This manuscript presents an exceptionally comprehensive soil geochemical dataset that addresses a critical gap in global biogeochemical databases by systematically characterizing 1,300+ samples across 30 mountain regions spanning five climatic zones in China. The authors’ methodological rigor is evident in their stratified sampling design across three pedogenic horizons (A, B, and C), standardized analytical protocols for 24 macro- and microelements, and integration with ancillary environmental variables including climatic indices, vegetation parameters, and human activity factor. The dataset’s particular strength lies in its unprecedented spatial coverage of montane ecosystems, combined with vertical resolution that captures pedogenic gradients crucial for understanding soil formation processes and biogeochemical cycling.

Overall, the authors’ efforts in assembling this high-resolution, multi-horizon, and climatically contextualized soil dataset are timely and scientifically significant for researchers in soil science, biogeochemistry, ecology, and Earth system modeling. Moreover, the manuscript is generally well organized, and it is suitable for publication in the journal after some minor revisions. Please find my comments below.

Specific comments:

I recommend the authors should stored the valuable data in the Zendo website.

Line 123: Replace “was” with “were”. Please check other grammar issues in the manuscript.

Line 132: Please specify the extraction method for pH measurement (e.g., water, KCl, or CaCl₂). This is essential for comparability with other pH datasets and can influence interpretation of cation exchange and element mobility.

Lines 154-158: The calculation of the Chemical Index of Alteration (CIA) should be more explicitly explained. Please clarify how CaO* was estimated, and whether the method has followed that of Nesbitt & Young (1982) directly or been corrected.

Lines 164-165: The strict coordination has been carried out, but it was not clearly defined. Does this refer to harmonization of sampling protocols across sites, or post-hoc statistical adjustments (e.g., normalization, transformation, unit standardization) to ensure cross-site comparability?

Line 103: The manuscript would benefit from a concise description of the statistical or visualization methods used to generate Figures 2-6. This addition will help readers better interpret the trends and distributions presented.

Line 260: The authors provided horizon-level sampling and vertical stratification but did not elucidate the implications for soil development modeling. Given the presence of C-horizon data and CIA indices, this dataset could serve as a valuable benchmark for soil formation modeling (e.g., using SoilGen or CLORPT frameworks). A short paragraph in Section 4 may highlight this point.

Line 316: Add a sentence summarizing the dataset structure (e.g., file formats, variable descriptions, metadata schema) to assist users in quickly understanding how to work with the data.

Line 249: The value “Fe (>200%)” as explanatory power in redundancy analysis seems inconsistent (R² cannot exceed 100%). Please double-check this statement or clarify if it refers to cumulative variance.

Tables 1 and 2: Several abbreviations used in these tables (e.g., MAT, MAP) are not defined within the table notes. As tables should be interpretable independently of the main text, please add a legend or footnotes explaining all abbreviations.

Figures 2 and 3: Both figures lack x-axis labels, which impairs interpretability. Ensure all figures include complete and clear axis annotations, including units.

Wu and coauthors present a robust and geographically extensive dataset of more than 20 soil elements derived from 1,314 samples across 30 mountain regions in China. This dataset covers diverse bioclimatic zones and three soil development horizons, which offers a valuable vertical and horizontal resolution for understanding the large-scale biogeochemical patterns. This comprehensive, spatially-explicit dataset from mountainous regions is timely and necessary considering the sampling difficulty and terrain complexity. The methodology alongside the rigorous quality control, open-access availability, and comprehensive metadata can significantly increase the reusability and scientific value of this dataset. In general, the manuscript is well-organized and written in fluent academic English, and scientifically rigorous. There are some concerns on the current manuscript that may help further clarify and improve the dataset’s accessibility and documentation. Please find my specific comments and suggestions below.

Please specify the sampling time in this work, which will help well use the dataset.

The sampling strategies need to be described more specific, considering such a large spatial scale and soil stratification. Were the samples composited from multiple subsamples or taken as single cores? How many replicates were collected per horizon at each site? Were replicate samples analyzed separately or composited before analysis? This information will help to assess the spatial resolution and statistical robustness of the dataset.

The authors have emphasized lithogenic and biogenic controls on soil elemental patterns in this study, and relevant lithology data have been used in their prior publications (e.g., Wu et al., 2025; Yang et al., 2022). However, such information is not included in the dataset. I strongly encourage the authors to incorporate this information as an additional column in the main dataset or in the supplementary materials. This will substantially enhance the dataset’s applicability in Earth system modeling.

