbon mass data represent forested areas only, whereas the model simulations cover the entire landscape, including non-forested vegetation. This mismatch may partly explain why the simulated carbon mass does not capture some extreme observed values. Several high-biomass observations at latitudes above ~28.5° occur at relatively low elevations (e.g., valley bottoms), where distinct topoclimatic conditions favor high biomass accumulation. Incorporating this perspective could help explain the observed relationship between biomass and simulated patch age (L324–327).

I also note that bole height evaluation relies on observations reported by Maharjan et al. (2021), but it is not entirely clear how these observations correspond to the simulated bole height derived from the “dynamic bole height scheme for each cohort” described in Lines 166–167 using the Tallo database. Clarifying this linkage would be helpful. Additionally, in Figure 5, simulated bole heights appear systematically lower than observed values for most PFTs. A brief discussion of potential causes such as parameterization choices, structural constraints in the model, or uncertainties in observational data would strengthen the evaluation.

For Figures 6, 9, and 10, the x-axis appears to represent elevation rather than latitude. If so, the axis labels should be corrected. In Figure 6, using contrasting line types and colors (rather than a color gradient alone) could improve readability and facilitate matching legend entries to plotted lines. In Figure 10, arranging panels in a single column with a shared x-axis may allow clearer comparison among classes.

In Lines 388–389, the text refers to latitude, while the figure shows elevation on the x-axis. Although latitude and elevation are correlated in this region, consistency between text and figures is important. Additionally, the color scheme used to distinguish PFTs makes it difficult to visually separate groups. Using similar hues for tropical and subtropical PFTs, or alternatively a more strongly contrasting color palette, may improve interpretability and alignment with the results narrative.

I also noticed some inconsistencies in terminology, such as the use of “above-ground carbon biomass,” “above-ground carbon mass,” and “above-ground biomass.” Standardizing terminology throughout the manuscript would improve clarity.

Finally, the discussion could be strengthened by incorporating additional relevant literature. For example, statements in Lines 501–504 could be supported by recent studies demonstrating the role of topoclimatic variability in shaping forest carbon stocks in the Central Himalayas, including studies that use the same or comparable plot-level datasets as those employed here.

Overall, this is a promising and well-conceived study. Addressing the points above would further enhance the clarity, robustness, and broader relevance of the manuscript.