Mapping Key Soil Properties of Cropland in a Mountainous Region of Southwestern China

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors apply well-tested methods for mapping key soil properties across Yunnan Province, China.  Overall, the article was grammatically well written with minor editorial changes required.  See line item suggestions below.

There is one major requested revision.  The authors do not include the importance of soil parent material (geology and surficial deposits - alluvium, eolian, etc) as a driver of soil physical and chemical characteristics.  Perhaps there are no geology maps available for this region?  If there are no maps available, I would suggest the authors review publications that discuss the role of geology (soil parent material) on soil characteristics and incorporate those concepts in your discussion as to the geospatial variability observed in your sampling. 

If soil parent material source can be associated with the 8K+ sample locations, this information needs to be considered in the RF model.  Frankly, 8K+ samples is impressive!!  My thoughts are that if you include soil parent material as a covariate in your model, you would see a significant improvement, given your sampling intensity.   This would allow you to not rely on surrogates like Elevation/Temp/Precip for pattern analysis.  While these climatic and topographic factors moderate weathering and mineralization, they are not responsible for the mineralogy of the soil parent material.  Only the mineralogy of a soil parent material is directly responsible for the particle sizes that it will weather to, and the relative nutrition found in the soil such as P and K (sans anthro influences).  Elevation/temp/moisture are important model variables as they do influence vegetation communities (species richness/diversity) and rate of organic matter decomposition, which in turn influences C and N found in the soil.  In other words, you need to account both for climate/topography, and soil parent material source.

If broad-scale geologic maps are unavailable for this region, the authors should have a good idea of the relative soil parent material types found throughout the region.  At least enough to speculate on how this information could impact models.

In short, if geology is available, a model re-run is necessary.  If not available, include in the discussion its importance and how future geologic mapping could help refine the current model.  The data collected is still excellent and can be further developed with future geologic mapping efforts.

Editorial suggestions:

L26: drop the "and" starting this line, and place it before C/N ratio

L46: The "solid" and objective assessment .... What do the authors mean by "solid"?  That is unclear use of the term.

L54: [9,10], supporting basic growth of crops and also affecting agricultural ....

L56: ... can guide a search for nutrient deficit areas and inform rational ....

L58: drop "the" at the start of this line.

L66: replace "it also plays" with "playing"

L68-69: drop "based on the consideration of soil nutrients"

L76: .... and thermal patterns, which affects ...

L84: drop "the" at the end of this line.

L86: One would argue that 1 km soil property mapping is not fine scale.  It all depends on use and context.  From a soil perspective, topographic features such as slope and aspect affect soil development at a much finer scale than 1 km.  So one must be careful when using the term fine scale.

L91-92: .... the importance of different variables to response variables.  This ending is very confusing.  Please provide clarity.

L98: Complex topography can certainly control weathering of soils, but soil parent material and its mineralogical composition is what determines particle sizes and relative nutrient availability.  Please see my long-winded paragraphs above.

L101: Replace "On the other hand" with "Alternatively"

L102: delete "in turn"

L104: replace "underline" with "address"

L107: place a period at the end of this line.

L108: capitalize "In".

L110: the trade-offs.  What do the authors mean here?  This sentence needs clarity.

L123: ... and with significant elevational gradients.

L147-149: This sentence is passive voice.  Take the last half after the comma following "environmental factors", and place it at the beginning of the sentence (and drop "have" from "we have").

L169: You were auguring the soil, not breaking through the soil.  Just say you used a stainless-steel augur to collect a bulk soil sample to a depth of 20 cm from the soil surface.

L169: It's unclear if the 8,571 samples were from 1,714 sampling locations, where you randomly collected 5 subsamples.  Or are the 8,571 samples, 8,571 unique locations where you collected 5 subsamples, which were then composited for a single location sample.  If the former sampling method was used, it could be argued that the five subsamples are pseudo-replication from a modeling perspective.  If you used the latter sampling approach, just clarify that they were composited.

L187-189: This sentence is passive voice.  An alternative could be:  "The significant advantage of kriging is unbiased predictions with minimal variance relative to the spatial correlation in weather station data used within this study".

