Rotational Tillage Practices to Deal with Soil Compaction in Carbon Farming

Round 1

Reviewer 1 Report

This study investigated the effects of combinations of various tillage methods (conventional tillage, 3 reduced tillage and no tillage) on soil organic matter (SOM), soil penetration resistance (PR) and dry bulk density (DBD) using six year period of a long-term tillage experiment. After carefully reading,

(1)  The concept of conservation tillage technology should be re-checked. No-till seeding and subsoiling are two core elements of the conservation tillage technology (conservation agriculture).  The authors stated that “many farmers still approach no-tillage with skepticism and regard soil compaction as one of the primary concerns”, which is misleading.

(2)  The manuscript investigated the effects of diverse combinations of various soil tillage methods on soil properties. However, the tillage depths of various reduced tillage methods were smaller than 30 cm in the experiments; as a result, soil compaction could not be effectively removed using the proposed reduced tillage methods as per many previous studies in relation to subsoiling.  

(3)  The main geometrical and working parameters of various tillage equipment in both conventional and reduced tillage methods should be provided as they can significantly affect the subsequent tillage effects, please supplement them.

(4)  Figure 5: no standard deviation was found, please supplement them.

Minor editing of English language is required.

Author Response

Dear Reviewer,

Please find our corresponding responses in line with your comments:

• The concept of conservation tillage technology should be re-checked. No-till seeding and subsoiling are two core elements of the conservation tillage technology (conservation agriculture). The authors stated that “many farmers still approach no-tillage with skepticism and regard soil compaction as one of the primary concerns”, which is misleading.

Dear Reviewer

The concepts of conservation tillage and conservation agriculture are clearly defined by [7 – added as new citation] and [2,3] respectively. Ee redefined the concept of conservation tillage in the first paragraph of the introduction to clarify better our statement.

We believe that the statement “many farmers still approach no-tillage with skepticism and regard soil compaction as one of the primary concerns” is not misleading. This opinion is also supported by the authors in [45,46,58,59] and {7 - added as new citation}. No-tillage is a core element of conservation tillage and conservation agriculture but its adoption in Europe is very small [2,7], For Greece, no tillage may be considered completely unknown (communication with The Hellenic Association for the Promotion of Conservation Agriculture). From our own experience, soil compaction and weed control are the main concerns of the farmers. So this article examines some alternative paths for introducing and familiarizing farmers with the no-till practice.

• The manuscript investigated the effects of diverse combinations of various soil tillage methods on soil properties. However, the tillage depths of various reduced tillage methods were smaller than 30 cm in the experiments; as a result, soil compaction could not be effectively removed using the proposed reduced tillage methods as per many previous studies in relation to subsoiling.

Dear reviewer

As mentioned in our previous comment, the conservation tillage is a management approach that aims to minimize the frequency or intensity of tillage operations. All the alternative tillage methods we implemented comply with this concept, that’s why they are shallower or equal depth with ploughing. Of course, we expected that they would induce compaction issues, so we examined the tillage combinations to deal with it. In our study, we avoided to use deep tillage like subsoiling because it is very intensive and in terms of energy consumption, may be equal or even more demanding compared to the conventional ploughing. Nevertheless, we used a chisel plough which operates almost at the same depth with mouldboard plough but less intensive.

• The main geometrical and working parameters of various tillage equipment in both conventional and reduced tillage methods should be provided as they can significantly affect the subsequent tillage effects, please supplement them.

Dear reviewer. We added extra information about the used implements in the description of the tillage methods (lines 112-147). Thank you for your mention.

•  Figure 5: no standard deviation was found, please supplement them.

Error bars depicting confidence interval range were added. Thank you.

Reviewer 2 Report

1. The introduction does not provide the necessary overview of this study, nor its hypotheses. The authors should consider providing general information about the experiment and the tested hypotheses.

2. The experimental design of this study suffers from the lack of randomness  and replication for the 5th year tillage practice. The authors claim the design to be a split-plot but this reviewer doubts that it is the correct description of the design. Also, the authors claims that there are 25 treatments. In reality, there are two treatments each with 5 treatment levels, leading to a total of 25 combinations. The authors should revise this part of the materials and methods. 

3. The first alphabet of the name of a soil order should be capitalized (Vertisols)

4. How many soil samples for bulk density and SOM were collected from each experimental unit? Were they homogenized for SOM? 

5. What was the statistical model for ANOVA? 

6. Figure 4: line graph does not seem to be the appropriate format when measurements were taken at discrete depths, and not continuously. Also, what do the errorbars indicate? SEM? SD?

