Determining Optimal Sampling Numbers to Investigate the Soil Organic Matter in a Typical County of the Yellow River Delta, China

Round 1

Reviewer 1 Report

 

The method of sampling and the representativeness of the analyzes carried out are the key issues in terms of the reliability of the conducted research. The presented paper is interesting but contains several points that need to be clarified or supplemented. In general, the publication omits the assessment of the impact of land use and sampling methods, which is a key element in determining the number of samples necessary for a correct assessment of SOM variability in the soils of a research area.

The abstract should be written in more detail because it is too general. The aim of the research and specific conclusions should be specified.

Keywords should not be in the title, so please rewrite them.

Please explain why the authors decided to take samples using the 's-shape' method (L. 103)

In my opinion, the better and more accurate method in SOM determinations is the C-analysis with the C-N analyzer. Were the authors able to perform this type of determination and estimate the differences? Performing the determinations with the use of the analyzer allows to reduce the error resulting from the analysis itself and the analyst's mistakes, which is extremely important with the subsequent use of the results for modeling. (L.109-112)

What were the soil usage of sampling area/sampling points? In my opinion this is a key factor which determines the variation in the SOM content in the samples.

If table 1 is presented in the form of descriptive statistics, then table 2 should be presented in the form of a bar graph with the normality curve

In the case of the accuracy of SOM assessment, the uncertainty of measurement including the sampling processes, preparation for analysis and determination methods should be performed. (L.207)

Author Response

Responses to the comments of Reviewer #1

Thank you very much for your valuable suggestions on our paper, and we have tried our best to revise our manuscript based on the comments. We hope the revised paper would satisfy you. The following is my reply point by point.

1. The method of sampling and the representativeness of the analyzes carried out are the key issues in terms of the reliability of the conducted research. The presented paper is interesting but contains several points that need to be clarified or supplemented. In general, the publication omits the assessment of the impact of land use and sampling methods, which is a key element in determining the number of samples necessary for a correct assessment of SOM variability in the soils of a research area.

Response: The consideration of land use factors in the sampling process has been described in the original article, but it may be not precise enough. So we adjusted the representation in the manuscript to “In the actual sampling process, the samples in each grid is determined according to the main crop types in the grid. When one crop type is in the grid, we sample at the center of the plot. When there are multiple crop types in the grid, we select the plot with the larger area for sampling. At the same time, some sample points were slightly adjusted by referencing the land use, road traffic situation.” in 112 line.

2. The abstract should be written in more detail because it is too general. The aim of the research and specific conclusions should be specified.

Response: Thanks for the reviewer's suggestion. We have revised the content of the abstract. We have added the specific process of soil sampling and the specific results of the experiment to the abstract.

3. Keywords should not be in the title, so please rewrite them.

Response: Thanks for your valuable suggestion. We have revised the keywords of the article, removed and corrected the keywords that are duplicated in the title.

4. Please explain why the authors decided to take samples using the 's-shape' method (L. 103)

Response: The S-shaped sampling method, also known as the serpentine sampling method, is suitable for plots with large sampling areas and irregular shapes. Soil sampling was carried out for the entire county in the study area, and the area of the sampling unit was about 150 hm², so we chose the S-shaped sampling method in the central plot of the survey.

5. In my opinion, the better and more accurate method in SOM determinations is the C-analysis with the C-N analyzer. Were the authors able to perform this type of determination and estimate the differences? Performing the determinations with the use of the analyzer allows to reduce the error resulting from the analysis itself and the analyst's mistakes, which is extremely important with the subsequent use of the results for modeling. (L.109-112)

Response: Thanks for your suggestion. As you said, soil organic matter determination with a C-N analyzer provides greater accuracy. While, our soil samples were collected earlier, and the experimental conditions such as C-N analyzer were not available at that time in our laboratory. According to the agricultural industry standard (NY/T 1121.6-2006) issued by the Ministry of Agriculture of the People's Republic of China, the content of soil organic matter can be obtained by K₂Cr₂O₇ oxidation-titration method(Officially recommended). So we chose this method. For ensuring the smallest error in the determination process, we set up three repetitions for each soil sample, and performed two blank experiments after every 90 soil samples during the determination of soil organic carbon. As the experimental conditions improve, we will adopt your opinion and use the C-N analyzer in the future experimental analysis to further improve the accuracy of the experiment.