In Figures 5 and 6, the abbreviation "AI" (aridity index) is not defined. Please ensure that all variables and indices (e.g., AI, CIA, NDVI) are spelled out at first mention, including in figure captions and abstract, to support clarity for multidisciplinary readers.

Carefully check all the figures to ensure that axis labels, units, and legends are present, standardized, and clearly legible. Some figures appear to lack axis units or use inconsistent font sizes. Improving figure formatting will significantly enhance the readability and usability of the manuscript.

Although the DOI is cited, the manuscript would benefit from explicitly stating the name of the data hosting platform (i.e., "National Tibetan Plateau Data Center") and providing a summary of available file formats (e.g., .CSV, .XLSX) and data structure.

In the dataset files, columns such as "Vegetation" and "Horizons" use abbreviated codes. Please ensure these codes are clearly documented in the metadata or in a separate codebook/readme file.

In addition to its clear value for biogeochemical modeling and soil quality assessment, the dataset also offers considerable potential for applications in soil development and weathering modeling. The inclusion of vertically stratified soil horizons, chemical weathering indices, and a range of environmental covariates, combined with the recommended addition of lithological data, provide a strong basis to simulate pedogenesis and mineral nutrient weathering and release across climate gradients. These potential applications are needed by highlighting them in the discussion to better reflect the broader relevance of the dataset.

This manuscript describes the multi-elemental composition of soil collected in locations in different mountain regions across China. It also describes the associated environmental parameters in these locations using remote sensing data. The authors then used these environmental parameters, along with geospatial and climatic data, to explain the variations in the soil multi-elemental composition.

 The database potentially contains important data which can be used for various applications as outlined by the authors in Sec. 4 of the manuscript. However, there are several important pieces of information that are missing in the manuscript, as well as serious concerns regarding data quality and completeness of the database, which I have described below.

Database

1. It is better to store the database in a more open-access platform like Zenodo that does not prompt users of data to log-in.
2. The database lacks metadata that explains what the columns mean. The units are also not given. The authors stated that a “Description of the dataset.docx” document accompanies the dataset, but it is not included when I downloaded the database several times. All these render the data in the database practically useless.
3. There are unrealistic values in the database. For example, the bulk densities are too low even in the surface mineral soil. I would not expect the values to be below 0.5 g/cm3, assuming this is the correct unit based on the data description paper. Another example is the moisture content which reaches up to over 200%. How were these values calculated? It would help if the authors explicitly state the measurement protocols and equations used.
4. What does depth mean in the database, and if I am right to assume that this represents the incremental depth, why are some organic layers lie below the surface mineral layer? Or does depth mean thickness here?
5. Why are some values missing for depth and bulk densities? This has to be explained by the authors.
6. It will also help if the authors include the soil type for each soil layer when available. This helps make sense of the elemental composition, CIA, and many other variables contained in the database.
7. The dataset does not contain the uncertainties of reported values.
8. An existing dataset of soil properties has recently been published in ESSD (https://doi.org/10.5194/essd-17-517-2025). How does your dataset compare to this?

Manuscript

L1 – A more appropriate title would be “Multi-Element dataset of soil profiles across (diverse) climatic zones in China’s mountains” since the focus is on the soil.

L17-19 – This has to be stated the other way around as this is a data description paper. Talk about how the dataset could contribute to the better understanding of China's mountain ecosystems. Also, include a statement about how this is related to soil.

L103 – The materials and methods lack the details needed especially in the analytical part. It also lacks a description of the statistical process that does not prepare readers about what to expect in the following parts of the paper.

L106 – This figure includes a disputed territory which is not relevant to the data because the authors did not sample in these areas. The authors are therefore advised to either remove the disputed territory from the map or include a statement regarding this in the figure caption. Make sure that it is aligned with ESSD's policy on neutrality regarding jurisdictional claims. I leave it to the journal editors to decide on this.

L113 – Provide the dates when the sampling was done.

L135 – How much soil was used in the analyses and how many replicates?

L156 – It is not clear how the oxide values were obtained from the elemental analysis. We're certain ratios used? If so, it has to be clearly stated in the manuscript.

L173-177 – This doesn't mean anything and is already given information for almost all soil types.

L246 – How can you have explanation by environmental factors exceeding 100%?