7. Figure 8. Why does this figure suddenly have depth on X axis unlike the previous figures?

8. This reviewer thinks that this manuscript will benefit if there were tables that summarizes the ANOVA results and the calculated means for measured variables. 

9. There is virtually no description of the results from statistical analysis except for in the very beginning about no difference between the two experimental fields. Therefore, all of the claims made in this manuscript is not supported statistically. 

This manuscript can benefit from in-depth revision on the English. There are odd use of the language throughout the manuscript that are not necessarily wrong grammatically but are difficult to read. 

Author Response

Dear Reviewer,

Please find our corresponding responses in line with your comments

1. The introduction does not provide the necessary overview of this study, nor its hypotheses. The authors should consider providing general information about the experiment and the tested hypotheses.

Dear reviewer,

We revised the last text on the introduction to make the overview and the hypothesis clearer.

Original text

“Nonetheless, hybrid systems that intermittently combine conservation tillage with conventional plough-based methods could serve as an alternative first step for the traditional farmers to become accustomed to the nuances of regenerative practices, easing their apprehensions and facilitating a smoother transition toward a sustainable, long-term system [43,44,56,57]. The present study capitalizes on the above perspective and explores data from the first period of a long-term tillage experiment established in Greece, during which, some changes were imposed to the experimental design leading to diverse combinations of conventional and conservation tillage practices. The study examines the impacts on soil compaction along with the changes in soil carbon.”

New text

“Nonetheless, hybrid systems that intermittently combine intensive and less intensive tillage practices could provide a practical solution for soil compaction easing the farmers apprehensions and facilitating a smoother transition toward regenerative soil management systems [45,46,58,59]. The present study capitalizes on the above hypothesis and explores data from the first period of a long-term tillage experiment established in Greece, during which, some changes were imposed to the experimental design leading to diverse combinations of conventional and conservation tillage practices. The study examines the impacts on soil compaction along with the changes in soil carbon and aims to identify the optimum tillage schemes that compromise benefits and drawbacks.”

We hope this new description defines better the main objectives of our work.

2. The experimental design of this study suffers from the lack of randomness  and replication for the 5th year tillage practice. The authors claim the design to be a split-plot but this reviewer doubts that it is the correct description of the design. Also, the authors claims that there are 25 treatments. In reality, there are two treatments each with 5 treatment levels, leading to a total of 25 combinations. The authors should revise this part of the materials and methods. 

Dear reviewer

After discussion with a statistician, we were advised that the experimental design was a split plot during the first four years but changed to a strip-plot after introducing a vertical factor B in year five. Strip plots are common in tillage experiments that require large space for machinery maneuvers. After that, we repeated the whole statistical analysis using the SPSSS suite (that has more capabilities compared to JASP). We used the common One-way Anova model for the first four years and (according to suggestion from the statistician) built a custom univariate linear model for the fifth and sixth year which had the following sources of variations:

• Location
• Replication
• Tillage for years 1-4 and 6 (Horizontal factor A)
• Error 1
• Tillage for year 5 (Vertical factor B)
• Error 2
• AxB
• Error 3
• Total

The new analysis didn’t alter the major findings of our work (there were some changes in the F statistic and the significance). The description in material and methods and all the graphs and tables were updated.

We also agree that the number of treatments are 2 with 5 levels each leading to 25 combinations. We corrected accordingly the description of the experiment.

All the changes are incorporated in the text. Thank you for the comments which we found very fruitful.

3. The first alphabet of the name of a soil order should be capitalized (Vertisols)

Corrected. Thank you.

4. How many soil samples for bulk density and SOM were collected from each experimental unit? Were they homogenized for SOM? 

In lines 249-50 it is referred that: “From each plot, three 0-0.30 m samples were taken randomly. From these, average dry bulk densities were estimated for the three depth intervals”

Also in lines 269-270 we refer that : “Three samples were taken from every plot. The same depths were thoroughly mixed to give one composite sample for each plot and depth”

I think they answer to your question, nevertheless we changed the phrase to “thoroughly mixed and homogenized” to become more clear.

5. What was the statistical model for ANOVA? 

Dear Reviewer. We answered this in comment 2 and adjusted appropriately the description in section 2.3.

6. Figure 4: line graph does not seem to be the appropriate format when measurements were taken at discrete depths, and not continuously. Also, what do the errorbars indicate? SEM? SD?

The lines in graphs 4, 6 and 7 were changed from b-spline to straight lines. The errorbars indicate 95% confidence intervals (mentioned in all the graph captions)

7. Figure 8. Why does this figure suddenly have depth on X axis unlike the previous figures?

Dear Reviewer. Figure 8 is similar to figures 9 and 10. They indicate dry bulk density for three different depths. To avoid confusion, we considered your suggestion and moved the depth labels in the top of Figure 8, as in Figures 9 and 10.