6. What were the soil usage of sampling area/sampling points? In my opinion this is a key factor which determines the variation in the SOM content in the samples.

Response: Thanks for your suggestion. The main soil types in Kenli County are fluvo-aquic soil and coastal saline soil. We've added relevant soil types and main planted crop types to the “Study area”.

7. If table 1 is presented in the form of descriptive statistics, then table 2 should be presented in the form of a bar graph with the normality curve

Response: Thanks for your suggestion. We selected 900 sampling point data as a representative, and plotted the normal distribution histogram after logarithmic transformation (line 211).

8. In the case of the accuracy of SOM assessment, the uncertainty of measurement including the sampling processes, preparation for analysis and determination methods should be performed. (L.207)

Response: Thanks for your suggestion. There are many factors that affect the accuracy of organic matter content, while the main purpose of this paper is to explore the influence of the number of soil sampling points on the prediction accuracy of the overall organic matter in the region. Influences of other factors on the spatial variation of soil organic matter were added in the discussion part. And also, a new article is being written on the influencing factors of soil organic matter spatial variation, which involves the analysis of land use types, parent material, topography, meteorological climate and other factors.

Author Response File: Author Response.docx

Reviewer 2 Report

L17: “lower middle degree of variation” Please provides the reference for the classification of variation mentioned in the manuscript

L157: Which method was used for the determination of SOM content in soil?

SOM is a very sensitive to the time and space. The study area is dominated with cotton based cropping system and very less contribution by the other crop (L94-95). Does the minimum sampling number (150 no.) is sufficient enough for every land use situation specifically where diversity of cropping system is very high with almost equal frequency? Does grid based sampling method is not appropriate krigging or co-krigging method for development of thematic maps? Author have mentioned about the removal of outliers data (L113-115). I suspect there is a possibility that the rice ecosystem having vary high content of SOM. Removing them as outliers will be wrong interpretation. In view of the heterogeneity in soil, it will be very difficult to fix the minimum sampling size for SOM in a very diverse ecosystem. 

 

Author Response

Responses to the comments of Reviewer #2

Thank you very much for your valuable suggestions on our paper, and we have tried our best to revise our manuscript based on the comments. We hope the revised paper would satisfy you. The following is my reply point by point.

1. L17: “lower middle degree of variation” Please provides the reference for the classification of variation mentioned in the manuscript

Responses: According to the classification indicator of coefficient of variation (CV) proposed by Nielsen and Bouma (1985) [1], we have add the reference in line178 to support our view of “the SOM content in the study area had a lower middle degree of variation”. In their opinion, the coefficient of variation (CV) can be categorized into low(CV < 0.1), moderate (CV 0.1–1) and high (CV > 1)variable classes. So I think it is suitable that the soil organic matter data in the study area has a lower middle degree of variation.

[1] D. R. Nielsen and J. Bouma, “Soil spatial variability,” in Proceedings of a Workshop of the International Society of Soil Science and Soil Science Society of America, Las Vegas, USA (Wageningen, 1985), p. 243.

2. L157: Which method was used for the determination of SOM content in soil?

Responses: Line 125, the method of SOM determination were corrected as “The soil organic carbon (SOC) was determined by the K₂Cr₂O₇ oxidation-titration method. The content of SOM was obtained by multiplying the content of SOC by the Van Benmmelen coefficient (1.724)” to more accurately describe the determination method of SOM.

3. SOM is a very sensitive to the time and space. The study area is dominated with cotton based cropping system and very less contribution by the other crop (L94-95). Does the minimum sampling number (150 no.) is sufficient enough for every land use situation specifically where diversity of cropping system is very high with almost equal frequency?

Responses: We are sorry for the incorrect writing in terms of the main crops due to the older reference before. Due to the large changes in planting structure over the years in Kenli County, combined the field investigation and new references, we known that main crops are winter wheat and summer maize, along with rice and cotton. We have corrected the main crop types for line 106. In this paper, we have got the conclusion that collecting about 107 soil sample points per 1000km² can meet the demand for the expression of soil organic matter spatial variation in places with similar physical and geographical environment to Kenli County. In the discussion, we mentioned that 82 sampling points per 1000 km² were needed to meet the sampling requirements when Sun et al. conducted a study on the optimal sampling density of soil organic carbon in Yucheng County, which is located in the Huang-Huai-Hai Plain (line 344). They mentioned that "Yucheng is an important grain-production area of China, and crop rotation is predominantly wheat–maize." This is similar to Kenli County, where this article is located. The 150 sample points mentioned in our manuscript were synthesized on the results of the spatial distribution characteristics of soil organic matter and the accuracy of spatial prediction. And it is sufficient for the geographical environment of the Yellow River Delta region. As you said, SOM is a very sensitive to the time and space. Therefore, the 150 sampling points are suitable for Kenli County but may not be the optimal number for other geographical environment regions.