L197:  .... and the importance of environmental variables?  Not sure what is meant by "their affecting variables".

L197-199: This sentence is quite unclear.  Are you using a model developed in a prior study?  Please clarify meaning.

L201-L202: "The RF modelling predicted ...." This sentence is out of place.  Move to after you discuss the second objective of the study.

L205: What significance level was used to evaluate correlated variables?

L264: Table 2 - Just wish to clarify that an r value of 0.06 was indeed significant at an p-value of 0.01?  That is a very weak r-value ......

L276-277: In general, higher average temperatures were observed at lower elevations.

L284-285: .... important environmental variables influencing soil SOC ....

L286: .... precipitation, sand and clay

L294: Figure 6. The predictive importance of environmental variables relative to soil organic carbon .....

L298: The measured soil C/N ratios ....

L319: ... distribution of precipitation and temperature patterns in an area, as well as the soil fertility, significantly impacting the distribution of local vegetation characteristics [49.50].  

L326: replace ventilation with aeration.

L329: replace good with important.

L338-339: Sandy soils generally have very poor nutrition, which is not what I think the authors are getting at with the comment "superior fertilizer supply".  I think you are saying that fertilizer is more readily available to plants as it is not adsorbed to clay/OM particles.  Just clarify this point.

L357: delete As

L360: Place a period after [66].

L382: Conclusions:  If no geological mapping has been completed for this region, make that a major recommendation of your study.  You cannot rely on topography and climate alone to map important soil variables.  The maps will never be useful at the application (field) level without this input.

Comments on the Quality of English Language

Overall, well done!  Just see my notes above.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The ms id agronomy-3023161 "Mapping key soil properties of cropland in a mountainous region of southwestern China" is very interesting and it has positive contributions in the field of soil science. Find below the specific comments regarding the ms:

1. What is the main question addressed by the research?

      The authors have analyzed the data of 8571 topsoil samples (a robust database). Additionally, they  carried out electronic mapping of the spatial distribution of soil texture (sand, clay, and silt content), SOC, TN, TP, TK content, and C/N ratio. It is very interesting since the authors have considered 1km resolution using the random forest model. 

2. What parts do you consider original or relevant for the field? What specific gap in the field does the paper address?

      The results, the maps (Fig. 2, 4, 5, and 8), and the discussion sections (subsections 4.1 and 4.3) are original and relevant to the field of soil science. Additionally, I recommend grouping Fig. 3, 5, and 7 into a single figure. In the discussion, subsection 4.2 provides a regional impact analysis. It would be better if the authors could use their regional results to predict similar issues in other ecoregions. They could use a PCA to show the main factors contributing with data variance in a single figure. Data from Fig. 3 and 7 could be transferred to a Table.

3. What does it add to the subject area compared with other published material?

      The results of this ms effectively enhance the understanding of mapping key soil properties and random forest model. It adds interesting maps about key soil properties of croplands by including data about C and plant nutrients, and these results are very interesting for soil scientists.

4. What specific improvements should the authors consider regarding the methodology? What further controls should be considered?

       Experimental design is robust and a PCA could improve their results and reduce the number of Figures. The authors have used a well-planned experimental design and they have adequately used the random forest model.

5. Please describe how the conclusions are or are not consistent with the evidence and arguments presented. Please also indicate if all main questions posed were addressed and by which specific experiments.

      The conclusions are consistent with the aims of the study by considering the the Random Forest  model and its prediction about soil nutrient distribution using data from 8571 topsoil samples. All questions were addressed in the results and disscusion sections accordingly.      

6. Are the references appropriate?

      The references are appropriate, as they were wisely utilized to support the results and main hypotheses regarding soil key properties.

7. Please include any additional comments on the tables and figures and quality of the data.

      Tables and figures follow authors' guidelines. However, they used to many figures. I would suggest to group Fig. 3, 5, and 7 into a single figure or/and transfer data from Fig. 3 and 7 to a Table.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

My suggestions regarding your article are given below.