8. This reviewer thinks that this manuscript will benefit if there were tables that summarizes the ANOVA results and the calculated means for measured variables. 

Dear Reviewer. We agree that this information is important, but it would extremely increase the length of the manuscript. Therefore we added an extra “Supplementary material” document with extra tables containing numerical means and other statistic parameters.

9. There is virtually no description of the results from statistical analysis except for in the very beginning about no difference between the two experimental fields. Therefore, all of the claims made in this manuscript is not supported statistically. 

Dear reviewer. After providing the statistical results on the Supplementary material, we believe that our findings have a statistic support. We also tried to enrich our document with comments upon the statistic findings wherever was suitable (please check track changes).

Comments on the Quality of English Language

This manuscript can benefit from in-depth revision on the English. There are odd use of the language throughout the manuscript that are not necessarily wrong grammatically but are difficult to read. 

Dear reviewer. The document was proofread by a native English speaker.  Thank you for your prompt.

Reviewer 3 Report

The manuscript “Rotational tillage practices to deal with soil compaction in car-2 bon farming” deals the soil tillage effect on soil compaction and soil organic carbon. The data are generated from a long-term field experiment.

The finding results and the conclucsions are useful for the soil manangement.

 

Some comments to M&M (Section 2):

Did you collect the data form the Field 1 (silty-clayey vertisol) and Field 2 (clayey vertisol)?

If it is done, then the results for the two fields must be presented, because the soil texture is different.

The equation 1 for the normalisation of the of penetration resistance is used in the figure 5. That is okay.

The comparsion of the soil penetration resistance in the figure 4 between the years is not suitable, because the penetration resistance is also affected by the soil moisture content. You should present also the soil moisture content in the figures. Maybe from the determination of the soil dry bulk density.

Comments for Results (Section 3):

I recommend also a deeper statistical analysis. You mentioned in secion 2.3, that you used a post hoc test (Tuckey). I can not see the applied post hoc test statistics in the results (e.g. dry bulk density, soil penetration resistance, soil organic matter). The dry bulk density data and soil organic matter data in the figures would be better shown in tables with ANOVA-results and Post-hoc-test results.

 

Overall please check for the several typing errors.

Author Response

Dear Reviewer,

Please find our corresponding responses in line with your comments

Did you collect the data form the Field 1 (silty-clayey vertisol) and Field 2 (clayey vertisol)?

If it is done, then the results for the two fields must be presented, because the soil texture is different.

Dear Reviewer. Indeed, the results presented here are the means from two fields. In order to conserve space, and because the statistic analysis didn’t reveal any field x treatment interactions (despite the differences in texture), we avoided to showcase the results from the two fields separately (it would require the double number of figures and much more conversation). Nevertheless, after complying with the prompt of reviewer 2, we added detailed tables with numerical means and other statistical parameters in the Supplementary material.  Hope this is sufficient to explain the similar behaviour in the two sites.

The comparsion of the soil penetration resistance in the figure 4 between the years is not suitable, because the penetration resistance is also affected by the soil moisture content. You should present also the soil moisture content in the figures. Maybe from the determination of the soil dry bulk density.

Dear reviewer. Figure 4 doesn’t intend to make a comparison between the years, because as you correctly mentioned, the measurements are presumably made in different soil moisture contents.  This figure depicts “snapshots” of the differences among the five treatments for each year. The within years comparisons are made in Figure 5, which is based on normalized data to overcome the obstacle of soil moisture differences.

Unfortunately, we can’t utilize the dry bulk density data, as you suggested, because these measurements were made at different dates and, some years don’t coincide. Nevertheless, we have information about the average soil water content during the penetration resistance measurements from soil samples obtained in parallel, but these measurements were indicative, so they weren’t made on each plot, but we took one sample from each replication (at three depths, 0-0.15m, 0.15-0.30m and 0.30-0.45m). We avoided to add this information on our previous version to conserve space. Now that we include a Supplementary material document, we added a corresponding Table S5. Corresponding information is referred in lines 210-212 and comments included also in lines 311-313.

Comments for Results (Section 3):

I recommend also a deeper statistical analysis. You mentioned in secion 2.3, that you used a post hoc test (Tuckey). I can not see the applied post hoc test statistics in the results (e.g. dry bulk density, soil penetration resistance, soil organic matter). The dry bulk density data and soil organic matter data in the figures would be better shown in tables with ANOVA-results and Post-hoc-test results.

As mentioned above, this information was added in the supplementary material to avoid making the manuscript too long.