4. Does grid based sampling method is not appropriate krigging or co-krigging method for development of thematic maps?

Responses: Kriging method is to use known sample points to make unbiased optimal estimation of unsampled areas. In the process of studying the spatial distribution of soil organic matter in the region, soil sampling is an extremely important task. In the layout of sample points, we should pay attention to the coverage of the study area, as well as the objectivity, representativeness and feasibility of sampling. The grid distribution method can not only cover the research area more evenly, but also can adjust the number of sample points according to the land use inside the sampling unit. In the process of using the Kriging interpolation method to predict the soil property space, many scholars also use the grid-based sampling method. For example, the grid-based sampling method were applied in the spatial prediction of SOC [2-5] and SOM [6], which the kriging method were used to map the spatial distribution of soil properties.

In our research, we use grids to distribute points, but the sampling points may not meet the actual requirements (for example, some points are close to the road), and they are not typical for the soil in this area. Therefore, the collection of soil sample points has been adjusted according to the actual condition such as land use, crop and traffic condition. So the objective requirement of sampling has been fully considered.

 

[2] Zhang, Z.Q.; Yu, D.S.; Shi, X.Z.; Weindorf, David C.; Wang, X.X.; Tan, M.Z. Effect of sampling classification patterns on SOC variability in the red soil region, China. Soil Tillage Res. 2010, 110(1):2-7. https://10.1016/j.still.2010.05.007

[3] Yu, D.S.; Zhang, Z.Q.; Yang, H.; Shi,S.Z.; Tan, M.Z.; Sun, W.S.; Wang, H.J. Effect of Soil Sampling Density on Detected Spatial Variability of Soil Organic Carbon in a Red Soil Region of China. Pedosphere 2011, 21(2), 207-213. https://doi.org/10.1016/S1002-0160(11)60119-7

[4] Sun, W.X.; Zhao, Y.C.; Huang, B.; Shi, X.Z.; Darilek, J.L.; Yang, J.S.; Wang, Z.G.; Zhang, B.E. Effect of sampling density on regional soil organic carbon estimation for cultivated soils. J. Plant Nutr. Soil Sci. 2012, 175(5), 671–680. https://doi.org/10.1002/jpln.201100181

[5] Zhang, Z.M.; Zhou, Y.C.; Huang, X.F. Exploring the optimal sampling density to characterize spatial heterogeneity of soil carbon stocks in a Karst Region. Agron. J. 2020, 113(): 99– 110. https://doi.org/10.1002/agj2.20467

[6] Long, J.; Liu, Y.L.; Xing, S.H.; Qiu, L.X.; Huang, Q.; Zhou, B.Q.; Shen, J.Q.; Zhang, L.M. Effects of sampling density on in-terpolation accuracy for farmland soil organic matter concentration in a large region of complex topography. Ecol. In-dic. 2018, 93(), 562–571. https://doi.org/10.1016/j.ecolind.2018.05.044

 

5. Author have mentioned about the removal of outliers data (L113-115). I suspect there is a possibility that the rice ecosystem having vary high content of SOM. Removing them as outliers will be wrong interpretation. In view of the heterogeneity in soil, it will be very difficult to fix the minimum sampling size for SOM in a very diverse ecosystem.

Responses: Thanks a lot for your suggestions here. The exclusion of outliers is a must in all data analysis. The reviewers mentioned possible high values or high variability in the rice ecosystem. So we counted the SOM values of the six main vegetation types using the data collected in Kenli County without removing sampling sites. It can be seen from the figure that the difference between the SOM content under various vegetation coverage conditions is not very large. The vegetation type with the highest organic matter content is wheat.