Abstract Part:

1. The sentence "The results showed that the average content of each nutrient in the soils of Yunnan Province was SOC (18.78 ± 0.09 g/kg), TN (1.78 ± 0.01 g/kg), TP (0.98 ± 0.01 g/kg), and TK (13.89 ± 0.08 g/kg), C/N ratio (10.56 ± 0.02) respectively." should be changed as “The average contents of SOC, TN, TP, TK and C/N ratio in the soils of Yunnan Province were found as 18.78 ± 0.09 g/kg, 1.78 ± 0.01 g/kg, 0.98 ± 0.01 g/kg, 13.89 ± 0.08 g/kg, and C/N 10.56 ± 0.02 respectively."

2. Generally, the abstract section was written well. However, some sentences were not found appropriate for the abstract section. They should be deleted and added to needed sections. The first sentence is "The spatial distribution of soil fertility indicators and their influencing factors have not been adequately evaluated due to the lack of soil sampling in previous studies." It should be added at the end of the Introduction part. The second sentence is "It was also indicated that elevation, temperature, precipitation, clay content, sand content and silt content were the most important factors affecting SOC, TN, TP and TK content." should be given in Discussion part.

3. A sentence should be given in Abstract part about result of Pearson correlation coefficients

Introduction Part:

1. The Introduction part should start with the identification of the Yunnan Province.

2. The "RF" should be defined with its long name in introduction part.

Materials and Methods Part:

1- In such studies, soil sample collection times are very important. The authors stated that they took soil samples from 2011 to 2013. This information is not enough. Were 12 months of the year taken as a sample during these 2 years or which months were taken? In fact, it would be much better if seasonal changes were given separately in soil sample results.

2- The authors reported the ranged of the elevation of experimental area. However, the elevation of each area from which soil samples were taken was not specified.  The elevations from which soil samples are taken vary in terms of soil properties and are important. They should be given in the manuscript or as a supplementary table.

3. The sub title "2.1 Overview of the trial site" should be rewritten for clarity. It is not understandable and it is so complex.

4. Why did the sentence "The predominant cropping systems are multiple cropping with two crops in one year or three crops in two years." was given? Also, it should be rewritten for the clarity.

5. What method was used to take soil samples?

6. Some methods section of the manuscript was explained by giving some references. It is not enough for the manuscript. The authors should be rewritten methods section by giving details for the methods.

 Results Part:

1. The subtitles as "3.1 Predicted spatial distribution of soil sand, clay and silt content of Yunnan Province; 3.3 Changes in average annual precipitation and average annual temperature; 3.4 Variable Importance Based on Random Forest (RF) Model; 3.5 C/N distribution in soils of Yunnan Province" should be rewritten by giving some important results according to results of the study. They were written so short.

For example, if soil samples were taken from the same places in different years, similarities and differences in terms of features such as months, seasons, topographic structure, environmental factors should be written and compared. Additionally, information can be given about the vegetation of the place where soil samples were taken.

Discussion Part:

1- The discussion part of the manuscript was written so shortly. However, authors gave many results. The given results in Results section should be explained in the Discussion part.

2- The subtitle "4.3 Limitations and Implications" was not enough discussed . Also the given results in this subsection should be added to "Results part"

Conclusion Part:

1. Like the other parts, the conclusion is also written very briefly. Instead of general information, important results should be emphasized in line with the results obtained from the study.  Also, the suggestions for the researchers and contributions of the manuscript should be added to this section.

References Part:

1. There are many mistakes in the reference part. They should be rechecked. For example, “O’Rourke, Sharon.; Angers, Denis.; Holden, Nicholas.; McBratney, Alex. Soil Organic Carbon across Scales. Glob. Change Biol. 2015, 21, 3561–3574, doi:10.1111/gcb.12959.” reference was rechecked. And the reference “Alhassan, A.-R.M.; Ma, W.; Li, G.; Jiang, Z.; Wu, J.; Chen, G. Response of Soil Organic Carbon to Vegetation Degradation along a Moisture Gradient in a Wet Meadow on the Qinghai-Tibet Plateau. Ecol. Evol. 2018, 8, 11999–12010, doi:10.1002/ece3.4656.” should be rechecked.

 Best regards,

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

Dear Authors,

I think your article is much better with the revisions you made. 

Best regards,