Overall please check for the several typing errors.

Checked. Thank you.

Round 2

Reviewer 1 Report

(1)  In the response, the authors stated that “In our study, we avoided to use deep tillage like subsoiling because it is very intensive and in terms of energy consumption, may be equal or even more demanding compared to the conventional ploughing.”. In fact, the subsoiling operations are applied only when the issue of soil compaction occurs. Removing soil compaction is quite necessary as both soil properties and crop growth conditions can be improved significantly although the draught force and energy are relatively high. Also, this practice is generally performed every 2-4 years. Therefore, subsoiling is a very necessary practice and not intensive. However, the tillage depths of various reduced tillage methods were smaller than 30 cm in the experiments; as a result, soil compaction could not be effectively removed using the proposed reduced tillage methods. 

Moderate editing of English language is required.

Author Response

Dear reviewer

We agree that subsoiling is an essential tillage practice for removing deep soil compaction or breaking a hardpan. In our case, we showed that a mouldboard plough or a chisel plough was also capable to alleviate soil compaction issues. For sure, it would have been interesting to have included a subsoiler on our comparisons, but currently, we have these findings which we consider important and deserving to be published. But, we keep in mind your prompt to test a subsoiler in the near future since our experiment is still running. Thank you for the advice.

We have improved our Introduction by adding extra references to the subsoiling practice [44],[45],[46] and also some extra references to rotational tillage [48],[49],[50]. We have also extended the discussion upon the importance of subsoiling in the “Discussion” section (3^d paragraph).

Reviewer 2 Report

Overall, this reviewer appreciates the appropriate changes made to the manuscript.

As of the current version:

1. Authors do not refer to the p-value when claiming that the difference was statistically significant (or not significant) (ex. line 416-417). Even if the table that summarized the results of statistical analysis is included, the authors should still report the p-values and refer to the table (Supplementary table X, for example) where this information can be found whenever such claim is made. 

2. For figures 5, 6, and 7, if the error bars indicate the 95% CI of the tillage rotation effect, I would expect each treatment to have its own error bar at each depth. But I only see one at each depth in the figures, which looks more like 95% CI of the depth main effect. Can the authors verify which 95% CI the error bars are representing? If they are indeed 95% CI of the tillage treatments, it would be most beneficial for each treatment line to have an error bar per depth so that the readers can easily compare among the treatments. 

3. For figures 10 and 11, the error bars are not that visible. The images should be refined and made more higher resolution

4. For the supplementary tables, does the "Sign." refer to the p-value? If so, please say p-value. If the interaction effect of the field and tillage was tested, why are their mean not shown? The means should have both the means of main effects and the means from interaction effects. Means that did not have significant statistical effects are conventionally not followed by a letter. For example, On Table S1, depth 0.40m, the means should not be followed by "a" as the p-value was not significant. There are also unnecessary and inconsistent number of spaces between the means and the letters. 

There are still some misspellings here and there. For example, figure 7 says "interupted" which should be "interrupted" 

Author Response

Dear reviewer,

Please check below our responses to your comments

1. We revised the results section and added references to the p-values and F-statistic from the corresponding supplementary tables.
2. The error bars are 95% CI for the tillage treatments. But if we add these bars on every treatment, it gets too hard to discern the differences (please check the attached “test 2001-02.jpg” as an example). Therefore, we find pore appropriate to keep one error bar per depth.
3. The images on the draft manuscript are jpegs, which is a compressed type. We used this type because when we pasted the original, high quality tiff formats on the manuscript, the document became greater than 2GB and wasn’t allowed to upload on the system. Nevertheless, we have provided the high-quality tiffs separately and we expect the editorial team to use these images during the preparation of the final article. On these tiffs, the error bars are visible.
4. The “Sign” in SPSS is actually the “p-value”. We changed the name in the supplementary tables.

Since the statistical analysis showed no-interactions between Fields and Tillage, we avoided presenting the interaction means to the supplementary tables to avoid a very large table size. But after your suggestion, we have updated the supplementary tables with the “Field” means and also, with the “Field” x “Current Tillage” and “Field” x “Previous Tillage” interaction effects, but we avoided to include also the interactions of “Field”x”Current tillage”x”Previous tillage since the tables would become extremely large (2x5x5 = 50 means for each depth). Nevertheless, we are open to provide these results as also all our primary data on demand.

We finally deleted the letters from the comparisons that did not have significant statistical effect and aligned the columns on the tables to keep consistency of the numbers and the letters.

We performed a final check for errors. A final spelling check anyway will be also performed by the MDPI editing team.

Author Response File: Author Response.pdf