 

According to the current research, complex terrain, diverse land use types and different crop growth conditions may be important factors affecting the spatial differentiation of SOM. However, in a certain area, the effect on the spatial differentiation of SOM is relatively small due to the large similarity of soil types and soil-forming parent materials. In our study, the location of the sample sites has taken into account the land use patterns, crop planting types and traffic conditions, etc. Therefore, from the overall consideration, the sampling number suitable for soil organic matter can basically be determined.

 

 

 

 

 

Author Response File: Author Response.docx

Reviewer 3 Report

The authors have performed a large amount of analytical and computational work. However, the work may contain a very significant methodological error. The authors nowhere give the names of the studied soil (or soils). There are also no data on the granulometric composition of soil samples. Indeed, if the soils were different and had different granulometric composition, then such results cannot be averaged. Therefore, the authors should be sure to give the name of the soil (or soils) according to the WRB and data on the granulometric composition. It is necessary to specify in more detail the methodology for determining soil carbon. Was carbon determined titrimetrically and photometrically? Also, in my opinion, in the title and in the text of the manuscript, use the terms "organic carbon", and not "organic matter". Since the authors determined precisely organic carbon.

Author Response

Responses to the comments of Reviewer #3

Thank you very much for your valuable suggestions on our paper, and we have tried our best to revise our manuscript based on the comments. We hope the revised paper would satisfy you. The following is my reply point by point.

1. The authors have performed a large amount of analytical and computational work. However, the work may contain a very significant methodological error. The authors nowhere give the names of the studied soil (or soils). There are also no data on the granulometric composition of soil samples. Indeed, if the soils were different and had different granulometric composition, then such results cannot be averaged. Therefore, the authors should be sure to give the name of the soil (or soils) according to the WRB and data on the granulometric composition.

Response: Thanks for your suggestion. We have added the main soil types in Kenli County to the "Study Area" in line 100. Fluvo-aquic soil and coastal saline soil were the major soil types in Kenli County, which can be compared with Cambisols and Solonchaks in WRB, respectively. During the experiment, only the particle size composition of some soil samples was measured. By referring to other studies, the relevant physical properties of the soil at the sampling point can be obtained (line 102).

2. It is necessary to specify in more detail the methodology for determining soil carbon. Was carbon determined titrimetrically and photometrically?

Response: Thank you for your comments on the SOM assay method. We have provided a more detailed description of the method for the determination of soil organic matter（in line125）. In this study, the K₂Cr₂O₇ oxidation-titration method were used to determine the SOC, and the content of SOM was obtained by multiplying the content of SOC by the Van Benmmelen coefficient (1.724).

3. Also, in my opinion, in the title and in the text of the manuscript, use the terms "organic carbon", and not "organic matter". Since the authors determined precisely organic carbon.

Response: Thanks for your suggestion. Soil organic carbon is one of the components of soil organic matter, which drives carbon (C) decomposition and nitrogen (N) mineralization in soil. We choose SOM as the research object instead of SOC, the purpose is to know the soil fertility status of the region as a whole. As well as, since the average carbon content in soil organic matter is 58%, the measured soil organic carbon is multiplied by 1.724 to obtain the content of soil organic matter. Following the advice of experts, I expressed it as SOC in another article on spatial prediction of soil organic carbon density.

Author Response File: Author Response.docx

Reviewer 4 Report

Small comments in the attached text of the article

Comments for author File: Comments.pdf

Author Response

Responses to the comments of Reviewer #4

Thank you very much for your valuable suggestions on our paper, and we have tried our best to revise our manuscript based on the comments. We hope the revised paper would satisfy you. The following is my reply point by point.

1.The line 94 need correction

Response: Thanks for your suggestion. Line106 the statements has been corrected as “The local crops mainly include winter wheat, corn, rice, cotton”. Three references with Kenli County as the study area have been cited here.

2. When there was a sampling? How long have a sampling (days, months, yeas)?

Response: Thanks for your suggestion. Line122 the statement has been corrected as “At last a total of 926 soil samples were collected in fall 2009 before and after crop harvest and the sampling area is approximately 1400km².” to show the time of soil sample collection.

3. In the legend would be better add "SOM content, g/kg" above brown pictograms

Response: Thanks for your suggestion. Based on the comments of the reviewers, we have revised the legend in Figure 4. The "SOM content, g/kg" has been adjusted above the brown pictograms and the positions of other legends have also been adjusted.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

I recommend the papaper for publication in Applied Sciences

Reviewer 3 Report

The authors corrected their manuscript according to the comments and the manuscript can be published.